<%BANNER%>

Transcriptional profiling on trees affected by citrus blight and identification of an etiological contrast potentially a...

University of Florida Institutional Repository

PAGE 1

TRANSCRIPTIONAL PROFILING ON TREES AFFECTED BY CITRUS BLIGHT AND IDENTIFICATION OF AN ETIOL OGICAL CONTRAST POTENTIALLY ASSOCIATED WITH THE DISEASE By EDUARDO FERMINO CARLOS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Eduardo Fermino Carlos

PAGE 3

This work is dedicated to my wife Darlen e, to my daughters Lusa and Lgia, to my parents Norival and Maria, to my aunt Josefa, and to all my friends and family that shared their love and care. Along the way, many things were surpassed and our dear aunt Neula, uncles Otvio and Waldemar, and friend Beatri z Nielsen will always be in our hearts.

PAGE 4

ACKNOWLEDGMENTS I thank my family and friends for the unconditional support and encouragement during all the steps of the program. I thank CNPq, a Brazilian Federal Agency, for sponsoring most of my graduate studies abroad and for their overall attention with the fellows. I thank Dr. Gloria Moore, my major professor, for the outstanding support during the whole process. During my academic and professional life I have been fortunate to meet people that helped me not only as scientific mentors but also as examples of positive attitudes. I especially thank my friends Eliana Lemos, Luiz Carlos Donadio, Marcos Machado, Julia Beretta, Ken Derrick, Gene Albrigo, Steve Futch and Jos Eduardo Lima, among others. I also thank Dr. Chris Chase and Dr. Nancy Denslow for their intelligent insights and enthusiasm, which always brought positive perspectives about the ongoing work and essential source of stimulus. I thank Dr. Ken Cline for his suggestions and scientific argumentation. A regular student life normally requires patience, patience and dedication. Fifield Hall demanded the same efforts, and more. I thank all my friends that shared their care and friendships. I specially want to thank Gary Barthe, Mukaddes Kayim, Marty Dekkers, Abdul Al-Saadi, Basma El Yacoubi, Ufuk Koca, Karen Champ, Kim Niblett, Manjunath Keremane, Methap Sahin, Vicente Febres, Abeer Khalaf, Maureen Meyerson, Juliana and Gustavo Monge, Ricardo Harakava, Cristina Moreira, Dirceu Mattos, Maria Luisa Targon, Eduardo Stuchi, Otvio Sempionato, Jos Antonio da Silva, Aline and iv

PAGE 5

Marcelo Carvalho, Georgia and Jos Carlos Dubeux, Lvia and Steel Vasconcelos, Jos Carlos Rodrigues, Juliana and Flavio Silvestre, Camila and L. Augusto de Paula, Ana Carolina and Vitor Lira, Carolina and Antonio Martins, Eliezer Martins, Maria de Lurdes Mendes, Marisol dAvila, Tania Querido, Nancy Velasquez, David Moraga, Li Liu, Bill Farmerie, Jason Blum, Natalia, Duy Nguyen, Patrick Larkin, Jannet Kocerha, Vishal Patel, Mike McCaffery, Carole Dabney-Smith, Fabien Garard, Richard Berger, Denise Tombolato, Fabricio Rodrigues, Gisele Martins, Angela and Wayne M. Jurick II, Adrianna Castaeda, Ronald French and many other friends. There is no end to this list... Thanks to all my friends. I started in the Department of Plant Pathology and I thank Dr. Bill Zettler and Dr. George Agrios for the attention and dedication that I received during the application process and as a new graduate student. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iv LIST OF TABLES ...........................................................................................................viii LIST OF FIGURES ...........................................................................................................ix ABSTRACT .......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 Importance of Citriculture............................................................................................1 Citrus in So Paulo and Florida....................................................................................3 2 IMPORTANCE OF CITRUS BLIGHT AND REASONS FOR ITS CONTROL.......5 Citrus Blight Etiology...................................................................................................6 Citrus Blight is Affected by the Employed Rootstock...............................................10 Objectives...................................................................................................................12 3 BUILDING THE CITRUS BLIGHT SUBTRACTED LIBRARIES........................13 Root Samples and cDNA Synthesis...........................................................................13 Building the Suppressive Subtracted cDNA Libraries................................................14 Sequence Analysis and Selection of Clones...............................................................17 4 IDENTIFICATION OF DIFFERENTIALLY TRANSCRIBED GENES UNDER NON AND CITRUS BLIGHT CONDITIONS...........................................20 Experimental Design..................................................................................................21 Array Design and Spot Features .................................................................................21 Sample Preparation and Labeling...............................................................................22 Hybridizations and Parameters...................................................................................24 Measurements and Specifications...............................................................................24 Normalization and Controls........................................................................................26 The Visual Evaluation of the Membranes..................................................................27 vi

PAGE 7

The Cluster Analysis...................................................................................................29 The Contrast of Means Analysis.................................................................................31 Comparing the Evaluation Methods...........................................................................36 Combining the Results for the Selected Clones.........................................................36 5 THE RELATIVE TRANSCRIPTIONAL RESPONSES OF THE SELECTED GENES UNDER DIFFERENT INCIDENCES OF CITRUS BLIGHT....................43 Quantitative Real Time PCR was the Method of Choice...........................................43 The Selected Clones and the Characteristics of the Probes........................................45 Collecting and Preparing the Samples........................................................................49 Reverse Transcriptase (RT) and PCR Reactions........................................................50 Relative Quantification of the Transcriptional Levels of the Selected Genes............51 The Potential Biological Meaning of the Clone 109..................................................57 A Tentative Test to Verify the Effect of Cold and Drought Stresses.........................58 The Potential Effect of Redundancy and Gene Families............................................61 6 IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY ASSOCIATED WITH THE CITRUS BLIGHT DISEASE.......................................63 The First Screening of the Subtracted Libraries.........................................................64 Other CTV Genes were also Found in both Subtracted Libraries..............................68 Quantitative Evaluation of the P27 Candidate Gene in the Blighted Trees...............68 7 CONCLUSIONS........................................................................................................71 APPENDIX: B LAST ANALYSIS BASED ON NUCLEOTIDE SEQUENCES ...........................73 LIST OF REFERENCES...................................................................................................84 BIOGRAPHICAL SKETCH.............................................................................................91 vii

PAGE 8

LIST OF TABLES Table page 4-1. Transcriptional pattern of the selected clones. ..........................................................37 4-2. Blast search analysis for some clones. .......................................................................38 5-1. Characteristics of the chosen probes. ........................................................................47 6-1. The clone E8-13 had homolog sequences matching different isolates of the citrus tristeza virus (CTV)........................................................................................66 6-2. Other sequences with homology to CTV genes found in the blighted minus healthy (B-H) and healthy minus blighted (H-B) subtracted libraries.....................68 A-1. Blast analysis based on nucleotide sequences of each individual clone....................73 viii

PAGE 9

LIST OF FIGURES Figure page 1-1. Citrus around the world................................................................................................2 1-2. The citrus belts in Florida and So Paulo.....................................................................3 2-1. Characteristics and symptoms of citrus blight..............................................................7 2-2. Transmission of citrus blight........................................................................................9 3-1. cDNA synthesis enriching for mRNA........................................................................15 3-2. Making the subtracted healthy and blighted libraries..........................................16 3-3. Analyses performed for each sequenced clone...........................................................18 4-1. Preparing individual clones for the cDNA array. ......................................................22 4-2. Array design. .............................................................................................................23 4-3. Quality parameter for each membrane. .....................................................................26 4-4. Visual evaluation of the membrane pairs. .................................................................28 4-5. Transcriptional profiling of the candidate genes. ......................................................30 4-6. Contrast of means analysis of selected clones............................................................35 4-7. Blast-X of the clone 6 using the non-redundant protein database..............................37 4-8. Clone 38 has sequence homology to citrus chitinases. ..............................................39 4-9. Transcriptional profiling of the clone 38 (a chitinase homolog) and other clones.......................................................................................................................40 4-10. The clone 109 had sequence homology to citrus ESTs. ..........................................41 5-1. Designing a candidate sequence for the P5 gene. ......................................................46 5-2. Relative PCR efficiency plot of the clone 109 against the normalizer 18S...............48 ix

PAGE 10

5-3. Examples of real time PCR reactions. .......................................................................52 5-4. Relative transcriptional levels of the selected genes. ................................................54 5-5. Contrasts comparing the effect of cold and drought treatments on the relative transcriptional level of the clone 109 and other clones. ..........................................60 6-1. The virtual northern blot of the B H clones. ...........................................................65 6-2. The transcripts of the p27 candidate gene were abundant in the roots.......................69 x

PAGE 11

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TRANSCRIPTIONAL PROFILING ON TREES AFFECTED BY CITRUS BLIGHT AND IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY ASSOCIATED WITH THE DISEASE By Eduardo Fermino Carlos August 2004 Chair: Dr. Gloria A. Moore Cochair: Dr. Kenneth S. Derrick Major Department: Plant Molecular and Cellular Biology Citrus blight is an important citrus disease in Florida (USA) and So Paulo (Brazil), primarily affecting yield in adult plants and compromising maintenance of entire commercial blocks. Blight is associated with rootstock choice, but, at present, is of unknown etiology. Therefore, the objectives of this study were to identify differentially transcribed genes and uncover etiological contrasts under non and citrus blight conditions. Roots of healthy and blighted Rough lemon (Citrus jambhiri Lush) rootstock supporting Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) canopy were collected from a central Florida area. Total RNA was obtained and RT-PCR was performed enriching for messenger transcripts. Subtracted cDNA libraries were created and around 140 clones were arrayed onto nylon membranes. Independent RNA sources from feeder roots of ten healthy and ten blighted trees were labeled with P33, hybridized overnight with the membranes and analyzed under an imager system. Selected clones xi

PAGE 12

were validated by RT quantitative real time PCR. The results indicated that citrus blight was able to affect the transcriptional levels of certain genes in a similar pattern among different replicates. The level of response was dependent of the assessed group of plants. One of the clones had sequence similarities to a citrus EST and to a potential ubiquitin subunit. Another one had similarities to a citrus chitinase, helping to deduce a candidate sequence for the blight associated P5 gene. Three genes had higher transcriptional levels under blight condition and did not respond to cold and drought stresses. The blight associated P12 had higher levels in mildly than in fully blighted trees. Further characterization of these genes may contribute to the understanding and control of citrus blight. In another experiment, citrus tristeza virus (CTV) genes were observed in the libraries. The transcriptional level of the P27 gene (divergent coat protein gene) of CTV was far more abundant in roots of blighted than in healthy Carrizo citrange, which is considered to be resistant to variant forms of CTVs. It remains to be investigated if CTV causes or enhances blight, or only grows better in feeder roots of already affected trees. xii

PAGE 13

CHAPTER 1 INTRODUCTION The importance of one industry in the agriculture segment and the context where science exists to help society were considered in this study. The result was a true excitement not only because the focus of this research is (still) an unresolved real problem of citrus, but also because the search for information involved molecular biology and related fields. Importance of Citriculture In addition to tasting good, citrus is a well known source of vitamin C and antioxidants. It is also believed to have anti-cancer properties (Rafter, 2002), and processed peel and pulp are frequently used for animal feed (Fegeros et al., 1995). Because of its nutritional relevance, citrus is an important industry world-wide, raising economies at macro and local levels by supporting social development directly with jobs and with secondary industries and services. Citrus is commercially present in more than a hundred countries in all five continents, primarily within tropical and subtropical regions. The total production has consistently increased in the last 40 years, and, more recently, reached one hundred million tons yearly (Figure 1-1). The early citrus pioneers would probably be astonished by this level of production, but the modern industry has through the years achieved efficiency, flexibility and high quality standards. Citrus was first brought to the new world by the Portuguese and Spanish explorers at the beginning of the XVI century (Moreira, 1980; Allen, 2000). By the second half of the XIX century, the USA and Brazil had established fresh fruit companies, and the 1

PAGE 14

2 frozen concentrated technology, developed in the 1940s in Florida (Lewandowski, 2000), increased the demand for citrus juices. A) Total production of citrus from 1961 to 2000 020406080100120616570758085909500YearsMillions of tons Brazil USA China Spain Others B) 102 million tons of citrus were produced in 2003 18%13%12%6%6%4%4%3%3%2%2%2%2%2%1%1%1%1%14% Brazil United States of America China Mexico Spain India Iran, Islamic Rep of Nigeria Italy Egypt Argentina Turkey Pakistan South Africa Japan Greece Morocco Thailand Others Figure 1-1. Citrus around the world. A) The total production of citrus has increased almost 5 fold in the last 40 years. Brazil took the lead in production over the USA only after the Florida freezes during the 1980s. B) Citrus is commercially present in 138 countries according to FAO, with China, Mexico and Spain ranking respectively 3 rd 4 th and 5 th in total production in 2003. (Source: FAOSTAT data, 2004, http://apps.fao.org last accessed April 07, 2004). Many changes in production systems have been necessary to meet the ongoing needs of growing markets and the demands of new challenges, such as changes in organoleptical concerns, unexpected drought and cold stresses, outbreaks of pests and diseases and raises of political barriers, to name just a few. Consequently, production constraints have been overcome by the use of grafted plants to replace seedlings, changes in rootstocks, selection of new cultivars, relocation of production fields and charges in diplomatic battles (Fawcett and Lee, 1926; Moreira, 1980; and others). Literature is available for all transitions suggesting that the need for research in citriculture is not a

PAGE 15

3 recent event. In fact, despite frequent new challenges, in the last decades citrus has surpassed other important fruit crops such as bananas, apples and grapes in production according to FAO (Source: FAOSTAT data, 2004, http://apps.fao.org last accessed April 07, 2004). Citrus in So Paulo and Florida The two major citrus producing areas in Brazil and the USA are the states of So Paulo and Florida (Figure 1-2). A) Florida B) So Paulo N Figure 1-2. The citrus belts in Florida and So Paulo. A) The industry in Florida has moved to warmer southern areas after the freezes of the 1980s, and the major producers are nowadays Polk and Hendry counties, respectively, in the central and southwestern regions. B) In So Paulo, the northern region is drier and warmer than the southeastern areas and the major citrus companies have their headquarters around Araraquara city. Supported by good environmental conditions, efficient systems and high quality products, both Florida and So Paulo have industries primarily devoted to processed citrus juice, especially oranges, supplying most of the world demand for this commodity.

PAGE 16

4 While Florida has focused its efforts on the North America market, So Paulo dominates the European one. The social and economic consequences of abrupt changes in the citrus industry of both So Paulo and Florida would affect, at least to a certain extent, the around 400,000 jobs and 3 billion dollars in exports yearly in Brazil, and the 89,000 jobs and 4 billion dollars in sales in the United States. More details and other statistics can be seen at ABECITRUS (Source: http://www.abecitrus.com.br last accessed April 07, 2004), FUNDECITRUS (Source: http://www.fundecitrus.com.br last accessed April 07, 2004), CREC (Source: http://www.lal.ufl.edu last accessed April 07, 2004), and USDA (Source: http://www.nass.usda.gov/fl last accessed April 07, 2004). Both socially and economically, there are similarities and also differences in citrus industries worldwide. However, all contribute to supply different markets with distinguished quality fruits: citrus fruits. Therefore, citrus is an important segment of agriculture and demanding of research, which, ultimately, may help in the stabilization of society.

PAGE 17

CHAPTER 2 IMPORTANCE OF CITRUS BLIGHT AND REASONS FOR ITS CONTROL There are many constraints to citrus production around the world and citrus blight, in Portuguese declnio, is one of them. It has caused impressive annual losses of around 60 million dollars in So Paulo and 100 million dollars in Florida (Derrick and Timmer, 2000). Since its appearance in Florida during the second half of the 1800s, it has been a top concern for the Florida citrus industry, probably only approached by the damage caused by freezes in the past. Swingle and Webber mentioned that blight was first reported by Underwood in 1891, and also, that affected trees had been seen two decades earlier (Swingle and Webber, 1896). Symptomatic plants undergo a slow tree decline, rather than a sudden disruption in development, as suggested by the given name in English. Citrus blight has distinct characteristics not observed in declines caused by other diseases, such as Phytophthora sp., Tristeza, and others. Early symptoms, which occur in mature productive trees normally older than five to six years, include loss of the intense color of leaves and other green tissues; the expression of specific proteins (Bausher, 1990; Derrick et al., 1990; Taylor et al., 1996; Paiva et al., 1997); and complexation and translocation of zinc from leaves to trunk bark tissues (Albrigo and Young, 1980);(Taylor et al., 2002). Affected plants have their xylem vessels gradually blocked by amorphous and filamentous particles (Cohen et al., 1983), which probably leads to the visual symptoms of overall decline: smaller tree size, twig die back, smaller yield and fruit size and general mineral unbalance (with the clear pattern for zinc deficiency on leaves). 5

PAGE 18

6 Major physiological changes also occur (Albrigo et al., 1986), such as off season flowering, longer flowering periods and shooting of sprouts inside of the canopy, as if the affected plant was attempting to keep its vital functions (under normal levels). It is common to have less water uptake into the trunk of the affected plants due to xylem blockage (Cohen, 1974; Lee et al., 1984). Affected plants rarely die but are often pushed out and replaced by a young tree. Thus, not only direct losses because of tree failure and replacement, but other costs in grove management are always incurred. Reset groves pose serious difficulties in horticultural maintenance programs, since they now have trees of different ages, sizes and possibly on different rootstocks, all with different needs. They have also increased the risk of other diseases that can be spread on the new young trees. The Figure 2-1 displays the described progress of citrus blight often seen in Florida and in So Paulo groves. In addition to the noted annual losses, a simple and dramatic way to see the impact of citrus blight is by just driving through the citrus areas in Florida and So Paulo and paying attention to symptomatic trees and reset groves. Citrus blight also occurs in many other countries (Wutscher et al., 1980), except those with Mediterranean and desert climates where it has not yet been reported. Citrus Blight Etiology Citrus blight is an unresolved problem and its origin and etiology are still matters of debate. Several characteristics of a plant disease, meaning the outcome between a host and a pathogen under a proper environment, are present in affected trees, except the presence of a confirmed pathogen. However, it is worthy to note that (Agrios, 1997)

PAGE 19

7 Figure 2-1. Characteristics and symptoms of citrus blight. A) Zinc deficiency on leaves. B) Water uptake test into the trunk of the affected plant. C) Xylem blockage. D) Overall plant decline. E) Off season flowering. F) A reset grove has trees with different ages and needs.

PAGE 20

8 defines plant disease as: .whenever the ability of the cells of a plant or plant part to carry out one or more of these essential functions is interfered with by either a pathogenic microorganism or an adverse environmental factor, the activities of the cells are disrupted, altered, or inhibited, the cells malfunction or die, and the plant become diseased. Thus, this study considers blight as a plant disease of citrus. In addition, the process seems to be infectious in nature (Timmer et al., 1992), since transmission by root grafting was achieved in experiments done in the USA (Tucker et al., 1984), in South Africa (Marais and Lee, 1990) and in Brazil (Rossetti et al., 1991). However, the etiology remains unconfirmed and transmission has not occurred by either grafting canopy branches (Albrigo et al., 1993), by bud grafting, or by soil replacement (Timmer and Graham, 1992). Figure 2-2 summarizes the experimental transmission of citrus blight obtained by root but not by canopy grafting. Citrus blight is also present in mature seedling trees, eliminating the possibility of scion/rootstock incompatibility as a cause. The pattern of spreading seems to be towards adjacent trees in the same planted row, rather than across rows, as often noted by farmers and also by (Swingle and Webber, 1896), who first raised the hypothesis of infectious spreading pattern. Over time, the progression of the disease seems not epidemic, following a pattern of incidence that can be described by a linear model, which is closer to abiotic abnormalities (Berger, 1998). (Swingle and Webber, 1896) recommended eradication of affected trees, fearing further spread of the problem. That was not implemented. Since then, candidates for causal agent and other theories have been examined, such as a non-parasitic origin (Rhoads, 1936), the initial transmission trials

PAGE 21

9 Figure 2-2. Transmission of citrus blight. A) Layout of the work done by (Albrigo et al., 1993), where the receptor trees that were root grafted to blighted trees displayed symptoms 6 years after that, while no limb grafted plant displayed symptoms during the same period; B) Receptor trees used by (Rossetti et al., 1991) showing symptoms of blight after the root grafting work done in Brazil, confirming that blight is transmitted by roots as first demonstrated by (Tucker et al., 1984).

PAGE 22

10 (Cohen, 1968), soil born candidate pathogens (Nemec, 1994), the vascular limited bacterium Xylella fastidiosa (Hopkins, 1987), nutritional related factors (Wutscher and Hardesty, 1979), and others, but all remain unconfirmed (Derrick and Timmer, 2000). Citrus Blight is Affected by the Employed Rootstock Resistance to citrus blight is dependent upon rootstock choice, and even though rootstock replacement has been a necessity for decades, the overall solution for citrus blight has not yet been achieved. Resistance to citrus blight is apparently less common than initially thought (Young et al., 1982). Rough lemon (Citrus jambhiri Lush), having excellent yield and drought tolerance but susceptible to citrus blight, was once the major rootstock in central Florida. Extensive plantings on this rootstock in the 1940s, replacing sour orange (Citrus aurantium L.), was probably the reason blight has ranked as one of the most common citrus diseases in the state since then. Volkamer Lemon (Citrus volkameriana Ten. and Pasq.) was another rootstock option in the past for similar reasons. However, replacement of both with other rootstocks more resistant to citrus blight contributed to major changes in citriculture, since the replacements were normally less vigorous and sustained lower yields (Young et al., 1982). Among other consequences were increases in grove maintenance needs such as more attention to grove fertilization, to tree density per area, to other disease susceptibility, and to increases in irrigation, which ultimately contributed to the urbanist concerns in water usage in Florida (Callies, 2000). In Florida, replacement of Rough lemon (Citrus jambhiri Lush) by Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.) was not as effective as anticipated because Carrizo proved to be almost as susceptible to blight as Rough lemon (Young et al., 1982). Sour orange (Citrus aurantium L.) never matched Rough lemon yield in

PAGE 23

11 central Florida, and it also succumbs to citrus tristeza virus. Cleopatra mandarin (Citrus reticulata Blanco) yields fruit later than Rough lemon and when in full production the fruit is smaller and the yield lower, primarily when supporting sweet orange and grapefruit cultivars. Sweet orange (Citrus sinensis L. Osbeck) and Sunki tangerine (Citrus sunki L.) are considered more resistant to citrus blight, but are not widely used because of high susceptibility to gummosis of Phytophthora sp. and also to drought. Swingle citrumelo (Citrus paradisi Macf. x Poncirus trifoliata L. Raf.), largely used nowadays, will probably be of restricted use in the near future because some groves on this rootstock have become affected by citrus blight. In So Paulo, rootstock replacement is also an unresolved issue, with no obvious alternative to Rangpur lime (Citrus limonia Osbeck), another rootstock susceptible to citrus blight, and more recently to citrus sudden death disease. Able to grow on non-irrigated areas, on soils with high aluminum, and in low input production systems, Rangpur lime is still the rootstock of choice in more than 80% of groves in the So Paulo area, despite citrus blight losses (Pompeu JR., 2001). Thus, groves that normally would last naturally for 50 years or more are being replaced within 10 to 15 years, or less. The consequence is a fast and expensive turnover of trees for big companies and the end of business for medium and small farmers. Just the return on investment in citrus normally takes 8 to 9 years after planting. Therefore in all aspects examined, citrus blight is an important concern for citrus industries, wherever it happens. The complexity and importance of citrus blight demands broad inter-institutional scientific cooperation on efforts to examine its etiology and gain insights into what causes the plant to block its own xylem and decline, what is transmitted from plant to plant (and why only by roots), how the plant responds and what

PAGE 24

12 is changed in the metabolism of the affected plant, why young non-bearing plants are normally not affected, and more. But these are overwhelming questions. The present study, more conservatively, seeks to uncover molecular and etiological contrasts between healthy and blighted trees. Objectives The major objectives of this study were to identify differentially transcribed genes under non and citrus blight conditions and contribute to its understanding and control, which is the long term goal of the citrus blight research at this institution. The specific objectives were to build subtracted cDNA libraries from root samples of non and blighted trees. to identify differentially transcribed genes using cDNA arrays. to obtain quantitative information about the transcriptional level of the selected genes using reverse transcriptase real time PCR. to evaluate the presence of potential causal agents found in the subtracted libraries.

PAGE 25

CHAPTER 3 BUILDING THE CITRUS BLIGHT SUBTRACTED LIBRARIES For research on citrus blight, several promising technologies were considered. Genome wide approaches and microarrays have offered great perspectives in several biological systems, but sequence information is limited for citrus. Thus, cDNA subtractive hybridization (Diatchenko et al., 1996) was undertaken, with the objective to obtain enriched cDNA libraries for each considered condition: healthy and blighted plants. Subtracted libraries can uncover genes up or down regulated under a specific condition, thus being a useful source of candidate clones for further studies. Similar approaches taken with rice (Xiong et al., 2001) and soybean (Colebatch et al., 2002) encourage such efforts, because genes differentially regulated under disease pressure and protein synthesis responses were identified. Root Samples and cDNA Synthesis Since blight dissemination is associated with the root system (Timmer et al., 1992) and morphological changes are seen in the roots of affected plants (Lindbeck and Brlansky, 1998), molecular responses are also expected to happen; therefore, root tissues were chosen for subsequent procedures. Superficial roots of Rough lemon (Citrus jambhiri Lush) rootstock supporting Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) canopy were collected and total RNA was extracted the same day from feeder root tissues using Qiagen RNeasy protocol, with DNase digestion (Source: http://www1.qiagen.com last accessed July 15, 2001). It was done during the summer of 2001. Rough lemon was chosen because of its known susceptibility to citrus blight. 13

PAGE 26

14 Feeder roots, also called fibrous roots, of around 2 to 4 mm in diameter, were chosen because they display xylem plugging and internal anatomic differences in affected trees compared with healthy ones (Lindbeck and Brlansky, 2000), similar to what happens in canopy tissues, suggesting that molecular contrasts may also be present. Samples were taken from healthy and fully blighted trees previously diagnosed by typical visual symptoms, by the water uptake test (Lee et al., 1984), and by the P12 based immunoassays on leaf tissues (Derrick et al., 1990). The chosen trees were in a 10-year-old block in the Lake Alfred area, Florida (USA). Reverse transcriptase (RT) polymerase chain reaction (PCR) was performed and cDNA was synthesized using Clontech procedures, enriching for messenger RNA transcripts (Source: http://www.bdbiosciences.com/clontech/ last accessed September 15, 2001). An experiment was performed to determine the number of PCR cycles necessary to normalize the cDNA synthesis of both samples (Figure 3-1). Building the Suppressive Subtracted cDNA Libraries The healthy and blighted cDNA samples were digested with Rsa I restriction enzyme to remove the Clontech Smart oligos at the terminal ends of the amplified cDNAs, generating blunt ends. After purification, different Clontech adaptors (1 and 2R) were ligated to split sets of cut healthy and blighted cDNAs. Healthy cDNAs ligated to each adaptor were hybridized with an excess adaptor-free blighted cDNA and with each other. The same procedure was done with blighted cDNA samples and PCRs were performed in both cases, enriching for differentially transcribed genes under each condition, following the Clontech PCR-Select cDNA subtraction procedures (Source:

PAGE 27

15 A B Figure 3-1. cDNA synthesis enriching for mRNA. A) Scheme of cDNA synthesis following the Clontech Smart cDNA procedures (Source: http://www.bdbiosciences.com/clontech/ last accessed April, 15, 2004). The 3-PolyT-oligo-5 binds to the poly-A tail of the mRNA and the final extension of the 1st cDNA strand leaves an overhang of Cs, which is bound by the 5-oligo-GGG-3. After template switching both ends are filled by polymerization of the complementary strands, and PCR starts, using the primer extension sites present at both oligos. B) Different cycles of PCR were performed on each sample to achieve tentative uniform cDNA accumulation before reaching the plateau of the reaction, following the manufacturer recommendations. Healthy samples were considered optimized at the 17th cycle, while for blighted ones, 23 cycles were run. http://www.bdbiosciences.com/clontech/ last accessed September 15, 2001). After PCR, only molecules with different adaptors were exponentially amplified, in theory, while others could only have linear amplifications or impaired products due to hairpin structures. This technique was developed by Diatchenko et al. (1996). Thus, subtracted healthy minus blighted (H B) and blighted minus healthy (B H) libraries were generated, in contrast to unsubtracted ones ( H and B) and controls ( + and ), Figure 3-2. PCR Cycles 15 18 21 24 15 18 21 24 Healthy Blighted 17 cycles 23 cycles

PAGE 28

16 A B Figure 3-2. Making the subtracted healthy and blighted libraries. A) Scheme of cDNA subtractive hybridizations following Diatchenko et al. (1996). Two sets of the same cDNA (called tester) are ligated to different adaptors and hybridized with in excess adaptor free cDNA (called driver) and with each other. PCR was performed in both cases, enriching for differentially expressed genes in both directions (called forward and reverse libraries). B) The same procedure was done with blighted and healthy cDNA samples. Thus, subtracted Healthy minus Blighted (H B) and Blighted minus Healthy (B H) libraries were generated, in contrast to unsubtracted ones ( H and B) and controls ( + and ). H-B B-H + B H Health y Blighted= Healthy__Adaptor 1 Healthy__ Adaptor 2R Blighted PCR ( H B ) In theory, Healthy only BBlliigghhtteeddHHeeaalltthhy y == BBlliigghhtteedd____AAddaappttoorr 11 BBlliigghhtteedd____ AAddaappttoorr 22RR HHeeaalltthhyy PPCCRR (( BB HH )) IInn tthheeoorryy,, BBlliigghhtteedd oonnllyy

PAGE 29

17 The differentially enriched cDNAs, (H B) and (B H), were cloned following the Promega protocol for the pGMTeasy vector (Source: http://www.promega.com/vectors/ last accessed on September 15, 2001) and around 500 clones were randomly selected for sequencing from both libraries (400 from the B-H library and 100 from the H-B library). Sequence Analysis and Selection of Clones The obtained sequences were trimmed of vector and adaptor sequences used for cloning until no more matches were found using the VecScreen search software (Source: http://www.ncbi.nlm.nih.gov/VecScreen/ last accessed April 10, 2002). The sequences were run in the automated system of the University of Florida sequencing core (Source: http://www.biotech.ufl.edu/ last accessed April, 15, 2003) but the Clontech adaptors were not hidden in many cases, requiring manual trimming to avoid contamination of vector sequences. To reduce redundancy of clones that may represent the same gene, the sequences were analyzed using all available Genebank sequences (all-sequences and non-human non-mouse ESTs) and the Brazilian Citrus-EST project as reference sequences for Blast searches (Altschul et al., 1997). With all sequences individually grouped by similar blasted match outcomes, further clusterization and alignment assemblies were made using the Vector NTI software assisted by Microsoft Office applications and other web-based tools (Source: http://searchlauncher.bcm.tmc.edu/seq-util/seq-util.html last accessed April, 10, 2003). The Figure 3-3 displays examples of the four steps performed for trimming and clustering all the 500 individual sequences. After that, the longest clone within the homolog group was the one considered for further studies.

PAGE 30

18 A) >CitrusBlight-K5R2-E09.g folder=CitrusBlight length=672 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXGAATTCACTAGTGATTAGCGTGGTCGCGGCCGAGGTACAAGATTGTTCACTGTCCTTCCGTTTGCGAATCCTGTGTATCTTTATGCAATGATGTTGGGGTTTCTAACGATCATGCTCGACGTTTGGCTCTCACTAATGGGCGCGCGCTTGCTGTTGTTCTAGTCCCTGGTAACGAGAGATCAGCCTCATGTGCGTCTTGAATGCCCTTGTATCATTACTCTAAATAATGGATCGGGCTAGCTCCTTGACGCTTATAACATAAACAAAAATGTCATTGCTAGCTAGCCTGTGTACCTGCCCGGGCGGCCGCTCGAAATCGAATTCCCGCGGCCXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX B) Match to Vector: Strong Moderate Weak C) clone bp Database Highest match description e-values 152 250 All Genebank 1,367,736 sequences >gi|19747177|gb|AC113414.2| Homo sapiens chromosome 5 clone RP11-541P9, complete sequence 1.2 Brazilian citrus CCSM ESTs 13,610 clusterized sequences CXContig831 3e-92 ESTs Genebank non-human non-mouse 4,893,238 sequences 1 gb|BM376262.2|BM376262 EBma01_SQ002_N11_R maternal, 4 DPA, no tr... 80 1e-12 1e-12 D) LEA,Z46824,EST,FASTA,599: 1 632 k1-25,: 69 481 K5-F9,: 83 352 K5-G5,: 83 352 (complementary) K5-C10,: 83 352 k1-5,: 84 361 (complementary) 1 50 100 150 200 250 300 350 400 450 500 550 600 262 bp, 1e-60 270 bp, 1e-127 270 bp, 1e-129 270 bp, 1e-129 400 bp, 2e-90 Citrus EST Reference, 599 bp Figure 3-3. Analyses performed for each sequenced clone. A) Looking for vector sequences, such as the Clontech adaptors highlighted in blue and red letters, after the sequences had been masked by the sequencing core system as Xs. B) VecScreen analysis were run for all available clones, confirming in many cases the need of further trimming. C) Blast-N analysis in 3 databases. Many clones had only poor matches in the Genbank search, but had good sequence homology in the citrus and in the non-human-non-mouse EST databases. D) Alignment of the homolog clones against a reference sequence. The green plot, underneath the bars representing the clones, reflects the sequence variability per region.

PAGE 31

19 This step was an intensive computational effort aimed at reducing potential future nonspecific variation from different clones that, in reality, might be representing a transcriptional pattern of only one gene. The average length of the selected clones reached 240 base pairs. Then, it was possible to narrow down all sequences to around 140 tentative unique clones. The Table A-1 in the appendix contains information about these clones and displays the tentative functions and other characteristics based on blast analysis of homolog sequences. It is worthy to note that several clones had high matches with different genes of citrus tristeza virus.

PAGE 32

CHAPTER 4 IDENTIFICATION OF DIFFERENTIALLY TRANSCRIBED GENES UNDER NON AND CITRUS BLIGHT CONDITIONS In order to evaluate the transcriptional pattern of the genes represented by the clones selected from the subtracted libraries, a cDNA array experiment was assembled and run. The major objective was to compare those clones under non and citrus blight conditions. The DNA array technology has been used in plants with success. There are examples studying defense reactions (Scheideler et al., 2002), stress responses (Rizhsky et al., 2002), and others. Reviews can also be seen at Seki et al. (2003) and at Donson et al. (2002). However, in general, one arduous step in DNA array technologies is to compare information between different experiments (Stoeckert et al., 2002). With a similar concern, Brazma et al. (2001) proposed the MIAME protocol, which stands for Minimum Information About Microarray Experiment. The intrinsic principles of the MIAME apply primarily to define the information that should always be included in microarray repository databases, allowing other researchers to understand the experiment and the data. The ultimate goal of the MIAME protocol is to establish a standard for recording and reporting microarray data based on detailed annotation of six experimental sections: 1) experimental design; 2) array design and spot features; 3) sample preparation and labeling; 4) hybridizations and parameters; 5) measurements and specifications; and 6) normalization and controls. The MIAME protocol does not specify the format in which the information should be provided, but only its content. In the present cDNA array 20

PAGE 33

21 study, much more modest than the commercially available microarrays, those guidelines were observed attempting to induce uniform experimental conditions and effective ways to analyze the data. Annotations during all experimental steps were recorded and are summarized below. Experimental Design The tentative non-redundant clones from the subtracted libraries, (H B) and (B H), described at the appendix A, plus ribosomal clones and other tentative controls, were standardized at the same concentration (160ng/ul) after PCR and purification steps. That was possible because all the clones had the same flanking sequences at the Clontech 1 and 2R adaptor regions. Figure 4-1 displays the initial PCR efforts for some of the clones used in this study. Most of the PCR products were in the range from 200 to 400 bp, and some clones had to be amplified more than once to achieve the standards above. Array Design and Spot Features The clones were uniformly arrayed on nylon membranes, by printing duplicated sets of each using the robot printer of the ICBR-UF (Interdisciplinary Center for Biotechnology Research of the University of Florida). Each membrane contained 188 double spotted probes ready for further quantitative assays (Figure 4-2). The description of each clone is given in the appendix. For positional landmarking, ribosomal genes were used. They were clones with sequence homology to a citrus ribosomal gene, Accession X05910.1. Controls for quality evaluations included detection of water spots for background assessments, plasmid with GFP for nonspecific hybridizations and aerial plant stress related genes, such as the cold responsive gene CORc115 (panel G12, spot 180), for variation determinations. Membranes were UV-cross linked and stored at around 6 o C.

PAGE 34

22 12 3 456 7 8 LW 9 10 11 12 ABEF1 2 3 4 5 6 7 8 LW 9 10 11 1212 3 4 5 6 7 8 LW 9 10 11 12CDGH (-) (-) 1 2 3 4 56 78 910 11 12 1 2 34 5 67 8 9 10 11 12 12 345 6789 10 11 12 12 3 4 5 6 7 8 LW 9 10 11 12 1 234 5 678 LW 9 10 11 12LW (bp)----2000----1200----800----400----200----100 Figure 4-1. Preparing individual clones for the cDNA array. Selected clones were amplified by PCR from the subtractred libraries and prepared to be arrayed on nylon membranes (LW for low DNA mass ladder from Invitrogen; (-) for no-template control). Sample Preparation and Labeling The membranes were hybridized with independent healthy and blighted RNA sources. Total RNA samples were extracted under the same conditions from feeder roots from adult groves of Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) on Rough lemon (Citrus jambhiri Lush), Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.), and Swingle citrumelo (Citrus paradisi Macf. x Poncirus trifoliata L. Raf.) rootstocks from the central Florida region, using Qiagen RNeasy procedures with DNase treatment. Plants had been previously diagnosed using the visual

PAGE 35

23 188 187186185 LM H2O LM LM969594939291908988878685H 180 179 178 177 176 175 174 173 172 17117016984838281807978777675 LM H2OG168167166165164163162161160159158157 LM LM70696867666564636261F156155154153152151150149148147146145605958575655545352515049E144143142141140139138137136135134133484746454443424140393837D132131130 LM H2O LM LM125124123122121363534333231302928272625C120119118117116115114113112111110109242322212019181716151413B108107106105104103102101100999897 12 11 10 9 8 765 LM H2O LM LMA121110987654321Clones ABCDEFGH GFP GFP Figure 4-2. Array design. Colored spots indicate different types of clones or controls: pink (LM) for landmark positing; blue (H2O) for water controls; yellow (spots 7 to 12) for probes from citrus leaf tissue; orange (spots 171 to 180) for probes from stress related genes; and green (GFP) for hybridization control. Those controls accounted for thirty four spots. Other spots were printed with clones from the subtracted libraries of citrus root tissue. Seventeen clones were from the H B library (spots 5 and 6; from 53 to 63; and from 67 to 70). All the remaining one hundred and fifty four clones were from the B H library. identification of typical symptoms in the canopy and the water injection test in the trunk (Lee et al., 1984). All samples were uniformly handled, using tissues from pairs of healthy and fully blighted trees of the same genotype, grown under the same condition and from the same grove. Some membranes were used to determine the best amount of RNA, type of primer for RT (reverse transcriptase) reaction, and RNA sources, comparing total versus messenger RNAs. After optimizations, samples with 2ug of total RNA had the 1 st strand cDNA labeled with dATP33 using random primers following procedures similar to that of the Ambion Strip-EZ kit (Source: http://www.ambion.com/

PAGE 36

24 last accessed Octpber 10, 2003). The RT efficiency was evaluated by counting the scintillation of the labeled samples. Hybridizations and Parameters All membranes were exposed to the same amount of labeled probe, based on the scintillation reading of one million counts/minute/ml of buffer. The samples were hybridized overnight at 64 o C with individual membranes, washed, and analyzed under the Typhoon Phosphoscreen Imager System (Source: http://www.amershambiosciences.com last accessed April 15, 2004), always uniformly manipulated in pairs of healthy and blighted samples to allow further pairwise comparisons. The employed imager system evaluates the fluorescence of each spot represented on the screen. Changes of the oxi-reductive status of a screen component, made by crystals of barium (Ba), fluorine (F), bromine (Br) and europium (Eu), are induced by the radiation of the samples after overnight exposure. Then, a laser beam from the reader induces excitation and subsequent fluorescence of each spot represented on the screen, proportionally to the initial radiation emitted from the labeled sample. The image of each membrane was captured by the scanner and the quantitative information was obtained using the software ImageQuant (Source: http://www.amershambiosciences.com last accessed April 15, 2004). All membranes were treated as uniformly as possible to minimize other sources of variation in the subsequent analyses. Measurements and Specifications The membranes were framed and had the image contrast optimized for better visualization using the ImageQuant, Microsoft PhotoEditor and PowerPoint softwares. The quantitative information of each clone was then evaluated. The parameter for quality

PAGE 37

25 of each membrane was verified looking at the ratio (Ratio) between the averaged absolute values, intensity of the fluorescence per pixel in the screen, of all the genes (AveGene) represented by the clones, over the average of all the water spots (AveH2O), used as negative controls for the hybridizations, and therefore considered as background or noise, as Ratio = AveGene / AveH2O Seventy-one membranes were printed and run. The expected true ratio was calculated to be within the confidence interval of 2.33 and 3.27, for p < 0.01. Therefore, individual membranes with a ratio smaller than the low limit of 2.33 were considered inappropriate for analysis and were discarded. One membrane that had nonspecific hybridization, measured by the spots with GFP genes, was also discarded. Forty unbiased membranes passed those criteria and were saved. Among those, twenty individual membranes represented ten pairs of uniformly manipulated healthy and blighted sampled trees. Each pair of samples was collected at the same day, from the same grove, and had their RNA extracted, labeled and hybridized under the same conditions, as uniformly as possible. Thus, membranes of any considered pair of healthy and blighted samples displayed an averaged absolute value (intensity/pixel) for all clones similar to the other membrane, in most of the cases, and both were respectively higher than their background and nonspecific hybridization controls by at least 2.33 fold. Root samples from trees on Carrizo were used on the pairs made by the membranes 1x2, 5x6, 67x54 and 65x66.

PAGE 38

26 From Lemon, on the pairs 9x10 and 61x58. From Swingle, on the pairs 29x30, 31x34, 51x52 and 69x70 (Figure 4-3). 024612566754656691061582930313451526970Cutoff AveGene/AveH2O 2.33 Ratiomembranes Carrizo Lemon Swingle Figure 4-3. Quality parameter for each membrane. The unitless ratio represents the proportion between the averaged absolute value for all clones (AveGene) in the membrane over the averaged absolute reading for the background level (AveH2O) of the same membrane. Membranes with a ratio smaller than 2.33 were discarded. Normalization and Controls The readings of the duplicated spots for each clone were averaged. All water spots (AveH2O), previously calculated for each membrane, were used to subtract the background from each clone (C). Negative values implied that the reading of the clone was smaller than the background value. Those values were discarded. Then, each clone was normalized (n) against the new mean of the clones (MeClone) in the membrane, rather than against one or two genes, such as a ribosomal or other ones, as follows: n = [ ( C AveH2O ) / MeClone ] 100

PAGE 39

27 The normalization method should reflect similar results comparing to the visual evaluation of the studied membranes. The mean of the clones was used in this study because it met this criteria better than other measures of central tendency, such as the median of the numbers. Normalization against a measure of variance, using the standard deviation, was also tried but no improvements were observed either. Besides the visual evaluation of the membranes, two other types of analyses were performed. To study the relative levels of transcription of the same clone comparing healthy and blighted samples from uniform conditions, cluster analysis was done. To study the variance of each clone, contrast of means analysis was done. The Visual Evaluation of the Membranes The tentative visual differences for each clone were highlighted in each framed membrane. This procedure helped to identify clones not previously seen under the initial visual evaluations. The three evaluated genotypes (a lemon, a citrange and a citrumelo) displayed differences across the species. But looking at one rootstock per time, some visual differences between healthy and blighted samples were also seen (Figure 4-4). Under the citrus blight condition, Carrizo citrange displayed stronger signal for the genes represented by the clones on the panels B1 (clone 109), C1 (clone 25), D4 (clone 136), F1 (clone 157), G7 (clone 79), and maybe others. Lemon had differences on the panels C1 (clone 25), C10 (clone 34), G7 (clone 79) and G8 (clone 80). Swingle had differences mainly on the panel C1 (clone 25). Very few clones displayed more signal on the healthy than on blighted samples: the panel A6 (clone 6) shows a tentative candidate of that on Carrizo and the panels E8 (clone 152) and E9 (clone 153) on Lemon.

PAGE 40

28 A) Carrizo roots healthy tree 1-15 blighted tree 1-11 B) Lemon roots healthy tree 3-11 blighted tree 6-9C) Swingle roots healthy tree H8R blighted tree B7R Figure 4-4. Visual evaluation of the membrane pairs. Membranes were hybridized with samples labeled with P33, using feeder roots of Valencia tree on different rootstocks. A) on Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.). B) on Rough lemon (Citrus jambhiri Lush). C) on Swingle citrumelo (Citrus paradisi Macf. x Poncirus trifoliata L. Raf.). Each panel has two clones horizontally replicated. Red arrows represent potential more signal on blighted samples. Blue arrows, on healthy samples. Green arrows, near identical. Blue circles represent spots filled only with water, for background control. Green circles, with GFP, for non-specific hybridization control. Black circles, with CORc115, for observations of a strees related gene.

PAGE 41

29 Nearly unchanged genes were seen in all samples and the panels D9 (clone 141) and E5 (clone 149) represent examples of those, in all three rootstocks. The Figure 4-4 shows one pair of membranes for each rootsctock. The Cluster Analysis The transcriptional levels of the normalized Healthy (nH) and Blighted (nB) samples were compared following the Log2 function, to allow equidistant visualization of the fold (F) induction or repression for each gene, as: F = Log2 [ (nB+0.01) / (nH+0.01) ] Gaasterland and Bekiranov (2000) describe two major types of analysis, the supervised and the unsupervised approach. In the supervised analysis, a particular context of each measurement is known and the end result is a list of individual genes behaving differently in each of the different contexts. In this study, the supervised model was accepted and the null hypothesis considered no differences in the transcriptional levels of the candidate gene under healthy and blighted conditions. Then, the cluster analysis and the graphic visualization for each clone was done employing respectively the Cluster and the TreeView softwares (Eisen et al., 1998). The Cluster software works on a random assignment of vectors for similarities found within the numbers. Subsequent new layers of vectors allow further clusterization of all samples. To test the significance of the differences, t-Tests were employed. Induced and repressed candidate clones were nominated, after analyzing the outcome of the ten independent pairs of healthy and blighted trees (Figure 4-5).

PAGE 42

30 GroupsIVIVIII II< -1 -0.75-0.5-0.25000.250.50.75> 1A) Cluster analysisB) Selected clones spring 03summer 02winter 03spring 03Clone t-Test p<0.307p<0.014p<0.156p<0.089p<0.368p<0.005p<0.832p<0.617p<0.269p<0.153p<0.450p<0.580p<0.007 Membrane pairs Figure 4-5. Transcriptional profiling of the candidate genes. A) Outcome of the ten pairs of trees under healthy and blighted conditions, being four pairs on Carrizo citrange, two on Rough Lemon and four on Swingle citrumelo rootstock. Each spot represented the ratio of the normalized blighted (nB) over the healthy (nH) sample given by Log2[(nB+0.01)/(nH+0.01). Green indicates negative ratio (more healthy than blighted) and red positive ratio (more blighted than healthy). B) Visualization of the individual clones and their p values for the t-Test, with the significant ones (p<0.05) being highlighted with a nearby star.

PAGE 43

31 After the cluster analysis, all membranes were visually re-evaluated. Clones that had p<0.05 from the group I, mostly upregulated candidates under the citrus blight condition, were located again. This process helped to spot differences not previously seen in the membranes. Similar approach was taken for the clones of the groups II, III and IV, mostly downregulated candidates under the citrus blight condition. However, it is possible that, for these clones, a more sensitive detection method would be more effective, because few visual differences were re-confirmed. The nearly unchanged genes of the group V, displayed the expected clones 141 and 149. Albeit small, the clone 141 did have a significant variation, with p<0.007, leaving the clone 149 alone as a candidate for an unchanged gene under citrus blight condition. These ten paired samples were collected and run from July of 2002 to May of 2003. Closer patterns can be observed comparing the Carrizo and Swingle samples from the summer of 2002 and the spring of 2003. However, more replicates would be needed to better study the real contribution of the season of the year on the studied model. The Contrast of Means Analysis Although the previous analysis is powerful for initial screening, it brings a natural chance of error by selecting potential false positives, since a clone with initial reading value (C) that approaches the background level (W) will generate a normalized number (nH or nB) that approaches zero, complicating any further conclusions for the pairwise or cluster analyses. On the other hand, simply discarding those clones may penalize the final outcome by leaving behind clones that may be indeed positives. Thus, besides exercising caution on apparently extremely induced or repressed clones, those were the reasons to perform the contrast of means analysis. For that, each clone was assessed individually

PAGE 44

32 within each uniform pair of plants (healthy and blighted), from the same rootstock type and from the same block of trees, as previously described. The assumed model for variance in this case implies that the variance of a clone (x) is a function of its averaged value (^x), plus the contribution of the membrane (M), plus the contribution of the uncontrolled residual factors (e), as x = ^x + M + e Other assumptions for the model above were also considered, as having additive factors, normal distribution of the data, independence between treatments and homocedasticity-also called uniformity of variance (Banzatto and Kronka, 1992). Additive contributions of each factor can be explained by the assumed model itself, since the outcome of each clone was the sum of those factors. Normality was only partially observed because the data aggregated around the mean, for the most part, but not perfectly. Independence of treatments in this biological system could not be fully accepted because that implies that the genes assessed by each clone operate independently from each other. Since metabolic pathways are complexes and regulated by counterpart genes, this is probably not true in this system either. Homocedasticity was not verified and the data was checked for transformation options, such as, RootSquare, Log and other functions. The smallest discrepancy in variances was obtained using the Log transformation of the previous normalized data (n), giving to each clone its final value (V), as

PAGE 45

33 V = Log2 (n + 1.5) The normalization for the contrast of means was similar to the initial normalization for the cluster analysis (nH and nB) previously described, except for considering individual values of all spots and adding the value 1.5 to avoid zeros in the computation. Once the transformation was observed, t-Test was employed on the established contrast of means, where the estimate of the healthy mean (Hm) and the estimate of the blighted mean (Bm) generated the estimate of the contrast (^Y) for each clone, as ^Y = Hm Bm Then, the relation between the absolute value for the estimate of the contrast (^Y) and its standard error, which is a function of the number of replicates (rH or rB) and the standard deviations (sH or sB) for healthy and blighted samples, gives the t-Test value (t) for each clone (Banzatto and Kronka, 1992). Subsequent comparisons against standard t-values determined at which level of probability (p) the null hypothesis, claiming no difference in the transcriptional level of the studied clone, can be accepted or rejected, in favor of the alternative hypothesis, claiming significant differences between healthy and blighted transcriptional levels for each clone. Some of the clones were variable but not at a level considered significant (t-Test) for this cDNA array system (p<0.05). However, other clones displayed significant

PAGE 46

34 response to citrus blight, suggesting that citrus blight is apparently able to affect the transcriptional level of certain genes in affected plants (Figure 4-6). In plants with citrus blight, the clone 6 was significantly downregulated (p<0.05) in three out of the ten pairs of evaluated trees. However, it was also significantly upregulated twice, implying in a great non-specific variation. The clone 25 was significantly upregulated (p<0.05) only in two studied pairs, but displayed higher averaged values on blighted samples in nine pairs of plants, and a strong significant difference on the pair 51x52. The clone 38 was significantly upregulated in two pairs of plants. It also had a higher averaged transcriptional value in three other pairs of plants. The clone 109 was significantly upregulated (p<0.05) in three out of the ten pairs, and had other three higher averaged values on blighted than on healthy samples. The clone 149 was nearly unchanged, however it had one significant difference (p<0.05) comparing the affected and the healthy plants on the pair 9x10. The clone 153 was significantly downregulated (p<0.05) in only two paired samples, but had higher averaged values on healthy than on blighted samples in other five pairs. The clone 157 had a higher averaged transcriptional level on affected than on healthy plants, but a large variance too, being significantly downregulated (p<0.05) in one pair of plants.

PAGE 47

35 02468Cz1x2Cz5x6Cz65x66Cz67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveClone healthy blightedClone 6 02468Cz1x2Cz5x6Cz65x66Cz67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveClone 02468Cz1x2Cz5x6Cz65x66Cz67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveCloneClone 109 02468Cz1x2Cz5x6Cz65x66Cz67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveCloneClone 149 02468Cz1x2Cz5x6CZ65x66CZ67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveCloneClone 153 02468Cz1x2Cz5x6Cz65x66Cz67x54Le61x58Le9x10Sw29x30Sw31x34Sw51x52Sw69x70AveH2OAveCloneClone 157Clone 6Clone 25 Carrizo Lemon Swingle p<0.013p<0.015p<0.004 p<0.010 p<0.019 p<0.010 p<0.000p<0.057 p<0.124p<0.095p<0.009p<0.068p<0.071p<0.001p<0.038p<0.009p<0.083p<0.000p<0.006p<0.129p<0.035p<0.063p<0.001 02468 Clone 38p<0.021p<0.000 Log2(n+1.5) Figure 4-6. Contrast of means analysis of selected clones. Pairs of healthy (blue columns) and blighted (purple columns) trees on Carrizo (Cz), Lemon (Le) and Swingle (Sw). The p value represents the significance of the t-Test, helping in the decision of rejecting or accepting the null hypothesis. The blue stars indicate significant differences when p<0.05. The bars are the standard deviation of each studied condition: healthy or blighted samples.

PAGE 48

36 Comparing the Evaluation Methods All three methods used to evaluate the transcriptional pattern of the genes represented by the clones displayed positive and negative aspects. The visual evaluation of the membranes required neither normalization nor computation and it was good for the major differences, besides confirming, or not, the results of the quantitative analysis. However it did not allow an easy identification of all the candidate clones in the first attempts. The cluster analysis was good for combining all clones from the available samples under the pairwise based approach. Inferences about the averaged fold induction, or repression, and about the aggregation group of each gene could be estimated. But not all quantitative information matched the visual evaluation of the membranes. The contrast of means was positive to reveal the significance of the differences within each pair of trees. It represented an unfolded view of the cluster analysis replicates. However, since the data was transformed, further inferences on the relative transcriptional levels within each pair of plants were no longer precise. Combining the Results for the Selected Clones Several clones displayed either visual or quantitative differences between the evaluated samples. The Table 4-1 displays the summarized results for those selected clones. Among the clones that displayed some differences in the cDNA array experiment, the clone 6 is from the healthy minus blighted (H-B) enriched library and displayed mostly lower transcriptional levels under the citrus blight condition. It is 325 bases long and has sequence homology (Table 4-2) with an Arabidopsis thaliana EST (e-value of

PAGE 49

37 Table 4-1. Transcriptional pattern of the selected clones. Observations from the three different types of analysis, comparing healthy (H) and blighted (B) samples. Clone (library) Visual differences Cluster analysis (averaged Log2*; p-value) Contrast of means analysis (number of significant differences, for p<0.05) 6 (H-B) H > B -0.46; p<0.153 3 (H > B) and 2 (H < B) 25 (B-H) H < B +1.82; p<0.156 2 (H < B) 38 (B-H) rare +2.44; p<0.089 2 (H < B) 109 (B-H) H < B + 3.12; p<0.014 3 (H < B) 149 (B-H) H ~ B +0.05; p<0.580 1 (H > B) 153 (B-H) H > B -3.13; p<0.005 2 (H > B) 157 (B-H) H < B +1.13; p<0.307 1 (H > B) and 1 (H < B) Log2 stands for Log2[(nB+0.01)/(nH+0.01)] 1e-180) in the Genbank and with a citrus EST (7e-11) in the Brazilian databank. It has an unknown function based on the translated query (Figure 4-7) to the protein databank (4e-30) and a high nucleotide homology to a mitochondrial sequence (1e-180), raising the question whether the clone 6 represents an open reading frame (ORF) of a gene or not. Figure 4-7. Blast-X of the clone 6 using the non-redundant protein database. Sequence of an environmental protein Accession EAI39113.1 (subject), with unknown function, had similarities (4e-30) to the translated version of the clone 6 (query). But because of its variance (Figure 4-6), the clone 6 is considered to represent a false positive outcome, not representing a true downregulated gene under the citrus blight condition.

PAGE 50

38 The Table 4-2 displays the accession numbers of the first match on the blast search analysis, using different databases, for the selected clones. Table 4-2. Blast search analysis for some clones. The search for homolog sequences was done using different databases. Clones e-values Highest match on sequence search analysis 6 1e -180 7e -11 4e -30 1e -180 Arabidopsis thaliana EST, Accession CF653082 Citrus EST, CCSM, Brazil, Contig 204 unknown protein, environmental sequence, Accession EAI39113.1 Arabidopsis thaliana mitochondrial genome, part A, Accession Y08501.1 25 1e -91 Citrus EST, CCSM, Brazil, Contig 416 38 8e -88 1e -88 3e -95 Citrus chitinase class II, Accession Z70032.1 Citrus chitinase class I, Accession AB081944.1 Citrus EST, CCSM, Brazil, Contig 1434 109 9e -56 9e -44 (2e -65 )* Citrus EST, Accession CK935651 Citrus EST, Accession CB293790 Arabidopsis thaliana, ubiquitin ligase SCF complex subunit, Accession NP568603.1 149 2e -54 Citrus EST, CCSM, Brazil, Contig CXJE02 153 Poor sequence information 157 Poor matches in the searched databases e-value of the citrus EST Accession CK935651 translated query on the protein databank The clone 25 is from the blighted minus healthy (B-H) enriched library and displayed higher transcriptional levels under the citrus blight condition in some of the tested pairs of plants. It is 300 bases long and has sequence homology with a citrus EST (1e-91) in the Brazilian databank, but only poor matches in Genbank. It also had a strong visual difference on the membranes of Swingle (panel C1, figure 4-4C). That difference was significant (p<0.05) in one out of the four tested pairs of plants on Swingle (Figure 4-6). However, considering all rootstocks, the overall transcriptional fold induction under the blight condition was around three and a half times, Log2[(nB+0.01)/(nH+0.01)=1.82, but with a poor p value of only 0.156 (Table 4-1 and Figure 4-5B). It is possible that the

PAGE 51

39 clone 25 represents a true upregulated gene under the citrus blight condition, but with higher transcriptional responses in Swingle citrumelo. The clone 38 is from the B-H library and is 235 base pairs long. It represents a citrus chitinase gene (Figure 4-8). Chitinase-I: 1 935 Clone38: 681 906 (complementary) 1 100 200 300 400 500 600 700 800 900 Chitinase-II: 1 1082 Clone38: 586 811 (complementary) 1 500 1000 A)B) e-value = 1e-88 e-value = 8e-88(Accession Z70032.1) (Accession AB081944.1) Figure 4-8. Clone 38 has sequence homology to citrus chitinases. Blast-N showed high similarities to (A) a class II chitinase; and to (B) a class I chitinase. The clone 38 had a transcriptional fold induction under the blight condition of around five times, Log2[(nB+0.01)/(nH+0.01)=2.44 (Table 4-1); but a strong visual difference within the membranes was not seen (panel D2, Figure 4-4A). Looking at the

PAGE 52

40 contrast of means (Figure 4-6), two pairs of trees displayed significantly (p<0.05) more transcripts of the gene represented by the clone 38 in blighted than in healthy trees. However, five pairs of trees responded only at a very low level, close to the background (AveH2O). It is possible that the clone 38 represents a responsive gene to citrus blight, but had poor printing on the membrane surfaces for those five abnormal readings. To test that hypothesis, a new series of membranes were printed and run. The figure 4-9 displays the results. 74-H76-B78-H80-BRibosomal LEA Others Chitinase Rep. 1Rep.2 Figure 4-9. Transcriptional profiling of the clone 38 (a chitinase homolog) and other clones. The membranes were manipulated and hybridized under similar conditions, using 5ug of total RNA labeled with P33, from feeder roots samples of a healthy (H) and a blighted (B) tree on Carrizo citrange. The process was repeated twice (rep.1 and 2) generating four membranes. The quantitative difference for the chitinase homolog clone, comparing the healthy and blighted samples (t-Test), was significant (p<0.01). Therefore, the clone 38 is considered to represent a true upregulated gene under the citrus blight condition. The clone 109 is also from the B-H enriched library and displayed mostly higher transcriptional levels under the citrus blight condition. It is 176 bases long and has sequence homology with citrus ESTs in the Genbank using the non-mouse and non

PAGE 53

41 human entry (Figure 4-10), but only poor direct matches as a translated query on the protein databank. CB293790,: 1 753 CK935651,: 78 828 CLONE109,: 662 847 (complementary) 1 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 e-value=9e-56 e-value=9e-44 Figure 4-10. The clone 109 had sequence homology to citrus ESTs. The citrus EST Accession CK935651 was obtained from a library using citrus fruit developing tissues. The CB293790, from a library using citrus cold acclimated tissues. However, blasting the longer citrus EST homolog (Accession CK935651) as a translated query gives a high match (2e-65) with an E3 ubiquitin ligase SCF complex subunit SKP1/ASK1 (At2)/UFO-binding protein (UIP2) of Arabidopsis thaliana (Accession NP568603.1). The clone 109 had visual differences in the membranes (panel B1, Figure 4-4A); had an averaged transcriptional fold induction of around eight times, Log2[(nB+0.01)/(nH+0.01)] =3.12, with p<0.014 (Table 4-1 and Figure 4-5B); and it was significantly upregulated in three out of the ten studied pairs of samples. Therefore, the clone 109 is considered to represent a true upregulated gene under the citrus blight condition. The clone 149 is 144 bases long and has only poor matches in Genbank either as a nucleotide query or as a translated query, but it has a good match with a citrus EST (2e-54) in the Brazilian databank. It is from the B-H enriched library. The clone 149 probably

PAGE 54

42 represents an unchanged gene under the citrus blight condition, considering the three methods used to evaluate its transcriptional pattern. The clone 153 is from the B-H enriched library, but probably represents a true downregulated gene under the citrus blight condition. It displayed visual differences in the membranes (panel E9, Figure 4-4B); significant differences in the cluster analysis with p<0.005 (Figure 4-5B); and two significant differences in the contrast of mean analysis (Table 4-1 and Figure 4-6). However, the sequence information available was poor, with low phred-quality parameter. It was re-sequenced twice with no further improvements. The clone 157 is 195 bases long but had no good match in all three searched databases. Because of its variance (Figure 4-6), it is considered to be another false positive outcome, not representing a true upregulated gene under the citrus blight condition.

PAGE 55

CHAPTER 5 THE RELATIVE TRANSCRIPTIONAL RESPONSES OF THE SELECTED GENES UNDER DIFFERENT INCIDENCES OF CITRUS BLIGHT The process of building cDNA arrays is marked by many steps that go from the collection of samples to make the libraries until the manipulation of the membranes. As a consequence, the final data can carry cumulative sources of nonspecific variation, which may compress or hide legitimate up or down regulated genes under study. With that concern, Chuaqui et al. (2002) stated that researchers must be aware whether the results in microarray based experiments are accurate and the data fundamentally describe the phenomenon being investigated. In addition to employing replicates as used in this study, Chuaqui et al. (2002) considers the need of a validation process. Therefore, an independent RNA assessment method was used to confirm, or not, the transcriptional level of the previously identified genes under non and citrus blight conditions. The effect of cold and drought stresses was also tested. Quantitative Real Time PCR was the Method of Choice Among the available methods, quantitative real time PCR was the method of choice, because of its sensitivity and reliability. It has been used under different platforms and for different purposes, such as confirmation of microarray data in human cancer research (van den Boom et al., 2003), and quantification of a citrus pathogen (Oliveira et al., 2002). It is based on the determination of a fluorescent signal produced during the PCR cycles, allowing the quantification of the amplified product at a real time and the subsequent estimation of its relative original template concentration (Mackay et al., 43

PAGE 56

44 2002). According to Dorak (source: http://dorakmt.tripod.com/genetics/realtime.html last accessed on January 10, 2004), real time PCR is a preferable alternative to study transcription to other forms of reverse transcriptase (RT) PCR, which only detect the final amount of the amplified product. The chemistry employed in real time PCR detection system is the key to the process. There are two general methods for the quantification of the PCR product: the DNA binding reagents (i.e. SYBR Green) and the fluorescent probes (i.e. TaqMan). Characteristics of each detection type and more details about the technique can be seen in Mackay et al. (2002). The TaqMan probe relies on the fluorescence resonance energy transfer (FRET) for quantification. TaqMan probes are oligonucleotides that contain a fluorescent dye, typically on the 5' base, and a quenching dye, typically located on the 3' base. When irradiated, the excited fluorescent dye transfers energy to the nearby quenching dye molecule rather than fluorescing, resulting in a nonfluorescent substrate. TaqMan probes are designed to hybridize to an internal region of a PCR product. During PCR, when the polymerase replicates a template on which a TaqMan probe is bound, the 5' exonuclease activity of the polymerase cleaves the probe. This separates the fluorescent and quenching dyes and FRET no longer occurs. Fluorescence increases in each cycle, proportional to the rate of probe cleavage. Normalization is done against an active internal control such as a ribosomal or other gene. Passive controls, compounds that do not participate in the PCR reaction, are used to adjust the reaction level and background. Other important parameter is the threshold cycle, or Ct, obtained by an arbitrary threshold line chosen within the linear phase of the PCR reaction, that gives the PCR cycle for that fluorescence value. The Ct values are

PAGE 57

45 used to compare different PCR reactions, allowing the quantitative estimation of the initial target template. Compared to DNA dyes, the TaqMan method also has the benefit of producing an amplified target with longer sequence specificity, since it is the product of the hybridization of two primers plus a probe, in a total length that normally stays around 150 bases. Real time PCR using dyes (i.e. SYBER Green) relies only on the specificity of the primers. Considering the described characteristics, the TaqMan probe was the method of choice to validate (or not) the initial results of the cDNA array experiment. The Selected Clones and the Characteristics of the Probes The selected clones observed in the cDNA array were used in this experiment. The clones considered to represent up-regulated genes under the citrus blight condition were the clones 25, 38 and 109. The clone 149 appeared to be nearly unchanged. The clone 153 was considered to be down-regulated; however the sequence information for this clone was poor, even after re-sequencing. Thus, the clone 6 was included in this experiment as a candidate to represent a down-regulated gene under the citrus blight condition. Those clones were chosen because genes with altered transcriptional levels under the citrus blight condition can be valuable tools on further comprehensive experiments. Genes with unaltered patterns can be useful references for normalizations or for other biological needs as well. Based on sequence homology, the clone 38 represents a citrus chitinase gene (Table 4-2, chapter 4). Chitinases are known to respond to different forms of stresses, such as fungal pathogens in citrus plants (Fanta et al., 2003). Citrus blight causes mineral unbalance with a clear pattern for zinc deficiency in leaves. Taylor et al. (1996), identified a small protein, named P5 (Accession AAB46813), that has sequence

PAGE 58

46 similarities to chitin binding proteins and functions complexing zinc in affected plants. It is probably associated with the known translocation of zinc from leaves to trunk tissues (Albrigo and Young, 1981). In addition to the clone 38, chitinase homologs were found in our libraries and in combination with the class I Chitinase (Accession AB081944.1) as reference, a putative nucleotide sequence for a P5 candidate gene was deduced and was used in this validation experiment (Figure 5-1). Citrus chitinaseclass I, Accession AB081944.1 1e-05 true P5 Highest match on sequence search analysis e-values protein 22 residues22 residues true P5AB081944.1 e-value=1e-05 Chitinase-I: 1 935 p5-candidate: 70 135 Clone38: 681 906 (complementary) 1 100 200 300 400 500 600 700 800 900 region of the citrus chitinasegene used in the real time PCR A)B)C) Figure 5-1. Designing a candidate sequence for the P5 gene. A) The true P5 protein has homology to the translated version of the citrus chitinase class I Accesion AB081944.1. B) The aminoacid identities was 80%, with an e-value of 1e-05. C) Based on the reference chitinase class I (Accession AB081944.1), which also had high homology to the clone 38, a candidate sequenced was designed for the P5 gene and was used in the real time PCR experiment.

PAGE 59

47 Another gene associated to citrus blight is the P12 (Accession AF015782). It is up-regulated in affected plants (Derrick et al., 1990) and has sequence similarities to expansins (Ceccardi et al., 1998), but has so far an unknown role in the decline of the trees. P12 was also included in this experiment. The Table 5-1 displays the characteristics of the chosen probes for each of the candidate clones and P12. Table 5-1. Characteristics of the chosen probes. The primers and probes were designed using the PrimerExpress Software (Source: Applied Biosystems Resources, http://www.appliedbiosystems.com/index.cfm last accessed December 10, 2003). Clone name Start (bp) Length (bp) Tm ( o C) GC (%) Taqman probe architecture (flourochrome-5sequence 3-quencher) 6 136 29 69 45 fam6-TCGCTACTTATGCGACAAGGAATTTCGCT-tamra 25 192 32 69 41 Fam6-TTGAAGGCAAGTTAGGAAATTAGCAAAGCCAG-tamra 109 94 39 69 33 fam6-ATGATACAGAGAAGGTTGGGATGATATGACATTAAAACA-tamra 149 213 26 69 42 fam6-CTGTATCATCTTACTTTACGCTTCCC-tamra P5 candidate 78 18 68 67 fam6-AAGCGGCGTTGTGTGCCC-tamra P12 179 25 69 56 tet-TGGAGTCATGATAGCCGCAGCAAGC-tamra Two major calculation methods are possible, the absolute and the relative quantification of the target gene (Source: bulletin # 2, Applied Biosystem, http://www.appliedbiosystems.com/index.cfm last accessed December 10, 2003). This study employed the relative method. Then, the 18S gene was chosen as the active internal control, or normalizer, because it is considered to be non altered under different conditions (Dorak, 2004, source: http://dorakmt.tripod.com/genetics/realtime.html last accessed on January 10, 2004). However, no previous information was seen about the 18S levels in feeder root samples of citrus hybrids under non and citrus blight conditions. It was assumed that the transcriptional level of the 18S gene was not altered by citrus blight.

PAGE 60

48 To test the efficiency of the amplification between the target gene and the normalizer, a RNA 2 fold dilution series was evaluated, from 3.9 to 1000 ng of a mixed pool of RNA templates. Under a similar efficiency, the graphic of the log input amount of the total RNA and Ct (difference between the chosen thresholds, the Ct of the target minus the Ct of the normalizer) would give a straight line, nearly parallel to the abscissa. Small deviations from that are accepted and the bulletin # 2 of Applied Biosystem (Source: http://www.appliedbiosystems.com/index.cfm last accessed December 10, 2003) recommends that the absolute value of the slope (of the estimated linear function) should be equal or smaller than 0.1. Figure 5-2 displays the outcome for the clone 109, which had a slope of 0.1055, for the tested range of 3.9 to 1000ng of RNA template. A) Plot of log input amount versus Ct y = 0.1055x + 26.138 2224262830-2.4-2.1-1.8-1.5-1.2-0.9-0.6-0.30.0 clone 109 x 18S Linear (clone 109 x 18S)CtLog ng Total RNA Figure 5-2. Relative PCR efficiency plot of the clone 109 against the normalizer 18S. In the same type of analysis, the clone 6 had a slope of 0.0104; the clone 25, 0.1029; the clone149, 0.0602; the P5 candidate, 0.0811; and the P12, 0.0196.

PAGE 61

49 Collecting and Preparing the Samples The same citrus groves in central Florida area, used to collect root samples for the cDNA array experiment, were revisited, with the expectation to re-assess the same trees. However, most of the previously healthy trees had started to display some blight symptoms and some of the affected trees had been pushed out. Therefore, some other plants were used in the real time PCR experiment, after verification of the typical symptoms and the water test (Lee et al., 1984) for diagnostic purposes. Three classes of plants were sampled: Healthy: with no visual symptoms in the canopy, and with a minimum uptake of water, manually injected into the trunk by a syringe, of 3 ml/10 seconds. Mildly affected: initial symptoms in the canopy, such as opaqueness of leaves and few twig dye backs on the top. Uptake of water from 1 to 2 ml/10 seconds. Blighted: fully symptomatic canopy. Uptake of water smaller than 0.5ml/10 seconds. Tentative different stages of the process (healthy, mildly and fully affected trees) were used because they may offer more information than only comparing healthy and fully declined trees, as done on the cDNA array experiment. Feeder root tissues from each tree were once again collected and the RNA extraction procedures had this time double DNA digestion step, using an adaptation of the Qiagen RNeasy kit (Source: Qiagen, http://www1.qiagen.com/Default.aspx last assessed January 4, 2004). No DNA contamination was observed when verifying RNA quality on formaldehyde gels. All sampling and extracting procedures were done as uniformly as possible. The sampled trees were Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) on the rootstock Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.) of around 10 years old. This choice was considered because

PAGE 62

50 trees of Valencia on Carrizo have been largely used in the Florida citrus industry. The combinations of Valencia on Rough lemon and on Swingle citrumelo were also initially considered. But the sampled trees were of different ages and had other non uniform conditions. Therefore, they were no longer considered. Reverse Transcriptase (RT) and PCR Reactions Duplicated RT reactions and two PCR reaction sets per each sample were used to minimize pipeting and other reaction errors, resulting initially, in four mechanical replicates per each biological sample. Both RT reactions for each sample were done at the same time and under the same conditions, using a 96 well plate and the protocol given by the Applied Biosystem RT kit (Source: http://www.appliedbiosystems.com/index.cfm last accessed January 10, 2004). Other conditions were also as uniform as possible. The RT plate was then stored at -20 o C. The 7700 Applied Biosystem sequence detection system was employed for all set of PCR reactions. The PCR conditions were 50 o C/2minutes, 95 o C/10minutes, and 45 cycles of 95 o C/15seconds plus 60 o C/1minute. An arbitrary threshold cycle (Ct) was chosen for each amplification plot. The data was analyzed using the baseline computation method on the Apple based Sequence Detection System (SDS) v1.9 software (Source: http://www.appliedbiosystems.com/index.cfm last accessed April 10, 2004). Three controls were employed in all real time PCR plates: the NTC (no template control, to verify non specific amplifications), the NRT (no reverse transcriptase control, to verify DNA contamination) and the NAC (no amplification control, to verify non specific readings). After the reactions, none of the NACs gave integral fluorescent readings. The great majority of the NTCs did not reach the chosen threshold level (Ct) for

PAGE 63

51 the target gene at the 40 th PCR cycle. The reactions that did, were discarded. All samples that had a difference in the Ct (the Ct of the sample minus the Ct of the NRT) for the reading of the 18S gene, smaller than 10 cycles were also discarded. The Applied Biosystem (Source: Taqman ribosomal RNA control reagents protocol, http://www.appliedbiosystems.com/index.cfm last accessed April 10, 2004) recommends this cut off level to avoid significant contribution of non-target templates to the measured gene. The remaining healthy, mildly affected and blighted samples were further analyzed. Figure 5-3 displays examples of amplification plots for the clone 109, P5 candidate and P12. Relative Quantification of the Transcriptional Levels of the Selected Genes All samples were analyzed in groups according to their similar levels of 18S, measured by the threshold (Ct) in each reaction. Four replicates, groups of independent healthy, mildly affected, and blighted trees, were analyzed, in a total number of twelve different sampled trees. Each clone was evaluated individually. The contrast to compare healthy-mildly affected-fully blighted conditions is not orthogonal. Therefore, the t-Test can not be used. The residual standard deviation (s) was chosen to indicate the non specific variation within each group of plants, or replicates. This measure of variation indicates the effect of non controlled factors, and therefore, seems adequate to estimate the contribution of errors in the experiments. It was calculated using the mean square of the residue (MSres), after the ANOVA (analysis of variance) for each group of plants, as: s = (MSres)

PAGE 64

52 A) The clone 109 probe used the fluorochromes FAM6 and TAMRAB) The P5 probe used the fluorochromes FAM6 and TAMRARnRnPCR cyclesPCR cyclesC) The P12 probe used the fluorochromes TET and TAMRA RnPCR cycles Figure 5-3. Examples of real time PCR reactions. A) For the clone 109. B) For the P5 candidate gene. C) For the P12. The reactions were done with samples from all the tested conditions and reflect the differences in the transcriptional level of the target gene according to each sample.

PAGE 65

53 Figure 5-4 displays the relative transcriptional levels of the selected clones. Within each group of plants, regression analysis was performed and a polynomial equation was shown. The three evaluated stages of the disease (healthy, mildly and fully blighted trees) are certainly not enough for better inferences about the progress of the disease, neither blight can be precisely estimated, since no pathogen is known to cause the problem. But an initial tendency of the relative transcriptional levels compared to the healthy samples, was observed for some of the clones. The clone 6 can probably be considered a mistaken choice taken from the cDNA array experiment. The tentative down-regulated pattern was not observed in the real time PCR experiment. Moreover, it seems to be up-regulated in affected comparing to healthy plants. More replicates, or maybe other techniques, would probably be needed for further elucidation The clone 25 was considered to be up-regulated in affected plants, especially when Swingle rootstock was used (panel C1, Figure4-4C). This pattern was also observed in the real time PCR experiment using Carrizo rootstock, confirming the previous expectation. In the group D of plants, the fully blighted tree displayed similar transcriptional levels to the healthy tree, but both were lower than the mildly affected one. It is possible that fully blighted trees reduce some of their metabolism affecting the transcriptional level of certain genes. That was apparently the case for this gene in this studied blighted tree. In the other three groups of plants, the relative amount of the transcripts of the gene represented by the clone 25 were more abundant in fully blighted than in healthy or mildly affected plants. Therefore, the clone 25 was considered to represent an up-regulated gene under citrus blight conditions.

PAGE 66

54 1.004,779.65 10,779.11 y = 610.4x2 + 2947.4x 3556.8R2 = 1-4,000.00-2,000.000.002,000.004,000.006,000.008,000.0010,000.0012,000.0014,000.00HczMczBcz P5 candi date-A Poly. (P5 candidate-A) 1.0 0 1.0 2 4.21Mcz y = 1.5845x2 4.7342x + 4.1497R2 = 10.000.501.001.502.002.503.003.504.004.505.00HczBcz P5 candidate-B Poly. (P5 candidat e-B) 1.000.27y = 2.5281x2 8.3178x + 6.7896R2 = 1-1.000.001.002.003.004.005.006.00HczMczBcz 4.59 P5 candidate-C Poly. (P5 candidate-C ) 1.001.2910.29y = 4.352x2 12.764x + 9.41212 = 1-4.00-2.000.002.004.006.008.0010.0012.0014.0016.00HczMczBcz R P5 candidate-D Poly. (P5 candidate-D) 1.009.10y = -327.09x2 + 1312.4x 984.33R2 = 1-50.000.0050.00100.00150.00200.00250.00300.00350.00400.00HczMczBcz 332.14 P12-A Pol y. (P12-A) 1.002.02y = -11.321x2 + 45.794x 33.473R2 = 10.002.004.006.008.0010.0012.0014.0016.00HczMczBcz 12.83 P12B Poly. (P 12-B) 1.00241.94195.73y = -143.58x2 + 671.68x 527.R2 = 1-100.00-50.000.0050.00100.00150.00200.00250.00300.00350.00HczMczBcz 1 P12-C Poly. (P12-C) 1.00205.642.28y = -204x2 + 816.63x 611.R2 = 1-100.00-50.000.0050.00100.00150.00200.00250.00300.00HczMczBcz 63 P12-D Poly. (P12-D)10.28+/-0.50 13.50+/-1.74 15.82+/-2.40 22.39+/-2.02 18S Ct +/-StDev P5 candidateP12Hcz Mcz Bcz Hcz Mcz Bcz Hcz Mcz Bcz Hcz Mcz Bcz trees 205 223 404 201 225 402 203 227 406 265 229 454 1.440.9900y = 0.224 0.6798x + 1.45510.000.200.400.600.801.001.201.401.601.80HczMcz 1.7x2 -R2 = 1Bcz c6-A Poly. (c 6-A) 2.541.0Bcz 1.360y = 0.4118x2 0.8761x + 1.4643R2 = 10.000.501.001.502.002.503.00HczMcz c6-B Poly. (c6-B) 4.501.551.00y = 1.2039x2 3.0645x + 2.8606R2 = 10.001.002.003.004.005.006.00HczMcz Bcz c6 -c Poly (c6-c) 15.3133.611.00y = -25.451x2 + 108.96x 82.507-20.00-10.000.0010.0020.0030.0040.0050.00HczMczBcz R2 = 1 c6-D Poly. (c6-D)rep. A rep. B rep. Crep. D 3.401.311.00y = 0.887x2 2.3489x + 2.4619R2 = 1-1.000.001.002.003.004.005.006.00HczMczBcz c25-A Poly. (c2 5-A) 4.031.621.00y = 0.8955x2 2.0653x + 2.1698R2 = 10.000.501.001.502.002.503.003.504.004.505.00HczMcz c25-B Poly. (c 25-B) 9.971.311.00y = 4.1736x2 12.21x + 9.0362R2 = 1-4.00-2.000.002.004.006.008.0010.0012.0014.00HczMczBcz c25-C Poly. (c25-C) Bcz 3.3325.701.00y = -23.538x2 + 95.317x 70.779R2 = 1-10.00-5.000.005.0010.0015.0020.0025.0030.0035.0040.00HczMczBcz c25-D Poly. (c25-D) 1.000.340.41y = 0.36R2 = 10.000.200.400.600.801.001.20HczMcz 44x2 1.7513x + 2.3869Bcz c109-A Poly. (c 109-A) 1.001.714.50y = 1.039x2 2.4045x + 2.3655R2 = 10.001.002.003.004.005.006.00HczMcz c109-B Poly. (c109-B) 1.0 0 3.8 5 21.2 0 y = 7.2564x218.924x + 12.66 7 R2 = 1-10.00-5.000.005.0010.0015.0020.0025.0030.0035.00HczMczBcz c109 -C Poly (c109-C) Bcz 1.0 0 53.4 9 25.4 8 y = -40.246x2 + 173.22x 131.9 8 R2 = 1-20.00-10.000.0010.0020.0030.0040.0050.0060.0070.00HczMczBcz c109-D Poly. (c109-D) 1.002 0.000.02y = 0.5072x 2.5177x + 3.0105R2 = 1-0.200.000.200.400.600.801.001.20HczMczBcz c149-A Poly. (c149-A) 1.001.873.05y = 0.1618x2 + 0.3797x + 0.4585 = 1-0.500.000.501.001.502.002.503.003.504.004.50HczMcz c149-B Poly. (c149-B) 1.000.1226.11y = 13.437x2 41.196x + 28.759R2 = 1-15.00-10.00-5.000.005.0010.0015.0020.0025.0030.0035.0040.00HczMczBcz c149R2Bcz C Poly. (c149-C) 1.0016.815.96y = -13.331x2 + 55.805x 41.475R2 = 1-10.00-5.000.005.0010.0015.0020.0025.00HczMczBcz c149-D Poly. (c149-D) clone 6clone 25clone 109clone 149 y = 0.7867x2 2.9304x + 3.1437R2 = 10.000.200.400.600.801.001.201.401.601.802.00HczMcz Bcz 33212241204107794.214.5910.290.023.0526.115.961.424.5021.2025.483.404.039.973.331.442.544.5015.31111111111111111111111111 fold change to healthy (1x) Figure 5-4. Relative transcriptional levels of the selected genes. Twelve individual plants were analyzed in four groups (rep. A, B, C and D), based on similar Ct values for the 18S readings. Feeder root samples of healthy (H), mildly affected (M) and fully blighted (B) trees on Carrizo (cz) rootstock were numbered (201 to 454) and used in this experiment. The columns represent the relative transcriptional level of each clone, comparing to healthy samples. The bars represent the residual standard deviation estimated for each group of plants.

PAGE 67

55 The clone 109 was up-regulated in three out of the four replicates, comparing the healthy to any of the two blighted conditions. The pattern was similar in the groups B and C. The mildly affected plant of the fourth replicate (group D) displayed even higher transcriptional level than the fully blighted tree. In addition, this pattern of the second and third replicates (groups B and C of plants) may represent only an initial or partial outcome for a latter blight development that has yet to come, and could reduce the transcripts of the gene represented by the clone 109 to a similar level observed in the group D of plants. That seems possible because citrus blight can not be precisely quantified so far, and therefore, the progression of the disease can not be completely pursued either. The first replicate (group A of plants) displayed a non expected result. The relative transcriptional level of the mildly affected plants was reduced to a range of 40% comparing to the healthy trees. To test whether or not that was an artifact of the method, a new RT and PCR reactions were run with the same RNA samples. No further discrimination was observed, however the residual standard deviation was higher. The mildly affected tree still had around 40% of the transcripts compared to the healthy sample while the blighted reached 1.4 the healthy sample. It is possible that this result may represent noise caused by blight itself, or by another factor not studied in the system. New samples from the field were not collected and tested because that would add a contribution of the season of the year and temperature to the present model. In addition, the sampled plants are probably in different stages of blight at this point, because all studied samples were collected and processed six months ago, from January 9th to the 15th of this year.

PAGE 68

56 Considering that the clone 109 was up-regulated in three out of four replicates; that a similar pattern was observed; and that the level of induction compared to healthy plants was above twenty fold in the groups C and D of plants; the clone 109 was considered to represent a gene that is upregulated under citrus blight conditions, confirming the expectation from the cDNA array experiment. The clone 149 was variable but with no defined pattern considering the outcome of the four groups of plants. On the cDNA array, the transcriptional level of the gene represented by this clone was nearly unchanged by citrus blight. However, since the real time PCR is more sensitive, discrepancies towards either way, being up or down-regulated according to individual organisms, could be unfolded because in the cDNA arrays the data tends to be compressed. That was apparently the case, and the clone 149 does seem variable, but according to other factor not studied. The P5 candidate gene was responsive to citrus blight with a similar pattern in all four replicates. It was more abundant in fully blighted trees than in any other studied condition. Apparently the gene represented by the P5 candidate displays an ascendant transcriptional response towards the fully blighted tree. This result may not parallel the initial complexation of zinc seen in leaves and trunks of mildly affected trees (Albrigo and Young, 1980). Therefore, the chitinase gene represented by the P5 candidate is considered to be upregulated under citrus blight condition, confirming the similar expectation observed with the clone 38 in the cDNA array experiment; however, it remains to be investigated whether or not this candidate gene is the true P5 identified by Taylor et al. (1996). It is worthy to note that the behavior of the true P5 is not known in feeder root tissues, as used in this experiment.

PAGE 69

57 The P12 displayed a significantly higher transcriptional level on mildly affected plants compared to healthy or fully blighted trees. The pattern was similar in all four groups of tested plants. It is possible that if more stages of blight were available, a wider range of conditions would be better evaluated, and maybe a different polynomial equation would better describe the phenomenon. But the pattern for a higher transcriptional level in mildly affected trees was clear. The lower amount of P12 transcripts in fully blighted trees may account for the previous failure in detecting P12 in the subtracted libraries and also in the cDNA array experiment. Both procedures employed samples from fully symptomatic trees. Although patterns were observed, the intensity of the responses was related to the group of studied plants. Therefore, a general, or averaged, transcriptional level for each gene was not estimated, because it may not meaningful. The Potential Biological Meaning of the Clone 109 Regardless of what causes citrus blight, affected plants visually go to a declining condition during the whole process. A closer look at the clone 109 reveals sequence homology to a citrus EST annotated as a cold acclimated responsive gene from the Washington Navel sweet orange (Citrus sinensis L. Osbeck cv. Bahia), in experiments done in California (Close et al., 2003; in press; assession number CB293790.1), shown in Figure 4-10 (Chapter 4), with an e-value of 9e-44. It is known that plants respond to cold and to drought periods activating and repressing responsive genes. Seki et al. (2001) found 5 drought specific inducible genes, 2 cold specific inducible genes and 16 drought and cold inducible genes using a microarray with around 1,300 full length cDNAs of Arabidopsis thaliana. Citrus blight is a xylem blockage problem, which ultimately may lead to water deficiency, and maybe,

PAGE 70

58 that gene represented by the clone 109 is involved in this process. In addition, the clone 109 displayed mostly higher transcriptional levels under citrus blight condition, using samples from different seasons of the year. An alternative hypothesis could claim specific response to the citrus blight process, rather than being induced only as an indirect effect of the internal water stress caused by citrus blight. The clone 109 also had potential similarities to an ubiquitin subunit (Table 4-2, chapter 4) of the SCF complex (Skp1-Cullin-F-box protein). This complex is involved in the ubiquitination of proteins and cell cycle regulation. The citrus blight process can be seen as an accelerator of senescence. Young affected trees, of 5 to 10 years old, start to display an overall decline and lack of vigor, normally only seen on healthy trees older than 50 years. The clone 109 could function in the ubiquitination process, targeting other proteins for degradation, and or, impairing the normal cell cycle in affected trees. Experimental confirmation is certainly needed either way regarding its function and involved pathways. But in the first scenario, if cold acclimation is a feasible goal in plant genetic engineering programs, maybe hardening rootstocks for citrus blight could be as well. In the second scenario, experimental effort on the cell cycle regulatory process in citrus can eventually offer perspectives to control citrus blight. A Tentative Test to Verify the Effect of Cold and Drought Stresses In order to evaluate whether or not the gene represented by the clone 109 responds to cold and drought stresses, another experiment was assembled and run. Other clones were also included. The major objective was to compare the transcriptional levels of the selected genes, and determine whether they seem to be a specific response to citrus blight, or only a secondary effect of the stress caused by the disease. However, moving

PAGE 71

59 adult citrus plants to controlled conditions was not attempted. Greenhoused young trees, of about one year old were used instead. They were exposed to cold (4 o C for 52 hours) and drought stresses (no water for one week plus additional root airing for 24 hours). Seedlings of Carrizo citrange (Citrus sinensis L.. Osb. x Poncirus trifoliata L. Raf.) were chosen to equal the genotype evaluated in the previous real time PCR experiment. The references, or control plants, were from the same lot of seedlings, but kept under normal greenhouse conditions for the same period of time. Four independent replicates (i.e. different seedling plants) were evaluated for each treatment. Feeder root tissues were once again used. Figure 5-5 shows the results. The contrast of interest, to compare the control, cold and drought treatments, is not orthogonal, and therefore, the t-Test can not be used. However, the number of replicates was uniform, allowing the application of a test to compare the means. Among the options, the Tuckey test was employed because of the better discrimination of the means compared to other tests, like Sheffe, Dunnett or Duncan (Banzatto and Kronka, 1992). The null hypothesis was no difference among the treatments. The clone 109 did not respond to cold and drought stresses under the studied conditions. No significant effect (p<0.05) was seen for most of the other clones either, including P12. The drought treatment only affected the gene represented by the clone 25. Its transcriptional level was reduced to around one-third compared to the control. This level of significance (p<0.05) implies that the maximum estimated chance of having similar results caused only by chance is of only 5%. Lowering this probability to 1% (p<0.01) makes this difference became not significant. In spite of which level of probability should be considered, this reduction of the transcriptional level was not

PAGE 72

60 expected. Blight induces xylem blockage; and consequently, a potential internal water stress. Since the clone 25 was considered to be up-regulated in the cDNA array and in the real time PCR experiment, it would also be expected to be unchanged or up-regulated 0.831.021.000.000.200.400.600.801.001.201.40seedlingsdroughtcold c6-A 0.900.381.000.000.200.400.600.801.001.201.40seedlingsdroughtcold c25-A 1.000.841.130.000.200.400.600.801.001.201.401.60seedlingsdroughtcold c109-A 0.640.661.000.000.200.400.600.801.001.201.401.601.80seedlingsdroughtcold c149-A 1.000.690.650.000.200.400.600.801.001.201.401.601.80seedlingsdroughtcold P5 candidate-A 1.000.321.23-1.50-1.00-0.500.000.501.001.502.002.503.00seedlingsdroughtcold P12-A clone 6clone 109P5 candidate clone 25clone 149P12 =0.41 =1.15=2.44=1.16=0.48=0.77 Figure 5-5. Contrasts comparing the effect of cold and drought treatments on the relative transcriptional level of the clone 109 and other clones. The Tuckey test was employed and an observed significant difference was highlighted with a star. The bars represent the minimum significant difference for the contrast ( values, for p<0.05) of each clone. under drought stress. Another experiment using adult trees under controlled conditions may address this question. Similar reasoning may be considered for the other genes as well. A definitive answer about the effect of major stresses would need adult trees under

PAGE 73

61 controlled conditions. Therefore, the responsive genes to citrus blight, and apparently not affected by drought and cold stresses, seem to be at this point, only the clone 109, the P5 candidate and the P12. The Potential Effect of Redundancy and Gene Families The selected clone 38 represents a citrus chitinase gene. Two genes (chitinase class I, accession AB081944.1; and chitinase class II accession Z70032.1) were matched to the 235 bases of the clone 38 with the same e-value of 1e-87. The homology was seen towards the 3end of the subjected sequences (Figure 4-8, chapter 4). Chitinases are enzymes that catalyze hydrolysis of chitin polymers, acting against plant intruders by destroying chitin-containing cell walls. The difference between the class I and the class II is based in the presence (I) or absence (II) of a N-terminal chitin binding domain (source: http://us.expasy.org/cgi-bin/prosite-search-ac?pdoc00620 last accessed June 1, 2004). Therefore, the clone 38 arrayed in the cDNA array membranes (chapter 4) could represent any of the two described genes. The 22 aminoacids of the P5 protein (accession AAB46813) had 80% identity to the same chitinase class I, accession AB081944.1, with an e-value of 7e-05. The matched region was however in the conserved 5 end of the subjected sequence. The Taqman probe and primers were designed to span this region, covering 88 nucleotides long. The Figure 5-1 displays the homolog regions between the sequences of the reference chitinase, clone 38 and P5 candidate. Therefore, it is not possible to discern precisely if the clone 38 and the P5 candidate represent the same gene, or not, within the chitinase family. In addition, the sequence for citrus is limited, but searching the genome of the model plant Arabidopsis thaliana, the reference chitinase class I, accession AB081944.1, had 15 matches in the protein

PAGE 74

62 databank (source: TAIR, http://www.arabidopsis.org/ last accessed June 1, 2004), with e-values varying from 3e-7 to 4e-95. It is possible that this situation is similar for citrus, with a high number of redundant genes, imposing another level of difficulty in the evaluation of transcriptomes. The situation can be similar for P12, which had 17 matches in the protein database of TAIR, all expansins, with e-values from 2e-04 to 5e-30. There are two gama expansins with noted similarity to P12. Therefore, it is possible that more copies of homolog P12 genes are also present in citrus plants.

PAGE 75

CHAPTER 6 IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY ASSOCIATED WITH THE CITRUS BLIGHT DISEASE A crucial question in citrus blight research is the origin of the problem. Several causal agent candidates and other theories have been examined, but none has proved to be definitive for blight. Each reported theory has been based on observed characteristics of the disease; however blight is a complex problem, and different views and perspectives are possible under different circumstances. For instances, the non-transmissibility of citrus blight by canopy tissues was observed long ago (Rhoads, 1936; and COHEN, 1968), but transmission by root grafting was later obtained (Tucker et al., 1984). Soils with higher pH and levels of Ca were associated with severe incidences of blight in Florida (Wutscher, 1989); however, blight does occur in acidic soils and examples are common, especially in the So Paulo citrus belt. Soil born pathogens were associated to blight (Nemec et al., 1982); but transmission of blight by soil replacement was not obtained (Timmer and Graham, 1992). Xylella fastidiosa was considered to be the causal agent of blight (Hopkins, 1988); but this xylem limited bacterium is naturally present in many vascular plants, including citrus. A higher incidence of blight was observed after the implementation of the nucellar programs in Florida and in So Paulo (Derrick and Timmer, 2000), probably reflecting the new endophytic balance in the nucellar plants but blight has also a high incidence on certain old line trees, and examples on Olmpia, and on other clones of the Pera sweet orange that were not pre-immunized against citrus 63

PAGE 76

64 tristeza virus (CTV) are commons in So Paulo. The theory about a molecular origin for blight suggested the involvement of defective signals transmitted from plant to plant (Carlos et al,2000), but no experimental evidence for that have been found. Strains of citrus tristeza virus (CTV) were reported to be associated with a potential variant of blight (named Rangpur lime decline) and to citrus sudden death diseases in Brazil (Derrick et al., 2003), but CTV is commonly present in stem-pitted sweet orange plants in So Paulo causing no major disturbances (Costa et al., 1954). In spite of the merit and investigations of each theory, one characteristic initially proposed by Swingle and Webber (1896) seems difficult still to be denied: citrus blight apparently has an infectious nature and dissemination. In order to investigate a potential causal agent for citrus blight, qualitative and quantitative experiments were performed. The First Screening of the Subtracted Libraries When the subtracted libraries were made (chapter 3), during the summer of 2001, the following question regarding which genes were represented there was first addressed using a virtual northern blot experiment. The differentially enriched cDNAs from the blighted minus healthy (B H) library were cloned and grown in the E. coli vector and plated on LB medium. Ninety six clones were randomly selected for re-growing on two equally printed nylon membranes. After growth, the membranes were rinsed with extensive washes and hybridized overnight with P32 labeled probes. The probes were also made from the subtracted blighted (B H) and healthy (H B) enriched cDNA libraries, uncovering tentatively more abundant transcripts represented there. A B H clone when revealed by the B H probes and not by the H B probes implied a larger presence of the transcript of that gene in the blighted samples that originated the library, from the affected trees. To the contrary, a B H clone

PAGE 77

65 when revealed by both types of probes ( B H and H B ) indicated similar amounts of the transcript in both libraries, and therefore, were of no interest. Virtual northern blots can be effective to reveal abundant transcripts in enriched libraries (Source: http://www.bdbiosciences.com/clontech/ last accessed January 15, 2004), which may include those of RNA based viruses, such as the majority of the plant viruses. Figure 6-1 displays the results for the virtual northern blot experiment. Among others, the arrows indicate that the clones E8-13 and E8-14 were far more abundant in the B H library. B -H probes H -B probesB H clones E8-13E8-14 Figure 6-1. The virtual northern blot of the B H clones. Each clone was printed in two membranes that were respectively probed with B H and H B probes labeled with P32. The subtracted libraries were made using the superficial roots of Rough lemon (Citrus jambhiri Lush) collected in the central Florida area, during the summer of 2001. The green arrows indicate the clones E8-13 and E8-14. Sixteen clones were considered to be differentially transcribed in the B H library and were selected for sequencing. Among the plant genes, possible up regulated ones included clones with similarities to metallothionein and several unknowns. However,

PAGE 78

66 based on sequence homology, the clones E8-13 and E8-14 were not plant genes, but part of the 3end of the P27 divergent citrus tristeza virus (CTV) coat protein gene. The clone E8-13 was 311 base pairs long and the E8-14 was revealed to be only a shorter and redundant version of the same gene, with 155 base pairs. Table 6-1 displays the first four nucleotide based homolog sequences found in all databases of Genbank for clone E8-13, using the Blast-N search analysis (Altschul et al., 1997). Similar results were found using the translated query searches. Table 6-1. The clone E8-13 had homolog sequences matching different isolates of the citrus tristeza virus (CTV). The clone E8-13 is 311 base pair long and was found in the citrus blighted minus healthy (B-H) subtracted library made from superficial roots of Rough lemon (Citrus jambhiri Lush). clone Homolog gene e-values E8-13 > gi|1732493|gb|U56902.1|CTU56902 Citrus tristeza virus p346, 54-kDa RNA dependent RNA polymerase, p33, p6, p65, p61, p27, 25-kDa coat protein (CPG), p18, p13, p20, and p23 genes, complete cds e-162 > gi|11414863|dbj|AB046398.1| Citrus tristeza virus genomic RNA, complete genome, seedling yellows strain e-147 > gi|2098825|gb|AF001623.1|CTAF001623 Citrus tristeza virus, complete genome e-136 > gi|3550999|dbj|AB011189.1| Citrus tristeza virus genomic RNA for 27K protein and coat protein, partial cds, isolate KS3A2 2e-87 Looking at the homolog sequences, the first match, accession number U56902, was isolated in Israel and represents the coat protein gene of CTV (Mawassi et al., 1993). The second, accession AB046398, was a seedling yellows strain from Japan (Suastika et al.,

PAGE 79

67 unpublished). The third, accession AF001623, was isolated in Texas causing severe symptoms in sweet orange (Yang et al., 1999). The fourth, accession AB011189 is the P27 gene also isolated from strains of Japan (Kano et al., unpublished). Sequence information is limited for citrus, but to verify other potential origins for the clone E8-13, other databases were searched. Homolog sequences were not found in different plant databases. The first matched outcome in the blast searches (Altschul et al., 1997) using the non-mouse and non-human ESTs and the viridiplantae databases of the Genebank yielded respectively matched sequences with e-values of only 5.7 and 1.1. Using the Brazilian citrus EST database, the result was not different, with an e-value of only 0.18 for the first outcome. Therefore, the clone E8-13 was considered to be part of the P27 gene of the CTV genome. The highest matched sequence indicated similarities to an Israeli isolate. In addition, both clones, E8-13 and E8-14, displayed polyadenylated 3ends in their sequences. The question whether that was an artifact of the method used to build the cDNA libraries (Smart cDNA kit, source: http://www.bdbiosciences.com/clontech/ last accessed June 20, 2001), or an adaptation of the CTV to the molecular machine of citrus remains to be investigated. The presence of CTV in citrus plants is not novel, and the solution for the Tristeza disease in sweet orange groves was reported long ago (Costa et al., 1954). However, the presence of CTV strains in roots of citrus plants affected by blight is intriguing. This result was similar to previous observations made by Derrick et al.(unpublished). It is also necessary to review that Rough lemon is susceptible to CTV and those findings alone may not add novelty to what is already known about CTV and citrus host interactions.

PAGE 80

68 Other CTV Genes were also Found in both Subtracted Libraries Later, around four hundred clones randomly selected from the blighted minus healthy (B-H) and 100 from the healthy minus blighted (H B) enriched libraries were sequenced. Clones with sequence similarities to other CTV genes were relatively common, matching different isolates and strains. Redundancy was observed, and normally, more than one clone matched the same analyzed gene. The Table 6-2 displays examples of pieces of CTV genes found in both libraries. Table 6-2. Other sequences with homology to CTV genes found in the blighted minus healthy (B-H) and healthy minus blighted (H-B) subtracted libraries. library CTV gene e-value clone B H p23 e-153 M1H7 p346RDRPol 6e-70 M2E2 p33 3e-43 M3A1 H B p61 e-131 E8-28 hsp90 p61 e-158 E8-21 p65 0.000 E8-22 Quantitative Evaluation of the P27 Candidate Gene in the Blighted Trees In order to evaluate the presence of the transcripts of the P27 candidate gene, reverse transcriptase quantitative real time PCR was run. The chosen Taqman probe employed the Fam6-5sequence 3-Tamra architecture and had an accepted PCR efficiency (Chapter 5). The absolute value for the slope of the linear function (y= -0.1008x +22.13), that described the log input amount of the total RNA (x) against the Ct (y) between the target and the normalizer 18S, was 0.1008. The tested range was a 2 fold dilution series from 3.9 to 1000 ng of total RNA. The

PAGE 81

69 comparative Ct method was used to avoid the need of in vitro synthesis and purification of the P27 RNA for absolute quantifications. For calculation purposes, reactions that did not display amplifications were considered having a Ct value of 45 cycles. To minimize non-specific variations, two reactions per sample were run. The same healthy, mildly affected and fully blighted plants that were used to evaluate other genes (Chapter 5) were used here. They were Valencia sweet orange (Citrus sinensis L.. Osbeck cv. Valencia) on Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L.. Raf.) of around 10 years and from the central Florida area. Four groups of plants, or replicates, were evaluated. The transcripts of the P27 candidate gene were present in far larger amounts in affected than in healthy trees (Figure 6-2). The fully blighted trees on the group B of plants still have around eight times more P27 transcripts than the healthy plants. 1.00.8 713,720-100,000.0200,000.0300,000.0400,000.0500,000.0600,000.0700,000.0800,000.0HczMczBcz P27-A 1.08.8 205,674-50,000.0100,000.0150,000.0200,000.0250,000.0300,000.0350,000.0400,000.0HczMczBcz P27-B 1.0142.5270.6-50.0100.0150.0200.0250.0300.0350.0400.0HczMczBcz P27-C 0.51.017.6-5.010.015.020.025.030.0HczMczBcz P27-DCt18S: 10.09+/-0.29 12.93+/-1.58 16.49+/-1.28 21.27+/-0.19 Figure 6-2. The transcripts of the p27 candidate gene were abundant in the roots of the affected trees, Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.). The Ct levels for the 18S were displayed below the graphics; columns of the blighted (B) and mildly affected (M) were compared to healthy (H) Carrizo (cz) trees.

PAGE 82

70 It is worthy to note that Poncirus trifoliata and its hybrids are considered to be resistant to CTV (Deng et al., 2001). The overall comparative amounts of the P27 transcripts indicated that CTV is present in feeder roots of Carrizo citrange affected by citrus blight. Further experiments can elucidate whether or not CTV causes or enhances citrus blight, or only grows better in roots of already debilitated plants. Eventual synergistic or antagonistic effect on the real causal agent of blight may also be investigated. Another possibility would be to consider the clone E8-13 as part of another virus, with sequence similarities to CTV. Another closterovirus would probably be the immediate suspect.

PAGE 83

CHAPTER 7 CONCLUSIONS Citrus blight has imposed consistent losses and changes to the citrus industry since its origin, in the late of the XIX century (Swingle and Webber, 1896). The molecular mechanisms involved in the plant responses are still unknown at this point. Evidences for differentially transcribed genes, under different citrus blight incidences, were observed in this study. Three genes, represented by the clones studied in this work, had higher transcriptional level in blighted than in healthy trees. The level of response was dependent of the evaluated group of plants but patterns were observed. This study employed transcriptional assessment of citrus genes in feeder roots of healthy, mildly affected and fully blighted trees. Under these studied conditions, P12 had higher transcriptional level in mildly than in fully blighted trees. The chitinase(s) represented by the P5 candidate sequence and by the clone 38 had higher levels in fully blighted than in mildly or healthy trees. The clones 25 and 109 showed higher levels in fully blighted compared to healthy samples. Research is needed to reveal the function of the P12 gene and the genes represented by the other clones. Cloning of the true P5 gene and the genes represented by the other clones can also be another task for future efforts. It is also possible that more clones from the cDNA array experiment are truly differentially transcribed under non and blighted plants, but more confirmation is needed. In addition, the created subtracted libraries can be a wealthy source of clones for further experiments. The suppressive subtraction method was an effective way to create enriched cDNA libraries. 71

PAGE 84

72 The finding of transcripts of CTV genes in roots of a blight-susceptible-CTV-resistant rootstock immediately raises further etiological questions. Under this context, CTV can intuitively be associated with citrus blight. However, whether or not it has a synergistic or antagonistic effect on blight, it is not known yet. Another potential consequence of this finding is regarded to options for tolerant rootstocks to be used not only for blight but also against citrus tristeza disease. Carrizo citrange may fall in discredit, contributing to the already known need of research on citrus rootstocks. Finally, it is possible to say that the end is certainly not near for blight. Research in roots seems the logical alternative to study the problem. In this way, it is hoped that the knowledge gained during this study can be of some use, helping to understand and control citrus blight.

PAGE 85

APPENDIX BLAST ANALYSIS BASED ON NUCLEOTIDE SEQUENCES OF EACH INDIVIDUAL CLONE Table A-1. Blast analysis based on nucleotide sequences of each individual clone. The clones were obtained from the subtracted libraries (chapter 3) and used in the cDNA array experiment (chapter 4). Three databases were used for the search and the outcomes were highlighted in black (for the All Genebank with 1,367,736 sequences), blue (for the Brazilian citrus CCSM ESTs with 13,610 clusterized sequences) and green (for ESTs of the non-human non-mouse Genebank with 4,893,238 sequences). The clones 7 to 12 were obtained from other libraries of citrus leaf tissues, and all the others were from the subtracted libraries of root tissues, as described in chapter 3. Clone Highest match description/score/e-value e-values Organism Function 5 >gi|806738|gb|U16304.1|CTU16304 Citrus tristeza virus complete genome e-132 CTV p61 6 >gi|1785673|emb|Y08501.1|MIATGENA A.thaliana mitochondrial genome, part A 0 Arabidopsis mitochondrial Contig204 7e-11 Citrus 7 >gi|167366|gb|L08199.1|COTPROXDS Gossypium hirsutum peroxidase mRNA, complete cds 6E-68 Gossypium peroxidase Contig373 e-173 Citrus 1 gb|BQ624415.1|BQ624415 USDA-FP_01506 Ridge pineapple sweet orange e-171 Citrus 8 >gi|6469118|emb|AJ275306.1|CAR275306 Cicer arietinum partial mRNA for mitochondrial phosphate 4E-26 Cicer mitochondrial CSJE01-038D08.g 6e-88 Citrus 1 gb|BQ623911.1|BQ623911 USDA-FP_00991 Ridge pineapple sweet orange 6e-83 Citrus 9 >gi|18401634|ref|NM_112657.1| Arabidopsis thaliana chromosome 3 CHR3v07142002 genomic sequence 6E-43 Arabidopsis unknown 73

PAGE 86

74 Contig1602 e-153 Citrus 1 dbj|C22180.1|C22180 C22180 Miyagawa-wase satsuma mandarin orange 4e-84 Citrus 10 cxContig94 e-146 Citrus 1 gb|BQ623982.1|BQ623982 USDA-FP_01073 Ridge pineapple sweet orange e-145 Citrus 11 CSJE01-038D08.g 2e-94 Citrus 1 gb|BQ623911.1|BQ623911 USDA-FP_00991 Ridge pineapple sweet orange 2e-84 Citrus 12 Contig2919 8e-27 Citrus 1 gb|BQ623455.1|BQ623455 USDA-FP_00546 Ridge pineapple sweet orange 3e-65 Citrus 13 Poor matches unknown 14 Poor matches unknown 15 CSContig70 2e-20 Citrus 16 cxAC01-022F03.g 52 7e-08 7e-08 Citrus 1 dbj|C22172.1|C22172 C22172 Miyagawa-wase satsuma mandarin orange... 1e-13 Citrus 17 Poor matches unknown 18 Poor matches unknown 19 Poor matches unknown 20 Contig1687 7e-64 Citrus 1 gb|BQ625123.1|BQ625123 USDA-FP_02214 Ridge pineapple sweet orange 9e-66 Citrus 21 1 gb|BQ624910.1|BQ624910 USDA-FP_02001 Ridge pineapple sweet orange e-135 Citrus 22 cxContig1377 6e-71 Citrus

PAGE 87

75 23 Poor matches unknown 24 Poor matches unknown 25 Contig416 1e-91 Citrus 26 Poor matches unknown 27 Poor matches unknown 28 LPAC00-013G11.g 2e-06 Citrus 29 Poor matches unknown 30 Poor matches unknown 31 Contig416 e-101 Citrus 32 Poor matches Human unknown 33 Contig2752 1e-10 Citrus 34 >gi|7228328|emb|Y18788.1|MSY18788 Medicago sativa mRNA for putative TFIIIA (or kruppel)-like zinc finger protein 2E-14 Medicago Zn finger cxContig1077 6e-07 Citrus 35 Contig2879 0.0 Citrus 36 Contig2919 9e-65 Citrus 37 Contig2879 3e-53 Citrus 38 Contig1434 3e-95 Citrus 39 Poor matches unknown 40 Poor matches unknown 41 Poor matches unknown 42 >gi|1220143|emb|Z70032.1|CSACHIT2 C.sinensis mRNA for class II acidic chitinase 6E-79 Citrus chitinase Contig1512 1e-81 Citrus

PAGE 88

76 43 >gi|1220143|emb|Z70032.1|CSACHIT2 C.sinensis mRNA for class II acidic chitinase 8E-40 Citrus chitinase Contig1434 1e-43 Citrus 44 Contig2879 2e-62 Citrus 45 FCContig127 5e-56 Citrus 46 Contig2919 2e-34 Citrus 47 Contig2879 5e-63 Citrus 48 Contig2879 6e-50 Citrus 49 Poor matches unknown 50 >gi|599725|emb|Z46824.1|CSLEA5PMB C.sinensis mRNA for Lea5 protein 2E-90 Citrus LEA Contig3802 4e-86 Citrus 51 No significant similarity found unknown 52 Contig2879 e-100 Citrus 53 Contig2919 2e-53 Citrus 54 >gi|5917784|gb|AF184068.1|AF184068 Citrus limon vacuolar membrane ATPase subunit G (LVMA10) mRNA, 2E-28 Citrus ATPase Contig2879 e-107 Citrus 55 Contig308 e-121 Citrus 56 Contig2879 e-130 Citrus 57 >gi|20809305|gb|BC029618.1| Homo sapiens, glyceraldehyde-3-phosphate dehydrogenase, clone e-128 Human G3PDH Contig80 1e-04 Citrus

PAGE 89

77 58 >gi|5815312|gb|AF176034.1|AF176034 Coliphage phiX174 isolate Anc, complete genome e-171 Coliphage unknown 59 cxJE01-111G08.g 0.0 Citrus 60 CXAC02-065D12.g 1e-42 Citrus 61 cxJE01-085A10.g 0.0 Citrus 62 >gi|21206806|gb|AY103728.1| Zea mays PCO142212 mRNA sequence 5E-19 Zea unknown Contig1287 e-162 Citrus 63 >gi|6653735|gb|AF209908.1|AF209908 Prunus dulcis unknown mRNA 3E-07 Prunus unknown Contig16 e-110 Citrus 64 >gi|1732493|gb|U56902.1|CTU56902 Citrus tristeza virus p346, 54-kDa RNA dependent RNA polymerase, p33, p6, p65, p61, p27, 25-kDa coat protein (CPG), p18, p13,p20, and p23 genes, complete cds, p20, and p23 genes, complete cds e-162 CTV p27 65 >gi|3308979|dbj|AB008100.1| Citrus unshiu mRNA for metallothionein-like protein, complete cds e-138 Citrus metallothionein Contig3121 e-142 Citrus 66 Poor matches unknown 67 >gi|806738|gb|U16304.1|CTU16304 Citrus tristeza virus complete genome e-130 CTV p25/p27 68 >gi|806738|gb|U16304.1|CTU16304 Citrus tristeza virus complete genome e-160 CTV hsp90 p61 69 >gi|4239714|emb|Y18420.1|CITV18420 Citrus tristeza virus complete genome, isolate T385 0 CTV p65 70 >gi|806738|gb|U16304.1|CTU16304 Citrus tristeza virus complete genome e-127 CTV p65

PAGE 90

78 75 Poor matches unknown 77 cxContig629 8e-90 Citrus 1 gb|BQ623985.1|BQ623985 USDA-FP_01076 Ridge pineapple sweet orange 3e-96 Citrus 78 Poor matches unknown 79 Poor matches unknown 80 Poor matches unknown 81 Poor matches unknown 82 Poor matches unknown 83 Poor matches unknown 84 Poor matches unknown 85 Poor matches unknown 86 Poor matches unknown 87 Poor matches unknown 88 Poor matches unknown 89 Contig2919 1e-31 Citrus 90 Poor matches unknown 91 Poor matches unknown

PAGE 91

79 92 Poor matches unknown 93 Poor matches unknown 94 Contig2919 2e-46 Citrus 1 gb|BQ623148.1|BQ623148 USDA-FP_00239 Ridge pineapple sweet orange 1e-77 Citrus 95 >gi|6728952|gb|AC020576.2|T12C22 Sequence of BAC T12C22 from Arabidopsis thaliana chromosome 1, complete 1E-08 Arabidopsis unknown 96 Poor matches unknown 97 Poor matches unknown 98 Poor matches unknown 99 >gi|9087297|dbj|AP000397.1|AP000396S2 Beta vulgaris mitochondrial genomic DNA, complete sequence, section 2/2 e-125 Beta unknown 100 >gi|9087297|dbj|AP000397.1|AP000396S2 Beta vulgaris mitochondrial genomic DNA, complete sequence, section 2/2 e-114 Beta mitochondrial 101 Poor matches unknown 102 Poor matches unknown 103 Poor matches unknown 104 Poor matches unknown 105 Poor matches unknown 106 Poor matches unknown 107 Poor matches unknown

PAGE 92

80 108 Poor matches unknown 109 >gi|28619247|gb|CB293790.1|CB293790 UCRCS01_06cg09_g1 Washington Navel orange cold acclimated flavedo & albedo cDNA library Citrus sinensis cDNA clone...8e-44 8e-44 Citrus 110 Poor matches unknown 111 Poor matches unknown 112 Poor matches unknown 113 Poor matches unknown 114 Poor matches unknown 115 1 dbj|C95562.1|C95562 C95562 Citrus unshiu Miyagawa-wase maturatio... 331 2e-88 2e-88 Citrus 117 Poor matches unknown 118 Poor matches unknown 119 Poor matches unknown 120 Poor matches unknown 121 Poor matches unknown 122 Poor matches unknown 123 Poor matches unknown 124 Poor matches unknown

PAGE 93

81 125 Poor matches unknown 131 Poor matches unknown 132 Poor matches unknown 134 Poor matches unknown 135 Poor matches unknown 136 FAAC01-035B05.g 1e-30 Citrus 137 >gi|12249|emb|X03775.1|CHSOATP1 Spinach plastid genes atpI-H-F for ATP synthase CF(O) subunits IV, 6E-86 Spinach ATP synthase 138 Poor matches unknown 139 Poor matches unknown 140 1 gb|BQ623784.1|BQ623784 USDA-FP_00875 Ridge pineapple sweet orang... 220 6e-55 6e-55 Citrus 141 >gi|6693795|gb|AF112970.1|AF112970 Daucus carota strain Imperator STS3A mitochondrial DNA segment 3E-74 Daucus mitochondrial 1 gb|BE460852.1|BE460852 EST412271 tomato breaker fruit, TIGR Lyco... 5e-68 Tomato 142 1-gb|BM371429.2|BM371429 EBma08_SQ002_L04_R maternal, 28 DPA, no t... 9e-11 unknown 143 Poor matches unknown 144 Poor matches unknown 145 Poor matches unknown

PAGE 94

82 146 >gi|18857892|dbj|AB061306.1| Citrus jambhiri mitochondrial ACRS gene for ACR toxin-sensitivity 4E-48 Citrus ACR toxin-sen. Contig4126 3e-16 Citrus 1 gb|BM358396.1|BM358396 GA__Ea0008M10r Gossypium arboreum 7-10 dp... 351 2e-94 2e-94 Gossypium 147 Poor matches unknown 148 >gi|5688942|dbj|AB017426.1| Oryza sativa (japonica cultivar-group) mitochondrial gene for ribosomal protein L5, complete cds 2E-59 Oryza mitochondrial 1 dbj|AV420567.1|AV420567 AV420567 Lotus japonicus young plants (t... 315 7e-84 7e-84 Lotus 149 CXJE02-097F04.g 2e-54 Citrus 150 CXJM02-089E08.g 2e-19 Citrus 151 1 gb|BM371381.2|BM371381 EBma08_SQ002_I05_R maternal, 28 DPA, no t... 86 1e-14 1e-14 152 CXContig831 3e-92 Citrus 153 Poor matches unknown 154 Poor matches unknown 155 1 gb|BI180911.1|BI180911 TY3H09 hepatocellular carcinoma expressio... 90 1e-15 1e-15 156 Poor matches unknown 157 1 gb|BI180911.1|BI180911 TY3H09 hepatocellular carcinoma expressio... 3e-14 158 1 gb|BM376133.1|BM376133 EBma01_SQ002_H07_R maternal, 4 DPA, no tr... 4e-12

PAGE 95

83 159 1 gb|BM371429.2|BM371429 EBma08_SQ002_L04_R maternal, 28 DPA, no t... 7e-08 160 Poor matches unknown 161 1 gb|T44610.1|T44610 7873 Lambda-PRL2 Arabidopsis thaliana cDNA cl... 359 1e-96 1e-96 unknown 162 1 gb|BM371429.2|BM371429 EBma08_SQ002_L04_R maternal, 28 DPA, no t... 84 9e-14 9e-14 163 1 gb|BM371429.2|BM371429 EBma08_SQ002_L04_R maternal, 28 DPA, no t... 72 2e-10 2e-10 164 >gi|12830831|gb|AF320906.1|AF320906 Citrus unshiu metallothionein-like protein (MT45) gene, complete cds 3E-90 Citrus metallothionein Contig2243 2e-92 Citrus 1 gb|BQ624047.1|BQ624047 USDA-FP_01138 Ridge pineapple sweet orang... 1e-86 Citrus 165 1 gb|BQ623549.1|BQ623549 USDA-FP_00640 Ridge pineapple sweet orang... 80 3e-13 3e-13 Citrus unknown 166 1 gb|BQ624621.1|BQ624621 USDA-FP_01712 Ridge pineapple sweet orange 1e-96 Citrus 167 Poor matches unknown 168 1 emb|AL729032.1|AL729032 AL729032 Danio rerio embryonic inner ear... 58 3e-06 3e-06 169 1 gb|BQ414420.1|BQ414420 GA__Ed0086E05r Gossypium arboreum 7-10 dp... 60 2e-07 170 Poor matches unknown

PAGE 96

LIST OF REFERENCES Agrios GN (1997) Introduction to plant pathology. In GN Agrios, ed, Plant Pathology, Ed 4. Academic Press Inc, pp 3-41 Albrigo LG, Syvertsen JP, Young RH (1986) Stress symptoms of citrus trees in successive stages of decline due to blight. J. Am. Soc.Hort. Sci. 111: 465-470 Albrigo LG, Timmer LW, Derrick KS, Tucker DPH, Graham JH (1993) Failure to transmit citrus blight by limb grafts. In International Organization of Citrus Virologists, Vol 12. University of California, New Delhi, pp 127-130 Albrigo LG, Young RH (1980) Phloem zinc accumulation in citrus trees with blight. Hortscience 15: 394-394 Albrigo LG, Young RH (1981) Phloem zinc accumulation in citrus trees affected with blight. Hortscience 16: 158-160 Allen M (2000) A seed is planted. In: The history of Florida citrus. In Florida Grower, Vol mid-August edition, pp 10-13 Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-3402 Banzatto DA, Kronka SN (1992) Experimentao Agrcola. FUNEP-UNESP, Jaboticabal Berger R (1998) A causa e o controle do declnio dos Citros. Laranja 19: 91-105 Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach J, Ansorge W, Ball CA, Causton HC, Gaasterland T, Glenisson P, Holstege FC, Kim IF, Markowitz V, Matese JC, Parkinson H, Robinson A, Sarkans U, Schulze-Kremer S, Stewart J, Taylor R, Vilo J, Vingron M (2001) Minimum information about a microarray experiment (MIAME)-toward standards for microarray data. Nat Genet 29: 365-371 84

PAGE 97

85 Bausher MG (1990) Electrophoretic and immunological evidence of unique proteins in leaves of citrus trees: application to citrus blight detection. Electrophoresis 11: 830-834 Callies T (2000) Reclaiming Florida's water. In Florida Grower, Vol September edition, pp 8-12 Ceccardi TL, Barthe GA, Derrick KS (1998) A novel protein associated with citrus blight has sequence similarities to expansin. Plant Mol Biol 38: 775-783 Chuaqui RF, Bonner RF, Best CJ, Gillespie JW, Flaig MJ, Hewitt SM, Phillips JL, Krizman DB, Tangrea MA, Ahram M, Linehan WM, Knezevic V, Emmert-Buck MR (2002) Post-analysis follow-up and validation of microarray experiments. Nat Genet 32 Suppl: 509-514 Cohen M (1968) Citrus blight and blight-like diseases. Citrus Industry 47: 12-26 Cohen M (1974) Diagnosis of young tree decline, blight and sand hill decline of citrus by measurement of water uptake using gravity injection. Plant Disease Reporter 58: 801-805 Cohen M, Pelosi RR, Brlansky RH (1983) Nature and location of xylem blockage structures in trees with citrus blight. Phytopathology 73: 1125-1130 Colebatch G, Kloska S, Trevaskis B, Freund S, Altmann T, Udvardi MK (2002) Novel aspects of symbiotic nitrogen fixation uncovered by transcript profiling with cDNA arrays. Mol Plant Microbe Interact 15: 411-420 Costa AS, Grant TJ, Moreira S (1954) Behavior of various citrus rootstock-scion combinations following inoculation with mild and severe strains of tristeza virus. In Florida State Horticultural Society, Vol 67, Miami Beach, pp 26-30 Deng Z, Huang S, Ling P, Yu C, Tao Q, Chen C, Wendell MK, Zhang HB, Gmitter FG, Jr. (2001) Fine genetic mapping and BAC contig development for the citrus tristeza virus resistance gene locus in Poncirus trifoliata (Raf.). Mol Genet Genomics 265: 739-747 Derrick KS, Beretta MJ, Barthe GA, Kayim M (2003) Strains of citrus tristeza virus predominately found in roots with implications for citrus blight and citrus sudden death. Phytopathology 93: S20 Derrick KS, Beretta MJ, Barthe GA, Kayim M, Harakava R (2003) Identification of strains of Citrus tristeza virus by subtraction hybridization. Plant Disease 87: 1355-1359

PAGE 98

86 Derrick KS, Lee RF, Brlansky RH, Timmer LW, Hewitt BG, Barthe GA (1990) Proteins associated with citrus blight. Plant Disease 74: 168-170 Derrick KS, Timmer LW (2000) Citrus blight and other diseases of recalcitrant etiology. Annu Rev Phytopathol 38: 181-205 Diatchenko L, Lau YF, Campbell AP, Chenchik A, Moqadam F, Huang B, Lukyanov S, Lukyanov K, Gurskaya N, Sverdlov ED, Siebert PD (1996) Suppression subtractive hybridization: a method for generating differentially regulated or tissue-specific cDNA probes and libraries. Proc Natl Acad Sci U S A 93: 6025-6030 Donson J, Fang Y, Espiritu-Santo G, Xing W, Salazar A, Miyamoto S, Armendarez V, Volkmuth W (2002) Comprehensive gene expression analysis by transcript profiling. Plant Mol Biol 48: 75-97 Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863-14868 Fanta N, Ortega X, Perez LM (2003) The development of Alternaria alternata is prevented by chitinases and beta-1,3-glucanases from citrus limon seedlings. Biol Res 36: 411-420 Fawcett HS, Lee HA (1926) Citrus Diseases and their Control, 1st. Ed. McGraw-Hill New York. [etc.] Fegeros K, Zervas G, Stamouli S, Apostolaki E (1995) Nutritive value of dried citrus pulp and its effect on milk yield and milk composition of lactating ewes. J Dairy Sci 78: 1116-1121 Gaasterland T, Bekiranov S (2000) Making the most of microarray data. Nat Genet 24: 204-206 Hopkins DL (1987) Xylem-limited bacteria cause blight symptoms in citrus. Phytopathology 77: 641-641 Hopkins DL (1988) Production of diagnostic symptoms of blight in citrus inoculated with Xylella fastidiosa. Plant Disease 72: 432-435 Lee RF, Marais LJ, Timmer LW, Graham JH (1984) Syringe injection of water into the trunk a rapid diagnostic-test for citrus blight. Plant Disease 68: 511-513 Lewandowski M (2000) Formulating frozen concentrate. In: The history of Florida Citrus. In Florida Grower, Vol mid-august edition, pp 44-48

PAGE 99

87 Lindbeck AGC, Brlansky RH (1998) Xylem plugging in feeder roots from blight-affected citrus trees. Phytopathology 88: S54 Lindbeck AGC, Brlansky RH (2000) Cytology of fibrous roots from citrus blight-affected trees. Plant Disease 84: 164-167 Mackay IM, Arden KE, Nitsche A (2002) Real-time PCR in virology. Nucleic Acids Res 30: 1292-1305 Marais LJ, Lee RF (1990) Experimental transmission of citrus blight in South Africa. In International organization of citrus virologists. University of California, pp 261-264 Mawassi M, Gafny R, Bar-Joseph M (1993) Nucleotide sequence of the coat protein gene of citrus tristeza virus: comparison of biologically diverse isolates collected in Israel. Virus Genes 7: 265-275 Moreira S (1980) Histria da citricultura no Brasil. In O Rodriguez, Vigas, F, ed, Citricultura Brasileir. Fund. Cargill, Campinas, pp 1-28 Nemec S (1994) Stress-related compounds in blight-diseased citrus xylem fluid associated with Fusarium solani naphthazarin toxins. Phytopathology 84: 870 Nemec S, Bustillo B, Obannon JH, Patterson M (1982) Effects of fungicides and nematicides on citrus blight in Florida. Phytopathology 72: 360-360 Oliveira AC, Vallim MA, Semighini CP, Araujo WL, Goldman GH, Machado MA (2002) Quantification of Xylella fastidiosa from citrus trees by real-time polymerase chain reaction assay. Phytopathology 92: 1048-1054 Paiva LV, DeSouza M, Lopes MA, Paiva E (1997) Identification and isolation of proteins found only in plants with citrus decline. Pesq. Agr. Bras. 32: 559-564 Pompeu JR. J (2001) Rootstocks and scions in the citriculture of the So Paulo state. In World Congress of the International Society of Citrus Nurseryman, 6th., pp 75-82 Rafter JJ (2002) Scientific basis of biomarkers and benefits of functional foods for reduction of disease risk: cancer. The British Journal of Nutrition 88: s219-s224 Rhoads AS (1936) Blight: a non-parasitic disease of citrus trees. University of Florida Agricultural Experiment Station, Gainesville, Fla. Rizhsky L, Liang H, Mittler R (2002) The combined effect of drought stress and heat shock on gene expression in tobacco. Plant Physiol 130: 1143-1151

PAGE 100

88 Rossetti VV, Beretta MJG, Teixeira ARR (1991) Experimental transmission of declinio by approach-root-grafting in So Paulo State, Brazil. In International organization of citrus virologists, Vol 11. University of California, Orlando, pp 250-255 Scheideler M, Schlaich NL, Fellenberg K, Beissbarth T, Hauser NC, Vingron M, Slusarenko AJ, Hoheisel JD (2002) Monitoring the switch from housekeeping to pathogen defense metabolism in Arabidopsis thaliana using cDNA arrays. J Biol Chem 277: 10555-10561 Seki M, Narusaka M, Abe H, Kasuga M, Yamaguchi-Shinozaki K, Carninci P, Hayashizaki Y, Shinozaki K (2001) Monitoring the expression pattern of 1300 Arabidopsis genes under drought and cold stresses by using a full-length cDNA microarray. Plant Cell 13: 61-72 Seki M, Shinozaki K, Ishida J, Nakajima M, Enju A, Sakurai T, Satou M, Akiyama K, Iida K, Oono Y (2003) [Arabidopsis functional genomics using full-length cDNAs]. Tanpakushitsu Kakusan Koso 48: 1890-1898 Stoeckert CJ, Jr., Causton HC, Ball CA (2002) Microarray databases: standards and ontologies. Nat Genet 32 Suppl: 469-473 Swingle WT, Webber HJ (1896) The principal diseases of citrus fruits in Florida. In GP Office, ed. USDA, p 50 Taylor KC, Albrigo LG, Chase CD (1996) Purification of a Zn-binding phloem protein with sequence identity to chitin-binding proteins. Plant Physiol 110: 657-664 Taylor KC, Ellis DR (1995) A zinc-binding protein in citrus with homology to plant chitinases. Hortscience 30: 900 Taylor KC, Ellis DR, Paiva LV (2002) Purification of a zinc binding protein from xylem of Citrus jambhiri. Journal of the Am. Soc. Hort. Sci. 127: 718-723 Timmer LW, Graham JH (1992) Nontransmission of citrus blight by soil. Plant Disease 76: 323-323 Timmer LW, Lee RF, Brlansky RH, Graham JH, L.G. A, Derrick KS, Tucker DPH (1992) The infectious nature of citrus blight. In Florida State Horticultural Society Annual Meeting, Vol 105, Orlando, pp 21-26 Tucker DPH, Lee RF, Timmer LW, Albrigo LG, Brlansky RH (1984) Experimental Transmission of Citrus Blight. Plant Disease 68: 979-980 van den Boom J, Wolter M, Kuick R, Misek DE, Youkilis AS, Wechsler DS, Sommer C, Reifenberger G, Hanash SM (2003) Characterization of gene

PAGE 101

89 expression profiles associated with glioma progression using oligonucleotide-based microarray analysis and real-time reverse transcription-polymerase chain reaction. Am J Pathol 163: 1033-1043 Wutscher HK (1989) Soil Ph and extractable elements under blight-affected and healthy citrus trees on six Florida USA soils. J. Am. Soc. for Hort. Sci. 114: 611-614 Wutscher HK, Hardesty CA (1979) Ammonium, nitrite, and nitrate nitrogen levels in the soil under blight-affected and healthy citrus trees. Co. in Soil Sci. Plant Anal. 10: 1495-1503 Wutscher HK, Schwarz RE, Campiglia HG, Moreira CS, Rossetti V (1980) Blight-like citrus tree declines in south-america and South-Africa. Hortscience 15: 588-590 Xiong LZ, Lee MW, Qi M, Yang YN (2001) Identification of defense-related rice genes by suppression subtractive hybridization and differential screening. Mol. Plant-Mic. Int. 14: 685-692 Yang ZN, Mathews DM, Dodds JA, Mirkov TE (1999) Molecular characterization of an isolate of citrus tristeza virus that causes severe symptoms in sweet orange. Virus Genes 19: 131-142 Young RH, Albrigo LG, Cohen M, Castle WS (1982) Rates of blight incidence in trees on Carrizo citrange and other rootstocks. Proc. of the Fl. State Hort. Soc. 95: 76-78 List of web-pages ABECITRUS (Source: http://www.abecitrus.com.br last accessed April 07, 2004). AMBION (Source: http://www.ambion.com/ last accessed Octpber 10, 2003). APPLIED BIOSYSTEMS RESOURCES (Source: http://www.appliedbiosystems.com/index.cfm last accessed December 10, 2003). BCM TOOLS (Source: http://searchlauncher.bcm.tmc.edu/seq-util/seq-util.html last accessed April, 10, 2003). CLONTECH (Source: http://www.bdbiosciences.com/clontech/ last accessed September 15, 2001).

PAGE 102

90 CREC (Source: http://www.lal.ufl.edu last accessed April 07, 2004). DORAK (Source: http://dorakmt.tripod.com/genetics/realtime.html last accessed on January 10, 2004). EXPASY CHITIN BINDING DOMAIN (source: http://us.expasy.org/cgi-bin/prosite-search-ac?pdoc00620 last accessed June 1, 2004). FAO (Source: http://apps.fao.org last accessed April 07, 2004). FUNDECITRUS (Source: http://www.fundecitrus.com.br last accessed April 07, 2004). IMAGER SYSTEM (Source: http://www.amershambiosciences.com last accessed April 15, 2004). PROMEGA (Source: http://www.promega.com/vectors/ last accessed on September 15, 2001) QIAGEN (Source: http://www1.qiagen.com last accessed July 15, 2001). TAIR DATABANK (source: http://www.arabidopsis.org/ last accessed June 1, 2004). UNIVERSITY OF FLORIDA SEQUENCING CORE (Source: http://www.biotech.ufl.edu/ last accessed April, 15, 2003). USDA (Source: http://www.nass.usda.gov/fl last accessed April 07, 2004). VECSCREEN SEARCH SOFTWARE (Source: http://www.ncbi.nlm.nih.gov/VecScreen/ last accessed April 10, 2002).

PAGE 103

BIOGRAPHICAL SKETCH Eduardo Fermino Carlos was born on February 27, 1967, in Votuporanga, State of So Paulo, Brazil. He was actively exposed to routine farming life and discovered a great interest in agriculture and general science. He graduated in 1990 in agronomic engineering at the University of Londrina (UEL), and started to work as consultant for citrus and other fruit growers. He received a master degree in 1996 at the State University of So Paulo (UNESP) working with molecular techniques in citrus, and a plant breeding specialist degree at the University of Wageningen (IAC) in 1998. He worked as scientific researcher for Fundecitrus from 1995 to 1998 and for Sylvio Moreira Citrus Center in 1998, before receiving a scholarship from CNPq, a Brazilian federal agency, to accomplish doctoral training abroad. This manuscript details his work in the Plant Molecular and Cellular Biology program at the University of Florida, under the supervision of Dr. Gloria Moore and Dr. Kenneth Derrick. 91


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110114_AAAABC INGEST_TIME 2011-01-14T09:21:29Z PACKAGE UFE0006602_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 6391 DFID F20110114_AAASTW ORIGIN DEPOSITOR PATH carlos_e_Page_020thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
9ab60fc93e32853c02d32a389a7786c3
SHA-1
7017f76ac765a9d8fa404f6eab98468b50d56fab
1051975 F20110114_AAASTX carlos_e_Page_020f.jp2
2957b6994a2f766e45bfe47c9953dc00
b3ee31fde3662ac24b95a3cc63f13e1565d791f5
6512 F20110114_AAASVA carlos_e_Page_043thm.jpg
54fc478ed26e48dc2c2bec747add56e4
27a0a27c4a6c929821ba456b8afb4eeed09540bc
9140 F20110114_AAASUL carlos_e_Page_031.QC.jpg
12f1d68c279caaeb31e379f2c5c2875a
56d1ab2e8aeaf1ced3c87797f6ee49b96dff8910
5359 F20110114_AAASTY carlos_e_Page_021thm.jpg
0f4dd9dd499bec4fc1bf95035b112baf
2e751bbc5d30aa1b16cfe3af254798e2d334f17a
19125 F20110114_AAASVB carlos_e_Page_044.QC.jpg
4fed323dd9ee7dc7c45329deb7460832
490174c8d582cc0a7fa9a3e6364ae48e9de74fa6
21847 F20110114_AAASUM carlos_e_Page_032.QC.jpg
ada2dc9db8fcacf4f863384ba0f33fb3
89f8f393a80f9c16dab80e97304c832cee5dd17e
1051922 F20110114_AAASTZ carlos_e_Page_021f.jp2
7296106ebcce6dc7aca305b918b40043
b9abf908155ac0729640fb52f88bf021b70fbc05
5287 F20110114_AAASVC carlos_e_Page_044thm.jpg
7bfe61e8574aa81cad652d01836d4324
117602764c50746da3a00799a6a8e069b31bbca5
5858 F20110114_AAASUN carlos_e_Page_032thm.jpg
039ceb61a0e536b28e86b66d2c325e83
98417c6608c278636b9c98b3917c270483692045
19170 F20110114_AAASVD carlos_e_Page_046.QC.jpg
73835082ad10cabaf9957c02e8a4c607
47b4fa3fc51c3bbbd25a0c5029b6df399331a166
23045 F20110114_AAASUO carlos_e_Page_033.QC.jpg
05abf9f3c55270ffef000ec11911cd73
70ed777fbe114d34734889ad9450f6171c86756f
844137 F20110114_AAASVE carlos_e_Page_046f.jp2
58a4d230aac2e24a939098c873055d76
b8c39c3eee1fa597c16cb9671b191fbab79477ba
4777 F20110114_AAASUP carlos_e_Page_034thm.jpg
5828fcfeea1ba05c16ab5b64c75d75e4
975163a90caf23cb1b5879001796de9d38c91cf0
21204 F20110114_AAASVF carlos_e_Page_047.QC.jpg
4ab07688dd68d029a091cd3aea6352d4
5b957af0e67acdcfe81b0ba88cb738d6cf59f7d6
23075 F20110114_AAASUQ carlos_e_Page_036.QC.jpg
66c752aac7cdda45f2f93fd4e900b0d0
730ce2dd4aeb5ff7d58985694059c6b2b5e2969b
998421 F20110114_AAASVG carlos_e_Page_048f.jp2
e02db0071c9e54fb26e4c2a6b15a6586
7ac96d97c5e6caf7da7a60b32748fe9ebce7eb75
6289 F20110114_AAASUR carlos_e_Page_036thm.jpg
157d51e2d48ad147ece8fce46d6a5108
8459d18446cccfa731b9f5dfb40ea2113e3ca84a
24671 F20110114_AAASVH carlos_e_Page_050.QC.jpg
933c94fa19e2a35f725da529d1097629
6cfd91548d6fbcadfff4a6ff86889beb09a5b5e1
1043408 F20110114_AAASUS carlos_e_Page_036f.jp2
dad5e26dbb0f5c1ea6253563b352963c
aeae6b63bdab152f2c6ba850d1af68339c9f912a
6584 F20110114_AAASVI carlos_e_Page_050thm.jpg
f507dd20792f0c598c8e956259dc3e52
42a9778ce5ab11a40e5374a37b32572ea6a62086
5633 F20110114_AAASUT carlos_e_Page_037thm.jpg
c1bb37bc7f10c807eb8f30c126751a86
ac29b183e32fe4802ce2d31d4417a35b5ab59e02
1051822 F20110114_AAASVJ carlos_e_Page_050f.jp2
93de19f7d50c951374aa01dc176513d2
30a026bde93a9a626c8294908891a44aeaa22f80
4918 F20110114_AAASUU carlos_e_Page_038thm.jpg
c7b62bde8cda8627b7302d84f19d2f31
877391486db7631f1da24c9fd06d5da336b392b0
543009 F20110114_AAASVK carlos_e_Page_051f.jp2
c2bba582327e8d204ed489d0214fcf35
b00f06d1965b372ff7c4e7fb6915afd5c349b5b3
22291 F20110114_AAASUV carlos_e_Page_039.QC.jpg
a78dbcfb7bba97a71b67f875496b4db7
1740b80a10c9e50b68ed168d4d3c8a0051116201
20627 F20110114_AAASVL carlos_e_Page_052.QC.jpg
d378b816e71d7c3012cbd30f4a76ca68
31ac72147bd8e1ea1c89461dd0ec3e638f9d2ae0
6275 F20110114_AAASUW carlos_e_Page_039thm.jpg
2c849e142db3d818406c7ed9decdffbd
83b89beb693646152ba2413321b47f6bd563b57c
983042 F20110114_AAASWA carlos_e_Page_063f.jp2
83a1cff8de3d67014ee5603a3f232372
3c4e0b05ffb0d23b79011934fb06945a190fef0d
26171 F20110114_AAASUX carlos_e_Page_040.QC.jpg
15aee7d3637502a0c48696cde0fcc5af
3a7f0b558c80165bc222cad2e5251dfb3415ed50
4012 F20110114_AAASWB carlos_e_Page_064thm.jpg
2e3970eacbd7363bc19fbe86378602f4
b0150093473f3d2d1a25b80da124def261458e96
5516 F20110114_AAASVM carlos_e_Page_052thm.jpg
c17dbaf557042d3ed725c25cd4c72d58
118022095898fd25e4768a4f60bb4c6b43aca0b2
20304 F20110114_AAASUY carlos_e_Page_041.QC.jpg
765fcbe3a810e6593ec8dfc50c53aa73
6c74cd9efb8230c9de6b6f3865898e12f2edb60b
960589 F20110114_AAASWC carlos_e_Page_064f.jp2
aa421af19fca154f1d696157b5f83ebb
184bce45e31b65b5cbbea4c4c512e1c6ae528a43
5615 F20110114_AAASVN carlos_e_Page_053thm.jpg
b765d7735112565ab532afda6a88b018
e5da8e3b32866f4383f63d3a149e3d1fc5a2941a
21829 F20110114_AAASUZ carlos_e_Page_042.QC.jpg
9ba2bc336638597dc504edea2329f62e
b0997af27dcc21016df650b1ee9b2a297063edb6
1051952 F20110114_AAASWD carlos_e_Page_065f.jp2
c294aa031db1f224cb5cede19ead2724
8d9dbaf5d5202e28b3e9bd554ea478c1ceb59b31
859756 F20110114_AAASVO carlos_e_Page_053f.jp2
dfa833a114041b85546759ad7f95eaf2
4a0c4141cb78f634c61f2168b050a76c3cf9862c
1036130 F20110114_AAASWE carlos_e_Page_066f.jp2
9d3928677a4f0cc865a664cbf4760371
0092a85850c9103b0ca87a45afaa2a353f4c14d8
6006 F20110114_AAASVP carlos_e_Page_055thm.jpg
dbf138e0fb2ac371f8778b4ed651b6de
e29bf37534c177ca078702e972512297a96f560b
1051958 F20110114_AAASWF carlos_e_Page_067f.jp2
0ffbf1885b5fbc9770c39065dd367287
fd50552aaaf534466560ba2a48fd51bc83648a06
23675 F20110114_AAASVQ carlos_e_Page_056.QC.jpg
a2b0221c9be7741e36a6ceb09a1a94f1
02d37b4f5bee1fed8a0b66e9efa28a8808dca359
1051984 F20110114_AAASWG carlos_e_Page_068f.jp2
a03902eb2e89fb185d27605dd90cb806
86303c3e8d453b715538b70c4e7f2794573c941b
1051979 F20110114_AAASVR carlos_e_Page_056f.jp2
cd890a0d8a711fe2360dad1889541af7
6c56b89037cd3d83d9815326383c3a1e39e1ab98
24579 F20110114_AAASWH carlos_e_Page_069.QC.jpg
3f1e800d611d46abfc27f91be6b18242
109f2d40e070069d21a2391cf3556456beacde1a
23432 F20110114_AAASVS carlos_e_Page_057.QC.jpg
8204076c6fe8fd98a2771604ccdb795e
91498867bfc70afabdec790e9421edc3806203cb
1051939 F20110114_AAASWI carlos_e_Page_069f.jp2
e60d314c29209750f273cb273dd4a9eb
615411c7f28a8481a644607e095132c94905bee6
6478 F20110114_AAASVT carlos_e_Page_057thm.jpg
20a377a2ffe6c477998288c412a13980
5a5908abefc8b52d080b5e608f268401e3f9a426
1051960 F20110114_AAASWJ carlos_e_Page_070f.jp2
ca25459ad780ebaf93ce6b105eb289d1
67bab3bba55e82f696b9a42aa8b9308936a7fe95
1051969 F20110114_AAASVU carlos_e_Page_057f.jp2
c0c7d3c99f4ce20cdf72815a34d55989
ecc7f9064c14fd0ea40c3fae8253724aaf267620
23752 F20110114_AAASWK carlos_e_Page_071.QC.jpg
d768e53c5778f57d66fc6e094c0ded28
a2f5d48440409bb0d0a5cf3a02612c3fc58af2d1
19600 F20110114_AAASVV carlos_e_Page_058.QC.jpg
2cbcaa035ed0388aa221c72afecff526
2d505fd7e5d2e49e32ea8ae10136d3ca52be01bf
5762 F20110114_AAASVW carlos_e_Page_059thm.jpg
6e084dd13ab6c66c75ddb4c2eb5ae241
741930a49a76e5c26e9fdd7faec6aaa34c2987b9
6460 F20110114_AAASWL carlos_e_Page_071thm.jpg
504f08d6f8f50485c4f817f7ba54cf18
cae60774da331c0601797707688cb7b8f21d8dbf
18436 F20110114_AAASVX carlos_e_Page_060.QC.jpg
b2875d5b76e543c73267a191b242ac2f
8a8d16f54c9a607bdf600fc0863d3149b9b90b1e
416012 F20110114_AAASXA carlos_e_Page_082f.jp2
cc51169342d08d64ae3152d32472138f
8ac99305db543c72bef7cecd491417613f231e25
1051943 F20110114_AAASWM carlos_e_Page_071f.jp2
13aaaa39acdaa3a6515ae2585524651b
7c05560fb1d3fb28550f2cad74618f98b082873f
6471 F20110114_AAASVY carlos_e_Page_062thm.jpg
ecb859aea22a02ca06ef240022e121b0
9fd7a7d3391c01c7264c2246538191d6efaf0252
5820 F20110114_AAASXB carlos_e_Page_083thm.jpg
914aba092d45b028842fcb224cd133ae
52e2e5dd51bfb330866c844c25735ae52d68565c
20911 F20110114_AAASVZ carlos_e_Page_063.QC.jpg
0383c8319930521d5b8c6a4e30e75698
400f81a87b65d2a7828cf3ee9d8558caf7e7be14
11936 F20110114_AAASXC carlos_e_Page_084.QC.jpg
67ac6768a29cf43ca796fdc8c6bddfb2
43d1e70b2b642eaa376d2cbb9b6270a1d2bff2c7
5400 F20110114_AAASWN carlos_e_Page_072thm.jpg
ba676c8307932ee68b7a98afdbe9b724
9c09afb485f93f17f96b0911f9f0b7964422afde
4974 F20110114_AAASXD carlos_e_Page_085thm.jpg
711b571a214fff96b49ab67ec5c18e80
541f6029425f2db6c788eb1439a5fcb54396b7e4
23661 F20110114_AAASWO carlos_e_Page_073.QC.jpg
3a92e23bbb431b3d49bf060d60f76726
03d2b4be706531db97b783c429ca07b733334df0
930875 F20110114_AAASXE carlos_e_Page_085f.jp2
a0a5030b946def5dfd28ee0e79190c68
130a9e8113d778c23df9350b7018beec8921b34d
5880 F20110114_AAASWP carlos_e_Page_075thm.jpg
d56627399726ab1ca619b96a5d258fb0
042554fa1de889f33d66ce2ae4c50afd09b498a2
14673 F20110114_AAASXF carlos_e_Page_086.QC.jpg
3d4cf02f1753bd9fb92124384e9565a1
c6b6e2f67d9242a28aa191c7372fd7ec1d5e7f94
25609 F20110114_AAASWQ carlos_e_Page_076.QC.jpg
36c066c1e63733308db4ed1cb07b3d4c
46c3bc6ab8435a997b23c8bf28e23dfc0de0e6e0
527807 F20110114_AAASXG carlos_e_Page_086f.jp2
0c39683ffce7be64a6270603831ae21e
afb516090825fe72dbf4d3dc810ad09bc01b614c
19047 F20110114_AAASWR carlos_e_Page_077.QC.jpg
4531895d245ec5eac2d49e553675dd78
b2885f2b18475ea68de8463ddcc41e3ee89a8d84
484520 F20110114_AAASXH carlos_e_Page_087f.jp2
e4f2893e6df898b215bd57df2182a286
941e54eea1a82695d1d15983beeb0e851baa5c46
5280 F20110114_AAASWS carlos_e_Page_077thm.jpg
1e1a871af4082aaa745ef631f6cf3f20
bbbb25399a415594a28dd5f0b695b99e01b88f4a
4134 F20110114_AAASXI carlos_e_Page_088thm.jpg
b9d052941f02523635fcc0f236528a5c
c2176ac465843e7902bccbfa5ded8f1d3849e8d5
859965 F20110114_AAASWT carlos_e_Page_077f.jp2
34d740d8185bda3e0624ab8df4514912
da46bbc23a93e1898871514024340a73a124421f
2018 F20110114_AAASAA carlos_e_Page_089.txt
aec3bc83af303ec3ae7d69cfef118e2f
67399d333e8fc63ef2017e021e10591b3f4a6ccb
17322 F20110114_AAASXJ carlos_e_Page_089.QC.jpg
dfb9b4f530825d73bc4c5f83f29295ab
a285cf00a8b8f316bd6c0a8c28e7600ef5c67c05
21817 F20110114_AAASWU carlos_e_Page_078.QC.jpg
a07cb13a431bfc03d8041ea8f9410118
e523707dca0fe6a968898692e46f707b01ba7953
1781 F20110114_AAASAB carlos_e_Page_094.txt
1864720f74d54c34b8f55c68483590a3
411eb7de13821caf948481793749030bb7bb3ce9
4868 F20110114_AAASXK carlos_e_Page_089thm.jpg
b0435b13c3144708cde8aff82e1de035
b7adbb6557f48d003a0db9757b3b24d71a54edb0
5940 F20110114_AAASWV carlos_e_Page_078thm.jpg
55f55eb2cf167fa299d5fb80c36b01d6
f503c341887b5e1cf54a2fa8675acc50c86395bc
34934 F20110114_AAASAC carlos_e_Page_094.pro
94576f39613a7e831d9ba43281445bda
f0348e24c68bc65de6fa83d1e3228fe97b1b5aa0
12772 F20110114_AAASXL carlos_e_Page_090.QC.jpg
66ca410bc5d39edf2c86c51d31ab697e
c1b8807c2c696878e5728d14ae6c679b8722ad59
6345 F20110114_AAASWW carlos_e_Page_079thm.jpg
0f6fa8c1ecb45177f7068430276f0a31
d613d4900ad5ff310e19c14577808c0b4b735c0a
2893 F20110114_AAASAD carlos_e_Page_030.txt
efd2b3610a4a5d94e1aad4fff6943ce2
2c50d83f21cf1ac2530a9bbc33a5ecab73f06262
3887 F20110114_AAASYA carlos_e_Page_103thm.jpg
9611fc1601acd393254824caae6d536e
a62d4f2d74c132bd5e6a4355801df2311119aa6d
3784 F20110114_AAASXM carlos_e_Page_090thm.jpg
9a08aed6a0186e87fa219370d066137e
7cff81d62b4c9558ef51b435113293f4bca28baa
5734 F20110114_AAASWX carlos_e_Page_080thm.jpg
f06681db3d1174f07c4935c2eb95e0eb
20ee1b4105899b324c0eb940c1e197ee65152d78
8423998 F20110114_AAASAE carlos_e_Page_041.tif
dad028cfdf5d3c5972c9a9051cccf1ed
47d13cb040d0629a1a927b8439cc788200404cd2
98528 F20110114_AAASYB UFE0006602_00001.mets FULL
04f2017b446b8faf509733a49adf3917
1c9fc3d912c4846b04eb63dd2e12d5f6d1d76fb7
4035 F20110114_AAASXN carlos_e_Page_092thm.jpg
21c29ccf56fcaa5c1bb8e59e35e0bccd
d5235fb75f79a39bb6381ac7443d2cd718b8d6f8
921509 F20110114_AAASWY carlos_e_Page_080f.jp2
70d0e04d130424183b9a70326f8837f5
6b9cf6d361d91f1404597f59ad0e4d7170878645
890623 F20110114_AAASAF carlos_e_Page_047f.jp2
677fb6f2df0c5facce70cf0a92c635e2
637c66f3522e375d8b09c56dfb3a8b40c7801675
5767 F20110114_AAASWZ carlos_e_Page_081thm.jpg
547730cac9c8e667b5a6931e13922b72
767e576ca2f25c08f43c5d50873ca2fb7e86ad56
9538 F20110114_AAASAG carlos_e_Page_082.QC.jpg
22f9df5053809f59395f0d4c40efec8f
c386c3b71bb37185c4b4baf3f3238de0454c8bbb
14792 F20110114_AAASXO carlos_e_Page_093.QC.jpg
f4c9e5cee86f3e37bc33909e6ff2f2f6
f33400874c64ce802e693ca20f52e2f61f8ddcda
16973 F20110114_AAASAH carlos_e_Page_045.QC.jpg
0936b96ec49feb1c6685d553709aa576
af081c4e1ad60b7b12b1e64f17113a83eddbc508
4384 F20110114_AAASXP carlos_e_Page_093thm.jpg
a5be64d662fa9d8632d28ffff473a195
64f4a605ab942cfff9f520e6d315db5f25c22e9f
F20110114_AAASAI carlos_e_Page_086.tif
afd41becb37c5c9ed3d5eb7e921df83b
ef8253bb7a7bfc1b126ca60c8ebf816bfc855df2
481054 F20110114_AAASXQ carlos_e_Page_093f.jp2
8e238c50709147f2fa9e949ba7522830
d8ec2e78e7a41b312310780dc386c1f52e987a52
21851 F20110114_AAASAJ carlos_e_Page_080.QC.jpg
3f6162e755a1d6d6478e3e78a879d006
a730f2be3ac1591727cae6f1010b63d3b257d2e9
571843 F20110114_AAASXR carlos_e_Page_094f.jp2
a938448caffa5470e9c74013eca2eeaf
5fe9066df9ebb026f96d3bedda3db65f0aed423c
F20110114_AAASAK carlos_e_Page_030.tif
0285e8da021294c0eab3aac569fe62cc
c4c491567b71db2d6e31e7bc0c935321f01c8dc6
14669 F20110114_AAASXS carlos_e_Page_095.QC.jpg
873cded92ac387fbd00afbe4478c8b6a
e53eae73b614968036c5d1221e96e65cf59621e2
4310 F20110114_AAASAL carlos_e_Page_087thm.jpg
ed51590e527d4288d7c85e1a4bb29a60
2d8d0c855b16e43062591a865104f1a9ee85dfaf
3912 F20110114_AAASXT carlos_e_Page_095thm.jpg
62898a35e97f441ebe9749b1ee45e5e9
72a2cc3eea4a9c3a63087ed89c28b470cc45606c
48683 F20110114_AAASBA carlos_e_Page_062.pro
cd9b316d7456b65c8c15749525126394
30f9061868486771fa3997cb25e2c6fbc8a01cb8
F20110114_AAASAM carlos_e_Page_050.tif
20b954b4ec64102717b068d6fac96c59
07466ee81eadf70ece1c4f71900c82495ebc390c
6539 F20110114_AAASXU carlos_e_Page_098thm.jpg
1d0a3d18acb445a07f23644147b0862d
5d03ea7f9ed6a4d5f20371e2824d0ee61de20519
F20110114_AAASBB carlos_e_Page_044.tif
22d9ca261746ed0a3dc5e86263341b0f
52091dbe0a7e537c0781b0cb72e3337664c31f79
52664 F20110114_AAASAN carlos_e_Page_076.pro
1a7927ddc55bfbb88119fd300124ef19
eca00f91213a6f04ebcf399900336fa03f302d88
1051977 F20110114_AAASXV carlos_e_Page_098f.jp2
8377403b88a8680bcb1a23ebf2b36722
18b5d05552b78467f4aed9da96fda99f1b80dd9a
5781 F20110114_AAASBC carlos_e_Page_004thm.jpg
1cf2406bfa22e99fc1202e7222f5240f
a9460a41cdfde397bc2bdf86f8105b021ebc9a9e
1242 F20110114_AAASAO carlos_e_Page_088.txt
fa7f6860a8dd16f752631ea7c55ea53a
f5703491b94772e1ef5cca5b399b1c66470f5e7c
6776 F20110114_AAASXW carlos_e_Page_099thm.jpg
8b01ca68744916f4f4f94652b288f5ad
f9bf3c2d2c475d4dbe5e81b1f8cc879ea985c12c
F20110114_AAASBD carlos_e_Page_039.tif
71c81a9ecd0710ac5bc81ee96f041dbc
3976e6d3a4a9a67bba0f544d8523160f640ef119
37868 F20110114_AAASAP carlos_e_Page_046.pro
1f1bba54477a82ad3c1207af041b5b07
c761d994ff250340ebc5e54d6b014cc57b96ca87
1051985 F20110114_AAASXX carlos_e_Page_099f.jp2
1b39745c8f6484635404c83f0d7d7560
d6b00b05c5e7cb22e0703b73d9cb1af5d2c9a116
5998 F20110114_AAASBE carlos_e_Page_063thm.jpg
18925f2c9380e28feff0601f1757367e
43356ad644fc1b7d5f7032fb9da82fcfaf48a477
F20110114_AAASAQ carlos_e_Page_091.tif
b663e44e21208e433ba3f80b328fa768
80dbb3cb85d87b79de946cf58d5eb09a9e9bdc0a
15708 F20110114_AAASXY carlos_e_Page_102.QC.jpg
1053abc7f403387d148ee0930266cfbf
69b67d816386978c0cc651da8f7985c411bc52a8
67667 F20110114_AAASBF carlos_e_Page_011.jpg
65799df58d59a06694238ef6e7f8d8ed
2448dee4d128639d5c298fb9d78d3f308bb68248
20524 F20110114_AAASAR carlos_e_Page_013.QC.jpg
4f2e9c36c8d736b210e05992e9965686
300ae6e81714dca43bf47822802e09121c956c00
687846 F20110114_AAASXZ carlos_e_Page_102f.jp2
8a2cb1ac648b72e2609d0ab95d586f14
c0dc44e6bc670b4ec54e7091e9662ae8103877f0
3635 F20110114_AAASBG carlos_e_Page_005thm.jpg
6ff06c920acab88f34f5253833728def
da8e6d818eb0215b43968fe033673b6f820f5345
50778 F20110114_AAASAS carlos_e_Page_059.pro
918f4ba4d97b119f414b251d097ae9ef
56e8eefe97825bce4ac4258e2562f39c0e035b98
F20110114_AAASBH carlos_e_Page_084.tif
96ab64c58f513a82a693dbd32a9747af
45055454c5036b786d80cc705ddbc379931ed820
6757 F20110114_AAASAT carlos_e_Page_076thm.jpg
3a610bd6ebd64a42ee9b36f8273f0478
888bd299c8334f6451c0d40b90531f9f4b608c1b
F20110114_AAASBI carlos_e_Page_064.tif
328307bc9bda206a266f77b3784e6cde
ed81e5fb3c549f95da4d68ebd1669114cfc19bfb
14614 F20110114_AAASAU carlos_e_Page_091.QC.jpg
66709e2e5ad10a222d21a64d8ac63807
79adc127292970fc0403ba7bba7679a9623a7d59
40389 F20110114_AAASBJ carlos_e_Page_053.pro
49c7668321711f87d551ecf726905827
bc1331ff7ededde881ddf62df7b05b7f5a458b3f
21665 F20110114_AAASAV carlos_e_Page_048.QC.jpg
7ec863a77dee1c40e1c3a1ee62b4d39d
f627a42bb2601ae0f0908c4fdcedc21bace265d4
24899 F20110114_AAASBK carlos_e_Page_023.QC.jpg
6367a27b6b9b7b801fbd8008445faceb
d230fb2ae8ecd39242e173b7e9d8fd912c14f64d
1829 F20110114_AAARWE carlos_e_Page_025.txt
752a37798ef31da13e991b744d4c3e99
1cf2c03dd74c2edc789d937cb80c42543b1e026b
17699 F20110114_AAASAW carlos_e_Page_038.QC.jpg
13684692e8002dca6ef5d373f308254d
63db57488570c024e8898d546feb038afd6862bf
75683 F20110114_AAASBL carlos_e_Page_057.jpg
b526d91da1843efe30645958b63154d0
9153de535e732d6aa40a50ab2865e2cf8675d5b3
F20110114_AAARWF carlos_e_Page_063.tif
06a2b7f60080f72bb0c53363d9ddbeb4
8c7d49edf33e0eefc67f1c9f272866a3521d10fc
74956 F20110114_AAASCA carlos_e_Page_014.jpg
35e93152314fa6fe2da29f8c258ab453
1b16b9e81095ceb70b9c0684d32f1c923e0a8341
1164 F20110114_AAASBM carlos_e_Page_102.txt
5c57fba53dae16d95815e3b565a81d29
57a500fe69cf83605dc5aeec3ef0d0e973204aaa
46269 F20110114_AAARWG carlos_e_Page_017.pro
1fae9a90bd9d5917495349e535c8987a
6ed0a292607e87128c6db3b0ea661ff95384a6e3
4126 F20110114_AAASAX carlos_e_Page_016thm.jpg
74f33f3c81def2cf3f4985fb01ae133a
443024cc595e99c23cb6f89254df923c14bb00d6
19659 F20110114_AAASCB carlos_e_Page_027.QC.jpg
5a65ca90c4557e4eba7a035a6d348f49
64beaa3afcb2124f68da4749b46fa0f402b5f91c
5885 F20110114_AAASBN carlos_e_Page_035thm.jpg
e74f4bff5bf8d296666fd13bcebf47d5
1c9771866e669d7e40b8ae8453b9911a01731678
22225 F20110114_AAARWH carlos_e_Page_075.QC.jpg
ccb27e5112767dbf5b6d6af56fde332d
6ca9c2a4b4402a30e5eb938e5db630b6d64f58cd
776432 F20110114_AAASAY carlos_e_Page_006f.jp2
fc797f78be2ddf7d82f565dbb5006c42
9cac390f4ee37a98631993c9f0f9957ace89dd25
90385 F20110114_AAASCC carlos_e_Page_097.jpg
135044b23730a883de7bd2d9d94d0c50
67e83fdabc358db45deef6204ce89658d66bf846
34935 F20110114_AAASBO carlos_e_Page_060.pro
d83e539d9a33167d9c646274f02950be
22538c85d47376986dd09135eeeabec77bd7bd10
39427 F20110114_AAARWI carlos_e_Page_058.pro
0574b3e3268751b1b4142a00fdf69fe3
8d7100ad89487b89dcc48c5ad56b68aeb384e098
996410 F20110114_AAASAZ carlos_e_Page_096f.jp2
1f38dc3a7cb27fa80aea9ceb7bf0bedc
943d909b98365e1cf6a76aa58f84b8f32b3878a1
26598 F20110114_AAASCD carlos_e_Page_103.pro
5047cb2c4473d05e3b30417b5d36b6cd
7cb4ea6c38be3695751b5d054d45d67f21190ae4
888038 F20110114_AAASBP carlos_e_Page_044f.jp2
905ec0c363c2dbba9ec3665148467dee
0ffb4fb1c38943adcc924af5f789380714c445c4
3779 F20110114_AAARWJ carlos_e_Page_006thm.jpg
5ab4d228a84e98589a8e66cb2630d704
79fd94d5bc56bed5895016ff0ae9a3b5fb05d22b
755 F20110114_AAASCE carlos_e_Page_082.txt
60d99915ab7d4a8e75fa97739e3130b8
3e20e08f10221e5416380bd451e2245046729258
50659 F20110114_AAASBQ carlos_e_Page_069.pro
0f1a08ba638b29d6c8b28680890ff7e2
daf6aa1ba220fa18b471d2a7bc5a84d860496e89
74828 F20110114_AAARWK carlos_e_Page_078.jpg
dc40a825fa83ef06136daa267709c9fa
f3f13d6bcefda9a5855150267fc8274b6f679e55
24120 F20110114_AAASCF carlos_e_Page_043.QC.jpg
2159981ccc00dbfe7fc73febf488ff13
2652bb64ae2ea881b30ecfe70d51959d46f6a7e1
1294 F20110114_AAASBR carlos_e_Page_012.txt
4c4a716ade709cc0ce3fa6f895f62414
08c385e001d06f157055e4d5323bd5881023fa4e
6577 F20110114_AAARWL carlos_e_Page_023thm.jpg
462fce0a66b6ef07aaa4d338d6383e1c
bd4035b0ca6ccabea2a0f86cbb22194b07ea4a2d
51704 F20110114_AAASCG carlos_e_Page_034.jpg
3b5d41c7f55fe700ae253524c22faac8
cc5708ac580713854f7ef9d0a5cd8bfca1368e2f
1051982 F20110114_AAARXA carlos_e_Page_039f.jp2
2eb88b49580306bc308ef891dec7e29d
2254d4ce397ccd6632b21329c7440b8b86e8006d
F20110114_AAASBS carlos_e_Page_029.tif
6834f79ed7aa4bbedc96dc1f208546a5
e89f09a584be348d17333df4f8b8a65d78e62ff5
F20110114_AAARWM carlos_e_Page_042.tif
f317aef10980dd614537e6556f3c0fbd
dca1d48c4a9db0f895ea361a16d7fbfbe1b49c42
983872 F20110114_AAASCH carlos_e_Page_052f.jp2
aa2c71e6fabea2acecfd6caaf65a1132
1b391dd34efe401a7393013c164d850358e3c582
F20110114_AAARXB carlos_e_Page_049.tif
82a88369d9d87890e37da6c1202b5998
f7df69c4afd6a01801c8a8db17b4a64cc6ba9edf
18096 F20110114_AAASBT carlos_e_Page_085.QC.jpg
941e80f6d8220efe2030ab75f53a5983
d7bca02c9b682dc58bbb76b6b27b0203e43a1296
925821 F20110114_AAARWN carlos_e_Page_041f.jp2
34c8ce58420f7918e8e3f952276fe455
4e8cbd33dbe5918b3384d9610e9cf1e5722a1677
57043 F20110114_AAASCI carlos_e_Page_006.jpg
f3c05199a53945b4017b38d49a1edf0f
785012b5ff3a1d037be27f56b2ff454f858e355f
1952 F20110114_AAARXC carlos_e_Page_020.txt
4c785a7f658669c0987e83c3f5fd09b0
6f02ce9364523aec561cf6f7767cc29b92012c2c
6224 F20110114_AAASBU carlos_e_Page_033thm.jpg
2190d1193014c90d44e6f3035f84aedf
fb386dd744c3eb653fd1718f9cdae81ec2d567a1
1925 F20110114_AAASCJ carlos_e_Page_010thm.jpg
0976adecdca6f9e3c469e084efcc9e2b
5c2feb1d5913bb24ed9235fec30cdfa8ddb97a88
2434 F20110114_AAARXD carlos_e_Page_100.txt
9d2f804683be773acf17777dd9ee7fc6
95931a71b041a5f35b889471ea8afb8899fedf1e
1104 F20110114_AAASBV carlos_e_Page_103.txt
33b87b5b7a6bf0c60cf2efebba01ed32
b4821e7c7d23e451ddacb2dfebd4f53cf82f0af8
1025276 F20110114_AAARWO carlos_e_Page_018f.jp2
8a0a77b731c4f2fb44af363524ba524f
148d3ac5a2d107bfb64feccbdbd6e7d542d94655
1051944 F20110114_AAASCK carlos_e_Page_043f.jp2
f89294eda798b879e7fed84476e8d2ca
f84085e07b4d0b871ff39c8bb028d337d0840b0f
46043 F20110114_AAARXE carlos_e_Page_085.pro
f64220c2b0fa768f7687ffb232a3019e
0278908cc4353664f784e2a0a8789601fb502024
805757 F20110114_AAASBW carlos_e_Page_045f.jp2
8ac34a148b00079eedcb02223fa3a649
50c4832f60234f7a7d4819eecd329ab29de6c46a
693866 F20110114_AAARWP carlos_e_Page_089f.jp2
d404a3d93e4c275e24654009998b8498
f19720281d4f3d4a65ef6c4c970892242b13dc0f
1856 F20110114_AAASCL carlos_e_Page_083.txt
e85076d24837ac64338fbf400f7fb15a
42f1cabd60bc588175fb001518e83174be826572
F20110114_AAARXF carlos_e_Page_033.tif
ee178d6341a1f001779e3cdc1bcbd851
a2b725043d02ab1c8bae580d51e81e7a7a24ed71
6179 F20110114_AAASBX carlos_e_Page_065thm.jpg
d45a1b4a8713b16bdd2fd64d3a90b5ee
184a516605396d678236aad96f2bc83c20c730f5
722309 F20110114_AAARWQ carlos_e_Page_012f.jp2
9967973c774b73bc1756b974ca76e6be
8c2995c016af66a5e3b71f1e6319b3f788234a04
39766 F20110114_AAASCM carlos_e_Page_011.pro
91a0c129b2de970b9838a71800c17f0d
66b14fe314c13534710a162835ae2e438d9927ba
1534 F20110114_AAARXG carlos_e_Page_060.txt
9a891e75cf69a224cedc696c080dff29
97cbfb110f8f034c828192309c516e6680e6b06f
21316 F20110114_AAARWR carlos_e_Page_010.jpg
3f84d792eee718f20f654dfb15f5c003
b1299098f483614a96dc978c5fd45fe23fa3f310
514393 F20110114_AAASDA carlos_e_Page_084f.jp2
480f31583da1e0b44cced6cb7e9fb368
b6b8461b82dc25738b2cd57462553594102c8317
F20110114_AAASCN carlos_e_Page_082.tif
588a407a2a465bd01cd4b4b9559fe4f3
725e02f8ec1661bd2ec82f5197d445c572178ff8
F20110114_AAARXH carlos_e_Page_075.tif
5de21628ce5d6ec1187c046897f63abe
3eb3179c081e795791a6ef7eacb26b335e453166
551971 F20110114_AAASBY carlos_e_Page_005f.jp2
a75f9fa53a44fbe60286d3fe6cb11982
1d0bf10ffca3cccdcf790e14fa87a5025b18476a
F20110114_AAARWS carlos_e_Page_048.tif
96f4293b83889f79c054f4d57309c00c
c3ad730291d9126d6276cb974a7cb9a540903552
52998 F20110114_AAASDB carlos_e_Page_081.pro
f26e93f52a985a5c9e7585c099fffda4
96e1f06b6d8f670a2b5af2f52d51c4aa5eb2f6d3
16264 F20110114_AAASCO carlos_e_Page_012.QC.jpg
bcbe108155acbab187d903b6ec9e8ee6
94ad307351d6063e8cc9cda820d179e125da5d95
2479 F20110114_AAARXI carlos_e_Page_081.txt
a148fefe70738f6a8d35a2a80a3d1ef7
631eee0f9bf39aa6a32bb062ff9982ce1c7fcda1
21341 F20110114_AAASBZ carlos_e_Page_035.QC.jpg
3e0121c6d99439c8c5f0c6012a42c714
026e948607663c691400e34d14cbb5d07b971630
F20110114_AAARWT carlos_e_Page_014.tif
8889a12ab50c01515167b7527fa755c6
fdd25d7a8c255aea327226ee2ad4bbf99ede0dff
6055 F20110114_AAASDC carlos_e_Page_029thm.jpg
10920bf63ba07db3f70f9e54d7161bc6
6c8fc4bd035ae53267d2f0179eabae489eb918a5
2590 F20110114_AAASCP carlos_e_Page_002.QC.jpg
d3479f3f5847b0c695e0582a91dee20f
e7c4f8afda56953278979db9b862465c5af784cf
3567 F20110114_AAARXJ carlos_e_Page_054thm.jpg
0c502323c9108d7c6f35d1234239ad4d
695d1e4e97fad506ed3bc81c3001ccd8f183957b
1935 F20110114_AAARWU carlos_e_Page_067.txt
c50706ad90d07816460a289b625711c7
1341f910b22d510afdb3b4af5e9b67d4aa0d5fb7
57155 F20110114_AAASDD carlos_e_Page_060.jpg
a3042fcec3119ec0ca3abae46512c796
310ab1c081d54aceb1aac14dcf9573a2dd59c492
4019 F20110114_AAASCQ carlos_e_Page_094thm.jpg
4637908b3ffae6b19b5a23b6af47d03f
442474c439b0a749b2181a5b2ffb08c64112ba27
45139 F20110114_AAARXK carlos_e_Page_095.jpg
e9ce142f57c4603e73631fb4debfdd82
34a46280881c9b9a5412a48a68b0a068f74dd90c
F20110114_AAARWV carlos_e_Page_067.tif
e6a0eca2a6a0bb5070383be3fbc4c1d0
d3951b1bfb56c37a1f47f70a52bcadd48d57842b
1702 F20110114_AAASDE carlos_e_Page_041.txt
bda6983f6baeaed9f221473613a59739
59e647621acbfa01324ace636dc36bb0069022b1
17228 F20110114_AAASCR carlos_e_Page_034.QC.jpg
1edc4e8a6cc5ce47cd5269f3de03c220
bad30e5dafa0e25b53c89d68466cc17bbf7ed720
57214 F20110114_AAARXL carlos_e_Page_097.pro
89626b0688ea27abf4a12cc4fb5fcf18
7eb70242dff6c45eae579731180a639b132a1c6c
68890 F20110114_AAARWW carlos_e_Page_083.jpg
113573f3363e61b04da92c1ccc931a2c
636fc2d57a2e463811eda5ec39318e573ccaf26f
5773 F20110114_AAASDF carlos_e_Page_041thm.jpg
dad326d9366b65c1b57153c53dfc2200
21d6260c27e7508c4638ad88619c2652f645ef22
799283 F20110114_AAASCS carlos_e_Page_060f.jp2
c34c01b02d8b77d456fba936756802c2
1a11e75b2b53485b4d9fae1b021c594515a4a94f
22668 F20110114_AAARXM carlos_e_Page_017.QC.jpg
fdbbfc8cb70f4834afc2536cf7ff5213
f9548bd001538c52dcca7469f74f9ecca876c0a8
76271 F20110114_AAARWX carlos_e_Page_071.jpg
81727ca452f449e8652c5cb6f9a8573a
67823da453dd74309df1e4278db29a106ab22bf3
23336 F20110114_AAASDG carlos_e_Page_030.QC.jpg
254e7bff4584762773cbe60942e21bb5
ba7d6c64e4e5dbe8c162264c0ac02faa49c07082
15011 F20110114_AAARYA carlos_e_Page_016.QC.jpg
fc5c8e158cbdad5b33f1505ff363fcfe
edecf71edec906a6997d61a7d6413ccde03ca5db
F20110114_AAASCT carlos_e_Page_057.tif
ff192d5c4e9fac884a37e0197b3b5d50
99481a3f0b73df4a348d4d2ace25265576001e6b
6291 F20110114_AAARXN carlos_e_Page_019thm.jpg
4a2cc3dbbe06e636e4a51165b37c6cde
4410a9226a5ae0cfab39398c690f1f1a193190ab
349080 F20110114_AAARWY carlos_e_Page_074f.jp2
2f0983a053c0b609da6b3ab3b8359bbc
a6959a8e9420b0422f678cbd17494db5c1d21417
22987 F20110114_AAASDH carlos_e_Page_008.pro
97630a849e5cf19f111b17ef403dc535
eeeb5296d2e07ce578608ec6e552e08c4fdc51db
1051859 F20110114_AAARYB carlos_e_Page_049f.jp2
66d04f98ee09a73848a3a15c2d619239
6d24c09a59a06e54a1dbe12057557c571fb3f0d8
6256 F20110114_AAASCU carlos_e_Page_067thm.jpg
d861b426d7d88b564df311a689591a61
80f06d4554d75cf4ac9a847454aaff7ccde2d0d6
1051964 F20110114_AAARXO carlos_e_Page_097f.jp2
fe3bf4808f7ffb420bc59b722506a505
024b84a484ece82f0cea9fce554dc8d8f3a5b6d4
1755 F20110114_AAARWZ carlos_e_Page_063.txt
5e3fd9aa9906d3486282895ead89a85c
0f1969b2c7e8051ccd7ed53fd64ab9382a34909d
21435 F20110114_AAASDI carlos_e_Page_081.QC.jpg
bc7651872a879da01f84d2368bf5744c
137d8f16cd44f4cf307a9970b0e4bc7f7255e0dd
F20110114_AAARYC carlos_e_Page_021.tif
e8180756807e7cd6f1f83a15cc7ad57e
e23c552e947acb6fc847e052727cd1a48f978e06
F20110114_AAASCV carlos_e_Page_035.tif
865cfd2a6e8bca2aa28ad6f1dc53d133
73a07288ef1efcc1e0daec3f246bb0842efe4bcf
21309 F20110114_AAASDJ carlos_e_Page_049.QC.jpg
ab170352bca3cc8949daca3cf53f69df
709b97eda136dc3cc9ebbb9d75bbfce0d9b61145
23302 F20110114_AAARYD carlos_e_Page_061.QC.jpg
607c672e6593e8cbc5527392ccd3e3d0
9398d628743ccf2f5b8efd0fb62df1089880902d
2380 F20110114_AAARXP carlos_e_Page_099.txt
8bc7f4bea457979fd4a2d24561c66b1c
40f7217e3407149dbf9ebdb758c38e9ef0570975
66969 F20110114_AAASDK carlos_e_Page_027.jpg
bc09e7ba6ed9f71036000bf2fa814aee
cc9fcb8a686f7161890dec80fa86341f77233920
32671 F20110114_AAARYE carlos_e_Page_012.pro
8fddcca58676d298eb0ed0d680a44b18
275aa49a790ba43261c52923de26a3989d82bc39
5616 F20110114_AAASCW carlos_e_Page_066thm.jpg
d541db0a153ff1e2ac8e51e875309f9a
21e8703c2c702cf9006e2ddc1ab09e1276aea74d
2397 F20110114_AAARXQ carlos_e_Page_007.txt
e97a3134cff164abc75ee5563a910792
51ea4f78a454864f9844eafa0ea4f6b6b2a9b15d
2209 F20110114_AAASDL carlos_e_Page_049.txt
83c72f8e8e5fe881b8632104eef97441
2cd7de35a85b8b86f26a60da22e5192c6be4c631
85467 F20110114_AAARYF carlos_e_Page_050.jpg
7ecc29d5aa5d8ab025b0de37b9f38ba8
277aa308e259236618755c856cb40a29af361dca
543804 F20110114_AAASCX carlos_e_Page_054f.jp2
0dc3e2fc37d00dc7d5e7ed35957210ad
a491291f396b1a9c249624ab7177f698cc0d1867
1954 F20110114_AAARXR carlos_e_Page_101.txt
ba1a8b1bc47ff5bd431d283830fac248
04ac3f61aa59edd9b3e6237cde36970c80450b7e
987419 F20110114_AAASEA carlos_e_Page_083f.jp2
edb7af4f4a95269d35b6fcec8d6c53e0
9b73be5eba7d8df08670d85052eed657ec8aa7d4
F20110114_AAASDM carlos_e_Page_096.tif
f739ba237c51cc0c20f281cd8666d0a4
23cb069e73ae4b6dc7846533e717dd1289f5b667
9167 F20110114_AAARYG carlos_e_Page_074.QC.jpg
3396997e788d6eb8b3ceee172feb7d24
c0b0226341519d3e92cf869616574815e1b4e830
44935 F20110114_AAASCY carlos_e_Page_048.pro
35bfc4f8ebbe125161e01ba44801442d
7c239dd5e3ef819ae684e80f1f213d99731992cb
15912 F20110114_AAARXS carlos_e_Page_087.QC.jpg
2db1dbfc17b96842dcac1452607e8c2d
e97516bdcf90d9c8150a209e2786c91d4c8b3c7a
44937 F20110114_AAASEB carlos_e_Page_083.pro
18c6f4333c8fa8bf440da20463bb6136
f39727c64b5b51f8be4088cd1e7d45468a9667a2
20035 F20110114_AAASDN carlos_e_Page_096.QC.jpg
f1d1dfa660526a582addd548efd533f9
b9c7ef9f49aa64204e756a0a9a147aaf4a5698e0
51054 F20110114_AAARYH carlos_e_Page_071.pro
1375eca72e009edb7aff4e48e4c0a40f
ea4f434729b99506bca816c08cf8718375b97c47
300099 F20110114_AAARXT carlos_e_Page_001f.jp2
638c5ce4763bae33586d2b6f8533a424
9d5776ecba6390154188bf3aafd8b5dfa28cf61b
14643 F20110114_AAASEC carlos_e_Page_094.QC.jpg
ef08f9e9f5539bf895c1837ddc85bdca
a591c93fb259f6e1c1cd4c041427958ea650765d
43996 F20110114_AAASDO carlos_e_Page_032.pro
57ccc13104d905e75548d0a20d308bf5
2b14de3552827c5fa68b93384b728d8da4d3f1ca
23705 F20110114_AAARYI carlos_e_Page_068.QC.jpg
0879ed6ab7cc79a843bec312ea0ddf50
b7b895f8e3a9034edf86a300147fa6ef09a6654a
43404 F20110114_AAASCZ carlos_e_Page_040.pro
de028582c81c81ba4be7affc18bd702b
c739b5b562422425c9321ff8ece7568a2affc55f
F20110114_AAARXU carlos_e_Page_018.tif
05b805853923f1d7e93ed2c82fd9f87a
1aa046c1975f9e7b48c4b0c7a921acac71bea8a9
F20110114_AAASED carlos_e_Page_065.tif
e3918d049666bafe2462d7eb835fe6b8
2494d6cf248f4c378c7a26974a8117701e71772e
F20110114_AAASDP carlos_e_Page_006.tif
c9d25ca4b13a561f7b93b1e8e4efecae
3b73aa7969cc56e7fd0fe5c4c8c78b8e05bb88e3
14143 F20110114_AAARYJ carlos_e_Page_088.QC.jpg
c4f3023e7f0e223dca1056d80fe6372a
b3ed59722b28e003d53d49a2511e541e19d282df
F20110114_AAARXV carlos_e_Page_103.tif
dad1ba5f9457a5f3c0c0b4786be024fa
ec045b3f18adbe94120ebacbd4d73883d8218cea
4425 F20110114_AAASEE carlos_e_Page_102thm.jpg
ed5ae174b959bc2c09d76dd322c9a326
fcf318b294d910eff00505876737faacf47fb826
1179 F20110114_AAASDQ carlos_e_Page_016.txt
7d8c5a402cf3522cc7e98420d1241144
ca249538827821d162f5379a82a7a5f965ae2c34
29525 F20110114_AAARYK carlos_e_Page_016.pro
90ddf0156290313e68297d1fef2a6b30
4fee48c85a650d66eb6e434bb89988b386e38d17
33681 F20110114_AAARXW carlos_e_Page_095.pro
d212b2cb74e7ca6d2c380a3d8e24ad88
fe514d25d312576171749627d7936bff1473d188
2273 F20110114_AAASEF carlos_e_Page_001thm.jpg
b6785ed374bde8eed63c5850d3710182
e272f1f910add7c7e617e7f347c0473f13a740fe
1301 F20110114_AAASDR carlos_e_Page_015.txt
8302fe426ce8306bcb7714505934d174
9b4d3fe651b796a5224336f7d4d9f92fca6fe9cc
F20110114_AAARYL carlos_e_Page_092.tif
ef749a6499f10bd0da57d557482eb0c8
5947b518fc45033703f5ac9d7b2e3fb6fcfb7ea8
672 F20110114_AAARXX carlos_e_Page_074.txt
2dc832c5813875a75d2860eb62b56002
c2ff0129185005e71e71da90fc92b3cb1525a5a2
24058 F20110114_AAASEG carlos_e_Page_098.QC.jpg
605c9d1da3e85e9d400163863c38f4bf
6b0afe0129564be80529b0d5d9f4dd83b026d893
74171 F20110114_AAARZA carlos_e_Page_101.jpg
5f8faa2a4266a717d8aaed862a45c2b3
9b34bfa73dba20faa2f3b4af38efa83b8b4d06d5
2876 F20110114_AAASDS carlos_e_Page_082thm.jpg
368db7d151fad113751a3155d4acf82b
327ad569770502e84eabf74e02a9f778277346e0
21857 F20110114_AAARYM carlos_e_Page_083.QC.jpg
ab64dabcebd7e652698b87a3fa9caf2b
a3e6c9580a60c5ad403de35534a4560b3dd4f7b2
F20110114_AAARXY carlos_e_Page_045.tif
83d9e02ca62c55ba6c218f27c204332f
861d594af7de9f75b73ea3029e37a675f482fcde
16334 F20110114_AAASEH carlos_e_Page_010.pro
37f0f5a2f87b37fd186a649bf45791b1
5a09506d01aa410b7b7aea29798f7d045fe8a806
41886 F20110114_AAARZB carlos_e_Page_037.pro
d087938e52a0492a2228820f8aaee4e9
9f6b81f0eae12361fd90d08aa83e168d5349c544
8985 F20110114_AAASDT carlos_e_Page_019.pro
fc2faf40ee0d69720b379a3b9ecacbe9
3af8a332b631603b105a1c0082d4d9cb4fe293a2
1813 F20110114_AAARYN carlos_e_Page_096.txt
025991c4ddb6047849aacc472f56ff8b
e1fc426a0d4af8a1287a1211838ac54034346701
1827 F20110114_AAARXZ carlos_e_Page_048.txt
15d4562c43494637d6de0f545f0439bc
c12b1e34a2be41bb0cdbed313f43be16274cf85a
24035 F20110114_AAASEI carlos_e_Page_005.pro
d57a9b9957ac10e8e367e76a97fa8c92
dbb1d68a576ed1cf698525418120dc0322ae599f
F20110114_AAARZC carlos_e_Page_020.tif
b59916d3750dd3798fef59f979aa7422
15536ef949277bf674de4912b01cbe7c949fe8ce
843 F20110114_AAASDU carlos_e_Page_021.txt
ef27df8c5a190b9cbc24c5b4737f257f
8c47b49a3150ab28bbd165cffdfeb5ff2468c6b3
24987 F20110114_AAARYO carlos_e_Page_070.QC.jpg
5f89e6a8b3c98c8da713a7eecba7005b
5617ed1cc9c8a84939b37ee81a02da8764065fa8
18721 F20110114_AAASEJ carlos_e_Page_003.jpg
d269eb6d85ad4d373636e451478ed082
e33b75c9280fa0a19cf12da0a356078f4fae910f
6268 F20110114_AAARZD carlos_e_Page_061thm.jpg
c3d7bc01f7ee2dd3851754565243514f
7d9d618bc46b274d9a5ba3523eddf7ce00785332
2220 F20110114_AAASDV carlos_e_Page_098.txt
8e16fbf24ace2f56c8a5009051e9dd89
3588d96c4ab15f8c9354aed3dc1eef6dc200269a
1795 F20110114_AAARYP carlos_e_Page_032.txt
f8a909ac21cae90382f382a461779f2a
1851abf78778d88b4e66daaae02855c1f0a590de
1826 F20110114_AAASEK carlos_e_Page_058.txt
1d0860e5f8374c05d8fd150717add572
ceb438d87ad98e6c8d66eefc7a535af1f4072a97
5993 F20110114_AAARZE carlos_e_Page_047thm.jpg
5223184976c0a2939485053f2664ec07
78df7c3a26dea65e46fe33adcf7097211b370b02
5596 F20110114_AAASDW carlos_e_Page_030thm.jpg
4b5d426c4c8bcc8ffaab56a8bbd9259b
29dafb5a2d618f7ffd5674383e542ebf1f97a658
51436 F20110114_AAASEL carlos_e_Page_012.jpg
140d844b9a7affc93733d4bd13d35997
79553548d3d9cad14f4fbb22194ef807b0145317
509681 F20110114_AAARZF carlos_e_Page_088f.jp2
45442a4ed3356a785f5f8174a6dff2c4
670deb70e2a4e4fc6fe214eee638af4b2fc7cd95
91907 F20110114_AAASDX carlos_e_Page_040.jpg
99e49652943450b3f40970717cd21293
91b11308a23392bd12dc5137b1da9449373079dc
22318 F20110114_AAARYQ carlos_e_Page_055.QC.jpg
cf1ea46e59460853953b34c55b167882
cf6456745159a6487132a5a1cf73b7975d1de58c
6315 F20110114_AAASEM carlos_e_Page_040thm.jpg
114a99dbd837d25d2566b6c3619aad4d
de945ececa85d4b822829b47b66562a721f9e9ec
72692 F20110114_AAARZG carlos_e_Page_026.jpg
73818c974ac1fbc47f90f6774fb53866
68aa66539efa03973aec6969d2e93a95fb9ac176
4978 F20110114_AAASDY carlos_e_Page_045thm.jpg
4008f81fd0358142be77158d6877d155
cd077826862b10fe3c77839325fa1f7e506c1e6b
22105 F20110114_AAARYR carlos_e_Page_059.QC.jpg
f6e8a63098ddcd2ece3000af58af55f8
631ce1ca2530830f94311130ec02e07dad5bbefd
F20110114_AAASFA carlos_e_Page_034.tif
7b01697dbd04afc13d6923d200452fae
a6c768f5adee4ea0aa5b0d394395c84d03a00ccc
74546 F20110114_AAASEN carlos_e_Page_061.jpg
4f72603e142b1f7669dbc8f034927b39
11c0d6ceab59faa87e0d3cd681005496b5c26523
77309 F20110114_AAARZH carlos_e_Page_069.jpg
4ed6c78c747594013d2502ca3a8959f5
8fa4da2469cccaa2db7ef7f6a8b76d94926332ca
F20110114_AAASDZ carlos_e_Page_058.tif
e018f86cfcb8f34d8f1f4ae5dd18ec5e
487b99ce5d2ad871dad6ac670a3302d074e9a940
23474 F20110114_AAARYS carlos_e_Page_084.pro
d83040f0256cea9a6eee8f8e582b51d8
05a48515d718717ec0074513b416e936b2559b0d
5328 F20110114_AAASFB carlos_e_Page_060thm.jpg
f6aeab10a8126faad2ba905e5ae22be5
6b81ff64f0c793e6b6fcce75c9824b3938f6233a
2705 F20110114_AAASEO carlos_e_Page_031thm.jpg
8d502914be6a0f250689601ceee5339c
e4f7ef53fb8d7447ac18c031a27bbf5276e3cc1d
5023 F20110114_AAARZI carlos_e_Page_028thm.jpg
30f1453bc4d03f4d84b1a7b3c7eabb87
0f13fbe344027de64951cd601b4f5ed9e0229485
54315 F20110114_AAARYT carlos_e_Page_014.pro
7c6c6ace098796c53afc557f1388733a
b48c99ea65ebb754c95d5d4bcd8d4f35a25f899c
968059 F20110114_AAASFC carlos_e_Page_081f.jp2
e10f513c4942077055ac705f6626e467
eb47501f5a03bff4e291ca2b0809cece1157f7bf
41692 F20110114_AAASEP carlos_e_Page_041.pro
a0d113be663de3eae6cce3ba3bd546f6
f49441bf43ac9b8d59bfaa6646883fa358aa3980
F20110114_AAARZJ carlos_e_Page_010.tif
6acdf55e14c57f9245c16617a9f7371d
a3fc69aa067655663479a41a770a29584ba34855
47621 F20110114_AAARYU carlos_e_Page_016.jpg
3cbc9416e5d5fbbe02dac28033006464
1748c1a8994e24ddc90e96d2a44fb8cfbc0aa9da
323935 F20110114_AAASFD carlos_e_Page_090f.jp2
7ed61fffa5e89f06bee75ee9a896b997
20c71ebc4d0cbb0cdbcad37821b716f4a0d2544b
1024 F20110114_AAASEQ carlos_e_Page_034.txt
d7b0e1b54d08e90acda5f060b358f3e8
a4005750497055d00477adca7214ba5216c1c9ab
36590 F20110114_AAARZK carlos_e_Page_045.pro
673e4fdb598e6e8557d8721f7a4f8a15
7036d99f103eba8b8a13b220a837395c96a16665
13594 F20110114_AAARYV carlos_e_Page_092.QC.jpg
4e6fffa0d1f7c394470422d11fd6d1a7
30907e47e02cacf6d07060a206c6bdfdf0ce6808
1051909 F20110114_AAASFE carlos_e_Page_101f.jp2
b93d32c6e1fcef6da1df73a02a550de2
201ec0f1f173028b3d2cc5ebd02c3c8301994b70
77007 F20110114_AAASER carlos_e_Page_070.jpg
804bf9c9944a98b1eaef68b0f765c445
2d7ea01ee6b5b53179d975a17ef2c8ae8c77cdce
71297 F20110114_AAARZL carlos_e_Page_096.jpg
d24d7cd6dbd785c4667a2f9367cbc53d
dc3451c1b4a840bb427f62df2a896e4a7e5bc0b6
4269 F20110114_AAARYW carlos_e_Page_051thm.jpg
5fa0601bfd127b380faed1689b14fd53
e4013819e9962f364fa65ff8ced0fe53a00a1433
42435 F20110114_AAASFF carlos_e_Page_051.jpg
488305651027e2cb445b046fbd988000
0090905f6bf6de927f91bd1d445e797ea9c6df8d
F20110114_AAASES carlos_e_Page_099.tif
c4ff3ec6cfe9e797fb947f6a0ba74aae
3f6ca82b18ede319075ba1e0550522f273c1b9ac
23018 F20110114_AAARZM carlos_e_Page_019.QC.jpg
fb8c0e4a117ec10002e3307c8bf2f94a
1c8e246c3934ae4e40f79f077acacc514928fd4b
571309 F20110114_AAARYX carlos_e_Page_095f.jp2
d2a865ac4d41c5c7fbf14acf037c75f6
1fdbfd5b1286ffb3d8b6e8dd2d686f843ac7dd4b
2006 F20110114_AAASFG carlos_e_Page_071.txt
5e55e950baff46de048255128fd22ba4
2f954b83988bbf141c3a0e52b175375043df9426
1020493 F20110114_AAASET carlos_e_Page_055f.jp2
21ffde1f0437f14cf05e0d2194dfc963
197abf510af595be0abba3a859545ac965ff1d44
855509 F20110114_AAARZN carlos_e_Page_072f.jp2
cb1c84fe9ec2f85525d05e42a9e153c8
fa60c5fe5633011cf2c2c1f2e09a6665631803b9
F20110114_AAARYY carlos_e_Page_076.tif
1f4f971e5d103a38e26ed537d644bcc2
c48e22dbe5d0a275e99618ee7b9334d9c50284b3
942345 F20110114_AAASFH carlos_e_Page_037f.jp2
94f900885d455f0678a66f8739e2059e
987969b61ae84e4fdaf37b09656b6eda62891759
41890 F20110114_AAASEU carlos_e_Page_004.pro
eb0e0a54a7c3d6139cc8330491c9494c
f4a408cc4c780e2ed5574aadc1071840bc48873c
60171 F20110114_AAARZO carlos_e_Page_077.jpg
14aa9a1b0aa4505e322f7fc058177c1e
baf02bfe9770dcdf0f265d08540feb914c650957
1994 F20110114_AAARYZ carlos_e_Page_078.txt
392b67f596e969f933949e5e09038ea5
4778ff2697a76cfad2eec9d06e0bcb00fa660d2c
44020 F20110114_AAASFI carlos_e_Page_063.pro
2168f92efd43275276a400619982f788
8d7b4eda8d4db66a1d1fd80b0991493a77231915
20331 F20110114_AAASEV carlos_e_Page_021.QC.jpg
a15619d2a2e0509a7afefb8c63db19c0
296512d37c88b7ac0cd557eb3dab6d8d672a7ba1
1756 F20110114_AAARZP carlos_e_Page_053.txt
2a34e847e1846388ea2a54f1f7c28ea3
03396d7ad413d9d21d2f00476caecbcf31714e92
67735 F20110114_AAASFJ carlos_e_Page_063.jpg
54cbfeec440e0c449aaa0d6f34427310
cf8c0d3ab2a403969a87ebeede8697e9f47ae16b
338382 F20110114_AAASEW carlos_e_Page_008f.jp2
379b11a5133e5d373e3879dd1615af7e
b51223d5a1f1ed1194d479b7d9fac987afd73427
F20110114_AAARZQ carlos_e_Page_062f.jp2
cc9bd40c0f031f38761ec93f2f4fe4bc
cca0d6d438001c681a94ea8a50864ed808c9cdf7
15907 F20110114_AAASFK carlos_e_Page_074.pro
255708f32abdb21420b4b79276bbc255
b3c31fd7ada1d92d6ba79e42c7ef93b882a19ccd
18606 F20110114_AAASEX carlos_e_Page_072.QC.jpg
e167b2496ee7b8e3b71cabcbf543106d
b8740b38049d8dcd60029bcc0db2c3fdd4e00aae
6497 F20110114_AAASFL carlos_e_Page_069thm.jpg
5983feb2233966b3c2378d9783661cf4
0975c8bf4a2314f3cf542b5d65cc572ca4274afd
2244 F20110114_AAASEY carlos_e_Page_085.txt
6126c4ca7a774400c01cde58b8d98992
2abb88ab3f1507cabb8e5432fd3616748fa712c5
15864 F20110114_AAARZR carlos_e_Page_007.QC.jpg
1db20cc288d7f120f1303d5cc2a99ab3
cffefe46bcdcb1c28cf387011620fdf752c14df8
5824 F20110114_AAASGA carlos_e_Page_101thm.jpg
61ef56e96a9324832f55a83e6133881e
5c39a595bced76423f95eacc42c6cacec02d019d
47583 F20110114_AAASFM carlos_e_Page_033.pro
95019df12f95f51aad08eb5c3e5de71e
f3faa65e2f76ffd1e799a05171384dd462fee028
F20110114_AAASEZ carlos_e_Page_001.tif
a91bb4be5c05516a13ed4b36a45a5950
84cb8aabebe29f01ec99e6866acf0fa6500eec69
1508 F20110114_AAARZS carlos_e_Page_003thm.jpg
7a74118057b64e7e003049c25573b2c9
236a501d9c66164a4932b057255170b7836265e7
F20110114_AAASGB carlos_e_Page_073.tif
f99fb3e18b7b460b24815ffeb621164a
359ec0c64a265be6ba344e8e713cb6fec19c4601
F20110114_AAASFN carlos_e_Page_061f.jp2
90512bf3687f40a1db146632e509c649
467e0ec3ca883aaae408b8c99311add294d63c3c
61285 F20110114_AAARZT carlos_e_Page_044.jpg
aea651b6658a37607a180c6fd84d2f53
f1b7582d662c80aaee9a3cc3d2501be72048c376
1818 F20110114_AAASGC carlos_e_Page_075.txt
7c18c12c0ecfd36321dd42555b9f07ea
66d4da3ecbd38e0eaa17c1f18f7742daa97ba744
1431 F20110114_AAASFO carlos_e_Page_066.txt
2b7769ee65dc80627e15fb64e5bc230b
ac8c555d046f6918237cb35a1b9ed394f66c7968
50543 F20110114_AAARZU carlos_e_Page_043.pro
d1012d8e3409bc778191476859221896
f4bc2eac698875f43031ba49d0b7d6568f7a32a8
1051938 F20110114_AAASGD carlos_e_Page_040f.jp2
9f1de1f48b47c6151815e9f06a36016a
52a0610822e2b5a28450d98e2038e17e4d92450f
67931 F20110114_AAASFP carlos_e_Page_081.jpg
76f32470bfbb67906e0812a5891238ec
5f99114050dabc030eff9b82fc5ebcf2c0f21e23
6436 F20110114_AAARZV carlos_e_Page_073thm.jpg
89fcf8170ca55550d27434f313bdf4cd
4c7b8def94ea6ea5f4fc534106c55bb491f0fe3a
2678 F20110114_AAASGE carlos_e_Page_035.txt
ad9913548466bb9fad3e6b4eba4613bf
0aa4846dcb5aecfb97d6fd11d0b88f4a39246d2f
23834 F20110114_AAASFQ carlos_e_Page_079.QC.jpg
c550deada9789079f2c43af5d66d1b8c
4be9ebe52540d7e9df28d5f9d53ad3d2f519455a
57817 F20110114_AAARZW carlos_e_Page_038.jpg
d9ea06b673847bb2ac94fcf7b7a6cc46
bcd24c78075338bb3d3e043fa55d507054fa78b3
1917 F20110114_AAASGF carlos_e_Page_072.txt
5a5b509d4024234a0473e84fe595b02c
0884556377e2ef68d8a8f1658a8697c6bcd4ca61
F20110114_AAASFR carlos_e_Page_056.tif
f0f85eb2b0f82a901b316c1847649344
a0aeac35ba740b409decdf7e366a2a9455099972
2091 F20110114_AAARZX carlos_e_Page_076.txt
db672b4736b9beddc42db8b7cb04425d
9f154ca25a6ca1a7bdcccff583a431f6576e5577
73090 F20110114_AAASGG carlos_e_Page_039.jpg
22ba9cb95c495723fb1f2564750424a3
204f91f165caba09d427fb52cc6a46ac8af84a96
32515 F20110114_AAASFS carlos_e_Page_042.pro
42ca86e69e8dfecbf46796e37f79699b
4efa37adb4576c653f60c2590a7b495aebb8524b
F20110114_AAARZY carlos_e_Page_026.tif
31fce47ea868af8708391d5a82cd2e83
3040322737e4e6f5f74383babcb71036fc5e8c5b
F20110114_AAASGH carlos_e_Page_094.tif
e7dbfedccda38ea24f4a9f3ddb3b3b02
dec54cd6e73833d4adc8245d2d777c33781641d5
20713 F20110114_AAASFT carlos_e_Page_101.QC.jpg
72e1dfb3ac746114cc61fe91eb1c87db
88ed378bee2b80464ef86df84cc6618c4e0098ee
75615 F20110114_AAARZZ carlos_e_Page_065.jpg
af8b587c4cad1fc4760a54373046ca79
cbbd3155e94c22acb39516e4e7d37931a376b9b1
F20110114_AAASGI carlos_e_Page_040.tif
fca1e43cdc02ec6a2eddc15547a72ddf
8da35ad1072b122030dc7642196b70a73a85662b
198837 F20110114_AAASFU carlos_e_Page_003f.jp2
e9f2b1623fa44a79ba8cd1df0e1f613b
88927a70eaa17632126dc2ded0276a2668f988d4
22123 F20110114_AAASGJ carlos_e_Page_029.QC.jpg
39779167bd159c714812ffa3b102b7fd
ba0c1c92c63027c3f4fd0a9f4078f8a20af9c3f0
F20110114_AAASFV carlos_e_Page_022.tif
7352ce4549d5a8d46e5e472fe1c33aa5
3285fe5a69be9c595d17278a4f15a3598d605b57
F20110114_AAASGK carlos_e_Page_013.tif
5b23b3940e49a6ad782233bfe321e75e
4bfc46a6bb9b72c8e65451ff1bd2eac0559d8b0c
494425 F20110114_AAASFW carlos_e_Page_091f.jp2
86ba85ff5fda8b6a7a8b027a5cb628e5
311aeb11f1cfcc0c26c5116ac40264a0a16c643a
72345 F20110114_AAASGL carlos_e_Page_035.jpg
8ba22134ec9007b7ce939e5388e60c8a
32005ad0d70a59aa888110f64a69a5c0bcdb7a86
F20110114_AAASFX carlos_e_Page_003.tif
ec1ed548dc03177c99457c6431f2cfab
a4e40dc2b1b2a7ab04866095c5bb814c3a9e6519
F20110114_AAASHA carlos_e_Page_032.tif
a48915646a14193a4b46c4575dabc5c4
7bc4424cd5ea8b1a249ddb560a399dca16a6b27b
1051900 F20110114_AAASGM carlos_e_Page_059f.jp2
3fb2bd20fa1d44544868ed24d20bc572
feb98684b6e956dc708214269c11b54b8009062a
56269 F20110114_AAASFY carlos_e_Page_035.pro
4e5e9709eba96dae0f581774ddafbc11
e4f445998e6579ee02fe940cbc6058bf5994216a
1903 F20110114_AAASHB carlos_e_Page_052.txt
1aa4fd83b7c33d1f35646f511cd29c79
8c65c5d4629f33f73f4471f8f4d2df90cf938c2e
F20110114_AAASGN carlos_e_Page_062.tif
61a35088bbee07cbdd2c2d56b4621384
ec645e5cb27dc25eb92a6dde186e042e50a019c1
49313 F20110114_AAASFZ carlos_e_Page_079.pro
eda497c874daff1ad53f58ab1f5365ca
c55edd57c702a4df87c86a3933daaedb5476bbc7
61256 F20110114_AAASHC carlos_e_Page_009.pro
f37a5598d8f24e4e57bb9b9104ea26a6
bcbf4f4164874addb769f64abfffa1b135877ef9
71850 F20110114_AAASGO carlos_e_Page_017.jpg
cd82830b1952421463f1298fda848a6f
d725623185d47ee5a2c4ebb18b83f221e80a2284
1045322 F20110114_AAASHD carlos_e_Page_029f.jp2
6c7e114a3ebee1354a7af1f80aedc812
cf1a0dea08d2e70976a556abfa721492c76ef611
7432 F20110114_AAASGP carlos_e_Page_100thm.jpg
e024c717fca07a9272732edec7a06658
0d006f558b62b9aebea71309b843755c33a6b7a7
66471 F20110114_AAASHE carlos_e_Page_004.jpg
b5a82800bd9164751c4e78a5ceac8dbf
d985c408da03bc748a52d8cb2eab2e94f11336e2
6749 F20110114_AAASGQ carlos_e_Page_097thm.jpg
4b0bc43e52f739b12f79fa4c6f392597
65eef501e76aaa8ed618243200cf8ce41d1a43c5
49595 F20110114_AAASHF carlos_e_Page_061.pro
d15ab6ef6d2c8b7f42ba0b7024605207
0bcb952a01d9a18681fef319b222d843dfe5bc2f
19328 F20110114_AAASGR carlos_e_Page_053.QC.jpg
54ba8dbca1f02b18da2f4aaa8471238e
5a6cd0d266e77084524ef66bbeed2661aba369e9
1949 F20110114_AAASHG carlos_e_Page_065.txt
e3707a1d0c611c85c998e42759d9e7b3
e5b075d68f2d096342a814884382ad91ebd1550a
46247 F20110114_AAASGS carlos_e_Page_026.pro
46b6665a7ce6f6a843197aeb99695050
0c21f07b2a2be95c0ba566a786726084d844bb1d
55863 F20110114_AAASHH carlos_e_Page_045.jpg
c4c01f2af97929c3a26c68550cd25baf
5b89cecdd28f9d443c237877b0b390df80baaa34
F20110114_AAASGT carlos_e_Page_026f.jp2
43e9faa936fb4a8b735413a4a17f451e
306424f74ee4849806238d986605d5b567b9395b
979238 F20110114_AAASHI carlos_e_Page_035f.jp2
75588bbd3aee10d470975e12001d5269
b686810b277a69018f6531a170524c686d19e68e
47439 F20110114_AAASGU carlos_e_Page_094.jpg
f67e704fd850d11e19efa27c5e5208ed
c9dfe69f8805e8c43120012341e257da963aecb6
4475 F20110114_AAASHJ carlos_e_Page_012thm.jpg
f35ef8d1b8e1c164ee26aea089705d28
9d036d240fac8e1b2fd94b895771c2181062ba3d
57339 F20110114_AAASGV carlos_e_Page_099.pro
ff090557f2326bf42ee957a9a9661e1c
ded39e56efdb35492a204b96095a38da8b814c4f
845933 F20110114_AAASHK carlos_e_Page_058f.jp2
8292e5153b218b4f10742c9b00c8769b
2e625258f2846cf5874827cf1fa0b42e4d3232d1
21705 F20110114_AAASGW carlos_e_Page_091.pro
613bb46cb220210698b5c013fb34f409
d3a173ca29e9c4cdd45b78406dfd949911b0c326
18659 F20110114_AAASHL carlos_e_Page_082.pro
e133f50442dc150214cb78e6f46a3485
a8692864b44f5f8dd144b53eef6d7581e7b9d30b
4346 F20110114_AAASGX carlos_e_Page_086thm.jpg
182f4b8325903e8c3d7120766de020a7
7d66e20422d500c719ef315617ab7c6e18044461
18058 F20110114_AAASHM carlos_e_Page_092.pro
568c99b38665f59e0089b5bfe3fa4cf6
1ce003765615b2b9e541ddefe46043eb8116a5b6
558817 F20110114_AAASGY carlos_e_Page_024f.jp2
95016a4e536fce9c37af535772dcf572
1f0c06ab7a1598257ed615f9c6fa5931d3e2dac4
416 F20110114_AAASIA carlos_e_Page_003.txt
b119199c0187b919e70c22e69ae91bf2
62dc7dc9d5108b30d2e869fc4f3cdf0eb626cdef
13425 F20110114_AAASHN carlos_e_Page_005.QC.jpg
b005788edf1703ec11689414c962346d
6dda74704579b201e46b6321b69e9aeca55d34e3
783809 F20110114_AAASGZ carlos_e_Page_028f.jp2
ddae30e10d20b5b8e221d8a56a10db51
0f158080ce1302b057c9bab695a05f2d26a74e6b
43879 F20110114_AAASIB carlos_e_Page_093.jpg
0c870fb3b0fc54186eb1365fe62e5a98
10ba4ae4b781d54ca5d1310037a28739c58cc544
F20110114_AAASHO carlos_e_Page_023.tif
c4258d2d48841cb334ad1f606ee0f8b2
d8961ffb191dcf678909e3d572d5afbcf269780c
F20110114_AAASIC carlos_e_Page_088.tif
0f8d8b0f3fdf198ae0101fc86ad7a102
5c786032bf1ced2b25fcdd8ffe1151673189cb4d
45354 F20110114_AAASHP carlos_e_Page_086.jpg
9999cd20c444ee32ff63caa8e0af3fcb
1ea015eef1cc2a3c6209cb0c7323886fae3c9dba
26887 F20110114_AAASID carlos_e_Page_047.pro
6728677a755c27b228ea8ed7ed5ea6d1
ab0bf9c9e835b6deeeb5a7450573675d92f57eac
1030619 F20110114_AAASHQ carlos_e_Page_078f.jp2
9e37d5db859de9f208c177d6b1fe1e19
badada3be66b031225444ef08712f521dcb47b91
53058 F20110114_AAASIE carlos_e_Page_050.pro
515ff8a5161e8e366d63da4d9af0f467
b57d0d301ab5cfb0717885fbdde71f7a7ab4518d
41601 F20110114_AAASHR carlos_e_Page_072.pro
56643045ca8f84cb6c66625c3d788598
71006676e155103e0376a0b9e92bea2a680fea13
980 F20110114_AAASIF carlos_e_Page_054.txt
d4d0d8c27657db9f8d19b6389e08a645
553c8fe4bca975d4aeccf1de44a84ddfce068516
760945 F20110114_AAASHS carlos_e_Page_034f.jp2
be599e9e20127039c414dc50868ffb0c
ced378b803fa0e7a76cd8c6cda102545f9bd884d
89564 F20110114_AAASIG carlos_e_Page_099.jpg
224f458cf74ad94355f041f91c9a8b98
e071fd5369d5ec23c450a9b836bdbd08646d0a52
77375 F20110114_AAASIH carlos_e_Page_042.jpg
73bf0806e777ad0f82d148b892059e2d
4d980d1b5cd953bb979e56b3d802ce8febba2dec
61210 F20110114_AAASHT carlos_e_Page_064.jpg
762a6b4dce921d5b529b8cdd3898c375
c8df89be3039145ac145c99df648b94d5388c853
F20110114_AAASII carlos_e_Page_076f.jp2
93a0f9b035b0ff64dd3eb62a16e9f86c
0ff29bae36a48459ca13f94fa70eeab4e2371f05
65714 F20110114_AAASHU carlos_e_Page_013.jpg
05b2d955b7beac95e14737154ab78154
ad2709d1415e4773c1b087faf254010d42d6f4e9
66059 F20110114_AAASIJ carlos_e_Page_052.jpg
c967a62352d70718ec328519b7ab7558
e97c96bb4013ac2193c1a00e383d45e6f6036fcc
22317 F20110114_AAASHV carlos_e_Page_018.QC.jpg
1bbb2109a62ba60e782913b9ca895b21
94ad0855dedd55dc5d15f4e06e11f5809ec18d98
F20110114_AAASIK carlos_e_Page_047.tif
e0c0703cc91484f8632bc0e480f7b10f
fcc93c68b46d5c3ec1a59a02168eeb43ad3bc5b0
26969 F20110114_AAASHW carlos_e_Page_100.QC.jpg
48e8b2d76b3d457a458a90cdf08c00d1
4607663b1aa1a622ea9dc128d97b6688dd025f89
75728 F20110114_AAASIL carlos_e_Page_068.jpg
db789cf346f944b8eea9a014f618ea4f
b4e6d3c9833dd5efe2240791cd7c7adb026421f3
17697 F20110114_AAASHX carlos_e_Page_028.QC.jpg
1e4896436f68d57d91b69b676f120e62
16b4da8cea89f8c6e72cf4db2b784c2f26e2eee3
F20110114_AAASJA carlos_e_Page_101.tif
321e29ab0a31204e22a727e2af126e71
43a6c365af032c994774d7b3561bfff11d05ee1a
53250 F20110114_AAASIM carlos_e_Page_098.pro
836eda3d4f6a0efb7a3604fcec8dd983
de3d79707d72ac91434a81da87339c91b18e47d7
62019 F20110114_AAASHY carlos_e_Page_072.jpg
984f9f140aa2bd83018fa04612b21263
0767b1c7d48ae6919fb299b5f98113a8d391222d
63569 F20110114_AAASJB carlos_e_Page_007.jpg
9f7bcc2bc77d8c050de57c2b05b07835
cb5ae8ebde34086efa753b2b4667bae41517ad11
5866 F20110114_AAASIN carlos_e_Page_049thm.jpg
9e478587db5d2198aa1190425cd49a8d
07de010726697186a558c12cc4e549f59abdf96b
1051978 F20110114_AAASHZ carlos_e_Page_030f.jp2
61fb1e13d4c92d12b5977d829e3c33e7
6fcf4bc64d95e5e6e9b36c981afd747807c52123
3977 F20110114_AAASJC carlos_e_Page_091thm.jpg
391ca00795c9f335f159d132cd06a211
95c0099fe53deaecd9d8eaff942d9f01d363cf08
44888 F20110114_AAASIO carlos_e_Page_055.pro
a20938ed93d8a641135a87f0734f6a32
676e3f81ad765ad3a91ece96ccb0d127c0840b25
43471 F20110114_AAASJD carlos_e_Page_096.pro
f4ce2d09a686a2ad269e4bc81d83d2aa
34f3e1c5b55fbbbfdd6dda398fd04d1669173bd1
53037 F20110114_AAASIP carlos_e_Page_023.pro
d4474b7f42fb61876046638bba5c5c0c
7ce2625f15fd722751e60981ba18b94e6eedff38
5842 F20110114_AAASJE carlos_e_Page_042thm.jpg
9219a6664668431367efd732773ccae2
909a8b95f1e7797ac80f4f25acf770d4d638861e
20225 F20110114_AAASIQ carlos_e_Page_037.QC.jpg
cf10e317df1e2a7c9bc367de53e12e6b
09cec6e41d1c860bae9a7877c15b77a234b38aa7
4408 F20110114_AAASJF carlos_e_Page_009thm.jpg
36b2bb43504c2b3d7d616cc1d6d09035
1e8bf301bbc21264cdbbb66f85ced267b40142b8
383532 F20110114_AAASIR carlos_e_Page_092f.jp2
3e7cb3d6bdb5b0c59f7e4ae6a23d9c4e
ae9d06caeed1d807cccd2a671874b4c606dee9d5
42915 F20110114_AAASJG carlos_e_Page_013.pro
c5b2bdb3a56a54666b69f32ddbfdf113
7b42912b8e84cc9f49a800246fa8eb102115fd69
2496 F20110114_AAASIS carlos_e_Page_014.txt
7fe0d81a48058dc4ffa40c676ab23650
a84487d5100f19ce770fa139ffbe074b03642f15
1016195 F20110114_AAASJH carlos_e_Page_075f.jp2
62d60e800ee1d20a8c8ccd9403ab63f7
71af27e912f6ec46fcd91bbd48d97cf6d6cf99ac
741120 F20110114_AAASIT carlos_e_Page_038f.jp2
3c7fc394b33f38ae757e503b7b98710a
eacd97e03d75527f379c8a2f954955a221fe373a
17292 F20110114_AAASJI carlos_e_Page_031.pro
3147224c530fd71e1d7b466025f76850
7f175696eb2d2eb0e127953a4dbe3c29bc960f46
17873 F20110114_AAASIU carlos_e_Page_015.QC.jpg
e3b3246811dbc08737ae5555c155203f
85ecf567d42ec436665bf969452c4ab077e4b9a6
698 F20110114_AAASJJ carlos_e_Page_031.txt
9deda808322b2adb8eb9765ad22453e1
f8485766d1ce8191d9c86bff17a7bc875a7e16e7
1912 F20110114_AAASIV carlos_e_Page_039.txt
bd8a2d06b85956204b8fa604fe75b9d7
d3ab79a0d1d68aaa53a624eac766b998e53919bb
967 F20110114_AAASJK carlos_e_Page_008.txt
c4cd12ff0d995a69f49a24247076c21f
cf6b3cdca5052764e4cd35d8946771a6faf8b256
F20110114_AAASIW carlos_e_Page_017.tif
353ac5b09e1114e6e63335bb308888e4
3643a507400403653844d5c95dfdb70cae76808b
13672 F20110114_AAASJL carlos_e_Page_051.QC.jpg
5ecdaff422b28e4545d51e4efcae60da
2873180b979045f2ade05508997f6bdc78854643
6205 F20110114_AAASIX carlos_e_Page_017thm.jpg
d0c4ea5491e266bbfc10a88f2b8d67d9
5d00e240f21b58f324d2640cab9c2f1404b39770
940470 F20110114_AAASKA carlos_e_Page_013f.jp2
bf0d4c31988913e846a7da28693f0274
2534cef0d522d9ca8068dfcd45ce1fb6918f4b75
27798 F20110114_AAASJM carlos_e_Page_074.jpg
5394fc7c8da82cf3e852048846c52093
9b5126b95cd6077ef9a6781a1aa3713f5ffe1da3
15438 F20110114_AAASIY carlos_e_Page_064.QC.jpg
eea649cb3fbfb8ee5c8941feb82c2d43
bd1cf479565d9bcd6a392b5d007a616d4785a197
F20110114_AAASJN carlos_e_Page_097.tif
735fbd9286a8072cf9c6f820903bcf43
a370af033e9134ccb9297b2d2b0f16c57c70f712
5062 F20110114_AAASIZ carlos_e_Page_058thm.jpg
741750c0a31d820c9861457a6b32ea04
df037a998e0be47191ba666e38ab8b27f5859745
54758 F20110114_AAASKB carlos_e_Page_009.jpg
d07c85a011fcbc02466ec090069cdf5f
cb786db6ca8d9fa74ccd535f19ae40cee67d3f3c
6474 F20110114_AAASJO carlos_e_Page_056thm.jpg
a536efbf3a2625c7d35125339f5522b3
1fdcf1f4137e77e0d41bebf0cf59481bd7831d5e
6598 F20110114_AAASKC carlos_e_Page_070thm.jpg
3c9a2cd18c879af471884195212db5a6
2bc98c78398847f0c38771b4e67d799a74d12f8e
3410 F20110114_AAASJP carlos_e_Page_084thm.jpg
c84dd92057c170a624c9ed17bc370541
c62713c277c484948d5eb16750f638f517d748fb
F20110114_AAASKD carlos_e_Page_085.tif
fb550a355b1fd14336fbb3846125620d
04bd13768285031b3d18650e366021c6e0659fa5
1030100 F20110114_AAASJQ carlos_e_Page_017f.jp2
2cca6535381094221346f750d5bd81e3
199fdf9168d0d48d6c58212eb302bc78ad4e210c
F20110114_AAASKE carlos_e_Page_102.tif
03f6fc28b0bc7ac3f628065506d9dac1
124be8f5acc7d9b0883937ea9f50d68865deffee
2698 F20110114_AAASJR carlos_e_Page_006.txt
a1ad14f683ce16bd310248225ed10369
77f3b0f343058d766840ddf22a0883761ff741b0
1009748 F20110114_AAASKF carlos_e_Page_032f.jp2
7178422f19850a4488dcf89d82c5af61
a13fa4fb6c632bf5c6af7572fd5fcfa97c5171fc
F20110114_AAASJS carlos_e_Page_081.tif
0ab55cd84993bbdc110e5a260fac3f7f
cc8c79592d47b0e1409aa3db538726aa209822d2
2545 F20110114_AAASKG carlos_e_Page_074thm.jpg
c5da3014e78e0b05fa7150ebdc41ccbd
8e5a9a7dc5a7693f12312e1b712f0bc172ed5080
1549 F20110114_AAASJT carlos_e_Page_047.txt
c27a02cea34b205bb55ce45b2ea8e933
4d513bae4af548121b51c017695f9ad16bf0abcc
6386 F20110114_AAASKH carlos_e_Page_068thm.jpg
aa179e6cbdfe513583aa88a4e88465c6
10893f6fdfeb2ad96436b0b5864cddb42374cd75
979606 F20110114_AAASJU carlos_e_Page_027f.jp2
8069a619e5e81a6c762fa0b4372b06c7
21c86020fb02cebe5678105d8e53e7bf3dcc99af
75901 F20110114_AAASKI carlos_e_Page_073.jpg
3a358b2847a1d4add5af81a427cf6db5
fd3598618c56ab28a8fdcee8e393d5d2a4a84237
974427 F20110114_AAASJV carlos_e_Page_015f.jp2
02ba22344651871e13f12772826f12b1
0735936626a197fc9de9c021d0e63fa423150749
5807 F20110114_AAASKJ carlos_e_Page_048thm.jpg
f19d1ea0000191349e5f0d4db3a9071e
3dcf223a9debd8ac0334ba17c30c5fec0db0998a
F20110114_AAASJW carlos_e_Page_037.tif
28829fd1cc7f977b2276273960359e4c
49cb4b1d9d26d1be77df206ff618911e2deb0abd
1161 F20110114_AAASKK carlos_e_Page_051.txt
edd88eae835ecd6ca0a378714a0fce16
bc39bcd0438ff03ac4f1829e752939a935176f74
F20110114_AAASJX carlos_e_Page_062.txt
2fdbaf723ff144bf82ae14b66efd9369
6ac69f780b549120a47ed6d6aec52a25c6478dc3
2250 F20110114_AAASLA carlos_e_Page_050.txt
e0dc17fb838c54888eb00bb7ebb3fbde
60a5ad115eb350c80b01ea2a3956e4d797e336fc
14154 F20110114_AAASKL carlos_e_Page_103.QC.jpg
2618aa4b17d7be927eb8b6da2c123dac
1df9f5683a6c3a9b15a2a1a67178ad626e30d59e
735 F20110114_AAASJY carlos_e_Page_090.txt
f7f23f8d17b6e86ce3b53e6c69640045
6fce40620bc8bcedb8f7a2d073e91343ec226e52
F20110114_AAASLB carlos_e_Page_054.tif
e0c32f8bfe15299b4d7f6dc42338154f
6a547dbc45b990dda4b15f5ba84e7936b3b6fb9f
67758 F20110114_AAASKM carlos_e_Page_006.pro
148bce0aec00122b7029ece25d65e75d
87912c474a7f072f8da397424a0902f02b8e8d4e
1051980 F20110114_AAASJZ carlos_e_Page_079f.jp2
68cd8c9a85b7de0881a1b892fb742e5d
1903a12cea4846768326ac2508a8447627e6e3fe
70663 F20110114_AAASKN carlos_e_Page_018.jpg
49986bb98b3ae58c742023e4bd62f4a0
d4b1b7fd8519537844406f52530314ce09e04810
8007 F20110114_AAASLC carlos_e_Page_001.QC.jpg
63d839c4ea2b91197d878a20b4bb8af3
4e39525413f16d260d19a840af565660d7875362
F20110114_AAASKO carlos_e_Page_090.tif
dc94fd22d5e0a6ec0d8b09b8684ba7b2
c0ebab83be4caaaca78ee563d9d5e2494c0e7cc8
F20110114_AAASLD carlos_e_Page_066.tif
1fcb68f1c1b6d4d2ce91d2104372ea7b
f05ab3d7c05c56801c11607f66137351ea6c2d8b
F20110114_AAASKP carlos_e_Page_053.tif
ddc11aeea12d2e6c33ac31d0d96ccd5b
3da614209fb3bbed053e5ad033a22300bad53ce9
F20110114_AAASLE carlos_e_Page_033f.jp2
8a4f00aede1cba0fb19a91a918608a86
a6e6602c52b48cccb38b6a67c2a507f8e4e9ea4f
24647 F20110114_AAASKQ carlos_e_Page_097.QC.jpg
936f49bdb499f3942de7f37194d5ed5e
0199886e0ba6e578cf01c5d1a90840231018292c
40151 F20110114_AAASLF carlos_e_Page_027.pro
eea7ae0ee63080684f6a20f5357a1191
74983af5d5aaa922b901c6a76b516115e0ad5567
12739 F20110114_AAASKR carlos_e_Page_054.QC.jpg
1d9350619d6ca75bb3b6de66659adb2b
5395a9eeabb7a746700c3bd789815c531d25f21c
54925 F20110114_AAASLG carlos_e_Page_089.jpg
2615bacf496b583f8f0b19202c4fcd3c
5e3090544bf61c672cc17acffc764cdf3fdac1d6
44640 F20110114_AAASKS carlos_e_Page_075.pro
358bc8de67b5ce745b0cb964a689a48e
8ed23e901902788dcfb6cf34fe2857a87d7f0849
1051949 F20110114_AAASLH carlos_e_Page_100f.jp2
2cb00f73a43bd6714b7c90f7017454e5
76706249f638c4d0905bd137775c6a60138adc8a
F20110114_AAASKT carlos_e_Page_091.txt
11f9ce1fb51b13d249597dadd01ea37b
cf7a0ae5b3f5c6ebf8ac41ae2cc5d6e43c85ffa6
384943 F20110114_AAASLI carlos_e_Page_031f.jp2
225b18c4a0ace8c95a9d5bba78e7c3b3
44e68c0f0da0fe4a9c21a85eb83b2348573a3d68
910 F20110114_AAASKU carlos_e_Page_064.txt
c44ad2da6c51268b0d14f96592b1fc78
6755508f32aa5a0412bc91a6a947baa423c31f65
23727 F20110114_AAASLJ carlos_e_Page_065.QC.jpg
6e7c3746ef89ba24170ecc3b0dc1d544
39a97d11577e2d67649faf9999430dc4b475bc89
F20110114_AAASKV carlos_e_Page_002.tif
fc7f1ab9cb12a932a036da53e9e9dd37
c4483f99a69f1a93348903ba15b81da817fbb676
F20110114_AAASLK carlos_e_Page_073f.jp2
fe9b0aaa646e5290db3b77388abe1eb9
b0c6d2c907c4ae2af8b88cc6aef1f42f55e0c517
80202 F20110114_AAASKW carlos_e_Page_076.jpg
ac06d1508858cbae773947481400a1ee
8492d0eb95dcd8923d0135fc0689328da10f5407
1973 F20110114_AAASMA carlos_e_Page_056.txt
f5b1d17a92678acf79ceeaaac4506213
532992a78778dfbd08c8899996cc39d423560b06
69792 F20110114_AAASLL carlos_e_Page_025.jpg
07b28ed73ecc6489596385293587253c
edba353f52f7342570defaef1201c64368b2b819
F20110114_AAASKX carlos_e_Page_004.tif
7afea256bb99f4b4e07611f639d5e256
a297a9a5e9efcad6b4a018964a82e16751e73895
23242 F20110114_AAASMB carlos_e_Page_020.QC.jpg
eda3dcd14d02d7304c96c24d1f0264dc
440948bc501fa663f47427676e3ad8c15ecef0b3
22978 F20110114_AAASLM carlos_e_Page_067.QC.jpg
e1b34346c72493a9a6c3f635ca8d6695
556b257104517c2baadcaff9fc1ffb996911524d
2080 F20110114_AAASKY carlos_e_Page_022.txt
c29cd7013672f5f14974521589f300b2
af2c47e2ee36fdba3ad791d0b49e698207131011
69436 F20110114_AAASMC carlos_e_Page_048.jpg
4dd6856ba2f9275c665b2bf3ad4ed59c
5e156d341d772f500b1e8e8b6e46651ed7e1d5bb
2020 F20110114_AAASLN carlos_e_Page_069.txt
e7ade36e91593c3a16b976d40a541e06
6da8da3232d43a1d246989e7c651da41ce6631eb
5285 F20110114_AAASKZ carlos_e_Page_046thm.jpg
294818646a65458d736e577343379190
d315968dc08539f48217d76ac10e02f53238b099
33665 F20110114_AAASLO carlos_e_Page_090.jpg
f3c56c70c0473f1099106dae482fce3e
aa18f59dea6ffcb85357f7d0e32c5944a8f79e71
20958 F20110114_AAASMD carlos_e_Page_066.QC.jpg
b30f2bb5b1ca94fef571be1f88b78d91
c4e88cd97f3f4b1aab3ada48c0481fe1e5d20c45
971061 F20110114_AAASLP carlos_e_Page_042f.jp2
018b78d0db5a794a66ac262e3c2dca0f
63da29bc0f85d8fe478779a637b0763ff6f31a13
38330 F20110114_AAASME carlos_e_Page_084.jpg
af72ce6cd50dd886e6dfa76da4142cda
71bcea08f20b501d24aaf8e83559f7a6746a3b78
1883 F20110114_AAASLQ carlos_e_Page_018.txt
3bfbad731bcb6971fd16515463e2a7db
516d0aa0e990a76e28c3c596f7566c83b1ab279b
F20110114_AAASMF carlos_e_Page_083.tif
b4c681af9c37afb501059aea43d2fb16
e3305fedbd36070e7b80b2753f4d5bd2e773868c
1594 F20110114_AAASLR carlos_e_Page_044.txt
91b272413794d99b12c48b88adea0aac
2aafb07ab807f4b1b597228d9184ab3f84bf61eb
23536 F20110114_AAASMG carlos_e_Page_034.pro
a3a200bd4ca509a4e9c09e33cc37a2ad
cf94ecc1825bb80162a9fdce13253e98a3f7c6f2
609561 F20110114_AAASLS carlos_e_Page_103f.jp2
2c0ce79afbf90cb1e7948cfc3c8b2e02
1d24afb7391b408420bab14bddc0530b2815ff13
530 F20110114_AAASMH carlos_e_Page_001.txt
d978c81ce7bf767d74436690b1cc72cc
b6db68628fd1e3a45b5dc5dd46fd71acc2ad1fcc
23559 F20110114_AAASLT carlos_e_Page_062.QC.jpg
5a46a08ccc0ef489e746e4195c7e948b
8c44acc7d82c6df263a50abedd20e0a8079522b4
28303 F20110114_AAASMI carlos_e_Page_001.jpg
b8bc769569e766ae4556602c2dba06af
ca68b065169c3a38f61cfa5c8ecb76061dff5b05
2104 F20110114_AAASLU carlos_e_Page_040.txt
05d7bb51f821210c5bd6e8f1134a32df
1804d2487072af1947acd5da7eacece34939f274
151173 F20110114_AAASMJ UFE0006602_00001.xml
c67e624da85b7fc0046699b094d1372a
5c6dbc445063430b8d91c96c11627318987db09a
24955 F20110114_AAASLV carlos_e_Page_099.QC.jpg
26f21e3818cbbebd6850f90a8f6feebb
da6c015d3904ebd4af50161b7deb3da3d0b370ea
5492 F20110114_AAASLW carlos_e_Page_096thm.jpg
01074487dd0f906b8749296335d651f2
be63995c3d27a60dbee705fd6aad9b25024a5ae5
1011812 F20110114_AAASLX carlos_e_Page_025f.jp2
e278a0129833810ee38868f13ae587f3
8adc3e6217823642bb66ba07d96c35beb54b2078
70381 F20110114_AAASNA carlos_e_Page_032.jpg
71de5e11e744069285152e253dde14bc
eb9b9d9687e6f2d11a6f46103feb4da97bb317f5
8399 F20110114_AAASMM carlos_e_Page_002.jpg
ce818f084569ddb4fb2c31726516729f
8cd016e9a84206bf6ea86d84cfe45b7cf062f2f9
F20110114_AAASLY carlos_e_Page_078.tif
32f0ca89ab49761a929efd69f3199905
4ab7f19e617520ddc4c288f31e57cc43fadd9a1a
73272 F20110114_AAASNB carlos_e_Page_033.jpg
616183a16ce2488a491e9577ed816567
5c7dceebdcc54288e192324b6e0125d95a218165
41171 F20110114_AAASMN carlos_e_Page_005.jpg
a96f097ab7bd1fc9ccb29c3891c20bbc
a587ebb4055d304e3ff622eb9122398dcf1a2cf5
1674 F20110114_AAASLZ carlos_e_Page_086.txt
bca645430ccfc740969d950c0b29234e
39b92f37e54fea2cc1c9d761a18da1215dbb384a
70699 F20110114_AAASNC carlos_e_Page_036.jpg
781e86b1c0b74fa70e9a5db80584d99b
488323610c817c296f0b7b8258d77b93e5905b01
26679 F20110114_AAASMO carlos_e_Page_008.jpg
8a89fb43b4723d07ab7d153539aa1003
335d7f303c4617eab518ad1659a6d2f5d2fd778f
65682 F20110114_AAASND carlos_e_Page_037.jpg
0dec087f691981c7265e39c6a2dfdfe9
15af4d2135f3406e4859e538c85c252fbe1c0376
61879 F20110114_AAASMP carlos_e_Page_015.jpg
1d36132aac049d58d377eb73e190c53a
1ac4fcf9508e14f359bdf4bddc9bd594244d5ac8
64727 F20110114_AAASNE carlos_e_Page_041.jpg
e6949405936571b0422e41ac0232c1f4
75f6a01158d0746fafa673b7f37c620a53573f50
79364 F20110114_AAASMQ carlos_e_Page_019.jpg
a801277c366ad369b972578d8bae2851
7671a6cf6bf9ffeef112df073e05802e32f1bd62
76036 F20110114_AAASNF carlos_e_Page_043.jpg
fb76da80653aecbe4f133c2093468db8
b4ae4483b44cef4649e4c0913e3949b84c67f220
74991 F20110114_AAASMR carlos_e_Page_020.jpg
4d1610fdd1d15a9d3f4c0aa2eda154b3
f0bc91f78e1a52075db915ed5aead14dd3cbffeb
58820 F20110114_AAASNG carlos_e_Page_046.jpg
fdab31f4ea6b261f7576dc7142f5d012
9e502891b6b7c18c60072ef0a785b560f04bec94
69530 F20110114_AAASMS carlos_e_Page_021.jpg
8c173e0c94ef82e42cd5a7799eecd2d8
e8e2fc1251efac1b9561cd60aa2e75b907b34b98
71724 F20110114_AAASNH carlos_e_Page_047.jpg
3a3f2a782e5c65f0c22aefda79df742b
d1d5bcec0a65b32beb041c4277e220e0fd879ada
78796 F20110114_AAASMT carlos_e_Page_022.jpg
afd852046495ee00f5841a83b69694c3
5ff14a78fda4156555616c812131da130a84d1f3
78142 F20110114_AAASNI carlos_e_Page_049.jpg
a9b65ad3830866a2b4f99efb242da322
587adb0a7128802b03bd0e7b80b1d0e1f72a1f01
78770 F20110114_AAASMU carlos_e_Page_023.jpg
32c2846587dacf902429774506ad33c8
bbbd53e69b2f474d66bec62fa00c2d3fb6a8c349
63061 F20110114_AAASNJ carlos_e_Page_053.jpg
50752a4e000ad240fcf2671166d1e2a9
3ca4738e412f5a1f75b38fabd53f28c02800a930
41812 F20110114_AAASMV carlos_e_Page_024.jpg
92808a9d00acb2f45436f961d1dd6c8c
463dd5398f1cfb7a66c2fe2e83dee01cbca1923d
39830 F20110114_AAASNK carlos_e_Page_054.jpg
a5cbb69bd6d570959ccc102a5917eb4c
75a09545931eea0357e33cfcf2f8620e5ca0b500
56871 F20110114_AAASMW carlos_e_Page_028.jpg
a0475b6c5456cd088a8a35d485078403
0383574f1628be676a3ab819bf01050353fc9f90
37512 F20110114_AAASOA carlos_e_Page_092.jpg
02aa67b25b1e216a44a1c4189850c72b
fb0338906a6f0191e04a0235e704406c773f2f90
71432 F20110114_AAASNL carlos_e_Page_055.jpg
289628db1c20ba00359c3b74cc858248
88115be7472bf8fc533f781b1208a85ac1a2b31a
72243 F20110114_AAASMX carlos_e_Page_029.jpg
f35652da9c5d031b798021327695ab7b
a4bb1e3d3712d3ea4a7c74ba1f09ff65fcefabc4
83486 F20110114_AAASOB carlos_e_Page_098.jpg
4ecac7130accc50370b225e36bdde922
0f6afc1259bc5b8e9abfd30ac1d3bd38cc1845a0
76716 F20110114_AAASNM carlos_e_Page_056.jpg
ffbd7a5bc79f5ae3f0ecb16f8a47566f
46f4ac2e826c88bcafba0dd2950b9812608e3d74
82327 F20110114_AAASMY carlos_e_Page_030.jpg
9a774ce9e20c428033a6e6b983202317
dbd86de7d8938d3faaaef1c07b4391f2c3c75db6
96482 F20110114_AAASOC carlos_e_Page_100.jpg
e129d4ae4c7ff738030a5e3124fa5340
3819242d80618e3f205bede52a5958249c92049d
63410 F20110114_AAASNN carlos_e_Page_058.jpg
3f33ae680f8fbd990b934b1d798696e5
0506d13f1482acd123a2dce45f0ba02c744fbd01
30262 F20110114_AAASMZ carlos_e_Page_031.jpg
a76025a8504d730478a69e0a23e7e81c
2c4868e86b1d890c19183877d240230138c4c8a0
50487 F20110114_AAASOD carlos_e_Page_102.jpg
f5ce923f415fad88dbbc604930f9688f
1c540923213a8d3018e4d84563368e7b0fe34e3e
77474 F20110114_AAASNO carlos_e_Page_059.jpg
2973dfa236f4aaf22ac11b73587a1f51
eac71bfd1e8f5432913cd39d7ca6bf68a46fc594
45412 F20110114_AAASOE carlos_e_Page_103.jpg
0c4c89ba6a67036881a4012fd511181f
72cc672fbf24598ffec5f4454b25a28208210bb6
74494 F20110114_AAASNP carlos_e_Page_062.jpg
89f4586e7f9964038e9d342ef974d823
4e27e9e7af2c40761234652e93c268f177ee4956
73636 F20110114_AAASNQ carlos_e_Page_066.jpg
6f99c82776cfcfa9b9c151bda777c94c
82f73a0059e74c8becd99a844738916c770f278e
F20110114_AAASOF carlos_e_Page_005.tif
4ccb74b0363420a170f6affc8d460972
db862dadf71a5b3837a44bf844948f6103cf78cf
74613 F20110114_AAASNR carlos_e_Page_067.jpg
522d7f0f4ebc080f8587f4280e36182c
028928c550785f504971e6998c19ff2c47c306a7
F20110114_AAASOG carlos_e_Page_007.tif
369fbbafc767bb51e07e04d005a59bbe
f5a225ea52ac1817bf5b0b527906d91234994d3b
71341 F20110114_AAASNS carlos_e_Page_075.jpg
f96fcd9282493c43ede29c571a4a1751
2b4f2ffc95de63eb2e4b3f3dbbb36f0aedad78f5
F20110114_AAASOH carlos_e_Page_008.tif
9478ebd7420f116434d61e7adcc0ba13
a4366dd15f524584cba392c008a615f98b110349
74432 F20110114_AAASNT carlos_e_Page_079.jpg
af376674930d2a3e00b1ebd9fb552664
a69c5af7a675e2ca61dc62d153b8d5a4bd0269c4
F20110114_AAASOI carlos_e_Page_009.tif
91e2f0192235e8e0131d049ce7aaf506
b87f321e49323fcd2461467af6423f0f316887f2
69897 F20110114_AAASNU carlos_e_Page_080.jpg
22d791383b16ed81c7be41bb9b797c3d
f97505b9890e6722eeab50e9b079238755d4516d
F20110114_AAASOJ carlos_e_Page_011.tif
5a527740f58bb108c33a53d4888e03a0
adc465597068ac971c10d417b09ae33c680b0114
31819 F20110114_AAASNV carlos_e_Page_082.jpg
323ce5714caa4d725e5ed486f122e95c
eec5940640cc1b7e5574faf54cb0018390db0aed
F20110114_AAASOK carlos_e_Page_012.tif
7bec4d27e88c133e69c3185b17a6abee
fa75fd3dab904126e2cb9769ee2ab679bb49bbb1
65463 F20110114_AAASNW carlos_e_Page_085.jpg
a6cd9baa3b69acae00713aaa27ed105d
c3ec3d8915ef4a2ab0352040b6bf946ce2ac5046
F20110114_AAASOL carlos_e_Page_015.tif
16256144a867233b629e1d4c34323055
a4827f35859851b5c9b972122281b44e80004f95
47259 F20110114_AAASNX carlos_e_Page_087.jpg
8115c8e48a320e59c910d9e73690b852
44f6ee04f32386b50012d89a1a28f5241589a6b5
F20110114_AAASPA carlos_e_Page_059.tif
566fce34119d88859a27e7d72806bd63
b1c665b437b74e088ad5774f4bcef2cd0e3d6840
F20110114_AAASOM carlos_e_Page_016.tif
a5b95d2293722faa88b0d9e5dc72fd0f
718816acd97cd51779e771208685ee4225127c59
42136 F20110114_AAASNY carlos_e_Page_088.jpg
e341beaac978060c3564154282899b47
81a284a0b9509be6b5d6a36899123856b85844b2
F20110114_AAASPB carlos_e_Page_060.tif
6daa6844e1da10242a9ad5ad18bfb714
ee47985b054199b0480e45a34fdab519a52aa683
F20110114_AAASON carlos_e_Page_019.tif
86fa5f547c1ca7faf3ac61f13e573177
328d77af8c5f4601a0e1da0052dc4168c06d0700
44645 F20110114_AAASNZ carlos_e_Page_091.jpg
5805abf796387d3e8af86b45f24bcdcd
1d46ed565083ba8c4c275ea515ad1c0d1610f005
F20110114_AAASPC carlos_e_Page_061.tif
283b7748ecfbff57db74fc394296c6ef
b581447ce920d9f6ee687dccf0024b3106f38e61
F20110114_AAASOO carlos_e_Page_024.tif
8ec3ee3d08e90ad871181664fab320a0
b98fb26d0540fc5266acab33d0d74eef6f268fd5
F20110114_AAASPD carlos_e_Page_068.tif
c979286ef0e5b546c63b86b3df6c8e42
fcfe75e1f6430e95a928b6dee53a1ae19fe9471c
F20110114_AAASOP carlos_e_Page_025.tif
5c2127dea91bbc3e76340f15b98e8f77
6c41dada1426bbae0813840f0f86aecf52bc8b15
F20110114_AAASPE carlos_e_Page_069.tif
c014b2b3dcf688a27d373ad3b1153fd1
9c62cb6596f75ae41a458a269604ae01d52bcebf
F20110114_AAASOQ carlos_e_Page_027.tif
ddc85a523bc41c1cd91016f4e7203ece
168f85e92110274cb7de4633206773deac66ec10
F20110114_AAASPF carlos_e_Page_070.tif
f5a08bcc900df1b062c32e4ff5f34b14
a6ee89ceb2c6ac913e66a949fe0b11a71314a6c7
F20110114_AAASOR carlos_e_Page_028.tif
dc9e5c1e96efc8f8cc4036a88156a2de
f6523e4840f73d8b4383989b5cc6c8ce497bb4c5
F20110114_AAASOS carlos_e_Page_031.tif
6127795cc6971cd0664af5bf2001645a
f353b326653c45cf86c2d65b5a9b2ab64fc5bc10
F20110114_AAASPG carlos_e_Page_071.tif
f91dbabab212e0d695c12c9a12c43af1
d199ec7f18e021ab4ad00e8b37556f240b55bc2d
F20110114_AAASOT carlos_e_Page_036.tif
16ea5827923e7c6417f5945d7794c61c
42183beea1f764b3a8a1a9211b0a5f2e56c8dddc
F20110114_AAASPH carlos_e_Page_072.tif
8b7c87da6f45964427a6eb0eb7718253
e54e7dbd7c673a39c010984c39048d4716297e06
F20110114_AAASOU carlos_e_Page_038.tif
76e6f5359d9666bffeee394f67e520d0
95f8a1c943f4840c10c5181b5c08f359ef942991
F20110114_AAASPI carlos_e_Page_074.tif
997f3c4ace8d2070f112594059f2d8bb
123f8f6e7e87f75c7fc35f1a80d9c8b6ff4919e1
F20110114_AAASOV carlos_e_Page_043.tif
3ed95c9ec31990d9f4896e4c95185a0b
99acb1ba0bfedad089d37b9e37aad159a9272503
F20110114_AAASPJ carlos_e_Page_077.tif
291e3d657b7fe532a4875b74f80b0f47
e7ca0ce2db84d796d56f187780e1da4a08c2a971
F20110114_AAASOW carlos_e_Page_046.tif
0bb098dfdc7349e0f0d1be44b054e9be
b55977b2c12e42c248e1786ffd41b6134fb855a8
F20110114_AAASPK carlos_e_Page_079.tif
bb9802ef3dcc10aa4fcea9dbfa47f29b
5bd140ab6b4456da7a1b0c1946bc6d1e7ce19d7d
F20110114_AAASOX carlos_e_Page_051.tif
cb4278210d3755c21de2966d383914bc
8194f014fd82b7f42e57141486f532ac0995d48f
52363 F20110114_AAASQA carlos_e_Page_022.pro
eaea40421e5cb175c7496005a5a27f80
e3e81c0db8fd2f64baa34084de371779ca6449a5
F20110114_AAASPL carlos_e_Page_080.tif
c6945d3332db15aafcf28fc175ab86bd
ca7cd774351fb26a40036147acd9a9b6e246eb8c
F20110114_AAASOY carlos_e_Page_052.tif
ac2e0274e999fea5b6fe9e87f8ac00fd
d0928287a3f428325d1e7cac415121850e04a96a
25273 F20110114_AAASQB carlos_e_Page_024.pro
877861c5e413a9c280bea3f05f6f4b70
58da29c7bcd3fa193e823c35b18c01d009bac6ae
F20110114_AAASPM carlos_e_Page_087.tif
256e391f0ed8b0d2b865cc3114e36278
9d7f345a3725534d970fcd2893d63238a046d1f4
F20110114_AAASOZ carlos_e_Page_055.tif
fce58f11499d1121bf748979f564703c
49304f98062318fb3369642d203cfa6a9468db8a
44073 F20110114_AAASQC carlos_e_Page_025.pro
a6322e1cb80ba5f6e3decb1a34e583ea
54a966d265da1a2edd47e6d59cf223e5577a02cc
F20110114_AAASPN carlos_e_Page_089.tif
7f36e4f5a451398b72d425cc82e634e4
52e7cb7d0acd02d982e629541de039a4c6a297df
31628 F20110114_AAASQD carlos_e_Page_028.pro
ffcb9ef0c289f87001207b7bc4efac6e
f2f891b5300fb19f6c14eeb5ff3b7687052350a4
F20110114_AAASPO carlos_e_Page_093.tif
6b8622e971f6a2fb9f1834037e399d05
b54525807b6e369a4a0edcbff62c058f64677839
46237 F20110114_AAASQE carlos_e_Page_029.pro
d1e086793048d6d6ed16b034df543459
37201d38bd871eb6f814887d768dcb4c2a86fd0b
F20110114_AAASPP carlos_e_Page_095.tif
8a13552163ee6e88c0ec0a6ca9fe8ec7
8040f86792c5e4af69db84003eef69c0ea48409a
60213 F20110114_AAASQF carlos_e_Page_030.pro
9479b8c188b8994cf267f6be586d3ce6
29c9f6e6b6191359d32dc3c4573854837ec7cd50
F20110114_AAASPQ carlos_e_Page_098.tif
79d29486b9be38c42e6450ea1fd7a709
d8103546f87f6d76e55f3fbfe6254a7a3f086279
45846 F20110114_AAASQG carlos_e_Page_036.pro
412a3f9b14f46f453053cde9c38e7c21
f743844d8898273f39534a4b396aa12da4a844c6
F20110114_AAASPR carlos_e_Page_100.tif
418b333ebb3bbd6495e1c73c9f09dc69
5ffa941b555254aa1805c4eac13a9566a7b111aa
10099 F20110114_AAASPS carlos_e_Page_001.pro
84d058ab40c1912365010c0e3174ed6e
4a2cdaba1d003c8223b903af0d58489e1de36811
32168 F20110114_AAASQH carlos_e_Page_038.pro
a9484a811efc282836f5aed0ef3de7ea
8de2735cd18d5cdfa78b3eb0db245a06f7c41e72
1357 F20110114_AAASPT carlos_e_Page_002.pro
12bb8f8e97f37d2ac7b1188d91145fb5
3a8f8580610ede40c3dc851387b6d1e9c76754d6
47599 F20110114_AAASQI carlos_e_Page_039.pro
f2359489b953aa1959346aa4708c83cf
da8751796650a3035c48a942dc759bd2c569a98d
9302 F20110114_AAASPU carlos_e_Page_003.pro
61f18cac9c0051d5e597ac38e09efebd
b89b5bb6c5eec6b6b4f5d881ade438433c0e9321
39770 F20110114_AAASQJ carlos_e_Page_044.pro
c32c02be436000e3f3bdf0d1ce976164
ceaad9567f6f608e0fe32f97d18840ca6ba92e90
59401 F20110114_AAASPV carlos_e_Page_007.pro
365c2168b0ebf64301b6ba79e035e19a
134aab19a6c9f60e689e13b065e17ecb8a437fcd
49631 F20110114_AAASQK carlos_e_Page_049.pro
f92187dc02f750a2993f333c8d0b678a
f3b31c215bcbeb987e4bc5cd0beb23b1748e9767
31010 F20110114_AAASPW carlos_e_Page_015.pro
d85b3d1ca5d98ba830a22fc77f2895e0
b5a13408e4b1b7d897631c820d3527a0877154fa
31681 F20110114_AAASRA carlos_e_Page_086.pro
231474f6765dd6ad8e441fdef0020ee1
c392cc24b86d32447df17f13f12af9c8c57598b4
27178 F20110114_AAASQL carlos_e_Page_051.pro
86d6d9db11b6a8aefec843ed3c3273b4
df421a22c4b256e3a646e1c01e275128b9ff9f72
46698 F20110114_AAASPX carlos_e_Page_018.pro
cc0c6467d14d3b2501ab8ba7b0184e3f
67cbb09a05b455a347e90d4a47f782c7770dc8ea
26819 F20110114_AAASRB carlos_e_Page_087.pro
7958d3bde4ce498c901346686a6647bc
898dcb728a8721ad1b4d5bd8d9bd7dad8cf7be0f
42519 F20110114_AAASQM carlos_e_Page_052.pro
37dc4e58a5a381663461f82e355b1eab
2b6afe89ffc5f11de5ef66a2c5cdf5e42e5b24e9
49606 F20110114_AAASPY carlos_e_Page_020.pro
40985bdf71b9e4a4337682c69132b8ac
d8958a30d999ab564f99f5870f86b2c3dcb794de
25340 F20110114_AAASRC carlos_e_Page_088.pro
111c7ea9e47bda308c410425f21c7735
da3d7548fa6560bda16559129fa9c1b5688cda4a
24258 F20110114_AAASQN carlos_e_Page_054.pro
16d33bb045fb98b0fa49eb2e24676fcc
6820c97e871e30dd14adb79f56610cf1f20f8544
20146 F20110114_AAASPZ carlos_e_Page_021.pro
3ec3bc53ff6b152a76f7fd98be8551f3
146e7a028b9298bc076b4974c2101fd4fbdccd6a
39968 F20110114_AAASRD carlos_e_Page_089.pro
33791842b81ddb92264b6cdca94a0afc
28fcc93cf92f0c3d79326df612582fc98a220f69
50327 F20110114_AAASQO carlos_e_Page_056.pro
157b5166eb234e8bc5c1f59202a25f99
7d9d31659c1682d15857bfeff1caef10caf6606e
14837 F20110114_AAASRE carlos_e_Page_090.pro
64b91fbabc42488621ee20b82c147337
ce94f71970b52934051725487272b5e14818c0bd
50484 F20110114_AAASQP carlos_e_Page_057.pro
ea3daf21892d8a86338cf3da42b05637
d01324c3f05e44a325968efc324aeb500123f34a
28523 F20110114_AAASRF carlos_e_Page_093.pro
3b6d9a0e3cb066de1e40de72df9fe36c
54054d903c676c59ed37446ce1dceaab938d73ff
14054 F20110114_AAASQQ carlos_e_Page_064.pro
06cbaad7657bbc213878d82738ead96d
c86dadf885efb4d99c5b2a1c91de378f1a673ae6
58613 F20110114_AAASRG carlos_e_Page_100.pro
b921311e1d049bc0955c53e4232cfeeb
7b249eafed88f07ad05947ba0de6640297999987
49292 F20110114_AAASQR carlos_e_Page_065.pro
a5ab009bf4f7d4d4fb247ae220799ec4
a19faf35a8e72e474ce42a3a7019e4fb8cb9a8ed
46091 F20110114_AAASRH carlos_e_Page_101.pro
cf82ed755518e2a887192625a090869e
46d9132a92305461fe8eb316318e9ff940643a2b
29811 F20110114_AAASQS carlos_e_Page_066.pro
83e453b1dcdb62fc228028f3e1655dd8
9131b5956078f1221e405e0fe532aff2430d773a
49121 F20110114_AAASQT carlos_e_Page_067.pro
8bb671dbb2151d8c122a6336a1105fbc
6ca337bad60e01486074a7aa55d09e236b6d1a79
28103 F20110114_AAASRI carlos_e_Page_102.pro
31294fd202110bdef965482b6d573d01
73c7c41e8787b40e1243a00fe408651f67ed068b
50000 F20110114_AAASQU carlos_e_Page_068.pro
c17efc8312cd948863686e84693d91de
156f6cf32dbf8cce0a8740fb0c5a842b17b22616
124 F20110114_AAASRJ carlos_e_Page_002.txt
4a4601e66af0a897156920d0cbff1213
a2e551ecd7cac85f78802454f134fe3bf0951dbe
52168 F20110114_AAASQV carlos_e_Page_070.pro
43432cc4bf286dd350cf859ec4b86d78
7415754643d5de54ac2006e30071a03b55d36902
1693 F20110114_AAASRK carlos_e_Page_004.txt
e5691b4f647bd8e6ec50ba4742a03c6e
3fc267c8a8d43a6b7bfef1106c863793b0505ecf
50195 F20110114_AAASQW carlos_e_Page_073.pro
cef75e5d5e334fa88b3aa2ee402687da
ae22c97b492daf9c5fd1853bed747708bff2a4e9
962 F20110114_AAASRL carlos_e_Page_005.txt
d1ff50c003f567f159415b39833b4806
f42f8a0a367d128a88059d42d52ec4032f9cfdde
35054 F20110114_AAASQX carlos_e_Page_077.pro
5ebacc6f48577c5568ef4005b317e593
57a044a482ba255854695e89021878c28a672292
1662 F20110114_AAASSA carlos_e_Page_037.txt
26c6d5cd989bc4e4f5213ac678c4c66d
e2d4befd360936805d7a7f596064d4aaab2193a3
2457 F20110114_AAASRM carlos_e_Page_009.txt
7b0184b581fd948e399ac549b2d1371e
f4e7cea55b97eef9f0e8bf23b6e9e549cfc43e88
45386 F20110114_AAASQY carlos_e_Page_078.pro
8e78c08856e301f9d3dbdf29fdfac8c3
c802a0228711e79eb41ec1ccc3a97bd06e5285df
1467 F20110114_AAASSB carlos_e_Page_038.txt
2254557379f47d34f4c35620ce902df7
abf6541a351cb0bb2b8b149dd3ba877f075c7506
41807 F20110114_AAASQZ carlos_e_Page_080.pro
1a937411be3c56fff30e43fdbf7c4ff8
a8759d4374051605f12b3a3855fac3b097fd2c2d
1648 F20110114_AAASSC carlos_e_Page_042.txt
6217c2462b83bfb1472fc7915232ee58
ce78d7864b006f807d6f3640a97f756870b5a103
651 F20110114_AAASRN carlos_e_Page_010.txt
50ca428eb4df3ab10d45b6f9dbdcdf75
f748dbba57bf7256691cc22be8520ad0be3510b3
2021 F20110114_AAASSD carlos_e_Page_043.txt
97de8035114008c571fabe2f66441c8e
05400235214966970ac140371e299786d56a738c
1741 F20110114_AAASRO carlos_e_Page_011.txt
4aaafcffdf57c6b0d7c00e9528664134
bfb0c992ef8c60cd6a6f146854a9747bbf42275c
1481 F20110114_AAASSE carlos_e_Page_045.txt
3d55f94c3644c8aaec8568268289a8e8
c30e836922e3ce50737a938eb9706c2bc1984637
1806 F20110114_AAASRP carlos_e_Page_013.txt
cd7041c45b9dc855fcfe21254be82d3a
0e5a3235cd5ce7127a06490407e327c711096338
F20110114_AAASSF carlos_e_Page_046.txt
428237e216da785b363cf4f72fc89d4f
3072b0322baf59ff23aa7e4f17bb46f3f95ea948
1865 F20110114_AAASRQ carlos_e_Page_017.txt
008ceb68b8cf98f3431616a1e43c549d
378918e4f4f21267374e395c3f1bea92cb3bf7c5
1836 F20110114_AAASSG carlos_e_Page_055.txt
904bcfafc5473b244ce2bf4db01b4039
7da7e9db2afd92da6d85cef21678886959a0e069
392 F20110114_AAASRR carlos_e_Page_019.txt
3e1065f3c8e8de376c4a741392aec11f
84f7b302d9b9c37396fbec3a0b3e2b740f5ca1cb
2003 F20110114_AAASSH carlos_e_Page_057.txt
22a5a9bea42a6235f986226c6422ddd4
61c07c1ef8a2194008ebc0c72ebfb4804d1d145a
2079 F20110114_AAASRS carlos_e_Page_023.txt
4e7f0fb888770037995b3b3ec87b113e
c2c893c70c31d5d6a75c1cfed39eecc0df69b52f
2158 F20110114_AAASSI carlos_e_Page_059.txt
381aa9e41fcf604c05815960a8d5fffd
d2ec2662e2e7f6441c66aced32444ab1d1a19bae
1073 F20110114_AAASRT carlos_e_Page_024.txt
0c220cf0dffcdcfcdbe64155ae7aba11
f9e956ca697297b42bb9b2aae5a3d3978c475c1f
1840 F20110114_AAASRU carlos_e_Page_026.txt
cb71ccabbc6f7467ef54fa35fc22e670
e06add6b036408144c6573f2ecf29f7f88a94004
2005 F20110114_AAASSJ carlos_e_Page_061.txt
f3d3eaf9f949a9b49eba5536fa372cd2
67ec9cc131282427d71dc0fe204df341a1e8d130
1916 F20110114_AAASRV carlos_e_Page_027.txt
62f80207314a4ef4ff6722948c153bb1
12c9fa403be687cc4eb3728a8495706105b517a8
1977 F20110114_AAASSK carlos_e_Page_068.txt
dd73e3a8344afb49881eaa8ae2322da1
bc8fee94d3a6276a9d83e10687edcd6cfe509a23
1566 F20110114_AAASRW carlos_e_Page_028.txt
c0d6ed32232415b0d885f6767764a3c9
0d60e2512923a01e07efc968ecccfb76280a59e7
20610 F20110114_AAASTA carlos_e_Page_004.QC.jpg
b7c9a10ad95c9f6a4581721485877a85
913bbbf3374685215a81f5fb486b74df5315abe1
2063 F20110114_AAASSL carlos_e_Page_070.txt
a6306de07056b4acfa46d6b311655725
f1463c45451593b206265b8d85bd0aa6899a733b
1851 F20110114_AAASRX carlos_e_Page_029.txt
1079ad4ce131bbbc395874267ac833bf
3768a9c0e86e3e5a83f6b7710cc359a7eaea1617
950668 F20110114_AAASTB carlos_e_Page_004f.jp2
f9e3f6ff0364182b8388f7931a4346e9
77b3378be7a24b14b3b1f2d1d044efa31fbff546
2001 F20110114_AAASSM carlos_e_Page_073.txt
89f23ee1db73d8ebd63a9814883944e2
6e94868ca69d9dc866446781dbcbdfca22013c5e
1941 F20110114_AAASRY carlos_e_Page_033.txt
f5389aa3c9b49f8be285c0dbe12e7fb6
96025b39623956ad01c245c41064eb33611e1d46
14820 F20110114_AAASTC carlos_e_Page_006.QC.jpg
0e5c94124e6ae64163887588a8b82646
31bff6a0196e67e6c6ce73aee83ccdc8e40c96f3
1484 F20110114_AAASSN carlos_e_Page_077.txt
6621fe1944f66fa8dbd24a5b01978342
5bfef3bccd3c4209454e0afb067ec6ad73c61f71
1867 F20110114_AAASRZ carlos_e_Page_036.txt
9cc578c5e9eae8d944e2ecfa827b66b5
f1014f5f6e125d2e5496acb2da69117637cd169f
4174 F20110114_AAASTD carlos_e_Page_007thm.jpg
015989a4b6d0ab1883426cc67f2d2bdc
d66b726735d475d0c12b6a2699bb8d6de5c08fa1
1948 F20110114_AAASSO carlos_e_Page_079.txt
055b43af2bf181c8c41e4fbe197992ea
5d1f6c31959efb387941d2186f18747deef0c6f0
848961 F20110114_AAASTE carlos_e_Page_007f.jp2
203650e37cb2ca28a90ae87a5aa4d0d5
ef0034b5214a695e06bb29d5d87539e38953242b
1961 F20110114_AAASSP carlos_e_Page_080.txt
2e0000fb36a774462d1c3cb0c7a743f2
3ca31583bf7c2a68bf58d351a54e35af77e35e15
8008 F20110114_AAASTF carlos_e_Page_008.QC.jpg
f1a498495346f9e88bba8b13c45d52bc
c90e5a5baf1f79d82e5daed9b4a918bf014180e8
945 F20110114_AAASSQ carlos_e_Page_084.txt
a82c3b15f874ccb92b38c31afbde3872
0202638ed85b18212a26764d5faefaa4b7500201
2356 F20110114_AAASTG carlos_e_Page_008thm.jpg
09568e3dac32d2d6067ab4b1b3648248
cdb30e419849493b9aa1b146b1588b86467f3c30
1374 F20110114_AAASSR carlos_e_Page_087.txt
c0fff12818b4f1e43bb7e73cc9887e29
e8bd2ade831da6d91f08c06103f11f536bc22d17
16048 F20110114_AAASTH carlos_e_Page_009.QC.jpg
4dc9b7bb1eeb0fcbec4c198896f81e1d
3a57dd05fd6169a85e647c61052fe8c6f535a641
873 F20110114_AAASSS carlos_e_Page_092.txt
9d6fd86a430867ac1e5beb2f8e5b6f60
424819f1432c770710eff25357475f8583d72b91
749919 F20110114_AAASTI carlos_e_Page_009f.jp2
0a280c11c7fe1bbb9285b32e9732d70f
c2a16a54ef2f9ef4e328c9964fed60cdca8462b0
1462 F20110114_AAASST carlos_e_Page_093.txt
17cfd30bd68487745ed6d24ef0280d23
66528e922632d11e11b9c239b6b8c2f8707afe93
6034 F20110114_AAASTJ carlos_e_Page_010.QC.jpg
75ab7da0c9bb20bb564fbec8a1675d62
4b1a29dfa62582363c5ae3f7661807093bbddeaa
F20110114_AAASSU carlos_e_Page_095.txt
049537e4ffafd71f29166419346bc834
aa4deca562b79fe66191a68174139f60e659a0a7
2387 F20110114_AAASSV carlos_e_Page_097.txt
debe5712c1b9f258e6b6cb5627b14b9f
456f0411a4ffdde52352b8cb1822a21492660d38
241091 F20110114_AAASTK carlos_e_Page_010f.jp2
43fa820f45f2c7efeff4ffd29df2c52e
d535731e4d2fcac49674c56c6d1407d3669ada41
2232659 F20110114_AAASSW carlos_e.pdf
7a2273dbb67f0d1ad279f0a42347c410
c958e6a5df4c1d1fb063f26091a840f2e9a0ae41
19824 F20110114_AAASTL carlos_e_Page_011.QC.jpg
da987bdb2d67556e9a83e7a1d195cfa8
b8b0de2178cd5e3c66fb9dc0ac56608f98758fa4
947 F20110114_AAASSX carlos_e_Page_002thm.jpg
43ae2e147b7de1fa258554d93a21397d
556a90c2db539e27777e7b179214b4c7abfb857f
25393 F20110114_AAASUA carlos_e_Page_022.QC.jpg
75665180dc450fcedbb0072de2c34276
d65fbfe9a881325c88c643ba431dbddfabf9e115
5312 F20110114_AAASTM carlos_e_Page_011thm.jpg
82cc48444f080f1f2a65e30e3f987db1
8e7e8aa02897db81b5ad31c736073e9274adcbc0
32062 F20110114_AAASSY carlos_e_Page_002f.jp2
724e3480cb810adc8cff5c123f189cf4
5c2df026d0b99fef84860428cbe29c1b7522f6d5
6628 F20110114_AAASUB carlos_e_Page_022thm.jpg
c56d3bf6f9aaf8c6b54a6c91dbed4cf1
f8a55504d6fe4fd0a24568bd137dee6e77695b59
927704 F20110114_AAASTN carlos_e_Page_011f.jp2
23e48c3dec0dd05b7e5444eb0cd07290
15879e0d18bb875eb77d2a997f9a4b9807da46f0
5321 F20110114_AAASSZ carlos_e_Page_003.QC.jpg
2f6797a3228f77abe685717539da7087
b48f80c5d8fab9795bccf3784f234085d1bc3d0c
F20110114_AAASUC carlos_e_Page_022f.jp2
b0cf476cfb62f4b316a1a669ef9ae900
56f2d3b89da36caa43731e4359069e83c1cd90f9
5639 F20110114_AAASTO carlos_e_Page_013thm.jpg
88974e4953662a5db57625fbfc915992
0818032ac9b1cfce1c84d88dd889203655795583
1051913 F20110114_AAASUD carlos_e_Page_023f.jp2
1203e759ccd485c002ec02327804492c
3aa8164cd5c7e872e5c5dad57cfb1ce191cbadf1
23061 F20110114_AAASTP carlos_e_Page_014.QC.jpg
cd87d26d2793c71d2f453f1d7c404b69
107a8a4fd1c711bfca3f8b0c443058d4a800f1e0
13593 F20110114_AAASUE carlos_e_Page_024.QC.jpg
635a0720142e4164ecc8624f78c0d013
d2354c07ddb5a9ec3f5402a156964562df949b3b
6178 F20110114_AAASTQ carlos_e_Page_014thm.jpg
5d318fa530d60a9ecb2ccb77f09be2cc
67f3ccd91b60d66b8b537d7981bb440ec1a9d358
3913 F20110114_AAASUF carlos_e_Page_024thm.jpg
0f53d756608488051dda811e95b3ed8c
5295ad0bce9d8a70a191399201188fc5743b588b
1031575 F20110114_AAASTR carlos_e_Page_014f.jp2
fb0e3e14864fa46fc8decad98ef95d23
3c018d65e0ca30b053050368a65de9f1435a7c08
22395 F20110114_AAASUG carlos_e_Page_025.QC.jpg
23e6a03bd036ecbfceed475aa4823a5c
9735ec23861f1832fb896297be099e6cfc563e5c
5118 F20110114_AAASTS carlos_e_Page_015thm.jpg
be4dbe4ce92b16b1112230cd7d112e9e
05a1eef7df3e41f3be0aa874aa002314cfffe814
5896 F20110114_AAASUH carlos_e_Page_025thm.jpg
254bb44218edde3ce4c9943b04c121a2
0ba5cb40fd46c23695e0c3cc470542b9c950b9b4
661351 F20110114_AAASTT carlos_e_Page_016f.jp2
f6dfd810087bd8f192b9cef43639c462
5a4bde40a6f01cb707bf6ffe411a697f52c74b35
22550 F20110114_AAASUI carlos_e_Page_026.QC.jpg
9257d245c4a3538ff58f9ba4a5fe4c46
e2cf3a25bfb2b9cca852f90a7d871e307caa2f0a
5991 F20110114_AAASTU carlos_e_Page_018thm.jpg
5375adc94c3bb07b371e01db16be98ef
c9948c99dc9100a3d4b60c0d45a08392a84d386c
6278 F20110114_AAASUJ carlos_e_Page_026thm.jpg
7542fceff298009a0270d5569c61d8c2
517bc8f253776c328026a3cf9cd9a132c6c34d8e
1051880 F20110114_AAASTV carlos_e_Page_019f.jp2
cbe8b810eb073f97bf46edf7a8abdcf7
d1f8532286a86502b696199ec56af22378ebe4a8
5373 F20110114_AAASUK carlos_e_Page_027thm.jpg
639b5a6b3bc768d04108a74b2c5ea92f
8d4863004d0e85a1333fa439ffdbd72835b64b89


Permanent Link: http://ufdc.ufl.edu/UFE0006602/00001

Material Information

Title: Transcriptional profiling on trees affected by citrus blight and identification of an etiological contrast potentially associated with the disease
Physical Description: Mixed Material
Language: English
Creator: Carlos, Eduardo Fermino ( Dissertant )
Moore, Gloria A. ( Thesis advisor )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Plant Molecular and Cellular Biology thesis, Ph.D
Dissertations, Academic -- UF -- Plant Molecular and Cellular Biology

Notes

Abstract: Citrus blight is an important citrus disease in Florida (USA) and São Paulo (Brazil), primarily affecting yield in adult plants and compromising maintenance of entire commercial blocks. Blight is associated with rootstock choice, but, at present, is of unknown etiology. Therefore, the objectives of this study were to identify differentially transcribed genes and uncover etiological contrasts under non and citrus blight conditions. Roots of healthy and blighted Rough lemon (Citrus jambhiri Lush) rootstock supporting Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) canopy were collected from a central Florida area. Total RNA was obtained and RT-PCR was performed enriching for messenger transcripts. Subtracted cDNA libraries were created and around 140 clones were arrayed onto nylon membranes. Independent RNA sources from feeder roots of ten healthy and ten blighted trees were labeled with P33, hybridized overnight with the membranes and analyzed under an imager system. Selected clones were validated by RT quantitative real time PCR. The results indicated that citrus blight was able to affect the transcriptional levels of certain genes in a similar pattern among different replicates. The level of response was dependent of the assessed group of plants. One of the clones had sequence similarities to a citrus EST and to a potential ubiquitin subunit. Another one had similarities to a citrus chitinase, helping to deduce a candidate sequence for the blight associated P5 gene. Three genes had higher transcriptional levels under blight condition and did not respond to cold and drought stresses. The blight associated P12 had higher levels in mildly than in fully blighted trees. Further characterization of these genes may contribute to the understanding and control of citrus blight. In another experiment, citrus tristeza virus (CTV) genes were observed in the libraries. The transcriptional level of the P27 gene (divergent coat protein gene) of CTV was far more abundant in roots of blighted than in healthy Carrizo citrange, which is considered to be resistant to variant forms of CTVs. It remains to be investigated if CTV causes or enhances blight, or only grows better in feeder roots of already affected trees.
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 103 pages.
General Note: Includes vita.
Thesis: Thesis (Ph.D.)--University of Florida, 2004.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0006602:00001

Permanent Link: http://ufdc.ufl.edu/UFE0006602/00001

Material Information

Title: Transcriptional profiling on trees affected by citrus blight and identification of an etiological contrast potentially associated with the disease
Physical Description: Mixed Material
Language: English
Creator: Carlos, Eduardo Fermino ( Dissertant )
Moore, Gloria A. ( Thesis advisor )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Plant Molecular and Cellular Biology thesis, Ph.D
Dissertations, Academic -- UF -- Plant Molecular and Cellular Biology

Notes

Abstract: Citrus blight is an important citrus disease in Florida (USA) and São Paulo (Brazil), primarily affecting yield in adult plants and compromising maintenance of entire commercial blocks. Blight is associated with rootstock choice, but, at present, is of unknown etiology. Therefore, the objectives of this study were to identify differentially transcribed genes and uncover etiological contrasts under non and citrus blight conditions. Roots of healthy and blighted Rough lemon (Citrus jambhiri Lush) rootstock supporting Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) canopy were collected from a central Florida area. Total RNA was obtained and RT-PCR was performed enriching for messenger transcripts. Subtracted cDNA libraries were created and around 140 clones were arrayed onto nylon membranes. Independent RNA sources from feeder roots of ten healthy and ten blighted trees were labeled with P33, hybridized overnight with the membranes and analyzed under an imager system. Selected clones were validated by RT quantitative real time PCR. The results indicated that citrus blight was able to affect the transcriptional levels of certain genes in a similar pattern among different replicates. The level of response was dependent of the assessed group of plants. One of the clones had sequence similarities to a citrus EST and to a potential ubiquitin subunit. Another one had similarities to a citrus chitinase, helping to deduce a candidate sequence for the blight associated P5 gene. Three genes had higher transcriptional levels under blight condition and did not respond to cold and drought stresses. The blight associated P12 had higher levels in mildly than in fully blighted trees. Further characterization of these genes may contribute to the understanding and control of citrus blight. In another experiment, citrus tristeza virus (CTV) genes were observed in the libraries. The transcriptional level of the P27 gene (divergent coat protein gene) of CTV was far more abundant in roots of blighted than in healthy Carrizo citrange, which is considered to be resistant to variant forms of CTVs. It remains to be investigated if CTV causes or enhances blight, or only grows better in feeder roots of already affected trees.
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 103 pages.
General Note: Includes vita.
Thesis: Thesis (Ph.D.)--University of Florida, 2004.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0006602:00001


This item has the following downloads:


Full Text












TRANSCRIPTIONAL PROFILING ON TREES AFFECTED BY CITRUS BLIGHT
AND IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY
ASSOCIATED WITH THE DISEASE














By

EDUARDO FERMINO CARLOS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Eduardo Fermino Carlos
































This work is dedicated to my wife Darlene, to my daughters Luisa and Ligia, to my
parents Norival and Maria, to my aunt Josefa, and to all my friends and family that shared
their love and care. Along the way, many things were surpassed and our dear aunt Neula,
uncles Otavio and Waldemar, and friend Beatriz Nielsen will always be in our hearts.















ACKNOWLEDGMENTS

I thank my family and friends for the unconditional support and encouragement

during all the steps of the program. I thank CNPq, a Brazilian Federal Agency, for

sponsoring most of my graduate studies abroad and for their overall attention with the

fellows. I thank Dr. Gloria Moore, my major professor, for the outstanding support

during the whole process.

During my academic and professional life I have been fortunate to meet people that

helped me not only as scientific mentors but also as examples of positive attitudes. I

especially thank my friends Eliana Lemos, Luiz Carlos Donadio, Marcos Machado, Julia

Beretta, Ken Derrick, Gene Albrigo, Steve Futch and Jose Eduardo Lima, among others. I

also thank Dr. Chris Chase and Dr. Nancy Denslow for their intelligent insights and

enthusiasm, which always brought positive perspectives about the ongoing work and

essential source of stimulus. I thank Dr. Ken Cline for his suggestions and scientific

argumentation.

A regular student life normally requires patience, patience and dedication. Fifield

Hall demanded the same efforts, and more. I thank all my friends that shared their care

and friendships. I specially want to thank Gary Barthe, Mukaddes Kayim, Marty

Dekkers, Abdul Al-Saadi, Basma El Yacoubi, Ufuk Koca, Karen Champ, Kim Niblett,

Manjunath Keremane, Methap Sahin, Vicente Febres, Abeer Khalaf, Maureen Meyerson,

Juliana and Gustavo Monge, Ricardo Harakava, Cristina Moreira, Dirceu Mattos, Maria

Luisa Targon, Eduardo Stuchi, Otavio Sempionato, Jose Antonio da Silva, Aline and









Marcelo Carvalho, Georgia and Jose Carlos Dubeux, Livia and Steel Vasconcelos, Jose

Carlos Rodrigues, Juliana and Flavio Silvestre, Camila and L. Augusto de Paula, Ana

Carolina and Vitor Lira, Carolina and Antonio Martins, Eliezer Martins, Maria de Lurdes

Mendes, Marisol d'Avila, Tania Querido, Nancy Velasquez, David Moraga, Li Liu, Bill

Farmerie, Jason Blum, Natalia, Duy Nguyen, Patrick Larkin, Jannet Kocerha, Vishal

Patel, Mike McCaffery, Carole Dabney-Smith, Fabien Garard, Richard Berger, Denise

Tombolato, Fabricio Rodrigues, Gisele Martins, Angela and Wayne M. Jurick II,

Adrianna Castafieda, Ronald French and many other friends. There is no end to this list...

Thanks to all my friends.

I started in the Department of Plant Pathology and I thank Dr. Bill Zettler and Dr.

George Agrios for the attention and dedication that I received during the application

process and as a new graduate student.
















TABLE OF CONTENTS



A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES .............................................. viii

LIST OF FIGURES ........................................ .............. ix

ABSTRACT .................................................... ................. xi


CHAPTER

1 IN TRODU CTION .................................................................. .. ...... ... ............... 1

Im portance of C itriculture ....................... ........................................................1......
C itrus in Sdo P aulo and F lorida ...................................... ...................... ...............3...

2 IMPORTANCE OF CITRUS BLIGHT AND REASONS FOR ITS CONTROL .......5

C itru s B light E tiology .. ......... ............... .......... ................................. ... ... 6
Citrus Blight is Affected by the Employed Rootstock .........................................10
O b je ctiv e s ............................................................................................................... .. 12

3 BUILDING THE CITRUS BLIGHT SUBTRACTED LIBRARIES .....................13

R oot Sam ples and cD N A Synthesis ........................ ....................... .....................13
Building the Suppressive Subtracted cDNA Libraries...........................................14
Sequence Analysis and Selection of Clones ................ ................................... 17

4 IDENTIFICATION OF DIFFERENTIALLY TRANSCRIBED GENES
UNDER NON AND CITRUS BLIGHT CONDITIONS ................ ..................... 20

E xperim mental D design ............... ................ .............................................. 21
A rray D esign and Spot Features............................................................ ................ 2 1
Sam ple Preparation and Labeling.................. .................................................... 22
H ybridizations and P aram eters .............................................................. ................ 24
M easurem ents and Specifications.......................................................... ................ 24
N orm alization and C controls ................................................................... ................ 26
The Visual Evaluation of the M em branes ............................................. ................ 27









T he C lu ster A naly sis ... ..................................................................... .. .......... ... 29
The C contrast of M eans A nalysis............................................................ ................ 31
Com paring the Evaluation M ethods ...................................................... ................ 36
Combining the Results for the Selected Clones .................................... ................ 36

5 THE RELATIVE TRANSCRIPTIONAL RESPONSES OF THE SELECTED
GENES UNDER DIFFERENT INCIDENCES OF CITRUS BLIGHT .................43

Quantitative Real Time PCR was the Method of Choice......................................43
The Selected Clones and the Characteristics of the Probes...................................45
Collecting and Preparing the Sam ples................................................... ................ 49
Reverse Transcriptase (RT) and PCR Reactions................................... ................ 50
Relative Quantification of the Transcriptional Levels of the Selected Genes............51
The Potential Biological M meaning of the Clone 109 ............................. ................ 57
A Tentative Test to Verify the Effect of Cold and Drought Stresses...................... 58
The Potential Effect of Redundancy and Gene Families.......................................61

6 IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY
ASSOCIATED WITH THE CITRUS BLIGHT DISEASE ..................................... 63

The First Screening of the Subtracted Libraries.................................... ................ 64
Other CTV Genes were also Found in both Subtracted Libraries...........................68
Quantitative Evaluation of the P27 Candidate Gene in the Blighted Trees ...............68

7 C O N C L U SIO N S .................................................. .............................................. 7 1


APPENDIX:

BLAST ANALYSIS BASED ON NUCLEOTIDE SEQUENCES .......................73

L IST O F R E F E R E N C E S ...................................................................................................84

BIO GR APH ICAL SK ETCH .................................................................... ................ 91















LIST OF TABLES


Table page

4-1. Transcriptional pattern of the selected clones. .................................... ................ 37

4-2. Blast search analysis for som e clones ................................................... ................ 38

5-1. Characteristics of the chosen probes. ............................................... ................ 47

6-1. The clone E8-13 had homolog sequences matching different isolates of the
citrus tristeza virus (C T V ) ....................................... ........................ ................ 66

6-2. Other sequences with homology to CTV genes found in the blighted minus
healthy (B-H) and healthy minus blighted (H-B) subtracted libraries ..................68

A-1. Blast analysis based on nucleotide sequences of each individual clone.................73















LIST OF FIGURES

Figure page

1-1. C itrus around the w orld ...................................................................... ...............2...

1-2. The citrus belts in Florida and Sdo Paulo ............................................... ...............3...

2-1. Characteristics and sym ptom s of citrus blight......................................... ...............7...

2-2. T ransm mission of citrus blight ....................................... ........................ ...............9...

3-1. cDN A synthesis enriching for m RN A ................................................... ................ 15

3-2. Making the subtracted "healthy" and "blighted" libraries.................................... 16

3-3. Analyses performed for each sequenced clone............... .....................................18

4-1. Preparing individual clones for the cDNA array. ............................. ................22

4 -2 A rray d e sig n ............................................................................................................. 2 3

4-3. Quality param eter for each m em brane ................................................. ................ 26

4-4. V isual evaluation of the m em brane pairs ............................................. ................ 28

4-5. Transcriptional profiling of the candidate genes. ............................. ................30

4-6. Contrast of m eans analysis of selected clones....................................... ................ 35

4-7. Blast-X of the clone 6 using the non-redundant protein database...........................37

4-8. Clone 38 has sequence homology to citrus chitinases ..........................................39

4-9. Transcriptional profiling of the clone 38 (a chitinase homolog) and other
clo n e s.................................................................................................... ....... .. 4 0

4-10. The clone 109 had sequence homology to citrus ESTs. .....................................41

5-1. Designing a candidate sequence for the P5 gene. .................................................46

5-2. Relative PCR efficiency plot of the clone 109 against the normalizer 18S. .............48









5-3. Exam ples of real tim e PCR reactions ................................................... ................ 52

5-4. Relative transcriptional levels of the selected genes. ..........................................54

5-5. Contrasts comparing the effect of cold and drought treatments on the relative
transcriptional level of the clone 109 and other clones. .....................................60

6-1. The virtual northern blot of the B H clones. ..................................... ................ 65

6-2. The transcripts of the p27 candidate gene were abundant in the roots....................69















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

TRANSCRIPTIONAL PROFILING ON TREES AFFECTED BY CITRUS BLIGHT
AND IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY
ASSOCIATED WITH THE DISEASE

By

Eduardo Fermino Carlos

August 2004

Chair: Dr. Gloria A. Moore
Cochair: Dr. Kenneth S. Derrick
Major Department: Plant Molecular and Cellular Biology

Citrus blight is an important citrus disease in Florida (USA) and Sao Paulo (Brazil),

primarily affecting yield in adult plants and compromising maintenance of entire

commercial blocks. Blight is associated with rootstock choice, but, at present, is of

unknown etiology. Therefore, the objectives of this study were to identify differentially

transcribed genes and uncover etiological contrasts under non and citrus blight

conditions. Roots of healthy and blighted Rough lemon (Citrus jambhiri Lush) rootstock

supporting Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) canopy were

collected from a central Florida area. Total RNA was obtained and RT-PCR was

performed enriching for messenger transcripts. Subtracted cDNA libraries were created

and around 140 clones were arrayed onto nylon membranes. Independent RNA sources

from feeder roots of ten healthy and ten blighted trees were labeled with P33, hybridized

overnight with the membranes and analyzed under an imager system. Selected clones









were validated by RT quantitative real time PCR. The results indicated that citrus blight

was able to affect the transcriptional levels of certain genes in a similar pattern among

different replicates. The level of response was dependent of the assessed group of plants.

One of the clones had sequence similarities to a citrus EST and to a potential ubiquitin

subunit. Another one had similarities to a citrus chitinase, helping to deduce a candidate

sequence for the blight associated P5 gene. Three genes had higher transcriptional levels

under blight condition and did not respond to cold and drought stresses. The blight

associated P12 had higher levels in mildly than in fully blighted trees. Further

characterization of these genes may contribute to the understanding and control of citrus

blight. In another experiment, citrus tristeza virus (CTV) genes were observed in the

libraries. The transcriptional level of the P27 gene (divergent coat protein gene) of CTV

was far more abundant in roots of blighted than in healthy Carrizo citrange, which is

considered to be resistant to variant forms of CTVs. It remains to be investigated if CTV

causes or enhances blight, or only grows better in feeder roots of already affected trees.














CHAPTER 1
INTRODUCTION

The importance of one industry in the agriculture segment and the context where

science exists to help society were considered in this study. The result was a true

excitement not only because the focus of this research is (still) an unresolved real

problem of citrus, but also because the search for information involved molecular biology

and related fields.

Importance of Citriculture

In addition to tasting good, citrus is a well known source of vitamin C and

antioxidants. It is also believed to have anti-cancer properties (Rafter, 2002), and

processed peel and pulp are frequently used for animal feed (Fegeros et al., 1995).

Because of its nutritional relevance, citrus is an important industry world-wide, raising

economies at macro and local levels by supporting social development directly with jobs

and with secondary industries and services. Citrus is commercially present in more than a

hundred countries in all five continents, primarily within tropical and subtropical regions.

The total production has consistently increased in the last 40 years, and, more recently,

reached one hundred million tons yearly (Figure 1-1). The early citrus pioneers would

probably be astonished by this level of production, but the modern industry has through

the years achieved efficiency, flexibility and high quality standards.

Citrus was first brought to the new world by the Portuguese and Spanish explorers

at the beginning of the XVI century (Moreira, 1980; Allen, 2000). By the second half of

the XIX century, the USA and Brazil had established fresh fruit companies, and the











frozen concentrated technology, developed in the 1940s in Florida (Lewandowski, 2000),

increased the demand for citrus juices.


A) Total production of citrus from 1961
to 2000


120

1 00



60 -

40 -


20

61 65 70 75 80 85 90 95 00


* Brazil
USA
China
*Spain
*Others


Years


B) 102 million tons of citrus were produced
in 2003
Brazil
o United States of America
14% 8% 0 China
4 U Mexico
ao Spain
1% 11 9 India
2% rtU Iran Islamic Rep of
2% 4wq [] Nigeria
o2 9% Italy
2% 13% Egypt
2% / Argentina
3% 5 Turkey
U Pakistan
3% South Africa
4% 5 12% Japan
4% U Greece
6% 6% U Morocco
o Thailand
Others


Figure 1-1. Citrus around the world. A) The total production of citrus has increased
almost 5 fold in the last 40 years. Brazil took the lead in production over the
USA only after the Florida freezes during the 1980s. B) Citrus is
commercially present in 138 countries according to FAO, with China, Mexico
and Spain ranking respectively 3rd, 4th and 5th in total production in 2003.
(Source: FAOSTAT data, 2004, http://apps.fao.org, last accessed April 07,
2004).




Many changes in production systems have been necessary to meet the ongoing

needs of growing markets and the demands of new challenges, such as changes in

organoleptical concerns, unexpected drought and cold stresses, outbreaks of pests and

diseases and raises of political barriers, to name just a few. Consequently, production

constraints have been overcome by the use of grafted plants to replace seedlings, changes

in rootstocks, selection of new cultivars, relocation of production fields and charges in

diplomatic battles (Fawcett and Lee, 1926; Moreira, 1980; and others). Literature is

available for all transitions suggesting that the need for research in citriculture is not a









recent event. In fact, despite frequent new challenges, in the last decades citrus has

surpassed other important fruit crops such as bananas, apples and grapes in production

according to FAO (Source: FAOSTAT data, 2004, http://apps.fao.org, last accessed

April 07, 2004).

Citrus in Sao Paulo and Florida

The two major citrus producing areas in Brazil and the USA are the states of Sao

Paulo and Florida (Figure 1-2).


A) Florida


B) Sao Paulo


CENTRAL
FLORIDA


Figure 1-2. The citrus belts in Florida and Sao Paulo. A) The industry in Florida has
moved to warmer southern areas after the freezes of the 1980s, and the major
producers are nowadays Polk and Hendry counties, respectively, in the central
and southwestern regions. B) In Sao Paulo, the northern region is drier and
warmer than the southeastern areas and the major citrus companies have their
headquarters around Araraquara city.



Supported by good environmental conditions, efficient systems and high quality

products, both Florida and Sao Paulo have industries primarily devoted to processed

citrus juice, especially oranges, supplying most of the world demand for this commodity.









While Florida has focused its efforts on the North America market, Sao Paulo dominates

the European one. The social and economic consequences of abrupt changes in the citrus

industry of both Sao Paulo and Florida would affect, at least to a certain extent, the

around 400,000 jobs and 3 billion dollars in exports yearly in Brazil, and the 89,000 jobs

and 4 billion dollars in sales in the United States. More details and other statistics can be

seen at ABECITRUS (Source: http://www.abecitrus.com.br, last accessed April 07,

2004), FUNDECITRUS (Source: http://www.fundecitrus.com.br, last accessed April 07,

2004), CREC (Source: http://www.lal.ufl.edu, last accessed April 07, 2004), and USDA

(Source: http://www.nass.usda.gov/fl, last accessed April 07, 2004).

Both socially and economically, there are similarities and also differences in citrus

industries worldwide. However, all contribute to supply different markets with

distinguished quality fruits: citrus fruits. Therefore, citrus is an important segment of

agriculture and demanding of research, which, ultimately, may help in the stabilization of

society.














CHAPTER 2
IMPORTANCE OF CITRUS BLIGHT AND REASONS FOR ITS CONTROL

There are many constraints to citrus production around the world and citrus blight,

in Portuguese "declinio", is one of them. It has caused impressive annual losses of around

60 million dollars in Sdo Paulo and 100 million dollars in Florida (Derrick and Timmer,

2000). Since its appearance in Florida during the second half of the 1800s, it has been a

top concern for the Florida citrus industry, probably only approached by the damage

caused by freezes in the past. Swingle and Webber mentioned that blight was first

reported by Underwood in 1891, and also, that affected trees had been seen two decades

earlier (Swingle and Webber, 1896). Symptomatic plants undergo a slow tree decline,

rather than a sudden disruption in development, as suggested by the given name in

English.

Citrus blight has distinct characteristics not observed in declines caused by other

diseases, such as Phytophthora sp., Tristeza, and others. Early symptoms, which occur in

mature productive trees normally older than five to six years, include loss of the intense

color of leaves and other green tissues; the expression of specific proteins (Bausher,

1990; Derrick et al., 1990; Taylor et al., 1996; Paiva et al., 1997); and complexation and

translocation of zinc from leaves to trunk bark tissues (Albrigo and Young, 1980);(Taylor

et al., 2002). Affected plants have their xylem vessels gradually blocked by amorphous

and filamentous particles (Cohen et al., 1983), which probably leads to the visual

symptoms of overall decline: smaller tree size, twig die back, smaller yield and fruit size

and general mineral unbalance (with the clear pattern for zinc deficiency on leaves).









Major physiological changes also occur (Albrigo et al., 1986), such as off season

flowering, longer flowering periods and shooting of sprouts inside of the canopy, as if the

affected plant was attempting to keep its vital functions (under normal levels). It is

common to have less water uptake into the trunk of the affected plants due to xylem

blockage (Cohen, 1974; Lee et al., 1984). Affected plants rarely die but are often pushed

out and replaced by a young tree. Thus, not only direct losses because of tree failure and

replacement, but other costs in grove management are always incurred. Reset groves pose

serious difficulties in horticultural maintenance programs, since they now have trees of

different ages, sizes and possibly on different rootstocks, all with different needs. They

have also increased the risk of other diseases that can be spread on the new young trees.

The Figure 2-1 displays the described progress of citrus blight often seen in Florida and

in Sao Paulo groves.

In addition to the noted annual losses, a simple and dramatic way to see the impact

of citrus blight is by just driving through the citrus areas in Florida and Sao Paulo and

paying attention to symptomatic trees and reset groves.

Citrus blight also occurs in many other countries (Wutscher et al., 1980), except

those with Mediterranean and desert climates where it has not yet been reported.

Citrus Blight Etiology

Citrus blight is an unresolved problem and its origin and etiology are still matters

of debate. Several characteristics of a plant disease, meaning the outcome between a host

and a pathogen under a proper environment, are present in affected trees, except the

presence of a confirmed pathogen. However, it is worthy to note that (Agrios, 1997)





















& UNF


t .. k


rZMt. .


Figure 2-1. Characteristics and symptoms of citrus blight. A) Zinc deficiency on leaves.
B) Water uptake test into the trunk of the affected plant. C) Xylem blockage.
D) Overall plant decline. E) Off season flowering. F) A reset grove has trees
with different ages and needs.









defines plant disease as: "....whenever the ability of the cells of a plant or plant part to

carry out one or more of these essential functions is interfered with by either a pathogenic

microorganism or an adverse environmental factor, the activities of the cells are

disrupted, altered, or inhibited, the cells malfunction or die, and the plant become

diseased...". Thus, this study considers blight as a plant disease of citrus. In addition, the

process seems to be infectious in nature (Timmer et al., 1992), since transmission by root

grafting was achieved in experiments done in the USA (Tucker et al., 1984), in South

Africa (Marais and Lee, 1990) and in Brazil (Rossetti et al., 1991). However, the etiology

remains unconfirmed and transmission has not occurred by either grafting canopy

branches (Albrigo et al., 1993), by bud grafting, or by soil replacement (Timmer and

Graham, 1992). Figure 2-2 summarizes the experimental transmission of citrus blight

obtained by root but not by canopy grafting.

Citrus blight is also present in mature seedling trees, eliminating the possibility of

scion/rootstock incompatibility as a cause. The pattern of spreading seems to be towards

adjacent trees in the same planted row, rather than across rows, as often noted by farmers

and also by (Swingle and Webber, 1896), who first raised the hypothesis of infectious

spreading pattern. Over time, the progression of the disease seems not epidemic,

following a pattern of incidence that can be described by a linear model, which is closer

to abiotic abnormalities (Berger, 1998). (Swingle and Webber, 1896) recommended

eradication of affected trees, fearing further spread of the problem. That was not

implemented. Since then, candidates for causal agent and other theories have been

examined, such as a non-parasitic origin (Rhoads, 1936), the initial transmission trials









A) ALBRIGO et al., 1993


*o


Receptor trees:
0 soil barrier, total physical exclusion
:soil barrier and limb grafted --...----........
no soil barrier, root grafted ............
no soil barrier, no root grafts


Figure 2-2. Transmission of citrus blight. A) Layout of the work done by (Albrigo et al.,
1993), where the receptor trees that were root grafted to blighted trees
displayed symptoms 6 years after that, while no limb grafted plant displayed
symptoms during the same period; B) Receptor trees used by (Rossetti et al.,
1991) showing symptoms of blight after the root grafting work done in Brazil,
confirming that blight is transmitted by roots as first demonstrated by (Tucker
et al., 1984).


r..









(Cohen, 1968), soil born candidate pathogens (Nemec, 1994), the vascular limited

bacterium Xylellafastidiosa (Hopkins, 1987), nutritional related factors (Wutscher and

Hardesty, 1979), and others, but all remain unconfirmed (Derrick and Timmer, 2000).

Citrus Blight is Affected by the Employed Rootstock

Resistance to citrus blight is dependent upon rootstock choice, and even though

rootstock replacement has been a necessity for decades, the overall solution for citrus

blight has not yet been achieved. Resistance to citrus blight is apparently less common

than initially thought (Young et al., 1982).

Rough lemon (Citrus jambhiri Lush), having excellent yield and drought tolerance

but susceptible to citrus blight, was once the major rootstock in central Florida. Extensive

plantings on this rootstock in the 1940s, replacing sour orange (Citrus aurantium L.), was

probably the reason blight has ranked as one of the most common citrus diseases in the

state since then. Volkamer Lemon (Citrus volkameriana Ten. and Pasq.) was another

rootstock option in the past for similar reasons. However, replacement of both with other

rootstocks more resistant to citrus blight contributed to major changes in citriculture,

since the replacements were normally less vigorous and sustained lower yields (Young et

al., 1982). Among other consequences were increases in grove maintenance needs such

as more attention to grove fertilization, to tree density per area, to other disease

susceptibility, and to increases in irrigation, which ultimately contributed to the urbanist

concerns in water usage in Florida (Callies, 2000).

In Florida, replacement of Rough lemon (Citrus jambhiri Lush) by Carrizo citrange

(Citrus sinensis L. Osb. x Poncirus trifoliata L. Raf.) was not as effective as anticipated

because Carrizo proved to be almost as susceptible to blight as Rough lemon (Young et

al., 1982). Sour orange (Citrus aurantium L.) never matched Rough lemon yield in









central Florida, and it also succumbs to citrus tristeza virus. Cleopatra mandarin (Citrus

reticulata Blanco) yields fruit later than Rough lemon and when in full production the

fruit is smaller and the yield lower, primarily when supporting sweet orange and

grapefruit cultivars. Sweet orange (Citrus sinensis L. Osbeck) and Sunki tangerine

(Citrus sunki L.) are considered more resistant to citrus blight, but are not widely used

because of high susceptibility to gummosis of Phytophthora sp. and also to drought.

Swingle citrumelo (Citrus paradise Macf. x Poncirus trifoliata L. Raf.), largely used

nowadays, will probably be of restricted use in the near future because some groves on

this rootstock have become affected by citrus blight. In Sao Paulo, rootstock replacement

is also an unresolved issue, with no obvious alternative to Rangpur lime (Citrus limonia

Osbeck), another rootstock susceptible to citrus blight, and more recently to citrus sudden

death disease. Able to grow on non-irrigated areas, on soils with high aluminum, and in

low input production systems, Rangpur lime is still the rootstock of choice in more than

80% of groves in the Sao Paulo area, despite citrus blight losses (Pompeu JR., 2001).

Thus, groves that normally would last naturally for 50 years or more are being replaced

within 10 to 15 years, or less. The consequence is a fast and expensive turnover of trees

for big companies and the end of business for medium and small farmers. Just the return

on investment in citrus normally takes 8 to 9 years after planting.

Therefore in all aspects examined, citrus blight is an important concern for citrus

industries, wherever it happens. The complexity and importance of citrus blight demands

broad inter-institutional scientific cooperation on efforts to examine its etiology and gain

insights into what causes the plant to block its own xylem and decline, what is

transmitted from plant to plant (and why only by roots), how the plant responds and what









is changed in the metabolism of the affected plant, why young non-bearing plants are

normally not affected, and more. But these are overwhelming questions. The present

study, more conservatively, seeks to uncover molecular and etiological contrasts between

healthy and blighted trees.

Objectives

The major objectives of this study were to identify differentially transcribed genes

under non and citrus blight conditions and contribute to its understanding and control,

which is the long term goal of the citrus blight research at this institution.

The specific objectives were

* to build subtracted cDNA libraries from root samples of non and blighted trees.

* to identify differentially transcribed genes using cDNA arrays.

* to obtain quantitative information about the transcriptional level of the selected
genes using reverse transcriptase real time PCR.

* to evaluate the presence of potential causal agents found in the subtracted libraries.














CHAPTER 3
BUILDING THE CITRUS BLIGHT SUBTRACTED LIBRARIES

For research on citrus blight, several promising technologies were considered.

Genome wide approaches and microarrays have offered great perspectives in several

biological systems, but sequence information is limited for citrus. Thus, cDNA

subtractive hybridization (Diatchenko et al., 1996) was undertaken, with the objective to

obtain enriched cDNA libraries for each considered condition: healthy and blighted

plants. Subtracted libraries can uncover genes up or down regulated under a specific

condition, thus being a useful source of candidate clones for further studies. Similar

approaches taken with rice (Xiong et al., 2001) and soybean (Colebatch et al., 2002)

encourage such efforts, because genes differentially regulated under disease pressure and

protein synthesis responses were identified.

Root Samples and cDNA Synthesis

Since blight dissemination is associated with the root system (Timmer et al., 1992)

and morphological changes are seen in the roots of affected plants (Lindbeck and

Brlansky, 1998), molecular responses are also expected to happen; therefore, root tissues

were chosen for subsequent procedures. Superficial roots of Rough lemon (Citrus

jambhiri Lush) rootstock supporting Valencia sweet orange (Citrus sinensis L. Osbeck

cv. Valencia) canopy were collected and total RNA was extracted the same day from

feeder root tissues using Qiagen RNeasy protocol, with DNase digestion (Source:

http://wwwl.qiagen.com, last accessed July 15, 2001). It was done during the summer of

2001. Rough lemon was chosen because of its known susceptibility to citrus blight.









Feeder roots, also called fibrous roots, of around 2 to 4 mm in diameter, were chosen

because they display xylem plugging and internal anatomic differences in affected trees

compared with healthy ones (Lindbeck and Brlansky, 2000), similar to what happens in

canopy tissues, suggesting that molecular contrasts may also be present. Samples were

taken from healthy and fully blighted trees previously diagnosed by typical visual

symptoms, by the water uptake test (Lee et al., 1984), and by the P12 based

immunoassays on leaf tissues (Derrick et al., 1990). The chosen trees were in a 10-year-

old block in the Lake Alfred area, Florida (USA). Reverse transcriptase (RT) polymerase

chain reaction (PCR) was performed and cDNA was synthesized using Clontech

procedures, enriching for messenger RNA transcripts (Source:

http://www.bdbiosciences.com/clontech/, last accessed September 15, 2001). An

experiment was performed to determine the number of PCR cycles necessary to

normalize the cDNA synthesis of both samples (Figure 3-1).

Building the Suppressive Subtracted cDNA Libraries

The healthy and blighted cDNA samples were digested with Rsa I restriction

enzyme to remove the Clontech Smart oligos at the terminal ends of the amplified

cDNAs, generating blunt ends. After purification, different Clontech adaptors (1 and 2R)

were ligated to split sets of cut healthy and blighted cDNAs. Healthy cDNAs ligated to

each adaptor were hybridized with an excess adaptor-free blighted cDNA and with each

other. The same procedure was done with blighted cDNA samples and PCRs were

performed in both cases, enriching for differentially transcribed genes under each

condition, following the Clontech PCR-Select cDNA subtraction procedures (Source:













A B
Poly A IRNA



FFrsi-stram syad hesis
oyled wi th (dC) tailing by RT
,rr.5' % f f poiyA

Ten plate swalching
and extenoeen by RI
5" GGG r JJ.raJr p lyA
CCC
I Prvm extension st Ib e






following the Clontech Smart cDNA procedures (Source:
http://www.bdbiosciences.com/clontech/, last accessed April, 15, 2004). The
3'-PolyT-oligo-5' binds to the poly-A tail of the mRNA and the final
extension of the 1 st cDNA strand leaves an overhang of Cs, which is bound
by the 5'-oligo-GGG-3'. After template switching both ends are filled by
polymerization of the complementary strands, and PCR starts, using the
primer extension sites present at both oligos. B) Different cycles of PCR were
performed on each sample to achieve tentative uniform cDNA accumulation
before reaching the plateau of the reaction, following the manufacturer
recommendations. Healthy samples were considered optimized at the 17th
cycle, while for blighted ones, 23 cycles were run.




http://www.bdbiosciences.com/clontech/, last accessed September 15, 2001). After PCR,

only molecules with different adaptors were exponentially amplified, in theory, while

others could only have linear amplifications or impaired products due to hairpin

structures. This technique was developed by Diatchenko et al. (1996). Thus, subtracted

healthy minus blighted (H B) and blighted minus healthy (B H) libraries were

generated, in contrast to unsubtracted ones ( H and B) and controls ( + and ), Figure 3-2.













Totl MA


Poly A- RM


SMART' Convnertionu
cOA!yth~ A cNAsyntwshu
/].din;sbt1cNA\


Tutu. cD94A withi Ad@PW~ I OrlywcDPIA (kI asemi) TesM c0Awhh AdaPtwr2N


FinrIy:16: =zIon


b ME- -








kb~c~d. e
Fil In the ends


B


d{



-I 0

b-. W wanoWpdmion
c Wer mWficatofl


Figure 3-2. Making the subtracted "healthy" and "blighted" libraries. A) Scheme of
cDNA subtractive hybridizations following Diatchenko et al. (1996). Two sets
of the same cDNA (called 'tester') are ligated to different adaptors and
hybridized with in excess adaptor free cDNA (called 'driver') and with each
other. PCR was performed in both cases, enriching for differentially expressed
genes in both directions (called 'forward' and 'reverse' libraries). B) The
same procedure was done with blighted and healthy cDNA samples. Thus,
subtracted Healthy minus Blighted (H B) and Blighted minus Healthy (B -
H) libraries were generated, in contrast to unsubtracted ones ( H and B) and
controls ( + and ).


.Adaptor I




Adaptor 211







PCR

(H-B)
In theory,
only


Z

Ell h Adaptor I





Adaptor 2R


I
,J zJj hy


PCR

(B-H)
In theory,
Ell -h only


2 ISM--
b


NV_-











The differentially enriched cDNAs, (H B) and (B H), were cloned following the

Promega protocol for the pGMTeasy vector (Source: http://www.promega.com/vectors/,

last accessed on September 15, 2001) and around 500 clones were randomly selected for

sequencing from both libraries (400 from the B-H library and 100 from the H-B library).

Sequence Analysis and Selection of Clones

The obtained sequences were trimmed of vector and adaptor sequences used for

cloning until no more matches were found using the VecScreen search software (Source:

http://www.ncbi.nlm.nih.gov/VecScreen/, last accessed April 10, 2002). The sequences

were run in the automated system of the University of Florida sequencing core (Source:

http://www.biotech.ufl.edu/, last accessed April, 15, 2003) but the Clontech adaptors

were not hidden in many cases, requiring manual trimming to avoid contamination of

vector sequences. To reduce redundancy of clones that may represent the same gene, the

sequences were analyzed using all available Genebank sequences (all-sequences and non-

human non-mouse ESTs) and the Brazilian Citrus-EST project as reference sequences for

Blast searches (Altschul et al., 1997). With all sequences individually grouped by similar

blasted match outcomes, further clusterization and alignment assemblies were made

using the Vector NTI software assisted by Microsoft Office applications and other web-

based tools (Source: http://searchlauncher.bcm.tmc.edu/seq-util/seq-util.html last

accessed April, 10, 2003). The Figure 3-3 displays examples of the four steps performed

for trimming and clustering all the 500 individual sequences. After that, the longest clone

within the homolog group was the one considered for further studies.













A)
>CitrusBlight-K5R2-E09.g folder-CitrusBlight length-672
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXGAATTCACTAGTGA
TTAGCGTGGTCGCGGCCGAGGTACAAGATTGTTCACTGTCCTTCCGTTTGCGAATCCTGTGTATCTTTATGCAATGATGT
TGGGGTTTCTAACGATCATGCTCGACGTTTGGCTCTCACTAATGGGCGCGCGCTTGCTGTTGTTCTAGTCCCTGGTAACG
AGAGATCAGCCTCATGTGCGTCTTGAATGCCCTTGTATCATTACTCTAAATAATGGATCGGGCTAGCTCCTTGACGCTTA
TAACATAAACAAAAATGTCATTGCTAGCTAGCCTGTGTACCTGCCCGGGCGGCCGCTCGAAATCGAATTCCCGCGGCCX
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

B)
1 a3 _1"_ 249


Match to Vector: U Strong U Moderate LII Weak


C)
clone bp Database Highest match description e-values
All Genebank 1,367,736 -_ I- 1- I '113414.2| Homo sapiens
152 250 sequences chromosome 5 clone RP11-541P9, complete sequence 1.2
Brazilian citrus CCSM ESTs CXContig831
13,610 clusterized sequences 3e-92
ESTs Genebank non-human 1 gb|BM376262.2|BM376262 EBma01 SQ002 N11 R
non-mouse 4,893,238 maternal, 4 DPA, no tr... 80 le-12 le-12
sequences
D)
kl-5,: 84 << 361 (complementary)
SI 262 bp, le-60
K5-C10,: 83 >> 352
I 270 bp, 1 e-127
K5-G5,: 83 << 352 (complementary)
C | 270 bp, 1 e-129
K5-F9,: 83 >> 352
| 270 bp, 1 e-129
kl-25,: 69 >> 481______ 270 b,
I 400 bp, 2e-90
LEA,Z46824,EST,FASTA599: 1 >>632 Citrus EST Reference, 599 bp
I -

11 r,50) ,100 150 ,200 )250 r300 350 (1400 (50 500 r550 00





Figure 3-3. Analyses performed for each sequenced clone. A) Looking for vector
sequences, such as the Clontech adaptors highlighted in blue and red letters,
after the sequences had been masked by the sequencing core system as Xs. B)
VecScreen analysis were run for all available clones, confirming in many
cases the need of further trimming. C) Blast-N analysis in 3 databases. Many
clones had only poor matches in the Genbank search, but had good sequence
homology in the citrus and in the non-human-non-mouse EST databases. D)
Alignment of the homolog clones against a reference sequence. The green
plot, underneath the bars representing the clones, reflects the sequence
variability per region.











This step was an intensive computational effort aimed at reducing potential future

nonspecific variation from different clones that, in reality, might be representing a

transcriptional pattern of only one gene. The average length of the selected clones

reached 240 base pairs. Then, it was possible to narrow down all sequences to around 140

tentative unique clones. The Table A-1 in the appendix contains information about these

clones and displays the tentative functions and other characteristics based on blast

analysis of homolog sequences. It is worthy to note that several clones had high matches

with different genes of citrus tristeza virus.














CHAPTER 4
IDENTIFICATION OF DIFFERENTIALLY TRANSCRIBED GENES UNDER NON
AND CITRUS BLIGHT CONDITIONS

In order to evaluate the transcriptional pattern of the genes represented by the

clones selected from the subtracted libraries, a cDNA array experiment was assembled

and run. The major objective was to compare those clones under non and citrus blight

conditions.

The DNA array technology has been used in plants with success. There are

examples studying defense reactions (Scheideler et al., 2002), stress responses (Rizhsky

et al., 2002), and others. Reviews can also be seen at Seki et al. (2003) and at Donson et

al. (2002). However, in general, one arduous step in DNA array technologies is to

compare information between different experiments (Stoeckert et al., 2002). With a

similar concern, Brazma et al. (2001) proposed the MIAME protocol, which stands for

'Minimum Information About Microarray Experiment'. The intrinsic principles of the

MIAME apply primarily to define the information that should always be included in

microarray repository databases, allowing other researchers to understand the experiment

and the data. The ultimate goal of the MIAME protocol is to establish a standard for

recording and reporting microarray data based on detailed annotation of six experimental

sections: 1) experimental design; 2) array design and spot features; 3) sample preparation

and labeling; 4) hybridizations and parameters; 5) measurements and specifications; and

6) normalization and controls. The MIAME protocol does not specify the format in which

the information should be provided, but only its content. In the present cDNA array









study, much more modest than the commercially available microarrays, those guidelines

were observed attempting to induce uniform experimental conditions and effective ways

to analyze the data. Annotations during all experimental steps were recorded and are

summarized below.

Experimental Design

The tentative non-redundant clones from the subtracted libraries, (H B) and (B -

H), described at the appendix A, plus ribosomal clones and other tentative controls, were

standardized at the same concentration (160ng/ul) after PCR and purification steps. That

was possible because all the clones had the same flanking sequences at the Clontech 1

and 2R adaptor regions. Figure 4-1 displays the initial PCR efforts for some of the clones

used in this study. Most of the PCR products were in the range from 200 to 400 bp, and

some clones had to be amplified more than once to achieve the standards above.

Array Design and Spot Features

The clones were uniformly arrayed on nylon membranes, by printing duplicated

sets of each using the robot printer of the ICBR-UF (Interdisciplinary Center for

Biotechnology Research of the University of Florida). Each membrane contained 188

double spotted probes ready for further quantitative assays (Figure 4-2). The description

of each clone is given in the appendix.

For positional landmarking, ribosomal genes were used. They were clones with

sequence homology to a citrus ribosomal gene, Accession X05910.1. Controls for quality

evaluations included detection of water spots for background assessments, plasmid with

GFP for nonspecific hybridizations and aerial plant stress related genes, such as the cold

responsive gene CORcl 15 (panel G12, spot 180), for variation determinations.

Membranes were UV-cross linked and stored at around 6C.

































Figure 4-1. Preparing individual clones for the cDNA array. Selected clones were
amplified by PCR from the subtractred libraries and prepared to be arrayed on
nylon membranes (LW for low DNA mass ladder from Invitrogen; (-) for no-
template control).



Sample Preparation and Labeling

The membranes were hybridized with independent "healthy" and "blighted" RNA

sources. Total RNA samples were extracted under the same conditions from feeder roots

from adult groves of Valencia sweet orange (Citrus sinensis L. Osbeck cv. Valencia) on

Rough lemon (Citrus jambhiri Lush), Carrizo citrange (Citrus sinensis L. Osb. x

Poncirus trifoliata L. Raf.), and Swingle citrumelo (Citrus paradisi Macf. x Poncirus

trifoliata L. Raf.) rootstocks from the central Florida region, using Qiagen RNeasy

procedures with DNase treatment. Plants had been previously diagnosed using the visual












1 2 3 4 5 6 7 8 9 10 11 12
5 6 7 8 9 10 11 12
A
97 98 99 100 101 102 103 104 105 106 107 108
13 14 15 16 17 18 19 20 21 22 23 24
B
109 110 111 112 113 114 115 116 117 118 119 120
25 26 27 28 29 30 31 39 33 34 35 36
C
121 122 123 124 125 130 131 132
37 38 39 40 41 42 43 44 45 46 47 48

133 134 135 136 137 138 139 140 141 142 143 144
E 49 50 51 52 53 54 55 56 57 58 59 60
145 146 147 148 149 150 151 152 153 154 155 156
F 61 62 63 64 65 66 67 68 69 70

157 158 159 160 161 162 163 164 165 166 167 168
G 7.5 76 77 7A 7P An AI A? A3 A4

169 170
H 85 86 87 88 89 90 91 92 93 94 95 96

185 186 GFP GFP


Figure 4-2. Array design. Colored spots indicate different types of clones or controls:
pink (LM) for landmark positing; blue (H20) for water controls; yellow (spots
7 to 12) for probes from citrus leaf tissue; orange (spots 171 to 180) for probes
from stress related genes; and green (GFP) for hybridization control. Those
controls accounted for thirty four spots. Other spots were printed with clones
from the subtracted libraries of citrus root tissue. Seventeen clones were from
the H B library (spots 5 and 6; from 53 to 63; and from 67 to 70). All the
remaining one hundred and fifty four clones were from the B H library.




identification of typical symptoms in the canopy and the water injection test in the trunk


(Lee et al., 1984). All samples were uniformly handled, using tissues from pairs of


healthy and fully blighted trees of the same genotype, grown under the same condition


and from the same grove. Some membranes were used to determine the best amount of


RNA, type of primer for RT (reverse transcriptase) reaction, and RNA sources,


comparing total versus messenger RNAs. After optimizations, samples with 2ug of total


RNA had the 1st strand cDNA labeled with dATP33 using random primers following


procedures similar to that of the Ambion Strip-EZ kit (Source: http://www.ambion.com/,









last accessed Octpber 10, 2003). The RT efficiency was evaluated by counting the

scintillation of the labeled samples.

Hybridizations and Parameters

All membranes were exposed to the same amount of labeled probe, based on the

scintillation reading of one million counts/minute/ml of buffer. The samples were

hybridized overnight at 64C with individual membranes, washed, and analyzed under

the Typhoon Phosphoscreen Imager System (Source:

http://www.amershambiosciences.com, last accessed April 15, 2004), always uniformly

manipulated in pairs of healthy and blighted samples to allow further pairwise

comparisons. The employed imager system evaluates the fluorescence of each spot

represented on the screen. Changes of the oxi-reductive status of a screen component,

made by crystals of barium (Ba), fluorine (F), bromine (Br) and europium (Eu), are

induced by the radiation of the samples after overnight exposure. Then, a laser beam from

the reader induces excitation and subsequent fluorescence of each spot represented on the

screen, proportionally to the initial radiation emitted from the labeled sample. The image

of each membrane was captured by the scanner and the quantitative information was

obtained using the software ImageQuant (Source:

http://www.amershambiosciences.com, last accessed April 15, 2004). All membranes

were treated as uniformly as possible to minimize other sources of variation in the

subsequent analyses.

Measurements and Specifications

The membranes were framed and had the image contrast optimized for better

visualization using the ImageQuant, Microsoft PhotoEditor and PowerPoint software.

The quantitative information of each clone was then evaluated. The parameter for quality









of each membrane was verified looking at the ratio (Ratio) between the averaged absolute

values, intensity of the fluorescence per pixel in the screen, of all the genes (AveGene)

represented by the clones, over the average of all the "water" spots (AveH20), used as

negative controls for the hybridizations, and therefore considered as background or noise,

as



Ratio = AveGene / AveH20



Seventy-one membranes were printed and run. The expected true ratio was

calculated to be within the confidence interval of 2.33 and 3.27, for p < 0.01. Therefore,

individual membranes with a ratio smaller than the low limit of 2.33 were considered

inappropriate for analysis and were discarded. One membrane that had nonspecific

hybridization, measured by the spots with GFP genes, was also discarded. Forty unbiased

membranes passed those criteria and were saved. Among those, twenty individual

membranes represented ten pairs of uniformly manipulated healthy and blighted sampled

trees. Each pair of samples was collected at the same day, from the same grove, and had

their RNA extracted, labeled and hybridized under the same conditions, as uniformly as

possible. Thus, membranes of any considered pair of healthy and blighted samples

displayed an averaged absolute value (intensity/pixel) for all clones similar to the other

membrane, in most of the cases, and both were respectively higher than their background

and nonspecific hybridization controls by at least 2.33 fold. Root samples from trees on

Carrizo were used on the pairs made by the membranes lx2, 5x6, 67x54 and 65x66.









From Lemon, on the pairs 9x10 and 61x58. From Swingle, on the pairs 29x30, 31x34,

51x52 and 69x70 (Figure 4-3).




Ratio 6
4ai 6-- ----------- ------- ---------------------------
H AveGene/AveH20






0
membranes 1 2 5 6 67 54 65 66 9 10 61 58 29 30 31 34 51 52 69 70 Cutoff

Carrizo Lemon Swingle


Figure 4-3. Quality parameter for each membrane. The unitless ratio represents the
proportion between the averaged absolute value for all clones (AveGene) in
the membrane over the averaged absolute reading for the background level
(AveH20) of the same membrane. Membranes with a ratio smaller than 2.33
were discarded.



Normalization and Controls

The readings of the duplicated spots for each clone were averaged. All "water"

spots (AveH20), previously calculated for each membrane, were used to subtract the

background from each clone (C). Negative values implied that the reading of the clone

was smaller than the background value. Those values were discarded. Then, each clone

was normalized (n) against the new mean of the clones (MeClone) in the membrane,

rather than against one or two genes, such as a ribosomal or other ones, as follows:


n = [ ( C AveH20 ) / MeClone ] 100









The normalization method should reflect similar results comparing to the visual

evaluation of the studied membranes. The mean of the clones was used in this study

because it met this criteria better than other measures of central tendency, such as the

median of the numbers. Normalization against a measure of variance, using the standard

deviation, was also tried but no improvements were observed either.

Besides the visual evaluation of the membranes, two other types of analyses were

performed. To study the relative levels of transcription of the same clone comparing

healthy and blighted samples from uniform conditions, cluster analysis was done. To

study the variance of each clone, contrast of means analysis was done.

The Visual Evaluation of the Membranes

The tentative visual differences for each clone were highlighted in each framed

membrane. This procedure helped to identify clones not previously seen under the initial

visual evaluations. The three evaluated genotypes (a lemon, a citrange and a citrumelo)

displayed differences across the species. But looking at one rootstock per time, some

visual differences between healthy and blighted samples were also seen (Figure 4-4).

Under the citrus blight condition, Carrizo citrange displayed stronger signal for the

genes represented by the clones on the panels Bl (clone 109), Cl (clone 25), D4 (clone

136), Fl (clone 157), G7 (clone 79), and maybe others. Lemon had differences on the

panels Cl (clone 25), C10 (clone 34), G7 (clone 79) and G8 (clone 80). Swingle had

differences mainly on the panel Cl (clone 25). Very few clones displayed more signal on

the healthy than on blighted samples: the panel A6 (clone 6) shows a tentative candidate

of that on Carrizo and the panels E8 (clone 152) and E9 (clone 153) on Lemon.













healthy tree 1-15


C *
o f...__.. _,9. @




F .. ** ". .-

e )ee *. ** ** --
H .* .
.9_ oS I __I

B) Lemon roots healthy tree 3-11
S 1 2 3 14 5 6 7 8 10in 11 12




C *



9. to 0
09 0* n 9 09 n n _* ..







5 1 2 3 14 6 17 9 10 11 12
9 *9


99'. .9_ __
A .9_W * *
*_ 99_9_ _9_ 9_





E- y .


A) Carrizo roots


Figure 4-4. Visual evaluation of the membrane pairs. Membranes were hybridized with
samples labeled with P33, using feeder roots of 'Valencia' tree on different
rootstocks. A) on Carrizo citrange (Citrus sinensis L. Osb. x Poncirus
trifoliata L. Raf.). B) on Rough lemon (Citrus jambhiri Lush). C) on Swingle
citrumelo (Citrus paradisi Macf. x Poncirus trifoliata L. Raf.). Each panel
has two clones horizontally replicated. Red arrows represent potential more
signal on blighted samples. Blue arrows, on healthy samples. Green arrows,
near identical. Blue circles represent spots filled only with water, for
background control. Green circles, with GFP, for non-specific hybridization
control. Black circles, with CORcl 15, for observations of a streets related
gene.


blighted tree 1-11
6 1 2 a 4 a 6 7 8 9 10 i 1 12

A ie(94






E
"'i"" "?^"-- t-------
C ** *. -






St *







A ** **a
0 ** .
** *" 9-
blighted tree 6-9

16 1 2 2 4 0 0 7 8 9 I0 11 12






. .
--- a---- -- ---_J _- --_





F 2 -. 5 6 .
F'* '40 .. 4o


Ho I W







U------^---^-----------

-- -- -- -- -- -- -- --



F 9 .9











Nearly unchanged genes were seen in all samples and the panels D9 (clone 141)

and E5 (clone 149) represent examples of those, in all three rootstocks. The Figure 4-4

shows one pair of membranes for each rootsctock.

The Cluster Analysis

The transcriptional levels of the normalized Healthy (nH) and Blighted (nB)

samples were compared following the Log2 function, to allow equidistant visualization of

the fold (F) induction or repression for each gene, as:



F = Log2 [ (nB+0.01) / (nH+0.01) ]



Gaasterland and Bekiranov (2000) describe two major types of analysis, the

supervised and the unsupervised approach. In the supervised analysis, a particular context

of each measurement is known and the end result is a list of individual genes behaving

differently in each of the different contexts. In this study, the supervised model was

accepted and the null hypothesis considered no differences in the transcriptional levels of

the candidate gene under healthy and blighted conditions.

Then, the cluster analysis and the graphic visualization for each clone was done

employing respectively the Cluster and the TreeView software (Eisen et al., 1998). The

Cluster software works on a random assignment of vectors for similarities found within

the numbers. Subsequent new layers of vectors allow further clusterization of all samples.

To test the significance of the differences, t-Tests were employed. Induced and repressed

candidate clones were nominated, after analyzing the outcome of the ten independent

pairs of healthy and blighted trees (Figure 4-5).












A) Cluster analysis



A -

i|











i7:
4-
4-









4-





4-
in_






3-





4-

)ii


B) Selected clones


winter spring summer spring
03m 03 02 l03


M M .
u4D CD

C2 2 C


N P P


ChD


N N @h
W9~ CD4
Q x'm c


M
a
2 Membrane pairs

a f

nO Clone t-Test


v W,, U, W U -
t' N .0 A
9 0 90 C 0 0;


Figure 4-5. Transcriptional profiling of the candidate genes. A) Outcome of the ten pairs
of trees under healthy and blighted conditions, being four pairs on Carrizo
citrange, two on Rough Lemon and four on Swingle citrumelo rootstock. Each
spot represented the ratio of the normalized blighted (nB) over the healthy
(nH) sample given by Log2[(nB+0.01)/(nH+0.01). Green indicates negative
ratio (more healthy than blighted) and red positive ratio (more blighted than
healthy). B) Visualization of the individual clones and their 'p' values for the
t-Test, with the significant ones (p<0.05) being highlighted with a nearby star.


157
109
25
38
136
153
34
152
80
6
79
149
141
scale


p<0.307
p<0.014
p<0.156
p<0.089
p<0.368
p<0.005
p<0.832
p<0.617
p<0.269
p<0.153
p<0.450
p<0.580
p<0.007


M M
M CD
a
LA
Ca Ln
04 C2
CD N
Ln

Q W









After the cluster analysis, all membranes were visually re-evaluated. Clones that

had p<0.05 from the group I, mostly upregulated candidates under the citrus blight

condition, were located again. This process helped to spot differences not previously seen

in the membranes. Similar approach was taken for the clones of the groups II, III and IV,

mostly downregulated candidates under the citrus blight condition. However, it is

possible that, for these clones, a more sensitive detection method would be more

effective, because few visual differences were re-confirmed. The nearly unchanged genes

of the group V, displayed the expected clones 141 and 149. Albeit small, the clone 141

did have a significant variation, with p<0.007, leaving the clone 149 alone as a candidate

for an unchanged gene under citrus blight condition.

These ten paired samples were collected and run from July of 2002 to May of 2003.

Closer patterns can be observed comparing the Carrizo and Swingle samples from the

summer of 2002 and the spring of 2003. However, more replicates would be needed to

better study the real contribution of the season of the year on the studied model.

The Contrast of Means Analysis

Although the previous analysis is powerful for initial screening, it brings a natural

chance of error by selecting potential false positives, since a clone with initial reading

value (C) that approaches the background level (W) will generate a normalized number

(nH or nB) that approaches zero, complicating any further conclusions for the pairwise or

cluster analyses. On the other hand, simply discarding those clones may penalize the final

outcome by leaving behind clones that may be indeed positives. Thus, besides exercising

caution on apparently extremely induced or repressed clones, those were the reasons to

perform the contrast of means analysis. For that, each clone was assessed individually









within each uniform pair of plants (healthy and blighted), from the same rootstock type

and from the same block of trees, as previously described.

The assumed model for variance in this case implies that the variance of a clone (x)

is a function of its averaged value (Ax), plus the contribution of the membrane (M), plus

the contribution of the uncontrolled residual factors (e), as



x= Ax+M+e



Other assumptions for the model above were also considered, as having additive

factors, normal distribution of the data, independence between treatments and

homocedasticity-also called uniformity of variance (Banzatto and Kronka, 1992).

Additive contributions of each factor can be explained by the assumed model itself, since

the outcome of each clone was the sum of those factors. Normality was only partially

observed because the data aggregated around the mean, for the most part, but not

perfectly. Independence of treatments in this biological system could not be fully

accepted because that implies that the genes assessed by each clone operate

independently from each other. Since metabolic pathways are complexes and regulated

by counterpart genes, this is probably not true in this system either. Homocedasticity was

not verified and the data was checked for transformation options, such as, RootSquare,

Log and other functions. The smallest discrepancy in variances was obtained using the

Log transformation of the previous normalized data (n), giving to each clone its final

value (V), as









V = Log2 (n + 1.5)


The normalization for the contrast of means was similar to the initial normalization

for the cluster analysis (nH and nB) previously described, except for considering

individual values of all spots and adding the value 1.5 to avoid zeros in the computation.

Once the transformation was observed, t-Test was employed on the established contrast

of means, where the estimate of the healthy mean (Hm) and the estimate of the blighted

mean (Bm) generated the estimate of the contrast (AY) for each clone, as



AY = Hm Bm



Then, the relation between the absolute value for the estimate of the contrast (AY)

and its standard error, which is a function of the number of replicates (rH or rB) and the

standard deviations (sH or sB) for healthy and blighted samples, gives the t-Test value (t)

for each clone (Banzatto and Kronka, 1992).

Subsequent comparisons against standard t-values determined at which level of

probability (p) the null hypothesis, claiming no difference in the transcriptional level of

the studied clone, can be accepted or rejected, in favor of the alternative hypothesis,

claiming significant differences between healthy and blighted transcriptional levels for

each clone.

Some of the clones were variable but not at a level considered significant (t-Test)

for this cDNA array system (p<0.05). However, other clones displayed significant









response to citrus blight, suggesting that citrus blight is apparently able to affect the

transcriptional level of certain genes in affected plants (Figure 4-6).

In plants with citrus blight, the clone 6 was significantly downregulated (p<0.05) in

three out of the ten pairs of evaluated trees. However, it was also significantly

upregulated twice, implying in a great non-specific variation.

The clone 25 was significantly upregulated (p<0.05) only in two studied pairs, but

displayed higher averaged values on blighted samples in nine pairs of plants, and a strong

significant difference on the pair 5 l1x52.

The clone 38 was significantly upregulated in two pairs of plants. It also had a

higher averaged transcriptional value in three other pairs of plants.

The clone 109 was significantly upregulated (p<0.05) in three out of the ten pairs,

and had other three higher averaged values on blighted than on healthy samples.

The clone 149 was nearly unchanged, however it had one significant difference

(p<0.05) comparing the affected and the healthy plants on the pair 9x10.

The clone 153 was significantly downregulated (p<0.05) in only two paired

samples, but had higher averaged values on healthy than on blighted samples in other five

pairs.

The clone 157 had a higher averaged transcriptional level on affected than on

healthy plants, but a large variance too, being significantly downregulated (p<0.05) in

one pair of plants.















Log2(n+.5) p<0.013 p<0.015 8




2 Clone 6


p<0.010 p<0. 057 p.OOO








48 ------- ---------------- -- -- --------- -- -- ------------------------ - -----
,, p<0.000
















Sp<09 p< 068 p





0<.8








6 --------------- -------- -- ---------------------

p<0.063 p<0.035
6 - -- - - - -- - - - - -- -

2 Clone 157















2 x2 Cz5x6 Cz65x66 Cz67x54 Le61x58 Le9x10 Sw29x30 Sw31x34 Sw51x52 Sw69x7 AveH20 AveClone










and blighted (purple columns) trees on Carrizo (Cz), Lemon (Le) and Swingle
(Sw). The 'p' value represents the significance of the t-Test, helping in the
decision of rejecting or accepting the null hypothesis. The blue stars indicate
significant differences when p<0.05. The bars are the standard deviation of
each studied condition: healthy or blighted samples.











Comparing the Evaluation Methods

All three methods used to evaluate the transcriptional pattern of the genes

represented by the clones displayed positive and negative aspects. The visual evaluation

of the membranes required neither normalization nor computation and it was good for the

major differences, besides confirming, or not, the results of the quantitative analysis.

However it did not allow an easy identification of all the candidate clones in the first

attempts. The cluster analysis was good for combining all clones from the available

samples under the pairwise based approach. Inferences about the averaged fold induction,

or repression, and about the aggregation group of each gene could be estimated. But not

all quantitative information matched the visual evaluation of the membranes. The contrast

of means was positive to reveal the significance of the differences within each pair of

trees. It represented an unfolded view of the cluster analysis replicates. However, since

the data was transformed, further inferences on the relative transcriptional levels within

each pair of plants were no longer precise.

Combining the Results for the Selected Clones

Several clones displayed either visual or quantitative differences between the

evaluated samples. The Table 4-1 displays the summarized results for those selected

clones.

Among the clones that displayed some differences in the cDNA array experiment,

the clone 6 is from the healthy minus blighted (H-B) enriched library and displayed

mostly lower transcriptional levels under the citrus blight condition. It is 325 bases long

and has sequence homology (Table 4-2) with an Arabidopsis thaliana EST (e-value of











Table 4-1. Transcriptional pattern of the selected clones. Observations from the three
different types of analysis, comparing healthy (H) and blighted (B) samples.
Clone Visual Cluster analysis Contrast of means analysis
(library) differences (averaged Log2*; (number of significant differences,
p-value) for p<0.05)
6 (H-B) H>B -0.46; p<0.153 3 (H > B) and 2 (H 25 (B-H) H 38 (B-H) rare +2.44; p<0.089 2 (H < B)
109 (B-H) H 149 (B-H) H ~ B +0.05; p<0.580 1 (H > B)
153 (B-H) H>B -3.13; p<0.005 2 (H > B)
157 (B-H) H B) and 1 (H * Log2 stands for Log2[(nB+0.01)/(nH+0.01)]


le-180) in the Genbank and with a citrus EST (7e-11) in the Brazilian databank. It has an

unknown function based on the translated query (Figure 4-7) to the protein databank (4e-

30) and a high nucleotide homology to a mitochondrial sequence (le-180), raising the

question whether the clone 6 represents an open reading frame (ORF) of a gene or not.



Query: 41 CRRQRGSRYTIRAGRYLCDKEFRYLRTVRVTAAVYRGFHSKLITLLLLTFQHRACVRLYT 220
C RQRGSRY IRAGR L DKEFRYLRTV VTAAVYRGF+S ++ LLLTF+HRA VR YT
Sbjct: 4 CWRQRGSRYAIRAGRMNLPKEFRYLRTVIVTAAVYRGENS-VIAiHLLLTFRHRAGVRPYT 62

Query: 221 SCYHLAESCVFNKQSLPPGMCRFPNQKIGEHPFSR 325
SCYH AESCVFNKQSLPPG+C PFS+
Sbjct: 63 SCYHFAESCVFNKQSLPPGLCHIALVAQHRSPFSQ 97

Figure 4-7. Blast-X of the clone 6 using the non-redundant protein database. Sequence of
an environmental protein Accession EAI39113.1 (subject), with unknown
function, had similarities (4e-30) to the translated version of the clone 6
(query).



But because of its variance (Figure 4-6), the clone 6 is considered to represent a

false positive outcome, not representing a true downregulated gene under the citrus blight


condition.









The Table 4-2 displays the accession numbers of the first match on the blast search

analysis, using different databases, for the selected clones.


Table 4-2. Blast search analysis for some clones. The search for homolog sequences was
done using different databases.
Clones e-values Highest match on sequence search analysis
6 le-'1o Arabidopsis thaliana EST, Accession CF653082
7e-'1 Citrus EST, CCSM, Brazil, Contig 204
4e-30 unknown protein, environmental sequence, Accession EAI39113.1
le-180 Arabidopsis thaliana mitochondrial genome, part A, Accession
Y08501.1
25 1e-91 Citrus EST, CCSM, Brazil, Contig 416
38 8e-88 Citrus chitinase class II, Accession Z70032.1
Se-88 Citrus chitinase class I, Accession AB081944.1
I3e-95 Citrus EST, CCSM, Brazil, Contig 1434
109 9e-56 Citrus EST, Accession CK935651
9e-44 Citrus EST, Accession CB293790
(2e-65)* Arabidopsis thaliana, ubiquitin ligase SCF complex subunit,
Accession NP568603.1
149 2e-4 Citrus EST, CCSM, Brazil, Contig CXJE02
153 Poor sequence information
157 Poor matches in the searched databases
* e-value of the citrus EST Accession CK935651 translated query on the protein databank


The clone 25 is from the blighted minus healthy (B-H) enriched library and

displayed higher transcriptional levels under the citrus blight condition in some of the

tested pairs of plants. It is 300 bases long and has sequence homology with a citrus EST

(le-91) in the Brazilian databank, but only poor matches in Genbank. It also had a strong

visual difference on the membranes of Swingle (panel Cl, figure 4-4C). That difference

was significant (p<0.05) in one out of the four tested pairs of plants on Swingle (Figure

4-6). However, considering all rootstocks, the overall transcriptional fold induction under

the blight condition was around three and a half times, Log2[(nB+0.01)/(nH+0.01)=1.82,

but with a poor 'p' value of only 0.156 (Table 4-1 and Figure 4-5B). It is possible that the










clone 25 represents a true upregulated gene under the citrus blight condition, but with

higher transcriptional responses in Swingle citrumelo.

The clone 38 is from the B-H library and is 235 base pairs long. It represents a

citrus chitinase gene (Figure 4-8).


A) Cone38: 586 ( 811 (corrplerrentary)


Chitinase-ll: 1 > 1082
(Accession Z70032.1)


i 5C00 ,1000


aone38: 681 << 906 (corrplenentary)


e-value = le-88


Chitinase-1: 1 > 935
1 > (Accession AB081944.1)


11 ,100 ,23 r300 1400 P 00 ,700 800 00


Figure 4-8. Clone 38 has sequence homology to citrus chitinases. Blast-N showed high
similarities to (A) a class II chitinase; and to (B) a class I chitinase.


The clone 38 had a transcriptional fold induction under the blight condition of

around five times, Log2[(nB+0.01)/(nH+0.01)=2.44 (Table 4-1); but a strong visual

difference within the membranes was not seen (panel D2, Figure 4-4A). Looking at the


e-value = 8e-88









contrast of means (Figure 4-6), two pairs of trees displayed significantly (p<0.05) more

transcripts of the gene represented by the clone 38 in blighted than in healthy trees.

However, five pairs of trees responded only at a very low level, close to the background

(AveH20). It is possible that the clone 38 represents a responsive gene to citrus blight,

but had poor printing on the membrane surfaces for those five abnormal readings. To test

that hypothesis, a new series of membranes were printed and run. The figure 4-9 displays

the results.




74-H Rep.

76-B 1

78-H
Rep.
80-B 2

Ribosomal LEA Others Chitinase

Figure 4-9. Transcriptional profiling of the clone 38 (a chitinase homolog) and other
clones. The membranes were manipulated and hybridized under similar
conditions, using 5ug of total RNA labeled with P33, from feeder roots
samples of a healthy (H) and a blighted (B) tree on Carrizo citrange. The
process was repeated twice (rep. 1 and 2) generating four membranes. The
quantitative difference for the chitinase homolog clone, comparing the healthy
and blighted samples (t-Test), was significant (p<0.01).


Therefore, the clone 38 is considered to represent a true upregulated gene under the

citrus blight condition.

The clone 109 is also from the B-H enriched library and displayed mostly higher

transcriptional levels under the citrus blight condition. It is 176 bases long and has

sequence homology with citrus ESTs in the Genbank using the non-mouse and non-









human entry (Figure 4-10), but only poor direct matches as a translated query on the

protein databank.




CLO~E1 09,: 662 <<847 (corrplerentary)

QC935651,: 78 > 828
1 e-value=9e-56
CB293790,: 1 > 753
1 -e-value=9e-44

,1 PO 100 1150 0 250 300 3 50 P0 0T P600 6 ,700 1,750 0



Figure 4-10. The clone 109 had sequence homology to citrus ESTs. The citrus EST
Accession CK935651 was obtained from a library using citrus fruit
developing tissues. The CB293790, from a library using citrus cold acclimated
tissues.



However, blasting the longer citrus EST homolog (Accession CK935651) as a

translated query gives a high match (2e-65) with an 'E3 ubiquitin ligase SCF complex

subunit SKP1/ASK1 (At2)/UFO-binding protein (UIP2)' of Arabidopsis thaliana

(Accession NP568603.1). The clone 109 had visual differences in the membranes (panel

B 1, Figure 4-4A); had an averaged transcriptional fold induction of around eight times,

Log2[(nB+0.01)/(nH+0.01)] =3.12, with p<0.014 (Table 4-1 and Figure 4-5B); and it was

significantly upregulated in three out of the ten studied pairs of samples. Therefore, the

clone 109 is considered to represent a true upregulated gene under the citrus blight

condition.

The clone 149 is 144 bases long and has only poor matches in Genbank either as a

nucleotide query or as a translated query, but it has a good match with a citrus EST (2e-

54) in the Brazilian databank. It is from the B-H enriched library. The clone 149 probably









represents an unchanged gene under the citrus blight condition, considering the three

methods used to evaluate its transcriptional pattern.

The clone 153 is from the B-H enriched library, but probably represents a true

downregulated gene under the citrus blight condition. It displayed visual differences in

the membranes (panel E9, Figure 4-4B); significant differences in the cluster analysis

with p<0.005 (Figure 4-5B); and two significant differences in the contrast of mean

analysis (Table 4-1 and Figure 4-6). However, the sequence information available was

poor, with low phred-quality parameter. It was re-sequenced twice with no further

improvements.

The clone 157 is 195 bases long but had no good match in all three searched

databases. Because of its variance (Figure 4-6), it is considered to be another false

positive outcome, not representing a true upregulated gene under the citrus blight

condition.














CHAPTER 5
THE RELATIVE TRANSCRIPTIONAL RESPONSES OF THE SELECTED GENES
UNDER DIFFERENT INCIDENCES OF CITRUS BLIGHT

The process of building cDNA arrays is marked by many steps that go from the

collection of samples to make the libraries until the manipulation of the membranes. As a

consequence, the final data can carry cumulative sources of nonspecific variation, which

may compress or hide legitimate up or down regulated genes under study. With that

concern, Chuaqui et al. (2002) stated that researchers must be aware whether the results

in microarray based experiments are accurate and the data fundamentally describe the

phenomenon being investigated. In addition to employing replicates as used in this study,

Chuaqui et al. (2002) considers the need of a validation process. Therefore, an

independent RNA assessment method was used to confirm, or not, the transcriptional

level of the previously identified genes under non and citrus blight conditions. The effect

of cold and drought stresses was also tested.

Quantitative Real Time PCR was the Method of Choice

Among the available methods, quantitative real time PCR was the method of

choice, because of its sensitivity and reliability. It has been used under different platforms

and for different purposes, such as confirmation of microarray data in human cancer

research (van den Boom et al., 2003), and quantification of a citrus pathogen (Oliveira et

al., 2002). It is based on the determination of a fluorescent signal produced during the

PCR cycles, allowing the quantification of the amplified product at a real time and the

subsequent estimation of its relative original template concentration (Mackay et al.,









2002). According to Dorak (source: http://dorakmt.tripod.com/genetics/realtime.html, last

accessed on January 10, 2004), real time PCR is a preferable alternative to study

transcription to other forms of reverse transcriptase (RT) PCR, which only detect the final

amount of the amplified product. The chemistry employed in real time PCR detection

system is the key to the process. There are two general methods for the quantification of

the PCR product: the DNA binding reagents (i.e. SYBR Green) and the fluorescent

probes (i.e. TaqMan). Characteristics of each detection type and more details about the

technique can be seen in Mackay et al. (2002). The TaqMan probe relies on the

fluorescence resonance energy transfer (FRET) for quantification. TaqMan probes are

oligonucleotides that contain a fluorescent dye, typically on the 5' base, and a quenching

dye, typically located on the 3' base. When irradiated, the excited fluorescent dye

transfers energy to the nearby quenching dye molecule rather than fluorescing, resulting

in a nonfluorescent substrate. TaqMan probes are designed to hybridize to an internal

region of a PCR product. During PCR, when the polymerase replicates a template on

which a TaqMan probe is bound, the 5' exonuclease activity of the polymerase cleaves

the probe. This separates the fluorescent and quenching dyes and FRET no longer occurs.

Fluorescence increases in each cycle, proportional to the rate of probe cleavage.

Normalization is done against an active internal control such as a ribosomal or other

gene. Passive controls, compounds that do not participate in the PCR reaction, are used to

adjust the reaction level and background. Other important parameter is the threshold

cycle, or Ct, obtained by an arbitrary threshold line chosen within the linear phase of the

PCR reaction, that gives the PCR cycle for that fluorescence value. The Ct values are









used to compare different PCR reactions, allowing the quantitative estimation of the

initial target template.

Compared to DNA dyes, the TaqMan method also has the benefit of producing an

amplified target with longer sequence specificity, since it is the product of the

hybridization of two primers plus a probe, in a total length that normally stays around 150

bases. Real time PCR using dyes (i.e. SYBER Green) relies only on the specificity of the

primers. Considering the described characteristics, the TaqMan probe was the method of

choice to validate (or not) the initial results of the cDNA array experiment.

The Selected Clones and the Characteristics of the Probes

The selected clones observed in the cDNA array were used in this experiment. The

clones considered to represent up-regulated genes under the citrus blight condition were

the clones 25, 38 and 109. The clone 149 appeared to be nearly unchanged. The clone

153 was considered to be down-regulated; however the sequence information for this

clone was poor, even after re-sequencing. Thus, the clone 6 was included in this

experiment as a candidate to represent a down-regulated gene under the citrus blight

condition. Those clones were chosen because genes with altered transcriptional levels

under the citrus blight condition can be valuable tools on further comprehensive

experiments. Genes with unaltered patterns can be useful references for normalizations or

for other biological needs as well.

Based on sequence homology, the clone 38 represents a citrus chitinase gene (Table

4-2, chapter 4). Chitinases are known to respond to different forms of stresses, such as

fungal pathogens in citrus plants (Fanta et al., 2003). Citrus blight causes mineral

unbalance with a clear pattern for zinc deficiency in leaves. Taylor et al. (1996),

identified a small protein, named P5 (Accession AAB46813), that has sequence









similarities to chitin binding proteins and functions completing zinc in affected plants. It

is probably associated with the known translocation of zinc from leaves to trunk tissues

(Albrigo and Young, 1981). In addition to the clone 38, chitinase homologs were found

in our libraries and in combination with the class I Chitinase (Accession AB081944.1) as

reference, a putative nucleotide sequence for a P5 candidate gene was deduced and was

used in this validation experiment (Figure 5-1).


true P5 Query: 2 CGKSWCPGGECCSRFGWCGL 22 22 residues
CG++ VVCPGGECCS++GWCGL
AB081944.1 Sbjct: 25 CGSGWCPGGECCSQYGWCGL 45 e-value=le-05


C) region of the citrus chitinase gene used in the real time PCR
SClone38: 681 (< 906 (corrplementary)
< I
p5-candidate: 70 ) 135
E->
Chitinase-I: 1 ) 935
I >

,1 ,100 O200 30O0 ,400 ,500 PO0 ,700 POO ,900



Figure 5-1. Designing a candidate sequence for the P5 gene. A) The true P5 protein has
homology to the translated version of the citrus chitinase class I Accesion
AB081944.1. B) The aminoacid identities was 80%, with an e-value of le-05.
C) Based on the reference chitinase class I (Accession AB081944.1), which
also had high homology to the clone 38, a candidate sequenced was designed
for the P5 gene and was used in the real time PCR experiment.


protein e-values Highest match on sequence search analysis

true P5 le-05 Citrus chitinase class I, Accession AB081944.1










Another gene associated to citrus blight is the P12 (Accession AF015782). It is up-

regulated in affected plants (Derrick et al., 1990) and has sequence similarities to

expansins (Ceccardi et al., 1998), but has so far an unknown role in the decline of the

trees. P12 was also included in this experiment.

The Table 5-1 displays the characteristics of the chosen probes for each of the

candidate clones and P12.




Table 5-1. Characteristics of the chosen probes. The primers and probes were designed
using the PrimerExpress Software (Source: Applied Biosystems Resources,
http://www.appliedbiosystems.com/index.cfm, last accessed December 10,
2003).
Clone Start Length Tm GC Taqman probe architecture
(bp) (bp) (C) (%)
name (flourochrome-5'sequence 3'-quencher)

6 136 29 69 45 fam6-TCGCTACTTATGCGACAAGGAATTTCGCT-tamra
25 192 32 69 41 Fam6-TTGAAGGCAAGTTAGGAAATTAGCAAAGCCAG-tamra
109 94 39 69 33 fam6-ATGATACAGAGAAGGTTGGGATGATATGACATTAAAACA-tamra
149 213 26 69 42 fam6-CTGTATCATCTTACTTTACGCTTCCC-tamra
P5 candidate 78 18 68 67 fam6-AAGCGGCGTTGTGTGCCC-tamra
P12 179 25 69 56 tet-TGGAGTCATGATAGCCGCAGCAAGC-tamra


Two major calculation methods are possible, the absolute and the relative

quantification of the target gene (Source: bulletin # 2, Applied Biosystem,

http://www.appliedbiosystems.com/index.cfm, last accessed December 10, 2003). This

study employed the relative method. Then, the 18S gene was chosen as the active internal

control, or normalizer, because it is considered to be non altered under different

conditions (Dorak, 2004, source: http://dorakmt.tripod.com/genetics/realtime.html, last

accessed on January 10, 2004). However, no previous information was seen about the

18S levels in feeder root samples of citrus hybrids under non and citrus blight conditions.

It was assumed that the transcriptional level of the 18S gene was not altered by citrus

blight.









To test the efficiency of the amplification between the target gene and the

normalizer, a RNA 2 fold dilution series was evaluated, from 3.9 to 1000 ng of a mixed

pool of RNA templates. Under a similar efficiency, the graphic of the log input amount of

the total RNA and ACt (difference between the chosen thresholds, the Ct of the target

minus the Ct of the normalizer) would give a straight line, nearly parallel to the abscissa.

Small deviations from that are accepted and the bulletin # 2 of Applied Biosystem

(Source: http://www.appliedbiosystems.com/index.cfm, last accessed December 10,

2003) recommends that the absolute value of the slope (of the estimated linear function)

should be equal or smaller than 0.1. Figure 5-2 displays the outcome for the clone 109,

which had a slope of 0.1055, for the tested range of 3.9 to 1000ng of RNA template.




A) Plot of log input amount versus ACt
30 ACt

28

26 -

24
clone 109 x 18S
S= 0.1055x + 26.138 Linear (clone 109 x 18S)
22 -....
-2.4 -2.1 -1.8 -1.5 -1.2 -0.9 -0.6 -0.3 0.0
Log ng Total RNA


Figure 5-2. Relative PCR efficiency plot of the clone 109 against the normalizer 18S.



In the same type of analysis, the clone 6 had a slope of 0.0104; the clone 25,

0.1029; the clonel49, 0.0602; the P5 candidate, 0.0811; and the P12, 0.0196.









Collecting and Preparing the Samples

The same citrus groves in central Florida area, used to collect root samples for the

cDNA array experiment, were revisited, with the expectation to re-assess the same trees.

However, most of the previously healthy trees had started to display some blight

symptoms and some of the affected trees had been pushed out. Therefore, some other

plants were used in the real time PCR experiment, after verification of the typical

symptoms and the water test (Lee et al., 1984) for diagnostic purposes. Three classes of

plants were sampled:

* Healthy: with no visual symptoms in the canopy, and with a minimum uptake of
water, manually injected into the trunk by a syringe, of 3 ml/10 seconds.

* Mildly affected: initial symptoms in the canopy, such as opaqueness of leaves and
few twig dye backs on the top. Uptake of water from 1 to 2 ml/10 seconds.

* Blighted: fully symptomatic canopy. Uptake of water smaller than 0.5ml/10
seconds.

Tentative different stages of the process (healthy, mildly and fully affected trees)

were used because they may offer more information than only comparing healthy and

fully declined trees, as done on the cDNA array experiment.

Feeder root tissues from each tree were once again collected and the RNA

extraction procedures had this time double DNA digestion step, using an adaptation of

the Qiagen RNeasy kit (Source: Qiagen, http://wwwl.qiagen.com/Default.aspx, last

assessed January 4, 2004). No DNA contamination was observed when verifying RNA

quality on formaldehyde gels. All sampling and extracting procedures were done as

uniformly as possible. The sampled trees were Valencia sweet orange (Citrus sinensis L.

Osbeck cv. Valencia) on the rootstock Carrizo citrange (Citrus sinensis L. Osb. x

Poncirus trifoliata L. Raf.) of around 10 years old. This choice was considered because









trees of Valencia on Carrizo have been largely used in the Florida citrus industry. The

combinations of Valencia on Rough lemon and on Swingle citrumelo were also initially

considered. But the sampled trees were of different ages and had other non uniform

conditions. Therefore, they were no longer considered.

Reverse Transcriptase (RT) and PCR Reactions

Duplicated RT reactions and two PCR reaction sets per each sample were used to

minimize pipeting and other reaction errors, resulting initially, in four mechanical

replicates per each biological sample. Both RT reactions for each sample were done at

the same time and under the same conditions, using a 96 well plate and the protocol given

by the Applied Biosystem RT kit (Source: http://www.appliedbiosystems.com/index.cfm,

last accessed January 10, 2004). Other conditions were also as uniform as possible. The

RT plate was then stored at -20C.

The 7700 Applied Biosystem sequence detection system was employed for all set

of PCR reactions. The PCR conditions were 50C/2minutes, 95C/10minutes, and 45

cycles of 95C/15seconds plus 60C/1minute.

An arbitrary threshold cycle (Ct) was chosen for each amplification plot. The data

was analyzed using the baseline computation method on the Apple based Sequence

Detection System (SDS) vl.9 software (Source:

http://www.appliedbiosystems.com/index.cfm, last accessed April 10, 2004).

Three controls were employed in all real time PCR plates: the NTC (no template

control, to verify non specific amplifications), the NRT (no reverse transcriptase control,

to verify DNA contamination) and the NAC (no amplification control, to verify non

specific readings). After the reactions, none of the NACs gave integral fluorescent

readings. The great majority of the NTCs did not reach the chosen threshold level (Ct) for









the target gene at the 40th PCR cycle. The reactions that did, were discarded. All samples

that had a difference in the ACt (the Ct of the sample minus the Ct of the NRT) for the

reading of the 18S gene, smaller than 10 cycles were also discarded. The Applied

Biosystem (Source: Taqman ribosomal RNA control reagents protocol,

http://www.appliedbiosystems.com/index.cfm, last accessed April 10, 2004) recommends

this cut off level to avoid significant contribution of non-target templates to the measured

gene. The remaining healthy, mildly affected and blighted samples were further analyzed.

Figure 5-3 displays examples of amplification plots for the clone 109, P5 candidate and

P12.

Relative Quantification of the Transcriptional Levels of the Selected Genes

All samples were analyzed in groups according to their similar levels of 18S,

measured by the threshold (Ct) in each reaction. Four replicates, groups of independent

healthy, mildly affected, and blighted trees, were analyzed, in a total number of twelve

different sampled trees.

Each clone was evaluated individually. The contrast to compare 'healthy-mildly

affected-fully blighted conditions' is not orthogonal. Therefore, the t-Test can not be

used. The residual standard deviation (s) was chosen to indicate the non specific variation

within each group of plants, or replicates. This measure of variation indicates the effect of

non controlled factors, and therefore, seems adequate to estimate the contribution of

errors in the experiments. It was calculated using the mean square of the residue (MSres),

after the ANOVA (analysis of variance) for each group of plants, as:


s = (MSres)














A) The clone 109 robe used the fluorochromes FAM6 and TAMRA


ARn








PCR cycles


B) The P12 probe used the fluorochromes FAM6 and TAMRA

ARn








PCR cycles


C) The P12 probe used the fluorochromes TET and TAMRA


ARn








PCR cycles




Figure 5-3. Examples of real time PCR reactions. A) For the clone 109. B) For the P5
candidate gene. C) For the P12. The reactions were done with samples from
all the tested conditions and reflect the differences in the transcriptional level
of the target gene according to each sample.









Figure 5-4 displays the relative transcriptional levels of the selected clones. Within

each group of plants, regression analysis was performed and a polynomial equation was

shown. The three evaluated stages of the disease (healthy, mildly and fully blighted trees)

are certainly not enough for better inferences about the progress of the disease, neither

blight can be precisely estimated, since no pathogen is known to cause the problem. But

an initial tendency of the relative transcriptional levels compared to the healthy samples,

was observed for some of the clones.

The clone 6 can probably be considered a mistaken choice taken from the cDNA

array experiment. The tentative down-regulated pattern was not observed in the real time

PCR experiment. Moreover, it seems to be up-regulated in affected comparing to healthy

plants. More replicates, or maybe other techniques, would probably be needed for further

elucidation

The clone 25 was considered to be up-regulated in affected plants, especially when

Swingle rootstock was used (panel Cl, Figure4-4C). This pattern was also observed in

the real time PCR experiment using Carrizo rootstock, confirming the previous

expectation. In the group D of plants, the fully blighted tree displayed similar

transcriptional levels to the healthy tree, but both were lower than the mildly affected

one. It is possible that fully blighted trees reduce some of their metabolism affecting the

transcriptional level of certain genes. That was apparently the case for this gene in this

studied blighted tree. In the other three groups of plants, the relative amount of the

transcripts of the gene represented by the clone 25 were more abundant in fully blighted

than in healthy or mildly affected plants. Therefore, the clone 25 was considered to

represent an up-regulated gene under citrus blight conditions.














rep. D 15.31

4K t clone 6



3.33


clone 25


25.48
--- clone 109


1 0.02 3.05 26.11 5.9c
--------- 2-4 -2O4



A A 10.29





332 9. 12 .. 241 .204
- --


clone 149



P5 candidate





P12


Hcz Mcz Bcz
205 223 404

10.28+/-0.50


Hcz Mcz Bcz
201 225 402

13.50+/-1.74


Hcz Mcz Bcz
203 227 406

15.82+/-2.40


Hcz Mcz Bcz trees
265 229 454

22.39+/-2.02 18S Ct +/-StDev


Figure 5-4. Relative transcriptional levels of the selected genes. Twelve individual plants
were analyzed in four groups (rep. A, B, C and D), based on similar Ct values
for the 18S readings. Feeder root samples of healthy (H), mildly affected (M)
and fully blighted (B) trees on Carrizo (cz) rootstock were numbered (201 to
454) and used in this experiment. The columns represent the relative
transcriptional level of each clone, comparing to healthy samples. The bars
represent the residual standard deviation estimated for each group of plants.


rep. A 1.44


rep. B


fold
change
to
healthy
(1x)


4.50


4--











The clone 109 was up-regulated in three out of the four replicates, comparing the

healthy to any of the two blighted conditions. The pattern was similar in the groups B and

C. The mildly affected plant of the fourth replicate (group D) displayed even higher

transcriptional level than the fully blighted tree. In addition, this pattern of the second and

third replicates (groups B and C of plants) may represent only an initial or partial

outcome for a latter blight development that has yet to come, and could reduce the

transcripts of the gene represented by the clone 109 to a similar level observed in the

group D of plants. That seems possible because citrus blight can not be precisely

quantified so far, and therefore, the progression of the disease can not be completely

pursued either. The first replicate (group A of plants) displayed a non expected result.

The relative transcriptional level of the mildly affected plants was reduced to a range of

40% comparing to the healthy trees. To test whether or not that was an artifact of the

method, a new RT and PCR reactions were run with the same RNA samples. No further

discrimination was observed, however the residual standard deviation was higher. The

mildly affected tree still had around 40% of the transcripts compared to the healthy

sample while the blighted reached 1.4 the healthy sample. It is possible that this result

may represent noise caused by blight itself, or by another factor not studied in the system.

New samples from the field were not collected and tested because that would add a

contribution of the season of the year and temperature to the present model. In addition,

the sampled plants are probably in different stages of blight at this point, because all

studied samples were collected and processed six months ago, from January 9th to the

15th of this year.









Considering that the clone 109 was up-regulated in three out of four replicates; that

a similar pattern was observed; and that the level of induction compared to healthy plants

was above twenty fold in the groups C and D of plants; the clone 109 was considered to

represent a gene that is upregulated under citrus blight conditions, confirming the

expectation from the cDNA array experiment.

The clone 149 was variable but with no defined pattern considering the outcome of

the four groups of plants. On the cDNA array, the transcriptional level of the gene

represented by this clone was nearly unchanged by citrus blight. However, since the real

time PCR is more sensitive, discrepancies towards either way, being up or down-

regulated according to individual organisms, could be unfolded because in the cDNA

arrays the data tends to be compressed. That was apparently the case, and the clone 149

does seem variable, but according to other factor not studied.

The P5 candidate gene was responsive to citrus blight with a similar pattern in all

four replicates. It was more abundant in fully blighted trees than in any other studied

condition. Apparently the gene represented by the P5 candidate displays an ascendant

transcriptional response towards the fully blighted tree. This result may not parallel the

initial complexation of zinc seen in leaves and trunks of mildly affected trees (Albrigo

and Young, 1980). Therefore, the chitinase gene represented by the P5 candidate is

considered to be upregulated under citrus blight condition, confirming the similar

expectation observed with the clone 38 in the cDNA array experiment; however, it

remains to be investigated whether or not this candidate gene is the true P5 identified by

Taylor et al. (1996). It is worthy to note that the behavior of the true P5 is not known in

feeder root tissues, as used in this experiment.









The P12 displayed a significantly higher transcriptional level on mildly affected

plants compared to healthy or fully blighted trees. The pattern was similar in all four

groups of tested plants. It is possible that if more stages of blight were available, a wider

range of conditions would be better evaluated, and maybe a different polynomial

equation would better describe the phenomenon. But the pattern for a higher

transcriptional level in mildly affected trees was clear. The lower amount of P12

transcripts in fully blighted trees may account for the previous failure in detecting P12 in

the subtracted libraries and also in the cDNA array experiment. Both procedures

employed samples from fully symptomatic trees.

Although patterns were observed, the intensity of the responses was related to the

group of studied plants. Therefore, a general, or averaged, transcriptional level for each

gene was not estimated, because it may not meaningful.

The Potential Biological Meaning of the Clone 109

Regardless of what causes citrus blight, affected plants visually go to a declining

condition during the whole process. A closer look at the clone 109 reveals sequence

homology to a citrus EST annotated as a cold acclimated responsive gene from the

Washington Navel sweet orange (Citrus sinensis L. Osbeck cv. Bahia), in experiments

done in California (Close et al., 2003; in press; assession number CB293790.1), shown in

Figure 4-10 (Chapter 4), with an e-value of 9e-44.

It is known that plants respond to cold and to drought periods activating and

repressing responsive genes. Seki et al. (2001) found 5 drought specific inducible genes,

2 cold specific inducible genes and 16 drought and cold inducible genes using a

microarray with around 1,300 full length cDNAs ofArabidopsis thaliana. Citrus blight is

a xylem blockage problem, which ultimately may lead to water deficiency, and maybe,









that gene represented by the clone 109 is involved in this process. In addition, the clone

109 displayed mostly higher transcriptional levels under citrus blight condition, using

samples from different seasons of the year.

An alternative hypothesis could claim specific response to the citrus blight process,

rather than being induced only as an indirect effect of the internal water stress caused by

citrus blight. The clone 109 also had potential similarities to an ubiquitin subunit (Table

4-2, chapter 4) of the SCF complex (Skpl-Cullin-F-box protein). This complex is

involved in the ubiquitination of proteins and cell cycle regulation. The citrus blight

process can be seen as an accelerator of senescence. Young affected trees, of 5 to 10

years old, start to display an overall decline and lack of vigor, normally only seen on

healthy trees older than 50 years. The clone 109 could function in the ubiquitination

process, targeting other proteins for degradation, and or, impairing the normal cell cycle

in affected trees.

Experimental confirmation is certainly needed either way regarding its function and

involved pathways. But in the first scenario, if cold acclimation is a feasible goal in plant

genetic engineering programs, maybe 'hardening' rootstocks for citrus blight could be as

well. In the second scenario, experimental effort on the cell cycle regulatory process in

citrus can eventually offer perspectives to control citrus blight.

A Tentative Test to Verify the Effect of Cold and Drought Stresses

In order to evaluate whether or not the gene represented by the clone 109 responds

to cold and drought stresses, another experiment was assembled and run. Other clones

were also included. The major objective was to compare the transcriptional levels of the

selected genes, and determine whether they seem to be a specific response to citrus

blight, or only a secondary effect of the stress caused by the disease. However, moving









adult citrus plants to controlled conditions was not attempted. Greenhoused young trees,

of about one year old were used instead. They were exposed to cold (4C for 52 hours)

and drought stresses (no water for one week plus additional root airing for 24 hours).

Seedlings of Carrizo citrange (Citrus sinensis L.. Osb. x Poncirus trifoliata L. Raf.) were

chosen to equal the genotype evaluated in the previous real time PCR experiment. The

references, or control plants, were from the same lot of seedlings, but kept under normal

greenhouse conditions for the same period of time. Four independent replicates (i.e.

different seedling plants) were evaluated for each treatment. Feeder root tissues were

once again used. Figure 5-5 shows the results. The contrast of interest, to compare the

control, cold and drought treatments, is not orthogonal, and therefore, the t-Test can not

be used. However, the number of replicates was uniform, allowing the application of a

test to compare the means. Among the options, the Tuckey test was employed because of

the better discrimination of the means compared to other tests, like Sheffee, Dunnett or

Duncan (Banzatto and Kronka, 1992). The null hypothesis was no difference among the

treatments.

The clone 109 did not respond to cold and drought stresses under the studied

conditions. No significant effect (p<0.05) was seen for most of the other clones either,

including P12. The drought treatment only affected the gene represented by the clone 25.

Its transcriptional level was reduced to around one-third compared to the control. This

level of significance (p<0.05) implies that the maximum estimated chance of having

similar results caused only by chance is of only 5%. Lowering this probability to 1%

(p<0.01) makes this difference became not significant. In spite of which level of

probability should be considered, this reduction of the transcriptional level was not








60



expected. Blight induces xylem blockage; and consequently, a potential internal water


stress. Since the clone 25 was considered to be up-regulated in the cDNA array and in the


real time PCR experiment, it would also be expected to be unchanged or up-regulated


120 o 1o
100 0- -_---- -
080
060 ----- ----- -
040

00
seedlings drought cold

1 60
1 40 -------- ------- ---
120 ---- ------ 113
100 --- 100 084
080 -- -- -
060 ---
040 --- -------- -
020 ---- --- -
000
seedlings drought cold
1l i -------------------


S100


I 069


clone 6

A=0.41









clone 109

A=0.77


P5
065 candidate
0-- A=1.15


seedlings


20


80 038
60-
40
20

seedlings drought cold





1 00
1 00 -
100 --- --- ..
00 60
040 -
020 --- ----
000
seedlings drought cold
nn


Figure 5-5. Contrasts comparing the effect of cold and drought treatments on the relative
transcriptional level of the clone 109 and other clones. The Tuckey test was
employed and an observed significant difference was highlighted with a star.
The bars represent the minimum significant difference for the contrast (A
values, for p<0.05) of each clone.





under drought stress. Another experiment using adult trees under controlled conditions


may address this question. Similar reasoning may be considered for the other genes as


well. A definitive answer about the effect of major stresses would need adult trees under


clone 25

A=0.48









clone 149

A=1.16









P12

A=2.44









controlled conditions. Therefore, the responsive genes to citrus blight, and

apparently not affected by drought and cold stresses, seem to be at this point, only the

clone 109, the P5 candidate and the P12.

The Potential Effect of Redundancy and Gene Families

The selected clone 38 represents a citrus chitinase gene. Two genes (chitinase class

I, accession AB081944.1; and chitinase class II, accession Z70032.1) were matched to

the 235 bases of the clone 38 with the same e-value of le-87. The homology was seen

towards the 3'end of the subjected sequences (Figure 4-8, chapter 4). Chitinases are

enzymes that catalyze hydrolysis of chitin polymers, acting against plant intruders by

destroying chitin-containing cell walls. The difference between the class I and the class II

is based in the presence (I) or absence (II) of a N-terminal chitin binding domain (source:

http://us.expasy.org/cgi-bin/prosite-search-ac?pdoc00620, last accessed June 1, 2004).

Therefore, the clone 38 arrayed in the cDNA array membranes (chapter 4) could

represent any of the two described genes.

The 22 aminoacids of the P5 protein (accession AAB46813) had 80% identity to

the same chitinase class I, accession AB081944.1, with an e-value of 7e-05. The matched

region was however in the conserved 5' end of the subjected sequence. The Taqman

probe and primers were designed to span this region, covering 88 nucleotides long. The

Figure 5-1 displays the homolog regions between the sequences of the reference

chitinase, clone 38 and P5 candidate.

Therefore, it is not possible to discern precisely if the clone 38 and the P5 candidate

represent the same gene, or not, within the chitinase family. In addition, the sequence for

citrus is limited, but searching the genome of the model plant Arabidopsis thaliana, the

reference chitinase class I, accession AB081944.1, had 15 matches in the protein






62


databank (source: TAIR, http://www.arabidopsis.org/, last accessed June 1, 2004), with

e-values varying from 3e-7 to 4e-95. It is possible that this situation is similar for citrus,

with a high number of redundant genes, imposing another level of difficulty in the

evaluation of transcriptomes. The situation can be similar for P12, which had 17 matches

in the protein database of TAIR, all expansins, with e-values from 2e-04 to 5e-30. There

are two gama expansins with noted similarity to P12. Therefore, it is possible that more

copies of homolog P12 genes are also present in citrus plants.














CHAPTER 6
IDENTIFICATION OF AN ETIOLOGICAL CONTRAST POTENTIALLY
ASSOCIATED WITH THE CITRUS BLIGHT DISEASE

A crucial question in citrus blight research is the origin of the problem. Several

causal agent candidates and other theories have been examined, but none has proved to

be definitive for blight.

Each reported theory has been based on observed characteristics of the disease;

however blight is a complex problem, and different views and perspectives are possible

under different circumstances. For instances, the non-transmissibility of citrus blight by

canopy tissues was observed long ago (Rhoads, 1936; and COHEN, 1968), but

transmission by root grafting was later obtained (Tucker et al., 1984). Soils with higher

pH and levels of Ca were associated with severe incidences of blight in Florida

(Wutscher, 1989); however, blight does occur in acidic soils and examples are common,

especially in the Sao Paulo citrus belt. Soil born pathogens were associated to blight

(Nemec et al., 1982); but transmission of blight by soil replacement was not obtained

(Timmer and Graham, 1992). Xylellafastidiosa was considered to be the causal agent of

blight (Hopkins, 1988); but this xylem limited bacterium is naturally present in many

vascular plants, including citrus. A higher incidence of blight was observed after the

implementation of the nucellar programs in Florida and in Sao Paulo (Derrick and

Timmer, 2000), probably reflecting the new endophytic balance in the nucellar plants but

blight has also a high incidence on certain old line trees, and examples on 'Olimpia', and

on other clones of the 'Pera' sweet orange that were not pre-immunized against citrus









tristeza virus (CTV) are commons in Sao Paulo. The theory about a molecular origin for

blight suggested the involvement of defective signals transmitted from plant to plant

(Carlos et al,2000), but no experimental evidence for that have been found. Strains of

citrus tristeza virus (CTV) were reported to be associated with a potential variant of

blight (named Rangpur lime decline) and to citrus sudden death diseases in Brazil

(Derrick et al., 2003), but CTV is commonly present in stem-pitted sweet orange plants in

Sao Paulo causing no major disturbances (Costa et al., 1954).

In spite of the merit and investigations of each theory, one characteristic initially

proposed by Swingle and Webber (1896) seems difficult still to be denied: citrus blight

apparently has an infectious nature and dissemination. In order to investigate a potential

causal agent for citrus blight, qualitative and quantitative experiments were performed.

The First Screening of the Subtracted Libraries

When the subtracted libraries were made (chapter 3), during the summer of 2001,

the following question regarding which genes were represented there was first addressed

using a virtual northern blot experiment.

The differentially enriched cDNAs from the blighted minus healthy (B H) library

were cloned and grown in the E. coli vector and plated on LB medium. Ninety six clones

were randomly selected for re-growing on two equally printed nylon membranes. After

growth, the membranes were rinsed with extensive washes and hybridized overnight with

P32 labeled probes. The probes were also made from the subtracted blighted (B H) and

healthy (H B) enriched cDNA libraries, uncovering tentatively more abundant

transcripts represented there. A B H clone when revealed by the B H probes and not

by the H B probes implied a larger presence of the transcript of that gene in the blighted

samples that originated the library, from the affected trees. To the contrary, a B H clone









when revealed by both types of probes ( B H and H B ) indicated similar amounts of

the transcript in both libraries, and therefore, were of no interest. Virtual northern blots

can be effective to reveal abundant transcripts in enriched libraries (Source:

http://www.bdbiosciences.com/clontech/, last accessed January 15, 2004), which may

include those of RNA based viruses, such as the majority of the plant viruses.

Figure 6-1 displays the results for the virtual northern blot experiment. Among

others, the arrows indicate that the clones E8-13 and E8-14 were far more abundant in the

B H library.




B H probes H B probes



B-H
clones






Figure 6-1. The virtual northern blot of the B H clones. Each clone was printed in two
membranes that were respectively probed with B H and H B probes labeled
with P32. The subtracted libraries were made using the superficial roots of
Rough lemon (Citrus jambhiri Lush) collected in the central Florida area,
during the summer of 2001. The green arrows indicate the clones E8-13 and
E8-14.



Sixteen clones were considered to be differentially transcribed in the B H library

and were selected for sequencing. Among the plant genes, possible up regulated ones

included clones with similarities to metallothionein and several unknowns. However,









based on sequence homology, the clones E8-13 and E8-14 were not plant genes, but part

of the 3'end of the P27 divergent citrus tristeza virus (CTV) coat protein gene. The clone

E8-13 was 311 base pairs long and the E8-14 was revealed to be only a shorter and

redundant version of the same gene, with 155 base pairs. Table 6-1 displays the first four

nucleotide based homolog sequences found in all databases of Genbank for clone E8-13,

using the Blast-N search analysis (Altschul et al., 1997). Similar results were found using

the translated query searches.


Table 6-1. The clone E8-13 had homolog sequences matching different isolates of the
citrus tristeza virus (CTV). The clone E8-13 is 311 base pair long and was
found in the citrus blighted minus healthy (B-H) subtracted library made from
sup erficial roots of Rough lemon (Citrus jambhiri Lush).
clone Homolog gene e-values

E8-13 >gil1732493 gb|U56902.1|CTU56902 Citrus tristeza e-162
virus p346, 54-kDa RNA dependent RNA polymerase,
p33, p6, p65, p61, p27, 25-kDa coat protein (CPG),
p18, p13, p20, and p23 genes, complete cds

>gil 11414863|dbj|AB046398.11 Citrus tristeza virus e-147
genomic RNA, complete genome, seedling yellows
strain

>gi|20988251gblAF001623.1|CTAF001623 Citrus e-136
tristeza virus, complete genome

>gi|3550999|dbj|AB011189.1| Citrus tristeza virus 2e-87
genomic RNA for 27K protein and coat protein, partial
cds, isolate KS3A2



Looking at the homolog sequences, the first match, accession number U56902, was

isolated in Israel and represents the coat protein gene of CTV (Mawassi et al., 1993). The

second, accession AB046398, was a seedling yellows strain from Japan (Suastika et al.,









unpublished). The third, accession AF001623, was isolated in Texas causing severe

symptoms in sweet orange (Yang et al., 1999). The fourth, accession AB011189 is the

P27 gene also isolated from strains of Japan (Kano et al., unpublished).

Sequence information is limited for citrus, but to verify other potential origins for

the clone E8-13, other databases were searched. Homolog sequences were not found in

different plant databases. The first matched outcome in the blast searches (Altschul et al.,

1997) using the non-mouse and non-human ESTs and the viridiplantae databases of the

Genebank yielded respectively matched sequences with e-values of only 5.7 and 1.1.

Using the Brazilian citrus EST database, the result was not different, with an e-value of

only 0.18 for the first outcome. Therefore, the clone E8-13 was considered to be part of

the P27 gene of the CTV genome. The highest matched sequence indicated similarities to

an Israeli isolate.

In addition, both clones, E8-13 and E8-14, displayed polyadenylated 3'ends in their

sequences. The question whether that was an artifact of the method used to build the

cDNA libraries (Smart cDNA kit, source: http://www.bdbiosciences.com/clontech/, last

accessed June 20, 2001), or an adaptation of the CTV to the molecular machine of citrus

remains to be investigated.

The presence of CTV in citrus plants is not novel, and the solution for the Tristeza

disease in sweet orange groves was reported long ago (Costa et al., 1954). However, the

presence of CTV strains in roots of citrus plants affected by blight is intriguing. This

result was similar to previous observations made by Derrick et al.(unpublished). It is also

necessary to review that Rough lemon is susceptible to CTV and those findings alone

may not add novelty to what is already known about CTV and citrus host interactions.









Other CTV Genes were also Found in both Subtracted Libraries

Later, around four hundred clones randomly selected from the blighted minus

healthy (B-H) and 100 from the healthy minus blighted (H B) enriched libraries were

sequenced. Clones with sequence similarities to other CTV genes were relatively

common, matching different isolates and strains. Redundancy was observed, and

normally, more than one clone matched the same analyzed gene. The Table 6-2 displays

examples of pieces of CTV genes found in both libraries.



Table 6-2. Other sequences with homology to CTV genes found in the blighted minus
healthy (B-H) and healthy minus blighted (H-B) subtracted libraries.

library CTV gene e-value clone

B H p23 e-153 M1H7
p346RDRPol 6e-70 M2E2
p33 3e-43 M3A1

H B p61 e-131 E8-28
hsp90 p61 e-158 E8-21
p65 0.000 E8-22


Quantitative Evaluation of the P27 Candidate Gene in the Blighted Trees

In order to evaluate the presence of the transcripts of the P27 candidate gene,

reverse transcriptase quantitative real time PCR was run.

The chosen Taqman probe employed the 'Fam6-5'sequence 3'-Tamra' architecture

and had an accepted PCR efficiency (Chapter 5). The absolute value for the slope of the

linear function (y= -0.1008x +22.13), that described the log input amount of the total

RNA (x) against the ACt (y) between the target and the normalizer 18S, was 0.1008. The

tested range was a 2 fold dilution series from 3.9 to 1000 ng of total RNA. The







69


comparative ACt method was used to avoid the need of in vitro synthesis and purification

of the P27 RNA for absolute quantifications. For calculation purposes, reactions that did

not display amplifications were considered having a Ct value of 45 cycles. To minimize

non-specific variations, two reactions per sample were run. The same healthy, mildly

affected and fully blighted plants that were used to evaluate other genes (Chapter 5) were

used here. They were Valencia sweet orange (Citrus sinensis L.. Osbeck cv. Valencia) on

Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata L.. Raf.) of around 10 years

and from the central Florida area. Four groups of plants, or replicates, were evaluated.

The transcripts of the P27 candidate gene were present in far larger amounts in

affected than in healthy trees (Figure 6-2). The fully blighted trees on the group B of

plants still have around eight times more P27 transcripts than the healthy plants.




800,0000 400 0000 4000 300
70000000 350,000 - 3500 -
250 --
6,0,000 -- 300,0000 - - 3000 0 -

50000000 ---------- 250,0000 --- -- -2500 200 -------- 76
400,0000 -- 200,0000 - -74 200 0 - - 150 -
300,000 - - 150,0000 - 1500 142
100 -
200,000 ---------- 100,0000 ----- ----- 1000 ----- --
100,0000 ---------- 50,0000 ----- 500----- -- 50 -- --
10 08 10 88 10 10 05
Hcz Mcz Bcz Hcz Mcz Bcz Hcz Mcz Bczcz Mc B

Ct18S: 10.09+/-0.29 12.93+/-1.58 16.49+/-1.28 21.27+/-0.19


Figure 6-2. The transcripts of the p27 candidate gene were abundant in the roots of the
affected trees, Carrizo citrange (Citrus sinensis L. Osb. x Poncirus trifoliata
L. Raf.). The Ct levels for the 18S were displayed below the graphics;
columns of the blighted (B) and mildly affected (M) were compared to healthy
(H) Carrizo (cz) trees.









It is worthy to note that Poncirus trifoliata and its hybrids are considered to be

resistant to CTV (Deng et al., 2001).

The overall comparative amounts of the P27 transcripts indicated that CTV is

present in feeder roots of Carrizo citrange affected by citrus blight. Further experiments

can elucidate whether or not CTV causes or enhances citrus blight, or only grows better

in roots of already debilitated plants. Eventual synergistic or antagonistic effect on the

real causal agent of blight may also be investigated. Another possibility would be to

consider the clone E8-13 as part of another virus, with sequence similarities to CTV.

Another closterovirus would probably be the immediate suspect.














CHAPTER 7
CONCLUSIONS

Citrus blight has imposed consistent losses and changes to the citrus industry since

its origin, in the late of the XIX century (Swingle and Webber, 1896). The molecular

mechanisms involved in the plant responses are still unknown at this point.

Evidences for differentially transcribed genes, under different citrus blight

incidences, were observed in this study. Three genes, represented by the clones studied in

this work, had higher transcriptional level in blighted than in healthy trees. The level of

response was dependent of the evaluated group of plants but patterns were observed. This

study employed transcriptional assessment of citrus genes in feeder roots of healthy,

mildly affected and fully blighted trees. Under these studied conditions, P12 had higher

transcriptional level in mildly than in fully blighted trees. The chitinase(s) represented by

the P5 candidate sequence and by the clone 38 had higher levels in fully blighted than in

mildly or healthy trees. The clones 25 and 109 showed higher levels in fully blighted

compared to healthy samples. Research is needed to reveal the function of the P12 gene

and the genes represented by the other clones. Cloning of the true P5 gene and the genes

represented by the other clones can also be another task for future efforts.

It is also possible that more clones from the cDNA array experiment are truly

differentially transcribed under non and blighted plants, but more confirmation is needed.

In addition, the created subtracted libraries can be a wealthy source of clones for further

experiments. The suppressive subtraction method was an effective way to create enriched

cDNA libraries.









The finding of transcripts of CTV genes in roots of a blight-susceptible-CTV-

resistant rootstock immediately raises further etiological questions. Under this context,

CTV can intuitively be associated with citrus blight. However, whether or not it has a

synergistic or antagonistic effect on blight, it is not known yet. Another potential

consequence of this finding is regarded to options for tolerant rootstocks to be used not

only for blight but also against citrus tristeza disease. Carrizo citrange may fall in

discredit, contributing to the already known need of research on citrus rootstocks.

Finally, it is possible to say that the end is certainly not near for blight. Research in

roots seems the logical alternative to study the problem. In this way, it is hoped that the

knowledge gained during this study can be of some use, helping to understand and

control citrus blight.















APPENDIX
BLAST ANALYSIS BASED ON NUCLEOTIDE SEQUENCES OF EACH
INDIVIDUAL CLONE


Table A-1. Blast analysis based on nucleotide sequences of each individual clone. The
clones were obtained from the subtracted libraries (chapter 3) and used in the
cDNA array experiment (chapter 4). Three databases were used for the search
and the outcomes were highlighted in black (for the All Genebank with
1,367,736 sequences), blue (for the Brazilian citrus CCSM ESTs with 13,610
clusterized sequences) and green (for ESTs of the non-human non-mouse
Genebank with 4,893,238 sequences). The clones 7 to 12 were obtained from
other libraries of citrus leaf tissues, and all the others were from the subtracted
libraries of root tissues, as described in chapter 3.

e-
Clone Highest match description/score/e-value values Organism Function

>gi|806738|gb|U16304.1|CTU16304 Citrus
5 tristeza virus complete genome e-132 CTV p61

>gi|1785673 emb|YO8501.1|MIATGENA
6 A.thaliana mitochondrial genome, part A 0 Arabidopsis mitochondrial

Contig204 7e-11 Citrus

>gi|1673661gb|L08199.1| COTPROXDS
Gossypium hirsutum peroxidase mRNA,
7 complete cds 6E-68 Gossypium peroxidase

Contig373 e-173 Citrus

1 gb|BQ624415.1|BQ624415 USDA-
FP_01506 Ridge pineapple sweet orange e-171 Citrus

>gi|6469118|emblAJ275306.1|CAR275306
Cicer arietinum partial mRNA for
8 mitochondrial phosphate 4E-26 Cicer mitochondrial

CSJE01-038D08.g 6e-88 Citrus

1 gblBQ623911.1|BQ623911 USDA-
FP_00991 Ridge pineapple sweet orange 6e-83 Citrus


>gi|184016341reflNM 112657.11
Arabidopsis thaliana chromosome 3
9 CHR3v07142002 genomic sequence


6E-43 Arabidopsis unknown











Contigl602


1 dbj|C22180.1|C22180 C22180 Miyagawa-
wase satsuma mandarin orange


e-153 Citrus


4e-84 Citrus


10 cxContig94 e-146 Citrus

1 gb|BQ623982.1|BQ623982 USDA-
FP_01073 Ridge pineapple sweet orange e-145 Citrus


11 CSJE01-038D08.g 2e-94 Citrus

1 gb|BQ623911.1|BQ623911 USDA-
FP_00991 Ridge pineapple sweet orange 2e-84 Citrus


12 Contig2919 8e-27 Citrus

1 gb|BQ623455.1|BQ623455 USDA-
FP_00546 Ridge pineapple sweet orange 3e-65 Citrus


13 Poor matches unknown

14 Poor matches unknown

15 CSContig70 2e-20 Citrus

cxAC01-022F03.g
16 52 7e-08 7e-08 Citrus

1 dbj|C22172.1|C22172 C22172 Miyagawa-
wase satsuma mandarin orange... le-13 Citrus


17 Poor matches unknown

18 Poor matches unknown

19 Poor matches unknown

20 Contigl687 7e-64 Citrus

1 gb|BQ625123.1|BQ625123 USDA-
FP_02214 Ridge pineapple sweet orange 9e-66 Citrus


1 gb|BQ624910.1|BQ624910 USDA-
21 FP_02001 Ridge pineapple sweet orange e-135 Citrus


22 cxContigl377


6e-71 Citrus












23 Poor matches

24 Poor matches

25 Contig416

26 Poor matches

27 Poor matches

28 LPACOO-013G11.g

29 Poor matches

30 Poor matches

31 Contig416

32 Poor matches

33 Contig2752

>gi|7228328|emb|Y18788.1| MSY18788
Medicago sativa mRNA for putative TFIIIA
34 (or kruppel)-like zinc finger protein

cxContigl077


unknown

unknown


le-91 Citrus


unknown

unknown


2e-06 Citrus


unknown

unknown


e-101 Citrus

Human

le-10 Citrus


2E-14 Medicago

6e-07 Citrus


unknown


Zn finger


35 Contig2879 0.0 Citrus

36 Contig2919 9e-65 Citrus

37 Contig2879 3e-53 Citrus

38 Contig1434 3e-95 Citrus

39 Poor matches unknown

40 Poor matches unknown

41 Poor matches unknown

>gi| 1220143 emb|Z70032.1| CSACHIT2
42 C.sinensis mRNA for class II acidic chitinase 6E-79 Citrus chitinase

Contigl512 le-81 Citrus










>gi| 1220143 emb|Z70032.1| CSACHIT2
43 C.sinensis mRNA for class II acidic chitinase

Contigl434


44 Contig2879


45 FCContigl27


46 Contig2919


47 Contig2879


48 Contig2879

49 Poor matches

>gi|599725|emb|Z46824. 1|CSLEA5PMB
50 C.sinensis mRNA for Lea5 protein

Contig3802


51 No significant similarity found


52 Contig2879


53 Contig2919

>gil59177841gblAF184068.1|AF184068
Citrus limon vacuolar membrane ATPase
54 subunit G (LVMA10) mRNA,

Contig2879


55 Contig308


56 Contig2879

>gi|208093051gb|BC029618.11 Homo
sapiens, glyceraldehyde-3 -phosphate
57 dehydrogenase, clone

Contig80


8E-40

le-43


2e-62


5e-56


2e-34


5e-63


6e-50


Citrus

Citrus


Citrus


Citrus


Citrus


Citrus


Citrus


2E-90 Citrus

4e-86 Citrus


chitinase


unknown


LEA


unknown


e-100


2e-53


Citrus


Citrus


2E-28 Citrus ATPase

e-107 Citrus


e-121 Citrus


e-130 Citrus



e-128 Human G3PDH

le-04 Citrus










>gi|58153121gb|AF176034.1|AF176034
Coliphage phiX174 isolate Anc, complete
58 genome e-171 Coliphage unknown


59 cxJE01-111G08.g 0.0 Citrus


60 CXAC02-065D12.g le-42 Citrus


61 cxJE01-085A10.g 0.0 Citrus

>gi|21206806|gb|AY103728.11 Zea mays
62 PCO142212 mRNA sequence 5E-19 Zea unknown

Contig1287 e-162 Citrus

>gi|6653735|gblAF209908.1|AF209908
63 Prunus dulcis unknown mRNA 3E-07 Prunus unknown

Contigl6 e-110 Citrus

>gi|17324931gb|U56902.1|CTU56902 Citrus
tristeza virus p346, 54-kDa RNA dependent
RNA polymerase, p33, p6, p65, p61, p27, 25-
kDa coat protein (CPG), p18, p I p2 1 and
p23 genes, complete cds, p20, and p23 genes,
64 complete cds e-162 CTV p27

>gi|3308979|dbj|ABOO8100.1| Citrus unshiu
mRNA for metallothionein-like protein,
65 complete cds e-138 Citrus metallothionein

Contig3121 e-142 Citrus



66 Poor matches unknown

>gi|806738|gb|U16304.1|CTU16304 Citrus
67 tristeza virus complete genome e-130 CTV p25/p27

>gi|806738|gb|U16304.1|CTU16304 Citrus
68 tristeza virus complete genome e-160 CTV hsp90 p61

>gi|4239714|emblY18420.1| CITV18420
Citrus tristeza virus complete genome, isolate
69 T385 0 CTV p65

>gi|806738|gb|U16304.1|CTU16304 Citrus
70 tristeza virus complete genome e-127 CTV p65





unknown


8e-90 Citrus


3e-96 Citrus


75 Poor matches


77 cxContig629

1 gb|BQ623985.1|BQ623985 USDA-
FP_01076 Ridge pineapple sweet orange


78 Poor matches


79 Poor matches


80 Poor matches


81 Poor matches


82 Poor matches



83 Poor matches


84 Poor matches


85 Poor matches


86 Poor matches


87 Poor matches


88 Poor matches


89 Contig2919


90 Poor matches


91 Poor matches


le-31 Citrus


unknown


unknown


unknown


unknown


unknown


unknown


unknown



unknown


unknown


unknown


unknown


unknown


unknown












Poor matches


Poor matches


Contig2919

1 gb|BQ623148.1|BQ623148 USDA-
FP_00239 Ridge pineapple sweet orange

>gi|67289521gb|AC020576.2|T12C22
Sequence of BAC T12C22 from Arabidopsis
thaliana chromosome 1, complete


Poor matches


Poor matches


Poor matches

>gi|9087297|dbj AP000397.1|AP000396S2
Beta vulgaris mitochondrial genomic DNA,
complete sequence, section 2/2

>gi|9087297|dbj AP000397.1|AP000396S2
Beta vulgaris mitochondrial genomic DNA,
complete sequence, section 2/2



Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


unknown


unknown


2e-46


le-77


Citrus


Citrus


1E-08 Arabidopsis


unknown


unknown



unknown


unknown


e-125 Beta unknown



e-114 Beta mitochondrial



unknown


unknown


unknown


unknown


unknown


unknown


unknown


99



100





unknown







8e-44 Citrus


108 Poor matches


>gil286192471gb|CB293790.1|C B "',,"
UCRCS01_06cg09_gl Washington Navel
orange cold acclimated flavedo & albedo
cDNA library Citrus sinensis cDNA
109 clone...8e-44


110 Poor matches


111 Poor matches


112 Poor matches


113 Poor matches


114 Poor matches

1 dbj|C95562.1|C95562 C95562 Citrus
unshiu Miyagawa-wase maturatio... 331 2e-
115 88


2e-88 Citrus


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


Poor matches


119


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown


unknown













Poor matches


Poor matches


Poor matches


Poor matches


unknown


unknown


unknown


unknown


135 Poor matches unknown


136 FAAC01-035B05.g le-30 Citrus

>gil 12249|emb|X03775.1|CHSOATP1
Spinach plastid genes atpl-H-F for ATP
137 synthase CF(O) subunits IV, 6E-86 Spinach ATP synthase


138 Poor matches unknown


139 Poor matches unknown

1 gblBQ623784.1|BQ623784 USDA-
FP 00875 Ridge pineapple sweet orang...
140 220 6e-55 6e-55 Citrus

>gi|66937951gblAF112970.1|AF112970
Daucus carota strain Imperator STS3A
141 mitochondrial DNA segment 3E-74 Daucus mitochondrial

1 gb|BE460852.1 BE460852 EST412271
tomato breaker fruit, TIGR Lyco... 5e-68 Tomato

1-gb|BM371429.21BM371429
EBma08_SQ002_L04_R maternal, 28 DPA,
142 no t... 9e-11 unknown


143 Poor matches unknown


144 Poor matches unknown


145 Poor matches unknown










>gi|18857892|dbj|AB061306.1| Citrus
jambhiri mitochondrial ACRS gene for ACR
146 toxin-sensitivity 4E-48 Citrus ACR toxin-sen.

Contig4126 3e-16 Citrus
1 gb|BM358396.1|BM358396
GA Ea0008M10r Gossypium arboreum 7-10
dp... 351 2e-94 2e-94 Gossypium


147 Poor matches unknown

>gi|5688942|dbj|ABO17426.11 Oryza sativa
(japonica cultivar-group) mitochondrial gene
148 for ribosomal protein L5, complete cds 2E-59 Oryza mitochondrial

1 dbj AV420567.1|AV420567 AV420567
Lotus japonicus young plants (t... 315 7e-84 7e-84 Lotus


149 CXJE02-097F04.g 2e-54 Citrus


150 CXJM02-089E08.g 2e-19 Citrus

1 gb|BM371381.21BM371381
EBma08_SQ002_I05 R maternal, 28 DPA, no
151 t... 86 le-14 le-14


152 CXContig831 3e-92 Citrus

153 Poor matches unknown



154 Poor matches unknown

1 gb|BI180911.1|BI180911 TY3H09
hepatocellular carcinoma expressio... 90 le-
155 15 le-15


156 Poor matches unknown


1 gb|BI180911.1|BI180911 TY3H09
157 hepatocellular carcinoma expressio... 3e-14

1 gb|BM376133.1|BM376133
EBma01_SQ002_H07_R maternal, 4 DPA, no
158 tr... 4e-12










1 gb|BM371429.21BM371429
EBma08_SQ002_L04_R maternal, 28 DPA,
159 no t... 7e-08


160 Poor matches unknown


1 gb|T44610.1|T44610 7873 Lambda-PRL2
161 Arabidopsis thaliana cDNA cl... 359 le-96 le-96 unknown

1 gb|BM371429.21BM371429
EBma08_SQ002_L04_R maternal, 28 DPA,
162 no t... 84 9e-14 9e-14

1 gb|BM371429.21BM371429
EBma08_SQ002_L04_R maternal, 28 DPA,
163 no t... 72 2e-10 2e-10

>gi112830831 gblAF320906.1|AF320906
Citrus unshiu metallothionein-like protein
164 (MT45) gene, complete cds 3E-90 Citrus metallothionein

Contig2243 2e-92 Citrus

1 gb|BQ624047.1|BQ624047 USDA-
FP_01138 Ridge pineapple sweet orang... le-86 Citrus

1 gblBQ623549.1|BQ623549 USDA-
FP_00640 Ridge pineapple sweet orang... 80
165 3e-13 3e-13 Citrus unknown


1 gb|BQ624621.1|BQ624621 USDA-
166 FP_01712 Ridge pineapple sweet orange le-96 Citrus


167 Poor matches unknown


1 emb|AL729032.1|AL729032 AL729032
168 Danio rerio embryonic inner ear... 58 3e-06 3e-06

1 gb|BQ414420.1|BQ414420
GA Ed0086EO5r Gossypium arboreum 7-10
169 dp... 60 2e-07


170 Poor matches unknown















LIST OF REFERENCES


Agrios GN (1997) Introduction to plant pathology. In GN Agrios, ed, Plant Pathology,
Ed 4. Academic Press Inc, pp 3-41

Albrigo LG, Syvertsen JP, Young RH (1986) Stress symptoms of citrus trees in
successive stages of decline due to blight. J. Am. Soc.Hort. Sci. 111: 465-470

Albrigo LG, Timmer LW, Derrick KS, Tucker DPH, Graham JH (1993) Failure to
transmit citrus blight by limb grafts. In International Organization of Citrus
Virologists, Vol 12. University of California, New Delhi, pp 127-130

Albrigo LG, Young RH (1980) Phloem zinc accumulation in citrus trees with blight.
Hortscience 15: 394-394

Albrigo LG, Young RH (1981) Phloem zinc accumulation in citrus trees affected with
blight. Hortscience 16: 158-160

Allen M (2000) A seed is planted. In: The history of Florida citrus. In Florida Grower,
Vol mid-August edition, pp 10-13

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ
(1997) Gapped BLAST and PSI-BLAST: a new generation of protein database
search programs. Nucleic Acids Res 25: 3389-3402

Banzatto DA, Kronka SN (1992) Experimentagdo Agricola. FUNEP-UNESP,
Jaboticabal

Berger R (1998) A causa e o control do declinio dos Citros. Laranja 19: 91-105

Brazma A, Hingamp P, Quackenbush J, Sherlock G, Spellman P, Stoeckert C, Aach
J, Ansorge W, Ball CA, Causton HC, Gaasterland T, Glenisson P, Holstege
FC, Kim IF, Markowitz V, Matese JC, Parkinson H, Robinson A, Sarkans
U, Schulze-Kremer S, Stewart J, Taylor R, Vilo J, Vingron M (2001)
Minimum information about a microarray experiment (MIAME)-toward
standards for microarray data. Nat Genet 29: 365-371









Bausher MG (1990) Electrophoretic and immunological evidence of unique proteins in
leaves of citrus trees: application to citrus blight detection. Electrophoresis 11:
830-834

Callies T (2000) Reclaiming Florida's water. In Florida Grower, Vol September edition,
pp 8-12

Ceccardi TL, Barthe GA, Derrick KS (1998) A novel protein associated with citrus
blight has sequence similarities to expansion. Plant Mol Biol 38: 775-783

Chuaqui RF, Bonner RF, Best CJ, Gillespie JW, Flaig MJ, Hewitt SM, Phillips JL,
Krizman DB, Tangrea MA, Ahram M, Linehan WM, Knezevic V, Emmert-
Buck MR (2002) Post-analysis follow-up and validation of microarray
experiments. Nat Genet 32 Suppl: 509-514

Cohen M (1968) Citrus blight and "blight-like" diseases. Citrus Industry 47: 12-26

Cohen M (1974) Diagnosis of young tree decline, blight and sand hill decline of citrus by
measurement of water uptake using gravity injection. Plant Disease Reporter
58: 801-805

Cohen M, Pelosi RR, Brlansky RH (1983) Nature and location of xylem blockage
structures in trees with citrus blight. Phytopathology 73: 1125-1130

Colebatch G, Kloska S, Trevaskis B, Freund S, Altmann T, Udvardi MK (2002)
Novel aspects of symbiotic nitrogen fixation uncovered by transcript profiling
with cDNA arrays. Mol Plant Microbe Interact 15: 411-420

Costa AS, Grant TJ, Moreira S (1954) Behavior of various citrus rootstock-scion
combinations following inoculation with mild and severe strains of tristeza
virus. In Florida State Horticultural Society, Vol 67, Miami Beach, pp 26-30

Deng Z, Huang S, Ling P, Yu C, Tao Q, Chen C, Wendell MK, Zhang HB, Gmitter
FG, Jr. (2001) Fine genetic mapping and BAC contig development for the
citrus tristeza virus resistance gene locus in Poncirus trifoliata (Raf.). Mol Genet
Genomics 265: 739-747

Derrick KS, Beretta MJ, Barthe GA, Kayim M (2003) Strains of citrus tristeza virus
predominately found in roots with implications for citrus blight and citrus
sudden death. Phytopathology 93: S20

Derrick KS, Beretta MJ, Barthe GA, Kayim M, Harakava R (2003) Identification of
strains of Citrus tristeza virus by subtraction hybridization. Plant Disease 87:
1355-1359









Derrick KS, Lee RF, Brlansky RH, Timmer LW, Hewitt BG, Barthe GA (1990)
Proteins associated with citrus blight. Plant Disease 74: 168-170

Derrick KS, Timmer LW (2000) Citrus blight and other diseases of recalcitrant
etiology. Annu Rev Phytopathol 38: 181-205

Diatchenko L, Lau YF, Campbell AP, Chenchik A, Moqadam F, Huang B,
Lukyanov S, Lukyanov K, Gurskaya N, Sverdlov ED, Siebert PD (1996)
Suppression subtractive hybridization: a method for generating differentially
regulated or tissue-specific cDNA probes and libraries. Proc Natl Acad Sci U S
A 93: 6025-6030

Donson J, Fang Y, Espiritu-Santo G, Xing W, Salazar A, Miyamoto S, Armendarez
V, Volkmuth W (2002) Comprehensive gene expression analysis by transcript
profiling. Plant Mol Biol 48: 75-97

Eisen MB, Spellman PT, Brown PO, Botstein D (1998) Cluster analysis and display of
genome-wide expression patterns. Proc Natl Acad Sci U S A 95: 14863-14868

Fanta N, Ortega X, Perez LM (2003) The development ofAlternaria alternate is
prevented by chitinases and beta-1,3-glucanases from citrus limon seedlings.
Biol Res 36: 411-420

Fawcett HS, Lee HA (1926) Citrus Diseases and their Control, 1st. Ed. McGraw-Hill,
New York. [etc.]

Fegeros K, Zervas G, Stamouli S, Apostolaki E (1995) Nutritive value of dried citrus
pulp and its effect on milk yield and milk composition of lactating ewes. J Dairy
Sci 78: 1116-1121

Gaasterland T, Bekiranov S (2000) Making the most of microarray data. Nat Genet 24:
204-206

Hopkins DL (1987) Xylem-limited bacteria cause blight symptoms in citrus.
Phytopathology 77: 641-641

Hopkins DL (1988) Production of diagnostic symptoms of blight in citrus inoculated
with Xylellafastidiosa. Plant Disease 72: 432-435

Lee RF, Marais LJ, Timmer LW, Graham JH (1984) Syringe injection of water into
the trunk a rapid diagnostic-test for citrus blight. Plant Disease 68: 511-513

Lewandowski M (2000) Formulating frozen concentrate. In: The history of Florida
Citrus. In Florida Grower, Vol mid-august edition, pp 44-48









Lindbeck AGC, Brlansky RH (1998) Xylem plugging in feeder roots from blight-
affected citrus trees. Phytopathology 88: S54

Lindbeck AGC, Brlansky RH (2000) Cytology of fibrous roots from citrus blight-
affected trees. Plant Disease 84: 164-167

Mackay IM, Arden KE, Nitsche A (2002) Real-time PCR in virology. Nucleic Acids
Res 30: 1292-1305

Marais LJ, Lee RF (1990) Experimental transmission of citrus blight in South Africa. In
International organization of citrus virologists. University of California, pp 261-
264

Mawassi M, Gafny R, Bar-Joseph M (1993) Nucleotide sequence of the coat protein
gene of citrus tristeza virus: comparison of biologically diverse isolates
collected in Israel. Virus Genes 7: 265-275

Moreira S (1980) Hist6ria da citricultura no Brasil. In 0 Rodriguez, Viegas, F, ed,
Citricultura Brasileir. Fund. Cargill, Campinas, pp 1-28

Nemec S (1994) Stress-related compounds in blight-diseased citrus xylem fluid
associated with Fusarium solani naphthazarin toxins. Phytopathology 84: 870

Nemec S, Bustillo B, Obannon JH, Patterson M (1982) Effects of fungicides and
nematicides on citrus blight in Florida. Phytopathology 72: 360-360

Oliveira AC, Vallim MA, Semighini CP, Araujo WL, Goldman GH, Machado MA
(2002) Quantification of Xylellafastidiosa from citrus trees by real-time
polymerase chain reaction assay. Phytopathology 92: 1048-1054

Paiva LV, DeSouza M, Lopes MA, Paiva E (1997) Identification and isolation of
proteins found only in plants with citrus decline. Pesq. Agr. Bras. 32: 559-564

Pompeu JR. J (2001) Rootstocks and scions in the citriculture of the Sdo Paulo state. In
World Congress of the International Society of Citrus Nurseryman, 6th., pp 75-
82

Rafter JJ (2002) Scientific basis of biomarkers and benefits of functional foods for
reduction of disease risk: cancer. The British Journal of Nutrition 88: s219-s224

Rhoads AS (1936) Blight: a non-parasitic disease of citrus trees. University of Florida
Agricultural Experiment Station, Gainesville, Fla.

Rizhsky L, Liang H, Mittler R (2002) The combined effect of drought stress and heat
shock on gene expression in tobacco. Plant Physiol 130: 1143-1151









Rossetti VV, Beretta MJG, Teixeira ARR (1991) Experimental transmission of
declinio by approach-root-grafting in Sdo Paulo State, Brazil. In International
organization of citrus virologists, Vol 11. University of California, Orlando, pp
250-255

Scheideler M, Schlaich NL, Fellenberg K, Beissbarth T, Hauser NC, Vingron M,
Slusarenko AJ, Hoheisel JD (2002) Monitoring the switch from housekeeping
to pathogen defense metabolism in Arabidopsis thaliana using cDNA arrays. J
Biol Chem 277: 10555-10561

Seki M, Narusaka M, Abe H, Kasuga M, Yamaguchi-Shinozaki K, Carninci P,
Hayashizaki Y, Shinozaki K (2001) Monitoring the expression pattern of 1300
Arabidopsis genes under drought and cold stresses by using a full-length cDNA
microarray. Plant Cell 13: 61-72

Seki M, Shinozaki K, Ishida J, Nakajima M, Enju A, Sakurai T, Satou M, Akiyama
K, lida K, Oono Y (2003) [Arabidopsis functional genomics using full-length
cDNAs]. Tanpakushitsu Kakusan Koso 48: 1890-1898

Stoeckert CJ, Jr., Causton HC, Ball CA (2002) Microarray databases: standards and
ontologies. Nat Genet 32 Suppl: 469-473

Swingle WT, Webber HJ (1896) The principal diseases of citrus fruits in Florida. In GP
Office, ed. USDA, p 50

Taylor KC, Albrigo LG, Chase CD (1996) Purification of a Zn-binding phloem protein
with sequence identity to chitin-binding proteins. Plant Physiol 110: 657-664

Taylor KC, Ellis DR (1995) A zinc-binding protein in citrus with homology to plant
chitinases. Hortscience 30: 900

Taylor KC, Ellis DR, Paiva LV (2002) Purification of a zinc binding protein from
xylem of Citrus jambhiri. Journal of the Am. Soc. Hort. Sci. 127: 718-723

Timmer LW, Graham JH (1992) Nontransmission of citrus blight by soil. Plant Disease
76: 323-323

Timmer LW, Lee RF, Brlansky RH, Graham JH, L.G. A, Derrick KS, Tucker DPH
(1992) The infectious nature of citrus blight. In Florida State Horticultural
Society Annual Meeting, Vol 105, Orlando, pp 21-26

Tucker DPH, Lee RF, Timmer LW, Albrigo LG, Brlansky RH (1984) Experimental
Transmission of Citrus Blight. Plant Disease 68: 979-980

van den Boom J, Wolter M, Kuick R, Misek DE, Youkilis AS, Wechsler DS,
Sommer C, Reifenberger G, Hanash SM (2003) Characterization of gene