|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
CHARACTERIZATION OF THE HSP70 PROTEIN HOMOLOG (HSP70h) OF CITRUS
INES-MARLENE ROSALES VILLAVICENCIO
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
To my parents, Ines and Moises
I would like to express my gratitude to all the people that supported my graduate
studies at the University of Florida. I am especially grateful to my former major professor
Dr. C.L.Niblett, for his encouragement and friendship along all these years. I would also
like to thank my present advisor, Dr. Richard Lee, who gave me guidelines and advice for
my research during my entire program, and for being one of the few people who showed
a permanent interest for my condition as graduate student. I would like to thank Drs. K.
Derrick, Dr. R. Brlansky, and G. Moore for serving in my committee and for their helpful
suggestions and comments and for reviewing this manuscript. I would like to extend my
gratitude to Debbie Howd and Dianne Anchor for the technical assistance during the EM
work, to Dr. J. Grosser for letting me use the facilities at his laboratory and to Dr. Rachel
Shireman for her special support and dedication to graduate students in our lab.
I want to give special thanks to the former and current members of the Niblett's
and Lee's labs, V. J. Febres, R. Chandrika, K. L. Manjunath, J. Vazquez, R. Harakava,
H.Genc, Y. Petersen and M. Dekkers for the limitless help, understanding and friendship
that they gave me in the past years. I want to extend my gratitude to all the sincere and
unconditional friends that I have made in Gainesville and Lake Alfred, who gave me
affection and unconditional support.
My deepest thanks and appreciation go to my parents and family, for the constant
support and unconditional love they have provided throughout my life.
My deepest thanks and appreciation go to my parents and family, for the constant
support and unconditional love they have provided throughout my life.
Finally, I thank INIA (Instituto de Investigaciones Agropecuarias) for its generous
financial support during my graduate studies at the University of Florida. A special
recognition goes to Dr. Sergio Bonilla, who took great care of all the administrative
issues with UF during this time. Thanks go to Carlos Mufioz and Emilio Ruz for their
support and help. My gratitude goes to the people working at "INIA-La Platina" for
being my friends from the distance.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ............................ ............................................ .....................iii
L IST O F T A B L E S ............................... .......................... ......... ..... .......... vii
LIST OF FIGURES ............................... ............ .............. ..... ....... viii
A B ST R A C T .......................................................... ................ .. x
1. CITRUS TRISTEZA VIRUS: THE DISEASE, THE CAUSAL AGENT AND ITS
C H A R A C T E R IS T IC S .................................................................... ............................
The Tristeza D disease and Its C ausal A gent .................................................................... 1
T h e D disease .................. ...................................................... ............... 2
Cytopathology of CTV-infected Tissue ........................................ .............. 3
N natural R resistance to C T V ................................... ............................................. 5
Molecular Characteristics of CTV ................... ........ ................. 7
Genomic Organization and Replication Strategies............................ .............. 7
RNA Populations in Infected Tissue ................................. ........................ 9
LMT RNAs ............................................ 10
Defective RNAs ..................................... .................. .. 10
Subgenom ic RN A s ................... ................................... ...... .. ............ .. 13
Population Structure and Genetic D iversity.................................... .................... 14
2. PRODUCTION OF A POLYCLONAL ANTISERUM AGAINST THE
CARBOXY-TERMINAL END OF THE CTV HEAT SHOCK PROTEIN
H O M O L O G (H SP 70h) ..................................................................... ..................... 20
Introduction....................... ............... ..... ............. 20
M material and M ethods................... ................................................ ........... .............. 22
R results .......... ...... ... .. .......... ........ ...................... .............. 29
Sequence A nalysis............................................... .. .... ........ ...... ............... ... .. .. 29
Expression and Purification of the Carboxy-Terminal End of the CTV-HSP70h.... 31
Production of the Polyclonal Antiserum....... .................. ............. 43
D isc u ssio n .......................................................... ............... 4 5
3. IN VIVO LOCALIZATION OF THE HSP70 PROTEIN HOMOLOG (HSP70h) IN
CITRUS TRISTEZA CLOSTEROVIRUS INFECTED PLANTS ..............................49
Introduction............................... ........... .......... 49
M material and M ethods................... ................................................ ........... .............. 52
R results .................................................................... ............... ... ...... 54
In Vivo Detection of the HSP70h by Tissue Printing............. .................. ......... 54
Immunoprecipitation of an HSP70h-CP Complex from CTV-infected Plants......... 56
SSEM -Im m unogold Labeling ...................... .... ............................... .............. 57
D isc u ssio n .............................................................................................. ............... ..... 6 0
4. THE CTV-HSP70h AS A COMPONENT OF CTV INCLUSION BODIES ...............64
Introduction .......................................... 64
M material and M ethods................... ................................................ ........... .............. 65
R results ........................................................ ................. 69
Light Microscopy and Inclusion Body Purification........................................... 69
Fluorescent Antibody Microscopy ................................................. .............. 71
Analysis by SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)...................... 73
W western B lot A nalysis.......................................... .......................... .... ............ .. 76
D iscu ssion.............................. .................... 77
5. AGROBACTERIUM-MEDIATED TRANSFORMATION OF DUNCAN
GRAPEFRUIT (Citrus paradise; Macf)............................................................. 80
Introduction............................... ........... .......... 80
M material and M ethods................... ................................................ ......................... 85
R results ................... ......... ..... ......... .......... ................ ......... 91
Constructs Used in the Transformation Experiments ........................................ 91
Transformation and Regeneration of Transgenic Plants........................................ 95
PCR Assay of the Putative Transgenic Plants..................... .................... .. 98
Conclusions ............................................. 100
L IST O F R E FE R E N C E S ........................................................................ ...................102
BIO GRAPH ICAL SK ETCH ......... ............................ ......................... ............... 16
LIST OF TABLES
2-1. Sequence of the primers used for the RT-PCR and cloning of the HSP70h from
C T V ................... ......... ................................................ ................ 2 4
2-2. Nucleotide identity among the CTV-HSP70h genes of different isolates of citrus
tristeza virus .............. ............ .... ..... ....... .. .... ..........32
2-3. Amino acid identity among the CTV-HSP70h proteins expressed by different
isolates of citrus tristeza virus CTV-HSP70h proteins .............. .................................. 32
2-4. Biological properties and origin of the citrus tristeza virus (CTV) isolates included
in the alignment of the amino acid sequences of their HSP70h proteins ................. 34
5-1. Set of primers used for PCR assay of the putative transgenic plants ...........................91
5-2. Summary of the results of Duncan grapefruit transformation experiments
performed with the constructs containing the CTV-HSP70h, the frameshift
mutant (HSP70h-HindIII), or the binary vector by itself .................... ............... 98
LIST OF FIGURES
1-1. Sym ptom s caused by citrus tristeza virus ....................................................................... 4
1-2. Representation of the gene expression and genome organization of citrus tristeza
virus .... ............................................................... ......... 8
2-1. Nucleotide sequence for the HSP70h gene from the grapefruit stem pitting CTV
isolate T3800 ..................................................33
2-2. Alignment of the amino acid sequences of the HSP70h proteins from several CTV
isolates. ................... ................... ............35...........
2-3. Cluster dendrogram based on the amino acid sequences of the translated p65 gene for
the various s C T V isolates........... .......................................................... .. .... ..... ... 4 1
2-4. Domain conservation between HSP70s and HSP70h proteins............................... 41
2-5. Silver stained SDS-polyacrylamide gel electrophoresis (PAGE) showing the over-
expression of the 149 amino acid fragment fusion protein in E.coli BL21 cells...........42
2-6. Western blots showing the reactivity of the test bleeds............................................... 46
2-7. Expression and analysis of the CTV-HSP70h induction in BL21 cells......................46
2-8. Western blot showing the reaction of the bacterial-expressed CTV-HSP70h protein
with the HSV-Tag monoclonal or chicken polyclonal antibody............................... 47
3-1. Representation of the citrus tristeza closterovirus genome ............................................. 51
3-2. Tissue prints of infected and healthy citrus stems after incubation with HSP70h and
coat protein specific antibodies. ........................................................................ ....... 55
3-3. Interaction of CTV-HSP70h and CTV-CP in CTV infected tissue............................ 58
3-4. Serologically specific electron microscopy (SSEM) of trapped citrus tristeza virus
(C T V ) particles ...................................... ................................ ................ 59
3-5. Immunogold labeling of citrus tristeza virus trapped particles using the coat protein
specific antibody........................... ......... .. .. .......... ................................ 59
3-6. Immunogold labeling of citrus tristeza virus particles using the CTV-HSP70h-specific
antibody. ............. .......... ....... .. ........... ......... .......... ........... 60
4-1. Azure A staining and light microscopy of leaf petiole sections of healthy and CTV
infected tissue ....................................................................................................70
4-2. Inclusion body purification ......................................................................... 71
4-3. Immunofluorescence of proteins contained in the CTV inclusions using a TRITC-
labeled conjugate. ....................................................................................................72
4-4. Silver staining of SDS-PAGE gels containing proteins from partially purified CTV
inclusion bodies (IB) from citrus tristeza virus (CTV) infected and healthy bark
tissue e ...................... ....... ................................. ................................. 74
4-5. Western blot detection of the CTV-CP in the purified inclusion bodies.....................75
4-6. Western blot detection of the CTV-HSP70h in the purified inclusion bodies (IB)........75
5-1. Agrobacterium-mediated transformation of grapefruit epicotyl segments................... 92
5-2. Nucleotide and amino acid sequences of the CTV-HSP70h and the frameshift mutant
cloned in the binary vectors used for transformation ....... ... ....................................93
5-3. Analysis of the putative transgenic Duncan grapefruit plants by PCR.......................... 97
5-4. Histochemical GUS-staining of the regenenerated shoots .......................................... 99
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
CHARACTERIZATION OF THE HSP70 PROTEIN HOMOLOG (HSP70h) OF
CITRUS TRISTEZA CLOSTEROVIRUS
Ines-Marlene Rosales Villavicencio
Chairman: Dr. Richard Lee
Major Department: Plant Pathology
Citrus tristeza virus (CTV), a member of the family Closteroviridae, is the causal
agent of one of the most destructive diseases of citrus, causing a diversity of symptoms
on various scion and rootstock combinations. The virus is a monopartite, single-stranded,
positive-sense RNA virus, with a genome of about 20 kb encapsidated by two capsid
proteins. The Closteroviridae is the only viral family known to encode a homolog of the
HSP70 family of cellular chaperones. The HSP70 homolog (HSP70h) of CTV is a 65
kDa protein (p65) with high homology to cellular chaperones. The carboxyl-end of the
p65 protein (3' end of the p65 gene) was chosen for study because of its lesser homology
with cellular chaperones, to avoid cross-reactivity of the antibody with host proteins. The
3' end of the p65 gene of CTV was cloned with a histidine tag fusion and expressed in
Escherichia coli. The purified fusion protein was used to raise a polyclonal antibody in
chicken. Using this antibody, the CTV-p65 gene product was specifically detected in
CTV-infected but not in healthy citrus plants. The localization pattern of the p65 and the
viral coat protein were similar in direct tissue print studies. The same antibody used for
immunogold labeling studies revealed a close association of the HSP70h protein with the
virion. This association was later confirmed by co-immuno-precipitation of the virion and
the p65 protein. The occurrence of the p65 protein in the inclusion bodies present in CTV
infected tissue was studied. Additionally, two different constructs containing a full-
length and a frameshift mutant of the HSP70h gene from CTV were transformed into
Duncan grapefruit seedlings to test the possibility of inducing pathogen derived
resistance against CTV.
CITRUS TRISTEZA VIRUS: THE DISEASE, THE CAUSAL AGENT AND ITS
Closteroviruses represent a group of emerging and re-emerging economically
important plant pathogens. Members of this group affect several crops of major
economic importance, such as sugar beet, citrus, tomato, lettuce, potato, sweet potato,
grapevine, pineapple, cherry, and some ornamentals (Karasev, 2000).
The family Closteroviridae comprises more than 30 plant viruses with
filamentous, flexuous virions and includes representatives of either mono or bipartite
positive sense single-stranded RNA genomes (Karasev, 2000). Closteroviruses are
transmitted semipersistently by insects, i.e. aphids, whiteflies, or mealybugs (Brunt et al.,
1996). Based on the virus particle structure, vector transmission, and genome
organization, the Closteroviridae family has been classified into two genera:
Closterovirus, containing monopartite viruses transmitted by aphids, mealybugs and
possibly whiteflies (Brunt et al., 1996), and the genus Crinivirus, containing bipartite
whitefly-transmitted viruses (Wisler et al., 1998).
The Disease and Its Causal Agent
Citrus tristeza virus (CTV), a member of the genus Closterovirus, is the causal
agent of one of the most destructive viral diseases of citrus, and it occurs in most of the
citrus producing areas of the world (Bar-Joseph, 1989). CTV has a positive-sense single-
stranded RNA genome encapsidated in flexuous particles about 2000 nm in length (Bar-
Joseph and Lee, 1989). The virions contain two capsid proteins (CP) arranged in a
"rattlesnake" structure: a 25 kDa CP that encapsidates -95% of the particle and a 27 kDa
minor CP that encapsidates -5% at one terminus. This morphology is considered a
hallmark for the closterovirus group (Agranovsky et al., 1995; Febres et al., 1996). CTV
occurs as a diverse complex of strains that vary greatly in aphid transmissibility and
severity of symptoms in different citrus hosts. The virus is transmitted by grafting, but it
is not seed-borne (Bar-Joseph and Lee, 1989). It also has been mechanically transmitted
by a knife-cut and slash inoculation but with some difficulty (Garnsey et al., 1977;
Garnsey and Muller, 1988). CTV is vectored by several aphid species in a semipersistent
manner with the aphid retaining the ability to transmit the virus for up to 24-48 hrs after
acquisition (Bar-Joseph, 1989; Raccah et al., 1989). The most efficient vector for CTV is
Toxoptera citricida Kirkaldy, commonly called brown citrus aphid (Bar-Joseph, 1989;
Yokomi et al., 1994). This aphid probably originated in China and now is distributed
throughout many regions of the world (Rocha-Pefia et al., 1995). In November 1995, the
brown citrus aphid was discovered in Florida in the Ft. Lauderdale area (Halbert, 1997)
and by the summer of 1997 was widely distributed throughout the South Central and
Coastal regions of Florida (Michaud, 1998), threatening the citrus industry of the area.
Tristeza, which means "sadness" in Spanish and Portuguese, is best thought of as
a family of diseases caused by different strains of CTV. A common tristeza disease is the
decline of citrus scion varieties grafted onto sour orange (C.aurantium L.) rootstock (Bar-
Joseph, 1989). In the field, decline may be rapid or gradual. The most dramatic
symptoms are observed with the combination of sweet orange on sour orange rootstock
where sudden wilting and death can occur following CTV infection. In the decline on
sour orange symptom, phloem necrosis develops at the bud union causing root starvation.
When the starch reserves are finally exhausted, the tree rapidly dies, often leaving a dead
tree with fruit hanging but no leaves remaining (Figure 1-1, A) (Lee et al., 1994; Rocha-
Pefia et al., 1995). The other important disease caused by CTV is stem pitting of the
scion regardless of the rootstock, which reduces tree vigor, yield and fruit quality of the
tree (Figure 1-1: B, C, G). Stem pitting symptoms of CTV are considered as the most
serious disease caused by the virus, because citrus production cannot be continued by just
replacing trees on a CTV tolerant rootstock. Other symptoms often associated with stem
pitting disease are vein clearing, vein corking and leaf cupping (Figure 1-1, D, E and H)
(Lee et al., 1994; Rocha-Pefia et al., 1995). Some isolates of CTV induce very mild
symptoms or are symptomless, even in the most sensitive citrus species (Bar-Joseph,
1989). Seedling yellows (SY) strains cause chlorosis and stunting in sour orange, acid
lemon, and grapefruit indicator plants (Figure 1-1, F) (Rocha-Pefia et al., 1995). The
seedling yellows reaction is mostly a greenhouse or nursery disorder that is used to detect
the presence of the more serious decline inducing or stem pitting strains of CTV (Lee et
al., 1994; Rocha-Pefia et al., 1995); however seedling yellows can cause problems in the
field when infected trees are topworked with susceptible varieties.
Cytopathology of CTV-Infected Tissue
Citrus tristeza virus, as other closteroviruses, is characteristically associated with
the phloem which is found most consistently in the phloem companion and parenchyma
cells; hence it is called a phloem-limited virus (Karasev, 2000). Cells with active
Figure 1-1. Symptoms caused by citrus tristeza virus. (A) Sweet orange tree on sour
orange rootstock undergoing tristeza decline (Lee, R.F.); (B) Stem pitting on Pera
sweet orange, occurring in Brazil (Lee, R.F.); (C) Stem pitting on grapefruit in
Venezuela (Lee, R.F.); (D) Cupping of the leaf in Mexican (Roistacher, C.N.); (E)
Vein corking symptoms on leaves of a Mexican lime seedling inoculated with a very
severe seedling-yellow tristeza isolate (Roistacher, C.N.); (F) Seedling yellows
reaction on grapefruit (left) and sour orange (right) seedlings in the greenhouse
(Roistacher, C.N.); (G) Grapefruit collected from a Marsh grapefruit tree on rough
lemon rootstock in Colombia which was affected by stem pitting strains of tristeza
(Lee, R.F.); (H) Vein-clearing symptoms in the leaf of a Mexican lime seedling (Lee,
RF.); (I) Pinholes in the bark, caused by bristles in the wood, cause honeycombing on
the back side of the bark patch over the sour orange rootstock (Lee, R.F.).
Photographs presented in this figure were downloaded from www.ecoport.org. The
author of the photograph is given in the parenthesis.
closterovirus replication display clusters of vesicles with a diameter of 80-120 nm which
show different levels of tonicity and contain a network of fine fibrils (Bar-Joseph et al.,
1997). CTV produces inclusion bodies that are confined mostly to the phloem and
associated tissue. The inclusion bodies appear to be large aggregates of virus particles
mixed with structures of unknown composition which might contain modified cell
constituents. The inclusion bodies have been detected by using light microscopy
(Brlansky and Lee, 1990; Garnsey et al., 1980), in situ immuno-fluorescence (Brlansky et
al., 1988), and by transmission electron microscopy (Kitajima and Costa, 1968; Gowda et
al., 2000). The detection of CTV inclusions using light microscopy can provide a rapid
method for diagnosis of CTV infection (Brlansky, 1987). Studies have shown differences
in the number of inclusion bodies caused by mild and severe CTV isolates in the various
host species (Brlansky and Lee, 1990). The effect of virus strain or host on the
morphology of the various CTV inclusion bodies is not known.
Natural Resistance to CTV
CTV infects all citrus species and varieties, most hybrids and some citrus relatives
(Mestre et al., 1997c). Recently, some pummelo (C. grandis (L.) Osb.) accessions were
found to be resistant to certain CTV strains (Garnsey et al., 1997). There are only three
citrus relatives that have been reported to be resistant to CTV (Garnsey et al., 1987;
Yoshida et al., 1983): Severinia buxolia (Poir.) Tenore, Swinglea glutinosa (Blanco)
Merr, and Poncirus trifoliata (L.) Raf The resistance found in P. trifoliata is conferred
by a single dominant gene, designated Ctv (Fang et al., 1998; Gmitter et al., 1996), which
has been mapped by using molecular markers (Gmitter et al., 1996; Mestre et al., 1997b).
Introgression of this resistance into rootstock cultivars has been successful via sexual
hybridization, but the development of CTV-resistant scions has been more difficult due to
the introgression of undesirable fruit characteristics from Poncirus (Deng et al., 2001b).
Molecular cloning of the Ctv gene would provide a means to develop resistant scion
cultivars using genetic transformation. The region containing this gene has been mapped,
and markers flanking and cosegregating with Ctv have been developed (Fang et al.,
Recently, Deng etal. (2000) identified 22 sequences similar to the NBS-LRR
(nucleotide binding site-leucine rich repeat) class resistance gene in the citrus genomic
DNA by using PCR amplification with degenerate primers. One of the fragments was
closely linked and another seems to co-segregate with Ctv, which might facilitate direct
landing on the resistance gene (Deng et al., 2001b). Different BAC libraries have been
developed to pursue this objective, and some BAC clones and BAC contigs containing R-
gene candidates were characterized (Deng et al., 2001b; Yang et al., 2001). The Ctv
locus was localized within a genomic region of approximately 180 kb, and efforts are
being made to assign this locus to a smaller genomic fragment whose function can be
confirmed by genetic complementation (Deng et al., 200 la).
Two interesting observations, the fact that the hypersensitive reaction has not
been observed in Ctv-bearing plants (Mestre et al., 1997b), and that the virus can
replicate in protoplasts from CTV-resistant plants (Albiachi-Marti et al., 1999), have
raised questions as to whether Ctv confers resistance by blocking virus replication or by
interfering with virus loading or unloading from the phloem (Mestre et al., 1997b). The
same group has suggested that at least two genes are responsible for CTV-resistance in P.
trifoliata var "Flying Dragon," based on the short distance accumulation that they
observed in some Ctv-Rr progeny segregant plants derived by self-pollination of this
resistant genotype. Bulked segregant analysis of this population identified five RAPD-
markers linked to a locus (Ctm) that is located in a different linkage group from Ctv
(Mestre et al., 1997a).
Molecular Characteristics of CTV
Genomic Organization and Replication Strategies
CTV is a member of the monopartite genus Closterovirus in the family
Closteroviridae (Bar-Joseph, 1989). The genomic RNA (gRNA) of this virus contains
from 19,226 to 19,302 nt, depending on the isolate (Karasev et al., 1995; Mawassi et al.,
1996; Suastika et al., 2001; Vives et al., 1999; Yang et al., 1999b), which occur as 12
open reading frames (ORFs), potentially encoding 19 protein products (Karasev, 2000a;
Karasev, 2000b; Karasev et al., 1995) (Figure 1-2). These include replication-associated
proteins, the homolog of the heat shock proteins 70 (HSP70h), the two coat proteins, and
several others products with unknown functions (Bar-Joseph et al., 1997). Computer-
assisted sequence analysis has identified two conserved groups or blocks of genes in the
genome of CTV and other closteroviruses. The first group (ORF la and lb) includes
replication-associated proteins (RNA polymerase, putative helicase, putative
methyltransferase, and two accessory processing papain-like proteases (Dolja et al.,
1994)). The replication-associated proteins of CTV are translated directly from the
gRNA and expressed as a large 400 kDa polyprotein that is further processed by virus-
0 2 4 I 8 10 12 14 1i is 19 21 rnA
OIRFla th 2 3 4 5 8 7 110 $1
PRO PRO 2 p33 HSPT7 p27 p18 p20
r r I l I /11EL- r- r- r
,U M. U I-I.LI Lj""
PWRdf p4 pal pVA~ pft"V21
'i %, Sg RNA 2 -----" ---
\ s* RNA 3---
I [ I SgRNA7
Multiple ODNAs Sg RNA 8
Figure 1-2 Representation of the gene expression and genome organization of citrus
tristeza virus. Open reading frames are shown as boxes and the putative domains on ORF
la and lb are separated by lines. PRO: papain-like proteases 1 and 2 RdRp: RNA-
dependent polymerase. The genomic and subgenomic RNAs are shown by solid lines,
the size, in kilobases, is indicated by the scale at the top. Defective-RNA strategy is
shown by dashed lines (Figure reproduced from Ochoa-Corona, 2001).
encoded proteases (Karasev et al., 1995; Mawassi et al., 1995a). ORFIb, which encodes
the RNA polymerase, is proposed to be expressed by +1 ribosomal frameshift (Cevik et
al., 1999; Cevik, 2001). The second group of genes, ORF 3 to 7, includes a five-gene
block unique for closteroviruses. This block encodes a small 6 kDa hydrophobic protein,
a 65 kDa homolog of the cellular HSP70 proteins (HSP70h), a 61 kDa protein and a
tandem of two structural proteins, a 27 kDa capsid protein (CP) duplicate, and the 25 kDa
CP itself (Karasev et al., 1995; Karasev et al., 1994; Pappu et al., 1994; Sekiya et al.,
1991). It has been proposed that this unique protein quintet is required for cell-to-cell
movement of a closterovirus, based on the fact that beet yellow closterovims (BYV)
requires this set of proteins for intercellular translocation in leaf epidermis and leaf
mesophyll of the local lesion host, Claytoniaperfoliata (Alzhanova et al., 2000).
The ORFs located at the 3' end potentially encode proteins p18, p13, p20 and p23
the functions of these proteins are not clear yet. It has been reported that p20 gene
product accumulated in infected tissue, exhibited a high affinity to itself in a yeast two
hybrid system, and was localized mainly in the CTV-infected cells within the amorphous
inclusion bodies (Gowda et al., 2000). The p23-kDa protein, encoded by the 3'-terminal
gene of CTV, is an RNA-binding protein which contains several basic amino acids and a
putative zinc-finger domain between positions 50-86 of its amino acid
sequence (Lopez et al., 2000).
RNA Populations in Infected Tissue
CTV-infected plants contain the large double-stranded (ds) replicative form (RF)
RNA molecule and a nested set of at least nine smaller 3'-coterminal subgenomic RNAs
(sgRNAs) corresponding to the 3'-terminal ORFs 2 to 11. Each sgRNA is present as a
single-stranded RNA (ssRNA) molecule and as a corresponding dsRNA species (Hilf et
al., 1995; Mawassi et al., 1995a). In addition, the use of 5'-end specific probes
demonstrated the presence of a considerable amount of low molecular single-stranded
positive-sense RNA fragments, designated Low Molecular Weight Tristeza (LMT)
(Mawassi et al., 1995c). Multiple defective RNAs (D-RNA) that vary in size and
abundance are associated with a majority of the CTV isolates (Ayllon et al., 1999a;
Mawassi et al., 1995b).
Recently an even more complex scenario was described by Gowda et al. (2001).
They proposed that each controller element in the CTV genome produced three sgRNAs:
a 5'-terminal positive-strand and both positive and negative-stranded 3'-terminal RNAs.
This implies that theoretically CTV could produce 30-33 different species of RNA in
The LMT RNAs make up a major proportion of the total virus-associated RNAs
(Che et al., 2001; Mawassi et al., 1995c). They mainly consist of a population of RNAs
having two modal lengths of 744-746 and 842-854 nucleotides. It has been suggested
that these LMT RNAs are produced by termination during the production of genomic
RNAs. Additionally, a second class of heterogeneous 5'-coterminal sgRNAs of -10 Kb,
designated Large Molecular Weight Tristeza (LaMT) has been found in infected plants,
but in much smaller amounts. LaMT were found less consistently in tissue from
chronically infected plants than in RNAs obtained from recent infections (Che et al.,
2001). It seems that none of the ten 3' genes encoded by the CTV genome are involved
in the synthesis of these 5'-co-terminal sgRNAs, since they appear to be produced
normally by a deletion mutant from the virus, CTV-A Cla, which has all the 3'-genes
deleted (Che et al., 2001; Satyanarayana et al., 1999)
When the CTV-VT isolate was cloned and sequenced, the presence of several D-
RNAs of various sizes was revealed. The D-RNA were composed of the 5' and 3'
termini of the genomic RNA with extensive internal deletions. The size of the termini
varied among species, with minimal lengths of 442 nt and 858 nt from the 3' and the 5'
termini, respectively, resulting in different sizes of D-RNAs with different junction sites.
The D-RNAs were encapsidated as shorter virions showing the typical heterodimeric
encapsidation pattern of CTV (Bar-Joseph et al., 1997; Mawassi et al., 1995b; Mawassi
et al., 1995c).
Later, it was found that multiple D-RNAs that vary in size and abundance are
present in a majority of CTV isolates (Ayllon etal., 1999a; Mawassi etal., 1995b). The
size of the CTV D-RNA species ranges from small (- 1.6 kb) to nearly genomic full
length (> 10 kb) (Bar-Joseph et al., 1997; Mawassi et al., 2000b; Yang et al., 1997).
Some isolates have one or two D-RNAs in major abundance, along with multiple minor
D-RNAs. Most of the characterized D-RNAs consist of simple fusions of the 5' and 3'
genomic termini, but their lengths and junction sites vary among species (Ayllon et al.,
1999a; Mawassi et al., 1995c; Yang et al., 1997). Although D-RNAs are usually
homologous to the helper, abundant D-RNA with sequences different from that of the
major component of the gRNA have been found, suggesting that the CTV-replicase
complex might be able to recognize and replicate heterologous sequences in trans
(Mawassi et al., 2000b).
The variation in abundance of the different D-RNAs in a population suggests
selection for those of higher fitness. Using in vitro constructed D-RNA, Mawassi et al.
(2000a) reported that the minimal sequence required for accumulation of the D-RNA was
within the genomic 5' proximal -1 kb, plus the 3' 270 nts, although internal sequences
also affect the accumulation. A continuous ORF through most of the sequence derived
from the 5' end of the genome was also a requirement for D-RNA amplification,
although its protein product did not affect the amplification of this replicon (Mawassi et
Defective interfering RNAs are usually found to compete with the non-defective
virus genome for the components of the replicase. This interference with viral replication
results in a reduced level of virus accumulation and a marked attenuation of viral
symptoms in infected plants (Simon and Bujarski, 1994). In contrast, CTV D-RNAs do
not noticeably affect levels of accumulation of the genomic or the subgenomic RNAs of
the helper virus, suggesting that the D-RNA did not compete for the same pool of
replicase as the helper virus (Mawassi et al., 2000b). The only relationship between
CTV D-RNAs and biological activity has been reported by Yang et al., (1999) who
described an association between the SY reactions of sour orange seedlings and the
genomic composition of the D-RNA. They found that among sub-isolates of the VT
strain, which were selected from chronically infected Alemow plants, there was an
association between the presence of 2.4, 2.7 and 4.5 kb D-RNA, and SY and non-SY
reactions, respectively. A similar pattern was obtained with the isolate Mor-T from
Israel. This study suggested that the non-SY reaction results either from the absence of
SY gene(s) in the genomes of certain CTV strains or through the suppression of the effect
of SY gene(s) by D-RNAs with 5' parts larger than 4000 nucleotides (Yang et al., 1999).
The mechanisms suggested to be involved in the generation of the CTV D-RNA
are the minus-strand jumping model and the template switching mechanism. The minus-
strand jumping model is supported by the finding of an extra C at the junction of the
common end D-RNA derived from three different CTV-isolates, and an extra G at the 3'-
terminus of the minus strand of the sgRNA for the ORF 11. The second model, the
template switching mechanism, is supported by the presence of direct repeats from two
separate genome locations in the virus which have been found flanking or in the vicinity
of the junction sites of the CTV D-RNAs (Ayllon et al., 1999a; Bar-Joseph et al., 1997;
Yang et al., 1997).
RNA viruses employ diverse strategies to express their genomes in their hosts.
These include sgRNAs which serve as messenger RNAs for the expression of internal
genes. The sgRNAs of CTV, which have been found to be encapsidated in particles,
consist of substantial amounts of both negative and positive strands (Mawassi et al.,
1995a; Gowda et al., 2001). The sequences involved in their production are known as
"controller elements" instead of promoters and the mechanism of their production is not
clear yet (Gowda et al., 2001).
Nine sgRNAs were identified when the CTV-specific RNAs from the CTV isolate
T36 were extracted from infected tissue and analyzed by Northern blot hybridization
using specific probes for the different ORFs. A probe derived from the ORF11 (p23)
(Figure 1-2) hybridized with all the sgRNAs, indicating that these were co-terminal. The
most abundant species were those corresponding to ORFs 10 and 11 (p20 and p23,
respectively), and the number of sgRNAs isolated did not change when the isolate was
passed through different hosts (Hilf et al., 1995).
The production of the sgRNA of CTV is regulated both temporally and
quantitatively. In studies of the kinetics of accumulation of the CTV-RNAs, it was found
that most of the abundant sgRNAs accumulated in parallel to the gRNA, and that the
sgRNAs that allow the expression of the 3'genes accumulated to higher levels than those
from the 5' end. The relative order of accumulation of the sgRNAs extracted from CTV-
infected tissue was p20>p23>pl3>p25>p27>p33>p65>p61>pl8, and this pattern of
accumulation was maintained either if the RNAs were extracted from CTV-infected
citrus tissue or CTV-inoculated N. bethamiana protoplasts (Navas-Castillo et al., 1997).
Recently, a new set of sgRNAs which are 5'-co-terminal and positive-stranded
has been described to occur in CTV-infected tissue. CTV apparently produces ten or
eleven 5'-terminal sgRNAs, one for each sgRNA controller element plus the highly
abundant -800 nt 5'-terminal sgRNA. 5'-terminal sgRNA production was correlated
with the ability of the controller element to produce 3'-terminal sgRNAs. It seems that
each controller element terminates positive-stranded RNA synthesis from the 5' end as it
induces synthesis of the 3'-terminal sgRNA (Gowda et al., 2001). It is not clear
whether these 5'-terminal sgRNAs are functional or not.
Population Structure and Genetic Diversity of CTV
CTV field isolates may contain multiple genomic variants, some of which can be
separated upon aphid (Tsai et al., 2000) or graft transmission to different host species
(Moreno et al., 1993). Uneven distribution of the genomic RNA variants within the
infected plant and acquisition of only certain variants by individual aphids may contribute
to the population changes during the transmission process (d'Urso et al., 2000). Single-
strand conformation polymorphisms (SSCP) and cDNA hybridization analyses have been
used to compare genomic populations of CTV isolates (Albiach-Marti et al., 2000a;
Ayllon et al., 1999b; d'Urso et al., 2000; Kong et al., 2000). When the population
diversity of the California CTV isolates was studied by SSCP, it was found that most of
the isolates were composed of a population of genetically related variants (haplotypes),
with one being predominant (usually accounting for 80-90%), and a few haplotypes in
very low frequency. In this study, the diversity between different isolates was greater
than within isolates (Kong et al., 2000). Ayllon et al.(1999b) studied the haplotype
distribution of the p20 and p18 genes from CTV after host change and aphid transmission
using Spanish and Japanese isolates. They reported that changes in haplotype
populations were more drastic for p20 than for p18, and that the variation within the
population was more significant than the variation between populations. This suggests
that adaptation to a new host (or other environmental conditions) could be as important as
the geographical origin at the moment of analyzing population diversities. Albiach-Marti
et al. (2000a), using hybridization with a panel of cDNA probes to different genomic
sequences, also detected changes in the CTV genomic and D-RNA population after aphid
All these findings provide evidence that changes in the viral population occur
during the transmission process, but they do not explain the mechanisms responsible for
these changes. Variations observed in SSCP profiles after aphid transmission of CTV
isolates would indicate titer increase of certain sequence variants in the aphid transmitted
isolate and/or drastic reduction or disappearance of other variants present in the viral
population. At least two factors could contribute to the altering of the genomic RNA
population in the transmission process: (I) uneven distribution of the genomic RNA
variants in different plant parts may result in the aphids acquiring a different viral
population, depending on the probing site; and (II) individual aphids might sort some of
the variants, even if these are not predominant, and transmit a sub-population different
from that of the source isolate (d'Urso et al., 2000). In either case, a minor variant of a
population could become predominant and give rise to a new and different population.
Because symptoms caused by CTV are probably dependent on the composition of its
viral population, aphid transmission may act as a bottleneck, sorting some genomic RNA
variants and giving rise to a different population that may also alter symptom expression
(d'Urso et al., 2000). These types of variation have been already reported for CTV.
Moreno et al. (1993) showed that sub-isolates obtained from mild CTV isolates by
various host passages were more severe and express stem pitting. A similar result was
reported by Broadbent et al (1996), who showed that single aphid transmissions of
Australian CTV isolates with the vector Toxoptera citricida separated some of the
subisolates based on biological indexing on three citrus indicators and the numbers of
inclusion bodies produced. Tsai et al.(2000) demonstrated the recovery of orange stem
pitting strains of CTV after doing single aphid transmissions with Toxoptera citricida
from a source plant infected with the Florida T66 decline isolate. This sorting of RNA
variants of CTV by the aphid could explain the appearance of virulent CTV isolates in
areas where they had not been observed before.
Different approaches have been taken to develop molecular techniques for the
rapid differentiation of CTV isolates and identification of molecular markers related to
the different strains of CTV. Variation in serological reactivity, peptide maps of the CP,
dsRNA patterns, hybridizations with cDNA probes, restriction fragment length
polymorphism, and SSCP have been described utilized to differentiate CTV isolates and
strains (Lee et al., 1988; Moreno and Guerri, 1997; Niblett et al., 2000).
Nucleotide sequence analysis is an accurate procedure for CTV strain
differentiation and estimation of molecular genetic variation (Rubio etal., 2001). To
date, the complete sequences of six CTV isolates have been reported: T30 (19,259 nt) and
T36 (19,296 nt) from Florida (Albiach-Marti et al., 2000b; Karasev et al., 1995; Pappu et
al., 1994), VT (19,226 nt) from Israel (Mawassi et al., 1996), SY568 (19,249 nt) from
California (Yang et al., 1999), T385 (19,259 nt) from Spain (Vives et al., 1999) and
Nuaga (19,302 nt) from Japan (Suastika et al., 2001). Analysis of these sequences
reveals that the genomic organization is similar in all the CTV isolates sequenced so far,
although the genomic sequences differ markedly, with as little as 50 to 80% nucleotide
identity in much of the genome (Mawassi et al., 1996; Vives et al., 1999). The identity
between some sequences is nearly uniform throughout the genome for some isolates
(T385 and VT for example), but for other isolates, the sequences are asymmetrical and
progressively decrease toward the 5' end, with as little as 42% within the 5'NTR (Lopez
et al., 1998). The highest identity between all isolates occurs at the 3' untranslated region
(UTR), where the identity is higher than 97% for all isolates (Vives et al., 1999).
Analysis of the polymorphism of the 5'UTRs allowed the classification of CTV
sequences into three discrete groups, with intragroup sequence identity higher than 88%
and intergroup sequence identity as low as 44%. T36 was the type isolate for group I, VT
for group II and various Spanish isolates belong to group III (Lopez et al., 1998). SY568
and T385 belong to group II and III, respectively.
It is not known whether the symptoms induced by CTV isolates in citrus are
induced by a predominant genomic sequence, the viral population, a combination of
genomic RNA and D-RNA, or other factors (Albiach-Marti et al., 2000b). Two mild
isolate sequences, from isolates that do not produce noticeable symptoms in the field
(T30 from Florida and T385 from Spain), were compared to determine whether different
isolates inducing similar phenotypes might also have similar sequences (Albiach-Marti et
al., 2000b). The RNA genome of both isolates was the same size (19,259 nt), and the
nucleotide identity between different ORFs ranged from 98.7 to 100% (Albiach-Marti et
al., 2000b; Vives et al., 1999). Because these two isolates were separated in time and
geography, this sequence similarity was unexpected. Comparison of additional mild
CTV sequences with the T30 isolate showed remarkable sequence similarity, with
variability less than 1% (Albiach-Marti et al., 2000b).
Comparison of the sequences of the mild Spanish isolate, T385, and the stem
pitting isolate, SY568, from California showed nucleotide identities close to 90% in the
5' and 3' terminal regions of the genome, whereas the central region had over 99%
identity. This suggests that the central region of the SY568 genome resulted from RNA
recombination between two CTV genomes, one of which was almost identical to the
master sequence of a mild isolate (Vives et al., 1999).
Overall, CTV is among the most diverse and complex plant RNA virus. The
numerous species of RNA species present in infected tissue, the multiple genomic
variants found in field samples, and the unknown function of most of its genes leave open
many questions about the virus biology and the infection process.
The main objective of this research was to characterize the HSP70 protein
homolog (HSP70h) of CTV, and to study the localization and function of this protein
during the process of viral infection. The possibility of generating pathogen-derived
resistance, using as transgene the full length HSP70h from CTV or mutated forms of this
gene, also was explored. The specific objectives were the following:
1. To develop an antibody for the CTV-HSP70h protein specific enough in its
reactivity to enable studies on the function of this viral protein
2. To study the in vivo localization and expression of the HSP70h from CTV in
infected citrus tissue, as well as the association of this protein with the virion.
3. To study the possible presence of HSP70h in the characteristic inclusion bodies
present in CTV-infected tissue.
4. To produce transgenic Duncan grapefruit plants by using Agrobacterium
tumefaciens-mediated transformation and different constructs of the HSP70h
gene from CTV.
PRODUCTION OF A POLYCLONAL ANTISERUM AGAINST THE CARBOXY-
TERMINAL END OF THE CTV HEAT SHOCK PROTEIN HOMOLOG (HSP70h)
Citrus tristeza virus (CTV), a member of the family Closteroviridae, causes one
of the most economically important diseases of citrus. The CTV symptoms are diverse
depending on various scion and/or rootstock combinations (Bar-Joseph and Lee, 1989).
The virus has a monopartite, single-stranded, positive-sense RNA genome of about 20 kb
encapsidated by two proteins. The CTV genome is organized into 12 open reading frames
(ORFs), potentially encoding for at least 19 polypeptides that are expressed through at
least three mechanisms: proteolytic processing, translational frameshifting, and
production of subgenomic RNAs (Hilfet al., 1995; Karasev et al., 1995).
The Closteroviridae is the only viral family encoding for a homolog of the 70-
kDa heat shock protein (HSP70) family of cellular chaperones (Agranovsky et al., 1991).
The HSP70s are members of a set of proteins which undergo increased synthesis in
response to a variety of physical and chemical stresses; and they play diverse roles in
successful folding, assembly, intracellular localization, secretion, regulation, and
degradation of other proteins (Lindquist and Craig, 1988). These chaperones were
originally identified as inducible proteins, but some HSP70s are constitutively expressed
and appear to be essential for physiological cell growth (Hartl, 1996; Lindquist and
Craig, 1988). They are highly conserved in all domains of life: Archae, eubacteria and
eukaryotes. Eukaryotic genomes encode multiple HSP70 versions that are localized to the
various cell compartments (cytosol, endoplasmic reticulum, mitochondria, and
chloroplasts) (Karlin and Brocchieri, 1998).
It is thought that the viral HSP70 homolog (HSP70h) was probably acquired by a
common ancestor of the closteroviruses by recombination with a host mRNA coding for
HSP70 (Dolja et al., 1994). Computer-assisted sequence analysis revealed that the
structural elements identified in the N-terminal ATPase domain of cellular HSP70s are
conserved in closteroviral homologs, while the more variable C-terminal domain showed
limited homology between cellular and closteroviral HSP70s proteins (Agranovsky et al.,
1991). Recently, the HSP70h of beet yellows closterovirus (BYV) was shown to be
involved in intercellular translocation, representing a new type of plant viral-movement
protein (Peremyslov et al., 1999). Additionally, the HSP70h from CTV was shown to be
necessary for efficient virion assembly (Satyanarayana et al., 2000).
Information on the possible function of the viral genes from CTV has been
inferred by comparative computer assisted and genetic analysis. An alternative way for
the functional characterization of a virus protein is to study its intracellular localization in
infected tissue. A basic requirement for immunolocalization of a protein is to have an
antibody reacting specifically with the target gene product. The objective of this research
was to develop an antibody for the CTV-HSP70h protein specific enough in its reactivity
to enable further study of the functions of this viral protein in the process of CTV
Material and Methods
CTV isolate T3800. The Florida CTV isolate T3800 was used as a virus source
for cloning and expression of the recombinant HSP70h protein. The T3800 source was
grapefruit (Citrus paradise; Macf.) plants in greenhouses of the Department of Plant
Industry (DPI) and the Plant Pathology Department at University of Florida, both located
in Gainesville, FL. CTV isolate T3800 was originated from a lemon tree in a home
dooryard in Delray Beach, FL. It causes severe stem pitting in grapefruit, no stem pitting
in sweet orange, and causes seedling yellows in sour orange and grapefruit (Manjunath et
Reverse transcription and PCR. The p65 gene from the Florida stem pitting
CTV isolate T3800 was reverse transcribed from double-stranded RNA (dsRNA). The
dsRNA was purified from infected bark tissue by using non-ionic cellulose (CF-11,
Whatman) column chromatography in the presence of 16% ethanol, according to the
procedure described by Moreno et al.(1990). For the annealing of the primer to the
template, 10 il of dsRNA (representing approximately 0.2 g of fresh tissue) and 1 il of
primer CN302 (0.1 ig/il) (Table 2-1) were incubated at 700 C for 10 min, then
transferred to an ice bucket for a minimum of 5 min. Then, a mixture containing 4 il of
5X first-strand buffer (250 mM Tris-HCl,pH 8.3 at 250C, 375 mM KC1, 15 mM MgCl2),
0.5 mm each dATP, dGTP, dCTP, and dTTP, 10 mM DTT, 1 il RNAsin (20-40 U/il)
(Promega Corp.), and 1 il Superscript II RT (Gibco-Life Technologies) was added to
each reaction to give a final volume of 20 il. After one hour incubation at 420C, the
reaction was held at 700C for 15 minutes, then transferred to ice or stored at -200C for
later use. This cDNA was used as template for the amplification of the full length
HSP70h or its carboxy-terminal end.
The PCR reaction was performed in a final volume of 50 il. The mixture
contained 5 il of 10X PCR buffer (500mM KC1, 100mM Tris-HCl(pH 9.0 at 250C) and
1.0% Triton X-100), 2.5 mM MgCbl, 0.4 mM each dNTP (dATP, dGTP, dCTP, and
dTTP), 0.1 ig each primer (CN 200 and CN 201) (Table 2-1), 2.5 U of Taq DNA
polymerase (Promega, Corp), and 2 il of the cDNA template. Thermocycling conditions
were 2 min at 940C, 40 cycles of 45 sec at 940C, 60 sec at 500C and 90 sec at 720C,
followed by a final extension of 5 min at 720C. RT-PCR products (size of approximately
1785 bp) were separated by electrophoresis in agarose gels and photographed using a
Fluor-S MAX Multilmager System (Bio-Rad). The RT-PCR amplified fragment was
cloned into a pGEM-T vector (Promega) (pGEM-T/HSP70h) and subsequently
sequenced at the DNA Sequencing Core Lab, at University of Florida using universal
(forward and reverse) M13 primers.
For the amplification of the carboxy-terminal end of the p65 gene from the CTV
isolate T3800, either the cDNA template previously described, or a clone of the pGEM-
/HSP70h was used as template (0.1 ig/ il). The PCR reaction was performed in a final
volume of 50 il. The mixture contained 5 il of 10X PCR buffer (500mM KC1, 100mM
Tris-HCl(pH 9.0 at 250C) and 1.0% Triton X-100), 2.5 mM MgC2, 0.4 mM each
dNTP (dATP, dGTP, dCTP, and dTTP), 0.1 ig each primer (CN 351 and CN 352) (Table
2-1), 2.5 U of Taq DNA polymerase (Promega, Corp), and 2 il of the cDNA or 1 il of
the HSP70h clone as template. Thermocycling conditions were 2 min at 940C, 40 cycles
of 30 sec at 940C, 30 sec at 500C and 45 sec at 720C, followed by a final extension of 5
min at 720C each. RT-PCR products (442 bp, nucleotides 13382 to 13824 in T36 isolate)
were separated by electrophoresis in agarose gels and photographed using a Fluor-S
MAX Multilmager System (Bio-Rad).
Table 2-1: Sequence of the primers used for the RT-PCR and cloning of the HSP70h
Primer Sequence Characteristics
CN302 5'AGNCGTCANTTCATGGGACGTCA-3' Sense alignment near nucleotide
CN 201 5'-AGATCTTCAGAGAGGTATTCTTTC C-3'
CN 468 5'-CATGCCATGGTGCTTTTGGGTTTAGAC -3'
14984 in CTV-T36 isolate. N for
Sense, EcoR I and Nde I sites.
Alignment at nucleotides 13380 to
13400 in the p65 gene of the CTV-
Antisense, Xho I site, non stop
codon. Alignment at nucleotides
13800 to 13824 in p65 gene of the
T36-HSP70h forward, sense primer
with NdeI site
T36-HSP70h reverse, anti sense
with Bgl II site.
p65 sense, from startcodon, Ncol
site at 5'end..
Cloning in pET-22b(+) and protein expression. The RT-PCR amplified
carboxy-terminal end fragment was cloned into the pGEM-T vector (Promega) and then,
subcloned into pET-22b(+) expression vector (Novagen) using the EcoRI and Xhol sites
that were incorporated into the primers CN351 and CN352. This cloning produced a
fusion of the fragment with a C-terminal His-tag sequence. This new construct was used
to transfom Escherichia coli strain BL21. Induction of the recombinant protein was
performed following the pET System Manual instructions (Novagen). Briefly, 500 ml of
Luria broth media (LB) (10g Bacto-tryptone, 5g yeast extract, 10g NaC1, adjust pH to 7.5
with NaOH, adjust volume to 1L) were inoculated with an overnight culture of the pET
recombinant in BL21. The media was incubated at 370C with shaking at 220 rpm to an
optical density at 600 nm (OD600) of approximately 0.5-1.0. After reaching this OD, the
protein was induced by adding isopropyl-beta-D-thiogalactopyranoside (IPTG) to a final
concentration of ImM. The culture was then allowed to grow for three hours under the
same conditions. The expression of the target protein was assessed by analysis of total
cell protein on a SDS-polyacrylamide gel followed by a Coomassie blue or silver
staining, according to standard procedures (Sambrook, 1989). A total cell protein sample
was also analyzed as described in the pET system Manual (Novagen), to study the
localization of the induced protein either in the media, periplasm, soluble cytoplasm or
insoluble cytoplasm fraction. The localization of the induced protein defines the
purification procedure to follow after the induction.
Protein purification under denaturing conditions. The induced culture was
harvested by centrifugation at 6,500 x g for 15 min at 40C. The pellet was then
resuspended in 0.1 culture volume of 1X Inclusion Bodies Wash Buffer (IBWB) (20 mM
Tris-HClpH 7.5, 10 mM EDTA, 1% Triton X-100), and lysozyme was added to a final
concentration of 100 ig/ml from a freshly prepared stock (10 mg/ml). Additionally, for
each gram bacteria harvested, 8 il of 50 mM phenylmethylsulfonylfluoride (PMSF) was
added to the cell suspension. After incubation at 300C for 15 min, the cells were
sonicated on ice (Misonix Inc., Model XL-2000 Microson Ultrasonic) with the power
level set between 4-5, at 40%-50% duty, until the cell solution was no longer viscous (15-
20 burst). The solution was always kept at 40C. The pellet was collected by
centrifugation at 10,000 x g for 10 min, and washed two times with 0.1 culture volume of
lX IBWB. The insoluble proteins present in the bacterial inclusion bodies were then
collected by centrifugation at 10,000 x g for 10 min, and stored overnight at -20 OC.
After thawing the pellet, the insoluble proteins were resuspended in 5 -10 ml of
Buffer A (6M guanidium chloride, 0.1 M NaH2PO4, 0.01 M Tris-HCl pH 8.0). The
lysate was stirred for approximately 1 hour at room temperature, until the solution
became translucent. The lysate was then centrifugated at 10,000 x g for 30 min at room
temperature to pellet cellular debris. The supernatant was mixed with 50% Ni-NTA resin
(Qiagen) at a ratio of 1 ml of resin for each 4 ml of lysate, and shaken for 30 min at room
temperature. The lysate-resin mixture was loaded into an empty column with the bottom
cap still attached. After removing the bottom cap, the flow-through was collected and
saved for later protein analysis. The column was then washed two times with 4 ml of
buffer C (8M urea, 0.1 M NaH2PO4, 0.01 M Tris-HC1, pH 6.3). The recombinant protein
was eluted with 2 ml of buffer D (8M urea, 0.1 M NaH2PO4, 0.01 M Tris-HC1, pH 5.9),
followed by 2 ml of buffer E (8M urea, 0.1 M NaH2PO4, 0.01 M Tris-HC1, pH 4.5). An
aliquot of each fraction was analyzed by standard SDS-polyacrylamide gel electrophore-
sis, followed by a Coomassie blue or silver staining.
Fractions containing the target protein were dialyzed in a multi-step way, against
a buffer (0.1 M NaH2PO4, 0.01 M Tris-HC1, pH 6.3) that contained decreasing concentra-
tions of urea (6M, 3M, and no urea) in a 24 hour period. Before storage of the induced
protein, 10 il of a broad-spectrum proteinase inhibitor (Sigma) was added to the
preparation to inhibit further degradation of the protein.
Production of the polyclonal antiserum. The services of Cocalico Biologicals,
Inc. (Reamstown, PA) were used for the production of the polyclonal antiserum. Chicken
was the host species selected for the project, and two young laying hens were injected
with the same immunogen at the same time. The immunogen was supplied as a purified
protein stored at -20 C, whose concentration was 100 ig/ml. The standard Cocalico
protocol for immunization was used for the company to raise the polyclonal antiserum in
chicken. Briefly it is:
Day 0: Prebleed and collection of pre-egg / initial inoculation.
Day 21: Boost / egg collection.
Day 31: Test bleed.
Day 42: Boost (same as day 21).
Day 52: Test bleed.
Day 63: Boost (same as day 21).
Day 73: Test bleed.
After day 73, a monthly boost followed by a test bleed 10 days later was
performed for the following six months. The eggs were collected daily, and evaluated as
pools from around the same time of the test bleeds. 100 ig of antigen were injected into
the breast muscle at each time. The first injection was in complete Freunds' adjuvant and
the remaining in a 1:1 dilution of complete and incomplete Freund's Adjuvants.
Sera collected during test bleeds were evaluated by Western blot analysis
according to standard procedures (Sambrook, 1989) using as controls the total cell extract
from induced and non-induced E.coli strain BL21, as well as bacteria BL21 that did not
carry the expression plasmid pET22b(+).
Extraction and purification of IgY from egg yolks. Isolation of polyclonal
chicken immunoglobulins from pooled eggs was performed using a simple two step
procedure (Camenisch et al., 1999). The egg yolks were separated from the egg white,
washed with deionized water, and placed in a funnel. The skins of the egg yolk were
removed with a forceps, and the yolks collected in a 50 ml screw cap tube. The yolk
volume was brought to 25 ml with sodium phosphate buffer (100 mM, pH 7.6), and
mixed vigorously. Subsequently, 20 ml of chloroform was added, and the mixture was
shaken until a semisolid phase was obtained. After centrifugation at 1,200 x g for 30 min,
the supernatant was decanted into a centrifugation tube, and solid polyethylene glycol
6000 was added to a final concentration of 12% (w/v). After centrifugation at 15,700 x g
for 10 min, the pellet was suspended and stored at -800C.
Cloning and expression of the full length CTV-HSP70h: The HSP70h gene
from CTV was amplified by PCR using as the template the clone pGEM- T/HSP70h from
the CTV isolate T3800 previously described, and the primers CN352 and CN468 (Table
2-1). A proofreading polymerase (Vent polymerase, New England Biolabs) was used
for the reaction to generate a blunt-end product following the manufacturer's
recommendations. The thermocycling conditions were similar to those previously
described for the amplification of the full p65 gene.
The blunt-end PCR product was then cloned in the pCR-BluntII TOPO vector
(Invitrogen), and subsequently subcloned in the expression vector pET27b(+) (Novagen),
using the Ncol and Xhol restriction sites that were incorporated in the primers. This
cloning produced a CTV-HSP70h with a C-terminal HSV-Tag and His-Tag sequences.
The HSV-Tag fusion allows the use of a HSV-Tag monoclonal antibody to follow the
expression of the target protein. The induction of the CTV-HSP70h was performed as
The sequence of the HSP70h from the Florida stem pitting CTV isolate T3800
was obtained by sequencing multiple clones of the RT-PCR fragment cloned in the
pGEMT vector. The gene was 1785 nt in length (Figure 2-1), and its deduced amino acid
sequence contains 594 residues (Figure 2-2), which agreed with the HSP70h previously
characterized from CTV (Karasev et al., 1995; Mawassi etal., 1996; Suastika etal.,
2001; Vives et al., 1999; Yang et al., 1999b). Analysis of the HSP70h sequences
contained in the Genebank indicated some variation occurs in the nucleotide sequences
for this gene among the various CTV isolates. The nucleotide sequence identity for the
CTV HSP70h ranged from 87 to 99% (Table 2-2), with a conservation of sequence
uniformly all along the gene, with randomly scattered variability in the sequence.
At the amino acid level, the identities among the HSP70h from CTV ranged from
90 to 99% (Table 2-3). For the isolate T-3800, the percentage of amino acid identity was
92% with the VT-CTV isolate, and 95% with all the others sequences analyzed. The
seedling yellows isolate from Israel (VT) showed the lowest percentage of sequence
similarity when compared to the other CTV-HSP70h sequences available in the
Genebank database (Table 2-4). The amino acid sequences of the translated p65 gene
from CTV (Figure 2-2) were used to construct a cluster dendogram. The clustering of the
isolates in this dendogram does not necessarily reflect their biological activities (Figure
2-3, Table 2-4) For example, the two mild isolates included in the analysis (T30 and
T385) do not cluster together and neither do the stem pitting isolates (T3800 and SY385).
Only the seedling yellows isolates (VT and Nuaga) shared the same cluster, showing
some association between amino acid sequence of the CTV-HSP70h and their biological
Computer assisted analyses have shown that the motifs identified in the ATPase
domain of the cellular HSP70s are conserved in the closteroviral HSP70h proteins
(Agranovsky et al., 1991) (Figure 2-4 A). If a CD-Search (Conserved Domain Database
and Search Service, v1.53) is conducted with the protein sequence of the HSP70h from
CTV, two conserved domains are retrieved by the system. These domains are both
located at the amino-terminal end of the CTV- HSP70h. One of them is an ATPase
domain contained at the N-terminal end of the cellular chaperones (HSP70) that expands
377 aa from the viral chaperone homolog and the other is a segment of about 167 aa
from the FtsA family of prokaryotic cell cycle proteins, which are predicted to contain an
ATPase activity (Figure 2-4 B). The C-terminal domain from the CTV-encoded homolog
shows limited homology, not only between the closteroviral HSP70h proteins, but also
between the closteroviral HSP70h and cellular chaperones. When the PROSITE database
of protein families and domains was searched using a chaperone homolog from CTV, two
signature sequences of the HSP70h protein family were retrieved. One corresponded to
the sequence expanded between amino acid 201 and 214 from the viral homolog
(VYDFGGGTFDVSIV), and the other to the sequence included between amino acids 325
and 339 (LVVVGGSSYLPGLLD) from the viral protein. In both cases there is 100%
identity between the signature sequence and the amino acid sequence of the CTV-
HSP70h (Figure 2-2). Considering these computer assisted analyses, the C-terminal end
of the HSP70h of CTV was chosen to raise a polyclonal antibody because of its lesser
homology with cellular chaperones. In this way, it was hoped to avoid cross reactivity
with host proteins. The sequence corresponding to the carboxy-terminal end of the
HSP70h gene from the Florida stem pitting isolate T3800 (nucleotides 13382 to 13824 in
the T36 isolate) was cloned in the pET22b(+) expression vector, producing a fusion
protein with a C-terminal histidine-tag sequence
Expression and Purification of the Carboxy-Terminal End of the CTV-HSP70h
The DNA fragment amplified using the primers CN351 and CN352 (Table 2-1)
was cloned into the pET22b(+) vector. This fragment of 444 bp corresponds to the
sequences spanned by the nucleotides 1333 and 1776 in Figure 2-1. The expression of
the cloned fragment produced a protein of 149 amino acids, which corresponds to the
amino acids located between positions 445 and 592 at the carboxy-terminal end of the
HSP70h protein of the CTV isolate T3800 (Figure 2-2), with a histidine-tag fusion
The recombinant protein was expressed in BL21 cells at high levels after
induction with ImM IPTG, and showed the expected size on SDS-polyacrylamide gels
(Figure 2-5 a). Using the program "Compute pi / Mw" at the ExPASy server, a
theoretical molecular weight of 16,681 Daltons was computed for the target protein. This
molecular weight does not consider the extra histidine tag from the carboxy-terminal
fusion or the effects of any possible post-translational modifications of the target protein.
The high levels of expression resulted in the formation of insoluble aggregates of
proteins (inclusion bodies) where the target protein was localized (Figure 2-5A). Under
Table 2-2. Nucleotide identity' among the CTV-HSP70h genes of different isolates
of citrus tristeza virus.
T-3800 T-36 SY-568 T-30 T-385 VT Nuaga
T-3800 100% 90%0 90% 90% 90% 89% 90%
T-36 100% 93% 93% 93% 88% 88%
Sy-568 100% 990/ 990 870 88%
T-30 100% 99% 87% 88%
T-385 100% 87% 88%
VT 100% 96%
1 Identity is defined as the extent to which two nucleotide sequences are invariant. This
table shows the percentage of identical nucleotides from the total nucleotide sequence
Table 2-3. Amino acid identity'
isolates of citrus tristeza virus
among the CTV-HSP70h proteins expressed by different
T-3800 T-36 SY-568 T-30 T-385 VT Nuaga
T-3800 100% 95% 95% 95% 95% 92% 95%
T-36 100% 95% 95% 95% 90% 93%
Sy-568 100% 99% 99% 91% 94%
T-30 100% 98% 90% 93%
T-385 100% 90% 93%
VT 100% 94%
1Identity is defined as the extent to which two amino acid sequences are invariant. The
percentage of identity from the total amino acid sequence (594 aa) are shown.
CTV-T3800 Length: 1785 nucleotides
tgggtttaga cttcggtacc acgttttcaa cagtggctat
Figure 2-1. Nucleotide sequence for the HSP70h gene from the grapefruit stem
pitting CTV isolate T3800.
Figure 2-1 continued.
Table 2-4. Biological properties and origin of the citrus tristeza virus (CTV)
isolates included in the alignment of the amino acid sequences of their HSP70h
CTV isolate Origin Symptoms induced
SY568 California Severe, stem pitting on grapefruit
and sweet orange
T385 Spain Mild
T30 Florida Mild
T36 Florida Severe, decline on sour orange
T3800 Florida Severe, stem pitting on grapefruit
Nuaga Japan Seedling yellows
VT Israel Seeding Yellows
I I DFTTSTAMATSEL 0KQ
I I DFTTSTAMATSEL 0KQ
GSFYKDLKRWVUU'i' KIN QTYLHKLSPSYKV
m m 3PNNLG
m m 3PNNLG
Figure 2-2. Alignment of the amino acid sequences of HSP70h protein from several
CTV isolates. The alignment was generated by the CLUSTAL X (1.8) program. The
sequences included were retrieved from the Genebank database, except for that of the
CTV T3800 isolate, which was obtained by direct sequencing.
* 140 *
C *A P
NTLQAFTQ*@ SGSCY INEPAAAY
NTLQAFTQ*@ SGTCY INEPAAAY
i NLQAFTQSSSGSCY INEPAAAY
i SLSG CVIIP F
S oT PK oSSA D .Y-AVY D w ,GT,,,VS.VVR,
S oT PK oSSA D .Y-AVY D w ,GT,,,VS.VVR,
S oT PK oSSA D .Y-AVY D w ,GT,,,VS.VVR,
S oT PKoLM .D.KYLA VY ,FGGG ,F o-.IVSV
S oTLPK oSSA.DKYLAVYDFGGG.,FD-.IVSV
SToPKLSAoKYLV"Y"F-G. T V SI .m-wVSV
Figure 2-2 continued.
* 260 28
Llllk* [.]C1 I~~E ,,""II4 'AIJ0]ID' I E
ADFIllI [.1C IPQENVSSLKE"ALS4LQT IIJ0,IIDPV-KYTVTH
ADnoFIPQoK.,LnVSSw ,r0 LKEALSoLQTD PnVKYp VTH
30I T A
Figure 2-2 continued
* 260 28
Llllk* [.]C1 I~~E ,,""II4 'AIJ0]ID' I E
ADFIllI [.1C IPQENVSSLKE"ALS4LQT IIJ0,IIDPV-KYTVTH
ADnoFIPQoK.,LnVSSw ,r0 LKEALSoLQTD PnVKYp VTH
30I T A
Figure 2-2 continued
* SALLDC3 LVTFA3 V AAAG
* SALLDC3 LVTFA3 V AAAG
* SALLDC3 LVTFA3 V AAAG
* SALLDC3 LVTFA3 V AAAG
* SALLDC3 LVTFA3 V AAAG
* RTALLDC3 LVTFA3 V AAAG
* gRCVTSYAP 3GD
* gRCVTSYAP 3GD
* gRCVTSYAP 3GD
* gICVTSYAP 3GD
* gRCVTSYAP 3GD
yeI 'v]B- g 'v m g ,I -[ li IVo
T m *Lyre L 0, VD -GV m, LVNo
I m DMKVFLKPGEVVgT
Figure 2-2 continued
L- YnIR IltI 3DLB' ![ 'I lI
LTRKYI LRTFR :
Figure 2-2 continued
Figure 2-3. Cluster dendrogram based on the amino acid sequences of the translated p65
gene for the various CTV isolates. The biological characteristics of the isolates are
summarized in Table 2-4.
I I I I I I I
Figure 2-4. Domain conservation between HSP70s and HSP70h proteins.
(A) Representation of the domain structure of members of the 70kDa stress proteins.
(B). Graphical overview of the putative conserved domains detected in the HSP70h
of CTV, using CD-Search. The numbers 1 to 594 represent the amino acid sequence of
this protein. HSP70 represents the conservation with the ATPase domain contained in the
family of cellular chaperones, and FtsA represents the conservation with the family of
prokaryotic cell cycle proteins
the same expression conditions, bacteria carrying only the vector pET22b(+) did not
express any exogenous protein with a size similar to the target protein (not shown).
After removing soluble proteins from the inclusion bodies by using sonication and
washes with Triton X-100 as described in the Materials and Methods section, the
inclusion bodies were solubilized in a buffer containing 6M guanidium chloride and 8M
urea. This solution was applied to a Ni-NTA resin (Qiagen) and purified by elution using
St 1 2 3 4 5 St 1 2 3 4 5 6 7
4 5 .5 3
19.7 I 20 -
Figure 2-5. Silver stained SDS-polyacrylamide gel electrophoresis (PAGE) showing the
over-expression of the 149 amino acid fragment fusion protein in E.coli BL21 cells.
(A) Protein expression at 370C: Lanes 1 and 2, non-induced bacteria; lanes 3 and 4,
IPTG-induced bacteria; lane 5, insoluble fraction containing the induced protein
(inclusion bodies). The arrow indicates the position of the induced protein in the gel.
(B) Purification of the fusion protein: Lanes 1 to 4 show different fractions eluted from
the Ni2+-NTI column after solubilization of the inclusion bodies. Lanes 5 to 7 show
proteins obtained after dialysis of fractions shown in lanes 1, 3, and 4, respectively.
a pH gradient (Figure 2-5 B). The purified protein was then dialyzed in a multi-step way,
against a buffer (0.1 M NaH2PO4, 0.01 M Tris-HC1, pH 6.3) that contained decreasing
concentrations of urea (6M, 3M, and no urea) over a 24 hour period. This procedure
allows the refolding of the protein by removing the denaturing agent from the medium.
The target protein obtained after dialysis (Figure 2-5 B) was used as antigen to raise
polyclonal antibody in chicken.
Production of the Polyclonal Antiserum
The purified His-tag/HSP70h fusion protein was injected into two young laying
hens for six months, according to the protocol described in Material and Methods section.
Sera collected during test bleeds were evaluated by Western blot analysis for their
reactivity against the recombinant protein used to raise the antiserum, against non-
induced bacteria, and against bacteria without the expression vector. Each of the sera
tested was diluted 1:1000.
A reaction of the antibody with the induced fusion protein was observed after day
73 (according to the scheme presented in the Material and Methods). The reactivity
increased with the new booster immunizations until it reached a plateau after the fourth
test bleed. The serum did not react with non-induced bacteria, or bacteria without the
plasmid pET22b(+) (Figure 2-6). The reactivity pattern of the antibody was similar for
the test bleeds obtained from the two hens immunized, although there was a slightly
stronger and more specific reaction with one of the hens (UF-C17).
After six months of immunizations, there was no further improvement in the
reactivity of the antibody. Therefore, it was decided to finish the booster immunizations
and proceed with the IgY purification from the eggs that were collected and pooled from
around the same time as the test bleeds from months three to six.
To further characterize this antibody, the full length CTV-HSP70h was over-
expressed as an HSV-Tag- HIS -Tag fusion protein in BL21 cells (Figure 2-7). The
induced protein was detected by a HSV-Tag monoclonal antibody, which has high
affinity and specificity for the 11 amino acid peptide derived from Herpes Simplex Virus
glycoprotein D. This antibody recognized a specific band that corresponds with the
expected size of the fusion protein (65kDa plus fusion peptides) after induction of the
bacterial cells with IPTG (Figure 2-7).
When similar western blots containing the lysate from induced and non-induced
bacteria were probed, either with the HSV-Tag monoclonal antibody, or the chicken
antibody raised for the carboxyl-terminal end of the viral chaperone (UF-C17), a similar
pattern of the reactivity of the proteins bands was observed (not shown). There also was
some non-specific reaction of the chicken antibody UF-C17 with other expressed
bacterial proteins. These were probably proteins that were co-purified with the viral
chaperone, or cellular chaperones from the bacteria that share homology with the viral
protein. In order to reduce non-specific reactions with plant tissue, the chicken antibody
was pre-adsorbed with healthy citrus tissue extract. This pre-adsortion reduced the
number of bands visualized after performing a Western blot with induced and non-
induced bacterial lysate (Figure 2-8 A, B).
When this pre-adsorbed antibody was used to probe the bacterial-expressed CTV-
HSP70h, a single prominent protein was detected from the induced culture. By Western
blot analysis, this protein corresponded in size with the protein detected by the HSV-Tag
monoclonal antibody (Figure 2-8 B). This confirms that the chicken antibody was able to
recognize the full-length protein of the viral chaperone analog expressed in BL21 cells.
The sequences of the HSP70s are highly conserved throughout a wide range of
organisms, from bacteria to mammals (Agranovsky et al., 1991). The cellular HSP70
sequences are more highly conserved in their N-terminal region, with the C-terminal end
being more variable (Craig et al., 1993). Computer-assisted analysis shows that the N-
terminal domain of the 65kDa protein from CTV has high levels of similarity to the
HSP70 ATPases, while the C-terminal portion shows no homology with the equivalent
domain in the cellular HSP70s. The C-terminal region of the HSP70h from CTV also
shows moderate homology among the closteroviral chaperones (Agranovsky et al., 1991;
Pappu et al., 1994).
The sequences available for the CTV-HSP70h protein have an amino acid identity
that ranges from 90 to 99%. When these sequences were arranged into a dendrogram, the
clustering obtained did not result in the clustering of isolates having similar biological
properties. This suggests that p65 probably is not involved in determining the
symptomatology of a particular CTV isolate. Relationships between sequence data and
biological properties have been found in the case of CP (Pappu et al., 1993a; Pappu et al.,
1993b), p27 (Febres et al., 1994) and p23 (Pappu et al., 1997) genes from this virus.
Previously there were attempts to produce an antibody to the p65 protein from
CTV by expressing the full sized recombinant protein in bacterial cells (C.L.Niblett,
1~~~ ~ ~ ~ 2 1 2 31 2
bleed day 73
Figure 2-6. Western blots showing the reactivity of the test bleeds. The serum
used in each blot is indicated at the bottom of the figure for all tests. Lane 1: non-
induced BL21 cells; Lane 2: IPTG-induced BL21 cells carrying the plasmid with
the carboxyl-terminal end of the CTV-HSP70h; lane 3: induced BL21 cells not
carrying the plasmid.
1 2 3 4 5 6 7 8 9
1 2 3 4 5 6 7 8 9
Figure 2-7. Expression and analysis of the CTV-HSP70h induction in BL21 cells.
(A) Coomassie blue-staining of a 10% SDS PAGE- gel. The total bacterial lysate was
analyzed before (lanes 2 and 4) and after (lanes 3 and 5) induction of protein
expression. The induced lysate from lanes 3 and 5 was fractionated in soluble (lanes 6
and 8) and insoluble fractions (lanes 7 and 9) to localize the target protein. The
positions of the molecular weight markers (10kDa protein ladder standard, Gibco-
BRL) are indicated at the left.
(B) Western blot analysis of the induced CTV-HSP70h protein. A gel identical to the
one shown in (A) was transferred to a nitrocellulose membrane and probed with a
HSV-Tag monoclonal antibody conjugated with alkaline phosphatase. The blot was
developed using the NBT-BCIP substrate.
1 2 3
1 2 3
1 2 3
1 2 3
1 2 3
A St 1 2 3 1 2 3 1 2 3 B 12 3 12 3
Left Center Right Left Right
Figure 2-8: Western blot showing the reaction of the bacterial-expressed CTV-HSP70h
protein with the HSV-Tag monoclonal or chicken polyclonal antibody.
(A) Left: Coomassie blue staining of a 10% PAGE-SDS gel. Center: Western blot using
the polyclonal antibody raised against HSP70h non-adsorbed and pre-adsorbed (right)
with healthy citrus extracts.
(B) Western blot showing the reaction of the bacterial expressed HSP70h (ull length
protein), using monoclonal HSV-Tag (left) or the pre-adsorbed chicken antibody (right).
For figures in A and B, lane 1: non-induced bacterial lysates; lane 2: IPTG-induced
bacterial lysate; and lane 3: healthy citrus tissue. The positions of the molecular weight
(10kDa protein ladder standard, Gibco- BRL) are indicated at the left.
personal communication). The antisera raised using this protein gave non-specific
reactions with healthy citrus tissue, as well as some other bacterial-expressed proteins.
This suggests that the p65 homolog from CTV has sequence similarity with the cellular
chaperones, and also structural similarities with them. These preliminary results
suggested that the C-terminal end of the chaperone homolog from CTV may be a better
candidate to raise an antibody for specific detection of this protein, due to its lower
homology with cellular chaperones. Utilization of the more conserved N-terminal
domain would more likely result in cross-reactivity of the serum with host heat shock
proteins. A fragment of 149 aa from the carboxyl-terminal end of the protein was
expressed in bacterial cells and used as antigen for chicken immunization. The
polyclonal antiserum obtained reacted specifically with the cognate antigen, and also with
a full-size recombinant CTV-p65 protein on Western blot analysis.
Using this antibody, it should be possible to study the immuno-localization of the
viral protein in CTV-infected tissue, and in this way, improve the understanding of the
function of this gene product in the process of viral infection as well as its interaction
with cell components.
IN VIVO LOCALIZATION OF THE HSP70 PROTEIN HOMOLOG (HSP70h) IN
CITRUS TRISTEZA CLOSTEROVIRUS INFECTED PLANTS
Citrus tristeza virus (CTV), the causal agent of the most economically important
virus disease of citrus, is a closterovirus with filamentous particles of about 2000 x 11 nm
in size. CTV is transmitted by man via infected buds and locally spread by various aphid
species (Bar-Joseph and Lee, 1989). Toxoptera citricida and Aphis gossypii are the most
efficient vectors, transmitting the virus in a semipersistent manner. The virus causes
different disease syndromes depending on the isolate and the scion/rootstock combination
of the citrus tree. Some CTV isolates cause decline and death of trees on the most
desirable horticultural rootstock, sour orange (Citrus aurantium). Other isolates induce
the stem pitting disease on the scion, which results in reduced vigor and small fruit on
infected trees. There also are some mild isolates, which do not elicit symptoms on
infected citrus plants (Bar Joseph, 1989; Bar-Joseph and Lee, 1989; Roistacher, 1991).
The CTV-genome is a single-stranded, positive sense RNA molecule of 19226-
19302 nt, organized into 12 open reading frames (ORFs), potentially encoding at least 19
protein products (Figure 3-1) (Albiach-Marti et al., 2000; Karasev et al., 1995; Mawassi
et al., 1996; Pappu etal., 1994; Vives etal., 1999). CTV also contains two untranslated
regions (UTRs) of about 105-108 and 273 nt at the 5' and 3' termini, respectively. ORF
la encodes a 349-kDa polyprotein with two protease-like domains, followed by a
methyltransferse-like and helicase-like domains. ORFlb contains an RNA-dependent
RNA polymerase-like domain that is thought to be expressed by a +1 translational
frameshift (Cevik et al., 1999; Karasev et al., 1995). ORFs la and lb are translated
directly from the genomic RNA, whereas ORFs 2 through 11 are expressed via sub-
genomic RNAs (sgRNAs) that are 3'co-terminal (Hilf et al., 1995; Karasev et al., 1995).
There is limited information available about the function of the CTV encoded
proteins (Manjuntah et al., 2000). Early reports about function established that the
product of the ORF6 and ORF7, which are the minor capsid protein (CPm) and the capsid
protein (CP), respectively, encapsidate the virion (Febres et al., 1996; Sekiya et al.,
1991). More recently, it was reported that the p23 (ORF11) is an RNA binding protein
(Lopez et al., 2000), p20 (ORF 10) accumulates in the amorphous inclusion bodies
(Gowda et al., 2000), and that p65 (ORF4) and p61 (ORF5) in addition to both CPs are
necessary for efficient virion assembly (Satyanarayana et al., 2000). Computer analysis
and in vitro translation experiments have shown that at least five proteolytic products are
predicted to be processed from the polyprotein product (ORFla) in CTV infected plants
(Erokhina et al., 2000; Karasev et al., 1995; Vazquez, 2001). Some of the CTV-encoded
proteins have been detected in vivo using antibodies developed against bacterial
expressed proteins. Using this strategy, the CP (Manjunath et al., 1993; Nikolaeva et al.,
1995), CPm (Febres et al., 1994), p20 (Price et al., 1996), RdRp (Cevik et al., 1999) and
p23 (Pappu et al., 1997) have been detected in CTV-infected but not in healthy tissue .
So far the CTV-HSP70h, the product of the ORF4, has not been detected in vivo.
The purpose of this research was to study the in vivo localization and expression of the
HSP70h from CTV infected citrus tissue, as well as the association of this protein with
the virion. A polyclonal antibody previously developed against this protein was used in
tissue printing, co-precipitation and immunogold labelling experiments. The results
obtained confirm that this protein is expressed upon virus infection, and that there is a
close association between the HSP70h and the filamentous virion of CTV.
1B 2 3 4
5 6 7 8 9 10 11
CPm CP 23
CPm CP p13 p23
Figure 3-1. Representation of the citrus tristeza closterovirus genome. Open reading
frames are shown as boxes. Putative domains in the ORFla are separated by lines. P-Pro:
Papain-like proteases 1 and 2; MTR: putative methyltransferase; HEL: putative helicase;
RdRp: RNA-dependent RNA polymerase; HSP70h: heat shock protein homolog; CPm:
minor coat protein; CP: coat protein.
Material and Methods
Tissue printing. Tissue blots were prepared as described Garnsey et al. (1993).
They were prepared from stem or petiole pieces from healthy or CTV-infected citrus
tissue. A smooth fresh cut was made with a razor blade, and the cut surface was pressed
gently and evenly to a nitrocellulose membrane. After drying the membrane for 10-30
minutes, it was blocked with phosphate-buffered saline (PBS) buffer (0.02 M sodium
phosphate buffer with 0.15 M sodium chloride, pH 7.4.) plus 1% bovine serum albumin
(BSA) for one hour. After three washes, the membrane was incubated with the primary
antibody overnight at room temperature (CREC 35 for CTV-CP, dilution 1:10,000; UF-
C17 for CTV-HSP70h, dilution 1:1000). The membrane was washed again, and then
incubated with the corresponding secondary antibody that was conjugated with alkaline
phosphatase (dilution 1:30,000) for four hours at 37C. After washing the membrane
three times, it was developed with the alkaline phosphatase substrate (NBT-BCIP). All
washes were done with PBS-Tween 20 (PBS + 0.5 ml Tween 20 per liter) (PBST), for 5
minutes with gentle agitation.
Immunoprecipitation CTV particles were immunoprecipitated using
paramagnetic beads (Dynal Co.) coated with sheep anti-rabbit IgG and CTV specific
antibodies (CREC35, produced in rabbit), or goat-anti-chicken and HSP70h specific
antibodies (UF-C17, produced in chicken). The Dynabeads-CTV virion complexes were
resuspended in 50 il of Western blot extraction buffer (0.125 M Tris HC1, pH6.8, plus
4% SDS, 20% glycerol, and 10% mercaptoethanol), and then boiled for three minutes.
Ten il of the resuspended virions were loaded and run in a 10% PAGE, electrophoresed
at 120 V for 2 hours, blotted to a nitrocellulose membrane, and then the membrane was
analyzed by Western blot using the polyclonal antisera raised against CTV-HSP70h (UF-
C17) or CTV-CP (CREC35) at a dilution of 1:1,000 and 1:5,000, respectively.
Serologically specific electron microscopy (SSEM). Adsorption of the CTV
particles to the grids was done following the SSEM procedure described by Derrick and
Brlansky (1976). Briefly, crude extracts from bark, stems or petioles from healthy or
CTV-infected citrus plants were obtained by chopping the tissue in extraction buffer
(0.05M Tris-HCl, pH 7.2, plus 0.15 M NaC1, 0.4M sucrose). CTV particles were
adsorbed to carbon-formvar coated copper grids using a CTV-CP specific antibody (1052
IgG, diluted 1:500) by floating the grid on drops of the crude extract for lh. After rinsing
the grids first in 0.05M Tris-HCl buffer, pH 7.2, and then with water, they were
positively stained with a 5% solution of uranyl acetate in 50% ethanol for 5 min, and
washed in 95% ethanol. The grids were viewed with a Philips 201 transmission electron
microscope at different magnifications.
SSEM-immunogold labeling. Adsorption of the CTV-particles to the grids was
done following the SSEM procedure described by Derrick and Brlansky (1976), and
outlined in the previous paragraph. Briefly, after adsorption of the particles to the grid,
they were first washed with 3 drops of buffer (0.05M Tris-HCl, pH 7.2) for three
minutes each, then in 2 drops of blocking buffer (0.05M Tris HC1, pH 7.2, with 0.1%
BSA) for 3 min each, and finally one extra drop of blocking buffer for 30 minutes. The
grids were then transferred to drops of primary antibody [UF-C17 (from chicken) and
chicken pre-immune antiserum were diluted 1:100, for the CTV-CP antibody (1052, from
rabbit) the dilution was 1:500 in blocking buffer]. The grids were rinsed with three drops
of blocking buffer for three minutes each, and then transferred to a drop of the secondary
antibody. For detection, a 10 nm gold-conjugated goat anti chicken IgG or a 15 nm gold-
conjugated goat anti rabbit IgG were used at a dilution 1:100 in blocking buffer. The
grids were rinsed first in blocking buffer, then in three drops of water, and finally stained
with 5% uranyl acetate in 50% ethanol and rinsed in 95% ethanol.
In Vivo Detection of the HSP70h by Tissue Printing.
An antibody raised against the carboxy-terminal end of the CTV-HSP70h was
previously produced in chicken (Chapter 2). This antibody showed a specific reaction
with the bacterial-expressed protein antigen (149 aa), as well as the full length bacterial
expressed chaperone homolog from CTV. To study whether the HSP70h protein was
expressed in infected plants, tissue-printing studies were conducted using this antibody
and a CP-specific antibody.
When the blots were treated with the CP antiserum, the imprint of the stem was
clearly visible with deep purple staining. The purple stained area was the phloem of
CTV-infected stems (Figure 3.2; row C, columns 2, 3 and 4). The healthy tissue imprint
showed a faint pink coloration that was easily distinguished from the intense purple
stained areas in the phloem of the CTV-infected samples (Figure 3.1; column 1, rows A,
When the antibody raised against the CTV-HSP70h protein (UF-C17) was used to
probe the blot, the chaperone homolog was specifically detected in CTV-infected but not
in uninfected citrus plants (Fig.3.2, membranes in row A and B, columns 2, 3, 4, as
compared to column 1). The localization pattern of the CTV-HSP70h and the CTV-CPs
were similar in the direct tissue printing studies, and this corresponded to the phloem
tissue. The amount of purple staining was less for the HSP70h than for the CP
suggesting a much lower level of expression for the CTV-HSP70h than for the CTV-CP.
The chicken antibody, UF-C17, did not react with healthy tissue, and only a faint pink
color similar to the reaction obtained with the CP antiserum against healthy tissue was
observed in the membrane (Figure 3-2, column 1, rows A and B).
Figure 3.2. Tissue prints of infected and healthy citrus stems after
incubation with HSP70h and coat protein specific antibodies.
Column 1 healthy citrus tissue; columns 2, 3 and 4 CTV-infected
tissue. Membranes in rows A and B were incubated with antibody
raised against the CTV-HSP70h (UF-C17). Membranes in row C,
were incubated with antibody raised against the CTV-CP (CREC35).
Immunoprecipitation of an HSP70h-CP Complex From CTV-Infected Plants.
Immunoprecipitation was used to determine whether the CTV-HSP70h could
interact with the CP in extracts of CTV infected plants. When the CTV-CP antibody
(CREC-35) was used to produce the precipitate, immunobloting of the precipitated
proteins with anti-CP antibody revealed the presence of CP in CTV-infected plants
(Figure 3.3 A, bottom, lanes 2 and 3), but not in the uninfected plants (Figure 3.2 A,
bottom, lanes 1). Furthermore, the HSP70h was also present in that precipitate. Its
presence was revealed by using the chicken antibody against the chaperone homolog in
the immunblot. A unique band, absent in extracts from uninfected plants, was detected in
extracts from CTV-infected tissue (Figure 3.3 A, top, lanes 2 and 3). This band has an
apparent molecular weight of 65 kDa. The antibody also showed a weak reaction with
some host proteins (probably cellular HSP70s) of greater molecular weight than the
closteroviral chaperone protein (Figure 3.3 A, top, lanes 1, 2 and 3).
When the extracts from CTV-infected plants were immunoprecipitated with the
CTV-HSP70h antibody, immunoblotting of the precipitated proteins with CP antibody
revealed the presence of the CP in the precipitate. In the left portion of Figure 3.3 B, the
proteins precipitated with the CP antibody and the proteins precipitated with the HSP70h
antibody are shown in a silver stained SDS-PAGE gel. An equivalent gel was transferred
to a membrane and blotted with the CP antibody (Figure 3.3 B, right portion). A
comparison of the intensity of the bands obtained with each antibody (right portion, lanes
2 and 3) demonstrates that the CP antibody was more efficient in the precipitation of the
complex then the CTV-HSP70h antibody. When the chicken preimmune serum was used
to immunoprecipitate proteins from healthy and CTV-infected citrus tissue, the CTV-CP
antibody did not react with any proteins transferred to the membrane.
To further examine the association between the HSP70h and the CP of CTV,
immunogold labeling of CTV particles was performed with the same antibody used in the
co-precipitation experiments. Before doing the decoration of the virus particles, the
relative amount of CTV virions present in the extracts was determined by SSEM (Figure
3.4). The decoration was performed only if enough particles were present. All the CTV
isolates used for this experiment, T3, T4, T36 and T3800 showed high virus titer and,
subsequently, there were sufficient virions on the grids for decoration. Labeling of the
virions using the CP antibody (1052) was done as a positive control to ensure that the
technique was working properly. Figure 3.5 shows the gold particles concentrated along
almost the entire length of the virion when these were labeled with the CP antiserum.
The number of gold particles were variable, but there were usually more than 13 for each
full length CTV-particle.
When the chicken antibody produced against the CTV-HSP70h was used to
decorate the trapped particles, the HSP70h protein was detected in close association with
the virions. There were a variable number of gold particles associated with each virion,
ranging from 2 to 11 (Figure 3.6). Most of the gold label was associated with full length
virions or with virion fragments. This can be better seen in Figure 3.6 E where the virion
is labeled with eleven gold particles and there also are some gold particles associated
with fragmented CTV virions. Thus, the CTV-HSP70h protein was detected in close
association with the virions and virion fragments.
Figure 3.3. Interaction of CTV-HSP70h and CTV-CP in CTV infected tissue.
Extracts from CTV-infected and uninfected citrus plants were incubated with
paramagnetic beads either coated with anti-CP or anti-HSP70h antibody. The
precipitated proteins were analyzed by immunoblotting of the PAGE-separated proteins
with either anti-HSP70h or anti-CP antibody.(A). Top: Precipitated proteins obtained
with the CP antibody, and then the membrane probed with the antibody to the CTV-
HPS70h. Bottom: The same precipitates probed with the anti-CP antibody. Lane 1:
healthy tissue, lanes 2 and 3, immunoprecipitated virions from CTV infected tissue (B).
Left half: Silver stained PAGE of the immunoprecipitated proteins. Right half: Immuno-
blotting of the PAGE-separated proteins using the CTV CP antibody. Lane 1: protein
standard, lane 2: precipitate obtained from CTV-infected tissue with the CTV CP
antibody, lane 3: precipitate obtained with the CTV HSP70h antibody, and lane 4:
healthy citrus tissue.
Figure 3.4. Serologically specific electron microscopy (SSEM) of trapped citrus
tristeza virus (CTV) particles. The CTV coat protein specific antibody was used
for trapping the particles on the grid. The magnification of the micrograph is
Figure 3.5. Immunogold labeling of citrus tristeza virus trapped particles
using the coat protein specific antibody. The CTV-CP specific antiserum
1052 was used for trapping of the particles to the grid. The magnification of
the micrograph is 45,000x
' ,c"' ,, A'
S -^ :. ai,/
D E '
Figure 3.6. Immunogold labeling of citrus tristeza virus particles using the CTV-
HSP70h-specific antibody. (A) CTV particles labeled with the preimmune serum.
(B) to (F). Decoration of CTV virions with the gold particles conjugated to the
CTV-HSP70h specific antibody. All of the micrographs are at 45,000X
The proteins from the HSP70 family of molecular chaperones are conserved
among unicellular and multicellular organisms (Chervitz et al., 1998; Guy and Li, 1998;
Tatusov et al., 1997). The major known functions of the HSP70 proteins are to mediate
the correct folding, assembly, intracellular localization, secretion, regulation and
degradation of other proteins (Gething, 1997). Closteroviridae members are the only
viruses known to encode an HSP70 homolog of cellular molecular chaperones-like
proteins (Karasev, 2000).
The purpose of this research was to study the in vivo expression of the CTV-
HSP70h and its possible association with CTV virions. Using an antibody specific for
this protein, it was possible to identify the presence of the CTV- HSP70h in infected
citrus tissue but not in non-infected tissue, using tissue printing studies. The localization
pattern of the CTV-HSP70h and the viral CP were similar in direct tissue printing, and
they corresponded with the location of the phloem tissue. Additionally using two
different experimental approaches, SSEM-immunogold labelling and
immunoprecipitation from plant extracts, provide strong evidence for the existence of
HSP70h-virion complexes in CTV-infected plants.
Co-localization of the HSP70h with virions, as reported in beet yellow virus
(BYV)-infected cells (Medina et al., 1999), and the detection of HSP70h in partially
purified preparations of lettuce infectious yellows virus (LIYV) (Tian et al., 1999),
suggest a physical association between the viral chaperones and the virions (Alzhanova et
al.., 2000). It also was shown for BYV that the HSP70h-virion complex is stable at high
salt concentrations, but the use of other dissociating agents, such as sodium dodecyl
sulfate, lithium chloride, or alkaline pH, resulted in at least partial virion disassembly.
The formation of the HSP70h-virion complex apparently does not involve covalent bonds
with any of the virion components (Napuli et al., 2000). These authors estimated the
number of viral chaperones bound to each virion. They separated virions and associated
proteins by SDS-PAGE, and the approximate amount of BYV-HSP70h was determinated
by comparison with standarized dilutions of a marker protein. By this approach, it was
found that each virion binds an average of 10 molecules of HSP70h protein. In this
present study, direct immunogold labeling of the CTV-HSP70h showed that there are a
variable number of molecules associated with each virion, ranging from 2 to 11. This
variation may result from the manipulations of the plant extracts during the process of
extraction and decoration of the particles, or to the fact that all the chaperone-specific
antibody binding sites were not saturated.
Complexes between the HSP70 and CP during viral infection have been reported
for viruses such as Sindbis virus (Garry et al., 1983), vesicular stomatitis virus (Garry et
al., 1983), adenovirus type 5 (Macejak and Luftig, 1991), poliovirus and coxsackievirus
(Macejak and Sarnow, 1992). The association of constitutively expressed cellular
chaperones and mature virions of animal RNA viruses has been reported for rabies virus,
vesicular stomatitis virus, Newcastle disease virus, influenza A virus (Sagara and Kawai,
1992) and canine distemper virus (Oglesbee et al., 1990). Because of the diverse
functions of the cellular chaperones and their increased expression and association with
viral proteins during viral infections, it is likely that HSP70 proteins assist in some
aspects of virion assembly as a cellular chaperone protein (Cripe et al., 1995).
Satyanarayana et al. (2000) have reported that the CTV-HSP70h appears to be
required for efficient virion formation, since mutations in the HSP70h gene resulted in
large decreases of the ability of the virus to be serially passage in Nicotiana
benthamiana protoplasts, as well as reductions in the proportions of full length particles.
Peremyslov et al. (1999) described that the chaperone homolog of BYV functions in
intracellular translocation and represents an additional type of plant viral movement
There are numerous possible functions that are feasible for the association of the
CTV-encoded chaperone with its virion during viral infection. It may be required not
only for virion formation, but also for virion disassembly during the process of viral
infection. The viral chaperone may play a role in aphid transmission, as has been
reported for some other plant viruses which require virus-encoded proteins other than
those on the virion for vector transmission (Harrison and Murant, 1984; Pirone, 1991).
CTV-HSP70h also can mediate the intercellular translocation of the virion as has been
reported for BYV. Finally, the CTV-HSP70h protein as a molecular chaperone could play
a role in preventing aggregation or assisting in the proper folding of the viral proteins
during the process of infection.
In summary, the data presented here confirm that the HSP70 chaperone homolog
encoded by CTV is expressed in CTV-infected plants and not in non-infected plants, and
that it forms a complex with the CTV virion or with the individual capsid proteins during
the process of CTV-infection.
THE CTV-HSP70h AS A COMPONENT OF CTV INCLUSION BODIES
The family Closteroviridae is comprised of more than 30 plant viruses with
flexuous, filamentous virions and includes representatives with either mono- or bipartite
positive sense single stranded RNA genomes. The Closteroviruses, a member of this
family, are a large and diverse group of viruses affecting several crops of major economic
importance, e.g. sugarbeet, citrus, tomato, sweet potato, grapevine, pineapple, cherry, and
some ornamentals (Karasev, 2000). Studies have shown that closterovirus infections
induce characteristic inclusion bodies (IB) within phloem associated cells, including
phloem parenchyma and companion cells (Lesemann, 1988). These IB include
cytoplasmic vesiculated membranous areas within infected cells, referred to as beet
yellow virus (BYV)-type vesicles, surrounded by lipid droplets (Medina et al., 1998).
Citrus tristeza virus (CTV) has been shown to produce IB that are confined
mostly to the phloem. The detection of CTV inclusions using light microscopy or in situ
immuno-fluorescence can provide a rapid method for diagnosis of CTV infection
(Brlansky, 1987; Brlansky et al., 1988). Early reports on CTV-infected tissue revealed
the presence of chromogenic bodies in the phloem parenchyma cells adjacent to sieve
tubes with strands of dark staining masses and needle-like structures (Kitajima and Costa,
1968; Schneider, 1959; Shneider and Sasaki, 1972). Schneider suggested that the
chromatic cells were the primary cytological symptoms from CTV infection, and that
they were involved in the development of wood pitting, vein clearing, and seedling
yellows symptoms. Staining of the inclusions with magenta and azure A suggested the
presence of nucleoproteins in fibrous and banded inclusions (Christie and Edwarson,
1977; Garnsey et al., 1980). Electron microscopic observations of thin sections revealed
the presence of large numbers of CTV particles packed in paracrystalline arrays in the
phloem of sieve elements (Bar-Joseph et al., 1979).
Studies have shown higher numbers of IB in various host species infected by
severe CTV isolates as compared to mild isolates (Brlansky and Lee, 1990). The effect of
virus strains or host on the morphology of the various CTV IB is not known. Recently,
Gowda et al (2000) reported the immunolocalization of the p20 protein (ORF 10) from
CTV with the amorphous IB present in CTV-infected tissue, suggesting that the p20
protein is a major component of this type of inclusion.
The purpose of this research was to determine if the CTV heat shock protein
analog (CTV-HSP70h) occurs in the characteristic IB present in CTV-infected tissue.
Material and Methods
Virus isolates and plant material. Florida CTV isolates T3, T66, T36, T30, T55
and T4, were used throughout the study. T3 causes severe decline on sweet orange
grafted on sour orange and severe stem pitting and vein clearing on Mexican lime.
Isolate T36 produces mild seedling yellows and decline on sour orange, T66 produces
strong reaction on Mexican lime and decline on sour orange, and T30 and T55 isolates
produce mild symptoms on Mexican lime, and no noticeable symptoms on commercial
citrus trees. T4 isolate produces a strong reaction on Mexican lime, but it is negative for
seedling yellows and decline on sour orange. The host selected for inclusion body
purification was Mexican lime. Healthy Mexican lime tissue control was included for
Light microscopy. Petiole samples approximately 0.5 cm long were excised
from the leaves at the abscission zone. Transverse sections were prepared using a Harris
WRC cryostat-microtome (Harris Manufacturing, Inc). Sections were stained according
to the method described by Garnsey et al (1980). Briefly, sections were stained for
approximately 15 minutes in 0.05% Azure A in 2-methoxyethanol and buffered with
0.2M Na2HPO4. Stained sections were washed sequentially in 95% ethanol and 2-
methoxy ethyl acetate, mounted in Euparal and observed under the light microscope.
Inclusion body purification. A protocol based on the purification procedure
described by Lee et al., (1982) was used in this study. Briefly, tender bark tissue was
homogenized in TSM buffer (0.05M Tris-HC1, pH 8.0, 10% sucrose, 0.5% 3-
mercaptoethanol), using Ig of fresh weight tissue and 15 ml of TSM buffer, and then
centrifuged at 10,000 x g for 15 min at 40C. The pellet was homogenized in 15 ml of
TSM buffer using a Polytron type homogeneizer and filtered through two layers of
cheesecloth. The filtrate was centrifuged at 4,000 x g for 15 min through a 5 ml pad of
20% sucrose made in TSM buffer. The pellet was resuspended in 5 ml of buffer (0.05M
Tris HC1, pH 8.0 plus 5% Triton X-100), let set for 30 min at room temperature, and then
centrifuged at 2,500 x g for 10 min. The resultant pellet was resuspended in 2 ml of TSM
buffer and layered onto a cesium sulfate step gradient prepared by layering 3 ml of 1.5
molal CsSO4, 3 ml 1.0 molal CsSO4, and 3.0 ml 0.5 molal CsSO4. All the CsSO4
solutions were prepared in 0.05M Tris HC1, pH 8.0 plus 30% sucrose. The step gradient
was centrifuged at 36,000 rpm for 3.5 hours at 120C using a SW41 Beckman rotor.
Following centrifugation in the cesium sulfate gradient, the IB (IB) were localized about
12 to 34 through the gradient. The bands containing the IB were collected, washed with
TSM buffer, and loaded into a new step centrifugation gradient, this time prepared by
layering 1.5 ml 1.5 molal CsSO4, 4 ml Imolal CsSO4, and 4 ml 0.5 molal CsSO4. This
step gradient was centrifuged at 36,000 rpm for 12 hours at 120C using a SW41 Beckman
rotor. The IB appeared as compact greenish bands located in the lower half of the
Fluorescent antibody microscopy. Antibodies raised against the following CTV
proteins were used as primary antibody : coat protein (CP), minor coat protein (p27), p20
and HSP70h (p65). A volume of 100 tl from the fractions containing the partially
purified IB was pipetted onto a black polycarbonate membrane (0.2 micron, Poretics
Corp). The membrane was overlayed with 3 drops of the primary antibody diluted 1:20 in
the antibody buffer (0.02M Tris-HCl pH 8.2, 0.9% NaC1, 1% gelatin, 0.1% BSA) for 30
minutes at room temperature. The membrane was then washed by flooding with a pipette
full (approximately 1 ml) of antibody buffer, let set for 1 minute and the buffer pipetted
off. The membrane was overlayed with 3-4 drops of goat anti-chicken antibody- (for
detection of HSP70h) or goat anti-rabbit antibody (for detection of CP, p27 and p20
proteins)- tetramethyl-rhodamine (TRITC) labeled IgG diluted 1:20 with antibody buffer
and incubated for another 30 min. The membrane was washed, and then mounted on a
glass microscope slide with the sample side up using 1-2 drops of Aqua Mount mounting
media (Lerner Laboratories) and covered with a cover slip. The samples were observed
using a Zeiss Dialux fluorescence microscope, at 40X magnification.
SDS- polyacrylamide gel electrophoresis (SDS-PAGE). An aliquot of 100 tl
from the fraction contained the IB was resuspended in 500 [l of ESB buffer (9M urea,
4.5% SDS, 7.5% a-mercaptoethanol and 75 mM Tris-HC1, pH 6.8) and gently agitated.
The samples were then boiled for 10 minutes and loaded onto a 10 or 12% SDS-PAGE
gel that was prepared following standard protocols (Sambrook, 1989). The gels were
electrophoresed at 120 Volts for 2 hours.
Protein silver staining. After electrophoresis, the gel was fixed in a solution
50% methanol, 5% acetic acid (v/v) in water for 30 minutes. Then the gel was washed
first in 50% methanol for 10 min, and then in water for other ten minutes. The gel was
sensitized by one minute incubation in 0.02% sodium thiosulfate, and rinsed with two
changes of distilled water for 1 minute each. After rinsing, the gel was submerged in a
chilled 0.1% silver nitrate solution and incubated for 30 minutes at 40C. After
incubation, the gel was rinsed twice with water for 1 minute and developed in 0.04%
formalin in 2% sodium carbonate with gentle shaking. After the desired staining
intensity was achieved, the development of the gel was stopped by decanting the reagent
and washing the gel with 5% acetic acid (v/v).
Western blot analysis. The protein samples were loaded onto a 10% SDS-PAGE
gel, and electrophoresed at 120 Volts for 2 hours. The proteins were then blotted to a
nitrocellulose membrane and blocked with 3% BSA in TBST buffer (25 mM Tris-HC1,
pH 8.0, 125 mM NaC1, 0.1% Tween 20). After, overnight incubation with the primary
antibody (UF-C17, a chicken antibody against the CTV- HSP70h, diluted 1:1000 in
blocking solution, or CP antibody, 1052 raised in rabbit against purified virus, at the
same dilution) and 4 hours with the secondary antibody (alkaline phosphatase conjugated
goat antichicken IgG or goat anti rabbit, diluted 1:30,000 in blocking buffer), the
membrane was developed using the NBT-BCIP substrate.
Light Microscopy and Inclusion Body Purification
The Azure A staining procedure was used to observe the IB in CTV-infected
tissue. Inclusions were observed in the phloem, phloem fiber and parenchyma cells of
CTV-infected tissue, but not in sections taken from healthy plants (Figure 4-1). Those
plants showing staining of IB by Azure A were selected for the IB purification procedure.
As previously reported by Brlansky et al.(1990), more IB were observed in tissue
infected with severe CTV isolates than in tissue infected with mild isolates.
A protocol for purification of IB based on the previous procedure described by
Lee et al. (1982) was used with an additional gradient centrifugation step added. Extracts
from healthy Mexican lime were included as a negative control to enable comparison of
the bands obtained from infected and healthy citrus tissue. After the first step gradient
centrifugation, the starch granules were pelleted at the bottom of the tube in all the
samples, but the bands containing the IB were not easily differentiated from the host
proteins when compared with the gradient loaded with healthy tissue extracts (Figure 4-2;
A throughout E). A second step gradient centrifugation allowed a more compacted
banding of the IB, with these bands located at different positions in the gradient
compared to the proteins present in the healthy tissue, which formed a band near the
bottom of the tube (Figure 4-2, F throughout J). The bands located in the lower half of
the cesium sulfate gradient were collected, and in cases where there was not a clear
differentiation between the IB and the host proteins (as for example, isolates T30 and T55
in Figure 4-2), the bottom fraction was split into two fractions. All fractions collected
were enumerated from top to bottom for further analysis. An estimation of the density at
the different fractions was calculated by weighing 100 il of volume collected for each
fraction. The band present in the cesium sulfate gradient loaded with healthy tissue had a
density (g/ml) of 1.63, and the densities for the fractions containing the CTV IB ranged
from 1.28 to 1.47 (Figure 4-2).
Figure 4-1. Azure A staining and light microscopy of leaf petiole sections of
healthy and CTV infected tissue. A and B, sections of healthy citrus tissue at a
magnification of 13.2 X and 80X respectively. From C throughout H, sections of
CTV-infected tissue at different magnifications. 25X in C, 40X in D and F, 80X
in E and G, and 160X in H. The isolates shown in the figure are T4 (C), T3 (D
through G) and T4 (H).
Fluorescent Antibody Microscopy
In an attempt to look for the presence of the HSP70h in the partially purified IB,
an aliquot of the IB was incubated with the UF-C17 antibody or the preimmune chicken
antiserum, and then with the secondary antibody (goat anti-chicken) labeled with TRITC.
Because the preimmune chicken antiserum showed reactivity with the proteins contained
in the IB fractions, the fluorescence observed when the antibody UF-C17 was used as
primary antibody could not be attributed to the presence of the CTV-HSP70h in the IB.
The secondary antibody, by itself, did not react with the IB. When a rabbit preimmune
Healthy T36 T30 T55 T3
Figure 4-2. Inclusion body purification. The procedure consisted of two
consecutive cesium sulfate step gradients. From A through E the bands obtained
after the first gradient are shown, and from F through J, the pattern obtained after
the second step gradient. The tissue loaded onto each gradient is shown at the top
of the figure. Densities (g/ml) calculated for the bands isolated from the gradient
are shown at the left.
was used as primary antibody, there was no fluoresce associated with the IB (Figure 4-3,
A). Antiserum raised against the p20 (ORF12), the CP (ORF7 ), and the p27 (ORF6),
previously developed in C.L. Niblett's lab, were used to study the composition of the
proteins isolated from the gradient centrifugation. The antibody reacting with the CP
(antibody 1052) showed a strong red-orange fluorescence when used with the IB (Figure
4-3, B and C). The antibody reacting with the p27 also showed some fluorescence when
it was used as primary antibody, but the number of foci were considerably smaller in
number and size than the fluorescence foci obtained with the CP antiserum (Figure 4-3,
D). When the antibody raised against the p20 protein was used as primary antibody,
there was no fluorescence associated with IB.
Figure 4-3. Immunofluorescence of proteins contained in the CTV inclusions
using a TRITC-labeled conjugate. The inclusion bodies were incubated with the
rabbit preimmune serum (A), the antibody (1052) raised against the CP-CTV (B
and C), or the p27 antibody (D). All pictures are shown at 40X magnification.
Analysis by SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)
The different fractions collected from the cesium sulfate step gradient
centrifugation were loaded onto a 10 or 12% SDS-PAGE gel. Proteins collected from
healthy tissue fractions were used to detect the presence of host proteins in the fractions
collected from CTV-infected bark. Three size groups of proteins were stained in the gels.
One group around 15 kDa, other group around 25 kDa, and the last group had a
molecular mass higher than 55 kDa (Figure 4-4).
The proteins present in the fractions collected from healthy tissue were also
detected in the fractions collected from CTV-infected plants. This was unexpected since
the bands occurred at different positions in the cesium sulfate gradient. However, several
proteins were present only in IB fractions from extract from infected tissue. These may
be virus-encoded proteins, or proteins from the host expressed as a response to virus
To ensure the reproducibility of the proteins patterns obtained, several IB
extractions were performed from plants infected with the different CTV isolates. The
proteins obtained were reproducible and consistent for the different extractions.
Interestingly, there was similarity in the pattern of the proteins obtained for the top
fractions from the severe isolates T36 and T66, and the mild isolates T55 and T30,
respectively (Figure 4-4).
The SDS-PAGE using a 10% gel enabled a better resolution of the proteins
having molecular masses higher that 55 kDa. In this size group of proteins there were
proteins which were exclusively present in the CTV-inclusions, as well as other proteins
that also were present in the healthy tissue. In general, less variation between isolates,
12 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 910
66- 31 --
55 21 -
C 2 3 4 5 67 8 910 D 1 2 3 45 67 8 9 10
Figure 4-4. Silver staining of SDS-PAGE gels containing proteins from partially
purfied CTV inclusion bodies (IB) from citrus tristeza virus (CTV) infected and
healthy bark tissue
Gels of 10% (B and D) and 12% (A and C) were loaded with a fraction of the
purified and ESB-solubilized IB proteins. Gels A and B: Lane 1: protein
standard; lane 2: fraction 1 from T66; lanes 3, 4 and 5: fractions 1, 2 and 3 from
T36; lanes 6 and 7: fraction 1 and 2 from T4; lanes 8 and 9: fractions 1 and 2 from
T3; lane 10: proteins from healthy tissue. Gels C and D: Lane 1: protein standard;
lanes 2 and 3: fractions and 2 from T55; lane 4: fraction 1 from T30, lane 5:
fraction 2 from T66; lane 6: fraction 4 from T36; lane7: fraction 3 from T4; lane
8: fraction 3 from T3, lane 9: fraction 3 from T55, lane 10 proteins from healthy
tissue. The numbers to the left indicate the molecular masses of the standards run
in lane 1
1 2 3 4 5 6
Figure 4-5. Western blot detection of the CTV-CP in the purified inclusion
bodies. Lane 1: protein standard; lanes 2 and 3: fractions 1 and 2 from T55; lanes
4, 5, and 6: fractions 1, 2, and 3 from T36; lane 7: fraction 1 from T30; lanes 8
and 9: fractions 1 and 2 from T3; lane 10: fraction 3 from T55; lane 11: fraction 4
from T36; lane 12: fraction 2 from T30; lane 13: fraction 3 from T3; lane 14:
proteins from healthy tissue.
Figure 4-6. Western blot detection of the CTV-HSP70h in the purified inclusion
bodies (IB). The B-proteins loaded in each lane are shown at the top of the
10 11 12 13 14
and the presence of more host proteins was found in this size group of proteins than in the
proteins of lower molecular mass, as detected by the silver staining.
Western Blot Analysis
Western blot analysis was performed with the proteins from fractions from the
cesium sulfate gradient. When the CP antibody was used to probe the membrane, the CP
was found at different concentrations in all fractions collected for the different isolates
used in this study. This band was absent in the proteins from fractions from healthy
citrus tissue (Figure 4-5)
The chicken antibody raised against the CTV HSP70h protein was used for
immunodetection of the HSP70h in the virus IB. The Western blots revealed a band with
a molecular mass higher than 60 kDa present in the viral IB fractions that was absent in
the fractions isolated from healthy tissue extracts. A membrane containing the
solubilized IB from the top fractions for isolates T36, T66 and T3 is shown in Figure 4-6.
Isolate T66 also contains an extra band of lower molecular weight, which may
correspond to a some proteolysis product of the HSP70h. No bands of similar molecular
mass were observed from fractions collected from healthy tissues (Fig. 4-6).
When the blots were treated with antibodies raised against the p20, p23, RdRp,
and p18 proteins, there was not detection of bands of the expected sizes associated with
the CTV IB. Only the p27 antibody gave a weak reaction when used for
immunodetection of the minor coat protein in the viral inclusions (not shown).
A protocol for purification of the IB present in CTV infected tissue was optimized
based on the previously reported procedure (Lee et al., 1982). This protocol, based on
two step gradient centrifugations, enabled recovery of fractions containing IB from the
CTV-infected tissue which sedimented at different positions in the gradient than the
proteins present in healthy citrus tissue extracts.
When these proteins were analyzed by fluorescent labeled antibody microscopy,
only the CP and p27 proteins were detected in the CTV inclusions. The non-specific
reaction by the chicken preimmune serum with the IB proteins precluded the use of this
technique to analyze the presence of the HSP70h in these structures.
The SDS-PAGE showed that there is no great variability in the protein
composition of the viral IB. Although there were not enough isolates included in this
study to associate a protein pattern with specific biological activity, there was similarity
in the patterns obtained between the top fractions of the severe isolates T36 and T66, and
the mild isolates T55 and T30, respectively. The silver stained gels also demonstrated
that the fractions containing the CTV IB also contained numerous host proteins, despite
their different locations following the step gradient centrifugation.
Western blot analysis indicated the presence of the HSP70h in the IB fractions
from CTV-infected tissue but present in low amounts. Therefore, it is necessary to
complement the present study with immunogold labeling analysis of ultrathin sections of
CTV-infected tissue to gain additional evidence about the presence of these proteins in
the viral IB.
The p20 protein from CTV was reported to be the major constituent of the
amorphous IB (Gowda et al, 2000). In that study, ultrathin sections from CTV-infected
tissue were examined by TEM, and the p20 antibody associated primarily with the
amorphous IB, while the CP-specific antibody associated primarily with the crystalline
IB which are thought to be composed of aggregates of virions.
We did not find evidence of the p20 protein in the purified CTV-IB. We do not
know if the purification procedure used in this study allowed the purification of both the
amorphous and crystalline inclusions which are associated with closterovirus infected
tissue. One possibility is that our procedure enhances the purification of the crystalline
inclusions, and not the amorphous structures, which could explain the absence of the p20
For many postive-stranded RNA viruses, replication is associated with cell
membranes (Buck, 1996). For assembly of the replicative complexes, some viruses
employ the use of pre-existing membrane organelles, whereas others induce their own
modification, leading to the formation of cytopathic ultrastructures. In closterovirus-
infected tissue, two types of IB have been described in phloem associated cells. These
are the amorphous and vesiculated inclusions, and the crystalline aggregates. The
proteins encoded by the first ORF from BYV, p63 (methyltransferase-like) and p100
(helicase-like), have been localized in the membranous-structures induced by the virus
and are referred as BYV-type vesicles. This ORF is known to encode the principal viral
products involved in RNA-replication and transcription in vivo. These results led to the
suggestion that the BYV-type of vesicle membranes are the specific sites of closterovirus
replication (Erokhina et al., 2000). However, Gowda et al. (2000) concluded that the
amorphous IB of CTV do not appear to be involved in virus assembly, since antibody
localization experiments showed little CP to be associated with these amorphous
The viral chaperone appears to be required for efficient virion formation
(Satyanarayana et al., 2000). Results from Chapter 3 indicate that the viral chaperone
protein (HSP70h) is expressed, and that it forms a complex with the CTV virion or with
the individual capsid proteins during the process of infection by CTV. In our study,
abundant CP was found in the purified IB, and the Western blot analysis also indicated
that the HSP70h was present in these structures. We do not know if the IB isolated using
the purification procedure described in this chapter are the site of CTV-replication, or if
the presence of the HSP70h in the inclusions is simply the result of the protein
association with the virion.
Further detailed ultrastructural analysis is necessary to provide information on the
specific localization of the HSP70h at the cellular level, and the types of IB purified by
the protocol described here.
AGROBACTERIUM-MEDIATED TRANSFORMATION OF DUNCAN GRAPEFRUIT
(Citrus paradisi; Macf)
Viral diseases cause serious losses worldwide in horticultural and agricultural
crops. Conventional breeding programs to develop resistance are effective, but the time
needed to release a new variety, and the possibility of integration of undesirable traits
make plant transformation technology an important approach for cultivar improvement.
The powerful combination of genetic engineering and conventional breeding programs
permits useful traits encoded by transgenes to be introduced into commercial crops within
a viable time frame (Hansen and Wright, 1999).
An alternative approach to engineering resistance was proposed by Sanford and
Johnson (1985); they suggested the possibility of engineering resistance by transforming
a susceptible plant with genes derived from the pathogen itself. This was named
"parasite-derived resistance", but subsequently the term "pathogen-derived resistance"
(PDR) became more commonly used. The PDR concept states that the expression of a
viral gene at either an inappropriate time or in inappropriate amounts or in an
inappropriate form during the infection cycle can perturb the ability of the pathogen to
sustain an infection. Therefore, plant viral transgenes can protect plants from infection
by the virus from which the transgene was derived.
Transgenic plants carrying plant virus-derived nucleotide sequences can exhibit
increased resistance to the viral diseases. Many viral sequences confer some level of
either resistance to infection or suppression of disease symptoms, which is better known
as tolerance (Fitchen and Beachy, 1993). The viral sequences used to developed PDR
include those encoding capsid or coat proteins (CP), viral replicase, movement proteins,
non- translatable sequences, defective interfering or satellite viruses, etc (Baulcombe,
In 1986, Powell et al. first demonstrated that transgenic tobacco plants expressing
the CP gene from tobacco mosaic virus (TMV) had an increased level of resistance to
TMV infection. Since then, numerous crop species have been genetically modified by
transformation with the viral CP with the intent of producing virus-resistant varieties
(Fitchen and Beachy, 1993). The majority of viruses for which PDR has been
successfully developed have positive-stranded RNA genomes. These include members of
the tobamo-, cucumo-, potex-, poty-, luteo-, carla-, ilar-, tobra-, nepo-, and alfalfa mosaic
groups (Lomonossoff, 1995).
The resistance mechanisms) are not yet completely understood, but it is known
that transgene products (RNA or protein) are involved (Jacquet et al., 1998). Evidence
that the accumulation of CP itself is responsible for at least some cases of coat protein
mediated resistance (CPMR) has been provided for both TMV (Nejidat and Beachy,
1989) and cucumber mosaic virus (CMV) (Okuno et al., 1993). Additionally, for
transgenic plants with heterologous or defective versions of movement proteins, there is
some evidence that the protein itself is responsible for the resistant phenotype (Lapidot et
al., 1993; Zhang et al., 1999). It also has been shown that only a dysfunctional form of
the movement protein can give rise to resistance, whereas the presence of wild type
protein may actually potentiate virus infection (Lomonossoff, 1995; Zhang et al., 1999).
Thus, the defect in the movement protein has the properties of a dominant negative
mutation (Herskovitz, 1987).
On the other hand, indications that the expression of a protein was not a requisite
for resistance came from the finding of the lack of correlation between protein levels and
resistance in plants containing CP genes from potato leafroll virus (PLRV), potato virus
Y (PVY) (Kawchuk et al., 1990), and tobacco etch virus (TEV) (Lindbo and Dougherty,
1992a; Lindbo and Dougherty, 1992b). In these cases, resistance did not require the
synthesis of any virus-derived protein, or protein fragment, but instead the expressed
RNA. This phenomenon became known as RNA-mediated resistance and was
characterized by a high level of resistance not easy to overcome by a high inoculum dose,
as compared to protein-mediated resistance (Lomonossoff, 1995). Its high sequence
specificity is another characteristic of RNA-mediated resistance, since this resistance
seems to be effective only against closely related viruses (Baulcombe, 1996;
Lomonossoff, 1995; Marano and Baulcombe, 1998). A possible mechanism to explain
the RNA-mediated resistance was proposed: the presence of the virus or the viral
homologous endogenous plant transcript is able to trigger a resistance mechanism active
in the cytoplasm which prevents virus replication in the cell (Dougherty et al., 1994;
Lindbo et al., 1993).
Recently it was proposed that the RNA-mediated virus resistance appears to
induce a form of post-transcriptional gene silencing (PTGS) (Baulcombe, 1996). The
PTGS mechanism is typified by the highly specific degradation of both the transgene
mRNA and the target RNA, which contains either the same or complementary nucleotide
sequences. If the transgene contains viral sequences, then virus genomic RNA cannot
accumulate in the plant (Lindbo and Dougherty, 1992a). In addition, it also was
proposed that PTGS is a manifestation of a natural virus resistance mechanism in plants
(Baulcombe, 1996; Pruss et al., 1997) since gene silencing can be induced by plant virus
infection in the absence of any known homology of the viral genome to host genes, and
because viruses can be initiators and targets of the gene silencing (Ratcliff et al., 1997).
Citrus tristeza virus (CTV) is a member of the Closteroviridae family. Since the
outbreak of decline in sour orange in the early thirties, CTV has caused widespread and
important economic losses because it kills citrus trees on sour orange rootstock or as a
result of stem pitting on the scion (Bar Joseph et al., 1989). Measures to control losses
caused by CTV include quarantine to avoid the introduction of exotic isolates,
certification schemes to prevent CTV spread and cross protection with mild isolates
(Cervera et al., 1998a; Rocha-Pena et al., 1995). Genetically engineered mild-strain
cross protection and RNA-mediated resistance are two strategies currently being
considered for management of CTV (Albiach-Marti et al., 2000b).
Genetic transformation and recovery of transgenic citrus trees has been achieved
in various species, hybrids and Citrus relatives such as sweet orange (Citrus sinensis (L)
Osbeck) varieties pineapple (Cervera et al., 1998a; Pena et al., 1995a) and Navel (Bond
and Roose, 1998), Carrizo citrange (Citrus sinensis L.Osbeck x Poncirus trifoliate)
(Cervera et al., 1998b; Gutierrezet al., 1997; Moore et al., 1992; Pefia et al., 1995b),
grapefruit (Citrus paradisi Macf.) (Luth and Moore, 1999; Yang et al., 2000), Poncirus
trifoliata (Kaneyoshi et al., 1994) and Mexican lime (Citrus aurantifolia Swing.)
(Gutierrez et al., 1997; Pefia et al., 1997). Citrus transformation procedures are often
inefficient due to the growth of non-transgenic or chimeric shoots during selection, low
frequencies of transformation and difficulties in rooting of transgenic shoots (Gutierrez et
al., 1997; Pefia et al., 1997).
There are some reports of the integration and/or expression of foreign genes other
than markers, in Citrus or its relatives. Among them, a chemically synthesized gene of
the human epidermal growth factor was transformed into Poncirus trifoliata under the
control of the cauliflower mosaic virus 35S RNA gene promoter, and the introduced
gene(s) were expressed in the young leaves of the regenerated plants (Kobayashi et al.,
1996). Constitutive expression of the Arabidopsis genes LEAFY and APETALAl
obtained by genetic transformation of citrange plants was associated with an appreciable
shortening of the juvenile phase of the citrus transformed plants (Pefia et al., 2001).
Attempts to develop PDR against CTV have been reported in the literature.
Gutierrez et al., (1997) produced transgenic Carrizo citrange, sour orange (C.arauntium
L.) and key lime plants expressing the CP gene from CTV. Later, Dominguez et al
(2000) reported the introduction of the CTV-CP into Mexican lime plants by using an
improved transformation protocol. This methodology used intemodal stem segments
from greenhouse-grown seedlings as explant material for transformation. Similarly,
Ghorbel et al. (2000) showed an enhancement of the transformation frequency of sour
orange by using explants from 4-month old seedlings grown in the greenhouse. This
method allowed them to introduce the CP-CTV gene into sour orange plants with an
efficiency of 3.6 +1 %. The p23 gene from CTV was also transformed into Mexican lime
plants to study whether the over-expression of this gene, or its truncated form, could
affect the normal CTV-infection process. Interestingly, the constitutive expression of
p23 induced a phenotype that resembled the CTV symptoms, whereas the plants
containing the truncated form of this gene were normal. They suggested that p23 gene
product is involved in symptom development and has a role in CTV pathogenesis
(Ghorbel et al., 2001b). None of the reports have shown data for evaluation for CTV-
The Closteoviridae is the only virus family that encodes a protein with similarity
to the cellular chaperones, a 70-kDa heat-shock protein homolog (HSP70h).
Satyanarayana et al. (2000) reported the involvement of HSP70h protein in CTV-
assembly. Mutations in the HSP70h-CTV gene resulted in a large decrease in the ability
of the virus to be passage in crude sap and in substantial reductions in the proportion of
full length particles. Recently, the HSP70h of BYV, a member of the Closteroviridae
family, was shown to be involved in intercellular translocation, representing a new type
of plant viral-movement protein (Peremyslov et al., 1999).
In this study, two different constructs (full-length and a frameshift mutant) of the
HSP70h gene from CTV were transformed into Duncan grapefruit seedlings to test the
possibility of inducing PDR, either by over expression of the CTV HSP70h protein or by
the expression of a truncated form of this protein.
Materials and Methods
Cloning and frameshift mutation of the CTV-HSP70h. The Florida CTV-
isolate T3800 was used as a virus source for the process of cloning the HSP70h gene.
The biological properties of this isolate have been described in the Material and Methods
in Chapter 2. The clone pGEM-T/HSP70h described in Chapter 2 was used as template
for the generation of the constructs described here. The restriction sites Apal and Xhol
were introduced at the 5' and 3' end of the CTV-HSP70h by PCR amplification, using
the primers CN394 (5'-GGGCCCATGGTGCTTTTAGGTTTAG-3') and CN401 (5'-
CTCGAGTCAGAGAGGT ATTCTTTCC-3'). Thermocycling conditions were 2 min at
940C, 40 cycles of 45 sec at 940C, 60 sec at 500C and 90 sec at 720C, followed for a final
extension of 5 min at 720C. The amplified product was cloned into the pGEMT vector,
generating the plasmid pGEM-T/HSP70h-1.
For the generation of the frameshift (FS) mutant, the vector pGEM-T/HSP70h-1
was digested with the restriction enzyme HindIII. After that, the linearized plasmid was
gel-purified using the Genclean II kit (BiolOl, Inc.) following the manufacturer
instructions; blunt-ended with Klenow fragment and religated to give pGEM-T/HSP70h-
1-Hindlllusing standard procedures (Sambrook, 1989). To confirm the frameshift
mutation, the plasmid was sequenced at the DNA Sequencing Core Lab, University of
Florida, using universal (forward and reverse) M13 primers.
A pUC 118-based plasmid vector containing the cauliflower mosaic virus 35S
promoter and termination signal (pUC 118-35S Poly 2-9) kindly provided by Dr. V.
Febres was used for the generation of the plant transformation constructs. First, the full
length and frameshift mutant of the CTV-HSP70h were subcloned into the ApaI-XhoI
cloning sites of the pUC 118-35S Poly 2-9 vector. This was done by consecutive
digestions of the plasmid vectors pGEM-T/HSP70h-1 and pGEM-T/HSP70h- 1-Hind III
with the restrictions enzymes Nco I, Xho I, and Apa I. This cloning located the sequence
of interest between the 35S promoter and the termination signal. These fragments were
then cloned into the Pst I sites of the binary plant transformation vectors pCambia-2201
and pCambia-2202, both containing the NPT-II gene as a selectable marker and GUS or
GFP as reporter genes, respectively. These modified pCambia vectors were then
introduced into Agrobacterium tumefaciens strain Agl using the cold shock
Agrobacterium co-culture, plant transformation and regeneration A protocol
previously described for transformation of epicotyl segments of Carrizo citrange and key
lime (Moore et al., 1992), and subsequently modified for transformation of etiolated
grapefruit seedling (Luth and Moore, 1999) was used for most of the transformation and
regeneration steps. A modification that involved an extra step in a shoot elongation
medium was included before grafting of the regenerated shoots, according to Yang et al.
Seed germination Citrusparadisi cv. Duncan seeds were peeled and sterilized,
first with 70% ethanol for 5 min and 0.525% hypochlorite solution plus 0.05% Tween-20
for 10 min, then they were rinsed five times with sterile distilled water. The seeds were
placed individually into 150 X 25 mm tubes containing half-strength MS medium (2.13
g/1 MS salts, 50 mg/1 myo-inositol, 15 g/1 sucrose and, pH 5.7) with 7 g/1 agar. The tubes
were kept in the dark at 28 C or at room temperature until the germinated seedlings were
used for transformation, approximately 4-6 weeks after planting (Figure 5-1).
Transformation of epicotyl segments. Agrobacterium tumefaciens strain Agl 1
containing the binary plasmid pCambia2201 or 2202, with either the CTV-HSP70h full
length or the mutant construct, was inoculated into YEP medium (10 g/1 Bactopeptone,
10 g/1 yeast extract and 5 g/1 NaC1, pH 7.0) containing the appropriate antibiotics. They
were grown overnight to log phase (OD600nm = 0.5-1.0) at 280 rpm and 280C. The cultures
were centrifuged at 40C and 5,000 rpm for 5 min, and the pellets were resuspended to a
final concentration of 5x10 8 cfu/ml in MS medium containing 100 mM acetosyringone.
The epicotyl portions of the etiolated seedlings were cut into 1 cm segments and
soaked in the Agrobacterium inoculum for 1 min. Then, the inoculated segments were
placed horizontally in petri plates containing co-cultivation medium (MS medium plus 7
g/1 agar, and 100 mM acetosyringone), the plates were sealed and kept in the dark at
room temperature for 2-3 days (Figure 5-1).
Selection and regeneration of transgenic shoots. After 2-3 days of co-
cultivation, the epicotyl segments were transferred to a shoot induction medium (MS
medium with 0.5mg/l benzyl adenine (BA) and 7g/l Bacto-agar) supplemented with 500
mg/1 Claforan to inhibit further growth of Agrobacterium and with 100 mg/1 kanamycin
sulfate for selection of transgenic shoots. The plates were maintained at 280C with a 16-
hour photo-period provided by cool-white fluorescent light for 6-8 weeks. Transfers to
fresh medium were made at 4 week intervals.
Rooting of transgenic shoots. When shoots appeared and reached about 5-10
mm in length, they were removed from the explants and placed on rooting medium (MS
medium with 0.5 mg/1 naphthalene acetic acid). Shoots remained on this medium for
approximately 6 weeks (Figure 5-1). Because after this time there was no evident root
formation in any of the shoots, they were transferred to a shoot elongation medium [BG
medium composed of MS salts and B5 vitamins, 0.2 mg /1 6-benzylaminopurine (BAP),
0.5 mg/1 gibbereic acid (GA3), 2.5% sucrose, pH 5.8 solidified with 8% agar]
containing 500 mg/1 Claforan before proceeding with the grafting.
Grafting of regenerated shoots. Two month old Carrizo citrange greenhouse-
grown seedlings were used as rootstocks for grafting ex vitrum. The seedlings were
decapitated, and a vertical excision was made 3-5 mm deep. The shoots were grafted by
cutting them into a V-shape, and inserting them into the incisions on the rootstock.
Leaves of the transformed shoots were removed. A standard 200 pl pipet tip was used to
hold the graft in place. After 2-4 weeks, the scion had grown new leaves, and the pipet tip
Analysis of regenerated shoots. Epicotyl segments transformed with the
different constructs previously described were examined periodically for the expression
of GFP and GUS, depending on the vector used. The expression of GFP in the
regenerated shoots was analyzed by using a dissecting microscope (Zeiss) with a
fluorescent light source with a 515 nm long pass emission filter transmitting red and
green light and a 450-490 nm excitation filter. GUS expression was analyzed by
histochemical staining. The leaves were placed in ELISA plates or 2-ml eppendorf tubes
containing the GUS staining solution (50 mM NaPO4, pH 7.2, 0.5% Triton X-100, ImM
5-bromo-4-chloro-3-indolyl- D-glucuronide (X-Gluc)) that was diluted from a 20mM
stock made in dimethylformamide). Vacuum was applied for 5 min to infiltrate the
leaves with the substrate, and then the tubes or plates were incubated overnight at 37C
with gentle agitation. The staining solution was removed, and the chlorophyll was
removed from the leaves with several washes in 70% ethanol.