<%BANNER%>

High Order Structure In The Babesia Bovis Locus Of Active ves Transcription

Permanent Link: http://ufdc.ufl.edu/UFE0024758/00001

Material Information

Title: High Order Structure In The Babesia Bovis Locus Of Active ves Transcription
Physical Description: 1 online resource (146 p.)
Language: english
Creator: Huang, Yingling
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: babesia, chromatin, dna, nuclease, nucleosome, structure, transcription, ves
Veterinary Medicine -- Dissertations, Academic -- UF
Genre: Veterinary Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Babesia bovis is an intraerythrocytic protozoan which causes severe disease in cattle following the transmission by its vector tick. It can persist life-long in animal hosts which survive acute infection. Cytoadhesion of parasite-infected red blood cells (IRBC) to the endothelium of host deep organs, and periodic antigenic variation of the adhesive ligands on the infected erythrocyte surface are two major survival strategies. Previous work had identified the heterodimeric variant erythrocyte surface antigen 1 (VESA1), which is encoded by ves1alpha/beta gene pairs among the ves multigene family, as a key bifunctional virulence factor due to its involvement in both cytoadhesion and antigenic variation. In the B. bovis C9.1 clonal line, the locus of active ves transcription (LAT) of ves1alpha/beta gene pairs has been revealed, and segmental gene conversion has been demonstrated to contribute greatly to B. bovis antigenic variation. The unique gene organization of the LAT, which contains two closely juxtaposed, divergent ves1alpha/beta genes sharing overlapping 5?untranslated regions (UTR), also exists at many other non-expressed ves gene loci. It is unknown how the LAT acquires its distinctive active transcription state among ves genes. To determine whether stable, higher order structure exists at the LAT which may be related to its active transcription, in vivo 3-dimensional DNA structural characteristics were examined initially through a combination of structure-protective genomic DNA extraction and DNA two-Dimensional (2-D) gel electrophoresis. In vivo psoralen crosslinking was also tried to stabilize LAT-associated higher-order DNA structure. Some higher-order DNA structures were detected with both structure-preserving methods and could most likely be associated with the LAT regarding active transcription although other evidences are essential to substantially prove and interpret their significance. In addition, an in vitro study also supported the potential of ves intergenic regions (IGr) to form higher-order structures when induced by thermal remodeling (TR). Several nucleases were used to further investigate the chromatin structural characteristics of ves genes. Several locus-specific nucleosomal repeat lengths were estimated found indistinguishable from bulk chromatin. In contrast, the LAT adopts an unusual chromatin structure featuring 5 MNase-hypersensitive sites (MH sites) in the IGr rather than three, and two protected IGr-flanking regions. The rates of hybridization signal losses from different gene loci upon the treatment with different concentrations of MNase or DNase I were estimated. Both enzymes produced three highly repeatable patterns of ves nuclease sensitivity. The LAT is significantly more sensitive than are some non-expressed ves genes or even the active B. bovis beta-tubulin and glyceraldehydes-3-phosphate dehydrogenase (GADPH) genes, consistent with their active state. In contrast, among non-expressed ves genes two populations may be defined: genes which are relatively insensitive to digestion, and those which are intermediate insensitivity between the LAT and the insensitive ves loci. All those facts suggested that different chromatin structures explain the differential transcriptional status among ves genes, which could be mediated by differential epigenetic modifications such as acetylation/methylation of histone H3.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Yingling Huang.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Allred, David R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024758:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024758/00001

Material Information

Title: High Order Structure In The Babesia Bovis Locus Of Active ves Transcription
Physical Description: 1 online resource (146 p.)
Language: english
Creator: Huang, Yingling
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: babesia, chromatin, dna, nuclease, nucleosome, structure, transcription, ves
Veterinary Medicine -- Dissertations, Academic -- UF
Genre: Veterinary Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Babesia bovis is an intraerythrocytic protozoan which causes severe disease in cattle following the transmission by its vector tick. It can persist life-long in animal hosts which survive acute infection. Cytoadhesion of parasite-infected red blood cells (IRBC) to the endothelium of host deep organs, and periodic antigenic variation of the adhesive ligands on the infected erythrocyte surface are two major survival strategies. Previous work had identified the heterodimeric variant erythrocyte surface antigen 1 (VESA1), which is encoded by ves1alpha/beta gene pairs among the ves multigene family, as a key bifunctional virulence factor due to its involvement in both cytoadhesion and antigenic variation. In the B. bovis C9.1 clonal line, the locus of active ves transcription (LAT) of ves1alpha/beta gene pairs has been revealed, and segmental gene conversion has been demonstrated to contribute greatly to B. bovis antigenic variation. The unique gene organization of the LAT, which contains two closely juxtaposed, divergent ves1alpha/beta genes sharing overlapping 5?untranslated regions (UTR), also exists at many other non-expressed ves gene loci. It is unknown how the LAT acquires its distinctive active transcription state among ves genes. To determine whether stable, higher order structure exists at the LAT which may be related to its active transcription, in vivo 3-dimensional DNA structural characteristics were examined initially through a combination of structure-protective genomic DNA extraction and DNA two-Dimensional (2-D) gel electrophoresis. In vivo psoralen crosslinking was also tried to stabilize LAT-associated higher-order DNA structure. Some higher-order DNA structures were detected with both structure-preserving methods and could most likely be associated with the LAT regarding active transcription although other evidences are essential to substantially prove and interpret their significance. In addition, an in vitro study also supported the potential of ves intergenic regions (IGr) to form higher-order structures when induced by thermal remodeling (TR). Several nucleases were used to further investigate the chromatin structural characteristics of ves genes. Several locus-specific nucleosomal repeat lengths were estimated found indistinguishable from bulk chromatin. In contrast, the LAT adopts an unusual chromatin structure featuring 5 MNase-hypersensitive sites (MH sites) in the IGr rather than three, and two protected IGr-flanking regions. The rates of hybridization signal losses from different gene loci upon the treatment with different concentrations of MNase or DNase I were estimated. Both enzymes produced three highly repeatable patterns of ves nuclease sensitivity. The LAT is significantly more sensitive than are some non-expressed ves genes or even the active B. bovis beta-tubulin and glyceraldehydes-3-phosphate dehydrogenase (GADPH) genes, consistent with their active state. In contrast, among non-expressed ves genes two populations may be defined: genes which are relatively insensitive to digestion, and those which are intermediate insensitivity between the LAT and the insensitive ves loci. All those facts suggested that different chromatin structures explain the differential transcriptional status among ves genes, which could be mediated by differential epigenetic modifications such as acetylation/methylation of histone H3.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Yingling Huang.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Allred, David R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024758:00001


This item has the following downloads:


Full Text

PAGE 1

1 HIGH ORDER STRUCTURE IN THE BABESIA BOVIS LOCUS OF ACTIVE Ves TRANSCRIPTION By YINGLING HUANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Yingling Huang

PAGE 3

3 To my husband, Wenqiang Tian, and my dear daughter, Alina Hanning Tian

PAGE 4

4 ACKNOWLEDGMENTS I sincerely thank m y major mentor Dr. Davi d Allred, who provides me with extensive support which is crucial to the completion of my work throughout my doctoral study. During this work, I benefited a lot from nume rous in-depth and inspiring scie ntific discussions between Dr. Allred and myself. With his great encouragement and assistance, I beca me more confident and independent as a grown re searcher needs to be. I also gratefully thank my supervisory co mmittee (Dr. John Dame, Dr. Anthony Barbet, Dr. David Ostrov and Dr. Jorg Bungert) who put fo rward many helpful suggestions to guide my research in this project. I would like to give many thanks to Dr Basima Al-Khedery, who offered me important help in teaching me southern blot ting procedures. Additionally, I want to sincerely thank Ms. Xuping Xiao and other labmates such as Xinyi Wang, Anne Bouchut and Alexia Berg and so on for providing me technical assistan ce and extensive laborat ory support. Finally, I would also like to thank my husband and my dear daughter for their love and support, which was the strongest drive for me to complete this work.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................7LIST OF FIGURES .........................................................................................................................8LIST OF ABBREVIATIONS ........................................................................................................ 10ABSTRACT ...................................................................................................................... .............12INTRODUCTION .................................................................................................................. .......14B. bovis Life Cycle and Pathogenesis .....................................................................................14Antigenic Variation in B.bovis ...............................................................................................15VESA1 Protein and C9.1 ves 1 / LAT .................................................................................. 16Inverted Repeats and Their Biological Importance ................................................................ 18DNA Two Dimensional Gel Electrophor esis and Its Application .......................................... 20Chromatin Structure and Gene Transcription ......................................................................... 21Nucleosomal Organization on Genes .............................................................................. 22Nucleases and Their Applications in Studying Chromatin Structure ..............................24Hypothesis ..............................................................................................................................26HIGH ORDER DNA STRUCTURE withIN BABESIA BOVIS LOCUS OF ACTIVE ves TRANSCRIPTION ................................................................................................................. 27Abstract ...................................................................................................................... .............27Introduction .................................................................................................................. ...........27Methods and Materials ...........................................................................................................29Parasite Culture ...............................................................................................................29Rupture of Infected Red Blood Cells (IRBCs) by Saponin Lysis ................................... 29Construction of Artificial DNA Cruciform Molecule .....................................................30In vitro DNA Thermal Remodeling Assay ......................................................................30DNA 2-D Gel Electrophoresis System ............................................................................ 31DNA Electron Microscopy ..............................................................................................33Detection of In vivo LAT-Associated DNA Structur e with Structure-Protective Extraction .................................................................................................................... .34Detection of In vivo LAT-Associated DNA Structure wi th Psoralen Crosslinking ........ 35Southern Blot Analysis ....................................................................................................36Results .....................................................................................................................................38Establishment of a DNA 2-D Gel System to Detect DNAs with Higher-order Structure ..................................................................................................................... ..38Construction of DNA cruc iform control molecule .................................................. 38Migration of thermally-remodeled cruciform molecule ........................................... 38DNA electron microscopy with constructured cruciform ........................................ 40

PAGE 6

6 Analysis of In vitro Structural Potential of Ves Genes .................................................... 41Migration of thermally remodeled ves gene fragments ............................................ 42DNA electron microscopy with structured ves fragments........................................43Detection of In vivo DNA Structure Associated with Ves Genes ................................... 43Detection of In vivo DNA Structure without DNA Crosslinking .................................... 44Detection of In vivo DNA Structure with DNA crosslinking .......................................... 49Discussion .................................................................................................................... ...........51Establishment of 2-D Gel System to Detect Higher-order DNA Structure in Chromatin .................................................................................................................... 51In vitro Structure Study of Ves Genes .............................................................................55Detection of In vivo LAT-Associated DNA Strucutre .................................................... 57CHROMATIN STRUCTURAL CHARACTERISTICS AMONG B. BOVIS ves GENES .......... 89Abstract ...................................................................................................................... .............89Introduction .................................................................................................................. ...........89Methods and Materials ...........................................................................................................91Parasite Culture ...............................................................................................................91Isolation of Parasite Nuclei .............................................................................................91Nuclease Digestion of Intact Nuclei ................................................................................91Micococcal nuclease digestion ................................................................................. 91DNase I digestion .....................................................................................................92S1 nuclease digestion ............................................................................................... 92Mung bean nuclease digestion ................................................................................. 92Southern blot analysis ..............................................................................................93Results .....................................................................................................................................93MNase Mapping ..............................................................................................................94DNase I Digestion ...........................................................................................................97Discussion .................................................................................................................... ...........99Nucleosomal Positioning Pattern among Ves Genes .....................................................100Nuclease Sensitivity Analysis ....................................................................................... 102SUMMARY ....................................................................................................................... ..........123PUTATIVE HISTONES AND HISTONE MODIFIING ENZYMES IN BABESIA BOVIS ... 130COMPARISON OF HISTONE H3 N-TERMINI BETWEEN BABESIA BOVIS AND OTHER ORGANISMS ........................................................................................................ 131THE VALIDATION OF COMMERIAL ANTIBOIDIES IN FIXED IFA ................................. 132LIST OF REFERENCES .............................................................................................................133BIOGRAPHICAL SKETCH .......................................................................................................146

PAGE 7

7 LIST OF TABLES Table page 2-1 List of primers used in cruciform control construction ..................................................... 612-2 List of oligonucleotide probes used in southern blot analysis ........................................... 612-3 List of the reagents us ed in structure-protective extraction of genomic DNA .................. 813-1 The sizes and AT content of examin ed gene loci in nuclease mapping. ......................... 119

PAGE 8

8 LIST OF FIGURES Figure page 2-1 Construction of cruciform control mo lecule B0.5 and the appearance of novel DNA for m(s) after thermal remodeling treatment. ...................................................................... 622-2 The novel forms produced after thermal re modeling are sensitive to S1 nuclease. .......... 632-3 The influence of concentrati on in the thermal remodeling of ves LAT gene sequence .... 642-4 The influence of concentrati on in the thermal remodeling of ves LAT gene sequence. ...652-5 Thermally remodeled B0.5 on DNA 2-D gel electrophoresis. .......................................... 662-6 2-D Southern blot analysis of therma lly remodeled B0.5. The 2-D gel (Figure 2-5) was blotted on Hybond N+ nylon me mbrane for hybridizations.. ..................................... 672-7 Structure loss during the pur ification of novel DNA forms. ............................................. 682-8 Visualization of the thermally-remo deled B0.5 through DNA electron microscopy ........ 692-9 The influence of melting temperatures (T m) in the thermal remodeling of the IGcontaining ves gene fragments. .......................................................................................... 702-10 The structure-forming potential of other ves IG-containing DNA fragments (1). ............. 712-11 The structure-forming potential of other ves IG-containing DNA fragments (2). ............. 722-12 Thermally remodeled ves IG-containing DNA fragments on DNA 2-D gel electrophoresis. .............................................................................................................. ....732-13 2-D Southern blot analysis of thermally remodeled ves IG-containing DNA fragments...................................................................................................................... ......742-14 Visualization of thermally remodeled ves IG-containing DNA fragments through DNA electron microscopy. ................................................................................................ 752-15 Optimization of saponin lysis fo r structure-protective extraction. .................................... 762-16 Test of oligonucleotide probes (1). ....................................................................................772-17 Test of oligonucleotide probes (2). ....................................................................................782-18 Migration of protectively extracted gDNA in TMSpe buffer and optimal buffer system for gDNA digestion. .............................................................................................. 792-19 Evaluation of a series of reagents in structure-protective extraction on the thermallyremodeled control molecules. ............................................................................................ 80

PAGE 9

9 2-20 2-D gel electrophoresis of C9.1 genom ic DNA e xtracted by structure-protective procedure. .................................................................................................................... .......822-21 Detection of in vivo DNA structures associated with ves loci (1). The 2-D blot was from the 2-D gel in Figure 3-20. ........................................................................................ 832-22 Detection of in vivo DNA structures associated with ves loci (2). ....................................842-23 Detection of in vivo DNA structures associated on the LAT with protective extraction. (A) Gel image of 2-D gel electrophoresis. ....................................................... 852-24 Detection of in vivo DNA structures associated on the LAT with psoralen crosslinking.. ................................................................................................................ ......862-25 Detection of in vivo DNA structures associated on the LAT with psoralen crosslinking .................................................................................................................. ......872-26 Higher-order DNA structures found on 2-D gels by two different methods. .................... 883-1 B. bovis bulk nucleosome organization. ..........................................................................1073-2 Examination of the overall nucleosomal organization of ves genes containing intergenic regions. ........................................................................................................... .1083-3 Gene-specific nucleosomal positio ning pattern detected by MNase. ..............................1093-4 Analysis of gene-specific nucleosomal positioning pattern detected by MNase.. ...........1103-5 LAT-associated nucleosomal posit ioning pattern dete cted by BAK52. .......................... 1113-6 Examination of LAT-specific probes. ..............................................................................1123-7 The confirmation of LAT-specific probes through regular southern blot analysis. Purified C9. ......................................................................................................................1133-8 C9.1-LAT specific nucleosomal positio ning pattern detected by MNase. ...................... 1143-10 Analysis of differential MNase sensitivity among ves genes. .........................................116311 Low resolution DNase I cleavage mapping. .................................................................... 1173-12 Analysis of differential DNase I sensitivity among ves genes.. ....................................... 1183-13 Low resolution of S1 cleavage mapping. .........................................................................1203-14 Low resolution of Mung bean nuclease cleavage mapping ............................................. 1213-15 Test of the activity of S1 nuclease and Mung bean nuclease. .........................................122

PAGE 10

10 LIST OF ABBREVIATIONS 6-CFDA 6-carboxyfluorescein diacetate ChIP Chromatin immunoprecipitation DAPI 4, 6-diamidino-2-phenylindol 2-D Two dimensional EB Ethidium bromide EDTA Ethylenediaminetetraacetic acid EGTA Ethylene glycol tetraacetic acid EM Electron microscopy FISH Fluorescence in situ hybridization Fixed IFA Fixed immunofluorescence assy GADPH D-glyceraldehyde-3phosphate dehydrogenase gDNA Genomic DNA HDAC Histone deacetylase IC50 Half maximal inhibitory concentration IGr Intergenic region IR Inverted repeat IRBC Infected red blood cell LAT Locus of active ves transcription MH site Micrococcal nucl ease hypersensitive site MNase Micrococcal nuclease MW Molecular weight NP-40 Nonidet-P40 NRL Nucleosomal repeat length PBS Phosphate-buffered saline

PAGE 11

11 PCR Polymerase chain reaction PIPES 1, 4-Piperazinediethanesulfonic acid ppe Percent parasitized erythrocytes qPCR quantitative polymerase chain reaction RBC Red blood cell RT Room temperature RT-PCR Reverse transcription po lym erase chain reaction TRThermally remodeledSDS Sodium dodecyl sulfate 5 UTR 5 untranslated region UVA Ultraviolet ves Variant erythrocyte surface gene VESA1 Variant erythrocyte surface antigen 1 VYMs Vega y Martinez solution

PAGE 12

12 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HIGH ORDER STRUCTURE IN THE B. BOVIS LOCUS OF ACTIVE ves TRANSCRIPTION By Yingling Huang December 2009 Chair: David R. Allred Major: Veterinary Medical Sciences Babesia bovis is an intraerythrocytic protozoan which causes severe disease in cattle following the transmission by its vector tick. It can persist life-long in animal hosts which survive acute infection. Cytoadhesion of pa rasite-infected red bl ood cells (IRBC) to the endothelium of host deep organs, and periodic an tigenic variation of the adhesive ligands on the infected erythrocyte surface are tw o major survival strategies. Previous work had identified the heterodimeric variant erythrocyte surface antigen 1 (VESA1), which is encoded by ves 1 / gene pairs among the ves multigene family, as a key bifunc tional virulence factor due to its involvement in both cytoadhesion and antigenic va riation. In the B. bovis C9.1 clonal line, the locus of active ves transcription (LAT) of ves 1 / gene pairs has been revealed, and segmental gene conversion has been demonstr ated to contribute greatly to B. bovis antigenic variation. The unique gene organization of th e LAT, which contains two cl osely juxtaposed, divergent ves 1 / genes sharing overlapping 5untra nslated regions (UTR), also ex ists at many other non-expressed ves gene loci. It is unknown how th e LAT acquires its distinctive active transcription state among ves genes. To determine whether stable, higher or der structure exists at the LAT which may be related to its ac tive transcription, in vivo 3-dimensional DNA structural characteristics were examined initially through a combination of structure-protective genomic DNA extraction and

PAGE 13

13 DNA two-Dimensional (2-D) gel electrophoresis. In vivo psoralen crosslinking was also tried to stabilize LAT-associated higher-order DNA structure. Some higher-order DNA structures were detected with both structure-pr eserving methods and could most likely be associated with the LAT regarding active transcripti on although other evidences are esse ntial to substantially prove and interpret their sign ificance. In addition, an in vitro study also supported the potential of ves intergenic regions (IGr) to form higher-order structures when induced by thermal remodeling (TR). Several nucleases were used to further investigate the chromatin structural characteristics of ves genes. Several locus-specific nucleoso mal repeat lengths were estimated found indistinguishable from bulk chromatin. In c ontrast, the LAT adopts an unusual chromatin structure featuring 5 MNase-hypersensitive sites (MH sites) in the IGr rather than three, and two protected IGr-flanking regions. The rates of hybridization signal lo sses from different gene loci upon the treatment with different concentrations of MNase or DNase I were estimated. Both enzymes produced three highly repeatable patterns of ves nuclease sensitivity. The LAT is significantly more sensitive than are some non-expressed ves genes or even the active B. bovis tubulin and glyceraldehydes-3-phos phate dehydrogenase (GADPH) genes, consistent with their active state. In contrast, among non-expressed ves genes two populations may be defined: genes which are relatively insensitive to digestion, a nd those which are intermediate insensitivity between the LAT and the insensitive ves loci. All those facts suggest ed that different chromatin structures explain the differentia l transcriptional status among ves genes, which could be mediated by differential epigenetic modifications such as acetylation/methylation of histone H3.

PAGE 14

14 CHAPTER 1 INTRODUCTION B. bovis Life Cycle and Pathogenesis B. bovis is a protozoal hemoparasite that invades m ammalian erythrocytes, establishing an infection through ixodid tick transmission. The disease caused by B. bovis infection occurs mainly in animals like cattle and buffalo (Bock et al., 2004). Bovine babe siosis accounts for significant cattle loss in subtr opical and tropical areas (Calvo de Mora et al., 1985; Kjemtrup and Conrad, 2000; Krause et al ., 2008). The parasite life cycle is composed of two major stages: an asexual stage in mammalian hosts, and a sexual stage in ticks. The asexual stage starts with the invasion of host erythrocytes by in fectious sporozoites, which are introduced into host tissues at the site of the bite of an infected tick. Th e parasites rapidly find their way into the host bloodstream. After successful entry via the vect or blood meal, the sporozoites multiply in erythrocytes through binary fi ssion to produce two daughter cel ls called merozoites, which rupture erythrocytes and thus are released into host bloodstream where they have the opportunity to invade new erythrocytes (Freidhoff, 1988). The merozoites can be ingested by non-infected ticks during the taking of a blood meal, and in the ticks midgut they start the sexual cycle by forming gametocytes and developing into ray bodies thereafter. Then, ray bodies experience multiplication and division to become gametes, two of which merge to form a zyote. Furthermore, zygotes infect the gut tissue of tick and develop in to kinetes through a series of complicated events, which get chance to invade tick oocytes to start new round of schizogony. As a consequence of larval tick attachment, sporozoites begin to form and invade mammalian host thereafter (Bock et al., 2004). The parasites can alte r the membrane of parasitized erythrocytes morphologically and antigenically in adaptation to the new living environment (Aikawa et al ., 1985; Allred et al.,

PAGE 15

15 1993). Some ridge-like structures can be developed on the surface of IRBCs. This modification assists in the cytoadhesion to th e capillary endothelium of some deep organs, including the liver, kidney and brain, and ther eby the sequestered parasites can avoid being eliminated during spleen passage by host inna te immune mechanisms (Aikawa et al., 1992; O'Connor et al., 1999). In addition, emerging variant antigens on IRBCs appear to go unrecognized by pre-existing host hu moral immunity, so that I RBCs expressing novel isoforms are not targeted for elimination (Allred et al., 1994). Most hosts can survive acute babesiosis, which is characterized by a variety of symp toms including high fever, severe anemia, hemoglobinuria, coagulatory disturbances, circulatory, respir atory, and sometimes cerebral complications, but sometimes death can occur, especially in the immunocompromised host (Yeruham et al., 2003; Krause et al., 2008). However, the parasites cannot be easily cleared from the infected host, hence long-lasting persistent infection is es tablished and maintained, which features density fluctuation of parasitemiae a nd is often asymptomatic (Calder et al., 1996; Krause et al., 1998). Antigenic Variation in B.bovis B. bovis has a very com plicated life cycle, which requires two hosts: ticks and mammalian hosts. To invade and proliferate asexually with in mammalian erythrocytes is crucial for the completion of life cycle. In spit e of the development of parasite-resistant host immunity, those who survive acute infection often have persistent and relapsing in fection established within their bodies, which features extremely l ong duration and fluctuating parasitemiae by several orders of magnitude (Calder et al., 1996; Allred et al., 2000). Both structural and morphological modifications are made on the IRBC surface to facilitate parasite survival within eryt hrocytes and expansion in the host. One important morphologic change is the development of short, ridge-like structures on the IRBC surface, which are similar

PAGE 16

16 to the knobs that are formed on human erythrocytes infected with another related Apicomplexan hemoparasite, Plasmodium falciprum (OConnor et al., 1997) The rapid antigen ic variation of VESA1 protein on the IRBC surface and cytoad hesion of IRBCs to the capillary and postcapillary venous endothelium are pivotal for succe ssful parasite long-term persistence in hosts. Antigenic variation may contribute to parasite pe rsistence in at least tw o aspects: host immune antibodies responding to the prior parasite popula tions fail to effectively recognize and thereby eliminate the emerging parasite populations occurring temporally later during the life-long infection; and, structural variability of the cytoadhesion-associated VESA1 also permit the IRBCs to cytoadhere to variable host cell li gands. VESA1 protein possesses dual roles in mediating both antigenic variation and cytoa dhesion (OConnor and A llred, 2000), and can be looked upon as a bifunctional virulence factor, analogous to the variant erythrocyte membrane protein 1 (PfEMP1) family in the malarial parasite, P. falciparum (Baruch et al., 1997; Scherf et al., 1998) Therefore, it would be of great value to investigate the molecular mechanism of antigenic variation in VESA1 protein, which would be very help ful in seeking novel strategy to reduce parasite persistence in the mammalian host preventin g the prevalence of bovine babesiosis. VESA1 Protein and C9.1 ves 1 / LAT VESA1 is a size-polym orphic, parasite-derived, heterodimeric protein that is trafficked to and expressed on the IRBC surface. VESA1 protein is a key parasite-derived virulence factor, playing crucial roles in antigen ic variation, cytoadhe rence and subsequent sequestion of IRBCs (Allred et al., 1993; Allred et al., 1994). Previous studies on isogenic clonal parasite lines indicated that the two VESA1 s ubunits, 1a and 1b, differ in both sizes and antigenicity. Similar amino acid composition and some common proteolyti c digestion fragments shared between them suggested genetic relatedness (OC onnor et al., 1997). Currently, available evidence indicates

PAGE 17

17 that these two polypeptides are encoded by two distinct genes, ves 1 and ves 1 respectively, belonging to two divergent but cl osely related branches of the ves multigene family (Allred et al., 2000; Al-Khedery and Allred, 2006; Xia o, Y and Allred, DR; unpublished data). The ves 1 gene encoding the VESA1a subunit belongs to a large polymorphic multigene family, ves Most members of this gene family seem to be invariant over time among several isogenic parasite lines (Allred et al., 2000). The LAT has been demonstrated and validated on the B. bovis chromosome 1 in the C9.1 cl onal line (Al-Khedery and Allr ed, 2006). In the C9.1 LAT, an actively transcribed ves 1 gene is juxtaposed divergently with an actively transcribed C9.1 ves 1 gene. Furthermore, these two head-to-head a pposing active genes share the 5 untranslated region (5 UTR) upstream of their respective transcription start c odons. This intergenic region is highly conserved among ves genes and averages approximate ly 433 bp long. It features two highly conserved pairs of quasi-palindromic sequence segments. This unique gene organization pattern at the LAT also exists many non-expressed ves gene loci (Al-khede ry and Allred, 2006; Brayton et al., 2007). However, t hose loci have variable combinations of genes apposing each other, as ves 1 ves 1 or ves 1 ves 1 Some of the genes are actually pseudogenes, although most appear not. It was shown th at, in the Texas T2Bo strain 24 ves gene loci have the same gene organization and a length simila r to the LAT (Brayton et al., 2007). Segmental gene conversion involving many ves sequence donor genes has been demonstrated to be one important m echanism of antigenic variation in B. bovis Sequence patches can be donated from ves donor loci to the LAT, which can cause the LAT to become variable and mosaic (Al-khedery and Allred, 2006). Ho wever, it is poorly understood how segmental gene conversion occurs at the LAT. The IGrs of the C9.1 LAT and many ves loci harbor two pairs of highly c onserved palindromic segments (Al-Khedery and Allred, 2006).

PAGE 18

18 Moreover, some preliminary evidence shows that the IGrs in the LAT and some non-expressed ves loci possess potent promoter activity (Wang, X and Allred DR; unpublished data). The entire IGr is essentially symmetrical regarding GC content and conser ved sequence patches. Based on this information, I hypothesized that different levels of higher order DNA structure within the LAT, especially the intergenic region, may exist in vivo distinguishing the LAT from other nonexpressed ves gene loci. It is inferred from its sequen ce characteristics that the LAT intergenic region has the potential to form extensive quasi-cruciform structure (Mfold algorithm, Genetics Computer Group). This type of uns table structure, once formed, may play an important role in transcriptional activation as we ll as in interand intra-chromo somal gene conversion events (Kim et al., 1998; Lisnic et al., 2009). Thus, it w ould be very informative to examine whether the stable higher-order structure can be detected at the LAT in vivo. Meanwhile, little is known about the chromatin structural characterisitics of ves genes, which could be very important in explaining their differential transcriptional status. Inverted Repeats and Their Biological Importance Many prokaryotic and eukaryotic genom es, including the human genome, contain widespread inverted repeats (IRs). Inverted repeats have the potential to form cruciform structures. They occur more often in genomes than do random sequences, and are predominantly distributed at or near genetic regulatory regions or other functionally important locations including replication origins, operator sites, pr omoter regions, hormone responsive elements, hot spots of chromosomal translocation and mitotic in terchromosomal recombination etc (Gellert et al., 1983; Tenen et al., 1985; Gordenin et al., 1993; Kato et al., 2003; Kurahashi et al., 2004). In some bacteria, cruciform DNA extrusion is a pre-requisite for initiating replication and transcription (Kato et al., 2003). Coliphage N4 virion-en capsidated, DNA-dependent RNA polymerase (vRNAP) became deactivated when mutations were introduced to impair the

PAGE 19

19 secondary structure of a conserved inverted repeat in the promoter region it recognizes (Glucksmann et al., 1992). These observations indicate that inverted repeats and/or the unusual DNA conformations they may form in vivo may serve important biological functions (Wells, 1988). Cruciform structure is thought to occur in genomic regions containing palindromic sequence or inverted repeats, fo rming two hairpin arms of equal length and a central four-way junction. It can be formed in vitro and has also been suggest ed to exist in living cells (Vologodskaia and Vologodskii, 1999; Wells, 1988; Tenen et al., 1985). DNA cruciforms are highly dynamic and exhibit a high degree of variability in the interarm angle. Inverted repeats could have many variable conformations includi ng B-form duplex, planar cruciform, X-type cruciform and stem-loop-loop-stem relying on nucleotide sequence homology and environmental requirements (Kato et al., 2003). These various conformations of IRs may be responsible for executing diverse biological functions under differe nt situations. The mol ecular basis for IRs importance could be established through creating special platforms for di fferent types of DNA binding proteins to help stabiliz e structure and/or recruit more functional protein factors. For example, one study suggested that cr uciforms could also be the target site for some activator or repressor proteins (Kim et al., 1997). Interestingly, cruciform stru cture is both stereochemically equivalent and topologically similar to the Hollid ay junction, the central intermediate in genetic recombination (Lilley and Kemper, 1984; Iwasaki et al., 1992; Zhang et al ., 1999). However, it is still unknown if they may have functional overl ap or if cruciforms may also play a recombination intermediate role in some organisms. Generally, IRs of variable length are thought to be one of the important sources of prokaryotic and eukaryotic genom ic instability (Gordenin et al., 1993; Collick et al., 1996).

PAGE 20

20 DNA IRs can become hot spots for DNA double strand breaks (DSB), which can create dangerous DNA lesions demanding error-free sensing and repairi ng (Kurahashi et al., 2004; Tanaka et al., 2009). Modified or unmodified IR s can be associated with many human diseases, such as Hereditary Angioneurotic Edema, Du chenne Muscular Dystrophy etc (Bissler, 1998). Additionally, because of their notable roles in inducing palindrome formation and gene amplification, short IRs can also be targeted in cancer therapy to prevent further genomic instability which normally associates with ma lignant tumor progression (Tanaka et al., 2002; Tanaka and Yao, 2009). Recent observations in plan ts also revealed a co rrelation between IRs and epigenetic modifications and their roles in gene regulation were then recognized. IRs can signal for gene silencing thr ough DNA methylation (Wolffe and Matzke, 1999). Methylated IRs could act in trans by possible sequence pairing to induce methylation, thereby transcriptionally silencing remote homologous sequences by indu cing chromatin-remodeling events via methylDNA-binding proteins (Wolffe and Ma tzke, 1999; Ebbs et al., 2005). Therefore, considering the biological importanc e of IRs and their potential involvement in transcriptional regulation and ge ne recombination, it would provide great value to the research community to focus a study on the ves intergenic regions, regardi ng its unusual location and sequence specificity. DNA Two Dimensional Gel Electrophoresis and Its Application The DNA 2D gel electrophoresis system wa s first developed in 1980 (Sundin and Varshavsky, 1980). Two dimensional (2-D) gel elect rophoresis can be used to efficiently and directly detect different forms of DNA. As early as 1983, Bell and By ers developed a DNA 2-D gel system to resolve branched DNA from linear DNA. This system has been modified thereafter to adapt different applicati ons. DNA 2-D gel system has now become a well-established, powerful tool commonly used in the study of re plication or recombina tion intermediates in

PAGE 21

21 eukaryotes (Bell and Byers, 1983; Arcangioli, 19 98; Allers and Lichten, 2000). It has not been demonstrated yet if DNA 2-D gel system is capable to separate cruciform although cruciform is stereochemically similar to Holliday junction mol ecules, which can be well separated on a 2-D gel (Iwasaki et al., 1992; Zhang et al., 1999; Allers and Lichten, 2000). The mobility of branched DNA molecules was found dependent upon agarose concentration and voltage (Bell and Byers, 1983) Basically, the first di mension of horizontal electrophoresis is run at low volta ge in a low concentration (0.5 ~0.8%) of agarose gel separating DNA molecules mainly in proportion to their mass. The second dimension is then run at higher voltage in a higher concentration (1.2~ 1.8%) of agarose gel with the presence of ethidium bromide (EtBr), which makes structured DNA more rigid and reduces branch migration (Bell and Byers, 1983). Therefore, DNA shape becomes the main factor affecting mobility in the second dimensional of gel electrophoresis. After electrophoresis, the 2-D gel is normally blotted onto nylon membrane for later hybridiza tions with locus-specific probes. In this project, I tried to establish a DNA 2-D gel system capable of separating cruciform from linear DNA for the purpose of detecting any in vivo LAT-associated cruciform-like DNA structure related to active transcription. Chromatin Structure and Gene Transcription In eukaryotic cells, nuclear DNA is packaged compactly into well-organized chrom atin composed of periodic arrays of nucleoprotein comp lexes, with the nucleosomes as the basic unit (Lusser and Kadonaga, 2003). Transcription in euka ryotic cells is stringe ntly controlled and regulated in an orderly fashion by numerous protein factors. The ge ne transcription status closely correlates with the permissive or non-permissive state of their chromatin structures for the entry of those regulators (Nalikar et al., 2002). To ma intain a state of being actively transcribed, the gene coding regions must be essentially made accessible to the enzymatic machinery of

PAGE 22

22 transcription through a series of chromatin-remodeling events. Cons equently, active genes can be distinguished from inactive gene s by differential chromatin stru ctures (reviewed in Rando and Chang, 2009). As an important characteristic of chromatin configuration, the nucleosome positioning pattern and orientation at gene promoters and coding regions can regulate transcriptional activities in many ways (Vega-Palas and Ferl, 1995; Svaren and Horz, 1996; Chen and Yang, 2001; Rando and Chang, 2009). Nucleosomal Organization on Genes Eukaryotic genom ic DNA is orga nized at its simplest level within 10 nm chromatin fibers, which feature phased arrays of nucleosomes (reviewed in Ramakrishnan, 1997). However, nucleosomes are not just randomly positioned onto genes. Preferential binding of nucleosomes on specific DNA regions has essential impact on th e regulations of gene activities such as DNA replication and repair, gene transcription and gene recombination (Pazin et al., 1997; Field et al., 2008). During those biological processes, nucleos omes have to frequently disassemble or reassemble on naked DNA to help achieve certain functions. The occupancy and rearrangement (or depletion) of nucleosomes from some regulatory DNA elements can regulate gene activities by preventing or allowing the targe ting of DNA-binding pr otein factors. Several aspects to describe nucleosomal orga nization on genes include nucleosome repeat length (NRL) or spacing, nucle osomal positioning, nucleosome-free regions and the composition of histones or histone varian ts in nucleosomes (Jin et al., 2009; Rando and Chang, 2009). Normally active genes adopt regular nucleosomal positioning patterns due to the frequency of transcriptional activities to expose some DNA regulatory elem ents which facilitates the transcriptional machinery and other regulator y protein factors (Chen and Yang, 2001). Despite different transcription status, inact ive genes or silent genes contai ned in heterochromatin can also have regular nucleosomal arrays (Sun et al., 2001; Rando and Chang, 2009). However, the

PAGE 23

23 positioning of nucleosomes on silent genes se em mostly rely on the underlying DNA sequence cues, while nucleosome locations and occupancy on active genes seem not follow sequence cues and may need ATP-dependent chromatin remode lers to actively deploy nucleosome binding (reviewed in Rando and Chang, 2009). Activation or re pression of genes can be associated with dynamic nucleosomal repositioning around gene pr omoter-proximal region, probably as a consequence of disposition of histone variants or epigenetic modification of gene sequences (reviewed in Schones et al., 2008) Meanwhile, the presence of some DNA-binding factors and ATP are also essential for aligning and main taining nucleosome-positioning patterns (Pazin et al., 1997; Corona et al., 2000). Moreover, switch of transcriptional status seem concomitant with the crucial changes in nucleosome-depleted region s due to the loss or occu pancy of transcriptionassociated protein factors (Kumar i et al., 1997; Sun et al., 2001). As one of the important struct ural characteristics of chroma tin structures, NRL reflects the compactness of chromatin, which is dynamic thr oughout the cell cycle to cope with different cellular activities (Misteli et al., 2000). Gene NRLs closely correla te with their transcriptional status, where active genes and inactive genes can differ significantly in their NRLs, and maintain the difference throughout cell cycles (De Ambrosis et al., 1987). Meanwhile, the NRL is not a static characteristic for gene chromatin struct ure; on the contrary, it can also be dynamic to respond to environmental changes (Castro, 1983). Different organi sms may feature different bulk NRLs with variable DNA linker regions. Some lowe r eukaryotes tend to ha ve shorter NRLs and may not possess a canonical histon e H1, but rather a small H1-lik e protein, which may mean that they have unique chromatin structural characte ristics different from hi gher eukaryotes (Thomas and Furber, 1976; Greaves and Borst, 1987; Lanz er et al., 1994; Mas on and Mellor, 1997; Hsiung and Kucherlapati, 1980).

PAGE 24

24 The dynamic organization of nucleosomes can serv e as inducible or re pressible signals and profoundly affect transcriptional behaviors (Pazin et al., 1997; Fi eld et al., 2008; Schones et al., 2008). The nucleosomes at the promoter region are particularly important in regulating transcription (Tirosh and Barkai, 2008; Ra ndo and Chang, 2009). Actively-transcribed genes usually have lower nucleosome occupancy levels at the promoter-proximal regions to allow for more frequent transcription initiation (Lee et al., 2004). Therefore, active genes usually show higher sensitivity to the digestion of variable nuc leases such as MNase or DNase I at the gene regions which are particularly nucleosome-free (Goriely et al., 2003). It has been suggested that actively expressed genes tend to have a transl ationally-positio ned nucleosome pattern at the promoter, while non-expressed genes do not (revi ewed in Schones et al., 2004). Therefore, nucleosomal organization pattern s can be probed by nucleases. Nucleases and Their Applications in Studying Chromatin Structure Nucleases w hich can cut the phosphodiester bonds between nucleic acids have wide applications in revealing in vivo chromatin configuration. A variet y of nucleases with different template specificities are available to provide different types of information on chromatin organization (Larsen and Wein traub, 1982; Drinkwater et al ., 1987; Gregory et al., 2001). Nuclease-hypersensitive sites are importantly indi cative of gene regulatory regions and therefore nucleases such as Micrococcal nuclease (MNase ) and DNase I are normally used to identify cis acting regulatory elements (Levy-Wilson et al., 1988; Kumari et al., 199 7; Sun et al., 2001). MNase is a relatively non-specific endoexonuclease enzyme which can cleave doublestranded nucleic acids but has a higher activity on single-stranded nucleic acids, preferentially at AT-rich regions (Telford and Stewart, 1989). It preferentially makes double-stranded cuts on chromatin substrate within the nucleos ome-free linker region occurring between nucleosomes, and can generate single-stranded nicks on the DNA wrapped on nucleosomes (Zaret et al., 2005).

PAGE 25

25 With partial MNase digestion, a series of nuc leosomal arrays of va rious lengths can be generated, which can form a regular ladder of DNA bands on agarose gel ranging from a mononucleosome to n-mer nucleosomes. These arrays are usually used to estimate average NRL or to detect the alteration of nucleosome positi oning pattern on specific genes with regard to gene activation or inactivation (G reaves and Borst, 1987; Anello et al., 1986; Mason and Mellor, 1997; Zhang and Reese, 2006). Additionally, MNase can be used to determine how a gene region of interest is assembled in nucleosomes, with the help of southern blot analysis combined with direct/ indirect end labl eling (Chen and Yang, 2001; Fryer and Archer, 2001). DNase I is an endonuclease that efficiently digests all free forms of single-stranded DNA or double-stranded DNA to short oligonucleotides or mononucleotides (Fox, 1997). It is sensitive to the structure of the AT-rich DNA minor gr oove facing away from the wrapped histone octamer, and generates a characteristic cleavag e pattern with ~ 10 bp in tervals (Chen and Yang, 2001). Due to their relatively open chromatin stru ctures, adapted in support of transcriptional activity, active genes are hypersensitive to DNase I digestion, especially th eir promoter-proximal region. Conversely, inactive genes normally are co mpactly packaged and relatively inaccessible for enzymatic digestion, and hen ce typically are insens itive to DNase I. Since DNase I does not work on DNA bound with regulatory pr oteins, it can also be used to do footprinting assays to map the location of gene regulatory elements and study protein-DNA interactions (Brenowitz et al., 2001). S1 nuclease is an endonuclease which uses single-stranded DNA or RNA as preferred substrates. It cleaves 5' and 3' overhanging single-stranded DNA and hairpin loops, and can sometimes produce single-stranded breaks to nick double-stranded DNA or RNA, or DNA-RNA hybrids (Wiegand et al., 1975). As a tool, it can be used to produce blunt ends for cloning work.

PAGE 26

26 It is widely used to do structure mapping on RNAs or DNA: RNA hybrids (Reyes and Wallace, 1987). Historically, its applica tion also included mapping intr ons and transcription termini (Harty et al., 1989). Importantl y, since it is very active at cleaving single-stranded DNA, it can be used to find transcriptional un its in genomes or to confirm tran scriptional status of specific genes, by detecting DNA bubbles which form dur ing transcription ini tiation or elongation. Mung bean nuclease has template specificity si milar to that of S1 nuclease, but nicks DNA duplexes more rarely than does S1 nuclease (Gub ler, 1987). Interestingly, it has been reported that when under carefully controlled conditions Mung b ean nuclease can cut naked Plasmodium genomic DNA, identifying the boundaries of transc riptional units and/or exons (McCutchan et al., 1984). Despite the reported ability to cut double-strand ed DNA when used at high concentrations and with long incubation times, th e predominant template of mung bean nuclease is single-stranded DNA (Sharp a nd Slater, 1993). Therefore, Mung bean nuclease can be a useful tool to help detect single-str anded regions in chromatin (Grom ova et al., 1995). In this study, I chose to use those four different types of nucleases to probe chromatin structures of ves genes to explore the transcriptional regulatory mechanism of this gene family. Hypothesis Based on th e differential transc ription status and the simila rity of gene organizations between the LAT and other non-expressed ves loci, the examination of high order structure of ves genes may provide useful information to eluc idate their control mode of mutually exclusive transcription. Our hypothesis is that differential chromatin structures may exist between the C9.1 ves1 / LAT and those non-expressed ves genes loci regarding thei r differential transcription status. Two different approaches for the detect ion of high order DNA/chromatin structure were tried at different structur al levels respectively desc ribed in chapter 2 and 3.

PAGE 27

27 CHAPTER 2 HIGH ORDER DNA STRUCTURE WITHIN B ABESIA BOVIS LOCUS OF ACTIVE VES TRANSCRIPTION Abstract Babesia bovis is an intraerythrocyte protozoan w hich can cause severe disease in cattle. It is poorly understood how ves polymorphic family, encoding the cr itical virulence factor VESA1, achieves mutually exclusive tr anscription. The highly conserve d intergenic regions among the head-to-head ves gene loci feature two pairs of hi ghly-conserved imperfect palindromic segments. The unique quasi-palindromic pr omoter region suggested the likelihood of involvement of higher-order DNA st ructure in the active transcription of the LAT. To explore the possible mechanism of transcriptional control over ves genes, the detection of in vivo LATassociated DNA structure in chromatin was carried out through a combination of DNA 2Dimensional gel (2-D) electrophore sis either with structure-prot ective gDNA extraction or with structure stablization by psoralen crosslinking, followed by hybrid izations with gene-specific oligonucleotide probes. Some offlinear arc signals were found and could be certain higher-order DNA structures in chromatin mos tly likely associated with the LAT. However, limited signal intensity and low repeatability re strained the significan ce and correct intrepre tation. Therefore, it is suggestive that high order DNA structure in chromatin may exist in vivo to assist active transcription of the LAT although further evidence would be esse ntial to substantially support this finding and correlate it with th e C9.1 LAT and its transcription. Introduction B. bovis is a hem oparasite which shares many trai ts with the human malarial parasite, P. falciparum VESA1 membrane protein, which is encoded by a multigene family named ves and is composed of two subunits, VESA1a and VESA 1b, is thought to be im portant for long-term persistence of the parasite in the host via its involvement in antigenic variation and IRBC

PAGE 28

28 cytoadherence (Allred et al., 1994; O Connor et al., 1997; O Connor and Allred, 2000). It is apparent that the ves multigene family is under mutually excl usive transcriptional control (Lau et al, 2007; Zupanska et al., 2009). However, the mechanism(s) by which transcriptional control over the ves multigene family is achieved is poorly understood. The LAT, has two active ves genes ( ves 1 and ves 1 ) apposing in a head-to-head orient ation, sharing a long 5 UTR region (Al-Khedery and Allred, 2006). It has been found in the genome of Texas T2Bo strain that ~ 24 ves gene loci share similar gene organization and lengths with the active LAT and yet remain inactive (Brayton et al., 2007). The intergenic regions (IGrs) among ves gene loci attract great attention in this study due to their promoter activity (Wang, X and Allred, DR; unpublished data) and their characteristic overall quasi-palindromic sequences. Inve rted repeats, such as those found in each side of the LAT IGrs, are known to be enriched in gene regulatory regions such as promoters and can play important roles in ge ne transcription, replication and recombination (Gellert et al., 1983; Tenen et al., 1985; Gordenin et al., 1993; Kato et al ., 2003; Kurahashi et al., 2004). Importantly, some indirect evidences suggested the regulat ory functions of palindromic sequences were due to their pote ntial to form cruciform structures (Kim et al., 1998; Kurahashi et al., 2004; Inagaki et al., 2009).Therefore, it was hypothesized that the LAT intergenic region may form higher order DNA structure to facilitate active transcription or possibly to act as an entryway for strand invasion during segmental gene conversion, which has been demonstrated to be an important mechanism in B. bovis antigenic variation (Al-Kh edery and Allred, 2006). To examine if there is in vivo higher order DNA structure ex isting in the C9.1 LAT IGr to maintain its active tran scription status, I took the approach of DNA 2-D gel electrophoresis to detect unusual structure associated with the LAT. DNA structure-protective extraction or psoralen-crosslinking both were tried to stabilize in vivo higher order struct ure prior to 2-D gel

PAGE 29

29 separation (Bell and Byers, 1983; Aller and Lichten, 2001). Some off-arc signals were observed on two of the 2-D gels and the significance was discussed. Methods and Materials Parasite Culture The B. bovis C9.1 clonal line was clo ned from its progenitor MO7 by limiting dilution (OConnor et al. 1997). Parasites were maintained unde r microaerophilous stationary-phase conditions as described (Levy and Ristic, 1980 ). Briefly, the culture was incubated at 37oC under 5% CO2, 5% O2, and 90% N2. 10% packed cell volume fresh bovi ne erythrocyte suspension was used to feed parasites. The cu lture medium contained 40% (v/v) of fresh normal bovine serum in Medium 199 supplemented with 20 mM TES buffer (N-[tris(hydroxymethyl)methyl]-2aminoethanesulfonic acid) and 1 antibiotics (antibiotic-antimycotic, Gibco BRL) to prevent contamination. When needed, the percent parasitized erythrocytes (ppe ) was boosted by dilution enrichment (OConnor et al., 1997), which is a method of reduc ing the packed cell volume from 10% to 0.625% in steps to obtain a ppe hi gher than 10%. In some experiments, bovine leucocytes were removed from parasite cu lture by passage through sterile Whatman CF-11 cellulose column (Ambrosio et al ., 1986). Rupture of Infected Red Blood Cells (IRBCs) by Saponin Lysis The C9.1infected red blood cells were r uptured by 0.05% saponin (i n phosphate-buffered saline, PBS, pH 7.4) lysis to reduce hem oglobin contamination. After lysi s of erythrocytes was essentially complete, parasites were collected by high speed centrif uge (16000 g, 10min at 4oC; Beckman). It was shown by vital fluorescen t staining with 6-CF DA (6-carboxyfluorescein diacetate; 0.5 g/ml) that the parasites reta ined this dye and appreare d viable following exposure to 0.05% saponin.

PAGE 30

30 Construction of Artifici al DNA Cr uciform Molecule PCR amplifications (TripleMaster@ PCR sy stem, Eppendorf) were carried out using primer pairs DA44 (forward) / PIRR-PinAI (reve rse), and PIRF-PinAI (forward)/ PIRR-BamHI (reverse) respectively with the C9.1 ves 1 cDNA (P9.6.2) serving as DNA template. The sequences of these primers are listed in Tabl e 2-1. The two specific DNA products were digested with PinAI, and ligated unidirectionally. The lig ation product was then cut by BamHI and cloned into the BamHI cloning site of the dephosporylated pBluescript II KS (+) vector (Invitrogen) to generate the plasmid pL3. The lig ation product was transformed into E. coli host strain DH5 (Invitrogen) to do blue/white selection (at 37oC) on X-gal agar plates White colonies were picked, and restriction enzyme analysis was done to confirm the correct insertion of desired DNA fragment. The inserted DNA fragment in pL3 has been sequenced and confirmed to have an internal 46-bp perfect inverted repeat. The sequences of the primers used in cloning were listed (Table 3-2). To retrieve the cruciform control molecule BssHII digestion was done on plasmid pL3 and generating a DNA fragment of 5 32 bp (B0.5). After gel purification (QIAquick Gel Extraction Kit, Qiagen) the fragment was di ssolved in TE buffer (10 mM Tris, 1 mM EDTA, pH 8.0) for later use. The primers used were listed (Table 2-1). In vitro DNA Thermal Remodeling Assay To m ake DNA molecules possess ing of cruciform potential to form cruciform structure based on sequence specificity, a temperature program to induce DNA thermal remodeling was adapted from a protocol used to construct s ynthetic Holliday juncti on (Constantinou and West, 2004). We chose to work with natural sequences to attempt to induce thermal remodeling when the molecules had many options for their final conformation. We called this procedure thermal remodeling (TR). Firstly, purified DNA in TE buffer was heated at 80.1oC for 8 min, then was gradually cooled in steps to 65oC, 37oC, and finally 25oC, and finally ending at 4oC, with 10-min

PAGE 31

31 incubation for each temperature stage. This specific DNA thermal remodeling program and conditions (melting temperature, Mg2+, NaCl) were optimized to give maximal output of novel structured form, relative to normal linear duplex, but without induc ing complete strand separation. DNA 2-D Gel Electrophoresis System A neutral / neutral DNA 2-D gel electrophoresis system was used to separate structured DNA molecules from their original linear forms. In order to distinguish DNA linear forms from structured forms, a smooth arc composed of linear DNA molecules of variable sizes was necessary on the 2-D gel to serve as linea r DNA control. The linear form of tested DNA molecules would fall on this arc whereas struct ured DNA forms were expected to migrate offarc. One hundred l (~200 ng) of kb plus DNA ladder (Invitrogen) was added to tested samples when being loaded on the first dimension of 2-D gel to form the linear DNA control arc. This amount of DNA ladder in samples also ga ve nice hybridization signals when hybridized with [32P]-ATP end labeled 1kb plus DNA ladder; th erefore, it would reveal the linear DNA control arc on autoradiographs of hybridized blots. To separate structured DNA molecules which had been thermally remodeled in vitro the first dimension of 2-D electrophoresis was perf ormed in a 0.75% agarose gel in 1XTBE buffer (0.089 M Tris-borate, 0.089 M Boric Acid, 0.002 M EDTA) with 3 mM Mg2+ at 1 volts cm-1 22 hours at room temperature. The sample lane was cut with a clean, sharp blade, stained with 0.5 g/ml EtBr and then photographed under UV. The sa mple lane was then placed horizontally on top of the gel tray, 1.5% agaros e gel was poured till it barely c overed the edge of the first dimensional sample lane. The second dimension of electrophoresis was run in 1X TBE buffer ( 3 mM Mg2+) supplemented with 0.5 g ml-1 EtBr to make DNA struct ures more rigid (Bell and

PAGE 32

32 Byers, 1983) at 3 volts cm-1 for 22 hours. The gel was photographed and exposed under UV for 5 min to nick the DNA facilitating transfer onto nylon membrane. In order to examine in vivo structures of C9.1 genomic DNA extracted in a structureprotective way, the DNA 2-D gel electrophoresis was done at 4oC to facilitate the stabilization of any DNA structures possibily present. The firs t dimension (0.6% agarose gel) was run at 1 volt/cm for 62 hrs in 1XTBE supplemented with 3 mM Mg2+ and the second dimension (1.2% agarose gel) was run at 3 volt/cm, 22 hr s in 1XTBE supplemented with 3 mM Mg2+ and 0.5 g/ml EtBr. To guarantee maximal protection of possible DNA structures, the 5X DNA loading buffer was also modified (25 mM Tris.HCl pH 7.5, 0.25 mM EDTA pH 8, 25 mM MgCl2, 0.25% bromophenol blue, 0.25% xylene cyanol FF, 15% Fi coll 400; Native/Native 2-D gel protocol provided by Dr. Michael Lichten from National Institutes of Health) The gel was exposed to UV light to fragment large DNA molecules to facilitate their transfer. Alkaline transfer of the DNA onto nylon membranes for late r southern blot analysis was followed by washes in a series of EDTA solutions to remove Mg2+ to allow DNA transfer. The gel was washed two times in 10 mM EDTA (pH 8.0), two times in 5 mM EDTA (pH8.0), then once with distilled water, 15 min per wash. Th e DNA was denatured by incubation of the gel in denaturation solution (1.5 M NaCl, 0.5 N NaOH) for 30 min prior to transfer onto Hybond XL nylon (GE healthcare) membrane in transferri ng buffer (1.5 M NaCl, 0.5 N NaOH) by classic capillary blotting. After tr ansfer was completed, the membrane was washed 3 times 5 min each in 2 SSC. The transferred DNA was fixed ont o the nylon membrane by UV crosslinking (Stratalinker 2400 UV Crosslinker; Stratagene). The membrane was kept wet and sealed in plastic bag and frozen at -20oC until it was used.

PAGE 33

33 DNA Electron Microscopy The 200-m esh cooper grids were cleaned, in order, with acetone, ethanol, and then deionized water in preparation for formvar coating. The formvar coating procedure was performed as described (Hayat, 2000). Selected grids were then thinly carbon-coated by the Electron Microscopy Center at the University of Florida. After thin car bon-coating, the formvar film-coated grids were stored at room temperature until use. The spreading of thermal remodeled DNA samp les onto carbon-coated formvar film grids was done by the Kleinschmidt procedure of DNA spreading (Kleinschmidt, 1968), with modifications. All steps were ex ecuted in a dust-free environment. A cleaned Petri dish (10 mm diameter), was filled with 20 25 ml hypophase solution ( 10 mM Tris, 1 mM EDTA, 14.8% formamide; pH 8.5; 0.45 m filter-sterilized). A cleaned glass slide was slanted by 30o angle against one edge of the dish. Ninety-five l of hyperphase solution (103 mM Tris, 10.3 mM EDTA, 54.5% formamide, 0.2 mg/ml cytochrome c, pH 8.5; 0.45 m filter-sterilized) and 5 l of DNA sample solution were mixed together well. The DNA mixture was dispersed on one side of the glass slide close to the interface between hypophase solution and slide. The carbon-coated formvar film grids were used to pick up a singl e layer of spread DNA on the spreading surface at the interface between hypophase and hyperphase. Th e grids were stained 30 seconds (s) in 1% uranyl acetate solution, then washed for 10 s in 90% ethanol and airdried. The grids were subjected to low angle rotary shadowing with Pt / Pd to enhance struct ural details. Rotary shadowing was carried out by University of Flor ida, Interdisciplinary Center for Biotechnology Research, Electron Micr oscopy Laboratroy (ICBR).

PAGE 34

34 Detection of In vivo LAT-Associated DNA Structure w ith Structure-Protective Extraction To examine the in vivo genomic DNA structures, the 2-D gel system needs to have controls for both linear and st ructured DNA forms. The 1kb plus DNA ladder served as a linear DNA control arc on 2-D gels and thermally remode led constructed cruciform control, described previously, was used as a struct ured DNA control. Specific probes we re generated to enable their detection on 2-D gel southern blots. The probe specific to the cruciform control (5GCGCGCGTAATACGACTCACTATA-3) was name d HP probe, and was end-labeled by [32P]-ATP (Southern, 2006). This probe recogni zes both major forms of cruciform control molecule B0.5 (linear form and structured forms after TR). The structure-protective extracti on protocol was adapted from that described by Allers and Lichten (Allers and Lichten, 2000). Briefly, B. bovis C9.l line cells at about 10% ppe were harvested and washed once with 1 Vega y Martinez solution (VYMs) by centrifuging at 16000 g /10 min /4 oC. Cell lysis in 0.05% saponin was done to reduce the packed cell volume to facilitate later handling. The cell pellet was then lysed in 300 l pre-warmed freshly-made extraction buffer (20 mM Co3+(NH3)6Cl 3, 2 M NaCl, 100 mM TrisHCl pH 7.5, 25 mM EDTA, 2% (w/v) polyvinyl pyrrolidone 40, and, finally, 3% (w/v) CT AB). With the addition of mercaptoethanol to 0.16% (v/v), proteinase K (500 g/ml) and RNase A (20 g/ml), the lysate was incubated at 37oC for one hour, then extracted with 200 l chloroform: isoamyl alcohol (24:1 v/v) by vigorous vort exing and centrifugation (16000 g 3 min, 4oC). The aqueous phase was transferred to a 5 ml round-bottomed tube, and 1.5 ml fresh dilution buffer (1% (w/v) CTAB, 50 mM TrisHCl pH 7.5, 10 mM EDTA, 4 mM Co3+(NH3)6Cl 3) was added, followed by 5 gentle inversions. After incuba tion for 10 min at room temperat ure and a further 20 inversions, gDNA spontaneously precipitated as a blob in the solution. The gDNA precipitate was carefully

PAGE 35

35 transferred to a new tube and all subsequent steps were carried out on ice. The wash-andprecipitation step was repeated several times e ssentially as described (Allers and Litchen, 2000) to replace Co(NH3)6Cl3 with Mg2+. Finally, purified gDNA was dissolved in cold TMSpe buffer (10 mM TrisHCl pH 7.5, 2 mM MgCl2, 50 M spermidine). The concentration of the isolated gDNA with protected in vivo structure was determinated in pr eparation for restriction enzyme digestion. Detection of In vivo LAT-Associated DNA Structure w ith Psoralen Crosslinking To stabilize in vivo DNA structure, 4 -Aminomethyl-4, 5 8-trimethylpsoralen hydrochloride (Sigma-aldrich; Catalog No. A4330) was used to crosslink DNA prior to DNA extraction and gel running. The pr ocedure was basically as descri bed (Bell and Byers, 1983) with some modifications. Briefly, 10 ml parasite cu lture at about 10% hematocrit and 10% ppe was harvested and washed once with cold 1XVYMs, then treated with 0.05% saponin in PBS as described above. The parasite pe llet was then suspended in 0. 1 ml cold pre-crosslinking cell suspension buffer (0.7 M sorbitol, 50 mM Tris-H Cl, 50 mM EDTA, pH 8.0) and incubated at 4oC for at least 15 min in the dark before crosslinking. To do in vivo DNA crosslinking, the parasite suspension was slowly added in a 35mm petri dish (Falcon) containing 0.9 ml of crosslinking solution (0.12 mg/ml 4 -Aminomethyl-4, 5 8-trimethylpsoralen hydrochloride, 50 mM Tris-HCl, 50 mM EDTA, pH 8.0) ; then the mixture in petri dish was placed on a benchtop 365-nm UV lamp (Blak-Ray, Model C50) at 4oC. After 15-min crosslinking under UV, parasites were lysed by the addition of 1 ml of 10% SDS to the crosslinked cell suspen sion. Then parasite gDNA was purified by phenol extracti on and isopropanol precipitati on as described (Tripp et al., 1989), digested by restriction enzymes and resolv ed by 2-D gel electrophoresis, as described earlier. To evaluate the crossli nking procedure, thermally-remodeled ves locus 15-1 DNA fragment (1.8 kb) or cruciform control molecule B0.5 was incorporated into samples as internal

PAGE 36

36 controls prior to psoralen crosslinking. The 2-D gel was photographed and then irradiated under short-wavelength UV light (120 mJ/side, UV Stra talinker, Stratagene) to reverse psoralen crosslinking. The gel was then transferred onto nylon membrane through alkaline transfer for hybridizations with specific probes (Sanz et al., 2007). Southern Blot Analysis C9.1-IRBCs were ruptured by 0.05% saponin in PB S and the parasite pellet was lysed in TE / 1% SDS. The lysates were incubated at 37oC in the presence of 100 g/ml proteinase K and 40 g/ml RNase A for 5 hr. Genomic DNA was ex tracted by phenol-chloroform procedure, precipitated by isopropanol and then finally dissolved in TE buffer (Tripp et al., 1989). About 2 g gDNA was digested by chos en restriction enzymes at 37oC for 3~5 hr, and an additional 0.5 unit of fresh restriction enzyme was added into the reactions to incuba te another 2 hr at 37oC to ensure complete digestion. Digested DNA was electrophoresed on a 0.8 % agarose gel, and the gel was later exposed to UV light for 5 minutes to nick DNA. Before blotting, the gel was first rinsed briefly in distilled water and then dena tured with denaturation solution (1.5 M NaCl, 0.5 N NaOH) with two15-min incubations. The gel was so aked in neutralizing so lution (1 M Tris-HCl, 1.5 M NaCl, pH 7.45) for 30 min, followed by 10 SSC (3 M NaCl, 0.3 M Na-citrate, pH 7.0) for 10 min. DNA was then transferred onto Hybond N+ nylon membrane (GE healthcare) in 10 SSC overnight. To fix the DNA on the membrane, UV crosslinking was done for the blots (120 mJ; Stratalinker 2400 UV Crosslinker; Stratagene). Three different probe labe ling methods were employed. Oligonucleotide probes were labeled using T4 polynucleotide kinase (Invitrog en), which catalyzes the transfer of the phosphate from -[32P]-ATP to the 5'-OH group of oligonucleotide probes to label the ends (Reddy et al., 1991). To highlight the linear DNA control arc (1kb plus DNA size ladder) on 2-D gel, 20~50 ng 1kb plus ladder was end-labeled in T4 polynucleotide kinase exchange reaction

PAGE 37

37 buffer by T4 polynucleotide kinase (protocol fr om Invitrogen). To make the IGr-containing DNA fragments labeled as specific probe s, random prime labeling was used ( DECAprime II Rando m Priming DNA Labeling Kit; Am bion) with random decamers. For probing with radioactive oligonucleotide probes, blot s were incubated with prehybridization solution supplemented with denatu red salmon sperm DNA (ssDNA; 50 g/ml) and yeast tRNA (20 g/ml), for 3 hrs at 55 C (1 liter prehybridization so lution was made by mixing 250 ml 1M NaH2PO4, pH 6.0, 300 ml 20 SSC, 15g Na4P2O710 H2O, 200 ml 50 Denhardts solution, 25ml 20% SDS and 225 ml distilled wate r, 0.45 m filter-sterilized). Blots were then hybridized overnight with -[32P]-ATP end-labeled oligonucle otide probes in 5~ 7 ml hybridization solution (same as pre-hybridizati on solution except for omitting 50 Denhardts solution) in the presence of denatured salm on sperm DNA (25 g/ml) and yeast tRNA (10 g/ml) at 50 C ~ 60 C (hybridization temperature depends on the Tm value of oligonucleotide probe). The membrane was washed 5 times in 6 SSC / 0.5 % SDS at 55oC, then washed one time for 5 min with 1 SSC / 0.5 % SDS at 60 C. Finally, the membrane was exposed to film (Hyperfilm MP; GE Healthcare) at -80oC, using an -emitter phosphor intensify screen For probing with end-labeled 1kb plus DNA ladder or random priming-labeled DNA probes, the pre-hybridization procedure was the same to that of oligonucleotide probing except the pre-hybridization solution omitted SDS. Hybridization was then done at 65oC in hybridization solution made the same as pre-hyb ridization solution, but supplemented only with denatured salmon sperm DNA (25 g/ml). The wash es were more stringent than that of oligonucleotide probing, involvi ng washing three times for 20 min with 0.2 SSC / 0.5 % SDS at 60 C. All the used oligonucleotid e probes were listed (Table 2-2).

PAGE 38

38 Results Establishment of a DNA 2-D Gel System to Detect DNAs with Higher-order Structure Construction of DNA cruciform control molecule A plasm id named pL3 was constructed by the ligation of the pBlues cript II (KS+) vector (Stratagene) and contains a DNA insert with an internal perfect 46-bp inverted repeat. The plasmid was transformed into E. coli strain DH5 To retrieve the DNA region containing the cloned inverted repeat, pL3 was cut by BssHII, producing a 0.5 kb BssHII fragment (B0.5) containing pL3 was produced (Figur e 2-1),which was then gel-pur ified. The insert sequence of pL3 was confirmed, and found to cont ain the perfect inverted repeat GGGTTCTTTAGTGCTGTCAGACCGGTCTGACAGCACTAAAGAACCC. Migration of thermally-remod eled crucifo rm molecule To assist the control molecule B 0.5 in forming a cruciform structure in vitro so that it could serve as a real structural control on a 2-D gel a special thermal remodling treatment was developed. During the process of strand se paration at elevated temperature (80.1oC) and gradual annealing, an opportunity is offered to DNA molecules to form s econdary structures dependent on internal sequence homology. This specific DNA thermal remodeling program and specific conditions (Mg2+, NaCl and melting temperature) were adapted to give maximal output of a novel structural form relative to the nor mal linear duplex. The results from this in vitro structureforming assay showed that a clearly identifiab le and novel DNA band with slower migration rate can be generated and is visible on a 1% agarose ge l (Figure 2-1). It is ve ry distinctive from the original linear duplex DNA band observed on agar ose gels. The retarded migration of this novel band indicated structure formation and the st ructured DNA molecules produced this way by in vitro thermal remodeling are designa ted and represented by TR.

PAGE 39

39 The novel TR-B0.5 was tested and found to be qu ite sensitive to S1 nuclease (Figure 2-2), suggesting the existence of singlestranded stretches in the DNA fragment. However, it is hard to study its sensitivity to T7 endonuclease I, a junction-specific enzyme (New England Biolab), since mock treatment suggested the novel form(s) were unstable and mostly lost in the treatment with reaction buffer only (10 mM Tris-HCl, 10 mM MgCl2, 50 mM NaCl and dithiothreitol; pH 7.9). There is equilibrium between structured molecules and linear molecules; therefore the proportion of structured B0.5 woul d not be 100% with the presence of branch migration, which can convert the structured DNA back to original linear form. To maximize the proportion of structured B0.5 as a result of TR treatment, [Na2+], [Mg2+] concentration and melting temperature (Tm) were optimized (Figure 2-9). Elevated concentrations of NaCl and Mg2+ were found not beneficial for this stru cture formation (Figur e 2-3), which was consistent with other researchers observations (Allers and Lichten, 2000). However, incomplete melting promotes its formation within a specific temperature range. The highest melting temperature observed was 94 oC, which failed to give the maximal yield of stru ctured forms (Figure 2-3). In contrast, the much lower melting temperature around 80.1oC gave a better proportion of structured: linear form. The novel band seem so discrete and clear (Figure 21), less likely being cau sed by irregular intermolecular annealing which may create various DNA st ructural forms migrating as more than one discrete band. The efficiency of structure formation in TR treatment of IGr-containing ves fragments was found to be concentration-dependent (Figur e 2-4). Thermally remodeled ph6-1/EcoRI 1.8 kb (LAT) produced two major structural forms, which mi grated faster than orig inal linear form. As the sample concentrations were reduced from 30 ng / l to 2.5 ng / l, the proportion of structured:

PAGE 40

40 linear forms increase accordingly and yet the ra tio between the two major structural forms was relatively constant without detectable changes in the migration of structural forms (Figure 2-4). Lower DNA concentration favors higher proportion of structured forms, which supports the notion that structure formation occurs intramolecularly. More concentrated DNAs in thermal remodeling seem disadvantageous for structure formation. One possible explanation is that DNA is negatively charged and therefore frequent colli sions between them may not be beneficial for the maintanence of higher-order DNA structure. In order to confirm the exis tence of novel DNA structure, thermally remodeled B0.5 was analyzed by DNA 2-D gel electrophoresis with a 1 kb plus DNA size ladder serving as a linear DNA control (Figure 2-5). The novel B0.5 confor mation migrated off the smooth parabolic arc formed by the 1kb plus DNA ladder, further suppor ting the formation of a special structure. Additionally, southern blot analysis confirmed that the offarc spot was B0.5-specific, by hybridizing with a synthesized B0.5-specific o ligonucleotide (HP probe) on the same blot prepared from this 2-D ag arose gel (Figure 2-6). DNA electron microscopy with constructured cruciform To elucidate the structural details of TR-B0.5 and prepare samples for electron m icroscopy, both electroelution and DNA spin-col umn had been attempted to purify TR-B0.5 from agarose gel. However, severe branch mi gration occurred even with the presence of Mg2+ during purification, as a significan t amount of linear form was pres ent after the purification of structured DNA when analysed by agarose gel el ectrophoresis (Figure 2-7) Therefore, after TR treatment the mixture of DNA forms was used directly for electron microscopy. The data allowed for the visualization of structural deta ils for conclusive proof of cruciform existence (Figure 2-8). The lengths of the DNA fragments observed under microscope were estimated according to the size scale and compared with expected lengths. The results showed that

PAGE 41

41 observed structured molecules were about the expected sizes, assu ming that they adopted B-form of DNA duplex. The production of a typical cruciform after TR tr eatment strongly qualified TRB0.5 as good structural control in 2-D gel electr ophoresis, proving the resolving capability of our established DNA 2-D gel system. Analysis of In vitro Structural Potential of Ves Genes Before the detection of in vivo LAT-associated DNA structure, it is necessary to examine in vitro if the two highly conserved quasi-palindromic sequence pairs within the 5 UTR of LAT and many other non-expressed ves gene loci have the potential to form cruciform-like structures, depending on their unusual sequence homology. Subcloned C9.1 LAT (phagemid 6-1; ph6-1) and ves 15-1 locus (phagemid 15-1; ph15-1) were used for this study. The phagemids were cut by EcoRI and gel purified to get two DNA fragments containing the complete intergenic regions (IGrs): ph6-1/EcoRI 1.8 kb (LAT) and ph15-1/EcoRI 1.8kb ( ves 15-1 locus). After TR treatment in TE buffer, both DNA fragments reliably generated certain novel forms which had m obilities distinct from that of the original linear forms in normal agarose gel electrophoresis (Figure 210, 11), suggesting structure formation. The novel forms migrated a little faster than original linear forms, whereas the structured form of cruciform c ontrol B0.5 migrated slower than its linear form (Figure 2-1). So whatever structures formed in the ph6-1/EcoRI 1.8 kb (LAT) or ph15-1/EcoRI 1.8kb ( ves 15-1 locus), the hindrance of migration caused by structured sequences may be overcome by the overall compactness of this DNA molecule, making it migrate faster. Other, larger subcloned ves gene loci fragments were also examined on a 2-D gel after TR (Figure 2-10, 11). Although novel forms also were clearly present for each one of th em, there were also smears trailing at positions below distinct novel bands. Faster migration th an the novel bands suggested that they were forming more compact structures. More variabil ity in structure formation was probably caused

PAGE 42

42 by more extensive homology shared by th e IG-flanking regions in the larger ves gene loci. The light smear shown between linear bands and novel ba nds are possibly the intermediate structural forms produced by branch migration. Migration of thermally remodeled ves gene fragments When exa mined by DNA 2-D gel electrophoresis, the novel forms of the selected ves gene loci migrated as part of an off-linear arc, fo rming a second, distinctive arc together with the structured cruciform control B0.5 (Figure 2-12). To identify and confirm the off-arc spots on the EtBr-stained agarose gel, the 2-D gel was blot ted onto membrane and hyb ridized with LATor 15-1-specific probes (Figure 2-13). As expected the linear forms of the two DNA fragments have co-localized hybridization si gnals since they were about th e same sizes (1.8 kb). However, their structural forms did not co-localize despit e few significant differenc es between their quasipalindromic sequence pairs, as structured ph15-1/ E1.8 migrated a bit faster than did the LAT sequence. This difference is likely due to slight sequence variations in the intergenic regions besides highly-conserved quasi -palindromic patches. It is not certain if struct ured TR-(6-1), TR-(15-1) and TR-B0.5 (Figure 2-10) were accidentally forming the distinctive arc or if diffe rent sizes of cruciforms would form an arc due to structural similarity. Since structured TR-(6-1 ) and TR-(15-1) were near the major linear arc, it was noted that the possible in vivo DNA structures, if cruciform-lik e, might be too near the linear control arc to be identified clearly. Therefore, it is necessar y to choose several enzymes to digest genomic DNA to give sizes different from the LAT in order to facilitate better exploration. However, the structural forms migrate clearly aw ay from their correspond ing linear forms due to the existence of significant unique structures, even though they i ndeed look close to the linear control arc.

PAGE 43

43 DNA electron microscopy with structured ves fragments To visualize those structured DNAs after TR tr eatm ent and elucidate their structure details, it is necessary to carry out DNA electron microscopy to provide direct evidence on the nature of the higher-order structure adapted. DNA electron microscopy was carried out on the thermally-remodeled IGr-containing DNA fragments, using low-angle rotary shadowing to visuali ze the molecules (Figure 2-14). Two major DNA forms were visualized: linear forms and cruciform-like forms. Assuming the linear parts of DNA fragments we re adopting the B form of the double helix, which is the most common DNA form in most utilized solvents, the observed DNA fragment s were estimated to have lengths very close to 1.8 kb, consistent with the sizes of the correct DNA fragments. Some variations in the lengths of th e cruciform arms and the overall shape of cruciforms can be visualized by electron microscopy, despite the fa ct that all of the observed structural forms appeared to be cruciform-like. The major observed cruciform isoform was Yshaped according to the angles between their hairpin arms No significant structural difference between structured TR-(6-1) and TR-(15-1) were observe d under microscope. Their different migration on 2-D gel could possibly be either caused by differential branch migration during gel electrophoresis or by minor structural differen ce in local stem-loops, induced by their sequence difference in or around the IGrs, which affected their partitioning with the agarose substrate during electrophoresis. Detection of In vivo D NA Structure Associated with Ves Genes To detect in vivo DNA structures that might be as sociated with the LAT or other ves gene loci, it is essential to protect potential higher-order DNA struct ures during gDNA extraction and electrophoresis because those br anched DNA molecules could be very unstable and easily

PAGE 44

44 damaged, possibly due to branch migration a nd the loss of protecti on from some DNA binding proteins (Allers and Lichten, 2000). Detection of In vivo DNA Structure with out DNA Crosslinking To extrac t the B. bovis gDNA with protected in vivo structure, a protocol developed for the isolation of recombination intermediates in Saccharomyces cerevisiae was em ployed (Allers and Lichten, 2000). Magnesium has been reported to be capable of restraining branch migration following the formation of Holliday junctions by preventing the loss of base stacking at the crossed portion (Vologodskaia and Vologodskii, 1999; Panyutin et al., 1995). Therefore, Mg2+ was incorporated into the 2-D gel running bu ffer at 3 mM to stabilize DNA structure. To make protective extraction work, the volume of starting material has to be small enough to have highly concentrated DNA at a specifi c step because the gDNAs must spontaneously precipitate without centrifugation, in order to protect in vivo structure. Since B. bovis parasite culture could only reach as high as ~10% parasi temiae, most of the starting materials were uninfected red blood cells. To reduce the volume fo r extraction, the cell culture was first treated by 0.05% saponin to rupture RBCs / IRBCs. The saponin treatment was optimized to not compromise parasite integr ity (Figure 2-15) since the in vivo DNA structure may be greatly damaged or altered in lysed parasites. 6-carboxyfluorescein diacetate (6-CFDA) was used to serve as a vital fluorescent staining dye to exam ine parasite integrity (Imai and Ohno, 1995; Weng et al., 2002). This dye will show green fluor escence after being passively transported into the cytoplasm of live cells and hydrolyzed by cytoplasmic esterases to produce 6carboxyfluorescein (Breeuwer et al., 1995). The op timial saponin concentration was found to be 0.05%, rupturing about 99% RBCs a nd IRBCs in 10 15 min on ice, while most of the parasites remained intact, as indicated by their bright green fluorescence.

PAGE 45

45 To detect LAT-associated in vivo DNA structures, it was also essential to validate genespecific probes, which could be used in 2-D southern blot analysis. Oligonucleotide BAK52 was available to recognize th e C9.1 LAT and an unknown ves locus (Figure 2-16, 17). Another oligonucleotide, LH 2, was later designed and validated to be specific to the LAT. Three different enzymes were used to cleave the LAT into different sizes: EcoRI (1.8 kb), KpnI (2.9 kb), NdeI plus EcoRI (1.3 kb) since it is unknown if structured LAT would migrate too near to its linear form to be distinguish ed. It is of concern that the st ructured LAT fragments generated by restriction enzyme digestion might be too near the linear arc to be seen. With the limited specific probes available for the detection of th e LAT, the choice of enzymes also became very restricted. However, three different enzymes ha ve been chosen to cut the LAT ranging from 1.32.9 kb, so the chances to get the po ssibly structured LAT clearly se parated from its linear form are much more enhanced. To ensure that the DNA structures possi bly observed on the 2-D gel are truly found in vivo the series of reagents were also examined on linear 6-1/EcoRI 1.8 kb (LAT) DNA fragments for structure creation (Figure 3-18). No special form was found after being treated by the reagents for 30 min at 37oC, which is the highest temperature the gDNA samples could experience during protective extraction. To test how this extraction procedure could protect cruciform structures, the series of buffers and solutions used in the extraction pr ocedure were tested on thermally remodeled DNA molecules (Figure 2-19; Table 23). The test was carri ed out by incubating thermally remodeled 6-1/EcoRI 1.8 kb (LAT) DNA fragme nts with excess solu tions or buffes individually at 37oC for 30 min. Treated DNA samples were run on 1% agarose gel to examine DNA migration, indicating structural preservation. However, DNA samples treated by some buffers, which

PAGE 46

46 contain the polyvalent cationic detergent, cety ltrimethyl ammonium bromide (CTAB), were shown to be nearly neutral and therefore migrat ed abnormally. Therefore those buffers could not be tested on structured contro l molecules. However, with the presence of both CTAB and hexamine cobalt chloride, S. cerevisiae in vivo recombination intermediates had a good chance to survive since they were polyvalent cations and could greatly rest rain branch migration (Allers and Litchen, 2000). Other buffers (Table 2-3) we re shown to have variable capabilities for preserving structures in a significant proportion, compared with untreated samples. A significant number of intermediates were observed after treatment with certain reagents. Those intermediates still maintain par tial features of the structures and still have the opportunity to restore some lost structures with stabiliza tion by some reagent(s) like 1 TBE (3 mM Mg2+) (Figure 2-19; Table 2-3). Although this test did not necessarily support that the in vivo DNA structure will surely be well prot ected by this procedure, it suggested that the buffers or solutions used in the procedure of protec tive extraction were to some exte nt protective for structural DNAs. To better evaluate the protective extr action, thermally remode led 6-1/EcoRI 1.8 kb (LAT) DNA fragments were tried to be used as an internal control incorporated in the whole DNA isolation procedure and then 2-D gel electr ophoresis was carried out to examine if there were remaining structural forms migrating off-arc. Unfortunately, most of the control molecules were lost during the procedure at that time. At a particular step, when gDNA was induced to precipitate spontaneously without centrifugation, the internal control was mostly lost due to the failure to co-precipitate with gDNA filaments. Therefore, no conclusion can be drawn from this form of evaluation whether the procedure was pr otective for the structured internal control. The protective extraction was then carried out to get purif ied gDNAs, which were finally dissolved in TMSpe buffer. Throughout the extraction procedure, the possible DNA structure

PAGE 47

47 was protected by different multivalent cations at various steps (CTAB, hexamine chloride, or Mg2+ together with spermidine). Spermidine could affect DNA migration when its concentration is higher than 0.5 mM (Correspondence; Dr. Mi chael Lichten from NI H). Therefore, it is necessary to check the migration of the DNA sample in TMSpe buffer (contains 50 M Spermidine) since DNA 2-D gel analysis is highly based on DNA migration. In addition, to ensure complete restriction digestion but minimize the incubation time at 37oC, which may adversely affect DNA structure pr eservation, optimization of the restriction buffer system was also necessary. This is especially true when the gDNA sample is in TMSpe buffer and may interfere with commercial buffer systems. The re sults (Figure 2-18) sugge sted that no migration difference occurred with or without supplementing with 10 l of TMSpe; and it was found that the basic restriction buffer system (Invitr ogen, Inc), containing only Tris-HCl, MgCl2 and NaCl, was better to get good cutting within the desired time (1~2 h), as compared with the more complicated buffer system of New E ngland Biolab (NEB buffer 1-4). Oligonucleotide LH1 can r ecognize two non-expressed ves loci: 15-1 and 1-1 and can be helpful in examing the li kelihood of existence of in vivo structured silent ves loci. Approximately 10 20 g of gDNA extracted under stru cture-protective conditions was fully digested by restriction enzyme(s), loaded onto a first dimensional ge l (0.6% agarose gel in 1TBE with 3 mM Mg2+) together with control molecules. The 1kb plus DNA size ladder was used as linear control, and thermally remodeled B0.5 as cruciform control to show the separation capacity of the 2-D gel system With the presence of Mg2+, the first dimension gel running with low voltage conditions at 4oC was extremely slow to finis h. The second dimension gel (1.2% agarose gel in 1TBE with 3 mM Mg2+, 0.5 g/ml EtBr) could be run faster with higher voltage. However, the heat when running at high voltage was a concern. It was observed that the gel

PAGE 48

48 temperature was lower than 10oC when being run for 30 hr (3 volt/cm) at 4oC. It has been shown that the structured control molecules maintained their higher-order stru ctures very well under these running conditions (Figure 2-20). EcoRI an d KpnI were chosen for restriction digestion because each of them cut the LAT to generate differently-sized DNA fragments containing the complete intergenic region. EcoRI digests the non-expressed ves gene locus, 15-1 recognized by LH1, into the same size fragment as the LAT (1.8 kb). Hence it could be very informative if the LAT were to adopt certain structure(s) while the 15-1 locus did not. Meanwhile, in a situation where each formed higher-order structures, it woul d also be easier to compare their respective structures. Although NdeI plu EcoRI would gene rate the smaller LAT fragments needed for detection, it greatly added to the difficulty of acquiring complete diges tion of sufficient gDNA for analysis within short time, and longer incubation at 37oC could be disadvantageous for in vivo structure preservation. The availability of LA T-specific probes greatly limited the choice of suitable enzymes. The 2-D gels were blotted onto nylon membrane and then a series of hybridizations with gene-specific oligonucleotide probes were carried out to detect the existence of in vivo DNA structures. However, no significant off-arc spot was found specific to the LAT or to several nonexpressed ves gene loci, including ph15-1, ph1-1, regard less of the specific enzyme(s) used to cut the gDNA (such as Figure 2-21, 22, 23). To imp rove detection sensitivity and meanwhile generally find out if any of the highlyconserved intergenic regions among many ves gene loci adopted certain structures, a 946-bp DNA fragment containing the intergenic region from the ves gene locus 1E10 was random-prime labeled by [32P]-dCTP (DECAprime II Random Priming DNA Labeling Kit; Ambion) and used as probe to hybridize with those same blots. The DNA probes recognized a series of DNA bands and high lighted almost half of the whole linear arc

PAGE 49

49 with much stronger signal inte nsity; however, there was still no off-arc spots found near the major linear arc. When overlaying two autora diographs generated fr om the hybridizations, respectively, with the HP probe (cruci form control B0.5) and 1E10 946-bp DNA probes respectively, the structured B0.5 was clea rly shown to be off-arc (Figure 2-21). Despite previous failures likely caused by limited signal intensity or inappropriate restriction enzymes to cut the LAT, a certain o ff-arc spot was found (Figure 2-26; panel A) at position lower than the LAT. However, the off-arc signal failed to be repeated with consistency on other blots prepared afterwards with structur e-protective extraction and the digestion of NdeI plus EcoRI for unknown reason (such as Figure 2-23). Detection of In vivo D NA Structure with DNA crosslinking It is not clear whether th e protective extraction is insufficient to maintain any in vivo structures, or if there simply is no LAT-associated structure existing in vivo Psoralencrosslinking is a powerful tool to fix in vivo DNA or chromatin structures and make them survive the detection procedure (Bell and Byers, 1983; Jasinskas et al., 1998; Wellinger et al., 1999; Sanz et al., 2007). Psoralen is part of a group of photoreactive DNA intercalators, which includes many derivatives, and can covale ntly crosslink oppositing pyrimi dines of double-stranded DNA to secure DNA or chromatin structures at prot ein-free regions upon exposur e to long-wavelength ultraviolet light of about 365 nm (Sastry et al ., 1997; Wellinger et al., 1999). The most popular form of psoralen used in chromatin structure re search is trimethylpsoral en (TMP). The psoralencrosslinking is reversible, wh ich can be achieved chemically (base-catalyzed reversal) or by photoreversal, using irradiation w ith short-wavelength ultraviolet light (Sastry et al., 1997; Sanz et al., 2007). The reversal of psoralen-crosslinking is essential for hybridizations in southern blotting analysis (Sanz et al., 2007).

PAGE 50

50 After 0.05% saponin lysis to rupture most of the RBC/IRBCs, 4, 5', 8-trimethylpsoralen crosslinking (Sigma-Aldrich) was carried out at UV365nm on ice for 10-15 min. Then gDNA was purified and digested to completion with hi gh concentrations of restriction enzymes (>2 unit/g of gDNA) till completion (Jasinskas et al., 1998). The gDNA fragments were run on a 2D gel and blotted onto nylon membrane. To ev aluate the crosslinking procedure, internal controls, which are thermally remodeled B0.5 or ves locus 15-1/E1.8 kb fragments, were included at the crosslinking step. The crosslinki ng time of psoralen was somewhat optimized according to the criteria that th e internal control should preser ve structures and complete restriction digestion was not interfered. The optimal crosslinking time was about 10-15 min under UVA 365 nm. In some experiments, hybridi zation with BAK52 failed to show off-arc structural form(s) (Figure 2-24, 25). However, hybridization with the HP probe showed a proportion of that the cruciform control molecule B0.5 retained a structural form (TR-B0.5; Figure 2-24, 25), which may suggest that the en tire procedure preserve d certain structural characteristics of control mol ecules. A 434-bp DNA fragment contai ning the intergenic region of LAT was random priming labeled with [32P]-dCTP and hybridized with the same blot (Figure 2-22). Since this DNA probe contains conserve d intergenic reigon, which is shared by the internal control TR-(15-1), it can also recognize the off-arc spot specific to TR-(15-1). No signicant off-arc signal was found on t hose 2-D gel blots (Figure 2-24, 25). Another approach was taken to demonstrate whether the LAT IGr was involved in the formation of higher order DNA structure associat ed with transcription, crosslinked gDNA was purified and split into equal parts, being digested by EcoRI, or NdeI plus KpnI, respectively and loaded onto 2-D gels individually to do southern blot analysis with LH 2 (Figure 2-25). With EcoRI digestion, LH 2 would recognize a 3.3-kb LAT fragment without the IGr (Figure 2-25

PAGE 51

51 (A)). With NdeI plus KpnI digestion, LH 2 would recognize a 1.6-kb LAT fragment containing the IGr. Comparing the hybridizat ion results obtained from two different regions of the LAT, no significant off-arc spot was found significan tly specific to the LAT (Figure 2-25). However, an off-arc spots was once observ ed on a 2-D gel prepared with psoralen crosslinking (Figure 3-26) although their specif icity and significance were both uncertain. The consistency and repeatability of the appearance of those off-arc signal(s) seem poor since it was not shown in another repeated experiment shown in Figure 3-25. However, it is possible that the internal control was too near the position where the structur ed LAT may locate and thereby block the direct view of it. Anyway, it is very difficult to capture and preserve in vivo high order DNA structure with many failed detections. Other direct approaches are also essential to complement the limitation and weakness of 2-D ge l analysis and thereby strictly interpret any off-arc signal suggesting certain higher-order DNA structure. Discussion Establishment of 2-D Gel System to Det ect Higher-order DNA Structure in Chromatin A previous comprehensive study in B. bovis reported that the intergenic regions among ves genes are, on average, approximately 433 bp long, highly conserved, and usually feature two pairs of quasi-palindromic DNA segments (Al-Kh edery and Allred, 2006). The inverted repeats have strong potential for struct ure formation, and are functionally important in many organisms, being involved in various biological aspects of cell function su ch as gene conversion, gene regulation and chromosome segregation (Gordeni n et al., 1993; Kato et al., 2003; Dorn and Maddox, 2008). Based on the theoretical predicti on provided by the Mfold algorithm (GCG), the conserved ves IGrs have the potential to adopt extensiv e secondary structure, featuring multiple minor stem-loops as tiny branches off a main cr uciform-like skeleton. It would be informative to find out if the conserved quasi-palin dromic pairs in 5 UTR region of ves genes regulate gene

PAGE 52

52 expression through a formation of certain DNA st ructures. We hypothesized that the LAT may form cruciform-like structure in its 5 UTR region and act exclusively as a strong promoter for both flanking ves 1 / gene pair. This hypothesis could e xplain the differential transcription status between the LAT and other non-expressed ves gene loci. In this study, we searched and endeavored to understand the DNA structure differences between the LAT and inactive ves gene loci regarding their differentia l transcription status. First, we established a DNA 2-D gel electrophoresis system to effectively separate DNA possessing higher-order structures, such as cruciforms. Then, the structure-protective gDNA extraction and in vivo psoralen crosslinking methods were both attempted to complement th e DNA 2-D gel system technique in order to detect in vivo DNA structures associated with th e LAT. The data collected on the in vitro structure(s) of the ves intergenic regions strong ly suggested that they a ll have similar potential for structure formation (Figure 2-10, 11). Some in vivo higher-order DNA structures were observed with uncertain si gnificance (Figure 2-26). DNA 2-D gel system has been applied to separate and visualize Holliday junction molecules in meiotic recombination, which can occur in high frequency in the yeast, Saccharomyces cerevisiae (Bell and Byers, 1983). In brief, it is composed of sequential agarose gel electrophoresis separations employing different agarose concentrations and voltage conditions (volt/cm) during the electrophoretic run. Ideally, the two dimensions separate DNA molecules based on their diverse mass and shape, respectively. Various modifications on this technique have been developed to expand its application. Curren tly, it is commonly used to study recombination intermediates a nd replication/repair intermed iates (Allers and Litchen, 2001; Courcelle et al., 2003; Bernstein et al., 2009). Although Holliday junctions are stereochemically similar to cruciforms (Zhang et al., 1999; Iwasaki et al., 1992), it was not clear if cruciforms

PAGE 53

53 could be separated on DNA 2-D gel system. It was important to confirm this since the detection of in vivo DNA structure would be highly dependent on the resolving capability of this separation system. To make cruciforms distinguishable from bulk linear DNA of different lengths, it is essential to have linear DNA mappe d on a gel. A 1 kb plus DNA size ladder was chosen as linear control on 2-D gel due to its wide range of si ze coverage and easily r ecognized pattern. Most importantly, 1kb plus ladder had been tested in southern blot analysis and was known not to interfere with the probing of B. bovis ves genes. On the 2-D gel system, it migrates to form a smooth arc, and can be highlighted by autora diography with an end-labeled 1kb plus ladder (Figure 2-5). Thus, it is useful to provide a control pattern to reve al any structured molecule that has distinctive migration properties. Meanwhile, it is also critical to have a known cruciform molecule as a control to validate the 2-D gel resolving capability. We constructed a small DNA molecule containing a 46-bp perf ect inverted repeat (Figure 2-1). The plasmid pL3 containing the 46-bp inverted repeat was stable, and could be transformed and maintained in the E. coli strain DH5 whereas attempts at construction of a 400bp inverted repeat we re unsuccessful due to stability issues. This construct was also f ound to be a difficult template in cycle-sequencing, resulting in missing sequence near the location of the inverted repeat, which suggested that structure formation might occur during thermal cycling which inhibits amplification. To make the inverted repeat form a cruc iform and separate off-arc on 2-D gels, a procedure referred as thermal remodeling was employ ed. It is generally comprised of heating DNA duplex to allow breathing of the strands, with gradual cooling an d reannealing to give DNA molecules chances to form alternative structures re lying on sequence specificity.

PAGE 54

54 After thermal remodeling, the control mo lecule B0.5 formed a novel band on normal agarose gel, which migrated more slowly and repeatedly migrated off-linear arc on a DNA 2-D gel (Figure 2-5, 6). Furthermore, the structural details observed by elec tron microscopy revealed that TR-B0.5 was cruciform with the small hairpi n portion internally locate d, consistent with the position of the inverted repeat (Figure 2-8). Interestingly, cruciform B0.5 migrates more slow ly on a 1D agarose gel, but a little faster on a 2-D gel compared to linear B0.5, and finall y, is positioned lower than the linear DNA arc (Figure 2-1; Figure 2-5). Recomb ination or replication interm ediates separated on the DNA 2-D gel system are larger than 5 kb and usually obs erved to migrate above the linear DNA arc (Allers and Litchen, 2001; Courcelle et al., 2003; Bern stein et al., 2009; Bell and Byers, 1983). This difference may be caused by the large differences in MWs of examined molecules, or simply structural differences. Although all the mitotic in termediates observed migrated above the linear arc, smaller intermediates of approximately 4 kb te nd to get much closer to the linear arc than do bigger intermediates (Friedman and Brewer, 1995; Bell and Byers, 1983). Hence, it is possible that the intermediates, if small enough, may migrat e off-arc, lower than the visible linear arc. Indeed, if one plots the migration of TR controls versus their mo lecular weight, the off-axis arc intersects the linear DNA arc at about 3 kb. Furthermore, when the 2-D gel blot was probed with a large ves probe that hybridized with numerous family members, off-axis spots above the linear arc were observed at sizes > 6 kb, consistent with detection of replication intermediates (Figure 2-21). To stabilize the conformation, 1XTBE (3 mM Mg2+) was used as the DNA 2-D gel running buffer, which was quite advantageous to the stabiliz ation of the structures. Southern blot analysis revealed that after 2-D gel electrophoresis, th e structural form of B0.5 comprised a larger

PAGE 55

55 proportion of total B0.5, in comparison with nor mal horizontal agarose gel electrophoresis (Figure 2-1; Figure 2-6). The increase of this structured form is probably due to the effect of Mg2+ in restraining branch migration which st abilized the crucifor m (Vologodskaia and Vologodskii, 1999; Panyutin et al., 1995). On th e autoradiograph (Figure 2-6), there are two minor trail lines forming a right angle, whic h connected the cruciform B0.5 and linear B0.5, suggesting the existence of intermediate structural forms. The good resolution and preservation of th e small cruciform B0.5 proved that the established DNA 2-D gel system can be effectively used to separa te cruciform from linear DNA, which was supported by the good separation of larger structured IG-containing ves fragments (Figure 2-11). To my knowledge, this is th e first time that DNA 2-D gel electrophoresis has clearly been demonstrated to separate typical cruciform from linear DNA. In vitro Structure Study of Ves Ge nes Although quasi-palindromic sequences are not perf ect inverted repeats, they still showed some potential to form cruciform-like structures (Kurahashi and Emanuel, 2001). To examine if the intergenic region (IGr, which is quasi-palindr omic) at the LAT can form some kind of higher order structure in vitro, the purified IGr-containing LAT fragment was thermally remodeled and resolved on a 2-D gel. Similarly to B0.5, it form ed novel band(s), which migrates off-linear arc (Figure 2-9). The LAT-specific structures were then observed by electron microscopy (Figure 213). These data demonstrated that the LAT inte rgenic region can form higher-order structure in vitro and has the potential to adopt such structure in vivo According to the alignment of th e intergenic regions among several ves gene loci, other loci could be expected to have the same structur al potential as that of th e LAT IGr. It was also observed that many ves gene loci, including S621, 1E10, 10-1 and 15-1, were able to form quasicruciform structures (Figure 2-9; Figure 2-13). All of these results indicated that the ves

PAGE 56

56 intergenic regions have simila r structural potentia l due to equivalent sequence specificity. Intriguingly, different sizes of ves IG-containing DNA fragments, after being thermally remodeled, produced off-arc structured forms, whic h together with the cr uciform B0.5 create a separate arc distinctive from the linear arc, and tend to merge into the linear arc at approximately 3~4 kb (Figure 2-11). The formation of this specia l arc could be caused by th eir shared structural similarity. If we assume that the analogously structured DNA molecule is large enough, it may migrate above the linear arc around the high mass portion. Indeed, the probing of a 2-D gel membrane with a randomly-primed ves IGr yields a second arc that crosses over that of the linear arc at around 3~4 kb (Figure 2-20). Another notable phenomenon regarding in vitro structure of ves IG-containing fragments is that fragments larger than 3 kb usually produ ced not only novel bands but also long smears trailing in front of the novel bands. This may be led by the more extensive homology shared between the two juxtaposed gene pairs flanking the intergenic regions such as ves loci 1-1 or 101 (Al-kehdery and Allred, 2006). The structural forms in the smear are probably more compact and variable due to the potentia lly extensive variety of higher or der DNA structures that may be adapted. The ves 15-1/E1.8 kb locus, which is not transcrip tionally active, was observed to adopt a higher order structure similar to that form ed by the 6-1(LAT)/1.8 kb locus (Figure 2-14). Considering the structural pot ential of such non-expressed ves gene loci, it is probable that adoption of higher-order structure is not necessarily associated with the LAT only. In addition, those cruciform-like structures may not necessa rily correlate with transcription only. The scenario might be even more complicated. Palin dromic sequences are an important source of genome instability (Kurahashi and Emanuel, 2001). It is conceivable that structure formation at

PAGE 57

57 the ves intergenic region might play a role in segmental gene c onversion at a specific stage, helping the LAT acquire donated sequence patches from other ves loci to become more mosaic (Al-Khedery and Allred, 2006). Any structure, if found at the LAT, might also have the likelihood to be related with ge ne conversion. However, the fr equency of gene conversion in B. bovis which most likely occurs during mitosis at the asexual stage, may not high enough to make intermediates detectable. In addition, du ring DNA replication/repair, the DNA strands can also be separated and some form of intermed iates may also appear. In mammalian organisms, cruciforms also have been demonstrated to f unction in the initiation of DNA replication (Novac et al., 2002). Since a DNA 2-D gel system can be used to resolve all kinds of structured molecules involved in recombination/replicati on intermediates, caution must be taken to correlate the possible DNA structure detected on a 2-D gel, with transcriptional status of the LAT. Detection of In vivo LAT-Associated DNA Strucutre To detect in vivo DNA structures, it is crucial to pr otect any higher-order structures fro m being lost during gDNA extraction and migration in the 2-D gel sy stem. I tried two approaches to maintain structures: the use of structure-protective extraction to restrain branch migration, and in vivo psoralen-crosslinking to fix DNA structures covalently prior to gDNA extraction. Most of the efforts failed to detect significan t higher-order DNA structure associated with the LAT or with non-expressed ves gene loci (Figure 2-21, 22, 23, 24, 25) probably due to the instability of DNA structure or method sensitiv ity. Thus, higher-order structure failed to discriminate the LAT from non-expressed ves loci. However, on some 2-D gels (Figure 2-26) either prepared with gDNA extracted structure-pr otectively or psoralen crosslinked gDNA, some off-arc signals were detected and mostly likely associated with the LAT due to its discriminative transcription status; therefor e still the possibility that in vivo higher-order DNA structure

PAGE 58

58 associated with the LAT assists ves transcription can not be ru led out. Nevertheless, caution should be taken to correlate those detected struct ures with LAT and its tr anscription. In Figure 226 panel B, it is clear that crosslinked g DNA was incompletely digested and therefore differently-sized LAT or another sequence-unknown ves locus showed up with BAK52 hybridization. Therefore, it is hard to strictly defi ne the off-arc signals. Those high order structure indicated by the off-arc signals we re most probably related to LAT transcription; however, the possibility of its associated with recombination involving the LAT cannot be ruled out due to the limitati ons of this study. The low frequency of gene conversion might explain the difficu lty to capture them. Straightfo rward evidence could be that the inactivation of the current LAT makes the structure not detectable anymore. Several possibilities exist to ex plain the poor repeatability of those off-arc signals if they are associated with the LAT. One is the instabil ity of possible LAT-associated structure, which was suggested by the failture to purify the st ructured TR forms for electron microscopy. Secondly, even though the B. bovis asexual cell cycle is on averag e 8 hr, it is really unknown when the ves 1 starts to be actively tr anscribed, and how long the ac tive transcription lasts. Moreover, the parasite culture used to do struct ure detection was not synchronized at all as no techniques currently exist to accomplish this, and therefore was a mixed cell population consisting of parasites at different stages. The de tection failure could thus be due to the short time of structure maintainence. It is possibly caused by short span of time when ves 1 is actively transcribed. Moreover, higher or der DNA structure associated w ith the LAT may probably be only needed transiently to induce active transcri ption and then not maintained any more. If it would be the case, the structured LAT would be too elusive to be captured. Big DNA structures are physiologically not favorable and are more likely to be sh ort-lived (Inagaki et al., 2009).

PAGE 59

59 Thirdly, structure extrusion and maintenance may need the involvement of topoisomerases and DNA-binding proteins. Hence, when any type of gDNA extraction denatures and removes all proteins, the structure may be irreversibly compro mised and has no chance to be detected in later steps. Fourthly, saponin lysis may still create problems in maintaining the parasite integrity since preliminary observations suggested the leaking of cytoplasmic pr oteins with saponin treatment (Mack, E and Allred, DR; unpublished data). Meanwhile, the stabilizatio n of chromatin structure by cross-linking the DNA with psoralen may be low efficiency since DNA bound with proteins is protected against psoralen crosslinking under long wave-l ength UVA (Widmer et al., 1988; Sastry et al., 1997; Wellinger et al., 1998). The formation of cruciform or other higher-order structure may comprise multiple steps includ ing strand separation, homologous strand pairing and structure extrusion, and DNA-binding proteins such as single-stranded DNA binding proteins (SSB) and RecA may participate and conse quently stabilize the stru ctures (Kotani et al., 1993). Meanwhile, the most character istic part of the cruciform is the four-way junction, which likely requires stabilization to avoid junction migration. Crucifor m-binding protein (CBP), which is a highly conserved regulatory protein among hi gher eukaryotes, can interact with cruciform non-specifically and essentially acts to regulate DNA replication (Todd et al., 1998; Novac et al., 2002). Some other proteins capable of binding cr uciform DNA, such as HMG1 and RepC, are also implicated in regulation of different cellular activities (Kotani et al., 1993; Bianchi et al., 1983; Pearson et al., 1996). Although a BLAST search failed to find CBP homologues in B. bovis, multifunctional proteins functionally similar to CBP might exist in B. bovis to recognize and bind cruciforms as a part of normal DNA homeostasis, including at the LAT or non-

PAGE 60

60 expressed ves loci where they could play roles in segm ental gene conversion, and/or transcription regulation of ves 1 genes. Despite any stated hypothesis, it is possible that no higher order DNA structure exists in vivo at the C9.1 LAT to contribute to its active tr anscription. Cruciform ex trusion from GC-rich palindromic sequence, unlike AT-rich palindrom ic sequence, is less physiologically and kinetically favorable in cells (Courey and Wang, 1983; Inagaki et al., 2009). The intergenic regions, and especially their c onserved palindromic segments are quite GC-rich (approximately 76.4%), whereas the overall GC content of the en tire intergenic regions is around 50%. Since the two quasi-palindromic sequence pair s in the intergenic regions of ves 1 genes are highly conserved, it would not be unreason able to suppose they have a certain functional importance at a non-structural level. For example, dimeric prot ein factors, such as steroid hormone receptors, can only recognize and cooperatively bind palindr omic sequences to regulate gene expression (Yannone and Burgess, 1997; Schwabe et al., 1995). Po ssibly, there are some unusual protein factors in B. bovis which can form palindromic DNA-protein complexes to exert specific promoter activity. To answer the question if there is higher orde r structure associated with the LAT to induce active transcription, future study in two aspects may be helpful. Fi rstly, testing promoter activity of thermally remodeled ves intergenic regions using in vitro transcription system may provide direct evidence if the formation of cruciformlike structure promotes transcription. Secondly, examination of LAT transcription after mutage nesis being introduced at the conserved quasipalindromic sequence patches would reveal indirec tly if the integrity of palindromic elements potentially involved in structur e formation is essential for ves 1 expression.

PAGE 61

61 Table 2-1. List of primers used in cruciform control construction Primer name Sequence (53) DA44 -CTGCACGCTAGCGAGGATAGGCTACCAGGGTTTAGTTGPIRR-PinAI -ACGTACCGGTGAGACTACCCCATTACCATACCPIRF-PinAI -ACGTACCGG TCTGACAGCACTAAAGAACCCPIRR-BamHI -ACGTGGATCCGAGACTACCCCATTACCATACCTable 2-2. List of oligonucleotide probes used in southern blot analysis Oligonucleotide name Sequence (53) BAK6 -GATTATCCAGTACCGGTAGATTCBAK52 -CAATCTATGTGTCTATGCCAACGLH1 -CCATACAAACTCTCCTTTCTGACLH2 -CCAGCACTATTACAGTTACAAGTAAZT02 -AACAACTGGGCCAAGGGTCBAK68 -GTAGGCGCTGAGAACCACCACHP -GCGCGCGTAATACGACTCACTATAGADPH-E1R -TGAACAGAATCGTACTTCAGCAAGTLH 1 -ACCTAGTAGTGGAGTAGTACTATCTLH 2 -ACTGATACACAAGTCACAGACCA

PAGE 62

62 Figure 2-1. Construction of cr uciform control molecule B0.5 and the appearance of novel DNA form(s) after thermal remodeling treatment. (A) Confirmation of the structure of plasmid pL3, which contains a 46-bp i nverted repeat. The cruciform control molecule B0.5 can be cut from pL3 by restriction enzyme BssHII to get a 532-bp fragment. After thermal remodeling, a novel DNA band showed up on 1% agarose gel, migrating significantly slower than original linear form. (B) Schematic map of B0.5 molecule and the recognition site of its specific probe (HP probe) used in southern blot analysis A. B.

PAGE 63

63 Figure 2-2. The novel forms produced after thermal remodeling are sensitive to S1 nuclease. (A) Thermally remodeled cruciform control molecule B0.5 or ves IG-containing DNAfragments were treated with 1 un it of S1 or S1 buffer only at 37oC for 10 min to examine the susceptibility of novel bands to S1, which prefer entially cuts singlestranded DNA template; (B) Schematic map of ves IG-containing DNA fragments including 1E10 and S621 loci. A. B.

PAGE 64

64 Figure 2-3. The influence of [Mg2+] and [NaCl] in the thermal re modeling of cruciform control molecule B0.5 or ves S621 IG-containing DNA molecule was originally in TE buffer (10 mM Tris, 1 mM EDTA, pH8.0), and then supplemented with different amount of MgCl2 or NaCl to be thermally remodeled under a melting temperature gradient.

PAGE 65

65 Figure 2-4. The influence of concentration in the ther mal remodeling of ves LAT gene sequence. 5 l of different concentrations of the LA T DNA fragment (ph(6-1)/EcoRI 1.8kb) in TE buffer were thermally remodeled and loaded onto a 1% agarose gel (A). The arrows represent different DNA forms. Th e ratios between linear form and NTR forms were designated under the gel image a nd plotted against sample concentrations (B). B. A.

PAGE 66

66 Figure 2-5.Thermally remodeled B0.5 on DNA 2D gel electrophoresis. 100 ng of thermally remodeled cruciform control molecule was mixed with 300 ng of 1kb plus DNA size ladder to run on DNA 2-D gel. First Dimens ion: 0.75% agarose gel, 1 TBE (3 mM Mg2+), 1v/cm for 62 hr at 4oC; second Dimension: 1.5% ag arose gel, 1 TBE (3 mM Mg2+, 0.5 g/ml ethidium bromide), 3v/cm for 22 hr at 4oC. The gel was then photographed under UV.

PAGE 67

67 Figure 2-6.2-D Southern blot analysis of thermally remodeled B0.5. The 2-D gel (Figure 2-5) was blotted on Hybond N+ nylon membrane for hybridizations. Two autorads were overlaid and scanned. The smooth 1kb plus DNA arc was highlighted by the hybridization with -[32P] ATP labeled 1kb plus DNA, and B0.5 was recognized by [32P] ATP labeled HP probe. The blue a rrow represents linear B0.5 on linear DNA control arc; purple arrow repr esents structured B0.5, which migrated off-linear arc.

PAGE 68

68 Figure 2-7. Structure loss during the purification of novel DNA forms. After thermal remodeling, the sample was run on 1% agarose gel (1XTBE, 3mM Mg2+) and two DNA forms were cut and purified individually by DNA spin-column. The purified DNAs were run on 1% agarose gel (1XTBE, 3mM Mg2+) to check mobility. The arrows represent different forms purified by DNA spin-column.

PAGE 69

69 Figure 2-8 Visualization of the thermally-remodeled B0.5 through DNA electron microscopy. Thermally remodeled cruciform control molecule B0.5 was spread onto carboncovered formvar film copper grids (200-mesh), which was done with Pt/Pd, 80/20 rotary shadowing. The size bar represents 0.2 nm.

PAGE 70

70 Figure 2-9. The influence of melti ng temperatures (Tm) in the thermal remodeling of the IGcontaining ves gene fragments. The temperature gr adient is (from lane 1 to 12): 70oC, 70.6 oC, 71.8 oC, 73.8 oC, 80.1 oC, 84.1 oC, 87.6 oC, 90.2 oC, 92.2 oC, 93.5 oC, 94 oC. IG represents i ntergenic region. A. B.

PAGE 71

71 Figure 2-10.The structure-fo rming potential of other ves IG-containing DNA fragments (1). (A) Schematic map of two ves loci used in vitro study; (B) Thermally remodeled ves loci and the novel bands were shown on 1% agaros e gel. U/L represents untreated linear DNA forms of ves loci; TR represents thermal remodeled DNA forms of ves loci. The arrows represent two different DNA fo rms with distinct migration rates. B. A.

PAGE 72

72 Figure 2-11.The structure-fo rming potential of other ves IG-containing DNA fragments (2). (A) Novel bands showed up on agarose gel afte r being thermally remodeled. The arrows represent different forms of DNA: blue for original linear DNA and purple for structured DNA forms; (B) schematic map of examined ves IG-containing gene fragments. B. A.

PAGE 73

73 Figure 2-12. Thermally remodeled ves IG-containing DNA fragments on DNA 2-D gel electrophoresis. A distinctive ar c can be visualized to be composed of all structured DNA forms after thermal remodeling. The colo red arrows represent structured forms of different ves IG-containing DNA molecules and cruciform control molecules. NTRrepresents novel form after thermal remodeling.

PAGE 74

74 Figure 2-13. 2-D Southern blot an alysis of thermally remodeled ves IG-containing DNA fragments. (A). Thermally remodeled ves IG-containing DNA mol ecules were run on 2-D gel together with 300 ng of 1kb plus DNA size ladder. (B). Southern blot analysis to confirm the locations of diffe rent DNA forms. Three individual autorads were overlaid and scanned. Colored arrow heads point at different forms of DNA molecules. LAT-(6-1/EcoRI1.8kb) was recognized by -[32P] ATP labeled oligonucleotide BAK52 and ves 15-1 was recognized by -[32P] ATP labeled oligonucleotide LH2. Cruciform mo lecule B0.5 was recognized by -[32P] ATP labeled HP probe. A. B.

PAGE 75

75 Figure 2-14. Visualization of thermally remodeled ves IG-containing DNA fragments through DNA electron microscopy. Thermally remodeled 6-1 and 15-1 molecules was spread onto carbon-covered formvar film copper grids (200-mesh), which was done with Pt/Pd, 80/20 rotary shadowing. The size bar represents 0.2 nm.

PAGE 76

76 Figure 2-15. Optimization of saponin lysis for structure-protective extraction. Different concentrations (0.02~0.07%) of saponin in PBS were used to treat parasite culture and after saponin lysis th e parasites were washed once with 1VYMs buffer and stained with 0.5 g/ml 6-CFDA in 1VYMs buf fer at room temperature for 15 min in the dark. 0.05% saponin treatment lysed 99% of IRBC/RBC without losing 6-CFDA fluorescence; however, 0.06% saponin trea tment lyses 100% of IRBC/RBC with some of the parasites not showing 6-CFDA fluorescence.

PAGE 77

77 Figure 2-16. Test of olig onucleotide probes (1). (A) ves gene-specific oligonucleotide probes used in southern blot analysis. Puri fied genomic DNAs from MO7, C9.1 and B9 parasite clonal lines were digested by Ec oRI, separated on 0.8% agarose gel and then blotted onto Hyond N+ nylon membrane for hybridizations. The hybridizations were done by using -[32P] ATP labeled oligonucleotide probes shown in (B). The arrows represent different ves loci. Subcloned phagemid DNAs of corresponding ves gene loci (15-1, 1-1 and 6-1/LAT) were used as positive control. (B) Schematic map showing probe recognition sites on the corresponding ves gene loci. A. B.

PAGE 78

78 Figure 2-17. Test of olig onucleotide probes (2). (A) ves gene-specific oligonucleotide probes used in southern blot analysis. Puri fied genomic DNAs from MO7, C9.1 and B9 parasite clonal lines were digested by K pnI, separated on 0.8% agarose gel and then blotted onto Hyond N+ nylon membrane for hybridizations. The hybridizations were done by using -[32 P] ATP labeled oligonucleotide pr obes shown in (B). The arrows represent different ves loci. (B) Schematic map showing probe recognization sites on the corresponding ves gene loci. B. A.

PAGE 79

79 Figure 2-18. Migration of protectively extrac ted gDNA in TMSpe buffer and optimal buffer system for gDNA digestion. C9.1 gDNA was di gested to completion with EcoRI or KpnI and then loaded on 0.8% agarose ge l together with or without 10 l TMSpe buffer, to test the influence of spermi dine on migration. On the other hand, C9.1 gDNA in TMSpe buffer was digested by EcoR I or KpnI using reaction systems from Invitrogen or New England Biolab in the same period of time. This is to optimize and test the restriction enzyme buffer syst em used in digestion of gDNA extracted structure-protectively. The gel was blotted onto Hybond N+ membrane and hybridized with -[32P]-ATP labeled oligonucleotide BAK 52. Lanes 1and 3 serve as positive controls for complete digestion of gDNA by EcoRI or KpnI.

PAGE 80

80 Figure 2-19. Evaluation of a series of reagents in structure-prot ective extraction on the thermallyremodeled control molecules. After thermal remodeling treatment, equal amount of TR-DNA molecules (20 ng/l; 5 l) were incubated with 15 ul of different reagents (Table 2-2) at 37oC for 30 min, and then loaded on 1% agarose gel immediately to check the structure loss (A, B) or gain (C). A. B. C.

PAGE 81

81 Table 2-3. List of the reagen ts used in structure-protec tive extraction of genomic DNA No. of Reagent Reagents used in structure-protective extraction S1 PBS Extraction buffer Chloroform Dilution buffer S2 1 VYMs S3 Wash buffer S4 NaCoHex buffer 70% ethanol + 0.3 mM Co(NH3)6Cl S5 2 colors S6 TMSpe solution S7 1TBE with 3 mM MgCl2S8 1 React No.1 buffer S9 1 React No.2 buffer S10 1 React No.3 buffer S11 1 React No.4 buffer S12 1 protective sample loading buffer

PAGE 82

82 Figure 2-20. 2-D gel electrophoresis of C9.1 genomic DNA extracted by structure-protective procedure. (A) 2-D gel electrophores is. C9.1 genomic DNA was extracted protectively, digested by EcoRI and then se parated on 2-D gel. First dimension: 0.6% agarose gel, 1XTBE (3mM Mg2+), 1 v/cm. 62 hr; 2nd dimension: 1.2% agarose gel, 1XTBE (3mM Mg2+; 0.5ug/ml EtBr), 3 v/cm, 22 hr. (B) schematic alignment of the LAT and two detectable ves loci, and the recognition sites of specific probes are shown in diagram. 1st D2 nd D B. A.

PAGE 83

83 Figure 2-21. Detection of in vivo DNA structures associated with ves loci (1). The 2-D blot was from the 2-D gel in Figure 3-20. (A) (C ) were scanned autogradiographs, which were respectively from the hybridizations with -[32P] ATP end-labeled BAK52 oligonucleotide recognizing th e 1.8-kb LAT and an unknown ves locus, -[32P] dCTP random priming labeled ves 1E10 locus (part; from1E10 phagemid, double digested by SmaI and PstI), and -[32P] ATP labeled LH1 oligonucle otide. (D) Schematic map of the DNA probe used in (C). The arrows point at examined ves loci or cruciform control. A. B. C. D.

PAGE 84

84 Figure 2-22. Detection of in vivo DNA structures associated with ves loci (2). (A) Gel image of 2-D gel electrophoresis. C9.1 genomic DNA wa s extracted protectively, digested by KpnI and then separated on 2-D gel as in Figure 3-19. (B) (D) were scanned autoradiography films, which were respectively hybridized with -[32P] ATP labeled BAK52, -[32P] dCTP random prime labeled 1E10 locus, and -[32P] ATP labeled LH1. The arrows point at examined ves loci or cruciform control. A. B. C. D.

PAGE 85

85 Figure 2-23 Detection of in vivo DNA structures associated on the LAT with protective extraction. (A) Gel image of 2-D gel elec trophoresis. C9.1 parasite culture was 0.05% saponin-treated and then irradiated with the presence of 100 g/ml 4-aminomethyl-4, 5, 8-trimethlpsoralen under 365-nm UVA (Black-Ray) for 15 min at 4oC. 60 ng of thermally remodeled cruciform control was incorporated at the crosslinking step serving as internal control. Purified DNAs were then separated on a 2-D gel. (B) The 2-D gel (A) was blotted onto nylon me mbrane and then hybridized with -[32P] ATP labeled-BAK52 and HP probe. The arrows point at examined ves loci or cruciform control. A. B.

PAGE 86

86 Figure 2-24 Detection of in vivo DNA structures associated on the LAT with psoralen crosslinking. C9.1 parasite culture was 0.05% saponin-treated and then irradiated with the presence of 100 g/ml 4-aminomethyl-4, 5, 8-trimethlpsoralen (Sigma) under 365-nm UVA (Black-Ray) for 10 min (pan el A) or 15 min (panel B). Purified gDNAs were digested by restriction enzymes and then respectively separated on 2-D gels, which were blotted onto nylon membrane for probe hybridizations. The blots were respectively hybridized with -[32P] ATP labeled BAK52, -[32P] dCTP random prime labeled Igr (434 bp) from the LAT locus, and -[32P] ATP labeled HP probes. The arrows point at examined ves loci or internal controls, which were thermally remodeled and incorporated into sample preparation at the crosslinking step. B. A.

PAGE 87

87 Figure 2-25 Detection of in vivo DNA structures associated on the LAT with psoralen crosslinking (2). (A) Schematic map showi ng the recognition site of oligonucleotide LH 2 on the C9.1 LAT; (B) psoralen-cross linked C9.1 gDNAs were respectively digested by EcoRI or KpnI plus NdeI, and then loaded on DNA 2-D gel to detect the LATassociated DNA structure(s). A. B.

PAGE 88

88 Figure 2-26 Higher-order DNA stru ctures found on 2-D gels by tw o different methods. (A) C9.1 gDNA which was extracted stru cture-protectively was cut with NdeI plus EcoRI and analyzed on DNA 2-D gel; (B) C9.1 gDNA which was psoralen-crosslinked under UV 365nm was cut with NdeI plus EcoRI incompletely and analyzed on DNA 2-D gel with an internal cruciform control TR-(B0.5). Off-arc spots marked by star symbols may represent transcription-a ssociated higher-order structure. A. B.

PAGE 89

89 CHAPTER 3 CHROMATIN STRUCTURAL CHARACTERISTICS AMONG B. BOVIS ves GENES Abstract Antigenic variation m ediated by VESA 1 protein, which is encoded by the ves multigene family, is a critical strategy for B. bovis to persistent in the mammalian host. Among this multigene family, a ves 1 ves 1 gene pair is actively tran scribed from LAT while other ves gene loci appear to be silent. The mechanis m of transcriptional control over the ves multigene family is not yet understood. To investig ate whether higher-order chroma tin structure might be involved in mutually exclusive transcription of ves genes, we used nucleases to examine the chromatin configuration of ves genes. The results revealed that th e LAT was significantly more accessible to nuclease digestion than ar e transcriptionally silent ves gene pairs and that the LAT adapts a unique nuclease-sensitivity pattern quite distinct from that of similarly assayed non-expressed ves loci and active housekeeping genes. Additionally, although non-expressed ves genes seem to have similar nucleosomal positioning patterns, unexpectedly, some showed intermediate sensitivity to nucleases, suggesting the co mplexity of chromatin structures among ves multigene family. Introduction Chrom atin structure plays an important role in mediating transcriptiona l control of genes in eukaryotes. The dynamic conversion between hete rochromatin and euchromatin can facilitate genes to switch from being transcriptionally pe rmissive to being transc riptionally repressive (reviewed in Smale and Fisher 2002; Schulze and Wallrath, 2007). In lower eukaryotes, such as yeast, P. falciparum, and African trypanosomes, heterochromatin silencing modulates monoalleic expression of gene families (Borst and Chaves, 1999; Duraisingh et al., 2005; Merrick and Duraisingh, 2006; Yang et al., 2009). To maintain the state of being actively

PAGE 90

90 transcribed, gene promoter-proximal regions must be principally made accessible to the transcription machinery and numerous regulators. Consequently, gene repr ession or inactivation can be achieved by packaging genes in heterochromatin through epigenetic modifications and subsequent chromatin remodeling (Greaves and Borst, 1987; Pfeifer and Riggs, 1991; Chen and Yang, 2001; Lejeune et al., 2007). In eukaryotic cells, nuclear DNA is packaged compactly into well-organized chromatin composed of periodic arrays of nucleoprotein comple xes as the first level of structure, with the nucleosome being the basic unit (reviewed in Ramakrishnan, 1997). As an important characteristic of chromatin configuration, nucle osomal position at gene promoters and coding regions can modulate transcripti onal activities in many ways {Vega-Palas and Ferl, 1995; Chen and Yang, 2001; Rando and Chang, 2009). Variable nucleases can be used to probe chromatin structure to provide information with respect to fine regulation of gene expression (Drinkwater, 1987; Greaves and Borst, 1987; Gross and Garard 1988; Kuhnert et al., 1992), although care should be taken in preparing chromatin substrates to avoid acquiring biased information (Jin et al., 2009). To explore how the LAT becomes a privile ged transcription locus among ~24 head-tohead ves 1 ves 1 loci which seem potentially competent in expression (reviewed in Brayton et al., 2007), here we took the approa ch of using different nucleases to analyze basic chromatin structures. Since both MNase and DNase I are si ngle-stranded or double-stranded DNA specific, they preferentially cleave chromatin in the linke r regions between nucleosomal particles (Gross and Garard, 1988; Telford and Steward, 1989). In th is study, both overall chromatin structures in B. bovis and ves gene-specific nucleosomal positioning patterns were examined by partial nuclease digestion. We found that the LAT adopt ed very distinctive chromatin structure

PAGE 91

91 especially at the intergen ic reigon, and non-expressed ves genes have regular nucleosome phasing comparable with active housekeeping genes -tubulin and GADPH. The results are discussed with regard to differe ntial transcriptional status among ves genes. Methods and Materials Parasite Culture Isolation of Parasite Nuclei Parasite nuclei were isolated from C9.1-in fected red blood cell culture essentially as described (Lanzer et al ., 1992). All steps were carried ou t on ice. Briefly, C9.1-infected erythrocytes were ruptured by 0.05% saponin ly sis to release hemoglobin. Parasites were collected by centrifugation at 10000 g for 10 min at 4oC. They were then washed and resuspended in solution A (20 mM PIPES pH7.5, 15 mM NaCl, 60 mM KCl, 14 mM 3mercaptoethanol, 0.5 mM EGTA, 4 mM EDTA 0.15 mM spermine, 0.5 mM spermidine, 0.125 mM PMSF), a hypotonic medium. After addition of 6.25% volume of 10% Nonidet P-40 solution, the parasite suspension was transf erred into chilled B dounce homogenizer and disrupted with 6~8 strokes. The nuclear pellet was collected by centrifugation (7000 g, 10 min at 4oC) and used immediat ely. DAPI staining (0.5 g/ml; Invitrogen) reveal ed that the isolated nuclei appeared to maintain proper condition. Nuclease Digestion of Intact Nuclei Micococcal nuclease digestion MNase digestion was done as described (Greaves and Bors t, 1987) with modifications. The nuclear pellet was washed and resuspended in cold MNase reaction buffer (10 mM NaCl, 5 mM MgCl2, 0.1 mM CaCl2, 10 mM Tris-HCl, pH 7.4), and ali quots incubated with different concentrations of MNase (0 ~ 6 unit /ml; Fe rmentas Life Science, Glen Burnie) at 37oC for 3

PAGE 92

92 minutes. The digestion was stopped by addition of stopping buffer, giving final concentrations of 0.2% (w/v) SDS and 10 mM EDTA, 5 mM EGTA. DNase I digestion Parasite nuclear pellets were washed and re suspended in cold DNase I reaction buffer (10 m M Tris-HCl, 2.5 mM MgCl2, 10 mM CaCl2, pH 7.5 at 25C). Aliquots were incubated with increasing concentrations of DNase I (0 ~ 1 unit/ml; Fermentas Life Science) at 37oC for 10 minutes. The digestions were stopped by addition of reaction stopping buffer (10; 50 mM Tris, 150 mM NaCl, 25 mM EDTA, 0.1% SD S, pH 8.0) (Chen and Yang, 2001). S1 nuclease digestion Nuclear pellets were washed and resuspended in cold S1 nuclease reaction buffer (50 mM sodium acetate, 280 mM NaCl, 4.5 mM ZnSO4, pH 4.5 at 25C; Promega Biosciences, Madison). Aliquots were incubated with increa sing concentrations of S1 nuclease (0 ~ 500 unit/ml; Promega Biosciences) at 37oC for 30 minutes. The reactions were stopped by addition of reaction stopping mix, giving fina l concentrations of 0.2% (w/v ) SDS and 10 mM EDTA, 5 mM EGTA (Greaves and Borst, 1987). To confirm that S1 nuclease works, it wa s tested on ph(15-1) pha gemid, which has its insert of the ves gene locus 15-1 terminated by EcoRI sites. This DNA molecule thermally remodeled (TR). Then single-stranded regions in the plasmids were produced, which were sensitive to S1 (Figure 3-14). To confirm that the cutting is due to enzymatic activity and not to the acidic buffer, a buffer treatm ent control was also included. Mung bean nuclease digestion Nuclear pellets were washed and re-suspende d in cold Mung bean nuclease reaction buffer (50 m M sodium acetate, 280 mM NaCl, 4.5 mM ZnSO4, pH 5.0 at 25C; Takara). Aliquots were incubated with increasing concentrations of Mung bean nuclease (0 ~ 1440 unit/ml; Takara) at

PAGE 93

93 37oC for 30 minutes. The reactions were stopped by placing reactions on ice and the addition of reaction stopping buffer to make final concentr ations of 0.2% (w/v) SDS, NaCl and 10 mM EDTA, 5 mM EGTA (Fryer and Ar cher, 2001; Gromova et al., 1995). To confirm the activity of Mung Bean nucleas e, it was tested on pL3 plasmid, which has the insert of B0.5 fragment and was thermally re modeled to form single-stranded regions in the plasmids (Figure 3-14). To confirm that the cutting is specific to enzymatic activity and not to the acidic buffer, a buffer treatm ent control was also included. Southern blot analysis DNA was purified, digested with restriction enzym es and lo aded on 1.2% agarose gel, which was blotted onto Hybond N+ nylon membrane (GE healthca re) and hybridized with a series [32P]-ATP or -[32P]-dCTP labeled probes. The labeling methods and the procedure of probe hybridization was the same as describe d (methods and material s; chapter one). Results Nucleases have proved to be very useful tools in the study of chrom atin structures (Greaves and Borst, 1987; Zhang and Reese, 2006). They can be em ployed to obtain a variety of fundamental information on chromatin configurat ion according to their template specificities. Among the available nucleases, MNase, DNase I, M ung bean nuclease, and S1 were used in this study to probe chromatin stru ctural characteristics at the LAT and several non-expressed ves loci. It was hypothesized that differential chromatin structures are adopted by the LAT and nonexpressed ves gene loci based upon their differential tr anscriptional status. Due to frequent transcription, active genes tend to have more open chromatin structures, and transcription bubbles could appear throughout whol e gene regions. Hence, active genes are more sensitive to nucleases, which preferentially cut naked doubl e-stranded or singlestranded DNA regions (Wang and Landick, 1997; Kuhnert et al., 1992). Therefore, MNas e and DNase I were used in

PAGE 94

94 this study to evaluate the overall chromatin structures among ves gene loci, whereas, Mung bean nuclease and S1 nuclease, which target sing le-stranded DNA templates, could detect ssDNA formed within any higher-order DNA structure associated with the ves loci. The potential for structural reorganization of LA T IGr during transcription could make it a better target for Mung bean or S1 nuclease digestion. MNase Mapping At lower concentrations, MNase typically cuts DNA pri miarily within linker regions between nucleosomes, which makes it very helpful in examining nucleosome-positioning patterns on specific genes or gene regions of in terest (Zaret, 2005; Telf ord and Stewart, 1989). Since the integrity of the isolat ed nuclei is crucial to acquire accurate information on chromatin structure, DAPI staining was used to check the condition of nuclei after is olation (Figure 3-1). Low-resolution MNase mapping employing part ial digestion only was first used to examine general chromatin configuration of B.bovis After partial MNas e digestion, n-mer nucleosomes were released to generate a regul ar ladder of low MW bands, forming a regular pattern (Figure 3-1), which was then used to estimate bulk NRL (Mason and Mellor, 1997). B. bovis bulk NRL was estimated to be around 156 bp by pl otting numbers of nucleosomes in n-mer nucleosomal bands against their estimated sizes a nd then calculating the linear slope. The short nucleosomal repeat length suggests that B. bovis may not have a canonical histone H1, similar to other lower eukaryotes like P. falciparum and Schizosaccharomyces pombe (Thomas and Furber, 1976; Lanzer et al., 1994; Mason and Mellor, 1997) Failure to find histone H1 homologues in B. bovis genome though BLASTP and Psi-Blast se arches supported this possibility. Gene-specific nucleosomal positioning pattern s were then examined by probing the MNase cutting pattern with radioactively labeled probes positioned near some restri ction sites, serving as an indirect end label. Long lo cus-specific DNA probes are not av ailable due to severe cross-

PAGE 95

95 hybridization and oligonucleotide probes are also difficult to desi gn because of frequent patches of homology shared among markers of the ves multigene family. Therefore, probe availability limited the examination of gene-specifi c nucleosomal arrays on many different ves gene loci. However, since many ves gene loci and the LAT have highly conserved IGrs (Al-Khedery and Allred, 2006), it is useful to use the whole IGr sequence as a DNA probe to check the general nucleosomal organization pattern on ves genes (Figure 3-2). A regul ar NRL pattern with an average step about 157 1 bp was observed at the lower MW part of n-mer DNA ladder generated from partial MNase trea tment, corresponding to the chromatin regions in or near the conserved IGrs of ves gene loci. The higher MW part of DNA ladder was irregular and less informative (Figure 3-2). Since all ves genes except the LAT are silent in the C9.1 clonal line, the observed regular pattern was predominantly contributed by those IGr-containing, inactive ves loci. The result revealed average nucleosoma l positioning pattern around IGrs in those IGrcontaining ves loci. Oligonucleotide BAK6, which target s the center of conserved intergenic regions and recognizes at least 3 non-expressed ves loci in addition to the LAT, also detected orderly nucleosomal orga nization (Figure 3-3). Despite difficulties in probe generation, oligonucle otide probes suitable for use, as indirect end labels are available for two ves genes of known sequences: 101 (probe: BAK68) and 15-1 (probe: LH2). The housekeeping genes including GADPH and -tubulin were used to show the nucleosome organization of active genes. Both those active genes and ves loci 10-1 and 15-1 all had regular n-mer nucleosome DNA ladders, fo rming good linear regression trendlines when plotted against nucleosome number (Figure 3-3, 4). Two probes were generated whic h are specific to the LAT (LH 1, LH 2). To validate their specificity, regular southern blot analysis was carried out (Figure 3-6, 7) and the results

PAGE 96

96 suggested that LH 2 was specific with the recognition only the LAT. Although LH 1 recognized only the LAT with EcoRI or PvuII di gestions, it hybridized with a second band of about 2.3 kb in addition to the LAT when KpnI digestion alone was performed. Since the second band is smaller than the LAT, it could not be cau sed by incomplete restriction digestion (Figure 3-6, 7). When the hybridization te mperature was raised from 59oC to 61oC, LH 1 still recognized two ves loci. If its recognition of the ves locus other than the LAT was caused by cross hybridization of im perfect target sequence with seve ral nucleotide differences, higher hybridization temperature and mo re stringent washing conditions should erase the non-specific signal. It is unknown why LH 1 only hybridized with the LAT on a blot prepared earily (Feb. 2009) and then recognized with a second locus on blots preapared th ereafter (June 2009) under similar conditions (Figure 3-6, 7, 8). However, the nucleosomal positioning patterns detected by LH 1 and LH 2 were essentially identical around the LAT IGr (Figure 3-8). Since LH 2 hybridizes more closely to the KpnI site than does LH 1 (Figure 3-8), it is mo re appropriate than LH 1 in mapping MNase hypersensitive (MH) bands shown in the LAT interg enic region. In contrast to the 3 nucleosomal bands observed from silent ves loci, 5 significant MH sites were distributed asymmetrically in the LAT IGr, flanked by two short protected regions, which corresponded to the 5 coding regions of the ves 1 and ves 1 genes (Figure 3-8). It is qu ite clear that the LAT adopts a chromatin organization easily distinguished nucleosomal organizati on from non-expressed ves gene loci, and which probably correlates with its active transcription status. A second approach for evaluating the overall chromatin c onfiguration of ves genes is to quantify locus-specific MNase se nsitivity. To do this, different concentrations of MNase (0.0 6.0 unit/ml) were used to treat isolated nuclei for 10 min at 37oC to achieve progressive partial

PAGE 97

97 digestion. The DNA was then purified and digested to completion with EcoRI or KpnI. Equal amounts of DNA were separated in 1.2% agarose gels and transf erred onto nylon membrane for a series of hybridizations with loci-specific probes. The fade-out rates of hybridization signals from various gene loci were then determined (Fi gure 3-9; Table 3-1). The intensities of all genespecific bands were determined (ImageJ) and plotte d as the proportion of in itial signal lost as a function of MNase concentration (unit/ml). Since the examined gene loci are variable in size and longer fragments theoretically have a greater ch ance to be cut by MNase, it is necessary to correct the fade-out rates for ba nd size (Table 3-1), as the propor tion of initial signal loss per kb. It was found that the LAT reproducibly is by fa r the most sensitive to MNase when compared with other ves gene loci, and remarkably is ev en more sensitive than GADPH or -tubulin (Figure 3-10). The high sensitivity of the LA T to MNase is consistent with its active transcriptional status In contrast, the ves locus 10-1 is reproducibly the least sensitive, whereas the two non-expressed ves loci S621 and 15-1 were reproducibly intermediate in sensitivity. It is intriguing that S621, 10-1, and 15-1 all belong to inactive ves genes and have similar nucleosomal organization (Figure 3-3), yet S621 and 15-1 are significantly more sensitive to MNase digestion than 10-1. DNase I Digestion Exam ination of the differential sensitivity of ves gene loci to DNase I digestion was carried out in a similar fashion as done for MNase. In contrast with MNase, DNase I can cut chromatin with very high efficiency; even at a low concentration (i.e. 0.5 unit/ml) most of the gene loci were digested completely (Figure 3-11). Remarkably, the results are qualitatively consistent with the MNase sensitivity assay in that the LAT is the most sensitive and 10-1 locus the least sensitive to DNase I (Figure 3-12). However, S621 and 15-1 seem more sensitive to DNase I than they were to MNase. This may be because DNase I cannot only cut internucleosomal DNA

PAGE 98

98 but also nucleosomal DNA, especially at AT -rich regions (Chen and Yang, 2001). Thus, DNase I is somewhat DNA sequence-dependent in additio n to nucleosomal organization-dependent. An attempt to correct for AT content was done by calculating proportion of initial signal loss per AT% (Figure 3-12; Table 3-1). After the correcti on, the overall digestion sensitivity pattern of DNase I, which refers to the differential accessibility of gene loci to DNase I, is similar to that of MNase (Figure 3-10, 12). Additionally, it is concei vable that treatment w ith a low concentration of DNase I treatment (0.1 unit/ml) results in a preferential attack of internucleosomal DNA, whereas at higher concentration (0.4 unit/ml) DNase I tended to have more chance to attack nucleosomal DNA, and therefore greater depende nce on AT content of targeted regions. Consequently, the sensitivity-plo tting pattern detected around low DNase I concentration (Figure 3-12; boxed area) is quite consistent with the sensitivity-plotting pattern of MNase. Active genes can be distinguished from inac tive genes by having some single-stranded regions during the formation of transcription-related bubbles (M ilne et al., 2000). S1 nuclease often can be used to support the data acquired by MNase or DNase cleavage mapping. However, no significant cutting was found with any examined gene locus even with very high S1 concentration (2000 unit/ml). To clarify that th e enzyme was still functional and had sufficient activity, the plasmid ph(15-1), which c ontained the EcoRI fragment of the ves 15-1 locus including the internal intergen ic region, was thermally remodeled and treated by S1. The results showed that S1 efficiently cut the novel form (s) after TR in 10 min (Figure 3-13). The bufferonly control suggested that the cutting was caused by S1, not the acidic reaction buffer (pH 4.6). Similarly, no significant cutting was found with any examined gene locus even with very high concentrations (400 unit/ml) of M ung bean nuclease. To ensure that the enzyme was functional and possessed sufficient activity, the plasmid pL3, which contained the cruciform control

PAGE 99

99 molecule with the short internal perfect invert ed repeat, was thermally remodeled and treated by Mung bean nuclease. Again, the data demonstrat ed that Mung bean nuclease cut the novel form(s) after TR in 10 min (Figure 3-14), and th e buffer-only control conf irmed that the cutting was caused by Mung bean nuclease, not the acidic reaction buffer (pH 5.0). The reasons for the failure of S1 and mung bean nucleases to cut B. bovis chromatin may have to do with low pH, which may cause contraction of nucle ar envelope, preventing access since B.bovis nuclei are extremely tiny in sizes compared wi th that of higher eukaryotes. Discussion Even if the form ation of high order structur e might assist the LAT transcription, other mechanisms may essentially ex ist to repress non-expressed ves loci since it is not understood why other non-expressed ves loci cannot form high order st ructure despite having similar potential of structure formation. In addition, episomally-introduced ves conserved intergenic regions subcloned from some non-expressed ves loci showed potent prom oter activity in driving marker genes such as luciferase (Wang, X a nd Allred, DR; unpublished data), which suggested that they might be freed of the restraint of chromatin structure and became capable of forming structures to assi st transcription. One common mechanism shared by Leishmania major Trypanosoma brucei and P. falciparum is that the silencing of their variant multigene families is partly achieved through being packaged into heterochro matin (Merrick and Duraisingh, 2006). Since so many traits are shared between the two intraerythrocytic parasites, B. bovis and P. falciparum it would be valuable to examine the chromatin organization of ves genes to explore if similar mechanism is employed by B. bovis MNase and DNase I were then used to probe chromatin structures, and it was found that the LAT was clearly distinguished from non-expressed ves loci by differential chromatin structures. Several aspe cts of this study are discussed.

PAGE 100

100 Generally, the first level of chromatin organi zation in higher eukaryotes is the nucleosome filament (reviewed in Ramakrishnan, 1997). However, exceptions do exist in lower eukaryotes such as Gyrodinium cohnii which lacks nucleosomes (Bodans ky et al., 1979). Another example is Entamoeba histolytica, which seems to lack typical histones, and also will not generate regular n-mer nucleosomal ladders when digested by MNase (Torres-Guerrero et al., 1991). Little is known about B. bovis chromatin organization and respective histones. Since chromatin structure has been shown to be important in the multilayer modulation and harmonization of gene expression in many organi sms (Felsenfeld et al., 1996; Agarwal and Rao, 1998; Merrick and Duraisingh, 2006), it would be necessary to determine if canonical nucleosomal organization exists in B. bovis. BLAST searches have revealed that B. bovis has putative histones H2A/B, H3 and H4, which are quite conserved compared with the histone genes of P. falciparum and many other organisms (AppendixB). After partial MNase digestion, regular nucleosomal ladders were visible in agarose gels with the estimated nucleosomal repeat length around 156-157 bp. These data strongly indicated that B. bovis bulk chromatin has a wellordered nucleosomal framework. The short nuc leosomal repeat length suggested that B. bovis like P. falciparum which possesss a nucleosome periodicity of approximately 155 bp (Cary et al., 1994; Lanzer et al., 1994), may lack a canonical H1 histone. Attempts to identify H1 sequences within the B. bovis genome by bioinformatic searching were unsuccessful, consistent with this possibility. Nucleosomal Positioning Pattern among Ves Genes MNase can only cut DNA linker regions non-spec ifically between nucleosomes although it can also nick nucleosomal DNA, which make s it very helpful in examining nucleosomepositioning patterns on specific genes or gene re gions of interest (Z aret, 2005; Telford and Stewart, 1989). To examine the ch romatin compactness on non-expressed ves loci, 15-1 and 10-1

PAGE 101

101 were chosen for study. GADPH and -tubulin were also chosen to show chromatin structural characteristics of transcriptionally activ e genes. Surprisingly, both the non-expressed ves loci and the actively expressed housekeeping genes are organized into orderly nucl eosomal arrays with similar periodicity (Figure 3-3) The regular pattern detected by the DNA probe generated from the conserved ves intergenic reigon suppor ted that non-expressed ves IGr-containing loci may be bound by orderly, phased nucleosome particles (Fi gure 3-2). This means that non-expressed ves loci may not locate in typical constitutive heterochromatin, which is quite inaccessible to nucleases compared to facultative heterochroma tin (Sun et al., 2001; Sperling et al., 1985). Another possibility is that Ba besia heterochromatin may not be organized as in other higher eukaryotes. The LAT-specific nucleosome-positioning patte rn was obtained through the hybridization with oligonucleotide LH 2 (Figure 3-8), to compare to that of non-expressed ves loci. Considering the active transcription at the LAT, it was expected that the nucleosomal pattern would be similar to that of the examined housek eeping genes (Figure 3-3). Surprisingly, the LAT was found to have an unusual nucleosomal pattern featuring 5 MH sites at the IGr and short MNase-inaccessible regions which flank the IGr (F igure 3-8). The different chromatin structure found at the LAT correlates with its active transcription. Availabl e evidence suggested that the IGrs of the LAT and some non-expressed ves loci such as S621 and 1E10 all had potent promoter activity (Wang, X and Alllred, DR; unpub lished data). According to the alignment of the ves IGrs (Al-Khedery and Allred, 2006) a pair of imperfect invert ed repeats is significantly conserved and could be important elements fo r promoter activity. However, the asymmetric locations of MH sites shown in the LAT IG r (Figure 3-8) sugges ted that the apposing ves 1 and ves 1 at the LAT may not equivalently manipulate the IGr to facilitate their respective

PAGE 102

102 transcription. According to the comparison bewteen LH 2and BAK52specific nucleosomal patterns, it is apparent that the regular pattern detected by B AK52 was mostly contributed by the non-expressed ves donor loci (Figure 3-5, 8). It is wort h noting that MH sites at the LAT IGr were found coincidently between those imperfect i nverted repeats with sh ort intervals. It was also reported in T. brucei and P. falciparum that active genes and silent genes share similar nucleosomal periodicity (Greaves and Borst, 1987; Lanzer et al., 1994) About 24 gene loci scattered in the B. bovis genome share similar gene organi zation and conserved intergenic regions with the LAT (Brayton et al., 2007). The regular nucleosomal pattern observed around the ves IGrs suggested they have si milar chromatin structure at th e nucleosomal level, which is obviously different from that of the LAT (Fi gure 3-3, 8). The finding that similar chromatin structures were observed among variable non-expressed ves IGrs within different chromosome contexts suggests that epigenetic regulation may be involved to coor dinately repress them (Halme et al., 2004). Nuclease Sensitivity Analysis Since MNase sensitivity of a specific gene locus is pr oportional to its length, size correction h as been done on the plotting of th e MNase sensitivity by calculating the proportion of initial signal loss per kb (Figure 3-10). However, the results from MN ase sensitivity assays showed that, under all conditions, the LAT was the most sensitive gene locus (Figure 3-12), consistent with its active transcription status. Two ves loci, S621 and 15-1, are intermediate in sensitivity compared with the LAT and 10-1 lo cus. As it is known that 15-1 and 10-1 have similar nucleosomal patterns (Figure 3-3) and neit her is transcribed (Zupan ska et al., 2009), it is unexpected that 15-1 has higher accessi bility to MNase (Figure 3-11).

PAGE 103

103 Similarly to MNase digestion, an alysis of DNase I digestion pa tterns revealed that the LAT was again the most sensitive overall. Since DNase I can cut not only the linker region but also the nucleosomal DNA, especially within pyrimidine-rich regions, it is more DNA sequencedependent. Therefore, the sensitiv ity plotting needs correction not only for target size, but also the AT content of examined loci. However, since DNase I can only cut nucleosomal DNA at the minor grooves where it can access easily (Chen a nd Yang, 2001), the accessibility of DNase I to AT-rich nucleosomal DNA is dependent upon how it is wrapped on the nucleosomes. Therefore, even if correction of AT cont ent is needed to get more accu rate comparisons of DNase I accessibility between gene loci, it can not be simply done. Without correction for AT%, S621, 10-1 and 15-1 have comparable sensitivity to DNa se I, which seems inconsistent with MNase sensitivity results (Figure 3-10, 12). However, S621 and 15-1 all have lower AT% than 10-1 (Table 2-3), therefore they can simply be less sensitive due to a lower AT%. With the correction of AT%, the overall sensitivity pattern is similar to that of MNase, suggesting the importance of AT% correction to overcome the influence of DNA sequence on DNase I sensitivity. However, it is not certain if the correction of AT content is appropriate when calcu lating the proportion of initial signal loss per AT%, as it is not known how AT-rich regions are wrapped into nucleosomes. Mung bean nuclease and S1 nuclease specificall y cut single-stranded DNA and were also tried to provide information on chromatin structur e. However, they did not show any discernible digestion pattern, even on ac tive housekeeping genes such as GADPH (Figure 3-32, Figure 333), and so there is a concern that the accessibilit y of those enzymes to nuclear chromatin may be problematic in our DNA of study. Some efforts had been taken to make nuclei suspension less viscous by passing the nuclei through an insulin gauge needle several times. Subsequent DAPI

PAGE 104

104 staining showed that nuclei still maintained good shape and became freer in solution; however, there was still no digestion shown at the active ge ne loci. It was possible that the S1 or Mung bean nuclease reaction buffer was too acidic (pH 4.6-5.0) and the nuclear lamina may contract, preventing enzymes from getting contact with the chromatin. Though they did not work to provide information from chromatin, MNase and DNase I were shown capable of cutting singlestranded DNA (Figure 3-14, 15). The reasons they could not digest B. bovis chromatin substrate are not clear. It is conceivable that non-expressed ves loci such as S621 and 15-1, which are intermediate in sensitivity, adopt relatively ope n chromatin structure, yet are not detectable at the protein level. Several possible reasons may explain thes e results. First, they may have stalled RNA polymerase II poising at the promoter-proximal region to make nucleases more accessible. RNA polymerase II pausing has been demonstrated to be widespread at those genes ready to cope with environmental or developmental changes (Mus e et al., 2007; Zeitlinge r et al., 2007; Wu and Snyder, 2008). RNA polymerase II is recruited to th e promoters of such genes and stalls after a short elongation (Muse et al., 2007; Wu and Snyd er, 2008). Considering the periodic antigenic variation occurring in B. bovis -infected cattle, it is possibl e that some non-expressed ves loci are in a state of readiness in order to respond to certain stimuli aris ing from the interaction of host and parasite. Hence, the parasites could prom ptly manipulate gene expression and thereby generate new phenotypes to evade host immunity. Second, the non-expressed ves loci may have leaky transcription. When T. brucei is at the blood stage, some telomere-linked monocistronic VSG expression sites were found not to be fully s ilenced instead were tran scribed at a very low level, producing rare VSG transcripts. In contrast, the active vsg gene was predominantly expressed at a high level (Ala rcon et al., 1999). A similar situ ation could possibly exist in B.

PAGE 105

105 bovis, since previous work in popul ation RT-PCR has found rare tr anscripts which appeared to arise from non-LAT ves loci (Zupanska et al., 2009). Third, they may be situated at the boundary between euchromatin and heterochromatin and accordingly be affected by neighboring active genes to have relatively open chroma tin structures (Sun et al., 2001). In summary, the results obtained so far from the study of the LAT and non-expressed ves loci at structural and biological activity levels collectively suggest that higher order chromatin structure is associated with their differe ntial transcription status. Additionally, the hypersensitivity of the LAT to MNase might be due to some transiently formed higher order DNA structure at the LAT IGr although those struct ures were elusive to be significantly and repeatably detected. Therefore, future study in other aspects, such as epigenetic regulation, may provide better explanation on how the LAT achieve s its privileged active transcriptional state while other ves loci, however potentially capable, are ma intained in the inactive state. Bioinformatics results showed that B. bovis has highly conserved histones (Appendix-B) and enzymatic machinery for histone modifications (Appendix-A). A series of modified forms of antibodies to histone H3 has been validated in fixed immunofluores cence assays (IFA); therefore, it is feasible to apply them in the study of H3 modifications on ves 1 genes (AppendixC). Apicidin has been shown to have broad in vi tro antiprotozoal activity in inhibiting histone deacetylases and preventing proliferation of several Apicomplexan parasites, which include Cryptosporidium parvum, Toxoplasma godii, Eimeri a tenella and the Plasmodium species etc (Darkin-Rattray et al., 1996). So far anti-babesia activity of Apicidin has not been demonstrated and how it can be related to the inhibition of histone deacetylases. The growth inhibition of Apicidin on B.bovis has been examined and its IC50 was estimated to be 464 nm (Huang, Y and

PAGE 106

106 Allred, DR; unpublished data). That informati on could be useful in studying epigenetical regulation of the expression of ves multigene family. Preliminary results with chromatin immunopr ecipitation through anti bodies to modified forms of histone H3 also suppor ted the notion that non-expressed ves genes may have different epigenetic markers from activ e housekeeping genes GADPH and -tubulin (Huang, Y and Allred, DR; unpublished data). Therefore, further investigation of epigen etic marks on ves genes may provide satisfactory answers for the mutually exlusive transcripti onal control over the ves multigene family.

PAGE 107

107 Figure 3-1. B. bovis bulk nucleosome organization (A) Lo w-resolution MNase cleavage mapping of ves gene loci. B.bovis C9.1 IRBCs were collected and filtered through CF-11 assembled column to get rid of leucocytes. Then 0.05% saponin lysis was used to release parasites. The nuclei were isol ated by homogenization and then divided equally into five portions, being treated w ith different concentrations of MNase, 37oC, 3min. After traditional DNA extraction, DNA was digested by EcoRI or KpnI and loaded with equal amount on 1.2% agaros e gel to separate. The gel was stained with 0.5 g/ml EtBr and photographed under UV. (B) Isolated nuclei of C9.1 strain was stained with DAPI. (C) Estimation of bulk nucleosome repeat length. Band sizes (2-7) observed from the nuclei digestion (6 unit/ ml) were plotted. The slope was determined to be 156 bp per band, representing bulk nucleosome repeat length. C. B. A.

PAGE 108

108 Figure 3-2. Examination of the overall nucleosomal organization of ves genes containing intergenic regions.

PAGE 109

109 Figure 3-3. Gene-specific nucleosomal pos itioning pattern detected by MNase. B. bovis nuclei were digested by MNase partially. Puri fied DNA fragments were digested by restriction enzymes, separated on 1.2% agarose gel and then blotted onto nylon membrane for hybridizations. The -[32P] ATP end-labeled oligonucleotide probes used to detect gene-specific nucleosomal organizations were: (from left to right) BAK6, LH2, BAK68, AZT02 and GADPHE1R The blue bar shows approximately the location of the intergenic region of 10-1 or 15-1 loci.

PAGE 110

110 Figure 3-4. Analysis of gene-specific nucleosomal positioning pattern detected by MNase. (A) Schematic map of examined ves loci and active housekeeping genes. The specific probes are shown on the top of their recognition sites; (B) Analysis of gene-specific NRLs by plotting n-mer nucleosomal band si zes against the nucleosome numbers. A. B.

PAGE 111

111 Figure 3-5. LAT-associated nucleosomal posit ioning pattern detected by BAK52. (A) BAK52specific nucleosomal organization. The blue bar shows approximately the location of the intergenic region of the LAT; (B) Schematic map of examined ves loci and active housekeeping genes. The green bar in panel B represents BAK6 recognition site. A. B.

PAGE 112

112 Figure 3-6. Examination of LAT-sp ecific probes. (A) Schematic ma p of indirect end labeling of LAT fragments with o ligonucleotides (BAK52, LH 1 and LH 2). (B) Initial validation of LH 1 and LH 2 on the blots used in MNase sensitivity analysis. A. B.

PAGE 113

113 Figure 3-7. The confirmation of LAT-specific probes through regular southern blot analysis. Purified C9.1 gDNA was digested by singl e restriction enzyme (EcoRI, KpnI or PvuII) and then electrophoresed on 0.8% agarose gel. The gel was blotted onto nylon membrane and hybridized with oligonucleotide probes LH 1 and LH 2. The red arrow pointed at a LH 1-recognized ves locus other than the LAT.

PAGE 114

114 Figure 3-8. C9.1-LAT specific nucleosomal positioning pattern detected by MNase. (A) Nucleosome organization on the C9.1 LAT. The thick green and purple bars represent estimated location of the LAT intergenic region. (B) Schematic map of indirect end labeling of the LAT by oligonucleotide probes LH 1 or LH 2. (C) Schematic map of MH sites distributed in the LAT IGr. A. B. ves1 ves1 MH1 MH2 MH3 MH4 MH5 bpGGTGGAATCC GCGCGCCC.

PAGE 115

115 Figure 3-9. Examination of diffe rential MNase sensitivity among ves genes. (A) Schematic Diagram showing examine ves loci and active housekeeping genes. (B) isolated nuclei were trea ted with different concentrations of MNase (0~6 unit/ml) at 37oC, 3 min. Purified DNAs were run on 1.2% agarose gel and blotted onto nylon membrane for hybridization with -[32P] ATP labeled BAK6. The arrows represent different ves loci.B. A.

PAGE 116

116 Figure 3-10. Analysis of differential MNase sensitivity among ves genes. (A) Southern blot analysis to show the fa de-out of full sizes of ves loci upon different concentrations of MNase treatment (Figure 3-24). The -[32P] ATP end-labeled oligonucleotide probes used to detect different loci were B AK6 (S621, 15-1, 10-1 and the LAT) and AZT02 ( -tubulin); (B) Differential MNase sensitiv ity showed by plotti ng the proportion of initial signal loss per kb against MNase concentrations. B. A.

PAGE 117

117 Figure 311. Low resolution DNase I cleavage mapping. (A) Equal amount of nuclei were treated with different concen trations of DNase I at 37oC for 10 min. Purified DNA fragments were digested by restriction enzymes, run on 1.2% agarose gel; (B) The blot was prepared form the gel (A ) and hybridized with a series of -[32P] ATP endlabeled probes: BAK6 (S621, 15-1, 10-1 and LAT), AZT02 ( -tubulin gene) and GADPHE1R (GADPH gene). A. B.

PAGE 118

118 Figure 3-12. Analysis of differential DNase I sensitivity among ves genes. Southern blot analysis to show the progressive loss of full size bands derived from ves loci upon different concentrations of DNase I treatment (Figur e 3-11). Differential DNase I sensitivity was showed by plotting the proportion of initial signal loss per kb against MNase concentrations with or without the co rrection of AT content (Table 3-2).

PAGE 119

119 Table 3-1. The sizes and AT content of examined gene loci in nuclease mapping. Gene loci / KpnI Size (kb) [AT] content (%) ves locus S621 4.8 51.8 ves locus 15-1 4.6 46.8 ves LAT 2.9 48.6 ves 10-1 1.7 54.8 GADPH 5.1 60.2 -tubulin 5.4 58.2

PAGE 120

120 Figure 3-13. Low resolution of S1 cleavage ma pping. (A) Equal amount of nuclei were treated with different concentrations of S1 (0~500 unit/ml, pH 4.6) at 37oC for 30 min. Purified DNA fragments were digested by restriction enzymes, run on 1.2% agarose gel; (B) The blot was prepared form the gel (A) and hybridized with a series of -[32P] ATP end-labeled probes: BAK6 (1 -1, 10-1), BAK52 ( LAT), AZT02 ( -tubulin) and GADPHE1R (GADPH). A. B.

PAGE 121

121 Figure 3-14. Low resolution of Mung bean nuclease cleavage mapping. (A) Equal amount of nuclei were treated with di fferent concentrations of Mung bean nuclease (0~1440 unit/ml, pH 5.0) at 37oC for 30 min. Purified DNA fr agments were digested by restriction enzymes, run on 1.2% agarose gel; (B) The blot was prepared form the gel (A) and hybridized w ith a series of -[32P] ATP end-labeled probes: BAK6 (S621, 151, 10-1, LAT), AZT02 ( -tubulin) and GADPHE1R (GADPH). A. B.

PAGE 122

122 Figure 3-15. Test of the activ ity of S1 nuclease and Mung b ean nuclease. (A) Thermally remodeled phagemid p(15-1) was treated with 1 unit of S1 (Promega) or S1 buffer only at 37oC for 10 min to test the activity of S1 on single-stranded template. The right panel showed the structur al potential of the plasmid with the insert containing a palindromic region. (B) Thermally remode led pL3 was treated with 1unit of Mung bean nuclease (Takara) or Mung bean nuclease buffer only at 37oC for 10 min to test the activity of Mung bean nuclease on si ngle-stranded template. The right panel showed the structural potential of the pl asmid with the insert containing a short inverted repeat. A. B.

PAGE 123

123 CHAPTER 4 SUMMARY Antigenic variation exists as a crucial su rvival strategy for m any mammalian pathogens such as Neisseria gonorrhoeae, Borrelia spp., Plasmodium spp., Babesia spp., African trypanosomes etc (Barbour and Restrepto, 2000). It helps those pathogens to modify their surface coats antigenically to evade host immune system; therefore, the pathogens can persist in the host and consequently the diseases caused by those organisms become refract ory (Krause et al., 2008; Figueiredo et al., 2009). The variant gene familie s involved in antigenic variation are usually under monoallelic or mutually exclusive transc riptional contro l (Merrick and Duraisingh, 2006; Figueiredo et al., 2009). It will greatly help the control of those pathogens by studying the mechanisms of transcriptional control over those virulence gene families and therefore preventing antigenic variation. Several models of monoallelic transcriptional have been prop osed in different organisms (Figueiredo et al., 2009). Firstly, the active allele is exclusively associated with a specialized compartment in nucleus to facil itate its transcription, such as vsg genes in African trypanosomes (Navarro and Gull, 2001). Secondly, trans or cis regulatory elements ma y help decide which allele can be activated, such as odor recepto r genes in mammalian olfactory neuron cells (Lomvardas et al., 2006 ; Nguven et al., 2007). Thirdly, chromatin structures specified by histone modifications can help maintain the established transcriptional control pattern of gene families and numerous chromatin remodeling factors have been characterized that are thought to be involved in gene state mainatence (Freitas-Jun ior et al., 2005; Merrick and Duraisingh, 2006; Bartova et al., 2008; Figueiredo et al., 2009). Fourthly, promoter activity at competent gene loci can be restrained by cis -linked elements, such as var gene introns in P.falciparum (Frank et al., 2006).

PAGE 124

124 According to those proposed models so far, it is conceivable if some of them are used by B.bovis to achieve mutually exclusiv e transcriptional control over ves gene family which includes more than 150 members (Brayton et al., 2007). In P. falciparum PfSir2 protein (homolog of yeast silent information regulator 2) has been iden tified and demonstrated to be involved in the repressi on of subtelomeric var genes. This repression occurs through mediating epigenetic modifications, thereby packaging th em into conditional heterochromatin (FreitasJunior et al., 2005; Duraisingh et al., 2005). It is possible that B. bovis may in part repress its silent ves genes through similar mechanisms. Consistent with this possibility, BLAST searches suggested the existence of a putative Sir2 homologue in B. bovis genome. Some subtelomeric ves genes may be silenced by Sir2-induced chromatin remodeling events, although a number of the ves genes are not located in or near telomeresor centromeres, suggesting the existence of other silencing mechanisms (Brayton et al., 2007). So far, a search for trans-acting enhancers hypothesized to be specifically associated with LAT active transcription has not been successful (Bouchut, A and Allred, DR, unpublished data). Ho wever, some preliminary data showed the possibility that ves intron1 may play such a role (Wa ng, X and Allred, DR, unpublished data). It is notable that 24 ves loci are organized similarly to th e LAT, and thus appear competent to express in situ although they appear to be silent (Brayton et al., 2007). Furthermore, some ves loci, such as 1E10 and S621, have already been shown to possess a potent bidirectional promoter (Wang, X and Allred, DR, unpublished data). No e xplanation is currently available for their repression status. The de monstrated LAT is a ves gene locus that is actively expressed in the C9.1 clonal line, although a rare cDNA transcri pt not related to the LAT has been found (Zupanska et al., 2009). Therefore, it can not be ruled out that ther e exists some leaky or very low level of transcription at other ves gene loci, which may be in an inducible state, ready to

PAGE 125

125 become dominant in response to certain circum stances or stimuli. In African trypanosomes, leaky transcription was suggest ed to be associated with st age change (Alarcon et al., 1999). Additionally, recent progressed in the study of var gene transcriptional control suggested the strategy that P.falciparum tends to express all var genes at a very low level at the initation of the asexual cycle (Wang et al., 2009). B. bovis may have a similar mechanism for ves genes and the transcriptional level of ves genes other than the LAT may be ju st too low to be easily detected. To explore the possible mechanism(s) of ves gene transcriptional control, examinations on chromatin structures of ves genes were carried out. Two differe nt approaches were used based on distinctive methodologies. The first approach employed invol ved the direct detection of in vivo higher order DNA structure in chromatin. This approach utili zed an established DNA 2-D gel system combined with two types of in vivo DNA structure-preservi ng methods. This is to find out if the LAT putative promoter, which shares two pairs of cons erved quasi-palindromic segments with that of many other ves gene loci (Al-khedery and Allred, 2006), adopts unusual configuration in adaption to bidirectional active transcriptions of ves 1 and ves 1 This hypotheis was supported by the sequence specificity of the ves IGrs and the observations that the ves gene fragments containing the conserved IGrs had great potent ial to form relatively stable cruciform-like structures in vitro In vivo structure detection was then ca rried out. Certain higher order DNA structures were found, using a 2-D gel techni que, in chromatin obtained by two different structure-preserving methods. These structures were most likely found in the LAT due to its distinctive transcription status. However, further evidence is needed to draw a conclusion due to problems in data repeatability and the inherent detection limits of the 2-D gel analysis. The difficulty in B.bovis cell culture synchroni zation made it impossible to learn when ves genes start

PAGE 126

126 to be actively transcribed in the asexual life cy cle and how long the active transcription lasts. Moreover, the enrichment of parasites undergoing active ves transcription assisted by the possible high order DNA structure in chromatin at the promoter region cannot be done simply as it is with yeast. It is possible with yeast to enrich for DNA recombination intermediates or replication intermediates by culture synchroni zation (Pearson et al., 1996; Aller and Litchen, 2001). Therefore, the off-arc signals found by 2-D gel analysis were very low in intensity and not conclusive enough. With the help of other approaches, such as in vitro transcription or transient transfection of B. bovis with in vitro -formed structured LAT promot er driving a reporter gene, there is some promise of supporting our observa tions obtained by2-D gel analysis. The unusually high sensitivity of the LAT IGr region to MNas e digestion and the MNase hypersensitive site pattern also supported the possi ble existence of high order D NA structure related to LAT transcription. For the first time, we tried to propose a novel and testable model in the transcriptional control of a gene family: th e specific association of unusual promoter configuration to active allele, although further evidence needs to be pursued.. However, one question would be why those other silent ves genes with such structure-forming potential in their putative promoter regions can not form this tr anscription-related DNA st ructures in chromatin. The possible explanation may be found in their n on-permissive chromatin structures specified by silent epigenetic markers, which was supporte d by preliminary chromatin immunoprecipitation data (not shown). The second approach taken to identify structure involved confirming that bulk B. bovis chromatin showed characteristics of canonical nu cleosome arrays. Subsequently, differential chromatin structures were revealed between the LAT and inactive ves loci. This was not only supported by their differential nucleosomal organizations probed by MNase, but also their

PAGE 127

127 overall chromatin compactness evaluated by nuc lease accessibility. Not surprisingly, the LAT has a very unusual nucleosome positioning pattern qu ite distinctive from that of other examined ves genes, consistent with its active transcription status and relatively open chromatin configuration to facilitate the recruitment of tr anscriptionary machinery. It was unexpected that five significant MH sites and tw o relatively protected IGr-flanki ng regions would be found at the LAT promoter region. It is a little unusual to have so frequent MH sites present in a 433-bp gene promoter region with peak-to-peak interv els shorter than 120 bp. Many mammalian gene promoters typically have been shown to be re latively nucleosome-free and to be somewhat protected from nuclease digestion due to the binding of transc ription factors (Chen and Yang, 2001; Gregory et al., 2001). The asymmetric distri bution of the MH sites of the LAT suggested that the transcription of ves 1 and ves 1 at the LAT might not equivalently employ the two pairs of conserved quasi-palindromic segments at the LA T IGr. Consistent with this possibility, some preliminary results sugg ested that the apposing ves genes at the LAT might not have similar transcription rates or levels, even though they are both subunits of VESA1 protein (Xiao, Y and Allred, DR., unpublished data). Sinc e the two pairs of quasi-palindromic segments are the most conserved elements in the ves IGrs (Al-khedery and Allre d, 2006; Brayton et al., 2007; additional unpublished observations), they are presumably key to ves bidirectional promoter activity. Therefore, it is surprising that the re latively symmetric distribu tion of those conserved quasi-palindromic segments in the ves IGrs did not appear to be re flected in the symmetry of the MH sites pattern, suggesting the flexibility of the ves promoters in transcriptions of both directions. The experiments have been repeat ed several times to avoid interference from heterogeneous parasite populations used for probing MH patterns while the obtained MH patterns were consistent.

PAGE 128

128 On the contrary, non-expressed ves loci, together with the active genes -tubulin and GADPH, all seem to have similar nucleosoma l organizations featur ing regularly-ordered nucleosomes and similar NRLs. The overall in termediate sensitivity of some non-expressed ves loci to nucleases may suggest the complex ity of transcriptiona l control over the ves multigene family. It could be caused by leaky transcription or relatively open chromatin structures caused by stalled RNA polymerase II, or their special ch romosome contexts (Ala rcon et al., 1999; Muse et al., 2007; Zupanska et al., 2009). Further study is needed to explain the intermediate sensitivity of those ves loci to nucleases. The regular nucleosome positioning patterns in/around the ves IGrs observed on those silent IGr-containing ves gene loci, which is consistent with their inactive states, suggested the i nvolvement of epigenetic regula tion in modulating their similar chromatin structures, since all si lent but apparently competent ves loci are dispersed on four chromosomes and reside within di ffering chromosomal contexts. This study cannot rule out the possibility th at the LAT may be mu tually exclusively associated with a special subcompartment in th e nucleus to facilitate its active transcription. DNA/RNA Fluorescence in situ hybri dization (FISH) may be useful in detecting the distribution of different ves genes in the nucleus and their association with RNA polymerases. However, the difficulty of carrying out FISH may rest in gene-specific probe design. The information drawn from this study contributes to our understand ing of the mechanism of mutually transcription control of ves gene expression in B. bovis Higher order DNA structure in chromatin might be adopted in vivo by the LAT to facilitate its transcription. Therefore, epigenetic modifications may be introduced to those ves genes for repression through recruiting chromatin-remodeling factors to make them relativ ely inaccessible to tran scriptional machinery. Meanwhile, the formation of heterochromatin might also help ves gene repression by preventing

PAGE 129

129 the formation of higher order DNA struct ure at those competent but silent ves loci. The preliminary results acquired from IFA and ChIP experiments strongly su pported the involvement of epigenetic modifications such as histone acetylation and methylation.

PAGE 130

130 APPENDIX A PUTATIVE HISTONES AND HISTONE MODIF IING ENZ YMES IN BABESIA BOVIS Putative Histones Chromosome E-value (Identity) Organisms H3 1 1e-66 (93%) Plasmodium falciparum 2 2e-68(89%) H4 3 7e-40(92%) H2A 2 9e-52(75%) H2B 2 5e-37(82%) 4 3e-32(71%) Sequence ID Putative HATs E-value (Identity) Organisms B_bovis* chr*1*930*19 GCN5 family e-165 (56%) Plasmodium falciparum e-84 (44%) Saccharomyces cerevisiae B_bovis*chr*4* 1104837696308*30 Unclassified e-174 (60%) Plasmodium falciparum MYST family 4e-72 (52%) Homo sapiens B_bovis *chr*4*930*19 MYST family 2e-80(53%) Toxoplasma gondii Sequence ID Putative HDACs E-value (Identity) Organisms B_bovis *chr*3*920*17 Unclassified e-129 (53%) Saccharomyces cerevisiae ClassII-HDAC9 e-140 (58%) Arabidopsis thaliana ClassII-HDAC6 e-142 (60%)

PAGE 131

131 APPENDIX B COMPARISON OF HISTONE H3 N-TERMINI BETWEEN BABESIA BOVIS AND OTHE R ORGANISMS

PAGE 132

132 APPENDIX C THE VALIDATION OF COMMERIAL ANTIBOIDIES IN FIXED IFA

PAGE 133

133 LIST OF REFERENCES Agarwal, S., and A. Rao. 1998. Modulation of chrom atin structure regulates cytokine gene expression during T cell differe ntiation. Immunity 9:765-775. Aikawa, M., E. Pongponratn, T. Tegoshi, K. Naka mura, T. Nagatake, A. Cochrane, and L. S. Ozaki. 1992. A study on the pathogenesis of human cerebral ma laria and cerebral babesiosis. Mem Inst Oswa ldo Cruz 87 Suppl 3:297-301. Aikawa, M., J. Rabbege, S. Uni, M. Ristic, and L. H. Miller. 1985. Structural alteration of the membrane of erythrocytes infected with Babesia bovis. Am J Trop Med Hyg 34:45-49. Alarcon, C. M., M. Pedram, and J. E. Donelson. 1999. Leaky transcription of variant surface glycoprotein gene expression sites in bloodstream african trypanosomes. J Biol Chem 274:16884-16893. Al-Khedery, B., and D. R. Allred. 2006. Antigenic variation in Babesia bovis occurs through segmental gene conversion of the ves multigene family, within a bidirectional locus of active transcription. Mo l Microbiol 59:402-414. Allers, T., and M. Lichten. 2000. A method for pr eparing genomic DNA th at restrains branch migration of Holliday junctions. Nucleic Acids Res 28:e6. Allers, T., and M. Lichten. 2001. Intermediate s of yeast meiotic recombination contain heteroduplex DNA. Mol Cell 8:225-231. Allred, D. R., R. M. Cinque, T. J. Lane, and K. P. Ahrens. 1994. Antigenic variation of parasitederived antigens on the surface of Babesia bovis-infected erythrocytes. Infect Immun 62:91-98. Allred, D. R., S. A. Hines, and K. P. Ahrens. 1993. Isolate-specific parasite antigens of the Babesia bovis-infected erythrocyte surf ace. Mol Biochem Parasitol 60:121-132. Ambrosio, R. E., F. T. Potgieter, and N. Ne l. 1986. A column purification procedure for the removal of leucocytes from parasite-inf ected bovine blood. Onderstepoort J Vet Res 53:179-180. Anello, L., I. Albanese, C. Casano, F. Palla, F. Gianguzza, M. G. Di Bernardo, R. Di Marzo, and G. Spinelli. 1986. Different micrococcal nuclease cleavage patterns characterize transcriptionally active and inactive sea-urchin histone genes. Eur J Biochem 156:367-374. Arcangioli, B. 1998. A siteand strand-specifi c DNA break confers asymmetric switching potential in fission yeas t. EMBO J 17: 4503-4510. Barbour, A.G. and Restrepto, B.I. 2000. Antigenic variation in vector-borne pathogens. Emerg Infect Dis 6: 449-457.

PAGE 134

134 Bartova, E., J. Krejci, A. Harn icarova, G. Galiova, and S. Kozubek. 2008. Histone modifications and nuclear architecture: a review. J Histochem Cytochem 56:711-721. Baruch, D. I., X. C. Ma, H. B. Singh, X. Bi, B. L. Pasloske, and R. J. Howard. 1997. Identification of a region of PfEMP1 that mediates adherence of Plasmodium falciparum infected erythrocytes to CD 36: conserved function with variant sequence. Blood 90: 37663775. Bell, L., and B. Byers. 1983. Separation of br anched from linear DNA by two-dimensional gel electrophoresis. Anal Biochem 130:527-535. Bell, L. R., and B. Byers. 1983. Homologous association of chromosomal DNA during yeast meiosis. Cold Spring Harb Sy mp Quant Biol 47 Pt 2:829-840. Bernstein, K. A., E. Shor, I. Sunjevaric, M. Fumas oni, R. C. Burgess, M. Foiani, D. Branzei, and R. Rothstein. 2009. Sgs1 function in the re pair of DNA replication intermediates is separable from its role in homologous r ecombinational repair. Embo J 28:915-925. Bianchi, M. E., M. Beltrame, and G. Paonessa 1989. Specific recognition of cruciform DNA by nuclear protein HMG1. Science 243:1056-1059. Bissler, J. J. 1998. DNA inverted repeats and human disease. Front Biosci 3:d408-418. Bock, R., L. Jackson, A. de Vos, and W. Jorg ensen. 2004. Babesiosis of cattle. Parasitology 129 Suppl:S247-269. Bodansky, S., L. B. Mintz, and D. S. Holmes 1979. The mesokaryote Gyrodinium cohnii lacks nucleosomes. Biochem Bi ophys Res Commun 88:1329-1336. Bon, M., S. J. McGowan, and P. R. Cook. 2006. Many expressed genes in bacteria and yeast are transcribed only once per cel l cycle. Faseb J 20:1721-1723. Borst, P., and I. Chaves. 1999. Mono-allelic expre ssion of genes in simple eukaryotes. Trends Genet 15:95-96. Borst, P., and S. Ulbert. 2001. Control of VSG gene expression sites. Mol Biochem Parasitol 114:17-27. Brayton, K. A., A. O. Lau, D. R. Herndon, L. Hannick, L. S. Kappmeyer, S. J. Berens, S. L. Bidwell, W. C. Brown, J. Crabtr ee, D. Fadrosh, T. Feldblum, H. A. Forberger, B. J. Haas, J. M. Howell, H. Khouri, H. Koo, D. J. Ma nn, J. Norimine, I. T. Paulsen, D. Radune, Q. Ren, R. K. Smith, Jr., C. E. Suarez, O. White, J. R. Wortman, D. P. Knowles, Jr., T. F. McElwain, and V. M. Nene. 2007. Genome se quence of Babesia bovis and comparative analysis of apicomplexan hemoprotozoa. PLoS Pathog 3:1401-1413.

PAGE 135

135 Breeuwer, P., J. L. Drocourt, N. Bunschoten, M. H. Zwietering, F. M. Rombouts, and T. Abee. 1995. Characterization of uptake and hydrol ysis of fluorescein diacetate and carboxyfluorescein diacetate by intracellular este rases in Saccharomyces cerevisiae, which result in accumulation of fluorescent product. Appl Environ Microbiol 61:1614-1619. Brenowitz, M., D. F. Senear, and R. E. Kings ton. 2001. DNase I footprint analysis of proteinDNA binding. Curr Protoc Mol Biol Chapter 12:Unit 12 14. Calder, J. A., G. R. Reddy, L. Chieves, C. H. Co urtney, R. Littell, J. R. Livengood, R. A. Norval, C. Smith, and J. B. Dame. 1996. Monitoring Babe sia bovis infections in cattle by using PCR-based tests. J Clin Microbiol 34:2748-2755. Calvo de Mora, A., J. M. Garcia Castellano, C. Herrera, and J. Jimenez-Alonso. 1985. [Human babesiosis: report of a case with fata l outcome]. Med Clin (Barc) 85:515-516. Cary, C., D. Lamont, J. P. Dalton, and C. Doerig. 1994. Plasmodium falciparum chromatin: nucleosomal organisation and histone-like proteins. Parasitol Res 80:255-258. Castro, C. E. 1983. Nucleosomal repeat length in rat liver nuclei is decreased by a high carbohydrate, fat-free diet. J Nutr 113:557-565. Chen, C., and T. P. Yang. 2001. Nucleosomes are translationally positioned on the active allele and rotationally positioned on the inactive alle le of the HPRT promoter. Mol Cell Biol 21:7682-7695. Collick, A., J. Drew, J. Penberth, P. Bois, J. Luck ett, F. Scaerou, A. Jeffreys, and W. Reik. 1996. Instability of long inverted repeats within mouse transgenes. Embo J 15:1163-1171. Constantinou, A., and S. C. West. 2004. Holliday junction branch migration and resolution assays. Methods Mol Biol 262:239-253. Corona, D. F., A. Eberharter, A. Budde, R. Deur ing, S. Ferrari, P. Varga-Weisz, M. Wilm, J. Tamkun, and P. B. Becker. 2000. Two histone fold proteins, CH RAC-14 and CHRAC-16, are developmentally regulated subunits of chromatin accessibility complex (CHRAC). Embo J 19:3049-3059. Courcelle, J., J. R. Donaldson, K. H. Chow and C. T. Courcelle. 2003. DNA damage-induced replication fork regression and processing in Escherichia coli. Science 299:1064-1067. Courey, A. J., and J. C. Wang. 1983. Cruciform formation in a negative ly supercoiled DNA may be kinetically forbidden under physiological conditions. Cell 33:817-829. Darkin-Rattray, S. J., A. M. Gurnett, R. W. Myers, P. M. Dulski, T. M. Crumley, J. J. Allocco, C. Cannova, P. T. Meinke, S. L. Colletti, M. A. Bednarek, S. B. Singh, M. A. Goetz, A. W. Dombrowski, J. D. Polishook, and D. M. Schm atz. 1996. Apicidin: a novel antiprotozoal agent that inhibits parasite histone deacet ylase. Proc Natl Acad Sci U S A 93:13143-13147.

PAGE 136

136 De Ambrosis, A., N. Ferrari, S. Bonassi, and G. Vidali. 1987. Nucleosomal repeat length in active and inactive gene s. FEBS Lett 225:120-122. Desjardins, R. E., C. J. Canf ield, J. D. Haynes, and J. D. Chulay. 1979. Quantitative assessment of antimalarial activity in vitro by a semiau tomated microdilution technique. Antimicrob Agents Chemother 16:710-718. Dorn, J. F., and P. S. Maddox. 2008. Chromosome se gregation: centromeres get bent. Curr Biol 18:R159-161. Drinkwater, R. D., P. J. Wilson, J. D. Skinner, and L. A. Burgoyne. 1987. Chromatin structures: dissecting their mixed patte rns in nuclease digests. Nucleic Acids Res 15:8087-8103. Duraisingh, M. T., T. S. Voss, A. J. Marty, M. F. Duffy, R. T. Good, J. K. Thompson, L. H. Freitas-Junior, A. Scherf, B. S. Crabb, and A. F. Cowman. 2005. Heterochromatin silencing and locus repositioning linked to regulation of virulence genes in Plasmodium falciparum. Cell 121:13-24. Ebbs, M. L., L. Bartee, and J. Bender. 2005. H3 lysine 9 methylation is maintained on a transcribed inverted repeat by combined act ion of SUVH6 and SUVH 4 methyltransferases. Mol Cell Biol 25:10507-10515. Felsenfeld, G., J. Boyes, J. Chung, D. Clark, and V. Studitsky. 1996. Chromatin structure and gene expression. Proc Natl Acad Sci U S A 93:9384-9388. Field, Y., N. Kaplan, Y. Fondufe-Mittendorf, I. K. Moore, E. Sharon, Y. Lubling, J. Widom, and E. Segal. 2008. Distinct modes of regulat ion by chromatin encoded through nucleosome positioning signals. PLoS Comput Biol 4:e1000216. Figueiredo, L.M. Cross G. M., Janzen, C. J. 2009. Epigenetic regulation in African trypanosomes: a new kid on the block. Nat Rev Microbiol 7:504-13. Fox, K. R. 1997. DNase I footprin ting. Methods Mol Biol 90:1-22. Frank, M., Dzikowski, R., Costantini, D., Amul ic, B., Berdougo, E. and Deitsch, K. 2006. Strict pairing of var promoters and introns is required for var gene silencing in the malaria parasite Plasmodium falciparum J. Biol. Chem. 281: 9942. Freidhoff, K. T. 1988. Transmission of Babesia. In Babesiosis of domestic animals and man. CRC press. Freitas-Junior, L. H., R. Hernandez-Rivas, S. A. Ralph, D. Montiel-Condado, O. K. RuvalcabaSalazar, A. P. Rojas-Meza, L. Mancio-Silva, R. J. Leal-Silvestre, A. M. Gontijo, S. Shorte, and A. Scherf. 2005. Telomeric heterochromatin propagation and hi stone acetylation control mutually exclusive expression of antig enic variation genes in malaria parasites. Cell 121:25-36.

PAGE 137

137 Friedman, K. L., and B. J. Brewer. 1995. Anal ysis of replication intermediates by twodimensional agarose gel electrophor esis. Methods Enzymol 262:613-627. Fryer, C. J., and T. K. Arch er. 2001. Analyzing the contributions of chromatin structure in nuclear hormone receptor activ ated transcription in vivo. Methods Mol Biol 176:283-296. Gellert, M., M. H. O'Dea, and K. Mizuuchi. 198 3. Slow cruciform transitions in palindromic DNA. Proc Natl Acad Sci U S A 80:5545-5549. Glucksmann, M. A., P. Markiewicz, C. Mal one, and L. B. Rothman-Denes. 1992. Specific sequences and a hairpin structure in the te mplate strand are requi red for N4 virion RNA polymerase promoter recognition. Cell 70:491-500. Golderer, G., and P. Grobner. 1991. ADP-ribosylati on of core histones an d their acetylated subspecies. Biochem J 277 (Pt 3):607-610. Gordenin, D. A., K. S. Lobachev, N. P. Degtyareva, A. L. Malkova, E. Perkins, and M. A. Resnick. 1993. Inverted DNA repeats: a source of eukaryotic genomic instability. Mol Cell Biol 13:5315-5322. Grant, P. A. 2001. A tale of histone m odifications. Genome Biol 2:REVIEWS0003. Greaves, D. R., and P. Borst. 1987. Trypanosoma brucei variant-specif ic glycoprotein gene chromatin is sensitive to single-strand-s pecific endonuclease digestion. J Mol Biol 197:471-483. Gregory, R. I., S. Khosla, and R. Feil. 2001. Probing chromatin structure with nuclease sensitivity assays. Meth ods Mol Biol 181:269-284. Gromova, II, O. F. Nielsen, and S. V. Razin. 199 5. Long-range fragmentation of the eukaryotic genome by exogenous and endogenous nucleases proceeds in a specific fashion via preferential DNA cleavage at matrix a ttachment sites. J Biol Chem 270:18685-18690. Gross, D. S., and W. T. Garrard. 1988. Nucl ease hypersensitive sites in chromatin. Annu Rev Biochem 57:159-197. Gubler, U. 1987. Second-strand cDNA synthesis: mRNA fragments as primers. Methods Enzymol 152:330-335. Halme, A., S. Bumgarner, C. Styles, and G. R. Fink. 2004. Genetic and ep igenetic regulation of the FLO gene family generates cell-su rface variation in yeast. Cell 116:405-415. Harty, R. N., C. F. Colle, F. J. Grundy, and D. J. O'Callaghan. 1989. Mapping the termini and intron of the spliced immediate-early transcript of equine herpes virus 1. J Virol 63:51015110. Hashem, V. I., E. A. Klysik, W. A. Rosche, and R. R. Sinden. 2002. Instability of repeated DNAs during transformation in Escherichia coli. Mutat Res 502:39-46.

PAGE 138

138 Hayat, M. A. 2000. Principles and Techniques of Electron Microscopy: Biological Applications. Cambridge University Press. Hsiung, N., and R. Kucherlapati. 1980. Histone gene expression and chromatin structure in mammalian cell hybrids. J Cell Biol 87:227-236. Imai, T., and T. Ohno. 1995. The relationship between viability and intrace llular pH in the yeast Saccharomyces cerevisiae. A ppl Environ Microbiol 61:3604-3608. Inagaki, H., T. Ohye, H. Kogo, T. Kato, H. Bolor, M. Taniguchi, T. H. Shaikh, B. S. Emanuel, and H. Kurahashi. 2009. Chromosomal instab ility mediated by non-B DNA: cruciform conformation and not DNA sequence is responsible for recurrent transl ocation in humans. Genome Res 19:191-198. Iwasaki, H., M. Takahagi, A. Nakata, and H. Shinagawa. 1992. Escherichia coli RuvA and RuvB proteins specifically interact with Holliday junctions and promote branch migration. Genes Dev 6:2214-2220. Jasinskas, A., N. Jasinskiene, and J. P. La ngmore. 1998. Psoralen cros slinking of active and inactive sea urchin histone and rRNA ge nes. Biochim Biophys Acta 1397:285-294. Jiang, C., and B. F. Pugh. 2009. Nucleosome positioning and gene regulation: advances through genomics. Nat Rev Genet 10:161-172. Jin, C., C. Zang, G. Wei, K. Cui, W. Peng, K. Zhao, and G. Felsenfeld. 2009. H3.3/H2A.Z double variant-containing nucleosomes mark nucleosome-free regions' of active promoters and other regulatory regi ons. Nat Genet 41:941-945. Kato, M., S. Hokabe, S. Itakura, S. Minoshima Y. L. Lyubchenko, T. D. Gurkov, H. Okawara, K. Nagayama, and N. Shimizu. 2003. Interarm interaction of DNA cruciform forming at a short inverted repeat sequence. Biophys J 85:402-408. Kim, E. L., H. Peng, F. M. Esparza, S. Z. Maltchenko, and M. K. Stachowiak. 1998. Cruciformextruding regulatory element controls cell-sp ecific activity of the tyrosine hydroxylase gene promoter. Nuclei c Acids Res 26:1793-1800. Kjemtrup, A. M., and P. A. Conrad. 2000. Human babesiosis: an emerging tick-borne disease. Int J Parasitol 30:1323-1337. Kleinschmidt, A. K. 1968. Monolayer technique s in electron microscopy of nucleic acid molecules. Methods in En zymology Volume 12:361-377 Kotani, H., E. B. Kmiec, and W. K. Holloman. 1993. Purification and prope rties of a cruciform DNA binding protein from Ustilago ma ydis. Chromosoma 102:348-354.

PAGE 139

139 Krause, P. J., B. E. Gewurz, D. Hill, F. M. Marty, E. Vannier, I. M. Foppa, R. R. Furman, E. Neuhaus, G. Skowron, S. Gupta, C. McCalla, E. L. Pesanti, M. Young, D. Heiman, G. Hsue, J. A. Gelfand, G. P. Wormser, J. Dickas on, F. J. Bia, B. Hartman, S. R. Telford, 3rd, D. Christianson, K. Dardick, M. Coleman, J. E. Girotto, and A. Spielman. 2008. Persistent and relapsing babesiosis in immunocomprom ised patients. Clin Infect Dis 46:370-376. Krause, P. J., A. Spielman, S. R. Telford, 3rd, V. K. Sikand, K. McKay, D. Christianson, R. J. Pollack, P. Brassard, J. Magera R. Ryan, and D. H. Persi ng. 1998. Persistent parasitemia after acute babesiosis. N Engl J Med 339:160-165. Kuhnert, P., E. Peterhans, and U. Pauli. 1992. Chromatin structure and DNase I hypersensitivity in the transcriptionally active and inactive po rcine tumor necrosis factor gene locus. Nucleic Acids Res 20:1943-1948. Kumari, M., J. C. Stroud, A. Anji, and J. R. McCarrey. 1996. Differential appearance of DNase I-hypersensitive sites correlates with diffe rential transcription of Pgk genes during spermatogenesis in the m ouse. J Biol Chem 271:14390-14397. Kurahashi, H., and B. S. Emanuel. 2001. Long AT-rich palindromes and the constitutional t(11;22) breakpoint. Hum Mol Genet 10:2605-2617. Kurahashi, H., H. Inagaki, K. Yamada, T. Ohye, M. Taniguchi, B. S. Emanuel, and T. Toda. 2004. Cruciform DNA structure underlies the e tiology for palindrome-mediated human chromosomal translocati ons. J Biol Chem 279:35377-35383. Lanzer, M., D. de Bruin, and J. V. Ravetch. 1992. Transcription mapping of a 100 kb locus of Plasmodium falciparum identifies an intergen ic region in which transcription terminates and reinitiates. Embo J 11:1949-1955. Lanzer, M., D. de Bruin, S. P. Wertheim er, and J. V. Ravetch. 1994. Organization of chromosomes in Plasmodium falciparum: a model for generating karyotypic diversity. Parasitol Today 10:114-117. Lanzer, M., D. de Bruin, S. P. Wertheimer and J. V. Ravetch. 1994. Transcriptional and nucleosomal characterization of a subtelom eric gene cluster flanking a site of chromosomal rearrangements in Plasmodi um falciparum. Nucleic Acids Res 22:41764182. Larsen, A., and H. Weintraub. 1982. An altere d DNA conformation detected by S1 nuclease occurs at specific regions in active chick globin chromatin. Cell 29:609-622. Lau, A. O., D. L. Tibbals, and T. F. McElwa in. 2007. Babesia bovis: the development of an expression oligonucleotide micr oarray. Exp Parasitol 117:93-98. Leach, D. R. 1994. Long DNA palindromes, crucif orm structures, genetic instability and secondary structure repa ir. Bioessays 16:893-900.

PAGE 140

140 Lejeune, E., M. Bortfeld, S. A. White, A. L. Pidoux, K. Ekwall, R. C. Allshire, and A. G. Ladurner. 2007. The chromatin-remodeling fact or FACT contributes to centromeric heterochromatin independently of RNAi. Curr Biol 17:1219-1224. Levy, M. G., and M. Ristic. 1980. Babesia bovis: continuous cultivation in a microaerophilous stationary phase culture. Science 207:1218-1220. Levy-Wilson, B., C. Fortier, B. D. Blac khart, and B. J. McCarthy. 1988. DNase Iand micrococcal nuclease-hypersensitive sites in the human apolipoprotein B gene are tissue specific. Mol Cell Biol 8:71-80. Lilley, D. M., and B. Kemper. 1984. Cruciform-reso lvase interactions in supercoiled DNA. Cell 36:413-422. Lisnic, B., I. K. Svetec, A. Stafa, and Z. Zgaga. 2009. Size-dependent palindrome-induced intrachromosomal recombination in y east. DNA Repair (Amst) 8:383-389. Liu, S. C., L. H. Derick, and J. Palek. 1987. Visualization of the he xagonal lattice in the erythrocyte membrane sk eleton. J Cell Biol 104:527-536. Ljunglof, A., P. Bergvall, R. Bhikhabhai, and R. Hjorth. 1999. Direct vi sualisation of plasmid DNA in individual chromatography adsorben t particles by confocal scanning laser microscopy. J Chromatogr A 844:129-135. Lomvardas, S., G. Barnea, D. J. Pisapia, M. Mendelsohn, J. Kirkland, and R. Axel. 2006. Interchromosomal interactions and olfactory receptor choice. Cell 126:403-413. Lusser, A., and J. T. Kadonaga. 2003. Chroma tin remodeling by ATP-dependent molecular machines. Bioessays 25:1192-1200. Mason, J. A., and J. Mellor. 1997. Isolation of nuclei for chromatin analys is in fission yeast. Nucleic Acids Res 25:4700-4701. McCutchan, T. F., J. L. Hansen, J. B. Dame, and J. A. Mullins. 1984. Mung bean nuclease cleaves Plasmodium genomic DNA at sites before and after genes. Science 225:625-628. Merrick, C. J., and M. T. Duraisingh. 2006. Heterochromatin-mediated control of virulence gene expression. Mol Microbiol 62:612-620. Milne, L., Y. Xu, D. M. Perrin, and D. S. Sigman. 2000. An approach to gene-specific transcription inhibition using oligonucleotides complementary to the template strand of the open complex. Proc Natl Acad Sci U S A 97:3136-3141. Misteli, T., A. Gunjan, R. Hock, M. Bustin, and D. T. Brown. 2000. Dynamic binding of histone H1 to chromatin in living cells. Nature 408:877-881.

PAGE 141

141 Muse, G. W., D. A. Gilchrist, S. Nechaev, R. Sha h, J. S. Parker, S. F. Grissom, J. Zeitlinger, and K. Adelman. 2007. RNA polymerase is poised fo r activation across the genome. Nat Genet 39:1507-1511. Narlikar, G. J., H. Y. Fan, and R. E. Ki ngston. 2002. Cooperation between complexes that regulate chromatin structure a nd transcription. Cell 108:475-487. Navarro, M. and Gull, K. 2001. A pol I transcriptional body associated with VSG mono-allelic expression in Trypanosoma brucei Nature 414: 759. Novac, O., D. Alvarez, C. E. Pearson, G. B. Price, and M. Zannis-Hadjopoulos. 2002. The human cruciform-binding protei n, CBP, is involved in DNA repl ication and associates in vivo with mammalian replication origins. J Biol Chem 277:11174-11183. Nguyen, M. Q., Zhou, Z., Marks, C. A., Ryba, N. J. and Belluscio, L. 2007. Prominent roles for odorant receptor coding sequences in allelic exclusion. Cell 131 : 1009. O'Connor, R. M., and D. R. Allred. 2000. Selecti on of Babesia bovis-infected erythrocytes for adhesion to endothelial cells coselects for altered variant erythrocyte surface antigen isoforms. J Immunol 164:2037-2045. O'Connor, R. M., T. J. Lane, S. E. Stroup, and D. R. Allred. 1997. Charact erization of a variant erythrocyte surface antigen (VESA1) expr essed by Babesia bovis during antigenic variation. Mol Biochem Parasitol 89:259-270. O'Connor, R. M., J. A. Long, and D. R. Allre d. 1999. Cytoadherence of Babesia bovis-infected erythrocytes to bovine brain capillary endothelial cells provi des an in vitro model for sequestration. Infect Immun 67:3921-3928. Panyutin, I. G., I. Biswas, and P. Hsieh. 1995. A pi votal role for the structure of the Holliday junction in DNA branch mi gration. Embo J 14:1819-1826. Pazin, M. J., P. Bhargava, E. P. Geiduschek, and J. T. Kadonaga. 1997. Nucleosome mobility and the maintenance of nucleosome positioning. Science 276:809-812. Pearson, C. E., H. Zorbas, G. B. Price, and M. Zannis-Hadjopoulos. 1996. Inverted repeats, stem-loops, and cruciforms: significance for in itiation of DNA replication. J Cell Biochem 63:1-22. Peterson, C. L., and M. A. Laniel. 2004. Histone s and histone modifications. Curr Biol 14:R546551. Pfeifer, G. P., and A. D. Riggs. 1991. Chromatin differences between ac tive and inactive X chromosomes revealed by genomic footprinting of permeabilized cells using DNase I and ligation-mediated PCR. Genes Dev 5:1102-1113. Ramakrishnan, V. 1997. Histone structure and th e organization of the nucleosome. Annu Rev Biophys Biomol Struct 26:83-112.

PAGE 142

142 Rando, O. J., and H. Y. Chang. 2009. Genome-wid e views of chromatin structure. Annu Rev Biochem 78:245-271. Reddy, G. R., D. Chakrabarti, C. A. Yowell, and J. B. Dame. 1991. Sequence microheterogeneity of the thre e small subunit ribosomal RNA genes of Babesia bigemina: expression in erythrocyte cultu re. Nucleic Acids Res 19:3641-3645. Reyes, A. A., and R. B. Wallace. 1987. Mappi ng of RNA using S1 nuclease and synthetic oligonucleotides. Methods Enzymol 154:87-94. Sanz, C., E. Gorab, M. F. Ruiz, J. M. Sogo, and J. L. Diez. 2007. Chromatin structure of ribosomal genes in Chironomus thummi (Dip tera: Chironomidae): tissue specificity and behaviour under drug treatme nt. Chromosome Res 15:429-438. Sastry, S. S., B. M. Ross, and A. P'Arraga 1997. Cross-linking of DN A-binding proteins to DNA with psoralen and psoral en furan-side monoadducts. Comparison of action spectra with DNA-DNA cross-linking. J Biol Chem 272:3715-3723. Scherf, A., R. Hernandez-Rivas, P. Buffet, E. Bo ttius, C. Benatar, B. Pouvelle, J. Gysin, and M. Lanzer. 1998. Antigenic variati on in malaria: in situ swit ching, relaxed and mutually exclusive transcription of var genes during in tra-erythrocytic devel opment in Plasmodium falciparum. Embo J 17:5418-5426. Schones, D. E., K. Cui, S. Cuddapah, T. Y. Roh, A. Barski, Z. Wang, G. Wei, and K. Zhao. 2008. Dynamic regulation of nucleosome positioning in the human genome. Cell 132:887898. Schulze, S. R., and L. L. Wallrath. 2007. Gene regulation by chromatin structure: paradigms established in Drosophila melanoga ster. Annu Rev Entomol 52:171-192. Schwabe, J. W., L. Chapman, and D. Rhodes. 1995. The oestrogen receptor recognizes an imperfectly palindromic response element through an alternative side-chain conformation. Structure 3:201-213. Sharp, A. J., and R. J. Slater. 1993. Mung-B ean Nuclease 1 (EC 3.1.30.1) Methods Mol Biol 16:253-261. Shlyakhtenko, L. S., V. N. Potaman, R. R. Sinden, and Y. L. Lyubchenko. 1998. Structure and dynamics of supercoil-stabilized DNA cruciforms. J Mol Biol 280:61-72. Smale, S. T., and A. G. Fisher. 2002. Chromatin structure and gene re gulation in the immune system. Annu Rev Immunol 20:427-462. Southern, E. 2006. Southern bl otting. Nat Protoc 1:518-525. Sperling, K., B. S. Kerem, R. Goitein, V. Ko ttusch, H. Cedar, and M. Marcus. 1985. DNase I sensitivity in facultative and constitu tive heterochromatin. Chromosoma 93:38-42.

PAGE 143

143 Staynov, D. Z. 2000. DNase I digestion reveals al ternating asymmetrical protection of the nucleosome by the higher order chromatin structure. Nucleic Acids Res 28:3092-3099. Sun, F. L., M. H. Cuaycong, and S. C. Elgin. 2001. Long-range nucleosome ordering is associated with gene silencing in Drosoph ila melanogaster pericentric heterochromatin. Mol Cell Biol 21:2867-2879. Sundin, O., and A. Varshavsky. 1980. Terminal stages of SV40 DNA replication proceed via multiply intertwined catenated dimers. Cell 21:103-114. Svaren, J., and W. Horz. 1996. Regulation of gene expression by nucleosomes. Curr Opin Genet Dev 6:164-170. Tanaka, H., S. J. Tapscott, B. J. Trask, and M. C. Yao. 2002. Short inverted repeats initiate gene amplification through the formation of a large DNA palindrome in mammalian cells. Proc Natl Acad Sci U S A 99:8772-8777. Tanaka, H., and M. C. Yao. 2009. Palindromic gene amplification--an evolutionarily conserved role for DNA inverted repeats in th e genome. Nat Rev Cancer 9:216-224. Telford, D. J., and B. W. Stewart. 1989. Micr ococcal nuclease: its specificity and use for chromatin analysis. Int J Biochem 21:127-137. Tenen, D. G., L. L. Haines, U. M. Hansen, R. G. Martin, and D. M. Livingston. 1985. Formation of a cruciform structure at the simian viru s 40 replication origin abolishes T-antigen binding to the origin in vitro. J Virol 56:293-297. Thomas, J. O., and V. Furber. 1976. Yeast chromatin structure. FEBS Lett 66:274-280. Tirosh, I., and N. Barkai. 2008. Two strategies for gene regulation by promoter nucleosomes. Genome Res 18:1084-1091. Todd, A., N. Cossons, A. Aitken, G. B. Price, and M. Zannis-Hadjopoulos. 1998. Human cruciform binding protein belongs to the 14-3-3 family. Biochemistry 37:14317-14325. Torres-Guerrero, H., D. A. Peattie, and I. Meza. 1991. Chromatin organization in Entamoeba histolytica. Mol Bioc hem Parasitol 45:121-130. Tripp, C. A., G. G. Wagner, and A. C. RiceFicht. 1989. Babesia bovis: gene isolation and characterization using a mung bean nucleasederived expression lib rary. Exp Parasitol 69:211-225. Vega CA, Buening GM, Green TJ, and C. CA. 1 985. In vitro cultivation of Babesia bigemina. American Journal of vete rinary medicine 46:416-420. Vega-Palas, M. A., and R. J. Ferl. 1995. The Ar abidopsis Adh gene exhibits diverse nucleosome arrangements within a small DNase I-s ensitive domain. Plant Cell 7:1923-1932.

PAGE 144

144 Vologodskaia, M. Y., and A. V. Vologodskii. 19 99. Effect of magnesium on cruciform extrusion in supercoiled DNA. J Mol Biol 289:851-859. Wang, D., and R. Landick. 1997. Nuclease cleavage of the upstream half of the nontemplate strand DNA in an Escherichia coli transc ription elongation complex causes upstream translocation and transcriptiona l arrest. J Biol Chem 272:5989-5994. Wang, H., Hermsen, C.C., Sauerwei n, R.W.., Arnot, D.E., Theande r, T.G. and Lavststen, T. 2009. The plasmodium falciparum var gene transcription strategy at the blood stage infection in a human volunteer Parasitol in t 58: 478-480. Wellinger, R. E., R. Lucchini, R. Dammann, and J. M. Sogo. 1999. In vivo mapping of nucleosomes using psoralen-DNA crosslinking and primer extension. Methods Mol Biol 119:161-173. Wells, R. D. 1988. Unusual DNA struct ures. J Biol Chem 263:1095-1098. Weng, S. L., S. L. Taylor, M. Morshedi, A. Schuffner, E. H. Duran, S. Beebe, and S. Oehninger. 2002. Caspase activity and apoptotic marker s in ejaculated human sperm. Mol Hum Reprod 8:984-991. Widmer, R. M., T. Koller, and J. M. Sogo. 1988. Anal ysis of the psoralen-crosslinking pattern in chromatin DNA by exonuclease digestion. Nucleic Acids Res 16:7013-7024. Wiegand, R. C., G. N. Godson, and C. M. Ra dding. 1975. Specificity of the S1 nuclease from Aspergillus oryzae. J Biol Chem 250:8848-8855. Wolffe, A. P., and M. A. Matzke. 1999. Epigenetics: regulation th rough repression. Science 286:481-486. Wu, J. Q., and M. Snyder. 2008. RNA polymerase II st alling: loading at the start prepares genes for a sprint. Genome Biol 9:220. Yang, X., L. M. Figueiredo, A. Espinal, E. Okubo, and B. Li. 2009. RAP1 is essential for silencing telomeric variant surface glycoprot ein genes in Trypanosoma brucei. Cell 137:99-109. Yannone, S. M., and B. K. Burgess. 1997. Identif ication of a palindromic sequence that is responsible for the up-regulati on of NAPDH-ferredoxin reductas e in a ferredoxin I deletion strain of Azotobacter vinela ndii. J Biol Chem 272:14454-14458. Yeruham, I., Y. Avidar, I. Aroch, and A. Hadani. 2003. Intra-uterine Infection with Babesia bovis in a 2-day-old Calf. J Vet Med B In fect Dis Vet Public Health 50:60-62. Zaret, K. 2005. Micrococcal nuclease analysis of chromatin structure. Curr Protoc Mol Biol Chapter 21:Unit 21 21.

PAGE 145

145 Zeitlinger, J., A. Stark, M. Kellis, J. W. Hong, S. Nechaev, K. Adelman, M. Levine, and R. A. Young. 2007. RNA polymerase stalling at developmental control genes in the Drosophila melanogaster embryo. Nat Genet 39:1512-1516. Zhang, Y., and D. Reinberg. 2001. Transcription re gulation by histone methylation: interplay between different covalent modifications of the core histone tails. Genes Dev 15:23432360. Zhang, Z., and J. C. Reese. 2006. Isolation of yeast nuclei and micrococcal nuclease mapping of nucleosome positioning. Methods Mol Biol 313:245-255. Zupanska, A. K., P. B. Drummond, D. M. Swet nam, B. Al-Khedery, and D. R. Allred. 2009. Universal primers suitable to assess populati on dynamics reveal apparent mutually exclusive transcription of the Babesia bovi s ves1alpha gene. Mol Biochem Parasitol 166:47-53.

PAGE 146

146 BIOGRAPHICAL SKETCH Yingling Huang earned a Bachelor of Scie nce in microbial pharmacy from China Pharmaceutical University. Her Master of Science degree in immunology was earned from Chinese Center for Disease Control and Preventi on. After that, Yingling attended University of Florida where she earned a Doctor of Ph ilosophy in veterinary medical sciences.