INVESTIGATION INTO SUBGROUP C FeLV INTERACTION WITH ITS HOST RECEPTOR FLVCR1 AND THE ROLE OF FLVCR1 IN DIAMOND BLACKFAN ANEMIA
By
Michelle Antoinette Rey
A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Graduate Department of Molecular Genetics University of Toronto
© Copyright by Michelle Antoinette Rey 2009 Investigation in subgroup C FeLV interaction with its host receptor FLVCR1 And the role of FLVCR1 in Diamond Blackfan Anemia For the degree of Doctor of Philosophy, 2009 By Michelle Antoinette Rey Graduate Department of Molecular Genetics University of Toronto
ABSTRACT
Retroviral infection requires an initial interaction between the host cell and the virion. This interaction is predominantly mediated by an envelope (env) protein exposed on the external face of the virion. For gammaretroviruses, such as feline leukemia virus
(FeLV), the receptor-binding domain (RBD) is located in the N terminus of env. The RBD forms a distinct domain that is sufficient for binding to the host receptor, but is inefficient in the absence of the corresponding C terminal env, Cdomain, sequence in viral infection studies. I developed a series of hybrid constructs between subgroup C, A and T FeLVs that use distinct receptors for infection to determine the role of Cdom in FeLV binding and infection. Using this approach, I have shown that the C domain (Cdom) of FeLV-C env forms a second receptor-binding domain, distinct from its RBD, which is critical for efficient binding and infection of FeLV-C to host cells expressing FLVCR1. I propose that this mechanism of interaction is conserved for all gammaretroviruses. My results could have important implications for designing gammaretrovirus vectors that can efficiently infect specific target cells. Upon infection with FeLV-C virus, cats develop a disease known as pure red cell aplasia (PRCA). This disease is characterized by a defect in erythropoeisis that results in a decreased number of mature erythroid cells. PRCA has been suggested to be caused by the FeLV-C env binding to and disrupting the host receptor, FLVCR1.
Interestingly, feline PRCA is clinically identical to Diamond Blackfan Anemia (DBA), a
ii fatal congenital anemia characterized by a specific disruption in erythroid progenitor cellular development. I show that erythroid cells from five DBA patients exhibit low levels of total
FLVCR1 transcript expression. In addition, the DBA patients express unique alternatively spliced FLVCR1 transcripts. These alternatively spliced transcripts encode FLVCR1 proteins that are defective in their cellular expression, cell surface localization, and receptor function. Taken together, I propose that the specific anemia observed in DBA is caused by decreased levels of functional FLVCR1 protein due to lowered and alternative splicing of
FLVCR1 transcript.
iii ACKNOWLEDGEMENTS
During the course of my graduate studies, I have overcome many obstacles and challenges with the help and support of family, friends, and co-workers. I would like to extend many thanks and appreciation to those who have encouraged and supported me to continue. First, to my supervisor, Dr. Chetan Tailor, an excellent scientist who gave me a chance to work an exciting project. Thank you for your continued support and guidance in completion of my projects.
To the members of the Tailor Lab, thank you for your scientific expertise and brilliant scientific discussions. This research could not have been completed without your help.
To Nadia, Johnny and Candy who cheered me on from the sidelines. Thank you for your continued encouragement and praise.
To my mother, who was always there for me when I needed her most. Without your continuing support and understanding, I would not have made it this far. Thank you for being my rock solid foundation.
Probably one of the greatest benefits during my doctoral degree was the chance to meet the man of my dreams. Richard, you support has helped me so much over the past few years as I have dealt the stress and challenge of finishing my PhD. Thank you for listening, helping and keeping me grounded.
Finally, to my darling daughter, Maleeka, who could turn a bad day at the lab, to a wonderful evening at home. The one who always makes me laugh when I felt like crying. As you inch ever nearer to being taller than me, entering your teenage years, know that I so proud of you and that I love you.
iv TABLE OF CONTENTS
ABSTRACT…………………………………………………………………………...... ii
ACKNOWLEDGEMENTS…..……………………………………………………...... iv
TABLE OF CONTENTS………………..………………………………………..………...v
LIST OF TABLES…..……………………………………………………………………...ix
LIST OF FIGURES……..………………………………………………………………….ix
LIST OF ABBREVIATIONS……….……………………………………………………..xi
1. INTRODUCTION…………………………..……………………………….…………..1
1.1 Classification of the retroviridae………………………………………………1 1.2 Retrovirus genome structure…………………………………………………..1 1.3 Life cycle………………………………………………………………………...5 1.4 Lentivirus………………………………………………………………………..6 1.5 Gammaretroviruses...... 7 1.5.1 Historical perspectives……………………………………………………..7 1.5.2 Categorization of gammaretroviruses……………………………………7 1.5.3 Retrovirus interference…………………………………………………….8 1.5.4 Gammaretrovirus envelope………………………………………………10 1.5.5 Gammaretroviral envelope-receptor interactions…………………….11 1.5.6 Gammaretroviruses and gene therapy………………………………….12 1.6 Feline Leukemia Virus………………………………………………………..13 1.6.1 Origins of FeLV……………………………………………………………13 1.6.2 FeLV structure……………………………………………………………..14 1.6.3 Epidemiology and immune response……………………………………14 1.6.4 FeLV replication and life cycle………………………………………….15 1.6.5 FeLV subgroups…………………………………………………………...16 1.6.5.1 FeLV-A……………………………………………………..16 1.6.5.2 FeLV-T……………………………………………………...17 1.6.5.3 FeLV-B……………………………………………………...18 1.6.5.4 FeLV-C……………………………………………………...18 1.7 Gammaretroviral receptors…………………………………………………..19 1.7.1 Ecotropic MLV receptor, CAT-1...... 20 1.7.2 Xenotropic and polytropic MLV receptor, XPR1...... 21 1.7.3 RD114, BaEV and type D retrovirus receptor, ASCT2...... 21 1.7.4 Receptors for FeLV………………………………………………………..22 1.7.4.1 FeLV-B receptor, Pit1...... 22 1.7.4.2 FeLV-A receptor, THTR1...... 23
v 1.7.5 FLVCR1…………………………………………………………………….23 1.7.6 Summary……………………………………………………………………26
1.8 Diamond Blackfan Anemia…………………………………………………...26 1.8.1 Clinical diagnosis and treatment methods of DBA……………………27 1.8.2 Erythropoiesis…………………………………………...…………………28 1.8.2.1 Erythropoietin……………………………………………….29 1.8.2.2 Erythroid cell surface protein expression…………………...31 1.8.2.3 Erythropoiesis in DBA patients……………………………..32 1.8.3 Inheritance and genetics …………………………………………………33 1.8.4 Rps19…………………….………...……………...... 33 1.8.4.1 Extraribosomal functions of Rps19…………………………34 1.8.5 Other ribosomal proteins linked to DBA……………………………….35 1.8.6 Ribosomal dysfunction and DBA………………………………………..36 1.8.7 FLVCR1 in DBA…………………………………………………………...36 1.8.8 Other bone marrow failure disorders…………………………………...37 1.8.9 Summary…………………………………………………………………….38
2. THE C DOMAIN IN THE SURFACE ENVELOPE GLYCOPROTEIN OF SUBGROUP C FELINE LEUKEMIA VIRUS IS A SECOND RECEPTOR-BINDING DOMAIN…………………………………………………………………………………...39
2.1 Abstract……………………………………….……………………………….40 2.2 Introduction……………………….…………………………………………..40 2.3 Materials and Methods………………………………………………...... 42 2.3.1 Cell lines…………………………………………………………………….42 2.3.2 Construction of hybrid envelopes………………………………………..42 2.3.3 Generation of feline cells expressing hybrid FeLV envs……………...45 2.3.4 Viruses and infection studies……………………………….…...………..45 2.3.5 Analysis of surface expression of hybrid FeLV envs…...…….………..46 2.3.6 Analysis of env proteins…………………………………………………...46 2.3.7 SU binding assay…………………………………………………….……..47 2.3.8 Immunoprecipitation of soluble SU………………………….…………..48 2.3.9 C2 peptide synthesis…………………………………………..……………49 2.4 Results…………………………………………………………………………49 2.4.1 Cdom is necessary for efficient interference with FeLV-C virus……49 2.4.2 Cdom controls efficiency of FeLV-C SU binding to CR1….…………52 2.4.3 Cdom is critical for FeLV-C infection of target cells……….………...54 2.4.4 C2 peptide is not sufficient for inhibition of FeLV-C infection….…..58 2.4.5 Co-expression of Cdom with hybrid env enhances infection…….…..58 2.4.6 Soluble Cdom binds to CR1 in the absence of FeLV-C RBD...... 58 2.4.7 Cdom does not interfere with FeLV-C infection…………………..…..61 2.5 Discussion……………………………………………………………...... 62
vi 3. THE ROLE OF FLVCR1 in DBA…………………………………………………...69
3.1 Abstract………………………………………………………………………....70 3.2 Introduction………………………………………………………………….....70 3.3 Methods and Materials...... 72 3.3.1 Bone marrow samples...... 72 3.3.2 RNA isolation from CD71high cells...... 73 3.3.3 Isolation of FLVCR1, Pit1 and EpoR sequences from DBA and normal erythroid cells...... 73 3.3.4 Construction of HA tagged E3- and E3-E6- retroviral expression constructs…...... 74 3.3.5 Protein expression profile...... 74 3.3.6 Quantitative real-time PCR...... 75 3.3.7 Genomic DNA analysis...... 75 3.4 Results...... 76 3.4.1 Isolation of alternatively spliced FLVCR1 isoforms from DBA erythroid cells……...... 77 3.4.2 Alternative splicing may be specific for FLVCR1……...... 80 3.4.3 Non-erythroid cell also express alternatively spliced FLVCR1…...... 80 3.4.4 Alternatively spliced FLVCR1 transcripts encode proteins that are disrupted in their cellular and surface expression, and in their receptor function……...... 83 3.4.5 Alternative splicing of FLVCR1 transcript is enhanced in DBA immature erythroid cells……...... 86 3.4.6 RPS19 expression levels vary in DBA patients……...... 86 3.4.7 DBA patients may exhibit mutations in their FLVCR1 gene affecting splicing…………...... 88 3.5 Discussion………………………………………………………...... 90
4. DISCUSSION AND FUTURE DIRECTIONS...…………………………………..... 93
4.1 The role of Cdom in FeLV-C env infection………..……………………...... 93 4.2 Application for specific cell targeting……………………………………...... 95 4.3 FLVCR1 in DBA………………………………………………...... 98 4.3.1 DBA patients exhibit decreased expression of FLVCR1 transcrpt...... 99 4.3.2 DBA patients exhibit increased alternative splicing of FLVCR1...... 100 4.3.3 A potential link between RPS19 and FLVCR1...... 102 4.4 Concluding remarks……………………………………………………...... 104
5. REFERENCES………………………………………………………………………..105
vii ADDENDUM. TARGETING RETROVIRAL INTEGRATION.…………….……...124 A.1 Abstract……………………………………………………………………....125 A.2 Introduction……………………………………………………………….....125 A.3 Methods……………………………………………………………………....128 A.3.1 Hybrid integrase construction…...... 128 A.3.2 Hybrid integrase function……...... 128 A.3.3 Clonal cell line production……...... 130 A.3.4 Integration site analysis………...... 131 A.3.4.1 Inverse PCR...... 131 A.3.4.2 Tailor’d LAM-PCR...... 131 A.3.4.3 Blockerette-ligated capture T7-amplified RT-PCR…...... 132 A.4 Results………………………………………………………………………..136 A.4.1 Hybrid integrase viruses are functional……...... 136 A.4.2 Integration site analysis…………...... 136 A.5 Discussion……………………………………………………………………139
viii LIST OF TABLES
Chapter 1
Table 1-1. Classification of the retroviridae...... 5 Table 1-2. Gammaretroviruses...... 20
Addendum
Table A.1 Infectivity of hybrid integrase virus on Te671 cells...... 136
LIST OF FIGURES
Chapter 1
Figure 1-1. Schematic of retroviral particle………………..……………………...... 2 Figure 1-2. Genomic structure of simple retrovirus …………………………………...... 3 Figure 1-3. Model of viral interference………………………………………...... 9 Figure 1-4. Predicted topology of FLVCR1………………………...... 23 Figure 1-5. Outline of hematopoiesis……………………...... 30
Chapter 2
Figure 2-1. Alignment of SU protein sequence from FeLV-A, -C and -T and construction of hybrid FeLV Envs...... 44 Figure 2-2. Interference property of hybrid FeLV Env...... 50 Figure 2-3. Binding of soluble hybrid FeLV SU proteins to MDTF cells expressing human FLVCR1...... 53 Figure 2-4. Susceptibility of MDTF cells expressing human FLVCR1 to pseudotype virus bearing hybrid FeLV Envs...... 55 Figure 2-5. Effect of C2 loop peptide on FeLV-C infection...... 56 Figure 2-6. Susceptibility of MDTF cells expressing human FLVCR1 to pseudotype virus bearing ΔRBD FeLV-C Envs...... 57 Figure 2-7. Binding of soluble Cdom to FLVCR1 expressing cells...... 59 Figure 2-8. Interference property of soluble Cdom...... 61 Figure 2-9. Proposed models envelope-receptor interaction for FeLV-C infection...... 67
Chapter 3
Figure 3-1. RT-PCR analysis of FLVCR1 from normal and DBA patient bone marrow samples...... 78 Figure 3-2. Alternatively spliced FLVCR1 isoforms isolated from DBA and normal immature erythroid cells...... 79
ix Figure 3-3. RT-PCR analysis of EpoR and Pit1...... 81 Figure 3-4. RT-PCR analysis of RNA from erythroid and non-erythroid DBA patient bone marrow sample...... 82 - - - Figure 3-5. Functional analysis of the E3 and E3 E6 FLVCR1 encoded proteins...... 84 - - - Figure 3-6. Expression analysis of the E3 and E3 E6 FLVCR1 encoded proteins...... 85 Figure 3-7. Quantification of normal and alternatively spliced FLVCR1 transcript expression in DBA and normal immature erythroid cells...... 87 Figure 3-8. Genomic sequencing analysis of FLVCR1 in a DBA patient...... 89
Chpater 4
Figure 4-1. Proposed method of viral targeting to specific cells...... 97
Addendum
Figure A.1. Schematic diagram showing the positions of the various TFIIIC and TFIIIB subunits on tRNA and 5S RNA genes...... 130 Figure A.2. Outline of methods used for integration site analysis Figure A.2A Inverse PCR...... 134 Figure A.2B Tailor’d LAM-PCR...... 135 Figure A.2C Blockerette-ligated capture T7-amplified RT-PCR...... 135 Figure A.3. Inverse PCR did not yield integration site amplification...... 137 Figure A.4. Tailor’d LAM-PCR requires more tailoring...... 138 Figure A.5. Blockerette-ligated capture T7-amplified RT-PCR may identify integration sites...... 140
x LIST OF ABBREVIATIONS
BFU-E Erythroid burst forming unit BM Bone marrow CA Capsid Cdom C domain CB Cordblood CD Cluster of differentiation CFU-E Erythroid colony-forming unit DBA Diamond Blackfan Anemia EDTA Ethylenediaminetetraaceticacid Env Envelope protein Epo Ertyhropoietin EpoR Erythropoietin receptor ESE Exonic splicing enhancers ESS Exonic splicing silencer FACS Fluorescence-activated cell sorting FBS Fetal bovine serum FeLV Feline leukemia virus FLVCR1 Feline leukemia virus subgroup C receptor 1 GALV Gibbon ape leukemia virus GlyA Glycophorin A Gp70 Glycoprotein 70 His Histidine HIV Human immunodeficiency virus hnRNP Heterogenous ribonuclear protein HSC Hematopoietic stem cell IN Integrase ISE Intronic splicing enhance ISS Intronic splicing silencer Lin- Lineage depleted LTR Long terminal repeat MA Matrix MLV Murine leukemia virus NC Nucleocapsid PBS Phosphate buffered saline Pit1 Inorganic phosphate symporter 1 PCR Polymerase chain reaction Pol Polymerase PR Protease PRR Proline rich region PRCA Pure red cell aplasia qPCR Quantitative polymerase chain reaction RBC Red blood cell RBD Receptor binding domain
xi RPS19 Ribosomal protein S19 RT Reverse transcriptase RT-PCR Reverse-transcriptase polymerase chain reaction SCID Severe combined immunodeficiency SDS-PAGE Sodium dodecyl sulfate polyacrylamide gel electrophoresis SU Surface protein TM Transmembrane VRA Variable region A Wt wildtype
xii 1. INTRODUCTION
1.1 Classification of the retroviridae
Retroviruses are infectious agents that are unable to reproduce outside of a host cell. They are known mostly for their severe effects on human health, but interest in their use as therapeutic agents has increased due to the advent of retroviral gene therapy. The term ‘retro’ was coined to describe the way in which retroviruses transmit their genome. Unlike some pathogens, whose genome consists of DNA that is replicated for their progeny, retroviruses contain an RNA genome that is reverse transcribed into DNA by the enzyme ‘reverse transcriptase’ for incorporation into the host cell genomic DNA (Temin and Mizutani, 1970, Baltimore, 1970). This method of genome propagation is backwards to the normal use of RNA as a transient copy of genomic DNA that is used mostly for coding protein. The retrovirus virion is enveloped by a membrane derived from the host cell from which it budded (Quigley et al., 1971) giving rise to a virion 80-200 nm in diameter with a spherical morphology. On the surface of this membrane exists virally encoded envelope proteins which are key in the initial interaction between the virion and its host cell; this interaction is a key feature to my studies and will be discussed in further detail in Chapter 1.4.
1.2 Retrovirus genome and structure
The retroviral genome is consists of two single positive stranded RNA molecules that range in size from 8-11 kilobases (kb) (Billeter et al., 1974, Kung et al., 1975, Beemon et al., 1976). The RNA molecules have a 5’ cap and 3’ poly adenylation tail equivalent to eukaryotic mRNA and encode for the gag, pol and env genes of the virus. The structural coat components, encoded within the gag gene, consist of the capsid, nucleocapsid and matrix proteins (Vogt and Eisenman, 1973). The capsid proteins (CA) form an
1 2
nucleocapsid integrase RBD PRR matrix Reverse Cdom RNA transcriptase capsid protease TM Figure 1-1. Schematic of retroviral particle. Cross section of a retrovirus particle. Virion contains two copies of single-stranded positive RNA. Nucleocapsid protein binds to the viral RNA genome. Capsid protein forms an icosahedral complex surrounding the viral genome. Matrix protein surrounds the icosahedral capsid underneath the lipid bilayer which is derived from the host cell membrane. The virion carries its own enzymes to complete reverse transcription of the viral RNA to cDNA and integration into the host cell genome. The envelope protein exists as a trimer at the surface of the virion lipid bilayer. The surface unit (SU) is a comprised of the receptor binding domain (RBD), proline rich region (PRR), and Cdomain (Cdom), and is anchored to the membrane non-covalently by the transmembrane (TM) region. 3
ψ 5’ LTR gag pol env 3’ LTR
MA CA NC
RT IN PR
SU TM
H TM
VRA VRC VRB PRR C dom LRR
RBD C1-C2-C3
Figure 1-2. Genomic structure of a simple retrovirus. The gag gene encodes for the matrix, capsid and nucleocapsid proteins, while the pol gene encodes for the protease, reverse transcriptase, and integrase genes. The env gene encodes for the viral envelope protein that consists of a surface protein (SU) and a transmembrane segment (TM). Upon alignment of the MLV and the FeLV gammaretroviral envelope protein sequences, sequence conservation and variability became apparent. Three variable regions (VRA, VRB and VRC) comprise the receptor binding domain (RBD). The proline rich region (PRR) contains many pro residues and is believed to act as a hinge between the RBD and C terminal end of the SU. The C region (C1, C2 and C3) is of unknown function, but proposed to be involved in virus infection. The transmembrane domain contains a highly hydrophobic leucine-rich region (LRR) involved in fusion with host cell membranes. N-terminal histidine (H) required for fusion is indicated.
4 icosahedral structure surrounding the viral genome. Approximately 2000 molecules of nucleocapsid (NC) protein bind the two copies of the RNA genome. The matrix protein (MA) assembles just beneath the lipid bylayer of the membrane (Pepinsky and Vogt, 1979) (Figure 1-1). The pol gene encodes for proteins involved in viral replication such as the aforementioned reverse transciptase enzyme (RT), the integrase (IN) used to integrate the viral DNA into the host genome, and the protease (PR) required to cleave the viral proteins into their individual components (reviewed in Katz and Skalka, 1994). The env gene encodes for the surface subunit (SU) and transmembrane (TM) subunit of the envelope protein (Duesberg et al., 1970). These proteins exist on the surface of the virion and are responsible for the initial virus to host cell interaction required for virus entry. The SU is required for binding to the retroviral receptor while the TM is involved in fusion of the host cell and retroviral membranes. The SU exists as a trimer of the SU-TM throughout the surface of the virion. A schematic of the genomic structure is depicted in Figure 1-2. The gag, pol, and env genes comprise the main genomic structure of a simple retrovirus, such as the gammaretrovirus, but a number of other retroviruses (such as the lentivirus) contain additional genes that are critical in regulating virus replication and are, therefore, referred to as complex retroviruses. Retroviruses were previously identified and classified due to their appearance under the electron microscope; however, other criteria have been applied to classify them based on their structure and genomic organization (Table1-1).
5
Table 1-1. Classification of the retroviridae
GENUS EXAMPLES ISOLATION/DISCOVERY α-retrovirus Rous sarcoma virus (RSV) Rous, 1911 (simple) Avian leukosis Virus (ALV) Ellermann and Bang, 1908
β-retrovirus Jaagsiekte sheep retrovirus (JSRV) Verwoerd et al., 1983 (simple) Mouse mammary tumour virus (MMTV) Bittner, 1936
γ-retrovirus Feline leukemia virus (FeLV) Jarrett et al. 1964 (simple) Murine leukemia virus (MLV) Gross, 1951 Gibbon ape leukemia virus (GALV) Theilen et al., 1971
δ-retrovirus Bovine leukemia virus (BLV) van der Maaten et al., 1972 (complex) Human T-cell leukemia virus (HTLV) Poiesz et al., 1980
ε-retrovirus Walleye dermal sarcoma virus (WDSD) Walker, 1969 (complex) Lentivirus Human immunodeficiency virus (HIV) Barre-Sinoussi et al., 1983 (complex) Simian immunodeficiency virus (SIV) Letvin et al., 1985 Feline immunodeficiency virus (FIV) Pedersen et al., 1987
Spumavirus Human foamy virus (HFV) Achong et al., 1971 (complex) Simian foamy virus (SFV) Enders and Peebles, 1954
1.3 Life cycle Retroviruses bind to distinct cell surface receptors on a host cell for entry. Upon viral binding and fusion of the host and viral membranes, the virus enters the cell and uncoats through loss of the capsid protein. The RNA genome is then reverse transcribed into cDNA by the viral RT, transported to the nucleus, and then integrated into the host genome by the viral integrase protein. The viral cDNA is flanked by two long terminal repeats (LTR) that contain enhancers to drive transcription of the viral genes and processing signals for the cleavage and polyadenylation of the viral RNA transcripts. Upon translation of the viral transcripts in the cytoplasm, the polypeptides are transported to the cell surface where the virus buds from the host cell. After budding the virus matures and the polypeptides are cleaved into their respective proteins by the viral protease (PR). The env gene is processed 6 in the endoplasmic reticulum, glycosylated and transported to the cell membrane by the host cell machinery. The viral RNA genome contains packaging signals that induce its binding by nucelocapsid (NC) and encapsulation by capsid (CA) during virion maturation. The viral particle is assembled at the cell surface just beneath the host cell membrane and then buds from the host cell gaining a cellular derived outer lipid bilayer membrane. The virus buds from the host cell without cell lysis, enabling a singly infected cell to produce many virus particles over its lifetime and enhance the rate of infection.
1.4 Lentivirus
The term lentivirus comes from the Latin term lenti which means slow. This term aptly describes a lentiviral infection as the virus often integrates within the host genome and remains latent for many years. The best characterized lentivirus and perhaps the most studied retrovirus is the human immunodeficiency virus (HIV) which causes acquired immune deficiency syndrome (AIDS). HIV is a complex retrovirus with many accessory proteins to aid in viral replication making it different from the simple retroviruses, but the envelope protein (env) of HIV while structurally different, displays mechanistic similarities in binding to its receptor as that of the gammaretroviruses. HIV env consists of two glycoproteins gp120 (the surface unit, SU) and gp41 (the transmembrane unit, TM) that interact with the HIV receptor and co-receptors, CD4 (Dagelish et al, 1984, Klatzmann et al, 1984), and CCR5, & CXCR4 (Feng et al., 1996, Dragic et al., 1996, Deng et al., 1996), respectively. gp120 sequences isolated from various strains of HIV show relatively conserved regions, with five variable regions interspersed through the envelope protein (Starcich, 1986, Myers et al., 1994). These studies led to attempts to characterize the specific regions of gp120 that interacted with its receptor and co-receptors. It was apparent upon co-crystallization of gp120 and CD4 (Kwong, 1998, Kwong et al. 2000, Huang et al. 2005) that the variable loop 3 (V3) of HIV env protruded from the core gp120 protein and was in close proximity with the CD4 receptor. The interaction between these two proteins initiated research towards the production a therapeutic target capable of disrupting this contact; thus antibodies antigenic to V3 were made and shown to disrupt HIV fusion with its target cell in vitro (Rusche et al., 1988), but not the binding between gp120 and CD4. 7
Interestingly, these V3 antibodies have also been shown to disrupt gp120-CD4 interaction with the chemokine co-receptors CCR5 (Wu et al., 1996, Trkola, et al., 1996). Furthermore, the presence of peptides consisting of V3 amino acid sequence can inhibit HIV infection of CD4+ HeLa cells expressing CXCR4 or CCR5, or both, co-receptors (Verrier, et al., 1999). These experiments demonstrated that the V3 epitope of gp120 is a significant component of the HIV env interface involved in receptor binding and virion interaction with the host cell. These results taken together show the importance of studying the mechanism of envelope- receptor interactions as they can lead to the production of therapeutic targets in the efforts to fight the many different diseases caused by retroviruses.
1.5 Gammaretroviruses
1.5.1 Historical perspectives
The γ-type genus of the retroviridiae family includes the feline leukemia virus (FeLV), murine leukemia virus (MLV), porcine endogenous retrovirus (PERV), Gibbon ape leukemia virus (GALV) and the endogenous feline retrovirus RD114 (Table 1.1). Of this group of simple retroviruses, MLV has been the most studied perhaps due to its early identification and infection of a frequently used model organism, the mouse. MLV was first described by Furth and colleagues in 1943 by breeding high-leukemia-incidence mouse strains (Furth et al. 1943). MLV infection induces disease in mice that are related to human disease such as leukemia and acquired immunodeficiency syndrome (AIDS). Also, some retroviral vectors used in gene therapy have been based on an MLV genome.
1.5.2 Categorization of gammaretrovirus
MLV is found predominantly in the hematopoietic tissues such as the bone marrow and lymph nodes, and the virus can remain latent from 2 to 18 months depending on the strain of the virus and the mouse. MLV has been classified into different subtypes based on host range infection, and subsequently the principles of viral interference, which will be discussed further in Chapter 1.5.3. Interference studies led to the identification of the 8 following subtypes of MLV: ecotropoic MLV (E-MLV) which is an exogenous virus capable of infecting only murine cells; xenotropic MLV-(X-MLV) which is an endogenous virus capable of infecting non-murine cells only in vitro; amphotropic and polytropic MLV (A-MLV, P-MLV) which are capable of infecting both murine & non-murine cells. Feline leukemia virus consists of four subgroups FeLV-A, -B, -C, and –T and they will be discussed in further detail in section 1.5. The porcine endogenous retrovirus (PERV) genomic sequence is present in many copies within the pig genome, but it has yet to be associated with any disease (Patience, et al. 1997). There are three PERV subgroups (A,B,C). The virus can be activated to produce virions that are infectious of several human cell lines in vitro and thus raises several issues regarding xenotransplantation of pig organs into humans (Akiyoshi et al., 1998). Gibbon-ape leukemia virus (GALV) is another species of the gammaretrovirus genus that infects the gibbon ape causing T-cell leukemia. Although no subgroups have been identified, GALV does have different strains such as GALV-Seato (Kawakami and Buckley 1974) or GALV-SF (Kawakami et al., 1972). RD-114 is another endogenous, xenotropic virus whose viral genome is present in all domestic cats (McAllister et al., 1969). Similar to PERV, RD-114 is not known to cause any disease in cats, and it has not yet been shown to recombine with FeLV to create an infectious virus (Hardy, 1983).
1.5.3 Retrovirus interference
The interaction between the SU of the virion and a receptor on the surface of a host cell is a crucial determinant for the infectivity of that virus for that cell. Receptor usage is a major determining factor in the host range and tissue tropism of a retrovirus. Changes in the SU sequence can alter receptor specificity, and subsequent host range. These changes have been used to categorize the gammaretroviruses into their various subgroups. Receptor specificity also determines the ability of a retrovirus to interfere with other viruses that use the same receptor for infection (Figure 1-3). Interference occurs when endogenous expression of the retroviral SU blocks infection by binding to its cognate receptor in the rough endoplasmic reticulum before it is transported to the cell surface. This binding forms a complex that is unable to be shuttled to 9
uninfected cell Cell infected with virus bearing env A
Legend
Receptor A
Receptor B
Envelope A
Envelope B
Figure 1-3. Model of retroviral interference. Interference occurs when the retroviral SU blocks available cellular receptors at the surface by binding to the receptors in the rough endoplasmic reticulum. In this scenario, a cell is infected with a virus bearing env A which leads to endogenous expression of the envelope protein and its interaction with receptor A. This binding forms a complex unlikely to be shuttled to the surface, thus down regulating expression of the receptor at the cell surface. Furthermore, any receptor A shuttled to the cell surface may be bound by env A. Subsequent infection by the same virus or a different virus that also utilizes receptor A is blocked. Conversely, infection with a virus bearing a different env B that binds to receptor B is still able to infect the cell. 10 the surface, thus down regulating expression of the receptor at the cell surface. Therefore, with the decreased expression of the receptor at the cell surface and any receptor at the surface being bound by SU, subsequent infection by a virus that also utilizes that receptor is blocked. Additionally, the SU-receptor interaction also disrupts the normal cellular function of the receptor, which if critical can cause cell death (Kavanaugh et al., 1994, Quigley et al., 2004). Thus, these interactions can ultimately determine the disease specificity of the virus.
1.5.4 Gammaretrovirus envelope
Upon alignment amongst FeLV subgroups and the closely related MLV, it became apparent that several regions of the env protein were conserved, and those residues outside these regions are highly variable. There exist several regions of high variability in the MLV and FeLV env, three of which (VRA, VRB and VRC) are collectively defined as the receptor binding domain (RBD) of the SU located in the N-terminus. These variable regions differ considerably in sequence and in length; therefore they were suggested to contain the primary receptor recognition residues. The RBD contains eight conserved cysteine residues that form disulfide bonded loops, and this region has been characterized to be critical for initial binding of the viral env to the host cell receptor (Battini et al., 1995, Battini et al, 1992 Tailor, et al., 1997, Tailor et al., 2000). Furthermore, some residues are critical for post-binding events such as fusion and will be discussed in Chapter 1.5.5. The crystal structure for the RBD of FeLV-B and ecotropic MLV have been solved and they reveal the variable regions as loops that extrude from the globular domain of the RBD (Barnett et al., 2003, Fass et al., 1997). Thus, the variable regions are more exposed for important contact to their specific receptor. Although the crystal structure has not been solved for FeLV-A, - C, or –T, it is hypothesized that they will also have an RBD with exposed variable regions on the outer surface. The fourth variable region occurs in the C domain (Cdom), which is comprised of three regions, C1-C3. The majority of this variation is found in C2, suggesting the importance of this region in some envelope function. The proline rich region (PRR) is less conserved between MLV and FeLV, but it has been suggested to maintain the correct conformational shape between the N-terminal RBD and the C-terminal domain of the SU (Battini et al., 1992). It is shown to be necessary for efficient viral entry as it provides 11 structural support for MLV RBD, but is not involved in receptor recognition (Gray and Roth, 1993, Battini et al., 1995). Furthermore, a critical histidine residue in the N-terminal of the envelope protein is crucial for triggering fusion (Bae et al., 1997, Lavillette et a., 2000, Zavorotinskya and Albritton, 1999). The TM of gammaretrovirus envelopes are highly related exhibiting over 95% sequence similarity.
1.5.5 Gammaretroviral envelope-receptor interactions
The mechanism behind gammaretroviral envelopes binding to their cognate receptors and triggering virus fusion has not been fully revealed. The mechanism is of great interest for scientists attempting to target retroviral infection to cells expressing a specific receptor. It is known that a critical histidine residue in a PHQ motif of the RBD is critical in triggering virus fusion (Bae et al., 1997, Zavorotinska et al., 1999, Lavillette et al., 2000), but dispensable for receptor binding. Intriguingly, this fusion defective envelope (ΔH) can be rescued with soluble RBD or SU with an intact histidine when provided in trans (Barnett et al., 2001, Lavillette et al., 2001). Furthermore, for some ΔH viruses, the soluble RBD need not be homologous to the ΔH virus, but both receptors for the ΔH and soluble RBD must be present. This suggests that the rescuing histidine residue may not be exposed until the env (or soluble RBD) has made contact with its cognate receptor. The exposed histidine is now available to aid fusion of the defective envelope protein. The current model for MLV env fusion activation involves interaction of RBD with the host receptor followed by interaction between RBD and Cdom (Barnett et al., 2001, Lavillette et al., 2001). This secondary interaction has been proposed to initiate fusion between the virus and host cell; however, a direct interaction between RBD and Cdom has yet to be demonstrated. There have been several reports that the C-terminal region of the gammaretrovirus may also control receptor recognition and virus infection. Studies characterizing PERV and FeLV-B have shown the importance of the C-terminal region for efficient virus binding and/or infection (Sugai et al., 2001, Faix et al., 2002, Gemeniano et al., 2006). In addition, studies have shown that the C2 loop in Cdom of FeLV-A and FeLV- T env control infection of target cells (Cheng et al., 2006). Due to the similarities in their 12 env sequence and mechanism for binding to their cognate receptors, it has proven to be possible to switch the receptor used for infection simply by switching the RBD of the env. These switching studies have allowed researchers to map the regions of the receptor involved in interacting with MLV and FeLV-B env (Tailor, et al., 1997, Tailor er al., 2000, Bae et al., 1997, Battini et al., 1995, Battini et al., 1992).
1.5.6 Gammaretroviruses and gene therapy
Recombinant retroviral vectors have been of major interest for gene delivery systems due to features that are conducive to efficient and stable gene transfer. Receptor-mediated uptake of the membrane-coated viral particle along with cytoplasmic assembly of the particles, and the use of RNA as the mobile form for transfer of genetic information are some of the unique sought after features of retroviruses. Perhaps the most interesting aspect of retroviral gene therapy is the ability of the reverse-transcribed double stranded DNA to be integrated into the host genome establishing an active or latent infection. Also, the ability of retroviruses to carry non-viral genes for genetic transfer to host cells is crucial in gene therapy. Due to their simple genomic structure and non-overlapping cis-acting genes, gammaretroviruses (especially MLV) have been used extensively in the construction of vectors for gene therapy. Furthermore, MLV vectors can be easily pseudotyped with different env glycoproteins without affecting virus production or efficacy (Sandrin et al., 2004). Of particular note is the increased efficiency of FeLV pseudotyped virus to infect hematopoietic stem cells over its gammaretroviral counterparts, MLV and GALV (Lucas et al., 2005, Josephson et al., 2000) suggesting that FeLV may be a suitable counterpart in retroviral vector construction. The ability to pseudotype retroviral vectors is crucial when attempting to target specific cells for gene delivery. Most retroviral gene transfer protocols follow an ex vivo method whereby the target cells are removed from the body, infected with the virus, and then retransplanted. This approach can be inconvenient and inefficient especially if the targeted cells exist normally in the heart or lung, for example. Thus, the concept of an injectable and targetable retroviral vector is of current interest to scientists. Limiting the host range properties of retroviral vectors has been attempted through modification of the 13 pseudotyped envelope glycoprotein that renders it unable to infect non-target cells and/or enhances its attraction towards target cells. Peptides with known specific binding activities have been added onto N-terminally truncated envelope glycoproteins in attempts to target infection to a particular cell type. Hepatocyte growth factor-envelope fusion proteins expressed in an amphotropic retroviral vector improved transduction efficiency of primary hepatocytes. Another experiment involving a chimeric erythropoietin (Epo) joined to a truncated ecotropic MLV envelope protein was several times more infectious of murine cells expressing the Epo receptor (EpoR) (Nguyen et al., 1998), and human cells with EpoR (Kasahara et al., 1994). Unfortunately, while increasing specificity of cellular targets, the efficiency of viral incorporation and subsequent virus titre has been low with these hybrid retroviral envelope approaches. Interactions between the SU and TM of the envelope protein must be conserved through correct folding, processing and trafficking. Increasing our understanding of envelope and receptor interaction will provide an avenue for further experimentation in hybrid envelope construction that increases specificity for a particular host cellular target while maintaining proper virion function.
1.6 Feline leukemia virus (FeLV)
1.6.1 Origins of FeLV
In 1964, a household in Scotland containing multiple cats displayed an increased incidence of lymphosarcomas suggesting that some form of a contagious agent was causing the disease. Four kittens were infected with cells from a lymphosarcoma sample and all four developed symptoms of the disease confirming the presence of an infectious agent (Jarret et al., 1968). Virus-like particles were identified within intracellular vesicles upon further analysis of the infected tissues. These virus-like particles closely resembled those observed in mice and chicken infected with leukemia-causing virus and thus were named feline leukemia virus (FeLV) (Jarret et al. 1968). Due to the great similarity (~50%) between FeLV and MLV, it is thought that FeLV may have originated from an ancestral rodent virus and crossed over when cats preyed on rodents (Benveniste et al., 1975, Neil et al. 1985). 14
1.6.2 FeLV structure
FeLV structure is like that of all the γ-type retroviruses; it contains two linear single stranded RNA molecules surrounded by a proteinaceous core, an inner protein coat, and a viral envelope (Figure 1-1). The FeLV genome is approximately 8 kilobases (kb) in length and it encodes for the gag, pol, and env genes. All FeLV subgroups contain similar gag genes, and the capsid protein of 27 kDa (p27) is commonly used as a diagnostic tool when confirming the presence of FeLV. The FeLV env gene encodes for the viral precursor glycoprotein of 85 kDa (gp85) which is proteolytically processed into gp70, a glycosylated extracellular surface protein (SU) of 70 kilodaltons (kDa), and a non glycosylated hydrophobic transmembrane (TM) protein, p15, which non-covalently holds the SU adjacent to the lipid bylayer of the membrane.
1.6.3 Epidemiology and immune response
Many pet cats become infected with FeLV thus making it a significant cause of death amongst young adult cats. FeLV can be found in 13% of sick cats, and 4% of healthy cats in the US (Shelton et al., 1990). The incidence of viral infection increases amongst households with numerous pet cats to 30% due to the close proximity of the pets and their shared use of bowls and litter pans. Greater than 35% of cats test positive for antibodies against the virus, indicating they have come in contact with an infected cat and developed immunity rather than an infection (Jarrett, 1994). This may be due to the weak infectivity of the virus for adult cats, kittens younger than eight months of age are at the greatest risk for acquiring FeLV infection. Older cats may have an advantage in being able to resist infection from the weak virus due to a stronger immune system. Some cats develop a latent FeLV infection whereby the virus is present in the bone marrow, but not replicating, and p27 is not present in the serum (Hardy, 1991). This period of latency was found to be short as the majority of cats cleared the infection from the bone marrow within months, and only 14% of cats remained latently infected after 138 weeks (Pacitti and Jarret, 1985). It has been suggested by Hoover and colleagues that the latently infected bone marrow cells are eventually eliminated by the immune response of the cat; or, as the virus infects only a small proportion 15 of the immature cells of the bone marrow, those cells over time will differentiate, leave the bone marrow and die (Hoover at al., 1976). Taken together, the ability of a FeLV infected cat to fight off infection is dependent on several factors including the age of the cat, the status of the immune system, and the amount of virus.
While there is evidence of FeLV in other non-domestic cat populations such as the lynx and the lion, there is no epidemiological evidence suggesting that FeLV can be transmitted to humans or dogs. Approximately one in five pet dogs live in a household with a pet cat, and all pet cats co-habitate with humans; thus there appears to be no species jumping within the natural environment for this virus.
1.6.4 FeLV replication and life cycle
FeLV virus is found in both the saliva and the blood. Thus, it is believed that biting, scratching and licking are the predominant methods of transmission for the FeLV-A. Upon exposure to the virus, the surface of the eye, mouth and nose may be penetrated by the virus which then travels to the lymph nodes for replication. At this early stage in viral entry, most cats are able to develop an immune response to resist infection and remain immune to the virus. After replication in lymphocytes, the virus travels to the bone marrow which is a site of high titre replication. It is at this stage that new viral subgroups may arise through recombination with endogenous FeLV env sequences or point mutations. Also, the p27 capsid protein can now be detected in the blood of a FeLV infected cat, and can be carried out to the salivary glands to be transmitted to another cat. This cycle has been reported to take approximately four weeks (Snyder et al., 1982). Transmission through the placenta of an infected female cat is very rare; in addition, most cats infected with the virus become infertile. However, transmission can occur from an infected mother to her newborn kitten via the milk (Hardy et al., 1976).
16
1.6.5 FeLV subgroups
Feline leukemia viruses (FeLVs) cause a wide range of diseases in cats from non- aggressive leukemia to severe leukemia and immune deficiency (Table 1-3). There are four subgroups of the FeLVs, FeLV-A, FeLV-B, FeLV-C, and FeLV-T, each of which causes a distinctive disease in cats (Jarrett, O., et al., 1973, Moser, M., et al., 1998, Sarma, P.S. and Log, T., 1973). Each of the four subgroups use a different receptor for entry into the host cell and this difference in receptor usage has been suggested to cause the specific diseases seen in cats. Viral interference and neutralization assays have confirmed different receptor usage by FeLV-A, -B, and –C (Sarma and Log, 1973). The subgroups differ only in the env sequence (Figure 1-4), further supporting their use of different receptors for viral entry as each env protein is specific for a particular receptor on the host cell surface. The changes in amino acid sequence of the env can also dictate their host range as some receptors are expressed only in certain cells. Interest in the study of FeLV has been peaked due to their candidacy for gene therapy studies as they have been shown to infect and introduce a desired gene into human hematopoietic stem cells much more efficiently than the previously used MLV-based viral vectors. (Josephson, N., et al. 2000). FeLV provide an excellent model for studying envelope-receptor interactions due to their highly related env sequence coupled with the use of distinct receptors for viral entry. The small regions of variability between the subgroups can be utilized to determine the regions of the env protein that are involved in binding to its specific receptor, and which regions play a role in post-binding events such as fusion of the viral and host membranes.
1.6.5.1 FeLV-A FeLV-A is the progenitor virus from which the others arise. It is the most commonly found FeLV and thus is present in all viral samples isolated from infected cats (Jarrett et al, 1984). FeLV-A is mildly pathogenic in the absence of the other subgroups, though it infrequently causes lymphomas in cats. It is the only subgroup transmitted from cat to cat through saliva, blood and stool samples, and is found most often in stray cats due to their increased propensity to bite and fight. FeLV-A can only infect feline cell lines, but it has 17 been shown to infect some canine and human cell lines in vitro, albeit weakly (Rey, M. et al, unpublished, Moser et al., 1998).
1.6.5.2 FeLV-T FeLV-T virus was originally isolated from a cat with FeLV induced immunodeficiency (FAIDS) similar to human AIDS. This newly discovered viral subgroup was found to share approximately 96% similarity with FeLV-A, although it is quite rare in the infected cat population (Overbaugh et al., 1988, Donahue et al., 1991). FeLV-T arises from point mutations in the env sequence that renders it unable to use feTHTR1 as a receptor. These mutations are almost exclusively within the variable region A (VRA) and the C domain (Cdom) of the env gene. Specifically, there is a six amino acid deletion in the VRA of FeLV-T along with a few amino acid changes which are suggested to be a part of the RBD of FeLV-A required for binding to its receptor. Interestingly, there is a six residue insertion in the Cdom of FeLV-T env that may have implications in envelope structure and receptor binding. These point mutations in FeLV-T env collectively lead to alternate receptor use and restriction to feline T cells and fibroblasts (Overbaugh et al, 1988). How FeLV-T enters these cells is not clear, as the receptor for this subgroup has yet to be found. An N-terminal amino acid change of a histidine residue present in FeLV-A to a proline in FeLV-T has been suggested to render the virus fusion-defective (Figure 1-4) (Bae et al., 1997). This histidine residue has been proven critical for fusion of the viral and host cell membranes for the closely related gammaretrovirus MLV (Bae et al., 1997, Zavorotinskaya et al., 1999). When mutated in MLV env the virus can bind to the receptor, but is unable to initiate the fusion events necessary for infection. Although it is clear the lack of fusion caused by this mutation in FeLV-T env may be a reason for its inability to efficiently infect feline cells, there have been no publications regarding a Pro to His mutant of FeLV-T env that can infect feline cells in vitro. Indeed, experiments in our lab indicated that this particular mutant can infect some feline cells in vitro, albeit weakly (Rey, et al., unpublished), suggesting that the defect in fusion can be partially rescued. It is also possible that another factor is required for efficient infection of feline cells in vitro and feline T cells in vivo. Intriguingly, this defective fusion event can be rescued by the presence of soluble wildtype (wt) SU, which contains the crucial His residue in trans within the culture media 18
(Lavillette et al., 2000). Furthermore, this soluble wt SU needs not be of the same subgroup, as E-MLV SU has been shown to rescue fusion defective virus from A-MLV (Lavillette et al., 2000). Taken together, these experiments raise further questions regarding the virus and host cell interactions that take place during binding and fusion.
1.6.5.3 FeLV-B FeLV-B arises from recombination between FeLV-A and endogenous retroviral sequences within the feline genome (Elder and Mullins, 1983, Overbaugh et al., 1988, Stewart et al., 1986), and thus maintains the least similar protein sequence identity (Figure 1-4). Using probes against the divergent regions of FeLV-B env, it was discovered that the env was derived from endogenous proviral FeLV env DNA of which there are approximately fifteen different copies in the domestic cat genome (Todaro et al., 1978, Bonner and Todaro, 1979). However, these endogenous FeLV env sequences do not generate infectious virus or cause disease (Soe et al., 1983). C-terminal to the proline rich region (PRR), FeLV-B env exhibits very few amino acid changes from FeLV-A env. Due to these changes in the RBD of FeLV-B env, a different receptor is used for viral entry. FeLV- B is known to cause a mild, non-fatal anemia, but has been linked to a variety of pathogenic diseases such as lymphosarcomas and other myeloproliferative diseases (Jarrett et al., 1987 Tzavaras et al., 1989, Sheets et al., 1993, Tsatsanis et al., 1994). FeLV-B cannot be transmitted from cat to cat, as FeLV-A can, but it exists in 40% of all isolates from infected cats (Jarrett et al., 1984). This suggests that recombination events with endogenous FeLV env sequences in the feline genome of domestic cats can produce a new subgroup with alternate receptor usage at a higher frequency than simple point mutations in the FeLV-A env gene.
1.6.5.4 FeLV-C Like FeLV-T, FeLV-C arises through point mutations in the FeLV-A env, and has the broadest host range of the subgroups with the ability to infect feline, human and guinea pig and mink cells in vitro (Sarma et al., 1975). The majority of these point mutations are located within the variable regions of the RBD, the PRR, and the C2 region of Cdom of the env protein, with three residue changes located elsewhere in the SU. A study by Rigby and 19 colleagues determined that interchanging VRA of FeLV-A for that of FeLV-C could confer FeLV-C-like host range and pathogenesis while retaining some of the FeLV-A infectivity phenotype. Conversely, they also concluded that regions outside of VRA were also involved in FeLV-C infectivity and pathogenicity due to lack of a complete phenotype switch (Rigby et al., 1992). Thus, replacing the RBD of one subgroup with another does not completely confer receptor switching indicating that there are regions outside of the RBD that may be involved in receptor binding, post-binding and fusion events. FeLV-C is the most pathogenic of all the subgroups causing a severe aplastic anemia characterized by a pure red cell aplasia (PRCA) to be discussed in greater detail in Section 1.7.2. Luckily for pet owners, it is also the rarest subgroup as this subgroup is present in only 1% of isolates from infected cats (Jarrett et al., 1984). Pathogen free kittens infected with FeLV-C virus develop PRCA within eight weeks; further supporting the notion that endogenous FeLV-C virus causes PRCA (Hoover et al., 1987, Riedel et al., 1986). However, kittens that have reached eight weeks of age do not become infected with FeLV-C upon inoculation with the virus, nor do they develop PRCA (Hoover et al., 1984, Riedel et al., 1986). This suggests that infection with FeLV-C is age related such that infection must occur in the early stages of life for proper production of FeLV-C virus and disease manifestation. This could also allude to the low presentation of the virus in naturally occurring isolates, as very few FeLV- C viruses can efficiently infect enough cells and therefore cannot be detected in the samples taken from infected domestic cats.
1.7 Gammaretroviral receptors
Retroviral receptors and viral receptors in general, are the first sight of interaction between a virus and its potential host cell. For this reason, their identification, cellular expression, and role in binding of the virus have and will continue to be a major subject for virus research. The expression of the receptor in various tissues and cell types is often what determines the host range and tropism of the virus. This provides a target for therapeutic agents to interfere with the interaction between the virion and the cell to prevent viral infection. Virus binding and fusion is a multi-step process in which conformational changes 20 occur in the viral env protein and/or in the receptor upon initial contact. This can lead to secondary interactions between the virus and the target cell to strengthen the adhesion and induce fusion of the viral and host membranes, or uptake of the virus by endocytosis. An interesting feature of receptors used by gammaretroviruses is that they are all multiple membrane spanning proteins (Table 1-2). Furthermore, they have all been identified as transporters of important solutes. Characterization of gammaretroviral receptors has shown that multiple extracellular domains are required for receptor function.
Table 1-2. Gammaretroviruses and their receptors
γ-TYPE RECEPTOR # OF FUNCTION OF RETROVIRUS TM RECEPTOR
FeLV-C FLVCR1 (Tailor et al., 1999) 12 Heme exporter FeLV-B Pit1 (Takeuchi et al., 1992) 10 Inorganic phosphate GALV (O’hara et al., 1990) transporter
FeLV-A THTR1 (Ghosh et al., 1992) 12 Thiamine transporter FeLV-T Pit1 + FeLIX ? (Anderson et al., 2000) A-MLV Pit2 (Miller et al., 1994) 10 Inorganic phosphate transporter
E-MLV mCAT-1 (Albritton et al., 1989) 14 Basic amino acid transporter RD114 ASCT2 (Tailor et al., 1999) 10 Neutral amino acid transporter
X-MLV, P-MLV X-receptor (Tailor et al., 1999) 8 G-protein coupled signaling?
1.7.1 Ecotropic MLV receptor, CAT-1
The receptor for ecotropic MLV was the first to be molecularly cloned using gene- transfer techniques by Albritton and colleagues in 1989 (Albritton et al., 1989). The 21 receptor (MCAT) was proven to be an integral membrane protein that spanned the plasma membrane 14 times. It was later shown to function in sodium-independent transport of the basic amino acids lysine, arginine and ornithine (Kim et al., 1991, Wang et al., 1991). Endogenous expression of the retroviral envelope glycoprotein in infected cells reduces the efficacy of amino acid transport (Wang et al., 1992). Furthermore, mutations in CAT-1 that disrupts amino acid transport do not disrupt receptor function, suggesting that the receptor and cellular functions of CAT-1 are independent of each other (Wang et al., 1994).
1.7.2 Xenotropic and polytropic MLV receptor, XPR1
The xenotropic MLV (X-MLV) receptor was cloned independently by three different groups (Battini et al., 1999, Tailor et al., 1999, Yang et al., 1999). XPR1 is used by both X- MLV and polytropic MLV (P-MLV) for infection and is predicted to contain eight or nine TM domains. This receptor is related to yeast proteins SYG1, of unknown function, and PHO81, which has been implicated in phosphate transport, but the true function of XPR1 is still undetermined.
1.7.3 RD114, BaEV and type D retrovirus receptor, ASCT2
A multitude of gammaretroviruses including the feline endogenous virus RD-114, some avian retroviruses, the human endogenous retrovirus family (HERV-W), and the baboon endogenous virus (BaEV) all use ASCT2 (for RD-114 and D-type retrovirus receptor) as a common receptor (reviewed in Overbaugh et al., 2001). ASCT2 has been identified as a sodium-dependent neutral amino acid transporter, predicted to contain nine or ten TM domains (Rasko et al., 1999, Tailor et al., 1999). ASCT2 is expressed in most human tissues with the liver and brain being the only exceptions (Tailor et al., 1999). The use of this receptor by such diverse retroviruses has led to speculation regarding the evolution of these viruses from a common ancestor. Upon analysis of the envelope proteins from these and other retroviruses it became apparent that while diverse in their host range and virus type, their envelope proteins are relatively similar and cluster together apart from other retroviral envelope proteins (Overbaugh et al., 2001). 22
1.7.4 Feline leukemia virus receptors
1.7.4.1 FeLV-B receptor, Pit1 The receptor for FeLV-B was identified to be the same receptor used by gibbon ape leukemia virus (GALV) for infection corresponding to previous interference assays between the two viruses (Sommerfelt and Weiss, 1990, Takeuchi et al., 1992). The GALV receptor gene was cloned in 1990 (O’hara et al., 1990), and later proven to function as a sodium- dependent inorganic phosphate symporter (Kavanaugh et al., 1994) with ten predicted TM domains. These discoveries led to the renaming of the GALV receptor as Pit1. Although FeLV-B naturally infects feline cells in vivo, the virus has expanded its host range to human cells in vitro through use of the human homologue of feline Pit1, hPit1. Through analysis of the differences between the functional human Pit1 and the non-functional mouse homologue, the fourth extracellular loop was identified as the region bound by both GALV and FeLV-B during infection (Tailor et al., 1993). In some feline cells, FeLV-B has been shown to use a related receptor, Pit2 for entry (Anderson, et al., 2001, Boomer et al. 1997). Pit2 was initially characterized as a receptor for amphotropic MLV (A-MLV) in addition to Pit1 (Miller et al. 1994). Previous reports had shown non-reciprocal interference between a strain of MLV known as 10A1 and GALV which was confusing due to their usage of the same Pit1 receptor, but identification of Pit2 allowed an explanation for this phenomenon. MLV 10A1 could use both Pit1 and Pit2 for entry whilst GALV could use only Pit1.
FeLV-T appears to be the only gammaretrovirus that can require additional host cell factor secretion for infection. This T-cell tropic FeLV has been shown to infect cells expressing the FeLV-B receptor, Pit1, but only in the presence of a soluble co-factor similar to the FeLV-B SU named FeLIX (Anderson et al., 2000). The authors argue that the presence of this soluble factor renders the fusion defective FeLV-T able to bind and fuse with Pit1, allowing FeLV-T entry into the cell. This mechanism is similar to that seen with fusion defective MLV virus, whereby addition of soluble RBD enables MLV virus to infect cells. The mechanism including use of a soluble cofactor while not novel in function was the first example of a potential in vivo mechanism of entry for a virus. Unfortunately, this 23 mechanism of infection has yet to be proven in vivo, and other in vitro studies suggest the use of the FeLV-C receptor, FLVCR1, by FeLV-T in the presence of FeLV-C SU (Cheng et al., 2006). Furthermore, the addition of a soluble RBD from any gammaretrovirus has yet to be proven necessary as wt FeLV-T virus can infect feline cells without the addition of FeLIX or overexpression of Pit1 (Shojima et al., 2006). Thus, the receptor used for in vivo entry into feline T cells or in vitro entry into other feline cells by FeLV-T virus remains unclear.
1.7.4.2 FeLV-A receptor, THTR1 The receptor for FeLV-A by which the virus enters feline cells was recently cloned by Mendoza et al. and identified as a thiamine transporter, feTHTR1 (Mendoza et al. 2006, Ghosh et al., 1992). Although there is a human homologue, hTHTR1, for this receptor, it has not yet been determined if FeLV-A uses this receptor for entry into human cells, and if so, why it is not as efficient as the feline version in FeLV-A infection. Overexpression of the hTHTR1 cells did not elicit a large increase in infection of human cells above background (Mendoza et al., 2006) suggesting that additional factors are required for efficient FeLV-A infection. It is also possible that human cells may secrete some inhibitory factor that prevents efficient binding of FeLV-A virus to hTHTR1.
1.7.5 FeLV-C receptor, FLVCR1
Both the human and feline homologues of FLVCR1 have been identified and characterized (Tailor et al., 1999, Quigley et al., 2000). The human homologue has a predicted cDNA length of 2.0 kb coding for a protein with 12 transmembrane spanning domains as determined by a hydrophobicity plot (Figure 1-4). FeLV-C virus can not infect murine mus dunni tail fibroblast (MDTF) cells, despite the endogenous expression of a mouse homologue of FLVCR1, mdFLVCR1, which exhibits 77% sequence identity (Tailor et al., 2000). It has been suggested that low expression of mdFLVCR1 at subthreshold levels inhibit FeLV-C infection. Overexpression of the endogenous receptor rendered these cells permissive to infection which suggests that not only presence of the homologous receptor 24 within the genome, but also receptor expression levels, is critical in determining the host range of a virus (Tailor et al., 2000). Additionally, there may be other masking factors expressed that inhibit infection either by blocking the binding sites of the incoming env or the receptor. Our laboratory has previously identified the regions of the receptor that are critical for FeLV-C virus infection as extracellular loops 1 and 6 (Brown et al., 2006). These regions were identified through the creation of receptor hybrids between FLVCR1 and its paralog, FLVCR2. FLVCR2 contains 52% sequence identity and similar topology to FLVCR1, but its overexpression in MDTF cells does not render them permissive to FeLV-C infection. Thus, using mouse cells overexpressing the hybrid receptors, Brown and colleagues were able to identify the specific regions of FLVCR1 that were involved in FeLV-C infection. Furthermore, they showed that a single asparagine in extracellular loop 6 of FLVCR2 when mutated to the corresponding acidic aspartate from FLVCR1 is sufficient to render FLVCR2 as a functional receptor for FeLV-C virus (Brown et al., 2006). Northern blot analysis indicated expression of FLVCR1 mRNA in a wide variety of hematopoietic tissues including peripheral blood lymphocytes, bone marrow, and fetal liver, and very little expression in other tissues (Tailor et al., 1999). Both the mRNA and protein expression of FLVCR1 have been demonstrated to decrease as hematopoietic CD34+ stem cells differentiate towards mature erythroid cells (Quigley et al., 2004). These observations are consistent with FeLV-C virus infection in vivo of predominantly hematopoietic cells (Dean et al., 1992). It also coincides with, but does not explain, the pure red cell aplasia (PRCA) that is induced in cats infected with FeLV-C virus. FLVCR1 has been identified to function as a heme exporter (Quigley et al., 2004, Keel et al., 2008). While the specific function of heme export in erythropoiesis has not yet been eluded, it has been shown that block of FLVCR1 through αFLVCR1 antibody can disrupt erythroid development of the human erythroleukemia cell line, K562 (Quigley et al., 2004). The idea of FeLV-C infection leading to production of FeLV-C env that could act in a dominant negative fashion to inhibit FLVCR1 function inducing PRCA has been raised (Tailor et al., 1999, Quigley et al., 2004). Experiments performed in our lab provide evidence that FeLV-C env does indeed bind to, and inhibit the function of FLVCR1 in hematopoietic stem cells (HSCs), leading to a block in erythropoiesis similar to that found in PRCA, and will be discussed in 25
Chapter 3. The function of a heme exporter as being critical during erythropoiesis may seem counterintuitive as it has been well established that heme synthesis is required for the production of mature erythroids. Quigley and colleagues argue that FLVCR1 acts as a cell membrane export channel (overflow valve) to provide a safety mechanism for maintaining the correct levels of heme at subtoxic levels within early erythrocytes for proper maturation (Quigley et al., 2004). Furthermore, a recent research article by the same group shows that FLVCR1 knockout mice lack definitive erythropoiesis, and die mid-gestation (Keel et al., 2008). Conditional FLVCR1 knockout mice develop a severe macrocytic anemia with early erythroblast maturation (Keel et al., 2008). Taken together, these results suggest that heme export is critical for correct erythroid maturation in mice. Considering how PRCA develops in cats that are infected with FeLV-C virus, and that FLVCR1 knockout mice exhibit defects in erythropoiesis, it is clear that FLVCR1 plays a significant role in the development of erythroid cells.
Extracellular loops 1 2 3 4 5 6
Heme exporter
N C Human Chromosome 1q31.3
37Kbp
E1 E2 E3 E4 E5 E6 E7 E8 E9 E10
Figure 1-4. Predicted topology of FLVCR1. FLVCR1 contains 12 transmembrane loops with six extracellular loops. Both N- and C- termini are intracellular. The FLVCR1 gene is located on chromosome 1q31.3 and is encoded by ten exons that correspond to a cDNA length of 1.7Kbp.
26
1.7.6 Summary
Binding of the retrovirus to the target cell is critical for gene delivery in efficient viral infection. The surface glycoprotein (SU) of retroviruses mediates the interaction between the virion and its receptor on the host cell surface. By studying this interaction we could be able to create designer retrovirus envelopes that are targeted to a receptor of interest. This method could be used for targeting retroviral gene therapy to a particular tissue or organ that expresses that particular receptor. The four subgroups of the FeLVs (FeLV-A,-C.-T and B) differ only in their envelope proteins with each group binding to a different host cell surface receptor. By investigation into these subtle amino acid sequence differences, it may be possible to identify the regions of the envelope protein that are important for binding to their receptors outside of the receptor binding domain (RBD). The specific goal of this project is to comprehensively map the regions of the envelope glycoprotein of subgroup FeLV-C that are involved in binding to its cognate receptor, FLVCR1.
1.8 Diamond Blackfan Anemia
In 1938 L.K. Diamond and K.D. Blackfan described the etiology of a perplexing disease as a “congenital insufficiency of red marrow tissue and inability on the part of the hematopoietic system to respond to the need for more blood as the erythrocytes wear out” (Diamond, LK and Blackfan, KD, 1938). While there have been more specific categories used to diagnose DBA, this description of the disease holds true to this date as clinicians across the world encounter this disease. Today DBA is diagnosed often in infancy with children presenting a congential hypoplastic anemia, macrocytic anemia, reticulocytopenia, and low numbers of erythroid precursors in the bone marrow (Lipton et al., 2006). More than fifty percent of DBA patients also have short stature and/or physical abnormalities.
27
1.8.1 Clinical diagnosis and treatment methods of DBA
A pale skin colour is often the only initial symptom of DBA seen within the first year of life. Upon laboratory analysis, other findings such as decreased hemoglobin levels, elevated mean corpuscular volume (MCV) and reticulocytopenia can also be used in diagnosis of DBA in patients. A bone marrow aspirate is often performed to verify erythroid underdevelopment with normal development of myeloid and lymphoid cells. Other hematological features of DBA include elevated erythrocyte adenosine deaminase (eADA) acitivty (Glader et al., 1988), elevated fetal hemoglobin, and elevated serum erythropoietin (EPO) (Lipton et al., 2006). Although not present in every patient, growth retardation and craniofacial anomalies have also been used in the clinical diagnosis of DBA. There appears to be an increased risk of myeloid leukemia in DBA patients although a clear correlation between the diseases has not yet been shown (Vlachos et al., 2001). Transient erythroblastopenia of childhood (TEC) which is a slowly developing anemia of early childhood characterized by gradual onset of pallor is the main clinical symptom used in diagnosis of DBA (Freedman, MH, 2000). Therapuetic treatment of DBA is commenced immediately after a confirmed diagnosis. Blood transfusions followed by corticosteroid treatment are the most often prescribed methods of treatment for DBA patients. Blood transfusions re-populate the blood stream with healthy mature erythroid cells while the corticosteroid treatment induces a general change of gene expression that promote erythroid differentiation (Ohene-Abukawa et al., 2005). There exists great variability in the response to these therapeutic treatments in DBA patients. After many years of transfusions and steroid treatment, up to twenty percent of patients go into a complete remission and can maintain normal hemoglobin levels without any further treatment. Forty percent of patients become transfusion-independent, but maintain steroid treatment, while the remaining forty percent become insensitive to steroids and require constant transfusion therapies (Lipton et al., 2006). While transfusion independence in the presence of corticosteroid treatment may seem to be a somewhat ideal substitute for complete remission, there are many long-term effects of prolonged use including growth retardation, cataracts and induced Cushing’s disease in patients. Long- term blood transfusions also exhibit some drawbacks due to additional therapy required to 28 maintain normal levels of iron in the body. Iron chelating therapy is often required to counteract the long term iron acquisition caused by frequent transfusions that can lead to fatal iron overload in the liver, heart, pancreas, testicles, and the pituitary, thyroid and parathyroid glands. The only existing curative treatment is allogenic bone marrow transplantation. Unfortunately, this treatment is associated with a high risk of mortality due to graft rejection or rejection, and infection. Most importantly, this method is only available for DBA patients that can find a matching donor to supply the healthy bone marrow. Indeed, some parents have embarked upon the difficult decision to bear another child that through modern in vitro fertilization techniques has been HLA-typed to ensure a healthy child and a potential HLA- matched bone marrow donor, capable of curing the older sibling (Wasserman, D. 2003). Recently, a group in Czech Republic has published the use of leucine as a therapeutic agent in the treatment of DBA (Pospisilova, D., et al., 2007). A 7-year old girl of short stature, who was transfusion dependent and undergoing iron-chelation therapy from the age of 4, was administered leucine as a supplement to her diet. Within three weeks, there was a notable increase in appetite and weight, and within 6 months, she became transfusion independent. She has remained in remission for almost a year thus far (Pospisilova, D., et al., 2007, personal communication, 2008). While the authors are cautious to point out that they cannot remove the idea of a spontaneous remission, they have carried out their study of leucine administration to three other patients who have all shown noticeable improvements in their health and well being (Pospisilova, D., et al., 2007). While this therapeutic treatment is in its infancy, the simplicity and ease of treatment is a welcome addition to the current methods, and a full-scale clinical trial is warranted to confirm these results.
1.8.2 Erythropoiesis
Erythropoiesis is the process by which mature red blood cells are produced from hematopoietic stem cells (HSCs) (Figure 1-5). A HSC produces all cells of the blood lineage which included the myeloid, lymphoid and erythroid cells. Early on at the embryonic stage, erythropoiesis takes place in the mesodermal yolk sac, a few months later, the majority of the prenatal red blood cells are produced in the liver and spleen. During 29 adulthood, the majority of erythropoiesis occurs in the bone marrow. Erythropoiesis begins when a HSC receives some signal, intrinsic or extrinsic, to differentiate into a mature RBC. The first cell that is distinguishable as being part of the erythroid lineage and specifically leading down the RBC pathway is the proerythroblast . As development progresses, the nucleus becomes somewhat smaller and the cytoplasm becomes more basophilic (hence termed a basophilic erythroblast), due to an increase in the amount of ribosomes that are needed for the surge in translation required for proliferation. As the cell begins to produce hemoglobin, the cytoplasm attracts both basic and eosin stains, and is called a polychromatophilic erythroblast; the cell also begins to decrease in size. The cytoplasm eventually becomes more eosinophilic, and the cell is called an orthochromatic erythroblast . This orthochromatic erythroblast will then extrude its nucleus and enter the circulation as a reticulocyte . Reticulocytes are so named because these cells contain reticular networks of polyribosomes. As reticulocytes loose their polyribosomes they become mature red blood cells.
1.8.2.1 Erythropoietin A multitude of epigenetic factors control the differentially expressed genes involved in production of mature erythroid cells. Different transcription factors and signaling molecules have been identified to play a role in erythropoiesis, the spatio-temporal expression patterns of GATA-1, GATA-2 and SCL have proven critical (Flygare and Karlsson, 2007). One of the most specific and well studied factor to be linked to erythropoiesis to date is erythropoietin (Epo). Epo’s principle function in the hematopoietic system is to regulate red blood cell (RBC) production. It is a glycoprotein hormone that acts as a cytokine by binding to the erythropoietin receptor (EpoR) which has been shown to dimerize and signal through protein kinases, anti-apoptotic proteins and transcription factors. Epo is produced by the liver and the kidney and acts in a feedback mechanism to maintain red blood cell homeostasis. Furthermore, circulating RBCs can bind Epo, thus in cases of anemia where there are low numbers of circulating RBCs, the increase in unbound Epo can act as a signal to initiate production of more RBCs. It is interesting to note that while Epo is critical for mature red cell ouput through activation of a transcription factor GATA-1, it has little role in determining an HSC commitment to the erythroid lineage (Wu, 30
HSC
Erythroid Lymphoid Myeloid Stem Cell BFU-E Stem Cell
CFU-E Monocyte Granulocyte Early Eosinophil erythroblasts Basophil Megakaryocyte B-cell T-cell Late erythroblasts
Erythrocytes
Figure 1-5. Outline of hematopoiesis. Schematic of hematopoiesis starting from the hematopoietic stem cell (HSC) through the myeloid, erythroid, and lymphoid lineages. DBA and PRCA exhibit a block in erythropoiesis characterized y a decrease in the number of erythroid colony-forming units (CFU-E). et al., 1995). Knockout of either Epo or EpoR gives rise to the same levels of cells committed to the erythroid lineage as wildtype (Wu et al., 1995). This suggests that Epo plays a role in the proliferation and maintenance of mature RBC numbers, but does not act as a signal to HSCs to differentiate down the erythroid lineage. Epo has other non- hematopoietic functions; it has a role in the brain’s response to neuronal injury (Siren et al., 2001) and is also involved in the healing process of wounds (Haroon et al. 2003). With the 31 advent of recombinant DNA technologies, human Epo is now produced in mass quantities for both experimental use in cell culture, and therapeutic use in patients with anemia. It is used in treating anemia resulting from chronic kidney disease, and those who undergo dialysis. Recombinant human Epo is also used for treatment of those with autoimmune or malignant diseases, and to aid in recovery from chemotherapy or radiation treatments.
1.8.2.2 Erythroid cell surface protein expression The external membrane of an RBC contains numerous proteins that either cross the lipid bilayer one or more times or are anchored to it through a lipid tail. These proteins are loosely divided into four categories based on their functions: membrane transporters; adhesion molecules and receptors; enzymes; and structural proteins that link the membrane with the membrane skeleton. Some proteins may complete their major functions during erythropoiesis or may only be important under adverse physiological conditions, while some carry out multiple functions outside of erythropoiesis. Furthermore, some might be evolutionary relics and may no longer have significant functions. Of note, is expression of some of the specific cell surface markers that have been associated with cells of the erythroid lineage. Early erythroid progenitor cells can be enriched from whole blood or bone marrow samples through their expression of CD71, a transferrin receptor. CD71 is needed for the import of iron into the cell and it's regulated in response to intracellular iron concentration. It binds to transferrin, a glycoprotein that tightly binds iron in the blood plasma. The iron-bound transferrin is transported to the intracellular space via CD71 which due to the pH change causes the iron to become dissociated. This process of iron uptake is critical for developing erythrocytes. Prior to the erythroblast stage there are two distinguishable forms of erythroid progenitor cells, namely the erythroid burst forming and colony forming units (BFU-E and CFU-E, respectively) which are responsible for the differentiation and proliferation required in erythropoiesis. The CFU-E, which is erythropoietin-sensitive, amplifies the differentiation process in response to erythropoietic stress. The in vitro culture of erythroid cell lines has revealed that the burst-forming unit (BFU-E) is not particularly sensitive to erythropoietin stimulation, but gives rise to the CFU- E and, when stimulated, produces morphologically identifiable erythroid colonies (Stephenson et al., 1971). CD36 is cell surface marker that is appears on CFU-E, but not 32
BFU-E cells and thus can be used to distinguish the two cells. Its expression is downregulated on the more mature erythroblasts, thus it can be used to mark the appearance and disappearance of CFU-E as HSCs undergo erythropoiesis. It is important to note that while CD36 has been used extensively to monitor CFU-E in liquid culture, the marker is expressed on other lineages of the hematopoietic system including platelets and monocytes. CD36 functions as a thrombospondin receptor with roles in fatty acid and glucose metabolism. Mature erythroblasts lose CD36 and CD71 expression and acquire expression of glycophorin A (GlyA). GlyA is a membrane-spanning glycoprotein that carries the blood group antigens as N- and O-linked glycans on its extracellular surface. These markers can be used to monitor the differentiation of HSCs down the erythroid lineage when cultured.
1.8.2.3 Erythropoiesis in DBA patients DBA patients exhibit a block in erythropoiesis similar to that seen in PRCA of cats (Figure 1-5). Low numbers of erythroid blast forming units (BFU-E) or colony forming units (CFU-E) have been reported in samples taken from DBA patients suggesting a specific block for erythroid precursors to mature into erythrocytes. This blockage may be due to insufficient precursors to repopulate the blood system with mature RBCs or may be due to increased apoptosis and inability to differentiate of these early erythroid precursors. As DBA patients exhibit a block at an early stage of erythropoiesis when Epo is necessary for proliferation of mature erythroid cells, it was thought that DBA patients could be rescued by treatment with repeated doses of Epo. Unfortunately, peripheral blood mononuclear cells (PBMCs), taken from DBA patients, when cultured with Epo did not respond with a wave of erythroid proliferation as did the PBMCs taken from normal patients (Ohene-Abuakwa et al., 2005). Also, when cultured in semi-solid medium, erythroid colony size was drastically smaller in cultures with DBA cells when compared to normals, although the number of erythroid colonies that formed was similar between the two (Ohene-Abuakwa et al., 2005). These observations along with the insensitivity of BFU-E to Epo stimulation suggest that the defect in erythropoiesis seen in DBA patients is downstream of the initial cues for HSCs to differentiate into mature RBCs. Erythropoiesis cannot be rescued by a signal of early erythroid cells to proliferate, and may be manifested by an inability to differentiate and proliferate beyond a certain point, or apoptosis of those early progenitor cells. 33
1.8.3 Inheritance and genetics
The incidence of DBA ranges from 5-7 per 1 million live births across the US, France, and the United Kingdom, but has been reported to be as high as 10 per 1 million in some Scandinavian countries (Willig et al. 1999). While most cases appear to be sporadic, up to fifteen percent of patients have a family member with DBA (Orfali et al., 2004), suggesting some familial inheritance for the disease. Due to the elevated levels of Epo in DBA patients, and their lack of response to supplemental Epo treatment, it was initially postulated the DBA patients may exhibit mutations in the Epo receptor (EpoR) gene rendering it non-functional. However, analysis of the EpoR genomic DNA sequence in 23 patients from Canada and Italy along with the absence of segregation of the EpoR with DBA in two families ruled out this gene as a potential genetic cause (Dianzani et al., 1996).
1.8.4 RPS19
The first DBA gene was identified through linkage analysis to be located at chromosome 19q13.2, and was mutated in 10 of 40 patients analyzed (Gustavsson et al. 1997, Draptchinskaia et al. 1999). This gene was cloned and identified as a ribosomal protein, RPS19. More than 60 different RPS19 mutations have been discovered in 25% of DBA patients. They fall into two major categories, the first class affect expression of the Rps19 protein through gene disruption via insertions, deletions, splice site mutations, and nonsense mutations. Missense mutations comprise the second class of Rps19 mutations, and they can disrupt assembly of the protein into the 40S ribosomal subunit (Orfali et al., 2004, Chatr-Aryamontri et al., 2004, Campagnoli et al., 2004, Gazda, et al., 2004, Willig et al., 1999). Family studies suggest that haploinsufficiency may be the genetic mechanism in most DBA patients, as most patients inherit only one mutated copy of the Rps19 gene (Gazda et al. 2004). It is also plausible that some mutated forms of Rps19 may act in a dominant-negative fashion to prohibit the wildtype Rps19 protein from functioning. A murine Rps19-deficient DBA disease model was attempted by targeted disruption of the Rps19 gene (Matsson, et al., 2004). However, the homozygous knockout for the Rps19 gene was lethal as determined by the lack of blastocyst formation. The heterozygous 34
RPS19+/- mice were viable with a normal phenotype and normal erythropoiesis. Furthermore, Rps19 mRNA and protein levels were similar between the RPS19+/+ and the heterozygous RPS+/- mice (Matsson, et al., 2006). These findings raise questions to the plausibility of the haploinsufficiceny in Rps19 mutated-DBA humans, as the mouse model appears to compensate Rps19 expression with the wt RPS19 allele. Other studies using RNA interference to silence Rps19 expression in cultured human CD34+ bone marrow cells have shown that the DBA phenotype can be induced through RPS19 inhibition (Ebert et al., 2005, Flygare et al., 2005). This induced defect in erythropoiesis could be rescued by expression of a RNAi-resistant Rps19 transcript. Similar in vitro models of DBA have been attempted with erythroid leukemia TF-1 cell line containing RNAi-mediated Rps19 silencing (Miyake et al., 2005) or overexpression of Rps19 mutants in erythroleukaemic K562 cells (Kuramitsu et al., 2008, Cmejlova, et al., 2006). As RNAi mediated knockdown of Rps19 provides a decent in vitro model for DBA, it opens the field for a mechanism for a potential mouse model. Many of the experiments to date have accomplished a defect in ribosome maturation through enhanced knockdown of Rps19 function, but it is important to note that many patients do not exhibit a complete or near complete inhibition of Rps19 expression or function. Many missense mutations identified in patients have not caused any disease phenotype, especially considering patient heterozygosity (Campagnoli et al, 2008). It is important for these mutations to be verified as true mutations and not polymorphisms that exist in the general population. Indeed, one group has found that a four base pair insertion in exon 1 of the RPS19 gene that has been found in five DBA patients is in fact a common polymorphism in healthy African American individuals (Huang et al., 2006). The authors tested 186 African American samples and found 33 individuals heterozygous for this polymorphism and one homozygous (healthy) individual. This result suggests that this polymorphism may not be the cause of DBA in these patients, and care should be taken when advising potential parents or donors of the presence of this allele.
1.8.4.1 Extraribosomal functions of Rps19
The identification of a ribosomal protein that is involved in translation as being the primary cause for the pathophysiology of DBA created a great deal of controversy regarding 35 its role. In this regard, many scientists postulated that Rps19 may have functions outside of ribosome maturation that may lead to specific defects in erythropoiesis seen in DBA patients. Early yeast two hybrid analysis by two different groups have shown the Rps19 interacts with a novel nucleolar protein S19-binding protein (Maeda et al., 2006), and the Pim1 oncoprotein (Chiocchetti et al., 2005). Additionally, fibroblast growth factor 2 pull- down assays yielded RPS19 as a potential binding partner (Soulet et al., 2001). Rps19 has also been implicated in the post-apoptotic mediation of monocyte chemotaxis (Yamamoto, 2000). Taken together, these interactions suggest a wide variety of extraribosomal functions for Rps19 that may be important in determining its role in the manifestation of DBA.
1.8.5 Other Ribosomal Proteins linked to DBA
Linkage analysis has also suggested another DBA gene on chromosome 8p23.3 (Gazda et al., 2001). These results suggest that several genes outside of RPS19 may play an important role in DBA. The identification of other ribosomal proteins being mutated in DBA patients has reinforced the idea of defective ribosome synthesis/function being the cause for DBA. Mutations in Rps24 (Gazda et al., 2006), Rps17 (Cmejla et al., 2007), and Rpl35A (Farrar et al., 2008) are extremely rare. Some patients carrying mutations in RPS24 exhibit defects in pre rRNA-maturation similar to those defects seen in RPS19 mutated patients (Choesmel et al., 2008). Identification of the large ribosomal subunit, Rpl35A as being necessary for correct ribosome maturation solidified the classification of DBA as a defective ribosomal disease (Farrar et al, 2008). Precaution must be taken in the identification of these and other ribosomal proteins as being involved in DBA. The majority of the mutations are single nucleotide polymorphisms (SNPs) that cause missense mutations yet to be verified as true mutations and not polymorphisms that exist in the general population. Indeed, some of the mutations identified to date do not cause any defect in ribosome maturation and prudence should be taken when informing families of genetic analysis of these genes. Furthermore, when accounting for all the potential ribosomal mutations that have been associated with DBA, more than 60% of DBA patients are still without any genetic link for their disease. This suggests that other genes involved in functions outside of ribosomal function may be involved in the progression of DBA. 36
1.8.6 Ribosome dysfunction and DBA
The selective pure red cell aplasia seen in DBA patients arising from a ubiquitous feature such as ribosome maturation seems unlikely. It has been argued that the extensive ribosome synthesis required for proliferation and differentiation of RBC precursors makes them overly sensitive to any perturbation in ribosome synthesis and function (Morimoto et al. 2007, Ohene-Abuakawa et al., 2005). In addition, Rps19 mRNA levels is high in primitive progenitor cells and decreases in more mature erythroid cells as extensive ribosome synthesis is rapidly increased, suggesting Rps19 importance in the early stages of erythropoeisis. Thus, it is postulated that Rps19 acts as a limiting factor in early erythroid cells. During erythropoiesis, Rps19 deficiency exhibited in DBA patients may become most limiting in proerythroblasts where Rps19 expression is low (Ellis and Massey, 2006), but the demand for ribosome biogenesis is high. Thus, the cells are unable to compensate for the insufficient Rps19 expression. This insufficienct ribosome biogenesis in the erythroid precursors might trigger cell cycle arrest or apoptosis (Gazda and Sieff, 2006), or failure to properly differentiate due to dysregulated translation.
1.8.7 FLVCR1 in DBA
The earliest potential genetic linkage of FLVCR1 to DBA was described in 1974 when karyotype analysis of a 14 year old boy with DBA showed an abnormal chromosome 1 structure (Heyn et al., 1974). Due to the similarities between PRCA which is caused through a block in FLVCR1 function, and DBA, it was suggested that FLVCR1 may play a role in the progression of the disease in humans. Linkage analysis of a small group of DBA families showed segregation of the disease to chromosome 1q31, which is the location for FLVCR1 gene (Quigley et al., 2005). Sequence analysis of the exons for FLVCR1 in two of these patients did not yield any mutations that may suggest a dysfunctional FLVCR1 protein, but a complete analysis of the FLVCR1 locus is warranted to confirm these findings. Furthermore, the authors did not analyze mRNA or protein levels of FLVCR1 in their study, which could also affect the development of erythroid cells in DBA patients. 37
1.8.8 Other Bone Marrow Failure Disorders
DBA belongs to a small, but growing group of congenital bone marrow failure syndromes that have been linked to deficiencies in the translation machinery (Liu and Ellis, 2006). Other bone marrow failures do not include pure red cell aplasia as with DBA, but have similarities with other indications of the disease such as physical abnormalities, growth retardation and an increased risk for cancer. Shwachman-Diamond syndrome (SDS) patients suffer from neutropenia, mild anemia, skeletal abnormalities, and an increased risk for leukemia (Boocock et al., 2006). Linkage analysis in SDS patients led scientists to the SBDS gene which codes for a nucleolus-associated protein (Boocock et al., 2003) suggested to be involved in translation. Mutated Dyskerin protein impairs maturation of ribosomal RNA in X-linked Dyskeratosis Congenita (DKC) which leads to bone marrow failure. Most studies have linked the BM failure to malfunctioning telomerase acitivty (Montanaro et al., 2002), but it is believed that defective ribosomoal biogenesis also contributes to DKC pathophysiology (Liu and Ellis, 2006). Cartilage-Hair Hypoplasia (CHH) is another disease that involves skeletal abnormalities. CHH is associated with hypoplastic hair, lymphopenia, anemia and dwarfism (Liu and Ellis, 2006). Mutations in the RNase mitochondrial RNA processing gene (RMRP) cause CHH through dysfunctional ribosomeal assembly, altered mRNA turnover and mitochondrial DNA replication (Martin and Li, 2007). Interestingly, DBA shares some of its pathophysiology with a congenital disorder not related to bone marrow failure known as Treacher-Collins syndrome (TCS). Patients with this disease exhibit the same cranio-facial malformations seen in DBA patients. Furthermore, the disease manifests due to a defect in 18S rRNA methylation (Gonzales et al., 2005). The molecular mechanism behind these diseases and ribosomal failure is unclear, as translation is required for every cell type. It has been suggested that ribosome biogenesis may be an accessory to the progression of these diseases thus giving rise to similar additional phenotypes, but that other specific genetic factors exist that have yet to be identified that produce the specific pathophysiologies.
38
1.8.9 Summary
Although ~30% of DBA individuals have mutations in the ribosomal proteins suggesting that DBA may be caused by a defect in ribosome biogenesis and translation, the molecular mechanism behind DBA remains unknown. Pure Red Cell Aplasia (PRCA) is a disease found in cats characterized by a block of FLVCR1 function and is clinically identical to DBA. Furthermore, linkage analysis in DBA families may pinpoint a genetic defect to FLVCR1. The goal of this project was to analyze the role of FLVCR1 in DBA through genetic and expression analysis of FLVCR1 in DBA patients to ascertain any defects that may contribute to the pathophysiology of the disease.
2. THE C DOMAIN IN THE SURFACE ENVELOPE GLYCOPROTEIN OF SUBGROUP C FELINE LEUKEMIA VIRUS IS A SECOND RECEPTOR-BINDING DOMAIN.
A portion of the work presented in this chapter has been previously published:
Rey MA, Prasad R, Tailor CS. The C domain in the surface envelope glycoprotein of subgroup C feline leukemia virus is a second receptor-binding domain. Virology. 2008 Jan 20;370(2):273-84. Epub 2007 Oct 22.
Author contributions / Acknowledgements:
All constructs, cell lines, infection assays, binding assays and protein expression assays were designed and performed by MAR with the exception of the analysis of the surface expression of hybrid envelopes by FACS analysis (Figure 2B) and protein expression analysis (Figure 5A) which was performed by RP. We are grateful to Maribeth V. Eiden and James Ellis for their critical reading of the manuscript. We are also grateful to Brian Willett (University of Glasgow, Scotland, UK) for providing the FeLV-C/Sarma envelope construct, to Yasuhiro Takeuchi and Francois Cosset for providing the TELCeB6 packaging cells. We would like to thank Julie Overbaugh for providing the FeLV-A receptor cDNA and for providing pCS-HA expression vector for expression of FeLV SUs. We are also grateful to Simon Duffy, Naveen Hussain, and Zvi Shalev for helpful suggestions and to Sabah Asad for his assistance in using the fluorescence-activated cell sorter.
39 40
2.1 Abstract
The receptor-binding domain (RBD) in the surface (SU) subunit of gammaretrovirus envelope glycoprotein is critical for determining the host receptor specificity of the virus. This domain is separated from the carboxy terminal C domain (Cdom) of SU by a proline- rich region. In this study, I show that the Cdom region in the SU from subgroup C feline leukemia virus (FeLV-C) forms a second receptor-binding domain that is distinct from its RBD, and which can independently bind to its host receptor FLVCR1, in the absence of RBD. Furthermore, my results suggest that residues located in the C2 disulfide-bonded loop in FeLV-C Cdom are critical for SU binding to FLVCR1 and for virus infection. I propose that binding of FeLV-C SU to FLVCR1 involves interaction of two receptor-binding domains (RBD and Cdom) with FLVCR1, and that this mechanism of interaction is conserved for other gammaretroviruses. My results could have important implications for designing gammaretrovirus vectors that can efficiently infect specific target cells.
2.2 Introduction Infection of a cell by a retrovirus is initiated by binding of the surface (SU) subunit of the viral envelope glycoprotein (Env) with a specific host cell surface receptor, followed by fusion of the virus and host cell membranes, which is mediated by the transmembrane (TM) subunit of Env. For gammaretroviruses (γ-retroviruses), receptor binding and specificity is controlled by the receptor-binding domain (RBD) located in the N terminal 200–250 amino acids of γ-retrovirus SU (Battini et al., 1992 and Battini et al., 1995). The RBDs from murine leukemia virus (MLV) and feline leukemia virus (FeLV) are highly conserved in sequence and differ predominantly in distinct variable regions defined as VRA, VRB and VRC (Battini et al., 1992 and Fass et al., 1997). These variable regions contain critical residues that control receptor binding and specificity (Battini et al., 1998, Brojatsch et al., 1992, Rigby et al., 1992, Tailor and Kabat, 1997 and Tailor et al., 2000). The RBD forms a distinct globular domain (Barnett et al., 2001 and Fass et al., 1997) that is anchored to the carboxy terminal C domain (Cdom) of MLV and FeLV SU, by a proline-rich region (PRR) that has been suggested to form a “hinge” structure. PRR and Cdom are also highly variable between MLVs and FeLVs.
41
The intricate mechanism of how γ-retrovirus SUs bind to their receptors and trigger the virus fusion mechanism has yet to be fully elucidated. A histidine residue in a PHQ motif, located in the N terminus of RBD, has been shown to be critical for triggering virus fusion (Bae et al., 1997, Lavillette et al., 2000 and Zavorotinskaya and Albritton, 1999). Mutation or deletion of this single residue disrupts virus fusion but not receptor binding. Interestingly, infection of these fusion defective viruses (ΔH viruses) can be rescued if soluble SU or RBD is provided in trans (Barnett and Cunningham, 2001, Barnett et al., 2001, Lavillette et al., 2001 and Lavillette et al., 2000). For some ΔH viruses, fusion can be activated in the presence of heterologous soluble RBDs, but only if both the receptors for ΔH virus and the soluble RBD are present. Subsequent studies have suggested a model for MLV Env fusion activation that involves an initial interaction of RBD with the host receptor followed by a second interaction of RBD with a disulfide-bonded loop (C2 loop) located in MLV Cdom (Barnett and Cunningham, 2001, Barnett et al., 2001 and Lavillette et al., 2001). This model is based on studies using soluble ecotropic MLV RBD which can activate fusion of ΔH amphotropic MLV with the C2 loop substituted for ecotropic C2 loop, and of MLVs with the RBD deleted (Barnett and Cunningham, 2001 and Barnett et al., 2001). However, a direct interaction between RBD and the C2 loop containing Cdom has yet to be demonstrated. Furthermore, the MLV fusion model is inconsistent with other MLVs because soluble RBDs from amphotropic and xenotropic MLV are unable to activate fusion of any RBD deleted MLVs including RBD-deleted amphotropic and xenotropic MLVs (Barnett and Cunningham, 2001). This raises the possibility of an alternative mechanism of envelope–receptor interaction for these γ-retroviruses. Recent reports suggest that, in addition to RBD, the C terminal domain of γ- retrovirus SU may also control receptor recognition and virus infection. Studies characterizing SUs from subgroup B FeLV and from pig endogenous retroviruses have shown that the C terminal domain of these SUs control recognition of receptor homologues from certain species (Boomer et al., 1997, Faix et al., 2002, Gemeniano et al., 2006 and Sugai et al., 2001). Additional studies have shown that the C2 loop in C terminal domain of FeLV-A and FeLV-T SUs controls infection of target cells (Cheng et al., 2006). In this study, I investigated the role of the Cdom region of FeLV-C SU in binding to the FeLV-C receptor FLVCR1 (Quigley et al., 2000 and Tailor et al., 1999), and in virus
42 infection. FLVCR1 has been identified as an exporter of heme and has been suggested to be critical for development of erythroid progenitor cells (Quigley et al., 2004). Hybrid Envs were generated between FeLV-C Env and the closely related FeLV-A or -T Envs, and subsequently tested for their ability to interfere with FeLV-C infection, and to bind to and infect FLVCR1-expressing cells. My results suggest that FeLV-C Cdom forms a receptor- binding domain that is distinct from FeLV-C RBD. Furthermore, I show that Cdom is critical for efficient SU binding and for virus infection. I propose that Cdom in FeLV-C SU functions as second receptor-binding domain, which in addition to RBD, interacts with the host receptor to initiate virus infection.
2.3 Materials and Methods
2.3.1 Cell lines
TELCeB6, feline kidney CCC S+L− and murine M. dunni tail fibroblast (MDTF) cells were maintained in Dulbecco's minimal essential medium with low glucose (1000 mg/ml) supplemented with 10% heat-inactivated fetal bovine serum (FBS). Phoenix ampho packaging cells (provided by Garry Nolan, Stanford University) producing replication- defective amphotropic MLV were maintained in Dulbecco's minimal essential medium with high glucose (4500 mg/ml) supplemented with 10% FBS. TELCeB6 cells are retroviral- packaging cells that do not contain retroviral envelope genes but produce non-infectious virus (Cosset et al., 1995). These cells were maintained using 6 μg/ml of blasticidine to ensure selection of gag–pol-expressing cells. MDTF cells expressing human FLVCR1 with a HA epitope tag were generated as previously described (Brown et al., 2006). These cells were used for FeLV SU binding and for virus infection assays.
2.3.2 Construction of hybrid FeLV envelopes cDNAs encoding hybrid FeLV envelopes containing the first 155 residues of receptor- binding domain (RBD155) from FeLV-C Env and the last 40 residues of RBD, proline-rich region (PRR), C domain (Cdom) and transmembrane (TM) segment from FeLV-A or FeLV-
43
T Envs were generated by first introducing an EcoRI restriction site (GAATTC) just downstream of cDNA encoding VRB (see Figure 2-1A). Introduction of this site caused glutamine 155 in the LQFT motif to be mutated to a glutamate residue to give a LEFT motif. The LEFT motif is present in all MLV Envs (Battini et al., 1992), and has previously been introduced in FeLV-B Env to generate functional hybrid FeLV-B/amphotropic MLV Envs (Tailor and Kabat, 1997). cDNA encoding FeLV-C RBD155 was amplified by PCR using the upstream primer 5′ GGGGGGATCCATCAAGATGG AAAGTCCAACGCACCCA-3′, which contains a BamHI restriction site and the downstream primer 5′- CCCCCTGGTCTTGGAATTCACCCAGAAG-3′ containing an EcoRI site. cDNA encoding the C-terminal region of FeLV-A, or -T Envs encompassing the last 40 residues of RBD, PRR, Cdom and TM was amplified by PCR from the FeLV-A 61E, and EECC (FeLV-T) Env cDNAs (provided by Julie Overbaugh, University of Washington, Seattle, WA). The upstream primer 5′-CTTCTGGGTGAATTCCAAGACCAGGGG-3′ contains an EcoRI site where the downstream primer 5′-GGGGCTCGAGTCATGGTCGGTCCGAATCGTATTG- 3′ contains a XhoI site. The amplified cDNAs was digested with EcoRI and XhoI and ligated to FeLV-C RBD155 cDNA digested with BamHI and EcoRI and cloned into the pFBneo retroviral vector (Stratagene) digested with BamHI and XhoI. This generated the respective CA and CT hybrid Env cDNAs. The FeLV-C Env cDNA was also cloned into the pFBneo expression vector to give pFBneo FeLV-C Env plasmid. cDNA encoding CAC and CTC Envs were generated by PCR. Briefly, cDNA encoding the C domain and TM from FeLV-C Env was amplified by PCR using pFBneo FeLV-C Env plasmid as a template and the upstream primer 5′-TATTACGAAGGGATTGCAATCTTAGGTAACTAC-3′, which contains a BsBI site, and the downstream primer 5′-GCCAGGTTTCCGGGCCTCAC-3′ which primes to the 3′ end of the multiple cloning site in pFBneo. The subsequent amplified cDNA was digested with BsBI and XhoI and ligated into a BsBI and XhoI cut pFBneo CA or pFBneo CT plasmids. Digestion of these plasmids with BsBI and XhoI removes the cDNA encoding the FeLV-A or FeLV-T Cdom and TM sequences and allows ligation of cDNA encoding FeLV-C Cdom and TM sequences. The cDNA encoding CTCC2 Env (Fig. 1B) and CΔC2 Env was also generated by PCR using mutant forward and complementary reverse primers that spanned the C2 loop cDNA sequence. The CTCC2 Env was also cloned
44
A. signal peptide start of SU
FeLV-A MESPTHPKPSKDKTLSWNLVFLVGILVGILFTIDIGMANPSPHQIYNVTWVITNVQTNTQANAT LFTIDIGMANPSPHQIYNVTWVITNVQTNTQANAT 27 FeLV-T MESPTHPKPSKDKTLSWNLVFLVGI LFTIDIGMANPSPPQMYNVTWVITNVQTNTQANAT 27 FeLV-C MESPTHPKPSKDKTFPWNLVFLVGI LFQIDMGMANPSPHQVYNVTWVITNVQTNSRANAT 27 VRA FeLV-A 28 SMLGTLTDVYP TLHVDLCDLVGDTWEP I VL S P TNVKHGARYP S SKYGCKTTDRKKQQQTY 87 FeLV-T 28 SMLGTLTDVYP TLHVDLTLHVDLCDLVGDTWEPMVLSCDLVGDTWEPMVLSPT- P T ------G-YPPSKYGCKTTDRKKQQQTY G - YP P SKYGCKTTDRKKQQQTY 81 FeLV-C 28 SMLGTLTDAYP TLYVDLCDLVGDTWEP I APDPR S - - -WARYS S S THGCKTTDRKKQQQTY 84 VRC VRB FeLV-A 88 P FYVC PGHAP S LGPKGTHCGGAQDGFCAAWGC ETTGEAWWKP S S SWDYI TVKRGS SQDNN 147 FeLV-T 82 P FYVC PGHAP S LGPKGTHCGGAQDGFCAAWGC ETTGEAWWKP S S SWDYI TVKRGSTVKRGSSQDNN SQDNN 141 FeLV-C 85 P FYVC PGHAP SMGPKGTYCGGAQDGFCAAWGC ETTGEAWWKP TS SWDYI TVKRGSNQDNS 144 VRB PRR FeLV-A 148 CEGKCNPLI LQFTQKGKQASWDGPKMWGLRLYRTGYDP* I ALFTVSRQVSTI TPPQAMGPN 207 FeLV-T 142 CEGKCNPLI LQFTQKGKQASWDGPKMWGLRLYRTGYDP I ALFTVSRQVSTI TPPQAMGPN 201 FeLV-C 145 CKGKCNPLVLQFTQKGRQASWDRPKMWGLRLYRSGYDPCKGKCNP LVLQF TQKGRQASWDR PKMWGLR LYR S GYD P I ALFSVSRQVMTIA L F S V S R QVMT I TPPQAMGPNT P P QAMG P N 204 hybrid junction PRR C domain FeLV-A 208 LVLPDQKPPSRQSQTGSKVATQRPQTNESAPRSVAPTTVGPKR IGTGDRLINLVQGTYLA 267 FeLV-T 202 LVLPDQKPPSRQSQTGSKVATQRPQTNESAPRSVAPTTVGPKR IGTGDRLINLVQGTYLA 261 FeLV-C 205 LVLPDQKPPSRQSQTKSKVTTQRPQI TSSTPRSVASATMGPKRIGTGDRLINLVQGTYLA 264 C domain FeLV-A 268 LNATDPNKTKDCWLCLVSRPPYYEGIAILNATDPNKTKDCWLCLVSRPPYYEGI A I LGNYSNQTNPPPSCLS I PQHKLTIPQHKLT I SEVSGQGL 327 FeLV-T 262 LNATDPNKTKDCWLCLVSRPPYYEGIAI LGNYSNQTNPPPSC I S I PPHKLTI SEVSGQGL 321 FeLV-C 265 LNATDPNKTKDCWLCLVSRPPYYEGIAVLGNYSNQTNPPPSCLSTPQHKLTI SEVSGQGL 323 S-S C2 loop FeLV-A 328 C IGTVPKTHQALCNKTQQGHTGAH------YLAAPNGTYWACNTGLTPC I SMAVLNWTSD 381 FeLV-T 322 C IGTVPKTHQALCNKTHQGHTGADYLTAPRYLAAPNGTYWACNTGLTPCIGTVPKTHQALCNKTHQGHTGADYLTAPRYLAAPNGTYWACNTGLTPC I SMAVLNLTSD 381 FeLV-C 324 C IGTVPKTHQALCKKTQKGHKGTH------YLAAPNGTYWACNTGLTPC I SMAVLNWTSD 378 C domain FeLV-A 382 FCVL I ELWPRVTYHQPEYVYTHFAKAVRFR ------411 FeLV-T 382 FCVL I ELWPRVTYHQPEYVYTHFAKAGRFR ------411 FeLV-C 379 FCVL I ELWPRVTYHQPEYI YTHFDKAVRFR ------408
RBD155 PRR Cdom TM RBD155 PRR Cdom TM B. C Env CT Env
A Env CTC Env
T Env CTCC2 Env
CA Env C-SU
CAC Env CT-SU
CTC-SU
AC-SU
Cdom SU
Figure 2-1. Alignment of SU protein sequence from FeLV-A, -C and -T and construction of hybrid FeLV Envs. Alignment of FeLV-A, -C and -T surface (SU) protein sequence. Identical amino acids are shaded and the respective variable regions VRA, VRB, VRC, proline rich region (PRR) and C domain are boxed and labeled. The C2 disulfide bonded loop is also shown. The numbering of the residues begin from the start of SU. The hybrid junction represents the site used to generate specific hybrid Envs and SUs. The glutamine residue (*) in the LQFT motif was mutated to a glutamate residue to generate a LEFT motif. This mutation introduces a unique EcoRI site in the encoded cDNAs encoding the FeLV-A, -C and -T Envs that was used to generate hybrid FeLV Env constructs. The dots represent potential N-linked glycosylation sites in FeLV-C SU. B. Diagram of hybrid FeLV Envs and SUs. Hybrid Envs contain the first 155 residues of FeLV-C RBD fused to the C terminal PRR, C domain (Cdom), and transmembrane (TM) regions from FeLV-A Env or FeLV-T Env.
45
C Env, A Env, T Env respectively represents FeLV-C, -A and -T Envs. CTCC2 Env was generated by replacing the FeLV-T C2 loop sequence in CT Env to FeLV-C C2 loop. The FeLV-C SU (C-SU) and all other SUs contain a double HA epitope fusedat the C terminus to allow detection of the protein in Western and SU binding assays. The Cdom SU contains the first 16 residues of FeLV-C SU fused to the C domain. into the pFBneo vector. For generation of pseudotype viruses bearing hybrid FeLV Envs, the cDNAs encoding the hybrid FeLV Envs were cloned into the pFBsalf retroviral expression vector (Cosset et al., 1995), which was subsequently introduced into the TELCeB6 retroviral packaging cells using the PolyFect transfection reagent (Qiagen). FeLV-C env Cdom and Cdom+TM constructs were constructed by PCR with the addition of the leader sequence as previously described to ensure proper trafficking to the cell surface (Grange et al., 2000).
2.3.3 Generation of feline cells expressing hybrid FeLV Envs
Phoenix-ampho retrovirus packaging cells were transfected with pFBneo expression constructs containing hybrid FeLV Env cDNAs. Virus supernatant was harvested 48 h post transfection, filtered with a 0.45-μm filter, and then used for infection of feline CCC cells. Infected cells were selected using G418 (1.5 mg/ml) and resistant cells were pooled and assayed for susceptibility to β-galactosidase encoding FeLV-C Sarma pseudotype virus (Tailor et al., 1999).
2.3.4 Viruses and infection studies
β-Galactosidase encoding hybrid FeLV Envs were generated by transfection of TELCeB6 cells with the respective pFBsalf Env expression constructs. Transfectants were selected using phleomycin (50 μg/ml), resistant colonies were pooled, virus supernatant harvested, filtered using a 0.45-μm filter and then subsequently used for infection studies. Target cells were seeded in a 24-well plate at 1 × 104 cells/well one day prior to the infection study. The following day, target cells were incubated with 1 ml of serially diluted lacZ pseudotype virus supernatant for 4 h in the presence of polybrene (8 μg/ml). The virus supernatant was then replaced with fresh growth medium, and cells were allowed to incubate for a further 2 days before X-gal (5-bromo-4-chloro-3-indoyl-β-d- galactopyranoside) (Sigma-Aldrich, Canada) staining. LacZ pseudotype titers were
46 determined by counting the number of blue colony-forming units (CFU's), and titers were expressed as the number of CFUs obtained per milliliter of undiluted virus supernatant. Virus titer results reported are an average of three independent experiments.
2.3.5 Analysis of surface expression of hybrid FeLV Envs
To confirm surface expression of FeLV-C and hybrid FeLV Envs on feline CCC cells, we used the monoclonal C11D8 anti FeLV gp70 antibody (provided by Custom Monoclonals International, Sacramento, CA) (Grant et al., 1983) and a fluorescein-conjugated donkey anti-mouse antibody for detection of hybrid FeLV Env proteins. Briefly, approximately 5 × 105 cells were dislodged from cell culture plate using a cell dissociation buffer (Invitrogen). Cells were then incubated with the respective C11D8 antibody followed by the secondary antibody (see above). Cells were then analyzed by flow cytometry. An increase in fluorescence denotes expression of SU on the surface of the feline cells.
2.3.6 Analysis of Env proteins
Env proteins were also analyzed by Western blot analysis. Target cells were dislodged from culture flask using Trypsin/EDTA (Invitrogen) and subsequently pelleted by centrifugation at 1000 rpm for 5 min. Lysates of cells were prepared by incubating cells with 200 μl of cell lysis buffer (20 mM Tris–HCl pH 7.5, 1% Triton X-100, 0.05% SDS, 5 mg/ml sodium deoxycholate, 150 mM NaCl, 1 mM PMSF) for 10 min on ice. The lysate was centrifuged at 13,000×g for 10 min at 4 °C to remove cellular genomic DNA, and the supernatant was stored at − 80 °C for later use. Approximately 100 μg of total protein, as determined by Bradford protein assay, was loaded onto a 10% SDS–polyacrylamide gel. Proteins were transferred to a nitrocellulose membrane (Pall, Pensacola, FL) and analyzed using a 1:500 diluted monoclonal C11D8 anti FeLV gp70 antibody, followed by a 1:1000 diluted goat anti-mouse antibody conjugated to HRP (Serotec). Signals were detected using chemiluminescence reagent (Perkin-Elmer, Boston, MA), followed by exposure to Kodak Biomax MR film. For loading control of cell lysate samples, the nitrocellulose membrane was incubated with 1 in 1000 diluted anti-actin monoclonal antibody (Sigma-Aldrich)
47 followed by 1 in 1000 diluted goat anti-mouse HRP (Sigma-Aldrich). Env protein incorporated into lacZ pseudotype virus particles was analyzed by first harvesting 10 ml of virus-containing supernatant, which was subsequently filtered using a 0.45-μm filter. Virus particles were then pelleted by centrifugation in a Beckman SW60Ti rotor at 28,500×g for 90 min at 4 °C. Virus pellets were resuspended in 100 μl of PBS and protein concentration determined by Bradford Assay (Bio-Rad Laboratories). Approximately 20 μg of protein was loaded in each well of a 10% SDS–polyacrylamide gel and protein transferred to a nitrocellulose membrane. Env protein was visualized as described above. The MLV capsid protein, which was used as loading control, was detected using the anti-p30(capsid) antibody (gift from Dr. Francois-Loic Cosset).
2.3.7 SU binding assay
For binding assay shown in Figure. 3, cDNAs encoding FeLV-C, CT or CTC SU were amplified using the upstream primer 5′- GGGGCTCGAGCACCATGGAAAGTCCAACGCACCCAAAA-3′, and the downstream primer 5′- GGGGGCGGCCGCTCAAGCGTAATCTGGTACGTCGTATGGGTAAGCGTAATCTGG TACGTCGTATGGGTATCTAGAGTAAATATATTCGGGTTGATGGTA-3′, which contained sequence encoding a double hemagglutinin (HA) epitope. These three constructs were subsequently cloned into an XhoI–NotI digested pFBneo vector. Soluble HA-tagged SUs were generated by transfection of HEK293T cells with the respective pFBneo expression constructs, and culture medium containing the SUs was isolated 48 h post transfection. Soluble SU containing culture medium was filtered using a 0.45-μm filter and subsequently used for SU binding assay or stored at − 80 °C for later use. For SU binding assays shown in Fig. 5, the respective SU encoding cDNAs were cloned into the pCS-HA expression vector that contains a double HA sequence using procedures previously described (Sugai et al., 2001). cDNA encoding FeLV-C RBD containing a V5 epitope tag was generated using the upstream primer 5′-GGGGGGGAATTCGGCA CCATGGAAAGTCCA-3′ and downstream primer 5′- CCCCCCGCGGCCGCTCAGGTGCTGTCCAGGCCCAGCAGGGGGTTGGGGATGGGC
48
TTGCCATCAGGTAAGACTAA-3′ primer which contains the V5 epitope. FeLV-C RBD cDNA was subsequently cloned into EcoRI–NotI digested pFBneo retroviral expression vector. All soluble SUs and FeLV-C RBD were expressed as described above. Binding of soluble SUs or FeLV-C RBD to target cells was carried out as previously described (Brown et al., 2006). Briefly, target cells were dislodged from culture flask using a cell dissociation buffer (Invitrogen) and approximately 1 × 106 cells were incubated with 1 ml of SU containing medium in the presence of polybrene (8 μg/ml) for 30 min at 37 °C. Cells were then centrifuged at 4000 rpm for 3 min and then washed twice with wash buffer (PBS containing 2% FBS). Cells were incubated with a 1 in 200 diluted monoclonal anti HA.11 antibody (Covance, Berkley, CA) or monoclonal anti-V5 antibody (Invitrogen) for 30 min at 4 °C with gentle agitation every 10 min. Cells were then subsequently washed twice with wash buffer and then incubated for 30 min at 4 °C with a fluorescein-conjugated donkey anti-mouse antibody. The cells were then subsequently washed twice, fixed with 1% paraformaldehyde, and subsequently analyzed by flow cytometry.
2.3.8 Immunoprecipitation of soluble SU protein
Approximately 1 ml of the HA-tagged SU or V5-tagged FeLV-C RBD protein was precleared with 100 μg/ml of 50% protein G-Sepharose (Amersham Biosciences) suspension for 3 h at 4 °C. The supernatants were then immunoprecipitated for 2 h at 4°C with bound anti-gp70 antibody–bead complex for isolation of full-length soluble SU proteins or by using bound anti-HA or anti V5 antibody–bead complex for respective isolation of soluble Cdom SU and soluble FeLV-C RBD. Immunoprecipitates were washed three times in PBS containing 0.1% Triton X-100 and 0.1% NP40. The bound SU proteins were eluted in 50 μl of sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) sample buffer and heating the sample to 100°C. Approximately 25 μl of sample was analyzed by Western blot analysis using 1:100 diluted anti-HA antibody conjugated to horseradish peroxidase (Sigma).
49
2.3.9 C2 peptide synthesis
Chemical synthesis of C2 peptide was performed by solid-phase synthesis at Chemicon. The synthesized peptide comprised amino acids 336 to 359 (CKKTQKGHKGTHYLAAPNGTYWAC) of the C2 loop of FeLV-C env. Analytical reversed-phase HPLC of the purified C2 peptide showed a homogeneity of >98%. Peptide was dissolved in PBS and added to culture media at final concentrations ranging from 10 to 150 μM.
2.4 Results
2.4.1 Cdom is necessary for efficient interference with FeLV-C virus
To comprehensively ascertain the role of the C terminal region of FeLV SU in receptor binding and virus infection, I generated specific hybrid Envs and SUs between the closely related Envs from FeLV-A, -C and -T, in which the C terminal region encompassing PRR and C domain was substituted. These FeLV Envs were selected for this study because first, their SUs share 86–95% amino acid identity (Figure 2-1A) yet they recognize distinct receptors (Anderson et al., 2000, Mendoza et al., 2006, Quigley et al., 2000 and Tailor et al., 1999), and second, sequence divergence between these Envs is primarily confined to VRA, located in RBD, and to the C2 loop located in the C domain (Cdom) (Figure 2-1A). In this study, I focused on characterizing the Cdom of FeLV-C SU for its role in binding to FLVCR1, and its role in virus infection of murine Mus dunni tail fibroblast (MDTF) cells expressing human FLVCR1 (MDTF/hCR1). I initially generated the CA and CT hybrid Envs that are spliced just downstream of VRB (Figures 2-1A and B). These Envs contain the first 155 residues of FeLV-C RBD (RBD155), encompassing VRA, VRB and VRC, which is fused to the C terminal domain from FeLV-A or FeLV-T Envs that encompasses the last 40 residues of RBD, PRR, Cdom and TM regions (Figure 2-1B). The design of the hybrid FeLV Envs were based on our previous studies characterizing the functional domains of FeLV-B envelope protein ([Tailor and Kabat, 1997] and [Tailor et al., 2000]).
50
A.
5
4
3 CFU/ml) 10 2 (log
Titer of lacZ(FeLV-C) 1 C C2 CT CA CTC CAC CTC No Env Env B.
C CAC CA Cell number Cell
CTC CT CTCC2
Surface expression
Figure 2-2. Interference property of hybrid FeLV Env. A. Specific hybrid Envs were expressed in feline kidney CCC cells, which were subsequently tested for susceptibility to β-galactosidase encoding FeLV-C pseudotype virus. Titers are mean of three infection experiments and are represented as colony forming units (CFU) per milliliter of virus medium. FeLV-C Env is represented as C Env B. Surface expression of hybrid FeLV Env proteins on feline CCC cells. Env protein expression was analyzed by flow cytometry using the monoclonal C11D8 antibody followed by a donkey anti-mouse antibody conjugated to FITC. Black histogram represents control feline CCC cells. White histogram represents surface FeLV Env expression.
51
To assess the receptor-binding properties of CA and CT Envs, and of all subsequent hybrid Envs, I first tested their ability to interfere with FeLV-C infection. I expressed the CA and CT Envs in feline kidney CCC cells and assessed their ability to interfere with infection of β-galactosidase encoding FeLV-C [lacZ(FeLV-C)] (Figure 2-2A). Interference of FeLV-C infection would give indirect evidence of the ability of CA and CT Envs to bind to FLVCR1. As expected, feline CCC cells expressing FeLV-C Env (C Env) were weakly susceptible to lacZ(FeLV-C) infection, whereas feline cells expressing no Env were highly susceptible (Figure 2-2A). FeLV-C infection titers were approximately 1000-fold lower on C Env-expressing cells than titers on parental feline cells. As shown in Figure 2-2A, feline cells expressing CA Env were 50-fold more susceptible to lacZ(FeLV-C) than C Env- expressing cells, whereas cells expressing CT Env were approximately 500-fold more susceptible. These findings suggest that substitution of C terminal of FeLV-C Env with the corresponding related domains from FeLV-A or FeLV-T Envs disrupts FeLV-C interference.
To assess the minimal sequence that would restore FeLV-C interference, I first introduced FeLV-C Cdom and TM sequence in CA and CT Envs to respectively generate the CAC and CTC Envs (Figure 2-1B). Expression of CAC or CTC Envs in feline cells enhanced FeLV-C interference by 10-fold when compared to interference caused by CA and CT Envs, respectively. Because the TM sequences from FeLV-A, -C, and -T are highly conserved and are not involved in receptor binding, I inferred from these results that re- introduction of FeLV-C Cdom in CA and CT Envs was responsible for the enhancement in FeLV-C interference. Because the C2 loop from FeLV-A, -C and -T Envs are highly divergent (Figure 2-1A), I hypothesized that the C2 loop of FeLV-C Env was responsible for enhancing FeLV-C interference. Thus, I generated the CTCC2 hybrid Env, which contains the C2 loop from FeLV-C in a CT Env backbone (Figure 2-1B). Expression of CTCC2 Env in feline cells also enhanced interference by 10-fold when compared to interference caused by CT Env (Figure 2-2A). These results suggest that the residues within the C2 loop are critical for enhancing virus interference.
52
The differences in interference properties of the hybrid Envs could be caused by differences in the expression level or processing of the Envs in feline cells. Thus, I analyzed Env expression on the surface of feline cells. Detection of surface Env expression would indicate that the hybrid Envs are expressed and correctly processed. Surface expression of hybrid Envs was analyzed by flow cytometry using the C11D8 FeLV gp70-specific antibody, which recognizes the MGPNL motif in FeLV PRR (Grant et al., 1983). Figure 2- 2B shows surface expression of hybrid FeLV Envs on feline cells. The white histogram represents cells expressing FeLV Env, whereas the black histogram represents feline cells expressing no Env. An increase in fluorescence denotes increase in surface Env expression. As shown in Figure 2-2B, all hybrid FeLV Envs were expressed efficiently on the cell surface suggesting that they are correctly processed. Using the NIH Image J software, I calculated the percent surface Env expression by determining the percent change in Env expression between cells expressing Env (Figure 2-2C, distance Z) compared to cells expressing no Env (Figure 2-2C, distance Y). We found that surface expression of CTC Env was approximately 29–45% lower than CA, CAC, CT, and CTCC2 surface Envs but comparable with C Env surface expression. Taken together, because interference is governed by the ability of Env to bind to the host receptor, I inferred from these results that the C2 loop residues in Cdom of FeLV-C Env were critical for Env binding to FLVCR1.
2.4.2 Cdom controls efficiency of FeLV-C SU binding to FLVCR1
To provide a more direct evidence that Cdom is involved in FeLV-C SU binding to FLVCR1, I generated soluble SU proteins of FeLV-C, CT and CTC that were tagged with a double hemagglutinin (HA) epitope at the C terminus (see Figure 2-1B, C-SU, CT-SU and CTC-SU). Using flow cytometry, I measured the ability of 1 ml of SU containing culture medium to bind to MDTF/hCR1 cells (see Materials and methods). An increase in fluorescence is indicative of SU binding. As shown in Figure 2-3A, both C-SU and CTC-SU bound to FLVCR1-expressing cells, with approximate equal level. However, I observed a marked reduction in binding of CT-SU to MDTF/hCR1 as shown by the reduced level of fluorescence intensity (Figure 2-3A). I determined the amount of soluble SU present in 1 ml of SU-containing medium by Western analysis of immunoprecipitated SU proteins. As
53
Figure 2-3. Binding of soluble hybrid FeLV SU proteins to MDTF cells expressing human FLVCR1. A. FLVCR1 expressing MDTF cells were incubated with (white histogram) or without (black histogram)soluble SU proteins tagged with a double HA epitope. Bound SU protein was detected using monoclonal anti-FeLV gp70 antibody and a fluorescein conjugated donkey anti-mouse antibody. An increase in fluorescence intensity correlates with an increase in SU binding efficiency. C-SU represents FeLV-C SU. B. Western blot of immunoprecipitated SU proteins. Approximately one milliliter of medium containing soluble SU protein was immunoprecipitated using anti FeLV gp70 bound to a Sepharose G column. SU proteins were analyzed by Western using anti-HA antibody conjugated to horse radish peroxidase. The control is immunoprecipitation using one milliliter of medium from untransfected HEK293 cells.
54 shown in Figure 2-3B, I detected approximate equivalent amounts of C, CT and CTC SU proteins. These results suggest that the presence of FeLV-C is critical for efficient binding of SU to FLVCR1.
2.4.3 Cdom is critical for FeLV-C infection of target cells
I next assessed the infection properties of β-galactosidase encoding retroviruses bearing C, CA, CT CTC or CTCC2 Envs on MDTF/hCR1. As expected, pseudotype virus bearing C Env (C virus) efficiently infected MDTF/hCR1 (Figure 2-4A). Virus bearing CA Env were 300-fold less infectious than C virus, whereas virus bearing CT Env was approximately 1,000,000-fold less infectious. However, the CTC and CTCC2 viruses were approximately 200-fold and 1000-fold, respectively, more infectious than CT virus. Western blot analysis showed that all Envs were efficiently processed and incorporated into virus particles (Figure 2-4B). I also observed that the molecular weight of CA and CT Envs were marginally higher than C, CTC and CTCC2 Envs. To further test the critical nature of FeLV-C C2 loop in virus infection, I generated a mutant FeLV-C Env in which the entire C2 loop was deleted (see Figure 2-1A for C2 loop). However, I found that this Env (CΔC2) failed to be incorporated into virus particles (Figure 2-4B), which is consistent with the lack of infection observed on MDTF/hCR1 cells (Figure 2-4A). This construct was being produced intracellulary as expression can be seen in the cell lysate of these cells (Figure 2- 4C). Taken together, my results suggest that the presence of FeLV-C Cdom or C2 loop is critical for FeLV-C infection.
55
2 C 6 Δ C. T C C
72 kDa
Actin
Figure 2-4. Susceptibility of MDTF cells expressing human FLVCR1 to pseudotype virus bearing hybrid FeLV Envs. A. Infection titers of virus bearing FeLV-C (C) or hybrid FeLV Envs on FLVCR1 expressing cells. Titers are average of three infection experiments and are represented as colony forming units (CFU) per milliliter of virus medium. B. Western blot of virus pellets bearing hybrid Envs. A mouse anti- FeLV gp70 antibody was used to detect hybrid FeLV Env proteins and a mouse anti-p30 antibody was used to detect MLVcapsid protein in virus particles. T6 represents virus particles produced from TELCeB6 retrovirus packaging cells that do not express retrovirus Env. C. Cell lysate expression analysis of CDC2.
56
7 6
5
4 CFU/ml)
10 3 (log 2 Infection of titer lacZ virus 1
0 150μM Peptide concentration
FeLV-C Lac Z
Figure 2-5. Effect of C2 loop peptide on FeLV-C infection. Infection titers of virus bearing FeLV-C Env in the absence or presence of 150μM C2loop peptide. Titers are average of three infection experiments and are represented as colony forming units (CFU) per milliliter of virus medium.
57
7 6
5
4 CFU/ml)
10 3 (log 2 Infection titer of lacZ virus 1 RBD RBD RBD CA CT CA + + CA CT + Δ Δ Δ
Lac Z pseudotype virus
Figure 2-6. Susceptibility of MDTF cells expressing human FLVCR1 to pseudotype virus bearing ΔRBD FeLV-C Envs. Infection titers of virus bearing hybrid FeLV Envs or hybrid envs + ΔRBD on FLVCR1 expressing cells. Titers are average of three infection experiments and are represented as colony forming units (CFU) per milliliter of virus medium.
58
2.4.4 C2 peptide is not sufficient for inhibition of FeLV-C infection
Inhibition of HIV infection by a peptide corresponding to the V3 loop of the lentivirus envelope gp120 has been well documented. The peptide can compete for env binding sites on the co-receptor, and block subsequent infection by the virus (Verrier et al., 1999). I constructed peptides corresponding to the C2loop of FeLV-C Cdom and attempted to inhibit FeLV-C virus infection of MDTF/hCR1 cells. No significant change in infection was noted at peptide concentrations of up to 150 μM (Figure 2-5).
2.4.5 Co-expression of Cdom with hybrid env enhances infection
Previous experiments with soluble MLV RBD had shown their ability to rescue fusion defective or RBD deleted (ΔRBD) virus, rendering them capable of successful infection (Barnett et al., 2001). I attempted to rescue CA and CT hybrid env virus infection through co-expression of ΔRBD FeLV-C env. No change in infection of MDTF/hCR1 cells was seen with virus expressing both CT and ΔRBD env on its surface (Figure 2-6). Conversely, pseudotype virus expressing both CA and ΔRBD env exhibited a ten-fold increase in infection in comparison to virus expressing CA env alone. This suggests that co- expression of the FeLV-C Cdom on the virus surface can enhance binding and/or fusion of hybrid envelopes to FLVCR1.
2.4.6 Soluble Cdom binds to FLVCR1 in the absence of FeLV-C RBD
Based on the results presented above, I hypothesized that Cdom forms a distinct domain that can bind to FLVCR1. To test this, I generated a modified FeLV-C SU (Figure 2-1B, see Cdom SU), in which the RBD and PRR were deleted. The Cdom SU contains the first 16 residues of FeLV-C SU fused to FeLV-C Cdom. The design of this construct is modified from a previously reported study describing RBD deleted MLV Envs (Barnett et al., 2001). The Cdom SU was also tagged with a double HA epitope for detection in receptor binding and for Western blot analysis. I expressed the Cdom SU in HEK293T cells and used
59
Figure 2-7. Binding of soluble Cdom to FLVCR1 expressing cells. A. Western analysis showing the cell lysate fraction and culture medium containing C-SU (FeLV-C SU), C- RBD (FeLV-C RBD195), AC-SU (hybrid SU containing RBD155 from FeLV-A and PRR and C domain from FeLV-C) or Cdom SU (FeLV-C C domain only). C-SU and AC-SU were isolated by immunoprecipitation using anti FeLV gp70 antibody on a Sepharose G column. Soluble Cdom SU was isolated from cell culture medium by immunoprecipitation using anti-HA antibody. Soluble C-RBD was isolated from culture medium using anti-V5 antibody. Proteins were detected on Western using anti-HA HRP monoclonal antibody or anti- V5 monoclonal to detect C-RBD. B. Calculation of the percentage increase in SU binding. The histograms represents target cells (MDTF/CR1) incubated with (white) or without (grey)HA tagged SU. Binding of SU to cells causes an increase in fluorescence. Using the NIH Image J Software, the difference in the shift between y and z was noted and the percentage increase in fluorescence/SU binding calculated. C. Bar graph showing the % change in SU binding on cells incubated with SU compared to cells without SU. Target cells (MDTF/hCR1 or MDTF) were incubated with or without soluble SU. All SU’s contain a double HA tag. The C-RBD is tagged with a V5 epitope. Bound SU’s or RBD were detected by flow cytometry using a monoclonal anti-HA or anti-V5 antibody followed by a fluorescein conjugated anti-mouse antibody.Arrow represents no detectable change in SU binding. ND denotes not determined.
60
1 ml of Cdom SU containing medium for binding to MDTF/hCR1 cells. As shown in Figure 2-7A, Cdom SU expression (see cell lysate) was comparable to expression of FeLV-C SU (C-SU), and appears as a broad protein band of approximately 40K molecular weight. I also determined the level of Cdom SU protein in culture medium as analyzed by Western analysis of immunoprecipitated SU protein from 1 ml of culture medium. I found that the level of soluble Cdom SU was significantly lower than soluble C-SU protein level (Figure 2- 7A).
Using flow cytometry, I analyzed binding of HA-tagged soluble Cdom to MDTF/hCR1 cells and to MDTF cells and compared binding of C-SU to these cells. The inset in Figure 2-7B shows a histogram representing MDTF/hCR1 cells incubated with C- SU (white histogram) and MDTF/hCR1 cells incubated without C-SU (grey histogram). As described above, an increase in fluorescence is indicative of SU binding. Using the NIH Image J software, I calculated the percent change in SU binding between cells incubated with SU (Figure 2-7B inset, distance z) compared to cells incubated without SU (Figure 2- 7B inset, distance y). As shown in Figure 2-7B, C-SU efficiently bound to MDTF/hCR1 cells. I also observed significant but a reduced level of C-SU binding to MDTF cells. Binding of C-SU to MDTF cells may be explained by the presence of the endogenous MDTF FLVCR1. We have previously shown that MDTF FLVCR1 functions as a FeLV-C receptor when over expressed (Tailor et al., 2000). Interestingly, I observed significant binding of Cdom SU to MDTF/hCR1, whereas no binding was detected on MDTF cells. I next tested whether Cdom SU binding could be enhanced in the presence of soluble FeLV-C RBD (C-RBD). C-RBD is secreted into the culture medium (Figure 2-7A), and can bind to MDTF/hCR1 (Figure 2-7C). However, I observed a reduction in Cdom SU binding to MDTF/hCR1 cells in the presence of C-RBD, when compared to Cdom binding to MDTF/hCR1. I also tested Cdom binding in the context of a full-length SU. I generated the AC-SU (Figure 2-1B), which contains FeLV-A RBD155 fused to the PRR and Cdom from FeLV-C SU, and tested its ability to bind to MDTF/hCR1. I found that AC-SU also weakly bound to MDTF/hCR1 cells whereas no binding was detected on MDTF cells.
61
2.4.7 Cdom does not interfere with FeLV-C infection
The ability of Cdom to interfere with FeLV-C infection through endogenous binding to FLVCR1 was tested. I expected to see a diminuitive decrease in FeLV-C virus infection of CCC cells as expression of Cdom should lead to an interaction with FLVCR1 endogenously, similar to whole env expression, to downregulate its expression at the cell surface and prohibit virus binding at the cell surface. Surprisingly, I observed a consistent ten-fold increase in FeLV-C infection of CCC cells expressing soluble Cdom (Figure 2-8). This effect was specific for Cdom expression, as control cells expressing FeLV-A env exhibited no change in FeLV-C infection. I also tested the ability of Cdom to bind and fuse MDTF/hCR1 cells for a successful infection. Using a ΔRBD FeLV-C env psuedotyped virus I was able to repeatedly obtain infection titres above background (Figure 2-5). This suggests that in the absence of RBD, in spite of the weak binding exhibited by Cdom, there is sufficient interaction to induce fusion and create a productive infection. Taken together, our results clearly suggest that Cdom forms a distinct receptor-binding domain that can independently bind to FLVCR1 in the absence of C-RBD.
5
4
3 CFU/ml) 10 2 (log
Titer of lacZ(FeLV-C) of Titer 1 C Cdom No Env Figure 2-8. Interference property of soluble Cdom. Cdom or FeLV-C env were expressed in feline kidney CCC cells, which were subsequently tested for susceptibility to β-galactosidase encoding FeLV-C pseudotype virus. Titers are mean of three infection experiments and are represented as colony forming units (CFU) per milliliter of virus medium.
62
2.5 Discussion
This is the first report to show that the Cdom region from FeLV-C SU forms a novel receptor-binding domain that is distinct from the amino terminal RBD, and which binds to its host receptor FLVCR1 in the absence of its RBD. My results raise the possibility that SUs from other related gammaretroviruses also consist of two distinct receptor-binding domains, RBD and Cdom, which interact with their host receptor to initiate virus infection. In this study, I show that FeLV-C Cdom controls the efficiency of FeLV-C SU binding to FLVCR1 and is critical for virus infection of receptor-expressing cells. Furthermore, my results suggest that residues in the C2 loop, located in Cdom, control receptor binding and virus infection. Replacement of FeLV-C PRR and Cdom with the respective domains from the related FeLV-A or FeLV-T Envs (see Figure 1B, CA and CT Envs) significantly disrupts FeLV-C interference (Figure 2A), FeLV-C SU binding to FLVCR1 (Figure 3A), and FeLV- C infection of receptor-expressing cells (Figure 4A). However, re-introduction of only the FeLV-C Cdom or the C2 loop in CA or CT Envs is sufficient to enhance interference, restore SU binding to FLVCR1, and significantly enhance virus infection. Except for the CΔC2 Env, all hybrid FeLV Envs tested were efficiently expressed, processed and incorporated into virus particles (Figures 2 and 4). However, I did observe some subtle but notable differences between some FeLV Envs. First, interference caused by CAC Env was approximately 5-fold greater than interference caused by CTC Env, yet both Envs have identical sequence. This may be explained by differences in Env expression as I found surface expression of CTC Env was approximately 30% lower than CAC Env surface expression (Figure 2B). Second, I found that virus bearing CTC Env, which contains the entire FeLV-C C domain, was 5-fold less infectious than virus bearing CTCC2 Env, which contains only the FeLV-C C2 loop. Analysis of Env incorporation showed that CTC virus had less Env incorporated than CTCC2 virus (Figure 4B). However, in addition to difference in Env incorporation, subtle variations in Env sequence outside the C2 loop must affect infection efficiency as observed by difference between CA and CTCC2 viruses. Sequences downstream of the C2 loop in the C domain of FeLV-A, -C and -T Envs (Figure 1A), although not required for infection, have been shown to modulate infection efficiency
63
(Cheng et al., 2006). Finally, I found that the molecular weight of CA and CT Envs was greater than Envs containing the FeLV-C C2 loop (e.g. C, CTC, CTCC2 Env). The most probable explanation for the higher molecular weight is the presence of a potential N-linked glycosylation site in FeLV-A and FeLV-T C2 loop, which is absent in the FeLV-C C2 loop (see Figure 1A).
The observations that CA Env can still interfere with FeLV-C infection and that the subsequent CA virus can infect MDTF/hCR1, albeit more weakly than FeLV-C virus, would suggest that FeLV-A Cdom can to some extent also recognize FLVCR1. This is not surprising because the Cdom sequence of FeLV-A and FeLV-C SUs are highly conserved in sequence and in length, with the majority of variation confined to the C2 loop. Thus, it is possible that in addition to C2 loop residues, other Cdom residues that are conserved between A and C are also involved in FLVCR1 receptor binding and infection. Taken together, my results clearly show a critical role for FeLV-C Cdom in binding to FLVCR1 and in virus infection, and that the C2 loop residues control receptor binding and infection. My finding that C2 loop is critical for virus infection is consistent with a previous report showing that insertion of FeLV-T C2 loop residues in FeLV-A Env significantly disrupts infection of cells expressing feline FeLV-A receptor ThTR1 (Cheng et al., 2006). Co- expression of Cdom with CA env virus enhanced infection by tenfold (Figure 2-5). This suggests that virus particles expressing a mixture of CA and ΔRBD env in the env trimers exposed on the surface of the virion were collectively able to bind and fuse more efficiently for effective infection. However, I observed no increases in infection for co-expression of CT and ΔRBD env virus. This may be due to the six amino acid insertion within the C2 loop of FeLV-T sequence that may cause steric hindrance in the trimer formation to the Cdom of FeLV-C, preventing it from binding FLVCR1 and aiding in infection. Thus, the size of the C2 loop may play an important role in binding to the cognate receptor for FeLV- A and FeLV-C.
Another feature of the C2 loop that may be critical for binding to a receptor is the three dimensional structure. While there have been many reports of peptides corresponding to the V3 loop of gp120 HIV env inhibiting infection, I could not replicate these results with a
64 peptide corresponding to the C2 loop of FeLV-C (Figure 2-6). It is possible that the concentration required for inhibition by C2 loop peptides were not tried in these experiments and further experimentation with a wider variation in peptide concentration are required. Another possible explanation lies in the structure of the C2 loop. In the context of the Cdom, the C2 loop is surrounded by two cysteine residues that in the oxidizing environment of the outer cell surface are proposed to be disulfide bonded. This bond may provide a structural requirement in providing the correct configuration of the C2 loop for binding to FLVCR1. The C2 loop peptide may not have a similar structure, or may be too flexible to bind efficiently to FLVCR1 for inhibition of virus infection. Conversely, it is also possible that the C2 loop does not inhibit infection, but rather facilitates it. There exists a possibility the presence of C2 loop binding may enhance FeLV-C infection through its interactions with FLVCR1 and is discussed below.
Enhancement of FeLV-C virus infection by endogenous expression of soluble Cdom was a surprising finding of this study (Figure 2-8). These results suggest that soluble Cdom caused increased efficiency in binding and/or fusion of FeLV-C virus at the cell surface. As with other interference studies, endogenous expression of the env causes a downregulation of the receptor at the cell surface and masks available binding sites. It appears that Cdom may have opened an available binding site for FeLV-C virus at the cell surface, perhaps by causing minute conformational changes in FLVCR1, exposing extracellular loop 1 or 6 that are necessary for infection (Brown et al., 2006). These unexpected results warrant further investigation into the mechanism behind Cdom binding to FLVCR1.
It is interesting to note that despite a significant enhancement in FeLV-C interference and virus infection when FeLV-C Cdom or C2 loop is introduced in CT Env, the levels of interference and virus infection are still significantly lower than respective FeLV-C Env interference and virus infection (Figures 2 and 4). These lower interference and virus infection levels may be explained by differences in the PRR sequences of CTC and FeLV-C Envs or differences in the last 40 residues of its RBD. The last 40 residues of CTC RBD are derived from FeLV-T and differ by five residues. Two of these changes are non-conserved, which could conceivably affect interference and infection. Alternatively, or in addition,
65 interference and infection could be affected by the nine amino acid difference in PRR. The PRR from MLVs have been suggested to modulate conformational changes in the envelope that is required for virus fusion and infection (Lavillette et al., 2002). Similarly, the presence of FeLV-C PRR may be required for the correct orientation of Cdom for its optimal interaction with FLVCR1 to promote efficient virus infection.
I show in this study that a modified FeLV-C SU containing only Cdom is soluble and secreted into the culture medium (Figure 2-7A). Cdom SU runs as a broad protein band with an approximate molecular weight of 40 kDa. The calculated molecular weight of Cdom is approximately 22 kDa, which suggests that Cdom may be heavily glycosylated. The presence of six potential N-linked glycosylation sites in FeLV-C Cdom (see Figure 1A) and the broad protein band running at 40 kDa would support Cdom to be N-linked glycosylated. I further show in this study that Cdom binds to FLVCR1 independently of FeLV-C RBD through binding and infection assays. Interestingly, Cdom binding to FLVCR1 is reduced in the presence of soluble C-RBD. It is not clear from our results how C-RBD affects Cdom binding but it is conceivable that there may be some interaction between C-RBD and Cdom that affects binding when the two proteins are expressed independently. My conclusion of a possible interaction between RBD and Cdom is consistent with previous studies suggesting an interaction between MLV RBD and its respective Cdom (Barnett et al., 2001 and Lavillette et al., 2001). It is also interesting that I observed binding, albeit reduced, of Cdom to FLVCR1 when expressed as a full SU containing FeLV-A RBD (Figure 2-7B, AC SU). This finding suggests the possibility that in the context of the full SU, receptor-binding sites within Cdom may be only partially exposed. Additional experiments are required to fully elucidate receptor binding by Cdom in the context of a full SU, and of the potential interaction between RBD and Cdom.
Taken together, I propose a model of envelope–receptor interaction for FeLV-C in which SU binding to FLVCR1 involves discrete interactions of two distinct receptor-binding domains with FLVCR1. One interaction involves RBD, which, in agreement with previous studies (Battini et al., 1995, Battini et al., 1992, Brojatsch et al., 1992 and Rigby et al., 1992), primarily governs receptor specificity, and a second interaction involves Cdom that is
66 critical for promoting virus infection. It is unclear from these studies whether RBD and Cdom simultaneously bind to FLVCR1 or they bind sequentially in a step-wise manner. My observation that FeLV-C Cdom binds to FLVCR1 independent of C-RBD would suggest that RBD and Cdom simultaneously binds to FLVCR1 (see Figure 2-9A). However, it is conceivable that RBD and Cdom interact with FLVCR1 in a two-step manner that involves an initial interaction of RBD with the receptor, which upon conformational changes in the Env, allows a second interaction of Cdom with the receptor (Figure 2-9B). A two-step interaction would be intriguing, as it would show a surprising analogy to the two-step envelope–receptor interaction in HIV infection, in which the gp120 Env primarily binds to CD4 (Dalgleish et al., 1984 and Klatzmann et al., 1984) followed by a second interaction of gp120, via the V3 loop, with chemokine receptors that act as co-receptors (reviewed in Hartley et al., 2005).
My envelope–receptor interaction model may be conserved for other γ-retroviruses. A report by Donahue et al. (1991) suggests that the replacement of FeLV-A C2 loop with FeLV-T C2 loop reduces the ability of the subsequent virus to establish superinfection interference, which leads to enhanced killing of T-cells by the virus. Another report by Lavillette et al. (2002) shows that soluble amphotropic MLV SU, which contains both RBD and Cdom, is up to 100-fold more efficient than the respective soluble RBD at blocking subsequent infection of amphotropic MLV. These previous studies, in addition to studies reporting the importance of C terminal domain of FeLV-B and PERV SUs in controlling recognition of certain receptor homologues (Boomer et al., 1997, Faix et al., 2002, Gemeniano et al., 2006 and Sugai et al., 2001), suggest that our model of envelope–receptor interaction, involving two receptor-binding domains (RBD and Cdom), is conserved for other γ-retroviruses.
Previous attempts to generate specific cell targeting retroviruses relied on introduction of specific ligands in the retrovirus RBD to target use of alternative cell surface proteins as receptors (reviewed by Sandrin et al., 2003). However, these attempts have led to poor infection efficiencies. Furthermore, some cell surface proteins are unable to mediate virus
67
A. (1) Simultaneous binding of RBD and Cdom to receptor (2) Virus infection virus initiated
out
in receptor B. (1)Primary binding of RBD to receptor (2) Secondary binding of virus Cdom to receptor (3) Virus infection virus initiated
TM Cdom PRR RBD
Figure 2-9. Proposed models envelope-receptor interaction for FeLV-C infection. A. Both the RBD (white circle) and Cdom (white square) simultaneously bind to receptor to activate virus infection. B. In the two-step model RBD initially binds to receptor which is subsequently followed by a second interaction of Cdom with the receptor.
68 infection despite efficient binding of modified virus to the surface protein (Cosset et al., 1995). I propose that γ-retroviruses have evolved to using specific host receptor molecules that contain the necessary determinants to allow efficient virus binding and allow additional interactions that promote subsequent virus fusion and infection. Based on my findings in this study, I propose that generation of efficient specific cell targeting γ-retroviruses will require a modified envelope protein that has a fully functional RBD and Cdom, in addition to the specific ligand. Furthermore, target cells will require expression of the targeted receptor for specific attachment, and the expression of the natural retroviral receptor to trigger virus fusion and infection.
3. ROLE OF FLVCR1 IN DIAMOND BLACKFAN ANEMIA
A portion of the work presented in this chapter has been previously published:
Michelle A. Rey, Simon P. Duffy, Jennifer K. Brown, James A. Kennedy, John E. Dick, Yigal Dror and Chetankumar S. Tailor. Enhanced alternative splicing of the FLVCR1 gene in Diamond Blackfan Anemia disrupts FLVCR1 expression and function that is critical for erythropoiesis. Haematologica (In Press, 13359)
Author contributions / Acknowledgements:
All experiments were performed by MAR. Quantitative PCR was initiated, carried out and analyzed by MAR, qPCR primers and experimental procedures were designed by SPD. Varun Charkravorty provided assistance to MAR in cloning of splice variants constructs used in Figure 3.5 and 3.6. We are also grateful to Rati Prasad, Naveen Hussain, and Zvi Shalev for helpful suggestions.
69 70
3.1 Abstract
Diamond Blackfan Anemia (DBA) is a fatal congenital anemia characterized by a specific disruption in erythroid colony-forming units (CFUs-E). Approximately, 25% of DBA individuals have mutations in the ribosomal protein RPS19 suggesting that DBA may be caused by a defect in ribosome biogenesis and translation. However, it is unclear how these defects specifically disrupt CFU-E development. Recent studies have shown that the retroviral receptor/heme exporter FLVCR1 is critical for CFU-E development and that FLVCR1 null mice show craniofacial and limb deformities similar to those found in some DBA patients. However, unlike DBA, FLVCR1 null mice are embryonic lethal and have high reduced levels of myeloid and lymphoid cells. In this study, I show that CD71 cells, which are enriched for immature erythroid cells, from DBA patients negative for RPS19 gene mutations express alternatively spliced isoforms of FLVCR1 transcript. Normal patient samples exhibited very little of the alternatively spliced FLVCR1 as determined by quantitative PCR. I also amplified another gene involved in erythropoeisis, the erythropoietin receptor (EpoR), and a non-related phosphate symporter, Pit1. No splice variants of these genes were found in normal or DBA patients, suggesting the specificity for alternative splicing of FLVCR1. In addition, normalized FLVCR1 transcript levels were significantly lower in the DBA patients in comparison to normal patients. My results suggest that DBA patients may lack sufficient full length FLVCR1 for proper function due to decreased transcription and alternative splicing of the FLVCR1 gene. This insufficient FLVCR1 may lead to the block in erythropoiesis seen in DBA patients. Although these DBA patients did not have mutations in their RPS19 genes, we tested RPS19 transcript levels and found decreased levels of RPS19 transcript in some patients. Taken together, I propose enhanced alternative splicing of FLVCR1 transcript and subsequent FLVCR1 insufficiency as a major contributing factor for the erythropoietic defect observed in DBA.
3.2 Introduction
Diamond Blackfan anemia (DBA) is a rare inherited red blood cell disorder characterized by a reduced production of erythroid colony forming units (CFUs-E) (Ball, 71
SE, et al., 1996, Diamond LK. 1978, Dianzani, I., et al., 1996). Current therapies such as steroids, blood transfusion, or bone marrow transplantation often have severe side effects or are ineffective for many DBA patients (Ball, SE., et al., 1996, Dianzani, I. et al., 2000, Vlachos, A et al., 2001, Willig, TN., et al., 1999). Approximately 40% of DBA patients have additional abnormalities that include short stature, and cranofacial and urogential malformations (Willig, T. et al., 1999). DBA patients are also predisposed to acute myeloid leukemia, lymphoma and solid tumors (Willig, T et al., 1999, Aquino VM.., et al., 1996, Lipton JM., et al. 2001). Approximately 25% of DBA patients have mutations in the gene encoding the S19 ribosomal protein (RPS19) (Draptchinskaia, N. et al., 1999, Willig, TN., et al., 1999), which is one of 33 ribosomal proteins that make up the 40S ribosomal subunit that is involved in translation (Flygare, J. et al., 2007). Down regulation of RPS19 in CD34+ human hematopoietic stem cells disrupts CFU-E development, but not myeloid development (Ebert BL., et al., 2005, Flygare J., et al., 2005), which mimics the hematological features observed in DBA. Transfer of the RPS19 gene in RPS19 deficient DBA bone marrow cells increases the number of erythroid progenitor cells (Hamaguchi, I. et al., 2002), demonstrating the importance of RPS19 in erythropoiesis. Establishment of an RPS19 mouse model has proven difficult. Mice homozygous in the disruption of RPS19 are embryonic lethal whereas mice heterozygous for RPS19 alleles are normal with normal growth of all hematopoietic stem cells (Matsson H., et al. 2004). Recent studies have identified mutations in two additional genes encoding RPS24 and RPS17, respectively in 2% and 4% of RPS19 negative DBA individuals (Cmejla, R. et al., 2007, Gazda, HT. et al., 2006). These findings suggest that DBA may be caused by a disruption in ribosome biogenesis and impairment in translation. However, it remains unclear how these defects specifically disrupt CFU-E development. Interestingly, the hematological features of DBA show a striking similarity to the feline pure red cell aplasia found in domestic cats infected with the subgroup C feline leukemia virus (FeLV-C) (Jarrett, O. et al., 1973, Neil JC. et al. 1991, Sarma PS and Log, T. 1973). The feline anemia has been suggested to be caused by the FeLV-C envelope (Env) protein disrupting the cellular function of the heme exporter FLVCR1 (Quigley JG et al., 2005), which also is used as receptor for entry by FeLV-C (Quigley, JG et al., 2000, Tailor 72
CS, et al., 1999, Brown, JK. et al. 2006). Disruption of FLVCR1 heme export function in human K562 erythroid cells induces apoptosis and disrupts K562 erythroid differentiation (Quigley JG., et al. 2004). Furthermore, FLVCR1 null mice, despite being embryonic lethal and showing reduced levels of myeloid and lymphoid cells (Keel, SB. et al., 2008), show a severe disruption in CFU-E development and have craniofacial and limb deformities found in some DBA patients (Ball, SE et al., 1996, Willig, TN et al., 1999, Keel SB., et al., 2008). These findings raise the possibility that DBA may be caused by a dysfunction in the human homologue of FLVCR1. Consequently, a previous linkage analysis identified four families with DBA linked to human Chr1q31 (Quigley, JG. et al., 2005), the FLVCR1 gene locus (Quigley, JG., et al., 2000). However, no genetic mutations were found in the exons of the FLVCR1 gene. In this study, I provide evidence of enhanced alternative splicing of FLVCR1 transcript, which disrupts FLVCR1 protein expression and function, in DBA cells negative for RPS19 gene mutations. I propose FLVCR1 insufficiency as a major factor for the anemia in DBA.
3.3 Materials and methods
3.3.1 Bone marrow samples
Bone marrow aspirate samples were collected into preservative-free heparinized syringes. Marrow mononuclear cells were separated using Ficoll-Hypaque and cryopreserved as previously described (Dror, Y. et al., 2001). Patients (D1-D5) were diagnosed with DBA based on published criteria (Willig, et al., 1999), which included chronic anemia, erythroid hypoplasia and either early onset (less than one year) or a first degree relative with DBA. No patient had evidence of severe aplastic anemia or malignant transformation at the time of bone marrow sample collection. Ten hematologically healthy normal (N1-N10) bone marrows donors served as controls. The studies were approved by the Institutional Research Ethics Board, and informed written consent was obtained from subjects or their legal guardians. An additional six bone marrow samples (N11-N16) were purchased from Stem Cell Technologies. 73
high 3.3.2 RNA isolation from CD71 cells.
Bone marrow aspirates from normal and DBA patients were thawed in RPMI media (Sigma) containing 10% fetal calf serum (FCS) (Sigma) and 0.1μg/ml DNAse (Sigma). Approximately one million cells were washed twice in cold phosphate buffered saline (PBS) containing 0.5% bovine serum albumin (BSA, Sigma), and stained with 20μl anti-CD71-PE (BD Biosciences) conjugated antibody at 4oC for one hour in the dark. Cells were then washed twice to remove unbound antibody and sorted into CD71high or CD71low and CD71- cells using a fluorescence activated cell sorter (BecktonDickson). Total RNA was extracted from sorted cells using RNeasy RNA preparation kit (Qiagen) and total cDNA was prepared using Thermoscript reverse transcriptase (Invitrogen).
3.3.3 Isolation of FLVCR1, Pit1 and EpoR sequences from DBA and normal erythroid cells.
FLVCR1 sequences were amplified by PCR using Expand Hi-Fidelity Taq DNA polymerase (Roche) using cDNA isolated from five DBA (D1-D5) and ten normal (N1- N10) cell samples (see above), and specific primers that primed to FLVCR1 cDNA encoding the transmembrane (TM) 1 sequence (FLVCR1-TM1 primer: 5’- TACTCGCTGGTCAACGCCTTTCAGTGG-3’) and TM12 sequence (FLVCR1-TM12 primer: 5’-TCTTCGCAGATCAGACTTGATTAATGCTGTTAA-3’). The PCR was run according to manufacturers protocol. Amplified products were cloned into pCR2.1TOPO vector (Invitrogen) and subsequently sequenced (ACGT Corporation, Toronto, Canada). Sequences encoding the Pit1 phosphate symporter and the erythropoietin receptor (EpoR) were also amplified by PCR using Pit1 primers (Pit1-F primer 5’- CCGCCGCTTCTGGTCCTTTGGTGG-3’; Pit1-R primer 5’- CGCCAGTCAACAGCCTTCTTGGACC-3’) and EpoR (EpoR-F primer 5’- ATGGACCACCTCGGGGCGTCCCTCTGG-3’; EpoR-R primer 5’- CTAAGAGCAAGCCACATAGCTGGG-3’).
74
- - - 3.3.4 Construction of HA tagged E3 and E3 E6 retroviral expression constructs.
- - - cDNA encoding the N-terminal of FLVCR1 was ligated to the E3 or E3 E6 FLVCR1 cDNAs by first digesting pFBneoFLVCR1HA (Brown JK. et al., 2006), containing the full-length FLVCR1 cDNA with 3’ HA tag, with BamHI restriction enzyme, - - - which removes the C terminal two-thirds of the FLVCR1 sequence. The E3 or E3 E6 FLVCR1 sequence, cloned in pCR2.1TOPO vector, was isolated by a BamHI digest and was subsequently cloned in the BamHI digested pFBneoFLVCR1HA to respectively - - - generate the E3 -N and E3 E6 -N vectors. cDNA encoding the HA tagged C-terminal sequence of FLVCR1 was isolated from the pFBneoFLVCR1HA vector by digestion with - - - AseI and was then ligated to a AseI digested E3 -N or E3 E6 -N vectors to respectively - - - generate E3 HA and E3 E6 HA FLVCR1. Phoenix ampho cells were then transfected with the respective expression constructs, and virus supernatant from the transfected phoenix- - - - ampho cells was used to introduce the HA tagged E3 or E3 E6 FLVCR1 cDNAs in murine MDTF cells. Transduced MDTF cells were selected using G418 (1.5mg/ml) and tested for susceptibility to lacZ virus infection.
3.3.5 Protein expression profile.
Cellular expression of HA tagged FLVCR1 proteins was analyzed by Western as previously described (Brown JK., et al., 2006). Cellular localization of the HA tagged proteins in target cells was analyzed by immunofluorescence. Briefly, target cells were fixed with 2.5% paraformaldehyde and permeabilized with 0.05% Triton-X-100 in PBS, after overnight growth. Staining was performed with a 1:500 dilution of monoclonal anti-HA.11 antibody (Covance, QC, Canada) followed by incubation with a 1:500 diluted goat anti- mouse antibody conjugated to Cy3. Cell DNA was stained with 0.1μg/ml DAPI DNA counter stain (Sigma). Fluorescence was visualized with a Zeiss LSM 510 Axiovert confocal microscope with 63X oil objective magnification.
75
3.3.6 Quantitative real-time PCR.
Gene expression was quantitated by Q-PCR using the ABI7900 real-time PCR machine with SYBR Green master mix (Applied Biosystems, CA, USA). Q-PCR was high performed using total cDNA isolated from DBA and normal CD71 cells and the following primers. Gene Expression was calculated relative to actin B gene expression (ActB-F primer: 5’-TGCGTGACATTAAGGAGAAG-3’; ActB-R primer: 5’- AGGAAGGAAGGCTGGAAGAG-3’). Total FLVCR1 transcript expression was quantitated using exon 1 specific primers (E1 primer: 5’- TCCATGGTGTACATGCTGGCCTA-3’; αE1 primer: 5’- AGGAGATGTTGTTGCACAGTGCCG-3’). E2 and E3 containing FLVCR1 transcripts were quantitated using the E2 primer 5’-GAACATCAGCTGTTGCCACA-3’, and the αE3 primer 5’-TGTCTTGAAGAGCTGCTTGA-3’. To eliminate the possibility of genomic contamination, Q-PCR was also performed using FLVCR1 intron 1 specific primers (intron1-F primer: 5’-AGTTCCTGGCACATAAGGGACACA-3’; intron1-R primer: 5’- GCCTCCCAAAGTGCTGAGATTACA-3’). RPS19 gene expression was quantitated using the RPS19-F primer 5’-AAAGACGTGAACCAGCAGGAGTTC-3’, and the RPS19-R primer 5’-AGGAAGGAAGGCTGGAAGAG-3’.
3.3.7 Genomic DNA analysis
A single DBA patient genomic DNA was used for brief sequence analysis of the FLVCR1 locus. Control genomic DNA from a normal patient was not readily available, but genomic DNA from a human cell line, TE671 was used for comparison. The intron/exon boundaries including 250+ bp of intronic sequence of exon 1, 2 & 3 were targeted for PCR amplification and sequencing using the following primers. The numbering of the nucleotides is relative to the initiation codon of the FLVCR1 gene, which is located at 2098010 on chromosome 1. Region 561-1233 corresponding to the exon1/intron1 boundary was amplified with forward primer 5’-AGCCTGTACTCGCTGGTCAA-3’, and reverse primer 5’-ACTCCTGACTTGAAGTGATCCACCCA-3’. Region 5161-5923 corresponding to the intron1/exon2/intron2 boundary was amplified with forward primer 5’- 76
TTCGACACCAGCCTGACCAACATGGTGAAA-3’ and reverse primer 5’CCTGGCTAGGTAGGCATAAGAGGCATAAGA-3’. Region 14040-14966 corresponding to the intron2/exon3/intron3 boundary was amplified with forward primer 5’- CACTGTGCCCAGCCTTCATGTTTATCTC-3’ and reverse primer 5’ACCTCTGCCTTCTAGGTTCAAGCGATTCTC-3’. PCR products were sequenced (ACGT Corp., Toronto, Canada) and analyzed for any potential mutations. Of particular interest were intronic regions that may act as intronic splicing silencers (ISS) or enhancers (ISEs), as previous studies has suggested no mutations in the exonic sequence of FLVCR1 (Quigley, et al., 2005). SNPsplicer (ElSharawy et al., 2006) and PESX (Zhang and Chasin, 2004, Zhang et al., 2005) were used to find putative ISS and ISE sequences in the introns of FLVCR1 to pinpoint any regions that may be mutated and affect efficient splicing.
3.4 Results
Previously in our laboratory, we were able to recapitulate the hematological features - of DBA in vitro by disrupting human FLVCR1 function in lineage depleted cord blood (Lin CB) cells using the FeLV-C Env protein. Expression of the FeLV-C Env protein caused a significant but more importantly a partial block in FLVCR1 function. Our results suggested - that a significant but partial block of FLVCR1 function by FeLV-C Env in Lin CB cells is critical for erythropoiesis but not myelopoiesis. We observed a dramatic reduction in the + - percent of GlyA and CD45 cells in FeLV-C Env transduced culture compared to vector only and FeLV-B Env transduced cultures, which suggests a reduction in mature erythroid cells. We also observed a significant reduction in the relative cell number in the FeLV-C Env transduced cell culture when compared to vector and FeLV-B Env transduced cultures suggesting a block in proliferation, which is also consistent with features observed in DBA. This is the first report showing that FeLV-C Env protein directly disrupts early erythropoiesis and provides more direct evidence that the FeLV-C induced anemia in cats is caused by FeLV-C Env protein disrupting feline FLVCR1. More importantly, our findings - clearly suggest that a significant but partial block in human FLVCR1 function in Lin CB 77 cells can recapitulate the DBA hematological features raising the possibility of an important role for FLVCR1 in DBA.
3.4.1 Isolation of alternatively spliced FLVCR1 isoforms from DBA erythroid cells.
To further investigate a role for FLVCR1 in DBA, I analyzed FLVCR1 sequences from immature erythroid cells isolated from five DBA (D1-D5), and from ten normal (N1- N10) bone marrow samples. All DBA patients were identified as being negative for RPS19 gene mutations. DBA patient samples that were positive for RPS19 gene mutations were not available for this study. Using antibodies specific to CD71, we isolated a population of cells expressing high levels of the CD71 surface protein (CD71high), which are enriched for immature erythroid cells (Loken et al., 1987). Consistent with the low levels of erythroid progenitor cells in DBA patients, we isolated fewer CD71high cells from each of the five DBA samples than from the ten normal samples. Using PCR and primers specific to FLVCR1 transmembrane (TM) 1 and 12 (see Figure3-2, white arrows) coding regions, I amplified FLVCR1 cDNA of approximately 1.2kbp in size from DBA and normal CD71high cells (Figure 3-1A). I subsequently cloned the amplified cDNA in a PCR cloning vector and sequenced clones from each of the DBA samples D1-D5 and from each of the normal samples N1-N10 (Figure 3-1B). Interestingly, in addition to amplifying the expected 1.2 kilobase pair (kbp) FLVCR1 cDNA, I also amplified cDNAs of 1.0-1.1kbp in size in each of the DBA and normal samples (Figure 3-1B). Subsequent sequencing showed that the 1.2kbp sequence was FLVCR1 with no mutations. Interestingly, the 1.0-1.1kbp sequences were also FLVCR1 but contained specific exon (E) sequence deletions. I isolated normal spliced - FLVCR1 and an E6 (exon 6 deleted) FLVCR1 from all five DBA and ten normal samples - - (Figure 3-2). In addition, I isolated four alternatively spliced FLVCR1 isoforms, E3 , E3 - - - - E6 , E2 , and the E2 E6 from DBA samples but not from normal samples. Figure 3-2 also shows a schematic of the potential proteins encoded by the alternatively spliced FLVCR1 isoforms. Splicing of FLVCR1 E3 or E6 causes in frame deletions that encode potential proteins with major deletions (Figure 3-2, see hashed lines). Interestingly, splicing of E2 causes a frame-shift deletion, which results in the generation of 78
A M - N D
1.6kb 1.0kb
B
Normal DBA
Figure 3-1. RT-PCR analysis of FLVCR1 from normal and DBA patient bone marrow samples. 3-1A. Representative agarose gel of RT-PCR from normal (N) and DBA (D) samples using primers against TM1 and TM12 of FLVCR1. 3-1B. Sample agarose gels of pCR2.1 TOPO cloning of extracted PCR bands. These clones were then sent for sequencing to verify their sequence as FLVCR1. 79
cDNA Encoded Sample protein 123456 E1 E2 E3 E4 E5 E6 E7 E8 E9 E10 D1-D5 FLVCR1 N1-N10
N C PTC * D1-D5 E2-
PTC * E2-E6- D1-D5
D1 E3-
D1 E3-E6-
D1-D5 E6- N1-N10
Figure 3-2. Alternatively spliced FLVCR1 isoforms isolated from DBA and normal immature erythroid cells. Alternatively spliced FLVCR1 isoforms were isolated from five DBA (D1-D5) and ten normal (N1-N10) samples. FLVCR1 exons (E) 1-10 are shown as boxes. Primers used to amplify FLVCR1 sequence are depicted by white arrows. FLVCR1 sequences are labeled E2- (exon 2 deleted), E2-E6- (exon 2 and 6 deleted), E3- (exon 3 deleted), E3-E6- (exon 3 and 6 deleted) and E6- (exon 6 deleted). The premature termination codon (PTC) is denoted by an asterisk. Also shown is a schematic of the topology of potential proteins encoded by the alternatively spliced FLVCR1 sequences. Normal spliced FLVCR1 encodes a cell surface protein predicted to contain 12 transmembrane spanning segments with six presumptive extracellular loops (loop numbers are indicated above the loops). White arrows denote the sequence encoded by the primers used to amplify FLVCR1 sequences. Also denoted are the DBA and normal samples from which the FLVCR1 sequences were isolated. 80 a premature termination codon (Figure 3-2, see PTC) that potentially encodes a truncated high protein. Together, these findings suggest that DBA CD71 cells express alternatively spliced FLVCR1 sequences that are not expressed, or weakly expressed (see below), in high normal CD71 cells.
3.4.2 Alternative splicing may be specific for FLVCR1
To ascertain the specificity of the aberrantly spliced FLVCR1 transcript, I amplified cDNA sequences encoding two additional human cell surface proteins. I isolated sequence encoding the inorganic phosphate symporter Pit1 (Kavanaugh MP et al., 1994), which functions as a receptor for FeLV-B (Takeuchi Y, et al., 1992), and I isolated sequence encoding the erythropoietin receptor (EpoR) (Maouche L, et a., 1991), which is a major protein expressed on surface of erythroid cells. I isolated the expected respective cDNAs of 1.6kb and 2.0kb from each of the six DBA and eight normal erythroid samples. I subsequently cloned these cDNAs into a PCR cloning vector, and sequenced 12 independent clones for each receptor cDNA from each of the six DBA and eight normal erythroid samples (Figure 3-3). I found that all clones were full-length Pit1 or full-length EpoR sequence. I did not identify alternatively spliced variants of Pit1 or EpoR sequence. These results suggest that within the limitation of these experiments that aberrant splicing is unique to FLVCR1 in DBA erythroid cells.
3.4.3 Non-erythroid cells also express alternatively spliced FLVCR1
To ascertain whether aberrant splicing of FLVCR1 is restricted to DBA erythroid cells, I isolated FLVCR1 sequence from DBA non-erythroid (i.e.CD71low and CD71-) cells. I isolated the E2- and E6- FLVCR1 sequences from two DBA (D3 and D6) non-erythroid cell samples tested (Figure 3-4). This finding suggests that both DBA erythroid (CD71high) and non-erythroid (CD71low and CD71-) cells express aberrantly spliced FLVCR1.
81
A. B. M - TE N1 N2 N6 N3 N4 N5 M M D5 D1 D2 D4 M N6 N3 N4 N5 D5 D1 D2 D4 –
C. D.
Figure 3-3. RT-PCR analysis of EpoR and Pit1. Representative agarose gel of RT-PCR from normal (N) and DBA (D) samples using primers against the 5’ and 3’ ends of Pit1 (A) or the EpoR (B). Sample agarose gels of pCR2.1 TOPO cloning of extracted PCR products (Pit 1, C, EpoR, D). These clones were then sent for sequencing to verify their sequence. * represent clones not containing Pit1R or EpoR sequences. 82
C = Control NE = Non-erythroid DBA #3 E = Erythroid C NE E
1.6kb Δexon6 Δexon2&6 1.0kb
Figure 3-4. RT-PCR analysis of RNA from erythroid and non-erythroid DBA patient bone marrow sample. Bone Marrow sample from DBA patient #3 was sorted for erythroid and non-erythroid cells using CD71. RNA from both samples were subjected to RT-PCR analysis with primers for TM1 and TM12 of FLVCR1 as described previously. Arrows indicate expected sizes of splice variants of FLVCR1. 83
3.4.4 Alternatively spliced FLVCR1 transcripts encode proteins that are disrupted in their cellular and surface expression, and in their receptor function.
- - I next characterized the expression profile and receptor function of the E3 and E3 - E6 encoded proteins. As described above, splicing of E3 or E6 in FLVCR1 transcript causes an in frame deletion that could potentially encode a functional protein. I expressed - - - HA tagged E3 , E3 E6 or full-length FLVCR1 cDNAs in murine Mus. dunni tail fibroblast (MDTF) cells, and tested the susceptibility of these cells to FeLV-C infection. Consistent with our previous findings (Brown JK., et al., 2006), parental MDTF cells were resistant, and MDTF cells expressing full-length human FLVCR1 were highly susceptible, to FeLV-C - infection (Figure 3-5A, see control and hFLVCR1). Interestingly, I found that the E3 and - - E3 E6 transduced MDTF cells were weakly susceptible to FeLV-C. Infection titers were 100,000-500,000 fold less than titers on MDTF/hFLVCR1 cells (Figure 3-5A). Furthermore, the FeLV-C SU did not display efficient binding to the MD cells over- expressing splice variants as is seen with MDTF/hFLVCR1 cells (Figure 3-5B, and see Chapter 2, Figure 2-7). This decrease in binding coincides with the low infection titers of FeLV-C lacZ virus. - - - To investigate the relative resistance of E3 and E3 E6 FLVCR1 transduced murine cells to FeLV-C, I analyzed the cellular expression of the encoded proteins in cell lysate - - - fractions. I found that the E3 and E3 E6 encoded proteins were weakly expressed compared to full-length FLVCR1 (Figure 3-6A). Using confocal microscopy, I further found - - - that the E3 and E3 E6 encoded proteins were predominantly dispersed within the cell (Figure 3-6B). In contrast, full-length FLVCR1 was predominantly localized at the cell membrane (see white arrows in Figure 3-6B) consistent with FLVCR1 being a cell surface - - - protein. Taken together, these findings suggest that the E3 and E3 E6 encoded proteins are disrupted in their cellular and surface expression, and in their receptor function.
84
A. 6
5
4 cfu/ml) 10 3 (log Infection titer of 2 lacZ encoding FeLV-C encoding lacZ
1
- - l 1 3 tro R E - E6 on C 3 c LV E hF B.
100 100 MD CR1 E3- MD CR1 E3-E6-
80 80
60 60 % of Max % of Max 40 40
20 20
0 0
Figure 3-5. Functional analysis of the E3- and E3-E6- FLVCR1 encoded proteins. A. Susceptibility of murine MDTF cells expressing human FLVCR1, E3- or E3-E6- encoded proteins to β-galactosidase expressing FeLV-C. Control represents parental MDTF cells. Infection titers are averages of three experiments and are represented as colony forming units (cfu) per milliliter of virus supernatant. Arrow indicates zero infection titer. B. Surface binding assay of FeLV-CSU envelope protein to bind to cells expressing FLVCR1 splice variants was tested as described previously (Chapter 2.3.7). Red histogram shows fluorescence in absence of FeLV-C SU, blue histogram shows binding in presence of FeLV-C SU. 85
1 R - l C 6 ro V - E A. t L - 3 on F 3 E kDa c h E
72
36 Actin
Cell lysate
B.
MDTF/hFLVCR1 control
MDTF/E3- MDTF/E3-E6-
Figure 3-6. Expression analysis of the E3- and E3-E6- FLVCR1 encoded proteins. A. Western blot analysis of HA tagged E3-, E3-E6-, and hFLVCR1 proteins in cell lysate fractions from FLVCR1 transduced MDTF cells. B. Confocal immunofluorescence microscope image of MDTF cells (control), and MDTF cells expressing HA tagged human FLVCR1, E3- or E3-E6- FLVCR1. White arrows depict surface localization. Nucleus of cell is stained with Dapi. 86
3.4.5 Alternative splicing of FLVCR1 transcript is enhanced in DBA immature erythroid cells.
To investigate the significance of the alternative spliced FLVCR1 in DBA, I quantitated the expression levels of alternatively spliced FLVCR1 transcripts in DBA samples and compared it to expression in normal samples. I used six new normal samples high (N11-N16) enriched for CD71 expressing cells in addition to samples N5, N7, N9 and N10. Samples N1-N4, N6 and N8 were not used because of insufficient sample material. Using total cDNA prepared from each DBA and normal samples and using the E1/αE1 primers (see Figure 3-7A, inset), I quantitated, by real-time PCR, total FLVCR1 (normal and alternatively spliced) transcript expression relative to actin B gene expression. I found that total FLVCR1 transcript expression was significantly down regulated in DBA samples D1- D5 compared to expression levels in the ten normal samples (Figure 3-7A). The exception was normal sample N15, which had a reduced level of FLVCR1 transcript expression. I next used the E2/αE3 primers (Figure 3-7A, inset) to quantify expression levels of FLVCR1 transcripts containing both E2 and E3, and calculated the percent of E2/E3 containing FLVCR1 transcript relative to total FLVCR1 transcript. As shown in figure 3-7B, I found that approximately 5-45% of FLVCR1 transcript in the DBA samples contained both E2 and E3 compared to 76-96% E2/E3 containing transcripts in the normal samples. These findings clearly suggest a significant disruption in FLVCR1 E2 and E3 in the DBA samples, and implicate a dramatic enhancement in alternatively spliced FLVCR1 transcript in the DBA samples compared to the normal samples.
3.4.6 RPS19 expression levels vary in DBA patients
I also analyzed RPS19 gene expression relative to actin B gene expression in DBA and normal samples. I found that RPS19 gene expression in three of the five DBA samples (D1, D2, and D3) was significantly lower compared to the ten normal samples (Figure 3- 7C), whereas gene expression in samples D4 and D5 was comparable to expression levels observed in normal samples. These results show that RPS19 gene expression is down regulated in three DBA samples despite the samples being RPS19 negative. 87
A. FLVCR1 E1 E2 E3
E1 αE1 E2 αE3 250 200 DBA 150 100 Normal 50
30
20 relative to actin B to actin relative
FLVCR1 gene expressionFLVCR1 10 N7 N9 D1 D2 D3 D4 D5 N5 N10 N15 N11 N12 N13 N14 N16 B. 100 90 80 70 60 50 40 30 20
containing and E3 E2 10 % of FLVCR1 transcript of FLVCR1 % N7 N9 N5 D1 D2 D4 D5 D3 N10 N12 N11 N13 N14 N15 N16 C. 200 150 100 50
30
20 Relative to actin B actin to Relative
RPS19 gene expressionRPS19 10 D1 D2 D3 D4 D5 N5 N7 N9 N12 N15 N10 N11 N13 N14 N16 Figure 3-7. Quantification of normal and alternatively spliced FLVCR1 transcript expression in DBA and normal immature erythroid cells. A. The specific E1/αE1 and E2/αE3 primers used in real-time PCR assay is shown. Also shown is the total FLVCR1 transcript expression, relative to actin B gene expression, in the five DBA (striped) and ten normal (black) cell samples. B. Percent of E2 and E3 containing FLVCR1 transcripts. C. RPS19 gene expression relative to actin B gene expression. 88
3.4.7 DBA patients may exhibit mutations in their FLVCR gene affecting splicing
To investigate a cause for the alternative splicing of FLVCR1 I amplified regions corresponding to the intron/exon boundaries of E1, E2 and E3 to ascertain any potential splice site mutations. Previous work had determined no gross malformations or mutations in the exonic sequence of FLVCR1 (Quigley et al., 2005), but the presence of mutations affecting potential splice sites was not determined. Repeated attempts at amplifying the intron/exon boundaries of exon 3 with multiple PCR conditions were not successful (Figure 3-8A). New primer design is warranted in amplification of this region from DBA patients exhibiting exon3 deletions. Amplified products of the exon1/intron1 and intronic sequence surrounding exon 2 (Figure 3-8A) were verified to be the correct amplicons by pCR2.1 TOPO cloning and sequencing. All insertion/deletion and substitution mutations were verified with bidirectional sequencing of multiple clones from two individual PCR amplicons. A T5714C single nucleotide mutation was found in intron 2 just downstream of exon 2 in the DBA patient genomic DNA, but not in the TE671 or Genbank genomic sequences (Figure 3-8B), suggesting this may be a potential SNP. Analysis of this region for ISS or ISE sequences determined this thymine to be part of an ISS sequence that may regulate splicing of this exon. In addition, a two adenine insertion in a string of adenines at positon 5329 just upstream of exon 2 found only in the DBA patient sample (Figure 3-8B) may also be involved in altering exon 2 silencing as adenine stretches have been associated with silencing splicing of adjacent exons (Zhang et al., 2005).
89
A. intron1/exon2/ exon1/intron1 intron2 intron2/exon3/intron3 M - TE D6 – TE D6 – TE D6
B.
5131 ttcgacacca gcctgaccaa catggtgaaa 5161 ccccgtctgt actaaaatac aaaaattagc tgggcgttgt ggcgcgcgcc tgtaatccca 5221 gctactcagg aggctgaggc aggagaattg cttgaacctg ggagggagag gttgcagtga 5281 gccaagattg cgtcactgca ctccagcctg ggagacagac tttgtctcaa aaaaaaaaaa 5341 aagaacataa tagtttaatt ttcaaagcac ctttataaac actagctgtc ctctttatgt 5401 tctgttaatt gccagcattt acttttttct ctaatgataa tagctgttaa caggactatg 5461 tgtttttcag cttggaactg cagttggctt tttgctacca ccagttttag tacccaacac 5521 acagaatgac acaaatctcc tggcttgtaa tatcagcacc atgttttatg gaacatcagc 5581 tgttgccaca cttttattta ttttaacagc aattggtaag tgaattactt tccctaaagc 5641 ttaaatgaat gcatgataga aatttgagaa taactaactc tggaaatttt tgttgataat 5701 atctcaagtg tca T gtgctt tgtttcattc ttccatgccc tgtcttgctc aggtaatgaa 5761 tgctacaacc ctagggatgt tatggcccac tacatttatt ttcacattgt gaagggtttg 5821 aagactaatt catttatccc tctactaact tttcttcttg gtaactgttg gtttcaaata 5881 tattacctgc ctcctcttcc catattcgcc tacatggtct gcatcttatg cctcttatgc 5941 ctacctagcc agg 5953
Figure 3-8. Genomic sequencing analysis of FLVCR1 in a DBA patient. A. Representative agarose gel of PCR amplification of intron.exon boundaries for exons 1, 2 and 3 using genomic DNA from DBA patient #6 and cultured TE671 cells. Arrows indicate correct amplified sequences. B. Sequence analysis of mutations found surrounding exon 2. Shown is the intron1/exon2/intron1 sequence amplified by PCR. Underlined sequence represents primer sequence used for amplification. Bolded sequence represents exon2 coding region. (*) indicates potential two adenine insertion site and capitalized/italicized T5714 indicates potential point mutation found in DBA patient FLVCR1 gene locus. 90
3.5 Discussion
Mounting evidence indicates that alternative splicing of transcripts leading to exon skipping plays a major role in many human diseases (Faustino, NA., et al., 2003, Venables, JP. 2004, Pettigrew, CA., et al., 2008). In this study, I provide evidence that suggests enhanced alternative splicing of the FLVCR1 transcript as a major contributing factor for the erythropoietic defect observed in DBA. My results suggest enhanced alternative splicing of FLVCR1 transcript, which disrupts FLVCR1 protein expression and function, in DBA immature erythroid cells negative for RPS19 gene mutations. Previous studies reporting the critical role of FLVCR1 in erythropoiesis only partially mimicked the hematological features of DBA. Disruption of FLVCR1 heme export function in human K562 cells was shown to disrupt erythroid differentiation of K562 cells but the effect of FLVCR1 disruption in development of other hematopoietic lineages was not determined (Quigley, JG, et al., 2004). Furthermore, FLVCR1 null mice are disrupted in CFU-E development and show craniofacial and limb deformities similar to those found in some DBA patients (Ball, SE., et al., 1996, Keel SB., et a., 2008). However, unlike DBA, FLVCR1 null mice are embryonic lethal and have a significant reduction in lymphoid and myeloid cell numbers. Previously, as a part of this study, we were able to recapitulate the - hematological features of DBA in vitro by disrupting human FLVCR1 function in Lin CB cells using the FeLV-C Env protein. These results suggested that a significant but partial - block of FLVCR1 function by FeLV-C Env in Lin CB cells is critical for erythropoiesis but not myelopoiesis. The ability to recapitulate the DBA hematological features with these experiments raises the possibility of an important role for FLVCR1 in DBA. To further investigate a potential role for FLVCR1 in DBA, I analyzed FLVCR1 sequence expressed in CD71high cells (enriched for immature erythroid cells) isolated from DBA bone marrow that were negative for RPS19 gene mutations, and from normal bone marrow. I show that both DBA and normal samples express alternatively spliced FLVCR1 transcript but more importantly our results suggests that alternative splicing of FLVCR1, specifically E2 and/or E3 skipping, is significantly enhanced in the DBA cell samples 91 compared to normal cell samples (Figures 3-2 and 3-7). The significance of alternatively spliced FLVCR1 in the DBA samples is further compounded by the dramatic reduction in FLVCR1 gene expression (Figure 3-7A). Moreover, the alternatively spliced transcripts that we identified in the DBA samples, specifically the E2-, E3- and E3-E6-, encode proteins that are defective. The E3- and E3-E6- encoded proteins are disrupted in their cellular expression, surface localization and in their receptor function. The heme export function of the E3- and E3-E6- encoded proteins was not determined in this study. However, a previous study has shown that the corresponding E3 deletion in murine FLVCR1 completely abrogates heme export function (Keel, SB., et a., 2008) suggesting that the E3- and E3-E6- encoded proteins in this study are also defective in their heme export function. The E2- transcript, although not characterized in this study, is also likely to be non-functional because splicing of E2 creates a premature termination codon (PTCs) (Figure 3-2) resulting in the potential expression of a truncated protein that lacks the E3 encoded sequence required for heme export (Kell, SB., et al., 2008). Moreover, the E2- transcript could be degraded by the nonsense-mediated decay pathway, which is a quality control mechanism that selectively degrades mRNAs containing PTCs (Chang, YF., et al., 2007). These findings suggest that expression of normal spliced FLVCR1 transcript is substantially reduced in DBA samples as a result of enhanced alternative splicing. I have calculated that on average there is an approximate 58-fold less normal spliced FLVCR1 transcript in the DBA samples compared to normal samples. Taken together, my findings suggest a dramatic disruption in FLVCR1 protein expression and function as a consequence of enhanced alternative FLVCR1 splicing in the DBA samples. Because of the limited availability of sample material, I was not able to assess FLVCR1 protein expression in the DBA or normal cells. I show in this study that RPS19 gene expression is down regulated in three of the five DBA samples despite the samples being RPS19 negative (Figure 3-7C). My findings are consistent with a previous report showing reduced RPS19 gene expression in two RPS19 negative DBA patients (Koga, Y., et al., 2006). This previous report also suggested that an underlying mechanism in DBA could be reduced RPS19 expression. However, my results do not support this previous conclusion because two of the DBA samples, D4 and D5, showed RPS19 gene expression comparable to levels found in normal samples. Our results 92 clearly suggest enhanced alternative splicing of FLVCR1 transcript and disruption of FLVCR1 protein expression and function as an underlying mechanism. A multitude of alternatively spliced genes have been identified as being involved in disease manifestation (Srebrow A. and Kornblihtt A., 2006, Faustino, et al., 2003, Venables, JP. 2004). The mechanism behind genes that are alternatively spliced differs as mutations can occur in both the intronic and exonic sequence that affect splicing of particular exons. The discovery of exonic and intronic splicing enhancer and silencer sequences has broadened the field in predicting the splicing of a particular gene. The mechanism behind alternative splicing of FLVCR1 in DBA patients remains to be seen. Preliminary data with just one DBA patient genomic DNA sample suggests there may be SNPs or indel mutations not previously characterized within the FLVCR1 gene that may affect splicing. Mutations in ISS sequences that alter binding of heterogenous nuclear ribonucleoproteins (hnRNP) splicing accessory proteins may increase or decrease skipping of the adjacent exon (Zhang, X., et al. 2005). It is possible that these mutations enhance hnRNP binding to increase exon 2 skipping in DBA patients, and perhaps give rise to faulty joining of exon 1 and exon 3 leading to the frameshift mutation. These results are by no means exhaustive, and other more stringent methods of genomic DNA sequencing are necessary to confirm these mutations. Furthermore, these mutations must be confirmed to be involved in splicing through in vitro splicing assays with minigene constructs. In conclusion, based on my findings in this study, I propose that FLVCR1 insufficiency caused by enhanced alternative splicing of the respective transcript, specifically to skipping of FLVCR1 E2 and/or E3, plays a major contributing role to the erythropoietic defect observed in DBA. How FLVCR1 specifically regulates erythropoiesis remains unclear. Abkowitz and colleagues have proposed that the heme export function of FLVCR1 regulates intracellular heme levels, which is critical for CFU-E development where FLVCR1 expression is high (Quigley, JG. et al., 2004, Keel et al., 2008). The authors have proposed that a disruption in FLVCR1 heme export leads to heme toxicity and apoptosis of CFUs-E. While further investigations are needed to establish the “heme toxicity” mechanism, it is evident from our study that in DBA there is enhanced alternative splicing of FLVCR1 transcript, which disrupts FLVCR1 expression and function that is critical for erythroid progenitor cell development. 4. DISCUSSION AND FUTURE DIRECTIONS
4.1 A new model of gammaretroviral infection
The role of the Cdom of gammaretroviral envelope proteins has been shown to be involved in efficient binding of the virus to its host receptor. Previous works had suggested that Cdom functioned by interacting with RBD through conformational changes that enhanced or intiated virus fusion. I have provided evidence that Cdom may serve as an additional receptor binding domain to enhance interaction with the cognate receptor for that envelope. The Cdom of FeLV-C env exhibits weak binding to FLVCR1 in the absence of RBD (Figure 2-7). Furthermore, virus bearing the ΔRBD envelope can confer infection of cells overexpressing FLVCR1 (Figure 2-5). These results provide conclusive evidence for an interaction between FeLV-C Cdom and its cognate receptor, FLVCR1. While the interaction between Cdom and FLVCR1 has been established, the purpose of this interaction remains unclear. It is possible that Cdom induces minute conformational changes within FLVCR1 creating a microenvironment more conducive to fusion between the viral and host membranes. The inability to interfere with FeLV-C env pseudotyped virus suggests that Cdom can only act in a complementary manner for enhancing infection. Indeed, the slight increase in FeLV-C infection of cells expressing Cdom suggests that Cdom has altered FLVCR1 expression levels or conformation at the cell surface thereby enhancing FeLV-C env binding or fusion.
The structural component of Cdom that is required for enhancement has not yet been discovered. The inability of the C2loop peptide to enhance (or inhibit) FeLV-C virus infection alludes to a structural requirement for C2loop in binding to FLVCR1. The presence of a disulfide bond between cysteine residues that surround the C2loop may provide some structure to this region of the envelope protein. This bond may force the C2 region to extrude from the globular surface of the Cdomain which may enhance its ability to interact with FLVCR1. A linear peptide structure may not suffice for this intricate interaction. The presence of peptides consisting of V3 amino acid sequence can inhibit HIV infection of CD4+ HeLa cells expressing CXCR4 or CCR5, or both, co-receptors (Verrier, et al., 1999).
93 94
This and other experiments demonstrate that the variable region 3 (V3) epitope of gp120 is a significant component of the HIV env interface involved in receptor binding and virion interaction with the host cell. Thus, the gammaretroviral envelope and lentiviral envelope receptor interactions are comparable, as the V3 loop makes interaction with HIV’s co- receptors, and C2loop makes contact with its cognate receptor. Conversely, there are contrasting effects of the presence of V3 peptide which acts to neutralize HIV infection while C2loop may act to enhance FeLV-C infection.
An interesting study would be to identify what regions of FLVCR1 are bound by Cdom. Extracelular loops 1 and 6 of FLVCR1 have been identified as critical for efficient FeLV-C binding and infection in our laboratory (Brown et al. 2006). These experiments involved hybrid construction between FLVCR1 and its paralog FLVCR2 that is not used for cell entry by FeLV-C virus. Using these hybrid receptors and the hybrid FeLV-C envelopes, we could determine which region of the SU binds to the extracellular loops of FLVCR1. This method has been used to characterize the points of interaction with A-MLV and FeLV-B pseudotyped viruses and their cognate receptors, Pit1 and Pit2 (Tailor et al. 2000). Through this analysis the points of interaction between FeLV-C RBD or Cdom and FLVCR1 may be determined. Confirming a potential binding site for Cdom may increase our understanding of the mechanism behind a second receptor-binding domain.
The interaction of two receptor-binding domains with their cognate receptor may be a common mechanism for all gammaretroviruses. It is important to test this hypothesis with other gammaretroviruses. As the receptor for FeLV-A has been cloned (Mendoza et al., 2006), similar experiments could be performed with hybrids containing the RBD of FeLV-A and Cdom of other FeLV subgroups to test their ability to bind and infect cells overexpressing feline THTR1. Of the upmost importance would be to assess FeLV-A Cdom binding to THTR1 in the absence of FeLV-A RBD. In conjunction with these experiments, another gammaretrovirus from a different class such as MLV or PERV is warranted. Recent experiments with PERV subgroup hybrids have also shown an importance of the Cdom in infectivity and tropism of the virus (Argaw, et al., 2008). The authors narrowed down the region responsible for efficient infection to the Cregion, but did not perform any experiments
95
to elucidate any intrinsic binding with its cognate receptor in the absence of RBD. Experiments showing Cdom binding to its cognate receptor for multiple gammaretroviruses would designate the two receptor-binding theory as a true mechanism for virus binding and infection with its host cell. Elucidation of this mechanism will enhance studies aiming to design novel envelope proteins targeted to a specific receptor.
4.2 Application for specific cell targeting
The host range of viruses has been changed through pseudotyping for many years. Indeed, all retroviruses used in these studies were pseudotyped viruses. However, the alternative envelope proteins used were always derived from naturally occurring envelope proteins. Thus, the host range of the pseudotyped virus would be limited by the host range of that naturally occurring virus. In some cases, it has been shown that viral targeting can be achieved through ligand-receptor interactions mediated by bivalent antibodies linked by biotin-streptavidin (Roux P et al., 1989). These manipulations often result in low infection efficiency of the virus. Various attempts have been made in the past to create retrovirus vectors that have specific ligand-receptor interactions to enhance specific cell targeting. Avian retroviruses have been shown to infect mammalian cells through addition of an integrin sequence into their envelope protein (Valsesia-Wimmann et al., 1994). Kan and colleagues successfully created an ecotropic Moloney MLV pseudotyped with both Mo- MLV env and a hybrid erythropoietin (Epo)-MLV envelope virus that was able to infect human erythroid K562 cells unlike the wildtype Mo-MLV control (Kasahara et al., 1994). Due to the presence of the natural Mo-MLV env on the surface of the virion, the hybrid virus maintained its natural host range while acquiring the ability to infect cells expressing the erythropoietin receptor (EpoR). This was a major step towards targeting of specific cells.
One of the goals of this research was to exploit the secondary binding site of FeLV-C env in the design of a target-specific retrovirus that would infect cells only expressing a particular set of receptors. In preliminary experiments, we tested the ability of ΔRBD to rescue infection of CT env psuedotyped virus. This CT virus exhibited extremely weak infectivty on cells expressing FLVCR1 due to a six amino acid insertion within the C2loop of
96
FeLV-T (Figure2-4). Unfortunately, little rescue of this weak virus was seen presumably due to faulty trimer formation of the env at the cell surface (Figure 2-5). The extra amino acids within the C2loop of FeLV-T may provide a steric hindrance to the proper structural conformation that is maintained by Cdom for efficient binding to FLVCR1. While addition of ΔRBD to CA env pseudotyped virus did provide a ten fold increase in infection (Figure 2- 5), additional experiments are required to ensure that ΔRBD can indeed rescue a defective virus. We postulate that there exists a threshold for the number of additional amino acid residues that can be tolerated within the C2 region. Further experiments to test this hypothesis involving alanine addition to C2 of FeLV-C env will determine the ideal Cdom sequence that maintains a low infection, but can be rescued by addition of ΔRBD. We proposed to use this hybrid env which shows extremely low levels of infectivity and binding, in conjunction with another receptor-specific protein, erythropoietin (Epo). In this setting a virus expressing both of these hybrid env on its surface could efficiently infect cells expressing both FLVCR1 and the erythropoietin receptor (EpoR) (Figure 4-1). The C domain could enhance hybrid FeLV-C env binding to FLVCR1 while Epo binds to its receptor. Without the C region, virions expressing the hybrid envelope are virtually ineffective at infecting cells expressing FLVCR1 (Figure 4-1). Thus, this method may provide a very specific targeting of erythroid cells as they are known to have expression of both EpoR and FLVCR1, and only cells expressing both receptors would be infected. This method of targeting specific cells for infection may provide a better mechanism at creating efficient tissue specific retroviruses.
97
Stage 1 Stage 2 Stage 3 Binding of virus to FLVCR1 and EpoR Interaction of FeLV-C C region with FLVCR1
Virus Fusion
EpoR FLVCR1 EPO
T C-terminus
C RBD
Figure 4-1. Proposed method of viral targeting to specific cells. Using a virus expressing the CT hybrid envelope and EPO fused to the C-terminal region of FeLV-C we propose to create a virus targeted to cells that express both FLVCR1 and EpoR. We want to increase the low infectivity of the CT virus by providing the functional C-terminal region for efficient binding to FLVCR1 while maintaining specificity for the EpoR. Thus, only cells expressing both receptors would be susceptible to infection by this virus.
98
4.3 FLVCR1 in DBA
There have been many ribosomal genes associated with DBA, but none of these proteins has provided insight into the molecular mechanism of DBA. It has been postulated that ribosomal proteins, while crucial for the efficient translation needed during erythropoiesis, only play a secondary or accessory role in the pathogenesis of the disease. A gene whose function is distinctly involved in erythropoiesis and whose dysregulation or lack of function could lead to the same phenotype seen in DBA is warranted. FLVCR1 has been identified as an exporter of heme (Quigley et al., 2004), which is a major component of hemoglobin involved in oxygen transport of erythroid cells. Heme has been shown to regulate globin protein synthesis and induce differentiation of erythroid cells. However, excess heme is toxic and can lead to membrane damage (Ryter and Tyrell, 2000). The Abkowitz group argue that FLVCR1 functions as a safety valve by exporting excesss heme out of the cells to prohibit toxicity (Quigley et al., 2004, Keel et al., 2008). This group has shown that inhibition of FLVCR1 in the K562 erythroleukemic cell line leads to accumulation of heme and death of erythroid cells by apoptosis (Quigley et al., 2004). Additionally, they established an FLVCR1 mouse model (Keel et al., 2008). Similar to the RPS19 knockout mouse model, FLVCR1 -/- mice die during embryogenesis, and heterozygous mice are normal with no defects in hematopoiesis (Matsson et al., 2004, Keel et al., 2008). Conversely, FLVCR1 null mice lack erythropoiesis, and have craniofacial and limb deformities similar to those found in DBA patients (Keel et al., 2008). Therefore, these FLVCR1 null and conditional knockout mice provide a much closer model to DBA than the current Rps19 mouse models. Combined with the potential genetic linkage of FLVCR1 to DBA (Quigley et al. 2004, Heyn et al. 1974), there is compelling evidence of a significant and more importantly, specific role for FLVCR1 in the manifestation of DBA. Recent research in our lab showed a defect in erythropoeisis of HSCs infected with FeLV-C env similar to that seen in DBA patients. Expression of FeLV-C env partially blocked FLVCR1 function which gave rise to a significant decrease in the number of CFU-E in comparison to cells infected with FeLV-B env, or vector only controls. These results provided strong evidence that a defect in FLVCR1 function gives rise to DBA-like symptoms. Upon analysis
99
of FLVCR1 expression levels in five DBA patients, along with the discovery of alternative splicing of the gene, the mounting evidence for the role of FLVCR1 in DBA demands further scrutiny of more DBA patients. A potential new genetic link has been found for DBA that answers the need for a gene with a more specific role in erythropoiesis than that of ribosomal proteins.
4.3.1 DBA patients exhibit decreased expression of FLVCR1 transcript
Comparison of the total amount of transcribed FLVCR1 to transcribed actin in DBA and normal patients produced an interesting discovery. All normal patients tested except for one had over one hundred fold higher expression of FLVCR1 transcript than all five DBA patients (Figure 3-7). This suggests that DBA patients may exhibit a defect in the transcription of FLVCR1 in addition to the alternative splicing (discussed below). This decrease in FLVCR1 expression in DBA patients is similar to the ex vivo cell culture model depicted with FeLV-C env expression in HSCs. These DBA patients may not have sufficient levels of FLVCR1 protein at their cell surface needed to export the increasing levels of heme causing a block in erythropoiesis due to apoptosis. There must exist some threshold amount of FLVCR1 protein required for proper erythropoiesis. As noted before, the FLVCR1 +/- heterozygous mice are completely normal (Keel et al., 2008), suggesting that transcription of one wildtype FLVCR1 gene is sufficient. A question remains regarding the threshold amount as we note one normal patient who exhibited decreased levels of FLVCR1 similar to DBA patients (Figure 3-7). It is possible that this patient has acute levels of anemia not yet diagnosed by a physician. Or this person may compensate for low expression levels through increased expression of another gene involved in erythropoiesis. The answer may lie in the second feature of DBA patients, increased alternative splicing of FLVCR1. This normal patient exhibited very little splicing of his FLVCR1 transcript, which suggests he may have surpassed the threshold amount of transcript needed to maintain healthy erythropoiesis. FLVCR1 protein expression analysis of these patients could clarify the conundrum surrounding the amount of functional FLVCR1 in the cells.
100
There are many possible reasons for the apparent decrease in transcription of FLVCR1 in DBA patients. The decrease could be due to point mutations in the promoter/enhancer region of the gene affecting efficient binding of transcription factors needed for efficient RNA production. Other genes such as NEIL2, a gene involved in DNA repair, have been shown to exhibit mutations affecting transcription leading to low level expression of the gene and subsequent manifestation of the disease, in this case a predisposition for genomic instability leading to cancer (Kinslow et al., 2008). Sequencing the promoter/enhancer regions controlling FLVCR1 expression could potentially elucidate the mechanism for this decrease. Obtaining genomic DNA samples from the five patients tested would be an excellent start to defining the mechanism behind low FLVCR1 mRNA levels.
4.3.2 DBA patients exhibit increased alternative splicing of FLVCR1
A multitude of alternatively spliced genes have been identified as being involved in disease manifestation such as the cystic fibrosis gene, CFTR, and the neurofibromatosis gene, NF1 (Srebrow A. and Kornblihtt A., 2006, Faustino, et al., 2003, Venables, JP. 2004). The cause for alternative splicing of the gene can lie within the genomic sequence encoding the 5’ or 3’ splice site as with the mutation in an intron in the ATM gene that is linked to breast cancer which causes a proportion of transcripts to skip an exon leading to truncation of the encoded protein (Broeks et al., 2003). In this study, alternative splicing of exon 2 or exon 3 of FLVCR1 was greatly enhanced in DBA patients in comparison to normal patients. The cause of the exon 2 or 3 skipping may be due to mutations in the splice acceptor or donor sites as has been seen with other diseases. The association of a frameshift mutation with exon 2 skipping and an extra amino acid insertion with exon 3 skipping in DBA patients suggest that these transcripts are not meant to be transcribed. Genomic sequencing of the intron/exon boundaries of FLVCR1 could confirm if DBA patients exhibit any point mutations in this region. Initial results from analysis of one DBA patient in comparison to genomic DNA from a human cell line and GenBank suggest that there maybe SNPs in these areas. Further analysis using more stringent methods of gene sequencing that provide
101
information about both copies of the gene such as multiplex ligation-dependent probe amplification (MLPA, Schouten et al., 2002) or quantitative multiplex PCR of short fluorescent fragments (Casilli et al., 2002) are required. These methods would provide a more indepth analysis of the intronic sequence surrounding these exons on both copies of the patient’s FLVCR1. In addition, these methods could be used in determining if any mutations exist in the promoter region of FLVCR1 that may affect transcriptional levels as discussed previously. The presence of exon2 or 3 skipping in normal patients is evident, albeit at a low frequency. The question remains as to whether these splice variants are always present with frameshifts and insertions included, and mutations in FLVCR1 of DBA patients enhance expression of the splice variants; or, if DBA patients exhibit some other mutation that leads to increased splicing of these exons or alternative splice junctions.
Splice site sequence is not the only determinant in exon skipping. In recent years, numerous studies have identified several cis-acting factors that provide an additional role in influencing exon inclusion or exclusion. These sequences occur in both introns and exons and can have a positive or negative effect on the splicing of a given exon. The sequences contained within an exon have been named exonic splicing enhancers (ESEs) or silencers (ESSs). Intronic sequences have been name intronic splicing enhancers (ISEs) or silencers (ISSs). Serine-and arginine-rich (SR) proteins and heterogenous nuclear ribonucleoproteins (hnRNP) are essential pre-mRNA splicing factors that can bind to these sequences (Sanford et al., Martinez-Contreras et al., 2007). hnRNP proteins can repress splicing by directly antagonizing the recognition of splice sites, or can interfere with the binding of proteins bound to enhancers. Conversely, several reports have described a positive role for some hnRNP proteins in pre-mRNA splicing. Thus, hnRNP proteins are important for both alternative and constitutive pre-mRNA splicing. SR proteins have been shown to be required for constitutive splicing, they bind to exon splicing enhancer (ESE) sequences and exert a positive effect on splice site selection by aiding the recruitment of spliceosome components to the 5’ and 3’ splice sites. SR proteins are also key players in controlling alternative splicing. Many alternative exons are designated by weak splicing signals, such as poor polypyrimidine (Py) tracts, which are recognized inefficiently by U2AF splicing factor. This often results in exon skipping. However, SR proteins bound to ESE’s compensate for the
102
weak Py tract by recruiting U2AF (Sanford et al., 2005), which significantly reduces exon skipping. Many diseases have been identified with various mutations that have affected splicing of a gene. These mutations can occur within non-coding regions of a gene, introns and within exons as synonomous point mutations. Upon initial assessment, these mutations may appear harmless, but mounting evidence indiciates that they can affect the binding of pre-mRNA binding proteins involved in splicing. The appearance of potential point mutations in FLVCR1 adjacent to exon 2 suggests a mechanism by which exon 2 splicing may be enhanced in DBA patients, although these SNPs need to be verified by more stringent methods as discussed above. Future experiments involving a more detailed analysis of the FLVCR1 gene in these and other patients is warranted to determine if any mutations are present that may increase binding of a splicing enhancer sequence or decrase binding of a splicing silencer sequence leading to increased exon 2 or 3 skipping.
4.3.3 A potential link between Rps19 and FLVCR1
There is increasing evidence that ribosomal proteins can exhibit functions outside of protein synthesis. Ribosomal proteins have been suggested to be involved in replication, transcription, development and splicing (Wood, 1997). A recent study using affinity purification and mass spectrometry identified 159 proteins that interact with RPS19 (Orru et al. 2007). RPS19 interacting proteins included other ribosomal proteins found in the 40S ribosomal subunit, but some proteins identified did not have roles in ribosome maturation and translation suggesting extraribosomal functions for RPS19. For example, RPS19 interacts with proteasome components, kinases and integrins (Orru et al., 2007). RSP19 was also shown to interact with subunits of splicing factor SF3B, and with the splicing factors Tra2-β and SRp30c, which belong to a family of SR proteins involved in controlling alternative splicing. The interaction of RPS19 with Tra2-β and SRp30c SR proteins is of significant interest because three of the five DBA patients tested exhibited decreased expression levels of RPS19. Furthermore, recent experiments in our laboratory have shown that down regulation of RPS19 in human cells enhances alternative splicing of the FLVCR1 transcript (Duffy, S. et al., unpublished). If Rps19 binds to SR proteins to enhance their
103 binding for exon inclusion, it is plausible that with decreased levels of Rps19, through mutation or low transcription, exon inclusion decreases as seen with exon 2 and 3 in DBA patients.
As described above, RPS19 has been shown to interact with the SR proteins Tra2- β and SRp30c. Interestingly, we have found that the FLVCR1 exon 3, which is spliced out in RPS19 reduced cells (Duffy, S et al., unpublished), and in DBA erythroid cells, contains three conserved Tra2-β binding motifs. This raises the possibility that RPS19 may control alternative splicing of FLVCR1 transcript through the action of Tra2-β. Tra2-β has been shown to be involved in various other human diseases. Tra2-β regulates alternative splicing of exon 10 of the tau gene, which is mutated in the frontotemporal dementia with Parkinsonism (FDTP), and plays a role in the manifestation of the disease (Kondo et al., 2004). Similarly, Tra2-β is induced in breast cancer and regulates alternative splicing of the CD44 gene (Watermann et al., 2006). Thus, an interaction between Tra2-β and Rps19 may be necessary for proper splicing of FLVCR1. Future experiments determining if Tra2-β is indeed involved in splicing of FLVCR1 through gel shift assays that show binding between the two could be informative. In addition, mini gene constructs containing FLVCR1 sequence could be used as a splicing reporter to assess an increase in exon 3 inclusion in the presence of co-transfected Tra2-β compared to control. Moreover, the addition of Rps19 may further enhance exon 3 inclusion while mutated Rps19 does not.
104
4.4 Concluding remarks
Research efforts to create designer retroviral vectors capable of binding to and infecting specific cell types of choice have yielded specific, but inefficient viruses. Using a group of very similar retroviruses that bind to different receptors through minute changes in their envelope protein, we have revealed a potential missing link in envelope-receptor interactions. Through our work scientists may now generate more efficient specific cell targeted γ- retroviruses that contain modified envelope protein with a fully functional RBD and Cdom, in addition to the ligand of choice. We have also identified a new specific genetic link for DBA that may surpass the current genes associated with the disease. FLVCR1 insufficiency may be present in all DBA patients, and may provide clinicians with a measurable target with which to diagnose patients or advise parents who are carriers of FLVCR1 mutations.
105
5. REFERENCES
Achong BG, Mansell PW, Epstein MA, Clifford P. An unusual virus in cultures from a human nasopharyngeal carcinoma. J Natl Cancer Inst 1971; 46 (2):299-307.
Akiyoshi DE, Denaro M, Zhu H, et al. Identification of a full-length cDNA for an endogenous retrovirus of miniature swine. J Virol 1998; 72 (5):4503-7.
Albritton LM, Tseng L, Scadden D, Cunningham JM. A putative murine ecotropic retrovirus receptor gene encodes a multiple membrane-spanning protein and confers susceptibility to virus infection. Cell 1989; 57 (4):659-66.
Anderson MM, Lauring AS, Burns CC, Overbaugh J. Identification of a cellular cofactor required for infection by feline leukemia virus. Science 2000; 287 (5459):1828-30.
Anderson MM, Lauring AS, Robertson S, Dirks C, Overbaugh J. Feline Pit2 functions as a receptor for subgroup B feline leukemia viruses. J Virol 2001; 75 (22):10563-72.
Aquino VM, Buchanan GR. Osteogenic sarcoma in a child with transfusion-dependent Diamond-Blackfan anemia. J Pediatr Hematol Oncol 1996; 18 (2):230-2.
Argaw T, Figueroa M, Salomon DR, Wilson CA. Identification of residues outside of the receptor binding domain that influence the infectivity and tropism of porcine endogenous retrovirus. J Virol 2008; 82 (15):7483-91.
Bae Y, Kingsman SM, Kingsman AJ. Functional dissection of the Moloney murine leukemia virus envelope protein gp70. J Virol 1997; 71 (3):2092-9.
Ball SE, McGuckin CP, Jenkins G, Gordon-Smith EC. Diamond-Blackfan anaemia in the U.K.: analysis of 80 cases from a 20-year birth cohort. Br J Haematol 1996; 94 (4):645-53.
Baltimore D. RNA-dependent DNA polymerase in virions of RNA tumour viruses. Nature 1970; 226 (5252):1209-11.
Barnett AL, Cunningham JM. Receptor binding transforms the surface subunit of the mammalian C-type retrovirus envelope protein from an inhibitor to an activator of fusion. J Virol 2001; 75 (19):9096-105.
Barnett AL, Davey RA, Cunningham JM. Modular organization of the Friend murine leukemia virus envelope protein underlies the mechanism of infection. Proc Natl Acad Sci U S A 2001; 98 (7):4113-8.
Barnett AL, Wensel DL, Li W, Fass D, Cunningham JM. Structure and mechanism of a coreceptor for infection by a pathogenic feline retrovirus. J Virol 2003; 77 (4):2717- 29.
106
Barre-Sinoussi F, Chermann JC, Rey F, et al. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science 1983; 220 (4599):868-71.
Battini JL, Heard JM, Danos O. Receptor choice determinants in the envelope glycoproteins of amphotropic, xenotropic, and polytropic murine leukemia viruses. J Virol 1992; 66 (3):1468-75.
Battini JL, Danos O, Heard JM. Receptor-binding domain of murine leukemia virus envelope glycoproteins. J Virol 1995; 69 (2):713-9.
Battini JL, Danos O, Heard JM. Definition of a 14-amino-acid peptide essential for the interaction between the murine leukemia virus amphotropic envelope glycoprotein and its receptor. J Virol 1998; 72 (1):428-35.
Battini JL, Rasko JE, Miller AD. A human cell-surface receptor for xenotropic and polytropic murine leukemia viruses: possible role in G protein-coupled signal transduction. Proc Natl Acad Sci U S A 1999; 96 (4):1385-90.
Beemon KL, Faras AJ, Hasse AT, Duesberg PH, Maisel JE. Genomic complexities of murine leukemia and sarcoma, reticuloendotheliosis, and visna viruses. J Virol 1976; 17 (2):525-37.
Benveniste RE, Sherr CJ, Todaro GJ. Evolution of type C viral genes: origin of feline leukemia virus. Science 1975; 190 (4217):886-8.
Billeter MA, Parsons JT, Coffin JM. The nucleotide sequence complexity of avian tumor virus RNA. Proc Natl Acad Sci U S A 1974; 71 (9):3560-4.
Bittner JJ. SOME POSSIBLE EFFECTS OF NURSING ON THE MAMMARY GLAND TUMOR INCIDENCE IN MICE. Science 1936; 84 (2172):162.
Bonner TI, Todaro GJ. Carnivores have sequences in their cellular DNA distantly related to the primate endogenous virus, MAC-1. Virology 1979; 94 (1):224-7.
Boocock GR, Morrison JA, Popovic M, et al. Mutations in SBDS are associated with Shwachman-Diamond syndrome. Nat Genet 2003; 33 (1):97-101.
Boocock GR, Marit MR, Rommens JM. Phylogeny, sequence conservation, and functional complementation of the SBDS protein family. Genomics 2006; 87 (6):758-71.
Boomer S, Eiden M, Burns CC, Overbaugh J. Three distinct envelope domains, variably present in subgroup B feline leukemia virus recombinants, mediate Pit1 and Pit2 receptor recognition. J Virol 1997; 71 (11):8116-23.
Broeks A, Urbanus JH, de KP, et al. IVS10-6T>G, an ancient ATM germline mutation linked with breast cancer. Hum Mutat 2003; 21 (5):521-8.
107
Brojatsch J, Kristal BS, Viglianti GA, et al. Feline leukemia virus subgroup C phenotype evolves through distinct alterations near the N terminus of the envelope surface glycoprotein. Proc Natl Acad Sci U S A 1992; 89 (18):8457-61.
Brown JK, Fung C, Tailor CS. Comprehensive mapping of receptor-functioning domains in feline leukemia virus subgroup C receptor FLVCR1. J Virol 2006; 80 (4):1742-51.
Bushman FD, Craigie R. Integration of human immunodeficiency virus DNA: adduct interference analysis of required DNA sites. Proc Natl Acad Sci U S A 1992; 89 (8):3458-62.
Campagnoli MF, Garelli E, Quarello P, et al. Molecular basis of Diamond-Blackfan anemia: new findings from the Italian registry and a review of the literature. Haematologica 2004; 89 (4):480-9.
Campagnoli MF, Ramenghi U, Armiraglio M, et al. RPS19 mutations in patients with Diamond-Blackfan anemia. Hum Mutat 2008; 29 (7):911-20.
Casilli F, Di Rocco ZC, Gad S, et al. Rapid detection of novel BRCA1 rearrangements in high-risk breast-ovarian cancer families using multiplex PCR of short fluorescent fragments. Hum Mutat 2002; 20 (3):218-26.
Chalker DL, Sandmeyer SB. Ty3 integrates within the region of RNA polymerase III transcription initiation. Genes Dev 1992; 6 (1):117-28.
Chang YF, Imam JS, Wilkinson MF. The nonsense-mediated decay RNA surveillance pathway. Annu Rev Biochem 2007; 76:51-74.
Chatr-Aryamontri A, Angelini M, Garelli E, et al. Nonsense-mediated and nonstop decay of ribosomal protein S19 mRNA in Diamond-Blackfan anemia. Hum Mutat 2004; 24 (6):526-33.
Check E. Gene therapy put on hold as third child develops cancer. Nature 2005; 433 (7026):561.
Cheng HH, Anderson MM, Hankenson FC, et al. Envelope determinants for dual-receptor specificity in feline leukemia virus subgroup A and T variants. J Virol 2006; 80 (4):1619-28.
Chiocchetti A, Gibello L, Carando A, et al. Interactions between RPS19, mutated in Diamond-Blackfan anemia, and the PIM-1 oncoprotein. Haematologica 2005; 90 (11):1453-62.
Choesmel V, Bacqueville D, Rouquette J, et al. Impaired ribosome biogenesis in Diamond- Blackfan anemia. Blood 2007; 109 (3):1275-83.
108
Choesmel V, Fribourg S, guissa-Toure AH, et al. Mutation of ribosomal protein RPS24 in Diamond-Blackfan anemia results in a ribosome biogenesis disorder. Hum Mol Genet 2008; 17 (9):1253-63.
Cmejla R, Cmejlova J, Handrkova H, Petrak J, Pospisilova D. Ribosomal protein S17 gene (RPS17) is mutated in Diamond-Blackfan anemia. Hum Mutat 2007; 28 (12):1178- 82.
Cmejla R, Cmejlova J, Handrkova H, Petrak J, Pospisilova D. Ribosomal protein S17 gene (RPS17) is mutated in Diamond-Blackfan anemia. Hum Mutat 2007; 28 (12):1178- 82.
Cmejlova J, Cerna Z, Votava T, Pospisilova D, Cmejla R. Identification of a new in-frame deletion of six amino acids in ribosomal protein S19 in a patient with Diamond- Blackfan anemia. Blood Cells Mol Dis 2006; 36 (3):337-41.
Cosset FL, Takeuchi Y, Battini JL, Weiss RA, Collins MK. High-titer packaging cells producing recombinant retroviruses resistant to human serum. J Virol 1995; 69 (12):7430-6.
Cosset FL, Morling FJ, Takeuchi Y, et al. Retroviral retargeting by envelopes expressing an N-terminal binding domain. J Virol 1995; 69 (10):6314-22.
Craigie R. Hotspots and warm spots: integration specificity of retroelements. Trends Genet 1992; 8 (6):187-90.
Dalgleish AG, Beverley PC, Clapham PR, et al. The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus. Nature 1984; 312 (5996):763-7.
Dean GA, Groshek PM, Mullins JI, Hoover EA. Hematopoietic target cells of anemogenic subgroup C versus nonanemogenic subgroup A feline leukemia virus. J Virol 1992; 66 (9):5561-8.
Deng H, Liu R, Ellmeier W, et al. Identification of a major co-receptor for primary isolates of HIV-1. Nature 1996; 381 (6584):661-6.
Diamond LK, Blackfan KD. Hypoplastic Anemia. In: 1938:464-7.
Dianzani I, Garelli E, Dompe C, et al. Mutations in the erythropoietin receptor gene are not a common cause of Diamond-Blackfan anemia. Blood 1996; 87 (6):2568-72.
Dianzani I, Garelli E, Ramenghi U. Diamond-Blackfan anemia: a congenital defect in erythropoiesis. Haematologica 1996; 81 (6):560-72.
Dianzani I, Garelli E, Ramenghi U. Diamond-Blackfan Anaemia: an overview. Paediatr Drugs 2000; 2 (5):345-55.
109
Dildine SL, Sandmeyer SB. Integration of the yeast retrovirus-like element Ty3 upstream of a human tRNA gene expressed in yeast. Gene 1997; 194 (2):227-33.
Dildine SL, Respess J, Jolly D, Sandmeyer SB. A chimeric Ty3/Moloney murine leukemia virus integrase protein is active in vivo. J Virol 1998; 72 (5):4297-307.
Donahue PR, Quackenbush SL, Gallo MV, et al. Viral genetic determinants of T-cell killing and immunodeficiency disease induction by the feline leukemia virus FeLV-FAIDS. J Virol 1991; 65 (8):4461-9.
Dragic T, Litwin V, Allaway GP, et al. HIV-1 entry into CD4+ cells is mediated by the chemokine receptor CC-CKR-5. Nature 1996; 381 (6584):667-73.
Draptchinskaia N, Gustavsson P, Andersson B, et al. The gene encoding ribosomal protein S19 is mutated in Diamond-Blackfan anaemia. Nat Genet 1999; 21 (2):169-75.
Dror Y, Freedman MH. Shwachman-Diamond syndrome marrow cells show abnormally increased apoptosis mediated through the Fas pathway. Blood 2001; 97 (10):3011-6.
Duesberg PH, Martin GS, Vogt PK. Glycoprotein components of avian and murine RNA tumor viruses. Virology 1970; 41 (4):631-46.
Ebert BL, Lee MM, Pretz JL, et al. An RNA interference model of RPS19 deficiency in Diamond-Blackfan anemia recapitulates defective hematopoiesis and rescue by dexamethasone: identification of dexamethasone-responsive genes by microarray. Blood 2005; 105 (12):4620-6.
Elder JH, Mullins JI. Nucleotide sequence of the envelope gene of Gardner-Arnstein feline leukemia virus B reveals unique sequence homologies with a murine mink cell focus- forming virus. J Virol 1983; 46 (3):871-80.
Ellerman C, Bang O. Centralbl Bakteriol 1908; 46:595-609.
Ellis SR, Massey AT. Diamond Blackfan anemia: A paradigm for a ribosome-based disease. Med Hypotheses 2006; 66 (3):643-8.
ElSharawy A, Manaster C, Teuber M, et al. SNPSplicer: systematic analysis of SNP- dependent splicing in genotyped cDNAs. Hum Mutat 2006; 27 (11):1129-34.
Faix PH, Feldman SA, Overbaugh J, Eiden MV. Host range and receptor binding properties of vectors bearing feline leukemia virus subgroup B envelopes can be modulated by envelope sequences outside of the receptor binding domain. J Virol 2002; 76 (23):12369-75.
Farrar JE, Nater M, Caywood E, et al. Abnormalities of the large ribosomal subunit protein, Rpl35A, in diamond-blackfan anemia. Blood 2008.
110
Fass D, Davey RA, Hamson CA, et al. Structure of a murine leukemia virus receptor-binding glycoprotein at 2.0 angstrom resolution. Science 1997; 277 (5332):1662-6.
Faustino NA, Cooper TA. Pre-mRNA splicing and human disease. Genes Dev 2003; 17 (4):419-37.
Feng Y, Broder CC, Kennedy PE, Berger EA. HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor. Science 1996; 272 (5263):872-7.
Flygare J, Kiefer T, Miyake K, et al. Deficiency of ribosomal protein S19 in CD34+ cells generated by siRNA blocks erythroid development and mimics defects seen in Diamond-Blackfan anemia. Blood 2005; 105 (12):4627-34.
Flygare J, Karlsson S. Diamond-Blackfan anemia: erythropoiesis lost in translation. Blood 2007; 109 (8):3152-4.
Freedman MH. Diamond-Blackfan anaemia. Baillieres Best Pract Res Clin Haematol 2000; 13 (3):391-406.
Furth J, Boon MC. ENHANCEMENT OF LEUKEMOGENIC ACTION OF METHYLCHOLANTHRENE BY PRE-IRRADIATION WITH X-RAYS. Science 1943; 98 (2536):138-9.
Gazda H, Lipton JM, Willig TN, et al. Evidence for linkage of familial Diamond-Blackfan anemia to chromosome 8p23.3-p22 and for non-19q non-8p disease. Blood 2001; 97 (7):2145-50.
Gazda HT, Zhong R, Long L, et al. RNA and protein evidence for haplo-insufficiency in Diamond-Blackfan anaemia patients with RPS19 mutations. Br J Haematol 2004; 127 (1):105-13.
Gazda HT, Sieff CA. Recent insights into the pathogenesis of Diamond-Blackfan anaemia. Br J Haematol 2006; 135 (2):149-57.
Gazda HT, Grabowska A, Merida-Long LB, et al. Ribosomal protein S24 gene is mutated in Diamond-Blackfan anemia. Am J Hum Genet 2006; 79 (6):1110-8.
Gemeniano M, Mpanju O, Salomon DR, Eiden MV, Wilson CA. The infectivity and host range of the ecotropic porcine endogenous retrovirus, PERV-C, is modulated by residues in the C-terminal region of its surface envelope protein. Virology 2006; 346 (1):108-17.
Ghosh AK, Bachmann MH, Hoover EA, Mullins JI. Identification of a putative receptor for subgroup A feline leukemia virus on feline T cells. J Virol 1992; 66 (6):3707-14.
111
Glader BE, Backer K. Elevated red cell adenosine deaminase activity: a marker of disordered erythropoiesis in Diamond-Blackfan anaemia and other haematologic diseases. Br J Haematol 1988; 68 (2):165-8.
Gonzales B, Henning D, So RB, et al. The Treacher Collins syndrome (TCOF1) gene product is involved in pre-rRNA methylation. Hum Mol Genet 2005; 14 (14):2035- 43.
Grange MP, Blot V, Delamarre L, et al. Identification of two intracellular mechanisms leading to reduced expression of oncoretrovirus envelope glycoproteins at the cell surface. J Virol 2000; 74 (24):11734-43.
Grant CK, Ernisse BJ, Jarrett O, Jones FR. Feline leukemia virus envelope gp70 of subgroups B and C defined by monoclonal antibodies with cytotoxic and neutralizing functions. J Immunol 1983; 131 (6):3042-8.
Gray KD, Roth MJ. Mutational analysis of the envelope gene of Moloney murine leukemia virus. J Virol 1993; 67 (6):3489-96.
GROSS L. "Spontaneous" leukemia developing in C3H mice following inoculation in infancy, with AK-leukemic extracts, or AK-embrvos. Proc Soc Exp Biol Med 1951; 76 (1):27-32.
Gustavsson P, Willing TN, van HA, et al. Diamond-Blackfan anaemia: genetic homogeneity for a gene on chromosome 19q13 restricted to 1.8 Mb. Nat Genet 1997; 16 (4):368- 71.
Hacein-Bey-Abina S, Le DF, Carlier F, et al. Sustained correction of X-linked severe combined immunodeficiency by ex vivo gene therapy. N Engl J Med 2002; 346 (16):1185-93.
Hacein-Bey-Abina S, Fischer A, Cavazzana-Calvo M. Gene therapy of X-linked severe combined immunodeficiency. Int J Hematol 2002; 76 (4):295-8.
Hacein-Bey-Abina S, von KC, Schmidt M, et al. A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N Engl J Med 2003; 348 (3):255-6.
Hacein-Bey-Abina S, von KC, Schmidt M, et al. LMO2-associated clonal T cell proliferation in two patients after gene therapy for SCID-X1. Science 2003; 302 (5644):415-9.
Hardy WD, Jr., Hess PW, MacEwen EG, et al. Biology of feline leukemia virus in the natural environment. Cancer Res 1976; 36 (2 pt 2):582-8.
Hardy WD, Jr. Naturally occurring retroviruses (RNA tumor viruses). II. Cancer Invest 1983; 1 (2):163-74.
112
Hardy WD, Jr., Zuckerman EE. Development of the immunofluorescent antibody test for detection of feline leukemia virus infection in cats. J Am Vet Med Assoc 1991; 199 (10):1327-35.
Haroon ZA, Amin K, Jiang X, Arcasoy MO. A novel role for erythropoietin during fibrin- induced wound-healing response. Am J Pathol 2003; 163 (3):993-1000.
Hartley O, Klasse PJ, Sattentau QJ, Moore JP. V3: HIV's switch-hitter. AIDS Res Hum Retroviruses 2005; 21 (2):171-89.
Heyn R, Kurczynski E, Schmickel R. The association of Blackfan-Diamond syndrome, physical abnormalities, and an abnormality of chromosome 1. J Pediatr 1974; 85 (4):531-3.
Hoover EA, Olsen RG, Hardy WD, Jr., Schaller JP, Mathes LE. Feline leukemia virus infection: age-related variation in response of cats to experimental infection. J Natl Cancer Inst 1976; 57 (2):365-9.
Hoover EA, Mullins JI, Quackenbush SL, Gasper PW. Experimental transmission and pathogenesis of immunodeficiency syndrome in cats. Blood 1987; 70 (6):1880-92.
Huang CC, Tang M, Zhang MY, et al. Structure of a V3-containing HIV-1 gp120 core. Science 2005; 310 (5750):1025-8.
Huang Q, Robledo S, Wilson DB, Bessler M, Mason PJ. A four base pair insertion in exon 1 of the RPS19 gene is a common polymorphism in African-Americans. Br J Haematol 2006; 135 (5):745-6.
Jarrett O, Laird HM, Crighton GW, Jarrett WF, Hay D. Advances in feline leukemia. Bibl Haematol 1968; 30:244-54.
Jarrett O, Laird HM, Hay D, Crighton GW. Replication of cat leukemia virus in cell cultures. Nature 1968; 219 (5153):521-2.
Jarrett O, Laird HM, Hay D. Determinants of the host range of feline leukaemia viruses. J Gen Virol 1973; 20 (2):169-75.
Jarrett O, Edney AT, Toth S, Hay D. Feline leukaemia virus-free lymphosarcoma in a specific pathogen free cat. Vet Rec 1984; 115 (10):249-50.
Jarrett O, Golder MC, Toth S, Onions DE, Stewart MF. Interaction between feline leukaemia virus subgroups in the pathogenesis of erythroid hypoplasia. Int J Cancer 1984; 34 (2):283-8.
Jarrett O. Vaccination against feline leukaemia virus. Vet Rec 1994; 134 (8):198.
113
Josephson NC, Sabo KM, Abkowitz JL. Transduction of feline hematopoietic cells by oncoretroviral vectors pseudotyped with the subgroup A feline leukemia virus (FeLV- A). Mol Ther 2000; 2 (1):56-62.
Kasahara N, Dozy AeM, Kan YW. Tissue-Specific Targeting of Retroviral Vectors Through Ligand-Receptor Interactions. Science 1994; 266 (5189):1373-6.
Katz RA, Skalka AM. The retroviral enzymes. Annu Rev Biochem 1994; 63:133-73.
Kavanaugh MP, Miller DG, Zhang W, et al. Cell-surface receptors for gibbon ape leukemia virus and amphotropic murine retrovirus are inducible sodium-dependent phosphate symporters. Proc Natl Acad Sci U S A 1994; 91 (15):7071-5.
Kavanaugh MP, Wang H, Zhang Z, et al. Control of cationic amino acid transport and retroviral receptor functions in a membrane protein family. J Biol Chem 1994; 269 (22):15445-50.
Kavanaugh MP, Wang H, Boyd CA, North RA, Kabat D. Cell surface receptor for ecotropic host-range mouse retroviruses: a cationic amino acid transporter. Arch Virol Suppl 1994; 9:485-94.
Kawakami TG, Huff SD, Buckley PM, et al. C-type virus associated with gibbon lymphosarcoma. Nat New Biol 1972; 235 (58):170-1.
Kawakami TG, Buckley PM. Antigenic studies on gibbon type-C viruses. Transplant Proc 1974; 6 (2):193-6.
Keel SB, Doty RT, Yang Z, et al. A heme export protein is required for red blood cell differentiation and iron homeostasis. Science 2008; 319 (5864):825-8.
Kim JW, Closs EI, Albritton LM, Cunningham JM. Transport of cationic amino acids by the mouse ecotropic retrovirus receptor. Nature 1991; 352 (6337):725-8.
Kinslow CJ, El-Zein RA, Hill CE, Wickliffe JK, bdel-Rahman SZ. Single nucleotide polymorphisms 5' upstream the coding region of the NEIL2 gene influence gene transcription levels and alter levels of genetic damage. Genes Chromosomes Cancer 2008.
Kirchner J, Connolly CM, Sandmeyer SB. Requirement of RNA polymerase III transcription factors for in vitro position-specific integration of a retroviruslike element. Science 1995; 267 (5203):1488-91.
Klatzmann D, Champagne E, Chamaret S, et al. T-lymphocyte T4 molecule behaves as the receptor for human retrovirus LAV. Nature 1984; 312 (5996):767-8.
Koga Y, Ohga S, Nomura A, Takada H, Hara T. Reduced gene expression of clustered ribosomal proteins in Diamond-Blackfan anemia patients without RPS19 gene mutations. J Pediatr Hematol Oncol 2006; 28 (6):355-61.
114
Kondo S, Yamamoto N, Murakami T, et al. Tra2 beta, SF2/ASF and SRp30c modulate the function of an exonic splicing enhancer in exon 10 of tau pre-mRNA. Genes Cells 2004; 9 (2):121-30.
Kung HJ, Bailey JM, Davidson N, et al. Electron microscope studies of tumor virus RNA. Cold Spring Harb Symp Quant Biol 1975; 39 Pt 2:827-34.
Kuramitsu M, Hamaguchi I, Takuo M, et al. Deficient RPS19 protein production induces cell cycle arrest in erythroid progenitor cells. Br J Haematol 2008; 140 (3):348-59.
Kwong PD, Wyatt R, Robinson J, et al. Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor and a neutralizing human antibody. Nature 1998; 393 (6686):648-59.
Kwong PD, Wyatt R, Majeed S, et al. Structures of HIV-1 gp120 envelope glycoproteins from laboratory-adapted and primary isolates. Structure 2000; 8 (12):1329-39.
Lavillette D, Ruggieri A, Russell SJ, Cosset FL. Activation of a cell entry pathway common to type C mammalian retroviruses by soluble envelope fragments. J Virol 2000; 74 (1):295-304.
Lavillette D, Boson B, Russell SJ, Cosset FL. Activation of membrane fusion by murine leukemia viruses is controlled in cis or in trans by interactions between the receptor- binding domain and a conserved disulfide loop of the carboxy terminus of the surface glycoprotein. J Virol 2001; 75 (8):3685-95.
Lavillette D, Ruggieri A, Boson B, Maurice M, Cosset FL. Relationship between SU subdomains that regulate the receptor-mediated transition from the native (fusion- inhibited) to the fusion-active conformation of the murine leukemia virus glycoprotein. J Virol 2002; 76 (19):9673-85.
Lenvik T, Lund TC, Verfaillie CM. Blockerette-ligated capture T7-amplified RT-PCR, a new method for determining flanking sequences. Mol Ther 2002; 6 (1):113-8.
Letvin NL, Daniel MD, Sehgal PK, et al. Induction of AIDS-like disease in macaque monkeys with T-cell tropic retrovirus STLV-III. Science 1985; 230 (4721):71-3.
Lipton JM, Federman N, Khabbaze Y, et al. Osteogenic sarcoma associated with Diamond- Blackfan anemia: a report from the Diamond-Blackfan Anemia Registry. J Pediatr Hematol Oncol 2001; 23 (1):39-44.
Lipton JM. Diamond blackfan anemia: New paradigms for a "not so pure" inherited red cell aplasia. Semin Hematol 2006; 43 (3):167-77.
Lipton JM, Atsidaftos E, Zyskind I, Vlachos A. Improving clinical care and elucidating the pathophysiology of Diamond Blackfan anemia: an update from the Diamond Blackfan Anemia Registry. Pediatr Blood Cancer 2006; 46 (5):558-64.
115
Liu JM, Ellis SR. Ribosomes and marrow failure: coincidental association or molecular paradigm? Blood 2006; 107 (12):4583-8.
Loken MR, Shah VO, Dattilio KL, Civin CI. Flow cytometric analysis of human bone marrow: I. Normal erythroid development. Blood 1987; 69 (1):255-63.
Lucas ML, Seidel NE, Porada CD, et al. Improved transduction of human sheep repopulating cells by retrovirus vectors pseudotyped with feline leukemia virus type C or RD114 envelopes. Blood 2005; 106 (1):51-8.
Maeda N, Toku S, Kenmochi N, Tanaka T. A novel nucleolar protein interacts with ribosomal protein S19. Biochem Biophys Res Commun 2006; 339 (1):41-6.
Maouche L, Tournamille C, Hattab C, et al. Cloning of the gene encoding the human erythropoietin receptor. Blood 1991; 78 (10):2557-63.
Marin M, Tailor CS, Nouri A, Kozak SL, Kabat D. Polymorphisms of the cell surface receptor control mouse susceptibilities to xenotropic and polytropic leukemia viruses. J Virol 1999; 73 (11):9362-8.
Marin M, Tailor CS, Nouri A, Kabat D. Sodium-dependent neutral amino acid transporter type 1 is an auxiliary receptor for baboon endogenous retrovirus. J Virol 2000; 74 (17):8085-93.
Marshall E. Gene therapy. Second child in French trial is found to have leukemia. Science 2003; 299 (5605):320.
Martin AN, Li Y. RNase MRP RNA and human genetic diseases. Cell Res 2007; 17 (3):219- 26.
Martinez-Contreras R, Cloutier P, Shkreta L, et al. hnRNP proteins and splicing control. Adv Exp Med Biol 2007; 623:123-47.
Matsson H, Davey EJ, Draptchinskaia N, et al. Targeted disruption of the ribosomal protein S19 gene is lethal prior to implantation. Mol Cell Biol 2004; 24 (9):4032-7.
Matsson H, Davey EJ, Frojmark AS, et al. Erythropoiesis in the Rps19 disrupted mouse: Analysis of erythropoietin response and biochemical markers for Diamond-Blackfan anemia. Blood Cells Mol Dis 2006; 36 (2):259-64.
McAllister RM, Melnyk J, Finkelstein JZ, Adams EC, Jr., Gardner MB. Cultivation in vitro of cells derived from a human rhabdomyosarcoma. Cancer 1969; 24 (3):520-6.
Mendoza R, Anderson MM, Overbaugh J. A putative thiamine transport protein is a receptor for feline leukemia virus subgroup A. J Virol 2006; 80 (7):3378-85.
116
Miller DG, Edwards RH, Miller AD. Cloning of the cellular receptor for amphotropic murine retroviruses reveals homology to that for gibbon ape leukemia virus. Proc Natl Acad Sci U S A 1994; 91 (1):78-82.
Miyake K, Flygare J, Kiefer T, et al. Development of cellular models for ribosomal protein S19 (RPS19)-deficient diamond-blackfan anemia using inducible expression of siRNA against RPS19. Mol Ther 2005; 11 (4):627-37.
Montanaro L, Chilla A, Trere D, et al. Increased mortality rate and not impaired ribosomal biogenesis is responsible for proliferative defect in dyskeratosis congenita cell lines. J Invest Dermatol 2002; 118 (1):193-8.
Morimoto K, Lin S, Sakamoto K. The functions of RPS19 and their relationship to Diamond- Blackfan anemia: a review. Mol Genet Metab 2007; 90 (4):358-62.
Moser M, Burns CC, Boomer S, Overbaugh J. The host range and interference properties of two closely related feline leukemia variants suggest that they use distinct receptors. Virology 1998; 242 (2):366-77.
Myers G. Molecular investigation of HIV transmission. Ann Intern Med 1994; 121 (11):889- 90.
Neil JC, Onions DE. Feline leukaemia viruses: molecular biology and pathogenesis. Anticancer Res 1985; 5 (1):49-63.
Neil JC, Fulton R, Rigby M, Stewart M. Feline leukaemia virus: generation of pathogenic and oncogenic variants. Curr Top Microbiol Immunol 1991; 171:67-93.
Nguyen TH, Pages JC, Farge D, Briand P, Weber A. Amphotropic retroviral vectors displaying hepatocyte growth factor-envelope fusion proteins improve transduction efficiency of primary hepatocytes. Hum Gene Ther 1998; 9 (17):2469-79.
O'Hara B, Johann SV, Klinger HP, et al. Characterization of a human gene conferring sensitivity to infection by gibbon ape leukemia virus. Cell Growth Differ 1990; 1 (3):119-27.
Ochman H, Gerber AS, Hartl DL. Genetic applications of an inverse polymerase chain reaction. Genetics 1988; 120 (3):621-3.
Ohene-Abuakwa Y, Orfali KA, Marius C, Ball SE. Two-phase culture in Diamond Blackfan anemia: localization of erythroid defect. Blood 2005; 105 (2):838-46.
Orfali KA, Ohene-Abuakwa Y, Ball SE. Diamond Blackfan anaemia in the UK: clinical and genetic heterogeneity. Br J Haematol 2004; 125 (2):243-52.
Orru S, Aspesi A, Armiraglio M, et al. Analysis of the ribosomal protein S19 interactome. Mol Cell Proteomics 2007; 6 (3):382-93.
117
Overbaugh J, Donahue PR, Quackenbush SL, Hoover EA, Mullins JI. Molecular cloning of a feline leukemia virus that induces fatal immunodeficiency disease in cats. Science 1988; 239 (4842):906-10.
Overbaugh J, Miller AD, Eiden MV. Receptors and entry cofactors for retroviruses include single and multiple transmembrane-spanning proteins as well as newly described glycophosphatidylinositol-anchored and secreted proteins. Microbiol Mol Biol Rev 2001; 65 (3):371-89, table.
Pacitti AM, Jarrett O. Duration of the latent state in feline leukaemia virus infections. Vet Rec 1985; 117 (18):472-4.
Pedersen NC, Ho EW, Brown ML, Yamamoto JK. Isolation of a T-lymphotropic virus from domestic cats with an immunodeficiency-like syndrome. Science 1987; 235 (4790):790-3.
Peng WJ, Chang CM, Lin TH. Target integration by a chimeric Sp1 zinc finger domain- Moloney murine leukemia virus integrase in vivo. J Biomed Sci 2002; 9 (2):171-84.
Pepinsky RB, Vogt VM. Identification of retrovirus matrix proteins by lipid-protein cross- linking. J Mol Biol 1979; 131 (4):819-37.
Pettigrew CA, Brown MA. Pre-mRNA splicing aberrations and cancer. Front Biosci 2008; 13:1090-105.
Poiesz BJ, Ruscetti FW, Gazdar AF, et al. Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci U S A 1980; 77 (12):7415-9.
Porter CD, Collins MK, Tailor CS, et al. Comparison of efficiency of infection of human gene therapy target cells via four different retroviral receptors. Hum Gene Ther 1996; 7 (8):913-9.
Pospisilova D, Cmejlova J, Hak J, Adam T, Cmejla R. Successful treatment of a Diamond- Blackfan anemia patient with amino acid leucine. Haematologica 2007; 92 (5):e66- e67.
Quigley JG, Burns CC, Anderson MM, et al. Cloning of the cellular receptor for feline leukemia virus subgroup C (FeLV-C), a retrovirus that induces red cell aplasia. Blood 2000; 95 (3):1093-9.
Quigley JG, Yang Z, Worthington MT, et al. Identification of a human heme exporter that is essential for erythropoiesis. Cell 2004; 118 (6):757-66.
Quigley JG, Gazda H, Yang Z, et al. Investigation of a putative role for FLVCR, a cytoplasmic heme exporter, in Diamond-Blackfan anemia. Blood Cells Mol Dis 2005; 35 (2):189-92.
118
Quigley JP, Rifkin DB, Reich E. Phospholipid composition of Rous sarcoma virus, host cell membranes and other enveloped RNA viruses. Virology 1971; 46 (1):106-16.
Rasko JE, Battini JL, Gottschalk RJ, Mazo I, Miller AD. The RD114/simian type D retrovirus receptor is a neutral amino acid transporter. Proc Natl Acad Sci U S A 1999; 96 (5):2129-34.
Rey MA, Prasad R, Tailor CS. The C domain in the surface envelope glycoprotein of subgroup C feline leukemia virus is a second receptor-binding domain. Virology 2008; 370 (2):273-84.
Riedel N, Hoover EA, Gasper PW, Nicolson MO, Mullins JI. Molecular analysis and pathogenesis of the feline aplastic anemia retrovirus, feline leukemia virus C-Sarma. J Virol 1986; 60 (1):242-50.
Rigby MA, Rojko JL, Stewart MA, et al. Partial dissociation of subgroup C phenotype and in vivo behaviour in feline leukaemia viruses with chimeric envelope genes. J Gen Virol 1992; 73 ( Pt 11):2839-47.
Rous P. A SARCOMA OF THE FOWL TRANSMISSIBLE BY AN AGENT SEPARABLE FROM THE TUMOR CELLS. Journal of Experimental Medicine 1911; 13:397-411.
Roux P, Jeanteur P, Piechaczyk M. A versatile and potentially general approach to the targeting of specific cell types by retroviruses: application to the infection of human cells by means of major histocompatibility complex class I and class II antigens by mouse ecotropic murine leukemia virus-derived viruses. Proc Natl Acad Sci U S A 1989; 86 (23):9079-83.
Rusche JR, Javaherian K, McDanal C, et al. Antibodies that inhibit fusion of human immunodeficiency virus-infected cells bind a 24-amino acid sequence of the viral envelope, gp120. Proc Natl Acad Sci U S A 1988; 85 (9):3198-202.
RUSTIGIAN R, JOHNSTON P, REIHART H. Infection of monkey kidney tissue cultures with virus-like agents. Proc Soc Exp Biol Med 1955; 88 (1):8-16.
Ryter SW, Tyrrell RM. The heme synthesis and degradation pathways: role in oxidant sensitivity. Heme oxygenase has both pro- and antioxidant properties. Free Radic Biol Med 2000; 28 (2):289-309.
Sandrin V, Russell SJ, Cosset FL. Targeting retroviral and lentiviral vectors. Curr Top Microbiol Immunol 2003; 281:137-78.
Sandrin V, Muriaux D, Darlix JL, Cosset FL. Intracellular trafficking of Gag and Env proteins and their interactions modulate pseudotyping of retroviruses. J Virol 2004; 78 (13):7153-64.
Sanford JR, Ellis J, Caceres JF. Multiple roles of arginine/serine-rich splicing factors in RNA processing. Biochem Soc Trans 2005; 33 (Pt 3):443-6.
119
Sarma PS, Log T. Subgroup classification of feline leukemia and sarcoma viruses by viral interference and neutralization tests. Virology 1973; 54 (1):160-9.
Sarma PS, Jain D, Hill PR. In vitro host range of feline leukemia virus. Bibl Haematol 1975 (40):489-92.
Schmidt M, Zickler P, Hoffmann G, et al. Polyclonal long-term repopulating stem cell clones in a primate model. Blood 2002; 100 (8):2737-43.
Schmidt M, Schwarzwaelder K, Bartholomae C, et al. High-resolution insertion-site analysis by linear amplification-mediated PCR (LAM-PCR). Nat Methods 2007; 4 (12):1051- 7.
Schouten JP, McElgunn CJ, Waaijer R, et al. Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res 2002; 30 (12):e57.
Schramm L, Hernandez N. Recruitment of RNA polymerase III to its target promoters. Genes Dev 2002; 16 (20):2593-620.
Schroder AR, Shinn P, Chen H, et al. HIV-1 integration in the human genome favors active genes and local hotspots. Cell 2002; 110 (4):521-9.
Sheets RL, Pandey R, Jen WC, Roy-Burman P. Recombinant feline leukemia virus genes detected in naturally occurring feline lymphosarcomas. J Virol 1993; 67 (6):3118-25.
Shelton GH, Grant CK, Cotter SM, et al. Feline immunodeficiency virus and feline leukemia virus infections and their relationships to lymphoid malignancies in cats: a retrospective study (1968-1988). J Acquir Immune Defic Syndr 1990; 3 (6):623-30.
Shojima T, Nakata R, Miyazawa T. Host cell range of T-lymphotropic feline leukemia virus in vitro. Biochem Biophys Res Commun 2006; 345 (4):1466-70.
Siren AL, Knerlich F, Poser W, et al. Erythropoietin and erythropoietin receptor in human ischemic/hypoxic brain. Acta Neuropathol 2001; 101 (3):271-6.
Snyder HW, Jr., Jones FR, Day NK, Hardy WD, Jr. Isolation and characterization of circulating feline leukemia virus-immune complexes from plasma of persistently infected pet cats removed by ex vivo immunosorption. J Immunol 1982; 128 (6):2726-30.
Soe LH, Devi BG, Mullins JI, Roy-Burman P. Molecular cloning and characterization of endogenous feline leukemia virus sequences from a cat genomic library. J Virol 1983; 46 (3):829-40.
Sommerfelt MA, Weiss RA. Receptor interference groups of 20 retroviruses plating on human cells. Virology 1990; 176 (1):58-69.
120
Soulet F, Al ST, Roga S, Amalric F, Bouche G. Fibroblast growth factor-2 interacts with free ribosomal protein S19. Biochem Biophys Res Commun 2001; 289 (2):591-6.
Srebrow A, Kornblihtt AR. The connection between splicing and cancer. J Cell Sci 2006; 119 (Pt 13):2635-41.
Starcich BR, Hahn BH, Shaw GM, et al. Identification and characterization of conserved and variable regions in the envelope gene of HTLV-III/LAV, the retrovirus of AIDS. Cell 1986; 45 (5):637-48.
Stephenson JR, Axelrad AA, McLeod DL, Shreeve MM. Induction of colonies of hemoglobin-synthesizing cells by erythropoietin in vitro. Proc Natl Acad Sci U S A 1971; 68 (7):1542-6.
Stewart MA, Warnock M, Wheeler A, et al. Nucleotide sequences of a feline leukemia virus subgroup A envelope gene and long terminal repeat and evidence for the recombinational origin of subgroup B viruses. J Virol 1986; 58 (3):825-34.
Sugai J, Eiden M, Anderson MM, et al. Identification of envelope determinants of feline leukemia virus subgroup B that permit infection and gene transfer to cells expressing human Pit1 or Pit2. J Virol 2001; 75 (15):6841-9.
Tailor CS, Takeuchi Y, O'Hara B, et al. Mutation of amino acids within the gibbon ape leukemia virus (GALV) receptor differentially affects feline leukemia virus subgroup B, simian sarcoma-associated virus, and GALV infections. J Virol 1993; 67 (11):6737-41.
Tailor CS, Kabat D. Variable regions A and B in the envelope glycoproteins of feline leukemia virus subgroup B and amphotropic murine leukemia virus interact with discrete receptor domains. J Virol 1997; 71 (12):9383-91.
Tailor CS, Willett BJ, Kabat D. A putative cell surface receptor for anemia-inducing feline leukemia virus subgroup C is a member of a transporter superfamily. J Virol 1999; 73 (8):6500-5.
Tailor CS, Nouri A, Zhao Y, Takeuchi Y, Kabat D. A sodium-dependent neutral-amino-acid transporter mediates infections of feline and baboon endogenous retroviruses and simian type D retroviruses. J Virol 1999; 73 (5):4470-4.
Tailor CS, Nouri A, Lee CG, Kozak C, Kabat D. Cloning and characterization of a cell surface receptor for xenotropic and polytropic murine leukemia viruses. Proc Natl Acad Sci U S A 1999; 96 (3):927-32.
Tailor CS, Nouri A, Kabat D. A comprehensive approach to mapping the interacting surfaces of murine amphotropic and feline subgroup B leukemia viruses with their cell surface receptors. J Virol 2000; 74 (1):237-44.
121
Tailor CS, Nouri A, Kabat D. Cellular and species resistance to murine amphotropic, gibbon ape, and feline subgroup C leukemia viruses is strongly influenced by receptor expression levels and by receptor masking mechanisms. J Virol 2000; 74 (20):9797- 801.
Tailor CS, Marin M, Nouri A, Kavanaugh MP, Kabat D. Truncated forms of the dual function human ASCT2 neutral amino acid transporter/retroviral receptor are translationally initiated at multiple alternative CUG and GUG codons. J Biol Chem 2001; 276 (29):27221-30.
Tailor CS, Lavillette D, Marin M, Kabat D. Cell surface receptors for gammaretroviruses. Curr Top Microbiol Immunol 2003; 281:29-106.
Takeuchi Y, Vile RG, Simpson G, et al. Feline leukemia virus subgroup B uses the same cell surface receptor as gibbon ape leukemia virus. J Virol 1992; 66 (2):1219-22.
Temin HM, Mizutani S. RNA-dependent DNA polymerase in virions of Rous sarcoma virus. Nature 1970; 226 (5252):1211-3.
Theilen GH, Gould D, Fowler M, Dungworth DL. C-type virus in tumor tissue of a woolly monkey (Lagothrix spp.) with fibrosarcoma. J Natl Cancer Inst 1971; 47 (4):881-9.
Todaro GJ. RNA tumor virus genes and the transforming genes: genetic transmission, infectious spread, and modes of expression. Natl Cancer Inst Monogr 1978 (48):199- 213.
Trkola A, Dragic T, Arthos J, et al. CD4-dependent, antibody-sensitive interactions between HIV-1 and its co-receptor CCR-5. Nature 1996; 384 (6605):184-7.
Tsatsanis C, Fulton R, Nishigaki K, et al. Genetic determinants of feline leukemia virus- induced lymphoid tumors: patterns of proviral insertion and gene rearrangement. J Virol 1994; 68 (12):8296-303.
Tzavaras T, Testa N, Neil J, Onions D. Isolation and characterisation of a myeloid leukaemia inducing strain of feline leukaemia virus. Haematol Blood Transfus 1989; 32:347-51.
Valsesia-Wittmann S, Drynda A, Deleage G, et al. Modifications in the binding domain of avian retrovirus envelope protein to redirect the host range of retroviral vectors. J Virol 1994; 68 (7):4609-19.
Van der Maaten MJ, Boothe AD, Seger CL. Isolation of a virus from cattle with persistent lymphocytosis. J Natl Cancer Inst 1972; 49 (6):1649-57.
Venables JP. Aberrant and alternative splicing in cancer. Cancer Res 2004; 64 (21):7647-54.
Verrier F, Borman AM, Brand D, Girard M. Role of the HIV type 1 glycoprotein 120 V3 loop in determining coreceptor usage. AIDS Res Hum Retroviruses 1999; 15 (8):731- 43.
122
Verwoerd DW, Payne AL, York DF, Myer MS. Isolation and preliminary characterization of the jaagsiekte retrovirus (JSRV). Onderstepoort J Vet Res 1983; 50 (4):309-16.
Vlachos A, Klein GW, Lipton JM. The Diamond Blackfan Anemia Registry: tool for investigating the epidemiology and biology of Diamond-Blackfan anemia. J Pediatr Hematol Oncol 2001; 23 (6):377-82.
Vlachos A, Federman N, Reyes-Haley C, Abramson J, Lipton JM. Hematopoietic stem cell transplantation for Diamond Blackfan anemia: a report from the Diamond Blackfan Anemia Registry. Bone Marrow Transplant 2001; 27 (4):381-6.
Vogt VM, Eisenman R. Identification of a large polypeptide precursor of avian oncornavirus proteins. Proc Natl Acad Sci U S A 1973; 70 (6):1734-8.
Walker R. Virus associated with epidermal hyperplasia in fish. Natl Cancer Inst Monogr 1969; 31:195-207.
Wang H, Kavanaugh MP, North RA, Kabat D. Cell-surface receptor for ecotropic murine retroviruses is a basic amino-acid transporter. Nature 1991; 352 (6337):729-31.
Wang H, Dechant E, Kavanaugh M, North RA, Kabat D. Effects of ecotropic murine retroviruses on the dual-function cell surface receptor/basic amino acid transporter. J Biol Chem 1992; 267 (33):23617-24.
Wang H, Kavanaugh MP, Kabat D. A critical site in the cell surface receptor for ecotropic murine retroviruses required for amino acid transport but not for viral reception. Virology 1994; 202 (2):1058-60.
Wasserman D. Having one child to save another: a tale of two families. Philos Public Policy Q 2003; 23 (1-2):21-7.
Watermann DO, Tang Y, Zur HA, et al. Splicing factor Tra2-beta1 is specifically induced in breast cancer and regulates alternative splicing of the CD44 gene. Cancer Res 2006; 66 (9):4774-80.
Weiss RA, Tailor CS. Retrovirus receptors. Cell 1995; 82 (4):531-3.
Willig TN, Ball SE, Tchernia G. Current concepts and issues in Diamond-Blackfan anemia. Curr Opin Hematol 1998; 5 (2):109-15.
Willig TN, Niemeyer CM, Leblanc T, et al. Identification of new prognosis factors from the clinical and epidemiologic analysis of a registry of 229 Diamond-Blackfan anemia patients. DBA group of Societe d'Hematologie et d'Immunologie Pediatrique (SHIP), Gesellshaft fur Padiatrische Onkologie und Hamatologie (GPOH), and the European Society for Pediatric Hematology and Immunology (ESPHI). Pediatr Res 1999; 46 (5):553-61.
123
Willig TN, Draptchinskaia N, Dianzani I, et al. Mutations in ribosomal protein S19 gene and diamond blackfan anemia: wide variations in phenotypic expression. Blood 1999; 94 (12):4294-306.
Withers-Ward ES, Kitamura Y, Barnes JP, Coffin JM. Distribution of targets for avian retrovirus DNA integration in vivo. Genes Dev 1994; 8 (12):1473-87.
Wood PN. A functional model for the ribosome. J Theor Biol 1997; 185 (1):97-118.
Wu H, Liu X, Jaenisch R, Lodish HF. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor. Cell 1995; 83 (1):59-67.
Wu L, Gerard NP, Wyatt R, et al. CD4-induced interaction of primary HIV-1 gp120 glycoproteins with the chemokine receptor CCR-5. Nature 1996; 384 (6605):179-83.
Wu X, Li Y, Crise B, Burgess SM. Transcription start regions in the human genome are favored targets for MLV integration. Science 2003; 300 (5626):1749-51.
Yamamoto T. Molecular mechanism of monocyte predominant infiltration in chronic inflammation: mediation by a novel monocyte chemotactic factor, S19 ribosomal protein dimer. Pathol Int 2000; 50 (11):863-71.
Yang YL, Guo L, Xu S, et al. Receptors for polytropic and xenotropic mouse leukaemia viruses encoded by a single gene at Rmc1. Nat Genet 1999; 21 (2):216-9.
Yieh L, Kassavetis G, Geiduschek EP, Sandmeyer SB. The Brf and TATA-binding protein subunits of the RNA polymerase III transcription factor IIIB mediate position-specific integration of the gypsy-like element, Ty3. J Biol Chem 2000; 275 (38):29800-7.
Zavorotinskaya T, Albritton LM. Suppression of a fusion defect by second site mutations in the ecotropic murine leukemia virus surface protein. J Virol 1999; 73 (6):5034-42.
Zhang XH, Chasin LA. Computational definition of sequence motifs governing constitutive exon splicing. Genes Dev 2004; 18 (11):1241-50.
Zhang XH, Kangsamaksin T, Chao MS, Banerjee JK, Chasin LA. Exon inclusion is dependent on predictable exonic splicing enhancers. Mol Cell Biol 2005; 25 (16):7323-32.
Zhu Y, Zou S, Wright DA, Voytas DF. Tagging chromatin with retrotransposons: target specificity of the Saccharomyces Ty5 retrotransposon changes with the chromosomal localization of Sir3p and Sir4p. Genes Dev 1999; 13 (20):2738-49.
ADDENDUM. TARGETING RETROVIRAL INTEGRATION
Author contributions/acknowledgements: All work in this addendum was performed by MAR. We would like to thank Dr. Nouria Hernandez for donation of the hBrf1 and hBRF2 constructs. We would also like to thank Dr. Frederic Bushman for his protocol and helpful suggestions.
124 125
A.1 Abstract
Retroviruses have been used in human gene therapy as a corrective mechanism to permanently introduce healthy genes into diseased cells. A major dilemma in use of this technique is the randomness with which the desired gene is integrated into the host cell genome. Some human retroviral gene therapy trials have led to insertional mutagenesis of the corrected cells causing the clinical trial to be halted. There are 513 tRNA genes encoding the 64 tRNAs needed for protein synthesis indicating a great deal of redundancy in the human genome. tRNA genes are transcribed by the RNA PolIII machinery with the aid of a multitude of transcription factors. hBrf1 is a component of TFIIIB that binds directly to the promoter regions of tRNA genes. By creating a hybrid MLV integrase with hBrf1 at its C-terminus I attempted to design a method of targeting retroviral integration to tRNA genes. The pseudotyped hybrid integrase retrovirus was used to infect TE671 cells. Integration site analysis was attempted from clonal infected cells. An increased number of infected neomycin-resistant clones with integration sites adjacent to the redundant genes as compared to the wild type integrase control could be indicative of more specific integration by the hybrid integrase. Integration site analysis was attempted to determine targeting to polIII genes.
A.2 Introduction
One major caveat of retroviral integration is its lack of specificity for integration sites in the host genome. Analyses of integration sites by a variety of viruses in human cells has shown that while most regions of cellular DNA are accessible, certain regions, termed ‘hot spots’, are used at a much higher rate (Craigie, 1992, Withers-Ward, et al., 1994, Schroder, et al., 2002). Thus, retroviral integration is non-specific, but not completely random. Non-specific integration can lead to insertional mutagenesis that can disrupt normal cell functions; integration may occur within an essential host gene inactivating it. Conversely, the virus may integrate its genome upstream of a normally inactive proto- oncogene that when under control of a strong viral promoter can lead to over expression of that gene and subsequent cell transformation. This unfortunate scenario was observed in
126 clinical trials for introduction of a functional IL2Rγ chain gene into children with X-linked severe combined immunodeficiency (X-SCID) (Hacein-Bey-Abina et al., 2002, Hacein-Bey- Abina, et al., 2003, Hacein-Bey-Abina et al., 2003, Marshall, 2003, Check, 2005) as the virus integrated the healthy gene upstream of an oncogene. The strong viral promoter caused overexpression of LMO2 leading to leukemia. Although the deficiency in functional expression of this gene was restored and disease manifestation blocked (Hacein-Bey-Abina et al., 2002, Hacein-Bey-Abina et al., 2002), the occurrence of leukemia in four of the ten patients treated emphasized that insertional mutagenesis can be an unexpectedly frequent adverse effect of retrorival gene transfer. To understand retroviral integration site selection, a number of scientists have embarked upon analyzing the specificities (or lack thereof) in site selection for retroelements including retroviruses and retrotransposons. Replication of retroviruses and retrotransposons is dependent on the selection of a favourable site for integration of their genomic DNA. These sites do differ among the elements with yeast retrotransposon element Ty3 preferring to integrate at transcriptional start sites of genes transcribed by Pol III (Kirchner et al., 1995) while Ty5 intregrates within heterochromatin at telomeres (Zhu et al., 1999). The HIV-1 lentivirus and MLV gammaretrovirus prefer to integrate adjacent to genes that are actively transcribed (Schroder et al.. 2002, Wu et al. 2003, respectively), although MLV prefers the promoter regions of these genes. With so much variation in integration site preference, it was surprising to see a commonality in the targeting mechanism for these elements especially as integrases have not been shown to contain any DNA binding activity. Integration complexes have been shown to be tethered to regions of the genome bound by specific proteins. Through interaction with proteins that are specific for their genomic DNA binding, retroelements have been able to somewhat target their integration to their favourite sites. For example, Ty3 elements only integrate at the 5’ end of PolIII transcripts through its interaction with the TFIIIB component of the PolIII transcription apparatus (Kirchner et al., 1995). These integration sites have been shown to be within 1-2bp of these genes, indiciating a very specific targeting for tRNA genes. The integrase enzyme encoded by Ty5 elements binds to the heterochromatin protein silent information regulator 4 (Sir4p) that is located at the heterochromatin sites with which it
127 integrates (Zhu et al., 1999). The integrase was further characterized to demonstrate the Sir4p binding domain to be in the carboxy-terminal half of the protein. This research has set the stage for creating a retroviral integrase protein that can target specific sites in the genome through its protein-protein interactions with specific DNA binding proteins. A chimeric Ty3/Moloney murine leukemia virus integrase was constructed by removing the C-terminal end of the Mo-MLV integrase and replacing it with the full Ty3 integrase protein producing a fully functional integrase protein incorporated within the virion (Dildine et al., 1998). This chimeric integrase was expected to target tRNA and other polIII genes for retroviral integration as with Ty3 retrotransposons. Unfortunately, none of the three integration sites sequenced from the infected cells were adjacent to tRNA genes (Dildine et al., 1998). More integration sites may have been needed to be verified to ensure no increase in targeting to tRNA genes. The same group later showed that Ty3 integrase associates with transcription factor IIIB (TFIIIB) factors and TATA-binding protein (TBP) to mediate Ty3 position specific integration (Yieh et al., 2000). Another attempt to target integration using a hybrid MLV integrase involved fusing the integrase and the zinc finger domains of the Sp1 transcription factor (Peng, et al., 2002). This hybrid integrase was able to target integration to Sp1 sites within the mouse genome 13% (5 out of 39 clones). The authors noted their success to be due to the small size of the added zinc finger domains (60 aa) in comparison to the previous study containing the entire Ty3 integrase gene. The downfall for this project was the limited number of Sp1 sites within the genome which the authors argue decreased the efficiency of targeted retroviral integration (Peng et al., 2002). We propose to target retroviral integration to sites within the human genome that are present in higher copy number such as tRNA genes. There are 513 tRNA genes encoding the 64 tRNAs needed for protein synthesis (Lowe, T. 2006). This creates a great deal of redundancy between the tRNA genes. A single integration site within the genome may not drastically affect the presence of that tRNA molecule and subsequent protein synthesis. In addition, there would be multiple regions for the integrase to select for integration unlike the Sp1 sites, thus enhancing efficiency of infection. tRNA genes are transcribed by the RNA pol III machinery that also transcribes snRNA and 5S rRNA genes. The RNA pol III has no specific DNA binding activity to the promoter of tRNA genes, but rather uses its associated
128 transcription factors to target it to the correct genes. hBrf1 (~74 kDa) and hBrf2 (~46 kDa) are small components of TFIIIB that bind directly to the promoter sequence of tRNA and snRNA genes, or snRNA genes alone, respectively, and recruit the rest of the machinery including the TATA-box binding protein (TBP) (Figure A.1) (Reviewed in Schramm and Hernandez, 2002). Both of these transcription factors contain an N-terminal zinc-finger DNA binding domain similar to Sp1. Brf2 is a homolog of Brf1 and binds to the promoter region of snRNA genes, while Brf1 binds to promoters of both snRNA and tRNA genes. We propose that by creating hybrid integrase proteins with these transcription factors, we can target retroviral integration to one of the many redundant non-coding RNA genes within the human genome.
A.3 Methods
A.3.1 Hybrid integrase construction
We have created murine leukemia virus (MLV) integrase hybrids with either hBrf1 (IN-hB1) or hBrf2 (IN-hB2) tagged onto the C-terminus of the integrase. The hybrid integrases were amplified by PCR from constructs donated by Dr. Nouria Hernadez. The amplicons were diegested with HpaI and BsbI for ligation onto the C-terminal end of the MLV IN wihin the CMV vector that contains the Moloney MLV (MMLV) gag and pol genes. The CMV plasmid was co-transfected into the virus packaging cell line HEK 293T along with the FeLV-B pFBSalf env, and pFB-Neo plasmid vector for neomycin (G418) antibiotic selection of infected cells, retroviral gene delivery, and expression to produce high-titre viral stocks or LacZ or determining virus titre.
A.3.2 hybrid integrase function
β-Galactosidase encoding non-replicative hybrid IN-TF psuedotyped viruses were generated by transfection of HEK293T cells with pFBsalf FeLV-B env, CMV gag/pol, and LacZ. Virus supernatant harvested 48 hours post-transfecton, filtered using a 0.45μm filter and then subsequently used for infection studies. Target TE671 cells were seeded in a 24-
129
TFIIIB
tRNA gene
TFIIIB
snRNA gene
TBP Brf1
Bdp1 Brf2
Figure A.1. Schematic diagram showing the positions of the various TFIIIC and TFIIIB subunits on tRNA and 5S RNa genes. Brf1 makes direct contact with the TATA box binding protein (TBP) and the promoter sequence of these genes. Brf2 is a homolog of Brf1 displaying strong identity in the N-terminal region, but not the C-terminus. Brf2 binds in a similar manner as Brf1 to the promoter sequence of snRNA genes.
130 well plate at 3 × 104 cells/well one day prior to the infection study. The following day, target TE671 cells were incubated with 1 ml of serially diluted lacZ pseudotype virus supernatant for 4 h in the presence of polybrene (8μg/ml). The virus supernatant was then replaced with fresh growth medium, and cells were allowed to incubate for a further 2 days before X-gal (5-bromo-4-chloro-3-indoyl-β-d-galactopyranoside) (Sigma-Aldrich, Canada) staining. LacZ pseudotype titers were determined by counting the number of blue colony-forming units (CFU's), and titers were expressed as the number of CFUs obtained per milliliter of undiluted virus supernatant. Virus titer results reported are an average of three independent experiments.
A.3.3 Clonal cell line production
To obtain retrovirally infected clones for integration site analysis, I infected the human medulloblastoma clonal line, TE671 with harvested wt MLV IN or hybrid IN-hBrf1 or IN-hBrf2 psuedotyped virus. Target Te671 cells were seeded in a 24-well plate at 3 × 104 cells/well one day prior to the infection study. The following day, target cells were incubated with 1 ml of psuedtoyped virus supernatant for 4 h in the presence of polybrene (8 μg/ml). The virus supernatant was then replaced with fresh growth medium, and cells were allowed to incubate for a further 2 days before transfer to 100mM dishes for G418 selection. Neomycin resistance was confirmed though 100μM G418 selection of cells for a period of over one week until visibility of individual colonies. 24 individual colonies were transferred to the wells of a fresh 24 well plate for approximately one week before transfer to a T25 flask, and then final subculture to a T75 flask for genomic DNA preparation. This procedure was repeated twice for each construct to produce over 25 individual retrovirally infected clones for integration site analysis. Genomic DNA was recovered from the clonal cell lines using a genomic DNA preparation kit (Qiagen) following the manufacturer’s protocol.
131
A.3.4 Integration site analysis
Integration site analysis on the clonal genomic DNA samples was attempted by three different methods depicted schematically in Figure A.2.
A.3.4.1 Inverse PCR Inverse PCR is a method for the rapid in vitro amplification of DNA sequences that flank a region of known sequence designed by Ochman and colleagues in 1988 (Ochman et al., 1988). Approximately 1μg of genomic DNA was digested with XbaI (Fermentas) at 37oC for 4 hours followed by inactivation of the enzyme at 65oC for 10 minutes. 10 ng or 100 ng of digested DNA was ligated overnight in final volumes of 10 or 100μl with ligase enzyme (NEB). Enzyme was removed from the ligated DNA through DNA cleanup kit (Qiagen), and elution in 50μl water. 10μl of the eluted DNA was used in the following PCR reaction conditions 94oC for 2’ initial denaturation followed by 35 cycles of denaturation at 94oC 20’’, annealing at 58oC 1’, extension at 68oC 5’, followed by a final extension at 68oC for 10’. Nested PCR was performed with the same conditions. Amplified PCR fragments were separated with 1% agrose gel electrophoresis. Amplicons of interest were gel extracted (Qiagen), cloned into pCR2.1 TOPO cloning vector (Invitrogen) and sequenced (ACGT Corp.). Expand Long PCR Taq polymerase enzyme and buffers (Roche) werse used according to the manufacturers protocol. A schematic of the procedure is represented in Figure A.2.A.
A.3.4.2 Tailor’d LAM-PCR Due to unavailability of the protocol for LAM-PCR (Schmidt et al., 2002), a few adjustments were made (F. Bushman, personal communication). Briefly, linear PCR was performed as per the protocol using a biotinylated LTR primer. Amplified ssDNA products were purified with streptavidin coated magnetic beads (Miltenyi Biotec). Random primers were annealed to the ssDNA, and ligated overnight with T4 DNA ligase (NEB). dsDNA was digested with PstI (Fermentas) and a poly adenine added using Terminal deoxynucleotidyl transferase in the presence of dATP. Two rounds of nested PCR was carried out using OligodT primer and nested LTR II and III primers. PCR reactions were
132 separated with 1% agarose gel electrophoresis. A schematic of the procedure is represented in Figure A.2.B.
A.3.4.3 Blockerette-ligated capture T7-amplified RT-PCR Integration site analysis was carried out following the methods of Lenvik et al.,(Lenvik et al., 2002) with the exception of the change in sequence of the biotiylated primer to repsresent the neomycin gene (5’-ACCCGTGATATTGCTGAAGAGCTTGG-3’) (Integrated DNA technologies). Briefly, 1μg of genomic DNA was digested with AcsI (NEB), incubated for 2 hous at 50oC. Linker was ligated onto the digested DNA with an overnight incubation at 16oC with ligase (NEB). Linear PCR with the biotinylated primer was performed to amplify the integration site sequence. Amplified DNA was captured with magnetic Streptavidin coated beads (Roche). Eluted DNA was subjected to in vitro transcription with T7 megascript, reverse transcription of the RNA to cDNA with Thermoscript (Invitrogen), and two rounds of nested PCR. Amplified products were visualized with ethidium bromide staining of 1% agarose gel electrophoresis. Selected amplicons were gel extracted (Qiagen), cloned into pCR2.1 TOPO sequencing vector (Invitrogen), and sequenced (ACGT Corp.). A schematic of the procedure is represented in Figure A.2.C.
133
5’ Genomic DNA 5’LTR R 3’LTR 1 Neo gene 3’ Genomic DNA Digest genomic DNA with XbaI
x x 2 x x x
Circularize digested DNA through ligation
3
First and second round PCR with Neo gene specific nested primers
4
Amplified products contain genomic DNA sequence adjacent to the integration site
Figure A.2. Outline of methods used for integration site analysis. A. Inverse PCR: (1) Restriction enzyme digestion of genomic DNA using XbaI. (2) Circularization of DNA fragments through ligation. (3) First and second round PCR with complementary outward primers to neomycin resistance gene.
134
R 5’ Genomic DNA 5’ Neo gene 3’ 3’ Genomic DNA
1. Linear PCR Streptavidin beads
2. Random Priming
3. DNA digest PstI
4. PolyA tail addition AAA
5. Nested TTTT PCR LTR II TTTT
LTR III
B. Tailor’d LAM-PCR: (1) Linear PCR with a long terminal repeat (LTR)–specific biotinylated primer was performed by repeated primer extension. Subsequently, the amplified fragments of target DNA were enriched by magnetic tag selection of extension primers. (2) A second DNA strand of each enriched target sequence was synthesized by random hexanucleotide priming. (3) Resulting double-stranded DNA was specifically digested with the restriction enzyme PstI, which cuts within genomic DNA approximately every 256 bp. The length of each fragment is thus dependent on the distance of the vector insertion site from the next Sse9I recognition sequence. (4) A poly(A) tail was added to the end of the PstI fragments using terminal deoxynucleotidyl transferase (TdT) + dATP. (5) Nested exponential PCR amplifications were then performed with oligo dT and LTR-specific reverse primers (LTR II followed by LTR III).
135
R 5’ Genomic DNA 5’ Neo gene 3’ 3’ Genomic DNA • Restriction enzyme AcsI AcsI digestion
• Blockerette linker ligation linke • PCR with biotinlated T7 promoter-vector T7VSP specific an linker specific primer • Capture on Streptavidin magentic beads
• T7 Amplification
• Reverse transcription
• Nested PCR RNA
C. Blockerette-ligated capture T7-amplified RT-PCR: (1) Restriction enzyme digestion of genomic DNA using AcsI. (2) Cohesive-end ligation of blockerette linkers to the genomic DNA fragments. (3) PCR with a biotin-coupled T7 promoter sequence-vector-specific primer (T7VSP) and a linker-specific primer (LSP1). (4) Enrichment of the biotin-labeled T7VSP amplicons by a streptavidin magnetic capture system. (5) In vitro transcription using the T7 promoter that was incorporated into the viral-specific primer followed by a DNase. (6) Reverse transcription using the linker-specific primer (LSP1). (7) The final step of nested RT-PCR with vector (VSP2) and linker specific primers (LSP2). Adapted from Lenvik et al., 2002.
136
A.4 Results
A.4.1 Hybrid integrase viruses are functional
The efficiency of infection by the wild type and hybrid integrase containing retroviruses was determined (Table A.1). There is a significant decrease in the efficiency of infection for the hybrid integrase MLV’s as compared to the wild type MLV. MLV IN-hB2 exhibited over a hundred fold decrease in infection from six thousand down to twenty infectious virus particles per ml of virus when compared to the wildtype IN psudotyped virus. The MLV IN-hB1 was also not as infectious as the wildtype with approximately 240 infectious particles per ml of virus supernatant. Both hybrid constructs yielded functional virus capable of delivering the LacZ gene into the target Te671 cells. Furthermore, the infections with the pseudotyped viruses yielded sufficient colonies for integration site analysis.
Table A.1 Infectivity of hybrid integrase virus on Te671 cells.
IN-HYBRID VIRUS TITRE (CFU/ml) MLV (wt) 6 x 103 IN-hB1 240 IN-hB2 20
A.4.2 Integration site analysis
Over 20 single clones for each construct (MLV, MLV-hB1, MLV-hB2) have been obtained. Each clone represents a single colony of neomycin-resistant cells containing one integration site within the genomic DNA. Genomic DNA was extracted from each clone to assess for the site of integration by three different methods of integration site analysis. Inverse PCR provided amplicons of differing sizes for each clone (Figure A.3), but the amplicons were too small to represent the correct neomycin gene+LTR+genomic DNA
137
Figure A.3. Inverse PCR did not yield integration site amplification Representative gel showing the second round of PCR amplification using nested primers against the neomycin resistance gene. Individual clones are labeled. –ve represents no DNA control. Circulization ligation was performed in 10 or 100 μl reactions. Arrows represent amplicons TOPO cloned and sent for sequencing.
138
Figure A.4. Tailor’d LAM-PCR requires more tailoring. Representative agarose gel of second round amplication using nested primer against the LTR and oligodT primer. Individual clones are labeled. –ve represents no DNA control. Arrow represents amplicon that appeared in all clones that was TOPO cloned and sent for sequencing.
139
sequences. Upon sequence analysis of the amplified DNA, it was confirmed they did not represent integration sites due their lack of LTR sequence. Furthermore, attempts at inducing circularization of the DNA fragments as opposed to concatamers by increasing the reaction volume, while successful at creating different amplicons, did not add sufficient specificity for the integration site amplification. The efforts to replicate the LAM-PCR methods were unsuccessful as no specific PCR products were amplified (Figure A.4). A PCR product of the same small size can be clearly seen in all genomic DNA samples indicating lack of specificity for the adjusted protocol. Furthermore, attempts at further amplification with additional nested LTR primers did not yield any new products from the pooled PCR. Also sequence analysis of these bands did not display the critical LTR sequence required to confirm the correct amplification of genomic integration sites. The blockerette-ligated capture T7-amplified RT-PCR method devised by Lenvik and colleagues proved somewhat successful as one of the amplified products corresponded to what appears to be a real integration site (Figure A.5). PCR amplification of clone IN-hB2 #9 contained both the LTR and linker sequence and sufficient genomic DNA sequence for integration site verification upon TOPO cloning and sequencing. While this clone was not within 1-2 bp of a polIII transcribed gene, appearing at position 9p22.3 of chromosome number 9 adjacent to a pseudogene to Pescadillo homolog 1, it does confirm the ability of the hybrid integrase to efficiently infect TE671 cells. Other amplified products of this PCR procedure did not represent true integration sites as determined through sequence analysis and lack of LTR or linker sequence.
A.5 Discussion
We attempted to design a retrovirus capable of targeting specific sites in the genome for integration of its genomic DNA. Hybrid integrases were constructed through addition of transcription factors, hBRF1, and hBrf2, to the MLV integrase. These factors bind specifically to the promoter regions of tRNA and snRNA genes, and were used in the targeting of the retroviral IN to these genes during integration. Surprisingly, the smaller the size of the fusion protein led to a greater decrease in virus infection. One might expect that
140
Figure A.5. Blockerette-ligated capture T7-amplified RT-PCR may identify integration sites. Representative agarose gel of second round amplication using nested primers against the LTR and linker. Individual clones are labeled. –ve represents no DNA control. Arrows represents amplicons that were TOPO cloned and sent for sequencing. Arrow denotes amplicon corresponding to true integration site.
141 the addition of a smaller protein to the integrase would lead to less interference with integrase function, and thus, infection efficiency similar to that of wild type integrase. It is possible that the addition of these proteins to the C-terminal end of MLV-IN rendered the IN less functional. This would produce fewer LacZ integration events following virus entry, and subsequent decrease in infected cells. Western blot analysis using an antibody against the MLV IN protein could provide useful information about the expression of the hybrid IN constructs. A decrease in expression of the hybrid IN could also account for a decreased infection rate of the pseudotyped viruses. Previous experiments have shown integration efficiency to be unaffected by addition of a large protein, Ty3 IN (Dildine et al., 1998) or the smaller Sp1 transcription factor (Peng et al., 2002). This suggests that MLV IN should withstand the addition of small proteins without major adverse effects on its function. A possible explanation for the decrease in virus infection may lie in the increase targeting of the IN for tRNA and/or snRNA genes. Perhaps the MLV IN-hBrf2 fusion protein has created a more specific integrase that is actively searching for the snRNA genes for integration of the viral genome. This active searching for a binding site within the genome may lead to decreased ability to integrate and subsequent 100 fold decrease in infection rates in comparison to the wild type MLV IN. Similarily, the multiple available tRNA genes available for integration of the IN-hB1 construct might render a slightly higher infection rate for IN-hB1 than the IN-hB2 virus. It was discovered that MLV preferentially integrates within the promoters of active genes (Wu et al., 2003), thus its possible that the hybrid IN is no longer integrating at the first available actively transcribed gene, but instead searching for its target. There are less available snRNA genes than tRNA genes which may be the cause for the change in efficiency of infection between the two hybrid constructs. The MLV IN- hBrf1 construct exhibited ten fold more infection which may be due to the presence of more tRNA target genes for integration. Thus, the decreasing efficiency in the hybrid IN-TF constructs may be due to the increasing specificity of targeted integration. To determine if integration occurred at the targeted sites, clonal integration sites within the genome must be determined. Several attempts were made to determine if the sites of integration by the hybrid IN- TF pseudotyped virus were adjacent to the targeted tRNA or snRNA genes. Determining retroviral integration sites has been attempted by many different laboratories using a
142 multitude of methods. Each method must provide to main features of retroviral integration site analysis, namely specificity and efficiency. When determining the site of integration for a small number of clones, efficiency is not a critical as many adaptations to the protocol may be evaluated to produce a product for sequencing analysis. Indeed, the most time consuming aspect of some of the methods tested was in the mere production of a PCR product for analysis. Specificity is also important in integration site analysis. PCR products produced from the several methods tested did not represent integration site analysis indicating a lack of specificity for the correct amplicon in the procedure. LAM-PCR was developed as an efficient, sensitive and specific technique to identify integration sites in retroviral infected cells (Schmidt et al., 2002). The technique has not been used by many labs due to its complexity and unavailability of the complete protocol. Fortunately for those studying retroviral integration, von Kalle and colleagues have recently published the complete protocol in Nature Methods due to many requests and successes with this method (Schmidt et al, 2007). It would be advantageous to test the complete protocol on all clonal genomic DNA samples obtained through this experiment as improved success should be observed in identifying the integration sites. The blockerette-ligated capture T7-amplified RT-PCR method of determining integration sites was efficient in amplifying the correct amplicon corresponding to an integration site, and may be used in determining more integration sites (Lenvik et al., 2002). Due to its costly, and lengthy protocol and the number of integration sites to be determined, it might be more cost effective to the LAM- PCR method. Many amplified PCR products from all attempted methods produced sequences that did not correspond to integration sites. These clones did not contain the LTR or linker sequences needed to verify the presence of the viral genome along with the genomic DNA expected to be present upon retroviral integration. These clones may represent mispriming to random regions of the chromosome, or pseudo-integration sites. The LTR sequence of many retroelements is present in multiple copies within the human genome. The use of LTR primers is required for these methods, but may represent a drawback in identifying true integration sites due to priming of LTR primers to similar LTR sequences within the genome. True integration clones containing the 3’ LTR, genomic DNA and linker sequence can be confirmed for their authenticity by amplifying upstream of the suggested genomic
143
DNA sequence with primers against the 5’ integration site sequence and the 5’LTR. These products should amplify the adjacent integration site host genomic DNA and would verify the insertion of the retroviral genome at that site.
Future goals include the use of other transcription factors that are part of the pol III machinery to tether the MLV integrase to particular sites in the human genome such as components of the RNA pol III transcription factor TFIIIC. Human TFIIIC63 binds specifically to the tRNA A-box and associates with TFIIIC102, hBrf1, and hTBP (Schramm and Hernandez, 2002). These constructs can also be made, and tested for their ability to target retroviral integration. Yeast retrotransposon TY3 contains an integrase protein that inserts the TY3 genome 1-2bp upstream of tRNA genes without altering the tRNA expression (Chalker and Sandmeyer, 1992). The TY3 integrase uses the yeast Brf transcription factors to guide it to the promoter regions of tRNA genes in yeast cells (Yieh, et al., 2000). In addition, yeast TY3 integrase has been shown to target the human tRNA Lys gene when it is co-transfected into yeast cells (Dildine and Sandmeyer, 1997). We propose to use an MLV-TY3 integrase hybrid, containing only the C-terminal region of Ty3 responsible for interacting with TFIIIB, to target tRNA genes in human cells in attempt to increase the interaction between the TY3 integrase and human Brf1 protein. This fusion IN- TF might provide a better targeting mechanism than the entire Ty3 IN fusion proteins. Studies of the C-terminal domain of retroviral IN will be extremely useful in identifying what regions are involved in binding to TFs and may be compatible with replacements or insertions of small DNA-binding domains. Much work remains to be explored regarding the mechanisms, implications, and applications of targeting retroviral integration. Integration by design remains an elusive, yet important concept for study.