Genetic Mobility and Instability of Retroviral Vector in Vector Producer Cells for Gene Therapy Won-Bin Young Iowa State University

Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

1999 Genetic mobility and instability of retroviral vector in vector producer cells for gene therapy Won-Bin Young Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Genetics Commons, Microbiology Commons, and the Molecular Biology Commons

Recommended Citation Young, Won-Bin, "Genetic mobility and instability of retroviral vector in vector producer cells for gene therapy" (1999). Retrospective Theses and Dissertations. 12196. https://lib.dr.iastate.edu/rtd/12196

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Leaming 300 North Zeeb Road, Ann ArtDor, Ml 48106-1346 USA 800-521-0600

Genetic mobility and instability of retroviral vector in vector producer cells

for gene therapy

Won-BIn Young

A dissertation submitted to the graduate faculty

In partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY

Co-majors: Molecular, Cellular and Developmental Biology: Nutritional Physiology

Major Professors: Gary L. Lindberg and Charles J. Link, Jr.

Iowa State University

Ames, Iowa

1999

This microform edition is protected against unauthorized copying under Title 17, United States Code.

tnvn 300 North Zeeb Road Ann Arbor, MI 48103 ii

Graduate College

Iowa State University

This is to certify that the Doctoral dissertation of

Won-Bin Young has met the dissertation requirements of Iowa State University

Signature was redacted for privacy.

CD Major Professor

Signature was redacted for privacy.

o Majc^ Professor

Signature was redacted for privacy.

Signature was redacted for privacy. For the Graduate College iii

TABLE OF CONTENTS

LIST OF FIGURES v

LIST OF TABLES vil

ABSTRACT vlii

CHAPTER 1. GENERAL INTRODUCTION 1

Introduction 1

Dissertation Organization 11

References 12

CHAPTER 2. RESTRICTION MAPPING OF RETROVIRAL VECTOR MUTATION BY ANALYZING EPISOMAL DNA 19

Introduction 20

Materials and Methods 21

Results 22

Discussion 24

References 24

CHAPTER 3. DNA METHYLATION INCREASES GENETIC INSTABILITY OF RETROVIRAL VECTOR PRODUCER CELLS 33

Abstract 34

Introduction 34

Materials and Methods 37

Results 40

Discussion 48

Acknowledgements 53

References 54 iv

CHAPTER 4. CHIMERIC RETROVIRAL HELPER VIRUS AND PICORNAVIRUS IRES SEQUENCE TO ELIMINATE DNA METHYLATION FOR IMPROVED PACKAGING CELLS 80

Abstract 81

Introduction 82

Materials and Methods 84

Results 87

Discussion 92

Acknowledgements 94

References 94

CHAPTER 5. HIGH FREQUENCY RETROTRANSPOSITION OF RETROVIRAL VECTOR IN CULTURED VECTOR PRODUCER CELLS 114

Abstract 115

Introduction 116

Materials and Methods 118

Results 121

Discussion 126

Acknowledgements 128

References 129

CHAPTER 6. GENERAL CONCLUSIONS 142

References 148

APPENDIX A. HUMAN RETROVIRAL PACKAGING CELLS BASED ON A CHIMERIC HELPER VIRUS FOR HIGH VECTOR PRODUCTION 151

ACKNOWLEDGEMENTS 154 V

LIST OF FIGURES

CHAPTER 1

Figure 1. Retrovlrus-mediated gene transfer 2

Figure 2. Reverse transcription and recombination 4

Figure 3. Endogenous reverse transcriptase activity and retroelements 6

Figure 4. Env-receptor interference and superinfection of vectors on vector producer cells 8

Figure 5. Intracellular retrotransposition of an Env-defective Moloney murine leukemia virus 9

CHAPTER 2

Figure 1. Detection of second vector from LTKOSN.2 vector production 28

Figure 2. Location of episomal DNA in vector producer cells 30

Figure 3. Characterization of HSVtk deleted LTKOSN (ALTKOSN) by episomal DNA restriction mapping 32

CHAPTER 3

Figure 1. Increased copy number of LTKOSN and ALTKOSN is observed in LTKOSN.2 subclones 64

Figure 2. Genetic instability of integrated LTKOSN in VPC genome 66

Figure 3. Differential gene expressions of vectors and helper virus in LTKOSN.2 VPC subclones 68

Figure 4. Superinfection susceptibility of LTKOSN.2 VPC subclones 70

Figure 5. DNA methylation statuses of helper virus and vectors 72

Figure 6. Treatment with 5-aza-C partly restores the vector production ability in subclones 73

Figure 7. Treatment with 5-aza-C reverses the DNA methylation of helper virus 75

Figure 8. Increased DNA methylation in helper virus of subclone #4 over time 77 vi

Figure 9. Transcription activity of fielper virus and retroviral vectors in LTK0SN.2 VPC subclone #4 79

CHAPTER4

Figure 1. Construction and cap-independent translation mechanism of chimeric pAM3-iRES-Zeo helper virus 103

Figure 2. Gene expression of pAM3-IRES-Zeo and vectors in LTK0SN.2 VPC subclones 105

Figure 3. DNA methylation of helper virus 5' LTR over time with and without Zeocin™ selection 107

Figure 4. The effectiveness of Zeocin selection on helper virus gene expression in AM IZ cells over time 109

Figure 5. Gene expression of pAM3-IRES-Zeo helper virus and LEIN vector in AMIZ cells transfected with LEIN vector 111

Figure 6. Drug selection eliminates DNA methylation of helper virus and vector from VPC population 113

CHAPTER 5

Figure 1. Retrotransposition assay with pELNIH vector 137

Figure 2. Retrotransposed ELNIH vector integrated into different chromosomal locations 139

Figure 3. Scheme of serial deletion constructs of MoMLV 141

CHAPTER 6

Figure 1. Scheme of cascade reactions initiated by increased DNA methylation of helper virus 5' LTR 144

Figure 2. Alternative pathway of retroviral replication 147

APPENDIX A

Figure 1. Variable sensitivity to human serum of vectors produced from different packaging cells 153 VII

LIST OF TABLES CHAPTER 3

Table 1. Titer of LTKOSN.2 sub-clones treated with 5-aza-C 62

Table 2. Decreased titer of LTKOSN.2 VPC subclone #4 in continuous cell culture 62

CHAPTER 4.

Table 1. Titer of LTKOSN.2 VPC subclones before and after transfection of pAM3-IRES-Zeo helper virus 101

Table 2. Increased resistance of superinfection by Zeocin™ selection 101

Table 3. Titer of AMIZ cells transfected with LEIN vector with continuous drug selection 101

CHAPTER 5.

Table 1. Retrotransposition frequencies of pELNIH vector in cultured VPC 135

Table 2. Retrotransposition frequencies of human L1 retroelement (pJMIOI) in VPC 135 viii

ABSTRACT

A primary biosafety issue of retroviral vector-mediated gene therapy is the genetic

instability of retroviral vectors. Reverse transcription of vector RNA genome is initiated by

viral reverse transcriptase (RT) in a virion particle after infection of a target celi. During

reverse transcription, abnormal template switches between vector and occasionally co-

packaged helper virus in a virion particle can therefore enable helper virus to regain

replication elements from the vector and revert to replication-competent retrovirus (RCR).

This research was undertaken to study the origins of RT enzyme activities and test the

hypothesis that RT enzyme activities are contributed by both exogenous RT, which is

imported by the re-infection of vector virions, and endogenous RT, which is provided by

Intracellular vector virion particles. Superinfection of vector in vector producer cells (VPC)

can increase the number of integrated vectors in VPC, and is mainly caused by decreased

Env-receptor interference, a consequence of helper virus gene inactivation by host DNA

methylation at the 5' LTR promoter region. Suppression of helper virus gene expression by

DNA methylation also reduces vector production. A chimeric retroviral helper virus

combined with picomavirus IRES (internal ribosome entry site) sequence and a selection

marker was therefore constructed to eliminate DNA methylated helper virus from cell

populations to maintain active gene expression and enhance Env-receptor interference.

Packaging cells established by this chimeric helper virus exhibit high titer production without

detectable superinfection. In addition to exogenous RT activities imported by superinfection,

significant endogenous RT activities (0.2% to 7.8% of exogenous RT activity), which

originates from viral precursor protein PrlSO^®^"'*", were also demonstrated by intracellular retrotransposition of a retroviral vector in VPC. Retrotransposed vectors can integrate into different chromosomal locations to increase the plasticity of the VPC genome as observed ix from superinfected vectors. The outcomes from this study provide important Insights for further strategies designed to prevent RCR formation by inactivation of RT activity.

Demonstration of Intracellular retrotransposition of retroviral vectors in VPC may reveal an alternative replication pathway of retrovirus. 1

CHAPTER 1. GENERAL INTRODUCTION

Introduction

The use of retroviral vector in gene therapy. Moloney murine leukemia viais-

(MoMLV) based retroviral vectors are being used in 71% of the protocols approved by the

U. S. Recombinant DNA Advisory Committee (23). Permanent modification of a targeted

cell's genotype by the integration of retroviral vector genetic information (28) and the

consistent production of retroviral vectors from permanent packaging cells without helper

virus (4, 25) are the most significant advantages of retroviral vector systems over other viral

vectors. The stable integration of transferred genes is an attractive feature that may permit

treatment of hereditary and chronic disorders, although it may potentially result in insertional

mutagenesis and oncogenesis (10, 35, 38). The reproducibility of therapeutic retroviral

vector production from a stable VPC is very important for safety and quality assurance

when conducting clinical trials on human subjects (41).

Overview of retroviral vector packaging systems. MoMLV is a RNA retrovirus

that packages two copies of retroviral genomic transcripts associated with cellular tRNA

primers and essential viral replication components, such as protease, RT and integrase

(IN), into its virion particles (Figure 1A). The packaging signal, T, and the redundant

sequences termed long terminal repeats (LTR), which flank both the 5' and 3' ends of

genomic transcripts, are essential for the packaging and reverse transcription of viral RNA

transcripts during replication (3). Currently applied retroviral vector systems contain two

basic components (Figure 1B): the retroviral vector itself, which contains ^ signal and LTR

but does not encode any viral proteins, and the helper virus, which provides gag, pol and env gene products in trans for retroviral vector assembly and transfer (28). The T signal 2

(A) RNA Genome

Nucleocapsid Integrase Matnx Reverse Transcriptase Cellular Membrane

(B) Wild Type Virus

LTR gag pol LTR

Helper Virus ~l LTR gag pol AAA

Vector j LTR Gene IHHn Neo LTR (C) Vector Producer Cell

Helper Virus

Vector

Target Cell

Proviral RNA

Proviral DNA Product of Transferred

FIG. 1. Retrovirus-mediated gene transfer. (A) Scheme of retrovirus. (B) Helper virus and vector. (C) Gene delivery of retroviral vector. 3 and 3' LTR of the helper virus can be deleted to prevent the helper virus RNA transcripts fironn being packaged and subsequently reverse transcribed during the production of the recombinant retroviral vectors (25, 27).

Production problems with retroviral vector. A major consideration for the safety of retrovirus vector application is the detection and elimination of RCR in the vector preparations, although murine MoMLV based vector was not considered as an acute pathogen for primates (5). Recently, Donahue et al. (8) reported that 3 of 10 rhesus monkeys developed T cell lymphoma after receiving bone marrow transplants which were transduced with RCR-contaminated MoMLV-based retroviral vector. None of the analyzed lymphoma cells carried vector sequences, but they all contained multiple copies of RCR.

In addition to the T cell lymphoma, the infection of wild type MLV, which is replication competent (e.g. RCR), can also cause abnormal T cell proliferation and result in leukemia

(37). This kind of T cell proliferation is possibly induced by the binding of retrovirus to the receptors on T cells (11), Integration of a viral promoter into the host genome resulting in the overexpression of proto-oncogenes (18, 38) or a complex chromosome rearrangement

(14). Since RCR can also replicate in the target cells, such as the primate T cells described above, RCR can also cause acute viremia without activating any cellular oncogenes by genetic recombination (13). The formation of RCR has been demonstrated to occur by the abnormal template switches between helper virus and retroviral vector sequences during reverse transcription (Figure 2).

Retroviral RT enzyme intends to 'jump' from one template to the other during reverse transcription of RNA genome. Indeed, this 'jumping' feature is essential for retrovirus replication while the 5' LTR region Is reverse transcribed into DNA and serves as a primer to polymerize first strand DNA started from 3' LTR. This primer jumping is needed 4

(A)

m v* 5' 3" RNA

First Primer Jumpping ^ g«fl QOi «nv 3* 5- H -t 3" RNA

PBS aaa pK «iv U3 B osena Rrst Stranti DNA 5- H 1 I I 3'

PBS g«o pel env 3' Rrst Strand ONA ^ 3' Second Strand DNA pfst US R US PBS Second Primer Jumpping ^

U3 n us PBS U3 R OS

5-1 1^3' Oouble-Stranded 3-1 *5' Proviras

(B)

Vector •- Helper RNA genomes Virus

Strand exchange Qp- during DNA synthesis •

Recombinant single- • strand DNA H' Recombinant double- strand DNA

FIG.2. Reverse transcription and recombination. (A) Reverse transcription. (B) Template switchings between helper virus and vector transcripts enable helper virus to regain replication elements for replication-competent retrovirus (RCR) formation. 5 again while second strand DNA is initiated by a poiypurine tract (ppt) as a primer at 3' LTR and then jumps onto 5' LTR to complete double-stranded provirus followed by integration.

Once a helper virus is co-packaged into a virion particle with vector, the nature of RT enzyme enables the helper virus to pick up essential sequences from vector transcript to regain its replication ability by switching templates between helper virus and vector.

Compared to helper virus which contains gag, pol and env in one transcript, a more advanced helper virus has been developed by separating the env gene from the gag-pol genes. This is in an attempt to reduce the probability of RCR formation, since at least three recombination events are required before RCR can be produced (22). However, RCR outbreaks still occur in these 'split-function' packaging cells (2). Therefore, the prevention of

RCR outbreak from VPC in the future may require attempts to block or decrease RT enzyme activities inside of VPC instead of simply separating helper virus genes.

The phenotype of VPC can be altered by insertional mutagenesis of proviral vector re-integration. In some gene therapy clinical trials, VPC have been implanted into tumor lesions in the brain (30) and intraperitoneal cavity (20) of human subjects, therefore, the stability of VPC genotypes is a biosafety concem for gene therapy with this approach.

Possible origins of RT enzyme activity. Within VPC, there are two possible categories of RT enzyme activity: 1) exogenous viral RT. which is brought into the cell by virus infection. In cultured VPC, the only source of exogenous RT enzyme activity would be derived from the re-infection of vector virions; 2) endogenous RT. which is the combined RT enzyme activity of the helper virus-encoded PrlSO^^^"'"'or po/gene product of virion particles within the cell (6, 17, 33, 45), endogenous retroviruses (9, 19) and retrotransposons (12, 15,

16). These different sources of endogenous RT enzyme activity can result in intracellular retrotransposition activity (Figure 3) (for review see reference 21). 6

Integration cDNA

Reverse Transcription

AAA Spliced mRNA Translaton

Integration cDNA

Reverse Transcription

•• Proteins (Gag)

Translation •AAA mRNA

cDNA

Reverse Transcription Proteins • • • .• Budding

Translation AAA No Reinfection mRNA

FIG.3. Endogenous reverse transcriptase activity and retroelements. (A) Generation of pseudogenes. (B) Transposition of retrotransposons. (C) Expression and amplification of endogenous retrovirus. 7

During the vector assembly process, viral Env protein produced by the helper virus is transported to the cell membrane where it binds the cellular receptor used for virus entry

(24, 43, 49). This Env-receptor interference can dramatically reduce the re-infection of vector virions for imported exogenous RT (Figure 4). The re-infection ratio of amphotropic

MoMLV (Am-MoMLV) vector virion on target cells, which were previously infected with wild type Am-MoMLV, is as low as 1 x 10"^ events per cell (26).

In contrast, intracellular retrotransposition of an Env-defective C-type MoMLV can occur at a higher frequency of 2.7 x 10"* events per cell on feline G355-5 cells which are not susceptible to MoMLV infection (Figure 5). The endogenous RT enzyme activity required for retrotransposition originated from intracellular particles of gag-pol gene products (46). Un- cleaved PrlSO^^^'has been demonstrated with 30% to 50% of RT enzyme activity compared to cleaved, matured RT in virion particle (6, 17, 33). In turn, the core proteins of virion particles inside host cells without the maturing process, which usually involved cleavage of Gag and Pr180^^', can initiate reverse transcription without a budding process and is the major portion of endogenous RT activity in a retrovirus infected cell (45, 46).

Consequences of RT enzyme activity inside cells. The RT enzyme activity in a cell will produce more proviral vector (cDNA) by reverse transcription and possibly increase integrations into the host genome. Approximately, 10% of the human genome is estimated to consist of reverse transcribed and transposed sequences (1). These abundant endogenous RT activities can be reproduced in experimental systems. Episomal proviral

DNA, the product of the reverse transcription reaction during retrotransposition process, has been estimated as one copy (event) per 10"* cells from LINE retrotransposition (39).

Integration of proviral DNA can result in deletion (15) or inactivation (44) of targeted DNA, 8

VPC Env-Receptor Interference

Chromosomal DNA Vector RNA

Vector

integration

Pro virus Reverse Transcription Superinfection

FIG. 4. Env-receptor interference and superinfection of vectors on vector producer cells.

Resistance to superinfection of vector depends on threshold gene expression of helper virus

(Chapter 3). 9

Chromosomal DNA Vector RNA

Budding

Intracellular Vector Retrotransposition Receptor

Integration Proviais Reverse Transcription

FIG. 5. Intracellular retrotransposition of an Env-defective Moloney murine leukemia virus.

Note that an Env-defective viral particle is non-infectious and, therefore no superinfection occurs. 10

greater genome fluidity (29, 36, 48), as well as in the genesis of functional genes (42)(for a

review see reference 34).

In gene transfer studies, exogenous RT imported by retrovirus infection is correlated

with a risk of RCR formation which is initiated by template switches between co-packaged

helper virus and vector transcripts during reverse transcription. The recombination feature

of reverse transcription also enables retroviruses to modulate their genome to survive under

artificial or natural selection pressure. In a single retroviral replication cycle, approximately

8.8% of herpes simplexvirus thymidine kinase (HStk) genes in MoMLV-based vectors can

be deleted when transduced cells are placed under the selection pressure (e.g. neomycin

analog, G418) for another selectable gene (neomycin phosphotransferase, neo") present in

the vector (31, 47). This represents a mutation rate of 3% per kb per retroviral replication

cycle (31).

Stability of integrated provirus. Once a provirus integrates into a host genome,

this integrated provirus becomes part of host genome and can replicate with host

chromosomal replication during mitosis. Integration of provirus has been generally

considered very stable and can be transmitted from one generation to the next if a provirus

integrates into gamete cells. A previous study has only shown a mutation (deletion) rate of integrated provirus as low as 4.5 x 10"® events per gamete by studying mouse coat color revertants. The deletion of this provirus in the mouse genome was hypothesized to occur by homologous recombination between the two LTR and deletion of the provirus to restore the dilute (d'O allele (40). In this study, a deletion of provirus integration site was observed in one of six subclones of LTKOSN.2 VPC (Chapter 3), which is significantly higher than above mentioned deletion rates. Several mechanisms have been considered for the instability of integrated provirus. One proposal is an auto-integration process in which 11 proviral vector integrates into a previously-integrated provirai vector. This has been observed in a complex episomal provirus of human immunodeficiency virus (HIV) (32). A translocation or amplification mechanism of a cellular gene allele by unequal sister- chromatid exchange or disproportionate replication (for review see reference 37) may also be involved.

Dissertation Organization

The work described in this dissertation has focused on the study of genetic instability of retroviral sequences and the host genome caused by host-virus interactions. Because of the insertion of an antibiotic selection marker gene in retroviral vector construct, this selection marker gene becomes a useful DNA marker to distinguish retroviral vector sequences from endogenous retroviral elements. Therefore, one can study the evolution of viral sequences in the host cell during retrovirus replication. A VPC designed for gene transfer purposes is an excellent model to represent a retrovirus-infected cell for studying the host-retrovirus interaction on genetic stability of both retroviral sequences and host genome. The first manuscript, "Restriction mapping of retroviral vector mutation by analyzing episomal DNA", describes the details of a deletion of a theraputic gene from a retroviral vector. The finding of this second vector in our VPC initially led to the study of genetic instability of VPC in this dissertation. This manuscript will be submitted to

Biotechniques. The second manuscript, "DNA methylation increases genetic instability of retroviral vector producer cells", describes cascade reactions initiated by DNA methylation of retroviral helper virus. The consequences are genome fluidity of host cells and decreased vector production. This manuscript will be submitted to the Journal of Virology. The third manuscript, "Chimeric retroviral helper virus and picornavirus IRES sequence to eliminate 12

DNA methylation for improved retroviral packaging ceils", is to confirm the role of de novo

DNA methylation involved in host-virus interaction and describes an innovation designed to

Improve retroviral vector production for gene transfer study. This manuscript will be submitted to Human Gene Therapy \o support the previous manuscript. The last manuscript, "High frequency retrotransposition of retroviral vectors in cultured vector producer cells", is using a retroviral vector system to demonstrate abundant retrotransposition activities of vector in VPC. Results from this study also reveal an alternative replication pathway of exogenous retrovirus in addition to well studied cell-to-cell transmission pathway of retrovirus replication. This manuscript will be submitted to Human

Gene Therapy as well. Following the manuscripts is a "general conclusions" section to describe further studies and possible tasks to prevent genetic instability in VPC. Appendix A is an application of chimeric retroviral helper virus to establish a human packaging ceil population to evade human immune surveillance /n vitro.

Won-BIn Young was the principal investigator on the research reported in each of the manuscripts and Appendix, and was the first author of the manuscripts. The work in this dissertation was advised by co-major professors Gary L. Lindberg and Charies J. Link, Jr.

References

1. Baltimore, D. 1985. Retroviruses and retrotransposons: the role of reverse transcription

in shaping the eukaryotic genome. Cell. 40:481-2.

2. Cheng, H., and R. G. Vile. 1996. Replication-competent retrovirus produced by a ' split-

function' third generation amphotropic packaging cell line. Gene Ther. 3: 624-629. 3. Coffin, J. M. 1996. Retroviridae: The viruses and their replication, p. 1767-1847. In B.

N. Fields and D. M. Knipe and R. M. Chanock and M. S. Hirsch and J. L. Melnick and T.

P. Monath and B. Roizman (ed.), Virology, 3 ed, vol. 58. Lippincott-Raven, Philadelphia.

4. Cone, R. D., and R. C. Mulligan. 1984. High-efficiency gene transfer into mammalian

cells: generation of helper-free recombinant retrovirus with broad mammalian host

range. Proc Natl Acad Sci. 81:6349-6353.

5. Cornetta, K., R. C. Moen, K. Culver, R. A. Morgan, J. R. Mclachlin, S. Strum, J.

Selegue, W. London, R. M. Blaese, and W. F. Anderson. 1990. Amphotropic murine

leukemia retrovirus is not an acute pathogen for primates. Hum. Gene Ther. 1:15-30.

6. Crawford, S., and S. P. Goff. 1985. A deletion mutation in the 5' part of the pol gene of

Moloney murine leukemia virus blocks proteolytic processing of the gag and pol poly-

proteins. J. Virol. 53:899-907.

7. Crystal, R. G. 1995. Transfer of genes to humans: early lessons and obstacles to

success. Science. 270:404-10.

8. Donahue, R. E., S. W. Kessler, D. Bodine, K. McDonagh, C. Dunber, S. Goodman,

B. Agricola, E. Byrne, M. Raffeld, R. Moen, J. Bacher, K. M. Zsebo, and A. W.

Nienhuis. 1992. Helper virus induced T cell lymphoma in nonhuman primates after

retroviral mediated gene transfer. J. Exp. Med. 176:1125-35.

9. Goodchild, N. L., J. D. Freeman, and D. L. Mager. 1995. Spliced HERV-H

endogenous retroviral sequences in human genomic DNA: evidence for amplification via

retrotransposition. Virology. 206:164-73.

10. Gray, D. A. 1991. Insertional mutagenesis: neoplasia arising from retroviral integration.

Cancer Invest. 9:295-304. 11. Hardy, W. D., Jr. 1984. Retrovirus-lnduced Leukemias Of Animals and Humans. The

Role of Viruses in Human Cancer. Volume II. Giraldo G, Beth E. eds. N:311-29.

12. Heidmann, O., and T. Heidmann. 1991. Retrotransposition of a mouse lAP sequence

tagged with an indicator gene. Cell. 64:159-70.

13. Ihfe, J. N., A. Rein, and R. Mural. 1984. Immunologic and virologic mechanisms in

retrovirus-induced murine leukemogenesis. Adv Viral Oncol. 4:95-137 1984.

14. ihle, J. N., B. Smith-White, B. Sisson, D. Parker, D. G. Biair, A. Schuitz, C. Kozak,

R. D. Lunsford, D. Askew, Y. Weinstein, and R. J. Isfort. 1989. Activation of the c-H-

ras proto-oncogene by retrovirus insertion and chromosomal rearrangement in a

Moloney leukemia virus-induced T-cell leukemia. J Virol. 63:2959-66.

15. Jensen, S., IM. P. Gassama, and T. Heidmann. 1994. Retrotransposition of the

Drosophila LINE I element can induce deletion in the target DNA: a simple mode! also

accounting for the variability of the normally observed target site duplications. Biochem

Biophys Res Commun. 202:111-9.

16. Jensen, S., and T. Heidmann. 1991. An indicator gene for detection of germline

retrotransposition in transgenic Drosophila demonstrates RNA-mediated transposition of

the LINE I element. EMBO J. 10:1927-37.

17. Katoh, i., Y. Yoshinaka, A. Rein, IM. Shybuya, T. Odaka, and S. Oroszlan. 1985.

Murine leukemia virus mutations: protease region required for conversion from

"immature" to "mature" core form and for virus infectivity. Virology. 63:280-292.

18. Laze, P. A., and P. N. Tsichlis. 1988. Recombination between two integrated

proviruses, one of which was inserted near c-myc in a retrovirus-induced rat thymoma:

implications for tumor progression. J Virol. 62:788-94. 15

19. Lefebvre, S., B. Hubert, F. Tekaia, M. Brahic, and J. F. Bureau. 1995. Isolation from

human brain of six previously unreported cDNAs related to the reverse transcriptase of

human endogenous retroviruses. AIDS Res Hum Retroviruses. 11:231-7.

20. Link, C. J., Jr., D. Moorman, T. Seregina, J. P. Levy, and K. J. Schaboid. 1996. A

phase I trial of in vivo gene therapy with the herpes simplex thymidine kinase/ganciclovir

system for the treatment of refractory or recurrent ovarian cancer. Hum Gene Ther.

7:1161-79.

21. Lower, R., J. Lower, and R. Kurth. 1996. The viruses in all of us: characteristics and

biological significance of human endogenous retrovirus sequences. Proc Natl Acad Sci

USA. 93:5177-84.

22. Markowitz, D., S. Goff, and A. Bank. 1988. A safe packaging line for gene transfer:

separating viral genes on two different plasmids. J. Virol. 62:1120-4.

23. Marshall, E. 1995. Gene therapy's growing pains. Science. 269:1050-1055.

24. Miller, A. D. 1996. Cell-surface receptors for retroviruses and implications for gene

transfer. Proc Natl Acad Sci USA. 93:11407-13.

25. Miller, A. D., and C. Buttlmore. 1986. Redesign of retrovirus packaging cell lines to

avoid recombination leading to helper virus production. Mol Cell Biol. 6:2895-902.

26. Miller, A. D., and F. Chen. 1996. Retrovirus packaging cells based on 10A1 murine

leukemia virus for production of vectors that use multiple receptors for cell entry. J Virol.

70:5564-71.

27. Miller, A. D., and G. J. Rosman. 1989. Improved retroviral vectors for gene transfer

and expression. Biotechniques. 7:980-2, 984-6, 989-90.

28. Mulligan, R. C. 1993. The basic science of gene therapy. Science. 260:926-932. 16

29. Nouvei, P. 1994. The mammalian genome shaping activity of reverse transcriptase.

Genetica. 93:191-201.

30. Oldfield, E. H., Z. Ram, K. W. Culver, R. M. Blaese, H. L. DeVroom, and W. F.

Anderson. 1993. Gene therapy for the treatment of brain tumors using intra-tumoral

transduction with the thymidine kinase gene and intravenous ganciclovir. Hum Gene

Ther. 4:39-69.

31. Parthasarathi, S., A. Varela-Echavarria, Y. Ron, B. D. Preston, and J. P. Dougherty.

1995. Genetic rearrangements occum'ng during a single cycle of murine leukemia virus

vector replication: characterization and implications. J Virol. 69:7991-8000.

32. Pauza, C. D., and J. Galindo. 1989. Persistent human immunodeficiency virus type 1

infection of monoblastoid cells leads to accumulation of self-integrated viral DNA and to

production of defective virions. J Virol. 63:3700-7.

33. Peng, C., B. K. Ho, T. W. Chang, and N. T. Chang. 1989. Role of human

immunodeficiency virus type 1-specific protease in core protein maturation and viral

infectivity. J Virol. 63:2550-6.

34. Robins, D. M., and L. C, Samuelson. 1992. Retrotransposons and the evolution of

mammalian gene expression. Genetica. 86:191-201.

35. Rodgers, J. R., C. Smith, S. M. Shawar, and R. R. Rich. 1990. Retrovirus-mediated

insertional mutagenesis: phagemid rescue of flanking DNA by selecting plasmid ori.

DNA Cell Biol. 9:139-47.

36. SanMiguel, P., A. Tikhonov, Y.-K. Jin, N. Motchoulskaia, D. Zakharov, A. Melake-

Berhan, P. S. Springer, K. J. Edwards, M. Lee, Z. Avramova, and J. Bennetzen.

1996. Nested retrotransposons in the intergenic regions of the maize genome. Science.

274:765-8. 17

37. Schiff, R. D., and A. Ollff. 1986. The pathophysiology of murine retrovirus-induced

leukemias. Crit Rev Oncol Hematol. 5:257-323.

38. Schwartz, J. R., S. Duesberg, and P. Duesberg. 1995. DNA recombination is

sufficient for retroviral transduction. Proc Natl Acad Sci USA. 92:2460-4.

39. Segal-Bendirdjian, E., and T. Heidmann. 1991. Evidence for a reverse transcription

intermediate for a marked LINE transposon in tumoral rat cells. Biochem Biophys Res

Commun. 181:863-70.

40. Seperack, P. K., M. C. Strobel, D. J. Ccrrcw, N. A. Jenkins, and N. G. Copeland.

1988. Somatic and germ-line mutations rates of the retrovirus-induced dilute coat-color

mutation of DBA mice. Proc. Natl. Acad. Sci. USA. 85:189-192.

41. Smith, K. T., A. J. Shepherd, J. E. Bcyd, and G. M. Lees. 1996. Gene delivery

systems for use in gene therapy: an overview of quality assurance and safety issues.

Gen. Ther. 3:190-200.

42. Scares, M. B., E. Schon, A. Henderson, S. K. Karathanasis, R. Cate, S. Zeitlin, J.

Chirgwin, and A. Efstratiadis. 1985. RNA-mediated gene duplication: the rat

preproinsulin I gene is a functional retroposon. Mol. Cell. Biol. 5:2090-2103.

43. Sommerfelt, M. A., and R. A. Weiss. 1990. Receptor interference groups of 20

retroviruses plating on human cells. Virology. 176:58-69.

44. Steinemann, M., and S, Steinemann. 1992. Degenerating Y chromosome of

Drosophila miranda: a trap for retrotransposons. Proc Natl Acad Sci USA. 89:7591-5.

45. Tchenic, T., and T. Heidmann. 1991. Defective retroviruses can disperse in the human

genome by intracellular transposition. J Virol. 65:2113-8. 18

46. Tchenio, T., and T. Heidmann. 1992. High-frequency intracellular transposition of a

defective mammalian provirus detected by an in situ colorimetric assay. J Virol.

66:1571-8.

47. Varela-Echavarria, A., C. M. Prorock, Y. Ron, and J. P. Dougherty. 1993. High rate

of genetic rearrangement during replication of a Moloney murine leukemia virus-based

vector. J Virol. 67:6357-64.

48. Voytas, D. 1996. Retroelements in genome organization. Science. 274:737-8.

49. Weiss, R. A., and C. S. Tailor. 1995. Retrovirus receptors. Cell. 82:531-3. 19

CHAPTER 2. RESTRICTION MAPPING OF RETROVIRAL VECTOR MUTATION

BY ANALYZING EPISOMAL DNA

A paper to be submitted to BioTechniques

Won-Bin Young'-^, E. Jeffery Beecham^, Gary L. Llndberg"^, and Charles J. Link, Jr.'-^*

ABBREVIATIONS:

VPC: vector producer ceil HSVf/c Herpes simplex tiiymidine l

LTR: long terminal repeat MoMLV: Moloney murine leukemia virus

KEYWORDS: vector producer cell; genetic instability; gene therapy; retrovirus; recombination; retroviral vector; packaging cell; iierpes simplex virus thymidine kinase;

Moloney murine leukemia virus; deletion.

^Human Gene Therapy Research Institute, John Stoddard Cancer Center, 1415 Woodland

Ave., Des Moines, Iowa 50309. ^Molecular, Cellular and Developmental Biology Program,

Iowa State University, Ames, Iowa 50011. ^Department of Medicine, Harvard Medical

School, Boston, Massachusetts 02115. '^Department of Animal Science, iowa State

University, Ames, Iowa 50011. *To whom reprint requests should be sent, (515) 241-8787;

Fax: (515) 241-8788; e-mail: [email protected]. 20

INTRODUCTION

The Infidelity of RT enzyme activity during the replication cycle of retrovirus can cause retroviruses to rapidly mutate and this raises several biosafety issues, such as deletion of therapeutic gene from vector and replication-competent retrovirus (RCR). In a single retroviral replication cycle, a therapeutic herpes simplex virus (HSV) thymidine kinase

(tk) gene was inactivated in approximately 8% of Moloney murine leukemia virus (MoMLV) - based vectors and mutation rates within a single retroviral vector were calculated as high as

3% per kb (7, 12). In addition to the deletion of HSVtk from retroviral vector, retroviral vectors carrying nerve growth factor receptor gene (5), a-rev sequence (1), or human glucocerebrosidase (14), have also been observed with a deletion mutation. To distinguish mutant vector from primary vector from genomic DNA is difficult since both vectors share most of the DNA sequences except the deletion region. To characterize the primary structure of deletion mutated vector and define the deletion region requires several probings on Southern blot analysis for restriction endonuclease mapping, which is easily interfered by endogenous retroviral element sequences in the mammalian genome, since these sequences are highly homologous to vector sequences.

High frequency of superinfection (15) and retrotransposltion activity (16) of retroviral vector in cultured VPC results in a detectable amount of episomal DNA which is suitable for

Southern blotting analysis of mutated vector without the above mentioned complications.

The observations of episomal vector or retroviral sequences were also previously reported with other retroviruses, including mouse mammary tumor virus (9), avian sarcoma virus (13), avian leukosis virus (10), HIV-1 (8) and in avian packaging cells (4). In this study, we successfully identified a HSVtk-deleted vector from VPC by analyzing episomal DNA instead of complicated genomic DNA restriction mapping. 21

MATERIALS AND METHODS

Construction of the pLTKOSN retroviral vector and LTKOSN.2 vector producer cell line. The pLTKOSN plasmid DNA (kindly provided by J. P. Levy) was constructed in a

LXSN retroviral backbone (kindly provided by A. D. Miller) and has been designed to express both the neo"^ and HStk gene (6). The HStk open reading frame was amplified with the upstream primer 5'CCA AGC TTC GGC CAG CGC CTT GTA G 3' (25 mer) and downstream primer 5'CCA AGC TTC CGG TAT TGT CTC CTT CC3' (26 mer). The HStk gene was ligated into the PCR cloning vector, PCR II (Invitrogen, Carlsbad, CA), to produce pTKO-PCRII which contains EcoR I sites flanking the TKO PCR product A 1,201 bp TKO

EcoR I fragment from pTKO-PCR II was isolated and cloned into the EcoR I (X) site of pLXSN to produce pLTKOSN (Fig. 1A).

LTKOSN.2 VPC was established by transient transfection of pLTKOSN plasmid DNA into the ecotropic packaging cell line GP+E86 (kindly provided by A. Bank). Supemates from these cells were used to transduce the amphotropic retroviral packaging line PA317 (kindly provided by A. D. Miller) and selected in G418 (1 mg/ml) for 2 weeks. Twenty different VPC clones were subcloned from the original pool. LTKOSN.2 VPC produces a high titer of approximately 1 x 10® colony formation unit (cfu) per ml by G418 selection on transduced target cells (3).

RNA analysis of LTKOSN vector supernatants. Viral supernatant was subjected to 20% sucrose gradient ultracentrifugation (125,000 X g) for 2hr at 4°C for virion pelleting and then extracted by RNAzol (Biotecx, Houston, TX). Viral RNA was subjected to Northern blot analysis on a 1% agarose-0.4M formaldehyde gel. Vector transcripts were detected by

Neo''probe and HStk probe.

Episomal DNA analysis. Episomal DNA was extracted from the cytoplasmic fraction of 1 X 10^ LTKOSN.2 VPC. Briefly, cells were trypsinized and subjected to 1% Triton X-100 22 detergent for 5 min at room temperature to lyse only the cellular membrane but not nuclear membrane. Nuclei were separated from the cytoplasmic fraction by centrifugation at 9,500 X g for 5 min at 4°C (2). Cytoplasmic fractions were subjected to phenol/chloroform extraction

(11). These episomal DNA samples were treated with ribonuclease A (RNase A, 1 mg/ml,

Boehringer Mannheim, Indianapolis, IN) to eliminate RNA contamination prior to restriction enzyme digestion. Southern blot analysis of episomal proviral LTKOSN and ALTKOSN DNA was performed and the membrane was hybridized with various probes, Including long terminal repeat (LTR; SacW/Kpnl), extended packaging signal sequence SpeUEcoRl),

HSVtk (EcoRl fragment), SV40 promoter (BamHUStul), and Neo*" {Hind\\\/Bpm\),

32 respectively (Fig. 1A). These probes were labeled with P-dCTP by a random primer method according to the manufacturer's protocol (Boehringer Mannheim).

DNA ampiification and sequence analysis of ALTKOSN. To analyze the deletion region of ALTKOSN proviral vector, genomic DNA was PGR amplified by forward primer, 5'-

CTG TGT CTG TCC GAT TGT CTA GTG TC-S within the extended packaging signal region, and reverse primer, 5'-CCC TTC CCG CTT CAG TGA CAA CG-3' within the

Neo'" gene. The PGR product was purified by gel separation and subjected to sequencing analysis (DNA Sequencing Facility, Iowa State University, Ames, Iowa) by using the same primers.

RESULTS

A deletion of the HStk gene in our vector production was first detected in viral RNA collected from pelleted viral particles in Northem blot analysis by using different probes (Fig.

1). To avoid the detection of endogenous retroviral elements and helper virus present In the cellular genomes, unintegrated, episomal form of viral DNA was therefore employed for 23

Southern blot analysis. Small amounts of episomal DNA derived from vector sequences has been routinely detected within VPC from both PAS17 derived VPC. These episomal DNA were mainly detected in cytoplasmic fraction rather than nuclear fraction of VPC (Fig. 2).

Analysis of LTKOSN.2 VPC demonstrated a single LTKOSN vector after probing for

HSV^gene sequence on these episomal DNAs. However, probing with the Neo''gene on extracted episomal DNA from LTKOSN.2 VPC demonstrated a second smaller vector (Fig.

3A, lane 1). To identify the primary structure of this second viral vector, restriction mapping and Southern blot analysis of the retroviral vectors was performed on the episomal DNA extracted from the cytoplasmic fraction of LTKOSN.2 VPC. Without restriction enzyme digestion, two respective sizes of episomal vector DNA were detected, 4.5-kb and 3.0-kb.

Linear LTKOSN is represented by the 4.5-kb DNA band (Fig. 3A, lane 1). Ba/vHI digestion of the episomal DNA resulted in two fragments, 2.7-kb and 1.8-kb, being generated from the

4.5-kb linear LTKOSN, while the second episomal proviral vector (3.0-kb) was resistant to

BamHI digestion (Fig. 3A, lane 2). This suggests that the 3.0-kb DNA was possibly a mutant of LTKOSN with the BamHI site deleted. Serial re-hybridizations with different probes were employed to map the deletion region (Fig. 3A). The primary structure of this truncated 3.0-kb

LTKOSN vector was found to include the S LTR, extended packaging signal (^ and portion of gag sequence), the Neo*" gene and the 3 LTR, but did not contain the HSVtk gene and

SV40 promoter in this assay. PCR primers flanking the deletion region within extended packaging signal region and Neo''gene were used to amplify this deletion region from both

LTKOSN.2 VPC genomic DNA and episomal DNA for sequencing analysis. The deletion region in the mutant vector, ALTKOSN, included the entire HSVtk gene (except the EcoRI polylinker region at its 5 end) and most of the SV40 promoter sequence (Fig. 3B). 24

DiSSCUSION

The mechanism of origin of this ALTKOSN is still not defined and might represent simultaneous infections with two vectors when LTK0SN.2 VPC was established. The deletion of the HSVtk gene from the LTKOSN vector may have occurred during the transduction process of LTKOSN vector into PA317 packaging cells from GP+E86, and this conclusion Is supported by the fact that ALTKOSN vector was not detected in LTKOSN transfected GP+E86 VPC (15).

REFERENCES

1. Junker, U., E. Bohniein, and G. Veres. 1995. Genetic instability of a MoMLV-based

antisense double-copy retroviral vector designed for HIV-1 gene therapy. Gene Ther.

2:639-46.

2. Lindberg, G. L., 0. K. Koehier, J. E. Mayfield, A. M. Myers, and D. 0. Beitz. 1992.

Recovery mitochondrial DNA from blood leukocytes using detergent lysis. Biochem

Genet. 30:27-33.

3. Link, C. J., Jr., D. Moorman, T. Seregina, J. P. Levy, and K. J. Schabold. 1996. A

phase I trial of in vivo gene therapy with the herpes simplex thymidine kinase/ganciclovir

system for the treatment of refractory or recurrent ovarian cancer. Hum Gene Ther.

7:1161-79.

4. Lum, R., and M. L. LiniaL 1998. Retrotransposition of nonviral RNAs in an avian

packaging cell line. J Virol. 72:4057-64.

5. Mavilio, F., G. Ferrari, 8. Rossini, N. Nobili, 0. Bonini, G. Casorati, 0. Traversari,

and C. Bordignon. 1994. Peripheral blood lymphocytes as target cells of retroviral

vector-mediated gene transfer. Blood. 83:1988-97. 6. Miller, A. D., and G. J. Rosman. 1989. Improved retroviral vectors for gene transfer and

expression. Biotechniques. 7:980-2, 984-6, 989-90.

7. Parthasarathi, S., A. Varela-Echavarria, Y. Ron, B. D. Preston, and J. P. Dougherty.

1995. Genetic rearrangements occurring during a single cycle of murine leukemia virus

vector replication: characterization and implications. J Virol. 69:7991-8000.

8. Pauza, C. D., and J. Galindo. 1989. Persistent human immunodeficiency virus type 1

infection of monoblastoid cells leads to accumulation of self-integrated viral DNA and to

production of defective virions. J Virol. 63:3700-7.

9. Ringold, G. M., K. R. Yamamoto, P. R. Shank, and H. E. Varmus. 1977. Mouse

mammary tumor virus DNA In Infected rat cells; characterization of unlntegrated forms.

Cell. 10:19-26.

10. Robinson, H. L., and B. D. Miles. 1985. Avian leukosis virus-induced osteopetrosis Is

associated with the persistent synthesis of viral DNA. Virology. 141:130-43.

11. Sambrook, J,, E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory

manual., p. 9.14 - 9.23, Chapter 9: Analysis and cloning of eukaryotic genomic DNA,

2nd. ed. Cold spring Harbor Laboratory Press, Cold Spring Harbor, N. Y.

12. Varela-Echavarria, A., C. M. Prorock, Y. Ron, and J. P. Dougherty. 1993. High rate of

genetic rearrangement during replication of a Moloney murine leukemia virus-based

vector. J Virol. 67:6357-64.

13. Varmus, H. E., and P. R. Shank. 1976. Unlntegrated viral DNA is synthesized in the

cytoplasm of avian sarcoma virus-transformed duck cells by viral DNA polymerase. J

Virol. 18:567-73.

14. Weinthal, J., J. A. Nolta, X. J. Yu, J. Lilley, L. Uribe, and D. B. Kohn. 1991.

Expression of human glucocerebrosldase following retroviral vector- mediated

transduction of murine hematopoietic stem cells. Bone Manrow Transplant. 8:403-12. 26

15. Young, W.-B., G. L. Lindberg, and C. J. Link, Jr. 1999. DNA methylation increased

genetic instability of retroviral vector producer cells. In preparation.

16. Young, W.-B., G. L. Lindberg, and C. J. Link, Jr. 1999. High frequency

retrotransposition of retroviral vector in cultured vector producer cells. (In preparation). FIG. 1. Detection of second vector from LTK0SN.2 vector production. Viral RNA was extracted from virion particle pellet and subjected to Northern blot analysis. (A) Schematic diagram of LTKOSN vector and the probes. LTKOSN contains a HSVtk gene, which was cloned into the EcoRI site of LXSN. W, extended packaging signal region; Neo, neomycin phosphate transferase gene; SV, promoter sequence of simian virus 40 early gene. The probes were produced from LTKOSN vector by the following restriction endonucleases: B,

BamH\; Bp, Bpml: E, EcoRI; H, Hindlll; K, h^n\; Sa, Sacll; Sp, Spel; St, Stul. (B) Two populations (4.5-kb and 3.0-kb) of vectors were detected from packaged viral transcripts by packaging signal probe (\j/) and Neor probe (Neo). Only one population (4.5-kb) of vector was detected by HStk probe (tk) probe. 28

LTKOSN vector 4.5 kb K E E B K V LTR ' -F HSVtk •TS^ Neo H LTR 1 I Sa K Sp E E B St H Bp

Probe: FIG.2. Location of episomal DNA in vector producer cells. Nuclear and cytoplasmic fractions were separated by centrifugation after 1% Triton X-100 detergent treatment (2)

(see Material and Methods). Vast majority of episomal DNA was detected in cytoplasmic fraction by Neo probe. 30

4.0-

2.0-

1.0- FIG. 3. Characterization of HSVtl< deleted LTKOSN (ALTKOSN) by episomal DNA

restriction mapping. (A) Episomal DNA was extracted from the cytoplasm fraction of

LTKOSN.2 VPC (see Materials and Methods). Without restriction digestion (lane 1),

digestion with BamHl (lane 2) or SamHI/EcoRI (lane 3), these episomal DNA were

transferred onto a nylon membrane and hybridized with different probes, LTR, tk, SV40

promoter (SV40) and Neo"^, respectively. Without endonuclease treatment, LTKOSN vector

is 4.5-kb and ALTKOSN vector is 3.0-kb in size. Note the absence of tk or SV40 probe

hybridization to the ALTKOSN vector. (B) Schematic map of ALTKOSN was constructed as determined by Southern blot analysis. PGR primers were then designed to amplify the deletion region for sequencing analysis. A 1.5-kb region including the HSVtk gene and the 5 portion of SV40 promoter was deleted. The EcoRI polylinker now joins with the 3 portion of

SV40 promoter. Drawings are not to scale. 32

kb 1 2 3 123 123 123 123

W tk SV40 Neo

B ALTKOSN

l-<- 3.0 kb ->-l K E K LTR Neo - LTR

SV40 promoter EcoRI CGCCTCGGCTTCTGAGCTA GCCGGAATTCGGCTTCCAAGCTTCG ' HSVtk 33

CHAPTER 3. DNA METHYLATION INCREASES GENETIC INSTABILITY OF RETROVIRAL VECTOR PRODUCER CELLS

A paper to be submitted to the Journal of Virology

Woh-Bin Young^"^, Gary L Lindbergh, and Charles J. Link, Jr.^'^*

RUNNING TITLE: DNA methylation and VPC genetic instability

ABBREVIATIONS:

GaLV: gibbon ape leukemia virus HSVffc Herpes simplex thymidine kinase

LTR: long terminal repeat MoMLV: Moloney murine leukemia virus

RCR: replication competent retrovirus RT: reverse transcriptase

VPC: vector producer cell Env: envelope

KEYWORDS: vector producer cell; genetic instability; gene therapy; retrovirus; DNA methylation; recombination; retroviral vector; packaging cell; integration; superinfection; retroviral interference; herpes simplex virus thymidine kinase; gibbon ape leukemia virus;

Moloney murine leukemia virus; deletion; 5-azacytidine; demethylation.

'Human Gene Therapy Research Institute, John Stoddard Cancer Center, 1415 Woodland Ave., Des

Moines, Iowa 50309. ^Molecular, Cellular and Developmental Biology Program, Iowa State University,

Ames, Iowa 50011. ^Department of Animal Science, Iowa State University, Ames, Iowa 50011. *To

whom reprint requests should be sent, (515) 241-8787; Fax: (515) 241-8788; e-mail: [email protected]. 34

ABSTRACT

Retroviral vector producer cells (VPC) have been considered genetically stable with a uniform vector integration pattem in the cell population for sustained vector production. In this study, we observed that the copy number of vectors is increased and varied in the population of established LTKOSN.2 VPC (Hum. Gene Ther. 7:1161-1179). A superinfection experiment demonstrated that this results from decreased Env-receptor interference by suppression of helper virus gene expression to allow increased superinfection by its own vector. In a particular subclone, the number of integrated vector was increased from 1 copy per cell to 9 copies per cell during a 31-day period of cell culture.

Vector titer fell from 1.5 X 10® colony forming unit (cfu)/ml to 0 cfu/ml. To understand the involved mechanism, DNA methylation status of helper virus and vectors were examined by methylation-sensitive restriction enzyme digestion. We demonstrated that DNA methylation of helper virus 5' LTR occurred at 2% population per day and correlated closely with inactlvation of helper virus gene expression. In contrast, retroviral vectors did not exhibit significant methylation and show consistent transcription activity. Treatment with 5- azacytidine (5-aza-C), a methylation inhibitor, partially reversed the helper virus DNA methylation and restored vector production. The preference for methylation of helper virus sequences over vector sequences may have important implications for host-virus interaction. Designing a helper virus to overcome cellular DNA methylation may therefore improve vector production and increase viral envelope-receptor interference to prevent replication-competent retrovirus (RCR) formation. 35

INTRODUCTION

Retroviral vectors are the most commonly used gene delivery vehicles in human gene transfer trials (49). Permanent modifications of a targeted cell's genotype by the integration of a retroviral vector and the consistent production of retroviral vectors from permanent packaging cells (5, 31) offer significant advantages over other gene delivery vectors. Retroviruses convert their single-stranded RNA genome to double-stranded DNA

(proviral DNA) by the activity of virus-associated reverse transcriptase (RT) after infection of target cells. The proviral DNA is permanently integrated into the cellular chromosomes as a provirus by viral integrase activity and then replicates as part of the host genome. The proviral DNA is then transmitted from one generation to the next if the virus is integrated into a germ line cell. The deletion of a provirus from its integration site is only observed in approximately one of 10® cells (52, 58).

The major biosafety concern with retroviral vectors is replication-competent retrovirus (RCR) outbreak (1). RCR has been correlated with the occurrence of T-cell lymphomas in monkeys that had received RCR-contaminated bone marrow cell transplants

(7, 57). Vector and helper virus rearrangement caused by abnormal template switches during the reverse transcription process has been considered the major cause of RCR formation (45, 57). Without Env-receptor interference, co-cultivation of VPCs with different host ranges, such as ecotropic and amphotropic VPC, can cause RCR formation in 3 to 10 days depending on the number of recombination events necessary. The driving force for these recombination events is RT enzyme activity imported by the infection of the vector from the other VPC (37).

During the vector assembly process, viral Env protein produced by the helper virus is transported to the cell membrane where it binds the cellular receptor used for virus entry 36

(30, 53, 59). This Env-receptor interference has been considered an efficient barrier to block the superinfection of vectors. The transduction efficiency of amphotropic vector on amphotropic PA317 packaging cells is reduced by six orders of magnitude from the amphotropic Env-receptor interference mechanism, compared to the transduction efficiency of the same vector on N1H3T3 (32). However, Env-receptor interference is not a complete block to superinfection. Most of cun-ently used retroviral packaging cells in gene transfer studies are based on either murine cells [/. e. NIH3T3 cells in GP+E86 (28) and PA317 (31)] or human cells [/. e. 293T cells in BOSC (46)] and express virus receptors. Thus, these packaging cells can still be infected by the vector they produce, albeit at low frequency (32,

35). These vector re-entry events may lead to RCR formation.

Suicide gene therapy against cancer cells by the combination of HSVf/c and ganciclovir (GCV) is a well-developed approach in human gene therapy trials (36, 42). We previously established a murine LTK0SN.2 VPC to transfer the HSVtfc gene by transducing a LTKOSN vector from GP+E86 to PA317 cells and demonstrated effective suicide gene therapy against breast cancer cells in vitro (26), colon carcinoma in vitro and in vivo (40), and human ovarian adenocarcenoma xenografts in vivo (27). In the current study of

LTKOSN.2 VPC and derivative subclones, we observed significant genetic instability of several types, including a HSVf/c-deleted mutant vector found in addition to the original

LTKOSN vector in the same VPC, increased copies of both vectors, a deleted provirus integration site, and a complete loss of vector production. To understand the mechanisms involved in these observations, we analyzed the DNA methylation status and the gene expression of vectors and helper virus. We demonstrated that transcriptional inactivation occurred of helper virus but not vectors. Significant DNA methylation of the helper virus 5'

LTR region was observed in VPC subclones that lacked helper virus gene expression. 37

These same VPC subclones also had a loss of vector production and high vector copy number. A superinfection experiment demonstrated that these subclones are significantly susceptible to the re-entry of vector, compared to the other subclones. These findings suggest that DNA methylation may lead to VPC genome instability by reducing Env-receptor interference and permitting high levels of vector re-entry. (W.-B. Young is a doctoral candidate at Iowa State University, and this work represents partial fulfillment of his Ph. D. dissertation requirement.)

MATERIALS AND METHODS

Generation of LTK0SN.2 VPC and derived subclones. The construction of pLTKOSN and LTK0SN.2 VPC has been previously described (27). Briefly, a 1,201 bp of

PCR-amplified HSVtk gene was cloned into the EcoRI site of pLXSN (34) (kindly provided by A. Dusty Miller, Fred Hutchinson Cancer Research Center, Seattle, WA) to produce pLTKOSN retroviral vector (Fig.1 A). Ecotropic packaging cell line GP+E86 (28), a generous gift from Arthur Bank, Columbia University, was transiently transfected with pLTKOSN.

Supemates from these GP+E86 cells were used to transduce the amphotropic retroviral packaging line PA317 (31), kindly provided by A. Dusty Miller. Forty different VPC clones were isolated and the LTK0SN.2 VPC produced the highest titer of 1.6 X 10® cfu/ml (27).

The subclones of LTK0SN.2 in this study were obtained by limiting dilution of LTKOSN.2

VPC onto two 96-well plates and examined by light microscopy to ensure that only single- cell subclones were further analyzed. Six single-cell clones (#1-5 and #10) were identified and expanded for further study.

Cell culture and drug treatment. Cell cultures were maintained in Dulbecco modified Eagle medium (DMEM; GIBCO BRL Life Technology Co., Gaithersburg, MD), 10% 38

fetal calf serum with 5% CO2, at 37°C. To reverse the DNA methylation, cells were grown in

DMEM containing 5 of 5-aza-C (Sigma Chemical Co., SL Louis, MO) for 72 hr (25) and then subjected to the titer assay, vector RNA slot blotting and DNA methylation analysis.

Retroviral infection. Vector titers were determined by the transduction of NIH3T3 tk{-) cells [American Type Culture Collection (ATCC) CRL1658]. A375 cells (ATCC

CRL1619, human melanoma) and IGROV cells [human ovarian carcinoma (56)] with 10-fold serial dilutions of vector stocks. Transduction was performed by incubating 1 X 10® cells/well in 6-well plates with vector in the presence of 10 ng/ml of protamine sulfate (Fujisawa USA

Inc., IL) in 1 ml of DMEM. After transduction for 24 hrs, the cells were selected in medium containing G418 (1 mg/ml) for 10-14 days. Titers were obtained by multiplying the number of resistant colonies by the dilution factor.

To perform superinfection assay on each subclone of LTKOSN.2 VPC, a LEIN retroviral vector carrying an enhanced green fluorescent protein (EGFP, (6, 10)) reporter gene was used to transduce LTKOSN.2 VPC subclones. This LEIN vector is based on an

LXSN vector but with the replacement of intemal ribosomal entry site (IRES, (18)) for the

SV40 promoter in LXSN. The EGFP gene was inserted in front of IRES to demonstrate transduction efficiency in target cells. Cells from NIH3T3, PA317, parental LTKOSN.2 VPC and LTKOSN.2 VPC subclones were seeded at 5 X 10® cells per well on 12-wells plates and exposed to 0.5-ml supemate of LEIN vector (3.4 X 10® cfu/ml) in the presence of 10 ^g/ml of protamine sulfate the day after seeding. Two days after a single exposure to LEIN vector, the transduction efficiency of these cell lines and VPC subclones was determined by fluorescence-activated cell sorter (FACS) analysis of EGFP expression (38) on an EPICS®

Profile II Analyzer (Coulter Co., Miami, FL, USA). 39

Southern blot analysis of genomic DNA. Total cellular DNA was extracted from

cell cultures using phenol/chloroform extraction and then dissolved in TE buffer (pH 8.0)

overnight at 55°C (50). To study the copy number of integrated vector in these subclones,

genomic DNA was digested by Kpnl to release nearly full length LTKOSN (4.0-kb) and

ALTKOSN (2.5-kb) fragments (Fig. 1) and then Southem blotted for Neo'" probe detection. A

0.7-kb BsfKl fragment of pPAM3 was used as a probe to detect a 2.5-kb DNA fragment of endogenous retroviral elements to demonstrate similar loading of DNA samples. To analyze for multiple integration sites in each subclone, SamHI was used to generate DNA fragments containing both proviral vector sequences and flanking chromosomal sequences. Southem blot analysis was performed and the duplicates of membrane were hybridized with Neo"^ and

HSVtk probes, respectively (Fig. 2).

RNA analysis of LTKOSN VPC and vector supernatants. Total cellular RNA was isolated from individual subclones using RNAeasy kit (Qiagen Inc., Valencia, CA). For virion

RNA extraction, viral supernatant was subjected to 20% sucrose gradient ultracentrifugation

(125,000 X g) for 2hr at 4°C for virion pelleting and then extracted by RNAzol (Biotecx,

Houston, TX). Both cellular and virion RNA from each subclone were subjected to Northern blot analysis on a 1% agarose-0.4M formaldehyde gel. Vector transcripts were detected by a Neo''probe and helper viral transcripts were detected by a 1.4-kb env probe, which was digested from pPAMS by Xho\. RNA slot blotting of titer determination (39) was performed by applying 200 jal of supemates, directly collected from VPC medium and centrifuged at

2,800 X g to eliminated the cellular debris, onto a slot blot apparatus (SlotBlot, Hoefer

Scientific Instruments, San Francisco, CA). Vector RNA was UV-cross-linked onto this membrane and detected by a Neo*^ probe. 40

Methylation analysis. The methylation status of provlrus and vectors at the Smal site In the 5' LTR as determined by digestion of genomic DNA with Oral and EcoRV to reduce the DNA fragment size. The DNA was then precipitated with ethanol, re-dlssolved in sterile water and divided into two equal portions, one of which was subjected to methylation- sensltive Smal restriction endonuclease digestion. The Southern blot membrane was hybridized with a 428-bp fragment of gag sequence {Pvu\\/Dra\) from pPAM3 to detect helper virus, a 273-bp EcoRI/EcoRV fragment of HSVtff sequence for LTKOSN vector, and a Neo*" probe to detect both LTKOSN and ALTKOSN, respectively (Fig. 7E). This Southern blot was also probed with an env probe to determine the methylation status of the Smal site

In the helper virus RT region. DNA derived from NIH3T3 and PG13 (ATCC CRL 10686), a packaging cell line for gibbon ape leukemia virus (GaLV) pseudotyped vector (33), were used as negative controls. Densitometric analyses were performed with a Hoefer

Densitometer GS300 (Hoefer Scientific Instruments) to measure the relative densities of

Smal-sensitive band compared to DraUEcoRV band. Due to the interference from endogenous retroviral elements, the fraction of Smal methylation in 5' LTR was calculated as one minus the Intensity ratio of the Smal sensitive band (1.5-kb) divided by DraUEcoRS/ band (1.8-kb) as depicted in Fig. 5.

RESULTS

Dynamics of proviral DNA integration. We have previously detected a HSVtk- deleted mutant vector (ALTKOSN) co-existing with full-length LTKOSN vector (Fig. 1A) in our LTKOSN.2 VPC population. To Investigate whether these ALTKOSN and LTKOSN vectors co-existed within the same individual cellular genome or existed separately in portions of the LTKOSN.2 VPC population, limiting dilution cloning on LTKOSN.2 VPC were 41 performed to isolate six single cell subclones (#1-5, and #10). After expanding the subclones to a sufficient cell population, each clone was then analyzed for integrated vectors by Kpnl digestion on both LTR regions to release the integrated proviral vectors of

ALTKOSN (2.5-kb, Fig. 1A) and LTKOSN (4.0-kb, Fig 1A). Southern blot analysis demonstrated that both ALTKOSN and LTKOSN co-existed in the same individual cellular genome, since both vectors were detected in all six subclones (Fig. 1B). Varying copy numbers of ALTKOSN and LTKOSN were detected in subclones #2, #3 and #5. In contrast, only one single copy each of ALTKOSN and LTKOSN vector were detected in parental

LTKOSN.2 VPC and subclones #1, #4, and #10. To estimate the copy numbers of integrated vectors, genomic DNA from each subclone was subjected to BamHl digestion and Southern blot analysis for the integration sites of both ALTKOSN and LTKOSN (Fig. 2).

The BamH\ site is unique in the LTKOSN sequence, and restriction digestion yields two fragments of the LTKOSN vector with adjacent chromosomal sequences. Since the BamH\ site was deleted with the HSVf/c gene in ALTKOSN vector, only one genomic DNA fragment of integrated ALTKOSN vector with flanking chromosomal sequence is observed after

SamHl digestion (Fig. 2B). The copy number of ALTKOSN was approximately six copies per cell in subclone #3, eight copies per cell in subclone 2, while only one copy of LTKOSN exists per cell in these two subclones. An increased copy number of LTKOSN was only observed in subclone #5, which showed approximately nine copies of LTKOSN and approximately 20 copies of ALTKOSN per cell. Therefore, significant changes in the vector copy numbers have occurred between the parental VPC and its subclones.

Deletion of proviral integration site, in subclone #5, a possible deletion of an integration site of LTKOSN vector was observed (Fig. 2A). A 2.2-kb fragment (arrow) containing the 3' portion of proviral LTKOSN fragment with adjacent chromosomal 42

sequences was detected by the Neo*" DMA probe on parental LTKOSN.2 VPC and all of the

subclones except subclone 5 (Fig. 2A, lane 7). A 4.5-kb fragment containing the entire

ALTKOSN vector with adjacent cellular sequences on both ends was detected by the same

Neo'"DNA probe on all of the subclones and parental LTKOSN.2 VPC. In addition to the 4.5-

kb and 2.2-kb bands detected in subclones #2 and #3, other bands that are larger than 2.2-

kb should be ALTKOSN vector, since only one integration site of LTKOSN (3.7-kb) was

detected by HSVtk probe on the same membrane in subclones #2 and #3. This 3.7-kb

fragment (arrow) containing the 5' portion of LTKOSN with adjacent chromosomal

sequences was detected in the LTKOSN.2 VPC and all subclones except subclone #5 (Fig.

2A, lane 7). Taken together, the original LTKOSN vector integration site, present in the

parental LTKOSN.2 VPC and the other subclones, appears absent in subclone #5 (Fig. 2A,

lane 7). Since subclone #5 still contains the same integration site (4.5-kb band)

of ALTKOSN as seen in parental LTKOSN.2 VPC and the other subclones, subclone #5 is a

true subclone derived from parental LTKOSN.2 VPC. We also suspect that Subclone #5

may be derived from subclones #2 and #3, since subclone #5 shares a very similar integration pattern of ALTKOSN with subclones #2 and #3.

Gene expression and packaging of ALTKOSN and LTKOSN. After Northern blot analysis of RNA transcripts from subclones #1, #3, #4, #5 and #10 (subclone #2 was lost to contamination), the results revealed that the RNA transcript ratio of ALTKOSN/LTKOSN in all of the subclones was approximately 2:1 or greater (Fig. 3A). This ratio was observed in subclones #1, #4 and #10 that contain only one copy of each ALTKOSN and LTKOSN vector. This could result from differences in the integration sites or the smaller size of ALTKOSN. Previous reports demonstrated that the internal SV40 promoter, present in

LTKOSN but deleted from ALTKOSN, can interfere with transcription from the 5' LTR (47). 43

Subclone #5, which contains the highest copy nunnber of both ALTKOSN and LTKOSN vectors, showed more abundant vector transcripts than the other subclones. Unexpectedly, these abundant vector transcripts in subclones #1, #3 and #5 were not packaged into virions (Fig. 3B).

Transcriptional inactivation of helper virus. To determine the mechanism for the observed reduction of vector production in subclones #1, #3 and #5, Northern blots of subclones were probed with amphotropic env DNA to evaluate the level of helper virus gene expression (Fig. 3). Spliced env transcripts were detected in all subclones, but significant levels of full-length helper virus transcripts were only observed in subclones #4 and #10

(Fig. 3A) that were previously shown to produce virions containing ALTKOSN and LTKOSN vectors (Fig. 3B). It has been reported that spliced RNA (e.g., env transcript) is more stable than precursor RNA (e.g., full-length helper virus) as a result of increased polyadenylation efficiency on RNA 3' processing (13). Therefore, the overall low level gene expression of helper virus and faster degradation of full length helper virus transcripts may explain why only env transcripts were observed in subclones #1, #3, and #5, but not full-length helper virus (Fig. 3A). These results demonstrated that the reduction of packaged ALTKOSN and

LTKOSN vectors in subclones #1, #3 and #5 correlated closely with low levels of helper virus gene expression.

Superinfection of LTKOSN.2 VPC subclones. To test our hypothesis that the supression of helper virus gene expression by DNA methylation reduces the Env-receptor interference, and then results in vector re-entry (superinfection), LTKOSN.2 VPC subclones were transduced with an EGFP-expression vector, LEIN (see Materials and Methods). The transduction efficiency of LEIN vector on these subclones was determined by FACS analysis of EGFP expression compared to NIH3T3. PA317 and parental LTKOSN.2 VPC 44

(Rg. 4). NIH3T3 cells that do not express Env protein on the cell surface were 21.9% EGFP positive two days after single exposure to LEIN vector. LTK0SN.2 VPC subclones with significant DNA methylation in the 5' LTR region were transduced at cpmparable rates to

NIH3T3 cells (15.7% to 19.3% in subclones #1, #3 and #5). Subclones #4 and #10 that exhibited less methylation of the 5' LTR and more detectable helper virus transcripts were transduced at lower rates of only 0.6% and 1.4% which were comparable to PA317 (0.2%) and parental LTKOSN.2 VPC (0.4%). These results demonstrate that the DNA methylation status and gene expression of helper virus in these subclones correlated closely with the superinfection of LEIN vector. These results also suggested that the increased copy number of LTKOSN and ALTKOSN in subclones #3 and #5 is due to superinfecion. Subclone #1 showed high DNA methylation, helper virus gene inactivation and superinfection of LEIN vector as well as subclones 3 and 5, but did not exhibit increased copy number of vectors.

We suspect that the methylation occurred during or after the subcloning procedure, since we did observe that the titer of subclone #4 was completely reduced to non-detectable by

DNA methylation in one month period of cell culture (Table 2, see below). Subclone #1 therefore did not produce any vector from then and consistently with one single copy of each vector without any vector re-entry.

Methylation status of helper virus and retroviral vectors. DNA methylation has been demonstrated to be associated with the repression of retroviral vector gene expression in different cells and tissues in vitro (12, 21, 23) and in vivo (3). The methylation of CpG within a Smal site by mammalian methylase will prevent Smal digestion (48). The methylation status of the Smal site in 5' LTR has been employed to demonstrate an inverse correlation with the gene expression levels of retroviral vector (3). Therefore, the same 45

analysis was perfomned to detect helper virus DNA methylation (Fig. 5) compared to helper

virus gene expression in VPC subclones (Fig. 3A).

Genomic DNA was first digested with EcoRV and Dral to generate a DNA fragment

convenient for Southern blot analysis. An equal portion of this EcoRVfDral digested DNA

was subjected to Sma\ digestion. Since the gag region is not present in the retroviral vector,

a 428-bp gag probe hybridizes only to the helper virus DNA (Fig. 5A and 5E). All of the

subclones and parental VPC showed a 1.8-kb EcoRV/Dral band (lanes 11-18), which was

reduced to a 1.5-kb band upon Smal digestion if no methylation occurred at the Smal site.

This 1.5-kb band was not detected in either NIH3T3 cells or subclone #5 (Fig. 5A, lanes 1

and 7), but was observed in all other samples. NIH3T3 cells were used to show the

presence of endogenous retroviral elements that hybridized with the gag probe (Fig. 5A,

lanes 1 and 10). After Smal digestion of NIH3T3-derived DNA, a band of approximately 1.8-

kb was generated from endogenous retroviral elements that complicates our analysis (Fig.

5A, lane 1). Therefore, the methylation status of the Smal site was evaluated by the relative

intensities of the observed 1.5-kb bands from Smal/EcoRV/Dral digestions (lanes 2-9)

compared to 1.8-kb bands from EcoRV/Dral digestions (lanes 11-18), which do not overiap

in size with endogenous sequences (Fig. 5A, lane 10). Although 40% of helper virus 5' LTR

in PA317 is methylated (Fig. 5A, lane 2), complete digestion by Smal shows the apparent

absence of methylation of the helper virus 5' LTR In parental LTKOSN.2 VPC (Fig. 5A, lane

3). This suggests that the prior selection of LTKOSN.2 VPC, that was chosen for its high

titer compared to the other VPC clones, is likely the result of both high level expression of

vector and low methylation of helper virus. A high degree (60%) of helper virus methylation

is also found in the 5' LTR of PG13, a GaLV pseudotyped VPC (Fig. 5A, lanes 9 and 18).

This indicates that the methylation of helper virus 5' LTR In PA317 is not a single packaging 46 cell line phenomenon. In contrast to the low level of helper virus methylation in parental

LTKOSN.2 VPC, higher degrees of helper virus DNA methylation were observed in the subclones (Fig. 5A). This high level of methylation was inversely proportional to the helper virus gene expression (Fig. 3A) and subsequent packaged vector titer (Fig. 3A and Table 1).

Subclone 5 has an absence of the 1.5-kb band, which indicates complete methylation of the helper virus 5' LTR (Fig. 5A, lane 7).

The methylation status of the Smal site in the RT region of the helper virus was evaluated by the re-hybridization of this Southem blot with a 1.4-kb env probe (Fig. 5B). In subclones #1, #3 and #5, that had highly methylated helper viral 5' LTR, the methylation level of Smal in the RT region ranged from 46% to 100%. In contrast, only 3% to 10% methylation were observed in parental LTKOSN.2 VPC and subclones #4 and #10. Thus, a strong correlation exists between the observed methylation of the Smal site in 5' LTR and within the RT gene (Fig. 5A and 5B). In addition to the 4.6-kb band in the DNA samples without Smal digestion (lanes 10-18), three additional bands were observed in parental

LTKOSN.2 VPC and all VPC subclones (lanes 2-8; lanes 10-17), but not in NIH3T3 and

PG13 (lanes 1 and 10; lanes 9 and 18). These results demonstrate that the env probe is specific for the env sequence of helper virus pPAMS and does not detect GaLV env or

MoMLV gag-pol from PG13 cells. Furthenmore, at least four copies of env sequence are present in the PA317 genome. These multiple copies of helper virus env sequence were confirmed by Southem blot analysis of PA317 genomic DNA with different methylation- insensitive restriction endonuclease digestions (data not shown). In contrast, hybridization with probes for either the HSVtfc (Fig. 5C) or Neo"" gene (Fig. 5D) reveals no significant methylation within either LTKOSN or ALTKOSN vectors. Extremely low methylation of the

5' LTR of LTKOSN and ALTKOSN vectors was only observed in subclone #5 (Fig. 5D, lane 47

7), which showed 100% methylation in both the helper vlais 5' LTR and RT region (Rg. 5A and 5B, lane 7).

Parental LTKOSN.2 VPC does not show significant methylation of either helper virus or vectors. However, the subclones derived from LTKOSN.2 VPC show various methylation levels of the helper virus but not of their retroviral vectors. These results suggest a sequential timeline of methylation that developed within the helper virus DNA in these five subclones. These data also support our previously theorized cell lineage relationship, based upon the similarity of vector integration patterns observed in Fig. 2A.

Reversal of DNA methylation. To re-activate silenced helper viruses, these subclones were treated with 5-aza-C, a cytidine analog that can inhibit DNA methylase and reverse methylation to restore gene expression (8, 20, 22, 25). After co-culturing with 5-aza-

C for 72 hr, the rescue of vector RNA from these VPC subclones was demonstrated by

RNA slot blot analysis of supernates (Fig. 6). Partial recovery of vector production was also demonstrated by titer assay (Table 1) along with the partial reversal of DNA methylation by

5-aza-C treatment (Fig. 7). The titer of subclones #1, #3 and #5 were increased by two orders of magnitude from <10 cfu/ml to up to 6 X lO^cfu/ml. No significant change was observed in subclones 4 and 10 that exhibited high vector production and less helper virus methylation before 5' aza-C treatment. The most significant reduction of Smal resistance was observed in subclone 5 In which the methylation of the 5' LTR and RT regions was reduced from 100% to about 45% on Southern blot analysis (Fig. 7. lane 4). In this model system, we conclude that the methylation of helper virus Is responsible for the reduction of vector production.

DNA methylation rate of helper virus 5' LTR. To study the DNA methylation rate of helper virus 5' LTR, we examined the DNA methylation of subclone #4, which showed the 48 least DNA methylatlon with most activate gene expression among these subclones. During

31 days of continuous cell culture, DNA methylation of helper virus 5' LTR in subclone #4 increased rapidly from 36% to 100% (Fig. 8A). Therefore the methylation rate of Smal site in helper virus 5' LTR was calculated as high as 2% population per day. The copy numbers of vector were increased from 1 copy (parental LTK0SN.2 VPC and subclone #4 on Day 0) to 5 copies (Day 7) and subsequently to 9 copies (Days 23 and 31) per cell (Fig. SB and

8C). Titer of this subclone #4 was reduced from 1.5 X 10^ cfu/ml (Day 0) to no detectable titer by Day 23 (Table 2). The reduction in vector titer correlated with increasing inactivation of helper virus gene expression (Fig. 9A). However, vector gene expression still remained at high levels (Fig. 9B). Only minimal DNA methylation (10%) of 5' LTR was observed in

LTKOSN vector and no detectable DNA methylation was noted in the ALTKOSN vector (Fig.

80).

DISCUSSION

In this study, the genetic stability of retroviral vectors and a VPO genome was examined in LTKOSN.2 VPO and derived subclones. We also compared the methylation status of helper virus and vectors in these individual subclones. Our results demonstrate that DNA methylation occurred of helper viral sequences rather than vector in subclones derived from LTKOSN.2 VPO (Fig. 5). The increase in methylation of the helper virus (Fig. 5 and Fig. 8) correlated directly with the inactivation of helper virus gene expression (Fig. 3 and Fig. 9), which resulted In a loss of vector production. In contrast, vector gene expression was consistent and no significant methylation of vector sequences was detected. Therefore, we conclude that the key limitation for vector production in the

LTKOSN.2 VPO is the inactivation of helper virus gene expression by DNA methylation. 49

This conclusion was supported by the partial restoration of vector production (Table 1) and the reversal of DNA metiiylati'on (Fig. 7) from VPC subclones treated with 5-aza-C. In previous studies, de novo DNA methylation has been suggested to cause transcriptional inacti'vation of retroviral vectors in infected embryonic cells (3,12, 21). Thus far, DNA methylation had not been a focus of investigation with regard to the genetic instability of retroviral VPC that are designed for continuous vector production. The methylation of helper virus sequence in a VPC may initiate a cascade of more complex events compared to the inactivation of vector gene expression noted in infected embryonic cells. The methylation of helper virus 5' LTR resulting In the genetic instability of VPC is proposed by this study to occur in the following sequence: (i) repression of helper virus 5' LTR transcriptional activity by methylation reduces the production of vector, (ii) decreased Env protein synthesis reduces Env-receptor interference and allows vector superinfection to cause increased vector copy number in VPC, and (iii) random integration of vectors caused by superinfection results in a more fluid VPC cellular genome.

The reason for the preference for methylation of the helper virus compared to the retroviral vectors is not clear and requires further studies. However, chromosomal location has been suggested as a key factor for facilitating methylation (12, 19). In PA317, four copies of env sequences of pPAMS helper virus were observed (Fig. 5B). Further restriction analysis of these helper virus sequences demonstrates that the sequences are located at different chromosomal sites (data not shown). The probability that all four copies of helper virus sequences are located within a hypermethylation region of PA317 seems remote. No significant methylation of vectors was observed in subclones 3 and 5, even though these subclones contain multiple copies of vectors, which were distributed into different integration sites. About 100 copies of integrated MoMLV provirus have been observed in infected 50

murine embryonic carcinoma cells and these proviruses were hypermethylated with no

significant gene expression (54). Presumably, these 100 copies of provirus were not all

located In hypermethylation regions. Therefore, the preference of methylation for helper

virus is probably related to both helper virus sequence and chromosomal location. The idea

that sequence has an important role is also suggested by the findings that the presence of

retrotransposon and provirus sequences enhance the de novo DNA methylation of adjacent

sequences (11, 16). Most recently, experiments provided evidence that DNA methylation

acts as a defense system against retrotransposons invading the mammalian genome (43).

infection of retrovirus, such as HIV, increased the overall DNA methyltransferase gene

expression and activity, which resulted in supression of cytokine gene expression for

evading immune surveillance (29). The spreading of de novo methylation from a provirus into adjacent cellular sequences was observed in either the 5' or 3' region of the integration site in Mov-derived mice carrying a MoMLV provirus in a distinct chromosomal location (16).

In the pPAM3 LTR, the significant methylation observed could be the result of spreading of methylation from adjacent gag sequences, since the LTR in either LTKOSN or ALTKOSN vector did not initiate any significant methylation. However, we did not rule out the role of chromosomal position and preferred DNA methylation (12). These subclones of VPC may provide a useful model to study DNA methylation preferences that are key in host-retrovinjs interactions (11, 16, 17, 29, 43).

One may argue that G418 selection of VPC could have eliminated VPC cells with methylated vectors. After the establishment of LTK0SN.2 VPC, VPC were not maintained on G418 selection on a continual basis. Other LTKOSN parental lines only require a single copy of the Neo"^ gene for G418 resistance. Subclones #3 and #5 contained multiple copies of LTKOSN and ALTKOSN. Even if partial methylation of the vector as noted in the helper 51

virus had occurred, adequate Neo*^ gene expression would nevertheless provide G418

resistance. However, almost no methylation was observed in either LTKOSN or ALTKOSN.

Extrenriely low degree of nnethylation was only observed in subclone #5 which contains

approximately 29 copies of integrated vectors, and this could be contributed by the random

integration of these vectors into hypermethylation regions of cellular genome. In contrast to

the moderate degree of methylation (40%) of the helper virus 5' LTR in PA317, the helper

virus in parental LTKOSN.2 VPC did not show significant methylation of Smal sites in the 5'

LTR. This demonstrates that a diverse range of methylation status exists within PA317 cells

and the selection of the highest titer LTKOSN VPC (LTKOSN.2) from among the other 39

clones may have selected for a clone with hypomethylated helper viruses.

It has been well documented that Env-receptor interference is important to reduce

the risk of superinfection (30, 32, 53, 59). Recent data demonstrated that at least four

copies of retrovirus per cell are required for sufficient Env protein synthesis to interfere with

the cellular receptor (41). Superinfection experiments on these LTKOSN.2 subclones (Fig.

4) demonstrated that DNA methylation (Fig. 5) and helper virus mRNA level (Fig. 3) are

highly correlated with the reduction of Env-receptor interference. The reduction of env gene

expression by DNA methylation would logically result in the re-entry of multiple vectors into

subclones #3 and #5. Subclone #1 may represent a transitional state between the parental

LTKOSN.2 and subclone #3 since only one copy of each LTKOSN and ALTKOSN vector is present while helper virus gene expression is low.

DNA methylation in subclones #4 and #10 (38% and 50%, respectively) and PA317

(40%) is significantly higher than in parental LTKOSN.2 cells (0%). However, the resistance to superinfection in subclones #4, #10 and PA317 is still comparable to that observed in parental LTKOSN.2 cells (0%, Fig. 5). A significant increase in superinfection was only 52 observed in subclones when at least 60% methylation occurred of the helper virus 5' LTR.

This suggests a possible threshold effect. It has been dennonstrated that the amount of Env product per cell, the interference and virion release did not increase lineariy but increased abruptly once an infected cell reached the threshold of proviral copy number which correlated with mRNA level (41). This is due to the fact that oligomerization of Env protein is essential for functional virus assembly and interference (2, 14, 41). Since the concentration of Env protein determines the kinetics of oligomerization, reduction of helper virus mRNA level by DNA methylation resulted in low Env protein concentration and oligomerization, and, therefore significantly decreased the Interference and virion release. Compared to the previously reported high level of resistance to the superinfection of PA317, which was transduced with the vector collected from another PA317 VPC (32), the apparent re-entry frequency of vectors in LTKOSN.2 VPC is relatively high (50% of population, three out of six subclones. Fig. 1). This superinfection was enhanced by both the DNA methylation of helper virus and the presence of extensive vector exposure in cell culture. It has also been observed that HIV provirus was accumulated in infected cells to multiple copies by superinfection of 70% of cells in the culture over time (44).

Amphotropic murine leukemia virus (Am-MLV) vectors are used widely for human gene transfer. However, most Am-MLV VPCs are derived from murine cells and are susceptible to vector re-entry. The consequence of vector re-entry is the import of active RT enzyme which may lead to the RCR formation and the random integration of additional vectors that can interrupt functional genes or activate proto-oncogenes (9, 15, 24, 51). Our results infer that DNA methylation decreases the Env-receptor interference, which results in vector re-integration and subsequent further instability of the VPC genome. Recently, progress in this direction has been made: PG13 VPC were established using NIH3T3 cells 53 for packaging xenotropic GaLV-pseudotyped vector. GaLV packaged vector will not re-infect

PG13 cells that lack the receptor for GaLV envelope (33, 60). PG13 VPC may diminish the vector re-entry problem; however, helper virus gene expression in PG13 is still not monitored and selected by a selection marker. Our data with PG13 show significant DNA methylation (60%) of helper virus 5' LTR (Fig. 5A, lanes 9 and 18). The observations in this study strongly suggest that improvement of the helper virus gene expression to overcome

DNA methylation in VPC is important for sustained, stable vector production. Treatment with

5-aza-C on LTK0SN.2 subclones did not completely restore the titer (Table 1). The effect of

5-aza-C treatment is transient (8) and can not be applied to long term cell culture since 5- aza-C is toxic to the treated cells (20). Several alternative strategies that ensure sustained vector gene expression in target cells might improve the gene expression of helper virus in

VPC. Insertion of a demethylation fragment of murine Thy-1 gene in front of 5' LTR can inhibit methylation (4, 55) and may be useful for helper virus. Another strategy is to ligate an internal ribosome entrance site [IRES, (18)] with a selection marker gene downstream of the helper virus so that drug selection would ensure active enhancer/promoter function of helper virus for high vector packaging and preventing superinfection by enhancing Env-receptor interference (61).

ACKNOWLEDGMENTS

We thank Dr. John Levy, Dr. A. Dusty Miller and Dr. Tatiana Seregina for helpful discussions. We also thank Ginger Dreifurst and Julie Seiwert for technical assistance with flow cytometry. This work was supported in part by a grant from the Iowa Health System,

Des Moines, Iowa. W.-B. Young is a recipient of a competitive research fellowship from

Molecular Cellular and Developmental Biology Program, Iowa State University, Ames, Iowa. 54

REFERENCES

1. Anderson, W. F., G. J. McGarrity, and R. C. Moen. 1993. Report to the NIH

Recombinant DNA Advisory Committee on murine replication-competent retrovirus

(RCR) assays (February 17, 1993). Hum Gene Ther. 4:311-21.

2. Bernstein, H. B., S. P. Tucker, S. R. Kar, S. A. McPherson, D. T. McPherson, J. W.

Dubay, J. Lebowitz, R. W. Compans, and E. Hunter. 1995. Oligomerization of the

hydrophobic heptad repeat of gp41. J Virol. 69:2745-50.

3. Chailita, P. M., and D. B. Kohn. 1994. Lack of expression from a retroviral vector after

transduction of murine hematopoietic stem cells is associated with methylation in vivo.

Proc Natl Acad Sci USA. 91:2567-71.

4. Chailita, P. M., D. Skelton, A. el-Khoueiry, X. J. Yu, K. Weinberg, and O. B. Kohn.

1995. Multiple modifications in cis elements of the long terminal repeat of retroviral

vectors lead to increased expression and decreased DNA methylation in embryonic

carcinoma cells. J Virol. 69:748-55.

5. Cone, R. D., and R. C. Mulligan. 1984. High-efficency gene transfer into mammalian

cells: generation of helper-free recombinant retrovirus with broad mammalian host

range. Proc Natl Acad Sci. 81:6349-6353.

6. Cormack, B. P., R. H. Vaidivia, and S. Falkow. 1996. FACS-optimized mutants of the

green fluorescent protein (GFP). Gene. 173:33-8.

7. Donahue, R. E., S. W. Kessler, D. Bodine, K. McDonagh, C. Dunber, S. Goodman,

B. Agricola, E. Byrne, M. Raffeld, R. Moen, J. Bacher, K. M. Zsebo, and A. W.

Nienhuis. 1992. Helper virus induced T cell lymphoma in nonhuman primates after

retroviral mediated gene transfer. J. Exp. Med. 176:1125-35. 8. Gram, G. J., S. D. Nielsen, and J. E. Hansen. 1998. Spontaneous silencing of

humanized green fluorescent protein (hGFP) gene expression from a retroviral vector by

DNA methylation [In Process Citation], J Hematother. 7:333-41.

9. Gray, D. A. 1991. Insertional mutagenesis: neoplasia arising from retroviral integration.

Cancer Invest. 9:295-304.

10. Haas, J., E. 0. Park, and B. Seed. 1996. Codon usage limitation in the expression of

HIV-1 envelope glycoprotein. Curr Biol. 6:315-24.

11. Kasse, A., and W. A. Schulz. 1994. Enhancement of reporter gene de novo

methylation by DNA fragments from the alpha-fetoprotein control region. J Biol Chem.

269:1821-6.

12. Hoeben, R. 0., A. A. Migchielsen, R. 0. van der Jagt, H. van Ormondt, and A. J.

van der Eb. 1991. Inactivation of the Moloney murine leukemia virus long terminal

repeat in murine fibroblast cell lines Is associated with methylation and dependent on its

chromosomal position. J Virol. 65:904-12.

13. Huang, M. T., and 0. M. Gorman. 1990. Intervening sequences increase efficiency of

RNA 3' processing and accumulation of cytoplasnnic RNA. Nucleic Acids Res. 18:937-

47.

14. Hunter, E. 1994. Macromolecular interactions in the assembly of HIV and other

retroviruses. Semi Virol. 5:71-83.

15. Ihle, J. N., B. Smith-White, B. Sisson, D. Parker, D. G. Blair, A. Schuitz, C. Kozak,

R. D. Lunsford, D. Askew, Y. Weinstein, and R. J. Isfort. 1989. Activation of the c-H-

ras proto-oncogene by retrovirus insertion and chromosomal rearrangement in a

Moloney leukemia virus-induced T-cell leukemia. J Virol. 63:2959-66. 16. Jahner, D., and R. Jaenisch. 1985. Retrovirus-induced de novo methylation of flanking

host sequences correlates with gene inactivity. Nature. 315:594-7.

17. Jahner, D., H. Stuhimann, C. L. Stewart, K. Harbers, J. Lohler, I. Simon, and R.

Jaenisch. 1982. De novo methylation and expression of retroviral genomes during

mouse embryogenesis. Nature. 298:623-8.

18. Jang, S. K., T. V. Pestova, C. U. Heilen, G. W. Witherell, and E. Wimmer. 1990. Cap-

Independent translation of picomavlrus RNAs: structure and function of the internal

ribosomal entry site. Enzyme. 44:292-309.

19. Jolly, D. J., R. C. Willis, and T. Friedmann. 1986. Variable stability of a selectable

provirus after retroviral vector gene transfer Into human cells. Mol Cell Biol. 6:1141-7.

20. Juttermann, R., E. Li, and R. Jaenisch. 1994. Toxicity of 5-aza-2'-deoxycytidine to

mammalian cells is mediated primarily by covalent trapping of DNA methyltransferase

rather than DNA demethylation. Proc Natl Acad Sci USA. 91:11797-801.

21. Kempler, G., B. Freitag, B. Berwin, O. Nanassy, and E. Barklis. 1993.

Characterization of the Moloney murine leukemia virus stem cell- specific repressor

binding site. Virology. 193:690-9.

22. Kuriyama, S., T. Sakamoto, M. Kikukawa, T. Nakatani, Y. Toyokawa, H. Tsujinoue,

K. Ikenaka, H. Fukui, and T. Tsujii. 1998. Expression of a retrovirally transduced gene

under control of an Internal housekeeping gene promoter does not persist due to

methylation and is restored partially by 5-azacytidlne treatment. Gene Ther. 5:1299-

1305.

23. Laker, 0., J. Meyer, A. Schopen, J. Friel, 0. Heberlein, W. Ostertag, and C.

Stocking. 1998. Host cis-mediated extinction of a retrovirus permissive for expression

In embryonal stem cells during differentiation. J Virol. 72:339-48. 57

24. Lazo, P. A., and P. N. Tsichiis. 1988. Recombination between two integrated

proviruses, one of which was inserted nearc-myc in a retrovirus-induced rat thymoma:

impfications for tumor progression. J Virol. 62:788-94.

25. Lengauer, C., K. W. Kinzler, and B. Vogelstein. 1997. DNA methylation and genetic

instability in colorectal cancer cells. Proc Natl Acad Sci USA. 94:2545-50.

26. Link, C. J., Jr., E. M. Kolb, and R. R. Muldoon. 1995. Preliminary in vitro efficacy and

toxicities studies of the herpes simplex thymidine kinase gene system for the treatment

of breast cancer. Hybridoma. 14:143-7.

27. Link, C. J., Jr., D. Moorman, T. Seregina, J. P. Levy, and K. J. Schabold. 1996. A

phase 1 trial of in vivo gene therapy with the herpes simplex thymidine kinase/ganciclovir

system for the treatment of refractory or recurrent ovarian cancer. Hum Gene Ther.

7:1161-79.

28. Markowitz, D., S. Goff, and A. Bank. 1988. A safe packaging line for gene transfer:

separating viral genes on two different plasmids. J. Virol. 62:1120-4.

29. Mikovits, J. A., H. A. Young, P. Vertino, J. P. Issa, P. M. Pitha, S. Turcoski-

Corrales, D. D. Taub, C. L. Petrow, S. B. Baylin, and F. W. Ruscetti. 1998. Infection

with human immunodeficiency virus type 1 upregulates DNA methyltransferase,

resulting in de novo methylation of the gamma interferon (IFN-gamma) promoter and

subsequent downregulation of IFN-gamma production. Mol Cell Biol. 18:5166-77.

30. Miller, A. D. 1996. Cell-surface receptors for retroviruses and implications for gene

transfer. Proc Natl Acad Sci USA. 93:11407-13.

31. Miller, A. D., and C. Buttimore. 1986. Redesign of retrovirus packaging cell lines to

avoid recombination leading to helper virus production. Mol Cell Biol. 6:2895-902. 58

32. Miller, A. D., and F. Chen. 1996. Retrovirus packaging cells based on 10A1 murine

leukemia virus for production of vectors that use multiple receptors for cell entry. J Virol.

70:5564-71.

33. Miller, A. D., J. V. Garcia, N. Von Shur, C. M. Lynch, C. Wilson, and M. V. Eiden.

1991. Construction and properties of retrovirus packaging cells based on gibbon ape

leukemia virus. J Virol. 65:2220-4.

34. Miller, A. D., and G. J. Rosman. 1989. Improved retroviral vectors for gene transfer

and expression. Biotechniques. 7:980-2, 984-6, 989-90.

35. Miller, A. D., D. R. Trauber, and C. Buttimore. 1986. Factors involved in production of

helper virus-free retrovirus vectors. SomatCell Mol Genet. 12:175-83.

36. Moolten, F. L. 1986. Tumor chemosensitivity confenred by inserted herpes thymidine

kinase genes: paradigm for a prospective cancer control strategy. Cancer Res. 46:5276-

81.

37. Muenchau, D. D., S. M. Freeman, K. Cornetta, J. A. Zwiebei, and W. F. Anderson.

1990. Analysis of retroviral packaging lines for generation of replication- competent

virus. Virology. 176:262-5.

38. Muldoon, R. R., J. P. Levy, S. R. Kain, P. A. Kitts, and C. J. Link, Jr. 1997. Tracking

and quantitation of retroviral-mediated transfer using a completely humanized, red-

shifted green fluorescent protein gene. Biotechniques. 22:162-7.

39. Nelson, D. M., J. J. Wahlfors, L. Chen, M. Onodera, and R. A. Morgan. 1998.

Characterization of diverse viral vector preparations, using a simple and rapid whole-

virion dot-blot method. Hum Gene Ther. 9:2401-5. 59

40. Nielsen, C. S., D. W. Moorman, J. P. Levy, and C. J. Link, Jr. 1997. Herpes simplex

thymidine kinase gene transfer is required for complete regression of murine colon

adenocarcinoma. Am Surg. 63:617-20.

41. Odawara, T., Il/I. Oshima, K. Doi, A. Iwamoto, and H. Yoshikura. 1998. Threshold

number of provirus copies required per cell for efficient virus production and interference

in moloney murine leukemia virus- infected NIH 3T3 cells. J Virol. 72:5414-24.

42. Oldfield, E. i-i., Z. Ram, K. W. Culver, R. M. Blaese, H. L. DeVroom, and W. F.

Anderson. 1993. Gene therapy for the treatment of brain tumors using intra-tumoral

transduction with the thymidine kinase gene and intravenous ganciclovir. Hum Gene

Ther. 4:39-69.

43. O'Neill, R., M. O'Neill, and J. A. Graves. 1998. Undermethylation associated with

retroelement activation and chromosome remodelling in an interspecific mammalian

hybrid. Nature. 393:68-72.

44. Ott, D. E., S. M. Nigida, Jr., L. E. Henderson, and L. O. Arthur. 1995. The majority of

cells are superinfected in a cloned cell line that produces high levels of human

Immunodeficiency virus type 1 strain MN. J Virol. 69:2443-50.

45. Otto, E., A. Jones-Trower, E. F. Vanin, K. Stambaugh, S. N. Mueller, W. F.

Anderson, and G. J. McGarrity. 1994. Characterization of a replication-competent

retrovirus resulting from recombination of packaging and vector sequences. Hum Gene

Ther. 5:567-75.

46. Pear, W. S., G. P. Nolan, M. L. Scott, and D. Baltimore. 1993. Production of high-titer

helper-free retroviruses by transient transfection. Proc Natl Acad Sci USA. 90:8392-6. 47. Robbins, P. B., X. J. Yu, D. M. Skelton, K. A. Pepper, R, M. Wasserman, L. Zhu, and

D. B. Kohn. 1997. Increased probability of expression from modified retroviral vectors in

embryonal stem cells and embryonal carcinoma cells. J Virol. 71:9466-74.

48. Roberts, R. J. 1990. Restriction enzymes and their isoschizomers. Nucleic Acids Res.

18 Suppl:2331-65-

49. Roth, J. A., and R. J. Cristiano. 1997. Gene therapy for cancer: what have we done

and where are we going? J Natl Cancer Inst. 89:21-39.

50. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory

manual., p. 9.14 - 9.23, Chapter 9: Analysis and cloning of eukaryotic genomic DNA,

2nd. ed. Cold spring Harbor Laboratory Press, Cold Spring Harbor, N. Y.

51. Schwartz, J. R., S. Duesberg, and P. Ouesberg. 1995. DNA recombination is

sufficient for retroviral transduction. Proc. Natl. Acad. Sci. USA. 92:2460-4.

52. Seperack, P. K., M. 0. Strobel, D. J. Corrow, N. A. Jenkins, and N. G. Copeiand.

1988. Somatic and germ-line mutations rates of the retrovirus-induced dilute coat-color

mutation of DBA mice. Proc. Natl. Acad. Sci. USA. 85:189-192.

53. Sommerfelt, M. A., and R. A. Weiss. 1990. Receptor interference groups of 20

retroviruses plating on human cells. Virology. 176:58-69.

54. Stewart, 0. L., H. Stuhlmann, D. Jahner, and R. Jaenisch. 1982. De novo

methylation, expression, and infectivlty of retroviral genomes introduced into embryonal

carcinoma cells. Proc Natl Acad Sci USA. 79:4098-102.

55. Szyf, M., G. Tanigawa, and P. L. McCarthy, Jr. 1990. A DNA signal from the Thy-1

gene defines de novo methylation patterns in embryonic stem cells. Mol Cell Biol.

10:4396-400. 61

56. Teyssier, J. R., J. Benard, D. Ferre, J. Da Silva, and L. Renaud. 1989. Drug-related

chromosomal changes in chemoresistant human ovarian carcinoma cells. Cancer Genet

Cytogenet. 39:35-43.

57. Vanin, E. F., M. Kaloss, C. Broscius, and A. W. Nienhuis. 1994. Characterization of

replication-competent retroviruses from nonhuman primates with virus-induced T-cell

lymphomas and observations regarding the mechanism of oncogenesis. J Virol.

68:4241-50.

58. Varmus, H. E., N. Quintreli, and S. Ortiz. 1981. Retroviruses as mutagens: insertion

and excision of a nontransforming provirus alter expression of a resident transforming

provirus. Cell. 25:23-36.

59. Weiss, R. A., and 0. S. Tailor. 1995. Retrovirus receptors. Cell. 82:531-3.

60. Wilson, C., M. S. Reitz, H. Okayama, and M. V. Eiden. 1989. Formation of infectious

hybrid virions with gibbon ape leukemia virus and human T-cell leukemia virus retroviral

envelope glycoproteins and the gag and pol proteins of Moloney murine leukemia virus.

J Virol. 63:2374-8.

61. Young, W.-B., and C. J. Link, Jr. 1999. Chimeric retroviral helper virus and

picomavirus IRES sequence to eliminate DNA methylation for improved retroviral

packaging cells. In preparation. 62

TABLE 1. Titer of LTKOSN.2 sub-clones treated with S-aza-C

Target cell

5-aza-C NIH3T3 A375 IGROV

+ 4X10^ 5X10^ 4X 10^ Sub-clone 1 1 X10^ 2X10° 0

-1- 4X10^ 6X10^ SXIO"" Sub-clone 3 0 0 0

+ 5X10® 4X10® 1 X 10® Sub-clone 4 2.2X10® 4X10® axio'^

+ 2X10^ 3X10^ 1.8 X 10^ Sub-clone 5 0 0 0

+ 3X10"^ 1 X10® 1.2X10"^ Sub-clone 10 4X10"^ 2X10® 1.6 X lO'^

TABLE 2. Decreased titer of LTKOSN.2 VPC subclone #4 in continuous cell culture

Subclone #4

LTKOSN.2 Day 0 Day 7 Day 23 Day 31

1.1 X 10® 1.5X10® 7X10"^ 0 0 FIG. 1. Increased copy number of LTKOSN and ALTKOSN in LTKOSN.2 subclones.

(A) Schematic diagram of LTKOSN and ALTKOSN vectors. LTKOSN contains a HSVtfc gene, which was cloned Into the EcoRI site of LXSN. extended packaging signal region;

Neo, neomycin phosphate transferase gene; SV, promoter sequence of simian virus 40 early gene. B, SamHI; Bp, Bpm\; E, EcoRI; H, Hindlll; K, Kpni. (B) Genomic DNA was extracted from GP+E86 transfected with LTKOSN (lane 1), parental LTKOSN.2 (lane 3) and

Its derived subclones #1, #5 and #10 (lanes 4-9). PA317 derived DNA (lane 2) was used as a confrol. Kpnl digests within both U5 regions of LTR and releases a full-length provlral

LTKOSN without the 5 LTR (4.0-kb) and a ALTKOSN without the 5 LTR (2.5-kb). Integrated

LTKOSN and ALTKOSN were detected from parental LTKOSN.2 and all subclones, but not

PA317 cells. Only LTKOSN vector was detected in LTKOSN-transfected GP+E86 cells.

Note the relative variance in ALTKOSN copy number of different subclones compared to

LTKOSN. (C) Hybridization with a BsfKl fragment probe (0.7-kb, located at gag coding region) that detects a 2.5-kb fragment of Kpnl digested endogenous retroviral sequences showed similar loading of DNA samples. 64

LTKOSN vector

4.0 kb

LTR HSVtk Neo LTR

TK probe Neo probe

ALTKOSN 2.5 kb K K LTR Neo f-fTjR

CO S

Sub-clones 1 2 3 4 5 10

- LTKOSN

-ALTKOSN 2.0-

3.0-

2.0-

1 23456789 FIG. 2. Genetic instability of integrated LTKOSN in VPC genome. (A) Genomic

DNA extracted from PA317 (lane 1), LTK0SN.2 VPC (lane 2) and its subclones (lanes 3-

8) were subjected to SamHI digestion and detected by Neo*^ and/or HSVf/c probes in

Southern analysis. Only two integration sites were detected on parental LTKOSN.2, subclones #1, #4, and #10, but subclones #2, #3 and #5 showed more integration sites.

The arrows indicate the positions of integrated LTKOSN vector. Note the absence of these bands from subclone 5 on botii Neo"^ and HSVtfc probe hybridizations. (B) Schema of LTKOSN and ALTKOSN integration sites showing the locations of SamHI sites on the flanking chromosomal sequences. The SamHI site on ALTKOSN is deleted, therefore, digestion with SamHI excises the entire ALTKOSN at the 5' and 3' flanking chromosomal

DNA sequence (4.5-kb) of the integration site. SamHI digestion excises the LTKOSN backbone 5' to the SV40 promoter and results in two fragments of 3.7-kb and 2.2-kb. which includes 5' and 3' junctional chromosomal DNA sequences, respectively. 66

CVI CVJ

CO g — Sub-clones 1 — Sub-clones 1 H: 2 3 4 5 10 2 3 4 5 10 kb

6.0- 5.0- 4.0-

3.0-

2.0-

1 2 3 4 5 6 7 8 2 3 4 5 6 7 8 Probes Neo HSVtk

B B B B LfRl-^^^-r~HSVtk ks^ Neo hlTfR LTKOSN l-<- 3.7 kb 2.2 kb

B B _L Neo hTlTR ALTKOSN

4.5 kb FIG. 3. Differential gene expressions of vectors and lieiper virus in LTKDSN.2

VPC subclones. (A) Total RNA was extracted from LTKOSN.2 VPC subclones (lanes 3-

7) and detected by Neo^ env and human glyceraldehyde 3-phosphate dehydrogenase

(GAPDH) cDNA probe. RNA transcripts of LTKOSN and ALTKOSN are 4.0-kb and 2.5- kb in size, respectively. RNA extracted from NIH3T3 (lane 8) was used as a negative confrol. Hybrid'czation with env probe detected helper virus gene expression. Compared to the spliced env transcript observed in all of subclones, full-length MoMLV transcript (gag- pol-env) was only detected in LTKOSN.2 VPC (lane 2) and subclones #4 and #10 (lanes

5 and 7) and barely seen in otiier subclones and PA317 cells. (B) Viral RNA extracted from viral particles was detected by Neo"" probe and showed that LTKOSN and

ALTKOSN transcripts were detected only in the supemates collected from subclones #4 and #10. Thus, despite abundant LTKOSN and ALTKOSN transcripts in the VPC, little or no vector was packaged and released from subclones #1, #3 and #5. 68

— gag-pol-env

env — env

GAPDH

1 2 3 4 5 6 7 8

CM Sub-clones

LTKOSN ALTKOSN

1 2 3 4 5 6 7 FIG. 4. Superinfection susceptibility of LTK0SN.2 VPC subclones. Cells from

NIH3T3. PA317, parental LTKOSN.2 and LTK0SN.2 VPC subclones were seeded at 5 X

10® cells per well on 12-wells plate and exposed to 0.5 ml supemate of EGFP-expressing vector, LEIN (3.5 X 10® cfu/ml, see Materials and Methods). FACS analysis of EGFP expression was performed 2 days after the single exposure of LEIN vector. The average of three transduction rates of each target cell line was, NIH3T3, 21.9%; PA317, 0.2%;

LTKOSN.2, 0.4%; subclone #1,18.4%; subclone #3, 15.7%; subclone #4, 0.6%; subclone

#5, 19.3%; and subclone #10, 1.4%. Error bars were calculated from triplex transduction data of each target cell line. 70

Target cell line FIG. 5. DNA methylation statuses of helper virus and vectors. Genomic DNA was sequentially digested with Dra\ and

EcoRV, precipitated with ethanol, resolved in sterile water, divided into two equal portions for Sma\ digestion (lanes 1-9) or without Smal digestion (lanes 10-18). (A) Hybridization of gag DNA probe to detect helper virus 5 LTR. Sma\ digestion reduced the 1.8-kb band (lanes 11-18) to 1.5-kb (lanes 2-9). DNA from NIH3T3 cells was used to demonstrate the presence of endogenous retroviral elements (lanes 1 and 10). Since a 1.8-kb band was generated from endogenous retroviral element after

Smal digestion (lane 1), the values for Smal resistance were measured by densitometry as the relative intensities of the 1.8-kb bands without Smal digestion (lanes 10-18), and 1.5-kb bands generated after Smal digestion (lanes 1-9). Note the absence of helper virus methylation in parental LTK0SN.2 (lane 3) and completely methylated helper virus 5' LTR in subclone 5 (lane 7). (B)

Re-hybridization of env DNA probe for the Smal methylation status in helper virus RT region. Smal resistance was measured as the relative intensities of the 4.6-kb and (3.3-kb + 4.6-kb) bands generated after Smal digestion. Note multiple copies of helper virus env sequences were detected in PA317 and derived cells. (C) HSVf/c DNA probe was used to detect the 5 LTR of LTKOSN vector. Smal digestion reduced the 1.5-kb fragment to 1.2-kb. The varied signal intensities are secondary to the copy number of

LTKOSN vector in each subclone. (D) Neo' probe was used to detemriine the Smal methylation status in both the HSVtk gene of

LTKOSN and the 5 LTR region of ALTKOSN. Smal reduced the 2.4-kb fragment of ALTKOSN to 2,1-kb, and the 2.1-kb band of

LTKOSN to 1.5-kb. (E) Schemes of the helper virus and vectors showing the locations of restriction enzyme sites and the probes used for the methylation analysis. D, Oral; E, EcoRV; S, Smal; AAA, SV40 poly (A) signal. Drawings are not to scale. E + Smal - Smal Helper virus pPAM3

CM CM EES D D S E e N. m £2 h- « « r: 5 Sub-clones J2 « £ 9 Sub-clones $2 1 aaa ' ' do! —I env"•^AAA Z^bl34510 az^5l34510 Q; 1.8 kb 1,5 kb .'Win!** gag probe 1-5- • „ 4.6 kb 3.3 kb env probe %Smal- gggggggg Resistance to

LTKOSN vector • -* 4.6 - ** -- m "» M M' M M- EES EE S EES « J" T^' -g M •• im hjRl—I Hk(/lk 'm Re5 NJ 3,3 - 't 1.5 kb 1.2 kb % Smal- o) ^ m ^ o b CO HSS/tk probe Resistance ^ r- o - 2,1 kb - 1.5 kb 1,5 - mm HI" Neo probe 1,2 -

ALTKOSN vector

2.4 •> 2.-1 - IW Wlli • mm9\ EES EES ,M I, , , , >I-L E0—QioZZKSE 1,5- - 2,4 kb - 2.1 kb 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1718 Neo probe 73

CO I— j— Sub-clones ~| CO 5-aza-C

FIG. 6. Treatment with 5-aza-C partly restores the vector production ability in subclones. Cell cultures were co-incubated with 5 ^iM of 5-aza-C for 72 hrs. Supernate (200 fil) collected from each of the cell culture with or without 5-aza-C treatment was loaded onto the slot blot apparatus and UV-cross-linked onto a nylon membrane. Neo"" probe was used to detect the LTKOSN and ALTKOSN vector transcripts to quantify the vector production ability of each subclone, supernate collected from NIH3T3 was used as a negative control. FIG. 7. Treatment with 5-aza-C reverses the DNA methylation of helper virus.

Genomic DNA digestion and hybridization with the gag (A) and env(B) DNA probes were

the same as described in Fig. 5 to evaluate the Smal methylation status in the helper virus

5 LTR and RT regions. Note the new appearance of a 1.5-kb band (A) and a 3.3-kb band

(B) in subclone 5 (lane 4), that were not seen in Fig. 5 (lane 7) before 5 aza-C treatment. 75

— + Smal — Smal —

I— Sub-clones —| [— Sub-clones —|

% Smal- ^ ^ ^ ^ Resistance £8 S co ^ ^

1 2345678910 % Smal- ^ ^ ^ ^ 5^ Resistance lOcm oaO) •«-LO Oto O FIG 8. Increased DNA methylatlon In helper virus of subclone #4 overtime. Genomic

DNA was first digested with Dral and EcoRV, and divided into two equal portions for Smal digestion as described in Figure 5. (A) Hybridization of gag DNA probe to detect helper virus

5 LTR. Smal digestion reduced the 1.8-kb band (even number lanes 4-14) to 1.5-kb (odd number lanes 3-13). Smal resistance was calculated as previous in Figure 5. (B) HSVtk

DNA probe was used to detect the 5 LTR of LTKOSN vector. Smal digestion reduced the

1.5-kb fragment to 1.2-kb. The values for Smal resistance were calculated as described In

Figure 5. The varied signal Intensities are secondary to increasing copy numbers of

LTKOSN vector over time. Copy number was estimated by comparing relative intensity of

HSVtk signals and intensities detected by SsfiCI fragment probe (D) to standardize loading.

Increased copy numbers were observed on Days 7 (5 copies, lane 10), 23 and 31 (9 copies, lane 12 and 14), while only one copy In parental LTKOSN.2 and subclone #4 on Day 0

(lanes 6 and 8). (C) Neo"" probe was used to determine the Smal methylatlon status in both the HSVf/c gene of LTKOSN and the 5 LTR region of ALTKOSN. (D) A 0.68-kb DNA fragment firom gag gene digested by BsfK\ was used as a probe to detect a 1.2-kb band of endogenous retroviral sequence to demonstrate equivalent DNA loading In paired samples of Smal digestion. 77

subclone #4 CVJ

% Smal- 54% 15% 36% 57% 82% 100% Resistance

% Smal- 0% 0% 0% 7% 10% Resistance

2.4 - 2.1 '

1.5 - FIG. 9. Transcription activity of helper virus and retroviral vectors in LTK0SN.2 VPC subclone #4. (A) Northern blot analysis of cellular RNA extracted from LTKOSN.2 subclone

#4 at different time points was hybridized with env probe. Unspliced MoMLV transcript (gag- pol-env) were detected at different sizes as a result of the presence of at least four different integrated pPAM3 detected in PA317 with different adjacent sequences. (B) Gene expression of LTKOSN and ALTKOSN vectors. Re-hybridlzation of this Northern blot membrane with Neo"^ probe to detect LTKOSN vector (4.0-kb), ALTKOSN vector (2.5-kb) and a Neo"^ RNA transcript (1.2-kb) derived fi-om intemal SV40 promoter activity. (C)

Hybridization with GAPDH cDNA probe demonstrated similar levels of RNA loading. 79

CVi z Sub-clone #4 CO I 1 s Day 0 7 23 31 kb

unspliced* Helper Virus 4.4-

2.4- env

4.4- LTKOSN Vector ALTKOSN neo

GAPDH

1 2 3 4 5 80

CHAPTER 4. CHIMERIC RETROVIRAL HELPER VIRUS AND PICORNAVIRUS IRES

SEQUENCE TO ELIMINATE DNA METHYLATION FOR IMPROVED RETROVIRAL

PACKAGING CELLS

A paper to be submitted to Human Gene Therapy

Won-Bin Young'-^ and Charles J. Link, JrJ'^

RUNNING TITLE: Chimeric retroviral helper virus for high vector production.

ABBREVIATIONS:

LTR: long temninal repeat MoMLV: Moloney murine leukemia virus

RCR; replication competent retrovirus MTase: methyltransferase

VPC: vector producer cell Env: envelope

KEYWORDS: vector producer cell; genetic instability; gene therapy; retrovirus; DNA methylation; retroviral vector; packaging cell; superinfection; retroviral interference; Moloney murine leukemia virus.

^ Human Gene Therapy Research Institute, John Stoddard Cancer Center, 1415 Woodland Ave.,

Des Moines, lA 50309. ^ Molecular, Cellular and Developmental Biology Program, Iowa State

University, Ames, lA 50011. *To whom reprint requests should be sent, (515) 241-8787; Fax:

(515) 241-8788; e-mail: [email protected]. 81

ABSTRACT

Most retroviral packaging cell lines were established by a helper virus plasmid co-

transfected with a separate plasmid encoding a selection marker. Since this selection marker co

existed in trans with the helper virus sequence, helper virus gene expression could be inactivated

by host DMA methylation despite selection for Uie co-transfected selection marker. We found

previously that DNA methylation could occur in tiie LTR region of helper virus in vector producer cells (VPC) up to 2% of tine population per day (Young et al., in preparation). To overcome host cell DNA methylation that causes suppression of viral gene expression, we constixicted a chimeric retroviral helper virus, pAM3-IRES-Zeo, that consists of MoMLV helper virus and a picomavirus intemal ribosome entiy site (IRES) sequence followed by Zeocin^e lection marker at the 3 end of env sequence. This pAM3-IRES-Zeo permitted selection for intact and functional helper virus in transfected cells without subcloning. By selection with Zeocin™a m ixed population of pAM3-IRES-Zeo ti^nsfected NIH3T3 cells (AMIZ cells) were maintained witii little or no DNA methylation of the helper viois 5 LTR. The high level of pAM3-IRES-Zeo gene expression resulted in no detectable vector superinfection and high vector titers (2 X10® to 1.5 X lO^cfu/ml) after introduction of a reti-oviral vector. After Zeocin™ selection was withdrawn from

AMIZ cells, methylation of 5 LTR increased from 17% to 36% of tiie population during 67 days of continuous culture and became susceptible to superinfection. During this period, gene expression of pAM3-IRES-Zeo decreased and vector titer production was reduced to 2 X10"^ cfli/ml. These data demonsti^te an important role of DNA methylation in genetic instability of VPC. Thus, the chimeric helper virus design allows establishment of a packaging cell line capable of high vector production without cloning procedures. 82

INTRODUCTION

Our laboratory is interested in the genetic instability of retroviral VPC caused by host cell

DNA methylation. We previously obsen/ed that extensive DNA methylation could occur in

LTK0SN.2 VPC of retroviral helper virus sequences in 2% of cell population per day. The DNA metiiylation of helper vinjs 5 LTR in LTKOSN.2 VPC correlated with reduced helper virus gene expression. These cells therefore had significantiy reduced Env-receptor interference and became target cells for vector re-entry (superinfection). The VPC developed increasing genetic instability manifest by increasing vector copy numbers within the cells. The decreased helper virus gene expression secondary to DNA metinylation dramatically reduced vector titer of VPC

(53). In order to overcome these observed limitations, we designed a modified retroviral helper virus to overcome limitations caused by host DNA methylation.

Mammalian DNA methyltiBnsferase (MTase) catalyses the ti^nsfer of a metiiyl group to cytosine located 5 to guanosine (CpG dinucleoti'de) and causes epigenetic effects which usually involves gene silencing (reviewed in reference 47). Metiiylated CpG dinucleoti'des inactivate gene expression by altering DNA confomiation (7, 22, 36) or binding methylated CpG-binding proteins

(12,23, 38, 39) that impede ti^nscripti'on. The majority of DNA methylation patterns in mammalian genomes are found in retirovirus-related sequences, such as retrotransposons and endogenous or exogenous retroviruses (52). Evidence suggests that DNA methylation may act as a host defense system against retroviral invasion of the cellular genome (3, 51, 52). DNA metiiylation can be triggered by insertion of viral DNA sequence into chromosomes regardless of whether tiransfection (2) or viral infection (16,29,45) was used to introduce viral DNA sequences.

In several other experimental systems, host cell methylation of retiroviral provirus or reti-otiransposons has been evaluated. In a ti^nsgenic mouse model, a retroviral provirus altered metiiylation pattern witiiin 1 kb of tiie reti-oviral integration site and the provirus was metiiylated 83

leading to an inactivaton of transcription (17,18). Sequences of small interspersed repetitive

elements contained in the rat a-fetoprotein promoter region were associated with increased DNA

methylation and decreased downstream reporter gene expression (11). Reduction of host DNA

methylation leads to amplification and retrotransposition of KERV-1 (kangaroo endogenous

retroviral element-1) and xenologous recombination of chromosomes in interspecific mammalian

hybrids of tiie Australian wallaby (41). Interestingly, retix)viruses may benefit fixjm host DNA methylation as well. HIV-1 infection may induce host DNA methylation activity and, as a consequence, the promoter region of gamma interferon was downregulated by DNA methylation

(29). This might in theory alter tiie balance of cytokines and reduce immune sun/eillance (28,29).

The Inactivation of HIV-1 or HTLV-1 gene expression by host DNA methylation of viral LTR regions may also induce latency of HIV-1 or HTLV-1 infection (1, 28,44,45).

These prior experiments did not evaluate DNA methylation of helper virus 5 LTR in VPC.

Many commonly used retroviral vector packaging cell lines were established by co-transfecb"on of two plasmids, one plasmid contains a helper virus genome and the other encodes a drug selection marker (26, 31-33). In this co-b^nsfecti'on system, selection for dnjg resistance does not require active helper virus gene expression, so tine 5 LTR promoter region can be silenced by

DNA methylation. In tiie current study, a chimeric helper virus, pAM3-IRES-Zeo'^ was designed containing an IRES sequence of tiie encephalomyocarditis virus (ECMV) (19), a member of picomaviruses (43), and a Zeocin™resistance gene (Zeo") (9) to allow selection against DNA methylation that might occur in the helper virus 5 LTR region.

For translation of mosteukaryoti'c mRNAs, ribosomes scan mRNAfi-om tine 5 cap sequence until an initiation codon is reached. In contiBst, in picomavirus mRNA, ribosomes initiate translation by an altemative mechanism tiiat involves internal initiation ratiner than scanning. The IRES sequences of picomavirus enable ribosomes to bind in a cap-independent fashion and start ti^ansiation at the next AUG codon downsti'eam (20). Ligation of IRES sequence 84

followed by Zeo"" at the 3 end of env gene pemnits the translation of helper virus open reading

frames and a selection marker from the same mRNA (Rg. 3). Selection with Zeocin'^fel iminates

cells witii metiiyfated helper vinjs 5 LTRftiom the population. This design should ensure more

stable helper virus gene expression which would increase virion production and create sufficient

Env-receptor interference to prevent superinfection. The prevention of superinfection may in turn reduce RCR formation (34, 35). One additional advantage is that pAM3-IRES-Zeo allows for establishment of packaging cell lines within a shorter time period. This advantage might be criti'cal when making human VPC from a primary cell culture or stem cells to avoid immune rejection (48,

49) while transplantation of VPC into patients is necessary for continuous gene ti^nsfer (42). (W.-

B. Young is a doctoral candidate at Iowa State University, and this work represents partial fulfillment of dissertation requirement)

MATERIALS AND METHODS

Construction of helper virus pAM3-IRES-Zeo and LEIN vector. An IRES sequence of

ECMV was isolated from the LXIN retrowral vector (Clontech, Palo Alto, CA) witii Nsil and Psfl digestions and tiien inserted into a Psfl-linearized pZeoSV mammalian expression vector

(Invifrogen, Carishbad, CA) immediately 5 of EM-7 prokaryoti'c promoter/Zeodn™ resistance gene (Zeo^) to create an IRES-Zeo expression cassette. This plasmid was named pIRES-Zeo-

SV40. San digestion on plRES-Zeo-SV40 deleted tiie SV40 promoter and downsfream polyadenylation signal to generated pIRES-Zeo. A 2.8-kb fragment, consisting of IRES-Zeo expression cassette, SV40 polyA signal, bacterial replication origin (ColEI Ori) and phage replication origin (F1 Ori), was excised from plRES-Zeo-SV40 by sequential steps, including EagI digestion, freatment with Klenow enzyme (GIBCO BRL, Life Technology Co., Gaithersburg, MD) to fill in the EagI restiiction site, and Xbal digestion. To construct pAM3-IRES-Zeo, an amphotropic helper virus pPAM3 (31) (kindly provided by A. Dusty Miller, Fred Hutchinson 85

Cancer Research Center, Seattle, WA) was digested by Hpal at 3 end of env gene and Nhe\ at

5 end of LTR to delete the C0IEI On and amplcilin resistance gene (AmpR). This deleted region

was replaced with the 2.8-kb IRES-Zeo-containing fragment described above (Rg. 3). The

resulting chimeric helper virus plasmid, pAM3-IRES-Zeo, allows selection with Zeocin™ in

bacterial culture and mammalian cells.

The LEIN retroviral vector carrying an enhanced green fluorescent protein (EGFP, (6,

10), Clontech, Palo Alto, CA) reporter gene was constructed by replacing the SV40 promoter- neomycin phosphotransferase gene (Neo^ cassette of pLESN (27) with a 1.4-kb IRES-Neo"^ cassette, excised from pIRES-Neo by A/ael and Afe/l digestions.

Cell culture and transfection. Cell cultures were maintained in Dulbecco modified Eagle medium (DMEM; GIBCO BRL, Life Technology Co., Gaithersburg, MD), 10% fetal calf serum with 5% CO2, at 37°C. The subclones of LTKOSN.2 VPC were obtained by limiting dilution of parental LTKOSN.2 VPC onto two 96-well plates from previous study (53). Helper virus and vector gene expression, DMA methylation status and vector production ability in these subclones have been previously characterized. Transfections of pAM3-lRES-Zeo into these LTKOSN.2

VPC subclones and/or NIH 3T3 tk- cells to produce AMIZ packaging cells were perfomied using

Fugene 6™ Transfection Reagent (Roche Molecular Biochemlcals, Indianapolis, IN). Prior to tiransfection, pAM3-IRES-Zeo plasmid was linearized by BspHI digestion and 6 to 10 |ig of pAM3-IRES-Zeo was then transfected to each well in 6-well plates. Selection with Zeocin™(350 lag/ml, Invitrogen) started 48 hr after ti^nsfection and continued for at least two weeks.

Transfection of LEIN vector into the AMIZ cell pool and GP+E86 packaging cells (26) (kindly provided by Artiiur Bank, Columbia University) was perfomned by using DOTAP Liposomal

Transfection Reagent (Roche Molecular Biohemicals) witii 5 |ag of LEIN plasmid for each well in

6-well plates. Selection with G418 (1 mg/ml; GIBCO BRL, Life Technology Co.) started 48 hr after transfection and continued for two weeks. 86

Retroviral infection, superinfection and titer assays. Titer assays were performed using stendard procedures. Supematants collected from pAM3-IRES-Zeo-transfected LTK0SN.2

VPC subclones were diluted in 10-fold serial dilutions to transduce NIH3T3 tk{-) cells [American

Type Culture Collection (ATCC) CRL1658], A375 cells (ATCC CRL1619, human melanoma) and

IGROV cells [human ovarian cardnoma (50)], which were plated at 1 X10® cellsA/vell in 6-well plate witiilO ^g/ml of protamine sulfate. Twenty-four hours after ti^nsduction, cells were selected in medium containing G418 (1 mg/ml) for 10-14 days. Titers were calculated by multiplying tiie number of resistant colonies by the dilution factor.

To perform superinfection assays on AMIZ cells, supematants containing LEIN vector collected from LEIN-transfected AMIZ cells were passed through a 0.4-^im syringe filter and diluted 10-fold and 100-fold. NIH3T3 and PA317 cells were transduced as Env-receptor interference positive and negative conti-ols, respectively, along with AMIZ cells. Selection witii

G418 (1 mg/ml) on tiiese transduced cells started 24 hr after a single exposure to LEIN vector and continued for 10-14 days. The number of G418 resistant colonies was used as the index for superinfection on PA317 and AMIZ cells. Transduction of a LEIN vector from GP+E86 cells into

AMIZ cell pool was perfomned as in tiie above superinfection procedure without dilution of supematants.

RNA analysis of helper virus and vector gene expression. Total cellular RNA was isolated from ti^nsfected cells and VPC by using RNAeasy kit (Qiagen Inc., Valencia, CA) and subjected to Northern blot analysis on a 1% agarose-0.4M formaldehyde gel. Vector ti^nscripts were detected by Neo''probe and helper viral transcripts were detected by a 1.4-kb env probe, which was digested from pPAM3 by X/70I. Human glyceraldehyde 3-phosphate dehydrogenase

(GAPDH) cDNA probe was used to demonsti:Bte similar RNA loading.

DNA Methylation analysis. In AMIZ cells ti^nsfected witii LEIN vector, the methylafa'on status of provirus and vectors at the Smaf site in the 5 LTR as determined by digestion of 87

genomic DNA with Dra\ and EcoRV to reduce the DNA fragment size. The DNA was then

precipitated with ethanol, re-dissolved in sterile water and divided into two equal portions, one of

which was subjected to methylation-sensitive Smal restriction endonuclease digestion. The

Southern blot membrane was hybridized wttii a 428-bp fragment of gag sequence {Pvu\\/Dra\)

fixDm pPAM3 to detect helper virus, a 273-bp EcoRI/EcoRV Augment of HSVtfc sequence for

LTKOSN vector, and a GFP probe to detect LEIN vector. Densitometric analyses were performed with a Hoefer Densitometer GS300 (Hoefer Sdentific Instruments) to measure the relative densities of Smal-sensitive band compared to Dral/EcoRV band. Due to interference from endogenous retroviral elements, the fraction of Smal methylation in 5 LTR was calculated as one minus tine intensity ratio of the Smal sensitive band (1.5-kb) divided by Dral/EcoRV band

(1.8-kb) as depicted in Rg. 6.

In AMIZ cells, no interference from eithervector or endogenous reti-oviral sequences was present. The DNA methylation status of 5 LTR region of pAM3-IRES-Zeo was determined by digesting genomic DNA witii EcoRV, Ss£ll and Smal. If methylation occurred at the Smal site, a

608-bp fragment would be excised instead of a 348-bp fragment when probed with a 261-bp fragment excised from pAM3-IRES-Zeo by Kpnl and AMI digestions. The degree of DNA methylation was calculated as the intensity ratio of tiie Smal insensitive band (608-bp) divided by the sum of tiie intensity of tills 608-bp fragment and Smal sensitive fragment (348-bp) as depicted in Rg. 3.

RESULTS

Construction of a chimeric retrovirai helper virus with IRES and selection marker to allow direct selection of helper virus gene expression. We previously determined that

DNA methylation occurred in 2% of the cell population per day within 5 LTR region of helper virus to inactivate helper virus gene expression in VPC (53). To eliminate methylated helper virus 88

SLTRfrom packaging cell population, a chimeric retroviral helper virus, pAM3-iRES-Zeo (Rg. 1), was constructed. The pAM3-IRES-Zeo construction allows Zeodn™ selection of cells with 5 LTR promoter function, since helper virus and Zeo"^ gene expression are transcribed fi'om the 5 LTR promoter (Rg. 1B). The selection with Zeocin™ maintains cells that also express helper virus and therefore counteract DMA methylati'on effects. Packaging cells based on this pAM-IRES-Zeo helper virus should maintain high titer production. Secondarily, tfie evaluation of tiiese pAM3-

IRES-Zeo ti^nsfected cells witii or without Zeodn™ selection allows study of metiiylati'on of helper virus 5 LTR and helper virus gene expression over time.

Analysis of chimeric pAM3-iRES-Zeo vector packaging ability. To test tine packaging ability of pAM3-IRES-Zeo helper virus, pAM3-IRES-Zeo was ti^nsfected into three individual subdones of LTK0SN.2 VPC. The pPAM3 helper virus gene expression in tinese tiiree

LTKOSN.2 VPC subclones #1, #3 and #5 (Rg. 2A, lanes 4-6), were inactivated by DNA methylati'on with impeded vector production ability (Table 1)(53). However, LTKOSN (4.0-kb) and

ALTKOSN (2.8-kb) vectors in these subdones were still tinanscribed (Rg. 2B, lanes 4-6) and no significant DNA methylati'on of these vectors was observed (53). This indicates tiiat a key limitation of vedor production in LTKOSN.2 VPC subdones is the lack of helper virus gene expression. Rescue of LTKOSN and ALTKOSN vedors from these three subclones was perfonned by transfection of pAM3-IRES-Zeo followed by two weeks of Zeocin™ selection. This restored high level of vedor production was shovwi by titer analysis on human IGROV ovarian cardnoma, human A375 melanoma and murine NIH3T3 tk- target cells. The titers ranged fi'om

1.6 X10^ to 4 X10® colony fomriing units (cfu)/ml (Table 1). In addition, tills increased packaging activity witin pAM3-IRES-Zeo resulted in a reduction of retained LTKOSN and ALTKOSN vedors inside VPC when evaluated by Northern blot analysis (Rg. 2B, lanes 7-9).

Analysis of gene expression in pAM3-IRES-Zeo transfeded LTKOSN.2 VPC subclones demonstiBted significantiy greater helper virus gene expression compared to pPAM3 in PA317 89

packaging cells and parental LTKOSN.2 VPC (Rg. 2A). In addition to env'transcripts, only one

jsopulation of unspliced helper virus (gagiX)f-env-\RES-Zeo) was detected in pAM3-IRES-Zeo

transfected cells, which indicates that the Integration of pAM3-IRES-Zeo should be intact in

transfected cells after selection. In contrast co-transfection of pPAM3 without direct selection for

pPAM3 gene expression but other selection marker in trans could result in randonnly interrupted

pPAMS for integration. This was represented by two more ti^nscripts of lower molecular weights

detected in PA317 and LTKOSN.2 VPC (Rg. 2, lanes 2 and 3) (53). Overall, tiiese results

demonstrate that enhanced and maintained helper virus gene expression can be obtained in

pooled packaging cells when pAM3-IRES-Zeo is used to allow Zeocin™ selection witiiout the

need to perform time-consuming cell subcloning. This implies a potential use of pAM3-IRES-Zeo

to establish new packaging cells from other cells such as human primary cells.

pAM3-IRES-Zeo transfected cells provides a model to study DNA methyiation of

retroviral sequences. DNA methyiation In mammalian cells is site-dependent within the genome

(14). Therefore, a mixed population of pAM3-IRES-Zeo ti^nsfected cells would be required to

study DNA methyiation of helper virus 5 LTR to minimize the effects of positional interference. A

pool of packaging cells, AMIZ, was established by transfection of pAM3-IRES-Zeo into NIH3T3

tfc-cells and selection with Zeocin™ without subcloning. To allow DNA methyiation to occur, AMIZ

cells were released from Zeodn™ selection for one month and then placed in continuous culture

witii or without Zeocin™selection for 78 days (about 10 passages). DNA methyiation and gene

expression of pAM3-IRES-Zeo were examined at 15,54, and 78 days after being released from

selection. Over the first 54 days of cell culture period, DNA methyiation of 5 LTR was increased

from 8% to 19% and by Day 78 reached 61% (Rg. 3). The DNA methyiation rate of helper virus

5 LTR averaged 0.7% of population per day during 78 day period. AMIZ cells witii continued

Zeocin™ selection did not exhibit any detectable DNA methyiation (Rg.3). This demonsti^tes tiie 90 effectiveness of daig selection to eliminate methylated pAM3-IRES-Zeo fl-om the pooled AMIZ population.

Retroviral superinfection is bloci(ed by enhanced helper virus gene expression. To study the effect of Zeodn™selection on AMIZ cells, we analyzed gene expression of pAM3-

IRES-Zeo in AMIZ cells from the above experiment Gene expression of pAM3-IRES-Zeo in

AMIZ cells with constant Zeocin™ selection showed a 2-fold increase compared to AMIZ cells without selection on Day 15 and a 4-fold increase on Days 54 and 78 (Rg. 4). In contrast, pAM3-

IRES-Zeo gene expression in AMIZ cells withoutZeodn^selection declined overtime (Rg. 4, lanes 3, 5 and 7). Continuous Zeodn™ selection may have selected integration sites that are highly transcriptionally active and tiiat have less DMA metiiylati'on activity (5,22).

We next directiy detemnined whether decreased pAM3-IRES-Zeo gene expression reduced Env-receptor interference and increased vector superinfection. The susceptibility to superinfection was measured by exposing AMIZ cells with or without Zeodn™ selection from the above experiment to amphoti-opic LEIN vector supematants followed by G418 selection. G418 resistant colony number obtained fi'om AMIZ cells with continued Zeocin™ selection was reduced from 2.3 X10^ on Day 15 to no superinfection observed on Days 54 and 78 (Table 2). In contrast, G418 resistant colonies obtained from AMIZ cells without Zeodn™ selection ranged from 1.2 X10^ to 5.6 X10^. These results demonstrate tiiat increased gene expression of helper vims correlates with reduced susceptibility to superinfection.

High level of vector production was maintained by Zeocin™ selection. To study vector production ability of tinis AMIZ cell pool, we transfected LEIN vector into AMIZ cells followed by G418 selection to establish a VPCfor titer assay. Zeocin™ selection was temporally withdrawn from AMIZ ceil culture during the first tiiree weeks of G418 selection after ti^nsfection with LEIN vector. Titer obtained fi-om tills newly established uncloned population of AMIZ cells was 3.5 X10® cfti/ml, which is 100-fold higher than the titer observed firom a mixed population of 91

PA317 transfected with LEIN vector (4 X10^ cfu/ml). In addition. AMIZ cells were transduced with

LEIN vector collected from LEIN-transfected GP+E86 cells and an improved titer of 9 X10®

cfii/ml was obtained from a mixed cell population. To investigate whether selection with both

Zeocin™ and G418 would adversely affect vector production, LEIN transfected AMIZ cells were

evaluated 56 (8 passages) and 67 days (10 passages) after transfection. Titers obtained from

AMIZ cells transfected with LEIN and placed under continuous selection with Zeocin™ and G418,

were maintained between 2 X10® to 1.5 X10^ cfu/ml (Table 3). In contrast, titers obtained from

the same AMIZ cells transfected with LEIN but not maintained with G4-18 and Zeocin™ selection

decreased to between 2 to 4 X10"* cfu/ml. The reduced titer correlated with a significant decrease

of both helper virus and vector gene expression (Rg. 5).

DNA methylation status of 5 LTRs of helper virus and vector were significantiy increased

in AMIZ cells transfected with LEIN vector and cultured without either G418 or Zeocin™ selection

(Rg. 6). This increased methylation corresponded to decreased vector titer (Table 3) and

significantiy reduced gene expression of helper virus and vector (Rg. 5). The DNA methylation of

helper virus 5 LTR increased from 17% (Day 0) to 30% and 36% by Days 56 and 67,

respectively. The average DNA metiiylati'on rate of helper virus 5 LTR in AMIZ cells transfected

with LEIN was estimated as low as 0.3% of tiie cell population per day during 67 days of

continuous cell culture. In contrast DNA methylation was not detected in AMIZ cells transfected

with LEIN vector and placed under continuous G418 and Zeocin™ selection. No detectable DNA

methylation occurred of LEIN vector on Day 0 (Rg. 6B, lanes 3 and 4) while 17% DNA methylation of 5 LTR helper virus was detected (Rg. 6A, lanes 3 and 4). This may be secondary to the timing of G418 and Zeodn™ selection. AMIZ cells transfected with LEIN vector were placed under G418 for three weeks to select for LEIN-positive population and Zeocin™ selection was not applied until Day 0 in tiie experiment 92

DISCUSSION

The experimental model described in this study demonstrates an approach using a retroviral helper virus combined with picomavirus (RES sequence and a selecbon mariner gene that allows efficient elimination of methylated helper virus from packaging cell populations. This strategy of using drug selection maintained high levels of helper virus gene expression and high titer vector production (1.5 X10^ cfu/ml) from populations of VPC without subcloning. The presence of greater Env-receptor interference blocks vector superinfection and may reduce other potential problems with retroviral vectors such as RCR formation.

The potential application of this chimeric helper virus to gene therapy might allow packaging cells to be established fi"om primary cell culture with less time and without subcloning.

Transfection of the pAM3-IRES-Zeo into cells followed by selection for positive populations takes only two weeks or less, depending on transfection efRdency. This is particulariy useful for transplantation of VPC into patients (42) for continuous in vivo gene transfer without the immune elimination of murine VPC and virions (48,49). Several stijdies have aimed to establish a retroviral packaging cell line by using other viral vectors, such as adenovirus (4, 8,25) and herpes simplex virus (46), to import retroviral helper virus genome into target cells in vivo or ex vivo. These approaches have significant limitations such as immunogenicity of tiie additional virus and direct cytotoxicity to the infected cells (13, 37). Since tine transferred helper virus sequences are not integrated into chromosomes, these systems generally have lower and unstable titers secondary to a loss of transferred helper virus. In this study, we demonsti^te that the titer of pooled population of pAM3-IRES-Zeo-transfected cells, AMIZ cells, was maintained between 3.5 X10® to 1.5 X10' cfu/ml for at least 67 days. These titers are comparable to reported titers fiiom individually cloned VPC, which generally ranged from to 10^ (30).

Transducing LEIN vectors Into AMIZ cells instead of transfection also resulted in a uncloned population of VPC with high titer production (9X10® cHi/ml). 93

The selection of transfected cells (AMIZ cells) with Zeodn™ to maintain pAM3-IRES-Zeo

gene expression eliminated DMA methylation firom AMIZ cells and may also select cells with pAM3-!RES-Zeo helper virus integrated in optimal and active chromosomal regions. Ratios of pAM3-IRES-;&o gene expression in selected AMIZ cells compared to non-selected AMIZ cells were about 4:1 on Days 54 and 78. The percentage of 5 LTR methylation was 0% in cells under continuous selection and 19% (Day 54) and 61% (Day 78) in cells without selection. The differences in vector titers were more dramatic with a 10OO-fold greater, 1.5X10' cfu/ml, Day 78, in LEIN-transfected AMIZ cells under continuous selection, compared to 4 XIC* cfu/ml. Day 78, in cells witiiout selection. This dramatic difference probably reflects the fact that structural proteins of viruses function as multi'mers (15). Therefore, the fomnation of multimers occurs in a sigmoid dose-response fashion, rather than a linear dose-response to protein concentration that correlates more directiy with helper virus gene expression. To maintain efficient Env-receptor interference and active viral production, a threshold level of helper virus gene expression is required. In reti^ovirus infection, this threshold level of gene expression is established by the accumulation of sufficient copy number of virus through superinfection until efficient Env-receptor interference is achieved and maintained (40). In our study, the threshold level of helper virus gene expression was achieved by Zeocin™ selection rather than increasing tiie copy number of helper virus. When selection pressure was released, superinfection was observed after helper virus gene expression declined. These results support a conclusion that continuous selection of helper virus in VPC might enhance Env-receptor interference and reduce tiie possibility of RCR fomiati'on.

For continuous virus production without interfering witii host cell growth and differentiation, reti-oviral gene expression has to be regulated at a level just sufficient in retrovirus infected cells. Increased levels of viral RNAs and proteins in infected cells can cause cytopathic effects, usually at the cost of cell death, by intenxipti'ng tiie production ortiBnslation of host 94 mRNA (47) (for a review see reference 46). Althougin we obsen/ed that AMIZ cells under continuous selection did proliferate more slowly than AMIZ cells without selection, under continuous selection for 67 days, AMIZ cells transfected witii LEIN vector proliferated without difRculty (data not shown).

We did not attempt to select for pPAM3 gene expression by drug selection against the

HSVtk selection marker plasmid co-transfected into PA317 cells. This approach is unlikely to be successful since the selection marker plasmid is separate firom pPAM3. An alternative approach to reverse methylation is ti'eatment with 5 -aza-cyti"dine (5-aza-C) to reverse DNA metiiylation (21,

24). In our previous experiments, we found that only a minor portion of pPAMS helper virus expression could be restored by 5-aza-C (53). Treatment with 5-aza-C does not specifically reverse helper viais DNA metiiylation, but inhibits Uie cellular DNA methyltransferase and causes cytotoxicity to treated cells (21). The data from this study suggest that a combination of helper virus and IRES sequences witii selectable markers is a viable option to eliminate host DNA methylation of helper virus from VPC.

ACKNOWLEDGMENTS

We thank Dr. John P. Levy for providing the LESN vector. This work was supported in partby a grant from Iowa Methodist Medical Center, Des Moines, Iowa. W.-B. Young is a recipient of a pre-doctoral fellowship grant fixim the Molecular, Cellular and Developmental

Biology Program, Iowa State University, Ames, Iowa.

REFERENCES

1. Bednarik, D. P., J. A. Cook, and P. M. Pitha. 1990. Inactivation of tiie HIV LTR by DNA

CpG methylation: evidence for a role in latency. Embo J. 9:1157-64. 95

2. Bednarik, D. P., J. D. Mosca, and N. B. Raj. 1987. Methylation as a modulator of

expression of human immunodeficiency virus. J Virol. 61:1253-7.

3. Bestor, T. H., and B. Tycko. 1996. Creation of genomic methylation patterns. Nat Genet

12:363-7.

4. Caplen, N. J., J. N. Higginbotham, J. R. Scheei, N. Vahanian, Y. Yoshida, and H.

Hamada. 1999. Adeno-retroviral chimeric viruses as In vivo ti^nsducing agents. Gene Ther.

6:454^59.

5. Cedar, H. 1988. DNA methylation and gene activity. Cell. 53:3-4.

6. Cormack, B. P., R. H. Valdivia, and S. Falkow. 1996. FACS-optimized mutants of the green

fluorescent protein (GFP). Gene. 173:33-8.

7. Fell, R., M. D. Boyano, N. 0. Allen, and G. Kefsey. 1997. Parental chromosome-specific

chromatin conformation in tiie imprinted U2af1-rs1 gene in the mouse. J Biol Chem.

272:20893-900.

8. Feng, M., W. H. Jackson, Jr., C. K. Goldman, C. Rancourt, M. Wang, S. K. Dusing, G.

Siegal, and D. T. Curiel. 1997. Stable in vivo gene ti^ansduction via a novel

adenoviral/retroviral chimeric vector. Nat Biotechnol. 15:866-70.

9. Gatignol, A., H. Durand, and G. Tiraby. 1988. Bleomydn resistance conferred by a drug-

binding protein. FEBS Lett. 230:171-5.

10. Haas, J., E. C. Park, and B. Seed. 1996. Codon usage limitation in the expression of HIV-1

envelope glycoprotein. Curr Biol. 6:315-24.

11. Hasse, A., and W. A. Schulz. 1994. Enhancement of reporter gene de novo metiiylation by

DNA fragments firom the alpha-fetoprotein contiDi region. J Biol Chem. 269:1821-6.

12. Hendrich, B., and A. Bird. 1998. Identification and characterization of a family of

mammalian metiiyl-CpG binding proteins. Mol Cell Biol. 18:6538-47. 96

13. Ho, D. Y., S. L. Fink, M. S. Lawrence, T. J. Meier, T. C. Saydam, R. Dash, and R. IM.

Sapolsicy. 1995. Herpes simplex viais vector system: analysis of Its in vivo and in vitro

cytopathic effects. J Neurosd Methods. 57:205-15.

14. Hoeisen, R. C., A. A. IMigchielsen, R. C. van der Jagt, H. van Ormondt, and A. J. van der

Eb. 1991. Inactivation of the Moloney murine leukemia virus long tenninal repeat in murine

fibroblast cell lines is associated with methyiation and dependent on its chromosomal

position. J Virol. 65:904-12.

15. Hunter, E. 1994. Macromolecular interactions in the assembly of HIV and other retroviruses.

Semi Virol. 5:71-83.

16. Jahner, D., and R. Jaenisch. 1985. Retrovirus-induced de novo methyiation of flanking host

sequences connelates with gene inactivity. Nature. 315:594-7.

17. Jahner, D., and R. Jaenisch. 1985. Chromosomal position and specific demethylation in

enhancer sequences of germ line-transmitted retiioviral genomes during mouse development

Mol Cell Biol. 5:2212-20.

18. Jahner, D., H. Stuhlmann, C. L. Stewart, K. Harbers, J. Lohler, I. Simon, and R.

Jaenisch. 1982. De novo metinylation and expression of retroviral genomes during mouse

embryogenesis. Nature. 298:623-8.

19. Jang, S. K., H. G. Krausslich, M. J. Nicklin, G. M. Duke, A. C. Paimenberg, and E.

Wimmer. 1988. A segment of the 5' nontranslated region of encephalomyocarditis virus RNA

directs internal entiy of ribosomes during in viti'o translation. J Virol. 62:2636-43.

20. Jang, S. K., and E. Wimmer. 1990. Cap-independent translation of encephalomyocarditis

virus RNA: struchjral elements of tine internal ribosomal entry site and involvement of a

cellular 57-kD RNA-binding protein. Genes Dev. 4:1560-72. 21. Juttermann, R., E. Li, and R. Jaenisch. 1994. Toxicity of 5-a2a-2'-deoxycytidine to

mammalian cells is mediated primarily by covalent trapping of DMA methyltransferase rather

than DMA demethylation. Proc Natl Acad Sci U S A. 91:11797-11801.

22. Keshet, I., J. Lieman-HunvHz, and H. Cedar. 1986. DNA methylation affects the formation

of active chromatin. Cell. 44:535-43.

23. Lamb, B. T., K. Satyamoorthy, L. Li, D. Softer, and C. C. Howe. 1991. CpG metiiylation of

an endogenous retroviral enhancer inhibits transcription factor binding and activity. Gene

Expr. 1:185-96.

24. Lengauer, C., K. W. Kinzier, and B. Vogelstein. 1997. DNA methylation and genetic

instability in colorectal cancer cells. Proc Nati Acad Sd U S A 94:2545-50.

25. Lin, X. 1998. Constixiction of new retroviral producer cells ft-om adenoviral and reti^oviral

vectors. Gene Ther. 5:1251-8.

26. IMarkowitz, D., S. Goff, and A. Bank. 1988. A safe packaging line for gene ti^nsfen

separating viral genes on two different plasmids. J. Virol. 62:1120-4.

27. Maze, f. A., J. P. Levy, R. R. Muldoon, C. J. Link, Jr., and S. R. Kain. 1999. Retroviral

expression of green fluorescent protein. Methods in Enzy. 302:329-341.

28. Mikovits, J. A., Raziuddin, M. Gonda, M. Ruta, N. C. Lohrey, H. F. Kung, and F. W.

Ruscetti. 1990. Negative regulation of human immune deficiency virus replication in

monocytes. Distinctions betiAreen restricted and latent expression in THP-1 cells. J Exp Med.

171:1705-20.

29. Mikovits, J. A., H. A. Young, P. Vertino, J. P. Issa, P. M. Pitha, S. Turcoski-Corraies, 0.

D. Taub, C. L. Petrow, S. B. Baylin, and F. W. Ruscetti. 1998. Infection with human

immunodeficiency virus type 1 upregulates DNA methyltiBnsferase, resulting in de novo

metiiylati'on of the gamma interferon (IFN-gamma) promoter and subsequent downregulation

of IFN-gamma production. Mol Cell Biol. 18:5166-77. 98

30. Miller, A. D. 1990. Retrovirus packaging cells. Hum Gene Ther. 1:5-14.

31. Miller, A. D., and C. Buttimore. 1986. Redesign of retrovirus packaging cell lines to avoid

recombination leading to helper virus production. Mol Cell Biol. 6:2895-902.

32. Miller, A. D., and F. Chen. 1996. Retrovirus packaging cells based on 10A1 murine

leukemia virus for production of vectors that use multiple receptors for cell entry. J Virol.

70:5564-71.

33. Miller, A. D., J. V. Garcia, N. Von Shur, C. M. Lynch, C. Wilson, and M. V. Eiden. 1991.

Construction and properties of retrovirus packaging cells based on gibbon ape leukemia

virus. J Virol. 65:2220^.

34. Miller, A. D., D. R. Trauber, and 0. Buttimore. 1986. Factors involved in production of

helper virus-free retrovirus vectors. Somat Cell Mol Genet 12:175-83.

35. Muenchau, D. D., S. M. Freeman, K. Cometta, J. A. Zwiebel, and W. F. Anderson. 1990.

Analysis of retroviral packaging lines for generation of replication- competent virus. Virology.

176:262-5.

36. Muiznieks, I., and W. Doerfier. 1994. The topology of the promoter of RNA polymerase II-

and Ill-transcribed genes is modified by the methylation of 5'-CG-3' dinucleotides. Nucleic

Acids Res. 22:2568-75.

37. Muruve, D. A., M. J. Barnes, I. E. Stillman, and T. A. Libermann. 1999. Adenoviral gene

therapy leads to rapid induction of multiple chemokines and acute neutrophil-dependent

hepatic injury in vivo [In Process Citation]. Hum Gene Ther. 10:965-76.

38. Nan, X., F. J. Campoy, and A. Bird. 1997. MeCP2 is a transcriptional repressor witii

abundant binding sites in genomic chromatin. Cell. 88:471-81.

39. Nan, X., H. H. Ng, C. A. Johnson, C. D. Laherty, B. M. Turner, R. N. Eisenman, and A.

Bird. 1998. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a

histone deacetylase complex [see comments]. Nature. 393:386-9. 40. Odawara, T., M. Oshima, K. Doi, A. Iwamoto, and H. Yoshikura. 1998. Threshold number

of provirus copies required per cell for efRdent virus production and interference in moloney

murine leukemia virus- infected NIH 3T3 cells. J Virol. 72:5414-24.

41. O'Neill, R., M. O'Neill, and J. A. Graves. 1998. Undermethyiation associated with

retroelement activation and chromosome remodelling in an interspecific mammalian hybrid.

Nature. 393:68-72.

42. Ram, Z., K. W. Culver, E. M. Oshiro, J. J. Viola, H. L. DeVroom, E. Otto, Z. Long, Y.

Chiang, G. J. McGarrity, L. M. Muul, D. Katz, R. M. Blaese, and E. H. Oldfield. 1997.

Therapy of malignant brain tumors by intratumoral implantation of retroviral vector-producing

cells. Nat Med. 3:1354-61.

43. Rueckert, R. R. 1996. Virology. Chapter 21 "Picomaviridae: The viruses and their

replication". 1:609-654.

44. Saggioro, D., M. Forino, and L. Chieco-Bianchi. 1991. Transcriptional block of HTLV-I LTR

by sequence-specific methyiation. Virology. 182:68-75.

45. Saggioro, D., M. Panozzo, and L. Chieco-Bianchi. 1990. Human T-lymphotropic virus type

I transcriptional regulation by metiiylati'on. Cancer Res. 50:4968-73.

46. Savard, N., F. L. Cosset and A. L. Epstein. 1997. Defective herpes simplex virus type 1

vectors harboring gag, pol, and env genes can be used to rescue defective retrovirus vectors.

J Virol. 71:4111-7.

47. Somasundaran, M., and H. L. Robinson. 1988. Unexpectedly high levels of HIV-1 RNA

and protein synthesis In a cytocidal infection. Sdence. 242:1554-7.

48. Takeuchi, Y., F. L. Cosset P- J- Lachmann, H. Okada, R. A. Weiss, and M. K. Collins.

1994. Type C retrovirus inactivati'on by human complement is determined by both the viral

genome and the producer cell. J Virol. 68:8001-7. 100

49. Takeuchi, Y., C. D. Porter, K. M. Strahan, A. F. Preece, K. Gustafsson, F. I_ Cosset, R.

A. Weiss, and M. K. Collins. 1996. Sensitization of cells and retrovimses to liuman seaim

by (alpha 1-3) galactosyftransferase. Nature. 379:85-8.

50. Teyssier, J. R., J. Benard, D. Ferre, J. Da Silva, and L. Renaud. 1989. Drug-related

chromosomal changes in chemoresistant human ovarian carcinoma cells. Cancer Genet

Cytogenet 39:35-43.

51. Yoder, J. A., and T. H. Bestor. 1996. Genetic analysis of genomic methylation patterns in

plants and mammals. Biol Chem. 377:605-10.

52. Yoder, J, A., C. P. Walsh, and T. H. Bestor. 1997. Cytosine methylation and the ecology of

intragenomic parasites [see comments]. Trends Genet 13:335-40.

53. Young, W.-B., G. L. Lindberg, and C. J. Link, Jr. 1999. DNA metiiylation increased genetic

instability of retroviral vector producer cells. In preparation. 101

TABLE 1 .Titer of LTKOSN.2 VPC subclones before and after transfection of pAM3-IRES-Zeo helper virus

Target cells

pAMS-IRES-Zeo NIH3T3 A375 IGROV

+ 1-1 X10^ 1 X 10'' 1 X10^ Subclone #1 1 XIO"* 2X 10° 0

+ 5X10® 5.5X10® 4X10® Subclone #3 0 0 0

+ 1.6X10^ 5X10® 4X10® Subclone #5 0 0 0

TABLE 2. Increased resistance of superinfection by Zeocin selection

Target cells

AMIZ

NIH3T3 PA317 No Selection Selection

Day 15 7X10® 3.2 X 10"^ 4.3 X 10^ 2.3 X lO"*

Day 54 6X10"^ 1 XIO'^ 1.2 X 10=^ 0

Day 78 5X10® 1.9X10"^ 5.6X10^ 0

TABLE 3. Titer of AMIZ cells transfected wrth LEIN with continuous drug selection

Day 0 Day 56 Day 67

Selection No Selection Selection No Selection

3.5X10® 2X10® 2X10"^ 1.5X10^ 4X10"^ FIG. 1. Constmct'on and cap-independent translation nnechanism of chimeric pAM3-

IRES-Zeo helper vims. (A) For details see Materials and Methods. Briefly, a 2.8-kfa fragment

including [RES-Zeo expression cassette, a SV40 polyA signal sequence, bacterial replication

origin (C0IEI On) and phage replication origin (F1 On) was exdsed from pIRES-Zeo. The C0IEI

Ori and ampicilin resistance gene region (AmpR) of pPAM3 were replaced with above 2.8-kb

IRES-Zeo-containing fiBgmentfirom pIRES-Zeo. The EM7 prokaryoti'c promoter located at 5 end

of Zeo"^ gene permits selection for pAM3-IRES-Zeo in bacterial. (B) Genomic RNA of MoMLV

contained two internal stop codons at 3 ends of gag and pol genes thattemninate cap-dependent translation and allow appropriate ratios of viral structural proteins. In pAM3-IRES-Zeo derived transcripts, ribosomes also recognize IRES sequence and initiate translation from the first ATG codon of Zeo*^ downstream of IRES sequence. A portion of genomic RNA is spliced into env transcripts tiiat are translated in cap-dependent mechanism. SD, splicing donor; SA, splidng acceptor. 103

SV40 pA SV40 pA Hpai

Anip(R) EM7

piRES-Zao pPAM3 3455 bp 10668 bp S*LTR IRES

FI Oft Xbst

SV40pA CGIEI Oft

pAM3-IRES-Zao 10743 bp

B pAM3-IRES-Zeo

SD SA I LU IRES Zeo [BEr—I \ pgi r env i i iaaa

UAG UAAUAG H • AAA FIG. 2. Gene expression of pAM3-IRES-Zeo and vectors in LTK0SN.2 VPC subclones.

(A) Northern blot analysis of cellular RNA extracted from pAM3-IRES-Zeo transfected LTK0SN.2

subclones hybridized with env probe. Unspliced MoMLV transcript (gag-po/-env-IRES-Zeo) and

spliced RNA (env-IRES-Zeo) were significantly greater than pPAM3 gene expression, which only

exhibited spliced env transcripts. (B) Gene expression of LTKOSN and ALTKOSN vectors.

Hybridization of the same Northern blot membrane with Neo"^ probe to detect RNA transcripts of

LTKOSN (4.0 kb), ALTKOSN (2.5 kb) expressed fiom 5 LTR and a Neo"^ transcript (1.2 kb)

expressed from the Internal SV40 promoter. Fewer vector transcripts were retained In pAM3-

IRES-Zeo transfected cells since transcripts were packaged into virions. (C) Hybridization of the

same Northem blot membrane with human GAPDH cDNA probe demonstrates fairiy equivalent

RNA loading. 105

LTK0SN.2 sub-clones

pAM3-(RES-Zeo transfection

— unspliced

Helper Virus — env-IRES-Zeo

— env

LTKOSN

Vector ALTKOSN

neo

GAPDH

123456789 FIG. 3. DNA methylation of helper virus S LTR over time with and without Zeocin™ selection. (A) Schema of pAM3-IRES-Zeo helper virus showing restriction enzyme sites and the probe used for the methylation analysis. B, SsfEII; E, EcoRV; S, Sma\: AAA, SV40 poly

(A) signal. Drawing is not to scale. (B) Genomic DNA Southern blot membrane was probed with a 261-bp fragment excised from pAM3-IRES-Zeo with Kpn\ and Afl\\ digestions. If methylation was present at the Smal site, a 608-bp fragment would result instead of a 348- bp fragment The degree of DNA methylation was calculated as the intensity ratio of Sma\ insensitive band (608-bp) divided by the sum of the intensity of this 608-bp band and Sma\ sensitive fragment (348-bp). 107

E S B r-J—L I , , I LTR H 1 I P9l r env IIRES I Zeo I AAA 608 bp 348 bp • LTR/gag probe

B CO o ^ 1-

I I I 1 Zeocin Selection - - - + - + FIG. 4. The effectiveness of Zeocin™s election on helper virus gene expression in

AMIZ cells overtime. (A) Unspliced MoMLV transcript (gag-po/-env-IRES-Zeo) and spliced

RNA (env-IRES-Zeo) were detected by an env probe In cellular RNA extracted from AMIZ cells with and without continuous Zeocin™s election on Days 0, 15, 54 and 78. (B)

Hybridization with GAPDH cDNA probe demonstrates the relative RNA loading. 109

DayO Day 15 Day 54 Day 78 Zeocin selection unspliced

GAPDH FIG. 5. Gene expression of pAM3-IRES-Zeo helper virus and LEIN vector in AMIZ ceils transfected with LEIN vector. (A) Northern blot analysis of cellular RNA extracted from

AMIZ cells transfected with LEIN vector evaluated at passage 3 (day 0, lane 1) and on days

56 and 67 (lanes 2-5) by hybridization with env probe. Zeocin^ n d G418 selection resulted in greater levels of unspliced MoMLV RNA transcript {gag-pol-env-lRES-Zeo) and spliced

RNA transcripts (env-IRES-Zeo). (B) The level of LEIN vector was also greater In AMIZ cells under Zeocin ™and G418 selection. (C) Re-hybridization with GAPDH cDNA probe was used to demonstrate fairiy equivalent RNA loading. 111

LU _i

§5 Day 56 Day 67 Zeo/G418 Selection

9.5- unspliced 7.5- Helper Virus 4.4- env-IRES-Zeo 2.4-

B 4.4- Vector i§ #• f

GAPDH

1 2 3 4 5 FIG. 6. Drug selection eliminates DNA methylation of helper virus and vector from

VPC population. Genomic DNA extracted from AMIZ cells transfected with LEIN vector with

Zeocin™ and G418 or without drug selection on days 0 (lanes 3 and 4), 56 and 67 (lanes 5-

12) were first digested with Oral and EcoRV and divided into two equal portions. One portion was subjected to methylation-sensitive Smal restriction endonuclease digestion and the other without Smal digestion. (A) Hybridization of gag DNA probe to detect helper virus 5

LTR. Smal digestion reduced the 1.8-kb band (even numbered lanes 4-12) to 1.5-kb (odd numbered lanes 3-11). DNA from NIH3T3 cells was used to show the presence of endogenous retroviral elements (lanes 1 and 2). Since a 1.8-kb band was generated from endogenous retroviral element after Smal digestion (lane 1), the values for Smal resistance were measured by densitometry as the relative intensities of the 1.8-kb bands without Smal digestion, and 1.5-kb bands after Smal digestion. (B) GFP DNA probe was used to detect the 5 LTR of LEIN vector. Smal digestion reduced the 3.7-kb fragment to 3.4-kb, 2.7-kb and

2.4-kb fi'agments dependent upon the methylation status of Smal site In LEIN vector. (C) A

0.68-kb DNA fragment digested from gag gene of pPAM3 by Bs{K\ was used as a probe to detect a 1.2-kb band of endogenous retroviral element to demonstrate relative loading in paired Smal+/- digestions. (D) Schema of the helper virus and vectors showing the locations of restriction enzyme sites and probes used for methylation analysis. B, 8sO

EcoRV; 8, Smal; AAA, SV40 poly (A) signal. Drawings are not to scale. 113

Helper virus pAM3-IRES-Zeo LEIN vector

EES D D S E EES 8 EES 1 gag' I ' Ul I ' snv ^|'"^V^°Iaaa hrRT-rgSPPVi^ Neo Klt^ ——— 1-8 kb 3.7 kb 1.5 kb 3.4 kb • probe 2.7 kb 2.4 kb GFP probe

2 111 B I—CO CO g Day 56 Day 67 < Zeo/G418 Unselected Zeo/G418 Unselected

% Smal- 17% 0% 30% 0% 36% Resistance 114

CHAPTER 5. HIGH FREQUENCY RETROTRANSPOSITION OF RETROVIRAL

VECTORS IN CULTURED RETROVIRAL VECTOR PRODUCER CELLS

A paper to be submitted to Human Gene Therapy

Won-Bin Young^'^, Gary L. Lindbergh, and Charles J. Link, JrJ'^*

RUNNING TITLE: Retrotransposition in VPC

ABBREVIATIONS:

GaLV: gibbon ape leukemia virus LTR: long terminal repeat

MoMLV: Moloney murine leukemia virus RT: reverse transcriptase

RCR; replication competent retrovirus VPC: vector producer cell

Env: envelope

KEYWORDS: vector producer cell; genetic instability; gene therapy; retrovirus; retroviral vector; packaging cell; integration; superinfection; gibbon ape leukemia virus; Moloney murine leukemia virus; retrotransposition; reverse transcriptase.

'Human Gene Therapy Research Institute, John Stoddard Cancer Center, 1415 Woodland

Ave., Des Moines, Iowa 50309. ^Molecular, Cellular and Developmental Biology Program,

Iowa State University, Ames, Iowa 50011. ^Department of Animal Science, Iowa State

University, Ames, Iowa 50011. 'To whom reprint requests should be sent, (515) 241-8787;

Fax: (515) 241-8788; e-mail: [email protected]. 115

ABSTRACT

The superinfection of retroviral vectors into vector producer cells (VPC) is the major source of reverse transcriptase (RT) that may results in replication-competent retrovirus

(RCR) formation. PG13 packaging cells are based on murine NIH3T3 cells and generate pseudotyped retroviral virions containing gibbon ape leukemia virus (GaLV) Env proteins to improve gene transfer into hematopoietic cells. PG13 does not express GaLV receptor on cell surface and therefore is not susceptible to superinfection of its own pseudotyped GaLV vectors. This is considered an advantage for PG13 cells. In this report, we found evidence of intracellular retrotransposition of retroviral vectors in PG13. This finding suggests that the potential of RCR formation might occur inside PG13 cells despite a lack of retroviral vector superinfection. In this study, PG13 was transfected with a vector containing a neomycin selection marker which functions only after the intron sequence is excised during retrotransposition. The ratio of intracellular retrotransposition was determined by the presence of G418-resistant colonies and was calculated as high as 1 x 10 and 3.9 X 10'^ events per cell in PG13 and NIH3T3 cells expressing a gag-po/construct, respectively. In contrast, superinfection of this vector into PA317 occun-ed at 5.1 x 10 events per cell. A complementary study of serial deletion mutants of MoMLV on NIH3T3 cells expressing this retroviral vector demonstrates that the RT activity required for retrotransposition of retroviral vectors originates from the intracellular particles of helper virus within the VPC. These results indicate that retroviral vectors frequently replicate as retrotransposons in murine VPC regardless of the presence of the Env protein. Demonstration of endogenous reverse transcriptase (RT) activity in VPC suggests the potential for RCR formation during retrotransposition of vectors without superinfection. 116

INTRODUCTION

The presence of RCR is a key biosafety issue when using retroviral vectors in gene therapy applications. The formation of RCR has been shown to occur by abnormal template switches between helper virus and retroviral vector sequences during reverse transcription

(32, 41). Therefore, the prevention of RCR outbreak from VPC may require attempts to block or decrease RT enzyme activities inside of VPC.

There are two categories of RT enzyme activity inside of VPC. Exogenous viral RT is imported into cells by virus infection. In a cultured VPC, the only source of exogenous RT enzyme activity would be derived from the re-entry of virions produced by the VPC

(superinfection of vector). Endogenous RT is a combination of retrovlrus-encoded Pr180^®®'

or po/gene product of virion particles within the cell (2,18, 33, 38), endogenous retroviruses (10, 19) and retrotransposons (12, 16,17). These different sources of endogenous RT enzyme activity can result In intracellular retrotransposition activity

(reviewed In reference 22).

Exogenous RT activity can be limited by Interference between viral Env proteins and viral receptors on the cell surface (Env-receptor interference) (31). The re-infection ratio of amphotropic Moloney murine leukemia virus (Am-MoMLV) vector virions on target cells, which were previously infected with wild type Am-MoMLV, is as low as 10"^ events per cell on rat 208 cells and 10"® events per cell on NIH3T3 cells (26). In contrast to Am-MoMLV re infection, intracellular retrotransposition of an Env-defective C-type MoMLV has been reported at a higher frequency of 2.7 x 10^ events per cell on feline G355-5 cells (39).

The retrotransposition pathway of this Env-defective retrovirus includes the transcription of provlral vector (e.g. integrated MoMLV-based vector), RNA splicing, reverse transcription and re-integration of processed provlral vector (13, 38, 39). This process is similar to the other previously described events such as the amplification of endogenous 117 retroviais (10) and retrotransposons (4, 12, 36), and the fonnation of pseudogenes (24, 40).

Retrotransposition is only observed when cells were complemented with retroviral gag-pol genes. This indicates that the RT enzyme activity involved in retrotransposition of retroviral vector is derived from gag-pol and pol gene products (38, 39). This proposes a possible involvement of immature gag-pol and pol gene products in the formation of RCR in VPC without superinfection of vector.

A specific interaction between core proteins and Env is dependent on the N-terminal domain of matrix (MA) protein and C-terminal domain of the Env transmembrane subunit (8,

21). Env proteins have been demonstrated to direct the release of human immunodeficiency virus (HIV) (21). This Gag and Env interaction may therefore eliminate intracellular retrotransposition of an Env-intact retrovirus by accelerating the release of virion particles.

To explore the possibility of retrotransposition of retroviral vector in currently used VPC irrespective of Env expression, we choose PA317 (25) and PG13 (27) packaging cell lines as models to compare with gag-pol transfected NIH3T3 cells. PG13 was established fi-om murine NIH3T3 and expresses MoMLV gag-pol genes and a GaLV env gene for packaging xenotropic GaLV-pseudotyped vector. Since NIH3T3 cells do not express PiTI receptor for

GaLV infection, PG13 is not susceptible to superinfection of its own pseudotyped GaLV vector (27,44). PA317 is also based on NIH3T3 ceils, but susceptible to infection of amphotropic MoMLV vector that it produces. Therefore, RT enzyme activities inside of

PA317 will be derived from both re-infection of vector virions and endogenous RT provided by intracellular virion particles or a cellular source.

In this study, we compared RT enzyme activity inside of PA317 and PG13 cells with a vector containing a neomycin selection marker disrupted by an intron sequence. This selection marker only functions after the Intron sequence Is excised during retrotransposition that required RT to have occured. This comparison allows an estimation of the ratio of 118

exogenous RT and endogenous RT in a VPC which represents a retrovirus infected cell.

The difference of RT activity between PG13 and gag-pol transfected NIH3T3 cells can

therefore reveal the effect of the interaction between GaLV Env and MoMLV core proteins

on intracellular retrotransposition. The results from PG13 packaging cells indicate a possible

role of endogenous RT enzyme activity in RCR formation and genetic instability of retroviral

vectors in VPC. (W.-B. Young is a doctoral candidate at Iowa State University, and this work

represents partial fulfillment of dissertation requirement)

MATERIALS AND METHODS

Plasmid constructions. To construct pELNIH vector, a reporter cassette (neo-int)

was first isolated from pNeo-!nt (9) (kindly provided by D. Mager at Terry Fox Lab.,

Vancouver, Canada) by BamHl and Sail digestions and then inserted into a retroviral shuttle

vector, pBabe-Hyg (30) (kindly provided by H. Land, Imperial Cancer Research Center,

London, UK) for pBabe-NIH (Neo-intron-Hyg) construct. A fragment containing neo-int

cassettee, hygromycin resistant gene (Hyg""), SV40 promoter and 3 LTR was isolated from

pBabe-NIH by Eco47lll and Pvul digestions and then subcloned into Eco47lll and Pvul

digested pELNCX. This pELNCX was previously constructed by digesting pLNCX (28) (a

kind gift from A. D. Miller at Fred Hutchinson Cancer Research Center, Seattle, WA) with

Pvul and Seal and then ligation of this pLNCX into Pvull and Pvul digested pREPIO

(InVitrogen) which contains OriP/EBNA-1 for extrachromosomal replication.

To generate deletion mutants of MoMLV, a pREP-XC vector was constructed to express the mutants. To construct pREP-XC, pREPIO was digested by Xbal and C/al, and the restriction sites were filled in by Klenow enzyme (BRL, Life Technology Co.) and self- ligated to remove Ori/EBNA-1 sequences from pREPIO. Since pREP-XC contains a hygromycin selection marker and the mutants carried by this vector become selectable to 119

hygromycin in mammalian cell culture. An Env-deletion mutant of MoMLV was first

constructed byAflWl digestion of pPAM3 (25) (kindly provided by A. D. Miller) and DNA

fragment ends were filled in by Klenow enzyme followed by Nhe\ digestion to release a

5" LTR-gag-pol fragment. This 5 LJR-gag-pol fragment was then inserted into PvuW and Nhe\

digested pREP-XC after RSV promoter to result in pGP construct, therefore, the gag-pol

construct is driven by both 5 LTR of MoMLV and RSV promoter (Fig. 3). To construct pAG, a

gag-deletion mutant, pGP was digested by Mscl and Afl\\ to delete matrix (MA) and capsid

(CA) open reading frames (ORFs) fi-om the gag gene. After digestion sites were filled in by

Klenow enzyme and self-ligated, the nucleocapsid (NC) ORF remained intact and in ft-ame

with first ATG codon of Gag protein and the downstream pol gene. To construct pAP, a

deletion of RT region, pGP was digested by Stu\ and self-ligated. Downstream IN gene

remained in frame after the deletion of RT. To construct pAl, an IN mutant, pGP was partially

digested by C/al and then two extra nucleotides were inserted into the C/al site in IN gene

by Klenow enzyme fill-in to induce a frame-shift mutation of the IN gene. These mutants

were sequenced by the Nucleic Acid Facility at Iowa State University to ensure that the

ORFs are con-ect as anticipated after construction.

Human L1 retroelement, pJMIOI, is provided by J. V. Moran at U. of Pennsylvania.

This pJMIOI is selectable (hygromycin) and can replicate in the cell nucleus by an origin of

replication (OriP/EBNA) from pCEP4 plasmid in mammalian cell culture. This LI .2 element

(pJMIOI) has been demonstrated with detectable retrotransposition in human Hela cells

and mouse fibroblast cell line, LTK- (29).

Cell culture and transfection. Cell cultures were maintained in Dulbecco modified

Eagle medium (DMEM; GIBCO BRL, Life Technology Co., Gaithersburg, MD), 10% fetal calf

serum with 5% CO2, at 37°C. Transfections of pJMIOI, pELNIH, MoMLV mutants, pGP, pAG, pAP and pAl, were performed by liposome transfection reagent, DOTAP (Roche 120

Molecular Biochemlcals, Indianapolis, IN) with 5 ^ig of a single plasmid DNA or 2.5 ng of each of two co-transfected plasmids for each well in 6-well plates. Selection with

Hygromycin B (100 |ig/ml, Roche Molecular Biochemicals) started 48 hr after transfection and continued for at least two weeks. Selection with G418 (1 nng/ml; GIBCO BRL, Life

Technologies) for retrotransposed ELNIH vector was applied for 10-14 days on above hygromycin positive cell populations.

NIH3T3 cells expressing p-galatosidase were established by transfection of pZeo-

SV-LacZ (InVitrogene) and selection with Zeocin™ (350 |ag/ml) for two weeks. A single clone was isolated with functional p-galatosidase gene expression. NIH3T3 cells expressing p- galactosidase were mixed with pELNIH-transfected VPC at a 1:1 ratio and co-cultured for one to two weeks before G418 selection. After two weeks of G418 selection, G418-positive colonies were rinsed with phosphate-buffered saline (PBS) and fixed for 5 min at room temperature in 2% formaldehyde, 0.3% glutaraldehyde in PBS. Following two additional rinses with PBS, cells were stained with a chromophore solution containing 0.1% 5-bromo-4- chloro-3-indolyl-b-D-galactopyranoside (X-gal) (Promaga, Madison, Wl), 5mM

K4Fe(CN)6-3H20, 5mM K3Fe(CN)6 and 2mM MgCb in PBS at 37°C for ovemight. p- galactosidase positive cells were visualized blue under microscopic examination.

Southern blot analysis of genomic DNA. Total cellular DNA was extracted from cell cultures using phenol/chloroform extraction and then dissolved in TE buffer (pH 8.0) for ovemight at 55°C (35). To detect the integration locations of retrotransposed vector, genomic DNA was digested by Sphl and then evaluated by Southern blot analysis. A 0.35- kb ft-agment (Neo probe) isolated from pNeo-Int by SamHI and Sphl digestions was used to detect the integration sites of retrotransposed pELNIH (Fig. 2A). The same Southern blot membrane was also probed with a 0.5-kb ft-agment (PTK probe), isolated from pNeo-Int by 121

Sa/I and Sad digestions, to detect the transfected pELNIH containing intron sequence (1.8- kb) and retrotransposed ELNIH laci

RESULTS

Construction of a retroviral vector to detect retrotransposition. To detemnine whettier retroviral vector could retrotranspose in VPC, we constructed a retroviral vector, pELNIH, containing a reporter cassette (neo-int), designed to detect rare retrotransposition events (9) (Rg. 1A). This reporter cassette consists of an antisense copy of an antibiotic resistant marker gene (neo), a TK pronnoter with polyoma enhancer, and a polyadenylation signal. The neo gene is disrupted by an intron (IVS 2 of y-globin gene) in the opposite transcriptional orientation. The reporter gene, neo, will be expressed only after (i) the retrotransposable vector has been transcribed, (ii) the intron sequence has been removed by RNA splicing and (iii) the spliced transcript has been reverse transcribed and integrated into the genome (Fig. 1A). To distinguish the integration of retrotransposed vector from transfected vector, this pELNIH was also constructed with an origin of replication

(OriP/EBNA 1) to allow extrachromosomal replication. Similar cassettes have been employed to study the formation mechanism of pseudogenes (24, 40) and detect the intracellular transposition of an Env-defective MoMLV retrovirus (13, 38, 39), mouse intracistemal A particles (lAP) (4, 5, 12), long interspersed nucleotidic elements (LINE) (16,

17, 36) and human L1 element (7, 29).

Retrotransposition of retroviral vector in VPC. To detect retrotransposition in

VPC without the possibility of vector superinfection, we transfected pELNlH into PG13 packaging cells and co-transfected pELNIH and pGP, a MoMLV gag-pol construct (see

Materials and Methods) into NIH3T3 cells. PA317, an amphotropic packaging cell line susceptible to superinfection of its own vector, was also transfected with pELNIH for vector 122 superinfection study and to compare with PG13 and NIH3T3 cells transfected with pGP

(NIH3T3-GP). NIH3T3 transfected with pELNIH was used as a control to determine the baseline endogenous RT enzyme activity that has been observed previously with pseudogene formation (24, 40). Transfected cells were under hygromycin selection for 14 days. Hygromycin positive population was expanded for further retrotransposition assay by selection with G418 for 10-14 days. The number of G418-res(stant colonies in plated cells was used to calculate retrotransposition frequency (Fig. IB). In an initial experiment, 10® hygromycin-resistant cells derived fi'om PA317, PG13 and NIH3T3-GP cells were plated in medium containing G418 for 10 days, the colonies were too numerous to count. Therefore, the cell populations were diluted serially before G418 selection.

Results fi'om three experiments performed in triplicate (Table 1) showed a relatively high frequency of retrotransposition of pELNIH vector in PG13 cells (about 1 event in lO'' cells or 1 X 10^ per cell). No G418-resistant colonies were observed in NIH3T3 cells expressing pELNIH alone when up to 3 X 10® cells were plated. The retrotransposition frequency observed in NIH3T3-GP cells (38.8 X 10^ per cell) was significantly greater than that observed In PG13 cells (P< .001, N=9). In PA317 cells that are permissive to superinfection, G418-reistant colonies occurred at 5.1 X 10'^ per cell. This frequency reflects both exogenous RT and endogenous RT enzyme activity. The relative contribution of endogenous RT enzyme activity compared to total RT enzyme activity (exogenous RT from superinfection plus endogenous RT) can therefore be estimated by comparing the frequencies of G418-resistant colonies in NIH3T3-GP. PG13 with the frequency observed in

PA317 cells. The ratio of endogenous RT enzyme activity detected in NIH3T3-GP is about

7.8% of total RT observed in PA317, and, in PG13, the ratio is reduced to 0.2 % of total RT in PA317. The reduction of endogenous RT might be contributed by Gag-Env interaction in

PG13cells. 123

To ensure that G418 resistant foci observed in transfected PG13 and NIH3T3-GP

cells were not derived from nonspecific transmission of virions from neighboring cells, such

as cell-to-cell contact infection (23), pELNIH transfected cells were co-cultured with NIH3T3

cells expressing a p-galactosldase gene. After being co-cultured for one or two weeks, these

co-cultures were then subjected to G418 selection for retrotransposition. If non-specific

infection had occurred, p-galactosidase positive colonies would have been observed after

G418 selection. As anticipated, no blue colonies were observed after X-gal staining of G418

resistant colonies obtained from co-cultured NIH3T3 expressing p-galatosidase with either

pELNIH-transfected PG13 or NIH3T3-GP cells. This indicates that these G418 resistant

colonies authentically resulted from intracellular retrotransposition of pELNIH vector but not

a result of exogenous RT. In contrast, some p-galatosidase positive G418-resistant colonies

were obtained from the cells co-cultured with PA317/pELNIH. This demonstrates that the

infection of ELNIH vector on NIH3T3 cells expressing p-galatosidase did occur (data not

shown).

We also calculated the splicing efficiency of this neo-Int cassette by transducing

human melanoma A375 cells with supernatant collected from PG13 transfected with

pELNIH, and then selection with either hygromycin or G418. The ratio of G418-resistant

colonies (spliced vector) and hygromycin-resistant foci (spliced + unspliced vector) was

42%. The titer of PG13 transfected with pELNIH was approximately about 1X10^ cfu/ml.

This indicated that only 42% of vector transcripts packaged into virions were spliced. This

may reflect the vector populations inside of VPC. Since the complication caused by

sensitivity of different selections and possible altemative splicing of IVS sequence causes a

defective neo gene (see below), we did not recalculate our previous data of retrotransposition events with this splicing efficiency. However, this selection model with a 124 retrotransposible vector couid be a useful tool to study the splicing regulation of HIV transcripts or other cellular intron splicing sequences, such as IVS of y-globin in this case.

Retrotransposed vectors are integrated into the genomic DNA. To determine whether these retrotransposed ELNIH vectors were integrated into chromosomes or maintained in episomal fomi, individual G418 resistant colonies were isolated from PG13 cells transfected with pELNIH and NIH3T3-GP transfected with pELNIH. Genomic DNA was extracted from each individual clone and digested with Sph\, a DNA methylation-insensitlve endonuclease for complete digestion without DNA methylation interference. Southern blot analysis was performed with a Neo probe and revealed the integration patterns of retrotransposed ELNIH vector (Fig. 2A). Integrated copies of ELNIH within the cell clones were resolved in Southem blot analysis. The presence of multiple bands suggested that

ELNIH vector had retrotransposed into different chromosomal locations within one cell. In 23 clones isolated from PG13 cells transfected with pELNIH, 16 (70%) clones exhibited different integration patterns in Southem blot analysis. The copy number of retrotransposed

ELNIH vectors ranged from 1 to 6 copies with a mean of 2.3 copies per cell (Fig. 2B). In

NIH3T3-GP transfected with pELNIH cells, 13 out of 21 colonies (62%) exhibited different integration pattems and the copy numbers of retrotransposed ELNIH vectors ranged from 1 to 8 copies with a mean of 3.2 copies per cell (Fig. 2C). These Southem blot membranes were re-probed with a PTK probe (Fig. 2A) to detect both retrotransposed vector and transfected pELNIH. Only 3 out of 23 G418-resistant clones derived from PG13 cells transfected with pELNIH were detected with the Integrations of both transfected pELNIH (a

1.8-kb band with intron sequence after Sphl digestion) and retrotransposed ELNIH vector (a

0.9-kb intronless fragment after Sphl digestion) (Fig. 2B). Only one out of 21 G418-resistant colonies derived from NIH3T3-GP transfected with pELNIH exhibted with the integrations of both transfected pELNIH (1.8-kb) and retrotransposed ELNIH (0.8-kb, Fig. 20). This 125 suggests that the vast majority of detected integration sites are a result of retrotransposed

ELNIH but not from the integration of transfected episomal pELNIH vector. Of note, a band less than 0.9-kb was also detected by the PTK probe. This small band may indicate an low level of alternative splicing event of IVS sequence that results in a defective neo gene and therefore can only be observed in some G418-resistant colonies with multiple integrations of retrotransposed ELNIH.

Retrotransposltion of pELNIH vector requires gag-pol gene products. To test whether the G418 resistant foci after pELNIH transfection could be a result of other endogenous cellular RT enzyme activity, three mutants of MoMLV gag-pol region were constpjcted. pGP was conducted by deleting the envgene from pPAM3 helper virus. The first mutant (pAG) has a deletion of matrix protein and capsid protein ORFs from the gag region of pGP. The second mutant (pAP) contains a deletion of RT from pol region of pGP.

The third mutant (pAl) contains a frame-shifted mutation in the integrase (IN) ORF of pGP

(Fig. 3). These three mutants and pGP were co-transfected with pELNIH into NIH3T3 and selected with hygromycin. Hygromycin resistant populations derived from these transfected cells were then plated under G418 selection to determine retrotransposltion frequencies as before. We also co-transfected pAG, pAP and pELNIH together to determine if retrotransposltion occurs when capsid structural proteins and pol gene products are translated in trans. Hygromycin-resistant cells (3 X 10® to 5 X 10®) were plated in triplicate under G418 selection for retrotransposltion events. In three different experiments, retrotransposltion of pELNIH vector was only observed in cells transfected with pGP construct (about 23 events in 10® cells). No G418-resistant colonies were observed in

NIH3T3 cells transfected with mutant constructs or the combination of pAG and pAP plasmids. These results suggest that the RT enzyme activity required for retrotransposltion of pELNIH originates from MoMLV Pri 80^®®""°' but not from other cellular sources. 126

Retrotransposition of vectors likely requires normal particle assembly for vectors to be packaged inside of these intracellular particles, since complementation experiment of pAG and pAP in the same ceil did not show detectable retrotransposition activity.

Retrotransposition in VPC is vector specific. To study whether the observed RT enzyme activity of helper virus in these retroviral packaging ceils is specific to pELNIH vector, a plasmid, pJM101, containing a human LI element (L1.2) and the same neo-int cassette (7, 29) was used as an altemative retrotransposition assay in VPC. This pJM101 is selectable with hygromycin and can replicate in the nucleus extrachromosomally as our pELNIH vector. However, the LI element In pJMIOI is a non-LTR retrotransposon and encodes its own nucleic acid-binding protein and reverse transcriptase for functional retrotransposition (11,14). Retrotransposition activity of LI.2 element has been well demonstrated in human HeLa cells and mouse fibroblast cell line, LTK- cells (29).

In two experiments, similar frequencies of G418-resistant foci were observed in

PA317, PG13 and NIH3T3 cells transfected with pJMIOI (Table 2). Since no significant differences were observed in retrotransposition events between NIH3T3 and packaging cells, PA317 and PG13, we conclude that the retrotransposition of pJMIOI depends mostly on its own protein components and not helper virus present in PA317 and PG13 cells.

Therefore, the high frequency of retrotransposition observed in PA317, PG13 and NIH3T3-

GP cells transfected with pELNIH is retroviral vector specific (Table 1).

DISCUSSION

In this study, the retrotransposition of retroviral vectors has been demonstrated in

VPC. These retrotransposition events required the presence of helper virus gag-pol gene products and are vector specific. Retrotransposed vectors can integrate into different chromosomal locations within one cell. The fi-equencies of vector retrotransposition inside 127

VPC are approximately 100-fold lower than vector superinfection, however, demonstration

of retroti^nsposition inside VPC may imply that RCR could develop firom recombination from

vector and occasionally co-packaged helper virus transcripts within intracellular particles.

To distinguish endogenous RT (derived from gag-pol precursor protein) from

exogenous RT (mature RT activity imported by superinfection) inside of a virus-infected cell

has practical difficulties. However, by comparing retrotransposition activity in PG13 and

NIH3T3-GP to the total RT activity in PA317 cells, the percentage of endogenous RT activity

that caused retrotransposition was estimated at 0.2% in PG13 and 7.8% in NIH3T3-GP

cells. The increase of percentage in NIH3T3-GP is significant and may reflect more

intracellular particles retained inside of cells compared to PG13 cells. If s important to note

that NIH3T3-GP is a population of cells and not an individual cell clone like PG13.

Therefore, we anticipate that the endogenous RT enzyme activity in NIH3T3-GP cells

should be greater in an NIH3T3-GP clone with high gene expression. The elevation of RT

enzyme activity in NIH3T3-GP cells will be expected to enhance effects of Env-Gag

interaction on the reduction of vector retrotransposition.

How Env-Gag interaction might affect retrotransposition of retroviral vectors needs

more detailed study. This question is interesting since assembly and release of a virion

particle does not require Env protein (reviewed in reference 15). However, Gag-Env

Interaction has been observed in dorsal root ganglion (DRG) cells expressing both Gag and

Env proteins of MoMLV and HIV respectively. Gag proteins were found to be co-localized

with Env proteins in the somatic region of DRG cells and not scattered within entire DRG

cell. Apparently, this interaction is specific and therefore was not observed from the heterologous pairs of Env and Gag proteins from HIV and MoMLV (42). HIV gp41 also has been reported to induce polarized budding of HIV from MDCK cells (21). These observations suggested that Env-Gag Interaction might attract assembling or assembled 128

virion particles with viral genome transcripts toward the cellular membrane and reduce the

level of virion particles within the cells. Such a theory could explain our observation of

reduced retrotransposition in PG13 cells compared to NIH3T3-GP cells. Although the

demonstration of interaction between GaLV Env proteins and Gag proteins from MoMLV

construct was not conducted in this study, several prior studies have indicated a possible

interaction between GaLV Env and MLV Gag proteins. First, GaLV was originally transmitted

to primates by rodents and not from a human retrovirus (20, 34). This is supported by the homology between GaLV and MLV in nucleic acid and protein sequences (1, 3) and an unusual degree of immunological cross-reactivity between the capsid proteins and RT of

GaLVs and MLVs (37, 45). In gene transfer experiments, the packaging signals of both

GaLV and MLV can be recognized by either GaLV or MLV core proteins (6). When GaLV

Env protein is used to pseudotype MoMLV core particles in PG13 (27) and other cells (6,

43), fairiy equivalent titers to PA317 are observed. In contrast, pseudotyping with HIV Env protein does not complement MoMLV core particles and produce infectious virions (43).

Taken together, these studies indicate that GaLV Env protein can specifically interact with

MoMLV Gag proteins. In this study, we also observed that this interaction might affect the frequency of vector retrotransposition. The outcomes of this study imply that an exogenous retrovirus can process Intracellular retrotransposition, a shortcut of replication cycle, In addition to cell-to cell infection.

ACKNOWLEDGEMENTS

We thank Dr. John V. Moran for helpful discussions and plasmid JM101. This work was supported in part by a grant from the Iowa Health System, Des Moines, Iowa. W.-B.

Young is a recipient of a competitive research fellowship from Molecular Cellular and

Developmental Biology Program, Iowa State University, Ames, Iowa. 129

REFERENCES

1. Benveniste, R. E., and G. Todaro. 1973. Homology between type-C viruses of various

species as determined by molecular hybridization. Proc Natl Acad Sci USA. 70:3316-

3320.

2. Crawford, S., and S. P. Goff. 1985. A deletion mutation In the 5' part of the pol gene of

Moloney murine leukemia virus blocks proteolytic processing of the gag and pol poly-

proteins. J. Virol. 53:899-907.

3. Delassus, S., P. Sonigo, and S. Wain-Hobson. 1989. Genetic organization of gibbon

ape leukemia virus. Virology. 173:205-213.

4. Dupressoir, A., and T. Heidmann. 1996. Germ line-specific expression of intracistemal

A-particle retrotransposons in transgenic mice. Mol Cell Biol. 16:4495-503.

5. Dupressoir, A., A. Puech, and T. Heidmann. 1995. lAP retrotransposons in the mouse

liver as reporters of ageing. Biochim Biophys Acta. 1264:397-402.

6. Eglitis, M. A., R. D. Schneiderman, P. M. Rice, and M. V. Eiden. 1995. Evaluation of

retroviral vector based on the gibbon ape leukemia virus. Gene Ther. 2:486-492.

7. Feng, Q., J. V. Moran, H. H. Kazazian, Jr., and J. D. Boeke. 1996. Human L1

retrotransposon encodes a conserved endonuclease required for retrotransposition. Cell.

87:905-16.

8. Freed, E. O., and M. A. Martin. 1995. Virion incorporation of envelope glycoproteins

with long but not short cytoplasmic tails is blocked by specific, single amino acid

substitutions in the human immunodeficiency virus type 1 matrix. J Virol. 69:1984-9.

9. Freeman, J. D., N. L. Goodchild, and D. L. Mager. 1994. A modified indicator gene for

selection of retrotransposition events in mammalian cells. Biotechniques. 17:46, 48-9,

52. 130

10. Goodchild, N. L., J. D. Freeman, and D. L. Mager. 1995. Spliced HERV-H

endogenous retroviral sequences in human genomic DNA: evidence for amplification via

retrotransposition. Virology. 206:164-73.

11. Hattori, M., S. Kuhara, O. Takenaka, and Y. Sakaki. 1986. LI family of repetitive DNA

sequences in primates may be derived from a sequence encoding a reverse

transcriptase-related protein. Nature. 321:625-8.

12. Heidmann, O., and T. Heidmann. 1991. Retrotransposition of a mouse lAP sequence

tagged with an indicator gene. Cell. 64:159-70.

13. Heidmann, T., O. Heidmann, and J. F. Nicolas. 1988. An indicator gene to

demonstrate intracellular transposition of defective retroviruses. Proc Natl Acad Sci U S

A. 85:2219-23.

14. Hoimes, S. E., M. F. Singer, and G. D. Swergold. 1992. Studies on p40, the leucine

zipper motif-containing protein encoded by the first open reading frame of an active

human LINE-1 transposable element. J Biol Chem. 267:19765-8.

15. Hunter, E, 1994. Macromolecular interactions in the assembly of HIV and other

retroviruses. Semi Virol. 5:71-83.

16. Jensen, S., M. P. Gassama, and T. Heidmann. 1994. Retrotransposition of the

Drosophila LINE I element can induce deletion in the target DNA; a simple model also

accounting for the variability of the normally observed target site duplications. Biochem

Biophys Res Commun. 202:111-9.

17. Jensen, S., and T. Heidmann. 1991. An indicator gene for detection of germline

retrotransposition in transgenic Drosophila demonstrates RNA-mediated transposition of

the LINE I element. EMBO J. 10:1927-37. 131

18. Katoh, I., Y. Yoshinaka, A. Rein, M. Shybuya, T. Odaka, and S. Oroszlan. 1985.

Murine leukemia virus mutations: protease region required for conversion from

"immature" to "mature" core form and for virus infectivity. Virology. 63:280-292.

19. Lefebvre, S., B. Hubert, F. Tekaia, M. Brahic, and J. F. Bureau. 1995. Isolation from

human brain of six previously unreported cDNAs related to the reverse transcriptase of

human endogenous retroviruses. AIDS Res Hum Retroviruses. 11:231-7.

20. Lieber, M. M., C. J. Sherr, G. J. Todaro, R. E. Benveniste, R. Callahan, and H. G.

Coon. 1975. Isolation from the asian mouse Mus caroll of an endogenous type 0 virus

related to infectious primate type 0 viruses. Proc Natl Acad Sci USA. 72:2315-9.

21. Lodge, R., H. Gottlinger, D. Gabuzda, E. A. Cohen, and G. Lemay. 1994. The

intracytoplasmic domain of gp41 mediates polarized budding of human

immunodeficiency virus type 1 in MDCK cells. J Virol. 68:4857-61.

22. Lower, R., J. Lower, and R. Kurth. 1996. The viruses In all of us; characteristics and

biological significance of human endogenous retrovirus sequences. Proc Natl Acad Sci

USA. 93:5177-84.

23. Lynch, W. P., W. J. Brown, G. J. Spangrude, and J. L. Portis. 1994. Microglial

infection by a neurovirulent murine retrovirus results in defective processing of envelope

protein and Intracellular budding of virus particles. J Virol. 68:3401-9.

24. Maestre, J., T. Tchenio, O. Dhellin, and T. Heidmann. 1995. mRNA retropositlon in

human cells: processed pseudogene formation. EMBO J. 14:6333-8.

25. Miller, A. D., and C. Buttimore. 1986. Redesign of retrovirus packaging cell lines to

avoid recombination leading to helper virus production. Moi Cell Biol. 6:2895-902.

26. Miller, A. D., and F. Chen. 1996. Retrovirus packaging cells based on 10A1 murine

leukemia virus for production of vectors that use multiple receptors for cell entry. J Virol.

70:5564-71. 132

27. Miller, A. D., J. V. Garcia, N. Von Shur, C. M. Lynch, C. Wilson, and M. V. Eiden.

1991. Construction and properties of retrovirus packaging cells based on gibbon ape

leukemia virus. J Virol. 65:2220-4.

28. Miller, A. D., and G. J. Rosman. 1989. Improved retroviral vectors for gene transfer and

expression. Biotechniques. 7:980-2, 984-6, 989-90.

29. Moran, J. V., S. E. Holmes, T. P. Naas, R. J. DeBerardinis, J. D. Boeke, and H. H.

Kazazian, Jr. 1996. High frequency retrotransposition in cultured mammalian cells. Cell.

87:917-27.

30. Morgenstem, J. P., and H, Land. 1990. Advanced mammalian gene transfer high titer

retroviral vectors with multiple drug selection markers and a complementary helper-free

packaging cell line. Nucleic Acids Res. 18:3587-96.

31. Odawara, T., M. Oshima, K. Doi, A. Iwamoto, and H. Yoshikura. 1998. Threshold

number of provirus copies required per cell for efficient virus production and interference

in moloney murine leukemia virus- infected NIH 3T3 cells. J Virol. 72:5414-24.

32. Otto, E., A. Jones-Trower, E. F. Vanin, K. Stambaugh, S. N. Mueller, W. F.

Anderson, and G. J. McGarrity. 1994. Characterization of a replication-competent

retrovirus resulting fi-om recombination of packaging and vector sequences. Hum Gene

Ther. 5:567-75.

33. Peng, C., B. K. Ho, T. W. Chang, and N. T. Chang. 1989. Role of human

immunodeficiency virus type 1-specific protease in core protein maturation and viral

infectivity. J Virol. 63:2550-6.

34. Reitz, M. S., Jr., M. Voltin, and R. C. Gallo. 1980. Characterization of a partial provirus

from a gibbon ape naturally infected with gibbon ape leukemia virus. Virology. 104:474-

81. 133

35. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory

manual., p. 9.14 - 9.23, Chapter 9: Analysis and cloning of eukaryotic genomic DNA,

2nd. ed. Cold spring Harbor Laboratory Press, Cold Spring Harbor, N. Y.

36. Segal-Bendirdjian, E., and T. Heidmann. 1991. Evidence for a reverse transcription

intermediate for a marked LINE transposon in tumoral rat cells. Biochem Biophys Res

Commun. 181:863-70.

37. Sherr, 0. J., L. A. Fedele, R. E. Benveniste, and G. J. Todaro. 1975. Interspecies

antigenic determinants of the reverse transcriptases and p30 proteins of mammalian

type C viruses. J Virol. 15:1440-8.

38. Tchenio, T., and T. Heidmann. 1991. Defective retroviruses can disperse in the human

genome by Intracellular transposition. J Virol. 65:2113-8.

39. Tchenio, T., and T. Heidmann. 1992. High-frequency intracellular transposition of a

defective mammalian provirus detected by an in situ colorimetric assay. J Virol. 66:1571-

40. Tchenio, T., E. Segal-Bendirdjian, and T. Heidmann. 1993. Generation of processed

pseudogenes in murine cells. EMBO J. 12:1487-97.

41. Vanin, E. F., M. Kaloss, 0. Broscius, and A. W. Nienhuis. 1994. Characterization of

replication-competent retroviruses from nonhuman primates with virus-induced T-cell

lymphomas and observations regarding the mechanism of oncogenesis. J Virol.

68:4241-50.

42. Weclewicz, K., M. Ekstrom, K. Kristensson, and H. Garoff. 1998. Specific interactions

between retrovirus Env and Gag proteins in rat neurons. J Virol. 72:2832-45.

43. Wilson, C., M. S. Reitz, H. Okayama, and M. V. Eiden. 1989. Formation of infectious

hybrid virions with gibbon ape leukemia virus and human T-cell leukemia virus retroviral 134

envelope glycoproteins and the gag and pol proteins of Moloney murine leukemia virus.

J Virol. 63:2374-8.

44. Wilson, C. A., K. B. Farreil, and M. V. Eiden. 1994. Comparison of cDNAs encoding

the gibbon ape leukaemia virus receptor from susceptible and non-susceptible murine

cells. J Gen Virol. 75:1901-8.

45. Wong-Staal, F., R. C. Gallo, and D. Gillespie. 1975. Genetic relationship of a primate

RNA tumour virus genome to genes in normal mice. Nature. 256:670-72. 135

TABLE 1. Retrotransposition frequencies of pELNIH vector in cultured VPC

Experiment

1 2 3

PA317 423 X 10"^ 468X10"* 639X10"*

PG13 1 XIO"^ 1 XIO"* 1 X 10"*

NIH3T3-GP 48X10"^ 18X10"* 51 X10"*

NIH3T3 <1 XIO"® <3X10"® < 1 X 10'®

TABLE 2. Retrotransposition frequencies of fiuman LI retroelement (pJMIOI) in VPC

Experiment

1 2

PA317 54X10"® 36X 10"®

PG13 58 X 10"® 9X10"®

NIH3T3 67X10"® 43X10"® FIG. 1. Retrotransposltion assay with pELNIH vector. (A) An overview of retrotransposition process of pELNIH vector. An indicator cassette containing an antisense copy of TK promoter (P) with the neo gene disrupted by intron 2 (IVS) of the y-globin gene was incorporated into pELNIH vector. The splicing donor (SD) and splicing acceptor (SA) of the intron (IVS) are indicated. A polyadenylation signal (A') is also in antisense orientation at

3 end of neo gene. A hygromycin resistant gene (Hyg) is driven by both SV40 promoter

(SV40) and a proi

(Ori) enables the cloning of integration sites of retrotransposed pELNIH. Transfected pELNIH could replicate extrachromosomally by a mammalian replication origin (OriP/EBNA

1). Transcripts originating from 5 LTR can splice the intron, but contain an antisense copy of the neo gene. G418-resistant (G418'') colonies should arise only after this transcript is reverse transcribed into double-stranded DNA, integrated into chromosomes, and expressed by its own TK promoter (P). (B) Outline of retrotransposition assay. Hygromycin- resistant (Hyg'') cell population was established on pELNIH-transfected cells and plated at different densities for G418 selection. G418-resistant foci represent retrotransposition events. 137

-fp^ OriP" EBNA1 AmpR G418^

HygR so SA LTR oe ^1 IVS > t\ |

(SV40>r*"Hyg

Transcription

-AAA I RNA Spliang -AAA

Reverse Transcription I Integration G418'^ Ori LTR 09t\ LTR Chrosomal Chrosomal Hygf DNA

(I)TRANSFECTION (2) HYGROMYCIN SELECTION (3) RETROTRANSPOSITION (4) G418 SELECTION FOR CELLS WITH pELNIH (+)

pELNIH^^^ G418'^ G418'^

HygR I i I

G418S FIG. 2. Retrotransposed pELNIH vector integrated into different chromosomal locations. (A) Scheme of pELNIH and probes for detecting integrations of retrotransposed pELNiH and transfected pELNIH. After Sph[ digestion, a PTK probe (0.5-kb of Sal\/Sac\ fragment) detects a 1.9-kb fragment of transfected pELNIH with intron sequence and a 0.8- kb fragment of retrotransposed pELNIH without intron. A Neo probe (0.35-kb of BamHUSpM fragment) detects 5 LTR with adjunct chromosomal DNA. (B) DNA extracted from G418- resistant colonies isolated from PG13 transfected with pELNIH was digested with Sph\ and sequentially detected by either Neo probe or PTK probe on the same Southern blot membrane. Multiple bands on Neo probe detection represent multiple integrations at different chromosomal locations and copies of retrotransposed ELNIH in one cell were estimated from the multiple bands. These 1.9-kb and 0.8-kb bands are marked on Southern blot detected by PTK probe for transfected pELNIH and refrofransposed ELNIH vectors, respectively. A 1.4-kb band (indicated by "•") detected by PTK probe arises from the other

TK promoter used for HStk selection marker gene in PG13, but is not seen in DNA extracted from NIH3T3-GP cells transfected with pELNIH (see panel C). A band smaller than 0.9-kb was detected on some of colonies and might originate from an alternative splicing of IVS intron. (C) DNA extracted from NH3T3-GP cells transfected with pELNIH was digested by

SphI and sequentially detected by either Neo probe and PTK probe on the same Southern blot membrane. The integration of transfected pELNIH was only observed in one G418- resistant colony. 139

BamHI Sphi Sad Sail SphI so SA Ori LTR 09 ] IVS > n K^SV40>r*"Hyg" LTR

— I.Skb Neo probe I PTK probe

SphI SphI SphI

Ori Chrosomal LTR 09t\ LTR DNA Chrosomal 0.9 kb DNA

Neo PTK

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 kb kb 12.0 — 12.0 — 5.0 _ Neo 5.0 — 4.0 ~ 4.0 — probe 3.0 — 3.0 — 2.0 — 2.0 — 1.6 ~ 1.6 -

1.8 PTK 1.6 — probe 1.0 — 0.9 0.9

20 21 12.0 —

5.0 _ 4.0 — probe 3.0 3.0 — 2.0 — 1.6 ""

• 1.8 PTK I 1.6 — probe 1.0- 1.0 — -0.9 FIG. 3. Scheme of serial deletion constaicts of MoMLV. These constaicts were carried by a modified pREP-10 (see Materials and Methods) and driven by RSV promoter and MoMLV 5 LTR. In pGP construct, env gene was deleted by AflWl restriction endonuclease and subcloned into modified pREP-10. Following deletion mutants are based on pGP. In pAG construct, MA and CA genes were deleted by Ami and Msc\ and then digestion sites were filled in to keep downstream ORFs in frame. In pAP, major portion of RT gene was deleted by Stu\ digestion without changing ORF of IN gene. In pAl construct, frame shifting of IN gene was conducted by insertion of two nucleotides into the C/al digestion site located at IN gene. 141

MscI StuI StuI Clal Afl III L IN pPAM3 EQ^EBOa RT r env lAAA I gag pol

pGP [RS^LTR I 1 gag | poi HAAAI

pAG [RS^LTRI INCT pol HTAAAI

pAP [RS^ LTR 1—H gag IPRIARTI IN HAAAI

pAl |RSV>| LTR gag IPRI ~ IN AAAJ 142

CHAPTER 6. GENERAL CONCLUSIONS

Human gene transfer has been demonstrated to be feasible and can evoke biologic responses to human diseases. Most gene transfer clinical trials use retroviral vectors for the advantages of sustained vector production and integration of vectors. Although adverse events have been uncommon, safety considerations are an ever present concern. A risk associated with the use of retroviral vectors is the inadvertent generation of replication- competent retroviruses (RCR) which may result from the recombination events between vector and the helper virus sequences present in packaging cell lines (1, 12, 15). The untoward effect of RCR is viremia (2) which in the long term can lead to neoplastic changes resulting from insertional mutagenesis (5). If s important to note that the integration of provirus is relatively stable and the mutagenesis caused by these events is a driving force of evolution. T-cell lymphoma has been observed in monkeys receiving bone marrow transplants intentionally infected with RCR (3, 15). Circulation of RCR from the blood stream into gonads may genetically alter germ cells, which allows RCR themselves with the mutations to be transmitted to generations. RCR can therefore be spread into human population by these individuals and may raise the risk of public health.

Detection and prevention of RCR break-out has always been a major biosafety concern during preparation of retroviral vectors. Unfortunately, RCR occurs quite often and has been reported by several groups (1, 3, 12, 15) including our own. RCR originates from template switching between helper virus and vector transcripts. The template switching is caused by RT enzyme activity when a helper virus transcript is occasionally co-packaged into a virion particle with vector. The target cell for this potential RCR is the VPC itself in the cell culture or the human cells in vivo. VPC usually are not susceptible as a target for its own virus infection because the receptors on the VPC surface are supposedly occupied by viral 143

Env proteins on cell membrane thus preventing superinfection (Env-receptor interference).

In this dissertation study, DNA methyiation was found to play an important role in superinfection of VPC by its own vector through inactivation of helper virus gene expression thereby reducing Env-receptor interference (Figure 1). The copy number of Integrated vector is also increased in VPC by superinfection and this causes variant genotypes in VPC population. The study of cell lineage in VPC cell culture initiated by continuous vector integration might be used to elucidate genome evolution of mammalian species by retrovirus infection.

Is DNA methyiation activity increased specifically by intruding viral sequences or just a common phenomenon In a cell? In LTKOSN.2 VPC and Its subclones, DNA methyiation is only detected In helper virus sequences, but not vector. Furthermore, comparing the DNA methyiation status of the RT gene and the 5 LTR show a methyiation preference for the

5 LTR over the RT region. However, the observations of DNA methyiation preference on helper virus over vector may be secondary to chromosomal locations. After a mixed population of VPC was established by using our chimeric retroviral helper virus (pAMS-

IRES-Zeo) and a vector, we examined the DNA methyiation status of both helper virus and vector. We did not observe significant differences of DNA methyiation between these two sequences. However, we should acknowledge that the method employed in this study is only detecting one or two restriction sites in approximately 8-kb of sequences and does not analyze DNA methyiation at all CpG islands on helper virus and vector. Several studies have indicated that DNA methyiation is a cellular defense system against retroelements (11,

16, 17) and can be activated by viral infection (10). Serial deletion mutants of MoMLV constructed with IRES sequence and a selection marker will be useful to understand which region of MoMLV sequence Is responsible for the observed differences in DNA methyiation status between LTRs. 144

Env-receptor interference Supennfectfon

I Vector production

*assembling \ ^ fI f.

Env protein PiiSO'™^ Vector integration

Transcnption

DNA Methylation

Helper virus

VPC

FIG. 1. Scheme of cascade reactions initiated by increased DNA methylation of helper virus

5 LTR. DNA methylation inactivates helper virus gene expression and then vector production is reduced. Without sufficient Env-receptor interference, susceptibility to superinfection of vector from its own or other VPC in the same population is increased. 145

Since increased DNA metliylation leads to superinfection that nnight permit RCR

formation, removing methylated helper virus from VPC populations should reduce RCR

formation. A chimeric retroviral helper virus consisting of MoMLV helper virus, picomavirus

IRES sequence and a selection marker gene was constructed and used to demonstrate that

helper virus gene expression can be maintained through selection pressure with the cap-

independent translation mechanism of IRES sequence. Under continuous selection,

superinfection was not detected in packaging cells established by using this chimeric helper

virus. The potential use of this chimeric helper virus is to establish a packaging cell

population from desired cell type without using a subcloning procedure. This is important

while transplantation of VPC fn vivo is required for sustained gene transfer. Transplanted

VPC need to evade host immune surveillance and therefore the best candidate for

establishing VPC might be a primary cell culture from an individual to be treated. To

demonstrate the usefulness of chimeric helper virus, we established a human packaging cell

population from this helper virus and demonstrated that the vector packaged by human

cells, but not by murine cells, is resistant to normal human serum that contains antibody

against the a1-3 galactosyl epitopes on glycoproteins on non-primate cells (Appendix A).

Superinfection assay on PA317 and LTKOSN.2 VPC demonstrated that only 0.2% to

0.4% of population of PA317 and LTKOSN.2 VPC are susceptible to superinfection. These

cells were exposed to GFP-expression vector only once in the experiment and do not represent culture conditions of VPC. An intron-containing vector used to detect retrotransposrtion activities in VPC is a very useful tool to determine superinfection in cell culture. The results show that about 10% of PA317 population is susceptible to superinfection when PA317 cells are consistently exposed to their own vectors. In addition to the RT enzyme activities imported by superinfection, intracellular virion particles also exhibit RT enzyme activity at a level equal to 7.8% of RT enzyme activity detected in 146 matured virions. The detection of RT enzyme activity in VPC with the evidence of vector retrotransposition poses a risk that RCR might occur from an intracellular virion particle instead of superinfection. The cause of RCR from retrotransposition events compared to superinfection events should be proportional to the ratio of endogenous RT enzyme activity.

MoMLV is classified in the category of exogenous type C retrovirus which does not exhibit retrotransposition during their replication cycle (8). Retrotransposition has only been observed in an Env-defective MoMLV (14). With the presence of GaLV Env protein in PG13 packaging cells, we found that MoMLV can process retrotransposition as a retrotransposon albeit at lower frequency (one log reduction). This provokes a possibility that exogenous retroviruses also use the retrotransposition route for replication as their ancestors do (Figure

2). Formation of particle-like structures of HIV in the cytoplasm without detectable extracellular release of viral particles has been observed after deletion of the N-terminal 15 amino acids of the MA protein (13). Observed retrotransposition of retrovirus might be resulted from intracellular budding of virion particles into cellular vacuoles (4, 9) caused by either mutations in matrix protein (4) or defective, unprocessed Env proteins (9). Therefore, the high frequency of retrotransposition observed in this study may partially be contributed by the errors arising from translation of Gag proteins.

Taken together, an improved packaging cell line for human gene transfer should be based on a human cell line, have a selectable helper virus to eliminate host DNA methylation for high gene expression and vector production, and would package a xenotropic vector that does not superinfect its own VPC. A helper virus with regulated RT will be necessary to prevent RCR formation from intracellular retrotransposition of viral sequences. Mutagenesis of RT to generate a temperature-sensitive (ts) mutant (7) will permit the regulation of RT enzyme activity within a VPC. An HIV variant with a temperature- 147

Chromosomal DNA Vector RNA Buddin(

Retrotranspositioir Vector

Provirus Reverse Integration DNA Transcription

FIG. 2. Alternative pathway of retroviral replication. A viral particle can retain inside host cell and reverse transcribe its RNA genome into double-stranded DNA (provirus DNA) for integration without being budded off host cell. This process enables retrovirus to amplify its copy number intracellularly as observed with other retroelements. 148 sensitive RT mutant (6) has been constaicted and can replicate at 34.5°C but not 39.5°C.

With the same theory, a (ts) mutant of RT can therefore be constructed to be active in 37°C for in vivo gene transfer and inactive in higher or lower temperatures for in vitro cell culture and vector production. A VPC containing a (ts) RT mutant should be useful for safer gene transfer application and viral replication study.

REFERENCES

1. Chong, H., and R. G. Vile. 1996. Replication-competent retrovirus produced by a'split-

function' third generation amphotropic packaging cell line. Gene Ther. 3: 624-629.

2. Cornetta, K., R. C. Moen, K. Culver, R. A. Morgan, J. R. Mclachlin, S. Strum, J.

Selegue, W. London, R. M. Blaese, and W. F. Anderson. 1990. Amphotropic murine

leukemia retrovirus is not an acute pathogen for primates. Hum Gene Ther. 1:15-30.

3. Donahue, R. E., S. W. Kessler, D. Bodine, K. McDonagh, C. Dunber, S. Goodman,

B. Agricola, E. Byrne, M. Raffeld, R. Moen, J. Bacher, K. M. Zsebo, and A. W.

Nienhuis. 1992. Helper virus induced T cell lymphoma in nonhuman primates after

retroviral mediated gene transfer. J Exp Med. 176:1125-35.

4. Freed, E. O., and M. A. Martin. 1995. Virion incorporation of envelope glycoproteins

with long but not short cytoplasmic tails is blocked by specific, single amino acid

substitutions in the human immunodeficiency virus type 1 matrix. J Virol. 69:1984-9.

5. Gray, D. A. 1991. Insertional mutagenesis: neoplasia arising from retroviral integration.

Cancer Invest. 9:295-304.

6. Huang, M., R. Zensen, M. Cho, and M. A. Martin. 1998. Construction and

characterization of a temperature-sensitive human immunodeficiency virus type 1

reverse transcriptase mutant. J Virol. 72:2047-54. 149

7. Kim, B., T. R. Hathaway, and L. A. Loeb. 1996. Human immunodeficiency virus

reverse transcriptase. Functional mutants obtained by random mutagenesis coupled with

genetic selection in Escherichia colL J Biol Chem. 271:4872-8.

8. Lower, R., J. Lower, and R. Kurth. 1996. The viruses in all of us: characteristics and

biological significance of human endogenous retrovirus sequences. Proc Natl Acad Sci

USA. 93:5177-84.

9. Lynch, W. P., W. J. Brown, G. J. Spangrude, and J. L. Portis. 1994. Microglial

infection by a neurovirulent murine retrovirus results in defective processing of envelope

protein and intracellular budding of virus particles. J Virol. 68:3401-9.

10. Mikovits, J. A., H. A. Young, P. Vertino, J. P. Issa, P. M. Pitha, S. Turcoski-Conrales,

D. D. Taub, C. L. Petrow, S. B. Baylin, and F. W. Ruscetti. 1998. Infection with human

immunodeficiency virus type 1 upregulates DNA methyltransferase, resulting in de novo

methylation of the gamma interferon (IFN-gamma) promoter and subsequent

downregulation of IFN- gamma production. Mol Cell Biol. 18:5166-77.

11. O'Neill, R., M. O'Neill, and J. A. Graves. 1998. Undermethylation associated with

retroelement activation and chromosome remodelling in an interspecific mammalian

hybrid. Nature. 393:68-72.

12. Otto, E., A. Jones-Trower, E. F. Vanin, K. Stambaugh, S. N. Mueller, W. F.

Anderson, and G. J. McGarrity. 1994. Characterization of a replication-competent

retrovirus resulting from recombination of packaging and vector sequences. Hum Gene

Ther. 5:567-75.

13. Spearman, P., J. J. Wang, N. Vander Heyden, and L. Ratner. 1994. Identification of

human immunodeficiency virus type 1 Gag protein domains essential to membrane

binding and particle assembly. J Virol. 68:3232-42. 150

14. Tchenio, T., and T. Heidmann. 1991. Defective retroviaises can disperse in the human

genome by intracellular transposition. J Virol. 65:2113-8.

15. Vanin, E. F., M. Kaloss, C. Broscius, and A. W. Nienhuis. 1994. Characterization of

replication-competent retroviruses from nonhuman primates with virus-induced T-cell

lymphomas and observations regarding the mechanism of oncogenesis. J Virol.

68:4241-50.

16. Yoder, J. A., and T. H. Bestor. 1996. Genetic analysis of genomic methylation patterns

In plants and mammals. Biol Chem. 377:605-10.

17. Yoder, J. A., C. P. Walsh, and T, H. Bestor. 1997. Cytosine methylation and the

ecology of intragenomic parasites. Trends Genet. 13:335-40. 151

APPENDIX A. HUMAN RETROVIRAL PACKAGING CELLS BASED ON A

CHIMERIC HELPER VIRUS FOR HIGH VECTOR PRODUCTION

Won-Bin Young and Charles J. Link, Jr.

The inhibition of the infectivity of murine type C retrovirus by human serum is mediated by human complement against a carbohydrate epitope from other mammals (3).

This epitope is synthesized by an enzyme, a-(1,3)gaiactosyltransferase. The gene encoding this enzyme in human is a firameshifted mutant, therefore, human cells do not express this carbohydrate epitope (1). Pre-existing humoral immunity in humans recognize this epitope as a foreign antigen and neutralize invading micro-organisms such as bacteria and viruses.

Retroviral vectors packaged by murine cells express those epitopes on their envelope proteins, which are susceptible to human serum inactivation. After being exposed to human serum, vector titer could therefore be reduced more than 90% (2). In contrast, retroviral vectors packaged from human cells are resistant to inactivation by human complement and human serum. We have established a population of human packaging cells, 293AM, by transfection of a chimeric retroviral helper virus, pAM3-IRES-Zeo (Chapter 4), into human embryonic kidney 293 cell line. Vectors packaged by 293AM cells were resistant to human serum inactivation (Fig. 1) and should increase the efficiency of gene transfer in human subjects.

REFERENCES

1. Larsen, R. D., C. A. Rivera-Marrero, L. K. Ernst, R. O. Cummings, and J. B. Lowe.

1990. Frameshift and nonsense mutations in a human genomic sequence homologous 152

to a murine UDP-Gal:beta-D-Gal(1,4)-D-GlcNAc alpha(1,3)- galactosyltransferase

cDNA. J Biol Chem. 265:7055-61.

2. Takeuchi, Y., F. L. Cosset, P. J. Lachmann, H. Okada, R. A. Weiss, and M. K.

Collins. 1994. Type 0 retrovirus inactivation by human complement is determined by

both the viral genome and the producer cell. J Virol. 68:8001-7.

3. Takeuchi, Y., C. D. Porter, K. M. Strahan, A. F. Preece, K. Gustafsson, F. L. Cosset,

R. A. Weiss, and M. K. Collins. 1996. Sensitization of cells and retroviruses to human

serum by (alpha 1-3) galactosyltransferase. Nature. 379:85-8. 153

(A)

e 150 S 100 • Serum m : • Inactivated Serum I 50 • Medium Q£e "n A375 IGROV Target Cells

(B)

•Serum • Inactivated Serum • Medium

A375 IGROV Target Cells

FIG. 1. Variable sensitivity to human serum of vectors produced from different packaging cells. Vectors produced from murine AMIZ cells (A), and human 293AM cells (B) were examined. Surviving titers of G418-resistant colonies after incubation with fresh human serum and heat-inactivated human serum or culture medium at 37 °C for 40 min were measured. Relative titers for normal and heat-inactivated human serum treatment versus medium treatment are shown. Titers for medium treatments on human A375 cells were 2.5

X 10® cfu/ml (293AM cells) and 2 X10® cfu/ml (AMIZ cells); on human IGROV cells were 5

XIO'' (293AM cells) and 3.5 X 10® (AMIZ cells). 154

ACKNOWLEDGEMENTS

I would like to thank Gary for his guidance and patience through all years of my graduate school at Iowa State University. I sincerely appreciate Chuck s support and advise on my research work at Human Gene Therapy Research Institute at Des Moines. Without their encouragement and support, none of this would have been possible. I would also like to thank my POS committee members, Dr. Norman Paradise, Dr. Chris Tuggle and Dr.

Daniel Voytas, for helping with my education and advise. I also thank Dr. Lloyd L. Anderson at Animal Science for my first experience of education in the United States and. Dr. E.

Jeffery Beecham and Dr. Kenneth Culver for the contribution while they were serving as my academic advisors at Human Gene Therapy Research Institute (HGTRI).

I would most like to thank my family for their love and endurance, especially my wife,

Chia-Yueh. She cares for our whole family and enables me to pursue this degree.