Studies on the construction and suppression of nonsense mutants in mammalian cells


Material Information

Studies on the construction and suppression of nonsense mutants in mammalian cells
Physical Description:
Wright, Richard M., 1957-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 12000073
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Figures
        Page vi
        Page vii
    List of abbreviations
        Page viii
        Page ix
        Page x
        Page xi
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
    Chapter 2. Materials and methods
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
    Chapter 3. Simian virus 40 amber suppression
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
    Chapter 4. Use of AAV as a vector for the xenopus try tRNAsup gene
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
    Chapter 5. Creation of amber mutants in AAV and adenovirus
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
    Chapter 6. Isolation of suppressor cell lines using AAV
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
    Chapter 7. AAV as a vector
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
    Chapter 8. Significance
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
    Biographical sketch
        Page 119
        Page 120
        Page 121
        Page 122
Full Text







To my family, for their interest and support.


I would like to thank Dr. Nicholas Muzyczka for giving me the

opportunity to work in his laboratory. Nick's patience made the

laboratory a very relaxed place in which to work. In addition, I

particularly admire his apolitical approach to science.

I would also like to thank Drs. Ed Siden, William Holloman,

and Bert Flanegan for their support and suggestions concerning

this work.

Phil, Jon, Sue, Anna and Jennifer made the laboratory a very

helpful, stimulating, and entertaining place in which to work.

The friendship of Brian Masters, one of the most helpful and

interesting individuals I have ever known, will be sadly missed.

I would also like to thank Sandy Ostrofsky for typing this


Finally, I would like to thank my wife, Rhonda, for her

unending support and love. She was extremely helpful in the

laboratory and patient and understanding during some very

stressful times.


ACKNOWLEDGEMENTS ......................................... iii

LIST OF FIGURES .......................................... vi

KEY TO ABBREVIATIONS ..................................... viii

ABSTRACT ................................................. x


ONE INTRODUCTION ................................... 1

TWO MATERIALS AND METHODS ........................... 8
Cells and Tissue Culture Conditions ........... 8
Virus Stocks and Infection .................... 8
Enzymes and Reaction Conditions ............... 10
Preparation of Plasmid DNAs ................... 10
Preparation of AAV DNA ........................ 11
Transformation of HB101 ....................... 12
Rapid Plasmid Isolation ....................... 12
Colony Hybridization .......................... 13
Nick Translation .............................. 14
Agarose Gel Electrophoresis ................... 14
Polyacrylamide Gel Electrophoresis ............ 15
Southern Blotting ............................. 15
Extraction of DNA from Agarose and
Polyacrylamide .............................. 15
Nitrocellulose Filter Hybridization ........... 16
Isolation of tRNA ............................. 17
Detection of AAV and SV40 Replication In Vivo 18
Diethylaminoethyl (DEAE)-Dextran and Calcium
Phosphate Transfection of Mammalian Cells ... 19
E. coli Spheroplast Fusions ................... 20
Immunofluorescence and Titering of AAV and SV40 21
Assay for AAV or SV40 Packaging ............... 22
Making the Amber Linker ....................... 22
35S Labeling and Extraction of Viral Proteins 23
Adenovirus Plaquing ........................... 24
Adenovirus Mutagenesis ........................ 25
Neomycin and XGPRT Selection .................. 26

Introduction .................................. 28
Results ....................................... 30
Discussion .................................... 36


tRNASUP GENE ................................... 39
Introduction ................................. 39
Results ...................................... 40
Discussion ................................... 50

Introduction ................................. 55
Results ...................................... 57
Discussion .................................... 65

Introduction ................................. 77
Results ...................................... 78
Discussion ................................... 79

SEVEN AAV AS A VECTOR ................................. 84
Introduction ................................. 84
Results ...................................... 88
Discussion .................................... 100

EIGHT SIGNIFICANCE ................................... 107

REFERENCES ............................................... 109

BIOGRAPHICAL SKETCH ...................................... 119


Figure 1. Physical Maps of Plasmids, pDR404,
pSV-tT-2(SU+), and pWM107 ...................... 32

Figure 2. In Vivo Suppression of DR404 by WM107 .......... 35

Figure 3. Physical Map of pSM620 ......................... 42

Figure 4. Physical Maps of Plasmids pWCM103, pWCM109,
PWCM106, and pWCMIO8 ........................... 44

Figure 5. Replication Assay for AAV/sup Plasmids pWCM103
and pWCM109 .................................... 47

Figure 6. Packaging Assays for AAV/sup Plasmids pWCM109
and pWCM103 ................................... 49

Figure 7. Comparison of tRNA Species Produced in AAV103
and AAV Infected Cells ......................... 52

Figure 8. Physical Maps of pWSM620-21, pWCM103-1, and
pWCM106-130 .................................... 60

Figure 9. Sequence Analysis of Amber Linker Inserts in
pWCM103-1 and pWCM106-130 ...................... 62

Figure 10. Inhibitory Effect of AAV103 on Adenovirus
ND1-SV40 Fusion Protein Production ............. 67

Figure 11. Analysis of the Four Possible Amber Linker
Orientations .................................. 69

Figure 12. Physical Maps of Plasmids d152-91/Neo,
d1_52-91/NeoX, and d152-91/NeoB ................. 71

Figure 13. Rescue of AAV/tRNASUP Virus from d152-91/NeoX
and dl52-91/NeoB Transduced D-6 ceTls .......... 81

Figure 14. Physical Maps of Plasmids pSV2gpt, d13-22,
pWMgptl, and pWMgpt2 .......................... 91

Figure 15. Estimation of d1l52-91/i..:.-pJr1.r. pi Virus Stock
Titer .......................................... 94

Figure 16. Sall Restriction Map of the ColIb Plasmid ...... 97

Figure 17. Sal Restriction Analysis of Plasmids pWMCo13
and pWMCol2 .................................... 99

Figure 18. Replication Assay for pWMCol1, pWMCol2, and
pWMCol3 ....................................... 102


AAV adeno-associated virus

Ad adenovirus

bp base pairs

BPV bovine papillomna virus

d daltons

D-6 Detroit 6 cell line

dATP deoxyadenosine 5' triphosphate

dCTP deoxycytidine 5' triphosphate

DEAE diethylaminoethyl

dGMP deoxyguanosine 5' monophosphate

dGTP deoxyguanosine 5' triphosphate

DHFR dihydrofolate reductase

dl deletion

DNA deoxyribonucleic acid

DNase deoxyribonuclease

DTT dithiothreitol

dTTP deoxythymidine 5' triphosphate

E. coli Escherichia coli

EDTA ethylenediaminetetraacetic acid

FCS fetal calf serum

fm form

GLN glutamine

HEPES N-2-hydroxyethylpiperazine'N'-2-
ethanesulfonic acid

HPRT hypoxanthine phosphoribosyltransferase

IgG immunoglobulin G

ins insertion

Kb kilobases

L Luria

MEM minimum essential medium

M.O.I. multiplicity of infection

mRNA messenger ribonucleic acid

M.U. map unit

Neo Neomycin

PBS phosphate buffered saline

poly(A) polyadenylic acid

RNA ribonucleic acid

RNase ribonuclease

SDS sodium dodecyl sulfate

sup suppressor

SV40 simian virus 40

tRNA transfer RNA

ts temperature sensitive

TYR tyrosine

VP viral protein

XGPRT xanthine guanine phosphoribosyl

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



Richard M. Wright

December 1984

Chairman: Nicholas Muzyczka, Ph.D.
Major Department: Immunology and Medical Microbiology

Our interest was to construct and study a mammalian cell

system capable of suppressing amber nonsense mutations in vivo.

To do this we constructed a replication deficient Xenopus TYR

tRNAsup gene/SV40 hybrid. This molecule was then

cotransfected into monkey cells with another replication negative

SV40 molecule, DR404, which contained an amber mutation in the Tag

gene. Cells cotransfected with both the SV40/sup gene and the

SV40/amber gene showed suppression of the amber mutation while

cells receiving only one of the SV40 genomes showed no


To create more efficient mammalian suppressor cell lines, we

cloned the 269 base pair Xenopus TYR tRNASUp gene into adeno-

associated virus (AAV). Due to AAV's ability to integrate into

the host genome upon infection, we were able to make cell lines

stably transformed with the suppressor tRNA gene. To check the

suppression efficiency of AAV/sup infected cells, we introduced

amber mutations into those AAV genomes that carried the suppressor

tRNA gene. To make the amber mutants, we inserted a novel 9 base

pair linker containing an amber stop codon into the AAV genome.

Those AAV molecules that contained both the suppressor tRNA gene

and the amber linker were assayed for suppression; however, no

suppression was observed. The failure of this system to show

suppression could be due to the inactivation of the amber "target"

protein by the addition of extra amino acids from the amber


We were also interested in determining if AAV was capable of

being used as a mammalian cloning vector for DNA fragments larger

than the suppressor tRNA gene. Therefore, we replaced AAV's

replication genes with the gene coding for the selectable marker

XGPRT. We showed the molecule could be packaged into AAV virions

and this virus used to confer XGPRT expression to mammalian cells.

In addition, we showed that the AAV genome containing a 20 Kb

insert was still capable of replication after transfection into

mammalian cells.


Some of the basic questions about the molecular biology of

the mammalian cell include the assignment of a particular protein

to a given locus, defining the function of that protein, and

understanding the regulation of that protein's synthesis. These

questions were approached as much as 15-20 years ago in prokary-

otes when rapidly developing techniques allowed geneticists to

deduce relatively precise genomic maps that related protein

function to genetic loci in the chromosomes of Escherichia coli

(E. coli) and bacteriophages. Among the most important class of

techniques available to bacterial geneticists was the ability to

isolate nonsense mutants, a potentially useful type of conditional

lethal mutant.

Several years ago these questions were only approachable in

prokaryotic systems but with the introduction of molecular clon-

ing, the ability to site specifically mutagenize genes in vitro,

and the increased ability to transform mammalian cells, these same

questions are being asked in mammalian systems as well. Although

the recent cloning and mapping techniques have made it possible to

begin asking these questions in eukaryotes, some of the nost

powerful techniques available to prokaryotic geneticists are still

not available to mammalian geneticists. These include the ability

to isolate and propagate nonsense mutants.

Bacterial geneticists have used three types of mutants for


1) Deletion mutants have a portion of the gene missing such

that either the reading frame is now incorrect and the protein

truncated and inactive or a crucial portion of the protein is

missing and the folding of the protein is altered and the protein

is nonfunctional.

2) Insertion mutants, like deletion mutants, can change the

reading frame or alter the protein's conformation and thus

inactivate it.

3) Point mutants consist of missense and nonsense mutants.

Missense mutants consist of a single base change in a codon, thus,

a single amino acid change in the protein. If the original amino

acid is crucial to the structure or function of the protein, the

protein is inactivated. The nonsense mutants have a single base

change which changes an amino acid codon to a stop codon. This

premature stop codon truncates the protein and may inactivate it.

These three types of mutants can be created both in vivo and in

vitro in a variety of ways and as mentioned above have been used

to map the gene position and define the functions of proteins in

prokaryotes. All but the nonsense mutants have also been used to

study animal viruses and more recently eukaryotic cellular genes.

Nonsense mutants are a class of point mutants in which a DNA

codon triplet coding for an amino acid is mutagenized to one of

the three stop codons: amber UAG, opal UGA, and ochre UAA. This

type of mutant was first found in the E. coli system using

bacteriophage T4, f-2, and the alkaline phosphatase gene of E.

coli (Yanofsky and St. Lawrence, 1960; Epstein et al., 1963;

Benzer and Champs, 1962; Garen and Siddiqi, 1962; Sarabhai et al.,

1963; Notani et al., 1965; Weigert and Garen, 1965; Stretten and

Brenner, 1964; Smith et al., 1966). Suppression levels of up to

50% were seen in some of these naturally occurring E. coli stocks.

It was later found that lysogenic bacteriophages carrying

suppressor tRNA genes could produce an in vivo E. coli system also

capable of producing high levels of suppression (Signer et al.,

1965; Garen et al.., 1965; Brenner et al., 1965; Henning ettaal.,

1965). The current list of different prokaryotic nonsense

suppressor tRNA genes includes over 60 different E. coli cell

lines and bacteriophages in which the suppressor tRNA has been

sequenced and the base or bases that differ from the normal

isoaccepting species determined (Celis, 1980). It is clear from

these data that the strongest suppressors have base changes in the

anticodon although a single base change as much as 30 bases from

the anticodon can change a normal tRNA to a tRNAsup. It is

also clear that mutations in the tRNA gene not involved with the

suppressor activity itself can cause aminoacylation of different

amino acids on the same tRNA molecule. In addition, temperature

sensitive E. coli suppressor tRNA genes have been isolated that

are functional only at a permissive temperature (Smith et al.,

1971; Smith et al., 1970; Ozeki et al., 1969; Gallucci et al.,

1970; Oeschger and Berlyn, 1973).

This wide variety of isoaccepting E. coli tRNASUP genes

with differing suppression levels has contributed a great deal to

the understanding of prokaryotic genetics. For instance, an in

vivo system capable of suppressing nonsense mutations allowed

bacterial geneticists to make pure stocks of bacteriophages

containing lethal nonsense mutations (Steege and Soil, 1978).

These nonsense bacteriophage stocks were crucial in that they

allowed complementation with other nonsense mutants and in this

way the mutant lesion could be mapped to a particular locus. The

stocks were also particularly useful in identifying and studying

the functions of the mutant protein. By producing quantities of

the truncated protein without contaminating wild type protein, the

truncated protein could be studied in vitro and in vivo (Steege

and Sol, 1978).

More recently suppressor tRNA genes have been isolated in

lower eukaryotes such as yeast (Sherman et al., 1973; Capecchi et

al., 1975; Brandiss et al., 1975; Liebman et al., 1976) and

Caenorhabitis elegans (Waterston and Brenner, 1978; Waterson,

1981). These eukaryotic suppressors raised the possibility that

tRNASup genes might be found for mammalian cells. This could

be extremely valuable in studying mammalian genetics since most

tempterature-sensitive mutants found in mammalian cells are

"leaky" due to the mammalian cells inability to grow at

sufficiently high temperatures. This lack of temperature-

sensitive mutants increases the need to develop a "stringent"

mammalian nonsense suppressor cell system.

The search for a mammalian nonsense suppressor system began

several years ago when several mammalian genes capable of acting

as selectable markers were isolated with amber mutations. Some of

these include the herpes thymidine kinase gene (Cremer etal .,

1979), large T antigen of SV40 (Rawlins and Muzyczka, 1980), human

B-globin gene (Chang and Kan, 1979) and the fusion protein of

SV40 adenovirus (ND1) hybrids (Gesteland et al., 1977). In

addition, several prokarytoic genes under mammalian promoter

control have been isolated which contain nonsense mutants (for

example, XGPRT and the aminoglycoside 3' phosphotransferase

resistance genes) (Hudziak et al., 1982). Some of these were

shown to be suppressed in yeast in vitro suppressor systems

(Gesteland et al., 1977; Chang etal., 1979) or in a Xenopus

oocyte system containing yeast suppressor tRNAs (Bienz et al.,

1980). However, all attempts to isolate mammalian suppressor cell

lines using these nonsense mutant selectable markers have failed.

More recent attempts to isolate mammalian and suppressor cell

lines have focused on introducing cloned tRNASUp genes into

mammalian cells. Initial attempts involved the microinjection of

yeast or E. coli aminoacylated suppressor tRNAs into mammalian

cells but as expected only a transient suppression was observed

(Temple et al., 1982; Capecchi et al., 1977; Celis et al., 1979).

Other attempts involved the use of cloned yeast or E. coli

tRNASUp genes using a mammalian cloning vector but these tRNA

genes failed because they did not splice or function in the

mammalian cells (Goff and Berg, 1979; Hamer et al., 1977).

The most recent and successful attempts at making a mammalian

suppressor system have involved the use of cloned higher eukary-

otic or mammalian tRNA genes. These mammalian tRNASUP genes
were created by M13 site specific mutagenesis (Kudo et al., 1981;

Hutchison et al., 1978; Razin et al., 1978) of the anticodon of

previously cloned cellular tRNA genes. To date, the three higher

eukaryotic tRNASuP genes are the Xenopus tyrosine tRNA gene

(Laski et al., 1982a) and the human glutamine and lysine tRNA

genes (Temple et al., 1982). These were chosen for mutagenesis

for three reasons: 1) they were among the few higher eukaryotic

cloned cellular tRNA genes at the time, 2) they were known to be

transcribed and spliced correctly in mammalian cells (Laski et

al., 1982b), and 3) these three tRNA genes could theoretically

become structural amber suppressor tRNAs by a single base change.

Undoubtedly, as more mammalian tRNA genes are cloned, a much wider

variety of tRNAs will be mutagenized and more tRNASUp genes

will become available.

The suppressor activity of cells containing these tRNASuP

genes, when assayed either in vitro or in vivo, varies dramati-

cally with the number of copies of the tRNA gene within the cells.

Cells infected with an Simian Virus 40 (SV40) vector containing

the tRNA genes show little or no suppressor activity before 24

hours postinfection. However, if these same cells are assayed

after 60 hours, suppression levels up to 30% are observed (Young

et al., 1983). This is probably due to the fact that SV40 begins

to replicate about 24 hours after infection and at 60 hours

postinfection the cell can contain as many as 105 copies of the

SV40-sup+ genome. The only drawback to using SV40 to create
these cell lines is that the SV40 infection lyses the cell at

about 72 hours postinfection. Therefore, to make permanent

suppressor cell lines, microinjection of the tRNAsuP genes has

been tried. So far, this method has produced permanent cell lines

with suppressor activities of only about 4% (Young et al., 1983).

It is very possible that suppression levels of greater than 4% are

lethal for mammalian cells over many cell divisions.

To date, no mammalian nonsense mutant has been suppressed to

a high enough level to be genetically useful, as, for example, for

the preparation of mutant viral stocks. Nevertheless, these

experiments have raised the hope that suppressor cell systems like

those in prokaryotes could be developed for mammalian cells. Such

suppressor systems would give geneticists a powerful tool for the

study of mammalian genes.


Cells and Tissue Culture Conditions

The CV-C and BSC-40 monkey cell lines were used to propagate

both wild type SV40 virus and SV40 virus carrying the Xenopus TYR

tRNASUp gene. The BSC-40 line was created from the BSC-2 cell

line by conditioning for growth at 40'C (Brockman and Nathans,

1974). The CV-C cell line was derived from TC7 cells (Robb and

Martin, 1972). The COS-1 cell line was derived by transformation

of the CV-1 cell line with a nonreplicating SV40 mutant (Gluzman,

1981). The hybridoma clone 412 cell line, capable of secreting

atibody to the carboxy terminus of Tag (Gurney et a1l., 1980), was

obtained from the cell Distribution Center of the Salk Institute.

The CV-C, BSC-40, HeLa and Detroit 6 cell lines were grown as

monolayers in Eagle's minimum essential media (MEM) supplemented

with 1% glutamine and 10% fetal calf serum (FCS). The COS-1, 293

and KB cell lines were grown in suspension in Debecco's modified

Eagle's media (DME) suplemented with 1% glutamine and 10% FCS.

HeLa cells were sometimes propagated in suspension in Joklik

modified MiEM suplemnented with 2% glutamine and 10% calf serum.

Virus Stocks and Infection

Wild type SV40, SV-tT-2(SU+) and am404 were propagated in

either CV-C or COS-1 cell lines respectively in MEM with 2% FCS.

Plaque purified Ad2 was a gift from Dr. Peter McGuire. Plaque

purified Ad5 was a gift from Dr. Thomas Shenk. Ad2 and Ad5 stocks

were propagated by infecting HeLa cells in suspension at an M.O.I.

of 5, centrifuging the infected cells down 48 hours postinfection,

resuspending the cells in 30 ml MEM, and freezing and thawing

three times. Virus was separated from cell debri by centrifuga-

tion at 6000 times gravity for 10 minutes at 0C.

All wild type AAV and mutant AAV stocks were propagated in

one of two ways. The first procedure simply involved freezing and

thawing Ad plus AAV infected cells three times 48 hours after

infection and centrifuging the cell debris from the virus. Other

AAV virus stocks were prepared essentially as described by Berns

and Rose (Berns and Rose, 1970). Three 10 cm dishes of HeLa cells

were transfected with AAV and infected with Ad5 or Ad2 and the

plates frozen and thawed three times 48 hours later. This lysate

along with 5 x 109 additional Ad2 or Ad5 was added to 5 x 108

HeLa cells growing in suspension culture. These cells were

harvested by centrifugation and washed in phosphate buffered

saline (PBS). The cell pellet was frozen and thawed and

resuspended in 30 ml PBS, and deoxycholate and trypsin were added

to a final concentration of 2% and 0.02% respectively. This

mixture was incubated at 37C for 30 minutes and then homogenized

with 15-20 strokes of a Dounce homogenizer. Cesium chloride was

added to this homogenate to obtain a density of 1.4 g/ml and the

AAV banded by centrifuging at 40,000 rpm in a SW50 rotor for 20

hours at 4C. The AAV virus band was visualized and collected

using a high intensity white light and the gradient repeated. The

virus bands collected were then dialyzed twice against one liter

of 0.3 times MEM.

Enzymes and Reaction Conditions

All restriction endonucleases were purchased from Bethesda

Research Laboratories (BRL), Biolab Laboratories, or International

Biotechnology Incorporated (IBI), and were used under conditions

suggested by the supplier. The DNA ligase, E. coli polymerase I

large fragment, DNA polymerase, and T4 kinase were purchased from

IBI or BRL and used under conditions suggested by supplier. The

calf alkaline phosphatase was purchased from Collaborative

Research. Pancreatic ribonuclease A (RNase A) and deoxyribo-

nuclease I (DNase I) were supplied by Worthington Biochemicals.
The RNase A was resuspended at a concentration of 1 mg/ml and

heated to 80C prior to use. RNase A was used at a concentration

of 10 ig/ml. Lysozyme and Pronase were purchased from Calbiochem-

Behring Corporation. Pronase resuspended in 0.1X SSC was allowed

to self digest for 1 hour at 37c prior to use.

Preparation of Plasmid DNAs

To obtain preparative amounts of plasmid DNAs, E. coli HB101

containing the desired plasmid was grown at 37C to a density of

50-70 Klett units (green filter, Klett-Sumerson colorimeter). The

culture was then induced for 12-20 hours by the addition of

chloramphenicol to a concentration of 170 ig/ml. The cells were

harvested by centrifugation at 3000 times gravity. The pellet was

resuspended on ice in 50 mM Tris-HCl (pH 8.0), 10 mM

ethylendinitrotrilo-tetraacetic acid (EDTA), and 25% sucrose.

Lysozyme was added to a final concentration of 1% and the solution

incubated on ice for 30 minutes. Sodium dodecyl sulfate (SOS) was

then added to a final concentration of 1% and after 30 minutes

sodium chloride added to a final concentration of one mole/liter.

After 12 hours at 4C, this gelatinous mixture is centrifuged at

15,000 times gravity for 60 minutes and the supernate collected.

Polyethylene glycol 6000 (Kodak) was then added to a final

concentration of 10% and this mixture was allowed to sit at 4C

for 3-8 hours. This mixture was then centrifuged at 10,000 times

gravity for 30 minutes. The resulting pellet was then dissolved

in 10 mM Tris-HCl, pH 7.5, and 10 mM EDTA and extracted once with

chloroform and then phenol. The supernatant was then precipitated

by the addition of one-tenth volume of 3.0 M sodium acetate,

pH 7.0, and 2.5 volumes of 95% ethanol. After 60 minutes at -20C

this solution was centrifuged at 10,000 times gravity and the

pellet dissolved in Tris-HCl, pH 7.5. RNase A was then added to a

final concentration of 20 pg/ml and the solution incubated at 37C

for 30 minutes. This solution was now extracted with phenol and

reprecipitated with 95% ethanol. The pellet was dissolved in

water and the form I plasmid DNA banded by equilibrium centrifuga-

tion in a cesium chloride gradient (density 1.56 g/ml)

containing 0.1 mg/ml of ethidium bromide.

Preparation of AAV DNA
AAV was prepared as described by Berns and Rose (Berns and

Rose, 1970). AAV virus was isolated as described above except

dialysis was against 10 nmM Tris-HCl, pH 7.5. The virus was then

denatured in 0.1 M NaOH, neutralized with an equal volume of 1 M

Tris, pH 7.5. This mixture was then extracted with phenol and

precipitated with ethanol. This pellet was resuspended in 1 ml of

0.1 M Tris-HCl, pH 7.5 and 50% deionized formamide and allowed to

stand at 25C for 12 hours. This solution was now precipitated

with ethanol.

Transformation of HB101

E. coli HB101 was grown to density of 110-130 Klett units

(green filter, Klett-Sumnmerson colorimeter) in Luria broth. Two

milliliter aliquots were centrifuged at 3000X gravity and the

pellets resuspended in 100 mM CaCl2 and allowed to sit for 12-24

hours at 4C. Between 0.1 ug and 1.0 pg DNA was added to the cell

mixture and allowed to sit at 4C for 30 minutes. This sample was

then gently mixed in a water at 42C for 3 minutes. Ten milli-

liters of Luria broth was then added and this mixture held at 37C

for one hour. The cells were now pelleted at 3000X gravity and

resuspended in 200 u1 of Luria broth. These cells were then

plated onto Luria agar plates containing either 100 ig/ml

ampicillin or 20 pg/ml tetracycline. Positive clones were identi-

fied by either rapid plasmid isolation or colony hybridization

(see below).

Rapid Plasmid Isolation

Potential plasmid containing E. coli HB101 clones were grown

to stationary phase in 2 ml of Luria broth containing the

appropriate antibiotic. The samples were now pelleted by centri-

fuging at 2000X gravity. These pellets were resuspended in 200 p1

STET buffer (25% sucrose, 50 mM Tris, pH 8.0, 10 nmM EDTA) and

lysozyme added to a final concentration of 1 mg/ml. After sitting

on ice for 10 minutes, these samples were placed in a vigorously

boiling water bath for 40 seconds and then spun at 13,000X gravity

for 10 minutes. The pellets were then removed with a toothpick

and an equal volume of isopropanol added to the supernatant and

placed at -20C. After 5 minutes, the samples were then centri-

fuged at 13,000X gravity for 10 minutes. The samples were then

resuspended in water, extracted with phenol and precipitated with

ethanol. The yield of plasmid DNA from the 2 ml sample is about

1.5 pg.

Colony Hybridization

Colony hybridization was done as described by Grunstein and

Hogness (Grunstein and Hogness, 1975). E. coli HB101 transfor-

mants were grown for 20-24 hours and then transferred to a

nitrocellulose filter by replica-plating. The filters were then

placed on Whatman 3 mm paper saturated with 1 M NaOH. After 7

minutes the filters were blotted on dry 3 mm paper and neutralized

2 times by placing them on 3 mm paper saturated with 1 M Tris-HCl,

pH 7.5. The filters were again blotted dry and moved to 3 mm

paper saturated with 1 M Tris-HCl, pH 7.5 and 1.0 M NaCl. The

filters were then dried and immersed in a solution of pronase for

in a solution of pronase for 2 minutes at a concentration of

2 mg/ml in 0.1X SSC at room temperature. The filters were then

washed in 0.1X SSC, dried, and submerged in chloroform for 2

minutes and dried. These filters were then washed in O.1X SSC and

baked in a vacuum oven at 80 for 2 hours.

Nick Translation

One microgram of the substrate DNA was mixed with a 100 pIl

reaction mixture containing 200 pM GTP, ATP and TTP, 200 ICi
a-32p CTP (specific activity of 200-500 Curies/mMole) and

10 ul of IOX nick translation buffer (200 nmM Tris-HCl, pH 7.5,

100 mM MgS04, 100 pg/ml BSA). To this solution 2 ji (4 units)

of E. coli polymerase I and 1 ng DNase I were added. This mixture

was incubated at 15C for 1-4 hours and incorporation of 32p

CTP checked by monitoring the acid precipitable counts. The

unincorporated label was removed by chromatographing the reaction

mixture on a Sephadex (Pharmacia Fine Chemicals) G-50 column. The

specific activity of the nick translated DNA was between 5-10 x
107 cpm/ug.

Agarose Gel Electrophoresis

Separation of DNA fragments was accomplished by electrophore-

sis through agarose gels varying in concentration from 0.8% to

2.0% agarose. The agarose was purchased from BRL and the gels run

in either IX TEA buffer (0.04 M Tris-HCI, pH 7.8, 0.02 M sodium

acetate, 2 t'M EDTA or O.5X TEA buffer). The gels were stained for

20 minutes in 2 ug/ml ethidium bromide and the bands visualized

using a 450 nM wave length ultraviolet light after


Polyacrylamide Gel Electrophoresis

Polyacrylamide gel electrophoresis was used for preparative,

analytical, and sequencing purposes. For preparative and

analytical purposes acrylamide gels contained 4-8% acrylamide

(ratio of acrylamide to bis-acrylamide was 20:1) with the lower

percentage gels used for separating larger fragments of 500 to

1000 base pairs. Electrophoresis times varied with the DNA

fragment size and the resolution required. Sequencing gels ranged

in concentration from 5% to 20% (ratio of acrylamide to

bis-acrylamide 50:1) and contained 8 M urea. They were used in

the sequencing of DNA fragments and in the resolution of tRNA


Southern Blotting

This procedure was performed essentially as described by

Southern (Southern, 1975). The DNA fragments were resolved on an

agarose gel and the fragments denatured by soaking the gel in 1.0

M potassium hydroxide for 40 minutes. The gel was then neutra-

lized to pH 6.5 with blotting buffer (1 M Tris-HCl + 1 M HCI) for

40 minutes and the DNA transferred up into nitrocellulose using

absorbant paper towels.

Extraction of DNA from Agarose and Polyacrylamide

Extraction of DNA fragments from agarose gels was done as

follows. The DNA fragment to be isolated was visualized by

staining the gel with ethidium bromide and a slice of the gel

containing the desired fragment excised. This fragment was then

placed in dialysis tubing (Fisher, Spectrapore) containing 2 ml of

0.5X TEA buffer. The gel fragment in the bag was then placed

perpendicular to the current in a horizontal gel apparatus and the

DNA electroeluted for 4 hours at 100 volts. After 4 hours the

current was reversed for 1 minute and the contents of the bag

emptied into a siliconized polypropylene tube. The gel slice and

bag were both thoroughly rinsed with water into this same tube.

This solution was then concentrated using butanol, extracted with

phenol 2X and precipitated with ethanol.

Extraction of DNA from polyacrylamide also involved cutting

out the desired gel slice. The gel slice was weighed and placed

in a siliconized tube and pulverized with a siliconized glass rod.

Now the gel was added to a solution of Magic X (0.3 M sodium

acetate, pH 4.8, 0.5% SDS, 10 mM EDTA) at a ratio of 1:5 wt/wt.

This was allowed to shake at 37C for 12 hours, centrifuged at

10,O00X gravity and the supernatant extracted with phenol and

precipitated with ethanol. This pellet was subsequently

resuspended and the DNA separated from the residual acrylamide

using an Elutip purchased from BRL.

Nitrocellulose Filter Hybridization

The same procedure was followed here whether hybridizing

Southern blots of filters from the colony hybridization. The

nitrocellulose filters were dipped in a solution of 6X SSC and

placed in hybridization bottles and 15 mls of hybridization

solution added (6X SSC, 50 ntmM Tris-HCl, and 4X Denhardt's solution

(0.02% Ficoll, 0.02% polyvinylpyrolidine, and 0.02% bovine serum

albumin)). Also added was 100 pg/ml of denatured Salmon sperm

DNA (Sigma). The filters were prehybridized at 65C in this

solution and after 6-24 hours it was poured out and 15 mis of new

hybridization solution added. Now 0.2 ig of the denatured nick

translated DNA and 1 mg of denatured salmon sperm DNA were added

and the filters hybridized from 12-24 hours at 65C. The filters

were then washed successively in 500 mis of 2X SSC, 500 ml of 0.2X

SSC with 50 mM KP04, and 0.2X SSC at 65C. These filters were

then dried and autoradiographed at -70%C with a Dupont HiPlus

intensifying screen.

Isolation of tRNA

The procedures used to isolate tRNA from mammalian cells was

described by Littauer and Stern (Littauer and Stern, 1971). Six

liters of HeLa cells were grown in suspension and pelleted by

centrifugation at 3000X gravity. The cell pellet (wet weight 2.5

grams) was resuspended in 5 ml cold water and an equal volume of

phenol added. The solution was shaken every 5 minutes for 1 hour

on ice, then allowed to come up to 20C and centrifuged at 20,O00X

gravity for 15 minutes. The upper layer was extracted, 0.2

volumes of NaCl and 2.5 volumes of ethanol added and precipitated

at -70C for 2 hours. We centrifuged this solution and the

resulting pellet was then resuspended in the cold in 3 mls of cold

1 M NaCl. This was then centrifuged at 15,000X gravity for 15

minutes at 4C and the supernatant was precipitated with ethanol.

This precipitate was then resuspended in enough 0.3 M NaAc (pH

7.0) to give the solution an optical density at 260 nM of between

3.0-3.5 mg/ml nucleicc acid concentration at an O.D. of 1.0 is

50 ug/ml). The solution was then warmed to 20C and 0.54 volumes

of isopropanol added slowly over 15 mintues. This solution was

then centrifuged at IO,O00X gravity for 10 minutes at 20C. The

supernatant was then collected and 0.34 volumes of isopropanal

added at 4C. After 10 minutes at 4C this solution was centri-

fuged at 10,O00X gravity for 10 minutes and the pellet resuspended

in water. This was then extracted with phenol and precipitated

with ethanol. The yield was about 1 mg of pure tRNA. All

solutions and glassware were be treated with diethyl pyrocarbonate

(Eastman) to insure the tRNA isolated was not partially degraded

by RNases.

Detection of AAV and SV40 Replication In Vivo

Eight hours after infection with AAV and Ad2, the media was

removed and replaced with phosphate-free MEM supplemented with 2%

dialyzed FCS and 50 UCi/ml 32p orthophosphate (Amersham). (If

replication is to be assayed by Southern blot, this step can be

omitted.) Thirty hours later the media was removed and the viral

DNA was extracted essentially as described by Hirt (1967). One

milliliter of the HIRT extraction buffer (0.5% SDS, 10 mM EDTA,

25 mM Tris-HCl, pH 7.5) was added to each plate and the lysed,

gelatinous mixture was scraped from the plates into tubes using a

rubber policeman. The solution was then made 1 M in NaCl and

allowed to sit at 4C for 12 hours. This mixture was then centri-

fuged at 13,OOOX gravity for 1 hour at 40C and the supernatant

extracted with phenol and precipitated with ethanol. The

procedure is the same for SV40 replication except the 32p

orthophosphate is added 20 hours postinfection and allowed to

remain on the cells for 48 hours.

Diethylaminoethyl (DEAE)-Dextran and Calcium Phosphate
Transfection of Mammalian Cells

All DEAE transfections into mammalian cells were done essen-

tially as described by McCutchan and Pagano (McCutchan and Pagano,

1968). Between 10 ng and 10 pg of DNA were added to 1 ml of

transfection solution (1 mg/ml DEAE-dextran in 2X MEM) and 1 ml of

water added. This solution was added to cells that were between

50-90% confluent after the media had been removed. This dextran

solution was allowed to sit on the cells for 20 minutes at room

temperature and was then washed off with one rinse of MEM. Media

containing 100 pM chloroquin (Sigma) was added to the cells and

the cells incubated at 37C. Four hours later the cells were

rinsed with MEM and MEM plus FCS added.

All calcium phosphate transfections were done essentially as

described by McKnight (McKnight et al., 1981). The DNA to be

transfected (1 vg to 20 ug) was made 40 ug/ml in nucleic acid by

the addition of salmon sperm DNA. This DNA solution was now made

200 mM in CaCl2. This CaCl2-DNA solution was then added

dropwise to an equal volume of 2X HEBS buffer (280 mM NaCl, 1.5 mM

sodium phosphate, pH 7.1, and 50 mM N-2-hydroxylethyl piperazine-

N-2-ethane sulfonic acid, pH 7.1). A fine precipitate was allowed

to form and this precipitate was added directly to the media of

the cells. This precipitate was allowed to adhere for 4 hours and

then PBS containing 15% glycerol was added for 1 minute. This

solution was removed, the cells washed with MEM and media


E. coli Spheroplast Fusions

The E. coli spheroplasts were fused to CV-C cells using the

procedures essentially as described by Weiss (1978) and

Rassoulzadegan et al. (1982). E. coli containing the plasmid to

be transferred was grown to a klett of between 0.8 and 1.4 and

induced with chloramphenicol for 12 hours. The bacterial cells

were then pelleted by centrifugation and resuspended in 1:10

volume of a solution of 20% sucrose and 50 nM Tris-HCI, pH 8.0,

and the suspension was warmed to 37C. Fresh lysozyme (1 mg/ml)

and DNase (1 mg/ml) were then added to the suspension to a

concentration of 100 ig/ml and 10 pg/ml respectively. After

incubating at 37C for 12 minutes, 1:10 volume of EDTA was added

very slowly and this suspension incubated at 37C for 10 minutes.

This mixture was then added to an equal volume of MEM and 2 ml of

this suspension added to a 70% confluent 6 cm plate of CV-C cells

prewashed with PBS one time. The E. coli spheroplasts were then

spun down onto the cells by centrifuging the culture dish at 2000X

gravity for 3 minutes at room temperature. The supernatant was

then removed and 1 ml of a 50% solution of polyethylene glycol

1000 (Fisher) in MEM gently added. After one minute the plates

were gently washed 3 times with PBS and fresh media containing

0.1% gentamyacin.

Immrunofluorescence and Titering of AAV and SV40

Immunofluorescence and titering of AAV and SV40 were done

using the same method. Hamster antisera specific to SV40 Tag was

a gift from J.S. Cole, III, Biological Carcinogenesis Branch of

the National Cancer Institute. Guinea pig antisera specific to

the AAV capsid proteins was a gift from James Rose. CV-C cells

were infected or transfected with SV40 and 30 hours later fixed by

replacing the media with -20C 95% ethanol and kept at -20C for

20 minutes. The ethanol was removed and the plate was then washed

with PBS. One milliliter of a 1:50 dilution of the SV40 antisera

was then added and the plate incubated at 37C for 1 hour in a

moisture chamber with occasional rocking. This solution was then

removed and the plate was given three 5-minute washes with PBS.

Now, 1 ml of a 1:50 dilution of fluoroscein labeled goat

anti-hamster IgG antisera (Cappel Laboratories) was added to the

plate and it was incubated for 1 hour with occasional rocking.

This solutions was removed and again the plate washed 3 times with

PBS. The plate was then viewed under a Zeiss fluorescence

microscope. This procedure was basically the same for AAV with

the following changes. The cells infected with AAV were also

infected with adenoviruses at an M.O.I. of 5. The fluoroscein

labeled goat anti-guinea pig IgG antisera was purchased from

Cappel Laboratories. In using this fluorescent antibody technique

for titering an AAV stock, 70% fluorescence was taken as the

maximum achievable percentage (Chamberlin, 1979).

Assay for AAV or SV40 Packaging

Forty-eight hours after adenovirus infection and AAV trans-

fection the cells were frozen and thawed three times in the plate.

This lysate was then centrifuged at 3000X gravtiy for 5 minutes

and the supernatant saved. An aliquot of this supernatant was

heated to 56C for 15 minutes and used to infect another plate of

cells and the cells assayed for AAV replication as described

above. The same procedure was used for SV40 virus except that

adenovirus was not added and the first set of plates were frozen

and thawed 72 hours postinfection.

Making the Amber Linker

Both strands of the amber linker were synthesized on a Systec

Oligonucleotide synthesizer. After synthesis the resin holding

the oligomers was added to 1 ml of dioxane and washed twice. The

resin was then added to a solution of 200 uI thiophenol (Fisher),

400 il triethylamine (Fisher), and 400 pl dioxane (Fisher). This

mixture was allowed to stand for 1 hour at room temperature with

occasional vortexing. The support was then washed 2 times with

1 ml dioxane, then 2 times with 1 ml methanol, and then 1 time

with 1 ml of ether. After drying, the resin was then added to

1 ml of concentrated ammonium hydroxide for 3 hours with occa-

sional shaking. The ammonium hydroxide containing the oligomer

was then taken off and saved and the resin washed with an addi-

tional 0.5 mls ammonium hydroxide which was added to the 1 ml

ammonium hydroxide already collected. This ammonium hydroxide-

oligomer solution was then heated at 55C for 12 hours and the

ammonium hydroxide then removed by evaporation. The oligomers

were then resuspended in 300 pl distilled water and extracted with

butanol 2 times. The oligomers were then precipitated by making

the solution 0.3 M in ammonium acetate and adding 3 volumes of

ethanol. The oligomers were resuspended in distilled water and the

yield determined by checking the optical density at a wavelength

of 260 nM.

To use these oligomers as a linker both oligomers were

kinased using T4 kinase (BRL) under conditions recommended by

the supplier. These kinased oligomers were then combined and 1:50

volume of 5 M NaCl added. These oligomers were then heated to

45C for 15 minutes and allowed to slowly cool to roan temperature

over 3 hours. The linkers were then precipitated with ethanol.

35s Labeling and Extraction of Viral Proteins

Adenovirus infected cells were kept in MEM supplemented with

10% FCS for 30 hours. The media was then taken off and the plate

washed with MEM lacking methionine but supplemented with 10%

dialyzed FCS. Three milliliters of the MEM minus methionine media

was now added to the plate and 60 uCi/ml 35S methionine

(Amersham specific activity 70 Ci/mMole) added. The cells were

now incubated at 37C from 5 to 30 minutes, the plate washed with

cold PBS and 1 ml of protein extraction buffer (10 ntM Tris-HCl,

pH 8.0, 140 mM NaCl, 1 mM CaCl2, 1 mM MgCl2, 10% glycerol,

1.0% Monidet-P40, and 200 pg/nml phenylmethylsulfonylfluoride)

added. After 5 minutes the cells were scraped off the plate and

centrifuged at 13,OOOX gravity for 15 minutes. The supernatant

was then made 500 ug/ml in BSA and this solution centrifuged at

30,OOOX gravity for 60 minutes. The supernatant was then

collected and added to an equal volume of antisera from hybridoma

412 (a gift from E. Gurney) and incubated with occasional shaking

for 12 hours at 4C. One half milliliter of Staphylococcus aureus

10% solution (a gift from W. Wakeland) were centrifuged at 10,O00X

gravity for 10 seconds. This pellet was then resuspended in

0.5 ml protein extraction buffer and 100 ul added to the antisera

solution at 4 for 4 hours with occasional shaking. This mixture

was then centrifuged at 10,00X gravity and resuspended and rinsed

3 times with extraction buffer. The pellet was then resuspended

in 100 u1 of protein sample buffer (62.5 mM Tris-HCl, pH 6.8, 2%

SOS, 20% glycerol, 70 mM 2-mercaptoethanol and 0.001% bromphenol

blue) and boiled for 5 minutes. This mixture was centrifuged at

10,000X gravity for 2 minutes and an aliquot loaded on a discon-

tinuous SDS-plyacrylamide protein gel as described by Laemmli

(1970). After electrophoresis, the proteins were fixed in the gel

by submerging the gel in fixing solution (10% acetic acid, 40%

ethanol, 50% water) for 1 hour. The gel was then rinsed by

submerging it in water for 1 hour and then fluorographed by

soaking the gel in a solution of 1 M sodium salicylate for 1 hour

(Young et al., 1983). The gel was then dried down using a gel

dryer (Hoefer Scientific Instruments) and autoradiographed.

Adenovirus Plaguing

Human 293 cells in a 6 cm dish at 70% confluence were

infected for 2 hours with a stock of adenoviruses and the media

then removed and replaced with 3 ml of plaquing media (DMEM

containing 5% FCS, 1% glutamine, 20 mM MgCl2 and 1% low melting

temperature agarose (BRL)) at 37C. The plates were then allowed

to stand at room temperature for 30 minutes to allow the agarose

to harden. After two days 3 ml more plaquing media was added;

however, this plaquing media contained 2% FCS. After 5 days, 1 ml

of 1% agarose containing 0.003% neutral red (Gibco) was added to

the plates and allowed to harden. The plates were incubated for

24 hours at 37C and the plaques were then counted on a lightbox.

Adenovirus Mutagenesis

In vitro nitrous acid mutagenesis of adenovirus was done

essentially as described by Ensinger and Ginsberg (1972). A

100 pl stock of adenovirus 5 (1 x 109) plaque forming units

(pfu) per ml was added to 0.5 mls of Hanks buffer (0.4 mg/ml

CaCl2, 0.4 mg/ml potassium chloride, 0.06 mg/nl potassium

phosphate monobasic, 8 mg/ml sodium chloride, and 0.09 mg/ml

sodium phosphate dibasic) and mixed with 1.0 ml of 0.2 M acetic

acid, pH 4.2, containing 0.2 M NaNO3. This solution was then

allowed to sit at room temperature and 200 pl aliquots were

removed every 10 minutes. These aliquots were immediately neutra-

lized with 4 volumes of 1 M Tris-HCl, pH 7.8, and diluted with 40

mls of ice cold MEM. These aliquots were now titered by plaquing

as described before.

In vitro hydroxylamine mutagenesis was carried out essen-

tially as described by Williams et al. (1971). One hundred
microliters of an adenovirus 5 stock (1 x 109 pfu) was added to

0.9 ml PBS and diluted 1:1 in PBS containing 2 M hydroxylamine

(Sigma). This solution was allowed to incubate at 37C for 24

hours. At 4 hour intervals over the 24 hour incubation 200 V1

aliquots were removed and dialyzed against MEM. These aliquots

were then titered by plaquing as described before.

In vivo nitrosoguanidine mutagenesis was done essentially as

described by Ensinger and Ginsberg (1972). Plates of HeLa cells

were infected with adenovirus 5 and 6-18 hours later nitrosoguani-

dine (Sigma) was added to the plates to achieve concentrations

between 2 Ug/ml and 50 ig/ml. Forty-eight hours postinfection the

plates were frozen and thawed 3 times and the plates titered for

adenovirus as described before. Those virus stocks that were

reduced 10- to 50-fold by the mutagenesis were used in the AAV

suppression assay.

In vivo ethane methanesulfonate (EMS) mutagenesis was done as

described below. Plates of HeLa cells were infected with

adenovirus 5 and 6-18 hours later EMS (Sigma) was added to the

plates to achieve concentrations between 150 ug/ml and 300 pg/ml.

Forty-eight hours postinfection the plates were frozen and thawed

three times and titered for adenovirus.

Neomycin and XGPRT Selection

Neomycin (G-418) selection was done essentially as described

by Southern and Berg (1982). Approximately 105 Detroit 6 cells

were infected with the virus stocks of d152-91/NeoX, d152-91/NeoB

or a mixed stock of pHM1320Neo/pWMGPT-1. These cells were allowed

to grow in MEM plus 10% FCS for 72 hours and each dish was then

split into ten 10 cm dishes with MEM plus 10% FCS supplemented

with 1 mg/ml G-418 (Gibco). This media was changed every 3 days

and after 7 days cell clones resistant to the G-418 were apparent.

These colonies were split into separate dishes and maintained in

MEM containing 200 pg/ml G-418.

The XGPRT selection was done as described by Mulligan and

Berg (1981). One hundred thousand Detroit 6 cells were infected

with either pHM1320Neo/pWMGPT-1 or ins96/A-M/pWMGPT-1 and grown in

MEM plus 10% FCS for 3 days. Each plate was then split into ten

10 cm dishes and MEM containing 25 pg/ml mycophenolic acid

(Gibco), 10 iy/ml aminopterin (Sigma), 250 ig/ml xanthine (Sigma),
15 ug/ml hypoxanthine (Behring Corp.), 150 pg/ml glutamine

(Gibco), 10 ig/ml thymidine (Calbiochem), and 10% dialyzed FCS.

Twenty four hours later the media was removed and fresh XGPRT

media added. Colonies were visible after 6 days. New XGPRT media
was then added that contained one half the concentration of

mycophenolic acid and aminopterin.


The race to isolate the first functional mammalian

tRNASUP gene ended when the Xenopus TYR tRNA gene was site

specifically mutagenized to a tRNASUP gene (Laski et al.,

1982a). The Xenopus TYR tRNA was known to differ in only 3 bases

from the human TYR tRNA, its gene had been isolated (Kuchino and

Nishimura, unpublished results) and it had been shown to be

expressed and active in mammalian cells (Laski et al., 1982b). In

bacterial systems, it had been clearly shown that the simple

replacement of the guanine with cytosine in the first position of

the anticodon was all that was necessary to allow a TYR tRNA to

now recognize and suppress the amber codon UAG. For these reasons

the Xenopus TYR tRNA gene was cloned into M13 and using a

oligonucleotide primer the guanine was replaced with cytosine in

the third position of the anticodon. The new Xenopus TYR

tRNASUp gene was then cloned into the late region of SV40 and

an SV40 virus stock was made containing the suppressor gene. The

SV40 molecule containing the Xenopus TYR tRNASuP gene was then

inserted into pBR322 and referred to as pSV-tT-2(SU+) (Laski et

al., 1982a).

The Xenopus TYR tRNASuP gene was first assayed in vitro.

An SV40 tRNASUP viral stock was used to infect monkey cells

and 36 hours postinfection the tRNAs from these cells were added

to a in vitro translation mix containing the genomic RNA of

Tobacco Moisaic Virus (TMV). One strain of this plant virus has a

naturally occurring stop codon that truncates a polyprotein in

vivo. This stop codon had been shown to be suppressable in

reticulocyte in vitro translation systems (Pelham, 1978). When

this TMV RNA template was used to test for suppression in the in

vitro mammalian translation system containing exogeneous Xenopus

TYR tRNASup, suppression was observed (Laski et al., 1982a).

To test the Xenopus TYR tRNASuP gene in vivo, an amber

marker gene capable of acting as the template for suppression in

mammalian cells was needed. The SV40-pBR322 plasmid, pDR404,

contains an amber mutation at M.U. 0.271 on the SV40 genome that

was created by specifically mutagenizing a CAG glutamine codon to

a TAG amber stop codon using sodium bisulfite mutagenesis (Rawlins

and Muzyczka, 1980; Shortle et al., 1979). This SV40 amber muta-

tion was located in the region coding for the large T antigen

(Tag) of SV40. T antigen is known to be involved in SV40 replica-

tion (Tegtmeyer and Ozer, 1971; Chou et al., 1974), the transfor-

mation of nonpermissive cells (Graham et al., 1975; Carter et al.,

1979; Pipas et al., 1980; Martin, 1981) and the stimulation of

cellular DNA synthesis in SV40 infected cells (Kit et al., 1967;

Hiscott and Defendi, 1979; Hiscot and Defendi, 1981). As

expected, the amber SV40 genome DR404 contained in pDR404 was

shown to be incapable of autonomous replication. It was also

shown to be immunofluorescence negative when cells transfected

with this SV40 molecule were incubated with antisera specific to

Tag (Rawlins and Muzyczka, 1980). Since these two properties of

SV40 were relatively simple to assay in mammalian cells, pDR404

was used to assay for suppressor activity in monkey cells

transfected with the SV40 TYR tRNASUp gene.


To test the suppression efficiency of pSV-tT-2(SU+) in vivo

we used the SV40 plasmid pDR404 containing an amber mutation at

M.U. 0.271 in the coding region for Tag (Fig. la). (To simplify

the nomenclature all bacterial plasmids are preceded by p as in

pDR404 and pSV-tT-2(SU+). When the cloned genome has been

excised, they are referred to as simple viral mutants DR404 or

SV-tT-2(SU+).) In constructing this mutant the PstI site at

M.U. 0.271 was destroyed by the site specific mutagenesis of the

glutamine codon to the amber codon. Because the replication and

immunofluorescence assays are dependent upon the suppression of

the amber mutation in Tag of DR404 by the SV-tT-2(SU+) molecule,

the SV-tT-2(SU+) molecule itself must have a nonfunctional Tag.

To accomplish this we made a deletion in pSV-tT-2(SU+) with Hpal

which cleaves at M.U. 0.169 and 0.372 on the SV40 genome

(Fig. Ib). The resulting molecule (pWM107; Fig. Ic) contained a

deletion that included the Pst site at 0.271 in Tag and accomp-
lished two goals: 1) it made the SV-tT-2(SU+) defective for

autonomous replication arid immunofluorescence and 2) it made the

reformation of a wild type Tag impossible through recombination of

the complementing SV40 molecules. Thus, neither DR404 nor WM107

could replicate or show irmunofluorescence unless the Xenopus TYR

Figure 1. Physical Maps of Plasmids, pDR404, pSV-tT-2(SU+), and
Broken line represents coding region for Tag. pBR322 and
tRNA gene inserts are represented by open triangles and are
labelled accordingly. Deletions are denoted by open boxed


a) f -

SI pDR 404

\ AMBER Pst PsaI
\ 0.271 0.039 0.0/1.00
HaeII E '- T
ORIGIN 2 189 bp pBR322-
.>---T~-SL RN GNENE ^ --^ /P"2
^ -'^ <\Ns^~-'INSERT
b) / DELETION Cfo 0.72

B0.72 0.82 \ S--HaeI 2

a 0.86-. 1.
\\ PSV-tT-2(SUt) j .86 1.0

\^^ BamHl vHaeII

0.169 pBR322GENE
,-L ^~ ""'*".1 INSERT
PBR322 C) // ^^ Hae--

I Tao 0.86-Y
I 1 PWM107 M.86 1.0
HT p al- /HindM -
0.37 Pstl 0.0/1.0/ EcoRl
\ k\0.271A
-.V"^ BamHl /
Hpal \ S^-0.143 .

Hpalz ^-^
0.169 pBR322

tRNASUP gene from pWM107 suppressed the amber mutation in the

Tag of am404. To test if the pWM107 could suppress in vivo, we

cut pWM107 with BamHI and pDR404 with PstI to free the SV40 mole-

cules from the pBR322 vector DNA. These SV40 molecules were now

religated to monomer circles and cotransfected into CVC cells.

Twenty-four hours later the cells were labeled with 32p ortho-

phosphate for an additional 48 hours. As can be seen in Figure 2,

lanes a-d, the WM107 alone and DR404 alone were unable to repli-

cate. But when these two molecules were cotransfected into the

same cells, replication of both molecules was seen (Fig. 2, lane

h). When digested with PstI both WMI107 and DR404 cut to a linear

form since both only have one Pstl site (Fig. 2, lane g).

This same replication experiment was repeated with a human

glutamine (GLN) tRNASUP gene (Temple et al., 1982). This

suppressor gene was inserted between M.U. 0.72 and M.U. 0.82 on

the SV40 genome in the clone pJYM-GLN41-116. In addition, the

Hpal fragment of Tag (M.U. 0.169 to M.U. 0.372) was removed

rendering it replication negative in clone pJYM-GLN41-116Hpa. The

SV40-GLN suppressor molecule showed less suppression than the

WM107 molecule when assayed with DR404 for replication (data not


To reaffirm that pSV-tT-2(SU+) could suppress the amber
mutation in pDR404, we assayed for immunofluorescence using anti-

sera to Tag in cells containing both pSV-tT-2(SU+) and pDR404.

For this assay it was not necessary to cut the SV40 molecule from

the plasmid DNA. Instead it was simply necessary to achieve high

copy numbers of both plasmids in the same cell. To accomplish

Figure 2. In Vivo Suppression of DR404 by WM107.
Molecules DR404 and WM107 were transfected either separately
or together into CV-C cells and labeled in vivo using 32p
orthophosphate. After HIRT extraction the DNA was fractionated on
a 1.4% agarose gel, the gel dried down, and the gel exposed to
film. Lanes a and b, mock transfection of CV-C cells uncut and
PstI cut respectively. Lanes c and d, WM107 transfected cells
uncut and PstI cut respectively. Lanes e and f, DR404 transfected
els uncut and PstI cut respectively. Lanes g and h, DR404 and
WM107 co-transfected cells uncut and PstI cut respectively.

a b c d e f g h

FmiII DR404

FmII WM107
S- FmI Ir#107
Fml DR404

--Fml WM107

this we transfected pSV-tT-2(SU+) and pDR404 into a nonsuppres-

sing strain of bacteria, E. coli C. Bacterial clones that

contained pSV-tT-2(SU+) or pDR404 were selected with ampicillin

or tetracycline, respectively. A clone containing both

pSV-tT-2(SU+) and pDR404 were selected using plates containing

both antibiotics. These three E. coli C isolates were then grown

to high density induced with chloramphenicol to amplify resident

plasmids, and digested with lysozyme to form spheroplasts. The

spheroplasts were then fused to CVC cells using polyethylene

glycol (Weiss, 1978; Rossoulzadegan et al., 1982). By using

antisera to Tag it was clear that cells fused with only pDR404 or

pSV-tT-2(SU+) did not show any immunofluorescence. But the CVC

cells fused with E. coli C containing both plasmids showed approx-

imately 2% immunofluroescence. The normal transfection efficiency

using E. coli spheroplasts and polyethylene glycol fusion is about

2-5%. This reaffirmed that the amber mutation in pDR404 was being

suppressed in vivo by the Xenopus TYR tRNASUP gene in



We wanted to show that the suppressor tRNA produced by

pSV-tT-2(SU+) was capable of suppressing an amber mutant in

vivo. To do this we utilized an existing SV40 amber mutant,

pDR4O4, which contained an amber mutation in Tag. The phenotype

of this mutant was replication negative and Tag immunofluorescence

negative and so restoration of these activities would indicate

suppression had occurred.

We first assayed for suppression by using the replication

assays. It was very clear from the replication assay that a low

level of suppression was occurring (Fig. 2). Unfortunately, the

amount of replication occurring from suppression was estimated to
be at least 100-fold less than the amount of replication occurring

by transfection of wild type SV40. In addition, other CVC plates

transfected with WMI107 and DR404 were assayed for having produced

virus by using a cell line supplying the Tag, the COS cell line

(Gluzman, 1981) (see Materials and Methods). No SV40 virus was

observed, probably due to the low level of suppression.

This low suppression level could be explained in two ways.

First, the amino acid glutamine that was mutagenized to the amber

codon in am404 was suppressed with a tRNA that inserted the

aromatic amino acid tyrosine. This change in amino acids, if in a

crucial area of the protein, could have affected the function of

Tag so drastically that even if present in large quantities repli-

cation would have occurred only at a low level. However, this

seems unlikely in that the functional human glutamine tRNASUp

gene showed less replication than the TYR tRNASuP gene.

Another possibility is that because WM107 lacked autonomous

replication there were never enough copies of the Xenopus TYR

tRNASUp to suppress DR404. If the DR404 molecules is not

suppressed due to lack of available TYR tRNAsup, then no Tag

is produced and neither molecule can replicate efficiently. This

possibility is supported by the E. coli spheroplast fusion data.

The fusion data showed that when the two plasmids pSV-tT-s(SU+)

and pDR404 were amplified in the same E. coli C and fused to

monkey cells, enough suppression took place such that the fused

cells were positive for Tag by immunofluorescence. When this

immunofluorescence experiment was carried out with WM107 and

DR404 transfected into monkey cells by DEAE dextran no immuno-

fluorescence was observed. This suggested that the level of

suppression could be greatly enhanced by increasing the number of

Xenopus TYR tRNASUp genes in the cell.

More recently, other data have shown very clearly that the

level of suppression in mammalian cells is greatly affected by the

number of Xenopus TYR tRNASuP genes present (Young et al.,

1983). Experiments showed that a 30 hour preinfection with

SV-tT-s(SU+) virus was necessary to suppress an amber mutation

in the adenovirus-SV40 virus hybrid, ND1-140 (Laski et al.,

1982a). This 30 hour preinfection gave the SV-tT-2(SU+) time to

replicate up to 105 copies per cell and provide the cell with

the same number of Xenopus TYR tRNASUP gene copies. The high

level of suppression was actually quantitated using a naturally

occurring influenza amber mutant (Young et al., 1983). In this in

vivo suppressor system no suppression was seen until 30 hours

after infection with SV-tT-2(SU+) virus. The maximum level of

suppression observed with SV-tT-2(SU+) in this influenza system

is 20%. More recently, two other mammalian amber suppressors, the

human lysine and glutamine tRNA genes, have been mutagenized to

read the amber stop codon. These two human suppressor tRNA genes

cloned into SV40 give in vivo suppression levels of about 20% if

cells are preinfected and then assayed in the amber influenza

system (Palese, personal communication).



The availability of a vector or system in which the number of

suppressor genes or expression of tRNASuP genes in a cell

could be controlled would be very useful. It would also be very

useful to have a vector capable of carrying suppressor tRNA genes

into mammalian cells without causing pathology to the cell. As

mentioned in the introduction, suppressor tRNA genes have been

introduced into cells by a variety of methods. Microinjection has

been used to isolate permanent suppressor cell lines containing

low levels of suppression. This is probably due to the cells' low

tolerance of high suppression levels over an extended period of

time. SV40 virus has been used to carry suppressor tRNA genes

into mammalian cells and replicate to high copy number. This

system is capable of high suppression levels but the cell is

rapidly killed by the replicating SV40 vector.

Adeno-Associated Virus (AAV) possesses some of the character-

istics that could make it a useful vector for tRNASuP genes.

AAV is a single stranded DNA virus capable of integrating into the

cellular genome thereby creating a latent infection (Hoggan et

al., 1972; Berns et al., 1982; Berns et al., 1975). This trans-

fected or latently infected cell is capable of being rescued and

amplified after the addition of adenovirus which provides a helper

function for the replication of AAV (Atchison et al., 1965; Parks

et al., 1967). This means that a cell line carrying an

AAV/tRNASUP recombinant genome could be induced to amplify the

suppressor gene by infection with adenovirus. Thus, if a

suppressor tRNA was inserted into AAV, it might be possible to

isolate amber mutants in either adenovirus or AAV itself. For

this reason it was important to determine if AAV is capable of

replication and packaging with the Xenopus TYR tRNASUp gene

inserted into the AAV genome.

AAV/pBR322 recombinants are infectious; this means that an

AAV recombinant virus stock can be isolated if AAV/pBR322 plasmid

DNA is transfected into adenovirus-infected cells. Therefore, the

simplest way to determine whether an AAV/sup recombinant was

viable was to clone the sup gene into the AAV sequences of infec-

tious plasmid, pSM620 (Fig. 3). This was accomplished by isolat-

ing the SV40 sequences from the plasmid pSV-tT-2sup+ Figure lb.

The tRNA gene was then isolated from the SV40 sequences on a 269

base pair Cfol fragment. Following isolation, the Cfol fragment

was blunted and ligated into pSM620 blunt-ended linears which had

been produced by a partial digestion with NcoI. The resulting

clones were screened by rapid plasmid isolation and clones

carrying a single copy of the Xenopus TYR tRNAsup gene in one

of the three Ncol sites were isolated, Figure 4. Plasmids pWCM106

and pWCM108 had copies of the tRNA gene located in the capsid gene

Figure 3. Physical Map of pSM620.
Thin line represents AAV DNA and thick solid line represents
pBR322 DNA. Open boxes represent the replication and capsid loci
for AAV.


INTRON m-u-48
\ M.U.60


N oNcol
/// ~m.u.80 \
// *N col \
M.Up7- PSM620 Nco -M.u.92

psti mt9 ipm.u.96
m.u .0.0 il M.U. 100

Figure 4. Physical Maps of Plasmids pWCM103, pWCM109, pWCM106,
and pWCM108.
The thin line represents AAV DNA, closed boxes are the AAV
inverted terminal repeats, and the wavy thin line represents
pBR322 sequences. The open boxes represent the AAV replication
and capsid loci. The tRNA gene inserts are shown as the open


a) -'-= p0 < _cT=_^_
t T

m.u.O pWCM103

m u.96

b) cZZZIz 0 < ZIZIDT.


c) < 30 < II-


d) = ,4 -- 0 < -- -

I 5 0
0 25 50 75 100

and the replication loci respectively. Plasmid pWCMI08 showed a
replication negative phenotype as expected. Plasmid pWCM106

showed a packaging negative phenotype as expected, but also

defined a new phenotype due to its inability to generate single

stranded monomer AAV DNA during replication (Hermonat et al.,

1984). pWCM103 contained a single copy and pWCM109 two copies of

the Xenopus TYR tRNASuP gene in the far right side Ncol site.

This particular Ncol site exists outside all known AAV reading

frames and 3' to the polyadenylation sequence for the AAV capsid

gene and therefore these pWCM103 and pWCM109 clones held the most

potential for replicating and packaging without needing AAV


Plasmids pWCM103 and pWCM109 were assayed for their ability

to replicate by transfecting them into HeLa cells and infecting

with adenovirus. Figure 5 shows that both pWCM103 and pWCM109

have the ability to excise from the pBR322 and replicate and that

they both seem to replicate as well as wild type AAV. To deter-

mine if the pWCM103 and pWCM109 would also produce virus, the

virus from transfected plates was amplified by second passage in

HeLa cells. The resulting virus was concentrated by equilibrium

CsCl centrifugation and analyzed by a Southern blot. Figure 6

shows that both the WCM103 and WCM109 genomes were capable of

packaging in spite of their increase in genome length. The virus

produced is referred to as AAV103 or AAV109.

The next logical step was to show that the Xenopus TYR

tRNASUp gene was being expressed in cells infected with

AAV103. This was accomplished by infecting Ad-infected HeLa cells

Figure 5. Replication Assay for AAV/sup Plasmids pWCM103 and
HeLa cells were infected with adenovirus and transfected with
either pWCM103 or pWCM109 and allowed to replicate for 30 hours.
The cells were then HIRT extracted, the DNA resolved on a 1.0%
agarose gel, the gel Southern blotted and hybridized to nick-
translated pSM620. The pWCM103 and pWCM109 transfected lanes are
labeled accordingly. The lengths of the three molecules (AAV,
WCM103, and WCM109) and their resolution on the 1% gel is shown
next to the figure. The band labeled SS represents the single
stranded AAV progeny.

C_) CD)

WCM103 I
4944 bpN WCM109

AAV 5213 bp
4675 bp


Figure 6. Packaging Assays for AAV/sup Plasmids pWCM109 and
HeLa cells were transfected with either pWCM103 or pWCM109
and infected with adenovirus. Forty-eight hours post-transfection
the AAV virus produced was used to infect 1 liter of HeLa cells in
suspension. The resulting virus was then isolated and the DNA
extracted, run on a 1.6% agarose gel, Southern blotted and
hybridized with 32p nick translated pSM620. In Figure 6a the
lanes labeled WCM109 and WCM109 x PstI are the viral DNAs from the
infected HeLa suspension cells. The lanes labeled WCM103 and
WCM103 x PstI are the viral DNAs from the infected HeLa suspension
cells. The lane labeled SM620 x PstI is a marker. All samples
cut with PstI before loading are labeled accordingly. Figure 6b
shows a map of the pSM620 cut with PstI and the resulting
fragments; A, B, C, and D. The location of the tRNA inserts in
fragment "D" is shown.

o 0. 0 0
x x


Pstl A

PstI B

S-- Pstl D + tRNA GENE
S690 bp

495 bp PstI C-- *
421 bp PstI D--

b) PstI Pstl PstI
pBR322 AAV DNA pBR322

with AAV103 or wild type AAV at a M.O.I. of five. The cells were

labeled in vivo with 32p orthophosphate and the cellular tRNA

was isolated and fractionated on a 8 molar urea, 12% polyacryla-

mide gel. Figure 7 shows that all the tRNA species from the

AAV-adenovirus lane except one are darker than the AAV103-

adenovirus lane. This single band that is darker in the AAV103

plus adenovirus lane represents the new Xenopus TYR tRNASUp

being synthesized from the Xenopus TYR tRNASUp gene in AAV103.

This experiment was repeated and the same result achieved.


There were several goals to be reached by putting the Xenopus

TYR tRNAsup gene into AAV. First, it was very likely that AAV

could be used as a mammalian cell vector to efficiently move the

Xenopus TYR tRNASuP gene into mammalian cells to create a

mammalian nonsense suppressor cell system. As mentioned earlier

the SV4O system developed was very short lived and the cell lines

that have been isolated to date have had very low suppression

levels. Secondly, we wanted to use AAV carrying the Xenopus

tRNASUp gene to isolate amber AAV and adenovirus mutants. Our

attempts to isolate these amber mutants will be discussed later.

Thirdly, we were interested in knowing if AAV was capable of

carrying any foreign gene into a mammalian cell, replicating, and

packaging normally, thus producing an AAV virus stock carrying the

inserted gene. This was of particular interest in that the only

gene that had been inserted to date, the adenovirus VA RNA gene,

had been partly or completely excised sometime before packaging

Figure 7. Comparison of tRNA Species Produced in AAV103 and AAV
Infected Cells.
HeLa cells were infected with adenovirus and either AAV or
AAV103. Twelve hours after infection the cells were labeled with
32p orthophosphate and the tRNAs from each infection harvested
18 hours later. The tRNAs were resolved on a 12%, 8 M urea
sequencing gel. The lanes loaded with tRNA from AAV or AAV103
infected cells are labeled accordingly. The new tRNA species is
shown by the arrow.


P ..C:

(Samulski and Shenk, personal communication). Finally, we hoped

the tRNA gene inserts would provide further information about the

genetics of AAV.

In order to insure that the AAV/tRNA recombinant would be

capable of replicating and packaging without complementation, the

TYR tRNASUP gene was inserted into the Ncol site at M.U. 96.

This new AAV/sup genome was clearly capable of replicating and

packaging without complementing wild type AAV virus. Plasmid

pWCM103 seemed to replicate as well as wild type but because of

its increase in genome length it only produced about one-tenth the

viable virus produced in a normal pSM620 transfection. In

contrast to the VA RNA results, no wild type AAV virus could be

detected after transfection with pWCMI03. This was probably

because the VA RNA gene is capable of forming a very stable

hairpin (Zain et al., 1979) and the mammalian DNA polymerase alpha

could be capable of replicating by DNA hairpins without copying

them. It was interesting that AAV plasmid pWCM109, which

contained two copies of the Xenopus TYR tRNASUp gene at M.U.

96, seemed to lose one copy sometime during replication and

produced only virus containing one full length copy. We believe

the tRNA genes were inserted in a head-to-tail manner in pWC109

and this formed a perfect substrate for legitimate recombination

and, therefore, deleted one copy of the tRNA gene.

Finally, it appeared that a new tRNA species was synthesized

in cells infected with AAV103. In this regard, it is not surpris-

ing that the Xenopus TYR tRNASUP comigrated with a normal HeLa

cell tRNA because with the exception of three base changes the


Xenopus TYR tRNA is the same size and sequence as the human TYR

tRNA. The fact that a new tRNA species was synthesized in AAV103

infected cells indicated that the TYR sup gene cloned into AAV was

probably being expressed.

The genetic data obtained from the tRNA gene inserts

supported the existing data about left and right side AAV mutants.

However, one new phenotype was observed for mutant pWCM106. This

AAV molecule was not capable of generating single strands after

replication. Presumably, one function of the capsid proteins is

the capture of single-strand progeny by encapsidation.



One problem of studying suppression in mammalian systems is

the lack of nonsense mutations in genes capable of being used as

markers in mammalian cells. As mentioned earlier there are

several genes containing amber mutations capable of being used as

amber mammalian markers. Some of these markers include the Tag of

SV40 (Rawlins and Muzyczka, 1980), the human beta globin gene

(Chang and Kan, 1979), the fusion protein of SV40-adenovirus NDI

hybrids (Gesteland et al., 1977), the Herpes thymidine kinase gene

(Cremer et al., 1979) and the E. coli xanthine-guanine phospho-

ribosyl transferase and aminoglycoside 3' phosphotransferase genes

(Hudziak et al., 1982). This list of amber mutants is very small

in comparison with the amber mutants available to bacterial

geneticists. Any new amber mutations in mammalian genes will be

useful in studying the mechanism and efficiency of mammalian

suppression along with providing information about the expression

and biochemistry of that particular gene product.

For these reasons it was useful to create a series of amber

mutations in the cloned AAV molecule pSM620. Amber mutations in

pSM620 would allow us to study the efficiency of suppression by

the Xenopus TYR tRNASUp gene in AAV and they would also

hopefully lead to the identification of new MAAV phenotypes.

Finally, the amraber mutations would be in a genome whose life cycle

was not inhibited by the addition of adenovirus (Giorno and Kates,

1971; Linberg et al., 1972).

A considerable amount is known about the genetics of AAV in

part due to the study of the AAV sup mutants described in the

previous chapter. The map of pSM620, Figure 2, shows the open

reading frames as determined by sequence analysis. This figure

also shows the areas that map to one of two known AAV phenotypes,

DNA replication or virus capsid production (Hermonat et al.,

1984). An amber mutation created between M.U. 7 and M.U. 43 would

produce an AAV genome defective for DNA replication and an amber

mutation created between M.U. 60 and M.U. 92 would produce an AAV

molecule defective for virion packaging. Finally, mutants within

map unit 48-52 were known to have the phenotype that they

replicated as wild type but produced tenfold lower amounts of

infectious particles.

Two strategies were used for isolating the amber mutants in

AAV. Mutants in all potential reading frames could be isolated in

pSM620 and potentially suppressed by cotransfection with another

AAV/sup molecule. Alternatively, the mutants could be isolated in

pWCM103 or pWCM106. In this case the strategy was to isolate

nonsense mutants and then select for those that remained viable.

This would be a class of mutants that would be suppressible by the

suppressor gene in pWCM103 or pWCM106. In general, our approach

would produce various AAV-pBR322 clones that contained a single

amber mutation, some of which would also contain a Xenopus TYR

tRNASuP gene. For either of these strategies to work it was

necessary to devise a method for creating amber mutants that had a

high success rate and would lead to mutations that could be quick-

ly and easily mapped and sequenced. The method that was chosen

was to insert a 9 bp oligonucleotide linker that contained stop

codons in two reading frames and single-cut restriction site.

The availability of the AAV sup virus AAV103 also suggested

the possibility of isolating amber mutations in adenovirus. The

idea was that the amber adenovirus mutants would allow AAV103

replication which in turn would suppress the adenovirus amber

mutation and allow the adenovirus to plaque. These plaques would

be picked and replaqued in the presence and absence of AAV103.

Mutants which plaqued only in the presence of AAV103 would be

identified a potential amber mutants. Amber mutations were intro-

duced into adenovirus by treating the virus in vitro with nitrous

acid or hydroxylamine causing base transitions (Ensinger and

Ginsberg, 1972; Williams et al., 1971) or treated while replicat-

ing in vivo with nitrosoguanidine or ethyl methanesulfonate

causing base transitions and transversions (Ensinger and Ginsberg,

1972; Williams et al., 1972; Crowe et al., 1978). These mutated

virus stocks were titered and replaqued in the presence of AAV103.


Isolating Amber Mutants in AAV

Amber mutations were created in genomes of pSM620, pWCM106,

and pWCM103. The pSM620 and pWCM103 genomes contained functional

replication and capsid loci and so amber mutations were created in

both of these loci. These molecules were then assayed for either

replication or virus production depending on the location of the

amber mutation. The pWCM106 genome only needed amber mutations in

the left half of the molecule since the suppressor tRNA gene

insert truncated the capsid genes on the right half. pWCM106 was

assayed for replication in the presence of the suppressor gene.

To make the amber mutations in the AAV molecules we synthe-

sized a 9 base pair linker that contained an internal Xbal site

and an amber stop codon in two reading frames. (There are no Xbal

sites in pSM620.) This linker was ligated into pSM620, pWCM103

and pWCM106 which had been partially cleaved with HaeIII to full

length linear molecules. These molecules were now cut with XbaI

and religated to monomer circles. pBR322 clones containing AAV

molecules which contained the amber linker were identified by

cleaving with Xbal and Pstl. This restriction digest also

identified the region containing the linker as either a left or

right side mutant. Maps of pSM620-21, pWCM103-1, and pWCM106-130

and the location of the useful amber linker inserts are shown in

Figure 8.

To check that mutants pSM620-21, pWCM106-130 and pWCM103-1

contained legitimate amber mutations, each of the amber inserts

and adjacent nucleotides were sequenced. This was accomplished by

cutting the mutants with Xbal, labeling the ends, and separating

the labeled ends by cutting with PstI. The sequence of the AAV

molecule before and after the amber linker insertion is shown in

Figure 9. This figure shows both pWCM103-1 and pWCM106-130

contain legitimate amber mutations.

Figure 8. Physical Maps of pWSM620-21, pWCM103-1, and
Thin straight lines represent AAV sequences and the filled
boxes, AAV inverted terminal repeats. Wavy lines represent pBR322
sequences. The open boxes represent AAV coding regions. The
amber linker and tRNA gene inserts are labeled accordingly.

m.u.O m.u.22 or m.u.29 m.U. 100

a) = ( )0 ( = > ="





m.u.80 m.u.96


T .

25 50 75 100


Figure 9. Sequence Analysis of Amber Linker Inserts in pWCM103-1
and pWCM106-130.
The sequences show the AAV HaeIII site and adjacent bases
before and after insertion of the amber linker. The amino acids
coded for by the sequence are shown below the sequence. The new
amino acids coded for by the amber linker are shown in bold face
type. The vertical arrows show the AAV HaeIII site in which the
amber linker was inserted.

M.U. 80.0





M.U. 32.0





Mutant pWCM103-1 (Fig. 8b) contained a Xenopus tRNA gene

outside M.U. 96 and an amber mutation at M.U. 80 in the capsid

loci. This amber mutation would presumably affect all three AAV

capsid proteins, VP1, VP2, and VP3, and would inhibit this virus

from packaging. This mutant was, therefore assayed for suppres-

sion by its ability to produce infectious virus. Mutant

pWCM106-130, Figure 8c, contained a Xenopus tRNA gene at M.U. 80

in the capsid coding region and an amber mutation at M.U. 32 in

the coding region for the replication protein. This amber muta-

tion in the replication protein coding region would inhibit this

virus from replicating and Suppression was detected by assaying

for replication. The reason for putting the amber mutation in the

same molecule as the Xenopus TYR tRNAsuP was to avoid any

recombination that could occur between the separate AAV molecules

containing the amber mutation and the Xenopus TYR tRNASUP

gene. However, keeping the amber mutation and Xenopus TYR

tRNAsup gene on separate molecules was attempted with one

amber mutant. Mutant pSM62U-21, Figure 8a, contained only an

amber mutation at M.U. 22 or M.U. 29 in the coding region for the

replication function. This plasmid was cotransfected with pWCMI08

which contained the Xenopus TYR tRNAsup gene at M.U. 13 of the

replication locus. Plasmid pWCM108 is capable of providing the

tRNASUp function while unable to supply the replication

function itself.

Plasmid pWCM103-1 was assayed for packaging and pSM620-21 and

pWCM106-130 assayed for replication (data not shown). Plasmid

pWCM106-130 and pWCM620-21 showed no replication and pWCM103-1

showed packaging only at a very low level. To prove the packaging

of pWCM103-1 was due to suppression, another AAV plasmid,

pWCM103-1-1, containing the same amber mutation but without a

suppressor tRNA gene was assayed for packaging. Plasmid

pWCM103-1-1 did not show evidence of packaging but the original

packaging result using pWCM103-1 could not be repeated.

To check the possibility that the lack of suppression in our

AAV system was a fault of the amber mutations themselves,

SV-tT-2(SU+) was used to suppress the AAV amber mutants. This

was done by cotransfecting the AAV amber mutant with SV-tT-2(SU+)

into HeLa or CVC cells. The SV-tT-2(SU+) was allowed to

replicate for 28 hours. At this point adenovirus was added to

provide the helper function for the AAV. The amberAAV mutants

were then assayed for suppression as before but no suppression was


Attempt to Isolate Adenovirus Amber Mutants

AAV103 was also used to isolate potential adenovirus amber

mutants. Adenovirus stocks were mutagenized with nitrosoguani-

dine, nitrous acid, ethyl methanesulfonate or hydroxylamine to

create amber adenovirus mutants. These mutagenized adenovirus

stocks were titered and the stocks whose titers had dropped one to

four logs were then used to infect cells which were later infected

with AAV103. (AAV103 was added 12 hours after the adenovirus to

avoid AAV's inhibition of adenovirus (Casto et al., 1967).)

Several thousand of the resulting adenovirus plaques were

isolated, the virus isolates diluted in PBS, and replaqued in the

presence and absence of additional MAAV103. It was expected that a

lethal amber adenovirus mutant would plaque only in the presence

of suppressor tRNAs supplied by AAV103. No adenovirus plaque

isolates were found that grew only in the presence of AAV103.

Since we failed to isolate any adenovirus amber mutants, we

checked if AAV103 was capable of suppressing a known adenovirus

mutant. We used the adenovirus-SV40 virus hybrid ND1-140

(Gesteland et al., 1977) amber mutant to assay AAV103 for suppres-

sion. This amber mutation is located in a SV40-adenovirus fusion

protein consisting of the carboxy terminus of Tag of SV40 fused to

a late adenovirus protein. It was isolated by mutagenizing the

wild type SV40-adenovirus hybrid, ND1 (Grodzicker et al., 1976).

Human cells were infected with either ND1 or ND1-140 and then

labeled with 35S-methionine. Protein extracts isolated from

these Ad-infected cells were then immunoprecipitated with hybri-

doma 412 antisera specific to the carboxy terminus of Tag. These

immunoprecipitates were then compared on a SDS PAGE gel. ND1

alone produced the full length 30 Kd fusion protein while ND1 plus

AAV103 produced almost none. The amber mutant ND1-140 produced no

30 Kd fusion protein with or without the AAV103, Figure 10. This

protein gel could not detect the truncated 19 Kd fusion protein

expected in the ND1-140 lane due to contaminating cellular



Making AAV Amber Mutants

There were two very important reasons for wanting to make

amber mutations in AAV genome. First, we needed an amber target

Figure 10. Inhibitory Effect of AAV103 on Adenovirus ND1-SV40
Fusion Protein Production.
HeLa cells were infected with either N01 alone, ND1-140
alone, ND1+AAVI03 or ND1-140+AAV103. Thirty hours after infection
the cells were labeled with 35s methionine and the Ad-SV40
fusion protein immunoprecipitated. Lanes are labeled accordingly.
The protein molecular weight markers are beside the gel. The
Ad-SV40 fusion protein is clearly inhibited in AAV103 infected


+ o

o "Iw
,-4 ,-4

.-4 -
U,, -;







Figure 11. Analysis of the Four Possible Amber Linker
Shown on top is the sequence of the amber linker and its
orientation designated by a horizontal arrow. The vertical arrows
show the location of the Xbal cut in the linkers. The four
possible insertion orientations are labeled a, b, c, and d. The
sequence and orientation of the two linker additions are shown
underneath resulting linker insert sequence. Amber stop codons
are underlined. The broken lines represent adjacent DNA



3 5 3' 5 3 5
I 1 1 .'. IGATC.TATC'" r
5. 3' 5- 3' 5 3 '


3' 5' 3' 5' 3' 5'
5' 3' T s' 3' f 5' 3'


3' 5' 3' 5' 3' 5'

5. 3-" (5 y '" 5. 3


3' 5' 1 3' T5' T r, 3' 5
1 AGATC.TATC'"~ ^ ~\ '1 r -
5' 3' .' 3' 5-" 3"


Figure 12. Physical Maps of Plasmids d152-91/Neo, d152-91/NeoX,
and d152-91/NeoB.
Thin straight lines represent AAV sequences and solid black
boxes represent the AAV inverted terminal repeats. Wavy thin
lines represent pBR322 sequences. Open boxes represent AAV coding
regions. Open dotted boxes represent SV40 promoter sequences and
hatched boxes represent neomycin coding sequences. The tRNA
inserts are labeled accordingly.

m.u.48 m.u.60

m.u.O dd52-91/Neo m.u.loo

b) -=* > o^ T r r' '"" -^_




0 25 50 75 100

gene that was uninhibited by adenovirus to test the suppression

efficiency of AAV carrying the Xenopus TYR tRNASUp gene. The

extremely strong inhibition of transcription by the helper adeno-

virus replication left few mammalian genes capable of being used

in this system with the exception of MAAV or adenoviral genes. We

eventually decided to attempt to make amber mutations in both of

these viruses. The second reason for making the amber mutations

in AAV was to study the effects these mutations might have on the

biology of AAV.

Amber mutations have been created in genes in several ways.

Random chemical mutagenesis has been used to create amber muta-

tions in genes both in vivo and in vitro. Specific chemical

mutagenesis was also used to create an amber mutation in the large

T antigen of SV40. If the gene to be mutagenized is well mapped

and sequenced, all or part of this gene could be cloned into the

bacteriophage M13 and site specifically mutagenized. Unfortu-

nately, these three methods involve tedious mutagenesis procedures

or time consuming screening procedures. For these reasons we

tried to devise a simple more efficient method of making amber


The method we used to create amber mutants involved synthe-

sizing a double stranded DNA oligomer that contained a restriction

site not found in AAV or pBR322 (Xbal). If this amber linker was

inserted into the correct reading frame it could create an amber

mutation. This was accomplished by lighting the blunt ended amber

linker into the linearized pSM620 partially cut with a restriction

endonuclease that created a blunt end such as HaeIII. These

molecules could now be cut with Xbal and religated to insure that

only one copy of the linker was inserted. The amber linker

containing clones resulting from the transfection of these ligated

molecules could now easily be identified by their susceptibility

to cleavage with XbaI.

We wanted to create an amber linker that contained amber

mutations in at least two reading frames, contained a site not

found in AAV or pBR322, and added as few new amino acids to the

protein as possible. One of the few restriction enzymes that did

not cut pBR322 or AAV was XbaI. The Xbal recognition sequence is

5' TCTAGA 3' which already contains one TAG amber stop codon.

Therefore, we needed to add only one more TAG amber stop codon to

the 3' end of the recognition sequence to have a DNA linker

containing amber stop codons in two reading frames and the recog-

nition sequence for XbaI. Figure 11 shows the sequence of the

amber linker, the four possible variations in which the linkers

could have ligated to the site, and the insert sequence that would

result from the four variations. This figure shows that 25% of

the molecules will contain one stop codon and insert three

additional amino acids; 25% of the molecules will contain two stop

codons and insert three additional amino acids; 25% of the

molecules will contain two stop codons and insert four additional

amino acids; and 25% of the molecules will contain one stop codon

and insert two additional amino acids. The sequence data of two

of the three amber AAV mutants that were assayed for suppression

showed that pWCM106-130 was in frame and added three amino acids

and pWCM103-1 was in frame and added three amino acids (Fig. 9).

In neither case were the flanking amino acids changed by the


Assay for Suppression of AAV

As mentioned before, the strong inhibitory effect of

adenovirus on cellular transcription persuaded us to use AAV as

the amber marker when assaying the AAV carrying the Xenopus TYR

tRNASUp gene for suppressor activity. Therefore, the assay

for suppression was to transfect into human cells AAV molecules

that contained both an amber mutation and a suppressor gene or

cotransfect mixtures of AAV molecules in which one contained an

amber mutation and the other a Xenopus TYR tRNASUP gene.

These cells would then be assayed for AAV replication or packaging

by Southern blotting as described in Materials and Methods.

When these plasmids were transfected into human cells by

either DEAE dextran or calcium phosphate there was no indication

of any suppression occurring with one exception. This exception

occurred at a low level with pWCM103-1 and this result could never

be repeated. It is unclear why the AAV suppressor systems did not

work. The problem could be that the Xenopus TYR tRNASUP gene

was not being expressed in the AAV vector. This idea is supported

by an experiment in which AAV103 was used at a M.O.I. of 500 to

infect cells containing the amber SV40 mutant DR404 and yet no

suppression as assyed by immunofluorescence or replication was

observed (data not shown). However, as shown in the previous

chapter, a new species of tRNA is produced in cells infected with

a AAV103 virus stock.

Therefore, the most likely explanation for our inability to

suppress the AAV amber mutants is that the linkers lead to the

addition of two to four new amino acids into the target protein.

So, even when the stop codon is suppressed, the suppressed protein

still has an insertion in it. To convince ourselves that the AAV

amber mutants were nonsuppressible, we did the following experi-

ment. The amber AAV mutants were cotransfected into both monkey

and human cell lines with SV-tT-2(SU+) and allowed to replicate

for 30 hours before adenovirus was added. No evidence of

suppression was observed for any of the AAV amber mutants even

though this preinfected SV-tT-2(SU+) system should give 20%

suppression (Young etal., 1983). There is no way to predict if

small insertions of two to four amino acids will affect the

function of a particular protein. The effect would vary from one

protein to another and greatly depend on where the insertion

occurred in a particular protein.

Using AAV to Isolate Adenovirus Amber Mutants

One of the original intentions of making an AAV molecule

containing the Xenopus TYR tRNASuP gene was to use it to

isolate amber adenovirus mutants. The scheme was to lower the

plaquing titer of a adenovirus stock from one to four logs by

random chemical mutayenesis hoping some of the mutants created

would be amber mutants. The methods of random mutagenesis by

nitrosoguanidine, nitrous acid, hydroxylamine, and ethylmethane

sulfonate were used to create amber mutations in adenovirus

(Grodzicker et al., 1976) and bacteriophages T-4, T-2, X174

(Drake and Kock, 1976). The mutagenized adenovirus stocks

were used to infect human 293 cells and 12 hours later AAV103

added at an M.O.I. of 5 and the adenovirus allowed to plaque.

Several thousand adenovirus plaques were isolated and replaqued in

the presence or absence of AAV103. We hoped to find adenovirus

isolates that were now unable to grow in the absence of AAV103 but

none were found. Three possible explanations could explain this

data: 1) we failed to make adenovirus amber mutants, 2) the

AAV103 was not producing sufficient amounts of the Xenopus TYR

tRNAsup, and 3) the AAV103 was inhibiting the production of

virus in the normal adenovirus lytic cycle. The third possibility

seems to be the most likely in view of the fact that the AAV virus

has been shown to inhibit the production of adenovirus in a

coinfection (Casto et al., 1967). In this regard, we showed

specifically that AAV103 was capable of almost completely

inhibiting the synthesis of the fusion protein of an adenovirus

SV40 (ND1) hybrid.



Very important to the study of mammalian nonsense suppressors

is having cell lines stabily transformed with suppressor tRNA

genes. Stable suppressor cell lines would avoid the problem of

using a lethal viral vector to carry the suppressor tRNA genes

into the cells to achieve high copy number such as with

pSV-tT-2(SU+). But to date, microinjection of up to 10,000

copies of pSV-tT-2(SU+) has produced cell lines with only about

a 4% suppression level (Hudziak, 1982; Young et al., 1983). We

believe the reason for this is that suppression levels over 4% are

lethal in mammalian cells. Therefore, it would be advantageous to

have a nonlethal vector that could carry in one copy of the

suppressor tRNA gene into the cell and that could be specifically

induced to amplify when a high copy number of the suppressor was


One viral vector that could fulfill the requirements of being

inducible and nonlethal is AAV. AAV is capable of stable

nonlethal integration into the cells genome and it can be rescued

and replicated to high copy number upon infection with adenovirus

(Atchison et al., 1965; Parks et al., 1967). Of course the

adenovirus necessary for induction is also lethal to the cell.

Therefore, the regions of the adenovirus genome responsible for

AAV replication would have to be identified and these regions

introduced into cells carrying the suppressor AAV genomes. This

has not yet been accomplished although work in this area has shown

some encouraging results (McLaughlin, S. and Muzyczka, N.,

personal communication).

In anticipation of adenovirus transformed stable cell lines

capable of replicating AAV, we made stable cell lines transformed

with AAV carrying a selectable marker and a suppressor tRNA gene.


Our intention was to put the available Xenopus TYR

tRNASUp gene into the existing AAV molecule, d152-91/Neo,

which contains the neomycin resistance gene (Southern and Berg,

1982) in place of the AAV capsid loci (M.U. 52 to M.U. 91)

(Hermonat and Muzyczka, in press) (Fig. 12a). Mutant d152-91/Neo

is capable of replicating in human cells upon infection with

adenovirus. It is also capable of being packaged if complemented

with another AAV virus that can provide the capsid genes (Hermonat

and Muzyczka, in press).

The Xenopus TYR tRNASUP gene was excised from

pSV-tT-2(SU+) (as described previously) on a 269 bp Cfol

fragment. BglII or Xhol linkers were attached to this 269 base

pair fragment inserted into dl_52-91/Neo that had been cut with

either Xhol or BgII. Figure 12b,c shows the location of the

inserted sup TYR tRNA gene with respect to the neomycin gene in

d152-91/NeoB and d152-91/NeoX.

Mutants dl_52-91/NeoX and dl52-91/NeoB were now transfected

with a cloned AAV molecule ins96/A-M, which provides the capsid

genes for dl152-91/NeoX and dl52-91/NeoB but is too large to

package itself (Hermonat and Muzyckza, in press). The resulting

virus stock was then used to infect a human Detroit 6 cell line

and the cells were selected in medium containing neomycin. The

cell clones that survived selection were picked and expanded as

separate cultures. We showed that several of these cell cultures

could amplify the integrated AAV genome carrying the Xenopus TYR

tRNAsup gene in response to adenovirus infection, Figure 13.


Suppressor Cell Lines

The creation of mammalian suppressor cell lines is now

recognized as a difficult problem for several reasons. The major

problem appears to be the mammalian cell's inability to tolerate

high levels of suppression. While this idea has not been proven,

it is implied by Xenopus TYR tRNASUP gene microinjection

results. These results show that whether 100 copies or 10,000

copies of the tRNASuP gene are injected into mammalian cells

only a 4% suppression level is observed in the dividing clonal

colonies that result. Those cells with a suppression level higher

than 4% are thought to die after very few generations. This

problem could be alleviated by using an inducible promoter to

control expression as could be done with a polymerase II gene.

But the tRNA genes are expressed by polymerase III and to date no

inducible promoter has been isolated that functions with the

Figure 13. Rescue of AAV/tRNASUP Virus from d152-91/NeoX and
dl152-91/NeoB Transduced D-6 cells.
Neomycin resistant D-6 cells clones which had been infected
with a virus stock of either dl_52-91/NeoX or dl52-91/NeoB were
grown to confluency in separate 6 cm dishes. These cell clones
were now infected with adenovirus and 48 hours later the cells
lysed and the low molecular weight DNA extracted. The DNA was cut
with Pstl, electrophoresed on a 1.2% agarose gel, Southern
blotted, and hybridized with 32p nick translated dl52-91/Neo
DNA. Figure A shows an autoradiograph of the rescueable
AAV/tRNASUp from representative dl_52-91/NeoX and d152-91/NeoB
cell lines. The lanes are labeled accordingly. Normally,
1463 bp, 1350 bp, and 800 bp fragments would result from
d152-91/Neo cut with Pstl. However, since the 1350 bp fragment
contains the 269 bp tRNAsuP gene in either the Bglll or Xhol
site, it should increase in size to 1619 bp. Figure B shows a map
of the Pstl sites in dl_52-91/Neo and the location of the
tRNASUp inserts.

4-1 "j

x "
o 0
(1 (3)
z z

^ a,
~I ~I


1619bp- -

1463bp i
1350bp -

800bp -- .

insert insert

I 1463 I 1350 T 800 I
az ) m m

polymerase III enzyme. One final way to alleviate this problem

could be to use a viral vector to carry the tRNASuP gene that

could be induced to amplify, thus amplifying the tRNASUP gene.

This type of system may be possible with AAV.

As mentioned earlier, adenovirus gene products are critical

to AAV replication but adenovirus itself is very lethal to the

cell. But temperature sensitive mutants in adenovirus have been

isolated which, when infected into cells at the nonpermissive

temperature, replicate extremely poorly and allow the cells to

divide and survive. Preliminary results show these same cells are

capable of supporting AAV replication at a low level (McLaughlin,

S. and Muzyczka, N., personal communication). If the adenovirus

genes responsible for AAV replication could be identified and

their gene products controlled with inducible polymerase II

promoter (such as the heat shock promoter), a cell line capable of

inducing AAV replication is possible. This type of AAV inducible

system could be ideal for creating an inducible suppressor cell


In anticipation of this possibility we inserted the Xenopus

TYR tRNASUP gene into the Xhol site at M.U. 48 of d152-91/Neo

and into a BglII site only 5 bases 5' of the start codon for the

neomycin resistance gene. The resulting virus stocks made from

these AAV plasmids were used to infect Detroit 6 cells and

neomycin (G-418) resistant cell clones were isolated. It seems

unlikely that these tRNA yenes were lost due to thile recombination

that always occurs when making mixed virus stocks (Hermonat et

al., 1984; Hermonat and Muzyczka, in press) since in one of the


clones d152-91/NeoB, the tRNA gene is within 5 bases of the

neomycin resistance gene start codon. In addition, some of the

cell lines carrying the AAV-suptRNA gene hybrid molecules have

been shown to excise (rescue) from the cellular DNA and replicate

upon addition of adenovirus. The AAV molecules that do rescue and

replicate have been shown to contain the tRNASUp gene.



One of the major pitfalls of mammalian genetics today is the

inability to transfer genes into mammalian cells. Using mechan-

ical methods such as DEAE dextran or calcium phosphate, transient

gene expression levels can be achieved in about 10% of the cells

treated. But in many instances a permanent transformation of the

cell is desired and the frequency of stable transformation is only

1 in 105 for calcium phosphate and nonexistent for DEAE dextran.

Microinjection has a much higher transformation frequency and

produces stable transformants in about 1% of the cells injected.

This method is very useful for some purposes, but if the object,

for example, was to "shotgun" DNA into a mammalian cell, this

technique would be much too inefficient.

To avoid the problems mentioned above, different mammalian

viruses are being used as vectors to transfer DNA into mammalian

cells. SV40 virus was the first mammalian virus to be engineered

to successfully transfer selectable markers into mammalian cells

(Goff and Berg, 1976; Goff and Berg, 1979; Uchida and Watanabe,

1969), but it was quickly discovered that SV40 tended to rearrange

quite rapidly after its introduction into the cells (Uchida et

al., 1968; Yoshiike, 1968; Uchida and Watanabe, 1969; Lavi and

Winocour, 1972; Tai et al., 1972; Martin et al., 1973; Brockman et

al., 1973; Fareed et al., 1974; Khoury et al., 1974; Ganem et al.,

1976). In addition, it would tolerate only small insertions (due

to its small size if packaging was required) and seemed to have a

toxic effect on the cell when it was replicating episomally (Lai

and Nathans, unpublished data). While SV40 has been used in some

cases to stably transform mammalian cells, the problems mentioned

above preclude it's use as an efficient mammalian viral DNA


Another viral vector, this one having been used very success-

fully, is Bovine Papilloma Virus (BPV) (DiMaio et al., 1982;

Kushner et al., 1982; Sarver et al., 1982; Law et al., 1983;

Maroteaux et al., 1983) BPV is a double stranded DNA virus that

appears to replicate to between 20 and 100 copies (Law et al.,

1981) per cell and has been shown to stably transform several

mammalian cell types while remaining in an episomal form. BPV

also does not appear to go through severe rearrangements after its

introduction into the cell. In addition, BPV is capable of being

propagated in E. coli in a pBR322 derivative that has the pBR322
"poison sequences" deleted (Kushner et al., 1982). This molecule

is now capable of episomal replication when transfected into

mammalian cells. Unfortunately, several genes inserted into BPV

have been shown to be inhibitory to BPV replication (DiMaio et

al., 1982). This means each gene to be inserted into BPV must be

checked for it's compatibility with BPV replication. In addition,

BPV's host range is limited so far to rodent cell lines.

The retroviruses are another class of animal virus that have

been used to efficiently transfer DNA into mammalian cells. The

retroviruses exist as a linear single stranded dimer RNA molecule

in the viral capsid and code for a reverse transcriptase which

copies the single stranded RNA into DNA in the cell. This double

stranded DNA now possesses the ability to integrate into the

cell's genome through its ends. The coding regions of the virus

contain genes for the reverse transcriptase, the capsid protein,

an RNA binding protein and depending on the variety of retrovirus,

an oncogenic gene product. The virus also contains long terminal

repeats (LTRs) on it's ends which are extremely strong promoters

for transcribing both strands of the virus in the provirus. These

terminal repeat promoters can also be used to efficiently express

inserted genes. However, the LTR promoters can be a problem if

the expression of the inserted gene's own promoter is to be

studied. Retroviruses have been shown to transfer and efficiently

express foreign genes in several different cell lines (Weeks et

al]., 1982; Joyner and Bernstein, 1983; Joyner et al., 1983; Miller

et al., 1983). This achievement was aided by retrovirus

transformed cell lines that can supply all replication and

packaging functions for the recombinant retrovirus (Mann et al.,


As discussed earlier adeno-associated virus is a single

stranded DNA virus that exists in a of double-stranded linear form

after replication in the cell. AAV, like the retroviruses, seems

to have the capability to integrate into the cell's genome. The

integration seems to manifest itself as a set of tandemly repeated

full length AAV molceules with the integration occurring seemingly

randomly within the junction AAV molecules (Hermonat and Muzyczka,

in press; Collis, P. and Muzyczka. N., personal communication).

As mentioned before, the addition of adenovirus to cells latently

infected with AAV or transfected with pSM620 causes MAAV replica-

tion and virus production. In addition, MAAV has a very broad host

range which is a valuable characteristic for a mammalian cloning


For the reasons mentioned above, AAV was recognized as a

possible mammalian cloning vector. As discussed earlier, AAV was

used to effectively transfer the small 269 base pair DNA fragment

containing the Xenopus TYR tRNAsuP gene into several human

cell lines. Since then much more extensive experiments have been

completed in which a large portion of the capsid coding region of

AAV was deleted and a 2000 base pair DNA fragment containing the

neomycin resistance gene inserted under the control of an SV40

promoter. When this new AAV construct, d152-91/Neo, is trans-

fected with an AAV molecule supplying the capsid genes, an AAV

virus stock can be made which contains the neomycin resistance

gene. This virus stock has been used to infect and confer

neomycin resistance to Detroit 6 cells. This stable transduction

occurs at a frequency of about 1% at a M.O.I. of 1 (Hermonat and

Muzyczka, in press; Collis, P., personal communication).

It was clear that AAV would be useful as a mammalian cloning

vector, however, there were several points that needed clarifica-

tion. It was unclear whether the left hand side of AAV, involved

in replication, was necessary for integration into the cellular

genome. If the left side was not critical, the entire coding

region of about 4000 base pairs could be excised and that space

used to insert a gene of equal size. It was also unclear whether

the cells that had been transduced by dl152-91/Neo were the same

phenotypically as the cells that had remained untransduced or

whether they possessed some characteristic that conferred a

"highly transducible" phenotype. Finally, since it was clear that

the ends of AAV were crucial to replication it was not known

whether a large fragment inserted in between the left and right

reading frames, thus separating the ends by several extra

kilobases, would affect the replication of the molecule. This

large fragment could also affect the processing of one AAV

transcript which spans the entire length of the genome between the

inverted terminal repeats.



The question of whether the left side of AAV, providing the

replication function, was crucial for stable integration was

addressed by inserting the E. coli xanthine-guanine phosphoribo-

syltransferase (XGPRT) yene into the left side of AAV. The XGPRT

gene is the prokaryotic equivalent of the hypoxanthine phosphori-

bosyltransferase (HPRT) in mammalian cells. Their functions are

to convert hypoxanthine and guanine to inosine monophosphate

(IMP) and guanosine monophosphate respectively (GMP). Unlike the

HPRT gene, the XGPRT gene is also capable of converting xanthine

to xanthosine monophosphate (Krenitsky et al., 1969; Miller et

al., 1972). If aminopterin and mycophenolic acid are added to

mammalian cell culture media the mannmmalian cells are incapable of

making GMP and die. However, if the E. coli XGPRT gene is

functioning in the mammalian cell the cell can make GMP from

xanthosine monophosphate in the presence of aminopterin and

mycophenolic acid and the cell continues to divide (Mulligan and

Berg, 1980).

The XGPRT fragment to be cloned into AAV was isolated from

pSV2-GPT (Mulligan and Berg, 1981) by cutting with either PvuII

and AatII or PvuII and PstI, Figure 14a. The PvuII-AatIIl fragment

was blunted and inserted into a BglII-BstEII deletion created in

di_3-22, Figure 14b. This would remove the left side promoter at

M.U. 3 and replace it with a SV40 promoter and remove the

polyadenylation signal present in pSV2-GPT using instead the AAV

polyadenylation signal present at M.U. 47 or 94. The inverted

terminal repeats, the promoter for the right side capsid

transcripts and the right side capsid genes would be left intact.

This AAV plasmid termed pWMGPT-1 is shown in Figure 14c. The

PvuII-PstI fragment was blunted and inserted into a deletion in

d1_3-22 made by cutting d13-22 to completion with BglII and Ncol,

Figure 14b. This deletion encompasses the AAV sequences between

M.U. 3 and M.U. 96. The PvuII-PstI XGPRT fragment contains the

SV40 polyadenylation signals. This AAV plasmid, pWMGPT-2, is

missing almost all AAV sequences except the inverted terminal

repeats, Figure 14d.

Plasmid pWMGPT-1 was now transfected with either d152-91/Neo

or ins96/X-M by DEAE dextran to provide the replication function

for pWMGPT-1. The resulting virus stocks were now used to infect
105 Detroit 6 cells in culture at an M.O.I. of 0.01 and after 72

hours these cells were split 1:5 and selected in HAT media