JOURNAL OF VIROLOGY, OCt. 1991, p. 5237-5243

Vol. 65, No. 10

0022-538X/91/105237-07$02.00/0 Copyright © 1991, American Society for Microbiology

Mutational Analysis of the Simian Immunodeficiency Virus SIVmac nef Gene DOROTHEA BINNINGER, JOACHIM ENNEN, DANIELA BONN, STEPHEN G. NORLEY, AND REINHARD KURTH*

Paul-Ehrlich-Institut, Paul-Ehrlich-Strasse 51-59, 6070 Langen, Germany Received 12 March 1991/Accepted 24 June 1991

We are using site-directed mutagenesis of single viral genes to identify and analyze the genetic determinants of human and simian immunodeficiency virus pathogenicity. In a first approach, we have constructed a series of simian immunodeficiency virus SlVmac nef mutants by partial deletion and insertions in the nef gene, as this gene is a candidate gene for the establishment and maintenance of latency. nef insertion mutants replicated faster than wild-type SIVmac, suggesting that the nef gene product acts as a negative factor for replication. Surface phenotyping revealed that cultures permanently infected with nef mutants exhibit an enhanced expression of viral proteins on the outer cell surface. We have analyzed the properties of the mutant viruses in cell culture and intend to use rapidly replicating mutants (putatively unable to undergo latency) as model vaccine viruses in the rhesus monkey.

nef, a regulatory gene of human and simian immunodeficiency viruses (HIV and SIV), is a unique feature of the primate lentiviruses and does not exist in other animal lentiviruses (17). Conservation of nef in HIV and SIV suggests that the nef protein confers a biological advantage in the evolution and the life cycle of these viruses. The HIV/SIV nef gene encodes a protein of 27 kDa, which is myristylated, phosphorylated, localized in the cytoplasm, and at least partly membrane associated (2). The nef gene product is not essential for HIV replication (24). Deletion of nef in an infectious provirus increases its replication in human lymphoblastoid cells (1, 16, 18), and this observation led to the concept that the nef gene product negatively regulates HIV replication. It was suggested that the target sequences, mediating the suppressive effect of nef, are localized 5' of the RNA initiation site on the long terminal repeat (LTR) between positions -340 and -156 (1, 8) in a region containing negative regulatory elements (17), although recent publications disagree with these conclusions (10, 12). Bachelerie et al. (4) observed that constitutive expression of the HIV nef protein in a human astrocyte cell line does not influence basal or induced HIV LTR activity. Nef shares sequence similarities with members of the proto-oncogene protein family such as p2lras and p60rc (9, 23). It was furthermore shown that p21 is able to induce transactivation of the HIV-1 LTR (3). These findings led to the prediction that Nef acts in conjunction with other cellular regulatory proteins in a manner similar to that of G proteins (23). The idea gained support by the demonstration that GTP-binding and GTPase activities are associated with partially purified HIV-1 nef protein expressed in Escherichia coli (9). In contrast, other authors could not substantiate that purified HIV-1 nef protein contained GTP-binding or GTPase activity (11). Cheng-Mayer et al. (6) recently observed different effects of nef on HIV replication, using stable lymphoid cell lines expressing the HIV-1 nef gene product. The presence of nef protein suppressed replication of some strains of HIV-1 and HIV-2, whereas fast-replicating and highly cytopathic HIV-1 *

Corresponding author. 5237

isolates recovered from patients with advanced disease states were not affected. These results suggest that progression of disease in the host correlates with appearance of virus variants that are less responsive to negative regulation by nef. Thus, Nef may play a key role in the establishment and maintenance of latent viral infections and in HIV pathogenesis. The establishment of a latent infection during the long asymptomatic incubation period may be part of the mechanism of viral pathogenicity. Continuously growing retroviruses insert viral neoantigens into the infected cell membrane, resulting in recognition and elimination by the immune system. In contrast, cells carrying intracellularly suppressed proviruses are not recognized, permitting the maintenance of a virus reservoir. Latent lentiviruses may thus avoid immune attack and clearance from the organism. Hypothetically, if mutant viruses are rendered incapable of establishing latency, they will continue to replicate and remain vulnerable to the host's immune response. This hypothesis is testable in the SIV/rhesus monkey system, currently the best animal model for human AIDS. We therefore constructed a series of SIVmac nef deletion and insertion mutants for the evaluation of the mutants both in vitro and in vivo. MATERIALS AND METHODS

Construction of SIVmac nef mutants. Starting material was the biologically active proviral SIVmac251 DNA clone BK28 (14), kindly provided by J. Mullins (Stanford University). BK28 also contains flanking host cell sequences and is cloned into the EcoRI and HincIl restriction sites of pUC18. A 1.7-kb NheI-EcoRI fragment was ligated with EcoRI linkers (the NheI site remains intact) and inserted into the EcoRI restriction site of pUC18. The resulting plasmid, pSIVmacBK28-EE, starting at nucleotide 8218, comprises the complete 3' end of the viral genome. pSIVmacBK28-EE was used to construct all SIVmac nef mutants. Clone pBK28-EE/PCRwt was constructed by the polymerase chain reaction (22), using a primer pair to amplify the region from position 8218 (primer env+; 5'-GAATTCGCTAGCTAAGT TAAGGCAGGGGG-3') to the 3' end of the 3' LTR US

5238

J. VIROL.

BINNINGER ET AL. #9296

#8553

nef

2

env

#8717

#8279

IR

U3

U5s k

ZZ

#8937 EcoRI

NheI

Bglll

NcoI

NdeI

#8218

#8594

#8757

#9226

pBK28/wt

pBK28/PCR/wt

L

CTAGTCIrAGACrAG pBK28nefBS

CTAGTCCAGACIrAG i

____

. . __,

pBK28nefNS pBK28 Anef3

#8929

#8698

pBK28 Anef 13

#

l

#8863

#8679

i

#8885 #8679 FIG. 1. SIVmac251 DNA clone BK28 nef subclones. A premature stop codon is located in the

region (primer LTR-; 5'-GAATTCATGCTAGGGATTT TCCTGCTTCGG-3'). This clone was needed to investigate the influence of the 3'-flanking host cell sequences on virus expression.

Construction of nef insertion mutants. Stop codons (SMURFT linker; Pharmacia) were introduced into the BgIII (8594) and NcoI (8757) restriction sites (after conversion of protruding 5' ends to blunt ends with E. coli DNA polymerase I Klenow fragment) of pSIVmacBK28-EE, resulting in plasmids pBK28neJBS and pBK28neJNS, respectively. Construction of nef deletion mutants. pSIVmacBK-EE was linearized by restriction with NcoI, subjected to digestion with exonuclease Bal3l, religated, and transformed into E. coli DH5cx competent cells. For all transformants with intact BglII and NdeI (9226) restriction sites, the exact extent of the deletion was determined by dideoxynucleotide sequencing. Three clones carry deletions that do not extend into the U3 region of the LTR: pBK28AneJ3, pBK28zAneJ5, and pBK28Anefl3. Transfection. SW480 cells (American Type Culture Collection [ATCC]) were grown at 37°C in Dulbecco's modification of Eagle's medium supplemented with 10% (vol/vol) fetal calf serum, 500 jig of penicillin per ml, and 100 ,ug of

pBK28 Anef5 env gene

at position 8279.

streptomycin per ml. The different NheI-EcoRI fragments of the mutated nef gene fragments were ligated to the 15-kb pBK28 NheI-EcoRI vector fragment and used to transfect subconfluent cultures of SW480 cells as described previously (5). One day posttransfection, cultures were cocultivated with Hut78 cells (ATCC) for about 12 h. Infected Hut78 cells were subsequently removed and maintained in RPMI 1640 medium supplemented with 10% (vol/vol) fetal calf serum, 500 jig of penicillin per ml, and 100 pg of streptomycin per ml. Production and titration of virus stocks. Cell-free culture supernatant from permanently infected Hut78 cell cultures was passed through a 0.22-.Lm-pore-size filter and stored as aliquots in liquid nitrogen. Titration of the virus stocks was performed by endpoint dilution on MT4 cells (ATCC; human T-cell lymphotropic virus type I-transformed T-cell line), and the endpoint was determined by the presence of cytopathic effects in cell culture as described previously (19). Reverse transcriptase (RT) assays. The assays were performed according to published procedures (13). Immunofluorescence staining and fluorescence-activated cell sorting (FACS) analysis. For cell surface staining viable infected Hut78 were resuspended at 106 cells per ml in

TABLE 1. Mutations introduced into the nef gene Mutant

Mutant

pBK28nefBS pBK28nefNS pBK28AneJ3 pBK28Anefl3 pBK28Anej5

Deletion from:

Deletion from:

Position of

~introduction

Frameshift

aa

before

Frmsitframeshift

aa

after

frameshift

8594 (stop) 8757 (stop)

8699-8928 8680-8862 8680-8884

2+ 0 2+

a Additional amino acid introduced because of the position of the deletion.

49 42 42 + 1"

15 144 5

Length of NEF (aa) 13 68 49 186 42

SIVmac nef GENE MUTANTS

VOL. 65, 1991

5239

were incubated with PBS supplemented with 2% BSA, 5% goat serum, and 0.002% Triton-X100 (Sigma) for 30 min to reduce nonspecific binding. Cells were resuspended for 30 min in a 1/1,000 dilution of serum from an HIV-2-positive individual. An HIV antibody-negative human serum served as a control. Results are given as percentage of positively stained cells. Mean channel of fluorescence intensity was calculated (logarithmic amplification of fluorescence signal).

108

2. 1._

d.p.i.

FIG. 2. Kinetics of RT activity in the cell-free supernatant of Hut78 cells infected with SIVmac nef mutants at an MOI of 0.002. Cells were split 1:2 every 2 days, and cell-free supernatant was assayed for RT activity.

Ca2+/Mg2+-free phosphate-buffered saline (PBS) supplemented with 0.1% NaN3 and 2% bovine serum albumin (BSA). Aliquots (0.1 ml) were treated with a monoclonal antibody specific for either the CD4 receptor (Leu3a; Becton Dickinson), LFA-1 adhesion molecule (IOT16; Immunotech), or HLA-ABC (IOT2; Immunotech) for 15 min. For the staining of viral proteins expressed on the cell membrane, serum from an HIV-2-positive individual was used (with serological reactivity against env and gag of HIV-2 as detected in immunoblots). As second antibody, fluorescein isothiocyanate conjugates of goat anti-mouse immunoglobulin G (Jackson, Immunoresearch Laboratories USA) or phycoerythrin conjugates of goat anti-human immunoglobulin G (Jackson) were used. After washing, cells were fixed with PBS containing 1.5% paraformaldehyde and analyzed in a flow cytometer (FACStar Plus; Becton Dickinson). To quantitate both membrane and cytoplasmic expression of viral proteins, infected Hut78 cells were fixed with methanol at -20°C for 20 min. After being washed with PBS, cells

RESULTS Kinetics of replication. To investigate the replicative behavior of the mutant SIVmac nef viruses, Hut78 cells (human T-lymphoma cell line) were infected (constant multiplicity of infection [MOI] of 0.002) with either wild-type viruses BK28wt and BK28/PCRwt, nef insertion mutants BK28nefBS and BK28neJNS, or deletion mutants BK28 AneJ3, -5, and -13 (Fig. 1; Table 1). Every 2 days, cells were split 1:2 and RT activity was determined in the cell-free supernatant. Figure 2 shows the time course of RT activity in the supernatant of the infected cells. The growth kinetics of the different nef mutants were surprising: all SIVmac nef mutants except BK28Anef3 exhibited faster replication kinetics in the initial phase of virus growth than did the wild-type, reaching a plateau at 16 days postinfection (dpi). The growth rates of BK28wt and of BK28/PCRwt (data not shown) were in good agreement. Compared with the wild type, insertion mutant BK28 nefBS, whose nef protein comprises only the 13 N-terminal amino acids (aa), exhibited an 11-fold-higher RT activity at 10 dpi and mutant BK28neJNS (nef protein of 68 aa) exhibited a 3-fold-higher RT activity (Fig. 3). In the case of the insertion mutants, the nef protein had been truncated by the introduction of a stop codon in the nef gene although they retained all nef gene-specific nucleotide sequences. The improved replicative capacity of the insertion mutants compared with the wild type is therefore likely to be due to the missing or impaired function of the truncated nef protein. Because of the high mutation rate of the immunodeficiency viruses, insertion mutants are unsuitable for in vivo studies, particularly as selection pressures may exist in vivo

1400 1-1

1300 1200

-4

1100

C en0

1000 900 F

800 700

Ez

600 500 _

400 300

200 100

-| BK28wt BK28PCRwt BK28nefBS BK28nefNS BK28wncf3 B' Virus

FIG. 3. RT activity at 10 dpi.

Mock

5240

J. VIROL.

BINNINGER ET AL.

for the presence of a functionally active gene product. Therefore, nef deletion mutants were constructed and tested in tissue culture. Deletion mutant BK28Anefl3, whose Nef has an internal deletion of 61 aa, with the 42 N-terminal aa and 144 C-terminal aa intact, grew as well as did insertion mutant BK28nefBS (Nef of 13 aa). Mutant BK28Anef5, with a 42-N-terminal-aa nef protein, showed by far the fastest growth rate. At 10 dpi in standardized Hut78 culture, BK28AneJ5 exhibited RT activities at least 4-fold higher than those of BK28Lnefl3 and 46-fold higher than those of BK28/PCRwt (Fig. 3). Surprisingly, deletion mutant BK28AneJ3, which encodes a nef protein of 49 aa, replicated more slowly than the wild type. The reduced growth rate can hardly be explained by the altered nef protein, because both a shorter nef protein (mutant Anejf5: 42 aa) and a longer nef protein (mutant NS: 68 aa) result in mutants that replicate better than the wild type. An explanation for the impaired replication can be deduced from the nucleotide sequence changes of this mutant. The deletion of BK28AneJ3 extends from positions 8699 to 8928 (Table 1). In the SIVmac BK28 genome, a polypurine tract (PPT) (20) is localized between positions 8920 and 8937 immediately adjacent to the U3 region of the 3' LTR. The PPT is highly conserved in all lentiviruses and is thought to function as a primer for plus-strand DNA synthesis in the viral replication cycle. In mutant BK28Anef3, 9 of the 16 nucleotides of the PPT are missing, suggesting that this defect is responsible for the reduced growth rate. Pullen and Champoux (21) mapped the start site for HIV-1 plus strands within the PPT (which is identical to the SIVmac251 PPT) to the sequence 5'-ACTG.... It is assumed that the RNase H activity of the RT is responsible for the endonucleolytic cleavage that creates the primer RNA. Apparently, deletion of the nine 5' nucleotides of the PPT did not completely abolish the effect of RNase H, but the efficiency was clearly reduced. The replicative capacity of three mutants (BK28wt, BK28nefBS, and BK28Ane]5) was tested on a second human T-cell line, MT4 (Fig. 4A). Growth rates for the individual mutants were comparable with those observed on Hut78. For example, deletion mutant BK28AneJ5 grew fastest on the MT4 cell line, as it did on Hut78. Simultaneously, the percentage of infected cells during the time course of infection was measured by FACS analysis (Fig. 4B), using a cross-reactive serum from an HIV-2 patient. At 4 dpi, only 1% of the cells in the wild-type virus-infected culture expressed viral antigens, whereas in the BK28nefBS- and BK28Anef5-infected cultures, the rates were 5 and 30%, respectively. Since MT4 cells are killed as the result of infection by SIV, no viable cells could be found at 8 dpi. Our results suggest strongly that inactivation of the nef gene product in SIVmac DNA clone BK28 leads to virus mutants that replicate faster than the wild type and thus support the idea that Nef acts as a negative regulator for SIVmac and HIV replication in human lymphoid cell lines. The fact that deletion mutant BK28Ane.f5 exhibited a growth rate even higher than that of insertion mutant BS suggests that this effect is not simply due to a lack of nef protein function. In contrast, the deletion may have secondary effects on the structure of virion RNA or viral mRNA. Phenotyping of mutant virus-infected host cells. In a first approach to detect possible novel effects of the nef mutants on the host cell phenotype, FACS analysis was used to examine permanently infected Hut78 cultures for the quantitative expression of CD4 receptor molecules and of SIV

antigens.

A

1i02

O

1

2

3

4

I

5

6

.

.

5

6

7

8

7

8

d.p.i. 100

B

, . BK28wt

.

2

3

.

BK28nefBS/

L50

o

1

4

d.p.i. FIG. 4. (A) Time course of RT activity in the supernatant of infected MT4 cells. MT4 cells were infected with SIVmac viruses BK28wt, BK28nefBS, and BK28Anef5 at an MOI of 0.002. Cells were split every 2 days, and cell-free supernatant was assayed for RT activity. (B) Detection of percentage of infected MT4 cells during the time course of infection measured by FACS analysis. MT4 cells were infected with SlVmac viruses BK28wt, BK28nefBS, and BK28Ane]5 at an MOI of 0.002. Cells were split every 2 days, and aliquots were fixed as described in Materials and Methods; the percentage of infected cells was determined by cross-reactivity with an HIV-2-positive serum by FACS analysis.

Investigation of the percentage of cells that express viral

proteins on their surface gave an unexpected result: whereas 18% of the cells of the wild-type-infected permanent Hut78 culture stained positively for HIV-2, the percentage of viral neoantigen-expressing cells was higher for all tested nef mutant viruses (BK28nefBS, 26%; BK28AneJ5, 37%; and BK28Anef3, 49%) (Fig. SA). The observation that even in the case of the deletion mutant Anef3, which has a reduced growth rate, 49% of the cells expressed viral protein on their surface is surprising but not unexplainable. This mutant with its incomplete PPT also integrates into the host's genome, and subsequent virus expression is obviously no longer hampered. CD4 expression is down-regulated to the same extent regardless of whether the cells are infected by the wild type or by the mutant SIVmac strains (Fig. 5B). Reduction of CD4 is characteristic and unique, as other cell surface antigens (LFA-1 adhesion molecule or class I histocompatibility antigens such as HLA-ABC) do not show a down-modulation (data not shown).

VOL. 65, 1991

SIVmac nef GENE MUTANTS

A

B

_ ,,PPt, PP,l.

BK28 nefBS

R,,,,--

F,i

C

26%

I' a

50%2 0

spcp

IID

co'

5.4%

IaI1.1 9

iIam

iIamt

.I-e

9..

A,

~~50%

_

LP99, pp9 *.

,

t

^'1R

a

BK28 Anef3

5241

a

.'I

4.4% 14 tons

I

see

14P@,, lCpe

P9@ .

BK28 Anef5

pag

.'1

'1

12

C4

&as

h

3.4%

-

lot

an

I*2

lmg

BK28 wt

1L

IU*

r

-I p.... PP9.

aI

as1

_

pry I- *rvloY I ,

S.

l4p9l.

gas

F0q

,111

IaI

....

iI2

PP9 IP 9.

P9., PP9.

1.2

sea

Log Fluorescence

as

1'4 I

asDe

83%

.

1.5%

--I

is'

1U4

PP9. , 1lege

Ias

.-

a

,SS

.

-

t@s,on

ts

too

3.8%

*. .I

*a

Mock

ad

i$

19%

1

1j

too

U4 aqus

ad

.

I

38% a

II

* as

£4

II

-.-

in,I*1

I

1

as

sea

Sam

14a

Log Fluorescence

FIG. 5. Surface phenotyping by detection of SIVmac antigens (A) and CD4 molecules (B) on the cell surface of Hut78 cells persistently infected with BK28wt, with nef deletion mutants, or with nef insertion mutants. SIVmac proteins were detected by using a human serum from an HIV-2-positive individual. CD4 receptor expression was measured by using monoclonal antibody anti-CD4 (Leu3a). An indirect immunofluorescence assay for surface antigens was performed as indicated in Materials and Methods. Percentages of positive stained cells were calculated by FACS analysis.

5242

BINNINGER ET AL.

DISCUSSION Most HIV/SIV vaccine research concentrates on the use of inactivated particles or of subunit, recombinant proteins (15). However, it is likely that a strong, broad, and longlasting immune response would be most successfully stimulated by using a live homologous or a related virus, albeit one attenuated in its pathogenicity. The immune response induced by such an attenuated virus vaccine would mimic very closely that induced by HIV itself, with B- and T-cell reactivities against all antigens. The experiments described here were designed to explore, using the SIVmac model system, this remote possibility of a live attenuated HIV vaccine, based on the hypothesis that a variant of HIV able to replicate very quickly in the initial stages of infection (and unable to undergo latency) would stimulate a strong immune response which would then eliminate the virus (and virusinfected cells). Furthermore, these studies are suited to explore the function of mutated genes such as nef. We show here that premature termination or deletion in the nef gene increases the replication rate of SIVmac, as measured by the release of virus particles (RT) and by the increased number of infected cells (FACS). This finding supports the hypothesis that nef is involved in down-regulation of SIV expression, possibly an essential function for establishment of

latency. It is not clear why there is disagreement in the literature concerning the supposed negative regulatory function of nef. Our results using nef insertion and deletion mutants clearly indicate such a role, but similar results were not found by others using HIV-1 (11, 13). However, it should be noted that in our case the faster growth kinetics could be seen only when cultures were infected with a very low MOI, and even under these conditions RT levels were only significantly different between 5 and 12 dpi, after which each virus culture reached an equivalent plateau of virus production. This finding suggests that the mutants replicate and spread through a culture faster than does the wild type but that each culture has a maximum capacity for virus production which all viruses eventually achieve. The results for the mutants showing increased virus expression as measured by RT activity were supported by the FACS analysis giving the number of infected cells in the acutely infected cultures (Fig. 5). However, although all chronically infected cultures contain similar percentages of infected cells (>90%) as demonstrated by intracellular staining for viral antigen, an increased percentage of cells infected with the SIVmac mutants expressed high levels of viral antigen on the cell membrane. It therefore appears either that the mutant-infected cells produce more antigen or that there is an accumulation of antigen at the cell surface as a result of enhanced transport. The fact that changes in the expression of cellular antigens on the cell surface (i.e., down-regulation of CD4) are the same after infection with the mutants or with the wild type indicates that other cell functions are not affected differently. We plan to infect rhesus macaques with nef deletion mutants to determine whether this virus is indeed incapable of establishing a latent state and to measure the strength of the resulting immune response, both humoral and cellular (19). However, the use of mutant BK28Anej5 in vivo has one drawback; the nef deletion also removes 38 bases from the 3' end of the env gene. Although this does not affect the virus in vitro (because of the presence of a premature stop codon upstream of the deletion), it is known that in vivo, SIVmac rapidly reverts to expression of the full length env gene

J. VIROL.

product (7), and the mutant in question would therefore lack the 12 C-terminal aa. We are therefore constructing SIVmac deletion mutants, comparable to BK28AneJ5, that do not affect the env gene. In summary, our results with SIVmac insertion and deletion mutants strongly indicate that nef indeed plays a negative regulatory role in the virus life cycle. Mutants deficient in nef replicate more rapidly and show a higher expression of surface viral antigen. Whether such mutants indeed replicate faster in vivo and, if so, whether they are cleared by the immune response before entering a stage of latency remain to be determined. ACKNOWLEDGMENTS We thank J. Mullins for the gift of pBK28 and H. Konig and M. Baier for stimulating discussions. This work was supported by a grant from the Federal Ministry of Research and Technology. REFERENCES 1. Ahmad, N., and S. Venkatesan. 1988. Nef protein of HIV-1 is a transcriptional repressor of HIV-1 LTR. Science 241:14811485. 2. Allan, J. S., J. E. Coligan, T. H. Lee, M. F. McLane, P. J. Kanki, J. E. Groopman, and M. Essex. 1985. A new HTLV-III/ LAV encoded antigen detected by antibodies from AIDS patients. Science 230:810-813. 3. Arenzana-Seisdedos, F., N. Israel, F. Bachelerie, U. Hazan, J. Alcami, F. Dautry, and J. L. Virelizier. 1989. c-Ha-ras transfection induces human immunodeficiency virus (HIV) transcription through the HIV-enhancer in human fibroblasts and astrocytes. Oncogene 4:1359-1362. 4. Bachelerie, F., J. Alcami, U. Hazan, N. Israel, B. Goud, F. Arenzana-Seisdedos, and J.-L. Virelizier. 1990. Constitutive expression of human immunodeficiency virus (HIV) nef protein in human astrocytes does not influence basal or induced HIV long terminal repeat activity. J. Virol. 64:3059-3062. 5. Binninger, D., F.-J. Ferdinand, and H. Rubsamen-Waigmann. 1989. Inhibition of SV40 DNA replication by Rous sarcoma virus LTR enhancer. Arch. Virol. 107:291-299. 6. Cheng-Mayer, C., P. lannello, K. Shaw, P. A. Luciw, and J. A. Levy. 1989. Differential effects of nef on HIV replication: implications for viral pathogenesis in the host. Science 246: 1629-1632. 7. Cranage, M. P., M. Almond, A. Jenkins, and P. A. Kitchin. 1989. Transmembrane protein of SIV. Nature (London) 342: 349. 8. Guy, B., R. B. Acres, M. P. Kieny, and J.-P. Lecocq. 1990. DNA binding factors that bind to the negative regulatory element of the human immunodeficiency virus 1: regulation by nef. J. Acquired Immune Defic. Syndr. 3:797-809. 9. Guy, B., M. P. Kieny, Y. Riviere, C. LePeuch, K. Dott, M. Girard, L. Montagnier, and J.-P. Lecocq. 1987. HIV F/3'-orf encodes a phosphorylated GTP-binding protein resembling an oncogene product. Nature (London) 330:266-269. 10. Hammes, S. R., E. P. Dixon, H. Malim, B. R. Cullen, and W. R. Greene. 1989. Nef protein of human immunodeficiency virus type 1: evidence against its role as a transcriptional inhibitor. Proc. Natl. Acad. Sci. USA 86:9549-9553. 11. Kaminchik, J., N. Bashan, D. Pinchasi, B. Amit, N. Sarver, M. I. Johnston, M. Fischer, Z. Yavin, M. Gorecki, and A. Panet. 1990. Expression and biochemical characterization of human immunodeficiency virus type 1 nef gene product. J. Virol. 64:34473454. 12. Kim, S., K. Ikeuchi, R. Byrn, J. Groopman, and D. Baltimore. 1989. Lack of a negative influence on viral growth by the nef gene of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 86:9544-9548. 13. Konig, H., E. Behr, J. Lower, and R. Kurth. 1989. Azidothymidine triphosphate is an inhibitor of both human immunodeficiency virus type 1 reverse transcriptase and DNA polymerase

VOL. 65, 1991 gamma. J. Antimicrob. Agents Chemother. 33:2109. 14. Kornfeld, H., N. Riedel, G. Viglianti, V. Hirsch, and J. I. Mullins. 1987. Cloning of HTLV-4 and its relation to simian and human immunodeficiency viruses. Nature (London) 326:610613. 15. Kurth, R., G. Kraus, A. Werner, S. Hartung, P. Centner, M. Baier, S. Norley, and J. Lower. 1988. AIDS: animal retrovirus models and vaccines. J. AIDS 1:284-294. 16. Luciw, P. A., C. Cheng-Mayer, and J. A. Levy. 1987. Mutational analysis of the human immunodeficiency virus: the orf-B region down-regulates virus replication. Proc. Natl. Acad. Sci. USA 84:1434-1438. 17. Myers, G., A. B. Rabson, J. A. Berzofsky, T. F. Smith, and F. Wong-Staal. 1990. Human retroviruses and AIDS. Los Alamos National Laboratory, Los Alamos, N.M. 18. Niederman, T. M. J., B. J. Thielan, and L. Ratner. 1989. Human immunodeficiency virus type 1 negative factor is a transcriptional silencer. Proc. Natl. Acad. Sci. USA 86:1128-1132. 19. Norley, S. G., G. Kraus, J. Ennen, J. Bonilla, H. Konig, and R. Kurth. 1990. Immunological studies into the basis for the apathogenicity of simian immunodeficiency virus from African

SIVmac nef GENE MUTANTS

5243

green monkeys. Proc. Natl. Acad. Sci. USA 87:9067-9071. 20. Panganiban, T., and D. Fiore. 1988. Ordered interstrand and intrastrand DNA transfer during reverse transcription. Science 241:1064-1069. 21. Pullen, K. A., and J. J. Champoux. Plus-strand origin for human immunodeficiency virus type 1: implications for integration. J. Virol. 64:6274-6277. 22. Saiki, R. K., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Higuchi, G. T. Horn, K. B. Mullis, and H. A. Erlich. 1988. Primerdirected enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-491. 23. Samuel, K. P., A. Seth, A. Konopka, J. A. Lautenberger, and T. S. Papas. 1987. The 3'-orf protein of human immunodeficiency virus shows structural homology with the phosphorylation domain of human interleukin-2 receptor and the ATPbinding site of the protein kinase family. FEBS Lett. 218:81-86. 24. Terwilliger, E., J. S. Sodroski, C. A. Rosen, and W. A. Haseltine. 1986. Effects of mutations within the 3' open reading frame region of human T-cell lymphotropic virus type III (HTLV III/LAV) on replication and cytopathogenicity. J. Virol. 60:754760.