Macaques Infected with Cloned Simian Immunodeficiency Virus Show RecurringnefGene Alterations

VIROLOGY

249, 260±274 (1998) VY989325

ARTICLE NO.

Macaques Infected with Cloned Simian Immunodeficiency Virus Show Recurring nef Gene Alterations G. Heidecker,*,1 H. MunÄoz,* P. Lloyd,* D. Hodge,² F. W. Ruscetti,² W. R. Morton,§ S.-L. Hu,§ and R. E. Benveniste³ *Intramural Research Support Program, SAIC Frederick, National Cancer Institute±Frederick Cancer Research and Development Center, Frederick, Maryland 21702; ²Laboratory of Leukocyte Biology and ³Laboratory of Genome Diversity, National Cancer Institute, Frederick, Maryland 21702; and §Regional Primate Research Center, University of Washington, Seattle, Washington 98195 Received February 11, 1998; returned to author for revision April 10, 1998; accepted July 7, 1998 We compared nef gene sequences isolated by polymerase chain reaction from peripheral blood lymphocyte DNA of macaques that had been inoculated with either biologically (E11S) or molecularly (clone 8) cloned SIV/Mne. Two samples from each animal obtained either early (weeks 2±8) or late (weeks 21±137) after infection were analyzed. Three substitutions in the predicted Nef amino acid sequence were seen in all animals at the late time point, and two other substitutions were seen in all except one. Two of the common exchanges are located ;40 residues apart in the Nef core sequence but are juxtaposed on the tertiary structure as judged by computer modeling using the structure of the HIV Nef core protein as a guide. Most recurrent in vivo changes replaced a residue found in the cloned Nef sequence with one present in a consensus derived by aligning the Nef sequences of the SIV/Sm clade. Recombinant virus containing a macaque-adapted (MA nef) nef on the clone 8 backbone was 3-fold more infectious on SMAGI cells than the original virus. A lymphocyte line infected with SIV-clone 8-MAnef contained a large proportion of cells carrying provirus with defective nef genes. These findings suggest that the nef gene of the cloned SIV/Mne had undergone attenuating mutations during propagation in tissue culture that were ªcorrectedº in vivo. © 1998 Academic Press

viral replication through downregulation of transcription from the LTR (Ahmad and Venkatesan, 1988; Luciw et al., 1987; Luo and Garcia, 1997). In addition, nef was reported to interfere with transcription of other nuclear factor-kBs, as well as activated protein (AP)-1-regulated promoters and with interleukin-2 production (Reinhart et al., 1996; Niederman et al., 1993, 1992; Tokunaga, 1996). Other reports demonstrated that there was no consistent negative effect on viral replication but that the viral yield was critically dependent on the cell line and, in particular, virus strain used (de Ronde et al., 1992; Kim et al., 1989; Hammes et al., 1989; Terwilliger et al., 1991). Recently, evidence has been presented that nef affects the infectivity of viral particles because viruses produced in the absence of nef show defects in the synthesis of proviral DNA or in incorporation of env-SU into the particle (Wiskerchen and Cheng-Mayer, 1996; Pandori et al., 1996; Chowers et al., 1995). At the cellular level, Nef has been shown to downregulate CD4 by enhancing its turnover rate (Guy et al., 1987; Garcia and Miller, 1991; Mariani and Skowronski, 1993). Recent studies have shown an involvement of Nef and CD4 with the adapter protein AP-1 in clathrin-coated pits, further elucidating the role of Nef in CD4 downregulation (Greenberg et al., 1997; Foti et al., 1997). Interference in the signal transduction system was suggested by its effect on transcriptional activities. Nef has been reported to affect and even to interact directly with a plethora of

INTRODUCTION The nef gene is an accessory gene unique to primate lentiviruses. It overlaps the 39 long terminal repeat (LTR) and in some virus species also the 39-end of env (Harris, 1996; Michael et al., 1995). It is transcribed early in infection from multiply spliced RNAs (Kaminchik et al., 1990), and depending on the virus strain, nef encodes a myristoylated protein of 25±32 kDa that is located on the cytoplasmic side of the cell membrane (Guy et al., 1987; Franchini et al., 1986). The role of the nef gene in the virus life cycle is still enigmatic. Early reports showed that a significant proportion of virus isolates contained inactivating mutations in nef but were found to replicate as efficiently as isolates with apparently intact nef genes, suggesting that the gene was not needed for replication in tissue culture cells (Blumberg et al., 1992; Fisher et al., 1986). Indeed, initial findings suggested that it had a negative effect on

The content of this publication does not necessarily reflect the views of policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. By acceptance of this article, the publisher or recipient acknowledges the right of the U.S. Government to retain a nonexclusive royaltyfree license in and to any copyright covering the article. 1 To whom reprint requests should be addressed. Fax: (301) 8467034. E-mail: [email protected]. 0042-6822/98 $25.00 Copyright © 1998 by Academic Press All rights of reproduction in any form reserved.

260

RECURRING nef GENE ALTERATIONS

components of the signal transduction network of various cell types; again, some results from different laboratories contradict each other (review in Luo et al., 1997)). Most studies of Nef function have been done with Nef proteins derived from virus isolates propagated in tissue culture cell lines, and in several instances, strain differences could be shown to account for some of the inconsistencies in the results (Garcia and Foster, 1996). Propagation of virus in tissue culture might lead to attenuated nef genes as a consequence of genetic drift of a sequence that is no longer under the selective pressure that would be present in vivo. In addition, there is evidence that Nef has a negative effect on cell growth, which would lead to selection of attenuating mutations: cells that are forced to express high levels of a marker linked to nef often acquire inactivating mutations in nef (Baur et al., 1994; Trono, 1995; Su et al., 1997). The importance of nef for viral fitness and pathogenicity in vivo has been irrefutably demonstrated: infection of macaques with a molecularly cloned SIV virus with a deletion in the nef gene resulted in very low levels of viremia and no disease for the time interval tested after infection (Kestler et al., 1991). Similarly, deletions of nef were found in some long-term survivors of HIV infection (Mariani et al., 1996; Deacon et al., 1995). Inhibitory point mutations, such as premature stop codons, acquired in tissue culture, are rapidly reverted in vitro, underlining the different selective pressures exerted on the nef gene in vivo and in vitro (Kestler et al., 1991). In parallel, the very rapid time course from infection to time of death of a different strain of SIV (SIV/sm/Pbj14) could, at least in part, be attributed to one amino acid substitution in Nef (Du et al., 1996). Incorporation of this exchange into the SIV/ Mac239 clone resulted in a virus that caused fatal disease in 7±10 days rather than the normally observed 6±9 months. This latter strain also caused activation of resting mononuclear cells and transformed NIH 3T3 cells. The molecular clone of SIV/Mne used in this report has no obvious inactivating mutations in the nef gene and was obtained from a single-cell clone of infected Hut 78 cells that produced high levels of infectious virus. We hypothesized that the virus had acquired attenuating missense mutations in nef, which should be reverted in vivo, thus highlighting residues that are important for Nef function. To test this hypothesis, we analyzed nef sequences obtained from macaques that had been inoculated with biologically (single-cell clone) or molecularly cloned SIV/Mne (Benveniste et al., 1989, 1988). We found that five Nef residues had undergone consistent substitutions in almost all sequences isolated late in infection from the animals studied. In addition, other residues were found to have changed in more than one animal. The macaque-adapted (MA) nef conferred increased in-

261

fectivity to the virus, but mutated versions rapidly accumulated in a tissue culture environment. RESULTS Biological cloning of SIV/Mne SIV/Mne was originally isolated on HuT 78 cells from a lymph node of a pigtailed macaque (Macaca nimestrina) that had died from lymphoma in 1982 at the University of Washington Primate Research Center (Benveniste et al., 1986). SIV/Mne clone E11S is a single-cell clone of SIV/Mne-infected HuT 78 cells that was obtained by plating dilutions of infected HuT 78 cells onto a monolayer of sheep choriod plexus cells in microtiter plates (Benveniste et al., 1990). Every HuT 78 cell in clone E11S is identical and contains two SIV/Mne proviruses integrated at two sites (as shown in this report). While SIV/Mne E11S causes disease more slowly than uncloned virus, it contains large amounts of gp120 envelope glycoprotein and gag proteins in the viral particle and has been used for the characterization of SIV viral proteins (Henderson et al., 1988a, 1988b), as well as for in vivo infectivity studies (Benveniste et al., 1994; Kuller et al.. 1994) and in various AIDS vaccine studies (Shafferman et al., 1991; Hu et al., 1992; Arthur et al., 1995). Molecular clones of SIV/Mne Because all E11S cells are identical and produce infectious and pathogenic virus, this cell line was chosen as the starting material for molecular cloning. DNA fragments of ;15±20 kb generated by partial digest of highmolecular-weight genomic DNA with MboI were cloned into lambda vector EMBL3 (Frischauf et al., 1983). A library of 106 plaque forming units was screened using the internal 7-kb fragment of SIV/Mac251 clone as a probe (Hirsch et al., 1986). Nine clones were isolated and mapped by restriction analysis, which revealed that they derived from two different proviruses (Fig. 1). This was verified by Southern blotting (data not shown). Only clone 8 contained a full-length copy of one provirus. DNA sequencing determined that all reading frames of clone 8 were open with the exception of env, which carried the typical premature stop codon at position 8277 seen in the transmembrane part of the env genes of tissue culture isolates of SIV (Kestler et al., 1989). The sequence for SIV/Mne clone 8 was deposited as M32741 in GenBank. Transfection of clone 8 DNA into HuT 78 cells resulted in the production of infectious virus (data not shown). This SIV/Mne clone 8 virus preparation was used to infect two pigtailed macaques (F87047 and M87044) intravenously. One animal (F87047) first showed a decrease in CD41 lymphocytes at 10 months after infection and was euthanized at 19 months because of bleeding caused by severe thrombocytopenia (Table 1). The other macaque (M87004) first showed CD4 de-

262

HEIDECKER ET AL.

FIG. 1. Restriction maps of molecular clones 3, 8, and 9. Clones 3 and 9 contain sequences from one of the two proviruses present in cell clone E11S, whereas clone 8 represents the other provirus. Restriction sites defining the difference between the integration sites of the proviruses are circled. Shaded boxes represent the left and right arms of the EMBL3 cloning vector. (B) BglII. (Bw) BsiWI. (Bb) BsmBI. (E) EcoNI. (H) HindIII. (S) SacI. (S1) SalI. (Sc) ScaI. (X) XbaI.

pletion at 1 year and was euthanized at 52 months with Campylobacter lari septicemia (Table 1). The second provirus present in E11S was reconstituted from clones 3 and 9 (Fig. 1). Transfection of the resulting construct into 293 cells yielded virus that was infectious on CEM and AA2CL5 cells (data not shown). The sequence of this provirus was determined around the nef gene between residues 8140 and 9310 and found to be identical to that of clone 8. Comparison of Nef sequences from animals inoculated with cloned SIV/Mne virus We compared the nef sequence of the SIV/Mne clone 8 with the sequences obtained directly from peripheral blood mononuclear cells (PBMCs) of macaques infected intravenously or intrarectally with clone E11S or clone 8 virus. Two species of macaques (M. fascicularis and M. nemestrina) were used in this study. The dose of virus, route of infection, time until death after infection, and principal pathological findings at necropsy are shown in Table 1. Of the nine macaques examined, seven had been infected directly with either molecular clone 8- or clone E11S-derived virus, whereas animal 90076 received PBMCs and lymph node cells from another macaque (89152) that had been infected with E11S virus 43 weeks previously. Macaque 92170 was a control animal in a vaccine study and had been infected with uncloned virus. All macaques showed CD4 depletion with an average time to onset of 50 weeks for the animals that had

received clone 8 or clone E11S virus and for which continuous CD4 data were available. The average time to death for these animals was 123 weeks (2.4 years). In contrast, macaques 90076 and 92170 died after 39 weeks and 65 weeks, respectively, and had developed metastatic lymphosarcoma. To evaluate the sequence divergence accumulated in the nef gene during the course of infection, we isolated genomic DNA from PBMCs obtained shortly after inoculation (2±8 weeks p.i.) and at later times (21±137 weeks p.i.). All except one of the later time points (92170) were obtained from animals that were already CD4 depleted. nef gene sequences were amplified by nested polymerase chain reaction (PCR), starting with 500 ng of DNA isolated from PBMCs. To determine the frequency of misincorporation during nested PCR, we sequenced five clones obtained from independent PCRs using the same primer sets on DNA isolated from E11S and found one divergence from the clone 8 sequence (data not shown), indicating an error rate of 1:3500 nucleotides. Sensitivity tests using DNA isolated from mixtures of E11S cells with uninfected HuT 78 cells showed that the detection limit was 2±5 proviruses in 104 cells. Pairwise comparison of 58 nef open reading frames (792 nucleotides) from the animal isolates to the original clone 8 nef revealed between 1 (0.1%) and 29 (3.3%) nucleotide substitutions (Fig. 2). Seven sequences had premature termination codons, one sequence had a three-nucleotide insertion (90184wk137 sequence 1), one

RECURRING nef GENE ALTERATIONS

263

TABLE 1 Macaques Used for the Isolation of nef Clones Animal #

Macaque species

Date of challenge

Date of death (weeks p.i.)

Weeks of sampling

F87047

M. nemestrina

Clone 8, 103 TCID, IV

11/5/87

6/15/89 (wk 83)

81

M87004

M. nemestrina

Clone 8, 103 TCID, IV

11/5/87

3/3/92 (wk 225)

135

89152

M. fascicularis

Clone E11S, 10 TCID (;10 AID), IV

8/16/90

8/20/92 (wk 104)

2, 65

90076

M. fascicularis

6/14/91

3/11/92 (wk 39)

4, 21

Malignant lymphoma

F90184

M. nemestrina

7/17/91

M. fascicularis

3/24/94 (wk137) 11/28/94 (wk 120)

8, 137

89144

PBMC 1 lymph node cells from 89152 (wk 43 p.i.)c Clone E11S, 104 TCID (100 AID) IRd Clone E11S, 20 TCID (20 AID), IV

CD4 depletion (wk 30) anemia, wasting CD4 depletion (wk 61) wasting

90125

M. fascicularis

Clone E11S, 20 TCID (20 AID), IV

8/3/92

9/12/94 (wk 110)

8, 84

CD4 depletion (wk 55), leg paralysis, anemia

91419

M. fascicularis

Clone E11S, 20 TCID (20 AID), IV

8/3/92

2/17/94 (wk 80)

4, 71

CD4 depletion (wk 61) bacteremia

92170

M. fascicularis

SIV/Mne uncloned, 100 TCID (10 AID), IV

6/8/93

9/9/94 (wk 65)

4, 31

CD4 depletion (wk 34), malignant lymphoma

Inoculum, dose, routea

8/3/92

2, 85

Pathology (onset), cause of deathb CD4 depletion (wk 40), severe thrombocytopenia CD4 depletion (wk 52), anemia, weight loss, Campylobacter septicemia CD4 depletion, anemia, wasting

Study

Reference

Pathogenicity of molecular clone Pathogenicity of molecular clone

Rudensey et al. (1995)

Control animal in vaccine study Recipient of cells from 89152 Intrarectal titration Control in vaccine study Control in vaccine study Control in vaccine study Control in vaccine study

Hu et al. (1989, 1993)

Rudensey et al. (1995)

Hu et al. (1989, 1993) Clerici et al. (1994) Hu et al. (1989) Hu et al. (1989) Hu et al. (1989)

a Animals were infected with either molecular clone 8 or biological clone E11S of SIV/Mne as shown. The tissue culture infectious dose (TCID) was determined with 10-fold serial dilutions on AA-2 cells; the animal infectious dose (AID) is listed when known. IV, intravenous, IR, intrarectal inoculation. b The most remarkable clinical findings are listed as is the onset of CD4 depletion. c 2 3 107 inguinal lymph node cells and 2 3 107 PBMC obtained from macaque 89152 at 43 weeks post inoculation (p.i.) were inoculated IV into this animal. d The minimal infectious dose required for intrarectal infection is ;103 times greater than the minimum infectious dose after an IV challenge.

had a three-nucleotide deletion (90125wk84 sequence 2), and one had a 13-nucleotide deletion (90125wk8 sequence 4). A total of 549 nucleotide differences were observed; $351 of these had occurred independently in different animals. Ninety-four percent (518) of the nucleotide exchanges resulted in amino acid substitutions (Fig. 2). Within the 167-nucleotide-long region of overlap between env and nef, 162 nucleotide substitutions were found, giving rise to 57 amino acid changes in Env (data not shown) and 102 in Nef. Only 2 of the changes in Env were common to two animals, but they also resulted in sequence alteration in Nef. In contrast, 5 changes in this region of nef were found in at least two independent isolates and resulted in amino acid substitutions in Nef but were silent in Env. One of these common changes is the A±G transition at nucleotide position 101, which gave rise to the glutamic acid±to±glycine exchange at position 34 (E34G) and occurred in late isolates from all animals. The prevalence of recurring changes that resulted in differences at the protein level in Nef but not in Env suggests that Nef, rather than Env, function was under selective pressure. Nucleotide sequences from early time point samples differed by 0.1±1.3% from clone 8 nef, whereas late

isolates had diverged by 0.8±3.3%. At the amino acid level, the differences were 0±4% at early times and 2±8% at late time points (Figs. 2 and 3). Although it is possible that some of these differences are caused by PCR artifacts, the number and overall pattern of the divergences argue that the majority are due to real divergence in the amplified virus population. Most of the differences observed early were evenly distributed throughout the gene and transient (i.e., they either were observed in only one clone or were not maintained in the virus population of a given animal to the later time point) (Figs. 2 and 3). The exception was the I111M exchange, which was present in both cases in which samples had been taken 8 weeks p.i. In contrast to the random distribution of mutations in early samples, late isolates contained several amino acid substitutions that were consistently observed in most samples: E34G, I87M, and I111M were present in at least one clone derived from the late sampling of each animal, and M153I/V and N95D were seen in all except one (89144) or two (89144, F87047) animals, respectively (Figs. 2 and 3). No early time point samples were available for the animals that had been inoculated with the molecularly cloned virus, but the

264

HEIDECKER ET AL.

FIG. 2. Comparison of the predicted amino acid sequences for the nef ORFs isolated from animals after infection with cloned virus. The animal number and week of sampling are given. The amino acid sequence of SIV/Mne clone 8 is shown in single-letter code. Residues that diverge from the SIV/Sm consensus are shown above the sequence. In the alignment, only amino acids that differ are shown, whereas identities are indicated (dots). (Exclamation marks) Synonymous changes in a codon. (Dashes) Gaps. The conserved boxes described for HIV/SIV Nef sequences are also indicated (Shugars et al., 1993). Also indicated is the location of the myristoylation signal. (PPT) Primer binding site is located in the coding sequence for the Nef protein. (env and LTR) Boundaries for the env gene and the LTR, respectively.

RECURRING nef GENE ALTERATIONS

265

FIG. 2ÐContinued

late time samples show the same pattern of mutations as the others. In addition to the five most common substitutions, several other changes were observed in

more than one animal: T5I, G75R, P101S, F122L, M144I/V, S212T, and R246K (Fig. 2). Clones from two animals showed different patterns

266

HEIDECKER ET AL.

FIG. 2ÐContinued

of divergence; early isolates from animal 90076, which had been infected with PBMCs and lymph node cells from macaque 89152 drawn at week 43, already showed the mutations typical for late isolates and

differed by 1.5±1.9% from the clone 8 nucleotide sequence and by 4% in the amino acid sequence. In addition, all of the early isolates from animal 92170, which had been infected with uncloned virus, had the

RECURRING nef GENE ALTERATIONS

267

FIG. 3. The frequency of substitutions early and late after infection. The number of substitutions at a given position is indicated by the height of the bar relative to the total number of sequences analyzed at each time point given by the y axis of the graph. All substitutions are scored regardless of whether they occurred independently. The changes that occurred repeatedly are given in single-letter format.

hallmark features of the MA nef gene rather than those of clone 8 Nef. Sequence analysis of the uncloned population revealed that most viruses contain a nef gene similar to the MA nef. We analyzed 15 nef sequences of proviral genomes present in the HuT 78/ SIV/Mne culture used to generate the virus stock inoculated into this animal. Ten of the genes were inactivated by nonsense or frameshift mutations, and one was inactivated by a missense mutation that resulted in loss of the myristoylation site. The majority (14 of 15) encoded the same residues as the MA nef at positions 34, 87, 95, and 111. Only residue 153 was generally a methionine as in clone 8 (data not shown). Biological activities of the MA nef gene To evaluate whether the mutations in nef present in MA E11S virus have any biological effect, we incorporated the nef gene of the third clone of the week 4 sample of animal 90076 into the clone 8 DNA, thus generating clone 8-MA nef (see Materials and Methods). This clone carried all the hallmark substitutions of an MA

gene with only three additional changes. The DNA was transfected into 293 cells, and supernatants were harvested after 48 h and used to infect AA2CL5 cells to generate virus stocks. The stocks were titered using an antigen capture kit, and equal amounts were used to infect SMAGI cells. Two independently generated sets of virus stocks yielded the same result (Table 2), showing that the virus containing the MA nef gene from 90076 was 3-fold more infectious than clone 8 virus. In addition, we generated a clone 8 derivative with a deletion in nef similar to that described by Kestler et al. (1991). This virus, clone 8 Dnef, was 3- and 6-fold less infectious than clone 8 and clone 8-MA nef viruses, respectively. We observed no difference in the ability of clone 8 Nef and MA Nef to downregulate CD4 from the cell surface. CEM cells infected with retroviral vector constructs expressing either clone 8-Nef or MA Nef presented similar levels of CD4 on their cell surface, which were ;5% of the level observed in cells infected with retroviral vector carrying the nef gene in antisense orientation (data not shown).

268

HEIDECKER ET AL. TABLE 2 Infectivity of Clone 8 and Clone 8-MA nef on SMAGI Cells

b-Gal.-positive cells/fielda Virus strain

Experiment 1

Experiment 2

Clone 8 Clone 8 MA nef Clone 8 Dnef

5 15 5

44 110 20

Note. Virus stocks used in the experiment were titered by antigen capture assay. In the first experiment, virus equivalent to 9000 pg of p26 was used; in the second, to 200,000 pg. The background in both experiments was approximately two stained cells per field. a Average of three fields were counted.

The residues in clone 8 that had consistently been replaced in animals could have accumulated in vitro as a consequence of genetic drift in a neutral sequence, or they could be due to negative selective pressure on the nef gene. To address this question, we compared the evolution of the MA nef gene, of the tissue cultureadapted clone 8 nef gene, and of the deleted nef gene in culture. AA2CL5 cells were infected with supernatants of 293 cells that had been transfected with the respective constructs. The viruses grew with comparable kinetics as determined by core antigen capture assay and reverse transcriptase activity measurements (data not shown). The first time point was taken after 3 weeks, by which time the cells were maximally infected, and further samples were analyzed at 2±3-week intervals. Proviral nef sequences were amplified in a single round, cloned, and sequenced. As shown in Figure 4A, MA nef genes showed a higher divergence from the input sequence, and a higher proportion of the changes were nonsynonymous and resulted in stop codons, indicating selective pressure on the gene. At the first time point, one of the two clones obtained from the clone 8-MA nef-infected culture contained a premature stop codon in nef. This mutant persisted in the culture and was found again in a clone from week 7. Similarly, a nonsense mutation found first at week 5 was again isolated in the week 9 sample. At week 5, one in four genes were inactivated; at week 7, two in four; and at week 9, three in four. Overall, 39 nucleotide exchanges were observed in 14 sequences, which resulted in 31 differences at the amino acid level, including seven premature stop codons. In contrast, 14 proviral nef sequences isolated from an AA2CL5 culture infected with clone 8 in the same manner showed only 13 nucleotide changes that resulted in seven nonsynonymous substitutions, including one premature stop codon (Fig. 4B). Even fewer changes were observed in a culture of AA2CL5 cells infected with clone 8 Dnef: seven nucleotide substitutions resulting in four amino acid differences were observed in 11 sequences isolated over the same time course (Fig. 4C). The ratio of nonsynonymous

to synonymous exchanges and the frequency of stop codons seen in the clone 8 Dnef- and nef-derived sequences are those expected for randomly drifting sequences, whereas the pattern of changes found in MA nef indicates selective pressure. DISCUSSION We compared DNA and predicted amino acid sequences for the nef genes obtained from seven animals that had been infected with virus that had been cloned biologically (SIVMne C1E11S) or molecularly (SIVMne Cl8). We found random and transient changes up to 4 weeks p.i., whereas by week 8, the isoleucine at position 111 in clone 8-Nef was consistently changed to methionine. This substitution was maintained in all late isolates in addition to two other substitutions that arose consistently: E34G and I87M. The substitution M153I/V was seen in all except one, and N95D was seen in all except two animals. Several other changes were present in at least two animals. The nef gene sequences obtained early from an animal that had been infected with uncloned virus showed only the residues at these positions acquired late in infection in animals inoculated with cloned virus. A comparison of the clone 8 sequence to a consensus sequence derived by aligning several SIVsm/HIV-2 Nef sequences reveals that clone 8 diverges from the consensus at 39 positions (see Fig. 2). Thirteen of the substitutions repeatedly found in the late isolates reconstituted the SIV/Mne Nef to the consensus sequence. Only four substitutions occurred in more than one animal, which involved residues that did not diverge from the consensus sequence. This finding suggests that the clone 8 sequence had accumulated attenuating mutations during in vitro passages that were reversed in vivo. Three of the five common substitutions would be considered conservative and involve changes from methionine to isoleucine, or vice versa, and are consequences of transitions. Nevertheless, they involve residues that are highly conserved among SIV/Sm and derivative strains; the methionine at position 111, which is the earliest consistent ªreversion,º aligns with a consensus methionine in Nef proteins of all primate immunodeficiency viruses (Zhu et al., 1993). Three (N95D, I111M, and M153I) of the five most common substitutions occurred in the Nef highly conserved core sequence. The HIV Nef core structure was recently determined by NMR analysis (Grzesiek et al., 1996, 1997) and, bound to the SH3 domain of Src family kinase Fyn, by x-ray crystallography (Lee et al., 1996). The residue corresponding to 95D lies outside the area that could be crystallized. Residues 111 and 153 correspond to residues 79M and 121F of HIV-1 Nef, which are juxtaposed in the tertiary structure. They define one edge of the face of the molecule that interacts with Fyn, and the methionine

RECURRING nef GENE ALTERATIONS

269

FIG. 4. Alignments of Nef sequences derived from AA2CL5 cells infected with clone 8 (A), clone 8-MA nef (B), and clone 8-nef (C). The predicted amino acid sequences in each group are compared with the input sequences. Only divergences are shown. (Dots) Identities.

is considered essential for maintaining the structural integrity of this region (Lee et al., 1996) Modeling of the SIV/Mne Nef sequence via the SWISS Model routine according to the coordinates of the NMR structure reveals that residues 111 and 153 are opposite each other in the three-dimensional structure. However, this program did not identify any disruption in the structure when the residues at these positions did not match the consensus. Although no prior study has evaluated the fate of the nef gene of a cloned provirus in this many animals, Zhu

et al. (1996) investigated Nef changes in different tissues of three animals that had been infected with SIV/mac239. They showed that in addition to reversion of the stop codon that occurred in all animals, several residues showed consistent substitutions in two of three animals. Interestingly, none of these occurred at positions at which the mac239 Nef diverges from a consensus Nef. A possible explanation for this different pattern is that the mac239 nef was completely inactivated early during propagation in tissue culture by nonsense mutation and did not need to acquire further attenuating mutations.

270

HEIDECKER ET AL.

FIG. 4ÐContinued

The results of our experiment in which we analyzed mutations that accumulate in nef genes in a transformed cell line support this notion. After growth in culture, cells infected with an SIV construct with the MA nef gene harbored a high proportion of proviruses (50%) with overtly inactivated nef genes. In contrast, proviruses with clone 8 nef or with a nef gene inactivated by deletion showed very few changes compared with the input virus sequence. Although none of the changes observed in the MA nef gene exactly recapitulated those seen in the original clone 8 sequence, the glutamic acid±to±lysine exchange at position 92 was

seen in two isolates. This change results in the disruption of a conserved acidic motif similar to the asparagine residue in clone 8 at position 95 that is consistently replaced by aspartic acid in vivo. Because mutations occur randomly, it is not likely that the clone 8 gene configuration would be found again considering the many positions one would expect to contribute to functionality, either directly by interaction with other cell components or by supplying the structural framework of the protein. Substitutions at any of these positions singly or in combination would attenuate Nef function. These findings show that selection in the

RECURRING nef GENE ALTERATIONS

271

FIG. 4ÐContinued

tissue culture environment favors cells carrying provirus with a nef gene that is less functional in vivo. This study has highlighted five residues in Nef that are important to its function. Although the MA nef has about the same activity for CD4 downregulation, it conferred 3-fold higher infectivity to the clone 8 virus. It is known that animals inoculated with uncloned virus have a shorter life expectancy than those infected with cloned virus (R. Benveniste, et al., manuscript in preparation). It also is noteworthy that the two animals in this study that were infected with virus without the attenuating mutations in nef developed aggressive lymphosarcoma. The original SIV/Mne isolate was obtained from an animal with lymphosarcoma. Although several lines of data suggest that Nef is a major contributor to pathogenicity, it is not likely that the increased virulence of the in vivo passaged virus can be entirely attributed to the changes in nef. Several other genes have been shown to undergo consistent mutations in vivo, such as the reversion of the nonsense codon in TM. Analysis of the env sequences from the animals that had been infected with clone 8 virus showed several consistent substitutions (Rudensey et al., 1995; Overbaugh et al., 1991). The effects of the changes on other Nef functions that can be assayed in

vitro, such as protein interactions and effects on signal transduction, as well as the contribution of the tissue culture attenuated and in vivo reverted Nef to the pathogenicity of SIV/Mne, are currently under investigation. MATERIAL AND METHODS Molecular cloning, restriction, and sequence analysis of SIV/Mne proviral sequences High-molecular-weight SIV/Mne E11S DNA (500 mg) was partially digested with restriction endonuclease MboI. Fragments of 12±20 kb were isolated on 10±40% sucrose gradients in 100 mM NaCl, 10 mM Tris, pH 8, 1 mM EDTA and ligated to lEMBL3 arms generated with BamHI (Stratagene, La Jolla, CA). After in vitro packaging into lambda particles according to the supplier's (Stratagene) protocol, 106 plaque forming units (pfu) were plated on E. coli K802 (Sambrook et al., 1989) and screened with radiolabeled SIV/mac DNA (108 cpm/mg). Positive plaques were purified to homogeneity and lambda DNA was purified using lambdasorb (Promega, Madison, WI). Restriction digests and Southern blotting were done following standard protocols (Sambrook et al.,

272

HEIDECKER ET AL.

1989). DNA sequencing was done using shotgun (Messing et al., 1981) or primer directed sequencing. Generation of recombinant virus stocks To obtain virus from clone 8, lambda DNA was transfected directly into HuT 78 cells using DEAE dextran. To reconstitute provirus 3/9, the SalI--SacI fragments from lambda clones 3 and 9 representing the 59 and 39 halves of the provirus were subcloned into pZero2 (InVitrogen, Carlsbad, California). DNAs from both plasmids were cut with SacI and ligated before transfection into 293 cells. To obtain clone 8-MA nef virus, two plasmids of clone 8 were generated: one (59-end plasmid) contained a 10-kb fragment extending from a BsiWI site in the lambda vector to the XhoI site in the nef gene (Fig. 1). The other (39-end plasmid) contained a fragment extending from a ScaI site just 59 of the nef gene to an XbaI site in the genomic DNA flanking the 39 side of the provirus. BglII and EcoNI sites unique to the 39-end plasmid were used to exchange the nef gene sequences. To recombine the 59- and 39-end plasmids, both were cut at a unique BsmBI site that is just 59 of the BglII site and is present on both plasmids and were ligated. Similar approaches were used to generate clone 8 Dnef and a control stock of clone 8 for this experiment. Supernatants of 293 cells were used to infect lymphocyte lines to obtain virus stocks for further experiments. Animals used in the study Macaques used in this study were housed at the Washington Regional Primate Research Center, which is accredited by the American Association for Accreditation of Laboratory Animal Care. Animals were anesthetized with ketamine before all inoculations and blood draws. Table 1 lists the macaques used in this study by species, dates and identity of virus inoculation, time of sampling, onset of CD4 depletion, and principal clinical findings at the time of death. PCR amplification and nucleotide sequencing of nef sequences Total cellular DNA was extracted from 1±5 3 106 PBLs according to standard protocol (Higuchi, 1989). Nested PCR was performed starting with 500 ng of DNA. The first primer pair (59-GAGGTACACTACTGGTGGCACCTCA-39 and 59CTGTTCTCCCTTTGATCGAC-39) amplified the region between residues 7616 and 9372 of SIV/Mne. Two different primer pairs were used for the second reaction. The first set, 59-CTTGACTTGGTTATTCAGCAACTGCAG-39 and 59GGAATTCCTGTTCTCCCTTTGATCGAC-39, was used on samples from animals 89152, 90076, and 91419 and amplified residues 8420±9372, adding an EcoRI site to the 39-end of the fragment. PCR products were cloned in Bluescript KS (Stratagene, La Jolla, California) after EcoRI±PstI digest. The second pair (59-CACGACCCTA-

CAGAGAATCCGAGAAG-39 and 59-ATTTTTACCGACTGTTCTCCCTTTGA-39) was used for samples from animals 89144, 90125, 90184, 92170, 92176, 87004, and 87047 and amplified fragments between nucleotides 8495 and 9336, which were cloned via TA cloning into pCRII (InVitrogen). The nef genes from cell culture lines were amplified in a single PCR using 59-CTTGACTTGGTTATTCAGCAACTGCAG-39 and 59-CCCATGACAGCCCTTATGGAAAGTC-39 primers. All reactions were carried out in 100 ml of 20 mM Tris±HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 200 mM concentration of each dNTP (Life Technologies, Gaithersburg, Maryland), 0.5 mM concentration of each primer, and 1.25 units of Taq polymerase (Life Technologies). Reactions were overlaid with 100 ml of mineral oil (Life Technologies) and amplified using a thermal cycler (Perkin-Elmer, Foster City, California). The program consisted of 30 s at 95°C; 40 cycles of 1 min at 95°C, 1 min at 55°C, 3 min at 72°C; and 7 min at 72°C. Fragments were cloned as indicated above. We obtained at least two independent clones from all samples. Sequencing was performed using T7 and M13 primers, as well as an internal primer that hybridized to residues 9090±9110. Sequences were assembled and analyzed with the Lasergene software package from DNASTAR (Madison, Wisconsin). Alignments were done using the CLUSTAL algorithm (Higgins and Sharp, 1989). Infectivity assay SMAGI cells were cultured and stained for b-galactosidase expression as previously described (Chackerian et al., 1995). The concentration of virus in stocks was determined by antigen capture assay using Coulter SIV Core Antigen assay (Coulter Corp., Miami, Florida) or as published (Tsai et al., 1993). ACKNOWLEDGMENTS We thank Dr. J. Overbaugh for DNA samples of animals F87047 and M87004; R. Hill, W. Knott, L. Kuller, and G.-K. Pei for expert assistance; and Drs. D. Derse and T. Fanning for critical reading of the manuscript. This work was supported in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. NO1-CO-56000 and in part by U.S. Public Health Sservice NIH Grant RR-00166 to the University of Washington. SMAGI cells were obtained from J. Overbaugh through the AIDS Research and Reference Reagent Program (Division of AIDS, NIAID, NIH).

REFERENCES Ahmad, N. and Venkatesan, S. (1988). Nef protein of HIV-1 is a transcriptional repressor of HIV-1 LTR. Science 241, 1481±1485. Arthur, L. O., Bess, J. W., Jr., Urban, R. G., Strominger, J. L., Morton, W. R., Mann, D. L., Henderson, L. E., and Benveniste, R. E. (1995). Macaques immunized with HLA-DR are protected from challenge with Simian immunodeficiency virus. J. Virol. 69, 3117±3124. Baur, A. S., Sawai, E. T., Dazin, P., Fantl, W. J., Cheng-Mayer, C., and Peterlin, B. M. (1994). HIV-1 Nef leads to inhibition or activation of T cells depending on its intracellular localization. Immunity 1, 373±384. Benveniste, R. E., Arthur, L. O., Tsai, C. C., Sowder, R., Copeland, T. D.,

RECURRING nef GENE ALTERATIONS Henderson, L. E., and Oroszlan, S. (1986). Isolation of a lentivirus from a macaque with lymphoma: comparison with HTLV-III/LAV and other lentiviruses. J. Virol. 60, 483±490. Benveniste, R. E., Morton, W. R., Clark, E. A., Tsai, C. C., Ochs, H. D., Ward, J. M., Kuller, L., Knott, W. B., Hill, R. W., and Gale, M. J. (1988). Inoculation of baboons and macaques with Simian immunodeficiency virus/Mne, a primate lentivirus closely related to human immunodeficiency virus type 2. J. Virol. 62, 2091±2101. Benveniste, R. E., Raben, D., Hill, R. W., Knott, W. B., Drummond, J. E., Arthur, L. O., Jahrling, P. B., Morton, W. R., Henderson, L. E., and Heidecker, G. (1989). Molecular characterization and comparison of Simian immunodeficiency virus isolates from macaques, mangabeys, and African green monkeys. J. Med. Primatol. 18, 287±303. Benveniste, R. E., Hill, R. W., Eron, L. J., Csaikl, U. M., Knott, W. B., Henderson, L. E., Sowder, R. C., Nagashima, K., and Gonda, M. A. (1990). Characterization of clones of HIV-1 infected HuT 78 cells defective in gag gene processing and of SIV clones producing large amounts of envelope glycoprotein. J. Med. Primatol. 19, 351±366. Benveniste, R. E., Roodman, S. T., Hill, R. W., Knott, W. B., Ribas, J. L., Lewis, M. G., and Eddy, G. A. (1994). Infectivity of titered doses of Simian immunodeficiency virus clone E11S inoculated intravenously into rhesus macaques (Macaca mulatta). J. Med. Primatol. 23, 83±88. Blumberg, B., Epstein, L., de Ronde, A., and Goudsmit, J. (1992). HIV1 39 ORF is open in pathological tissue and accelerates viral replication in primary lymphocytes. Res. Virol. 143, 63±65. Chackerian, B., Haigwood, N. L., and Overbaugh, J. (1995). Characterization of a CD4-expressing macaque cell line that can detect virus after a single replication cycle and can be infected by diverse Simian immunodeficiency virus isolates. Virology 213, 386±394. Chowers, M. Y., Pandori, M. W., Spina, C. A., Richman, D. D., and Guatelli, J. C. (1995). The growth advantage conferred by HIV-1 nef is determined at the level of viral DNA formation and is independent of CD4 downregulation. Virology 212, 451±457. Clerici, M., Clark, E. A., Polacino, P., Axberg, I., Kuller, L., Casey, N. I., Morton, W. R., Shearer, G. M., and Benveniste, R. E. (1994). T-cell proliferation to subinfectious SIV correlates with lack of infection after challenge of macaques. AIDS 8, 1391±1395. de Ronde, A., Klaver, B., Keulen, W., Smit, L., and Goudsmit, J. (1992). Natural HIV-1 NEF accelerates virus replication in primary human lymphocytes. Virology 188, 391±395. Deacon, N. J., Tsykin, A., Solomon, A., Smith, K., Ludford-Menting, M., Hooker, D. J., McPhee, D. A., Greenway, A. L., Ellett, A., and Chatfield, C. (1995). Genomic structure of an attenuated quasi species of HIV-1 from a blood transfusion donor and recipients. Science 270, 988±991. Du, Z., Ilyinskii, P. O., Sasseville, V. G., Newstein, M., Lackner, A. A., and Desrosiers, R. C. (1996). Requirements for lymphocyte activation by unusual strains of Simian immunodeficiency virus. J. Virol. 70, 4157± 4161. Fisher, A. G., Ratner, L., Mitsuya, H., Marselle, L. M., Harper, M. E., Broder, S., Gallo, R. C., and Wong-Staal, F. (1986). Infectious mutants of HTLV-III with changes in the 39 region and markedly reduced cytopathic effects. Science 233, 655±659. Foti, M., Mangasarian, A., Piguet, V., Lew, D. P., Krause, K. H., Trono, D., and Carpentier, J. L. (1997). Nef-medciated clathrin-coated pit formation. J. Cell Biol. 139, 37±47. Franchini, G., Robert-Guroff, M., Ghrayeh, J., Chang, N. T., and WongStall, F. (1986). Cytoplasmic localization of the HTLV-III 39 orf protein in cultured T-cells. Virology 155, 593±599. Frischauf, A. M., Lehrach, H., Poustka, A., and Murray, N. (1983). Lambda replacement vectors carrying polylinker sequences. J. Mol. Biol. 170, 827±842. Garcia, J. V., and Miller, A. D. (1991). Serine phosphorylation-independent downregulation of cell-surface CD4 by nef. Nature 350, 508±511. Garcia, J. V., and Foster, J. L. (1996). Structural and functional correlates between HIV-1 and SIV Nef isolates. Virology 226, 161±166. Greenberg, M. E., Bronson, S., Lock, M., Neumann, M., Pavlakis, G. N., and Skowronski, J. (1997). Co-localization of HIV-1 Nef with the AP-2

273

adaptor protein complex correlates with Nef-induced CD4 downregulation. EMBO J. 16, 6964±6976. Grzesiek, S., Bax, A., Clore, G. M., Gronenborn, A. M., Hu, J. S., Kaufman, J., Palmer, I., Stahl, S. J., and Wingfield, P. T. (1996). The solution structure of HIV-1 Nef reveals an unexpected fold and permits delineation of the binding surface for the SH3 domain of Hck tyrosine protein kinase. Nat. Struct. Biol. 3, 340±345. Grzesiek, S., Bax, A., Hu, J. S., Kaufman, J., Palmer, I., Stahl, S. J., Tjandra, N., and Wingfield, P. T. (1997). Refined solution structure and backbone dynamics of HIV-1 Nef. Protein Sci. 6, 1248±1263. Guy, B., Kieny, M. P., Riviere, Y., Le Peuch, C., Dott, K., Girard, M., Montagnier, L., and Lecocq, J. P. (1987). HIV F/39 orf encodes a phosphorylated GTP-binding protein resembling an oncogene product. Nature 330, 266±269. Hammes, S. R., Dixon, E. P., Malim, M. H., Cullen, B. R., and Greene, W. C. (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. Harris, M. (1996). From negative factor to a critical role in virus pathogenesis: the changing fortunes of Nef. J. Gen. Virol. 77, 2379±2392. Henderson, L. E., Benveniste, R. E., Sowder, R., Copeland, T. D., Schultz, A. M., and Oroszlan, S. (1988a). Molecular characterization of gag proteins from Simian immunodeficiency virus (SIVMne). J. Virol. 62, 2587±2595. Henderson, L. E., Sowder, R. C., Copeland, T. D., Benveniste, R. E., and Oroszlan, S. (1988b). Isolation and characterization of a novel protein (X-ORF product) from SIV and HIV-2. Science 241, 199±201. Higgins, D. G., and Sharp, P. M. (1989). Fast and sensitive multiple sequence alignments on a microcompouter. Comput. Appl. Biosci. 5, 151±153. Higuchi, R. (1989). Simple and rapid preparation of samples for PCR. In ªPCR Technologyº (H. A. Erlich, Ed.), pp. 31±38. Stockton Press, New York. Hirsch, V., Riedel, N., Kornfeld, H., Kanki, P. J., Essex, M., and Mullins, J. I. (1986). Cross-reactivity to human T-lymphotropic virus type III/ lymphadenopathy-associated virus and molecular cloning of Simian T-cell lymphotropic virus type III from African green monkeys. Proc. Natl. Acad. Sci. USA 83, 9754±9758. Hu, S. L., Abrams, K., Barber, G. N., Moran, P., Zarling, J. M., Langlois, A. J., Kuller, L., Morton, W. R., and Benveniste, R. E. (1992). Protection of macaques against SIV infection by subunit vaccines of SIV envelope glycoprotein gp160. Science 255, 456±459. Kaminchik, J., Bashan, N., Pinchasi, D., Amit, B., Sarver, N., Johnston, M. I., Fischer, M., Yavin, Z., Gorecki, M., and Panet, A. (1990). Expression and biochemical characterization of human immunodeficiency virus type 1 nef gene product. J. Virol. 64, 3447±3454. Kestler, H. W., Naidu, Y. N., Kodama, T., King, N. W., Daniel, M. D., Li, Y., and Desrosiers, R. C. (1989). Use of infectious molecular clones of Simian immunodeficiency virus for pathogenesis studies. J. Med. Primatol. 18, 305±309. Kestler, H. W., Ringler, D. J., Mori, K., Panicali, D. L., Sehgal, P. K., Daniel, M. D., and Desrosiers, R. C. (1991). Importance of the nef gene for maintenance of high virus loads and for development of AIDS. Cell 65, 651±662. Kim, S., Ikeuchi, K., Byrn, R., Groopman, J., and Baltimore, D. (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. Kuller, L., Benveniste, R. E., Tsai, C. C., Clark, E. A., Polacino, P., Watanabe, R., Overbaugh, J., Katze, M. G., and Morton, W. R. (1994). Intrarectal inoculation of macaques by the Simian immunodeficiency virus, SIVmne E11S: CD41 depletion and AIDS. J. Med. Primatol. 23, 397±409. Lee, C. H., Saksela, K., Mirza, U. A., Chait, B. T., and Kuriyan, J. (1996). Crystal structure of the conserved core of HIV-1 Nef complexes with a Src family SH3 domain. Cell 85, 931±942. Luciw, P. A., Chang-Mayer, C., and Levy, J. A. (1987). Mutational analysis

274

HEIDECKER ET AL.

of the human immune deficiency virus: the ORF-B region down regulates virus replication. Proc. Natl. Acad. Sci. USA 84, 1434±1438. Luo, T., Foster, J. L., and Garcia, J. V. (1997). Molecular determinants of Nef function. J. Biomed. Sci. 4, 132±138. Luo, T., and Garcia, J. V. (1997). Regulation of Nef function. Res. Virol. 148, 64±68. Mariani, R., and Skowronski, J. (1993). CD4 down-regulation by nef alleles isolated from human immunodeficiency virus type 1-infected individuals. Proc. Natl. Acad. Sci. USA 90, 5549±5553. Mariani, R., Kirchhoff, F., Greenough, T. C., Sullivan, J. L., Desrosiers, R. C., and Skowronski, J. (1996). High frequency of defective nef alleles in a long-term survivor with nonprogressive human immunodeficiency virus type 1 infection. J. Virol. 70, 7752±7764. Messing, J., Crea, R., and Seeburg, P. H. (1981). A system for shotgun DNA sequencing. Nucleic Acids Res. 9, 309±321. Michael, N. L., Chang, G., d'Arcy, L. A., Ehrenberg, P. K., Mariani, R., Busch, M. P., Birx, D. L., and Schwartz, D. H. (1995). Defective accessory genes in a human immunodeficiency virus type 1-infected long-term survivor lacking recoverable virus. J. Virol. 69, 4228±4236. Niederman, T. M., Garcia, J. V., Hastings, W. R., Luria, S., and Ratner, L. (1992). Human immunodeficiency virus type 1 Nef protein inhibits NF-kappa B induction in human T cells. J. Virol. 66, 6213±6219. Niederman, T. M., Hastings, W. R., Luria, S., Bandres, J. C., and Ratner, L. (1993). HIV-1 Nef protein inhibits the recruitment of AP-1 DNAbinding activity in human T-cells. Virology 194, 338±344. Overbaugh, J., Rudensey, L. M., Papenhausen, M. D., Benveniste, R. E., and Morton, W. R. (1991). Variation in Simian immunodeficiency virus env is confined to V1 and V4 during progression to Simian AIDS. J. Virol. 65, 7025±7031 Pandori, M. W., Fitch, N. J., Craig, H. M., Richman, D. D., Spina, C. A., and Guatelli, J. C. (1996). Prodicer-cell modification of human immunodeficiency virus type 1: Nef is a virion protein. J. Virol. 70, 4283±4290. Reinhart, T. A., Rogan, M. J., and Haase, A. T. (1996). RNA splice site utilization by Simian immunodeficiency viruses derived from sooty mangabey monkeys. Virology 224, 338±344. Rudensey, L. M., Kimata, J. T., Benveniste, R. E., and Overbaugh, J. (1995). Progression to AIDS in macaques is associated with changes in the replication, tropism, a and cytopathic properties of the Simian immunodeficiency virus variant population. Virology 207, 528±542. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ªMolecular Cloning,º

2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. Shafferman, A., Jahrling, P. B., Benveniste, R. E., Lewis, M. G., Phipps, T. J., Eden-McCutchan, F., Sadoff, J., Eddy, G. A., and Burke, D. S. (1991). Protection of macaques with a Simian immunodeficiency virus envelope peptide vaccine based on conserved human immunodeficiency virus type 1 sequences. Proc. Natl. Acad. Sci. USA 88, 7126±7130. Shugars, D. C., Smith, M. S., Glueck, D. H., Nantermet, P. V., SeillierMoiseiwitsch, F., and Swanstrom, R. (1993). Analysis of human immunodeficiency virus type 1 nef gene sequences present in vivo. J. Virol. 67, 4639±4650. Su, L., Kaneshima, H., Bonyhadi, M. L., Lee, R., Auten, J., Wolf, A., Du, B., Rabin, L., Hahn, B. H., Terwilliger, E., and Mccune, J. M. (1997). Identification of HIV-1 determinants for replication in vivo. Virology 227, 45±52. Terwilliger, E. F., Langhoff, E., Gabuzda, D., Zazopoulos, E., and Haseltine, W. A. (1991). Allelic variation in the effects of the nef gene on replication of human immunodeficiency virus type 1. Proc. Natl. Acad. Sci. USA 88, 10971±10975 Tokunaga, K. (1996). Function of human immunodeficiency virus type 1 nef gene product in virus replication. Hokkaido. Igaku. Zasshi. 71, 345±357. Trono, D. (1995). HIV accessory proteins: leading roles for the supporting cast. Cell 82, 189±192. Tsai, C. C., Follis, K. E., Grant, R. F., Nolte, R. E., Wu, H., and Benveniste, R. E. (1993). Infectivity and pathogenesis of titered dosages of Simian immunodeficiency virus experimentally inoculated into longtailed macaques (Macaca fascicularis). Lab. Anim. Sci. 43, 411±416. Wiskerchen, M., and Cheng-Mayer, C. (1996). HIV-1 Nef association with cellular serine kinase correlates with enhanced virion infectivity and efficient proviral DNA synthesis. Virology 224, 292±301. Zhu, G. W., Mukherjee, S,. Sahni, M., Narayan, O., and Stephens, E. B. (1996). Prolonged infection in rhesus macaques with Simian immunodeficiency virus (SIVmac239) results in animal-specific and rarely tissue-specific selection of nef variantes. Virology 220, 522±529. Zhu, T., Mo, H., Wang, N., Nam, D. S., Cao, Y., Koup, R. A., and Ho, D. D. (1993). Genotypic and phenotypic characterization of HIV-1 patients with primary infection. Science 261, 1179±1181.