Proc. Natl. Acad. Sci. USA Vol. 88, pp. 3195-3199, April 1991 Microbiology Effect of mutations affecting the p6 gag on human immunodeficiency particle release ( assembly/budding/cell surface attachment/interferon a) HEINRICH G. GOTTLINGER, TATYANA DORFMAN, JOSEPH G. SODROSKI, AND WILLIAM A. HASELTINE Laboratory of Human Retrovirology, Dana-Farber Cancer Institute, Department of Pathology, Harvard Medical School, Boston, MA 02115 Communicated by Bernard Fields, January 2, 1991

ABSTRACT Mutations in sequences at the C terminus of MATERIALS AND METHODS the capsid precursor protein ofhuman immunodeficiency virus type 1 that affect the viral p6 protein prevent release ofbudded Construction of Mutant Proviruses. Single-stranded DNA virus particles from the cell surface. The experiments reported was prepared from plasmid pGEMgag-pol (5), which contains here define an important step in the life cycle of the virus, the a 4.3-kilobase (kb) Sph I-Sal I gag-pol fragment from the release of the budded particle from a tether that binds the infectious proviral clone HXBc2 (15), and was used as assembled particle to the cell surface. Inhibition of the release template for the annealing of oligonucleotides and primer of the viral capsid by interferon a indicates that this extension (16). A 473-base-pair (bp) Apa I-PpuMI fragment step of virus maturation may be sensitive to inhibition by carrying the correct mutation as confirmed by DNA sequence antiviral drugs. analysis was reinserted into the parental vector pHXB-SV, which harbors the full-length HXBc2 provirus and provides a simian virus 40 (SV40) origin of replication (17). The The capsid proteins of human immunodeficiency virus type primers used for the construction of mutant clones L1/s, 1 (HIV-1), like those of other , are initially made Li/c, S17/s, and S17/c are given in Fig. 1. The following in the form of a polyprotein precursor (1, 2). The precursor oligonucleotides were used for the construction ofthe mutant molecules assemble at the cell membrane in a process that clones Y36/s (5'-GGAACTGTAAGCTTTAACTTC-3'), eventually leads to the release of an immature virion (3). No P11/L (5'-CAGCCCCACTCGAAGAGAGC-3'), and other virally encoded proteins are required for the formation P10,11/L (5'-CAACAGCCCTACTCGAAGAGAGC-3'). ofthe HIV-1 capsid structure (4). Capsid formation ofHIV-1 Cell Culture and Transfection. Jurkat cells (human T-cell and of several other retroviruses requires the attachment of line) were maintained in RPMI-1640 medium supplemented a myristic acid residue to the N terminus of the capsid with 10%o fetal bovine serum. COS-7 cells (SV40-transformed precursor (5, 6). monkey cell line) were grown in Dulbecco's modified Eagle's Upon their release from the infected cell the HIV-1 virions medium with 10%6 fetal bovine serum. Jurkat cells (5 x 106) undergo a morphological maturation associated with the were transfected by the DEAE-dextran procedure (18) with cleavage of the capsid precursor polyprotein into smaller 5 pg of CsCl-banded plasmid DNA. COS-7 cells (4 x 105) polypeptides by a virally encoded (7). The smaller were seeded into 50-ml tissue culture flasks 24 hr before polypeptides comprise the capsid ofthe mature virus particle. transfection. The cells were transfected with 10 pg ofproviral In the absence of a functional protease, the virus particles DNA by a calcium phosphate precipitation technique (19). form and are released but are not infectious (5, 8, 9). Recombinant human interferon a was purchased from Jan- In the mature virus particles the individual components of ssen Biochimica (Beerse, Belgium). Equivalent results were the capsid precursor segregate to different compartments of obtained with a2-interferon kindly provided by Jerome the virion. The p17 matrix protein, which is derived from the Schwartz (Schering-Plough). N terminus ofthe precursor molecule and retains the myristic Metabolic Labeling and Immunoprecipitation. COS-7 cell acid moiety (10), forms an envelope-associated outer capsid cultures and aliquots of transfected Jurkat cells were meta- (11). The major core protein, p24, forms the shell of the bolically labeled for 12 hr with [35S]cysteine (60 ,Ci/ml; 1 characteristic cone-shaped core structure in the mature vir- ,pCi = 37 kBq). Labeling ofCOS-7 cells was started 48 hr after ion (11). The nucleocapsid protein p9 has an essential role in transfection. Equivalent numbers of labeled cells were lysed the packaging of genomic viral RNA into the viral core (12). in RIPA buffer (140 mM NaCl/8 mM Na2HPO4/2 mM A proline-rich peptide of =6 kDa has been found in HIV-1 NaH2PO4/1% Nonidet P-40/0.5% sodium deoxycholate/ virions (13, 14). The peptide, designated p6, is derived from 0.05% SDS) and virus-encoded proteins were immunopre- the C terminus ofthe capsid precursor. The entirety ofthe p6 cipitated from the cell lysates with HIV-1-infected patient coding sequence overlaps thepol open reading frame (Fig. 1). serum and separated in 10% polyacrylamide gels as described In contrast to the matrix, core, and nucleocapsid proteins, a (20). The labeling medium was collected and cleared of cell peptide equivalent to p6 is not generally present in retrovi- debris by centrifugation. Virus was lysed by adding Sx RIPA ruses. Consequently, studies of other retroviruses do not buffer and viral proteins were immunoprecipitated from provide insight into the possible functions of p6 in the viral equivalent volumes. life cycle. The Assay. Particle-associated reverse mutational analysis ofthe p6 coding sequences transcriptase activity in culture supernatants was determined reported here demonstrates that amino acid sequences pres- as described (20). ent in p6 are required for a critical late step in virus particle Electron Microscopy. Processing oftransfected COS-7 cells maturation. for thin-section and scanning electron microscopy were done as described (5). The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviations: HIV, human immunodeficiency virus; SV40, simian in accordance with 18 U.S.C. §1734 solely to indicate this fact. virus 40.

3195 Downloaded by guest on September 24, 2021 3196 Microbiology: Gottlinger et al. Proc. Natl. Acad. Sci. USA 88 (1991)

A substitution mutants were made that replaced one (P11/L) or v Li S17 Y36 both (P10,11/L) of the two adjacent proline residues within the 5-amino acid stretch Pro-Thr-Ala-Pro-Pro, a sequence HC | | P6 |gag-frame that is conserved between the p6 sequences of HIV-1 and T1PR gpol-frame HIV-2. Effect of p6 Mutations on HIV-1 Replication. Parallel cul- B tures of Jurkat cells were transfected with the wild-type Li/s: 5'-CA GGG MT m taa CAG AGC AGA C-3' HXBc2 provirus or equivalent amounts of the mutated pro- gag-frame------G N F * . To measure virus replication, expres- pol-frame .-.--R E F L E Q sion was analyzed by metabolic labeling with [35S]cysteine and immunoprecipitation of the cell lysates and supernatant Li/c: 5'-CA GGG MT TTc tTa CAG AGC AGA C-3' fractions with serum from an HIV-1-infected patient. Syn- gag-frame------G N F L Q S R cytium formation in the transfected cultures was also mon- pol-frame .--- R E F T E Q itored and the release of reverse transcriptase activity into S1 7/s: 5'-G AGC TTC AGG Tag GGG GTA GAG -3' the culture supernatant was determined. Mutations present in gag-frame. S F R * the proviruses Li/c and S17/c, which affect the pol reading pol-frame ------E L Q V G G R frame only, had no significant effect on viral protein synthesis (Fig. 2A, lanes 4 and 6) or the other parameters monitored S1 7/c: 5'-G AGC TTC AGG TCg GGG GTA GAG -3' (data not shown). The Y36/s and S17/s mutants, which were gag-frame.-.--.S F RS G V E expected to synthesize a capsid precursor with a truncated pol-frame ...... E L Q V [G G R p6, showed a delay in peak syncytium formation; the delay was more pronounced for the S17/s mutant, which has a FIG. 1. p6 mutations. (A) View ofthe gag-pol overlap region. NC more extensive deletion (Table 1). Viral protein expression and p6 are the C-terminal components ofthe capsid precursor, which is encoded by the gag reading frame. Expression of the products was significantly reduced for both mutants 3 days posttrans- encoded by the pol frame requires a -1 ribosomal frameshift during fection (Fig. 2A, lanes 5 and 7), but reached almost wild-type translation of the full-length viral RNA at the site marked by a levels at day 8 posttransfection (data not shown). By con- triangle. The portion ofthe gag-pol overlap that is sufficient to direct trast, the Li/s mutant, which lacks p6 entirely, was defective frameshifting with wild-type efficiency is marked by a solid bar. The for syncytium formation (Table 1) and synthesis of virus locations of mutations introduced into the p6 coding sequence are protein (Fig. 2A, lane 3, and data not shown). Reverse indicated by arrows. NC, nucleocapsid protein; PR, protease. (B) transcriptase activity was not detected in the cell supernatant The oligonucleotides shown were used for the construction of pairs for the mutant during the 18-day observation period of isogenic mutant proviruses that differed only by the presence or Li/s absence of a termination codon in the gag reading frame. Altered (data not shown). nucleotides are indicated by lowercase letters and the resulting amino The two mutant proviruses that carried substitutions of acid sequence (single-letter code) in both the gag and the pol frame adjacent proline residues in p6 differed from one another as is shown. Amino acids differing from the parental sequence are follows. The replacement of proline-11 with leucine had no boldfaced and boxed. Premature termination signals are marked by noticeable effect on the efficiency of syncytium formation asterisks. upon transfection ofJurkat cells (Table 1), nor did it diminish viral protein synthesis (Fig. 2B, lane 2) or export of virus RESULTS protein into the medium (Fig. 2B, lane 5). The replacement of Construction of Mutant Proviruses. Changes in the coding both proline-10 and -11 with leucine residues resulted in a sequence of the virus were introduced by site-directed mu- prolonged appearance of syncytia in the cultures that were tagenesis into a fragment of an infectious proviral DNA CELLS SUP derived from the IIIB strain of HIV-1. The corresponding A 1 2 3 4 5 6 7 B 1 2 3 4 5 6 DNA fragment of the infectious provirus was then replaced '- "- by an altered fragment. To minimize the likelihood of sec- 20Ol~tlwgpl *19193,R$ gpl260-O*tgplO-'"W ondary mutations, only a 473-nucleotide-long region of the provirus was derived from the DNA on which mutagenesis p55 0 p 5 was performed. p55 To test whether all or part of the p6 sequence is required for virus replication, terminating codons were introduced to p24 * - *_0 replace the codon specifying the N-terminal leucine of p6 p17 * * *I (Ll/s) as well as the codons specifying serine-17 (S17/s) and p24 _U0 -m tyrosine-36 (Y36/s). These mutations were predicted to cre- ate truncated capsid precursors that lack p6 entirely or partially. The premature termination codons are located 3' to D1 7 4 the sequences directing ribosomal frameshifting (21), which allows the gag-pol precursor to be synthesized (Fig. 1). Only FIG. 2. (A) Viral protein expression in CD4+ Jurkat cells after 16 nucleotides at the 5' end of the gag-pol overlap are transfection with wild-type and mutant proviruses. Equivalent num- required for this frameshift, and the region encoding p6 has bers of cells from each culture were labeled overnight on day 3 after been shown to be dispensable for frameshifting (21). transfection, and the viral proteins were immunoprecipitated from In case of the Li/s and S17/s mutants, it was not possible the cell lysates. The transfected were salmon sperm (lane 1), to create termination codons without also changing the HXB-SV (lane 2), Li/s (lane 3), Li/c (lane 4), S17/s (lane 5), S17/c overlapping pol reading frame. To determine whether the (lane 6), and Y36/s (lane 7). (B) Level ofcell-associated vs. released viral capsid protein. Jurkat cells were transfected with the parental alterations in the pol frame themselves affect virus replica- provirus HXB-SV (lanes 1 and 4) or the proline substitution mutants tion, the mutants Li/c and 517/c were made to introduce the P11/L (lanes 2 and 5) and P10,11/L (lanes 3 and 6). Aliquots from same amino acid substitutions into the pol frame as do the each culture were labeled overnight on day 6 after transfection and Li/s and S17/s mutations without affecting the amino acid the viral proteins were immunoprecipitated from the cell lysates sequence of the capsid precursor (Fig. 1). Additionally, two (lanes 1-3) or culture supernatants (lanes 4-6). Downloaded by guest on September 24, 2021 Microbiology: G6ttlinger et al. Proc. Natl. Acad. Sci. USA 88 (1991) 3197 Table 1. Syncytium formation in Jurkat cells A B C D Syncytium formation on days 1-10 CELLS SUP CELLS SUP Provirus 1 2 3 4 5 6 7 8 9 10 12345 6 1 2 3 456 123 1 2 3 HXB-SV - - + ++ +++ ++ + - - - : 0,.-o" 21 Ll/s ------

S17/s - -- + + ++ +++ +++ + - p66- a a ji4 S s .... . Y36/s -- + ++ ++ +++ ++ ++ - - p55- q.

P11/L - - + ++ +++ ++ + -- - p41 -

P10,11/L - - + ++ +++ +++ ++ + - - l, Jurkat cells were transfected on day 0 with the parental proviral plasmid HXB-SV; the p6 truncation mutants L1/s, S17/s, and p24- _ - a Y36/s; or the p6 proline substitution mutants P11/L and P10,11/L. Syncytium formation in the cultures was scored as + [1-5 syncytia per high-power field (hpf)], + + (6-10 syncytia per hpf), or + + + p1 7- A _ (>10 syncytia per hpf). transfected with the mutant (Table 1), indicating a slightly FIG. 3. Viral protein synthesis and release in CD4- COS-7 cells. diminished ability to replicate. Analysis of protein synthesis COS-7 cells were transfected with the wild-type and mutant proviral at day 6 posttransfection showed that in the cell-associated DNAs and the cultures were metabolically labeled from 48 to 60 hr fraction, the mature capsid proteins p24 and p17 produced by posttransfection. Labeled viral proteins were precipitated from the the mutant reached levels similar to those observed cell lysates (A and C) or from the cell-free culture supernatants (B P10,11/L and D). The transfected proviruses were as follows: in A and B, after transfection of wild-type or P11/L mutant DNA (Fig. HXB-SV (lanes 1), Y36/s (lanes 2), S17/s (lanes 3), S17/c (lanes 4), 2B, lane 3). However, the amount ofp24 and p17 precipitable Li/s (lanes 5), and Ll/c (lanes 6); in C and D, HXB-SV (lanes 1), from the supernatant of Jurkat cells transfected with the P11/L (lanes 2), and P10,11/L (lanes 3). ps5, p41, p24, and p17 are P10,11/L mutant was reduced when compared to that of the gag-related products; p66 is encoded by the pol frame. parental virus (Fig. 2B, lane 6). In agreement with these data, peak reverse transcriptase activity for the P10,11/L mutant (Fig. 3B, lanes 2, 3, and 5). This effect was most pronounced reached only 20-50% of the level obtained after transfection for the p6-negative mutant Li/s, which was almost entirely with the parental provirus in repeated experiments (data not defective for virus particle release (Fig. 3B, lane 5), despite shown). the presence of ample amounts of viral protein in the cyto- Effect of p6 Mutations on Capsid Protein Release. The plasm. In repeated experiments, the amount ofmature capsid discrepancy between the amount of cell-associated capsid protein precipitable from the medium ofcells transfected with protein and the amount of fully processed capsid protein the Li/s mutant was <10% of the amount detectable in the released by the P10,11/L mutant prompted an investigation supernatant of cells transfected with the wild-type virus, as of the possibility that viruses defective in p6 are also defec- measured by immunoprecipitation of viral proteins from the tive for virus particle release. To analyze the mutant provi- cell-free supernatant. The reverse transcriptase activity de- ruses independently of their relative ability to replicate, the tectable in the medium of COS-7 cells transfected with the proviral DNAs were introduced into COS-7 cells. Although Li/s mutant was <5% of that obtained for wild-type DNA the CD4-negative COS-7 cells cannot be infected by HIV-1, (data not shown). Few particles were released from cells the plasmid DNA, which contains an SV40 origin, is ampli- transfected with the mutant P10,11/L, in which two adjacent fied within the transfected cells as a consequence of SV40 proline residues were replaced by leucines (Fig. 3D, lane 3). T-antigen expression. Amplification of the proviral DNA p25 accumulated in the cells transfected with the P10,11/L yields high levels of expression of HIV-1 viral proteins and mutant almost to the exclusion of p24 (Fig. 3C, lane 3). In production of virus particles from the transfected cells. contrast, the P11/L mutant behaved similarly to the wild- The viral proteins produced in COS-7 cells after transfec- type provirus (Fig. 3 C and D, lanes 2). tion of the proviruses Li/c and S17/c were indistinguishable Effect on Virus Budding. The marked reduction in virion from those obtained upon transfection with the parental release displayed by the Li/s and P10,11/L mutants might provirus DNA (Fig. 3 A and B, lanes 1, 4, and 6). A shortened have been due to an inability ofthe mutated capsid precursor capsid precursor was present in the lysates of COS-7 cells molecules to aggregate at the membrane, as was shown to be transfected with the three mutant proviruses that contain stop the case for mutants defective for myristoylation ofthe capsid codons in p6 (Fig. 3A, lanes 2, 3, and 5). Proteolytic proc- precursor (5), or to a defect at a later stage of virus budding. essing of the truncated precursor molecules into the mature To determine whether particles were assembled at the cell capsid proteins was slightly retarded as judged from a mod- surface by the mutants defective for release of viral proteins, erate increase in the prominence of a p4l processing inter- COS-7 cells transfected with wild-type and mutant proviruses mediate that is known to contain p17 and p24/25 sequences were examined by transmission and scanning electron mi- (5). A more pronounced processing defect was apparent in croscopy. The studies of cells transfected with the Li/s or the altered ratio between the p25 and p24 forms of the major P10,11/L mutants revealed a remarkable, and novel, viral core protein. The amounts of the p24 and p25 forms of this phenotype. Numerous particles were evident on the outer protein were approximately equal in cells transfected with the surface of the cell membrane (Fig. 4 A and E). Very few, if S17/s and Y36/s mutants (Fig. 3A, lanes 2 and 3). The p25 any, of the particles appeared to leave the cell surface. protein was present almost to the exclusion of p24 in cells Inspection of multiple fields of such cells revealed that many transfected with the Li/s mutant (Fig. 3A, lane 5). ofthe virus particles were attached to the cell via a thin tether Immunoprecipitation of the viral proteins from cell-free (Fig. 4C). The particles produced by the Li/s mutant were supernatants collected after transfection with the mutant of immature appearance, with a thick electron-dense outer DNAs showed that the p25 form ofthe major core protein was shell and an electron-lucent center (Fig. 4B). Only a few not efficiently released into the medium. The altered intra- particles were found to have electron-dense material in the cellular p25/p24 ratios evident in the mutants containing stop core. None had the cone-shaped cores typical of mature codons in p6 were accompanied by a corresponding reduction wild-type virions (5). The thick electron-dense outer shell in the amount of capsid proteins released into the medium appeared to be thinner at the point of attachment to the cell. Downloaded by guest on September 24, 2021 3198 Microbiology: G6ttlinger et al. Proc. Natl. Acad. Sci. USA 88 (1991)

CELLS MEDIUM A 12 3 4 5 6 7 B 1 2 3 4 5 6 7 gpl 60 _~~~~~-m______- gpl 20 Igpl20 Wf

ai.. tSs p55 .. 46 ID~~~~~~~~ P41

p24 - p24

p17 ~__ ~_,.. plp77 _0 _

FIG. 5. Comparison of the effects of interferon a treatment and deletion of p6 in the absence or presence of vpu on virus protein release. COS-7 cells were transfected with the parental HXB-SV proviral construct (lanes 1-4), the p6-deletion mutant Li/s (lanes 5), the vpu+ HXBH1O provirus (lanes 6), or the vpuI p6-deletion mutant HXBH1O-L1/s. Fifteen hours after transfection, interferon a at 102 units/ml (lanes 2), 103 units/ml (lanes 3), or 104 units/ml (lane 4) was added to cultures transfected with HXB-SV. The cultures were metabolically labeled from 48 to 60 hr posttransfection and viral proteins were immunoprecipitated from the cell lysates (A) or culture supernatants (B). and HXBH10-Li/s proviruses into COS-7 cells showed that the magnitude of the defect in capsid protein release caused by the p6 deletion mutation was similar in the presence and FIG. 4. Membrane association of p6-mutant particles. Transmis- absence of vpu (Fig. 5B, lanes 6 and 7). Shearing forces sion (A-C and E) and scanning (D) electron micrographs of COS-7 generated by brief vortex mixing did not substantially in- cells transfected with the mutant proviruses Li/s (A-D) and P10,11/L (E). (A, X14,800; B, x31,900; C, x19,500; D, x6600; E, crease the amount of capsid protein released from COS-7 X3200.) cells transfected with the HXBH10-L1/s provirus, indicating a rather firm attachment of the p6-deleted particles to the cell The virions that accumulated at the surface of the COS-7 surface (data not shown). cells transfected with the P10,11/L mutant generally were of more mature appearance (Fig. 4E). Only a minority of the DISCUSSION had an electron-dense outer shell as is characteristic for nascent, immature virions. A condensation of core ma- The results show that deletion or alteration of the sequences terial was seen more frequently than for the Ll/s mutant. of the C terminus of the capsid precursor polyprotein affects However, typical cone-shaped cores could not be found in virus particle maturation. In the absence of p6 sequences, the samples analyzed. For both mutants, the average number virus particles assemble at the cell membrane but are not of virus capsids visible at the surface of the transfected cells efficiently released and consequently accumulate at the cell appeared to exceed the number of particles detected upon surface. This conclusion is based on the finding that although transfection with the parental provirus. Scanning electron vigorous budding was observed on the cell surface by elec- microscopy of COS-7 monolayers confirmed that virus par- tron microscopy, the viral capsid protein synthesized by the ticles produced by the mutants Ll/s and P10,11/L accumu- p6-deletion mutant remained almost entirely cell-associated lated on the surface at a high density (Fig. 4D). as demonstrated by immunoprecipitation analysis. The Vpu Interferon a Treatment. A defect in a late stage of virus protein, which facilitates assembly of the wild-type virus maturation was previously reported upon treatment of HIV- particle (23, 24), does not contribute to the release of virus 1-infected cells with interferon a (22). To determine whether particles for mutants lacking p6. interferon a mimics defects in p6, recombinant human inter- It seems likely that p6 sequences are required to permit feron a was added to COS-7 cells 15 hr after transfection with correct, final assembly of the virus particle. Minor pertur- the parental HXBc2 provirus. In the range tested (102-104 bations in the three-dimensional structure of individual cap- units/ml), interferon a treatment resulted in a similar alter- sid precursor subunits could lead to the formatiqn of a ation of the intracellular/extracellular capsid protein ratio as nascent capsid structure that does not close perfectl. Perfect was observed in the absence of p6 (Fig. 5). However, in closure of the capsid structure may be required to sever the contrast to the effect observed with the Li/s mutant, the budding virus from the cell surface. Deletion or substantial p25/p24 ratio was not substantially altered by interferon a alteration of p6 sequences is proposed to create such alter- treatment. Electron microscopy revealed that particles of ations in the three-dimensional structure of the capsid pre- mostly mature appearance were produced by COS-7 cells cursor. treated with interferon a (data not shown). The cell-associated virus particles produced in the absence Effect of vpu. A cell-associated accumulation of progeny ofp6 sequences have an immature morphology. Although the HIV-1 virions has been observed in the absence of a func- appearance of the outer shell of the virions resembles that tional vpu product (23, 24). As the HXBc2 parental produced by protease-defective mutants of HIV-1 (5, 9), it is provirus is defective for vpu expression, the Ll/s p6 deletion likely that proteolytic processing of the capsid proteins in mutation was introduced into the vpul HXBH10 provirus these attached particles is almost complete, as most of the (23) to yield HXBHiO-Li/s. The HXBHiO-Li/s virus did not capsid precursor is cleaved in the p6-defective mutants. replicate in Jurkat cells (data not shown). A comparison of The notable exception to complete maturation processing protein synthesis after transfection of the isogenic HXBH10 is the reduction in the efficiency of the posttranslational Downloaded by guest on September 24, 2021 Microbiology: G6ttlinger et al. Proc. Natl. Acad. Sci. USA 88 (1991) 3199 modification event that converts p25 into the mature p24 Beaumont for electron microscopy studies and Dr. Ernest Terwil- protein (25). Defective core condensation observed for the liger for kindly providing the pHXBH10 plasmid. H.G.G. is a Li/s mutant may result from the inefficient conversion ofp25 Sheldon/Andelson AmFAR Scholar. J.G.S. is a Scholar of the to Kinetic Leukemia Society ofAmerica. This work was supported by National p24. studies using in vitro synthesized capsid Institutes of Health polyprotein and purified HIV-1 protease are consistent with Grants A129873, A124845, and GM39599. p25 being a processing intermediate that is eventually con- 1. Schupbach, J., Popovic, M., Gilden, R. V., Gonda, M. A., verted to mature p24 as a consequence of a slow C-terminal Sarngadharan, M. G. & Gallo, R. C. (1984) Science 224, 503- cleavage event (26). The defect in p25 to p24 conversion may 505. be a manifestation of slightly less efficient proteolytic proc- 2. Ratner, L., Haseltine, W., Patarca, R., Livak, K. J., Starcich, essing of the p6-deleted capsid precursor, which would B., Josephs, S. F., Doran, E. R., Rafalski, J. A., Whitehorn, preferentially affect p25 to p24 conversion due to the partic- E. A., Baumeister, K., Ivanoff, L., Petteway, S. R., Pearson, slow nature of this event. It seems that M. L., Lautenberger, J. A., Papas, T. S., Ghrayeb, J., Chang, ularly cleavage likely N. T., Gallo, R. C. & Wong-Staal, F. (1985) Nature (London) a reduced cleavage of p25 to p24 is not directly related to 313, 277-284. defects in release of virus particles from the cell surface, 3. Barr6-Sinoussi, F., Chermann, J. C., Rey, F., Nugeyre, M. T., because virus particles formed by mutants defective in pro- Chamaret, S., Gruest, J., Dauguet, C., Axler-Blin, C., Vdzinet- tease are shed into the medium (5). Cleavage of the capsid Brun, F., Rouzioux, C., Rozenbaum, W. & Montagnier, L. precursor does not occur in such mutants. Conversion ofp25 (1983) Science 220, 868-871. to p24 is not inhibited by interferon a treatment, yet capsid 4. Gheysen, D., Jacobs, E., De Foresta, F., Thiriat, C., Fran- protein release into the supernatant is substantially dimin- cotte, M., Thines, D. & De Wilde, M. (1989) Cell 59, 103-112. ished. 5. Gottlinger, H. G., Sodroski, J. G. & Haseltine, W. A. (1989) In the absence of no virus was observed. Proc. Natl. Acad. Sci. USA 86, 5781-5785. p6 replication 6. Bryant, M. & Ratner, L. (1990) Proc. Natl. Acad. Sci. USA 87, However, the substitution of two conserved proline residues 523-527. in the P10,11/L mutant, which behaves similarly to the 7. Kramer, R. A., Schaber, M. D., Skalka, A. M., Ganguly, K., p6-deletion mutant in COS-7 cells, does not abolish virus Wong-Staal, F. & Reddy, E. P. (1986) Science 231, 1580-1584. replication in Jurkat cells. Very little protein is released into 8. Kohl, N. E., Emini, E. A., Schleif, W. A., Davis, L. J., the supernatant of COS-7 cells transfected with the P10,11/L Heimbach, J. C., Dixon, R. A., Scolnick, E. M. & Sigal, I. S. mutant virus. In contrast, although the amount of viral (1988) Proc. Nati. Acad. Sci. USA 85, 4686-4690. protein released from Jurkat cells transfected with the mutant 9. Peng, C., Ho, B. K., Chang, T. W. & Chang, N. T. (1989) J. is less than that released from cells transfected with the Virol. 63, 2550-2556. substantial amounts of are re- 10. Veronese, F. D., Copeland, T. D., Oroszlan, S., Gallo, R. C. wild-type provirus, protein & Sarngadharan, M. G. (1988) J. Virol. 62, 795-801. leased by this mutant. This observation indicates that the 11. Gelderblom, H. R., Hausmann, E. H., Ozel, M., Pauli, G. & viral functions required for particle release may differ in Koch, M. A. (1987) Virology 156, 171-176. different cell types. 12. Aldovini, A. & Young, R. A. (1990) J. Virol. 64, 1920-1926. An inhibition of viral antigen release accompanied by an 13. Veronese, F. D., Rahman, R., Copeland, T. D., Oroszlan, S., accumulation of virus particles at the cell surface has been Gallo, R. C. & Sarngadharan, M. G. (1987) AIDS Res. Hum. observed during interferon a treatment of HIV-1-producing Retroviruses 3, 253-264. cell lines (22). Such inhibition has been attributed to an 14. Henderson, L. E., Copeland, T. D., Sowder, R. C., Schultz, inhibitory effect of interferon a on the release of preformed A. M. & Oroszlan, S. (1988) in Human Retroviruses, Cancer, virions from the membrane The data and AIDS: Approaches to Prevention and Therapy, ed. Bo- plasma (22). presented lognesi, D. (Liss, New York), pp. 135-147. here show that interferon a treatment of COS-7 cells trans- 15. Fisher, A. G., Collalti, E., Ratner, L., Gallo, R. C. & Wong- fected with the HXBc2 provirus results in an increase in the Staal, F. (1985) Nature (London) 316, 262-265. ratio of-cell-associated to cell-free viral capsid proteins, 16. Kunkel, T. A., Roberts, J. D. & Zakour, R. A. (1987) Methods consistent with the hypothesis that interferon a affects virus Enzymol. 154, 367-382. particle release. There are some differences between the 17. Dayton, A. I., Terwilliger, E. F., Potz, J., Kowalski, M., phenotype induced by interferon a treatment and the phe- Sodroski, J. G. & Haseltine, W. A. (1989) J. AcquiredImmune notype observed with the mutations affecting p6. The virus Defic. Syndr. 1, 441-452. particles produced during interferon a treatment are mostly 18. Queen, C. & Baltimore, D. (1983) Cell 33, 741-748. mature in whereas the 19. Cullen, B. R. (1987) Methods Enzymol. 152, 684-704. appearance, p6-deficient particles 20. Dayton, A. I., Sodroski, J. G., Rosen, C. A., Goh, W. C. & appear immature. The conversion of p25 to the p24 major Haseltine, W. A. (1986) Cell 44, 941-947. core protein is not apparently affected by interferon a treat- 21. Wilson, W., Braddock, M., Adams, S. E., Rathjen, P. D., ment, whereas this modification is much less efficient for Kingsman, S. M. & Kingsman, A. J. (1988) Cell 55, 1159-1169. viruses containing alterations in p6. 22. Poli, G., Orenstein, J. M., Kinter, A., Folks, T. M. & Fauci, In summary, these results reveal that the terminal se- A. S. (1989) Science 244, 575-577. quences of the HIV-1 capsid precursor contain important 23. Terwilliger, E. F., Cohen, E. A., Lu, Y., Sodroski, J. G. & determinants for the release of the assembled virus from the Haseltine, W. A. (1989) Proc. Natl. Acad. Sci. USA 86, 5163- cell surface. Inhibitors ofvirus replication that may be useful 5167. for clinical application, such as interferon a, may modify 24. Klimkait, T., Strebel, K., Hoggan, M. D., Martin, M. A. & critical Orenstein, J. M. (1990) J. Virol. 64, 621-629. elements of the release reaction. Further exploration 25. Mervis, R. J., Ahmad, N., Lillehoj, E. P., Raum, M. G., of the requirements for virus particle release from the cell Salazar, F. H., Chan, H. W. & Venkatesan, S. (1988) J. Virol. surface appears appropriate. 62, 3993-4002. 26. Tritch, R. J., Cheng, Y.-S. E., Yin, F. H. & Erickson-Viita- We thank Drs. Henry Slayter, Simon Watkins, and Elisabeth nen, S. (1991) J. Virol. 65, 922-930. Downloaded by guest on September 24, 2021