Jan 22, 1991 - KATHRYN PARTINt, GABRIELE ZYBARTHO, LORNA EHRLICHf, MARIE DECROMBRUGGHEt, ECKARD WIMMERt,. AND CAROL CARTERt§.
Proc. Nati. Acad. Sci. USA Vol. 88, pp. 4776-4780, June 1991 Biochemistry
Deletion of sequences upstream of the proteinase improves the proteolytic processing of human immunodeficiency virus type 1 (asprtic protelnase/zymogen activatIon/Iral maturation) KATHRYN PARTINt, GABRIELE AND CAROL CARTERt§
ZYBARTHO, LORNA EHRLICHf, MARIE DECROMBRUGGHEt, ECKARD WIMMERt,
*Department of Microbiology, State University of New York at Stony Brook, Stony Brook, NY 11794; and tHoward Hughes Medical Institute, Duke University, Durham, NC 27710 Communicated by Dorothy M. Horstmann, January 22, 1991 (received for review September 14, 1990)
Hum immunodeficiency virus type 1 exABSTRACT presses struct proteins and replicative enzymes within gag and gag-pol precursor polyproteins. Specific proteolytic processing of the precursors by the viral proteinase is essential for maturation of infectious viral particles. We have studied the activity of proteinase in its immature form, as part of a gag-pot fusion protein, in an in vitro expression system. We found that deletion of p6*, the region in po up m of proteinase, resulted in improved processing of the precursor. A modified proteinase is released, but it functions less ecen than wild type. Improved autoprocessing correlates with increased accessibility of the active site region in the polyprotein carrying the p6* deletion. Our results suggest that p6* is involved in the regulation of proteinase activation, perhaps as a region Umiting the interaction of the active site and substrate binding domain with the remainder of the polyprotein. Release of p6* inhIbiton may be an activation step necessary for infectious particle maturation.
all processing of viral polyproteins and lead to budding of noninfectious particles (2-4). Mutations that block processing at specific gag sites (i.e., Tyr-Pro or Leu-Ala) also result in production of noninfectious particles, with aberrant morphology (4). Thus, proteolytic processing by HIV-1 PR appears to be intimately linked with morphogenesis. However, it is not known by what mechanism assembly, processing, and budding are coordinately regulated. All known mammalian aspartic proteases are encoded as zymogens, which are activated after cleavage from an upstream regulatory region (15). Frameshifting within the NC domain of the gag gene leads to the synthesis of a gag-pol polyprotein in which the proteins are ordered MA-CA-NCp6*-PR-RT-IN. The 68-amino-acid sequence, called p6*, encoded upstream from PR in the pol reading frame (16), has no ascribed function. We report here a study of the participation of p6* in the regulation of PR activity. We find that deletion of p6* from the polyprotein improves proteolytic processing.
Human immunodeficiency virus type 1 (HIV-1) encodes two precursor polyproteins, gag and gag-pol, which must be proteolytically processed to produce mature structural proteins [matrix (MA), capsid (CA), nucleocapsid (NC), and p6] and replicative enzymes [proteinase (PR), reverse transcriptase (RT), and integrase (IN)] (1). Maturation of infectious particles requires accurate processing of more than 10 cleavage sites recognized by the virus-encoded PR (2-4). Unlike substrates that are recognized by cellular proteases and proteinases encoded by nonretroviral RNA viruses, neither the amino acids constituting the scissile bond nor the flanking amino acids that interact with the PR binding cleft are identical in primary amino acid sequence (5, 6). HIV-1 PR is encoded as an 11-kDa domain within the pol region of the
MATERIALS AND METHODS Construction of Wild-Type (wt) and Mutant rai. The construction of the parent plasmid (pHIV-FSIII) has been described (9, 16). Briefly, pHIV-FSIII was made by trans-
ferring nucleotides 221-2130 of the HIV cDNA (from BH10) (17) into pBS/KS' (Stratagene) (16). FS-III contains a 4-base-pair insertion at nucleotide 1640 of the HIV cDNA and two stop codons 3' to the PR coding sequence. FS-DTG (16) contains a mutation in the PR domain (D25A) that inactivates PR, resulting in accumulation of precursor polyproteins. Plasmids were propagated in Escherichia coli strain BH7118, C600, or MV 1109. Site-directed mutagenesis was performed by using the method of Kunkel (18) with minor modifications (using the Bio-Rad mutagenesis kit) (16). Deletion of p6* was done by introducing a second BgI II site at nucleotide 1837 (AGATCA to AGATCT) and then digesting that plasmid with Bgl II. Removal of the nucleotide 16401837 fragment and religation of the vector causes deletion of p6*, leaving an intact Bgl II site. Mutation was confirmed by sequencing. Expression in Vitro. Expression of this construct in vitro was performed as described (16). After in vitro transcription using T7 polymerase, synthetic RNAs were translated in rabbit reticulocyte lysates (RRLs; Promega), following the protocol of the supplier, in the presence of [35S]methionine (specific activity = 1100 Ci/mmol; 1 Ci = 37 GBq; NEN). Reactions were analyzed by SDS/13.5% PAGE and fluorog-
gag-pol polyprotein (7-10). Sequence analysis (11), mutational dissection (12), and crystallographic studies (13, 14) of purified PR have shown that this enzyme is a member of a family of mammalian aspartic proteases. However, it encodes only half of a catalytic center (Asp-Thr-Gly). Furthermore, proteolytic activity is found in association with a 22-kDa protein. Thus, mature PR activity requires both the scission of the PR domain from the precursor polyprotein at virus-specific cleavage sites and the association of two monomers to form an enzymatically active dimer. It remains to be determined whether HIV-1 PR has enzymatic activity only as a mature dimer and by what mechanism its dimerization and maturation occurs, as no cellular proteinase has been identified that catalyzes the release of PR from the polyprotein. Processing by HIV-1 PR is essential in the viral life cycle. Mutations that destroy the catalytic site result in absence of
Abbreviations: HIV-1, human immunodeficiency virus type 1; PR, proteinase; MA, matrix; CA, capsid; NC, nucleocapsid; RT, reverse transcriptase; IN, integrase; wt, wild type; RRL, rabbit reticulocyte
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
lysates. §To whom reprint requests should be addressed.
4776
Proc. Natl. Acad. Sci. USA 88 (1991)
Biochemistry: Partin et al. raphy (EN3HANCE; NEN). Where indicated, an LKB Bromma UltroScan laser densitometer was used for densitometric analyses. Immunoprecipitation Analysis. The immunoprecipitation analyses shown were performed by using antibodies against PR [provided by Phillip Barr (Chiron) and Mary Graves (Hoffman-LaRoche)]. One to 5 ,ul of RRL was diluted 100-fold, and immunoprecipitation was done as described (16). Purification of PR Expressed in E. coli and Proteolytic Processing in Trans. Purification of PR and assay of activity in trans reactions was performed as described (16, 19).
RESULTS Deletion of p6*, the Region Upstream of PR, Improves Proteolytic Processing. To investigate whether the p6* domain exerted any effect on proteolytic processing, we constructed a mutant with a deletion of this region, called Ap6*. This mutant lacks all of p6* and also the C-terminal processing site in NC (Phe-Leu) and the N-terminal cleavage site of PR (Phe-Pro). However, it maintains an Asn-Phe site near the C terminus of NC. Expression of Ap6* should yield a polyprotein, which may be cleaved at the Asn-Phe site, resulting in a PR with a modified N terminus (Phe-Leu-Gly-Lys-PR, or FLGK-PR) that is two residues larger than authentic PR. Fig. 1 shows the products after in vitro translation of wt and WT FSIII
CA
Bgl 11 Bg1 1l PR II NC 6 N/F
r' rr &p6'* FMAI
F/P
CA 1 "u---PR I
RRL 2 3 4
M
9669-
-
a,-
a'_i!
*
Ap6* and schematically defines the precursor, intermediates,
and products based on immunological analysis as described (16). These translation reactions were done with equimolar amounts of synthetic mRNA, as determined by incorporation of [a-32P]ATP during in vitro translation. The wt pattern (Fig. 1, lane 2) was identical to that previously described (9, 16). wt lysates characteristically contain residual full-length precursors, initiated at the first methionine residue (product a, precursor protein pr69), and molecules initiated internally at the second methionine residue (Met-142) and methionine residues further downstream (ref. 16; product c, pr50 and smaller, respectively). The identity of primary translation products was confirmed by translation of FS-DTG (Fig. 1, lane 3), which contains an inactivating mutation in the PR domain (16, 20). The major products in wt lysates (Fig. 1, lane 2), MA-CA (product d, p41) and CA (product f, p24/25), were detected in expected amounts based on their content of [35S]methionine. PR (product g, p1l) was obscured by hemoglobin in RRL lysates but was detected after immunoprecipitation (cf. Fig. 2). Antibody for detection of mature NC (p7) was not available. Mature MA and p6* have no methionine residues and, therefore, were not detectable. Translation of mRNA encoding polypeptide Ap6* (Fig. 1, lane 4) resulted in a pattern of synthesis and proteolytic processing that was similar to that obtained following translation of wt mRNA, with the exception that wt contained residual precursors and intermediates (products a and c) that were reduced or entirely absent in Ap6* (products a' and c'). The mutant produced mature CA (product f) and a new 18-kDa species (product e'), which was anti-PR reactive (see below). Eighteen kilodaltons is the calculated molecular mass of the NC and PR (NC-PR) fusion protein, which would be expected to accumulate if the Asn-Phe site were not efficiently utilized. Apparently, this site was not efficiently utilized under these conditions, since NC-PR was detected more readily than PR (band g in Fig. 2, lane 6). The wt and Ap6* precursors and products detected after =45 min were stable, and the pattern remained essentially unchanged for at least 5 hr. Consistent with the lower level of residual precursor,
a
0-
c
-
d-
w.a
Ap6*
WT
-a
46-
+
M
-"c-c
PR
+
--
+
r*
C
--
;
C
--
_
... ------
_
30- _
6946-
dFT" f-*W
s-f e
f
9 14-
4777
I
r*
30;_
,
-
*s
;
b cl d
*-
--_
* -. *:
FIG. 1. In vitro translation products of wt and Ap6*. The autoradiograph shows products after translation in RRL of no RNA (lane 1) or equimolar amounts of wt (lane 2), FS-DTG (lane 3), or p6* (lane 4). Lane M, molecular size markers (in kDa). The schematic drawing shows construction of wt and Ap6* and identification of the translation products based on their immunoreactivity to antibodies against HIV proteins (data not shown). Products are identified by lowercase letters. Note that products c and c' are initiated at internal AUG codons (16); product f represents CA; product e' represents the NC-PR fusion product; and product g represents PR, which, on this gel, cannot be seen. Dashes, region of deletion; bent arrows, sites of initiation of translation (5' proximal or internal); straight arrows, cleavage sites; filled circles, termination of translation. The letters in the elongated box indicate the amino acid sequence at the NC-PR fusion. Boldfaced letters, PR sequence.
'4-J*4f9
.4
T: 9*
-e -
t
2 3 4 5 6 7 8
FIG. 2. Characterization of wt and Ap6* products. wt (lanes 1-4) and Ap6* (lanes 5-8) products were characterized by imnunoreactivity to antibodies against PR (aPR) before and after treatment with purified PR in trans assays. Lanes 1 and 5, sample applied directly; lanes 2 and 6, immunoprecipitation with anti-PR; lanes 3 and 7, digestion with PR; lanes 4 and 8, immunoprecipitation with anti-PR following PR digestion; lane M, molecular size markers (in kDa). Letters refer to bands indicated in Fig. 1.
4778
Biochemistry: Partin et al.
Proc. Natl. Acad. Sci. USA 88 (1991)
WT 0.886
'x
A
x/y- 1.2 1.858.
Ap6* x x/y=
7.2
x/y-
5.0
compare lanes 1 and 3 as well as lanes 5 and 7). Complete digestion of translation products with PR followed by immunoprecipitation with anti-PR indicated that wt and Ap6* lysates contained approximately the same amount of labeled PR (band g in Fig. 2, lanes 4 and 8). Note that band g in Fig. 2, lane 8 migrated more slowly than wt PR, possibly due to the N-terminal modification. We conclude that processing of the Ap6* precursor polyprotein is more efficient than wt. To quantitate the higher efficiency of Ap6* processing relative to wt, translation was done over a range of RNA concentrations (50-200 nM), and the ratio of CA (x) to residual precursor (y) was calculated after densitometry of autoradiographs (Fig. 3). Fig. 3 shows products obtained with increasing mRNA concentrations. Since some forms of c' and d comigrate, the entire peak at y (a', c', and d) was taken as Ap6* precursor. By this criterion, deletion of p6* improved proteolytic processing -2-fold at 200 nM mRNA (Fig. 3 G and H) and 6-fold at 50 nM mRNA (Fig. 3 A and B) compared with wt. Interestingly, with increasing amounts of added mRNA, x/y increased slightly for wt up to 150 nM mRNA (Fig. 3 A, C, and E) but was dramatically decreased for Ap6* (Fig. 3 B, D, F, and H). For wt, an increase in mRNA concentration from 50 to 200 nM resulted in a 3-fold increase in precursor and a similar 2.5-fold increase in CA. In contrast, increasing Ap6* mRNA increased precursor by 5-fold and had little affect (-1.2-fold) on the amount of CA produced. These results suggest that wt and Ap6* differ in the underlying processing mechanism and that the mutation affected the initial cleavage event. Activity of the Altered PR Produced by Ap6*. We investigated the possibility that improved processing by Ap6* was due to formation of an altered form of PR that was more active and/or more stable than wt PR. Two proteins containing PR sequences can be produced from Ap6*, NC-PR (product e') and FLGK-PR (product g). Their relative amounts are dependent on the efficiency of cleavage at the Asn-Phe site. Since NC-PR failed to autoprocess, or to be processed further by FLGK-PR in the lysate, it seemed unlikely that these products were efficient enzymes. To synthesize the processed forms of PR exclusively, we expressed Ap6* or FS-WT in E. coli, where we have previously demonstrated rapid and complete wt polyprotein processing with accumulation of mature PR (9, 21). We then directly compared the activities of the mutant and wt enzymes in trans assays (Fig. 4; refs. 16 and 19). For these experiments, we
0.269
0.269
1.187. C
/y
.4
x/y
1.7
x/y
09
1.949
D
-n
Qf) 1 0 i9 cl
O.
1.590 E
2 64 r \vU_
2.07
F
x/y 3.0
2.2421
H
x/y
0.281
'
C)
0.295 1.824
0. 2 76
0.286
1.9
'
Top
Migration
Migration
Top
FIG. 3. Densitometric analysis of the wt and Ap6* products synthesized in translation reactions using 50-200 nM mRNA. Translation reactions were done using wt (A, C, E, and G) or Ap6* (B, D, F, and H) mRNA at concentrations of 50 (A and B), 100 (C and D), 150 (E and F), or 200 nM (G and H). Translation products were quantitated following SDS/PAGE by densitometry and autoradiography. x, migration position of CA; y (for wt), migration positions of a, b, and c; y (for Ap6*), a', c', and d. Note that the range on the y axis is different in the panels.
relatively more CA accumulated in Ap6* lysates than wt lysates (band f in Fig. 1, lanes 2 and 4). Treatment of both wt and Ap6* lysates with exogenously added PR resulted in formation of additional CA, as expected (band f in Fig. 2, A M
2 3 4 M40406m
46-..3..0-... a. :
5 6 7 omplww
t
_
11 12 13
8 9 10 wmw
-
~
14 15 16
17 18 19
-pr 53 - pr4
'M.m-
-
B
30- 4
_.Nwqp
A
4omw
_
all
-CA
( f) 14-
M
23
-s. ...-
:
0 .-
30 -
14-
FIG. 4. (A) Proteolytic processing of in vitro-synthesized HIV gag precursor proteins. Samples of translation mixtures programmed with gpII RNA were analyzed directly after translation (lane 1) or incubated with bacterial lysates from uninduced (lanes 2-4) or induced (lanes 5-19) cells. Cleavage reactions were carried out in a 40-,u final volume by using 1 ,ul of translation mixture as substrate and various amounts of lysate as enzyme. Incubation was at 30°C. Reaction products were analyzed on 12.5% gels. Lanes 2-4, incubation for 60 min with 5 A of uninduced lysates from Ap6* (lane 2), 10 1,u of uninduced lysates from Ap6* (lane 3), or 5 ,ul of uninduced lysates from wt (lane 4). Lanes 5-7, incubation with S ,ul of induced lysates from Ap6* for 15, 30, or 60 min. Lanes 8-10, incubation with 10 IL of induced lysates from Ap6* for 15, 30, or 60 min. Lanes 11-13, incubation with 1 jl of induced lysates from FS for 15, 30, or 60 min. Lanes 14-16, incubation with 2.5 1A of induced lysates from FS for 15, 30, or 60 min. Lanes 17-19, incubation with 5.0 IL of induced lysates from FS for 15, 30, or 60 min. Lane M, molecular size markers (in kDa). (B) Western analysis of PR from cells expressing FS and Ap6*. Lane 1, partially purified mature wt PR (11 kDa; ref. 19); lane 2, lysate of induced cells carrying FS; lane 3, lysate of induced cells carrying Ap6*. The arrow indicates the PR position.
Biochemistry: Partin et al.
Proc.
I
MA
CA
cQ
z
L MLO
f
M
DTG - ATG +
CC
CC
M
c
translation products were incubated with lysates of induced cells containing wt PR. The results support the conclusion that improved processing by Ap6* was not due to formation of a more efficient enzyme. If the effect of Accessibility of the Active Site in is in to improve the sequences Ap6* upstream of deletion efficiency of the initial cleavage event, the deletion may make the active site and binding cleft more accessible. One way to test this hypothesis is according to the strategy of Rajogopalan et (22). These investigators have shown that the active site of pepsin is more available for modification in the absence of its upstream region than in its presence. In their study, for which purified pepsinogen was available, the modification was done with diazo compounds. We used studies to attempt to show the same effect on the accessibility of the PR active site in the Ap6* gag-PR pre-
F
U
U
c
z
p6*
A
WT DTG-ATG
Ml
CU
2
En
0
Ap6*.
LO
96 69
4779
uninduced
R iI
i1l
111111
Acad. Sci. USA 88 (1991)
cells carrying Ap6* (Fig. 4A, lanes 2 and 3) or FS (Fig. 4A, lane 4), or incubated with lysates of induced cells carrying Ap6* (Fig. 4A, lanes 5-10) or FS (Fig. 4A, lanes 11-19). The precursors were completely stable in buffer alone or in bacterial lysates from uninduced cells. Very limited extracts from cleavage occurred upon incubation with mutantonly induced cells. Efficient cleavage was observed when the
DTG
I
NatL.
al.
-
6 4^
antibody
30-
cursor.
U 14-
:tive hwt
FIG. 5. Accessibility of the PR ac site in wt and Ap6* polyproteins. (Upper) Mutations of boti and Ap6* were made, which altered the active site Asp resin idue to Ala, rendering PR enzymatically inactive (wt-DTG andAp(5*-DTG). The double arrow indicates the site of the mutation in both4 clones. (Lower) Products of translation from both were immunopreicipitated with an antibody directed against the active site region of PR. Immunoprecipitations were performed in 10mM Tris under various conditions: the sample was applied directly (0) or the sample w, as precipitated and washed with 0.05 M NaCl (nondenaturing), 1.0] M NaCl (disruptive), RIPA buffer (denaturing), or RIPA buffer with dithiothreitol (DTJ) (denaturing). Lane M, molecular size marker s (in kDa).
duct
RNA, which
used as substrate the translation pro( of gpII encodes gag precursor polyproteii ns pr53 and pr4l (19). Western blot analysis (Fig. 4B) of the crude Ap6* lysate (Fig. 4B, lane 3) identified PR sequences in a band that migrated close to the position of purified (Fig. 4B, lane 1) or crude (Fig. 4B, lane 2) wt mature PR, as expects -d for FLGK-PR. RRLs, programmed with gpII RNA, were either analyzed directly after translation (Fig. 4A, lane 1), itncubated with lysates of
HIV2
__ _ TPSGSSSGSTGE
HIV1
F F RE DL AF
apparent 2.5-fold increase in active site accessibility.
Algnment of the Proenzyme Fragment of Pepsinogen with Lentivirus p6*. Fig. 6 shows an alignment of the proenzyme
segment of porcine pepsinogen and upstream sequences of several lentiviruses (23, 24). The proenzyme fragment shows a strong dominance of basic amino acids. These are thought to play an important role in its function as an inhibitory domain (23). In contrast, in p6*, acidic and basic residues are is interspersed, making it unlikely that a similararemechanism utilized. For HIV-1 and HIV-2, whose PR very closely conrelated, 63% of the residues in p6* were functionally served (26) by the alignment, although only 17% of positions were identical. In comparison, 81% functional conservation (43% identity) is observed for the PR domains (27). The conserved residues in p6* are clustered in two regions. These
IA R AGO DT
REETI OGSD RD L
YAR-R K T E
- --
EIAV
.E TNF PDTTE E SAQQ I
----------------
-IT
CCT-Rf
L VKVPL VRKKSL RONL I
K "ISIVPRSE RNKIKG
SCKYG E GSIS--RGSIQPGIC
KDHKI L K FLKTHKHN) iS K YF P I 0
0
00
0
0
L AA
E T AGKQOQ EIG NT C GIAVR A
R N NH PISIL WK K R T Y AKOG L S S
0 G A T NR
S E A GIA D R OIG TIV_-SIF N F
LOGKARE F SSE -QITI p AANS PITIRRE L O VWG R Di
VISN
PIGP
Fig. 5 shows immunoprecipitation of the translation prodAp6*-DTG, respectively) under native and denaturing conditions. Antibody directed against 22 residues around and including the active site (20) had a higher affinity for theAp6* polyprotein than the wt polyprotein. Quantitation by densitometry of the wt-DTG and Ap6*-DTG lanes after immunoin 0.05 M NaCl indicated that this effect was precipitation 2.5-fold greater for Ap6* compared to wt. Thus, the %3-fold exhibited by Ap6* at improvement of polyprotein processing this mRNA concentration (=70 nM) correlates well with the ucts of wt gag-PR and Ap6* gag-PR precursors (wt-DTG and
0
0
IPL El jA
o
I
YF VL
UN
0 00
00
0
0 00
00
0
YL D
FIG. 6. Alignment of the N-terminal fragment of porcine pepsinogen (23) with lentivirus "p6*" (24). Sequences were aligned by using a progressive alignment algorithm (ALIGN; ref. 25). The boxed regions represent positions of conserved residues; one circle, residues conserved through the virus family; two circles, residues conserved through viral and pepsinogen sequence; arrowhead, cleavage site in viral sequence; arrows, cleavage sites in pepsinogen. Groups functionally conserved are as described by Dayhoff et al. (26). VISN, Visna virus; EIAV, equine
anemia infectious virus; PIGP, porcine pepsinogen.
4780
Biochemistry: Partin et al.
regions also exhibit the greatest similarity to the proenzyme fragment and are positioned by alignment near the activation cleavage sites in the proenzyme fragment (23). The lack of strong sequence conservation makes it unlikely that specific residues in p6* are critical for interaction with PR. However, the similarities suggest that p6* may play an important structural role in regulation of PR activity.
DISCUSSION The restriction of proteolytic processing of the gag-pol polyprotein prior to assembly and budding of the retrovirus particle is a major conundrum. We have shown that sequences upstream of those encoding mature PR can inhibit proteolytic processing in an in vitro translation/proteolytic processing system. Certainly, this inhibition is not complete, as processing occurs in wt constructs. Moreover, oUr constructs lacked the C portion of pol, including all of RT and IN. Definitive analysis of the contributions of various segments of'the polyprotein towards processing will require inclusion of these sequences. Nevertheless, the function of the p6* region in the gag-pol polyprotein may be solely to delay processing of newly formed polypeptide chains. One explanation for improved processing by Ap6* is that this mutation affects autocatalysis. Improved processing upon removal of upstream sequences is a characteristic feature of the cellular homologues of HIV PR. In some (e.g., pepsinogen), the activation step is independent of the proenzyme concentration, suggesting an intramolecular activation process in which the precursor molecule cleaves itself.' James and Sielecki (23), in an elegant study of the crystal structure of the zymogen pepsinogen, proposed that upstream sequences bind near the active site of pepsin, rendering the binding cleft inaccessible to substrates. We demonstrated here that removal of p6* increased the accessibility of the active site and binding cleft in the PR domain of the polyprotein. Alignment of the proenzyme fragment of'pepsinogen and the p6* regions of HIV-1 and other lentiviruses provides'a basis for speculation that a functional similarity exists between these regions. p6* may function as a spacer region in the gag-pol precursor, between PR and distal domains containing Asp residues that can potentially interact with the catalytic Asp-25 in PR. The slow step in the wt activation cleavage may be a bimolecular interaction that requires alteration of the p6* spacer function. Deletion of p6* may permit PR-substrate interaction to take place more freely, resulting in improved processing. Improved autocatalysis could also result in release of a more efficient altered enzyme. However, our results, including analyses in vitro, in trans assays with crude enzyme, and insertional mutagenesis of Ap6* to change the structure of the modified enzyme (data not 'shown), do not support this conclusion. Other possibilities, such as conformational changes in the polyprotein induced by deletion or alteration of possible ordered processing of cleavage sites (28), have not been 'eliminated. The molecular events that coordinately regulate assembly have not'been elucidated. p6* inhibition of PR could play a role in this process by preventing PR from cleaving itselffrom the polyprotein prematurely. During the late stages of infection, gag and gag-pol polyproteins aggregate under the cell membrane of an infected cell. When the concentration at the surface becomes high, cooperative gag-gag interactions, possibly driven by the self-association of the CA domain (21), may cause alteration of a p6*-PR interaction, relieving repression and permitting PR dimerization. This could trigger the cascade of proteolytic events leading to formation of infectious virions.
Proc. Natl. Acad. Sci. USA 88 (1991) We thank Sung Key Jang and S. F. J. Le Grice for suggesting some of the experiments included here. We thank Peter Kissel for synthesis of oligonucleotides, John Dunn for providing T7 polymerase, Chris Helmke for photography, Joyce Schirmer for artwork, and Christopher Hellen and Patrick Hearing for scientific discussions and critical reading of the manuscript. This work was supported by Grant AI-25993 from the National Institutes of Health.
1. Leis, J., Baltimore, D., Bishop, J. M., Coffin, J., Fleissner, E., Goff, S. P., Oroszlan, S., Robinson, H., Skalka, A. M., Temin, H. M. & Vogt, V. (1988) J. Virol. 62, 1808-1809. 2. Kohl, N. E., Emini, E. k., 'Schleif, W. A., Davis, L. J., Heimbach, J. C., Dixon, R. A. F., Scolnick, E. M. & Sigal, I. S. (1988) Proc. Nat!. Acad. Sci. USA 85, 4686-4690. 3. Peng, C., Ho, B. K., Chang, T. W. & Chang, N. T. (1989) J. Virol. 63, 2550-2556. 4. Gottlinger, H. G., Sodroski, J. G. & Haseltine, W. A. (1989) Proc. Nat!. Acad. Sci. USA 86, 5781-5785. 5. Hellen, C. U. T., Krausslich, H.-G. & Wimmer, E. (1989) Biochemistry 28, 9881-9890. 6. Krausslich, H.-G. & Wimmer, E. (1988) Annu. Rev. Biochem. 57, 701-754. 7. Debouck, C., Gorniak, J. G., Strickler, J. E., Meek, T. D., Metcalf, B. W. & Rosenberg, M. (1987) Proc. Natl. Acad. Sci. USA 84, 8903-8906'. 8. Graves, M. C., Lim, J. J., Heimer, E. P. & Kramer, R. A. (1988) Proc. Nat!. Acad. Sci. USA 85, 2449-2453. 9. Krausslich, H.-G., Schneider, H., Zybarth, G., Carter, C. & Wimmer, E. (1988) J. Virol. 62, 4393-4397. 10. Le Grice, S. F. J., Mills, J. & Mous, J. (1988) EMBO J. 7, 2547-2553. 11. Pearl, L. H. & Taylor, W. R. (1987) Nature (London) 329, 351-354. 12. Loeb, D. D., Hutchinson, C. A., III, Edgell, M. H., Farmerie, W. G. & Swanstrom, R. (1989) J. Virol. 63, 111-121. 13. Navia, M. A., Fitzgerald, P. M. D., McKeever, B. M., Leu, C.-T., Heimbach, J. C., Herber, W. K., Sigal, I. S., Darke, P. L. & Springer, J. P. (1989) Nature (London) 337, 615-620. 14. Wlodawer, A., Miller, M., Jaskolski, M., Sathyananayana, B. K., Baldwin, E., Weber, I. T., Selk,' L. M., Clawson, L., Schneider, J. & Kent, S. B. H. (1989) Science 245, 616-621. 15. Tang, J. & Wong, R. N. S. (1987) J. Cell Biol. 33, 53-63. 16. Partin, K., Krausslich, H.-G., Ehrlich, L. S., Wimmer, E. & Carter, C. (1990) J. Virol. 64, 3938-3947. 17. Ratner, L., Haseltine, W., Patarca, R., Livak, K. J., Starcich, B., Josephs, S. F., Doran, E. R., Rafalski, J. A., Whitehorn, E. AK, Baumeister, K., Ivanoff, L., Petteway, S. R., Jr., Pearson, M. L., Lautenberger, J. A., Papas, T. S., Ghrayeb, J., Chang, N. T., Gallo, R. C. & Wong-Staal, F. (1985) Nature (London) 313, 277-283. 18. Kunkel, T. (1985) Proc. Natl. Acad. Sci. USA 82, 488-492. 19. Krausslich, H.-G., Ingraham, R. H., Skoog, M. T., Wimmer, E., Pallai, P. V. & Carter, C. A. (1989) Proc. Natl. Acad. Sci. USA 86, 807-811. 20. Mous, J., Heimer, E. P. & Le Grice, S. F. G. (1988) J. Virol. 62, 1433-1436. 21. Ehrlich, L. S., Krausslich, H.-G., Wimmer, E. & Carter, C. (1990) AIDS Res. Hum. Retroviruses 6, 1169-1175. 22. Rajogopalan, T. G., Stein, W. H. & Moore, S. (1966) J. Biol. Chem. 241, 4295-4297. 23. James, M. N. G. & Sielecki, A. R. (1986) Nature (London) 319, 33-38. 24. Myers, G., Josephs, S., Berzofski, J. A., Rabson, A. B., Smith, T. F. & Wong-Staal, F. (1989) Theoretical Biology and Biophysics (Los Alamos Natl. Lab., Los Alamos, NM). 25. Feng, D.-F. & Doolittle, R. F.' (1990) Methods Enzymol. 183, 375-387. 26. Dayhoff, M. O., Eck, R. V. & Park, C. M. (1972) in Atlas of Protein Sequence and Structure, ed. Dayhoff, M. 0. (Natd. Biomed. Res. Found., Washington), Vol. 5, pp. 89-99. 27. Weber, I. T. (1989) Gene 85, 565-566. 28. Henderson, L. E., Benveniste, R. E., Sowder, R., Copeland, T. D., Schultz, A. M. & Oroszlan, S. (1988) J. Virol. 62, 2587-2595.
