Identification of protein intermediates in the processing of the p55 HIV ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 264 No. 15. Issue of May 25, p 84598462,1989 0 1989 by The American Soeiet; for Biochemistry and M o k l a r Biology, Inc. Printed in U.S.A.

Communication

viral proteins by a virally encoded protease (1-6,12). The gag open reading frame (ORF) of HIV-1 encompasses 1.7 kilobases of the 5’ portion of the genome and is translated as a 55-kDa (p55) precursor polyprotein (1-6, 12). The p55 precursor polyprotein comprised, in order, of NHz-p17-p24-p7p6-COOH is cleaved post-translationallyintomature 24(p24), 17- (p17), and 15- (p15) kDa proteins through at least (Received for publication, January 19, 1989) two intermediate proteins of approximately 41 kDa (7-10). Recently, the p15 protein hasbeen shown to be further cleaved Shantharaj D. GowdaS, Barry S.Stein, and to yield two peptides, 7 (p7) and 6 (p6) kDa in size (7). Edgar G . Engleman Despite the identification of these gag gene products, there From the Department of Pathology, Stanford University remains a relative paucity of information relating to the School of Medicine, Stanford, California 94305 and the processing pathway of the p55 precursor. For example, the Stanford University Bfood Bank, origin of the two intermediate proteins is uncertain. In this Palo Alto, California94304 study we have used a recombinant HIV/vaccinia virus We haveusedarecombinantvaccinia virus (VV) (VV:gag) to analyze the processing pathway of the p55 precursor protein. In a previous report we described the construcwhich expresses high levels ofhumanimmunodefition of VVgag and demonstrated that VV:gag-infected cells ciency virus-1 (HIV-1) gag proteinstoanalyzethe processing pathway of the gag p66 precursor. HIV-1 rapidly and predictably express high levels of the gag polyprogagproteins were isolated from [SH]leucine-labeled tein (10). In the current report we provide evidence to show VV:gag-infected H9 T lymphocytes by immunoprecip- that the initial cleavage of the p55 precursor occurs a t both itation with either anti-p24, anti-pl7, or anti-p6 an- the NH2 and COOH termini of the p24 protein, yielding three tibodies.Sodiumdodecylsulfate-polyacrylamide gel processing intermediates of 39 (p39), 41 (p41a), and 41 (p41b) electrophoresisanalysis revealed that processing of the kDa. One (p41b) is the product of NHr-terminal processing p56 precursor involves three major intermediates of the p55 precursor, while the other two apparently result (p41a, p41b, and p39). The p41a and p39 proteins from COOH-terminal cleavage of the precursor a t two distinct contain the p17 and p24 protein segments, and the sites. p41b is comprised of p24 and p16 segments. On twodimensional gels, each intermediateas well as the maMATERIALS AND METHODS ture p24 and p17 proteins migratedas distinct species. VV:gag virus stock preparation, the propagation of the T lympho[‘HIMyristic acid labeling of the HIV-1 gag proteins blastic cell line (HS), metabolic labeling with either ~-[3,4,5-~H] revealed that in addition to p66 and p17, the p41a and leucine or [9,10-3H]myristicacid (Du Pont-New England Nuclear), p39 intermediates, butnot p41b,aremyristylated, radioimmunoprecipitation, and SDS-PAGE were performed as deconfirming that myristylation occurs at the NHa ter- scribed previously (10). Nonequilibrium pH gradient electrophoresis minus before cleavage of the pi55 precursor protein. (NEPHGE) was performed according to O’Farrell et al. (11). Rabbit p24 antiserum (anti-p24; IgG2) was kindly provided by Dr. We conclude that the myristylated HIV- 1 gagp65 precursor is initially cleaved at random either at the p17/ Kathelyn S. Steimer (ChironCorp., Emeryville, CA). Anti-pl7 monop24 junctionorat two sites between p24 and p16 clonal antibody (IgG2) was purchased from Cellular Products, Inc. proteins, resulting in three intermediates (p41a, p41b, (Buffalo, NY). Anti-p6 monoclonal antibody (IgG1)(7) was a generous gift from Dr. Fulvia di Marzo Veronese (Bionetics Research, Rockand p39) which are subsequently cleaved to yield ma- ville, MD). ture gag proteins.

Identification of Protein of Intermediates in the Processing the p55 HIV-1 gag Precursor in Cells Infectedwith Recombinant Vaccinia Virus*

RESULTS AND DISCUSSION

The genome of the human immunodeficiency virus (HIV1)‘consists of three major genetic elements that are arranged in the order 5‘-gag-pol-env-3‘. The structuralproteins of the nucleocapsid (encoded by the gag gene) and the replication enzymes (encoded by the pol gene) are exclusively synthesized as polyproteins that are proteolytically processed to mature *This research was supported in part by Grants A125922 and A172657 from the National Institutes of Health and by the Medical Research Service of the Veterans Administration. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “udvertisement” in accordance with 18 U.S.C.Section 1734 solely to indicate this fact. 4 To whom correspondence should be addressed Stanford Blood Center, 800 Welch Rd., Palo Alto, CA 94304. ’The abbreviations used are: HIV-1, human immunodeficiency virus-1; NEPHGE, nonequilibrium pH gradient gel electrophoresis; W, vaccinia virus; SDS, sodium dodecyl sulfate, PAGE, polyacrylamide gel electrophoresis.

The gag precursor, synthesized as a 55-kDa polyprotein, is eventually cleaved to yield the mature core proteins (p17, p24, p7, and p6)(7-9). Recent studies have suggested the existence of two 41-kDa intermediate proteins from which the mature gag products are ultimately derived (7-9). We reasoned that if the two intermediate forms result from random cleavage of either the NHZ-terminal p17 or the COOH-terminal p15, then antibodies to p24 should recognize both forms, whereas antibodies to p17 (NHz-terminal component of p55) and p6 (COOH-terminal peptide of p55) should precipitate one but not bothof the two forms. To explore these possibilities, H9 cells infected previously with VV:gag (2 plaque-forming units/cell) for 1 h were metabolically labeled with [3H]leucine for 6 h. The lysates were subjected to radioimmunoprecipitation and SDS-PAGE analysis as described previously (10) using anti-p24, anti-pl7, and anti-p6antibodies. The results are summarized in Fig. 1. As predicted, the anti-p24 antibody precipitated two intermediate forms of 41 and 39 kDa in addition to p55 and p24

84519

Characterization of HIV-1 gag Processing Intermediates

8460

C

B

A 1

1

2

2 1

2 kD

- 160

-

0 *

0

-

55

-41/39

am

- 25/ 24

*r)

-11/15

FIG. 1. RadioimmunoprecipitationofHIV-1 gag p r o t e i n s expressed in VV:gag-infected cells. Infection of H9cellswith V\':gag. radio1al)eling with [,'H]leucine(0.25 mCi/ml), and immunoprecipitation were performed as described (10). Theprecipitated proteins were fractionated in 9-11'; linear gradient gels. A, lane 1, proteins eluted from nonimmune rahbit serum resin incubated with an extract of radiolaheled VV:gag-infected H9 cells (VV:gag lysate); lnnc, 2. proteinseluted from rabhit anti-p24 resin incubated with VV:gag lysate. H , lnnc I , proteins elutedfrom isot-ype-matched control monoclonal antibody resin incuhated with VV:gag lysate; lane 2, proteins eluted from anti-ply resin incubated with VV:gag lysate. C, Innr I , proteins eluted from goat anti-mouse I& resin incubated with VV:gag lysate; lnnc 2, proteins eluted from goat anti-mouse IgG resin incuhated with VV:gag lysate previously treated with anti-p6 specific murine monoclonal antihody ( 7 ) for 16 h at 4 "C.

p41a and p39 contain the p17 and the p24 protein segments, and p41b is comprised of p24 and p15 proteins. Since thep41a and p41b proteins differ in length by only 7 amino acids, these intermediate proteins are probably not resolvable by SDS-PAGE. T o distinguish these intermediate proteins from one another on the basis of their charge aswell as size, lysates from VV:gag-infected H9 cells labeled with ['Hlleucine were immunoprecipitated with anti-p24, anti-pl7, and anti-p6 antibodies, and the immunoprecipitates were subjected to NEPHGE according to O'Farrell et al. (11).The results are summarized in Fig. 2. The p55 precursor precipitated by all three antibodies migrated as a single species to a pH of 7.0 (Fig. 2, A, B, and C). p24 migrated as three distinct species between pH 6.2 and 6.8 (Fig. 2 A ) . A minor p25 protein precipitated by anti-p24antibodyalso resolved intothree species identical in charge to p24 (Fig. 2 A ) . Since the p24 protein is reported tobe phosphorylated (8,9), the migration of p24 and p25 as three species could be due to differential phosphorylation. The p41 (a and b) intermediate proteins precipitated by the anti-p24 antibody migrated as multiple species between pH 6.5 and 7.4 (Fig. 2 A ) . The p39 protein migrated as multiple species between pH 6.5 and 6.9 (Fig. 2A ). As shown in Fig. 2B, the p41a intermediate precipitated by the anti-pl7 antibody migrated as multiple species between pH 6.5 and 7.0. The migration of the p39 intermediate was essentially identical to that of the p39 species precipitated by the anti-p24 antibody (compare Fig. 2, A and B ) . The anti-

(Fig. I A , lone 2). The 160-kDa ( ~ 1 6 0gag-pol ) fusion protein (10) precipitated by anti-p24 was visible after prolonged auA toradiography (data not shown). Anti-p24 antibody also precipitated a minor 25-kDa (p25) protein (Fig. l A , lane 2) which has alsobeen seen in HIV-1-infected cells (9).' Thisp25 could represent a post-translational modification or an intermediate of 1124 resulting from alternate cleavage at the COOH-terminal p24/p15 cleavage site. As shown in Fig. lB, the anti-pl7 antibodyprecipitatedthe p160, p55, and p17 proteinsas 25/24 expected. Surprisingly, it also precipitated proteins of 41 and - I7 39 kDa (Fig. 1H, lane 2). Three different anti-pl7monoclonal antibodies gave identical results (data not shown). Two minor proteins(approximately 45 and 49 kDain size) were also precipitated by anti-p24 and anti-pl7 antibodies (Fig. 1, A and H ) . Although a 45-kDa protein has been observed in HIV1-infected cells, the origin of this and the 49-kDa protein is unknown. As shown in Fig. lC, anti-p6 antibody (7) precipitated only a p41 intermediate protein in addition to p55 and putative pl.5 (Fig. lC, lane 2). The p6 peptide did notresolve in 9-14% gradient gels (Fig. IC, lane 2) but was seen in 10-20% gradient gels (data not shown). Thep160 species precipitated by anti1 1 1 1 1 1 1 1 1 1 1 1 pfi antibody (Fig. 1R) was visible only after prolonged auto11.0 1 . 6 1 . 2 6 . 8 6 . 4 8 .60. 0 1.6 7 . 2 6.11 6.4 6 . 0 PH radiography (data not shown). The anti-p6 antibody did not FIG. 2. NEPHGE analysis of HIV-1 gag proteins expressed precipitate the p.79 protein (Fig. IC, lane 2). Although un- in VV:gag-infected cells. Infection of H9 cells with VV:gag, radilikelv, we cannot rule out the possibility that epitopic differ- olabeling, and immunoprecipitationwere performed as described (10). ences account for the inability of anti-p6 antibody to precip- The immunoprecipitates were eluted with isoelectric focusing buffer itate p89. Nonimmune rabbit serum (Fig. l A , lane I), isotype- ( 5 mM KPOI, 2% Triton X-100, 5% if-mercaptoethanol, 9 M urea) matched control monoclonal antibody (Fig. 1B, lane I), and and subjected to NEPHGE according to O'Farrell et al. (11). The proteins were focused for 4.5 h a t 400 V. The separation of the goat anti-mouse antibody (Fig. lC, lane I ) did not precipitate proteins in the seconddimension was performed in 9-14% linear any of these proteins, confirming specificity. The anti-p24, gradient gels. Proteins elutedfrom ( A )rabbit anti-p24resin incubated anti-pl7, and anti-p6 antibodies did not precipitate any of with VV:gag lysate, ( R )anti-pl7 resin incubated with VV:gag lysate, these bands in radiolabeled wild t-ype vaccinia virus-infected ( C ) goat anti-mouse IgG resin incubated with VV:gag lysate previously treated with anti-p6 monoclonal antibody (7) for 16 h a t 4 "C, H9 cells (data not shown).Theseresults suggest that the initial cleavage of the p.55 gag precursor is random and gives and ( D )control antibody resin incubated with VV:gag lysate. Following the electrophoresis, the tubes loaded withisoelectricfocusing rise to three (p41a, p41b. and p39) processing intermediat,es. huffer alone were sliced into 1-cm pieces, equilibrated with 3 ml of

.

~

~

~~~

' H. S.Stein. unpul)lished results.

~

distilled water for 16-18 h, andthepH determined.

value of each slice was

Characterization of H I V - 1 gag Processing Intermediates p17 antibody precipitated two proteins migrating in the 17kDa size range(Fig. 2B). While the slower formmigrated predominantly as a singlespecies to pH 7.3 (Fig. 2B), the faster migrating protein resolved into multiple species spreading between pH 6.9 and 7.3 (Fig. 2B). The relationship between these proteins is unknown. As shown in Fig. 2C, the p41b intermediate recognized by the anti-p6 antibody migrated as a single species to pH 7.4. The migration of this intermediate species coincided exactly with the more basic species precipitated by anti-p24 antibody (compare Fig. 2, A and C). This species was not precipitated by the anti-pl7 antibody (Fig. 2B). The p15 and p6 proteins did not resolve in this gel (Fig. 2C).Fig. 2 0 is a representative nonspecific control confirming the specificity. Twominor proteins (approximately 45 and 49 kDa in size) precipitated by anti-p24 and anti-pl7 antibodies were more basic than p55 precursor and both migrated between pH 7.2 and 7.3 (Fig. 2, A and B ) . These proteins were not precipitated by anti-p6 antibody (Fig. 2C). Theorigin of these proteins are unknown. These results confirm the existenceof two p41 (41a and 41b) intermediate proteins, one containing p17 and p24 protein determinants(41a)andtheothercontaining p24 and p6 protein determinants (41b). The p39 intermediate contains only p17 and p24 determinants. Recently, Schneider and Kent (13) identified two cleavage sites for the HIV-1 protease defined at the Met-Met (376377) and Leu-Ala (362-363) sites at the junction between the p24 and p15 structural proteins (13). The cleavage of p15 from the p55 precursor at these COOH-terminal sites would result in two intermediates varyingin length by 14 amino acids. The subsequent cleavage of these two intermediates at the junction of p17 and p24 would result in two p24 core proteins differing by 14 amino acids. In accordance with these predictions, two intermediate proteins (p41a and p39) recognized by anti-pl7 and anti-p24 antibodies were identified in cells infected with VV:gag (Figs. 1, 2, and 3). We have also identified both p41 and p39 intermediates in HIV-1-infected cells.' Furthermore, two core proteins, a slower (p25) and a faster(p24)migrating species,havealsobeen detected in VV:gag-infected cells (Figs. 1A and 2 A ) as well as HIV-1infected cells (9).2However, mature virions contain only the

*.

U__"_~ ' I

~ _ _ _ ._ J

..> I.

a:'.:

. . +E : -1

K - 2

. ....

.. .... :

" "

~

:,? .:

I.,,

846 1

1

::' "3

,

~

+?L-57.:-

I

.

,: !