Gag proteins of the highly replicative MN strain of human immunodeficiency virus type 1: posttranslational modifications, proteolytic processings, and complete amino acid sequences.

(1)

Gag Proteins of the Highly Replicative

MN

Strain of

Human

Immunodeficiency Virus Type 1: Posttranslational Modifications,

Proteolytic Processings, and Complete Amino Acid Sequences

LOUIS E.HENDERSON,1* MICHELLE A. BOWERS,' RAYMOND C. SOWDER II,' STEFAN A.

SERABYN,1

DONALD G. JOHNSON,' JULIAN W. BESS, JR.,' LARRY0. ARTHUR,' DUNCAN K. BRYANT,2

ANDCATHERINE FENSELAU2

AIDS Vaccine Program, Program Resources, IncorporatedlDynCorp, National CancerInstitute-Frederick Cancer Research and Development Center,

Frederick,

Maryland21702-1201,1 andDepartmentof Chemistry

and Biochemistry, University of Maryland, Baltimore County, Baltimore, Maryland 212282 Received 3 September 1991/Accepted 17December 1991

The MN strain of human immunodeficiency virus type 1 was grown in H9 cells, concentrated by

centrifugation, and disrupted, and proteinswerepurified byreversed-phase high-pressure liquid

chromatog-raphy. Complete amino acid sequences were determined for the mature Gag proteins, showing natural proteolytic cleavage sitesand the orderof proteins (pl7-p24-p2-p7-pl-p6)in the Gagprecursors.At leasttwo

sequencevariants ofp24 and eightsequencevariants ofp17weredetected. Thetwo mostabundant variantsof

p24 andp17 representedatleast 50%Yo 5% and20% ± 5%oftheirtotals, respectively. These datasuggest

heterogeneityin thevirus population,with 50% of the total viruscontainingthemostabundantforms of p17

and p24 and 20% of the virus containing the second most abundant forms. The Gag precursors ofthese suggested viruses differ from each otherby only3amino acid residuesbut differ from theprecursorspredicted

by the published MN proviral DNA sequence by 10 residues. Electrospray ionization mass spectrometry

analysisof the purified p24forms showed that the measured molecularweightof theproteinwas200 ± 50

atomic mass units greater than the calculated molecularweight. The source ofadditional massfor the p24 forms was not determined, but the observation is consistent with previous suggestions that the protein is phosphorylated. Greater than 98% of the total recovered p17 was myristylated at the N-terminal glycine

residue,and the measured molecularweights (as determined byelectrosprayionizationmass spectrometry)of themostabundant formswerewithin 3atomic mass units of the calculated molecularweights (15,266). Human immunodeficiency virus type 1 (HIV-1) is the

etiologic agent ofAIDS (1, 25, 29), and the MN strain of HIV-1 (HIV-lMN) (13) has been shown to be a prevalent

prototypeinthe U.S.population,bothbyserologicalstudies (7)andbyvirusisolation andsequencingofthe V3regionof theenv gene(23). Because of itsprevalence, HIV-lMNhas beensuggestedastheprototypevirus for thedevelopmentof vaccines and other antiviral strategies. Standardized stocks of this strain will be neededaschallengereagentsfortesting the effectiveness of various developed procedures; it is highly desirable that these stocks be well characterized. Challenge stocks ofHIV-lMNgrown in H9 cells have been prepared andarebeingevaluatedat thebiologicallevel. At themolecularlevel,atleastonecompleteinfectiousprovirus

has been isolated from cells infected withHIV-1MN, and its nucleotide sequence has been determined (28). However, infected cells generally contain multiple sites of proviral integration, and it is difficult to determine which of the integrated proviruses produces the mostabundant virus in tissue cultures. One method for characterizing the most abundant virus istoisolate viralproteinsand determine their amino acidsequences.

The HIV-1viralgenomecontains threelargeopenreading

frames, designatedgag,pol,andenv,that code forproteins

ultimatelydestined forincorporation into thematurevirus. Gagand Polproducts togetherwith RNAmakeup the viral core, and Envproducts arepart of the lipid envelope. Pol

*Corresponding author.

proteins are enzymes necessary for viral replication. Gag

proteins are often referred to as structural proteins and

representamajor portionof thecorestructure.Theproteins analyzedhere areproductsofthegag genepurified directly fromHIV-lMNgrownin tissue cultures. Theprimary trans-lationalproductof thegag gene isapolyproteinprecursor,

designated Pr55Fag, that is ultimately cleaved by the viral protease tothematureGag proteinsfound in the virus. Here weshow thatPr559agofHIV-1MNis cleavedtosixproducts, including the matrix antigen (MA), core antigen (CA),

nu-cleocapsid (NC) protein,and threepeptides, designatedpl, p2,andp6,and the order oftheproductsin theprecursoris MA-CA-p2-NC-pl-p6, aspreviously suggested (15, 18).We reporton the purification andcomplete chemical structural analysisof the sixmatureGag proteins purifieddirectlyfrom HIV-1MNviralparticles.

MATERIALS AND METHODS

Viruspreparation.H9cellschronicallyinfected with

HIV-1MNwere

provided by

R. C. Gallo

(Laboratory

of Tumor

CellBiology, Division ofCancerEtiology,NationalCancer Institute, Bethesda, Md.) and were grown in roller bottles

underbiosafetylevel 3laboratoryconditionsatthe National Cancer Institute-Frederick Cancer Research and

Develop-ment Center, Frederick, Md. Virus production was in-creasedapproximately50-foldby supplementinginfectedH9 cells withanequalnumber of uninfectedH9cellsafter each harvest (2). Virus was harvested from the culture medium

every 3 to 4 days, isolated by continuous-flow sucrose

1856

on November 9, 2019 by guest

http://jvi.asm.org/

(2)

(saturated at room temperature), the pH was adjusted to 8.5 by the addition of concentrated Tris hydrochloride buffer,

and 2-mercaptoethanolwas added to a final concentration of

2% (vol/vol). The resulting clear solution was transferred to a clean sterile tube and removed from the biosafety level 3 environment.

Protein and peptide purification. All samples for reversed-phasehigh-pressure liquid chromatography (rp-HPLC) were dissolved in saturated guanidine hydrochloride containing 2%2-mercaptoethanol at pH 8.5 and sonicated at

50°C

for 15 min before injection into the chromatographic column. Sep-arations were performed on,uBondapak C18 (Waters Asso-ciates) rp-HPLC supports by use of a liquid chromatograph (Pharmacia LKB) equipped with a rapid spectral detector (model 2140). Elutions were accomplished with 0.05% (vol/

vol) trifluoroacetic acid at pH 2 and with a gradient of increasing acetonitrile concentrations. Eluted proteins and peptides were detected by UV absorption at 206, 280, and 294 nm and collected, and the solvents were removed by lyophilization.

Endoproteinase digestion. Proteins to be digested were dissolved in 50 mM NH4HCO3 at pH 8 to a final concentra-tion of 1 mg/ml. When a protein was initially insoluble, the pH was raised to 11 (concentrated NaOH) to solubilize the protein and thenadjusted to 8 to 9 (concentratedHCl) before addition of the enzyme and buffer. Digestions were initiated by adding theendoproteinase to a final concentration of 0.02 mg/ml and were continued for 18 h at room temperature. When a precipitate developed during the course of the digestion, the process was repeated for a second 18-h period. Digestions were stopped by adding concentrated hydrochlo-ric acid to a final pH of 2, and the peptides were separated by rp-HPLC on ,uBondapak C18 columns (3.9 by 300 mm). The endoproteinases used were endoproteinase Glu-C, endopro-teinase Lys-C, and trypsin (Boehringer, Mannheim, Germa-ny).

Amino acidanalysis. Approximately 1 ,ug of purified pro-tein or peptide washydrolyzed in 200,ulof 6 M hydrochloric acidcontaining 0.1% phenol in vacuo at 110°Cfor 24 h. The samples were dried in vacuo and analyzed on a Beckman model 6300 automated amino acid analyzer in accordance with themanufacturer's recommendations.

Gel electrophoresis. Sodium dodecyl sulfate-polyacryla-mide gel electrophoresis (SDS-PAGE) was performed as

previously described (22).

Edmandegradation. Automatic N-terminal Edman degra-dationswereperformed onapproximately 1-nmol samples of purified protein or peptide by use of a pulsed, liquid-phase protein sequencer (model 477A; Applied Biosystems Inc.) equipped with an on-line phenylthiohydantoin analyzer (model 120A) inaccordance with themanufacturer's recom-mendations.

FAB mass

spectrometry.

Fast atom bombardment (FAB) massspectrometry of Gag proteins was performed by use of the firsttwosectors of anHX110/HX110massspectrometer (JEOL, Tokyo, Japan) at an accelerating voltage of 10 kV,

tween the two mass spectrometers (4), with helium used as the collision gas at a pressure sufficient to attenuate the precursor ion beam by80%. The collision cell was floated at 4kV. The collision-induced dissociation spectra were re-corded at a resolution of 1:1,000 and with 100-Hz filtering. For both scanning modes, 50 to 100 pmol of peptide in 0.1% trifluoroacetic acid (1

p,l)

and 3-nitrobenzyl alcohol (1

p,l)

were used for the analysis.

The computer program RESIDUES, written by David Heller (Middle Atlantic Mass Spectrometry Center, Johns Hopkins University, Baltimore, Md.), was used to assist with the interpretation of tandem mass spectrometry data, while the computer program MacProMass (written by Terry D. Lee and Sunil Vemuri, Beckman Research Institute, City of Hope, Duarte, Calif.) was used to identifyp17 proteolysis peptides in the conventional magnetic scans.

ESI-MS.Electrospray ionization mass spectrometry (ESI-MS) was performed with a Vestec electrospray source fitted to a Hewlett-Packard 5988A quadrupole mass spectrometer. Sample p17 (from pool A) was dissolved in an aqueous solution containing 1.5% acetic acid and 50% methanol. This solution was introduced into the source with a Harvard 22

syringe infusion pump.

RESULTS

Isolation of HIV.lMN proteins. As part of an ongoing program to characterize HIV-lMN and to prepare challenge stocks for future studies, the virus was produced in the AIDS Vaccine Program virus production facility at the National Cancer Institute-Frederick Cancer Research and Development Center by cocultivating infected and unin-fected H9 cells. Compared with tie amount of virus pro-duced from H9 cells chronically infected with

HIV-1MN,

this cocultivation method results in approximately a 50-fold increase in the amount of virus 7ecovered and has been adopted as a standard production method (2). Concentrated purified virus was disrupted and solubilized in saturated guanidine hydrochloride (pH 8.5) containing 2-mercaptoeth-anol, and the resulting clear solution was injected into a high-pressure liquid chromatograph for separation and puri-fication of the viral proteins. Figure 1 shows a typical chromatogram for the rp-HPLC separation of the proteins

associ.ted

with the purified virus. Each peak of eluted protein was analyzed by SDS-PAGE, amino acid analysis, and N-terminal Edman degradation to qualitatively identify the purified proteins. Many of the peaks were found to contain cellular proteins, and these will be discussed in a separate communication (18a). The peaks containing purified HIV-lMN Gag proteins p17, p24, p2, p7, pl, and p6 are indicated in

Fi,,.

1. Protein in each of the labeled peaks (Fig. 1) was analyzed by SDS-PAGE, and the results for peaks p7, pl, p6,pl7athrough pl7e, and p24 are shown in Fig. 2. Peak pl7f was not included in this gel but yielded results indistin-guishable from those yielded by pl7e (data not shown); peak p2 yielded no detectable bands in the SDS-PAGE analysis

on November 9, 2019 by guest

http://jvi.asm.org/

(3)

E

C w

(0

0

400 800 1200 1600

Minutes (5.0 mllmin)

1923

FIG. 1. Protein separation by rp-HPLC. HIV-1MN virions were disrupted, and the viral proteins and peptides were separated by preparative rp-HPLC.Approximately175 mg ofpurifiedHIV-1MN wasdissolvedand reduced in 40 ml of saturated guanidine hydrochloride containing2% (vol/vol) 2-mercaptoethanolandinjectedintoa,uBondapak C18 column (19 by150mm). Proteinsandpeptideswereeluted with agradientofincreasingacetonitrile concentrations atpH 2.0(0.05% trifluoroaceticacid)and detected by UV absorption at 206 nm. The UV peaksassociated withpurifiedviralproteinspl7a throughp17f,p24, p2, p7, pl, and p6 are indicated. The pool of p24 taken for further analysis is indicatedby a bracket. OD, optical density.

p24- p17-p6+p7

-14 6.5

a b c d e

pl 7s

z

FIG. 2. Coomassie blue-stained gel of purified proteins.

HIV-1MNproteins purified as shown in Fig. 1 were analyzed by

SDS-PAGE on6 to 18% polyacrylamide gradient gelsand detected by staining with Coomassie brilliant blue R-250(18). Lane HIV-1MN

contained wholedisrupted virus,and lanesp7,pl, p6, p24,andpl7a through pl7econtainedapproximately3to5,ugofprotein purified

inthecorresponding peaklabeled inFig. 1.

(datanotshown).These results(Fig. 1 and 2) show that pl,

p2, p6, and p7 are eluted as highly purified proteins in well-separated and nearly symmetrical peaks. p24 is also elutedas ahighly purified protein,butthepeakis asymmet-rical. The trailing edge of the p24 peak contained

contami-nating proteins (visualized by SDS-PAGE analysisof

indi-vidual fractions) and was omitted from the pool of highly

purified p24 (bracketmarked Pool inFig. 1)taken for further

analysis. On the basis of UVabsorption,we estimated that

thepool containedatleast70% of the total p24 eluted in the

asymmetrical peak. In contrast, p17 was eluted from the

column as a group of partly separated peaks labeled a

through f, as shown in Fig. 1. The protein in each peak

migratedas asinglebandatabout 17 kDa in theSDS-PAGE

analysis(Fig. 2)andreacted withapolyclonalrabbit

antise-rumprepared against theprotein purifiedinpeakaand also

with a mouse monoclonal antibody to HIV-1 p17 (data not

shown). Of the total p17 eluted from the column, we

esti-mated that 66%was eluted inpeak pl7a, 5%waseluted in

peak pl7b, 9% was eluted in pl7c, 9%was eluted in peak

pl7d, 8%waseluted inpeakpl7e, and 3%waseluted inpeak

pl7f. Tobetter understand the natureof the apparent chro-matographic heterogeneityof the p17forms and to charac-terize purified pl, p2, p6, p7, and p24, we subjected the proteins to complete amino acid sequence analysis and examined them for posttranslational modification by mass

spectraland chemical methods.

The protein in peak pl7a (Fig. 1) was digested with

endoproteinase Glu-C, and the peptides wereseparated by

rp-HPLC (Fig. 3). Each purified peptide was analyzed for amino acidcomposition andamino acid sequenceby N-ter-minal Edman degradation and mass spectrometry. The re-sults are summarized in Fig. 4, in which the determined

on November 9, 2019 by guest

http://jvi.asm.org/

[image:3.612.149.486.73.321.2] [image:3.612.82.298.419.660.2]

(4)

E(

is~~~~~~~~~~~~~~~~~~~~~ICbJ

40 80I1201

o 0~~~~~~~~~~~~~~~c

I-co~~~~~~~~t

ejmIlLO 0

40 80 120 160 200

[image:4.612.123.488.74.346.2]

ml

FIG. 3. Separationofpeptidesresultingfromendoproteinase Glu-Cdigestionofpl7a.Theproteinpurifiedinpeakpl7a(Fig.1 and2)was

digestedwithendoproteinaseGlu-C,andthepeptideswereseparated by rp-HPLC.Elutedpeptidesweredetectedby UV absorptionat206 nm.Thepeptidematerialin each UVpeakwasidentifiedbyaminoacidsequencing (see the text),andpeptidesarenumberedaccordingto thepositionsoftheirfirstand last residues in the amino acid sequencealignment showninFig.4.Peptidesrepresenting identical regions of p17(Fig. 4) butdifferingby amino acid substitutions aredistinguished by the lettersa and b (seethe text). Peptides containing cysteine residues(positions56 and86; Fig. 4)wereisolated intheir monomeric form,indicatedby(1),and intheirdisulfide form, indicated by (2). The peptide designated (55-73)+(74-98) is a heterologous disulfide dimer. When two peptides coeluted in the samepeak, both peptides are

indicated andseparated byacomma.OD, opticaldensity.

amino acid sequences for thepeptides arecomparedwitha

predicted amino acid sequence based on the nucleotide

sequence of codons 2through 135 of theHIV-1MN proviral

gag gene(28).Thealignmentshows that the determined and

predicted amino acid sequencesarein agreementfor 130 of

134 compared positions and prove that the protein in peak

pl7a(Fig. 1)is derived from the N-terminal end of the Gag

precursor. These data also show that the protein in peak

pi7a

is amixture ofat least twoclosely related p17forms that differ from each other by substitutions of valine for

isoleucine at positions34 and 46.

The two forms ofp17 present in peak pi7a (Fig. 1) are revealed by the data in Fig. 3, which shows sets of two

peptides that differ from each otherbya single amino acid

residue. Onesetiscomposedofpeptides17-39a and 17-39b

(Fig. 3).Thesepeptideshave identical amino acid sequences

corresponding to positions 17

through

39 in Fig. 4, except

thatpeptide 17-39a has isoleucine atposition34 andpeptide

17-39bhasvalineatposition34.Anothersetiscomposed of

peptide 42-Sla, with a valine at position 46, and peptide

42-51b, with an isoleucine at position 46. The peak

desig-nated 74-98(1),42-51b in Fig. 3is amixture containingboth

peptide 74-98 and peptide 42-51b. Edman degradation

re-vealedtwo residues at each step, confirming thesequences of the twopeptidesandindicatingthatpeptide74-98 repre-sented greater than50% of the totalpeptide in themixture.

In addition, mass spectral analysis of peak 74-98(1),42-51b

(Fig. 3) confirmed the sequence of peptide 42-51b. The

determined mass spectra of peptide 42-51b and peptide

42-Sia are shown in Fig. 5. The two spectra are identical,

exceptthat themassesofionsd4to

dg

and a4to

ag

in 42-Sia

(Fig. Sa) are14 massunits greaterthan those of the

homol-ogousions in42-Sib(Fig. Sb),confirmingthesubstitution of valine for isoleucine. The amount ofpeptide 17-39a recov-ered is about four times the amount of peptide 17-39b

recovered(seepeakheightsinFig. 3). Similarly,theamount

ofpeptide42-51a recovered is about four times the amount

ofpeptide 42-Sib recovered(Fig. 3). On thebasis of these

recoveries,weconclude that about 70to80%of theprotein

in pl7a (Fig. 1) has isoleucine at position34 and valine at

position 46 and that 20 to 30% of the protein hasvaline at

position34andisoleucineatposition46; together, thesetwo

formsmake up atleast66%of the totalp17eluted from the column. Inaddition,the determinedamino acid sequencesof

thesetwoforms differ from thepredictedsequencesatthree

otherpositions, atwhich theproteins havelysine in placeof asparagine(position 17) (peptides 17-39a and17-39b),

argin-ine in place of leucine (position 74) [peptides 74-98(1) and

74-98(2)], and lysine in place ofglutamic acid (position 92)

(peptide 74-98). Since both the 17-39a and the 17-39b

pep-tides havelysineatposition17 andnootherforms ofpeptide

74-98 were detected, we conclude that both forms of p17

present in thepl7a peak (Fig. 1)havelysineatpositions 17 and 92 and arginineatposition74.

Peptide 1-11 (Fig. 3) had an amino acid composition consistent with thepredictedsequence (based onthe DNA

on November 9, 2019 by guest

http://jvi.asm.org/

(5)

10 20 30 40 50 60

[image:5.612.160.472.68.336.2]

---l---KR---I---I"I--V---

|l---KV----'''!'--I

K---l---.I---N---I---I----

--!

!---' ---v----'

----N

.---I.---.

---IK---1

pool

] A B

I C

I--- RK---I----I I--V--- i_ --- D

K .

i

~~~~-K---I---',--I---

11---70 80 90 100 110 120 130

SIQTGSEELKSLYNTVATLYCVHQKIEITKEALEKIEEEQNKSKKKAQQAAADTGNRGNSSQVSQNY

K.---I'""

I1 ]

I

-K---K_I-

'-'__----' '-1--'

---K.---R

K----

. -R

-R---E D

l---K---Hl---l

---K----FIG. 4. Amino acidsequenceofp17.The amino acid sequences of thechromatographic forms ofp17purifiedasshown inFig. 1(forms pl7atopl7d)weredetermined byanalysis ofpeptidesasdescribed in the legendtoFig. 3. The deduced sequence based on the proviral DNA sequenceis shown in boldface type inthetopline. The determined aminoacidsequencesof the peptides are shown in pool A for pl7a, pool Bfor pl7b,pool C forpl7c,andpoolD forpl7d.Dashes indicate positionsatwhichthe deduced and determined sequences agree.

sequence) butwas "blocked" to N-terminal Edman

degra-dation. The peptide sequence was determined by mass

spectralanalysis.All sequence ions derived from the amino

terminus of thepeptide were 211massunits greater than the

sum of their amino acid residues. In addition, the spectra

showedamyristyl-Gly

[CH3(CH2)12CONHCH2CO']

ion at 268mass units and amyristyl

[CH3(CH2)12CO']

ion at 211 mass units, as previously observed for other myristylated

peptides (16, 19). The N-terminal peptide eluted late from

the column, consistent with the covalent attachment of a fattyacid,andwasimmediately preceded byanotherpeptide

(unlabeled in Fig. 3). Analysis of this preceding peptide

revealed that itwas apartial cleavage productconsisting of residues 1 to 16 and was also myristylated. No other

peptidesrecovered from thedigestcontained the N-terminal

residue. The molaryields of the N-terminal peptides (1-11

plus 1-16) were equal to the molar yields of the other

peptides

recovered inFig. 3. Therefore, weconcluded that

at least 95% of the protein in peak pl7a(Fig. 1), including

both forms ofp17, contained the N-terminal myristyl moi-ety. Except for disulfide-bonded peptides (Fig. 3)that may haveformed during the digestion, noother posttranslation-allymodified forms of theproteinweredetected.

Thetechniques (endopeptidase digestion, peptide

separa-tion andanalysis byEdmandegradation, andmassspectral

analysis)usedabove for theanalysisofproteininpeakpl7a

wereused for theanalysisofprotein purifiedinpeaks

pi7b,

pl7c, andpl7d (Fig. 1),and theresultsof theseanalyses are

alsosummarized inFig. 4.These results confirmedthatthe proteins purified inthepeakswereviralp17but alsoshowed that eachpeakcontainedamixture ofmultiple forms of the

protein. Peakpl7bcontained the forms ofp17 identified in

peak

pl7a

plus a form of the protein with asparagine at

position17and isoleucineatposition34. Inaddition, peptide

93-98was identified inpeakpl7b, suggestingthat aportion

of the proteincontainedglutamic acid at position92.

Anal-ysis of peak pl7c revealed additional forms with leucine,

serine, orasparagine at position74. The leucine at position

74 is in agreement with the nucleotide sequence, but the

peptidealso containedlysineatposition92, indisagreement

with the nucleotide sequence. The recoveries ofpeptides

from the analysis of peak pl7c suggested that this peak

contained a form of the protein with valine residues at

positions 34 and 46, in agreement with the nucleotide

se-quence. Peakpl7calsocontained forms of theproteinwith

glycine at position 89 and glutamine at position 104. Peak

pl7dcontained additional forms of theproteinwithlysineat

position 108and possibly glutamicacid atpositions 74, 84,

and 92(also indicated inpl7b).

Thetotalnumber of amino acid sequencevariantsof

p17

present in the virus population cannot be estimated from these results, because some variants may be present in amountsbelowourlimits of detection. However, the

analy-sis of the variouschromatographic forms of

p17

revealedat least eight different amino acid sequence variants of the protein. In addition, the proteins in peaks pl7e and

pl7f

werenotanalyzed butprobablycontained additional variant forms. Nevertheless, at least one-half of the total

p17

is contributed by the major form in peak pl7a, and about one-fifth of thetotal

p17

iscontributedby theminor formin pl7a. None of the forms identified were found to be in complete agreement with theproteinsequencededuced from

theproviralDNAsequence ofHIV-1MN.Atleast98% of the

protein analyzed contained the N-terminal myristyl

post-translationalmodification,butpeptidesthatlacked the

mod--

---_I

ZV-

---

on November 9, 2019 by guest

http://jvi.asm.org/

(6)

* 40

u

a3

A.1-d5

L F

1. b,

a3

as

11 .1

I. I

.

a7

a8 dg

a9

a

1.11

200 400 6Be BOB 10B0

m/z

[M+H]H

Arg

Phe

lAla

Val Asn Pro Gly Leu Leu Glu J

R

II,

as

F

b,

a8 d9 ag

-1

.LI a

dUim6IWIA&h i a,I~Ii& I1IAL,r II.L IL

a ..

288 408

. I . I

689

a7

m/z

FIG. 5. Tandemmassspectraofpeptides 42-51a and 42-Sib(Fig. 3). Both peptidesrepresentpositions42to51 inFig. 4; residue 45 is isoleucinein42-51a(spectrum a) and valine in42-Sib(spectrum b). The fragment ions that contain residue 45are14massunits heavier in

spectrumathan in spectrum bbecauseofthepresenceofthe isoleucineinplace ofavaline.

ified N-terminalglycineresiduewerenoted in theanalysisof

pl7bandpl7c.

ESI-MS ofpl7.Theprotein purified inpeak pl7a (Fig. 1)

was analyzed byESI-MS. This isa highlyaccuratemethod for determining the molecular weight of a protein and is useful forverifyingamino acidcompositionsofproteinsand fordetectingthepresenceofposttranslationalmodifications presentbeforeproteolysis (26). Figure6 showstheESI-MS spectrum obtainedfrom the analysisofpl7a (Fig. 1). Anal-ysisof the iongroupswith differentcharged statesrevealed a molecularweight of 15,269.3 mass units for the protein. This value is in near-perfect agreement with the molecular weightof15,266.3massunits calculated from thedetermined

structuresof bothforms ofp17 (Fig. 4)present inpeak p17a (Fig. 1). Close agreement between the determined and calculated molecularweights strongly suggests that thep17 analyzed here (pl7a) is posttranslationally modified bythe addition of an N-terminal myristyl group but contains no other modifications. However, this conclusion cannot be extendedtothe otherchromatographicforms ofp17 (pl7bto

pl7f; Fig. 1), since they were not analyzed by ESI-MS. Phosphorylated forms of p17 have been detected by 32P

labeling of

HIV-lvAv

viral proteins (27), and it is quite possible that one or more of the minor chromatographic forms ofp17 contain phosphorylatedresidues. The finding that thepredominantforms ofp17arenotphosphorylatedis consistent with an earlier analysis of p17 purified from

HIV-1IIIB in which we did not detect the presence of phosphorylatedamino acids(15). Our previous analysiswas done with nonradioactive methods and would not have detected phosphoproteins present in <20% of the total protein analyzed.

Apoolofhighly purified p24 (markedinFig. 1) containing

at least 70% of the total p24 eluted from the columnwas

digested with endoproteinase Glu-C or trypsin. For each

digest, the resulting peptides were separated by rp-HPLC

(datanotshown)and analyzedasdescribed for thevarious p17 forms. The determined amino acid sequences were

aligned with the amino acid sequence predicted by the proviral DNA sequence (Fig. 7). Only the relevant and

(b)

8_

c

"a 68

.0

> 41

Cu L1

R

I I 11 -1 L AlI i..ai

I

.

NM

AA Ak1

11

s

ZSi

1-I

rs

+

1-a6

I

on November 9, 2019 by guest

http://jvi.asm.org/

[image:6.612.119.494.76.470.2]

(7)

+13

+14

+15

+16

11. 1.|11A

MWfound=15269.3

+12

I.

+11

-J IIa* JiL,....LhMa.LELI.I.A,&ll"Milp*.l NM 1000 1200 1400

nmlz 1600 1800

FIG. 6. ESI mass spectrum of pl7a (Fig. 1). Each peak in the envelope is a molecular ion with the indicated charge state. A mass

spectrometermeasuresmass/charge; hence,themolecularweight ofalarge analyte, suchasp17,canbedetermined if the analyte hasahigh

chargestate(26). These datayieldadeterminedmolecularweight of 15,269.3 for pl7a (Fig. 1).

nonredundant peptides are shown. These analyses revealed

twoamino acid sequencevariants ofp24. Thepredominant form represented about 70% of the total protein analyzed and contained alanine at position 86; theother form repre-sented about30%of the totalproteinanalyzedand contained valineatposition86. Both forms differed from thepredicted amino acid sequenceat threeotherpositions: position 7, at which theprotein containedglutamine in place ofglutamic acid; position 184,atwhichtheproteincontainedtryptophan in place ofarginine; andposition 185, atwhich the protein containedmethionine inplaceof threonine.

Protein in the p24 pool (Fig. 1) was also analyzed by

ESI-MS, andamolecularweight of25,752 + 50was deter-mined (datanotshown). Thecalculatedweightsfor the two

forms ofp24 (Fig. 7)were25,551 (alanineatposition 86)and 25,579 (valineatposition 86). Thus,themeasured molecular weight is about 200 greater than the calculated molecular weight. Previous reportshave shown that HIV-1 p24 (from strain IIIB or BRU) contains phosphoserine and

phospho-threonine (15, 27) and exhibits multiple isoelectric forms containing phosphate (24). The ESI-MSdatadonotdirectly determine phosphates; however, the results of the analysis

are consistent with the previous proposals that the p24

protein isphosphorylated. The observedvariation in molec-ularweight (±50)isgreaterthantheprecisionof the ESI-MS method(26)and is inpartduetothesequenceheterogeneity in thesamplebut also couldsuggestthat thesamplecontains a mixture of diphosporylated and triphosphorylated

pro-teins. Unfortunately, mass spectral analysis of the proteo-lytic peptidesderived fromp24 (asinFig. 7)failedtoreveal any evidence of posttranslational modifications, including phosphorylations. Itseemspossiblethatputative phosphate

groups were lost through hydrolysis during the enzymatic

digestionofthe protein orseparationof the peptides. The otherGag proteins purifiedinFig. 1 (pl, p2, p6, and p7) were also analyzed by Edman degradation of purified

peptides and mass spectral analysis of peptides (data not

shown). The results showed that the amino acid sequences

1 10 20 30 40 50 60 70

PIVQNIEGQMVHQAISPRTLNAWVKVVEEKAFSPEVIPKFSALSEGATPQDLNTIIMTVGGHQAAMQNLKETINEEA

j Q---I- I---

---_____Q---

l-0-80 90 100 110 120 130 140 150

AEWDRLHPVHAGPITPGQNREPRGSDIAGTTSTLQEQIGWNTNNPPIPVGEIYKRWIILGINICVRNYSPSSILDIR

- --|j--A--

_{--V---A----I---}

---A----1 ---1 ---H-H---'

---A----A-160 170 180 190 200 210 220 230

QGPKEPFRDYVDRFYKTLRAEQASQEVINRTTETLLVQNANPDCKTILKI.IQPAATLEEMKTACQGVGGPGHKARVL

---I II II--- WM---H

---FIG. 7. Amino acidsequenceofp24. Peptidesderived fromtrypsinandendoproteinaseGlu-Cproteolytic digestionsofp24 (Fig. 1)were

isolatedby rp-HPLCandsequenced byEdmandegradationandmassspectral analysis.The determined amino acidsequencesarealignedwith

the sequencededuced from the proviral DNA sequence shown in boldface type in the top line. Dashesindicate locations at which the

determinedsequences agreewith the deducedsequence.Two variantsofpeptide83-96werefound(aminoacidsequencevariations atposition 86). Themostabundant form ofpeptide83-96(approximately70% of thetotal) contained alanine atposition86.

100-

90-

80-

70-

50-402

302

202

101

_I III EL ME1111111-1_MNL&M

1--WN-H

---i

I.M- ti

i.

I-1

.11.

on November 9, 2019 by guest

http://jvi.asm.org/

[image:7.612.145.491.73.277.2] [image:7.612.156.479.551.677.2]

(8)

75

---pl7--- --_---________---________--_--Myr-GARASVLSGGELDRWEKIRLRPGGKKKYKLKHVVWASRELERFAINPGLLETSEGCRQILGQLQPSLQTGSEERK

(I) (V)

150

.- - - - -p17 . --1p24 .

SLYNTVATLYCVHQKICKIKIKEALEXIEEEQNKSKKKAQQAAADTGNRGNSSQVSQNYPIVQNIQGQMVHQAIS

. . . 225

PRIMAWVXVVEERAFSPEVIPNFSALSEGATPQDLNTQMLNGGHQAAMQNLETINEEAAEWDRIHPAHAGPI

(V)

. . . 300

APGQMREPRGSDIAGTTmSTLQEQIGWMTNNPPIPVGEIYKMWIIGINKIVRNYSPSSIDIRQGPKEPFRDYVD

. . . 375

.--- p24---

p2--RFYKTIRAEQASQEVKNWNTETLLVQNANPDCITILKALGPAATLEEMNTACQGVGGPGHKARVIAEAVQVTNS

p7.

. . . 450

-p7 ---

-.pI

----ATI!Q(QRGNFRNQRKIIKCFNCGKEGHIADNIcRAP IKRCNFCGKEGHQDRQ lKIWPSCKGRPGNF

. . . 506

---.p6----_._._._._-_

PQSRTEPTAPPEESFRFGEETTmTPYQKQEKKQETIDKDLYPLASLKSLFGNDPLSQ

FIG. 8. (Top)Schematic diagramshowing the maturationproteolyticprocessingof theHIV-lMNGag precursor to theGag proteins(p17, p24, p2, p7, pl, andp6)found inthe mature virus.(Bottom)Amino acid sequence of a functionalGagprecursor from the most infectious replicative variants of HIV-1MN grown in H9 cells. Amino acid residues in parentheses indicate substitutions that distinguish another competentGagprecursor also present in the infectiousreplicative variants.

ofpurified p2,

p6,

andp7were identical to the amino acid

sequences predicted by the proviral DNA sequence, and therewas no evidence of amino acid sequence variants or

posttranslational modifications for these proteins. Purified

pl (Fig. 1) was found to have the amino acid sequence

FLGKIWPSCKGRPGNF, which differs from the predicted sequence at position 13, at which the pl protein contains

proline in place of a deletion, indicated by the DNA

se-quence, andatposition 14,atwhichthe plproteincontains glycine in place of arginine. No posttranslational modifica-tionsorsequencevariants of pl were detected.

In summary, the analysis of the purified Gag proteins shows the complete amino acid sequences and maturation

proteolytic cleavage sites for the Gag precursors that

con-stitute themajorportionof

HIV-lMN

producedbyinfection

andreplicationin H9 cells(Fig.8).The amino acid sequence

of the most abundant form of the Gag precursor

(approxi-mately50% of the total) is shown inFig. 8, and that of the

nextmostabundant form(approximately 20%ofthetotal)is

alsoindicated. The schematicdiagram showing the

matura-tionproteolyticcleavage sites of the Gag precursors (Fig. 8)

indicates the cleavage sites for generating the proteins de-scribed here (p17, p24, p2, p7, pl, and p6). The cleavage sites shown in Fig. 8 are highly efficient sites, since no appreciable partial cleavage products were identified. How-ever,otherpeptidefragments of the Gag proteins indicating less efficientcleavage sites were also noted during the course of this work. These included fragments of p2, pl, and p6, indicating cleavages between residues 369 and 370, 439 and 440, and 490 and 491 inthediagram shown inFig. 8.

DISCUSSION

Products of the retroviral gag gene perform highly com-plex orchestrated tasksduringthe assembly,budding,

mat-uration, and infection stagesof the viral replication cycle.

Duringviralassembly, the proteins form membrane

associ-ations(12,30,32)and self-associations thatultimatelyresult

inbuddingofanimmature virion fromthe infectedcell, and

point mutationsorsmalldeletionsin the gag gene canexert a dominant repressive effect on viral production (35). Gag

precursorsalsofunctionduringviralassembly toselectively

bind andpackage two plus strands of genomic RNA(10, 11).

The assembled immature virion contains approximately

2,000 to3,000 Gag precursors(3, 20) and becomes infectious after activation of theviral protease(21), proteolytic cleav-ageofthe Gag and Gag-Pol precursors, and rearrangement of the products to form the mature virion. In the mature

virion, the Gag proteins associate to form specifically

de-fined structures (9) that are probably necessary for the infection process.

The amino acid sequences ofGag precursors have been deduced from the DNA sequences of proviruses derived from numerousstrains ofHIV-1 (28). Many of these provi-ruses are known tobe infectious; however, it is difficult or impossible to deduce the exact amino acid sequences of highly efficient Gag precursors from the study of proviral DNA sequences. The most direct way todetermine struc-tures of Gag precursors that function efficiently in viral

assemblyistodetermineexactamino acid sequences ofGag

proteins incorporated into highly replicative viruses. H9

on November 9, 2019 by guest

http://jvi.asm.org/

[image:8.612.131.491.79.371.2]

(9)

cellschronically infected with HIV-lMN contain many inte-grated proviruses (31) but do not shed large amounts of virions into the culture medium (2). However, when chron-ically infected H9 cells are continuously cocultivated with uninfected H9 cells, the amounts of virus shed into the mediumincrease by at least 50-fold. This biological amplifi-cation selects for the most infectious replicative viruses.

Here wehavepurified mature Gag proteins directly from large-scale production lots of HIV-1MN and determined their

completeamino acid sequences andposttranslational

modi-fications, includingthematuration proteolytic cleavagesites

in the precursors. We have usedcomplimentary, but inde-pendent, methods (mass spectrometry and automated Ed-man degradation plus amino acid analysis) (4, 8) to deter-mine the amino acid sequences and posttranslational modifications and have confirmed the results by indepen-dently determining the exact molecular weights of the puri-fied proteins by using ESI-MS. This combined approach creates alarge body of overlapping and redundant data that results inavery high level of confidence in the final deter-mined structures.

Theresults show that during maturation, the HIV-1 Gag

precursorsare cleavedtoyieldsixmatureGagproducts,as indicated in Fig. 8, and that the order of cleavage products in the precursors is p17-p24-p2-p7-pl-p6. The cleavage sites and number ofproteolytic products are in agreement with previous findings for

HIV-111,,

(15) and are also highly

homologous to the cleavage sites and number of products previously identified for simianimmunodeficiencyvirus(14)

and related viruses, such as HIV-2 (18). The proteolytic

processing scheme proposed here reveals five cleavage sites

and differs from other proposals indicating fewer cleavage

sites (27, 36). All sixcleavage products wereisolated from

HIV-1MN (Fig. 1) and, on the basis of the amounts of the

proteins, we estimate that approximately equal molar

amountsof eachGagproteinwererecovered fromthevirus. These findings suggestthatcleavage at each site occursby

an efficient process and are in agreement with the equal

molar recoveries ofGagprecursorcleavageproducts

previ-ously reportedfor other retroviruses(14, 17).

Each Gag protein was recovered as a single

chromato-graphic entity,with theexceptionofp17, forwhichatleast

six chromatographically distinct forms were noted (peaks

pl7atop17finFig.1). Analysisofpl7atopl7drevealed that

aminimumofeightdifferent amino acid sequencevariantsof

this protein were present in the total

HIV-lMN population

(Fig. 4). Quantitative recoveries indicated that a single

amino acid sequence variant of

p17 (Fig. 4)

contributed at

least50%to the total

p17

recoveredfrom the virus and that another variant (also identified in Fig. 4) contributed an

additional20to30%tothe total.The remainderof the total

p17

was divided among the other amino acid sequence variants ofp17 (Fig. 4),with no singlevariantrepresenting

more than about 5 to 8% of the total. Analysis of p24

revealed two amino acid sequence variants (Fig. 7). The mostabundant variant contributed at least50% tothetotal p24 recovered from thevirus,and the other variant contrib-utedatleast20% to thetotal (30%of the totalp24 was not

analyzed). No other amino acid sequence variants were

revealedby analysis of the totalprotein recoveredforeach ofthe otherGag proteins(pl,p2, p6, and p7;Fig. 1).Onthe basis ofquantitative recoveries and amino acid sequences, we suggest that approximately 50 to 60% of the total Gag

proteinsrecovered are proteolytic cleavage fragments of a

single

Gagprecursorand thatanadditional 20to

30%

of the total Gag

proteins

may

originate

from another

Gag

precur-sor, as indicated in Fig. 8. The sequence heterogeneity

probably results from the expression of several different

proviruses integratedin the chronicallyinfected cell

popu-lation. The total number ofdifferent virusesexpressedin the mixture cannot be estimated, but the most replicative vi-ruseswillbeexpanded in the cocultivationprocedure. The

data provided here help characterize the more replicative

viruses and show the amino acid sequences ofGag precur-sors that function

efficiently

in the assembly and budding

process.

In addition to amino acid sequences, posttranslational

modificationscangreatly influence biological propertiesand mustbe considered when describingthe molecular compo-sitions ofproteins. A

previous quantitative analysis

ofthe IIIB(15) andLAV(27) strains ofHIV-1suggestedthat most

MAproteinsin thematurevirusesweremyristylated. Here

we analyzedfive chromatographicvariants of the MA

pro-tein produced by the MN strain and showed that each

variantwasmyristylated.Together, thesevariantsrepresent

thevastmajorityof the total MAproteinpresent inthe bulk

virus; therefore, these data suggestthat themost prevalent

forms ofthe viruses

produced

from the MN strain utilize

myristylated Gag precursors

during

viral

assembly.

The

importance of the myristylation modification (30, 33) is

underscored by the recent finding that analogs of myristic acid caninhibit virus

production (5, 6).

Asearch for other

possible posttranslational

modifications was conducted

by

comparing experimentally determined molecular weights

(mass

spectrometry)

and calculated molecular

weights

(se-quencing)of the

purified

Gag

proteins.

The observed

molec-ular

weight

of the

protein

in the

pool

of

p24

analyzed

was200 ± 50

weight

units

higher

than the calculated molecular

weight, indicatingthe presence ofposttranslational

modifi-cations. The nature of the modifications was not

directly

determined; however,

these data are consistent with the

phosphoserine

and

phosphothreonine

(15, 27).

The results

presented

here were obtained from a

single

production lot of HIV-1MN and clearly show amino acid

sequence

heterogeneity

in the isolated virus. The methods

used here can be

applied

to other

production

lots to

help

determine sequence variations in the virus that may arise

throughcontinued

culturing.

Other lots have been

analyzed

by rp-HPLC andshow

chromatographic

evidence of

heter-ogeneity.

However, a morerefined

chromatographic

proce-dure must be developed to analyze and compare the

se-quence

heterogeneity

in each lot. ACKNOWLEDGMENTS

Wethank Clara M. Dinterman and Patricia Coulter Grove for editorial and clericalassistance.

This research was sponsored, at least in part, bythe National Cancer Institute under contract N01-CO-74102with Program Re-sources, Incorporated/DynCorp. This workwasalsosupported,in part,bygrantBUS87-14238 from theNationalScience Foundation. Mass spectral measurements were performed at the Structural BiochemistryCenter,anNSF-supportedBiologicalInstrument

Cen-ter attheUniversityofMaryland,BaltimoreCounty. REFERENCES

1. Barre-Sinoussi, F., J.C.Chermann,F. Rey,M.T. Nugeyre,S. Chamaret, J. Gruest, C. Dauguet, C. Axier-Blin, F. Vezinet-Brun,C.Rouzioux,W.Rozenbaum, andL.Montagnier. 1983. Isolation ofaT-lymphotropic retrovirusfromapatientat risk for acquired immune deficiency syndrome (AIDS). Science

220:868-871.

2. Bess, J. W., Jr. (Program Resources, Incorporated/Dyn Corp,

on November 9, 2019 by guest

http://jvi.asm.org/

(10)

Natl. Acad. Sci. USA 86:8655-8659.

6. Bryant, M. L., L. Ratner, R. J. Duronio, N. S. Kishore, B. Devadas, S. P.Adams, and J.I.Gordon. 1991.Incorporationof 12-methoxydodecanoate into the human immunodeficiency vi-rus1gagpolyproteinprecursorinhibits its proteolytic process-ingandvirus production in achronically infected human lym-phoidcell line. Proc. Natl. Acad. Sci. USA88:2055-2059. 7. Devash, Y., T. J. Matthews, J. E.Drummond, K. Javaherian,

D.J.Waters, L.0.Arthur, W. A. Blattner, and J. R.Rusche. 1990. C-terminalfragmentsofgp120 andsyntheticpeptidesfrom fiveHTLV-III strains: prevalenceofantibodies to the HTLV-III-MNisolate in infectedindividuals. AIDS Res. Hum. Retro-viruses6:307-316.

8. Fenselau,C. 1991.Beyondgenesequencing: analysis ofprotein

structure with mass spectrometry. Annu. Rev. Biophys. Biophys. Chem. 20:205-220.

9. Gelderbloom,H. R., M.Ozel, and G.Pauli. 1989. Morphogen-esis andmorphology ofHIV,structure-functionrelations. Arch. Virol. 106:1-13.

10. Gorelick, R., L. Henderson, J. Hanser, and A. Rein. 1988. Point mutantsofMoloneymurine leukemiavirusthatfail topackage viral RNA: evidence forspecific RNArecognition by a "zinc finger-like" protein sequence. Proc. Natl. Acad. Sci. USA 85:8420-8424.

11. Gorelick, R.J.,S.M.Nigida, Jr., J. W. Bess, Jr., L.0. Arthur,

L. E. Henderson, and A. Rein. 1990. Noninfectious human immunodeficiency virus type 1 mutants deficient in genomic RNA.J. Virol. 64:3207-3211.

12. Gottlinger, H. G., J. G. Sodroski, and W. A. Haseltine. 1989. Role of capsid precursor processing and myristoylation in morphogenesis and infectivity of humanimmunodeficiency

vi-rustype 1. Proc. Natl. Acad. Sci. USA86:5781-5785. 13. Gurgo, C.,H.-G.Guo,G.Franchini, A.Aldovini,E.Collalti, K.

Farrell,F.Wong-Staal, R. C. Gallo, and M. S. Reitz, Jr. 1988. Envelope sequences oftwonewUnited States HIV-1 isolates. Virology164:531-536.

14. Henderson, L. E., R. E.Benveniste, R. Sowder, T. D. Copeland, A. M. Schultz, and S. Oroszlan. 1988. Molecular characteriza-tion of gag proteins from simian immunodeficiency virus (SIVMne). J. Virol.62:2587-2595.

15. Henderson,L.E., T. D.Copeland, R. C.Sowder,A. M.Schultz, andS. Oroszlan. 1988.Analysis ofproteins and peptides purified from sucrose gradient banded HTLV-III, p. 135-147. In D. Bolognesi (ed.), Human retroviruses, cancer, and AIDS: ap-proaches toprevention andtherapy. Alan R. Liss, Inc., New York.

16. Henderson, L. E., H. C. Krutzsch, and S. Oroszlan. 1983. Myristyl amino-terminal acylation of murine retrovirus pro-teins: anunusual post-translational proteinmodification. Proc. Natl. Acad.Sci.USA 80:339-343.

17. Henderson,L.E., R. Sowder, T. D. Copeland, G. Smythers, and

S. Oroszlan. 1984. Quantitative separation of murine leukemia virus proteins by reversed-phase high-pressure liquid chroma-tographyreveals newly described gag and env cleavage

prod-ucts.J.Virol. 52:492-500.

18. Henderson, L. E., R. C. Sowder, T. D. Copeland, S. Oroszlan, andR.E.Benveniste.1990.Gagprecursorsof HIVand SIV are cleaved into sixproteins found in the mature virions. J. Med. Primatol. 19:411-419.

viral infectivity. Proc. Natl. Acad. Sci. USA 85:4686-4690. 22. Laemmli, U. K. 1970. Cleavage of structural proteins during the

assembly of the head of bacteriophage T4. Nature (London) 227:680-685.

23. LaRosa, G. J., J. P. Davide, K. Weinhold, J. A. Waterbury, A. T.Proty,J. A.Lewis,A. J.Langlois, G. R. Dreesman, R. N. Boswell, P. Shadduck, L. H. Holley, M. Karplus, D. P. Bolognesi, T. J.Matthews, E. A. Emini, and S. D. Putney. 1990. Conserved sequence and structural elements in the HIV-1 principal

neu-tralizing determinant. Science249:932-935.

24. Laurent, A. G., B. Krust, M.-A. Rey, L. Montagnier, and A. G. Hovanessian. 1989. Cell surface expression of several speciesof human immunodeficiency virus type 1 major core protein. J. Virol. 63:4074-4078.

25. Levy, J. A., A. D. Hoffman, S. M. Kramer, J. A.Landis,J. M. Shimabukuro, and L. S. Oshiro. 1984. Isolation of lymphocyto-pathic retroviruses from San Francisco patients with AIDS. Science 225:840-842.

26. Loo, J. A., C. G. Edmonds, and R. D. Smith. 1990. Primary sequence information from intact proteins by electrospray ion-ization tandem mass spectrometry. Science 248:201-204. 27. Mervis, R. J., N. Ahmad, E. P. Lillehoj, M. G. Raum, F. H. R.

Salazar, H. W. Chan, and S. Venkatesan. 1988. The gag gene products of human immunodeficiency virus type 1: alignment within the gag open reading frame, identification of posttrans-lational modifications, andevidencefor alternative gag precur-sors. J. Virol. 62:3993-4002.

28. Myers, G., J. A.Berzofsky, B. Korber, R. F. Smith, and G. N. Pavlakis (ed.). 1991. Human retroviruses and AIDS 1991, a compilation and analysis of nucleic acid and amino acid se-quences. Los Alamos National Laboratory, Los Alamos, N.Mex.

29. Popovic, M., M. G. Sarngadharan, E. Read, and R. C. Gallo. 1984. Detection, isolation, and continuous production of cyto-pathic retroviruses (HTLV-III) from patients with AIDS and pre-AIDS. Science 224:497-500.

30. Rein, A., M. R. McClure, N. R. Rice, R. B. Luftig, and A. M. Schultz. 1986. Myristylation site inPr65ragis essential for virus particle formation by Moloney murine leukemia virus. Proc. Natl. Acad. Sci. USA 83:7246-7250.

31. Reitz, M. (National Cancer Institute). 1991. Personal communi-cation.

32. Rhee, S. S., and E. J. Hunter. 1990. Structural role of the matrix protein of type D retroviruses in Gag polyprotein stability and capsid assembly. J. Virol. 64:4383-4389.

33. Schultz, A., and A. Rein. 1989. Unmyristylated Moloney murine leukemia virus Pr65gag is excluded from virus assembly and maturation events. J. Virol. 63:2370-2373.

34. Toplin, I., and P. Sottong. 1972. Large-volume purification of tumor viruses by use of zonal centrifuges. Appl. Microbiol. 23:1010-1014.

35. Trono, D., M. B. Feinberg, and D. Baltimore. 1989. HIV-1 Gag mutants can dominantly interfere with the replication of the wild-type virus. Cell 59:113-120.

36. Veronese, F. D., T. D.Copeland,S. Oroszlan, R. C.Gallo, and M. G. Sarngadharan. 1988. Biochemical and immunological analysis of human immunodeficiency virus gag gene products p17 and p24. J. Virol. 62:795-801.