© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES
Virus-specific cytotoxic T-lymphocyte responses select for amino-acid variation in simian immunodeficiency virus Env and Nef DAVID T. EVANS1, DAVID H. O’CONNOR1, PEICHENG JING1, JOHN L. DZURIS2, JOHN SIDNEY2, JACK DA SILVA3, TODD M. ALLEN1, HELEN HORTON1, JOHN E. VENHAM1, RICHARD A. RUDERSDORF4, THORSTEN VOGEL1, C. DAVID PAUZA1,5, RONALD E. BONTROP6, ROBERT DEMARS4, ALESSANDRO SETTE2, AUSTIN L. HUGHES3 & DAVID I. WATKINS1,5 1
© 1999 Nature America Inc. • http://medicine.nature.com
Wisconsin Regional Primate Research Center, University of Wisconsin, 1220 Capitol Court, Madison, Wisconsin 53715, USA 2 Epimmune, 5820 Nancy Ridge Drive, San Diego, California 92121, USA 3 Institute of Molecular Evolutionary Genetics, The Pennsylvania State University, University Park, Pennsylvania 16802, USA 4 Laboratory of Genetics, Genetics Building, University of Wisconsin, Madison, Wisconsin 53715, USA 5 Department of Pathology and Laboratory Medicine, University of Wisconsin, Madison, Wisconsin 53715, USA 6 Biomedical Primate Research Centre-TNO, P. O. Box 5815, 2280 HV Rijswijk, The Netherlands D.T.E. present address: New England Regional Primate Research Center, Harvard Medical School, One Pine Hill Drive, Southborough, Massachusetts 01772, USA J.D. present address: Department of Biology, East Carolina University, Greenville, North Carolina 27858, USA D.T.E. and D.H.O. contributed equally to this study. Correspondence should be addressed to D.I.W.; email:
[email protected]
Cytotoxic T-lymphocyte (CTL) responses to human immunodeficiency virus arise early after infection, but ultimately fail to prevent progression to AIDS. Human immunodeficiency virus may evade the CTL response by accumulating amino-acid replacements within CTL epitopes. We studied 10 CTL epitopes during the course of simian immunodeficiency virus disease progression in three related macaques. All 10 of these CTL epitopes accumulated amino-acid replacements and showed evidence of positive selection by the time the macaques died. Many of the amino-acid replacements in these epitopes reduced or eliminated major histocompatibility complex class I binding and/or CTL recognition. These findings strongly support the CTL ‘escape’ hypothesis.
Cytotoxic T-lymphocyte (CTL) responses arise early after human immunodeficiency virus (HIV) or simian immunodeficiency virus (SIV) infection and are important in controlling viral replication throughout the course of infection1–8. Despite vigorous CTL responses, most infected individuals eventually develop AIDS. One hypothesis to explain the inability of the CTL response to contain HIV proposes that selective pressure by CTLs ultimately results in the emergence of viral ‘escape’ variants. In support of this viral ‘escape’ hypothesis, genetic changes have been found in the virus populations in infected individuals that reduce or eliminate CTL recognition in vitro9–18. However, the importance of CTL selection in the course of virus evolution in infected individuals remains controversial19. A chief criticism of the viral ‘escape’ hypothesis is the absence of data showing statistically significant positive Darwinian selection within CTL epitopes in infected hosts19,20. To explore the relationship between virus-specific CTL responses and the evolution of SIV in vivo, we did a longitudinal study in five SIV-infected offspring from a family of major histocompatibility complex (MHC)-defined rhesus macaques. Three of these animals made strong CTL responses to five different epitopes of the SIV Env and Nef proteins, thereby enabling us to follow the evolution of their virus populations in response to CTL selection. We also tested 1270
the ability of CTL epitope variants to bind to their restricting MHC class I molecules and to sensitize targets for CTL recognition.
Emergence of SIV Env and Nef CTL epitope variants We infected five members of a family of MHC-defined rhesus macaques that were identical at their MHC class II DRB loci (Fig.
Fig. 1 Pedigree of an MHC-defined family of rhesus macaques. Within symbols, MHC class I haplotypes for each macaque. Below, MHC class I molecules that bind peptides from each macaque (in parentheses, haplotype encoding these molecules). The MHC class II DRB molecules in all of the offspring of this family were identical. NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES Macaque C recognized a Mamu-A*11 Env-derived CTL epitope and a MamuB*17 Nef-derived CTL epitope. There were amino-acid replacements in both Macaque CTL epitopes Restricting Survival time Disease progression (class I haplotype) MHC class I molecule (days) of these CTL epitopes at the time of (haplotype) death (Fig. 4). For the Env epitope bound by Mamu-A*11, we found a C (b/e) GDYKLVEIEnv497-504 Mamu-A*11 (b) 299 intermediate leucine-to-valine change at position 5 Mamu-B*17 (b) IRFPKTFGWNef165-173 (GDYKVVEI) and an isoleucine-to-vaD (b/c) GDYKLVEIEnv497-504 Mamu-A*11 (b) 889 slow line change at position 8 (GDYKLVEV). Mamu-B*17 (b) IRFPKTFGWNef 165-173 We also found amino-acid changes at KRQQELLRLEnv575-583 Mamu-B*03 (c) the first and third positions of the Nef Mamu-B*03 (c) ARRHRILDIYLNef136-146 epitope bound by Mamu-B*17, includQGQYMNTPWNef62-70 Mamu-B*04 (c) ing threonine at position 1 and tryptoMamu-B*03 (c) 511 slow A (a/c) KRQQELLRLEnv575-583 phan at position 3 (TRWPKTFGW). ARRHRILDIYLNef136-146 Mamu-B*03 (c) These epitope variants constituted most Mamu-B*04 (c) QGQYMNTPWNef62-70 of the late-stage plasma virus populaB (a/d) Not detected None 77 rapid tion, indicating that they might confer B’(a/d) Not detected None 78 rapid a selective advantage to the virus. Thus, both of the CTL epitopes mapped in 1) intravenously with the same dose and stock of SIV (refs. macaque C accumulated amino-acid substitutions by the time this 21,22). Two of these, macaques B and B’, rapidly progressed to animal was killed with AIDS-associated wasting on day 299 after indisease without making a strong CTL response to the virus. The fection. other three, macaques A, C and D, had either slow or intermediMacaque A developed CTL responses directed against three difate courses of disease progression and recognized multiple SIV ferent CTL epitopes derived from the Env and Nef proteins which CTL epitopes (D.T.E et al., manuscript in preparation, and Table were restricted by Mamu-B*03 and Mamu-B*04. There was a 1). The MHC alleles of the rapid progressors B and B’ were iden- change from leucine to methionine at position 9 of the Mamutical, and differed from those of the slow progressor A by a sin- B*03-restricted Env epitope (KRQQELLRM; Fig. 4b). There were also gle MHC class I haplotype. This haplotype (haplotype c) was changes in the first and second positions of the Mamu-B*04-reshared by both slow progressors A and D, and encoded two mol- stricted Nef epitope (EGQYMNTPW; QEQYMNTPW; SGQYMecules (Mamu-B*03 and Mamu-B*04; Mamu (Maccaca mulatta), NTPW and PGQYMNTPW; Fig. 4a). Two prominent changes were nomenclature for the rhesus MHC) used to present three differ- also present in the Mamu-B*03-restricted Nef epitope. At the time ent CTL epitopes derived from the SIV Env and Nef proteins of the macaque’s death, the alanine and the isoleucine of this epi(Table 1). Likewise, two MHC class I molecules (Mamu-A*11 tope had been replaced completely by proline and methionine and Mamu-B*17) encoded by the b haplotype of the intermedi- (PRRHRILDMYL) (Fig. 4a). Thus, by the time macaque A died, the ate progressor C bound two additional Env and Nef CTL epi- plasma virus population showed amino-acid changes in all three topes. Macaque D inherited the b haplotype in addition to the c haplotype, and thus made CTL responses to all five epitopes (Table 1). If MHC class I-restricted CTL responses were responsible for exerting antiviral pressure in the intermediate and slow progressors, positive selection for sequence variation would be expected to occur in the CTL epitope-coding regions of the SIV Env and Nef genes of plasma virus populations in these individuals. To test this, we sequenced Env and Nef cDNA clones, amplified by RT–PCR from cryopreserved plasma virus samples taken at selected times after infection. Compared with the inoculum, there were amino-acid substitutions in all of the five CTL epitopes recognized by these macaques before they succumbed to AIDS (Figs. 2–5, Env and Nef CTL epitope variation).
© 1999 Nature America Inc. • http://medicine.nature.com
Table 1
CTL epitopes, restricting MHC class I molecules, and disease progression in a family of SIV-infected macaques
Fig. 2 Evolution of the plasma virus population in Nef from macaque D. This macaque’s CTLs recognized all three Nef epitopes: B*04 (pink), B*03 (blue) and B*17 (yellow). Each of these epitopes sustained amino-acid replacements by the time of death (889 d). Left (d. PI), day after infection on which samples were obtained. Each line represents different combinations of amino-acid replacements within CTL epitopes, and the frequency of these replacements is in parentheses. For example, 13 of 17 clones sequenced at time of death had Asp (D) substitutions within the B*04 epitope, Glu (E) and Lys (K) substitutions within the B*03 epitope, and Thr (T), Tyr (Y) and Ile (I) substitutions within the B*17 epitope. In regions flanking the CTL epitopes, the consensus amino-acid sequence is shown. NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
1271
© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES
© 1999 Nature America Inc. • http://medicine.nature.com
Fig. 3 Evolution of the plasma virus population in Env from macaque D. This macaque’s CTLs recognized two Env epitopes: A*11 (purple) and B*03 (green). Left (d. PI), day after infection on which samples were obtained. Each line represents different combinations of amino-acid replacements within CTL epitopes and the frequency of these replacements are in parentheses.
Statistical evidence for selection on CTL epitopes To determine whether the amino-acid replacements had accumulated in the CTL epitopes by chance or because of selection, we calculated the synonymous (silent) nucleotide substitution rate (dS), and the non-synonymous (amino-acid replacement) nucleotide substitution rate (dN) in the CTL epitope-coding regions of Env and Nef. Because of the low fidelity of reverse transcription, there is a high rate of mutation associated with the replication of SIV and HIV (ref. 23). As most non-synonymous mutations are deleterious and will be selected against, dS is almost always higher than dN (ref. 24). However, there are rare examples in which dN is greater than dS, and this is because of positive selection for increased sequence variation25. Analysis of dN versus dS in the CTL epitope coding regions showed evidence of CTL selection on both Env and Nef. For the regions of Env and Nef encoding CTL epitopes, the mean dN was significantly higher than the mean dS when each timepoint was compared with previous sample, and when the time-of-death sample for each macaque was compared with the viral inoculum sequences (P < 0.01; Table 2). As expected, dS was higher than dN in the regions flanking the epitopes. Furthermore, there was no significant increase in dN for epitope-encoding sequences in virus from macaques that did not recognize the corresponding peptide (nonrestricted CTL epitopes; Table 2). Thus, statistically significant evidence for positive selection was present exclusively in the regions of Env and Nef coding for the restricted CTL epitopes. To determine whether the most variable regions of the viral Nef and Env genes coincided with the CTL epitope coding regions, we scanned the sequenced regions of Nef and Env with a continually shifting ‘window’ of nine codons. We compared the mean number of synonymous and non-synonymous nucleotide substitutions within each ‘window’ for the last time point taken for each macaque and the viral inoculum, to generate sequence variability plots of dN and dS. The Nef frames showing the greatest dN values coincided with the Nef CTL epitopes recognized by each macaque
defined CTL epitopes. All five of the CTL epitopes recognized by macaque D had also sustained amino-acid replacements by the time this macaque was killed on day 889 after infection (Fig. 4). The most common variants of each CTL epitope in this macaque at that time included GDYKLIEV (Mamu-A*11 Env), KRQHELLRL (Mamu-B*03 Env), ERRHRILDKYL (Mamu-B*03 Nef), QGQYMNNPW (Mamu-B*04 Nef), and TRYPKIFGW (Mamu-B*17 Nef). Most of these substitutions differed from those observed in macaques A and C. However, there were similarities in the position of the substitutions in two of the epitopes. The Mamu-B*03-restricted Nef epitope had amino-acid changes in the first and ninth positions in macaques A and D (Figs. 2–4). Similarly, the most common variants of the Mamu-B*17-restricted Nef epitope in macaques C and D included substitutions in the first and third positions. The Env and Nef regions encoding CTL epitopes that are not recognized in one anTable 2 Mean numbers of synonymous (dS) and nonsynonymous (dN) nucleotide imal, but are recognized in the other anisubstitutions per site in SIV epitope and non-epitope (remainder) regions. mal, provide an internal control for Comparisons with previous sample changes caused by factors other than CTL dS (s.e.m.) dN (s.e.m.) P* (dS = dN) selection. The three CTL epitopes recognized by macaques A and D were not rec- Remainder 0.0150 (0.0055) 0.0092 (0.0048) < 0.001 ognized by macaque C, and the two CTL Epitopes: Restricted 0.0121 (0.0140) 0.0276 (0.0317) < 0.01 epitopes recognized by macaques C and D Non-restricted 0.0071 (0.0089) 0.0035 (0.0082) n.s. were not recognized by macaque A. These P (restricted = remainder) n.s < 0.01 non-restricted CTL epitopes did not undergo the replacements found in the reP (restricted = non-restricted) n.s. < 0.01 stricted epitopes, supporting the conclusion that the amino-acid changes Comparisons between last and inoculum samples in the restricted CTL epitopes were not the dS (s.e.m.) dN (s.e.m.) P (dS = dN) result of other factors. Furthermore, in one of the two rapid progressors (macaque B) Remainder 0.0288 (0.0104) 0.0150 (0.0089) < 0.01 that did not make a detectable CTL Epitopes: Restricted 0.0097 (0.0168) 0.0817 (0.0363) < 0.01 response, substantial variation was not Non-restricted 0.0015 (0.0030) 0.0108 (0.0202) n.s. seen in any of the five CTL epitopes recognized by the other macaques P (restricted = remainder) n.s. < 0.01 (http://www.primate.wisc.edu/people/ doconnor/SuppData1.pdf). These observa- P (restricted = non-restricted) n.s. n.s. tions support the conclusion that CTL Restricted, CTL epitope recognized in an animal that expresses an MHC class I molecule that binds the corresponding selection was responsible for the amino- peptide; non-restricted, CTL epitopes that are not recognized by an animal (because the animal does not express the apacid substitutions within each restricted propriate MHC class I molecule), yet are recognized in another animal (which has the appropriate MHC class I molecule). n.s., not significant (P > 0.05); *, Paired-sample t-test with pairing within each macaque (two-tailed). epitope. 1272
NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
© 1999 Nature America Inc. • http://medicine.nature.com
© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES Fig. 4 Analysis of sequences isolated from MHC-defined rhesus macaques at time of death. a, Nef sequences. b, Env sequences. Shaded, epitopes recognized by CTLs; boxed, CTL epitopes not recognized in each macaque (control, non-restricted CTL epitopes; Table 2). Virus was quantified by branched DNA analysis. The inoculum sequences were amplified from an input pool of 5.98 × 106 virions/ml. Plasma SIV sequences from macaque C were amplified from 1.58 × 106 input virions; macaque D from 4.3 × 104 input virions; macaque A from 2.80 × 105 input virions. Each line represents different combinations of amino-acid replacements within CTL epitopes, and the frequency of these replacements is in parentheses. Variability plots below each sequence alignment show the nucleotide sequence variability in the CTL epitope coding versus flanking regions. dS (green, line) and dN (red, solid) were computed by comparing sequence from the most recent time to the inoculum for the sequenced region of Nef using a sliding ‘window’ of nine codons.
a
C
D
A
(Fig. 4a). Similarly, the Env frames corresponding to CTL epitopes in macaques A, C and D also showed increases in dN (Fig. 4b). These results provide additional evidence that CTL selection is responsible for increased sequence diversity in the studied regions of the SIV Nef and Env genes.
b
C
MHC I binding and CTL recognition of variants To determine whether the amino-acid replacements contributed to evasion of the CTL response, we tested variant peptides for MHC class I binding and CTL recognition. We measured MHC class I D binding by competition with 125I-labeled epitope or epitope-analog peptides in a live cell binding assay26 and assessed SIV-specific CTL activity using standard 51Cr-release assays. In all but a single epitope, the most frequent variants in the plasma virus population at time of death diminished MHC class I binding and/or CTL recognition (Table 3 and Fig. 5). The most profound A example of CTL evasion was in the Mamu-A*11restricted Env response. The single amino-acid substitution predominating at time of death in macaque C resulted in both an 89% reduction in peptide binding and a complete loss of CTL recognition at a peptide concentration between 1 and 10 nM (Table 3 and Fig. 5a). In contrast, the single replacement of methionine for leucine in the Mamu-B*03-restricted Env epitope (KRQQELLRM) reduced peptide binding by 56%, yet had only a minimal effect on CTL lysis (Table 3 and Fig. 5c). The substitution of histidine for glutamine in the same epitope in macaque D NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
(KRQHELLRL) also had a limited effect on peptide binding and CTL recognition (Table 3 and Fig. 5h). For the Nef epitopes, most of the variants reduced both peptide binding and CTL lysis (Table 3 and Fig. 5b and d–f). The TRWPKTFGW variant did not reduce peptide binding, yet failed to sensitize targets for CTL lysis (Table 3 1273
© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES
© 1999 Nature America Inc. • http://medicine.nature.com
and Fig. 5b), indicating that the mutant peptide may interfere with T-cell receptor recognition. As expected, replacement of glutamic acid at the ‘anchor’ residue of the QGQYMNTPW peptide (J.L.D. et al., manuscript submitted) completely abrogated peptide binding and reduced CTL recognition (Table 3 and data not shown).
Discussion In both HIV- and SIV-infected individuals, MHC class I-restricted CTL responses are important in controlling virus replication. In the first few weeks after infection, the appearance of virus-specific CTLs coincides with the resolution of primary viremia1,3,4,27. CTL responses also seem to contribute to the containment of HIV replication during later stages of infection. In certain individuals, high frequencies of HIV-specific CTL are present throughout the asymptomatic phase, and only begin to decline as the individuals progress to AIDS (refs. 28–30). A correlation between the strength of the CTL response and plasma virus loads has been shown by the use of MHC class I tetramers for directly quantifying virus-specific CTLs (ref. 5). Those studies demonstrated an inverse correlation between the frequency of HIV-specific CTLs and steady-state plasma virus loads, which are the best predictor of survival after HIV infection31. Moreover, depletion of CD8 cells from nonhuman primates infected with SIV or simian–human immunodeficiency virus results in higher viral loads, consistent with the importance of CD8 cells in controlling viral replication6–8. However, despite the presence of strong virus-specific CTL responses, most HIV-infected individuals eventually succumb to AIDS. Although several mechanisms could contribute to the ultimate failure of CTLs to control HIV infection (review, ref. 32) our data support the CTL ‘escape’ hypothesis by providing clear evidence for CTL selection on Env and Nef epitopes. Escape from CTL recognition seems to be a common and direct result of CTL selection, as the most frequent CTL epitope variants late in infection in each animal substantially reduced CTL recognition and/or MHC class I binding. In most cases, amino-acid changes were selected that greatly diminished MHC class I binding and resulted in a measurable decrease in CTL recognition in vitro. One exception was the Mamu-B*03 Env epitope. The variants KRQQELLRM and KRQHELLRL had little effect on CTL recognition, as assessed by multiple chromium release assays. Although these results are difficult to interpret, it is possible that subtle changes in CTL recognition of an epitope in vitro may have greater consequences on in vivo viral fitness. Alternatively, it is also possible that these substitutions may interfere with peptide processing and therefore may not be efficiently presented on the surface of SIV-infected cells for CTL recognition. Several studies have examined the viral ‘escape’ hypothesis without demonstrating conclusive evidence for selection on CTL epitopes. Infection of related, MHC-defined rhesus macaques with a sequence-defined virus allowed us to follow the evolution of the virus population in infected hosts with few confounding conditions. The reduced length of disease course in SIV-infected macaques compared with that in HIV-infected humans and the absence of anti-retroviral treatment probably facilitated our ability to observe selection by CTL on the virus. Positive selection on Env and Nef CTL epitopes supports the viral ‘escape’ hypothesis and emphasizes the importance of CTLs in controlling viral replication. Future challenge experiments with MHC-defined macaques and SIV CTL ‘escape’ variants can now be used to evaluate the contribution of CTL ‘escape’ to viral fitness. 1274
Table 3
Peptide binding to rhesus MHC class I molecules Relative binding Average
± s.d.
% Binding reduction
Epitopea Analog Variant Variant
KRQQELLRL Y———— ————M —-H——-
1.000 0.154 0.435 1.090
0.00 0.090 0.127 0.370
0.0 84.6 56.5 0.0
Epitopeb Epitope Epitope Variant Variant Variant Variant Variant Variant Variant
RRHRILDMYL A————— A———-I— ———-I— V————— E————— P————— E———-T— ———-T— E———-K—
1.000 2.390 2.550 2.480 4.263 0.790 0.060 0.052 1.398 0.012
0.00 0.902 1.090 1.780 4.983 0.829 0.069 0.037 0.520 0.005
0.0 0.0 0.0 0.0 0.0 21.0 94.0 94.8 0.0 98.8
Epitopec Variant Variant
GDYKLVEI ——V———-I-V
1.000 0.107 0.920
0.00 0.006 0.20
0.0 89.3 8.0
Epitoped Epitope Variant Variant Variant Variant Variant Variant
IRFPKTFGW —Y——— —L——— —W——— T-Y——— T-W——— —Y—I—T——I—-
1.000 0.410 0.690 2.728 3.760 1.340 0.155 0.105
0.000 0.149 0.014 1.463 3.147 0.761 0.097 0.049
0.0 59.0 31.0 0.0 0.0 0.0 84.5 89.5
Epitopee Variant Variant Variant Variant Variant Variant Variant Variant Variant Variant
QGQYMNTPW P———— E———— S———— -E———-R————-H———-C—————N— ———K— ——I-N—
1.000 0.012 0.033 1.470 0.0007 0.0005 0.171 0.301 0.425 0.310 0.920
0.000 0.008 0.008 1.195 0.001 0.000 0.050 0.258 0.343 0.080 0.120
0.0 98.9 96.7 0.0 >99.9 >99.9 82.9 69.9 57.3 69.0 8.0
Peptide binding is expressed relative to the binding affinity of the index peptide for each particular rhesus macaque MHC class I molecule. Percent binding reduction is calculated as (1.0–relative binding ratio) × 100. 125I labeled peptides: bYRQQELLRL and RRHRILDMYL; cRRHRILDMYL and YRQQELLRL; dGDYKLVEI; eIRYPKTFGW; fQGQYMNTPW.
Methods Viruses and infections. Rhesus macaques were infected intravenously with 40 TCID50 (TCID50: half-maximal tissue culture infectious dose) of a heterogeneous SIV stock (originally provided by R.C. Desrosiers, Harvard Medical School and New England Regional Primate Research Center). The stock was amplified by growth on rhesus peripheral blood mononuclear cells with a final passage on CEMX174 cells to increase the titer21–33. Virus dilutions were prepared in 1 ml of sterile saline, and were slowly injected into the saphenous veins of anesthetized macaques at a rate of 1 ml per minute. SIV-infected macaques were cared for according to an experimental protocol approved by the University of Wisconsin Research Animal Resource Committee. CTL epitope mapping. CTL cultures were established from peripheral blood samples (collected in heparinized tubes) of SIV-infected rhesus macaques. Peripheral blood lymphocytes were isolated on Ficoll-Hypaque and stimulated 1:1 with 5 × 106 paraformaldehyde-fixed, autologous B-lymphoblastoid cell lines infected overnight with vaccinia virus constructs expressing the SIVmac251 Gag, Pol, or Env proteins or the SIVmac239 Nef protein (provided by Therion Biologics, Cambridge, Massachusetts)34,35. Half of the medium was replaced after 2 d with R10 medium supplemented with 20 U/ml recombinant IL-2, a gift from Hoffman-LaRoche (Nutley, New NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
© 1999 Nature America Inc. • http://medicine.nature.com
© 1999 Nature America Inc. • http://medicine.nature.com
Fig. 5 Recognition of epitope variants by CTLs from macaques C, A and D. Epitope variants that reduced or eliminated CTL recognition were identified in all three macaques. Amino-acid differences between the epitope sequences of the inoculum and these variants are underlined. Although we tested many mutant peptides in CTL assays, only those peptides corresponding to the most frequent epitope variants identified in the plasma virus populations of each animal at time of death are presented. CTL recognition of these mutant peptides ( and ) were compared to the mapped peptide () and an irrelevant control peptide () for each epitope. Autologous B-lymphoblastoid cell line targets were pulsed with 10fold dilutions of each peptide and were tested at an effector:target ratio of 20:1 E:T in chromium release assays.
% Specific 51Cr release
ARTICLES
Peptide concentration [nM]
Jersey). On day 7, viable cells were isolated on Ficoll-Hypaque and again stimulated 1:1 with paraformaldehyde-fixed, vaccinia-infected B-lymphoblastoid cell lines. CTLs were then expanded in the presence of recombinant IL-2, and were tested for CTL activity after a total of 13 d in culture. CTL epitopes of the SIV Env and Nef proteins were mapped using sets of peptides synthesized according to the predicted amino-acid sequences of these proteins for SIVmac251. Autologous B-lymphoblastoid cell lines were first pulsed with ‘pools’ of overlapping peptides 20 amino acids in length, and used as CTL targets in 51Cr-release assays. In subsequent CTL assays, individual peptides 20 amino acids in length from positive pools were tested for CTL recognition. These were then followed by testing overlapping peptides 9 amino acids in length within each positive peptide 20 amino acids in length. Fine mapping was completed by ‘titrating’ additional peptides with single amino-acid additions or subtractions at the Nand C-termini (peptides 8–10 amino acids in length) to precisely define the ends of each CTL epitope. SIV-specific CTL activity was assessed using a standard 51Cr-release assay36. Plasma virus sequencing. SIV was isolated from 1- to 2-ml frozen plasma samples collected in tubes containing EDTA. After being thawed, each sample was centrifuged at 10,000g for 10 min to remove any contaminating cellular debris. The virus was then pelleted in an ultracentrifuge at 35,000 rpm for 33 min at 4 °C, using a Beckman SW60 rotor. Viral RNA was extracted from the pellets using 0.5 ml RNA STAT-60 (Tel-Test) according to the manufacturer’s instructions. After being precipitated overnight at –20 °C with an equal volume of isopropanol and 20 µg glycogen, the viral RNA was collected by centrifugation at 12,500g, washed with 75% ethanol and resuspended in 20 µl RNAse-free water. Total SIV cDNA was synthesized from 2–3 µl viral RNA in a 20-µl reverse transcription reaction using Superscript II according to the manufacturer’s instructions (Life Technologies). Fragments approximately 700 bp in length of the viral Env and Nef genes were amplified in a 100-µl PCR reaction by the addition of 80 µl PCR reaction buffer containing 1 mM MgCl2, 2.5 U Taq polymerase (Perkin-Elmer, Norwalk, Connecticut), and 25 pmol primers (Env-specific: ENVF793, 5’–GCCTCCAAGAGAGGGAGA–3’ and ENVR8637, 5’–AAGGTCAAACCAATTGCC–3’; or Nef-specific: NEFF9047, 5’–CTACCTACAATATGGGTG–3’ and NEFR9762, 5’–GCTTCCAACTCTTCTGGG–3’). After an initial denaturation step at 94 °C for 1 min, reactions consisted of 40 cycles of denaturation (30 s at 94 °C), annealing (30 s at 53 °C) and extension (45 s at 72 °C), followed by a final extension step at 72 °C for 10 min. To preserve more material for future analysis, viral RNA from macaque D was isolated using a QIAgen Viral RNA kit (QIAgen, Valencia, California) according to the manufacturer’s instructions. Four 140-µl reactions were combined onto a single purification column to increase the viral RNA yield. NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
Viral RNA (2 µl) was then used in the SuperScript One-Step RT–PCR System (Life Technologies) with the Env and Nef primers described above. The cDNA was obtained in a 30-minute step at 50 °C and was amplified with a 2-minute 94 °C denaturation step followed by 35 cycles of denaturation (15 s at 94 °C), annealing (30 s at 53 °C), and extension (90 s at 72 °C). The PCR was completed with a final extension at 72 °C for 10 min. PCR products were cloned into the pCRII vector in a 5-minute ligation reaction at room temperature using the TOPO TA Cloning kit according to the manufacturer’s instructions (Invitrogen, Carlsbad, California). The ligation reactions were then used to transform Escherichia coli strain TOP10 (Invitrogen, Carlsbad, California), and colonies were grown overnight on LB plates containing 50 µg/ml ampicillin and 50 µg/ml X-gal. Plasmid DNA was prepared by ‘miniprep’ from selected white colonies, and sequenced using the Taq DyeDeoxy Terminator Cycle Sequencing kit and an ABI 373 automated sequencer (both from Perkin-Elmer, Norwalk, Connecticut). Primers used for sequencing included the PCR primers ENVF7935, ENVR8637, NEFF9047 and NEFR9762, and nested primer pairs (ENVF7977, 5’–CAGTCTCATAGCAAACAT–3’ and ENVR8446, 5’–CAGTAGTGTGGCAGACTTG–3’; and NEFF9115, 5’–ACTCTTGCGGGCGCGTGG–3’ and NEFR9689, 5’–GAACCTCTCCCCAAGGGT–3’). The consensus was derived by manually determining the predominant nucleotide at a given position in the inoculum. MHC class I binding assay. Peptides were obtained as lyophilized products from Chiron Mimotopes (San Diego, California). Peptides were resuspended at a concentration of either 10 mg/ml or 20 mg/ml in 10% DMSO and PBS. Peptide GDYKLVEI was used for binding to Mamu-A*11,; peptide IRYPKTFGW. for Mamu-B*17 binding; peptide QGQYMNTPW, for MamuB*04 binding, and peptides RRHRILDMYL; and YRQQELLRL, for binding to Mamu-B*03. The live cell binding assay was done as described26,36. Macaque MHC class I-transfected 721.221 cells (1 × 106 cells/ml) were preincubated overnight in RPMI 1640 supplemented with 15% FBS, L-glutamine, 100 IU/ml penicillin and 100 µg/ml streptomycin at room temperature. Cells were then washed twice in RPMI 1640 without supplements and resuspended at a concentration of 1.25 × 107 cells/ml in RPMI supplemented with 3 µg/ml β2-microglobulin (Scripps Clinic and Research Foundation, La Jolla, California). Peptides purified with high-performance liquid chromatography were radiolabeled with 125I according to the chloramine-T method37. Cells (2 × 106 cells/well in a 96-well U-bottom microtiter plate) were incubated at 20 °C for 4 h in the presence of 1 × 105 counts per minute (c.p.m.) 125I-radiolabeled peptides, titrating concentrations of unlabeled inhibitor peptide (10-fold dilutions, 100 µg–0.1 ng), and protease inhibitors (final concentrations: 250 µg/ml PMSF, 1.07 mg/ml EDTA, 62.5 µg/ml pepstatin A, 60 µg/ml TLCK and 325 mg/ml phenanthroline). After this incubation, unbound peptide was removed by washing 1275
© 1999 Nature America Inc. • http://medicine.nature.com
ARTICLES cells three times in serum-free medium, then passed through a FBS gradient. Labeled peptide bound to the cells was then determined by measuring the radioactivity of pelleted cells on a gamma scintillation counter, and the concentration of peptide yielding 50% inhibition of the binding of the radiolabeled probe peptide was calculated (IC50).
© 1999 Nature America Inc. • http://medicine.nature.com
Statistical analysis of sequence data. Numbers of synonymous nucleotide substitutions per synonymous site (dS) and of nonsynonymous nucleotide substitutions per site (dN) were estimated by a published method24, and mean dS and dN were estimated between each sample and the previous sample and between the last sample and inoculum. These were estimated for separate regions of the sequenced portions of Env and Nef: epitopes restricted by the host; epitopes not restricted by the host; and the flanking regions. Mean dS and dN were computed in a sliding ‘window’ of nine codons along the sequence for comparisons between the last sample taken and the original inoculum.
Acknowledgments We thank L. Smith for help in preparing this manuscript and B. Becker for help with illustration. We thank J. Scheffler for initially identifying this family, and J. Mitchen, M. Dykhuizen and L. Acker for infecting the macaques, collecting blood and monitoring disease progression. We also thank J. Malter, S. Wolinsky and G. Watkins for critical review. This work was supported by grants from the National Institutes of Health (AI32426, AI42641, and AI41913 to D.I.W.; AI36643 to C.D.P.; AI15486 to R.D.; GM34940 to A.L.H.; AI38081 to Epimmune; and RR00167 to the Wisconsin Regional Primate Research Center). D.I.W. is an Elizabeth Glaser Scientist.
RECEIVED 24 MAY; ACCEPTED 31 AUGUST 1999 1. Borrow, P., Lewicki, H., Hahn, B., Shaw, G. & Oldstone, M. Virus-specific CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunodeficiency virus type 1 infection. J. Virol. 68, 6103–6110 (1994). 2. Koup, R. & Ho, D. Shutting down HIV. Nature 370, 416 (1994). 3. Yasutomi, Y., Reimann, K., Lord, C., Miller, M. & Letvin, N. Simian immunodeficiency virus-specific CD8+ lymphocyte response in acutely infected rhesus monkeys. J. Virol. 67, 1707–1711 (1993). 4. Reimann, K. et al. Immunopathogenic events in acute infection of rhesus monkeys with simian immunodeficiency virus of macaques. J. Virol. 68, 2362–2370 (1994). 5. Ogg, G.S. et al. Quantitation of HIV-1-specific cytotoxic T lymphocytes and plasma load of viral RNA. Science 279, 2103–2106 (1998). 6. Schmitz, J.E. et al. Control of viremia in simian immunodeficiency virus infection by CD8+ lymphocytes. Science 283, 857–860 (1999). 7. Jin, X. et al. Dramatic rise in plasma viremia after CD8+ T cell depletion in simian immunodeficiency virus-infected macaques. J. Exp. Med. 189, 991–998 (1999). 8. Matano, T. et al. Administration of an anti-CD8 monoclonal antibody interferes with the clearance of chimeric simian/human immunodeficiency virus during primary infections of rhesus macaques. J. Virol. 72, 164–169 (1998). 9. Couillin, I. et al. Impaired cytotoxic T lymphocyte recognition due to genetic variations in the main immunogenic region of the human immunodeficiency virus 1 NEF protein. J. Exp. Med. 180, 1129–34 (1994). 10. Goulder, P.J.R. et al. Late escape from an immunodominant cytotoxic T-lymphocyte response associated with progression to AIDS. Nature Med. 3, 212–217 (1997). 11. McMichael, A.J. & Phillips, R.E. Escape of human immunodeficiency virus from immune control. Annu. Rev. Immunol. 15, 271–296 (1997). 12. Borrow, P. et al. Antiviral pressure exerted by HIV-1-specific cytotoxic T lympho-
1276
cytes (CTLs) during primary infection demonstrated by rapid selection of CTL escape virus. Nature Med. 3, 205–211 (1997). 13. Wolinsky, S.M. et al. Adaptive evolution of human immunodeficiency virus-type 1 during the natural course of infection. Science 272, 537–542 (1996). 14. Koenig, S. et al. Transfer of HIV-1-specific cytotoxic T lymphocytes to an AIDS patient leads to selection for mutant HIV variants and subsequent disease progression. Nat. Med. 1, 330–336 (1995). 15. Dai, L.C., West, K., Littaua, R., Takahashi, K. & Ennis, F.A. Mutation of human immunodeficiency virus type 1 at amino acid 585 on gp41 results in loss of killing by CD8+ A24-restricted cytotoxic T lymphocytes. J. Virol. 66, 3151 (1992). 16. Haas, G. et al. Dynamics of viral variants in HIV-1 Nef and specific cytotoxic T lymphocytes in vivo. J. Immunol. 157, 4212–4221 (1996). 17. Mortara, L. et al. Selection of virus variants and emergence of virus escape mutants after immunization with an epitope vaccine. J. Virol. 72, 1403–1410 (1998). 18. Price, D.A. et al. Positive selection of HIV-1 cytotoxic T lymphocyte escape variants during primary infection. Proc. Natl. Acad. Sci. USA 94, 1890–1895 (1997). 19. Balter, M. Modest Briton stirs up storm with views on role of CTLs. Science 280, 1860–1861 (1998). 20. Brander, C. et al. Lack of strong immune selection pressure by the immunodominant, HLA-A*0201-restricted cytotoxic T lymphocyte response in chronic human immunodeficiency virus-1 infection. J. Clin. Invest. 101, 2559–2566 (1998). 21. Dykhuizen, M. et al. Determinants of disease in the simian immunodeficiency virus-infected rhesus macaque: characterizing animals with low antibody responses and rapid progression. J. Gen. Virol. 79, 2461–2467 (1998). 22. Pauza, C.D. et al. Pathogenesis of SIVmac251 after atraumatic inoculation of the rectal mucosa in rhesus monkeys. J. Med. Primatol. 22, 154–61 (1993). 23. Coffin, J.M. HIV population dynamics in vivo: Implications for genetic variation, pathogenesis, and therapy. Science 267, 483–489 (1995). 24. Nei, M. & Gojobori, T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Molec. Biol. Evol. 3, 418–426 (1986). 25. Hughes, A.L. & Nei, M. Pattern of nucleotide substitution at major histocompatibility complex class I loci reveals overdominant selection. Nature 335, 167–170 (1988). 26. delGuercio, M.F. et al. Binding of a peptide antigen to multiple HLA alleles allows definition of an A2-like supertype. J. Immunol. 154, 685–693 (1995). 27. Koup, R.A. et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J. Virol. 68, 4650–4655 (1994). 28. Walker, B.D. et al. HIV-specific cytotoxic T lymphocytes in seropositive individuals. Nature 328, 345–348 (1987). 29. Koup, R.A. et al. Limiting dilution analysis of cytotoxic T lymphocytes to human immunodeficiency virus gag antigens in infected persons: In vitro quantitation of effector cell populations with p17 and p24 specificities. J. Exp. Med. 174, 1593–1600 (1991). 30. Carmichael, A., Jin, X., Sissons, P. & Borysiewicz, L. Quantitative analysis of the human immunodeficiency virus type 1 (HIV-1)-specific cytotoxic T lymphocyte (CTL) response at different stages of HIV-1 infection: Differential CTL responses to HIV-1 and Epstein-Barr virus in late disease. J. Exp. Med. 177, 249–256 (1993). 31. Mellors, J.W. et al. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 272, 1167–1170 (1996). 32. McMichael, A. T cell responses and viral escape. Cell 93, 673–676 (1998). 33. Trivedi, P. et al. Selective amplification of simian immunodeficiency virus genotypes after intrarectal inoculation of rhesus monkeys. J. Virol. 62, 7649–7653 (1994). 34. van Baalen, C.A. et al. Selective in vitro expansion of HLA class I-restricted HIV-1 gag-specific CD8+ T cells: Cytotoxic T-lymphocyte epitopes and precursor frequencies. AIDS 7, 781–786 (1993). 35. Voss, G. et al. Human immunodeficiency virus type 1 envelope glycoprotein-specific cytotoxic T lymphocytes in simian-human immunodeficiency virus-infected rhesus monkeys. Virology 208, 770–775 (1995). 36. Allen, T.M. et al. Characterization of the peptide-binding motif of a rhesus MHC class I molecule (Mamu-A*01) that binds an immunodominant CTL epitope from SIV. J. Immunol. 160, 6062–6071 (1998). 37. Greenwood, F., Hunter, W. & Glover, J. The preparation of 131I-labeled human growth hormone of high specific radioactivity. Biochem. J. 89, 114 (1963).
NATURE MEDICINE • VOLUME 5 • NUMBER 11 • NOVEMBER 1999
