Journal of General Virology (1998), 79, 2191–2201.

Printed in Great Britain

...................................................................................................................................................................................................................................................................................

Characterization of HLA-B57-restricted human

immunodeficiency virus type 1 Gag- and RT-specific cytotoxic T
lymphocyte responses
Miche' l R. Klein,1 Sjoerd H. van der Burg,2 Egbert Hovenkamp,1 Agnes M. Holwerda,1
Jan Wouter Drijfhout,2 Cornelis J. M. Melief 2 and Frank Miedema1, 3
1
Department of Clinical Viro-Immunology, Central Laboratory of the Netherlands Red Cross Blood Transfusion Service and Laboratory of
Experimental and Clinical Immunology, University of Amsterdam, Plesmanlaan 125, 1066 CX Amsterdam, The Netherlands
2
Department of Immunohaematology and Blood Bank, Leiden University Hospital, Leiden, The Netherlands
3
Department of Human Retrovirology, Academic Medical Center, Amsterdam, The Netherlands

HLA-B57 has been shown to be strongly associated epitope could be recognized in the context of both
with slow disease progression in human immuno- HLA-B*5701 and HLA-B*5801. Interestingly, three
deficiency virus type 1 (HIV-1)-infected patients epitope variants of IVLPEKDSW were observed,
from the Amsterdam Cohort. Since HIV-1-specific which coincided with the strongest detectable CTL
CTL can control and eliminate virus-infected cells, response to RT. One variant (T2E7) was not recog-
we sought to characterize the dominant HLA-B57- nized by IVLPEKDSW-specific CTL despite the fact
restricted CTL responses at the epitope level. It was that this variant bound to HLA-B*5701 with a similar
found that HLA-B57-restricted CTL responses were affinity as the index peptide. Finally, only viruses
targeted at multiple proteins of HIV-1, with CTL which contained the epitope index sequence were
specific for Gag and RT being the most pronounced. obtained suggesting efficient virus control by CTL.
Gag-specific CTL recognized peptides ISPRTLNAW In conclusion, we report the characterization of
(aa 147–155) and STLQEQIGW (aa 241–249), dominant HIV-1 Gag- and RT-derived, HLA-B57-
which had previously been reported as HLA-B57- restricted CTL epitopes which are associated with
restricted. The RT-specific CTL response in one long- longer time to AIDS. Further characterization of
term survivor studied in great detail persisted for CTL responses restricted by HLA-B57 and other
" 10 years and was dominated by HLA-B57- protective HLA alleles may contribute to the
restricted CTL that recognized the newly defined development of effective AIDS vaccines.
epitope IVLPEKDSW (RTLAI, aa 244–252). This

Introduction long-term survivors appears heterogeneous and consists

The clinical course and outcome of human immuno- largely of long-term asymptomatic patients ; however, in most
deficiency virus type 1 (HIV-1) infection are highly variable. cases minor signs of disease progression can be observed.
The time to AIDS appears to follow a quasi-Gaussian These individuals may harbour peculiar combinations of virus
distribution as a consequence of multiphasic and multifactorial characteristics, host genetic determinants, antiviral immune
virus–host interactions (Klein & Miedema, 1995 ; Haynes et al., responses and environmental co-factors that could favour slow
1996). In a small proportion of HIV-1-infected individuals, who disease progression. Among the group of long-term survivors,
seem to represent the extreme of the right-hand tail of the there may also be some individuals who are protected from
distribution, an extraordinarily benign disease course beyond progression to AIDS for life. It seems likely that such a rare
the median time to AIDS is observed. This group of so-called phenomenon can only result from unique features not observed
in other patients progressing to AIDS (Klein & Miedema,
1995 ; Haynes et al., 1996).
Author for correspondence : Miche' l Klein. Present address : Medical It is widely held that HIV-1-specific CTL are among the
Research Council Lab., Fajara, PO Box 273, Banjul, The Gambia. favourable determinants for delaying disease progression,
Fax ­220 496513. e-mail petiet!2-cool.com since in vitro experiments have shown that HIV-1-specific CTL

0001-5570 # 1998 SGM CBJB

M. R. Klein and others
kill infected target cells via MHC class I-restricted recognition
(Plata et al., 1987 ; Walker et al., 1987) and can suppress virus
replication via secretion of various antiviral cytokines (Walker
et al., 1986 ; Tsubota et al., 1989 ; Cocchi et al., 1995 ; Buseyne
et al., 1996). This view is further underscored by in vivo
associations of emerging HIV-1-specific CTL responses with
the profound reduction of viraemia during the acute phase of
infection (Borrow et al., 1994, 1997 ; Koup et al., 1994 ; Price et
al., 1997) and with the relatively sustained control of viraemia
during the asymptomatic phase (Klein et al., 1995). Interest-
ingly, several HLA alleles are consistently associated with
either rapid disease progression (Kaslow et al., 1990 ; Itescu et
al., 1992 ; Sahmoud et al., 1993 ; Klein et al., 1994) or longer
time to AIDS (Klein & Miedema, 1995 ; Goulder et al., 1996 ;
Haynes et al., 1996 ; Kaslow et al., 1996). This suggests that
there may be qualitative differences in MHC class I-restricted
HIV-1-specific CTL responses that could variably impact on
the rate of virus replication and hence on time to AIDS.
Characterization of HIV-1 CTL responses restricted by these
so-called risk and protective HLA alleles should be one of the
first steps taken to reveal the molecular basis of the correlates
of immune protection (Harrer et al., 1996 ; Van der Burg et al.,
1997).
Since the association between HLA-B57 and long-term
survival appeared to be the strongest correlate of slow disease
progression among patients from the Amsterdam Cohort and
independently confirmed observations in other cohort studies
(Haynes et al., 1996 ; Kaslow et al., 1996), we sought to
characterize the dominant HLA-B57-restricted CTL responses
to HIV-1 Gag and RT proteins at the epitope level.
Methods
+ Study population
Nested case-control study on HLA and long-term survival of HIV-1 in-
fection. Within the Amsterdam Cohort Studies on HIV-1 infection and
AIDS, a nested case-control study was designed to examine the re-
lationship between host genetics and duration of HIV-1 infection.
Long-term survivors (cases) were HIV-1-seropositive homosexual men
(n ¯ 23) with : (i) at least 9 years of active follow-up ; (ii) no clinical
AIDS-defining conditions according to the Centers for Disease Control
and Prevention (1993) classifications ; and (iii) mean CD4+ T cell counts
" 400 cells}µl. Control men (n ¯ 86) were randomly selected from
199 HIV-1-seropositive participants who developed clinical AIDS within
8 years after seroconversion or seropositive entry in the Cohort Study.
Local Caucasian controls (n ¯ 804) were unrelated healthy male blood
donors (Klein et al., 1994).
Case report of two patients from the Amsterdam Cohort. Patient
H090 is a Caucasian homosexual man with HLA type : A1,*0201 ;
B41,*5701 ; Cw*0602,*1701 ; DR13 ; DR52 ; DQ1,3 (Klein et al., 1995).
This patient was selected because he displayed the most stable disease
course among the 238 HIV-1-seropositive participants recruited be-
tween October 1984 and May 1985, and was among the 32 men who
seroconverted for HIV-1 during that same period. Patient H090 entered
the Cohort Study in November 1984 and was found seropositive for
CBJC
HIV-1 at the next visit in February 1985. He subsequently attended
the Municipal Health Service in Amsterdam at three-monthly intervals
(at least 50 visits) until October 1997 and so far he has remained
asymptomatic. Standard laboratory measurements were routinely per-
formed and were found to be in the normal range : CD4+ T cell counts
(mean³SD), 860³190 cells}µl ; CD8+ T cell counts, 520³170
cells}µl ; and CD4+}CD8+ T cell ratio, 1±7³0±37 (Klein et al., 1995).
HIV-1 RNA was always undetectable in serum samples (! 10$ RNA
copies}ml). HIV-1 could only be isolated from 3 out of 16 different
blood samples tested with a frequency of ! 1 tissue culture infectious
dose per 10' CD4+ T cells.
Patient H433 was a Caucasian homosexual man with HLA type :
A2,3 ; B7,*5701 ; Cw6,7 ; DR1,3 ; DR52 ; DQ1,2. He was found to be HIV-
1-seropositive when he entered the Cohort Study in January 1985. After
7 years, he developed a generalized cytomegalovirus infection of which
he died. Since six months before AIDS diagnosis his CD4+ T cell counts
were not severely decreased (310 cells}µl), no anti-retroviral therapy was
considered at that time. This patient is one of three HLA-B57-positive
participants from the Cohort who have developed AIDS so far.
Both long-term survivor H090 and progressor H433 did not carry the
32 bp deletion in the C-C chemokine receptor-5 (CCR5) gene (De Roda
Husman et al., 1997), and both individuals were infected with non-
syncytium-inducing HIV-1 variants only as determined in the MT2 assay
(Koot et al., 1993).
+ In vitro restimulation and expansion of HIV-1-specific CTL.
HIV-1-specific CTL in PBMC samples of HIV-1-seropositive individuals
were expanded in vitro using an antigen-specific stimulation protocol as
previously described (Van Baalen et al., 1993 ; Klein et al., 1995). This
method results in the generation and expansion of CTL which are biased
towards recognition of HIV-1LAI-derived sequences, sequences highly
similar to clade B or highly conserved epitopes among all clades of
HIV-1.
T cells were stimulated with autologous Epstein–Barr virus-trans-
formed B-lymphoblastoid cell lines (B-LCL) infected at an m.o.i. of 5 with
recombinant vaccinia viruses (rVV) expressing single genes of HIV-1LAI
(Myers et al., 1995). Cells were harvested after 20–22 h infection at 37 °C
and 5 % CO , and fixed with 1 % (w}v) paraformaldehyde for 15 min at
#
room temperature. Subsequently, fixed cells were incubated with 0±2 M
glycine in PBS for 15 min and washed once with complete medium
(RPMI 1640 ; Life Technologies) containing antibiotics and heat-
inactivated, pooled human serum (10 %) from 8–10 healthy HIV-1-
seronegative blood donors. Fixed stimulator cells were stored at 4 °C and
kept for a maximum period of 4 weeks.
HIV-1-specific CTL responses were quantified using standard
methods for limiting dilution analysis (Lefkovits & Waldmann, 1984).
Briefly, 24 replicate wells of eight serial dilutions of PBMC ranging from
20 000 to 745 cells per well, were co-cultivated in 96-well round-bottom
microtitre-plates with 10% fixed stimulator cells and 10% irradiated (30 Gy)
autologous PBMC per well (100 µl). At days 2 and 9, micro-cultures were
fed with complete medium supplemented with recombinant IL-2 (rIL-2 ;
10 U}ml) (Proleukin ; kindly provided by R. Rombouts, Chiron Benelux
BV, Amsterdam, The Netherlands). The cultures were restimulated at day
7 with complete medium supplemented with rIL-2 (10 U}ml) and 10%
fixed stimulator cells. On day 15, wells were split and tested for
cytotoxicity (see below).
Cultures from positive wells were further expanded in 96-well round-
bottom microtitre-plates by non-specific restimulation as previously
described (Van de Griend et al., 1984). Briefly, HIV-1-specific CTL bulk
cultures or clones were co-cultivated with a cocktail of (i) 2¬10' per ml
irradiated (30 Gy) PBMC of 3–5 healthy HIV-1-seronegative blood
donors and (ii) 1±5¬10' per ml irradiated (50 Gy) B-LCL (mixture of
 HLA-B57-restricted CTL and slow progression to AIDS

equal amounts of allogeneic B-LCL APD, BSM and JY) in complete Table 1. Associations between HLA and long-term
medium supplemented with rIL-2 (10 U}ml) and leucoagglutinin (Leuco survival of HIV-1 infection
A, 1 µg}ml ; Sigma). When sufficient amounts of cells were obtained,
effector cells were tested for cytotoxicity at various effector : target ratios HLA phenotype frequencies are given.
(see below).
+ Recombinant vaccinia viruses (rVV). Recombinant vaccinia Long-term Local
viruses (rVV) have been constructed from the Copenhagen strain of survivors AIDS cases controls
vaccinia virus. The following rVV expressing single genes of HIV-1LAI HLA (n ¯ 23) (n ¯ 86) RR* P (n ¯ 804)
(Myers et al., 1995) were used for this study : TG.4163 (RT) ; TG.1144
(Gag) (Rautmann et al., 1989) ; TG.3183 (Env) (McChesney et al., 1990) ; Cw6 50±0 12±5 6±5 0±007 20±0
TG.1147 (Nef) (Guy et al., 1987) ; TG.3196 (Tat) ; TG.4113 (Rev) ; Bw4 78±3 43±8 4±3 0±005 52±0
TG.1160 (Vif) ; and control rVV 186poly, containing no specific insert. B57 26±1 2±3 12±6 0±0006 5±2
M. P. Kieny (Transge' ne SA, Strasbourg, France) and Y. Rivie' re (Institut B27 13±0 3±5 4±1 † 8±4
Pasteur, Paris, France) kindly provided all the rVV. B51 21±7 11±8 2±1  10±3
B18 8±7 9±3 1±1  7±2
+ Peptides. Peptides were synthesized by solid-phase strategies on an
automated peptide synthesizer (Abimed AMS 422) using Fmoc chem-
istry. Peptides were analysed by reverse-phase HPLC ; they were * Relative risks (RR) were odds ratios according to Haldane (1955). RR
dissolved in DMSO at 10–50 mg}ml, aliquotted and stored at ®70 °C and corresponding probability (P) indicate the likelihood of long-term
until use. survival relative to progression to AIDS in the presence of particular
HLA alleles. The Chi-square (χ#) test was used to evaluate whether the
+ 51Cr-release assays. Standard &"Cr-release assays were performed
calculated RR differed significantly from unity.
as previously described (Van Baalen et al., 1993 ; Klein et al., 1995 ; Van
, Not significant ; indicates that the RR did not differ from unity
der Burg et al., 1995 a). Briefly, autologous or (partially) HLA-matched B- [taken at the 5 % level except where marked (†, P ¯ 0±06)].
LCL were infected at an m.o.i. of 5 with rVV expressing single genes of
HIV-1LAI or control rVV 186poly and labelled with 100 µCi Na &"CrO
# %
(Amersham) for 16 h at 37 °C and 5 % CO . After three wash steps, 4 000
# Caucasian populations differed significantly between the total
target cells were added to each well. Alternatively, &"Cr-labelled and
uninfected B-LCL were pre-incubated for 30 min at room temperature HIV-1-infected population studied here and the local healthy
with synthetic peptides at a final concentration of 5–10 µg}ml or in 10- control group (data not shown) (Klein et al., 1994). The Bw4
fold serial dilutions in titration experiments as indicated. After 4 h, group of HLA-B alleles was observed more frequently among
supernatants were harvested and radioactivity was measured with a γ- long-term survivors compared to subjects who developed
counter (Cobra-II ; Canberra Packard Benelux). &"Cr release was expressed AIDS. This association was mainly due to a significantly
as specific lysis (%) using the following formula : ([experimental release® increased frequency of HLA-B57 (P ¯ 0±0006) and, to a lesser
spontaneous release]}[maximum release®spontaneous release])¬100.
Spontaneous &"Cr release was always ! 15 % of the maximum release.
extent, a slightly increased frequency of HLA-B27 (P ¯ 0±06)
Wells were considered positive when &"Cr release exceeded 10 % specific (Table 1). In addition, HLA-Cw6 was also over-represented in
lysis. This arbitrary threshold was always greater than the the group of long-term survivors as a result of linkage
mean­(3¬SD). CTL precursor (CTLp) frequencies were calculated disequilibrium with HLA-B57. Furthermore, similar associa-
using methods previously described (Strijbosch et al., 1987). CTLp tions were observed in a prospective study of 148 cohort
frequencies for specific antigens were computed as differences between participants with a known date of HIV-1 seroconversion.
CTLp frequencies determined on specific versus control targets. Kaplan–Meier survival analysis showed in this case that
individuals with HLA-B57 had a relative risk of 0±163 (P ¯
0±07) for developing AIDS compared to individuals without
Results HLA-B57 (data not shown). Since the association between
Association of HLA-B57 with long-term AIDS-free HLA-B57 and long-term survival appeared to be the strongest
survival correlate in the Amsterdam Cohort and independently con-
To address which HLA alleles contribute to delayed disease firmed observations made in other cohort studies (Kaslow et al.,
progression in HIV-1 infection, HLA phenotype frequencies 1996 ; Goulder et al., 1996), we sought to further characterize
were analysed in a nested case-control study among homo- the HLA-B57-restricted HIV-1-specific CTL responses.
sexual men from the Amsterdam Cohort. Long-term survivors Participant H090 was selected for further study since he
(cases) were HIV-1-seropositive participants who remained displayed the most stable disease course among the persons
AIDS-free for over 9 years (median 10±8 years, range 9±2–11±1) recruited between October 1984 and April 1985. Despite more
with normal CD4+ T cell counts (mean 538 cells}µl, range than 12 years of documented HIV-1 infection, virus load was
408–953 in the ninth year of follow-up). Control subjects were very low to undetectable and CD4+ T cell counts remained
HIV-1-seropositive participants who developed clinical AIDS stable and in the normal range of healthy individuals (Klein et
within 8 years after seroconversion or seroprevalent entry in al., 1995). Patient H433 was also selected for this study because
the study (median 3±7 years, range 0±8–7±9). None of the HLA he was one of three HLA-B57-positive patients who had
alleles which are frequently present in north European developed AIDS at the start of this study. Patient H433 was

CBJD
M. R. Klein and others

(a) (b)
Relative HIV-1LAI CTLp frequency

Fig. 1. Distribution of HLA-B57-restricted CTL responses. Using limiting dilution analysis, CTLp frequencies for Gag, RT, Env,
Tat, Rev, Vif and Nef were quantified in a blood sample from February 1990 of patient H090 (a) and from April 1988 of
patient H433 (b). The contribution of HLA-B57-restricted CTL responses was determined in parallel using HLA-B*5701-
matched B-LCL as targets in a split-well 51Cr-release assay. For each protein, the relative contribution of HLA-B57-restricted
CTL (+) and the remaining CTL (*) are expressed as a fraction of the total HIV-1-specific CTLp frequency, which is
determined by summation of the individual CTLp frequencies for the seven proteins tested.

(a) (b)

Specific lysis (%) Specific lysis (%)

Fig. 2. Immunodominant HLA-B57-restricted GagLAI-specific CTL epitopes. Bulk CTL cultures of patient H090 (a) and patient
H433 (b) were specifically stimulated for HIV-1 GagLAI. Cytotoxicity was tested in a standard 51Cr-release assay with
autologous B-LCL pulsed with 5 µg/ml of the indicated Gag-derived 9-mer peptides carrying the motif for binding to HLA-
B*5701. Effector cells were added to target cells in a ratio of 5 : 1.

HIV-1-seropositive when he entered the Cohort Study in analysis of a blood sample from long-term survivor H090
January 1985 and died of AIDS 7 years later. obtained in February 1990. In addition, the contribution of
HLA-B57-restricted CTL responses was determined in parallel
with HLA-B*5701-matched B-LCL as targets in split-well &"Cr-
HLA-B57-restricted CTL responses against multiple release assays (Fig. 1 a). In agreement with our previous
HIV-1 proteins observations, dominant CTL responses were targeted at Gag,
HIV-1-specific CTL responses against Gag, RT, Env, Tat, RT and Tat (Klein et al., 1995 ; Van der Burg et al., 1997 ; Van
Rev, Vif and Nef were quantified using limiting dilution Baalen et al., 1997). Minor responses were detected against

CBJE
 HLA-B57-restricted CTL and slow progression to AIDS

(a) (b)

(c) (d)

Fig. 3. Characterization of RTLAI-specific CTL from long-term survivor H090. (a) HLA-restriction of RTLAI-specific CTL. Effector
cells were obtained from cultures of T cell lines with consistent specificity for HIV-1 RTLAI-expressing target cells. Effector cells
were tested for cytotoxicity on autologous or allogeneic B-LCL matched at one or more HLA class I allele and infected with rVV.
Effector cells were added to target cells in a ratio of 2±5 : 1. Specific lysis was determined by subtracting lysis of targets infected
with rVV 186poly (control) from lysis of those infected with rVV TG.4163 (RTLAI). (b) Mapping of epitope specificity of RTLAI-
specific CTL. Target cells were prepared by incubating autologous B-LCL with 5 µg/ml of the indicated synthetic 10-mer
peptides that were derived from the set of overlapping RT peptides. Effector cells were added to target cells in a ratio of 5 : 1.
(c) Titration of 9- and 10-mer peptides recognized by RTLAI-specific CTL. Effector cells were tested for their ability to lyse
autologous B-LCL infected with rVV TG.4163 (RTLAI) or rVV 186poly (control) or pulsed with varying concentrations of peptide
p244/9 IVLPEKDSW (E) or p243 PIVLPEKDSW (+). Effector cells were added to target cells in a ratio of 5 : 1. Horizontal
dotted line indicates half-maximum lysis. (d) Recognition of RTLAI IVLPEKDSW in the context of HLA-B17 alleles. RTLAI
IVLPEKDSW-specific CTL from patient H090 were tested for their ability to recognize autologous HLA-B*5701-positive B-LCL
(E) or HLA-mismatched HLA-B*5801-positive B-LCL (+) pulsed with various concentrations of peptide p244/9 IVLPEKDSW.
Effector cells were added to target cells in a ratio of 2±5 : 1.

Env, Nef, Rev and Vif. The contribution of HLA-B*5701-
restricted CTL was 50 % in the case of Gag-specific CTL and
85 % in the case of RT-specific CTL. HLA-B*5701-restricted
CTL responses against the other proteins tested represented
only minor components of the total HIV-1-specific CTL
response (Fig. 1 a).
Similar experiments were carried out for patient H433 who
developed AIDS in about 7 years. For this purpose, a blood
sample from April 1988 was analysed which was obtained
about 3 years after he entered the Cohort Study. At that time,
he was still asymptomatic and both CD4+ T cell numbers and
the CD4+}CD8+ T cell ratio were only slightly decreased (450
cells}µl and 0±31, respectively). Dominant CTL responses of
progressor H433 comprised CTL against Gag (2165 per 10'
PBMC) followed by the responses to Env and Nef (both 1180
per 10' PBMC) and a minor response to RT (157 per 10'
PBMC) (Fig. 1 b). Using HLA-matched targets, it was de-
termined that 84 % of the CTL response against Gag was HLA-
B*5701-restricted. CTL responses against Env and Nef con-
sisted of about 20 % of HLA-B57-restricted CTL. The fre-
quency of HLA-B57-restricted RT-specific CTL was 104 per
10' PBMC and no CTL responses were detected against Tat,
Rev and Vif (Fig. 1 b). Since HLA-B57-restricted HIV-1 Gag-
and RT-specific CTL constituted dominant responses in the

CBJF
M. R. Klein and others
long-term survivor with the most stable disease course, we
decided to characterize these further at the epitope level.
Identification of HLA-B57-restricted Gag-specific CTL
epitopes
The Gag-specific CTL epitopes were identified with a series
of peptides containing the HLA-B57 peptide-binding motif
(Falk et al., 1995 ; Barber et al., 1997). Despite the fact that a
phenylalanine has been reported as the C-terminal anchor
residue (Goulder et al., 1996) and that this residue has been
found in some endogenous peptide sequences eluted from
HLA-B17 alleles (Falk et al., 1995 ; Barber et al., 1997), we
decided to focus on a motif with tryptophan at position 9. The
rationale behind this was based on previous observations in
our laboratory using functional binding assays (Van der Burg
et al., 1995 b) which showed that the alternative anchor residue
(phenylalanine) was less able to promote binding to HLA-
B*5701 than tryptophan (not shown).
Gag-specific CTL bulk cultures were expanded from a
blood sample of patient H090 from August 1988, and were
subsequently tested against the panel of Gag-derived HLA-
B57 motif-bearing peptides. We found that peptide GagLAI
STLQEQIGW (aa 241–249) was recognized by CTL from this
patient (Fig. 2 a). Positive wells from the limiting dilution
experiments of patient H433 that showed consistent cytolytic
activity for GagLAI were also non-specifically expanded in
vitro. Bulk cultures were subsequently tested on the panel of 9-
mer peptides and strongly recognized peptide STLQEQIGW
(aa 241–249) and, to lesser extent, peptide ISPRTLNAW (aa
147–155) (Fig. 2 b). Both peptides are highly conserved among
the published HIV-1 sequences (Myers et al., 1995) and have
also previously been observed as being part of dominant HLA-
B57-restricted CTL responses to HIV-1 (Johnson et al., 1991 ;
Goulder et al., 1996).
Characterization of a novel HLA-B57-restricted
epitope in HIV-1 RT
From a blood sample of subject H090 obtained in October
1986, five oligoclonal cytotoxic T cell lines were obtained
which showed consistent specificity for HIV-1LAI RT. These
CTL were CD8+ and granzyme B-positive (data not shown).
Using B-LCL that were partially matched for the HLA type of
patient H090, it was shown that HLA-B*5701 was the
restriction element for all of these RT-specific CTL (Fig. 3 a).
The CTL epitope specificity was determined using a set of 536
consecutive 10-mer peptides each with a nine residue overlap
and spanning the entire sequence of HIV-1 RTLAI (Van der
Burg et al., 1997) (Fig. 3 b). It was shown that peptides 243–245
were recognized by all RT-specific CTL from patient H090,
with RT peptide p243 (aa 243–252 ; PIVLPEKDSW) being the
optimal. Of note, RT-specific CTL from another HLA-B57-
positive long-term survivor previously studied in our lab-
oratory also recognized the 10-mer peptides p243 and p244
(Van der Burg et al., 1997).
CBJG
Specific lysis (%)
Fig. 4. HLA-B57-restricted RT-specific CTL from progressor H433.
Effector cells were derived from bulk CTL cultures specifically stimulated
for GagLAI and added to target cells in a ratio of 5 : 1. Target cells were
autologous B-LCL pulsed with 5 µg/ml of the indicated RT-derived 9-mer
peptides.
Both the 10-mer sequence (p243) and the peptide sequence
lacking the N-terminal proline (p244}9) were titrated in order
to assess the optimal epitope length (Fig. 3 c). The 9-mer
sequence IVLPEKDSW was recognized at much lower con-
centrations than the 10-mer peptide p243 (median value half-
max. killing 4±7 nM vs 251 nM, respectively). Therefore, it was
concluded that HIV-1 RTLAI peptide p244}9 (aa 244–252 ;
IVLPEKDSW) constituted the optimal epitope sequence.
Positive wells from the limiting dilution experiments of
patient H433 which showed consistent cytolytic activity for
HIV-1 RTLAI were non-specifically expanded in vitro. Bulk
cultures were subsequently tested on the panel of RT-derived
HLA-B57 motif-bearing 9-mer peptides. Bulk cultures of RT-
specific CTL showed consistent reactivity toward peptides
IVLPEKDSW and RTLAI (aa 375–383 ; ITTESIVIW) (Fig. 4).
Of note, two 10-mer peptides containing the latter sequence
were also recognized by another HLA-B57 patient previously
studied in our laboratory (Van der Burg et al., 1997).
According to reported similarities in peptide-binding motifs
for HLA-B17 alleles (Falk et al., 1995 ; Barber et al., 1997), it is
to be expected that a number of HLA-B57-restricted CTL
epitopes will also be recognized in the context of HLA-B58
alleles (Goulder et al., 1996). Similarly to the situation
demonstrated for peptide HIV-1 Gag TSTLQEQIGW
(Goulder et al., 1996 ; Bertoletti et al., 1998), we found here that
HLA-B*5701-restricted RT-specific CTL from patient H090
could also recognize peptide RTLAI IVLPEKDSW in the
context of HLA-B*5801 (Fig. 3 d).
Longitudinal analysis of RT-specific CTL and HIV-1
variants
The kinetics of the RT-specific CTL responses of long-term
survivor H090 were studied at the epitope level (Fig. 5). The
RT-specific CTL response reached peak levels 44 months after

Fig. 5. Kinetics of RTLAI-specific CTL responses at epitope level. Using
limiting dilution analysis, the CTLp frequencies for RT were quantified in
six blood samples of long-term survivor H090. RT-specific CTLp
frequencies (+) were determined by subtracting CTLp frequencies
determined with targets infected with rVV 186poly (control) from CTLp
frequencies calculated with targets infected with rVV TG.4163 (RTLAI). In
addition, the contribution of CTL specific for RTLAI (aa 244–252)
IVLPEKDSW (E) was determined in parallel using peptide-pulsed HLA-
B*5701-matched B-LCL as targets in a split-well 51Cr-release assay.
Epitope-specific CTLp frequencies were calculated by subtracting CTLp
frequencies determined with targets without peptide from CTLp
frequencies determined with targets pulsed with 10 µg/ml RTLAI peptide
p244/9 IVLPEKDSW (index). In addition, observed virus variants of this
epitope are boxed and time-points of successful virus isolation are
indicated with a down-pointing arrow. Ranges at each point represent the
standard error of the calculated frequencies.
HIV-1 seroconversion (frequency³standard error is 1132³
282 CTLp per 10' PBMC). Thereafter, CTLp frequencies
gradually declined, but clearly remained detectable until 10
years after seroconversion (160³74 per 10' PBMC). Sim-
ultaneous testing of RT peptide p244}9 revealed that CTL
directed against the HLA-B*5701-restricted epitope IVLPE-
KDSW dominated the RT-specific CTL response for many
years (Fig. 5). Although the CTL response against this epitope
showed largely the same kinetics as the total response against
the entire RT protein, a slight difference was observed at some
time-points. Separate experiments with HLA-matched targets
indeed showed a minor RT-specific CTL response (Fig. 1 a)
which appeared to be restricted by HLA-A*0201 (data not
shown).
To investigate the relationship between dominant HIV-1-
specific CTL responses and selective pressure on the virus
quasispecies, we decided to sequence the vicinity of epitope
IVLPEKDSW in replication-competent strains we had obtained
via biological cloning procedures (Schuitemaker et al., 1991).
HLA-B57-restricted CTL and slow progression to AIDS
Serial dilutions of variant peptides were tested for their ability
to inhibit binding of the fluorescein-labelled index peptide
p244}9 to HLA-B*5701 (Van der Burg et al., 1995 b). To
determine whether these peptides could still be recognized by
IVLPEKDSW-specific CTL, they were also tested in standard
&"Cr-release assays.
Due to the low numbers of infected CD4+ T cells in patient
H090, biological variants of HIV-1 could only be obtained
from three out of 16 different blood samples tested. Eight
clones were obtained from PBMC donated in February 1989,
a single clone was from October 1990, and eight clones were
from August 1994. With respect to the epitope sequence, the
isolates from February 1989 consisted of three variants : four
clones contained the index sequence of the epitope, three
clones harboured amino acid substitutions at positions 2 and 7
(V2 ! T2}D7 ! E7), and one clone was observed with a
single mutation at position 2 (V2 ! M2) (Fig. 5). Interestingly,
we did not observe any sequence variation in the stretch of 30
nucleotides flanking either side of the epitope sequence (not
shown).
The clone from October 1990, and the eight remaining
isolates from August 1994 all contained the index sequence
IVLPEKDSW (Fig. 5). The M2 variant of p244}9 bound to
HLA-B*5701 with somewhat lower affinity than the index
sequence, but was very well recognized by IVLPEKDSW-
specific CTL of patient H090 (not shown). The T2E7 variant
was not recognized by IVLPEKDSW-specific CTL of patient
H090 (half-max. killing " 10 µM), despite the fact that this
variant bound to HLA-B*5701 with similar affinity (IC ¯
&!
6 µM) as the index peptide.
In total, seven virus isolates were obtained from patient
H433, four from April 1988 and three from October 1989.
Sequencing revealed that all biological isolates contained a
glutamic acid (E) at position 2 (V2 ! E2). The latter variant
was clearly recognized by IVLPEKDSW-specific CTL of patient
H090, albeit at a lower level than the index peptide p244}9
(not shown). In addition, previous binding experiments also
indicated that the E2 variant bound significantly less strongly
to HLA-B*5701 compared to the index peptide (not shown).
Discussion
To start unravelling the correlates of immune protection to
AIDS, we have analysed the distribution of HLA class I alleles
in a nested case-control study of long-term survivors and
progressors from the Amsterdam Cohort. HLA-B57 was
strongly associated with longer time to AIDS and inde-
pendently confirmed observations made in other study
populations (Klein et al., 1995 ; Goulder et al., 1996 ; Haynes et
al., 1996 ; Kaslow et al., 1996). Since MHC class I-restricted
CTL responses to HIV-1 can interfere with virus replication via
secretion of antiviral cytokines (Walker et al., 1986 ; Tsubota et
al., 1989 ; Cocchi et al., 1995 ; Buseyne et al., 1996) or via cell-
mediated cytotoxicity (Plata et al., 1987 ; Walker et al., 1987),
CBJH
M. R. Klein and others
we aimed at the identification of HLA-B57-restricted CTL
responses to HIV-1.
HLA-B57-restricted CTL responses were vigorous and
involved broad recognition of epitopes from multiple proteins
of HIV-1. Dominant responses appeared to be directed against
HIV-1 Gag and RT, but minor responses to Env, Nef, Tat, Rev,
and Vif were also observed. The major CTL responses directed
at Gag and RT were subsequently characterized at the epitope
level. The Gag-derived epitopes that were recognized by CTL
from the patients we studied involved two highly conserved
sequences, namely ISPRTLNAW (aa 147–155) and STLQE-
QIGW (aa 241–249). Previously, these epitopes have been
described by others as being part of dominant HLA-B57-
restricted CTL responses to HIV-1 (Johnson et al., 1991 ;
Goulder et al., 1996).
The major component of the RT-specific CTL response in
long-term survivor H090 was targeted at a newly defined
HLA-B57-restricted epitope IVLPEKDSW (aa 244–252) and
persisted for more than 10 years. These observations add to
the findings of Kalams et al. (1994) who studied clonotypes of
TcR-Vβ CDR3 sequences of a dominant CTL response to an
HLA-B14-restricted epitope in gp41 that persisted for more
than 5 years. However, it remains unclear whether our findings
here were due to persistent clones or persistent immuno-
dominance.
The epitope data obtained here using functional assays
clearly show that peptides with a valine at position 2, maybe
in conjunction with other auxiliary residues such as a proline at
position 4 (Falk et al., 1995 ; Barber et al., 1997), can clearly bind
to HLA-B*5701. So far, this has not been observed in the
sequences of endogenous peptides eluted from HLA-B17
alleles (Falk et al., 1995 ; Barber et al., 1997). From previous in
vitro binding assays in our laboratory, we also concluded that
a phenylalanine as C-terminal anchor residue was less efficient
than tryptophan (not shown) (Van der Burg et al., 1995 b). So
far, only one CTL epitope has been reported with a
phenylalanine as C-terminal anchor residue in contrast to at
least seven with a tryptophan (Johnson et al., 1991 ; Goulder et
al., 1996 ; Van der Burg et al., 1997).
There is accumulating evidence that HIV-1 escape mutants
emerge in situations where the CTL response is vigorously
directed at a single epitope (Koenig et al., 1995 ; Borrow et al.,
1997 ; Goulder et al., 1997 ; Price et al., 1997 ; Mortara et al.,
1998). In situations where CTL responses involve broad
recognition of many epitopes, virus escape mutants may be
observed only temporarily (Haas et al., 1996). Our data may
add to the latter situation since we noticed three distinct
variants of the IVLPEKDSW sequence at the time when the
strongest CTL response to RT was measured. One variant
(T2E7) was not recognized by CTL specific for the index
sequence despite the fact that it did bind to HLA-B*5701 with
similar affinity. Since a threonine (T) at position 2 is considered
to be one of the primary anchor residues for binding (Falk et al.,
1995 ; Barber et al., 1997), the lack of recognition of this variant
CBJI
is most likely due to the glutamic acid (E) at position 7.
Although only limited numbers of virus isolates could be
obtained from this individual and sequence variation could
have accumulated randomly, it appeared that only viruses with
the index sequence persisted. Of interest is the fact that
variation was confined to the epitope sequence as no mutations
were observed in 30 neighbouring nucleotides on either side.
As similarly demonstrated by Haas et al. (1996), our obser-
vations suggest that the TcR escape mutants may have been
eliminated from the virus quasispecies by the vigorous and
polyclonal CTL responses we observed in this patient.
In patient H433, we observed a very marginal HLA-B57-
restricted CTL response to RT. In all seven biological isolates
of HIV-1, we observed a mutation in the first anchor residue for
binding to HLA-B57 (i.e. E2). This situation may again point
towards virus escape, but could also suggest that this variant
sequence was not sufficient to elicit strong RT-specific CTL
responses. This observation also adds to the growing notion to
include virus sequencing data in order to fully appreciate
functional epitope information. The latter has been elegantly
demonstrated by Sipsas et al. (1997) who reported that the
actual epitope sequence in autologous viruses may be quite
different from the virus index sequence (e.g. HIV-1LAI) which
is widely used to screen for CTL activity.
What remains controversial is the fact that viruses with the
index sequence do persist in the face of vigorous and broadly
directed CTL responses. This strongly suggests that there may
be virus sanctuaries where HIV-1-specific CTL are less capable
of controlling infected cells (Klein et al., 1998). Furthermore, it
shows that mechanisms of immune protection to disease are
more complex than just the recognition of specific epitopes per
se (Van der Burg et al., 1997). HLA-B57-positive patient H433,
who developed AIDS after 7 years, illustrates this since his
CTL recognized four HLA-B57-restricted epitopes. Moreover,
these epitopes are shown here, as well as in other studies, to be
part of dominant CTL responses of long-term survivors
(Goulder et al., 1996 ; Van der Burg et al., 1997). The obvious
difference in disease course may be explained by taking into
account the longevity of these responses as well as the entire
clonal composition of dominant and sub-dominant HIV-1-
specific CTL. For example, in progressor H433, we did not
observe CTL responses to Tat, Rev and Vif and only a weak
response against RT was seen. It could well be that CTL
responses to regulatory proteins expressed early during the
virus life-cycle are more efficient in controlling virus load, since
they are thought to eliminate virus-infected cells before the
major release of virus progeny (Van Baalen et al., 1997).
For future AIDS vaccines, it may be worthwhile to include
the type of immune responses that mimic the dominant and
persistent responses observed in long-term survivors. With
respect to the HLA restriction elements of HIV-1-specific CTL
epitopes, these would ideally have to be alleles common in
ethnic populations with a high prevalence of HIV-1 infection.
Coincidentally, HLA-B17 alleles are quite common in particular

sub-Saharan populations living in areas with a high prevalence
of HIV-1 infection (Imanishi et al., 1992). As with the findings
for peptide Gag TSTLQEQIGW (Goulder et al., 1996 ;
Bertoletti et al., 1998), we observed here that the HLA-B*5701-
restricted epitope RTLAI IVLPEKDSW could also be recog-
nized by CTL in the context of HLA-B*5801. This finding
again underscores the striking similarity of the peptide-binding
motifs of the structurally related HLA-B17 alleles (Falk et al.,
1995 ; Barber et al., 1997). Unravelling of the molecular basis of
the correlates of immune protection, in particular the role of
HIV-1-specific CTL responses restricted by HLA-B57 as well
as other ‘ protective ’ HLA class I alleles, will hopefully
contribute to the further development of effective AIDS
vaccines.
The authors are greatly indebted to all cohort participants for their
continuous dedication to the study ; to Oscar Pontesilli, Susana Kerkhof-
Garde and Neeltje Kootstra for expert technical assistance ; to Willemien
Benckhuijsen for synthesis of peptides ; to Ana-Maria de Roda Husman
for CCR5 genotyping ; to Dr Y. Rivie' re and M. P. Kieny for all the
recombinant vaccinia viruses ; to Dr R. Rombouts for generously
providing Proleukin (rIL-2) ; to Dr N. M. Lardy and Dr M. Bunce for
performing HLA-typings ; and to Professor Andrew McMichael for
providing facilities for carrying out some of the experiments described.
This study was performed as part of the Amsterdam Cohort Studies
on HIV-1 infection and AIDS, a collaboration between the Municipal
Health Service, the Academic Medical Center and the Central Laboratory
of the Netherlands Red Cross Blood Transfusion Service, Amsterdam,
The Netherlands.
This study was financially supported by the Dutch AIDS Fund, the
Dutch Ministry of Public Health on advice of the Dutch Program
Committee of AIDS Research in the context of the National AIDS
Research Stimulation Program (PccAo 94014, 95017) and in part by The
Netherlands Organization for Scientific Research (NWO).
M. R. Klein and S. H. van der Burg contributed equally to this study.
References
Barber, L. D., Percival, L., Arnett, K. L., Gumperz, J. E., Chen, L. &
Parham, P. (1997). Polymorphisms in the α1 helix of the HLA-B heavy
chain can have an overriding influence on peptide-binding specificity.
Journal of Immunology 158, 1660.
Bertoletti, A., Cham, F., McAdam, S., Rostron, T., Rowland-Jones, S.,
Sabally, S., Corrah, T., Ariyoshi, K. & Whittle, H. (1998). Cytotoxic T
cells from human immunodeficiency virus type 2-infected patients
frequently cross-react with different human immunodeficiency virus type
1 clades. Journal of Virology 72, 2439–2448.
Borrow, P., Lewicki, H., Hahn, B. H., Shaw, G. M. & Oldstone, M. B. A.
(1994). Virus-specific CD8­ cytotoxic T-lymphocyte activity associa-
ted with control of viremia in primary human immunodeficiency virus
type 1 infection. Journal of Virology 68, 6103–6110.
Borrow, P., Lewicki, H., Wei, X., Horwitz, M. S., Peffer, N., Meyers, H.,
Nelson, J. A., Gairin, J. E., Hahn, B. H., Oldstone, M. B. A. & Shaw,
G. M. (1997). Antiviral pressure exerted by HIV-1 specific cytotoxic T
lymphocytes (CTLs) during primary infection demonstrated by rapid
selection of CTL escape virus. Nature Medicine 3, 205–211.
Buseyne, F., Fevrier, M., Garcia, S., Gougeon, M. L. & Rivie' re, Y.
(1996). Dual function of a human immunodeficiency virus (HIV)-specific
HLA-B57-restricted CTL and slow progression to AIDS
cytotoxic T lymphocyte clone : inhibition of HIV replication by nonlytic
mechanisms and lysis of HIV-infected CD4­ cells. Virology 225,
248–253.
Centers for Disease Control and Prevention (1993). 1993 revised
classification system for HIV infection and expanded surveillance case
definition for AIDS among adolescents and adults. Mortality and
Morbidity Weekly Report 41, 1–19.
Cocchi, F., DeVico, A. L., Garzino-Demo, A., Arya, S. K., Gallo, R. C. &
Lusso, P. (1995). Identification of RANTES, MIP-1α, and MIP-1β as the
major HIV-suppressive factors produced by CD8­ T cells. Science 270,
1811–1815.
De Roda Husman, A. M., Koot, M., Cornelissen, M., Keet, I. P. M.,
Brouwer, M., Broersen, S. M., Bakker, M., Roos, M. T. L., Prins, M., De
Wolf, F., Coutinho, R. A., Miedema, F., Goudsmit, J. & Schuitemaker, H.
(1997). Association between CCR5 genotype and the clinical course of
HIV-1 infection. Annals of Internal Medicine 127, 882–890.
Falk, K., Ro$ tzschke, O., Takiguchi, M., Gnau, V., Stevanovic, S., Jung,
G. & Rammensee, H. G. (1995). Peptide motifs of HLA-B58, B60, and
B62 molecules. Immunogenetics 41, 165–168.
Goulder, P. J. R., Bunce, M., Krausa, P., McIntyre, K., Crowley, S.,
Morgan, B., Edwards, A., Giangrande, P., Phillips, R. E. & McMichael,
A. J. (1996). Novel, cross-restricted, conserved, and immunodominant
cytotoxic T lymphocyte epitopes in slow progressors in HIV type-1
infection. AIDS Research and Human Retroviruses 12, 1691–1698.
Goulder, P. J. R., Phillips, R. E., Colbert, R. A., McAdam, S., Ogg, G.,
Nowak, M. A., Giangrande, P., Luzzi, G., Morgan, B., Edwards, A.,
McMichael, A. J. & Rowland-Jones, S. (1997). Late escape from an
immunodominant cytotoxic T-lymphocyte response associated with
progression to AIDS. Nature Medicine 3, 212–217.
Guy, B., Kieny, M. P., Rivie' re, Y., LePeuch, C., Dott, K., Girard, M.,
Montagnier, L. & Lecocq, J. P. (1987). HIV F}3« orf encodes a
phosphorylated GTP-binding protein resembling an oncogene product.
Nature 330, 266–269.
Haas, G., Plikat, U., Debre! , P., Lucchiari, M., Katlama, C., Dudoit, Y.,
Bonduelle, O., Bauer, M., Ihlenfeldt, H. G., Jung, G., Maier, B.,
Meyerhans, A. & Autran, B. (1996). Dynamics of viral variants in HIV-
1 Nef and specific cytotoxic T lymphocytes in vivo. Journal of Immunology
157, 4212–4221.
Haldane, J. B. S. (1955). The estimation and significance of the
logarithm of a ratio of frequencies. Annals of Human Genetics 20,
309–311.
Harrer, T., Harrer, E., Kalams, S. A., Elbeik, T., Staprans, S. I., Feinberg,
M. B., Cao, Y., Ho, D. D., Yilma, T., Caliendo, A. M., Johnson, R. P.,
Buchbinder, S. P. & Walker, B. D. (1996). Strong cytotoxic T cell and
weak neutralizing antibody responses in a subset of persons with stable
non-progressing HIV type-1 infection. AIDS Research and Human
Retroviruses 12, 585–592.
Haynes, B. F., Pantaleo, G. & Fauci, A. S. (1996). Toward an
understanding of the correlates of protective immunity to HIV infection.
Science 271, 324–328.
Imanishi, T., Akaza, T., Kimura, A., Tokunaga, K. & Gojobori, T. (1992).
Allele and haplotype frequencies for HLA and complement loci in various
ethnic groups. In HLA 1991, Proceedings of the Eleventh International
Histocompatibility Workshop and Conference held in Yokohama, pp.
1065–1220. Edited by K. Tsuji, M. Aizawa & T. Sasazuki. Oxford :
Oxford University Press.
Itescu, S., Mathur, W. U., Skovron, M. L., Brancato, L. J., Marmor, M.,
Zeleniuch, J. A. & Winchester, R. (1992). HLA-B35 is associated with
accelerated progression to AIDS. Journal of Acquired Immune Deficiency
Syndromes 5, 37–45.
CBJJ
M. R. Klein and others
Johnson, R. P., Trocha, A., Yang, L., Mazzara, G. P., Panicali, D. L.,
Buchanan, T. M. & Walker, B. D. (1991). HIV-1 gag-specific cytotoxic
T lymphocytes recognize multiple highly conserved epitopes : fine
specificity of the gag-specific response defined by using unstimulated
peripheral blood mononuclear cells and cloned effector cells. Journal of
Immunology 147, 1512–1521.
Kalams, S. A., Johnson, R. P., Trocha, A. K., Dynan, M. J., Ngo, H. S.,
D’Aquila, R. T., Kurnick, J. T. & Walker, B. D. (1994). Longitudinal
analysis of T cell receptor (TCR) gene usage by human immunodeficiency
virus 1 envelope-specific cytotoxic T lymphocyte clones reveals a limited
TCR repertoire. Journal of Experimental Medicine 179, 1261–1271.
Kaslow, R. A., Duquesnoy, R., Van Raden, M., Kingsley, L., Marrari, M.,
Friedman, H., Su, S., Saah, A. J., Detels, R., Phair, J. P. & Rinaldo, C.
(1990). A1, Cw7, B8, DR3 HLA antigen combination associated with
rapid decline of T-helper lymphocytes in HIV-1 infection. Lancet 335,
927–930.
Kaslow, R. A., Carrington, M., Apple, R., Park, L., Munoz, A., Saah,
A. J., Goedert, J. J., Winkler, C., O’Brien, S. J., Rinaldo, C., Detels, R.,
Blattner, W., Phair, J. P., Erlich, H. & Mann, D. L. (1996). Influence of
combinations of human major histocompatibility complex genes in the
course of HIV-1 infection. Nature Medicine 2, 405–411.
Keet, I. P. M., Klein, M. R., Just, J. & Kaslow, R. A. (1996). The role of
host genetics in the natural history of HIV-1 infection : the needles in the
haystack. AIDS 10 (Suppl. A), S59–S67.
Klein, M. R. & Miedema, F. (1995). Long-term survivors of HIV-1
infection. Trends in Microbiology 3, 386–391.
Klein, M. R., Keet, I. P. M., D’Amaro, J., Bende, R. J., Hekman, A.,
Mesman, B., Koot, M., de Waal, L. P., Coutinho, R. A. & Miedema, F.
(1994). Associations between HLA frequencies and pathogenic features
of human immunodeficiency virus type 1 infection in seroconverters from
the Amsterdam Cohort of homosexual men. Journal of Infectious Diseases
169, 1244–1249.
Klein, M. R., Van Baalen, C. A., Holwerda, A. M., Kerkhof-Garde, S. R.,
Bende, R. J., Keet, I. P. M., Eeftinck Schattenkerk, J. K. M., Osterhaus,
A. D. M. E., Schuitemaker, H. & Miedema, F. (1995). Kinetics of Gag-
specific CTL responses during the clinical course of HIV-1 infection : a
longitudinal analysis of rapid progressors and long-term asymptomatics.
Journal of Experimental Medicine 181, 1365–1372.
Klein, M. R., Van der Burg, S. H., Pontesilli, O. & Miedema F. (1998).
Cytotoxic T lymphocytes in HIV-1 infection, a killing paradox?
Immunology Today (in press).
Koenig, S., Conley, A. J., Brewah, Y. A., Jones, G. M., Leath, S., Boots,
L. J., Davey, V., Pantaleo, G., Demarest, J. F., Carter, C., Wannebo, C.,
Yannelli, J. R., Rosenberg, S. A. & Clifford Lane, H. (1995). Transfer of
HIV-1 specific cytotoxic T lymphocytes to an AIDS patient leads to
selection for mutant HIV variants and subsequent disease progression.
Nature Medicine 1, 330–336.
Koot, M., Keet, I. P. M., Vos, A. H. V., De Goede, R. E. Y., Roos, M. T. L.,
Coutinho, R. A., Miedema, F., Schellekens, P. T. A. & Tersmette, M.
(1993). Prognostic value of human immunodeficiency virus type 1
biological phenotype for rate of CD4+ cell depletion and progression to
AIDS. Annals of Internal Medicine 118, 681–688.
Koup, R. A., Safrit, J. T., Cao, Y., Andrews, C. A., McLeod, G.,
Borkowsky, W., Farthing, C. & Ho, D. D. (1994). Temporal associations
of cellular immune responses with the initial control of viremia in primary
human immunodeficiency virus type 1 syndrome. Journal of Virology 68,
4650–4655.
Lefkovits, I. & Waldmann, H. (1984). Limiting dilution analysis of the
cells of immune system. I. The clonal basis of the immune response.
Immunology Today 5, 265–268.
CCAA
McChesney, M., Tanneau, F., Regnault, A., Sansonetti, P., Montagnier,
L., Kieny, M. P. & Rivie' re, Y. (1990). Detection of primary cytotoxic T
lymphocytes specific for the envelope glycoprotein of HIV-1 by deletion
of the env amino-terminal signal sequence. European Journal of Immunology
20, 215–220.
Mortara, L., Letourneur, F., Gras-Masse, H., Venet, A., Guillet, J. G. &
Bourgault-Villada, I. (1998). Selection of virus variants and emergence
of virus escape mutants after immunization with an epitope vaccine.
Journal of Virology 72, 1403–1410.
Myers, G., Hahn, B. H., Henderson, L. E., Korber, B. T. M., Jeang, K. T.,
McCutchan, F. E. & Pavlakis, G. N. (1995). Human Retroviruses and
AIDS 1995 : A Compilation and Analysis of Nucleic Acid and Amino Acid
Sequences. Los Alamos, NM : Los Alamos National Laboratory.
Plata, F., Autran, B., Martins, L. P., Wain-Hobson, S., Raphael, M.,
Mayaud, C., Denis, M., Guillon, J. M. & Debre, P. (1987). AIDS virus-
specific cytotoxic T lymphocytes in lung disorders. Nature 328, 348–351.
Price, D. A., Goulder, P. J., Klenerman, P., Sewell, A. K., Easterbrook,
P. J., Troop, M., Bangham, C. R. & Phillips, R. E. (1997). Positive
selection of HIV-1 cytotoxic T lymphocyte escape variants during
primary infection. Proceedings of the National Academy of Sciences, USA 94,
1890–1895.
Rautmann, G., Kieny, M. P., Brandely, R., Dott, K., Girard, M.,
Montagnier, L. & Lecocq, J. P. (1989). HIV-1 core proteins expressed
from recombinant vaccinia viruses. AIDS Research and Human Retroviruses
5, 147–157.
Sahmoud, T., Laurian, Y., Gazengel, C., Sultan, Y., Gautreau, C. &
Costagliola, D. (1993). Progression to AIDS in French haemophiliacs :
association with HLA-B35. AIDS 7, 497–500.
Schuitemaker, H., Kootstra, N. A., De Goede, R. E. Y., De Wolf, F.,
Miedema, F. & Tersmette, M. (1991). Monocytotropic human immuno-
deficiency virus 1 (HIV-1) variants detectable in all stages of HIV
infection lack T-cell line tropism and syncytium-inducing ability in
primary T-cell culture. Journal of Virology 65, 356–363.
Sipsas, N. V., Kalams, S. A., Trocha, A., He, S., Blattner, W. A., Walker,
B. D. & Johnson, R. P. (1997). Identification of type-specific cytotoxic
T lymphocyte responses to homologous viral proteins in laboratory
workers accidentally infected with HIV-1. Journal of Clinical Investigation
99, 752–762.
Strijbosch, L. W. G., Buurman, W. A., Does, R. J. M. M., Zinken, P. H. &
Groenewegen, G. (1987). Limiting dilution assays. Experimental design
and statistical analysis. Journal of Immunological Methods 97, 133–140.
Tsubota, H., Lord, C. I., Watkins, D. I., Morimoto, C. & Letvin, N. L.
(1989). A cytotoxic T lymphocyte inhibits acquired immunodeficiency
syndrome virus replication in peripheral blood lymphocytes. Journal of
Experimental Medicine 169, 1421–1434.
Van Baalen, C. A., Klein, M. R., Geretti, A. M., Keet, I. P. M., Miedema,
F., Van Els, C. A. C. M. & Osterhaus, A. D. M. E. (1993). Selective in
vitro expansion of HLA class I-restricted HIV-1 gag specific CD8­ T
cells from seropositive individuals : identification of CTL epitopes and
precursor frequencies. AIDS 7, 781–786.
Van Baalen, C. A., Pontesilli, O., Huisman, R. C., Geretti, A. M., Klein,
M. R., De Wolf, F., Miedema, F., Gruters, R. A. & Osterhaus, A. D.
(1997). Human immunodeficiency virus type 1 Rev- and Tat-specific
cytotoxic T lymphocyte frequencies inversely correlate with rapid
progression to AIDS. Journal of General Virology 78, 1913–1918.
Van de Griend, R. J., Van Krimpen, B. A., Bol, S. J. L., Thompson, A. &
Bolhuis, R. L. H. (1984). Rapid expansion of human cytotoxic T cell
clones : growth promotion by a heat-labile serum component and by
various types of feeder cells. Journal of Immunological Methods 66,
285–298.

Van der Burg, S. H., Klein, M. R., Van de Velde, C. J. H., Kast, W. M.,
Miedema, F. & Melief, C. J. M. (1995 a). Induction of primary human
CTL response against a novel conserved epitope in a functional sequence
of HIV-1 reverse-transcriptase. AIDS 9, 121–127.
Van der Burg, S. H., Ras, E., Drijfhout, J. W., Benckhuijsen, W. E.,
Bremers, A. J. A., Melief, C. J. M. & Kast, W. M. (1995 b). An HLA class
I peptide-binding assay based on competition for binding to class I
molecules on intact human B cells : identification of conserved HIV-1
polymerase peptides binding to HLA-A*0301. Human Immunology 44,
189–198.
Van der Burg, S. H., Klein, M. R., Pontesilli, O., Holwerda, A. M.,
Drijfhout, J. W., Kast, W. M., Miedema, F. & Melief, C. J. M. (1997).
HLA-B57-restricted CTL and slow progression to AIDS
HIV-1 reverse transcriptase-specific cytotoxic T lymphocytes targeted at
epitopes under structural or functional constraints do not protect against
progression to AIDS. Journal of Immunology 159, 3648–3654.
Walker, C. M., Moody, D. J., Stites, D. P. & Levy, J. A. (1986). CD8­
lymphocytes can control HIV infection in vitro by suppressing virus
replication. Science 234, 1563–1566.
Walker, B. D., Chakrabarti, S., Moss, B., Paradis, T. J., Flynn, T., Durno,
A. G., Blumberg, R. S., Kaplan, J. C., Hirsch, M. S. & Schooley, R. T.
(1987). HIV-1 specific cytotoxic T lymphocytes in seropositive
individuals. Nature 328, 345–348.
Received 18 March 1998 ; Accepted 12 May 1998
CCAB