1

Virology 272, 347–356 (2000)doi:10.1006/viro.2000.0404, available online at http://www.idealibrary.com on

Direct Measurement of CD81 T Cell Responses in Macaques Infectedwith Simian Immunodeficiency Virus

Sean M. Donahoe,* Walter J. Moretto,* Rachel V. Samuel,† Karin J. Metzner, Preston A. Marx,‡ Tomas Hanke,†Ruth I. Connor,* and Douglas F. Nixon*,1

*Aaron Diamond AIDS Research Center, The Rockefeller University, 455 First Avenue, New York, New York 10016; †MRC Human Immunology Unit,Institute of Molecular Medicine, The John Radcliffe, Headington, Oxford OX3 9DS, United Kingdom; and ‡Department of Tropical Medicine

and Tulane Regional Primate Research Center, Tulane University Medical Center, Covington, Louisiana 70433

Received October 25, 1999; returned to author for revision April 14, 2000; accepted April 17, 2000

The simian immunodeficiency virus (SIV) macaque model system has been used extensively to study AIDS pathogenesisand to test candidate vaccines for their ability to protect against homologous or heterologous challenge with pathogenic SIVor SHIV. Recent studies suggest that stimulation of HIV-1-specific CTL responses is important for effective vaccinationagainst HIV-1. While quantitative measurements of SIV-specific cytotoxic T lymphocyte (CTL) responses have been facilitatedby the use of tetrameric peptide complexes, this technique is currently limited to the study of Mamu-A*01-positive rhesusmacaques. Furthermore, very few SIV-specific CTL epitopes have been identified, and there is limited identification of otherMHC alleles in macaques. In this study, cytokine flow cytometry (CFC) was used to quantify SIV-specific CD81 antigen-reactive T cells in macaques infected with SIV. We found a strong correlation (r 5 0.96, P , 0.001) between CD81antigen-reactive T cells stained with the Mamu-A*01 p11C, C-M tetramer and production of intracellular TNF-a in the CFCassay. Furthermore, the CFC assay was used to identify a novel SIV-specific CTL epitope in Envelope (SIV Env, a.a. 486–494,

sequence AEVAELYRL). The use of the CFC assay facilitates the study of antigen-reactive T cell responses in SIV infectionand vaccination. © 2000 Academic Press

INTRODUCTION

Recent studies have highlighted the critical role thatcellular immune responses may play in protectionagainst infection with HIV-1 (Akridge et al., 1999; Bernardet al., 1999; Cheynier et al., 1992; De Maria et al., 1994;Fowke et al., 1996; Rowland-Jones et al., 1993, 1995), andstimulation of virus-specific cellular responses have be-come a priority in HIV-1 vaccine strategies (Letvin, 1998).The simian immunodeficiency virus (SIV) challenge-pro-tection model has been extensively used to test andcompare experimental vaccines, and in some cases,protection against pathogenic SIV challenge has beendocumented in the apparent absence of neutralizing an-tibodies (Lehner et al., 1998; Nilsson et al., 1998; Robin-son et al., 1999). However, comparative quantification ofSIV-specific CTL responses elicited by these vaccineshas been difficult, principally due to the laborious natureof limiting dilution analysis.

The use of tetrameric MHC peptide complex technol-ogy has provided a new methodology to rapidly identifyand quantify antigen-reactive T cells (Altman et al., 1996;Kuroda et al., 1998; Ogg et al., 1998, 1999). A tetramericpeptide complex of the Mamu-A*01-restricted CTL

1 To whom reprint requests should be addressed. Fax: (212) 725-126. E-mail: DNixon@ADARC.org.

347

epitope in SIV Gag, p11C, C-M (Allen et al., 1998; Hankeet al., 1999; Miller et al., 1991) has been used to measurethe kinetics of SIV Gag p11C, C-M-specific CTL in pri-mary SIV infection (Kuroda et al., 1999b), in vaccinatedanimals (Seth et al., 1998) and in lymphoid tissues (Jor-dan et al., 1999; Kuroda et al., 1999a). While this tech-nique has provided a major advance in the identificationof antigen-reactive T cell responses in Mamu-A*01-pos-itive monkeys, it is currently limited because few SIV-specific CTL epitopes have been identified and otherMHC alleles in macaques have not been fully character-ized or expressed (Vogel et al., 1999). In addition, tet-ramer staining provides no measure of effector function(lysis or cytokine secretion) of the antigen-reactiveCD81 T cells.

Virus-specific CTL can mediate effector functionsthrough perforin- or Fas-mediated lysis of infected targetcells or by release of cytokines such as IFN-g and TNF-a.An alternative method to quantify virus-specific T cells isto detect intracellular cytokines produced after antigen-specific stimulation (Kern et al., 1998; Murali-Krishna etal., 1998). In this study, we have used a cytokine flowcytometry (CFC) assay to measure the production ofTNF-a in SIV-specific CD81 T cells. We compared thequantification of Mamu-A*01-restricted SIV Gag p11C,

C-M-reactive T cells using tetrameric-peptide complexstaining and CFC, and showed a strong positive corre-

0042-6822/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

(

ppwa

dnisfA

T

w

2e ning ann

348 DONAHOE ET AL.

lation between these two assays. We also demonstratedthat the CFC assay can be used to identify novel SIV-specific CTL epitopes using SIV-derived peptides, whichwill be of value in the assessment of candidate SIVvaccines.

RESULTS

Enumeration of SIV Gag p11C, C-M epitope-specificresponses by detection of intracellular TNF-a

Quantification of SIV-specific CTL epitopes throughlimiting dilution assays is technically difficult and re-quires the prior generation of autologous BCL. Wesought to simplify the procedure by measuring the pro-duction of intracellular cytokines from CD81 T cells inresponse to specific antigenic stimulation. As a positivecontrol for cytokine production, we incubated PBMC froman SIV-infected Mamu-A*01-positive monkey (Rh 1484)with the superantigen Staphylococcus enterotoxin BSEB). Direct staining of PBMC showed TNF-a produced

from 16.59% of CD81 T cells in response to SEB stimu-lation (Fig. 1A). CD81 T cells responsive to the SIV Gag

11C, C-M epitope were then enumerated after directeptide stimulation with the p11C, C-M peptide in vitro. Itas found that 1.31% of CD81 T cells upregulated CD69nd produced intracellular TNF-a in response to peptide

stimulation (Fig. 1B). Background release of TNF-a was0.19% (Fig. 1C). The reproducibility of the assay was

FIG. 1. Specific antigenic stimulation with peptide p11C, C-M inducesas measured in the CFC assay from a Mamu-A*01 1ve SIV-infecte

TNF-a-secreting cells is shown top right in each panel. Stimulation wiproduced minimal TNF-a (C). Stimulation with peptide p11C, C-M indu

0-mer overlapping peptides from SIV Gag (Table 1), including a 20-merpitope peptide was added to Pool 1 (D). Another pool, Pool 2, contaiegative control (F).

tested by comparing the production of cytokine frommultiple aliquots of PBMC samples from two macaques.

The standard deviation of responses were 0.09 and 0.3%,respectively. IFN-g was also produced on peptide spe-cific stimulation of CD81 cells and correlated with pro-

uction of TNF-a (correlation coefficient r 5 0.998; dataot shown). No TNF-a was produced in CD3-CD81 cells

n response to peptide stimulation (data not shown). Thepecificity of the CFC assay was confirmed using PBMC

rom an SIV-uninfected Mamu-A*01-positive macaqueT-80. Only background levels of TNF-a were produced

(data not shown).Because the quantity of PBMC obtained from a single

bleed of a rhesus macaque is often limited, it would beadvantageous to screen pools of peptides in the CFCassay to identify novel peptide epitopes. We made a poolof SIV Gag peptides (Pool 1, which contained the p11C,C-M epitope encoded within a 20-mer peptide, as well asfour other 20-mer Gag peptides; Table 1) and comparedcytokine production after stimulation with either the poolor individual peptides. Three conditions were then com-pared: Pool 1 alone, Pool 1 plus the addition of theminimal peptide p11C, C-M, and Pool 2 which containedfive Gag 20-mer peptides not recognized in the CFCassay (Table 1). Direct measurement of TNF-a release inthe CFC assay showed 0.74% CD81 T cells produced

NF-a after stimulation with pool 1 (Fig. 1E) compared to1.31% for the minimal peptide alone. When the p11C, C-Mpeptide was added to Pool 1, the percentage of TNF-aproduced increased to 1.01% (Fig. 1D) but was still lower

production from CD81 CD691 T cells. Intracellular TNF-a productionaque. Cells were stained with anti-CD8. The percentage of CD691uperantigen SEB induces production of TNF-a (A). Unstimulated cellsF-a production in the SIV-infected macaque (B). Pool 1 contained five

e which encompassed the minimal epitope p11C, C-M (E). The minimalother five 20-mer overlapping Gag peptides (Table 1), was used as a

TNF-ad mac

th the sced TNpeptid

than the optimal peptide (Fig. 1B), suggesting some de-gree of antigenic competition within the peptide pools.

Ta

scPiMdps

349SIV CTL EPITOPE IDENTIFICATION BY CYTOKINE STAINING

Pool 2 (Fig. 1F) had background levels of cytokine re-lease. This demonstrates that peptide pools can be usedto screen for SIV CTL epitopes in the CFC assay.

Quantification of SIV p11C, C-M response with CFCstaining is comparable to staining with a Mamu-A*01p11C, C-M tetrameric peptide complex

SIV Gag p11C, C-M-Mamu-A*01 tetrameric complexeshave been used to analyze CTL responses early after SIVinfection and localize antigen-reactive T cells to tissuessuch as lymph node and semen (Kuroda et al., 1998,1999a,b). To compare the quantity of antigen-reactive Tcells measured by tetrameric peptide staining to thatmeasured by CFC, PBMC were stained by both methodsand the results analyzed by FACS. The positive controlfor the tetramer was a p11C, C-M-derived CTL line andthe negative control was a non-Mamu-A*0–1-positiveanimal (data not shown). On direct staining with tetramer,1.60% of CD81 T cells were p11C, C-M positive (Fig. 2A).By comparison, 0.96% of PBMC produced TNF-a afterp11C, C-M stimulation as measured in the CFC assay

TABLE 1

SIV-Sequence Derived Peptides

Peptide orpool Protein Peptide Amino acid sequence

Pool 1 gag p26 714.1 PVQQIGGNYVHLPLSPRTLNgag p26 714.2 HLPLSPRTLNAWVKLIEEKKgag p26 714.3 AWVKLIEEKKFGAEVVPGFQgag p26 714.4 FGAEVVPGFQALSEGCTPYDgag p26 714.5 ALSEGCTPYDINQMLNCVGD

p11C, C-M gag p26 9-mer CTPYDINQMPool 2 gag p26 714.18 RAEQTDAAVKNWMTQTLLIQ

gag p26 714.19 NWMTQTLLIQNANPDCKLVLgag p26 714.20 NANPDCKLVLKGLGVNPTLEgag p26 714.21 KGLGVNPTLEEMLTACQGVGgag p26 714.22 EMLTACQGVGGPGQKARL

Pool 3 env gp120 774.46 WTDGNQTNITMSAEVAELYRenv gp120 774.47 MSAEVAELYRLELGDYKLVEenv gp120 774.48 LELGDYKLVEITPIGLAPTDenv gp120 774.49 ITPIGLAPTDVKRYTTGGTS

p30 env gp120 9-mer SAEVAELYRp31 env gp120 9-mer AEVAELYRLp32 env gp120 9-mer EVAELYRLEPool 4 gag p26 714.1 PVQQIGGNYVHLPLSPRTLN

gag p26 714.2 HLPLSPRTLNAWVKLIEEKKgag p26 714.3 AWVKLIEEKKFGAEVVPGFQgag p26 714.4 FGAEVVPGFQALSEGCTPYDgag p26 714.6 INQMLNCVGDHQAAMQIIRD

p5 gag p26 714.5 ALSEGCTPYDINQMLNCVGDPool 5 env gp120 774.44 PREGDLTCNSTVTSLIANID

env gp120 774.45 TVTSLIANIDWTDGNQTNITenv gp120 774.46 WTDGNQTNITMSAEVAELYRenv gp120 774.48 LELGDYKLVEITPIGLAPTDenv gp120 774.49 ITPIGLAPTDVKRYTTGGTS

p47 env gp120 774.47 MSAEVAELYRLELGDYKLVE

(Fig 2B). A comparison between tetramer and CFC stain-ing using PBMC samples obtained over a 2-year time

course from Animal Rh 1484 showed a strong positivecorrelation (r 5 0.96; P , 0.001; Fig. 2C).

he CFC assay can be used to detect specificntigen-reactive cells after in vitro culture

A period of in vitro stimulation with antigen has beenhown to increase the percentage of tetramer positiveells within a bulk PBMC culture (19). We stimulatedBMC for 14 days in vitro with fixed autologous BCL

nfected with rVV-Gag and then stained either with theamu-A*01/p11C, C-M tetramer or measured TNF-a pro-

uction in the CFC assay. Comparable numbers of CD8111C, C-M-specific T cells were detected by tetramertaining (24.4%) and TNF-a production (21.87%; Figs. 3A

and 3B). To test the function of the CTL, bulk culture cellswere also tested in a standard 51Cr-release assayagainst autologous or mismatched target cells infectedwith a recombinant vaccinia virus expressing SIV Gag orpulsed with the p11C, C-M peptide. A high degree ofspecific lysis against the p11C, C-M peptide and SIVGag-rVV was observed (Fig. 3C), indicating that the an-tigen reactive cytokine producing cells detected in theCFC assay had lytic potential after restimulation in vitro.

Identification of a novel SIV-specific CTL epitope inEnv using the CFC assay

We identified an SIV Env-specific CTL response inmacaque Rh 1488, immunized with SIVDnef mac239 (Fig.4). To identify which peptide was recognized, we testedpools of overlapping SIV Env peptides as antigenic stim-ulants in the CFC assay. One pool, Pool 3, showed apositive result using CFC. Staining was repeated withindividual peptides within Pool 3, and one peptide, p47,(sequence MSAEVAELYRLELGDYKLVE) stimulatedTNF-a production (data not shown). We confirmed rec-ognition of this peptide using a 51Cr release assay. Apanel of overlapping 9-mer peptides was then used tomap the epitope, and the 9-amino-acid peptide p31 (se-quence AEVAELYRL) was identified as the 9-mer peptiderecognized, by the CFC assay (Fig. 4D) and by sensiti-zation of target cells in a 51Cr release assay (data notshown). Peptides 30 and 32, which overlapped the p31peptide by one amino acid (see Table 1), failed to sen-sitize targets or stimulate the production of TNF-a (Figs.4C and 4E). Preliminary assessment of MHC restrictionof this response shows that this peptide is not presentedby Mamu-A*01 or B*02 (data not shown).

Different concentrations of the optimal peptide Envp31 were used to stimulate the production of cytokine(Fig 5A). For this optimal peptide, a concentration of 1 or0.1 mM stimulated over 1.3% cytokine production. Withreduction in peptide concentration, less cytokine wasproduced with background levels seen at 0.1 nM peptide.

The optimal peptide was compared with the 20-merpeptide (Env p47), and the minimal peptide (Env p31)

1Crrmetwsmit

taininga -a pro

350 DONAHOE ET AL.

stimulated more cytokine production (Fig. 5B). This wasalso seen with a comparison of the minimal peptidep11C, C-M and a 20-mer in Rh 1484 (data not shown). Alow level of cytokine was detected from Peptide Pool 5plus Env p47. No specific cytokine was detected with amixture of peptides, p5 1 p47 1 Pool 4 1 Pool 5 (Fig.5B). The sensitivity of the assay was determined bymixing peptide stimulated with unstimulated autologousPBMC at various ratios. The sensitivity was around 1 in5000, similar to that reported for the tetramer assay (Ogget al., 1998) (data not shown).

DISCUSSION

The SIV/macaque model system has been used ex-tensively in the study of AIDS pathogenesis and to testcandidate vaccines for their ability to protect againsthomologous or heterologous challenge with pathogenicSIV or SHIV. Recent studies have suggested that stimu-lation of HIV-1-specific CTL responses is important forvaccination against HIV-1, and several candidate vac-

FIG. 2. Comparison of quantitation of SIV Gag p11C, C-M-specific CDof PBMC with Mamu-A*01/p11C, C-M tetramer. The percentage of tetraC-M stimulated TNF-a production (B). Correlation between tetramer s

ssays. The correlation coefficient between tetramer staining and TNF

cines tested in primates have elicited protection againstchallenge without stimulation of neutralizing antibodies

(Lehner et al., 1998; Nilsson et al., 1998; Robinson et al.,999). While quantitative measurements of SIV-specificTL responses have been facilitated by the use of tet-

americ MHC peptide complexes, this technique is cur-ently limited to the study of Mamu-A*01-positive rhesus

acaques. Furthermore, very few SIV-specific CTLpitopes have been identified, and there is limited iden-

ification of other MHC alleles in macaques. In this study,e used cytokine flow cytometry (CFC) to quantify SIV-

pecific antigen-reactive CD81 T cells in SIV-infectedacaques. We found a strong correlation between stain-

ng with the Mamu-A*01/p11C, C-M tetramer and produc-ion of intracellular TNF-a. Furthermore, the CFC assay

was used to identify a novel SIV-specific CTL epitope inthe SIV Envelope protein. Identification of new epitopeswill increase the usefulness of the macaque model in theevaluation of CTL epitope vaccines.

Quantification of antigen-reactive T cells following SIVinfection or vaccination is essential in the investigation ofthe correlates of protection against SIV challenge and

cell responses using tetramer staining and CFC assay. Direct stainingsitive CD81 T cells is shown in the top right corner (A). Peptide p11C,and CFC assay (C). Aliquots of PBMC were split and tested in both

duction in the CFC assay (r) was 0.96, and P , 0.001.

81 Tmer-po

understanding the role of CTL responses in pathogene-sis of SIV-induced disease. Until recently, measurements

e(

s

351SIV CTL EPITOPE IDENTIFICATION BY CYTOKINE STAINING

FIG. 3. Quantitative correlation between tetramer staining and CFC assay on in vitro expanded p11C, C-M-specific CTL. Bulk cultures werestablished as described (Materials and Methods) and tested either with tetramer staining (A) or CFC assay (B). Cells were tested against autologous

1484) or mismatched (1488) BCL, infected with SIV gag rVV, or vaccinia control, TK2, without peptide (unpulsed) or pulsed with peptide p11C, C-M

at 10 mg/ml.

FIG. 4. Identification of a novel SIV Env CTL epitope using the CFC assay. Intracellular TNF-a production was measured in the CFC assay on PBMC

from Animal Rh 1488. Stimulation with the superantigen SEB is shown in (A). Unstimulated cells made low amounts of TNF-a (B). SIV Env peptide p31

timulated TNF-a production (D). SIV Env peptides p30 and p32 had background levels of TNF-a production (C and E).

SiBIimmoq

caaaMamel

pacpiaWwCaaue

hs(1Ia

g

to timallym he add

352 DONAHOE ET AL.

of the numbers of virus-specific CD81 T cells followinginfection or after vaccination have been derived fromlimiting dilution analysis (LDA), a method that is ex-tremely tedious, technically demanding, and notoriouslyvariable (Doherty, 1998). Three new methods have shownthat the LDA assay greatly underestimates the number ofantigen-reactive T cells in viral infection. The first methodidentified antigen-specific T cells through use of MHCpeptide tetrameric complexes (Altman et al., 1996) andshowed that following a virus infection, the number ofvirus-specific T cells was greater than previously thought(Kuroda et al., 1999a; Murali-Krishna et al., 1998; Ogg etal., 1998). The other two methods detect the productionof cytokine after stimulation with viral antigen, either inan ELISPOT assay (Butz and Bevan, 1998; Lalvani et al.,1997; Larsson et al., 1999; Murali-Krishna et al., 1998;

piegel et al., 1999) or by staining cytoplasmic cytokinen fixed cells after stimulating for 6 h in the presence ofrefeldin A (Kern et al., 1998; Murali-Krishna et al., 1998).

n one study of Epstein–Barr virus (EBV)-specific T cellsn chronic EBV infection, a comparison between these

ethods showed the number of virus-specific T cellseasured by ELISPOT were 5.3 times higher than that

btained from LDA, but were slightly less than the fre-uency measured by tetramer (Tan et al., 1999).

The number of studies that quantified SIV-specific Tell responses before the advent of tetramer technologyre limited due to the prolonged tissue culture and re-gents required for an SIV-specific CTL limiting dilutionssay. Tetramers have now been constructed with theamu-A*01 allele and used to study the frequency of

ntigen-reactive T cells in SIV-infected or vaccinatedacaques (Hanke et al., 1999; Jordan et al., 1999; Kuroda

FIG. 5. The magnitude of TNF-a production is dependent upon peptihe CFC assay on PBMC from Animal Rh 1488 with peptide Env p31 atf different length or mixtures was compared (B). Peptides were the opinimal epitope, a mixture of six 20-mer env peptides, and also with t

t al., 1999a,b; Seth et al., 1998). However, this method isimited in two important respects: the requirement for

rior identification of the minimal peptide epitope region,nd the expression of the MHC allele of interest. It is nowonsidered essential to stimulate virus-specific CTL asart of an effective HIV vaccine (Letvin, 1998), and there

s urgent need to make comparisons in the quantity ofntigen-reactive T cells elicited by candidate vaccines.e therefore sought to develop a novel method thatould allow the direct measurement of SIV-specificD81 T cells. Such an assay could be used to quantifyntigen-reactive T cells following infection or vaccinationnd to identify new epitope regions for use in the man-facture and testing of epitope-based vaccines (Hanket al., 1999).

The detection of intracellular cytokine production fromuman antigen-reactive T cells has been used to mea-ure cytokines released from CMV-specific CD41 T cells

Komanduri et al., 1998; Pitcher et al., 1999; Waldrop et al.,997), and CMV-specific CD81 T cells (Kern et al., 1998).n the present study, we measured SIV-specific CD81ntigen-reactive T-cell responses by detection of TNF-a

production from CD81 T cells after specific peptidestimulation. Antigen-reactive CD81 T cells produce anumber of cytokines after antigenic stimulation, includ-ing TNF-a, and we chose to measure the production ofTNF-a because it is produced earliest after antigenicstimulation (Mascher et al., 1999). We also showed thatthe production of IFN-g could be detected in this assayand that there was a strong correlation between theresults obtained with the two cytokines.

The CFC assay identifies the number of CD81 anti-en-reactive T cells that produce TNF-a after a specific

antigenic stimulus. CD81 T cells (0.96%) in AnimalRh1484 produced TNF-a after stimulation with SIV-Gag

entration and length. Intracellular TNF-a production was measured inilutions (A). The production of cytokine after stimulation with peptidesdefined minimal epitope Env p31, a 20-mer Env p47 encompassing theition of six gag peptides (12 peptides from gag and env).

de conclog10 d

peptide p11C, C-M (Fig. 2B); this animal had been in-fected with SIVmac239Dnef and protected against chal-

itssrtucbiusmmd

TCstS(fpfplupaudfr

se1phociwi

kqiepma

wMTLgt

353SIV CTL EPITOPE IDENTIFICATION BY CYTOKINE STAINING

lenge with pathogenic SIV (Connor et al., 1998). Interest-ngly, no plasma viremia was detected at the time point ofhe PBMC sample used in the CFC assay (data nothown). The maintenance of virus-specific CTL re-ponses at high frequency despite undetectable viraleplication suggests that continuous antigenic stimula-ion may not be necessary to maintain the effector pop-lation, although it is possible that low-level viral repli-ation from virus sequestered in lymphoid tissues maye stimulating these effector T cells. In some individuals

nfected with an attenuated nef-deleted virus who havendetectable plasma viremia (Dyer et al., 1999), HIV-1-pecific CTL can be detected at high frequency. Theechanisms by which an effector CTL response can beaintained at high frequency will be important to eluci-

ate.In addition to quantifying SIV-specific antigen-reactivecell responses, the identification of new SIV-specific

TL epitopes is essential in understanding immune re-ponses induced following vaccination and in response

o pathogenic SIV infection. Only a limited number ofIV-specific-CTL epitopes have been identified thus far

Allen and Watkins, 1998), and this method will be usefulor additional epitope identification, particularly usingools of peptides to simplify viral epitope identification in

uture vaccine studies. From our data, the use of optimaleptides appears preferable to longer peptides, although

onger peptides used at high concentration may still beseful for screening purposes. In addition, the number ofeptides in a peptide pool should be kept to a minimums we saw a loss of sensitivity when peptide pools weresed compared to single peptides. As vaccine candi-ates may stimulate SIV-specific CD81 T cells at low

requency, the use of rVV restimulated cultures to identifyesponses should be considered (Fig. 3B).

Recent evidence suggests that under certain circum-tances CD81 antigen-reactive T cells may fail to mountffector functions, including cytokine release (Lee et al.,999; Zajac et al., 1998). In this study, we show that theopulation of PBMC stained with tetramer is slightlyigher than that identified by CFC but indicates that mostf the tetramer-positive cells are probably able to secreteytokine. The combination of intracellular cytokine stain-

ng and tetramer analysis will assist in identifyinghether such immunosilent T cells are present in SIV-

nfected animals with low CD41 T cells.In conclusion, the main use of this intracellular cyto-

ine flow cytometry staining technique will be in theuantitative assessment of SIV-vaccine candidates and

dentification of novel dominant and nondominant viralpitopes. Stimulation of a vigorous CTL response willlay a role in the development of an HIV-1 vaccine. This

ethod will allow comparison of CTL immunogenicity

nd effector function in the primate model.

MATERIALS AND METHODS

Animals. Female rhesus macaques (Macaca mulatta)ere housed either at the Laboratory for Experimentaledicine and Surgery in Primates, Tuxedo, NY, or at the

ulane Regional Primate Research Center, Covington,A. Animals were maintained in accordance with theuidelines of the local institutional animal use commit-

ee. Animals were infected intravenously with 2 3 104

TCID50 SIVmac239Dnef and challenged with 10 animalinfectious doses (AID) of uncloned SIVmac251 (kindlyprovided by Dr. R. Desrosiers) using the same route 10weeks later (Connor et al., 1998). Control animals wereuninfected with SIV. Rhesus monkeys were screened forthe presence of the Mamu-A*01 allele by a PCR-basedtechnique (Dr. D. Watkins, Wisconsin Regional PrimateCenter) (Evans et al., 1997).

Cell lines. Autologous B-lymphoblastoid cell lines(BCL) from study animals were transformed by incubat-ing peripheral blood mononuclear cells (PBMC) at 37°Cwith RPMI 1640 (Biowhittaker, Walkersville, MD) supple-mented with R-15 (15% fetal calf serum (FCS), 10 mMHEPES, 2 mM L-glutamine and 50 IU of Pen./Strep.) andHerpesvirus papio from the supernatant of S594 cell line(kindly provided by Dr. D. Watkins, Wisconsin RegionalPrimate Center, Madison, WI). PBMC were isolated fromheparinized or EDTA blood by centrifugation over a Fi-coll-sodium diatrizoate gradient (Phamacia, Uppsala,Sweden).

Peptides. Peptides were provided by the MRC AIDSDirected Program (Potters Bar, UK) or the Endocrine Unit,Massachusetts General Hospital, MA, and are listed inTable 1. The p11C, C-M peptide was made in Oxford, UK(T.H.).

Stimulation of effector cells. Cryopreserved PBMCwere thawed, washed twice in R-15, and resuspended at2 3 106 cells/ml in R-15. For superantigenic stimulation,PBMC were incubated with staphylococcal enterotoxin B(SEB, Sigma, St. Louis, MO) at 5 mg/ml for 3 days,washed twice, and resuspended into R-15 plus 100 in-ternational units (IU) human recombinant interleukin-2(R-15–100) (Kakimoto et al., 1999). For antigen-specificstimulation, autologous BCL were infected at a m.o.i. of 2with recombinant vaccinia virus expressing SIV geneproducts Env, Gag, or control vaccinia (Therion Biologics,Cambridge, MA). Following overnight incubation, in-fected BCL were separated on a Ficoll gradient, washedtwice with R-15, counted, and incubated with 1.5% (w/v)paraformaldehyde (Sigma, St. Louis, MO) in PBS (Bio-whittaker, Walkersville, MD) at room temperature for 30min. Cells were then pelleted and incubated with 0.2 Mglycine (Fisher, Fairlawn, NJ) in PBS at room temperaturefor 15 min. After glycine incubation, cells were pelleted,washed and resuspended in filter-sterilized FCS (1 3

7

10 /ml), and stored at 4°C. PBMC were then culturedwith stimulators at a responder to stimulator ratio of 1:1

ab1(pet

mR5B

ih(tttFt

B

C

354 DONAHOE ET AL.

in R-15, with R-15–100 added on the third day. CTL as-says were performed 10–14 days after stimulation.

51Chromium-release assay. SIV-specific cytolytic activ-ity was measured using a standard 51chromium-release

ssay with cold target inhibition and autologous recom-inant vaccinia-infected BCL as targets (Johnson et al.,997). Recombinant vaccinia viruses were rVV-GagvAbT252, encoding the SIVmac251 p55gag and proteaseroteins), rVV-Env (vAbT2531 1126 Mg the SIVmac251nvelope), or control (Vaccinia Thymidine kinase nega-

ive, TK2) (Therion Biologics, Cambridge, MA).Tetramer staining. The soluble Mamu-A*01 p11 C, C-M

tetramer was made as previously described (Hanke etal., 1999). To quantify tetramer-reactive cells, cryopre-served PBMC were thawed and washed twice in R-15.Phycoerythrin (PE)-conjugated Mamu-A*01 p11C, C-Mtetramer was added for 20 min at 37°C, and the suspen-sion was then incubated with anti-CD8 TriColor (Caltag,Burlingame, CA, clone 3B5) for 30 min at 4°C. Cells werewashed twice and fixed with 2% paraformaldehyde inPBS. PBMC were analyzed for the expression of cellsurface markers and tetramer-positive cells using a Flu-orescent Activated Cell Sorter (FACS) Calibur (Becton-Dickinson, San Jose, CA) with CellQuest Software (Bec-ton-Dickinson, San Jose, CA). The mononuclear cell pop-ulation was gated on the basis of forward and sidescattering properties and subsequently evaluated for thefluorochromes conjugated to the tetramer and anti-CD8antibody. Staining of PBMC identified a distinct popula-tion of CD8/tetramer double-positive cells. The tetramerwas used to stain a known SIV Gag p11C, C-M CTL lineas positive control, and the negative control was PBMCfrom an SIV-uninfected Mamu-A*01-positive animal.

Cytokine flow cytometry. To quantify cells which pro-duce TNF-a, cryopreserved PBMC were thawed andwashed twice in R-15. Cells were resuspended in 200 mlof R-15 and incubated with either staphylococcal Entero-toxin B (SEB) (5 mg/ml) (Sigma), peptide (10 mg/ml), or

edia for 6 h in a 96-well v-bottom microtiter plate (Nunc,oskilde, Denmark). The incubation was done at 37°C in% CO2 with the last 5 h performed in the presence ofrefeldin A (10 mg/ml) (Sigma), a potent inhibitor of in-

tracellular transport that prevents secretion of any pro-duced cytokines. Antibody to CD28 (1 mg/well) (cloneL293, Becton-Dickinson, San Jose, CA) was also addedduring the incubation to facilitate costimulation. Afterstimulation, the cells were washed in PBS, incubated for10 min at 37°C in PBS10.02% EDTA, and again washedn PBS. Cells were then briefly fixed in 4% paraformalde-yde (Sigma) for 5 min at 37°C, washed in PBS11% BSA

1% BSA) (Sigma), and stored at 280°C in 1% BSA con-aining 10% DMSO (Sigma). The next day, cells werehawed and washed in 1% BSA. They were then sequen-ially incubated for 10 min each in FACSLyse and

ACSPerm (Becton-Dickinson) solutions at room tempera-

ure and washed once in 1% BSA. Following permeabiliza-

tion, cells were stained with monoclonal antibodies specificfor CD8 [conjugated to fluoroscein isothiocyanate (FITC),Becton-Dickinson], intracellular TNF-a [conjugated to phy-coerythrin (PE), clone 6401.111, No. 340512, Becton-Dickin-son], IFN-g (conjugated to PE, Biosource Int’l, Camarillo,CA, clone MD-1, part: AHC 4337), and CD69 [conjugated toperidinin chlorophyll protein (PerCp), Becton-Dickinson]. Af-ter staining, cells were washed twice and resuspended in1% paraformaldehyde in PBS. Samples were collected by aFACS Calibur flow cytometry instrument and analyzed us-ing CellQuest software. Fifty thousand total events wereacquired for each sample with sequential gating of mono-nuclear cells into the CD41 and CD81 subpopulations.These T cell subsets were further evaluated for the expres-sion of intracellular TNF-a or IFN-g and the early activationmarker, CD69.

ACKNOWLEDGMENTS

We thank Drs. K. Komanduri and J.M. McCune (U.S.C.F.) for helpfuladvice, teaching, and discussions on the cytokine flow cytometry stain-ing technique, Dr. D. Watkins (Wisconsin Regional Primate Center) forMHC typing of the animals, Dr. M. Gately (Hoffman-La-Roche) forrecombinant IL-2, Dr. G. Mazzara (Therion) for recombinant vacciniaviruses, Agegnehu Gettie and Amara Luckay for provision of primateblood samples, the MRC AIDS Directed Program for peptides, the NIHAIDS Research and Reference Reagent Program for IL-2 and SIVrecombinant proteins, and Wendy Chen and Maryanne Small for ad-ministrative assistance. This work was supported by NIH grants R01-AI-44595, AI-45510 and AI-38573, the Irene Diamond Fund and theMedical Research Council, UK (T.H.). Douglas F. Nixon is an ElizabethGlaser Scientist of the Elizabeth Glaser Pediatric AIDS Foundation.

REFERENCES

Akridge, R., Hladik, F., Markee, J., Alef, C., Kelley, H., Collier, A., andMcElrath, M. J. (1999). Cellular immunity and target cell suscep-tibility in persons with repeated HIV-1 exposure. Immunol. Lett.66, 15–19.

Allen, T. M., Sidney, J., del Guercio, M. F., Glickman, R. L., Lensmeyer,G. L., Wiebe, D. A., DeMars, R., Pauza, C. D., Johnson, R. P., Sette, A.,and Watkins, D. I. (1998). Characterization of the peptide bindingmotif of a rhesus MHC class I molecule (Mamu-A*01) that binds animmunodominant CTL epitope from simian immunodeficiency virus.J. Immunol. 160, 6062–6071.

Allen, T., and Watkins, D. (1998). SIV and SHIV CTL epitopes identifiedin macaques. http://hiv-web.lanl.gov/immunology/, IV8-IV13.

Altman, J. D., Moss, P. A. H., Golder, P. J. R., Barouch, D. H., McHeyzer-Williams, M. G., Bell, J. I., McMichael, A. J., and Davis, M. M. (1996).Phenotypic analysis of antigen-specific T lymphocytes. Science 274,94–96.

Bernard, N., Yannakis, C., Lee, J., and Tsoukas, C. (1999). Humanimmunodeficiency virus (HIV)-specific cytotoxic T lymphocyte ac-tivity in HIV-exposed seronegative persons. J. Infect. Dis. 179,538–547.

utz, E. A., and Bevan, M. J. (1998). Massive expansion of antigen-specific CD81 T cells during an acute virus infection. Immunity 8,167–175.

heynier, R., Langlade, D. P., Marescot, M. R., Blanche, S., Blondin, G.,Wain, H. S., Griscelli, C., Vilmer, E., and Plata, F. (1992). Cytotoxic Tlymphocyte responses in the preipheral blood of children born to

human immunodeficiency virus-1-infected mothers. Eur. J. Immunol.22, 2211–2217.

D

E

K

L

L

L

L

L

M

M

M

N

O

O

P

R

R

355SIV CTL EPITOPE IDENTIFICATION BY CYTOKINE STAINING

Connor, R., Montefiori, D., Binley, J., Moore, J., Bonhoeffer, S., Gettie, A.,Fenamore, E., Sheridan, K., Ho, D., Dailey, P., and Marx, P. (1998).Temporal analyses of virus replication, immune responses, and ef-ficacy in rhesus macaques immunized with a live, attenuated simianimmunodeficiency virus vaccine. J. Virol. 72, 7501–7509.

e Maria, A., Cirillo, C., and Moretta, L. (1994). Occurrence of humanimmunodeficiency virus type 1 (HIV-1)-specific cytolytic T cell activityin apparently uninfected children born to HIV-1-infected mothers.J. Infect. Dis. 170, 1296–1299.

Doherty, P. C. (1998). The numbers game for virus-specific CD81 Tcells. Science 280, 227.

Dyer, W. B., Ogg, G. S., Demoitie, M. A., Jin, X., Geczy, A. F., Rowland-Jones, S. L., McMichael, A. J., Nixon, D. F., and Sullivan, J. S. (1999).Strong human immunodeficiency virus (HIV)-specific cytotoxic T-lymphocyte activity in Sydney Blood Bank Cohort patients infectedwith nef-defective HIV type 1. J. Virol. 73, 436–443.

vans, D. T., Piekarczyk, M. S., Allen, T. M., Boyson, J. E., Yeager, M.,Hughes, A. L., Gotch, F. M., Hinshaw, V. S., and Watkins, D. I.(1997). Immunodominance of a single CTL epitope in a primatespecies with limited MHC class I polymorphism. J. Immunol. 159,1374–1382.

Fowke, K. R., Nagelkerke, N. J., Kimani, J., Simonsen, J. N., Anzala,A. O., Bwayo, J. J., MacDonald, K. S., Ngugi, E. N., and Plummer,F. A. (1996). Resistance to HIV-1 infection among persistentlyseronegative prostitutes in Nairobi, Kenya [see comments]. Lan-cet 348, 1347–1351.

Hanke, T., Neumann, V. C., Blanchard, T. J., Sweeney, P., Hill, A. V., Smith,G. L., and McMichael, A. (1999). Effective induction of HIV-specificCTL by multi-epitope using gene gun in a combined vaccinationregime. Vaccine 17, 589–596.

Johnson, R., Glickman, R., Yang, J., Kaur, A., Dion, J., Mulligan, M., andDesrosiers, R. (1997). Induction of vigorous cytotoxic T-lymphocyteresponses by live attenuated simian immunodeficiency virus. J.Virol.71, 7711–7718.

Jordan, H. L., Kuroda, M. J., Schmitz, J. E., Steenbeke, T., Forman,M. A., and Letvin, N. L. (1999). Detection of simian immunodefi-ciency virus Gag-specific CD8(1) T lymphocytes in semen ofchronically infected rhesus monkeys by cell staining with a tet-rameric major histocompatibility complex class I-peptide com-plex. J. Virol. 73, 4508–4512.

Kakimoto, W. M., Gettie, A., Smith, S., Donahoe, S. M., Jin, X., Marx, P.,Connor, R., and Nixon, D. F. (1999). Comparison of restimulationmethods to elicit SIV specific cytotoxic T-lymphocytes (CTL) in vitro:Staphylococcal enterotoxin B (SEB) provides a novel method for thequantification of SIV specific CTL precursors. Immunol. Lett. 66,135–140.

Kern, F., Surel, I. P., Brock, C., Freistedt, B., Radtke, H., Scheffold, A.,Blasczyk, R., Reinke, P., Schneider-Mergener, J., Radbruch, A., Wal-den, P., and Volk, H. D. (1998). T-cell epitope mapping by flowcytometry. Nat. Med. 4, 975–978.

Komanduri, K. V., Viswanathan, M. N., Wieder, E. D., Schmidt, D. K.,Bredt, B. M., Jacobson, M. A., and McCune, J. M. (1998). Restorationof cytomegalovirus-specific CD41 T-lymphocyte responses afterganciclovir and highly active antiretroviral therapy in individualsinfected with HIV-1. Nat. Med. 4, 953–956.

Kuroda, M., Schmitz, J., Barouch, D., Craiu, A., Allen, T., Sette, A.,Watkins, D., Forman, M., and Letvin, N. (1998). Analysis of Gag-specific cytotoxic T lymphocytes in simian immunodeficiency virus-infected rhesus monkeys by cell staining with a tetrameric majorhistocompatibility complex class I-peptide complex. J. Exp. Med. 187,1373–1381.

Kuroda, M., Schmitz, J., Charini, W., Nickerson, C., Lord, C., Forman, M.,and Letvin, N. (1999a). Comparative analysis of cytotoxic T lympho-cytes in lymph nodes and peripheral blood of simian immunodefi-ciency virus-infected rhesus monkeys. J. Virol. 73, 1573–1579.

uroda, M. J., Schmitz, J. E., Charini, W. A., Nickerson, C. E., Lifton, M. A.,Lord, C. I., Forman, M. A., and Letvin, N. L. (1999b). Emergence of CTL R

coincides with clearance of virus during primary simian immunode-ficiency virus infection in rhesus monkeys. J. Immunol. 162, 5127–5133.

alvani, A., Brookes, R., Hambleton, S., Britton, W. J., Hill, A. V., andMcMichael, A. J. (1997). Rapid effector function in CD81 memory Tcells. J. Exp. Med. 186, 859–865.

arsson, M., Jin, X., Ramratnam, B., Ogg, G. S., Engelmayer, J., Demoitie,M. A., McMichael, A. J., Cox, W. I., Steinman, R. M., Nixon, D., andBhardwaj, N. (1999). A recombinant vaccinia virus based ELISPOTassay detects high frequencies of Pol-specific CD8 T cells in HIV-1-positive individuals. AIDS 13, 767–777.

ee, P. P., Yee, C., Savage, P. A., Fong, L., Brockstedt, D., Weber, J. S.,Johnson, D., Swetter, S., Thompson, J., Greenberg, P. D., Roederer, M.,and Davis, M. M. (1999). Characterization of circulating T cells spe-cific for tumor-associated antigens in melanoma patients. Nat. Med.5, 677–685.

ehner, T., Wang, Y., Cranage, M., Bergmeier, L. A., Mitchell, E., Tao, L.,Hall, G., Dennis, M., Cook, N., Jones, I., and Doyle, C. (1998). Protec-tive mucosal immunity elicited by targeted lymph node immunizationwith a subunit SIV envelope and core vaccine in macaques. Dev. Biol.Stand. 92, 225–235.

etvin, N. L. (1998). Progress in the development of an HIV-1 vaccine.Science 280, 1875–1880.ascher, B., Schlenke, P., and Seyfarth, M. (1999). Expression andkinetics of cytokines determined by intracellular staining using flowcytometry. J. Immunol. Methods 223, 115–121.iller, M. D., Yamamoto, H., Hughes, A. L., Watkins, D. I., and Letvin,N. L. (1991). Definition of an epitope and MHC class I moleculerecognized by gag-specific cytotoxic T lymphocytes in SIVmac-in-fected rhesus monkeys. J. Immunol. 147, 320–329.urali-Krishna, K., Altman, J., Suresh, M., Sourdive, D., Zajac, A., Miller,J., Slansky, J., and Ahmed, R. (1998). Counting antigen-specific CD8T-cells: A reevaluation of bystander activation during viral infection.Immunity 8, 177–187.

ilsson, C., Makitalo, B., Thorstensson, R., Norley, S., Binninger-Schin-zel, D., Cranage, M., Rud, E., Biberfeld, G., and Putkonen, P. (1998).Live attenuated simian immunodeficiency virus (SIV)mac in ma-caques can induce protection against mucosal infection with SIVsm.AIDS 12, 2261–2270.

gg, G. S., Jin, X., Bonhoeffer, S., Dunbar, P. R., Nowak, M. A., Monard,S., Segal, J. P., Cao, Y., Rowland-Jones, S. L., Cerundolo, V., Hurley, A.,Markowitz, M., Ho, D. D., Nixon, D. F., and McMichael, A. J. (1998).Quantitation of HIV-1-specific cytotoxic T lymphocytes and plasmaload of viral RNA. Science 279, 2103–2106.

gg, G. S., Jin, X., Bonhoeffer, S., Moss, P., Nowak, M. A., Monard, S.,Segal, J. P., Cao, Y., Rowland-Jones, S. L., Hurley, A., Markowitz, M.,Ho, D. D., McMichael, A. J., and Nixon, D. F. (1999). Decay kineticsof human immunodeficiency virus-specific effector cytotoxic Tlymphocytes after combination antiretroviral therapy. J. Virol. 73,797–800.

itcher, C. J., Quittner, C., Peterson, D. M., Connors, M., Koup, R. A.,Maino, V. C., and Picker, L. J. (1999). HIV-1-specific CD41 T cells aredetectable in most individuals with active HIV-1 infection, but declinewith prolonged viral suppression. Nat. Med. 5, 518–525.

obinson, H. L., Montefiori, D. C., Johnson, R. P., Manson, K. H., Kalish,M. L., Lifson, J. D., Rizvi, T. A., Lu, S., Hu, S. L., Mazzara, G. P., Panicali,D. L., Herndon, J. G., Glickman, R., Candido, M. A., Lydy, S. L., Wyand,M. S., and McClure, H. M. (1999). Neutralizing antibody-independentcontainment of immunodeficiency virus challenges by DNA primingand recombinant pox virus booster immunizations. Nat. Med. 5,526–534.

owland-Jones, S., Sutton, J., Ariyoshi, K., Dong, T., Gotch, F., McAdam,S., Whitby, D., Sabally, S., Gallimore, A., Corrah, T., and et al. (1995).HIV-specific cytotoxic T-cells in HIV-exposed but uninfected Gam-bian women [published erratum appears in Nat. Med. 1, 598, 1995].

Nat Med 1, 59–64.

owland-Jones, S. L., Nixon, D. F., Aldhous, M. C., Gotch, F., Ariyoshi, K.,

S

V

W

Z

356 DONAHOE ET AL.

Hallam, N., Kroll, J. S., Froebel, K., and McMichael, A. (1993). HIV-specific cytotoxic T-cell activity in an HIV-exposed but uninfectedinfant. Lancet 341, 860–861.

Seth, A., Ourmanov, I., Kuroda, M., Schmitz, J., Carroll, M., Wyatt, L.,Moss, B., Forman, M., Hirsch, V., and Letvin, N. (1998). Recombinantmodified vaccinia virus Ankara-simian immunodeficiency virus gagpol elicits cytotoxic T lymphocytes in rhesus monkeys detected by amajor histocompatibility complex class I/peptide tetramer. Proc. Natl.Acad. Sci. USA 95, 10112–10116.

piegel, H. M., DeFalcon, E., Ogg, G. S., Larsson, M., Beadle, T. J., Tao,P., McMichael, A. J., Bhardwaj, N., O’Callaghan, C., Cox, W. I., Kra-sinski, K., Pollack, H., Borkowsky, W., and Nixon, D. F. (1999).Changes in frequency of HIV-1-specific cytotoxic T cell precursorsand circulating effectors after combination antiretroviral therapy in

children. J. Infect. Dis. 180, 359–368.

Rowland-Jones, S., McMichael, A. J., Rickinson, A. B., and Callan,M. F. (1999). A re-evaluation of the frequency of CD81 T cells specificfor EBV in healthy virus carriers. J. Immunol. 162, 1827–1835.

ogel, T. U., Evans, D. T., Urvater, J. A., O’Connor, D. H., Hughes, A. L.,and Watkins, D. I. (1999). Major histocompatibility complex class Igenes in primates: Co-evolution with pathogens. Immunol. Rev. 167,327–337.

aldrop, S., Pitcher, C., Peterson, D., Maino, V., and Picker, L. (1997).Determination of antigen-specific memory/effector CD41 T cell fre-quencies by flow cytometry: Evidence for a novel, antigen-specifichomeostatic mechanism in HIV-associated immunodeficiency.J. Clin. Invest. 99, 1739–1750.

ajac, A. J., Blattman, J. N., Murali-Krishna, K., Sourdive, D. J., Suresh,M., Altman, J. D., and Ahmed, R. (1998). Viral immune evasion due to

persistence of activated T cells without effector function. J. Exp. Med.

Tan, L. C., Gudgeon, N., Annels, N. E., Hansasuta, P., O’Callaghan, C. A., 188, 2205–2213.