Virology 378 (2008) 48–57

Contents lists available at ScienceDirect

Virology

j ourna l homepage: www.e lsev ie r.com/ locate /yv i ro

Distinctive in vitro effects of T-cell growth cytokines on cytomegalovirus-stimulatedT-cell responses of HIV-infected HAART recipients

Julie Patterson a, Renee Jesser a, Adriana Weinberg a,b,c,⁎a Department of Pediatrics, University of Colorado School of Medicine, Denver, CO, USAb Department of Medicine, University of Colorado School of Medicine, Denver, CO, USAc Department of Pathology, University of Colorado School of Medicine, Denver, CO, USA

⁎ Corresponding author. 4200 E. 9th Avenue, CampuUSA. Fax: +1 303 315 1787.

E-mail address: Adriana.Weinberg@uchsc.edu (A. W

0042-6822/$ – see front matter © 2008 Elsevier Inc. Aldoi:10.1016/j.virol.2008.05.018

A B S T R A C T

A R T I C L E I N F O

Article history:

Functional immune reconst Received 8 February 2008Returned to author for revision8 March 2008Accepted 16 May 2008Available online 24 June 2008

Keywords:CytomegalovirusCell-mediated immunityHIV infectionHAARTT-cell growth cytokines

itution is limited after HAART, maintaining the interest in adjunctive immune-modulators. We compared in vitro the effects of the γ-chain T-cell growth cytokines IL-2, IL-4, IL-7 and IL-15on cytomegalovirus-stimulated cell-mediated immunity. IL-2 and IL-15 increased cytomegalovirus-specificlymphocyte proliferation in HAART recipients, whereas IL-4 and IL-7 did not. The boosting effect of IL-2 andIL-15 on proliferation correlated with their ability to prevent late apoptosis. However, IL-2 increased thefrequency of cells in early apoptosis, whereas IL-15 increased the frequency of fully viable cells. Both IL-2 andIL-15 increased cytomegalovirus-induced CD4+ and CD8+ T-cell proliferation and the synthesis of Th1 andpro-inflammatory cytokines and chemokines. However, only IL-2 increased the frequency of regulatory Tcells and Th2 cytokine production, both of which have the potential to attenuate antiviral immune responses.Overall, compared to other γ-chain cytokines, IL-15 had the most favorable profile for boosting antiviral cell-mediated immunity.

© 2008 Elsevier Inc. All rights reserved.

Introduction

Highly active antiretroviral therapy (HAART) decreases theincidence of opportunistic infections and increases CD4+ T cells inHIV-infected patients. However, reconstitution of adaptive immuneresponses is often incomplete as demonstrated by persistently poorresponses to vaccines (Lange et al., 2003; Weinberg et al., 2006a) andthe occasional occurrence of opportunistic infections, such as cyto-megalovirus (CMV), herpes zoster, and disseminated tuberculosis inindividuals with N200 CD4+ cells/µl (Jacobson et al., 1997; Komanduriet al., 2001; Lawn et al., 2005).

We previously reported that HAART recipients had multipleabnormalities of the CMV-specific cell-mediated immunity (CMI)(Weinberg et al., 2006b, 2001, 2003), including decreased prolifera-tion of CMV-stimulated peripheral blood mononuclear cells (PBMC),which inversely correlated with apoptosis of both resting and CMV-stimulated PBMC (Weinberg et al., 2004). Both proliferation andapoptosis improved after in vitro treatment with the viral serpincrmA, which blocks caspase 8 and 9 activities, but not after cell deathreceptor blockade, suggesting that the mitochondrial apoptotic path-way was involved in the decreased CMV CMI. HIV infection has beenshown to directly or indirectly increase apoptosis via the intrinsicmitochondrial pathway (Genini et al., 2001; Holm and Gabuzda,

s Box C227, Denver, CO 80262,

einberg).

l rights reserved.

2005), but other factors, such as insufficient induction of bcl-2 oranalogous anti-apoptotic factors by T-cell growth cytokines may alsocontribute to this effect (Pahwa et al., 2006; Petrovas et al., 2004;Zaunders et al., 2003). Several cytokines, including the Th1 cytokinesIL-2 and IL-15, the Th0 cytokine IL-7, and the Th2 cytokine IL-4,increase bcl-2 production and activate other cell survival mechanismsby signaling through the common γ-chain receptor (Adachi et al.,1996; Leonard, 2001). In addition, HIV infection has been associatedwith abnormal IL-2, IL-7 and IL-15 production and/or function,suggesting that exogenous addition of these cytokines may improveimmune functions (d'Ettorre et al., 2002; De Paoli, 2001; Ostrowskiet al., 2003; Sasson et al., 2006).

Due to the incomplete immune reconstitution resulting fromHAART, cytokines, chemokines and other immune mediators havebeen administered to enhance immune responses of HIV-infectedsubjects (De Paoli, 2001; Katlama et al., 2002; Keh et al., 2006;Kuekrek et al., 2005; Lindemann et al., 2004; Martinez-Marino et al.,2004; Sereti et al., 2004, 2002). In vivo administration of IL-2, inaddition to HAART, consistently increased CD4+ T-cell numbers, butgenerally failed to augment pre-existing or to replace previously lostantigen-specific CMI (De Paoli, 2001; Lindemann et al., 2004). IL-2administration rarely increased responses to recall or neo antigens(Kuekrek et al., 2005; Levy et al., 2006; Nacsa et al., 2005; Valdez et al.,2003). Although the presence of IL-2 during primary antigenicexposure may be required for the development of immunologicmemory (Weinberg and Merigan, 1988; Williams et al., 2006), IL-2administration is also associated with an increase in the number of

49J. Patterson et al. / Virology 378 (2008) 48–57

circulating regulatory T cells (Treg) and activation-induced cell death(AICD) in the presence of persistent antigen presentation (Andersonet al., 2005; Malek and Bayer, 2004; Sereti et al., 2002), which maycontribute to the limited effect of IL-2 on reconstitution of adaptiveimmunity of HIV-infected individuals (Nacsa et al., 2005).

IL-15 is a Th1 cytokine that shares many properties with IL-2,including β- and γ-chain cellular receptors (Cornish et al., 2006; Duboiset al., 2003). However, IL-15 also has specific effects including a markedincrease of memory CD8+ T-cell survival (Sato et al., 2007). IL-15 hasnot been extensively studied in HIV-infected subjects, but IL-15administration to SIV-infected macaques enhanced their immunereconstitution (Picker et al., 2006).

IL-7 induces lymphopoiesis and peripheral T-cell homeostasis andprolongs survival of memory and effector CD4+ and CD8+ cells (Fry andMackall, 2005; Picker et al., 2006; Rathmell et al., 2001). In addition,IL-7 may increase cytotoxicity (Lum et al., 2004) and antigen-specificIFNγ production (Jennes et al., 2002).

IL-4 also stimulates T-cell growth and proliferation and, like IL-2and IL-15, can prevent AICD (Kennedy and Celis, 2006). IL-4, however,is more likely to dampen than boost antiviral CMI (Wherry, 2004).

The goal of this study was to compare the effect of γ-chaincytokines on apoptosis and CMV-specific CMI in HIV-infectedindividuals and to determine the function of the T cells expandedwith the use of exogenous cytokines.

Results

Differential effect of T-cell growth cytokines on PBMC apoptosis andCMV-specific proliferation

To test the hypothesis thatγ-chain receptor signaling is sufficient todecrease apoptosis and increase proliferation of PBMC from HIV-infected subjects, we examined the ex-vivo effect of IL-2, IL-4, IL-7 andIL-15 γ-chain cytokines on CMV-stimulated PBMC from HIV-infectedanduninfected subjects. Although these cytokines have commonγ-chainreceptor activity, they differ with respect to their overall function: IL-2and IL-15 have Th1 properties, IL-7 has Th0 properties and IL-4 has Th2properties. The experiments using PBMC from healthy volunteersshowed that all 4 cytokines significantly increased proliferationstimulated with mock-infected control antigen (average increase of65,240, 4802, 9251 and 105,800 cpm for IL-2, IL-4, IL-7 and IL-15,respectively; not depicted), but also CMV-specific proliferation, assessedby the difference between 3H-thymidine incorporation in CMV andcontrol antigen-stimulated cultures (Fig. 1). IL-2, IL-7 and IL-15, but notIL-4, increased proliferation of control antigen-stimulated PBMC fromHIV-infected subjects (averages of 16,300, 1836, 29,780 and 97cpm,respectively; not depicted). Only the Th1 cytokines IL-2 and IL-15significantly increased CMV-specific proliferation of PBMC from HIV-infected subjects on HAART (Fig. 1). Although there was no apparentdifference in thenet effects of IL-4 and IL-7 onCMV-specific proliferationof PBMC from HIV-infected subjects, they resulted from differentmechanisms: IL-4 failed to increase proliferation of either CMV orcontrol antigen-stimulated PBMC, whereas IL-7 equally increasedproliferation of CMV- and control antigen-stimulated PBMC.

In cytokine-untreated cultures, CMV-stimulated proliferation ofPBMC from HIV-infected subjects was significantly lower comparedwith proliferation of PBMC from healthy individuals, with a meandifference±SEM of 30,260±10,350cpm, p=0.008. To determine ifexogenous addition of cytokines brought the CMV-specific prolifera-tive responses of PBMC from HAART recipients closer to the levelsobserved in healthy subjects, we compared the CMV-specific cpm incytokine-treated cultures of HIV-infected subjects with the CMV-specific cpm in cytokine-untreated cultures from healthy subjects. IL-2 and IL-15 reduced the gap between HIV-infected and uninfectedsubjects to 3808±13,170cpm (p=0.77) and 11,000±17,350cpm(p=0.53), respectively. In contrast, exogenous IL-4 or IL-7 did not

appreciably alter the CMV-specific proliferation difference betweenHIV-infected and uninfected subjects (p of 0.01 and 0.05, respectively).

IL-15 ex-vivo treatment significantly decreased apoptosis of CMV-stimulated PBMC from HIV-infected (mean difference±SEM of 27±4%,pb0.0001 paired T test) and healthy subjects (10±3%, p=0.006 pairedT test). IL-2 marginally decreased apoptosis of CMV-stimulated PBMCfrom HIV-infected individuals (14±5%, p=0.06 paired T test), but notof PBMC from healthy subjects (p=0.13 paired T test). IL-4 and IL-7tended to decrease apoptosis of PBMC from HIV-infected anduninfected subjects, but their effects did not reach statisticalsignificance (p of 0.10 to 0.23, paired T test).

In cytokine-untreated cultures, apoptosis of CMV-stimulatedPBMC from HIV-infected subjects was significantly higher comparedwith uninfected controls with mean difference±SEM Annexin V+

PBMC of 23±3% (pb0.0001, unpaired T test). IL-15 ex-vivo treatmentreduced the gap in apoptosis between CMV-stimulated PBMC fromHIV-infected vs. uninfected individuals to mean difference±SEM of6±5% (p=0.28, unpaired T test). IL-2 treatment had a less potent effecton the excess apoptosis of HAART recipients compared withuninfected controls (mean difference of 8±4%, p=0.06, unpaired Ttest), whereas supplementation with exogenous IL-4 or IL-7 did notappreciably change the difference in apoptosis between CMV-stimulated PBMC cultures from HAART recipients vs. healthy subjects,which remained highly significant (p of 0.03 and 0.007, respectively).These data indicated that IL-15 had the most potent anti-apoptoticeffect among the 4 cytokines evaluated in this study.

To investigate the association between the proliferative and anti-apoptotic effects of ex-vivo cytokine treatment, we performedcorrelation analyses between the cytokine-induced increase in log10CMV-specific cpm with the decrease of Annexin V+PI+% PBMC.Significant positive correlations were found for IL-2 and IL-15(p=0.01 each) but not for IL-4 or IL-7. These results suggested thatthe lack of survival and proliferation of PBMC from HAART recipientsin response to CMV stimulation might be due to a Th1 cytokinespecific deficit that could be corrected by the exogenous administra-tion of IL-2 and evenmore so of IL-15. IL-4 and IL-7 did not appreciablyimprove apoptosis or proliferation of CMV-stimulated PBMC fromHAART recipients, therefore all subsequent experiments addresseddifferences and similarities between the effects of IL-2 and IL-15 only.

Effect of IL-2 and IL-15 on CMV-stimulated CD4+ and CD8+

cell proliferation and apoptosis

It was previously reported that the effect of IL-15 was moreprominent on CD8+ compared with CD4+ T cells. To determine if IL-15preferentially elicited CD8+ cell proliferation in the presence of CMVantigenic stimulation and to understand how this compared with theeffect of IL-2 on CMV-specific T-cell proliferation, we used CFSE dyedilution, which enabled the differentiation between CD4+ and CD8+

cell proliferation.We used the CMV lysate as a stimulant after showingthat this particulate antigen induces both CD4+ and CD8+ cellproliferation (Fig. 2).

The analysis of CMV or control antigen-stimulated PBMC culturesfrom 8 HIV-infected subjects showed that there were no appreciabledifferences in the effect of IL-2 or IL-15 on CD4+ vs. CD8+ cellexpansions (pN0.5 for either cytokine). Both cytokines increased celldivisions in CMV or control antigen-stimulated cultures by approxi-mately an order of magnitude. Compared with IL-2, IL-15 treatmentgenerated higher CMV-specific proliferation of CD4+ (mean±SEM of11±3% vs. 4±3%, p=0.05) and CD8+ cells (8±2% vs. 1±2%, p=0.05),suggesting that IL-15 may have a more discriminative antigen-enhancing effect compared with IL-2 (Fig. 3).

Neither IL-2 nor IL-15 exogenous administration appreciablychanged viability of control antigen-stimulated PBMC from 6 donorsequally distributed between HIV-infected and uninfected individuals(mean±SEM viability of 44±7%, 46±5% and 51±5% in untreated, IL-2-

50 J. Patterson et al. / Virology 378 (2008) 48–57

treated and IL-15-treated cultures, respectively; 0.12, ANOVA forrepeated measures; not depicted). In contrast, in CMV-stimulatedcultures, IL-15 exogenous administration significantly increased thepercentage of viable T cells to 48±5% compared with 40±6% incytokine-untreated cultures (p=0.005) and 43±5% in IL-2-treated

cultures (p=0.002; Fig. 4). In contrast, viability of IL-2-treated andcytokine-untreated T cells was not appreciably different. IL-2 treat-ment resulted in a significantly higher percentage of cells in earlyapoptosis (15±1%) compared with cytokine-untreated and IL-15-treated lymphocytes (13±1% each, p=0.05). IL-2 and IL-15 treatment

Fig. 2. CD4+ and CD8+ cell expansions in cultures stimulatedwith inactivated CMV antigen. The plots represent typical examples of CFSE-measured proliferation in HIV-infected (right2 columns) and uninfected (left 2 columns) subjects. PBMC are labeledwith CFSE prior to in vitro culture and uptake of CD4+ and CD8+ cells is verified after individually gating on eachcell population (upper row). CFSE content is measured again after 6 days of in vitro stimulation with CMV (bottom row) and mock-infected control antigen (middle row). The plotsshow very few dividing cells (CFSElo) in mock-infected control-stimulated cultures, but a significant proportion of CD4+ and CD8+ cell divisions in CMV-stimulated PBMC.

51J. Patterson et al. / Virology 378 (2008) 48–57

equally decreased the percentage of T cells in late apoptosis from 38±6% in untreated cultures to 23±3% and 23±4% in IL-2- and IL-15-treated cultures, respectively (p=0.005 and 0.006, respectively).

Effect of IL-2 and IL-15 on the frequency of regulatory T cells (Tregs)

IL-2 was previously shown to increase CD4+CD25+FoxP3+ Tregs.Wehypothesized that a diverse effect of IL-2 vs. IL-15 on Tregs maycontribute to the differential effect of these cytokines on apoptosis andantigen-specific proliferation of CMV-stimulated PBMC. The effect ofIL-2 and IL-15 on the frequency of Tregs was investigated using PBMCfrom 8 HIV-infected subjects on HAART (median CD4+ numbers=340

Fig. 1. Effect of T-cell growth cytokines on CMV-specific proliferation and apoptosis of PBMsubjects (median CD4=352 cells/µl; median plasma HIV RNAb400 copies/ml) and 18 healbetween PWM cpm/control antigen cpm ≥5. Bars represent means and SEM of untreated (ware depicted in the first, second, third and fourth row, respectively. The first column showsantigen- from CMV-stimulated PBMC. The second column shows apoptosis, measured byassociation between the cytokine-induced increase in proliferation (delta log10CMV cpm, calPBMC cultures) vs. decrease in late apoptosis (delta % apoptosis, calculated by the differencultures). The data in the third columnwere derived from 9 to 12 experiments using PBMC fcytokine-treated with untreated cultures in the same group of subjects (continuous horizo(interrupted lines); and linear regression for association analysis. All data passed normality

cells/µl and plasma HIV RNAb400 copies/ml; Fig. 5). We limited theseexperiments to HIV-infected donors because these constitute thetarget population for a potential therapeutic application. IL-2-treatedcultures of CMV-stimulated PBMC had higher frequencies of CD4+

CD25+FoxP3+ Tregs compared with untreated (mean difference=5%,p=0.03) or with IL-15-treated cultures (mean difference=4.6%,p=0.04). IL-2 also increased the frequency of CD8+CD25+FoxP3+ cellscompared with untreated (mean difference=8%, p=0.03) or with IL-15-treated cultures (mean difference=7.4%, p=0.05). Addition ofexogenous IL-15 did not appreciably increase the frequency of CD4+

CD25+FoxP3+ or CD8+CD25+FoxP3+ Tregs compared with untreatedcultures (p=0.4 for either comparison).

C from HIV-infected and uninfected subjects. Data were derived from 36 HIV-infectedthy individuals. All PBMC cultures had a positive response to PWM defined by a ratiohite) and cytokine-treated (black) PBMC cultures. The effects of IL-2, IL-4, IL-7 and IL-15proliferation (CMV cpm) calculated by subtracting the 3H-Thy incorporation of controlthe frequency of Annexin V+PI+ CMV-stimulated PBMC. The third column shows theculated by the difference between the log10CMV cpm in cytokine-treated and untreatedce between % Annexin V+PI+ CMV-stimulated PBMC in cytokine-untreated vs. treatedrom HIV-infected subjects. p values were derived from paired T tests for comparisons ofntal lines), unpaired T tests for comparisons of HIV-infected with uninfected subjectsdistribution test.

Fig. 3. Effect of IL-2 and IL-15 on CMV or control antigen-stimulated expansion of CD4+

and CD8+ PBMC from HIV-infected subjects on HAART. The bars represent means andSEM of results derived from 8 HAART recipients (median CD4=322 cells/µl; medianplasma HIV RNAb400 copies/ml) of CMV-specific CFSElo T cell %. p values, calculated bypaired T test, are noted above the horizontal lines that connect the bars representing thecell populations involved in the comparison.

52 J. Patterson et al. / Virology 378 (2008) 48–57

Effect of IL-2 and IL-15 on cytokine and chemokine production ofCMV-stimulated PBMC from HIV-infected subjects

Cytokine production was used to assess additional functional char-acteristics of CMV-stimulated PBMC. Th1 and Th2 cytokines and chemo-kines were measured in supernatants of CMV or control antigen-stimulated PBMC from 22HIV-infected subjects onHAARTwithmedianCD4+ cells=315 cells/µl and median plasma HIV RNAb400 copies/ml

Fig. 4. Comparative effect of IL-2 and IL-15 on apoptosis of CMV-stimulated PBMC from HIV-ilate apoptotic cells were measured by the % events in lower left, lower right and upper righlower and right upper quadrant. Data in panel B represent paired experiments of CMV-stimulRNAb400 copies/ml) treated ex-vivowith IL-2 (light grey), IL-15 (dark grey), and no exogenoulate apoptotic PBMC were defined as Annexin V−PI−, Annexin V+PI−, and Annexin V+PI+ cells

(Fig. 6). The data, expressed as differences between CMV or controlantigen-stimulated cultures treated with IL-2, IL-15 or no cytokine,showed that both IL-2 and IL-15 significantly increased the Th1cytokines IFNγ, TNFα and the chemokine RANTES (pb0.05, paired Ttests). For MIP-1α, only the boosting effect of IL-15 reached statisticalsignificance (p=0.04). In contrast, IL-2, but not IL-15, significantlyincreased the production of Th2 cytokines including IL-4, IL-5, IL-10 andIL-13 (pb0.01, paired T tests). IL-2 had its largest effect on IL-5 pro-duction, which increased frommean±SEM of 5±0.7 pg/ml in untreatedcultures to 114±34 in IL-2-treated PBMC cultures, followed by IL-13, IL-10 and IL-4. The amount of IL-4 measured in culture supernatants wassmall and so was the difference between untreated and IL-2 treatedPBMC. However, the consistency of the effect of IL-2 across all Th2cytokines together with the fact that there was no IL-4 detected in theculture medium or in the supernatant of the PBMC stimulated withcontrol antigen suggests that these differences were real. The biologicsignificance of such small differences is difficult to ascertain.

The effect of IL-2 and IL-15 on PBMC from 4 healthy individuals(data not depicted) was consistent with the findings of HIV-infectedsubjects. Both cytokines significantly increased the secretion of TNFα,RANTES and IL-10, but only IL-2 significantly increased IL-5 and IL-13production.

Discussion

In vitro measures of CMV-specific CMI are significantly lower inHIV-infected subjects on HAART compared with healthy individuals.We sought to determine if any of 4 γ-chain cytokines IL-2, IL-4, IL-7

nfected subjects on HAART. Panel A shows a typical scatter plot. Live, early apoptotic andt quadrants, respectively. Total apoptosis was measured by the sum of % events in rightated PBMC from 8 HIV-infected subjects (median CD4=322 cells/µl; median plasma HIVs cytokines (white). Bars and lines representmeans and SEM. Viable, early apoptotic and, respectively. p values were derived by paired T test.

Fig. 5. Comparative effect of IL-2 and IL-15 on CD4+ and CD8+ Treg frequencies in PBMC fromHIV-infected subjects on HAART. Panel A shows a typical scatter plot. Panel B presents thedata from 8 subjects (median CD4=340 cells/µl; median plasma HIV RNAb400copies/ml). Lines connect results of the same donor.

53J. Patterson et al. / Virology 378 (2008) 48–57

Fig. 6. Comparative effect of IL-2 and IL-15 on CMV-stimulated cytokine and chemokine production of PBMC from HIV-infected subjects on HAART. Bars represent means and SEM ofcytokine or chemokine concentrations in CMV-stimulated PBMC culture supernatants after subtraction of concentrations in control antigen-stimulated PBMC from 22 HAARTrecipients (median CD4=315 cells/µl; median plasma HIV RNAb400 copies/ml). Asterisks indicate significant differences between cytokine-treated and untreated cultures (pb0.05by paired T test).

54 J. Patterson et al. / Virology 378 (2008) 48–57

and IL-15 could boost CMV-specific CMI responses of HAARTrecipients. We found that IL-2 and IL-15 increased CMV-specificproliferation measured by 3H-thymidine incorporation and CFSEdilution. Furthermore, IL-2 and IL-15 in vitro supplementation

substantially decreased the gap in proliferation between CMV-stimulated PBMC from HIV-infected subjects on HAART and healthyindividuals. In contrast, IL-4 or IL-7 did not appreciably increaseCMV-specific proliferation of PBMC from HIV-infected subjects.

55J. Patterson et al. / Virology 378 (2008) 48–57

Interestingly, all 4 γ-chain cytokines increased CMV-specific pro-liferation of PBMC from healthy donors, underscoring some of thedifferences in immune functions between HIV-infected individualson HAART and healthy subjects.

IL-4 failed to enhance proliferation of CMV- or control antigen-stimulated PBMC from HIV-infected subjects, but enhanced CMV-stimulated proliferation of PBMC from healthy donors. This suggeststhat the IL-4 receptors or their downstream pathways may besaturated in cultured T cells from HIV-infected subjects. Others havenoted a Th2 bias in the immune response of HIV-infected subjectsbefore the advent of HAART (Wasik et al., 1997) and it is conceivablethat this may not be reverted with therapy.

IL-7 had a similar proliferative effect on antigen-stimulated vs.unstimulated PBMC from HIV-infected subjects. This finding may notbe unexpected considering that IL-7 is important for the survival ofboth naive and memory T cells (Kondrack et al., 2003; Li et al., 2003;Sun et al., 2006; Vassena et al., 2007). However, IL-7 favored CMV-specific proliferation of PBMC from healthy individuals, suggestingthat HIV infection alters the ability of PBMC to respond to IL-7. Severallines of investigation also support this conclusion: 1) HIV-infectedpatients have increased levels of circulating IL-7 associated withreduced IL-7 receptor-α expression (Sasson et al., 2006); 2) IL-7increases Fas-mediated apoptosis in the context of HIV infection(Fluur et al., 2007); and 3) IL-7 is not necessary for the survival ofmemory CD8+ cells (Klonowski et al., 2006), which represent animportant contingent among the PBMC that proliferate in response toCMV stimulation in HIV-infected subjects. These and our data indicatethat therapeutic use of IL-7 is unlikely to increase antigen-specificimmune responses of HAART recipients.

The increase of CMV-specific proliferation of PBMC from HIV-infected subjects generated by IL-2 or IL-15 in vitro supplementationsignificantly correlated with the ability of each cytokine to protect thePBMC against late apoptosis. However, IL-15 had a more robust anti-apoptotic effect than IL-2 in PBMC cultures from HIV-infected anduninfected subjects. In addition, IL-15 arrested PBMC progression toapoptosis in earlier stages compared with IL-2, increasing theproportion of fully viable cells, whereas IL-2 prevented progressionfrom early to late apoptosis, suggesting a mechanistic differencebetween the 2 cytokines. Since apoptosis is increased by in vitroantigenic stimulation (Weinberg et al., 2003), the ability of IL-15 toprevent apoptosis may account for its greater effect on CMV-inducedproliferation by comparison with IL-2. This observation is alsoconsistent with the notion that IL-15 has a unique role in prolongingT-cell survival and maintaining immunologic memory (Alpdoganet al., 2005; Berard et al., 2003; Picker et al., 2006; Schluns et al.,2002). The robust anti-apoptotic effect of IL-15 may allow HIV-infected patients with chronic antigenic stimulation to maintain highlevels of antigen-specific T cells.

IL-2, but not IL-15 significantly increased the proportion of CD4+

Tregs in CMV-stimulated cultures from HIV-infected or uninfectedsubjects. This is in accordance with previous reports of increased inCD4+ Tregs in patients treated with IL-2 (Sereti et al., 2002). We alsoshowed that IL-2 increased CD8+CD25+FoxP3+, which may also haveTreg properties (Endharti et al., 2005). The ability of IL-15 to stimulateCMI without inducing excess Treg expansion may be an advantage forits use as an immune-stimulator in HIV-infected individuals.

IL-15, compared with IL-2, increased the production of a slightlybroader array of Th1 cytokines and chemokines by CMV-stimulatedPBMC from HIV-infected subjects, suggesting that IL-15 may have amore robust boosting effect on antiviral CMI. In contrast, IL-2, but notIL-15, significantly increased Th2 cytokine production by CMV-stimulated PBMC from HIV-infected subjects on HAART. Since Th2cytokines have been previously shown to divert the immune responsefrom Th1 to Th2 and IL-10 is also known to mediate Treg activity, theTh2-enhancing effect of IL-2 may attenuate antiviral defenses. Thiseffect may also contribute to the failure of IL-2 administration to

HAART recipients to boost their antigen-specific CMI when theantigen predominantly stimulates Th1 responses.

Taken together, our data indicate that IL-15 therapeutic adminis-tration is more likely than IL-2 or IL-7 to facilitate recovery of adaptiveimmune responses of HIV-infected individuals. Clinical studies arewarranted to further assess its therapeutic value.

Subjects and methods

Subjects

This study enrolled 62 CMV-seropositive individuals, including 36HIV-infected subjects on HAART and 26 HIV-uninfected controls,according to the norms of the Colorado Multiple Institutional ReviewBoard. Other inclusion criteria for HIV-infected subjects were HAART,defined as ≥3 antiretrovirals including ≥2 classes, for ≥3months; CD4+

b100 cells/µl before HAART and ≥100 cells/µl at enrollment. At blooddraws, the median duration of HAART was 18 months; the medianCD4+ cell count was 340 cells/µl with a maximum of 1309 cells/µl; andthe median and upper quartile plasma HIV RNA was b102.6 copies/mlwith a maximum of 105.3.

Reagents

CMV or control antigens were prepared by glycine extraction andUV inactivation of CMV AD169- and mock-infected human lungfibroblasts, respectively. Although inactivated, the CMV-infected celllysate stimulates both CD4+ and CD8+ T cells similar to the responsesto other inactivated virions (Burgdorf et al., 2007; Larsson et al., 2002;Rutebemberwa et al., 2007). Recombinant human IL-2, IL-4, IL-7, andIL-15 were purchased from R&D Systems. Staining reagents were asfollows: carboxyfluorescein diacetate, succinimidyl ester (CFSE,Molecular Probes), PerCP-conjugated anti-CD4 (anti-CD4-PerCP) andanti-CD8-PerCP mouse mAbs (Becton Dickinson); and anti-CD25-APCand anti-FoxP3-FITC (eBiosciences).

Specimen processing and cryopreservation

PBMC from heparinized blood, separated by Ficoll/Hypaquegradient (Sigma) centrifugation, were cryopreserved as previouslydescribed (Weinberg et al., 2000) at 107 cells/ml in 10% dimethylsulfoxide (Sigma)-containing fetal bovine serum (Gemini Bio-Products).

Lymphocyte proliferation

This was performed using cryopreserved PBMC as previouslydescribed (Weinberg et al., 2001). Stimulation medium consistedof RPMI 1640 with glutamine (Gibco), 10% human AB serum (Nabi)and 1% antibiotics (Gibco). 106 PBMC/ml were stimulated withCMV antigen at a 1:200 final dilution in quadruplicate wells orculture tubes. Mock-infected antigen and pokeweed mitogen(Sigma) were used as negative and positive controls at finaldilutions of 1:200 and 10 µg/ml, respectively. Recombinant humanIL-2, IL-4, IL-7, and IL-15 in concentrations pre-determined toprovidemaximum stimulation (3.3 ng/ml, 2.5 ng/ml, 3.3 ng/ml and2.5 ng/ml, respectively) were added to the stimulation medium. Tomeasure proliferation, 6-day cultures maintained at 37 °C and 5%CO2, were pulsed with 3H-thymidine (Perkin Elmer) for 6 h,harvested and counted using Unifilter plates (Perkin Elmer) and amicroplate scintillation counter (Packard). CMV-specific prolif-eration was calculated by subtracting the median counts perminute (cpm) measured in the mock-infected control wells fromthe median cpm of the CMV-stimulated wells. For flow cytometricassays, cells were harvested, washed and stained as describedbelow.

56 J. Patterson et al. / Virology 378 (2008) 48–57

Apoptosis assay

TACS Annexin V-FITC apoptosis detection kit (R&D Systems) wasused for measurement of PBMC apoptosis. Samples containing 105 to106 PBMC were centrifuged at 500 g for 10 min, washed with coldphosphate-buffered saline (PBS), and resuspended in 100 µl of AnnexinV Incubation Reagent containing Annexin V-FITC and propidiumiodide (PI). After 15 min of incubation in the dark, 400 µl of bindingbuffer was added. Samples were analyzed using a FACSCalibur flowcytometer and CellQuest software (Becton Dickinson).

T-cell phenotypic characterization

Stimulated PBMC were washed in PBS and incubated with theappropriate surface marker mAb for 30 min at room temperature. Forintracellular staining, PBMC were then permeabilized and fixed usingCytofix/Cytoperm (BD Pharmingen) for 20 min at 4 °C, washed andstained with the appropriate reagents. After washing with PBS, PBMCwere analyzed in real timewith a FACSCalibur instrument and CellQuestsoftware (BD Biosciences). The number of events collectedwas ≥10,000.

CFSE dilution assays

PBMC were labeled with 0.5 µM CFSE (Molecular Probes) in 0.1%bovine serum albumin PBS for 10 min at 37 °C. Staining was quenchedwith cold 10% human AB serum in RPMI. An aliquot of PBMC wasremoved for baseline flow analysis and the remaining cells werestimulated with CMV or control antigen for 6 days, after which PBMCwere washed in PBS, stained with CD4 and CD8 surface markers andanalyzed by flow cytometry as described above.

Cytokine and chemokine detection

Inducible cytokines and chemokines were measured in the super-natants of 6-day CMV or mock-infected control antigen-stimulatedPBMC using Human Cytokine Antibody Bead Kits (Biosource) accord-ing to the manufacturer's instructions. CMV-specific production wasexpressed as the difference in concentration between CMV and mock-infected control antigen-stimulated cultures.

Statistical analysis

Analysis was performed using Instat3 and Prism4 (GraphPad)software. Two-tailed parametric T tests or ANOVA were used for thestatistical analyses after verifying that the data were normallydistributed. Significance was defined by p≤0.05.

Acknowledgments

This work was supported by grant 1R21AI073121 and facilitated bythe infrastructure and resources provided by the Colorado Center forAIDS Research Grant P30 AI054907. The authors thank Anna Vazquezfor assisting in manuscript preparation.

References

Adachi, Y., Oyaizu, N., Than, S., McCloskey, T.W., Pahwa, S., 1996. IL-2 rescues in vitrolymphocyte apoptosis in patients with HIV infection: correlation with its ability toblock culture-induced down-modulation of Bcl-2. J. Immunol. 157 (9), 4184–4193.

Alpdogan, O., Eng, J.M., Muriglan, S.J., Willis, L.M., Hubbard, V.M., Tjoe, K.H., Terwey, T.H.,Kochman, A., van den Brink, M.R., 2005. Interleukin-15 enhances immunereconstitution after allogeneic bone marrow transplantation. Blood 105 (2),865–873.

Anderson, P.O., Sundstedt, A., Yazici, Z., Minaee, S., O'Neill, E.J., Woolf, R., Nicolson, K.,Whitley, N., Li, L., Li, S., Wraith, D.C., Wang, P., 2005. IL-2 overcomes theunresponsiveness but fails to reverse the regulatory function of antigen-inducedT regulatory cells. J. Immunol. 174 (1), 310–319.

Berard, M., Brandt, K., Bulfone-Paus, S., Tough, D.F., 2003. IL-15 promotes the survival ofnaive and memory phenotype CD8+ T cells. J. Immunol. 170 (10), 5018–5026.

Burgdorf, S., Kautz, A., Bohnert, V., Knolle, P.A., Kurts, C., 2007. Distinct pathways ofantigen uptake and intracellular routing in CD4 and CD8 T cell activation. Science316 (5824), 612–616.

Cornish, G.H., Sinclair, L.V., Cantrell, D.A., 2006. Differential regulation of T-cell growthby IL-2 and IL-15. Blood 108 (2), 600–608.

d'Ettorre, G., Forcina, G., Lichtner, M., Mengoni, F., D'Agostino, C., Massetti, A.P.,Mastroianni, C.M., Vullo, V., 2002. Interleukin-15 in HIV infection: immunologicaland virological interactions in antiretroviral-naive and -treated patients. Aids 16(2), 181–188.

De Paoli, P., 2001. Immunological effects of interleukin-2 therapy in human immuno-deficiency virus-positive subjects. Clin. Diagn. Lab. Immunol. 8 (4), 671–677.

Dubois, S., Shou, W., Haneline, L.S., Fleischer, S., Waldmann, T.A., Muller, J.R., 2003.Distinct pathways involving the FK506-binding proteins 12 and 12.6 underlie IL-2-versus IL-15-mediated proliferation of T cells. Proc. Natl. Acad. Sci. U. S. A. 100 (24),14169–14174.

Endharti, A.T., Rifa, I.M.s., Shi, Z., Fukuoka, Y., Nakahara, Y., Kawamoto, Y., Takeda, K.,Isobe, K., Suzuki, H., 2005. Cutting edge: CD8+CD122+ regulatory T cells produce IL-10 to suppress IFN-gamma production and proliferation of CD8+ T cells. J. Immunol.175 (11), 7093–7097.

Fluur, C., De Milito, A., Fry, T.J., Vivar, N., Eidsmo, L., Atlas, A., Federici, C., Matarrese, P.,Logozzi, M., Rajnavolgyi, E., Mackall, C.L., Fais, S., Chiodi, F., Rethi, B., 2007. Potentialrole for IL-7 in Fas-mediated T cell apoptosis during HIV infection. J. Immunol. 178(8), 5340–5350.

Fry, T.J., Mackall, C.L., 2005. The many faces of IL-7: from lymphopoiesis to peripheral Tcell maintenance. J. Immunol. 174 (11), 6571–6576.

Genini, D., Sheeter, D., Rought, S., Zaunders, J.J., Susin, S.A., Kroemer, G., Richman, D.D.,Carson, D.A., Corbeil, J., Leoni, L.M., 2001. HIV induces lymphocyte apoptosis by ap53-initiated, mitochondrial-mediated mechanism. FASEB J. 15 (1), 5–6.

Holm, G.H., Gabuzda, D., 2005. Distinct mechanisms of CD4+ and CD8+ T-cell activationand bystander apoptosis induced by human immunodeficiency virus type 1 virions.J. Virol. 79 (10), 6299–6311.

Jacobson, M.A., Zegans, M., Pavan, P.R., O'Donnell, J.J., Sattler, F., Rao, N., Owens, S.,Pollard, R., 1997. Cytomegalovirus retinitis after initiation of highly activeantiretroviral therapy. Lancet 349 (9063), 1443–1445.

Jennes, W., Kestens, L., Nixon, D.F., Shacklett, B.L., 2002. Enhanced ELISPOT detection ofantigen-specific T cell responses from cryopreserved specimens with addition ofboth IL-7 and IL-15—the Amplispot assay. J. Immunol. Methods 270 (1), 99–108.

Katlama, C., Carcelain, G., Duvivier, C., Chouquet, C., Tubiana, R., De Sa, M., Zagury, L.,Calvez, V., Autran, B., Costagliola, D., 2002. Interleukin-2 accelerates CD4 cellreconstitution in HIV-infected patients with severe immunosuppression despitehighly active antiretroviral therapy: the ILSTIM study—ANRS 082. Aids 16 (15),2027–2034.

Keh, C.E., Shen, J.M., Hahn, B., Hallahan, C.W., Rehm, C.A., Thaker, V., Wynne, S.M., Davey,R.T., Lane, H.C., Sereti, I., 2006. Interruption of antiretroviral therapy blunts but doesnot abrogate CD4 T-cell responses to interleukin-2 administration in HIV infectedpatients. Aids 20 (3), 361–369.

Kennedy, R., Celis, E., 2006. T helper lymphocytes rescue CTL from activation-inducedcell death. J. Immunol. 177 (5), 2862–2872.

Klonowski, K.D., Williams, K.J., Marzo, A.L., Lefrancois, L., 2006. Cutting edge: IL-7-independent regulation of IL-7 receptor alpha expression and memory CD8 T celldevelopment. J. Immunol. 177 (7), 4247–4251.

Komanduri, K.V., Feinberg, J., Hutchins, R.K., Frame, R.D., Schmidt, D.K., Viswanathan,M.N.,Lalezari, J.P.,McCune, J.M., 2001. Lossof cytomegalovirus-specific CD4+Tcell responsesin human immunodeficiency virus type 1-infected patients with high CD4+ T cellcounts and recurrent retinitis. J. Infect Dis. 183 (8), 1285–1289.

Kondrack, R.M., Harbertson, J., Tan, J.T., McBreen, M.E., Surh, C.D., Bradley, L.M., 2003.Interleukin 7 regulates the survival and generation of memory CD4 cells. J. Exp.Med. 198 (12), 1797–1806.

Kuekrek, H., Schlingmann, T., Valdez, H., Boehm, B.O., Pollard, R.B., Mitsuyasu, R.,Goebel, F.D., Lederman, M.M., Lehmann, P.V., Tary-Lehmann, M., 2005. Differentialeffect of interleukin-2 treatment on primary and secondary immunizations in HIVinfected individuals. Aids 19 (17), 1967–1974.

Lange, C.G., Lederman, M.M., Medvik, K., Asaad, R., Wild, M., Kalayjian, R., Valdez, H.,2003. Nadir CD4+ T-cell count and numbers of CD28+ CD4+ T-cells predictfunctional responses to immunizations in chronic HIV-1 infection. Aids 17 (14),2015–2023.

Larsson, M., Fonteneau, J.F., Lirvall, M., Haslett, P., Lifson, J.D., Bhardwaj, N., 2002.Activation of HIV-1 specific CD4 and CD8 T cells by human dendritic cells: roles forcross-presentation and non-infectious HIV-1 virus. Aids 16 (10), 1319–1329.

Lawn, S.D., Bekker, L., Wood, R., 2005. How effectively does HAART restore immuneresponses to mycobacterium tuberculosis? Implications for tuberculosis control.AIDS 19, 1113–1124.

Leonard,W.J., 2001. Cytokines and immunodeficiency diseases. Nat. Rev. Immunol. 1 (3),200–208.

Levy, Y., Durier, C., Lascaux, A.S., Meiffredy, V., Gahery-Segard, H., Goujard, C., Rouzioux,C., Resch, M., Guillet, J.G., Kazatchkine, M., Delfraissy, J.F., Aboulker, J.P., 2006.Sustained control of viremia following therapeutic immunization in chronicallyHIV-1-infected individuals. Aids 20 (3), 405–413.

Li, J., Huston, G., Swain, S.L., 2003. IL-7 promotes the transition of CD4 effectors topersistent memory cells. J. Exp. Med. 198 (12), 1807–1815.

Lindemann, M., Witzke, O., Winterhagen, T., Ross, B., Kreuzfelder, E., Reinhardt, W.,Roggendorf, M., Mann, K., Philipp, T., Grosse-Wilde, H., 2004. T-cell function afterinterleukin-2 therapy in HIV-infected patients is correlated with serum cortisolconcentrations. Aids 18 (15), 2001–2007.

Lum, J.J., Schnepple, D.J., Nie, Z., Sanchez-Dardon, J., Mbisa, G.L., Mihowich, J., Hawley, N.,Narayan, S., Kim, J.E., Lynch, D.H., Badley, A.D., 2004. Differential effects of

57J. Patterson et al. / Virology 378 (2008) 48–57

interleukin-7 and interleukin-15 on NK cell anti-human immunodeficiency virusactivity. J. Virol. 78 (11), 6033–6042.

Malek, T.R., Bayer, A.L., 2004. Tolerance, not immunity, crucially depends on IL-2. Nature4, 665–674.

Martinez-Marino, B., Shiboski, S., Hecht, F.M., Kahn, J.O., Levy, J.A., 2004. Interleukin-2therapy restores CD8 cell non-cytotoxic anti-HIV responses in primary infectionsubjects receiving HAART. Aids 18 (15), 1991–1999.

Nacsa, J., Edghill-Smith, Y., Tsai,W.P., Venzon,D., Tryniszewska, E.,Hryniewicz, A.,Moniuszko,M., Kinter, A., Smith, K.A., Franchini, G., 2005. Contrasting effects of low-dose IL-2 onvaccine-boosted simian immunodeficiency virus (SIV)-specific CD4+ andCD8+ Tcells inmacaques chronically infected with SIVmac251. J. Immunol. 174 (4), 1913–1921.

Ostrowski, S.R., G.J., Pedersen, B.K., Ullum, H., 2003. Impaired production of cytokines isan independent predictor of mortality in HIV-1-infected patients. AIDS 17, 521–530.

Pahwa, R., McCloskey, T.W., Aroniadis, O.C., Strbo, N., Krishnan, S., Pahwa, S., 2006. CD8+ Tcells in HIV disease exhibit cytokine receptor perturbation and poor T cell receptoractivation but are responsive to gamma-chain cytokine-driven proliferation. J. Infect.Dis. 193 (6), 879–887.

Petrovas, C., Mueller, Y.M., Dimitriou, I.D., Bojczuk, P.M., Mounzer, K.C., Witek, J., Altman,J.D., Katsikis, P.D., 2004. HIV-specific CD8+ Tcells exhibit markedly reduced levels ofBcl-2 and Bcl-xL. J. Immunol. 172 (7), 4444–4453.

Picker, L.J., Reed-Inderbitzin, E.F., Hagen, S.I., Edgar, J.B., Hansen, S.G., Legasse, A., Planer,S., Piatak Jr., M., Lifson, J.D., Maino, V.C., Axthelm, M.K., Villinger, F., 2006. IL-15induces CD4 effector memory Tcell production and tissue emigration in nonhumanprimates. J. Clin. Invest 116 (6), 1514–1524.

Rathmell, J.C., Farkash, E.A., Gao, W., Thompson, C.B., 2001. IL-7 enhances the survivaland maintains the size of naive T cells. J. Immunol. 167 (12), 6869–6876.

Rutebemberwa, A., Bess, J., J.W., Brown, B., Arroyo, M., Eller, M., Slike, B., Polonis, V.,McCutchan, F., Currier, J.R., Birx, D., Robb,M.,Marovich,M., Lifson, J.D., Cox, J.H., 2007.Evaluation of aldrithiol-2-inactivated preparations of HIV type 1 subtypes A, B, andD as reagents to monitor T cell responses. AIDS Res. Hum. Retrovir. 23 (4), 532–542.

Sasson, S.C., Zaunders, J.J., Zanetti, G., King, E.M., Merlin, K.M., Smith, D.E., Stanley, K.K.,Cooper, D.A., Kelleher, A.D., 2006. Increased plasma interleukin-7 level correlateswith decreased CD127 and Increased CD132 extracellular expression on T cellsubsets in patients with HIV-1 infection. J. Infect Dis. 193 (4), 505–514.

Sato, N., Patel, H.J., Waldmann, T.A., Tagaya, Y., 2007. The IL-15/IL-15Ralpha on cellsurfaces enables sustained IL-15 activity and contributes to the long survival of CD8memory T cells. Proc. Natl. Acad. Sci. U. S. A. 104 (2), 588–593.

Schluns, K.S., Williams, K., Ma, A., Zheng, X.X., Lefrancois, L., 2002. Cutting edge:requirement for IL-15 in the generation of primary and memory antigen-specificCD8 T cells. J. Immunol. 168 (10), 4827–4831.

Sereti, I., Martinez-Wilson, H., Metcalf, J.A., Baseler, M.W., Hallahan, C.W., Hahn, B.,Hengel, R.L., Davey, R.T., Kovacs, J.A., Lane, H.C., 2002. Long-term effects ofintermittent interleukin 2 therapy in patients with HIV infection: characterizationof a novel subset of CD4(+)/CD25(+) T cells. Blood 100 (6), 2159–2167.

Sereti, I., Anthony, K.B., Martinez-Wilson, H., Lempicki, R., Adelsberger, J., Metcalf, J.A.,Hallahan, C.W., Follmann, D., Davey, R.T., Kovacs, J.A., Lane, H.C., 2004. IL-2-inducedCD4+ T-cell expansion in HIV-infected patients is associated with long-termdecreases in T-cell proliferation. Blood 104 (3), 775–780.

Sun, J.C., Lehar, S.M., Bevan, M.J., 2006. Augmented IL-7 signaling during viral infectiondrives greater expansion of effector T cells but does not enhance memory. J. Immunol.177 (7), 4458–4463.

Valdez, H., Mitsuyasu, R., Landay, A., Sevin, A.D., Chan, E.S., Spritzler, J., Kalams, S.A.,Pollard, R.B., Fahey, J., Fox, L., Namkung, A., Estep, S., Moss, R., Sahner, D., Lederman,M.M., 2003. Interleukin-2 Increases CD4+ lymphocyte numbers but does notenhance responses to immunization: results of A5046s. J. Infect Dis. 187 (2),320–325.

Vassena, L., Proschan, M., Fauci, A.S., Lusso, P., 2007. Interleukin 7 reduces the levels ofspontaneous apoptosis in CD4+ and CD8+ T cells from HIV-1-infected individuals.Proc. Natl. Acad. Sci. U. S. A. 104 (7), 2355–2360.

Wasik, T.J., Jagodzinski, P.P., Hyjek, E.M., Wustner, J., Trinchieri, G., Lischner, H.W.,Kozbor, D., 1997. Diminished HIV-specific CTL activity is associated with lower type1 and enhanced type 2 responses to HIV-specific peptides during perinatal HIVinfection. J. Immunol. 158 (12), 6029–6036.

Weinberg, A., Edelstein, C.L., Jesser, R.D., et al., 2003. Ninth International Cytomega-lovirus Workshop and First International Betaherpesvirus Workshop. Maastricht,Netherlands.

Weinberg, A., Merigan, T.C., 1988. Recombinant interleukin 2 as an adjuvant for vaccine-induced protection. Immunization of guinea pigs with herpes simplex virus subunitvaccines. J. Immunol. 140 (1), 294–299.

Weinberg, A., Wohl, D.A., Brown, D.G., Pott, G.B., Zhang, L., Ray, M.G., van der Horst, C.,2000. Effect of cryopreservation on measurement of cytomegalovirus-specificcellular immune responses in HIV-infected patients. J. Acquir. Immune Defic. Syndr.25 (2), 109–114.

Weinberg, A.,Wohl, D.A., Barrett, R.J., van der Horst, C., 2001. Inconsistent reconstitutionof cytomegalovirus-specific cell-mediated immunity in human immunodeficiencyvirus-infected patients receiving highly active antiretroviral therapy. J. Infect. Dis.184 (6), 707–712.

Weinberg, A., Wohl, D.A., MaWhinney, S., Barrett, R.J., Brown, D.G., Glomb, N., Van DerHorst, C., 2003. Cytomegalovirus-specific IFN-gamma production is associated withprotection against cytomegalovirus reactivation in HIV-infected patients on highlyactive antiretroviral therapy. Aids 17 (17), 2445–2450.

Weinberg, A., Jesser, R.D., Edelstein, C.L., Bill, J.R., Wohl, D.A., 2004. Excess apoptosis ofmononuclear cells contributes to the depressed cytomegalovirus-specific immu-nity in HIV-infected patients on HAART. Virology 330 (1), 313–321.

Weinberg, A., Gona, P., Nachman, S.A., Defechereux, P., Yogev, R., Hughes, W., Wara, D.,Spector, S.A., Read, J., Elgie, C., Cooper, M., Dankner, W., 2006a. Antibody responsesto hepatitis A virus vaccine in HIV-infected children with evidence of immunologicreconstitution while receiving highly active antiretroviral therapy. J. Infect. Dis. 193(2), 302–311.

Weinberg, A., Wiznia, A.A., Lafleur, B.J., Shah, S., Levin, M.J., 2006b. Cytomegalovirus-specific cell-mediated immunity in HIV-infected children on HAART. AIDS Res.Hum. Retrovir. 22 (3), 283–288.

Wherry, E.J., A.R., 2004. Memory CD8 T cell differentiation during viral infection. J. Virol.78 (11), 5535–5545.

Williams, M.A., Tyznik, A.J., Bevan, M.J., 2006. Interleukin-2 signals during priming arerequired for secondary expansion of CD8+ memory T cells. Nature 441 (7095),890–893.

Zaunders, J.J., Moutouh-de Parseval, L., Kitada, S., Reed, J.C., Rought, S., Genini, D., Leoni,L., Kelleher, A., Cooper, D.A., Smith, D.E., Grey, P., Estaquier, J., Little, S., Richman, D.D.,Corbeil, J., 2003. Polyclonal proliferation and apoptosis of CCR5+ T lymphocytesduring primary human immunodeficiency virus type 1 infection: regulation byinterleukin (IL)-2, IL-15, and Bcl-2. J. Infect. Dis. 187 (11), 1735–1747.