iavd

d

Virology 286, 225–236 (2001)doi:10.1006/viro.2001.0984, available online at http://www.idealibrary.com on

Nef Is Required for Efficient HIV-1 Replication in Coculturesof Dendritic Cells and Lymphocytes

Caroline Petit,* Florence Buseyne,† Claire Boccaccio,‡ Jean-Pierre Abastado,‡Jean-Michel Heard,* and Olivier Schwartz*,1

*Unite Retrovirus et Transfert Genetique, †Laboratoire d’Immunopathologie Virale, URA CNRS 1930, Institut Pasteur,28 rue du Dr Roux, 75724 Paris Cedex 15, France; and ‡ImmunoDesigned Molecules Research Laboratories,

Institut des Cordeliers, 15 rue de l’Ecole de Medecine, 75006 Paris, France

Received March 6, 2001; returned to author for revision April 16, 2001; accepted April 30, 2001

Dendritic cells (DCs) are thought to play a crucial role in the pathogenesis of HIV-1 infection. DCs are believed to transportvirus particles to lymph nodes before transfer to CD41 lymphocytes. We have investigated the role of Nef in these processes.HIV-1 replication was examined in human immature DC-lymphocyte cocultures and in DCs or lymphocytes separately. Usingvarious R5-tropic and X4-tropic HIV-1 strains and their nef-deleted (Dnef) counterparts, we show that Nef is required foroptimal viral replication in immature DC-T cells clusters and in T lymphocytes. Nef exerts only a marginal role on viralreplication in immature DCs alone as well as on virion capture by DCs, long-term intracellular accumulation and transmissionof X4 strains to lymphocytes. We also show that wild-type and Dnef virions are similarly processed for MHC-I restrictedexogenous presentation by DCs. Taken together, these results help explain how HIV-1 Nef may affect viral spread and

immune responses in the infected host. © 2001 Academic Press

Key Words: HIV-1; Nef; dendritic cells; lymphocytes; virus transmission; MHC-I exogenous presentation.

epa1ftScprDrl

cCtscnpt

INTRODUCTION

The nef gene is required for efficient in vivo replicationand pathogenicity of human and simian immunodefi-ciency viruses (Cullen, 1998; Harris, 1996). Experimentalinfection of macaques with SIVmacDnef is characterizedby a low level of viral replication (Chakrabarti et al., 1995;Kestler et al., 1991). In humans, several long-term non-progressors have been identified, in which only nef-deleted proviruses were detected (Deacon et al., 1995;Kirchoff et al., 1995). These individuals display low viralloads and remained healthy for more than a decade,although some HIVDnef-infected patients show signs ofmmune defects (Greenough et al., 1999). The cellularnd virological mechanisms responsible for this role inivo remain unclear. Multiple activities of Nef have beenescribed in vitro: Nef interferes with cellular signal

transduction pathways and interacts with a number ofkinases (Saksela, 1997). Nef down-regulates the cellsurface expression of CD4, the primary receptor for HIVand SIV (Guy et al., 1987; Piguet et al., 1999). ReducedCD4 surface levels could prevent superinfections, in-crease the life span of viral envelope-expressing cells,enhance envelope incorporation into virions, and alteractivity of CD41 T cells (Benson et al., 1993; Lama et al.,1999; Ross et al., 1999; Schwartz et al., 1993; Skowronski

1

cTo whom correspondence and reprint requests should be ad-

ressed. Fax: 133 1 45 68 89 40. E-mail: schwartz@pasteur.fr.

225

et al., 1993). Nef also down-regulates the surface expres-sion of major histocompatibility complex class I mole-cules (MHC-I) (Le Gall et al., 1998; Schwartz et al., 1996),preventing the recognition and lysis of infected cells byboth CD81 cytotoxic T lymphocytes (CTLs) and NK cells(Cohen et al., 1999; Collins et al., 1998). In addition, Nef

xerts direct positive effects on the virus life cycle inrimary human lymphocytes and in macrophages (Aikennd Trono, 1995; de Ronde et al., 1992; Glushakova et al.,999; Miller et al., 1994; Spina et al., 1994), probably by

acilitating early (entry or postentry) steps of the replica-ive cycle (Aiken and Trono, 1995; Miller et al., 1994;chwartz et al., 1995). Finally, Nef mediates lymphocytehemotaxis and activation when expressed in macro-hages, indicating that the viral protein may also play a

ole in lymphocyte recruitment (Swingler et al., 1999).endritic cells (DCs) are also natural targets of HIV-1

eplication. However, the importance of Nef for viral rep-ication in these cells remains poorly documented.

DCs are the most potent antigen presenting cells thatan stimulate resting naive T lymphocytes and initiateTL responses (Banchereau and Steinman, 1998). Imma-

ure DCs residing in peripheral tissues, including thekin and several mucosal surfaces, are specialized inapturing and processing antigens. Immature DCs doot express high levels of surface molecules required forotent T-cell stimulation. After antigen capture, DCs ma-

ure and travel toward secondary lymphoid organs, pro-

essing antigens for presentation and acquiring the ca-

0042-6822/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

np

1mr

at

1at

t(amHtp(aStmihcHp

l

226 PETIT ET AL.

pacity to attract and activate resting CD81 CTLs duringthat journey. DCs are believed to play a crucial role in thepathogenesis of HIV-1 infection (Banchereau and Stein-man, 1998; Blauvelt, 1997; Knight and Patterson, 1997;Rowland-Jones, 1999). Immature DCs in mucosal tissuesmay represent the initial targets for HIV during primaryinfection, and these cells are suspected to be responsi-ble for the transport of virus to lymph nodes and transferto CD41 lymphocytes (Banchereau and Steinman, 1998;Blauvelt, 1997; Knight and Patterson, 1997; Rowland-Jones, 1999).

DCs express low levels of the HIV receptors CD4 andCCR5 (Ayehunie et al., 1997; Granelli-Piperno et al., 1996;Klagge and Schneider-Schaulies, 1999; O’Doherty et al.,1993; Rubbert et al., 1998). DCs are considered either

egative or weakly positive for CXCR4 expression, de-ending on the study (Ayehunie et al., 1997; Granelli-

Piperno et al., 1996; Klagge and Schneider-Schaulies,999; Rubbert et al., 1998). These contradictory resultsay be due to different maturation states of DCs, which

egulate chemokine receptor surface expression (Lin etal., 1998). HIV-1 replicates rather inefficiently in DCs.R5-tropic, but not X4-tropic, HIV-1 strains induce chemo-taxis and replicate in immature DCs (Granelli-Piperno et

l., 1998; Lin et al., 2000). However, both R5- and X4-ropic HIV-1 readily bind and enter DCs (Ayehunie et al.,

1997; Granelli-Piperno et al., 1999, 1996; Klagge andSchneider-Schaulies, 1999). HIV-1 transmission fromDCs to T cells involves a DC-specific receptor, DC-SIGN,which binds viral envelope glycoproteins and retains theattached virus in an infectious state (Geijtenbeek et al.,2000). Interestingly, DC-T cell clusters are the sites ofextremely efficient HIV-1 (Cameron et al., 1992; Granelli-Piperno et al., 1998, 1999; Pope et al., 1994) and SIV(Ignatius et al., 1998; Messmer et al., 2000; Pope et al.,

997) replication in culture, and evidence exists for anctive HIV-1 replication within DC-T clusters in lymphoid

issues (Frankel et al., 1996).The observation that Nef affects viral loads as soon as

he early stages of infection in humans and monkeysChakrabarti et al., 1995; Deacon et al., 1995; Kestler etl., 1991; Kirchoff et al., 1995) suggests that this factoray play an important role in the interactions betweenIV-1, DCs, and T cells. Recently, Messmer et al. reported

hat SIVmac239Dnef displayed decreased replicative ca-acity in cocultures of simian immature DCs and T cells

Messmer et al., 2000). However, because immature DCsre not susceptible to SIVmac239 replication, whetherIV Nef is required for viral replication in DCs or for viral

ransmission from DCs to T cells could not be deter-ined in this study (Messmer et al., 2000). We have

nvestigated here the role of Nef on HIV-1 replication inuman immature DCs-T cell cocultures and in separateultures of DCs or lymphocytes. Using various R5 and X4

IV-1 strains and their nef-deleted counterparts, we re-ort that in mixtures of immature DCs and T cells, the

evel of replication of Dnef was significantly lower thanthat of the wild-type. Altogether, our results indicate thatNef is required for optimal viral replication in immatureDC-T cell clusters and in T lymphocytes, but exerts onlya marginal role in immature DCs. We also show thatwild-type and Dnef virions are similarly captured andprocessed for MHC-I restricted exogenous presentationby DCs.

RESULTS

Role of Nef on HIV-1 replication in immature DC-Tcell cultures

To investigate the role of Nef on HIV-1 replication inprimary immature DC-T cell cultures, we compared thekinetics of viral growth of WT and Dnef isogenic viruses.Immature DCs have been reported to express CD4 andCCR5, albeit at lower levels than on T cells (Ayehunie etal., 1997; Granelli-Piperno et al., 1996; O’Doherty et al.,1993). Flow cytometry analysis indicated that monocyte-derived immature DCs used in this study expressed CD4,low levels of CCR5, and virtually no CXCR4 at the cellsurface (Fig. 1A). Since HIV replication in DCs dependson the tropism of the virus (Granelli-Piperno et al., 1998),we analyzed the replication of two R5 isolates, NLAD8and YU-2, and one X4 isolate, NL43. Immature DCs wereexposed to low virus doses (7 ng of p24 per 2 3 105 cells)in an overnight incubation; cells were then washed toremove extracellular virus and mixed with autologous Tlymphoblasts that had been activated 3 days beforecoculture. Viral replication was assessed by measuringp24 production in cell supernatants. All three WT virusesreplicated in these mixed cultures (Fig. 1B). With NLAD8and YU-2, the levels of p24 production reached 100–200ng/ml at days 8–12 postinfection (pi). They were lowerwith NL43 (peak at 20 ng/ml), suggesting that R5 strainsreplicate more efficiently than X4 strains in DC-T cellcultures. Interestingly, Dnef replication was significantlyimpaired when compared to WT viruses (Fig. 1B). More-over, the defect of Dnef viruses varied with the viralstrain. NLAD8Dnef replication was moderately affected,with about a threefold reduction in p24 production whencompared to WT virus (Fig. 1B). With YU-2 and NL43isolates, Dnef replication was more dramatically im-paired, reaching only 5–10 ng p24/ml at day 12 pi. Thiscorresponded to about a 10-fold reduction in p24 pro-duction when compared with WT viruses. Therefore, HIVreplication is significantly affected in primary DC-T cellcultures in the absence of Nef. The replicative defect ofDnef viruses is observed with both R5 and X4 strains andis thus independent of viral tropism. Of note, a similardefect of Dnef viruses was observed when cells wereinfected with a 10-fold higher viral inoculum (not shown)or when immature DCs were mixed with heterologous

activated T cells (not shown).

One attractive feature of the DC-T coculture model is

v moveonactivdays po

227Nef AND HIV REPLICATION IN DENDRITIC T-CELL COCULTURES

that it promotes the growth of HIV without prior activation

FIG. 1. Surface expression of HIV receptors in immature DCs and replevels of HIV receptors CD4, CXCR4, and CCR5 in immature DCs. CellmAbs and revealed with a phycoerythrin-labeled secondary antibody. Cisotypic mAb. (B) Replication of isogenic WT and Dnef isogenic HIV-1 in

iruses (7 ng of p24). After overnight incubation, cells were washed to rehad been activated 3 days before (top) or with 8 3 105 autologous nproduction in the culture supernatants over a 12-day period. Days pi:

of T cells. We therefore examined the effects of Nefdeletion on viral replication in cocultures of DCs and

autologous nonactivated PBLs. All three WT viruses rep-

of HIV-1 WT and Dnef in immature DCs-T cell cocultures. (A) Surfacestained with Leu3A (anti-CD4), 12G5 (anti-CXCR4), or 2D7 (anti-CCR5)ere analyzed by flow cytometry. CTRL, cells stained with an irrelevantre DCs-T cell cocultures. DCs (2 3 105) were exposed to the indicated

unbound viruses and mixed with 8 3 105 autologous lymphoblasts thatated lymphocytes. Viral replication was assessed by measuring p24stinfection. Data are representative of three experiments.

lications were

ells wimmatu

licated in these mixed cultures (Fig. 1B). Levels of p24production were much lower than when PBMCs were

ecgt(wc2oGrDarr

o

tdiepNt

228 PETIT ET AL.

activated before coculture. With NLAD8 and YU-2, p24production reached 10 ng/ml at days 12–14 pi and waseven lower with NL43 (peak below 5 ng/ml). Of note,none of the three viral strains replicated in nonactivatedPBMCs alone (not shown), confirming that viral replica-tion in this experimental system requires the presence ofDCs. As for DC-activated T-cell cocultures, Dnef replica-tion was significantly impaired when compared to WTviruses, and the Dnef defect varied with the viral strain(Fig. 1B). When compared to WT, NLAD8Dnef replicationwas moderately affected, with about a twofold reductionin p24 production virus. With YU-2 and NL43 isolates,Dnef defect was more marked (Fig. 1B). Altogether, theseexperiments indicate that Nef is required for efficientHIV-1 replication in DC-T cell cultures irrespective of theactivation state of T lymphocytes.

Influence of Nef on HIV-1 replication in immature DCs

We then characterized the replicative defect of Dnefviruses in immature DC-T cell cocultures. More precisely,we examined the role of Nef on viral replication in im-mature DCs and in T cells separately, and on viral trans-mission from immature DCs to T lymphocytes. To studythe role of Nef on HIV-1 replication in immature DCs, wecompared the viral growth of WT and Dnef isogenicNLAD8, YU-2, or NL43 strains. We used two viral inputs(7 and 50 ng of p24 per 2 3 105 target cells, respectively),since the effect of Nef may be more pronounced at lowviral doses (Miller et al., 1994; Spina et al., 1994). What-

ver the viral input, both NLAD8 and YU-2 strains repli-ated in immature DCs, whereas NL43 was unable torow in these cells (Fig. 2A). This confirmed the finding

hat immature DCs selectively replicate R5-tropic HIV-1Granelli-Piperno et al., 1998). At the lower viral input,

hich was the same as that used in immature DCs-T cellocultures, replication of WT NLAD8 or YU-2 reached–3 ng p24/ml at day 12 pi (Fig. 2). Therefore, as previ-usly reported (Cameron et al., 1992; Frank et al., 1999;ranelli-Piperno et al., 1998; Pope et al., 1995), viral

eplication is much less efficient in immature DCs than inCs-T cell cocultures (compare p24 levels in Figs. 2And 1B). At the higher viral input, virus production

eached 30 and 80 ng of p24 for NLAD8 and YU-2,espectively (Fig. 2A).

Interestingly, the replication of Dnef viruses was not, ornly slightly, affected in DCs. NLAD8 WT and Dnef

yielded similar p24 production levels at high viral input.Even at the lower viral input, replication of NLAD8Dnefwas not significantly affected (Fig. 2A). Similarly, only aslight difference (less than twofold) was observed be-tween YU-2 WT and Dnef viruses (Fig. 2A). Altogether,these results indicate that R5 HIV-1 strains are able toreplicate in immature DCs, albeit with a low efficiency.

Moreover, Nef is only marginally necessary for HIV-1replication in immature DCs.

Role of Nef on HIV-1 replication in primarylymphocytes

Since HIV-1 replication in immature DCs is not signifi-cantly affected by the absence of Nef, we next studied thegrowth of WT and Dnef isogenic NLAD8, YU-2, or NL43strains in primary lymphocytes. The importance of Nef onHIV replication in PBMCs has been demonstrated (Aikenand Trono, 1995; de Ronde et al., 1992; Glushakova et al.,1999; Miller et al., 1994; Spina et al., 1994). In some studies,it has been reported that HIVDnef replicated poorly only inCD41 cells stimulated after infection, when virus inoculumwas low (Miller et al., 1994; Spina et al., 1994). In others, thereplicative defect was also manifest in fully activated PB-MCs and was independent of the multiplicity of infection(Aiken and Trono, 1995; de Ronde et al., 1992). In theimmature DC-T cell cultures in which HIVDnef replicationwas impaired (Fig. 1), we used lymphocytes which hadbeen activated before incubation with virus-exposed DCs.We therefore used the same protocol and analyzed viralreplication in PBLs that had been activated with PHA 3 daysbefore infection. As for immature DCs, we used two viralinputs (7 and 50 ng of p24 per 8 3 105 target cells, respec-ively). Wild-type NLAD8, YU-2, and NL43 induced p24 pro-uction in supernatants of infected cells (Fig. 2B), confirm-

ng that R5 and X4 strains replicate in PBLs. Both strainsfficiently replicate and reached equivalent peaks of p24roduction at each viral inoculum (100–150 ng p24/ml forLAD8, 25–50 ng p24/ml for YU-2 and NL43). In contrast,

he replicative capacity of Dnef mutants was dramaticallyaffected. As in immature DC-T cell cultures, the defect ofDnef viruses varied with the viral strain. NLAD8Dnef repli-cation was only two- to threefold lower than WT virus (Fig.2B). With YU-2 and NL43 isolates, Dnef replication wasmore significantly impaired (Fig. 2B). Nef-mutated virusesreplicated very poorly, even at the higher viral inoculum,plateauing at only 5 ng/ml p24. Taken together, these re-sults demonstrated a consistently positive influence of Nefon HIV-1 replication in primary lymphocytes.

The absolute levels of viral replication, as well as thedefects of nef-deleted viruses in primary cells, may varyaccording to donors. To obtain a better insight of theinfluence of Nef on viral replication, results from three tofive independent experiments are shown in Fig. 3. Dif-ferences between WT and Dnef replication levels wereexpressed as ratios of p24 concentrations at the peak ofviral production, with 100% corresponding to WT values.Calculations were performed for each viral strain and foreach culture condition, including DCs and activated Tcells separately and DC-activated T cell cocultures. Irre-spective of cell culture conditions, Dnef virus defect wasminimal with NLAD8 strain, whereas it was much morepronounced with YU-2 and NL43 strains (Fig. 3). There-fore, the replicative defect associated with the absence

of Nef is strain-dependent. Since NLAD8 and NL43 areisogenic viruses differing only in the env gene, our result

iD

229Nef AND HIV REPLICATION IN DENDRITIC T-CELL COCULTURES

indicates that the Nef defect may be modulated by Envproteins. Data also indicates that Nef requirement foroptimal viral growth varied with cell culture conditions. Inimmature DCs-T cell cocultures, Dnef replicated lessefficiently than WT (2-fold, 10-fold, and 5-fold reduction of

FIG. 2. HIV-1 WT and Dnef replication in separate cultures of immat(8 3 105 cells, B) were exposed to either 50 ng (high inoculum panel) oncubation, cells were washed to remove unbound virus. Viral replicat

ata are representative of three experiments.

p24 peak levels for NLAD8, YU-2, and NL43, respectively).In immature DCs alone, the effect of Nef was less obvious

(Dnef p24 levels reduced to 75 and 60% of those of WT forNLAD8 and YU-2, respectively). In contrast, Nef was re-quired for efficient virus production in lymphocytes, as p24levels reached by Dnef viruses were 4-, 10-, and 20-foldlower than for WT NLAD8, NL43, and YU-2, respectively

s and activated PBLs. DCs (2 3 105 cells, A) and PHA-activated PBLs(low inoculum panel) p24 of the indicated viral strains. After overnight

assessed by measuring p24 production in the culture supernatants.

ure DCr 7 ng

ion was

(Fig. 3). Taken together, these experiments indicate a moreprominent role of Nef in lymphocytes than in DCs.

(b

epende

230 PETIT ET AL.

Role of Nef during transmission of X4-tropic HIV-1from DC to T cells

Since Nef is required for efficient HIV-1 replication inDCs-T cell cocultures, the transmission step betweenDCs and T cells may be impaired with Dnef viruses. Wetherefore developed an experimental system allowingthe study of the capacity of DCs to transmit infectiousviral particles. Since R5 strains replicate in immature

FIG. 3. Comparison of HIV-1 WT and Dnef replication in immature Dwere infected as described in Figs. 1–2 with the indicated viral strainssupernatants. WT and Dnef replication levels were compared by calccorresponding to WT values. Data are mean 6 SD of three to five indimmature DCs.

DCs even in the absence of T cells (see Fig. 2A), it wasimportant to use an X4 strain (NL43) for this analysis.

7 ng of p24) (C). A 5-day incubation period at 37°C of cell-free NL43 WT or Dy measuring p24 production in the culture supernatants over a 14-day perio

With the aim to overcome possible artifacts that couldresult from the altered replication of NL43Dnef in PBLs(see Fig. 2B), DCs were cocultivated with immortalized Tcells, in which Nef is dispensable for viral replication(Aiken and Trono, 1995; Miller et al., 1994; Schwartz et al.,1995; Spina et al., 1994). As we previously reported (LeGall et al., 1997; Schwartz et al., 1995), cell-free infectionof the CD41 lymphoid line MT4 with NL43 WT or Dnef

5

ated PBLs cocultures, in immature DCs, and in activated PBLs. Cellsreplication was assessed by measuring p24 production in the culture

the ratio of p24 released at the peak of viral production with 100%nt experiments. N.A., not applicable since NL43 does not replicate in

virions (7 ng of p24 per 8 3 10 cells) produced equiva-lent amounts of p24 with similar kinetics (Fig. 4C). To

FIG. 4. Viral transmission and stability of HIV-1 WT and Dnef viruses after capture by immature DCs. Immature DCs (2 3 105) were exposed to NL43WT or Dnef (7 ng of p24). After overnight incubation, cells were washed to remove unbound virus. Cells were then mixed with 8 3 105 MT4 cells (A)or incubated 5 days before addition of 8 3 105 MT4 cells (B). As a control, MT4 cells (8 3 105) were directly exposed to cell-free NL43 WT or Dnef

C-activ. Viralulating

nef fully abrogates viral infectivity (D). Viral replication was assessedd. Data are representative of two experiments.

taiotf

HDrbstts

231Nef AND HIV REPLICATION IN DENDRITIC T-CELL COCULTURES

monitor the transmission of viral particles from DCs toMT4 cells, DCs were pulsed overnight with NL43 WT orDnef virions (7 ng of p24 per 2 3 105 cells). Cells werehen washed extensively to remove extracellular virionsnd mixed with MT4 cells. Efficient replication occurred

n recipient MT4 cells (Fig. 4A). The same kinetics werebserved for WT and Dnef viruses (Fig. 4A), indicating

hat Nef is dispensable for transmitting viral particlesrom DCs to MT4 cells.

It has been reported recently that transmission ofIV-1 from DCs to T cells involves a specific DC receptor,C-SIGN, which binds the viral envelope protein and

etains the captured virus in an infectious state (Geijten-eek et al., 2000). Using THP1 monocytoid cell linestably expressing DC-SIGN, it was demonstrated that

his molecule is able to capture and retain HIV-1 for morehan 4 days, after which the virus can still infect permis-ive target cells (Geijtenbeek et al., 2000). We thus ex-

amined whether immature DCs are able to capture viralparticles for long period of times and the influence of Nefin this process. DCs were pulsed overnight with NL43WT and Dnef virions (7 ng of p24 per 2 3 105 cells). Cellswere then washed extensively and 5 days later they weremixed with MT4 cells. Efficient viral replication occurredonly after addition of recipient MT4 cells (Fig. 4B). Again,similar kinetics of viral production were observed for WTand Dnef viruses (Fig. 4B). Of note, a 5-day incubation at37°C in culture medium of cell-free NL43 WT and Dnefabolished viral infectivity (Fig. 4D). Therefore, primaryimmature DCs are able to capture and retain HIV-1 in aninfectious state for at least 5 days. NL43Dnef particleswere captured and transmitted to target cells as effi-ciently as their WT counterparts, even after 5 days. Alto-gether, these experiments strongly suggest that Nefdoes not affect virion capture, long-term intracellularaccumulation, and transmission of X4 strains to T lym-phocytes by DCs.

Role of Nef on MHC-I restricted presentation of HIV-1virion antigens by DCs

A major function of DCs is to stimulate resting naive Tlymphocytes and to initiate primary CTL responses(Banchereau and Steinman, 1998). We have recentlyshown that immature DCs capture and process incomingHIV-1 virions through the MHC-I-restricted presentationpathway, leading to CTL activation in the absence of viralprotein neosynthesis (Buseyne et al., 2001). Presentationof exogenous HIV-1 antigens required adequate virus-receptor interactions and fusion of viral and cellularmembranes. We compared the ability of WT and Dnefvirions to be processed for presentation by this exoge-nous MHC-I-restricted pathway. To this aim, immatureHLA-A21 DCs were exposed to various HIV-1 virions in

the presence of AZT, to inhibit reverse transcription andsubsequent synthesis of viral proteins. After overnight

incubation, extracellular particles were washed awayand DCs were cocultivated with the HLA-A2-restrictedCD81 CTL cell line EM71-1. EM71-1 cells were derivedfrom an HIV-infected patient and recognize a well-char-acterized immunodominant epitope of the Gag p17 pro-tein (Buseyne et al., 2001). DCs pulsed with the syntheticGag epitope (SL9) activated EM71-1 cells specifically, asmeasured by IFN-g production in an Elispot assay (Fig.5A). This epitope is present in BRU (which is a X4-tropicstrain) and YU-2, but not in NL43 or NLAD8 isolates(Buseyne et al., 2001). We first tested the ability of imma-ture DCs to present the Gag epitope after exposure tothe BRU isolate. We observed that DCs pulsed witheither WT or Dnef viruses efficiently activated EM71-1cells (Fig. 5A). Immature DCs also efficiently activatedeffector cells upon exposure to the YU-2 strain (Fig. 5B).Again, no significant difference was noted between WTand Dnef viruses, at either high or low viral inoculum (Fig.5B). Of note, HLA-A22 DCs exposed to various HIVstrains were not recognized by EM71-1 cells (not shown),indicating that the exogenous presentation of HIV-1epitopes is appropriately MHC-I restricted. Altogether,these experiments indicate that DCs present a Gagepitope upon exposure to incoming virions in the ab-sence of viral protein neosynthesis. The exogenous pre-sentation is observed with similar efficiencies with WTand nef-deleted virions. Therefore, the presence of Nef invirus producer cells does not significantly influence theability of progeny virions entering DCs to be processedfor exogenous MHC-I presentation.

DISCUSSION

We demonstrate here deficient replication of variousnef-deleted HIV strains in immature DC-T cell cocultures.DCs are thought to play a crucial role in HIV-1 transmis-sion and propagation. Immature DCs residing in the skinand the mucosa might be the first cell targets of HIV-1.They are believed to transport virus particles and trans-mit a vigorous infection to T cells in lymph nodes (Cam-eron et al., 1996; Knight and Patterson, 1997). Althoughthere is considerable evidence that DCs from blood,skin, or lymphoid organs harboring HIV-1 can be found inpatients, the contribution of infected DCs to viral burdenis probably marginal (Cameron et al., 1996). In contrast,HIV-1 replication is intense in conjugates of DCs and Tcells (Cameron et al., 1996) and active viral productionhas been demonstrated in lymphoid tissues in DC–lym-phocyte clusters (Frankel et al., 1996). In vivo, the Nefgene product confers a replicative advantage as early asthe acute phase of infection (Chakrabarti et al., 1995;Kestler et al., 1991). The main conclusion of our study isthat high levels of viral replication would occur whenDCs exposed to wild-type virus encounter CD41 T cells,

whereas Dnef virus would disseminate much less effi-ciently in this environment.

ppsPcic

i

ow

ment dD of du

232 PETIT ET AL.

In vitro, HIV-1 produced in DC–lymphocyte coculturesrobably originates from different sources. A high viralroduction from DC-T cell syncytia, due to the coexpres-ion of transcription factors, has been reported (Granelli-iperno et al., 1995). However, analysis of the cellularomponents of virions indicated that virus produced from

mmature DC-T cell cocultures primarily originates from Tells (Frank et al., 1999). We have dissected the replica-

tive defect of Dnef HIV-1 in immature DC-T cocultures.The replication rates of HIV-1 in immature DCs alonewere equivalent for wild-type and Dnef, whereas themost dramatic replicative defect of Dnef HIV-1 was ob-served in lymphocytes. Therefore, the compromised rep-lication of Dnef in DCs-T cells mixture is most likely dueto the lymphocytic component of the coculture, ratherthan to immature DCs. Although initially controversial, itis currently believed that purified DCs do not replicateHIV-1 with high efficiency (Cameron et al., 1996; Granelli-Piperno et al., 1998, 1999). Accordingly, we report herethat the X4 strain NL43 does not grow in immature DCs,whereas levels of viral production of the R5 stains YU-2and NLAD8 were about 10–100 times lower in DCs alonethan in T cells or in DCs-T cell cocultures. Furthermore,this low virus production was almost unchanged withDnef viruses. This marginal effect of Nef in immatureDCs could be due to a reduced number of replicativecycles of HIV-1, which will be consistent with the lowlevels of viral production in these cells. Alternatively,immature DCs might be “permissive” for Dnef replication,in a situation reminiscent of that of Vif (Madani andKabat, 1998; Simon et al., 1998). On the other hand, ourobservation that Nef is not required for viral replication in

FIG. 5. MHC-I presentation of a Gag epitope derived from incomindividuals were used as stimulator cells in an IFN-g-Elispot assay. T

restricted epitope (SL9) from the Gag p17 protein. Stimulating cells werr in an independent experiment, to YU-2 WT and Dnef at either low orith EM71-1 cells. Activity of EM71-1 cells is depicted as the number of

cells were pulsed with the cognate peptide (SL9, 1 mg/ml). In the experiwhen the cognate peptide was used (not shown). Data are mean 6 S

immature DCs provides indirect evidence for the limitedcontribution of these cells to the viral burden in vivo.

DCs bind and capture HIV-1 through the DC-specifictype C lectin, DC-SIGN, which is the receptor for ICAM-3(Geijtenbeek et al., 2000). Transmission of DC-associ-ated virus to T cells involves the interaction of otheradhesion molecules, including LFA1/ICAM-1 and LFA3/CD2 (Tsunetsugu-Yokota et al., 1997), indicating thatclose contacts between the two cell types are requiredfor this process. We have investigated the role of Nefduring viral transmission from DCs to T cells. However,R5 HIV-1 strains replicate in immature DCs in the ab-sence of T cells, whereas Dnef viruses are altered intheir replication in primary lymphocytes. These consid-erations precluded the study of the transmission of R5strains and the use of primary lymphocytes for this partof our study. We therefore followed the transmission ofthe X4 strain NL43 from DCs to MT4 lymphoid cells, inwhich Nef is dispensable for viral replication. We ob-served similar replication kinetics for WT and Dnef vi-ruses, strongly suggesting that Nef is not required forX4-tropic HIV-1 transmission from DCs to T cells. Theproductive infection of DCs by HIV-1 and their ability tocapture virus are thought to be mediated through sepa-rate pathways (Blauvelt et al., 1997). Captured particlesaccumulate in a trypsin-resistant compartment, in whichthey retain their infectivity for several days (Blauvelt et al.,1997; Geijtenbeek et al., 2000). We show here that im-mature DCs similarly retain WT and Dnef HIV-1 strainNL43 in an infectious state for at least 5 days. Altogether,our experiments strongly suggest that Nef is not requiredfor virion capture, long-term intracellular accumulation,and transmission of X4 strains to T lymphocytes by DCs.

Replication of HIV-1 in vivo is highly productive and

-1 virions. Immature DCs prepared from HLA-A21 HIV-seronegativeector was the CD81 CTL line EM71-1, which recognizes an HLA-A2-eated with AZT (5 mM), exposed to BRU WT and Dnef (250 ng p24) (A),oculum (25 or 250 ng p24, respectively) (B). Cells were then incubatedositive cells for 500 or 2500 effectors. As a positive control, stimulating

epicted in B, the IFN-g positive cells were too numerous to be countedplicates and are representative of two independent experiments.

ng HIVhe eff

e pretrhigh inIFN-g p

continuous. Productively infected lymphocytes have ahalf-life of less than 2 days and contribute to approxi-

itT

csppAD

233Nef AND HIV REPLICATION IN DENDRITIC T-CELL COCULTURES

mately 99% of the viremia (Perelson et al., 1996). Weshow here that Dnef replication is severely compromisedn activated lymphocytes, confirming the importance ofhe viral protein on HIV replication in PBMCs (Aiken androno, 1995; Brown et al., 1999; de Ronde et al., 1992;

Glushakova et al., 1999; Miller et al., 1994; Spina et al.,1994). Thus, the low viral loads observed in SIVDnef or inHIVDnef-infected hosts (Deacon et al., 1995; Kestler etal., 1991; Kirchoff et al., 1995) are most likely the conse-quence of the replicative defect of mutated viruses inlymphocytes. In unstimulated lymphocytes, it has beenreported that Nef may provide an activating signal nec-essary for reaching a high level of productive infection(Alexander et al., 1997; Brown et al., 1999; Glushakova etal., 1999; Wang et al., 2000). That Nef is also required inactivated T cells suggests that other mechanisms areinvolved in the positive effects of this viral protein. Inter-estingly, we observed strain-specific differences in thedependence on Nef for replication in activated PBMCs.The defect of Dnef virus was minimal with NLAD8,whereas it was much more pronounced with YU-2 andNL43. Coreceptor usage apparently does not account forthese differences, because both YU-2 and NLAD8 usesCCR5 for entry. In primary macrophages, opposite re-sults concerning the requirement of Nef for HIV-1 repli-cation have been reported. Whereas Dnef replication ofYU-2 or SF162 strains is severely compromised in mac-rophages (Brown et al., 1999; Miller et al., 1994), similarreplication efficiencies were observed with Dnef andwild-type viruses for the ADA isolate or for a NL43/BALchimera (Meylan et al., 1998; Swingler et al., 1999). Con-sistent with our observation in lymphocytes, these differ-ences may also be virus strain-dependent. Molecularand cellular mechanisms underlying these differencesare not fully understood. The NLAD8 and NL43 virusesthat we used here are isogenic strains differing only inthe env gene. This indicates that the Nef defect may bemodulated by Env proteins. We previously reported thatNef modulates the surface expression and cytopathiceffects of gp120/gp41 complexes (Schwartz et al., 1993).Nef also enhances envelope glycoprotein incorporationinto virions (Lama et al., 1999; Ross et al., 1999). There-fore, it will be of interest to compare the effects of Nef ongp120/gp41 function using different isogenics strains.

Messmer et al. recently reported a positive role of SIVNef on SIVmac239 replication in cocultures of simianimmature DCs and T cells (Messmer et al., 2000). Ourresults indicate that SIV and HIV Nef share functionalhomologies in supporting viral replication in immatureDC-T cells mixtures. However, our study differs from thatof Messmer et al. in several respects. First, the effects ofSIV Nef on viral replication in DCs alone or on viraltransmission from DCs to T cells were not examined byMessmer et al. because immature DCs are not suscep-

tible to SIVmac239 replication (Messmer et al., 2000). Weshow here that the prominent role of HIV-1 Nef on viral

replication in this experimental system occurs in lympho-cytes. Second, we observed that Nef is required forefficient HIV-1 replication in DC-T cell cultures irrespec-tive of the activation state of T lymphocytes. In contrast,Messmer et al. reported that when DC-T cell cultureswere activated with the superantigen SEB, both wt andDnef SIVmac239 replicated with identical levels and ki-netics (Messmer et al., 2000). This discrepancy could bedue to experimental differences, or to inherent differ-ences between human and monkey cells or virus strainsused. On the other hand, the functional organization ofSIV and HIV-1 Nef is different and includes variations inthe ability to bind the TCR (Howe et al., 1998; Schaefer etal., 2000). This could influence the respective abilities ofSIV and HIV-1 Nef to stimulate viral replication in acti-vated or inactivated lymphocytes. Moreover, by usingthree HIV-1 isolates, we point out that the requirement forNef in DC-T cell cocultures is viral strain-specific anddepends on Env proteins.

DCs can stimulate resting, naive T lymphocytes and,as a consequence, initiate CTL immune responses invivo (Banchereau and Steinman, 1998). Following antigencapture in peripheral tissues, immature DCs migrate tosecondary lymphoid organs. As cells travel, they mature,process antigens for presentation, and acquire the abilityto attract and activate resting CD81 cytotoxic T lympho-

ytes. In most cells, MHC-I molecules associate exclu-ively with peptides derived from endogenous cytosolicroteins. However, stimulation of CTLs against trans-lants, tumors, bacteria, or viruses that do not infectPCs requires presentation of exogenous antigens byCs (Albert et al., 1998; Sigal et al., 1999; Yewdell et al.,

1999). We have shown that immature DCs process anti-gens from incoming HIV-1 virions through the MHC-I-restricted presentation pathway, leading to CTL activa-tion in the absence of viral protein neosynthesis(Buseyne et al., 2001). Presentation of exogenous HIV-1antigens required adequate virus-receptor interactionsand fusion of viral and cellular membranes (Buseyne etal., 2001). Thus, the only step of the viral cycle that isrequired for CTL stimulation by DCs is virus entry. Pre-sentation of epitopes before the synthesis of viral pro-teins may be essential for CTL activation, since Nefdown-regulates MHC-I expression and decreases im-mune recognition of cells that become productively in-fected (Collins et al., 1998; Schwartz et al., 1996). Weshow here that presentation of a Gag epitope derivedfrom incoming virions is observed with similar efficien-cies for WT and nef-deleted viruses. Our results stronglysuggest that Nef does not influence the ability of incom-ing virions to be processed for MHC-I exogenous pre-sentation. Moreover, they provide a likely explanation ofhow anti-HIV CTLs may be induced despite the down-modulating activity of Nef on MHC-I.

In summary, we have shown here that Nef is requiredfor optimal HIV-1 replication in mixtures of immature DCs

vUao

F

cbsmotD(asSa

V

a(wc

234 PETIT ET AL.

and T cells, in a system relevant for the physiopathologyof the infection. Furthermore, wild-type and Dnef virionsare similarly captured and processed for MHC-I re-stricted exogenous presentation by DCs. Altogether,these results help explain how Nef affects viral spreadand immune responses in vivo.

MATERIALS AND METHODS

Generation of mononuclear subsets

DCs were prepared as described using a VacCellprocessor (IDM, Paris, France) (Goxe et al., 1998). Briefly,peripheral blood mononuclear cells (PBMCs) from leu-kapheresis were cultured 7 days in serum-free AIM-Vmedium (Gibco) supplemented with 500 U/ml GM-CSF (akind gift from Novartis) and 50 ng/ml IL-13 (Sanofi), andDCs were isolated by elutriation. This isolation proce-dure yielded CD1a1 MHC-I1, MHC-II1, CD642, CD832,CD80 low, CD86 low cells, a phenotype corresponding toimmature DCs. DC purity was .95%. PBMCs were acti-

ated with PHA and cultivated in the presence of IL-2 (50/ml, Chiron). Nonactivated PBMCs were allowed todhere for 2 h at 37°C. Nonadherent cells were washedff and used for further studies.

low cytometry analysis

DCs were washed in PBS and incubated with mono-lonal antibodies (mAbs) for 30 min at 4°C in PBA (1%ovine serum albumin, 0.1% sodium azide in PBS). Aftertaining, cells were fixed in PBA containing 1% parafor-aldehyde and analyzed with a FACScalibur cytoflu-

rometer (Becton–Dickinson). Cells were stained withhe following mAbs: anti-CD4: Leu 3A (IgG1, Becton–

ickinson), anti-CXCR4 (IgG2a, 12G5), and anti-CCR5IgG2a, 2D7), obtained through the NIH AIDS Researchnd Reference Reagent Program. As a control, cells weretained with an isotypic IgG2a mAb (Becton–Dickinson).econdary antibody was a phycoerythrin-labeled goatnti-mouse antibody (Southern Biotechnology).

iruses and infections

The use of X4-tropic NL43 WT and Dnef, and BRU WTnd Dnef HIV-1, has been described elsewhere

Schwartz et al., 1995). The R5-tropic NLAD8 WT strainas a kind gift of Eric Freed (Englund et al., 1995). It

arries a portion of the env gene from the R5-tropic cloneAD8 on a pNL4–3 backbone (Englund et al., 1995).NLAD8 Dnef was generated by inserting a frameshiftmutation at the unique XhoI site of the viral genome. TheR5-tropic YU-2 WT and Dnef strains were a kind gift ofMark Feinberg (Miller et al., 1994). Viruses were pro-duced by transient transfection and used for infectionsas described (Schwartz et al., 1998). For all strains, in

1 1

single-cycle replication assays using HeLa-CD4 CCR5LTR–LacZ indicator cells (Marechal et al., 1998), Dnef

infectivity was reduced by 20 to 60% compared to that ofWT (not shown). Infections were carried out as described(Schwartz et al., 1995). When stated, PBMCs were acti-vated 3 days before infection with the indicated viralstrains. PBMCs were maintained in the presence of IL-2throughout the experiments. For coculture infections,DCs were first exposed to the indicated doses of viruses.After an overnight incubation at 37°C, cells were washedand target cells (PBMCs or MT4 cells) were then addedat a ratio of four targets for one DC. After infection, DCswere cultivated in RPMI medium (Gibco) supplementedwith 10% human AB1 serum. Similar viral replicationcurves were obtained when DCs were cultivated in se-rum-free AIM-V medium supplemented with GM-CSFand IL-13 (not shown). Viral replication was assessed bymeasuring p24 production by ELISA (Du Pont deNemours-NEN) in culture supernatants.

CTL line

The CTL line EM71-1 has been described elsewhere(Buseyne et al., 2001). It was derived from a child perinatallyinfected with HIV-1 by repeated stimulations of PBMC withirradiated autologous B-EBV cells coated with the p17 Gagpeptide SLYNTVATL (SL9, originally described by Tsomideset al., 1994) in the presence of allogeneic irradiated PBMC.Peptide recognition was HLA-A2 restricted, with a SD50 of0.5 ng/ml in 51Cr release assays (not shown). This epitopeis present in BRU and YU-2 HIV-1 strains. Ninety-sevenpercent of EM71-1 cells were CD81 and CD31, and 94%were stained with SL9-HLA-A2 tetramers (Buseyne et al.,2001). CTL culture conditions were as described (Buseyneet al., 1998).

Elispot assay

The reverse-transcriptase inhibitor AZT (5 mM, Sigma)was added to cells 3 to 5 h before exposure to virusesand maintained throughout the assay. DCs (2 3 106 cells)were exposed to the indicated viruses for 1 h in a volumeof 1 ml and diluted twice in fresh medium before over-night incubation. Viral inoculum was 25 or 250 ng ofp24/2 3 106 cells. Stimulator cells (DCs) were washedtwice before incubation with effectors. IFN-g productionwas measured in a Elispot assay as described else-where (Buseyne et al., 2001). Briefly, targets and effectorswere incubated overnight in nitrocellulose-bottomed 96-well plates (Millipore) coated with anti-IFN-g mAb 1-D1K(15 mg/ml, Mabtech). IFN-g production was revealed bysequential incubations with biotinylated anti-IFN-g mAb7-B6-1 (1 mg/ml, Mabtech), streptavidine-alkaline phos-phatase (0.5 U/ml, Boehringer Mannheim), and BCIP-NBT substrate (Promega). Positive spots were countedusing a binocular microscope.

ACKNOWLEDGMENTS

We thank Susan Michelson and Nathalie Sol-Foulon for critical read-ing of the manuscript. We thank Eric Freed, Mark Feinberg, and the NIH

A

A

A

B

B

F

G

G

G

G

G

G

G

G

G

H

H

I

235Nef AND HIV REPLICATION IN DENDRITIC T-CELL COCULTURES

AIDS Research and Reference Reagent Program for the kind gift ofreagents. This work was supported by grants from the Agence Natio-nale de Recherche sur le SIDA (ANRS), SIDACTION, and the PasteurInstitute. C.P. is a fellow of the ANRS.

REFERENCES

Aiken, C., and Trono, D. (1995). Nef stimulates human immunodefi-ciency virus type 1 proviral DNA synthesis. J. Virol. 69, 5048–5056.

lbert, M. L., Sauter, B., and Bhardwaj, N. (1998). Dendritic cells acquireantigen from apoptotic cells and induce class I-restricted CTLs.Nature 392, 86–89.

lexander, L., Du, Z., Rosenzweig, M., Jung, J. U., and Desrosiers, R. C.(1997). A role for natural simian immunodeficiency virus and humanimmunodeficiency virus type 1 nef alleles in lymphocyte activation.J. Virol. 71(8), 6094–6099.

yehunie, S., Garcia-Zepeda, E. A., Hoxie, J. A., Horuk, R., Kupper, T. S.,Luster, A. D., and Ruprecht, R. M. (1997). Human immunodeficiencyvirus-1 entry into purified blood dendritic cells through CC and CXCchemokine coreceptors. Blood 90(4), 1379–1386.

anchereau, J., and Steinman, R. M. (1998). Dendritic cells and thecontrol of immunity. Nature 392(6673), 245–252.

enson, R. E., Sanfridson, A., Ottinger, J. S., Doyle, C., and Cullen, B. R.(1993). Downregulation of cell-surface CD4 expression by simianimmunodeficiency virus nef prevents viral super infection. J. Exp.Med. 177, 1561–1566.

Blauvelt, A. (1997). The role of skin dendritic cells in the initiation ofhuman immunodeficiency infection. Am. J. Med. 102, 16–20.

Blauvelt, A., Asada, H., Saville, M. W., Klaus Kovtun, V., Altman, D. J.,Yarchoan, R., and Katz, S. I. (1997). Productive infection of dendriticcells by HIV-1 and their ability to capture virus are mediated throughseparate pathways. J. Clin. Invest. 100, 2043–2053.

Brown, A., Wang, X., Sawai, E., and Cheng-Mayer, C. (1999). Activationof the PAK-related kinase by human immunodeficiency virus type 1Nef in primary human peripheral blood lymphocytes and macro-phages leads to phosphorylation of a PIX-p95 complex. J. Virol.73(12), 9899–9907.

Buseyne, F., Chaix, M. L., Fleury, B., Manigard, O., Burgard, M., Blanche,S., Rouzioux, C., and Riviere, Y. (1998). Cross-clade-specific cytotoxicT lymphocytes in HIV-1-infected children. Virology 250(2), 316–324.

Buseyne, F., Le Gall, S., Boccaccio, C., Abastado, J. P., Lifson, J. D.,Arthur, L. O., Riviere, Y., Heard, J. M., and Schwartz, O. (2001).MHC-I-restricted presentation of HIV-1 virion antigens without viralreplication. Nat. Med. 7, 344–349.

Cameron, P., Pope, M., Granelli-Piperno, A., and Steinman, R. M. (1996).Dendritic cells and the replication of HIV-1. J. Leukoc. Biol. 59,158–171.

Cameron, P. U., Freudhenthal, P. S., Barker, J. M., Gezelter, S., Inaba, K.,and Steinman, R. M. (1992). Dendritic cells exposed to human im-munodeficiency virus type-1 transmit a vigorous cytopathic infectionto CD41 cells. Science 257, 383–387.

Chakrabarti, L., Baptiste, V., Khatissian, E., Cumont, M. C., Aubertin,A. M., Montagnier, L., and Hurtrel, B. (1995). Limited viral spread andrapid immune response in lymph nodes of macaques inoculated withattenuated simian immunodeficiency virus. Virology 231, 535–548.

Cohen, G. B., Gandhi, R. J., Davis, D. M., Mandelboim, O., Chen, B. K.,Strominger, J. L., and Baltimore, D. (1999). The selective down-regulation of class I major histocompatibility complex proteins byHIV-1 protects HIV-infected cells from NK cells. Immunity 10, 661–671.

Collins, K. L., Chen, B. K., Kalams, S. A., Walker, B. D., and Baltimore, D.(1998). HIV-1 Nef protein protects infected primary cells againstkilling by cytotoxic T lymphocytes. Nature 391, 397–401.

Cullen, B. R. (1998). HIV-1 auxiliary proteins: Making connections in a

dying cell. Cell 93, 685–692.

de Ronde, A., Klaver, B., Keulen, W., Smit, L., and Goudsmit, J. (1992). K

Natural HIV-1 Nef accelerates virus replication in primary humanlymphocytes. Virology 188, 391–395.

Deacon, N. J., Tsykin, A., Solomon, A., Smith, K., Ludford-Menting, M.,Hooker, D. J., McPhee, D. A., Greenway, A. L., Ellet, A., Chatfield, C.,Lawson, V. A., Crowe, S., Maerz, A., Sonza, S., Learmont, J., Sullivan,J. S., Cunninghan, A., Dwyer, D., Dowton, D., and Mills, J. (1995).Genomic structure of an attenuated quasi species of HIV-1 from ablood transfusion donor and recipients. Science 270, 988–991.

Englund, G., Theodore, T. S., Freed, E. O., Engelman, A., and Martin,M. A. (1995). Integration is required for productive infection of mono-cyte-derived macrophages by human immunodeficiency virus type 1.J. Virol. 69, 3216–3219.

Frank, I., Kacani, L., Stoiber, H., Stossel, H., Spruth, M., Steindl, F.,Romani, N., and Dierich, M. P. (1999). Human immunodeficiency virustype 1 derived from cocultures of immature dendritic cells withautologous T cells carries T-cell-specific molecules on its surfaceand is highly infectious. J. Virol. 73, 3449–3454.

rankel, S. S., Wenig, B. M., Burke, A. P., Mannan, P., Thompson, L. D.,Abbondanzo, S. L., Nelson, A. M., Pope, M., and Steinman, R. M.(1996). Replication of HIV-1 in dendritic cell-derived syncytia at themucosal surface of the adenoid. Science 272(5258), 115–117.

eijtenbeek, T. B., Kwon, D. S., Torensma, R., Van Vliet, S. J., VanDuijnhoven, G. C., Middel, J., Cornelissen, I. L., Nottet, H., KewalRa-mani, V., Littman, D., Figdor, C. G., and Van Kooyk, Y. (2000). DC-SIGN,a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell 100, 587–597.

lushakova, S., Grivel, J. C., Suryanarayana, K., Meylan, P., Lifson, J. D.,Desrosiers, R., and Margolis, L. (1999). Nef enhances human immu-nodeficiency virus replication and responsiveness to interleukin-2 inhuman lymphoid tissue ex vivo. J. Virol. 73(5), 3968–3974.

oxe, B., Latour, N., Bartholeyns, J., Romet-Lemonne, J. L., and Chokri,M. (1998). Monocyte-derived dendritic cells: Development of a cel-lular processor for clinical applications. Res. Immunol. 149, 643–646.

ranelli-Piperno, A., Delgado, E., Finkel, V., Paxton, W., and Steinman,R. M. (1998). Immature dendritic cells selectively replicate macro-phagetropic (M-tropic) human immunodeficiency virus type 1, whilemature cells efficiency transmit both M- and T-tropic virus to T cells.J. Virol. 72, 2733–2737.

ranelli-Piperno, A., Finkel, V., Delgado, E., and Steinman, R. M. (1999).Virus replication begins in dendritic cells during the transmission ofHIV-1 from mature dendritic cells to T cells. Curr. Biol. 9(1), 21–29.

ranelli-Piperno, A., Moser, B., Pope, M., Chen, D., Wei, Y., Isdell, F.,O’Doherty, U., Paxton, W., Koup, R., Mojsov, S., Bhardwaj, N., Clark-Lewis, I., Baggiolini, M., and Steinman, R. M. (1996). Efficient inter-action of HIV-1 with purified dendritic cells via multiple chemokinecoreceptors. J. Exp. Med. 184(6), 2433–2438.

ranelli-Piperno, A., Pope, M., Inaba, K., and Steinman, R. M. (1995).Coexpression of NF-kappa B/Rel and Sp1 transcription factors inhuman immunodeficiency virus 1-induced, dendritic cell-T-cell syn-cytia. Proc. Natl. Acad. Sci. USA 92(24), 10944–10948.

reenough, T. C., Sullivan, J. L., and Desrosiers, R. C. (1999). DecliningCD4 T-cell counts in a person infected with nef-deleted HIV-1. NewEngl. J. Med. 340, 236–237.

uy, B., Kieny, M. P., Riviere, Y., Le Peuch, C., Dott, K., Girard, M.,Montagnier, L., and Lecoq, J. P. (1987). HIV F/39orf encodes a phos-phorylated GTP-binding protein resembling an oncogene product.Nature 330, 266–269.

arris, M. (1996). From negative factor to a critical role in virus patho-genesis: The changing fortune of Nef. J. Gen. Virol. 77, 2379–2392.

owe, A. Y., Jung, J. U., and Desrosiers, R. C. (1998). Zeta chain of theT-cell receptor interacts with nef of simian immunodeficiency virusand human immunodeficiency virus type 2. J. Virol. 72(12), 9827–9834.

gnatius, R., Isdell, F., O’Doherty, U., and Pope, M. (1998). Dendritic cellsfrom skin and blood of macaques both promote SIV replication with

T cells from different anatomical sites. J. Med. Prim. 27(2–3), 121–128.

estler, H. W., Ringler, D. J., Mori, K., Panicali, D. L., Sehgal, P. K., Daniel,

K

K

K

L

L

M

M

236 PETIT ET AL.

M. D., and Desrosiers, R. C. (1991). Importance of the nef gene formaintenance of high virus loads and for development of AIDS. Cell65, 651–662.

irchoff, F., Greenough, T. C., Brettler, D. B., Sulllivan, J. L., and Desro-siers, R. C. (1995). Absence of intact nef sequences in a long-termsurvivor with non-progressive HIV-1 infection. N. Engl. J. Med. 332,228–232.

lagge, I. M., and Schneider-Schaulies, S. (1999). Virus interactionswith dendritic cells. J. Gen. Virol. 80, 823–833.

night, S. C., and Patterson, S. (1997). Bone marrow-derived dendriticcells, infection with human immunodeficiency virus, and immunopa-thology. Annu. Rev. Immunol. 15, 593–615.

Lama, J., Mangasarian, A., and Trono, D. (1999). Cell-surface expressionof CD4 reduces HIV-1 infectivity by blocking Env incorporation in aNef- and Vpu-inhibitable manner. Curr. Biol. 9, 622–631.

e Gall, S., Erdtmann, L., Benichou, S., Berlioz-Torrent, C., Liu, L. X.,Benarous, R., Heard, J. M., and Schwartz, O. (1998). Nef interacts withm subunits of clathrin adaptor complexes and reveals a crypticsorting signal in MHC-I molecules. Immunity 8, 483–495.

Le Gall, S., Prevost, M. C., Heard, J. M., and Schwartz, O. (1997). Humanimmunodeficiency virus type I Nef independently affects virion incor-poration of major histocompatibility complex class I molecules andvirus infectivity. Virology 229, 295–301.

in, C. L., Sewell, A. K., Gao, G. F., Whelan, K. T., Phillips, R. E., andAustyn, J. M. (2000). Macrophage-tropic HIV induces and exploitsdendritic cell chemotaxis. J. Exp. Med. 192(4), 587–594.

Lin, C. L., Suri, R. M., Rahdon, R. A., Austyn, J. M., and Roake, J. A. (1998).Dendritic cell chemotaxis and transendothelial migration are in-duced by distinct chemokines and are regulated on maturation. Eur.J. Immunol. 28(12), 4114–4122.

Madani, N., and Kabat, D. (1998). An endogenous inhibitor of humanimmunodeficiency virus in human lymphocytes is overcome by theviral Vif protein. J. Virol. 72(12), 10251–10255.arechal, V., Clavel, F., Heard, J. M., and Schwartz, O. (1998). CytosolicGag p24 as an index of productive entry of human imunodeficiencyvirus type 1. J. Virol. 72, 2208–2212.essmer, D., Ignatius, R., Santisteban, C., Steinman, R. M., and Pope,M. (2000). The decreased replicative capacity of simian immunode-ficiency virus SIVmac239Dnef is manifest in cultures of immaturedendritic cells and T cells. J. Virol. 74, 2406–2413.

Meylan, P. R., Baumgartner, M., Ciuffi, A., Munoz, M., and Sahli, R.(1998). The nef gene controls syncytium formation in primary humanlymphocytes and macrophages infected by HIV type 1. AIDS Res.Hum. Retroviruses 14(17), 1531–1542.

Miller, M. D., Warmerdam, M. T., Gaston, I., Greene, W. C., and Feinberg,M. B. (1994). The human immunodeficiency virus-1 nef gene product:A positive factor for viral infection and replication in primary lympho-cytes and macrophages. J. Exp. Med. 179, 101–113.

O’Doherty, U., Steinman, R. M., Peng, M., Cameron, P. U., Gezelter, S.,Kopeloff, I., Swiggard, W. J., Pope, M., and Bhardwaj, N. (1993).Dendritic cells freshly isolated from human blood express CD4 andmature into typical immunostimulatory dendritic cells after culture inmonocyte-conditioned medium. J. Exp. Med. 178, 1067–1076.

Perelson, A. S., Neumann, A. U., Markowitz, M., Leonard, J. M., and Ho,D. D. (1996). HIV-1 dynamics in vivo: Virion clearance rate, infectedcell life-span, and viral generation time. Science 271(5255), 1582–1586.

Piguet, V., Schwartz, O., Le Gall, S., and Trono, D. (1999). The down-regulation of CD4 and MHC-I by primate lentiviruses: A paradigm forthe modulation of cell surface receptors. Immunol. Rev. 168, 51–63.

Pope, M., Betjes, M. G., Romani, N., Hirmand, H., Cameron, P. U.,Hoffman, L., Gezelter, S., Schuler, G., and Steinman, R. M. (1994).Conjugates of dendritic cells and memory T lymphocytes from skinfacilitate productive infection with HIV-1. Cell 78, 389–398.

Pope, M., Elmore, D., Ho, D., and Marx, P. (1997). Dendrite cell–T cell

mixtures, isolated from the skin and mucosae of macaques, supportthe replication of SIV. AIDS Res. Hum. Retroviruses 13(10), 819–827.

Pope, M., Gezelter, S., Gallo, N., Hoffman, L., and Steinman, R. M.(1995). Low levels of HIV-1 infection in cutaneous dendritic cellspromote extensive viral replication upon binding to memory CD41 Tcells. J. Exp. Med. 182, 2045–2056.

Ross, T. M., Oran, A. E., and Cullen, B. R. (1999). Inhibition of HIV-1progeny virion release by cell-surface CD4 is relieved by expressionof the viral Nef protein. Curr. Biol. 9, 613–621.

Rowland-Jones, S. L. (1999). HIV: The deadly passenger in dendriticcells. Curr. Biol. 9, R248–R250.

Rubbert, A., Combadiere, C., Ostrowski, M., Arthos, J., Dybul, M.,Machado, E., Cohn, M. A., Hoxie, J. A., Murphy, P. M., Fauci, A. S., andWeissman, D. (1998). Dendritic cells express multiple chemokinereceptors used as coreceptors for HIV entry. J. Immunol. 160(8),3933–3941.

Saksela, K. (1997). HIV-1 Nef and host protein kinases. Front. Biosci. 2,606–618.

Schaefer, T. M., Bell, I., Fallert, B. A., and Reinhart, T. A. (2000). The T-cellreceptor zeta chain contains two homologous domains with whichsimian immunodeficiency virus Nef interacts and mediates down-modulation. J. Virol. 74(7), 3273–3283.

Schwartz, O., Marechal, V., Danos, O., and Heard, J. M. (1995). Humanimmunodeficiency virus type 1 Nef increases the efficiency of re-verse transcription in the infected cell. J. Virol. 69, 4053–4059.

Schwartz, O., Marechal, V., Friguet, B., Arenzana-Seisdedos, F., andHeard, J. M. (1998). Antiviral activity of the proteasome on incomingHIV-1. J. Virol. 72, 3845–3850.

Schwartz, O., Marechal, V., Le Gall, S., Lemonnier, F., and Heard, J. M.(1996). Endocytosis of MHC-I molecules is induced by HIV-1 Nef. Nat.Med. 2, 338–342.

Schwartz, O., Riviere, Y., Heard, J. M., and Danos, O. (1993). Reducedcell surface expression of processed HIV-1 envelope glycoprotein inthe presence of Nef. J. Virol. 67, 3274–3280.

Sigal, L. J., Crotty, S., Andino, R., and Rock, K. L. (1999). Cytotoxic T-cellimmunity to virus-infected non-hematopoietic cells requires presen-tation of exogenous antigen. Nature 398, 77–80.

Simon, J. H., Gaddis, N. C., Fouchier, R. A., and Malim, M. H. (1998).Evidence for a newly discovered cellular anti-HIV-1 phenotype. Nat.Med. 4(12), 1397–1400.

Skowronski, J., Parks, D., and Mariani, R. (1993). Altered T cell activationand development in transgenic mice expressing the HIV-1 nef gene.EMBO J. 12, 703–713.

Spina, C. A., Kwoh, T. J., Chowers, M. Y., Guatelli, J. C., and Richman,D. D. (1994). The importance of nef in the induction of humanimmunodeficiency type 1 replication from primary quiescent CD4lymphocytes. J. Exp. Med. 179, 115–123.

Swingler, S., Mann, A., Jacque, J., Brichacek, B., Sasseville, V. G.,Williams, K., Lackner, A. A., Janoff, E. N., Wang, R., Fisher, D., andStevenson, M. (1999). HIV-1 Nef mediates lymphocyte chemotaxisand activation by infected macrophages. Nat. Med. 5(9), 997–1003.

Tsomides, T. J., Aldovini, A., Johnson, R. P., Walker, B. D., Young, R. A.,and Eisen, H. N. (1994). Naturally processed viral peptides recog-nized by cytotoxic T lymphocytes on cells chronically infected byhuman immunodeficiency virus type 1. J. Exp. Med. 180(4), 1283–1293.

Tsunetsugu-Yokota, Y., Yasuda, S., Sugimoto, A., Yagi, T., Azuma, M.,Yagita, H., Akagawa, K., and Takemori, T. (1997). Efficient virus trans-mission from dendritic cells to CD41 T cells in response to antigendepends on close contact through adhesion molecules. Virology239, 259–268.

Wang, J. K., Kiyokawa, E., Verdin, E., and Trono, D. (2000). The Nefprotein of HIV-1 associates with rafts and primes T cells for activa-tion. Proc. Natl. Acad. Sci. USA 97(1), 394–399.

Yewdell, J. W., Norbury, C. C., and Bennink, J. R. (1999). Mechanisms ofexogenous antigen presentation by MHC class I molecules in vitroand in vivo: Implications for generating CD81 T cell responses to

infectious agents, tumors, transplants, and vaccines. Adv. Immunol.73, 1–77.