Synergistic activation of NFAT by HIV-1 nef and the Ras/MAPK pathway.

Nef is a lentiviral protein involved in pathogenesis of AIDS, but its molecular mechanisms of action remain incompletely understood. Here we report a novel effect of Nef on lymphocyte signaling, which is mediated via a T cell receptor (TCR)-independent contribution of Nef to induction of nuclear factor of activated T cells (NFAT), a transcription factor that plays a central role in coordinating T cell activation. Expression of Nef did not significantly alter the basal level of NFAT activity in Jurkat cells nor the increased activity following T cell receptor stimulation by anti-CD3 or anti-CD3 + anti-CD28. We also mimicked NFAT induction by TCR triggering by simultaneous activation of the Ras and calcium signaling pathways with phorbol 12-myristate 13-acetate and ionomycin, respectively. Strikingly, whereas activation of either of these pathways individually did not induce NFAT activity in control cells, in Nef-expressing cells phorbol 12-myristate 13-acetate treatment alone resulted in a 100-fold increase in NFAT-directed gene expression. Experiments with different dominant negative mutant signaling proteins, inhibitory chemicals, and Lck-deficient Jurkat cells revealed that this effect was mediated via activation of calcineurin by Nef-induced changes in calcium metabolism, but was independent of TCR-associated signaling events. This ability of Nef to substitute for triggering of the calcium pathway in induction of NFAT could promote activation of human immunodeficiency virus (HIV)-infected T cells in response to stimuli mediated via TCR or other cell surface receptors under conditions when activation of Ras rather than calcium signaling would otherwise predominate.

persons and in SIV-infected monkeys. Studies in cell culture have revealed a number of functions for Nef, including enhancement of HIV replication kinetics and particle infectivity, down-regulation of cell surface expression of CD4 and major histocompatability complex I, and modulation of intracellular signaling. In agreement with a role in signal transduction, Nef has been reported to bind to a large number of proteins involved in cellular signaling cascades (for a review, see Refs. [1][2][3]. An extreme example of the potential of Nef to alter cellular signaling is provided by the malignant transformation of fibroblasts by the mutant Nef allele of the SIVpbj strain (4). Also, co-expression in rat fibroblasts of native Nef and the Src family tyrosine kinase Hck can lead to malignant transformation of these cells via an increased enzymatic activity of Hck (5). Remarkably, Stevenson and colleagues (6) recently demonstrated that in infected macrophages Nef could cause secretion of chemotactic and mitogenic cytokines, which in turn could render resting T cell susceptible to HIV infection. In addition to acting via such paracrine mechanisms, Nef has also been shown to alter signal transduction of T cells directly. However, most such studies have indicated that the role of Nef is to inhibit T cell receptor (TCR)-induced signaling events, such as induction of the transcription factors NF-B and activating protein-1 (AP-1) and cell surface expression of CD69 (7)(8)(9)(10)(11)(12). On the other hand, some reports have concluded that Nef has neither positive nor negative effects on T cell signaling (13,14).
Nevertheless, because T cell activation positively correlates with HIV replication, it would seem logical that the effects of Nef on T cell activation pathways would be positive, and it could be speculated that the opposite results might have somehow arisen via a paradoxical manifestation of the regulatory potential of Nef. Although this hypothesis is attractive, there has been only limited evidence that Nef could under any conditions contribute to T cell activation. Hyperactivation of T lymphocytes has been observed in Nef transgenic mice (15). The SIVpbj Nef allele has been shown to enable this virus to productively infect resting peripheral blood mononuclear cell cultures and activate T cells in these cultures (4). This unusual property of SIVpbj Nef is because of an immunoreceptor tyrosine-based activation motif created by mutations in its amino terminus (16,17). Forced expression on the cell surface of a Nef chimera containing the extra-and transmembrane parts of CD8 can result in activation of several signaling cascades in Jurkat T-lymphoid cells (18). However, as in the case of SIVpbj Nef, it is not clear how well the effects caused by the CD8-Nef fusion protein reflect the normal function of Nef. Desrosiers and colleagues (19) have demonstrated that wild-type Nef can allow SIV to replicate in an IL-2-dependent T cell clone even upon IL-2 withdrawal, apparently because of Nef-induced secretion of IL-2. Perhaps related to this observation, two groups have recently shown that under certain experimental condi-tions HIV-1 Nef can clearly synergize (rather than interfere) with TCR stimulation as measured by TCR-induced IL-2 secretion or other indicators of T cell activation (20,21).
In this study we have examined the ability of HIV-1 Nef to influence basal and stimulated levels of transcription directed by NFAT, a pivotal transcription factor that coordinates the effects of two important signaling pathways upon activation of T cell gene expression (22)(23)(24). We did not find evidence of positive or negative modulation by Nef of TCR-stimulated NFAT activation. Instead, we report here that Nef can activate calcium signaling in T lymphocytes independently of TCR, and thereby potently synergize with inducers of the Ras pathway to cause a dramatic (ϳ100-fold) increase in NFAT-dependent transcription.
Cells, DNA Transfections, and Luciferase Assays-J.CaM-1 (kindly provided by Tomas Mustelin, The Burnham Institute, La Jolla, CA), Jurkat E-6 (JE-6; from ATCC), A3.01 (a CEM, human T cell line, derivative obtained from National Institutes of Health AIDS Research and Reference Reagent Program), and MT-4 (kindly provided by Anssi Lagerstedt from our Institute) cell lines were maintained in RPMI 1640 (BioWhittaker) medium supplemented with 2 mM glutamine (HyClone) and 10% fetal bovine serum without antibiotics. Cultures were diluted one day prior to transfection into 5-6 ϫ 10 5 cells/ml. Two to three million cells were transfected using Fugene (Roche Molecular Biochemicals) according to the manufacturer's instructions. Twenty hours later cells were either directly stimulated with 50 ng/ml anti-CD3 antibody and/or 1 g/ml anti-CD28 antibody for 6 h or with PMA and/or A23187 for 4 h. When harvested the cells were washed once with 0.5 ml of standard phosphate-buffered saline and lysed in 200 l of 1ϫ cell culture lysis buffer (Promega, Madison, WI) luciferase activity with Promega luciferase reagents and a Bio-Orbit (Turku, Finland) luminometer. For measurement of ␤-galactosidase activity in the same samples, 100 l of lysate was mixed with 10 l of 10xLacZ buffer (500 mM NaCl, 100 mM MgCl 2 , 100 mM ␤-mercaptoethanol) followed by the addition of 100 l of 10 mM 0-nitrophenyl ␤-D-galactopyranoside (Sigma). Reactions were incubated overnight at 37°C, after which their absorbances were measured at 420 nm. The absorbance values of each sample were divided by the mean value of all samples from that transfection experiment. The corresponding luminometer readings were then divided by this ratio to normalize for transfection efficiency.

RESULTS
Triggering of the TCR results in induction of NFAT-dependent transcription via simultaneous activation of calcium/calcineurin and Ras/MAPK pathways (22,24,27). To investigate if expression of HIV-1 Nef could substitute for TCR stimulation in activating these pathways, we transfected Jurkat T lymphoid cells with a reporter plasmid expressing luciferase under the control of three tandem short binding sites for NFAT (-286 to -257 of human IL-2 enhancer; Ref. 28) with or without a Nef expression vector. Twenty hours after transfection NFAT-dependent luciferase activity in these cells was analyzed either without further treatment or following a 4-h stimulation with a calcium/calcineurin pathway-activating ionophore (A23187; 2 M), a Ras/MAPK pathway-activating phorbol ester (PMA, 100 nM), or a combination of these ( Fig. 1).
As expected, stimulation of either pathway alone did not result in significant NFAT activation, but co-stimulation with PMA plus ionophore caused a robust increase (ϳ150-fold) in luciferase activity. A small (ϳ2-fold) but reproducible increase in NFAT-dependent transcription was observed in Nef-transfected cells that were left unstimulated. However, when Neftransfected cells were treated with PMA a 100-fold increase in NFAT-driven transcription was observed. The addition of ionophore to Nef-transfected and PMA-stimulated cells caused only a small (ϳ2-fold) additional increase in NFAT activity. These results suggested that Nef could activate the calcium arm but not the Ras arm of TCR signaling.
We also examined the effect of Nef expression on NFAT activity after anti-CD3-mediated TCR triggering. Nef expression did not enhance or inhibit NFAT activation in cells stimulated by anti-CD3 or anti-CD3 ϩ anti-CD28 co-ligation ( Fig.  1). Similar negative results were observed regardless of whether suboptimal or saturating concentrations of anti-CD3 were used (data not shown). Thus, despite the ability of Nef to synergize with PMA stimulation in inducing NFAT, no potentiation with TCR stimulation was observed, suggesting that the Ras pathway rather than the calcium pathway is limiting in TCR-mediated NFAT induction.
To confirm that our observations are not unique to the NL4 -3/R71 allele of Nef, which is derived from a laboratory-adapted isolate of HIV-1 (25), we also subcloned into the same potent expression vector (pEF-BOS) the Nef gene of the HIV-1 SF2 isolate (29), which has not been passaged extensively before its molecular cloning and whose Nef gene is more similar to sequences obtained directly from patients (30). We found that SF2 Nef was equally if not more potent than NL4 -3/R71 Nef in activating NFAT in PMA-treated Jurkat cells and also did not positively or negatively modulate TCR-induced NFAT activity (data not shown). Thus, the ability to contribute to NFAT activation via the calcium signaling appears to be a general property of different HIV-1 Nef alleles. Although Jurkat cells are a widely used model for studies on T lymphocyte activation, we also examined the effect of Nef in other T lymphocytic human cell lines. Like in Jurkat cells, PMA treatment of A3.01 cells did not induce the expression of the NFAT-luciferase reporter when transfected with the control vector (empty pEF-BOS) but did result in a marked NFAT activation (ϳ10-fold) when co-transfected with Nef (data not shown). Whereas the magnitude of this effect was not as great as that of Nef ϩ PMA in Jurkat cells, it was in accordance with a comparably smaller activation of NFAT-mediated transcription (30 -40-fold) seen in A3.01 cells also in response to PMA ϩ ionophore treatment (data not shown). Likewise, we were also able to consistently observe a small (2-3-fold) induction of NFAT by Nef ϩ PMA in MT-4 cells, even if PMA ϩ ionophore stimulation of these cells resulted in only a modest 6-fold increase in NFAT-dependent luciferase expression (data not shown). Thus, whereas the contribution of Nef to NFAT activation was most pronounced in Jurkat cells, this effect could also be observed in other T cell lines, suggesting that it is mediated by a mechanism that has general biological significance.
To study the molecular mechanisms underlying the synergistic NFAT activation by Nef and PMA, we co-transfected Jurkat cells with dominant negative (DN) mutants of signaling molecules known to act at distinct positions in TCR-initiated signaling cascades (Fig. 2). Expression of either DN-Ras or DN-MEKK1, two potent blockers of Ras signaling leading to the c-Jun NH 2 -terminal kinase and extracellular signal-regulated kinase family of mitogen-activated protein kinases (31) efficiently blocked NFAT activation induced by Nef plus PMA. This result confirmed that the synergistic effect of PMA stimulation with Nef was indeed mediated via the Ras pathway. As expected, by interfering with the Ras arm of TCR-signaling, DN-Ras and DN-MEKK1 also inhibited anti-CD3-stimulated NFAT induction.
Weiss and colleagues (32) have recently shown that PAK1 plays a critical role in TCR signaling by acting at a proximal step downstream of Lck, Cdc42, and Vav but before the calcium and Ras pathways bifurcate. Although we could confirm the ability of DN-PAK1 to block anti-CD3-stimulated NFAT activity, it had only a very minor effect on Nef-induced NFAT activation (Fig. 2). As expected, DN-PAK1 did not inhibit NFAT-directed gene transcription in response to PMA ϩ ionophore treatment (data not shown). This observation suggested that Nef did not mediate its co-stimulatory function by interacting with TCR or associated factors. To substantiate this conclusion, we tested the ability of Nef to activate NFAT in the Jurkat-derived J.CaM1 cells, which are deficient in Lck (Fig. 3) (33). Lck is the major Src kinase of T cells, plays a critical role as the initiation of intracellular TCR signaling (34), and has also been suggested as one of the cellular partners of Nef (10,35). As expected, no NFAT activation was seen in anti-CD3stimulated J.CaM1 cells, whereas a response similar to that seen in Jurkat cell was observed following PMA plus ionophore treatment. Notably, the ability of Nef expression to induce NFAT in these Lck-negative cells was not compromised, providing further proof that this effect is independent of TCR signaling.
To further characterize Nef-induced NFAT activation, we tested a panel of chemicals for their ability to block this effect. To confirm that these agents did not decrease cell viability or the ability to respond to the PMA co-stimulation, we compared in parallel the effect of the same compounds on PMA-induced activation of NF-B-dependent transcription. As shown in Fig.  4 preincubation with cyclosporin A (200 nM), a potent inhibitor of calcineurin (36) efficiently blocked NFAT activation induced by Nef plus PMA but caused only a modest reduction in PMAstimulated NF-B-driven luciferase expression (35% decrease from 1,918,224 to 1,250,188 relative units; data not shown).
Together with the observation that ionophore did not significantly potentiate the effect of Nef (Fig. 1), this finding suggested that Nef acts in the calcium pathway upstream of calcineurin and probably by regulating intracellular calcium concentration. To test if prevention of calcium influx from extracellular sources could block the effect of Nef on NFAT, we supplemented the medium of the transfected Jurkat cells with 500 M calcium chelating agent EGTA. As seen in Fig. 4., this efficiently blocked NFAT activation in Nef-transfected plus PMA-stimulated cells but did not significantly affect PMAinduced NF-B-activated luciferase expression (17% decrease; data not shown).
Phosphatidylinositol 3-kinase has been implicated as an effector molecule in TCR signaling (37) and in regulation of calcium metabolism (38,39). The addition of 200 M wortman- nin, an irreversible inhibitor of the enzymatic activity of phosphatidylinositol 3-kinase, resulted in significant but not complete blocking of anti-CD3-stimulated NFAT activity (Fig. 4). By contrast, wortmannin failed to decrease the effect of Nef on NFAT, indicating that the catalytic activity of phosphatidylinositol 3-kinase was not involved in this process.
Finally, we tested PP1, a tyrosine kinase inhibitor with relative specificity toward the Src family kinases (40). We found PP1 (5 nM) to be an efficient inhibitor of anti-CD3-induced NFAT activation, whereas it only had a minimal effect on NFAT activity induced by Nef plus PMA (Fig. 4). This observation was in good agreement with our results using the Lcknegative J.CaM1 cells and indicated that activation of the calcium/calcineurin pathway by Nef was also independent of other Src kinases, such as Fyn, which has been implicated in calcium regulation via the TCR (41,42). DISCUSSION The present study describes a strong positive effect by wildtype HIV-1 Nef on T cell signaling. This effect was found to be mediated via activation of the calcium/calcineurin pathway and thus was strongly synergistic with the Ras pathway in inducing NFAT-dependent gene expression. Notably, this contribution of Nef to NFAT activation did not require TCR stimulation and was not dependent on the activity of the TCR-associated signaling protein complex.
Although some previous studies have also reported positive effects by Nef on T cell activation, in these cases the action of Nef has been mediated through modulation of TCR signaling. Expression of certain mutant Nef alleles (Nef-pbj (4,17,43) and CD8-Nef (18,44)) has been reported to be sufficient to activate T cells, apparently by mimicking extracellular TCR ligation via action of Nef on the intracellular components of the TCR signaling complex. Although a similar dominant function for wild-type Nef has not been shown, it has recently been clearly demonstrated that native HIV-1 Nef has the potential to cooperate with TCR stimulation in promoting T cell activation (20,21). The mechanism of this enhancement is still un-known but was suggested to involve a lowered threshold for T cell activation caused by Nef-mediated priming of the TCR signaling complex (20,21). It remains unclear, however, why other studies have concluded that the effect of Nef on TCR signaling is negative (such as Ref. 12) or that Nef does not modify the outcome of TCR stimulation at all (such as Ref. 14). Our present data concur best with the latter conclusion, as we saw little or no effect by Nef on NFAT activity induced by anti-CD3 or anti-CD3 ϩ anti-CD28 stimulation. The variables that account for these differential effects by Nef on TCR signaling will need to be clarified by future research and could provide important insights into Nef function.
Whatever the role of Nef in modulating TCR signaling will prove to be, our current data indicate that Nef can also promote T cell activation by a different mechanism that involves a more direct effect on calcium/calcineurin-regulated cellular processes. One reason why this novel function of Nef has not been revealed by earlier studies could be that these may have focused on transcriptional effects induced by Nef alone or addressed a possible co-stimulatory role of Nef only in the context of TCR activation, which in itself results in the activation of calcium signaling. Furthermore, the NFAT-inducing effect of Nef described here requires a relatively high level of Nef expression. We have noted that Nef expression in transiently transfected Jurkat cells was significantly lower if a cytomegalovirus promoter-driven vector (pcDNA3) was used instead of pEF-BOS and that this lower level of Nef expression was insufficient to cause any increase in NFAT activity in PMAtreated Jurkat cells. 2 Thus, the expression levels achieved in earlier studies may not have been as high as the relatively strong expression of Nef used here, which might better recapitulate the abundance of Nef thought to be produced in acutely HIV-infected lymphocytes (45).
The detailed mechanism that connects Nef to cellular calcium metabolism remains to be characterized. Interestingly, two recent studies (46,47) have shown that chronic Nef expression in certain stably transfected cells resulted in the enlargement of their intracellular calcium stores, although no changes in cellular signaling or steady state intracellular free calcium concentrations were observed. It is possible, however, that such covert changes in calcium homeostasis could have been caused by an adaptive response to constitutive Nef expression. Our experiments on direct measurement of calcium levels in transiently Nef-transfected Jurkat cells have so far been complicated by the low numbers of positively transfected cells but do support the idea that acute Nef expression would result in an increased concentration of intracellular calcium. 3 The ability of Nef to cooperate with the Ras pathway in activation of NFAT-dependent gene expression independently of TCR signaling raises the intriguing possibility that HIVinfected T lymphocytes expressing Nef might become abnormally activated in response to extracellular stimuli that would induce Ras but not the calcium pathway. We have observed a modest up-regulation of NFAT in Nef-transfected cells treated with IL-1, as compared with Nef or IL-1 alone (ϳ2-fold), 2 but are looking for soluble or cell-associated molecules that would show more potent synergy with Nef in activating NFAT. On the other hand, it is also possible that lymphocytes under some conditions might react to TCR stimulation with a signaling response that would be dominated by activation of the Ras pathway (unlike the anti-TCR-activated Jurkat cells studied here) and that the independent effect of Nef on calcium signaling could then contribute to complete T cell activation. 2 A. Manninen, unpublished results. 3 A. Manninen, manuscript in preparation.

FIG. 4. Comparison of the sensitivity of NFAT activation by
Nef plus PMA or anti-CD3 stimulation to chemical inhibitors of calcium/calcineurin signaling, phosphatidylinositol 3-kinase, or Src kinases. Jurkat cells were transfected with Nef or with a control vector and stimulated with PMA or anti-CD3, respectively, as described in Fig. 2, except that no DN plasmids were included. Instead, 30 min before PMA or anti-CD3 stimulation was applied the transfected cultures were supplemented with cyclosporin A (200 nM), EGTA (500 M), wortmannin (WM; 200 M) or PP1 (5 nM). As in Fig. 2, the mean luciferase activities from duplicate transfections of a representative experiment are expressed relative (%) to values from similarly transfected cells that were not treated with inhibitors, and normalized to 100% from 11486 (anti-CD3-stimulated control cells) and 19171 (PMAstimulated Nef-transfected cells).
Besides merely promoting T cell activation, Nef-mediated up-regulation of NFAT signaling could also contribute to HIV replication and AIDS pathogenesis by other mechanisms. A recent study by Nolan and colleagues (48) reported that ectopic expression of NFATc in resting CD4ϩ T lymphocytes induced a permissive state, which despite the lack of evidence of T cellactivating effects caused by NFATc overexpression, supported HIV replication in these cells in the absence of further stimulation. The target genes of NFATc that were responsible for this permissive state were not identified but presumably were distinct from those that depend on NFAT complexes formed upon co-stimulation of the calcium/calcineurin and Ras/MAPK pathways during T cell activation.
Besides IL-2, a known member of the latter category of NFAT targets is the gene encoding FasL, the ligand for the proapoptotic receptor CD95/Fas, which has been suggested to be involved in HIV-induced immune destruction and in interfering with elimination of HIV-infected cells by cytotoxic T cells (reviewed in Ref. 49). Interestingly, up-regulation of cell surface FasL expression has also been correlated with Nef expression both in SIV and HIV infection as well as in transfection studies (44, 50 -52). Although the regulation of FasL is complex, it is conceivable that Nef-induced NFAT activity may contribute to this phenomenon. Finally, in addition to cellular genes involved in T cell activation, NFAT has also been shown to activate HIV-1 long terminal repeat-directed transcription by interacting with an unusual binding site that overlaps with the NF-B-responsive element (53).
Identification of the relevant target genes regulated by Nef via NFAT that are involved in Nef-induced enhanced HIV replication and pathogenicity in vivo, as well as a better understanding of the molecular mechanisms by which Nef induces the calcium/calcineurin pathway provide important challenges for future studies. Elucidation of these outstanding issues should help to clarify the enigmatic role of Nef in T cell physiology and could provide a novel means for interfering with the pathogenic function of Nef in AIDS.