HIV Nef Increases T Cell ERK MAP Kinase Activity*

The human immunodeficiency regulatory protein Nef enhances viral replication and is central to viral pathogenesis. Although Nef has displayed a capacity to associate with a diverse assortment of cellular molecules and to increase T cell activity, the biochemical activity of Nef in T cells remains poorly defined. In this report we examine the bioactivity of Nef in primary CD4 T cells and, in par-ticular, focus on the biochemical pathways known to be central to T cell activity. The extracellular signal-regulated kinase (ERK) mitogen-activated protein (MAP) kinase pathway was dramatically affected by Nef expression with increases in ERK, MEK, and Elk induction. The capacity of Nef to increase the MAP kinase pathway activity was dependent on T cell receptor stimulation. By increasing ERK MAP kinase activity, Nef is functionally associated with a kinase known to affect T cell activity, viral replication, and viral infectivity. Most of the viral gene products of HIV 1 have established structural or biochemical functions. Nef, which is the predom-inant early transcript (1, 2), is largely defined by cellular and viral phenotypes and by in vivo effects. Nef expression modu-lates cell surface receptors 4), enhances virion infectivity and enhances in vivo viral replication and pathogenesis CD4 and the major antigen I down-modulation most of cellular moieties,

The human immunodeficiency regulatory protein Nef enhances viral replication and is central to viral pathogenesis. Although Nef has displayed a capacity to associate with a diverse assortment of cellular molecules and to increase T cell activity, the biochemical activity of Nef in T cells remains poorly defined. In this report we examine the bioactivity of Nef in primary CD4 T cells and, in particular, focus on the biochemical pathways known to be central to T cell activity. The extracellular signalregulated kinase (ERK) mitogen-activated protein (MAP) kinase pathway was dramatically affected by Nef expression with increases in ERK, MEK, and Elk induction. The capacity of Nef to increase the MAP kinase pathway activity was dependent on T cell receptor stimulation. By increasing ERK MAP kinase activity, Nef is functionally associated with a kinase known to affect T cell activity, viral replication, and viral infectivity.
Most of the viral gene products of HIV 1 have established structural or biochemical functions. Nef, which is the predominant early transcript (1,2), is largely defined by cellular and viral phenotypes and by in vivo effects. Nef expression modulates cell surface receptors (3,4), enhances virion infectivity (5)(6)(7), and enhances in vivo viral replication and pathogenesis (8). The study of CD4 and the major histocompatibility antigen I down-modulation has been the most productive and has resulted in identification of numerous cellular moieties, largely restricted to endocytotic machinery, that can associate with Nef (for reviews, see Refs. 9 and 10). From this work, it has been suggested that Nef serves as an adapter for coupling endocytotic molecules to the targeted membrane receptors.
The study of Nef-mediated effects on activation pathways has been less conclusive. Studies addressing Nef function in T cells define capacities that both inhibit and enhance T cell activity (for review, see Refs. 10 and 11). From work with a CD8-Nef fusion protein, it was proposed that cellular location defined whether Nef expression resulted in negative or positive effects on T cell activity (12). Recent efforts from our laboratory have demonstrated that the opposing effects of Nef on T cell activation are also mediated by different Nef concentrations (13). Nef can increase T cell interleukin-2 synthesis in both T cell lines and human primary CD4 T cells (14 -17). Furthermore, Nef has been shown to increase both T cell nuclear factor of activated T cells (17,18) and NF-B (17) reporter activities. Thus, there is evolving support for Nef as a positive T cell factor.
Attempts to define the biochemical capacity of Nef are numerous. Nef can bind to a number of cellular signaling moieties. For example, Nef binds to and activates the tyrosine kinase Hck (19,20), and the co-expression of Nef and Hck in Rat-2 fibroblasts resulted in cellular transformation, although Nef expression alone had no effect (19). In a macrophage cell line, Nef, through Hck and MAP kinase, induces the transcription factor activating protein-1 (21). Hck is not expressed in T cells, but with regard to T cell kinases, Nef has been shown to bind and inhibit Lck and MAP kinase activity (22). Nef expression has also been demonstrated to alter calcium signaling. When expressed in NIH3T3 cells, Nef inhibited inositol trisphosphate-mediated calcium flux (23), an effect similar to that seen in a Jurkat cell expressing a Nef-CD8 fusion protein (12). However, in Nef-expressing transgenic murine thymocytes, T cell stimulation resulted in elevated calcium responses (24), a finding more consistent with T cell activation enhancement. Nef also binds to an activated serine kinase, p21-activated kinase or Pak (25), and this association occurs in HIVinfected primary T cells (26). In a recent exploration of Nef activity, it was demonstrated that the co-expression of Nef with Vav, through Pak, increased c-Jun N-terminal kinase (JNK) activity in NIH3T3 cells (27). However, it is unclear how relevant these studies are to the activity of Nef in peripheral CD4 T cells, the main target of HIV infection.
The biochemistry of T cell activation is highly complex, but many of the molecular pathways leading to IL-2 expression have been characterized (for reviews, see Refs. 28 and 29). As used here, IL-2 is both a reporter for the metabolic activity of a T cell and the end-product of a highly characterized and dissectible biochemical process of the CD4 T cell. Per se, an increase in IL-2 levels does not significantly contribute to HIV replication (30). However, with a recent demonstration that Nef, as expressed from HIV infection, increased T cell activity (as defined by IL-2 secretion) and viral production (31), an understanding of the biochemical activity of Nef in these contexts appears relevant.
Events mediated by engagement of the T cell receptor and CD28 co-receptor result in activation of the MAP kinases extracellular signal-regulated kinase (ERK), JNK, and p38, in addition to phosphorylation of IB, elevation of cytosolic calcium, and activation of the kinase Akt. The pathways leading to these cytosolic signaling moieties are briefly outlined in Fig.  1. The choice of kinases and signaling molecules is based on the assumption that Nef, as a cytosolic protein, would affect pathways in this cellular compartment. The choice was also to look at late cytosolic events, yielding the greatest opportunity to "capture" Nef effects. We make use of a system that expresses * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 1 The abbreviations used are: HIV-1, human immunodeficiency virus type 1; ERK, extracellular signal-regulated kinase; MAP, mitogen-activated protein; MEK, extracellular signal-regulated kinase kinase 1/2; MKK, MAP kinase kinase; JNK, c-Jun N-terminal kinase; SEK1, stress-activated kinase kinase 1; IL-2, interleukin-2; CCD, charge-coupled device; AM, acetoxymethyl ester.
HIV Nef at concentrations similar to those seen in HIV infection (13) in hopes of identifying an indigenous pathway in primary CD4 T cells that would permit relevant molecular dissection. In this report, we demonstrate that Nef expression in primary CD4 T cells specifically increases activity of the ERK MAP kinase cascade.

EXPERIMENTAL PROCEDURES
T Cell Cultures-The peripheral lymphocyte fraction from healthy donors was obtained by leukapheresis and countercurrent centrifugal elutriation from the Department of Transfusion Medicine at the National Institutes of Health (32) and purified as previously described (16). Cells were grown in complete growth medium (RPMI 1640 supplemented with 10% heat-inactivated fetal calf serum, 25 mM Hepes, 2 g/liter sodium bicarbonate, 1 mM nonessential amino acids, 10 mM sodium pyruvate, 4 l/liter ␤-mercaptoethanol, and 50 g/ml gentamicin, adjusted to pH 7.4). Proliferation of purified CD4 T cells was achieved by the addition of CD3 plus CD28 antibody immobilized on magnetic beads (16).
Transduction and Detection of Nef-Primary CD4 T cells were transduced with the PA-317 retroviral LXSN system (3) expressing either the NL4 -3 Nef or the nonmyristylated NL4 -3 Nef mutant, which was generated by a glycine to alanine switch at residue position 2 (G2A) (33). Following selection in G418, Nef was detected by Western analysis (16). For T cell functions, the G2A cells were found to be similar to the nontransduced cells, as previously noted (16). All transductions were tested and found positive for Nef expression.
Kinase Assays-For cell activation studies, magnetic beads were removed from proliferating cell cultures after gentle pipetting of the cells, followed by re-exposure to a magnetic field. Cells no longer bound to beads were removed. The bead-containing fraction was cycled through this process two or three times. The cells were resuspended at 2 million cells/ml in fresh RPMI with fetal calf serum and then rested overnight. The next day 4 million cells were resuspended in 1 ml of serum-free RPMI 25 mM Hepes, pH 7.4. At time 0, either 10 g of anti-CD3 (clone HIT3A) or anti-CD3 plus anti-CD28 (clone CD28.2) beads (five beads per cell) was added at 37°C with mixing. At defined times, cells were centrifuged at 4°C for 30 s, and the cell pellet was resuspended in 200 l of lithium dodecyl sulfate sample buffer (Novex), heated for 10 min at 70°C, and sonicated for 20 s to shear DNA, and similar aliquots were applied to a 10% NuPage gel (Novex). The electrophoresis gel run was then transferred to nitrocellulose and probed with kinase-, substrate-, or phosphospecific antibodies (Cell Signaling Technology or Santa Cruz Biotechnology, Inc., Santa Cruz, CA). Binding of antibodies was assayed by secondary peroxidase conjugate antibody (Kirkegaard and Perry) and developed with West Dura (Pierce) substrate. Measurements of generated light were achieved on an Alpha Innotech ChemiImager with a cooled CCD camera. For reprobing, blots were treated with Restore stripping solution (Pierce).
For analysis of ERK phosphorylation of recombinant Elk-1, 4 million cells were suspended in 1 ml of serum-free RPMI, 25 mM Hepes, pH 7.4. At time 0, cells were activated by the addition of antibody as described above. At defined times, cells were centrifuged at 4°C for 30 s, and the cell pellet was resuspended into 1 ml of lysis buffer (20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerolphosphate, 1 mM Na 3 VO 4 , 1 g/ml leupeptin, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride). Lysates were sonicated and centrifuged at 14 K for 10 min, and the supernatant was aliquoted for kinase assays or frozen at Ϫ70°C. Total cellular protein in the lysate was determined by micro-BCA protein assay, and then all samples to be compared were made equivalent in protein concentration. Immobilized phospho-ERK kinase antibody (Thr 202 /Tyr 204 ) was used to purify active ERK kinase from the cell lysate, followed by an in vitro kinase assay utilizing recombinant Elk-1 protein (as purchased in assay kit form from New England Biolabs). A Western analysis for phospho-Ser 383 Elk-1 was then performed as above. For this assay, we followed the protocols of the manufacturer.
Cytosolic Calcium-Free cytosolic calcium was determined by the procedure outlined for fluo-3 acetoxymethyl ester (AM) (34), but using the fluo-4 AM plus Pluronic F-127 reagent from Molecular Probes, Inc. (Eugene, OR) with minor modifications. Cell loading of 4 M fluo-4 AM was achieved in Hanks' balanced salt buffer with Hepes 25 mM, pH 7.4, 37°C, 1 h. Loaded and washed cells were aliquoted in a 96-well plate (10 5 cells). Cells were stimulated through the addition of 2 g of anti-CD3 antibody, and calcium levels were determined by measurement of emitted fluorescence at 37°C on a CytoFluor 4000 (Perseptive Biosystems) fluorimeter fitted with a 485-nm excitation filter and a 530-nm emission filter. Calculation of intracellular calcium was achieved following the sequential addition of ionomycin and EGTA, as previously described (35,36), but a value of K d ϭ 345 nM was used for fluo-4. For each experiment, cell number and volume were determined on a Coulter Z2 particle analyzer.

RESULTS
We transduced purified primary CD4 T cells with either wild type NL4 -3 Nef or the nonmyristylated G2A mutant of NL4 -3 Nef retroviral expression vectors. The G2A mutant has previously been demonstrated to lack T cell enhancement capacity (16,17) but serves as a control for the cellular manipulations of transduction and selection. Under the conditions used in this report, the G2A Nef-transduced cell performed identically to a nontransduced cell. Both native and G2A Nef-transduced cells specifically expressed Nef protein (Refs. 13 and 16; data not shown).
Activation of the ERK MAP kinase pathway is essential for IL-2 synthesis (37,38). ERK is also central to the induction of cellular transcription and translation factors and activation of proliferation machinery (39). In T cells, stimulation of the T cell receptor (CD-28 co-stimulus is not necessary) leads to ERK activity (40). T cell activation involves two ERK species (ERK1 and ERK2) of different molecular weights (ϳ44,000 and 42,000, respectively). Activation of this kinase is achieved by MEK1/2 phosphorylation of a pair of threonine and tyrosine residues of ERK (41). This active dual phosphorylated species was quantitated by phosphospecific antibody binding on Western blots developed from electrophoresis gels of cell lysates. Control and Nef-expressing primary CD4 T cells were maintained by stimulation with CD3-CD28 antibodies immobilized on magnetic beads (see "Experimental Procedures"). Prior to activation, the cells were removed from beads and rested overnight. As shown in Fig. 2A, active phosphorylated forms of both ERK1 (p44) and ERK2 (p42) are generated following stimulation with soluble CD3 antibody. The level of activated ERK was dramatically increased in Nef-expressing cells, but not in the nonmyristylated G2A Nef mutant transduced cell. Probing this blot for total ERK protein (activated plus nonactivated) displayed no differences, demonstrating that the Nef-mediated increase in ERK1 and -2 activity was not due to changes in total ERK protein. For comparative purposes, we have plotted the induction of ERK activity with time (Fig. 2B). The inability of the nonmyristylated G2A mutant Nef to mediate the Nef increase in ERK activity is consistent with its lack of effect on T cell activation enhancement (16,17). The ERK activity was increased by Nef in cells from three different donors and, in addition, was evident in cells stimulated by CD3-CD28 beads (data not shown).  (41,(43)(44)(45). Phosphatidylinositol 3-kinase (PI3K) activity is required for induction of the ERK cascade pathway (76,77) as well as for activation of 3-phosphoinositide-dependent kinase (PDK), which phosphorylates protein kinase B/Akt (46,47,78,79). Protein kinase Cactivates the IKK/IB/NF-B cascade through phosphorylation of IKK (80 -82), which phosphorylates IB and promotes NF-B activity (49,50,83,84). IKK can also be phosphorylated by MEKK1 (85,86) and by COT-activated NIK (87)(88)(89). Cytosolic calcium is released from intracellular stores by an inositol trisphosphate-sensitive receptor (90).
T cells also possess two other inducible MAP kinases, JNK and p38. JNK is induced in T cells by CD3 plus CD28 costimulation (42). JNK activation in T cells is achieved by SEK1 (also known as MKK4)-mediated phosphorylation of a proximal threonine and tyrosine pair in JNK (43). Rested primary human CD4 T cells were stimulated with beads containing immobilized anti-CD3 and CD28 antibodies. As shown in Fig. 3A, JNK activation was dependent on stimulation. Unlike the activity of MAP kinase ERK, there is no measurable effect on JNK activity in primary CD4 T cells by HIV Nef. Comparison of total JNK protein is also displayed in Fig. 3A.
A third MAP kinase, p38, is also activated by threoninetyrosine phosphorylation (44,45). Relative to ERK, the phosphorylation of p38 was less intense and required higher exposure. As shown in Fig. 3B, activation of p38 in the absence or presence of Nef was similar.
Akt kinase (also known as protein kinase B) is downstream of phosphoinositide 3-kinase activity and is phosphorylated on Thr 308 by a phosphatidylinositol trisphosphate-dependent Akt/ protein kinase B kinase (46) and on Ser 473 by autophosphorylation (47). Engagement of the surface receptors resulted in the anticipated phosphorylation of Akt, as shown in Fig. 3B, and Nef expression appeared to have no effect.
The NF-B transcription factor is inactive when bound to the inhibitory factor IB (48). The phosphorylation of Ser 32 /Ser 36 (49,50) leads to release and degradation of IB␣, which correlates with NF-B nuclear localization following T cell activation (51,52). To examine the effect of Nef on IB␣ phosphorylation, Nef-expressing and control cells were stimulated with CD3-CD28 beads. Cells were lysed and applied to gel electrophoresis as above, blotted onto cellulose nitrate, and probed by anti-phospho-IB␣ antibody (see Fig. 3C). Stimulation of cells led to rapid phosphorylation of IB␣, and expression of Nef appeared to play no role in enhancing this process.
Stimulation of the T cell receptor also results in a rapid elevation of intracellular calcium, a downstream effect of phospholipase C␥1 activity. Cytosolic calcium was determined by fluorimetric measurement of free calcium binding by fluo-4. Response of control and Nef cells to T cell receptor stimulation resulted in similar cellular calcium elevations as shown in Fig. 3D.
Nef increased the induction of ERK MAP kinase but did not appear to affect other pathways. To further define pathway specificity, we then compared the phosphosphorylation state of MEK1/2 and SEK1 (MKK4), the specific upstream T cell kinases for ERK and JNK, respectively. As with the previous studies, T cells were rested overnight, then activated by antibody engagement of the T cell receptor. Like ERK, MEK1/2 activationassociated phosphorylation was found to be transient, with the highest levels around 5 min (Fig. 4A). Compared with the nontransduced controls and the G2A mutant Nef cells, the Nefexpressing cells displayed an increase in the induction of MEK1/2 activity, as measured by antibody recognition of the dual serine (Ser 217 /Ser 221 ) phosphorylation by Raf. As with the ERK induction, we did not see Nef-mediated activation in the absence of T cell receptor stimulation. An examination of the induction of the corresponding kinase in the JNK pathway, SEK1 (MEKK4), demonstrated a similar dependence on surface receptor stimulation (Fig. 4B). Consistent with the lack of Nef on JNK activation, SEK1 activity was unchanged by Nef expression.
The increased phosphorylation of ERK Thr 202 /Tyr 204 by MEK (Fig. 2) and of MEK Ser 217 /Ser 221 by Raf (Fig. 4) defines the pathway specifically enhanced by Nef expression. These phosphorylation measurements define activities upstream to ERK MAP kinase. Once active, ERK MAP kinase induces numerous downstream transcription factors, including the phosphorylation (Ser 383 ) of Elk-1 (53). We performed Western analysis on the lysate of activated control and Nef-expressing primary CD4 T cells. Increased specific phosphorylation of the Elk-1 Ser 383 was evident in the Nef-expressing cells, compared with the control G2A Nef mutant cell (Fig. 5A) or nontransduced cells (data not shown). The increased phosphorylation of Elk was not due to an increase in Elk protein levels in the Nef-positive cell. In addition to MAP kinase ERK, Elk-1 can be phosphorylated by the MAP kinases JNK and p38 (54 -56), although activation of JNK or p38 in these studies does not appear to be affected by Nef. We then measured, in an in vitro kinase assay, Elk-1 phosphorylation by ERK kinase activity that was purified (separated from JNK and p38) from lysates of activated cells. As shown in Fig. 5B, there was an increased induction of ERK MAP kinase activity in Nef-expressing T cells following T cell stimulation, consistent with increased Elk-1 phosphorylation in Nef-expressing cells (Fig. 5A). DISCUSSION We conclude that the ERK MAP kinase activation cascade is enhanced in Nef-expressing primary CD4 T cells. The lack of effect on other pathways emanating from the T cell surface receptor suggests that this activity is responsible for the enhancement in T cell activity. In addition, the mediated increase in ERK MAP kinase activity also connects Nef to previously characterized HIV infection processes. ERK activity has been demonstrated to increase HIV retroviral long terminal repeat expression (57) and HIV replication (58). Furthermore, viral infectivity has been found to be adversely affected by ERK pathway-specific inhibitors and enhanced by constitutive activation of this pathway (59,60). Vif (viral infectivity factor) increases infectivity of HIV (61, 62), and its phosphorylation by ERK remarkably increases the activity of this viral regulatory protein (63). Thus, the capacity of Nef to increase ERK activity
suggests that a role of this viral protein in enhancing virion infectivity may, in part, be upstream to Vif activity.
Our work with CD4 T cells suggests that Nef expression by itself does not activate the MAP kinase pathway. In our system, there is a requirement for stimulation of the surface T cell receptor. This is consistent with previous demonstrations that the Nef-mediated IL-2 induction in primary human CD4 T cells requires activation (16,17) and that the HIV infection-induced hypersensitization of T cell IL-2 secretion is also dependent on T cell receptor stimulation (31,64).
Our finding appears to differ from a recent demonstration that co-expression of Nef with Vav, through Pak, increases JNK activity in NIH3T3 cells (27). We do not find an increase in JNK activity with Nef in the absence of or following T cell receptor stimulation. Two systematic differences may be responsible. First, the injection of plasmids into NIH3T3 cells would probably result in higher expression levels of Nef and Vav than those seen in the cells examined here. Second, while Pak can induce JNK in numerous cells (65)(66)(67), Pak activity in T cells, following receptor stimulation, has been shown to be essential for ERK activity but not JNK (68). The role of Pak in ERK activity is unresolved, and experimental approaches in-  5-8). The lack of enhancement in JNK induction was seen in cells from three donors. B, induction of Akt and p38 kinase activity is unaffected by Nef. NL4 -3 Nef and control cells were activated as described above and simultaneously probed with either antibody detecting (Active Species) phospho-p38 (Thr 180 /Tyr 182 ) or phospho-Akt (Ser 473 ). Heavy chain of IgG (HC) is present on this blot as well as the blot for total Akt and p38 protein (Total Kinase). C, Nef does not alter induction of IB␣ phosphorylation. Cells were activated and analyzed as above. Detection of IB␣ phosphorylation was achieved with a phospho-IB (Ser 32 )-specific antibody (phospho-IB). The Western analysis for total IB␣ protein is also shown. D, measurement of intracellular calcium following T cell receptor stimulation. Cells were rested overnight, loaded with fluo-4 AM, and then stimulated by anti-CD3 soluble antibody. Fluorescence (485-nm excitation/530-nm emission) was measured in quadruplicate samples every minute. Plotted data represent the mean Ϯ S.D. Data for control cells are depicted as squares; NL4 -3 Nef cells are circles. Open symbols are activated populations, whereas nonstimulated populations are represented by the filled symbols.
FIG. 4. Nef affects MEK1/2 but not SEK1 activity. A, Nef expression increases T cell receptor-induced phosphorylation of MEK1/2. The assay was performed as in Fig. 2, except the blot was probed with antibodies recognizing the dual serine (Ser 217 /Ser 221 ) phosphorylation (Active MEK1/2) or total MEK1/2 protein. B, Nef expression does not alter SEK1 activation. Cells were activated as in Fig. 3A and probed with phospho-SEK1 (Thr 261 ) (Active Sek1) or antisera that recognize total SEK1.

FIG. 5. Nef increases Elk-1 phosphorylation (Ser 383 ).
A, Western analysis of cell lysate of activated control and Nef-expressing primary CD4 T cells, performed as in Fig. 2, comparing the control G2A Nef mutant cell to wild type Nef. G2A control and nontransduced cells were equivalent. The Western for total Elk-1 is also shown. B, Western analysis for Elk-1 phosphorylated by immunoprecipitated active ERK1/2 in an in vitro kinase assay.
volving Nef are complicated by the potential for Nef to physically associate with different Pak isoforms and different Pakassociated nucleotide exchange factors (26, 69 -71).
The ability of Nef to affect ERK but not the JNK or p38 MAP kinase pathways suggests biochemical specificity for Nef molecular targeting, but our present understanding of the biochemical nature of T cell activation suggests broad possibilities. Differentiation of ERK from JNK and p38 MAP kinase activation can occur as early as the CD3-␦ chain (72) and, obviously, as late in the pathway as the kinase species studied here. In fact, the effect of Nef at a distal biochemical site that is responsible for the enhanced ERK response could affect cellular or viral function not associated with the ERK activity. Previous efforts have demonstrated that recombinant Nef can associate with purified ERK MAP kinase (22); however, this interaction inhibited MAP kinase activity. Nef has also been reported to bind Raf (73), but the effect of this association on kinase activity is unknown.
The positive effect of Nef on T cell IL-2 secretion following stimulation has been documented in murine, simian, and human T cell lines, as well as in primary human T cells (14 -17). The use of this system to resolve the biochemical activity of Nef in T cells is validated by the recent demonstration that Nef from HIV infection also increases this T cell response (31). Given that Nef increases both murine and human T cell activity, we believe the conclusion that Nef expression results in increased ERK activity in human T cells is corroborated by the transgenic study of Hanna et al. (74). These authors found that Nef-expressing, CD3-stimulated murine thymocytes display enhanced phosphorylation of several proteins, including the Thr 202 /Tyr 204 phosphorylation of ERK, characteristic of its activation by MEK1 (41,75). Thus, the findings for Nef bioactivity in primary CD4 T cells presented here appear to be reproducible in other cellular systems. Although the actual pathogenesis may indeed be different, this Nef activity is correlated with a state in these mice that is remarkably similar to the disease seen in humans; thus, further studies of the molecular activity of Nef that affects the MAP kinase pathway is warranted. In conclusion, the specificity for ERK MAP kinase activity over other T cell activation pathways offers an opportunity for biochemical resolution of Nef molecular activity.