We show that the pertussis toxin B oligomer (PTX-B), and the PTX mutant PT9K/129G, which is safely administered in vivo, inhibit both transcription and secretion of TGF-β elicited by HIV-1 Tat in NK cells. Tat-induced TGF-β mRNA synthesis is also blocked by the ERK1 inhibitor PD98059, suggesting that ERK1 is needed for TGF-β production. Moreover, Tat strongly activates the c-Jun component of the multimolecular complex AP-1, whereas TGF-β triggers c-Fos and c-Jun. Of note, treatment of NK cells with PTX-B or PT9K/129G inhibits Tat- and TGF-β-induced activation of AP-1. TGF-β enhances starvation-induced NK cell apoptosis, significantly reduces transcription of the antiapoptotic protein Bcl-2, and inhibits Akt phosphorylation induced by oligomerization of the triggering NK cell receptor NKG2D. All these TGF-β-mediated effects are prevented by PTX-B or PT9K/129G through a PI3K-dependent mechanism, as demonstrated by use of the specific PI3K inhibitor, LY294002. Finally, PTX-B and PT9K/129G up-regulate Bcl-xL, the isoform of Bcl-x that protects cells from starvation-induced apoptosis. It is of note that in NK cells from patients with early HIV-1 infection, mRNA expression of Bcl-2 and Bcl-xL was consistently lower than that in healthy donors; interestingly, TGF-β and Tat were detected in the sera of these patients. Together, these data suggest that Tat-induced TGF-β production and the consequent NK cell failure, possibly occurring during early HIV-1 infection, may be regulated by PTX-B and PT9K/129G.

Infection with HIV has been shown to induce the production of several cytokines both in vitro and in vivo; in turn, cytokines modulate the levels of HIV-1 expression in infected cells (1, 2, 3). This double-edged mechanism is supposed to play an important role in the pathogenesis of AIDS and AIDS-related syndromes (4, 5). HIV-1 products, among which is endogenous HIV-1 Tat, are known to induce the transcription of different cytokines, including TGF-β (2); moreover, exogenous Tat can be taken up and translocate to the nucleus, where it can function as a transactivating agent acting on both viral products and cytokines (5, 6, 7). Regulation of TGF-β gene transcription by Tat has been claimed to contribute to immunosuppression in AIDS (8, 9). Along this line, it has been shown that TGF-β inhibits the cytotoxic function of the effectors of innate immunity involved in the antiviral response represented by NK cells (10, 11). In turn, we and others (12, 13, 14, 15, 16, 17, 18) have reported on the possible immunosuppressive effects of extracellular HIV-1 Tat, both in vitro and in vivo, once taken up by bystander cells, including NK cells.

HIV-1 Tat promotes activation of the multimolecular complex AP-1 (19, 20, 21, 22, 23, 24), which mediates the transactivating effects of Tat. Interestingly, the same pathway is activated by TGF-β, which acts mainly by up-regulating the Jun family of AP-1 transcription factors (25, 26, 27). In addition, TGF-β has been reported to induce apoptosis of different cell types through AP-1-dependent activation of SHIP, which, in turn, determines dephosphorylation of Akt, a PI3K-dependent enzyme deeply involved in protection from apoptosis (20, 28, 29, 30). Indeed, phosphorylated Akt (pAkt) 3 induces the transcription of antiapoptotic proteins such as Bcl-2 and Bcl-x, whereas the dephosphorylated form of Akt is inactive, leading to down-regulation of Bcl-2 and Bcl-x transcription (31, 32). Of note, suppression of TGF-β-induced apoptosis can be reached by up-regulating the PI3K/Akt pathway (33). In contrast, pAkt down-regulates AP-1 activation (29, 30), thus providing cells with a mechanism controlling both Tat- and TGF-β-mediated effects.

Interestingly, it has been shown that the pertussis toxin B oligomer (PTX-B) displays several anti-HIV-1 mechanisms: it inhibits the entry of R5 HIV-1 in human macrophages, it interferes with postentry events in the HIV-1 life cycle and counteracts the transactivating effects of Tat, and as a consequence, it can inhibit HIV-1 replication (34, 35, 36). In this regard, the anti-HIV effect of PTX has been linked to its ability to activate PI3K (34, 37). Like PTX-B, the PTX mutant PT9K/129G, approved for human use as a component of a vaccine against Bordetella pertussis infection (38, 39, 40), have signal transduction properties (41, 42).

In this study we show that 1) HIV-1 Tat induces the transcription and secretion of TGF-β in NK cells; 2) TGF-β accelerates starvation-induced NK cell apoptosis, and this is apparently dependent on the inhibition of Akt phosphorylation and down-regulation of Bcl-xL and Bcl-2 transcription; 3) all these effects are suppressed by PTX-B or the genetically modified mutant of PTX, PT9K/129G; 4) in NK cells from patients with early HIV-1 infection, Bcl-xL and Bcl-2 mRNA expression was consistently lower than that in healthy donors; and 5) TGF-β and Tat were detected in the sera of these patients.

The anti-NKG2D mAb (IgG1) and the neutralizing anti-TGF-β MAB240 (IgG) were obtained from R&D Systems Europe. Recombinant IL-2 (100U/ml) and TGF-β (0.1–10 ng/ml) were purchased from PeproTech EC and BioSource Europe, respectively. Cells were cultured in RPMI 1640 medium supplemented with 10% FCS and with glutamine and penicillin-streptomycin (Biochrom). PHA, the PI3K inhibitor LY294002 (10 μM), goat anti-mouse Ig (GAM), murine Ig, propidium iodide (PI), and FITC-annexin V were obtained from Sigma-Aldrich. The ERK1/2 inhibitor PD98059 (10 μM) and the p38 inhibitor SB203580 (10 μM) were obtained from Calbiochem-Novabiochem. Synthetic Tat, used at a 50-nM concentration (>99% pure, as determined by chromatography), and anti-Tat rabbit antiserum were purchased from Tecnogen. Biotinylated rabbit anti-Tat IgG was obtained from DBA. PTX-B was purchased from Calbiochem, whereas the modified holotoxin PTX, PT-9K/129G, was a gift from Dr. R. Rappuoli (Chiron Spa, Siena, Italy); both were used at 1 nM (34, 38).

Peripheral blood was drawn from 20 HIV-1-infected patients and from 15 healthy donors, matched for sex and age, provided informed consent, at the Infectious Diseases Department (University of Genoa). Patients were in stage A of the disease, defined according to the Center for Disease Control and Prevention criteria, free of concurrent infections, and naive to antiretroviral treatment. HIV-1 RNA was quantitated using the commercial branched DNA (bDNA Ultrasensitive Assay; Chiron) with a lower limit of detection of 50 RNA copies/ml. PBMC from the HIV-1-infected patients and from healthy donors were isolated by Ficoll-Hypaque gradient. Highly purified NK cells were obtained from whole heparinized blood by negative immunodepletion with RosetteSep for NK cell enrichment (StemCell Technologies); the resulting cell population was 95–98% CD16+CD56+CD3. To obtain NK cell lines, purified NK cells were seeded under limiting dilution conditions in 96-well microplates (Greiner Labortecknic) and cultured in RPMI 1640 medium supplemented with 200 mM l-glutamine, 10% FCS, 1 μg/ml PHA, 100 U/ml rIL-2, and 105 irradiated feeder cells (3000 rad) as previously described (13).

NK cells, untreated or pretreated with 1 nM PTX-B or PT9K/129G, were exposed to TGF-β (0.1–10 ng/ml), in the absence of rIL2, and recovered at different time points (24, 48, 72, and 96 h). In some experiments the supernatant derived from NK cells treated with Tat for 24 h and containing TGF-β (10 ng/ml) was added to autologous NK cells and seeded in 96-well microwell plates, and apoptosis was analyzed at 24 and 48 h. To assess apoptosis, cells were stained with 50 μg/ml PI, containing 100 U/ml RNase type A, 10 mM EDTA, and 0.0015% Nonidet-P40 as previously described (43). In other experiments cells were stained with FITC-conjugated annexin V to show exposure of phospholipids at the external side of the plasma membrane, which occurs in early apoptosis. Cells were analyzed on a FACSort (BD Biosciences) equipped with laser excitation at 488 and 610 nm. A longpass filter was used for the PI fluorescence detection. Calibration was assessed with CALIBrite particles using the CellQuest computer program (BD Biosciences). At least 104 cells/sample were analyzed, and results were plotted as a percentage of annexin V+ cells or as a percentage of apoptotic cells (DNA content <2n diploid assessed by PI staining). As a control, fully apoptotic cells were obtained by incubation with the anti-Fas mAb CH-11 (IgM; MBL; 1 μg/ml) for 24 h (data not shown) (43).

Total RNA was isolated from cell pellets using the RNAzol B method (Biotecx Laboratories). cDNA (corresponding to 2 μg of RNA) was synthesized from oligo(dT)-primed RNA in 20 μl of reverse transcriptase buffer and 200 U of Moloney murine leukemia virus reverse transcriptase (PerkinElmer-Cetus) incubated at 42°C for 45 min and at 52°C for 45 min. TGF-β, β-actin, GAPDH, Bcl-xL, and Bcl-2 were amplified with 2.5 U of AmpliTaq Gold Polymerase (PerkinElmer-Cetus) in 2 mM dNTP and 50 μM 5′ and 3′ oligonucleotide primers containing 2.5, 5, and 1 mM MgCl2, respectively. The sequences (5′ to 3′) of the specific primers used are as follows: GAPDH: upstream, 5′-gCC AAA Agg gTC ATC ATC TC; downstream, 5′-ggC CAT CCA Cag TCT TCT (225 bp); Bcl-xL: upstream, 5′-TTg gAc AAT ggA CTg gTT gA; downstream, 5′-gTA gAg Tgg Atg gTC AgT g (780 bp); and Bcl-2: upstream, 5′-CgC CTT CgC CgA gAT gTC Cag Cca g; downstream, 5′-ACT TgT ggC CCA gAT Agg CAC CCA g (385 bp). Amplification was performed in a DNA thermal cycler (Eppendorf Italia) as follows: a first cycle of 94°C for 10 min, 65°C for 45 s, and 72°C for 45 s, followed by 29 cycles of 95°C for 45 s, 65°C for 45 s, and 72°C for 1 min. An extension of all products was performed for 7 min at 72°C at the end of the last cycle, then products were held at 4°C. The sequences (5′ to 3′) of the β-actin- and TGF-β1-specific primers used are as follows: β-actin, 5′-CAT ACT CCT gCT TgC TgA TCC and 5′-ACT CCA TCA TgA AgT gTg ACg (228-bp fragment); and TGF-β1, 5′-Gcc CTg gAC ACC AAC TAT TgC and 5′-gCT gCA CTT gCA ggA gCg CAC (336-bp fragment). Amplification conditions were as follows: 94°C for 1 min, 60°C for 1 min, 72°C for 90 s (35 cycles), and 72°C for 10 min. PCR products were size-fractionated by agarose gel electrophoresis, stained with ethidium bromide, and normalized according to the amount of β-actin or GAPDH detected in the same mRNA sample (43). Briefly, images were acquired by Chemi 550 (Alpha Innotech) and analyzed by Gel Pro Analyzer 3.1 (Media Cybernetics). Results of the densitometric analysis, measured as pixel number × 10−3 arbitrary units, were expressed as a percentage of β-actin for TGF-β or as a percentage of GAPDH for Bcl-xL, Bcl-xS, and Bcl-2.

Phosphorylation of the serine/threonine kinase Akt1/PKBα (pAkt) in cell lysates of NK cells was assessed with the ELISA kit (BioSource Europe) upon ligation of NKG2D, obtained with the specific mAb (R&D Systems; 5 μg/ml) followed by GAM, in the presence or the absence of the PI3K inhibitor LY294002 (10 μM) or of the ERK1/2 inhibitor PD98059 (10 μM). As a control, cells were exposed to an unrelated isotype-matched mAb (BD Biosciences) plus GAM. The same samples were also analyzed for the content of total Akt with a specific ELISA kit (BioSource Europe). Results were expressed as the percentage of pAkt, normalized for total Akt (units per 106 cells). AP-1 activation (c-Jun, c-Fos, and JunD) was determined in nuclear extracts from NK cells with the TransAM AP-1 Family Transcription Factor Assay Kit (Active Motif Europe). This is an ELISA-based kit with a 96-well plate to which oligonucleotide containing a 12-O-tetradecanoylphorbol-13-acetate-responsive element was immobilized. AP-1 dimers in the nuclear extracts bound specifically to this oligonucleotide and were detected with specific Abs directed against c-Fos, c-Jun, or Jun-D. After addition of a secondary Ab conjugated to HRP, samples were quantified by spectrophotometry, and results were expressed as OD at 450 nm. Nuclear extracts from HeLa cells, untreated or serum-stimulated after serum starvation, provided by the commercial kit were used as positive and negative controls (not shown) (43).

TGF-β was determined in the serum of HIV-1 patients by a commercial ELISA Kit (BioSource Europe); HIV-1 Tat was measured using a rabbit anti-Tat antiserum (10 μg/ml), directed to synthetic Tat, as capture Ab and a biotinylated rabbit anti-Tat antiserum (1 μg/ml), raised against recombinant Tat, as detection Ab, followed by avidin-peroxidase (Av-HRP; 1/2000; Sigma-Aldrich) and by the specific substrate (2,2′-azino-bis(3-ethylbenzthyazoline-6-sulfonic acid; Sigma-Aldrich) (44). Plates were read at OD 450 nm, and results were expressed as nanograms per milliliter referred to TGF-β standards of the commercial kit or as nanomolar concentrations referred to synthetic Tat (44). The specificity of the antisera was checked by dot blot (not shown). TGF-β and Tat contents in sera from non-HIV viral-infected patients or healthy donors were 10 ng/ml and <0.5 nM, respectively.

Data are presented as the mean ± SD. Statistical analysis was performed using ANOVA, calculating the F ratio and, when corrected, applying the Bonferroni-Dunnett test with α = 0.01; regression analysis was used for the data in Table I.

Table I.

Increase in TGF-β and detection of HIV-1 Tat in the serum of patients with high HIV-1 mRNA

Patient No.CD4+ Cells (%)aCD4/CD8 RatioCD3CD16+ Cells (%)aHIV-1 RNA (copies/ml)bTGF-β (ng/ml)cTat (nM)d
01 10 0.2 24000 95 40 
02 22 0.5 17000 90 50 
03 36 0.4 9000 100 50 
04 41 1.0 8100 85 10 
05 40 1.0 6800 80 20 
06 26 0.5 6000 100 40 
07 0.1 10 5000 105 30 
08 23 0.8 2700 100 50 
09 31 0.9 900 95 40 
10 28 1.0 12 900 50 20 
11 36 1.0 11 900 35 10 
12 30 1.2 10 800 55 10 
13 20 1.0 12 <80 100 20 
14 38 1.3 18 <80 28 nd 
15 27 1.7 14 <80 20 nd 
16 57 2.9 18 <80 15 nd 
17 40 1.2 12 <80 80 10 
18 44 1.1 15 <80 20 nd 
19 46 1.5 13 <80 15 nd 
20 42 1.3 16 <80 30 nd 
Patient No.CD4+ Cells (%)aCD4/CD8 RatioCD3CD16+ Cells (%)aHIV-1 RNA (copies/ml)bTGF-β (ng/ml)cTat (nM)d
01 10 0.2 24000 95 40 
02 22 0.5 17000 90 50 
03 36 0.4 9000 100 50 
04 41 1.0 8100 85 10 
05 40 1.0 6800 80 20 
06 26 0.5 6000 100 40 
07 0.1 10 5000 105 30 
08 23 0.8 2700 100 50 
09 31 0.9 900 95 40 
10 28 1.0 12 900 50 20 
11 36 1.0 11 900 35 10 
12 30 1.2 10 800 55 10 
13 20 1.0 12 <80 100 20 
14 38 1.3 18 <80 28 nd 
15 27 1.7 14 <80 20 nd 
16 57 2.9 18 <80 15 nd 
17 40 1.2 12 <80 80 10 
18 44 1.1 15 <80 20 nd 
19 46 1.5 13 <80 15 nd 
20 42 1.3 16 <80 30 nd 
a

CD4+ or CD8+ and CD3CD16+ cells were evaluated by immunofluorescence with the specific mAbs and FACS analysis. The percentage of CD3CD16+ cells in the peripheral blood of 15 healthy donors tested for comparison was 15 ± 3.

b

HIV-1 RNA was quantitated using the commercial branched DNA (bDNA Ultrasensitive Assay; Chiron) with a lower limit of detection of 50 RNA copies/ml.

c

TGF-β was measured in the sera of HIV-1-infected patients by an ELISA commercial kit. Results, expressed as nanograms per milliliter, refer to the standard. TGF-β in the sera of 15 healthy donors tested for comparison was 8 ± 2 ng/ml.

d

Tat was detected using a polyclonal rabbit anti-Tat antiserum (10 μg/ml) as capture Ab and a biotinylated rabbit anti-Tat antiserum (1 μg/ml) as detection Ab, followed by Av-HRP (1/2000) and the specific substrate. Results, expressed as nanomolar concentrations refer to synthetic Tat (Tecnogen) used as standard. nd, not detectable. Tat measured in the sera of 15 healthy donors tested for comparison was <1 nM.

Endogenous Tat is known to induce the transcription of several cytokines, whereas the exogenous protein can be taken up and translocated to the nucleus, where it can function as a transactivating agent (5, 6, 7). In this study we show that synthetic Tat induced very early transcription of TGF-β in NK cells, reaching a peak between 3 and 6 h (Fig. 1,A). This induction resulted in a 10-fold increase in TGF-β mRNA, as evaluated by densitometric analysis, expressed as a percentage of β-actin (Fig. 1,B shows the mean ± SD of six independent experiments, including that depicted in Fig. 1,A). Moreover, the cytokine was released in the culture medium by 12 h, with a maximum of 25 ng/ml by 24 h (Fig. 1,C). The time points corresponding to the peak of transcription (6 h) and secretion (24 h) were chosen for additional studies. Because we and others have reported that PTX can counteract some effects of extracellular Tat (45, 46), we tested the hypothesis that it could also prevent Tat-induced TGF-β production. As shown in Fig. 1, both transcription (Fig. 1,D, lane 3 vs lane 2; Fig. 1,E shows the densitometric analysis and is the mean ± SD of six independent experiments, including the experiment shown in Fig. 1,D) and secretion (Fig. 1,F) of TGF-β were strongly inhibited (65 and 75% inhibitions, respectively) by a short exposure (10 min) of NK cells to the PTX-B oligomer (1 nM). Interestingly, the genetically modified mutant of PTX, PT9K/129G, was able to inhibit the effects of Tat on TGF-β synthesis (Fig. 1,D, lane 4 vs lane 2; Fig. 1,E for densitometric analysis; 67% inhibition) and secretion (Fig. 1 F; 75% inhibition).

FIGURE 1.

PTX-B and PT9K/129G inhibit transcription and secretion of TGF-β induced by HIV-1 Tat in NK cells. A, Kinetics of TGF-β mRNA transcription, evaluated by PCR and compared with β-actin, after treatment of NK cells with Tat (50 nM). B, Densitometric analysis, expressed as the percentage of β-actin in the same sample, of the TGF-β mRNA transcription kinetics as described in A; values are the mean ± SD from six independent experiments. ∗, p < 0.01. C, Kinetics of TGF-β secretion (nanograms per milliliter per 106 cells), determined by ELISA, after exposure of NK cells to Tat (50 nM); values are the mean ± SD from six independent experiments. ∗, p < 0.01. D, TGF-β mRNA evaluated by PCR in untreated NK cells (lane 1) or 3 h (lane 2) after treatment with Tat (50 nM) compared with β-actin. Lanes 3 and 4, Pretreatment of NK cells with 1 nM PTX-B or 1 nM PT9K/129G, respectively, for 10 min before exposure to Tat for 3 h. Lanes 5 and 6, PTX-B or PT9K/129G, respectively, for 10 min on untreated NK cells. E, Densitometric analysis (TGF-β percentage of β-actin) of the experiment depicted in D; values are the mean ± SD from six independent experiments. F, TGF-β secretion (nanograms per milliliter per 106 cells) after exposure of NK cells to Tat (50 nM), without or with PTX-B or PT9K/129G (1 nM); values are the mean ± SD from six independent experiments. MW, m.w.; None, untreated NK cells. ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PTX-B or Tat plus PT9K/129G vs Tat.

FIGURE 1.

PTX-B and PT9K/129G inhibit transcription and secretion of TGF-β induced by HIV-1 Tat in NK cells. A, Kinetics of TGF-β mRNA transcription, evaluated by PCR and compared with β-actin, after treatment of NK cells with Tat (50 nM). B, Densitometric analysis, expressed as the percentage of β-actin in the same sample, of the TGF-β mRNA transcription kinetics as described in A; values are the mean ± SD from six independent experiments. ∗, p < 0.01. C, Kinetics of TGF-β secretion (nanograms per milliliter per 106 cells), determined by ELISA, after exposure of NK cells to Tat (50 nM); values are the mean ± SD from six independent experiments. ∗, p < 0.01. D, TGF-β mRNA evaluated by PCR in untreated NK cells (lane 1) or 3 h (lane 2) after treatment with Tat (50 nM) compared with β-actin. Lanes 3 and 4, Pretreatment of NK cells with 1 nM PTX-B or 1 nM PT9K/129G, respectively, for 10 min before exposure to Tat for 3 h. Lanes 5 and 6, PTX-B or PT9K/129G, respectively, for 10 min on untreated NK cells. E, Densitometric analysis (TGF-β percentage of β-actin) of the experiment depicted in D; values are the mean ± SD from six independent experiments. F, TGF-β secretion (nanograms per milliliter per 106 cells) after exposure of NK cells to Tat (50 nM), without or with PTX-B or PT9K/129G (1 nM); values are the mean ± SD from six independent experiments. MW, m.w.; None, untreated NK cells. ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PTX-B or Tat plus PT9K/129G vs Tat.

Close modal

HIV-1 Tat is known to activate ERK1/2 (19) and, as a consequence, to promote activation of the multimolecular complex, AP-1 (21, 22, 23, 24), which, in turn, induces TGF-β production (25). Accordingly, we found that the ERK1 inhibitor PD98059 (10 μM) strongly reduced (by 70%) Tat-induced TGF-β mRNA synthesis (Fig. 2,A, lane 4 vs lane 2; densitometric analysis in Fig. 2,B is the mean ± SD of six independent experiments, including that depicted in Fig. 2,A) and secretion (Fig. 2,C) in NK cells; conversely, the p38 inhibitor SB203580 (10 μM) had no effect (Fig. 2,A, lane 3 vs lane 2; see also Fig. 2, B and C).

FIGURE 2.

Involvement of ERK1 in TGF-β production induced by HIV-1 Tat in NK cells. A, TGF-β mRNA compared with β-actin in NK cells untreated (lane 1) or treated for 3 h with 50 nM Tat, without (lane 2) or with the p38 inhibitor SB203580 (10 μM; lane 3) or the ERK1 inhibitor PD98059 (10 μM; lane 4). MW, m.w. B, Densitometric analysis of the experiment depicted in A and three more experiments (TGF-β percentage of β-actin); values are the mean ± SD from four experiments. None, untreated NK cells. C, TGF-β secretion (nanograms per milliliter per 106 cells) after exposure of NK cells to Tat (50 nM), without or with SB203580 (10 μM) or PD98059 (10 μM); values are the mean ± SD from six independent experiments. ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PD98059 vs Tat.

FIGURE 2.

Involvement of ERK1 in TGF-β production induced by HIV-1 Tat in NK cells. A, TGF-β mRNA compared with β-actin in NK cells untreated (lane 1) or treated for 3 h with 50 nM Tat, without (lane 2) or with the p38 inhibitor SB203580 (10 μM; lane 3) or the ERK1 inhibitor PD98059 (10 μM; lane 4). MW, m.w. B, Densitometric analysis of the experiment depicted in A and three more experiments (TGF-β percentage of β-actin); values are the mean ± SD from four experiments. None, untreated NK cells. C, TGF-β secretion (nanograms per milliliter per 106 cells) after exposure of NK cells to Tat (50 nM), without or with SB203580 (10 μM) or PD98059 (10 μM); values are the mean ± SD from six independent experiments. ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PD98059 vs Tat.

Close modal

We also observed that Tat induces the activation of AP-1, preferentially involving c-Jun in NK cells (∼30-fold increase in six independent experiments; Fig. 3,A). Of note, this activation was inhibited (Fig. 3,A; mean ± SD of six independent experiments) by either PTX-B (66% inhibition) or PT9K/129G (73% inhibition). In control experiments, we found that PTX-B and PT9K/129G do not affect activation of AP-1 by themselves (not shown). Because TGF-β also induces AP-1 (28, 29), thus possibly creating a positive loop on its synthesis and secretion, we investigated the effects of PTX-B and PT9K/129G on c-Jun, c-Fos, and JunD activation upon TGF-β treatment of NK cells. We found that exposure of NK cells to 10 ng/ml TGF-β led to an 18-fold activation of c-Fos and, to a lesser extent, of c-Jun and JunD (9- and 6-fold increases; Fig. 3,A); again, pretreatment of NK cells with either PTX-B or PT9K/129G (1 nM) abolished c-Fos and JunD activation and strongly inhibited c-Jun activation (74% inhibition with PTX-B and 56% inhibition with PT9K/129G; Fig. 3,B; mean ± SD of six independent experiments). Interestingly, the blocking effect of PTX-B or PT-9K/129G on Tat-induced (Fig. 3,C) or TGF-β-induced (Fig. 3 D) c-Jun activation was abolished by preincubation of NK cells with LY294002, but not with PD98059, suggesting that both PTX-B and its nontoxin mutant act mainly through PI3K activation.

FIGURE 3.

PTX-B or PT9K/129G inhibit AP-1 activation induced by HIV-1 Tat or by TGF-β. AP-1 activation (c-Jun, c-Fos, and JunD) was determined with the TransAM AP-1 Family Transcription Factor Assay Kit in nuclear extracts from NK cells untreated (none) or exposed to 50 nM Tat (A and C) or 10 ng/ml TGF-β (B and D), with or without PTX-B or PT9K/129G (1 nM). In some experiments PTX-B was added to NK cells in the presence of 10 μM LY294002 or 10 μM PD98059 (C and D). Results, expressed as OD at 450 nm, are the mean ± SD of four independent experiments. Nuclear extracts from HeLa cells, untreated or serum-stimulated after serum starvation, provided by the kit and used as positive and negative controls gave OD450 values of 1.8 and 0.2, respectively, for c-Fos, 1.5 and 0.1 for c-Jun, and 1.0 and 0.1 for JunD (not shown in the figure). A: ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PTX-B or Tat plus PT9K/129G vs Tat. B: ∗∗, p < 0.01, TGF-β vs none; ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β alone. C: ∗∗, p < 0.01, Tat plus PTX-B vs Tat; ∗, p < 0.01, Tat plus PTX-B plus LY294002 vs Tat plus PTX-B. D: ∗∗, p < 0.01, TGF-β plus PTX-B vs TGF-β; ∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

FIGURE 3.

PTX-B or PT9K/129G inhibit AP-1 activation induced by HIV-1 Tat or by TGF-β. AP-1 activation (c-Jun, c-Fos, and JunD) was determined with the TransAM AP-1 Family Transcription Factor Assay Kit in nuclear extracts from NK cells untreated (none) or exposed to 50 nM Tat (A and C) or 10 ng/ml TGF-β (B and D), with or without PTX-B or PT9K/129G (1 nM). In some experiments PTX-B was added to NK cells in the presence of 10 μM LY294002 or 10 μM PD98059 (C and D). Results, expressed as OD at 450 nm, are the mean ± SD of four independent experiments. Nuclear extracts from HeLa cells, untreated or serum-stimulated after serum starvation, provided by the kit and used as positive and negative controls gave OD450 values of 1.8 and 0.2, respectively, for c-Fos, 1.5 and 0.1 for c-Jun, and 1.0 and 0.1 for JunD (not shown in the figure). A: ∗∗, p < 0.01, Tat vs none; ∗, p < 0.01, Tat plus PTX-B or Tat plus PT9K/129G vs Tat. B: ∗∗, p < 0.01, TGF-β vs none; ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β alone. C: ∗∗, p < 0.01, Tat plus PTX-B vs Tat; ∗, p < 0.01, Tat plus PTX-B plus LY294002 vs Tat plus PTX-B. D: ∗∗, p < 0.01, TGF-β plus PTX-B vs TGF-β; ∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

Close modal

It is known that TGF-β is a mediator of apoptosis in different cell types (28, 29, 30); we found that it is able to accelerate starvation-induced apoptosis in NK cells. Indeed, annexin V staining, which identifies phospholipid exposure at the external side of the membrane, i.e., the earliest event of apoptosis, showed that ∼15% of NK cells cultured in the absence of growth factors (IL-2) were annexin V+ after 48 h, 25% were annexin V+ after 72 h, and 40% were annexin V+ at 96 h (Fig. 4,A); in the presence of TGF-β, the same effect was evident earlier, with ∼40% of cells undergoing apoptosis at 48 h (Fig. 4,A). Titration of TGF-β showed that the effect of the cytokine on starvation-induced NK cell apoptosis was maximal at 10 ng/ml, but was still detectable at 1 ng/ml (Fig. 4,A). Of note, this effect of TGF-β was blocked by PTX-B or the mutant PT9K/129G; indeed, after 48 h of starvation, even in the presence of 10 ng/ml TGF-β, the percentage of annexin V+ NK cells pretreated with either PTX-B or PT9K/129G was comparable to that detected in the absence of TGF-β (∼15 vs 37% in the presence of TGF-β alone; Fig. 4,B). Moreover, the effect of PTX-B on apoptosis is probably dependent on PI3K activation, because it was blocked by LY294002 (Fig. 4 B).

FIGURE 4.

TGF-β enhances and accelerates starvation-induced NK cell apoptosis: inhibition by PTX-B or PT9K/129G. Apoptosis of NK cells was evaluated by FACS analysis after annexin V staining (A and B) or PI staining (C and D) in NK cells cultured in the absence of IL-2 and in the presence of TGF-β (0.1–10 ng/ml) for the indicated time periods (A) or for 48 h (B), with or without PTX-B (1 nM), in the absence or the presence of the PI3K inhibitor LY294002 (LY; 10 μM) or PT9K/129G (1 nM). Values are the mean ± SD from six independent experiments. C, Apoptosis evaluated by FACS analysis of NK cells stained with PI after 48 h of culture without IL-2 (as in B), untreated (left histogram; none), treated with 10 ng/ml TGF-β (central histogram), or treated with TGF-β, 10 min after incubation with 1 nM PTX-B (right histogram). The percentage of apoptotic cells (cells with DNA content <2n) is indicated in each panel. The data shown are from one representative experiment of six. D, Apoptosis evaluated as in C in NK cells after 48 h of culture without IL-2, in the absence (left histogram; none) or the presence of supernatant from autologous NK cells cultured for 24 h with Tat (central histogram), without or with (right histogram) the neutralizing mAb anti-TGF-β (5 μg/ml) added at the onset of 48-h incubation. A: ∗, p < 0.01, TGF-β (48, 72, and 96 h) vs none. B: ∗∗, p < 0.01, TGF-β vs none; ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β; ∗∗∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

FIGURE 4.

TGF-β enhances and accelerates starvation-induced NK cell apoptosis: inhibition by PTX-B or PT9K/129G. Apoptosis of NK cells was evaluated by FACS analysis after annexin V staining (A and B) or PI staining (C and D) in NK cells cultured in the absence of IL-2 and in the presence of TGF-β (0.1–10 ng/ml) for the indicated time periods (A) or for 48 h (B), with or without PTX-B (1 nM), in the absence or the presence of the PI3K inhibitor LY294002 (LY; 10 μM) or PT9K/129G (1 nM). Values are the mean ± SD from six independent experiments. C, Apoptosis evaluated by FACS analysis of NK cells stained with PI after 48 h of culture without IL-2 (as in B), untreated (left histogram; none), treated with 10 ng/ml TGF-β (central histogram), or treated with TGF-β, 10 min after incubation with 1 nM PTX-B (right histogram). The percentage of apoptotic cells (cells with DNA content <2n) is indicated in each panel. The data shown are from one representative experiment of six. D, Apoptosis evaluated as in C in NK cells after 48 h of culture without IL-2, in the absence (left histogram; none) or the presence of supernatant from autologous NK cells cultured for 24 h with Tat (central histogram), without or with (right histogram) the neutralizing mAb anti-TGF-β (5 μg/ml) added at the onset of 48-h incubation. A: ∗, p < 0.01, TGF-β (48, 72, and 96 h) vs none. B: ∗∗, p < 0.01, TGF-β vs none; ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β; ∗∗∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

Close modal

These results were confirmed by PI staining, which identified apoptotic cells (DNA content, <2n; Fig. 4,C); indeed, ∼8% of NK cells were apoptotic, evaluated as PI+ cells after 48 h of starvation (left histogram), being ∼35% in the presence of 10 ng/ml TGF-β (central histogram), whereas <10% of apoptotic cells were detected after pretreatment with PTX-B (Fig. 4,C, right histogram) or with PT9K/129G (not shown) before exposure to TGF-β. Of note, NK cell apoptosis was also detected after 48-h culture of NK cells in the presence of supernatant from autologous NK cells treated for 24 h with Tat (Fig. 4,D, central histogram) containing 20 ng/ml TGF-β (Fig. 1,C) and diluted 1/2; this apoptosis was blocked by adding neutralizing anti-TGF-β mAb (5 μg/ml) (Fig. 4 D, right histogram), supporting the idea of a Tat-induced, TGF-β-mediated autocrine pathway.

We investigated the molecular mechanism by which PTX-B antagonizes TGF-β. This cytokine is known to induce dephosphorylation of Akt, thus impairing the protection from apoptosis (29, 30). Interestingly, PTX is an activator of PI3K, which, in turn, leads to phosphorylation and activation of Akt (34, 37). In keeping with this, we found that the percentage of pAkt vs total Akt increased from 5 to 25% at 8 min upon treatment of NK cells with 1 nM PTX-B or PT9K/129G (Fig. 5,A). This effect was blocked in the presence of the PI3K blocker LY294002 (Fig. 5,A), supporting the hypothesis that PTX-B acts through the activation of PI3K. We also analyzed the effect of TGF-β on NKG2D oligomerization in NK cells. Indeed, NKG2D, beside triggering NK cell cytolytic function, is known to activate PI3K and Akt in NK cells (11, 47), thus possibly representing a molecule delivering a survival signal. We found that oligomerization of NKG2D, achieved by a specific mAb followed by GAM, induced Akt phosphorylation (6-fold increase at 4 min; Fig. 5,B), and this effect was inhibited by TGF-β (Fig. 5,B). Of note, TGF-β-mediated inhibition of Akt phosphorylation induced by NKG2D engagement was abolished by PTX-B or PT9K/129G (Fig. 5,C). Again, the PI3K blocker LY294002 deleted the protective effect exerted by PTX-B and PT9K/129G (Fig. 5 C).

FIGURE 5.

PTX-B and PT9K/129G activate Akt and prevent TGF-β inhibition of Akt phosphorylation through a PI3K-dependent pathway. Phosphorylation of Akt (pAkt) in cell lysates of NK cells was assessed by ELISA. The same samples were also analyzed for the content of total Akt by ELISA; results are expressed as the percentage of pAkt, normalized for total Akt, and are the mean ± SD of four independent experiments. A, NK cells exposed to PTX-B or PT9K/129G (both at 1 nM) for the indicated periods of time. None, untreated NK cells. B, NK cells upon ligation of NKG2D, obtained with the specific mAb (5 μg/ml), followed by addition of GAM and incubation at 37°C for 2, 4, or 8 min in the presence or the absence of TGF-β (10 ng/ml). As a control, cells were exposed to an unrelated isotype-matched mAb (BD Biosciences) plus GAM (none). Values are the mean ± SD of four independent experiments. C, pAkt measured in NK cells untreated (none) or 4 min after engagement of NKG2D in the absence (medium) or the presence of TGF-β alone (10 ng/ml), TGF-β with either PTX-B or PT9K/129G (both at 1 nM), or PTX-B or PT9K/129G alone. Values are the mean ± SD of four independent experiments. The effect of the PI3K inhibitor LY294002 (10 μM) is shown in A and C. A: ∗, p < 0.01, PTX-B or PT9K/129G vs none; ∗∗, p < 0.01, PTX-B plus LY294002 vs PTX-B. B: ∗, p < 0.01, NKG2D vs none; ∗∗, p < 0.01, NKG2D plus TGF-β vs NKG2D. C: ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β; ∗∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

FIGURE 5.

PTX-B and PT9K/129G activate Akt and prevent TGF-β inhibition of Akt phosphorylation through a PI3K-dependent pathway. Phosphorylation of Akt (pAkt) in cell lysates of NK cells was assessed by ELISA. The same samples were also analyzed for the content of total Akt by ELISA; results are expressed as the percentage of pAkt, normalized for total Akt, and are the mean ± SD of four independent experiments. A, NK cells exposed to PTX-B or PT9K/129G (both at 1 nM) for the indicated periods of time. None, untreated NK cells. B, NK cells upon ligation of NKG2D, obtained with the specific mAb (5 μg/ml), followed by addition of GAM and incubation at 37°C for 2, 4, or 8 min in the presence or the absence of TGF-β (10 ng/ml). As a control, cells were exposed to an unrelated isotype-matched mAb (BD Biosciences) plus GAM (none). Values are the mean ± SD of four independent experiments. C, pAkt measured in NK cells untreated (none) or 4 min after engagement of NKG2D in the absence (medium) or the presence of TGF-β alone (10 ng/ml), TGF-β with either PTX-B or PT9K/129G (both at 1 nM), or PTX-B or PT9K/129G alone. Values are the mean ± SD of four independent experiments. The effect of the PI3K inhibitor LY294002 (10 μM) is shown in A and C. A: ∗, p < 0.01, PTX-B or PT9K/129G vs none; ∗∗, p < 0.01, PTX-B plus LY294002 vs PTX-B. B: ∗, p < 0.01, NKG2D vs none; ∗∗, p < 0.01, NKG2D plus TGF-β vs NKG2D. C: ∗, p < 0.01, TGF-β plus PTX-B or TGF-β plus PT9K/129G vs TGF-β; ∗∗, p < 0.01, TGF-β plus PTX-B plus LY294002 vs TGF-β plus PTX-B.

Close modal

Phosphorylated Akt is known to induce transcription of antiapoptotic proteins, such as Bcl-2 and Bcl-x; the latter, and in particular its isoform Bcl-xL, are involved in the protection from starvation-induced apoptosis (31, 32). Moreover, TGF-β-induced apoptosis is a consequence of Akt dephosphorylation induced by this cytokine, and this effect can be counteracted by up-regulating the PI3K-dependent Akt pathway (32, 33). Accordingly, in NK cells exposed to TGF-β, a reduction by 80% of Bcl-2 transcript was observed (Fig. 6,A, lanes 1 and 2; Fig. 6,B for densitometric analysis, expressed as a percentage of GAPDH; mean ± SD of four independent experiments, including that shown in Fig. 6,A). Although not statistically significant compared with basal levels, in TGF-β-treated NK cells, Bcl-xL mRNA was undetectable (Fig. 6,A, compare lane 2 with lane 1). Treatment with PTX-B (Fig. 6) or PT9K/129G (not shown) not only prevented the inhibitory effect of TGF-β on Bcl-2 (Fig. 6,A, lane 4 vs 2; Fig. 6,B for densitometric analysis), but also induced Bcl-xL transcription, even in the presence of TGF-β (7-fold increase vs basal levels; Fig. 6).

FIGURE 6.

TGF-β inhibition of Bcl-2 transcription in NK cells is counteracted by PTX-B. A, Bcl-2, Bcl-xL, and Bcl-xS mRNA were evaluated by PCR in NK cells, untreated (none; lane 1) or treated with either 10 ng/ml TGF-β (lane 2) or 1 nM PTX-B alone (lane 3) or before adding TGF-β (lane 4) and compared with GAPDH. B, Densitometric analysis, expressed as Bcl-2, Bcl-xL, or Bcl-xS percentage of GAPDH in the same sample, of the experiment depicted in A and three other experiments; values are the mean ± SD of four independent experiments. None, untreated NK cells. ∗, p < 0.01, TGF-β vs none; ∗∗, p < 0.01, PTX-B vs none; ∗∗∗, p < 0.01, TGF-β plus PTX-B vs TGF-β.

FIGURE 6.

TGF-β inhibition of Bcl-2 transcription in NK cells is counteracted by PTX-B. A, Bcl-2, Bcl-xL, and Bcl-xS mRNA were evaluated by PCR in NK cells, untreated (none; lane 1) or treated with either 10 ng/ml TGF-β (lane 2) or 1 nM PTX-B alone (lane 3) or before adding TGF-β (lane 4) and compared with GAPDH. B, Densitometric analysis, expressed as Bcl-2, Bcl-xL, or Bcl-xS percentage of GAPDH in the same sample, of the experiment depicted in A and three other experiments; values are the mean ± SD of four independent experiments. None, untreated NK cells. ∗, p < 0.01, TGF-β vs none; ∗∗, p < 0.01, PTX-B vs none; ∗∗∗, p < 0.01, TGF-β plus PTX-B vs TGF-β.

Close modal

To verify the potential pathological relevance of our findings, we studied 20 HIV-1-infected patients at stage A of the disease. We evaluated the mRNA for Bcl-2, Bcl-xL, and Bcl-xS in purified NK cells and PBMC obtained from these patients compared with that in NK cells and PBMC from healthy donors matched for sex and age. As shown in Fig. 7 (A and B show eight representative patients and eight representative healthy donors), the level of Bcl-xL mRNA transcripts was consistently lower in purified NK cells of HIV-1-infected patients (Fig. 7,A) than in healthy donors (Fig. 7,B). In particular, densitometric analysis performed on experiments obtained with NK cells from all 20 HIV-1 patients revealed that the Bcl-xL mRNA content is barely detectable, being reduced by 90% compared with that of NK cells from healthy donors (Fig. 7,C). Moreover, Bcl-2 mRNA was decreased by 70% in NK cells from HIV-1 patients (Fig. 7, A and C) compared with healthy donors (Fig. 7, B and C); conversely, no significant difference in Bcl-xS was found (Fig. 7,C). Similar results were obtained by analyzing unfractionated PBMC from HIV-1+ patients and healthy donors (Fig. 7).

FIGURE 7.

Low expression of Bcl-x and Bcl-2 in NK cells from HIV-1-infected patients. A and B, Bcl-2, Bcl-xL, and Bcl-xS mRNA were evaluated by PCR in purified NK cells or PBMC from eight representative of 20 HIV-1-infected patients (Pt.; A, numbers refer to the patients listed in Table I) or eight representative of 15 healthy donors (HD; B) and compared with GAPDH. Lane 1: MW, m.w. C, Densitometric analysis, expressed as Bcl-2, Bcl-xL, or Bcl-xS percentage of GAPDH in the same sample of NK cells or PBMC; values are the mean ± SD of 20 HIV-1-infected patients (NK-Pt or PBMC-Pt) and the mean ± SD of 15 healthy donors (NK-HD or PBMC-HD) including those depicted in A and B. ∗, p < 0.01, Pt. vs HD; ∗∗, p < 0.01, NK-Pt. vs NK-HD.

FIGURE 7.

Low expression of Bcl-x and Bcl-2 in NK cells from HIV-1-infected patients. A and B, Bcl-2, Bcl-xL, and Bcl-xS mRNA were evaluated by PCR in purified NK cells or PBMC from eight representative of 20 HIV-1-infected patients (Pt.; A, numbers refer to the patients listed in Table I) or eight representative of 15 healthy donors (HD; B) and compared with GAPDH. Lane 1: MW, m.w. C, Densitometric analysis, expressed as Bcl-2, Bcl-xL, or Bcl-xS percentage of GAPDH in the same sample of NK cells or PBMC; values are the mean ± SD of 20 HIV-1-infected patients (NK-Pt or PBMC-Pt) and the mean ± SD of 15 healthy donors (NK-HD or PBMC-HD) including those depicted in A and B. ∗, p < 0.01, Pt. vs HD; ∗∗, p < 0.01, NK-Pt. vs NK-HD.

Close modal

TGF-β was present in the serum of all HIV-1 patients; in particular, in 13 of 20 patients, serum levels of TGF-β ranged between 50 and 100 ng/ml (Table I), whereas in healthy donors, TGF-β was always <10 ng/ml (n = 15; not shown). In six patients, Tat was detectable in the serum at 10–50 nM, and viremia was found (HIV-1 mRNA ranging between 1,000 and 24,000 copies/ml) in 11 patients (Table I). Interestingly, regression analysis showed that RNA copy number was significantly associated with TGF-β (r = 0.70) and Tat (r = 0.77) serum levels. Moreover, in these patients the percentage of CD3CD16+ (NK) cells was consistently lower (7 ± 3%) than that in patients with low or undetectable TGF-β or Tat (14 ± 4%) or in healthy donors (15 ± 3%; Table I).

In this paper we provide evidence that HIV-1 Tat induces both transcription and secretion of TGF-β, which, in turn, accelerates starvation-induced NK cell apoptosis, inhibits Akt phosphorylation, and down-regulates Bcl-xL and Bcl-2 transcription. All these effects are suppressed by PTX-B and by the nontoxic mutant of PTX, PT9K/129G, through a PI3K/Akt-dependent mechanism (Fig. 8).

FIGURE 8.

Proposed signaling pathways of Tat and TGF-β and role of PTX-B and PT-9K/129G. Tat is known to induce ERK1/2, which activates AP-1 and is responsible for TGF-β production. In turn, TGF-β itself can activate ERK1/2 and AP-1 through SMAD3/4; both these steps are inhibited by ERK1/2 inhibitors, such as PD98059. Another effect of TGF-β is the dephosphorylation of Akt, elicited through a double mechanism: activation of SHIP and/or of calcineurin. The down-regulation of Akt leads to a decrease in the synthesis of antiapoptotic proteins, such as Bcl-2 and Bcl-x. On the contrary, PTX-B or its nontoxic mutant, PT9K/129G, activates PI3K, maintaining pAkt and the synthesis of antiapoptotic proteins; this pathway is blocked by PI3K inhibitors, such as LY294002.

FIGURE 8.

Proposed signaling pathways of Tat and TGF-β and role of PTX-B and PT-9K/129G. Tat is known to induce ERK1/2, which activates AP-1 and is responsible for TGF-β production. In turn, TGF-β itself can activate ERK1/2 and AP-1 through SMAD3/4; both these steps are inhibited by ERK1/2 inhibitors, such as PD98059. Another effect of TGF-β is the dephosphorylation of Akt, elicited through a double mechanism: activation of SHIP and/or of calcineurin. The down-regulation of Akt leads to a decrease in the synthesis of antiapoptotic proteins, such as Bcl-2 and Bcl-x. On the contrary, PTX-B or its nontoxic mutant, PT9K/129G, activates PI3K, maintaining pAkt and the synthesis of antiapoptotic proteins; this pathway is blocked by PI3K inhibitors, such as LY294002.

Close modal

The finding that Tat stimulates TGF-β production is in keeping with previous data from the literature showing the same effect on cells of the CNS and bone marrow macrophages (6, 7, 8, 9); we confirm this effect on NK cells, which are considered crucial in the first-line defense against viral infections (10, 11, 12). TGF-β itself has a potential immunosuppressive function (7, 9); in addition, we found that this cytokine can inhibit the signal transduction triggered in NK cells via the NKG2D receptor, which is deeply involved in activating NK cell function (11, 47). TGF-β is also able to trigger proapoptotic signals (28, 29, 30, 33); in agreement with this, we observed that the cytokine can accelerate starvation-induced apoptosis of NK cells. These mechanisms, if operating in vivo, would lead to either functional impairment or even elimination of cells involved in the early antiviral response. Of note, PTX-B and its nontoxic mutant, PT9K/129G, are able to counteract all these biochemical events in NK cells.

PTX is a pentameric protein that is functionally composed of two subunits: the A (active) subunit is an adenine diphosphate ribosyl-transferase responsible for ribosylation and inactivation of Gi-like proteins, whereas the role of PTX-B was initially thought to be limited to the binding of PTX to target cells (48). However, it is now accepted that PTX-B alone can trigger signal transduction in monocytes, T and B lymphocytes, via a 43-kDa receptor and via either the CD11a/CD18 integrin or the CD14 molecule, leading to cell activation (35, 37, 49, 50). Although it is not yet clear whether this interaction is integrin-mediated in NK cells, we found that PTX-B prevents the synthesis and secretion of TGF-β induced by HIV-1 Tat in NK cells, mainly up-regulating Akt phosphorylation via PI3K activation and thus interfering with the ERK/AP-1 signal transduction pathway activated by Tat (19, 20, 21, 22, 23, 24). In this regard, the mechanism of action of PTX-B would be independent of the ADP-ribosylation of Gi-like proteins, which is selectively due to the PTX-A subunit (48, 50), but would be related to the ability of PTX-B to activate the PI3K/Akt pathway (37). Indeed, both PTX-B and the PTX mutant, PT9K/129G, lead to PI3K-dependent phosphorylation of Akt, which, in turn, prevents activation of the AP-1 components c-Jun, c-Fos, and JunD induced by Tat in NK cells. As a consequence, Tat-mediated transactivating effects are blocked, including those on synthesis and secretion of TGF-β (Fig. 8). Interestingly, the same pathway is activated by TGF-β, which acts mainly by up-regulating the Jun family of AP-1 transcription factors (25, 26, 27); thus, activation of PI3K/Akt by PTX-B or PT9K/129G leads to the inhibition of TGF-β’s biological effects. In particular, TGF-β-mediated apoptosis, which is based on the ability of the cytokine to induce Akt dephosphorylation (28, 29), is prevented. In addition, Tat itself has been reported to induce apoptosis through this mechanism (16, 17). In muscle cells, TGF-β has also been shown to activate calcineurin, a serine-threonine kinase that is responsible for mitochondrial-dependent apoptosis in cardiomyocytes, possibly through Akt dephosphorylation (51, 52). In turn, PTX-B- or PT9K/129G-mediated up-regulation of phosphorylated Akt would induce the transcription of antiapoptotic proteins such as Bcl-2 and Bcl-x, mainly the Bcl-xL isoform that is responsible for protection of NK cells from starvation-induced apoptosis (31, 32). It is noteworthy that the signal transduction triggered, in NK cells, via the activating molecule NKG2D is restored by maintaining Akt phosphorylation. By restoring the PI3K/Akt pathway, PTX-B or the nontoxic PTX mutant (Fig. 8) may contribute in vivo to the maintenance of NK cell function and survival, thus counteracting the potential immunosuppressive role of TGF-β in AIDS (8, 9). Indeed, NKG2D is a lectin-like type 2 transmembrane immunoreceptor expressed by CD8+ T lymphocytes, NK, and γδ T cells that contains a charged transmembrane residue interacting with signaling adaptor molecules, mainly DNAX-activating protein of 10 kDa, which, in turn, recruit and activate PI3K (47). NKG2D recognizes a number of MHC class I-related molecules, including MHC class I chain-related protein A and the UL16-binding proteins, that are generally poorly expressed by normal cells, but are up-regulated in transformed and infected cells. The engagement of NKG2D by its ligands is a sufficient stimulus to activate cytolysis and cytokine production by NK cells, but it also provides a costimulatory signal for the activation of T cells (47).

The pathological relevance of our data is also supported by the finding that in NK cells from patients with early HIV-1 infection, Bcl-xL and, to a lesser extent, Bcl-2 mRNA expressions were consistently lower than those in healthy donors; moreover, TGF-β and, in two-thirds of the cases, Tat were detected in the sera of these patients at concentrations that are biologically active in vitro. We cannot exclude that in these patients, TGF-β is produced by other cells than NK cells, including APCs, endothelial cells, or stromal cells in the lymph nodes, where similar Tat-mediated effects might be operative as well, thus possibly amplifying the immunosuppressive effect. Administration of PT9K/129G, which is already approved for human use as a component of a vaccine against B. pertussis infection (38, 39, 40), as a component of a Tat-based vaccine in HIV-1-infected patients might be of interest not only as an adjuvant (42), but also as a component potentially able to interfere with HIV-1 replication (36), with unwanted Tat effects (16, 17, 18), and with cytokine-mediated immunosuppressive action (7, 8, 9), allowing the first antiviral defense to be maintained.

We thank Rino Rappuoli (Chiron, Siena, Italy) for providing the PT9K/129G mutant; Chiara Rizzi (Scientific Institute San Raffaele, Milan, Italy) for technical help and suggestions, and Giuseppe Murdaca (Department of Internal Medicine, University of Genoa, Genoa, Italy) for helpful discussion. We also thank the Center for Statistics, University of Genoa, for help with statistical analyses.

The authors have no financial conflict of interest.

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 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported in part by the Italian Ministry of Health (Coordinated Project 2002–2004; to A.P.) and Grants 40.D84 (to M.R.Z.) and 40.D5 (to M.A.) from of the IV National Program of Research on AIDS of the Istituto Superiore di Sanità (AIDS Project).

3

Abbreviations used in this paper: pAkt, phosphorylated Akt; Av-HRP, avidin-peroxidase; GAM, goat anti-mouse; PI, propidium iodide; PTX-B, pertussis toxin B oligomer.

1
Poli, G., A. S. Fauci.
1993
. Cytokine modulation of HIV expression.
Semin. Immunol.
5
:
165
-173.
2
Reinhart, T. A..
2003
. Chemokine induction by HIV-1: recruitment to the cause.
Trends Immunol.
24
:
351
-353.
3
Murphy, P. M..
2001
. Viral exploitation and subversion of the immune system through chemokine mimicry.
Nat. Immunol.
2
:
116
-122.
4
Rappaport, J., J. Joseph, S. Croul, G. Alexander, L. Del Valle, S. Amini, K. Khalili.
1999
. Molecular pathways involved in HIV-1-induced CNS pathology: role of viral regulatory protein Tat.
J. Leukocyte Biol.
65
:
458
-465.
5
Ensoli, B., G. Barillari, R. C. Gallo.
1992
. Cytokines and growth factors in the pathogenesis of AIDS-associated Kaposi’s sarcoma.
Immunol. Rev.
127
:
147
-155.
6
Cupp, C., J. P. Taylor, K. Khalili, S. Amini.
1993
. Evidence for stimulation of the transforming growth factor β1 promoter by HIV-1 Tat in cells derived from CNS.
Oncogene
8
:
2231
-2236.
7
Zauli, G., B. R. Davis, M. C. Re, G. Visani, G. Furlini, M. La Placa.
1992
. Tat protein stimulates production of transforming growth factor-β1 by marrow macrophages: a potential mechanism for human immunodeficiency virus-1-induced hematopoietic suppression.
Blood
80
:
3036
-3043.
8
Sawaya, B. E., P. Thatikunta, L. Denisova, J. Brady, K. Khalili, S. Amini.
1998
. Regulation of TNFα and TGFβ-1 gene transcription by HIV-1 Tat in CNS cells.
J. Neuroimmunol.
87
:
33
-42.
9
Reinhold, D., S. Wrenger, T. Kahne, S. Ansorge.
1999
. HIV-1 Tat: immunosuppression via TGF-β1 induction.
Immunol. Today
8
:
384
-385.
10
Bellone, G., M. Aste-Amezaga, G. Trinchieri, U. Rodeck.
1995
. Regulation of NK cell functions by TGF-β1.
J. Immunol.
155
:
1066
-1073.
11
Castriconi, R., C. Cantoni, M. Della Chiesa, M. Vitale, E. Marcenaro, R. Conte, R. Biassoni, C. Bottino, L. Moretta, A. Moretta.
2003
. Transforming growth factor β1 inhibits expression of NKp30 and NKG2D receptors: consequences for the NK-mediated killing of dendritic cells.
Proc. Natl. Acad. Sci. USA
100
:
4120
-4125.
12
Rubartelli, A., R. Sitia, A. Poggi, M. R. Zocchi.
1998
. HIV-1 Tat: a polypeptide for all seasons.
Immunol. Today
19
:
543
-545.
13
Zocchi, M. R., A. Rubartelli, P. Morgavi, A. Poggi.
1998
. HIV-1 Tat inhibits human natural killer cell function by blocking L-type calcium channels.
J. Immunol.
161
:
2938
-2943.
14
Poggi, A., R. Carosio, M. G. Spaggiari, C. Fortis, G. Tambussi, G. Dell’Antonio, E. DalCin, A. Rubartelli, M. R. Zocchi.
2002
. NK cell activation by dendritic cells is dependent on LFA-1-mediated induction of calcium-calmodulin kinase II: inhibition by HIV-1 Tat C-terminal domain.
J. Immunol.
168
:
95
-101.
15
Frankel, A. D., C. O. Pabo.
1988
. Cellular uptake of the Tat protein from human immunodeficiency virus.
Cell
55
:
1189
-1193.
16
Li, C. J., D. J. Friedman, C. Wang, V. Metelev, A. B. Pardee.
1995
. Induction of apoptosis in uninfected lymphocytes by HIV-1 Tat protein.
Science
268
:
429
-431.
17
Westendorp, M. O., R. Frank, C. Ochsenbauer, K. Stricker, J. Dhein, H. Walczak, K. M. Debatin, P. H. Krammer.
1995
. Sensitization of T cells to CD95-mediated apoptosis by HIV-1 Tat and gp120.
Nature
375
:
497
-500.
18
Cohen, S. S., C. Li, L. Ding, Y. Cao, A. B. Pardee, E. M. Shevac, D. J. Cohen.
1999
. Pronounced acute immunosuppression in vivo mediated by HIV-1 Tat challenge.
Proc. Natl. Acad. Sci. USA
96
:
10842
-10847.
19
Mischiati, C., F. Pironi, D. Dilani, M. Giacca, P. Mirandola, S. Capitani, G. Zauli.
1999
. Extracellular HIV-1 Tat protein differentially activates the JNK and ERK/MAPK pathways in CD4 T cells.
AIDS
13
:
1637
-1645.
20
Manna, S. K., B. B. Aggarwal.
2000
. Vesnarinone suppresses TNF-induced activation of NF-κB, c-Jun kinase, and apoptosis.
J. Immunol.
164
:
5156
-5166.
21
Gu, Y., R. F. Wu, Y. C. Xu, S. C. Flores, L. S. Terada.
2001
. HIV Tat activates c-Jun amino-terminal kinase through an oxidant-dependent mechanism.
Virology
286
:
62
-71.
22
Roebuck, K. A., D. A. Brenner, M. F. Kagnoff.
1993
. Identification of c-fos-responsive elements downstream of TAR in the long terminal repeat of human immunodeficiency virus type-1.
J. Clin. Invest.
91
:
1336
-1348.
23
Roebuck, K. A., M. F. Rabbi, M. F. Kagnoff.
1997
. HIV-1 Tat protein can transactivate a heterologous TATAA element independent of viral promoter sequences and the trans-activation response element.
AIDS
11
:
139
-146.
24
Ghibellini, D., A. Caputo, S. Capitani, M. La Placa, G. Zauli.
1997
. Upregulation of c-Fos in activated T lymphoid and monocytic cells by human immunodeficiency virus-1 Tat protein.
Blood
89
:
1654
-1664.
25
Virolle, T., M. N. Monthouell, Z. Djabari, J. P. Ortonne, G. Meneguzzi, D. Aberdam.
1998
. Three activator protein-1 binding sites bound by the Fra-2.JunD complex cooperate for the regulation of murine laminin α3A (lama3A) promoter activity by transforming growth factor-β.
J. Biol. Chem.
273
:
17318
-17325.
26
Liberati, N. T., M. B. Datto, J. P. Fraderick, X. Shen, C. Wong, E. M. Rougier-Chapman, X. F. Wang.
1999
. Smads bind directly to the Jun family of AP-1 transcription factor.
Proc. Natl. Acad. Sci. USA
96
:
4844
-4849.
27
Yue, J., K. M. Mulder.
2000
. Requirement of Ras/MAPK pathway activation by transforming growth factor β for transforming growth factor β1 production in a Smad-dependent pathway.
J. Biol. Chem.
275
:
30765
-30773.
28
Yamamura, Y., X. Hua, S. Bergelson, H. F. Lodish.
2000
. Critical role of Smads and AP-1 complex in transforming growth factor-β-dependent apoptosis.
J. Biol. Chem.
275
:
36295
-36302.
29
Arsura, M., G. R. Panta, J. D. Bilyeu, L. G. Cavin, M. A. Sovak, A. A. Oliver, V. Factor, R. Heuchel, F. Mercurio, S. S. Thorgeirsson, et al
2003
. Transient activation of NF-κB through a TAK1/IKK kinase pathway by TGF-β1 inhibits AP-1/SMAD signaling and apoptosis: implications in liver tumor formation.
Oncogene
22
:
412
-425.
30
Valderrama-Carvajal, H., E. Cocolakis, E. H. Lacerte, A. Lee, G. Krystal, S. Ali, J. J. Lebrun.
2002
. Activin/TGF-β induce apoptosis through Smad-dependent expression of the lipid phosphatase SHIP.
Nat. Cell. Biol.
4
:
963
-969.
31
Pugazhenthi, S., A. Nesterova, C. Sable, K. A. Heidenreich, L. M. Boxer, J. E. Reusch.
2000
. Akt/protein kinase B up-regulates Bcl-2 expression through cAMP-response element-binding protein.
J. Biol. Chem.
275
:
10761
-10766.
32
Chao, D. T., S. J. Korsmeyer.
1998
. Bcl-2 family: regulators of cell death.
Annu. Rev. Immunol.
16
:
395
-419.
33
Chen, R. H., Y. H. Su, R. L. Chuang, T. Y. Chang.
1998
. Suppression of transforming growth factor-β-induced apoptosis through a phosphatidylinositol 3-kinase/Akt-dependent pathway.
Oncogene
17
:
1959
-1968.
34
Alfano, M., G. Vallanti, P. Biswas, C. Bovolenta, E. Licenzi, B. Mantelli, T. Pushkarsky, R. Rappuoli, A. Lazzarin, M. Bukrinsky, et al
2001
. The binding subunit of pertussis toxin inhibits HIV replication in human macrophages and virus expression in chronically infected promonocytic U1 cells.
J. Immunol.
166
:
1863
-1870.
35
Alfano, M., H. Schmidtmayerova, C. A. Amella, T. Pushkarsky, M. Bukrinsky.
1999
. The B-oligomer of pertussis toxin deactivates CC chemokine receptor 5 and blocks entry of M-tropic HIV-1 strains.
J. Exp. Med.
190
:
597
-605.
36
Alfano, M., T. Pushkarsky, G. Poli, M. Bukrinsky.
2000
. The B-oligomer of pertussis toxin inhibits human immunodeficiency virus type 1 replication at multiple stages.
J. Virol.
74
:
8767
-8770.
37
Li, H., W. S. Wong.
2001
. Pertussis toxin activates tyrosine kinase signaling cascade in myelomonocytic cells: a mechanism for cell adhesion.
Biochem. Biophys. Res. Commun.
283
:
1077
-1082.
38
Podda, A., L. Nencioni, M. T. De Magistris, A. Di Tommaso, P. Bossu, S. Nuti, P. Pileri, S. Peppoloni, M. Bugnoli, P. Ruggiero, et al
1990
. Metabolic, humoral, and cellular responses in adult volunteers immunized with the genetically inactivated pertussis toxin mutant PT-9K/129G.
J. Exp. Med.
172
:
861
-868.
39
Rappuoli, R..
1997
. Rational design of vaccines.
Nat. Med.
3
:
374
-376.
40
Del Giudice, G., R. Rappuoli.
1999
. Genetically derived toxoids for use as vaccines and adjuvants.
Vaccine
17
:(Suppl. 2):
S44
-S52.
41
Nencioni, L., M. G. Pizza, G. Volpini, M. T. De Magistris, F. Giovannoni, R. Rappuoli.
1991
. Properties of the B oligomer of pertussis toxin.
Infect. Immun.
59
:
4732
-4734.
42
Roberts, M., A. Bacon, R. Rappuoli, M. Pizza, I. Cropley, G. Douce, G. Dougan, M. Marinaro, J. McGhee, S. Chatfield.
1995
. A mutant pertussis toxin molecule that lacks ADP-ribosyltransferase activity, PT-9K/129G, is an effective mucosal adjuvant for intranasally delivered proteins.
Infect. Immun.
63
:
2100
-2108.
43
Ferrero, E., D. Belloni, P. Contini, C. Foglieni, M. E. Ferrero, M. Fabbri, A. Poggi, M. R. Zocchi.
2003
. Transendothelial migration leads to protection from starvation-induced apoptosis in CD34+CD14+ circulating precursors: evidence for PECAM-1 involvement through Akt/PKB activation.
Blood
101
:
186
-193.
44
Poggi, A., R. Carosio, D. Fenoglio, S. Brenci, G. Murdaca, M. Setti, F. Indiveri, S. Scabini, E. Ferrero, M. R. Zocchi.
2004
. Migration of Vδ1 and Vδ2 T cells in response to CXCR3 and CXCR4 ligands in healthy donors and HIV-1-infected patients: competition by HIV-1 Tat.
Blood
103
:
2205
-2213.
45
Poggi, A., R. Carosio, A. Rubartelli, M. R. Zocchi.
2002
. β3-Mediated engulfment of apoptotic tumor cells by dendritic cells is dependent on CAMKII: inhibition by HIV-1 Tat.
J. Leukocyte Biol.
71
:
531
-537.
46
Rizzi, C., M. Alfano, A. Bugatti, M. Camozzi, G. Poli, M. Rusnati.
2004
. Inhibition of intra- and extra-cellular Tat function and HIV expression by pertussis toxin B-oligomer.
Eur. J. Immunol.
34
:
530
-536.
47
Raulet, D. H..
2003
. Roles of the NKG2D immunoreceptor and its ligands.
Nat. Rev. Immunol.
3
:
781
-790.
48
Wong, W. S., P. M. Rosoff.
1996
. Pharmacology of pertussis toxin B-oligomer.
Can. J. Physiol. Pharmacol.
74
:
559
-564.
49
Rosoff, P. M., R. Walker, L. Winberry.
1987
. Pertussis toxin triggers rapid second messenger production in human T lymphocytes.
J. Immunol.
139
:
2419
-2423.
50
Kolb, J. P., E. Genot, E. Petit-Koskas, N. Paul-Eugene, B. Dugas.
1990
. Effect of bacterial toxins on human B cell activation. I. Mitogenic activity of pertussis toxin.
Eur. J. Immunol.
20
:
969
-976.
51
Mann, K. M., J. L. Ray, E. S. Moon, K. M. Sass, M. R. Benson.
2004
. Calcineurin initiates smooth muscle differentiation in neural crest stem cells.
J. Cell Biol.
165
:
483
-491.
52
Molketin, J. D..
2001
. Calcineurin, mitochondrial membrane potential, and cardiomyocyte apoptosis.
Circ. Res.
88
:
1220
-1222.