Pertussis toxin B-oligomer dissociates T cell activation and HIV replication in CD4 T cells released from infected lymphoid tissue

Objective: To investigate, in human lymphoid tissue infected with HIV-1 ex vivo, the immunostimulatory and HIV inhibitory properties of pertussis toxin B oligomer (PTX-B) and of the genetically modified non-toxic PT-9K/129G.

Methods: Human tonsils from uninfected donors were infected ex vivo with R5 or X4 HIV-1 in the presence or absence of PTX-B. Virus replication was evaluated in culture supernatants; cells emigrated from tissue blocks were immunostained for lymphocytic and activation markers. HIV DNA and cell proliferation were evaluated with real-time PCR and [³H]thymidine incorporation, respectively.

Results: Both PTX-B and PT-9K/129G inhibited HIV-1 replication. These compounds activated and stimulated the proliferation of emigrated cells, most of which were CD4 T lymphocytes. Cells emigrated from infected tissues did not produce detectable virus in unstimulated or in PTX-B- or PT-9K/129G-stimulated cultures whereas robust virus production was triggered by phytohemagglutinin (PHA) or interleukin-2 (IL-2). Analysis of HIV DNA content indicated that infected cells were present among emigrated cells and that their number greatly increased following IL-2 stimulation, whereas it remained constant in the presence of PTX-B or PT-9K/129G.

Conclusions: PTX-B and PT-9K/129G inhibit both R5 and X4 HIV-1 replication in human lymphoid tissue ex vivo. In contrast to PHA and IL-2, they promote the proliferation of CD4 T lymphocytes emigrated from tissue, including HIV-infected cells, without triggering virus replication. Therefore, these emigrated CD4 T cells represent a novel model of a latent inducible HIV reservoir. Thus, PTX-B and the clinically approved PT-9K/129G are potential antiretroviral agents endowed with immunostimulatory capacity.

Introduction

Although viremia and CD4 T-cell numbers are routinely evaluated in peripheral blood, crucial events of HIV disease occur in lymphoid tissue ^[1-3]. In order to study these events under controlled laboratory conditions, we developed and characterized a system of human lymphoid tissue ex vivo. In this system, the in vivo-like cytoarchitecture and cellular repertoire ^[4,5], including heterogeneic lymphocyte subsets and the network of follicular-dendritic cells, are retained ^[6]. Also, ex vivo human lymphoid tissues support productive infection by both CCR5-dependent (R5) and CXCR4-dependent (X4) HIV-1 without exogenous stimulation ^[6-8].

For the present study, we adopted this system to further investigate the anti-HIV properties of pertussis toxin B-oligomer (PTX-B), the nontoxic component of PTX ^[9], which mediates cell binding and signaling of PTX ^[10,11]. PTX-B is a well-known mitogen ^[12] that activates protein kinase C and calcium fluxes leading to cytokine production ^[13,14]. A genetically modified non-toxic form of PTX, PT-9K/129G, has been developed as an acellular human vaccine against Bordetella pertussis infection ^[15,16] and retains major PTX-B properties including its anti-HIV activity in isolated cells ^[17]. It has been previously shown that PTX-B prevents entry of R5 HIV-1 in isolated T cells and macrophages and that it also inhibits X4 HIV-1 replication at a post-entry level ^[17-19].

In the study we report on here, we demonstrated that: PTX-B and PT-9K/129G inhibit both R5 and X4 HIV-1 replication in ex vivo-infected human lymphoid tissue; PTX-B and PT-9K/129G stimulate emigration of lymphocytes out of the tissue and promote their proliferation; and the cells that emigrate include HIV-infected CD4 T cells that maintain viral latency in spite of being activated by PTX-B or PT-9K/129G.

Methods

Reagents

PTX-B (Calbiochem, San Diego, California, USA) was dissolved in sterile phosphate-buffered saline (PBS; BioWhittaker, Verviers, Belgium) and stored at 4°C; PT-9K/129G (Chiron Corporation, Emeryville, California, USA) was dissolved in 50% glycerol containing NaCl at a final concentration of 0.5 M.

Ex vivo human lymphoid tissue preparation and infection

Human tonsils were obtained from patients undergoing tonsillectomy, or cardiac surgery in the case of thymus removal, at the Children's Hospital in Washington, DC, and at the San Raffaele Hospital in Milan, Italy, under an IRB-approved protocol, within 3 h of the operation. Tissues were kept in PBS and washed in complete culture medium composed of RPMI 1640 (GIBCO BRL, Grand Island, New Jersey, USA) containing 15% heat-inactivated fetal calf serum (FCS) (Summit Biotechnology, Fort Collins, Colorado, USA), non-essential amino acids (1 mM), sodium pyruvate (1 mM), L-glutamine (292 μg/ml), amphotericin B (2.5 μg/ml; GIBCO BRL), timentin (310 μg/ml; SmithKline Beecham, Philadelphia, Pennsylvania, USA), and gentamicin (50 μg/ml; Quality Control Inc., Rockville, Maryland, USA), and sectioned into 2-mm³ blocks. Nine tissue blocks were placed on top of medium-hydrated collagen sponge gel (Gelfoam, Johnson & Johnson Medical Ltd, Gargrave, Skipton, UK) in complete medium at the air-liquid interface in each well of a six-well plate (Costar, Cambridge, Massachusetts, USA), for a total of 54 tissue blocks for each experimental condition. For each experiment, we used tissue from one donor, and the experiments were repeated n times, where n is indicated in the text. After 24 h, the culture supernatant was replaced with fresh medium containing 1% Pen/Strep (GIBCO BRL), and tissue blocks were incubated for 20 min with or without PTX-B at different concentrations and then infected with HIV-1. In a few experiments mentioned below, we utilized thymus tissue processed according to an identical protocol.

Infection of lymphoid tissue blocks was carried out with 300 50% tissue culture infectious doses (TCID₅₀), corresponding approximately to 1 ng of HIV-1 p24 Gag antigen (p24_Gag) in 3-5 μl of viral stock previously filtered through a 0.2-μm filter (Millipore, Bedford Center, Massachusetts, USA) loaded on the top of each tissue block. Infection was performed with prototypic tissue culture laboratory-adapted HIV-1, either R5 (SF162 and Ba-L) or X4 (LAV.04) ^[6]; HIV-1 replication was assessed by means of an HIV-1 p24_Gag Ag enzyme-linked immunosorbent assay (Beckman Coulter, Fullerton, California, USA) or an Mg²⁺-dependent reverse transcriptase (RT) activity assay measured in culture supernatants ^[17]. In particular, 24 ml of culture supernatant bathing six collagen sponge gels and 54 tissue blocks was collected for each condition in a single plastic tube (Falcon; Becton Dickinson Labware, Lincoln Park, New Jersey, USA), and p24_Gag or RT activity determination was performed on an aliquot of this fluid. Culture supernatants were completely removed every 2-3 days, stored at -80°C pending analysis of viral content, and replaced with fresh complete medium containing or not containing PTX-B at the indicated concentrations. The collected fluid was centrifuged at 250 g for 10 min before being tested for virus content, in order to avoid contamination by cell-associated viral Ag. Tissue cultures were checked daily by means of light microscopy (Zeiss, Göttingen, Germany) for bacterial contamination.

Analysis of cells emigrated from lymphoid tissues

The cells that were spontaneously released into the culture supernatants from the 54 tissue blocks cultured for each condition were harvested after 72 h of culture, centrifuged, and resuspended at 1 × 10⁶ cells/ml in complete medium. In some experiments, 200 μl of cell suspension was seeded in 96-well plastic plates (Costar) and stimulated with recombinant interleukin-2 (IL-2, Chiron Corporation, Emeryville, California, USA) at 400 IU/ml or with different concentrations of PTX-B. After an additional 72 h, cells were resuspended in RPMI 1640 medium (GIBCO BRL) containing 10% FCS, and approximately 50% of the supernatant was removed every 72 h, stored at -80°C pending analysis of virus content, and replaced with fresh complete medium supplemented or not supplemented with PTX-B.

Cell proliferation and flow cytometry

Emigrated cells were resuspended at 1 × 10⁶ cells/ml in 200 μl of their own culture supernatant and incubated in the presence of 0.5 μCi of [³H]thymidine with a specific activity of 25 Ci/mmol (Amersham Biosciences AB, Piscataway, New Jersey, USA). Cells were harvested after 16 h of culture, and [³H]thymidine incorporation was measured with a β-counter (Top Count, Packard, Downers Grove, Illinois, USA). Lymphocytes were identified according to their light-scattering properties and then further analyzed for expression of typical lymphocyte markers. Cells were resuspended in cold PBS at 1 × 10⁶ cells/ml and incubated with propidium iodide (PI, an indicator of cell viability; Boehringer Mannheim, Indianapolis, Indiana, USA) at 2 μg/ml for 5 min on ice. More than 95% of emigrated cells excluded PI. A second aliquot of cells was stained for determination of lymphocyte subsets and for their state of activation with various combinations of CD3-TriColor, CD8-fluorescein isothyocyanate (FITC), CD19-phycoerythrin (PE), CD3-FITC, CD4-PE, CD69-TriColor, HLA-DR-allophycocyanin (APC), HLA-DR-FITC, and CD25-APC (Caltag Laboratories, Burlingame, California, USA) monoclonal antibodies (MAbs). Cells gated on PI-negative scatter, previously identified, and staining positive for CD45-peridinin chlorophyll protein (PerCP) (Anti-Hle-1; Becton Dickinson, San Jose, California USA) were acquired on a four-color FACSCalibur (Becton Dickinson). The CellQuest software was utilized for both acquisition and analysis; 25 000 cells were acquired in the PI-negative gate for each analysis. To enumerate cells in tissue blocks, we used Truecount tubes (Beckton Dickinson) containing a known number of fluorescent beads. The absolute number of cells in the sample was determined by normalization to the number of acquired beads. To pool the data from different experiments and experimental conditions, we weighed blocks and used their weight as a normalizing parameter.

Real-time PCR quantification of HIV-1 DNA

We determined the number of HIV DNA copies in emigrated cells using the TaqMan assay with an ABI7700 Prism instrument (Applied Biosystems, Foster City, California, USA). HIV-1 strains were treated with 70 U/ml DNAse (Roche Diagnostic Corp., Indianapolis, Indiana, USA) for 30 min at room temperature, and infection was carried out as described above. Aliquots of 1 × 10⁶ cells were harvested at the indicated time points, resuspended in PBS, and centrifuged at 12 000 rpm twice. Cell pellets were then lysed in a proteinase K-containing buffer, and the DNA was extracted by phenol-chloroform and precipitated with ethanol. Total DNA (250 ng) was amplified with a primer pair and probe derived from the HIV-1 gag sequence ^[20]: forward 5′-ACATCAAGCAGCCATGCAAAT-3′; reverse 5′-ATCTGGCCTGGTGCAATAGG-3′; probe 5′-(FAM)CATCAATGAGGAAGCTGCAGAATGGGATAGA(TAMRA)-3′. The HIV-1 copy numbers were normalized to the number of GAPDH copies (as evaluated with the primer pair forward 5′-ACCACAGTCCATGCCATCACT-3′; reverse 5′-GGCCATCACGCCACAGITT-3′) and quantified with real-time PCR using SYBR Green ^[21]). The thermal cycling conditions were 50°C for 2 min, 95°C for 12 min, and 40 cycles of 95°C for 15 s and 65°C for 1 min.

Statistical analysis

All results are expressed as means ± SEM of experiments performed on tissues from several different donors. The exact numbers of the donors (n) are stated for each individual experiment. Statistical analysis was performed with a one-way ANOVA paired test.

Results and discussion

PTX-B and PT-9K/129G inhibit R5 and X4 HIV-1 replication in lymphoid tissues

Inoculation of blocks of human tonsils from normal donors with either R5 or X4 HIV-1 resulted in a productive infection, as evaluated from either p24_Gag or RT activity content in the culture supernatants ^[4]. Productive infection usually became evident approximately 6 days after infection, peaked at day 9, and then decreased (Fig. 1). PTX-B inhibited replication of both R5 and X4 viruses in these ex vivo-infected human tonsillar cultures (Fig. 1a). We confirmed these results with blocks of thymic tissue (data not shown).

Fig. 1

There was no difference between the antiviral activities of PTX-B and PT-9K/129G, in that both molecules were equally efficient in inhibiting X4 and R5 replication. The maximal antiviral activities of PTX-B (Fig. 1a) and PT-9K/129G (Fig. 1b) were achieved at the submitogenic concentration of 10 pM, which reduced virus replication to 25 ± 10% of control levels. At mitogenic concentrations (1 and 5 nM, the highest concentrations used), both PTX-B and PT-9K/129G inhibited HIV replication to 52 ± 9% and 66 ± 9% of control levels (n = 20), respectively.

PTX-B triggers activation and proliferation of CD4 T lymphocytes emigrated from lymphoid tissue

Typically, a fraction of lymphocytes in ex vivo-cultured tissue blocks leaves them and can be recovered in the bathing medium within 3 days of the beginning of the culture ^[22]. Two main cellular populations, 'lymphocytes' and 'lymphoblasts', have been described on the basis of their light-scattering properties among cells isolated from lymphoid tissues ^[8]. Similar cells were identified here among those emigrated from lymphoid tissue ex vivo. Approximately 85% of these cells were 'lymphocytes', whereas 15% were 'lymphoblasts' (Fig. 2a). Similarly to what was found in tissue blocks, T (CD45CD3) and B (CD45CD19) cells were equally represented in emigrated cells (Fig. 2b). Among emigrated T cells in both the 'lymphocyte' and 'lymphoblast' subsets, 80 ± 6% were CD45CD3CD4, whereas 20 ± 4% were CD45CD3CD8 (Fig. 2b and Table 1). Infection of tissue blocks by R5 or X4 HIV-1 did not grossly affect the distribution of B and T 'lymphocytes' and 'lymphoblasts' among emigrated cells.

Fig. 2

Table 1

PTX-B stimulation increased the total numbers of emigrated cells with a 'lymphoblast' phenotype in a concentration-dependent manner (Fig. 2a). An average twofold increase in the total number of emigrated cells relative to control tissues was also observed with the lowest tested concentration of PTX-B, i.e. 0.2 nM (data not shown), reaching a fourfold increase at 5 nM (38 ± 31×10⁴ versus 9.3 ± 7.6×10⁴ cells per tissue block in PTX-B-stimulated and matched unstimulated tissues, respectively; n = 7, P = 0.039). Stimulation with PTX-B (5 nM) led to a preferential increase in the number of emigrated CD4 T 'lymphoblasts' from 42 ± 2% in controls to 70 ± 5% after stimulation; (P < 0.05, Table 1). Concomitantly, the fraction of B cells within 'lymphoblasts' decreased from 48 ± 3% in controls to 22 ± 4% in PTX-B treated tissues (P < 0.01, Table 1). In contrast, PTX-B stimulation did not affect the distribution of B, CD4, and CD8 T cells among emigrated 'lymphocytes' (Table 1).

Cells emigrated from PTX-B-stimulated tissues infected with either R5 or X4 HIV-1 exhibited phenotypes similar to those described above for uninfected tissues. However, for HIV-infected tissues the PTX-B-induced expansion of CD4 T 'lymphoblasts' was less pronounced (70 ± 5% in matched uninfected tissues versus 60 ± 3% in infected tissues; n = 8, P = 0.15), with a proportional increase of CD8 T cells (P < 0.01, Table 1). HIV-1 infection did not change the distribution of B, CD4, and CD8 T cells among 'lymphocytes' (Table 1).

The increased numbers of cells found outside the PTX-B-treated tissue blocks could be the result either of emigration or of enhanced proliferation of emigrated cells. To distinguish between these two possibilities, we assessed the numbers of CD4 and CD8 T cells that remained associated with the tissue after PTX-B stimulation ^[7]. The numbers of T lymphocytes remaining in the tissue blocks after PTX-B stimulation, as evaluated from the known number of added fluorescent beads and tissue weight for normalization, were similar to those observed in matched control tissues (98 ± 5% and 97 ± 7% for CD4 and CD8 T cells, respectively, n = 3). These results suggested that the increase in the numbers of cells found in the bathing medium of PTX-B-stimulated tissues was mostly accounted for by the proliferation of emigrated cells. This was confirmed by [³H]thymidine uptake: cells emigrated from unstimulated control tissues incorporated 150-250 c.p.m. of [³H]thymidine per 1 × 10⁶ cells; PTX-B stimulated proliferation of these cells in a concentration-dependent manner, with 5 nM PTX-B increasing [³H]thymidine uptake 12 ± 3 times (n = 5). Thus, the increased numbers of 'lymphoblasts' emigrated from PTX-B-stimulated tissue were mostly a consequence of the well-established mitogenic activity of PTX-B ^[12].

Furthermore, cells emigrated from PTX-B-stimulated tissues also upregulated the expression of typical activation antigens, such as CD25, HLA-DR, and CD69; this upregulation occurred predominantly in CD4 T 'lymphoblasts' (Table 2) rather than in 'lymphocytes'. X4 infection of PTX-B-stimulated tissue prevented the upregulation of these activation markers on emigrated cells, whereas their upregulation in R5-infected PTX-B-stimulated tissues was similar to that observed in matched uninfected controls (Table 2). Thus, X4 HIV-1 inhibited PTX-B-induced CD4 T-cell activation while R5 HIV-1 had no effect on it, likely as a consequence of differential signaling by R5 and X4 gp120 Env interacting with CD4 cells and chemokine receptors ^{[20,21,23-25]}.

Table 2

Emigrated cells from infected lymphoid tissue constitute a latently infected, inducible HIV reservoir

Cells emigrated from the tissue did not replicate HIV for at least as long as 3 weeks in culture. Neither the treatment of these cells with mitogenic concentrations of PTX-B induced HIV replication, as evaluated from RT activity (Fig. 3). Therefore, we investigated whether these cells did indeed include latently infected ones or simply were not infected at all. For this purpose, emigrated cells were stimulated with either IL-2 or phytohemagglutinin (PHA). High levels of HIV replication, comparable with those in acutely infected mitogen-activated PBMC, were readily induced (Fig. 3), indicating that a fraction of these cells was latently infected with a replication-competent virus. Of note, lower levels of viral replication were observed in IL-2-stimulated cells emigrated from PTX-B-treated tissue blocks than in cells emigrated from control tissues (Fig. 3).

Fig. 3

To further demonstrate that cells emigrated from infected tissue include infected cells, HIV DNA was quantified with real-time PCR ^[20]. Indeed, between 100 and 800 copies of HIV DNA were found per 1 × 10⁶ cells emigrated from control or PTX-B-treated tissue (Fig. 3). The HIV DNA copy number remained essentially constant in unstimulated emigrated cells for at least a week after their isolation (Fig. 3). PTX-B stimulation of emigrated cells lead to viral DNA dilution due to cell proliferation, while RT activity remained unchanged (Fig. 3). IL-2 stimulation induced viral spreading in emigrated cells: the number of HIV DNA as well as RT activity was much higher than in control or PTX-B treated cells. (Fig. 3b). Similar results were obtained with the genetically modified holotoxin PT-9K/129G (data not shown), further demonstrating the functional homology between PT-9K/129G and PTX-B ^[15-17].

In summary, our study shows that: PTX-B inhibits viral replication in ex vivo-infected human lymphoid tissue, in agreement with what has been reported for isolated cell systems ^[17,18]; cells emigrated from ex vivo HIV-infected human lymphoid tissue include non-productively infected ones, as demonstrated by their HIV DNA content and by the robust viral production induced by PHA or IL-2 (therefore, they represent a new cellular model for the study of viral latency); and PTX-B promotes activation and proliferation of cells emigrated from lymphoid tissue but, unlike PHA or IL-2, does not induce viral replication, thus uncoupling T-cell activation from HIV replication.

The effects of PTX-B and PT-9K/129G on ex vivo-infected human lymphoid tissue described here are consistent with those observed in SIV-infected rhesus macaques treated with the holotoxin PTX ^[26]. In these animals, a single injection of PTX resulted both in a decrease of viral replication in lymphoid tissues and in fewer productively infected cells in peripheral blood ^[26]. Our current findings suggest that the effects of PTX holotoxin in SIV-infected macaques were likely mediated by PTX-B. In support of this interpretation, as we observed in a parallel study, PTX-B inoculation of SCID-hu PBL mice either prevented or curtailed HIV infection and/or virus replication as a function of the schedule of administration and of the dose of PTX-B ^[27].

Although we do not know which molecular mechanisms and signaling pathways are differentially triggered by PTX-B and IL-2 (or PHA) in CD4 T cells emigrated from infected lymphoid tissue, their detailed investigation, although important, is beyond the scope of the present study. It has been reported that in isolated lymphocyte or macrophage cultures PTX-B stimulation modifies the ratio of unspliced to spliced viral transcripts ^[17] and inhibits NF-kB ^[28]. We have not investigated whether PTX-B stimulation induces viral latency. However, we show here that in vitro PTX-B maintains viral latency despite triggering lymphocyte activation and proliferation. In vivo, a reservoir of latently infected CD4 T lymphocytes is established early during primary HIV infection ^[29,30], and viral production from this reservoir can be induced both by T-cell activation and by pro-inflammatory cytokines ^[31-33]. This reservoir represents the major obstacle to the eradication of HIV in vivo, since it is resistant to highly active antiretroviral therapies, and viremia rebounds to high levels after therapy interruption in most individuals ^[34]. The above-described new ex vivo model of latent inducible R5 and X4 HIV-1 infection of CD4 T lymphocytes emigrated from lymphoid tissue may help us to better understand the mechanisms of establishment and maintenance of latent viral reservoirs and to design more efficient antiviral strategies.

In conclusion, PTX-B, and especially its clinically approved equivalent PT-9K/129G, both of which exhibit anti-HIV activities in human lymphoid tissue, induce activation and proliferation of CD4 T cells while maintaining viral latency and thus represent new potential antiviral agents endowed with CD4 T-cell immunostimulating capacity.

Acknowledgments

Sponsorship: This study was supported by Grant n. 40D.5 (to M.A.) of the IV° National Program of Research on AIDS of the Istituto Superiore di Sanità, Rome, Italy. The work of J.-C.G. and L.M. was supported, in part, by the NASA/NIH Center for Three-Dimensional Tissue Culture.