Viability of the placental histocultures throughout the culture period was evaluated by quantifying total hCG production in histoculture supernatants every 3 to 4 days from day 1 to day 18 or 21. The maximum level of total hCG was observed at day 4 or 7. A decrease was observed between day 7 and day 11 of culture, and the minimum level was reached after day 18. These levels are comparable to those obtained in histocultures from previously reported term placentas 34. Kinetics of hCG production are shown in Figure 1.

After the set-up of the histoculture system and before the start of the infection protocols, the tissue response to an external stimulus such as LPS was evaluated. Placental histocultures were stimulated with 0.1 and 10 μg/ml LPS at day 0, 3 and 6 for 24 h before supernatant collection. Placental histocultures showed a response to this stimulus in a dose-dependent manner secreting high amounts of TNF-α at all the times tested (data not shown).

After viral transduction with or without parasite infection, tissue function of placental histocultures was analyzed by measuring total hCG secretion levels in histoculture supernatants from each experiment. As shown in Figure 2, neither viral nor parasite treatment significantly modified hCG secretion, indicating that the outcome of infection was not due to direct cytotoxicity of the inocula for the histocultures.

The effect of blood trypomastigotes (BT) on HIV-1 replication was assessed on BeWo cells, a model of early trophoblast cells which are the first placental layer in direct contact with maternal blood. Previous data indicated that the HIV-1 R5 (BaL) or X4 (HXB2) pseudotyped reporter virus did not replicate in BeWo cells 35, thus we used only VSV-G pseudotyped HIV-1 reporter virus. Cells were incubated with BT and/or pseudotyped virus, and transduction of the luciferase reporter gene as an indicator of viral replication was evaluated at the end of the experiment. As shown in Figure 3 (left bar), viral replication was decreased by BT (-86%, p < 0.005).

As trypomastigotes shed several soluble factors 36, we wanted to determine whether the trypomastigote supernatant could interfere with HIV-1 replicative cycle, or if an active T. cruzi infection was necessary to achieve the previously described effect. Thus, BeWo cells were incubated with 24 h-supernatants from BT (BTSn) and VSV-G pseudotype virus. Similar effect on luciferase activity as in the case of BT was observed for BTSn (-76%, p < 0.005) (Figure 3, right bar).

Transduction with pseudotyped virus harboring VSV-G, HIV-1 R5 (BaL) or HIV-1 X4 (HXB2) envelope protein were performed on placental histocultures, with or without infections with BT. Results were normalized in each sample by total protein concentration. When placental histocultures were transducted with HXB2 pseudotyped virus, no luciferase activity was detected (data not shown). When BaL pseudotyped virus was used, even at higher doses than VSV-G pseudotyped virus, levels of luciferase activity were lower. However, for both pseudotyped virus luciferase activities were significantly decreased in coinfection with BT (mean ± SD; -90.98% ± 5.83, p < 0.001 for VSV-G and -94% ± 5.02, p < 0.001 for BaL) (Figure 4A).

Purification of BT might carry other components from mouse blood, mainly white cells and platelets, which could interfere with HIV-1 replication. Thus, similar experiments were performed using trypomastigotes purified from Vero cell culture supernatants (CT). Similarly, live CT significantly decreased virus-driven luciferase activity in placental histocultures when they were coinfected with VSV-G pseudotyped HIV-1 reporter virus (mean ± SD; -97.36% ± 0.98, p < 0.001) (Figure 4B).

Additionally, placental explants were incubated with 24 h-supernatants from either BT (BTSn) or CT (CTSn) and VSV-G pseudotype virus. A similar effect on luciferase activity as in the case of BT was observed for BTSn (mean of diminution ± SD; -81.48% ± 8.15, p < 0.001), while CTSn also decreased luciferase activity although at a lower degree (-61.21% ± 1.94, p < 0.001) (Figure 4B).

In an attempt to determine whether changes in the placental microenvironment due to parasite-viral interaction are responsible for inhibiting HIV replication, cytokine/chemokine secretion was measured in histoculture supernatants at day 1 and at day 4 post-infection or coinfection. Results for day 1 presented in Figure 5 demonstrate that T. cruzi acts as a potent inhibitor of IL-6 (p < 0.01), IL-8 (p < 0.05), IP10 (p < 0.01) and MCP-1 (p < 0.02), while no significant changes were observed in RANTES, MIP-1α, MIP-1β, G-CSF, GRO-α, GM-CSF, IFN-γ, IL-10, IL-13, IL-1β, IL-2, IL-4 and IL-5 production (data not shown). The effect of HIV-T. cruzi interaction on cytokine/chemokine secretion seems to be parasite-driven since their levels correlated with those induced by the parasite alone.

When placental histocultures were treated with BTSn, significant decreases only in IL-6 (p < 0.04), IL-8 (p < 0.05), MCP-1 (p < 0.01), GM-CSF (p < 0.05), MIP-1α (p < 0.05), and MIP-1β (p < 0.05) secretion were detected in histoculture supernatants collected at day 1 post-infection or coinfection (Figure 6). Surprisingly, this diminution was detected only in BTSn treated histocultures but no changes were observed in HIV-BTSn treated samples.

In all experiments, the differences seen at day 1 were no longer seen at day 4 post-infection or coinfection.

Term placentas were obtained after programmed cesarean section at the Obstetrics Unit of the Fernández and Ramos Mejía Hospitals in the city of Buenos Aires, in accordance with Argentinean ethics guidelines. This study was approved by the Ethics Committee from the School of Medicine, University of Buenos Aires. Histocultures of chorionic villi were performed as previously described 34 with slight modifications. Briefly, placental villi were isolated, washed extensively with RPMI 1640 (CellGro, USA) and dissected into 2–3 mm blocks. After infection or coinfection, chorionic villi were placed on 1.5 cm² collagen sponge gels (Espongostan, Johnson & Johnson, USA) at medium-air interface into the wells of 6- or 12-well plates (Greiner, Germany) with 3 or 2 ml media per well respectively, at 9 tissue blocks per collagen sponge and per well. Histoculture medium was RPMI 1640 (CellGro) supplemented with 15% heat-inactivated fetal calf serum (FCS, PAA-Bioser, Argentina), 1% penicillin-streptomycin, 0.1% gentamicin, 1% L-glutamine, 1% non-essential amino acids, 1% sodium pyruvate; (Gibco BRL Ltd., USA). Placental histocultures were maintained in 5% CO₂ atmosphere/95% air at 37°C. Each experimental point means a duplicate histoculture well.

Viability of histocultures was monitored by detection of total hCG secretion determined with a chemiluminescence method (Immulite 1000, detection limit 1.1 mIU/ml, Siemens Medical Solutions Diagnostics, USA) in supernatants from both histoculture system setup protocols (from day 1 to day 18) and in infection protocols (day 4 post-infection or coinfection). Tissue responses to LPS were analyzed by incubating tissue fragments with 0.1 and 10 μg/ml LPS (from E. coli serotype 055:B5, Sigma-Aldrich, USA) at day 0, 3 and 6 of histoculture. After 24 h, levels of Tumor Necrosis Factor-alpha (TNF-α) secretion were quantified by ELISA (Peprotech, Mexico) in histoculture supernatants.

The human choriocarcinoma BeWo cell line 54, used as a model for early trophoblast cells 555657, was obtained from the American Type Culture Collection (ATCC # CCL98, Rockville, Md.). These cells were maintained in Dulbecco's Modified Eagle Medium (DMEM, CellGro) containing 25 mM glucose, supplemented with 20% heat-inactivated FCS, 20 mM glutamine, 50 IU/ml penicillin and 50 μg/ml streptomycin, in 5% CO₂ atmosphere/95% air at 37°C.

Luciferase reporter viruses were produced as previously described 35 by transiently cotransfecting (SuperFect; Qiagen, Germany) 293T cells with the proviral pNL-Luc-E^-R⁺ vector 58, which lack the env gene and has the firefly luciferase gene inserted into the nef gene, and the expression vector pCMV harboring the gene coding for either the VSV-G envelope protein or the HIV-1 R5 (BaL) envelope protein 59, or the expression vector pSV harboring the gene coding for HIV-1 X4 (HXB2) envelope protein 60. Supernatants from 293T cells were harvested 72 h after transfection and p24gag levels were measured using a commercial ELISA kit (Murex, UK).

T. cruzi VD strain (isolated from a case of congenital Chagas' disease, lethal for mice, lineage II) was used 61. This subpopulation was maintained by serial passages in 21-day old CF1 mice. Either bloodstream forms (BT) or tissue culture-derived (CT) trypomastigotes were employed for the coinfection assays. BT were collected from blood of T. cruzi infected mice at the peak of parasitemia by cardiac puncture. To enrich blood supernatants with BT, the centrifuged blood was incubated for 1 h at 37°C and the supernatant was collected. Thus, BT were pelleted by centrifugation for 30 min at 10,000 × g, counted in a Neubauer hematocytometer and diluted to 10⁷ BT/ml in BeWo medium or to 4 × 10⁷ BT/ml in histoculture medium for further use in coinfection assays. In order to obtain CT, Vero cell monolayers were allowed to interact with BT in a parasite/cell ratio of 5:1 for 24 h. CT harvested from the second passage in Vero monolayers were pelleted by centrifugation for 30 min at 10,000 × g, counted in a Neubauer hematocytometer and diluted to 4 × 10⁷ CT/ml in histoculture medium for further use in coinfection assays.

In order to obtain parasite supernatants, 10⁷ BT diluted in BeWo medium or 4 × 10⁷ BT or CT diluted in histoculture medium were incubated for 24 h at 37°C in 5% CO₂. To remove parasites and cellular debris, both parasite suspensions were pelleted as described above and the supernatants were filtered through a 0.22 μm pore-size filter. Filtrated aliquots were stored at -80°C until use for coinfection assays.

BeWo cells were seeded in 96-well plates (2 × 10⁴cells per well) 24 h before infection and then incubated with VSV-G pseudotyped HIV-1 (100 ng of p24 per well) and 2 × 10⁵ BT or 24 h-supernatant of BT (BTSn) overnight at 37°C in 5% CO₂. Controls included transduction with only the pseudotyped virus or Δenv pseudotype, infection with parasites or treatment with parasite supernatants, and also mock infected cells with culture medium. After culture for an additional 72 h, 100 μl of luciferase lysis buffer (Promega) per well was added and luciferase activity as an indicator of viral replication was measured in 10 μl of lysate with a luminometer (Veritas), using the commercially available substrate; data were expressed as RLU/sec.

After dissection, 18 blocks of placental villi were placed in 24-well plates and transducted overnight with BaL (250 ng p24/placental block), HXB2 (250 ng p24/placental block) or VSV-G pseudotyped HIV-1 (100 ng p24/placental block) and infected with BT or CT (10⁶ parasites/placental block), or parasite supernatants. Controls included transduction with only the pseudotyped virus or Δenv pseudotype, infection with parasites or treatment with parasite supernatants, and also mock infected histocultures with culture medium. The following day, placental blocks were washed 6 times in 6-well plates with PBS 1× (Gibco BRL Ltd.) using a cell strainer (BD Biosciences, USA), placed on collagen sponges and cultured as described above for an additional 72 h. Protocols of overnight infection and then supernatant collection were also performed.

For further cytokine/chemokines quantification, at the end of each experiment, supernatants were collected, clarified by centrifugation at 1,000 × g for 10 minutes, filtered (0,22 μm), aliquoted and stored at -80°C. Placental fragments were collected and preserved at -80°C until homogenization.

For luciferase activity quantification, placental fragments were homogenized in 500 μl of luciferase lysis buffer (Promega, USA) with an Ultra-turrax homogenizer (IKA, USA). Luciferase activity was measured at day 4 post-infection or coinfection in 20 μl of lysate with a luminometer (Veritas, USA), using a commercially available substrate (Dual-Luciferase Reporter Assay System, Promega), and expressed as relative light units (RLU). Results were normalized to total protein concentration measured on lysates from each sample using a Micro BCA™ Protein Assay Kit (Pierce, USA). Final data were expressed as RLU/μg prot (RLU/prot).

Supernatants of stimulated histocultures were diluted with 10% FCS in RPMI and used for the simultaneous determination of 29 cytokines and chemokines with Luminex technology as previously described 6263. The coated bead/biotinylated antibody combinations used were: G-CSF (LINCOplex human G-CSF, Linco Research, St. Charles. MO), GM-CSF (Beadlyte human GM-CSF, Upstate USA, Charlottesville, VA), GRO-α (Beadlyte human GRO-α, Upstate), IFN-α (anti-human IFN-α clones MMHA-11 and MMHA-2, PBL Biomedical Laboratories, Piscataway, NJ), IFN-γ (Beadlyte primate IFN-γ, Upstate), IL-1β (Monkey IL-1β ELISPOT reagents, U-Cytech), IL-1Ra (Fluorokine MAP human IL-1Ra/IL-1F3, R&D System, Minneapolis, MN), IL-2 (Beadlyte primate IL-2, Upstate), IL-4 (LINCOplex human IL-4, Linco), IL-5 (LINCOplex human IL-5, Linco), IL-6 (LINCOplex human IL-6, Linco), IL-7 (anti-human IL-7 clone 7417 and polyclonal anti-human IL-7, R&D), IL-8 (Beadlyte human IL-8, Upstate), IL-9 (Beadlyte human IL-9, Upstate), IL-10 (anti-human IL-10 clones BN-10 and QS-10, Cell Sciences Inc., Canton, MA), IL12(p40) (anti-human IL-12 clones IL-12I and IL-12II, Mabtech Inc., Mariemont, OH), IL-12(p70) (anti-human IL-12 p70 clone 20C2, Endogen, and IL-12II, Mabtech), IL-13 (Beadlyte human IL-13, Upstate), IL-15 (anti-human IL-15 clone 34505 and polyclonal anti-human IL-15, R&D), IL-17 (Human IL-17, Biosource International, Camarillo, CA), IL-18 (anti-human IL-18 clones 125-2H and 159-12B, MBL International, Woburn, MA), IP-10 (anti-human IP-10 clone 33036 and polyclonal anti-human IP-10, R&D), MCP-1 (Human MCP-1, Biosource), MIP-1α (Human MIP-1α, Biosource), MIP-1β (Human MIP-1β, Biosource), RANTES (Beadlyte human RANTES, Upstate), sCD40L (Fluorokine MAP human sCD40L, R&D System), TNF-α (Beadlyte human TNF-α, Upstate), and TNF-β (Beadlyte human TNF-β, Upstate). Cytokine concentrations were determined using human cytokines (Upstate) as standards and the Masterplex QT software from Mirabio.

Results of luciferase activity and cytokine/chemokine production from each group (infected or coinfected tissue/cells) are presented as mean ± SD. Comparison of their distributions between 2 groups for luciferase activity was performed by a t-Student test and distribution between more than 2 groups for cytokine/chemokine expression was performed by a non-parametric Friedman test and a post-test Dunns, using the Graph Pad Prism 4 software.