The present report shows that CCR5 antagonists and fusion inhibitors block, but with a moderate efficacy, HIV-1 cell-to-cell transmission during the contact between R5 or R5X4 infected PBMCs in contact with a polarized monolayer of BeWo cells. The CXCR4 antagonist evaluated in the present study also potently inhibited HIV-1 transmission from X4-viruses infected PBMCs interacting with BeWo cells. All these drugs were efficient with IC50 in the range of 4 to 20,000 nM. The present study also uncovered an intriguing feature of T20 that we found to increase the transmission of X4 viruses in our in vitro model of interaction between BeWo/PBMC-HIV-1X4 that mimics the human trophoblast barrier.
Earlier studies have shown that cell-to-cell HIV infection is more efficient than cell-free virus infection of permissive cells . Thus, our data on increased IC50s of T20, TAK779, SCH-350581 to block R5 and R5X4 HIV-1 cell-to-cell transmission as compared to cell-free PBMC infection is consistent with these observations.
The two CCR5 antagonists failed to inhibit PBMC infection with the dualtropic isolate A204, but were efficient, though at high concentrations, to block cell-to-cell viral transmission of the same isolate. This difference could be explained by the homogeneity of BeWo cells, as compared to the mixture of cell populations with different properties present in PBMCs preparation. It has been recently shown a preferential use of CXCR4 by R5X4 HIV-1 isolates in primary lymphocytes . In the present study, we characterized by a functional assay , the coreceptor usage of the clinical isolate A204. Of the 20 clones we analyzed, all were able to use both CCR5 and CXCR4 to infect U87.CD4-CXCR4 or U87.CD4-CCR5 (data not shown).
Another feature of the two CCR5 antagonists is the difference in their distinct efficacies to block cell-free HIV-1 PBMC infection and cell-to-cell transmission of the virus in our system. The IC50 for SCH-350851 necessary to block BeWo/PBMC-HIV-1 Ba-L passage is 18 to 22 folds higher than that needed to inhibit HIV-1 Ba-L/PBMC infection. On the contrary, for TAK779, only 1 to 1.5 times more drugs were needed to achieve 50% inhibition of BeWo/PBMC-HIV-1 Ba-L passage. The difference could be due to particular physico-chemical properties of the molecules, or to a subtle difference in the mode of action of the two molecules. It has been shown that both drugs inhibit infection of PBMCs by R5 viruses by binding to various amino acids residues of the transmembrane domains 1, 2, 3, 5 and 7 of the CCR5 molecule, preventing viral entry [39, 40]. Given these latter observations, the difference in activity we observed between the two CCR5 antagonists should lay in their intrinsic physico-chemical properties rather than the targets amino-acids on the CCR5 molecule.
In line with published data obtained with different infection systems, our data here show that the prototype CXCR4 antagonist AMD3100 can also efficiently inhibit the viral passage across the BeWo cell monolayer in contact with PBMCs infected by X4 or R5X4 HIV-1. Given the progressive emergence of X4 viruses in the course of AIDS disease even under HAART , the availability of a potent CXCR4 antagonist as part of salvage therapy is of utmost interest. However, AMD3100 does not seem to be the best candidate for this purpose because of its paucity in reducing plasma viral load in a clinical trial . A concern also remains on the effects of AMD3100 on haematopoiesis, as this molecule has been shown to be a powerful and fast mobilizer of CD34+ progenitor cells, beside its potential anti X4-HIV-1 activity .
The mechanism by which T20 blocks HIV-1/target cell membrane fusion has been extensively explored [44, 45]. The general consensus is that T20 (and C34) are effective once HIV-1 gp120-CD4 has taken place, inducing a conformational change of the heterodimer gp120-gp41 . This conformational change allows, to its turn, the exposition of otherwise masked regions of gp41, especially the N- and C-terminal Heptad repeats (NHR and CHR, respectively) domains [36, 46]. This so-called pre-hairpin intermediate is stable enough to be targeted by antibodies and inhibitors . Instead of the normal interaction between N- and C-terminal HRs to yield a six-helix bundle for membrane fusion and viral genetic material transfer into the host cell, T20 (and C34) binds to the NHR, preventing the formation of the six helix bundle and thus membrane fusion. Our present results with HIV-1 Ba-L and HIV-1 A204 are consistent with that model for BeWo-CD4 positive cells, as a dose-dependent inhibition of viral entry was observed, though with IC50s that are 1,000 fold higher than those reported for classical HIV-1/PBMC infection. Half maximal inhibitory concentrations (3–5 μM for CD4 negative target cells) remain, however, in the range of T20 pharmacokinetics values with standard subcutaneous dose of 90 mg twice daily . The fact that T20 was efficient in preventing R5 and R5X4 passage even if the target BeWo cells were CD4 negative suggests that another mechanism of inhibition distinct from the one outlined above might be involved in the context of HIV-1 cell-to-cell transmission. This alternative mechanism of T20 action during HIV-1 R5 and R5X4 infected PBMC and CD4 negative BeWo cells, might imply other receptors like Galactosyl Ceramide, expressed on the surface of these cells [49, 50]. Although efficient, though at high concentrations, on R5 and R5X4 viruses, T20 was not effective in blocking viral passage in X4 HIV-1 infected PBMC in interaction with BeWo cells. Rather, T20 (and C34), enhanced it.
This enhancement of X4 HIV-1 passage by T20 and C34 following cell-to-cell contact was unexpected. This enhancement is CD4 independent. In a recent work by Blanco and colleagues , a high level of coreceptor-independent HIV transfer was observed between primary CD4+ T-cells. This transfer was enhanced up to 300 folds by C34 but also by the CXCR4 antagonist AMD3100. The enhancement was only observed with the X4 virus NL4-3. Our present observations are different in the way that AMD3100 blocked viral passage across the BeWo cell monolayer in contact with X4 HIV-1 infected PBMCs. We cannot, however, exclude that AMD3100 could also increase X4 HIV-1 passage in our model at higher doses (10 μg/ml) like in the work reported by Blanco and colleagues. Also, entry enhancement we observed is coreceptor dependent and CD4-independent, the exact opposite of transfer enhancement reported by Blanco and colleagues. It has also been shown that X4 virus were more sensitive to inhibition by T20 and C34 in cell-free PBMC infection . This higher sensitivity of X4 virus to fusion inhibitors is thought to lay, at least in part, to the direct interaction of T20, but not C34, with the CXCR4 molecule on the cell surface . This "chemokine" aspect of T20 needs further investigation, complementary to published data .
From our results, we cannot exclude a direct interaction of T20 with CXCR4 present on target BeWo cells because the enhancement was only observed with X4 virus and could be inhibited in the presence of AMD3100 (not shown). This entry enhancement was not due to the polarized state of BeWo cells, as we had similar observations with non-polarized cells on slides. Rather, T20 (and C34), could first attract effector cells infected with an X4 HIV-1 towards target cells, followed by an increased of cell-membrane fusion. It has been reported that T20 can act as a chemo attractant of monocytes, but not of lymphocytes . This observation partly justifies our experiment with purified CD4 positive T cells.
Interestingly, when we used these purified CD4 positive T cells (instead of PBMCs) infected with HIV-1/LAI as effector cells in our system, T20 was found to be highly potent in inhibiting viral passage across the BeWo monolayer (Fig. 8). The enhancement of viral passage across the BeWo monolayer by T20 could thus be a result of direct interaction between X4 HIV-1 infected CD4+ T cells, another cell population (CD4 negative) within the PBMCs and the BeWo cells. Alternatively, enhancement of X4 HIV-1 passage by T20 and C34 could be due to the synergistic effect of T20 (and C34) and a soluble factor secreted par CD4 negative cells present in the PBMCs. In our last experiment (9), 24h-cultures supernatants from CD4 negative cells were added during the interaction between CD4 positive T cells infected with HIV-1/LAI and CD4 negative BeWo cells. In these conditions, the enhancing effect of T20 was restored, suggesting that a soluble factor secreted by CD4 negative cells contributes to the increase of X4 HIV-1 passage in the presence of T20. These findings need, however, further investigations.
The precise mechanism by which the two C-terminal Heptad Repeat peptides (T20 and C34) enhanced X4 HIV-1 passage across the BeWo cell monolayer remains to be elucidated. However, this observation is of real concern, if confirm by others, with regards to HIV-1 transmission at the level of epithelial surfaces, during sexual or mother-to-child transmission for example, in patients under Enfuvirtide therapy. Other abnormal peculiarities of HIV-1 in the presence of T20 have been reported. Some HIV-1 strains whose abnormal envelope folding, leading to viruses unable to properly matured, were rescued by T20 addition .
In the specific case of HIV-1 PMTCT, the use of Enfuvirtide is very limited to date. Recently however, studies reported the use of T20, on a backbone of combined therapies [56–58]. The study by Cohan and colleagues reported the case of a mother infected with a multidrug resistant HIV-1 and who transmitted the virus to her child, despite plasma viral load suppression by T20 treatment. The phenotype of the transmitted virus was not discussed, but the eight clones randomly selected and sequenced in gp41 HR1 and HR2 were all T20 sensitive. The study by Brennan-Benson and co-workers, reported on two cases of pregnant women infected by multidrug resistant HIV-1, with the delivery of two uninfected children. The plasma viral load was fully suppressed for one case and lowered down to 300 copies/ml for the other. In this study too, the viral phenotypes were not discussed. The other interesting observation from this study is the lack of transplacental passage of T20. Thus, from these studies, it is difficult to conclude on the potency of T20, with a weak if any, transplacental transfer, in preventing HIV-1 mother-to-child transmission.