br Discussion Our data demonstrate that the
Discussion Our data demonstrate that the bicyclam AMD3100 strongly interacts with CXCR-4, whereas DS and AR177, two other compounds that also interfere with virus entry into the glutamate receptor antagonist (and interfere with virus binding), do not. AMD3100 is a unique compound in that it interferes with the chemokine receptor CXCR-4. AMD3100 does not appear to interact with CCR-5, or other CC chemokine receptors, as monitored by its inability to prevent the binding of biotinylated MIP-1α or biotinylated RANTES to SUP-T1 cells, MT-4 cells or freshly isolated PBMC (data not shown). Furthermore there is no inhibitory effect of AMD3100 on the replication of SIV or HIV-1 M-tropic viruses in PBMC (Table 1), which all seem to use CCR-5, perhaps in addition to other coreceptors such as CCR-2 or CCR-3 (but certainly not CXCR-4), to enter the target cells (Alkhatib et al., 1996, Choe et al., 1996, Doranz et al., 1996, Feng et al., 1996). The specific interaction of AMD3100 with CXCR-4 explains why the bicyclams are not active against M-tropic HIV strains in human PBMC. DS and AR177 directly interfere with binding of a mAb (9284) to the V3 loop (Schols et al., 1990, Esté et al., 1997a) but the bicyclams do not (De Clercq et al., 1992). However, when the cells are persistently infected with the AMD3100-resistant virus strain, the mAb 9284 does not longer bind to the cells. In contrast this mAb is able to bind to cells persistently infected with the WT strain (De Vreese et al., 1996a). Furthermore, mutations in the gp120 glycoprotein have been described for the AMD3100-, DS- and AR177-resistant virus strains, especially in the V3–V4 region (De Vreese et al., 1996a, Esté et al., 1997a, Esté et al., 1997b), but unlike the DS- and AR177-resistant strains, the AMD3100-resistant strain has up to seven mutations clustered in the V3 loop. These alterations may induce profound changes in the V3 loop, the putative binding site of HIV with the coreceptors and also alter the affinity of the V3 loop for the mAb 9284. Our results demonstrate that mutations in the V3 loop are probably not caused by the direct interaction of AMD3100 with the viral envelope glycoprotein, but more likely emerge following the interaction of AMD3100 with the HIV coreceptor CXCR-4. It took 25 passages in cell culture before the NL4-3 strain became resistant to JM2763 (172-fold less sensitive) (IC50: 127 μg/ml) and 60 passages before resistance to AMD3100 (300-fold less sensitive) (IC50: 546 ng/ml) developed (De Vreese et al., 1996a). It is important to mention that we never obtained complete resistance against AMD3100 with the NL4-3 strain. However, for the HIV-1 T-tropic strain HE only 6 passages were required for JM2763, and 8 passages for AMD3100, to achieve complete resistance (IC50: >250 μg/ml) (De Vreese et al., 1996a). Unlike the NL4-3 strain, the HE strain, may use other coreceptors besides fusin to infect the cells. In support of this assumption, the CC chemokine, RANTES, was found to be active against the HE strain in PHA-activated PBMC, whereas the NL4-3 strain was completely insensitive to RANTES (Schols et al., 1997). The bicyclams are inhibitory to syncytium formation, although their IC50 in these assays are about 10–100 fold higher than their IC50 in the viral replication assays (De Clercq et al., 1994). This can also be explained by their interaction with fusin on the CD4+ target cells or on the persistently HIV-infected cells. The AMD3100-resistant virus is cross-resistant to DS and AR177 (De Vreese et al., 1996a, Esté et al., 1997a), whereas, vice versa, the AR177- and DS-resistant strains are still sensitive to JM2763 and AMD3100 (De Vreese et al., 1996a, Esté et al., 1997a). This suggest that fusion inhibitors may be still active against virus that has become resistant to an adsorption inhibitor. The reverse does not hold, however, as resistance to fusion inhibitors may be accompanied by resistance to virus adsorption inhibitors as well. What remains to be resolved is why RANTES, MIP-1α and MIP-1β, that are active against M-tropic viruses in PM1 cells (a T-cell line) and PBMC (Cocchi et al., 1995) have no anti-HIV activity whatsoever in purified monocytes/macrophages (Deng et al., 1996, Dragic et al., 1996, Schmidtmayerova et al., 1996). In fact, some of these chemokines even enhance viral replication in these cells (Dragic et al., 1996, Schmidtmayerova et al., 1996). Although monocytes express CXCR-4 (McKnight et al., 1997), they cannot be infected with T-tropic viruses such as IIIB or NL4-3 for reasons which are not clear yet.