Background: The calcium-dependent lectin, DC-SIGN, binds to HIV (and SIV) gp120 and mediates the binding and transfer of HIV from monocyte-derived dendritic cells (MDDCs) to permissive T-cells. However, it has been recently reported that DC-SIGN binding to HIV gp120 may be carbohydrate independent. Here, we formally demonstrate that gp120 binding to DC-SIGN and MDDCs is largely if not wholly carbohydrate dependent, and report progress towards the fine definition of the envelope determinants of DC-SIGN binding.

Methods: Enzymatic deglycosylation in concert with a differential lectin binding assay was used to formally prove that gp120 binding to DC-SIGN was high mannose dependent. Mutational analyses in conjunction with competition studies with 2G12 and cyanovirin were used to further delineate the N-linked glycosylation sites responsible for Dc-SIGN binding.

Results: EndoH treatment of gp120-Fc under conditions that maintained wild-type CD4 binding, and the full complement of complex glycans (as shown by differential lectin binding), abrogated binding to DC-SIGN expressing cell lines as well as MDDCs. Mutational analysis indicated that no single glycosylation site affected the ability of gp120-Fc to bind DC-SIGN. To further guide our efforts in mapping the DC-SGN binding sites on gp120, we used 2 well-characterized HIV inhibitory agents (2G12 Mab and cyanovirin) that bind to high mannose sugars on gp120. We showed that 2G12 and DC-SIGN bound to non-overlapping sites in gp120 because 1) 2G12 did not block soluble gp120 or virion binding to DC-SIGN; 2) 2G12 bound to gp120-Fc that was pre-bound to cell surface DC-SIGN; and 3) gp120-Fc mutants that lack glycosylation sites involved in 2G12’s epitope were also fully capable of binding DC-SIGN. These data were substantiated by the inability of cyanovirin to block gp120-Fc binding to DC-SIGN. Cyanovirin has been shown to effectively compete for 2G12 binding to gp120. Indeed, high concentrations of cyanovirin dramatically enhanced gp120-Fc binding to cell surfaces in the presence or absence of DC-SIGN. We provide evidence that this enhancement may be due to cyanovirn’s ability to bridge gp120 to mannosylated cell surface proteins.

Conclusions: HIV envelope (gp120) binding to DC-SIGN and primary dendritic cells is carbohydrate dependent but does not involve 2G12 or cyanovirin binding sites. These results have implications for on-going efforts to finely map the DC-SIGN binding sites on gp120.