Recent evidence has suggested that plasma membrane sphingolipids and cholesterol spontaneously coalesce into raft-like microdomains and that specific proteins, including CD4 and some other T-cell signaling molecules, sequester into these rafts. In agreement with these results, we found that CD4 and the associated Lck tyrosine kinase of peripheral blood mononuclear cells and H9 leukemic T cells were selectively and highly enriched in a low-density lipid fraction that was resistant at 0 degrees C to the neutral detergent Triton X-100 but was disrupted by extraction of cholesterol with filipin or methyl-beta-cyclodextrin. In contrast, the CXCR4 chemokine receptor, a coreceptor for X4 strains of human immunodeficiency virus type 1 (HIV-1), was almost completely excluded from the detergent-resistant raft fraction. Accordingly, as determined by immunofluorescence with confocal microscopy, CD4 and CXCR4 did not coaggregate into antibody-induced cell surface patches or into patches of CXCR4 that formed naturally at the ruffled edges of adherent cells. The CXCR4 fluorescent patches were extracted with cold 1% Triton X-100, whereas the CD4 patches were resistant. In stringent support of these data, CD4 colocalized with patches of cholera toxin bound to the raft-associated sphingoglycolipid GM1, whereas CXCR4 did not. Addition of the CXCR4-activating chemokine SDF-1 alpha did not induce CXCR4 movement into rafts. Moreover, binding of purified monomeric gp120 envelope glycoproteins from strains of HIV-1 that use this coreceptor did not stimulate detectable redistributions of CD4 or CXCR4 between their separate membrane domains. However, adsorption of multivalent gp120-containing HIV-1 virion particles appeared to destabilize the local CD4-containing rafts. Indeed, adsorbed HIV-1 virions were detected by immunofluorescence microscopy and were almost all situated in nonraft regions of the cell surface. We conclude that HIV-1 initially binds to CD4 in a raft domain and that its secondary associations with CXCR4 require shifts of proteins and associated lipids away from their preferred lipid microenvironments. Our evidence suggests that these changes in protein-lipid interactions destabilize the plasma membrane microenvironment underlying the virus by at least several kilocalories per mole, and we propose that this makes an important contribution to fusion of the viral and cellular membranes during infection. Thus, binding of HIV-1 may be favored by the presence of CD4 in rafts, but the rafts may then disperse prior to the membrane fusion reaction.