Kinetic studies of HIV-1 and HIV-2 envelope glycoprotein-mediated fusion

Abstract

Background: HIV envelope glycoprotein (Env)-mediated fusion is driven by the concerted coalescence of the HIV gp41 N-helical and C-helical regions, which results in the formation of 6 helix bundles. Kinetics of HIV Env-mediated fusion is an important determinant of sensitivity to entry inhibitors and antibodies. However, the parameters that govern the HIV Env fusion cascade have yet to be fully elucidated. We address this issue by comparing the kinetics HIV-1IIIB Env with those mediated by HIV-2 from two strains with different affinities for CD4 and CXCR4.

Results: HIV-1 and HIV-2 Env-mediated cell fusion occurred with half times of about 60 and 30 min, respectively. Binding experiments of soluble HIV gp120 proteins to CD4 and co-receptor did not correlate with the differences in kinetics of fusion mediated by the three different HIV Envs. However, escape from inhibition by reagents that block gp120-CD4 binding, CD4-induced CXCR4 binding and 6-helix bundle formation, respectively, indicated large difference between HIV-1 and HIV-2 envelope glycoproteins in their CD4-induced rates of engagement with CXCR4.

Conclusion: The HIV-2 Env proteins studied here exhibited a significantly reduced window of time between the engagement of gp120 with CD4 and exposure of the CXCR4 binding site on gp120 as compared with HIV-1IIIB Env. The efficiency with which HIV-2 Env undergoes this CD4-induced conformational change is the major cause of the relatively rapid rate of HIV-2 Env mediated-fusion.

Figure 1

Kinetics of HIV-1 and HIV-2 Env-mediated fusion. HIV-1_IIIB (squares), HIV-2_SBL (triangles), and HIV-2_ROD (circles) Env proteins were expressed in HeLa cells using vaccinia recombinants as described in Materials and Methods. Target SupT1 cells, labeled with calcein, were added to the plated HeLa cells, labeled with CMTMR, at various times during a two hour period at 37°C. The cells were then examined by fluorescence microscopy for dye transfer indicating cell-cell fusion. Lines represent fits to the sigmoidal equation f = a/(1-exp [-b(t - t_1/2)]) using Sigmaplot (SPSS, Chicago). Values of time for half maximal fusion (t_1/2) are 63 ± 6, 28 ± 2 and 23 ± 4 minutes for HIV-1_IIIB, HIV-2_SBL and HIV-2_ROD, respectively.

Figure 2

Binding of soluble gp120 to cellular CD4 orCXCR4. Receptor binding efficiencies of gp120s were determined using a cell surface-binding assay in which bound protein was detected by Western blot analysis as described in Materials and Methods. Binding to CXCR4 was performed in the presence of sCD4 to expose the coreceptor binding site. HIV-1_IIIB and HIV-2_SBL gp120 exhibited relatively high CD4 binding efficiencies but weak CXCR4 binding. HIV-2_ROD gp120 exhibited relatively weak binding to CD4 and relatively high binding to CXCR4. Numbers indicate mean band intensity following subtraction of background pcDNA3 lane intensity.

Figure 3

Time course for escape from HIV-2_SBLEnv-mediated fusion inhibition. Fusion was measured as described in the legend to figure 1. Leu3A (3 μg/ml, triangles), AMD3100 (40 μM, circles) and SIV C34 (2 μM, squares) were added at different times following co-culture at 37°C and residual fusion was measured following their time of addition. The curves were fitted by the sigmoidal function given in the legend to figure 1; values of time for half maximal fusion (t_1/2) are 26.9 ± 4.1, 24.8 ± 6.4 and 26.4 ± 2.3 minutes for Leu3A (green), AMD3100 (red) and C34 (blue), respectively.

Figure 4

Sequence comparison of HIV-1 and HIV-2 Envs. The HIV-1 subtype B infectious molecular clone HXB2 Env gp41 sequence (Database accession number K03455) is aligned to the HIV-2 ROD (M15390) and ISY (J04498) Env gp41 sequences. Amino Acid sites conserved in all 3 sequences are shaded black, and those conserved in 2 of the 3 are shaded grey. Numbering is based on the HXB2 amino acid sequence. The sequences corresponding approximately to the different regions of HIV-1 gp41 have been highlighted as follows: Membrane Anchor (180–203), red; Lentivirus Lytic Peptide-3 (268–286), green; Lentivirus Lytic Peptide-2 (287–313), cyan; Lentivirus Lytic Peptide-1 (326–354), turquoise. Note the lack of homology in LLP-2 and 3 sequences, whereas LLP-1 (337–354) appears to be very homologous between the three strains.