397
Tissue-specific Adaptive Changes in R5X4 HIV-1 Env Variants Compartmentalized in Brain and Lymphoid Tissues of Individuals with AIDS
Lachlan Gray*1,2, M Churchill1, D Cowley1,3, J Sterjovski1,3, A Ellett1, N Saksena4, P Poumbourios1, S Wesselingh1,2,3, D Gabuzda5,6, and P Gorry1,2,3
1Burnet Inst, Melbourne, Australia; 2Univ of Melbourne, Australia; 3Monash Univ, Melbourne, Australia; 4Westmead Millenium Inst, Australia; 5Dana-Farber Cancer Inst, Boston, MA, US; and 6Harvard Med Sch, Boston, MA, US
Background: Most neurotropic HIV-1 strains are
CCR5-restricted, but R5X4 variants have been identified infrequently in brain.
Methods: HIV-1 Env were cloned from R5X4 primary
viruses isolated from brain (n = 6) and spleen (n = 6) of subject
MACS1. Single R5X4 Env cloned from brain and blood of another subject were
included (aBR01, aBL01). Env were sequenced and structural modeling was
performed to analyze amino acid variants. Env fusogenicity and CD4/co-receptor
dependence was tested in fusion assays with wild type and mutant co-receptors.
Results: MACS1 brain (M1br) and the brain-derived
aBR01 Env were more fusogenic than Env from matched spleen or blood in cells
expressing CD4/CCR5, whereas MACS1 spleen (M1sp) and the blood-derived aBL01
Env were more fusogenic than Env from matched brain in cells expressing
CD4/CXCR4. Compared to M1sp and aBL01 Env, 6 of 6 M1br and the aBR01 Env had
reduced dependence on CCR5 and CD4 levels in fusion assays with cells
expressing CD4/CCR5. Compared to M1br and aBR01 Env, a subset of M1sp and the
aBL01 Env had reduced dependence on CXCR4 and CD4 levels in fusion assays with
cells expressing CD4/CXCR4. Studies with co-receptor mutants showed that,
compared to M1sp Env, M1br Env had reduced dependence on residues in the CCR5
N-terminus (Y15), ECL1 (H88), and ECL3 regions (E262, F264), but increased
dependence on H181 in the ECL2 region for CCR5-mediated fusion. Compared to
M1br Env, M1sp Env had reduced dependence on residues 4-36 in the CXCR4
N-terminus and R183 in the ECL2 region for CXCR4-mediated fusion. Sequence
analysis identified R306 in the V3 loop of 6 of 6 M1sp Env and S306 in 6 of 6
M1br Env, and a unique Ser variant at the highly conserved P369 position in the
gp120 CD4 binding site (CD4bs) in 6 of 6 M1sp and 0 of 6 M1br Env. Structural
models predict S369 may form additional H-bonds with K421 in the gp120 b-19 strand.
Conclusions: Enhanced CCR5 or CXCR4 usage by brain-
or spleen/blood-derived R5X4 Env, respectively, was associated with altered
dependence on co-receptor and CD4 levels, and altered mode of co-receptor
usage. The presence of R306 or S306 in V3 may contribute to preferential co-receptor
usage via charge alterations. S369 may reduce CD4 dependence of M1sp Env by
increasing stability or exposure of the CD4bs via intramolecular interactions.
Thus, tissue-specific adaptive changes may enhance the tropism of
compartmentalized R5X4 strains for cells expressing CCR5 in brain and CXCR4 in
lymphoid tissues.
|