HIV-2 interaction with cell coreceptors: amino acids within the V1/V2 region of viral envelope are determinant for CCR8, CCR5 and CXCR4 usage

Abstract

Background: Human immunodeficiency virus 1 and 2 (HIV-1 and HIV-2) use cellular receptors in distinct ways. Besides a more promiscuous usage of coreceptors by HIV-2 and a more frequent detection of CD4-independent HIV-2 isolates, we have previously identified two HIV-2 isolates (HIV-2MIC97 and HIV-2MJC97) that do not use the two major HIV coreceptors: CCR5 and CXCR4. All these features suggest that in HIV-2 the Env glycoprotein subunits may have a different structural organization enabling distinct - although probably less efficient - interactions with cellular receptors.

Results: By infectivity assays using GHOST cell line expressing CD4 and CCR8 and blocking experiments using CCR8-specific ligand, I-309, we show that efficient replication of HIV-2MIC97 and HIV-2MJC97 requires the presence of CCR8 at plasma cell membrane. Additionally, we disclosed the determinants of chemokine receptor usage at the molecular level, and deciphered the amino acids involved in the usage of CCR8 (R8 phenotype) and in the switch from CCR8 to CCR5 or to CCR5/CXCR4 usage (R5 or R5X4 phenotype). The data obtained from site-directed mutagenesis clearly indicates that the main genetic determinants of coreceptor tropism are located within the V1/V2 region of Env surface glycoprotein of these two viruses.

Conclusions: We conclude that a viral population able to use CCR8 and unable to infect CCR5 or CXCR4-positive cells, may exist in some HIV-2 infected individuals during an undefined time period, in the course of the asymptomatic stage of infection. This suggests that in vivo alternate molecules might contribute to HIV infection of natural target cells, at least under certain circumstances. Furthermore we provide direct and unequivocal evidence that the usage of CCR8 and the switch from R8 to R5 or R5X4 phenotype is determined by amino acids located in the base and tip of V1 and V2 loops of HIV-2 Env surface glycoprotein.

Figure 1

HIV-2 _MIC97 and HIV-2 _MJC97 use CCR8 as coreceptor to infect GHOST-CD4 cells and PBMC. PBMC and GHOST-CD4 cell lines expressing different coreceptors were exposed to 100 TCID₅₀ of each virus; viral replication was quantified by RT activity in culture supernatants during a 12-day period after infection and the highest value of RT activity observed during this time period was used. Results are expressed as the mean of three independent experiments performed in duplicate. Error bars represent the standard error of the mean. A star (*) indicates statistical significant difference (p < 0.001) between the means of peak RT activity measured in culture supernatants of GHOST-CD4/CCR8 inoculated with HIV-2_MIC97 and HIV-2_MJC97, compared with GHOST-CD4/CCR5, GHOST-CD4/CXCR4, GHOST-CD4/CX3CR1, and GHOST-CD4/Hi5 inoculated with the same viruses.

Figure 2

Specific inhibition of HIV-2 _MIC97 and HIV-2 _MJC97 by I-309. GHOST-CD4/CCR8 cell line and CD8-depleted PBMCs were inoculated with 100 TCID₅₀ of HIV-2_MIC97 and HIV-2_MJC97 either in the presence or absence of CCR8 ligand, I-309. HIV-2_ROD and HIV-1_Ba-L were also included as controls. The data are expressed as the mean of three independent experiments performed in duplicate. Error bars represent the standard error of the mean.

Figure 3

Amino acid residues changed in V1/V2 region of env glycoprotein by site-directed mutagenesis. Amino acids are denoted by single-letter code. (A) For a better localization of mutated amino acids, the sequence alignment between HIV-2_ROD (GenBank accession number: M15390) and HIV-2_MJC97 (GenBank accession number: EU021092) was included. The red boxes indicate the conserved regions between HIV-2_ROD and HIV-2_MJC97 amino acids sequences. (B) The first set of mutants (MJC97mt1 to MJC97mt7) was obtained by sequential mutagenesis starting in the non-mutated recombinant virus, MJC97wt. The second set of mutants (MJC97mt5′ to MJC97mt7′) was derived from the V1/V2 of MJC97mt4. For each sequential mutant, underlined red letters represents the newly changed amino acids residues, while the non-underlined red letters denote mutations that were previously added. Amino acids residues (in panel A and B) were numbered according to HIV-2_ROD (GenBank accession number: M15390) or HIV-2_MJC97 sequence (GenBank accession number: EU021092).

Figure 4

Effect of sequential mutations in V1/V2 region of HIV-2 _MJC97 env glycoprotein on coreceptor usage. Stocks of each mutated virus (the details of these mutants are described in Figure 3) were used to infect PHA-activated PBMCs and GHOST-CD4 cell lines individually expressing CCR8, CCR5, and CXCR4 coreceptors. (A) Viral replication was followed-up for 12 days by assessing RT activity in culture supernatants of infected cells. The highest value of RT activity observed during this time period was used. A star (*) indicates statistical significant difference (p < 0.001) between the means of peak RT activity measured in culture supernatants of GHOST-CD4/CCR8, GHOST-CD4/CCR5 and GHOST-CD4/CXCR4 inoculated with MJC97mt7 compared to the same cells inoculated with MJC97wt. Conversely, a double star (**) denotes statistical significant difference (p < 0.001) between the means of peak RT activity measured in culture supernatants of GHOST-CD4/CCR8, GHOST-CD4/CCR5 and GHOST-CD4/CXCR4 inoculated with MJC97mt7′ compared to the same cells inoculated with MJC97wt. The strains HIV-2_ALI (R5), HIV-2_ROD (R5X4) and HIV-2_MJC97 (R8) were used as controls. Replication kinetics of MJC97wt was compared to mutant viruses that switch from CCR8 usage to CCR5/CXCR4 (MJC97mt7) or to CCR5 (MJC97mt7′); HIV-2ROD and HIV-2ALI strains were also included as controls. The replication kinetics, assessed by RT activity in culture supernatants, was followed up during 21 days and was performed in PHA-activated PBMCs (B), GHOST-CD4/CCR8 (C), GHOST-CD4/CCR5 (D) and GHOST-CD4/CXCR4 (E). In all experiments, results are expressed as the mean of three independent experiments performed in duplicate. Error bars represent the standard error of the mean.

Figure 5

Location of amino acids residues involved in coreceptor usage. Schematic representation of the envelope SU glycoprotein of HIV-2_MJC97 putative secondary structure spanning from C1 to V3 regions. The amino acid sequence of HIV-2_MJC97 (MJC97wt; R8) are represented in black; the mutated amino acids present in MJC97mt7 (R5X4; panel A) and MJC97mt7′ (R5; panel B) are represented in red. Amino acids are denoted by single-letter code. The underlined amino acids represent potential glycosylation sites linked to asparagine (N) as defined using the LANL N-glycosite program (http://www.hiv.lanl.gov/content/sequence/GLYCOSITE/glycosite.html). Amino acids residues were numbered according to HIV-2_MJC97 sequence (GenBank accession number: EU021092). This schematic representation was based on previous data regarding epitope and cysteine loops mapping [58,59].

Figure 6

Amino acid sequences of mutants targeting the tip of V1 region and the base of V2 loop. (A) For a better localization of mutated amino acids, the sequence alignment between HIV-2_ROD (GenBank accession number: M15390) and HIV-2_MJC97 (GenBank accession number: EU021092) was included. The red boxes indicate the conserved regions between HIV-2_ROD and HIV-2_MJC97 amino acids sequences. (B) The tip of V1 of MJC97wt has the sequon ISTTDYSL (amino acids residues 113–120) that was changed to IPTDQEQE (MJC97mtV1) or PGSTLKPL (MJC97mtV1′). The MJC97mtV1V2 was obtained by replacing the sequon PTNET (amino acids residues 172–176) in the base of V2 loop of MJC97mtV1 by TNNES; MJC97V1V2′ was originated by changing the referred sequon of MJC97mtV1′ by PFNTT. Amino acids residues (in panel A and B) were numbered according to HIV-2_ROD (GenBank accession number: M15390) or HIV-2_MJC97 sequence (GenBank accession number: EU021092).

Figure 7

Coreceptor usage of mutants targeting the tip of V1 region and the base of V2 loop. Recombinant viruses with mutations targeting the tip of V1 region (MJC97mtV1 and MJC97mtV1′) or with additional mutations in the base of V2 loop (MJC97mtV1V2 and MJC97mtV1V2′). The details of these two sets of mutants are described in Figure 6. Viral replication was followed-up for 12 days by assessing RT activity in culture supernatants of infected cells. The highest value of RT activity observed during this time period was used. Results are expressed as the mean of three independent experiments performed in duplicate. Error bars represent the standard error of the mean. A star (*) indicates statistical significant difference (p < 0.001) between the means of peak RT activity measured in culture supernatants of GHOST-CD4/CCR8, GHOST-CD4/CCR5 and GHOST-CD4/CXCR4 inoculated with MJC97mt6 and MJC97mtV1 or MJC97mt7 and MJC97mtV1V2. A double star (**) indicates statistical significant difference (p < 0.001) between the means of peak RT activity measured in culture supernatants of GHOST-CD4/CCR8 and GHOST-CD4/CCR5 inoculated with MJC97mt6′ and MJC97mtV1′ or MJC97mt7′ and MJC97mtV1V2′.