CTL escape in B27-KK10 is associated with multiple mutations. Viral sequences of Gag p24 in subjects expressing HLA-B27 are aligned to consensus clade B and NL4-3. Areas in gray highlight amino acid position 173 and the KK10 epitope (amino acids 263 to 272). The CypA binding domain is shown in the consensus clade B sequence (underlined), and mixed amino acids are indicated by lowercase letters. ND indicates “not determined.” cl indicates sequence derived from a single full-length clone representative of the quasispecies. Where available, viral loads are indicated. (A) Longitudinal viral sequencing in four HLA-B27⁺ subjects illustrates that the L₂₆₈M mutation arises within the first 1 to 3 years after infection and prior to the R₂₆₄K mutation. Sample time points are indicated as days postpresentation (pp). Fractions in parentheses indicate the fractions of clones representing displayed sequence. (B) Cross-sectional sequencing of 26 chronically infected B27⁺ subjects illustrates the propensity for viral escape through the R₂₆₄K and L₂₆₈M mutations. The S₁₇₃A mutation is strongly associated with the R₂₆₄K mutation. Numbers in parentheses indicate the number of clones sequenced for which the consensus sequence is provided.

Location of KK10-associated mutations in the N-terminal domain of p24 capsid. The mutations associated with viral escape in B27-KK10 were mapped onto the structural model of the N-terminal domain of p24 capsid using Cn3D (National Center for Biotechnology Information). R₂₆₄K and L₂₆₈M (red) are located inside the KK10 epitope (green) on helix VII, while the S₁₇₃A mutation (red) is located on helix II on the same planar surface of the folded p24 molecule as R₂₆₄K and L₂₆₈M.

RK infectivity is restored in the absence of CypA. (A) CEM-GXR cells were infected with VSV-g-pseudotyped WT, RK, RKLM, and SARKLM viruses (400 ng p24) in the presence or absence of 0.5 μM CsA. The percentage of GFP⁺ cells was determined by flow cytometry at 48 h postinfection and is shown for each variant. Infectivities of WT (black) and SARKLM (green) were reduced by 2.2- and 2.5-fold, respectively, in the presence of 0.5 μM CsA, while the percentages of infected cells for RK (red) and RKLM (orange) increased by 5.7- and 3.4-fold, respectively, in the presence of drug. The introduction of A₂₃₇T increased the infectivity of RKLM (gray) by 4.2-fold in the absence of CsA. In the presence of the drug, the infectivity of this strain was reduced by threefold. Results displayed are mean values for duplicate cultures and are representative of three independent experiments. To further evaluate the role of CypA, 1 million CEM-GXR (B), JKT (C), or JKT CypA^−/− (D) cells were infected with 100 ng p24 equivalent of the WT (black), RK (red), or SARKLM (green). The p24 concentration in the supernatant at the indicated days postinfection was determined by ELISA. Cultures with WT and SARKLM viruses generated similar amounts of p24 in CEM-GXR and JKT cells. In the JKT CypA^−/− lines, p24 production was reduced by 1 log, and peak p24 levels were delayed by 4 days. RK virus failed to increase p24 values in CEM-GXR cells. Compared to the WT and SARKLM, p24 production by RK in JKT cells was reduced by 2.5 logs, while all three viruses displayed similar p24 values for JKT Cyp^−/− cells. Results shown in B, C, and D are representative of three independent experiments.

Escape mutation R₂₆₄K substantially impacts viral replication in vitro. (A) CEM-GXR cells were infected with WT or variant viruses (100 μl of viral stocks), and the percentage of infected GFP-positive (GFP⁺) cells/ng p24 was determined at 48 h by flow cytometry. Values were normalized to WT values (black). The RK (red) and RKLM (orange) variants exhibited substantial deficiencies in infectivity, while only minor differences were observed for LM (blue), SA (purple), and SARKLM (green). Results shown are representative of three or more independent experiments. (B) CEM-GXR cells were infected (multiplicity of infection of 0.0015), and viral spread was measured over 7 days using flow cytometry to quantify GFP⁺ cells. The percentage of infected cells was normalized to the value at day 2 (0.15%). Again, RK (red) and RKLM (orange) exhibited deficiencies in replication, while SARKLM virus (green) replicated efficiently. Results shown are representative of three independent experiments. (C) The replicative defect of the RK variant was confirmed using primary cells. PBMC were inoculated with virus (10 ng p24), and viral spread was measured by p24 ELISA over 16 days. WT (black) and SARKLM (green) replicated similarly, while the RK variant (red) generated less p24 at all times tested. Results shown are representative of three independent experiments using different donors.

The RK variant is defective in the production of late viral reverse transcripts without affecting capsid-uncoating kinetics. (A) CEM-GXR cells were infected with VSV-g-pseudotyped WT, RK, and SARKLM viruses (400 ng p24). Cells were harvested at 3, 12, 24, and 48 h postinfection and analyzed for late reverse transcription (RT) products using real-time PCR, and values were normalized to the host cell actin gene. Late RT products were readily generated by WT (black) and SARKLM (green) within 12 h and then increased through 48 h, while the levels of reverse transcription products for RK (red) were significantly reduced at all times tested. Results are representative of two independent experiments. (B) The stabilities of WT (black) and RK (red) capsids were examined using a previously described assay (27). Purified cores were incubated in STE buffer at 37°C for the times shown. Following incubation, the samples were subjected to ultracentrifugation. The extent of disassembly was determined as the percentage of the total capsid protein in the reaction mixture present in the supernatant. In this in vitro assay, both viruses disassemble at the same rate. The results shown are mean values of duplicate determinations and are representative for two independent experiments.