Cyclophilin A regulates HIV-1 infectivity, as demonstrated by gene targeting in human T cells

Abstract

The human immunodeficiency virus type 1 (HIV-1) Gag polyprotein binds most members of the cyclophilin family of peptidyl-prolyl isomerases. Of 15 known human cyclophilins, cyclophilin A (CypA) has been the focus of investigation because it was detected in HIV-1 virions. To determine whether CypA promotes HIV-1 replication, we deleted the gene encoding CypA (PPIA) in human CD4(+) T cells by homologous recombination. HIV-1 replication in PPIA(-/-) cells was decreased and not inhibited further by cyclosporin or gag mutations that disrupt Gag's interaction with cyclophilins, indicating that no other cyclophilin family members promote HIV-1 replication. The defective replication phenotype was specific for wild-type HIV-1 since HIV-2/SIV isolates, as well as HIV-1 bearing a gag mutation that confers cyclosporin resistance, replicated the same in PPIA(+/+) and PPIA(-/-) cells. Stable re-expression of CypA in PPIA(-/-) cells restored HIV-1 replication to an extent that correlated with steady-state levels of CypA. Finally, virions from PPIA(-/-) cells possessed no obvious biochemical abnormalities but were less infectious than virions from wild-type cells. These data formally demonstrate that CypA regulates the infectivity of HIV-1 virions.

Fig. 1. Gene targeting of the PPIA locus by homologous recombin ation. (A) Strategy for deleting all but the fifth exon of both copies of PPIA in Jurkat T cells by two consecutive rounds of gene targeting. Linearized targeting plasmids are shown above and below a schematic of the genomic locus. ‘Neo’, neomycin (G418) resistance gene; ‘Hygro’, hygromycin resistance gene; S, SacI; X, XbaI. (B) Southern blot hybridization of PPIA^+/+, PPIA^+/– and PPIA^–/– cell lines; total cellular DNAs were digested with both SacI and XbaI and probed with the 5′ probe (shown in A). (C) Western blot of total cell lysates from PPIA^+/+, PPIA^+/– and PPIA^–/– cell lines, probed with polyclonal anti-actin and anti-CypA antibodies.

Fig. 2. HIV-1 replication is decreased in PPIA^–/– Jurkat T cells. (A) Replication of HIV-1_NL4-3 in PPIA^+/+ (filled circles), PPIA^+/– (shaded triangles) and PPIA^–/– (open squares) cell lines, demonstrating delayed kinetics in PPIA^–/– cells. (B) Replication of HIV-1_Eli in PPIA^+/+ (filled circles) and PPIA^–/– (open squares) cell lines, demonstrating an even greater defect in replication than HIV-1_NL4-3.

Fig. 3. The decreased replication phenotype in PPIA^–/– Jurkat T cells is specific to wild-type HIV-1. Replication of HIV-1_NL4-3/A224E (A), HIV-2_GB122 (B), HIV-2₇₇₆₁₈ (C) or HIV-1_NL4-3/P222A (D) is nearly the same in PPIA^+/+ (filled circles) and PPIA^–/– cells (open squares), indicating that replication of these viruses is independent of CypA.

Fig. 4. HIV-1 replication in PPIA^–/– cells is insensitive to a competitive inhibitor of the Gag–cyclophilin interaction. Effect of methyl-Ile4-cyclosporin on replication of HIV-1_NL4-3 (A) and HIV-2₇₇₆₁₈ (B) in PPIA^+/+ cells. Effect of methyl-Ile4-cyclosporin on replication of HIV-1_NL4-3 (C) and HIV-2₇₇₆₁₈ (D) in PPIA^–/– cells. No drug (filled symbols), 1.25 µM (shaded symbols) and 2.5 µM (open symbols) methyl-Ile4-cyclosporin.

Fig. 5. HIV-1 replication kinetics depend on the level of CypA expression. (A) Western blot of total cell lysates showing the relative level of CypA in the cell lines. (B) Replication of HIV-1_NL4-3 in PPIA^+/+ cells (filled circles), in three cell lines derived from PPIA^–/– cells stably re-expressing CypA (shaded symbols) and in PPIA^–/– cells (open squares). Higher CypA expression levels correlate with more rapid replication kinetics.

Fig. 6. Virions produced from PPIA^–/– cells are less infectious than those produced from PPIA^+/+ cells. Virus stocks (wild-type HIV-1_NL4-3 and HIV-1_NL4-3/G221A) were produced in either PPIA^+/+ or PPIA^–/– cells (as indicated) and used to infect equal numbers of GHOST cells. (A) Low molecular weight DNA was isolated 12 h post-infection. Real-time PCR was performed to quantitate full-length viral cDNA and 2-LTR circles. The relative levels of products (percentage of wild-type signal) are shown. (B) The percentage of GFP⁺ cells (of total cells) was determined by flow cytometry 48 h post-infection.