Cooperative effect of gag proteins p12 and capsid during early events of murine leukemia virus replication

Abstract

The Gag polyprotein of murine leukemia virus (MLV) is processed into matrix (MA), p12, capsid (CA), and nucleocapsid (NC) proteins. p12 affects early events of virus replication and contains a PPPY motif important for virus release. To probe the functions of p12 in the early steps of MLV replication, we tested whether p12 can be replaced by spleen necrosis virus (SNV) p18, human immunodeficiency virus type 1 p6, or Rous sarcoma virus p2b. Analyses revealed that all chimeras generated virions at levels similar to that of MLV gag-pol; however, none of them could support MLV vector replication, and all of them exhibited severely reduced DNA synthesis upon virus infection. Because a previously reported SNV gag-MLV pol chimera, but not the MLV hybrid with SNV p18, can support replication of an MLV vector, we hypothesized that other Gag proteins act cooperatively with p12 during the early phase of virus replication. To test this hypothesis, we generated three more MLV-based chimeras containing SNV CA, p18-CA, or p18-CA-NC. We found that the MLV chimera containing SNV p18-CA or p18-CA-NC could support MLV vector replication, but the chimera containing SNV CA could not. Furthermore, viruses derived from the MLV chimera with SNV CA could synthesize viral DNA upon infection but were blocked at a post-reverse-transcription step and generated very little two long terminal repeat circle DNA, thereby producing a phenotype similar to that of the provirus formation-defective p12 mutants. Taken together, our data indicate that when p12/p18 or CA was from different viruses, despite abundant virus production and proper Gag processing, the resulting viruses were not infectious. However, when p12/p18 and CA were from the same virus, even though they were from SNV and not MLV, the resulting viruses were infectious. Therefore, these results suggest a cooperative effect of p12 and CA during the early events of MLV replication.

FIG. 1.

General structures of the MLV-based gag-pol expression constructs used to study the functions of p12. (A) Wild-type MLV gag-pol expression construct and its derivatives containing a deletion or substitution of the p12 domain. (B) Chimeric MLV-based gag-pol expression constructs containing CA and other domains from SNV. Open boxes, MLV-derived sequences; shaded boxes, SNV-derived sequences; hatched box, HIV-1 p6; stippled box, RSV p2b; dashed line, deleted region.

FIG. 2.

Western analyses of wild-type and mutant MLV gag gene products in transfected cells (A) and cell-free virions (B). Anti-MLV CA antibody was used to detect gag gene products in transfected cell lysates, whereas both anti-MLV CA and anti-MLV MA antibodies were used to detect gag gene products in virion lysates.

FIG. 3.

Western analyses of cell-free virions harvested from cells transfected with wild-type MLV (A) or MLV/SNV chimeric gag-pol expression constructs (B). Anti-MLV MA and anti-MLV CA antibodies were used in the Western analysis shown in panel A, whereas anti-MLV MA and anti-SNV CA antibodies were used for panel B.

FIG. 4.

Quantitation of MLV vector RNA encapsidated in cell-free virions. Levels of MLV vector SR2 RNA encapsidated in the cell-free virions derived from various gag-pol expression constructs were determined by real-time RT-PCR and normalized to the RT activity of the input virus. For direct comparison, the amount of RNA encapsidated by virions derived from wild-type MLV (pWZH30) was set as 100%. Data from three independent experiments are summarized and are shown as means ± standard errors.

FIG. 5.

Analyses of viral DNA synthesis in 293T cells infected with viruses derived from wild-type or mutant gag-pol expression constructs. (A) Synthesis of early reverse transcription product R-U5. (B) Synthesis of late reverse transcription product U5-Ψ. Real-time PCR was performed using primer and probe sets specific to R-U5, U5-Ψ, or PBGD sequences. PBGD quantitation served as a standard to control for variation in DNA recovery. For direct comparison, the amount of viral DNA detected from wild-type MLV was set as 100%. Data from three independent experiments are summarized and shown as means ± standard errors.

FIG. 6.

Analyses of 2-LTR circle DNA in 293T cells infected with virions derived from wild-type (WT) (A) or wild-type or mutant gag-pol expression constructs (B). DNA was isolated from infected cells 24 h postinfection, and the amounts of viral DNA in these samples were quantified by detecting U5-Ψ sequences. Unless specified in the dilution (dil.) studies in panel A, samples containing equal amounts of viral DNA were used as the template for PCR analyses to detect 2-LTR circle DNA. nts, nucleotides.

FIG. 7.

EM analyses of wild-type and mutant MLV particles. Virions were derived from pWZH30 (A), pMΔp12 (B), pMΔp12/PY (C), pMΔp12/Hp6 (D), pMΔp12/Rp2b (E), pMΔp12/Sp18 (F), pM/SCA (G), pM/SPC (H), or pM/SPCN (I). The black bar in each electron micrograph indicates 100 nm. Tube-like structures are indicated by white arrows; elongated structures are indicated by solid arrows.

FIG. 7.

EM analyses of wild-type and mutant MLV particles. Virions were derived from pWZH30 (A), pMΔp12 (B), pMΔp12/PY (C), pMΔp12/Hp6 (D), pMΔp12/Rp2b (E), pMΔp12/Sp18 (F), pM/SCA (G), pM/SPC (H), or pM/SPCN (I). The black bar in each electron micrograph indicates 100 nm. Tube-like structures are indicated by white arrows; elongated structures are indicated by solid arrows.