Virus specificity and nucleoporin requirements for MX2 activity are affected by GTPase function and capsid-CypA interactions

Abstract

Human myxovirus resistance 2 (MX2/MXB) is an interferon-induced GTPase that inhibits human immunodeficiency virus-1 (HIV-1) infection by preventing nuclear import of the viral preintegration complex. The HIV-1 capsid (CA) is the major viral determinant for sensitivity to MX2, and complex interactions between MX2, CA, nucleoporins (Nups), cyclophilin A (CypA), and other cellular proteins influence the outcome of viral infection. To explore the interactions between MX2, the viral CA, and CypA, we utilized a CRISPR-Cas9/AAV approach to generate CypA knock-out cell lines as well as cells that express CypA from its endogenous locus, but with specific point mutations that would abrogate CA binding but should not affect enzymatic activity or cellular function. We found that infection of CypA knock-out and point mutant cell lines with wild-type HIV-1 and CA mutants recapitulated the phenotypes observed upon cyclosporine A (CsA) addition, indicating that effects of CsA treatment are the direct result of blocking CA-CypA interactions and are therefore independent from potential interactions between CypA and MX2 or other cellular proteins. Notably, abrogation of GTP hydrolysis by MX2 conferred enhanced antiviral activity when CA-CypA interactions were abolished, and this effect was not mediated by the CA-binding residues in the GTPase domain, or by phosphorylation of MX2 at position T151. We additionally found that elimination of GTPase activity also altered the Nup requirements for MX2 activity. Our data demonstrate that the antiviral activity of MX2 is affected by CypA-CA interactions in a virus-specific and GTPase activity-dependent manner. These findings further highlight the importance of the GTPase domain of MX2 in regulation of substrate specificity and interaction with nucleocytoplasmic trafficking pathways.

Figure 1.. The effects of CsA on HIV-1 sensitivity to MX2 are the direct result of blocking CA-CypA interactions.

A) Structure of an HIV-1 capsid monomer in complex with CypA (PDB 1AK4). Critical active -site residues that interact with CA are indicated in dark green, amino acids outside the active site that affect capsid recognition highlighted in pink/magenta. The cyclophilin-binding loop of HIV-1 CA indicated in yellow. B) Infection of control or CypA-mutant HT1080 cells expressing doxycycline-inducible MX2 in the presence (open bars) or absence (filled bars) of doxycycline (Dox) and present or absence of CsA with the indicated GFP reporter viruses. Titers are represented as mean + sem of infectious units (IU) per pg of reverse transcriptase (RT), n≥6 technical replicates combined from two-three independent experiments. Statistical analysis in S1 File. C) Western blot analysis of doxycycline-inducible MX2, CypA, and tubulin loading control.

Figure 2.. Summary of MX1 and MX2 mutants used in this study.

A) Diagram representing the various domains of MX2, with the N-terminal 25 amino acids shown in detail. BSE: bundle-signaling element. Residues of particular interest are indicated. B) Diagram representing the domains of MX1 (top), with residues corresponding to those highlighted in MX2 indicated, and diagram of chimeric fusion of MX1 containing the N-terminal 91 amino acids of MX2 (bottom). C) Overview of mutations in MX2 known to affect CA-binding, phosphorylation, GTPase activity, and oligomerization. Known consequences of mutations/functions of specific residues and relevant citations are shown.

Figure 3.. Determinants for MX2 activity in the absence of CA-CypA interactions.

A) Infectivity of GFP reporter viruses in HT1080 cells expressing doxycycline-inducible C-terminally myc-tagged MX2, MX2 mutants, or MX1_MX2-NTD in the presence (open bars) and absence (filled bars) of doxycycline (Dox). Cells were infected in the presence of CsA where indicated. Titers are represented as mean + sem of infectious units (IU) per pg of reverse transcriptase (RT), n≥8 technical replicates combined from two-six independent experiments. Statistical analysis in S1 File. B) Western blot analysis of doxycycline-inducible MX2-myc and tubulin loading control.

Figure 4.. Phosphorylation at residue T151 does not determine CypA-independent MX2 activity.

A) Infection HT1080 cells stably transduced with doxycycline-inducible myc-tagged MX2 or MX2_T151D in the presence (open bars) or absence (filled bars) of doxycycline and presence or absence of CsA with the indicated GFP reporter viruses. Titers are represented as mean + sem of infectious units (IU) per pg of reverse transcriptase (RT), n≤9 technical replicates combined from three-five independent experiments. Statistical analysis in S1 File. B) Western blot analysis of doxycycline-inducible MX2-myc and tubulin loading control.

Figure 5.. Restriction by GTPase-deficient MX2 in the absence of CA-CypA binding is not mediated by known CA-GTPase domain interactions.

A) Infection of HT1080 cells stably transduced with doxycycline-inducible myc-tagged MX2 with or without mutations in the N-terminal triple-arginine motif (₁₁AAA₁₃), T151A, or CA-binding residues in the GTPase domain (SSE) in the presence (open bars) or absence (filled bars) of doxycycline and presence or absence of CsA with the indicated GFP reporter viruses. Titers are represented as mean + sem of infectious units (IU) per pg of reverse transcriptase (RT), n≥9 technical replicates combined from three-five independent experiments. Statistical analysis in S1 File. B) Western blot analysis of doxycycline-inducible MX2-myc and tubulin loading control.

Figure 6.. Antiviral activity of GTPase-deficient chimeric MX1_MX2-NTD proteins.

A) Infection of HT1080 cells stably transduced with doxycycline-inducible myc-tagged MX1 or MX1_MX2-NTD with the indicated mutations in the presence (open bars) or absence (filled bars) of doxycycline and presence or absence of CsA with the indicated GFP reporter viruses. Titers are represented as mean + sem of infectious units (IU) per pg of reverse transcriptase (RT), n≥9 technical replicates combined from three-seven independent experiments. Statistical analysis in S1 File. B) Western blot analysis of doxycycline-inducible MX1_MX2-NTD or MX1-myc and tubulin loading control.

Figure 7.. GTPase-deficient MX2 has distinct nucleoporin requirements for antiviral activity.

A) Schematic representation of the nuclear pore complex and genes included in siRNA library color coded by subcomplex. (*- NUP98 is listed as a member of Nup107 subcomplex, however. Nup98 and Nup96 are produced following autoproteolytic cleavage of a polyprotein precursor (57, 58), this siRNA will target both Nups). Importins/nuclear transport receptors (NTRs) included in the siRNA library listed in grey. Also included, siRNA targeting CypA, CPSF6, and a non-targeting control siRNA. B-D) Infectivity of HIV-1 GFP reporter viruses in HT1080 cells stably transduced with doxycycline-inducible MX2 or MX2_T151A in the presence (open circles) or absence (filled circles) of doxycycline 64-72h after transfection with siRNA (color coded by subcomplex as in A). Titers are mean ± sem, n=3 technical replicates, representative of four independent experiments. Statistical analysis in S1 File.