The G1/S Specific Cyclin D2 Is a Regulator of HIV-1 Restriction in Non-proliferating Cells

Abstract

Macrophages are a heterogeneous cell population strongly influenced by differentiation stimuli that become susceptible to HIV-1 infection after inactivation of the restriction factor SAMHD1 by cyclin-dependent kinases (CDK). Here, we have used primary human monocyte-derived macrophages differentiated through different stimuli to evaluate macrophage heterogeneity on cell activation and proliferation and susceptibility to HIV-1 infection. Stimulation of monocytes with GM-CSF induces a non-proliferating macrophage population highly restrictive to HIV-1 infection, characterized by the upregulation of the G1/S-specific cyclin D2, known to control early steps of cell cycle progression. Knockdown of cyclin D2, enhances HIV-1 replication in GM-CSF macrophages through inactivation of SAMHD1 restriction factor by phosphorylation. Co-immunoprecipitation experiments show that cyclin D2 forms a complex with CDK4 and p21, a factor known to restrict HIV-1 replication by affecting the function of the downstream cascade that leads to SAMHD1 deactivation. Thus, we demonstrate that cyclin D2 acts as regulator of cell cycle proteins affecting SAMHD1-mediated HIV-1 restriction in non-proliferating macrophages.

Fig 1. Susceptibility to HIV-1 infection in primary macrophages depends on differentiation stimuli.

(A) Evaluation of cell proliferation by Ki67 staining. Histograms of a representative donor showing Ki67 staining in M-CSF and GM-CSF derived macrophages. Percentage of positive cells is shown in each case. (B) Dot plots representing cell cycle profile showing DNA (7-AAD) and RNA (Pyronin Y) content of M-CSF and GM-CSF differentiated macrophages quantified by flow cytometry. Data from a representative donor is shown. (C) Dot plots of infected M-CSF macrophages (upper panels) or GM-CSF macrophages (lower panels) untreated (left) or after treatment (right) with VLP_Vpx. Data from a representative donor is shown. (D) Percentage replication in M-CSF MDM and GM-CSF MDM untreated (white bars) or transduced with VLP_Vpx (black bars). Mean ± SD of 4 different donors performed in triplicate is shown. (E) and (F) Drug sensitivity in M-CSF and GM-CSF differentiated macrophages. Antiviral activity of zidovudine (AZT, 1 μM), nevirapine (NVP, 5 μM) and palbociclib (PD, 4 μM) was evaluated in M-CSF (E) and GM-CSF (F) MDM, untreated (white bars) or transduced with VLP_Vpx (black bars). Antiviral potency was decreased with AZT and completely lost with PD after degradation of SAMHD1 with VLP_Vpx. Mean ± SD of at least 3 different donors performed in triplicate is shown. ND; no-drug, ns; not significant, * p<0.05; ** p<0.005.

Fig 2. M-CSF and GM-CSF macrophages present differential expression of cell cycle-related proteins and SAMHD1 activation.

(A) Gene expression of cell cycle-related genes and SAMHD1 restriction pathway. mRNA levels of SAMHD1, CDK1, CDK2, CDK4 and CDK6, the corresponding cyclins (A2, B1, B2, D1, D2, D3, E1 and E2) and CDK2 inhibitor p21 (CDKN1A) was quantified in M-CSF (white bars) and GM-CSF (black bars) differentiated macrophages. Data is normalized to M-CSF relative expression. Mean ± SD of 3 independent donors is shown. ** p<0.005; *** p<0.0005. (B) Western blot showing protein expression of different cell cycle proteins, SAMHD1 expression and activation and Hsp90 as loading control. Two representative donors are shown. M; M-CSF MDM, GM; GM-CSF MDM. (C) Induction of gene expression following LPS (100 ng/ml) treatment of M-CSF and GM-CSF macrophages. Expression of IFNB1, CCL-2 and IL-10 were evaluated. Data is normalized to untreated M-CSF condition. Mean ± SD of 3 independent donors is shown.

Fig 3. RNAi of cyclin D2 and cyclin D3 differently impact viral replication depending on MDM type.

(A) Effective and specific knockdown of CCND2 and CCND3 expression by siRNA. Relative mRNA expression of CCND2 (left panel) and CCND3 (right panel) in M-CSF (white bars) and GM-CSF (black bars) macrophages. mRNA of the corresponding cyclin was measured by quantitative PCR and normalized to GAPDH expression. Data represents mean ± SD of 3 different donors and is normalized to Mock-transfected M-CSF macrophages. (B) Protein expression and SAMHD1 activation in cyclin D2 and cyclin D3 knockdown macrophages. Western blot showing cyclin protein expression, SAMHD1 expression and activation in siRNA-treated M-CSF and GM-CSF macrophages. SAMHD1 inactivation is reduced in siCCND3 (D3) M-CSF macrophages and increased in siCCND2 (D2) GM-CSF macrophages compared to the corresponding non-targeting siRNA (NT). Hsp90 was used as loading control. A representative donor is shown. (C) Evaluation of cell proliferation in cyclin D2 and cyclin D3 knockdown macrophages. Ki67 staining of siRNA-treated M-CSF (upper panels) and GM-CSF (lower panels) derived macrophages. Histograms from a representative donor are shown. (D) HIV-1 replication in siRNA-treated M-CSF or GM-CSF macrophages. Transfected MDM were infected with a VSV-pseudotyped, GFP-expressing HIV-1 and infection measured 72h later by flow cytometry. Data represent percentage replication relative to mock-transfected cells in M-CSF (white bars) or GM-CSF (black bars) macrophages. Mean ± SD of 3 different donors performed in triplicate is shown. * p<0.05; ** p<0.005; *** p<0.0005.

Fig 4. Cyclin D2 expression restricts HIV-1 replication by controlling SAMHD1 activation in GM-CSF macrophages.

(A) Effective and specific knockdown of CCND2, CCND3, CDKN1A and CDK4 expression by siRNA in GM-CSF macrophages. mRNA of the corresponding gene was measured by quantitative PCR and normalized to GAPDH expression. Data represents mean ± SD of at least 3 different donors and is normalized to Mock-transfected macrophages. (B) Protein expression and SAMHD1 activation in siRNA-treated GM-CSF macrophages. Western blot showing cyclin D2, cyclin D3, CDK4 and p21 protein expression and SAMHD1 expression and activation in siRNA-treated GM-CSF macrophages. SAMHD1 inactivation by phosphorylation is increased in cyclin D2 and p21 knockdown macrophages, compared to Mock-transfected or macrophages treated with a non-targeting siRNA (siNT). (C) Evaluation of cell proliferation and cell cycle analysis in siRNA-treated GM-CSF macrophages. Histograms showing Ki67 staining (upper panels) and dot plots representing cell cycle profile by DNA (7-AAD) and RNA (Pyronin Y) staining (lower panels) in siRNA-treated macrophages. Percentage of positive cells in a representative donor is shown in each case. (D) HIV-1 replication in siRNA-treated GM-CSF macrophages. Transfected MDM were infected with a VSV-pseudotyped, GFP-expressing HIV-1 and infection measured 72h later by flow cytometry. Data represent percentage replication relative to mock-transfected macrophages. Mean ± SD of at least 3 different donors performed in duplicate is shown. (E) Proviral DNA formation after 16h infection with HIV-1 BaL of GM-CSF macrophages transfected with the indicated siRNA or treated with AZT (3 μM) or raltegravir (RAL; 2 μM). Proviral DNA was normalized to mock-treated macrophages. Mean ± SD of at least 3 different donors is shown. * p<0.05; ** p<0.005; *** p<0.0005.

Fig 5. Co-immunoprecipitation (IP) of cyclin D2 with p21.

(A) Co-IP assay of 293T cells transfected with a plasmid expressing Flag-tagged p21. Lysates from mock- transfected (M) HEK293T cells or transfected with a Flag-p21 expression plasmid (p21) were subjected to immunoprecipitation with anti-Flag antibodies attached to sepharose or sepharose alone (B; beads). Whole cell lysates (WCL) and immunoprecipitates were analyzed by immunoblotting with an anti-cyclin D2 antibody or different CDK antibodies (CDK4, CDK6 and CDK1). Anti-p21 and anti-Hsp90 antibodies were used as controls. (B) Co-IP assay of 293T cells transfected with a plasmid expressing HA-tagged cyclin D2. As in (A) Lysates from mock- transfected (M) HEK293T cells or transfected with a HA-cyclin D2 expression plasmid (D2) were subjected to immunoprecipitation with anti-HA antibodies attached to sepharose or sepharose alone (B; beads). Whole cell lysates (WCL) and immunoprecipitates were analyzed by immunoblotting with anti-p21, anti-CDK4 and anti-CDK1 antibodies. Anti-cyclin D2 and anti-Hsp90 antibodies were used as controls. (C) Co-IP assay of endogenous cyclin D2 and p21 in GM-CSF primary macrophages. Lysates from primary macrophages where subjected to immunoprecipitation with anti-cyclin D2 or anti-p21 antibodies, attached to sepharose or sepharose attached to IgG alone (B, beads). Whole cell lysate (WCL) and immunoprecipitates were analyzed by immunoblotting with anti-cyclin D2, anti-p21, anti-CDK4 and anti-CDK1 antibodies. Anti-Hsp90 antibody was used as a control. For all Co-IP assays, a representative experiment out of 3 is shown. (D) Quantification of endogenous immunoprecipitated proteins with anti-cyclin D2 (white bars) or anti-p21 antibodies (black bars). Mean band density values ± SD of three independent donors is shown.

Fig 6. Cyclin D2 expression controls subsequent activation of CDK1 and CDK2.

(A) CDK1 and CDK2 gene expression in CCND2 and CDKN1A knockdown GM-CSF macrophages. CDK1 and CDK2 gene expression was measured by quantitative PCR and normalized to GAPDH expression. Data represents mean ± SD of 3 different donors and is normalized to Mock-transfected macrophages. (B) Western blot showing CDK1 and CDK2 expression and phosphorylation of CDK2 in siRNA-treated GM-CSF macrophages. CDK1 expression is upregulated after cyclin D2 knockdown as well as CDK2 activity measured as phosphorylation at Thr130, compared to Mock-transfected or macrophages treated with a non-targeting siRNA (siNT). (C) Proposed regulatory model of Cyclin D2 mediated control of HIV-1 restriction in non-proliferating cells. The Cyclin D2/CDK4/p21 complex is responsible for the lack of active CDK2 and CDK1 expression in GM-CSF macrophages. This situation is reversed in the absence of cyclin D2, leading to the activation of CDK2 and CDK1 with the subsequent phosphorylation of its substrates, including SAMHD1.