Microbial Protein Binding to gC1qR Drives PLA2G1B-Induced CD4 T-Cell Anergy

Abstract

The origin of the impaired CD4 T-cell response and immunodeficiency of HIV-infected patients is still only partially understood. We recently demonstrated that PLA2G1B phospholipase synergizes with the HIV gp41 envelope protein in HIV viremic plasma to induce large abnormal membrane microdomains (aMMDs) that trap and inactivate physiological receptors, such as those for IL-7. However, the mechanism of regulation of PLA2G1B activity by the cofactor gp41 is not known. Here, we developed an assay to directly follow PLA2G1B enzymatic activity on CD4 T-cell membranes. We demonstrated that gp41 directly binds to PLA2G1B and increases PLA2G1B enzymatic activity on CD4 membrane. Furthermore, we show that the conserved 3S sequence of gp41, known to bind to the innate sensor gC1qR, increases PLA2G1B activity in a gC1qR-dependent manner using gC1qR KO cells. The critical role of the 3S motif and gC1qR in the inhibition of CD4 T-cell function by the PLA2G1B/cofactor system in HIV-infected patients led us to screen additional microbial proteins for 3S-like motifs and to study other proteins known to bind to the gC1qR to further investigate the role of the PLA2G1B/cofactor system in other infectious diseases and carcinogenesis. We have thus extended the PLA2G1B/cofactor system to HCV and Staphylococcus aureus infections and additional pathologies where microbial proteins with 3S-like motifs also increase PLA2G1B enzymatic activity. Notably, the bacteria Porphyromonas gingivalis, which is associated with pancreatic ductal adenocarcinoma (PDAC), encodes such a cofactor protein and increased PLA2G1B activity in PDAC patient plasma inhibits the CD4 response to IL-7. Our findings identify PLA2G1B/cofactor system as a CD4 T-cell inhibitor. It involves the gC1qR and disease-specific cofactors which are gC1qR-binding proteins that can contain 3S-like motifs. This mechanism involved in HIV-1 immunodeficiency could play a role in pancreatic cancer and several other diseases. These observations suggest that the PLA2G1B/cofactor system is a general CD4 T-cell inhibitor and pave the way for further studies to better understand the role of CD4 T-cell anergy in infectious diseases and tumor escape.

Keywords: CD4 T cell; HCV; HIV; PLA2G1B; gC1qR; infectious disease; porphyromonas gingivalis; staphylococcus aureus.

Figure 1

HIV gp41 protein increases PLA2G1B enzymatic activity on CD4 T-cell membranes. (A, B) The sPLA2 inhibitors varespladib, pentapeptide, and sPLA2R1 and the anti-PLA2G1B mAb 14G9 strongly inhibit PLA2G1B activity in HIV viremic plasma. Purified HD CD4 T cells from two donors (A) and three donors (B) were treated with 3% of VP plasma from five viremic HIV patients together with varespladib (50 µM), pentapeptide (15 µM), sPLA2R1 (160 nM), or not (A) or control isotype (ctrl iso) or 14G9 mAb (667 nM, B) and the pSTAT5-NT cell response to IL-7 was analyzed by confocal microscopy. (C) PLA2G1B inhibits the pSTAT5-NT cell response to IL-7 in a dose-response manner. HD CD4 T cells were purified from four donors. The IC50 value is indicated. Results are shown as the mean ± SD of the percentage of pSTAT5-NT cell inhibition for A-C. (D) PLA2G1B activity can be followed in a dose-response manner on human CD4 T cells labeled with tritiated arachidonic acid. CD4 T cells purified from four donors were incubated with several concentrations of PLA2G1B (0.1 nM-5 µM) for 2 h. Then, the release of [3H]-AA in the cell supernatant due to PLA2G1B activity was quantified with a radioactivity counter (tri-Carb 2800 TR liquid scintillation analyzer, Perkin Elmer). Results are shown as the mean ± SD of PLA2G1B activity in cpm/mL from a pool of seven experiments. The EC50 value is indicated. (E) pSTAT5-NT cell inhibition positively correlates with PLA2G1B enzymatic activity on [3H]-AA-labeled CD4 T cells. Potential correlations were analyzed using the Pearson r test and a linear regression is presented. (F) [3H]-AA release is dependent on PLA2G1B enzymatic activity. [3H]-AA-labeled CD4 T cells were treated with several concentrations (10, 63, 200 nM) of WT human (hPLA2G1B) or WT or the catalytic-site mutant H48Q porcine PLA2G1B (pPLA2G1B). Results are shown as the mean of PLA2G1B activity ± SD of one experiment with the treatment in triplicate. (G) Varespladib, and pentapeptide strongly inhibit PLA2G1B enzymatic activity on [3H]-AA-labeled CD4 T cells. Results are shown as the percentage of inhibition of PLA2G1B activity on cells treated with PLA2G1B (65 nM) and several doses of inhibitors. IC50 values are presented. (H) ProPLA2G1B is inactive in the [3H]-AA release assay on CD4 T-cell membranes. [3H]-AA-labeled CD4 T cells were treated in triplicate with PLA2G1B or proPLA2G1B at 200 nM. Results are shown as the percentage of activity with proPLA2G1B relative to that of PLA2G1B. (I, J) HIV gp41 increases PLA2G1B activity in a dose-dependent manner on human CD4 T cells. [3H]-AA-labeled CD4 T cells were incubated alone or with PLA2G1B (63 nM or 200 nM) in the presence or not of several concentrations of recombinant gp41 protein (0.68 nM-340 nM). Results are shown as the mean ± SD from a pool of four experiments for I and n = 2-3 experiments for (J) **p < 0.01 and ***p < 0.001 by two-way ANOVA with Tukey’s correction for multiple comparisons (A, B, F, H) and the Kruskal-Wallis test, followed by the Mann-Whitney test with p-values adjusted for multiple comparisons between groups (J). For F, only comparisons between H48Q and WT hPLA2G1B or pPLA2G1B at each PLA2G1B concentration are shown.

Figure 2

PLA2G1B directly binds to gp41. (A, B) gp41 binds to coated PLA2G1B by ELISA. Binding to coated PLA2G1B of serial dilutions (0-1 µg/well) of gp41 (D117III strain on A and MN strain on B), or unrelated proteins (CTL1: EFF-1 on A and B or CTL2: IF38 on A). (C, D) PLA2G1B binds to coated gp41 by ELISA. Binding of serial dilutions (0-1 µg/well) of PLA2G1B to coated gp41 (D117III on C, or MN on D) was tested in triplicate by ELISA. Results are shown as the mean ± SD of the OD value of one representative experiment among two (A), five (B), one (C) and five (D). (E) gp41 binds to PLA2G1B in pull-down assays. Recombinant PLA2G1B protein was incubated with strep-tagged-gp41 (D117III strain, gp41+PLA2G1B) or not (PLA2G1B only). PLA2G1B-gp41 complexes were pulled-down with strep-tactin XT-beads that pull-down gp41. Pulled-down products and unbound proteins were revealed by immunoblotting with goat polyclonal anti-gp41 Ab or mouse anti-PLA2G1B mAb. One representative experiment of five experiments with similar results with D117III gp41 is presented. Similar results were obtained in three experiments with MN gp41 ( Supplementary Figure 2 ).

Figure 3

PLA2G1B activity on CD4 T cells is increased by several gC1qR-binding microbial proteins. (A, B) HCV core increases PLA2G1B activity on human CD4 T cells in a dose-dependent manner. [3H]-AA-labeled CD4 T cells were incubated alone or with PLA2G1B (63 nM or 200 nM) in the presence or not of several concentrations of recombinant HCV Core protein (11.9 nM-1190 nM) or equivalent concentrations of buffer alone (Buffer for A or 0 nM for B). (C) Staphylococcus aureus (SA) protein A increases PLA2G1B activity on human CD4 T cells in a dose-dependent manner. [3H]-AA-labeled CD4 T cells were incubated alone or with PLA2G1B (63 nM or 200 nM) in presence or not of several concentrations of SA protein A (23.8 nM-1190 nM). Results are shown as the mean ± SD from a pool of five experiments for A, and three experiments for B and C. **p < 0.01 and ***p < 0.001 by the Kruskal-Wallis test followed by the Mann-Whitney test, with p-values adjusted for multiple comparisons between groups.

Figure 4

The 3S/gp41 peptide regulates PLA2G1B activity in a gC1qR-dependent manner. (A, B) The HIV 3S peptide increases PLA2G1B activity on human Jurkat T cells. [3H]AA-labeled Jurkat T cells were pretreated with 3S gp41 (3S) or scrambled 3S (Scr 3S) peptides (11 µM) for various times (2, 4, or 21 h for A) or 21 h (for B) and incubated alone or with PLA2G1B (200 nM). (C) The gC1qR protein is detected in WT but not in gC1qR-deficient (gC1qR KO) Jurkat T cells by immunoblot. Two different anti-gC1qR mAbs were used (60.11 and 74.5.2), as well as anti-ß-actin mAb as an endogenous control of protein load. (D) The 3S peptide increases PLA2G1B activity on WT but not on gC1qR KO cells. WT and gC1qR KO [3H]-AA-labeled Jurkat T cells were pretreated with 3S or Scr 3S peptides at 11 μM for 21 h. Then, peptide-pretreated cells were incubated with PLA2G1B (200 nM) for 2 h. Results are expressed as the mean ± SEM of a pool of three (A, D) and seven (B) experiments performed in triplicate and as ΔPLA2G1B activity (activity with the 3S peptide minus that with the Scr 3S peptide). The level of [3H]AA released in the cell supernatant was quantified in cpm/mL. *p < 0.05, **p < 0.01 and ***p < 0.001 by two-way ANOVA with Tukey’s correction for multiple comparisons (A, D) or an unpaired t-test (B).

Figure 5

PLA2G1B is involved in PDAC plasma inhibition of the CD4 T-cell pSTAT5-NT response. (A) A 3S-like peptide from P. gingivalis (OmpA Pg) increases PLA2G1B enzymatic activity. [3H]-AA-labeled Jurkat T cells were pretreated with 3S, OmpA Pg or scrambled 3S (Scr 3S) peptides (11µM) for 21 h and incubated alone or with PLA2G1B (200 nM). Results are shown as the mean ± SD of PLA2G1B activity with the 3S or OmpA Pg minus PLA2G1B activity with Scr 3S from one representative experiment of two with similar results. (B) ELISA quantification of active PLA2G1B and proPLA2G1B in plasma from HD and PDAC donors (the median is shown). (C) Anti-PLA2G1B mAb inhibits PLA2G1B activity in PDAC plasma. HD CD4 T cells from three donors were treated with PLA2G1B (75 nM), 3% of PDAC plasma (PDACp) alone (w/o Ab, n = 9 plasma samples), with control isotype (ctrl iso, 667 nM, n = 6 plasma samples) or anti-PLA2G1B mAb (14G9, 667 nM, n = 8 plasma samples) and the pSTAT5-NT cell response to IL-7 was analyzed by confocal microscopy. Results are shown as the mean ± SD of percentage of pSTAT5-NT cells inhibition. (D) Heterogeneity of anti-PLA2G1B mAb 14G9 inhibition of PLA2G1B activity in PDAC plasma. Results are shown as the mean ± SD of the percentage of inhibition of PDACp activity from six patients on pSTAT5-NT by 14G9 relative to that of control isotype-treated plasma. *p < 0.05, and **p < 0.01, by two-way ANOVA with Tukey’s correction for multiple comparisons for A, by the Mann-Whitney test for B and D, and by the Kruskal-Wallis test followed by the Mann-Whitney test, with p-values adjusted for multiple comparisons between groups, for C. ns, non significant.