Hyperactive PLCG1 drives non-canonical signaling to promote cell survival

Abstract

One of the long-standing questions in cell signaling field to identify and characterize key signaling nodes out of a complex network. Phospholipase Cγ1 (PLCG1) was identified as the most frequently mutated gene in adult T-cell leukemia/lymphoma, suggesting a critical function of PLCG1 in driving T cell activation. However, it remains unclear how these mutations regulate T cell physiology and pathology. Here we investigated three common leukemia/lymphoma associated mutations (R48W, S345F, and D1165H). We discovered that these mutations induced hyperactive T cell signaling and caused pro-survival phenotypes. PLCG1 mutants enhanced LAT condensation, calcium influx, and ERK activation. They promoted T cell proliferation, induced cell aggregation, and rendered resistance to vorinostat, an FDA-approved drug for cutaneous T-cell lymphoma. The resistance to vorinostat depended on ERK signaling and can be reversed with an ERK inhibitor. Mechanistically, alpha smooth muscle actin, which was specifically induced by PLCG1 mutants, directly bound PLCG1 to promote its activation. Together, these results demonstrated that hyperactive PLCG1 promoted T cell survival and drug resistance through inducing non-canonical signaling.

Keywords: Actin; Condensation; ERK; PLCG1; T cell.

Fig. 1:. PLCG1 acquiring ATLL-associated mutations promoted LAT condensation in vitro.

(A) Location of the three T-cell leukemia/lymphoma-associated mutations in the structure of PLCG1 (PDB: 6PBC). (B) Schematics of biochemical reconstitution of LAT condensation on supported lipid bilayers. (C) TIRF microscopy revealed that PLCG1 mutations enhanced LAT condensation on bilayers at physiological concentrations. Cy3B-labeled LAT at 300 molecules / μm² was incubated with 300 nM Sos1, 3,000 nM Grb2 and 50 nM PLCG1 for 0.5 h before imaging. Scale bar: 5 μm. (D) Quantification of LAT clustering as normalized variation. Shown are mean ± SD. Unpaired two-tail t-test was used. *p < 0.05, **p < 0.01.

Fig. 2:. PLCG1 mutants increased TCR-triggered T cell activation.

(A) Schematics of imaging T cell activation in live cells. Jurkat T cells expressing LAT-mCherry and PLCG1 WT or mutants were plated and activated on OKT3 antibody-coated glass. The formation of LAT condensates on plasma membranes were monitored by time-lapsed TIRF microscopy. (B) Representative images of Jurkat T cells stimulated with glass-coated OKT3. Images shown were 45 sec after cell landing. Scale bar: 5 μm. (C) Quantification of LAT clustering during T cell activation. Shown are mean ± SEM. N=24 to 30 cells. Unpaired two-tail t-test was used for mutation groups compared to WT. *: p < 0.05, **: p < 0.01. (D) Immunoblot analysis of Jurkat T cells harboring PLCG1 WT or indicated mutated variant stimulated with anti-CD3 and anti-CD28 antibodies. (E) Quantification of (D). Shown are mean ± SD. Unpaired two-tail t-test was used. *p < 0.05, ***p < 0.001, ns: not significant. (F) Calcium flux monitored by flow cytometry following TCR activation. Jurkat T cells expressing calcium sensor GCaMP6s and PLCG1 WT or mutation were stimulated by OKT3 antibody and monitored by flow cytometry in a continuous recording mode. (G) Activation of human primary T cells expressing PLCG1 WT or mutants. The expression of CD69 was determined by flow cytometry 14 days after T cells were infected with lentivirus encoding PLCG1 WT or mutants. (H) Proliferation of human primary T cells expressing PLCG1 WT or mutants. The cell number was quantified 14 days after T cells were infected with lentivirus encoding PLCG1 WT or mutants. Shown are mean ± SD. **p < 0.01, ***p < 0.001.

Fig. 3:. Hyperactive PLCG1 was sufficient to trigger T cell activation and cytokine production without TCR engagement.

(A) Immunoblot analysis of signaling in Hut78 cells stably expressing PLCG1 WT or mutants (without TCR activation). (B) Quantification of (A). Shown are mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. (C) Profiling of protein kinase phosphorylation. Hut78 cells expressing PLCG1 WT or mutations were lysed, and applied to the proteome profiler human phospho-kinase array kit for analysis. The three kinases with altered phosphorylation were highlighted by blue arrows. (D) Quantification of the phosphorylation level of kinases. Shown are mean ± SD. **p < 0.01, ***p < 0.001. (E) ELISA analysis of IL2 and TGF-beta secretion by Hut78 cells. Shown are mean ± SD. **p < 0.01, ***p < 0.001.

Fig. 4:. Hyperactive PLCG1 signaling induced aggregation of Hut78.

(A) Aggregation of Hut78 cells expressing PLCG1 mutants. Scale bar: 100 μm. (B) Flow cytometry revealed ICAM-1 expression on Hut78 expressing PLCG1 WT or mutants. (C) Quantification of (B). GOM is geometric mean fluorescence intensity. Shown are mean ± SD. ***p < 0.001. (D) Blocking ICAM-1and LFA-1 interaction inhibited ERK activation. LFA-1 blocking antibody was added to Hut78 expressing PLCG1 D1165H for 36 h. (E) Quantification of (D). Shown are mean ± SD.

Fig. 5:. Hyperactive PLCG1 conferred Hut78 resistance to HDAC inhibitors.

(A) PLCG1 mutations conferred Hut78 resistance to vorinostat. The plain group is Hut78 cells without ectopically expressing PLCG1. The CCK8 assay was used to detect viable cell number after vorinostat treatment for 72 h. (B) PLCG1 mutations decreased vorinostat-induced apoptosis. Hut78 cells were incubated with 1.25 μM vorinostat for 60 h. (C) Quantification of (B). Shown are mean ± SD. **p < 0.01 (WT vs R48W, S345F, or D116H). (D) ERK inhibition mitigated vorinostat resistance in Hut78 cells. ERK inhibitor LY3214996 at noncytotoxic concentration 0.37 μM abolished the resistance to vorinostat in Hut78 expressing PLCG1 D1165H. Shown are mean ± SD. **p < 0.01, ns: not significant.

Fig. 6:. Hyperactive PLCG1 signaling induces a distinct gene profile from TCR activation.

(A) Venn plot illustrates the gene expression difference in the R48W, S345F and D1165H group as compared to the WT PLCG1. RNA-seq analysis was used to profile gene expression in Hut78 cells expressing WT or mutant PLCG1. Threshold: Log2 fold change>1 and FDR < 0.001 when comparing to the WT group. (B) Comparing gene expression profile between hyperactive PLCG1 signaling and TCR signaling. In the TCR group, Hut78 expressing PLCG1 WT was activated by anti-CD3/CD28 antibodies for 3 days. (C) Qiagen ingenuity pathway analysis (IPA) revealed pathways enriched in the gene set uniquely triggering by hyperactive PLCG1 signaling but not TCR signaling. (D) Heatmap illustrated genes uniquely triggered by hyperactive PLCG1 signaling. Shown are the 72 genes which are above 3 folds change of D1165H vs TCR stimulation from the unique 163 genes. Red rectangles indicate genes belong to the smooth muscle contraction pathway. The full list of genes is in supplemental table2. (E) Immunoblot analysis of actin isoform expression in Hut78 cells. WT+Stimulation: PLCG1 WT was stimulated by anti-CD3/CD28 antibodies for 2 days. (F) Quantification of the actin expression. Shown are mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant.

Fig. 7:. Alpha smooth muscle actin-dependent activation of ATLL mutations of PLCG1.

(A) Immunoprecipitation assay to identify PLCG1-binding partners. Hut78 cells expressing PLCG1-GFP (WT or mutant) were lysed, pulled down by beads coated with protein G and an anti-GFP antibody-coated, and applied for SDS-PAGE. The specific bands stained by Coomassie Blue were cut and identified using mass spectrometry. The plain sample is Hut78 cells without exogenously expressed PLCG1. (B) Pelleting assay to determine the direct interaction between PLCG1 and filamentous actin in vitro. Input: 4 μM α-SMA (pre-assembled into filaments), 0.4 μM PLCG1, and 1 μM myosin. (C) Quantification of PLCG1 binding to F-actin. Shown are mean ± SD. ***p < 0.001, ns: not significant. (D) Regulation of PLCG1 phosphorylation by filamentous actin in Hut78 cells. Actin filaments were destabilized and stabilized with the treatment of 0.5 μM latrunculin A (Lat A) or 0.15 μM jalapinolate (Jas), respectively for 0.5 h. (E) Quantification of PLCG1 phosphorylation. Shown are mean ± SD. *p < 0.05.