Myeloid-derived suppressor cells inhibit T cell activation through nitrating LCK in mouse cancers

Abstract

Potent immunosuppressive mechanisms within the tumor microenvironment contribute to the resistance of aggressive human cancers to immune checkpoint blockade (ICB) therapy. One of the main mechanisms for myeloid-derived suppressor cells (MDSCs) to induce T cell tolerance is through secretion of reactive nitrogen species (RNS), which nitrates tyrosine residues in proteins involved in T cell function. However, so far very few nitrated proteins have been identified. Here, using a transgenic mouse model of prostate cancer and a syngeneic cell line model of lung cancer, we applied a nitroproteomic approach based on chemical derivation of 3-nitrotyrosine and identified that lymphocyte-specific protein tyrosine kinase (LCK), an initiating tyrosine kinase in the T cell receptor signaling cascade, is nitrated at Tyr394 by MDSCs. LCK nitration inhibits T cell activation, leading to reduced interleukin 2 (IL2) production and proliferation. In human T cells with defective endogenous LCK, wild type, but not nitrated LCK, rescues IL2 production. In the mouse model of castration-resistant prostate cancer (CRPC) by prostate-specific deletion of Pten, p53, and Smad4, CRPC is resistant to an ICB therapy composed of antiprogrammed cell death 1 (PD1) and anticytotoxic-T lymphocyte-associated protein 4 (CTLA4) antibodies. However, we showed that ICB elicits strong anti-CRPC efficacy when combined with an RNS neutralizing agent. Together, these data identify a previously unknown mechanism of T cell inactivation by MDSC-induced protein nitration and illuminate a clinical path hypothesis for combining ICB with RNS-reducing agents in the treatment of CRPC.

Keywords: LCK; immune checkpoint blockade; myeloid-derived suppressor cells; prostate cancer; protein nitration.

Fig. 1.

Intratumoral MDSCs induce protein nitration. (A) Comparison of relative nitrite concentration in Pten/p53/Smad4 spontaneous tumor with wild-type (WT) prostate (Left) and LLC s.c. tumor with WT lung (Right) (n = 3). (B) Western blot of 3-NT of WT prostate and Pten/p53/Smad4 spontaneous tumor. (C) Western blot of 3-NT of WT lung and LLC s.c. tumor. Na₂S₂O₄ treatment of the blot membrane converted 3-NT to 3-aminotyrosine and eliminated signals. (D) Primary T cells isolated from WT mouse spleen were stimulated by anti-CD3/CD28 condition for 48 h in the absence or presence of MDSCs isolated from Pten/p53/Smad4 tumor or spleen. Peroxynitrite at 50 μM or uric acid at 50 μM were added as indicated. Whole cell lysates of isolated T cells were immunoblotted for 3-NT. (E) Primary T cells were treated as in D, and the IL2 concentration of each sample was measured in the medium using ELISA (n = 2). In A and E, data represent mean ± SD. *P < 0.05, ***P < 0.001, Student’s t test.

Fig. 2.

Methods for nitropeptide enrichment and identification of nitro-Tyr394 peptide LIEDNEYTAR in LCK from both mouse models. (A) The workflow of the chemical derivation method for detecting nitrated Ang II. (B) The MS of the modified LCK peptide from T cells of the Pten/p53/Smad4 tumors. The monoisotope peak at m/z 706.33 matched the duple-charged, CH₂O-labeled peptide. (C) The MS of the modified LCK peptide from T cells of the LLC tumors. The monoisotope peak at m/z 708.35 matched to the duple-charged, CD₂O-labeled peptide. (D and E) The MS/MS of B and C, respectively. (F) The intensities of y₆, y₇, and y₈ ions of LIEDNEY(nitro)TAR in stimulated T cells (black), stimulated T cells cocultured with MDSCs (red) and stimulated T cells treated with peroxynitrite (blue) (n = 3). The data were acquired by MRM-MS, and each sample was performed by two technical replicates. (G) Normalized abundance of LIEDNEY(nitro)TAR in three cell conditions as in F by using two reference peptides from actin (see Methods). Data represent mean ± SD. **P < 0.01, Student’s t test.

Fig. 3.

Effects of nitration on LCK and T cell activation. (A) Brief scheme of T cell activation pathway. (B) Relative IL2 concentration in medium of Jurkat cells stimulated with PMA/ionomycin (blue) or anti-CD3/CD28 antibodies (red) in the presence of different concentrations of peroxynitrite, measured by ELISA (n = 2). (C) Immunoblot analysis of phosphotyrosine levels of a number of signaling proteins in the T cell activation pathway, with lysates purified from Jurkat cells stimulated with anti-CD3/CD28 antibodies in the presence of peroxynitrite. (D) The intensities of y₆, y₇, and y₈ ions of LIEDNEY(nitro)TAR in stimulated primary human T cells (black), stimulated T cells cocultured with MDSCs from Pten/p53/Smad4 tumors (red), and stimulated T cells treated with peroxynitrite (blue) (n = 3). The data were acquired by MRM-MS, and each sample had two technical replicates. (E) Normalized abundance of LIEDNEY(nitro)TAR in three cell conditions as in D by using two reference peptides from actin. (F) The MS/MS of tryptic nitro–Tyr394-containing peptide in recombinant LCK. The parent ion is at 634.79 m/z in duple charge. (G) Relative in intro enzymatic activity of recombinant human LCK inducted with different concentrations of peroxynitrite. Under each concentration of peroxynitrite treatment, the enzyme activity was measured twice. (H) Effect on IL2 production by J.CaM1.6 cells, which were stimulated with anti-CD3/CD28 antibodies and transfected with unmodified or nitrated LCK. Data were normalized to PMA/inomycin stimulation (n = 3). In B, E, G, and H, data represent mean ± SD. **P < 0.01, ***P < 0.001, Student’s t test.

Fig. 4.

Combinational therapy of the mouse CRPC and LLC model. (A) Representative 3-NT IHC images of adjacent normal or CPRC samples (n = 15) and the costaining pattern of 3-NT with pan-cytokeratin (pan-CK) or CD3. (Scale bar, 100 μm.) DAB and VIP refer to 3, 3-diaminobenzidine and violet peroxidase substrate, respectively. The black open circle in top right image was an accidentally trapped air bubble. (B) Preclinical trial result of prostate tumor mass of the Pten/p53/Smad4 CRPC model (n = 6, 5, 6, 5, respectively). (C) Representative prostate images of the preclinical trial in B. (Scale bar, 3 mm.) (D) Immunoblot analysis of 3-NT levels in whole prostate lysates from the preclinical trial in B. (E) Quantification of infiltrating CD8⁺ T cells based on CD8 IHC of CRPC samples from the preclinical trial in B. (F) Quantification of infiltrating T_reg cells based on Foxp3 IHC of CRPC samples from the preclinical trial in B. (G) Ratios of CD8⁺ T cell numbers over T_reg cell numbers for each treatment condition in the preclinical trial in B. (H) Preclinical trial result of the LLC s.c. tumor model (n = 8–10 for each group). In B, ***P < 0.001, Mann–Whitney U test. In E, F, and H, ***P < 0.001, ^#P > 0.05, Student’s t test.