Tryptophan Metabolite Indole-3-Aldehyde Induces AhR and c-MYC Degradation to Promote Tumor Immunogenicity

Abstract

The role of tryptophan (Trp) and its metabolites in immune regulation is well-established, however, whether and how they regulate immunogenicity of tumor cells is not completely understood. In this study, a range of Trp metabolites are evaluated for their potential to modulate tumor immunogenicity using a co-culture assay with tumor cells and T cells. Indole-3-aldehyde (I3A) is identified as an indole derivative that significantly enhances tumor immunogenicity both in vitro and in vivo. This enhancement is attributed to the upregulation of antigen presentation and immunogenic molecules on tumor cells by I3A, thereby promoting their immunogenicity. Mechanistically, I3A induces the activation and degradation of Aryl hydrocarbon receptor (AhR), leading to increased expression of MHC-I molecules on tumor cell surfaces. Meantime, I3A induces rapid degradation of c-MYC in tumor cells and further enhances T cell activation. In mouse melanoma and lymphoma models, I3A demonstrates immune-dependent antitumor effects and enhances the efficacy of adoptive OT-I T cell therapy. Moreover, overexpression of the Trp metabolic enzyme interleukin-4-induced gene-1 (IL4I1) in tumor cells increases the intracellular level of I3A and enhances tumor immunogenicity. In summary, I3A is identified as a tumor immunogenicity inducer, which holds the potential to enhance antitumor immunotherapy efficacy.

Keywords: AhR; c‐Myc; indole‐3‐aldehyde; tryptophan metabolism; tumor immunogenicity.

Figure 1

I3A pre‐treated tumor cells induce T cell activation. A) The schematic diagram illustrating the tumor immunogenicity screening assay based on IL‐2 promoter‐driven LacZ reporter. EG7 cells were treated with different metabolites for 18 h, followed by PBS washing and co‐cultured with B3Z T cells in 96‐well plates in a ratio of 1:1 for 24 h, then the OD590 reading value was detected as LacZ reporting activity; B) EG7 cells were treated with tryptophan metabolites (500 µM and 1000 µM) for 18 h, followed by PBS washing and co‐cultured with B3Z T cells for additional 24 h. T cell activation was reflected by IL‐2 driven LacZ activity and IL‐2 ELISA assay. C) EG7 and B16‐OVA cells were treated with I3A (500 and 1000 µm) for 18 h, followed by PBS washing and co‐cultured with naïve OT‐I T cells for 24 h, then the IFNγ production of naïve OT‐I T cells was measured by ELISA. D) EG7 cells were treated with I3A (500 and 1000 µm) for 18 h, followed by co‐culturing with naïve OT‐I T cells labeled with CFSE for an additional 3 days; CFSE signal was detected by FACS. E,F) EG7 cells were pretreated with I3A and then co‐cultured with naïve OT‐I T cells for 24 h, the expression of GZMB, IFNγ, CD25, and CD69 on naïve OT‐I cells was detected by FACS. Bar graphs represent the average ± SEM. p‐values were derived from unpaired Student's t‐test or one‐way ANOVA analysis of variance with Bonferroni's post‐test for panels B–D. ns, not significant; ** p < 0.01; **** p < 0.0001. Panel B–F were representative results of at least 3 biological replicates.

Figure 2

I3A treatment induces immunogenic markers on tumor cells. A) EG7 and EL4 tumor cells were treated with I3A (1000 µm) for 18 h, followed by PBS washing and co‐cultured with B3Z T cells for 24 h. The LacZ level was measured; B) B16‐OVA and B16 tumor cells were treated with I3A for 18 h, followed by PBS washing and co‐cultured with naïve OT‐I T cells for 24 h. The IFNγ level was measured by ELISA assay; C,D) EG7 and B16‐OVA cells were treated with I3A for 18 h, then the expression of MHC‐I and MHC‐I‐SIINFEKL were detected by FACS. The expression of MHC‐I‐SIINFEKL complex on tumor cells was detected by staining with a monoclonal antibody recognizing OVA_257‐264 (SIINFEKL) peptide bound to H‐2Kb. E,F) EG7 and B16‐OVA cells were treated with I3A for 18 h, then the expression of B2M was detected by WB. G,H) EG7 cells were treated with I3A for 12 h, and the expression of antigen‐presenting machinery genes and lysosome‐related genes were measured by q‐PCR. Bar graphs represent the average ± SEM. p‐values were derived from unpaired Student's t‐test or one‐way ANOVA analysis of variance with Bonferroni's post‐test for panel A‐H. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. The results shown were representative results of at least 3 biological replicates.

Figure 3

IL4I1 catalyzes tumor‐intrinsic I3A production to induce tumor immunogenicity. A) The enzyme IL4I1 catalyzes tryptophan (Trp) and phenylalanine (Phe) to I3P and phenylpyruvic acid; B) EG7 tumor cells were treated with I3A for 18 h, then the intracellular concentrations of Phe, Trp and I3A in EG7 tumor cells were detected by LC‐MS/MS; C) The EG7 tumor cells which were overexpressed IL4I1 or empty vector (EV), were used for q‐PCR and WB assay to verify the overexpression of IL4I1. D) EG7 tumor cells expressing EV or IL4I1 were treated with I3A for 18 h, followed by PBS washing and co‐cultured with naïve OT‐I cells for 24 h, and the IFNγ level was measured by ELISA assay. E) Protein extracts of EG7 cells overexpressing IL4I1 WT or IL4I1 K351A mutant were collected, and the expression level of IL4I1 was measured by Western blot; F) EG7 tumor cells expressing EV/IL4I1‐WT/IL4I1 K351A were co‐cultured with naïve OT‐I cells for 24 h, then the IFNγ level was measured by ELISA assay. G) The intracellular concentrations of tryptophan metabolites of EG7 cells expressing EV/IL4I1‐WT/IL4I1 K351A mutant were detected by LC‐MS/MS and shown in fold change. H) Correlation of IL4I1 expression with MHC‐I and B2M in SKCM patients and correlation of IL4I1 expression with CTL activation signature (GZMA, GZMB, CD8A, and IFNG) in SKCM patients. I) The correlation of the IL4I1 expression level with the overall survival of SKCM patients, with p‐values calculated using the Log‐Rank test. Bar graphs represent the average ± SEM. P values were derived from unpaired Student's t‐test or one‐way ANOVA analysis of variance with Bonferroni's post‐test for panel B‐G. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. The results shown were representative results of at least 3 biological replicates.

Figure 4

I3A inhibits tumor growth via a T cell‐dependent manner. A–C) Schematic illustration of mouse tumor model for I3A treatment; EG7 cells (1 million per mouse) A), B16‐OVA cells (0.2 million per mouse) B) and B16 cells (0.2 million per mouse) C) were s.c. inoculated in B6 mice, followed by intraperitoneal (i.p.) injection of I3A (50 mg kg⁻¹) or vehicle solvent (Ctrl) on day 3 for EG7 tumor or day 5 for B16/B16‐OVA tumor, once every other day, then the tumor growth was recorded. EG7 tumor: n = 6 (Ctrl), n = 7 (I3A); B16‐OVA: n = 6 per group, B16: n = 7 per group; D,E) Mice were i.p. injected with I3A or vehicle from day 5 or day 3 after B16 and EG7 tumor inoculation, and the tumors were harvested on day 15 for tumor‐infiltrating lymphocytes (TILs) isolation for IFNγ ELISpot assay. The TILs from B16 tumors were stimulated by BMDCs primed with B16 tumor lysate D), or stimulated by OVA_257‐264 peptide for TILs from EG7 tumors E); n = 5 for EG7 tumors and n = 4 for B16 tumors; F) EG7 cells (1 million per mouse) were s.c. inoculated in nude mice, I3A (50 mg kg⁻¹) or vehicle were i.p. injected once every other day, then the tumor growth was recorded; n = 5 per group. G,H) EG7 cells (1 million per mouse) were s.c. inoculated in C57 mice, followed by I3A (50 mg kg⁻¹) or vehicle treatment (i.p. injection, every other day); anti‐CD8 depletion antibody (100 µg/mouse) were injected at days 3, 6, and 9, then the tumor growth was recorded G), and the tumors were weighed at the end of the experiment H); n = 8 per group. I) B16‐OVA cells (0.2 million per mouse) were s.c. inoculated in nude mice, I3A (50 mg kg⁻¹) or vehicle were i.p. injected once every other day, and then the tumor growth was recorded. n = 6 per group. J,K) B16‐OVA cells (0.2 million per mouse) were s.c. inoculated in B6 mice, followed by I3A (50 mg kg⁻¹) or vehicle treatment (i.p. injection, every other day); anti‐CD8 depletion antibody was injected at day 4, 7, and 10, then the tumor growth was recorded J) and tumors were weighed at the end of the experiment K). n = 7 for CD8 depletion group, n = 6 for CD8 depletion + I3A group. The tumor weight was recorded in K). Bar graphs represent the average ± SEM. p‐Values were derived from unpaired Student's t‐test or two‐way ANOVA analysis of variance with Bonferroni's post‐test for panel A‐K. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5

I3A downregulates AhR to boost T cell activation. A) B16‐OVA cells were transfected with XRE‐Luc for 24 h, then treated as indicated for additional 24 h, then the XRE‐luc luciferase activity was measured; CH:CH223191, AhR inhibitor, 50 µm; B,C) EG7 and B16‐OVA cells were treated by I3A or I3A together with AhR inhibitor for 18 h, then co‐cultured with T cells for 24 h, then the IL‐2 promoter activity and IFNγ production level was measured by LacZ assay B) and ELISA C); D) B16‐OVA cells were treated with AhR‐targeting siRNAs for 48 h or treated with I3A for 18 h, followed by co‐cultured with naïve OT‐I cells for 24 h, then the supernatant level of IFNγ was measured by ELISA; E) EG7 cells expressing AhR shRNA were co‐cultured with B3Z cells for 24 h, then the LacZ activity was measured; F) the expression level of MHC‐I‐SIINFEKL complex and MHC‐I molecule on B16‐OVA expressing siAhR were measured by FACS; G,H) B16‐OVA cells were treated with I3A at different concentrations for 18 h, or B16‐OVA cells were treated by I3A or I3A together with AhR inhibitor, then the expression of AhR was detected by Western blot. I) The expression of Cyp1a1 was measured by q‐PCR assay in B16‐OVA cells expressing AhR‐WT/dNLS/CA mutant; CA: constitutively activated, dNLS: nuclear localization signal deleted. J) Re‐expression of AhR‐WT, AhR‐dNLS, or AhR‐CA mutant protein in B16‐OVA siAhR cells was detected by Western blot. K) T cells were co‐cultured with indicated B16‐OVA cells for 18 h, and then the supernatant IFNγ levels were detected by ELISA. L) B16‐OVA cells were treated with AhR‐targeting siRNAs for 48 h and treated with I3A for 18 h, then the expression of proteins was detected by Western blot. M) B16‐OVA cells were treated with I3A and JAK inhibitor for 18 h, then the expression levels of proteins were detected by WB. JAK1/2 inhibitor: Ruxolitinib. N) EG7 cells were treated with I3A and JAK inhibitor for 18 h, then the expression levels of MHC‐I were detected by FACS. O) The shStat3 B16‐OVA cells were treated with I3A or not for 18 h, then co‐cultured with naïve OT‐I cells for 24 h, the level of IFNγ was measured by ELISA assay. ns, not significant; Bar graphs represent the average ± SEM. p‐Values were derived from unpaired Student's t‐test or one‐way ANOVA analysis of variance with Bonferroni's post‐test for panel A‐O. ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. The results shown were representative results of at least 3 biological replicates.

Figure 6

I3A treatment induces c‐MYC degradation in tumor cells. A) Enrichment plots of c‐MYC pathways in EG7 cells which were treated with I3A; B) Gene set enrichment analysis (GSEA) of c‐MYC pathways in EG7 cells treated with I3A; C) B16‐OVA and EG7 cells were treated by I3A for 1 hour and c‐Myc mRNA expression were detected by q‐PCR; D) EG7 and B16‐OVA cells were treated by I3A for 18 h and c‐MYC expression were detected by Western blot. E,F) EG7 cells expressing scramble (SCR) or c‐Myc targeting shRNAs (sh1, sh2) and B16‐OVA cells were transfected with siMyc for 48 h, these cells were treated with I3A or not for 18 h, then co‐cultured with naïve OT‐I cells for 24 h, the level of IFNγ was measured by ELISA assay. G,H) B16‐OVA cells were transfected with different siRNA for 48 h, and co‐cultured with naïve OT‐I T cells for 24 h, then the supernatant levels of IFNγ were measured; I) B16‐OVA cells were transfected with siAhR and siMyc for 48 h, and treated with I3A for 24 h, the p‐STAT3 and STAT3 protein levels were detected. Bar graphs represent the average ± SEM. p‐values were derived from unpaired Student's t‐test or one‐way ANOVA analysis of variance with Bonferroni's post‐test for panel C‐H ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. The WB results shown were representative results of at least 3 independent experiments.

Figure 7

I3A treatment potentiates the antitumor efficacy of adoptive OT‐I T cell therapy. A–D) Schematic illustration of I3A and OT‐I cell transfer treatment plan on mouse melanoma and lymphoma tumor models; B16‐OVA cells (0.2 million per mouse) A,C) and EG7 cells (1 million per mouse) B,D) were s.c. inoculated on B6 mice, then treated with I3A or in combination with activated OT‐I (2 million, tail vein injection), tumor growth was recorded, For B16‐OVA model, n = 5 for Ctrl and I3A+OT‐I groups, and n = 6 for I3A and OT‐I groups; For EG7 model, n = 6 for each group. E) The tumors were isolated from mouse, and tumor weights were recorded. F) Another batch of experiments was performed as in (B), and the Kaplan‐Meier survival curves were plotted and the significance was analyzed by log‐rank test, the endpoint is when the tumor volume reached 2000 mm³, n = 6 per group. Bar graphs represent the average ± SEM. p‐values were derived from one‐way or two‐way ANOVA analysis of variance with Bonferroni's post‐test for panel C–E. *p < 0.05; ***p < 0.001; ****p < 0.0001.