NLRC5-CIITA Fusion Protein as an Effective Inducer of MHC-I Expression and Antitumor Immunity

Abstract

Aggressive tumors evade cytotoxic T lymphocytes by suppressing MHC class-I (MHC-I) expression that also compromises tumor responsiveness to immunotherapy. MHC-I defects strongly correlate to defective expression of NLRC5, the transcriptional activator of MHC-I and antigen processing genes. In poorly immunogenic B16 melanoma cells, restoring NLRC5 expression induces MHC-I and elicits antitumor immunity, raising the possibility of using NLRC5 for tumor immunotherapy. As the clinical application of NLRC5 is constrained by its large size, we examined whether a smaller NLRC5-CIITA fusion protein, dubbed NLRC5-superactivator (NLRC5-SA) as it retains the ability to induce MHC-I, could be used for tumor growth control. We show that stable NLRC5-SA expression in mouse and human cancer cells upregulates MHC-I expression. B16 melanoma and EL4 lymphoma tumors expressing NLRC5-SA are controlled as efficiently as those expressing full-length NLRC5 (NLRC5-FL). Comparison of MHC-I-associated peptides (MAPs) eluted from EL4 cells expressing NLRC5-FL or NLRC5-SA and analyzed by mass spectrometry revealed that both NLRC5 constructs expanded the MAP repertoire, which showed considerable overlap but also included a substantial proportion of distinct peptides. Thus, we propose that NLRC5-SA, with its ability to increase tumor immunogenicity and promote tumor growth control, could overcome the limitations of NLRC5-FL for translational immunotherapy applications.

Keywords: B16-F10; EL4; MHC-I; MHC-I associated peptides; NLRC5; NLRC5-SA; tumor immunogenicity.

Figure 1

MHC-I induction by NLRC5-SA in mouse and human cancer cell lines. (A) The NLRC5-SA construct. NLRC5 superactivator (NLRC5-SA) was generated by fusing the N-terminus atypical CARD (aCARD) domain of full-length human NLRC5 (NLRC5-FL) to the central NACHT and C-terminus LRR regions of NLRA (CIITA), thereby retaining its tripartite structure. (B) Cell surface MHC-I expression in mouse B16.F10 melanoma cells stably expressing control vector, NLRC5-FL or NLRC5-SA expression vectors evaluated by flow cytometry using a bi-specific H-2Kb/Db antibody. (C) MHC-I expression in mouse EL4 lymphoma cells expressing control vector, NLRC5-FL or NLRC5-SA expression constructs. (D) Cell surface expression of HLA-ABC in A549 lung cancer and T47D breast cancer cell lines stably expressing NLRC5-FL or NLRC5-SA labeled with a pan-HLA antibody.

Figure 2

Control of B16 melanoma and EL4 lymphoma tumor growth by NLRC5-SA. Syngeneic C57BL/6 mice were inoculated subcutaneously with 1 × 10⁵ B16-vector, B16-NLRC5-FL or B16-NLRC5-SA (A–C; n = 7–11 per group) or 5 × 10⁵ EL4-vector, EL4-NLRC5-FL or EL4-NLRC5-SA cells (D–F; 4 mice per group). Tumor growth curve, monitored every 2–3 days (A,D). Tumor mass at the endpoint (B,E). Representative tumors formed by each tumor line at the endpoint (C,F). Statistical comparisons (in A,B,D,E): mean ± SE; two-way ANOVA with Tukey’s multiple comparison test. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

Figure 3

Loss of MHC-I expression in EL4-NLRC5-SA cells following elution of MHC-I-associated peptides (MAPs). (A) Selective destabilization of MHC-I expression by acid elution of MAPs. Parental EL4 cells were incubated with citrate phosphate buffer for 2 min, and the expression of MHC-I and the indicated cell surface markers was examined by flow cytometry. Representative data from three experiments are shown. Untreated (red) and unstained samples (grey) cells were used as controls. (B) Loss of MHC-I upon acid elution in NLRC5-expressing EL4 cells. EL4 cells expressing empty vector, NLRC5-FL or NLRC5-SA, stimulated or not with 20 ng/mL mIFNγ for 24 h, were subjected to acid elution, and MHC-I expression was measured by flow cytometry. MFI, Mean fluorescence intensity. Mean and standard deviation from three experiments. Turkey’s multiple comparison test: * p ≤ 0.05, *** p ≤ 0.001, **** p ≤ 0.0001.

Figure 4

MHC-I-associated peptide repertoire of NLRC5-expressing EL4 cells. (A) MAPs were eluted from EL4 cells expressing empty vector, NLRC5-FL or NLRC5-SA, stimulated or not with 20 ng/mL IFNγ for 24 h. The eluted MAPs were identified by the PEAKS software to include peptides from known protein database as well as de novo peptides. Number of all MAPs and 8–11 amino acid long MAPs are shown in (A,B), and the peptide size distribution in (C). The number of 12–13aa long peptides are indicated within the 12–15aa long MAPs. The lists of 8–13aa long peptides from all samples are given in Supplementary Table S2.

Figure 5

NLRC5-SA and NLRC5-FL differentially modulate MAPs in EL4 cells. (A,B) Shared and unique 8–13aa MAPs in EL4 cells expressing NLCR5-FL, NLRC5-SA or control vector at steady state (A) and after IFNγ stimulation (B). The lists of 8–13aa peptides used in (A,B) are given in Supplementary Table S3. (C–E) Pairwise comparison of 8–13aa MAPs eluted from unstimulated and IFNγ-stimulated EL4-vector (C), EL4-NLRC5-FL (D) and EL4-NLRC5-SA (E) cells.

Figure 6

NLRC5-SA significantly differs from NLRC5-FL in modulating steady-state and IFNγ-stimulated MAPs. (A) Comparison of 8–13aa MAPs that are not affected by IFNγ stimulation in each of the EL4-derived cell lines. The lists of 8–13aa peptides used in (A) are given in Supplementary Table S4. (B) The 8–13aa MAPs shared between EL4-NLRC5-FL and EL4-NLRC5-SA cells under steady state or IFNγ-stimulated conditions but absent from EL4-vector cells. The lists of peptides are given in Supplementary Table S5.

Figure 7

Potential immunogenic MAPs that could be identified using NLRC5-SA. Six-way Venn diagram showing the comparison of MAPs from all the three cell lines at steady state and after IFNγ stimulation.