Dynamic proteomic change of tumor and immune organs in an immune-competent hepatocellular carcinoma mouse model

Abstract

Subcutaneous implantation of a human cancer cell line in immune-deficient mice (CDX) is a commonly used tool in preclinical studies for the assessment of potential anti-cancer drugs. As immunotherapy is transforming cancer treatment, tumor models in immunocompetent mice are necessary for us to understand the immune aspects of tumor biology. However, the systemic immune response to the implantation of cancer cells at proteome level is unclear. In this study, we characterized the dynamic proteomic changes of subcutaneous tumors and 5 immune organs (draining lymph node, mesenteric lymph node, spleen, thymus and marrow) at six time points after implantation using a Hepa1-6 derived allograft mouse model. Our data suggest that interaction of the implanted tumor cells with mouse immune system followed the trajectory of "tumor rejection" to "immune evasion" in that the tumor gained the ability to evade the immune system for growth. Furthermore, anti-PDL2 antibody was validated here as an optional immunotherapy strategy to inhibit the growth of Hepa1-6 subcutaneous tumors. These findings from our study provided valuable information for the understanding of tumor and immune interaction and shed light on the rational design for clinical cancer treatment and other preclinical experiments.

Keywords: Hepa1-6; PDL2; allograft mouse model; proteomics.

Figure 1

The general workflow and proteomic landscape of developing tumor and immune organs from Hepa1-6 subcutaneous allograft mouse model. A. I. The establishment of HCC mouse model and experiment outline. Subcutaneous allograft was established with Hepa1-6 cells in 2-3 m old female mice (n=18). Six tissue types were collected at different days after cancer cell inoculation as indicated in the figures. Icons was made by Freepik from www.flaticon.com. II. Extraction and digestion of tissue proteome. III. MS-based identification and quantitative proteomics. IV. Bioinformatics analysis of multiple datasets. B. Multiple data sets with different filtering criteria and venn diagram of proteins in six tissues across 6 timepoints. C. Heatmap analysis of temporal proteomic data in six tissues by unsupervised clustering and Correlation’s hierarchical clustering. D. PCA analyses of the proteome data on the five immune organs’ samples. E. Heatmap of the correlation between the module eigengenes and traits of the tissue by WGCNA.

Figure 2

Protein expression dynamics during tumor growth. (A) Protein trajectories of the 5 clusters in tumor, Fuzzy C-Means Clustering. (B) The Metascape term enrichments for proteins of each cluster in (A). (C) The iFOT of Banf1 in 6 tissues.

Figure 3

Upregulated protein expression during Hepa1-6 cell derived tumor growth in immune-competent mice. A. Venn diagram of proteins in cluster 4 of the proteins from tumor and immune organs in GO: 002376. B. The iFOT of 9 proteins in tumor including Cxcl10, Nos2, Pdcd1lg2, Ido1, Il1rn, Oxsr1, C1qc, Clec4n, Gpsm3. C. Validation of anti-tumor efficacy by inhibiting PDL2 in Hepa1-6 mouse model. The tumor growth was inhibited by intraperitoneal injection of anti-PDL2 antibody. Error bars are created by mean ± SD of four replicates.

Figure 4

PDL2 co-expression modules during tumor growth. (A) Heatmap of the 23 proteins in MEpink module of WGCNA. ME, module eigengene. (B) Co-expression modules of PDL2. Spearman’s correlation coefficient (R ≥ 0.7, t-test, P-value ≤ 0.05). (C) Venn diagram of proteins in two PDL2 modules and immune proteins in GO: 002376. (D) String analysis (medium confidence) of the common immune protein in (C). (E) The iFOT of Padi4 in 6 tissues.