Introduction:

M2 polariztaion of tumor associated macrophages contribute to T cell exhaustion and worse clinical outcomes in ccRCC patients. The factors that enable macrophage polarization, and subsequent tumor growth, remain unknown. Recently, our group has identified a novel splice variant of RNASET2, a protein normally contributing to anti-tumor, M1 macrophage activity. This splice variant was associated with worse clinical outcomes in ccRCC patients. We set out to determine the distribution of M2 macrophages (CD163), expression patterns of RNASET2, and their relationship with the RNASET2 splice variant. We hypothesize that the splice variant leads to abnormal protein formation, allowing for a tumor microenvironment that favors growth.

Methods:

We obtained primary tumor samples from 72 patients with ccRCC. All patients in this cohort underwent either partial or radical nephrectomy for clinically localized, locoregional or oligo-metastatic ccRCC. A database containing clinical characteristics, pathologic and histologic data, and survival outcomes was constructed to provide clinical correlates. Tissue sections from these specimens were obtained from the tumor core, tumor edge (tumor-stromal interface), and stroma (Figure 1). Immunohistochemistry was performed, staining for RNASET2, M2 macrophages (CD 163), and markers for T cell exhaustion (TIM3, CD8). In addition, RNA sequencing data were utilized to identify the RNASET2 splice variant expression in the tumors of 69 patients. Expression patterns of these cellular markers were analyzed, and then stratified based on the presence/absence of the splice variant.

Results:

Our analysis of the tumor core and tumor-stromal interface demonstrated novel molecular associations. The expression of CD163 was inversely correlated with expression of RNASET2 in the tumor core and tumor edge. This relationship can be seen in Figure 2 with the associated correlation coefficient provided within each plot (** represents statistical significance). Furthermore, when patients were stratified based on RNASET2 splice variant expression, the strong correlation between CD163 and RNASET2 was weakened. Finally, when evaluating expression as a prognostic indicator, high co-expression of CD163 and RNASET2 led to significantly worse overall survival relative to low co-expression. This finding was observed in the tumor (47.0 mo vs 138.2 mo, p<0.001), tumor edge (79.5 mo vs 174.6 mo, p=0.001) and stroma (15.3 mo vs 125.2 mo, p<0.001).

Conclusion:

These data represent the first evidence that a relationship exists between M2 macrophages and RNASET2 in ccRCC. We have demonstrated through staining of patient tissue samples that high RNASET2 expression is correlated with low M2 polarization of macrophages, and vice versa. Furthermore, in the presence of splice variant expression, the relationship between RNASET2 and M2 polarization is weakened, suggesting that this event may disrupt the normal immune-mediated inhibition of tumor growth. Further work is underway to elucidate the mechanism(s) behind this disruption in the tumor microenvironement.

Funding: Moffitt Division of Quantitative Sciences Team Science AwardP30-CA076292