Integrative proteomics and genomics supports a role for interferon gamma in the pathogenesis of Kaposi sarcoma and finds multiple candidate diagnostic proteins for early detection or prevention

Background

Kaposi sarcoma (KS) is a common and morbid condition among persons with HIV infection. Strategies for preventing KS or designing better treatment regimens would be aided by the identification of biomarkers for development or progression of KS. Due to the vascular and often disseminated nature of KS, proteomic signatures detected from and specific to KS tumors may yield viable candidate diagnostic protein markers and insights into the pathogenesis of KS.

Methods

Flash-frozen punch biopsies from cutaneous samples of tumor and normal skin of individuals having epidemic (HIV-positive) or endemic (HIV-negative) KS were profiled using tandem mass spectrometry. Protein data were integrated with previously existing databases relevant for prioritizing diagnostic marker candidates, including plasma proteome data of cancer-free individuals, normal endothelial cells, and microarrays profiling the mRNA of KS, normal skin, KSHV-infected endothelial cells, and uninfected cells.

Results

In a comparison of 13 HIV+ tumor punch biopsies to 5 HIV+ normal samples, 5100 total proteins were identified. We have identified proteins abundant in the majority of tumor samples and absent from normal skin and plasma of cancer-free individuals that may serve as candidate diagnostic markers. Aligning the data with KS tissue expression data produced 72 proteins that were over-abundant in KS samples in both proteomic and transcriptomic datasets. Most notably, a third of those proteins (24 total) are known to interact with interferon gamma (IFNγ) (enrichment p-value<10^-8). Gene symbols are shown in Figure 1. This finding is consistent with previous observations on the importance of IFNγ in endothelial cell proliferation and expression in KS tumors. We are currently evaluating these results further, including conducting other experiments intended to compare the KS proteomics signatures that distinguish 6 HIV+ and 6 HIV- KS tumors. Preliminary results from these analyses will be presented as well.

Figure 1

Tumor proteins with direct or indirect relationships to IFNγ.

Conclusions

The combined use of genomic and proteomic interrogation of biopsy material from KS tumors has revealed a large set of proteins that are overexpressed in KS compared to normal skin and provides a set of candidate diagnostic proteins for the prevention or early detection of KS. These data are also useful in exploring hypotheses regarding the pathogenesis of KS and relating those mechanisms to their role in endemic and epidemic disease.