The perinucleolar compartment is a nuclear substructure that is associated with, but structurally distinct from the nucleolus. Previous studies have shown that the PNC preferentially forms in cancer cells and PNC prevalence (% of non-mitotic and non-apoptotic cells in a population with one or more PNC) is a prognostic in human breast cancer. Since PNC prevalence is a prognostic marker for breast cancer and can be high in tumor cell lines derived from other cancerous tissues, it is hypothesized that PNC prevalence may be developed as a pan-cancer marker of malignancy and that PNC prevalence reduction can be utilized for chemical biology and drug discovery screening strategies to elucidate PNC function and to discover novel anti-cancer compounds.

The first goal of this thesis was to determine the validity of PNC prevalence as a pan-cancer prognostic marker. The results show that the PNC forms preferentially in solid tumor cells independent of origin, and PNC prevalence selectively correlates with the malignancy of solid tumor cells, which supports further development of PNC prevalence as a prognostic marker for other solid tumors. To elucidate the function of the PNC, the second goal of this thesis was to employ a chemical biology approach to complement already used molecular and cellular biology strategies in order to elucidate the cellular pathways involved in the formation and maintenance of the PNC. The chemical biology studies have led to the discovery of two proteins that localize to the PNC and, together with other data from our lab, support a model that the PNC is associated with a DNA locus and may participate in trafficking of newly made pol III RNA. The final goal of this thesis is to determine the feasibility of utilizing PNC prevalence reduction as a drug discovery screening strategy to discover broadly efficacious and cancer cell selective compounds. To address this goal, an unbiased library of 6,400 structurally diverse compounds was screened for PNC prevalence reduction. This hit rate and screening rate from this library determined that this strategy is feasible for drug discovery screening. In addition, PNC prevalence reduction was preliminarily validated as a drug discovery marker since clinical and experimental anti-cancer compounds are much more enriched in PNC reducing compounds than is the 6,400 compound library and also validated based on the fact that novel cancer cell selective compounds were discovered. From the screening of anti-cancer compounds, we identified amonafide, a PNC reducing experimental anti-cancer drug. Amonafide has demonstrated pharmacogenetic variability in previous clinical trials; therefore, derivatives were synthesized to address this variability. Three derivatives of amonafide performed equally well or better in several in vitro assays than the parental compound and are suitable for further development.