This study reports the HMGB1 interactomes in prostate and ovary cancer cells lines. Affinity purification coupled to mass spectrometry confirmed that the HMGB1 nuclear interactome is involved in HMGB1 known functions such as maintenance of chromatin stability and regulation of transcription, and also in not as yet reported processes such as mRNA and rRNA processing. We have identified an interaction between HMGB1 and the NuRD complex and validated this by yeast-two-hybrid, confirming that the RBBP7 subunit directly interacts with HMGB1. In addition, we describe for the first time an interaction between two HMGB1 interacting complexes, the septin and THOC complexes, as well as an interaction of these two complexes with Rab11. Analysis of Pan-Cancer Atlas public data indicated that several genes encoding HMGB1-interacting proteins identified in this study are dysregulated in tumours from patients diagnosed with ovary and prostate carcinomas. In PC-3 cells, silencing of HMGB1 leads to downregulation of the expression of key regulators of ribosome biogenesis and RNA processing, namely BOP1, RSS1, UBF1, KRR1 and LYAR. Upregulation of these genes in prostate adenocarcinomas is correlated with worse prognosis, reinforcing their functional significance in cancer progression.

Epithelial ovarian cancer (EOC) is one of the deadliest gynecological cancers worldwide, mainly because of its initially asymptomatic nature and consequently late diagnosis. Long non-coding RNAs (lncRNA) are non-coding transcripts of more than 200 nucleotides, whose deregulation is involved in pathologies such as EOC, and are therefore envisaged as future biomarkers. We present a meta-analysis of available gene expression profiling (microarray and RNA sequencing) studies from EOC patients to identify lncRNA genes with diagnostic and prognostic value. In this meta-analysis, we include 46 independent cohorts, along with available expression profiling data from EOC cell lines. Differential expression analyses were conducted to identify those lncRNAs that are deregulated in (i) EOC versus healthy ovary tissue, (ii) unfavorable versus more favorable prognosis, (iii) metastatic versus primary tumors, (iv) chemoresistant versus chemosensitive EOC, and (v) correlation to specific histological subtypes of EOC. From the results of this meta-analysis, we established a panel of lncRNAs that are highly correlated with EOC. The panel includes several lncRNAs that are already known and even functionally characterized in EOC, but also lncRNAs that have not been previously correlated with this cancer, and which are discussed in relation to their putative role in EOC and their potential use as clinically relevant tools.