Application of PathoChip to urine-derived nucleic acids for broad microbial profiling in men with suspected prostate cancer: setup of a methodological workflow and pilot feasibility study.
BACKGROUND: Urine-based liquid biopsy is an attractive non-invasive source of prostate cancer (PCa) biomarkers, but urinary microbiome studies have mainly relied on 16S rRNA sequencing or shotgun metagenomics. This pilot study optimized and evaluated a practical workflow using PathoChip - a broad-spectrum microarray designed to detect bacterial, viral, fungal, and parasitic signatures - for microbial profiling of urine sediments from men with suspected PCa, an application not previously established. METHODS: First-morning urine was collected without prostatic massage from 35 men scheduled for biopsy; 19 were diagnosed with PCa and 16 were biopsy-negative. Different urine volumes and extraction strategies were evaluated to optimize DNA/RNA recovery. A setup phase compared 25 ng versus 50 ng of urine DNA and RNA input. DNA/RNA isolated from human B cells was used as reference control. An analysis pipeline was developed to detect outlier probes and create a presence/absence matrix. Reproducibility was assessed via library yield, Pearson correlation, blank-control subtraction, outlier probe detection. Prevalence comparisons were performed between clinical groups. RESULTS: An 8 mL starting volume was chosen as consistently available from self-collected urine. Sequential DNA/RNA extraction using the AllPrep DNA/RNA Micro Kit from sediment provided the best balance between nucleic-acid recovery, purity, and clinical compatibility. Reducing the input from 50 ng to 25 ng preserved highly concordant hybridization profiles, with matched samples clustering together with strong correlations. Exploratory analysis revealed PCa- and grade-associated patterns involving Actinomycetaceae, Aerococcaceae, and Streptococcaceae, with Streptococcaceae enriched in PCa of higher grades (ISUP GG ≥ 2). Other signatures, including Mobiluncus, Prevotella, Rhodotorula, Hymenolepis, and JC polyomavirus, were broadly detected but not PCa-discriminating. CONCLUSIONS: PathoChip can be adapted to urine sediments, generating reproducible microbial profiles from limited DNA/RNA input without prostatic massage. This platform provides a quick and accessible approach to broad screening, extending beyond 16S rRNA sequencing by enabling simultaneous multi-kingdom detection. The observed PCa- and grade-associated patterns are hypothesis-generating and require validation in larger independent cohorts.