Metagenomic Assay for Identification of Microbial Pathogens in Tumor Tissues

Baldwin, Don A.; Feldman, Michael; Alwine, James C.; Robertson, Erle S.

Aliquots of the TransPlex products used for PathoChip screening were combined into six capture pools as indicated in Fig. 6A . The pools were mixed with a panel of biotinylated DNA probes that included HPV16 probes. Pooled DNA that hybridized to the probes was captured by magnetic streptavidin beads. Deep sequencing of libraries derived from the captured material produced a number of reads that map with high homology to the HPV16 genome (Fig. 6A) , and these reads were enriched for the regions targeted by the capture probes but also included distal sequences. Mapping individual reads from each library showed that very few HPV16 templates were present in the sample pool containing only tumors that were HPV16 negative by PathoChip or p16 assays (pool 1, Fig. 6A and C) . More HPV16 templates, up to 73 in pool 3, were observed in the libraries from HPV16positive tumors. Interestingly, there were regions of high reads which showed high intensities in the majority of the tumors analyzed. The E4 region which was prevalent in the majority of the tumors (Fig. 6A to C) may provide a window into the transcription profiles of HPV16 in these particular type of OSCCs. The oncogenes E6 and E7 also showed a high number of signals in the tumors, but not as dramatic and surprising as the E4 open reading frame (ORF) region. However, this may be an interesting discovery which suggests a higher number of transcripts for the E4 ORFs in these tumors than was previously thought. Notably, the E1 region also showed greater signals in the tiled probes across the HPV16 genome. Predominantly, E4 seems to be the most prominent signal for the OSCC tumors and suggests a greater involvement of E4 in maintenance of the tumor by HPV than previously indicated.