Frequent overexpression of klotho in fusion-negative phosphaturic mesenchymal tumors with tumorigenic implications

Phosphaturic mesenchymal tumors (PMT) are tumors that cause hypophosphatemia/osteomalacia chiefly by secreting FGF23. We have identified FN1-FGFR1/FGF1 fusion genes in nearly half of PMT, suggesting a central role of FGFR1 pathways in the pathogenesis of PMT. Tumorigenic drivers are unknown for tumors where previous study detected neither fusion, including many in bone, where FISH failed because of tissue decalcification. To identify alternative fusions in PMT without known fusions, as well as to validate the positive FISH results and characterize the fusion junctions, 34 PMT were studied, including 12 with known FN1-FGFR1 fusion by FISH (Group A), 2 with FN1-FGF1 (B), 12 with neither fusion (C), and 8 with previous acid-based decalcification and hence unknown fusion status (D). In total, 23 archival samples were subjected to anchored multiplex PCR-based RNA-sequencing (AMP-seq) with primers targeting FN1, genes encoding the FGF/FGFR families, and KL (alpha-Klotho); five Group C cases were also studied with whole-transcriptomic and exome-captured RNA sequencing, respectively. The AMP-seq results were consistent with previous FISH and/or transcriptomic sequencing data, except in one old Group A sample. One case had a novel FGFR1 exon 9 breakpoint, confirmed by genomic DNA sequencing. One Group D bone tumor was found to harbor FN1-FGF1. All 3 RNA-sequencing platforms failed to identify convincing fusion genes in Group C (N = 10), which instead expressed significantly higher levels of either KL or KLB. This result was further confirmed with KL and KLB RNA CISH semi-quantification (RNAscope). Our results demonstrated the utility of AMP-seq, which was compromised by decalcification and prolonged archiving. Of potential importance, fusion-negative PMT frequently overexpressed alpha-Klotho (or instead beta-Klotho less commonly), whose role as an obligatory co-receptor for FGF23-FGFR1 binding suggests its aberrant expression in osteocytes/osteoblasts might result in an FGF23-FGFR1 autocrine loop that in turn drives the overexpression of FGF23 and tumorigenesis through activated FGFR pathways.