Papillary and muscle invasive bladder tumors with distinct genomic stability profiles have different DNA repair fidelity and KU DNA-binding activities.

Low-grade noninvasive papillary bladder tumors are genetically stable whereas muscle invasive bladder tumors display high levels of chromosomal aberrations. As cells deficient for nonhomologous end-joining (NHEJ) pathway components display increased genomic instability, we sought to determine the NHEJ repair characteristics of bladder tumors and correlate this with tumor stage and grade. A panel of 13 human bladder tumors of defined stage and grade were investigated for chromosomal aberrations by comparative genomic hybridization and for NHEJ repair fidelity and function. Repair assays were conducted with extracts made directly from bladder tumor specimens to avoid culture-induced phenotypic alterations and selection bias as only a minority of bladder tumors grow in culture. Four noninvasive bladder tumors (pTaG2), which were genetically stable, repaired a partially incompatible double-strand break (DSB) by NHEJ-dependent annealing of termini and fill-in of overhangs with minimal loss of nucleotides. In contrast, four muscle invasive bladder cancers (pT2-3G3), which displayed gross chromosomal rearrangements, repaired DSBs in an error-prone manner involving extensive resection and microhomology association. Four minimally invasive bladder cancers (pT1G3) had characteristics of both repair types. Error-prone repair in bladder tumors correlated with reduced KU DNA-binding and loss of TP53 function. In conclusion, there were distinct differences in DSB repair between noninvasive papillary tumors and higher stage/grade invasive cancers. End-joining fidelity correlated with stage and was increasingly error-prone as tumors became more invasive and KU binding activity reduced; these changes may underlie the different genomic profiles of these tumors.

Alternative splicing of fibroblast growth factor receptor 3 produces a secreted isoform that inhibits fibroblast growth factor-induced proliferation and is repressed in urothelial carcinoma cell lines.

Fibroblast growth factor receptors (FGFRs) are a family of receptor tyrosine kinases that play key roles in proliferation, differentiation, and tumorigenesis. FGFR3 was identified as the major family member expressed in both normal human urothelium and cultured normal human urothelial (NHU) cells and was expressed as the IIIb isoform. We also identified a splice variant, FGFR3 Delta8-10, lacking exons encoding the COOH-terminal half of immunoglobulin-like domain III and the transmembrane domain. Previous reports have assumed that this is a cancer-specific splice variant. We showed that FGFR3 Delta8-10 is a normal transcript in NHU cells and is translated, N-glycosylated, and secreted. Primary urothelium expressed high levels of FGFR3 transcripts. In culture, levels were reduced in actively proliferating cells but increased at confluence and as cells approached senescence. Cells overexpressing FGFR3 IIIb showed FGF1-induced proliferation, which was inhibited by the addition of FGFR3 Delta8-10. In bladder tumor cell lines derived from aggressive carcinomas, there were significant alterations in the relative expression of isoforms including an overall decrease in the proportion of FGFR3 Delta8-10 and predominant expression of FGFR3 IIIc in some cases. In summary, alternative splicing of FGFR3 IIIb in NHU cells represents a normal mechanism to generate a transcript that regulates proliferation and in bladder cancer, the ratio of FGFR3 isoforms is significantly altered.

Functional analysis of discoidin domain receptor 1: effect of adhesion on DDR1 phosphorylation.

Discoidin domain receptor 1 (DDR1), a receptor tyrosine kinase (RTK), has been shown to be activated mainly by soluble fibrillar collagen. Unusually, the kinetics of phosphorylation of the receptor is slow, with maximal phosphorylation observed after 90 min. To understand the reasons for slow phosphorylation of the receptor, we examined several cell lines under different conditions. We confirm that endogenous DDR1 is phosphorylated slowly by collagen in adherent T47D and HCT116 cells. In detached and resuspended cells, collagen induced rapid phosphorylation of DDR1. This was further confirmed with a semiadherent cell line (COLO201) and one that grows as a suspension (K562), both of which express endogenous DDR1. Replating K562 on fibronectin to mimic adherent conditions altered the kinetics of phosphorylation from rapid to slow, similar to those of adherent cells. The slow kinetics of phosphorylation in the adherent state was probably not due to cell-cell contacts because EDTA had no major effect. However, pervanadate in the absence of collagen was able to induce strong DDR1 phosphorylation, indicating that a phosphatase may inhibit or delay the phosphorylation of DDR1. Further, downstream signals after phosphorylation of DDR1 by collagen were not transmitted through the classical mitogen-activated protein kinase pathway. In addition, a chimeric TrkA-DDR1 receptor failed to become phosphorylated on stimulation with nerve growth factor (NGF), although it dimerized normally. This is the first RTK whose kinetics of phosphorylation is dependent on cellular context. The interaction of the cells with the matrix, rather than cell-cell contact, is probably responsible for the inhibition of phosphorylation.

Cell responses to FGFR3 signalling: growth, differentiation and apoptosis.

FGFR3 is a receptor tyrosine kinase (RTK) of the FGF receptor family, known to have a negative regulatory effect on long bone growth. Fgfr3 knockout mice display longer bones and, accordingly, most germline-activating mutations in man are associated with dwarfism. Somatically, some of the same activating mutations are associated with the human cancers multiple myeloma, cervical carcinoma and carcinoma of the bladder. How signalling through FGFR3 can lead to either chondrocyte apoptosis or cancer cell proliferation is not fully understood. Although FGFR3 can be expressed as two main splice isoforms (IIIb or IIIc), there is no apparent link with specific cell responses, which may rather be associated with the cell type or its differentiation status. Depending on cell type, differential activation of STAT proteins has been observed. STAT1 phosphorylation seems to be involved in inhibition of chondrocyte proliferation while activation of the ERK pathway inhibits chondrocyte differentiation and B-cell proliferation (as in multiple myeloma). The role of FGFR3 in epithelial cancers (bladder and cervix) is not known. Some of the cell specificity may arise via modulation of signalling by crosstalk with other signalling pathways. Recently, inhibition of the ERK pathway in achondroplastic mice has provided hope for an approach to the treatment of dwarfism. Further understanding of the ability of FGFR3 to trigger different responses depending on cell type and cellular context may lead to treatments for both skeletal dysplasias and cancer.