Nick Wright : Histopathology

 We have continued to study the genetic evolution of pre-neoplastic and neoplastic lesions, mainly in the gastrointestinal tract. In the human stomach, we have previously shown that metaplastic glands from patients undergoing resection for gastric carcinoma can be entirely deficient in cytochrome c oxidase (CCO) and are thus clonal populations. We have now shown that mutated glands are found in patches that had developed through gland fission. Metaplastic and dysplastic glands can be genetically related, demonstrating the clonal origin of dysplasia from metaplasia. Entire dysplastic fields show a founder mutation from which multiple, distinct sub-clones develop. Therefore we show a distinct clonal evolution from metaplasia to dysplasia in the human stomach. By field cancerization, a single clone can expand to form an entire dysplastic lesion. Over time this field appears to become genetically diverse, suggesting that gastric cancer may arise from a sub-clone of the founder mutation.

In patients who develop Crohn’s disease the development of cancer is a complication. Using a technique known as clonal ordering, which utilizes the relative spatial position of mutations within a lesion to infer the order in which the mutations were acquired, we have shown that Crohn’s carcinogenesis is usually initiated by mutations in p53, and followed by subsequent alterations to p16 and chromosome arm 18q. Furthermore, we have demonstrated widespread field cancerization in these patients whereby the normal gut lining is replaced by a histologically equivalent but mutated tumorigenic clone. The clinical implications of this result, both for patient screening and treatment and surgery regimes are very significant, and as a result we are now undertaking a larger study to confirm the generality of our findings.

We have also been investigating the role of genetic diversity in the development of Barrett’s oesophagus-associated adenocarcinoma. Interestingly, the genetic variation present within endoscopic mucosal resections tends to be low, whereas across a large Barrett’s segment many genetically distinct clones can be observed. Genetic diversity within Barrett’s segments predisposes to cancer development, and our new data have prompted questions of how these genetic distinct and spatially isolated clones can promote tumour growth. We are also detailing the epigenetic mechanisms involved in Barrett’s oesophagus and Barrett’s adenocarcinoma.

By analyzing methylation patterns of normal human colon crypts, known to be related by their sharing clonal point mutations in mitochondrial DNA, we have found that the methylation patterns of two clonally related normal human colon crypts were as dissimilar to one another as two unrelated crypts. This indicates that in normal tissue the rate of niche succession is slow, taking a number of years, and that crypt fission is infrequent, occurring approximately once every 25 years. We have also demonstrated that using only methylation patterns to infer the clonality of tissues over long time periods, as a number of authors recently have, is flawed. Using these techniques we have gained insights into stem cell dynamics and crypt fission rates in dysplastic human colon crypts. Our work demonstrates that the rates of stem cell niche succession, monoclonal conversion and crypt fission are massively increased in dysplastic crypts, suggesting that these up-regulated processes enable rapid fixation and spread of successive oncogenic mutations within adenoma during tumorigenesis.

We have also investigated the clonal architecture of human normal urothelium and demonstrated the presence of CCO-deficient patches of very variable size. Clonal ordering of microdissected bladder carcinomas showed a ‘passenger’ role for FGFR3 mutations, but additional evaluation of potential ‘driver mutations’ (RB, H-RAS, PTEN) still has to be considered. Using the same methods involving mtDNA mutations, we are investigating the histogenesis of pancreatic islets and hepatic regenerative nodules in circumstances of chronic organ injury that often precedes neoplastic development. We continue to question the ability of gastrointestinal cancer cell lines to express so-called stem cell markers and the cell biological effects of factors such as oxygen tension, substratum and co-culture with sub- epithelial myofibroblasts in the presence of Wnt proteins. We are examining normal, premalignant (DCIS) and malignant breast tissue in a collaborative study with the Division of Cancer Studies at King’s College (Prof. Gabrielle Dontu and Prof. Lars Holnberg) to investigate if premalignant tissue also harbours clonal populations suggestive of so-called cancer stem cells. We have yet to be able to show a credible contribution of bone marrow stem cells to the epithelial component of carcinomas.

Our work to express GKN2 in mammalian cells continues. The three TFF proteins are also being expressed with a view to assessing their separate and simultaneous effects on gastric and intestinal cell biology.

We have collaborated with key groups in the Netherlands, Spain, Australia, and more locally in Oxford and CR-UK funded groups in Glasgow and Cambridge in studies seeking to understand the relevance of certain newly acclaimed mouse stem cell markers to normal biology and premalignant changes in the aerodigestive tract. As no useful antisera exist capable of detecting LGR5, OLMF4, ASCL2 or many other targets, we used in situ hybridization (ISH) to localize expression of these genes in routine formalin-fixed paraffin embedded tissues and were able to provide the first demonstration of how LGR5/Lgr5 expression varies during development, premalignant change and tumour formation. In other studies, we assessed the importance of CUX1 expression in pancreatic cancer and several effectors of phenotype in Fh knockout renal tumours. We also observed that tyramide-enhanced fluorescent ISH could generate significant probe-specific artefacts that renewed our concerns about the reliability of many NISH results. We have made progress towards understanding whether adult human kidney epithelium has multipotent progenitors by seeking evidence for the clonal origin of patches of cells spanning nephron segment junctions.