The Department of Biomedicine

CCBIO seminar: Norman J. Maitland

Modelling of cell fate and differentiation using tissue-derived human prostate epithelial cells

Main content

Norman J. Maitland
Cancer Research Unit, Department of Biology, University of York, UK

My laboratory uses primary cells of epithelial and stromal lineages disaggregated from human tissues to reconstruct cell-cell interactions in the human prostate (and prostate cancers). At the core of these studies was a desire to understand the change in the proportion of cell types from an equal mixture of basal and luminal cells in normal prostate acini, to a population of cancer cells composed of >99% luminal cells. Whereas cell mass in normal and benign prostate tissues is driven by replicating basal cells, cancers consist largely of replicating luminal cells. In many tissues, most notably skin and colon, the mouse provides an excellent system for cell fate tracking, as a model for human tissues. Unfortunately, the mouse prostate is quite different from the human equivalent in both structure and function, and there are relatively few existing cell line models available for human prostate.

Using freshly disaggregated biopsy material, we were able to enrich for sub-populations, using multiple cell surface markers and differential cell adhesion properties (Collins et al, Cancer Res 2005, 65:10946). Reconstitution of cell types in 3D could regenerate PSA-expressing polarised acini from basal precursor cells. Unlike most contemporary studies, we produced detailed gene expression profiles using the enriched subpopulations, rather than phenotyping heterogeneous cell mixtures from large tissue biopsies. We also identified, for the first time, highly clonogenic stem-like (SC) populations from both normal and malignant tissues. Cancers were distinguished by a strongly pro-inflammatory phenotype (Birnie et al, Genome Biol 2008, 9:R83).

Given the distinct cell populations in benign and cancer tissues, we have attempted to resolve the control mechanisms, which drive cellular differentiation in prostate epithelium.

We have tracked cell fates using 3 independent techniques. Firstly using lentiviral transduction of the SC with a differentiation-sensitive fluorescent marker, we determined that luminal cells derived from a basal precursor. Secondly, by measuring telomere length and telomerase activity we showed that the simple linear relationship from SC to differentiated cells in normal epithelium was changed in benign prostatic hyperplasias (BPH; Rane et al, Eur Urol 2016, 69:551) and finally we exploited the heterozygous state of the characteristic fusion between the prostate-specific TMPRSS2 gene and the ERG oncogene to identify for the first time allelic preferences in TMPRSS2 expression (Polson et al, Nat Commun 2013, 4:1623). A novel co-expression analysis monitored genes whose expression was linked not only in prostate but other human tissues to identify 4 distinct, non-overlapping gene families (Rane et al, Stem Cell Reports 2014, 2:180).

Our most recent studies have focussed on diffusible factors, e.g. cytokine signalling, where both pro- and anti-inflammatory molecules provided by autocrine and paracrine interactions, were exploited by cancer stem-like cells to establish distal metastases, in the absence of effects on tumour cell growth. However, few previous and contemporary studies take account of differentiation, focusing instead on mutation profiles. A true combination therapy would seek to eliminate not only the replicating luminal cells, but also the underlying, largely quiescent cancer stem-like cells.

Chairperson: Karl-Henning Kalland <kalland@uib.no>, CCBIO