Elucidating the mechanism by which the SHROOM2 locus on Xp22.2 contributes to colorectal cancer aetiology.

Cancer of the large bowel (colorectal cancer - CRC) is common, with >41,000 new cases and 16,000 deaths annually in the UK and an estimated 1 million new cases worldwide annually. Thus, CRC is a major clinical and public health problem. However, available evidence indicates that CRC incidence can be substantially reduced by early detection and prevention using drugs and/or dietary components. Understanding the causes of CRC would enable the rational development of preventative agents, biomarkers of risk and potentially new chemotherapy drugs. Twin studies have shown that there is a heritable contribution to CRC (~ 35%). However, until recently, much of the heritable component remained undiscovered. Over the last 5 years, we have successfully studied many thousands of CRC patients and controls using genomic based (genome-wide association studies - GWAS) approaches. We have discovered 20 new genetic regions that are inherited markers of the risk of developing CRC and have shown that genetic marker data, family history, gender and age can be combined into risk profiling models that can identify high risk subgroups within the general population. Whilst the risk imparted by each of the common variants we have identified is individually modest (~10-20% increase in risk), common variants that impart modest effects on human phenotypes can have major therapeutic implications. One example is that of the cholesterol-lowering agents, statins, which are in widespread use and are attributed with a dramatic reduction in coronary heart disease in Britain. Genetic approaches identified a gene (HMGCR) that influences blood cholesterol levels and it is the gene product of HMGCR (HMGcoA reductase) which is inhibited by statins.
We recently discovered a region on the X chromosome that is associated with CRC susceptibility. In pilot studies, we now show that there is a deletion/insertion in the DNA sequence that appears to explain the effect on CRC risk. The region also has a dramatic effect on expression of a gene called SHROOM2 which effects a variety of cell functions. We studied these changes in human large bowel epithelium and showed a very strong relationship between the genotype and SHROOM2 expression, indeed so strong that the DNA sequence explained most of the variation in SHROOM2 expression in the colon. SHROOM2 is also involved in pigmentation of the human retina. Intriguingly, we have shown an association between retinal pigmentation and CRC. We now link these retinal pigment effects with genotype at the SHROOM2 region. Finally, we show that altering the level of SHROOM2 in CRC cell lines substantially reduces cancer cell growth.
The rationale is compelling for study of SHROOM2 as a gene responsible for susceptibility to a common cancer killer (CRC). The research aims to shed new light on cancer biology and susceptibility. Our pilot studies already suggest SHROOM2 (and related pathway components) is a drug target for developing cancer prevention agents. Hence we plan to:
- Identify the causative DNA sequence variant(s) on chromosome X and define frequency and effect on CRC risk by comparing CRC cases and controls.
- Elucidate the specific molecular mechanism by which SHROOM2 impacts on CRC risk and on gene expression in human large bowel epithelium and the retina.
- Identify genes downstream of SHROOM2 in the pathway that are indirectly affected by variation.
- Define the effect of modifying SHROOM2 expression (knockdown & over-expression) on cancer-relevant cell biological phenotypes in: cancer cell lines, human large bowel epithelium.
- Generate in vivo models to describe the phenotype of Shroom2 deletion and to study the role of Shroom2 in tumour initiation and progression.
- Identify targets for chemoprevention and biomarkers of CRC risk in humans.
Above we outline a coherent body of work that will allow us to understand the role of SHROOM2 in large bowel cancer and to apply this in the future to prevent the disease.

Colorectal cancer (CRC) is a common cancer but incidence could be reduced by early detection or prevention with drugs and/or dietary constituents/additives. Understanding CRC causation would enable development of preventative agents and risk biomarkers. Through GWAS, we discovered a locus on Xp22.2, harbouring SHROOM2, the focus of this proposal. We plan to take this GWAS "hit" from sequence interrogation, through to understanding molecular mechanism and function in relevant tissues, into in vivo human and model organisms. Pilot data provide compelling rationale for study of SHROOM2 in CRC initiation and biology. The association between Xp22.2 and CRC risk is robust (p=7.46x10-10). Although effect size is modest (OR 1.07), we identified an eQTL with striking effects on expression, explaining most of the variance in SHROOM2 expression in human normal large bowel epithelium. By re-sequencing, we found a putative causative variant exhibiting enhanced association with CRC risk and even stronger eQTL effects. We show siRNA knockdown of SHROOM2 inhibits CRC cell growth in vitro. We found Xp22.2 is associated in males with retinal pigment changes, previously linked with CRC risk. These exciting data represent the first discovery of an eQTL at a cancer risk locus with such a dramatic effect on gene expression. We plan a coherent body of work to define the role of SHROOM2 in CRC carcinogenesis and aim to identify drug targets for prevention and new risk variants for use in risk profiling strategies:- resequencing SHROOM2 region; characterising risk and expression correlates of newly identified variants; study molecular mechanisms of SHROOM2 regulation by various techniques - RNAseq, ChIP, methylation; deplete/over-express SHROOM2 in CRC lines and cultured human colonic epithelial organoids to study cancer-relevant biological phenotypes; generate SHROOM2 deleted model organisms to study phenotypic effects, effect in chemical carcinogenesis and other genetic (APCmin) models.

This project seeks to determine how a colorectal cancer risk allele on chromosome X affects gene expression and regulation. The work also seeks to understand, how when dysregulated, the gene product of SHROOM2 appears to act in cancer initiation and progression.

Benefits to Nations' Health and General Public awareness:
Lifetime risk of colorectal cancer (CRC) is 5.2%, resulting in >41,000 new cases annually in the UK and >16,000 dying from the disease. CRC is particularly amenable to clinical intervention guided by risk, because there are differing levels of intervention and surveillance that can be tailored to the degree of risk a person has. Therefore identifying and understanding how risk alleles act, and what genes and pathways they act on, can ultimately aid in the development of personalised medicine. Hence various clinical departments, such as genetics, surgical and oncology departments may benefit from the outcome of the research with respect to improved risk profiling and prediction models (see Academic Beneficiaries for further details on the interest in risk and prediction model development), better understanding of disease initiation and progression providing new and improved therapeutic targets and preventative measures. The investigators are involved in focused meetings and working groups and therefore have continued contact with these clinical departments, to ensure the impact is realised and taken forward for further clinical assessment.
Increasing public awareness to disease risk would be of huge benefit - as these risk alleles are common, a large proportion of the population carry risk alleles and therefore population screening approaches, along with information on prevention incorporating simple life-style changes, could be utilised to reduce the patient numbers presenting with untreatable disease. As described elsewhere, the Investigators are committed to informing the public of their work and how it impacts on public health and have played major roles in providing and creating newsletters, information sheets and clinical guidelines.
In addition to refining risk associated with the SHROOM2 locus, we consider that the major impact of the work will relate to identification of targets for intervention in cancer initiation and progression. The dramatic effects we observed in the eQTL analyses combined with the effects on CRC cell lines on depletion of SHROOM2 suggest this is likely to be a productive effort in the medium to long term.

Economic beneficiaries:
There is the potential that findings from this research may lead to be patentable discoveries such as:
i) risk profiling and prediction models
ii) new CRC risk allele tests
iii) development of biomarkers for cancer initiation or progression
iv) development of targeted therapeutic agents

Benefits to Public Sector:
Development of material which is patentable will ultimately mean benefit for the University of Edinburgh and MRC, both from the patents themselves and also increasing funding to the University. The University has onsite contacts for discussion of such patents and therefore the delay in impact can be minimal. As stated above, the results could lead to improved screening and cancer management, which can lead to reduced NHS costs.

Private and other sector beneficiaries:
Development of therapeutic agents and even risk allele tests will interest various industrial companies and may ultimately lead to interactions between industry and grant giving bodies, to influence and increase similar research at a National level.

Further information on how we envisage these benefits being implemented are described in the Pathways to Impact document.