GLASGOW — Certain bacteria within the microbiome are associated with an increased risk of colorectal cancer, potentially by increasing cell proliferation and migration, particularly in malignant tissues, suggests two studies presented at NCRI Cancer Conference 2019.
Kaitlin Wade, PhD, an early career fellow at the University of Bristol, and colleagues analysed data on almost 4000 individuals taking part in microbiome studies and a further 120,000 from a colorectal cancer study.
Using Mendelian randomisation, they preformed a genome-wide association study (GWAS), finding that people with a member of the Bacteroidales group of bacteria have an increased risk of colorectal cancer.
Kaitlin Wade said in a news release that the results mean that, “on average, people with this type of bacteria within their gut may have a slightly higher risk of bowel cancer compared to those who don’t”.
She added: “Our findings support previous studies that have shown that Bacteroidales bacteria are more likely to be present, and in larger quantities, in individuals with bowel cancer compared to those without the disease.”
Prof Ian Tomlinson, is incoming director of the Cancer Research UK Edinburgh Centre, and a member of the NCRI conference scientific committee.
He commented that, “this is one of the first studies” to use Mendelian randomisation “to provide insights into the reasons for the postulated and plausible, but largely unproven, links between the microbiome and bowel cancer”.
While Prof Tomlinson said that it was “too early to ascribe causality to the findings reported…similar larger studies have the potential for greatly improving our understanding of how bowel cancer develops”.
However, other experts questioned the methodology used and said the available information was “insufficient” to assess the findings properly.
In a separate poster, James Robson, a PhD student in the Centre for Research in Biosciences at the University of the West of England, Bristol, and colleagues looked at the impact of several bacterial species linked to colorectal cancer on both benign and malignant cell lines.
They showed that, while the majority enhanced cell proliferation, two species, albeit not ones identified by Wade, significantly increased cell migration in malignant, but not benign, cells, potentially contributing to metastasis.
For the first study, Kaitlin Wade and colleagues relied on Mendelian randomisation, a technique that overcomes the limitations of observational studies by using variations in the human genome as ‘proxies’ for exposure to a variable.
They used data form the Flemish Gut Flora Project, a dataset of 16S ribosomal RNA sequencing and genotyping in 2223 individuals, as well as the German FoCus (n=950) and PopGen (n=717) cohorts.
A GWAS was then conducted to examine the human genetic contribution to gut microbiome variation, from which 1018 independent genetic associations with 114 microbial traits (MT) were identified.
Thirteen SNP-MT associations reached the level of genome-wide significance, and these were used to determine the impact of the microbiome on colorectal cancer in the three original cohorts, as well as in 120,328 individuals from the Genetics and Epidemiology of Colorectal Cancer Consortium.
The results showed that a classified bacterium within the Bacteroidales family, with an allele frequency in the European population of 0.04, was associated with an increased risk of colorectal cancer, at an odds ratio of 1.08 (p=0.02).
Kaitlin Wade said, however, that more work needs to be done to understand the implications on human health.
She said: “We need to classify the exact species or strain of bacteria in the Bacteroidales group, and we need to do more work to understand how and why human genetic variation can alter the gut microbiome.
“Even if these results show that these bacteria may cause bowel cancer, we don’t know whether trying to alter them in an effort to reduce the risk of bowel cancer might have other, unforeseen effects on other aspects of health.”
However, Dr Mangesh Thorat, from the Centre for Cancer Prevention, Wolfson Institute of Preventive Medicine, Queen Mary University of London, took issue with the findings.
“The information available about this study is insufficient to evaluate it properly,” he said in a statement. “However, based on the available information, this appears to be a preliminary type of study, prone to several limitations.”
In particular, he pointed out that the Mendelian randomisation “is dependent on how accurately the instrumental variable represents the exposure” and the results are “less meaningful” if that is not “highly accurate”.
He added: “The lower limit of effect size quoted is also close to zero, indicating possibility of no effect.”
Prof Robin Weiss, emeritus professor of viral oncology, University College London, said: “I think Dr Wade should have identified” the species of bacteria, adding that there “may be issues” with the way the research was presented.
In the second poster, James Robson and colleagues examined the impact of gut bacteria associated with colorectal cancer on cell behaviour by co-culturing benign RG/C2 colorectal adenoma cells and HCT116 carcinoma cell lines with various bacterial species.
He explained that the issue with understanding the relationship between the gut microtome and colorectal cancer is that it’s not clear “where in the process of tumour development and progression” the bacteria exert their effects.
“Someone who’s been diagnosed with bowel cancer has possibly had that tumour for 20, 30, 40 years,” he told Medscape News UK.
“Just because we’re isolating a different bacteria form healthy controls versus cancer patients, or isolating different bacteria at the tumour site in a cancer patient versus perhaps normal tissue further on in the colon, we don’t know whether these bacteria are different because they’re colonising the healthy individual and then driving the tumour or do they just colonise tumours better than the normal gut bacteria?
“So it’s a bit of a chicken and egg situation.”
Their experiments showed that all of the bacterial species they studied attached to the tumour cells and were typically internalised.
However, Bacteroides fragilis and Enterococcus ibercalis attached more readily to RG/C2 than HCT116 cells, while Fucobacterium nucleatum was substantially more likely to be internalised by HCT116 cells.
For both benign and malignant cells, being in contact with the bacteria caused a “massive stimulation of tumour cell growth”, Robson said, although the effect was more pronounced in the malignant cells due to them being faster growing by nature.
Turning to cell migration, the researchers found that co-culture with the bacteria had no effect on benign cells.
However, Escherichia coli Nissle and F. nucleatum both significantly increased cell migration in the malignant cell line.
James Robson said that the results suggest that all of the species they looked at affect progression, at least in terms of increasing cell growth, but with those two species “only increasing migration in the late stage”.
Consequently, the bacteria are “having different effects at different stages but contributing to tumour progression nonetheless”.
The results therefore suggest that, in non-cancerous tissue, the proliferative effect of the microbiome is held in check by the surrounding tissue environment, which does not apply in malignant tissue.
However, he noted that, “obviously, this is 2D cell culture model so it’s a bit removed from in vivo mouse models”.
Poster 2532 was funded by the University of Bristol, Wellcome Trust and Elizabeth Blackwell Institute.
No conflicts of interest declared.
NCRI Cancer Conference 2019: Posters 2532 & 2437. Presented 4 November.