Bacteria from the intestinal tract can be used to predict the occurrence of colorectal cancer

The microbiome, a collection of microorganisms in the gut, can be used to predict the occurrence of colon cancer – the second most common type of tumor in women and the third most common in men. A study by an international team, with Brazilian participation, identified patterns in the intestinal microbioma – independent of the food culture of the populations studied – and discovered a link between changes in this pattern and the occurrence of colorectal cancer. The finding opens the way for the development of non-invasive tests that are able to predict the occurrence of the disease.

---

The study, published this Monday (01/04) in the journal Naturopathy, is the first author to be Andrew Maltez Thomas, PhD in Bioinformatics from the University of São Paulo (USP), and was supported by FAPESP through an Overseas Research Fellowship, held at the University of Trento, Italy.

The researchers combined metagenomics, bioinformatics and machine learning (using artificial intelligence) to correlate the occurrence of colorectal cancer with data from 969 people in Germany, France, Italy, China, Japan, Canada and the United States. This is one of the largest and most varied studies on the subject.

The results not only identify sets of microorganisms associated with colorectal cancer in all populations studied, but also indicate signatures in the microbial metabolism (standard of metabolites produced by the microorganisms) that also have the ability to act of the disease.

The analysis yielded two other important findings. One refers to the higher prevalence of bacteria that are often found in the mouth and in the airways in the intestines of patients with colorectal cancer. The other finding indicates the link between disease and the presence of a microbial enzyme that degrades choline, a nutrient that is part of the vitamin B complex.

In the study, patients with colorectal cancer had a higher presence of Fusobacterium nucleatum bacteria compared to healthy individuals. This bacterium is usually located in the & # 39; s mouth region. Until that time, the acidic environment of the stomach was expected to be fatal to such microorganisms.

"There is a wider range of oral species that end up in the gut in colorectal cancer patients. Perhaps this migration causes an inflammation in the gut from the tumor. However, we do not yet know the real reason that they are being transported to the gut "We only know that there is a connection between its presence in the gut and colon cancer. It's something that still needs to be better understood," said Thomas.

The detection of a greater abundance of the microbial enzyme choline-trimethyl lyase (cutC) gene in the fecal samples from cancer patients shows a possible link between the microbiota and the results of previous studies on the relationship of disease with a high-fat diet.

"This enzyme breaks down choline, a metabolite that is present in diets with high concentrations of red meat and other fatty foods. After breaking down the choline, it releases the acetaldehyde, a known carcinogen," Thomas said.

In the study, the researchers analyzed data on the composition and abundance of all bacteria found in 969 faecal samples. For a simpler method of analysis that can be used extensively in clinics and hospitals, researchers have been able to select the most important bacteria in the analysis.

"With 16 species we achieved results that are comparable to the analyzes made with all species. So it is a step towards a simple diagnostic tool, without having to follow the entire microbiota and without having the necessary precision," said Thomas.

Cause or effect

The relationship between intestinal microbiota and public health is a research area that has grown, especially in the last 10 years. However, the new study follows an innovative concept of the use of bacteria as markers for disease development.

"The most common is to look for markers that are directly related to tumor cells. In our work, the concept is a different one. The analysis is made of changes in a relatively small set of bacteria within a spectrum of hundreds of bacteria in the intestines and that can indicate a disease, "said Emmanuel Dias-Neto of the International Research Center (Cipe) at ACCamargo Cancer Center, also author of the article.

With the analysis of DNA sequences obtained from the microbiota, it is possible to find out which bacteria are present in the microbiota of each sample, as well as to determine the amount of each of the bacteria and variants of the genome of these microorganisms identify which may be related to different outcomes – For example, the occurrence or increased risk of colorectal cancer.

However, it should be noted that the study has not revealed that an altered microbiota causes colorectal cancer.

"An association has been found, which does not necessarily imply a causal relationship. The question remains: are these particular bacteria that cause cancer or is it the cancer that creates a different environment in the dermetal that makes certain bacteria preferable to others? We have still not this answer, which will be fundamental, so the results of this article can help in the future to develop therapies for the treatment of colorectal cancer, "said João Carlos Setubal, coordinator of the Post-Graduation Interunit Program in Bioinformatics with USP and another author of the article. Setubal and Dias-Neto have supervised Thomas & # 39; s doctorate.

Computational analysis

According to the researchers, this is perhaps the largest analysis of colorectal cancer with data from fecal samples and with such diverse populations. The team of researchers analyzed data from five public studies with data from two other studies conducted by researchers from the University of Trento.

With the information obtained in the seven studies, it was possible to identify enzymes, bacteria and how the microbiota is able to predict the presence of colorectal cancer. Data from two other studies plus 200 samples were used to validate the findings.

"DNA sequencing of the samples – the analysis of which requires a distinction between DNA from the microbiota and human DNA – was a way to identify and quantify the types of microorganisms and their genes in those samples. We have the DNA from the fecal samples were extracted and sequenced, then we analyzed the data using computational methods and were able to identify and quantify which species were present and what the abundance of genes was, "said Thomas.

Because they are data from different studies, the researchers used advanced statistical methods to analyze them together.

"We used statistical analyzes that are used to perform meta-analyzes, and machine learning techniques were also used to understand how predictive the results are," Thomas said.

In addition to validating the data, the Nicola Segata group, from the University of Trento and leader of the project, was reinforced by another study conducted at the European Molecular Biology Laboratory (EMBL), based in Germany. The EMBL scientists also studied the relationship between microbioma and cancer and had published an article in the same issue of Naturopathy.

"While preparing the articles, we exchanged data and information with the other group, in a partnership that proved very important to reinforce our findings. Although we use machine learning techniques and different statistical methods, we come to the same result : that the intestinal microbiome can predict the presence of colorectal cancer in different populations and studies, "Thomas said.

The article Metagenomic analysis of colorectal cancer data sets identifies cross-cohort microbial diagnostic signatures and a relationship with choline degradation, by Andrew Maltez Thomas, Paolo Manghi, Francesco Asnicar, Edoardo Pasolli, Federico Armanini, Moreno Zolfo, Francesco Beghini, Serena Manara, Nicolai Karcher, Chiara Pozzi, Sara Gandini, Davide Serrano, Sonia Tarallo, Antonio Francavilla, Gaetano Gallo, Mario Trompetto , Giulio Ferrero, Sayaka Mizutani, Hirotsugu Shiroma, Satoshi Shiba, Tatsuhiro Shibata, Shinichi Yachada, Takuji Yamada, Jakob Wirbel, Petra Schrotz-King, Cornelia M. Ulrich, Hermann Brenner, Manimozhiyan Arumugam, Pear Bork, Georg Zeller, Francesca Cordero, Emmanuel Dias-Neto, João Carlos Setubal, Adrian Tett, Barbara Pardini, Maria Rescigno, Levi Waldron, Alessio Naccarati and Nicola Segata, can be read at: www.nature.com/articles/s41591-019-0405-7.

FAPESP Agency

class = "eliadunit">