Deciphering individual bacterial strains in the gut is important to understanding the microbiome

Phylogenetic trees go beyond just genus and species, but also to strains.

When scientists determine the bacterial population of one’s microbiome using genetic sequencing they are forced to make determinations of the populations’ phylogeny.  Some scientists will determine the abundance of each genus, like all the bacteria that belong to the Clostridium genus, while others will narrow their scope to specific species, like all the bacteria that belong to the Clostridium difficile versus Clostridium scindens species.  These distinctions can be immensely important.  For example, C difficile can cause colitis, but C. scindens can prevent colitis, so measuring the amount of Clostridium in the gut does not paint as clear a picture as measuring the amount of C. difficile (potentially harmful) and C. scindens (potentially helpful).  There is, however, a deeper level of differentiation within species: the strain (e.g. C. difficile A90 vs. C. difficile AA1).  These strains are very similar genetically, especially in the genomic regions most important to determining phylogeny, there are however, potentially important genetic differences. 

Scientists have long realized and understood that by characterizing the species population of the microbiome they were neglecting possible important strain-specific effects.  There have even been specific examples within the microbiome that differences in strains are important, like strains within the Staphylococcus aureus that differ in their antibiotic resistance (e.g. MRSA).  Just how important these strain-specific differences are is unknown, but there is mounting evidence they need to be taken into account.  Last week in the journal Cell, researchers from the University of Washington published results that showed strain-specific differences can be vast and immense, and that this is even more so true in the microbiome where genetic mutations and genetic transfers happen at a high rate.

The researchers used metagenomics data from patients with IBD from a previously published data set.  They took this data, compared it with previously published species’ genomes, and did a lot of fancy bioinformatics to measure strain-specific genetic differences within species.  I don’t mean to neglect the bioinformatics aspect of the paper, which is critically important their results, but the details of their ‘pipeline’ are beyond the scope of this short blog.  In any event, they learned that there were many examples of different strains that coded for as much as 20% more copies of specific genes.  As it turns out, these differences were prevalent in genes that coded for important functions, like transport, signaling, biosynthesis, motility, secretion, and virulence.  These are important processes in the gut environment, and each may have important impacts on the host.

Before this paper, the level of genomic resolution and bioinformatics needed to make strain specific determinations was difficult, and beyond the expertise of most labs.  Now that these researchers have published their methods, this type of strain analysis can be incorporated into many more experiments.  It does have its drawbacks, as full genomes are needed for each species that is analyzed, and inserted/deleted genes are not analyzed, but overall it is a very important paper.  As we move forward in microbiome research, this type of analysis that incorporates specific strains will become critical in associating diseases and phenotypes with the microbiome.

Gut bacteria protect against malaria transmission

Malaria is a deadly disease transmitted through mosquitoes and most widespread in tropical and subtropical regions around the world, especially in Africa. According to the Center for Disease Control and Prevention (CDC), 627,000 people died in 2012 and there were a total of 207 million cases worldwide. Through studying the microbiome, scientists last week published a major discovery in Cell that may lead to better vaccinations for malaria that could help prevent the disease from being transmitted.

Scientists in Portugal, collaborating with colleagues in the United States, Australia, and Mali, found that the parasite the causes malaria, Plasmodium, expresses the same sugar molecule that is seen in a type of Escherichia coli (E. coli).  This sugar molecule from the E. coli called alpha-gal (a-gal) results in the body’s immune system producing antibodies against this molecule and therefore also protecting against the malaria parasite. It is known that adults who are exposed to malaria are at lower risk of contracting the disease than children under the age of 5 and the researchers hypothesized that this was due to the children lacking this specific E. coli in their body and therefore unable to fight back against Plasmodium exposure.   

The scientists studied the gut bacteria of a group of individuals in Mali who had very high rates of malaria transmission. They found that those who had higher levels of anti-a-gal antibodies had lower risk of transmitting malaria and those with low levels of these antibodies had greater risk of transmitting the disease.  This showed that children are at greater risk for the disease because they do not produce enough anti-a-gal antibodies to prevent the parasite from infecting the body.

The scientists also found that the transmission of the parasite is blocked almost immediately following its introduction into the body through the skin. The antibodies against a-gal attach to the Plasmodium as soon as it is exposed to the body, and a part of the immune system called the component cascade is activated, killing the parasite before it can leave the skin and reach the blood stream.   

They found that by vaccinating mice against a kind of a-gal, the mice produced enough anti-a-gal antibodies that were highly efficient in protecting the mice from malaria transmission.  The scientists believe that it may now be possible to translate this work to humans and develop vaccines that would increase anti-a-gal antibodies and prevent malaria transmission. If successful, vaccinations could be given to children who are at high risk for the disease and could prevent hundreds of thousands of deaths every year.  These findings also illustrate the protective aspects of the microbiome in regulating immunity, and the potential treasure-trove of molecules produced by the microbiome that could be used in therapeutics.

IBD, Crohn's disease, and the microbiome

There have been several large scale studies that show a definite association between the microbiome and inflammatory bowel diseases like Crohn’s and colitis.  While it is becoming clear that the microbiome plays a pivotal role in these diseases, the exact mechanisms of pathogenesis are not known.  One of the reasons it has been so hard to uncover this link is because of a lack of robust studies, that usually contain small sample sizes.  In response, a team of leading scientists in the field, including multiple scientists from our own scientific advisory board, assembled a brand new cohort, the largest of its kind, to study the disease.  The cohort consisted of 447 children with newly diagnosed, still untreated, Crohn’s disease and 221 healthy children.   They then combined this data with two other cohorts that included adult patients to bring the total number of samples to 1,742.  The first results of this study were published in Cell Host and Microbe, and shed many insights into IBD and its relationship to the microbiome.

In the study the researchers sampled and sequenced the microbiome of the gut mucosa and stool.  From this data, they identified specific bacteria that had higher abundances in diseased patients, like Enterobacteriaceae and Veillonellaceae, and others that had lower than normal abundances, such as Clostridiales and Bacteroidales.  According to the samples, these relative abundances were more pronounced in the mucosal samples, rather than the stool samples, meaning that the mucosa may play a more important role in Crohn’s pathogenesis and diagnosis.  Moreover, children under age 10 did not have large populations of the ‘bad’ bacteria, which were negatively correlated with age.

Another important finding from the study is that antibiotic treatment of the Chrohn’s disease further exacerbated the microbial imbalances (dysbiosis), and caused the bad bacteria to proliferate and the good bacteria to die off.  The researchers are quick to dismiss antibiotics as the cause of the disease, rather hypothesizing that the antibiotics allow for opportunistic bacteria to grow which may cause dysbiosis.  They also encourage further research on the subject and question whether antibiotics should be prescribed to IBD patients.

The researchers used this knowledge of the dysbiosis to create a potential diagnostic for IBD.  They were able to calculate a microbiome diversity index by sampling the mucous for the existence of certain bacteria and for overall microbiome diversity.  The index shows remarkable ability to not only accurately predict the existence of IBD, but also the severity of it.  Also, the researchers showed their microbiome dysbiosis index could be combined with clinical data to better predict the future outcome of the disease.

A final conclusion of the paper is that the gut mucous is a more accurate signal for IBD.  They conclude that stool samples are composed of higher levels of aerobic, oxygen using, bacteria than those in gut, and that stool is not representative of the overall microbiome.

This study is the first to rigorously tackle IBD, specifically, Crohn's disease, and the microbiome.  We now know what bacteria are most highly associated (both positively and negatively) with IBDs, and how this this information can be incorporated into early diagnostic screens.

Human genetics, the microbiome, and obesity

Editors note: Happy birthday to our intern, Becky Siegert.  She has been helping to bring you the blogs for the past 3 weeks, and we at the AMI deeply appreciate her help with them.  Enjoy the day, Becky!

An article was published last week in Cell out of Ruth Ley's lab at Cornell University that discovered a connection between host genetics and the microbiome.  In doing so, the researchers discovered which aspects of the microbiome were heritable, and how some heritable bacteria in the microbiome are related to obesity.

The researchers obtained stool samples from almost a thousand humans that were twins.  They discovered that twins had more similar microbiomes than non related individuals, and that identical twins had even more similar microbiomes than fraternal twins.  Then, by meticulously analyzing their data sets along with other published twins data they discovered which specific types of bacteria were the most heritable.  They discovered that the bacteria from the family Christensenellaceae were the most heritable.  Interestingly, this family seems to exert great influence over the existence or non existence of other important bacteria in the gut, and it often occurs with methane producing archaea.  Additionally, this family of bacteria appeared to be associated with a low body mass index (BMI), or leanness.  To test the effect of this family of bacteria on obesity the researchers performed a series of microbiome transplants from humans to germ free mice.  In one test, an obese microbiome that was amended with a Christensenellaceae bacteria prior to transplantation resulted in weight loss for the mouse.  In other tests the amount of weight gain in mice mirrored the amount of Christensenellaceae.

This study is important in many ways.  First, it unambiguously connects host genetics to the microbiome.  Second, it connects a specific family of bacteria, Christensenellaceae, to BMI.  This family is heritable and seems to play a large role in shaping the rest of the microbiome.  Altogether this paper adds a new dimension to the microbiome and nutrition.  Many studies associate certain genes with obesity, but perhaps the microbiome is actually responsible.

Current challenges in microbiome research

Scanning electron micrograph of E. Coli bacterium.

Curtis Huttenhower, Rob Knight, Owen White, Jacques Ravel (all members of our SAB), and others recently published a commentary titled "Advancing the Microbiome Research Community" in Cell.  The paper was a 'State of the Microbiome' address and outlined the current challenges and the future outlooks of the microbiome field.

Among the many challenges they outlined include how to design causality studies, and how to attribute causality of phenotypes to specific bacteria, proteins, or metabolites.  The paper also discussed the bioinformatics bottleneck that is occurring in the field, as the necessary expertise required to properly analyze sequencing data is severely lacking.  There also lacks a proper centralized online repository for all sequencing data that is generated.  This sequencing data needs to include metadata that would describe how the data was generated.  Furthermore, the authors describe the need for guidance on how to properly define and comprehensively describe phenotypes, such as inflammatory bowel disease (IBD), in microbiome studies.  Without this standardization, proper comparison between studies can not take place.  Finally, institutional review boards (IRBs) need guidance on how to properly regulate human microbiome studies in order for scientists to properly prepare and carry out their projects.

While the above areas are only a few of the challenges facing the field, each of them reinforces the mission and programs of the AMI.  The MBQC is standardizing techniques and creating rigorously tested analytical protocols by investigating the sources of variation in microbiome testing.  In addition, the AMI is focusing on future data repository efforts.  These are just a few of the areas we hope to serve the field in the coming years.  We believe a nonprofit, specifically the AMI, can help fill these voids and address many of the issues outlined in this paper.