The magic of microbiology

Self-reported survey data is very useful, but it's even better when backed up with biology. Nom Nom links survey data with microbiome sequencing data, providing biological associations and potential mechanisms of action for a wide variety of health issues. In turn, the survey data collected can help explain and understand variation in the microbiome that can lead to big discoveries.

Higher microbiome diversity has been correlated with a variety of positive health outcomes, thus understanding which factors are associated with diversity can help us determine possible risk factors for low diversity as well as identify actionable changes to help increase it. Alpha diversity is commonly measured by calculating the Shannon Diversity Index (SDI), which takes into account two factors: species richness (how many organisms there are) and species evenness (how they are distributed). In ongoing analyses of alpha diversity we are using machine learning approaches to help identify what factors are best at differentiating dogs with a high SDI from those with a low SDI.
We observed a wide variety of interesting correlations between various behaviors and lifestyles and microbial diversity. Interestingly some of these have been observed in human and animal studies, and many of them make biological sense. For example, we see associations and changes by type of diet, consumption of probiotics, and sharing food and water with people and other dogs. Each of these makes intuitive sense, since each has a conceivable impact on the microbiome. Diet is a known major driver of compositional shift, as are probiotics and fermented foods (we’ve actually published some studies on these.) And direct interaction with the microbiome of other people and pets can cause sharing of microbes, which could be reflected in their diversity. Other correlations can lead to constructing risk profiles for decreased diversity, as well as inform lifestyle decisions to rectify it. Decreased diversity correlates with obesity, and conversely, increased diversity associates with increased activity. So interestingly, the classic advice to maintain an optimal weight and get regular exercise even holds true at the level of the microbiome. Decreased diversity with age has been observed in various species, and it’s something we’re very interested in mitigating, with hopefully concomitant restoration of any lost health benefits.
Microbiome beta diversity refers to how individuals can be compared and clustered together by their overall compositional similarity. This is commonly accomplished by using multi-dimensional approaches such as principal coordinate analysis (PCoA) or Uniform Manifold Approximation and Projection (UMAP). Understanding how samples cluster together and what the factors are that drive clustering is useful in various ways. Similar to alpha diversity, it can prophylactically enable the identification of risk factors for unwanted health outcomes, identify pathological populations, and track their resolution with treatment and recovery. Ideally it can also function as a tool for personalized interventions based on the preexisting status of the microbiome, an area we and others are actively pursuing. Here is a subset of our data graphed using UMAP, with colors indicating significant clusters. We are working hard to better understand the drivers of this clustering. Some drivers that we have found to be associated with these cluster assignments are age, body condition and weight.
In addition to diversity, compositional differences between groups of dogs or cats in specific bacterial groups can be useful and interesting for a variety of reasons. For example, are there specific bacteria that are under- or over-represented in dogs with GI issues, or in senior versus adult cats. And we utilize cutting-edge shotgun sequencing approaches, enabling superior taxonomic resolution compared to other sequencing approaches such as 16S. It has been well-established that dietary change leads to changes in the gut microbiome. However, linking these differences to positive or negative health outcomes can be difficult, especially considering that different pets might have different responses depending on a number of other demographic and lifestyle factors, including their microbiome.
In this study, we sampled dogs before and after switching from kibble to a fresh diet to better understand what changes were occurring. We found that individuals who experienced the greatest changes in overall health after the diet change had a higher abundance of certain species and lower abundance of others before, and that microbiome differences observed at the end of the study depended partially on microbiome differences between individuals at the beginning of the study. Findings such as this will help drive future advances in personalized nutrition.
In this analysis, we compared the gut bacteria of dogs that demonstrated a large shift in bacterial composition after switching from kibble to a fresh diet (high responders), to dogs that did not demonstrate such a shift (low responders). We found that the samples from high responders were dominated by Bacteroidales and Clostridiales while the samples from low responders were dominated by Enterobacterales and Lactobacillales. Studies like this can help us understand which pets may benefit the most from fresh food diets, and why.
Shotgun sequencing has additional benefits aside from increased taxonomic resolution. Importantly, this technique also allows us to derive functional interpretations of the microbiome utilizing the Kyoto Encyclopedia of Genes and Genomes (KEGG). A significant amount of compositional variation exists between individuals. However there is much less functional variation that occurs. This is due to functional redundancy that different microbial populations may possess. Looking at functional differences between cohorts can be very illuminating and aid in understanding possible causes of pathology driven by loss or gain of specific metabolic capabilities.
Using machine learning approaches we identified a significant association between microbiome function and pet age. The microbiome is known to be dynamic throughout the aging process in people and pets. However, understanding the functional changes that are occurring provides intriguing targets for study and product development. This builds on the concept that keeping normal microbiome functions intact throughout aging will aid in maintaining health and decreasing age-related pathologies.