The COVID-19 pandemic is spreading to every corner of the world. But not everyone is falling sick at the same rate. A new study published on the preprint server medRxiv in April 2020 suggests that the composition of the gut microbiome could partially explain the difference in susceptibility. This adds a new dimension to what is currently known about the disease.
Is COVID-19 linked to the gut?
Clinicians have observed that more than 60% of patients with COVID-19 have diarrhea, nausea, and vomiting, and these symptoms predict a worse outcome overall.
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19 disease, enters the human host cell by binding to the angiotensin-converting enzyme (ACE) 2, which acts as the viral receptor. This molecule is found at higher concentrations in the ileum and colon and regulates intestinal inflammation. ACE2 directly affects the gut microbiome, and indirectly the cardiopulmonary risk.
How was the study done?
Older and sicker individuals are more likely to fall sick when exposed to this virus. The current study investigates the potential link between the gut microbiome and the clinical course and features of COVID-19.
The researchers selected a set of proteins that could act as biomarkers to forecast the progression to severe disease. However, they also examined whether these proteins could help understand what makes a person more or less susceptible to the disease, and what role the gut microbiota plays in regulating the levels of these biomarkers in healthy people.
Finding the PRS
Blood proteomics data from 31 patients with COVID-19 was used, along with multi-omics data from 2,400 uninfected individuals. The former dataset was used to create a risk score to predict whether a case of COVID-19 would progress to a severe or critical level. This is called the blood proteomic risk score (PRS).
The blood PRS was then related to inflammatory biomarkers to see whether the PRS was capable of predicting disease susceptibility in healthy individuals. They used both proteomic and blood biomarkers for inflammation, from 990 people, for this investigation.
The next step was to identify the specific features of the gut microbiota, which would predict the significant proteomic biomarkers for COVID-19, using machine learning. The metabolic profile of the feces was also analyzed to uncover other mechanisms that might be key to the linkage of the gut microbiome to disease vulnerability.
The last step was to assess how 40 different host factors and environmental factors shaped these gut microbiome factors.
PRS predicts severity of COVID-19
In earlier work on COVID-19 patients, the researchers identified 22 serum proteomic biomarkers that helped predict progression to severe disease. This set of biomarkers was pruned to 20, leaving out two which were not available for the healthy patients.
This set was used to set up a blood PRS for the 31 patients in the current study. Eighteen of the cases were not severe, while 13 were severe.
As the PRS increased by 10%, the risk of severe disease went up by 57%. The researchers interpreted this as being evidence that the risk score could predict the progression of COVID-19.
PRS linked to inflammation in the elderly
In an extension of their work, they created the PRS using the same 20 proteins but based on a database of healthy participants, including both proteomics and inflammatory markers. The blood proteomic data came from the serum samples taken at baseline. The inflammatory markers included IL-1β, IL-6, TNF-α, and hsCRP.
They found that the PRS was positively correlated with hsCRP and TNF-α among this set. In a subgroup analysis stratified by age, they found a significant link between a higher PRS and higher blood levels of all inflammatory markers in older but not in younger subgroups (above and below 58 years, respectively).
The question the researchers face is whether the changes in these proteins underlie the immune activation seen in these patients, or are the results of it. Whatever the answer, the study shows a clear link between immune imbalance and a higher PRS, especially in older adults, supporting its role as a biomarker.
PRS linked to gut microbiome and inflammatory changes
The next step was performed on a cohort of approximately 300 participants. The researchers measured the correlation between the microbial profile of the gut microbiota and the blood proteomics. They performed both a cross-sectional or snapshot study of the participants in one sample (n=132), while another prospective study was carried out in 169 participants. Here, the proteomics was analyzed about three years later than the stool collection.
In the first case, machine learning was used to reveal the top 20 bacterial clusters (operational taxonomic units, OTUs) to predict COVID-19 susceptibility. Most of the OTUs came from the following genera and families: Bacteroides, Streptococcus, Lactobacillus, Ruminococcaceae 119, Lachnospiraceae, and Clostridiales. Using this, over a fifth of the variability in PRS could be explained.
The analysis showed a superior correlation between the core OTU-predicted PRS and the actual PRS. There was also a close association between the 20 core OTUs and the 20 proteomic biomarkers predicting severe COVID-19. When stratified by age, the association was seen to be significant only in older age groups.
The results were duplicated in the prospective study. This shows that gut microbiome changes occur earlier than the change in the blood proteomics. If so, they could have a causal role.
A confirmatory study in a larger subcohort of 366 participants showed that 11 OTUs had a significant association with the inflammatory cytokines, either negative (Bacteroides, Streptococcus, and Clostridiales), or positive (Ruminococcus, Blautia, and Lactobacillus).
Fecal metabolites may link PRS, gut microbiome and inflammation
The researchers examined the link between the core gut microbial profile and the fecal metabolites among approximately 1,000 participants. They found that 45 metabolites in urine were significantly associated with over half the core OTUs.
Most of them were either amino acids, fatty acids, or bile acids, involved in three pathways. The amino acid level in the tissues is vital in maintaining healthy immunity because it depends on the metabolic stress pathways and the availability of nutrients. Reduced levels of specific amino acids suppress inflammation.
Thus, these metabolic pathways, modulated by diet and the dependent bacterial populations, may drive the effects of the gut microbiota on the metabolism of the host and inflammation.
The relative abundance of ACE2 and its function in regulating amino acid levels in response to dietary intakes, and in innate immunity, might be a second link between the gut microbiome and inflammation, which in turn predicts COVID-19 severity.
Do host and environmental factors matter?
The study also shows that 2.4% of the difference in susceptibility to the disease is explained by 9 factors related to the demographics and clinical characteristics of the sample, such as educational level, sex, and various physiological parameters like blood pressure and biochemistry. These indirectly modulate the composition of the gut microbiome.
Can the gut microbiome predict severe COVID-19?
The researchers suggest that in healthy people, the composition of the gut microbiome is highly predictive of the blood proteomic biomarkers that are linked to severe COVID-19.
The ‘cytokine storm’ (excessive levels of pro-inflammatory chemicals in the body) associated with severe COVID-19 should be treated effectively to reduce the fatality of the condition. The association of proteomic biomarkers with inflammatory molecules, especially in older age groups, indicates that the cytokine storm is a result of the underlying inflammation seen to be more common among this subgroup.
The researchers sum up: “The discovered core gut microbial features and related metabolites may serve as a potential preventive/treatment target for intervention, especially among those who are susceptible to the SARS-CoV-2 infection. They could also serve as potential therapeutic targets for drug development.”
medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.
Gou, W. et al. (2020). Gut microbiota may underlie the predisposition of healthy individuals to COVID-19. medRxiv. doi: https://doi.org/10.1101/2020.04.22.20076091. https://www.medrxiv.org/content/10.1101/2020.04.22.20076091v1.