Authors:  Julius Goepp MD, James Geyer MD, Thomas Patton MD

Progress in developing rationally-designed and replicable diagnostics and therapeutics for microbiome-associated disorders (MAD) has been impeded by an excessive focus on the taxonomic makeup of microbial ecosystems in disease and health¹. Until quite recently, definitions of “microbiomes” were primarily census-like, wherein the human microbiome is the “collection of all the microorganisms living in association with the human body^2-4.” Such a definition entirely ignores the actual functional contributions of each taxon to the overall biochemical milieu of the gut lumen, and can provide only general predictions about how an imbalanced, or dysbiotic microbiome produces observed phenotypic features of a disease.

The nearly unitary focus on taxa in MADs produces the so-called “Anna Karenina Effect,” which has confounded attempts to link MAD phenotypes with microbiome features. Tolstoy’s novel by that name opens with the phrase “Happy families are all alike; every unhappy family is unhappy in its own way.” In the taxon-centric “census” definition of microbiome, this is manifested by the apparent fact that “Dysbiotic individuals vary more in microbial community composition than healthy individuals⁵.” For example, people with certain phenotypes of Inflammatory Bowel Disease (IBD) have more variable microbial communities than do healthy individuals⁶, a finding that makes it difficult to reliably characterize IBD patients by microbiome taxonomic composition alone⁷.

A superior means of analyzing the impact of gut microbiota on the “host” phenotype is to examine the microbiome ecosystem from the genetic and molecular, rather than the intact organism, standpoint, and to apply principles of Systems Biology in a “Taxonomy Agnostic” fashion to understand how individual components (cells, molecules, pathways) of biological systems interact and ultimately give rise to an observed phenotype^1,8.

Strong support for this approach is rapidly accumulating.

A recent study by Tierney et al⁹, for example, showed that gene-level analyses of microbiome-disease relationships provide more robust associations with several microbiome-associated diseases (MADs) than do taxonomic-level analyses. At a still-higher level of resolution, bioactive microbial proteins were shown to be differentially enriched in IBD patients, many of which were carried by multiple individual genera.⁷ Indeed, in some cases in this analysis, proteins that were both enriched and depleted in IBD patients  were found within the same species (specifically Ruminococcus gnavus and Faecalibacterium prausnitzii). Functionally, bacterial pilin proteins (typically carried by Proteobacteria) were among those differentially enriched in IBD patients⁷; pilins are of crucial importance for many of these bacterial opportunistic pathogens (or “pathobionts”), mediating their attachment to human intestinal epithelial cell membranes and facilitating transmission of toxins and other metabolites that contribute to increased inflammation, loss of intestinal barrier integrity, and genomic damage^7,10,11. Similarly, amyloid fibers called curli, important biofilm components, are produced by multiple members of the Family Enterobacteriaceae, and may contribute to disease phenotype in several gut and extraintestinal disorders, including some neurodegenerative diseases^12-17. These findings are consistent with the “insurance hypothesis” notion that biological function of a microbial community can be maintained even in the absence of specific taxa, so long as others can provide that function¹⁸.

Taken together, a more nuanced and inclusive view is that a microbiome consists of “The genes and genomes of the microbiota, as well as the products of the microbiota and the host environment^2,19.”

In short, a microbiome ecosystem in health comprises a molecular milieu that promotes homeostasis and self-regulates to sustain metabolic equilibrium (Figure 1).

Figure 1. Microbiome Ecosystem in Health¹.  In health, a self-regulated ecosystem forms, driven by diet and environmental factors that promote a diverse, pathogen-resistant microbiota, which in turn produces predominantly beneficial metabolites and limited exposure to deleterious microorganism-produced bioactive molecules. A normal mucous layer is maintained, and epithelial cell barrier function is sustained, limiting the amounts of lumenal materials translocated into the circulation and maintaining a low-inflammation, low oxidant-stress environment. Created with BioRender.com

By contrast, a dysbiotic microbiome can be viewed as a state of ecological “molecular pollution,” wherein multiple taxa contribute to a disease phenotype in a MAD, via their collective output of bioactive proteins (Figure 2).015;12(2):81-90.

Figure 2. Microbiome Ecosystem in Dysbiosis¹. Dysbiosis is triggered by dietary and other environmental factors, which differentially support opportunistic pathogens/pathobionts. “Blooms” of pathobionts exclude beneficial bacteria from ecological niches, resulting in low-diversity populations dominated by bacteria contributing to “molecular pollution” in the ecosystem. Such “pollution” consists of toxic or damaging bacterial metabolites, virulence factors, and microbe-associated molecular patterns (MAMPs). These disruptions contribute to a thinning mucous layer, biofilms that can harbor and protect pathobionts, and damaged epithelial junctional proteins resulting in influx of bioactive molecules through epithelial cells and into circulation, where they drive elevated levels of inflammation. Created with BioRender.com

Those bioactive proteins contribute in predictable ways to ultimate phenotype, by their global effects on fundamental properties as inflammation, loss of gut barrier function, increased oxidant stress, etc.  In such a framework, it becomes possible to identify with considerable confidence specific bioactive proteins – independent of their taxonomic origin – whose reduction within the system will yield a favorable modification of the molecular milieu, potentially resulting in a return to a phenotype free of the MAD in question.

This final observation holds the key to future microbiome diagnostics and therapeutics advances – when we understand our targets and their contextual roles in producing disease phenotypes, we can rationally design taxon-agnostic, systems biology-based methods for their detection, analysis, and ultimately modification to produce multiple new lines of microbiome therapeutic drugs.

REFERENCES

1. Brumfield KD, Cox P, Geyer J, J G. A Taxonomy-Agnostic Approach to Targeted Microbiome Ther-apeutics – Leveraging Principles of Systems Biology Pathogens (Basel, Switzerland). 2023;In Press.
2. Heinzel S, Aho VTE, Suenkel U, et al. Gut Microbiome Signatures of Risk and Prodromal Markers of Parkinson Disease. Ann Neurol. 2020;88(2):320-331.
3. Dekaboruah E, Suryavanshi MV, Chettri D, Verma AK. Human microbiome: an academic update on human body site specific surveillance and its possible role. Archives of microbiology. 2020;202(8):2147-2167.
4. Sender R, Fuchs S, Milo R. Revised Estimates for the Number of Human and Bacteria Cells in the Body. PLoS Biol. 2016;14(8):e1002533.
5. Zaneveld JR, McMinds R, Vega Thurber R. Stress and stability: applying the Anna Karenina principle to animal microbiomes. Nature microbiology. 2017;2:17121.
6. Holmes I, Harris K, Quince C. Dirichlet multinomial mixtures: generative models for microbial metagenomics. PloS one. 2012;7(2):e30126.
7. Zhang Y, Bhosle A, Bae S, et al. Discovery of bioactive microbial gene products in inflammatory bowel disease. Nature. 2022.
8. Snoep JL, Bruggeman F, Olivier BG, Westerhoff HV. Towards building the silicon cell: a modular approach. Biosystems. 2006;83(2-3):207-216.
9. Tierney BT, Tan Y, Kostic AD, Patel CJ. Gene-level metagenomic architectures across diseases yield high-resolution microbiome diagnostic indicators. Nature communications. 2021;12(1):2907.
10. Middeldorp JM, Witholt B. K88-mediated binding of Escherichia coli outer membrane fragments to porcine intestinal epithelial cell brush borders. Infection and immunity. 1981;31(1):42-51.
11. Sémiramoth N, Gleizes A, Turbica I, et al. Escherichia coli type 1 pili trigger late IL-8 production by neutrophil-like differentiated PLB-985 cells through a Src family kinase- and MAPK-dependent mechanism. J Leukoc Biol. 2009;85(2):310-321.
12. Nicastro LK, Tursi SA, Le LS, et al. Cytotoxic Curli Intermediates Form during Salmonella Biofilm Development. Journal of bacteriology. 2019;201(18).
13. Barnhart MM, Chapman MR. Curli biogenesis and function. Annu Rev Microbiol. 2006;60:131-147.
14. Tursi SA, Puligedda RD, Szabo P, et al. Salmonella Typhimurium biofilm disruption by a human antibody that binds a pan-amyloid epitope on curli. Nature communications. 2020;11(1):1007.
15. Buret AG, Motta JP, Allain T, Ferraz J, Wallace JL. Pathobiont release from dysbiotic gut microbiota biofilms in intestinal inflammatory diseases: a role for iron? Journal of biomedical science. 2019;26(1):1.
16. Chew SS, Tan LT, Law JW, et al. Targeting Gut Microbial Biofilms-A Key to Hinder Colon Carcinogenesis? Cancers (Basel). 2020;12(8).
17. Moran-Garcia N, Lopez-Saucedo C, Becerra A, et al. A Novel Adult Murine Model of Typical Enteroaggregative Escherichia coli Infection Reveals Microbiota Dysbiosis, Mucus Secretion, and AAF/II-Mediated Expression and Localization of β-Catenin and Expression of MUC1 in Ileum. Frontiers in cellular and infection microbiology. 2022;12:885191.
18. Yachi S, Loreau M. Biodiversity and ecosystem productivity in a fluctuating environment: the insurance hypothesis. Proceedings of the National Academy of Sciences of the United States of America. 1999;96(4):1463-1468.
19. Whiteside SA, Razvi H, Dave S, Reid G, Burton JP. The microbiome of the urinary tract–a role beyond infection. Nat Rev Urol. 2015;12(2):81-90.