DAY 3 - The Future of Microbiome Therapeutics: Bright or Bleak?

Weston Whitaker, PhD, from the Stanford University School of Medicine, in Stanford, California, discussed his experiences with a microbiome startup, Novome Biotechnologies, and development of engineered therapeutic microbes. Fecal microbiota transplant (FMT) studies have demonstrated the therapeutic power of reprogramming the microbiome, but the field has stalled because of uncertainty about what the bacteria are doing. As an alternative to FMT, his group develops engineered microbes to be introduced (engrafted) on top of the existing microbiome. They selected hyperoxaluria as the lead indication to prove their approach; hyperoxaluria is addressable with colonic activity, is a significant unmet need, there is very little disease biology risk, it provides an inexpensive and early efficacy readout, and development can be financed and commercialized without pharma. Decreasing oxalate in the colon before it gets absorbed reduces the risk of kidney stone formation and renal failure. They used an engineered strain that used porphyran as a food source and contained a porphyran-inducible promoter-driven essential gene to allow control over strain growth. They found that the engineered strain led to clinically meaningful reduction in urine oxalate in rats with diet-induced hyperoxaluria, and phase 1 results in healthy humans (single dose, once a day every day for 2 weeks) found it was safe and well tolerated and strain abundance slowly decreased over time. In phase 2 trials in patients with hyperoxaluria, there was a decrease in urine oxalate, and correspondence between engrafted bacteria and the oxalate pathway; however, engraftment was not stable, which would limit treatment durability. They are looking at other indications: bile acid transformation, inflammatory bowel disease in partnership with Genentech, and immune-oncology. 

The scientific lessons learned were: very high levels of engineered bacteria are needed for efficacy, delivering prebiotics to create an exclusive niche enables high colonization levels in healthy volunteers, eliminating detectable preclinical fitness defects may be necessary to avoid stability problems, and it is important to closely match preclinical animal models with expected clinical conditions.

The startup lessons learned were: FDA takes a scientific approach to GMO risks that fits within the existing standard framework, manufacturing challenges are mitigated by collecting early efficacy data with simpler formulation (e.g., dosing a glycerol stock), it is difficult to find many microbiome indications that have well-defined enough mechanisms in humans for venture capital investment, and fundraising for microbiome startups is quite difficult.

Matthew Odenwald, MD, PhD, from The University of Chicago in Illinois, reviewed the future of microbiome therapeutics. He described the evidence that symbiotic organisms in the gut enhance human resistance to disease, as can be seen in antibiotic-associated dysbiosis. A diverse microbiome with beneficial metabolites loses diversity and metabolic concentration upon exposure to antibiotics, leaving them susceptible to colonization by harmful strains. IBD, IBS, allergy, obesity, and metabolic syndrome, and alcohol-related hepatitis have also been shown to have distinct microbiome differences. Unfortunately, in vitro and murine models do not completely recapitulate human biology, microbiome complexity, diet variability, and disease; human clinical trials are needed, but are complex due to population heterogeneity.

Current FDA-approved microbiome therapies (Rebyota from Ferring and VOWST by Seres) were developed using a top-down approach, with many unanswered questions regarding mechanisms of action, engraftment success, and the effect of variable starting material. As a result, the bottom-up approach has become more attractive, such as probiotics that contain selected strains. In the absence of rigorous testing of probiotics in clinical trials, only general claims can be made about gut health, with no disease-specific claims. Probiotics may also interfere with the established microbiome or make it more difficult to recover from antibiotic-associated dysbiosis.

As a more precise assembly of bacterial consortia, VE303 by Vendata Therapeutics contains 8 strains and is effective for treatment of C. difficile colitis (Menon et al. Nature Med. 2025). Efficacy depends on early engraftment. 

There is also a need to improve identification of patients who would benefit from microbiome therapeutics. Dr. Odenwald developed a rapid metabolomics assay to identify severe dysbiosis based on fecal material butyrate and deoxycholic acid (DCA) (Mullowney et al. bioRxiv 2025 and Odenwald et al. Gast Hep Adv. 2025). The assay can also be used to identify missing strains from the microbiome to target new therapeutics. From the DFI Symbiotic Strain Bank of 1695 unique strains from 120 species derived from human samples, they picked a panel of strains to create different consortia based on their metabolomes (e.g., butyrate producers) and ability to cross-feed. An investigational new drug application was developed to manufacture the consortia at a cGMP facility at University of Chicago using rigorous quality control mechanisms. The planned MARCO adaptive trial will start soon; patients with liver disease will be screened for metabolites (i.e., butyrate and DCA), and those that are deficient will be given consortium A to determine the effect on metabolites, engraftment, and safety. Organisms can be substituted in to get the metabolites to the desired level.

Amir Zarrinpar, MD, PhD, from the University of San Diego, in La Jolla, California, discussed current and new targets for microbial therapeutics. The excitement around microbiome therapeutics is based on evidence of the microbiome’s involvement in a myriad of disorders and the possibility that it may have curative potential rather than just a disease-modifying effect. The goal of microbiome therapy is to engraft bacteria that can be controlled, can detect disease, perform a therapeutic function, and does what the gut microbiome already does. SER-109 (VOWST) and RBX2660 (Rebyota) are focused on compositional restoration, but better controlled therapies are possible through bacterial engineering (described by Dr. Whitaker) or compositional selection (described by Dr. Odenwald). There are several living cell diagnostics and therapeutics that are being used to respond to and detect inflammation, bleeding, pathogens, and tumors; the question now is whether they can be used for treatment. Studies have examined engineered IL-10–expressing Lactococcus lactis strains for the treatment of chronic DSS. Others have looked at L. lactis strains that express 1L-27 or E. coli that secrete anti-TNF neutralizing nanobodies. Creative and complex circuits are being introduced to the gut with engineered probiotics and tested in germ-free mice, but it may be time to test these when added to the existing microbiome. Engraftment faces challenges in the luminal environment because the microbiome therapeutic must compete with far better adapted competitors, essentially “letting a hamster out into the forest to survive.” Engineered probiotics that have been shown to be safe yielded disappointing results in humans because they were not being engrafted in the context of the existing microbiome. Dr. Zarrinpar described new approaches to improving the survival of microbiome therapeutics, including Metagenomic Alteration of Gut Microbiome by In Situ Conjugation (MAGIC), where the introduced strain shares genomic information with other strains before it dies out. Inspired by CAR-T, his group is also looking at taking E. coli already adapted to a host, engineering it, and then transplanting it back; they found that bacteria engineered to express bile salt acids and then transplanted back into mice remained engrafted for the entire lifetime of the mouse, with retention of function and BSH expression for 2 years. Dr. Zarrinpar concluded that engineered microbial therapies can stably colonize the gut to deliver precise therapeutic functions. While successful microbial therapeutics focused on composition have been approved, future therapies will need to focus on microbial function and the use of engineered bacteria to robustly modulate the microbiome in controlled ways. Leveraging native bacteria and unique metabolic niches will enable predictable long-term engraftment in complex microbiomes.