When it comes to the next big thing in drug development, biotechnology mogul Emily Drabant Conley has always been one step ahead of the curve. During her time at the DNA testing company 23andMe, she witnessed genetic data expand from a tool to uncover an individual’s ancestry and genetic health risks to a powerful drug discovery platform. “I saw firsthand how drug developers are able to leverage genomic information for things like discovering novel therapeutic targets, profiling subsets of patients, and determining how they ought to structure their trials,” Conley said. “Really important aspects of drug development were changed by human genetics, and I foresaw a similar transformation coming with the microbiome.” 

While scientists have long understood pathogenic microbes that invade the body to drive infection, researchers have only more recently come to understand the critical activities that native microbes perform for human health and to appreciate the role of microbiome disruption in disease. Now as CEO of the company Federation Bio, Conley develops drugs that restore and enhance microbiome function for a broad range of therapeutic applications. By harnessing advances in microbial genetic sequencing and culturing technology, she and her team design and manufacture communities of microbes with unprecedented complexity. The company’s lead program focuses on developing a gut microbiome replacement therapy for enteric hyperoxaluria, a metabolic disorder that leads to chronic kidney stones. The researchers are currently testing their drug in a phase 1 clinical trial, marking the largest rationally designed bacterial consortium to reach this point in the pipeline. The trial will help uncover the potential of this novel treatment approach, answer key questions about its feasibility in humans, and inform the development of therapeutic microbial communities for other microbiome-linked indications.   

A gut microbiome in a pill

As a dynamic and interconnected network of microbes, the gut microbiome influences health in myriad ways. Due to the intricate interactions within the collective microbial community, scientists have explored several diseases associated with microbiome disruption as candidates for fecal microbiota transplant (FMT) treatment, in which a stool sample capturing a donor’s healthy gut microbiome is administered into a patient (1). FMT, however, faces limitations such as extreme variability in the composition of the donor microbiome, the risk of transferring pathogens to the patient, and the inability to manufacture a drug at scale due to donor dependence. 

The team at Federation Bio aimed to combine the comprehensiveness of FMT with the control provided by traditional drug development. “If we were able to manufacture something that was just as complex as a microbiome, we could make a drug that's essentially a microbiome replacement but in a pill, and we could design that drug so that it targets multiple mechanisms of action in a very fit-for-purpose way,” Conley said. “This is the power of a large, rationally designed consortium: that we can go after these really complex indications where we need to hit multiple [mechanisms of action] to be successful.” 

One such indication is enteric hyperoxaluria, which affects approximately 250,000 people in the United States — frequently those with impaired intestinal absorption of fat due to underlying gastrointestinal conditions or gastric bypass surgery — but has no approved treatments. People with enteric hyperoxaluria generally have a low-diversity gut microbiome where the microbes cannot effectively degrade the oxalate they ingest from dietary sources such as chocolate, spinach, and raspberries. Excess fat in the gut lumen due to malabsorption traps calcium, preventing it from binding to and sequestering the oxalate for excretion. Patients also typically have poor gut epithelial barrier function, helping free oxalate to pass through to the bloodstream. The oxalate then travels to the kidneys, where it meets up with calcium to form calcium oxalate complexes. These complexes create recurrent kidney stones that ultimately lead to renal damage.  

To develop a microbial consortium that could tackle these intersecting disease pathways, the team realized that they needed to clear out the patient’s existing microbiome. “Conceptually, you can think of this as a way to make space inside [the gastrointestinal tract] — both physical space and metabolic space,” Conley said. “You wouldn't give someone a new liver and leave the old one in.” 

Andreas Grauer, the chief medical officer at Federation Bio, added, “If the dysbiosis persists and we're trying to just put the new microbiome on top, then that may not be therapeutically as efficient as if we create that clean slate.” The researchers developed an antibiotic pretreatment regimen to clear out the original gut microbiome, which is also a common practice in FMT. 

Facing the opportunities and challenges provided by a blank slate, the team set out to develop a complete gut microbiome replacement fortified with specific activity for enteric hyperoxaluria. They collected fecal samples from healthy volunteers and immediately placed them in an anaerobic chamber to preserve the oxygen-averse microbes that make up the majority of the gut microbiome. They then analyzed these samples using advanced metagenomic sequencing. 

Traditionally, microbiome genetic sequencing entailed analyzing only the 16S gene in each organism to identify the species of microbes that comprise the community. With this technique, however, “No one knew how to define what a microbiome was because the diversity of life there at the name level, the 16S locus, was quite large across humans,” said Lee Swem, the chief scientific officer at Federation Bio. Thanks to advances in sequencing technology, the researchers could sequence and assemble entire chromosomes of genetic information from each organism, moving beyond species-level data to capture metabolic function. “We were able to look at thousands of microbiome samples that had been deposited and sequenced that deeply to figure out the metabolic pathways that were present in each healthy individual,” Swem said. 

The team exposed the fecal samples to various culture media that promote the growth of different types of bacteria, allowing them to enrich specific populations and iteratively isolate each colony-forming strain to create a library of strains. They sequenced the genome of each strain in the library to elucidate its metabolic activity, enabling them to identify microbes that fulfill the conserved metabolic functions across all the healthy human microbiomes. In this way, they designed a backbone for their therapeutic community that spans all six phyla found in the normal gut microbiome. 

The researchers then set out to design a cohort of microbes that could specifically combat enteric hyperoxaluria. In searching the literature, they encountered the species Oxalobacter formigenes, which is present in about 20 percent of healthy human microbiomes. It relies on oxalate for its sole carbon source, making it extremely efficient at degrading oxalate in the gut lumen. This microbe also secretes a molecule that reverses the molecular pumps that push oxalate out of the gut and into the bloodstream, an adaptive survival mechanism for when oxalate is in short supply, enabling it to suck oxalate from the blood back into the gut. The team identified three unique strains of O. formigenes from their donor fecal samples to incorporate into their consortium. 

To maintain maximum activity, O. formigenes depends on a series of chemical reactions from fellow microbes. The oxalate degradation reaction produces the compound formate as a byproduct, which slows down subsequent oxalate consumption. Formate reducing microbes break down the formate byproduct, allowing O. formigenes to chew through oxalate at full speed. However, these formate reducers require hydrogen gas to do their job. Other microbes that ferment polysaccharides produce hydrogen gas in the process, thereby fueling the formate reducers. The researchers mined their microbial genomic data for genes encoding these enzymatic reactions and identified formate reducers and polysaccharide fermenters to add to the community, closing the oxalate degradation chemical loop. In the same way, they identified another class of microbes to incorporate that produces short chain fatty acids, which fortify the gut epithelial barrier, reducing oxalate penetration into the bloodstream. “This, I think, really speaks to the multi-nodal capability of a consortium,” Conley said. “We actually are able to do a lot of different things because we can use different constellations of microbes to do them.”

Because different microbes thrive depending on the food their host eats, the researchers included multiple microbes that perform the same function using different energy sources. For example, to sustain the polysaccharide fermentation reaction that feeds the downstream oxalate degradation activity, the team incorporated strains with polysaccharide preferences spanning the dominant dietary starches, such as those found in potatoes or rice. “We have designed the consortium to be able to metabolize lots of different kinds of dietary inputs and provide all the core outputs,” Conley said. 

We actually are able to do a lot of different things because we can use different constellations of microbes to do them. 
- Emily Drabant Conley, Federation Bio 

The team tested several candidate communities in a mouse model of enteric hyperoxaluria and identified one that reduced urinary oxalate levels, which correlate with kidney stone formation, by up to 65 percent. They observed this high degree of oxalate reduction whether they fed the mice a healthy, grain-based diet or a high fat, high sugar diet, suggesting that the community features enough redundancy to form a functional ecosystem under a wide variety of human diets. 

The final consortium, termed FB-001, consists of 148 individual bacterial strains. Using naturally occurring microbes rather than engineered ones allowed the researchers to bypass Good Laboratory Practice toxicity studies and to overcome challenges related to genetic drift. “If someone engineered me to run five miles a day, but that’s not advantageous for me to do, I’m going to drop that somehow; I’m going to mutate away from that and not do that. And that’s what microbes do when you engineer them,” Swem said. In contrast, natural microbes are unlikely to budge from the function they have specifically evolved to perform. 

Manufacturing a consortium of this scale from purified cell lines necessitated new breakthroughs in microbe culturing technology. Each cell line typically requires its own fermentation unit, but fermenting such a complex community one strain at a time would take far too long and cost far too much. The team therefore devised a novel method to grow tens of cell lines together in a single fermenter, which required special approval from the FDA. They had to determine how to balance the optimal culture conditions for multiple microbes simultaneously to get them to efficiently grow when combined. In the end, each fermentation family yields a powder, and the researchers blend these powders at a certain ratio to make up the final oral drug capsule.

Long-term lessons

With the great effort required to create a full microbiome replacement comes great value. “I am a big proponent of it,” said Eric Alm, a microbiome engineer at the Massachusetts Institute of Technology (MIT). “With really small, well-defined cocktails of organisms, you're not replacing much of the biomass in the gut, generally. ... So, if you only have a few species that you're putting in, unless a few of those just happen to be the most dominant members of the microbiome, then you're not affecting things a lot,” he said. “The tricky part will be engraftment. You can put bacteria into the gut, but whether they stably colonize, that’s a question that we don’t fully understand yet.”

While engraftment is a common challenge, transplanting a complete microbial ecosystem may also be advantageous by this measure. “If you put in a couple of liver cells, that's very different than taking out the whole liver and putting in a brand new one,” Conley said. “The concept is that you get that stability because now you have a set of microbes that can function synergistically and harmoniously, and they can occupy their metabolic niches.” 

Still, analogous FMT communities have shown considerable variability in which donor microbes engraft in the recipient (2). The team, however, has prepared for this possibility in the design of their consortium. “We don't expect that all 148 of these strains will engraft in every patient. We've built it so that there's multiple strains that perform similar functions so that we can get a subset of these strains engrafting and still get the complete metabolic function that we need,” Conley said.

 One potential challenge with using a manufactured microbial community is that, unlike in FMT, the microbes have not all cohabitated together inside one human body. “There have been a couple of studies that have shown that cross-feeding or syntrophies are more likely to occur in strains of bacteria that have coevolved with each other in the same donor. And so, by assembling consortia where you pull strains from many different donors, you could be missing that,” said Alice Cheng, a gastroenterologist and hepatologist at Stanford Medicine (3). In a recent study in Cell, Cheng and her colleagues took an approach similar to Federation Bio’s to construct a model of the gut microbiome using data from the NIH Human Microbiome Project and human fecal samples (4). Initially, the researchers were concerned that combining strains from multiple donors could result in an unstable, potentially antagonistic community. On the contrary, they observed that even a collection of strains lacking a shared history settled into a stable configuration. “Now, the trick is to better understand why. So, it calls into question, how important are these donor-specific interactions?” Cheng said. 

The team at Federation Bio also hasn’t had any trouble with a microbial consortium assembled from different donors. When the researchers used metagenomic sequencing to analyze stool samples from mice pretreated with antibiotics and then administered FB-001, they observed that “a full microbiome occurred. You see a plethora of strains across multiple phyla engraft in the mouse, and that’s an ecosystem,” Swem said. “But really, what's exciting is can you test it in humans, which is where we're at now.” 

In part one of the phase 1 clinical trial, the researchers induced hyperoxaluria-like conditions in healthy volunteers by feeding them a high oxalate, low calcium diet. After the antibiotic pretreatment regimen, they orally administered either the FB-001 capsules or a placebo and evaluated reduction in urinary oxalate levels and other urine properties that signal a propensity to form kidney stones. They also assessed the community’s metabolic activity by measuring key metabolites such as short chain fatty acids. The researchers monitor the participants for any adverse effects, looking closely at bloating, diarrhea, and other gastrointestinal symptoms of FMT. They also metagenomically sequence the participants’ stool samples to see which members of the community engraft. If all goes well, the team will test FB-001 in people with enteric hyperoxaluria in part two.

The potential for a therapy with lasting power is especially attractive for a disorder where patients have chronic kidney stones. “Sometimes, those kidney stones are separated by months or years, and to ask a patient with that kind of a condition to take a pill every day, multiple times a day, I think can be difficult versus doing a single course of treatment that then can durably engraft,” Conley said. “There’s just a tremendous benefit in terms of compliance.” 

However, it is possible that taking antibiotics for an infection after receiving FB-001 could kill off important members of the consortium. The team plans to continue to follow patients treated with FB-001 to monitor the effect of routine antibiotic use on oxalate levels and learn whether they need to readminister the drug.   

While they face a long road ahead, the researchers appreciate how far they’ve already come. “This is a totally novel therapeutic modality,” Conley said. “When the company was founded, it was unclear if there was a regulatory path and a manufacturing feasibility path. We have now proven both of those things. We can indeed design and manufacture at scale under [Good Manufacturing Practice] this type of drug product, and we've gotten alignment with the FDA to be able to test it in humans.”

Obtaining human data on this first-in-class therapeutic is a valuable milestone that can guide the development of other complex microbial consortia. “Is it safe and well tolerated? And do the bugs engraft? Those are two key questions coming out of this trial that will be, I think, generalizable and help inform future studies,” Conley said. 

Grauer added, “If we can reliably produce a metabolically complete microbiome and use it in clinical trials, we can try using it in all kinds of diseases where dysbiosis has been associated with disease, and that list is long.”

If we can reliably produce a metabolically complete microbiome and use it in clinical trials, we can try using it in all kinds of diseases where dysbiosis has been associated with disease, and that list is long. 
- Andreas Grauer, Federation Bio 

The team at Federation Bio can readily adapt the design of the consortium to target other diseases, selecting microbes with the desired metabolic activity from their library to place on their backbone of strains with core microbiome functions. They are interested in expanding to other metabolic disorders and cholestatic diseases, which are characterized by defective bile acid transport, and have already launched programs in inflammatory bowel disease and cancer immunotherapy. The researchers recently partnered with the MD Anderson Cancer Center to design and manufacture a microbial community that mimics the microbiome of a unique cancer patient. This person was effectively treated by a checkpoint inhibitor and has served as an FMT donor to improve responses to immunotherapy in other cancer patients.    

“Everybody's seen the potential of fecal material transplant, but nobody could figure out how to actually control it, scale it, manufacture it, and turn it into a drug product,” Swem said. “To be able to manufacture something of this tremendous diversity for the first time, akin to the diversity found in FMT, I think is revolutionary.”

References

1. Ser, H.-L., Letchumanan, V., Goh, B.-H., Wong, S.H., & Lee, L.-H. The use of fecal microbiome transplant in treating human diseases: too early for poop? Front Microbiol  12, 519836 (2021). 
2. Ianiro, G. et al. Variability of strain engraftment and predictability of microbiome composition after fecal microbiota transplantation across different diseases. Nat Med  28, 1913-1923 (2022).
3. Goodman, A.L. et al. Extensive personal human gut microbiota culture collections characterized and manipulated in gnotobiotic mice. Proc Natl Acad Sci USA  108, 6252-6257 (2011). 
4. Cheng, A.G. et al. Design, construction, and in vivo augmentation of a complex gut microbiome. Cell  185, 3617-3636 (2022).