Hidden Gut Bacteria CAG-170 Discovered as Key Health Marker

The Improving Gut Health Team • 2026-04-02 • updated Mon Nov 16 • 8 min read
Editorial byline — how this site is written · educational content, not medical advice

Cambridge researchers identify CAG-170, a largely uncultured gut bacteria group repeatedly enriched in healthy people across 39 countries. This 'hidden microbiome' lineage has Vitamin B12 biosynthesis capacity and may become a next-generation probiotic target.

Quick answer

CAG-170 is a genus of gut bacteria that scientists have almost never grown in a lab, yet it keeps showing up as the strongest health-linked lineage in the human gut. In a 2026 global meta-analysis of 11,115 gut microbiome samples from 39 countries, CAG-170 was the single most health-associated bacterial group — lower levels tracked with Crohn's disease, ulcerative colitis, obesity, and myalgic encephalomyelitis/chronic fatigue syndrome, and higher levels marked stable, balanced microbiomes. Its genomes carry an unusually complete vitamin B12 biosynthesis pathway, and its abundance plus subspecies diversity correlated negatively with dysbiosis over time. Important caveat: this is an association study. CAG-170 is a candidate health marker, not a proven cause of good health, and no CAG-170 probiotic exists yet.

This article is one piece of the microbiome-research cluster on this site — the stool-testing guide covers how these findings reach commercial tests, and whether at-home gut microbiome tests work covers what they can and can't detect today.

What the study actually did

A team led by Dr. Alexandre Almeida at the University of Cambridge (first author Ana C. da Silva) published the analysis in Cell Host & Microbe in 2026. Instead of new lab work on live bacteria, the study re-analyzed 11,115 existing human gut metagenomes — shotgun-sequenced stool samples — collected across 39 countries from healthy people and patients with 13 different noncommunicable diseases.

The researchers matched every detectable species against the Unified Human Gastrointestinal Genome (UHGG) catalog and split them into two piles: species with a cultured, laboratory-grown representative, and species known only from their DNA. That second pile is large — roughly 60% of known gut species have never been cultured, and in the average stool sample in this study about 31% of the species present were uncultured, representing around 12% of the microbial reads.

The headline result: uncultured bacteria were significantly overrepresented among the species enriched in healthy people. Of 373 health-associated species, 189 (51%) were uncultured, versus 128 of 342 disease-associated species (37%). A third of the microbiome's species list is nearly as clinically informative as the 88% of reads we can culture — and we have barely studied it.

Why CAG-170 stood out

The genus CAG-170 — a branch of the Oscillospiraceae family represented by 13 species and 1,046 genomes in the UHGG catalog — topped every ranking the researchers applied:

  1. Disease comparisons. 11 of the 13 CAG-170 species detected were associated with health (i.e., higher in controls than in cases) across Crohn's disease, ulcerative colitis, obesity, and myalgic encephalomyelitis. It was the only genus in the whole catalog that simultaneously had more than 10 health-associated species, an effect size above 0.5, and more than 75% of its species still uncultured.
  2. Ecological networks. In co-occurrence networks built from healthy populations on every continent, CAG-170 was the most interconnected genus of all — cultured or uncultured. Of the 31 "hub" species at the top 1% of network centrality, 27 were uncultured, and 12 of them were CAG-170.
  3. Longitudinal stability. In a reanalysis of the Human Microbiome Project's time-series data, individual CAG-170 levels stayed stable over a full year, and both its abundance and its subspecies diversity were negatively correlated with dysbiosis scores — the more CAG-170 and the more genetically varied it was, the less imbalanced the rest of the ecosystem.

Worth keeping in proportion: CAG-170 was detected in about 20% of samples at a median relative abundance of just 0.07%. This is not a dominant organism like Bacteroides or Faecalibacterium — it is a low-abundance, specialist lineage whose signal is consistent enough across diseases and continents to matter statistically.

The vitamin B12 connection

Comparing 1,046 CAG-170 genomes against 4,025 other Oscillospiraceae genomes, the study found 14 genes of the cobalamin (vitamin B12) biosynthesis pathway significantly enriched in CAG-170 — a more complete B12-making toolkit than its close relatives have. Genomic prediction also flagged strong cross-feeding potential: the capacity to supply metabolites, including B12, to neighboring bacteria.

That fits a broader pattern. A systematic survey of B-vitamin biosynthesis across gut genomes found that the human microbiome operates as a co-operating B-vitamin factory, with most organisms producing only part of any given pathway and exchanging the intermediates — so a B12-producing member of a healthy network plausibly feeds the whole community rather than the human host directly.

Two more genome-level findings round out the picture. CAG-170 was depleted of pro-inflammatory machinery (LPS/rhamnose biosynthesis genes) relative to its relatives, and it showed signs of arginine auxotrophy — it lacks key arginine-synthesis genes. That last detail has a practical consequence: the one successful cultivation of a CAG-170 isolate (strain POC01, described in 2022) used arginine-supplemented culture medium, which now looks like the reason it worked.

Why "hidden" is the right word

Most of these bacteria have never been grown in a laboratory dish. The landmark demonstration that the "unculturable" label is premature came from the Browne lab's 2016 work, which grew 320 previously unknown species from human stool by mimicking gut conditions — including oxygen-free handling and long incubations — and found extensive spore formation in this uncultured reservoir. Sporulating, oxygen-sensitive organisms simply die in standard aerobic plating, which is why reference libraries missed them for a century.

The UHGG catalog that made the CAG-170 analysis possible came out of that same push. The 2019 "genomic blueprint" reconstructed 1,952 new species-level genomes from metagenomes alone; the 2021 unified catalog extended it to 204,938 reference genomes covering 4,744 prokaryotic species, most with no cultured representative. Almeida, who led both catalog efforts and the 2026 analysis, put it this way in the University of Cambridge announcement: "Our earlier work revealed that around two-thirds of the species in our gut microbiome were previously unknown. No-one knew what they were doing there – and now we've found that some of these are a fundamental and underappreciated component of human health."

Uncultured doesn't mean unimportant: the cholesterol evidence

CAG-170 is the newest headline, but the clearest proof that "hidden" bacteria do real metabolic work predates the meta-analysis. In 2020, researchers showed that bacteria never grown in a lab — identifiable only by their genes — carry a cholesterol-metabolizing program, and that people carrying more of these organisms had measurably lower cholesterol levels Kenny DJ, Plichta DR, Shungin D, 2020. That study is the template for why the 2026 finding matters: organisms we couldn't even name a few years ago are performing jobs with direct health consequences, and the meta-analysis suggests there are far more of those jobs than anyone had tallied.

A scoping review of the microbiome's "dark matter" makes the same point from the survey side: low-abundance, hard-to-culture organisms may play outsized roles in both health and disease precisely because they've been invisible to culture-based methods de Cena JA, Zhang J, Deng D, 2021. The CAG-170 result is the most systematic quantification of that blind spot so far.

Next-generation probiotics: the honest version

The Cambridge team explicitly suggests CAG-170 could eventually inform "probiotics aimed at supporting them" — Almeida: "The probiotic industry hasn't kept up with gut microbiome research; people are still using the same probiotic species that were being used decades ago." But the same paper's limitations section says plainly what the industry marketing will omit: these results are statistical associations and genomic predictions. They do not establish causality, and with roughly one cultured isolate in existence, no one can yet test whether raising CAG-170 improves anyone's health.

The strain-specific trial evidence we do have concerns exactly the culturable organisms probiotics already use — for example, a 2024 randomized double-blind placebo-controlled trial of Bifidobacterium longum CECT 7347 in diarrhea-predominant IBS Srivastava S, Basak U, Naghibi M, 2024. Nothing in the uncultured fraction has anything comparable, and cannot until someone learns to grow it at scale. So treat any future "hidden microbiome" supplement claiming CAG-170 content with the same skepticism the marketing deserves — and the same skepticism applies to the probiotic aisle generally; see how to choose a probiotic for histamine issues and how to choose a microbiome test before paying for anyone's claims.

There is one clinical procedure that already delivers the complete community, uncultured microbes included: fecal microbiota transplantation. Its best-established use is recurrent C. difficile infection, where donor-feces infusion beat vancomycin head-to-head in a landmark randomized trial van Nood E, Vrieze A, Nieuwdorp M, 2013 — success that may partly rest on transferring hidden microbes no supplement can deliver. It remains a clinical procedure for a specific indication, unproven and unregulated for general gut health.

What you can act on today is the ecological, not the bacterial, lesson. CAG-170 sits at the center of healthy gut networks as a fiber-fermenting, B12-exchanging team player. The inputs that grow a diverse, stable microbiome — varied plant fiber, minimal unnecessary antibiotics — are the same ones that support rare specialist taxa like this one. Our prebiotic fiber comparison and the piece on microbiome diversity vs. symptom control cover the practical side; if you want to see where your microbiome stands, the dysbiosis index explainer covers how network-level measures are moving into testing, and microbiome patterns linked to early disease detection shows where sequencing-based health markers are heading.

Key takeaways

  • CAG-170 is an almost entirely uncultured gut bacterial genus that ranked as the strongest health-associated lineage in a 2026 meta-analysis of 11,115 gut metagenomes from 39 countries.
  • Lower CAG-170 levels tracked with Crohn's disease, ulcerative colitis, obesity, and ME/chronic fatigue syndrome across 13 diseases studied.
  • It is the most interconnected genus in healthy gut networks and stayed individually stable over a year, with abundance and subspecies diversity tracking low dysbiosis.
  • Its genomes carry an enriched vitamin B12 biosynthesis pathway and cross-feeding capacity — predicted from DNA, not yet shown in live organisms.
  • Uncultured bacteria already have a proven metabolic role: cholesterol-metabolizing organisms known only from their genes track with measurably lower host cholesterol.
  • Fecal transplants work clinically (strongest evidence: recurrent C. difficile) precisely because they transfer the complete community, uncultured microbes included.
  • The finding is associative; no CAG-170 probiotic or validated clinical test exists, and anyone selling one now is ahead of the science.

FAQ

Can I test my CAG-170 levels or take a CAG-170 probiotic? No. Commercial stool tests report cultured-species panels and generally do not quantify CAG-170, and since only one isolate has ever been grown, no live CAG-170 product exists. The realistic lever today is diet: a varied, fiber-rich eating pattern that supports overall microbial diversity — which is what CAG-170's health association most likely reflects.


This article is for educational purposes only and isn't medical advice. If you have severe or persistent symptoms, work with a qualified clinician.

Citations

  1. da Silva AC et al. Meta-analysis of the uncultured gut microbiome across 11,115 global metagenomes reveals a candidate signature of health. Cell Host & Microbe 2026 — DOI: 10.1016/j.chom.2026.01.013 — PMID: 41666920
  2. Almeida A et al. A new genomic blueprint of the human gut microbiota. Nature 2019 — DOI: 10.1038/s41586-019-0965-1 — PMID: 30745586
  3. Almeida A et al. A unified catalog of 204,938 reference genomes from the human gut microbiome. Nature Biotechnology 2021 — DOI: 10.1038/s41587-020-0603-3 — PMID: 32690973
  4. Browne HP et al. Culturing of 'unculturable' human microbiota reveals novel taxa and extensive sporulation. Nature 2016 — DOI: 10.1038/nature17645 — PMID: 27144353
  5. Magnúsdóttir S et al. Systematic genome assessment of B-vitamin biosynthesis suggests co-operation among gut microbes. Frontiers in Genetics 2015 — DOI: 10.3389/fgene.2015.00148 — PMID: 25941533
  6. Turnbaugh PJ et al. The Human Microbiome Project. Nature 2006 — DOI: 10.1038/nature06244
  7. Kenny DJ et al. Cholesterol metabolism by uncultured human gut bacteria influences host cholesterol level. Cell Host & Microbe 2020 — DOI: 10.1016/j.chom.2020.05.013 — PMID: 32544460
  8. de Cena JA et al. Low-abundant microorganisms: the human microbiome's dark matter, a scoping review. Frontiers in Cellular and Infection Microbiology 2021 — DOI: 10.3389/fcimb.2021.689197 — PMID: 34136418
  9. van Nood E et al. Duodenal infusion of donor feces for recurrent Clostridium difficile. New England Journal of Medicine 2013 — DOI: 10.1056/NEJMoa1205037 — PMID: 23323867
  10. Srivastava S et al. Randomized double-blind, placebo-controlled trial of Bifidobacterium longum CECT 7347 (ES1) in diarrhea-predominant IBS. Gut Microbes 2024 — DOI: 10.1080/19490976.2024.2338322 — PMID: 38630015