The AtlasAmino acidsWritten in a stop codon
PyrrolysinePyl · O
The twenty-second amino acid, which no human cell makes or uses, and which matters to us through the microbes that do.
Molecule · Pyl · 18 heavy atoms
Pyrrolysine
C12H21N3O3255.32 g/mol
The twenty-second amino acid, which no human cell makes or uses, and which matters to us through the microbes that do.
Built fromthe charted ones open their own entry
- Codes
- Pyl · O
- Formula
- C12H21N3O3
- Molar mass
- 255.32 g/mol
- Systematic name
- (2S)-2-amino-6-{[(2R,3R)-3-methyl-3,4-dihydro-2H-pyrrole-2-carbonyl]amino}hexanoic acid
- Side chain
- Lysine’s four-carbon chain, ending in an amide to a methylated pyrroline ring.
- Class
- Written in a stop codon
- In the diet
- Not used by humans
- Carbon skeleton
- Not applicable
- Codons
UAG
Formula and mass computed from the structure.
In brief
What it is
Lysine with a ring bolted onto the end of its side chain: a methylated pyrroline carboxylate, joined by an amide bond to lysine’s ε-amino group 1. It is the twenty-second amino acid of the genetic code, written as UAG 2,3.
Why it matters
Archaea that make methane from methylamines cannot do it without pyrrolysine 1. Some of them live in the human gut, where they consume trimethylamine, the compound the liver turns into TMAO 4,5.
Where it runs short
There is no pyrrolysine deficiency in a person: the trait is confined to a few microbes, and in human genes UAG is only a stop 6.
Where it turns
Nor is there an excess. What varies between people is whether their gut carries the archaea that use it; the Methanomassiliicoccales turned up in microbiome data from eight countries, and are associated with older age 4,7.
An amino acid humans do not use, which matters through the microbes that do.
The molecule
Pyrrolysine is two amino acids’ worth of atoms in one: lysine whose side-chain amino group is joined, through an amide bond, to a five-membered ring, a methylated pyrroline carboxylate 1. At 255 daltons it is the largest of the twenty-two genetically encoded amino acids.
It was discovered by looking. In 2002 the structure of a methanogen’s monomethylamine methyltransferase, at 1.55 Å, showed electron density at the position of an in-frame UAG codon that matched none of the twenty-one known amino acids 3. In the same issue of Science, a second group showed that a tRNA with the anticodon CUA, which pairs with UAG, and a synthetase to charge it were encoded beside the methyltransferase genes 2.
StereochemistryThree stereocentres. The lysine’s α-carbon is S, as in every L-amino acid but cysteine and selenocysteine; the two on the ring are both R, as the crystal structure that found it showed 3.
| 1st ↓ 2nd → | U | C | A | G | 3rd |
|---|---|---|---|---|---|
| U | UUUPhe | UCUSer | UAUTyr | UGUCys | U |
| UUCPhe | UCCSer | UACTyr | UGCCys | C | |
| UUALeu | UCASer | UAAStop | UGAStop | A | |
| UUGLeu | UCGSer | UAGPyl | UGGTrp | G | |
| C | CUULeu | CCUPro | CAUHis | CGUArg | U |
| CUCLeu | CCCPro | CACHis | CGCArg | C | |
| CUALeu | CCAPro | CAAGln | CGAArg | A | |
| CUGLeu | CCGPro | CAGGln | CGGArg | G | |
| A | AUUIle | ACUThr | AAUAsn | AGUSer | U |
| AUCIle | ACCThr | AACAsn | AGCSer | C | |
| AUAIle | ACAThr | AAALys | AGAArg | A | |
| AUGMet | ACGThr | AAGLys | AGGArg | G | |
| G | GUUVal | GCUAla | GAUAsp | GGUGly | U |
| GUCVal | GCCAla | GACAsp | GGCGly | C | |
| GUAVal | GCAAla | GAAGlu | GGAGly | A | |
| GUGVal | GCGAla | GAGGlu | GGGGly | G |
| Pyrrolysine | Selenocysteine | |
|---|---|---|
| Codon | UAG | UGA |
| How it is made | As a free amino acid, from two lysines, then loaded onto its tRNA 1,8 | On its tRNA: serine is loaded first and converted in place 9 |
| What the message needs | A nearby signal helps but is not required 10,11 | A SECIS hairpin; without it UGA is read as stop 11,12 |
| Who uses it | Some methane-making archaea and a few bacteria; not humans 2,6 | All three domains of life, humans included 13 |
| How many proteins | More than 1,800 known in archaea 14 | 25 in humans 15 |
| What it does there | Takes the methyl group off methylamines, the first step in making methane from them 1 | Redox chemistry: peroxides, thioredoxin, thyroid hormone 13 |
Only one organism is known to use both 6.
Where it comes from
No human cell makes or uses pyrrolysine. The trait is confined to a few microbes, mostly methane-making archaea, with related genes in a Gram-positive bacterium 2,6.
Those organisms make it from lysine alone, by three enzymes, PylB, PylC and PylD 1. A dedicated synthetase, PylS, then attaches the finished amino acid to its tRNA; it activates pyrrolysine with ATP and does not accept lysine 8,16. Moving five genes, pylTSBCD, into the bacterium Escherichia coli is enough to make it build pyrrolysine and put it into protein 1.
Making it in a flask has got easier, as a by-product of general method development. A 2025 copper/ruthenium relay catalysis merging hydrogen-borrowing with asymmetric Michael addition between inert allylic alcohols and a ketoimine ester builds α-amino δ-hydroxy acids with two adjacent tertiary stereocentres in good yield, high diastereoselectivity and excellent enantioselectivity — and among its demonstrations is a streamlined formal synthesis of L-pyrrolysine 17. The three-stereocentre problem that made the first chemical synthesis an achievement is now a worked example in somebody else's methods paper.
Not used by humansNot a nutrient. Humans have none of the machinery to make it or build it into protein; the organisms that use it make it from lysine 1,6.
How much
None. Humans do not use it 6.
Where it is in food
- Not a dietary amino acid: the organisms that use it make their own from lysine 1.
In the bottle · laboratory onlyNot sold as a supplement. Chemists synthesized it in 2004 to prove that its synthetase loads it directly 8,16, and laboratories now use the pyrrolysine machinery to build proteins carrying new amino acids 18.
What the body does with it
Pyrrolysine is necessary for every known pathway by which methane is made from methylamines. Three methyltransferases start those pathways, one for each of the methylamines, and each is encoded with pyrrolysine at an in-frame UAG 1. That 'every known pathway' has now acquired its first exception. Proteomics of a quaternary-amine-utilizing methanogen, strain Q3c, grown on tetramethylammonium or choline rather than trimethylamine, identified the methyltransferase MtqB and its corrinoid partner MtqC as the highly produced pair — and MtqB turned out to be a NON-pyrrolysine homolog of the monomethylamine methyltransferase MtmB, with its activity on both substrates confirmed by enzymatic assay and co-docking. The authors call it the first function ever demonstrated for a non-pyrrolysine MtmB 19. The methylamine methyltransferase family is not uniformly pyrrolysine-dependent, and the quaternary amines are where the exception lives.
For two decades it looked like a specialty of those enzymes: a 2005 survey of genomes found conserved pyrrolysine in only four protein families, chiefly the methylamine methyltransferases 6. In 2025, proteomics showed that some archaea read every UAG as pyrrolysine, an alternative genetic code with 62 sense codons for 21 amino acids, and found more than 1,800 archaeal proteins containing it 14.
That is where it touches human health. Trimethylamine is made only by gut microbes, from choline, carnitine, lecithin and TMAO in food, and the liver oxidizes it to TMAO, which is associated with cardiovascular disease 5,7. A group of gut archaea, the Methanomassiliicoccales, use methylamines to make methane; one of them, Methanomassiliicoccus luminyensis, first isolated from human faeces, depleted trimethylamine in culture by reducing it with hydrogen 5,20. The association that motivates all of this has itself thinned. Pooling nine prospective cohort studies and 267,569 participants, elevated circulating choline (HR 1.33, 95% CI 1.07 to 1.66) and phosphatidylcholine (HR 1.25, 95% CI 1.16 to 1.34) were each associated with incident heart failure, and in established heart failure elevated betaine tracked a composite of adverse outcomes (HR 1.15, 95% CI 1.02 to 1.30) — while the analysis reports no pooled effect estimate for TMAO itself 21. The precursors and the neighbors carry the signal in the largest current pooled analysis of this pathway; the metabolite the story is named after does not get a number.
The claim is now measurable in a working reactor rather than only in genomes. Across 220 days of continuous anaerobic digestion of protein-rich food waste, a semi-thermophilic reactor at 41 to 49 °C beat both mesophilic and thermophilic digestion — 768 ± 109 mL of biogas per gram of volatile solids per day against 710 ± 103 and 520 ± 138 — with the lowest volatile fatty acid accumulation and the highest alkalinity. Competitive isotope tracing showed 81.83% of the methane in that reactor came from methylated intermediates, and multi-omics tied the routing shift to active methylotrophic methanogenesis, PYRROLYSINE-DEPENDENT methyl transfer, and ammonia-tolerant protein-utilization networks 22. An amino acid that exists in a handful of microbes is a design parameter for municipal waste treatment.
In three sentences each
Made from two lysines
PylB rearranges one lysine into 3-methylornithine, PylC joins it to a second lysine, and PylD oxidizes the result into pyrrolysine 1.
Loaded as a finished amino acid
Unlike selenocysteine, pyrrolysine is made free and then attached to its own UAG-reading tRNA by a dedicated synthetase, which ignores lysine 8,11,16.
A tool for writing new amino acids
Pyrrolysyl-tRNA synthetase is unusually tolerant of what it loads, and engineered versions have put more than 100 non-canonical amino acids into proteins at UAG codons 18.
How it is made, moved and broken down
Pyrrolysine’s pathway is short and entirely its own: lysine in, pyrrolysine out, onto its tRNA, into a methyltransferase 1,8.
- Lysine
- Lysine mutase (radical SAM)pylB · SAM
- 3-Methylornithine
- LigasepylC
- 3-Methylornithyl-lysine
- OxidasepylD
- Pyrrolysine
- Pyrrolysine + tRNA(Pyl)
- Pyrrolysyl-tRNA synthetasepylS · ATP
- Pyrrolysyl-tRNA(Pyl)
- The ribosomepylT · PYLIS helps
- Pyrrolysine at UAG
Where it matters most
- The human gut
- Home to the Methanomassiliicoccales, found in microbiome data from eight countries and associated with older age 4,7. Across three independent human cohorts, methanogenic archaea were reproducibly enriched in people WITHOUT obesity, while bacteriophages of the class Caudoviricetes were enriched in obesity 23.
- Other guts, soil and sediment
- The same order lives outside the human gut, and the nearest relatives of the first human isolate came from the digestive tracts of a cockroach, a chicken and other mammals 7,20.
- The laboratory
- The pyrrolysine synthetase and its tRNA are a standard tool for putting non-canonical amino acids into proteins 18.
When it goes wrong
Inherited
Trimethylaminuria (fish-odor syndrome)
FMO3 · hereditary
A hereditary defect in flavin-containing monooxygenase 3 leaves trimethylamine unoxidized, and bacterial trimethylamine from food is believed to contribute to the odor 5. Archaea that consume trimethylamine with pyrrolysine enzymes have been proposed as a treatment, so-called archaebiotics; it has been shown in culture, not in patients 5. Three other approaches are further along than the archaea, and none has reached a patient. A preparation of tyndallized probiotics and their postbiotics lowered trimethylamine in human faecal slurry and in mice without killing the producing bacteria 24. A growth-based phenotypic screen plus medicinal chemistry produced compounds structurally unlike choline and unlike existing choline trimethylamine-lyase inhibitors that lowered serum TMAO in gnotobiotic mice without visibly disturbing the microbiome 25 — stopping the bacteria making trimethylamine rather than adding archaea to consume it. And a pharmacokinetic model validated against clinical choline and engineered-probiotic data PREDICTS that an engineered probiotic given three to four hours before meals would cut systemic trimethylamine by 95%, matching healthy hepatic clearance, with enzyme expression the dominant design parameter 26. That last figure is a prediction from a model, not a measurement from a trial, and it is written here that way because it will not stay that way elsewhere.
Association
TMAO and cardiovascular risk
Trimethylamine from the gut becomes TMAO in the liver, and TMAO is associated with atherosclerosis 5,7. Methanomassiliicoccales abundance correlated with lower faecal trimethylamine in an elderly cohort; whether these archaea lower TMAO or risk in people is untested 4. The nearest thing to human data points at a different organism. In 25 hemodialysis patients with chronic kidney disease, plasma TMAO was high against reference values and the most abundant gut archaea were Methanobrevibacter, with a significant NEGATIVE correlation between TMAO and Methanobrevibacter abundance 27. The authors read it as support for the archaebiotic hypothesis; the correlated genus is not the pyrrolysine-using Methanomassiliicoccales the hypothesis was built on, and 25 patients with a correlation is not a trial.
Association
Colorectal cancer
In 2,101 stool metagenomes from eleven cohorts, the order Methanomassiliicoccales was depleted in colorectal cancer while Methanobrevibacter smithii was enriched 28. An association, not a cause, and not a test.
How it is foundResearch only 28.
Association
Multiple sclerosis
Earlier small studies linked gut methanogenic archaea to multiple sclerosis. A harmonized re-analysis of International Multiple Sclerosis Microbiome Study metagenomes — 115 patients against 115 healthy HOUSEHOLD controls across Buenos Aires, Edinburgh, New York and San Francisco — found no significant difference in overall archaeome profiles. Treated patients did carry more of two methanogen taxa than untreated patients, and one of those correlated with lower disease severity scores among the treated. The authors conclude gut methanogens are not directly associated with onset or progression, and may instead reflect microbiome health during treatment 29.
How it is foundResearch only 29.
How it is measured
Pyrrolysine is identified in proteins by crystallography and mass spectrometry 3,14. In people, what can be measured is the archaea that make it, and only in research.
- Stool metagenomicsSequencing the DNA in stool can find the Methanomassiliicoccales and their genes for using trimethylamine 4.Research only.In an elderly cohort, their abundance went with lower faecal trimethylamine 4.Shows who is present, not what they are doing at the time.
- Breath methaneMethane in the breath is how gut methanogens were first noticed 7.It reports all methanogens together. The best-known methane producers of the gut, Methanobrevibacter smithii and Methanosphaera stadtmanae, belong to a different order from the pyrrolysine-using Methanomassiliicoccales 7.
Food, supplements and the evidence
Establishedreplicated in people, for a named outcome
- Nothing in this column.
Uncertainsmall, short, mixed, surrogate or preclinical
- Archaebiotics, gut archaea given to consume trimethylamine and so lower TMAO or treat trimethylaminuria: proposed in 2013, supported in culture, untested in people 4,5. The one human correlation that exists is in 25 hemodialysis patients and is with Methanobrevibacter rather than with the pyrrolysine-using order 27; meanwhile the better-funded approach is to inhibit the bacterial enzyme that makes trimethylamine in the first place, which is at the gnotobiotic-mouse stage 25.
- Whether the apparent decline of gut methanogens with industrialization has metabolic consequences 30.
- Postbiotics sold for trimethylamine, where the evidence is a slurry and two mouse strains. A preparation of tyndallized probiotics and their postbiotics, fermented with garlic extract and senna leaf, reduced trimethylamine release from human faecal slurry and lowered TMAO in wild-type mice and trimethylamine in FMO3-knockout mice, without killing the trimethylamine-producing genera — the authors position it as potential support rather than proven therapy, and note that trimethylaminuria still has no definitive treatment, only controlled diet, antibiotics and acidic soaps 24. The work was done by a pharmaceutical company's own research group, and no patient has taken it in a trial.
Sold asthe claim on the label, against the evidence
- Nothing in this column.
What is strange about it
Some archaea have their own genetic code in which UAG never means stop. It was the first alternative genetic code found in archaea, and it appears to have arisen independently several times 14; one 2026 analysis argues instead that it may be ancestral 31. In the best-studied pyrrolysine user the question is not which meaning UAG has, but that it has both. In the model methanogen Methanosarcina, UAG carries dual meaning as stop AND pyrrolysine, and no mechanism for conditionally suppressing it had ever been identified. The data suggest that expression of the pyrrolysine biosynthesis and incorporation genes is tuned to the cell's demand for pyrrolysine, which the authors propose is how these archaea live with ambiguous stop decoding and adjust it to environmental cues 32. The ambiguity is not a problem the organism solved; it is a dial it keeps.
Its synthetase is so undemanding about what it loads that it became one of the main tools of synthetic biology, used to write more than a hundred amino acids that nature never made into proteins 18. The tool has moved a long way since. Machine learning applied to the synthetase's tRNA-binding domain — a statistical model over pairwise combinations of 12 single mutations, then deep learning to find more sites — produced variants with 11-fold and then 30.8-fold increases in stop-codon suppression efficiency and up to 7.8-fold better catalytic efficiency, and transplanting the mutations into seven derived synthetases improved yields for six classes of non-canonical amino acid 33. Directed evolution by in vivo hypermutation, across eight campaigns covering 13 non-canonical amino acids, produced systems that read a single amber codon about as efficiently as natural translation reads a sense codon, plus one synthetase that regulated its own expression to sharpen its dependence on the added amino acid 34. And the prevailing design rule — start from a hyperstable scaffold so it survives diversifying mutations — has been inverted: a cold-adapted synthetase from Methanococcoides burtonii outperformed established mesophilic and thermophilic variants at single and multiple sites, held its efficiency at low concentrations of the added amino acid, and did better in absolute terms at lower growth temperatures 35. How the enzyme actually works is also clearer: electron paramagnetic resonance in vitro and IN CELLS, with a paramagnetic non-canonical amino acid as the probe, found secondary amino acid binding sites at the interface between the two monomers, far from the catalytic site, supporting a model in which the two halves of the dimer act in alternation 36.
Industrialized guts may be losing their methanogens along with the rest of their microbial diversity; whether that matters for human metabolism is an open question 30.
The orthogonality that made the pyrrolysine system a laboratory standard is not perfect, and the people building the replacement say so plainly. Reassigning a stop codon, they write, “is not completely orthogonal to translation termination for cellular transcripts” — so they built an alternative that does not touch stop codons at all: pseudouridine codons (ΨGA, ΨAA, ΨAG) installed on chosen messenger RNAs by a guide RNA, read by an engineered decoder tRNA and synthetase, incorporating non-canonical amino acids in mammalian cells. The ΨGA system showed higher translatome-wide and proteomic specificity than genetic code expansion, the three pseudouridine codon pairs were mutually orthogonal, and the two approaches work together for dual encoding 37. The tool built on an archaeal quirk now has a competitor that works around the quirk.
Where it connects
In the Atlas
Topics on the map
On the map
A star in The rest of the amino acids, one of 6. The twenty-second, and not ours. Some methane-making archaea read UAG as an amino acid built from two lysines; in our cells that codon is still stop.
Sources
37 sources, numbered as they are cited. Every one was checked against PubMed or its publisher before it was cited here; the note under each says what it shows and what it does not.
- 1Gaston MA, Zhang L, Green-Church KB, Krzycki JA. The complete biosynthesis of the genetically encoded amino acid pyrrolysine from lysine.doi:10.1038/nature09918 · PMID 21455182
PylB, PylC and PylD make pyrrolysine from two lysines; required for methane from methylamines.
- 2Srinivasan G, James CM, Krzycki JA. Pyrrolysine encoded by UAG in Archaea: charging of a UAG-decoding specialized tRNA.doi:10.1126/science.1069588 · PMID 12029131
The pylT tRNA with a CUA anticodon and the pylS synthetase; homologs in a Gram-positive bacterium.
- 3Hao B, Gong W, Ferguson TK, et al. A new UAG-encoded residue in the structure of a methanogen methyltransferase.doi:10.1126/science.1069556 · PMID 12029132
The 1.55 Å structure of MtmB that revealed pyrrolysine.
- 4Borrel G, McCann A, Deane J, et al. Genomics and metagenomics of trimethylamine-utilizing Archaea in the human gut microbiome.doi:10.1038/ismej.2017.72 · PMID 28585938
Methanomassiliicoccales in data from eight countries; abundance against faecal TMA in the ELDERMET cohort.
- 5Brugère JF, Borrel G, Gaci N, et al. Archaebiotics: proposed therapeutic use of archaea to prevent trimethylaminuria and cardiovascular disease.doi:10.4161/gmic.26749 · PMID 24247281
Proposal; M. luminyensis depleted TMA in culture. Not tested in people.
- 6Zhang Y, Baranov PV, Atkins JF, Gladyshev VN. Pyrrolysine and selenocysteine use dissimilar decoding strategies.doi:10.1074/jbc.M501458200 · PMID 15788401
The pyrrolysine trait in a few microbes; one organism with both; four protein families.
- 7Gaci N, Borrel G, Tottey W, et al. Archaea and the human gut: new beginning of an old story.doi:10.3748/wjg.v20.i43.16062 · PMID 25473158
Review of gut archaea, the Methanomassiliicoccales, trimethylamine and TMAO.
- 8Blight SK, Larue RC, Mahapatra A, et al. Direct charging of tRNA(CUA) with pyrrolysine in vitro and in vivo.doi:10.1038/nature02895 · PMID 15329732
Synthetic pyrrolysine attached as a free molecule by PylS, with ATP.
- 9Xu XM, Carlson BA, Mix H, et al. Biosynthesis of selenocysteine on its tRNA in eukaryotes.doi:10.1371/journal.pbio.0050004 · PMID 17194211
Selenocysteine is made on its tRNA from serine.
- 10Longstaff DG, Blight SK, Zhang L, et al. In vivo contextual requirements for UAG translation as pyrrolysine.doi:10.1111/j.1365-2958.2006.05500.x · PMID 17140411
A downstream PYLIS enhances UAG translation; a UAG with no evolved context was still read at 20%.
- 11Peiter N, Rother M, Krzycki JA. Recent developments in (archaeal) pyrrolysine and selenocysteine specification and metabolism.doi:10.1101/cshperspect.a041979 · PMID 41997725
Review contrasting how the two recoded amino acids are made and read.
- 12Chavatte L, Lange L, Schweizer U, et al. Understanding the role of tRNA modifications in UGA recoding as selenocysteine in eukaryotes.doi:10.1016/j.jmb.2025.169017 · PMID 39988117
SECIS and the machinery of UGA recoding.
- 13Labunskyy VM, Hatfield DL, Gladyshev VN. Selenoproteins: molecular pathways and physiological roles.doi:10.1152/physrev.00039.2013 · PMID 24987004
Selenoproteins in all three domains of life; mostly oxidoreductases.
- 14Kivenson V, Peters SL, Borrel G, et al. An archaeal genetic code with all TAG codons as pyrrolysine.doi:10.1126/science.adu2404 · PMID 41264687
The Pyl code: 62 sense codons, 21 amino acids; more than 1,800 pyrrolysine proteins.
- 15Kryukov GV, Castellano S, Novoselov SV, et al. Characterization of mammalian selenoproteomes.doi:10.1126/science.1083516 · PMID 12775843
Twenty-five human selenoproteins.
- 16Polycarpo C, Ambrogelly A, Bérubé A, et al. An aminoacyl-tRNA synthetase that specifically activates pyrrolysine.doi:10.1073/pnas.0405362101 · PMID 15314242
PylS charges pyrrolysine and not lysine; named pyrrolysyl-tRNA synthetase.
- 17Xiong Q, Chen BB, Dong XQ, Wang CJ. Asymmetric access to δ-hydroxy α-amino acids bearing two adjacent stereocenters from inert allylic alcohols via Cu/Ru relay catalysis.doi:10.1021/jacs.5c07896 · PMID 40690401
A formal synthesis of L-pyrrolysine falls out of a general method for making two adjacent stereocentres.
- 18Wan W, Tharp JM, Liu WR. Pyrrolysyl-tRNA synthetase: an ordinary enzyme but an outstanding genetic code expansion tool.doi:10.1016/j.bbapap.2014.03.002 · PMID 24631543
More than 100 non-canonical amino acids incorporated with engineered PylRS.
- 19Kashyap J, Das D, Ticak T, Creighbaum AJ, Khatri R, Ferguson DJ. Methanogenesis from tetramethylammonium and choline in Q3c requires a nonpyrrolysine monomethylamine methyltransferase homolog.doi:10.3389/fmicb.2026.1739651 · PMID 41939719
The genus is missing from the indexed title; the organism is the quaternary-amine-utilizing methanogen strain Q3c. The first function ever shown for a methylamine methyltransferase that carries no pyrrolysine.
- 20Dridi B, Fardeau ML, Ollivier B, et al. Methanomassiliicoccus luminyensis gen. nov., sp. nov., a methanogenic archaeon isolated from human faeces.doi:10.1099/ijs.0.033712-0 · PMID 22859731
The first human isolate of the order; relatives from cockroach, chicken and mammal guts.
- 21Shokravi A, Luo Y, Rabkin SW. Unraveling the choline pathway in heart failure risk and outcomes: a systematic review and meta-analysis.doi:10.1016/j.bbadis.2025.168103 · PMID 41202941
Nine prospective cohorts, 267,569 participants. Note which metabolite does NOT get a pooled estimate.
- 22Zhao C, Mo J, Xu X, et al. Methylotrophic rewiring enhances methane recovery and stability in semi-thermophilic anaerobic digestion of protein-rich food waste.doi:10.1016/j.watres.2026.126367 · PMID 42365724
220 days of continuous operation, with isotope tracing to say where the methane came from. A working reactor rather than a genome.
- 23Gallardo-Nuell L, Rosell-Díaz M, Garre-Olmo J, et al. Ecological restructuring of the nonbacterial fecal microbiome in obesity across human cohorts.doi:10.1080/19490976.2026.2701446 · PMID 42458961
Three independent human cohorts, covariate-adjusted, with the archaea looked at rather than filtered out.
- 24Giannini G, Soldi S, Elli M, et al. A mixture of postbiotics/tyndallized probiotics reduces trimethylamine (TMA) in trimethylaminuria models: evidence from in vitro and in vivo studies.doi:10.3389/fphar.2025.1591825 · PMID 41132530
Faecal slurry and two mouse strains, from a pharmaceutical company's own research and development group. No patients.
- 25Woo AYM, Sandoval-Espinola WJ, Bollenbach M, et al. Phenotypic high-throughput screening identifies modulators of gut microbial choline metabolism.doi:10.1128/mbio.01172-25 · PMID 41728982
The competing approach: stop the bacteria making trimethylamine, rather than add archaea to eat it. Gnotobiotic mice.
- 26DeVito VL, Karamched BR. Signed, sealed, delivered: a generalizable model for living biotherapeutic dosing and metabolism.doi:10.1038/s41540-026-00685-4 · PMID 41872195
The 95% figure is a model's prediction, not a trial's result. It is recorded here as such because it will be repeated as if it were measured.
- 27Kemp JA, Schultz J, Modolon F, Ribeiro-Alves M, Rosado AS, Mafra D. Is there a correlation between TMAO plasma levels and archaea in the gut of patients undergoing hemodialysis?doi:10.1007/s11255-024-04273-5 · PMID 39562414
25 patients, and the correlated genus is not the pyrrolysine-using one the hypothesis was built on.
- 28Li T, Coker OO, Sun Y, et al. Multi-cohort analysis reveals altered archaea in colorectal cancer fecal samples across populations.doi:10.1053/j.gastro.2024.10.023 · PMID 39490771
2,101 metagenomes; Methanomassiliicoccales depleted in colorectal cancer.
- 29Woh PY, Chen Y, Kumpitsch C, Mohammadzadeh R, Schmidt L, Moissl-Eichinger C. Reevaluation of the gastrointestinal methanogenic archaeome in multiple sclerosis and its association with treatment.doi:10.1128/spectrum.02183-24 · PMID 39998261
115 patients against 115 household controls in four cities, with one method applied to all of it. A null result that replaces smaller positive ones.
- 30Subrahmanian A, Patel A, Veerus L, et al. Methanogens: vital but threatened members of the human microbiome?doi:10.1016/j.tim.2026.08.008 · PMID 42692861
Review: evidence that methanogens may be declining with industrialization.
- 31Di Giulio M. The archaeal genetic code with all UAG codons as pyrrolysine might be an ancestral and not derived trait: the non-monophyletic origin of the genetic code.doi:10.1016/j.biosystems.2026.105787 · PMID 41985704
Theoretical analysis; a minority view.
- 32Shalvarjian KE, Chadwick GL, Pérez PI, Woods PH, Orphan VJ, Nayak DD. Methanogenic archaea encoding pyrrolysine maintain ambiguous amber codon usage.doi:10.1073/pnas.2517473122 · PMID 41196353
In the best-studied pyrrolysine user, UAG means both things at once, and the ambiguity is tuned rather than resolved.
- 33Zhang Q, Jiang L, Niu Y, et al. Machine learning-guided evolution of pyrrolysyl-tRNA synthetase for improved incorporation efficiency of diverse noncanonical amino acids.doi:10.1038/s41467-025-61952-2 · PMID 40681550
The limiting factor was the enzyme's activity all along, and the fold-changes say how much was being left on the table.
- 34Furuhata Y, Rix G, Van Deventer JA, Liu CC. Directed evolution of aminoacyl-tRNA synthetases through in vivo hypermutation.doi:10.1038/s41467-025-60120-w · PMID 40413191
Eight evolution campaigns, thirteen non-canonical amino acids, and a synthetase that regulated its own expression.
- 35Koch NG, Goettig P, Nash MA, Rappsilber J, Budisa N. Cold orthogonal translation: a psychrophilic pyrrolysyl-tRNA synthetase boosts genetic code expansion in E. coli.doi:10.1002/advs.202513600 · PMID 41984518
The opposite of the usual design rule: start from a flexible cold enzyme rather than a rigid hot one.
- 36Dröden J, Globisch C, Landwehr E, Braun TS, Summerer D, Peter C, Drescher M. Unraveling synthetase's mode of action: the pyrrolysyl-tRNA synthetase dimer uses secondary binding sites in the cell.doi:10.1002/anie.202514065 · PMID 41873850
EPR spectroscopy in cells, with a paramagnetic non-canonical amino acid as the probe. The first in-cell evidence for the secondary sites.
- 37Liu J, Yan X, Wu H, et al. RNA codon expansion via programmable pseudouridine editing and decoding.doi:10.1038/s41586-025-09165-x · PMID 40562925
Written by people building the alternative, which is why their statement about the amber system's limits is worth quoting.
This is education, not medical advice. Nothing on this page is written with knowledge of your history, your medications or your risks, and nothing here is a dose. Do not start or stop any treatment on the basis of it — talk to your own physician. Read the full medical disclaimer.