Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes

Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes View ORCID Profile Juan Manuel Medina-Méndez , Paula Iruzubieta , View ORCID Profile Raúl Fernández-López , Javier Crespo , View ORCID Profile Fernando de la Cruz doi: https://doi.org/10.1101/2025.05.03.652020 Juan Manuel Medina-Méndez 1 Instituto de Biomedicina y Biotecnología de Cantabria (IBBTEC), Universidad de Cantabria-CSIC , Santander, Spain 2 Servicio de Gastroenterología y Hepatología. Grupo de Investigación Clínica y Traslacional en Enfermedades Digestivas. Instituto de Investigación Valdecilla (IDIVAL). Hospital Universitario Marqués de Valdecilla , Santander, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Juan Manuel Medina-Méndez Paula Iruzubieta 2 Servicio de Gastroenterología y Hepatología. Grupo de Investigación Clínica y Traslacional en Enfermedades Digestivas. Instituto de Investigación Valdecilla (IDIVAL). Hospital Universitario Marqués de Valdecilla , Santander, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Raúl Fernández-López 1 Instituto de Biomedicina y Biotecnología de Cantabria (IBBTEC), Universidad de Cantabria-CSIC , Santander, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Raúl Fernández-López Javier Crespo 2 Servicio de Gastroenterología y Hepatología. Grupo de Investigación Clínica y Traslacional en Enfermedades Digestivas. Instituto de Investigación Valdecilla (IDIVAL). Hospital Universitario Marqués de Valdecilla , Santander, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site Fernando de la Cruz 1 Instituto de Biomedicina y Biotecnología de Cantabria (IBBTEC), Universidad de Cantabria-CSIC , Santander, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Fernando de la Cruz For correspondence: delacruz{at}unican.es Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Metabolic dysfunction-associated steatotic liver disease (MASLD) is a multifactorial condition influenced by the gut microbiome (GM). While previous studies have reported inconsistent associations between MASLD and key microbial clades using low-resolution 16S rRNA profiling, we employed high-resolution metagenomic sequencing and multi-marker taxonomic classification across three independent cohorts to identify robust microbial and functional signatures of MASLD. We consistently detected a depletion of Agathobacter rectalis , a known butyrate producer, in MASLD patients. Functionally, MASLD was characterized by a depletion of genes involved in butyrate and methane biosynthesis-particularly within the crotonyl-butyryl-CoA axis-alongside an enrichment of genes driving the production of endogenous alcohols such as ethanol and 1-propanol. Genes encoding these fermentative pathways, often organized in operons like pdu and tor , were more abundant in MASLD samples, indicating a potential shift toward alcohol-producing metabolism. These geno-metabolic changes were accompanied by a broader displacement of beneficial taxa and an increase in accessory gene content across the GM, underscoring the limitations of taxonomy-based disease associations. Many of the differentially abundant genes were also found on plasmids, suggesting that horizontal gene transfer contributes to strain-level metabolic variability relevant to MASLD progression. Our findings support a model in which GM-driven metabolic shifts-rather than taxonomic changes alone-play a central role in MASLD pathogenesis, highlighting the importance of functional and mobile genetic element (MGEs) profiling for uncovering mechanistic links between the microbiome and liver disease. 1. Introduction Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common liver disease worldwide, affecting approximately one quarter of the global population 1 . Clinically, MASLD is defined by the presence of steatosis (i.e., accumulation of intrahepatic triglycerides) in more than 5% of hepatocytes, in association with metabolic risk factors (particularly, obesity and type 2 diabetes) and in the absence of excessive alcohol consumption (≥30 g per day for men and ≥20 g per day for women) or other chronic liver diseases 2 . However, the term encompasses a spectrum of pathological situations, ranging from simple steatosis to metabolic dysfunction-associated steatohepatitis (MASH), which can further progress to fibrosis, cirrhosis and hepatocellular carcinoma 3 . MASLD also significantly increases the risk of cardiovascular disease, extrahepatic malignancies (e.g., colorectal or breast cancer), liver-related complications or chronic kidney disease 4 . The pathogenesis of MASLD is multifactorial, involving genetic, environmental, inflammatory and metabolic factors, and it has a bidirectional association with components of the metabolic syndrome, particularly insulin resistance 5 . The gut microbiome (GM) interacts with the liver via the gut-liver axis (GLA) through the portal vein, which transports GM-derived metabolites to the liver. This axis can be disrupted by multiple interconnected factors, affecting the progression of MASLD and fibrosis 6 . Alterations of the GM composition and/or their associated metabolic production can lead to an increased production of harmful metabolites such as pathogen-associated molecular patterns, which induce inflammatory responses in the liver when they reach it through the bloodstream, leading to liver injury 7 , 8 . The transient of these endotoxins into portal circulation is facilitated by an increased intestinal permeability caused by a disruption of tight junction proteins in the intestinal membrane 9 . Additionally, alterations of the bile acid composition and circulation affect liver metabolism and function. Since bile acids play a role in the regulation of glucose and lipid metabolism, their dysregulation contributes to gut inflammation and MASLD 10 , 11 . Several studies have attempted to establish a causal link between bacterial metabolism and MASLD 12 , 13 . However, metabolic levels can also be influenced by host metabolism, diet or environmental factors. Similarly, metataxonomic analysis have explored the relationship between GM composition and MASLD 14 , 15 . Although 16S rRNA gene sequencing provides taxonomic resolution at the genus or species level, it fails to capture strain-level metabolic variability. Bacterial genomic plasticity allows key metabolic genes, such as those potentially relevant to MASLD pathogenesis, to be encoded in their accessory genome, including MGEs, further complicating the association between taxonomy and metabolic function 16 . In this study, we used the genes encoding metabolic enzymes involved in the final steps of microbial pathways leading to the production of butyrate, short-chain alcohols (SCAs) ethanol and propanol, methane, trimethylamine (TMA) and trimethylamine N-oxide (TMAO) as proxies to determine their abundance in MASLD. To achieve this, we isolated these gene families and quantified them in over 550 metagenomes obtained from fecal samples of MASLD and healthy patients, stratified across three cohorts. We also explored taxonomic signatures to gain a comprehensive view of bacterial diversity and abundance. Finally, we assessed the presence of these genes within the human GM genomes and plasmids. This approach provided insights into the role of microbial pathways in MASLD progression and highlighted key genetic signatures in both healthy and MASLD patients. 2. Results Agathobacter rectalis is depleted in MASLD Certain bacterial clades have been positively correlated with MASLD progression, including Enterobacteriaceae 15 , 17 such as Escherichia / Shigella spp . 17 , 18 , Bacteroidaceae like Bacteroides spp. 18 , 19 , Veillonellaceae 15 or Streptococcus spp. 20 . Conversely, other clades have been reported to exhibit an inverse correlation with MASLD progression, including Ruminococcaceae 15 such as Faecalibacterium prausnitzii 18 , 21 , Lachnospiraceae like Eubacterium rectale 18 , 21 or Dorea longicatena 21 , Clostridiaceae like Clostridium spp. 17 , and even Pseudomonas spp. 20 . However, these studies often yield conflicting results. For example, a reported a positive correlation between Ruminoccocus spp. abundance and MASLD progression 19 contradicts one of the main findings in two posterior studies 18 , 21 . Similarly, Prevotella spp. was found to decrease with MASLD development 22 and fibrosis stage 19 , yet another study reported an increase in this genus in MASLD patients 20 . There are additional examples of discordant taxonomic signatures in MASLD, such as Blautia spp. or Roseburia spp. 23 . To investigate whether any bacterial genera is systematically altered in MASLD, we profiled the taxonomic composition of the GM communities across all samples from the three cohorts at both genus and species level using MetaPhlAn 24 , as detailed in Materials and Methods. This analysis identified 28 bacterial genera shared across all three cohorts ( Fig. 1A ), which accounted for 59%, 31.5% and 49% of the GM fraction in each cohort, respectively ( Fig. 1B ). Thus, nearly half of the bacterial genera identified in at least one cohort were shared across all three (28 out of 63), representing between one-third and two-thirds of the total GM abundance. Download figure Open in new tab Download figure Open in new tab Figure 1. Taxonomic profiling of GM in MASLD. (A) Number of bacterial genera common to the three cohorts. (B) Average relative abundance of bacterial genera predicted by MetaPhlAn across the three study cohorts. The blue section of the pie charts reflects the average relative abundance of bacterial genera common to the three cohorts (n=28), whereas the grey and green sections reflect, respectively, the abundances of non-common and minor genera (i.e., with an average relative abundance of 0 in at least one of the cohorts). (C) Top abundant GM genera (average absolute abundance >0.5%) (D-E) Relative abundance of bacterial genera (D) and species (E) in the metagenomes of the fecal samples, shown on a logarithmic scale. Blue and red boxplots correspond to healthy and MASLD groups, respectively. Each facet represents one cohort. Differences in abundance were evaluated using pairwise Mann-Whitney tests with Benjamini-Hochberg adjustment for multiple testing. Horizontal bars represent median values, while boxes and whiskers represent the interquartile range and the full data spread (minimum to maximum), respectively. Clades depicted in green across the x-axis have a consistent differential abundance pattern between groups over the three cohorts. Only Agathobacter rectalis , depicted in red, has a consistent and statistically significant decrease in the MASLD group across cohorts. (*) p<0.05, (**) p<0.01, (***) p<0.001, (****) p<0.0001. Non-starred genera present non-significant differences in abundance between groups. Among these shared genera, Gemmiger , Streptococcus , Mediterraneibacter , Ruminococcus and Agathobacter exhibited a consistent pattern of differential abundance between groups across all cohorts, reaching statistical significance in at least two ( Fig. 1D ). This exclusion criterion was applied to enhance the reliability of the results, ensuring that differentially abundant genera were not solely driven by heterogeneous demographic variables (e.g. geographical location, age, sex, or ethnicity) or environmental factors (e.g. diet, physical activity, comorbidities, or drug exposure), rather than being directly related to the MASLD phenotype 25 , 26 . However, this threshold was not applied to Cohort 2 because all patients in this group had MASLD, and applying it could have been overly restrictive, masking potential taxonomic differences associated with the disease. Of these five genera, only Agathobacter spp. showed statistically significant differences across the three cohorts ( Fig. 1D ). Note that Agathobacter , along with Ruminococcocus and Mediterraneibacter , belongs to the fraction of the highest abundant GM bacterial genera ( Fig. 1C ), highlighting their potential contribution to the phenotype in specific patient cohorts. Analogously, the most abundant bacterial species within the GM were analyzed for differences in abundances between groups across the three cohorts ( Fig. 1E ). Only Agathobacter rectalis and Bacteroides uniformis , two of the most abundant GM bacterial species (Supplementary Fig. 1), exhibited a consistent pattern of differential abundance across all cohorts, with statistical significance in at least two. However, only A. rectalis showed a significant difference between groups in all cohorts, identifying it as a robust taxonomic marker depleted in MASLD. Butyrate-producing genes are depleted in MASLD The identification of A. rectalis as a consistent taxonomic signature across multiple MASLD cohorts underscores its potential protective role in MASLD development. However, taxonomy alone may not fully capture the functional contributions of the GM. Certain microbial metabolites regulate GLA integrity and have been suggested to influence MASLD progression 27 – 29 . These compounds derive from dietary carbohydrates and proteins that are not fully digested in the upper gastrointestinal tract and are instead catabolized by the GM through fermentative pathways that are predominant in the anaerobic environment of the colon 30 , 31 . Once absorbed into the bloodstream, they are transported to the liver via the portal vein, where they can exert both beneficial and detrimental effects 32 , 33 . The most well-characterized of these metabolites are short chain fatty acids, produced by the colonic fermentation of complex carbohydrates present in dietary fiber that escape digestion in the small intestine. These metabolic pathways have been relatively well understood for nearly thirty years 34 and their impact on GM homeostasis has been reviewed in several articles 35 – 37 . For this reason, we interrogated the metagenomic fecal samples from the patient cohorts for metabolic signatures that could explain the MASLD phenotype. First, we focused on the genes encoding enzymes involved in butyrate formation as a proxy for the bacteria that carry them, independently of their taxa. To validate our hypothesis, we analyzed the genic abundance of the final enzymes involved in butyrate production. We selected butyrate as a target because it is crucial for maintaining GLA homeostasis by supporting the colonic epithelial barrier 38 , 39 . It also promotes lipid oxidation 40 , 41 , induces hepatic gluconeogenesis 42 , and prevents insulin resistance 43 , all of which are key factors in MASLD. In mice, both butyrate-producing probiotics 44 and sodium butyrate administration 45 – 47 have been shown to prevent steatosis and slow MASLD progression. Additionally, fecal microbiota transplantation (FMT) from healthy to MASH mice increased cecal butyrate levels and reduced steatosis 48 . Bacterial butyrate can be formed through two major terminal reactions ( Fig. 2A ). The first involves a dephosphorylation of butyryl-P that yields one ATP through an energetically favorable reaction that is catalyzed by a butyrate kinase encoded by buk. In the precursor reaction, butyryl-P is formed through a phosphorylation of butyryl-CoA catalyzed by a phosphate butyryl transferase encoded by ptb . Butyrate also accumulates after the transfer of a CoA cofactor from butyryl-CoA to acetate through butyryl-CoA:acetate CoA transferase, encoded by but . This CoA-transferase pathway also conserves the energy of the CoA bond in the newly formed CoA-moiety of the co-substrate. Butyryl-CoA is produced from crotonyl-CoA through two different enzymes: butyryl-CoA trans-2-oxidorreductase, encoded by fabV , and butyryl-CoA dehydrogenase, encoded by bcd. This pathway is further reviewed in the human GM elsewhere 49 . Download figure Open in new tab Figure 2. Abundance of butyrate-producing genes in patient cohorts. (A) Characterization of the terminal reactions involved in the formation of bacterial butyrate from crotonyl-CoA. Enzymes, EC numbers, coding genes and KOs, as well as reaction types, are indicated for each metabolic step according to the legend (lower part). (B) Abundance of butyrate-producing genes in Cohorts 2 and 3, shown on a logarithmic scale. Boxplots represent the gene abundance from the metagenomic measures of the fecal samples, and differences in abundances were evaluated using pairwise Mann-Whitney tests with Benjamini-Hochberg adjustment for multiple testing. Upper and lower facets represent abundances in Cohorts 2 and 3, respectively. Left and right facets represent abundances of precursor and terminal coding genes, respectively, as described in (A). RPKSM: reads per kilobase per genic size per million reads. (*) p<0.05, (**) p<0.01, (***) p<0.001, (****) p<0.0001, ns: non-significant. We quantified the abundance of these genes in the metagenomic fecal samples of both groups across the three MASLD cohorts, as described in Materials and Methods. Genes involved in the production of butyrate and its precursors, butyryl-P and butyryl-CoA, were significantly depleted in the GM of patients with cirrhosis MASLD compared to the healthy group, and their abundance also decreased significantly with MASLD progression ( Fig. 2B ). The precursor gene fabV was nearly an order of magnitude less abundant than ptb and bcd in both cohorts, underscoring the greater prevalence of bcd over fabV in the CoA-transferase route. Additionally, ptb did not exhibit significant differences in Cohort 2. Gene but was significantly depleted in the control group in both cohorts, whereas buk did not show differences in Cohort 2 either. Notably, the depletion of butyrate-related genes was not observed in MASLD patients from Cohort 1. SCA-producing genes are increased in MASLD Short-chain alcohols (SCAs) are a metabolic family also suggested to regulate GLA integrity, with several studies linking GM endogenous alcohol production to MASLD. One of the first reports, published over a decade ago, described elevated peripheral blood ethanol and increased fecal Escherichia spp. 16S rRNA levels in patients with MASH 50 . Similar findings were reported in a pediatric cohort, where children with MASLD exhibited higher levels of fecal ethanol, Prevotella spp. , and Gammaproteobacteria 16S rRNA 51 . A more recent study showed that postprandial peripheral blood ethanol and fecal Lactobacillaceae 16S rRNA levels increased in MASH patients after selective inhibition of host ADH. This effect was reversed with broad-spectrum antibiotic treatment 52 . In 2019, Yuan and colleagues identified a high-alcohol-producing strain of Klebsiella pneumoniae capable of causing MASLD through ethanol production 53 . Specifically, they found that FMT from a MASH patient containing this strain induced MASLD in healthy mice, whereas removing the strain before FMT prevented disease development. Furthermore, ethanol produced by this K. pneumoniae strain was shown to disrupt lipid homeostasis in hepatic cells and to cause mitochondrial dysfunction, oxidative stress, and the accumulation of reactive oxygen species 54 . These factor are critical to MASH development through their role in lipid peroxidation and inflammation 55 . Additionally, a recent computational study identified a positive correlation between the abundance of two fermentative pathways in certain families from class Clostridia and fatty-liver disease in patients. These pathways include the production of ethanol from pyruvate and the heterolactic fermentative pathway 56 . With this ample evidence, we hypothesized that genes encoding enzymes involved in the formation of endogenous SCAs may represent another metabolic mechanism contributing to MASLD pathology. We focused on ethanol and propanol, which are produced by the reduction of their respective aldehydes through ethanol and propanol dehydrogenases. These enzymes help to reduce the levels of toxic aldehydes in the GM 57 , 58 and replenish NAD + levels to maintain fermentative growth. Bacterial ethanol is formed from acetaldehyde through a fermentative reaction that can be catalyzed by 1) an NADH-dependent, bifunctional acetaldehyde-ethanol dehydrogenase encoded by adhE , 2) an NADPH-dependent alcohol dehydrogenase encoded by adhB 59 , 60 and 3) a NADPH-dependent aldehyde reductase encoded by two different genes: ahr and yahK , with the corresponding liberation of phosphorylated reducing power in form of NADP + 61 , 62 . The precursor acetaldehyde can be formed from 1) ethanolamine degradation, through a reaction catalyzed by ethanolamine ammonia-lyase which is encoded by eutB and eutC genes and liberates ammonia as a by-product 63 , 2) acetyl-CoA, through a fermentative reaction catalyzed by bifunctional AdhE and by a CoA-acetylating acetaldehyde dehydrogenase encoded by mhpF 64 , 65 and 3) acetate, through a dehydration catalyzed by aldehyde dehydrogenase B, encoded by aldB 66 ( Fig. 3A , top). Bacterial 1-propanol is formed from propionaldehyde through a fermentative reaction catalyzed by a propanol dehydrogenase encoded by two different genes: adhP (a propanol-preferring alcohol dehydrogenase) and pduQ. Propionaldehyde can be generated through two distinct reactions: 1) a CoA transference from propionyl-CoA catalyzed by a propionaldehyde dehydrogenase encoded by pduP ; and 2) a dehydration of 1,2 propanediol catalyzed by a propanediol dehydratase composed by three subunits encoded by pduC , pduD and pduE. These genes are co-located within the same operon and encode the corresponding enzymes responsible for the metabolism of 1,2-propanediol within the propanediol-utilizing (Pdu) microcompartment, a specialized bacterial structure found in certain enteric species 67 , 68 ( Fig. 3A , bottom). Download figure Open in new tab Download figure Open in new tab Figure 3. Abundance of SCA-producing genes in patient cohorts. (A) Characterization of the terminal reactions involved in the formation of bacterial ethanol (top) and propanol (bottom). Enzymes, EC numbers, coding genes and KOs, as well as reaction types, are indicated for each metabolic step according to the legend in Fig. 2A . Non-candidate genes are colored in grey. (B) Abundance of mhpF, aldB, yahK and ahr genes involved in the last two reactions from pyruvate to ethanol in the three cohorts. (C) Abundance of pduC-E, pduP, pduQ and adhP genes, involved in the last two reactions from 1,2-propanediol and propionyl-CoA to 1-propanol in the three cohorts. Boxplot layout, distribution in facets and statistical tests were performed as in Fig. 2C . We quantified the abundance of genes involved in bacterial SCA formation in the metagenomic samples of the three cohorts as described above. The genes mhpF and aldB , involved in ethanol production, and yahK and ahr , involved in acetaldehyde production, were significantly enriched in the GM of MASLD patients from Cohort 1 and Cohort 3 compared to healthy groups. Their abundance also increased with MASLD progression ( Fig. 3B ). Notably, mhpF and ahr were depleted compared to aldB and yahK respectively in all three cohorts. All genes involved in the formation of 1-propanol and propionaldehyde were also significantly enriched in MASLD patients by more than half order of magnitude, increasing with MASLD progression ( Fig. 3C , Supplementary Fig. 4A). These findings reveal a systematic increase of SCA-producing genes in MASLD. Genes involved in methane production are decreased in MASLD, whereas tor operons driving TMA accumulation are elevated Trimethylamine N-oxide (TMAO) is a metabolite derived from the oxidation of trimethylamine (TMA), which is produced by the microbial degradation of dietary compounds such as choline and L-carnitine. Circulating TMAO levels have also been linked to MASLD severity 69 and all-cause mortality in MASLD patients 70 . Higher plasmatic TMAO levels also increase cardiovascular disease risk-a condition associated with MASLD-in both mice 71 and clinical patients 72 . Therefore, we investigated whether bacterial TMAO could represent another metabolic mechanism contributing to MASLD. Methane production in the GM occurs through a two-step process: a simultaneous TMA demethylation and CoM methylation, followed by the reduction of methyl-CoM ( Fig. 4A ). TMA originates from choline degradation via the cut operon 73 and is oxidized in the liver to TMAO 74 or reduced back to TMA under anaerobic conditions via the torCAD and torZY operons. Specifically, torA and torZ encode two TMAO reductases, whose molecular mechanisms are reviewed elsewhere 75 . Download figure Open in new tab Download figure Open in new tab Figure 4. Abundance of methane and TMA-producing genes in patient cohorts. (A) Characterization of the terminal reactions involved in the formation of microbial methane. Enzymes, EC numbers, coding genes and KOs, as well as reaction types, are indicated for each metabolic step according to the legend annexed to Fig. 2A . (B) Abundance of genes encoding methyl-coM reductase I ( mcr ) and methyl-CoM reductase II ( mrt ). (C) Abundance of genes encoding CoB-CoM heterodisulfide reductase ( hdr ) and tetrahydromethanopterin S-methyltransferase ( mtr ). (D) Abundance of tor operons. Facets represent genic operons. Boxplot organization and statistical tests were conducted as in Fig. 2C . CoM methylation occurs through two well-characterized enzymatic systems: a TMA-specific (composed of mttC , mtbA and mttB ) and a methanol-specific pathway (composed of mtaC, mtaA and mtaB ) 76 . The final reduction of methyl-CoM to methane is catalyzed by methyl-CoM reductase (MCR) I, which is encoded by the mcrBDCGA operon 77 . MCR I has an alternative isozyme termed MCR II, encoded by mrtBGA operon 78 . The resulting heterodisulfide (CoB-S-S-CoM) is then reduced to CoM and CoB by a heterodisulfide reductase encoded by operon hdrABC , thereby regenerating both cofactors 79 . Additionally, methyl-CoM is replenished via tetrahydromethanopterin CoM-methyltransferase (encoded by operon mtrABCDEFGH ), which methylates CoM 80 . We measured the abundance of these genes in the samples of the three MASLD cohorts as described above. Genes encoding MCR I and II, which catalyze the reduction of methyl-CoM to methane, were significantly decreased in the GM of MASLD patients. Their abundance also declined with MASLD progression ( Fig. 4B ). Concordantly, hdr and mtr genes, responsible for replenishing CoM and transferring it to a methyl group to form the methyl-CoM precursor, were significantly decreased in MASLD and showed a progressive depletion with MASLD advancement ( Fig. 4C ). Interestingly, hdrA was two orders of magnitude more abundant than the rest of the genes from the operon. Genes mtbA, mttB, mtaA and mtaB were significantly decreased in MASLD. However, they did not exhibit differences between the initial and advanced stages of MASLD (Supplementary Fig. 2). Notably, the higher abundance of methane-producing genes observed in the healthy group was not replicated in Cohort 3. In contrast, bacterial genes involved in TMA regeneration showed the opposite trend. The torCAD and torZY operons were significantly increased in the GM of MASLD patients, and their abundance increased with MASLD progression ( Fig. 4D , Supplementary Fig. 4B). These findings suggest that methane production from TMA and methanol, which typically occurs in a healthy state, is disrupted in MASLD, revealing another potential microbial pathway involved in MASLD pathogenesis. Finally, we quantified the abundance of five universal single-copy genes (USCGs), as described in Material and Methods, to confirm that observed differences in genic abundance between sample groups were truly associated with the MASLD phenotype and not confounded by biological variability or spurious technical artifacts related to the quantification method. No significant differences were found in gene abundance between groups for any USCG in any of the three cohorts (Supplementary Fig. 1). Candidate metabolic genes are accessory in the GM Although this analysis identifies A. rectalis and multiple metabolic genes involved in the formation of butyrate, SCAs, TMAO and methane as consistent taxonomic and functional signatures associated with MASLD, these markers alone are not sufficiently robust to establish a causal link between GM composition and MASLD phenotype. GM bacterial species have pleomorphic genomes, consisting of up to thousands of genomically distinct strains that share a conserved genetic core but possess an accessory genome that is highly variable among them 81 , 82 . This accessory genome is often encoded in plasmids and other MGEs which are subject to frequent changes and can readily transfer between different bacterial strains 16 , 83 . To test this hypothesis, we assessed whether candidate metabolic genes involved in MASLD are core or accessory across some of the most abundant human GM genomes. Genome selection and genic presence were determined as described in Materials and Methods. Briefly, the presence of a target gene in a clade was defined as the percentage of strains encoding it, classifying it as core if present in >80% of congeneric or conspecific genomes, as accessory if found in 20-80%, and as highly accessory if present in <20% genomes. Most of the selected GM genera (n=24) belong to class Clostridia, specifically to orders Lachnospirales and Oscillospirales ( Fig. 5A ). Butyrate-producing genes from the butyryl-CoA pathway ( bcd and but ) were core in nearly half of these genera, while those from the butyryl-P pathway ( ptb and buk ) were core only in Flavonifractor, Clostridium and Coprococcus. Notably, butyrate-producing genes were highly accessory across several genera, including Lachnospira, Ruminococcus_B and E, Ruthenibacterium, Blautia , Fusicatenibacter , Dorea, and Mediterraneibacter , and were completely absent in Ruminococcus_C and D . In contrast, acetaldehyde and ethanol-producing genes were highly accessory across all genera except Streptococcus , an heterolactic genus that constitutively encoded aldhB in all its genomes (specifically in S. salivarius and S. thermophilus, Fig. 5B ) and exhibited moderated presence (20-80%) of yahK / ahr . The acetaldehyde-producing gene mhpF was nearly absent in the GM, except for a few genomes of Flavonifractor and Agathobacter . Propionaldehyde and propanol-producing genes from the pdu operon were core in three genera: Ruminococcus_B ( R_B. gnavus ), Anaerobutyricum ( A. hallii ) and Flavonifractor ( F. plautii ), though pduE was unexpectedly absent in Ruminococcus_B . These genes were highly accessory in Blautia ( B. wexlerae ), Streptococcus , Medierraneibacter or Agathobacter ( A. rectalis ). Notably, A. hallii and F. plautii were the only species encoding both core butyrate and propanol-producing genes. Download figure Open in new tab Download figure Open in new tab Figure 5. Presence profile of candidate metabolic genes in the human GM. (A-B) Heatmap showing the presence of candidate genes involved in MASLD across some of the most abundant clades in the human GM. The x-axis represents genes associated with butyrate and SCAs production, as well as USCGs. The y-axis lists the bacterial clades inspected. Gene presence is expressed as the percentage of (A) congeneric and (B) conspecific genomes encoding each target gene, with colors indicating classification: black for core genes (>80% of genomes), dark blue for accessory genes (20-80% genomes), light blue for highly accessory genes (<20% genomes) and white for genes completely absent in the clade. Hierarchical clustering was applied to the y-axis to group genera (top) based on gene-presence similarity. (C) Presence of candidate genes in RefSeq200 annotated plasmids. Black bars represent the number of different plasmids that encode at least one copy of each gene. Note that the nomenclature used in Fig. 5 follows the Genome Taxonomy Database standards, which redefine taxonomic ranks based on genome-wide phylogenetic analyses 84 . As a result, traditional genera such as Ruminococcus and Clostridium , which were found to be polyphyletic (i.e., composed of species that do not share a single common ancestor) are split into multiple monophyletic clades (e.g., Ruminococcus_A, Ruminococcus_B or ClostridiumQ ). Beyond their accessory nature within GM pangenomes, candidate metabolic genes were also encoded on plasmids ( Fig. 5C ). Genes involved in ethanol formation such as aldhB, yahK and ahr were among the most widely distributed, being present in hundreds of different plasmids. In contrast, butyrate-producing genes ( fabV, but, ptb, and buk ) and propanol-related genes ( pduC, pduD, pduE, pduP, and pduQ ) were detected in fewer plasmids, further supporting their variable presence across GM genomes. A notable exception was bcd , which exhibited a broader presence across both chromosomal and plasmid-encoded sequences. These findings reinforce the idea that non-essential metabolic functions, especially ethanol production, are encoded in MGEs, contributing to their accessory nature within the GM. 3. Discussion MASLD is a liver pathology influenced by multiple contributing factors, and the GM has been long suspected to play a role in its progression. Numerous studies have tried to ascribe a causality between GM composition and MASLD, but taxonomic signatures in key GM clades, such as Blautia and Roseburia spp. , have been inconsistently associated with MASLD progression 23 . These issues may be linked to the use of 16S rRNA-based profiling in metataxonomic studies, which frequently lacks resolution at species and strain level 85 . However, the heterogeneity concomitant to bacterial genomes poses another possible cause for the lack of association between disease progression and the taxonomic distribution. Bacterial genomes are highly plastic, and strains from the same species often harbor different genes, many of which are involved in different metabolic pathways 81 . Fermentative pathways in lactobacilli, for example, are often strain-specific 86 , and the products of the enterobacterial mixed acid fermentation also exhibit wide variability between strains 87 . If MASLD progression is linked to the metabolic functions of the GM, then the association between the disease and the taxonomic profile may be obscured by the inherent metabolic diversity of bacterial species. To study these potential alternatives, we followed two distinct strategies. To identify species associated with MASLD and overcome the limitations of 16S rRNA analysis, we utilized multiple marker genes for taxonomic classification, in an attempt to provide a more comprehensive view of microbial abundances in health and disease states. Using MetaPhlan, we profiled the taxonomic composition of the GM in metagenomic fecal samples of healthy and MASLD donors across three independent cohorts. We identified 28 bacterial genera shared across all cohorts ( Fig. 1A ), collectively accounting for 30-60% of the total GM abundance ( Fig. 1B ). Among these, five candidate genera- Gemmiger , Streptococcus , Mediterraneibacter , Ruminococcus and Agathobacter ( Fig. 1C )- and two species, Agathobacter rectalis and Bacteroides uniformis ( Fig. 1E ), exhibited consistent differences in their relative abundance between groups across all the cohorts. These taxa rank among the most abundant clades in the GM ( Fig. 1C , Supplementary Fig. 1). However, only Agathobacter rectalis showed statistically significant differences. Given its high abundance in the GM, its consistent abundance trend at both genus and species level, and its significant depletion in MASLD, A. rectalis emerges as a robust taxonomic marker for the disease. Although this observation has been previously reported 18 , 21 , this result should be interpreted with caution due to ongoing debates in bacterial nomenclature, as A. rectalis is a newer designation that has replaced Eubacterium rectale 88 . Updates in taxonomic assignments pose challenges, as newer species or genera may not be consistently reflected across different databases or studies 84 , 89 . Moreover, taxonomic classifications themselves, whether based solely on 16S rRNA or on multiple markers, limit our ability to reliably infer GM changes. Bacterial genome plasticity, driven by MGEs, can blur the lines of taxonomy-phenotype associations, as they contribute to a vast strain-level diversity 81 that remains undetected by taxonomic profiling methods. In fact, we have previously reported that geno-metabolic shifts in GM propionate production associated with inflammatory bowel disease do not align with species-level variations 90 . GM associated metabolism contributes to MASLD by disrupting the integrity of the GLA. For this reason, we focused on identifying abundance differences in the genes responsible for producing metabolites potentially implicated in MASLD across several patient cohorts, i.e. butyrate, SCAs, TMA and methane. We isolated the GM gene families encoding the enzymes involved in the biosynthesis of these metabolites and aligned them against the metagenomic samples to quantify their abundance. We found that butyrate-producing genes are depleted in the GM of MASLD patients ( Fig. 2B ), particularly those involved in the crotonyl-butyryl CoA axis: fabV (butyryl-CoA oxidoreductase), bcd (butyryl-CoA dehydrogenase) and but (butyryl-CoA:acetate-CoA transferase) ( Fig. 2A ). While ptb and buk were also higher in both healthy subcohorts, their levels did not differ significantly between initial and advanced MASLD groups within Cohort 2. This suggests that CoA transfer may be the predominant pathway for butyrate production in the GM, as previously reported 91 . The subtler differences in ptb and buk in Cohort 2 may reflect a common MASLD phenotype across different fibrosis stages, potentially influenced by the smaller sample size relative to Cohort 3. Additionally, two other enzymes-structured as two-subunit systems-, can transfer CoA from different cofactors to butyrate to produce butyryl-CoA: butyryl-CoA:acetoacetate-CoA transferase (encoded by atoD and atoA, EC 2.8.3.9) and acetate CoA-transferase (encoded by ctfA and ctfB, EC 2.8.3.9) ( Fig. 2A ). These were not included in this study since they catalyze the reverse reaction, consuming butyrate to form butyryl-CoA 92 . Notably, the depletion of butyrate genes was not observed in MASLD patients from Cohort 1. Contrarily, genes involved in the formation of ethanol and propanol through connected fermentative reactions are generally increased in MASLD ( Fig. 3 ). Genes mhpF and aldB -encoding two aldehyde dehydrogenases that produce ethanol-, and yahK and ahr -encoding two NADPH-dependent aldehyde reductases that form acetaldehyde-are significantly increased in the GM of MASLD patients compared to the healthy groups ( Fig. 3B ). Genes involved in the production of 1-propanol structured in operon pdu are also increased in MASLD ( Fig. 3C , Supplementary Fig. 4A). Specifically, pduCDE -encoding a propanediol dehydratase-, pduP -encoding a propionaldehyde dehydrogenase- and pduQ and adhP -encoding two distinct propanol dehydrogenases-. The pronounced differential abundance of these propanol-producing genes, organized in the pdu operon and adhP , which have been largely unexplored in the context of liver disease, suggests that 1-propanol produced in bacterial microcompartments may play a more significant role than ethanol in MASLD onset and progression. The abundance of genes in the TMA-methane metabolic axis is also disrupted in MASLD. Genes involved in methane formation ( Fig. 4A ) are depleted in the GM of individuals with the pathology. This reduction includes both MCR I and II coding-genes ( Fig. 4B ), which reduce methyl-CoM to methane and mtr , hdr ( Fig. 4C ), and Mtb / Mta-coding genes (Supplementary Fig. 2), dedicated to regenerating methyl-CoM. In contrast, TMAO-reductive genes, organized in torCAD and torZY operons and responsible for regenerating TMA, are increased in MASLD ( Fig. 3D , Supplementary Fig. 4B). These results suggest that methane production is impaired in MASLD. The opposing trend in the abundance of tor operons further indicates that this impairment stems from a metabolic bottleneck that leads to TMA accumulation instead of its conversion to methyl-CoM, effect that is driven by TMAO reductases I and II. Taken together, these results reveal a geno-metabolic alteration associated with MASLD, characterized by an increase in genes involved in the production of SCAs and TMA, and a concurrent depletion of genes responsible for butyrate and methane production ( Fig. 6 ). We confirmed that differences in gene abundance were driven by the MASLD phenotype by showing no differences in USCG abundance between groups (Supplementary Fig. 3). These findings support the hypothesis that the accumulation of endogenous alcohols may represent a key metabolic feature of MASLD, occurring alongside the displacement of butyrate-producing taxa, as previously proposed 33 . Moreover, the observed decrease in A. rectalis abundance partially explains the reduction in butyrate levels, as this organism is a known GM butyrate producer 88 . Download figure Open in new tab Figure 6. Geno-metabolic scheme involved in MASLD. Proposed scheme showing the GM genic alterations associated with butyrate, SCAs, TMA and methane production in MASLD. Genes involved in the production of ethanol, propanol and TMA are increased in the pathology (top), whereas those associated with butyrate and methane are decreased (bottom). Blue boxes indicate the catalyzing genes whose abundance is altered in each metabolic step. Download figure Open in new tab Figure 7. Overview of the methodology used in this study. This diagram summarizes the complete computational workflow for processing whole paired-end metagenomic samples, including data retrieval from the Short Read Archive (SRA), read quality control and taxonomic profiling. Target gene families were isolated and their abundance quantified by alignment against the sequencing libraries. Additionally, the presence of target genes was examined across GM genomes and annotated plasmids was inspected. Blue arrows indicate the progression between workflow steps, with the corresponding software shown in black beneath each arrow. Filtering steps are highlighted in pink boxes. Red-labeled references indicate the corresponding subsection in Materials and Methods (A-H), where each step is explained in detail. Notably, results obtained at the single-gene level using curated gene families were partially reproduced with the whole-genome approach applied to random sub-cohorts, particularly for genes involved in the TMA-methane metabolic axis and propanol production (Supplementary Fig. 4). This partial overlap supports the accuracy of the single-gene analysis. However, the lack of complete concordance may be attributed either to the reduced sample size in the co-assemblies or to mismatches between our custom gene families and the broader Kegg Orthology (KO) annotations required to detect the same functional signals. Importantly, we found that metabolic genes associated with MASLD exhibit a highly accessory profile within the GM ( Fig. 5 ). Butyrate-producing genes are highly accessory in nearly half of the most abundant GM genera, while SCA-producing genes display an even more pronounced accessory pattern ( Fig. 5A ). Among the two main butyrate biosynthesis routes, genes from the butyryl-CoA pathway are more pervasive in the GM than those from the butyryl-P pathway. Notably, Phocaeicola and Bacteroides - the two most abundant GM genera-were excluded from this analysis due to a limited number of high-quality genomes passing our completeness filters, despite their extensive genomic representation in the UHGG. Interestingly, the two key genes in the butyryl-CoA pathway ( bcd and but ) are core in A. rectalis . Nonetheless, some strains of this known butyrate-producing species also harbor genes for ethanol and propanol production, which our findings suggest may contribute to MASLD onset. This highlights a critical point: taxonomic signatures alone are insufficient for associating bacterial clades with disease. Genome plasticity introduces substantial strain-level metabolic variability in the GM, leading to distinct phenotypic traits even within the same species: an effect well-documented in both pathogenic and commensal E. coli strains 93 , 94 . Further supporting their mobility, candidate genes are frequently encoded in plasmids ( Fig. 5C ). Among these, ethanol formation genes ( aldhB , yahK , and ahr ) are the most widely distributed, while butyrate- and propanol-related genes are found in plasmids less frequently. Since non-essential, adaptive metabolic pathways are often encoded in plasmids across many bacterial species 94 – 96 , their presence likely drives phenotypic differences in MASLD without altering the overall species composition detected by conventional metataxonomic marker genes. If MASLD is actually driven by bacterial metabolism, strains from different species, but harboring similar metabolic pathways, may produce similar liver-impacting metabolites, potentially contributing to MASLD pathogenesis regardless of their specific taxonomic position. 4. Conclusions Our study reveals that MASLD is marked by distinct geno-metabolic alterations in the GM that are not captured by conventional taxonomic profiling. Specifically, we observed an increase in genes involved in the production of SCAs and TMA, coupled with a depletion of genes responsible for butyrate and methane production. These results confirm multiple previous findings and suggest a broader metabolic shift that may contribute to the pathogenesis of MASLD, with the displacement of beneficial butyrate-producing taxa such as Agathobacter rectalis . Importantly, our work underscores the value of analyzing metagenomes at gene level to extract additional phenotypic signatures beyond what 16S rRNA-based or marker gene-based approaches can offer. By integrating metabolic pathway analysis with the isolation of specific gene families, and by examining gene abundance, accessory gene profiles and plasmid-mediated gene mobilization, we have significantly advanced our understanding of bacterial metabolism in MASLD. This comprehensive approach not only enhances our understanding of the microbial etiology of MASLD, but also paves the way for improved bacterial analysis in clinical settings, offering new avenues for diagnosis and therapeutic intervention in liver pathologies. Materials and Methods A. Downloading of sequencing metagenomic data A total of 554 metagenomic sequencing libraries generated from the fecal samples of MASLD and healthy subjects were downloaded from the SRA with SRA Toolkit’s fastq-dump 97 . These samples were retrieved using the accession numbers corresponding to five different studies, and they were stratified in three MASLD cohorts, as detailed in Table 1 . Information on the collection and processing of fecal samples, bacterial DNA extraction and sequencing protocols can be found in the original studies. View this table: View inline View popup Download powerpoint Table 1: MASLD patient cohorts analyzed in this study. Each cohort (column 1) was divided into two phenotypic comparison groups: control vs. disease (column 2). The cohorts include N MASLD patients and corresponding whole-metagenomic stool samples derived from them (column 3). The disease group in Cohort 1 consists of 73 99 and 48 100 MASLD patients with initial fibrosis. Cohort 2 was divided into initial fibrosis (control) and advanced fibrosis groups. Patients were originally recruited in the listed studies (column 4), with metagenomic data and medical metadata downloaded from the indicated SRA accession numbers (column 5). Additionally, we retrieved the medical metadata associated with each sample donor to confirm their MASLD status and classify samples into two phenotypic groups of comparison within each cohort. Samples lacking a clearly defined MASLD status were excluded from this study. All data was downloaded between March 2022 and March 2024 B. Description of MASLD patient cohorts Cohort 1 consists of two sets of European, obese MASLD patients aged 20-64 years with initial fibrosis (F0-F2) and varying degrees of steatosis: 73 patients from Hoyles et al. 99 and 10 patients from Mardinoglu et al. 100 . In the latter group, 48 fecal samples were collected at five time points during the study (two samples from day 1 were not obtained). Diagnoses were confirmed through hepatic biopsy performed during obesity surgery, and none of the subjects had alcoholic liver disease, hepatitis or diabetes. The control group for comparison includes 117 healthy patients from Poyet et al. 98 . Cohort 2 comprises 86 American MASLD patients from Loomba et al. 18 , including 72 with initial (F0-F2) and 14 with advanced fibrosis (F3-F4). Diagnoses were confirmed through hepatic biopsy and magnetic resonance imaging, and none of the patients had alcoholic liver disease, hepatitis or diabetes. Note that in this study, the initial fibrosis group is referred to as “healthy” for group comparison purposes, despite including patients with early-stage MASLD. Cohort 3 includes 230 Han Chinese patients from Qin et al. 101 : 125 diagnosed with cirrhotic MASLD and 105 with no hepatic pathology. None of these subjects had hepatitis. All cohorts consist of one metagenomic stool sample per patient, except for the 48 samples collected at multiple time points from 10 patients in the study from Mardinoglu et al. 100 . C. Quality control of metagenomic reads Sequencing reads were trimmed to remove Illumina adapter remnants and low-quality regions (<Q25) using BBDuk from the BBTools suite (version 37.62) 102 . Read quality was assessed with FastQC (version 0.12.1) 103 and summarized using MultiQC (version 1.22.3) 104 . D. Taxonomic signatures associated with MASLD To determine whether any bacterial clade is systematically increased or decreased in MASLD across the three cohorts, we profiled the taxonomic composition of GM communities in fecal samples using MetaPhlAn (version 4.0.6), which uses a wide-range of clade-specific maker genes to estimate taxonomic abundances 24 . The diversity of these markers allows MetaPhlAn to provide highly accurate taxonomic classification across a wide spectrum of bacterial clades, from the species level to broader taxonomic ranks, without being limited to 16S rRNA alone. For this study, we used the database version mpa_vOct22_CHOCOPhlAnSGB_202212 , downloaded on August 22, 2023, and added the --ignore_eukaryotes flag to skip profiling eukaryotic organisms. Post-processing of the MetaPhlAn output involved a) removing low-abundant clades (i.e., those with an average relative abundance of 0 in at least one cohort), b) excluding taxa predicted as “unclassified”, and c) verifying that excluded clades were not solely present in only one of the comparison groups, in order to avoid removing what could be very clear taxonomic markers. To identify the most abundant clades present in the GM, we selected genera with an average relative abundance above 0.5% ( Fig. 1C ) and species above 0.2% ( Fig. 1E ). Notedly, in this step we faced an issue regarding the status of taxonomic assignments, as several species required manual updates to their nomenclature assigned by MetaPhlAn. Specifically, Eubacterium rectale was reassigned to Agathobacter rectalis 88 , Ruminococcus torques to Mediterraneibacter torques 105 and Roseburia faecis to Agathobacter faecis 106 . E. Isolation of target gene families in the GM We constructed families of the human GM metabolic genes encoding the enzymes involved in the production of butyrate, ethanol, propanol, TMA, methane and their precursor metabolites, as represented in Figs. 2A , 3A and 4A. The Unified Human Gastrointestinal Genome (UHGG, version 2.0.2) 81 served as the reference database for this purpose. This resource, comprising 289,232 bacterial and archaeal genomes and more than 665 million proteins, represents the most comprehensive collection of characterized human GM genes to date. Each gene family was built by retrieving a seed set of sequences based on gene and protein annotation from the UHGG, assessing their similarity with lab-validated enzymatic sequences when available. Validated gene sequences were retrieved from MetaCyc 107 and used as controls to assess their phylogenetic proximity to the retrieved gene seeds. Gene families were aligned using MAFFT (version 7.271) 108 with options “--localpair” and “--maxiterate 1000” and phylogenetic trees were constructed with IQ-TREE (version 2.0.3) 109 with the ultrafast bootstrap option (1000 bootstraps) 110 to assess branch support. The best fitting model for each metabolic gene was estimated using ModelFinder Plus 111 , according to the Bayesian information criterion. Low phylogenetic distances between control and retrieved sequences served as a criterion for defining gene families. Duplicated sequences, gene fragments and phylogenetically distant genes were excluded after inspecting the trees. To improve gene identification and classification, we then built profile Hidden Markov Model (HMM) for each target gene family using HMMER hmpress (version 3.4) 112 . MetaCyc 107 and KEGG 113 were also used to validate gene and enzyme names, KO groups, the biochemical reactions involved in the formation of each metabolite, and their enzyme commission (EC) numbers. The same protocol was applied to isolate five USCGs used as controls. These USCGs and their encoded proteins were: argS (arginyl-tRNA synthetase), dnaA (chromosomal replication initiator protein DnaA), rpoA , rpoB and rpoC (alpha, beta and gamma subunits of the DNA-directed RNA polymerase, respectively). F. Analysis of gene abundance by read-based metagenomic quantification We quantified the abundance of target genes in the metagenomic samples from the three cohorts by aligning the quality-controlled reads against the isolated gene families using DIAMOND (version 2.0.14) 114 . Alignments were filtered to retain only the best match for each read and gene family sequence by using the “--max-hsps 1” option, with an E-value threshold of <1E-10. These parameters were based on experimentally validated recommendations for detecting genes in fecal metagenomic samples with DIAMOND 115 . The total number of aligned reads for each gene family was normalized using a variation of the RPKM formula, referred to as RPKSM (reads per kilobase per size per million reads). This measure was used as an indicator of gene abundance ( Figs. 2B , 3B -C and 4B-D ; and Supplementary Figs. 2 and 3). It is defined as the number of aligned reads per kilobase per gene family size per million reads, to account for variations in gene length, gene family size (i.e., the number of sequences within each gene family) and library size, respectively. G. Whole metagenomic analysis We applied a whole-metagenome analysis workflow to validate our results at single-gene level. To this end, SqueezeMeta (version 1.6.0) 116 was used to predict both the taxonomic and functional composition of the metagenomic samples from the three cohorts. Due to computational limitations, it was not feasible to co-assemble all available samples, so approximately 24-30 samples per cohort were randomly selected, with equal representation from each phenotypic group. Briefly, SqueezeMeta assembled the metagenomes from each sub-cohort into a single co-assembly using MEGAHIT (version 1.2.9) 117 by pooling the sequencing reads from every sample. ORFs were then predicted from contigs 118 , and reads from individual samples were mapped back against the co-assembly to quantify the abundance of each KO group per metagenome. We further processed these results with SQMtools (version 1.6.0) 119 to identify the most differentially abundant KOs between phenotypic groups within each cohort through DESeq2 (version 1.34.0) 120 , applying a log fold-change >2 and a p-adjusted value <0.05 as thresholds (Supplementary Fig. 4). H. Presence of candidate metabolic genes across the human GM genomes and annotated plasmids To determine whether candidate metabolic genes involved in MASLD were core or accessory across the human GM, their presence was quantified in both UHGG genomes and curated RefSeq200 plasmids 121 . Curation involved eliminating sequences corresponding to partial plasmid DNA sequences, unassignable hosts, or PacBio internal control sequences, as done previously 16 . Since many UHGG genomes originate from metagenome-assembled or single-cell-amplified genomes, they may be incomplete, potentially leading to false positive annotations. To minimize this risk, we included only UHGG genomes with >95% completeness and containing the five USCGs described above. ORFs were predicted using Prodigal (version 2.6.3) 122 , retaining only those covering >90% of the HMM size. We then used HMMER hmmscan with the following parameters: protein-pHMM coverage threshold >90%, E-value <0.01 and independent E-value (i-E value) <0.01, strictly enforcing the 90% threshold to avoid capturing ORFs with similar protein domains. This approach yielded 31,227 genomes classified as complete and containing the five USCGs. To classify genes as core or accessory, we defined their presence as the percentage of strains within a given genus and species that encode the gene. Genes were categorized as core if present in >80% of congeneric ( Fig. 5A ) or conspecific ( Fig. 5B ) genomes, as accessory if present in 20-80%, and as highly accessory if present in 100 genomes meeting the completeness criteria were included. Additionally, we assessed whether these genes could be encoded in annotated plasmids. Using the same constraints, we predicted ORFs from the RefSeq200 plasmid collection, which includes 23,309 plasmids encoding over 2 million ORFs. ORFS were retained if they met the >90% HMM size threshold, with hmmscan parameters identical to those applied in the UHGG genome analysis ( Fig. 5C ). I. R packages and statistical analysis Data manipulation was performed using R (version 4.1.3) and the tidyverse package (version 1.3.2) 123 . Figures were generated with the R packages ggplot2 (version 3.4.2) 124 , ggpubr (version 0.6.0) 125 and pheatmap (version 1.0.12) 126 . Mann-Whitney tests with a Benjamini-Hochberg adjustment for multiple testing were applied on the relative abundance of bacterial clades ( Figs. 1D and 1E ) and gene abundance ( Figs. 2B , 3B -C, 4B-D , and Supplementary Figs. 2 and 3) between phenotypic groups in the metagenomic fecal samples of each cohort to assess the significance of the differences. Statistical analysis and graphical representation on the plots were conducted using the R package rstatix (version 0.7.2) 127 . Footnotes Updated figure captions Revision of supplementary material Updated discussion Fixed typos References 1. ↵ Riazi , K. et al. The prevalence and incidence of NAFLD worldwide: a systematic review and meta-analysis . Lancet Gastroenterol. Hepatol . 7 , 851 – 861 ( 2022 ). OpenUrl CrossRef PubMed 2. ↵ EASL . EASL–EASD–EASO Clinical practice guidelines for the management of non-alcoholic fatty liver disease . J. Hepatol . 64 , 1388 – 1402 ( 2016 ). OpenUrl CrossRef PubMed 3. ↵ Allen , A. M. et al. Clinical course of non-alcoholic fatty liver disease and the implications for clinical trial design . J. Hepatol . 77 , 1237 – 1245 ( 2022 ). OpenUrl CrossRef PubMed 4. ↵ Powell , E. E. , Wong , V. W.-S. & Rinella , M . Non-alcoholic fatty liver disease . The Lancet 397 , 2212 – 2224 ( 2021 ). OpenUrl CrossRef 5. ↵ Yki-Järvinen , H . Non-alcoholic fatty liver disease as a cause and a consequence of metabolic syndrome . Lancet Diabetes Endocrinol . 2 , 901 – 910 ( 2014 ). OpenUrl CrossRef PubMed 6. ↵ Martín-Mateos , R. & Albillos , A . The role of the gut-liver axis in metabolic dysfunction-associated fatty liver disease . Front. Immunol . 12 , 660179 ( 2021 ). 7. ↵ Carpino , G. et al. Increased liver localization of lipopolysaccharides in human and experimental NAFLD . Hepatology 72 , 470 – 485 ( 2020 ). OpenUrl CrossRef PubMed 8. ↵ Li , Q. , Rempel , J. D. , Yang , J. & Minuk , G. Y . The effects of pathogen-associated molecular patterns on peripheral blood monocytes in patients with non-alcoholic fatty liver disease . J. Clin. Exp. Hepatol . 12 , 808 – 817 ( 2022 ). OpenUrl CrossRef PubMed 9. ↵ Miele , L. et al. Increased intestinal permeability and tight junction alterations in non-alcoholic fatty liver disease . Hepatology 49 , 1877 – 1887 ( 2009 ). OpenUrl CrossRef PubMed Web of Science 10. ↵ Wang , W. et al. Tauroursodeoxycholic acid inhibits intestinal inflammation and barrier disruption in mice with non-alcoholic fatty liver disease . Br. J. Pharmacol . 175 , 469 – 484 ( 2018 ). OpenUrl CrossRef PubMed 11. ↵ Kuang , J. et al. Hyodeoxycholic acid alleviates non-alcoholic fatty liver disease through modulating the gut-liver axis . Cell Metab . 35 , 1752 – 1766.e8 ( 2023 ). OpenUrl CrossRef PubMed 12. ↵ Yang , C. et al. Altered gut microbial profile accompanied by abnormal short chain fatty acid metabolism exacerbates nonalcoholic fatty liver disease progression . Sci. Rep . 14 , 22385 ( 2024 ). 13. ↵ Zhang , R. et al. Gut microbial metabolites in MASLD: implications of mitochondrial dysfunction in the pathogenesis and treatment . Hepatol. Commun . 8 , 1 – 21 ( 2024 ). OpenUrl 14. ↵ Del Chierico , F. et al. Gut microbiota profiling of pediatric non-alcoholic fatty liver disease and obese patients unveiled by an integrated meta-omics-based approach . Hepatology 65 , 451 – 464 ( 2017 ). OpenUrl CrossRef PubMed 15. ↵ Lee , G. et al. Distinct signatures of gut microbiome and metabolites associated with significant fibrosis in non-obese NAFLD . Nat. Commun . 11 , 4982 ( 2020 ). OpenUrl CrossRef PubMed 16. ↵ Redondo-Salvo , S. et al. Pathways for horizontal gene transfer in bacteria revealed by a global map of their plasmids . Nat. Commun . 11 , 3602 ( 2020 ). OpenUrl CrossRef PubMed 17. ↵ Lanthier , N. et al. Microbiota analysis and transient elastography reveal new extra-hepatic components of liver steatosis and fibrosis in obese patients . Sci. Rep . 11 , 659 ( 2021 ). 18. ↵ Loomba , R. et al. Gut microbiome-based metagenomic signature for non-invasive detection of advanced fibrosis in human non-alcoholic fatty liver disease . Cell Metab . 25 , 1054 – 1062.e5 ( 2017 ). OpenUrl CrossRef PubMed 19. ↵ Boursier , J. et al. The severity of non-alcoholic fatty liver disease is associated with gut dysbiosis and shift in the metabolic function of the gut microbiota . Hepatology 63 , 764 – 775 ( 2016 ). OpenUrl CrossRef PubMed 20. ↵ Caussy , C. et al. A gut microbiome signature for cirrhosis due to non-alcoholic fatty liver disease . Nat. Commun . 10 , 1406 ( 2019 ). OpenUrl CrossRef PubMed 21. ↵ Oh , T. G. et al. A universal gut-microbiome-derived signature predicts cirrhosis . Cell Metab . 32 , 878 – 888.e6 ( 2020 ). OpenUrl CrossRef PubMed 22. ↵ Shen , F. et al. Gut microbiota dysbiosis in patients with non-alcoholic fatty liver disease . Hepatobiliary Pancreat. Dis. Int . 16 , 375 – 381 ( 2017 ). OpenUrl CrossRef PubMed 23. ↵ Aron-Wisnewsky , J. et al. Gut microbiota and human NAFLD: disentangling microbial signatures from metabolic disorders . Nat. Rev. Gastroenterol. Hepatol . 17 , 279 – 297 ( 2020 ). OpenUrl CrossRef PubMed 24. ↵ Blanco-Míguez , A. et al. Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4 . Nat. Biotechnol . 41 , 1633 – 1644 ( 2023 ). OpenUrl CrossRef PubMed 25. ↵ McPherson , S. et al. Age as a confounding factor for the accurate non-invasive diagnosis of advanced NAFLD fibrosis . Am. J. Gastroenterol . 112 , 740 – 751 ( 2017 ). OpenUrl CrossRef PubMed 26. ↵ Ponziani , F. R. , Gasbarrini , A. & Pompili , M . NAFLD or comorbidities, that is the question . J. Hepatol . 73 , 723 ( 2020 ). 27. ↵ Chu , H. , Duan , Y. , Yang , L. & Schnabl , B . Small metabolites, possible big changes: a microbiota-centered view of non-alcoholic fatty liver disease . Gut 68 , 359 – 370 ( 2019 ). OpenUrl Abstract / FREE Full Text 28. Dai , X. et al. Microbial metabolites: critical regulators in NAFLD . Front. Microbiol . 11 , 567654 ( 2020 ). 29. ↵ Martin-Grau , M. & Monleón , D . The role of microbiota-related co-metabolites in MASLD progression: a narrative review . Curr. Issues Mol. Biol . 46 , 6377 – 6389 ( 2024 ). OpenUrl CrossRef PubMed 30. ↵ Oliphant , K. & Allen-Vercoe , E . Macronutrient metabolism by the human gut microbiome: major fermentation by-products and their impact on host health . Microbiome 7 , 91 ( 2019 ). 31. ↵ Donaldson , G. P. , Lee , S. M. & Mazmanian , S. K . Gut biogeography of the bacterial microbiota . Nat. Rev. Microbiol . 14 , 20 – 32 ( 2016 ). OpenUrl CrossRef PubMed 32. ↵ Albillos , A. , De Gottardi , A. & Rescigno , M . The gut-liver axis in liver disease: pathophysiological basis for therapy . J. Hepatol . 72 , 558 – 577 ( 2020 ). OpenUrl CrossRef PubMed 33. ↵ Iruzubieta , P. , Medina , J. M. , Fernández-López , R. , Crespo , J. & De La Cruz , F . A role for gut microbiome fermentative pathways in fatty liver disease progression . J. Clin. Med . 9 , 1369 ( 2020 ). OpenUrl CrossRef PubMed 34. ↵ Miller , T. L. & Wolin , M. J . Pathways of acetate, propionate, and butyrate formation by the human fecal microbial flora . Appl. Environ. Microbiol . 62 , 1589 – 1592 ( 1996 ). OpenUrl Abstract / FREE Full Text 35. ↵ Koh , A. , De Vadder , F. , Kovatcheva-Datchary , P. & Bäckhed , F . From dietary fiber to host physiology: short-chain fatty acids as key bacterial metabolites . Cell 165 , 1332 – 1345 ( 2016 ). OpenUrl CrossRef PubMed 36. Morrison , D. J. & Preston , T . Formation of short chain fatty acids by the gut microbiota and their impact on human metabolism . Gut Microbes 7 , 189 – 200 ( 2016 ). OpenUrl CrossRef PubMed 37. ↵ Blaak , E. E. et al. Short chain fatty acids in human gut and metabolic health . Benef. Microbes 11 , 411 – 455 ( 2020 ). OpenUrl CrossRef PubMed 38. ↵ Donohoe , D. R. et al. The microbiome and butyrate regulate energy metabolism and autophagy in the mammalian colon . Cell Metab . 13 , 517 – 526 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 39. ↵ Wang , H.-B. , Wang , P.-Y. , Wang , X. , Wan , Y.-L. & Liu , Y.-C . Butyrate enhances Intestinal epithelial barrier function via up-regulation of tight junction protein claudin-1 transcription . Dig. Dis. Sci . 57 , 3126 – 3135 ( 2012 ). OpenUrl CrossRef PubMed 40. ↵ Gao , Z. et al. Butyrate improves insulin sensitivity and increases energy expenditure in mice . Diabetes 58 , 1509 – 1517 ( 2009 ). OpenUrl Abstract / FREE Full Text 41. ↵ Li , Z. et al. Butyrate reduces appetite and activates brown adipose tissue via the gut-brain neural circuit . Gut 67 , 1269 – 1279 ( 2018 ). OpenUrl Abstract / FREE Full Text 42. ↵ De Vadder , F. et al. Microbiota-generated metabolites promote metabolic benefits via gut-brain neural circuits . Cell 156 , 84 – 96 ( 2014 ). OpenUrl CrossRef PubMed Web of Science 43. ↵ Sanna , S. et al. Causal relationships among the gut microbiome, short-chain fatty acids and metabolic diseases . Nat. Genet . 51 , 600 – 605 ( 2019 ). OpenUrl CrossRef PubMed 44. ↵ Endo , H. , Niioka , M. , Kobayashi , N. , Tanaka , M. & Watanabe , T . Butyrate-producing probiotics reduce non-alcoholic fatty liver disease progression in rats: new insight into the probiotics for the gut-liver axis . PLoS ONE 8 , e63388 ( 2013 ). OpenUrl CrossRef PubMed 45. ↵ Jin , C. J. , Sellmann , C. , Engstler , A. J. , Ziegenhardt , D. & Bergheim , I . Supplementation of sodium butyrate protects mice from the development of non-alcoholic steatohepatitis (NASH) . Br. J. Nutr . 114 , 1745 – 1755 ( 2015 ). OpenUrl CrossRef PubMed 46. Zhou , D. et al. Sodium butyrate attenuates high-fat diet-induced steatohepatitis in mice by improving gut microbiota and gastrointestinal barrier . World J. Gastroenterol . 23 , 60 ( 2017 ). 47. ↵ Zhou , D. et al. Sodium butyrate reduces high-fat diet-induced non-alcoholic steatohepatitis through upregulation of hepatic GLP-1R expression . Exp. Mol. Med . 50 , 1 – 12 ( 2018 ). OpenUrl CrossRef PubMed 48. ↵ Zhou , D. et al. Total fecal microbiota transplantation alleviates high-fat diet-induced steatohepatitis in mice via beneficial regulation of gut microbiota . Sci. Rep . 7 , 1529 ( 2017 ). OpenUrl CrossRef PubMed 49. ↵ Louis , P. & Flint , H. J . Formation of propionate and butyrate by the human colonic microbiota . Environ. Microbiol . 19 , 29 – 41 ( 2017 ). OpenUrl CrossRef PubMed 50. ↵ Zhu , L. et al. Characterization of gut microbiomes in non-alcoholic steatohepatitis (NASH) patients: a connection between endogenous alcohol and NASH . Hepatology 57 , 601 – 609 ( 2013 ). OpenUrl CrossRef PubMed Web of Science 51. ↵ Michail , S. et al. Altered gut microbial energy and metabolism in children with non-alcoholic fatty liver disease . FEMS Microbiol. Ecol . 91 , 1 – 9 ( 2015 ). OpenUrl CrossRef PubMed 52. ↵ Meijnikman , A. S. et al. Microbiome-derived ethanol in non-alcoholic fatty liver disease . Nat. Med . 28 , 2100 – 2106 ( 2022 ). OpenUrl CrossRef PubMed 53. ↵ Yuan , J. et al. Fatty liver disease caused by high-alcohol-producing Klebsiella pneumoniae . Cell Metab . 30 , 675 – 688.e7 ( 2019 ). OpenUrl CrossRef PubMed 54. ↵ Chen , X. et al. Endogenous ethanol produced by intestinal bacteria induces mitochondrial dysfunction in non-alcoholic fatty liver disease . J. Gastroenterol. Hepatol . 35 , 2009 – 2019 ( 2020 ). OpenUrl CrossRef PubMed 55. ↵ Fromenty , B. & Roden , M . Mitochondrial alterations in fatty liver diseases . J. Hepatol . 78 , 415 – 429 ( 2023 ). OpenUrl CrossRef PubMed 56. ↵ Ruuskanen , M. O. et al. Links between gut microbiome composition and fatty liver disease in a large population sample . Gut Microbes 13 , 1888673 ( 2021 ). 57. ↵ Malaguarnera , G . Gut microbiota in alcoholic liver disease: pathogenetic role and therapeutic perspectives . World J. Gastroenterol . 20 , 16639 ( 2014 ). 58. ↵ Schut , G. J. , et al. Tungsten enzymes play a role in detoxifying food and antimicrobial aldehydes in the human gut microbiome . Proc. Natl. Acad. Sci . 118 , e2109008118 ( 2021 ). OpenUrl Abstract / FREE Full Text 59. ↵ Pony , P. , Rapisarda , C. , Terradot , L. , Marza , E. & Fronzes , R . Filamentation of the bacterial bi-functional alcohol/aldehyde dehydrogenase AdhE is essential for substrate channeling and enzymatic regulation . Nat. Commun . 11 , 1426 ( 2020 ). OpenUrl CrossRef PubMed 60. ↵ Hitschler , L. , Nissen , L. S. , Kuntz , M. & Basen , M . Alcohol dehydrogenases AdhE and AdhB with broad substrate ranges are important enzymes for organic acid reduction in Thermoanaerobacter sp. strain X514 . Biotechnol. Biofuels 14 , 187 ( 2021 ). 61. ↵ Pick , A. , Rühmann , B. , Schmid , J. & Sieber , V . Novel CAD-like enzymes from Escherichia coli K-12 as additional tools in chemical production . Appl. Microbiol. Biotechnol . 97 , 5815 – 5824 ( 2013 ). OpenUrl CrossRef 62. ↵ Merchel Piovesan Pereira , B. , Adil Salim , M. , Rai , N. & Tagkopoulos , I. Tolerance to glutaraldehyde in Escherichia coli mediated by overexpression of the aldehyde reductase YqhD by YqhC . Front. Microbiol . 12 , 680553 ( 2021 ). 63. ↵ Moreira De Gouveia , M. I. , Daniel , J. , Garrivier , A. , Bernalier-Donadille , A . & Jubelin , G. Diversity of ethanolamine utilization by human commensal Escherichia coli . Res. Microbiol . 174 , 103989 ( 2023 ). 64. ↵ Dellomonaco , C. , Clomburg , J. M. , Miller , E. N. & Gonzalez , R . Engineered reversal of the β-oxidation cycle for the synthesis of fuels and chemicals . Nature 476 , 355 – 359 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 65. ↵ Eram , M. & Ma , K . Decarboxylation of pyruvate to acetaldehyde for ethanol production by hyperthermophiles . Biomolecules 3 , 578 – 596 ( 2013 ). OpenUrl CrossRef PubMed 66. ↵ Ho , K. K. & Weiner , H . Isolation and characterization of an aldehyde dehydrogenase encoded by the aldB gene of Escherichia coli . J. Bacteriol . 187 , 1067 – 1073 ( 2005 ). OpenUrl Abstract / FREE Full Text 67. ↵ Ravcheev , D. A. , Moussu , L. , Smajic , S. & Thiele , I . Comparative genomic analysis reveals novel microcompartment-associated metabolic pathways in the human gut microbiome . Front. Genet . 10 , 636 ( 2019 ). 68. ↵ Dank , A. et al. Bacterial microcompartment-dependent 1,2-propanediol utilization of Propionibacterium freudenreichii . Front. Microbiol . 12 , 679827 ( 2021 ). 69. ↵ Chen , Y. et al. Associations of gut-flora-dependent metabolite trimethylamine-N-oxide, betaine and choline with non-alcoholic fatty liver disease in adults . Sci. Rep . 6 , 19076 ( 2016 ). 70. ↵ Flores-Guerrero , J. L. et al. Circulating trimethylamine- N -oxide is associated with all-cause mortality in subjects with non-alcoholic fatty liver disease . Liver Int . 41 , 2371 – 2382 ( 2021 ). OpenUrl CrossRef PubMed 71. ↵ Wang , Z. et al. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease . Nature 472 , 57 – 63 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 72. ↵ Roncal , C. et al. Trimethylamine-N-oxide (TMAO) predicts cardiovascular mortality in peripheral artery disease . Sci. Rep . 9 , 15580 ( 2019 ). 73. ↵ Martínez-del Campo , A., et al. Characterization and detection of a widely distributed gene cluster that predicts anaerobic choline utilization by human gut bacteria . mBio 6 , e00042 – 15 ( 2015 ). OpenUrl CrossRef PubMed 74. ↵ Goris , M. et al. Increased thermostability of an engineered flavin-containing monooxygenase to remediate trimethylamine in fish protein hydrolysates . Appl. Environ. Microbiol . 89 , e00390 – 23 ( 2023 ). OpenUrl PubMed 75. ↵ Leimkühler , S . The biosynthesis of the molybdenum cofactors in Escherichia coli . Environ. Microbiol . 22 , 2007 – 2026 ( 2020 ). OpenUrl CrossRef PubMed 76. ↵ Kurth , J. M. et al. Methanogenic archaea use a bacteria-like methyltransferase system to demethoxylate aromatic compounds . ISME J . 15 , 3549 – 3565 ( 2021 ). OpenUrl CrossRef PubMed 77. ↵ Gendron , A. & Allen , K. D . Overview of diverse methyl/alkyl-coenzyme M reductases and considerations for their potential heterologous expression . Front. Microbiol . 13 , 867342 ( 2022 ). 78. ↵ Thauer , R. K . Methyl (alkyl)-coenzyme M reductases: nickel F-430-containing enzymes involved in anaerobic methane formation and in anaerobic oxidation of methane or of short chain alkanes . Biochemistry 58 , 5198 – 5220 ( 2019 ). OpenUrl CrossRef PubMed 79. ↵ Wagner , T. , Koch , J. , Ermler , U. & Shima , S . Methanogenic heterodisulfide reductase (HdrABC-MvhAGD) uses two noncubane [4Fe-4S] clusters for reduction . Science 357 , 699 – 703 ( 2017 ). OpenUrl Abstract / FREE Full Text 80. ↵ Aziz , I. , et al. Structural and mechanistic basis of the central energy-converting methyltransferase complex of methanogenesis . Proc. Natl. Acad. Sci . 121 , e2315568121 ( 2024 ). OpenUrl CrossRef PubMed 81. ↵ Almeida , A. et al. A unified catalog of 204,938 reference genomes from the human gut microbiome . Nat. Biotechnol . 39 , 105 – 114 ( 2021 ). OpenUrl CrossRef PubMed 82. ↵ Rosconi , F. et al. A bacterial pan-genome makes gene essentiality strain-dependent and evolvable . Nat. Microbiol . 7 , 1580 – 1592 ( 2022 ). OpenUrl CrossRef PubMed 83. ↵ Yu , M. K. , Fogarty , E. C. & Eren , A. M . Diverse plasmid systems and their ecology across human gut metagenomes revealed by PlasX and MobMess . Nat. Microbiol . 9 , 830 – 847 ( 2024 ). OpenUrl CrossRef PubMed 84. ↵ Parks , D. H. et al. GTDB: an ongoing census of bacterial and archaeal diversity through a phylogenetically consistent, rank normalized and complete genome-based taxonomy . Nucleic Acids Res . 50 , D785 – D794 ( 2022 ). OpenUrl CrossRef PubMed 85. ↵ O’Callaghan , J. L. , Willner , D. , Buttini , M. , Huygens , F. & Pelzer , E. S . Limitations of 16S rRNA gene sequencing to characterize Lactobacillus species in the upper genital tract . Front. Cell Dev. Biol . 9 , 641921 ( 2021 ). 86. ↵ Tomita , S. , Saito , K. , Nakamura , T. , Sekiyama , Y. & Kikuchi , J . Rapid discrimination of strain-dependent fermentation characteristics among Lactobacillus strains by NMR-based metabolomics of fermented vegetable juice . PLOS ONE 12 , e0182229 ( 2017 ). OpenUrl CrossRef PubMed 87. ↵ Forster , A. H. & Gescher , J . Metabolic engineering of Escherichia coli for production of mixed-acid fermentation end products . Front. Bioeng. Biotechnol . 2 , ( 2014 ). 88. ↵ Rosero , J. A. et al. Reclassification of Eubacterium rectale (Hauduroy et al. 1937) in a new genus Agathobacter gen. nov. as Agathobacter rectalis comb. nov., and description of Agathobacter ruminis sp. nov., isolated from the rumen contents of sheep and cows . Int. J. Syst. Evol. Microbiol. 66 , 768 – 773 ( 2016 ). OpenUrl CrossRef PubMed 89. ↵ Parks , D. H. et al. A standardized bacterial taxonomy based on genome phylogeny substantially revises the tree of life . Nat. Biotechnol . 36 , 996 – 1004 ( 2018 ). OpenUrl CrossRef PubMed 90. ↵ Medina , J. M. , Fernández-López , R. , Crespo , J. & Cruz , F. D. L . Propionate fermentative genes of the gut microbiome decrease in inflammatory bowel disease . J. Clin. Med . 10 , 2176 ( 2021 ). OpenUrl CrossRef PubMed 91. ↵ Louis , P. , McCrae , S. I. , Charrier , C. & Flint , H. J . Organization of butyrate synthetic genes in human colonic bacteria: phylogenetic conservation and horizontal gene transfer . FEMS Microbiol. Lett . 269 , 240 – 247 ( 2007 ). OpenUrl CrossRef PubMed Web of Science 92. ↵ Cui , Y. et al. Production of butyl butyrate from lignocellulosic biomass through Escherichia coli - Clostridium beijerinckii G117 co-culture . Process Biochem . 128 , 58 – 67 ( 2023 ). OpenUrl CrossRef 93. ↵ Gonzalez-Alba , J. M. , Baquero , F. , Cantón , R. & Galán , J. C . Stratified reconstruction of ancestral Escherichia coli diversification . BMC Genomics 20 , 936 ( 2019 ). 94. ↵ Palomino , A. et al. Metabolic genes on conjugative plasmids are highly prevalent in Escherichia coli and can protect against antibiotic treatment . ISME J . 17 , 151 – 162 ( 2023 ). OpenUrl CrossRef PubMed 95. Abriouel , H. et al. New insights into the role of plasmids from probiotic Lactobacillus pentosus MP-10 in Aloreña table olive brine fermentation . Sci. Rep . 9 , 10938 ( 2019 ). 96. ↵ Pudlo , N. A. et al. Phenotypic and genomic diversification in complex carbohydrate-degrading human gut bacteria . mSystems 7 , e00947 – 21 ( 2022 ). OpenUrl PubMed 97. ↵ Leinonen , R. , Sugawara , H. , Shumway , M. , & on behalf of the International Nucleotide Sequence Database Collaboration. The Sequence Read Archive . Nucleic Acids Res . 39 , D19 – D21 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 98. ↵ Poyet , M. et al. A library of human gut bacterial isolates paired with longitudinal multiomics data enables mechanistic microbiome research . Nat. Med . 25 , 1442 – 1452 ( 2019 ). OpenUrl CrossRef PubMed 99. ↵ Hoyles , L. et al. Molecular phenomics and metagenomics of hepatic steatosis in non-diabetic obese women . Nat. Med . 24 , 1070 – 1080 ( 2018 ). OpenUrl CrossRef PubMed 100. ↵ Mardinoglu , A. et al. An integrated understanding of the rapid metabolic benefits of a carbohydrate-restricted diet on hepatic steatosis in humans . Cell Metab . 27 , 559 – 571.e5 ( 2018 ). OpenUrl CrossRef PubMed 101. ↵ Qin , N. et al. Alterations of the human gut microbiome in liver cirrhosis . Nature 513 , 59 – 64 ( 2014 ). OpenUrl CrossRef PubMed 102. ↵ Bushnell , B. BBMap . ( 2014 ). 103. ↵ Andrews , S. FastQC: A quality control tool for high throughput sequence data . ( 2010 ). 104. ↵ Ewels , P. , Magnusson , M. , Lundin , S. & Käller , M . MultiQC: summarize analysis results for multiple tools and samples in a single report . Bioinformatics 32 , 3047 – 3048 ( 2016 ). OpenUrl CrossRef PubMed 105. ↵ Togo , A. H. et al. Description of Mediterraneibacter massiliensis , gen. nov., sp. nov., a new genus isolated from the gut microbiota of an obese patient and reclassification of Ruminococcus faecis , Ruminococcus lactaris , Ruminococcus torques , Ruminococcus gnavus and Clostridium glycyrrhizinilyticum as Mediterraneibacter faecis comb. nov., Mediterraneibacter lactaris comb. nov., Mediterraneibacter torques comb. nov ., Mediterraneibacter gnavus comb. nov. and Mediterraneibacter glycyrrhizinilyticus comb. nov. Antonie Van Leeuwenhoek 111 , 2107 – 2128 ( 2018 ). OpenUrl PubMed 106. ↵ Murotomi , K. , Tourlousse , D. M. , Hamajima , M. & Sekiguchi , Y . Complete genome sequence of Roseburia faecis M72/1 T . Microbiol. Resour. Announc . 14 , e01253 – 24 ( 2025 ). OpenUrl PubMed 107. ↵ Caspi , R. et al. The MetaCyc database of metabolic pathways and enzymes - a 2019 update . Nucleic Acids Res . 48 , D445 – D453 ( 2020 ). OpenUrl CrossRef PubMed 108. ↵ Katoh , K. & Standley , D. M . MAFFT multiple sequence alignment software version 7: improvements in performance and usability . Mol. Biol. Evol . 30 , 772 – 780 ( 2013 ). OpenUrl CrossRef PubMed Web of Science 109. ↵ Minh , B. Q. et al. IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era . Mol. Biol. Evol . 37 , 1530 – 1534 ( 2020 ). OpenUrl CrossRef PubMed 110. ↵ Hoang , D. T. , Chernomor , O. , Von Haeseler , A. , Minh , B. Q. & Vinh , L. S . UFBoot2: improving the ultrafast bootstrap approximation . Mol. Biol. Evol . 35 , 518 – 522 ( 2018 ). OpenUrl CrossRef PubMed 111. ↵ Kalyaanamoorthy , S. , Minh , B. Q. , Wong , T. K. F. , Von Haeseler , A. & Jermiin , L. S . ModelFinder: fast model selection for accurate phylogenetic estimates . Nat. Methods 14 , 587 – 589 ( 2017 ). OpenUrl CrossRef PubMed 112. ↵ Eddy , S. HMMER . ( 2019 ). 113. ↵ Kanehisa , M. , Furumichi , M. , Sato , Y. , Matsuura , Y. & Ishiguro-Watanabe , M . KEGG: biological systems database as a model of the real world . Nucleic Acids Res . 53 , D672 – D677 ( 2025 ). OpenUrl CrossRef PubMed 114. ↵ Buchfink , B. , Reuter , K. & Drost , H.-G . Sensitive protein alignments at tree-of-life scale using DIAMOND . Nat. Methods 18 , 366 – 368 ( 2021 ). OpenUrl CrossRef PubMed 115. ↵ Treiber , M. L. , Taft , D. H. , Korf , I. , Mills , D. A. & Lemay , D. G . Pre-and post-sequencing recommendations for functional annotation of human fecal metagenomes . BMC Bioinformatics 21 , 74 ( 2020 ). 116. ↵ Tamames , J. & Puente-Sánchez , F . SqueezeMeta, a highly portable, fully automatic metagenomic analysis pipeline . Front. Microbiol . 9 , 3349 ( 2019 ). OpenUrl CrossRef PubMed 117. ↵ Li , D. , Liu , C.-M. , Luo , R. , Sadakane , K. & Lam , T.-W . MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph . Bioinformatics 31 , 1674 – 1676 ( 2015 ). OpenUrl CrossRef PubMed 118. ↵ Hyatt , D. , LoCascio , P. F. , Hauser , L. J. & Uberbacher , E. C . Gene and translation initiation site prediction in metagenomic sequences . Bioinformatics 28 , 2223 – 2230 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 119. ↵ Puente-Sánchez , F. , García-García , N. & Tamames , J . SQMtools: automated processing and visual analysis of ‘omics data with R and anvi’o . BMC Bioinformatics 21 , 358 ( 2020 ). 120. ↵ Love , M. I. , Huber , W. & Anders , S . Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol . 15 , 550 ( 2014 ). 121. ↵ Haft , D. H. et al. RefSeq and the prokaryotic genome annotation pipeline in the age of metagenomes . Nucleic Acids Res . 52 , D762 – D769 ( 2024 ). OpenUrl CrossRef PubMed 122. ↵ Hyatt , D. et al. Prodigal: prokaryotic gene recognition and translation initiation site identification . BMC Bioinformatics 11 , 119 ( 2010 ). 123. ↵ Wickham , H. et al. Welcome to the Tidyverse . J. Open Source Softw . 4 , 1686 ( 2019 ). OpenUrl CrossRef 124. ↵ Wickham , H. Ggplot2: Elegant Graphics for Data Analysis . ( Springer-Verlag New York , 2016 ). 125. ↵ Kassambara , A. ggpubr : ggplot2 based publication ready plots . ( 2023 ). 126. ↵ Kolde , R. pheatmap : pretty heatmaps . ( 2019 ). 127. ↵ Kassambara , A. rstatix : pipe-friendly framework for basic statistical tests . ( 2023 ). View the discussion thread. Back to top Previous Next Posted June 06, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes Juan Manuel Medina-Méndez , Paula Iruzubieta , Raúl Fernández-López , Javier Crespo , Fernando de la Cruz bioRxiv 2025.05.03.652020; doi: https://doi.org/10.1101/2025.05.03.652020 Share This Article: Copy Citation Tools Bacterial metabolic signatures in MASLD predicted through gene-centric studies in stool metagenomes Juan Manuel Medina-Méndez , Paula Iruzubieta , Raúl Fernández-López , Javier Crespo , Fernando de la Cruz bioRxiv 2025.05.03.652020; doi: https://doi.org/10.1101/2025.05.03.652020 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Microbiology Subject Areas All Articles Animal Behavior and Cognition (7642) Biochemistry (17715) Bioengineering (13907) Bioinformatics (42005) Biophysics (21472) Cancer Biology (18624) Cell Biology (25534) Clinical Trials (138) Developmental Biology (13391) Ecology (19935) Epidemiology (2067) Evolutionary Biology (24356) Genetics (15617) Genomics (22529) Immunology (17753) Microbiology (40437) Molecular Biology (17200) Neuroscience (88697) Paleontology (667) Pathology (2840) Pharmacology and Toxicology (4829) Physiology (7653) Plant Biology (15171) Scientific Communication and Education (2046) Synthetic Biology (4304) Systems Biology (9827) Zoology (2272)