Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 96 of the 96 references it cites were found in a registry or an index.

Host genetic regulation of human gut microbial structural variation

Open access
Zhernakova, Daria V., Wang, Daoming, Liu, Lei, Andreu-Sánchez, Sergio, Zhang, Yue, Ruiz-Moreno, Angel J., Peng, Haoran, Plomp, Niels, Del Castillo-Izquierdo, Ángela, Gacesa, Ranko, Lopera-Maya, Esteban A., Temba, Godfrey S., Kullaya, Vesla I., van Leeuwen, Sander S., Lifelines Cohort Study, Aguirre-Gamboa, Raul, Deelen, Patrick, Franke, Lude, Kuivenhoven, Jan A., Nolte, Ilja M., Sanna, Serena, Snieder, Harold, Swertz, Morris A., Visscher, Peter M., Vonk, Judith M., Xavier, Ramnik J., de Mast, Quirijn, Joosten, Leo A. B., Riksen, Niels P., Rutten, Joost H. W., Netea, Mihai G., Sanna, Serena, Wijmenga, Cisca, Weersma, Rinse K., Zhernakova, Alexandra, Harmsen, Hermie J. M., Fu, Jingyuan · 2024 · Nature

Abstract

Although the impact of host genetics on gut microbial diversity and the abundance of specific taxa is well established 1–6 , little is known about how host genetics regulates the genetic diversity of gut microorganisms. Here we conducted a meta-analysis of associations between human genetic variation and gut microbial structural variation in 9,015 individuals from four Dutch cohorts. Strikingly, the presence rate of a structural variation segment in Faecalibacterium prausnitzii that harbours an N -acetylgalactosamine (GalNAc) utilization gene cluster is higher in individuals who secrete the type A oligosaccharide antigen terminating in GalNAc, a feature that is jointly determined by human ABO and FUT2 genotypes, and we could replicate this association in a Tanzanian cohort. In vitro experiments demonstrated that GalNAc can be used as the sole carbohydrate source for F. prausnitzii strains that carry the GalNAc-metabolizing pathway. Further in silico and in vitro studies demonstrated that other ABO -associated species can also utilize GalNAc, particularly Collinsella aerofaciens . The GalNAc utilization genes are also associated with the host’s cardiometabolic health, particularly in individuals with mucosal A-antigen. Together, the findings of our study demonstrate that genetic associations across the human genome and bacterial metagenome can provide functional insights into the reciprocal host–microbiome relationship.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41586-023-06893-w

Cites

96 of 96 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 An overview of the workflow and microbial SV data. a, The workflow of this GWAS on gut microbial SVs. In this study we integrated data for 9,015 individuals from whom both gut microbial metagenomic and host genetic data are available from four cohorts. Data from a Tanzanian cohort of 279 individuals were included as a replication cohort. We generated gut microbial SV profiles based on metagenomic sequencing, with the dSVs and vSVs then subjected to association analysis with the genotypes of more than 6 million common SNPs in the human genome. The genetic associations are presented in box plots for vSVs, bar plots for dSVs and Manhattan plots for the whole-genome level. Created with BioRender.com. b, The number of common microbial SVs detected in 49 species for GWAS. Each bar represents a species. The y axis refers to the number of common SVs detected in that species. dSVs and vSVs are coloured in green and blue, respectively. c, A pie chart of the 3,552 common SVs, including 1,666 dSVs and 1,886 vSVs, involved in the GWAS. Source Data
Fig. 1 An overview of the workflow and microbial SV data. a, The workflow of this GWAS on gut microbial SVs. In this study we integrated data for 9,015 individuals from whom both gut microbial metagenomic and host genetic data are available from four cohorts. Data from a Tanzanian cohort of 279 individuals were included as a replication cohort. We generated gut microbial SV profiles based on metagenomic sequencing, with the dSVs and vSVs then subjected to association analysis with the genotypes of more than 6 million common SNPs in the human genome. The genetic associations are presented in box plots for vSVs, bar plots for dSVs and Manhattan plots for the whole-genome level. Created with BioRender.com. b, The number of common microbial SVs detected in 49 species for GWAS. Each bar represents a species. The y axis refers to the number of common SVs detected in that species. dSVs and vSVs are coloured in green and blue, respectively. c, A pie chart of the 3,552 common SVs, including 1,666 dSVs and 1,886 vSVs, involved in the GWAS. Source DataCC BY · PubMed Central
Fig. 2 A Manhattan plot of genome-wide associations of human SNPs and gut microbial SVs. GWAS results for dSVs (top) and vSVs (bottom). The x axis shows the genomic position on the human chromosomes (chromosomes 1–22) for both the top and bottom panels. The y axes in both panels show statistical significance as −log10[P] estimated using a linear mixed model by fastGWA. The plotted P values are not adjusted for multiple testing. The red horizontal lines indicate the study-wide significance cutoffs determined using the Bonferroni method: 3.00 × 10−11 for dSV and 2.65 × 10−11 for vSV associations. Significantly associated loci are highlighted in yellow and labelled with the nearby human gene and the corresponding species name. Source Data
Fig. 2 A Manhattan plot of genome-wide associations of human SNPs and gut microbial SVs. GWAS results for dSVs (top) and vSVs (bottom). The x axis shows the genomic position on the human chromosomes (chromosomes 1–22) for both the top and bottom panels. The y axes in both panels show statistical significance as −log10[P] estimated using a linear mixed model by fastGWA. The plotted P values are not adjusted for multiple testing. The red horizontal lines indicate the study-wide significance cutoffs determined using the Bonferroni method: 3.00 × 10−11 for dSV and 2.65 × 10−11 for vSV associations. Significantly associated loci are highlighted in yellow and labelled with the nearby human gene and the corresponding species name. Source DataCC BY · PubMed Central
Fig. 3 GalNAc utilization underlies the ABO association with F. prausnitzii SVs. a, Association between ABO blood type and the presence or absence of genomic segment 577–579 kb (SV 577–579) of F. prausnitzii is dependent on FUT2 secretor status. Individuals were grouped into different groups on the basis of blood types (A versus B or O blood types) and FUT2 secretor status (secretors versus non-secretors). The y axis refers to the fraction of individuals with the 577–579-kb region in the DMP dataset. Unadjusted association P values are reported, based on linear mixed models. b, A scheme of the GalNAc pathway identified in the associated SV region, which is divided into the initial cleavage from A-antigen (step 0) and five key steps of GalNAc utilization (steps 1–5; Supplementary Note 3). agaF + V + C+ D genes encode four subunits of the GalNAc PTS system II complex protein. DHAP, dihydroxyacetone phosphate; FBP, fructose-1,6-bisphosphate; F6P, fructose 6-phosphate; GAP, glyceraldehyde 3-phosphate; PTS, phosphotransferase system; TBP, tagatose 1,6-bisphosphate. Created with BioRender.com. c, A phylogenetic tree of the strains used for the culture experiments and organization of genes involved in GalNAc utilization in the SV region. The x axis indicates the base pair position starting from the flanking gene dinB. The different coloured lines indicate the location of the same genes in different strains. The names of strains with and without the SV region are in red and black, respectively. d,e, Growth curves of F. prausnitzii strains with or without SV 577–579 on medium with GalNAc (d) and galactose (e). The x axis refers to time points in hours. The y axis refers to cell density as measured as the optical density at 600 nm (OD600nm). Points with bars on the growth curves represent means ± s.d. of three replicates. f, Fold change of gene expression following GalNAc induction compared to glucose induction. The y axis shows the fold change of the expression of the agaC, agaD, agaV and agaF genes in the strains HTF-495 and ATCC 27768 following GalNAc induction relative to glucose induction. Each dot indicates one replicate. Bars and error bars indicate means ± s.e. of three replicates. Source Data
Fig. 3 GalNAc utilization underlies the ABO association with F. prausnitzii SVs. a, Association between ABO blood type and the presence or absence of genomic segment 577–579 kb (SV 577–579) of F. prausnitzii is dependent on FUT2 secretor status. Individuals were grouped into different groups on the basis of blood types (A versus B or O blood types) and FUT2 secretor status (secretors versus non-secretors). The y axis refers to the fraction of individuals with the 577–579-kb region in the DMP dataset. Unadjusted association P values are reported, based on linear mixed models. b, A scheme of the GalNAc pathway identified in the associated SV region, which is divided into the initial cleavage from A-antigen (step 0) and five key steps of GalNAc utilization (steps 1–5; Supplementary Note 3). agaF + V + C+ D genes encode four subunits of the GalNAc PTS system II complex protein. DHAP, dihydroxyacetone phosphate; FBP, fructose-1,6-bisphosphate; F6P, fructose 6-phosphate; GAP, glyceraldehyde 3-phosphate; PTS, phosphotransferase system; TBP, tagatose 1,6-bisphosphate. Created with BioRender.com. c, A phylogenetic tree of the strains used for the culture experiments and organization of genes involved in GalNAc utilization in the SV region. The x axis indicates the base pair position starting from the flanking gene dinB. The different coloured lines indicate the location of the same genes in different strains. The names of strains with and without the SV region are in red and black, respectively. d,e, Growth curves of F. prausnitzii strains with or without SV 577–579 on medium with GalNAc (d) and galactose (e). The x axis refers to time points in hours. The y axis refers to cell density as measured as the optical density at 600 nm (OD600nm). Points with bars on the growth curves represent means ± s.d. of three replicates. f, Fold change of gene expression following GalNAc induction compared to glucose induction. The y axis shows the fold change of the expression of the agaC, agaD, agaV and agaF genes in the strains HTF-495 and ATCC 27768 following GalNAc induction relative to glucose induction. Each dot indicates one replicate. Bars and error bars indicate means ± s.e. of three replicates. Source DataCC BY · PubMed Central
Fig. 4 GalNAc utilization capacity of strains of F. prausnitzii and other ABO-associated species. a, Completeness of the GalNAc pathway in four gut microbial species associated with human ABO blood type: F. prausnitzii, C. aerofaciens, F. lactaris and B. bifidum. The upper bar plot shows the number of strains (intersection size), and the combination of black dots in each vertical column underneath represents the presence of the genes in the corresponding step. The bar plot on the left represents the number of strains (set size) containing the genes of each corresponding GalNAc metabolism step. b,c, Proportion of strains with a complete GalNAc pathway in F. prausnitzii (b) and C. aerofaciens (c). d–g, Growth curves of C. aerofaciens strains containing GalNAc pathway genes on medium supplemented with different sugars. The x axis indicates the hours after culturing in the medium. The y axis indicates cell density measured as the OD600nm value. Points on the growth curves represent means ± s.d. of three replicates. YCFA medium and YCFA–glucose (Glc) medium were used as negative and positive controls, respectively. GalNAc and galactose (Gal) were supplied in YCFA medium to test whether C. aerofaciens can grow with the monosaccharides released from A-antigen and B-antigen as their sole carbohydrate sources. Source Data
Fig. 4 GalNAc utilization capacity of strains of F. prausnitzii and other ABO-associated species. a, Completeness of the GalNAc pathway in four gut microbial species associated with human ABO blood type: F. prausnitzii, C. aerofaciens, F. lactaris and B. bifidum. The upper bar plot shows the number of strains (intersection size), and the combination of black dots in each vertical column underneath represents the presence of the genes in the corresponding step. The bar plot on the left represents the number of strains (set size) containing the genes of each corresponding GalNAc metabolism step. b,c, Proportion of strains with a complete GalNAc pathway in F. prausnitzii (b) and C. aerofaciens (c). d–g, Growth curves of C. aerofaciens strains containing GalNAc pathway genes on medium supplemented with different sugars. The x axis indicates the hours after culturing in the medium. The y axis indicates cell density measured as the OD600nm value. Points on the growth curves represent means ± s.d. of three replicates. YCFA medium and YCFA–glucose (Glc) medium were used as negative and positive controls, respectively. GalNAc and galactose (Gal) were supplied in YCFA medium to test whether C. aerofaciens can grow with the monosaccharides released from A-antigen and B-antigen as their sole carbohydrate sources. Source DataCC BY · PubMed Central
Fig. 5 Comparison of GalNAc associations between two groups of individuals with or without mucosal A-antigen. a,b, Comparison of the abundance of GalNAc genes, lacC (a) and gatY–kbaY (b), between two groups of individuals (nwithout = 3,866 and nwith = 1,868). The violin plots show ln-transformed gene abundance with units of reads per kilobase million (RPKM). The inner box plots represent summary statistics: the centre line represents the median, the box hinges represent the lower and upper quartiles of the distribution, whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P values are reported, based on linear mixed models. c, Comparison of the correlations of GalNAc metabolism gene abundance with gut microbiome α-diversity and richness between two groups: individuals with mucosal A-antigen (x axis; n = 1,868) and those without (y axis, n = 3,866). Each dot represents a Spearman correlation coefficient between a GalNAc gene with the Shannon index (green) or richness (blue). The error bars indicate the confidence interval of the correlation R estimation based on s.e. d,e, The association between GalNAc metabolism gene abundance and host phenotypes in individuals with (d; n = 1,868) and without (e; n = 3,866) mucosal A-antigens (unadjusted P values are estimated from linear regression). Hash symbols indicate group-specific significant associations. Asterisks indicate significant associations shared by the two groups (Bonferroni-corrected P < 0.05). Exact P values are listed in Supplementary Table 19. Positive associations are in red. Negative associations are in blue. The colour gradients reflect the effect size. BMI, body mass index; BristolType, Bristol stool type; GeneralHealth, general health score; HDL, high-density lipoprotein cholesterol; TG, triglyceride levels. Source Data
Fig. 5 Comparison of GalNAc associations between two groups of individuals with or without mucosal A-antigen. a,b, Comparison of the abundance of GalNAc genes, lacC (a) and gatY–kbaY (b), between two groups of individuals (nwithout = 3,866 and nwith = 1,868). The violin plots show ln-transformed gene abundance with units of reads per kilobase million (RPKM). The inner box plots represent summary statistics: the centre line represents the median, the box hinges represent the lower and upper quartiles of the distribution, whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P values are reported, based on linear mixed models. c, Comparison of the correlations of GalNAc metabolism gene abundance with gut microbiome α-diversity and richness between two groups: individuals with mucosal A-antigen (x axis; n = 1,868) and those without (y axis, n = 3,866). Each dot represents a Spearman correlation coefficient between a GalNAc gene with the Shannon index (green) or richness (blue). The error bars indicate the confidence interval of the correlation R estimation based on s.e. d,e, The association between GalNAc metabolism gene abundance and host phenotypes in individuals with (d; n = 1,868) and without (e; n = 3,866) mucosal A-antigens (unadjusted P values are estimated from linear regression). Hash symbols indicate group-specific significant associations. Asterisks indicate significant associations shared by the two groups (Bonferroni-corrected P < 0.05). Exact P values are listed in Supplementary Table 19. Positive associations are in red. Negative associations are in blue. The colour gradients reflect the effect size. BMI, body mass index; BristolType, Bristol stool type; GeneralHealth, general health score; HDL, high-density lipoprotein cholesterol; TG, triglyceride levels. Source DataCC BY · PubMed Central
Extended Data Fig. 1 Overview of gut microbial SVs detected in the Dutch cohorts before filtering. a, Number of dSVs and vSVs detected per species. Each bar represents a species, and the y-axis shows the number of SVs detected per species. The number of dSVs and vSVs are colored in green and blue, respectively. b, Percentages of the overall number of dSVs and vSVs detected. Pie chart shows the number of dSVs and vSVs and the corresponding proportions. c, Number of samples with detected SVs per species, colored by cohort. X-axis indicates the different species. Y-axis indicates the number of samples with detectable SVs in that species, colored by different cohorts. d, Distribution of the total relative abundance of species detected with SVs per cohort. X-axis is the total abundance in terms of proportion of a microbial community. Y-axis is density. The red dashed vertical line indicates the average total relative abundance of species with sufficient coverage to detect SVs in all Dutch samples. Source Data
Extended Data Fig. 1 Overview of gut microbial SVs detected in the Dutch cohorts before filtering. a, Number of dSVs and vSVs detected per species. Each bar represents a species, and the y-axis shows the number of SVs detected per species. The number of dSVs and vSVs are colored in green and blue, respectively. b, Percentages of the overall number of dSVs and vSVs detected. Pie chart shows the number of dSVs and vSVs and the corresponding proportions. c, Number of samples with detected SVs per species, colored by cohort. X-axis indicates the different species. Y-axis indicates the number of samples with detectable SVs in that species, colored by different cohorts. d, Distribution of the total relative abundance of species detected with SVs per cohort. X-axis is the total abundance in terms of proportion of a microbial community. Y-axis is density. The red dashed vertical line indicates the average total relative abundance of species with sufficient coverage to detect SVs in all Dutch samples. Source DataCC BY · PubMed Central
Extended Data Fig. 2 Heritability of gut microbial SVs and corresponding species abundance. Bar height indicates heritability h2 values for vSVs (blue) and dSVs (cyan) that are nominally significant (P < 0.05), estimated using GCTA. Error bars in gray are the 95% confidence intervals (1.96 × standard error) of the estimated heritability. Solid red vertical lines indicate the heritability of the corresponding species abundance previously reported in Gacesa et al. Nature (2020), and the red opaque rectangles indicate the corresponding 95% confidence interval. Species names marked with * indicate those with significant abundance heritability at P < 0.05 level. The number of samples used for SV heritability estimation is given on the right. Source Data
Extended Data Fig. 2 Heritability of gut microbial SVs and corresponding species abundance. Bar height indicates heritability h2 values for vSVs (blue) and dSVs (cyan) that are nominally significant (P < 0.05), estimated using GCTA. Error bars in gray are the 95% confidence intervals (1.96 × standard error) of the estimated heritability. Solid red vertical lines indicate the heritability of the corresponding species abundance previously reported in Gacesa et al. Nature (2020), and the red opaque rectangles indicate the corresponding 95% confidence interval. Species names marked with * indicate those with significant abundance heritability at P < 0.05 level. The number of samples used for SV heritability estimation is given on the right. Source DataCC BY · PubMed Central
Extended Data Fig. 3 Comparison of F. prausnitzii dSV associations between the Dutch and Tanzanian cohorts. a, Q-Q plots of the associations between human SNPs and F. prausnitzii SVs in the Dutch cohorts. X-axis is the expected log-transformed P-value. Y-axis is the observed log-transformed unadjusted P-value derived from the linear mixed model. Each dot represents an association P-value. The diagonal is represented by the solid red line. The lambda (λ) value represents the genomic inflation factor. b, Comparison of the presence/absence ratio of dSVs between the DMP cohort (N = 5044) and the Tanzanian cohort (N = 201). Bar plots show the fraction of samples with (dark blue) or without (cyan) the dSV region. c, Association of F. prausnitzii dSV 575–577 with rs550057 in both the Dutch and Tanzanian cohorts. Bar plots show the number of samples with (dark blue) and without (cyan) the dSV region per genotype. The P-value indicates the unadjusted association significance assessed by fastGWA using linear mixed models. d, Association of F. prausnitzii 577–579 with rs1633513 in both cohorts. Bar plots show the number of samples with (dark blue) and without (cyan) the dSV region per genotype. The P-values indicate the unadjusted association significance assessed by fastGWA based using linear mixed models. Source Data
Extended Data Fig. 3 Comparison of F. prausnitzii dSV associations between the Dutch and Tanzanian cohorts. a, Q-Q plots of the associations between human SNPs and F. prausnitzii SVs in the Dutch cohorts. X-axis is the expected log-transformed P-value. Y-axis is the observed log-transformed unadjusted P-value derived from the linear mixed model. Each dot represents an association P-value. The diagonal is represented by the solid red line. The lambda (λ) value represents the genomic inflation factor. b, Comparison of the presence/absence ratio of dSVs between the DMP cohort (N = 5044) and the Tanzanian cohort (N = 201). Bar plots show the fraction of samples with (dark blue) or without (cyan) the dSV region. c, Association of F. prausnitzii dSV 575–577 with rs550057 in both the Dutch and Tanzanian cohorts. Bar plots show the number of samples with (dark blue) and without (cyan) the dSV region per genotype. The P-value indicates the unadjusted association significance assessed by fastGWA using linear mixed models. d, Association of F. prausnitzii 577–579 with rs1633513 in both cohorts. Bar plots show the number of samples with (dark blue) and without (cyan) the dSV region per genotype. The P-values indicate the unadjusted association significance assessed by fastGWA based using linear mixed models. Source DataCC BY · PubMed Central
Extended Data Fig. 4 The associations of F. prausnitzii SVs with ABO blood groups depend on FUT secretor status. Each row of panels represents a different SV region. Each column of panels represents a different cohort. The associations of dSVs are visualized using bar plots where the y-axis shows the fraction of samples with the dSV region. Number of samples for each group is given at the bottom of each bar. The associations of vSVs are visualized using violin plots where the y-axis shows the standardized coverage of vSVs. Violin plots show density distribution, whereas the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, the whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. The x-axis of all plots shows the ABO blood group (A/AB or B/O) split by FUT2 status. Unadjusted P-values show the significance of the association estimated using linear mixed models. Source Data
Extended Data Fig. 4 The associations of F. prausnitzii SVs with ABO blood groups depend on FUT secretor status. Each row of panels represents a different SV region. Each column of panels represents a different cohort. The associations of dSVs are visualized using bar plots where the y-axis shows the fraction of samples with the dSV region. Number of samples for each group is given at the bottom of each bar. The associations of vSVs are visualized using violin plots where the y-axis shows the standardized coverage of vSVs. Violin plots show density distribution, whereas the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, the whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. The x-axis of all plots shows the ABO blood group (A/AB or B/O) split by FUT2 status. Unadjusted P-values show the significance of the association estimated using linear mixed models. Source DataCC BY · PubMed Central
Extended Data Fig. 5 Population genetic structure of F. prausnitzii and its association with ABO-associated SVs. a, The x-axis and y-axis show the top two PCs. Each dot represents a sample. In each panel figure, dots are colored differently based on the presence/absence status of dSVs or the standardized coverage of vSVs following the color key shown in the figure. The significance of associations between population genetic structure of F. prausnitzii and SVs was estimated by permutational multivariate analysis of variance (PERMANOVA), and all associations for the five ABO-associated SVs are significant (unadjusted P < 0.05). b, Association of F. prausnitzii abundance with ABO blood groups in the DMP cohort. Species abundance (CLR-transformed) of F. prausnitzii (y-axis) plotted against ABO-encoded blood groups for all DMP individuals (x-axis). Violin plots show density distribution, whereas the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, the whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P-value reflects the significance of association of ABO blood groups (A/AB vs B/O) with F. prausnitzii abundance after adjusting for covariates assessed using linear mixed models. Source Data
Extended Data Fig. 5 Population genetic structure of F. prausnitzii and its association with ABO-associated SVs. a, The x-axis and y-axis show the top two PCs. Each dot represents a sample. In each panel figure, dots are colored differently based on the presence/absence status of dSVs or the standardized coverage of vSVs following the color key shown in the figure. The significance of associations between population genetic structure of F. prausnitzii and SVs was estimated by permutational multivariate analysis of variance (PERMANOVA), and all associations for the five ABO-associated SVs are significant (unadjusted P < 0.05). b, Association of F. prausnitzii abundance with ABO blood groups in the DMP cohort. Species abundance (CLR-transformed) of F. prausnitzii (y-axis) plotted against ABO-encoded blood groups for all DMP individuals (x-axis). Violin plots show density distribution, whereas the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, the whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P-value reflects the significance of association of ABO blood groups (A/AB vs B/O) with F. prausnitzii abundance after adjusting for covariates assessed using linear mixed models. Source DataCC BY · PubMed Central
Extended Data Fig. 6 Transmission of F. prausnitzii SVs. a, Different phylogenetic distances between F. prausnitzii strains based on marker genes and GalNAc genes. Phylogenetic trees based on the phylogenetic distance of marker genes (left) and GalNAc genes (right). b, Scatter plot of between-strain genetic distance based on marker genes (x-axis) and GalNAc genes (y-axis). Each dot represents pairwise phylogenetic distance between two strains. c, Comparison of between-strain genetic distance based on marker genes and GalNAc genes. Each box plot represents the distribution of pairwise phylogenetic distances of strains based either on GalNAc genes or on species marker genes. The center line of the box is the median, the box hinges are the lower and upper quartiles of the distribution, and each dot represents a pair of two strains (npair = 21). d, Fraction of samples with or without SV 577–579 dependent on household members. Each blue bar represents the fraction of samples with the GalNAc-containing 577–579 dSV region in individuals with/without A-antigen and with/without household members with the dSV. e, Gain or loss of the F. prausnitzii dSV 577–579 region after 4 years. Alluvial plot shows the number of individuals with and without the 577–579 region in F. prausnitzii at two time-points 4-years apart. The number of individuals with the same presence or absence status for this region after 4 years is indicated in gray. The number of individuals who gained the region is indicated in red. The number of those who lost the region is indicated in blue.
Extended Data Fig. 6 Transmission of F. prausnitzii SVs. a, Different phylogenetic distances between F. prausnitzii strains based on marker genes and GalNAc genes. Phylogenetic trees based on the phylogenetic distance of marker genes (left) and GalNAc genes (right). b, Scatter plot of between-strain genetic distance based on marker genes (x-axis) and GalNAc genes (y-axis). Each dot represents pairwise phylogenetic distance between two strains. c, Comparison of between-strain genetic distance based on marker genes and GalNAc genes. Each box plot represents the distribution of pairwise phylogenetic distances of strains based either on GalNAc genes or on species marker genes. The center line of the box is the median, the box hinges are the lower and upper quartiles of the distribution, and each dot represents a pair of two strains (npair = 21). d, Fraction of samples with or without SV 577–579 dependent on household members. Each blue bar represents the fraction of samples with the GalNAc-containing 577–579 dSV region in individuals with/without A-antigen and with/without household members with the dSV. e, Gain or loss of the F. prausnitzii dSV 577–579 region after 4 years. Alluvial plot shows the number of individuals with and without the 577–579 region in F. prausnitzii at two time-points 4-years apart. The number of individuals with the same presence or absence status for this region after 4 years is indicated in gray. The number of individuals who gained the region is indicated in red. The number of those who lost the region is indicated in blue.CC BY · PubMed Central
Extended Data Fig. 7 Growth curves for F. prausnitzii strains in media supplemented with different sugars. a, Growth curve of F. prausnitzii strain HTF−383 in GalNAc medium over a longer growth time (40 h). X-axis refers to culturing time (in hours). Y-axis refers to the cell density measured as OD600 value. The red points with bars on the growth curve represent the means ± standard deviation of three replicates. b, Red lines indicate the growth curves of strains with SV 577–579. Black lines indicate the growth curve of strains without this SV. YCFA is Yeast Casitone Fatty Acids (YCFA) basic medium without supplemental sugars. The other graphs are for the YCFA medium supplemented with the sugar indicated: glucose (Glc), lactose (Lac), mannose (Man), N-acetylglucosamine (GlcNAc), fructose (Fruc), N-acetylneuraminic acid (NeuNAc) and 2-fucosyllactose (2FL). X-axis is the culturing time (in hours). Y-axis is the cell density measured as OD600 value. The red and black points with bars on the growth curves represent the means ± standard deviation of three replicates. Source Data
Extended Data Fig. 7 Growth curves for F. prausnitzii strains in media supplemented with different sugars. a, Growth curve of F. prausnitzii strain HTF−383 in GalNAc medium over a longer growth time (40 h). X-axis refers to culturing time (in hours). Y-axis refers to the cell density measured as OD600 value. The red points with bars on the growth curve represent the means ± standard deviation of three replicates. b, Red lines indicate the growth curves of strains with SV 577–579. Black lines indicate the growth curve of strains without this SV. YCFA is Yeast Casitone Fatty Acids (YCFA) basic medium without supplemental sugars. The other graphs are for the YCFA medium supplemented with the sugar indicated: glucose (Glc), lactose (Lac), mannose (Man), N-acetylglucosamine (GlcNAc), fructose (Fruc), N-acetylneuraminic acid (NeuNAc) and 2-fucosyllactose (2FL). X-axis is the culturing time (in hours). Y-axis is the cell density measured as OD600 value. The red and black points with bars on the growth curves represent the means ± standard deviation of three replicates. Source DataCC BY · PubMed Central
Extended Data Fig. 8 Comparison of GalNAc-induced expression fold changes in GalNAc genes between HTF-495 and ATCC27768. Bar plots represent the average fold changes of gene expression induced by GalNAc compared to glucose induction. Each dot represents a sample, and the bar indicates the standard error of three replicates. The light blue and dark blue bars represent the expression fold change of glucose induction and GalNAc induction in relation to glucose induction, respectively. Thus, the fold change in glucose was set to 1. Unadjusted P-values estimated using the two-sided t-test are given for each gene. * Indicates a difference significant at unadjusted P < 0.05 level. Source Data
Extended Data Fig. 8 Comparison of GalNAc-induced expression fold changes in GalNAc genes between HTF-495 and ATCC27768. Bar plots represent the average fold changes of gene expression induced by GalNAc compared to glucose induction. Each dot represents a sample, and the bar indicates the standard error of three replicates. The light blue and dark blue bars represent the expression fold change of glucose induction and GalNAc induction in relation to glucose induction, respectively. Thus, the fold change in glucose was set to 1. Unadjusted P-values estimated using the two-sided t-test are given for each gene. * Indicates a difference significant at unadjusted P < 0.05 level. Source DataCC BY · PubMed Central
Extended Data Fig. 9 Association of nine key GalNAc pathway genes with A-antigen presence in FUT2 secretors and non-secretors. Y-axis of the violin plots is the gene abundance in the DMP cohort (n = 1868, 2496, 610 and 760 for FUT2 secretor A/AB, FUT2 secretor B/O, FUT2 non-secretor A/AB and FUT2 non-secretor B/O, respectively). The RPKM value of gene abundance (number of reads per kilobase of transcript per million reads mapped) was profiled using ShortBRED. The y-axis unit ln(RPKM) stands for normalized gene abundance after ln(RPKM + 1) transformation, adjusted for covariates (age, sex, total read number, species abundance and the five F. prausnitzii SVs associated with ABO). The violin plots for FUT2 secretors are colored in blue. Plots for FUT2 non-secretors are in gray. Violin plots show density distribution, while the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P-values show the significance of the association between the ABO blood group and gene abundance while adjusting for covariates estimated using linear mixed models. Source Data
Extended Data Fig. 9 Association of nine key GalNAc pathway genes with A-antigen presence in FUT2 secretors and non-secretors. Y-axis of the violin plots is the gene abundance in the DMP cohort (n = 1868, 2496, 610 and 760 for FUT2 secretor A/AB, FUT2 secretor B/O, FUT2 non-secretor A/AB and FUT2 non-secretor B/O, respectively). The RPKM value of gene abundance (number of reads per kilobase of transcript per million reads mapped) was profiled using ShortBRED. The y-axis unit ln(RPKM) stands for normalized gene abundance after ln(RPKM + 1) transformation, adjusted for covariates (age, sex, total read number, species abundance and the five F. prausnitzii SVs associated with ABO). The violin plots for FUT2 secretors are colored in blue. Plots for FUT2 non-secretors are in gray. Violin plots show density distribution, while the inner boxplots represent summary statistics: the center line is the median, the box hinges are the lower and upper quartiles of the distribution, whiskers extend no further than 1.5× interquartile range from the hinges, and data beyond the end of the whiskers are outliers plotted as individual points. Unadjusted P-values show the significance of the association between the ABO blood group and gene abundance while adjusting for covariates estimated using linear mixed models. Source DataCC BY · PubMed Central
Extended Data Fig. 10 Association of GalNAc pathway genes with gut microbiome diversity and host phenotypes. a, Community-level GalNAc utilization capacity is positively associated with gut microbiome diversity and richness. Bar plot shows the Spearman correlation coefficients (R) between the abundance of GalNAc genes in a microbial community with the alpha diversity (Shannon index) (top panel) and richness (bottom panel) of the community for individuals with (red, n = 1,868) and without (blue, n = 3,866) mucosal A-antigen. Y-axis is the Spearman correlation coefficient. X-axis indicates different GalNAc genes. b, Heterogeneity of associations between gut microbial GalNAc utilization genes and human phenotypes in individuals with and without mucosal A-antigen. Scatter plot shows the association effect size (standardized beta-coefficient from linear regression) between GalNAc metabolism gene abundance and host phenotypes in individuals with mucosal A-antigen (x-axis) and those without (y-axis). Error bars indicate the confidence interval of the beta-coefficient estimation. The associations between GalNAc metabolism gene abundance and host phenotypes are significantly higher in individuals with mucosal A-antigen (n = 1,868) compared to those without (n = 3,866) (Unadjusted Pheterogeneity < 0.05; Cochran’s Q test). Dots are colored differently for different phenotypes. Source Data
Extended Data Fig. 10 Association of GalNAc pathway genes with gut microbiome diversity and host phenotypes. a, Community-level GalNAc utilization capacity is positively associated with gut microbiome diversity and richness. Bar plot shows the Spearman correlation coefficients (R) between the abundance of GalNAc genes in a microbial community with the alpha diversity (Shannon index) (top panel) and richness (bottom panel) of the community for individuals with (red, n = 1,868) and without (blue, n = 3,866) mucosal A-antigen. Y-axis is the Spearman correlation coefficient. X-axis indicates different GalNAc genes. b, Heterogeneity of associations between gut microbial GalNAc utilization genes and human phenotypes in individuals with and without mucosal A-antigen. Scatter plot shows the association effect size (standardized beta-coefficient from linear regression) between GalNAc metabolism gene abundance and host phenotypes in individuals with mucosal A-antigen (x-axis) and those without (y-axis). Error bars indicate the confidence interval of the beta-coefficient estimation. The associations between GalNAc metabolism gene abundance and host phenotypes are significantly higher in individuals with mucosal A-antigen (n = 1,868) compared to those without (n = 3,866) (Unadjusted Pheterogeneity < 0.05; Cochran’s Q test). Dots are colored differently for different phenotypes. Source DataCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 96 references

Research works printed but not linked: 1

  1. Schwengers, O. et al. Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microb. Genom. 7, 000685 (2021).

Cited by

See all 174

Metadata

Authors
Zhernakova, D. V., Wang, D., Liu, L., Andreu-Sánchez, S., Zhang, Y., Ruiz-Moreno, A. J., Peng, H., Plomp, N., Del Castillo-Izquierdo, Á., Gacesa, R., Lopera-Maya, E. A., Temba, G. S., Kullaya, V. I., van Leeuwen, S. S., Study, L. C., Aguirre-Gamboa, R., Deelen, P., Franke, L., Kuivenhoven, J. A., Nolte, I. M., Sanna, S., Snieder, H., Swertz, M. A., Visscher, P. M., Vonk, J. M., Xavier, R. J., de Mast, Q., Joosten, L. A. B., Riksen, N. P., Rutten, J. H. W., Netea, M. G., Sanna, S., Wijmenga, C., Weersma, R. K., Zhernakova, A., Harmsen, H. J. M., & Fu, J.
Affiliations
  • University Medical Center Groningen (NL)
  • Radboud University Medical Center (NL)
  • Kilimanjaro Christian Medical Centre (TZ)
  • Broad Institute (US)
  • Iuliu Hațieganu University of Medicine and Pharmacy (RO)
  • University of Bonn (DE)
Corresponding author
  • Harmsen, Hermie J. M.
  • Fu, Jingyuan
Venue
Nature
Published
2024-01-03
Volume
625
Issue
7996
Pages
813-821
ISSN
  • 0028-0836 (ISSN-L)
  • 1476-4687
OpenAlex
W4390570973
Cited by
174
References
96
Citations by year
2026: 59 · 2025: 78 · 2024: 36
FWCI
7.07
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (submittedVersion, cc-by)
  • Radboud Repository (Radboud University) (submittedVersion)
Funders
Nederlandse Organisatie voor Wetenschappelijk Onderzoek (VI.Veni.194.006, 024.003.001, VI.C.202.022, 016.178.056, 016.Vidi.178.056, 024.004.017, VI.Veni.222.016)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gut microbiota and health
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Nutrition, Genetics, and Disease
Keywords
host genetics, gut microbiota, Faecalibacterium prausnitzii, ABO blood group system, FUT2 genotype, Collinsella aerofaciens, host-microbiome interactions, cardiometabolic health
MeSH terms
ABO Blood-Group System, Gastrointestinal Microbiome, Faecalibacterium prausnitzii, Host Microbial Interactions, Galactoside 2-alpha-L-fucosyltransferase, Acetylgalactosamine, Bacteria, Computer Simulation, Galactosyltransferases, Gene Expression Regulation, Multigene Family, Genotype, Humans, Netherlands, Tanzania, Genetic Variation, Cohort Studies, Genome, Human, Actinobacteria, Metagenome, Microbiota, In Vitro Techniques

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Host genetic regulation of human gut microbial structural variation. https://app.cochat.ai/papers/10.1038/s41586-023-06893-w
The paper
Zhernakova, D. V., Wang, D., Liu, L., Andreu-Sánchez, S., Zhang, Y., Ruiz-Moreno, A. J., Peng, H., Plomp, N., Del Castillo-Izquierdo, Á., Gacesa, R., Lopera-Maya, E. A., Temba, G. S., Kullaya, V. I., van Leeuwen, S. S., Study, L. C., Aguirre-Gamboa, R., Deelen, P., Franke, L., Kuivenhoven, J. A., . . . Fu, J. (2024). Host genetic regulation of human gut microbial structural variation. Nature. https://doi.org/10.1038/s41586-023-06893-w

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.