Tracing NAD+ metabolism uncovers adaptive coordination between host and microbiome during colitis

Abstract

Host-microbiota metabolic interactions critically regulate nicotinamide adenine dinucleotide (NAD⁺) homeostasis, and their disruption is increasingly linked to chronic diseases including inflammatory bowel disease (IBD). However, it remains unclear whether NAD⁺ dysregulation in IBD arises from impaired production, enhanced consumption, or both. Using multi-omics approaches and stable isotope-labeled NAD⁺ precursors administered via intravenous infusion in a murine model of dextran sulfate sodium (DSS)-induced colitis, we mapped tissue- and lumen-specific NAD⁺ metabolism under inflammatory stress. Our results reveal tissue-specific rewiring of NAD⁺ metabolism, with increased flux through the salvage pathway compensating for reduced de novo NAD⁺ synthesis from tryptophan. In parallel, microbial de novo NAD⁺ production was elevated, highlighting a cooperative host-microbiota response to inflammatory stress. These findings demonstrate differential regulation of NAD⁺ biosynthesis during acute colitis and underscore the dynamic interplay between host and microbial metabolism in maintaining NAD⁺ homeostasis under inflammatory conditions.

Keywords: DSS colitis; IBD; NAD+; gut microbiota; nicotinamide; tryptophan.

Fig.1|. Acute-induced colitis decreases NAD⁺ levels in colonic tissues.

(a) Experimental setup illustrating the administration of 2.5% DSS dissolved in water for 5 days, followed by a switch to normal drinking water until the end of the experiment (day 11) in 11–12 weeks old C57BL/6 male mice. Longitudinal sampling of fecal and blood samples (shown in black arrows) was collected for LC-MS metabolomics and metagenomics analysis. At the end of the experiment mice were sacrificed by cervical dislocation, tissues and luminal samples were dissected for LC-MS metabolomics. (b) Daily recorded disease activity index (DAI) in DSS-treated and control mice including weight loss compared to the initial weight, stool consistency, bloody stool, rectal bleeding, overall activity, posture, and fur. (c) Representative image of the H&E staining of Swiss roll sections of the distal colon segment (at magnification 20x, scale bar 100μm), and (d) a corresponding histopathological score. (e) Fecal lcn-2 was measured by ELISA in fecal samples collected on days 8 and 11 during the flare phase of intestinal inflammation and controls. LC-MS relative total ion count (TIC) levels of NAD⁺ pooled from all experiments in this study of (f) tissues NAD⁺ levels and (g) luminal NAD⁺ levels. Data are presented as mean ± SEM; in (f) (n= 18–40) and in (g) (n=20–40). Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups and Mann-Whitney U test for two-groups comparisons. § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (a) was created with Biorender.com

Fig.2|. Enhanced NAD⁺ production from tryptophan during acute DSS-induced intestinal inflammation.

(a) Experimental schematic of intravenous infusion of universally labeled-tryptophan (¹³C₁₁-Trp) into pre-catheterized male C57BL/6 mice (11–12 weeks old) to assess NAD⁺ flux from tryptophan in host tissues and luminal samples. Tissue fractional labeling following 20-hour intravenous infusion of (b) kynurenine, (c) 3-hydroxyanthranilic acid, (d) quinolinic acid, (e) NAD⁺ and (f) nicotinamide. (g) Serum fractional labeling of nicotinamide over 20-hour intravenous infusion during the active flare phase of intestinal inflammation. mRNA expression of genes involved in de novo NAD⁺ synthesis from tryptophan via the kynurenine pathway was measured by qRT-PCR and normalized to TATA-box binding protein (Tbp) in (h) distal colon; indoleamine 2,3-dioxygenase 1(Ido1) (i) liver; indoleamine 2,3-dioxygenase 2 (Ido2) and tryptophan 2,3-dioxygenase (Tdo2), during the active flare phase (days 8 and 11) compared to untreated mice. Data are presented as mean ± SEM, (n=10–20) in b-g and (n=4–5) in h-j. Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups and Mann-Whitney U test for two-groups comparisons. NS, not significant, § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (a) was created with Biorender.com

Fig.3|. Elevated flux of tryptophan to luminal NAD⁺ during the active flare phase of acute DSS-induced colitis.

LC-MS luminal fraction of metabolites labeled from the tryptophan tracer (shown in figure 2) following 20-hour intravenous infusion of (a) tryptophan, (b) 3-hydroxyanthranilic acid, and (c) NAD⁺. Data are presented as mean ± SEM, (n= 10–20). Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups. NS, not significant, § <0.1, *P<0.05, **P<0.01,***P<0.001, and ****P<0.0001.

Fig.4|. Integrative pathway and taxa analysis of the gut microbiota alterations in acute colitis

(a) Bray–Curtis PCoA of species-level profiles shows clear separation between control (gray) and DSS (red) groups during the active phase of colitis (days 8 and 11) (PERMANOVA, R² = 0.216, p = 0.002), indicating significant remodeling of the microbial functional potential during DSS-induced colitis. (b) Species-level heatmap of the top 50 most variable taxa, clustered by relative abundance Z-scores across samples. Hierarchical clustering reveals distinct community compositions between groups, with several Clostridium and Lachnospiraceae species enriched in DSS-treated mice during the active flare phase. (c) Volcano plot of differential species abundance between DSS and control groups. 33 species were significantly altered (FDR < 0.1, |log₂FC| > 0.5), including 11 species enriched in DSS-treated group and 22 reduced relative to control. (d) Heatmap of significantly altered metabolic pathways (FDR<0.1). DSS-induced colitis group shows enrichment of lipid metabolism and NAD⁺ salvage pathway, and depletion of amino acid and purine biosynthetic pathways, suggesting a shift toward energy metabolism and stress-adaptive functions. (e) Predicted metabolite–taxon network linking 20 microbial taxa to 9 tryptophan-related metabolites via 110 high-confidence edges. (f) Altered tryptophan-metabolizing taxa including four differentially abundant species (Lachnospiraceae spp, Adlercreutzia equolifaciens, Ruminococcus gauvreauii, Parabacteroides goldsteinii), potentially modulating tryptophan metabolism during acute intestinal inflammation.

Fig.5|. Acute colitis alters tryptophan-dependent pathways in the host tissues and gut microbes.

(a) Overview of tryptophan metabolism in host tissues and gut microbiota. Metabolites shown in bold are compounds measured in the study. The fractional labeling of (¹³C₁₁-tryptophan) over 20-hour intravenous infusion (as in figure 2), measured by LC-MS in different tissues for (b) anthranilic acid, (c) serotonin; and luminal regions of the gastrointestinal tract: (d) indole-propionic acid, (e) indole-acetic acid, (f) tryptamine, (g) quinaldic acid. Data are presented as mean ± SEM, (n= 10–20). Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups. NS, not significant, § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (a) was created with Biorender.com

Fig.6|. Inflammation enhances NAD⁺ production from nicotinamide via the salvage pathway in a tissue-specific manner.

(a) Overview schematic of experimental setup for intravenous infusion of deuterium-labeled nicotinamide (2,4,5,6-²H₄-NAM) into pre-catheterized male C57BL/6 mice (11–12 weeks old) to assess NAD⁺ flux from nicotinamide (NAM) via the salvage pathway in host tissues. Tissue fractional labeling following 20-hour intravenous infusion of (b) NAD⁺ and (c) nicotinamide. (d) Fractional labeling of nicotinamide in the serum post 20-hour intravenous infusion during the flare phase of intestinal inflammation (M+0: unlabeled NAM, M+3: recycled NAM, M+4: infusate NAM). (e) qRT-PCR analysis of mRNA expression normalized to Tbp in distal colon and liver tissues during the active flare phase (days 8 and 11) of nicotinamide phosphoribosyltransferase (NAMPT) compared to control. (f) Fractional labeling of methylnicotinamide in host tissues. qRT-PCR analysis of mRNA expression normalized to Tbp in distal colon and liver tissues during the active flare phase (days 8 and 11) compared to control for (g) nicotinamide N-methyltransferase (NNMT), (h) nicotinamide mononucleotide adenylyltransferase 1 (NMNAT1; nuclear), and (i) nicotinamide mononucleotide adenylyltransferase 3 (NMNAT 3; mitochondrial). Data are presented as mean ± SEM, in b-d and f (n=8–18), in e,g,h (n=4–5). Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups and Mann-Whitney U test for two-groups comparisons. NS, not significant, § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (a) was created with Biorender.com

Fig.7|. Acute intestinal inflammation does not compromise the metabolic cycling of NAD⁺ precursors between the host tissues and gut microbiota.

(a) Schematic of the metabolic exchange of NAD⁺ precursors nicotinamide (NAM) and nicotinic acid (NA) between the host tissues and gut lumen. Host-derived NAM enters the gut lumen and contributes to microbial NAD⁺ either through conversion to NA and subsequently to NAD⁺, which supports both host and microbial NAD⁺ biosynthesis, or is directly converted to NAD⁺. Complex carbohydrates also contribute to microbial NAD⁺. LC-MS measurements of luminal microbial metabolites in the NAD⁺ salvage pathway after 20-hour intravenous infusion of the nicotinamide tracer (described in figure 6); (b) luminal nicotinamide, (c) luminal nicotinic acid, (d) luminal NAD⁺. Data are presented as mean ± SEM (n=8–18). Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups. § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (a) was created with Biorender.com

Fig.8|. Acute colitis triggers systemic metabolic adaptation to restore NAD⁺ levels through activation of the salvage pathway.

(a) Percent contributions of tryptophan and nicotinamide to (a) host NAD⁺ levels in different tissues and (b) luminal NAD⁺, across different phases of intestinal inflammation. In (a) and (b), the contribution of nicotinamide (M+4) to NAD⁺ (M+3) was calculated as [fraction of labeled NAD⁺ (M+3)/ the sum of (NAM+3 and NAM+4) in serum]. Tryptophan (M+11) contribution to tissue NAD⁺ (M+6) was quantified as [fraction of labeled NAD⁺ (M+6)/ fraction of labeled Trp (M+11) in the serum), following 20-hour intravenous infusion of the labeled-NAD⁺ precursors (depicted in figure 2 and 6). (c) Summary of the compensatory metabolic adaptation of host and microbial NAD⁺ metabolism during experimental acute colitis. Data are presented as mean ± SEM. Statistical significance was determined by Kruskal-Wallis test with post hoc Dunn’s test for comparisons among more than two groups. § <0.1, *P<0.05, **P<0.01, ***P<0.001, and ****P<0.0001. (c) was created with Biorender.com