Characteristics of the animal model and overview of intestinal RNAseq analysis
To examine the impact of chronic EtOH administration and elevated levels of endogenous n3 PUFAs on the intestinal transcriptome, we performed RNAseq analysis and compared the transcriptomes of intestinal epithelial tissue obtained from WT and fat-1 mice subjected to control (pair-fed mice, PF) or EtOH-containing diets (EtOH-fed mice, see Supplemental Fig. S1a for study design). Animal daily food consumption and body weights can be found in Supplemental Fig. S1b,c, respectively. While there were no differences in food consumption between genotypes, as well as between PF and EtOH-fed animals, WT EtOH, WT PF, and fat-1 PF mice gained body weight by the end of the experiment, whereas fat-1 EtOH mice did not. Compared to WT, fat-1 animals exhibited elevated intestinal tissue n3 PUFA levels in both PF and EtOH-fed mice (see ref.14 and Supplemental Fig. S1d). There were no differences in n6 PUFAs between WT and fat-1 control PF mice; however, elevated intestinal n6 PUFA levels were observed in response to EtOH in WT but not in fat-1 littermates. In addition, the ratio of n6/n3 PUFAs was significantly lower in fat-1 compared to WT mice in both PF and EtOH-fed groups.
Intestinal RNAseq analysis performed on these mice revealed that both the genotype and EtOH administration caused alterations in the intestinal transcriptome; the full list of significant differentially expressed genes (DEGs) can be found in Supplemental Materials Excel File S1. We next applied a cutoff of ≥ two-fold change for the DEGs used in Volcano Plots, GO Processes, and STRING Cluster Analyses. This threshold was used to identify robust gene expression changes and the accompanying biological processes. There were 630 DEGs between fat-1 PF and WT PF animals (420 up and 210 down). Of those genes, 276 DEGs between fat-1 PF and WT PF animals met the two-fold change threshold for further analysis (193 up and 83 down). EtOH administration resulted in 1144 DEGs (481 up and 663 down) in the WT EtOH vs WT PF group, and of these genes, 439 met the two-fold change threshold (163 up and 276 down). While 2107 genes were differentially expressed in fat-1 EtOH vs fat-1 PF mice (1045 up and 1062 down), 1501 met the two-fold change threshold for further analyses (812 up and 689 down). 94 genes were significantly different (80 up and 14 down) in the WT EtOH group when compared to fat-1 EtOH mice (all of these DEGs met the two-fold change threshold).
Differentially expressed genes between fat-1 and WT control pair-fed animals
The most prominent changes among genes differentially expressed in fat-1 PF compared to WT PF mice (≥ two-fold change, Fig. 1a and Supplementary Table S1) were observed for Onecut2, Reg1, Igkv4-80, Cym, Ighv5-12, and Afp (increased expression), and Defa-rs7, Igk4-51, Slc6a14, and Ighv8-8, (decreased expression). In order to best classify fat-1 PF vs WT PF mice with respect to all DEGs that met the two-fold threshold, we performed Gene Ontology (GO) pathway analysis in CytoScape. The top enriched and diminished pathways are shown in Fig. 1b. Next, gene clusters identified by STRING were used to visualize gene interactions and identify processes affected by the endogenous increase in n3 PUFAs in fat-1 compared to WT mice. The gene and protein interactions inferred from differentially expressed transcripts between fat-1 PF and WT PF included down-regulated gene networks (Fig. 1c) for leukocyte tethering (Chst8, Gcnt1, Sell), inflammation (Gsdmc2, Gsdmc4, Pla2g4c), immunity (Saa2, Saa1, Lcn2), bile acid metabolism (Slc10a2, Hao2, Baat), development (Hoxb9, Hoxb6, Hoxb8), oxoacid metabolism (Scd2, Suox2, Fa2h), and amino acid transport (Slc6a14). Up-regulated gene networks (Fig. 1d) included metabolism of steroids/xenobiotics (Cyp2c65, Cyp2b10, Cyp2b66), glycolysis and gluconeogenesis (Fbp1, G6pc, Pck1), glutathione metabolism (Gstt1, Gstm3, Gsta1), peroxisome proliferator-activated receptor (PPAR) signaling (Apoa2, Fabp1, Acaa1b), immunity (Igkv4-80, Ighv5-12), digestion (Reg1, Cym), and transcription (Onecut2).
EtOH-mediated alterations in the intestinal transcriptome: similarity in transcriptional responses between WT and fat-1 mice
EtOH administration resulted in substantial alterations in intestinal gene expression. Notably, there were almost 2 times more genes differentially expressed in fat-1 mice compared to WT littermates in response to EtOH (2107 DEGs vs 1144 DEGs, respectively), suggesting that the alterations in the associated lipid profile and, possibly, lipid homeostasis in fat-1 mice resulted in elevated gene transcription changes in response to EtOH challenge. Further analysis revealed a set of 835 genes in WT and fat-1 mice with similarly changed expression in response to EtOH (321 ≥ two-fold change), while 1272 and 309 genes were exclusively changed by EtOH in fat-1 and WT mice, respectively (180 and 118 ≥ two-fold change). Volcano plot analysis of DEGs that met the two-fold threshold in WT-EtOH relative to WT-PF (Fig. 2a) and in fat-1 EtOH relative to fat-1 PF (Fig. 2b) demonstrated that the top down-regulated genes shared by both genotypes included Lct, Cyp2b10, Ugt2a3, Gata4, and Plb1. Among the genes up-regulated by EtOH in both WT and fat-1 mice were Cyp2d34, Retnlb, Slc6a14, and Fa2h. The top DEGs that changed their expression in response to EtOH in WT and fat-1 mice, as well as shared genes between WT and fat-1 mice in response to EtOH are listed in Supplementary Tables S2, S3, and S4, respectively. Cytoscape GO analysis identified downregulated processes in WT-EtOH vs WT-PF mice including lipid metabolism, oxoacid metabolism, anion transport, fatty acid metabolism, and cholesterol homeostasis; while upregulated processes included development and response to hormones (Fig. 2c). GO processes diminished in fat-1 EtOH vs fat-1 PF mice included oxoacid metabolism, lipid metabolism, drug metabolism, and nucleobase metabolism, while cell communication, cell migration, signal transduction, transport, cell adhesion, and cell differentiation were enriched (Fig. 2d).
GO processes representative of the decreased genes in response to EtOH shared between genotypes included lipid metabolism, carboxylic acid metabolism, drug metabolism, and anion transport. The elevated shared genes due to EtOH included hormone and developmental processes (Fig. 3a). Gene clusters identified by STRING were used to visualize gene interactions and pathways among DEGs commonly affected by EtOH in both genotypes. Common pathway clusters decreased by EtOH in WT and fat-1 mice included amino acid metabolism (Arg2, Dao, Pipox), sugar metabolism (Aldob, G6pc, Fbp1), purine and pyrimidine metabolism (Ada, Cda, Nt5e), steroid and xenobiotic metabolism (Rdh7, Adh1, Cyp2b10), and the peroxisome (Ephx2, Dao, Acsl5) (Fig. 3b). Pathway clusters representative of common genes elevated by EtOH in WT and fat-1 mice included immunity (Pparγ, Muc2, Tlr2), development (Hoxb4, Hoxb8, Hoxb5), lipid metabolism (Cyp2c55, Cyp3a44, Far1), sulfur compound metabolism (Cbs, Ethe1, Fa2h), and the phagosome (Itga2, Thbs2, Sparcl1) (Fig. 3c).
EtOH-mediated alterations in the intestinal transcriptome: unique transcriptional responses to EtOH in WT or fat-1 mice
Next, we identified the specific transcriptional responses to EtOH in the intestinal mucosa of WT and fat-1 mice. There were a set of genes most prominently up and down-regulated in response to EtOH treatment exclusively in WT but not in fat-1 mice including Slc37a2, Ighv14-4, Ighv1-63, Abca12, and Rn7sk (~ 47-, 15-, 14-, 13-, and 11- fold increases, respectively), and Fabp1, Ighv2-3, Igkv4-51, Gip, and Ighv8-8 (~ 70-, 16-, 15-, 14-, and 8- fold decreases, respectively, Supplementary Table S5). The genes most highly changed in response to EtOH treatment exclusively in fat-1 but not in WT mice included: Car1, Marco, Nov, Plet1, and Cdhr1 (~ 1454-, 151-, 97-, 31-, and 30-fold increases, respectively), and Bbox1, Slc4a5, Dnah2, Rec8, and Afp (~ 59-, 25-, 20-, 19-, and 16-fold decreases, respectively, Supplementary Table S6).
Cytoscape cluster analysis identified genotype-specific pathway alterations in response to EtOH. Pathways that were downregulated in response to EtOH in WT mice exclusively included hormones (Igf2, Gip), hemoglobin (Hba-a1, Hba-a2, Hbb-bs), and innate immunity (Lcn2, Slpi, Ly6d). Pathway clusters upregulated in response to EtOH in WT mice only included tuft cell markers (Itpr2, Dclk1, Sucnr1), lectin recognition (Chodl, Siglec5, Reg4), and fibroblast growth factors (FGFs) (Fgf13, Fgf15) and inflammation (Fosb, Egr1) (Fig. 4a). Clustered pathways decreased in response to EtOH in fat-1 mice exclusively include cell death (Gsdmd, Il18, Casp7), the peroxisome (Hadh, Crat, Acaa1a), arachidonic acid metabolism (Cyp2c66, Cyp2j6, Cyp2u1), cytokine signaling (Il15, Cxcl16, Gzma), lipid metabolism (Nr1h3, Dgat1, Dgat2), and sugar metabolism (Pklr, Eno1, Mdh1) (Fig. 4b). Pathway clusters increased in response to EtOH in fat-1 mice only included defense response (Marco, Il1rn, Tlr4), PPAR signaling (Pnpla3, Plin4, Plin1), endocytosis (Syt1, Dab2, Epn2), mucin O-glycan synthesis (B3gnt6, Gcnt1, Gaint12), xenobiotic metabolism (Gstm2, Cbr2, Cbr3), and GPCR signaling (Ffar2, Ffar4, Tacr2) (Fig. 4b).
Differential transcriptional responses to EtOH between fat-1 and WT mice
The differences described above in transcriptional responses to EtOH in fat-1 and WT mice resulted in a total of 94 genes that were differentially regulated between fat-1 EtOH and WT EtOH animals. Among 80 up-regulated DEGs in fat-1 EtOH vs WT EtOH, the greatest increase was observed for Nov, Plet1, Cd209b, Ighv1-85, Grin3a, Nxpe2 (~ 61-, 46-, 20-, 16-, 14-, and tenfold increases, respectively, Fig. 5a, Supplementary Table S7). Immunoglobulins, including Ighv3-8 and Igkv2-109 were among the 14 most decreased DEGs in fat1-1 EtOH vs WT EtOH mice. Processes representative of genes decreased in fat-1 EtOH mice included negative regulation of tyrosine phosphorylation and JAK-STAT signaling, while leukocyte migration, stress response, negative regulation of cell death, and lipid metabolism were increased (Fig. 5b). Gene clusters identified by STRING analysis and the pathways they represent were used to visualize ileum DEGs between fat-1 EtOH and WT EtOH mice. Clustered pathways downregulated in fat-1 EtOH vs WT EtOH mice included immunity (Igkv8-21, Igkv2-109, Igkv5-17), the JAK-STAT pathway (Ptk6, Socs3), inflammation (Nos2), GTP-binding protein (Arl4a), bile acid metabolism (Fgf15), and phospholipase (Pla2g4c) (Fig. 5c). Pathway clusters upregulated in fat-1 EtOH vs WT EtOH mice included glycoproteins (St3gal3, St6galnac6, Ces1d), lipid metabolism (Acaa1b, Cyp2c65, Acacb), cysteine-methionine metabolism (Tst, Cbs), tryptophan metabolism (Tph1), sugar metabolism (Hk1, Akr1b8), immune response (Cd209b, Ighv1-85), glutamate receptor (Grin3a), and tissue homeostasis (Plet1, Nov) (Fig. 5d).
Targeted analysis of selected intestinal metabolic pathways relevant to diet and EtOH-mediated alterations of gut homeostasis and overall well-being
Effects of EtOH on the expression of free fatty acid receptors in WT and fat-1 mice
Nutrient sensing in the gut epithelium is fundamental for intestinal health. Since dietary fatty acids (FAs) can impact intestinal homeostasis, and n3 and n6 PUFAs act as signaling molecules, we were interested in evaluating the expression of receptors recognizing these molecules, including the FFAR and PPAR families. We observed that Pparβ/δ was the most abundantly expressed among all FA receptors (Fig. 6a). Ffar4, which recognizes MCFAs and LCFAs15,16, had the highest expression among the FFARs in the intestinal mucosa (Fig. 6a). There were no evident differences in FA receptors between WT PF and fat-1 PF mice. EtOH administration resulted in elevated levels of Ffar2 and Ffar4 in fat-1 but not WT mice, and an increase in Pparγ, which recognizes LCFAs and LCFA metabolites, in both WT and fat-1 EtOH-fed animals (Fig. 6b,c). The expression of Cd36, which mainly recognizes LCFAs was significantly elevated in fat-1 vs WT PF but not EtOH-fed animals (Fig. 6d).
Changes in the expression of microbial sensing genes in WT and fat-1 mice exposed to EtOH
The intestinal epithelium expresses a range of pattern recognition receptors (PRRs) that sense and respond to a variety of microbial signals to maintain mucosal homeostasis17. Among the PRRs, toll-like receptors (TLRs) are key players in microbe and microbial-product recognition. TLRs are also significantly involved in host defense and tissue repair responses18, and play a critical role in EtOH-mediated immune response19. In addition, saturated and unsaturated FAs may exert their effects via activation of TLRs like TLR2 and TLR420. In our study, we observed that Tlr3 was the highest expressed TLR in the intestinal mucosa, followed by Tlr1 and Tlr12 (Fig. 7a). Further, there was a significant increase in Tlr1 and Tlr2 expression in both fat-1 and WT EtOH-fed mice compared to control PF animals, while Tlr4 and Tlr5 were significantly up-regulated by EtOH only in fat-1 mice. Interestingly, the expression of Tlr2 and Tlr5 were significantly higher in fat-1 EtOH compared to WT EtOH (Fig. 7b). Tlr12 was the only family member decreased by EtOH in fat-1 mice (Fig. 7c). There were no significant differences between any experimental groups in the expression of other Tlrs or other PRR families, e.g., nucleotide-binding oligomerization domain-like receptors (Nlrs, data not shown).
Impact of EtOH on the intestinal genes involved in the adenosine signaling pathway in WT and fat-1 mice
Adenosine signaling is recognized as an important endogenous anti-inflammatory pathway in various diseases, including intestinal injury and inflammation21,22. Adenosine triphosphate (ATP) is released by stressed, apoptotic, and necrotic intestinal epithelial cells (IECs) as well as bacteria during inflammation23. ATP is converted to adenosine by cell surface ectonucleotidases ectonucleoside triphosphate diphosphohydrolase-1 (CD39) and ecto-5′ nucleotidase (CD73). Adenosine signals via several receptors and is inactivated by equilibrative nucleoside transporters (ENTs) or by adenosine deaminase (ADA)24 (Fig. 8a). RNAseq analysis in our study revealed no significant differences in Cd39 due to EtOH consumption or genotype, while Cd73 was significantly down-regulated by EtOH in both WT and fat-1 mice (although the expression was not different between the fat-1-EtOH and WT EtOH groups) (Fig. 8b). Further, A2bR, a predominant adenosine receptor expressed in IECs21,25, was also significantly down-regulated by EtOH in both genotypes. Interestingly, A2bR expression was markedly higher in fat-1 vs WT PF but not EtOH-fed animals (Fig. 8b). There were no noticeable changes in the levels of Ent1 and Ent2 in response to EtOH administration or genotype (Fig. 8c). Ada expression was significantly increased in fat-1 PF vs WT PF mice but was down-regulated by EtOH in both genotypes (Fig. 8c). Overall, our results suggest that n3 PUFAs may enhance adenosine signaling initiating an anti-inflammatory signaling cascade in response to EtOH.