Systematic Investigations on the Metabolic and Transcriptomic Regulation of Lactate in the Human Colon Epithelial Cells

Abstract

Lactate, primarily produced by the gut microbiota, performs as a necessary "information transmission carrier" between the gut and the microbiota. To investigate the role of lactate in the gut epithelium cell-microbiota interactions as a metabolic signal, we performed a combinatory, global, and unbiased analysis of metabolomic and transcriptional profiling in human colon epithelial cells (Caco-2), using a lactate treatment at the physiological concentration (8 mM). The data demonstrated that most of the genes in oxidative phosphorylation were significantly downregulated in the Caco-2 cells due to lactate treatment. Consistently, the levels of fumarate, adenosine triphosphate (ATP), and creatine significantly decreased, and these are the metabolic markers of OXPHOS inhibition by mitochondria dysfunction. The one-carbon metabolism was affected and the polyol pathway was activated at the levels of gene expression and metabolic alternation. In addition, lactate significantly upregulated the expressions of genes related to self-protection against apoptosis. In conclusion, lactate participates in gut-gut microbiota communications by remodeling the metabolomic and transcriptional signatures, especially for the regulation of mitochondrial function. This work contributes comprehensive information to disclose the molecular mechanisms of lactate-mediated functions in human colon epithelial cells that can help us understand how the microbiota communicates with the intestines through the signaling molecule, lactate.

Keywords: NMR-based metabolomics; lactate; mitochondrial dysfunction; one-carbon metabolism; polyol pathway; transcriptomics.

Figure 1

Determination of the cytotoxic activity of lactate on Caco-2 Cells using the MTT assay. N = 4, values are means ± SEM, *** p < 0.0001.

Figure 2

Typical ¹H NMR spectra from the cell extracts. Keys: PA, Pantothenate; Ile, Isoleucine; Leu, Leucine; Val, Valine; α-KIV, α-Keto-isovalerate; α-KMV, α-keto-beta-methyl-valerate; Lac, Lactate; Ala, Alanine; Thr, Threonine; Glu, Glutamate; Gln, Glutamine; NAG, N-Acetyl-Glutamine; Suc, Succinate; Met, Methionine; DMG, Dimethylglycine; Cre, Creatine; Cho, Choline; PC, Phosphocholine; GPC, Glycerophosphocholine; Tau, Taurine; GSH, Glutathione; Glc, Glucose; Fru, Fructose; Gal-1-p, Galactose-1-phosphate; Tyr, Tyrosine; Phe, Phenylalanine; Fum, Fumaric acid; NAD⁺, Nicotinamide adenine dinucleotide; FA, Formate; His, Histidine; AMP, Adenosine monophosphate; ATP, Adenosine triphosphate; UMP, Uridine 5′-monophosphate; UDP, Uridine-5′-diphosphate; UDPG, UDP Glucuronate; UDP-N-Ace, UDP-N-acetylglucosamine; C-alpha, The central carbon atom in amino acids.

Figure 3

Differential metabolomics of the Caco-2 cells between the lactate and the control group. (A). Principal component analysis (PCA) scores of the complete metabolic profiles of all samples (N = 4). The light blue region is for the 95% confidence interval of the lactate treatment; the light orange portion is for the 95% confidence interval of the control group; and the light blue and orange dots are representative of the samples of the lactate and the control treatments, respectively. (B). Color-coded correlation coefficient loadings plots generated by comparing the spectra of the intracellular metabolites between the lactate and the control treatments.

Figure 4

Comparison of the transcriptomes of the Caco-2 cells of the lactate and the control treatments. (A). Principal component analysis (PCA) scores of the complete expression profiles of all the samples (N = 3). The light blue region is for the 95% confidence interval of the lactate treatment; the light orange region is for the 95% confidence interval of the control group; and the light blue and orange dots are representative of the samples of the lactate and the control treatments, respectively. (B). Volcano plot showing the differentially expressed genes (DEGs) in the Caco-2 Cells of the lactate and the control treatments. The FDR-corrected p < 0.05; the absolute fold change >= 1.5. The blue dots are the downregulated DEGs, and the red dots are the upregulated DEGs caused by the lactate treatment. The gray dots are the genes with no difference in expression.

Figure 5

Supervised heat maps of genes in the enriched gene sets. The gene set enrichment analyses determined that the enriched gene sets and were performed using supervised hierarchical clustering in Heatmapper including DEGs with an FDR-corrected p < 0.05 and fold change ≥1.5, upregulated (red) gene, and downregulated (green) gene in the lactate vs. the control treatments.

Figure 6

Pathways significantly enriched in lactate treatment compared with control treatment. NES: normalized enrichment score. Negative NES (blue dot) shows downregulated pathways and positive NES (red dot) shows upregulated pathways in lactate treatment, green-highlighted pathways are significantly altered both at the metabolic and transcriptomic levels, * Adjusted p < 0.05, ** Adjusted p < 0.01.

Figure 7

Metabolic and transcriptomic alternations of pathways under lactate exposure. Blue (downregulated) and red (upregulated) in the lactate vs. the control treatments.