Butyrate metabolism in human colon carcinoma cells: implications concerning its growth-inhibitory effect

Butyrate and acetate are bacterial metabolites present in the large intestine lumen. Although butyrate is well known to inhibit the in vitro proliferation of human colon carcinoma cells in a process involving the hyperacetylation of specific nuclear histones, little is known about the possible link between butyrate metabolism and its growth-inhibitory effect. In a previous study (Leschelle et al., 2000, Eur J Biochem 267: 6435-6442), we showed that butyrate accumulates and is metabolized in HT-29 Glc(-/+) cells without increasing oxygen consumption. In the present study, using the same cell line incubated with (14)C-labeled butyrate, we determined that a minor part of (14)C from butyrate was recovered in nuclear histones. Unlike butyrate, acetate exerted no effect on cell growth but was a precursor for overall net histone acetylation. Although butyrate was able to increase the cellular AMP/ADP ratio, it did not affect the ATP cell content or the adenylate charge or the oxidation of endogenous L-glutamine. Butyrate oxidation was found to be markedly sensitive to the presence of other substrates with D-glucose decreasing this oxidation and L-malate stimulating it. Furthermore, in the presence of L-malate, the growth-inhibitory effect of butyrate was significantly weaker than in its absence. From these data, we conclude that the metabolism of butyrate downstream acetyl-CoA synthesis is not involved in the butyrate antiproliferative effect. The suggestion that butyrate metabolism in mitochondria is not used in these cells as a fuel but acts as a regulator of butyrate free concentrations (thus limiting its action upon cellular targets), is discussed.     

Spectroscopic and kinetic evidence for a cyclic geminal diamine intermediate in the reaction of O-aminoserine with pyridoxal 5'-phosphate in alkali

HT-29, a cell line derived from a human colon carcinoma, exhibits very low alkaline phosphatase activity. The enzyme is thermolabile and is of the intestinal type. Hyperosmolality and/or sodium butyrate induce increased levels of activity. The increase is most pronounced with HT-29 cells growing in hyperosmolar medium containing sodium butyrate. Under these conditions specific activity rises over 1000-fold. The effect of hyperosmolality is blocked by cycloheximide and that of sodium butyrate by thymidine, cordycepin, and cycloheximide. By contrast to other human cancer cell lines, the enzyme of HT-29 is not influenced by cell density and glucocorticoid hormones. 5-Bromo-2′-deoxyuridine and inhibitors of DNA synthesis cause a slight increase in specific activity.     

Upregulation of activin A gene by butyrate in human colon cancer cell lines

Activin A has been reported to play a role in the progression of colorectal cancer. Because dietary fiber protects against colorectal cancer, we hypothesized that butyrate, a fermentation product of dietary fiber, may affect the expression of activin A in colon cancer cells. Semiquantitative RT-PCR demonstrated that the activin A gene was upregulated by sodium butyrate in the human colon cancer cell lines HT-29 and Caco-2 in a concentration- and time-dependent manner. However, the activin A gene did not respond to sodium butyrate in the human normal colonic cell line FHC, rat normal intestinal epithelial cell (IEC) line IEC-6, and the explant of rat colon. Flow cytometry and agarose gel electrophoresis of genomic DNA revealed that cell cycle arrest and apoptosis were induced by sodium butyrate but not exogenous activin A in HT-29 cells, indicating that activin A could not act as an autocrine factor in colon cancer cells. By assuming that activin A promotes colorectal cancer spread as a paracrine factor, our findings suggest that butyrate could act as a tumor promoter in some circumstances.     

The effects of TNF-alpha and inhibitors of arachidonic acid metabolism on human colon HT-29 cells depend on differentiation status

The level of differentiation could influence sensitivity of colonic epithelial cells to various stimuli. In our study, the effects of TNF-alpha, inhibitors of arachidonic acid (AA) metabolism (baicalein, BA; indomethacin, INDO; niflumic acid, NA; nordihydroguaiaretic acid, NDGA), and/or their combinations on undifferentiated or sodium butyrate (NaBt)-differentiated human colon adenocarcinoma HT-29 cells were compared. NaBt-treated cells became growth arrested (blocked in G0/G1 phase of the cell cycle), and showed down-regulated Bcl-xL and up-regulated Bak proteins and increased expression of cyclooxygenase-2 (COX-2) and 5-lipoxygenase (5-LOX). These cells were more perceptive to anti-proliferative and apoptotic effects of TNF-alpha. Both inhibitors of LOX (BA and NDGA) and COX (INDO and NA) in higher concentrations modulated cell cycle changes accompanying NaBt-induced differentiation and induced various level of cell death in undifferentiated and differentiated cells. Most important is our finding that TNF-alpha action on proliferation and cell death can be potentiated by co-treatment of cells with AA metabolism inhibitors, and that these effects were more significant in undifferentiated cells. TNF-alpha and INDO co-treatment was associated with accumulation of cells in G0/G1 cell cycle phase, increased reactive oxygen species production, and elevated caspase-3 activity. These results indicate the role of differentiation status in the sensitivity of HT-29 cells to the anti-proliferative and proapoptotic effects of TNF-alpha, AA metabolism inhibitors, and their combinations, and imply promising possibility for novel anti-cancer strategies.     

Effects of short chain fatty acids on colonic Na+ absorption and enzyme activity

Short chain fatty acids (SCFA) stimulate colonic Na+ absorption and inhibit cAMP and cGMP-mediated Cl- secretion. It is uncertain whether SCFA have equivalent effects on absorption and whether SCFA inhibition of Cl- secretion involves effects on mucosal enzymes. Unidirectional Na+ fluxes were measured across stripped colonic segments in the Ussing chamber. Enzyme activity was measured in cell fractions of scraped colonic mucosa. Mucosal 50 mM acetate, propionate, butyrate and poorly metabolized isobutyrate stimulated proximal colon Na+ absorption equally (300%). Neither 2-bromo-octanoate, an inhibitor of beta-oxidation, nor carbonic anhydrase inhibition affected this stimulation. All SCFA except acetate stimulated distal colon Na+ absorption 200%. Only one SCFA affected proximal colon cGMP phosphodiesterase (PDE) (18% inhibition by 50 mM butyrate). All SCFA at 50 mM stimulated distal colon cAMP PDE (24-43%) and decreased forskolin-stimulated mucosal cAMP content. None of the SCFA affected forskolin-stimulated adenylyl cyclase in distal colon or ST(a)-stimulated guanylyl cyclase in proximal colon. Na+-K+-ATPase in distal colon was inhibited 23-51% by the SCFA at 50 mM. We conclude that all SCFA (except acetate in distal colon) stimulate colonic Na+ absorption equally, and the mechanism does not involve mucosal SCFA metabolism or carbonic anhydrase. SCFA inhibition of cAMP-mediated secretion may involve SCFA stimulation of PDE and inhibition of Na+-K+-ATPase.     

Transient vs. prolonged histone hyperacetylation: effects on colon cancer cell growth, differentiation, and apoptosis

The role of histone hyperacetylation in regard to growth, differentiation, and apoptosis in colon cancer cells was assessed in an in vitro model system. HT-29 cells were grown in +/-10% fetal bovine serum with either 5 mM sodium butyrate or 0.3 microM trichostatin A [single dose (T) or 3 doses 8 h apart (TR)] for 24 h. Serum-starved HT-29 cells were further treated with epidermal growth factor or insulin-like growth factor I for an additional 24 h. Apoptosis was quantified with propidium iodide and characterized by electron microscopy. Northern blot analyses were performed with cDNA probes specific for intestinal alkaline phosphatase, Na-K-2Cl cotransporter, the cell cycle inhibitor p21, and the actin control. Flow cytometric analysis revealed a time-dependent growth suppression along with early induction of p21 mRNA in the butyrate, T, and TR groups. Histone hyperacetylation, assessed by acid-urea-triton gel electrophoresis, was transient in the T group but persisted for up to 24 h in the butyrate and TR groups. Induction of apoptosis, growth factor unresponsiveness, and differentiation occurred in the butyrate- and TR-treated cells but not those treated with a single dose of trichostatin A. Thus transient hyperacetylation of histones is sufficient to induce p21 expression and produce cellular growth arrest, but prolonged histone hyperacetylation is required for induction of the programs of differentiation, apoptosis, and growth factor unresponsiveness.     

Histone deacetylase inhibition results in a common metabolic profile associated with HT29 differentiation

Cell differentiation is an orderly process that begins with modifications in gene expression. This process is regulated by the acetylation state of histones. Removal of the acetyl groups of histones by specific enzymes (histone deacetylases, HDAC) usually downregulates expression of genes that can cause cells to differentiate, and pharmacological inhibitors of these enzymes have been shown to induce differentiation in several colon cancer cell lines. Butyrate at high (mM) concentration is both a precursor for acetyl-CoA and a known HDAC inhibitor that induces cell differentiation in colon cells. The dual role of butyrate raises the question whether its effects on HT29 cell differentiation are due to butyrate metabolism or to its HDAC inhibitor activity. To distinguish between these two possibilities, we used a tracer-based metabolomics approach to compare the metabolic changes induced by two different types of HDAC inhibitors (butyrate and the non-metabolic agent trichostatin A) and those induced by other acetyl-CoA precursors that do not inhibit HDAC (caprylic and capric acids). [1,2-¹³C₂]-d-glucose was used as a tracer and its redistribution among metabolic intermediates was measured to estimate the contribution of glycolysis, the pentose phosphate pathway and the Krebs cycle to the metabolic profile of HT29 cells under the different treatments. The results demonstrate that both HDAC inhibitors (trichostatin A and butyrate) induce a common metabolic profile that is associated with histone deacetylase inhibition and differentiation of HT29 cells whereas the metabolic effects of acetyl-CoA precursors are different from those of butyrate. The experimental findings support the concept of crosstalk between metabolic and cell signalling events, and provide an experimental approach for the rational design of new combined therapies that exploit the potential synergism between metabolic adaptation and cell differentiation processes through modification of HDAC activity.     

Acidic extracellular pH shifts colorectal cancer cell death from apoptosis to necrosis upon exposure to propionate and acetate,major end-products of the human probiotic propionibacteria

The human probiotic Propionibacterium freudenreichii kills colorectal adenocarcinoma cells through apoptosis in vitro via its metabolites, the short chain fatty acids (SCFA), acetate and propionate. However, the precise mechanisms, the kinetics of cellular events and the impact of environmental factors such as pH remained to be specified. For the first time, this study demonstrates a major impact of a shift in extracellular pH on the mode of propionibacterial SCFA-induced cell death of HT-29 cells, in the pH range 5.5 to 7.5 prevailing within the colon. Propionibacterial SCFA triggered apoptosis in the pH range 6.0 to 7.5, a lethal process lasting more than 96 h. Indeed at pH 7.5, SCFA induced cell cycle arrest in the G2/M phase, followed by a sequence of cellular events characteristic of apoptosis. By contrast, at pH 5.5, the same SCFA triggered a more rapid and drastic lethal process in less than 24 h. This was characterised by sudden mitochondrial depolarisation, inner membrane permeabilisation, drastic depletion in ATP levels and ROS accumulation, suggesting death by necrosis. Thus, in digestive cancer prophylaxis, the observed pH-mediated switch between apoptosis and necrosis has to be taken into account in strategies involving SCFA production by propionibacteria to kill colon cancer cells. This work has been supported by a grant from CRITT santé Bretagne.     

Butyrate reduces colonic paracellular permeability by enhancing PPARgamma activation

Butyrate may have a role in preventing ulcerative colitis, but its precise mechanism is unknown. Also, PPARgamma (peroxisome proliferator-activated receptor gamma) is expressed at high levels both in the colonic epithelium and colon cancer cell lines, but no report was shown on the relationship between PPARgamma activation and the effect of butyrate. We investigated the effects of butyrate and PPARgamma agonist on paracellular permeability. To discover whether PPARgamma expressed in the cell lines treated with butyrate was functional or not, we transfected HT-29 cells with an acyl-CoA oxidase promoter-luciferase reporter plasmid containing a PPRE (peroxisome proliferator responsive element) and analyzed the luciferase activity. Butyrate and PPARgamma agonist significantly reduced paracellular permeability of the colon cell line (p<0.05) and this effect indicated that butyrate and PPARgamma agonist decreased HT-29 cell growth and increased differentiation (p<0.01). PPRE activation treated with butyrate was approximately four and a half times that in untreated cells (p<0.01). These findings suggest that the effect of butyrate on paracellular permeability has apparently taken place through PPARgamma activation and this effect attributes to preventing inflammation of the colon.     

Activation of autophagy and PPARγ protect colon cancer cells against apoptosis induced by interactive effects of butyrate and DHA in a cell type-dependent manner: The role of cell differentiation

The short-chain and n-3 polyunsaturated fatty acids exhibit anticancer properties, and they may mutually interact within the colon. However, the molecular mechanisms of their action in colon cancer cells are still not fully understood. Our study focused on the mechanisms responsible for the diverse effects of sodium butyrate (NaBt), in particular when interacting with docosahexaenoic acid (DHA), in distinct colon cancer cell types, in which NaBt either induces cell differentiation or activates programmed cell death involving mitochondrial pathway. NaBt activated autophagy both in HT-29 cells, which are sensitive to induction of differentiation, and in nondifferentiating HCT-116 cells. However, autophagy supported cell survival only in HT-29 cells. Combination of NaBt with DHA-promoted cell death, especially in HCT-116 cells and after longer time intervals. The inhibition of autophagy both attenuated differentiation and enhanced apoptosis in HT-29 cells treated with NaBt and DHA, but it had no effect in HCT-116 cells. NaBt, especially in combination with DHA, activated PPARγ in both cell types. PPARγ silencing decreased differentiation and increased apoptosis only in HT-29 cells, therefore we verified the role of caspases in apoptosis, differentiation and also PPARγ activity using a pan-caspase inhibitor. In summary, our data suggest that diverse responses of colon cancer cells to fatty acids may rely on their sensitivity to differentiation, which may in turn depend on distinct engagement of autophagy, caspases and PPARγ. These results contribute to understanding of mechanisms underlying differential effects of NaBt, when interacting with other dietary fatty acids, in colon cancer cells.     

Modulation of lipid synthesis,apolipoprotein biogenesis,and lipoprotein assembly by butyrate

Short-chain fatty acids (SCFAs) are potent modulators of the growth, function, and differentiation of intestinal epithelia. In addition, high-fiber diets may protect against the development of atherosclerosis because of their cholesterol-lowering effects due, in large part, to SCFA production, liver sterol metabolism, and bile acid excretion. Although the small gut plays a major role in dietary fat transport and contributes substantially to plasma cholesterol and lipoprotein homeostasis, the impact of SCFAs on intestinal lipid handling remains unknown. In the present study, the modulation of lipid synthesis, apolipoprotein biogenesis, and lipoprotein secretion by butyrate was investigated in Caco-2 cells plated on permeable polycarbonate filters, which permit separate access to the upper and lower compartments of the monolayers. Highly differentiated and polarized cells (20 days of culture) were incubated for 20 h with 20 mM butyrate in the apical medium. In the presence of [14C]oleic acid, butyrate led to a significant reduction of secreted, labeled triglycerides (27%; P < 0.01) and phospholipids (25%; P < 0.05). Similarly, butyrate significantly decreased the incorporation of [14C]acetate into exported cholesteryl ester (49%; P < 0.005). As expected from these results, with [14C]oleic acid as a precursor, butyrate significantly (P < 0.05) diminished the delivery of radiolabeled chylomicrons and very low-density lipoproteins. In parallel, [35S]methionine pulse labeling of Caco-2 cells revealed the concomitant inhibitory effect of butyrate on the synthesis of apolipoproteins B-48 (28%; P < 0.05) and A-I (32%; P < 0.01). Collectively, our data indicate that butyrate may influence lipid metabolism in Caco-2 cells, thus suggesting a potential regulation of intestinal fat absorption and circulating lipoprotein concentrations.     

Effects of agmatine accumulation in human colon carcinoma cells on polyamine metabolism, DNA synthesis and the cell cycle

Putrescine, spermidine and spermine are low molecular polycations that play important roles in cell growth and cell cycle progression of normal and malignant cells. Agmatine (1-amino-4-guanidobutane), another polyamine formed through arginine decarboxylation, has been reported to act as an antiproliferative agent in several non-intestinal mammalian cell models. Using the human colon adenocarcinoma HT-29 Glc(-/+) cell line, we demonstrate that agmatine, which markedly accumulated inside the cells without being metabolised, exerted a strong cytostatic effect with an IC50 close to 2 mM. Agmatine decreased the rate of L-ornithine decarboxylation and induced a 70% down-regulation of ornithine decarboxylase (ODC) expression. Agmatine caused a marked decrease in putrescine and spermidine cell contents, an increase in the N1-acetylspermidine level without altering the spermine pool. We show that agmatine induced the accumulation of cells in the S and G2/M phases, reduced the rate of DNA synthesis and decreased cyclin A and B1 expression. We conclude that the anti-metabolic action of agmatine on HT-29 cells is mediated by a reduction in polyamine biosynthesis and induction in polyamine degradation. The decrease in intracellular polyamine contents, the reduced rate of DNA synthesis and the cell accumulation in the S phase are discussed from a causal perspective.     

Stimulatory effect of short chain fatty acids on the epithelial cell proliferation in rat large intestine

1. Short chain fatty acids [SCFA: acetate 75, propionate 35, butyrate 20 (microM)] introduced intraluminally into the colon increased the mitotic index and the labeling index of the large intestinal epithelial cells within 60 min in fasted rats. 2. Epithelia that lacked direct contact with SCFA were also stimulated. 3. SCFA did not have such a stimulatory effect either in vagotomized rats or in sympathectomized rats. 4. These results suggest that SCFA stimulate epithelial cell proliferation in the large intestine of fasted rats via the autonomic nervous system.     

Occurrence, absorption and metabolism of short chain fatty acids in the digestive tract of mammals

Short chain fatty acids (SCFA) also named volatile fatty acids, mainly acetate, propionate and butyrate, are the major end-products of the microbial digestion of carbohydrates in the alimentary canal. The highest concentrations are observed in the forestomach of the ruminants and in the large intestine (caecum and colon) of all the mammals. Butyrate and caproate released by action of gastric lipase on bovine milk triacylglycerols ingested by preruminants or infants are of nutritional importance too. Both squamous stratified mucosa of rumen and columnar simple epithelium of intestine absorb readily SCFA. The mechanisms of SCFA absorption are incompletely known. Passive diffusion of the unionized form across the cell membrane is currently admitted. In the lumen, the necessary protonation of SCFA anions could come first from the hydration of CO2. The ubiquitous cell membrane process of Na+-H+ exchange can also supply luminal protons. Evidence for an acid microclimate (pH = 5.8-6.8) suitable for SCFA-protonation on the surface of the intestinal lining has been provided recently. This microclimate would be generated by an epithelial secretion of H+ ions and would be protected by the mucus coating from the variable pH of luminal contents. Part of the absorbed SCFA does not reach plasma because it is metabolized in the gastrointestinal wall. Acetate incorporation in mucosal higher lipids is well-known. However, the preponderant metabolic pathway for all the SCFA is catabolism to CO2 except in the rumen wall where about 80% of butyrate is converted to ketone bodies which afterwards flow into bloodstream. Thus, SCFA are an important energy source for the gut mucosa itself.     

Src activity increases and Yes activity decreases during mitosis of human colon carcinoma cells.

Src and Yes protein-tyrosine kinase activities are elevated in malignant and premalignant tumors of the colon. To determine whether Src activity is elevated throughout the human colon carcinoma cell cycle as it is in polyomavirus middle T antigen- or F527 Src-transformed cells, and whether Yes activity, which is lower than that of Src in the carcinoma cells, is regulated differently, we measured their activities in cycling cells. We observed that the activities of both kinases were higher throughout all phases of the HT-29 colon carcinoma cell cycle than in corresponding phases of the fibroblast cycle. In addition, during mitosis of HT-29 cells, Src specific activity increased two- to threefold more, while Yes activity and abundance decreased threefold. The decreased steady-state protein levels of Yes during mitosis appeared to be due to both decreased synthesis and increased degradation of the protein. Inhibition of tyrosine but not serine/threonine phosphatases abolished the mitotic activation of Src. Mitotic Src was phosphorylated at novel serine and threonine sites and dephosphorylated at Tyr-527. Two cellular proteins (p160 and p180) were phosphorylated on tyrosine only during mitosis. Tyrosine phosphorylation of several other proteins decreased during mitosis. Thus, Src in HT-29 colon carcinoma cells, similar to Src complexed to polyomavirus middle T antigen or activated by mutation at Tyr-527, is highly active in all phases of the cell cycle. Moreover, Src activity further increases during mitosis, whereas Yes activity and abundance decrease. Thus, Src and Yes appear to be regulated differently during mitosis of HT-29 colon carcinoma cells.     

Butyrate suppresses Cox-2 activation in colon cancer cells through HDAC inhibition

Cox-2 plays an important role in colon carcinogenesis and inflammation. Studying the HT-29 colon cancer cell line as a model, we found that Cox-2 expression and activity is increased approximately 25-fold by TNF-alpha. As previously reported for other Cox-2 inducers, this activation appears to result from a p38-mediated mRNA stabilization rather than an increase in promoter activity. The HDAC inhibitors butyrate and TSA blocked the TNF-alpha activation of Cox-2 protein and mRNA synthesis, and dramatically suppressed Cox-2 activity in HT-29 cells. The suppression of Cox-2 synthesis did not involve promoter inactivation and could be achieved even when applied after the TNF-alpha stimulus. The effect of the HDAC inhibitors was observed prior to the activation of p21 expression and did not require new protein synthesis. Finally, butyrate did not prevent p38 phosphorylation, so the block is likely to occur at a later step in the activation pathway. We propose that a component of the cytokine-induced Cox-2 mRNA stabilization pathway is sensitive to acetylation.