Dietary calcium as a possible anti-promoter of colon carcinogenesis

SUMMARYColon cancer is the second-most common malignancy in both males and females and is strongly related to environmental factors of which diet seems to be the most important one. Dietary fat is positively correlated with the incidence of colon cancer whereas dietary fibre and dietary calcium seem to be negatively associated with the risk of colon cancer. The development of colon cancer can genetically and histologically be characterized in different phases which may offer several oppertunities for dietary intervention. A high intake of dietary fat promotes colon cancer in tumor-induction models whereas dietary calcium counteracts this promotive effect of dietary fat. The promotive effect of dietary fat may be mediated by bile acids and fatty acids in colon which induce hyperproliferation of the colonic epithelium. Colonic hyperproliferation is generally considered as a biomarker of an increased susceptibility to colon cancer. The mechanism by which the surface-active bile acids and fatty acids induce hyperproliferation consists probably of damage of the cell-membrane resulting in cytolysis and cell-death. Luminal binding of these surfactants by insoluble calcium phosphate may decrease the solubility of these surfactants, their cytotoxic effects and the induction of hyperproliferation (Figure 1).The experimental studies described in this thesis investigated this hypothesized mechanism by which dietary calcium inhibits proliferation of the colonic epithelium and the risk of colon cancer using the following step-wise approach. First, in in vitro studies (chapter 3) it was shown that bile acids and fatty acids are toxic for erythrocytes and for colonic epithelial cells in vitro (CaCo-2 cells). Moreover, cytotoxicity of these surfactants was similar in both cell-types. Fatty acids were as cytotoxic as bile acids with a similar dependency on hydrophobicity, but in contrast to bile acids, cytotoxicity of fatty acids was apparently not associated with micelle-formation. Low submicellar, sublytic concentrations of bile acids synergistically stimulated fatty acid-induced cytotoxicity. Cytotoxicity of these mixtures of bile acids and fatty acids was, analogous to cytotoxicity of bile acids alone, dependent on micelle formation. The synergistic effect of bile acids on fatty acid-induced cytotoxicity was stimulated by an increasing hydrophobicity of the bile acids. Lowering of the fatty acid concentration of these mixtures drastically decreased their cytotoxic effects. This in vitro study showed that physiologically relevant mixtures of bile acids and fatty acids are toxic to enterocytes as well as erythrocytes.Secondly, it was investigated in rats whether a diet-induced increase in colonic bile acids stimulated cytotoxicity of fecal water and colonic proliferation (chapter 4). Therefore, fecal water was isolated from feces. Fecal water is that fraction of feces which contains the soluble surfactants which are assumed to damage cells. A dietinduced increase in colonic bile acids drastically stimulated cytotoxicity of fecal water analogous to the increase in concentration of bile acids in fecal water. Also an increase in colonic proliferation measured as [ 3H]-incorporation into DNA was observed. Cytotoxicity of fecal water and colonic proliferation were highly correlated (r=0.85, n=24, P<0.001) indicating cause-and-effect relationships. Thus, this study showed that diet could affect the risk of colon cancer by modulation of cytotoxicity of the intestinal contents. Subsequently, the next step in the hypothesis, suggesting that dietary supplementation with CaHPO 4 stimulated the intestinal formation of insoluble CaP i , decreased the concentrations of soluble surfactants and lowered cytotoxicity of fecal water, was investigated (chapter 5). Supplemental CaHPO 4 resulted in a dosedependent decrease in bile acid concentration of fecal water and an increased precipitated fraction of bile acids. Consequently a drastic, dose-dependent fall in cytotoxicity was observed. Correlations between precipitated bile acids and cytotoxicity with precipitated calcium and inorganic phosphate (r>0.90) showed that theseparameters were closely associated. Thus, dietary supplementation with CaHPO 4 decreased cytotoxicity of fecal water by intestinal precipitation of bile acids.Because bile acids and fatty acids are products of fat digestion, the type of dietary fat may influence the cytotoxic effects of the intestinal contents and the interaction with calcium phosphate. Therefore, three different purified types of dietary fat (milk fat, palm oil and corn oil) were used each at a low (25 μmol CaHPO 4 /g diet) and a high (225 μmol CaHPO 4 /g diet) level (chapter 6). Milk fat contains rather large amounts of short- and medium-chain triglycerides (C 4:0 -C 12:0 ), palm oil contains triglycerides rich in palmitate (Ca 16:0 ) and oleate (C 18:1 ) and corn oil contains linoleate (C 18:2 ) rich triglycerides. The excretion of fatty acids was dependent on the type of dietary fat and increased in the order corn oil << milk fat << palm oil. Supplemental CaHPO 4 drastically stimulated fatty acid excretion with the same fat-type dependency. In contrast, concentrations of soluble fatty acids and bile acids were drastically decreased by supplemental CaHPO 4 . This effect was dependent on the type of dietary fat with palm oil resulting in the highest concentration of fatty acids. Cytotoxicity of fecal water decreased analogous to the decrease in concentrations of soluble surfactants with the same fat-type dependency. Multiple regression analysis showed that the concentrations of soluble bile acids and fatty acids are equally important determinants of cytotoxicity and explain 80% (R = 0.89) of the cytotoxicity of fecal water. Intestinal alkaline phosphatase activity in fecal water as a marker of intestinal epitheliolysis showed that the effects on cytotoxicity of fecal water were reflected in effects on intestinal epitheliolysis (r=0.92, P<0.001). Thus, despite increases in total fecal concentrations, dietary calcium phosphate decreased concentrations of soluble surfactants dependent on the type of dietary fat. The decrease in soluble surfactants resulted in a decreased cytotoxicity of fecal water and a decreased intestinal epitheliolysis.To investigate whether the luminal effects of dietary calcium resulted in a change in response of the colonic epithelium, the anti-proliferative effect of calcium was studied in rats fed Western-type high risk diets containing low CaHPO 4 (25 μmol Ca/g), low fibre and high fat (40 en-%) levels (chapter 7). It should be noted that this low calcium diet mimicked a human diet with a calcium consumption of 500 mg/day (mean intake in The Netherlands is about 1000 mg/day). Types of fat used were commercially available fats: butter, mimicking a diet with a high saturated mediumchain triglycerides content, saturated margarine, reflecting a diet with a high long-chain saturated fat content, and polyunsaturated margarine, ressembling a diet with a high polyunsaturated fat content. Diets were also supplemented with CaHP0 4 to 225 μmol/g diet. Dietary calcium phosphate decreased soluble surfactant concentrations and cytotoxicity of fecal water. These luminal effects of dietary calcium resulted in a lower intestinal epitheliolysis. On the butter and saturated margarine diets, these protective effects of calcium resulted in a decreased colonic proliferation. On the polyunsaturated diet, proliferation was not decreased by supplemental calcium which may suggest an additional mechanism in the induction of hyperproliferation. Multiple regression analysis of soluble surfactants with cytotoxicity, epitheliolysis and colonic proliferation showed highly significant associations. Cytotoxicity and epitheliolysis as well as epitheliolysis and colonic proliferation were highly correlated (resp. r=0.97 and r=0.88) for control and Ca-supplemented groups. This is consistent with the proposed mechanism and may indicate cause-and-effect relationships. Thus the anti-proliferative effect of dietary calcium is mediated by luminal surfactants and dependent on the type of dietary fat.Finally, the luminal effects of supplemental dietary calcium were studied in a pilot intervention trial with young healthy volunteers. In this study supplemental calcium complexated with phosphate to insoluble calcium phosphate, but in contrast to the animal studies the (already low) concentration of soluble bile acids was not further decreased. The free fatty acid concentration in fecal water was lowered by supplemental calcium. Gaschromatographic analysis of fecal water showed that the composition of the bile acids in fecal water was altered from hydrophobic to hydrophilic. The overall effect of dietary calcium supplementation was a decrease in hydrophobicity of fecal water measured as a decrease in free fatty acid concentration, hydrophobic bile acids and neutral sterols. This resulted in a significant decrease in cytotoxicity of fecal water in these healthy volunteers. Thus, luminal effects of dietary calcium supplementation in humans are analogous to the effects in rats. These effects may offer a molecular explanation of the protective effects of dietary calcium supplementation on colonic epithelial proliferation in patients at risk for colon cancer.CONCLUDING REMARKSThe experimental studies described in this thesis lead to the following conclusions:1. Bile acids and fatty acids are cytotoxic surfactants which cause damage of cellular membranes resulting in cell-death. Lysis of erythrocytes is a convenient and relevant model system to determine cytotoxicity of these surfactants (chapter 3).2. Diet may modulate the proliferative behavior of the colonic epithelium by luminal effects. Cytotoxicity of fecal water measured as lysis of erythrocytes providesrelevant information about the luminal effects of diet (chapter 4 & 5).3. The type of dietary fat influences the concentrations of luminal surfactants, cytotoxicity, intestinal epitheliolysis (chapter 6 & 7) and colonic proliferation (chapter 7). Dietary calcium phosphate supplementation causes precipitation of luminal surfactants. As a consequence, concentrations of soluble cytotoxic surfactants are lowered, which results in a decreased cytotoxicity of fecal water (chapter 5-7). These luminal effects of dietary calcium phosphate decrease intestinal epitheliolysis (chapter 6 & 7) and lower colonic epithelial proliferation (chapter 7).4. In humans, supplemental calcium lowers the hydrophobicity of fecal water and consequently decreases cytotoxicity of fecal water.With regard to these conclusions some remarks have to be made concerning the protective effects of calcium. It should be stressed that cytotoxicity is not a measure of intestinal cell-damage. Luminal cytotoxicity only refers to the potency of the intestinal contents to damage cells. Cytotoxicity is therefore solely determined by the physicochemical characteristics of the luminal surfactants. Lysis of epithelial cells is dependent on this cytotoxicity, but also on the susceptibility of the plasmamembrane to the luminal surfactants. Determinants of this epithelial susceptibility are at present largely unknown. In rats consuming identical diets only differing in the amount of CaHPO 4 and type of dietary fat, the proposed sequence of effects (figure 1) has been demonstrated. However, in the experiment with human volunteers only the luminal effects of dietary calcium supplementation were studied (chapter 8). Whether the observed protective effects result, analogous to the effects in rats, in alower colonic proliferation is at present unknown. Therefore, the results of chapter 8 should only be interpreted as a first indication how dietary calcium in humans could decrease the risk of colon cancer.Further investigations regarding calcium and colon cancer should focuss on the effects of calcium supplementation on colonic epithelial proliferation using a combined biochemical and physiological design. Only these combined efforts may shed some light on the complex role of diet in colorectal carcinogenesis. Many questions have still to be answered. Dose-response and time-relationships need to be ascertained. Whether an optimal dose of calcium supplementation exists, is not clear. Whether dairy products, which are rich in calcium, have the same effects should also be ascertained. Possible interactions not only influenced by the the type of dietary fat, but also interactions between the protective agents fibre and calcium in the intestinal lumen may be important. It should be clear that luminal colonic effects of diet are very complex to study, because of intestinal interactions between the different dietary components. However, these interactions are important in the dietary prevention of colon cancer.In conclusion, the experiments described in this thesis show that dietary supplementation with CaHPO 4 decreases cytotoxicity of fecal water resulting in a decreased colonic epithelial proliferation. The relative importance of dietary calcium as an anti-promoter of colon carcinogenesis has to be established in the following years. Several multi-center intervention trials have started during the last few years studying the effects of calcium supplementation on proliferation-markers and recurrence of polyps. However, one should be cautious not to neglect biochemical and animal studies to investigate mechanisms how calcium could affect the risk of colon cancer. Dietary advice with regard to calcium-intake and the prevention of colon cancer has to await knowledge of relevant mechanisms combined with consistent results in in vitro studies, experiments with animals and human intervention trials.