- Email: [email protected]
Deoxycholate Inhibitsin VivoButyrate-Mediated BrDU Labeling of the Colonic Crypt
JOURNAL OF SURGICAL RESEARCH ARTICLE NO.
69, 344–348 (1997)
JR975075
Deoxycholate Inhibits in Vivo Butyrate-Mediated BrDU Labeling of the Colonic Crypt Omaida C. Vela´zquez, M.D.,*† Rene´e W. Seto, B.A.,* Allison M. Bain, B.A.,* Jason Fisher, B.A.,* and John L. Rombeau, M.D.†,1 *Harrison Department of Surgical Research, University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104; and †Department of Surgery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19104 Presented at the Annual Meeting of the Association for Academic Surgery, Chicago, Illinois, November 13–16, 1996
The short-chain fatty acid butyrate (NaBu) selectively increases colonic crypt base proliferation and inhibits ‘‘premalignant’’ crypt surface hyperproliferation while the secondary bile acid deoxycholate (DCA) induces surface hyperproliferation, in vitro. We hypothesized that NaBu and DCA have similar selective and antagonistic effects on the colonic crypt proliferative pattern, in vivo. Fifty-six adult SD rats underwent surgical isolation of the colon and 24-hr intraluminal instillation with physiological (10 mM) and pharmacological (25 mM) levels of butyrate alone or combined with a physiological DCA level (5 mM). Bromodeoxyuridine-labeling indices (LI) were determined as labeled cells divided by total cells, for the whole crypt and five crypt compartments from base to surface. Treatment with NaBu increased total LI when compared to NaCl. This effect was significant only at the crypt base. Both doses of NaBu resulted in similar LI with no further response at the higher concentration. In contrast to prior in vitro studies, DCA alone at this concentration did not affect LI, but when combined with NaBu, DCA inhibited the effects of NaBu at the crypt base and surface. The conclusions are: (1) the in vivo proliferative effects of NaBu are selective to the crypt base, (2) an in vivo low physiological DCA level does not promote crypt surface hyperproliferation but does inhibit butyrate’s proliferative effect, and (3) NaBu and DCA interact in a complex and antagonistic manner to selectively modulate crypt base and surface proliferation, in the rat colon, in vivo. These findings may have clinical relevance since colonic levels of NaBu and DCA are affected by diet. q 1997 Academic Press
INTRODUCTION
Total crypt and crypt surface hyperproliferation occur in the colonic mucosa of humans at increased risk of colon cancer [1–4]. A shift in the proliferative zone 1 To whom correspondence should be addressed at Department of Surgery, 4 Silverstein Pavilion, 3400 Spruce Street, Hospital of the University of Pennsylvania, Philadelphia, PA 19104. Fax: 215-3498996.
0022-4804/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
JSR 5075
/
6n1e$$$441
from the base to the surface of the crypt as assessed by increased fh value (labeled cells in the upper 40% of the crypt divided by labeled cells in the whole crypt) is an excellent predictor of cancer risk in humans [1]. In animal models of chemically induced carcinogenesis, similar cytokinetic abnormalities have been observed [5]. The short-chain fatty acid butyrate and the secondary bile acid deoxycholate are diet by-products normally present within the colon luminal environment and are known to modulate human colonic crypt proliferation in vitro in an interactive and compartmentspecific manner [6]. While butyrate enhances crypt base proliferation, deoxycholate increases total crypt and crypt surface proliferation [7]. Moreover, co-incubation with butyrate and deoxycholate reverses the deoxycholate-induced surface hyperproliferation in proximal and distal human colon in vitro [6]. These findings have generated great interest since butyrate and deoxycholate have been implicated in colon cancer prevention and promotion, respectively [8–11] and since colon cancer incidence has been strongly associated with dietary practices [12, 13]. The purpose of this study was to determine whether the observed in vitro antagonistic effects of butyrate and deoxycholate on colonic crypt labeling indices also occur in vivo. MATERIALS AND METHODS Animals. The experimental protocol was approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Fifty-six adult healthy male Sprague–Dawley rats (250–350 g) were housed in conventional animal facilities, three rats per cage. The rats were fed fiber-free diet (Diet No. 112636, Dyets Inc., Bethlehem, PA) and water ad libitum for 5 days prior to surgery. Cages contained raised bottoms so that animals could not reach and ingest their cage bedding. Solid food was withheld the night prior to surgery. All animals were operated on in groups of four to six per day, taking care to include one to two animals from each treatment group at each day of surgery until all groups were completed. Surgical procedures. Rats were anesthetized with Nembutal (50 mg/kg) intraperitoneally. The abdomen was cleansed with Betadine. A laparotomy was performed with a full midline incision. A 3-0 silk ligature was used to ligate the proximal colon immediately distal to the cecum. The colon was flushed with saline (377C) injected into the proximal colon, just distal to the ileo-cecal junction, with a 27-gauge needle. All the fluid and particulate material were carefully evacu-
344
06-02-97 15:03:39
srga
´ ZQUEZ ET AL.: DEOXYCHOLATE AND BrDU LABELING VELA
FIG. 1. Diagram of colonic crypt compartments.
ated from the colonic lumen. A 3-0 silk purse string ligature was place at the anus to prevent escape of the treatment solution. The treatment solution (2 ml) was injected with a 27-gauge needle through the same puncture site used for the flushing solution. The puncture site was closed with a 6-0 prolene suture. Care was taken not to disturb the blood supply to the cecum or ascending colon. The animals were resuscitated with 10 cc of warm sterile saline intraperitoneally at the end of the procedure. The abdomen was closed in a single layer with a running 3-0 silk suture. Rats were allowed to recover from the general anesthetic under a warming light and then were returned to raised bottom cages with access to only water ad libitum. Twenty-two hours after instillation of the treatment solution (2 hr prior to sacrificing the rat and harvesting the colon), the rats received a 2 ml/100 g intraperitoneal injection with undiluted Cell Proliferation Labeling Reagent Bromodeoxyuridine (Amersham Life Science). Twenty-four hours after instillation of the treatment solution, rats were anesthetized with Nembutal (50 mg/ kg) intraperitoneally and the colon was harvested. Animals were then killed by cardiac puncture. The colon was immediately flushed with 10 cc of cold saline. A standard 2-cm segment starting 1-cm distal to the injection site (marked by the prolene suture) was placed in 10% formalin for bromodeoxyuridine immunohistochemistry [14]. Preparation of treatment solutions. Individual solutions were prepared fresh each day for each treatment group: (1) 10 mM NaCl, (2) 10 mM NaBu, (3) 5 mM DCA, (4) 10 mM NaBu plus 5 mM DCA, (5) 25 mM NaCl, (6) 25 mM NaBu, and (7) 25 mM NaBu plus 5 mM DCA (all 2 ml, pH 7.0). All chemicals were obtained from Sigma Chemical Co., St. Louis, MO. Analysis of whole crypt and crypt-compartment proliferation. Twelve longitudinally sectioned crypts were assessed per rat since it has been shown by the use of the running average statistical procedure that 12 crypts are sufficient to reliably determine proliferation rates [15]. Care was taken to assess crypts from at least nine different transverse levels of the colonic segment. The number of BrDU-labeled cells, and the total number of cells were counted according to the criteria described by Lipkin et al. [2].
FIG. 2. Butyrate–deoxycholate in vivo effects on bromodeoxyuridine (BrDU) labeling index (LI) by compartments. Base (1) to surface (5). Data expressed as means { SEM; n Å 8 for each group; *P £ 0.05 vs. NaCl; P £ 0.05 vs. NaBu, by ANOVA.
AID
JSR 5075
/
6n1e$$$442
06-02-97 15:03:39
345
FIG. 3. Butyrate–deoxycholate in vivo effects on bromodeoxyuridine (BrDU) labeling index (LI) by compartments. Base (1) to surface (5). Data expressed as means { SEM; n Å 8 for each group; *P £ 0.05 vs. NaCl; P £ 0.05 vs. NaBu, by ANOVA.
Proliferation of the whole colonic crypt (Total LI) was calculated as the number of labeled cells divided by the total number of cells. The crypt was subdivided into five longitudinal compartments (Fig. 1) from base (compartment 1) to surface (compartment 5) as described by Lipkin et al. [2]; see Fig. 2. Labeling indices were calculated for each of the five compartments as labeled cells within each compartment divided by total counted cells within the compartment. In addition, the combined lower 60% labeling index (Base LI) and the combined upper 40% labeling index (Surface LI) were separately determined as labeled cells divided by total cells in the base or surface proliferative zones, respectively. The fh value, regarded as an excellent discriminator for increased colon cancer risk [1], was calculated as the ratio of labeled cells in compartments 4 / 5 divided by the number of labeled cells in the entire crypt. Statistical analysis. All values are presented as means { SEM. The differences between calculated means among all study groups were examined by analysis of the variance (ANOVA). A matrix of pairwise comparison probabilities was calculated by Fisher’s leastsignificant difference test. The statistical software package SYSTAT (SYSTAT Inc.) was used for analysis. Differences were considered statistically significant at P £ 0.05.
RESULTS
Twenty-four-hour in vivo exposure of normal rat colonic mucosa to butyrate and deoxycholate significantly affected crypt proliferation in a compartment-specific manner (Figs. 2–4). Key comparative findings on proliferation effects are depicted graphically in Figs. 2–6. All groups showed a reproducible proliferative pattern
FIG. 4. Butyrate–deoxycholate in vivo effects on BrdU labeling index. Base, compartments 1–3; surface, compartments 4 and 5; total, compartments 1–5. Data expressed as means { SEM; n Å 8 per group; *P £ 0.05 vs. NaCl; **P £ 0.05 vs. NaBu, by ANOVA.
srga
346
JOURNAL OF SURGICAL RESEARCH: VOL. 69, NO. 2, MAY 1997
FIG. 5. In vivo effect of butyrate on the fh value by BrdU (bromodeoxyuridine) immunohistochemistry. Data expressed as means { SEM; n Å 8 for each group; *P Å 0.022 (t test).
according to crypt compartment (Figs. 2 and 3) with the highest labeling indices seen in compartment 3 and a progressive decline in labeling index in the surface compartments 4 and 5. Treatment with NaBu (10 and 25 mM) increased total LI when compared to NaCl (Figs. 2–4). This proliferative effect was significant only at the crypt base (Fig. 4). Both doses of NaBu resulted in similar LI with no further proliferative response at the higher concentration (Fig. 4). In contrast to prior in vitro studies [6, 7], DCA alone at this concentration (5 mM), did not affect LI, but when combined with NaBu, DCA significantly inhibited the proliferative effects of NaBu (10 and 25 mM) at the crypt base (Fig. 4). Interestingly, the surface LI was also significantly reduced by the combination of 10 mM NaBu plus 5 mM DCA compared to 10 mM NaBu alone (Fig. 5). Butyrate (25 mM) significantly decreased surface proliferation as measured by the fh value (Fig. 5), when compared to the 25 mM sodium chloride control. Interestingly the 25 mM NaCl group had a significantly increased fh value when compared to the 10 mM NaCl group (Fig. 6). DISCUSSION
This study indicates that butyrate and deoxycholate have selective and antagonistic, compartment-specific effects on colonic crypt proliferation in vivo. These findings may be relevant to colon cancer prevention. The fh value (an index of crypt surface hyperproliferation) is increased in the flat, ‘‘grossly normal’’ mucosa of human subjects at increased risk of developing colon cancer such as familial polyposis coli, Gardner’s syndrome, and ulcerative colitis [1, 3, 16] and in the mucosa of rodents treated with chemical carcinogens [5]. The data demonstrate that butyrate at a concentration of 25 mM significantly decreases the fh value compared to equimolar sodium chloride (25 mM NaCl) treatments. Our in vivo results, in normal rat colon, are consistent with in vitro findings in normal human colon biopsies [6, 7, 17]. We now confirm with this in vivo rat model the prior in vitro observation that butyrate exerts a proliferative effect that is confined to the base
AID
JSR 5075
/
6n1e$$$442
06-02-97 15:03:39
(lower 60%) compartments of the colonic crypt [17]. Also consistent with our findings are the in vivo effects seen in ulcerative colitis patients receiving butyrate enemas [18]. Ulcerative colitis represents a state of abnormal colonic hyperproliferation that leads to an increased risk of developing colon cancer. Treatment with butyrate enemas (100 mmol/L) for 2 weeks decreases the surface-labeling index in the colonic crypts of these ulcerative colitis patients to the level of that seen in normal subjects [18]. Therefore, we believe that the results observed in our in vivo rat model are consistent with those observed in human colon in vitro and in vivo. The observation that in vivo 25 mM intraluminal sodium chloride treatment results in significantly higher fh values than 10 mM sodium chloride exposure, in the rat colon, is a novel finding that requires further study. Prior studies have indicated that sodium levels greater than 25 mM may increase proliferation [19] in human colonic epithelial cells from high-risk patients. However, in vitro studies in normal human colon biopsies have documented that increasing levels of sodium exposure (up to 95 mM NaCl) have no effect on colonic crypt cell proliferation [17, 20]. To our knowledge this is the first report documenting in vivo effects of sodium on the colonic crypt proliferation pattern. The observed interaction between butyrate and deoxycholate at the concentrations tested are also novel findings. In contrast to prior in vitro findings [6], a low physiological level of deoxycholate (5 mM) does not significantly increase labeling indices in the colonic crypt. Moreover, our data indicate that deoxycholate inhibits the butyrate-induced increases in labeling indices. This effect may indicate that at low physiological levels, deoxycholate may have a role on maintaining the normal proliferative pattern of the colonic epithelium. Alternatively, deoxycholate’s cytotoxic effects [21, 22] may interfere with the known trophic effects of butyrate [23] on the colonic epthelium. The trophic effects of butyrate and other short-chain fatty acids on the colonic mucosa are thought to be physiologic and essential in maintaining the normal integrity of the colonic epithelium [17, 23]. Several prior studies [8, 17, 24] as well as this current work confirm the observation that butyrate-induced increases in crypt labeling indices are selective to the basal 60% of the crypt or ‘‘the
FIG. 6. In vivo effect of sodium on fh value by BrDU immunohistochemistry. Data expressed as means { SEM; n Å 8 per group; *P Å 0.004 (t test).
srga
´ ZQUEZ ET AL.: DEOXYCHOLATE AND BrDU LABELING VELA
normal proliferative zone,’’ while the labeling indices for the crypt surface are decreased. Prior studies suggest that these butyrate-induced effects are probably protective against colon carcinogenesis [8, 9]. While our in vivo model essentially creates a colonic closed loop obstruction for 24 hr, extensive preliminary studies from our laboratory have demonstrated that the baseline bromodeoxyuridine-labeling indices and the colonic crypt proliferation patterns by compartments are unchanged by the surgery compared with rats undergoing either no surgery or laparotomy only. Moreover, our preliminary evaluation of this surgical model showed that rats tolerate this 24-hr period of colonic obstruction without evidence of abdominal distention or systemic illness. The in situ isolated rat colon absorbs the injected 2-ml treatment solution, such that at the time of harvest, a decompressed colon has been consistently observed. Also, at the time of harvest, the small bowel is only mildly dilated. Therefore, we feel that for a short study period of 24 hr, this in vivo model is adequate and valid for the study of acute effects of the tested agents (butyrate, deoxycholate, and sodium) on colonic crypt proliferation patterns. Further studies using longer duration in vivo models and in vivo animal models of chemically induced colon carcinogenesis are indicated to delineate if these acute effects of butyrate, deoxycholate, and sodium on colonic crypt proliferation pattern bear relevance to colon cancer risk in humans. A recent study using butyrate enemas in experimental colitis has documented butyrate-induced protection against large bowel cancer in rats [8]. In this in vivo rat model [8] butyrate therapy reduced the incidence and size of colonic tumors and also affected colonic proliferation pattern with significant reductions in the labeling indices at the crypt surface (compartments 4 and 5). The mechanisms by which butyrate and deoxycholate modulate in vivo colon crypt proliferation in the observed compartment-specific manner are not known but are probably ultimately mediated by direct or indirect effects on gene expression [24–28]. Further studies are needed to determine if the observed interaction between butyrate and deoxycholate represents a colon cancer protective or promoting effect. Also, the mechanism(s) to explain these in vivo interactions requires further investigation. We conclude that: (1) the in vivo proliferative effects of sodium butyrate are selective to the crypt base, (2) an in vivo low physiological deoxycholate level does not promote crypt surface hyperproliferation but does inhibit butyrate’s proliferative effect, (3) sodium butyrate and deoxycholate interact in a complex and antagonistic manner to selectively modulate crypt base and surface proliferation, in the rat colon, in vivo, and (4) the level of intraluminal sodium may significantly affect the crypt proliferative pattern in vivo. These findings may have clinical relevance since colonic levels of butyrate, deoxycholate, and sodium can be affected by diet.
AID
JSR 5075
/
6n1e$$$443
06-02-97 15:03:39
347
REFERENCES 1. Lipkin, M., Blattner, W. E., and Fraumeni, J. F. Tritiated thymidine (fp , fh) labeling distribution as a marker for hereditary predisposition to colon cancer. Cancer Res. 43: 1899, 1983. 2. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinal cancer: Their development and application to studies of cancer prevention. Gastroenterology 92: 1083, 1987. 3. Lipkin, M., Enker, W., and Winawer, S. J. Tritiated-thymidine labeling of rectal epithelial cells in ‘‘non-prep’’ biopsies of individuals at increase risk for colonic neoplasia. Cancer Lett. 37: 153, 1987. 4. Pandey, S., Gordon, P. H., and Wang, E. Expression of proliferation-specific genes in the mucosa adjacent to colon carcinoma. Dis. Colon. Rectum 38: 462, 1995. 5. Deschner, E. E., and Maskens, A. P. Significance of the labeling index and labeling distribution as kinetic parameters in colorectal mucosa of cancer patients and DMH treated animals. Cancer 50: 1136, 1982. 6. Bartram, H-P., Englert, S., Scheppach, W., et al. Antagonistic effects of deoxycholic acid and butyrate on epithelial cell proliferation in the proximal and distal human colon. Z. Gastroenterol. 32: 389, 1994. 7. Bartram, H-P., Scheppach, W., Schmid, H., et al. Proliferation of human colonic mucosa as an intermediate biomarker of carcinogenesis: Effects of butyrate, deoxycholate, calcium, ammonia, and pH. Cancer Res. 53: 3283, 1993. 8. D’Argenio, G., Cosenza, V., Delle Cave, M., Iovino, P., Della Valle, N., Lombardi, G., and Mazzacca, G. Butyrate enemas in experimental colitis and protection against large bowel cancer in a rat model. Gastroenterology 110: 1727, 1996. 9. McIntyre, A., Gibson, P. R., and Young, G. P. Butyrate production from dietary fibre and protection against large bowel cancer in a rat model. Gut 34: 386, 1993. 10. DeCosse, J. J., Miller, H. H., and Lesser, M. L. Effect of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenopolyposis. J. Natl. Cancer Inst. 81: 1290, 1989. 11. Reddy, B. S., Weisburger, J. H., and Wynder, E. L. Colon cancer: bile salts as tumor promoters. In T. J. Salga, A. Sivak, and R. K. Boutwell (Eds.), Carcinogenesis. New York: Raven Press, 1978. Pp. 453–464. 12. Howe, G. R., Benito, E., Castelleto, R., et al. Dietary intake of fiber and decreased risk of cancers of the colon and rectum: Evidence from the combined analysis of 13 case-control studies. J. Natl. Cancer Inst. 84: 1887, 1992. 13. Yu, H., Harris, R. E., Gao, Y. T., et al. Comparative epidemiology of cancers of the colon, rectum, prostate and breast in Shanghai, China versus the United States. Int. J. Epidemiol. 20: 76, 1990. 14. Qin, Y., and Willems, G. Comparison of the classical autoradiographic and the immunohistochemical methods with BrDU for measuring proliferation parameters in colon cancer. Anticancer Res. 13: 731, 1993. 15. Welberg, J. W., deVries, E. G. E., Hardonk, M. J., et al. Proliferation rate of colonic mucosa in normal subjects and patients with colonic neoplasms: a refined immunohistochemical method. J. Clin. Pathol. 43: 453, 1990. 16. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinal cancer: New application to studies of cancer prevention in human subjects. Cancer Res. 48: 235, 1988. 17. Scheppach, W., Bartram, P., Richter, A., et al. Effect of shortchain fatty acids on the human colonic mucosa in vitro. J. Parenter. Enteral. Nutr. 16: 43, 1992. 18. Scheppach, W., Sommer, H., Kirchner, T., et al. Effect of butyrate enemas on the colonic mucosa in distal uncerative colitis. Gastroenterology 103: 51, 1992. 19. Friedman, E., Lightdale, C., and Winawer, S. Effects of psyllium fiber and short-chain organic acids derived from fiber break-
srga
348
JOURNAL OF SURGICAL RESEARCH: VOL. 69, NO. 2, MAY 1997 down on colonic epithelial cells from high-risk patients. Cancer Lett. 43: 121, 1988.
20. Bartam, H., Scheppach, W., Schmid, H., Hofmann, A., Dusel, G., Richter, F., Richter, A., and Kasper, H. Proliferation of human colonic mucosa as an intermediate biomarker of carcinogenesis: Effects of butyrate, deoxycholate, calcium, ammonia, and pH. Cancer Res. 53: 3283, 1993.
25.
21. Rafter, J. J., and Branting, C. Bile acids—Interaction with the intestinal mucosa. Eur. J. Cancer Prev. 1(Suppl. 2): 49, 1991.
26.
22. Craven, P. A., Pfanstiel, J., Saito, R., et al. Relationship between loss of rat colonic surface epithelium induced by deoxycholate and initiation of the subsequent proliferative response. Cancer Res. 46: 5754, 1986.
27.
23. Kripke, S. A., Fox, A. D., Berman, J. M., et al. Stimulation of intestinal mucosal growth with intracolonic infusion of shortchain fatty acids 1. J. Parenter. Enteral. Nutr. 13: 109, 1989. 24. Velazquez, O. C., Zhou, D., Seto, R. W., Jabbar, A., Choi, J.,
AID
JSR 5075
/
6n1e$$$444
06-02-97 15:03:39
28.
Lederer, H., and Rombeau, J. L. In Vivo crypt surface hyperproliferation is decreased by butyrate and increased by deoxycholate in normal rat colon: Associated in vivo effects on c-Fos and c-Jun expression. J. Parenteral. Enteral. Nutr. 20: 243, 1996. Velazquez, O. C., Seto, R., and Rombeau, J. L. The scientific rationale and clinical application of short-chain fatty acids and medium-chain triacylglycerols. Proc. Nutr. Soc. 55: 49, 1996. Souleimani, A., and Asselin, C. Regulation of c-myc expression by sodium butyrate in the colon carcinoma cell line Caco-2. FEBS Lett. 326: 45, 1993. Tang, S. J., Huang, Y. M., and Wang, F. F. Analysis of c-fos expression in the butyrate-induced F-98 glioma cell differentiation. Biochem. J. 306: 47, 1995. Nishina, Y., Sumi, T., Souichi, A., et al. The induction of jun genes during the reversible changes induced with sodium butyrate on the differentiation of F9 cells. Exp. Cell Res. 208: 492, 1993.
srga
69, 344–348 (1997)
JR975075
Deoxycholate Inhibits in Vivo Butyrate-Mediated BrDU Labeling of the Colonic Crypt Omaida C. Vela´zquez, M.D.,*† Rene´e W. Seto, B.A.,* Allison M. Bain, B.A.,* Jason Fisher, B.A.,* and John L. Rombeau, M.D.†,1 *Harrison Department of Surgical Research, University of Pennsylvania Medical Center, Philadelphia, Pennsylvania 19104; and †Department of Surgery, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19104 Presented at the Annual Meeting of the Association for Academic Surgery, Chicago, Illinois, November 13–16, 1996
The short-chain fatty acid butyrate (NaBu) selectively increases colonic crypt base proliferation and inhibits ‘‘premalignant’’ crypt surface hyperproliferation while the secondary bile acid deoxycholate (DCA) induces surface hyperproliferation, in vitro. We hypothesized that NaBu and DCA have similar selective and antagonistic effects on the colonic crypt proliferative pattern, in vivo. Fifty-six adult SD rats underwent surgical isolation of the colon and 24-hr intraluminal instillation with physiological (10 mM) and pharmacological (25 mM) levels of butyrate alone or combined with a physiological DCA level (5 mM). Bromodeoxyuridine-labeling indices (LI) were determined as labeled cells divided by total cells, for the whole crypt and five crypt compartments from base to surface. Treatment with NaBu increased total LI when compared to NaCl. This effect was significant only at the crypt base. Both doses of NaBu resulted in similar LI with no further response at the higher concentration. In contrast to prior in vitro studies, DCA alone at this concentration did not affect LI, but when combined with NaBu, DCA inhibited the effects of NaBu at the crypt base and surface. The conclusions are: (1) the in vivo proliferative effects of NaBu are selective to the crypt base, (2) an in vivo low physiological DCA level does not promote crypt surface hyperproliferation but does inhibit butyrate’s proliferative effect, and (3) NaBu and DCA interact in a complex and antagonistic manner to selectively modulate crypt base and surface proliferation, in the rat colon, in vivo. These findings may have clinical relevance since colonic levels of NaBu and DCA are affected by diet. q 1997 Academic Press
INTRODUCTION
Total crypt and crypt surface hyperproliferation occur in the colonic mucosa of humans at increased risk of colon cancer [1–4]. A shift in the proliferative zone 1 To whom correspondence should be addressed at Department of Surgery, 4 Silverstein Pavilion, 3400 Spruce Street, Hospital of the University of Pennsylvania, Philadelphia, PA 19104. Fax: 215-3498996.
0022-4804/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
JSR 5075
/
6n1e$$$441
from the base to the surface of the crypt as assessed by increased fh value (labeled cells in the upper 40% of the crypt divided by labeled cells in the whole crypt) is an excellent predictor of cancer risk in humans [1]. In animal models of chemically induced carcinogenesis, similar cytokinetic abnormalities have been observed [5]. The short-chain fatty acid butyrate and the secondary bile acid deoxycholate are diet by-products normally present within the colon luminal environment and are known to modulate human colonic crypt proliferation in vitro in an interactive and compartmentspecific manner [6]. While butyrate enhances crypt base proliferation, deoxycholate increases total crypt and crypt surface proliferation [7]. Moreover, co-incubation with butyrate and deoxycholate reverses the deoxycholate-induced surface hyperproliferation in proximal and distal human colon in vitro [6]. These findings have generated great interest since butyrate and deoxycholate have been implicated in colon cancer prevention and promotion, respectively [8–11] and since colon cancer incidence has been strongly associated with dietary practices [12, 13]. The purpose of this study was to determine whether the observed in vitro antagonistic effects of butyrate and deoxycholate on colonic crypt labeling indices also occur in vivo. MATERIALS AND METHODS Animals. The experimental protocol was approved by the University of Pennsylvania Institutional Animal Care and Use Committee. Fifty-six adult healthy male Sprague–Dawley rats (250–350 g) were housed in conventional animal facilities, three rats per cage. The rats were fed fiber-free diet (Diet No. 112636, Dyets Inc., Bethlehem, PA) and water ad libitum for 5 days prior to surgery. Cages contained raised bottoms so that animals could not reach and ingest their cage bedding. Solid food was withheld the night prior to surgery. All animals were operated on in groups of four to six per day, taking care to include one to two animals from each treatment group at each day of surgery until all groups were completed. Surgical procedures. Rats were anesthetized with Nembutal (50 mg/kg) intraperitoneally. The abdomen was cleansed with Betadine. A laparotomy was performed with a full midline incision. A 3-0 silk ligature was used to ligate the proximal colon immediately distal to the cecum. The colon was flushed with saline (377C) injected into the proximal colon, just distal to the ileo-cecal junction, with a 27-gauge needle. All the fluid and particulate material were carefully evacu-
344
06-02-97 15:03:39
srga
´ ZQUEZ ET AL.: DEOXYCHOLATE AND BrDU LABELING VELA
FIG. 1. Diagram of colonic crypt compartments.
ated from the colonic lumen. A 3-0 silk purse string ligature was place at the anus to prevent escape of the treatment solution. The treatment solution (2 ml) was injected with a 27-gauge needle through the same puncture site used for the flushing solution. The puncture site was closed with a 6-0 prolene suture. Care was taken not to disturb the blood supply to the cecum or ascending colon. The animals were resuscitated with 10 cc of warm sterile saline intraperitoneally at the end of the procedure. The abdomen was closed in a single layer with a running 3-0 silk suture. Rats were allowed to recover from the general anesthetic under a warming light and then were returned to raised bottom cages with access to only water ad libitum. Twenty-two hours after instillation of the treatment solution (2 hr prior to sacrificing the rat and harvesting the colon), the rats received a 2 ml/100 g intraperitoneal injection with undiluted Cell Proliferation Labeling Reagent Bromodeoxyuridine (Amersham Life Science). Twenty-four hours after instillation of the treatment solution, rats were anesthetized with Nembutal (50 mg/ kg) intraperitoneally and the colon was harvested. Animals were then killed by cardiac puncture. The colon was immediately flushed with 10 cc of cold saline. A standard 2-cm segment starting 1-cm distal to the injection site (marked by the prolene suture) was placed in 10% formalin for bromodeoxyuridine immunohistochemistry [14]. Preparation of treatment solutions. Individual solutions were prepared fresh each day for each treatment group: (1) 10 mM NaCl, (2) 10 mM NaBu, (3) 5 mM DCA, (4) 10 mM NaBu plus 5 mM DCA, (5) 25 mM NaCl, (6) 25 mM NaBu, and (7) 25 mM NaBu plus 5 mM DCA (all 2 ml, pH 7.0). All chemicals were obtained from Sigma Chemical Co., St. Louis, MO. Analysis of whole crypt and crypt-compartment proliferation. Twelve longitudinally sectioned crypts were assessed per rat since it has been shown by the use of the running average statistical procedure that 12 crypts are sufficient to reliably determine proliferation rates [15]. Care was taken to assess crypts from at least nine different transverse levels of the colonic segment. The number of BrDU-labeled cells, and the total number of cells were counted according to the criteria described by Lipkin et al. [2].
FIG. 2. Butyrate–deoxycholate in vivo effects on bromodeoxyuridine (BrDU) labeling index (LI) by compartments. Base (1) to surface (5). Data expressed as means { SEM; n Å 8 for each group; *P £ 0.05 vs. NaCl; P £ 0.05 vs. NaBu, by ANOVA.
AID
JSR 5075
/
6n1e$$$442
06-02-97 15:03:39
345
FIG. 3. Butyrate–deoxycholate in vivo effects on bromodeoxyuridine (BrDU) labeling index (LI) by compartments. Base (1) to surface (5). Data expressed as means { SEM; n Å 8 for each group; *P £ 0.05 vs. NaCl; P £ 0.05 vs. NaBu, by ANOVA.
Proliferation of the whole colonic crypt (Total LI) was calculated as the number of labeled cells divided by the total number of cells. The crypt was subdivided into five longitudinal compartments (Fig. 1) from base (compartment 1) to surface (compartment 5) as described by Lipkin et al. [2]; see Fig. 2. Labeling indices were calculated for each of the five compartments as labeled cells within each compartment divided by total counted cells within the compartment. In addition, the combined lower 60% labeling index (Base LI) and the combined upper 40% labeling index (Surface LI) were separately determined as labeled cells divided by total cells in the base or surface proliferative zones, respectively. The fh value, regarded as an excellent discriminator for increased colon cancer risk [1], was calculated as the ratio of labeled cells in compartments 4 / 5 divided by the number of labeled cells in the entire crypt. Statistical analysis. All values are presented as means { SEM. The differences between calculated means among all study groups were examined by analysis of the variance (ANOVA). A matrix of pairwise comparison probabilities was calculated by Fisher’s leastsignificant difference test. The statistical software package SYSTAT (SYSTAT Inc.) was used for analysis. Differences were considered statistically significant at P £ 0.05.
RESULTS
Twenty-four-hour in vivo exposure of normal rat colonic mucosa to butyrate and deoxycholate significantly affected crypt proliferation in a compartment-specific manner (Figs. 2–4). Key comparative findings on proliferation effects are depicted graphically in Figs. 2–6. All groups showed a reproducible proliferative pattern
FIG. 4. Butyrate–deoxycholate in vivo effects on BrdU labeling index. Base, compartments 1–3; surface, compartments 4 and 5; total, compartments 1–5. Data expressed as means { SEM; n Å 8 per group; *P £ 0.05 vs. NaCl; **P £ 0.05 vs. NaBu, by ANOVA.
srga
346
JOURNAL OF SURGICAL RESEARCH: VOL. 69, NO. 2, MAY 1997
FIG. 5. In vivo effect of butyrate on the fh value by BrdU (bromodeoxyuridine) immunohistochemistry. Data expressed as means { SEM; n Å 8 for each group; *P Å 0.022 (t test).
according to crypt compartment (Figs. 2 and 3) with the highest labeling indices seen in compartment 3 and a progressive decline in labeling index in the surface compartments 4 and 5. Treatment with NaBu (10 and 25 mM) increased total LI when compared to NaCl (Figs. 2–4). This proliferative effect was significant only at the crypt base (Fig. 4). Both doses of NaBu resulted in similar LI with no further proliferative response at the higher concentration (Fig. 4). In contrast to prior in vitro studies [6, 7], DCA alone at this concentration (5 mM), did not affect LI, but when combined with NaBu, DCA significantly inhibited the proliferative effects of NaBu (10 and 25 mM) at the crypt base (Fig. 4). Interestingly, the surface LI was also significantly reduced by the combination of 10 mM NaBu plus 5 mM DCA compared to 10 mM NaBu alone (Fig. 5). Butyrate (25 mM) significantly decreased surface proliferation as measured by the fh value (Fig. 5), when compared to the 25 mM sodium chloride control. Interestingly the 25 mM NaCl group had a significantly increased fh value when compared to the 10 mM NaCl group (Fig. 6). DISCUSSION
This study indicates that butyrate and deoxycholate have selective and antagonistic, compartment-specific effects on colonic crypt proliferation in vivo. These findings may be relevant to colon cancer prevention. The fh value (an index of crypt surface hyperproliferation) is increased in the flat, ‘‘grossly normal’’ mucosa of human subjects at increased risk of developing colon cancer such as familial polyposis coli, Gardner’s syndrome, and ulcerative colitis [1, 3, 16] and in the mucosa of rodents treated with chemical carcinogens [5]. The data demonstrate that butyrate at a concentration of 25 mM significantly decreases the fh value compared to equimolar sodium chloride (25 mM NaCl) treatments. Our in vivo results, in normal rat colon, are consistent with in vitro findings in normal human colon biopsies [6, 7, 17]. We now confirm with this in vivo rat model the prior in vitro observation that butyrate exerts a proliferative effect that is confined to the base
AID
JSR 5075
/
6n1e$$$442
06-02-97 15:03:39
(lower 60%) compartments of the colonic crypt [17]. Also consistent with our findings are the in vivo effects seen in ulcerative colitis patients receiving butyrate enemas [18]. Ulcerative colitis represents a state of abnormal colonic hyperproliferation that leads to an increased risk of developing colon cancer. Treatment with butyrate enemas (100 mmol/L) for 2 weeks decreases the surface-labeling index in the colonic crypts of these ulcerative colitis patients to the level of that seen in normal subjects [18]. Therefore, we believe that the results observed in our in vivo rat model are consistent with those observed in human colon in vitro and in vivo. The observation that in vivo 25 mM intraluminal sodium chloride treatment results in significantly higher fh values than 10 mM sodium chloride exposure, in the rat colon, is a novel finding that requires further study. Prior studies have indicated that sodium levels greater than 25 mM may increase proliferation [19] in human colonic epithelial cells from high-risk patients. However, in vitro studies in normal human colon biopsies have documented that increasing levels of sodium exposure (up to 95 mM NaCl) have no effect on colonic crypt cell proliferation [17, 20]. To our knowledge this is the first report documenting in vivo effects of sodium on the colonic crypt proliferation pattern. The observed interaction between butyrate and deoxycholate at the concentrations tested are also novel findings. In contrast to prior in vitro findings [6], a low physiological level of deoxycholate (5 mM) does not significantly increase labeling indices in the colonic crypt. Moreover, our data indicate that deoxycholate inhibits the butyrate-induced increases in labeling indices. This effect may indicate that at low physiological levels, deoxycholate may have a role on maintaining the normal proliferative pattern of the colonic epithelium. Alternatively, deoxycholate’s cytotoxic effects [21, 22] may interfere with the known trophic effects of butyrate [23] on the colonic epthelium. The trophic effects of butyrate and other short-chain fatty acids on the colonic mucosa are thought to be physiologic and essential in maintaining the normal integrity of the colonic epithelium [17, 23]. Several prior studies [8, 17, 24] as well as this current work confirm the observation that butyrate-induced increases in crypt labeling indices are selective to the basal 60% of the crypt or ‘‘the
FIG. 6. In vivo effect of sodium on fh value by BrDU immunohistochemistry. Data expressed as means { SEM; n Å 8 per group; *P Å 0.004 (t test).
srga
´ ZQUEZ ET AL.: DEOXYCHOLATE AND BrDU LABELING VELA
normal proliferative zone,’’ while the labeling indices for the crypt surface are decreased. Prior studies suggest that these butyrate-induced effects are probably protective against colon carcinogenesis [8, 9]. While our in vivo model essentially creates a colonic closed loop obstruction for 24 hr, extensive preliminary studies from our laboratory have demonstrated that the baseline bromodeoxyuridine-labeling indices and the colonic crypt proliferation patterns by compartments are unchanged by the surgery compared with rats undergoing either no surgery or laparotomy only. Moreover, our preliminary evaluation of this surgical model showed that rats tolerate this 24-hr period of colonic obstruction without evidence of abdominal distention or systemic illness. The in situ isolated rat colon absorbs the injected 2-ml treatment solution, such that at the time of harvest, a decompressed colon has been consistently observed. Also, at the time of harvest, the small bowel is only mildly dilated. Therefore, we feel that for a short study period of 24 hr, this in vivo model is adequate and valid for the study of acute effects of the tested agents (butyrate, deoxycholate, and sodium) on colonic crypt proliferation patterns. Further studies using longer duration in vivo models and in vivo animal models of chemically induced colon carcinogenesis are indicated to delineate if these acute effects of butyrate, deoxycholate, and sodium on colonic crypt proliferation pattern bear relevance to colon cancer risk in humans. A recent study using butyrate enemas in experimental colitis has documented butyrate-induced protection against large bowel cancer in rats [8]. In this in vivo rat model [8] butyrate therapy reduced the incidence and size of colonic tumors and also affected colonic proliferation pattern with significant reductions in the labeling indices at the crypt surface (compartments 4 and 5). The mechanisms by which butyrate and deoxycholate modulate in vivo colon crypt proliferation in the observed compartment-specific manner are not known but are probably ultimately mediated by direct or indirect effects on gene expression [24–28]. Further studies are needed to determine if the observed interaction between butyrate and deoxycholate represents a colon cancer protective or promoting effect. Also, the mechanism(s) to explain these in vivo interactions requires further investigation. We conclude that: (1) the in vivo proliferative effects of sodium butyrate are selective to the crypt base, (2) an in vivo low physiological deoxycholate level does not promote crypt surface hyperproliferation but does inhibit butyrate’s proliferative effect, (3) sodium butyrate and deoxycholate interact in a complex and antagonistic manner to selectively modulate crypt base and surface proliferation, in the rat colon, in vivo, and (4) the level of intraluminal sodium may significantly affect the crypt proliferative pattern in vivo. These findings may have clinical relevance since colonic levels of butyrate, deoxycholate, and sodium can be affected by diet.
AID
JSR 5075
/
6n1e$$$443
06-02-97 15:03:39
347
REFERENCES 1. Lipkin, M., Blattner, W. E., and Fraumeni, J. F. Tritiated thymidine (fp , fh) labeling distribution as a marker for hereditary predisposition to colon cancer. Cancer Res. 43: 1899, 1983. 2. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinal cancer: Their development and application to studies of cancer prevention. Gastroenterology 92: 1083, 1987. 3. Lipkin, M., Enker, W., and Winawer, S. J. Tritiated-thymidine labeling of rectal epithelial cells in ‘‘non-prep’’ biopsies of individuals at increase risk for colonic neoplasia. Cancer Lett. 37: 153, 1987. 4. Pandey, S., Gordon, P. H., and Wang, E. Expression of proliferation-specific genes in the mucosa adjacent to colon carcinoma. Dis. Colon. Rectum 38: 462, 1995. 5. Deschner, E. E., and Maskens, A. P. Significance of the labeling index and labeling distribution as kinetic parameters in colorectal mucosa of cancer patients and DMH treated animals. Cancer 50: 1136, 1982. 6. Bartram, H-P., Englert, S., Scheppach, W., et al. Antagonistic effects of deoxycholic acid and butyrate on epithelial cell proliferation in the proximal and distal human colon. Z. Gastroenterol. 32: 389, 1994. 7. Bartram, H-P., Scheppach, W., Schmid, H., et al. Proliferation of human colonic mucosa as an intermediate biomarker of carcinogenesis: Effects of butyrate, deoxycholate, calcium, ammonia, and pH. Cancer Res. 53: 3283, 1993. 8. D’Argenio, G., Cosenza, V., Delle Cave, M., Iovino, P., Della Valle, N., Lombardi, G., and Mazzacca, G. Butyrate enemas in experimental colitis and protection against large bowel cancer in a rat model. Gastroenterology 110: 1727, 1996. 9. McIntyre, A., Gibson, P. R., and Young, G. P. Butyrate production from dietary fibre and protection against large bowel cancer in a rat model. Gut 34: 386, 1993. 10. DeCosse, J. J., Miller, H. H., and Lesser, M. L. Effect of wheat fiber and vitamins C and E on rectal polyps in patients with familial adenopolyposis. J. Natl. Cancer Inst. 81: 1290, 1989. 11. Reddy, B. S., Weisburger, J. H., and Wynder, E. L. Colon cancer: bile salts as tumor promoters. In T. J. Salga, A. Sivak, and R. K. Boutwell (Eds.), Carcinogenesis. New York: Raven Press, 1978. Pp. 453–464. 12. Howe, G. R., Benito, E., Castelleto, R., et al. Dietary intake of fiber and decreased risk of cancers of the colon and rectum: Evidence from the combined analysis of 13 case-control studies. J. Natl. Cancer Inst. 84: 1887, 1992. 13. Yu, H., Harris, R. E., Gao, Y. T., et al. Comparative epidemiology of cancers of the colon, rectum, prostate and breast in Shanghai, China versus the United States. Int. J. Epidemiol. 20: 76, 1990. 14. Qin, Y., and Willems, G. Comparison of the classical autoradiographic and the immunohistochemical methods with BrDU for measuring proliferation parameters in colon cancer. Anticancer Res. 13: 731, 1993. 15. Welberg, J. W., deVries, E. G. E., Hardonk, M. J., et al. Proliferation rate of colonic mucosa in normal subjects and patients with colonic neoplasms: a refined immunohistochemical method. J. Clin. Pathol. 43: 453, 1990. 16. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinal cancer: New application to studies of cancer prevention in human subjects. Cancer Res. 48: 235, 1988. 17. Scheppach, W., Bartram, P., Richter, A., et al. Effect of shortchain fatty acids on the human colonic mucosa in vitro. J. Parenter. Enteral. Nutr. 16: 43, 1992. 18. Scheppach, W., Sommer, H., Kirchner, T., et al. Effect of butyrate enemas on the colonic mucosa in distal uncerative colitis. Gastroenterology 103: 51, 1992. 19. Friedman, E., Lightdale, C., and Winawer, S. Effects of psyllium fiber and short-chain organic acids derived from fiber break-
srga
348
JOURNAL OF SURGICAL RESEARCH: VOL. 69, NO. 2, MAY 1997 down on colonic epithelial cells from high-risk patients. Cancer Lett. 43: 121, 1988.
20. Bartam, H., Scheppach, W., Schmid, H., Hofmann, A., Dusel, G., Richter, F., Richter, A., and Kasper, H. Proliferation of human colonic mucosa as an intermediate biomarker of carcinogenesis: Effects of butyrate, deoxycholate, calcium, ammonia, and pH. Cancer Res. 53: 3283, 1993.
25.
21. Rafter, J. J., and Branting, C. Bile acids—Interaction with the intestinal mucosa. Eur. J. Cancer Prev. 1(Suppl. 2): 49, 1991.
26.
22. Craven, P. A., Pfanstiel, J., Saito, R., et al. Relationship between loss of rat colonic surface epithelium induced by deoxycholate and initiation of the subsequent proliferative response. Cancer Res. 46: 5754, 1986.
27.
23. Kripke, S. A., Fox, A. D., Berman, J. M., et al. Stimulation of intestinal mucosal growth with intracolonic infusion of shortchain fatty acids 1. J. Parenter. Enteral. Nutr. 13: 109, 1989. 24. Velazquez, O. C., Zhou, D., Seto, R. W., Jabbar, A., Choi, J.,
AID
JSR 5075
/
6n1e$$$444
06-02-97 15:03:39
28.
Lederer, H., and Rombeau, J. L. In Vivo crypt surface hyperproliferation is decreased by butyrate and increased by deoxycholate in normal rat colon: Associated in vivo effects on c-Fos and c-Jun expression. J. Parenteral. Enteral. Nutr. 20: 243, 1996. Velazquez, O. C., Seto, R., and Rombeau, J. L. The scientific rationale and clinical application of short-chain fatty acids and medium-chain triacylglycerols. Proc. Nutr. Soc. 55: 49, 1996. Souleimani, A., and Asselin, C. Regulation of c-myc expression by sodium butyrate in the colon carcinoma cell line Caco-2. FEBS Lett. 326: 45, 1993. Tang, S. J., Huang, Y. M., and Wang, F. F. Analysis of c-fos expression in the butyrate-induced F-98 glioma cell differentiation. Biochem. J. 306: 47, 1995. Nishina, Y., Sumi, T., Souichi, A., et al. The induction of jun genes during the reversible changes induced with sodium butyrate on the differentiation of F9 cells. Exp. Cell Res. 208: 492, 1993.
srga