Adult and pediatric gastroenterologists widely prescribe senna and cascara preparations, which act as laxatives because of their content of rhein and other anthraquinonic compounds (1–2). However, the prolonged use of or the frank misuse of these drugs may cause an inflamed colonic mucosa and an endoscopic picture of “brown bowel” or pseudomelanosis coli (3). The latter presents microscopically as leukocytes loaded with lipofuscinotic pigment and colonic epithelial cells undergoing apoptosis (3). However, it is unclear how rhein induces laxation, leukocyte recruiting, and colonocyte apoptosis and whether these effects share common pathogenic features. Recently, it has been suggested that nitric oxide, a reactive nitrogen species (RNS) mediates rhein-induced intestinal chloride secretion in a rat model (4). Reactive nitrogen species belong to the family of free radicals, that is, small, short-lived, molecules containing an unpaired electron to which they owe their oxidant power. Reactive nitrogen species recently have been shown to act as transducers of diverse cell functions (5). In this article, we hypothesize that rhein-generated RNS may be responsible for the effects of the drug on the intestinal mucosa, which can help to explain the composite picture that pathologists describe. To test this hypothesis, we used diverse electrophysiologic, biochemical, and cytochemical in vitro systems. Because prolonged culture of normal human intestine is not currently feasible, a well-characterized cell line such as human colonic adenocarcinoma (CaCo-2) has been chosen as a faithful model of human intestinal epithelium (6–7).
MATERIALS AND METHODS
Cell culture chemicals were obtained from GIBCO-Life Technologies (Milan, Italy). BAPTA/AM was purchased from Molecular Probes (Eugene, OR, U.S.A.); NG-nitro-L-arginine methyl ester (L-NAME) and highly purified rhein was obtained from Sigma (St. Louis, MO, U.S.A.). In this study, rhein powder dissolved in 2.5 mmol/L NaHCO3 and used shortly thereafter is referred to as native rhein, whereas the identical solution left to degrade for 18 hours is called degraded rhein. The latter was checked with the spectrophotometer for loss of absorbance peak at 442 nm before use in a biologic assay. Hoechst 33258 is a dye that specifically stains nuclei for morphologic evaluation of apoptotic features, such as chromatin condensation and nuclear fragmentation (8). Hoechst 33258 and other analytical-grade reagents also were purchased from Sigma.
Human colonic adenocarcinoma (CaCo-2) cells were purchased from the Istituto Zooprofilattico della Lombardia e dell' Emilia (Brescia, Italy). Cells were grown in Dulbecco modified Eagle medium containing 25 mmol/L glucose and supplemented with 10% fetal bovine serum, 1% nonessential amino acids, 2 mmol/L L-glutamine, 1% penicillin-streptomycin, and 1% sodium pyruvate. Cells were maintained in a humidified atmosphere of 5% CO2 in air at 37°C. Single-cell suspensions were obtained from 70% to 80% confluent cultures by incubation with 0.05% trypsin, and cells were then seeded at 105 cells/cm2 onto either 4-cm2 glass coverslips or detachable polycarbonate microporous cell culture inserts (Snapwells, 12 mm diameter, 0.4 μm pore size; Costar; Cambridge, MA, U.S.A.).
Ussing Chambers Experiments
Experiments were conducted as previously described by Raimondi et al. (9), but with some modifications. Snapwell inserts were mounted between the Perspex half-chambers of a modified Ussing chamber. Monolayers were bathed with Ringer solution (5 mL/half-chamber), kept at 37°C, and gassed with 5% CO2/95% O2. Ringer solution composition was (mmol/L): NaCl, 53; KCl, 5; Na2SO4, 30.5; mannitol, 30.5; Na 2HPO4, 1.69; NaH2PO4, 0.3; CaCl2, 1.25; MgCl2, 1.1; and NaHCO3, 25. In chloride-free experiments, chloride ions were replaced with an equivalent concentration of sulphate ions. Where appropriate, cell monolayers were preincubated at 37°C for 15 minutes with BAPTA/AM, a cell-permeant intracellular calcium buffer, or with L-NAME, a known inhibitor of nitric oxide synthesis. Transepithelial potential difference (PD) was recorded with a voltmeter-amperometer (World Precision Instruments, FL, U.S.A.), and electrical resistance (Rt) and short-circuit current (Isc) were calculated according to Ohm law. The maximal changes in these parameters recorded after the addition of an agent were defined as peak differences, and symbolized as ΔPD, ΔIsc, and ΔRt, respectively. At the end of each experiment, 10−4 mol/L epinephrine was added to the serosal side to check for a brisk response, to verify cell viability.
Nitrates Accumulation Experiments
Experiments were performed as described by Hevel and Marletta (10). Briefly, cells in phenol red-free Dulbecco modified Eagle medium were challenged with 50 μmol/L rhein for 30 minutes at 37°C in a humidified atmosphere of 5% CO2 in air. Culture supernatant was then collected, briefly spun at maximum speed in an Eppendorf tabletop centrifuge, and 105 μL of supernatant was then added in a 96-well plate to 10 μL of a solution composed of 25 μmol/L NADPH and 405 mU nitrate reductase. After incubation for 30 minutes at room temperature, 10 μL of a solution of 100 mmol/L ammonium chloride, 4 mmol/L α-ketoglutarate, and 10 mU glutamate dehydrogenase were added to each well and again incubated for 10 minutes at room temperature. Finally, 125 μL of reconstituted Griess reagent (0.05% naphtylenediamine and 0.5% sulphanilamide in 2.5% phosphoric acid) were added to each well and incubated at 37°C for 5 minutes. A standard curve was built using increasing concentrations of sodium nitrate. Results were read with a spectrophotometer (model DU640; Beckman Coulter Inc., Fullerton, CA, U.S.A.) at 543 nm.
Standard techniques were used to isolate human polymorphonuclear leukocytes (PMNs) from heparinized blood obtained by venipuncture from adult healthy donors. Briefly, blood was layered over Ficoll-Hypaque and centrifuged at 400 g for 30 minutes to remove lymphocytes and monocytes (11). Erythrocytes were removed by 2% dextran (m.w.: 480,000) sedimentation for 30 minutes at room temperature. Contaminating erythrocytes were lysed with cold 0.2% NaCl for exactly 30 seconds. Adding an equal volume of cold 1.6% NaCl restored isotonicity. Isolated PMNs were washed twice and suspended in RPMI supplemented with 10% FBS. Cell suspensions obtained by this technique were more than 95% pure and more than 98% viable, as determined by trypan blue exclusion dye test. Chemotaxis was performed in 200 μL masked well chambers (Neuro Probe Inc., Maryland, U.S.A.), using 3-μm polycarbonate filters. After incubation for 45 minutes at 37°C (5% CO2; 95% relative humidity), filters were removed and cells were fixed in 2.5% glutaraldehyde and stained with 10% Giemsa solution. The cells that migrated through the pores were counted at ×1,000 magnification from four randomly selected fields. Polymorphonuclear leukocytes that dropped off the membrane were recovered from the lower chamber, counted, and added to the total number of migrated cells. In a parallel series of experiments, cell viability at the end of incubation was verified by the trypan blue exclusion dye test. The known chemoattractant formyl-methionine-leucine-phenylmethionine peptide was used to normalize the data. Each sample was run in duplicate, and each experiment was performed at least five times on different days. In a separate set of experiments, neutralizing antibodies raised against chemokines IL-8 (monoclonal, IgG1, mouse antihuman antibody; Biosource, CA, U.S.A.) and ENA78 (polyclonal, IgG1, goat antihuman antibody; Santa Cruz, CA, U.S.A.) were used to inhibit rhein-induced chemotaxis, and isotypic anti–IL-15 antibody (Immunex, Washington, U.S.A.) was used as a negative control. More in detail, Caco-2 cell culture supernatants challenged with 50 μmol/L rhein were harvested and incubated for 20 minutes at 37°C with known antibody concentrations. Supernatants were then used in the chemotaxis assay.
Analysis of Rhein-induced Apoptosis
Experiments were performed as in Fiorentini et al. (8), with some modifications. CaCo-2 cells, 105, were seeded in a 60×15-mm Petri dish (Falcon, Plymouth, UK), cultured for 7 days, and then challenged with 50 μmol/L rhein with and without 100 μmol/L L-NAME for 14 hours. A 1-mL supernatant aliquot was spun and resuspended in 200 μL of a 1 mg/mL solution of Hoechst 33258 dye in PBS (pH 7.4). Fifty microliters of the cell suspension was then spread on a 3-aminopropylethoxysilane–treated glass slide, coverslips were mounted, and a masked, experienced morphologist (L.M.) scanned at least 10 fields at ×40 enlargement (total area, 0.5 mm2) for apoptotic enterocytes using a fluorescence microscope (Zeiss Axioscope, Federal Republic of Germany).
Results are presented as mean ± S.D. The data were analyzed using one-way analysis of variance, and P less than 0.05 was considered statistically significant.
Ussing Chambers Experiments
Historically, rhein is known as a laxative, and experimentally it has been shown to induce electrolyte secretion in animal models (12). Using a human-derived intestinal epithelium, as shown in Fig. 1A, progressive additions of rhein to the mucosal side of CaCo-2 monolayers produced a brisk increase in Isc with a maximum response at 50 μmol/L. The Isc increase is significantly less when substituting chloride ions with an equivalent amount of sulphate ions in the bathing buffer, an indirect but reliable way to demonstrate chloride secretion. Moreover, the electrical response is short lived, returning to baseline within 20 minutes. Figure 1B shows that the rhein-induced increase in Isc is inhibited by preincubation with both 2 μmol/L BAPTA/AM or 100 μmol/L L-NAME and that 50 μmol/L degraded rhein does not trigger any significant Isc increase.
The L-NAME-induced inhibition of rhein-mediated electrolyte secretion in a human-derived intestinal model led us to hypothesize that the drug could induce RNS in CaCo-2 cells. As showed in Figure 2, 30-minute incubation with 50 μmol/L native rhein elicited the accumulation of nitrates, as detected by the Griess reaction, whereas an equivalent concentration of degraded rhein gave a result not different from control values. Because a concentration of 50 μmol/L falls within the interval that many other investigators used (4,12–14), it can be defined as biologically relevant.
Leukocyte recruitment is a common pathologic finding of chronic anthraquinone abuse. We excluded the possibility that rhein by itself induced chemotaxis of PMNs. Figure 3A demonstrates that although rhein per se was unable to trigger PMN chemotaxis, CaCo-2 culture supernatants stimulated with 50 μmol/L rhein (but not with 50 μmol/L degraded rhein) elicited a time-dependent locomotion of PMNs. However, this phenomenon is minimal at 30 minutes, when chloride secretion is maximal, and it plateaued at 6 to 8 hours. As with ion secretion, chemotaxis also was inhibited by preincubating CaCo-2 cultures with 100 μmol/L L-NAME (Fig. 3B). The presence of a soluble inductor of chemotaxis and the relatively long time over which it occurred prompted us to think that chemoattractant interleukins (or chemokines) also may be involved. As shown in Figure 3C, neutralizing antibody anti–IL-8, a chemokine expressed by epithelial cells, inhibited chemotaxis elicited by culture supernatants of rhein-stimulated CaCo-2 cells. Chemotaxis was inhibited in a dose-dependent fashion by a neutralizing antibody anti-ENA78, an epithelium-specific chemokine. Comparing the two antichemokines antibody inhibition curves, a lower concentration of anti-ENA78 antibody was required to inhibit PMNs locomotion. These data demonstrate that a complex cross talk between the intestinal epithelium challenged with rhein and PMNs exists and recognizes several mediators.
Analysis of Rhein-induced Apoptosis
Colonocyte apoptosis is often observed in patients who abuse anthraquinones (3). To investigate in our in vitro model whether rhein could directly drive programmed cell death of intestinal cells, cultures were challenged for 14 hours with 50 μmol/L rhein with and without 100 mol/L L-NAME, and a 1-ml supernatant aliquot was spun, resuspended, and analyzed using a fluorescence microscope. Native rhein, 50 μmol/L, but not its degraded form, gave a significant number of apoptotic cells versus controls. Coincubation with L-NAME decreased rhein-induced apoptosis to a level comparable with controls (Fig. 4). In a separate series, incubation experiments for 6 hours (when the chemotactic effect reaches its plateau) with 50 μmol/L rhein showed no significant apoptosis induction (data not shown).
Senna preparations have been recommended for treating chronic constipation in childhood (2). Previous investigations on the laxative effects of senna, using different animal models, have shown that crude senna extracts and purified rhein cause brisk, short-lived, Ca2+-dependent intestinal chloride secretion (12). However, the detailed mechanism of action of the laxative is obscure. In particular, it is unclear whether a direct drug–epithelium interaction was sufficient to elicit intestinal secretion, and the pathway from which such a secretion arose also is unknown (15). Using a human epithelial model, our data demonstrate that rhein induces not only chloride secretion but also recruitment of PMNs, an early marker of inflammation and colonocyte apoptosis. Therefore, desired and undesired effects of rhein can occur when an intact molecule of the drug triggers the generation of RNS in the enterocyte. In fact, the undegraded form of rhein, like other naturally occurring quinones, interferes with the respiratory chain (4,14) and an altered electron flow can explain the formation of RNS documented by the nitrate-accumulation experiments. Reactive nitrogen species have already been shown to mediate separately intestinal ion secretion (7), human PMN chemotaxis (16), and enterocyte apoptosis (17). Intriguingly, our data clearly indicate that rhein-generated RNS modulate all three of these biologic effects of rhein, although with different time courses. In fact, ion secretion is immediate and short lived, caused by the direct effect of RNS. However, PMN chemotaxis reached its peak after a few hours and was abolished by neutralizing antibodies against the two chemokines IL-8 and ENA78. These observations strongly suggest that the role of nitrogen radicals in recruiting PMNs is indirect. This supports recent evidence that CaCo-2 cells produce both IL-8 and ENA78 (18), and that not only are these chemokines involved in chronic inflammation of the gut but they often are present simultaneously in the same pathologic specimen (19–20). Reactive nitrogen species also have been shown to modulate the production of cytokines (21).
Finally, a longer time is required for rhein-generated RNS to induce programmed cell death. Although we provide no molecular detail for rhein-induced apoptosis, there are several possible explanations for this event. Rhein interference with the respiratory chain leads to decreased adenosine triphosphate production in rat hepatocytes (22), and energy depletion has been linked to programmed cell death (18). Moreover, apoptosis can result from a nitrosate challenge characterized by the accrual of nitrosylated proteins (23) or by direct damage to cellular DNA by rhein-generated free radicals (18). Finally, RNS have also been postulated to mediate apoptosis through signals external to the cell such as the FAS pathway (18).
In conclusion, our in vitro data detail a unifying mechanism for both wanted and unwanted effects of a drug widely used in the management of infantile constipation and, therefore, may be of help to clinicians who routinely prescribe it.
The authors thank Dr. Alessio Fasano for helpful suggestions and for critically revising the manuscript.
1. Brunton LL. Agents affecting gastrointestinal water flux and motility, digestants and bile acids. In: Goodman Gilman A, Rall TW, Nies AS, et al, eds. The Pharmacological Basis of Therapeutics. New York: Pergamon Press; 1990:914–32.
2. Baker SS, Liptak GS, Colletti RB, et al. Constipation
in infant and children: evaluation and treatment. J Ped Gastroenterol Nutr 1999; 29:612–26.
3. Balazs M. Melanosis coli. Ultrastructural study of 45 patients. Dis Col Rectum 1986; 29:839–44.
4. Izzo AA, Gaginella TS, Mascolo N, et al. NG
-nitro-L-arginine methyl esther reduces senna- and cascara-induced diarrhoea and fluid secretion in the rat. Eur J Pharmacol 1996; 301:137–42.
5. Bredt DS. Endogenous nitric oxide
synthesis: biologic functions and pathophysiology. Free Radic Res 1999; 31:577–96.
6. Grasset E, Berenabeu J, Pinto M. Epithelial properties of human colonic carcinoma cell line CaCo-2: effect of secretagogues. Am J Physiol 1985; 248:C410–8.
7. Grasset E, Pinto M, Dussaulx E, et al. Epithelial properties of human colonic carcinoma cell line CaCo-2: electrical parameters. Am J Physiol 1984; 247:C260–7.
8. Fiorentini C, Fabbri A, Falzano L, et al. Clostridium difficile toxin B induces apoptosis
in intestinal cultured cells. Infect Immun 1998; 66:2660–5.
9. Raimondi F, Kao JP, Fiorentini C, et al. Enterotoxicity and cytotoxicity of Vibrio parahaemolyticus thermostable direct hemolysin in in vitro systems. Infect Immun 2000; 68:3180–5.
10. Hevel JM, Marletta MA. Nitric-oxide synthase assays. Methods Enzymol 1994; 233:250–8.
11. Viggiano D, Romano G, Caniglia M, et al. Impaired LTB4 release by neonatal polymorphonuclear leukocytes. Pediatr Res 1994; 36:60–3.
12. Goerg KJ, Wanitschke R, Schwarz M, et al. Rhein stimulates active chloride secretion in the short-circuited rat colonic mucosa. Pharmacology 1988; 36:111–9.
13. Bironaite D, Ollinger K. The hepatotoxicity of rhein involves impairment of mitochondrial functions. Chem Biol Interact 1997; 103:35–50.
14. Donowitz M, Wics J, Battisti L, et al. Effect of Senokot on rat intestinal electrolyte transport. Evidence of Ca2+
dependence. Gastroenterology 1984; 87:503–12.
15. Floridi A, Castiglione S, Bianchi C. Sites of inhibition of mitochondrial electron transport by rhein. Biochem Pharmacol 1989; 38:743–51.
16. Beauvais F, Lawrence M, Dubertret L. Exogenous nitric oxide
elicits chemotaxis of neutrophils in vitro. J Cell Physiol 1995; 165:610–4.
17. Murphy MP. Nitric oxide
and cell death. Biochim Biophys Acta 1999; 1411:401–14.
18. Remick DG, Villarete L. Regulation of cytokine gene expression by reactive oxygen and reactive nitrogen intermediates. J Leukoc Biol 1996; 59:471–5.
19. Keates S, Keates AC, Mizoguchi E, et al. Enterocytes are the primary source of the chemokine ENA-78 in normal colon and ulcerative colitis. Am J Physiol 1997; 273:G75–82.
20. MacDermott RP, Sanderson IR, Reinecker HC. The central role of chemokines (chemotactic cytokines) in the immunopathogenesis of ulcerative colitis and Crohn's disease. Inflamm Bowel Dis 1998; 4:54–67.
21. Muhl H, Nold M, Chang JH, et al. Expression of chemokines associated with apoptotic cell death in human promonocytic U937 cells and peripheral blood mononuclear cells. Eur J Immunol 1999; 29:3225–35.
22. Richter C, Schweizer M, Cosarizza A, et al. Control of apoptosis
by the cellular ATP level. FEBS Lett 1996; 378:107–10.
23. Eu JP, Liu L, Zeng M, et al. An apoptotic model for nitrosative stress Biochem 2000; 39:1040–7.