Main

The WHO has promoted the use of ORS as the first line of defense against diarrheal illness regardless of etiology(1). Neither the currently recommended formula WHO-ORS nor its variants shorten disease duration or decrease stool losses, despite their effectiveness in compensating for dehydration and redressing electrolyte imbalance(1,2). Numerous studies have been devoted to investigating ORS additives that might further increase water and electrolyte absorption(3). These studies have included clinical trials investigating the efficacy of ORS containing amino acids such as glutamine, maltodextrins, and other glucose polymers. However, these studies have generally produced limited results(4–7). Therefore, research efforts continue for the development of an ORS capable of enhancing water and electrolyte absorption with the ultimate hope of shortening disease duration and reducing mortality, especially in infants and children(2,8).

Home remedies for diarrheal illness, such as rice extracts and gruels, have been universally used in an attempt to increase water electrolyte absorption in children with diarrhea(9,10). The addition of rice and starch-rich cereal grains to ORS has received attention as a method of both rehydration and provision of greater energy than the standard WHO-ORS because these products have been shown to be proabsorptive without increasing the osmotic load. An additional benefit is that infants and children can tolerate and digest starch products with little difficulty(10–13). Previous studies from our laboratory indicate that other additives can accelerate water and solute availability from ORS, such as the poorly absorbed soluble fibers gum arabic and carboxymethylcellulose(14–16). In this study, we investigated whether a digestible modified tapioca starch, TX, when added to a standard ORS, could enhance absorption during acute jejunal perfusions in two models of diarrheal disease. In one, an acute secretory status was induced via a jejunal perfusion of THEO, a methylxanthine capable of increasing cAMP and cGMP levels, promoting water and electrolyte secretion(17–19). The second approach was a model of chronic diarrhea induced in rats by a Mg Cit solution provided as their sole source of fluid for 7 d. In addition to jejunal perfusions assessing the effects of TX on improving water and electrolyte absorption, we also investigated whether a TX-based ORS could accelerate the recovery of animals with chronic osmotic diarrhea.

METHODS

Animals. All animals used in this study were male Sprague-Dawley rats (Harlan-Sprague-Dawley, Indianapolis, IN) weighing 60-80 g with free access to food (Purina Lab Chow; Ralston Purina Co., St. Louis, MO) and water. They were acclimated for a minimum of 48 h in polycarbonate cages at 72°F with a 12-h light-dark cycle before dietary treatment or intestinal perfusion as described below. Food was withheld from the rats before laparotomy unless otherwise indicated.

Solutions used and experimental protocol. The basal WHO-ORS contained 60 mmol/L Na chloride and 10 mmol/L trisodium citrate for a final Na concentration of 90 mmol/L, 20 mmol/L K chloride, and 111 mmol/L glucose. The osmolality of each solution was 300 ± 4 mOsm/kg (Wescor vapor pressure osmometer, model 5500, Logan, UT). The modified tapioca starch, TX, was added to or omitted from the ORS at 5.0 or 10.0 g/L with gentle heating. Addition of TX did not contribute free glucose to the solutions, and there was no difference in osmolality between the preparations containing TX and those not containing TX. In addition, heating the solutions to facilitate TX dissolution did not affect the osmolality or viscosity of either ORS containing TX. TX was a gift from the National Food Starch and Chemical Company (Bridgewater, NJ). Viscosity of the solution was determined with a Brookfield viscometer (model DV-I; Brookfield Engineering Labs, Stoughton, MA). The relative viscosity (η) of each solution containing 0, 0.5, and 10.0 g/L TX was 1.00, 1.25, and 1.60, respectively. Chemicals for the ORS were purchased from Sigma Chemical Co. (St. Louis, MO). Tritiated water (3H2O, 2 µCi/L = 74 kBq/L; DuPont-New England Nuclear, Boston, MA) was added to all experimental solutions to monitor lumen-to-serosa water flux.

Anesthesia was induced with ketamine/xylazine (87/13 g/L, at a dose of 0.15 mL/100 g, intramuscularly). After induction, a midline incision was made, the jejunum was cannulated with polyethylene tubing at the ligament of Treitz, and an exit port was inserted 20-30 cm distally. Twelve rats were used in each perfusion and three solutions were perfused in each experiment to decrease interexperimental variability, although only one solution was perfused in each rat.

Solutions were heated in a water bath to reach body temperature (37°C) as they entered the rat intestine and were perfused with a peristaltic pump (model 1203; Harvard Instruments, Boston, MA) at a rate of 10-12 mL/h. The actual flow rate of each line was determined by collecting 30-min effluents pre- and postperfusion. The experimental protocol consisted of a 1-h equilibration to stabilize ion gradients, followed by 6-15 min collections. The weight differences between fluid entering and leaving the intestinal segment in each collected fraction were used to calculate net H2O absorption. Jin was measured by the disappearance of 3H2O from the perfusate. Jeff was determined by the difference between Jin and net H2O absorption. In addition, values of Jin/Jeff were calculated for each rat. Electrolyte and glucose absorption rates, as well as the algorithms used in the calculation of absorption rates and unidirectional water fluxes, have been published(16). This study was approved by the Institutional Animal Care Committee.

Secretory diarrhea model. Rats were fasted overnight, and anesthetized, and the jejunum was perfused with the WHO-ORS containing 10 mmol/L THEO, to which TX was added at concentrations of 0, 5.0, or 10.0 g/L. Weights of the control and THEO-treated rats, perfused with or without TX, did not significantly differ [control: 107.0 ± 3.2 g (n = 8); THEO: 104.0 ± 4.3 g (n = 30)]. Controls were perfused with the WHO-ORS without THEO.

Chronic osmotic diarrhea model. Juvenile rats (60-80 g) had free access to solid food for 1 wk and were offered as their only source of fluid a cathartic solution containing magnesium citrate USP diluted with an equal volume of tap water to a final magnesium concentration of 4.8 g/L with the addition of 0.1 g/L phenolphthalein. Control animals were fed the same solid diet and tap water. After 1 wk, rats were anesthetized and the jejunum was perfused, as described above. To compare transport in animals with the same final weight, initial weights of rats receiving the cathartic solution were selected such that their initial weight before Mg Cit treatment was greater than that of the controls [Mg Cit: 72.6 ± 5.2 g (n = 22); controls: 61.0 ± 3.7 g (n = 11)]. This was done because in previous experiments, rats ingesting the cathartic solution failed to grow as well as controls(16). At the end of the treatment period, immediately before perfusion, rat weights were indistinguishable [Mg Cit: 114.0 ± 6.7 g (n = 22); controls = 108.0 ± 7.3 g (n = 11)].

Recovery experiments. Osmotic secretory diarrhea was induced in juvenile rats as reported above. After the 7-d treatment with the cathartic solution, animals were divided into four treatment groups offive to six animal each. Each animal was individually housed in acid-washed steel metabolic cages during the recovery trial. Rats in group 1 (no recovery) were not placed on a recovery trial and was continued on Mg Cit for an additional 24 h. Group 2 (water recovery) received water, group 3 received the WHO-ORS (ORS recovery), and group 4 received the same WHO-ORS supplemented with 0.5 g/L TX (TX recovery). Fluid intake, food consumption, animal weights, and stool weights were monitored after 24 h of rehydration therapy. A flow chart outlining the experimental protocol for both the recovery and perfusion experiments can be found in Figure 1.

Figure 1
figure 1

Experimental protocol of perfusion and metabolic studies. Animals were acclimatized for 48 h and then either fasted overnight and perfused with 90 mmol/L Na+-111 mmol/L glucose ORS in the presence or absence of 10 mM THEO containing 0, 5.0, or 10.0 g/L TX or placed on Mg Cit for 7 d to induce diarrhea. Those rats drinking Mg Cit were either perfused with the ORS containing 0, 5.0, or 10.0 g/L TX or placed on a recovery trial in which each group of rats received either Mg Cit (no recovery), water, ORS, or ORS + 5.0 g/L TX for 24 h, after which metabolic parameters were measured (see "Methods").

Full size image

Statistics. Data were analyzed with a one-way analysis of variance with a post hoc Tukey's test for critical differences(20). Significance was set at the 95% level (p < 0.05). All data are expressed as means ± SEM.

RESULTS

Acute Jejunal Secretory Model

General characteristics. The jejunal perfusion of 10 mM THEO, as reported here and previously published, results in secretion of Na into the lumen and reduced water and K absorption(16).

Water, electrolyte, and glucose absorption. Net water, electrolyte, and glucose absorption were impaired by perfusion with THEO-containing ORS. The fluid and solute absorption effects described above were reversed by the addition of TX (Fig. 2 and Table 1). TX (5.0 g/L) added to ORS enhanced net water absorption by increasing Jin compared with THEO-treated rats without TX and untreated controls (Table 1). TX (10 g/L) added to ORS maintained the proabsorptive effects seen at 5.0 g/L; however, no additional benefit was apparent. TX (5.0 and 10.0 g/L) normalized Na absorption and enhanced K and glucose absorption above THEO-perfused and control values (Fig. 2). TX (10 g/L), although maintaining electrolyte absorption seen at 5.0 g/L, did not provide any additional benefit. Rats treated with THEO without TX had a reduced water Jin/Jeff compared with untreated controls (Table 1). This reduction in Jin/Jeff was reversed by the addition of TX to ORS to values indistinguishable from controls.

Figure 2
figure 2

Intestinal absorption rates of sodium, potassium, and glucose in THEO-treated rats. Data are means ± SEM. Controls, n = 8 (□); THEO + 0 g/L TX, n = 13 (▪); THEO + 5.0 g/L TX, n = 12 (□); THEO + 10.0 g/L TX, n = 13 (□). Values that do not share a common letter are significantly different (p < 0.05). The control group was perfused with 90 mmol/L Na+-111 mmol/L glucose ORS. The other groups were perfused with the same ORS containing 10 mmol/L THEO with or without the modified tapioca starch, TX.

Full size image
Table 1 Net water absorption and water influx/efflux ratio in rats treated with THEO and TX
Full size table

Chronic Osmotic Diarrhea Model

General characteristics. Rats drinking the Mg Cit solution for 7 d presented with loose stools, failed to grow as well as controls, and had fluid-filled ceca at laparotomy. Quantitative physical and biochemical characteristics of this model have been published(16,19).

Water, electrolyte, and glucose absorption. Rats drinking the cathartic solution did not show a decrease in net jejunal water absorption, Jin, or Jeff (Fig. 3, and Table 2); however, the addition of TX at 5.0 or 10.0 g/L increased net water absorption, attributable to a greater Jin than the Mg Cit-treated rats and control rats. Na transport was unchanged by the cathartic solution treatment compared with controls but was enhanced by the addition of TX (5.0 or 10.0 g/L) to the ORS (Fig. 3). Net K absorption was sharply reduced in the Mg Cit rats compared with controls, but the addition of TX to ORS (5.0 and 10.0 g/L) restored the absorption of this ion to normal levels. Although the cathartic treatment did not alter glucose transport, TX at 5.0 g/L increased glucose absorption compared with rats with diarrhea not perfused with TX and controls. Raising the TX concentration from 5.0 to 10.0 g/L maintained, but did not improve, ORS efficacy.

Figure 3
figure 3

Intestinal absorption rates of sodium, potassium, and glucose in rats with Mg Cit-induced chronic osmotic diarrhea. Data are means ± SEM. Controls, n = 8 (□); Mg Cit + 0 g/L TX, n = 13 (squlf]); Mg Cit + 5.0 g/L TX, n = 12 (□); Mg Cit + 10.0 g/L TX, n = 11 (□). Values that do not share a common letter are significantly different (p < 0.05). The control group was perfused with 90 mmol/L Na+-111 mmol/L glucose ORS. The other groups were offered Mg Cit for 7 d and perfused with the same ORS containing either 0, 5.0, or 10.0 g/L TX.

Full size image
Table 2 Net water absorption and water influx/efflux ratio in rats treated with magnesium citrate and TX
Full size table

Recovery Trials

Animals with Mg Cit-induced diarrhea not placed on a recovery trial presented with virtually no weight gain over 24 h despite eating and drinking similar amounts compared with other groups (Table 3). The rats receiving water or ORS as recovery solutions gained significantly more weight than those animals not placed on a recovery solution. However, rats receiving the TX containing ORS (5.0 g/L) had a greater weight gain compared with all other groups. The TX recovery group ate significantly greater amounts of solid food compared with all other groups. However, fluid intake, which we believe is an indicator of overall improvement in physiologic and nutritional status of the animals on the TX recovery regimen, did not differ among treatments. Stool output of animals not recovering from diarrhea, as well as those in the water and ORS groups, was indistinguishable after 24 h. By contrast, rats offered the TX containing ORS had a lower stool output compared with all other groups.

Table 3 Metabolic data from rats recovering from diarrhea
Full size table

DISCUSSION

This study extends our earlier work on the proabsorptive properties of TX in WHO-ORS to two animal models of diarrhea(21). Earlier findings showed that 5.0 g/L TX added to WHO-ORS stimulated small intestinal water and Na+ absorption in normal rats, compared with ORS without the additive(21). In this report, we show that the addition of TX enhanced Na, K, and glucose absorption during jejunal perfusions in the acute model of secretory diarrhea induced by THEO, as well as in the model of chronic diarrheal disease produced by administration of the Mg Cit solution. The physiologic transport effects produced by the addition of TX are paralleled by morphologic evidence of expansion of the villus extracellular spaces, a structural feature associated with enhanced lumen-to-serosa flow of water and electrolytes (data not shown)(22). Similar results were also seen in the intact animals who received TX during recovery from osmotic diarrhea. These animals showed a decrease in stool output and an increase in weight gain during recovery compared with other groups. These are precisely the goals sought after to improve the efficacy of ORS.

The THEO model is used to emulate certain aspects of the secretion induced by the toxins of Vibrio cholerae and some strains of Escherichia coli. These toxins, like THEO, elevate mucosal epithelial cyclic nucleotides, leading to the stimulation of crypt chloride secretion, the former by increasing guanylyl cyclase activity and the latter by phosphodiesterase inhibition resulting in increased cAMP(17,18,23,24). The experimental model of chronic osmotic diarrhea used in this study, induced by the administration of a Mg Cit solution, intends to reproduce some aspects of chronic osmotic diarrhea in infants associated with repeated cow's milk feedings and recurrent gastrointestinal infections linked to lactase deficiency and lactose intolerance. Poorly absorbed lactose increases the osmotic load in the gut resulting in persistent diarrhea(13). Rats drinking the Mg Cit solution present with marked fluid accumulation in the cecum and decreased rate of growth, paralleling common characteristics seen in recurrent clinical diarrheal episodes in infants and children. The mechanisms of phenolphthalein-induced catharsis involves the inhibition of (Na/K)-ATPase in the basolateral membrane of the enterocyte(25), either directly or through the stimulation of the inducible or constitutive form of nitric oxide synthase, resulting in increases in cGMP and subsequent smooth muscle relaxation and intestinal secretion(26). Although neither the acute nor chronic forms of experimental diarrhea are ideal models because the mode of inducing fluid and electrolyte loss differ from that triggering diarrhea in humans, they permit systematic studies on formula modifications or the utilization of additives to ORS(16,19).

The basis for the addition of TX to ORS stems from the considerable interest in the use of rice or other cereal derivatives as components of ORS to shorten the illness and ameliorate electrolyte loss during diarrheal episodes(27–29). Although preliminary studies appeared promising, the lack of standardization of the additives and the complexity of the differing products, including starchy gruels and amylase-treated starch, have produced largely inconclusive results(4). We reasoned that a standardized product that shares some chemical characteristics of partially hydrolyzed starch should yield positive proabsorptive effects comparable with those seen in the field with homemade heterogeneous solutions, thus facilitating analysis and interpretation of results. TX can remain in suspension without increasing viscosity, whereas non-modified types of starch tend to precipitate upon standing, which may confound their effectiveness in the ORS or make their evaluation inaccurate.

ORS with a greater energy content than that provided by glucose, without an osmotic penalty, has an added potential benefit for infants with diarrhea. Malnourished infants with chronic diarrhea commonly present with deficiencies in α-amylase, maltase, and isomaltase, which impairs or abolishes the digestion of complex carbohydrates provided by solid foods(30). Only salivary glucoamylase remains unaffected by gastroenteritis. This enzyme can hydrolyze starch into free glucose(31); therefore, modified starch products added to currently used ORS offer a readily useable energy source for infants with diarrhea. Although a TX-enriched ORS could not substitute for the timely introduction of regular feedings, it could serve as a bridge during recovery from gastroenteritis.

Although the mechanism(s) by which TX stimulates increased water and electrolyte absorption from ORS remains undetermined, we can offer several hypotheses to explain its proabsorptive effects. These include 1) TX-induced release of the proabsorptive gut hormone PYY or 2) induction of apical GLUT1 in the small intestine by TX. Recent studies have shown that colonic perfusions of fibers and gums such as pectin or gum arabic were secretagogues of PYY, a peptide known to stimulate water and electrolyte absorption in the intestine(32,33). Similarly, a glucose polymer such as TX, which resembles gum arabic chemically may stimulate the release of PYY, thus stimulating intestinal absorption.

Alternatively, TX may enhance electrolyte absorption through the activation of the GLUT1 transporter. This could occur by the recruitment of GLUT1 transporters from prestored intracellular sites to the apical plasma membrane or induction of GLUT1 mRNA synthesis, because it has been shown that the GLUT1 transporter can be up-regulated by feedings of a carbohydrate-rich diet(34–37). This potential mechanism, however, may be more significant in the recovery trials presented in this study rather than in acute perfusion studies. This mechanism along with the release of gut trophic factors(38) might improve intestinal morphology and function, and thereby play a positive role in the shortening of disease duration(39,40). Further research is required to evaluate the specific mechanisms underlying the improvement in transport function produced by the addition of TX to ORS and other long-chain polysaccharides to ORS. These studies could provide the physiologic and biochemical basis for the formulation of more effective ORS.