COLI DIARRHOEA

In a series of papers Hamilton et al. (1977, 1978a, 1978b) have investigated the effects of heat-stable and heat-labile E.coli entero-toxins, choleratoxin and theophylline on the small intestine of wean-ling pigs. The reason for the use of theophylline in these and other studies is, as noted, that it inhibits the phosphodiesterase and the-refore augments the concentration of cyclic nucleotides. In the pa-pers of Hamilton et al. ligated loops of jejunum were used in animals kept under anaesthesia throughout the experiment. PEG served as a marker of net volume movement. Most experiments were of 20 minutes duration. A marked secretion of water and NaCl was observed. Unidi-rectional fluxes of Cl and Na were carried out, but changing specific activities may limit the interpretation of the results. The toxins and theophylline showed additive effects, and glucose augmented ab-sorption from control loops and impaired the secretion. The mucosal concentration of cAMP was not changed significantly (Hamilton et al.

1978c); the heat-stable toxin affects however cGMP.

The passive permeability properties of duodenum and middle and distal jejunum were also measured after exposure to heat-stable coli toxin (Presnell et al. 1979). The osmotic permeability coefficient (by the authors denoted filtration coefficient) was estimated by ad-ding mannitol to a total osmolality of 600-700 mOsm to a solution con-taining 160 mM NaCl. Interpretation of the results is therefore li-mited by the likely presence of solute-linked water flow in the con-trol loops, but presumably absence in the toxin-treated loops. No change in apparent osmotic permeability was actually observed with mannitol as osmotic agent, but the osmotic permeability was slightly impaired by osmotic agents with reflection coefficients below unity (reflection coefficient for erythritol = 0.9, for urea = 0.7). The reflection coefficient was not affected by the toxin indicating that the toxin does not damage the intestinal wall as such, but af-fects the active NaCl transport and the accompanying water flow.

The effects of heat-stable E.coli toxin and theophylline were stu-died both in vivo and in vitro on pig jejunum and ileum (Argenzio et al. 1984). The effect of theophylline was as described in other spe-cies, a net secretion of Cl was induced. Heat-stable toxin induced

a rise in the mucosal concentration of cGMP and an increased secre-tion of HCO-, both in jejunum and ileum. In addisecre-tion a net absorpsecre-tion of NaCl was in both organs changed to secretion. The SCC as studied in vitro was, in agreement with these findings, augmented by heat sta-ble toxin and theophylline in an additive way.

In contrast to the findings in some other species the proximal colon of the pig is also influenced by the action of heat-stable E.

coli enterotoxin (and by theophylline and choleratoxin as well). Ar-genzio & Whipp (1981) found by loop experiments in vivo a large re-duction of water absorption, with net Na absorption reduced to near zero and increased secretion of HCCU. There was a clear dose-respon-se curve for heat-stable toxin. Similar effects were brought about by choleratoxin and theophylline but there was a difference. Coli-toxin selectively elevated mucosal cGMP . choleraColi-toxin cAMP, whereas theophylline augmented both. The lumen-negative transepithelial PD was augmented during the cause of action of the toxins. The mucosa was also mounted in vitro, and after theophylline PD and SCC were ob-served to rise in parallel. This shows, that the change of PD is due to movement of charge, presumably an ion secretion. In a subsequent study of the proximal spiral colon in vitro Argenzio & Whipp (1983b).

found an only moderate SCC equal to the small difference between lar-ge unidirectional fluxes of Na and Cl. An electroneutral coupled NaCl transport does therefore seem to be of importance. The effect of heat-stable toxin was apparently not that of changing the Cl flux in the serosa-mucosa direction as in other species (Field et al. 1980' but to reduce Na transport om the serosa-mucosa direction.

Of great interest in these studies is the fact that colitoxin in-hibits salt and water absorption in colon as such. The diarrhoea is therefore also of primarily colonie origin. As mentioned above HCO-, is secreted in great quantity due to the direct colonie action of the toxin and not only as a part of a compensatory mechanism. As corona-virus produces acid secretion there is a pathophysiological basis for the clinical observation that feces pH is high during infection with choleratoxin but low when transmissible gastroenteritis is the pathogenic agent (Bertschinger 1984, Bergeland & Henry 1982).

Recent advances in pathophysiological insight and in treatment of enterotoxigenic diarrhoea has been summarized by Powell (1984).

VIRAL DIARRHOEA

Transmissible gastroenteritis (TGE) is caused by a coronavirus infection. The intestinal malfunction leads to diarrhoea. Its patho-physiology has been studied in some detail in relation to salt and water transport in the small and large intestine. In vivo perfusion of the proximal jejunum, mid-jejunum, ileum and colon 40 hours after the infection in 3 week old piglets showed only impaired NaCl and water transport in the first segment (Butler et al. 1974). The proxi-mal jejunum was therefore analysed further in vitro by the Ussing-chamber technique (McClung et al. 1976). These authors observed Na and Cl secretion in all tissues. The secretion was changed to absorp-tion of Na by luminal glucose (3o m M ) . Chloride was less affected.

The control rates of net Na and Cl transport were not affected by TEG, but the glucose response was abolished. Addition of ouabain re-duced the SCC due to impaired Na transport, but theophylline increased SCC reflecting augmented Cl secretion. In a further study the time course of the infection was studied. It was progressing the first 40 hours, with complete recovery after 6 days (Kerzner et al. 1977).

This study, and a following study in the ileum (Shepherd et al. 1979), characterized the reaction of several enzymes of the enterocytes to the infection. The lactase, sucrase and Na,K-ATPase (in units/g pro-tein) were strongly reduced, whereas the concentration of thymidine kinase was quadrapled. This can readily be interpreted to indicate that the absorption defect is caused by the villus atrophy characte-ristic of TGE with a preponderance of undamaged crypt cells. These lack the Na-glucose cotransport system in their apical membrane as well as the sessile enzymes causing the final hydrolysis to amino acid and hexose. The latter study (Shepherd et al. 1979) demonstra-ted the same glucose stimulation of ileal SCC (presumably Na trans-port) as in jejunum. This demonstrates directly the physiological rationale for the effect of oral glucose also in this disease, albeit with a reduced response.

Since TGE may lead to a loss of glucose from the small intestine, due to the impaired absorption of carbohydrate, Argenzio et al. (1984) addressed the question of the nature of compensation in colon. This usually allows 3 week old piglets to survive the infection but leads regularly to death in younger animals (3 day old piglets were studied), Argenzio et al. slaughtered the animals 2 days after infection in the

younger, 4 days in the older group. PEG was used as a water marker.

As to be expected control animals had a lower PEG concentration in stomach and small intestine than given in the milk. With the PEG con-centration of stomach content as a baseline ileum of control pigs had a 5-fold higher concentration, colon 9-fold indicating 80 and near 90 per cent absorption of water, respectively. At 3 weeks of age these values were higher. In the infected piglets no water absorp-tion was apparent along the entire gut in the 3 day old group, but a pronounced colonie compensation (around 5-fold increase in marker concentration) was observed in the 3 week old pigs. The 3 day old pigs had a 6-fold higher carbohydrate concentration in colon during infection, in the 3 week groups it was reduced to 3/4, but with a slightly increased SCFA concentration; this was higher than in the 3 day old group. The main reason for the improved colonie compensa-tion in the older animals is therefore the development of microbial fermentation. This reduces the osmotic load imposed by undigested carbohydrate (due to the infection of the small intestine). The SCFA1

which are absorbed lead also to a high rate of Na absorption which augments the water absorption.

The summary, the TGE diarrhoea is osmotic in origin due to malab-sorption caused by the villus atrophy of the small intestine.

ROTAVIRUS DIARRHOEA

Two days after infection with rotavirus 6 day old piglets were prepared for in vivo infusion experiments of jejunum and ileum (Gra-ham et al. 1984). In 'addition stool analysis was made. The infectioi resulted in a major change of net absorption of water, Na and 3-0-me-thylglucose into secretion both in jejunum and ileum. In addition the levels of sucrase and lactase of the mucosa were strongly reduced

Over 3 days after infection fecal osmolality increased from 250 mOsm to 350 mOsm with the difference accounted for two thirds by lactose and one third by Na. In a previous study Telch et al. (1981) demon-strated impaired glucose absorption both in vivo and in vitro.

The pathogenecity of the rotavirus diarrhoea is just as for TGE:

impaired absorption and enzyme digestion in the small intestine leads to an overload of colon giving an osmotic diarrhoea made worse by a reduced absorption of Na, acetate and water in colon. The authors suggest that virus enteritis should be treated with oral solutions with less glucose than in the WHO formula so that no more

carbohy-drate is given than is actually absorbed. Some clinical studies point to greater advantage of a glucose-amino acid solution (Bywater & Woode 1980),and it is understandable that simple withholding of food may re-duce the scour.

SWINE DYSENTERIA (TREPDNEMA HYODYSENTERIE)

In this disease the morphological findings point to a primary af-fection of the colon with unaffected small intestine. These patholo-gical events have been amply confirmed by physiolopatholo-gical experiments.

Net solute and water absorption from the entire small intestine, stu-died by in vivo perfusion, was identical in control and infected pigs

(Argenzio 1980) whereas colonie water and electrolyte (Na, K, Cl, HCO,) absorption, studied by in vivo installation by the loop tech-nique (Argenzio et al. 1980),was abolished in infected animals. Uni-directional Na and Cl fluxes showed no change of serosa-mucosa move-ment, but severe reduction of fluxes in the mucosa-serosa direction.

In agreement with this the cellular concentration of cAMP was little affected (Schmall et al. 1983). The severe loss of electrolytes and water occurring in the disease is therefore caused exclusively by failing colonie absorption. It might be borne in mind also in this context that an amount of Na equal to nearly half the content of the extracellular volume is daily absorbed in the colon. With a primary colonie affection the efficacy of the oral glucose-electrolyte treat-ment is therefore in this disease due to small intestinal compensa-tion of a loss occurring further anally.

POSTWEANING DIARRHOEA

The large and complex problems of adjustment of the pig intestine to weaning are outside the scope of this survey but a few notes con-cerning the adaptation in general and the pathophysiology of diarrhoea in particular are relevant.

As will be pointed out in the section on adaptation any major va-riation in amount and type of food ingested will lead to morphologi-cal changes of the intestine, to different rates of flow and enzyma-tic composition of secretory juices, and to changed physiological transport parameters. A change from sows milk to other nutrients will therefore eo ipso, regardless of age of the piglet, lead to transient and permanent changes of the variables mentioned above.

To this comes the specific effects of the start of the weaning

pro-ded to distinguish clearly between the effect of the weaning process itself (which depends heavily on the time schedule of feeding change-over and the type and amount of food), and the time after birth when the weaning process is instituted. To this comes the problems of ef-fects of race, and differences in general maintenance of the sows and piglets; among these factors is environmental temperature (Pou-teaux et al. 1982, Dauncey 1983). All the parameters finally influen ce the chance of pathogenic agents to develope to a stage which cau-ses diarrhoea in the already diarrhoea-prone weaning piglet.

One major hypothesis concerning the ethiology of postweaning diarrhoea has been advanced by Miller et al. (1984a, b ) . These authors conclude that pigs have a transient immune hypersensitivity to dietary antigenes which leads to villus atrophy, crypt cell hyper-plasia and malabsorption. This causes increased susceptibility to E.coli enteritis and consequently diarrhoea (Miller et al. 1984c).

The malabsorption is likely to have a central position in the patho-genesis of the diarrhoea which may occur even in the absence of patho genie microorganisms. The general reduction of absorption in the small intestine, as demonstrated by Hamilton & Roe (1977), and the lack of enzymatic adaptation to the ingested foods, particularly carbohy-drates, may lead to an osmotic overload of the colon. The structural and enzymatic changes in the postweaning period were recently con-firmed by Hampson & Kidder (1984). As Miller's group these authors point to the antigenicity of the diet and substantiate this concept with the finding of smaller structural changes and less reduction of brush-border enzyme activity in a group weaned on to a less antigenic diet (Hampson et al. 1983). These effects were demonstrated very clearly in a study by Etheridge et al. (1984). They collected feces from three groups of piglets weaned at four weeks of age to two diets the first consisted of a corn-soybean meal starter diet, the second a steamed rolled oat groats-casein diet; a third group remained with the sow. Feces was collected daily and analysed for minerals (Na, K, Cl, Ca, P) and some organic substances (lactose, glucose and SCFA).

Total osmotic excretion was measured by dissolving the dried feces into a known quantity of water and measuring osmolality by freezing point depression. The three diets resulted in very different pattern (Table 1 ) . Osmolality was as noted not determined on fresh feces but from the authors' account of water content of the feces,approximate

Table 1. Osmotic composition of feces.

Food

Organic molecules Minerals

Unknown "osmolality"

Approximate osmolality

Corn-soybean 16 S 31 % 53 % 750 mOsm

Oats-casein 12 ?»

3 9 % 49 % 440 mOsm

Sow's milk 8 % 71 % 21 % 250 mOsm estimates of total osmolality can be given. These are included in the table.

The table shows that the postweaning diets present an osmotic stress to the colon as they lead to an increased output of organic molecules of low molecular weight and an osmolality which will force water movement in the serosa-mucosa direction. The large amount of unknown "osmolality" in feces is probably also due to organic mole-cules. The conclusion is that the decreased absorption of nutrients, particularly carbohydrates, in the small intestine leads to unnatural fermentation and loss of nutrients and water from the large intestine

As this pathogenicity of diarrhoea is similar to that of viral diarrhoea it is not surprising that restriction of feed intake has significantly reduced the incidence and severity of postweaning diarr hoea (Ball & Aherne 1982).

ADAPTATION OF INTESTINAL ABSORPTION

The word adaptation can be used in several contexts in relation to intestinal transport: 1. The change of various parameters as func-tions of time after a change in diet or other external or internal disturbances. 2. The difference between "before and after" when a new steady state has been achieved. 3. The causal relations between observed changes in transport functions and the releasing mechanisms.

Adaptation of intestinal absorption of nutrients and of salt and water is difficult to characterize even for one clinical or experimen-tal situation. The problems in describing adaptation to external and internal changes are manyfold larger. Dietary, hormonal, pathologi-cal, and environmental factors can all change one or several absorp-tion patterns on various segments of the gut.

The description of a transport pattern for a given intestinal seg-ment is difficult because a complete kinetic characterization requires

many experiments. Consider for example a carrier-mediated transport which, following Michaelis-Menten kinetics, changes in such a way

from state A to state B that the affinity decreases considerably (K is i n c r e a s e d ) , but the maximal transport rate (V ) is increased.

m 9 x

In this case transport measurements at low concentration of the sub-stance will show reduced rate of absorption from A to B whereas mea-surements at high concentration will show increased absorptionl

The description of adaptive changes is also difficult becauseit is significant which intestinal parameters the change are related to.

Absorption may be related to the intact organism (either kg body weight, surface area, basal metabolic rate, e t c . ) , or calculated per

2

cm along the intestine, per cm serosal or mucosal surface area, or per mg protein or DNA in the mucosa. These differences will often prevent meaningful comparison between experiments. Finally comes the difficulty in determining whether adaptation is best characterized by local rates of absorption at naturally occurring concentrations in the chyme, by maximal absorption capacity (as determined in vivo or in vitro) or by fractional absorption related to localized segments of the gut.

To these more fundamental problems comes technical difficulties and the inherent limitation in the various techniques available to the nutritionist or intestinal physiologist. There is always a dif-ference between in vivo and in vitro experiments (Smith 1 9 8 0 ) . This is at least partly due to the absence of normal blood flow in vitro (see Mailman 1982) and - for fast diffusing molecules - caused by different unstirred layers (see Section I in Skadhauge & Heintze 1983).

A large body of knowledge of intestinal adaptation is available to day but largely at the descriptive level. The time course of e-vents, the change in structure and transport function for transposi-tions of gut segments, changes in food intake, reaction to hormones, etc. are known in detail. The requirements for composition of chyme and pancreatico-biliary secretion to maintain normal structure and function are well known (Robinson et al. 1 9 8 1 ) . Knowledge about what causes these changes is, however, virtually non-existant. For the important adapative regulation of sugar and amino acid transport the-re is little knowledge about the the-regulation at a subcellular level.

The central question of whether increased transport depends on in-ducing more carriers in existing cells or proin-ducing new cells with a higher density of carriers is not known (Karasow & Diamond 1 9 8 3 ) .

The same ignorance applies to the effects of aldosterone on Na trans-port (Skadhauge 1983, Clauss et al. 1984).

In the case of the pig an important maturation takes place over the first 8 weeks of life. This involves gastric .digestion, acid se-cretion and proteolytic enzyme production (Cranwell 1984), as well as plasma gastrin concentration (Bunn et al. 1981), development of di-gestive enzymes (Aumaitre 1971, Hartman et al. 1961, Manners 1976) and the change in size of the organs (McCance 1974). The villus height and crypt depth and the amino acid-Na transport interaction are all affected by the energy intake (Dauncey et al. 1983). Early weaning may lead to morphological changes similar to an acute inflam-matory response (Kenworthy 1976).

Important information concerning correlation of structure and function in food utilization can be obtained from detailed physiolo-gical studies of the adaptative changes induced by selective breeding ans crossing of races. The classical papers of Hjalmar Clausen (1953,

Important information concerning correlation of structure and function in food utilization can be obtained from detailed physiolo-gical studies of the adaptative changes induced by selective breeding ans crossing of races. The classical papers of Hjalmar Clausen (1953,