Water transport revisited

Abstract

Nearly forty years ago Curran & MacIntosh (1962) presented experimental evidence for a model of water transport that would satisfy a question posed by such famous physiologists as Heidenhain and Reed before the turn of the century: how can the intestine accomplish transport of water from one isotonic compartment (the bowel lumen) to another (the blood)? Curran, who was a superb experimentalist as well as a theoretician, had previously shown that water transport bore a linear relationship to solute (Na+) transport. He and MacIntosh then demonstrated that a three-compartment model (see Fig. 1) would allow movement of fluid from compartment I to III as long as compartment II contained a solution hypertonic to I and III, and as long as the permeabilities (reflection coefficients) of the membranes A and B, which separated the compartments, were finite with B greater than A. They used cellophane for membrane A and sintered glass for membrane B in their model. The biological counterparts to these membranes were thought to be the tight junctions (TJ) of the epithelial cells for membrane A, the basement membrane (BM) for membrane B and the intercellular space (ICS) for compartment II. Active Na+ transport along the basolateral membrane of the epithelial cell into the ICS was proposed as the active solute's transport step driving the passive flow of water from I to III. Diamond's ‘standing gradient’ hypothesis of water movement was a later variation of this model (Diamond, 1978). Thus, the mystery of the special driving force, or ‘treibkraft’ moving water across the intestinal wall was solved: it was ‘compartmentalized’ osmotic pressure. Curran and MacIntosh's model system for water transport (right) is shown with proposed biological counterparts in the colonic mucosa (left). SMS, subepithelial myofibroblastic syncytium with reticular fibrils. These models of water transport were adequate to explain isotonic transport, i.e. an absorbant osmolality that was essentially isotonic with the luminal fluid and with the plasma. It was subsequently shown that the distal colon could transport fluid hypertonically, i.e. an absorbate with osmolalities of 500–1000 mosmol (kg H2O)−1. Furthermore, such hypertonic transport was necessary to dehydrate the faecal mass in the distal colon. Naftalin and colleagues have proposed that the development of hypertonic water transport allows the distal colonic crypts to act as suction devices capable of dehydrating faeces (Naftalin et al. 1999). Given known hydraulic permeabilities of the epithelial cell membranes and rates of Na+ transport, it has been difficult to imagine how the distal colon could create an osmotic gradient in the ICS sufficient to accomplish the formation of hard faeces. The two papers by Naftalin and colleagues (Naftalin et al. 1999; Naftalin & Pedley, 1999) in this issue of TheThe Journal of Physiology shed light on this question by proposing a role for the myofibroblast-reticular sheath (which they call the fibronexus). This ‘sheath’ forms a second fenestrated ‘membrane’ just under the fenestrated BM in the intestine and probably most epithelial tissues (Toyoda et al. 1997). Heretofore, this myofibroblastic syncytium had been thought only to influence intestinal secretion through prostaglandin (cyclic AMP)-mediated sensitization of intestinal epithelial cells to Ca2+-mediated secretogogues (Berschneider & Powell, 1992). Using confocal microscopy to identify this ‘sheath’ and a fluorescent probe for Na+, these investigators show that it presents a diffusion barrier to Na+ that correlates well with the gut segments that are able to transport hypertonically. These experiments suggest that the reflection coefficient of the BM is such that this barrier is adequate for isotonic transport, but that the myofibroblast-reticular sheath serves as an additional diffusion barrier in series with the BM in order to accomplish hypertonic transport. Alternatively, it may be that it is the myofibroblast-reticular sheath, and not the BM, which represents the basolateral diffusion barrier for either isotonic or hypertonic transport. The investigators have also shown a correlation between the barrier function of this sheath and a high renin-angiotensin II-aldosterone state brought about by dietary Na+ depletion. It has been previously known that such a state greatly increases distal colonic Na+ transport by increasing the resistance of the tight junctions and by increasing expression of amiloride-sensitive Na+ channels and Na+,K+-ATPase on the apical and basolateral membranes of the distal colonocyte. Naftalin and colleagues propose that this sodium depletion state also leads to greater osmotic transport, through angiotensin II- or aldosterone-induced activation of the myofibroblasts with increased synthesis and secretion of reticulin fibrils, resulting in changes in the reflection coefficient of this myofibroblast-reticular membrane. These investigators give increased verification and clearer biological counterparts to a model of transport proposed nearly half a century ago. It remains to be determined if this myofibroblast-reticular sheath has such a transport function in all electrolyte and water transporting epithelia. It appears that the syncytium of myofibroblasts exists under the epithelial BM in all transporting epithelia where it has contractile functions and growth, differentiation and wound repair functions. This syncytium will be the subject of considerable investigation over the next few years.