CHAPTER 7

REGULATORY VOLUME DECREASE IN SMALL INTESTINAL CRYPTS.

Cell volume regulation following exposure to hypotonic media is effected by the efflux of KCl and, in some cell types organic osmolytes, accompanied by the loss of osmotically obliged water from the cell. Although a KCl cotransporter has been proposed to mediate KCl efflux in erythrocytes and some epithelia (Hoffmann & Simonsen, 1989), the most commonly reported mechanism for mediating regulatory volume decrease (RVD) in epithelial cells is the parallel activation of separate K⁺ and Cl^- conductance pathways (for review see Lewis & Donaldson, 1990). The activation of an epithelial Cl^- conductive pathway in response to hypotonicity was intimated as early as 1961 by MacRobbie and Ussing. The parallel activation of separate conductive fluxes of K⁺ and Cl^- ions was first proposed for Ehrlich Ascite tumour cells (Hoffmann, 1978; Hoffmann et al., 1984), although the parallel activation of K⁺ and Cl^- channels following exposure to hypotonic media was only later demonstrated electrophysiologically in intestinal 407 cells (Hazama & Okada, 1988).

Due to the nature of intestinal transport activity there is considerable interest in the mechanisms of enterocyte volume regulation following cell swelling evoked by Na⁺-nutrient coupled cotransport or exposure to hypotonic media. Measurements of cell volume in isolated small intestinal villus enterocytes indicate that cell swelling evoked both by Na⁺-alanine cotransport and hypo-osmotic media induces an RVD that is mediated by the activation of separate K⁺ and Cl^- conductances, although the mechanisms of ion channel regulation may differ between the hypotonically and nutrient-induced cell swelling (MacLeod et al., 1992a; Brown and Sepulveda 1985).

To determine whether volume regulatory mechanisms are activated in the small intestinal crypt following exposure to hypotonic media, changes in crypt volume have been estimated by image analysis. The results presented in this chapter indicate that RVD occurs in small intestinal crypts and is inhibited by increasing extracellular [K⁺]_o and by inhibitors of K⁺ and Cl^- channels, consistent with the activation of separate K⁺ and Cl^- conductance pathways following exposure to hypotonic media.

Crypts were isolated as described previously in section 3.2.1. for electrophysiological measurements and in section 5.2.1. for crypt volume measurements. Electrophysiological recordings and data acquisition and analysis were performed as described in sections 3.2.3. and 3.2.4. respectively. Measurements of crypt volume were performed as described previously in section 5.2.3..

7.2.1. Solutions.

The composition of the pipette and extracellular solutions used are given in table 7.1. Patch pipettes were filled with a KCl-rich "intracellular" solution containing 100 m g/ml nystatin prepared as described in 3.2.2.. Hypotonic medium was prepared as the isotonic medium, omitting mannitol. Osmolarities were measured as 298 ± 5 and 225 ± 8 mosmole/l for the isotonic and hypotonic solutions respectively using a freezing point depression osmometer (3MO Advanced Instruments, Massachussetts, U.S.A.). The tonicity of the standard isotonic medium was increased to 400 ± 5 mosmole/l when required by addition of 100 mM D-mannitol, or to 499 ± 8 mosmole/l by the addition of 200 mM D-mannitol (means ± SD).

Stock solutions of the Cl^- channel blockers 9-anthracene carboxylic acid (9-AC) and 4,4'-diisothiocyanostilbene disulphonic acid (DIDS) were prepared in DMSO at 100mM concentration and stored frozen. Quinine was freshly prepared as a 10mM stock in distilled water and barium chloride was stored as a 100mM stock in distilled water. 5mM Ba²⁺ Hanks was prepared as standard Hanks medium, omitting 7.5mM NaCl. Precautions against light-sensitive reactions were adopted when using Cl^- channel blockers and quinine. Final concentrations were achieved by appropriate dilutions with normal Hanks medium.

All solutions and the chambers containing crypts were pre-equilibrated at 25° C for 30 minutes before the start of the experiment which was conducted at 25° C. All errors quoted are standard errors of the mean of n observations.

Table 7.1. Composition of solutions.

pH adjusted to 7.2 with Tris.

7.3.1 Determination of the accuracy of the volume assay.

Measurements of changes in crypt volume were performed to estimate the accuracy of the image analysis technique. The tonicity of the bathing medium was varied by the addition or removal of D-mannitol, exposing the crypt to a range of extracellular osmolarities. Fig.7.1. shows the relative change in crypt volume immediately after changing the tonicity of the bathing medium to the osmolarity indicated in the abscissa. These values are compared to the predicted relative volume change calculated assuming perfect osmometric behaviour (curve in Fig.7.1). The volume changes seen in response to changes in tonicity approached those predicted for ideal behaviour. Only at 500 mosmole/l did a significant deviation from Boyle-Van't Hoff law take place, although in general the tendency was for changes in relative volume to be less than those predicted by Boyle-Van't Hoff's law (Fig.7.1).

7.3.2. Effect of hypotonicity on crypt volume.

Experiments were performed to determine whether crypt enterocytes possess homeostatic regulatory mechanisms to recover their volume following exposure to hypo-osmotic media. Crypts maintained in isotonic solution showed a tendency to shrink slightly during incubation. This is shown in Fig.7.2A where the estimated volume, relative to initial volume, is plotted as a function of time after simply renewing the isotonic solution contained in the perfusion chamber. The osmolarity of the bathing solution was reduced at the time marked zero in the abscissa, without altering the ionic composition of the bathing medium, by removing 70 mM mannitol. This led to a rapid 25-30% increase in relative volume within the first minute after solution change. After the initial swelling, crypts slowly recovered their original volume over a period of 20-25 min.

7.3.3. Effect of increasing the extracellular K⁺ concentration upon crypt RVD.

If crypt RVD is due to the passive parallel loss of ions from the cytoplasm down their respective electrochemical gradients, then it should be prevented by the elimination of one of these gradients or by the inhibition of one of the conductive pathways. The effect of increasing extracellular K⁺ concentration by the equimolar substitution of NaCl with KCl upon crypt RVD is shown in Fig.7.2., where volume recovery is shown to occur following hypotonic swelling in the presence of 2 mM but not 20 mM extracellular K⁺. Altering extracellular K⁺ did not markedly affect crypt volume when maintained in isotonic solution (Fig.7.3).

7.3.4. Effect of K⁺ and Cl^- channel blockers upon RVD.

Established channel blockers were tested to determine whether K⁺ or Cl^- channels mediate crypt RVD. Fig.7.3A shows that the K⁺ channel blockers quinine and Ba²⁺, which both evoke membrane depolarisations in the unstimulated crypt, when added at the time of the change from iso- to hypotonicity abolished crypt RVD, the crypts remaining swollen for the duration of the experiment. Neither inhibitor markedly affected the rate or the extent of the osmotically-induced increase in volume. The addition of blockers under hypotonic conditions did not produce any significant change in crypt volume (the effects of the addition of Ba²⁺ and quinine are included in the mean values for the isotonic controls).

Cl^- permeability pathways are also known to mediate RVD in a variety of cell types. In order to investigate the possible involvement of Cl^- channels in mediating crypt RVD the effect of the Cl^- channel blockers 9-anthracene carboxylic acid (9-ACA) and 4-4' diisothiocyanostilbene disulphonic acid (DIDS) upon crypt RVD were tested. Fig.7.3B shows the effect of hypo-osmotic challenge in the presence and absence of 100 m M 9-ACA. Maximal swelling was observed within 4 min and RVD was complete within 16-20 minutes under control conditions. When the switch from iso- to hypotonicity was done in the presence of the blocker a rapid increase in crypt volume was observed, but unlike the control situation crypt volume did not recover for the duration of the experiment. Crypts maintained in isotonic medium did not change their volume markedly and 9-ACA had no effect under these conditions. A separate series of experiments was conducted to test the effect of DIDS upon crypt RVD. Hypotonic challenge increased crypt volume by 19 ± 2 and 15.5 ± 1% in the absence and presence of 100 m M DIDS respectively; the corresponding values after 12 minutes were -0.6 ± 1 and 10.2 ± 2% (means ± standard errors from 4 experiments).

7.3.6. Effects of cell swelling upon crypt membrane potential and conductance.

If parallel Cl^- and K⁺ conductances are activated by cell swelling then presumably one of these permeability pathways may be rate-limiting for the loss of KCl from the cytoplasm and consequently for RVD. Crypts shrunken by the addition of 100mM D-mannitol appear to recover their original volume (data not shown). A return to `isotonic' conditions by the removal of mannitol from the Hanks medium would induce crypt swelling. The effect of this manoeuvre is shown in Fig.7.4. where the removal of extracellular mannitol evokes a rapid 10 mV depolarisation of E_m associated with an increase in membrane conductance. This depolarisation declined over 15 minutes with an associated reduction in membrane conductance.

These experiments were designed to ascertain whether isolated small intestinal crypts are capable of regulating their volume following exposure to hypotonic media and to make initial investigations into the possible mechanisms underlying any volume regulatory response. The two most commonly employed techniques for estimating cell volume are the electronic sizing of dissociated cells (MacLeod & Hamilton, 1991a) and the analysis of timed photographic images (Suzuki et al., 1991). The method employed here assumes that the crypt morphology approximates to that of a cylinder. Using this approximation the estimated changes in relative crypt volume upon exposure to anisotonic media are close to those predicted for a perfect osmometer (Fig.7.1.). However, with the addition of 200mM D-mannitol, the calculated relative change in crypt volume is significantly smaller than the value predicted. This may not be surprising, as some 30% of cell volume is occupied by solid matter, and only cell water content could be expected to vary in direct proportion to changes in medium osmolarity (MacKnight, 1988). The volume of the luminal space is not accounted for in these experiments, although the approximation to osmometric behaviour indicates that it probably constitutes only a small proportion of total crypt volume.

The tendency for a slight reduction in crypt volume when maintained in isotonic medium may reflect the continuing adjustment of intracellular ionic activities following isolation in a Na⁺-rich Ca²⁺-free medium. After the initial swelling in hypotonic medium the crypt enterocytes appear to recover their initial volume within 20-25 minutes, with 50% RVD completed within 5-8 min of exposure to hypotonic medium. It is interesting to note that the process of RVD appears to be considerably slower in the intact crypt epithelium than has been generally reported for experiments with single cell preparations, such as the cultured Intestinal 407 cell line (Okada & Hazama, 1989) and isolated jejunal villus enterocytes (MacLeod & Hamilton, 1991a). Whether these time-dependent differences are due to surface area to volume ratio effects or specific differences in cellular physiology is not clear.

RVD in cells exposed to hypotonic media has been shown in other cells to be effected by the loss of organic osmolytes and/or KCl down an electrochemical potential gradient upon the activation of normally quiescent regulatory pathways (Hoffmann & Simonsen, 1989). Increasing the extracellular K⁺ concentration depolarises the crypt membrane, as the resting membrane potential is dominated by a basolateral K⁺ conductance, thus reducing the electrochemical gradient for Cl^- efflux and the concentration gradient for K⁺ efflux. Increasing extracellular K⁺ from 2.5 to 20 mM completely abolished RVD in small intestinal crypts. If membrane potential values of -58mV and -31mV are assumed when extracellular [K⁺]_o is 2.5 and 20 mM respectively (extrapolated from Fig. 3.1) and intracellular activities of K⁺ and Cl^- are taken as 125 and 40 mM respectively (from Sullivan & Field, 1991) then the driving forces on each ion can be calculated following exposure to a hypotonic medium of 0.76 x isotonic osmolarity. Assuming that small intestinal crypts behave as ideal osmometers intracellular Cl^- and K⁺ intracellular activities will be reduced to 31 and 98 mM respectively. From the Nernst equation we can therefore calculate that increasing [K⁺]_o from 2.5 to 20 mM reduces the driving force for K⁺ efflux from 36 to 10 mV (Emf = E_m - E_K). A similar calculation for Cl^- gives driving forces for Cl^- efflux of 26 mV and -1 mV, as E_Cl is calculated to be approximately -32mV. Thus the most probable explanation for the blockade of RVD with high extracellular K⁺ is the elimination of the electrochemical gradient for Cl^- efflux. However the possibility remains that the RVD may be mediated by a KCl cotransport mechanism. The thermodynamic favourability of KCl cotransport for a hypotonically swollen crypt can be calculated from the relationship:

([K⁺]_o x [Cl^-]_o) / ([K⁺]_i x [Cl^-]_i)

From this relation we obtain values of 0.09 and 0.73 for [K⁺]_o values of 2.5 and 20 mM respectively. Thus net KCl efflux would still be effected by a putative KCl cotransport mechanism even at 20mM [K⁺]_o as the cotransporter is still below equilibrium (calculated [K⁺]_o to achieve equilibrium is 27 mM). Volume measurements from villus enterocytes have similarly demonstrated that hypotonically-induced RVD was blocked by increasing extracellular [K⁺]_o (MacLeod & Hamilton, 1991a).

Both Ba²⁺ and quinine depolarise the unstimulated small intestinal crypt, an effect consistent with the inhibition of a basolateral K⁺ conductance (Fig.3.3). Both inhibitors have been demonstrated to abolish RVD in isolated villus enterocytes and in the Intestine 407 cell line (MacLeod & Hamilton, 1991a; Okada & Hazama, 1989). In the experiments described here quinine and Ba²⁺, used at 100m M and 5mM concentrations respectively, abolish RVD in guinea-pig small intestinal crypts. Measurements of RVD by electronic cell sizing in dissociated preparations of guinea pig crypt enterocytes have suggested that RVD in crypt cells is partial and is not inhibited by either quinine or Ba²⁺ employed at identical concentrations (MacLeod et al., 1992b). MacLeod and co-workers have suggested that a volume activated K⁺ conductance present in guinea-pig villus enterocytes is absent from crypt enterocytes.

Evidence for the involvement of a volume-activated Cl^- conductance in mediating crypt RVD comes from the abolition of volume recovery by 100m M 9-ACA, an established Cl^- channel inhibitor. RVD has been reported to be inhibited by 9-ACA in preparations of dissociated guinea-pig villus and crypt enterocytes (MacLeod et al., 1991a; MacLeod & Hamilton, 1992b). DIDS, a stilbene derivative, partially inhibited RVD in small intestinal crypts. This type of blocker is primarily an inhibitor of Cl^-/HCO^-₃ exchange but has also been reported to block intestinal Cl^- channels (Bridges et al., 1989; Giraldez et al., 1989). On the basis of the current evidence from the effects of channel inhibitors and ion substitutions it would appear that crypt RVD occurs as a result of the efflux of K⁺ and Cl^- ions through separate conductive membrane pathways.

Central to our understanding of the mechanism of RVD is whether cell swelling activates previously dormant conductances. Electrophysiological recordings have revealed the presence of a predominant Ba²⁺- and quinine-sensitive basolateral K⁺ conductance in unstimulated guinea-pig small intestinal crypts through both ion substitutions and experiments with ion channel inhibitors (Figs.3.1. & 3.3). It would then be possible to envisage that a Cl^- permeability is rate-limiting in RVD, with K⁺ efflux taking place through a tonically active K⁺ conductance. This hypothesis is given some support by the depolarisation evoked by the removal of mannitol from the bathing medium (Fig.7.4.). However, studies of RVD in dissociated preparations of crypt enterocytes have also indicated that a volume-activated Cl^- conductance is present, and that RVD is accelerated by the addition of the cation-selective ionophore gramicidin, suggesting that the cationic permeability limits both the rate and extent of crypt RVD (MacLeod et al., 1992b).

The addition of 100m M 9AC, which completely abolishes crypt RVD (Fig.7.2B), only partially inhibits crypt SVD. This observation implies that different Cl^- conductances mediate Cl^- exit during fluid secretion and RVD. This is consistent with the evidence obtained from direct measurements of whole-cell current in other epithelial cells (Wagner et al., 1991; Worrell et al., 1989).

Previous studies of volume regulation in mammalian intestinal cells have not elucidated which membrane domain, apical or basolateral, harbours the K⁺ and Cl^- conductive pathways that mediate RVD. In Necturus small intestine electrophysiological recordings suggest that a Ba²⁺-sensitive, swelling-activated K⁺ conductance is located in the basolateral membrane (Lau et al., 1984) whilst a swelling activated Cl^- conductance is located in the apical membrane (Giraldez et al., 1988b). Although it is not possible to ascribe a site for the conductive pathways postulated here, it is tempting to speculate that they are basolateral. Such a location would be consistent with the virtually instantaneous effect of the blockers, as the lumen of the crypts should not be readily accessible to the bathing solution. Measurements of volume and single channel activity in isolated rat colonic crypts following exposure to hypotonic media have shown that RVD is mediated by the parallel activation of basolateral K⁺ and Cl^- channels (Diener et al., 1992).

The results presented here are consistent with the activation of separate K⁺ and Cl^- channels in small intestinal crypt RVD, as RVD is prevented by specific blockers of these permeability pathways and by increasing the extracellular K⁺ concentration.

Return to top
Return to table of contents
Return to science homepage
References
Next chapter