CHAPTER 10

DISCUSSION AND SUMMARY

10.1.1. The evolution of small intestinal transport studies.

Early studies of the mechanisms of small intestinal secretion have principally involved measurements of short-circuit current across the intestinal epithelium in vitro, using the technique originally devised by Ussing and Zerahn in 1951, and of ion fluxes in vivo and in vitro (e.g. Schwartz et al., 1974). Whilst the capacity of the small intestine to secrete fluid and electrolytes in response to enterotoxins and secretagogues has been long established, the heterogeneity of the various cell populations that are present within the epithelium and the contamination of such preparations by intrinsic nerve fibres, fibroblasts and cells involved in mediating inflammatory responses such as neutrophils (Madara et al., 1991), has impeded our understanding of the regulation and compartmentalisation of fluid and electrolyte secretion at the cellular level in the intestinal epithelium.

Although a number of polypeptides, amines and eicosanoids have been shown to modulate small intestinal transport in vivo and in vitro, evidence for a direct action by these agents in modulating the secretion of fluid and electrolytes at the level of the epithelium is often limited. Studies of short-circuit current in polarised monolayers of colonic carcinoma cells (Dharmsathaphorn et al., 1985) and microelectrode impalements of the apical membrane of the small intestinal epithelium (Giraldez et al., 1988a; Stewart & Turnberg, 1989) have led to the identification of some of the membrane conductance pathways that are activated by secretagogues. However, the development of improved methods to isolate preparations of viable dissociated enterocytes was essential to further our understanding of small intestinal transport phenomena. The advent of the patch-clamp technique (Neher & Sakmann, 1976) permitted the detailed characterisation of the K⁺ and Cl^- channels present in enterocytes isolated from the small intestine of the amphibian, Necturus (Giraldez et al., 1989; Sheppard et al., 1988a), and provided a model for small intestinal secretion.

The subsequent characterisation of the K⁺ and Cl^- conductance pathways present in dissociated guinea-pig villus enterocytes (Sepulveda et al., 1991) extended the study of the cellular mechanisms of small intestinal secretion to the mammalian level. However secretagogues and agents that are known to increase levels of intracellular cAMP or Ca²⁺ did not appear to evoke increases in membrane conductance in freshly dissociated preparations of guinea-pig villus enterocytes (Sepulveda et al., 1991). These observations necessitated the study of the effect of secretagogues upon the membrane conductances present in identifiable populations of crypt enterocytes.

The results presented in this thesis indicate the presence of functional receptors for both VIP and acetylcholine in the basolateral membrane of the guinea-pig small intestinal crypt, neurotransmitters that are known to be released from the nerve endings of the sub-mucosal plexus in proximity to the epithelium (Costa et al., 1980; Brown & Miller, 1992).

The data provides direct evidence for the presence of voltage and second messenger modulated K⁺ and non-selective cation channels in the crypt basolateral membrane and affords indirect evidence for the presence of volume and agonist-activated Cl^- conductance pathways. However, further work is required to determine the location of functional receptors for autocrine mediators, such as PGE₂ (Stewart & Turnberg, 1989), for putative paracrine mediators released from enteroendocrine cells present in the epithelium and for circulating hormones such as ANP (Bianchi et al., 1989) in the small intestinal epithelium.

10.1.2. Regulation of cation-selective channels in the crypt basolateral membrane.

Muscarinic stimulation evokes an increase in K⁺ conductance in isolated small intestinal crypts that is apparently mediated by small conductance K⁺ channels located in the crypt basolateral membrane. However, only indirect evidence exists to support the contention that VIP and its intracellular mediators increase a K⁺ conductance in the small intestinal crypt epithelium. The VIP-induced depolarisation might in itself induce the activation of small conductance voltage-gated K⁺ channels, rather than via a direct action by cAMP and its intracellular mediators, providing a mechanism whereby increases in Cl^- conductance may be paralleled by an increase in basolateral K⁺ conductance. Changes in membrane potential and intracellular free Ca²⁺ similarly regulate non-selective cation channel activity, constituting a potential feedback mechanism to limit the extent of the carbachol-induced hyperpolarisation.

The dependence of E_m upon the concentration of extracellular K⁺ is decreased at low extracellular K⁺ concentrations (Fig.3.2.) This may be due to the direct modulation of basolateral K⁺ and cationic conductance pathways by changes in extracellular cation concentration and/or membrane potential. The conductance of the inward-rectifier and the selectivity of the sub-picosiemen cation-selective conductance of the crypt basolateral membrane both appear to be modulated with changes in the intracellular activities of Na⁺ and K⁺ ions, indicating that intracellular ionic activities may play an important role in the coupling of ion channel activity to transepithelial ion fluxes. The secretagogue-induced activation of basolateral K⁺ channels has been attributed to be the cause of the increased K⁺ content in the venous outflow from the cat submandibular gland following parasympathetic stimulation (Petersen & Poulsen, 1968). Local increases in extracellular K⁺ in proximity to the basolateral membrane have been proposed to activate inwardly-rectifying K⁺ channel activity in other epithelial cells (Stelling & Jacob, 1992; Kotera et al., 1991) and this may provide a mechanism whereby an increased basolateral K⁺ efflux results in the stimulation of K⁺ reuptake across the basolateral membrane of transporting epithelia. The present results suggest that a decrease in intracellular K⁺ activity and/or an increase in intracellular Na⁺ activity decreases the Na⁺ conductance or increases the K⁺ conductance of the crypt basolateral membrane. This may provide mechanisms whereby the membrane potential is repolarised following secretagogue action, and an intracellular K⁺ activity favourable to continued electrolyte secretion is maintained in conjunction with the operation of the Na⁺-K⁺ ATPase. A putative model showing the possible feedback pathways by which changes in membrane potential may govern the regulation of the constituent cation-selective channels present in the crypt basolateral membrane is shown in Fig.10.1.

Model of the regulation of cation-selective ion channel activity in the crypt basolateral membrane. Proposed model of the excitatory and inhibitory influences of changes in membrane potential and second messenger levels upon the activity of the various cation-selective single channel types present in the crypt basolateral membrane. The scheme is based upon the current-clamp and single-channel data presented in this thesis. Inhibitory feedback pathways are represented by dashed lines, positive feedback pathways are represented by solid lines and the effects of increasing channel conductance upon membrane potential is shown by the dotted arrows. The abbreviations used are RGK (resting K⁺ conductance), IR (inward rectifier), CAN (calcium-activated cation non-selective channel), SK (small conductance carbachol-activated K⁺ channel) and VIP (vasoactive intestinal polypeptide).

The decrease in cell volume that accompanies secretagogue action may serve not only to maintain intracellular ionic activities favourable to secretion, which in consequence may control the activity of ion channels and the thermodynamic gradients for carrier-mediated ion transport, but may also directly modulate ion channel activity and/or the levels of intracellular second messengers, events possibly transduced by changes in cytoskeletal conformation (see Watson, 1991 for review). Thus changes in intracellular second messenger levels, intracellular and extracellular ionic activities, membrane potential and cell volume evoked by hormone-receptor interaction may all serve to co-ordinate transcellular ion fluxes, effecting the parallel regulation of apical Cl^- and basolateral K⁺ conductances during stimulus-secretion coupling.

Given the variety of regulatory influences upon the cationic permeability of the basolateral membrane in exocrine acinar cells, it is perhaps not surprising to find such a diversity of cation and K⁺-selective channels in the crypt basolateral membrane. Ion channels that are inwardly-rectifying, cation non-selective, regulated by aniso-osmotic media and by changes in intracellular Ca⁺ and cAMP are ubiquitous in transporting epithelia (for review see Dawson, 1987). Whether the cationic permeability of the crypt basolateral membrane is also modulated by the levels of other second messengers, such as cGMP, which inactivates a voltage-gated K⁺ channel in winter flounder enterocytes (O'Grady et al., 1991), or by hypotonic media, which activates basolateral K⁺ channels in the colonic crypt epithelium (Diener et al., 1992) remains to be established.

10.1.3. Regulation of Ca²⁺ mobilisation in the crypt enterocyte.

The presence of a Ca²⁺-permeability in the crypt basolateral membrane can be inferred from the biphasic dependence of the carbachol-induced hyperpolarisation upon extracellular Ca²⁺ and the hyperpolarisation evoked by the Ca²⁺ ionophore A23187. Muscarinic receptors of the M1, M3 and M5 subtypes are known to be coupled to phosphoinositidase C (Wess et al., 1990), the activation of which results in an increased turnover of phosphatidyl-inositol bisphosphate and an increased production of inositol-(1,4,5)-trisphosphate (IP₃) and diacylglycerol, which in turn leads to the mobilisation of PKC and intracellular Ca²⁺ respectively (Berridge & Irvine, 1989). There is now clear evidence that channel-mediated Ca²⁺ entry is induced by the depletion of intracellular Ca²⁺ stores (Hoth & Penner, 1992) and by the action of inositol-(1,3,4,5)-tetrakisphosphate (IP₄, Luckhoff & Clapham, 1992), an inositol polyphosphate metabolite produced by the action of I-(1,4,5) P₃-3-kinase upon IP₃ (see Berridge 1993 for review). However, the spatial and temporal nature of the intracellular Ca²⁺ transients induced by muscarinic stimulation and the presence of Ca²⁺-permeable membrane channels remain to be established in the small intestinal crypt.

10.1.4. Membrane Cl^- conductance pathways of the intestinal epithelium and their central role in the pathophysiology of diarrhoea and cystic fibrosis.

Electrogenic Cl^- movements have been postulated to be central to small intestinal secretion. The results presented in this thesis provide indirect evidence for the presence of one or more Cl^- conductance pathways that are activated by VIP, forskolin, carbachol, phorbol esters and hypotonicity. Fundamental to our understanding of the physiology of small intestinal secretion is the location of the Cl^- conductance pathway(s) to either the apical or basolateral membrane domain and the determination of whether the second messenger pathways regulating changes in membrane Cl^- conductance converge on the same or separate Cl^- channel pathways. The volume-activated Cl^- conductance of Necturus enterocytes has been shown to be located in the apical membrane domain (Giraldez et al., 1988b), whilst in isolated colonic crypts regulatory volume decrease has been shown to be mediated by the activation of basolateral Cl^- channels by a mechanism that is dependent upon extracellular Ca²⁺ and prevented by inhibitors of the lipoxygenase pathway (Diener et al., 1992). The location and mechanism of regulation of the volume-activated Cl^- conductance present in the small intestinal crypt epithelium remains to be established, and may be of potential importance in the development of therapeutic approaches for the treatment of CF.

The intestinal epithelium is one of the tissues most severely affected in CF, a condition characterised by a deficit in fluid and electrolyte secretion, which was first demonstrated to be a consequence of impaired Cl^- transport by Quinton (1983) using sweat ducts isolated from normal and CF patients. Welsh and Liedtke (1986) and Frizzell and co-workers (1986) subsequently used the patch-clamp technique to show that the regulation of an apical Cl^- conductance at a site distal to the accumulation of cAMP was defective in CF, and implicated an intermediate conductance outwardly-rectifying Cl^- channel as central to the deficit in Cl^- transport. The failure of both cAMP-dependent phosphorylation (Li et al., 1988; Schoumacher et al., 1987) and PKC-mediated phosphorylation (Li et al., 1989) to activate a Cl^- conductance in cells homozygous for the CF mutation was subsequently reported.

However Gray and colleagues (1988) observed that secretin, acting via an increase in intracellular levels of cAMP, activated a small conductance Cl^- channels in the apical membrane of pancreatic duct cells. Later studies by Giraldez and coworkers (1989) showed that small conductance Cl^- channels activated by PKA were also present in membrane patches obtained from Necturus enterocytes. The over expression of CFTR cDNA in invertebrate cells (Kartner at al., 1991), the purification and reconstitution of the CFTR mRNA (Bear et al., 1992) and changes in anion selectivity of the cAMP-activated Cl^- conductance induced by site-directed mutagenesis of the CFTR gene (Anderson et al., 1991) have since provided conclusive evidence that the CFTR is itself a small conductance, phosphorylation-activated anion-selective channel.

The effects of VIP and forskolin upon crypt membrane potential and conductance are consistent with the presence of a cAMP-activated Cl^- conductance in the small intestinal crypt epithelium. The sensitivity of the volume and membrane conductance changes evoked by VIP to NPPB and the synergism between VIP and carbachol in modulating crypt membrane conductance are consistent with the activation of Cl^- channels associated with CFTR expression (Kartner, et al., 1991; Tabcharani et al., 1991). High levels of expression of the CFTR gene product have been detected in the crypt region of the rat small intestinal epithelium (Trezise & Buchwald, 1992), providing further evidence for the presence of a cAMP-activated Cl^- conductance in the crypt epithelium. Such a cAMP-activated Cl^- conductance may also be central to the mechanism by which enterotoxins, such as cholera toxin, evoke the hypersecretion of fluid and electrolytes across the small intestinal epithelium (Field & Semrad, 1993).

The heat-stable enterotoxin produced by Escherichia coli (STa) is another common cause of diarrhoea, stimulating an increase in intracellular cGMP levels (Field et al., 1978). In excised inside-out patches obtained from the T84 cell line cGMP-dependent protein kinase (PKG) activates a low conductance Cl^- channel, similar in properties to that activated by PKA (Lin et al., 1992). This evidence raises the possibility that the excessive intestinal Cl^- secretion induced by STa is due to the PKG-mediated activation of the Cl^- channel that is defective in CF. Increases in short-circuit current are evoked by the addition of phorbol esters, dibutyryl cGMP, dibutyryl cAMP and IBMX, a phosphodiesterase inhibitor, across normal, but not across human CF jejunum, consistent with the view that the kinases responsive to each of these intracellular `messengers' are unable to activate the Cl^- conductance known to be defective in CF (O'Loughlin et al., 1991). In the CF intestine both cAMP- and Ca²⁺-stimulated Cl^- secretion are defective (Taylor et al., 1987; Berschneider et al., 1988), whereas Ca²⁺-stimulated Cl^- secretion appears to be unaffected in CF airways (Wagner et al., 1991; Widdicombe, 1986). Thus although Ca²⁺ ionophore appears to evoke an increase in short-circuit current across normal human jejunum, the effect upon Cl^- conductance may be mediated by the action of PKC rather than by Ca²⁺ and its intracellular mediators, since the ionophore-induced increase in short-circuit current does not occur across the human CF jejunum (O'Loughlin et al., 1991).

Carbachol may activate a Cl^- conductance in isolated small intestinal crypts, as suggested by the sensitivity of the secretagogue-induced volume decrease to Cl^- channel inhibitors, although any effect upon a Cl^- conductance is masked in electrophysiological recordings by a predominant action upon a basolateral K⁺ conductance. The present body of evidence supports the contention that PKG, PKA and PKC all activate the intestinal CFTR, and unless CF mutations also result in the defective regulation of other types of Cl^- channel (Gabriel et al., 1993), there is no direct evidence for the presence of an alternative Cl^- conductance pathway activated by secretagogues in the small intestinal epithelium.

In the CF airway epithelium the Cl^- secretory deficit has been successfully circumvented by the lumenal application of nucleotide triphosphates such as ATP or UTP, (Stutts et al., 1992; Knowles et al., 1991). Thus an understanding of the regulation of alternative Cl^- conductive pathways that may be present in the apical membrane of the intestinal crypt epithelium, as for example a volume-activated Cl^- conductance, may lead to the development of therapeutic strategies in the treatment of the secretory deficit in the CF intestine. The excessive secretion of fluid and electrolytes elicited by enterotoxins is a major cause of morbidity in the developing world. The detailed characterisation of the pharmacology of the Cl^- conductances present in the apical membrane of the small intestinal epithelium may provide a basis for the development of suitable inhibitors to use in the treatment of diarrhoeal disease.

10.1.5. Proliferation, differentiation and secretion in the small intestinal crypt.

The small intestinal epithelium of Necturus consists of an apparently homogeneous population of enterocytes and is anatomically less complex in structure than the crypt and villus architecture of the mammalian small intestine. Thus the Necturus small intestine has been used to provide a generalised model of the principal mechanisms of small intestinal secretion. Electrophysiological studies have revealed a heterogeneity in the distribution of spontaneous resting membrane potentials and responses to muscarinic stimulation in different regions along the small intestinal crypt axis. This apparent heterogeneity indicates that there may well be a considerable variation in the functional expression of ion channels and transporters at different stages of enterocyte development within the different regions of the crypt. Therefore even enriched populations of dissociated crypt enterocytes may show a considerable variation in their transport properties and thus care must be taken in interpreting quantitative data obtained from such populations. The preparation used throughout this thesis provides a consistent means of establishing the position of enterocytes along the crypt axis and thus the potential for variation in the response to agonists can be minimised. Whilst there remains considerable controversy as to the possible role of villus enterocytes as a site of agonist-induced fluid and electrolyte secretion (Stewart & Turnberg, 1989; Sepulveda et al., 1991), the results presented in this thesis are consistent with role of the small intestinal crypt epithelium in mediating the agonist-induced secretion of fluid and electrolytes.

10.1.6. A possible model for small intestinal secretion.

The model originally proposed by Silva (1977) to account for fluid and electrolyte secretion by the shark rectal gland has been used as a basis to explain the mechanism of secretion in other epithelia, including the exocrine pancreas (Petersen, 1986). In this model, represented in Fig.10.2., K⁺ and Cl^- ions are accumulated intracellularly above electrochemical equilibrium by the parallel operation of a Na(K)Cl cotransporter and a Na⁺-K⁺ ATPase; the cotransporter, the Na⁺-K⁺ ATPase and the basolateral K⁺ conductance functioning together as an electrogenic Cl^- pump. The agonist-induced activation of apically sited Cl^- channels results in the electrogenic exit of Cl^- ions from the cytosol into the lumen, which is accompanied by a parallel increase in basolateral K⁺ permeability and the movement of cations from the interstitium into the lumen via the paracellular `shunt' pathway. However, this model does not adequately account for common features of many secretory epithelia and for some of the observations from the small intestinal crypt.

Regulation of secretion in the small intestinal crypt enterocyte. Model showing the component signalling pathways, membrane conductances and transporters necessary to effect the agonist-evoked stimulation of fluid and electrolyte secretion. The diagram represents the polarised monolayer of crypt enterocytes with tight junctions represented by filled rectangles. The location and regulatory pathways of the cationic, Cl^- and K⁺ membrane conductances apparent from the current experimental data are illustrated. Ionic fluxes and second messenger pathways are represented by solid arrows, whilst the modulating influences upon membrane conductance pathways are represented by the broken arrows. See text for discussion of the model.

Many investigators have reported the presence of cation non-selective channels in exocrine epithelia that are activated by elevations in cytosolic Ca²⁺ (Marty, 1984) or cAMP (Siemer & Gogelein, 1992). Such cation non-selective conductances would be expected to allow the entry of Na⁺ ions into the cytosol following agonist stimulation. The small intestinal crypt possesses Ca²⁺-activated cation non-selective channels in the basolateral membrane, but the potential contribution of a Na⁺ influx to the forskolin and VIP-induced depolarisation remains to be investigated. A depolarising influx of Na⁺ into a secretory epithelial cell would be expected to impair the efficiency of Na⁺-coupled Cl^- cotransport and to reduce the electrochemical gradient for Cl^- exit. Marty (1987) has proposed an alternative model for exocrine secretion wherein the movement of K⁺ ions across the apical membrane is accompanied by a parallel efflux of Cl^- ions. In this model the activation of a cation non-selective conductance pathway provides a continuous influx of Na⁺ ions, which in turn stimulates an increase in Na⁺-K⁺ ATPase activity and therefore results in an increased uptake of K⁺ ions into the cells to maintain an intracellular K⁺ activity favourable to sustained secretion. Many secretory epithelia possess a significant apical K⁺ conductance, although this has yet to be demonstrated for the small intestinal crypt. Such an apical K⁺ conductance has been proposed to enhance fluid and electrolyte secretion across `leaky' epithelia (Cook & Young, 1989).

The model proposed by Silva, Petersen and others does not predict that cell volume changes will accompany the onset of fluid and electrolyte secretion, as the agonist-evoked efflux of K⁺ and Cl^- ions is believed to be accompanied by a parallel increase in cotransport-mediated ion uptake. Following agonist addition small intestinal crypts undergo a considerable decrease in cell volume, which is apparently due to the net loss of K⁺ and Cl^- ions from the cytosol. It has previously been demonstrated that intracellular Cl^- activity falls towards equilibrium in salivary acinar cells following muscarinic stimulation, and that this quantitatively accounts for the agonist-induced decrease in cell volume (Foskett, 1990). Foskett proposed that this fall in intracellular Cl^- activity may be necessary to increase the thermodynamic gradient for cotransport-mediated Cl^- influx across the basolateral membrane, thus maintaining an elevated transepithelial Cl^- flux (Foskett, 1990).

10.1.7. Conclusions.

The characterisation of the effects of VIP and carbachol upon ion transport in the small intestinal crypt forms the core of this dissertation and provides a basis for further investigations into the coupling of intracellular signalling pathways to the regulation of ion channel and transporter activity during fluid and electrolyte secretion. The intracellular signals co-ordinating transcellular ion movements may include changes in cell volume and ionic activities, as well as changes in membrane potential and levels of intracellular second messengers such as cAMP and Ca²⁺. Further investigations using preparations of intact small intestinal crypts may advance our understanding of the cellular events underlying small intestinal secretion and intestinal proliferation and differentiation, with potential implications for the development of therapeutic approaches in the treatment of diarrhoeal disease, cystic fibrosis and cancer.

Viable and intact units of crypt epithelium were harvested from guinea-pig small intestine by vibration in Ca²⁺-free isolation media at 4° C. The presence of K⁺ and Cl^- channels in the small intestinal crypt epithelium could be inferred from membrane potential and conductance changes in both quiescent and agonist-stimulated small intestinal crypts using the nystatin perforated-patch whole-cell recording technique. The basolateral membrane of the mid-region of isolated small intestinal crypts is predominantly K⁺-selective, as inferred from experiments using ion substitutions and K⁺ channel inhibitors. Carbachol hyperpolarised crypt membrane potential in a manner that was partially dependent upon extracellular Ca²⁺. The carbachol-induced hyperpolarisation was attributed to the activation of a basolateral K⁺ conductance. No direct evidence for the activation of a Ca²⁺-activated Cl^- conductance by carbachol could be inferred, although phorbol esters evoked a slow biphasic depolarisation of crypt membrane potential consistent with the presence of a PKC modulated Cl^- conductance. VIP evoked a dose-dependent depolarisation of crypt membrane potential that was attributable at least in part to the activation of a Cl^- conductance. The effect of carbachol upon crypt membrane conductance was strongly potentiated by the addition of VIP. Electrophysiological recordings along the crypt axis also indicate that the resting membrane potential and changes in membrane potential evoked by carbachol vary with the stage of development of the crypt enterocyte.

Changes in crypt volume were estimated by image analysis, and revealed that both VIP and carbachol evoke a reversible decrease in crypt volume that is due to the activation of K⁺ and Cl^- channels. The participation of separate K⁺ and Cl^- conductance pathways in the recovery of crypt volume following exposure to hypotonic media, was also inferred from measurements of crypt volume.

Single-channel recordings from cell-attached and excised patches obtained from the mid-region of the crypt basolateral membrane have revealed the presence of an outwardly-rectifying, Ca²⁺-independent K⁺-selective sub-picosiemen conductance cation permeability and an inwardly-rectifying channel that is both voltage- and Ca²⁺-dependent in its activity. The inward-rectifier may participate in the recycling of cations across the basolateral membrane, whilst the K⁺-selective cationic permeability may underlie the resting K⁺ conductance present in small intestinal crypts. The addition of carbachol activates both a cation non-selective channel of intermediate conductance and a small conductance channel that is selective for K⁺ ions. The small conductance K⁺ channel probably underlies the carbachol-induced hyperpolarisation, whilst the intermediate conductance cation non-selective channel may serve to limit the extent of the carbachol-induced hyperpolarisation.

The present data strongly supports the contention that the crypt epithelial compartment mediates the secretion of fluid and electrolytes across the small intestinal mucosa. This preparation also has potential value in the study of the cellular mechanisms of enterocyte proliferation and differentiation and may be exploited to reveal the presence of other functional receptors and to provide a pharmacological basis for therapeutic approaches to cystic fibrosis and secretory diarrhoea.

Return to top
Return to table of contents
Return to science homepage
References
Chapter 1
Chapter 2
Chapter 3
Chapter 4
Chapter 5
Chapter 6
Chapter 7
Chapter 8
Chapter 9
All methods