CHAPTER 1

INTRODUCTION

The activation of K⁺ and Cl^- permeability pathways is known to be central to cellular homeostasis during solute absorption, volume regulation and fluid and electrolyte secretion in a variety of epithelia, including the small intestinal epithelium (Frizzell and Halm, 1990; Dawson & Richards, 1990). The development of viable enterocyte preparations and the advent of the patch-clamp single channel recording technique (Neher & Sakmann, 1976) has allowed the characterisation of K⁺ and Cl^- selective ion channels present in the small intestinal epithelium (Giraldez et al, 1989; Sepulveda et al, 1991).

The mechanisms of small intestinal ion transport are poorly understood at the cellular level, particularly in the crypt epithelial compartment, where its contribution to the secretion of fluid and electrolytes by the small intestinal epithelium is uncertain. The initial aim of this research was to isolate viable populations of identifiable crypt enterocytes from the guinea-pig small intestine suitable for electrophysiological characterisation. Studies of the effects of established secretagogues upon the membrane conductances present in small intestinal crypts using the patch-clamp technique would provide a means of testing the hypothesis that the crypt is a site of fluid and electrolyte secretion and establish the presence of any second messenger pathways that may be involved. The single channel mode of the patch-clamp technique has proved valuable in the characterisation of the selectivity, regulation and pharmacology of the single channel conductances that underlie the macroscopic conductance pathways present in epithelial cells (Sheppard et al., 1988a, 1988b) and will be employed to elucidate the nature of the conductances present in the crypt enterocyte. Measurements of cell volume in populations of isolated villus enterocytes has provided an insight into the homeostatic mechanisms that maintain enterocyte volume following exposure to aniso-osmotic media (MacLeod & Hamilton, 1991a), and thus populations of isolated crypt enterocytes will be exploited to determine the volume regulatory mechanisms present in the crypt epithelium.

In order to effect the net transport of fluid and electrolytes across the intestinal epithelium, a selective permeability barrier must be generated that regulates the exchange of ions and nutrients between the gut lumen and the internal environment. By the functional division of the plasma membrane into apical and basolateral domains, an asymmetrical distribution of surface enzymes, ionic channels and solute transporter proteins may be established, endowing the epithelium with the capacity to transport fluid and electrolytes vectorially between the serosal and lumenal compartments (for review see Steward & Case, 1989).

The small intestine is typical of `leaky', or low resistance, epithelia that are characterised by their high conductance paracellular pathway, permeability to water and inability to maintain steep transepithelial concentration gradients. Leaky epithelia can transport relatively large volumes of fluid isotonically. Because of their inability to sustain large transepithelial gradients of solutes, the salt concentrations of the fluid compartments that bathe these tissues are near to equilibrium. Their high conductance paracellular pathway serves to minimise voltage differences between the apical and basolateral membranes (Steward & Case, 1989). The cells of the intestinal epithelium display a relatively hyperpolarised apical membrane potential (V_a) that arises by virtue of a dominant basolateral K⁺ conductance and a high ratio of paracellular to apical membrane conductance, which permits the basolateral K⁺ conductance to hyperpolarise V_a. As a result, the diffusional driving force for Cl^- exit at the apical membrane opposes Cl^- absorption, as V_a is more negative than E_Cl. Thus if present, the activation of Cl^- channels would result in Cl^- exit from cell to lumen (for review see Frizzell & Halm, 1990).

The hyperpolarisation of V_a by a basolateral K⁺ conductance is necessary to maintain the electrochemical driving gradient for Na⁺-coupled solute entry during intestinal absorption, as these transport systems are electrogenic (Dawson & Richards, 1990). However, the small intestine is able to effect the net electroneutral absorption of NaCl by the coupling of Cl^- movements to Na⁺ entry, a process thought to be mediated by the parallel activation of Na⁺-H⁺ and Cl^--HCO₃^- exchangers which are located in the apical membrane domain of rabbit small intestinal villus cells (Knickelbein et al, 1988).

A basolateral membrane K⁺ permeability is present in both absorptive (Schultz, 1981) and secretory epithelia (Dawson & Richards, 1990). Small conductance K⁺-selective channels activated by L-alanine transport have been identified in Necturus enterocytes (Sheppard et al, 1988a) and large conductance K⁺ channels that are both Ca²⁺ and voltage independent have been described in guinea pig villus enterocytes (Mintenig et al, 1992) and in the basolateral membrane of crypts isolated from rat duodenum (Fraser et al, 1991). Most Cl^--secreting epithelia have been shown to possess a basolateral K⁺ conductance that is regulated by increases in intracellular free Ca²⁺. Ca⁺-activated K⁺ (maxi-K⁺) channels of both a large (Morris et al, 1986) and small conductance (Sepulveda & Mason, 1985) have been demonstrated to be present in isolated enterocytes but their role in regulating ion transport across the intestinal mucosa remains to be established.

The small intestinal epithelium is composed of several different cell types that mediate a variety of functions including nutrient absorbtion, fluid and electrolyte secretion, mucus secretion and the sampling of antigens from the lumenal environment. Whilst enterocytes from the amphibian intestine (e.g. Necturus) appear to have the capacity both to absorb and secrete fluid and electrolytes (Giraldez et al., 1988a; Sheppard et al., 1988a), it is suggested that anatomically separate epithelial cell populations may mediate absorption and secretion across the mammalian small intestine (Field, 1979). The architecture of the small intestinal epithelium is composed of finger-like villi that project into the small intestinal lumen and the invaginations of the small intestinal epithelium, known as the crypt compartments, which descend into the lamina propria, and both epithelial compartments can be clearly distinguished in haematoxylin and eosin stained sections of small intestine (see Fig.1.1). It has long been proposed that the enterocytes lining the villus tips mediate the processes of net nutrient, electrolyte and fluid absorption, whilst the enterocytes of the crypt region are proposed to actively secrete fluid and electrolytes (Donowitz & Welsh 1991). The absorptive role of the villus enterocytes is well established (Sullivan & Field, 1992). The consensus that secretion is a property of the crypt compartment is largely based on indirect evidence as few physiological studies have been able to address this question at the cellular level (Sullivan & Field, 1991). Although the selective hypertonic lysis of villus cells failed to attenuate cholera toxin-induced secretion in the rabbit jejunum (Roggin et al., 1972), Escherichia coli heat-stable enterotoxin has been demonstrated to stimulate Cl^- efflux from isolated porcine villus and crypt cells (Ahrens & Panichkriangkria, 1985). Intracellular microelectrode measurements of apical membrane potential changes evoked by the secretagogues 5-hydroxytryptamine (5-HT), acetylcholine (ACh) and prostaglandin E₂ (PGE₂) in the rat small intestinal epithelium have suggested that secretion might be a property of both villus and crypt cells (Stewart & Turnberg 1989). However whole-cell patch-clamp studies of epithelial cells isolated from the guinea-pig villus region indicate that these cells do not play a role in the secretory process as they show no conductance changes in response to secretagogues or agents that elevate intracellular cAMP or Ca²⁺ (Sepúlveda et al., 1991).

Direct experimental approaches at a cellular level are needed to demonstrate that small intestinal crypt enterocytes mediate the secretion of fluid and electrolytes. Progress to date has been limited by the comparative inaccessibility of the crypt compartment of the small intestinal epithelium to microelectrodes, the small size of crypt enterocytes and the difficulty in isolating `homogeneous' populations of identifiable crypt enterocytes. As the crypt is the site of stem cell division, cell proliferation and ultimately of differentiation it is possible that not all the enterocytes along the crypt axis will respond in the same manner to secretagogues and thus a novel strategy is required to study transport processes in the intact crypt epithelium.

Fluid secretion across the small intestine is thought to involve the activation of apically sited Cl^- channels allowing the electro-diffusive exit of Cl^- accumulated intracellularly by a bumetanide-sensitive NaK2Cl cotransport mechanism that couples the uptake of Cl^- and K⁺ ions to the electrochemical gradient for Na⁺ entry. Evidence for the presence of a bumetanide-sensitive ⁸⁶Rb⁺ uptake has recently been demonstrated in villus and crypt units isolated from rat duodenum (McNicholas et al., 1992). The cations taken up by the NaK2Cl cotransporter are recycled across the basolateral membrane via the electrogenic Na⁺-K⁺ ATPase and K⁺-selective ion channels to maintain an electrochemical gradient favourable to Cl^- secretion (Dawson, 1987). The intracellular accumulation of Cl^- above equilibrium in enterocytes has been demonstrated in several species (White, 1980; Giraldez & Sepúlveda, 1987). Na⁺ ions, crossing the epithelium via the cation-selective paracellular route, and water would accompany the flow of Cl^-.

The presence of an apically located Cl^- conductance that is activated as a result of increases in intracellular cAMP has been demonstrated in many secretory epithelia including the trachea (Welsh et al., 1982) and Necturus small intestine (Giraldez et al., 1988a). However the expression and localisation of such a cAMP-activated Cl^- conductance in the small intestinal epithelium between anatomically distinct epithelial cell populations has yet to be demonstrated. However, considerable interest and indirect evidence for such a conductance in the intestine is inferred from studies of the Cystic Fibrosis Transmembrane conductance Regulator (CFTR), the gene product that is defective in CF. Individuals who are homozygous for the mutant CFTR gene product suffer a deficit in conductive Cl^- transport across the epithelia of the lung, pancreas, sweat gland and intestine. In the intestine CF is characterised by the impaired hydration of secreted mucus, and defects in the Ca²⁺, cGMP and cAMP mediated activation of a Cl^- conductance at a site distal to their respective protein kinases have been shown to occur in the human CF jejunum (O'Loughlin, 1991). Speculation that a cAMP-activated Cl^- conductance may be present in the mammalian small intestine is supported by recent work showing high levels of CFTR mRNA expression in rat small intestinal crypts (Trezise & Buchwald, 1992).

Many epithelial cells are known additionally to express Cl^- channels that are activated by cell swelling and elevations in intracellular free Ca²⁺. The presence of volume-activated Cl^- channels has been demonstrated in the T84 colonic carcinoma cell line (Worrell et al., 1989), Necturus enterocytes (Giraldez et al., 1988b) and in isolated colonic crypts (M.Diener, 1992), and a volume-sensitive Cl^- conductance may be present in the small intestinal villus epithelium (MacLeod & Hamilton, 1991a). In the CF small intestine both cAMP and Ca²⁺-stimulated Cl^- secretion have been reported to be defective (Berschneider et al., 1988; Taylor et al., 1987; De Jonge et al., 1987) suggesting the absence of a Ca²⁺-activated Cl^- conductance in the small intestinal epithelium.

A variety of chemical factors secreted by the neural or endocrine elements present in the small intestine are believed to induce alterations in small intestinal electrolyte transport. Neurotransmitters and humoral factors are thought to interact with specific membrane bound receptors present in the epithelium to activate the intracellular signalling cascades which mediate changes in cellular ionic permeabilities. Several classes of biologically active substances have been demonstrated to regulate small intestinal ion transport including the amine neurotransmitters, purines, neurogastrointestinal peptides and circulating hormones that regulate body fluid and electrolyte homeostasis. Efferent neurones containing acetylcholine, 5-HT, bombesin and vasoactive intestinal polypeptide (VIP), agents which increase anion secretion across the guinea-pig small intestinal mucosa, have been demonstrated histochemically to be present in the small intestinal submucosa (Brown & Miller, 1992). However a number of other substances, such as atrial natriuretic peptide (ANP, Catto-Smith et al., 1991), ATP (Kohn et al., 1967, 1970) and histamine (Cooke et al., 1984b) have also been shown to evoke Cl^- secretion across the small intestinal mucosa. Histamine and ATP may be released from other cell types present in the mucosa, such as mast cells; whilst ANP induces changes in small intestinal ion transport when infused into the general circulation (Kanai et al., 1987).

Two of the most extensively characterised small intestinal secretagogues are acetylcholine and VIP. Both VIP and muscarinic agonists have been demonstrated to evoke Cl^- secretion and an increase in short-circuit current across the guinea-pig ileum (Carey et al., 1987; Cooke et al., 1984a). Muscarinic receptors, coupled to the mobilisation of intracellular Ca²⁺ and diacylglycerol (DAG), have been demonstrated to be present on both villus and crypt enterocytes isolated from the rat small intestine (Isaacs et al., 1982), and high affinity binding sites for VIP have been shown to be present on isolated guinea-pig enterocytes (Binder et al., 1980), suggesting that these secretagogues modulate epithelial electrolyte transport via a direct action on the epithelium. The cellular mechanisms of action of these classical secretagogues in modulating electrolyte transport across the small intestinal crypt epithelium remain to be established.

In the small intestine cyclic nucleotides, Ca²⁺ and arachidonic acid metabolites act as intracellular second messengers (Donowitz & Welsh, 1991) mediating the excitatory and inhibitory actions of hormones and neurotransmitters (including VIP, acetylcholine, substance P and 5-HT) and substances mediating inflammatory responses (e.g. the prostaglandins and histamine). Electrolyte and fluid secretion in the small intestine can also be initiated by certain bacterial toxins known to act by elevating intracellular levels of cAMP, cGMP or Ca²⁺ (reviewed by Field & Semrad, 1993). Many of these agents are believed to interact at the basolateral membrane of enterocytes to elevate or suppress levels of intracellular second messengers that modulate ion transport in both the apical and basolateral membranes. Intracellular messengers are believed to modulate protein activity either directly, by altering the conformation of receptive proteins, some of which subsequently act on target proteins or by activating protein kinases which then phosphorylate target proteins (Cohen, 1982). Evidence is accumulating that second messengers act synergistically (e.g. Ca²⁺ and cAMP, Malhotra et al., 1989; Larsson & Olgart, 1989) or antagonistically (e.g. cAMP and cGMP, MacFarland et al., 1991) to regulate ion transport.

There is a functional coupling between conductive processes present in the apical and basolateral membranes of absorptive and secretory epithelial cells. This means that electrogenic Cl^- efflux across the apical membrane must be accompanied by a proportional change in basolateral K⁺ efflux to maintain the electrochemical gradient for Cl^- exit. Changes in intracellular second messenger levels provide an integrative mechanism regulating the rate of apical Cl^- efflux and basolateral K⁺ efflux in parallel, a phenomenon known as membrane cross-talk.

As the electrochemical gradient for Cl^- efflux is believed to be maintained by increases in K⁺ permeability; the rate of apical Cl^- exit and basolateral K⁺ efflux are regulated in parallel and therefore potentially linked ("cross-talk"). Thus there must be a level of integration between the conductances modulated by various pathways, possibly at the level of second messenger metabolism, intracellular ATP concentration, membrane potential and intracellular ion activities.

The first second messenger reported to increase intestinal secretion was cAMP (Field, 1971). Secretagogues such as PGE₂ (Kimberg et al., 1971) and VIP (Schwartz et al., 1974) increase intracellular cAMP levels by stimulating adenylate cyclase activity. Adenylate cyclase activity may also be irreversibly stimulated by bacterial enterotoxins including cholera toxin and Escherichia coli heat-labile enterotoxin. The resulting increases in cAMP levels are believed to inhibit electroneutral NaCl transport in absorptive villus cells and to activate Cl^- channels in crypt cells thus inducing secretory diarrhoea (Gill & Woolkalls, 1985).

The proposal that cholinergic agonists mobilise intracellular Ca²⁺ in vivo is supported by observations that secretory responses evoked by carbachol in the rabbit and rat intestinal mucosa are partially dependent upon extracellular Ca²⁺ (Hardcastle et al., 1984) and TMB-8, an inhibitor of Ca²⁺ mobilisation from intracellular stores, abolishes the carbachol-induced elevations in I_sc across the rabbit ileal mucosa (Donowitz et al., 1986). Increases in cytosolic free Ca²⁺ levels following exposure to carbachol have been demonstrated in chicken villus enterocytes using intracellular fluorescent Ca²⁺ chelators (Chang et al., 1986), although direct evidence for intracellular Ca²⁺ mobilisation by secretagogues has yet to be demonstrated in crypt enterocytes.

Many animal cells alter their osmolyte content during exposure to aniso-osmotic media by the transient activation of ion transport pathways associated with osmotically obliged water movement (reviewed by Sarkadi & Parker, 1991). Cells of the intestinal epithelium will be exposed more readily to both intra- and extracellular anisotonicity due to the nature of their transport activity and will routinely encounter fluctuations in luminal osmolarity. Enterocytes also have to maintain a constant volume during the absorbtion and secretion of large amounts of solute and fluid (Schultz & Hudson, 1986). It may therefore be expected that enterocytes will possess mechanisms regulating the transport processes that mediate solute entry and exit (MacLeod & Hamilton, 1991b), and the measurement of cell volume provides a good indicator of the balance between these processes (Hoffmann & Simonsen, 1989).

In response to swelling in a hypotonic environment or as a result of an increased uptake of osmolytes by Na⁺/nutrient cotransport, cells effect the passive loss of organic osmolytes and/or KCl by the activation of normally quiescent transport pathways in a process termed regulatory volume decrease (RVD). Three different mechanisms (illustrated in Fig.1.3) have been proposed to mediate the net efflux of K⁺ and Cl^- ions during RVD (reviewed by Hoffman & Simonsen, 1989): the parallel activation of electroconductive channels selective to both anions and K⁺, electroneutral cotransport and in Amphiuma red blood cells functionally coupled H⁺/K⁺ and Cl^-/HCO₃^- exchange (Cala, 1985). The importance of the activation of Cl^- and K⁺ permeability pathways in enterocyte volume homeostasis has recently been confirmed by direct measurements of cellular volume in guinea-pig villus enterocytes (MacLeod & Hamilton, 1991a, 1992a). In response to cell shrinkage induced by a hyperosmotic challenge animal cells stimulate net ion and osmotically obliged water uptake leading to a recovery of original cell volume, a process known as regulatory volume increase (RVI). Three principal transport mechanisms, illustrated in Fig.1.3, have been proposed to mediate ion uptake following a decrease in cellular volume; electroneutral NaCl cotransport; electroneutral NaK2Cl cotransport and electroneutral Na⁺-H⁺ exchange functionally coupled to Cl^--HCO₃^- exchange (Sarkadi & Parker, 1991). These transporters are distinguished by the interdependency of K⁺, Na⁺ and Cl^- ions present on the same side of the membrane, their ion selectivity, and the specificity of pharmacological inhibition.

Whilst the exocrine model of secretion seeks to explain the stoichiometry of the ionic movements associated with electrogenic Cl^- transport, there are important consequences for the physiology of the secretory cell that are not considered in detail by this model. The activation of Cl^- and K⁺ selective conductances by secretagogues might result in a net loss of K⁺ and Cl^- ions down their respective electrochemical gradients. This net loss of ions from the cytosol would be accompanied by an osmotically obliged water efflux, resulting in a tendency for cell shrinkage. Such a phenomenon has been described for dissociated sweat coil gland cells challenged with a muscarinic agonist (Suzuki et al., 1991). The requirement for sustained Cl^- secretion would demand a subsequent increase in the rate of basolateral ion uptake, which in turn would be accompanied by an osmotically obliged uptake of water into the cell. Thus the cells of Cl^- secreting epithelia may possess adaptive transport processes regulating their volume during secretion. The effects of secretagogues upon enterocyte volume remain to be determined.

The development of techniques using conventional microelectrodes for the intracellular recording of cellular electrical phenomena has permitted the investigation of the electrical responses of the apical membrane of the intestinal epithelium to secretagogues (Stewart and Turnberg, 1989) and changes in extracellular osmolarity (Giraldez et al., 1988b). However, the resolution of intracellular microelectrode recordings is limited by the leak current resulting from membrane puncture of the small, relatively inaccessible cells and the inability to characterise the individual channels underlying the macroscopic conductances.

The discovery that blunt, polished glass micropipettes can form seals of a high electrical resistance when brought in contact with the cell membrane (Neher & Sakmann, 1976; Hamill et al., 1981) has made possible the resolution of picoampere-size currents flowing through individual ion channels. This technique of single-channel recording in the "cell-attached" configuration offers the advantages of electrically isolating a patch of membrane and controlling the composition of fluid bathing its extracellular face, whilst leaving the intracellular environment relatively undisturbed. In polarised epithelial monolayers, such as preparations of intact small intestinal crypts, ion channels may be localised to a specific membrane domain (see Fig.1.4.). Furthermore, the patch-pipette could be withdrawn ("excised") from a cell with the seal intact (Horn & Patlak, 1980), permitting exposure of both the intracellular ("inside-out") and extracellular ("outside-out") faces of a membrane patch to bathing solutions of defined composition, allowing the detailed resolution of the selectivity and regulation of individual ion channels.

The cell-attached mode of the patch-clamp technique allows the resolution of agonist-induced changes in the activity of individual ion channels in a patch constituting only a small proportion of the total membrane surface area of the cell (Fig. 1.4.). In order to resolve the nature of the macroscopic conductances present in the cell membrane, the membrane patch isolated by the patch pipette may be ruptured by the application of strong suction to the pipette, without a resultant loss of the gigaohm seal (Hamill et al., 1981). In this `conventional whole-cell' configuration the patch electrode effectively becomes an intracellular recording electrode with a low input resistance, without the cell impalement damage associated with the use of intracellular microelectrodes and can be employed both to inject current into the cell and record the intracellular potential. This permits the intra- and extracellular ionic environments of a cell to be defined and thus allows the determination of both the selectivity and regulation of the component conductances of the cell membrane. However electrical changes evoked by agonists have been shown to `run down' rapidly in this conventional "whole-cell" configuration (Horn & Marty, 1988). By employing polyene pore-forming antibiotics such as nystatin or amphotericin B in the patch-pipette it is possible to obtain low access resistance whole-cell recordings whilst still in the cell-attached configuration (Horn & Marty, 1988), avoiding the dialysis of intracellular macromolecules associated with the conventional whole-cell recording technique (Pusch & Neher, 1988). The perforated-patch whole-cell recording method is depicted in Fig. 1.4., where patch pipettes back-filled with a nystatin-containing solution permeabilise the isolated crypt basolateral membrane patch typically within 1-3 minutes of GÛ seal formation. Furthermore resting membrane potentials can be measured accurately in intact epithelial monolayers where the cells are in gap-junctional communication even where intracellular Na⁺ and K⁺ activities are substantially different from those present in the patch pipette (Rae & Fernandez, 1990). The electrophysiological characteristics of small intestinal villus cells have been studied using both the conventional whole-cell (Sepulveda et al., 1991) and single channel configurations (Mintenig et al., 1992) of the patch-clamp technique. The nature and regulation of the ionic permeabilities present in crypt enterocytes remains to be characterised in the whole-cell configuration, although single channel recordings of a basolaterally located large conductance K⁺ channel have recently been obtained from isolated rat duodenal crypts (Fraser et al., 1991).

Changes in crypt volume were estimated by image analysis to further characterise the actions of VIP and carbachol upon crypt membrane conductance. This technique was also employed to characterise the mechanisms of crypt cell volume regulation following exposure to hypotonic media. Finally single channel recordings from the crypt basolateral membrane were performed to establish the single channel correlates of the macroscopic conductances present in the quiescent and agonist-stimulated crypt.