■©.ju-c-nh 0 0



Sulfapyrimidine derivative

0 N


Fig. 21. Chemical structure of sulphonylureas.

Fig. 21. Chemical structure of sulphonylureas.

(Steinke et al., 1972), isolated perfused pancreas (Ammon and Abdel-Hamid, 1981), isolated B-cells (Gorus et al.. 1988) or hamster insulin-secreting tumour (HIT) cells (Nelson et al., 1987) results in rapid insulin release. Use of Bovine Serum Albumin (BSA) in Incubation Media and Sulphonylurea Actions. For all studies with sulphonylureas in vitro and in vivo the following should be considered. Sulphonylureas possess a high binding affinity for plasma proteins, but only the unbound form is pharmacologically active. In in vitro studies incubation media are frequently supplemented with BSA to stabilize cells. This holds also for experiments on insulin secretion. Bound and free sulphonylureas are in equilibrium. The use of BSA is therefore responsible for the fact that relatively high concentrations of sulphonylureas have been employed to study insulin secretion in vitro. Panten et al. (1989) developed a filtration assay for measuring free drug concentrations in the presence of BSA. In perfusion experiments with mouse pancreatic islets, they observed half-maximal insulin secretion for free glibenclamide (0.4 nM), glibizide (4nM) and tolbutamide (5/u.M), in excellent agreement with the equilibrium dissociation constants established for high-affinity binding.

2.1.2 Binding to and Uptake in Islet Cells

Sulphonylureas appear to induce their insulin secretory effect mainly from the extracellular site of the cytoplasmic membrane. Thus Hellman et al. (1984) have shown that, except for glibenclamide, sulphonylureas do not enter B-cells or only to a small extent (tolbutamide, carbutamide, chlorpropamide and glibizide).

This raises the question of whether sulphonylureas specifically bind to the cytoplasmic membrane and whether or not such binding induces a cascade of events which finally leads to exocytosis. In 1982 Kaubisch and colleagues reported specific binding of sulphonylureas to crude membrane fractions from brain and B-cell tumours. Studies performed by Geisen et al. (1985) showed similar results. Here, binding correlated with the hypoglycaemic action. Verspohl et al. (1990c) demonstrated the existence of more than one binding site for various sulphonylureas.

For rat B-cell tumour Kd values of 0.03 nM were reported by Geisen et al. (1985). The presence of high-affinity binding sites in B-cells was confirmed for RIN cells (Schmid-Antomarchi et al., 1987), HIT cells (Gaines et al., 1988) and mouse pancreatic islets (Panten et al., 1989). In all three studies, a high-affinity binding site was observed using [3H]glibenclamide (Kd) in the range 0.3-0.8 nM). Receptor isolation has been reviewed by Nelson et al. (1992). There is evidence that binding of sulphonylureas to their receptors occurs from the lipid phase of the B-cell membrane rather than from the cytoplasm (Zuenkler et al., 1989).

An important location of the sulphonylurea receptor is the ATP-sensitive K+ channel. Here, in contrast to ATP action, the binding site for sulphonylureas is not on the intracellular side but on the extracellular side of this channel (Niki et al., 1990). Niki et al. (1989, 1990) have provided evidence that ADP also binds to and competitively displaces glibenclamide from high-affinity HIT-cell sulphonylurea-binding sites. They also showed that ADP inhibited 86Rb+ efflux, elicited a rapid and sustained increase in [Ca2+]j and caused insulin secretion. Since ADP is unable to cross the cytoplasmic membrane, they concluded that ADP and sulphonylureas have common binding sites on the outer cell surface.

Recently, a- and /3-endosulfines, peptides isolated from brain, have been shown to bind to sulphonylurea receptors of brain membranes (Virsolvy-Vergine et al., 1992). The question arises whether they and/or ADP are physiological ligands of sulphonylurea receptors.

2.1.3 Binding to Other Tissues Brain. In rat cortical membranes [3H]gliquidone binding was specific, and could be displaced by other sulphonylureas. Dissociation constants were estimated for glibenclamide (0.06 nM), unlabelled gliquidone (0.9 nM), tolbutamide (1.4/u,M) and chlorpropamide (2.8/xM). The binding affinities were correlated with the rank order of the therapeutic doses (for a review see Nelson etal., 1992). A binding site for glipizide was also reported in rat cerebral cortex (Lupo and Bataille, 1987) with a Kd of 1.5 nM. Other areas of sulphonylurea (glibenclamide) binding in the brain (including substantia nigra, globus pallidus, hippocampus, etc.) have been identified (Treherne and Ashford, 1991).

In brain cells sulphonylurea-sensitive ATP-regulated K+ channels are present which play a role in neurosecretion at nerve terminals. ATP-regulated K+ channels in substantia nigra, a brain region that shows high sulphonylurea binding, are inactivated by high glucose concentrations and by antidiabetic sulphonylureas and are activated by ATP depletion and anoxia. ATP-regulated K+ channel inhibition leads to activation of GABA. These channels may be involved in the response of the brain to hyper- and hypo-glycaemia (in diabetes) and ischaemia or anoxia (Amoroso et al., 1990). Other Tissues. Sulphonylurea receptors have also been described for cardiac muscle, skeletal muscle and smooth muscle but do not appear to be of therapeutic benefit for lowering blood sugar (Panten et al., 1992). Specific binding to membranes isolated from other rat tissues (liver, lung, kidney, heart, spleen, diaphragm, duodenum, colon and stomach) was negligible. A sulphonylurea-binding protein in the plasma membrane of adipocytes has been proposed (Martz et al., 1989).

2.1.4 Membrane Potential and Ion Fluxes Membrane Potential. As early as 1970 Matthews and Dean reported depolarization of the B-cell by tolbutamide and glibenclamide. This observation was confirmed by others. Thus, in the presence of a non-stimulating glucose concentration (3 mM), tolbutamide and glibenclamide produced depolarization and spike activity (Meissner and Atwater, 1976; Meissner et al., 1979; Henquin and Meissner, 1982). These effects were not due to Na+ influx (Kawazu et al., 1980). K+ Efflux. As discussed above, inhibition of K+ efflux along the ATP-sensitive K+ channel causes depolarization. In fact, it has been shown that tolbutamide, in the presence of a glucose concentration (3 mM) that does not stimulate insulin secretion, inhibits HARb+ efflux, used as a measure of K+ efflux (Boschero and Malaisse, 1979; Henquin, 1980b; Henquin and Meissner, 1982). A similar effect was observed in the presence of glibenclamide (Gylfe et al., 1984). It was also reported that the addition of glibenclamide (20 /xM) or tolbutamide (1 mM) to the bathing medium of excised B-cell plasma membrane patches reduced the number of single ATP-sensitive K+-channel openings (Sturgess et al., 1985). Similar results were obtained in excised RIN-cell patches using 20 nM glibenclamide (Schmid-Antomarchi et al., 1987).

Mb 699, a benzoic acid derivative similar to the non-sulphonylurea moiety of glibenclamide, also inhibits 86Rb+ efflux, depolarizes the B-cell membrane and accelerates 45Ca2+ efflux from islet cells (Garrino et al., 1985). The authors suggest that a sulphonylurea group is not required to trigger the sequence of events finally leading to insulin release.

That inhibition of K+ efflux by sulphonylureas along ATP-sensitive K+ channels does not depend on ATP is evident from studies in which RIN cells were depleted of ATP by the use of 2-deoxy-d-glucose and oligomycin which block glycolysis and oxidative phosphorylation. Here, sulphonylureas such as glibenclamide inhibited 86Rb+ efflux (Schmid-Antomarchi et al., 1987). Similar results were obtained by Niki et al. (1989) using glibenclamide in HIT cells. Thus the evidence discussed so far indicates that sulphonylurea-induced depolarization is the result of inhibition of K+ permeability (Henquin, 1980b).

Chronic treatment of rats with glyburide (3mg per kg per day; intraperitoneal injection every hour for 9 days) increased its binding to heart and brain membranes. The authors concluded that KATP channels can be regulated after chronic treatment (Gopalakrishnan and Triggle, 1992). Whether this holds also for B-cells remains to be established. Ca2+ Fluxes. As can be expected, sulphonylureas increase net Ca2+ uptake along voltage-dependent Ca2+ channels (Henquin, 1980b; Amnion et al., 1986) and, as far as the chemical structure is concerned, only those sulphonylureas that produce insulin release enhance uptake of Ca2+ (Hellman, 1981). Uptake of Ca2+ is associated with increased [Ca2+]j (Abrahamson et al., 1985). In HIT cells, membrane depolarization effected by the addition of glibenclamide or tolbutamide increased intracellular Ca2+ by activating voltage-dependent Ca2+ channels (Nelson et al., 1987).

It thus seems clear that the initiating mechanism by which sulphonylureas promote insulin secretion is depolarization and subsequent uptake of Ca2+ which is similar to the mode of action of glucose. However, and here is the difference, sulphonylureas cannot replace the metabolism of glucose in B-cells. In contrast, their action on Ca2+ uptake, but not K+ efflux, depends on the concentration of glucose; in other words, it depends on the metabolism of glucose which seems to modulate the action of sulphonylureas on the B-cell.

2.1.5 Modulating Systems

The interrelationships between sulphonylureas and modulating systems of the B-cell can be seen from two aspects. First, do modulating systems interfere with the initiating action of sulphonylureas? Second, do sulphonylureas affect the modulating systems? As discussed in chapter 6, section 4, there are at least three modulating systems, i.e. glucose metabolism, the adenylate cyclase system and the PLC system. Glucose Metabolism. It has been extensively discussed in chapter 6, section 3.3.2, that glucose metabolism not only produces a signal (ATP) which, by interfering with the ATP-sensitive K+ channel, inhibits K+ efflux and thus causes depolarization, but also delivers one or more metabolic products which distally to the KAXP channel, and perhaps close to Ca2+ uptake, are involved in modulating Ca2+ influx and thus insulin secretion. Such metabolic products/factors were suggested to be the redox ratios of nicotinamide nucleotides and/or thiols (glutathione) (for a review see also Ammon and Wahl, 1994). In the case of sulphonylureas it was shown that Methylene Blue and thiol reagents that decrease these redox ratios diminished Ca2+ uptake and insulin secretion in response to tolbutamide and glibenclamide (Ammon et al., 1984, 1986). Since these compounds did not change ATP levels of pancreatic islets and failed to affect tolbutamide-induced inhibition of 86Rb+ efflux (Ammon and Wahl, 1994), it seems possible that the redox ratios of NAD(P)H/NAD(P)+ and GSH/GSSG are involved in sulphonylurea-mediated Ca2+ uptake and insulin secretion distal to depolarization.

The question of whether or not sulphonylureas interact with the metabo lism of pancreatic islets has also been studied. It was found that tolbutamide changed neither glucose metabolism nor the above redox ratios (Ammon, 1975), indicating that, in contrast with glucose, which initiates and modulates insulin release, sulphonylureas, as far as metabolism is concerned, possess only initiating activity. Moreover, they have even been found to decrease ATP levels (Hellmann et al., 1969; Kawazu et al., 1980). Adenylate Cyclase System. In the presence of sulphonylureas some increase in cAMP has been observed to occur in islet tissue. This effect seems, however, not to be of relevance for insulin secretion (Gylfe et al., 1984) because it is small (Taljedal, 1982). The effect on cAMP is probably due to Ca2+ influx (Malaisse and Malaisse-Lagae, 1984). Another possibility comes from the observation that sulphonylureas inhibit low-/Cm phosphodiesterase (Malaisse and Malaisse-Lagae, 1984).

As discussed above (see chapter 6, section 4.1), initiation of insulin secretion via depolarization can be modulated by compounds that affect the adenylate cyclase system. It is therefore not surprising that glucagon and db-cAMP potentiate tolbutamide-induced insulin secretion (Ammon, 1975). This also holds for methylxanthines which, at the concentrations used in vitro, inhibit phosphodiesterase and thus cAMP (Lambert et al., 1971; Ammon, 1975).

2.2 extrapancreatic effects

2.2.1 Introduction

While the pancreatic effects of sulphonylureas on their hypoglycaemic action are beyond doubt, the relevance of extrapancreatic effects is controversial. In this connection, the increased sensitivity of Type-II diabetics to insulin in response to sulphonylureas has been addressed repeatedly. One argument among others for the possibility of extrapancreatic effects is the observation that, on one hand, normalization of blood glucose levels after a long period of therapy with sulphonylureas does not necessarily correlate with increased plasma levels of insulin (Reaven and Dray, 1967; Barnes et al., 1974), but, on the other hand, hyperglycaemia returns after withdrawal of sulphonylureas. Moreover, sulphonylureas were observed to save some insulin in insulin-treated pancreactomized dogs (Beyer et al., 1972). In the same experimental animal sulphonylureas increased the effect of an intravenous insulin-tolerance test (Beck-Nielsen, 1988).

From a theoretical point of view there are three possible reasons for the extrapancreatic effects of sulphonylureas that lead to a lowering of blood glucose: (1) they increase insulin action; (2) they have insulin-like effects; (3) they have indirect effects.

COMPOUNDS ACTING ON INSULIN SECRETION: SULPHONYLUREAS 115 2.2.2 Increase in Insulin Actions and Insulin-like Effects

Target tissues for the enhancement of insulin effects and/or direct effects of sulphonylureas are adipose tissue, skeletal muscle and liver. Insulin Receptors. Insulin mediates its effects after binding to insulin receptors. Olefsky and Reaven (1976) reported that, in Type-II diabetes, treatment of patients with chlorpropamide increased the number of insulin receptors of monocytes. Similar results were achieved with human fibroblasts after treatment with glibenclamide (Prince and Olefsky, 1980). However, further studies carried out by others (for a review see BeckNielsen, 1988) have produced contradictory results. The major criticism was that some of the tissues studied are not involved in blood glucose lowering and that the questionable effect of sulphonylureas on insulin receptors may be of an indirect nature. Furthermore it was claimed that the doses of the sulphonylureas used were very high, leading to maximal secretion of insulin (Joost, 1985) or were even higher than necessary for blood glucose lowering (Maloff and Lockwood, 1981). It seems therefore unlikely that an increased number of insulin receptors is responsible for the potentiation of insulin action by sulphonylureas. Post-receptor Effects. Biological effects of insulin after binding to its receptor have been discussed elsewhere. Some of the post-receptor effects of insulin have been studied in the presence of sulphonylureas. The problem of dosage and binding of sulphonylureas to proteins added to incubation media has already been discussed. In connection with insulin release, first-generation sulphonylureas have been found to be effective in the micromolar range whereas second-generation sulphonylureas are effective in the nanomolar range. The studies discussed now must be seen in this light. Insulin Receptor Metabolism. Frank et al. (1985) determined the turnover rate of insulin receptors in liver of rats. They found a doubling after 6 days of treatment with 5mgkg_1 glibenclamide daily. If this is true, a longer-lasting insulin receptor binding would be conceivable. In this connection it should be mentioned that glibenclamide (10-1000 ng ml-1) inhibited degradation of insulin in endothelial cells (Kaiser et al., 1983). Glucose Transporter. Glucose is carried from the extracellular site to the interior of cells by so-called glucose transporters. One of them (GLUT-4) operates only after binding of insulin to its receptor. One measure for transporter-mediated hexose uptake is the determination of the cellular uptake of 3-O-methylglucose and 2-deoxyglucose. Unlike glucose which undergoes metabolism immediately after its uptake (and therefore measurement of tissue concentrations of glucose are not valid as a parameter of glucose uptake), 3-O-methylglucose and 2-deoxyglucose are transported by

GLUT-4 but are then not metabolized. In fact, in adipocytes glibenclamide (2/igmF1) increases insulin-induced translocation of glucose transporters from the microsomes to the cytoplasmic membrane (Jacobs and Jung, 1985). This might be an explanation for the increase in insulin-mediated uptake of 3-O-methyIglucose and 2-deoxyglucose into adipocytes caused by tolazamide (3-300 yu,g ml-1) and glibenclamide (l^gmr1) (Lockwood, 1983). Results similar to those in adipose tissue were obtained in skeletal muscle (Jacobs and Jung, 1985). Glucose Metabolism. As far as the metabolism of glucose is concerned it is interesting to note that sulphonylureas in liver and skeletal muscle per se (absence of insulin) increase fructose 2,6-bisphosphate, a metabolite of glycolysis. In perfused liver experiments, Hatao et al. (1985) reported stimulation of fructose 2,6-bisphosphate formation by first- and second-generation sulphonylureas, the maximum effect being produced by tolbutamide at 1 mM and glibenclamide at 1 /xM. This action is synergistic with that of insulin (50/liM glipizide, 300/¿g ml-1 tolbutamide) (Matsutani et al., 1984; Monge et al., 1985; Lopez-Alarcon et al., 1986). An increase in fructose 2,6-bisphosphate stimulates the activity of phosphofructokinase, increases glycogen synthesis and inhibits gluconeogenesis. Consequently, in hepatocytes an increase in insulin-mediated synthesis of glycogen, and inhibition of gluconeogenesis and glucose output from liver were seen in response to gliquidon (5 mgmf1) and glibenclamide (1.6/tig ml"1) (Fleig et al., 1984; Rinninger et al., 1984; McGuinness et al., 1987).

Insulin-stimulated 3-O-methylglucose uptake and glucose oxidation in adipocytes were potentiated when the tissue had been taken from rats given a daily dose of gliclazide (8mgkg_1) for 6 days (Hoich and Ng, 1986). Activation of glucose transport associated with activation of PKC in rat adipocytes by tolbutamide (1-2 mM) and glyburide (20-40 /¿M) was also reported (Farese et al., 1991). Similar effects of both have also been demonstrated in myocytes (Cooper et al., 1990b). In cultured hepatocytes, glyburide (2mgF!) directly inhibited glycogenolysis, stimulated glycogen synthesis and glycogen synthase and potentiated the action of insulin on glycogen synthesis at a post-binding site (Davidson and Sladen, 1987).

Since most of these extrapancreatic effects of sulphonylureas are coupled to the presence of insulin, it would be attractive to conclude that sulphonylureas possess an extrapancreatic action in Type-II diabetes where insulin secretion is still possible.

2.2.3 Critical Evaluation

In 1985 Joost pointed out that the extrapancreatic effects of sulphonylureas can be observed with tolbutamide at concentrations of 0.5-5 mM and with glibenclamide at 1-10 /u.M, whereas for maximal insulin secretion only 0.1 mM and 0.1 ¿iM respectively are necessary. Panten (1987), in a review, also raised this discrepancy considering that only a few per cent of a given drug are unbound and therefore biologically active. These facts make it difficult to conclude that in vitro data are transferable to the situation in vivo. This sceptical view is supported by the study of Mooradian (1987), who observed no improvement in insulin-mediated glucose uptake into liver, adipose tissue and skeletal muscle in rats given daily doses of 5 mg kg-1 glipizide.

If increasing insulin sensitivity is a common feature of sulphonylureas, then these compounds should act synergistically with insulin treatment in Type-I diabetes. However, this is not the case.

2.2.4 Indirect Effects

Taking into account that hyperglycaemia as such can cause insulin resistance, it is conceivable that, primarily through an increase in insulin secretion caused by sulphonylureas, hyperglycaemia and then normoglycaemia are decreased. Since sulphonylureas are more effective during hyperglycaemia and since after long-term treatment with sulphonylureas normoglycaemia can be achieved, it is also possible to understand why at a later phase of treatment no increase in plasma insulin is detectable.

3 Pharmacokinetics

Variations in the pharmacokinetics of the sulphonylureas are clinically relevant because of the differences in their rate of onset and their duration of action. Differences in the rate of onset are important because they relate to the capacity to reduce the delay in acute insulin release after nutritional challenge and therefore their capacity to reduce the evaluation and prolongation of the postprandial hyperglycaemia. Differences in the duration of action are important because they relate to the risk of causing chronic hyperin-sulinaemia, long-lasting hypoglycaemia, and possibly desensitization to sulphonylureas (Melander et al., 1990).

3.1 dosage and application

Treatment with a sulphonylurea is begun with a small dose of medication and a progressive build-up over a 1-2-week period (Asmal and Marble, 1984). This is important in order not to induce hypoglycaemia. Furthermore, the progress of the patient must be followed closely in the first weeks of treatment. The dose must be increased progressively in order to avoid side effects. After treatment for 6 months or 1 year, reduction of the dose or complete withdrawal of the treatment must be considered (Beck-Nielsen, 1991).

The most appropriate dosage schedules for the application of sul-phonylureas are not yet established. Available data suggest that once-daily (morning) administration 30 min before breakfast may improve the efficacy of sulphonylurea treatment, but only if the exposure to the drug is continuous (Samanta et al., 1984). If postprandial hypoglycaemia ensues in the early part of the day, or if inadequate glycaemic control occurs in the later part of the day, a divided dosage schedule could be tried (Melander et al., 1990).

The optimal daily dosage is difficult to define, as it is dependent on the degree of impairment of /3-cell function before treatment and also the degree of compliance with diet regulation. Furthermore, the dose-response curve may be bell-shaped, and the steady-state concentrations of several, if not all, sulphonylureas show a large interindividual variation following standard doses (Melander et al., 1989).

3.2 pharmacokinetic parameters (table 5)

3.2.1 Absorption

Absorption of both first- and second-generation sulphonylureas is rapid and complete except for gliclazide and tolazamide, which are absorbed more slowly. The maximal plasma concentrations are usually reached within 2-4 h. The kinetics of absorption depend on the formulation and crystalline structure of the drug. Absorption of chlorpropamide may also depend on pH and therefore on food intake, which appears not to be true for the other sulphonylureas (Sartor et al., 1980). The absorption of glibenclamide, although rapid and almost complete, can be improved by an appropriate formulation (Haupt et al., 1984) which leads to a reduction in the daily dosage required.

At least for some sulphonylureas (gliclazide and glipizide), intra- and inter-individual variations in absorption are pronounced, perhaps explaining the large variations observed in their responses (Hartling et al., 1987).

3.2.2 Distribution

The volume of distribution of the sulphonylureas is between 0.1 and 0.3 1 kg~'. Their plasma protein binding is about 95-99%, but decreases for tolbutamide with age in healthy volunteers showing concentration-dependence (Adir etal., 1982). Displacement from plasma-binding sites is uncharac-


Pharmacokinetic characteristics of oral hypoglycaemic agents


Pharmacokinetic characteristics of oral hypoglycaemic agents


Time of

Volume of




0 0

Post a comment