Troglitazone treatment increases plasma vascular endothelial growth factor in diabetic patients and its mRNA in 3T3-L1 adipocytes.

Troglitazone is one of the thiazolidinediones, a new class of oral antidiabetic compounds that are ligands of peroxisome proliferator-activated receptor-gamma. This study on vascular endothelial growth factor (VEGF), also known as vascular permeability factor, was prompted by our clinical observation that the characteristics of troglitazone-induced edema were very similar to those caused by vascular hyperpermeability. When Japanese diabetic patients were screened for plasma VEGF, we found levels to be significantly (P < 0.001) increased in troglitazone-treated subjects (120.1 +/- 135.0 pg/ml, n = 30) compared with those treated with diet alone (29.2 +/- 36.1 pg/ml, n = 10), sulfonylurea (25.8 +/- 22.2 pg/ml, n = 10), or insulin (24.6 +/- 19.0 pg/ml, n = 10). Involvement of troglitazone in increased VEGF levels was further supported by the plasma VEGF levels in five patients before treatment (20.2 +/- 7.0 pg/ml), after 3 months of troglitazone treatment (83.6 +/- 65.9 pg/ml), and 3 months after discontinuation (28.0 +/- 11.6 pg/ml). We further demonstrated that troglitazone, as well as rosiglitazone, at the plasma concentrations observed in patients, increased VEGF mRNA levels in 3T3-L1 adipocytes. VEGF is an angiogenic and mitogenic factor and is currently considered the most likely cause of neovascularization and hyperpermeability in diabetic proliferative retinopathy. Although increased VEGF may be beneficial for subjects with macroangiopathy and troglitazone is currently not available for clinical use, vascular complications, especially diabetic retinopathy, must be followed with great caution in subjects treated with thiazolidinediones.

Electrophysiological effects of FK664, a new cardiotonic agent, on preparations from guinea pig ventricle and from rabbit sino-atrial node.

Effects of the cardiotonic agent FK664, 6-(3, 4-dimethoxy-phenyl)-1-ethyl-4-mesitylimino-3-methyl-3,4-dihydro-2 (1H)-pyrimidone, on isolated guinea pig ventricular muscles and rabbit sinus node pacemaker cells were studied using micro-electrode techniques. In ventricular muscles driven at 0.5-1.0 Hz, FK664 above 3 mumol.litre-1 caused an increase in contractile force and a shortening of time to peak tension. This positive inotropic effect of FK664 was accompanied by a slight elevation of the early plateau phase of the action potential, while other action potential variables were unaffected. The change in contractile force induced by FK664 was abolished in a low Ca2+ medium (0.12 mmol.litre-1) or by treatment with ryanodine (2 mumol.litre-1), whereas it was relatively well preserved in the preparations pretreated with nefedipine (1 mumol.litre-1). The slow action potentials induced by isoprenaline (0.3 mumol.litre-1) in high K+ medium (30 mmol.litre-1) and the slow inward current measured by single sucrose gap voltage clamp at a holding potential of -40 mV were unaffected by FK664. In sinus node pacemaker cells, FK664 (1-10 mumol.litre-1) caused a dose dependent acceleration of phase 4 depolarisation and a shortening of spontaneous firing cycle length. This positive chronotropic effect of FK664 was markedly inhibited in a low Ca2+ medium (0.3 mmol.litre-1). These findings suggest that FK664 has positive inotropic and chronotropic effects on the heart, due to an enhancement of transsarcolemmal calcium influx through the low threshold, dihydropyridine insensitive Ca2+ channel population.

Relation of electrocardiographic left ventricular hypertrophy with and without T-wave changes to systemic blood pressure, body mass, and serum lipids and blood glucose levels in Japanese men.

Left ventricular (LV) hypertrophy, especially combined with an abnormal ST-T, is considered 1 of many coronary risk factors. Seven hundred forty-nine Japanese men were selected according to their electrocardiographic findings, i.e., normal electrocardiogram, LV hypertrophy without an abnormal ST-T segment, LV hypertrophy with a flat T wave, and LV hypertrophy with a negative T wave. Coronary risk factors were compared among these 4 age-matched groups. Groups with LV hypertrophy with negative or flat T waves had larger body mass index (24.9 vs 22.9 kg/m2), higher mean systemic blood pressure (111 vs 95 mm Hg), larger LV mass (265 vs 157 g), higher blood glucose (110 vs 100 mg/dl), higher serum triglyceride (148 vs 122 mg/dl), higher total cholesterol (206 vs 198 mg/dl), and lower high-density lipoprotein cholesterol (47 vs 54 mg/dl) than the normal group or the group with LV hypertrophy without T-wave change. Among these risk factors, blood pressure and glucose remained higher even after the adjustment by body mass index or by body mass index and blood pressure. Electrocardiographic LV hypertrophy with a changed T wave signified higher risk of coronary artery disease in Japanese men.

Electrophysiological effects of Ro 22-9194, a new antiarrhythmic agent, on guinea-pig ventricular cells.

1. Cardiac effects of Ro 22-9194 were examined in papillary muscles and single ventricular myocytes isolated from guinea-pigs and compared with those of moricizine. 2. In papillary muscles, both Ro 22-9194 (> or = 10 microM) and moricizine (> or = 1 microM) caused a significant dose-dependent decrease in the maximum upstroke velocity (Vmax) and a shortening of the action potential duration. 3. In the presence of either drug, trains of stimuli at rates > or = 0.2 Hz led to an exponential decline in Vmax. This use-dependent block was enhanced at higher stimulation frequencies. A time constant (tau R) for Vmax recovery from the use-dependent block was 9.3 s for Ro 22-9194 and 26.4 s for moricizine. 4. The curves relating membrane potential and Vmax in single myocytes were shifted by Ro 22-9194 (30 microM) or by moricizine (3 microM) in a hyperpolarizing direction by 8.4 mV and 8.0 mV respectively. 5. In myocytes treated with Ro 22-9194 (30 microM), a 10 ms conditioning clamp to 0 mV caused a significant decrease in Vmax of the subsequent test action potential; further prolongation of the clamp pulse duration resulted in a modest enhancement of the Vmax inhibition. In the presence of moricizine (3 microM), a similar conditioning clamp > 200 ms caused a significant Vmax reduction; the longer the clamp pulse duration, the greater the Vmax reduction. 6. Ro 22-9194 > or = 30 microM caused a slight decrease of calcium inward current (ICa) of myocytes without affecting the delayed rectifier potassium current (IK). 7. These findings suggest that the primary electrophysiological effect of Ro 22-9194 as an antiarrhythmicagent is, like moricizine, a use- and voltage-dependent inhibition of sodium channels. From the onset and offset kinetics of the use-dependent block, Ro 22-9194 belongs to the intermediate kinetic Class I drugs, while moricizine is a slow kinetic drug. From the state-dependence of sodium channel block, Ro 22-9194 may belong to activated channel blockers, while moricizine belongs to inactivated channel blockers.

Effects of calcium, calcium entry blockers and calmodulin inhibitors on atrioventricular conduction disturbances induced by hypoxia.

Effects of hypoxia on atrioventricular conduction were investigated in the Langendorff-perfused isolated heart of the rabbit with various extracellular calcium concentrations ([Ca2+]) as well as in the presence of verapamil, nifedipine, N-(6-aminohexyl)-5-chloro-1-naphthalenesulphonamide (W-7) and chlorpromazine. The prolongation of the atrio-His (AH) interval by hypoxia for 7 min was greater with increasing [Ca2+]o ranging from 1.2 to 5.2 mM. At [Ca2+]o of over 3.2 mM under hypoxic conditions, AH block of the Wenckebach type was observed in some cases. Verapamil (5 X 10(-8) M) and nifedipine (5 X 10(-8) M) caused a significant prolongation of AH intervals before hypoxia. However, the intensity of AH prolongation due to hypoxia was significantly attenuated in the presence of the calcium entry blocker, and AH block was not induced even at 3.2 mM [Ca2+]o. W-7 (5 X 10(-6) M) and chlorpromazine (10(-6) M) did not affect the AH intervals before hypoxia. The hypoxia-induced prolongation of the AH interval or AH block was prevented in the presence of these drugs. W-5, a chlorine-deficient derivative of W-7, showed no protection against hypoxia-induced AV nodal conduction disturbances. These findings suggest that hypoxia-induced AV nodal conduction disturbance is explained, at least in part, by the electrical uncoupling of nodal cells, probably due to the calcium overload. This conduction disturbance is protected by calcium entry blockers or by calmodulin inhibitors, but the mode of protective action is not the same for these different categories of drugs.

Electromechanical effects of bepridil on rabbit isolated hearts.

Electromechanical effects of a new antianginal agent, bepridil, on Langendorff-perfused rabbit hearts were compared with those of verapamil and lidocaine. Bepridil at concentrations above 2 X 10(-7)M caused a dose-related decrease in heart rate (HR), a prolongation of the atrio-His bundle conduction time (A-H interval) and a prolongation of the functional refractory period (FRP) of the atrioventricular (A-V) node. Similar changes in HR, A-H interval and the FRP of the A-V node were observed with verapamil at concentrations above 2 X 10(-8)M. Lidocaine at above 4 X 10(-5)M prolonged slightly the A-H interval and the FRP of the A-V node but did not decrease the HR. Bepridil at concentrations above 10(-6)M prolonged both the His bundle-ventricular conduction time (H-V interval) and the effective refractory period (ERP) of ventricular muscles. Similar changes in the H-V interval and in the ERP of ventricular muscles were observed with lidocaine at above 10(-5)M. Verapamil ranging between 5 X 10(-8)M and 8 X 10(-7)M had no effect on the H-V interval and appreciably shortened the ERP of ventricular muscles. Bepridil at concentrations above 2 X 10(-6)M reduced the developed tension as did verapamil at above 10(-7)M. On a molar basis, the depressant effect of bepridil on HR and the A-V nodal conduction, which may reflect inhibitory action on the slow channels, was less than one tenth that of verapamil. Bepridil was more potent than lidocaine in prolonging the H-V interval and the ERP of ventricular muscles indicating a possible inhibition of the fast sodium channels. Both of these electrophysiological effects of bepridil may afford significant bases for the antiarrhythmic action of the drug.

Effects of bepridil on the electrophysiological properties of guinea-pig ventricular muscles.

Effects of bepridil, a new antianginal and potential antiarrhythmic agent, on transmembrane action potentials of ventricular muscles were examined in isolated right ventricular papillary muscles of guinea-pig. Bepridil at concentrations above 5 X 10(-6)M caused a dose-dependent decrease in both the maximum upstroke velocity (Vmax) and the action potential duration from the upstroke to 30% repolarization ( APD30 ). On the other hand, the resting potential (RP), the amplitude of action potential (AMP), and the action potential duration from the upstroke to 90% repolarization ( APD90 ) were not affected even at the highest concentration applied (10(-5)M). The curves relating membrane potential and Vmax were shifted by bepridil at 5 X 10(-6)M along the voltage axis in the direction of more negative potentials. The recovery kinetics of Vmax assessed by premature stimuli were definitely slowed by bepridil at above 10(-6)M. This effect was more pronounced with higher [K+]o (10 mM) than normal [K+]o (5 mM). Bepridil at 5 X 10(-6)M caused a rate-dependent decrease of Vmax (use-dependent block) with rapid onset and offset, as did lidocaine. Slow responses, which had been induced by isoprenaline (5 X 10(-6)M) in K+-depolarized preparations, were suppressed significantly by additional application of bepridil at 10(-5)M. These findings suggest that bepridil has electrophysiological characteristics similar to those both of Class Ib and Class IV antiarrhythmic drugs.

[Molecular and biophysical aspects of cardiac Na channel].

Voltage-dependent ion channels mediate action potential and conduction in cardiac cells. In the past ten years, their single channel properties and protein components have been identified, isolated and restored to functional form in the purified state. Na channel has a primary alpha subunit, which is composed of four homologous domains, each of which has six transmembrane segments. Site-directed mutation has probed the function of specific regions or even of single amino acids. The ion-conducting pore, molecular voltmeter, ion filter and regulatory system of sodium channel have been characterized by coordination of biophysical, biochemical and molecular genetic methods.

Interactions of flecainide with guinea pig cardiac sodium channels. Importance of activation unblocking to the voltage dependence of recovery.

Effects of flecainide, a potent antiarrhythmic agent, on sodium channel availability was investigated in guinea pig single cardiac cells by the whole-cell voltage-clamp technique. Sodium current (INa) experiments were performed at 17 degrees C, and maximum upstroke velocity (Vmax) experiments were performed at 37 degrees C. Flecainide (3 microM) caused little tonic block, but reduced sodium channel availability in a use-dependent manner. The latter effect was accentuated by depolarization and attenuated by hyperpolarization. Long (200-msec) and short (10-msec) depolarizations yielded similar use-dependent block. These results indicate that flecainide has a low affinity for rested (R) and inactivated (I) channels but a high affinity for activated ones (A). In each of these states, the channels can bind to drug to form the corresponding RD, ID, and AD states. Recovery from flecainide block consisted of two components. An initial fast component was strongly voltage dependent: with increasing hyperpolarization, recovery developed more quickly and to a larger extent. At 17 degrees C, the mean time constant shortened from 132 +/- 81.6 msec at -120 mV to 46.9 +/- 34.1 msec at +/- 160 mV (kinetics were too fast for accurate measurement at 37 degrees C). A later slow component was largely voltage independent: at 37 degrees C, the mean time constant was 9.8 +/- 3.2 seconds at -100 mV and 10.7 +/- 3.8 seconds at -75 mV. The slow component of recovery was similarly independent of voltage at 17 degrees C. In terms of the modulated-receptor theory, our results indicate that the fast recovery depends on availability for unblocking (RD) but occurs during activation (AD----A). Indeed, when the RD state is maximized by strong hyperpolarization, activation unblock was also maximized. However, during depolarization to -100 mV, availability for activation unblock declined with a time constant of 98 +/- 12 msec (RD----ID). Therefore, the voltage-dependent fast unblocking is mostly due to priming of the RD state (RD----ID), and the voltage-independent slow unblock reflects dissociation of flecainide from closed states (RD----R and ID----I). We conclude that flecainide interacts with sodium channels preferentially in the activated state, whereas unblocking occurs via two separate pathways: activated and closed states. Furthermore, drug association with channels shifts the voltage dependence of closed-state transitions (RD in equilibrium ID) and their kinetics toward more negative potentials.

Electrophysiological effects of nipradilol (K-351) on isolated rabbit hearts and guinea pig ventricular muscles.

Effects of nipradilol (K-351), a newly developed antihypertensive agent, on electrophysiological properties of the heart were examined in Langendorff-perfused rabbit hearts and in superfused guinea pig ventricular muscles. In rabbit hearts, nipradilol at concentrations greater than 10(-5) M caused a significant prolongation of atrio-His bundle conduction time (AH) as well as His bundle-ventricular conduction time (HV), whereas heart rate (HR), functional refractory period of atrioventricular node (FRPVM) were unaffected. The intensity of the negative dromotropic effect of nipradilol was less than half that of propranolol. Isoproterenol-induced increase in HR and a shortening of AH, FRPAV, ERPVM were inhibited competitively by pretreatment with nipradilol at greater than 10(-8) M. Nipradilol was about two to three times more potent than propranolol in antagonizing isoproterenol-induced effects. In guinea pig ventricular muscles, nipradilol at a concentration greater than 10(-5) M caused a decrease of the maximum upstroke velocity (VMAX) of the action potential without affecting resting potential and action potential duration. In the presence of nipradilol, trains of stimuli at rates higher than 0.2 Hz led to an exponential decline in VMAX to achieve a new plateau. The time constant for the recovery from use-dependent block was 4.3 s. These findings suggest that nipradilol may have a potent cardiac beta-adrenoceptor blocking action and, at extremely high concentrations, quinidine-like inhibitory effects on the cardiac excitability and conductivity.

Inhibitory effects of nitroglycerin on contractile responses to methoxamine and clonidine in isolated rabbit renal and femoral arteries.

The inhibitory effect of nitroglycerin (NG) on contractile responses to methoxamine (MO) and clonidine (CL) was investigated in isolated rabbit renal and femoral arteries. NG (10(-7)-10(-5) mol/l) inhibited the responses to MO and CL in a noncompetitive manner and its inhibitory effect on CL-responses was much greater than that on the MO-responses. In the presence of phenoxybenzamine, however, NG (10(-5) mol/l) markedly inhibited the residual response to MO. Contractile responses to KCl or added Ca2+ in a Ca2+-free medium containing KCl were slightly inhibited by NG. In the presence of nifedipine, NG (10(-5) mol/l) markedly inhibited residual responses to CL but it only slightly inhibited the response to MO. In a Ca2+-free medium with EGTA, MO (10(-5) mol/l) or CL (10(-5) mol/l) induced a phasic contraction. NG had a greater inhibitory effect on the response to CL than MO. In a Ca2+-free medium with EGTA, nifedipine and MO (10(-5) mol/l) or CL (10(-5) mol/l), Ca2+ induced a tonic contraction. NG inhibited the Ca2+-response in the presence of CL, but it had little or no effect on the Ca2+-response in the presence of MO. These results suggest that in rabbit renal and femoral arteries, the potent inhibitory effect of NG on the responses to CL as compared to MO may be due to differences in the amount of receptor reserves that exist for both the agonists. In addition, the inhibition of voltage-dependent Ca2+ channels may not play a major role in the vasoinhibitory action of NG. Further, NG inhibits contractile responses due to mobilization of intracellular Ca2+ much more than the responses due to receptor-activated, nifedipine-insensitive Ca2+-movement.

Temperature and voltage dependence of sodium channel blocking and unblocking by O-demethyl encainide in isolated guinea pig myocytes.

O-demethyl encainide (ODE) is a metabolite of encainide with potent antiarrhythmic effects. We studied block of sodium channels by ODE in isolated guinea pig ventricular myocytes, using both direct measurement of whole cell sodium current and Vmax. Specifically, we examined whether block by ODE was use-dependent, whether ODE blocked activated or inactivated channels, and whether ODE's effect depended on membrane potential. At cold temperatures (15-20 degrees C), ODE was a potent blocker of activated cardiac sodium channels, with little or no effect on inactivated channels. At these temperatures, there was no detectable diastolic recovery from block. Repeated depolarizations produced use-dependent unblocking, which increased as the holding potential was made more negative. At warmer temperatures, use-dependent unblocking was increased, the availability curve of sodium channels for use-dependent unblocking was less shifted toward negative potentials, and diastolic recovery from block became detectable, albeit slow (tau approximately 25 s). Our findings demonstrate that electrophysiologic effects of agents such as ODE may be both quantitatively and qualitatively different at cold temperatures than at normal body temperature. One should be cautious about extrapolating electrophysiologic data obtained at cold temperatures to clinical situations.

Long-term study of recurrent vasospastic angina using coronary angiograms during ergonovine provocation tests.

Chronologic changes of coronary spasm were examined by repeated ergonovine provocation tests during angiography. A total of 322 patients who had variant angina without severe atherosclerosis demonstrated a positive response to the first test. Ninety of these patients had recurrent variant anginal symptoms after an angina-free period of 38 +/- 12 months (mean +/- SD). Of these 90 patients, 76 (84%) had symptoms or electrocardiographic (ECG) findings similar to those of the first test. The initial 9 of these 76 patients underwent a second provocation test and showed coronary responses analogous to those on the first test. Of the 90 patients, 14 (16%) had different symptoms or ECG findings from those elicited at the first episode. All 14 patients again had a positive response to a second ergonovine test and the following angiographic changes were observed in the three major vessels between the two tests. Of the 21 vessels that had spasm on the first test, eight vessels (19%) did not have spasm on the second test. Of the 21 vessels that did not demonstrate spasm on the first test, 10 (24%) demonstrated spasm on the second test. In the present study it is concluded that the majority of patients with recurrent angina seemed to have consistency in the location of coronary spasm, while in some patients the fluctuation of coronary spasm was confirmed by two ergonovine provocation tests.

Stage-dependent changes in membrane currents in rats with monocrotaline-induced right ventricular hypertrophy.

Sequential changes in action potential configuration, 4-amino-pyridine-sensitive transient outward current (Ito), and L-type calcium current (ICa) in association with hypertrophy were investigated in ventricular myocytes from rats with monocrotaline (MCT)-induced pulmonary hypertension. The tissue weight ratio of right ventricle (RV) to left ventricle plus septum 14 and 28 days after a subcutaneous injection of MCT increased by 29.7 and 77.2%, respectively. Action potential duration (APD) of RV cells from MCT rats increased progressively, prolonged by 73.2 and 92.2% on days 14 and 28, respectively. The current density of Ito in RV cells from MCT rats on day 14 (32.5 +/- 4.5 pA/pF, n = 13) was significantly larger than in controls (26.8 +/- 4.5 pA/pF, n = 8; P < 0.05). On day 28, however, Ito density in MCT rats (15.3 +/- 4.6 pA/pF, n = 9) was significantly less than in controls (27.3 +/- 4.2 pA/pF, n = 10; P < 0.05). There were no differences in the voltage dependence of steady-state activation and inactivation of Ito between MCT and control rats. ICa density in MCT rats on day 14 (15.7 +/- 2.6 pA/pF, n = 10) was significantly larger than in controls (10.0 +/- 2.3 pA/pF, n = 10; P < 0.05), but there was no significant difference in Ito density between MCT rats (8.3 +/- 3.7 pA/pF, n = 10) and controls (11.6 +/- 3.0 pA/pF, n = 10) on day 28. These findings suggest that hypertrophy of mammalian hearts may cause stage-dependent changes in Ito and ICa density of ventricular myocytes. The APD prolongation in the early stage of hypertrophy may be caused mainly by an increase in ICa density, whereas the APD prolongation in the late stage may be ascribed to a reduction in Ito density.

Inhomogeneity of cellular activation time and Vmax in normal myocardial tissue under electrical field stimulation.

To clarify the mechanism underlying the ectopic excitation after countershock, the cellular activation processes of cardiac tissue with a low-potential-gradient electric field (LPEF) were investigated in experiments using guinea pig papillary muscles and in computer simulation. Action potential upstrokes in papillary muscles during longitudinal propagation (LP) or transverse propagation (TP) were different from those of nonpropagating ones in single ventricular cells in terms of lower maximum upstroke velocity (Vmax) (LP, 231 V/s; TP, 309 V/s) and the presence of a linear ascending segment in the phase-plane plot. High Vmax (409 V/s) close to the single cell (512 V/s) was obtained in the muscle at the collision of LP (LC). Field stimulation of the muscles with LPEF < 5 V/cm caused inhomogeneous excitation suggesting multiple wave fronts, which collide with each other, and a wide spatial dispersion of Vmax (132-388 V/s). Phase-plane plots of action potential with lower Vmax were similar to LP or TP, whereas those with higher Vmax were similar to LC. In the two-dimensional discrete sheet composed of 51 x 51 elements of modified Beeler-Reuter model, the inhomogeneous excitation induced by LPEF is mimicked by setting a random variation of stimulus onset in each element. LPEF may induce inhomogeneous excitation with multiple wave fronts through a complex electrotonic interaction. This would provide a basis for the genesis of ectopic focal excitation.

Modulation of pacemaker activity of sinoatrial node cells by electrical load imposed by an atrial cell model.

To investigate the electrotonic modulation of sinoatrial (SA) node pacemaker activity by atrial muscle, single or multiple (2-7) SA node cells isolated from rabbit hearts were connected to a membrane model [resistance-capacitance (R-C) circuit] of an atrial cell through an external circuit that mimics the gap junctional conductance (Gc) between cells. When Gc was 0 nS (uncoupled conditions), all the preparations generated regular and stable spontaneous action potentials with a mean cycle length (SCL) of 263 +/- 45 ms (+/- SD, n = 35). Step increases of Gc were associated with a progressive prolongation of SCL. At sufficiently high values of Gc, the spontaneous activity became irregular and finally stopped. We defined the threshold Gc causing an appreciable SCL irregularity as the minimum Gc at which the ratio of SD to mean of SCL was > 0.3. The threshold Gc for a single SA node cell was calculated to be 0.58 nS. In the presence of acetylcholine (ACh; 0.05-0.2 microM), the coupling-induced inhibition of spontaneous activity was greatly increased, and the threshold Gc for a single SA node cell was decreased in a concentration-dependent manner. These findings show that the pacemaker activity of SA node cells is easily inhibited when the cells are coupled to a passive atrial cell model and the inhibition is amplified by ACh. Computer simulation using a modified Oxsoft HEART model indicates that the passive atrial cell model acts as a current sink, imposing a substantial outward current on the SA node cell, and ACh amplifies the effect by activating an additional outward current.