J Mel Cell Cardiol

Discordant Permeability

Ronald

23, 861-872 (1991)

Efkcts ofNisoldipine Changes during

G. Tilton, John Salvatore

on Coronaty Vascular Resistance and Reflow after Ischemia in Isolated Rabbit Hearts

A. Watts3, P. Sutera2,

Marishawn P. Laud, Kenneth and Joseph R. Williamson

B. Larson’,

Department of Patholou and’Biomedical Computer Laboratory, Washington Universi~ School of Medicine, and 2Department of Mechanical Engineering, Washington University St. Louis, MO, UsA. (Received 15 November 1990, accepted in revisedform 26 Februay 1991) RONALD G. TILTON, JOHN A. WATIS, MARISHAWN P. LAND, KENNETH B. LARSON, SALVATORE P. SUTERA AND JOSEPH R. WILLIAMSON. Discordant Effects of Nisoldipine on Coronary Vascular Resistance and Permeability Changes during Reflow after Ischemia in Isolated Rabbit Hearts. Journal ofhfofccufarand Cellular Cardiology (1991) 23, 861-872. Effects of a low dose (5 nM) of nisoldipine on vascular and ventricular function were assessed in isolated rabbit hearts during 2 h of reperfusion after 40 min of global, zero-flow &hernia. External detection of bolus injections of lz51-BSA and pressure data generated during the experiment provided repeated estimates of vascular volume, and fractional rate of albumin permeation and vascular hemodynamics (resistance, for 3.5 h, vascular resistance, intravascular washout of rz51-BSA (k,,,)). In control hearts perfused continuously vascular volume, LVEDP, and ke, remained constant, while maximum + dP/dl and - dP/df increased 25 % above baseline values, and estimates of albumin permeation increased 1.7 x baseline. Addition of 5nM nisoldipine to the perfusate after the baseline period produced sustained decreases in vascular resistance (16 % vs mean baseline value) without significantly affecting any other parameter. Postischemic perfusion of hearts increased vascular resistance and vascular volume -50% above baseline, decreased kc1 by 25% (intravascular washout of ‘251-BSA was prolonged), and increased albumin permeation -5 x baseline. While LVEDP remained elevated 3 x baseline, maximum + dP/df and -dP/dl recovered 100% of baseline values (75-80% of untreated control values at comparable time points). Addition of 5nM nisoldipine to the perfusate prior to ischemia prevented the increased vascular resistance during reflow, prevented the decrease in ke, and the increase in vascular volume, but did not affect the increased albumin permeation and, in general, did not affect the rate of recovery of left ventricle function. These results indicate that a low dose of nisoldipine preserves postischemic coronary vascular hemodynamics, but has little or no effect on the increased vascular leakage of albumin. KEY WORDS:

Endothelium;

External

detection;

Nisoldipine;

No-flow

ischemia;

Permeability;

Reperfusion.

Introduction A great deal of research has been focused on the use of calcium antagonists to salvage ischemic myocardium and to reduce the extent of myocardial necrosis following impaired coronary artery perfusion [I-3]. The beneficial effects of these drugs have been attributed to their ability to reduce myocardial oxygen demand by reducing contractile force, left ventricle afterload, and heart rate, as well as, their capacity to improve oxygen supply by vasodilating coronary arteries and collateral vessels. Several investigators have reported that PM concentrations of calcium antagonists can prevent increases in vascular permeability induced by ischemia and reperfusion [4-c],

but significant depression of left ventricle function also occurs at this dose. Recently, it has been reported that low concentrations of calcium antagonists, which do not depress cardiac function prior to ischemia, can reduce ischemic damage to the heart during reperfusion, suggesting that mechanisms independent of contractile function may provide the protection [ 7,8]. The high vascular selectivity of calcium antagonist dihydropyridines [ g-111, such as nisoldipine, allows investigation of effects of ischemia and reperfusion on coronary vascular hemodynamics and endothelial cell permeability to macromolecules not confounded by dose-related cardiodepressant

Please address all correspondence and reprint requests to: Ronald G. Tilton, Department of Pathology Washington University School of Medicine, 660 South Euclid Avenue, St. Louis MO 63110, USA 0022.2828/91/070861

+ 12$03.00/O

@ 1991 Academic

- Box 8118 Press Limited

862

R. G. Tilton

effects of these drugs on myocardial contractile function. Therefore, we have used a low concentration of nisoldipine (5 nM) to determine if this calcium antagonist can protect the coronary vasculature from global, zero-flow &hernia and reperfusion in the absence of significant pre-ischemic depression of myocyte contractile function. Our results indicate that this dose of nisoldipine prevents changes in coronary vascular hemodynamics induced by ischemia and reperfusion, but does not prevent the loss of endothelial cell barrier functional integrity.

Methods PerfUSion techniques Isovolumically beating, isolated hearts from male Dutch Belted rabbits (weighing 2-3 kg and heparinized intravenously with 1000 units sodium heparin 10min prior to killing) were perfused in a retrograde manner via the aorta with no recirculation at a paced rate of 180-200 beats/min with Krebs-Henseleit (KH) buffer containing 1% dialyzed bovine serum albumin (BSA), 0.5 % dialyzed polyvinylpyrrolidone (MW = 360,000), 3 mM pyruvate, 8 mM dextrose, and 100 mu/L insulin (pork; regular Iletin II; Lilly, Indianapolis, IN, USA). The salt composition (mM) of the KH buffer was as follows: NaCl, 120, KCl, 2.6; CaC12.2H20, 2.5; KHsPO,, 1.2; MgSO,.7HsO, 1.2; NaHCOs, 25. KH was filtered through 0.45~pm Gelman minicapsule filters, warmed to 37%, and oxygenated at pH 7.4 by dialysis against 95 % 0,/5% COs across medicalgrade silastic tubing (0.058-in i.d. x 0.077-in. o.d.; Dow Corning, Midland, MI, USA). Perfusate pH, pCO2, and ~02 were monitored with a blood gas analyzer (Instrumentation Laboratory, Inc.); arterial was PO2 maintained constant at-600 mmHg. Perfusion pressure and LV end-diastolic pressure (LVEDP) and developed pressure (obtained from a left ventricle balloon) were recorded continuously with P23Db Statham pressure transducers and a Gould recorder; maximum + dPldt was obtained by differentiation of the left ventricle pressure signal. Hearts were perfused apex up in an Oak-Tree Instruments (St. Louis, MO, USA) apparatus and were rinsed over the surface and through the right ventricle and atrium to facilitate rapid removal

of tracer that had passed through vasculature.

Iodination

the coronary

of BSA

Monomer bovine serum albumin (1 mg) was iodinated with 1 mCi lz51 for 30min at room temperature using the iodo-gen method (Pierce Chemical Co., Rockford, IL, USA), separated from free lz51 by use of column centrifugation and dialyzed against KH buffer at 4OC. Prior to use each day, column centrifugation was used to .remove any residual lz51.

Experimental

protocol

Perfusion pressure was adjusted to 60mmHg during the first hour of each experiment by changing the pump flow rate; this baseline flow rate was then used during subsequent perfusion in each heart to achieve constant flow perfusion conditions. Non-ischemic control hearts were perfused continuously for 3.5 h; ischemic hearts were perfused for a 1 h baseline period prior to 40 min of global, zero-flow ischemia, then reperfused at the same baseline flow rate for 2 h. A low dose (5 nM) of nisoldipine was added to the KH buffer and perfused through the coronary vasculature at the end of the baseline period in controls and 1Omin prior to ischemia and during the entire reperfusion interval in hearts subjected to ischemia. To assess effects of a higher dose of nisoldipine on lz51-BSA permeation, 50nM nisoldipine was added to the perfusate of selected hearts subjected to ischemia. Each day, the nisoldipine was dissolved in polyethylene glycol (MW = 400) to prepare a 1 mM stock solution, then diluted with 0.9 % NaCl and added to the perfusate to achieve a final concentration of 5 or 50nM. The nisoldipine was used under sodium vapor lighting with no exposure to fluorescent or natural lighting in order to minimize photodegradation. During zero-flow ischemia, heart temperature was maintained at 37% with heat lamps and surface rinses. Every 30min during the perfusion, a 25-~1 bolus of lz51-BSA in KH buffer was injected into the proximal aorta near the coronary ostia with a 50-~1 Hamilton syringe, and the single passage of this tracer through the coronary vasculature was monitored with a lead-

Coronary

Vascular

Responses

shielded and collimated Nal scintillation detector positioned to view the entire heart. The scintillation-detector output was amplified, processed by a programmable Ortec counting module and a pulse-height analyzer with an energy-acceptance window adjusted symmetrically around the 35-keV photopeak of lZ51, corrected for background, and then stored (on a Hewlett-Packard 7914 hard disk) for subsequent analysis with a HewlettPackard iOOOA-series computer. Sampling intervals were 200ms in the initial portion of each recording when count rates were changing rapidly.

Mathematical

model for assessment of tracer transport

The uptake and clearance of a bolus injection albumin in these heart of lz51-labeled preparations were interpreted with a compartmental model describing temporal changes of radiolabeled-tracer distributions due to transport of label within the vasculature and between the vascular and extravascular spaces. Tracer movements are interpreted in terms of diffusive and convective transport mechanisms operating simultaneously within a hypothetical membrane separating albuminaccessible vascular and extravascular spaces; no specific morphological identity is assigned to this membrane. Components of the model, complete mathematical derivations, and conventional requirements for radiotracer stimulus-response methodology have been described previously [L?]. The model parameters and their numerical estimates from the experimental data represent global averages. With the use of a parameter estimation technique based on maximum-likelihood estimation procedures for Poisson-distributed data [ 13- 151, the biexponential function r(t) =A,e-a+

+A2e-azt

(1)

is fitted to the experimental residue count-rate data. Here, t represents time elapsed after the instant the maximum count rate is observed following injection of the radiotracer bolus. The symbols Ai, A,, cri, and (~2 represent adjustable parameters and are related to the compartmental turnover rate constants of the

model,

863

to Nisoldipine

k~, , ki2, and k,,, according

k12

= ~1~2/kll

hl

= (a,

to

(3)

and + %)

- (hl

+ h2)

(4)

here roe r(t =O), i.e., the peak count rate, whose instant of occurrence defines zero time. Subscripts 1 and 2 denote, respectively, space space and extravascular vascular accessible to albumin; zero denotes the surroundings external to both compartments 1 and 2. An interpretation of the compartmental turnover rate constants that relates them to the perfusate flow, to the compartmental albumin and to the practical membranespace, transport coefficients [ 16- 181 is kol = F/V, (ss’)

(5)

&I* = [y,PS + l/2(1 - Oxf”]/Y~ (s- ‘)

(6)

k12= [y*PS - l/2(1 - a)J\,]lv,(S-‘)

(7)

and

In the above, F is the measured specific perfusate flow, P is the surface-area-averaged permeability coefficient, S is the total membrane surface available for permeation,J, is the net systemic fluid-filtration volume flow, u is the filtration-flow-averaged solvent-drag reflection coefficient [ 181, I’, is the vascularcompartment volume, I’, is the extravascular albumin-compartment volume, and yi is the thermodynamic activity coefficient of albumin in compartment i, for i = 1,2. The volume flow, J,,, is driven by an average pressure drop, due to combined hydraulic and osmotic effects, from compartment 1 to compartment 2. To allow comparison of results between preparations of different sizes, values of the global parameter estimates are computed and reported relative to the dry weight of each heart; this normalization is indicated below by tildes. Simultaneous solution of Eqs. 6 and 7 and division by the heart dry weight yields PS = 1/2(F2i + Fr2)/;G. (cm3/s/g)

(8)

R. G. Tilton

864

for the specific (i.e. weight-normalized) permeability coefficient x surface area product and v = l/2( 1 - C7)j” = l/2(&, (cm3/s/g)

Between runs Means and standard deviations, as well as standard errors of the mean (for data presented graphically), were calculated for each parameter assessed for control and ischemic groups at each perfusion time interval. Overall baseline (mean of 2 recordings obtained during the first hour of perfusion) differences among experimental groups for each parameter were assessed by the Van Der Waerden test [20]; if this test indicated that differences among groups were not statistically significant at P