Endothelium-dependent relaxation in pulmonary arteries after lung preservation and transplantation

Endothelium-Dependent Relaxation in Pulmonary Arteries After Lung Preservation and Transplantation Per Ola Kimblad, MD, Giorgio Massa, MD, Trygve Sjoberg, PhD, and Stig Steen, MD, PhD Department of Cardiothoracic Surgery, University Hospital, Lund, Sweden

Pulmonary hypertension is frequently seen after lung transplantation. To study how the release of the endothelium-dependent relaxing factor is affected by lung preservation and transplantation, porcine pulmonary arteries were investigated in organ baths. The arteries (1 mm in diameter) were taken from fresh nonperfused lungs (group I), lungs immediately after flush-perfusion with a low-potassium-dextran solution (group II), nonperfused lungs stored for 12 hours in low-potassiumdextran solution (group III), flush-perfused lungs stored for 12 hours in low-potassium-dextran solution (group IV), and group IV lungs after left lung transplantation and right pneumonectomy followed by 24 hours of reperfusion (group V). Stable contractions were induced with the thromboxane A, analogue U-46619. Acetylcholine was used to stimulate the release of endotheliumdependent relaxing factor. In vessel segments where the

P

ulmonary hypertension is frequently seen after lung transplantation. As the endothelium-dependent relaxing factor (EDRF) has been shown to play a key role in the modulation of pulmonary vascular tone [l, 21, it is of interest to study how EDRF is affected by lung preservation and transplantation.

Material and Methods Twenty-four Swedish native breed pigs with a mean weight of 40 kg (range 38 to 43 kg) were used (16 for donor procedure, 5 recipients, 3 for sham operation). All the animals received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" published by the National Institutes of Health (NIH publication no. 85-23, revised 1985). Anesthesia was induced with intramuscular administration of ketamine hydrochloride (Ketalar; Parke-Davis, Morris Plains, NJ), 30 mg/kg, and atropine sulfate (Atropin; Kabi Pharmacia AB, Uppsala, Sweden), 1 mg. Sodium thiopental (Pentothal; Abbot Laboratories, North Chicago, IL), 5 to 8 mg/kg, was given intravenously before a tracheostomy (tube No. 7) was done. Anesthesia and muscular relaxAccepted for publication Jan 22, 1993 Address reprint requests to Dr Steen, Department of Cardiothoracic Surgery, University Hospital of Lund, S-221 85 Lund, Sweden.

0 1993 by The Society of Thoracic Surgeons

endothelium had been removed, acetylcholine elicited no response. In segments with intact endothelium, acetylcholine induced concentration-dependent relaxation; the maximum relaxation obtained was 91% 2 3% (I), 86% 2 3% (10, 85% & 3% (1111, 69% & 5% (IV), and 69% 2 9% (V). Relaxation was significantly reduced in groups IV ( p < 0.01) and V ( p < 0.05) as compared with group I. Stable moderate pulmonary hypertension was present in all the transplanted lungs throughout the 24-hour observation period. It is concluded that the endotheliummediated relaxation is significantly reduced after flush perfusion combined with 12 hours of storage in lowpotassium-dextran solution. Lung transplantation, followed by 24 hours of reperfusion did not further impair the endothelium-dependent relaxation.

(Ann Thorac Surg 1993;56:1329-34)

ation were maintained with a continuous infusion of 40 mL/h of a mixture of 3 g of ketamine, 200 mg of pancuronium bromide (Pavulon; Organon Teknika, Boxtel, the Netherlands), and 10 mg of midazolam hydrochloride (Dormicum; Roche, Basel, Switzerland) dissolved in 500 mL of 10% glucose. The animals received ventilatory support with a Siemens-Elema Servoventilator 900B. Air enriched with oxygen to an inspired oxygen fraction of 0.4, and fixed-volume ventilation of 8.5 L/min, 20 breaths/ min, and a positive end-expiratory pressure of 5 cm H,O was used. All donors received heparin (500 U/kg). In six of the donor operations, where the lungs would not be used for transplantation, a segment of the right lung was taken and the pulmonary artery dissected for immediate contractility studies (group I), and for studies after 12-hour storage in cold (4"to 6°C) Perfadex (Kabi Pharmacia AB) (group 111). Perfadex is a commercially available lowpotassium-dextran solution (LPD). One thousand milliliters of Perfadex contains the following: dextran 40 (Kabi Pharmacia AB), 50 g; Na+, 138 mmol; K', 6 mmol; Mg', 0.8 mmol; C1-, 142 mmol; SO,'-, 0.8 mmol; H,PO,- + HPO,*-, 0.8 mmol; glucose, 0.91 g; and hydrochloric acid and sodium hydroxide to pH 6 with sterile water in a sufficient quantity. A pulmonary artery flush perfusion was performed in all donor lungs using 6 L of cold (4" to 0003-4975/93/$6.00

1330

Ann Thorac Surg 1993;561329-34

KIMBLAD ET AL ENDOTHELIUM AFTER TRANSPLANTATION

6°C) Perfadex (buffered to pH 7.40 with THAM) with a physiologic perfusion pressure (10 to 20 mm Hg), while the lungs were fully ventilated. After perfusion the lungs were excised in a collapsed state. The left lung was immersed in cold (4" to 6°C) LPD for 12 hours. One segment of the right lung was prepared for immediate contractility studies (group 11). The rest of the right lung was immersed in 4" to 6°C LPD for 12 hours (group IV). The recipient operation was performed 12 hours later in 5 animals. First a single-lung transplantation was done on the left side. Immediately thereafter, pneumonectomy of the contralateral right lung was done with the help of venoarterial extracorporeal membrane oxygenation (Carmeda AB, Stockholm, Sweden) [3], which was removed within 1 hour after the completion of the pneumonectomy. The pig was kept anesthetized and with ventilatory support for 24 hours, completely dependent for its survival on the transplanted lung. The animal was then killed. The transplant was removed and immediately used for contractility studies (group V). In 3 separate animals a sham operation was done, consisting of bilateral thoracotomy and right pneumonectomy. These animals were observed for 24 hours, under conditions equivalent to those of the transplanted animals.

Preparation and Mounting A small branch of an intralobar pulmonary artery (approximately 1 mm external diameter) was dissected free from surrounding tissue under an operation microscope and divided into 1.5-mm-long segments. Each of the ring segments was suspended between two parallel L-shaped metal holders, of which one was attached to a Grass FT03C force displacement transducer, which in turn was connected to a Grass polygraph to record the isometric tension. The other holder was attached to a movable unit allowing fine adjustments of vessel tension. The vessel segments were immersed in temperature-controlled (37°C) organ baths, containing 5 mL of Krebs' solution, continuously bubbled with a mixture of 95% 0, and 5% CO,, giving a pH of approximately 7.40. One thousand milliliters of Krebs' solution contains the following: NaCl, 119 mmol; NaHCO,, 15 mmol; KC1, 4.6 mmol; NaH,PO,, 1.2 mmol; MgCl,, 1.2 mmol; CaCl,, 1.5 mmol; and glucose, 11 mmol. In one vessel segment from each pig, the endothelium was removed by gently rubbing the intimal surface over a pair of microforceps before mounting in the organ baths. This technique has previously been shown to remove the endothelium without damaging the arterial smooth muscle [4, 51. The vessels were repeatedly stretched to a tension of 4 mN during an equilibration period of approximately 2 hours. Previous experiments have shown that an optimal contractile force is obtained around this tension in pig pulmonary arterial ring segments of similar size. The distance between the metal holders at basal tension, as measured by means of a microscope equipped with a scaler, was 1.91 0.36 mm (mean standard deviation, z = number of segments = 123), giving a diameter of

*

*

approximately 1.2 mm, and the length of the vessel segments was 1.79 0.40 mm (mean standard deviation, z = 123).

*

*

Experimental Procedure Contraction was induced in all vessel segments with the thromboxane A, analogue U-46619 (3 x lop7 mol/L) (Upjohn, Kalamazoo, MI). This contraction was allowed to stabilize until a steady state was reached. Acetylcholine (Sigma, St Louis, MO), sodium nitroprusside (Nipride; Roche), or papaverine sulfate (Kabi Pharmacia AB) was then added cumulatively (lop9 to mol/L). In separate experiments, indomethacin (3 x lop7 mol/L) (Dumex, Copenhagen, Denmark) was added to the bath before vasocontraction with U-46619 to exclude the interaction of prostanoids. Acetylcholine, indomethacin, nitroprusside, and papaverine were diluted in 0.9% NaCl with 1.0 mmol/L ascorbic acid, whereas U-46619 was diluted in phosphate buffer at neutral pH just before use. Nitroprusside was kept protected from light. One segment from each pig served as a control, where contraction was induced by U-46619, but no relaxing drug was given.

Analysis of Data The maximum contraction and the steady-state level of the response to U-46619 was determined. The responses to each concentration of the relaxing drugs were expressed in percent of the tension at steady state. All results are given as mean standard error of the mean, and n refers to the number of different pigs used. Only groups I and I11 allowed for paired statistical tests, and Wilcoxon's test for paired data was used for comparisons between these two groups. In all other comparisons, Wilcoxon's test for unpaired data was used. A probability level less than 0.05 was regarded as significant.

*

Results The five pigs receiving transplants, which were followed up throughout a 24-hour observation period, were all in good condition, with arterial oxygen tensions around 200 mm Hg (inspired oxygen fraction = 0.4), and arterial CO, tensions around 40 mm Hg. The mean pulmonary arterial pressure was stable around 30 mm Hg (range 20 to 35 mm Hg) in the pigs receiving transplants, as compared with 22 mm Hg (range 18 to 26 mm Hg) in the shamoperated pigs (Fig 1). The thromboxane A, mimic U-46619 induced strong contractions that remained stable over hours and thus allowed assessment of the induced relaxation. Acetylcholine induced a concentration-dependent relaxation in all the vessels with intact endothelium (Figs 2 and 3). The maximum relaxation was obtained with or mol/L acetylcholine, whereas higher concentrations (lop4 and lo-, mol/L) induced diminished relaxation. No relaxation was elicited with acetylcholine in vessels where the endothelium was removed (Fig 2C). Addition of indomethacin to the bath did not influence the effects of acetylcholine.

Ann Thorac Surg 1993;561329-34

60

KIMBLAD ET AL ENDOTHELIUM AFTER TRANSPLANTATION

Transplanted

-

%

g

.A

ld

o

-

%

20 10 -

0

30

0 -

g

Sham

60 -

20

: 40

.A

4

ld

50 -

4

$

40 -

60

30 -

80

20 -

100

10

-

-8

-9

0 l

l

l

.

l

l

l

l

l

l

l

l

l

l

l

l

l

l

l

I

I

.

.

I

Fig 1. Pulmona y arterial pressure (PAP) during the first 24 hours in the transplanted animals (upper panel) and in the sham-operated animals where only right pneumonectomy was done (lower panel). Filled circles show mean pulmona y arterial pressure, open circles show systolic and diastolic pressures.

The maximum endothelium-dependent relaxation elicited by acetylcholine was significantly reduced in groups IV ( p < 0.01) and V ( p < 0.05) compared with group I, whereas the endothelium-independent relaxation induced by nitroprusside and papaverine remained unaltered (Table 1 and Fig 4). Control

T

5 mN

1

*

B

With endothelium

C

No endothelium

1

25 mN

1

ACh -5 -4

-3

ACh

*

-9-8

-7

-6

! ! i !

-5

I

7

-4

I

-3 7

I

5 mN

t

U-46619 3 ,10-7mol/L

t (

5 min

-4

-3

There is evidence that the endothelium modulates the tone of the pulmonary vascular bed through the release of

- 9- 8- 7- 6

-5

Comment

I

1

-6

Fig 3. Effects of cumulative addition of acetylcholine (ACh) to pulmonuy artey precontracted with thromboxane analogue U-46619 in fresh nonperfused lungs (group I , n = 6), lungs immediately after flush-perfusion with low-potassiumtfextran solution (group 11, n = 1 9 , nonperfused lungs stored for 12 hours in low-potassium-dextran solution (group 111, n = 6)‘ flush-perfused lungs stored for 12 hours in low-potassium-dextran solution (group IV, n = 12), and group IV lungs after transplantation and 24 hours of reperfusion (group V , n = 5). Upper panel shows the controls where no relaxing drug was given.

*

k-7 I

-7

Log ACh conc (mol/L)

I

4 0 12 16 20 24 Time a f t e r reperfusion (hours)

0

A

3

4

x

50 40 -

r

1331

I

lomN

Fig 2. Tracings from concentrationresponse experiments in three segments of pulmonary artey. In (0 the endothelium was removed. After contraction with U-46619 in all segments, acetylcholine was added cumulatively in (B)and (C) whereas ( A ) served as a control. All drug concentrations are shown as the logarithm of moles per liter. (*Sensitivity diminished by a factor of 2. “Sensitivity diminished by a factor of 5.)

1332

KIMBLAD ET AL ENDOTHELIUM AFTER TRANSPLANTATION

Ann Thorac Surg 1993;561329-M

Table 1. Maximum Relaxation in Pulmonaru Artenf Acetylcholine Percent P

Group I I1

91 2 3 86 f 3 85 f 3 69 f 5 69 9

111

IV V

Nitroprusside Percent P

n 6

NS

15

NS

6 12 5

b

*

c

88 2 4 90 f 3 86 2 4 87 f 3

Papaverine n

Percent

6

101 2 2

NS

7

NS

6 7

102 f 1 98 f 4

NS

98 f 2

P

n

NS NS

6 7 6 7

NS

Maximal relaxation in pulmonary artery from fresh nonperfused lungs (group I), lungs immediately after flush-perfusion with low-potassium-dextran solution (group 11), nonperfused lungs stored for 12 hours in low-potassium-dextran solution (group 111), flush-perfused lungs stored for 12 hours in low-potassium-dextran solution (group IV), and group IV lungs after transplantation and 24 hours of reperfusion (group V). The maximum relaxation is p < 0.01, p < 0.05. All comparisons made expressed in percent of the tension at steady state and is given as mean 2 standard error of the mean. with group I.

a

NS = not significant.

an EDRF [l, 21. Nitric oxide has been identified as such a factor [6]. Nitric oxide is thought to stimulate guanylate cyclase, thus increasing the intracellular levels of cyclic guanosine monophosphate in smooth muscle cells, %

0 20

.? 4 (d

X

2

2

-

Nitroprusside

-

40 60-

80 -

100

c

4

%

0 20

40 (d

d

'ai

P: 60 80

100

1 d I:, A IV

1

-9

I

-8

I

I

I

I

-7 -6 -5 - 4 Log conc (moI/L)

I

-3

I

-2

Fig 4. Effects of cumulative addition of nitroprusside and papaverine to pulmonary artery precontracted with thromboxane analogue U-46619in fresh nonperfused lungs (group I , n = 6), lungs immediately after push-perfusion with low-potassium-dextran solution (group 11, n = 7), nonperfused lungs stored for 12 hours in low-potassiumdextran solution (group Ill, n = 6), and push-perfused lungs stored for 12 hours in low-potassium-dextran solution (group IV, n = 7).

thereby inducing relaxation [7-91. Acetylcholine has previously been shown to produce potent endotheliumdependent relaxation in peripheral as well as in pulmonary arteries in humans; the receptor is muscarinic and the response is mediated by EDRF [lo, 111. The release of EDRF is dependent on oxygen, and the pulmonary artery contraction induced by hypoxia has been shown to be caused largely by decreased EDRF activity in rat and dog [12, 131. Ischemia followed by reperfusion has also been shown to impair the endothehum-dependent relaxation to acetylcholine in the coronary artery in dogs [4]. Endothelium-dependent relaxing factor does not seem to be responsible for the low vascular tone of the normoxic human and rat lung, whereas it might modulate the hypoxic pressor response and act as a physiologic brake against pulmonary vasocontraction [5, 141. Pulmonary hypertension was seen in all the pigs receiving transplants. The mean pulmonary arterial pressure was 30 mm Hg as compared with 22 mm Hg in the sham-operated animals. This implies that some factor in the process of preservation or transplantation caused the pulmonary hypertension. A prerequisite for the study of vascular relaxation is a vessel preparation in a stable contractile condition. The thromboxane A,-mimic U-46619 has previously been shown to induce stronger contractions than noradrenaline and prostaglandin F,, in human pulmonary arteries [15], and was therefore chosen to elicit the contraction in this investigation. The endothelium-dependent relaxation, elicited by acetylcholine, appeared slightly reduced immediately after flush-perfusion (group 11), as well as in the arteries not subjected to flush-perfusion but subjected to storage for 12 hours (group 111). However, these differences were not statistically significant (Table 1). Struber and colleagues [16] have also previously reported that flush-perfusion alone of canine lungs with Euro-Collins solution does not seem to affect the endothelium-dependent relaxation caused by acetylcholine in the pulmonary arteries. Flush-perfusion and 12 hours of storage in LPD (group IV) caused a significantly reduced ( p < 0.01) endotheliumdependent relaxation that was unaltered 24 hours after reperfusion (group V). The relaxation induced by nitro-

Ann Thorac Surg 1993;56132934

prusside and papaverine through endothelium-independent mechanisms [17], however, was not affected (Table 1 and Fig 4). This suggests that the EDRF-producing capacity of the endothelium is reduced, and not the capability of the smooth muscle to relax. The 24-hour reperfusion period does not appear to have further influenced the capacity of the endothelium to release EDRF. Ultrastructural studies by Mills and colleagues [lS] on transplanted canine lungs have indicated that the pulmonary vasculature sustains no additional structural injuries after reperfusion. and lop3 mom) of acetylHigher concentrations choline induced less relaxation compared with concentrations of or lop7 moYL (Figs 2 and 3). In separate experiments, 3 x lop7 moVL of indomethacin was added to the bath before vasocontraction with U-46619. This did not affect the relaxation induced by the cumulative addition of acetylcholine. This indicates that the diminished relaxation elicited by lop4 and lop3 m o m acetylcholine was not induced by a stimulated release of vasocontracting prostanoids. The vessel segments we used were approximately 1mm in diameter. In coronary arteries, reperfusion has been shown to impair the endothelium-dependent relaxation in microvessels but not in larger arteries (1 to 3 mm) [19]. The pulmonary artery segments used in our study are located peripherally in the vascular tree, whereas coronary arteries with a diameter of 1 to 3 mm have a central location. In conclusion, the endothelium-dependent relaxing capacity in pulmonary arteries was significantly reduced after flush-perfusion and 12 hours of storage in LPD. This was unaltered 24 hours after lung transplantation. It is thus possible that impaired endothelial function, after flush-perfusion and ischemic storage, may be one explanation for the pulmonary hypertension seen after lung transplantation. This study was supported by a grant from Svenska Nationalforeningen mot Hjart- och Lungsjukdomar, Crafoordska Stiftelsen, T. Westerstroms Stiftelse, and the Medical Faculty, University of Lund.

References 1. Cherry PD, Gillis CN. Evidence for the role of endotheliumderived relaxing factor in acetylcholine-induced vasodilatation in the intact lung. J Pharmacol Exp Ther 1987;241:516-20. 2. Yamaguchi T, Rodman DM, OBrien RF, McMurtry IF. Potentiation of pulmonary vasoconstriction by inhibitors of

KIMBLAD ET AL ENDOTHELIUM AFTER TRANSPLANTATION

1333

endothelium derived relaxing factor. Am Rev Respir Dis 1987;135:A131. 3. Steen S, Sjoberg T, Massa G, Ericsson L, Lindberg L. Safe pulmonary preservation for 12 hours with low-potassiumdextran solution. Ann Thorac Surg 1993;55:43440. 4. VanBenthuysen KM, McMurtry IF, Horwitz LD. Reperfusion after acute coronary occlusion in dogs impairs endotheliumdependent relaxation to acetylcholine and augments contractile reactivity in vitro. J Clin Invest 1987;79:265-74. 5. Hasunuma K, Yamaguchi T, Rodman DM, OBrien RF, McMurtry IF. Effects of inhibitors of EDRF and EDHF on vasoreactivity of perfused rat lungs. Am J Physiol 1991;260: L97-104. 6. Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 1987;327524-6. 7. Holzmann S. Endothelium-induced relaxation by acetylcholine associated with larger rises in cyclic GMP in coronary arterial strips. J Cyclic Nucl Res 1982;8:409-19. 8. Rapoport RM, Murad F. Agonist-induced endotheliumdependent relaxation in rat thoracic aorta may be mediated through cGMP. Circ Res 1983;52:352-7. 9. Kukovetz WR, Holzmann S. Cyclic GMP in endotheliumdependent relaxation of coronary smooth muscle by acetylcholine. In: Bevan JA, et al, eds. Vascular neuroeffector mechanisms. Amsterdam: Elsevier Science Publishers, 1985: 115-21. 10. Luscher TF, Cooke JP, Houston DS, Neves RJ, Vanhoutte PM. Endothelium-dependent relaxations in human arteries. Mayo Clin Proc 1987;62:601-6. 11. Liischer TF. Endothelial vasoactive substances and cardiovascular disease. Basel: Karger, 1988365. 12. De Mey JG, Vanhoutte PM. Anoxia and endotheliumdependent reactivity of the canine femoral artery. J Physiol 1983;335:65-74. 13. Rodman DM, Yamaguchi T, Hasunuma K, O'Brien RF, McMurtry IF. Effects of hypoxia on endothelium-dependent relaxation of rat pulmonary artery. Am J Physiol 1990;258: L207-14. 14. Crawley DE, Liu SF, Evans TW, Barnes PJ. Inhibitory role of endothelium-derived relaxing factor in rat and human pulmonary arteries. Br J Pharmacol 1990;101:166-70. 15. Sjoberg T, Steen S. The strong contractile effect of the thromboxane receptor agonist U-46619 in isolated human pulmonary arteries and its competitive antagonism by BM13.505. Acta Physiol Scand 1989;136:161-5. 16. Struber M, McGregor CGA, Locke TJ, Miller VM. Effect of flush-perfusion with Euro-Collins solution on pulmonary arterial function. Transplant Proc 1990;22:2206-11. 17. Pearson PJ, Evora PRB, Schaff HV. Bioassay of EDRF from internal mammary arteries: implications for early and late bypass graft patency. Ann Thorac Surg 1992;54:107&84. 18. Mills AN, Hooper TL, Hall SM, McGregor CGA, Haworth SG. Unilateral lung transplantation: ultrastructural studies of ischemia-reperfusion injury and repair in the canine pulmonary vasculature. J Heart Lung Transplant 1992;11:5M7. 19. Quillen JE, Sellke FW, Brooks LA, Harrison DG. Ischemiareperfusion impairs endothelium-dependent relaxation of coronary microvessels but does not affect large arteries. Circulation 1990:82:586-94.

INVITED COMMENTARY The pulmonary vascular endothelium plays an important role in the determination of pulmonary vascular smooth muscle tone. Through the release of EDRF, the

endothelium not only influences basal smooth muscle tone, but also modulates the response to vasoconstricting agents. If the endothelium is damaged, the vasoconstrict-