Effect of benzylamine and its metabolites on the responses of the isolated perfused mesenteric arterial bed of the rat

J. Auton. Pharmacol. (1991) 11, 323-335

Effect of benzylamine and its metabolites on the responses of the isolated perfused mesenteric arterial bed of the rat Jonathan Elliott" & Brian A. Callingham Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 lQJ UK

1 Semicarbazide-sensitive amine oxidase (SSAO) is an enzyme activity which can be found in the plasma membrane of rat vascular smooth muscle cells. We have investigated the possibility that the products of deamination by this enzyme, namely ammonia, hydrogen peroxide and the aldehyde, may be important in the modulation of the responses of vascular smooth muscle to extracellular stimuli.

2 The isolated perfused mesenteric arterial bed of the rat was used and dose- pressure response curves (DRC) to bolus injections of adrenaline (Ad) or ATP were plotted by non-linear curve fitting. The relaxant effects of carbachol (CCh), which releases endothelium dependent relaxing factor (ERDF), were studied by co-administering CCh with Ad. The effects of including the preferred SSAO substrate, benzylamine ( B Z 25 p ~ ) , in the perfusion fluid throughout the experiment and of inhibition of SSAO by treatment of rats with (E)-2-(3',4'- dimethoxyphenyl)-3-fluoroallylamine (MDL 72 145; 1 mg kg-') 1 h before dissection, have been studied.

3 Neither BZ nor SSAO inhibition affected the DRC to ATP. BZ shifted Ad responses to the left, inhibition of SSAO increased this shift indicating that the amine, but not its metabolites, were responsible for the potentiation of the responses to Ad. DRC to CCh showed a shift to the left and a significant decrease in the Hill slope with BZ, indicative of a potentiation of low doses of CCh more than high doses. Inhibition of SSAO prevented this change and so the metabolites of BZ deamination appeared to be involved in the potentiation.

4 Ammonia generated by SSAO may contribute to the production of EDRF or hydrogen peroxide may sensitize guanylate cyclase to stimulation by EDRF and so explain these findings.

*Present address: Department of Veterinary Basic Sciences, Royal Veterinary College, Royal College Street, London NW1 OTU, UK. Correspondence: B. A. Callingham.

324 J. ELLIOTT & 6. A . CALLINGHAM

Introduction

Semicarbazide-sensitive amine oxidase (E.C. 1.4.3.6; SSAO) is highly active in vascular tissue of the rat (Lewinsohn, Bohm, Glover & Sandler, 1978), where it resides in the plasma membrane of vascular smooth muscle cells (Wibo, Duong & Godfraind, 1980; Lyles & Singh, 1985). This property is one which distinguishes SSAO from the mitochondria1 enzyme, monoamine oxidase (E.C. 1.4.3.4; MAO) (see Callingham & Barrand, 1987 for review).

The preferred substrate for SSAO is the synthetic amine, benzylamine (BZ), but the identity of the natural substrate remains to be determined. In homogenates of rat blood vessels, SSAO has a high affinity for, and metabolizes, primary aliphatic and aromatic monoamines (Elliott, Callingham & Shar- man, 1989a) many of which are also substrates for MAO. Inhibition of SSAO has been shown to potentiate the contractile effect of tryptamine on rings of rat aorta (Lyles & Taneja, 1987; Taneja & Lyles, 1988) and of the pressor effect of an infusion of tyramine in the isolated perfused mesenteric arterial bed of the rat (Elliott, Callingham & Sharman, 1989b). In both cases, this effect was seen only when MAO-A had also been inhibited, indicating that both enzyme activities are involved in scavenging monoamines in the blood vessels of this species.

In addition to the inactivation of circulat- ing amines, the cellular location of SSAO has prompted suggestions that the products of deamination, an aldehyde, ammonia and hydrogen peroxide, might be involved in modulating transduction processes of exter- nal stimuli, which regulate vascular tone by acting at the smooth muscle cell membrane (see Callingham, 1987). There is evidence from other cell systems that enzymatically generated hydrogen peroxide can modulate transduction processes. In fat cells, hydrogen peroxide can mimic the effects of insulin on glucose transport and metabolism (Czech, Lawrence & Lynn, 1974a), effects which appear to be due to oxidation of sulphydryl groups (Czech, Lawrence & Lynn, 1974b). Insulin can stimulate the activity of a membrane bound enzyme, nicotinamide adenine dinucleotide phosphate (NADP) oxidase,

which produces peroxide (Mukherjee & Lynn, 1977). Although many of the actions of insulin are now thought to be mediated by the tyrosine protein kinase activity of the insulin receptor (see Czech, 1983, peroxide can stimulate the activity of the insulin receptor kinase (Yu, Khalaf & Czech, 1987) and may be a modulator of the insulin response rather than being the second mes- senger.

The concept that peroxide modulates membrane events by oxidation of sulphydryl groups is applicable to the vascular smooth muscle cell. P-adrenoceptors and the enzyme they activate, adenylate cyclase, have been shown to contain an essential disulphide and sulphydryl group respectively and peroxide can affect binding of P-adrenoceptor agonists and the activity of adenylate cyclase in skeletal and ventricular muscle (Wright & Drummond, 1983; Haenen, Van Dansik, Vermeulen, Timmerman & Bast, 1988). The reversible oxidation of sulphydryl groups is thought to be important in stimulating guanylate cyclase activity (both soluble and particulate forms), although excessive oxidation can result in loss of activity (see Wald- man & Murad, 1987). Thus, the potential sites at which peroxide generated by SSAO could modulate transmembrane signalling events within vascular smooth muscle cells are many and this line of investigation seems justified. Before detailed investigations are carried out at the cellular level, it is necessary to establish that alterations in SSAO activity modify the responses of perfused blood vessels to vasoconstrictors and dilators.

The isolated perfused mesenteric arterial bed of the rat has been used to examine this hypothesis. Previous studies have shown that addition of SSAO substrates, such as benzyl- arnine (BZ), to fluid perfusing this vascular bed, results in rapid release of deaminated metabolites (Elliott, Callingham & Sharman, 1989~). In addition, the potent SSAO inhibitor, (E)-2-(3',4'-dimethoxyphenyl)-3-fluoroallylamine (MDL 72 145) used in an ex vivo manner prevents metabolism of BZ in this vascular bed. The responses of these resis- tance vessels to pressor and relaxant agents have been examined in the presence and absence of BZ in the perfusing fluid. MDL 72 145 has been used to inhibit the formation

BENZYLAMINE IN MESENTERIC ARTERY 325

of deaminated metabolites from BZ. These results have been presented in abstract form (Elliott & Callingham, 1990).

Methods

Animals Male Wistar rats (200-350 g) supplied by A. J. Tuck & Son, Rayleigh, Essex, were used. SSAO was inhibited in some rats by i.p. injection of 1 mg kg-' MDL 72145 (dis- solved in distilled water at 1 mg ml-I) 1 h before use. This dose achieved 95% inhibition of SSAO activity measured in homogenates of mesenteric blood vessels made after perfusion experiments and had no effect on the MAO-A activity present in these vessels (Elliott et al., 1989~). Administration of distilled water alone had no effect on the enzyme activities.

The isolated perfused mesenteric preparation Heparin (500 units, i.p.) was administered to the rats 10 min before they were anaesthe- tized with pentobarbitone sodium (60 mg kg-l, i.p.). The mesenteric arterial bed was then isolated and perfused as described by McGregor (1 965). Briefly, the cranial mesenteric artery was cannulated with a polyethyl- ene catheter (1.02 mm o.d., Portex Ltd, Hythe, Kent) and the ileocolonic arterial branches were ligated. Perfusion with modified Krebs-Henseleit solution maintained at 37" and gassed with 95% 0, and 5% C 0 2 was started as soon as possible after the blood flow had ceased. A constant flow rate of 2 ml min-l was delivered by a Harvard 1203A peristaltic pump. Perfusion pressure was recorded by a Washington PT400 pressure transducer connected via an amplifier to a Kipp & Zonen BD41 two-channel X/T pen recorder. Two preparations were tested at the same time. In some preparations, 25 p~ benzylamine hydrochloride was included in the perfusion fluid and was administered as an infusion throughout the experiment.

Following the 45-min equilibration period, dose-response curves to adrenaline (Ad), ATP or carbachol were constructed. Doses of Ad (0.3 to 20 nmoles) or ATP (0.06-14 pmoles) were administered (in 200 pl of

Krebs solution) through an injection port into the perfusing fluid. (Delay between injection and the beginning of the response was less than 5 s.) The maximum increase in pressure was recorded. The method chosen to study responses to carbachol was that adopted by Hiley, Phoon & Thomas (1987) where they co-administered a standard dose of pressor agent together with various doses of relaxant agent into the perfusion fluid. The reduction in the response to the pressor agent caused by the presence of the relaxant agent could then be measured. In the present experiments, doses of Ad that gave a sub- maximal response of about 80 mm Hg were chosen as the pressor agent against which the relaxant responses to carbachol were measured. The dose of Ad required to produce this response varied from one preparation to the next and a number of test doses of Ad were administered at the start of the experiment to determine the most appropriate one to use. After every two doses of Ad with carbachol, Ad was administered alone and the relaxation produced by each dose of carbachol was expressed as the percentage reduction in the maximum increase in pressure caused by Ad alone (see Fig. 1). Similar experiments were performed with noradrenaline as the pressor agent, but the dose (5 nmoles) was constant for all preparations.

One dose-response curve to one agonist was obtained in any one preparation. The composite dose-response curves for each agonist were measured in six to eight preparations from each group. Four groups of preparations were examined for each agonist. These were preparations from control rats (Control), preparations from rats treated with MDL 72 145 (MDL), preparations from control rats where 25 ~ U M BZ was included in the Krebs solution (Control+BZ) and those from MDL 72145 treated rats where BZ was added to the Krebs solution (MDL+BZ). The rats used were weight- and age-matched between each group.

Statistical analysis of the results The dose-response curves were fitted to the observations by non-linear regression analysis using a modified Marquardt procedure contained in the Harwell Library routine

326 J. ELLIOTT & B. A . CALLINGHAM

- 6 min

-A- 0.0 0.e

80

mm

00

- 6 mln

I i 0 . 0

0.08 0.2 0.0 2.0 8.0 14.0

Dosa of ATP (rmolar)

Dora 01 carbachol fnmoled

c I 0 . 3 '

Fig. 1. Examples of chart records from the isolated perfused mesenteric arterial bed of the rat showing responses to adrenaline, ATP and carbachol. (a) Doses of adrenaline (Ad) administered into the perfusing fluid at 5-min intervals at the points on the trace indicated by the solid circles. (b) Doses of ATP administered as described above. This regime was repeated after 20 min perfusion free of drug administration to give duplicate responses at each of the doses tested. (c) Doses of carbachol have been co-administered with Ad (3 nmoles) in a total volume of 200 pl at the points indicated by solid circles. At the points marked 'C', Ad (3 nmoles) has been given alone. The response to carbachol was expressed as the percentage reduction of the Ad responses either side of the carbachol response. In each of the experiments shown, the preparation is from a previously untreated rat and 25 p~ benzylamine has been included in the Krebs solution.

BENZYLAMINE I N MESENTERIC ARTERY 327

VBO 1 A, which calculated the best fit values of the Hill slope, theEC,,and the maximumpres- sure response. The equation employed was:

Response = Resp,,,. Dl(D + ECSOn)

where D is the dose of agonist added, n is the Hill coefficient, ECSo is the dose giving the half maximal response and Resp,,, is the maximum response. Each point was weighted according to the reciprocal of its variance (Aceves, Mariscal, Morison & Young, 1985). Since this is a non-linear fitting procedure employing dose of agonist (not log. dose), the mean EC5, values are provided with their estimated standard errors. The curves were also compared using the method of De Lean, Munson & Rodbard (1978), fitting two curves simultaneously with none of the parameters constrained and then constraining each in turn (NAG library routine E04 FDF). The Ptest was applied to the sum of the squares of residuals in each situation and a change in the parameter was taken to be significant when Pt0.05.

Drugs and solutions The modified Krebs-Henseleit solution had the following composition (mM): NaCl 11 8, KCl 4.57, CaCl, 1.27, KH2P04 1.19, MgSO, 1.19, NaHCO, 25 and glucose 5.55. Adrenal- ine bitartrate, adenosine 5’-triphosphate (ATP, sodium salt) were obtained from Sigma Chemical Co., Poole, Dorset, UK. Carbamoyl choline (carbachol) was purchased from Koch-Light, Haverhill, Suffolk, UK. MDL 72145 was a gift from Merrell Dow Research Institute, Strasbourg, France. Heparin injection BP was obtained from Paines & Byrne Ltd, Greenford, Middlesex, UK and pentobarbitone sodium (Sagatal) was obtained from May & Baker Ltd (now Rh6ne-Poulenc-Rorer), Eccles, Manchester, UK. Benzylamine was purchased from Sigma as the free base and the hydrochloride pre- pared and crystallized. All other reagents were of analytical grade where possible.

Results The baseline perfusion pressure for each preparation reached a stable level within 10 min of the start of perfusion with Krebs

solution. MDL 72145 treatment of the rats or inclusion of 2 5 p ~ BZ in the Krebs solution did not affect the baseline pressure, which did not usually fluctuate by more than 1 mm Hg throughout the experiment. The mean values for the baseline perfusion pressures for each group of preparations taken just before the first dose of agonist was added are shown in Table 1.

Table 1. The effect of inclusion of benzylamine (25 p ~ ) in the Krebs solution and treatment of rats with MDL 72 145, on the baseline perfusion pressure in the isolated perfused mesenteric arterial bed of the rat.

Baseline perfusion pressure Group (mean k SEM; mm Hg) n

Control 9.09 k 0.38 21 MDL 9.17 -t 0.42 21 Control + BZ 9.24 t 0.37 20 MDLfBZ 9.22 k 0.53 20

The baseline perfusion pressure reached a steady level within 10 min of the start of perfusion when it was recorded and did not vary by more than 1 mm Hg throughout the experiment. The four groups of preparations examined were from control rats (Control), from rats treated with 1 mg kg-1 MDL 72145 1 h before use (MDL), from control rats with the inclusion of benzylamine (25 p ~ ) in the Krebs solution throughout the experiment (Control+BZ), and from rats treated with MDL 72145 with the inclusion of BZ in the Krebs solution (MDL+BZ).

Responses to adrenaline Examples of the responses of this preparation to Ad are shown in Fig. la. At low doses of Ad there was a dose-dependent rise in perfusion pressure, which reached its maximum very quickly and returned to the baseline within 1 min. At higher doses, once the maximum response had been reached, the pressure returned to the baseline more slowly. No attempt was made to find the dose which gave the maximum area under the response. The dose-response curves to Ad resulting from these experiments are shown in Fig. 2 and the parameters derived from these computer-generated curves are given in Table 2.

328

20

J . ELLIOTT & B . A . C A L L I N G H A M

Responses to adrenaline:

11C.

1 o c -

90-

8C- - 0)

= 7c- E E - 6 0 - v) c : 50- a,

40-

100-

9 0.

80-

7 0- - W I

E E 60-

$ 5 0 -

v

0

2 40- a

3 c -

2 0-

1 c - :i 1

Effect of BZ on control preparations

T

I‘ /I I I )I

0.3 0.6 1 2 3 5 10 2‘ 0

Dose of adrenal ine (nmoles)

2 b Responses to adrenaline:

Effect of Ez on MDL 72145 treated preparations

’or

L * 0.3 0.6 1 2 3 5 10 20

Dose of adrenaline (nm0leS)

Fig. 2. Dose-response curves of the isolated perfused mesenteric arterial bed of the rat to adrenaline. Doses of adrenaline (0.3-20 nmoles) were injected into the perfusing fluid and the maximum rise in pressure has been plotted against the dose given (log scale). Each point represents the mean value k SEM from eight preparations. The lines of best fit have been drawn through the points using a weighted, non-linear regression computer program. The values obtained for Hill slope, ECS0 and maximum response for each group of preparations are given in Table 2. The curves are from (a) control rats (Control; solid squares) and from control rats with 25 ,UM benzylamine in the Krebs solution (Control-tBZ; open squares) and (b) from MDL 72145 treated rats (MDL; solid circles) and from MDL 72 145 treated rats with 25 ,UM BZ in the Krebs solution (MDL+ BZ; open circles).

MDL 72 145 treatment of the rats resulted in a small, but significant increase in the EC50 value and the maximum response (Control vs MDL). Inclusion of BZ in the Krebs solution perfusing the vascular bed caused a significant decrease in the EC50 value for both control preparations (1.3 fold; Control vs Control + BZ) and preparations from MDL 72145 treated rats (two fold; MDL vs MDL+BZ). Comparing the two curves measured in the presence of BZ, the MDL+BZ

group had a significantly lower EC50 when compared with the Control+BZ group. In- clusion of BZ in the Krebs solution also caused a significant increase in the maximum response (both Control vs Control+BZ and MDL vs MDL+BZ), the increase being of a greater magnitude for the Control +BZ group. The Hill slopes of the four curves showed very little variation between the groups with only one comparison showing a significant difference (MDL vs MDL+BZ).


Table 2. Responses of the isolated perfused mesenteric arterial bed of the rat to bolus doses of adrenaline.

EC50 Max. response Hill slope (nmoles) (mm Hg)

Control 1.32 k 0.08 1.53 k 0.08 89.3+ 1.59 MDL 1.38+0.03 1.96 2 0.05** 94.0k 0.65* Control + BZ 1.49 k 0.04 1.17 k 0.03** 106.0 k 1.32** MDL+BZ 1.62 + 0.07t 0.98 k 0.057?,# 99.2 k 1.327,#

The data given above are derived from the dose-response curves shown in Fig. 2 . The abbreviations for the groups of preparations used are the same as those given in Table 1 . A weighted non-linear regression computer program was used to determine the best fit for each curve as described in the methods section and the three parameters, Hill slope, EC5, and maximum response, together with their estimated standard errors, were generated (n= 8 for each group). A method for simultaneously fitting two curves (De Lean et al., 1978) with or without one of the parameters constrained was used to test for significant differences between the four curves. The levels of significance for these comparisons are:- *Pt0.05 and **PcO.Ol vs Control group; tPt0.05 and t tPt0.01 vs MDL group; #P<0.05 and ##P<O.Ol vs ControlfBZ and this convention has also been used in Tables 3 and 4.

Responses to ATP

An example of a chart record obtained from a preparation where ATP was used is given in Fig. lb. The computer-generated parameters for the dose-response curves to ATP are given in Table 3. Treatment of the rats with MDL 72145 did not alter any of these parameters (Control vs MDL) nor did the addition of BZ to the perfusing fluid (Control vs Control+BZ). The inclusion of BZ in the Krebs solution caused a significant decrease in the maximum response obtained in preparations from MDL 72145 treated animals (MDL vs MDL+BZ). In general, the Hill slopes from the control groups of preparations tended to be lower than the MDL groups but the difference was only significant when BZ was included in the Krebs solution (Control+BZ vs MDL+BZ).

Responses to carbachol Co-administration of carbachol with Ad caused a dose-dependent decrease in the maximum response attained when compared with the response to the same dose of Ad administered alone, as can be seen in the chart record shown in Fig. lc. The dose-

response curves to carbachol are shown in Fig. 3 and the computer-generated parameters from these curves are given in Table 4. The mean ‘pressor response’ referred to in the table is the ‘standard‘ response to Ad, against which the relaxations caused by carbachol were measured. This did not vary significantly between the four groups, indicating that the relaxant effects of carbachol have been compared against the same standard for each group.

MDL 72 145 treatment caused a significant decrease in the EC,, value for carbachol without altering the maximum response or the Hill slope (Control vs MDL). Addition of BZ to the Krebs solution caused a decrease in the Hill slope obtained from control preparations (Control vs Control + BZ), whereas MDL 72145 treatment prevented the addition of BZ causing any significant changes in the dose-response curve parameters (MDL vs MDL+BZ).

Essentially the same effect was obtained in control preparations when carbachol was co- administered with noradrenaline as the pressor agent. The Hill slope was reduced non-significantly and the EC5,, value was significantly reduced by benzylamine, which therefore potentiated the response to carbachol (Table 5).

330 J. ELLIOTT & 6 . A . CALLINGHAM

3a Responses to carbacol:

3b T ResDonses to carbachol: Effect of Bz on control preparations

1 2 3 5 10 20 50 1 ,

I Effect of Bz on MDL 72145 treated preparations

1 2 3 5 10 20 50 I.

Dose of carbachol (nmoles) Dose of carbachol (nrnoles)

Fig. 3. Dose-response curves of the isolated perfused mesenteric arterial bed of the rat to carbachol. Doses of carbachol (1-50 nmoles) were co-administered with a standard dose of adrenaline and the percentage inhibition of the response seen when the standard dose of Ad alone was administered has been plotted against the carbachol (log scale) dose (see Fig. Ic). Each point represents the mean value k SER from six or seven preparations. The lines of best fit have been drawn through the points using a weighted, non-linear regression computer programme. The values obtained for Hill slope, ECS0 and maximum response for each group of preparation are given in Table 4. The curves are (a) from control rats (Control; solid squares) and from control rats with 25 PM benzylamine (BZ) in the Krebs solution (ControlfBZ open squares) and (b) from MDL 72145 treated rats (MDL; solid circles), and from MDL 72145 treated rats with 25 PM BZ in the Krebs solution (MDLfBZ; open circles).

Discussion Inclusion of BZ in the Krebs solution perfusing the isolated mesenteric arterial bed of the rat ensured formation of deaminated metabolites at the smooth muscle cell membrane throughout the experiment in this in v i m preparation. We have shown that in this preparation, metabolism of BZ by SSAO continues for at least 150 min, the duration

of the present experiments (Elliott, 1989). Inhibition of the metabolism of BZ has proved possible by the use of the potent SSAO inhibitor, MDL 72145, which has been shown, when used in this ex vivo manner, to inhibit the release of BZ metabolites from the mesenteric preparation (Elliott et al., 1989~). From the present studies, it should be possible, therefore, to determine

BENZYLAMINE I N MESENTERIC ARTERY 331

Table 3. Responses of the isolated perfused mesenteric arterial bed of the rat to bolus doses of ATP

EGO Max. response Hill slope @moles) (mm Hg)

Control 0.95 t 0.04 1.23 f 0.14 MDL 1 .OO f 0.04 1.57f 0.16 Control+BZ 0.93 t 0.03 1.63 f 0.15 MDLSBZ 1.10 2 0.04# 1.42 2 0.12

81.6t2.26 89.6 2 2.34 84.4 t 2.27 79 .7 t 1.68t

The data given above are derived from the dose-response curves to ATP obtained by injecting 200 p1 of Krebs solution containing 0.06-14 pmoles into the perfusing fluid and recording the maximum increase in pressure for each dose. The abbreviations for the groups of preparations used are the same as those given in Table 1. A weighted non-linear regression computer program was used to determine the best fit for each curve as described in the methods section and the three parameters, Hill slope, EC50 and maximum response, together with their estimated standard errors, were generated (n= 6 for each group). Comparisons between the four curves were made as described by De Lean et al. (1978) and the level of significance of the differences found is indicated in the same way as in Table 2.

Table 4. Responses of the isolated perfused mesenteric arterial bed of the rat to carbachol co- administered as a bolus dose with adrenaline

Max. response EC50 (Yo inhibition of control

Hill slope (nmoles) response)

Control 1.77 k 0.16 5.73 2 0.40 49.8 f 1.74 (81.5)ll MDL 2.14 t 0.09 3.64 i 0.19** 47.22 0.42 (75.7) ControlSBZ 1.05 k 0.09** 3.58 t 0.40 55.4+ 2.34 (79.6) MDL+BZ 2.26 2 0.22# 4.1 I -t 0.22 47.5 f 1.36# (80.0)

The data given above are derived from the dose-response curves shown in Fig. 3. The abbreviations for the groups of preparations used are the same as those given in Table 1. A weighted non-linear regression computer program was used to determine the best fit for each curve as described in the methods section and the three parameters, Hill slope, EC5o and maximum response, together with their estimated standard errors, were generated (n=6 or 7 for each group). Comparisons between the four curves were made as described by De Lean et al. (1 978) and the level of significance of the differences found is indicated in the same way as in Table 2. lThe figure given in parentheses in the maximum response column is the mean pressor response (mm Hg) against which the relaxation to carbachol was measured.

Table 5. Responses of the isolated perfused mesenteric arterial bed of the rat to carbachol co- administered as a bolus dose with noradrenaline.

Max. response EC50 (% inhibition of

Hill slope (nmoles) control response

Control (NA+carbachol) 1.87 t 0.17 5.0820.31 67.6k 1.51 Control + benzylamine 1.17 f 0.22 2.69f 0.36 67.8 f 3.69 F-value 5.036 N.S. 7.87 P<0.05 0.004 N.S.

n=6 in all cases.


whether the observed changes in the DRCs to Ad, ATP and CCh were due to BZ itself or its deaminated products. All four groups of preparations had the same baseline perfusion pressure, indicating that neither of the ex- perimental procedures (treatment of the rats with MDL 72145 or inclusion of BZ in the perfusion fluid) influenced the resting tone in this preparation.

Overall, the evidence from the present studies does not support a major role for SSAO-generated metabolites in modulating the responses of this vascular bed to Ad. The leftward shift (decrease in ECS0) in the DRC caused by the inclusion of BZ in the Krebs solution (Control v.s Control+BZ) was greater when metabolism of BZ had been inhibited, indicating that the amine itself, rather than its deaminated metabolites, was responsible. Indeed, it appears that BZ metabolism by SSAO reduces the effective con- centration of BZ in the tissue and protects against this potentiating effect of the amine. The potentiation of the Ad response caused by BZ could be due to inhibition of neuronal uptake of Ad, a process which can be shown to be of importance in terminating the action of Ad in this vascular bed (Elliott et al., 1989b). Alternatively, competitive inhibition of the P-adrenergic response to Ad, which in this vascular bed opposes its a- adrenergic vasoconstrictor action (Borkow- ski & Porter, 1983; Nichols & Hiley, 1985) could result in the potentiation seen with BZ. There is some evidence for an az- adrenoceptor that is inhibitory in this vascular bed (Fiotakis & Pipili 1983) which may be situated on the endothelium and stimulate the release of endothelium-derived relaxing factor (EDRF) in large mesenteric blood vessels (White & Carrier, 1986). However, the influence of the endothelium on adrenergic vasoconstriction in the isolated perfused mesenteric arterial bed appears to be limited (Randall & Hiley, 1988; Randall, Kay & Hiley, 1988).

Addition of BZ to the Krebs solution also resulted in an increased maximum response to Ad but in this case the increase was less marked, although still significant, when metabolism of BZ had been inhibited. It is possible that at high levels of smooth muscle tone BZ metabolites have a small positive

effect on the processes which are important for generating and maintaining tone in vascular smooth muscle.

Treatment of rats with MDL 72145 gave rise to preparations which were less sensitive to Ad with a significantly higher ECS0 but an increased maximum response when compared with control preparations. This small but significant effect could be attributed to inhibition of SSAO if it were proposed that SSAO acted upon endogenously derived substrates which were supplied throughout the experiment and whose metabolism influenced the tissue’s response to Ad. Alterna- tively, MDL 72145 could have other effects on the tissue, which influenced its responses to Ad. These effects would have to be of an irreversible nature, since the tissue would be washed free of any reversibly bound drug early in the experiment.

In the isolated perfused mesenteric preparation, Ralevic & Burnstock (1988) recently demonstrated that exogenous ATP could produce both vasodilation and vasoconstriction. The vasodilation was endothelium- dependent and involved PZr receptors whilst the vasoconstriction involved Pzx receptors on the smooth muscle cells. These two subtypes of purinergic receptor were identified according to the criteria proposed by Burnstock & Kennedy (1 986). The vasoconstrictor responses of ATP measured in the present experiments involved the activation of Pzx receptors on the smooth muscle cells. Pzy receptors on the endothelium were prob- ably maximally activated at the doses used in these experiments (Ralevic & Burnstock, 1988) and so the dose-related constriction was measured against a constant stimulation of EDRF release. The results from the present studies indicate that the metabolic products generated by SSAO, either from endogenous substrates or from BZ added to the Krebs solution, do not influence the responses of this preparation to ATP, as neither inhibition of SSAO nor addition of BZ significantly altered the responses. If the same changes occurred in these experiments to ATP-induced EDRF release as occurred with carbachol-induced EDRF release (see below), changes in the dose-response curves to ATP might have been expected. The most likely explanation is that the influence of


EDRF release on the constrictor response to ATP was relatively small (see Ralevic & Burnstock, 1988) and so any changes in that response were below our limits of detection.

In the present experiments, co-administration of carbachol with Ad or noradrenaline caused a dose-dependent decrease in the response to Ad. Removal of the endothelium by the use of collagenase confirmed that this effect was dependent on the presence of the endothelium (data not shown). Inhibition of SSAO caused a small but significant parallel shift in the dose-response curve to carbachol with a decrease in the EC5,, but no change in the Hill slope or the maximum response (Control vs MDL). This effect could be attributable to inhibition of SSAO activity metabolizing some endogenous substrate or to an effect of MDL 72145 that has yet to be identified.

Inclusion of BZ in the Krebs solution caused a shift to the left ofthe dose-response curve to carbachol, with a significant decrease in the Hill slope (Control vs Control+BZ). BZ seemed to potentiate the responses at low doses of carbachol much more than at high doses. This effect was due tothemetabolismof thisamineasinhibitionofSSAOpreventedthe potentiating effect of BZ (Control+ BZ vs MDL+BZ). Indeed, there was no difference between the two dose-response curves in preparations where SSAO had been inacti- vated (MDL vsMDL+BZ),indicatingthat the amine itself did not influence the EDRF responses seen. Experiments using noradrenaline-induced tone also showed that inclusion of benzylamine in the perfusing fluid potentiated the action of carbachol.

Any interaction between the products generated by SSAO and the EDRF responses measured in these experiments could be due to an interaction at the level of EDRF production and release, a chemical interaction with EDRF, or an interaction at the level of the smooth muscle cell influencing the mechanisms involved in EDRF-induced relaxation. Strong evidence has accumulated that the relaxing factor released into the lumen of blood vessels by agonists which cause endothelium-dependent relaxations is nitric oxide, or a compound closely related to it which is formed from the terminal guan-

idino atoms of L-arginine (see Moncada, Palmer & Higgs, 1989). The exact reaction pathway which results in release of nitric oxide is not known. In the mouse macrophage, an L-arginine deiminase that liberates ammonia, which is then oxidized to NO2 and NO3, has been postulated (Hibbs, Taintor & Vavrin, 1987). An increase in the availability of ammonia for the synthesis of EDRF could explain the potentiation seen in these studies at low doses of carbachol in the presence of the metabolites of BZ.

Hydrogen peroxide has been shown to cause relaxation of bovine pulmonary arteries, an action which was resistant to indo- methacin but inhibited by methylene blue (Burke & Wolin, 1987). These authors also showed that hydrogen peroxide could stimulate guanylate cyclase activity, an action which depended upon its metabolism by catalase. It is possible that the hydrogen peroxide generated by the action of SSAO on BZ in the present experiments activated guanylate cyclase in a catalase-dependent manner and this potentiated the action of EDRF generated by low doses of carbachol. In conclusion, it has been possible to demon- strate some modulatory effects of the metabolites of BZ generated by SSAO at the smooth muscle cell membrane on the responses of this tissue to vasoactive agents, particularly EDRF. These results provide no information about the possible mechanisms involved at the cellular level but have provided support for the view that SSAO is not merely an amine-scavenging enzyme.

Acknowledgments J.E. was a Wellcome Veterinary Scholar.

References

ACEVES, J., MARISCAL, S., MORRISON, K.E. & YOUNG, J.M. (1985). The binding of doxepin to histamine HI-receptors in guinea-pig and rat brain. Br. J. Pharmacol., 84, 417-424.

BORKOWSKI, K.R. & PORTER, M. (1983). Modulation of vasoconstriction by P-adrenoceptors in isolated perfused rat mesenteric arteries. Br. J. Pharmacol., 80, 508P.

BURKE, T.M. & WOLIN, M.S. (1987). Hydro- gen peroxide elicits pulmonary arterial relax-


ation and guanylate cyclase activation. Am. J. Physiol., 252, H721-H732.

BURNSTOCK, G. & KENNEDY, C. (1986). Is there a basis for distinguishing two types of P,-purinoceptor? Gen. Pharmacol., 16,

CALLINGHAM, B.A. (1987). Some aspects of the enzymic inactivation of sympathomimetic amines. Blood Vessels, 24, 240-252.

CALLINGHAM, B.A. & BARRAND, M.A. (1987). Some properties of semicarbazide-sensitive amine oxidases. J. Neural Transm., [Suppl.]

CZECH, M.P. (1985). The nature and regulation of the insulin receptor: structure and function. Ann. Rev. Physiol., 47, 357-381.

CZECH, M.P., LAWRENCE, J.C. & LYNN, W.S. ( 1 974a). Hexose transport in isolated brown fat cells. A model system for investigating insulin action on membrane transport. J. Biol. Chem., 249, 5421-5427.

CZECH, M.P., LAWRENCE, J.C. & LYNN, W.S. (1974b). Evidence for the involvement of sulfhydryl oxidation in the regulation of fat cell hexose transport by insulin. Proc. Natl. Acad. Sci. USA, 71, 4173-4177.

DE LEAN, A., MUNSON, P.J. & RODBARD, D. ( 1 978). Simultaneous analysis of families of sigmoidal curves: application to bioassay, radioligand assay, and physiological dose- response curves. Am. J , Physiol., 235,

ELLIOTT, J. (1 989). Semicarbazide-sensitive amine oxidase of blood vessels: a functional study. PhD Thesis, University of Cam- bridge.

ELLIOTT, J. & CALLINGHAM, B.A. (1990). Effect of benzylamine on responses of the isolated perfused mesenteric arterial bed of the rat. J. Autonomic Pharmacol., 10, 18 (abst).

ELLIOTT, J., CALLINGHAM, B.A. & SHARMAN, D.F. (1 989a). Semicarbazide-sensitive amine oxidase of rat aorta: interactions with some naturally occurring amines and their structu- ral analogues. Biochem. Pharmacol., 39,

ELLIOTT, J., CALLINGHAM, B.A. & SHARMAN, D.F. (1989b). The influence of amine meta- bolising enzymes on the pharmacology of tyramine in the isolated perfused mesenteric arterial bed of the rat. Br. J. Pharmacol., 98,

433-440.

23, 37-54.

E97-E 102.

1507-1 5 15.

5 15-522.

ELLIOTT, J., CALLINGHAM, B.A. & SHARMAN, D.F. ( 1 989c). Metabolism of amines in the isolated perfused mesenteric arterial bed of the rat. Br. J. Pharmacol., 98, 507-514.

FIOTAKIS, D. & PIPILI, E. (1983). The effects of a,-adrenoceptor agonism and antagonism and of PG-inhibition of the responsiveness of the rat mesenteric arterial bed. Br. J. Pharmacol., 79, 234P.

HAENEN, G.R., VAN DANSIK, P., VERMEU- LEN, N.P., TIMMERMAN, H. & BAST, A. (1988). The effect of hydrogen peroxide on beta-adrenoceptor function in the heart. Free Radic. Res. Commun., 4, 243-249.

HIBBS, J.B. Jr, TAINTOR, R.R. & VAVRIN, Z. ( 1 987). Macrophage cytotoxicity: role for L- arginine deiminase and imino nitrogen oxidation to nitrite, Science, 235, 473-476.

HILEY, C.R., PHOON, C.K.L. & THOMAS, G.R. (1 987). Acetylcholine vasorelaxation in superior mesenteric arterial bed of the rat is endothelium-dependent and sensitive to antioxidants. Br, J. Pharmacol., 91, 378P.

LEWINSOHN, R., BOHM, K.-H., GLOVER, V. & SANDLER, M. (1978). A benzylamine oxidase distinct from monoamine oxidase B- widespread distribution in man and rat. Biochem. Pharmacol., 27, 1857-1 863.

LYLES, G.A. & SINGH, I . (1985). Vascular smooth muscle cells: a major source of the semicarbazide-sensitive amine oxidase of the rat aorta. J. Pharm. Pharmacol., 37,637-643.

LYLES, G.A. & TANEJA, D.T. (1987). Effects of amine oxidase inhibitors upon tryptamine metabolism and tryptamine-induced contractions of rat aorta. Br. J . Pharmacol., 90, 16P.

MCGREGOR, D.D. (1 965). The effect of sympa- thetic nerve stimulation on vasoconstrictor responses in perfused mesenteric blood vessels of the rat. J. Physiol., 177, 21-30.

MONCADA, S., PALMER, R.M.J. & HIGGS, E.A. (1 989). Biosynthesis of nitric oxide from L-arginine. A pathway for the regulation of cell function and communication. Bio- chem. Pharmacol., 38, 1709-1 7 15.

MUKHERJEE, S.P. & LYNN, W.S. (1977). Re- duced nicotinamide adenine dinucleotide phosphate oxidase in adipocyte plasma membrane and its activation by insulin. Possible role in the hormone’s effects on adenylate cyclase and the hexose monophosphate shunt. Arch. Biochem. Biophys., 184, 69-76.

NICHOLS, A.J. & HILEY, C.R. (1985). Identifi-


cation of adrenoceptors and dopamine receptors mediating vascular responses in the superior mesenteric arterial bed of the rat. J. Pharm. Pharmacol., 37, 1 10- 1 15.

RALEVIC, V. & BURNSTOCK, G. (1988). Actions mediated by P,-purinoceptorsubtypes in the isolated perfused mesenteric bed of the rat. Br. J. Pharmacol., 95, 637-645.

RANDALL, M.D. & HILEY, C.R. (1988). Deter- gent and methylene blue affect endothelium- dependent vasorelaxation and pressuretflow relations in rat blood perfused mesenteric arterial bed. Br. J. Pharmacol., 95, 108 1-1088.

RANDALL, M.D., KAY, A.P. & HILEY, C.R. (1 988). Endothelium-dependent modulation of the pressor activity of arginine vasopressin in the isolated superior mesenteric arterial bed of the rat. Br. J. Pharmacol., 95,

TANEJA, D.T. & LYLES, G.A. (1988). Further studies on the interactions between amine oxidase inhibitors and tryptamine-induced contractions of rat aorta. Br. J . Pharmacol., 93, 253P.

646-652.

WALDMAN, S.A. & MURAD, F. (1987). Cyclic GMP synthesis and function. Pharmacol. Rev., 39, 163-196.

WHITE, R.E. & CARRIER, G.O. (1986). al- and a*-adrenoceptor agonist-induced contraction in rat mesenteric artery upon removal of endothelium. Eur. J. Pharmacol., 122,

WIBO, M., DUONG, A.T. & GODFRAIND, T. (1980). Subcellular location of semicarbazide-sensitive amine oxidase in rat aorta. Eur. J. Biochem., 112, 87-94.

WRIGHT, M. & DRUMMOND, G.I. (1983). Inactivation of the P-adrenergic receptor in skeletal muscle by dithiols. Biochem. Phar- macol., 32, 509-5 15.

Yu, K.T., KHALAF, N. & CZECH, M.P. (1987). Insulin stimulates the tyrosine phosphoryla- tion of a Mr= 160,000 glycoprotein in rat adipocyte plasma membrances. J. Biol. Chem., 262, 7865-7873.

349-352.

(Received 6 August 1990 Revised 18 May 1991

Accepted 14 June 1991)

Documents

Effect of benzylamine and its metabolites on the responses of the isolated perfused mesenteric arterial bed of the rat