Contrasting Evects of Circulating Nitric Oxide and Nitrergic Transmission on Exocrine Pancreatic Secretion in Rats

Home , Indazole

684 Gut 1998;43:684–691 Contrasting eVects of circulating nitric oxide and nitrergic transmission on exocrine pancreatic Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from secretion in rats

E Vaquero, X Molero, V Puig-Diví, J-R Malagelada

Abstract Inhibition of endogenous nitric oxide (NO) Background—Nitric oxide (NO) blockade synthesis by L-nitroarginine methyl ester (L- by L-nitroarginine methyl ester (L-NAME) NAME) decreases basal and stimulated pan- inhibits pancreatic secretion in vivo and creatic secretion in vivo. This process has been aggravates caerulein induced pancreati- shown in several species including the rat, pig, tis. Nitric oxide synthase (NOS) is present cat, and dog,1–4 but its mechanism of action is in pancreatic islets, endothelium, and poorly understood. nerve fibres. L-NAME blocks all known In the normal pancreas, nitric oxide synthase NOS isoforms. (NOS) is present in the endothelium, nerve Aim—To investigate the source of NO fibres, and islets of Langerhans.5–7 Immuno- blocked by L-NAME that inhibits amylase staining techniques do not show the presence secretion. of NOS in exocrine cells,7 whereas functional Methods—Amylase output was measured studies indicate that NOS activity is present.18 in rats in response to caerulein (0.1–50 In fact, NOS may be involved in the control of µg/kg) alone or with indazole. Baseline agonist stimulated calcium influx.8 However, in secretion and the response to supramaxi- dispersed pancreatic acini, blockade of NO mal caerulein were also examined after production by L-NAME has not proven to be administration of indazole, L-NAME, detrimental to amylase secretion,1 an observa- haemoglobin, or aminoguanidine under tion that is consistent with the lack of eVect of continuous blood pressure measurement. intracellular modifications of cGMP concen- In separate experiments, pancreatic se- tration on amylase release.9 cretion was measured after blockade of L-NAME (and most other structurally re- aVerent nerve fibres by either systemic or lated compounds) exerts an inhibitory eVect local capsaicin. The eVect of neural NOS over all NOS isoforms, being eVective on a inhibition on caerulein induced pancrea- wide variety of cell types. Thus L-NAME elic- http://gut.bmj.com/ titis was also investigated. its a powerful and sustained vasopressor Results—L-NAME, haemoglobin, and su- response, increasing mean arterial pressure pramaximal caerulein (10 µg/kg) in- (MAP)10 (as the result of endothelial NOS creased blood pressure, whereas indazole blockade), inducing anti-nociception by inhib- and suboptimal caerulein (0.1 µg/kg) did iting neural NOS (nNOS) preventing NO not. Indazole and capsaicin decreased mediated neurotransmission (nitrergic basal amylase output. L-NAME and 11 transmission), and disrupting the regulation on September 30, 2021 by guest. Protected copyright. haemoglobin reduced basal amylase out- of islet hormone secretion612 by interfering put to a lesser extent and potentiated the with NOS activity in the islets of Langerhans. inhibitory response to supramaximal We have previously shown that L-NAME caerulein. In contrast, full neural NOS inhibits in vivo pancreatic secretion stimulated inhibition by L-NAME partially reversed by optimal and supramaximal doses of caer- the expected caerulein induced suppres- ulein when both drugs are infused sion of amylase output. This eVect was 1 G simultaneously. However, L-N -nitroarginine reproduced by indazole and capsaicin. (the compound that results from hydrolysis of Digestive System Indazole did not alter responses to either L-NAME and often regarded as equivalent to optimal (0.25 µg/kg) or suboptimal (0.1 13 Research Unit, L-NAME) shows progressive and irreversible Hospital General µg/kg) caerulein, nor, in contrast with 14 15 Universitari Vall inhibition of nNOS. We hypothesised that L-NAME, aggravate the outcome of caer- selective NOS inhibition may reveal the d’Hebron, Barcelona, ulein induced pancreatitis. Spain specific role of various NOS sites in pancreatic E Vaquero Conclusions—Reduction of circulating exocrine secretion. The aim of the present X Molero NO availability, probably of endothelial study was to establish the source of NO that, V Puig-Diví origin, is responsible for the decrease in when blocked, is associated with decreased J-R Malagelada amylase secretion observed in the early pancreatic secretion in vivo. A pharmacological response to L-NAME. Nitrergic neuro- strategy was employed: the pattern of amylase Correspondence to: transmission plays an important role in Dr X Molero, Servicio de secretion was examined in the anaesthetised rat Aparato Digestivo, Hospital the control of pancreatic secretion and in response to selective compounds that inter- General Vall d’Hebron, may induce opposite eVects to endothelial act with the normal physiology (traYcking) of Paseo Vall d’Hebron NOS activity. 119–129, 08035 Barcelona, NO. (Gut 1998;43:684–691) Spain. Indazole inhibits nNOS without interfering 16 Accepted for publication Keywords: nitric oxide; pancreatic secretion; with endothelial NOS activity in vivo. Thus 29 April 1998 pancreatitis; indazole; haemoglobin MAP is not modified by systemic indazole. In Contrasting eVects of circulating and neural NO 685

Table 1 Summary of study groups, route of drug administration, dose and schedule of (Alcobendas, Spain). Indazole was purchased caerulein challenges from Boehringer (Mannheim, Germany). Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from Experimental group Dose, route of administration Time before caerulein stimulation ANIMAL PREPARATION Control (vehicle) i.v./i.p. – Experiments were conducted on a total of 214 L-NAME 30 mg/kg, i.v. 2 min L-NAME 25 30 mg/kg, i.v. 25 min male Sprague-Dawley rats weighing between Indazole 50 + 50 mg/kg, i.p. 50 and 2 min 220 and 320 g. After an overnight fast, rats Indazole 100 100 mg/kg, i.p. 2 min were anaesthetised with ketamine (100 mg/kg Systemic capsaicin 150 mg/kg, s.c. 5 days Local capsaicin 0.1%, perivagal + 1% intraduodenal 5 days body weight intraperitoneally (i.p.)), with sub- Caerulein alone 0.1–50 µg/kg, i.v. – sequent intramuscular (i.m.) doses used as Haemoglobin 500 + 500 mg/kg, i.v. 2 min necessary to maintain anaesthesia. A jugular Aminoguanidine 40 mg/kg, i.v. 60 and 2 min vein was cannulated (polyethylene catheter i.v., intravenous; i.p., intraperitoneal; s.c., subcutaneous. PE-10; Clay Adams, Parsippany, New Jersey, USA) for drug administration and continuous our study, indazole was used to investigate the saline perfusion at 3 ml/h by means of a syringe eVects of selective inhibition of nNOS on pan- pump. A heparinised (10 U/ml) catheter creatic secretion.16–20 Capsaicin is a sensory (PE-50) was placed into the carotid artery to neurotoxin with highly selective eVects on a monitor blood pressure and to facilitate blood subset of primary aVerent sensory neurons, sampling. A pressure transducer was con- including nociceptive neurons. Systemic ad- nected to the carotid catheter for continuous ministration or high concentration local appli- recording of MAP in the anaesthetised resting cation produces anti-nociceptive eVects.21–23 state and in response to administration of NO has been shown to be formed and released drugs. in capsaicin sensitive aVerent fibres.24 25 Thus Animals were then laparotomised and the to demonstrate the contribution of nitrergic biliopancreatic duct was cannulated through its fibres to the control of pancreatic secretion duodenal opening (PE-50). The abdominal without interference from inhibition of other wound was covered with a saline moisturised NO sources, amylase output was measured in gauze, and body temperature was maintained response to caerulein after either inhibition of with a heating lamp. neural transmission of NO by indazole or aVerent sensorial nerve fibre ablation by capsaicin. EXPERIMENTAL PROTOCOLS Since a selective inhibitor for endothelial Measurement of pancreatic secretion Pancreatic secretion was measured as previ- NOS is not currently available, intravenous 1 (i.v.) infusion of haemoglobin was used to ously described. Briefly, after completion of decrease plasmatic free NO acutely. Reduced surgical procedures, biliopancreatic secretion 26 27 was allowed to drain freely for 30 minutes haemoglobin avidly binds free NO, and it

before initiation of the experiments (stabilisa- http://gut.bmj.com/ has been used as an NO scavenger to tion period). Then secretion was collected in investigate the eVects of acute NO removal 28 tared vessels in separate 10 minute fractions for from blood. It appears to have little or no 26 60 minutes and volumes estimated by weight. eVect on nitrergic transmission in vivo. The Amylase output was calculated by determining eVects of inhibition of the inducible isoform of amylase concentration in the collected frac- NOS (non-neural, non-endothelial) by amino- tions. The mean amylase output obtained from guanidine on pancreatic secretion were also two consecutive 10 minute fractions after the

examined. on September 30, 2021 by guest. Protected copyright. stabilisation period was taken as the initial In a previous study1 we have shown that “basal” secretion. Separate experiments were L-NAME, besides potentiating the inhibition of performed to calculate the secretory response pancreatic secretion elicited by supramaximal to vehicle, NOS inhibitors, haemoglobin, or doses of caerulein, also aggravates some of the caerulein at the indicated doses, administered parameters commonly used to assess pancreas an i.v. bolus injection at the end of the 20 atic injury in caerulein induced pancreatitis. In minutes of basal collection (time 0). Table 1 addition, L-NAME has also been shown to summarises the various groups entering the increase tissue levels of trypsinogen activation study. peptide, an index of the amount of tissue Basal secretions tend to fluctuate spontane- necrosis.29 To determine whether specific inhi- ously over time,30 and bursts of secretion occur bition of nNOS may play a role in the develop- 31 ment of acute pancreatitis in this model, the in a cyclic pattern every two hours or less. Accordingly, when unstimulated basal secre- eVects of indazole on three distinct features of acute pancreatic injury (amylasaemia, tissue tion was analysed, data on amylase output were oedema, and polymorphonuclear infiltration) standardised by expressing results as a percent- were evaluated. age of the variation from initial basal output.

NOS blockade by L-NAME The eVects of L-NAME were investigated after Materials and Methods i.v. administration of 30 mg/kg at time 0. Neu- MATERIALS ral NOS inhibition by i.p. L-NAME is best Ketamine was from Parke-Davis SL (EL Prat manifested starting 15 minutes after its 11 de Llobregat, Spain). Caerulein, L-NAME, administration, at a time when there are no bovine haemoglobin, bovine serum albumin, further increments in arterial blood pressure. aminoguanidine, capsaicin, and sodium Thus, to investigate most eVectively the eVect dithionite were from Sigma Chemical Co of nNOS inhibition by L-NAME on pancreatic 686 Vaquero, Molero, Puig-Diví, et al

secretion, amylase output was measured in ment was completed, all of the gauzes were response to supramaximal caerulein (10 µg/ removed and the incisions sutured. Animals

kg), infused either 2 or 25 minutes after were allowed to recover from surgery and Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from L-NAME (30 mg/kg). experiments conducted ﬁve days later.

Specific inhibition of neural NOS Removal of circulating NO Indazole (100 mg/kg, i.p.) was dissolved in one Reduced haemoglobin was prepared as previ- part of ethanol and five parts of peanut oil (v/v) ously described34 by adding to a 1 mmol/l solu- to a maximal volume of 1 ml. Pancreatic secretion of haemoglobin in distilled water, a 10-fold tion was measured as described above. Inda- molar excess of the reducing agent, sodium zole or vehicle was administered i.p. at the end dithionite. The sodium dithionite was then of the initial two 10 minute basal periods (time removed by dialysis against 100 vol of saline for 0). two hours at 4°C. This solution was warmed to Since the eVects of L-NAME on pancreatic 37°C before infusion. secretion were analysed both at 2 and 25 min- Reduced haemoglobin was infused through utes from its administration, two protocols the venous catheter at time 0. A dose of 500 were conducted to examine the eVects of inda- mg/kg in 1 ml saline was administered as a zole on caerulein stimulated pancreatic secre- bolus injection, followed by an infusion of the tion. Indazole (or vehicle) was given in two same amount over two minutes. To prevent equally divided doses (100 mg/kg total, i.p.) 50 acute changes in intravascular volume, an and 5 minutes before juice collection. Pancre- equal amount of blood from the arterial atic secretion was then measured as described catheter was withdrawn. MAP was continu- above in independent experiments in response ously recorded and pancreatic secretions col- to 0.1, 0.25, 1, 10, or 50 µg/kg caerulein (bolus lected in 10 minute fractions. i.v.). In a second set of experiments, indazole In separate studies, the eVect of haemo- (100 mg/kg) or vehicle was administered 2 globin on pancreatic response to supramaximal minutes before 10 µg/kg caerulein. caerulein was investigated. Haemoglobin was administered as described above. Two minutes Ablation of sensorial aVerent nerve fibres by after the end of haemoglobin infusion, caer- capsaicin ulein (10 µg/kg) was administered and pancre- To investigate the eVects of functional destruc- atic juice collected for measurement of amylase tion of aVerent nerve fibres on pancreatic output. secretory response to supramaximal caerulein, To exclude possible eVects of an increase in capsaicin was administered according to two oncotic pressure, albumin was infused in diVerent protocols (systemic or local adminis- amounts that matched the oncotic capacity of tration) five days before measurement of haemoglobin.35 Albumin (485.5 mg/kg) in 1 ml

pancreatic secretion in response to i.v. caer- saline was administered in the same way as http://gut.bmj.com/ ulein (10 µg/kg). haemoglobin, and MAP and amylase output measured. Systemic capsaicin administration The secretory pancreatic response was also Capsaicin as a 1% (w/v) solution was dissolved examined after aminoguanidine treatment. in ethanol/Tween 80/saline (10:10:80, by vol). Aminoguanidine (40 mg/kg body weight, i.v.) Under light ether anaesthesia, rats were inhibits the inducible NOS isoform and was injected subcutaneously (s.c.) with capsaicin used as a control for inhibition of non-neural in three divided doses on three consecutive non-endothelial NOS. on September 30, 2021 by guest. Protected copyright. days for a total dose of 150 mg/kg in a regimen shown to induce functional ablation of Induction of acute pancreatitis primary aVerent sensitive ﬁbres.32 Rats were Acute pancreatitis was produced by adminis- pretreated with terbutaline (0.1 mg/kg, i.m.) tering a total of four i.p. injections of caerulein and aminophylline (10 mg/kg, i.m.) before at the dose of 20 µg/kg body weight at one hour capsaicin injections to prevent acute respira- intervals in ketamine anaesthetised rats (n = 5 tory distress. On the day before the per group). Indazole (50 mg/kg) was solubi- experiments, defunctionalisation of aVerent lised by sonication and warming in 4% ethanol

neurons was ascertained by evaluating the in 0.5% Na2CO3, pH 7.4, and administered i.p. reduction of wiping movements in response to 30 minutes before induction of acute pancrea- intraocular instillation of a 0.1% solution of titis. To maintain adequate levels of indazole capsaicin. A reduction of more than 80% throughout the experiment, doses of 10 mg/kg compared with control rats was considered an indazole were administered i.m. at one hour adequate sensitive defunctionalisation.33 intervals. Control rats received either vehicle alone or vehicle plus caerulein. Blood was Local capsaicin administration obtained by cardiac puncture for determina- Rats were anaesthetised with ketamine. After a tion of amylase activity nine hours after the first midline incision, both subdiaphragmatic vagal caerulein injection. Amylase activity was deter- trunks were exposed and surrounded with a mined by the alpha amylase EPS test (Boe- small piece of gauze soaked with 100 µl 0.1% hringer) for BM/Hitachi system 717. Aliquots capsaicin for 30 minutes. In addition, a gauze of volume 200 µl of 1:800 to 1:2000 dilutions soaked with 200 µl 0.5% capsaicin was of the sample were used in the assay. advanced to the duodenum from the stomach To measure pancreatic oedema, excised through an antral incision and applied to the pancreata were wet weighed, desiccated at duodenal mucosa for 30 minutes. When treat- 160°C for 24 hours and reweighed. Pancreatic Contrasting eVects of circulating and neural NO 687

evaluated by analysis of variance. Statistical comparisons of data in ﬁg 2 were performed

75 using repeated measures analysis of variance Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from followed by Fisher’s protected least signiﬁcant diVerences test. DiVerences were regarded as * * signiﬁcant when p<0.05.

50 Results EFFECT OF NOS INHIBITORS, REDUCED HAEMOGLOBIN, AND CAERULEIN ON MAP Basal MAP was 113 (2) mm Hg. Infusion of 25 reduced haemoglobin produced a sharp and ** marked increase (ﬁg 1). This eVect was revers- ible, with MAP returning to baseline values in

Peak change in MAP (% control) 11 (0.5) minutes. Supramaximal caerulein in † † 200 µl saline also induced a sharp rise in MAP which matched that induced by haemoglobin, but with shorter duration (5.25 (0.4) minutes). –10 L-NAME Caerulein Albumin Caerulein In contrast, infusion of caerulein at submaxi- (10 µg/kg) (0.1 µg/kg) mal doses of 0.1 µg/kg in the same amount of Haemoglobin Haemoglobin Indazole saline (200 µl) induced a small but significant + caerulein and transient (5 (0.7) minutes) reduction in Figure 1 EVect of single doses of L-NAME (30 mg/kg, i.v.), haemoglobin (1 g/kg, i.v.), MAP. The eVect of i.v. bolus infusion of supramaximal caerulein (10 µg/kg, i.v.), haemoglobin plus supramaximal caerulein, L-NAME was characterised by a progressive albumin (970 mg/kg), indazole (100 mg/kg, i.p.), and suboptimal caerulein (0.1 µg/kg, i.v.) on mean arterial pressure (MAP) in anaesthetised rats (n=8foreach experimental increase in MAP during the first 7.6 minutes, condition). Bars represent average percentage variation from control in peak MAP induced followed by a slow decline to 1.6 times baseline by drug administration. *p<0.05 compared with L-NAME group. , and †‡p<0.05 values by the end of the observation period. compared with all other groups. Peak change in MAP was greater after water content was calculated as a percentage of L-NAME infusion than after haemoglobin or total wet weight. caerulein infusion alone. Interestingly, the peak Tissue myeloperoxidase (MPO) activity was rise in MAP was enhanced by the combined used as a biochemical marker for polymorpho- treatment of supramaximal caerulein and nuclear infiltration. It was determined using a haemoglobin, which then became similar to the previously described method36 37 that measures peak change in MAP observed after L-NAME infusion. MPO activity as the result of H2O2 dependent Treatment with albumin in amounts that

oxidation of 3,3',5,5'-tetramethylbenzidine and http://gut.bmj.com/ expressed as units per mg protein. matched the expected oncotic capacity of haemoglobin raised MAP to a much lesser extent and duration (6.25 (0.25) minutes) than STATISTICAL ANALYSIS Results are presented as means (SEM). For the response obtained by reduced haemoglobin or L-NAME. Indazole given either i.m. or i.p. statistical evaluations of diVerences between showed no significant eVects on MAP. two groups of animals, Students’s t test for unpaired data was used. DiVerences between more than two groups of animals were EFFECT OF NOS INHIBITORS AND REDUCED on September 30, 2021 by guest. Protected copyright. HAEMOGLOBIN ON BASAL AMYLASE OUTPUT 1 Albumin As previously described, an i.v. L-NAME 150 Vehicle bolus of 30 mg/kg decreased basal amylase Haemoglobin output (fig 2). Removal of circulating NO by L-NAME infusion of reduced haemoglobin caused a Drugs Indazole decrease in amylase output of similar magni- *† tude. In contrast, i.v. infusion of iso-oncotic amounts of albumin did not inhibit basal secretion. In fact, albumin produced a slight, 100 but significant, increment in amylase output. Systemic capsaicin treatment virtually abol- * ished the eye wiping response to a local irritant, indicating eVective defunctionalisation of sensory fibres. Inhibition of nNOS by indazole (fig * Amylase output (% basal) 2) or capsaicin treatment (both systemically or locally applied) resulted in a reduction in basal 50 *† amylase output. Capsaicin treatment reduced unstimulated amylase output relative to that of control rats (47.4 (8) v 85.8 (11) U/10 min; 0 10 20 30 40 p<0.05), which was not statistically diVerent Time (minutes) from amylase output measured after indazole Figure 2 Time course of the eVect of single bolus injections of L-NAME, haemoglobin, treatment (42.7 (7) U/10 min). albumin, indazole, or vehicle on amylase output expressed as percentage of variation from Aminoguanidine, a non-neural non- basal. Each point is the mean (SE) from five separate experiments for each experimental condition. Doses were as in fig 1. *p<0.05 compared with vehicle. †p<0.05 compared with endothelial inhibitor of NOS, had no detect- all groups. able eVect on basal amylase output (n = 5). 688 Vaquero, Molero, Puig-Diví, et al

1500 Sensorial nerve ablation by capsaicin reproduced the eVects of indazole treatment and rendered rats more sensitive to doses of Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from Caerulein alone caerulein that would ordinarily suppress amy- Caerulein + indazole lase secretion (ﬁg 4). This type of response was readily observed in all rats treated with capsai- 1000 cin, whichever way capsaicin was administered, that is either via systemic or local application. As expected, co-infusion of L-NAME and supramaximal caerulein induced a further reduction in amylase output as compared with * caerulein alone. However, when nNOS was 500 fully inhibited by L-NAME (administered 25 minutes before supramaximal caerulein) a paradoxical response occurred, since L-NAME * counteracted the suppression of amylase out-

Amylase output in 40 min (% control) put elicited by supramaximal caerulein alone * (ﬁg 4), in sharp contrast with the profound inhibition of amylase output observed when 0 0.1 0.25 10 50 L-NAME was co-administered with caerulein. Caerulein (µg/kg)

Figure 3 Amylase output in response to independent bolus doses of caerulein alone or EFFECT OF L-NAME AND REDUCED HAEMOGLOBIN caerulein plus indazole (100 mg/kg). Each bar represents percentage increment from ON PANCREATIC RESPONSE TO SUPRAMAXIMAL baseline amylase output 40 minutes after caerulein infusion (means (SE)). Six or more separate experiments were conducted for each experimental condition. *p<0.05 compared CAERULEIN with caerulein alone. The decrease in circulating NO caused by reduced haemoglobin or L-NAME, when EFFECT OF nNOS INHIBITION ON CAERULEIN administered 2 minutes before supramaximal STIMULATED PANCREATIC SECRETION caerulein, potentiated the inhibition of amylase Indazole did not modify amylase output in secretion produced by these doses of caerulein response to suboptimal and optimal doses of (ﬁg 4). Thus removal of circulating NO is caerulein (ﬁg 3). However, indazole treatment clearly detrimental to normal amylase secretion partially reversed the inhibitory eVect of a by the intact pancreas in vivo. supramaximal (10 µg/kg) dose of caerulein. Inhibition of the inducible form of NOS by Increasing the dose of caerulein to 50 µg/kg aminoguanidine did not modify the pancreatic reduced the secretory response in indazole response to supramaximal caerulein (260 (43) treated rats, suggesting that nNOS blockade U/40 min for caerulein alone v 192 (27) for induces a desensitisation of the inhibitory

caerulein plus aminoguanidine; p>0.05; n = 6). http://gut.bmj.com/ pathway activated by high doses of caerulein. No diVerences were observed in the amylase EFFECT OF nNOS INHIBITION ON CAERULEIN response to supramaximal caerulein, whether INDUCED ACUTE PANCREATITIS indazole was administered 2 or 50 minutes Repeated i.p. caerulein (20 µg/kg each) in- before caerulein (data not shown). creased serum amylase (32.2 (3.6) U/ml) and pancreatic water content (81.3 (0.9)% of pan- * creatic weight) as compared with unstimulated rats (4.3 (0.4) U/ml and 71.2 (1.3)% respec- on September 30, 2021 by guest. Protected copyright. 500 tively). Indazole treatment failed to either ameliorate or aggravate caerulein induced * increments in these two parameters of early 400 pancreatic injury (28.2 (2.6) U/ml and 82 * (1.1)%). However, indazole reduced the increase in MPO activity induced by caerulein 300 (1300 (148) v 793 (115) mU/mg protein; p<0.05), indicating a putative permissive role for neural NO in polymorphonuclear inﬁltra- 200 tion to the pancreas.

Discussion Amylase output (% control) 100 The issue investigated in our studies is which NOS isoform blocked by the non-speciﬁc * * inhibitor L-NAME is responsible for the observed inhibition of amylase output. To Caerulein +Haemoglobin +Indazole address this problem, we followed a pharmaco- +L-NAME +L-NAME* +Capsaicin Caerulein (10 µg/kg) logical approach: pancreatic secretory pattern was measured in response to NOS antagonists. Figure 4 Amylase output in response to an intravenous bolus dose of caerulein (10 µg/kg). To inhibit nNOS activity, indazole, a specific Caerulein was infused alone or after L-NAME (30 mg/kg), haemoglobin (1 g/kg), or indazole (100 mg/kg) administration or after aVerent nerve ablation by local application of inhibitor, was employed. Indazole and certain capsaicin. Drugs were given 2 minutes before caerulein except for L-NAME* (25 minutes) related derivatives have been reported to selec- or capsaicin (ﬁve days before the experiments). Each bar represents percentage increment tively inhibit nNOS in vivo.16 17 They show no from baseline amylase output 40 minutes after caerulein infusion (means SE)). Six or more separate experiments were conducted for each experimental condition. *p<0.05 compared eVects on endothelial NOS and accordingly do with caerulein alone. not raise arterial blood pressure.19 20 Certainly, Contrasting eVects of circulating and neural NO 689

indazole did not modify MAP in our experi- To obtain further support for this hypothesis ments. However, impairment of nitrergic neu- we acutely removed NO from the intravascular rotransmission by nNOS inhibition reduced bed by infusion of haemoglobin, an NO Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from baseline amylase output. In the experimental scavenger. Reduced haemoglobin avidly binds design of the present work, biliopancreatic and inactivates NO,27 39 but does not interfere secretion is diverted from the duodenal lumen with intracellular NO pathways or prevent for 50 minutes before addition of drugs (stabi- neural NO transmission in vivo.26 lisation period plus two consecutive 10 minute An inhibitory eVect of haemoglobin on NO periods to calculate “basal” secretion). There- mediated neurotransmission in vitro has been fore basal secretion was most likely the result of reported.40 However, our current understand- endogenous cholecystokinin stimulation at ing of the pharmacokinetics of native or artifi- picomolar (physiological) concentrations.38 cially modified haemoglobins and their relation Our results on basal pancreatic secretion after to NO binding argues against the conclusions indazole treatment are in close agreement with drawn by some investigators on the eVects of those obtained by Li 38 in response to bilio- i.v. haemoglobin on NO neurotransmission in et al 41 pancreatic juice diversion after truncal vivo. Owing to its high aYnity for NO, intra- vagotomy or perivagal application of capsaicin. venously infused haemoglobin rapidly reacts Although L-NAME inhibits all NOS isoen- with circulating free NO (the reaction is essen- zymes, its eVect on nNOS appears to be tially instantaneous) to form nitrate and met- haemoglobin, which is the basis for a well delayed with respect to endothelial NOS, since 42 it is fully expressed after 15 minutes of systemic known NO assay. administration,11 whereas MAP (resulting from Removal of free NO from the circulation induces a transient increase in MAP and an endothelial NOS inhibition) starts to increase 43 44 in two to three minutes. This may explain why increase in peripheral vascular resistance. To scavenge NO close to any synaptic cleft in indazole had a greater impact on baseline amy- vivo, where NO might be released, haemo- lase secretion, reflecting a regulatory function globin should reach the extravascular space of nitrergic transmission in the maintenance of while still keeping its NO binding capacity. the neural pathways that control the normal However, free circulating haemoglobin is likely physiology of baseline pancreatic secretion. to be rapidly processed into modified metabo- Indazole treatment showed no measurable lites that may lose their NO aYnity. In this eVect on the secretory response to suboptimal regard, some haemoglobin derivatives even and optimal doses of caerulein, but partially show vasodilator activity after having been reversed the suppression of secretory activity nitrosylated.44 It is conceivable that some of elicited by supramaximal doses of caerulein, the reported eVects of haemoglobin on neuro- just as L-NAME when fully inhibiting nNOS. transmission in vivo are related to eVects The pattern of response to indazole was repro-

caused by the activities of its catabolic http://gut.bmj.com/ duced by capsaicin. Basal secretion was products.45 46 reduced and supramaximal inhibition partially Under basal conditions, circulating NO reversed. Both capsaicin application and mostly reﬂects endothelial NOS activity. NO is nNOS inactivation result in an anti-nociceptive a major physiological regulator of basal sys- eVect and prevention of neuropeptide temic blood vessel tone. Continuous release of 11 20 22–24 release. This suggests that impairment NO by vascular endothelial cells controls MAP of neural NO transmission shifts the caerulein in the resting state.47 Variations in MAP in dose-response curve to the right, rendering response to NO inhibitors reﬂect, to a large on September 30, 2021 by guest. Protected copyright. anaesthetised rats less sensitive to cholecystoki- extent, variations in the concentration of circu- nin; they do not respond to low cholecystokinin lating free NO of endothelial origin. Indeed, concentrations (as induced by biliopancreatic haemoglobin infusion induced a sharp rise in juice diversion) but are capable of secretory MAP and a parallel reduction in amylase responses to supramaximal doses of caerulein output. In contrast, iso-oncotic amounts of that would otherwise suppress amylase output. albumin resulted in a smaller increment in We observed a dual eVect of L-NAME on MAP while pancreatic secretion actually in- pancreatic secretion: when it is evaluated creased. The reason for this divergent eVect immediately after L-NAME infusion, inhibi- may reside in the diVerent interactions between tion of basal, optimal, and supramaximal caer- both haemoglobin and albumin and free NO. ulein stimulation occurs, as previously Haemoglobin scavenges and inactivates described.1 In contrast, when the pancreas is NO27 39 44 whereas albumin acts as a carrier for stimulated allowing a short lag period after NO making it available to the micro- L-NAME administration (25–35 minutes), the vasculature.48 Amylase output in response to pattern of response parallels that observed after supramaximal caerulein stimulation was indazole or capsaicin treatment, two independ- further reduced when haemoglobin was given ent pharmacological approaches that impair i.v., just as L-NAME does when it is sensorial nerve function. Subsequently, no fur- co-administered with caerulein, and in sharp ther increases in arterial pressure are recorded contrast with the pattern of secretion induced (there is actually a slow decline in MAP) but by indazole or capsaicin. L-NAME accomplishes full irreversible inhibi- Interestingly, supramaximal caerulein in- tion of nNOS.11 This suggests that nNOS inhi- duced an increase in MAP and potentiated the bition may suppress or totally reverse the peak increase in MAP elicited by haemoglobin. eVects of endothelial NOS blockade on pancre- Suboptimal doses of caerulein (which would atic secretion. induce plasma caerulein concentrations that 690 Vaquero, Molero, Puig-Diví, et al

match postprandial plasma cholecystokinin) 1 Molero X, Guarner F, Salas A, et al. Nitric oxide modulates pancreatic basal secretion and response to cerulein in the did not increase MAP but, on the contrary, rat: eVects in acute pancreatitis. Gastroenterology induced a transient reduction in MAP. 1995;108:1855–62. Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from 2 Holst JJ, Rasmussen TN, Schmidt P. Role of nitric oxide in The eVects of caerulein on the microvascula- neurally induced pancreatic exocrine secretion in pigs. Am ture have been known for a long time, but the J Physiol 1994;266:G206–13. 3 Patel AG, Toyama MT, Nguyen TN, et al. Role of nitric mechanisms underlying this vasoreactivity are oxide in the relationship of pancreatic blood flow and exo- poorly understood. In dogs, i.v. bolus of crine secretion in cats. Gastroenterology 1995;108:1215–20. 4 Konturek SJ, Bilski J, Konturek PK, et al. Role of caerulein at doses that eVectively stimulate endogenous nitric oxide in the control of canine pancreatic pancreatic secretion induce a fall in blood secretion and blood flow. Gastroenterology 1993;104:896– 49 902. pressure. Likewise, the transient increase in 5 Shimosegawa T, Abe T, Satoh A, et al. Histochemical dem- MAP after 10 µg/kg caerulein is in agreement onstration of NADPH-diaphorase activity, a marker for nitric oxide synthase, in neurons of the rat pancreas. Neuro- with reports of an increase in MAP after these sci Lett 1992;148:67–70. 50 6 Schmidt HH, Warner TD, Ishii K, et al. Insulin secretion doses of caerulein. from pancreatic ß cells caused by L-arginine-derived nitro- It has been shown that NOS inhibition by gen oxides. Science 1992;225:721–3. L 7 Schmidt HHHW, Gagne GD, Nakane M, et al. Mapping of -NAME aggravates caerulein induced neural nitric oxide synthase in the rat suggests frequent pancreatitis.129 Since indazole counteracted co-localization with NADPH diaphorase but not with solu- the suppression of amylase secretion in re- ble guanylyl cyclase, and novel paraneural functions of nitrergic signal transduction. J Histochem Cytochem 1992; sponse to doses of caerulein that induce the 40:1439–56. 8 Gukovskaya AS, Pandol SJ. Nitric oxide production development of pancreatitis, we examined regulates cGMP formation and calcium influx in pancre- whether or not indazole treatment could atic acinar cells. Am J Physiol 1994;266:G350–6. 9 Menozzi D, Sato T, Jensen RT, et al. Cyclic GMP does not prevent pancreatic injury. inhibit protein kinase C-mediated enzyme secretion in rat Moreover, NO has been involved in the pancreatic acini. J Biol Chem 1989;264:995–9. 10 Gardiner SM, Compton AM, Kemp PA, et al. Regional and pathogenesis of oedema formation in animal cardiac hemodynamic eVects of NG-nitro-L-arginine me- models of neurogenic inflammation.20 If neural thyl ester in conscious, Long Evans rats. Br J Pharmacol 1990;101:625–31. NO contributes to oedema formation in 11 Moore PK, Oluyomi AO, Babbedge RC, et al.L-NG-nitro caerulein induced pancreatitis, liquid accumu- arginine methyl ester exhibits antinociceptive activity in the mouse. Br J Pharmacol 1991;102:198–202. lation in the pancreas would be expected to be 12 Salehi A, Carlberg M, Henningson R, et al. Islet constitutive prevented, at least in part, by indazole nitric oxide synthase: biochemical determination and regulatory function. Am J Physiol 1996;270:C1634–41. treatment. 13 Southan GJ, Szabó C. Selective pharmacological inhibition In contrast with L-NAME eVects when it is of distinct nitric oxide synthase isoforms. Biochem Pharma- 1 col 1996;51:383–94. co-infused with caerulein, impairment of 14 Dwyer MA, Bredt DS, Snyder SH. Nitric oxide synthase: nitrergic transmission by indazole did not irreversible inhibition by L-NG-nitroarginine in brain in vitro and in vivo. Biochem Biophys Res Commun 1991;176: increase hyperamylasaemia in caerulein in- 1136–41. duced pancreatitis. Also, indazole had no 15 Furfine ES, Harmon MF, Paith JE, et al. Selective inhibition of constitutive nitrergic oxide synthase by L-NG- measurable eVect on liquid accumulation in nitroarginine. Biochemistry 1993;32:8512–17.

the pancreas, suggesting that nitrergic trans- 16 Babbedge RC, Bland-Ward PA, Moore Pk. Inhibition of rat http://gut.bmj.com/ cerebellar nitric oxide syntase by 7-nitro indazole and mission is not significantly involved in oedema related substituted indazoles. Br J Pharmacol 1993;110: formation in this model of inflammation. The 225–8. 17 Moore PK, Wallace P, GaVen Z, et al. Characterization of fact that indazole reduced MPO activity is once the novel nitric oxide synthase inhibitor 7-nitro indazole again in opposition to the previously described and related indazoles: antinocioceptives and cardiovascular eVects. Br J Pharmacol 1993;110:219–24. eVects of L-NAME in this situation, since 18 Faraci FM, Brian JE. 7-nitroindazole inhibits brain nitric L-NAME has been shown to increase MPO oxide synthase and cerebral vasodilatation in response to 1 N-methyl-D-aspartate. Stroke 1995;26:2172–6. activity. 19 Moore PK, Wallace P, GaVen Z, . 7-nitroindazole, an et al on September 30, 2021 by guest. Protected copyright. NOS is widely distributed in the body and inhibitor of nitric oxide synthase, exhibits antinocioceptive activity in the mouse without increasing blood presure. Br may well be present in several cell types in a J Pharmacol 1993;108:296–7. given organ, as in the case of the pancreas. 20 Kajekar R, Moore PK, Brain SD. Essential role for nitric oxide in neurogenic inflammation in rat cutaneous General blockade of NOS by universal inhibi- microcirculation. Circ Res 1995;76:441–7. tors such as L-NAME may give rise to 21 Buck SH, Burks TF. The neuropharmacology of capsaicin: a review of some recent observations. Pharmacol Rev 1986; contrasting eVects in organ function, depend- 38:179–226. ing on the predominant pathway or cell 22 Bevan S, Szolcsnyi J. Sensory neuron-specific actions of capsaicin: mechanisms and applications. Trends Pharmacol function that is disrupted and on alternative Sci 1990;11:330–3. 23 Dray A. Mechanism of action of capsaicin-like molecules on activation of compensatory mechanisms. We sensory neurons. Life Sci 1992;51:1759–65. should reconsider conclusions drawn from in 24 Birder LA, Kanai AJ, Groat WC. DMSO: eVect on bladder aVerent neurons and nitric oxide release. J Urol 1997;158: vivo experiments, especially those in which 1989–95. L-NAME was given in repeated doses or for 25 Ren K, Ruda MA. Nitric oxide synthase-containing neurons in sensory ganglia of the rat are susceptible to capsaicin- long periods of time. The use of inhibitors that induced cytotoxicity. Neuroscience 1995;65:505–11. selectively target diVerent forms of NOS (best 26 Jenkinson KM, Reid JJ, Rand MJ. Hydroxocobalamin and haemoglobin diVerentiate between exogenous and neuro- if cell selective) should be recommended, since nal nitric oxide in the rat gastric fundus. Eur J Pharmacol they would produce more meaningful results. 1995;275:145–52. 27 Feelish M, Noack EA. Correlation between nitric oxide for- We conclude that nitrergic nerve fibres play mation during degradation of organic nitrates and an important role in the control of pancreatic activation of guanylate cyclase. Eur J Pharmacol 1987;139: 19–30. secretion, independently of other pancreatic 28 Salvemini D, Radziszewski W, Karbut R, et al. The use of NO sources. Pancreatic eVects of nNOS oxyhaemoglobin to explore the events underlying inhibition of platelet aggregation induced by NO or NO-donors. inhibition diVer from the eVects resulting from Br J Pharmacol 1990;101:991–5. endothelial NOS blockade, which we believe is 29 Werner J, Rivera J, Fernandez-del Castillo C, et al.DiVering roles of nitric oxide in the pathogenesis of acute edematous the NO related event that produces inhibition versus necrotizing pancreatitis. Surgery 1997;121:23–30. 30 Dale WE, Turkelson CM, Solomon TE. Role of cholecysto- of basal and stimulated secretion in vivo and kinin in intestinal phase of meal-induced pancreatic secre- the aggravation of acute pancreatitis. tion. Am J Physiol 1989;257:G782–90. Contrasting eVects of circulating and neural NO 691

31 Maouyo D, Guan D, Rivard N, et al. Stability of circadian 40 Rattan S, Rosenthal GJ, Chakder S. Human recombinant and minor cycles of exocrine pancreatic secretion in hemoglobin (rHb1.1) inhibits nonadrenergic noncholiner- atropine- and MK-329-infused rats. Am J Physiol 1995; gic (NANC) nerve-mediated relaxation of internal anal 268:G251–9. sphincter. J Pharmacol Exp Ther 1995;272:1211–16. Gut: first published as 10.1136/gut.43.5.684 on 1 November 1998. Downloaded from 32 Martling CR. Sensory nerves containing tachykinins and 41 Murray JA, Ledlow A, Launspach J, et al. The eVects of cGRP in the lower airways. Acta Physiol Scand Suppl 1987; recombinant human hemoglobin on esophageal motor 563:1–57. function in humans. Gastroenterology 1995;109:1241–8. 33 Girolomoni G, Tigelaar RE. Capsaicin-sensitive primary 42 Hevel JM, Marletta MA. Nitric-oxide synthase assays. sensory neurons are potent modulators of murine delayed- Methods Enzymol 1994;233:250–8. type hypersensitive reactions. J Immunol 1990;145:1105– 43 Hess JR, MacDonald VW, Brinkley WW. Systemic and pul- 12. monary hypertension after resucitation with cell-free 34 Martin W, Villani GM, Jothianandan D, et al. Selective hemoglobin. J Appl Physiol 1993;74:1769–78. blockade of endothelium-dependent and glyceryl 44 Jia L, Bonaventura C, Bonaventura J, et al. trinitrate-induced relaxation by hemoglobin and by meth- S-nitrosohaemoglobin: a dynamic activity of blood involved ylene blue in the rabbit aorta. 1985; J Pharmacol Exp Ther in vascular control. 1996; :221–6. 232:708–16. Nature 380 45 Somers M, Piqueras AI, Strange K, . Interactions of 35 Thompson A, McGarry AE, Valeri CR, et al. Stroma-free et al hemoglobin increases blood pressure and GRF in the ultrapure bovine hemoglobin with renal epithelial cells in hypotensive rat: role of nitric oxide. J Appl Physiol 1994;77: vivo and in vitro. Am J Physiol 1997;273:F38–52. 2348–24. 46 Everse J, Hsia N. The toxicities of native and modiﬁed 36 Grisham MB, Benoit JN, Granger DN. Assessment of hemoglobins. Free Radic Biol Med 1997;22:1075–99. leukocyte involvement during ischemia and reperfusion of 47 Rees DD, Palmer RMJ, Moncada S. Role of endothelium- the intestine. Methods Enzymol 1990;186:729–41. derived nitric oxide in the regulation of blood pressure. Proc 37 Yamada T, Sartor RB, Marshall S, et al. Mucosal injury and Natl Acad Sci USA 1989;86:3375–8. inﬂammation in a model of chronic granulomatous colitis 48 Stamler JS, Jaraki O, Osborne JA. Nitric oxide circulates in in rats. Gastroenterology 1993;104:759–71. mammalian plasma primarily as an S-nitroso adduct of 38 Li Y, Owyang C. Endogenous cholecystokinin stimulates serum albumin. Proc Natl Acad Sci USA 1992;89:7674–7. pancreatic enzyme secretion via vagal aVerent pathway in 49 Erspamer V, Bertaccini G, De Caro G, et al. Pharmacologi- rats. Gastroenterology 1994;107:525–31. cal actions of caerulein. Experientia 1967;23:702–3. 39 Moncada S, Palmer RMJ, Higgs EA. Nitric oxide: 50 Bertolini A, Guarini S, Ferrari W, et al. Caerulein and chole- physiology, pathophysiology, and pharmacology. Pharmacol cystokinin reverse experimental hemorrhagic shock. Neu- Rev 1991;43:109–42. ropeptides 1986;8:25–31. http://gut.bmj.com/ on September 30, 2021 by guest. Protected copyright.