Dual effects of α 2 -adrenoceptors in modulating myogenic tone in sheep isolated internal anal sphincter

Dual effects of a2-adrenoceptors in modulating myogenic

tone in sheep isolated internal anal sphincter

S. J. RAYMENT,* J. A. D. SIMPSON,* T. EAMES,* A. G. ACHESON,* M. R. DASHWOOD,† Y. HENRY,‡ H. GRUSS,‡J. H. SCHOLEFIELD* & V. G. WILSON§

*Division of GI Surgery, School of Medicine, University of Nottingham, Queen’s Medical Centre, Nottingham, UK

†Clinical Biochemistry, Royal Free and University College Medical School, London, UK

‡Clinical Development, Norgine Ltd, R&D Division, Harefield, UK

§School of Life Sciences, University of Nottingham, Queen’s Medical Centre, Nottingham, UK

Key Messages

• a2-adrenoceptors were found to act as both a constrictor and a neuromodulatory agent in sheep internal anal

sphincter (SIAS). This dual effect increases our understanding of how these pharmacological tools could be used

to potentially benefit patients with faecal incontinence.

• The aim of this research was to characterise the responses of SIAS to 3 imidazoline compounds by organ bath

pharmacology: identifying their affinity at different adrenoceptor subtypes using competitive radioligand

binding. Additionally the effect of one of these agents (brimonidine) on neurogenic responses using electrical

field stimulation (EFS) was established.

• Whilst all the agents tested caused constriction of smooth muscle, only with brimonidine was this mediated

through a2-adrenoceptors. Brimonidine inhibited adrenergic contractile responses to EFS.

Abstract

Background The role of a-adrenoceptors in promoting

continence through modulation of sphincter tone has

focused primarily on the effects of a1-adrenoceptors.We have used three clinically available agents, which

are selective for a2-adrenoceptors, to investigate their

role in contractile and neurogenic responses on the

internal anal sphincter (IAS). Methods IAS strips,

which had spontaneously generated tone, were used

to investigate the contractile effect of lofexidine,

brimonidine, and dexmedetomidine on muscle tone

in the presence or absence of subtype selective

antagonists. The effect of brimonidine on the magni-

tude and time course of neurogenic responses gener-

ated by electrical field stimulation (EFS) was also

examined. The affinity of test compounds at a1- anda2-adrenoceptors was established by competition

binding with [3H]-prazosin and [3H]-RX821002. Key

Results All agonists caused concentration-dependent

contraction of the IAS and lofexidine demonstrated

an enantiomeric difference in potency with a 10-fold

difference between the (�) and (+) isomers. Responses

to lofexidine and dexmedetomidine were inhibited in

the presence of the a1-adrenoceptor selective antago-

nist prazosin, but not in the presence of RX811059

(a2-adrenoceptor selective antagonist); brimonidine

responses were inhibited by RX811059 and, to a

lesser extent, by prazosin. Brimonidine affected both

magnitude and duration of neurogenic responses,

which was reversed in the presence of RX811059.

Conclusions & Inferences We conclude that a2-adrenoceptors can mediate contraction of IAS,

although this effect is most evident with efficacious

imidazoline agonists rather than the most selective

ligand. In addition, this receptor subtype can directly

inhibit noradrenergic contractile responses to EFS

and, indirectly, enhance nitrergic relaxatory

responses.

Address for Correspondence

Dr S. J. Rayment, Division of GI Surgery, School of Medicine,University of Nottingham, Queen’s Medical Centre, CliftonBoulevard, Nottingham NG7 2UH, UK.Tel: +44 (115) 823 0154; fax: +44(115) 823 1160;e-mail: [email protected]: 3 December 2013Accepted for publication: 17 April 2014

© 2014 John Wiley & Sons Ltd 1095

Neurogastroenterol Motil (2014) 26, 1095–1103 doi: 10.1111/nmo.12363

Neurogastroenterology & Motility

Keywords fecal incontinence, imidazoline, radioli-

gand binding.

INTRODUCTION

Over the past 15 years, developments in the pharma-

cological treatment of ano-rectal disorders have been

largely driven by improved understanding of the auto-

nomic control of the anal sphincter and the use of

existing drugs, with established systemic use in man,

for novel indications. Thus, the smooth muscle relax-

ants glyceryl trinitrate1 and nifedipine2 have been used

for the treatment of anal fissures, while the a1-adreno-ceptor agonists phenylephrine3 and L-erythromethox-

amine4 have been investigated as potential agents for

the management of fecal incontinence.

Despite the availability of a large number of clini-

cally evaluated imidazoline derivatives, with known

selectivity for a2-adrenoceptors, surprisingly little is

known about the role of these receptors in the control

of anal sphincter tone. It is clear, however, that this

subtype may have a role in the treatment of fecal

incontinence. While radioligand binding studies have

revealed the presence of a2-adrenoceptor binding sites

in the sheep anal sphincter membranes, the density

was only half of that detected for a1-adrenoceptorbinding sites.5 This marked difference in receptor

density may account for the finding that while cloni-

dine was ~10-fold more potent than L-erythromethox-

amine in the sheep internal anal sphincter, the

maximum contraction was smaller than that to the

selective a1-adrenoceptor agonist. A subsidiary role for

a2-adrenoceptors in pig and sheep anal sphincter con-

tractions was also suggested when subtype-selective

antagonists were used to examine responses to nor-

adrenaline6 and clonidine.5

Clonidine is clinically used as an antihypertensive

and antimigraine agent and is generally considered a

partial agonist at a2-adrenoceptors.7 Under certain

circumstances, however, it can also activate a1-adreno-ceptors.8 Interestingly, three other imidazoline deriv-

atives, dexmedetomidine,9 brimonidine,10,11 and

lofexidine12, are clinically used as a2-adrenoceptoragonists. These agents are applied in different indica-

tions: anesthetic sedative, glaucoma, and morphine

withdrawal, respectively, but are generally considered

to be either more potent, and/or more efficacious than

clonidine at a2-adrenoceptors.9–12

The aims of this study were threefold. Firstly, to

determine the selectivity of dexmedetomidine, brimo-

nidine, and lofexidine at a1- and a2-adrenoceptorbinding sites in the sheep internal anal smooth

muscle. As lofexidine occurs as a racemic mixture,

we have also determined the selectivity of the indi-

vidual (�) and (+) isomers.13,14 Secondly, to determine

the potency of the imidazoline derivatives as contrac-

tile agents, compared with that of L-erythromethox-

amine, and characterize the principal receptor subtype

involved. Finally, to use one of the agents to investi-

gate whether a2-adrenoceptors are able to modulate

nitrergic or adrenergic neurotransmission in the anal

sphincter.

MATERIALS AND METHODS

Sheep internal anal sphincter (IAS) was isolated by dissectionfrom sheep anus (obtained from a local abattoir), as previouslydescribed.15

For competition binding experiments, isolated sheep IAS weredissected and frozen (at �20 °C) on the day of collection. Whensufficient material had been collected, membrane preparations ofthe sheep IAS were freshly prepared using the method previouslydescribed5 and utilized in competition binding studies with [3H]-prazosin and [3H]-RX821002 to assess the relative affinities ofbrimonidine, dexmedetomidine, and lofexidine (isomers andracemate) at a1- and a2-adrenoceptors, respectively. [3H]-prazosinand [3H]-RX821002 were used at a fixed concentration of 0.7 and1 nM, respectively. Competitor concentration ranges were opti-mized for individual agonists, but ranged between 1 9 10�4 Mand 1 9 10�10 M.

Tissue used for contractile studies was stored overnight at 4 °Cin modified Krebs–Henseleit solution and prepared for isometrictension recording the following day as previously described.15,16

Segments were placed in a 20-mL isolated organ bath containingmodified Krebs–Henseleit solution maintained at 37 °C andgassed with 95% O2/5% CO2. After 40 min equilibration seg-ments were placed under ~20 mN resting tension and myogenictone developed over a further 60-min period. Once stablemyogenic tone was established, tissue strips were utilized forcontractile experiments with selective agonists or subject toelectrical field stimulation.

Cumulative concentration curves were constructed using half-logarithm increments of L-erythromethoxamine, norepinephrine,phenylephrine, brimonidine (UK14,304), dexmedetomidine, andlofexidine (isomers and racemate). Brimonidine, dexmedetomi-dine, and lofexidine responses were examined in the presence orabsence of 100 nM prazosin and 100 nM RX-811059: antagonistsselective for a1- and a2-adrenoceptors, respectively.

5,17 The antag-onists were allowed to equilibrate for 30 min prior to constructionof the agonist concentration curve. To assess the potential forsynergy between selective agonists, the effect of preincubationwith 300 nM brimonidine on subsequent concentration-dependent contractions evoked by L-erythromethoxamine wasinvestigated.

Transmural stimulation of the tissue with trains of pulses at 1or 10 Hz (pulse strength 300 mA, pulse width 0.3 ms, 30 s trainsevery 10 min) was achieved using a Digitimer Multistim SystemD330 (Digitimer Ltd, Welwyn Garden City, UK). Tissue segmentswere stimulated at 1 and 10 Hz until stable responses wereobserved and then allowed a period of equilibration with submax-imal concentrations of either brimonidine, phenylephrine, orsodium orthovanadate (a non-adrenoceptor constrictor). In addi-tion, the effect of brimonidine on electrically evoked relaxationsto 1 and 10 Hz transmural stimulation was reexamined in the

© 2014 John Wiley & Sons Ltd1096

S. J. Rayment et al. Neurogastroenterology and Motility

presence of either 100 nM RX-81105917 (to block a2-adrenocep-tors), 30 lM bretylium (to remove the influence of noradrenergicnerves), or 100 lM L-NAME (to remove nitrergic responses).

Data analysis

For competition binding curves, mean pKi values � SEM fortriplicate experiments were calculated. In contractile studies,changes in tension with the addition of drugs were determined asthe maximal increase or decrease in tone and are expressed as apercentage of the initial stable myogenic tone. Individual con-centration curves were constructed for each tissue and pEC50 andRmax values calculated using Graphpad Prism (GraphPad, SanDiego, CA, USA). Differences between mean pEC50 and Rmax

values for dexmedetomidine, brimonidine, and lofexidine in thepresence or absence of antagonists were compared using a one-way ANOVA with Dunnett’s post hoc test and consideredsignificant if p < 0.05.

For the electrical field stimulation experiments, both themagnitude of neurogenic relaxations (expressed as a percentage ofthe myogenic tone prior to the electrical stimulus) and timecourse of response were evaluated. Time course (t50) wasmeasured as previously described.15 The t50 value is calculatedas the time (in seconds) required for the maximal response todecline by 50% following the cessation of EFS. Therefore, apositive value for t50 denotes a response that was sustained for theperiod of stimulation and that declined following cessation ofstimulation, whereas a negative t50 indicates that the relaxatoryresponse was declining prior to the cessation of EFS.

Differences between mean values were compared using apaired Student’s t-test and considered significant if p < 0.05.

Materials

The composition of the modified Krebs–Henseleit saline was(mM) NaCl 118.4, KCl 4.7, CaCl2 1.25, MgSO4 1.2, NaHCO3 24.9,KH2PO4 1.2, glucose 11.1. TE buffer was (mM) Tris 50, EDTA 1,pH 7.4. The drugs used were obtained from Sigma-Aldrich (Poole,UK) with the exception of prazosin hydrochloride (Pfizer,Sandwich, UK), 2-(2-ethoxy-1,4-benzodioxan-2-yl)-2-imidazoline(RX-811059; Reckitts and Coleman, Hull, UK), bretylium tosylate(International Medication Systems, Slough, UK), brimonidinebitartrate (UK14,304; Pfizer), lofexidine for initial contractilestudies was obtained from Macfarlane Smith Ltd (Edinburgh, UK):subsequent studies on individual isomers used lofexidine isomers(+ and �) and its racemic mixture supplied by Norgine (Hengoed,UK). Dexmedetomidine and L-erythromethoxamine were alsosupplied by Norgine. [3H]-RX821002 (2-methoxyidazoxan) and[3H]-prazosin were obtained from GE Healthcare Ltd (LittleChalfont, UK).

RESULTS

Radioligand binding

Competition binding analysis was used to investigate

the affinity of chosen compounds at a1- and a2-adreno-ceptor binding sites on sheep internal sphincter mem-

branes labeled with [3H]-prazosin and [3H]-RX-811059,

respectively. The pKi values generated for competing

agents are summarized in Table 1. All Hill slope values

were close to unity (≥0.8) except for dexmedetomidine

when using [3H]-RX821002 (Hill slope �0.76 � 0.01).

Based on pKi values, the most potent lofexidine agonist

using both radioligandswas (�)-lofexidinewith the rank

order of affinity of (�)-lofexidine > racemic lofexidine >(+)-lofexidine. The ratio of a2/a1 observed for the lofex-

idine isomers was between 15 and 20. When comparing

all agonists tested, (�)-lofexidine was still the agent

with the highest affinity using [3H]-RX821002, with the

rank order of affinity being (�)-lofexidine ≥ dexmede-

tomidine > racemic lofexidine > (+)-lofexidine > brimo-

nidine (UK-14304). Using [3H]-prazosin, the compound

with the highest affinity was (�)-lofexidine with the

rank order of affinity being (�)-lofexidine > racemic

lofexidine > (+)-lofexidine > dexmedetomidine> brimo-

nidine. The a2/a1 ratio for brimonidine and dexmede-

tomidine was 30 and 73, respectively.

Agonist responses

Strips of the sheep isolated internal anal sphincter

developed myogenic tone equivalent to 30–60 mN. All

the a-adrenoceptor agonists examined caused concen-

tration-dependent contractions, with brimonidine, lof-

exidine, and dexmedetomidine causing sustained

contractions (Fig. 1). As shown in Table 2, the rank

order of potency of the agonists was lofexidine >brimonidine > L-erythromethoxamine ≥ dexmedetom-

idine > norepinephrine = phenylephrine, while the rank

order of the maximum responses was norepinephrine >L-erythromethoxamine = phenylephrine > dexmede-

tomidine > lofexidine = brimonidine.

Table 1 Mean pKi values of three experiments (� SEM) of compounds tested at both a1- ([3H]-prazosin) and a2-adrenoceptor ([

3H]-RX-821002) on sheep

internal anal sphincter membranes

pKi (Hill Slope)

[3H]-Prazosin

pKi (Hill Slope)

[3H]-RX-821002

a2/a1Ratio

(+)-lofexidine 6.60 � 0.12 (�1.00 � 0.10) 7.91 � 0.12 (�0.84 � 0.07) 20

(�)-lofexidine 7.13 � 0.11 (�1.07 � 0.10) 8.37 � 0.09 (�0.80 � 0.05) 17

Racemic lofexidine 6.92 � 0.09 (�0.97 � 0.10) 8.16 � 0.06 (�0.81 � 0.03) 17

Brimonidine 5.51 � 0.07 (�0.97 � 0.03) 7.00 � 0.09 (�0.80 � 0.07) 30

Dexmedetomidine 6.46 � 0.10 (�1.00 � 0.10) 8.32 � 0.19 (�0.76 � 0.01) 73


Volume 26, Number 8, August 2014 a2-adrenoceptor function in anal sphincter

The availability of the individual isomers of the

lofexidine prompted a further set of experiments

designed to assess the relative potency of these agents.

The maximal responses of sheep isolated internal anal

sphincter were not significantly different between the

different isoforms of lofexidine, but the pEC50 values

showed an eightfold difference between (+) and (�)

isomers with the racemate having a pEC50 value in

between the two isomers (pEC50 (+)-lofexidine

6.45 � 0.15, n = 16; (�)-lofexidine 7.27 � 0.11,

n = 12; racemate 7.05 � 0.17, n = 14).

Contractile responses in the presence ofantagonists

Fig. 2 shows cumulative concentration curves for

lofexidine, dexmedotomidine, and brimonidine in the

absence and presence of either 0.1 lM prazosin or

A

B

C

Figure 1 Representative digitized trace

recordings of contractions of

(A) brimonidine, (B) lofexidine,

and (C) dexmedetomidine in the sheep

isolated internal anal sphincter.



0.1 lM RX-811059. Contractile responses to lofexidine

(pEC50 6.87 � 0.09, n = 13) were not significantly

altered by the presence of 0.1 lM RX-811059 (pEC50

6.87 � 0.09, n = 14). The maximum response to lofex-

idine in the presence of 0.1 lM prazosin was signifi-

cantly reduced by ~60% (p < 0.01) and the contraction

response curve (pEC50 5.31 � 0.15, n = 9) was shifted

30-fold to the right (see Fig. 2A).

Fig. 2B shows that the presence of 0.1 lMRX-811059

failed to cause a significant change in the concentration

response curve to dexmedetomidine based on one-way

ANOVA with a Dunnett’s post hoc test (Control pEC50

6.54 � 0.17, n = 10; RX81059 pEC50 6.06 � 0.15,

n = 9). The pEC50 value for dexmedetomidine observed

in the presence of 100 nM prazosin was significantly

affected (an approximate 10-fold right shift) compared

to control values (prazosin pEC50 5.55 � 0.26, n = 8;

p < 0.01). No significant differences were observed in

Rmax values under any of the conditions used. Finally,

the concentration response curve to brimonidine was

significantly displaced (a 10-fold shift) by the presence

of 0.1 lM RX-811059 (Control pEC50 6.06 � 0.18,

n = 17; RX811059 4.77 � 0.14, n = 11; p < 0.01), while

0.1 lM prazosin failed to significantly alter either the

potency (Prazosin pEC50 6.48 � 0.11, n = 13) or the

maximum response (Fig. 2C).

In a separate series of experiments, the presence of

300 nM brimonidine caused a significant reduction in

the maximum response to L-erythromethoxamine

(Control 127 � 14%; Brimonidine 86 � 12%, n = 8;

p = 0.04), but failed to alter the potency of the agonist

(Control pEC50 6.09 � 0.10, n = 6; Brimonidine pEC50

6.12 � 0.07, n = 8).

Effect of a-adrenoceptor agonists on neurogenicresponses

Electrical field stimulation at 1 and 10 Hz for 30 s

caused a relaxation of myogenic tone that was abol-

ished by 0.1 lM tetrodotoxin (data not shown). Brimo-

nidine (10 lM) caused a sustained contraction that

reduced the magnitude of the neurogenic relaxation to

1 Hz, but a small increase in the response to 10 Hz. In

Table 2 Summary of the effect of various a-adrenoceptor agonists on

the sheep isolated internal anal sphincter

pEC50 Rmax n

Norepinephrine 5.02 � 0.08 159 � 41% 11

Phenylephrine 5.08 � 0.16 125.1 � 11.2% 12

L-erythromethoxamine 6.14 � 0.06 125 � 22% 11

Dexmedetomidine 6.09 � 0.30 99 � 11% 10

Lofexidine 6.84 � 0.11 61 � 7% 10

Brimonidine 6.40 � 0.18 50 � 14% 7

A

B

C

Figure 2 Cumulative concentration response curves to (A) lofexidine,

(B) dexmedetomidine, and (C) brimonidine in the absence or presence

of 0.1 lM prazosin and 0.1 lM RX-811059. The responses are the

mean � SEM of observations in sheep internal anal sphincter from 8 to

17 different animals.



addition, the duration of the response to 10 Hz in the

presence of 10 lM brimonidine, as determined by the

time required for the myogenic tone return to 50% of

the peak response (t50), was significantly increased

(Fig. 3 and Table 3). In contrast, contractions induced

by 10 lM phenylephrine significantly reduced the

magnitude of neurogenic relaxations to 1 and 10 Hz

(Table 3), but did not affect the duration of the

responses. Sustained contractions induced by the non-

adrenoceptor constrictor sodium orthovanadate (3 mM)

caused similar changes in neurogenic relaxations to

those produced by phenylephrine (Table 3; Fig. 3).

When the concentration of brimonidine was reduced

to 0.3 lM, no significant effect was observed on the

magnitude of response, although the duration of the

response at 10 Hz remained significantly prolonged

(Table 3). Interestingly, 0.3 lM brimonidine failed to

prolong the duration of the neurogenic response to

10 Hz stimulation in the presence of 0.3 lM RX-

811059 (which also reduced the increase in myogenic

tone) and 30 lM bretylium (Table 3).

In a separate series of experiments, the effect of

brimonidine on EFS responses in tissues treated with

100 lM L-NAME was tested. L-NAME inhibits the

release of nitric oxide and in sheep IAS, in addition to

affecting the EFS response, also results in an increase of

basal tone which was maintained after addition of

brimonidine. The effect of L-NAME and the subse-

quent addition of brimonidine on EFS responses at

10 Hz can be seen in Fig. 4 (lower). As can be clearly

Table 3 Changes in magnitude and duration of neurogenic relaxations of SIAS to EFS at 1 and 10 Hz following treatment. Data shown in bold with

asterix denoting values where drug treatment differs significantly from control based on paired student t-tests (* is shown where p < 0.05)

Neurogenic relaxation—% of myogenic tone

Time to 50% of response after cessation of

stimulation t50

1 Hz, 30 s 10 Hz, 30 s 1 Hz, 30 s 10 Hz, 30 s

Control Drug Control Drug Control Drug Control Drug

Brimonidine (10 lM) 54.7 � 6.8 48.5 � 8.1* 62.6 � 5.4 71.1 � 3.5* 7.7 � 0.5 7.3 � 1.5 �6.0 � 2.7 9.0 � 0.9*

Phenylephrine (10 lM) 77.7 � 10.4 45.3 � 8.1* 91.3 � 12.6 73.6 � 7.2* 8.0 � 0.9 6.0 � 1.1 �5.0 � 2.6 �6.0 � 2.4

Sodium Orthovanadate (3 mM) 52.2 � 4.3 26.3 � 5.1* 65.5 � 4.4 59.2 � 3.2* 7.2 � 0.7 4.9 � 1.3 �4.0 � 2.2 �1.9 � 2.0

Brimonidine (0.3 lM) 49.9 � 8.1 53.6 � 6.1 54.4 � 5.0 64.2 � 6.4 3.7 � 1.2 3.5 � 1.5 �8.2 � 2.2 0.7 � 1.9*

Brimonidine (0.3 lM)

RX-811059 (0.3 lM)

47.7 � 9.0 56.4 � 8.5 55.1 � 7.4 48.3 � 9.2 8.0 � 1.0 3.5 � 2.7 �7.7 � 3.3 �6.4 � 4.0

Brimonidine (0.3 lM)

Bretylium (30 lM)

42.1 � 7.1 50.9 � 7.1 57.3 � 6.2 64.8 � 6.3 3.0 � 1.9 4.0 � 2.5 0.5 � 2.3 1.2 � 2.6

A

B

CFigure 3 Representative digitized trace

recordings showing the response of the

sheep internal anal sphincter to electrical

field stimulation to 1 and 10 Hz (30 s,

0.3 ms, and 300 mA) before and after

exposure to (A) 3 mM sodium

orthovanadate, (B) 0.3 lM brimonidine

(UK14304), or (C) a combination of 0.3 lMbrimonidine (UK14304) and 0.3 lMRX811059.



seen in the trace, the addition of L-NAME resulted in

the conversion of the relaxation observed in control

tissues (relaxation 76.1 � 2.5% of basal tone; n = 9) to

a contraction (contraction 80.6 � 11.5% of basal tone;

n = 9). In the presence of brimonidine, this contraction

was significantly reduced compared with that observed

after the addition of L-NAME based on a Wilcoxon’s

matched pairs test (31.3 � 8.8%; n = 9; p = 0.0039).

A graphical representation of this data can be seen in

Fig. 4 (upper).

DISCUSSION

The starting point for this study was a desire to better

understand the potential value of existing imidazoline

derivatives, with known activity at a-adrenoceptors, aspossible drugs to influence anal sphincter tone and

fecal incontinence.

The radioligand binding studies revealed that all

three imidazoline-based agonists, dexmedetomidine,

lofexidine, and brimonidine, possessed greater selec-

tivity for anal sphincter a2-adrenoceptor binding sites

compared with a1-adrenoceptor binding sites, as iden-

tified by [3H]-RX821102 and [3H]-prazosin, respec-

tively. The degree of selectivity ranged from between

17-fold (lofexidine) and 73-fold (dexmedetomidine),

making these imidazoline derivatives comparable to,

or better than, those previously examined in this

tissue.5 Significantly, the pKi values observed with

dexmedetomidine in the sheep internal anal sphincter

were comparable to those previously determined in

Chinese hamster ovary cells transfected with human

a1- or a2-adrenoceptors and labeled with [3H]-prazosin

and [3H]-RX821002,18 respectively. Similarly, a2-adre-noceptor affinities for both brimonidine and dexmede-

tomidine in the sheep internal anal sphincter were

similar (but slightly lower) than published values for

the human cerebral cortex.19 With respect to lofexi-

dine, the rank order of affinity for the isomers at anal

sphincter binding sites labeled by [3H]-RX821102,

(�)-lofexidine > (�)-lofexidine > (+)-lofexidine, is sim-

ilar to that reported for the centrally mediated hypo-

tensive action of the imidazoline.20

However, comparison between a2-adrenoceptoraffinity and potency as constrictor agents (Tables 1

and 2) in internal anal sphincter revealed a 30- to 200-

fold discrepancy for lofexidine and dexmedetomidine,

respectively, yet only a fourfold discrepancy for brimo-

nidine. In the case of (�)-lofexidine, however, this

isomer is the most potent (by threefold) imidazoline

derivative examined in this tissue.5 Despite the greater

selectivity of these derivatives for a2-adrenoceptorbinding sites in the sheep internal anal sphincter,

contractions elicited by lofexidine and dexmedetomi-

dine were sensitive to low concentrations of prazosin

and largely unaffected by the selective a2-adrenoceptorantagonist RX811059.17 Thus, the activity of these

agonists is comparable to that previously reported for

clonidine in this tissue5 and consistent with the

finding that the maximum concentrations were less

than those elicited by either norepinephrine, phenyl-

ephrine, or L-erythromethoxamine. In marked con-

trast, contractions to brimonidine exhibited a different

profile with greater sensitivity to RX-811059 than

to prazosin, suggesting a major role for anal sphincter

a2-adrenoceptors.The most likely explanation for the difference

between lofexidine and dexmedetomidine, on the one

hand, and brimonidine, on the other, is that the latter

agent is known to be a full agonist at a2-adrenocep-tors.21,22 Furthermore, the activity of both dexmede-

tomidine and lofexidine at a1-adrenoceptors is not

without precedent. In a rat vascular model, contractile

responses to 30 nM–10 lM dexmedetomidine were

inhibited by the presence of prazosin.23 However, in

vivo administration of prazosin to rats attenuated the

Figure 4 (Upper) A graph representing the changes observed in

electrical field stimulation (EFS) responses in nine individual sheep

internal anal sphincter (IAS) tissue strips at 10 Hz (30 s, 0.3 ms, and

300 mA) for Control, L-NAME (100 lM), and L-NAME (100 lM) with

brimonidine (0.3 lM) sequentially added. (Lower) A representative

digitized trace recording showing effect of sequential addition of

L-NAME (100 lM) and brimonidine (0.3 lM) on EFS responses at 1 Hz

(period of stimulation shown by the black bar) and 10 Hz (period of

stimulation shown as a red bar) on a segment of the sheep IAS.



increase in diastolic pressure arising from acute admin-

istration of lofexidine.14 These observations highlight

that a2-adrenoceptors present on the internal anal

sphincter can mediate significant contractions, but

only when activated by a highly efficacious agonist.

The observation that contraction of IAS smooth

muscle by brimonidine is mediated predominantly

through a2-adrenoceptors raises the possibility that

both a-adrenoceptor subtypes may reside on the same

cells, with the potential for synergistic, constrictor

interaction.24 However, when we investigated the

effect of a ‘selective’ constrictor concentration of

brimonidine (0.3 lM) on response to L-erythromethox-

amine (a1-adrenoceptor agonist), we found no evidence

for enhancement of contractions.

In addition to being present on smooth muscle cells,

a2-adrenoceptors are also known to reside on nerve

endings and modulate neurotransmission.25 In the anal

sphincter, nitric oxide and norepinephrine are two

major neurotransmitters that mediate opposing effects

on myogenic tone.5,26 Using brimonidine, we estab-

lished that unlike other constrictor agents, this agonist

does not directly influence nitrergic neurotransmission

at low frequencies of electrical field stimulation, but

may indirectly enhance relaxatory response at higher

frequencies, by inhibiting the release of norepinephr-

ine. Confirmation of a role for a2-adrenoceptors was

provided by the finding that when the receptor was

inhibited by RX-810059,17 or when norepinephrine

release was prevented by bretylium, this indirect effect

on nitrergic neurotransmission was not observed.

The reduction in contractile responses observed in

L-NAME-treated tissue following application of brimo-

nidine lends further weight to this argument. The

finding that a2-adrenoceptors may be present both on

the sphincter muscle and on nerves may explain the

observation in the anesthetized opossum that naphaz-

oline (a mixed a1-/a2-adrenoceptor agonist) both

increased internal anal sphincter pressure and modu-

lates the recto-anal inhibitory response.5,27 Indeed, the

ability of a2-adrenoceptor agonists to inhibit neurogenic

constrictor responses in the internal anal sphincter is

similar to that reported for cholinergic contractions in

rat colonic segments25 and guinea-pig ileum.28

As far as we are aware, there have been no detailed

studies regarding the role of a2-adrenoceptors in the

rectum in vitro, yet this could be the ideal site for drugs

to inhibit constrictor influences if the therapeutic

objective is to enhance rectal compliance and delay

defecation. As previously shown by Malcolm et al.,29

orally administered clonidine (a partial agonist at

a2-adrenoceptors) increased rectal compliance in

human subjects. This finding raises the possibility

that local application of imidazoline derivatives may

produce a similar effect.30 As defecation is a physio-

logical event involving a coordinated response of the

rectum and anal sphincter, the possibility has to be

considered that a combination of agents may achieve a

better patient outcome than either agent alone. A

detailed study of the role of rectal a-adrenoceptorsubtypes appears to be warranted.

This study has provided evidence for a dual role for

a2-adrenoceptors on the sheep isolated internal anal

sphincter: direct constrictor effect and modulator of

noradrenergic transmission. While potent and selective

imidazoline derivatives caused a contraction of the

anal sphincter, only brimonidine (a highly efficacious

agonist) produced an unambiguous contraction via

a2-adrenoceptors. As the magnitude of these contrac-

tions is smaller than that mediated via a1-adrenocep-tors, it seems unlikely that imidazoline derivatives

could be used alone in a topical formation for the

treatment of fecal incontinence.

ACKNOWLEDGMENTS

The authors would like to thank M Najib and Sons at Derby forthe supply of animal tissue used in these experiments. Theauthors had access to all data relating to this manuscript and haveno competing interests.

FUNDING

The authors would like to thank Norgine Ltd for financialsupport.

DISCLOSURE

No competing interest declared.

AUTHOR CONTRIBUTION

SR, JS, TE, and AA performed the research; SR, MD, and VWanalyzed the data; JS, VW, HG, and YH designed the study; SR andVWwrote the manuscript; all authors contributed to proof readingand revising the manuscript.

REFERENCES

1 Scholefield JH, Bock JU,Maria B, Rich-ter HJ, Athanasiadis S, ProisM,Herold

A. A dose finding study with 0.1%,0.2% and 0.4%glyceryl trinitrate oint-ment in patients with chronic analfissures.Gut 2003; 52: 264–9.

2 Cook TA, Humphreys MM, McCMortensen NJ. Nifedipine for thetreatment of anal fissures. Br J Surg

2000; 86: 1269–73.



3 Cheetham MJ, Kamm MA, PhillipsRK. Topical phenylephrine increasesanal canal resting pressure in patientswith faecal incontinence. Gut 2001;48: 356–9.

4 Nisar P, Gruss HJ, Bush D, Barras N,Acheson A, Scholefield JH. Intra-analand rectal application of L-erythro-methoxamine gel increases anal rest-ing pressure in healthy volunteers. BrJ Surg 2005; 92: 1539–45.

5 Rayment SJ, Eames T, Simpson JAD,Dashwood MR, Henry Y, Gruss H,Acheson AG, Scholefield JH et al.

Investigation of the distribution andfunction of alpha-adrenoceptors in thesheep isolated internal anal sphincter.Br J Pharmacol 2010; 160: 1727–40.

6 Mills KA, Hausmann N, Chess-Wil-liams R. Characterization of thealpha1-adrenoceptor subtype mediat-ing contraction of the pig internalanal sphincter. Br J Pharmacol 2008;155: 110–7.

7 Lowenthal DT, Matzek KM, MacGr-egor TP. Clinical pharmacokinetics ofclonidine. Clin Pharmacokinet 1988;14: 287–310.

8 Barbieri A, Santagostini-Barbone MG,Zonta F, Lucchelli A. Pharmacologi-cal characterization of alpha-adreno-ceptors mediate contraction of thesplenic artery strips from the pig.Naunyn Schmiedeberg’s Arch Phar-macol 1998; 357: 654–61.

9 Mason KP, Lerman J. Review article:Dexmedetomidine in children - cur-rent knowledge and future applica-tion. Anesth Analg 2011; 113: 1129–42.

10 Wikberg-Matsson A, Simonsen H.Potent alpha(2A)-adrenoceptor-medi-ated vasoconstriction by brimonidinein porcine ciliary arteries. Invest

Ophthalmol Vis Sci 2001; 42: 2049–55.

11 Fung AT, Reid SE, Jones MP, HealeyPR, McCluskey PJ, Craig JC. Meta-analysis of randomised controlled tri-als comparing latanoprost withbrimonidine in the treatment ofopen-angle glaucoma ocular hyper-tension or normal tension glaucoma.Br J Opthalmol 2007; 91: 62–8.

12 Gish EC, Miller JL, Honey BL,Johnson PN. Lofexidine, an alpha2receptor agonist for opioid detoxifica-tion. Ann Pharmacother 2010; 44:343–51.

13 Biedermann J, Le�on-Lomel�ı A, BorbeHO, Prop G. Two stereoisomeric im-idazoline derivatives: synthesis andoptical and a2-adrenoceptor activities.J Med Chem 1986; 29: 1183–8.

14 Willffert B, Mathy MJ, Batink HD,de Jonge A, Thoolen MJ, Prop G, GrafE, Timmermans PB et al. Interferenceof enantiomers of lofexidine withalpha-adrenoceptors. Arch Int Phar-

macodyn Ther 1985; 273: 18–32.15 Acheson A, Rayment S, Eames T,

Mundey M, Nisar P, Scholefield J,Wilson VG. Investigation of the roleof adrenergic and non-nitrergic, non-adrenergic neurotransmission in thesheep isolated internal anal sphincter.Neurogastroenterol Motil 2009; 21:335–45.

16 Mundey MKM, Jonas M, Worthley T,Scholefield JH, Wilson VG. Pharma-cological characterization of the neu-rogenic response of the sheep isolatedinternal anal sphincter. Br J Pharma-

col 2000; 130: 489–94.17 Mallard NJ, Hudson AL, Nutt DJ.

Characterization and autoradiograph-ical localization of non-adrenoceptor,idazoxan binding sites in the ratbrain. Br J Pharmacol 1992; 106:1019–27.

18 MacLennan SJ, Reynen PH, MartinRS, Eglen RM, Martin GR. Character-ization of human recombinant alpha(2A)-adrenoceptors expressed in Chi-nese hamster lung cells using extra-cellular acidification rate changes. BrJ Pharmacol 2000; 129: 1333–8.

19 Miralles A, Olmos G, Sastre M,Barturen F, Martin I, Garcia-SevillaJA. Discrimination and pharmacolog-ical characterization of I2-imidazo-line sites with [3H]-Idazoxan andalpha-2 adrenoceptors with [3H]RX821002 (2-methoxy Idazoxan) inthe human and rat brains. J Pharm

Exp Ther 1993; 264: 1187–97.20 Fagan TC, Bloomfield SS, Cowart TD,

Corns-Hurwitz RH, Lipicky RJ, Con-radi EC, Hsu CY, Grossman WJ et al.Antihypertensive effects of lofexidinein patients with essential hyperten-sion. Br J Clin Pharmacol 1982; 13:405–10.

21 Jansson CC, Kukkonen JP, Nasman J,Huifang G, Wurster S, Virtanen R,Savola JM, Cockcroft V et al. Proteanagonism at alpha2A-adrenoceptors.Mol Pharmacol 1998; 53: 963–8.

22 MacLennan SJ, Luong LA, Jasper JR,To ZP, Eglen RM. Characterization ofalpha2-adrenoceptors mediating con-traction of dog saphenous vein: iden-tity with human alpha2A subtype. BrJ Pharmacol 1997; 121: 1721–9.

23 Wong ES, Man RY, Vanhoutte PM,Ng KF. Dexmedetomidine inducesboth relaxations and contractions,via different a2-adrenoceptor sub-types, in the isolated mesentericartery and aorta of the rat. J Pharma-col Exp Ther 2010; 335: 659–64.

24 Daly CJ, McGrath JC, Wilson VG.Pharmacological analysis of post-junctional a-adrenoceptors mediatingcontractions to (-)-noradrenaline inthe rabbit isolated saphenous veincan be explained by interactingresponses to simultaneous activationof a1- and a2-adrenoceptors. Br J Phar-macol 1988; 95: 485–500.

25 Blandizzi C, Fornai M, Colucci R,Baschiera G, Barbera G, De GiorgioR, De Ponti F, Del Breschi MC et al.

Altered prejunctional modulation ofintestinal cholinergic and noradrener-gic pathways by alpha2-adrenoceptorsin the presence of experimental coli-tis. Br J Pharmacol 2003; 139: 309–20.

26 Glavind EB, Forman A, Tøttrup A.Effects of transmural field stimula-tion in isolated smooth muscle ofhuman rectum and internal analsphincter. Am J Physiol 1997; 272:G1075–82.

27 Yamato S, Rattan S. Role of alphaadrenoceptors in opossum internalanal sphincter. J Clin Invest 1990;86: 424–9.

28 Herbert MK, Roth-Goldbrunner S,Holzer P, Roewer N. Clonidine anddexmedetomidine potently inhibitperistalsis in the guinea-pig ileum invitro. Anaesthesiology 2002; 97:1491–9.

29 Malcolm A, Camilleri M, Kost L,Burton DD, Fett SL, ZinsmeisterAR. Towards identifying optimaldoses for alpha-2 adrenergic modula-tion of colonic and rectal motor andsensory function.Aliment Pharmacol

Ther 2000; 14: 783–93.30 Bharucha AE, Seide BM, Zinsmeister

AR. Effect of clonidine on symptomsand anorectal sensorimotor functionin women with fecal incontinence.Aliment Pharmacol Ther 2010; 32:681–8.



Documents

Dual effects of α 2 -adrenoceptors in modulating myogenic tone in sheep isolated internal anal sphincter