British journal of pharmacology 2015 volume 172 part 5

This antiproliferative action was only present in H4 receptor-expressing colorectal cancer cell lines, but not inmock-transfected cells and could be prevented by pretreat-ment with the s

the gastrointestinal tract

A Deiteren1, J G De Man1, P A Pelckmans1,2and B Y De Winter1

1Laboratory of Experimental Medicine and Pediatrics, Division of Gastroenterology, University of

Antwerp, Antwerp, Belgium, and2Department of Gastroenterology and Hepatology, Antwerp

University Hospital, Antwerp, Belgium

Correspondence

Benedicte Y De Winter,University of Antwerp, CampusDrie Eiken, Laboratory ofExperimental Medicine andPediatrics, Division ofGastroenterology,Universiteitsplein 1, B-2610Antwerp, Belgium E-mail:

Histamine is a well-established mediator involved in a variety of physiological and pathophysiological mechanisms and exerts

histamine receptor family, and is expressed throughout the gastrointestinal tract as well as in the liver, pancreas and bile

inflammation such as in colitis, ischaemia/reperfusion injury, radiation-induced enteropathy and allergic gut reactions In

various gastrointestinal disorders such as inflammatory bowel disease, irritable bowel syndrome and cancer

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://

www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are

permanently archived in the Concise Guide to PHARMACOLOGY 2013/14 (a,b Alexander et al., 2013a,b).

British Journal of Pharmacology (2015) 172 1165–1178 1165

© 2014 The British Pharmacological Society

Histamine (2-[4-imidazole]-ethylamine) is a short-acting

endogenous amine, involved in several physiological and

pathophysiological processes (Jutel et al., 2009) It is present

in virtually all bodily organs, with high concentrations

reported in the stomach, lymph nodes and thymus (Kumar

et al., 1968; Zimmermann et al., 2011) Histamine is

synthe-tized from L-histidine by L-histidine decarboxylase and is

stored in the granules of mast cells (MCs) and basophils, the

main sources of histamine (Endo, 1982; Jones and Kearns,

2011) Enterochromaffin-like cells, histaminergic neurons,

lymphocytes, monocytes, platelets and neutrophils also

express L-histidine decarboxylase and are capable of

produc-ing, but not storproduc-ing, high amounts of histamine (Snyder and

Epps, 1968; Vanhala et al., 1994; Bencsath et al., 1998; Jutel

et al., 2009; Alcaniz et al., 2013) Histamine exerts its actions

by binding to four GPCRs that are differentially expressed

throughout the body and designated as the H1, H2, H3and H4

receptors Histamine H1receptors mediate sensorineural

sig-nalling, vascular dilatation and permeability and airway

smooth muscle contraction, and are involved in allergic

rhi-nitis, atopic dermatitis, conjunctivitis, urticaria, asthma and

anaphylaxis (Togias, 2003; Simons and Simons, 2011)

Hista-mine H2receptors are well-known for their role in gastric acid

secretion, but also exert immune modulatory properties

(Black et al., 1972; Jutel et al., 2009) Histamine H3receptors

are most abundantly present in the CNS and are implicated in

sleep–wake disorders, attention-deficient hyperactivity

disor-der, epilepsy, cognitive impairment and obesity (Kuhne et al.,

2011; Singh and Jadhav, 2013) Finally, histamine H4

recep-tors are predominantly expressed on immune cells, such as

lymphocytes, MCs and dendritic cells, and are currently

mainly under evaluation for immune-mediated disorders

such as allergic rhinitis, asthma and pruritus (Liu, 2014)

However, new roles for this receptor subtype are continuously

being discovered Here we provide an overview of the current

evidence of H4receptor involvement in multiple

gastrointes-tinal physiological and pathophysiological processes

H4 receptors

In the early 2000s, several groups reported on the discovery

and cloning of a fourth histamine receptor (Nakamura et al.,

2000; Oda et al., 2000; Liu et al., 2001a; Morse et al., 2001;

Nguyen et al., 2001; Zhu et al., 2001) The H4 receptor is

encoded by a single copy on chromosome 18q11.2 and

dem-onstrates an overall homology of 23% to H1receptors, 22% to

H2receptors and 37% to H3receptors (Oda et al., 2000; Coge

et al., 2001) The human full-length receptor consists of 390

amino acids, which form seven transmembrane helices, three

extracellular loops and three intracellular loops, with an

extracellular N-terminal and an intracellular C-terminal

peptide (Leurs et al., 2009) H4 receptors couple to Gαi/0

pro-teins, inhibiting downstream adenylyl cyclase and

forskolin-induced cAMP (Morse et al., 2001; Zhu et al., 2001) They are

mainly present in immune cells and highly expressed in bone

marrow and spleen; varying expression levels were also

reported in gastrointestinal tissues, testes, kidney, lung,

pros-trate and brain (Nakamura et al., 2000; Oda et al., 2000; Coge

et al., 2001; Strakhova et al., 2009) Tissue distribution is quite

similar across species (Liu et al., 2001b; Oda et al., 2005).

There is high homology in the amino acid sequence betweenhuman and monkey H4receptors (92%), whereas this is 72%between human and pig and 65–70% between human androdent H4receptors (Liu et al., 2001b; Oda et al., 2002; 2005).

These differences in amino acid sequence also affect thebinding profile of histamine towards H4receptors with highaffinity for human and guinea pig H4 receptors (KD4.8 and

6 nM) compared with rat and mouse H4 receptors (136 and

42 nM) (Liu et al., 2001b) Compared with H1and H2tors, histamine displays high affinity for H4receptors in bothhuman and rodents (Table 1)

recep-Soon after its discovery and cloning, attempts were made

to elucidate the pharmacological profile of H4receptors andidentify (selective) ligands to stimulate or inhibit H4receptorsignalling Early assessments indicated that several H3recep-tor ligands demonstrated significant affinity for H4receptors,such as clozapine, imetit and immepip (H3 and H4 receptoragonists) and clobenpropit (H3receptor antagonist, H4recep-

tor agonist) (Table 1) (Leurs et al., 2009; Smits et al., 2009).

Table 1

Ligands for the human H4receptor

Compound

H 4 R (pKi)

H 1 R (pKi)

H 2 R (pKi)

H 3 R (pKi)

et al (2004; 2014) Of note, ligand affinity may differ among

species Data presented as Kivalue (nM) for the human mine H4and H3receptors

hista-BJP A Deiteren et al.

1166 British Journal of Pharmacology (2015) 172 1165–1178

Since then, several new compounds have been developed

targeting H4receptors such as 4-methylhistamine, VUF8430

and OUP-16 (selective agonists) and A-943931, JNJ7777120

and VUF6002 (selective antagonists; Table 1) (Leurs et al.,

2009; Smits et al., 2009) However, it was recently

demon-strated that in addition to inhibition of Gαi/0proteins, many

H4receptor antagonists can also exert a partial agonist effect

at certain species H4 receptor orthologues via β-arrestin

recruitment and ERK activation (Rosethorne and Charlton,

2011), which may contribute to some of the species

differ-ences that have been reported for H4 receptor ligands (Liu

et al., 2001b; Seifert et al., 2011; Nijmeijer et al., 2013; Salcedo

et al., 2013).

Histamine and histamine receptors in

the gastrointestinal tract

In the gastrointestinal tract, histamine participates in

multi-ple physiological processes among which immunological

responses, visceral nociception, modulation of intestinal

motility and gastric acid secretion (Black et al., 1972; Poli

et al., 2001; Dawicki and Marshall, 2007; Takagaki et al., 2009;

Simon et al., 2011; van Diest et al., 2012) Histamine is also

involved in several gastrointestinal disorders such as

inflam-matory bowel diseases (IBD), irritable bowel syndrome (IBS),

malignancies, systemic mastocytosis, food allergy and gastric

ulcers (Black et al., 1972; He, 2004; Wood, 2004; Barbara et al.,

2006; Sokol et al., 2010; Kennedy et al., 2012) All four

hista-mine receptors are expressed in the gastrointestinal tract,

although the presence of H3 receptors in the human gut

remains controversial (Poli et al., 2001) Human H1receptors

are abundantly expressed throughout the gastrointestinal

tract on enterocytes as well as connective tissue cells,

immune cells, blood vessels, myocytes and enteric nerves

(Sander et al., 2006) H2receptors are present on gastric

pari-etal cells, enterocytes, immunocytes such as lymphocytes,

myenteric ganglia and smooth muscle cells (Fukushima et al.,

1999; Sander et al., 2006) H3 receptors were reported to be

expressed in gastrointestinal tissue of guinea pig and

func-tional data located them on nerve terminals in the myenteric

plexus and on pre- and post-ganglionic cholinergic and

non-adrenergic, non-cholinergic fibres (Poli et al., 2001) However,

human intestine seems to be devoid of H3 receptors

(Hemedah et al., 2001; Poli et al., 2001; Cianchi et al., 2005;

Sander et al., 2006).

Using a variety of techniques, several groups

demon-strated expression of H4 receptors throughout the

gastroin-testinal tract and in the pancreas, liver and bile ducts, not

only in humans, but also in other species such as rodents,

pigs, dogs and monkeys (Table 2) Sander et al (2006)

reported similar distribution of H4receptors along the human

duodenum, colon, sigmoid and rectum More specifically, H4

receptors were present on lamina propria mononuclear cells

and intestinal MCs, on leucocytes in mucosal and

submu-cosal blood vessels and to a lesser extent on tissue resident

leucocytes In addition, H4 receptor immunoreactivity was

seen in intraepithelial cells considered to be neuroendocrine

cells, in myenteric ganglion cell somata and neuronal fibres,

and on enterocytes in the crypt of Lieberkühn (Sander et al.,

2006; Chazot et al., 2007) Expression of H4 receptors oncolonic enterocytes was later confirmed by others, who alsoreported limited staining of non-specified submucosal and

connective tissue cells (Boer et al., 2008; Fang et al., 2011) A

caveat must be made when interpreting data obtained byimmunohistochemistry Recently the selectivity of commer-cially available H4 receptor-antibodies was questioned asseveral of these antibodies failed to yield a specific signalwhen evaluated in transfected or H4 receptor−/− cells

(Beermann et al., 2012).

Interestingly, gastrointestinal H4 receptor expression isaltered in several disease states Decreased H4receptor expres-sion was reported in gastric cancer specimens, whereas over-expression was demonstrated in cholangiocarcinoma andboth enhanced and decreased expression levels have been

reported in colorectal cancer (Cianchi et al., 2005; Boer et al., 2008; Fang et al., 2011; Meng et al., 2011; Francis et al., 2012; Zhang et al., 2012) Colonic inflammation seems to enhance

H4receptor expression as in two experimental models of IBD,namely murine trinitrobenzene sulphonic acid (TNBS)-induced colitis and spontaneous colitis in Gi α2 protein-deficient mice, active inflammation was associated with anincrease in colonic H4 receptor mRNA (Sutton et al., 2008; Kumawat et al., 2010) Also after complete resolution of

TNBS-colitis, colonic H4 receptor mRNA levels remained

increased (Deiteren et al., 2014) However, in colonic biopsies

of IBS patients with concomitant food allergy, no alterations

in H4 receptor mRNA levels were reported (Sander et al.,

in the gastrointestinal tract, their immune modulatory erties have been studied using models of colitis, ischaemia/reperfusion injury and allergic gut reactions (Table 3).MCs, an important source of gastrointestinal histamine,are key players of both the innate and adaptive immunesystems and congregate at the interface between the internaland external milieu (such as the gut mucosa), where theyexert immune modulatory effects Alterations in MCnumbers and activation state with excessive release of hista-

prop-mine have been reported in patients with IBD (Knutson et al., 1990; Bischoff et al., 1996; Farhadi et al., 2007) Moreover,

treatment with the MC stabilizer ketotifen prevented cally induced colitis in animal models and improved diseaseactivity in a small group of IBD patients; however, the under-lying mechanism of action was not investigated further

chemi-(Eliakim et al., 1992; Jones et al., 1998; Marshall and Irvine, 1998; Fogel et al., 2005) Ketotifen stabilizes MCs (in addition

to H1 receptor antagonist properties), and thus inhibits therelease of histamine in the gut; this may indirectly benefi-cially affect H4 receptor-mediated pathways activated byhistamine

British Journal of Pharmacology (2015) 172 1165–1178 1167

Immunofluorescence Mucosal cells Zhang et al (2012)

Rat Immunohistochemistry Ganglion cell somata and neuronal fibres in the

myenteric but not the submucous plexus;

A-like cells in the fundic epithelium

Chazot et al (2007); Morini et al (2008)

Duodenum

Small intestine

RT-PCR Coge et al (2001); Nakamura et al (2000); Oda et al (2000)

Rat Immunohistochemistry Ganglion cell somata and neuronal fibres in the

myenteric plexus

Chazot et al (2007)

Colon

RT-PCR Lamina propria mononuclear cells and MCs,

mucosa

Boer et al (2008); Cianchi et al (2005); Fang et al (2011); Oda

et al (2000); Sander et al (2006)

Western blot Mucosa Boer et al (2008); Fang et al (2011)

Immunohistochemistry Neuroendocrine-like cells, lamina propria,

intravascular granulocytes, enterocytes, non-epithelial mucosal cells, submucosal connective tissue cells

Boer et al (2008); Fang et al (2011); Sander et al (2006)

Immunohistochemistry Ganglion cell somata and neuronal fibres in the

myenteric, but not the submucous plexus

Chazot et al (2007)

Monkey RT-PCR Longitudinal muscle Kim et al (2011); Oda et al (2005)

Pancreas

Liver

Bile ducts

Immunohistochemistry Cholangiocytes Francis et al (2012); Meng et al (2011)

A caveat must be made when interpreting H 4 receptor expression data obtained by immunohistochemistry: recently the selectivity of commercially available antibodies for the H 4 receptor was questioned as several of these antibodies failed to yield a specific signal when evaluated in transfected or H 4 receptor −/− cells

(Beermann et al., 2012).

BJP A Deiteren et al.

1168 British Journal of Pharmacology (2015) 172 1165–1178

The selective H4 receptor antagonists JNJ7777120 and

JNJ10191584 and the H3/H4 receptor antagonist

thiopera-mide also reduced chemically induced colitis in different rat

models for IBD (Fogel et al., 2005; 2007; Varga et al., 2005;

Dunford et al., 2006b) More specifically, treatment with

these antagonists reduced macroscopic colonic injury,

neu-trophil influx and myeloperoxidase levels (a marker for

myeloid cell infiltration) (Fogel et al., 2005; 2007; Varga et al.,

2005; Dunford et al., 2006b) This is in line with previous

evidence demonstrating that blockade of H4 receptors

impedes neutrophil recruitment and cytokine release in other

models of inflammation, such as zymosan-induced pleuritis

and allergic airway inflammation (Takeshita et al., 2003;

Thurmond et al., 2004; Dunford et al., 2006a) The

anti-inflammatory effect of H4receptor antagonism resulted – at

least partly – from inhibition of aberrant Toll-like receptor

signalling via dendritic cells leading to reduced production of

TNF-α and IL-6 (Fogel et al., 2005; Varga et al., 2005; Dunford

et al., 2006b) In addition, colonic H4receptor expression was

reported to be increased in the colon of mice with

TNBS-induced colitis and during spontaneous colitis in Gαi2

protein-deficient mice (Sutton et al., 2008; Kumawat et al.,

2010) Whether H4 receptor expression is also increased in

IBD patients is an interesting question that has not been

investigated to our knowledge

Data have also emerged, suggesting a possible role for H4

receptors in mediating gastrointestinal inflammation in

ischaemia/reperfusion models However, in most of these

studies non-selective antagonists were used, making it cult to ascertain that this effect was indeed solely mediated bythe H4 receptor subtype In a mouse model of mesentericischaemia/reperfusion injury treatment with the H3/H4recep-tor antagonist thioperamide significantly reduced myeloper-

diffi-oxidase activity (Ghizzardi et al., 2009) In contrast, the

opposite effect was seen on hepatic ischaemia/reperfusiondamage: histamine, the H2/H4receptor agonist dimaprit andthe H3/H4 receptor agonist clozapine reduced post-ischemicliver damage, as shown by a reduction in serum transami-

nases (Adachi et al., 2006) This protective effect was

abol-ished by the H3/H4 receptor antagonist thioperamide butremained unaffected by the selective H2receptor antagonistcimetidine, suggesting a beneficial influence of H4 receptorstimulation in the prevention of ischaemia/reperfusion liverdamage Recently, the mechanism of action was further elu-

cidated by El-Mahdy et al (2013) They found that liver

damage was significantly reduced by pretreatment with tamine, remained unaffected by a selective H1or H2receptorantagonist, was abolished by the H3/H4 receptor antagonistthioperamide and was reproduced by the H3/H4 receptoragonist clozapine The protective effect of histamine andclozapine was mediated by attenuating TNF-α and IL-12secretion and consequently reduced reactive oxygen species

his-(El-Mahdy et al., 2013) As H3 receptors were absent from

adult mouse liver tissue (Heron et al., 2001), it seems

reason-able to assume that the protective effect of histamine andclozapine was indeed mediated by H4receptors However, it is

Table 3

Preclinical in vivo experiments with H4receptor ligands in models of inflammation

In vitro/

TNBS-induced colitis Rat In vivo JNJ7777120

JNJ10191584

JNJ7777120 and JNJ10191584 reducedTNBS-induced colitis

Varga et al (2005)

TNBS-induced colitis Rat In vivo JNJ10191584 JNJ10191584 reduced TNBS-induced colitis Dunford et al.

(2006b)TNBS-induced colitis Rat In vivo Thioperamide Thioperamide reduced TNBS-induced colitis Fogel et al (2007)

Thioperamide Thioperamide reversed the protective effect of

histamine and dimapritIschaemia/reperfusion

histamineRadiation-induced small

important to exclude the possibility that hepatic ischaemia/

reperfusion does not induce H3receptor expression to be sure

that the effect is due to H4receptor modulation

Pronounced gastrointestinal inflammation is seen after

radiation and results from reactive oxygen/nitrate species,

apoptosis and clonogenic cell death, mucosal breakdown and

transcription of proinflammatory cytokines, chemokines and

growth factors (Francois et al., 2013) In view of the

promis-ing results of H4receptor blockade on gastrointestinal

inflam-mation in other animal models, Martinel Lamas et al (2013)

evaluated the radioprotective potential of JNJ7777120, a

selective H4 receptor antagonist Preventive treatment with

JNJ7777120 preserved the villi and the number of crypts in

the small intestine and diminished mucosal atrophy after

radiation by reducing apoptosis and DNA damage in

entero-cytes (Martinel Lamas et al., 2013).

Finally, preliminary evidence also points towards a

possi-ble involvement of H4receptors in allergic gut reactions (Yu

et al., 2008) Actively sensitized guinea pigs were exposed to

inhaled 0.1% ovalbumin; MC and eosinophil infiltration into

the oesophagus was assessed 1 h later Pretreatment with the

H3/H4receptor antagonist thioperamide inhibited migration

of both cell types to the oesophageal epithelium (Yu et al.,

2008) As both MCs and eosinophils did not express H3

recep-tors, the effect was ascribed to blockade of H4 receptors,

which seems consistent with previous reports of H4

receptor-mediated chemotaxis of these cell types (Hofstra et al., 2003;

Thurmond et al., 2004; Yu et al., 2008).

In conclusion, these in vivo experiments suggest that H4

receptors participate in mediating gastrointestinal

inflamma-tion and immune responses in a variety of animal models

These findings are in line with previous preclinical

observa-tions from immune-mediated disorders in other organ

systems and underline the immunomodulatory role of H4

receptors However, further research confirming these

find-ings using highly selective ligands for H4receptors are much

needed before clinical trials can be initiated for

gastrointesti-nal inflammation and immune-mediated disorders

H4 receptors and carcinogenesis

Enhanced expression of L-histidine decarboxylase and high

histamine producing and secreting capabilities have been

reported in malignancies, such as melanoma, breast,

colorec-tal and pancreatic carcinoma both in experimencolorec-tal models

and in human tumour biopsies (Medina and Rivera, 2010;

Kennedy et al., 2012) Histamine, released by the malignant

cells themselves or by other histamine-secreting cells in the

environment such as MCs, acts as a growth factor in an

autocrine or paracrine fashion, regulating angiogenesis, cell

invasion, migration, differentiation, apoptosis and immune

suppression (Medina and Rivera, 2010) These results suggest

an important role for histamine in tumour development and

progression Histamine-induced cell proliferation seems to be

mediated via H2receptors as antagonists for these receptors

induced apoptosis in human colorectal and gastric cancer cell

lines and in experimental models (Rajendra et al., 2004; Jiang

et al., 2010) These findings culminated in clinical trials

evaluating the effect of H2receptor-targeted therapy in

colo-rectal cancer, indicating a beneficial effect when H2receptor

antagonists were given as therapy, adjuvant to curative gical resection (Deva and Jameson, 2012) Interestingly, H2

sur-receptor expression was comparable in colorectal cancer andadjacent normal mucosal specimens, whereas H1receptor and

H4receptor expression were significantly reduced in tumour

tissue (Boer et al., 2008; Fang et al., 2011) These findings

suggest that carcinogenesis might benefit from loss of H4

receptors (and H1 receptors) A potential antiproliferativeaction of H4receptors in colorectal cancer was further sub-

stantiated by in vitro experiments demonstrating that

stimu-lation of H4 receptors induced a cell cycle arrest in the G1phase via a cAMP-dependent pathway, resulting in reduced

cell proliferation and tumour growth (Table 4) (Fang et al.,

2011) This antiproliferative action was only present in H4

receptor-expressing colorectal cancer cell lines, but not inmock-transfected cells and could be prevented by pretreat-ment with the selective H4 receptor antagonist JNJ7777120,further corroborating involvement of these receptors Inaddition, H4 receptor stimulation enhanced apoptosisinduced by the chemotherapeutic agent 5-fluorouracil (Fang

et al., 2011) In contrast, Cianchi et al (2005) found that H4

receptor expression was increased in colorectal cancer mens Moreover, histamine-exposure stimulated cell prolif-eration and VEGF levels, which were reduced by the H4

speci-receptor antagonist JNJ7777120 (and the H2receptor nist cimetidine) This proliferative effect of H4receptor stimu-lation was mediated by COX 2-induced PGE2 as it wasonly evident in those cell lines that expressed COX 2

antago-(Cianchi et al., 2005) In addition, JNJ7777120 only reduced

histamine-induced cell proliferation, but did not affect basal

(non-histamine stimulated) cell growth (Coruzzi et al., 2012).

Attenuated H4receptor expression was reported in humangastric cancer specimens and was most prominent in

advanced malignancies (Zhang et al., 2012) Similarly to what

was previously demonstrated in colorectal cancer, reduced H4

receptor expression was linked to enhanced cell proliferation

as H4receptor stimulation with clobenpropit and histamine

reduced the growth of gastric cancer cells (Zhang et al., 2012).

Although neither ligand is an exclusive H4receptor agonist,the involvement of H4receptors was inferred from the factthat pretreatment with the selective H4 receptor antagonistJNJ7777120 completely abolished agonist-induced responses

(Zhang et al., 2012) In line with this, clobenpropit reduced

tumour cell proliferation in a pancreatic duct carcinoma cell

line (Cricco et al., 2008).

In contrast, H4 receptor expression was enhanced inmalignant cholangiocytes from patients with proven cholan-

giocarcinoma (Meng et al., 2011) H4 receptor stimulationwith the H3receptor antagonist/H4receptor agonist cloben-propit dose-dependently reduced proliferation of several

cholangiocarcinoma cell lines in vitro (Meng et al., 2011) This

cytostatic effect resulted from reduced growth potential anddisruption of the invading capacity of the cells As the effect

of clobenpropit was maintained in in vitro experiments in

which H3 receptors were knocked down, this indicates thatthe effects were indeed mediated via H4 receptors Impor-

tantly, in an elegant in vivo design, the authors demonstrated

the clinical potential of H4 receptor-modulation in thistumour type as treatment with clobenpropit inhibitedtumour growth and disrupted its invasive potential in a xeno-

graphic cholangiocarcinoma mouse model (Meng et al.,

BJP A Deiteren et al.

1170 British Journal of Pharmacology (2015) 172 1165–1178

2011) However, in another study, inhibition of H4receptors

by the H3/H4receptor antagonist thioperamide did not affect

cholangiocarcinoma cell line proliferation (Francis et al.,

2012)

Overall, these findings indicate that depending on the

type of tumour (gastric vs colorectal vs cholangiocarcinoma)

H4receptor-expression can either be decreased or enhanced

It is unclear whether H4 receptor expression differs in early

versus advanced stages, and this would be interesting to

investigate further In addition, although the data gathered

from in vitro experiments using different cell lines strongly

indicate that H4 receptors can potently modulate tumour

growth and progression, the results are not univocal To

complement these in vitro findings and increase our

under-standing of the role of H4 receptors in gastrointestinal

car-cinogenesis, additional research and in vivo experiments

using selective ligands seem crucial

H4 receptors and visceral

sensory signalling

Visceral hypersensitivity refers to an enhanced perception of

stimuli originating from the internal organs and is believed to

contribute to abdominal pain in multiple gastrointestinal

disorders among which IBD, IBS and functional dyspepsia

(Vermeulen et al., 2014) Sensitization of afferent nerve

endings in the gut wall is thought to underlie visceral

hyper-sensitivity (Anand et al., 2007) Several lines of evidence

indi-cate that histamine is involved in this process (Buhner and

Schemann, 2012; van Diest et al., 2012) For instance,

super-natant from IBS colonic biopsies contains increased levels of

histamine (Barbara et al., 2007) When applied to human

submucous neurons, this supernatant increased neuronalactivity and the degree of activation correlated with hista-

mine levels in the supernatant (Buhner et al., 2009) In

addi-tion, histamine induced murine jejunal afferent firing and

excited primary sensory neurons (Kreis et al., 1998; Brunsden

and Grundy, 1999) The pro-nociceptive effect of histamineseems to be mediated – at least partially – by H1 receptorsexpressed on sensory afferents, which is consistent with thefinding that excitation of rat jejunal afferents by IBS super-natant can be reduced by application of the H1 receptor

antagonist pyrilamine (Barbara et al., 2007) In addition, a

role for H4receptors in mediating visceral sensory signallingand nociception has emerged (Table 5)

Breunig et al (2007) reported that the H4receptor agonist4-methylhistamine excited human submucous plexusneurons, an effect that was inhibited by the selective H4

receptor antagonist JNJ7777120 Also, in vitro jejunal afferent

excitation by histamine was reversed by the H3/H4 receptorantagonist thioperamide (Brunsden and Grundy, 1999),although these results are in contrast to earlier reports in a

similar set-up (Kreis et al., 1998) Recently, our group vided in vivo evidence of reduced visceral nociception after

pro-blockade of H4 receptors Post-inflammatory visceral sensitivity was dose-dependently reduced by JNJ7777120 in arat model of post-inflammatory IBS, without affecting vis-

hyper-ceral sensitivity in healthy controls (Deiteren et al., 2014).

Although increased colonic expression of H4receptor mRNA

in hypersensitive rats points towards a peripheral mechanism

of action, it remains to be determined whether the ceptive effect is mediated by blockade of H4 receptors onsensory afferents directly or indirectly by modulation of H4

antinoci-receptors expressed elsewhere in the gut wall (Deiteren et al.,

Table 4

Preclinical in vitro and in vivo experiments with H4receptor ligands on carcinogenesis

In vitro/

Colorectal cancer cell

line

Human In vitro Clozapine

ClobenpropitJNJ7777120

Clozapine and clobenpropit reduced cell growthClozapine enhanced 5-FU induced apoptosis, whichwas reversed by JNJ7777120

Fang et al.

(2011)Colorectal cancer cell

line

Human In vitro JNJ7777120 JNJ7777120 prevented histamine-induced COX-2

expression/activity, cell proliferation and VEGFproduction

Cianchi et al.

(2005)Gastric cancer cell line Human In vitro Clobenpropit

carcinoma cell line

Human In vitro Clobenpropit Clobenpropit stimulation reduces cell growth Cricco et al.

(2008)Cholangiocarcinoma

cell line

Human In vitro Thioperamide No effect on histamine secretion and cell growth Francis et al.

(2012)Xenograft

cholangiocarcinoma

Mouse In vivo Clobenpropit Clobenpropit inhibited tumour growth Meng et al.

(2011)5-FU, 5-fluorouracil; clobenpropit, H3 receptor antagonist, H4receptor agonist; clozapine, H3 and H4receptor agonist; JNJ7777120, H4

receptor antagonist; thioperamide, H3and H4receptor antagonist

British Journal of Pharmacology (2015) 172 1165–1178 1171

2014) Nevertheless, these findings coincide with previous

reports of antinociceptive and analgesic effects of H4receptor

antagonists in models of somatic and neuropathic pain

(inde-pendent of their anti-inflammatory properties) (Coruzzi et al.,

2007; Hsieh et al., 2010) and emphasize that H4receptors are

also attractive targets in the modulation of visceral pain

H4 receptors and

intestinal contractility

Histaminergic control of gastrointestinal contractility and

motility is complex and involves all histamine receptor

sub-types H1receptors, located in smooth muscle cells,

contrib-ute to contractility by increasing calcium availability at the

sarcoplasmic level whereas H2 receptors mainly facilitate

cholinergic and non-cholinergic excitatory transmission in

intramural neurons (Poli et al., 2001) Although H3receptors

inhibit the release of excitatory and inhibitory

neurotrans-mitters from the myenteric plexus, their involvement in

enteric peristalsis remains unclear, as no effect of H3receptor

ligands on gastrointestinal transit was seen in in vivo models

and the presence of H3 receptors in the human digestive

tract remains controversial (Hemedah et al., 2001; Poli

et al., 2001; Cianchi et al., 2005; Sander et al., 2006) Recently,

H4 receptors were reported to be present on murine

myenteric neurons (Chazot et al., 2007) In addition,

4-methylhistamine excited human submucous plexusneurons, which could be blocked by the H4receptor antago-

nist JNJ7777120 (Breunig et al., 2007) As the enteric plexus

in highly involved in the regulation of reflex behaviour, stalsis and intestinal secretion, these findings suggest that H4

peri-receptors could be involved in gut motility and transit(Table 6) However, the H4receptor agonist VUF8430 did notaffect twitch responses induced by electrical field stimulation

in rat duodenum (Pozzoli et al., 2009) In addition, no effect

was seen from H4 receptor stimulation on the membranepotential of murine small intestinal interstitial cells of Cajal,the enteric pacemaker cells and conductors of electrical slow

waves in intestinal smooth muscle (Kim et al., 2013) In a

recent study, longitudinal smooth muscle preparations with

Table 5

Preclinical in vitro and in vivo experiments with H4receptor ligands in visceral sensory signalling and nociception

In vitro/

Submucous plexus neurons Human In vitro 4-methylhistamine

JNJ7777120

4-methylhistamine-induced excitationreduced by JNJ7777120

Breunig et al.

(2007)Jejunal afferent firing Rat In vitro Thioperamide Thioperamide reduced histamine-induced

jejunal afferent firing

Brunsden andGrundy (1999)Jejunal afferent firing Rat In vivo Thioperamide No effect on histamine-induced jejunal

Table 6

Preclinical in vitro experiments with H4receptor ligands in gastrointestinal contractility and transit

In vitro/

Submucous plexus neurons Human In vitro 4-methylhistamine

JNJ7777120

4-methylhistamine-induced excitationreduced by JNJ7777120

Breunig et al.

(2007)Whole mount duodenum

segments

Rat In vitro VUF8430 No effect on contractions Pozzoli et al.

(2009)Longitudinal smooth muscle

incl myenteric plexus

Guinea pig In vitro Thioperamide No effect on contractions induced by

IBS supernatant

Balestra et al.

(2012)Colonic smooth muscle strips Monkey In vitro 4-methyl-histamine 4-methylhistamine increased contractile

force

Kim et al.

(2011)Cultured small intestine

interstitial cells of Cajal

Mouse In vitro 4-methyl-histamine No effect on pace maker potentials Kim et al.

(2013)4-methylhistamine, H4receptor agonist; IBS, irritable bowel syndrome; JNJ7777120, H4receptor antagonist; thioperamide, H3and H4receptorantagonist; VUF8430, H4receptor agonist

BJP A Deiteren et al.

1172 British Journal of Pharmacology (2015) 172 1165–1178

an intact myenteric plexus were harvested from guinea pig

ileum and exposed to supernatants prepared from colonic

biopsies from IBS patients This supernatant enhanced

cho-linergic twitch contractions; however, the responses were

not affected by a mixture containing antagonists for H1–H4

receptors (Balestra et al., 2012) The H4 receptor agonist

4-methylhistamine increased contractile forces only in

longi-tudinal smooth muscle strips of monkey colon (Kim et al.,

2011) However, as these effects were only present when high

doses were used, these results need to be interpreted with

caution

H4 receptors and gastric acid secretion

and ulceration

Histamine is a potent activator of the acid secreting cells of

the stomach (Kopic and Geibel, 2010) Binding of histamine

to basolateral H4receptors activates adenylyl cyclase resulting

in accumulation of cAMP and H+secretion Before the

devel-opment of proton pump inhibitors, pharmacological

block-ade of H2receptors was the cornerstone of the treatment of

acid-related gastrointestinal disorders (Kopic and Geibel,

2010) In addition to H2 receptor antagonists, H3 receptor

stimulation also exerted gastroprotective effects via increased

mucus production in animal models (Coruzzi et al., 2001;

Barocelli and Ballabeni, 2003) The homology between H3

and H4receptors subsequently spurred interest in a possiblerole for H4 receptors in gastric acid secretion (Table 7).Overall, the data gathered to date suggest that H4receptors donot participate in gastric acid secretion under physiologicalconditions as neither H4receptor agonists such as VUF8430and VUF10460 nor H4 receptor antagonists such asJNJ7777120 and VUF5949 affected basal acid production or

the macroscopic appearance of the stomach (Lim et al., 2009; Coruzzi et al., 2011; Adami et al., 2012) However, when the

mucosal integrity was compromised such as in models ofchemically induced gastric ulceration, damage was signifi-cantly enhanced by H4 receptor stimulation and markedly

reduced by its blockade (Adami et al., 2005; 2012; Coruzzi

et al., 2009; 2011) In addition, enhanced chemically induced

mucosal damage by H4receptor agonists could be prevented

by concomitant H4 receptor antagonists and vice versa (Coruzzi et al., 2009; 2011) However, the findings are not

fully consistent as the H4 receptor agonists VUF8430 andVUF10460 had no effect on indomethacin/bethanecol-induced lesions in a mouse model whereas the HCl-induceddamage in rats was enhanced by both agonists, and in con-trast, indomethacin-induced ulcerations were reduced by

VUF10460 (Coruzzi et al., 2009; 2011; Adami et al., 2012) It

was hypothesized that species and strain differences mightcontribute to the differential effects as JNJ7777120 effectivelyreduced indomethacin/bethanecol-induced lesions in CD-1,

NMRI and BALB/c, but not in C57BL/6J mice (Adami et al.,

2012) However, it should be kept in mind that several of the

Lim et al (2009)

JNJ7777120 Induced gastric acid secretion was not

affected by JNJ7777120Indomethacin/

bethanechol-induced

gastric ulceration

Mouse In vivo JNJ7777120 JNJ7777120 reduces lesions in CD-1, NMRI

and BALB/c, but not in C57BL/6J mice

Adami et al (2012; Coruzzi et al.

(2009)VUF10460

(2009)VUF10460 VUF10460 reduced lesions

VUF8430 VUF8430 only reduced lesions in the presence

of a H2receptor antagonistHCl-induced gastric

ulceration

Rat In vivo Immepip

VUF8430VUF10460

Immepip, VUF8430 and VUF10460 enhancedHCl-induced gastric lesions

Coruzzi et al.

(2011)JNJ7777120 JNJ7777120 abolished the effect of immepip,

but not of VUF8430 and VUF10460Dimaprit, H2receptor agonist, H4receptor agonist; Immepip, H3and H4receptor agonist; JNJ7777120, H4receptor antagonist; VUF10460,

H4receptor agonist; VUF5949, H4receptor antagonist; VUF8430, H4receptor agonist

British Journal of Pharmacology (2015) 172 1165–1178 1173

compounds used also display considerable affinity for the H3

receptor, such as VUF8430 [pKifor rat H4receptors of 6.9 vs

6.5 for rat H3 receptors (Lim et al., 2009); Table 1], again

underscoring the need for selective H4 receptor ligands

Although these data seem promising, more research is needed

to further elucidate the effect of H4receptor modulation on

gastric ulcer disease If a beneficial effect of H4receptor

block-ade on gastric ulceration could be confirmed, this would be a

major advantage in terms of drug development, as H4

recep-tor antagonists are already under evaluation for their

anti-inflammatory and analgesic properties

Clinical development

To date, no clinical trials with H4receptor ligands have been

initiated in the field of gastroenterology However, several H4

receptor antagonists have already progressed to phase II

clinical trials for immune-mediated disorders such as

rheu-matoid arthritis, asthma, atopic dermatitis and allergic

rhinitis (Table 8) Currently registered clinical trials (http://

clinicaltrials.gov) include compounds from Johnson &

Johnson (JNJ39758979 and JNJ38518168), Ziarco Pharma

(ZPL3893787) and Palau Pharma (UR-63325)

JNJ39758979, derived from the H4 receptor antagonist

JNJ7777120, showed promising results in initial phase I trials,

with good pharmacokinetics upon oral dosing with a plasma

half-life of 124–157 h after a single oral dose (Thurmond

et al., 2014) In addition, the compound was well-tolerated up

to 1200 mg in single ascending dose studies and up to 300 mgbid in a multiple ascending dose study; dose-dependentgastrointestinal symptoms were the main adverse events(abnormal faeces, nausea, vomiting and abdominal pain)

(Thurmond et al., 2014) A single dose of 600 mg effectively

reduced histamine-induced itch in 23 healthy volunteers

(Kollmeier et al., 2014) However, a subsequent phase II trial

in patients with atopic dermatitis was discontinued due totwo cases of drug-induced agranulocytosis, leading to the

termination of JNJ39758979 (Thurmond et al., 2014) The

agranulocytosis was reported to be related to the chemicalstructure of JNJ39758979 and not to the H4receptor antago-

nism (Kollmeier et al., 2014; Liu, 2014; Thurmond et al.,

2014); further details are expected to be released in the near

future (Thurmond et al., 2014) Therefore, the development

of other H4receptor antagonists is currently being pursuedsuch as JNJ38518168, which has progressed to phase II forrheumatoid arthritis and asthma However, one of these trialswas terminated because of a single, unexpected serious event,which was not specified further Details on the underlyingmechanisms (H4 receptor-related or compound-specific) arenot yet available

ZPL3893787 (former PF03893787) is a lead compound ofZiarco Pharma and successfully completed phase I singleascending dose and 14 days multiple ascending dose studies

Table 8

Current clinical trials with H4receptor ligands

I Histamine-induced itch in healthy volunteers Completed NCT01068223

I Patients with normal or mild to moderate

hepatic impairment

*Terminated because of a single, unexpected serious event.†Terminated because of two cases of agranulocytosis.‡Terminated/withdrawnbecause of cases of agranulocytosis in trial NCT01497119.§Former PF03893787 Clinical trials as registered on http://clinicaltrials.gov on 11June 2014

BJP A Deiteren et al.

1174 British Journal of Pharmacology (2015) 172 1165–1178

in healthy volunteers (Liu, 2014) No results have been

pub-lished yet; however, Ziarco Pharma communicated on their

website that the compound displayed an excellent

pharma-cokinetic and safety profile Results from a subsequent proof

of concept trial in patients with asthma have not yet been

disclosed

The Palau Pharma compound UR-63325 successfully

com-pleted single and multiple dose ascending studies

demon-strating a linear pharmacokinetic profile and no safety

concerns according to Salcedo et al (2013); in addition, a

phase II clinical trial in allergic rhinitis patients was recently

completed and the data are eagerly awaited

Conclusions

Since their discovery and cloning almost 15 years ago,

knowl-edge on the role of H4 receptors has increased rapidly The

expression of H4receptors on immune cells has spurred

inter-est in H4 receptor antagonists as a potential new class of

anti-inflammatory drugs in the treatment of rheumatoid

arthritis and asthma among others Also in the

gastrointesti-nal tract, there is now strong preclinical evidence that H4

receptors modulate the inflammatory process, indicating that

these receptors could be interesting new targets in the

treat-ment of IBD, ischaemia/reperfusion injury, radiation-induced

enteropathy and allergic intestinal reactions It would be

interesting to investigate whether genetic polymorphisms

and copy gene number variations for H4receptors are linked

to gastrointestinal inflammation as was previously reported

for other immune-mediated disorders such asthma and

atopic dermatitis (Yu et al., 2010; Simon et al., 2012; Chen

et al., 2013) In addition, recent data indicate that H4

recep-tors also participate in carcinogenesis and gastric ulceration

and in mediating IBS-like visceral pain The preliminary data

gathered so far seem promising, but the effects of

pharmaco-logical H4 receptor modulation will need to be confirmed

using highly selective ligands, that are devoid of biased

sig-nalling and are extensively evaluated in both in vitro and in

vivo settings In addition, the results of ongoing trials with H4

receptor antagonists for immune-mediated disorders are

eagerly awaited and will be crucial for the future of any

therapy targeted at H4receptors

Acknowledgements

A Deiteren is an aspirant of the Fund for Scientific Research

(FWO), Flanders This work was supported financially by the

FWO (G.0341.13 and G.0249.09N)

Conflicts of interest

The authors report no conflicts of interest

References

Adachi N, Liu K, Motoki A, Nishibori M, Arai T (2006) Suppression

of ischemia/reperfusion liver injury by histamine H4 receptor

stimulation in rats Eur J Pharmacol 544: 181–187

Adami M, Coruzzi G, Guaita E, de Esch I, Leurs R (2005)

Antiinflammatory, analgesic and gastroprotective effects of thenovel and selective histamine H4-receptor antagonist VUF5949.34th Meeting of the European Histamine Research Society p 47.Adami M, Pozzoli C, Menozzi A, Bertini S, Passeri B, Cantoni AM

et al (2012) Effects of histamine H4 receptor ligands in a mouse

model of gastric ulceration Pharmacology 89: 287–294

Alcaniz L, Vega A, Chacon P, El Bekay R, Ventura I, Aroca R et al.

(2013) Histamine production by human neutrophils FASEB 27:2902–2910

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL,

Spedding M et al (2013a) The Concise Guide to PHARMACOLOGY

2013/14: G protein-coupled receptors Br J Pharmacol 170:

1459–1581

Alexander SPH, Benson HE, Faccenda E, Pawson AJ, Sharman JL

Spedding M et al (2013b) The Concise Guide to PHARMACOLOGY

2013/14: Enzymes Br J Pharmacol 170: 1797–1867

Alfon J, Sanchez-Gomez S, Salcedo C, Fernandez E, Gil-Torregrosa B,

Ardanaz N et al (2011) Efficacy of UR-63325, a new histamine H4

receptor antagonist in house dust mite-induced mouse asthmamodels Am J Respir Crit Care Med 183: A1302

Anand P, Aziz Q, Willert R, van Oudenhove L (2007) Peripheraland central mechanisms of visceral sensitization in man

Neurogastroenterol Motil 19 (1 Suppl.): 29–46

Andaloussi M, Lim HD, van der Meer T, Sijm M, Poulie CB, de Esch

IJ et al (2013) A novel series of histamine H4 receptor antagonists

based on the pyrido[3,2-d]pyrimidine scaffold: comparison of hERGbinding and target residence time with PF-3893787 Bioorg MedChem Lett 23: 2663–2670

Balestra B, Vicini R, Cremon C, Zecchi L, Dothel G, Vasina V et al.

(2012) Colonic mucosal mediators from patients with irritablebowel syndrome excite enteric cholinergic motor neurons

Neurogastroenterol Motil 24: 1118–e1570

Barbara G, Stanghellini V, De Giorgio R, Corinaldesi R (2006).Functional gastrointestinal disorders and mast cells: implicationsfor therapy Neurogastroenterol Motil 18: 6–17

Barbara G, Wang B, Stanghellini V, de Giorgio R, Cremon C,

Di Nardo G et al (2007) Mast cell-dependent excitation of

visceral-nociceptive sensory neurons in irritable bowel syndrome.Gastroenterology 132: 26–37

Barocelli E, Ballabeni V (2003) Histamine in the control of gastricacid secretion: a topic review Pharmacol Res 47: 299–304

Beermann S, Seifert R, Neumann D (2012) Commercially availableantibodies against human and murine histamine H(4)-receptor lackspecificity Naunyn Schmiedebergs Arch Pharmacol 385: 125–135.Bencsath M, Paloczi K, Szalai C, Szenthe A, Szeberenyi J, Falus A(1998) Histidine decarboxylase in peripheral lymphocytes ofhealthy individuals and chronic lymphoid leukemia patients.Pathol Oncol Res 4: 121–124

Bischoff SC, Wedemeyer J, Herrmann A, Meier PN, Trautwein C,

Cetin Y et al (1996) Quantitative assessment of intestinal

eosinophils and mast cells in inflammatory bowel disease

Histopathology 28: 1–13

Black JW, Duncan WA, Durant CJ, Ganellin CR, Parsons EM (1972).Definition and antagonism of histamine H2-receptors Nature 236:385–390

Boer K, Helinger E, Helinger A, Pocza P, Pos Z, Demeter P et al.

(2008) Decreased expression of histamine H1 and H4 receptors

British Journal of Pharmacology (2015) 172 1165–1178 1175

suggests disturbance of local regulation in human colorectal

tumours by histamine Eur J Cell Biol 87: 227–236

Breunig E, Michel K, Zeller F, Seidl S, Weyhern CW, Schemann M

(2007) Histamine excites neurones in the human submucous

plexus through activation of H1, H2, H3 and H4 receptors

J Physiol 583 (Pt 2): 731–742

Brunsden AM, Grundy D (1999) Sensitization of visceral afferents

to bradykinin in rat jejunum in vitro J Physiol 521 (Pt 2): 517–527.

Buhner S, Schemann M (2012) Mast cell-nerve axis with a focus on

the human gut Biochim Biophys Acta 1822: 85–92

Buhner S, Li Q, Vignali S, Barbara G, De Giorgio R, Stanghellini V

et al (2009) Activation of human enteric neurons by supernatants

of colonic biopsy specimens from patients with irritable bowel

syndrome Gastroenterology 137: 1425–1434

Chazot PL, Shenton FC, Waldvogel HJ, Grandi D, Morini G (2007)

The H4 histamine receptor is expressed in both the human CNS

and rodent PNS 36th Annual Meeting of the European Histamine

Research Society, Florence, Italy, p O4

Chen B, Ye T, Shao Y, Zhang J, Zhong Q, Hu X et al (2013).

Association between copy-number variations of the human

histamine H4 receptor gene and atopic dermatitis in a Chinese

population Clin Exp Dermatol 38: 295–300, quiz 300-301

Cianchi F, Cortesini C, Schiavone N, Perna F, Magnelli L, Fanti E

et al (2005) The role of cyclooxygenase-2 in mediating the effects

of histamine on cell proliferation and vascular endothelial growth

factor production in colorectal cancer Clin Cancer Res 11 (19 Pt 1):

6807–6815

Coge F, Guenin SP, Rique H, Boutin JA, Galizzi JP (2001) Structure

and expression of the human histamine H4-receptor gene Biochem

Biophys Res Commun 284: 301–309

Coruzzi G, Morini G, Adami M, Grandi D (2001) Role of histamine

H3 receptors in the regulation of gastric functions J Physiol

Pharmacol 52 (4 Pt 1): 539–553

Coruzzi G, Adami M, Guaita E, de Esch IJ, Leurs R (2007)

Antiinflammatory and antinociceptive effects of the selective

histamine H4-receptor antagonists JNJ7777120 and VUF6002 in a

rat model of carrageenan-induced acute inflammation Eur J

Pharmacol 563: 240–244

Coruzzi G, Adami M, Pozzoli C, Smits R, De Esch I, Leurs R (2009)

Gastroprotective effects of histamine H4 receptor ligands in rodent

ulcer models Presented at the British Pharmacological Society,

London, UK

Coruzzi G, Adami M, Pozzoli C, de Esch IJ, Smits R, Leurs R (2011)

Selective histamine H(3) and H(4) receptor agonists exert opposite

effects against the gastric lesions induced by HCl in the rat

stomach Eur J Pharmacol 669: 121–127

Coruzzi G, Adami M, Pozzoli C (2012) Role of histamine H4

receptors in the gastrointestinal tract Front Biosci (Schol Ed) 4:

226–239

Cowart MD, Altenbach RJ, Liu H, Hsieh GC, Drizin I, Milicic I et al.

(2008) Rotationally constrained 2,4-diamino-5,6-disubstituted

pyrimidines: a new class of histamine H4 receptor antagonists with

improved druglikeness and in vivo efficacy in pain and

inflammation models J Med Chem 51: 6547–6557

Cricco GP, Mohamad NA, Sambuco LA, Genre F, Croci M, Gutierrez

AS et al (2008) Histamine regulates pancreatic carcinoma cell

growth through H3 and H4 receptors Inflamm Res 57 (Suppl 1):

S23–S24

Dawicki W, Marshall JS (2007) New and emerging roles for mast

cells in host defence Curr Opin Immunol 19: 31–38

Deiteren A, De Man JG, Ruyssers NE, Moreels TG, Pelckmans PA,

De Winter BY (2014) Histamine H4 and H1 receptors contribute topostinflammatory visceral hypersensitivity Gut 63: 1873–1882.Deva S, Jameson M (2012) Histamine type 2 receptor antagonists

as adjuvant treatment for resected colorectal cancer CochraneDatabase Syst Rev (8): CD007814

van Diest SA, Stanisor OI, Boeckxstaens GE, de Jonge WJ,van den Wijngaard RM (2012) Relevance of mast cell-nerveinteractions in intestinal nociception Biochim Biophys Acta 1822:74–84

Dunford PJ, O’Donnell N, Riley JP, Williams KN, Karlsson L,Thurmond RL (2006a) The histamine H4 receptor mediates allergicairway inflammation by regulating the activation of CD4+ T cells

J Immunol 176: 7062–7070

Dunford PJ, Varga C, Thurmond RL, Whittle BJ (2006b) Histamine

H4receptor antagonism attenuates toll-like receptor signaling andinhibits experimental colitis in the rat Gastroenterology 130 (4Suppl 2): A689

Eisenschenk MN, Torres SM, Oliveira S, Been CS (2011) Theexpression of histamine H4 receptor mRNA in the skin and othertissues of normal dogs Vet Dermatol 22: 396–400

El-Mahdy NA, El-Sisi AE, Dewidar BI, El-Desouky KI (2013).Histamine protects against the acute phase of

experimentally-induced hepatic ischemia/re-perfusion

J Immunotoxicol 10: 9–16

Eliakim R, Karmeli F, Okon E, Rachmilewitz D (1992) Ketotifeneffectively prevents mucosal damage in experimental colitis Gut33: 1498–1503

Endo Y (1982) Simultaneous induction of histidine and ornithinedecarboxylases and changes in their product amines following theinjection of Escherichia coli lipopolysaccharide into mice BiochemPharmacol 31: 1643–1647

Fang Z, Yao W, Xiong Y, Li J, Liu L, Shi L et al (2011) Attenuated

expression of HRH4 in colorectal carcinomas: a potential influence

on tumor growth and progression BMC Cancer 11: 1–11

Farhadi A, Keshavarzian A, Fields JZ, Jakate S, Shaikh M, Banan A(2007) Reduced immunostaining for c-kit receptors in mucosalmast cells in inflammatory bowel disease J Gastroenterol Hepatol22: 2338–2343

Fogel WA, Wagner W, Sasiak K, Stasiak A (2005) The role ofhistamine in experimental ulcerative colitis in rats Inflamm Res 54(Suppl 1): S68–S69

Fogel WA, Jochem J, Lewinski A (2007) Influence of the H3/H4receptor antagonist, thioperamide on regional haemodynamics inrats with trinitrobenzene sulfonic acid-induced colitis Inflamm Res

56 (Suppl 1): S21–S22

Francis H, DeMorrow S, Venter J, Onori P, White M, Gaudio E et al.

(2012) Inhibition of histidine decarboxylase ablates the autocrinetumorigenic effects of histamine in human cholangiocarcinoma.Gut 61: 753–764

Francois A, Milliat F, Guipaud O, Benderitter M (2013)

Inflammation and immunity in radiation damage to the gutmucosa Biomed Res Int 2013: 123241

Fukushima Y, Ohmachi Y, Asano T, Nawano M, Funaki M, Anai M

et al (1999) Localization of the histamine H(2) receptor, a target

for antiulcer drugs, in gastric parietal cells Digestion 60: 522–527.Ghizzardi P, Gobbetti T, Bertoni S, Saccani F, Flammini L,Ballabeni V, Barocelli E (2009) Histamine H4 receptor antagonism

BJP A Deiteren et al.

1176 British Journal of Pharmacology (2015) 172 1165–1178

and mesenteric ischemia/reperfusion injury in mice.

Gastroenterology 136 (5 Suppl 1): A397

He SH (2004) Key role of mast cells and their major secretory

products in inflammatory bowel disease World J Gastroenterol 10:

309–318

Hemedah M, Loiacono R, Coupar IM, Mitchelson FJ (2001) Lack of

evidence for histamine H3 receptor function in rat ileum and

human colon Naunyn Schmiedebergs Arch Pharmacol 363:

133–138

Heron A, Rouleau A, Cochois V, Pillot C, Schwartz JC, Arrang JM

(2001) Expression analysis of the histamine H(3) receptor in

developing rat tissues Mech Dev 105: 167–173

Hofstra CL, Desai PJ, Thurmond RL, Fung-Leung WP (2003)

Histamine H4 receptor mediates chemotaxis and calcium

mobilization of mast cells J Pharmacol Exp Ther 305: 1212–1221

Hsieh GC, Chandran P, Salyers AK, Pai M, Zhu CZ, Wensink EJ

et al (2010) H4 receptor antagonism exhibits anti-nociceptive

effects in inflammatory and neuropathic pain models in rats

Pharmacol Biochem Behav 95: 41–50

Jiang CG, Liu FR, Yu M, Li JB, Xu HM (2010) Cimetidine induces

apoptosis in gastric cancer cells in vitro and inhibits tumor growth

in vivo Oncol Rep 23: 693–700.

Jiang W, Lim HD, Zhang M, Desai P, Dai H, Colling PM et al.

(2008) Cloning and pharmacological characterization of the dog

histamine H4 receptor Eur J Pharmacol 592: 26–32

Jones BL, Kearns GL (2011) Histamine: new thoughts about a

familiar mediator Clin Pharmacol Ther 89: 189–197

Jones NL, Roifman CM, Griffiths AM, Sherman P (1998) Ketotifen

therapy for acute ulcerative colitis in children: a pilot study Dig

Dis Sci 43: 609–615

Jutel M, Akdis M, Akdis CA (2009) Histamine, histamine receptors

and their role in immune pathology Clin Exp Allergy 39:

1786–1800

Kennedy L, Hodges K, Meng F, Alpini G, Francis H (2012)

Histamine and histamine receptor regulation of gastrointestinal

cancers Transl Gastrointest Cancer 1: 215–227

Kim BJ, Kwon YK, Kim E, So I (2013) Effects of histamine on

cultured interstitial cells of cajal in murine small intestine Korean J

Physiol Pharmacol 17: 149–156

Kim H, Dwyer L, Song JH, Martin-Cano FE, Bahney J, Peri L et al.

(2011) Identification of histamine receptors and effects of

histamine on murine and simian colonic excitability

Neurogastroenterol Motil 23: 949–e409

Knutson L, Ahrenstedt O, Odlind B, Hallgren R (1990) The jejunal

secretion of histamine is increased in active Crohn’s disease

Gastroenterology 98: 849–854

Kollmeier A, Francke K, Chen B, Dunford PJ, Greenspan AJ, Xia Y

et al (2014) The histamine H4 receptor antagonist, JNJ 39758979,

is effective in reducing histamine-induced pruritus in a randomized

clinical study in healthy subjects J Pharmacol Exp Ther 350:

181–187

Kopic S, Geibel JP (2010) Update on the mechanisms of gastric

acid secretion Curr Gastroenterol Rep 12: 458–464

Kreis ME, Haupt W, Kirkup AJ, Grundy D (1998) Histamine

sensitivity of mesenteric afferent nerves in the rat jejunum Am J

Physiol 275 (4 Pt 1): G675–G680

Kuhne S, Wijtmans M, Lim HD, Leurs R, de Esch IJ (2011) Several

down, a few to go: histamine H3 receptor ligands making the final

push towards the market? Expert Opin Investig Drugs 20:

1629–1648

Kumar V, Laddu AR, Lahiri PK, Sanyal RK (1968) The physiologicallevels of histamine in human tissues Int Arch Allergy ApplImmunol 34: 233–236

Kumawat AK, Götlind Y-Y, Fredin MF, Willén R, Strid H,Hultgren-Hornquist E (2010) Expression patterns of histaminereceptors in the Gi2alpha-deficient mouse model of colitis InflammRes 59 (4 Suppl.): S358–S359

Leurs R, Chazot PL, Shenton FC, Lim HD, de Esch IJ (2009).Molecular and biochemical pharmacology of the histamine H4receptor Br J Pharmacol 157: 14–23

Lim HD, Adami M, Guaita E, Werfel T, Smits RA, de Esch IJ et al.

(2009) Pharmacological characterization of the new histamine H4receptor agonist VUF 8430 Br J Pharmacol 157: 34–43

Liu C, Ma X, Jiang X, Wilson SJ, Hofstra CL, Blevitt J et al (2001a).

Cloning and pharmacological characterization of a fourthhistamine receptor (H(4)) expressed in bone marrow MolPharmacol 59: 420–426

Liu C, Wilson SJ, Kuei C, Lovenberg TW (2001b) Comparison ofhuman, mouse, rat, and guinea pig histamine H4 receptors revealssubstantial pharmacological species variation J Pharmacol Exp Ther299: 121–130

Liu WL (2014) Histamine H4 receptor antagonists for thetreatment of inflammatory disorders Drug Discov Today 19:1222–1225

Marshall JK, Irvine EJ (1998) Ketotifen treatment of active colitis inpatients with 5-aminosalicylate intolerance Can J Gastroenterol 12:273–275

Martinel Lamas DJ, Carabajal E, Prestifilippo JP, Rossi L, Elverdin

JC, Merani S et al (2013) Protection of radiation-induced damage

to the hematopoietic system, small intestine and salivary glands inrats by JNJ7777120 compound, a histamine H4 ligand PLoS ONE 8:e69106

Medina VA, Rivera ES (2010) Histamine receptors and cancerpharmacology Br J Pharmacol 161: 755–767

Meng F, Han Y, Staloch D, Francis T, Stokes A, Francis H (2011).The H4 histamine receptor agonist, clobenpropit, suppresses humancholangiocarcinoma progression by disruption of epithelialmesenchymal transition and tumor metastasis Hepatology 54:1718–1728

Morini G, Becchi G, Shenton FC, Chazot PL, Grandi D (2008).Histamine H3 and H4 receptors are expressed on distinct endocrinecell types in the rat fundic mucosa Inflamm Res 57 (Suppl 1):S57–S58

Morse KL, Behan J, Laz TM, West RE Jr, Greenfeder SA, Anthes JC

et al (2001) Cloning and characterization of a novel human

histamine receptor J Pharmacol Exp Ther 296: 1058–1066.Mowbray CE, Bell AS, Clarke NP, Collins M, Jones RM, Lane CA

et al (2011) Challenges of drug discovery in novel target space.

The discovery and evaluation of PF-3893787: a novel histamine H4receptor antagonist Bioorg Med Chem Lett 21: 6596–6602.Nakamura T, Itadani H, Hidaka Y, Ohta M, Tanaka K (2000).Molecular cloning and characterization of a new human histaminereceptor, HH4R Biochem Biophys Res Commun 279: 615–620

Nguyen T, Shapiro DA, George SR, Setola V, Lee DK, Cheng R et al.

(2001) Discovery of a novel member of the histamine receptorfamily Mol Pharmacol 59: 427–433

Nijmeijer S, Vischer HF, Sirci F, Schultes S, Engelhardt H, de Graaf

C et al (2013) Detailed analysis of biased histamine H(4) receptor

signalling by JNJ 7777120 analogues Br J Pharmacol 170: 78–88

British Journal of Pharmacology (2015) 172 1165–1178 1177

Oda T, Morikawa N, Saito Y, Masuho Y, Matsumoto S (2000).

Molecular cloning and characterization of a novel type of

histamine receptor preferentially expressed in leukocytes J Biol

Chem 275: 36781–36786

Oda T, Matsumoto S, Masuho Y, Takasaki J, Matsumoto M,

Kamohara M et al (2002) cDNA cloning and characterization of

porcine histamine H4 receptor Biochim Biophys Acta 1575:

135–138

Oda T, Matsumoto S, Matsumoto M, Takasaki J, Kamohara M, Soga

T et al (2005) Molecular cloning of monkey histamine H4

receptor J Pharmacol Sci 98: 319–322

Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP,

Buneman OP et al.; NC-IUPHAR (2014) The IUPHAR/BPS Guide to

PHARMACOLOGY: an expert-driven knowledge base of drug targets

and their ligands Nucl Acids Res 42 (Database Issue):

D1098–1106

Poli E, Pozzoli C, Coruzzi G (2001) Role of histamine H(3)

receptors in the control of gastrointestinal motility An overview

J Physiol (Paris) 95: 67–74

Pozzoli C, Adami M, Smits RA, Coruzzi G (2009) Effect of

histamine H4 receptor ligands on cholinergic neurotransmission of

the rat duodenum Inflamm Res 58 (Suppl 1): 59–60

Rajendra S, Mulcahy H, Patchett S, Kumar P (2004) The effect of

H2 antagonists on proliferation and apoptosis in human colorectal

cancer cell lines Dig Dis Sci 49: 1634–1640

Rosethorne EM, Charlton SJ (2011) Agonist-biased signaling at the

histamine H4 receptor: JNJ7777120 recruits beta-arrestin without

activating G proteins Mol Pharmacol 79: 749–757

Salcedo C, Pontes C, Merlos M (2013) Is the H4 receptor a new

drug target for allergies and asthma? Front Biosci 5: 178–187

Sander LE, Lorentz A, Sellge G, Coeffier M, Neipp M, Veres T et al.

(2006) Selective expression of histamine receptors H1R, H2R, and

H4R, but not H3R, in the human intestinal tract Gut 55: 498–504

Seifert R, Schneider EH, Dove S, Brunskole I, Neumann D, Strasser

A et al (2011) Paradoxical stimulatory effects of the ‘standard’

histamine H4-receptor antagonist JNJ7777120: the H4 receptor

joins the club of 7 transmembrane domain receptors exhibiting

functional selectivity Mol Pharmacol 79: 631–638

Simon T, Laszlo V, Falus A (2011) Impact of histamine on

dendritic cell functions Cell Biol Int 35: 997–1000

Simon T, Semsei AF, Ungvari I, Hadadi E, Virag V, Nagy A et al.

(2012) Asthma endophenotypes and polymorphisms in the

histamine receptor HRH4 gene Int Arch Allergy Immunol 159:

109–120

Simons FE, Simons KJ (2011) Histamine and H1-antihistamines:

celebrating a century of progress J Allergy Clin Immunol 128:

1139–1150 e1134

Singh M, Jadhav HR (2013) Histamine H3 receptor function and

ligands: recent developments Mini Rev Med Chem 13: 47–57

Smits RA, Leurs R, de Esch IJ (2009) Major advances in the

development of histamine H4 receptor ligands Drug Discov Today

14: 745–753

Snyder SH, Epps L (1968) Regulation of histidine decarboxylase in

rat stomach by gastrin: the effect of inhibitors of protein synthesis

Mol Pharmacol 4: 187–195

Sokol H, Georgin-Lavialle S, Grandpeix-Guyodo C, Canioni D,

Barete S, Dubreuil P et al (2010) Gastrointestinal involvement and

manifestations in systemic mastocytosis Inflamm Bowel Dis 16:

1247–1253

Strakhova MI, Nikkel AL, Manelli AM, Hsieh GC, Esbenshade TA,

Brioni JD et al (2009) Localization of histamine H4 receptors in

the central nervous system of human and rat Brain Res 1250:41–48

Sutton TL, Zhao A, Madden KB, Elfrey JE, Tuft BA, Sullivan CA et al (2008) Anti-Inflammatory mechanisms of enteric Heligmosomoides polygyrus infection against trinitrobenzene sulfonic acid-induced

colitis in a murine model Infect Immun 76: 4772–4782

Takagaki K, Osawa S, Horio Y, Yamada T, Hamaya Y, Takayanagi Y

et al (2009) Cytokine responses of intraepithelial lymphocytes are

regulated by histamine H(2) receptor J Gastroenterol 44: 285–296.Takeshita K, Sakai K, Bacon KB, Gantner F (2003) Critical role ofhistamine H4 receptor in leukotriene B4 production and mast

cell-dependent neutrophil recruitment induced by zymosan in vivo.

J Pharmacol Exp Ther 307: 1072–1078

Thurmond RL, Desai PJ, Dunford PJ, Fung-Leung WP, Hofstra CL,

Jiang W et al (2004) A potent and selective histamine H4 receptor

antagonist with anti-inflammatory properties J Pharmacol ExpTher 309: 404–413

Thurmond RL, Chen B, Dunford PJ, Greenspan AJ, Karlsson L, La D

et al (2014) Clinical and preclinical characterization of the

histamine H4 receptor antagonist JNJ-39758979 J Pharmacol ExpTher 349: 176–184

Togias A (2003) H1-receptors: localization and role in airwayphysiology and in immune functions J Allergy Clin Immunol 112(4 Suppl.): S60–S68

Vanhala A, Yamatodani A, Panula P (1994) Distribution ofhistamine-, 5-hydroxytryptamine-, and tyrosine

hydroxylase-immunoreactive neurons and nerve fibers indeveloping rat brain J Comp Neurol 347: 101–114

Varga C, Horvath K, Berko A, Thurmond RL, Dunford PJ, Whittle BJ(2005) Inhibitory effects of histamine H4 receptor antagonists onexperimental colitis in the rat Eur J Pharmacol 522: 130–138.Vermeulen W, De Man JG, Pelckmans PA, De Winter BY (2014).Neuroanatomy of lower gastrointestinal pain disorders World JGastroenterol 20: 1005–1020

Wood JD (2004) Enteric neuroimmunophysiology andpathophysiology Gastroenterology 127: 635–657

Yu B, Shao Y, Zhang J, Dong XL, Liu WL, Yang H et al (2010).

Polymorphisms in human histamine receptor H4 gene areassociated with atopic dermatitis Br J Dermatol 162: 1038–1043

Yu S, Stahl E, Li Q, Ouyang A (2008) Antigen inhalation inducesmast cells and eosinophils infiltration in the guinea pig esophagealepithelium involving histamine-mediated pathway Life Sci 82:324–330

Zhang C, Xiong Y, Li J, Yang Y, Liu L, Wang W et al (2012) Deletion

and down-regulation of HRH4 gene in gastric carcinomas: apotential correlation with tumor progression PLoS ONE 7: e31207

Zhu Y, Michalovich D, Wu H, Tan KB, Dytko GM, Mannan IJ et al.

(2001) Cloning, expression, and pharmacological characterization

of a novel human histamine receptor Mol Pharmacol 59: 434–441.Zimmermann AS, Burhenne H, Kaever V, Seifert R, Neumann D(2011) Systematic analysis of histamine and N-methylhistamineconcentrations in organs from two common laboratory mousestrains: C57Bl/6 and Balb/c Inflamm Res 60: 1153–1159

BJP A Deiteren et al.

1178 British Journal of Pharmacology (2015) 172 1165–1178

This review is based on the JR Vane Medal Lecture presented at the BPS Winter Meeting in December 2011 by J.C McGrath.

A recording of the lecture is included as supporting information It covers his laboratory’s work from 1990 to 2010 on the

several cell types in arteries, not only on medial smooth muscle, but also on adventitial, endothelial and nerve cells; (ii) all

capable of function and not merely expressed (iii) all of these cell types can take up an antagonist ligand into the intracellularcompartments to which endocytosing receptors move; (iv) each individual subtype can exist at the cell surface and

intracellularly in the absence of the other subtypes As functional pharmacological experiments show variations in the

involvement of the different subtypes in contractions of different arteries, it is concluded that the presence and disposition of

different cell types, suggest that differences between the distribution of subtypes in model systems do not directly correlatewith those in native tissues This review includes a historical summary of the alternative terms used for adrenoceptors

(adrenergic receptors, adrenoreceptors) and the author’s views on the use of colours to illustrate different items, given hispartial colour-blindness

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://

www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the Concise Guide to PHARMACOLOGY 2013/14 (Alexander et al., 2013b).

British Journal of Pharmacology (2015) 172 1179–1194 1179

© 2014 The British Pharmacological Society

A catecholamine, noradrenaline, is released from

sympa-thetic nerve endings to act as neurotransmitter, in the brain

and almost all peripheral organs Noradrenaline and another

catecholamine, adrenaline, are released from adrenal

medul-lary cells into the bloodstream to act as endocrine hormones

They activate a family of receptor molecules (the

adrenocep-tors; a note on terminology is included at the end of this

article) located on the cells of the target organs to initiate

physiological signals that regulate almost all organ systems in

the body

For almost 120 years, the chemical and biological

prop-erties of the adrenoceptors and their natural activators have

provided a central strand in the interaction between

physi-ology and drug discovery, each informing the other

However, the story is not complete A gulf remains between

our knowledge of the identity and structure of the main

chemical players and our understanding of how the receptors

operate physiologically One reason for this is our limited

understanding of the localization and distribution of the

receptors, in general, in heterogeneous natural tissue, and

how these relate to the nervous and hormonal regulation in

which the receptors participate Specifically, we do not have a

complete conceptual view of the localization of

adrenocep-tors on cells within tissues even though we are starting to

understand how they behave at the individual molecular and

cellular level

Adrenoceptors were defined initially by pharmacological

techniques that exploited the differences between various

drugs, hormones or neurotransmitters to mimic or block

these actions at the different receptors After the initial

iden-tification of two divisions,α and β, a family of nine

adreno-ceptors was defined by both functional and genetic means

Theα-family was split into two subfamilies α1andα2, each

comprising three members, with theα1subfamily comprising

three members,α1A,α1Bandα1D There was an original

con-vention to make the alphabetical subscript lower case, for

example, α1a, for receptors defined genetically, and upper

case, for example,α1A, for those defined by functional

phar-macology However, this convention has not been

main-tained In this review, the upper case subscript will be

employed throughout the text The focus of this review is on

α1-adrenoceptors

The theme for the lecture on which this review is based

was ‘Black Boxes’, indicating both the opaque nature of the

operation of receptors and the black background

ubiqui-tously employed in fluorescence imaging studies

Questions pursued

1 Where in blood vessels are α1-adrenoceptors located; in

which tissue layers and on which cell types?

2 Areα1-adrenoceptors at the cell surface capable of binding

ligands? Does this apply to all three subtypes? Does it

apply to all cell types?

3 Can vascular smooth muscle cells take up antagonist

ligands into the intracellular compartments to which

endocytosing receptors move Does this apply to all three

subtypes of adrenoceptor and all cell types expressing thereceptors?

4 Are the subtypes dependent upon each other for theircellular disposition? Is there interaction between subtypes

in their tissue expression?

Our laboratory has addressed these questions using rescent ligands and microscopy, to visualize the location ofreceptors in relation to tissue cell membrane and sub-cellularstructures, particularly in small arteries and particularly fortheα1-adrenoceptors, with excursions into the study ofβ- and

fluo-α2-adrenoceptors The most surprising and interestingoutcome was that receptors were found to be located onmany cell types that had not previously been considered

targets for adrenoceptor agonists or blockers (McGrath et al.,

2005; Daly and McGrath, 2011)

Combining knockouts with pharmacology to overcome poor drug selectivity

We decided in the early 1990s to approach the problem ofidentifying the subtypes of α-adrenoceptors involved inresponses to adrenoceptor agonists and sympathetic nerves,

by combining the use of the most ‘selective’ agonists andantagonists and of knockouts of the receptor subtypes Thisarose from earlier studies where the focus was to discrimi-nate α1 from α2-adrenoceptors; this pharmacological dis-crimination became more complex once the existence ofsubtypes started to emerge from functional and bindingproperties of both α1 and α2 families (McGrath, 1982;

McGrath and Wilson, 1988; Brown et al., 1990; Wilson

et al., 1991).

We employed mainly arteries in these studies and trated mainly on the threeα1-adrenoceptors, for which wewere eventually able to use all three single knockouts, allthree double knockouts and triple knockout animals Ourfirst example was actually a functional knockout of an

concen-α2-adrenoceptor generously provided by Dr Lee Limbird

(MacMillan et al., 1998), which we used to identify theα2Aadrenoceptor subtypes involved in endothelial vasodilatory

-responses to noradrenaline (Shafaroudi et al., 2005) We also

applied these techniques to prostatic tissue Benign prostatichyperplasia is currently the major therapeutic target for

α1-adrenoceptor antagonists, the other being blood vessels inhypertension and heart failure (still used in some countries;

Mackenzie et al., 1999; 2000; McGrath et al., 1999a,b) We

also applied them to hepatic tissue, which was the onlyexample we identified in which the knockouts showedevidence of substitution of another subtype; when theα1B-adrenoceptor was eliminated, it was ‘replaced’ by α1A-

adrenoceptors (Deighan et al., 2004).

In parallel with this, we employed fluorescent ligands to

identify the receptors (Daly et al., 1992; Daly and McGrath,

2003) This allowed us to map their presence and relate this tothe responses that we saw in arteries On the basis of previousliterature, using more indirect methods, on the associationbetween autonomic nerves and post-junctional receptors

(Hansen et al., 1999), we expected to find that receptors on

1180 British Journal of Pharmacology (2015) 172 1179–1194

smooth muscle would be located preferentially near to the

sympathetic nerves This turned out not to be the case: they

were fairly uniformly distributed over the muscle cell

popu-lation and throughout the depth of the muscle layer,

accord-ing to our analysis

In summary, we found three α1-adrenoceptor subtypes

throughout the arterial smooth muscle layers in both large

and small arteries but, unexpectedly to us, they were found

on the endothelial cells and also on several cell types of the

adventitial layer This heterogeneity of cell types containing

the receptors suggested a potential for heterogeneity of

physi-ological responses to catecholamines or adrenoceptor

ago-nists We pursued this idea using knockouts and selective

agonists and antagonists During the ensuing 20 year period,

several other groups also identified adrenoceptor-mediated

responses arising from adventitia and endothelium (Filippi

et al., 2001; Somoza et al., 2005; de Andrade et al., 2006;

Bulloch and Daly, 2014) allowing us to build up a more

complex scenario for adrenoceptor action in blood vessels

than hitherto existed (McGrath et al., 2005; Daly and

McGrath, 2011)

Distribution of α1-adrenoceptors

throughout the vascular wall

Because we were interested in identifying the location of

adrenoceptors in relation to noradrenergic nerves, in our first

experiments, we combined a fluorescent ligand with the

visu-alization of noradrenaline Rabbit saphenous arteries, which

contract toα1-adrenoceptor agonists (Dunn et al., 1989) were

incubated in the ligand, washed to remove non-bound

ligand, then freeze dried and exposed to formaldehyde to

convert noradrenaline to a fluorescent compound (Carlsson

et al., 1961; Gillespie and McGrath, 1974) The tissue was

then sectioned and the ligand and the sites of ‘noradrenaline

stores’ photographed separately on a conventional

fluores-cence microscope using appropriate excitation/emission

wavelengths The two images were then combined (Figure 1a;

McGrath et al., 1996a).

Unexpectedly, we found both diffuse and punctate ligand

fluorescence on virtually all tissues making up the artery It

was on smooth muscle cells, where we expected to find it,

except that it showed no preferential location near to nerves

It was also on endothelial cells, which was surprising, but

not entirely unexpected, since functional endothelial

α2-adrenoceptors had been demonstrated on blood vessels

much earlier (Miller and Vanhoutte, 1985; Angus et al., 1986).

In addition, it was found on cells in the adventitia, which was

a complete surprise (McGrath et al., 2005) Binding was

spe-cific as it was absent in the presence of the usual very high

concentration of phentolamine (10μM) as routinely used in

radioligand binding, but we were rather skeptical of whether

we were actually seeing the localization of receptors Precise

localization to a particular part of the cell was not possible

due to the damage done by the freeze drying process so we

were not surprised that our images did not show binding

confined to discrete plasmalemmal membranes, which would

also have been our expectation There seemed to be plenty of

fluorescence within the various cell types, which was also not

our expectation as we anticipated it to be confined to smoothmuscle cells

Later, once we had improved on resolution by usinggentler fixation, or none at all, and confocal microscopy, wewere able to confirm the rather surprising results inFigure 1a Figure 1b shows a later cross-sectional imagewhere the cellular detail in each layer of a small mesentericartery is clear, showing the presence of receptors both on thecell surface and at intracellular sites Figure 2 takes this astage further, showing the presence of binding throughoutthe vessel wall This figure shows ‘optical slices’ through a 3Dreconstruction of the artery so that several layers of the vesselare shown Binding was intense on the nerve plexus and onsome adventitial cells, with lower levels inside the smoothmuscle layer

It was obviously necessary to validate the techniqueand understand how the ligand accessed various parts of thecells

Figure 1

Examples of binding of the α1-adrenoceptor ligand prazosin (QAPB) to arteries A: Rabbit saphenous artery.Noradrenaline-derived fluorescence shows the sympatheticnoradrenaline-containing nerves (white) at the adventitia-medialborder The green colour, which appears in all parts of the artery,represents the binding of BODIPY-FL-prazosin (QAPB) In the medialsmooth muscle layers, this revealed punctate binding orientated in adirection consistent with the arrangement of smooth muscle cells;this was where binding was expected All of the green colour isligand-dependent The ligand-free control had no green at all Thus,the binding to cells other than smooth muscle was unexpected inthis early example of the fluorescent ligand-binding technique B: Ratthird-order mesenteric artery, isolated then fixed under 70 mmHgpressure, fixed with formaldehyde then sectioned and subsequentlyexposed to QAPB This shows binding to all smooth muscle cells andmost endothelial cells and binding to cells in the adventitia (mac-rophages and fibroblasts) The image is a collage and the brighterfluorescence at the upper left sector is not significant, reflecting less

BODIPY-FL-‘fade’ when this part was captured Courtesy of Jose Maria GonzalezGranado (PhD Thesis, Autonomous University of Madrid, 2004) Thislater example, employing confocal microscopy, shows in some detailthat there is ligand binding to various cell types and that this is both

on the surface and inside cells

British Journal of Pharmacology (2015) 172 1179–1194 1181

Validation of fluorescent ligand

binding to α-adrenoceptors

First, we had to validate the technique for the main ligand

that we used, QAPB (quinazoline piperidine Bodipy) This is

marketed by Molecular Probes, Inc (Eugene, OR, USA) under

the name ‘Bodipy-prazosin’, but we felt that this name was

misleading as the fluorescent moiety is substituted for the

furan group that makes this particular compound ‘prazosin’

rather than the many other members of the family such as

doxazosin or alfuzosin, which have different substituents at

this point So, it really represents this whole family of

antago-nists rather than prazosin (Figure 3)

We decided to use a combination of cell culture with

recombinant adrenoceptors and studies of tissues, the

adrenoceptor pharmacology of which we were familiar,

such as small mesenteric arteries and rat anococcygeus

muscle Because these tissues had been suggested to have

α1A-adrenoceptors from their functional pharmacology

(Mackenzie et al., 2000), we employed cells that we

under-stood had been transfected withα1A-adrenoceptors However,

at a late stage in our investigation, we tested the

ligand-binding properties of the receptors using a series of selective

α1-adrenoceptor antagonists and this showed that the

recep-tors had the pharmacological properties ofα1D-adrenoceptors

When we traced back the history of the cells, we found that

they had been transfected at a time when this clone had been

named asα1Abut it had subsequently been reclassified asα1D

(see Hieble et al., 1995) Fortunately, we discovered this

before submitting the data for publication but it must have

seemed odd to the readers that we chose this receptor rather

than theα1A-adrenoceptors present in our ‘real tissue’

com-parator (Fig 4)

These early studies also showed us an unexpected erty of the fluorescent ligand It fluoresced only when boundand not when free in solution This allowed us to studybinding in live cells in real time and at equilibrium betweenthe free ligand and that bound to the cell This gives anadvantage over conventional ligand binding, which employs

prop-‘post-wash’ measurement of bound ligand: this means thatthe cells or tissues are exposed to ligand to allow binding butthen have to be washed to remove unbound ligand; it isnecessary with radioligands and with compounds that arefluorescent in solution because they would give too muchbackground counts, swamping the bound ligand’s measure-ment With our method, extracellular non-fluorescent com-pound is not detected so we can carry out measurements inreal time and at equilibrium

In the first study, we concentrated on quantifying thebinding in different visual ‘domains’ Most binding was

‘diffuse’ but some was in ‘clusters’ However, both had similarbinding characteristics when we quantified the fluorescence

at different ligand concentrations and this was similar inblood vessels and in cells containing the recombinant recep-tors Image analysis methods had to be devised for quantify-ing receptor distribution at the subcellular level in isolated

cells and tissues (Luo et al., 1998; Shang et al., 2000a,b,c) We

then made the critical observation that antagonists could

Figure 2

Confocal images of BODIPY-FL-prazosin (QAPB) binding to

adventi-tia, nerve plexus at the adventitial-medial junction and within the

media in different focal planes of rat third-order mesenteric artery

Bottom right image is an extended focus showing binding in all

layers Courtesy of Anna Briones and Elisabet Vila

sub-515 nm The compound was obtained from Molecular Probes and islisted in their catalogue as ‘BODIPY-FL-prazosin’ but because it lacksthe furan group that defines prazosin, as opposed to other com-pounds that share the quinazolinyl piperazine group, such as doxa-zosin, we refer to it by an acronym, ‘QAPB,’ derived from its chemicalname (quinazolinyl piperazine borate-dipyrromethene) Reproduced

with permission from Daly et al (1998).

1182 British Journal of Pharmacology (2015) 172 1179–1194

prevent ligand binding in real time (Daly et al., 1998;

Mackenzie et al., 2000) By then, we had started to use

con-focal microscopy, but because the cultured cells were very

thin, we could not be certain whether the fluorescence and,

therefore, the receptors, were on the surface or inside the cell

Intracellular binding to adrenoceptors

As our studies progressed on cell culture of different cell

types, it became clear when visualizing thicker cells that the

clusters were mainly intracellular (Mackenzie et al., 1998;

2000; McGrath et al., 1999a,b) Because by then we were

using living cells, this gave us another series of phenomena

and questions It showed that binding could occur to

intrac-ellular receptors – but how did the ligand access these? If an

antagonist ligand could access them what would this mean

for the pharmacology of antagonists? Would the kinetics vary

according to whether the antagonist could enter the cell?

Would this also vary with the relative surface/intracellular

location of different receptors or receptor subtypes? We were

in uncharted territory here

Around that time, various suggestions emerged

concern-ing the differential localization of receptor subtypes that

could affect their pharmacology Our subsequent

investiga-tions did not validate many of these concepts, but they gave

us hypotheses to test and disprove

Hirasawa et al (1997) hypothesized that, in COS-7 cells, a

relatively greater proportion of α1A receptors were located

intracellularly than the proportion ofα1Breceptors that were

located intracellularly If generally true, this might influence

the access to ligands and hence the ability to activate or block

the receptors Hirasawa et al (1997) wrote ‘Together, the

results showed that a hydrophilic alkylating agent CEC erentially inactivatesα1-adrenoceptors on the cell surface irre-spective of their subtype, and that the subtype-specificsubcellular localization rather than the receptor structure

pref-is a major determinant for CEC inactivation of α1Badrenoceptors.’ This provided a possible explanation for whyCEC could blockα1B-adrenoceptors more effectively thanα1A-adrenoceptors However, this was incompatible with the sub-sequent finding that α1D-adrenoceptors, which are CEC-susceptible, are predominantly intracellular (Chalothorn

-et al., 2002; Hague -et al., 2003) However, with our approach,

we could not find evidence of this either In our hands, innative tissues and in cell cultures, all threeα1-adrenoceptorscould be found to be at both plasmalemmal and intracellularsites

All three human α1-adrenoceptors can

be plasmalemmal or intracellular

When we looked at the distribution of the three humanrecombinant subtypes in rat-1 fibroblasts, we did not find anymajor differences between them Furthermore, we madesimilar observations in various tissues including human pro-

static smooth muscle and blood vessels (McGrath et al., 1999a,b; Mackenzie et al., 2000) We also compared fluores-

cent ligand binding to intact cells and to membranes andfound no differential effects between the three subtypes Thefluorescent ligand QAPB also had similar affinity for all threesubtypes, which means that it labels all three subtypes

equally (Mackenzie et al., 2000; Fig 5) We also demonstrated

that the recombinant receptors were functional and hadappropriate agonist/antagonist pharmacology, using Ca2+sig-

nalling (Pediani et al., 2000).

As far as we could see, when we expressed the subtypes indifferent cell types, we saw similar distributions with a lot ofligand binding in the endoplasmic reticulum, which hadvarious degrees of ‘clustered’ or ‘punctate’ nature according

to the cell type

We were also confident that our ligand was getting access

to all the receptors because the distribution was similar tothat found when we labelled the receptors with GFP and the

ligand colocalized with the GFP (Pediani et al., 2005).

We showed that α1A-adrenoceptor agonist/antagonistpharmacology ‘worked’ at the single cell level with therecombinant receptors, using intracellular calcium as thereadout We even showed that phenylephrine could displacefluorescent antagonist from the cells (these experiments didnot have 3D spatial resolution so we could not distinguishwhether displacement was both from inside and from on the

cell surface; McGrath et al., 1995; Mackenzie et al., 1999; Pediani et al., 2000) This gave some hope that, in future

experiments, we might identify receptors activated by thenoradrenaline released from nerves and localize them by itsdisplacement of ligand However, we have not yet succeeded

in this because our ligands have an off-rate which is too slow,that is, they dissociate too slowly to be replaced by thetransiently high local concentration of neurotransmitternoradrenaline

Figure 4

Early QAPB binding experiments Top: single cell transfected with

α1D-adrenoceptors (QAPB 1 nM) Bottom: sheet of smooth muscle of

rat anococcygeus muscle (QAPB 10 nM) (Daly et al., 1998) J

Phar-macol Exp Ther 1998 286(2): 984–990, permission applied for

British Journal of Pharmacology (2015) 172 1179–1194 1183

The findings reported in Mackenzie et al (2000) were also

seminal to our subsequent work in that they established

the selectivity of the key non-fluorescent antagonists RS100

329 (selective against α1A, compared with α1B and α1D

-adrenoceptors) and BMY7378 (selective against α1D,

com-pared with α1A andα1B-adrenoceptors) We went on to use

both of these antagonists when imaging subtypes and when

correlating these data with functional pharmacology

Ligand is internalized by binding to

α1-adrenoceptors undergoing

spontaneous endocytosis

By 2000, we still did not understand how the ligand accessed

the intracellular receptors However, Morris et al (2004)

showed that α1A-adrenoceptors spontaneously internalise,

which generated a new concept (‘constitutive

internaliza-tion’) for the distribution of receptors and placed them in a

dynamic context Later, Stanasila et al (2008) demonstrated

this for recombinant α1B-adrenoceptors though the two

groups had contrary evidence on which, between α1A

-adrenoceptors and α1B-adrenoceptors, internalized to the

greater extent

Meanwhile we had been investigating the hypothesis that

the fluorescent ligand gained entry to the cell after binding to

the receptor (Pediani et al., 2005) but Morris et al (2004)

published the idea of constitutive internalization before we

published our work, so we confirmed their finding by a

dif-ferent approach They had dealt only with internalization of

the receptors We were interested in seeing whether the

ligand could be taken into the cell attached to the receptorand where it would go inside the cell

We had developed a hypothesis of how the ligand ended

up inside the cell This was based on observation of thedevelopment of fluorescence when cells were exposed to thefluorescent ligand It could be seen to bind first to the plas-malemmal membrane then to intracellular structures andfinally, when the ligand was removed, the process reversed,the intracellular fluorescence disappearing first Our hypoth-esis was that the ligand binds to the receptor on the externalcell surface, then, when the receptor and its surroundingmembrane undergo spontaneous endocytosis, the ligand-receptor complex is internalized; furthermore the bindingsite would now be on the inside of the endocytic vesicle Thevesicles then move inside the cell and fuse with the endo-plasmic reticulum; this explains the punctate nature of thefluorescence and how we could see fluorescent structuresmoving back and forward inside the cells Subsequently thestructures moved back to the cell surface to undergo exocy-tosis, allowing the ligand to escape and leaving the receptorready to operate once more in its extracellular location.The alternative possibility was that the lipophilic ligandentered the cell by diffusing through the plasmalemmalmembrane then found and bound to receptors that werealready inside the cell This requires that the ligand pen-etrates two membranes: the plasmalemmal membrane, thenthe membrane of the intracellular organelle since the recep-tors here are inward facing Spontaneous endocytosis ofmembrane-bound receptors provides a much simplerexplanation

The key experiment that proved our hypothesis and proved ‘entry by diffusion’ was to show that entry of the

ligand into the cell was absent in β-arrestin-deficient cells.

β-arrestin is necessary for endocytosis The

β-arrestin-deficient cells had been transfected withα1B-adrenoceptors

labeled with GFP, which were visible both inside the cell and

on the surface, so if the ligand could enter the cells it would

have bound to these intracellular adrenoceptors However,

the extracellularly applied ligand colocalized with

GFP-labelled receptors on the cell surface but did not bind to the

GFP-labelled receptors inside the cell This showed that the

ligand could not pass the plasmalemmal membrane unless

endocytosis of the receptors took place: endocytosis was its

only route into the cell This idea was confirmed by blocking

endocytosis with concanavalin A or hyperosmotic sucrose

Importantly, we also showed that the ligand partially

colo-calized withβ-arrestin in recycling and late endosomes,

indi-cating receptor transit without destruction (Pediani et al.,

2005)

We went on to follow the time course of internalisation

with the fluorescent ligand, showing that it took

approxi-mately an hour for all receptors in the cell to equilibrate with

the ligand and a similar time to leave the cell when ligand

was removed This provided an insight into the rate of cycling

of the receptors and changed our view of the stability of the

plasmalemmal membrane receptor population This is

illus-trated in simplified cartoon form in Figure 6

We could not contribute to the argument about whether

the subtypes were differently involved in agonist-induced

internalisation because our antagonist ligand, at any

concen-tration that is useful for receptor localisation, blocks

responses to even high concentrations of phenylephrine

(Pediani et al., 2000) However, by labeling the surface

recep-tors with an antibody that only fluoresces once internalised,

colleagues were able to show that bothα1A- andα1B-subtypescould be internalised after activation by high concentrations

of phenylephrine with only a small difference in time course

(Flacco et al., 2013; Perez-Aso et al., 2013; Segura et al., 2013).

Locating adrenoceptors in native tissues

Our work on recombinant receptors and then isolated ciated cells was merely background for our main interest oflocating adrenoceptors in native tissues using fluorescentligands In addition to our initial expectation that receptorswould be seen on the plasmalemmal membranes, it was ofinterest to know whether the ligands could penetrate cells innative tissues, as was the case in cell cultures These issues arecompletely different in fixed and live cells

disso-For fixed cells or tissues, the problems are similar to thoseencountered in the immunohistochemical localization ofreceptors The ligand or antibody must be able to penetrate tothe binding site inside the tissue, to the plasmalemmal mem-brane and to the intracellular organelles This can beachieved by chemical treatment to make membranes porousand/or by thin sectioning to expose the insides of the cells

We could achieve this as shown for example by Figure 1a with

freeze drying or Figure 1b with fixation of the in vitro surized vessel (see also Miquel et al., 2005) However, even our

pres-early experiments had shown that in live cells, the QAPBligand could bind to all receptors, as shown by colocalizationwith GFP-labelled receptors, so we decided to continue using

unfixed ‘live in vitro’ tissues.

Cell surface receptors

Extracellular facing binding sites

Agonist or antagonist ligands

Endocytosis

Exocytosis

Receptors on intracellular organelles; binding sites facing inwards

Nucleus Transcytosis

Constitutive recycling of α1A-AR in rat-1-fibroblasts

Figure 6

Cartoon of constitutive recycling ofα1-adrenoceptors Starting from the left, ligands bind to the outward facing recognition site on the receptorand are then taken into the cell by spontaneous endocytosis, becoming trapped inside the intracellular organelles These organelles then moveback to the cell surface where the receptors become re-incorporated into the plasmalemma and the ligands are released back into the extracellularspace This accounts for the punctate nature of the labelling of the intracellular receptors and the reversibility of the binding Our estimate of the

entire cycle is around 1 h and the receptors spend approximately one third of the time at the surface Principles from Pediani et al., 2005.

British Journal of Pharmacology (2015) 172 1179–1194 1185

Our basis, since no intracellular binding is seen in the

absence ofβ-arrestin, is that, in live cells, the ligand cannot

penetrate the cell unless carried in bound to the receptor

Thus, intracellular relocation in live tissue would happen

only if receptors start off at the surface, bind ligands and

spontaneously internalize, taking the ligand with them If we

are to use the fluorescent ligand to compare the locations and

properties of different subtypes ofα1-adrenoceptor, then we

need to know whether each subtype has the appropriate

properties, that is, located at the cell surface and able to

spontaneously internalize, as shown in the recombinant cells

(Pediani et al., 2005) The question was now whether the

same would occur in native tissues There was reason to

believe, from the literature, that the subtypes might differ,

favouring different ratios of surface to intracellular location

(Hirasawa et al., 1997; Chalothorn et al., 2002; Hague et al.,

2003) or having different tendencies towards spontaneous

internalization (Morris et al., 2004; Stanasila, 2008) However,

as discussed earlier, these authors found contradictory results

not parallel ones

α1-adrenoceptor subtypes in vascular

smooth muscle

Our work on vessels from a small artery (mesenteric first

order) and a large artery (carotid) from wild-type (WT) mice

showed images for smooth muscle cells similar to those that

we saw earlier in rat mesenteric third-order arteries (Figure 2)

and rat aorta (Miquel et al., 2005).

The next questions were whichα1-adrenoceptor subtypesare present and, if more than one subtype, do they differ intissue or cellular location and do they handle the liganddifferently?

To identify the subtypes, we combined two approachesaimed at isolating subtypes: selective ligands and selectivereceptor subtype knockout In each case, we sought tissuelocation microscopically and quantified this from averagefluorescence intensity of selected equivalent areas of smoothmuscle First, we identified the receptors to which the ligandsbound using subtype-selective pharmacological agents, toprevent binding selectively; secondly, we used receptorknockout strains, to isolate each subtype physically by elimi-nating the others; subsequently we then combined the twomethods for cross validation

Elimination of subtype binding by knocking out subtypes genetically

We generated mice with three single knockout, three doubleknockout and the triple knockout of the α1-adrenoceptorsubtypes The triple KO showed no QAPB fluorescent ligandbinding, acting as an excellent negative control The fluores-cent image in the presence of QAPB was, literally, a ‘blackbox’ Compared with the WT, each single KO showed reducedintensity of fluorescence and this was reduced further in eachdouble KO (Figure 7) This was demonstrated quantitatively

in images of the smooth muscle layer (Methven et al.,

Elimination of subtype binding with

antagonists to identify receptor

subtypes in smooth muscle of arteries

from WT mice

Using recombinant receptors expressed in cell culture, we

developed a protocol using selective antagonists to occlude

binding of particular subtypes We would have preferred to

have selective antagonists for each of the three subtypes

However, despite trying various putatively selective α1B

-adrenoceptor subtype-selective antagonists, we were unable

to find one that, in our hands, was selective, using

recombi-nant human α1-adrenoceptor subtypes Therefore, we used

only anα1A-adrenoceptor selective ligand, RS100 329 and an

α1D-adrenoceptor selective ligand, BMY7378 These ligands

were chosen to be potent as well as selective as they need to

bind more effectively than QAPB at the concentrations

employed We first showed that QAPB binds with similar

affinity to all three subtypes of receptor (Mackenzie et al.,

2000) Both selective antagonists were then shown to have

100-fold selectivity for their high affinity subtype compared

with the other two α1-adrenoceptor subtypes (Mackenzie

et al., 2000) Our criterion for identifyingα1B-adrenoceptors

was susceptibility of QAPB binding to an α1

-adrenoceptor-selective concentration of prazosin and resistance to the

α2-adrenoceptor antagonist, rauwolscine, coupled with

resistance to selective concentrations of RS100 329 and

BMY7378

Having developed the protocol on the known receptorsubtypes in culture, we applied it to analyse the unknownreceptor subtypes in the smooth muscle of blood vessels Wedid this first on smooth muscle cells isolated from prostate

and blood vessels (Mackenzie et al., 1998; 1999) then applied

it systematically to intact blood vessels

The results were essentially similar in the two types ofartery that we employed: carotid, representing a large con-ducting artery, and first-order mesenteric, representing a

‘resistance’ artery Theα1-adrenoceptor selective antagonist,prazosin, eliminated all fluorescence, whereas equivalentconcentrations of theα2-antagonist, rauwolscine, were inef-fective, indicating that all binding was toα1-adrenoceptors

(Fig 8; Methven et al., 2009a,b; data for the α1AD KO,Methven, unpublished)

Either of the subtype-selective antagonists (RS100 329and BMY7378.) reduced fluorescence intensity and the twocombined had a greater effect than each alone It wasassumed that α1B-adrenoceptors were responsible for theremaining fluorescence that was sensitive to prazosin (Fig 8).The simplest explanation for these observations is that allthree subtypes are present in arterial smooth muscle Therewas no detected difference between the antagonist treat-ments except for overall fluorescence intensity, that is, therewas no evidence for differences in distribution between sub-types This was not necessarily expected from functional data,since previous work, including our own pharmacologicalassessment of responses to the α1-agonist phenylephrine

(Daly et al., 2002; Deighan et al., 2005) had shown a tendency

Figure 8

Antagonist drugs identify binding toα1-adrenoceptor subtypes Images of QAPB binding to smooth muscle layers of mesenteric arteries from WTmouse and the two double knockouts, BD and AB Prazosin was effective in blocking all binding and rauwolscine blocked none, validating QAPBbinding asα1- and notα2-adrenoceptors The selectiveα1Dantagonist BMY 7378 reduced binding in WT and abolished it in ABKO when only the

α1D-adrenoceptor was present Conversely theα1Aantagonist RS100 329 reduced binding in the WT and abolished it in the BD KO when only the

α1A-adrenoceptor was present This validates the selective antagonists and shows that the WT harbours all three subtypes (the α1Bbeing

responsible for the prazosin-sensitive but RS and BMY-resistant binding in the WT) Based on Methven et al., 2009a,b.

British Journal of Pharmacology (2015) 172 1179–1194 1187

towards greater involvement of α1D-adrenoceptors in large

arteries and ofα1A-adrenoceptors in small arteries However,

previous work had indicated that mRNA for all three subtypes

could be detected (Hrometz et al., 1999) So we could now

confirm this for the protein and show that it was functional

up to the point of ligand binding

α1-adrenoceptor ‘knockouts’

Combining the selective antagonists with the knockouts had

the anticipated effects thereby confirming the selectivity of

the antagonists, an important finding In the single or double

knockouts, the selective antagonists reduced binding only in

the examples where their ‘selected’ subtype was present, for

example, BMY 7378 reduced fluorescence in WT, α1A

-adrenoceptor knockout (A KO),α1B-adrenoceptor knockout (B

KO) andα1A andα1Breceptor knockout (AB KO), but not in

α1D-adrenoceptor knockout (D KO) or the other double

knockouts (AD KO, BD KO)

The double knockouts provided the opportunity to

observe subtype distribution when the subtypes were present

individually, as for the recombinant receptors in cell culture,

which were shown to have different distributions (Hague

et al., 2003) However, we found no differences in the surface

or intracellular distribution of ligand binding between cells

that expressed the different subtypes Each was found both

on the surface and intracellularly (Fig 9)

This suggests that the ligand was handled similarly by all

subtypes, each of which must be expressed on the cell surface

and capable of spontaneous internalization, in order to

achieve the distribution at equilibrium that we saw Thisindicates that the differential distribution of subtypes seen inmodel systems does not apply in arterial smooth muscle.Furthermore, internalization of ligand occurred in everynative cell type that we investigated: hepatocytes, fibroblasts,endothelial cells, as well as smooth muscle from blood vesselsand prostate

Only one difference in fluorescence distribution wasdetected visually: in the BD KO, in which only α1A-adrenoceptors were present, fewer cells showed fluorescence(Fig 10) In other words, the receptor population was moreintermittent across the smooth muscle cell population inthese animals than when the other two subtypes wereexpressed alone This was reinforced by an unpublishedobservation made earlier: in a mouse strain in which eitherthe α1A-adrenoceptor or the α1B-adrenoceptor was labelled

with GFP (Papay et al., 2004; 2006), we had found thatα1Aadrenoceptors were expressed intermittently in mesentericarteries, whereasα1B-adrenoceptors had a faint but more evendistribution We did not publish these data, since we felt thatthe expression level of bothα1A- andα1B-adrenoceptors wastoo low in arteries to be usefully demonstrated by its fluores-cence but, in the light of the intermittent distribution of

-α1A-adrenoceptors in the BD KO, we now feel that theseobservations reinforce each other (In the published workusing theseα1-adrenoceptor-GFP fusion mouse strains immu-nohistochemistry with antibodies to GFP was used to mapthe α1-adrenoceptor subtypes in the brain because the GFP

fluorescence was insufficient for visualization; Papay et al.,

2004; 2006)

Adrenoceptors on endothelium

The idea that adrenoceptors are present on the endotheliumand mediate vasodilator responses is long known (Miller and

Vanhoutte, 1985; Angus et al., 1986) but receives little

atten-tion in the literature There is little informaatten-tion about thelocation of these receptors except for autoradiography, whichhas rather low resolution (Summers and Molenaar, 1995)

Figure 9

Optical isolation of single smooth muscle cells in mouse carotid

arteries shows that QAPB binding looks similar forα1Aandα1D

sub-types Based on Methven et al., 2009a,b.

Figure 10

High magnification of smooth muscle in mesenteric arteries from the

α1-BD KO (on right), which has onlyα1A-adrenoceptors, shows thatlabelling is intermittent compared with WT (on left) Reproduced

from Daly et al (2010).

1188 British Journal of Pharmacology (2015) 172 1179–1194

We first visualizedα1A-adrenoceptors on the endothelium

inadvertently We were seeking to locateα2-adrenoceptors on

aortic endothelium in a study of noradrenaline-induced

relaxation of mouse aorta We used the fluorescent ligand

QAPB on the basis that it has affinity forα2-adrenoceptors,

albeit at higher concentrations than for α1-adrenoceptors,

and in the mistaken belief that there would be no

α1-adrenoceptors on the endothelium (Shafaroudi et al.,

2005) In the event, we found a great deal of binding of QAPB

that was blocked by prazosin We could not identify the

endothelial binding site as a singleα1-adrenoceptor subtype

but did eliminate it by using aortae fromα1B-KO mice (in lieu

of an α1B-adrenoceptor antagonist) together with the same

α1A- andα1D-adrenoceptor antagonists that we had employed

to analyse smooth muscle receptors In this preparation, we

revealed binding to aα2-adrenoceptors that was blocked by

rauwolscine and by genetic elimination of the α2A

-adrenoceptor

We followed this up in more detail and the majority of

endothelial cells showed surface binding of QAPB (Figure 11)

This reinforces the work of several groups who have shown

α1-adrenoceptor-mediated endothelium derived

vasodilata-tion in the rat mesenteric bed and carotid (Filippi et al., 2001;

de Andrade et al., 2006) Thus, the aortic endothelium

pos-sesses all threeα1-adrenoceptor subtypes as well as, at least,

theα2A-adrenoceptor All seem to be involved in the release of

endothelial relaxant factors

Note on colour for comparing localization of different things: α-and β-adrenoceptors

In the course of working with fluorescent ligands, it is table that one will seek to compare localization of differentcell types or multiple ligands, using specific indicators Doingthis properly requires quantitative image analysis, which is

inevi-beyond the scope of this review (but see (Daly et al., 2012) To

illustrate the conclusions in a simple way, different coloursare applied to the images of the different indicators; inmodern microscopes, the signal is an intensity level and has

no actual colour, so any colours can be chosen

I am not well equipped for that since, like around 8% ofmales, I am colour-blind I use a green laser pointer becausethe dots from the red ones are not visible to me [Please bearthis in mind next time you choose a pointer for a presenta-tion.] Like most of those impaired in this way, irrespective oftheir precise colour pathology, my difficulty is with red/greencolour pairs I can see red and green when they are wellseparated but mixtures of these colours are very difficult todistinguish A 50/50 mix of red and green provides yellow sothere are already three colours to deal with Irritatingly, thedefault position of many microscopes is red-green So, Ibelieve we should follow the advice of Clifford Saper, former

Editor-in Chief of the Journal of Comparative Neurology, when

Figure 11

The coloured image in right hand lower panel shows endothelial cells on WT mouse carotid artery attached to folds in the internal elastic lamina(the parallel yellow lines) On the other side of the folds are the cigar-shaped smooth muscle cells: the optical section slices through the folds inthe luminal surface showing alternately, lumen (L), endothelium, lamina, smooth muscle, lamina, endothelium, lumen (L), which is then repeatedthrough several folds There is diffuse binding of QAPB over the surface of endothelial cells and bright yellow spots also near the surface Cell nucleiare labelled blue, which makes it obvious where the endothelial cells are, and QAPB binding is yellow The black and white images are 3Dreconstructions as if the luminal surface is viewed from inside the lumen (L) at right angles (left) or at an oblique angle (top) These are negativeimages where the ligand appears as black dots The endothelial cells appear as objects outside the smooth folds of the intimal elastic lamina (darkgrey) which, in contrast to the coloured section, covers up the smooth muscle layer Images courtesy of Laura Methven

British Journal of Pharmacology (2015) 172 1179–1194 1189

presenting such images He recommends magenta and green

rather than red/green Another possibility is blue/yellow

(Saper, 2007) For both magenta/green and blue/yellow, the

50/50 mix is white, which is easy to detect against the usually

black background

An example of a vascular image showing the separate

localization ofα- and β-adrenoceptors is shown in Figure 12

(from Daly et al., 2010) This is from mesenteric arteries from

the BD KO mouse where theα1A-adrenoceptors have an

inter-mittent distribution among the smooth muscle cells This

allows us to see thatβ-adrenoceptors were located in many

cells that did not contain α1A-adrenoceptors and vice versa.

This is very interesting from the point of view of

understand-ing howα- and β-adrenoceptors interact since this would be

quite different according to whether they are in the same cell

or not This example shows very limited colocalization, that

is, where the two indicators are found in the same pixels of

the image at high enough levels to produce the intermediate

colour A few yellow cells in the red/green image and white

cells in the other two colour combinations show the small

proportion where colocalization could be claimed

This is a good place to end this review of the evidence for

localization of adrenoceptors since the pictures in Figure 12

illustrate the remarkable and, as yet, unexplained, finding

thatα1A- andβ-adrenoceptors populate different cells in

mes-enteric arteries This is all the more remarkable becauseα1A

-adrenoceptors are the ones that are the least disputably

involved in neurovascular transmission, yet they are present

in only a minority of smooth muscle cells The presence ofintercellular gap junctions is essential for the coordinatedfunction of the smooth muscle syncytium

Current understanding of the relationship between the presence of

α1-adrenoceptor subtypes and physiological function in blood vessels

Over many years, estimates of the presence of the three

α1-adrenoceptor subtypes in blood vessels have been made by

a variety of means (mRNA, Western blot and chemistry for the protein) and attempts made to correlatethis with roles for the individual subtypes in vascular regula-

immunohisto-tion (Piascik et al., 1997; Yang et al., 1997; McCune et al., 2000; Chalothorn et al., 2002) The usual assumption has

been that these receptors are expressed on vascular smoothmuscle (though mRNA detection and Western blot do notlocalize this) and that they will mediate vasoconstriction Ingeneral, the presence of all three subtypes has been predictedwhatever the method used Functional responses, usuallysmooth muscle constriction to catecholamines or agonistsurrogates, have been analysed pharmacologically and adiversity of results has been found, normally explained as

‘mainly’ through one subtype, with possible contributionsfrom others

The advances from our work are that we can demonstrate

1 Receptors are present on several cell types in arteries, notonly smooth muscle cells, but also adventitial cells ofseveral types, nerves and probably Schwann cells, andendothelial cells

2 All three receptor subtypes are capable of binding ligands

at the cell surface, strongly indicating that they are capable

of function and not merely present

3 All three receptor subtypes can carry the antagonist ligandinto the intracellular compartments to which endocytos-ing receptors move

4 All of these arterial cell types that express the

α1-adrenoceptors can carry the antagonist ligand into theintracellular compartments to which endocytosed recep-tors move

5 Each individual subtype is capable of existing at the cellsurface and intracellularly in the absence of the othersubtypes and does not require an association with anothersubtype as has been suggested in some heterologousexpression systems

Unquestionably, several differences between the subtypeshave been observed in model systems in relation to sponta-neous or agonist-induced internalization or receptor disposi-tion Yet, the quantitative aspects have not been consistentbetween laboratories Our work shows more similarities thandifferences between the subtypes in cellular disposition andreceptor mobility in several native cell types in the arterialwall Internalization of the ligand occurred in adventitialfibroblasts, hepatocytes and vascular endothelial cells as well

as in smooth muscle of blood vessels and prostate We suggest

Figure 12

Effect of colours chosen for comparing the distribution of ligands for

α1-adrenoceptors (QAPB) and β-adrenoceptors (TMR-CGP12177)

(Baker et al., 2003) Smooth muscle layers of mesenteric arteries from

theα1-BD KO, which has onlyα1A-adrenoceptors (as in Figure 10)

that are not present in all cells From left to right, the colour pairs are

for α-(QAPB)/β-(TMR), respectively, the conventional green/red

(left), the recommended green/magenta (centre) and yellow/blue

(right) In the left image, colocalization shows up as yellow and in the

other images as white The colour-blind author finds it easiest to see

the distinction between the two ligands in the yellow/blue image but

the colocalization is clearer in the magenta/green image Modified

from Daly et al (2010) The conclusion of this illustration is very

straightforward: in this example,α- and β-adrenoceptors are present

in different cells with only a few cells where the two receptor types

are present

1190 British Journal of Pharmacology (2015) 172 1179–1194

that such properties are held in common among the subtypes

and, although they may be capable of being differentially

regulated, this does not seem to occur in the cell types that we

have investigated

It was a notable feature that when the intensity of the

fluorescent ligand was quantified in smooth muscle cells, the

‘amount’ and hence the density of receptors seemed to be

consistent for each subtype and independent of the others

For example, the level of fluorescence to which each selective

antagonist reduced the total was very similar to the level

found when the high affinity receptor for that antagonist was

eliminated genetically, for example, WT vascular smooth

muscle cells after BMY7378, and those from anα1D-KO had

similar fluorescence levels This contrasted with liver where

the WT mouse has only α1B-adrenoceptors but the α1B-KO

expressesα1A-adrenoceptors, which are not detectable until a

mature adult stage of 4 months old, indicating a delayed

compensation (Deighan et al., 2004) In contrast to liver,

vas-cular smooth muscle does not exhibit up-regulation of

α1-adrenoceptor subtypes in response to the absence of other

subtypes All our experiments have been conducted in 4

month old (mature) mice, unlike many other studies in the

literature

The next question is whether the presence of a subtype is

a predictor of its role in vascular function Our functional

data concur with the literature that α1D-adrenoceptors are

‘dominant’ in the carotid artery, whereasα1A-adrenoceptors

play a more ‘minor’ role; conversely,α1A-adrenoceptors

domi-nate in the mesenteric artery, andα1Dreceptors play a more

‘minor’ role None of this functional evidence correlates

directly with the demonstrable expression per se of all three

subtypes in each artery, that is, there is no evidence of ‘more’

receptors of the subtype that plays the ‘major’ role The

inescapable conclusion is that the presence of a subtype in an

artery is not a predictor of its contractile function per se There

are presumably other regulatory factors that determine the

contribution of the various subtypes

Furthermore, the presence of all three subtypes on cell

types other than smooth muscle serves to complicate matters

further, since we cannot yet predict the ‘functional’ roles of

the subtypes in these cells and, therefore, how they might

interfere with the contractile function of the smooth muscle,

or indeed, any other functions in the vessel wall

So much remains to be done

Note on terminology for receptors

activated by the catecholamines,

adrenaline and noradrenaline

The official approach to receptor terminology, laid out by

IUPHAR (International Union of Basic and Clinical

Pharma-cology), states that ‘The receptor should be named after the

endogenous agonist, or the appropriate collective term when

a family of related substances may interact with the receptor.’

(Vanhoutte et al., 1996) On this basis, the term should be

either ‘noradrenaline receptor’, reflecting the predominant

endogenous agonist or ‘adrenaline receptor’, reflecting

history and the gene symbol (ADRA1A, ADRB2 and so on)

Alternatively, and in line with the official definition,

‘cat-echolamine receptor’ recognizes both endogenous agonists.Indeed, ‘adrenaline receptor’ was employed by many distin-guished pharmacologists up until the early 1960s (e.g Dale,1943; Ing, 1943; Tickner, 1951; Fleckenstein, 1952; McDougaland West, 1953; Stafford, 1963)

So, how did it all get so confused that three alternativenames (adrenoceptor, adrenoreceptor, adrenergic receptor)are now used, none of them corresponding to these logicalnames?

The concept of receptors per se is usually attributed to J.N.

Langley, who suggested that nerves operated to contractmuscle through a receptive substance, ‘some substance which

is not the actual contractile molecule though capable ofacting upon it’ (Langley, 1905; 1907)

The next development of a naming system for receptorscame when Ahlquist (1948) established the idea that therewas more than one type of receptor for adrenaline/noradrenaline He hypothesized that there were two recep-tors, which he termed α- and β- adrenotropic receptors

‘Adrenotropic’, with a suffix more correctly used for stances releasing other substances’ was never employedagain, even by Ahlquist, who employed all the subsequentvariations in his later papers However, the idea of multiplereceptor types was taken up critically by Furchgott (1959) Hesuggested that two receptor types were inadequate and that atleast four receptors were needed to explain existing data onthe responses to exogenous catecholamines

‘sub-Going with Ahlquist’s methodology based on relativepotency of different agonists, and Greek prefixes, Furchgottsuggestedα, β, γ, δ; but this did not survive either What didcatch on were his suggestions for the nomenclature for thereceptor family, which he based broadly on previous terms forphysiological phenomena Dale had introduced the idea thatautonomic sympathetic nerves should be called ‘adrenergic’,meaning ‘works by release of adrenaline’ (Dale, 1935) andLangley had coined ‘receptive substances’ (see above) Furch-gott now suggested two terms for the receptors, viz ‘adren-ergic receptors’ (from Dale’s ‘adrenergic nerves’) and

‘adrenoceptive sites’ (from Langley’s ‘receptive substances’).The latter mutated into ‘adrenoreceptors’ or ‘adrenoceptors’.The usage has remained confused ever since

This journal, the BJP, has used ‘adrenoceptor’ fairly

con-sistently since 1968 and the IUPHAR nomenclature tee uses this term in defining this receptor family from the

commit-1990s up to the time of writing (Bylund et al., 1994; Hieble

et al., 1995; Alexander et al., 2013a,b) This is the term I have

used in this review Nevertheless, many authors use the terms

‘adrenergic receptors’ or ‘adrenoreceptors’ Fortunately, mostsearch engines now accept ‘adrenergic receptors’, ‘adrenocep-tors’ and ‘adrenoreceptors’ as pseudonyms, so that thereseems to be little point in trying to endorse one term againstanother at this stage However, it is confusing for studentsand people entering the field as well as inconsistent with theIUPHAR standard

Conflict of interest

No conflicts of interest

British Journal of Pharmacology (2015) 172 1179–1194 1191

Ahlquist RP (1948) A study of the adrenotropic receptors Am J

Physiol 153: 586–600

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL,

McGrath JC et al (2013a) The Concise Guide to PHARMACOLOGY

2013/14: overview Br J Pharmacol 170: 1449–1458

Alexander SP, Benson HE, Faccenda E, Pawson AJ, Sharman JL,

Spedding M et al (2013b) The Concise Guide to PHARMACOLOGY

2013/14: G Protein-Coupled Receptors Br J Pharmacol 170:

1459–1581

de Andrade CR, Fukada SY, Olivon VC, de Godoy MA, Haddad R,

Eberlin MN et al (2006).α1D-adrenoceptor-induced relaxation on

rat carotid artery is impaired during the endothelial dysfunction

evoked in the early stages of hyperhomocysteinemia Eur J

Pharmacol 543: 83–91

Angus JA, Cocks TM, Satoh K (1986) The alpha adrenoceptors on

endothelial cells Fed Proc 45: 2355–2359 PubMed [citation] PMID:

3015689

Baker JG, Hall IP, Hill SJ (2003) Pharmacology and direct

visualisation of BODIPY-TMR-CGP: a long-acting fluorescent

β2-adrenoceptor agonist Br J Pharmacol 139: 232–242

Brown CM, MacKinnon AC, McGrath JC, Spedding M, Kilpatrick

AT (1990).α2-adrenoceptor subtypes and imidazoline-like binding

sites in the rat brain Br J Pharmacol 99: 803–809

Bulloch JM, Daly CJ (2014) Autonomic nerves and perivascular fat:

interactive mechanisms Pharmacol Ther 143: 61–73

Bylund DB, Eikenberg DC, Hieble JP, Langer SZ, Lefkowitz RJ,

Minneman KP et al (1994) International Union of Pharmacology

nomenclature of adrenoceptors Pharmacol Rev 46: 121–136

Carlsson A, Falck B, Hillarp NA, Thieme G, Torp A (1961) A new

histochemical method for visualization of tissue catechol amines

Med Exp Int J Exp Med 4: 123–125

Chalothorn D, McCune DF, Edelmann SE, García-Cazarín ML,

Tsujimoto G, Piascik MT (2002) Differences in the cellular

localization and agonist-mediated internalization properties of the

α1-adrenoceptor subtypes Mol Pharmacol 61: 1008–1016

Dale H (1935) Pharmacology and nerve-endings (Walter Ernest

Dixon Memorial Lecture): (Section of Therapeutics and

Pharmacology) Proc R Soc Med 28: 319–332

Dale H (1943) Modes of drug action General introductory address

Trans Faraday Soc 39: 319b–322

Daly CJ, McGrath I (2003) Fluorescent ligands, antibodies, and

proteins for the study of receptors Phamacol Ther 100: 101–118

Daly CJ, McGrath JC (2011) Previously unsuspected widespread

cellular and tissue distribution of beta-adrenoceptors and its

relevance to drug action Trends Pharmacol Sci 32: 219–226

Daly CJ, Gordon JF, McGrath JC (1992) The use of fluorescent

nuclear dyes for the study of blood vessel structure and function:

novel applications of existing techniques J Vasc Res 29: 41–48

Daly CJ, Milligan CM, Milligan G, Mackenzie JF, McGrath JC

(1998) Cellular localization and pharmacological characterization

of functioningα1-adrenoceptors by fluorescent ligand binding and

image analysis reveals identical binding properties of clustered and

diffuse populations of receptors J Pharmacol Exp Ther 286:

984–990

Daly CJ, Deighan C, McGee A, Mennie D, Ali Z, McBride M et al.

(2002) A knockout approach indicates a minor vasoconstrictor role

for vascularα1B-adrenoceptors in mouse Physiol Genomics 9:

85–91

Daly CJ, Ross RA, Whyte J, Henstridge CM, Irving AJ, McGrath JC(2010) Fluorescent ligand binding reveals heterogeneousdistribution of adrenoceptors and ‘cannabinoid-like’ receptors insmall arteries Br J Pharmacol 159: 787–796

Daly CJ, Parmryd I, McGrath JC (2012) Visualization and analysis

of vascular receptors using confocal laser scanning microscopy andfluorescent ligands Methods Mol Biol 897: 95–107

Deighan C, Woollhead AM, Colston JF, McGrath JC (2004).Hepatocytes fromα1B-adrenoceptor knockout mice revealcompensatory adrenoceptor subtype substitution Br J Pharmacol142: 1031–1037

Deighan C, Methven L, Naghadeh MM, Wokoma A, Macmillan J,

Daly CJ et al (2005) Insights into the functional roles of

α1-adrenoceptor subtypes in mouse carotid arteries using knockoutmice Br J Pharmacol 144: 558–565

Dunn WR, McGrath JC, Wilson VG (1989) Expression offunctional postjunctionalα2-adrenoceptors in rabbit isolated distalsaphenous artery – a permissive role for angiotensin II? Br JPharmacol 96: 259–261

Filippi S, Parenti A, Donnini S, Granger HJ, Fazzini A, Ledda F(2001).α1D-adrenoceptors cause endothelium-dependentvasodilatation in the rat mesenteric vascular bed J Pharmacol ExpTher 296: 869–875

Flacco N, Parés J, Serna E, Segura V, Vicente D, Pérez-Aso M et al.

(2013).α1D-Adrenoceptors are responsible for the high sensitivityand the slow time-course of noradrenaline-mediated contraction inconductance arteries Pharmacol Res Perspect 1: e00001

Fleckenstein A (1952) A quantitative study of antagonists ofadrenaline on the vessels of the rabbit’s ear Br J PharmacolChemother 7: 553–562

Furchgott RF (1959) The receptors for epinephrine andnorepinephrine (adrenergic receptors) Pharmacol Rev 11: 429–441,discussion 441-422

Gillespie JS, McGrath JC (1974) The effect of pithing and of nervestimulation on the depletion of noradrenaline by reserpine in therat anococcygeus muscle and vas deferens Br J Pharmacol 52:585–590

Hague C, Chen Z, Uberti M, Minneman KP (2003).α1-adrenergicreceptor subtypes: non-identical triplets with different dancingpartners? Life Sci 74: 411–418

Hansen MA, Dutton JL, Balcar VJ, Barden JA, Bennett MR (1999).P2X (purinergic) receptor distributions in rat blood vessels J AutonNerv Syst 75: 147–155

Hieble JP, Bylund DB, Clarke DE, Eikenburg DC, Langer SZ,

Lefkowitz RJ et al (1995) International Union of Pharmacology X.

Recommendation for nomenclature ofα1-adrenoceptors: consensusupdate Pharmacol Rev 47: 267–270

Hirasawa A, Sugawara T, Awaji T, Tsumaya K, Ito H, Tsujimoto G(1997) Subtype-specific differences in subcellular localization of

α1-adrenoceptors: chlorethylclonidine preferentially alkylates theaccessible cell surfaceα1-adrenoceptors irrespective of the subtype.Mol Pharmacol 52: 764–770

Hrometz SL, Edelmann SE, McCune DF, Olges JR, Hadley RW, Perez

DM et al (1999) Expression of multipleα1-adrenoceptors onvascular smooth muscle: correlation with the regulation ofcontraction J Pharmacol Exp Ther 290: 452–463

Ing HR (1943) Chemical constitution and pharmacological action.Trans Faraday Soc 39: 372–380

Langley JN (1905) On the reaction of cells and of nerve-endings tocertain poisons, chiefly as regards the reaction of striated muscle tonicotine and to curari J Physiol 33: 374–413

1192 British Journal of Pharmacology (2015) 172 1179–1194

Langley JN (1907) On the contraction of muscle, chiefly in relation

to the presence of ‘receptive’ substances: part I J Physiol 36:

347–384

Luo D, Barker JR, McGrath JC, Daly CJ (1998) Iterative multilevel

thresholding and splitting for 3D segmentation of live cell nuclei

using laser scanning confocal microscopy J Comput Assist Microsc

10: 151–162

Mackenzie JF, Daly CJ, Luo D, McGrath JC (1998) Cellular

localisation ofα1-adrenoceptors in native smooth muscle cells

Online Proceedings of the 5th Internet World Congress on

Biomedical Sciences ‘98 at McMaster University, Canada download

Available at: http://www.mcmaster.ca/inabis98/cvdisease/

mackenzie0899/index.html (accessed 21/12/2014)

Mackenzie JF, Daly CJ, Luo D, McGrath JC (1999) Cellular

localisation ofα1-adrenoceptors in human primary smooth muscle

cells derived from the prostate of patients with benign prostatic

hypertrophy Eur Urol 36 (Suppl 1): 115

Mackenzie JF, Daly CJ, Pediani JD, McGrath JC (2000) Quantitative

imaging in live human cells reveals intracellularα1-adrenoceptor

ligand-binding sites J Pharmacol Exp Ther 294: 434–443

MacMillan LB, Lakhlani PP, Hein L, Piascik M, Guo TZ, Lovinger D

et al (1998) In vivo mutation of theα2A-adrenergic receptor by

homologous recombination reveals the role of this receptor subtype

in multiple physiological processes Adv Pharmacol 42: 493–496

No abstract available, PubMed [citation] PMID: 9327947

McCune DF, Edelmann SE, Olges JR, Post GR, Waldrop BA, Waugh

DJJ et al (2000) Regulation of the cellular localization and

signalling properties of theα1B- andα1D-Adrenoceptors by agonists

and inverse agonists Mol Pharmacol 57: 659–666

McDougal MD, West GB (1953) The action of drugs on isolated

mammalian bronchial muscle Br J Pharmacol Chemother 8: 26–29

McGrath JC, Arribas SM, Daly CJ, Gordon JF (1995) Confocal

microscopy for structure and real-time pharmacology in blood

vessels J Hum Hypertens 9: 645–647

McGrath J, Wilson V (1988) Alpha-adrenoceptor subclassification

by classical and response-related methods: same question, different

answers Trends Pharmacol Sci 9: 162–165

McGrath JC (1982) Evidence for more than one type of

post-junctional alpha-adrenoceptor Biochem Pharmacol 31:

467–484

McGrath JC, Arribas S, Daly CJ (1996a) Fluorescent ligands for the

study of receptors Trends Pharmacol Sci 17: 393–399

McGrath JC, Arribas S, Daly CJ (1996b) Fluorescent ligands for the

study of receptors Trends Pharmacol Sci 17: 393–399 PubMed

[citation], Erratum in: Trends Pharmacol Sci 1997 May;18:181,

Erratum was still incorrect, see figure 5 PMID: 8990953

McGrath JC, Mackenzie JF, Daly CJ (1999a) Pharmacological

implications of cellular localization ofα1-adrenoceptors in native

smooth muscle cells J Auton Pharmacol 19: 303–310

McGrath JC, Naghadeh MA, Pediani JD, Mackenzie JF, Daly CJ

(1999b) Importance of agonists in alpha-adrenoceptor classification

and localisation ofα1-adrenoceptors in human prostate Eur Urol

36 (Suppl 1): 80–88

McGrath JC, Deighan C, Briones AM, Shafaroudi MM, McBride M,

Adler J et al (2005) New aspects of vascular remodelling: the

involvement of all vascular cell types Exp Physiol 90: 469–475

Methven L, McBride M, Wallace GA, McGrath JC (2009a) The

α1B/D-adrenoceptor knockout mouse permits isolation of the

vascularα1A-adrenoceptor and elucidates its relationship to theother subtypes Br J Pharmacol 158: 209–224

Methven L, Simpson PC, McGrath JC (2009b) Alpha1A/B-knockoutmice explain the nativeα1D-adrenoceptor’s role in vasoconstrictionand show that its location is independent of the other

alpha1-subtypes Br J Pharmacol 158: 1663–1675

Miller VM, Vanhoutte PM (1985) Endothelialα2-adrenoceptors incanine pulmonary and systemic blood vessels Eur J Pharmacol 118:123–129

Miquel MR, Segura V, Ali Z, D’Ocon MP, McGrath JC, Daly CJ(2005) 3-d image analysis of fluorescent drug binding MolImaging 4: 40–52

Morris DP, Price RR, Smith MP, Lei B, Schwinn DA (2004) Cellulartrafficking of humanα1A-adrenergic receptors is continuous andprimarily agonist-independent Mol Pharmacol 66: 843–854.Papay R, Gaivin R, McCune DF, Rorabaugh BR, Macklin WB,

McGrath JC et al (2004) Mouseα1B-adrenergic receptor is expressed

in neurons and NG2 oligodendrocytes J Comp Neurol 478: 1–10.Papay R, Gaivin R, Jha A, McCune DF, McGrath JC, Rodrigo MC

et al (2006) Localization of the mouseα1A-adrenergic receptor (AR)

in the brain: alpha1AAR is expressed in neurons, GABAergicinterneurons, and NG2 oligodendrocyte progenitors J CompNeurol 497: 209–222

Pawson AJ, Sharman JL, Benson HE, Faccenda E, Alexander SP,

Buneman OP et al.; NC-IUPHAR (2014) The IUPHAR/BPS Guide to

PHARMACOLOGY: an expert-driven knowledge base of drug targetsand their ligands Nucl Acids Res 42 (Database Issue): D1098–1106.Pediani JD, MacKenzie JF, Heeley RP, Daly CJ, McGrath JC (2000).Single-cell recombinant pharmacology: bovineα1A-adrenoceptors inrat-1 fibroblasts release intracellular Ca2+, display

subtype-characteristic agonism and antagonism, and exhibit anantagonist-reversible inverse concentration-response phase JPharmacol Exp Ther 293: 887–895

Pediani JD, Colston JF, Caldwell D, Milligan G, Daly CJ, McGrath

JC (2005) Beta-arrestin-dependent spontaneousα1A-adrenoceptorendocytosis causes intracellular transportation of alpha-blockers viarecycling compartments Mol Pharmacol 67: 992–1004

Perez-Aso M, Segura V, Montó F, Barettino D, Noguera MA,

Milligan G et al (2013) The threeα1-adrenoceptor subtypes showdifferent spatio-temporal mechanisms of internalization andERK1/2 phosphorylation Biochim Biophys Acta 1833: 2322–2333.Piascik MT, Hrometz SL, Edelmann SE, Guarino RD, Hadley RW,Brown RD (1997) Immunocytochemical localization of the

α1B-adrenergic receptor and the contribution of this and the othersubtypes to vascular smooth muscle contraction: analysis withselective ligands and antisense oligonucleotides J Pharmacol ExpTher 283: 854–868

Saper CB (2007) In living color J Comp Neurol 502: 173–174.Segura V, Pérez-Aso M, Montó F, Carceller E, Noguera MA, Pediani J

et al (2013) Differences in the signaling pathways ofα1A- and

α1B-adrenoceptors are related to different endosomal targeting PLoSONE 8: e64996 doi: 10.1371/journal.pone.0064996; Print 2013.Shafaroudi MM, McBride M, Deighan C, Wokoma A, Macmillan J,

Daly CJ et al (2005) Two ‘knockout’ mouse models demonstrate

that aortic vasodilatation is mediated viaα2A-adrenoceptors located

on the endothelium J Pharmacol Exp Ther 314: 804–810

Shang C, Daly CJ, McGrath JC, Barker JR (2000a) Neural networkbased classification of cell images via estimation of fractaldimensions In: Malmgren H, Borga M, Niklasson L (eds)

British Journal of Pharmacology (2015) 172 1179–1194 1193

Perspectives in Neural Computing Springer Press: London, pp.

111–116 Proceedings of the International Conference on Artificial

Neural Networks in Medicine and Biology, Goteborg, Sweden

Shang C, Daly CJ, McGrath JC, Barker JR (2000b) Analysis and

classification of tissue section images using directional fractal

dimension features Proceedings of the IEEE International

Conference on Image Processing Vancouver, Canada 1, 164–167

Shang C, McGrath JC, Daly CJ, Barker JR (2000c) Modelling and

classification of vascular smooth muscle cell images Electronics

Letters – IEEE 36: 1532–1533

Somoza B, Gonzalez MC, Gonzalez JM, Abderrahim F, Arribas SM,

Fernandez-Alfonso MS (2005) Modulatory role of the adventitia on

noradrenaline and angiotensin II responses role of endothelium

and AT2 receptors Cardiovasc Res 65: 478–486

Stafford A (1963) Potentiation of some catechol amines by

phenoxybenzamine, guanethidine and cocaine Br J Pharmacol

Chemother 21: 361–367

Stanasila L, Abuin L, Dey J, Cotecchia S (2008) Different

Internalization Properties of theα1a- and α1b-Adrenergic Receptor

Subtypes: The Potential Role of Receptor Interaction with

β-Arrestins and AP50 Mol Pharmacol 74: 562–573

Summers RJ, Molenaar P (1995) Autoradiography of beta 1- and

beta 2-adrenoceptors Methods Mol Biol 41: 25–39

Tickner A (1951) Inhibition of amine oxidase by antihistaminecompounds and related drugs Br J Pharmacol Chemother 6:606–610

Vanhoutte PM, Humphrey PP, Spedding M (1996) X InternationalUnion of Pharmacology recommendations for nomenclature of newreceptor subtypes Pharmacol Rev 48: 1–2

Wilson VG, Brown CM, McGrath JC (1991) Are there more thantwo types of alpha-adrenoceptors involved in physiologicalresponses? Exp Physiol 76: 317–346

Yang M, Verfurth F, Buscher R, Michel MC (1997) Is

α1D-adrenoceptor protein detectable in rat tissues? NaunynSchmiedebergs Arch Pharmacol 119: 269–277

Supporting information

Additional Supporting Information may be found in theonline version of this article at the publisher’s web-site:http://dx.doi.org/10.1111/bph.13008

Prof Ian McGrath – Adrenoceptors – Black Boxes to blackboxes over my lifetime.MP4

1194 British Journal of Pharmacology (2015) 172 1179–1194

Yi-Fong Chen1,2, Yi-Chien Lin2, Susan L Morris-Natschke3,

Chen-Fang Wei2, Ting-Chen Shen2, Hui-Yi Lin2, Mei-Hua Hsu2,

Li-Chen Chou2, Yu Zhao3, Sheng-Chu Kuo1,2, Kuo-Hsiung Lee3,4* and

Li-Jiau Huang1,2*

1The Ph.D Program for Cancer Biology and Drug Discovery, China Medical University and

Academia Sinica, Taichung, Taiwan,2School of Pharmacy, China Medical University, Taichung,

Taiwan,3Natural Products Research Laboratories, UNC Eshelman School of Pharmacy,

University of North Carolina, Chapel Hill, NC, USA, and4Chinese Medicine Research and

Development Center, China Medical University and Hospital, Taichung, Taiwan

Correspondence

Li-Jiau Huang, School ofPharmacy, China MedicalUniversity, No.91 Hsueh-ShihRoad, Taichung, 40402, Taiwan.E-mail: ljhuang@mail.cmu.edu.tw

or Kuo-Hsiung Lee, NaturalProducts Research Laboratories,UNC Eshelman School ofPharmacy, University of NorthCarolina, Chapel Hill, NC27599-7568, USA E-mail:

BACKGROUND AND PURPOSE

4-Phenylquinolin-2(1H)-one (4-PQ) derivatives can induce cancer cell apoptosis Additional new 4-PQ analogs were

investigated as more effective, less toxic antitumour agents

EXPERIMENTAL APPROACH

Forty-five 6,7,8-substituted 4-substituted benzyloxyquinolin-2(1H)-one derivatives were synthesized Antiproliferative activities

were evaluated using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazoliun bromide assay and structure–activity relationship

correlations were established Compounds 9b, 9c, 9e and 11e were also evaluated against the National Cancer Institute-60

human cancer cell line panel Hoechst 33258 and Annexin V-FITC/PI staining assays were used to detect apoptosis, whileinhibition of microtubule polymerization was assayed by fluorescence microscopy Effects on the cell cycle were assessed byflow cytometry and on apoptosis-related proteins (active caspase-3, -8 and -9, procaspase-3, -8, -9, PARP, Bid, Bcl-xL andBcl-2) by Western blotting

KEY RESULTS

Nine 6,7,8-substituted 4-substituted benzyloxyquinolin-2(1H)-one derivatives (7e, 8e, 9b, 9c, 9e, 10c, 10e, 11c and 11e)

Mechanistic studies indicated that compound 11e disrupted microtubule assembly and induced G2/M arrest, polyploidy and

apoptosis via the intrinsic and extrinsic signalling pathways Activation of JNK could play a role in TRAIL-induced COLO 205apoptosis

British Journal of Pharmacology (2015) 172 1195–1221 1195

© 2014 The British Pharmacological Society

CONCLUSION AND IMPLICATIONS

New quinolone derivatives were identified as potential pro-apoptotic agents Compound 11e could be a promising lead

compound for future antitumour agent development

Abbreviations

4-PQ, 4-phenylquinolin-2(1H)-one; AGT, alkylguanine-DNA alkyltransferase; BG, benzylguanine; CDK, cyclin dependent

kinase; NCI, National Cancer Institute; PPA, polyphosphoric acid; SAR, structure–activity relationship

Introduction

Cancer is presently a worldwide health problem and the

leading cause of death in the United States and other

devel-oped countries (Rastogi et al., 2004) Cancer is a formidable

disease caused by disordered cell growth and invasion of

tissues and organs While various therapies and strategies have

been developed to treat cancer, most of them have limitations

Thus, new anticancer drugs are continually needed The main

challenge facing clinical cancer therapy is to find a specific

approach that kills malignant cells with no or few adverse

effects on normal tissues and considerable attempts have been

made to develop innovative, safe and effective methods to

defeat cancer While scientists have discovered many agents

with cytostatic action against cancer cells (Liu et al., 2007;

Folger et al., 2011), increasing understanding of the biological

processes involved in cancer cell survival has led to the design

and discovery of better targeted, novel therapeutic anticancer

drugs For several chemotherapeutic agents, a direct

correla-tion has been found between antitumour efficacy and ability

to induce apoptosis (Kaufmann and Earnshaw, 2000) Thus,

approaches aimed at promoting apoptosis in cancer cells have

gained importance in cancer research (Fesik, 2005; Fischer and

Schulze-Osthoff, 2005)

Heterobicycles are indispensable structural units in

com-pounds with a broad range of biological activities Among

various nitrogen-containing fused heterocyclic skeletons,

qui-noline and quinolone structures are important components

prevalent in a vast array of biological systems Compounds

with a quinoline nucleus exhibit various pharmacological

properties, including antioxidant (Chung and Woo, 2001;

Zhang et al., 2013), anti-inflammatory (Baba et al., 1996; Mukherjee and Pal, 2013), antibacterial (Cheng et al., 2013), anti-human immunodeficiency virus (Freeman et al., 2004; Hopkins et al., 2004), antimalarial (Cornut et al., 2013; Pandey

et al., 2013), antituberculosis (Lilienkampf et al., 2009),

anti-Alzheimer’s disease (Fiorito et al., 2013), anticancer (Wang

et al., 2011; Abonia et al., 2012; Chan et al., 2012) activities.

Accordingly, Solomon and Lee described containing subunits as ‘privileged structures’ for drug devel-opment (Solomon and Lee, 2011) 2-Quinolone [quinolin-

quinoline-2(1H)-one], also called 1-aza coumarin or carbostyril, and

4-quinolone are structural isomers The 2-quinolone skeleton

is a fertile source of biologically active compounds, including

a wide spectrum of alkaloids investigated for antitumour

activ-ity (Ito et al., 2004; He et al., 2005; Nakashima et al., 2012) In

our previous investigation, 6,7-methylenedioxy-4-substituted

phenylquinolin-2(1H)-one derivatives

(4-phenylquinolin-2(1H)-ones; 4-PQs) were identified as novel inducing agents (Figure 1) (Chen et al., 2013b) Recently, Arya and Agarwal reported that 4-hydroxyquinolin-2(1H)-one

apoptosis-derivatives, prepared efficiently through microwave tion, showed strong photo-antiproliferative activity (Arya andAgarwal, 2007) Thus, we have directed our focus onto 4-PQanalogues as inducers of apoptosis In our current study, wetargeted the 2-quinolone structure as a basic scaffold of newderivatives with different substituents

irradia-Purine-based compounds such as olomoucine and vitine (Figure 1), which contain other heterobicyclic ringsystems, are known ATP-binding site competitive inhibitors

rosco-Tables of Links

TARGETS

Aurora A kinase DR4, death receptor 4

Aurora B kinase DR5, death receptor 5

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://

www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are

permanently archived in the Concise Guide to PHARMACOLOGY 2013/14 (a,b Alexander et al., 2013a,b).

1196 British Journal of Pharmacology (2015) 172 1195–1221

of cyclin dependent kinase (CDK) and are useful cell

prolif-eration inhibitors in the treatment of cancer (Jorda et al.,

2011) Structure–activity relationship (SAR) studies on CDK

inhibitors demonstrated that a small hydrophobic group

such as a non-polar benzyl group at the O6- or N6-position

of the heterobicycle maximized CDK inhibition (Gibson

et al., 2002; Zatloukal et al., 2013) In addition, numerous

CDK inhibitor-related compounds that contain benzyl or

aryl methyl groups on different core scaffolds, such as

pyrazolo[1,5-a]pyrimidines (Paruch et al., 2007),

quinazolin-4-amines (Mott et al., 2009), pyrimidine (Coombs et al., 2013)

and aminopurine (Doležal et al., 2006) (Figure 1), have

been studied Furthermore, a series of

6-(benzyloxy)-2-(aryldiazenyl)-9H-purine derivatives were reported to act as

prodrugs of O6-benzylguanine (O6-BG; Figure 1), which

selec-tively targets O6-alkylguanine-DNA alkyltransferase (AGT) in

hypoxic tumour cells (Zhu et al., 2013) The AGT protein

plays a critical role in DNA repair, which can be exploited in

chemotherapeutic treatment of neoplastic cells (Dolan and

Pegg, 1997; Daniels et al., 2000) Alkylation of AGT with the

benzyl group of O6-BG results in complete depletion of the

alkyltransferase protein Consequently, numerous O6-BG

analogs have been developed as AGT inhibitors (Chae et al.,

1995; Terashima and Kohda, 1998) Ruiz et al (2008) reported

that a family of quinolinone compounds acted as novel

non-nucleosidic AGT inhibitors These quinolinones could reach

the critical catalytic residue Cys145 buried deep within the

binding groove, occupy the catalytic cleft of human DNA

repair AGT protein and act as substrate mimics of the

O6-guanine moiety

Furthermore, the activity of biologically proven cer pharmacophores can be enhanced by introducing appro-priate substitutions on the chemical scaffolds In medicinalchemistry, shortening or lengthening chain length is a usefultactic to improve the affinity of target binding Some reportshave demonstrated that the pro-apoptotic (anti-tumour)activity of certain compounds was dramatically improved byslightly changing the length and spacing of lateral branches,such as benzyl and other alkyl-aromatic side chains, on core

antican-skeletons (Al-Obaid et al., 2009; Font et al., 2011) Such

explo-ration and utilization of chemical diversity relative to macological space is an ongoing drug discovery strategy,referred to as privileged-substructure-based diversity-orientedsynthesis (Oh and Park, 2011) Based on this strategy, as well

phar-as the structures shown in Figure 1, we proposed addition of

a substituted benzyl (C ring) side chain linked at the

O4-position of 4-hydroxyquinolin-2(1H)-one (2-quinolone

scaffold) as a possible strategy for discovering new leads withpro-apoptotic bioactivity The flexibility of the benzyl moietymight provide better antitumour activity compared with ourearlier 4-PQ derivatives (Figure 1) Therefore, we designed a

series of 4-benzyloxyquinolin-2(1H)-one analogues 7a–e ∼ 15a–e, with the general structures of target compounds

depicted in Figure 1 To the best of our knowledge, this

is the first evaluation of 2-quinolone analogues bearing

an O4-benzyl moiety against cancer The goal of thecurrent study was to discover more effective and less toxicantitumour agents and contribute to the SAR profile of2-quinolones with anti-proliferative activity and pro-apoptotic activities in cancer cells

Figure 1

The structures of some anticancer agents and the general structure of the target compounds (7a–e ∼ 15a–e).

British Journal of Pharmacology (2015) 172 1195–1221 1197

Materials and physical measurements

All solvents and reagents were obtained commercially and

used without further purification The progress of all

reac-tions was monitored by TLC on 2× 6 cm pre-coated silica gel

60 F254plates of thickness 0.25 mm (Merck KGaA, Darmstadt,

Germany) The chromatograms were visualized under UV at

254–366 nm Column chromatography was performed using

silica gel 60 (Merck KGaA, particle size 0.063–0.200 mm)

Melting points (mp) were determined with a Yanaco

MP-500D melting point apparatus (Yanaco New Science Inc.,

Kyoto, Japan) and are uncorrected IR spectra were recorded

on Shimadzu IR-Prestige-21 spectrophotometers (Shimadzu

Corp., Kyoto, Japan) as KBr pellets The one-dimensional

NMR (1H and13C) spectra were obtained on a Bruker Avance

DPX-200 FT-NMR spectrometer (Bruker Corp., Billerica, MA,

USA) at room temperature The two-dimensional NMR

spectra were obtained on a Bruker Avance DPX-400 FT-NMR

spectrometer (Bruker Corp.) and chemical shifts were

expressed in parts per million (ppm,δ) The following

abbre-viations are used: s, singlet; d, doublet; t, triplet; dd, double

doublet and m, multiplet Mass spectra were performed at the

Instrument Center of National Science Council at National

Chung Hsing University (Taichung City, Taiwan) using a

Finnigan ThermoQuest MAT 95 XL (EI-MS) (Thermo Fisher

Scientific Inc., Waltham, MA, USA)

General procedure for the synthesis of

4-hydroxyquinolin-2(1H)-one

derivatives (5a–i)

4-Hydroxyquinolin-2(1H)-one derivatives 5a–i were prepared

by ‘one-pot’ cyclization in polyphosphoric acid (PPA) A

mixture of the appropriate substituted aniline 1a–i (1 equiv)

and diethylmalonate (2) (1.2 equiv) was heated with PPA (five

to six times by weight) at 130°C for 2–6 h (TLC monitoring)

Then, the mixture was cooled and diluted with water A gum

solidified upon standing overnight and the precipitate was

filtered, washed with water and air-dried to provide 5a–i with

sufficient purity for the next reaction Physical and

spectro-scopic data for 5a are given in the succeeding text; the data

for the remaining compounds are provided as Supporting

Information

4-Hydroxyquinolin-2(1H)-one (5a)

(Mohamed, 1991; Nadaraj et al., 2006; Arya

and Agarwal, 2007; Park et al., 2007; Zhang

et al., 2008)

Compound 5a (3.48 g, 21.59 mmol) was obtained from

aniline (1a) (3.82 g, 41.01 mmol) and diethylmalonate

(2) (7.88 g, 49.20 mmol); yield: 53%; light-yellow solid;

164.18; MS (EI, 70 eV) m/z: 161.1[M]+; HRMS (EI) m/z:

calcu-lated for C9H7NO2: 161.0477; found: 161.0472

General procedure for the synthesis of 6,7,8-substituted 4-substituted

benyloxyquinolin-2(1H)-one derivatives

(7a–e, 8a–e, 9a–e, 10a–e, 11a–e, 12a–e, 13a–e, 14a–e, 15a–e)

A mixture of 4-hydroxyquinolin-2(1H)-one derivatives 5a–i

(1 equiv) and K2CO3(2 equiv) in DMF (10–20 mL) was heated

at 90°C for 1–2 h The appropriate benzyl chloride or bromide

(6a–e, 1–1.4 equiv) was added and the mixture was heated at

80–90°C for 1–6 h Reaction completion was confirmed byTLC monitoring The mixture was poured into ice water(200 mL) and the precipitated solid was collected by filtrationand then washed with water The residue was treated withethyl acetate (EtOAc) and purified by recrystallization If nosolid was formed after the addition of ice water, then thereaction mixture was extracted with EtOAc (3× 100 mL) Thecombined organic layers were dried over anhydrous MgSO4

before evaporation of solvent in vacuo The residue was

iso-lated by column chromatography (silica gel, EtOAc as eluate)and then recrystallized to give the corresponding pure prod-

ucts, 4-benzyloxyquinolin-2(1H)-one derivatives 7a–e, 8a–e,

9a–e, 10a–e, 11a–e, 12a–e, 13a–e, 14a–e and 15a–e

Physi-cal and spectroscopic data for 11e are given as examples; the

data for the remaining compounds are provided as ing Information

Support-

4-(3′,5′-Dimethoxybenzyloxy)-6-methoxyquinolin-2(1H)-one (11e)

Compound 11e (0.70 g, 2.05 mmol) was obtained from 5e

(1.12 g, 5.86 mmol) and 3,5-dimethoxybenzyl bromide(1.48 g, 6.40 mmol); yield: 35%; white crystal; mp: 217–219°C; IR (KBr)ν (cm−1): 1674 (C= O);1H NMR (200 MHz,

DMSO-d 6) δ (ppm): 3.74 (s, 6H, 3′, 5′–OCH3), 3.76 (s, 3H,6–OCH3), 5.20 (s, 2H, –O–CH2–), 5.94 (s, 1H, H–3), 6.47 (dd,

J = 2.2,2.2 Hz, 1H, H–4′), 6.66 (d, J = 2.2 Hz, 2H, H–2′, H–6′),

7.14–7.26 (m, 3H, H–5,7,8), 11.32 (br s, 1H, NH);13C NMR

(50 MHz, DMSO-d 6)δ (ppm): 55.63 (2C), 55.78, 70.05, 98.71,100.04, 104.17, 105.62 (2C), 115.50, 117.18, 120.50, 133.55,138.76, 154.43, 161.07 (2C), 161.89, 163.23; MS (EI, 70 eV)

m/z: 341.0 [M]+; HRMS (EI) m/z: calculated for C19H19NO5:341.1263; found: 341.1257

diphenyltetrazoliun bromide (MTT) assay for antiproliferative activity

3-(4,5-dimethylthiazol-2-yl)-2,5-Human tumour cell lines (HTCLs) of the cancer screeningpanel were maintained in RPMI-1640 medium supplementedwith 10% FBS (GIBCO®, Life Technologies, Grand Island, NY,USA), penicillin (100 U·mL−1)/streptomycin (100μg·mL−1)(GIBCO, Life Technologies) and 1%L-glutamine (GIBCO, LifeTechnologies) at 37°C in a humidified atmosphere containing5% CO2 Human hepatoma Hep 3B and normal skin Detroit

551 cells were maintained in DMEM medium supplementedwith 10% FBS (GIBCO, Life Technologies), penicillin(100 U·mL−1)/streptomycin (100μg·mL−1) (GIBCO, Life Tech-nologies) and 1%L-glutamine (GIBCO, Life Technologies) at37°C in a humidified atmosphere containing 5% CO2 Loga-rithmically growing cancer cells were used for all experiments.The HTCLs were treated with vehicle or test compounds for

48 h Cell growth rate was determined by MTT reduction

1198 British Journal of Pharmacology (2015) 172 1195–1221

assay (Mosmann, 1983) After 48 h treatment, cell growth rate

was measured on anELISAreader at a wavelength of 570 nm

and the IC50values of test compounds were calculated

In vitro National Cancer Institute (NCI)-60

HTCL panel

In vitro cytotoxic activities were evaluated through the

Devel-opmental Therapeutic Program (DTP) of the NCI (Shoemaker,

2006) For more information on the anticancer screening

protocol, please see: http://dtp.nci.nih.gov/branches/btb/

ivclsp.html

Cell morphology and Hoechst 33258 staining

COLO 205 cells were plated at a density of 2.5× 105cells per

well in 12-well plates and then incubated with 50 nM of

compound 11e for 12 to 48 h Cells were directly examined

and photographed under a contrast-phase microscope Nuclei

were stained with Hoechst 33258 (bis-benzimide;

Sigma-Aldrich, St Louis, MO, USA) to detect chromatin

condensa-tion or nuclear fragmentacondensa-tion, features of apoptosis After 0,

12, 24, 36 and 48 h, 11e-treated cells were stained with

5μg·mL−1 Hoechst 33258 for 10 min After washing twice

with PBS, cells were fixed with 4% paraformaldehyde (PFA) in

PBS for 10 min at 25°C Fluorescence of the soluble DNA

(apoptotic) fragments was measured in a Leica DMIL Inverted

Microscope (Leica Microsystems GmbH, Wetzlar, Germany)

at an excitation wavelength of 365 nm and emission

wave-length of 460 nm

Apoptosis studies

Determination of apoptotic cells by fluorescent staining was

carried out as described previously (van Engeland et al., 1998;

Zhuang et al., 2013) The Annexin V-FITC Apoptosis

Detec-tion Kit was obtained from Strong Biotech CorporaDetec-tion

(Taipei, Taiwan) The COLO 205 cells (2× 105cells·per well)

were fluorescently labelled for detection of apoptotic and

necrotic cells by adding 100μL of binding buffer, 2 μL of

Annexin V-FITC and 2μL of propidium iodide (PI) to each

sample Samples were mixed gently and incubated at room

temperature in the dark for 15 min Binding buffer (300μL)

was added to each sample immediately before flow

cytomet-ric analysis A minimum of 10 000 cells within the gated

region was analysed

Flow cytometric analysis for cell cycle

Compound 11e (final concentration 50 nM) was added to

COLO 205 cells for 0, 12, 24, 36 and 48 h Cells were fixed in

70% EtOH overnight, washed twice and resuspended in PBS

containing 20μg·mL−1 PI, 0.2 mg·mL−1 RNase A and 0.1%

Triton X-100 in the dark After 30 min incubation at 37°C,

cell cycle distribution was analysed using ModFit LT Software

(Verity Software House, Topsham, ME, USA) in a BD

FACS-Canto flow cytometer (Becton Dickinson, San Jose, CA, USA)

Molecular modelling

The crystal structure of microtubules in complex with

N-deacetyl-N-(2-mercaptoacetyl)-colchicine (DAMA-colchicine)

was downloaded from the Protein Data Bank (PDB entry

1SA0: http://www.rcsb.org/pdb/home/home.do) (Ravelli

et al., 2004) Docking studies were performed for proposed

11e in the colchicine-binding site of tubulin The AutoDock

Vina (The Scripps Research Institute, Molecular GraphicsLab., La Jolla, CA, USA) was used to perform docking calcu-lations (Trott and Olson, 2010) The final results were pre-pared with PyMOL (v 1.3) (Schrödinger LLC., ShanghaiOffice, Shanghai, China) in Windows 7 After removing theligand and solvent molecules, hydrogen atoms were added toeach amino acid atom The three-dimensional structure ofcompound were obtained from ChemBioDraw ultra 12.0(PerkinElmer Inc., Waltham, MA, USA) followed by MM2energy minimization Docking was carried out by AutoDockVina in the colchicine-binding pocket Grid map in Auto-Dock 4.0 was used to define the interaction of protein andligand in the binding pocket For compound binding into thecolchicine-binding site, a grid box size of 25× 25 × 25 points

in x, y and z directions was built and the grid centre waslocated in x= 116.909, y = 89.688 and z = 7.904

Localization of microtubules

After treatment, cells were fixed with 4% PFA in PBS, blockedwith 2% BSA, stained with anti-tubulin monoclonal anti-body, and then with FITC conjugated anti-mouse IgG anti-body PI was used to stain the nuclei Cells were visualizedusing a Leica TCS SP2 Spectral Confocal System (LeicaMicrosystems GmbH)

Mitochondrial membrane potential analysis

Cells were plated (6 well plates) at 1.0× 106cells·per well and

treated with 50 nM 11e for 6–24 h Mitochondrial

mem-branes were stained with 0.5 mL JC-1 working solution (BDMitoScreen Kit; BD Biosciences Pharmingen, San Diego, CA,USA) added to each sample Samples were incubated for10–15 min at 37°C in the dark Mitochondrial membranepotential was measured using the BD FACSCanto flow cyto-meter (Becton Dickinson)

Western blot assay

The treated cells (1× 107 cells·10 mL−1 in 10 cm dish) werecollected and washed with PBS After centrifugation, cellswere lysed in a lysis buffer The lysates were incubated on ice

for 30 min and centrifuged at 12 000 g for 20 min

Superna-tants were collected and protein concentrations were thendetermined using the Bradford assay After adding a 5 ×sample loading buffer containing 625 mM Tris-HCl, pH= 6.8,

500 mM dithiothreitol, 10% SDS, 0.06% bromophenol blueand 50% glycerol, protein samples were separated by electro-phoresis on 10% SDS-polyacrylamide gel and transferred to anitrocellulose membrane Immunoreactivity was detectedusing the Western blot chemiluminescence reagent system(PerkinElmer, Boston, MA, USA)

The synthetic procedures for the new 4-substituted

benzyloxyquinolin-2(1H)-ones (7a–e∼ 15a–e) are illustrated British Journal of Pharmacology (2015) 172 1195–1221 1199

in Scheme 1 A general synthetic approach to the key

inter-mediate 4-hydroxyquinolin-2(1H)-one is the Knorr quinoline

synthesis, which involves cyclization and dehydration

of a transient β-ketoanilide, formed by condensation of a

β-keto ester and aniline at relatively high temperature

More specific synthetic approaches include cyclization of

N-acetylanthranilic acid derivatives (Buckle et al., 1975),

con-densation of malonates/malonic acid with anilines usingZnCl2and POCl3(Zhang et al., 2008; Priya et al., 2010), Ph2O

(Ahvale et al., 2008) and cyclization of malonodianilides with PPA (Cai et al., 1996; Park et al., 2007; Moradi-

e-Rufchahi, 2010), CH3SO3H/P2O5 (Kappe et al., 1988) and

p-toluenesulfonic acid (Nadaraj et al., 2006) In our study,

4-hydroxyquinolin-2(1H)-one derivatives (5a–i) were

synthe-O

OO

ONH

Reagents and conditions: (A) 130°C with PPA (B) K2CO3/DMF, 80–90°C

1200 British Journal of Pharmacology (2015) 172 1195–1221

sized by treatment of a substituted aniline (1a–i) with

diethylmalonate (2) in one flask (Mohamed, 1991; Arya and

Agarwal, 2007), followed by cyclization of the formed

mono-anilide (3a–i) or malondimono-anilide (4a–i) precursors in the

presence of PPA The target 4-benzyloxyquinolin-2(1H)-one

derivatives 7a–e, 8a–e, 9a–e, 10a–e, 11a–e, 12a–e, 13a–e,

14a–e and 15a–e were synthesized by reaction of the

inter-mediate 4-hydroxyquinolin-2(1H)-one derivatives 5a–i with

various benzyl halide 6a–e in the presence of K2CO3 and

DMF (Guo et al., 2009; Deng et al., 2010) All synthetic

prod-ucts were characterized by IR, 1H and 13C NMR and mass

spectroscopy

The 2-quinolones have a minor tautomeric structure

(2-hydroxyquinoline) because of protonation of the

carbonyl oxygen (Lewis et al., 1991) Deprotonation of the

2-quinolone would cause ring resonance and electron

shifting within the N-1, O-2, C-3 and O-4 positions of

the 4-hydroxyquinolin-2(1H)-one derivatives (Figure 2A)

(Pirrung and Blume, 1999) Consequently, earlier reports

have indicated that 4-hydroxyquinolin-2(1H)-ones could be

alkylated at the 1-NH, 2-OH, 4-OH or 3-CH position (Park and

Lee, 2004; Ahmed et al., 2010; 2011) Therefore, we

con-firmed the structures of our synthesized compounds using

NMR spectroscopic analyses The 1H NMR spectrum of

4-benzyloxyquinolin-2(1H)-one derivatives 7a–e ∼ 15a–e

featured a singlet for O-linked C(9)-H 2 methylene protons

between 5.13 and 5.27 ppm, a singlet for a C(3)-H proton

between 5.80 and 6.09 ppm and a broad singlet for an

exchangeable NH group between 10.47 and 11.54 ppm The

chemical shifts for the benzylic CH2 were consistent with

O-alkylation rather than N-alkylation (Park and Lee, 2004).

The 13C shifts for O-alkylated compounds are typically

downfield (higher ppm value; 52.7–68.4) compared with

N-alkylated compounds (lower ppm value; 28.6–45.0)

(LaPlante et al., 2013) The13C NMR spectra of 7a–e ∼ 15a–e

included an O-linked methylene carbon between 65.74 and

70.74 ppm, which again indicated O-alkylation Furthermore,

regioselective alkylation at the 4-OH position was confirmed

by two-dimensional NMR study via heteronuclear

multiple-quantum correlation and heteronuclear multiple-bond

corre-lation (HMBC) spectroscopy experiments that disclose the

relationship between 1H and 13C coupling In the case of

compound 11e, as shown in Figure 2B, the 4-O-linkage was

supported by observation of3J-HMBC correlations between

C(9)-H methylene protons (δH 5.20) on the

3′,5′-dimethoxybenzyloxy moiety with the carbon at C(4) position

(δc161.89) of the 2-quinolone core, which shows a further

correlation with the C(5)-H proton (δH 7.14–7.26,

over-lapped) In other words, O4-alkylation was determinedthrough the observation of H9/C4 and H5/C4 cross-peaks.These data proved that 3′,5′-dimethoxybenzyloxy moiety is

attached to the 4-O-position of the 2-quinolone core

struc-ture Furthermore, the IR spectra of 7a–e ∼ 15a–e possessed a

characteristic absorption band for an amido C = O group(1633–1674 cm–1)

Biological evaluation and SAR analysis

All newly synthesized target compounds (7a–e, 8a–e, 9a–e,

10a–e, 11a–e, 12a–e, 13a–e, 14a–e and 15a–e) were

assayed for growth inhibitory activity against Detroit 551cells (human normal skin fibroblast) and four cancer cell lines– HL-60 (leukaemia), Hep 3B (hepatoma), H460 (non-small-cell lung carcinoma) and COLO 205 (colorectal adenocarci-noma) Cells were treated with compounds for 48 h and cellproliferation was determined by MTT assay The antiprolif-erative activity of each compound was presented as the con-centration of compound that achieved 50% inhibition (IC50)

of cancer cell growth The results are summarized in Table 1.Collectively, the present series of novel 4-benzyloxyquinolin-

2(1H)-one derivatives exhibited a range of potencies against

the four tested tumour cell lines Among them, compounds

7e, 8e, 9b, 9c, 9e, 10c, 10e, 11c and 11e displayed high

potency against HL-60, Hep3B, H460 and COLO 205 cells,with IC50value less than 1μM (Table 1) Notably, 11e dis-

played the most prominent growth inhibitory activitiesagainst these four cell lines with IC50values ranging from 14

to 40 nM Moreover, none of the active compounds showedcytotoxicity (IC50> 50 μM) towards Detroit 551 cells Theseresults suggested that this new series of 4-benzyloxyquinolin-

2(1H)-one derivatives could effectively suppress tumour

growth without causing toxicity to normal somatic cells.Based on the biological data obtained so far, SAR correla-tions were determined Firstly, we evaluated the effects of

Figure 2

Alkylation of 4-hydroxyquinolin-2(1H)-ones (A) Tautomerism of 4-hydroxyquinolin-2(1H)-one derivatives (B) Key HMBC correlations (blue

arrows) of 11e indicated alkylation at the 4-OH position.

British Journal of Pharmacology (2015) 172 1195–1221 1201

1202 British Journal of Pharmacology (2015) 172 1195–1221

methoxy substitution of the C-4 benzyloxy ring (C ring) on

the cytotoxic activity Generally, compounds with 3

′,5′-dimethoxybenzyloxy side chain (7e–15e) showed the

highest potency in their respective series (7–15) Among

them, compounds 7e, 8e, 9e, 10e and 11e exhibited

signifi-cant activity against Hep 3B, H460 and COLO 205 cancer cell

lines (IC50< 1 μM) These results indicated

3′,5′-dimethoxy-benzyloxy substitution is preferred relative to other benzyl

substitution Compounds 9b, 9c, 10b, 10c, 11b, 11c with a

2′- or 3′-methoxybenzyloxy side chain demonstrated

moder-ate activity (IC50 0.2–5.0μM), whereas compounds bearing

side chains of benzyloxy or 4′-methoxybenzyloxy were

inac-tive (IC50> 50 μM) or exhibited only marginal activity (IC50

4.5–10μM)

Next, we explored the SAR of the 2-quinolone A ring

Compounds with a substituted benzyloxy moiety at C-4 and

various functional groups at C-6, -7 and -8 were studied and

different anticancer effects were found Regarding the C-6

substitution, compound 8e (6-fluoro), 9e (6-chloro), 10e

(6-methyl) and 11e (6-methoxy) were more potent than 7e

(no substitution) Moreover, compound 11e (IC50 0.014–

0.04μM) displayed the strongest growth inhibitory activity

among the C-6 substituted compounds, suggesting that the

C-6 methoxy group might play a pivotal role Moving the

methoxy group from C-6 to C-7 (12e, IC502.11–4.9μM) or

C-8 (13e, IC502.2–3.8μM) dramatically decreased inhibitory

activity Activity also decreased when the C-8 methoxy of 13e

was replaced with chlorine (14e), while activity was retained

when the methoxy was replaced with methyl (15e) Thus, in

this series of 4-benzyloxy-2-quinolones, optimal

antiprolif-erative effects were found with a 6-methoxy group on the

2-quinolone ring

In the present work, the earlier findings can be

summa-rized in the following two SAR conclusions:

(i) The in vitro anticancer activity of the substituted

benzy-loxy moiety (C ring) on the 4-position of 2-quinolone

derivatives can be ranked in the following order of

decreasing activity: 3′,5′-dimethoxybenzyloxy (7e–15e) >

3′-methoxybenzyloxy (7c–15c) ≧ 2′-methoxybenzyloxy

(7b–15b) > benzyloxy (7a–15a) ≧ 4′-methoxybenzyloxy

(7d–15d).

(ii) C-6 substituents on the 2-quinolone (A ring) resulted in

better activity compared with C-7 and C-8 substituent

The following rank order of in vitro anticancer activity

was found relative to the identity of the C-6 substituent:

6-methoxy > 6-chloro ≧ 6-methyl > 6-fluoro ≧ no

substitution

Anticancer drug screen panel of compound

9b, 9c, 9e and 11e against NCI-60 human

cancer cell lines

We selected four potent compounds 9b, 9c, 9e and 11e and

submitted them for screening against the NCI-60 HTCL panel

assay through the US NCI DTP (Boyd and Paull, 1995;

Shoemaker, 2006) The cell lines used in this assay represent

nine tumour subpanels, leukaemia, melanoma and cancers of

lung, colon, brain (CNS), ovary, kidney, prostate and breast

Initially, the compounds were added at a single dose (10μM)

and the culture was incubated for 48 h End-point

determi-nations were made with a sulforhodamine B assay Results foreach compound are given in Table 2, with a negative value inthe cell growth percentage indicating an antiproliferative

effect against that cell line Compound 9b displayed positive

cytotoxic effects towards 11 out of 60 cell lines, and the

positive cytotoxic proportions of 9c, 9e and 11e were 10/59, 18/60 and 26/57 Our prominent compound 11e exhibited

inhibitory effects ranging from –59% to −0.80% At theprimary single high dose 10μM (10−5M), 9b, 9c, 9e and 11e

showed greatest effects against colon carcinoma COLO 205with cell growth percentage of−55, −57, −64 and −59 respec-tively The melanoma MDA-MB-435 cell line was also sensi-tive to these compounds (growth percentages−46%, −43%,

et al., 2010) The GI50value (growth inhibitory activity) responds to the concentration of compound causing 50%decrease in net cell growth, the TGI value (cytostatic activity)

cor-is the concentration of compound resulting in total growthinhibition (100% growth inhibition) and LC50value (cyto-toxic activity) is the lethal dose of compound causing net50% death of initial cells The calculated results are presented

as log concentration (given in the Supporting Information),

as shown in Table 3 The NCI data revealed broad-spectrum

sensitivity profiles for 9b, 9c, 9e and 11e towards all nine

cancer subpanels with GI50values less than 1μM, and lessthan 0.01μM (log GI50< −8.0) against some cell lines for 9e and 11e The anticancer effects of these compounds were

comparable with those of fluorouracil (5-FU), which is widely

used clinically for treating cancer (Longley et al., 2003) These

screening results were in good agreement with the single dose

results, showing broad anticancer spectra for 9b, 9c, 9e and

from 0.01 to 8.08μM in 51 of the 56 cell lines, with GI50

values below 0.01μM in five cell lines (leukaemia K-562 and

SR, non-small-cell lung cancer NCI-H522, colon cancerCOLO 205, melanoma MDA-MB-435)

To further determine which cancer subtypes were moresensitive to these 4-benzyloxy-2-quinolones, we calculatedsubpanel-selectivity ratios based upon GI50values The calcu-lated results are shown in Table 4 and Figure 3 Selectivityratios less than 3 were rated non-selective, ratios rangingfrom 3 to 6 were termed moderately selective and ratiosgreater than 6 were designated highly selective (Boyd and

Paull, 1995; Noolvi et al., 2012; Chen et al., 2013a) With all

ratios less than 3, compounds 9b and 9c were rated selective towards all nine subpanels Interestingly, both 9e and 11e, which contain a 3′,5′-dimethoxybenzyloxy moiety, were much more selective than 9b and 9c (Figure 3) As shown in Table 4, the average selectivity ratios of 9e and 11e

non-(ratios= 6.51 and 4.05) were higher than those of 9b and 9c

(ratios= 1.12 and 1.28) Compound 9e exhibited selectivity

against leukaemia, colon cancer, CNS cancer, melanoma,renal cancer and prostate cancer In terms of the total MID

(an average sensitivity across all cell lines), compound 11e

displayed significant activity (0.31μM) and was moderately

British Journal of Pharmacology (2015) 172 1195–1221 1203

1204 British Journal of Pharmacology (2015) 172 1195–1221