The vasopressin Avprlb receptor: Molecular and pharmacological studies

The distribution, pharmacology and function of the arginine vasopressin (Avp) lb receptor subtype (Avprlb) has proved more challenging to investigate compared to other members of the Avp receptor family. Avp is increasingly recognised as an important modulator of the hypothalamic–pituitary–adrenal (HPA) axis, an action mediated by the Avprlb present on anterior pituitary corticotrophs. The Avprlb is also expressed in some peripheral tissues including pancreas and adrenal, and in the hippocampus (HIP), paraventricular nucleus and olfactory bulb of the rodent brain where its function is unknown. The central distribution of Avprlbs is far more restricted than that of the Avprla, the main Avp receptor subtype found in the brain. Whether Avprlb expression in rodent tissues is dependent on differences in the length of microsatellite dinucleotide repeats present in the 5′ promoter region of the Avprlb gene remains to be determined. One difficulty of functional studies on the Avprlb, especially its involvement in the HPA axis response to stress, which prompted the generation of Avprlb knockout (KO) mouse models, was the shortage of commercially available Avprlb ligands, particularly antagonists. Research on mice lacking functional Avprlbs has highlighted behavioural deficits in social memory and aggression. The Avprlb KO also appears to be an excellent model to study the contribution of the Avprlb in the HPA axis response to acute and perhaps some chronic (repeated) stressors where corticotrophin-releasing hormone and other genes involved in the HPA axis response to stress do not appear to compensate for the loss of the Avprlb.

The availability of several specific ligands for these receptorsh as considerably aided in the characterisation of the Avpr1a, Avpr2 and Oxtr but, until recently,r esearch on the Avpr1b has been hampered by al ack of Avpr1b-specific ligands. Avpr1br esearch has focussed primarily on the detection of receptor transcript levels (indirectly inferring receptor protein levels), genetic KO studies ande xperiments with the naturally Avp-deficient Brattleboro (di/di)r ats. This review details recent pharmacological and KO data on ther oleo ft he Avpr1b in brain, pituitarya nd peripheral tissues with particular emphasis on its function in theh ypothalamic-pituitary-adrenal (HPA) axisresponse to stress.

V1B receptor distribution
The Avpr1a and the Oxtr have been well characterised in the rodentb rain and are thought to be the likely substrates for central actions of Avpa nd/or Oxt (de Wied et al. 1993;Burbach et al. 1995; Barberis and Tribollet 1996).M orer ecentd ataf rom KO animals suggest specific behavioural deficits in social memory and aggression are directly due to the absence of central Avpr1bs (Wersinger et al. 2002).The search for central Avpr1bs has proved more elusive than thato ft he Avpr1as and Oxtrsa st he shortage of specific ligands has prevented binding studies to visualise the Avpr1b protein. Nevertheless, an umber of studies utilising immunohistochemistry and in situ hybridisation histochemistry( ISHH) haves hown that the Avpr1b is expressed centrally,w hile reverse transcriptionpolymerase chain reaction (RT-PCR) and functional studies haved emonstrated Avpr1bs in an umber of peripheral tissues (e.g. Lolait et al. 1995;Vaccarietal. 1998;Hernando et al. 2001;O'Carroll et al. 2008).
Although the highest concentration of Avpr1bs is found within anterior pituitaryc orticotrophs (Jard et al. 1986;Antoni 1993;Lolait et al. 1995), several studies suggest awide central (Barberis andT ribollet 1996;Vaccarieta l. 1998;Hernando et al. 2001) and peripheral (Lolait et al. 1995;Saitoe ta l. 1995;Ventura et al. 1999;Oshikawa et al. 2004) distribution in rodents. Analysis of various brain regions and peripheral tissues suggests that Avpr1bt ranscript levels mayb et oo low to be reliably detected by Northernblot analysis and often depends on RT-PCR to detect possible Avpr1b expression (Lolait et al. 1995).Arecent distribution study using ISHH with probes directed to 5 0 or 3 0 untranslated regions of the Avpr1bm RNA details am ore restricted patterno f Avpr1bm RNAi nm ouse brain than previously reported (Young et al. 2006). The riboprobesu sed by Yo ungand co-workershad low sequence identity to otherA vp receptorst om inimisec ross-reactivity with related mRNA sequences. This study shows Avpr1b mRNAt ob em ost prominent in the CA2 pyramidal neurons of the mouse and human HIP while receptor transcripts are also found in thep araventricular nucleus (PVN) and amygdala, albeit at al ower level. All of these studies infer receptor expression by determiningm RNAt ranscriptl evels rather than receptor protein levels. However, in the absence of specific radiolabelled ligands of high-specific activity (whichm ay notp rovide detailed anatomical resolution), mRNA expression coupled with immunohistochemical techniques can accurately reflect receptor protein distribution and quantity.
In the pancreas, Avph as been shown to act on Avpr1bs presenti ni slets to stimulate the secretion of insulinf rom b cells where it may act synergistically with corticotrophin-releasing hormone (Crh) (Oshikawa et al. 2004;O'Carroll et al. 2008). While Avpacts on the Avpr1b to decrease blood sugar levels through insulin release, it can also act in opposition to this by stimulating glucagonr elease (Yibchok-anun and Hsu 1998) and promoting hepatic glycogenolysis (Kirke ta l. 1979). Importantly,A vp-mediated glycogenolysis acts viat he Avpr1a subtype present in hepatic tissue (Morel et al. 1992) suggesting bifunctional but oppositeroles of Avpinglucose homeostasis by employing two different receptor subtypes. Exactly whichr ole Avpp lays in regulating glucose balance at the pancreatic level is dependent on the local glucose concentrationp resenti nt his tissue (Abu-Basha et al. 2002).One study on aglucagon-secreting hamster apancreatic cell lines uggests thatA vp-induced glucagon secretion is mediatedv ia the Avpr1b since the Avpr1ba ntagonist SSR149415 potentlya ntagonises Avp'se ffects in these cells (Folny et al. 2003). The specificity of SSR149415 has, however, been questioned with evidence of activity in Chinese hamstero vary celll ines expressing recombinant humanO xtrs ( Griffante et al. 2005;Hodgson et al. 2007).T his may be an important consideration, as Oxtrsinaddition to Avpr1bs are apparently present in pancreatic islets (Oshikawa et al. 2004) and havebeen shown to causei nsulin/glucagon release (Jeng et al. 1996;Yibchok-anun et al. 1999).One laboratorythat used agonists and antagonists to Avp/Oxt receptors (but notably none selective for Avpr1bs) in a -cell lines (Yibchok-anun and Hsu 1998) and perfused rat pancreas (Yibchok-anun et al.1 999) produced contrastingr esults, thel atters tudy suggesting glucagon secretion in response to Avp, andO xt is mediated through the activation of Avpr1bsa nd Oxtrs, respectively,rathert han any activity of Avpon Oxtrs. On the otherhand, studies in Avpr1b KO mice suggest both Avpa nd Oxt can stimulate glucagon secretion through Oxtrs ( Fujiwara et al. 2007b). Avpmay play arole in the hypersecretionofglucagon from the pancreas of diabetics (Yibchok-anun et al. 2004),w hich is likelyt oi nvolve theA vpr1b. Pancreatic b cells isolatedf romm icel acking functional Avpr1bs unsurprisinglyd isplay ab lunted insulins ecretion (Oshikawa et al. 2004),r einforcing the role of Avpr1bs in this tissue. Interestingly, further studies with this Avpr1b KO line show an increased sensitivity of Avpr1b KO mice to the metabolic effects of insulin ( Fujiwara et al. 2007a). To gether with the discoveryo fA vpr1b mRNA by RT-PCR in white adipose tissue, Fujiwara andc o-workerss uggest that the disruptioni ni nsulin -adipocyte signalling may lead to altered metabolism of glucose in Avpr1b KO mice. Whether this is due to alack of Avpr1b influence in the pancreas, white adipose tissue, or both, is unclear.
Grazzini and co-workersdemonstrated the presence of Avpr1b mRNAbyR T-PCR in the medullabut not in the cortex of the rat adrenal gland. The cortex expresses Avpr1at ranscripts primarily in the zona glomerulosa (Guillon et al. 1995;Grazzini et al. 1996) where this receptor regulates steroid secretion in vitro (Grazzini et al. 1998).T hese studies also show that Avpprecursor mRNA and Avppeptideare presentin the adrenal medullas uggesting Avpc an be released withinthe tissue, possibly acting in an autocrine/paracrine mannert or egulate adrenal function. Stimulation of the Avpr1b in the rat adrenal medulla causes catecholamine secretion ( Grazzini et al.1 998). The presence of Avpr1as in the adrenal cortex and Avpr1bs in thec hromaffin cellso ft he medulla provides strong evidence of an independent modulatoryrole of each receptor in discrete regions of adrenal tissue. The presence, however, of both Avpr1s in the human ( Grazzini et al.1 999) adrenalm edulla indicates ap ossible co-expression of Avpr1r eceptors in somes pecies. This suggests ap ossible overlap of function distinct from the roles already noted and may reflect the action of Avpo riginating from different sources (e.g. pituitaryv s. local tissuer elease (Gallo-Payetand Guillon 1998)), althoughthe medullarycell type that expresses Avpr1as has yett ob ei dentified. Notably, the plasma catecholamine response to forced swimming and sociali solations tressi sa ttenuated in Avpr1bK Om ice (Itoh et al. 2006).
The central, pituitarya nd peripherale xpression of the Avpr1bg enem ay be influenced by activity of elements in the upstream Avpr1bp romoter region. In vitro studies using cells transientlytransfected with a rat Avpr1bg ene promoter sequence have identified regulatoryGAGArepeats that influence Avpr1b gene transcription ( Vo lpi et al. 2002). This provides a possible mechanismo fp hysiological Avpr1b gene regulation that may enable different levels of Avpr1b expression in different tissues or species. When we compared the 5 0 microsatellite region in the 5 0 Avpr1b promoter sequenceofdifferent mouse strains, amajor size differencei nm icrosatellite length between the C57BL6J/OlaHsds traina nd Balb/cOlaHsd and 129S2/SvHsd strains was observed( see, Figure 1A). Further analysis of the sequence details differences in the number of CTandCArepeats between strains (see, Figure 1B)t hat mayc onfer changes in promoter activity.Basal promoter activity of aBalb/c 5 0 fragment is threefold greater than that of the C57BL/6 strain confirming an increase in Avpr1b promoter activity with the "long" formofmicrosatellite (see, Figure 1B and C), in areporter assay in COS-7cells. The impact of microsatellite DNA sequences on receptor expression andb ehavioural phenotypesh as been examined in studies on the effects of Avpr1a expression on socialbehaviour in voles. Affiliative behaviourssuch as pair bonding haveb een attributed to changes in Avpr1ae xpression patterns caused by microsatellite length variations in the 5 0 Avpr1a regulatoryr egion (Hammock andY oung 2002;Hammock and Yo ung 2005).D ifferences in Avpr1b protein expression that result from variancesi ng enep romotera ctivity between mouse strains maybeacontributing factor to varying susceptibilityt os tress. Several neurogenic, psychogenic and systemic stressorshavebeen tested in different strains of micet or eveal as train-dependant stressr esponse (Anisman et al. 2001). Interestingly, C57BL/6ByJ miced isplay higher levels of plasma CORTaswellasincreases in stress-related behaviours comparedt ot he Balb/cByJs train, strengthening supportf or Avpr1b'si nvolvement in mouse stress susceptibility (Anisman et al. 2001).Itisimportant to note,however, that similar 5 0 microsatellite sequences are not presentinthe human Avpr1b gene.
Ther elativelyr ecente mergence of data from genetic KO studies and the development of promising pharmacological compounds haveg iven the task of characterising the function of central andp eripheral Avpr1br enewed vigour. Avpr1b KO micet ogether with the long-standing subject of Avpr esearch, the Brattleboro rat, serve as robust systems with whichto study the role of the Avpr1b and Avpinthe HPAaxis response to stress.

The HPAa xisa nd stress
The complex homeostatic controlthat constantly acts to resist challenges and fluctuations in the internal environment that may threaten the survival of an organism is necessaryf or life. As ar esult of any deviation in conditions, ah ost of physiological and behavioural changes occur that allow an organism to adaptt os uch challenges and restore the homeostatic balance. One such neuroendocrine system that is activated in stressful circumstances is the HPAa xis. The end product of HPAaxis activation is an increase in circulating glucocorticoids that, togetherwith other stressm ediators, act on target cells to enable the organism to cope with the stress. Consequences of elevated glucocorticoids are widespreada sc ytosolic glucocorticoid receptorsa re present in most central and peripheral tissues. The most profound effects of elevated glucocorticoid levelsa re immunological and metabolic changes. Glucocorticoids ecretion, in concertw ith catecholamine release due to rapidly 301 bp 186 bp promoter of the mouse Avpr1b gene using genomic DNA from three different mouse strains (1, 2, Balb/cOlaHsd; 3, 4, 129S2/SvHsd; 5, 6, C57BL/6JOlaHsd). The PCR products weregenerated using 100 ng genomicD NA,0 .5 units of AmpliTaq polymerase (Applied Biosystems, Warrington, UK) and primers: upstream 5 0 GCGA GC TCTTTC ACACAT GCC TAGG3 0 incorporating SacI restriction site (underlined); downstream 5 0 CAGGAT CCACTG AGCACC AACTCA C3 0 incorporating aBamH1 restriction site (underlined) with cycling conditions of 958 C1min followed by 40 cycles of 948 C for 1min, 62.58 C1min, 728 C30sfollowed by afinal 728 Cstep for 10 min and a48 Csoak. Gel electrophoresis (2% agarose) of a15-m l sample from the 50-m lr eaction volume revealed a3 01 bp product (bp1755-2055) with Balb/cOlaHsd and 129S2/SvHsd DNA templates and a , 186 bp product with C57BL6J/OlaHsd strain DNA template. The PCR products were subcloned into pGEM4Z vector using Bam-H1 and Sac-1 restriction enzymes and sequenced.
(B) Alignment of 129SV and C57BL strain microsatellite sequences between bases 1801 and 1988 of Genbank Acc#AF152533 (ending 824 bases upstreamofthe initiating ATGcodon). Microsatellite CA and CT repeats are highlighted with dashes representing nucleotides that are absent in the C57BL mouse 5 0 promoter region that gives rise to the shorter sequence.T he longer formi sp resent in 129/Svj, J1, SWR/J, AKR/J, FVBa nd CD1( USA) 129S2/SvHsd and Balb/cOlaHsd( UK)s trains, whereast he shorter form is in C57BL/6Ncr (USA),C 57BL/6JOlaHsd( UK)a nd C57BL/6J (USA) strains.( C) Fragments of 5 0 Avpr1b gene promoter region incorporating the microsatellite repeats were generated by PCR using 100 ng genomic DNA, Herculase II Fusion DNA polymerase (Stratagene (Agilent Te chnologies), Stockport,U K), and primers corresponding to a , 1.1 kb region of the mouse Avpr1b gene (from bp1755-bp2823, Genbank accession number AF152533) using genomic DNA extracted from the Balb/c, C57BL/6 and 129 Sv mouse strains (Harlan, Bicester, UK). Amplified fragments from each strain weret hen subcloned into ap GL3 basic dualluciferase reporter assaysystem (Promega, Southampton, UK) for expression in COS-7 cells. Values (mean^SEM) are expressed as a percentage of control pGL3 basic activity.The promoter activity of vectorsc ontainingB alb/c and 129 Sv constructs is , 3 £ greater than that of C57BL/6 ( *** ¼ P , 0.001).W hent he Balb/c microsatellite region was replaced with the C57BL/6 microsatellite sequence in the Balb/c reporter construct, activity diminished to that of the C57BL/6 strain. activated sympatho-adrenomedullarys ystem, prepares tissues for the physical "load" that may be requiredbythe body as partofacoordinatedresponse to manageo rt ackle the stress. The HPAr esponse to stressr elies on several key mediatorsthat fine-tune glucocorticoid release, which is dependent on anumberoffactorssuch as the nature of the stressor and the immunologicals tateo ft he organism, tailoring the response to the specific stressor concerned. Whens ubjected to as tressor, the challenge is perceivedi nb rain regions appropriate to the nature of the stressor. Interpretation of these signals in relevant brain areas, such as various limbic or hindbrain regions, leads to activation (or inhibition) of the PVN (Hermane ta l. 2003).
The parvocellular subdivision of the PVN (pPVN) is the most important site among several hypothalamic nucleithat regulate the HPAresponse, as this is where stress signals are integrated anda djusted before peripheral signalling is initiated.O ncea ctivated, pPVNp rojections that terminatei nt he external zone of the median eminence release Avpand Crh into the hypophysial portal blood (Antoni 1993). These two ACTH secretagogues act on Avpr1bs and Crh type 1receptors(Crhr1; coupled to G s andadenylate cyclasea ctivation) present in pituitaryc orticotrophs causing the release of ACTH into the peripheral blood system. ACTH via melanocortin receptorsp resenti n zona fasciculata cellsofthe adrenal cortex stimulates a rapid secretion of glucocorticoids into the peripheral blood supply.G lucocorticoids such as corticosterone ((CORT) in rodents) and cortisol (in humans)inturn provide negative feedbackcontrolofthe HPAaxis via pituitary, pPVN and higher brain centreg lucocorticoid receptors(e.g. HIP).
The dominance of Crh as the primary ACTH secretagogue is still the prevailing view; however, numerous direct andi ndirect neuronal inputs into the pPVN, as well as humoral influences from blood or cerebrospinal fluid-bornes tress signals, dynamically modulatet he contribution of Crh to the stress response. The influence of Avp/Avpr1b mayb eo f greater significance than that of Crh/Crhr1 in some stressc ircumstances, such as in response to some chronic (repeated) stressors  or to novel stressorss uperimposed on ar epeated stress paradigm ).E vidence of as witch from Crh-ergic to vasopressinergic pPVN drive in response to some chronic stressorsr einforces the concept thate achr esponse is tailored to that specific stressor andt hatA vp mayb ea ni ncreasingly important mediator in these circumstances (Aguilera 1994;M ae ta l. 1999; Aguilerae ta l. 2008). Moreover, studies of some specific acutes tressors indicate that Avpm ay be preferentially released in favour of Crh (e.g. insulin-induced hypoglycaemia (IIH)i nr ats( Plotskye ta l. 1985), ketamine anaesthesia and IIH in sheep (Engler et al. 1989)).
It is important to emphasiset hat in some species (e.g. sheep andh orse) Avpr ather thanC rh appears to be the main ACTH secretagogue (Engler et al. 1989;Alexander et al. 1997).

V1B receptor KO mice
Witht he generation of Avpr1bK Om iceb yu si n 2002, attention was initially directed towards the deficits observed in cognitive andb ehavioural tests (Wersinger et al. 2002;We rsinger et al. 2004).I tw as notedt hat mice lacking functionalA vpr1bs showed impairments in social recognitiona nd reduced responses in some aggression paradigms whereas otherp hysiological and behavioural test responses were normal (Wersinger et al. 2002(Wersinger et al. , 2004(Wersinger et al. , 2007a. Findings from asecond Avpr1b KO line generated by Ta nouea nd co-workersi nitially focussed on the disruptiono ft he HPAa xis (Tanoue et al. 2004),a nd they reported that KO mice havel ower circulating ACTH levels under basal and acutestressconditions. This is in contrast to basal measurements of HPA axis activity seen in our Avpr1b KO mouse colony, whichm aintain normal restingA CTHl evels (Lolait et al. 2007a). Notwithstanding basal HPA axis differences, both Avpr1b KO lines have been used to generate al arge amounto fc ompatible evidence supportingt he involvement of the Avpr1b in stress anda ggression, whichi sa lsol argely consistent with in vivo Avpr1b antagonism with SSR149415 Blanchard et al. 2005; Stemmeline ta l. 2005).

Reduced aggression in KO mice
Avph as been implicated as am oderator of several central behaviourst hat were initially thought,b ased on pharmacological profiles andd ue to its much higherp revalence in the brain, to be mediatedb yt he Avpr1a. Experiments in rodents, particularly hamsters, with Avpr1a antagonists have consistently shown that Avpr1as facilitate somea ggressive behaviour (Ferrisetal. 1997(Ferrisetal. , 2006Caldwell and Albers2004), althought his is yet to be verified in Avpr1a KO mutants (Wersinger et al. 2007b). It is possible that the neuralc ircuitryu nderlying aggression compensates for the loss of the Avpr1a in the global Avpr1aKO. In contrast, evidence of Avpr1b involvement in aggression comesfrom both pharmacological and KO data, e.g. antagonism of the Avpr1b with SSR149415 lowers the frequency and duration of aggressive behaviour in hamsters ( Blanchard et al. 2005) while Avpr1b KO micedisplay reduced attacknumber and longerattack latencyc omparedt ow ild types. This latter observation has been further categorised as ad eficit in the attackc omponent of aggression, as defensive aggression remains intact in mutant animals (Wersinger et al. 2002(Wersinger et al. , 2007a. Furthermore, the reduced aggression phenotype persists when Avpr1bKOmice are crossed with am ore aggressivew ild-derived mouse strain, confirming that the reduced aggression observed is not simply ap eculiarity of laboratory mouses trains (Caldwell and Yo ung2 009).T he specific neural substrate(s) vital for Avpr1b'srole in aggression is not known, noristhe possible interaction between Avpr1a and Avpr1b (or Oxt and the Oxtr for that matter-see, Winslow et al. 2000) in vivo. It shouldb en oted that the central distribution of Avpr1a-andO xtr-binding sites and Avpr1b mRNA expression are clearly distinct but may overlap in some brain regions (e.g. olfactory system) (see, Ta ble IIIi nB eery et al. Caldwell et al. 2008a).
The changes in aggression,aswellasdifferences in socialm otivation (Wersinger et al. 2004) or social memorye videnti nA vpr1b KO mice may be due to deficits in the processing of accessoryolfactorystimuli that are needed to evoke such behaviours ( Caldwell et al. 2008b). It is suggestedt hat the role of central Avpr1bs may be to couple socially relevant cues detected in the accessoryo lfactorys ystem to the appropriate behavioural response (Caldwell et al. 2008b). Intriguingly,b oth Avpr1a andA vpr1b genes aree xpressed in thef orebraino lfactory system (Ostrowski et al. 1994;Hernando et al. 2001). Evidence of pyramidal CA2 Avpr1bs (Hernando et al. 2001;Yo unge ta l. 2006) also suggests a relationship between the deficits in social memoryand the uncoupling of socialc ues from the accessory olfactorys ystem andt he formation or recall of relevant memories (Caldwell et al. 2008b). Basedo n studiesi nA vpr1b KO animals, the central Avpr1b may also be involved in anumber of other behaviours (summarised in Ta ble I). Prepulse inhibition of the startler eflex is attenuated in Avpr1b KO mice (Egashira et al. 2005)  Notes: Note that defensiveaggression behavioursare less influenced by lack of Avpr1b than attacking behaviours. Furthermore, Avpr1b KO mice not only show clear deficits in social memory/recognition but also temporal order memory. Memoryd eficits are not attributable to a reduction in olfactoryperformance that suggests the detection of olfactorycues remains unaffected but an altered processing of olfactorycues results in the behaviour change.Interestingly,disruption in prepulse inhibition to acoustic startle seen in Avpr1b KO mice parallels that seen in some human conditions such as schizophrenia and panic disorders. Injection of Avpr1b KO mice with atypical antipsychotics used to treat schizophrenia appeared to reverse these deficits.A rrowsi ndicate changes observedi nt he behavioural phenotypeo fA vpr1b KO mice compared to wild-type animals.
mouse is nota na ppropriate model for examining stress-induced anxiety or depression (e.g. see Kalueff et al. 2007).

The stress response in KO mice
The impact of AvpinACTHrelease is often regarded as ancillaryt oC rh as Avpa lone is aw eak ACTH secretagogue but togetherw ith Crh it acts synergisticallyt of acilitate ACTH secretion( Gilliese ta l. 1982; Rivier andV ale 1983;Antoni 1993 Shaker stress (see Figure 2) (Lolait et al. 2007b) # 30 min and 4h ( C ) # 30 min, $ 4h ( C ) Ethanol (Lolait et al. 2007b) Notes: This table highlights how the nature of the stressor can influence the effects of the Avpr1b on ACTH and CORTrelease. Stressors that give rise to mismatches between ACTH and CORTdata, and stressors that have shownsome level of adaptation after chronic administration are shown. # ,H ormone values are reduced in KO mice compared to wild-type mice; $ ,h ormone values are unchangedi nK Om ice compared to wild-type mice; * Plasma hormone levels measured following the final acute stress in a1 0-14d aysd aily repeated acute stress paradigm; † "Severe" restraint, mice placed in a50mlfalcon tube with tissue paper packing inserted to restrict any movement of the animal; ‡ "Mild" restraint, mice placed in 50 ml falcon tube.
acute stressinBrattleboro rats Zelena et al. 2009) both often note adisparity between stress-inducedA CTHv ersusC ORTr elease in Avp/Avpr1b-deficienta nimals.I ti si mportantt o recognise that the CORTresponse to ACTH saturates at low circulating ACTH levels (Dallmanetal. 2002). In addition, plasma hormones levels were only measured at one time point in our studies so any incremental changei nC ORTl evels mayh aveb een missed. Furthermore,w eh avea ssumed that the dynamics of stress-induced Avp, Crh and/or other ACTH secretagogue is similar in Avpr1b and wild-type mice. Where stress-induced ACTH release in Avpr1b KO micei sn ot always followed by ap roportional CORTa ttenuation, suggests that CORTm ay be released independently of ACTH in some circumstances. Interestingly, this patterno fA CTH/CORT profilesi nr esponse to acutes tressa lso appearst ob e presenti nn eonatal Brattleboro rats (Zelenae ta l. 2008).Discrepancies between plasma stress hormone levels appear stressor specifics uggestingt hata particular stressor may possess specific characteristics that coulda ctivate an umbero fp athwayst op romote CORTr elease, e.g. viad irects planchnic or other neurali nnervation of the adrenal cortex or medulla (Ehrhart-Bornsteinetal. 1998).Alternatively,adrenal sensitivity to locally synthesised or humoral factors may be altered permitting CORTr elease even when low levelso fc irculating ACTH are present. Adrenal hypersensitivityand ACTH-independent pathways of CORTr elease would likelyb ypass the normal feedback controls that governA CTH release from the pituitary(forreview see Bornsteinetal. 2008). As mentioned above, chronic (repeated) stressh as often been associated with an alteration in the control of ACTH release from what is ap redominantly Crhmediated drive, seen in acutes tress, to an Avpmediated drive speculated to maintain ACTH levels during adaptation to repeateds tress (Harbuz and Lightman 1992;Aguilera 1994;Ma et al. 1997;Aguilera andR abadan-Diehl 2000).A daptation to repeated stress( i.e.l ower ACTH and/or CORT following repeated stress compared to asingle episode of acute stress) is species ands tressors pecific and is nota lways observed (Marti andA rmario 1998;Armario et al. 2004).T he apparent flip in control of ACTH release from Crh to Avpthat is seen in some repeated stressorsi nr ats (e.g. restraint) has made Avpa nd the Avpr1b attractive targets for pharmacological intervention in conditionso fr epeatedo r chronic stress, althoughi ts houldb en oted that Avpd oes notp lay ar ole in HPAa xis responses to all chronic stressors( e.g. chronic morphinei njection (Domokose ta l. 2008)). The preferential expression of Avpi nt he pPVN in adaptation to chronic stress observed in rats is associated with an upregulation of Avpr1bs (but not Crh receptors) in the anterior pituitaryg land (Aguilera 1994).C hronic stressf rom repeated (daily) exposure to an acute stressor leading to ACTH hyperresponsiveness to as ingle, novel stressor episode maya lso involvea ni ncrease in PVN Avp (Ma et al.1999) and pituitaryAvpr1b expression, and pituitaryA CTHh yperresponsiveness (Aguilera 1994).H owever, Avp/Avpr1bd oes not appear to be responsible for HPAa xish ypersensitivityt on ovel stressors .T he mechanism(s) by whichh ypothalamicA vp andp ituitary Avpr1b responsiveness is maintained during stress adaptation suggests the existence of numerous transcriptional and translational regulatory components involved in PVN and pituitaryplasticity that dynamically alter Avpand Avpr1bl evelsa ccording to demand ( Volpi et al. 2004a,b). One view is that Avp/Avpr1bm ay alter corticotroph proliferation and pituitaryr emodelling during prolonged activationo ft he HPAa xis (Subburaju and Aguilera 2007). However, studieso f repeated stress in Avpr1b KO mice andB rattleboro rats, suggest thatthe role of the Avpr1b and its cognate ligandi nt he adaptation of ACTH/CORTl evelst o chronic stressm ay not be as convincing as first thought.
Fort he chronic( repeated)s tressors tested in Avpr1bK Om ice, there is ar eduction in the ACTH response to afinal acutestressfollowing 10 -14days of stressr epeated once daily.T he responses of Avpr1b KO animalsa nd wild-typea nimals exposedt o repeated stressa re summarised in Ta ble II. These studiesh avean umber of salient features: firstly,a s observed in male Brattleboror atss ubjected to repeated restraint ( Zelena et al. 2004),t he reduction in the ACTH response following repeated stressi n Avpr1bK Oa nimals is not often accompaniedb ya similarr eductioni nC ORTr esponses-this mirrors what we have observedi nthese animals'responses to acutestress(see above). Secondly,with the exception of the ACTH response to repeated, severe restraint, the ACTH and CORTresponsestoacute or repeated stressare of equivalent magnitude. This suggests that the Avpr1b participates in the fast ACTH secretion (as seen in the response to acute stress) in repeatedly stressed mice. Andfinally,ofall the repeated stressors studiedinwild-type mice from ourApr1b KO colony, adaptation in ACTH responses was only seen with repeated exposure to shaker stress(SS) (Figure 2). No adaptation in CORTr esponsesw as observed in Avpr1bo rw ild-type mice following any repeated stressparadigm. There is arobust plasma ACTH and CORTi ncrease in response to as ingle, acuteS S episode in wild-type mice; however, after 10 days of repeated SS, the ACTH response is reduced in these animals (see, Figure 2-graphA :p lasma ACTH levels, wild-type acutes tress response vs. wild-type chronic stress response). The acuteA CTH response in Avpr1bK Om ice, while reduced from that seen in wild-type animals,i st he same for repeated SS. Furthermore, the ACTH response to repeated SS in Avpr1bK Om ice is not different from that seen in wild-type mice. It is tempting to speculate that there is aloss of adaptation to repeatedSSinAvpr1b KOsbut such ac onclusion would be tenuous since ACTH levels in acutely stressed Avpr1b KO mice are already veryl ow.T he lack of functional Avpr1b in the KO mice has such ap rofound effecto nt he ACTH response to SS in these mice that any non-Avpr1bmediated adaptation to repeated SS is probably negligible. The studies in Avpr1b KO mice and Brattleboro ratsh ighlight the discrepancyb etween Avp/Avpr1b-mediated ACTH and CORTs ecretion during acute and repeated stress-this mayh ave implications on the potential use of Avpr1b antagonists to ameliorate symptoms of HPAaxishyperactivity in stress-relateddisorders.

Gene expression in Avpr1b KO mice
Whether the deficits observedi nA vpr1b KO mice outlineda bovea re ad irectr esult of disruptiono f Avpr1bsignalling pathways or the result of an altered compensatorye xpression profile is not known. As far as HPAa xis function is concerned, clearly Crh (or Oxt) does notc ompensate for the loss of the Avpr1bi nA vpr1b KO mice. There are some direct changes that occur in KO mice that give rise to phenotypes such as altered glucose metabolisma nd attenuated ACTH release;h owever, indirect changes that affect behavioural systems that mayl eadt ot he deficits seen in Avpr1b KOsh aveyet to be identified. We have used ISHH to assess basal gene transcript levels of anumberofgenes that are closely linked with HPAa xis function and find no significant differences in gene expression between Avpr1bKOand wild-type mice (Figure 3). Comparisons of Oxtr mRNA levelsin Avpr1bK Oa nd wild-type mouse anterior pituitaries appear to suggest an upregulation of OxtrsinA vpr1b KO mice (Nakamurae ta l. 2008).A sO xt at high concentrations can elicit ACTHr elease viat he Avpr1b, and the Oxtr may be expressed in corticotrophs,ithas been suggested that increased expression of Oxtrsm ay be ac ompensatorymechanism through which Avpr1b KO mice and Brattleboro rats can, to some degree, makeu pf or the lack of Avpr1b/Avpmediated ACTH release (Nakamura et al. 2008). We cannott otally rule outt he possibilityt hat mechanisms compensating for the loss of Avpr1b are active in the Avpr1b KO.C hanges in neurochemical networks active centrally (e.g. projections to the PVN or signals within the PVN itself) or at the level of the anterior pituitaryand adrenal may have occurred. We also cannot exclude ar ole of central Avpr1bs (or for that matterA vpr1as) in directly or indirectly influencing HPAa xis activity.N evertheless, stress-induced ACTH levels in our Avpr1bK Os are consistently lower (often to basal levels) than wild-type controls, confirming any compensation (e.g. from the action of Crh) is not sufficient to fully counteract the loss of the Avpr1b (Lolait et al. 2007a,b;Stewarte ta l. 2008a,b). The role of potential Oxtr-mediated ACTH release in Avpr1b KO mice and Brattleboro rats remainst ob ec larified, however, as the Oxtr is predominantly located in lactotrophs rathert han corticotrophs in the anterior pituitary ( Breton et al. 1995) it is likely that the Oxtr plays am inor role Figure 2. Plasma hormone levelso fm ale wild type ( þ / þ )a nd KO ( 2 / 2 )m ice following acute and chronic (repeated) SS. Note that the ACTH response to acute SS in wild type but not Avpr1b KO animals is significantly reduced following repeated SS. Mice of appropriate genotype generated from parental crosses heterozygous for the Avpr1b werer andomly assigned to an experimental group containing7-8 mice. SS consisted of shaking mice in separate clean empty cages on as tandardb ench top orbital shaker (stroke: 38 mm at 150 rpm) for 10 min. Mice were returned to their home cage for afurther 10 min prior to sacrifice. Chronically shaken mice weres tressed once daily for 10 dayswhereas mice in acute groups werehandled for 9daysand stressed once on the 10th day.Control mice were handled once daily and all experiments were performed between 0900 and 1130 hinaccordancewith UK Home Office regulations (Animal Scientific Procedures Act (1986)) and approvedb yt he University of Bristol Ethical Review Board. Plasma ACTH (graph A) and CORTl evels (graph B) were determined by ELISA and EIA, respectively (IDS, Tyne and We ar, UK). Data are expressed as mean^SEM. Significant differences are denoted as ***: P , 0.001, NS: not significant. Figure 3. RepresentativeI SHH photomicrographs of gene expression in brain and pituitaryo fm ale Avpr1b KO and wild-type mice. Riboprobesc orrespondingt ot he following sequencesw ereu sed: 5-HT1a receptor (Htr1a) (bp1230-1597 of Genbank Accession number NM_008308),5 -HT2a receptor (Htr2a) (bp1884-2492 of Acc#NM_172812),5 -HT2cr eceptor( Htr2c) (bp1435-1965o f Acc#NM_008312), brain-derivedn eurotrophic factor (Bdnf) (bp310-822 of Acc#AK017559), cannabinoid CB1R (Cnr1) (bp728-1040 of Acc#Y18374), Crh (bp101-686 of Acc#205769), type 1Crh receptor (Crhr1) (bp160-528 of Acc#NM_007762), glucocorticoid receptor (Nr3c1) (bp439-958 of Acc#X04435), Oxt (bp1753-1865 of Acc# M88355), Avp( bp2966-3388 of Acc# M88354), intronic Avp(bp1965-2406 of Acc#M88354) and pro-opiomelanocortin (Pomc) (bp145-620 of Acc#V01529). The intronic Avpprobe was targeted towards intron 1ofthe Avp/neurophysin II gene to reflect heteronuclear RNA expression levels. These probes were obtained by PCR using genomic DNA, or brain (Crhr1, Crh, Bdnf,Avp and Oxt) or pituitary(Pomc) cDNA from 129 Sv mouse tissue as template. Restriction sites were incorporated into the 5 0 ends of PCR primers to facilitate cloning of the PCR product into vector pGEM4Z. The integrity of all riboprobes was verified by DNA sequencing. ISHH on 12 m ms ections with 35 S-labelledr iboprobes was performed as in ref. ( Lolait et al. 2007a).
compared to the Avpr1b in wild-type animals.The use of conditional KOsa nd emerging pharmacological developments will no doubt help clarify the contributiono ft he pituitaryO xtr in acutea nd chronic stress.

Emerging pharmacological data
Anumber of agonists (such as desmopressin,anAvpr2 agonist) and antagonists (such as relcovaptan, an Avpr1aa ntagonist) are available and frequently used to study the Avpr1a, Avpr2a nd Oxtr (forr eview see Lemmens-Gruber and Kamyar 2006;Manning et al. 2008).Some of these are in clinical trials or have been approved for use as treatments in disorderss uch as nocturnal enuresis andneurogenic diabetes insipidus. The search for Avpr1b pharmacological tools has gained much impetusd ue to theirp otential as treatments for conditions associated with chronic stresssuch as anxiety anddepression (forshortreview see Griebel et al. 2005;Arban 2007).The most widely used non-peptide Avpr1b antagonist, SSR149415,has been used extensively in researchs ince its characterisation in 2002 ) (see, Ta bles III andI V). SSR149415 is orally active and inhibits some butn ot alla cute stressor-induced ACTH release in rats Ramose ta l. 2006),a nd does nota ffect HPA hypersensitivityt on ovel stressors ). The compound acts at the human Avpr1b and also has some antagonist activity at the human Oxtr in vitro (Griffante et al. 2005) but has high selectivitya nd nanomolar affinity for rodent forms of the Avpr1b (Serradeil-LeG al et al.2 002). In micea nd rats, SSR149415 has been tested in av ariety of classical models of anxiety (e.g. elevated plus maze, light/dark box test) and depression (e.g. forced swim test, chronic mild stress) as well as otherm odels (e.g. olfactorybulbectomy,Flinder'ssensitiveline) that are used to determine the efficacyo fp otential antidepressant and anxiolytic drugs Overstreet and Griebel 2005;Stemmelin et al. 2005;Louis et al. 2006;Salomé et al. 2006;Shimazakietal. 2006;Hodgson et al. 2007;Iijima and Chaki 2007). Peripheraland central pretreatmentwith SSR149415 reduces anxiety andd epressive-related behaviour in theset ests with high compatibility between the findings of these studies. SSR149415 also reduces aggressioni nh amsters( Blancharde ta l. 2005), significantly reversest he reduction in dentate gyrus cell proliferation caused by chronic mild stressinmice (Alonsoe ta l. 2004) andb lockss tress-induced hyperalgesia in rats (Bradesi et al. 2009).I th as also been radiolabelled with tritium and used in receptor autoradiography to reveal low-resolution binding in the human and rat pituitary-no Avpr1b binding sites were observedi ns ections of rat brain (Serradeil-Le Gal et al. 2007).R ecently,S SR149415 has failed phase II clinical trials (Kirchhoff et al. 2009). Overall, the resultso fs tudies with SSR149415 evidence a possible role for the Avpr1binaffective disordersand point to animal model-validated targets with which to treat them. Around thes amet ime, SSR149415w as first reporteda sa nA vpr1b antagonist, the Avpp eptido mimetic[ 1-deamino-4-cyclohexylalanine]a rginine vasopressin (d[Cha 4 ]Avp) was described (Derick et al. 2002).T his agonist was the first to show efficacya t nanomolar concentrations and stimulate the release of ACTH/CORTw ithout exhibiting vascular or renal activity (Derick et al. 2002). Other peptide agonists selective for the rat Avpr1b haveb een generated by modifying positions 4a nd 8o ft he Avpa nalogue deamino-[Cys]arginine vasopressin (Penae ta l. 2007a).M any of thesem odified Avpa nalogues display high selectivity for the Avpr1b andb ind with sub-nanomolar affinities (Pena et al. 2007a) and thus could well be useful in the study of the Avpr eceptors in rodents; however, they may be of limited use as humant herapeutics due to their peptidergic nature. The agonists created since d[Cha 4 ]Avp do, however, haveanincreasingly refined agonist profile. One such memberofthe recent modified range of Avpanalogue agonists, d[Leu 4 ,Lys 8 ]Avp, is noted to be afull agonist at human, rat and mouse Avpr1bs in vitro as well as stimulates ACTH and insulin release at low doses from mouse pituitarya nd perfused rat pancreas, respectively ( Pena et al.2 007b). Thise ffecto f d[Leu 4 ,Lys 8 ]Avp on mouse and rat tissue is blocked when co-administered with SSR149415 (Pena et al. 2007b).
In conclusion, thed istribution andf unctional studieso ft he Avpr1b have established its major role in the pituitaryw here it plays ap ivotal parti nt he regulation of the HPAr esponse to stress, and in particular to acutestress. Additionally,wesee several, perhaps minor, metabolic ande ndocrine roles in the periphery. Behavioural changes generated from experiments in Avpr1bK Oa nimals,t ogether with recent Avpr1ba ntagonist data, have highlighted am ore elusive central role for this receptor. The behavioural implications of Avp, acting viat he Avpr1b,i n aggression and stress, andt he integral connection between stress, anxiety and depression make the Avpr1ba na ttractivet argetf or pharmacological intervention. Increased Avps ecretion and enhanced pituitaryresponsiveness to Avphavebeen reportedin some subtypeso fd epression (e.g.m elancholic depression)(see, Dinan andS cott 2005 for ar eview). Furthermore, polymorphisms in the Avpr1bgene have been associated with depression ( van We st et al. 2004), childhood-onset mood disorder (Dempster et al. 2007) and attention-deficit hyperactivity disorder( van We st et al.2 009).T he progressm adei ng enerating compoundss electivef or this receptorm ay have considerable implications for potential treatments for anumberofdiseasestates as well as for Avpresearch in general. The development of new compounds that can be radiolabelled to high specific activity is ac ritical step in Avpr1b research as determination of its central distribution may well provide an anatomical template to assign how changes in behaviourso rd iseases tates are influenced. Further developments in molecular (e.g. conditional Avpr1b KOs; the use of Avpr1bspecific small-interfering RNAst os electively silence Avpr1bgene expression in specific brain regions) and pharmacological tools for use in rodents and primates (e.g. positron emissiontopography (PET) ligands, see Schonberger et al. 2010) will help elucidate the full function of the Avpr1b andthus the therapeutic value researchinto this receptor may hold.

Noticeofc orrection
Please note that the following amendments have been made to this paper, first published Online Early 9 September 2010: the removal of ad uplication of text within Ta ble Icaptiononpage 6; in Ta ble IV on page 12, the heading 'Anxiety related behaviour' has been moved up to the 2nd line in columno ne from its previous position at line 12; and as pelling error, incorrect year of publication and incorrect volume details have been amended in the Zhou et al. 2008 reference list and citationsinT ables III and IV.