Somatostatin Receptors

S

Receptors

Yogesh C. Patel and Coimbatore B. Srikant

The diverse biological effects of somatostatin (SRIF) are mediated by a family of G protein-coupled receptors (termed sst) that are encoded by five nonallelic genes located on separate chromosomes. The receptors can be further divided into two subfamilies: sst2,3j5react with octapeptide and hexapeptide SRIF analogues and belong to one subclass; sst1j4 react poorly with these compounds and fall into another subclass. This review focuses on the molecular pharmacology and function of these receptors, with particular emphasis on the ligand-binding domain, subtype-selective analogues, agonist-dependent receptor regulation and desensitization responses, subtype-specific efector coupling, and signal t~ansductionpathways responsible for inhibiting cell secretion and cell growth or induction of apoptosis. (Trends Endocrinol Metab 1997;8: 398–405). @ 1998, Ekevier Science Inc.

●

The SomatostatinReceptor Family

Somatostatin (SRIF) is synthesized as the bioactive peptides, SRIF-14 and SRIF-28, which act on multiple targets, including the brain, gut, pituitary, endocrine and exocrine pancreas, adrenals, thyroid, kidneys, and immune cells (Reichlin 1983). The actions of SRIF include inhibition of virtually every known endocrine and exocrine secretion; motor, senso~, behavioral, cognitive and autonomic effects; as well as effects on intestinal motility, vascular contractility, cell proliferation, and intestinal absorption of nutrients and ions. These pleiotropic effects can be resolved into three cellular processes that are modulated by SRIF— neurotransmission, secretion, and cell proliferation—and are mediated via high-affinity plasma membrane receptors termed sst receptors (Patel et al. 1995, Reisine and Bell 1995, Lamberts et al. 1996, Patel 1997). Beginning in 1992,

Yogesh C. Patel and Coimbatore B. .%ikant are at the Fraser Laboratories, McGill University,Departmentsof Medicine and Neurology and Neurosurge~, Royal Victoria Hospital and the Montreal Neurological Institute, Montreal, Quebec H3A IA1, Canada.

398

the struclure of SSLreceptors was elucidated by molecular cloning (Yamada et al. 1992a). Five individual subtypes were rapidly identified and shown to consist of a family of heptahelical G protein– coupled receptors (GPCR) (Yamada et al. 1992a and b, Bruno et al. 1992, O’Carroll et al. 1992, Rohrer et al. 1993, Panetta et al. 1994, Patel et al. 1995, Reisine and Bell 1995, Patel 1997). Human sst receptors (hsst) are encoded by a family of five nonallelic genes located on separate chromosomes (Table 1). Four of the genes are intronless, the exception being sstz, which gives rise to spliced variants sst2A and sstz~, which differ only in the length of the cytoplasmic C-tail (Figure 1) (Patel et al. 1995, Reisine and Bell 1995). There are thus six putative sst subtypes of closely related size, each displaying a seven transmembrane domain (TM) topology. All sst isoforms that have been cloned so far from humans as well as from other species possess a highly conserved sequence motif YANSCANP1/VLY in the VIIth TM, which serves as a signature sequence for this receptor family (Figure 1) (Patel et al. 1995, Reisine and Bell 1995). Overall, there is 3$ V0–57% sequence identity among the various members of this fam-

ily, with sstl and ssta showing the highest sequence identity. The individual subtypes display a remarkable degree of structural conservation across species. Thus there is 94%–98% sequence identity between the human, rat, and mouse isoforms of sstl; 93%-96% sequence

o

identity between human, rat, mouse, porcine, and bovine isoforms of sstz; and 88v0 sequence identity between the rat and human isoforms of sstq; sst~ and sst~ are somewhat less conserved, showing 82Y~83% sequence identity between the human and rodent homologies. The nearest relatives of the sst receptors are the opioid receptors, whose s subtype displays 37% sequence similarity to mouse sstl.

●

Binding Affinity of Natural and SyntheticSomatostatinLigands

Over the years, many different analogues of SRIF have been synthesized for investigational and clinical use (Figure 2). Structure-function studies have shown that amino acid residues Phe7, Trp8, Lys9, and Thrl O,which comprise a ~ turn, are necessary for biological activity, with residues Trp8 and Lysy being crucial. The general strategy for designing SRIF analogues has been to retain the Phe7, Tkp8, Lys9, and Thrl” segment and to incorporate a variety of cyclic and bicyclic restraints to stabilize the P turn around the conserved residues. In this way, a library of short synthetic compounds has been synthesized, several of which show greater metabolic stability and some pharmacological selectivity compared with SRIF-14 (Patel and Srikant 1994, Patel et al. 1995, Reisine and Bell 1995, Bruns et al. 1996, Shimon et al. 1997). All five sst subtypes bind SRIF-14 and SRIF-28 with high affinity (Table 2). sst,-. bind SRIF-14 ~ SRIF28, whereas sst~ exhibits weak selectivity for SRIF-28 compared with SRIF-14. The octapeptide analogues SMS201 -995 (SMS, Octreotide) and BIM23014 (Lanreotide) that are in clinical use, as well as the octapeptide RC160 (Vapreotide) and the hexapeptide MK678 (Seglitide), bind to only three of the hsst subtypes, displaying high affinity for subtypes 2 and 5 and moderate affinity for type 3 (Table 2). Based on structural similarity and reactivity for octapeptide and hexapeptide SRIF analogues, the receptor family can be divided into two

01998, ElsevierScience Inc., 1043-2760/98/$17.00 PIIS1043-276O(97)OO168-9

TEM vol. 8, No. 10, 1997

Table 1. Characteristics of the cloned subtypes of human somatostatin rece~torsa

Chrornosomal localization Amino acids mRNA (kb) G-protein coupling Ef’lector coupling Adenvl@ cyclasc acl i\)ity Tyrosinc phosphatase activitv Ca2 + channels Na ‘ /H exchanget’ Phospbolipasc C/IPl act ivil}j Phospholipzrse A2 act ivitv MAP kinase activitv Tissue cfislt”ibution’)

Sst,

sst2A

sst3

14ql 3

17q24

22q13. 1

391 4.8 4

369 8.5 (?) +

418 5.0 +

sst5 16pl 3.3 363 4.0 +

38>’ 4.( I 4

J

‘r

‘T

‘r

‘r

T

J/’r ‘r

Brain, pituitary, stomach, liver, pancreas, kidneys

Brain, pituitary,, stomach, pancreas, kidneys

subclasses: SS12, 3,5react with these anaIogues and conslittrte members of one SLIbgIOLIp; sst,,4 react poorly with these con]pounds and fall into another sLlk~rOLlp, ‘1’he analogL1c! Dcs-AA ,,2,5[D-TrPx [AMP”] SRI F (CH275) has been t-ecently rcportecl as an sst, selective conlpoLlnd (Liapakis et al. 1996). In our hands, CH275 also binds to sstq and appears to be a prototypic agonist for the sst,,4 subc]ass (Table 2) (Pate] 1997). Sevcwl other SRIF analogues have been sim iIarly reported to be selective [“or one sst SLlbtVpC, f’01’ exzLmpk, sst2 (MK678), sstq (BIM23056), and sst5 (BIM23052, L362855) (Reisine and Bell 1995). Because of’ howc\e]-, lllethc)cic)lc)gic:~l variations, such claims of subtype se]ectivit~ of tb~se and other analogucs have not bum substantiated by others and shoolci be intetprctcd \vith caution (Patel and Srikant 1994, BrLInS et al. 1996). More recent binding analyses ol” these LX)npoLInds using the human sst clones have ident ilied only BI M23268 with modest selectivity for hssti (Table 2). L362855 binds \\wll to sst~ and sst2 and is ordv w,cakl} selective for- sst~. Likmvise, MK678 displays good binding affini[v

TEM ltd. 8, N(). 10, 1997

sst4

$ Brai 11, pitw tary, ston ]ach

‘T BI-., 11, st( )i ]ach, pa I I :1’eas, lull.

s,

st( I:

Iach

JBt-ain, pituitary, stomach

for both sstz and ssI subtypes. OveIall then, thes(. results SI1!,gcstthat the binding selec ivity 01 , [Irrently available SRIF arralogues fot’ I)e human sst sLlblypes k relative ratht. than absolute and agonists available lhat thel-e arc no pLI for the i,]dividual .Llbtypes. Very reccntlv, tht first pot~ \Iial sst peptidc am lagonists Ilave bee]) described (Bass et al. 1996, wilkinsoi ~ et al. 1997). One such co npound Ac-4-N02-Phe-c(DCVS-TVI--DTrp-Lys-l ! i-Cvs)-D-Tyr-NHJ binds to hsstz and h~ [, with nM affinity, but antap{)nizcs r-c ptor effcctor cx)upling to ~derrylyl . clase (Bass et al. 1996). A second pep de, BIM23056, appears to he an anta}’ )nist at the sst~ receptor (~’ilkinson ii il. 1997).

●

Ligand-Bindingl}omain

The ligan Li-biriding ite of’ peptidc agm nists comparable t, SRIF typically in\olves residues iJi the ext raceilu lat. loops (EC Ls) or bo I the ECLS and the TMs (Sch\vartz an,l Rosenkilde 1996). Bv exploi[ ing the d 1 ferential abilitv of SMS to bind to hs: ~ but not to hss[,, Kaupmal et a]. ( I ’95) Systernaticzdly

mulated hsst, to resemble hsstz. They found two crucial residues, Glnz.,l and Ser~05, in TMs VI and VII, respectively, o(’ hsst, , substitution of which for the corresponcli ng residues Asn276 an d Phe 2’)4 in bsstz increased the affi nitv of hsstl f’o[ SMS and other octapcptidc analogues 1000-fold (Figure 1). Based on these results, Kaupman et al. ( 1995) have postulated a binding ca\ilv for SMS involving hydrophobic and ch:irgecf residues located exclusively within TMs I II–Vi 1.Their findings predict that the core resiclues Phe7, TrpX, Lvs<’, and Thr of SMS interact \~J ith Asn 276 and Phe2’”4 located at the outer end of TMs VI and VI 1, rcspecl ively (prcscnl in SS12but not in sstl ), which provide a hydrophobic environment for lipophilic interactions with Pbc7, TIp8, Thr ”), and Asp’ in TM 111, \ubich anchors the ligarrd bv an electrostatic inter-action with LVS’J(Firgurc 1) (Nchring et :LI. 1995). SMS binds poorlj to hsst] because of tbe presence of rcs idues Gln2c} and SCI-~05located close to the extracclluht” rims of TM helices V1 and VII which prevent lhe short pept icle horn reaching deep \vithin the pockc[, wheteas the c-orrcspondi ng residues Asn276 and Phe2”4 in SSL2 provide for a stable interaction \vith the disulfidc bridge of’ SMS. Because of their greater length and I’lcxibility, the natur-al Iigands SRIF- 14 and SRIF-28 can presumably adopt a conf’ormat ion that allo\vs their entry in[o the binding pockc[ of’ all five sst receptors. The involvement of the extracellular domains for’ binding SRIF 1igands has been investigated by Greenwood et al. ( 1977), who used amino terminal delelion mutants OI- conservative segment exchange m utagcnesis I“ot’ the three ECLS of hssti. Tlreil rcsd[s predicl a potential contribution of” ECL2 (but not of ECL1, ECL3, or the amino term inal segment) to binding of” the natural SRIF Iigands (SRI. F- 14, SRIF28), as well as SMS. The overall model that emel’ges from lhese studies sugges~s a binding domain Ior SRIF Iigands made LIp of’ residues wi[hin TMs Ill–VII, \vith a potential contribution b} ECL2, and is consistent with other peptide-b ind ing GPCRS, for ex ample, rreurokini n I, angiotcrlsi n II, GnRH receptors, which interact with in both ECLS, and TMs residues (Sch\\artz and Rosenkilde

C 1998,Else, icr Science IIIC.,1043-270 98/$17.00 PU S1043-2760(97)001 68.9

1996).

399

Fe-r

Oamoq=== “

CHO

Figure1. Schematic depiction of the seven-transmembrane topology of the human sstl-5 receptors. CHO are the potential sites for N-linked glycosylation wilhin the amino terminal segment and second extracelkdar loops (ECL); P04 are the putative sites for phosphorylation by protein kinase A, protein kinase C, and casein kinase. The cysteinc residue 12 amino acids downstream from the VIIth TM is conserved in sst,,z,~,sand may be the site of a Potential palmitoyl membrane anchor. The YANSCAN PIiVLY sequence in [he VIIth TM is highly 294in TMs III, VI, consemed in all members of the sst family. Residues ASP’22,Asn276,and Phe and VII, respectively, of sst2Ahave been proposed to form part of a Iigand-binding pocket for Octreotide and are shown by the closed circles (From Patel 1997).

●

G-Protein Coupling and Signal Transduction

Ligand activation of endogenous sst receptors is associated with a reduction in intracellular cAMP and Ca2+ and stimulation of protein phosphatases because of receptor activation of four major effecter pathways, each involving a pertussis toxin–sensitive GTP-binding protein: (a) adenylyl cyclase, (b) K+ channels, (c) Ca2+ channels, and (d) protein phosphatases (Figure 3) (Patel et al. 1995, Reisine and Bell 1995, Lamberts et al. 1996,

400

Patel 1997). Receptor activation inhibits adenylyl cyclase, leading to a fall in intracellular cAMP. Ss1 receptors are coupled to several subsets 01K+ channels (delayed rectifier, inward rectifier, ATP-sensitive K+ channels, and large conductance Ca2+-activated BK channels). Receptor activation of K+ channels causes reversible hyperpolarization of the membrane, leading to the cessation of spontaneous action potential activity and secondary reduction in intracellular Ca2+i because of inhibition of the normal depolarization induced Ca2+ infhmvia voltage sensitive Ca2+ channels. In

addition to this indirect effect on Ca2+ entry sst receptors act directly on highvoltage-dependent Ca2+ channels to block Ca2+ currents. Stimulation of both K+ and Ca+ channels may also occur through dephosphorylation of the channel proteins secondag to sst activation of a serine threonine phosphatase. Furthermore, sst receptors may inhibit Caz+ currents through induction of cGMP, which activates cGMP protein kinase with further phosphorylation-dependent inhibition of Ca2+ channels. Sst receptors activate a number of phosphatases such as serine threonine phosphatases (White et al. 2+-dependent Phosphatase 1991), the Ca calcineurin (Renstrom et al. 1996), and protein tyrosine phosphatases (PTP) (Florio et al. 1994 and 1996, Buscail et al. 1995, Reardon et al. 1996). This action is dependenton activation of pertussis toxin–sensitive G proteins, but the nature of the G proteins involved and whether

01998, ElsevierScience Inc., 1043-276W98/$17.IICI PII S1043-276Q(97)OOI 68-9

TEM Vol. 8,

No.

10, 1997

~~lF-28

Ser-Ala-Asn-Ser-Asn-Pro-Ala-Met-Ala-Pro-Arg togen-acti\ated proi in kinase (MAPK) ~lu-Arg-LysAloGly Cys-LysAsn-Phe- Phe.Trp in CHO-K 1 cells, I> ~th via a pcrtussis dysSeEThhPhe-Thr/LYS

which mcmbrane ion channels are short circuitccf) through a distal effect, which is toxiwscn~itivc G 1~otein (Bito et al. bclicvcd to be lmediated via a G-proteill1994, Pat~l et al. 195, Reisine and Bell depcndcnt mechanism linking the recepAla-Gly-Cys-LysAsn-Phe-P he. TIP 1995). Ac[ ing throu:.: i sstq and sst~, howSRIF-14 tor to exocytotic vcsiclcs (Figure 3). Smch Cys-SerThr-PheThr ,LYS ever, SRI F also inh i ~its the MAPK sigdirect inhibition of cxocytosis is induced DPhe-Cys-Phe.DTrp naling cascade (R, l[don et al. 1996, through SRIF-ciependent activation of the SMS201-995 lhr(ol)-Cys-Thr/ Ly$ Cordclicr ct al. 199’7 Sst modulation of protein phosphatasc calcincurin (Renocireofide the PLC-I F’; path\va\ remains con[ro\elstrom et al. 1996). The profound ability of DPNal-:ys-Tyr.DTrp sial (Patel 1997). In t ansfected CHO-KI sst anci other inhibitmy receptors, for exBIM23014 Thr-Cys-Val/IYS brreotide cells, sstc inhibits 1P,~-lmcdiated C:(2+ ample, wadrenergic anci galanin rcccpmobil iza[ on, \vher(, s sst~ is without eftors, to bk)ck secretion via this distal DPhe-~yslyr.Dfip RC-160 I“ect.By a ]ntt-ast,in [ OS-7 cells, sst~ and mechanism suggests that phosphoryl~lvapreoiide Trp-Cys-val-lys sst2~ botl [ sli rnulat, IP; production, alevents rather tiotl-dephosphotylation beit only :It high ag ~list concentrations. MK678 than the Ca2”+signal play a key role in the (NMe)-Ala~yr.Dfip seglitide Major vo~ds still r( !lain in oLu” undel-- distal steps 01 exoqtosis. The specific sst Pke-Val’t Ys standing f)f sst subt) w-selectivity f’br-ion subtypes involved in this process remain Figure2. Natural and swtheticpeptic{e :ig(k channel c oupling a J of [be molecular to be cictcrminect.‘rhe :trltiproli[ktati~e el”nistsol’the SSI rcccp(orf”arn ilf. signals in the r-ccc]~ )rs responsible l’c)r- l“ects01”SRIF arc mediated both inc{irectly activatior of var. )LIS phosphatases. through inhibition of hormones and Much of ( ~urcurrel 1~ lkno\\]cdgcon sllbgrowth factors that promote cell growth, they couple sst receptors directly or inditvpe-selectivity for . ~naling is basecl on recdv to pbosphatascx is unknown. and clirectly via sst rcccptors prcwmt on transfecttd cells al ~I should be intetScneral laboratories have investigated tar-gel cells, leading to growth anw+t and preted wvth cautio given the limita[be coupling of individually expressed induction of apoptosis (Srikant 1995, Pations of tlese systet it . The cmcrgencc of’ te] 1997). Scvcra] sst subtypes ancl signal sst sLlbty”pes to G proteins and vari(ms select ive agonists and arr[agonists effcctors. Despite initial controversy, transduction path\\ays have been itmpliI [te the study of subshould gn ally facil i there is now a growing consensus that catcd. A SRIF-sensitive 66-kD SH2 dolype-selec[ ive effccl II coupling 01’ enall fi~’e SSLsubtvpes, and certainly the main containing PTP called PTP-f C, clogenous sst recepl t’s in normal cells. hulman isoforms, arc hnctionally couwhich clephospholylates and inactivates p]cd to inhibitiorr of adcnylvl cyclase growth factor rcccptor kinases, has been (Table 1) (Pate] et al. 1994). sst, ,2,3,4also shown to translocate from the cytosol to ● Signal TransductionPathwaysfor stimulate PTP (Table 1). sst, stimulates a the plasma membrane upon receptor acInhibitionof Sc*retion and Cell Na ‘ /H ‘ evcchangel via a per[ussis toxill– tivation ancl to associate with sst receplors Proliferation insensitive mechanism (HOLI et al. 1994). (Zeggari et al. 1994, Srikant and Shen Blockade (~f seer-cti(, by SRIF is in part Sstz s~lpprcsses \mltagc-cicpencfenl Ca2 ‘ 1996). SRIF also inhibits MAPK activit?r mediated lhrough i} 1 ~ibition of Ca2 ‘ and channels in islet RIN M5f cells and Nvia PTP-dependent inactivation of Ral’-l cAMP. Adclitionally; 11l\vever,SRIF can ir~- or-lhmugh inhibition of gLmnylate cyck~sc and P/Q-type Ca2”’ channels in cultured hibit horlnonc SCC,~[ion stimulated by t-atamygdaloici neurons (Viana and H ille (Rcarclon e[ al. 1996, Cordclier e[ al. cAMP, Ca ~+, and (,; )cr second messen1996). sstd activates PLA2-dependcmt 1997). PTP-dependent ~llltipl-olifelatioll by arach idcrnatc production as well as m i- gers, as \\ell as in p meabilized CCIIS(in SRIF in\olvcs both cytostatic anci cytotoxic (apoptosis) ac[ions and is ciosc-dependent and subtvpe-selective. Apoptosis occurs only in cycling cells, uniquely via IC50 (nM)” the sstl sLIblype, and is associated with induction of wild-type p53 and Bax (Shar-sst5 sst4 sst2A sst3 Sstl ma et al. 1996). The remaining foLu sst subtypes elicit a c~rtostaticI“esponse by im 0.2-0.9 0.3- 1.> 0.3- 1.6 0.2-1.3 0,1-2.26 SRIF-14 ducing G, cell cycle arrest associated with 0.05-0.4 0.3-7. ” 0.3-6.1 0.2-4.1 0.1-2.2 SR[F-28 activation of the retinoblastoma gene > 100( I 5.6-32 4.4-34,5 0.4-2.1 290-1140 Octrcoticfe product pRb and the cyclin-dependent ki66–2 1[! I 0.6-14 43-107 0.5-1.8 500-2330 Lanrcoticfe nase inhibitor. p21. Whereas apoptosis is 0.7 31 45 5.4 > 1000 Vapreotide tri.ggereclat low agonist concentt-ation (~ 127: 1000 2-23 27-36 > 1000 0.1-1.5 Seglitide 0.1 nM), cvtostasis is induced a[ much 1.2–7.3 16-14 I 2.1-5.6 1o–1 3.5 BIM 23052 6.3-100 higher (= 50 nM) agonist concentrations. 17-234 5.7-14.1 BIM 23056 110– > 1000 132– > 1000 10.8–177 Overall, this means that acting via sst~, 16.3 0.37 15.1 61.6 18.4 BIM 23268 SRIF \\’ill inc{uce apoptosis even when 63- ~- 000 0.1-0.016 1 6.2 > 1000 L 362855 >1000 present at low phwiological concentra>1000 4.3–874 >1000 3.2-4.3 CH 275” tions, whereas it will attenoate the nlito“Basecl 011Patcl anc] Srikant ( 1994), Reisinc and Bell ( I Y95), Bmns c1 al, ( 19%} Ind Shimon et al. ( 1997). genic sign:ll and trigger growth arrest via “C):II:I (II’ PLIte]and Sr-ikat][ ( 1994) Cxpi-esseclw Ki. sst, ,,, z ~ ~ only at pharmacological concen‘ FI-{Im Liapakis et al. (1996) and Pa[el ( 1997).

TEAI vol. 8, N() 10, 1997

Else, icr Science I IIC., (“)19LJ8,

1043-2711

98/$1 7.00 I’1I S1043-2760(97)00 168-9

401

A

Sst

‘i

trations. The molecular signal in sst~ that confers subtype selectivity for apoptosis as well as the downstream pathways for both the cytostatic and cytotoxic actions of SRIF remains to be unraveled but could involve phosphorylation-dependent modulation of p53, Raf-1, MAPK, and other substrates implicated in regulating cell proliferation.

●

/

~

1

GROWTH

In view of the many biological actions of SRIF, an important question that arises is whether these effects are subtype-selective or whether multiple subtypes are involved in mediating a given response. The lack of

Bax~

MAPK~

APOPTOSIS

INHIBITION

Sst

B

ADENYLYL CYCLASE K’/ Ca’+ CHANNELS

00 (7*-

‘vExocytOsi~ —— ——-

I INHIBITION OF SECRETION

I

Figure3. Schematic depiction of the key second-messenger systems involved in somatostatin (SRIF) modulation of cell secretion, cell proliferation, and apoptosis. (A) Receptor activation leads to a fall in intracellular cAMP (because of inhibition of adenylyl cvclase), a fall in Ca2+ influx (because of activation of K+ and Ca2+ ion channels) and stimulation of phosphatases such as calcineurin (which inhibits exocytosis), and serine threonine phosphatases (which dephosphorylate and activate Ca2+ and K+ channel proteins). Blockade of secretion by SRIF is in part mediated through inhibition of Ca2+ and cAMP (proximal effec[) and through a more potent distal effect involving direct inhibition of exocytosis via SRIF-dependent activation of calcineurin. (B) Induction of protein tvrosine phosphatase by SRIF plays a key role in mediating the antiproliferative response by dephosphorylating growth factor receptor kinases and other putative substrates such as P53.Raf-1 and mitogen-activated protein kinase (MAPK) implicated in regulating cell proliferation and apoptosis. 402

Subtype-SelectiveBiological Responses

suitable subtype-selective SRIF antagonists has so far proved to be a major impediment in elucidating subtype-specific tissue effects. Based on the pattern of expression of sst receptors in individual target cells as determined by double-label immunocytochemistry or in situ hybridization, there is growing evidence that single cells express several sst subtypes in varying densities (Day et al. 1995, O’Carroll and Krempels 1995, Kumar et al. 1997). For instance, rat somatotrophs feature all five sst isoforms, with sst~ being the preponderant species (Day et al. 1995, Kumar et al. 1997). On the basis of the differential effects of various SRIF agonists on pituitary islet, and intestinal responses, several studies have suggested sstz selectivity for inhibition of GH, glucagon, and gastric acid secretion in the rat; sst~selectivity for insulin inhibition in the rat; and sst~selectivity for gastric smooth muscle contractility in the guinea pig (Rossowski and Coy 1994, Wyatt et al. 1996, Gu et al. 1995). Unfortunately, because of questions concerning the subtype monoselectivity of the SRIF analogues used in these studies, these findings remain inconclusive (Patel and Srikant 1994, Bruns et al. 1996). Using a slightly different approach, Shimon et al. (1997) have proposed the mediation of both sst2and sst5 in regulating GH and TSH secretion from the human fetal pituitary. A sst2-deficient knockout mouse has been recently generated by Zheng et al. (1997) in which sst, receptors in arcuate GHRH neurons were found to be solely responsible for mediating GH-induced feedback inhibition of GH. Nonetheless, the animals grew normally and appeared healthy up to 15 months of age, excluding an essential role for sstz in embryogenesis

PII S1043-276O(97)OO168-9 01998, ElsevierScience Inc., 1043-2760/98/$17.00

TEM

Vol.

8,

No.

10, 1997

postnatal growth and Cfevcloplmen[.Although [L~rtherwork is requilccf with this model, the findings so far do not support a prel”erentia] ro]e of’ the Ss[z subtype in Lhe pi[uital~ or other peripheral targets in the rodent.

or

receptor regulatory Iilnction or to an alteration in (he patleri i and composition of the variolls subtvpe~ Jxpresscd or because of abnor] nal rccep[t, signaling. Agonist-specific cl,sensitization is conlmon to )nany GP(. 1
Agonist-DependentRegulation and DesensitizationResponses

of

Although the acute adrninistratiot] SR[F-14 ot-SMS produces a diverse range 01”biological effects, the initial response diminishes \\ith corrtinuccf exposLu’c to because of’ the dcwelopment the peptides of [olerancc (Lalmberts ct al. 1996). P:itients ~\ith SRtF-producing tumors display sustainccf hypersoma[ostatinemia, \\, bich, howcvu-, causes minimal s\lmptomatok)gy; notably’ mild steatorrhea, diabetes mellitus, and cholelithiasis secondaly to inbibilion of pancreatic exocx-ine secretion, insulin release, and ga]lbladdcrcontraction. Long-term lherapy with SMS is remarkably I“r-eeof side effects (Lan~berts et al. 1996). Most patients delelop tmilcf steatonhea to which they become toletant after 10–14 days. Likewise, the? adapt rapidly to some of the other ef’f’ects of SMS, for example, inhibition of insulin and thereby develop and TSH secl-dion, minima] signs of carbohydrate intolcl-ance or hvpoth,yroidisrn. Some cf’fects, however, do pcx-sisl,for example, inhibition of gallbladder emptying, which giwx rise to a significant increase in the incidence of cholesterol gallstones (Lalmbcrts ct al. 1996). What is most interesting is that hormone-prochlcing lumors such as GH adcmornas, carcinoicfs, and VJPomas continue to rcsporrd to SMS injections \\ith persistent suppression or hormone secretion, h-equently [“or several ?cars. This s~lggests a differential regulaliorr of sst r-eceptor-sin normal tissues and in kmors. Agonist-mediated receptor downrcgukrtion could explain [he desensitization responses of insulin, TSH, and pancreatic cxocrine secretion. The SS[ receptors involved in modulating biliary hlrrction are clearly different, because thev do not appear to clounregulatc with continued SRIF [reatmcnt. The ability o(” bor-mom-secreting tumors to withstand desensitization mav be due to several factors. Thrnot’s express a high density of’ sst receptors compared with surr-otrrlding normal tissues (Lambcrts et al. 1996). Conceivably, sst receptors in tumors behave diflkrentlv owing to a loss of nor-real

TEM Vol. 8, N<). 10, 1997

During the 6 years since the Iirst sst reccptot” WZLScloned, gr-eat progress has been made towarcl characterizing the stmcture and molecular’ pharmacology of this re-

ceptor I“amily, which now has fiw members. The rich pattm-nof expression of sst pling 01 [ I]e recepto from G proteins, rereceptors throughout the br-ainand ill peceptor in[ernalizati< ~, and receptor degripheral tissues, coupled with the potent radation. Shor-tet i I exposure 10 SRIF biological effects that they elicit, clcarlv has been shown to ncfLIce G-protein unsuggests that SS1r-eceptorsreprcscmta macoupling and sst r-(,cptor intcrmdizalion jor class o(’ inhibitory receptors that pla,v in pituita[y and isl~l cells. Prolonged ag(]an important role in moduklting higher nist exp( ,sure (24–4 ~ h) upregulates sst brain kmction, the secretoly process, cell receptors in GH4C’, and RIN M5f cells proliferation, and apoptosis. Synthetic ag(Presky and Schonl ‘~lnn 1988). BeC2LLlS~ onisls (ofSRIF have been in clinical use for over 10 years ancl now occupy an impol-S or their normal Ilituital~ a I ~.1 islet CC1l Lumol-CCIIderivati\ C>express multiple sst tant therapeutic niche both in the ciiagnosis and tr-eatment of tumor-s. These are, subtypes, it has bc~ ] necessary to sludv host cells stably tra sfectcd with each sst boweve[; first-gener’ation compounds that interact with only three of the five SS1sobsubtvpe ,n order 1( characterize Lhe retypes. The development of other more sesponse ~)f indivi(l ~al SSTR isotypes. lective agonists, as well as subtype-specific hsst~,~,d,.undergo 1. i~idinternalization in antagonists, ~hCjL1ldgreatly expand the a time- aild tempel. ~:ure-dependcmt nw~nscope of SRI F pharmacotherapy. FLltLlrc ner over 60 min i I I CHO-K1 cells upon studies will need to define the function of agonist activation i I [ukovic c1 al. 1996). incfiviclual subtypes, the role of multiple Maximul n intcI”na]i/Ition occurs with Sst+ subtypes in the same cell, the downstream (78%), followed by-~>t, (66%), sstl (29%), signaling pathways responsible for growth and sstz (20%). In . ,ntrast, hsstl fails to arrest and apoptosis, the molecular basis be interrlalizcd. Pt ttonged agonist treatof reccpt(m upregulation, and the nwch:~ment fol 22 h upt gulates hsst, at the nisms underlying cliffercmtial descnsitiz&~membrane by 11()( , hsst2 and hsst.l by lion responses in tumor cells compareci 2696 and 22[%, ) spcctivelv, whereas with normal cells. Finally, a great cieal et al. hsst~,s SI I(.)w no c1 I inge (HLlkovic’ needs to bc learned aboui the biology of 1996). Ayonist-prol t)ted receptor upregsst receptor civsfunction in neurological, with receptor- ph{ ISphorylation,

●

ConcludingRemarks

●

uncou-

ulation a Iso occLm= .vith other rcccptors, for example, those 1 r GnRH and cfopamine 2, aml involves II c sut-facerecruitmcm[ ol’ preexisting pook )[”receptors (Ng ct al. 1997). Tile llnderl) ng molecular signals remain (() be idcr1I~fied. Desensitization and intel nalization ~1’sst2Aand sst, receptors ar-e i~ssociated \ith phosphotyiation of cytoplasmic rcsit ~les,especially in the carboxvl terminal + gmcmt (Hipkin et al. 1997,Rot h et al. 19’) 7).The ability of SRIF to reguklte its rec( ~tors may provide a mechanism for ta1ieting selective SLIb trees f’01 chagnosi. and therapy For instance, \lpregulati( )

●

and immunological

Acknowledgments

The work cited from the authors’ labor& toty was supported by grants from the Canaclian Medical Research Council (MT10411, MT-601 1, MT-12603), the National Jnstitllles of Health (NS32160), the U.S. Department of Defense, and the National Cancer Jnstitute of Canada. The expcrl secretarial help of Maria Cor-mia and Tracy Wilson is grate[”ully acknow]edgeci.

of sstl and sstz by

appropria~e agoni.1 tr-catment

coLdc~

be

used fol enhancitn. sst receptor expression km !’eceptor’ s. 11s..% btvpes such as sst< ancl sst5, whicl are extensively internalizeci, ~ould be tit[
or (km

gastroerrterological, disorcicrs.

References 13ass RT, tluckwalter J3L,el al,: 19%. Idcntificalion and ctlzll~ictel-ixati(lll0(’ novel somzrtostatin antagonists.Mol Phatmmcol50: 709-715. Bito H, Mori M, Sakanaka C, ct al.: 1994. Functional coupling of SSTR4, a major hip-

[) 1W8, Elstwicl-Scieno Inc., 1043-2 098/$’17.00 Plr S1043-2760(97)001 68-{)

403

pocampal somatostatin receptor, to adenylate cyclase inhibition, arachidonate release and activation of the mitogen-activated protein kinase cascade. J Biol Chem 269:12,722-12,730. Bruno JF, Xu Y, Song J, Berelowitz M: 1992. Molecular cloning and functional expression of a brain-specific somatostatin receptor. Proc Natl Acad Sci USA 89:11,15 l– 11,155. Bruns C, Raulf F, Hoyer D, Schloos J, Lubbert H, Weckbecker G: 1996. Binding properties of somatostatin receptor subtypes. Metabolism 45( Suppl 1):17–20. Buscail L, Esteve J-P, Saint-Laurent N, et al.: 1995. Inhibition of cell proliferation by the somatostatin analogue RC-160 is mediated by somatostatin receptor subtypes SSTR2 and SSTR5 through different mechanisms. Proc Natl Acad Sci USA 92:1580–1584. Cordelier P, Esteve J-P, Bousquet C, et al.: 1997. Characterization of the antiproliferative signal mediated by the somatoslatin receptor subtype sst~. Proc Natl Acad Sci USA 94:9343-9348. Day R, Dong W, Panetta R, Kraicer J, Greenwood MT, Patel YC: 1995. Expression of mRNA for somatostatin receptor (sstr) types 2 and 5 in individual rat pituitary cells: a double labeling in situ hybridization analysis. Endocrinology 136:5232–5235. Florio T, Rim C, Hershberger RE, Loda M, Stork PJ: 1994. The somatostatin receptor SSTR1 is coupled to phosphotyrosine phosphatase activity in CHO-K1 cells. Mo] Endocrinol 8:1289-1297. Florio T, Scarziello A, Fattore M, et al.: 1996. Somatostatin inhibits PC C13 thyroid cell proliferation through the modulation of phosphotyrosine phosphatase activity-impairment of the somatostatinergic effects by stable expression of EIA viral oncogene. J Biol Chem 271:6129-6136. Greenwood MT, Hukovic N, Kumar U, et al.: 1997. Ligand binding pocket of the human somatostatin receptor 5 (hSSTR5): notational analysis of the extracellular domains. Mol Pharmacol 52:807-814. Gu ZF, Corleto VD, Mantey SA, et al.: 1995. Somatostatin receptor subtype 3 mediates the inhibitory action of somatostatin on gastric smooth muscle cells. Am J physiol 268:G739-G748. Hipkin RW, Friedman J, Clark RB, et al.: 1997. Agonist-induced desensitization, internalization and phosphorylation of the sst2A somatostatin receptor. J Biol Chem 27:13,869-13,876. Hou C, Gilbert RL, Barber DL: 1994. Subtypespecific signaling mechanisms of somatostatin receptors SSTR1 and SSTR2. J Biol Chem 269:10,357-10,362. Hukovic N, Panetta R, Kumar U, Patel YC: 1996. Agonist-dependent regulation of

404

cloned human somatostatin receptor types 1–5 (hSSTR1–5): subtype selective internalization or upregulation. Endocrinology 137:4046-4049.

Patel YC, Greenwood MT, Panetta R, Demchyshyn LL, Niznik HB, Srikant CB: 1995. The somatostatin receptor family. Life Sci 57:1249-1265.

Kaupman K, Bruns C, Raulf F, Weber HP, Mattes H, Lubbert H: 1995, Two amino acids, located in transmembrane domains VI and VH, determine the selectivity of the peptide agonist SMS201-995 for the SSTR2 somatostatin receptor. EMBO J 14:727– 735.

Presky DH, Schonbrunn A: 1988. Somatostatin pretreatment increases the number of somatostatin receptors in GH4C1 pituita~ cells and does not reduce cellular responsiveness to somatostatin. J Biol Chem 263: 714-721.

Kumar U, Laird Dd, Srikant CB, Escher E, Patel YC: 1997. Expression of the five somatostatin receptor (SSTRI-5) subtypes in rat pituitary somatotrophes: quantitative analysis by double-label immunofluorescence confocal microscopy. Endocrinology 138:4473-4476. Lamberts SWJ, Van Der Lely AJ, de Herder WW: 1996. Drug therapy: octreotide. N Engl J Med 334:246-254.

Reardon DB, Wood SL, Brautigan DL, Bell GI, Dent P, Sturgill TW: 1996. Activation of a protein tyrosine phosphate and inactivation of Raf-1 by somatostatin. Biochem J 314:401-404. Reichlin S: 1983. Somatostatin. N Engl J Med 309:1495-1501, 1556-1563. Reisine T, Bell GI: 1995. Molecular biology of somatostatin receptors. Endocr Rev 16: 427442.

Liapakis G, Haeger C, Rivier J, Reisine T: 1996. Development of’ a selective agonist at the somatostatin receptor subtype SSTR1. J Pharmacol Exp Ther 276:1089-1094.

Renstrom E, Ding WG, Bokvist K, Rorsman P: 1996. Neurotransmitter-induced inhibition of exocytosis in insulin-secreting (3 cells by activation of calcineurin. Neuron 17:513-522.

Nehring RB, Meyerhof W, Richter D: 1995. Aspartic acid residue 124 in the third transmembrane domain of the somatostatin receptor subtype 3 is essential for somatostatin-14 binding. DNA Cell Biol 14:939-944.

Rohrer L, Raulf F, Bruns C, Buettner R, Hofstaedter F, Schule R: 1993. Cloning and characterization of a fourth human somatostatin receptor. Proc Natl Acad Sci USA 90:4196-4200.

Ng GYK, Varghese G, Chung HT, et al.: 1997. Resistance of the dopamine D2Lreceptor to desensitization accompanies the upregulation of receptors onto the surface of Sf9 cells. Endocrinology 138:4199–4206.

Rossowski WJ, Coy DH: 1994. Specific inhibition of rat pancreatic insulin or glucagon release by receptor-selective somatostatin analogs. Biochem Biophys Res Commun 205:341-346.

O’Carroll AM, Krempels K: 1995. Widespread distribution of somatostatin receptor messenger ribonucleic acids in rat pituitary. Endocrinology 136:5224-5227.

Roth A, Kreienkamp H-J, Meyerhof W, et al.: 1997. Phosphorylation of four amino acid residues in the carboxyl terminus of the rat somatostatin receptor subtype 3 is crucial for its desensitization and internalization. J Biol Chem 38:23,769–23,774.

O’Carroll AM, Lolait SJ, Konig M, Mahan LC: 1992. Molecular cloning and expression of a pituitary somatostatin receptor with preferential affinity for somatostatin-28. Mol Pharmacol 42:939-946. Panetta R, Greenwood MT, Warszynska A, et al.: 1994. Molecular cloning, functional characterization, and chromosomal localization of a human somatostatin receptor (somatostatin receptor type 5) with preferential affinity for somatostatin-28. Mol Pharmacol 45:417-427. Patel YC: 1997. Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 20:348–367.

Schwartz TW, Rosenkilde MM: 1996. Is there a lock for all agonist keys in 7 TM receptors? Trends Pharmacol Sci 17:213–216. Sharma K, Patel YC, Srikant CB: 1996. Subtype selective induction of p53-dependent apoptosis but not bell cycle arrest by human somatostatin receptor 3. Mol Endocrinol 10:1688–1696. Shimon 1, Taylor JE, Dong JZ, et al.: 1997. Somatostatin receptor subtype specificity in human fetal pituitary culture. J Clin Invest 99:789–798.

Patel YC, Srikant CB: 1994. Subtype selectivity of peptide analogs for all five cloned human somatostalin receptors (hSSTRl5). Endocrinology 135:2814-2817.

Srikant CB: 1995. Cell cycle dependent induction of apoptosis by somatostatin analog SMS201-995 in AtT-20 mouse pituitary tumor cells. Biochem Biophys Res Commun 209:400-407.

Patel YC, Greenwood MT, Warszynska A, Panetta R, Srikant CB: 1994. All five cloned human somatostaiin receptors (hSSTR1-5) are functionally coupled to adenylyl cyclase. Biochem Biophys Res Commun 198: 605-612.

Srikant CB, Shen SH: 1996. Octapeptide somatostatin analog SMS 201-995 induces translocation of intracellular PTPIC to membranes in MCF-7 human breast cancer adenocarcinoma cells. Endocrinology 137: 3461-3468.

01998, ElsevierScienceInc., 1043-2760/98/$17.00 PIIS1043-276O(97)OO168-9

TEM

vol.

8,

No.

10, 1997

B:1996. MoLILII:lI ionut high lfjllage-;ictill:~[ccl c:ilci~]llqcllzillnels b\s~JInalosta[io inc~cL1(el\is(.)l:iteci l-alarmj$cfaJNc~lr;sci 16:6000 -60 11. oidnctlmns.

Viana

F, Hillc

White RE, %honbrunn A, Atrnstrong DL: I991. Soma(ostatin stimulates Ca2 -ac[iw(cd K ‘ channels lhrough pru[ein clephus pho]-yla[ioll. Nature 35 I :570-573. Wilkinson

CF, Fcniuk

W,

19Y7. Ch~~r:tcteriz:iti(]tl nant sonlat(m(atin

PPA:

}ILInIphI-Cy

of human

reconlbi-

sst5 t-ccc ptors mediating

~i falnll,,

of’phosphuinrxi tide metabolism Br J Pha[-mace] 121:91–96.

W\att MA, Jalxie E, Fcniuk W, Humphrey PPA: 1996. Somatostatin Sstz reccplol-tmediatd inbibitiun of parietal cell I’unctic)u in rat isolated gas[ric- mucmsa. Br J Pharmxo] 1 I9:905–9 I(). Yamada Y, Post SR, Wang K, et a].: 1992a. Cloning and l“unctional cllzil-z~ctcriz~ititln0{

:, j ICIrrroLlse somat(~s(atin

exprcssc(l II hrain, gastl.oi ntesti nal tr-ao, and kidt I, ~. Proc Nzrtl Acad Sci USA 89:?5 1–255.

rcccptrr)s

), Rcisine 1 i.L1\V SF, et al.: 1992b. Sorna(wtatin recelj IIS, an expanding Rcne family cloning art [’functional character.iz,ation t,f human S5 [’R3, a prolcin couplccl to adeI]vlatc cyLI, SC. Mcrl f3ndocrinol 6:2 136-.? 142.

Yamada

Zeggari

acti\a(ion

“f tlumzin

h’i, Esteve J 1‘ R:ILIly 1, e[ al.: 1994.

Cupurif’cation 0[”: phatase \vithacti\:, tors [’II,tm r-a[ p:, branes. E+irxl]cm J Zheuy H, Bailey A, .1 Srrmato. fatin rece[l mice aI c refractO negativ, feedback ~ Endocr Im)l 11: 170

>lotein lyrosine phosd somatostatin rccclI.reatic acinal. men)3:441-448. (IIg M-H, et al.: 1997. ~rsubtype 2 knock-out to growth hormone :U.CLIZite nLW1-Oll S. MO] 1717. TEM

Wimlandet 1994). A family of snRNAs interac[ by base pairing with sequence information defining the 5‘ and 3‘ splice sites of an cxon. The consensus scquenccs for these splice sites are loose. Tbere are many examples of constitw ti\ely LIscd splice sites that deviate considerably from the consensus, whereas sequences that match the consensus may be present on a pre-mRNA and yet never used lot-splicing. When alternative splicing occLIrs, the splicing apparatus must either use a splice site not genetally recognized or ignore a splice site that is usually used. Thus, the regulation of altelnativc splicing is often discussed in tetms of splice site selec[ion. The terms “s( rong” and “weak” splice sites are often applied to sites that match or deviate from the conscrrsus, respectively. In alternative splicing, different conlbinations of exons can be derived Irolm a pre-mRNA (for examples, see Figures 3, 4, and 5). There are numerous genes \\,ith~1] (elnzltive splicing patterns, and

AlternativeSplicing of mRNA as a Mode of Endocrine Regulation Shern L. Chew

Alternative splicing ofmessenger RNA (nzRAA) is a n) ‘an.%ofregula[ing geH6’eXpt’eSSiOll and occurs in nlany genes ol”the end( ) iine system. T/liS review

covers

an

introduction

into

mRNA splicing,:.. and the meclla -

fli.sms and regulation of alterfza[ive .splicin,q.So??le xamples are di.scus.seclin which alternativelyspliced genes encode fi[,!ctionally distinct proteins. Evidence that hormones and other metal dic signals may regula[e nism.s

alternative are con.sidwed.

splicing

events

(Trends

is revie~ veal, an[~ potential

Endocrinol

Metab

mecha-

I )$)7; s:405-4

13). o

1998, Ekevier Science tnc.

Split genes were cfiscxwercxi in 1977 (Em-get et al. 1977, ChovV et al. 1977, lefi-e>s and Flavell 1977). Exons are separated bv introns in the genom ic DNA, as \vell as in the pt-imary RNA transct-ipi (prc-mRNA) (Figut-e 1). Most eukal-votic genes contain irr[rons, and [hese musl be remcwed to gi\e a mature

Shern L. Chc\\, is at the Deparlmen( of’ End(.)ct’inology, St B~\l-t}l(~](>nlc\i’s H(.rspit2Ll, LoIIdon F.C IA 7BE, United Kingdom.

7’EM Vol. 8, N(). 10, 1997

mRNA [ ]anscript i r transport from the nucleus Io the Cvt{ IIasm, where transktion inn) protein ccurs. Splicing is a very ac~ urate pbt I ]menon, with exons en if separated by precisel) joined, many’ kilobases 1“ intmn (Haukins 1988). This precisi n is achieved by premRNA sequence ii I’onmation (Figure 2) and nuclear fact( 1,>. The nuclcal--splicing app:iratus is c:i CLIthe “splicemx)mc” and is a cornplcx ~rlade up of srna}} nLl clear RhAs (snRR S) and many proteins (Moore et al. !993, Bauren and

this is an important and common mechanism f’ot- generating protein diversity and regulating gene expression in higher cukaryotes (Smith et al. 1989). The regulation ot” altcn-native splicing events by developmental and tissue-specific lac[ors is well recognized, and in some cases, the molecular mechanisms have been c[ucidated. There are severa] examalternative splicing being ples of changed by hormonal or metabolic signals (Table 1). This is a relatively new area of research, and the mechanisms are nol urtdcrstood. The general strategy used for dissecting the mechanisms of” alternative splicing is w map the sequence elements on the premRNA through which regukltion is nledialed. In higher eukalyotes, this is by constrnc[ing and transecting a “minigene” into cell lines. The minigene usually contains [be upstream and downstream constitutivc exons that flank the alternatively spliced cxons, together with all or part of the introns. The minigenc is then mutated to idcntih key sequence elements, which, in turn, may allow purification of factors interacting with those sequence elements. Some examples oi” sequence elements (other than splice site sequences) shown to enhance or inhibil alternative splice site usage are listed in Table 2. In some cases, the proleins interacting with enhancers and inhibitors have been identified.

() 1998, Elsmic] Scicnc, [nc., 1043-2 >0/98/$17.00 PII S104.3-2760(97)001 67-6

405