当前位置：文档库 › 2003 Neuregulins functions, forms, and signaling strategies

2003 Neuregulins functions, forms, and signaling strategies

Neuregulins:functions,forms,and signaling strategies

Douglas L.Falls*

Center for Neurodegenerative Disease,Department of Neurology,Emory University,Atlanta,GA 30322,USA

Received 13December 2002,revised version received 17December 2002

Abstract

The neuregulins (NRGs)are cell-cell signaling proteins that are ligands for receptor tyrosine kinases of the ErbB family.The neuregulin family of genes has four members:NRG1,NRG2,NRG3,and NRG4.Relatively little is known about the biological functions of the NRG2,3,and 4proteins,and they are considered in this review only brie?y.The NRG1proteins play essential roles in the nervous system,heart,and breast.There is also evidence for involvement of NRG signaling in the development and function of several other organ systems,and in human disease,including the pathogenesis of schizophrenia and breast cancer.There are many NRG1isoforms,raising the question “Why so many neuregulins?”Study of mice with targeted mutations (“knockout mice”)has demonstrated that isoforms differing in their N-terminal region or in their epidermal growth factor (EGF)-like domain differ in their in vivo functions.These differences in function might arise because of differences in expression pattern or might re?ect differences in intrinsic biological characteristics.While differences in expression pattern certainly contribute to the observed differences in in vivo functions,there are also marked differences in intrinsic characteristics that may tailor isoforms for speci?c signaling requirements,a theme that will be emphasized in this review.?2003Elsevier Science (USA).All rights reserved.

Keywords:Neuregulin;Acetylcholine receptor-inducing activity;Glial growth factor;Heregulin;Neu differentiation factor;Sensory and motor neuron-derived factor;ErbB receptor tyrosine kinase;Schizophrenia;Neuromuscular synapse;Cell-cell signaling proteins;Juxtacrine signaling;Paracrine signaling;Transmembrane ligands;Proteolytic process;Shredding

The discovery of neuregulins (NRGs);NRG family genes;the focus of this review

Neuregulins (NRGs)are signaling proteins that mediate cell-cell interactions in the nervous system,heart,breast,and other organ systems.“Forward”signaling by NRGs—i.e.,signaling from a NRG-producing cell to a NRG-respon-sive cell—involves binding of NRG to the extracellular domain of the receptor tyrosine kinases ErbB3or ErbB4,which leads to formation of ErbB homo-or heterodimers (often including ErbB2),which in turn activates intracellu-lar signaling pathways leading to cellular responses that include stimulation or inhibition of proliferation,apoptosis (programmed cell death),migration,differentiation,and ad-hesion [1].

The ?rst identi?cations of NRGs were reported in 1992–

1993by four groups.Two of these groups sought a ligand for the oncogene ErbB2(a.k.a.neu,HER2)[2–4];the third sought a factor that stimulated the proliferation of Schwann cells [5,6],and the fourth sought a factor that stimulated the synthesis by muscle of receptors for acetylcholine,the ma-jor neurotransmitter at developing neuromuscular synapses [7].The neuregulin proteins isolated by each of these groups are encoded by the gene that would now be referred to as NRG1.It should be noted that though one approach leading to the identi?cation of NRGs was a search for ErbB2li-gands,in fact,it appears that NRG proteins interact with ErbB2only after binding ErbB3or ErbB4[1].

Subsequent to the identi?cation of the NRG1gene,three other genes encoding related proteins were discovered.These “other”NRGs are referred to as NRG2(a.k.a.Don-1,NTAK [8–11]),NRG3[12],and NRG4[13].The NRG1proteins effectively bind to both ErbB3and ErbB4;the protein products of these other NRG genes effectively bind one or the other or both of these ErbBs [1,14,15].Very little is yet known about the functions of the NRG2,3,and 4

*Rollins Research Building,Rm.2105,1510Clifton Road,Emory University,Atlanta,GA 30322,USA.Fax:?1-404-727-2880.E-mail address:dfalls@https://www.wendangku.net/doc/5d18999359.html, (D.L.

Falls).

Available online at https://www.wendangku.net/doc/5d18999359.html,

Experimental Cell Research 284(2003)14–https://www.wendangku.net/doc/5d18999359.html,/locate/yexcr

proteins.One intriguing recent ?nding is evidence of NRG4involvement in the differentiation of the somatostatin-ex-pressing delta cells of pancreatic islets of islets [16].Buo-nanno and Fischbach [17]compare the structure and expres-sion patterns of NRG2,3,and 4proteins to NRG1proteins and discuss the intriguing discovery that NRG2activation of a speci ?c ErbB receptor combination (i.e.,ErbB4ho-modimers)can elicit different patterns of receptor phos-phorylation and downstream consequences than activation of the same receptor combination in the same cell type by NRG1[18,19].

This review will focus on the NRG1proteins,and unless explicitly indicated otherwise,the terms “neuregulin ”and “NRG ”here refer to NRG1proteins.Since the discovery of NRGs 10years ago,the ?eld has grown rapidly.A search of Pub Med for “neuregulin ”or various names by which these proteins have also been known (see below)in early Decem-ber 2002returned a list of over 800publications.This rapid growth in the literature attests to the importance of NRGs,but also means that that this review cannot be in any sense comprehensive.Thus,I apologize in advance for the many important papers not referenced and for interesting areas of NRG research omitted or mentioned only in passing.A number of excellent reviews of NRGs and NRG-mediated cell-cell interactions have appeared over the years.Selected reviews published in the last 6years include those focused speci ?cally on NRGs [17,20–25],as well as reviews of neuromuscular synapse development [26–28],neuron-glial interactions [24,29–33]and cell interactions regulating heart development and function [169].Companion reviews in this issue discuss the NRG receptors ErbB2,ErbB3,and ErbB4and intracellular signaling pathways activated by these receptors (see reviews in this issue by Yarden,Car-penter,and Wiley;see also [1,17]).Other companion re-views in this issue describe the roles of ErbB family recep-tors and ligands in breast development,in cancer,and as therapeutic targets (reviews by Stern,Hynes,Arteaga,Fry,and Maihle).

While I will begin with an overview of NRG biology,the

availability of these other reviews allows me to emphasize recent literature and to focus on the normal biological func-tions of the NRG1proteins during development and in the adult,evidence for the involvement of NRG1signaling in neuropathology (other than cancer),and mechanisms regu-lating NRG signal production.Readers of the companion reviews on the EGFR and its ligands (see reviews by Coffey and Burgess)and the ErbB receptor/ligand homologues in invertebrates (see reviews by Shilo and Sternberg)will note both marked similarities and striking differences between the biology of NRG1proteins and these related signaling systems of vertebrates and invertebrates.

The NRG1gene;NRG1isoforms and nomenclature An important recent advance is the sequencing and as-sembly of the entire human NRG1gene (Fig.1A,[34]).The gene is ?1.4megabases long (?1/2000th of the genome);less than 0.3%of this span encodes protein.As a conse-quence of rich alternative splicing and multiple promoters,at least 15different NRG isoforms are produced from the single NRG1gene (Fig.1B and C [17,20]).The three struc-tural characteristics we know to importantly differentiate isoforms with respect to in vivo functions and cell biolog-ical properties are the type of EGF-like domain (?or ?),the N-terminal sequence (type I,II,or III),and whether the isoform is initially synthesized as a transmembrane or non-membrane protein;the import of these differences will be discussed below.The EGF-like domain contained in all bioactive NRG isoforms is alone suf ?cient for activation of ErbB receptor-tyrosine kinases (see [17]for comparison of NRG1EGF-like domain sequences to the EGF-like domain in NRG2,3,and 4and other EGF family members).To-gether the types I and II NRGs are sometimes referred to as “Ig-NRGs,”and the type III NRGs are sometimes referred to as “CRD-NRGs.”The names ?rst used in the literature to refer to various NRG isoforms —acetylcholine receptor-in-ducing activity (ARIA [7]),glial growth factor (GGF [5,6]),

Fig.1.3NRG1gene and isoform structure.(A)Human NRG1gene structure (Genbank accession no.BK000383).The NRG1gene is on the short arm of chromosome 8.On the expanded illustration of this region,the position of each exon included in reported NRG1isoforms is indicated by a vertical line.Lines descending along the edge of the green box delineate the boundaries of the core at-risk haplotype for schizophrenia.Only the exon encoding the type II-speci ?c N-terminal region lies within these bounds.Exon naming:Exons are named here for the structural region of the NRG1protein they encode.Abbreviations used closely correspond to names of NRG protein structural regions indicated in panel B.EGFc refers to the exon encoding the portion of the EGF-like domain sequence shared by NRGs with an ?-type and NRGs with a ?-type EGF-like domain.The exon labeled TMc also includes adjacent extracellular juxtamembrane sequence and cytoplasmic tail sequence.(B)Illustration of NRG “coding segments.”Isoforms differ in their coding segment composition due to initiation of transcription from different NRG1gene promoters and alternative splicing.The EGF-like domain alone is suf ?cient for high potency activation of the cognate ErbB receptor tyrosine kinases.Available evidence indicates that the NRGs most commonly expressed in the nervous system are transmembrane NRGs with a ?-type EGF-like domain and the 374-amino acid a-type tail.Not all potential combinations of coding segments have been reported.CRD ?cysteine-rich domain;EGF ?epidermal growth factor-like domain;Ig ?immunoglobulin-like domain;CTc and TMc ?cytoplasmic tail and TM domain C-terminal of the EGF-like domain;CTn and TMn ?cytoplasmic tail and TM domain N-terminal of the EGF-like domain.Only type III NRGs have the CTn and TMn.*?stop codon.(C)Structural regions of the I-?1a,II-?1a,III-?1a,and III-?3proproteins.The I-?1a,II-?1a,and III-?1a isoforms differ only in their N-terminal region;their sequence is identical from the EGF-like domain through the carboxy-terminus.The sequence of III-?1a and III-?2a is identical except that in III-?2a the last eight amino acids of the “1”subtype are absent (illustrated under III-?1a sequence).The sequence of III-?1a and III-?3is identical from the N-terminus to just beyond the ?nal cysteine of the EGF-like domain.Hydrophobic regions that serve as transmembrane domains are indicated by black boxes.A hydrophobic sequence in the type II N-terminal region is indicated by a hatched box.This is believed to serve as a noncleaved internal signal sequence but not as a transmembrane domain.

D.L.Falls /Experimental Cell Research 284(2003)14–30

16 D.L.Falls/Experimental Cell Research284(2003)14–30

17 D.L.Falls/Experimental Cell Research284(2003)14–30

heregulin(HRG[2]),neu differentiation factor(NDF[3,4]), and sensory and motor neuron-derived factor(SMDF[35]),—cannot be taken to indicate speci?c biological functions of the isoforms to which these names have been applied.For example,it now seems likely that the major NRG isoforms that act as“glial growth factors”in vivo are type III NRGs, not the type II NRGs originally called glial growth factor (GGF).

NRG1signaling in health;isoforms differing in their N-terminal region or EGF-like domain differ in their in vivo functions

Without NRGs life is not possible.However,even I—a con?rmed neuregulin fanatic—was surprised as I surveyed the literature in preparation for writing this review,at how pervasive NRG signaling appears to be(Table1).While I will here introduce the in vivo functions of NRGs by de-scribing the dramatic phenotypes of the knockouts,it must be emphasized that a number of other experimental ap-proaches have made important contributions to our current understanding of NRG functions,and it seems likely that many NRG functions remain to be discovered.

Studies of mice with targeted mutations of the NRG1 gene have been very valuable in elucidating functions of NRG1proteins[34,36–47].Analysis of pan-NRG1knock-out(KO)mice(mice in which all NRG isoforms are unable to bind to and activate ErbB receptors due to disruption of the EGF-like domain)revealed an essential role of NRGs in cardiac morphogenesis[36],a role unsuspected from pre-vious studies of NRG bioactivities.Due to the defect in cardiogenesis,these mice die midway through embryogen-esis(E10.5),the time at which mouse embryos switch from dependence on the maternal circulation to dependence on their own circulation.The pan-NRG1KO mice also have a severe reduction in several neural crest-derived cell popu-lations including Schwann cells,the glia of the peripheral nervous system,which—among other things—form the my-elin sheaths of peripheral nerves;neural crest-derived cra-nial sensory neurons;and sympathetic neurons[39,40]. Having a knockout with a severe phenotype is both a boon and a bane:a boon because it provides comforting reassur-ance that the gene of interest has essential roles and a bane because death or early developmental disruptions in the mutants preclude analysis of later developmental events. Such is the case with analysis of nervous system develop-ment in the pan-NRG1knockout mice:at E10.5the devel-opment of the nervous system is only beginning to unfold; and the role of NRGs in processes such as neuromuscular synaptogenesis,which begins around E14,and the devel-opment of oligodendrocytes,the cells that myelinate axons in the central nervous system,is inaccessible for direct in vivo examination in these mice.In some cases,clever strat-egies can allow this limitation to be partially circumvented. Thus,through analysis of spinal cord slices harvested from E9.5pan-NRG1embryos and maintained in organ culture for up to11days,a strong case has been made for an essential role of NRGs in oligodendrocyte lineage develop-ment[48].

Mice with targeted mutations that inactivate only certain classes of NRG isoforms have revealed differential in vivo functions of NRG proteins(Table2).Ig-NRG1?/?mice: Mice with all Ig-NRG isoforms inactivated(types I and II NRGs inactivated)—but in which CRD-NRG(Type III NRG)production is presumably normal—die at E10.5and have defects in cardiac,cranial sensory neuron,and sym-pathetic development similar to those of the pan-NRG knockouts[37,39,40].However,unlike the pan-NRG KOs, the Ig-NRG KO mice have normal development of Schwann cell precursors[39].CRD-NRG1?/?mice:In con-trast to the pan-and Ig-NRG KO mice,mice with all CRD-NRG isoforms(type III NRGs)inactivated—but with normal expression of Ig-NRGs—do not have defects in heart development[42].The CRD-NRG KO embryos sur-vive to birth.They die at birth because they cannot breathe;

Fig.2.Membrane orientation and proteolytic processing proposed for the NRG I-?1a isoform(A)and III-?1a isoform(B)based on studies in transfected ?broblastic cell lines.(A)Proteolytic cleavage of the I-?1a proprotein in the“stalk”region(arrow no.1)produces an N-terminal fragment(NTF)containing the bioactive EGF-like domain and a C-terminal fragment(CTF),also referred to as the“a-tail remnant.”The NTF is ef?ciently released into the medium. The protease(s)catalyzing stalk cleavage—at least in?broblasts—is likely to be a metalloprotease.All Type I and Type II proproteins with a transmembrane domain are expected to have similar topology and processing(see Fig.3).(B)Available evidence indicates that the III-?1a proprotein has two transmembrane domains and that proteolytic cleavage in the stalk region(arrow no.1)produces a transmembrane N-terminal fragment(NTF

)that accumulates at the cell surface.Cleavage of the III-?1a NTF m near the membrane(arrow no.2)can release a fragment(NTF s)containing the EGF-like domain into the medium,

but for NRGs expressed in?broblasts,the amount of this released Type III NTF

s is very small compared to the amount of released Type I NTF

.All Type

III proproteins with a transmembrane domain C-terminal of the EGF-like domain are expected to have topology and processing similar to that illustrated here for III-?1a(see Fig.3).

Fig.3.Proposed topology and mode of presentation to receptor for selected NRG1proprotein isoforms.The data and reasoning supporting the assigned topology and mode of presentation(paracrine,shed;paracrine,secreted;juxtacrine)for the types I,II,and III isoforms are described in the text.The proteins encoded by two other mRNA sequences are illustrated to facilitate their comparison with the full-length NRGs shown.HRG-?is a truncated Type I sequence. It is unlikely to be bioactive for two reasons:?rst,the EGF-like domain,which is necessary for activating ErbBs,is incomplete;and second,since it is a truncated version of I-?3,it is unlikely to be released.The rightmost diagram illustrates a protein named?-HRG( HRG-?).This protein is encoded by transcripts produced by the breast cancer cell line MDA-MB-175[166].It has now been shown that?-HRG is a fusion protein with N-terminal sequence from the human homologue of transcription factor DOC-4[167].This transcription factor is a member of the Oz/ten M family.On the basis of sequence similarity to the N-terminal portion of?-HRG,some sequences encoding Ten-m/Odz family members have been erroneously labeled as NRGs or“NRG-like.”There is as yet no evidence for a physiological role of the HRG-?sequences or?-HRG[168],and they are not further discussed in the text.

18 D.L.Falls/Experimental Cell Research284(2003)14–30

they cannot breathe because they do not have functional neuromuscular synapses.Unlike the Ig-NRG knockouts,the CRD-NRG knockouts have a marked reduction in Schwann cell precursors.Other prominent phenotypic characteristics of the CRD-NRG KO mice include degeneration of periph-eral and cranial nerves and an ?50%reduction in the number of spinal motor and sensory neurons.The reduction in motor and sensory neuron number appears to be due to abnormal neuron death,as the initial number of these (post-mitotic)neurons is normal.Like the spinal motor and sen-sory neurons,in the CRD-NRG1KO mice,cranial motor and sensory neurons (which together contribute most of the axons that make up the cranial nerves)appear to be reduced in number.However,unlike the Ig-NRG1KO mice,both placode-and neural crest-derived cranial sensory neurons are affected.In the Ig-NRG1KOs,the reduction in neural crest-derived cranial sensory neurons is due to a defect in initial accumulation of these neurons in the nascent ganglia (perhaps,like the defect in sympathetic neuron accumula-tion,caused by abnormal neuronal precursor migration?),but in the CRD-NRG1KO mice it seems likely that the reduction in cranial sensory neurons results from abnormal death of neurons,and,as for spinal sensory neurons,this increased death is probably a consequence of disruption in signaling between these neurons and their supporting Schwann cells or targets.NRG1?/?mice:Mice with a targeted mutation that inactivates all NRG isoforms with an ?-type EGF-like domain (NRG ?s)—but in which produc-tion of isoforms with ?-type EGF-like domain (NRG ?s)is presumably normal —have not been reported to have abnor-

Table 1

Selected proposed functions of NRGs a Organ/cell type/structure

Effect

Reference(s)

Nervous system Schwann cells Survival,proliferation,migration,differentiation,myelination [24,29,31–33]Oligodendrocytes

Proliferation,survival,differentiation,myelination

[48,129–134,153]Neuromuscular synapse

Nerve –muscle interaction controlling initial formation,acetylcholine receptor synthesis during development and in the adult,nerve terminal interactions with “terminal Schwann cells ”

[17,20,26–28,146]

Muscle spindle

Muscle spindle development (Muscle spindles are muscle length/stretch sensors.)

[172]

Cranial sensory neurons

Initial population of cranial sensory ganglia (ganglia of cranial nerves)with neural-crest derived sensory neurons (However,the initial population of cranial sensory ganglia with placode-derived sensory neurons appears to be unaffected by NRG1mutations.)[36–39,44]

Motor and sensory neurons Survival (spinal and probably also cranial)

[42]

Peripheral and cranial nerves Fasciculation (bundling)of axons and/or integrity of nerves

[36–39,42,44]Sympathetic

neurons/adrenal medulla Migration of sympathetic neuron/adrenal chromaf ?n precursors to the anlage of sympathetic ganglia/adrenal medullas

[40]Cerebellum

Production of cerebellar neuron precursors

[38]Cortical neuron precursors/cerebellar granule cells Migration of CNS neuronal precursors along radial glia

[71,72]Hypothalamus Hypothalamic control of mammalian female sexual maturation [135,173]Parasympathetic Enteric ganglia development

[136,174,175]Hippocampus

Inhibition of long-term potentiation (LTP)induction (LTP is a model for studying the neurophysiological basis of learning and memory.)

[97]

Various neurons of CNS and PNS Regulation of neuronal neurotransmitter receptors (NMDA,GABA,neuronal nicotinic acetylcholine receptors)and other neuronal ion channels

[64,74–76,176]

Heart Development of ventricular wall trabeculae,AV-septum,and cardiac valves [36–38,44,137]Heart Development of cardiac conduction system

[138]

Heart

Growth,repair,survival of adult cardiomyocytes;response to increased work load [89–91,139]Blood vessels Angiogenesis

[144]

Breast Breast development during pregnancy and lactation

[45];Review in this issue by D.Stern Lung Development of pulmonary epithelium (autocrine effect?)[142]Muscle Myogenesis (autocrine effect?)

[140]Muscle Muscle ?ber survival in neonatal period [141]Muscle Glucose uptake [144]Gonads Gonadogenesis

[145,177]Stomach

Proliferation of gastric epithelium;regulation of parietal and chief cell population size

[178,179]

This list is not intended to be comprehensive,and further investigation will be required to con ?rm the physiological signi ?cance of many of the proposed roles.Some proposed functions have been inferred principally from the effects of exogenously supplied recombinant NRG1,ErbB blockade,or ErbB knock-out;in these cases the physiological signal may actually be NRG2,3,or 4or another ErbB ligand.Caveat emptor.Functions proposed solely on the basis of mRNA/protein expression data are not included.Reviews have been cited for proposed functions of NRG1s in Schwann cell and neuromuscular synapse development due to the large body of relevant literature.

D.L.Falls /Experimental Cell Research 284(2003)14–30

malities in nervous system or cardiac development,but have marked defects in breast development [45].(Aside:In most assays,NRG1s with a ?-type EGF-like domain are 10–100times more potent than NRG1s with an ?-type EGF-like domain.It is a puzzle as to why NRG1s with both ?-and ?-type EGF-like domains exist and why the ?-type was selected by evolution for a critical role in breast develop-ment.)

Comparison of the phenotypes of the ErbB knockouts [34,38,40,41,43,45,49–57]to the NRG1knockouts has pro-vided insight into the ErbB receptor combinations mediat-ing early essential actions of NRGs.Furthermore,most all reported characteristics of the various NRG1knockouts are shared by one or more of the ErbB knockouts and vice versa,which suggests that interactions of NRG2,3,and 4with ErbBs are unlikely to play a prominent role in the developmental events that dominate the ErbB KO pheno-types;rather these developmental events are likely to be mediated principally by NRG1-ErbB interactions.There is one clear exception to this generalization;ErbB4knockout mice have a defect in hindbrain segmentation not seen in the NRG1knockout mice [49],indicating that a non-NRG1ligand interacts with ErbB4to guide this developmental event.

One further note,many of the phenotypic characteristics of the NRG1and ErbB KO mice center on abnormalities of ErbB expressing cell populations,suggesting that these de-fects result from disruption of forward signaling (NRG-producing cell signaling to ErbB-expressing cell).However,it has been proposed that NRG-ErbB signaling is bidirec-tional [58–60],similar to what has been demonstrated for

Eph-Ephrin signaling (see [61]and references therein).In this model,it is proposed that there is not only the conven-tional “forward signaling ”from NRG-producing cells to ErbB-expressing cells,but that there is also reverse signal-ing (or “back-signaling ”)from ErbB-expressing cells to NRG-expressing cells.In the latter case,NRG would serve as the receptor and ErbB the ligand.Some phenotypic char-acteristics of the NRG1and ErbB1knockouts involve NRG-producing cells,raising the possibility that these de-fects result from disruption of NRG 3ErbB reverse sig-naling.For example,motor neurons,which produce NRGs and which communicate with ErbB expressing Schwann and skeletal muscle cells,are reduced in number in the CRD-NRG KO mice compared to wild-type mice.This might be due to interruption of ErbB 3NRG reverse signaling,though alternatively,this abnormality could also result from interruption of forward signaling secondarily disrupting a separate “reciprocal ”signaling pathway back to motor neurons from Schwann cells and/or muscle.While the hypothesis of NRG1-ErbB reverse signaling is very attractive,de ?nitive evidence that this occurs has yet to be published.

NRG1signaling in disease:evidence for involvement in pathophysiology and potential therapeutic uses The many functions of neuregulins revealed through knockout and other studies (Tables 1and 2)attest to the importance of neuregulin signaling during development and in the adult.Are disorders of neuregulin signaling involved

Table 2

Comparison of NRG1knockout mice with respect to selected characteristics Development of:

Genotype and isoforms INactivated NRG1?/?All

(Pan-NRG1KO)

Ig-NRG1?/?Ig-NRG1

(Type I and II)CRD-NRG1?/?CRD-NRG1(Type III)NRG1??/?NRG1?

NRG1?CT/?CT Type I

NRG1(and others?)**Heart

f f

Schwann cell precursors f

f ??(Not described)

Neuromuscular

synapses

Breast (during

pregnancy)

Homozygotes die at E10.5

E10.5

Birth

Normal lifespan E10.5

Cause of death:Heart failure Heart failure Respiratory (neuromuscular)failure Old age Heart failure References

[36,38,39]

[37]

[42]

[45]

[44]

The major neurotransmitter receptor at neuromuscular synapses is the muscle nicotinic acetylcholine receptor.Although neuromuscular synapse development (which begins around E14)cannot be assessed in the Ig-NRG KOs,adult Ig-NRG ?/?mice (i.e.,heterozygous for the mutation inactivating Ig-NRGs)have a 50%reduction in the number of acetylcholine receptors at neuromuscular synapses [47],indicating that Ig-NRGs do function as an “acetylcholine receptor-inducing activity ”(ARIA).This “postsynaptic phenotype ”—i.e.,the effect on AChRs,which are concentrated in the postsynaptic muscle membrane —is different than the “presynaptic ”and “Schwann cell phenotype ”of the CRD-NRG KO mice (for review of neuromuscular synapse development and NRG functions at the neuromuscular synapse,see refs.[20,146]).b

The ?CT allele of the NRG1gene has an in-frame stop codon within the sequence encoding the cytoplasmic tail.This mutation causes all transmembrane NRG1proteins produced from this allele to have a cytoplasmic tail length of only 3amino acids.See text (“A tale of the heart ...”)for interpretation of the NRG1?CT/?CT phenotype.

Black ?abnormal;blank ?normal;shaded ?not accessible for analysis because mice die prior to occurrence of this developmental event.Mice are normally born at embryonic day 21or 22(E21or E22).

20 D.L.Falls /Experimental Cell Research 284(2003)14–30

in the pathogenesis of disease,and what are the prospects for disease therapy based on modulating neuregulin signal-ing?Table 3summarizes currently investigated pathological and therapeutic considerations with respect to NRG1.Here I will brie ?y describe only one:the recent evidence for NRG involvement in schizophrenia.

Schizophrenia is a disabling neuropsychological disorder with strong familial characteristics suggestive of a genetic component [62].A genome wide survey of patients with familial schizophrenia in Iceland,employing both linkage and association methodologies,uncovered the NRG1gene as a candidate susceptibility gene for schizophrenia,and this association was con ?rmed in a Scottish population [34,63].The known activities of NRGs ?t well with current hypotheses regarding the neurobiological basis of schizo-phrenia.One theory proposes that schizophrenia results from a de ?ciency of glutamatergic innervation relative to dopaminergic innervation.Consistent with the idea that impairment of NRG1signaling contributes to the pathology of schizophrenia,mice heterozygous for two different mu-tations in the NRG1gene or a null mutation of the ErbB4gene display hyperactivity in behavioral tests similar to hyperactivity observed in mice treated with the psychogenic drug phencyclidine (PCP)or with mutations that impair glutamatergic neurotransmission or enhance dopaminergic neurotransmission [34,41,49].The NRG1mutant mice ex-amined in these behavioral studies had an in-frame stop codon introduced within the sequence encoding the NRG1EGF-like domain [41](inactivating all NRG1products of the mutated allele)or introduced within the sequence en-coding the transmembrane domain [34].The phenotype of this latter strain has not yet been reported,but I suspect it will be similar to mice in which the NRG1cytoplasmic tail has been severely truncated by targeted mutagenesis (see “A tale of the heart ...”below).Response to the antipsychotic drug clozapine and levels of glutamate receptors were also studied in the mice heterozygous for mutation of the NRG1transmembrane domain [34].Treatment with clozapine re-versed the hyperactivity of these mice,and they had reduced levels of the NMDA type of glutamate receptors,as as-sessed by binding of the NMDA receptor ligand MK801.Furthermore,application of soluble NRG1to cultured neu-rons stimulates transcription of the NMDA receptor subunit NR2C [64].

Another theory proposes that abnormalities in glial biol-ogy contribute to the pathology of schizophrenia [65].Neu-regulins are required for initial differentiation of oligoden-drocyte precursors and for their survival [43,48].A variant of this idea is that a de ?ciency of glial growth factors —such as NRG —predisposes to synaptic destabilization [66].It is clear that NRG signaling is required for the stabilization of neuromuscular synapses [67,68],and evidence for NRG involvement in astrocyte biology might implicate NRGs in formation or stabilization of central synapses [69](see also [70]).A third idea is that schizophrenia results from abnor-malities in cortical wiring,and NRGs have been shown to regulate migration of neuronal precursors in culture [71,72].A fourth hypothesis is that schizophrenia results from ab-normalities in synaptic plasticity,and NRG1s inhibit induc-tion of long-term potentiation,a form of synaptic plasticity studied as a model for the neurophysiological substrates of learning and memory [97].

Fig.1A depicts the boundaries of the genetic haplotype associated with schizophrenia.The only exon of reported NRG isoforms within these bounds is the exon encoding the type II (“GGF2”)N-terminal sequence.While widely ex-pressed in the nervous system during development and postnatally [39,73],no in vivo functions of type II NRGs are yet known,and no mutation within this exon that segregates with schizophrenia susceptibility has yet been detected.This raises the possibility that if alterations in NRG signal-ing are indeed involved in the pathogenesis of schizophre-

Table 3

Diseases/injuries in which the pathophysiology may involve perturbations in NRG1signaling and/or in which NRG1s may be of therapeutic use a Organ/organ system Effect

Type of evidence with references Nervous system Schizophrenia

[34,62,63,66]Nervous Multiple sclerosis (pathology and treatment)

[147–153]Nervous Promotion of neural regeneration/proliferation of olfactory ensheathing glia for therapeutic use in neural regeneration

[154–157]Nervous system Protection against neuropathy induced by cancer chemotherapeutic drug cisplatin [158]Nervous Response to traumatic brain injury [159]

Heart Trastuzumab (Herceptin)cardiotoxicity [87,88,92]Vascular Angiogenesis (in tumor growth)[160,161]Breast Breast tumor formation

[162]Breast Paget ’s disease (?-NRG as motility factor stimulating spread of neoplastic cells)[163]Skin Wound Healing (?-NRG as motility factor for keratinocytes)[164]Gut Hirschsprung ’s disease [136]Limb

Regeneration (newt limb)

[165]

This compilation includes diseases/injuries for which NRG involvement is supported by genetic evidence and/or by manipulation of NRG1signaling in animal models.The one exception is response to traumatic brain injury and stroke,for which involvement of NRG1signaling has been proposed solely on the basis of changes in NRG1expression.

D.L.Falls /Experimental Cell Research 284(2003)14–30

nia,the causative mutation may be in intronic or upstream sequence that regulates transcription or splicing. Sometimes a kiss sent from a distance may be

suf?cient,but in other situations a kiss on the lips may be required:NRG1paracrine signaling by shedding and secretion and NRG1juxtacrine signaling The ErbB family of receptors and their ligands has been described as a“signaling network”with an input layer comprised of ligands,receptors,and transactivators;a signal processing layer comprised of adapters,cascades,and tran-scription factors;and an output layer comprised of the biological consequences of ligand-ErbB interaction,such as stimulation of proliferation,inhibition of apoptosis,and differentiation([1];see also[98–100]and companion re-views in this issue).In moving from level to level of the network,there is both convergence and divergence,and there are horizontal(lateral)interactions within each level. While we have learned much about the network,we are only beginning to understand how the components of the net-work are selected,arranged,modi?ed,and modulated in individual cells to achieve physiologically adaptive out-comes of cell-cell interactions.

Much of the recent excitement in the study of cell-cell interactions derives from advances in de?ning the intracel-lular signal transduction pathways that couple(translate) receptor activation to cellular responses.However,equally important to understanding cell-cell interactions is de?ning the mechanisms that regulate the presentation of signals to receptors;that is,understanding the events upstream of receptor activation.Here the role of isoform topology and proteolytic processing in governing NRG-mediated cell-cell interactions will be considered.

Paracrine signaling by Ig-NRGs

Paracrine signaling refers to short distance cell-cell com-munication mediated by diffusible signaling molecules. Communication mediated by such diffusible signals allows cells not in direct contact to“talk to”each other.Proteins that serve as“paracrine signals”are commonly synthesized as soluble proteins,which—following processing in the ER-Golgi system and transport—are released by secretion, the spilling out of the traf?cking vesicle’s contents when it fuses with the cell’s plasma membrane.Pre-1992,when NRGs were still molecularly unidenti?ed“factors,”it was assumed they would turn out to be such typical paracrine signaling proteins,for they were being puri?ed as soluble proteins from medium conditioned by cultured transformed cells and from aqueous(nondetergent)extracts of brain and pituitary and were bioassayed by dissolving the partially puri?ed protein preparations in medium bathing responsive cells[2–7].Indeed,one NRG isoform—the NRG II-?3isoform,commonly referred to as GGF2or simply GGF—is believed to conform to this model.However,most NRGs are synthesized as transmembrane proteins.So,(1)are these transmembrane NRGs released to serve as paracrine sig-nals?And(2)if so,how?The answers are(1)yes,the Type I and(probably)the Type II transmembrane NRGs do generate paracrine signals and(2)the ectodomain is“shed”from the membrane by proteolytic processing[101–107].

The topology and processing of NRGs has been studied principally in cultures of?broblastic cells in which NRG isoforms have been expressed by transfection[77,82,108–113].Through such studies,it has been shown that type I TMc-NRGs are expressed as Nout/Cin single-pass trans-membrane proteins(type I membrane proteins,not to be confused with a type I NRG)that are cleaved in the“stalk”region to produce a paracrine signal.Type I NRGs(and likely type II NRGs,the other class of Ig-NRGs)appear to act principally as paracrine Type I(short distance,diffus-ible)signals.Since most type I NRGs in the nervous system are synthesized as transmembrane proteins,paracrine sig-naling requires proteolytic cleavage of the TMc-NRG pro-protein in the stalk region to release the bioactive ectodo-main fragment(NTF;see Fig.2).One example of a cell-cell interaction that appears quite clearly to be mediated by a soluble bioactive NRG1fragment produced by shedding is the communication between endocardium and myocardium that was discussed below(“A tale of the heart...”). Though space precludes more than a passing remark,it must be noted that shedding can produce an autocrine signal (signal-producing cell talks to itself or cells of the same type),as well as a paracrine signal(signal-producing cell talks to cells of a different type).Thus,reported cases of autocrine signaling by NRGs(i.e.,[114,115,140,142]) might involve shed TMc-NRGs.

Similar to Ig-NRGs,EGF,TGF?,and most other ligands for the EGFR are synthesized as transmembrane proproteins that are shed to serve as paracrine signals.Knockout of ADAM17(a.k.a.TACE)has demonstrated that this metal-loprotease is essential for the processing of TGF?and likely one or more other ligands of the EGFR[103,116]. ADAM17and ADAM19(a.k.a.meltrin-?)have been shown capable of mediating shedding of NRGs from cul-tured cells[117,118],but their role in governing NRG signal production in vivo remains unknown.Since inhibit-ing or stimulating metalloproteases is being considered in the therapy of various diseases,including Alzheimer’s dis-ease and cancer it will be important to determine the effects of candidate drugs on NRG signaling.

Are CRD-NRGs(type III NRGs)specialized to serve as juxtacrine signals?

Initially it was assumed that like type I NRGs,the type III NRGs with a transmembrane domain C-terminal of the EGF-like domain(TMc-NRGs),such as III-?1a,would be

22 D.L.Falls/Experimental Cell Research284(2003)14–30

single pass transmembrane proteins and that stalk cleavage of Type III NRGs would shed a bioactive ectodomain frag-ment that includes both the“cysteine-rich domain”(CRD) and the EGF-like domain.However,a direct test of this model in which type I and type III NRGs were expressed by transfection in?broblastic cell lines[77]yielded surprising results:the topology of NRG III-?1a is unlike the topology of NRG I?1a,and,in fact,unlike the topology of any previously reported RTK ligand(however,see[119,120]for discussion of an RTK ligand with another interesting topol-ogy).The sequences of NRG III-?1a and I-?1a differ only in their N-terminal regions;their sequence from the EGF-like domain through the C-terminus—including the se-quence of the TM domain and juxtamembrane segments is identical.However,instead of being an Nout/Cin single-pass transmembrane protein like type I TMc-NRGs,the type III NRGs with a TM-domain C-terminal of the EGF-like domain are Nin/Cin two-pass transmembrane proteins,with a hydrophobic segment within the CRD serving as a second transmembrane domain(see Fig.2).This has two major consequences:(1)Instead of being an extracellular protein-protein interaction domain as originally suspected,the CRD domain is mostly intramembrane and intracellular.(2)Stalk cleavage of Type III TMc-NRGs does not shed a bioactive ectodomain fragment,but instead creates a transmembrane N-terminal fragment.As would be predicted from the topo-logical differences,when type III and type I NRGs are expressed in parallel cultures,the amount of type III NRG released into the medium is much less than the amount of type I NRG,but the amount of type III NRG exposed at the cell surface—most of which is the transmembrane N-termi-nal fragment—is much more than the amount of type I NRG [77].It should be noted also that NRG III-?3,a form lacking the TM-domain C-terminal of the EGF-like domain also accumulates on the cell surface([79];J.Wang and D. Falls,unpublished data).These results raise the possibility that type III NRGs are specialized for juxtacrine(direct-contact)signaling,whereas type I NRGs are specialized for paracrine signaling.

There is evidence that type III NRGs do in fact serve as juxtacrine signals in vivo.Schwann cells are the glia of the peripheral nerves.One well-known function of Schwann cells is to myelinate the axons of sensory and motor neu-rons,thereby dramatically speeding conduction of action potentials.In sensory neuron-Schwann cell cocultures, Schwann cells in contact with sensory neuron axons have a higher proliferation rate than Schwann cells not contacting axons and this growth-promoting activity is blocked by antibodies inhibiting NRG signaling[121,122].That type III NRGs are an essential component of this contact-depen-dent signal is suggested by the profound depletion of Schwann cell populations in the type III NRG KO mice ([42];see Table2and discussion above).

A study of mechanisms regulating differentiation of Schwann cells from neural crest progenitors in a cell culture model has both demonstrated juxtacrine signaling by type III NRGs and shown that the consequences of signaling by membrane-bound type III can differ from the effects of signaling by soluble NRG[123].In these experiments,NRG III-?3was expressed in a small proportion of the cultured cells using a retroviral vector.As a control,green?uores-cent protein(GFP)was expressed using the same vector in parallel cultures.Cells contacting the NRG expressing cells were positive for Schwann cell markers at a signif-icantly higher frequency than cells contacting the GFP-expressing cells,demonstrating the juxtacrine signaling capability of type III NRG.Intriguingly,in similar cul-tures,soluble(recombinant)type III or type II NRG applied at high concentration was incapable of inducing expression of Schwann cell markers.The proposed pro-protein topology and mode of signaling for various iso-forms is summarized in Fig.3.Taken together,the cur-rent evidence argues for juxtacrine signaling by type III NRGs and paracrine signaling by types I and II.

Even for a kiss sent from a distance,the tingle can linger:prolongation of Ig-NRG’s effect by heparin The retention of type III NRGs in the membrane of type III expressing cells may not only limit the range of signaling,but also effectively concentrate the signal by con?ning it to the two-dimensional plane of the mem-brane.There is recent evidence for an alternative strategy of signal enhancement employed by Ig-NRGs.Each of the protein puri?cation schemes by which NRGs were initially isolated employed a step of heparin chromatog-raphy,and each puri?ed an Ig-NRG.In retrospect this is not surprising,as it was subsequently shown the Ig-like domain binds heparin and other highly charged glycos-aminoglycans[124].In contrast,CRD-NRGs do not bind heparin[79].

Glycosaminoglycans are the carbohydrate side chains of proteoglycans,proteins found in the extracellular matrix and on the surface of cells.The af?nity of the NRG Ig-like domain for cell-surface and extracellular matrix proteoglycans may provide a mechanism for lim-iting diffusion of Ig-NRGs and/or creating extracellular reservoirs of NRGs.Ig-NRGs are deposited in the basal lamina of the neuromuscular synapse[125,126].It has been proposed that Ig-NRGs become bound to the basal lamina by interaction of the Ig-like domain with basal lamina proteoglycans and that the bioactive EGF-like domain is subsequently freed from the matrix by a pro-tease that cleaves the matrix bound N-terminal fragment between the EGF-like domain and the Ig-like domain [124,127].

A new twist to this story is evidence that binding of Ig-NRGs to the surface of cultured muscle through interac-tion with cell-surface proteoglycans enhances the potency of the Ig-NRG in inducing receptor phosphorylation com-pared to a recombinant form consisting only of the EGF-like

D.L.Falls/Experimental Cell Research284(2003)14–30

domain.Furthermore,compared to the EGF-like domain-only form,the Ig-NRG induced a longer period of ErbB receptor phosphorylation and more effectively stimulated synthesis of acetylcholine receptors[128].

A tale of the heart(and of paracrine signaling,

Ig-NRGs,the NRG cytoplasmic tail,and NRG

traf?cking)

As noted above,mice genetically altered so that they produce no bioactive Ig-NRGs(Ig-NRG KOs)have the same cardiac phenotype as the pan-NRG KOs.Mice ho-mozygous for NRG1mutation that causes all transmem-brane NRG1s(TMc-NRG1s)to have their tail truncated to a length of only three amino acids(NRG1?CT/?CT mice)also have the same cardiac phenotype([44];see Table2).How-ever mice that produce no bioactive CRD-NRGs have not been reported to have cardiac defects.What is the underly-ing cell biology responsible for these results?Several lines of evidence can be woven together to construct an expla-nation.(1)The geometry of cardiac development is such that for normal cardiac morphogenesis,the endocardium must signal to cells in the presumptive myocardium with which it is not in direct contact.This requires a paracrine (diffusible)type of signal.As discussed above,the Ig-NRGs (Types I and II)may be specialized for paracrine signaling; whereas,the type III/CRD-NRGs may be poorly released from CRD-NRG producing cells and instead specialized for juxtacrine(direct-contact)signaling[77].(2)Type I NRGs and low levels of type III NRGs are expressed by the endocardium during embryogenesis,but type II NRGs are not expressed by the embryonic myocardium([39];see also [78]).(3)While most released and transmembrane proteins have a classic N-terminal signal sequence that targets the nascent protein to the endoplasmic reticulum,none of the NRGs do.Instead various other sequences in NRG isoforms appear to serve as“cryptic”,noncleaved internal signals targeting nascent NRG proteins to the ER-Golgi-export pathway.The?3-NRGs lack the transmembrane domain C-terminal of the EGF-like domain(see Fig.1).When expressed by transfection in?broblastic cells,NRG II-?3is effectively released into the medium[6]and NRG III-?3is effectively traf?cked to the cell surface[79],but NRG I-?3 is not released[2,6].This suggests that the types II and III N-terminal sequences each contain an ER targeting signal, but that the type I N-terminal sequence does not.Both the type II and III N-terminal regions include a hydrophobic stretch of amino acids,and it is likely that this hydrophobic stretch is all or part of the signal.So does this mean that type I NRGs are a“dead”class of NRG isoforms?Indeed not! Unlike NRG I-?3,the transmembrane type I NRGs(i.e., I-?1a,I-?2a,and I-?4a)are effectively released.Since these type I NRGs with a transmembrane domain carboxytermi-nal of the EGF-like domain(“TMc-NRGs”)are identical with NRG I-?3from the N-terminus through the EGF-like domain,the TM-NRGs must contain an export pathway targeting sequence downstream of the EGF-like domain. Likely the transmembrane domain serves as a signal-anchor sequence[80,81].But it turns out that a substantial length of the NRG cytoplasmic tail is also required for traf?cking of type I TMc-NRG to the cell surface and their release([77]; see also[82]).Whether the cytoplasmic tail in conjunction with the TM domain is required for initial targeting type I NRGs to the ER,or—as is the case for TGF?[83–85]—for transport along the export pathway,is unclear.

Now we have suf?cient information on the table to synthesize an explanation for the similarity in the cardiac phenotypes of mice with all NRG1s inactivated,only Ig-NRG1s inactivated,and tail-truncated TMc-NRG1s.A paracrine NRG signal is required for endocardial induction of myocardial differentiation,and type III/CRD-NRGs—though expressed at low levels by the myocardium—may not be suitable for paracrine signaling.Type II NRGs may be suitable,but they are not expressed.Type I NRGs are expressed,but type I NRGs without a cytoplasmic tail are not released:i.e.,the NRG1?CT/?CT mice would—from today’s perspective—be expected to have exactly the same phenotype as a type I NRG null.In summary,what we have learned of NRG’s cell biology provides a satisfying explanation of the fact that both the Ig-NRG KO and NRG1?CT/?CT mice have the same cardiac phenotype as the pan-NRG1KO.

Do NRGs also have functions in the heart beyond early development?The cardiac toxicity of trastuzumab(Hercep-tin)suggests they do.Trastuzumab is used in the treatment of metastatic breast cancer.It is a humanized monoclonal antibody that binds to the NRG receptor ErbB2(HER2) which is overexpressed in many breast cancers,and it not only reduces the ligand-independent activation of ErbB2 that occurs in cells highly expressing this protein or express-ing mutated ErbB2by down-regulating ErbB2[1,86],but also blocks NRG activation of ErbB2/4and ErbB3/4het-erodimers[1].A fraction of patients treated with trastu-zumab develop dilated cardiomyopathy,re?ecting weaken-ing of cardiac muscle contractility[87,88].Most of the trastuzumab-treated patients that develop cardiomyopathy are also being treated with the chemotherapeutic agent an-thracycline.Mouse models with a targeted mutation of ErbB2affecting only the ventricular muscle[89,90]develop dilated cardiomyopathy closely resembling the pathology in trastuzumab-treated patients.Furthermore,in cell culture neuregulins promote survival and growth of cardiac myo-cytes,and protect them from anthracycline toxicity[91,92]. Thus,NRGs mediate critical signaling in the adult heart,as well as in the developing heart.In the adult,the endothelium of the cardiac microvasculature may be a source of the (paracrine)NRG signal[91].Just as NRGs appear to me-diate signaling in the adult heart,so NRGs are likely to mediate critical signaling events in the adult nervous sys-tem.For example,NRGs and their receptors are widely expressed in the postnatal nervous system[73,93–95],NRG

24 D.L.Falls/Experimental Cell Research284(2003)14–30

expression in the brain is upregulated by activity[96],and NRGs can inhibit long-term potentiation(LTP),a model of learning[97].Thus therapeutic strategies that involve per-turbing NRG signaling,such as the use of trastuzumab, must take careful cognizance of the normal functions of NRGs in adult,as well as embryonic,organ systems. Conclusion

We certainly are just at the beginning of deciphering the functions of NRGs and the mechanisms by which the NRG signaling is shaped and modulated to achieve physiologi-cally adaptive outcomes,but already it is clear the NRGs play critical roles in the functioning of a number of organ systems,both during embryonic development and postna-tally.Evidence that aberrations in NRG signaling contribute to the pathology of diseases such as schizophrenia and multiple sclerosis lend additional urgency to expanding our understanding of NRG biology.An appreciation of diversi-?cation of signaling through employment of combinations of receptors and variations in intracellular signaling cas-cades has grown over recent years.Now it seems clear that structural differences in NRG isoforms tailor them for dif-ferent signaling strategies and requirements,providing con-siderable additional diversi?cation upstream of receptor ac-tivation.

Acknowledgment

Preparation of this article was supported by a grant to D.L.F.from the National Institutes of Health(GM56337).

Note added in proof.A good entry point for access to the wealth of NRG isoform information freely available via the World Wide Web is LocusLink[170,171](http:www. https://www.wendangku.net/doc/5d18999359.html,/LocusLink/).To begin accessing the Lo-cusLink neuregulin information,on the LocusLink home page enter“neuregulin”in the query box(without quotation marks),and then click“Go.”This will take you to a page listing neuregulin loci.Note that not all species are included in the LocusLink database.For example,as of December 2002,the frog(Xenopus)and chicken(Gallus)NRG1se-quences are not referenced in LocusLink.The Mouse Ge-nome Informatics database(MGI)can be accessed from LocusLink or directly(https://www.wendangku.net/doc/5d18999359.html,).Among many other things,this valuable compilation of NRG data includes a listing of each reported targeted mutations of the NRG1gene,along with the of?cial nomenclature for each mutation.

References

[1]Y.Yarden,M.X.Sliwkowski,Untangling the ErbB signaling net-

work,Nat.Rev.Mol.Cell.Biol.2(2001)127–137.

[2]W.E.Holmes,M.X.Sliwkowski,R.W.Akita,W.J.Henzel,J.Lee,

J.W.Park,D.Yansura,N.Abadi,H.Raab,G.D.Lewis,H.M.

Shepard,W.-J.Kuang,W.I.Wood,D.V.Goeddel,R.L.Vandlen, Identi?cation of heregulin,a speci?c activator of p185erbB2,Sci-ence256(1992)1205–1210.

[3]E.Peles,S.S.Bacus,R.A.Koski,H.S.Lu,D.Wen,S.G.Ogden,

R.B.Levy,Y.Yarden,Isolation of the neu/HER-2stimulatory ligand:a44kd glycoprotein that induces differentiation of mam-mary tumor cells,Cell69(1992)205–216.

[4]D.Wen,E.Peles,R.Cupples,S.V.Suggs,S.S.Bacus,Y.Luo,G.

Trail,S.Hu,S.M.Silbiger,R.B.Levy,R.A.Koski,H.S.Lu,Y.

Yarden,Neu differentiation factor:a transmembrane glycoprotein containing an EGF domain and an immunoglobulin homology unit, Cell69(1992)559–572.

[5]A.D.J.Goodearl,J.B.Davis,K.Mistry,L.Minghetti,M.Otsu,M.D.

Water?eld,P.Stroobant,Puri?cation of multiple forms of glial growth factor,J.Biol.Chem.268(1993)18095–18102.

[6]M.A.Marchionni,A.D.J.Goodearl,M.S.Chen,O.Bermingham-

McDonogh, C.Kirk,M.Hendricks, F.Danehy, D.Misumi,J.

Sudhalter,K.Kobayashi,D.Wroblewski,C.Lynch,M.Baldassare,

I.Hiles,J.B.Davis,J.J.Hsuan,N.F.Totty,M.Otsu,R.N.McBur-

ney,M.D.Water?eld,P.Stroobant,D.Gwynne,Glial growth fac-tors are alternatively spliced erbB2ligands expressed in the nervous system,Nature362(1993)312–318.

[7]D.L.Falls,K.M.Rosen,G.Corfas,https://www.wendangku.net/doc/5d18999359.html,ne,G.D.Fischbach,

ARIA,a protein that stimulates acetylcholine receptor synthesis,is

a member of the neu ligand family,Cell72(1993)801–815.

[8]S.J.Bus?eld,D.A.Michnick,T.W.Chickering,T.L.Revett,J.Y.

Ma,E.A.Woolf,https://www.wendangku.net/doc/5d18999359.html,rack,B.J.Dussault,J.Woolf,A.D.J.

Goodearl,D.P.Gearing,Characterization of a neuregulin-related gene,don-1,that is highly expressed in restricted regions of the cerebellum and hippocampus,Mol.Cell.Biol.17(1997)4007–4014.

[9]K.L.Carraway,J.L.Weber,M.J.Unger,J.Ledesma,N.Yu,M.

Gassmann,https://www.wendangku.net/doc/5d18999359.html,i,Neuregulin-2,a new ligand of erbB3/erbB4-receptor tyrosine kinases,Nature387(1997)512–516.

[10]H.Chang,D.J.Riese,W.Gilbert,D.F.Stern,U.J.McMahan,Li-

gands for erbB-family receptors encoded by a neuregulin-like gene, Nature387(1997)509–512.

[11]S.Higashiyama,M.Horikawa,K.Yamada,N.Ichino,N.Nakano,

T.Nakagawa,J.Miyagawa,N.Matsushita,T.Nagatsu,N.Tanigu-chi,H.Ishiguro,A novel brain-derived member of the epidermal growth factor family that interacts with ErbB3and ErbB4,J.Bio-chem.(Tokyo)122(1997)675–680.

[12]D.Zhang,M.X.Sliwkowski,M.Mark,G.Frantz,R.Akita,Y.Sun,

K.Hillan, C.Crowley,J.Brush,P.J.Godowski,Neuregulin-3 (NRG3):a novel neural tissue-enriched protein that binds and acti-vates ErbB4,https://www.wendangku.net/doc/5d18999359.html,A94(1997)9562–9567.

[13]D.Harari,E.Tzahar,J.Romano,M.Shelly,J.H.Pierce,G.C.

Andrews,Y.Yarden,Neuregulin-4:a novel growth factor that acts through the ErbB-4receptor tyrosine kinase,Oncogene18(1999) 2681–2689.

[14]J.T.Jones,R.W.Akita,M.X.Sliwkowski,Binding speci?cities and

af?nities of egf domains for ErbB receptors,FEBS Lett.447(1999) 227–231.

[15]S.S.Hobbs,S.L.Cof?ng,A.T.Le,E.M.Cameron,E.E.Williams,

M.Andrew,E.N.Blommel,R.P.Hammer,H.Chang,D.J.Riese 2nd,Neuregulin isoforms exhibit distinct patterns of ErbB family receptor activation,Oncogene21(2002)8442–8452.

[16]M.A.Huotari,P.J.Miettinen,J.Palgi,T.Koivisto,https://www.wendangku.net/doc/5d18999359.html,tinov,D.

Harari,Y.Yarden,T.Otonkoski,ErbB signaling regulates lineage determination of developing pancreatic islet cells in embryonic organ culture,Endocrinology143(2002)4437–4446.

[17]A.Buonanno,G.D.Fischbach,Neuregulin and ErbB receptor sig-

naling pathways in the nervous system,Curr.Opin.Neurobiol.11 (2001)287–296.

D.L.Falls/Experimental Cell Research284(2003)14–30

[18]C.S.Crovello,https://www.wendangku.net/doc/5d18999359.html,i,L.C.Cantley,K.L.Carraway3rd,Differential

signaling by the epidermal growth factor-like growth factors neu-regulin-1and neuregulin-2,J.Biol.Chem.273(1998)26954–26961.

[19]C.Sweeney,https://www.wendangku.net/doc/5d18999359.html,i,D.J.Riese II,J.Diamonti,L.C.Cantley,K.L.

Carraway III,Ligand discrimination in signaling through an ErbB4 receptor homodimer,J.Biol.Chem.275(2000)19803–19807. [20]G.D.Fischbach,K.M.Rosen,ARIA—a neuromuscular junction

neuregulin,Annu.Rev.Neurosci.20(1997)429–458.

[21]M.Gassmann,G.Lemke,Neuregulins and neuregulin receptors in

neural development,Curr.Opin.Neurobiol.7(1997)87–92. [22]S.Burden,Y.Yarden,Neuregulins and their receptors:a versatile

signaling module in organogenesis and oncogenesis,Neuron18 (1997)847–855.

[23]K.Adlkofer,https://www.wendangku.net/doc/5d18999359.html,i,Role of neuregulins in glial cell development,

Glia29(2000)104–111.

[24]A.N.Garratt,S.Britsch,C.Birchmeier,Neuregulin,a factor with

many functions in the life of a schwann cell,Bioessays22(2000) 987–996.

[25]A.M.Davies,Neuronal survival:early dependence on Schwann

cells,Curr.Biol.8(1998)R15–18.

[26]S.J.Burden,The formation of neuromuscular synapses,Genes Dev.

12(1998)133–148.

[27]J.R.Sanes,J.W.Lichtman,Induction,assembly,maturation and

maintenance of a postsynaptic apparatus,Nat.Rev.Neurosci.2 (2001)791–805.

[28]L.Schaeffer,A.de Kerchove d’Exaerde,J.P.Changeux,Targeting

transcription to the neuromuscular synapse,Neuron31(2001)15–

22.

[29]K.R.Jessen,R.Mirsky,Origin and early development of Schwann

cells,Microsc.Res.Tech.41(1998)393–402.

[30]B.A.Barres,M.C.Raff,Axonal control of oligodendrocyte devel-

opment,J.Cell.Biol.147(1998)1123–1128.

[31]K.R.Jessen,R.Mirsky,Schwann cells and their precursors emerge

as major regulators of nerve development,Trends Neurosci.22 (1999)402–410.

[32]G.Lemke,Glial control of neuronal development,Annu.Rev.Neu-

rosci.24(2001)87–105.

[33]R.Mirsky,K.R.Jessen,A.Brennan,D.Parkinson,Z.Dong,C.

Meier,E.Parmantier,https://www.wendangku.net/doc/5d18999359.html,wson,Schwann cells as regulators of nerve development,J.Physiol.Paris96(2002)17–24.

[34]H.Stefansson,E.Sigurdsson,V.Steinthorsdottir,S.Bjornsdottir,T.

Sigmundsson,S.Ghosh,J.Brynjolfsson,S.Gunnarsdottir,O.Ivars-son,T.T.Chou,O.Hjaltason,B.Birgisdottir,H.Jonsson,V.G.

Gudnadottir,E.Gudmundsdottir,A.Bjornsson,B.Ingvarsson,A.

Ingason,S.Sigfusson,H.Hardardottir,R.P.Harvey,https://www.wendangku.net/doc/5d18999359.html,i,M.

Zhou,D.Brunner,V.Mutel,A.Gonzalo,G.Lemke,J.Sainz,G.

Johannesson,T.Andresson,D.Gudbjartsson,A.Manolescu,M.L.

Frigge,M.E.Gurney,A.Kong,J.R.Gulcher,H.Petursson,K.

Stefansson,Neuregulin1and susceptibility to schizophrenia,Am.J.

Hum.Genet.71(2002)877–892.

[35]W.H.Ho,M.P.Armanini,A.Nuijens,H.S.Phillips,P.L.Osheroff,

Sensory and motor neuron-derived factor—a novel heregulin variant highly expressed in sensory and motor neurons,J.Biol.Chem.270 (1995)14523–14532.

[36]D.Meyer,C.Birchmeier,Multiple essential functions of neuregulin

in development,Nature378(1995)386–390.

[37]R.Kramer,N.Bucay,D.J.Kane,L.E.Martin,J.E.Tarpley,L.E.

Theill,Neuregulins with an Ig-like domain are essential for mouse myocardial and neuronal development,https://www.wendangku.net/doc/5d18999359.html,A 93(1996)4833–4838.

[38]S.Erickson,K.Shea,N.Ghaboosi,L.Loverro,G.Frantz,M.Bauer,

L.Lu,M.Moore,ErbB3is required for normal cerebellar and cardiac development:a comparison with ErbB2-and heregulin-de?cient mice,Development124(1997)4999–5011.[39]D.Meyer,T.Yamaai,A.Garratt,E.Riethmacher-Sonnenberg,D.

Kane,L.E.Theill,C.Birchmeier,Isoform-speci?c expression and function of neuregulin,Development124(1997)3575–3586. [40]S.Britsch,L.Li,S.Kirchhoff,F.Theuring,V.Brinkmann,C.

Birchmeier,D.Riethmacher,The ErbB2and ErbB3receptors and their ligand,neuregulin-1,are essential for development of the sympathetic nervous system,Genes Dev.12(1998)1825–1836. [41]R.Gerlai,P.Pisacane,S.Erickson,Heregulin,but not ErbB2or

ErbB3,heterozygous mutant mice exhibit hyperactivity in multiple behavioral tasks,Behav.Brain Res.109(2000)219–227.

[42]D.Wolpowitz,T.B.Mason,P.Dietrich,M.Mendelsohn, D.A.

Talmage,L.W.Role,Cysteine-rich domain isoforms of the neuregu-lin-1gene are required for maintenance of peripheral synapses, Neuron25(2000)79–91.

[43]S.K.Park,https://www.wendangku.net/doc/5d18999359.html,ler,I.Krane,T.Vartanian,The erbB2gene is

required for the development of terminally differentiated spinal cord oligodendrocytes,J.Cell.Biol.154(2001)1245–1258.

[44]X.Liu,H.Hwang,L.Cao,M.Buckland,A.Cunningham,J.Chen,

K.R.Chien,R.M.Graham,M.Zhou,Domain-speci?c gene disrup-tion reveals critical regulation of neuregulin signaling by its cyto-plasmic tail,https://www.wendangku.net/doc/5d18999359.html,A95(1998)13024–13029.

[45]L.Li,S.Cleary,M.A.Mandarano,W.Long,C.Birchmeier,F.E.

Jones,The breast proto-oncogene,HRGalpha regulates epithelial proliferation and lobuloalveolar development in the mouse mam-mary gland,Oncogene21(2002)4900–4907.

[46]X.Yang,S.Arber,C.William,L.Li,Y.Tanabe,T.M.Jessell,C.

Birchmeier,S.J.Burden,Patterning of muscle acetylcholine receptor gene expression in the absence of motor innervation,Neuron30 (2001)399–410.

[47]A.W.Sandrock,S.E.Dryer,K.M.Rosen,S.N.Gozani,R.Kramer,

L.E.Theill,G.D.Fischbach,Maintenance of acetylcholine receptor number by neuregulins at the neuromuscular junction in vivo,Sci-ence276(1997)599–603.

[48]T.Vartanian,G.Fischbach,https://www.wendangku.net/doc/5d18999359.html,ler,Failure of spinal cord oligo-

dendrocyte development in mice lacking neuregulin,Proc.Natl.

https://www.wendangku.net/doc/5d18999359.html,A96(1999)731–735.

[49]M.Gassmann,F.Casagranda,D.Orioli,H.Simon,https://www.wendangku.net/doc/5d18999359.html,i,R.Klein,

G.Lemke,Aberrant neural and cardiac development in mice lacking

the erbb4neuregulin receptor,Nature378(1995)390–394. [50]K.F.Lee,H.Simon,H.Chen,B.Bates,M.C.Hung,C.Hauser,

Requirement for neuregulin receptor erbB2in neural and cardiac development,Nature378(1995)394–398.

[51]D.Riethmacher,E.Sonnenberg-Riethmacher,V.Brinkmann,T.

Yamaai,G.R.Lewin,C.Birchmeier,Severe neuropathies in mice with targeted mutations in the ErbB3receptor,Nature389(1997) 725–730.

[52]J.K.Morris,W.Lin,C.Hauser,Y.Marchuk,D.Getman,K.F.Lee,

Rescue of the cardiac defect in ErbB2mutant mice reveals essential roles of ErbB2in peripheral nervous system development,Neuron 23(1990)273–283.

[53]M.T.Woldeyesus,S.Britsch,D.Riethmacher,L.Xu,E.Sonnen-

berg-Riethmacher,F.Abou-Rebyeh,R.Harvey,P.Caroni,C.Birch-meier,Peripheral nervous system defects in erbB2mutants follow-ing genetic rescue of heart development,Genes Dev.13(1999) 2538–2548.

[54]A.N.Garratt,O.Voiculescu,P.Topilko,P.Charnay,C.Birchmeier,

A dual role of erbB2in myelination and in expansion of the

Schwann cell precursor pool,J.Cell Biol.148(2000)1035–1046.

[55]J.P.Golding,P.Trainor,R.Krumlauf,M.Gassmann,Defects in

path?nding by cranial neural crest cells in mice lacking the neu-regulin receptor ErbB4,Nat.Cell.Biol.2(2000)103–109. [56]W.Lin,H.B.Sanchez,T.Deerinck,J.K.Morris,M.Ellisman,K.F.

Lee,Aberrant development of motor axons and neuromuscular syn-apses in erbB2-de?cient mice,https://www.wendangku.net/doc/5d18999359.html,A97(2000) 1299–1304.

26 D.L.Falls/Experimental Cell Research284(2003)14–30

[57]R.Chan,W.R.Hardy,https://www.wendangku.net/doc/5d18999359.html,ing,S.E.Hardy,W.J.Muller,The

catalytic activity of the ErbB-2receptor tyrosine kinase is essential for embryonic development,Mol.Cell.Biol.22(2002)1073–1078.

[58]J.Y.Wang,K.E.Frenzel, D.Wen, D.L.Falls,Transmembrane

neuregulins interact with LIM kinase1,a cytoplasmic protein kinase implicated in development of visuospatial cognition,J.Biol.Chem.

273(1998)20525–20534.

[59]J.Bao,J.Gautier,L.Role,D.Talmage,Novel functions of the

cytoplasmic domain of neuregulin,Soc.Neurosci.Abstr.25(1999) 303.9.

[60]M.L.Hancock,L.Role,D.Talmage,Neuregulin1intracellular

domain-dependent signaling:requirements for intramembranous proteolysis and nuclear targeting,Soc.Neurosci.Abstr.28(2002) 822.11.

[61]D.G.Wilkinson,Multiple roles of EPH receptors and ephrins in

neural development,Nat.Rev.Neurosci.2(2001)155–164. [62]Cloninger,C.R.The discovery of susceptibility genes for mental

https://www.wendangku.net/doc/5d18999359.html,A99(2002)13365–13367. [63]Stefansson,H.,Sarginson,J.,Kong,A.,Yates,P.,Steinthorsdottir,

V.,Gud?nnsson,E.,Gunnarsdottir,S.,Walker,N.,Petursson,H., Crombie,C.,Ingason,A.,Gulcher,J.R.,Stefansson,K.,Clair,D.S., Association of neuregulin1with schizophrenia con?rmed in a Scottish population,Am.J.Hum.Genet.72(2003)83–87. [64]M.Ozaki,M.Sasner,R.Yano,H.S.Lu,A.Buonanno,Neuregulin-

beta induces expression of an NMDA-receptor subunit,Nature390 (1997)691–694.

[65]Y.Hakak,J.R.Walker,C.Li,W.H.Wong,K.L.Davis,J.D.Bux-

baum,V.Haroutunian,A.A.Fienberg,Genome-wide expression analysis reveals dysregulation of myelination-related genes in chronic schizophrenia,https://www.wendangku.net/doc/5d18999359.html,A98(2001)4746–4751.

[66]H.W.Moises,T.Zoega,I.I.Gottesman,The glial growth factors

de?ciency and synaptic destabilization hypothesis of schizophrenia.

BMC Psychiatry2(2002)8.

[67]J.T.Trachtenberg,W.J.Thompson,Nerve terminal withdrawal from

rat neuromuscular junctions induced by neuregulin and Schwann cells,J.Neurosci.17(1997)6243–6255.

[68]J.T.Trachtenberg,W.J.Thompson,Schwann cell apoptosis at de-

veloping neuromuscular junctions is regulated by glial growth fac-tor,Nature379(1996)174–177.

[69]R.Pinkas-Kramarski,R.Eilam,O.Spiegler,https://www.wendangku.net/doc/5d18999359.html,vi,N.Liu,D.

Chang,D.Wen,M.Schwartz,Y.Yarden,Brain neurons and glial cells express Neu differentiation factor/heregulin:a survival factor for astrocytes,https://www.wendangku.net/doc/5d18999359.html,A91(1994)9387–9391.

[70]E.M.Ullian,S.K.Sapperstein,K.S.Christopherson,B.A.Barres,

Control of synapse number by glia,Science291(2001)657–661.

[71]C.Rio,H.I.Rieff,P.Qi,G.Corfas,Neuregulin and erbB receptors

play a critical role in neuronal migration,Neuron19(1997)39–50.

[72]E.S.Anton,M.A.Marchionni,K.F.Lee,P.Rakic,Role of GGF/

neuregulin signaling in interactions between migrating neurons and radial glia in the developing cerebral cortex,Development124 (1997)3501–3510.

[73]M.S.Chen,O.Bermingham-McDonogh,F.T.Danehy,C.Nolan,

S.S.Scherer,J.Lucas,D.Gwynne,M.A.Marchionni,Expression of multiple neuregulin transcripts in postnatal rat brains,https://www.wendangku.net/doc/5d18999359.html,p.

Neurol.349(1994)389–400.

[74]H.I.Rieff,L.T.Raetzman,D.W.Sapp,H.H.Yeh,R.E.Siegel,G.

Corfas,Neuregulin induces GABA(A)receptor subunit expression and neurite outgrowth in cerebellar granule cells,J.Neurosci.19 (1999)10757–10766.

[75]Y.Liu,B.Ford,M.A.Mann,G.D.Fischbach,Neuregulins increase

alpha7nicotinic acetylcholine receptors and enhance excitatory syn-aptic transmission in GABAergic interneurons of the hippocampus, J.Neurosci.21(2001)5660–5669.

[76]X.Yang,Y.Kuo,P.Devay,C.Yu,L.Role,A cysteine-rich isoform

of neuregulin controls the level of expression of neuronal nicotinic

receptor channels during synaptogenesis,Neuron20(1998)255–270.

[77]J.Y.Wang,https://www.wendangku.net/doc/5d18999359.html,ler,D.L.Falls,The N-terminal region of neu-

regulin isoforms determines the accumulation of cell-surface and released neuregulin ectodomain,J.Biol.Chem.276(2001)2841–2851.

[78]G.Corfas,K.M.Rosen,H.Aratake,R.Krauss,G.D.Fischbach,

Differential expression of ARIA isoforms in the rat brain,Neuron14 (1995)103–115.

[79]A.Schroering,D.J.Carey,Sensory and motor neuron-derived factor

is a transmembrane heregulin that is expressed on the plasma mem-brane with the active domain exposed to the extracellular environ-ment,J.Biol.Chem.273(1998)30643–30650.

[80]K.E.Matlack,W.Mothes,T.A.Rapoport,Protein translocation:

tunnel vision,Cell92(1998)381–390.

[81]K.Ota,M.Sakaguchi,G.von Heijne,N.Hamasaki,K.Mihara,

Forced transmembrane orientation of hydrophilic polypeptide seg-ments in multispanning membrane proteins,Mol.Cell.2(1998) 495–503.

[82]X.Liu,H.Hwang,L.Cao,D.Wen,N.Liu,R.M.Graham,M.Zhou,

Release of the neuregulin functional polypeptide requires its cyto-plasmic tail,J.Biol.Chem.273(1998)34335–34340.

[83]J.M.Urena,A.Merlos-Suarez,J.Baselga,J.Arribas,The cytoplas-

mic carboxy-terminal amino acid determines the subcellular local-ization of proTGF-(alpha)and membrane type matrix metallopro-tease(MT1-MMP),J.Cell.Sci.112(1999)773–784.

[84]A.Kuo,C.Zhong,https://www.wendangku.net/doc/5d18999359.html,ne,R.Derynck,Transmembrane trans-

forming growth factor-alpha tethers to the PDZ domain-containing, golgi membrane-associated protein p59/GRASP55,EMBO J.19 (2000)6427–6439.

[85]J.Fernandez-Larrea,A.Merlos-Suarez,J.M.Urena,J.Baselga,J.

Arribas,A role for a PDZ protein in the early secretory pathway for the targeting of proTGF-alpha to the cell surface,Mol.Cell.3(1999) 423–433.

[86]E.Penuel,R.W.Akita,M.X.Sliwkowski,Identi?cation of a region

within the ErbB2/HER2intracellular domain that is necessary for ligand-independent association,J.Biol.Chem.277(2002)28468–28473.

[87]J.W.Schneider,A.Y.Chang,A.Garratt,Trastuzumab cardiotoxic-

ity:Speculations regarding pathophysiology and targets for further study,Semin.Oncol.29(2002)22–28.

[88]J.W.Schneider,A.Y.Chang,T.P.Rocco,Cardiotoxicity in signal

transduction therapeutics:erbB2antibodies and the heart,Semin.

Oncol.28(2001)18–26.

[89]C.Ozcelik,B.Erdmann,B.Pilz,N.Wettschureck,S.Britsch,N.

Hubner,K.R.Chen,C.Birchmeier,A.N.Garratt,Conditional mu-tation of the ErbB2(HER2)receptor in cardiomyocytes leads to dilated cardiomyopathy,https://www.wendangku.net/doc/5d18999359.html,A99(2002) 8880–8885.

[90]S.A.Crone,Y.Y.Zhao,L.Fan,Y.Gu,S.Minamisawa,Y.Liu,K.L.

Peterson,J.Chen,R.Kahn,G.Condorelli,J.Ross Jr.,K.R.Chien, K.F.Lee,ErbB2is essential in the prevention of dilated cardiomy-opathy,Nat.Med.8(2002)459–465.

[91]Y.Zhao, D.R.Sawyer,R.R.Baliga, D.J.Opel,X.Han,M.A.

Marchionni,R.A.Kelly,Neuregulins promote survival and growth of cardiac myocytes.Persistence of erbb2and erbb4expression in neonatal and adult ventricular myocytes,J.Biol.Chem.273(1998) 10261–10269.

[92]D.B.Sawyer,C.Zuppinger,https://www.wendangku.net/doc/5d18999359.html,ler,H.M.Eppenberger,T.M.

Suter,Modulation of anthracycline-induced myo?brillar disarray in rat ventricular myocytes by neuregulin-1beta and anti-erbB2:poten-tial mechanism for trastuzumab-induced cardiotoxicity,Circulation 105(2002)1551–1554.

[93]O.Bermingham-McDonogh,Y.T.Xu,M.A.Marchionni,S.S.

Scherer,Neuregulin expression in PNS neurons:isoforms and reg-ulation by target interactions,Mol.Cell.Neurosci.10(1997)184–195.

D.L.Falls/Experimental Cell Research284(2003)14–30

[94]S.L.Carroll,https://www.wendangku.net/doc/5d18999359.html,ler,P.W.Frohnert,S.S.Kim,J.A.Corbett,

Expression of neuregulins and their putative receptors,erbb2and erbb3,is induced during wallerian degeneration,J.Neurosci.17 (1997)1642–1659.

[95]K.M.Gerecke,J.M.Wyss,I.Karavanova,A.Buonanno,S.L.Car-

roll,Erbb transmembrane tyrosine kinase receptors are differentially expressed throughout the adult rat central nervous system,https://www.wendangku.net/doc/5d18999359.html,p.

Neurol.433(2001)86–100.

[96]R.Eilam,R.Pinkas-Kramarski,B.J.Ratzkin,M.Segal,Y.Yarden,

Activity-dependent regulation of neu differentiation factor/neuregu-lin expression in rat brain,https://www.wendangku.net/doc/5d18999359.html,A95(1998) 1888–1893.

[97]Y.Z.Huang,S.Won,D.W.Ali,Q.Wang,M.Tanowitz,Q.S.Du,

K.A.Pelkey,D.J.Yang,W.C.Xiong,M.W.Salter,L.Mei,Regu-lation of neuregulin signaling by PSD-95interacting with ErbB4at CNS synapses,Neuron26(2000)443–455.

[98]M.A.Olayioye,R.M.Neve,https://www.wendangku.net/doc/5d18999359.html,ne,N.E.Hynes,The ErbB

signaling network:receptor heterodimerization in development and cancer,EMBO J.19(2000)3159–3167.

[99]R.Pinkas-Kramarski,M.Shelly,B.C.Guarino,L.M.Wang,L.

Lyass,I.Alroy,M.Alimandi,A.Kuo,J.D.Moyer,https://www.wendangku.net/doc/5d18999359.html,vi,M.

Eisenstein,B.J.Ratzkin,R.Seger,S.S.Bacus,J.H.Pierce,G.C.

Andrews,Y.Yarden,M.Alamandi,ErbB tyrosine kinases and the two neuregulin families constitute a ligand-receptor network[pub-lished errata,including important error re switched labeling of NRG2-?and NRG2-?,appear in Mol.Cell.Biol.1998Dec;18(12): 7602and1999Dec;19(12):8695],Mol.Cell.Biol.18(1998)6090–6101.

[100]R.Pinkas-Kramarski,L.Soussan,H.Waterman,G.Levkowitz,I.

Alroy,L.Klapper,https://www.wendangku.net/doc/5d18999359.html,vi,R.Seger,B.J.Ratzkin,M.Sela,Y.

Yarden,Diversi?cation of neu differentiation factor and epidermal growth factor signaling by combinatorial receptor interactions, EMBO J.15(1996)2452–2467.

[101]J.Massague′,Transforming Growth Factor-?:a model for mem-brane-anchored growth factors,J.Biol.Chem.265(1990)21393–21396.

[102]R.A.Black,J.M.White,ADAMs:focus on the protease domain, Curr.Opin.Cell.Biol.10(1998)654–659.

[103]J.J.Peschon,J.L.Slack,P.Reddy,K.L.Stocking,S.W.Sunnarborg,

D.C.Lee,W.

E.Russell,B.J.Castner,R.S.Johnson,J.N.Fitzner,

R.W.Boyce,N.Nelson,C.J.Kozlosky,M.F.Wolfson,C.T.Rauch,

D.P.Cerretti,R.J.Paxton,C.J.March,R.A.Black,An essential role

for ectodomain shedding in mammalian development,Science282 (1998)1281–1284.

[104]Z.Werb,Y.Yan,A cellular striptease act[comment],Science282 (1998)1279–1280.

[105]A.J.Turner,N.M.Hooper,Role for ADAM-family proteinases as membrane protein secretases,Biochem.Soc.Trans.27(1999)255–259.

[106]C.P.Blobel,Remarkable roles of proteolysis on and beyond the cell surface,Curr.Opin.Cell.Biol.12(2000)606–612.

[107]P.Primakoff,D.G.Myles,The ADAM gene family:surface proteins with adhesion and protease activity,Trends Genet.16(2000)83–87. [108]D.Z.Wen,S.V.Suggs,D.Karunagaran,N.L.Liu,R.L.Cupples,Y.

Luo,A.M.Janssen,N.Benbaruch,D.B.Trollinger,V.L.Jacobsen, S.Y.Meng,H.S.Lu,S.Hu,D.Chang,W.N.Yang,D.Yanigahara, R.A.Koski,Y.Yarden,Structural and functional aspects of the multiplicity of neu differentiation factors,Mol.Cell.Biol.14(1994) 1909–1919.

[109]T.L.Burgess,S.L.Ross,Y.X.Qian,D.Brankow,S.Hu,Biosyn-thetic processing of neu differentiation factor—glycosylation,traf-?cking,and regulated cleavage from the cell surface,J.Biol.Chem.

270(1995)19188–19196.

[110]J.A.Loeb,E.T.Susanto,G.D.Fischbach,The neuregulin precursor proARIA is processed to ARIA after expression on the cell surface by a protein kinase C-enhanced mechanism,Mol.Cell.Neurosci.11 (1998)77–91.[111]B.Han,G.D.Fischbach,The release of acetylcholine receptor in-ducing activity(ARIA)from its transmembrane precursor in trans-fected?broblasts,J.Biol.Chem.274(1999)26407–26415. [112]E.Diaz-Rodriguez,A.Esparis-Ogando,J.C.Montero,L.Yuste,A.

Pandiella,Stimulation of cleavage of membrane proteins by calmod-ulin inhibitors,Biochem.J.346(Pt2)(2000)359–367.

[113]H.S.Lu,S.Hara,L.Wong,M.D.Jones,V.Katta,G.Trail,A.H.

Zou,D.Brankow,S.Cole,S.Hu,D.Z.Wen,Post-translational processing of membrane-associated neu differentiation factor proisoforms expressed in mammalian cells,J.Biol.Chem.270 (1995)4775–4783.

[114]M.A.Avila,J.A.Velasco,C.Cho,R.Lupu,D.Z.Wen,V.Notario, Hyperactive autocrine loop mediated by a Ndf-related factor in neoplastic hamster embryo?broblasts expressing an activated Cph oncogene,Oncogene10(1995)963–971.

[115]M.Alimandi,A.Romano,M.C.Curia,R.Muraro,P.Fedi,S.A.

Aaronson,P.P.Di?ore,M.H.Kraus,Cooperative signaling of Erbb3 and Erbb2in neoplastic transformation and human mammary car-cinomas,Oncogene10,(1995)1813–1821.

[116]S.W.Sunnarborg,C.L.Hinkle,M.Stevenson,W.E.Russell,C.S.

Raska,J.J.Peschon,B.J.Castner,M.J.Gerhart,R.J.Paxton,R.A.

Black,D.C.Lee,Tumor necrosis factor-alpha converting enzyme (TACE)regulates epidermal growth factor receptor ligand availabil-ity,J.Biol.Chem.277(2002)12838–12845.

[117]K.Shirakabe,S.Wakatsuki,T.Kurisaki,A.Fujisawa-Sehara,Roles of meltrin beta/ADAM19in the processing of neuregulin,J.Biol.

Chem.276(2001)9352.

[118]J.C.Montero,L.Yuste,E.Diaz-Rodriguez,A.Esparis-Ogando,A.

Pandiella,Differential shedding of transmembrane neuregulin iso-forms by the tumor necrosis factor-alpha-converting enzyme,Mol.

Cell.Neurosci.16(2000)631–648.

[119]R.L.Cagan,H.Kramer,A.C.Hart,S.L.Zipursky,The bride of sevenless and sevenless interaction:internalization of a transmem-brane ligand,Cell69(1992)393–399.

[120]Hart,A.C.,Kramer,H.,Van,D.L.Vactor,Jr,Paidhungat,M.Zi-pursky,S.L.Induction of cell fate in the Drosophila retina:the bride of sevenless protein is predicted to contain a large extracellular domain and seven transmembrane segments.Genes Dev.4(1990) 1835–1847.

[121]J.L.Salzer,R.P.Bunge,L.Glaser,Studies of Schwann cell prolif-eration.III.Evidence for the surface localization of the neurite mitogen,J Cell Biol84(1980)767-778.

[122]T.K.Morrissey,A.Levi,A.Nuijens,M.X.Sliwkowski,R.P.Bunge, Axon-induced mitogenesis of human schwann cells involves heregulin and p185(erbb2),https://www.wendangku.net/doc/5d18999359.html,A92(1995) 1431–1435.

[123]R.Leimeroth,C.Lobsiger,A.Lussi,V.Taylor,U.Suter,L.Som-mer,Membrane-bound neuregulin1type III actively promotes Schwann cell differentiation of multipotent progenitor cells,Dev.

Biol.246(2002)245–258.

[124]J.A.Loeb,G.D.Fischbach,Aria can be released from extracellular matrix through cleavage of heparin-binding domain,J.Cell Biol.

130(1995)127–135.

[125]A.D.J.Goodearl,A.G.Yee,A.W.Sandrock,G.Corfas,G.D.Fis-chbach,Aria is concentrated in the synaptic basal lamina of the developing chick neuromuscular junction,J.Cell Biol.130(1995) 1423–1434.

[126]S.A.Jo,X.J.Zhu,M.A.Marchionni,S.J.Burden,Neuregulins are concentrated at nerve-muscle synapses and activate Ach-receptor gene expression,Nature373(1995)158–161.

[127]J.A.Loeb,T.S.Khurana,J.T.Robbins,A.G.Yee,G.D.Fischbach, Expression patterns of transmembrane and released forms of neu-regulin during spinal cord and neuromuscular synapse development, Development126(1999)781–791.

[128]Q.Li,J.A.Loeb,Neuregulin-heparan-sulfate proteoglycan interac-tions produce sustained erbB receptor activation required for the

28 D.L.Falls/Experimental Cell Research284(2003)14–30

induction of acetylcholine receptors in muscle,J.Biol.Chem.276 (2001)38068–38075.

[129]Colognato,H.,Baron,W.,Avellana-Adalid,V.,Relvas,J.B.,Ever-cooren,A.B.,Georges-Labouesse,E.,Ffrench-Constant,https://www.wendangku.net/doc/5d18999359.html,S integrins switch growth factor signalling to promote target-depen-dent survival.Nat.Cell Biol.4(2002)833–841.

[130]A.I.Flores,B.S.Mallon,T.Matsul,W.Ogawa,A.Rosenzweig,T.

Okamoto,W.B.Macklin,Akt-mediated survival of oligodendro-cytes induced by neuregulins,J.Neurosci.20(2000)7622–7630. [131]P.A.Fernandez,D.G.Tang,L.Cheng,A.Prochiantz,A.W.Mudge, M.C.Raff,Evidence that axon-derived neuregulin promotes oligo-dendrocyte survival in the developing rat optic nerve[in process citation],Neuron28(2000)81–90.

[132]P.D.Canoll,R.Kraemer,K.K.Teng,M.A.Marchionni,J.L.Salzer, GGF/neuregulin induces a phenotypic reversion of oligodendro-cytes,Mol.Cell.Neurosci.13(1999)79–94.

[133]P.D.Canoll,J.M.Musacchio,R.Hardy,R.Reynolds,M.A.Mar-chionni,J.L.Salzer,Ggf/neuregulin is a neuronal signal that pro-motes the proliferation and survival and inhibits the differentiation of oligodendrocyte progenitors,Neuron17(1996)229–243. [134]V.Calaora,B.Rogister,K.Bismuth,K.Murray,H.Brandt,P.

Leprince,M.Marchionni,M.Dubois-Dalcq,Neuregulin signaling regulates neural precursor growth and the generation of oligoden-drocytes in vitro,J.Neurosci.21(2001)4740–4751.

[135]Y.J.Ma,D.F.Hill,K.E.Creswick,M.E.Costa,A.Cornea,M.N.

Lioubin,G.D.Plowman,S.R.Ojeda,Neuregulins signaling via a glial erbB-2-erbB-4receptor complex contribute to the neuroendo-crine control of mammalian sexual development,J.Neurosci.19 (1999)9913–9927.

[136]C.Paratore,D.E.Goerich,U.Suter,M.Wegner,L.Sommer,Sur-vival and glial fate acquisition of neural crest cells are regulated by an interplay between the transcription factor Sox10and extrinsic combinatorial signaling,Development128(2001)3949–3961. [137]T.D.Camenisch,J.A.Schroeder,J.Bradley,S.E.Klewer,J.A.

McDonald,Heart-valve mesenchyme formation is dependent on hyaluronan-augmented activation of ErbB2ErbB3receptors,Nat.

Med.8(2002)850–855.

[138]S.Rentschler,J.Zander,K.Meyers,D.France,R.Levine,G.Porter, S.A.Rivkees,G.E.Morley,G.I.Fishman,Neuregulin-1promotes formation of the murine cardiac conduction system,Proc.Natl.

https://www.wendangku.net/doc/5d18999359.html,A99(2002)10464–10469.

[139]S.Rohrbach,X.Yan,E.O.Weinberg,F.Hasan,J.Bartunek,M.A.

Marchionni,B.H.Lorell,Neuregulin in cardiac hypertrophy in rats with aortic stenosis.Differential expression of erbB2and erbB4 receptors,Circulation100(1999)407–412.

[140]D.Kim,S.Chi,K.H.Lee,S.Rhee,Y.K.Kwon,C.H.Chung,H.

Kwon,M.S.Kang,Neuregulin stimulates myogenic differentiation in an autocrine manner,J.Biol.Chem.274(1999)15395–15400. [141]J.T.Trachtenberg,Fiber apoptosis in developing rat muscles is regulated by activity,neuregulin,Dev.Biol.196(1998)193–203. [142]N.V.Patel,M.J.Acarregui,J.M.Snyder,J.M.Klein,M.X.

Sliwkowski,J.A.Kern,Neuregulin-1and human epidermal growth factor receptors2and3play a role in human lung development in vitro,Am.J.Respir.Cell.Mol.Biol.22(2000)432–440. [143]K.S.Russell,D.F.Stern,P.J.Polverini,J.R.Bender,Neuregulin activation of ErbB receptors in vascular endothelium leads to an-giogenesis,Am.J.Physiol.277(1999)H2205–H2211.

[144]E.Suarez,D.Bach,J.Cadefau,M.Palacin,A.Zorzano,A.Guma,

A novel role of neuregulin in skeletal muscle.Neuregulin stimulates

glucose uptake,glucose transporter translocation,and transporter expression in muscle cells,J.Biol.Chem.276(2001)18257–18264. [145]A.L.Kierszenbaum,L.L.Tres,Primordial germ cell-somatic cell partnership:A balancing cell signaling act,Mol.Reprod.Dev.60 (2001)277–280.

[146]D.L.Falls,Neuregulin-ErbB signaling:roles in regulating neuro-transmitter receptor synthesis and development of the neuromuscu-

lar junction.In G.Adelman,B.H.Smith,(Eds,),Encyclopedia of Neuroscience,3rd Edition.(2003)Elsevier.

[147]A.Viehover,https://www.wendangku.net/doc/5d18999359.html,ler,S.K.Park,G.Fischbach,T.Vartanian, Neuregulin:an oligodendrocyte growth factor absent in active mul-tiple sclerosis lesions,Dev.Neurosci.23(2001)377–386. [148]Deleted in proof.

[149]B.Cannella,C.J.Hoban,Y.L.Gao,R.Garcia-Arenas,https://www.wendangku.net/doc/5d18999359.html,wson, M.Marchionni, D.Gwynne, C.S.Raine,The neuregulin,glial growth factor2,diminishes autoimmune demyelination and en-hances remyelination in a chronic relapsing model for multiple sclerosis,https://www.wendangku.net/doc/5d18999359.html,A95(1998)10100–10105. [150]B.Cannella,D.Pitt,M.Marchionni,C.S.Raine,Neuregulin and erbB receptor expression in normal and diseased human white mat-ter,J.Neuroimmunol.100(1999)233–242.

[151]M.A.Marchionni,B.Cannella,C.Hoban,Y.L.Gao,R.Garcia-Arenas,https://www.wendangku.net/doc/5d18999359.html,wson,E.Happel,F.Noel,P.To?lon,D.Gwynne,C.S.

Raine,Neuregulin in neuron/glial interactions in the central nervous system.GGF2diminishes autoimmune demyelination,promotes ol-igodendrocyte progenitor expansion,and enhances remyelination, Adv.Exp.Med.Biol.468(199)283–295.

[152]M.Dubois-Dalcq,K.Murray,Why are growth factors important in oligodendrocyte physiology?,Pathol.Biol.(Paris)48(2000)80–86. [153]S.K.Park,D.Solomon,T.Vartanian,Growth factor control of CNS myelination,Dev.Neurosci.23(2001)327–337.

[154]H.Yan,M.B.Bunge,P.M.Wood,G.W.Plant,Mitogenic response of adult rat olfactory ensheathing glia to four growth factors,Glia33 (2001)334–342.

[155]D.J.Bryan,A.H.Holway,K.K.Wang,A.E.Silva,D.J.Trantolo,D.

Wise,I.C.Summerhayes,In?uence of glial growth factor and Schwann cells in a bioresorbable guidance channel on peripheral nerve regeneration,Tissue Eng.6(2000)129–138.

[156]A.V.Boruch,J.J.Conners,M.Pipitone,G.Deadwyler,P.D.Storer,

G.H.Devries,K.J.Jones,Neurotrophic and migratory properties of

an olfactory ensheathing cell line,Glia33(2001)225–229. [157]J.L.Zheng,G.Frantz,A.K.Lewis,M.Sliwkowski,W.Q.Gao, Heregulin enhances regenerative proliferation in postnatal rat utric-ular sensory epithelium after ototoxic damage,J.Neurocytol.28 (1999)901–912.

[158]M.P.ter Laak, F.P.Hamers, C.J.Kirk,W.H.Gispen,rhGGF2 protects against cisplatin-induced neuropathy in the rat,J.Neurosci.

Res.60(2000)237–244.

[159]Y.Tokita,H.Keino,F.Matsui,S.Aono,H.Ishiguro,S.Higash-iyama,A.Oohira,Regulation of neuregulin expression in the injured rat brain and cultured astrocytes,J.Neurosci.21(2001)1257–1264. [160]R.Bagheri-Yarmand,R.K.Vadlamudi,R.A.Wang,J.Mendelsohn, R.Kumar,Vascular endothelial growth factor up-regulation via p21-activated kinase-1signaling regulates heregulin-beta1-mediated angiogenesis,J.Biol.Chem.275(2000)39451–39457.

[161]L.Yen,X.L.You,A.E.Al Moustafa,G.Batist,N.E.Hynes,S.

Mader,S.Meloche,M.A.Alaoui-Jamali,Heregulin selectively up-regulates vascular endothelial growth factor secretion in cancer cells and stimulates angiogenesis,Oncogene19(2000)3460–3469. [162]I.M.Krane,P.Leder,Ndf/heregulin induces persistence of terminal end buds and adenocarcinomas in the mammary glands of transgenic mice,Oncogene12(1996)1781–1788.

[163]V.R.Schelfhout,E.D.Coene,B.Delaey,S.Thys,D.L.Page,C.R.

De Potter,Pathogenesis of Paget’s disease:epidermal heregulin-alpha,motility factor,and the HER receptor family,J.Natl.Cancer Inst.92(2000)622–628.

[164]V.R.Schelfhout,E.D.Coene,B.Delaey,A.A.Waeytens,L.De Rycke,M.Deleu,C.R.De Potter,The role of heregulin-alpha as a motility factor and amphiregulin as a growth factor in wound heal-ing,J.Pathol.198(2002)523–533.

[165]L.Wang,M.A.Marchionni,R.A.Tassava,Cloning and neuronal expression of a type III newt neuregulin and rescue of denervated, nerve-dependent newt limb blastemas by rhGGF2[in process cita-tion],J.Neurobiol.43(2000)150–158.

D.L.Falls/Experimental Cell Research284(2003)14–30

[166]G.Schaefer,V.D.Fitzpatrick,M.X.Sliwkowski,Gamma-heregulin:

a novel heregulin isoform that is an autocrine growth factor for the

human breast cancer cell line,MDA-MB-175,175,Oncogene15, (1997)1385–1394.

[167]X.Liu,E.Baker,H.J.Eyre,G.R.Sutherland,M.Zhou,Gamma-heregulin:a fusion gene of DOC-4and neuregulin-1derived from a chromosome translocation,Oncogene18(1999)7110–7114. [168]E.A.Sanchez-Valdivieso,J.J.Cruz,R.Salazar,M.del Mar Abad,A.

Gomez-Alonso, A.Gomez,R.Gonzalez-Sarmiento,Gamma-heregulin has no biological signi?cance in primary breast cancer, Br.J.Cancer86(2002)1362–1363.

[169]D.L.Brutsaert,Cardiac endothelial-myocardial signaling:its role in cardiac growth,contractile performance,and rhythmicity,Physiol.

Rev.83(2003)59–115.

[170]K.D.Pruitt,K.S.Katz,H.Sicotte,D.R.Maglott,Introducing RefSeq and LocusLink:curated human genome resources at the NCBI, Trends Genet.16(2000)44–47.

[171]K.D.Pruitt,D.R.Maglott,RefSeq and LocusLink:NCBI gene-centered resources,Nucleic Acids Res.29(2001)137–140. [172]S.Hippenmeyer,N.A.Schneider,C.Birchmeier,S.J.Burden,T.M.

Jessell,S.Arber,A role for neuregulin1signaling in muscle spindle differentiation,Neuron36(2002)1035–1049.

[173]V.Prevot,C.Rio,G.J.Cho,A.Lomniczi,S.Heger,C.M.Neville, N.A.Rosenthal,S.R.Ojeda,G.Corfas,Normal female sexual de-

velopment requires neuregulin-erbB receptor signaling in hypotha-lamic astrocytes,J.Neurosci.23(2003)230–239.

[174]S.Britsch,D.E.Goerich,D.Riethmacher,R.I.Peirano,M.Rossner, K.A.Nave,C.Birchmeier,M.Wegner,The transcription factor Sox10is a key regulator of peripheral glial development,Genes Dev.15(2001)66–78.

[175]S.A.Crone,A.Negro,A.Trumpp,M.Giovannini,K.F.Lee,Co-lonic epithelial expression of ErbB2is required for postnatal main-tenance of the enteric nervous system,Neuron37(2003)29–40. [176]J.S.Cameron,L.Dryer,S.E.Dryer,beta-Neuregulin-1is required for the in vivo development of functional Ca2?-activated K?channels in parasympathetic neurons,https://www.wendangku.net/doc/5d18999359.html,A98 (2001)2832–2836.

[177]H.Toyoda-Ohno,M.Obinata,Y.Matsui,Members of the ErbB receptor tyrosine kinases are involved in germ cell development in fetal mouse gonads,Dev.Biol.215(1999)399–406.

[178]H.Noguchi,C.Sakamoto,K.Wada,T.Akamatsu,T.Uchida,A.

Tatsuguchi,H.Matsui,H.Fukui,T.Fujimori,M.Kasuga,Expres-sion of heregulin alpha,erbB2,and erbB3and their in?uences on proliferation of gastric epithelial cells,Gastroenterology117(1999) 1119–1127.

[179]S.O’Shea,K.Johnson,R.Clark,M.X.Sliwkowski,S.L.Erickson, Effects of in vivo heregulin beta1treatment in wild-type and ErbB gene-targeted mice depend on receptor levels and pregnancy,Am.J.

Pathol.158(2001)1871–1880.

30 D.L.Falls/Experimental Cell Research284(2003)14–30