2319RESEARCH ARTICLE

INTRODUCTIONEstablishing precise neuronal circuits during development isessential for the proper execution of various neural activities.Dorsal root ganglion (DRG) neurons form a circuit that relayssignals from peripheral sensory organs to the central nervoussystem. DRG neurons that transmit different sensory modalitiesproject their axons to distinct laminae in the spinal cord (Mirnicsand Koerber, 1995b; Sanes and Yamagata, 1999), and thedevelopment of these projections is a complex process (Altman andBayer, 2001). First, DRG proximal afferents travel to the spinalcord, where they bifurcate and extend ascending and descendingbranches within the prospective dorsal funiculus. Then, collateralsinvade the gray matter of the spinal cord and terminate in the targetlamina. Finally, the ascending fibers of DRG neurons within thedorsal funiculus proceed to the dorsal column nuclei of themedulla. Peripheral projections of sensory afferents have beenexamined in detail in rat embryos (Mirnics and Koerber, 1995a).First, afferent fibers exit the lumbosacral DRG at E12, and by E14they are present in the epidermis of the proximal hindlimb. Fibersoriginating from L3 to L5 reach the paw by E14.5-15, and theepidermis of the most distal toes is innervated by E16-16.5. Giventhat the development of afferent projections in the spinal cord hasbeen shown in several studies to be delayed relative to theinnervation of the hindlimb, it has been proposed that peripheralinnervation may stimulate central axon growth (Smith and Frank,1988). However, recent experiments using carbocyanine dyes argueagainst this hypothesis (Mirnics and Koerber, 1995a).

DRG neurons can be divided into three groups based on sensorymodality: nociceptive, mechanoreceptive and proprioceptive(Marmigere and Ernfors, 2007). Nociceptive afferent neuronspenetrate into the dorsal horn of the spinal cord and terminate inlaminae I and II. Mechanoreceptive neuron afferents invade the

medial gray matter of the spinal cord, then turn and enter the dorsalhorn ventrally to terminate in laminae III and IV of the dorsal horn.Proprioceptive afferents pass through the medial part of the dorsalhorn without branching and reach the ventral spinal cord, wherethey finally synapse to motoneurons (Caspary and Anderson,2003). Whereas the axon terminals of nociceptive neurons arestrictly confined to the dorsal horn and do not enter the ventralspinal cord, proprioceptive central afferents do not branch in thedorsal horn and are able to invade the ventral spinal cord (Eide andGlover, 1997). Thus, it has been suggested that repulsive andattractive cues residing in the dorsal horn and/or the ventral spinalcord guide each afferent to the proper region (Perrin et al., 2001;Sharma and Frank, 1998). However, despite some progress (Frank,2006), the mechanisms governing central afferent projections to thespinal cord remain unclear.

Neurotrophins [Ngf, Bdnf, NT-3 (Ntf3) and NT-4 (Ntf5)] play acrucial role in the differentiation, innervation pattern and survivalof sensory neurons (Lewin and Barde, 1996; Markus et al., 2002).The sensory modality of a DRG neuron is tightly correlated withTrk subtype expression (Moqrich et al., 2004; Sun et al., 2008).Nociceptive neurons express TrkA (Ntrk1) and proprioceptiveneurons express TrkC (Ntrk3). However, only a subset ofmechanoreceptive neurons expresses TrkB (Ntrk2) (Gonzalez-Martinez et al., 2004), and no histochemical marker for allmechanoreceptive neurons has yet been identified (Snider andWright, 1996).

Ret is a receptor tyrosine kinase that binds to the glial cell line-derived neurotrophic factor (GDNF) family ligands (GFLs) Gdnf,neurturin (Nrtn), artemin and persephin (Airaksinen and Saarma,2002). Recent studies of genetically modified mice havedemonstrated that GFL/Ret signaling is important for cell migration,axonal outgrowth and cell survival during peripheral nervous systemdevelopment (Fundin et al., 1999; Kramer et al., 2006a). In thesensory nervous system, Ret expression first starts in TrkA+ small-diameter neurons during embryonic development, but these Ret+

neurons downregulate TrkA expression and eventually becomeTrkA–, nonpeptidergic, isolectin-B4 (IB4)+ nociceptive DRG neuronsafter birth, while Ret– TrkA+ neurons become peptidergic nociceptiveneurons. Ret is responsible for the acquisition of several properties

Development 137, 2319-2328 (2010) doi:10.1242/dev.046995© 2010. Published by The Company of Biologists Ltd

1Ogawa Research Unit, Brain Science Institute, RIKEN, Saitama 351-0198, Japan.2Department of Neurochemistry, National Institute of Neuroscience, Kodaira, Tokyo187-8502, Japan.

*Author for correspondence (yhonma@brain.riken.jp)

Accepted 10 May 2010

SUMMARYEstablishment of connectivity between peripheral and central organs is essential for sensory processing by dorsal root ganglion(DRG) neurons. Using Ret as a marker for mechanoreceptive DRG neurons, we show that both central and peripheral projectionsof mechanoreceptive neurons are severely impaired in the absence of Ret. Death of DRG neurons in Ret-deficient mice can berescued by eliminating Bax, although their projections remain disrupted. Furthermore, ectopic expression of the Ret ligandneurturin, but not Gdnf, in the spinal cord induces aberrant projection of mechanoreceptive afferents. Our results demonstratethat Ret expression in DRG neurons is crucial for the neurturin-mediated formation of precise axonal projections in the centralnervous system.

KEY WORDS: Neuroscience, Axon guidance, Mechanoreceptive neuron, Sensory neuron, Mouse

Axonal projections of mechanoreceptive dorsal rootganglion neurons depend on RetYutaka Honma1,*, Masako Kawano1, Shinichi Kohsaka2 and Masaharu Ogawa1

DEVELO

PMENT

2320

of nonpeptidergic sensory neurons (Luo et al., 2007; Molliver et al.,1997). Since Ret-deficient mice die perinatally due to renal agenesis,conditional deletion of Ret in sensory neurons has been performedto analyze the role of Ret in nonpeptidergic nociceptive neurondevelopment (Luo et al., 2007). Loss of Ret affects only theperipheral, and not the central, projections of nonpeptidergicnociceptive neurons. In addition, the number of cells expressingGfr1 and Gfr2 is significantly reduced in conditional Ret-nullmutants, whereas nonpeptidergic neuronal loss is not observed. Thus,Ret signaling regulates the expression of these Gfr co-receptorsduring nonpeptidergic nociceptive neuron development, withoutaffecting cell survival in these neuronal populations.

Furthermore, the expression of Ret in small TrkA+ nociceptiveneurons is regulated by the neurotrophic factor Ngf, which instructsa subset of TrkA+ sensory neurons to adopt a nonpeptidergicneuronal fate by suppressing TrkA expression (Luo et al., 2007).Several experiments have provided evidence that Ret is alsoexpressed in large-diameter DRG neurons from early in DRGdevelopment (Kramer et al., 2006b; Luo et al., 2007), and Ret+

fibers have been observed in the deep dorsal horn. However, thephysiological role of Ret in large-diameter DRG neurons is notwell understood (Ernsberger, 2008). In order to determine whetherRet signaling plays a role in large-diameter neurons, we examinedRet mutant mice in which tau-GFP was knocked into the Ret locus.We found that Ret is required for the projections ofmechanoreceptive neurons during development.

MATERIALS AND METHODSAnimalsThe generation of RetGFP/GFP and Gfra3–/– mice was described previously(Enomoto et al., 2001; Honma et al., 2002). Bax-deficient mice werepurchased from Jackson Laboratory (Knudson et al., 1995). Nrtn-GFPtransgenic mice were obtained from Mutant Mouse Regional ResourceCenters (Gong et al., 2003). This study was carried out in accordance withthe Guide for the Care and Use of Laboratory Animals from the Societyfor Neuroscience and was authorized by the Animal Care and UseCommittee of RIKEN.

In utero electroporation The introduction of DNA into E12.5 spinal cord was performed by in uteroelectroporation. The expression vector (1-5 g/l) was mixed with 0.05%Fast Green as a tracer and injected into the central canal of the embryothrough the uterine wall. After injection, electrodes (CUY 650-5, NepaGene) were placed on both sides of the embryo and electroporation wasperformed using a square-pulse electroporator CUY21 EDIT (Nepa Gene).

AntibodiesThe following antibodies were used: mouse anti-Isl1 (DSHB); rabbit anti-cleaved caspase 3 and rabbit anti-Aml1 (Runx1) (Cell Signaling); rabbitanti-Runx3 (Osaki et al., 2004); goat anti-Gfr1, goat anti-Gfr2, goatanti-Ret, goat anti-TrkB, goat anti-TrkC and goat anti-neurturin (R&D);rabbit anti-Ret (IBL); rabbit anti-TrkA, rabbit anti-NF-200, rabbit anti-peripherin and guinea-pig anti-vGlut1 (Millipore); mouse anti-cytokeratin20 and rabbit anti-S100 (DAKO); rabbit anti-PGP9.5 (Uchl1) (Lab Vision);rabbit anti-GFP (MBL and Invitrogen); rat anti-GFP (Nacalai); and chickanti-GFP (Aves Labs).

Histological and statistical analysesTo compare the phenotype of central afferents between Ret+/GFP andRetGFP/GFP mice, lower cervical level and upper thoracic level spinal cordswere prepared. To examine the Ret+ central and peripheral projection, 10-15 mice for each genotype at E12.5-18.5 were used, except where notedotherwise in the figure legends and experimental procedures.

To perform whole-mount immunostaining, we cut the skin of the leftupper forelimb in half; the radial side was stained with NF-200 antibody,while the ulnar side was stained with peripherin antibody.

To determine the percentage of neurons expressing Isl1, GFP, Gfra1 andGfra2 per ganglion, we prepared five adjacent sets of 12-m sections fromL4 and L5 DRG from three E14.5 embryos per group, and then stained eachseparately with an antibody or riboprobe. To compare the number of neuronslabeled with each marker, Isl1 staining was used to determine the totalnumber of L4 and L5 DRG neurons at E14.5. To quantitate the number ofneurons undergoing apoptosis, we prepared five adjacent sets of 12-msections, and the number of cells labeled with GFP and caspase 3 antibodieswas counted. The number of neurons expressing each marker in controls wasset to 100%. To investigate the cell size distribution, L3-L5 DRG of E18.5embryos were collected and dissociated with trypsin (n3). After plating andstaining with GFP antibody, the cell size distribution of the mean soma areafor GFP+ neurons was examined. To quantitate central and peripheralprojections, we prepared five adjacent sets of 16-m sections from the entireseries of axial spinal cord (corresponding to C5-C6 DRG levels) or of theforelimb. The positive fibers were framed and quantitated from ImageJ(version 1.41, NIH). We set the fluorescence intensity in controls to 100%.n3 for each genotype. To evaluate the attractive effect of ectopic Nrtnquantitatively, the lower edge of the dorsal horn to the upper edge of thecentral canal of spinal cord was divided horizontally into three bins, asindicated in Fig. 7. The average pixel intensity of fibers in each bin wasdetermined using ImageJ. Twenty sections from five embryos electroporatedwith Nrtn were used for this analysis.

Data are presented as mean ± s.e.m. and were compared using Student’st-test.

RESULTSTwo distinct types of Ret+ DRG neurons duringdevelopmentTo characterize Ret expression in the sensory nervous system, weused mice in which tau-GFP was knocked into the Ret locus(RetGFP/GFP) to map the projections of Ret+ DRG neurons in vivo(Enomoto et al., 2001). Consistent with previous data, the GFPreporter accurately reflected the Ret expression pattern. To assesswhich population of early embryonic DRG neurons expresses Retand to examine the central and peripheral axonal projectionpatterns of Ret+ neurons, we examined GFP expression in anextensive array of sensory neuron targets.

DRG neurons began expressing GFP at ~E11.5 (see Fig. S1A,Bin the supplementary material). Most of these GFP+ neurons hadlarger somata than TrkA+ small-diameter neurons. Few, if any,small-diameter DRG neurons, which express TrkA, were GFP+

even at E14.5 (see Fig. S1C in the supplementary material).However, by E18.5, the number of GFP+ small-diameter neuronshad increased, even though the percentage of GFP+ large-diameterDRG neurons did not change during embryonic development(E13.5-18.5), remaining at ~7-10% of Isl1+ neurons. This suggeststhat there are two distinct populations of Ret+ DRG neurons(Kramer et al., 2006b). One is a small population of large-diameterneurons that begin expressing Ret at E11.5, the percentage ofwhich, relative to the total number of DRG neurons, is stablethroughout embryonic development. The other populationcomprises small-diameter neurons that emerge at ~E13.5, expressTrkA (see Fig. S1D in the supplementary material), and graduallybecome the major population of Ret+ DRG neurons at later stagesof embryonic development.

Development of Ret+ central afferents in thespinal cordWe carefully examined GFP+ axonal projections in the spinal cordduring early embryonic stages (E12.5-16.5) in order to understandthe behavior of central afferent projections of Ret+ large-diameterneurons. The axon collaterals of GFP+ DRG neurons began topenetrate the spinal cord gray matter at E12.5 (Fig. 1C). At E12.5,

RESEARCH ARTICLE Development 137 (14)

DEVELO

PMENT

the oval bundle of His, which is the primordium of the dorsalfuniculus formed by GFP+ fibers, was distinct from both TrkA+ andTrkC+ oval bundles, suggesting that Ret+ afferents form a uniquepath within the dorsal funiculus (Fig. 1A,B). Central afferentsprojected laterally underneath the deep dorsal horn at E14.5 (Fig.1H), and finally terminated in the deep dorsal horn with upwardendings at E16.5 (Fig. 1J). GFP+ terminal plexuses were alsoobserved in the medial part of the gray matter (Fig. 1J). Even atpostnatal day (P) 6, GFP-labeled Ret+ axon collaterals were foundin the deep dorsal horn that were labeled with vesicular glutamatetransporter 1 (vGlut1; Slc17a7 – Mouse Genome Informatics),which marks the central afferents that transmit low-thresholdmechanoreceptive sensory inputs to the deep dorsal horn of thespinal cord (Todd et al., 2003; Yoshida et al., 2006) (Fig. 1L,M).We also observed weak GFP expression in the superficial layer ofthe dorsal horn (lamina II), which probably reflects labeling of theterminals of IB4+ Ret+ neurons. This signal grew more intense afterP7 (see Fig. S1E in the supplementary material).

Mechanoreceptive afferents terminating in thedeep dorsal horn are severely disturbed inRet-deficient miceTo determine whether loss of Ret affects the development of themechanoreceptive sensory neurons that normally express Retduring embryonic development, we examined the phenotype ofGFP+ fibers in RetGFP/GFP mice. GFP+ DRG neurons were detectedin RetGFP/GFP and Ret+/GFP mice at E11.5 (see Fig. S1A,B in the

supplementary material). At E12.5, the first GFP+ axonalprojections were observed, but no GFP+ axon collaterals wereobserved in the spinal cord gray matter in RetGFP/GFP mice (Fig.1D). At E13.5, GFP+ axon collaterals were found in the gray matterof RetGFP/GFP mice (Fig. 1G), but were few in number comparedwith those in Ret+/GFP mice (Fig. 1E-G). Although the number ofaxon collaterals invading the spinal cord increased from E13.5-14.5in RetGFP/GFP mice (Fig. 1G,I), the fibers proceeding to the medullawithin the dorsal funiculus were sparse at E14.5 (Fig. 1I). Theseresults suggest that there are fewer distal DRG central axons inRetGFP/GFP mice than in control mice. Furthermore, the axoncollaterals of RetGFP/GFP mice appeared to lose their directionwithin the spinal cord (Fig. 1I), whereas those of Ret+/GFP miceextended laterally along the developing dorsal horn with properdirectionality (Fig. 1H). Finally, GFP+ axon collaterals were almostentirely excluded from the deep dorsal horn in RetGFP/GFP mice atE16.5 (Fig. 1K), suggesting that Ret-ablated neurons might die dueto their inability to project axons to the proper targets. Collectively,these results indicate that the loss of Ret in large-diameter DRGneurons delays their central afferent projections and impairsdirectional axonal extension within the spinal cord.

We next examined whether all mechanoreceptive afferentsterminating in the deep dorsal horn are affected in RetGFP/GFP mice.We first performed DiI tracing to label all centrally projecting DRGafferents. DiI labeling at E16.5 revealed severe fiber loss in thedeep dorsal horn, despite a lack of obvious abnormalities inproprioceptive axonal projections to the ventral spinal cord (Fig.

2321RESEARCH ARTICLERet function in mechanoreceptive neurons

Fig. 1. Characterization of Ret+ central afferents in the spinal cord, which are aberrant in Ret mutants. (A,B)The oval bundle of Ret+ fibers(green in A) encircles the TrkC+ oval bundle (red in A) and differs from TrkA+ fibers (blue in B). (C,D)Central projections of Ret+ neurons are impairedin RetGFP/GFP mice. GFP+ axon collaterals invade the developing dorsal horn normally at E12.5 in RetGFP/GFP mice, but fail to penetrate into the spinalcord (D). Furthermore, GFP+ fibers are sparse within the oval bundle in RetGFP/GFP mice (D), as compared with those in Ret+/GFP mice (C). (F-K)Centralafferents invading the spinal cord of RetGFP/GFP mice are increased at E13.5 (G), but still sparse compared with those in Ret+/GFP mice (F). Axoncollaterals in Ret+/GFP mice project laterally (H), but those of RetGFP/GFP mice appear to lose their direction at E14.5 (I). Collaterals finally terminate inthe deep dorsal horn with upward endings at E16.5 in Ret+/GFP mice (J), but these fibers are severely impaired in RetGFP/GFP mice (K). Verticallyrunning fibers projecting to the medulla within the dorsal funiculus in Ret+/GFP mice (arrow in H) are severely impaired in RetGFP/GFP mice at E14.5(arrow in I). (L,M)Central afferents projecting into the deep dorsal horn are apparent even at P6 (L), and these afferents are also labeled withvGlut1, a marker for the deep dorsal horn (M). (E)Central afferent projections are significantly reduced in RetGFP/GFP mice at E12.5 and E13.5. n3for each group. *P<0.05; **P<0.01 (Student’s t-test). Scale bars: 50m.

DEVELO

PMENT

2322

2A,B). We then investigated the expression of vGlut1 because itmarks the central afferents that transmit low-thresholdmechanoreceptive sensory inputs to the deep dorsal horn of thespinal cord. We found that vGlut1 expression was exclusivelyeliminated from laminae III-IV, but was retained in the intermediateand ventral gray matter of spinal cord at E16.5 in Ret-deficientmice (Fig. 2C,D). Thus, all mechanoreceptive sensory afferents arelost from the deep dorsal horn in the absence of Ret. Consistentwith the preservation of vGlut1 expression in the intermediate andventral spinal cord in RetGFP/GFP mice, parvalbumin+

proprioceptive afferents were not affected (Fig. 2E,F). In addition,TrkA+ cutaneous nociceptive afferents in laminae I-II werecomparable in RetGFP/GFP and Ret+/GFP mice (Fig. 2G,H). We alsofound similar abnormalities at E18.5 in the deep dorsal horn ofRetGFP/GFP mice (Fig. 2I-N). These results indicate that Ret isexclusively required for mechanoreceptive neuron terminals in thedeep dorsal horn, and that loss of these terminals does not result ininvasion of proprioceptive and nociceptive afferents into the deepdorsal horn.

Since laminar structure could influence the lamina-specifictermination of DRG neurons, we examined the laminar structure ofthe dorsal horn using lamina-specific molecular markers, includingEbf1, Drg11 (Prrxl1 – Mouse Genome Informatics), Brn3a(Pou4f1) and c-Maf (Li et al., 2006). At E14.5, expression of thesemolecules in dorsal horn was normal in RetGFP/GFP mice,suggesting that the dorsal horn structure is not impaired by the lackof Ret (see Fig. S2A-H in the supplementary material), despiteexpression of Ret in the developing dorsal horn (see Fig. S2I in thesupplementary material). Thus, it is unlikely that the defects inmechanoreceptive central afferent projections are due to aberrantdorsal horn formation.

Ret+ fibers innervate mechanoreceptors in theperiphery and are severely impaired inRet-deficient miceNext, we examined GFP-labeled peripheral projections in Ret+/GFP

mice. Because we still observed GFP immunostaining in the deepdorsal horn, as described above (Fig. 1L,M), it seemed likely thatGFP staining marks early large-diameter Ret+ DRG sensory neuronfibers during early postnatal stages. Thus, we examined GFP+ axonterminals in the forelimb skin at P3-P6. GFP+ fibers with lanceolateor Ruffini endings were found around hair follicles (Fig. 3A).Furthermore, we found GFP+ fibers innervating several types ofmechanoreceptors, including Meissner corpuscles (Fig. 3D),Merkel cells (Fig. 3E) and Pacinian corpuscles (Fig. 3F) in the skinand the crural interosseous membrane. Thus, Ret+ DRG neuronsproject into the deep dorsal horn (laminae III-IV) and innervatemechanoreceptors in the skin, suggesting that large-diameter Ret+

DRG neurons are mechanoreceptive neurons.To explore whether Ret deficiency also influences peripheral

projections, we examined GFP+ fibers in the body wall skin atE12.5 and in the forelimb skin at E18.5 in RetGFP/GFP mice. GFP+

fibers were observed in the body wall at E12.5 (data not shown)and were abundant around hair follicles in forelimb skin at E18.5(Fig. 3G) in Ret+/GFP mice. These fibers were severely reduced inRetGFP/GFP mice (Fig. 3G), suggesting that Ret is required for boththe central and peripheral projections of these neurons.

Since Ret+ peripheral afferents are defined as mechanoreceptivefibers, which are usually myelinated, we next explored whetherloss of Ret affects NF-200+ fibers innervating the skin. Asexpected, NF-200 (Nefh – Mouse Genome Informatics)immunostaining, which marks prospective myelinated axons(Zylka et al., 2005), was significantly reduced in the skin of E18.5RetGFP/GFP as compared with Ret+/GFP mice (Fig. 3H). However,these axons were not completely lost, consistent with theobservation of a few GFP+ fibers in RetGFP/GFP skin (Fig. 3G). Bycontrast, fibers that are positive for peripherin, a marker forprospective unmyelinated axons (Kosaras et al., 2009; Lariviere etal., 2002), were unaffected in RetGFP/GFP mice (Fig. 3I), consistentwith the idea that Ret+ neurons are myelinated mechanoreceptiveneurons, although the staining specificity of NF-200 and peripherinis somewhat controversial (Jackman and Fitzgerald, 2000). Toexamine the cutaneous branch of the sensory nerve in Ret-deficientmice, we performed whole-mount NF-200 and peripherinimmunostaining of the dorsal skin of the left upper forelimb atE14.5. NF-200+ fibers in RetGFP/GFP mice were thinner than thosein Ret+/GFP mice, but their branching appeared to be intact (Fig.3J,K). By contrast, peripherin+ fibers were not impaired inRetGFP/GFP mice (Fig. 3L,M). Thus, our data show that theperipheral projections of NF-200+ sensory fibers are severelyreduced in RetGFP/GFP mice.

RESEARCH ARTICLE Development 137 (14)

Fig. 2. The selective loss of central afferents ofmechanoreceptive neurons in the deep dorsal horn. (A-D)DiIlabeling and vGlut1 immunostaining in E16.5 mice revealed thatmechanoreceptive central afferents projecting into the deep dorsal hornare exclusively eliminated in RetGFP/GFP mice (arrowheads).(E-H)Parvalbumin+ proprioceptive afferents and TrkA+ nociceptiveafferents are identical in RetGFP/GFP and Ret+/GFP mice. (I-N)Doubleimmunostaining for GFP and TrkA, vGlut1 and parvalbumin wasperformed at E18.5. GFP+ fibers are impaired (J,L,N) and doublestaining of GFP with vGlut1 is severely reduced in RetGFP/GFP mice(arrowheads in I,J). Punctate GFP staining reflects Ret+ cells in the deepdorsal horn (J,L,N). Scale bars: 50m.

DEVELO

PMENT

The role of Nrtn-Gfr2 in Ret+ mechanoreceptiveneuronsGFL signaling requires a receptor complex comprising Ret and aGfr co-receptor. Since the effects of Ret deficiency on theexpression of Gfr co-receptors have only been examined duringthe later stages of embryonic development (Luo et al., 2007), weexamined co-receptor expression in DRG neurons during earlyembryogenesis in Ret mutant mice. Because we did not observeany abnormalities in central and peripheral Ret+ DRG afferents in

Ret+/GFP; Gfra3–/– mice (see Fig. S3A-D in the supplementarymaterial), we focused on Gfr1 and Gfr2. Antibodies againsteither Gfr1 or Gfr2 and GFP were used to double label DRGneurons. At E13.5, Gfr2+ cells (93±1%) were GFP+ large-diameter neurons (Fig. 4D-F), whereas Gfr1+ cells (30±2.8%)were GFP+ small-diameter neurons (Fig. 4A-C). Thus, duringembryonic development, Gfr2 is the best candidate for a Ret co-receptor in large-diameter DRG neurons.

Because Gfr2 is a known co-receptor for Nrtn (Heuckeroth etal., 1999; Rossi et al., 1999), we next examined Nrtn expressionduring embryonic development. Since Nrtn expression is scarcelydetectable by in situ hybridization, we used transgenic micecarrying GFP under the control of the Nrtn promoter in order tofollow Nrtn expression during development. We found Nrtnexpression in the DRG and in peripherally projecting axons atE12.5 (Fig. 4J). At E13.5, Nrtn was expressed in the dorsal rootentry zone and in the path of the central afferents (Fig. 4K). Wealso detected signal in the prospective dorsal funiculus and deepdorsal horn at E13.5 (Fig. 4M). Although Nrtn expression in theproximal afferents and DRG became weak at E14.5 (data notshown), expression in the deep dorsal horn was clearly evident atE14.5 (Fig. 4O). The signal in the deep dorsal horn was scatteredat E13.5 (Fig. 4M), but was concentrated on the lateral side of thedorsal horn at E14.5 (Fig. 4O). The Nrtn signal appeared to precedethe GFP+ fiber projections in the deep dorsal horn of Ret+/GFP mice(Fig. 4L-O). Finally, the Nrtn expression level in the dorsal hornwas decreased by E16.5 (data not shown). At E15.5, Nrtn was alsoexpressed in the periosteum and ligaments of the developingforelimb at E12.5 (data not shown) and around the hair follicles, apotential target of mechanoreceptive neurons (Fig. 4P).Intriguingly, we observed that Ret+ central afferents terminated inthe deep dorsal horn in chick embryos (Fig. 4Q) and that Nrtnexpression was detected in the region where those fibers project(Fig. 4R), as in the mouse (Fig. 4O). Moreover, chick Nrtnexpression was also seen around the hair follicles, as in the mouse(Fig. 4P,S), suggesting that the mechanism by whichmechanoreceptive neuron projections find their path might beevolutionary conserved. These observations support the idea thatNrtn is involved in the development of the peripheral and centralprojections of mechanoreceptive neurons via Ret and Gfr2.

Characterization of DRG neuron abnormalities inRet-deficient miceSeveral lines of evidence indicate that functionally distinct DRGneurons express specific neurotrophic factor receptors, and that thisdiversification underlies the segregation of DRG neurons accordingto sensory modality (Woolf and Ma, 2007). To assess whether theloss of Ret influences the segregation of neurotrophic factorreceptor expression in DRG neurons, we performed double-labeling experiments at E13.5 with an anti-GFP antibody andantibodies against different neurotrophic factor receptors. Therewas little overlap between GFP expression and neurotrophic factorreceptor expression in the DRG neurons of E13.5 Ret+/GFP mice(Fig. 5C,E,G), in agreement with a previous report (Kramer et al.,2006b). The GFP and Trk receptor expression patterns inRetGFP/GFP mice were almost identical to those in Ret+/GFP mice(Fig. 5C-H), suggesting that Ret signaling is not involved in thesegregation of Trk receptor expression. Furthermore, we found thatthere was little overlap between GFP expression and that of Runx1(a marker for nociceptive neuron) or Runx3 and parvalbumin(markers for proprioceptive neurons) in Ret+/GFP mice (Patel et al.,2003; Sun et al., 2008; Yoshikawa et al., 2007), and the expression

2323RESEARCH ARTICLERet function in mechanoreceptive neurons

Fig. 3. Characterization of Ret+ peripheral afferents, which areimpaired in Ret-deficient mice. (A)Ret+ fibers with lanceolate andRuffini endings innervate hair follicles in hairy skin at P3. (B,C)MostRet+ fibers are positive for neurofilament 200 (NF-200), a myelinatedaxon marker (B), but not for peripherin, an unmyelinated axon marker(C). (D-F)Ret+ peripheral fibers also innervate Meissner corpuscles(positive for PGP9.5) (D), Merkel cells [positive for cytokeratin 20(Krt20)] (E) and Pacinian corpuscles (S100+) (F) in P6 peripheral tissues.(G-I)At E18.5, GFP+ and NF-200+ axons in the skin are severelyimpaired in RetGFP/GFP mice (G,H), but peripherin+ axons are spared (I).Quantitative analysis revealed significant reduction of GFP+ and NF-200+ peripheral projections in Ret-deficient mice at E18.5 (G,H),whereas peripherin+ fibers are intact (I). (J-M)Whole-mountimmunohistochemistry for NF-200 and peripherin in the upper forelimbskin at E14.5 revealed that NF-200+ fibers are dense in Ret+/GFP mice(arrowheads in J), but severely impaired in RetGFP/GFP mice (arrowhead inK). By contrast, peripherin+ fibers are unaffected even in RetGFP/GFP mice(L,M). n3 for each group. **P<0.01 (Student’s t-test). Scale bars:50m.

DEVELO

PMENT

2324

pattern in RetGFP/GFP mice was comparable to that in Ret+/GFP mice(Fig. 5I-N). These observations indicate that Ret is not involved inthe segregation of DRG neurons with respect to sensory modality.

Because GFP+ afferent projections were impaired during earlyembryogenesis in Ret mutants, we examined whether Ret functionsin mechanoreceptive neurons during early embryonic development.We first tested whether impaired central and peripheral DRGprojections in Ret-deficient early embryos affect sensory neurondevelopment. In situ hybridization revealed loss of Gfra2expression in the DRG of E14.5 RetGFP/GFP mice (Fig. 6C,D),whereas Gfra1 expression was identical to that of controls (Fig.6A,B). We then quantitatively examined the number of Isl1+,GFP+, Gfra1-expressing and Gfra2-expressing neurons (Fig. 6E).Isl1 labeling was used to evaluate the total number of DRGneurons. We detected a 37% reduction in the number of GFP+

neurons in Ret-deficient embryos at E14.5, even though thenumber of Isl+ neurons was not significantly reduced. The smallsubset of GFP+ DRG neurons at E14.5 (7% of Isl1+ neurons) mightexplain this discrepancy. The number of Gfra2-expressing neuronswas also markedly reduced in RetGFP/GFP embryos (by 44%,compared with Ret+/GFP embryos), whereas the number of Gfra1-expressing neurons was unchanged. We also observed a GFP+

neuron reduction in RetGFP/GFP mice at E18.5 (by 23%, comparedwith Ret+/GFP embryos). In addition, when we examined the meansoma area of GFP+ neurons, we found that the number of large-diameter GFP+ neurons in RetGFP/GFP mice at E18.5 was reduced(Fig. 5A,B). Cells with small somata also appeared to be increased

in number, which might reflect the observation that mostperipherin+ neurons are hypotrophic in Ret-deficient mice, aspreviously described (Luo et al., 2007). Collectively, these resultsindicate that large-diameter sensory neurons are severely reducedin the DRG of Ret-deficient mice.

Bax deficiency rescues loss of GFP+ andGfra2-expressing neurons, but fails to rescuemechanoreceptive neuron projection defectsIn order to determine whether the reduction in the number ofneurons expressing GFP or Gfra2 was caused by cell loss or bytranscriptional downregulation, we used an antibody againstcleaved caspase 3 to measure cell death. The number of labeledneurons in DRG of Ret+/GFP and RetGFP/GFP mice was counted atE13.5, before the stage at which we observed the massive reductionin GFP+ and Gfra2-expressing neurons (see above). There weremore cleaved caspase 3+ cells in RetGFP/GFP than in Ret+/GFP mice(121% of control) (Fig. 6F), suggesting that cell death isresponsible for the reduction in Ret+ DRG neurons at E14.5.

The axonal defects we observed might simply reflect DRGneuron loss in the absence of Ret. To eliminate the influence of celldeath, we generated mice that were double null for Ret and Bax. InBax knockout mice, cell death is virtually eliminated in peripheralganglia, including the DRG (White et al., 1998). The number ofGFP+ and Gfra2-expressing neurons in the DRG was normal inRetGFP/GFP; Bax–/– mice (Fig. 6G), suggesting that Bax-mediatedprogrammed cell death is involved in neuronal loss during

RESEARCH ARTICLE Development 137 (14)

Fig. 4. Role of Gfr2-Nrtn in Ret+ mechanoreceptive neuron development. (A-F)Double immunolabeling for GFP (B,E) and Gfr1 (A) orGfr2 (D) in E13.5 DRG sections. Most GFP and Gfr1 double-positive neurons have small somata (C), whereas almost all GFP and Gfr2 double-positive neurons have large somata (F). Notably, most Gfr2+ cells coexpress GFP (F), whereas many Gfr1+ cells do not (C). (G-I)GFP+ cells werestained for Isl1. (J-M)At E12.5, Nrtn is detected on the path of peripherally projecting sensory afferents (arrowheads in J). At E13.5, Nrtn is alsoexpressed in proximal axons (arrow in K) and the dorsal root entry zone of DRG neurons (arrowhead in K), as well as in the primordium of thedorsal funiculus (arrowhead in L,M) and deep dorsal horn (arrow in M). (N-P)At E14.5, Nrtn is expressed in the deep dorsal horn of the spinal cord(arrowhead in O), where Ret+ axon collaterals project (arrowhead in N). In the peripheral tissue, Nrtn is expressed in the hair follicles, which aretarget tissues of Ret+ mechanoreceptive neurons at E15.5 (P). (Q-S)Similar to mouse Nrtn, chick Nrtn (cNRTN) expression is also found in the deepdorsal horn (arrowhead in R), where Ret+ fibers innervate (arrowhead in Q), as well as in hair follicles (S). Scale bars: 50m.

DEVELO

PMENT

embryonic development of Ret-deficient DRG neurons.Remarkably, although the number of neurons was restored byknocking out Bax in RetGFP/GFP mice (Fig. 6G), the central andperipheral axonal projections of mechanoreceptive neurons werestill impaired (69% and 89% reduction, respectively, as comparedwith Ret+/GFP; Bax–/– mice) (Fig. 6H-M). Taken together, these datasuggest that cell death is not the direct cause of the axonal defects.Instead, the axonal projections themselves are dependent on Ret inmechanoreceptive neurons.

Ectopic Nrtn attracts Ret+ DRG central afferents toan aberrant region of the spinal cordGiven that Ret-ablated mechanoreceptive neurons in RetGFP/GFP

mice fail to extend axons in the proper direction and to projectinto the deep dorsal horn, we hypothesized that the Ret signalmight be involved in the attraction of mechanoreceptive afferentsto the deep dorsal horn of the spinal cord. To test this hypothesisdirectly, we introduced an expression vector for the Gfr2 ligandNrtn into Ret+/GFP mouse spinal cords by in utero electroporationat E12.5, and harvested the embryos 3 days later to determinewhether Ret+ fibers were attracted to exogenously expressedNrtn. Consistent with our hypothesis, GFP+ fibers were attractedto the region of ectopic Nrtn expression (Fig. 7A-D). By contrast,neither Gdnf nor DsRed attracted GFP+ afferents (Fig. 7H; datanot shown). TrkA+ nociceptive afferents and parvalbumin+

proprioceptive afferents in spinal cord did not respond to ectopicNrtn expression (Fig. 7F,G), indicating that mechanoreceptivecentral fibers are the only afferents that are able to respond toNrtn. Moreover, Nrtn had no attractive effect on central afferentsin Ret-deficient mice (Fig. 7I), indicating that the attractiveeffect of Nrtn is Ret dependent.

DISCUSSIONRet expression has been observed in large-diameter DRG neurons,but the sensory modality of these Ret+ large-diameter neurons wasunclear (Ernsberger, 2008). Here, we show that Ret+ DRG neuronssend peripheral axons to mechanoreceptors in the skin and centralaxon collaterals to the deep dorsal horn in the spinal cord. Theseresults indicate that Ret+ large-diameter neurons aremechanoreceptive neurons. Moreover, we found that Ret isrequired for proper laminar termination of mechanoreceptiveneuron central projections in the spinal cord. The peripheralprojections of mechanoreceptive neurons are also severelydisturbed in the absence of Ret. Finally, we show that Nrtn, a Retligand, is expressed in the region where Ret+ neurons send theiraxons, and that ectopic Nrtn attracts Ret+ fibers to an aberrantregion of the spinal cord. These results demonstrate that Retsignaling is required for mechanoreceptive neurons to establish thecircuit between peripheral and central organs, and also suggest thatcentral afferent growth within the spinal cord is independentlyregulated from the signal for peripheral projections.

Neurotrophic factors for mechanoreceptiveneuronsNeurotrophins (i.e. Bdnf, NT-3 and NT-4) play a developmentalrole in the survival, function and axonal projections ofmechanoreceptive DRG neurons (Airaksinen et al., 1996; Cronk etal., 2002; Stucky et al., 1998). However, neurotrophins have onlybeen implicated in the postnatal development of these neurons(Carroll et al., 1998), and no trophic factors involved in theirdevelopment during embryonic stages have thus far been described.Our data show that Ret is essential for the development ofmechanoreceptive neuron axonal projections during

2325RESEARCH ARTICLERet function in mechanoreceptive neurons

Fig. 5. DRG neuron abnormalities in Ret-deficient mice. (A,B)The number of GFP+ large-diameter neurons is dramatically reduced in RetGFP/GFP

mice at E18.5. Cell size distribution analysis using dissociated DRG neurons also revealed severe reduction of large-diameter neurons in RetGFP/GFP

mice. (C-N)Double labeling for GFP (green) and the proteins indicated (red) in E14.5 DRG sections from control Ret+/GFP (C,E,G,I,K,M) and RetGFP/GFP

(D,F,H,J,L,N) mice. A small number of GFP+ neurons from control Ret+/GFP and RetGFP/GFP mice express one of the Trk proteins. Notably, the number ofGFP+ DRG neurons was reduced in RetGFP/GFP mice (C-H). Staining with other molecular markers for nociceptive (Runx1) and proprioceptive (Runx3and parvalbumin) neurons also revealed no DRG neuron segregation abnormalities in RetGFP/GFP mice (I-N). Scale bar: 50m.

DEVELO

PMENT

2326

embryogenesis. Previous work showed that D-hairmechanoreceptors switch their neurotrophin dependence from NT-3 to NT-4 during postnatal development, even after DRG neuronsare terminally differentiated (Stucky et al., 2002). It thereforeappears that DRG neurons change their dependence on trophicfactors several times before maturation.

It has been suggested that a subset of mechanoreceptive neuronsexpress TrkB (Kramer et al., 2006b). Is there a correlation betweenRet and TrkB in mechanoreceptive neuron development beforebirth? In TrkB–/– mice, central afferents terminating in theintermediate zone of the spinal cord, where myelinatedmechanoreceptors innervate, are severely reduced, although thecentral afferents in the deep dorsal horn are intact (Silos-Santiagoet al., 1997). By contrast, central afferents terminating in theintermediate zone are not impaired in RetGFP/GFP mice. Togetherwith the fact that TrkB+ neuron number is not reduced inRetGFP/GFP mice, this suggests that TrkB-dependentmechanoreceptive neurons are distinct from Ret-dependentmechanoreceptive neurons. It is possible that mechanoreceptiveneurons that project to the intermediate gray matter of the spinalcord depend on TrkB, but are not Ret-dependent for theirinnervation, at least before birth.

RESEARCH ARTICLE Development 137 (14)

Fig. 7. Attractive effect of Nrtn on Ret+ central afferents in thespinal cord. (A-C)Ectopic Nrtn-induced aberrant extension of Ret+

afferents (arrowheads in A). Although a slight attraction of Ret+ fiberson the control side was observed (Bin 2), ectopic Nrtn significantlyinduced Ret+ fiber projection compared with the non-electroporatedcontrol side (Bin 3). Horizontal black dashed lines in B indicate the linesused to divide bins 1-3, as quantified in C. n5. *P<0.01 (Student’s t-test). (D,E)Higher magnification views also reveal the attractive effectof ectopic Nrtn on GFP+ fibers in the spinal cord (arrowheads). Doublestaining for GFP with DsRed , as superimposed in E, shows the GFP+

fibers winding around DsRed+ cells expressing Nrtn. (F,G)TrkA+ andparvalbumin+ central afferents do not respond to ectopic Nrtn. (H)Gdnfdoes not have any attractive effect on Ret+ afferents. (I)GFP+ fibers arenot attracted to ectopic Nrtn in Ret-deficient mice. EP, electroporated.Scale bars: 50m.

Fig. 6. Cell death of large-diameter neurons in Ret-deficient mice.(A-D)DRG neurons from E14.5 control Ret+/GFP and RetGFP/GFP micewere hybridized with Gfra1 and Gfra2 in situ probes. Representativesections show a reduction in Gfra2, but not Gfra1, expression inRetGFP/GFP mice compared with control mice. (E)The percentage of Isl1+,GFP+, Gfra1-expressing and Gfra2-expressing DRG neurons wasquantitatively analyzed at E14.5. The number of positive cells for eachmarker in control Ret+/GFP mice was set to 100%. n3 for each group.*P<0.05; **P<0.01 (Student’s t-test). (F)Quantitative analysis of thepercentage of cleaved caspase 3+ cells in E13.5 Ret+/GFP and RetGFP/GFP

mice shows an increase in apoptotic cells in the homozygous mutant.The number of positive cells for each marker in Ret+/GFP mice was set to100%. n4 for each group. **P<0.01 (Student’s t-test).(G)Quantitative analysis of the percentage of Isl1+, GFP+, Gfra1-expressing and Gfra2-expressing DRG neurons in E14.5 Ret+/GFP; Bax–/–

and RetGFP/GFP; Bax+/– and RetGFP/GFP; Bax–/– mice. Bax deficiency rescuedthe number of DRG neurons expressing GFP or Gfra2. n3 for eachgroup. **P<0.01 (Student’s t-test). (H-K)Ret+ fibers in the deep dorsalhorn and skin are severely impaired in E16.5 RetGFP/GFP; Bax+/– micecompared with control Ret+/GFP; Bax–/– mice. (L,M)Bax deficiency failedto rescue the axonal defect in the deep dorsal horn and skin of micelacking Ret. Scale bars: 50m.

DEVELO

PMENT

The role of Nrtn in the survival ofmechanoreceptive neuronsRet signaling is dispensable for the viability of nonpeptidergicDRG neurons in vivo (Luo et al., 2007). By contrast,mechanoreceptive neurons undergo cell death in Ret-deficient miceat E13.5. In mechanoreceptive neurons, Ret may function in thesurvival of Ret+ and Gfr2+ neurons, as Bax deficiency rescued thenumber of cells expressing GFP and Gfr2 co-receptors in Ret-deficient mice. However, we could not conclude whether or notNrtn is the survival factor for mechanoreceptive neurons becausea small number of GFP+ Gfra2-expressing neurons survive inRetGFP/GFP mice. Moreover, a few GFP+ fibers are already found inthe peripheral tissues and spinal cord of RetGFP/GFP mice at E12.5.These observations support the idea that once GFP+ neurons reachthe skin, a target-derived trophic factor other than Nrtn promotestheir survival. Indeed, GFLs fail to support embryonic DRGsensory neuron survival in the absence of Ngf in vitro, despite theexpression of Ret and its co-receptors in DRG neurons (Baudet etal., 2000). It is likely that Ret signaling is required only for axonalprojections, but not for the survival of mechanoreceptive neurons.However, it is still possible that other molecules, such asneurotrophins, induce mechanoreceptive neuron axonal growtheven before birth, similar to what occurs in postnatal development.Thus, Nrtn might be involved in both mechanoreceptive neuronsurvival and axonal projections.

Nrtn is required for the axonal projections of Ret+

mechanoreceptive neuronsDRG neurons that convey different sensory modalities send axonsinto distinct laminae within the spinal cord, but the molecularmechanisms underlying lamina-specific projection patterns arelargely unknown. Proprioceptive central afferents are disrupted inNt3-null mice (Ernfors et al., 1994), and NT-3 attracts axons inDRG explant cultures in vitro (Genc et al., 2004). We found thatmechanoreceptive neurons require Ret to establish their terminalprojections in the deep dorsal horn. Ectopic Nrtn induces aberrantaxonal extension of mechanoreceptive neurons via Ret in vivo. Nrtnis expressed in the dorsal horn, suggesting that Nrtn acts as adiffusible guidance cue for mechanoreceptive afferent neurons.Thus, we propose that a diffusible guidance cue exists for each classof DRG neuron that attracts afferents in the spinal cord. However,when ectopic Nrtn is introduced into the superficial dorsal horn byelectroporation, Ret+ afferents fail to invade the region (ourpreliminary observation). This raises the possibility that repulsivecues in the superficial lamina of the dorsal horn prevent Ret+ fibersfrom invading the region. Sema3a repels Ngf-responsive axons, buthas little effect on NT-3-responsive axons (Messersmith et al.,1995), suggesting that a repulsive cue might also exist for each classof DRG neuron. Mechanoreceptive afferents take a unique path intothe spinal cord, which might indicate that they use their ownrepulsion molecules to terminate in the deep dorsal horn. It will beinteresting to explore which repulsive cues act in concert with theattractive Nrtn cue to guide mechanoreceptive afferent projectionsand to reveal how precise neural circuits are established duringDRG development. Repulsive cues for mechanoreceptive afferentsremain to be identified, but histochemical Ret labeling should allowmechanoreceptive afferent behavior within the spinal cord to betraced. In the periphery, we have observed Nrtn expression alongthe peripheral sensory afferents and in their target tissues, but notalong the ventral motor efferents. These findings suggest that Nrtnis required for the peripheral afferents to find their target organs, asis the case for the central projections. The generation of mice that

ectopically express Nrtn in the spinal cord and/or the peripheraltissues might allow us to better understand the mechanism by whichNrtn regulates the projections of mechanoreceptive neurons.

Ret signaling in pathological conditionsPrevious studies have shown that Gfr2 is the main co-receptor ofRet in nonpeptidergic nociceptive neurons (Luo et al., 2007). Thus,Gfr2 appears to play a major role in both nonpeptidergicnociceptive and mechanoreceptive Ret+ sensory neurons. In somepathological conditions, non-nociceptive stimuli evoke pain; forexample, in allodynia, tactile stimuli cause pain (Sandkuhler, 2009).It is hypothesized that mechanoreceptive neurons undergo anabnormal phenotypic switch to nociceptive neurons in allodynia, andthat non-nociceptive stimuli might evoke pain sensation throughmechanoreceptive neurons. Although ligand-receptor binding maynormally activate different downstream targets in different cell typesand therefore elicit distinct actions, under pathological conditions itis possible that this distinction is lost. Thus, one intriguing hypothesisis that Ret-Gfr2 signaling is partly responsible for this phenotypicswitch of mechanoreceptive neurons to nociceptive neurons inallodynia.

AcknowledgementsWe thank Dr Jeffrey Milbrandt for providing Ret and Gfra3 knockout mice; DrMitsuhiko Osaki for providing Runx3 antibody; Drs Masato Hoshi and KazuyoKamitori for their technical support; Dr Takayoshi Inoue for providing the pCAexpression vector for in utero electroporation; Drs Tomomi Shimogori andToshiyuki Araki for their comments on the manuscript; and Dr Toshio Ohshimaand members of the M.O. laboratory for their helpful discussions.

Competing interests statementThe authors declare no competing financial interests.

Supplementary materialSupplementary material for this article is available athttp://dev.biologists.org/lookup/suppl/doi:10.1242/dev.046995/-/DC1

ReferencesAiraksinen, M. S. and Saarma, M. (2002). The GDNF family: signalling, biological

functions and therapeutic value. Nat. Rev. Neurosci. 3, 383-394.Airaksinen, M. S., Koltzenburg, M., Lewin, G. R., Masu, Y., Helbig, C., Wolf,

E., Brem, G., Toyka, K. V., Thoenen, H. and Meyer, M. (1996). Specificsubtypes of cutaneous mechanoreceptors require neurotrophin-3 followingperipheral target innervation. Neuron 16, 287-295.

Altman, J. and Bayer, S. A. (2001). Development of the Human Spinal Cord. NewYork: Oxford University Press.

Baudet, C., Mikaels, A., Westphal, H., Johansen, J., Johansen, T. E. andErnfors, P. (2000). Positive and negative interactions of GDNF, NTN and ART indeveloping sensory neuron subpopulations, and their collaboration withneurotrophins. Development 127, 4335-4344.

Carroll, P., Lewin, G. R., Koltzenburg, M., Toyka, K. V. and Thoenen, H.(1998). A role for BDNF in mechanosensation. Nat. Neurosci. 1, 42-46.

Caspary, T. and Anderson, K. V. (2003). Patterning cell types in the dorsal spinalcord: what the mouse mutants say. Nat. Rev. Neurosci. 4, 289-297.

Cronk, K. M., Wilkinson, G. A., Grimes, R., Wheeler, E. F., Jhaveri, S., Fundin,B. T., Silos-Santiago, I., Tessarollo, L., Reichardt, L. F. and Rice, F. L. (2002).Diverse dependencies of developing Merkel innervation on the trkA and bothfull-length and truncated isoforms of trkC. Development 129, 3739-3750.

Eide, A. L. and Glover, J. C. (1997). Developmental dynamics of functionallyspecific primary sensory afferent projections in the chicken embryo. Anat.Embryol. (Berl.) 195, 237-250.

Enomoto, H., Crawford, P. A., Gorodinsky, A., Heuckeroth, R. O., Johnson, E.M., Jr and Milbrandt, J. (2001). RET signaling is essential for migration, axonalgrowth and axon guidance of developing sympathetic neurons. Development128, 3963-3974.

Ernfors, P., Lee, K. F., Kucera, J. and Jaenisch, R. (1994). Lack of neurotrophin-3leads to deficiencies in the peripheral nervous system and loss of limbproprioceptive afferents. Cell 77, 503-512.

Ernsberger, U. (2008). The role of GDNF family ligand signalling in thedifferentiation of sympathetic and dorsal root ganglion neurons. Cell Tissue Res.333, 353-371.

Frank, E. (2006). Axon guidance in the spinal cord: choosin’ by exclusion. Neuron52, 745-746.

2327RESEARCH ARTICLERet function in mechanoreceptive neurons

DEVELO

PMENT

2328

Fundin, B. T., Mikaels, A., Westphal, H. and Ernfors, P. (1999). A rapid anddynamic regulation of GDNF-family ligands and receptors correlate with thedevelopmental dependency of cutaneous sensory innervation. Development126, 2597-2610.

Genc, B., Ozdinler, P. H., Mendoza, A. E. and Erzurumlu, R. S. (2004). Achemoattractant role for NT-3 in proprioceptive axon guidance. PLoS Biol. 2,e403.

Gong, S., Zheng, C., Doughty, M. L., Losos, K., Didkovsky, N., Schambra, U.B., Nowak, N. J., Joyner, A., Leblanc, G., Hatten, M. E. et al. (2003). A geneexpression atlas of the central nervous system based on bacterial artificialchromosomes. Nature 425, 917-925.

Gonzalez-Martinez, T., Germana, G. P., Monjil, D. F., Silos-Santiago, I., deCarlos, F., Germana, G., Cobo, J. and Vega, J. A. (2004). Absence of Meissnercorpuscles in the digital pads of mice lacking functional TrkB. Brain Res. 1002,120-128.

Heuckeroth, R. O., Enomoto, H., Grider, J. R., Golden, J. P., Hanke, J. A.,Jackman, A., Molliver, D. C., Bardgett, M. E., Snider, W. D., Johnson, E. M.,Jr et al. (1999). Gene targeting reveals a critical role for neurturin in thedevelopment and maintenance of enteric, sensory, and parasympatheticneurons. Neuron 22, 253-263.

Honma, Y., Araki, T., Gianino, S., Bruce, A., Heuckeroth, R., Johnson, E. andMilbrandt, J. (2002). Artemin is a vascular-derived neurotropic factor fordeveloping sympathetic neurons. Neuron 35, 267-282.

Jackman, A. and Fitzgerald, M. (2000). Development of peripheral hindlimb andcentral spinal cord innervation by subpopulations of dorsal root ganglion cells inthe embryonic rat. J. Comp. Neurol. 418, 281-298.

Knudson, C. M., Tung, K. S., Tourtellotte, W. G., Brown, G. A. andKorsmeyer, S. J. (1995). Bax-deficient mice with lymphoid hyperplasia and malegerm cell death. Science 270, 96-99.

Kosaras, B., Jakubowski, M., Kainz, V. and Burstein, R. (2009). Sensoryinnervation of the calvarial bones of the mouse. J. Comp. Neurol. 515, 331-348.

Kramer, E. R., Knott, L., Su, F., Dessaud, E., Krull, C. E., Helmbacher, F. andKlein, R. (2006a). Cooperation between GDNF/Ret and ephrinA/EphA4 signalsfor motor-axon pathway selection in the limb. Neuron 50, 35-47.

Kramer, I., Sigrist, M., de Nooij, J. C., Taniuchi, I., Jessell, T. M. and Arber, S.(2006b). A role for Runx transcription factor signaling in dorsal root ganglionsensory neuron diversification. Neuron 49, 379-393.

Lariviere, R. C., Nguyen, M. D., Ribeiro-da-Silva, A. and Julien, J. P. (2002).Reduced number of unmyelinated sensory axons in peripherin null mice. J.Neurochem. 81, 525-532.

Lewin, G. R. and Barde, Y. A. (1996). Physiology of the neurotrophins. Annu.Rev. Neurosci. 19, 289-317.

Li, M. Z., Wang, J. S., Jiang, D. J., Xiang, C. X., Wang, F. Y., Zhang, K. H.,Williams, P. R. and Chen, Z. F. (2006). Molecular mapping of developing dorsalhorn-enriched genes by microarray and dorsal/ventral subtractive screening. Dev.Biol. 292, 555-564.

Luo, W., Wickramasinghe, S. R., Savitt, J. M., Griffin, J. W., Dawson, T. M.and Ginty, D. D. (2007). A hierarchical NGF signaling cascade controls Ret-dependent and Ret-independent events during development of nonpeptidergicDRG neurons. Neuron 54, 739-754.

Markus, A., Patel, T. D. and Snider, W. D. (2002). Neurotrophic factors andaxonal growth. Curr. Opin. Neurobiol. 12, 523-531.

Marmigere, F. and Ernfors, P. (2007). Specification and connectivity of neuronalsubtypes in the sensory lineage. Nat. Rev. Neurosci. 8, 114-127.

Messersmith, E. K., Leonardo, E. D., Shatz, C. J., Tessier-Lavigne, M.,Goodman, C. S. and Kolodkin, A. L. (1995). Semaphorin III can function as aselective chemorepellent to pattern sensory projections in the spinal cord.Neuron 14, 949-959.

Mirnics, K. and Koerber, H. R. (1995a). Prenatal development of rat primaryafferent fibers: I. Peripheral projections. J. Comp. Neurol. 355, 589-600.

Mirnics, K. and Koerber, H. R. (1995b). Prenatal development of rat primaryafferent fibers: II. Central projections. J. Comp. Neurol. 355, 601-614.

Molliver, D. C., Wright, D. E., Leitner, M. L., Parsadanian, A. S., Doster, K.,Wen, D., Yan, Q. and Snider, W. D. (1997). IB4-binding DRG neurons switchfrom NGF to GDNF dependence in early postnatal life. Neuron 19, 849-861.

Moqrich, A., Earley, T. J., Watson, J., Andahazy, M., Backus, C., Martin-Zanca, D., Wright, D. E., Reichardt, L. F. and Patapoutian, A. (2004).Expressing TrkC from the TrkA locus causes a subset of dorsal root ganglianeurons to switch fate. Nat. Neurosci. 7, 812-818.

Osaki, M., Moriyama, M., Adachi, K., Nakada, C., Takeda, A., Inoue, Y.,Adachi, H., Sato, K., Oshimura, M. and Ito, H. (2004). Expression of RUNX3protein in human gastric mucosa, intestinal metaplasia and carcinoma. Eur. J.Clin. Invest. 34, 605-612.

Patel, T. D., Kramer, I., Kucera, J., Niederkofler, V., Jessell, T. M., Arber, S.and Snider, W. D. (2003). Peripheral NT3 signaling is required for ETS proteinexpression and central patterning of proprioceptive sensory afferents. Neuron38, 403-416.

Perrin, F. E., Rathjen, F. G. and Stoeckli, E. T. (2001). Distinct subpopulations ofsensory afferents require F11 or axonin-1 for growth to their target layers withinthe spinal cord of the chick. Neuron 30, 707-723.

Rossi, J., Luukko, K., Poteryaev, D., Laurikainen, A., Sun, Y. F., Laakso, T.,Eerikainen, S., Tuominen, R., Lakso, M., Rauvala, H. et al. (1999). Retardedgrowth and deficits in the enteric and parasympathetic nervous system in micelacking GFR alpha2, a functional neurturin receptor. Neuron 22, 243-252.

Sandkuhler, J. (2009). Models and mechanisms of hyperalgesia and allodynia.Physiol. Rev. 89, 707-758.

Sanes, J. R. and Yamagata, M. (1999). Formation of lamina-specific synapticconnections. Curr. Opin. Neurobiol. 9, 79-87.

Sharma, K. and Frank, E. (1998). Sensory axons are guided by local cues in thedeveloping dorsal spinal cord. Development 125, 635-643.

Silos-Santiago, I., Fagan, A. M., Garber, M., Fritzsch, B. and Barbacid, M.(1997). Severe sensory deficits but normal CNS development in newborn micelacking TrkB and TrkC tyrosine protein kinase receptors. Eur. J. Neurosci. 9, 2045-2056.

Smith, C. L. and Frank, E. (1988). Specificity of sensory projections to the spinalcord during development in bullfrogs. J. Comp. Neurol. 269, 96-108.

Snider, W. D. and Wright, D. E. (1996). Neurotrophins cause a new sensation.Neuron 16, 229-232.

Stucky, C. L., DeChiara, T., Lindsay, R. M., Yancopoulos, G. D. andKoltzenburg, M. (1998). Neurotrophin 4 is required for the survival of asubclass of hair follicle receptors. J. Neurosci. 18, 7040-7046.

Stucky, C. L., Shin, J. B. and Lewin, G. R. (2002). Neurotrophin-4: a survivalfactor for adult sensory neurons. Curr. Biol. 12, 1401-1404.

Sun, Y., Dykes, I. M., Liang, X., Eng, S. R., Evans, S. M. and Turner, E. E.(2008). A central role for Islet1 in sensory neuron development linking sensoryand spinal gene regulatory programs. Nat. Neurosci. 11, 1283-1293.

Todd, A. J., Hughes, D. I., Polgar, E., Nagy, G. G., Mackie, M., Ottersen, O. P.and Maxwell, D. J. (2003). The expression of vesicular glutamate transportersVGLUT1 and VGLUT2 in neurochemically defined axonal populations in the ratspinal cord with emphasis on the dorsal horn. Eur. J. Neurosci. 17, 13-27.

White, F. A., Keller-Peck, C. R., Knudson, C. M., Korsmeyer, S. J. and Snider,W. D. (1998). Widespread elimination of naturally occurring neuronal death inBax-deficient mice. J. Neurosci. 18, 1428-1439.

Woolf, C. J. and Ma, Q. (2007). Nociceptors-noxious stimulus detectors. Neuron55, 353-364.

Yoshida, Y., Han, B., Mendelsohn, M. and Jessell, T. M. (2006). PlexinA1signaling directs the segregation of proprioceptive sensory axons in thedeveloping spinal cord. Neuron 52, 775-788.

Yoshikawa, M., Senzaki, K., Yokomizo, T., Takahashi, S., Ozaki, S. and Shiga,T. (2007). Runx1 selectively regulates cell fate specification and axonalprojections of dorsal root ganglion neurons. Dev. Biol. 303, 663-774.

Zylka, M. J., Rice, F. L. and Anderson, D. J. (2005). Topographically distinctepidermal nociceptive circuits revealed by axonal tracers targeted to Mrgprd.Neuron 45, 17-25.

RESEARCH ARTICLE Development 137 (14)

DEVELO

PMENT