ONTOGENY, REGIONAL AND CELLULAR DISTRIBUTION OF THENOVEL METALLOPROTEASE NEPRILYSIN 2 IN THE RAT: ACOMPARISON WITH NEPRILYSIN AND ENDOTHELIN-CONVERTINGENZYME-1

P. FACCHINETTI, C. ROSE, J. C. SCHWARTZ ANDT. OUIMET*

Unite de Neurobiologie et Pharmacologie Moleculaire (U573) del’INSERM, Centre Paul Broca, 2ter rue d’Alesia, 75014 Paris, France

Abstract—The localisation of the gene transcripts of a re-cently discovered peptidase, neprilysin 2 (NEP2), was estab-lished by in situ hybridisation in rat tissues during develop-ment and adulthood. It was compared with those of neprilysin(NEP), a closely related enzyme in terms of sequence homol-ogy or substrate specificity, and of endothelin-convertingenzyme 1 (ECE-1) which, like the other two, belongs to theM-13 sub-family of zinc-dependent metallopeptidases.

The ontogeny of the three enzymes differed markedly, theexpression of NEP2 being restricted to developing and dif-ferentiating fields of the CNS, whereas NEP and ECE-1 geneswere broadly expressed early on in the CNS and periphery. Incontrast to the wide expression of NEP and ECE-1 in periph-eral adult tissues and in CNS, NEP2 was almost exclusivelyexpressed in selected neuronal populations of the brain andspinal cord. The only exceptions were the intermediate andanterior lobes of the pituitary as well as the choroid plexuses,where NEP2 was also strongly expressed. These localisa-tions as well as those in the hypothalamic nuclei, togetherwith the previously established pattern of cleaved peptides,suggest the involvement of NEP2 in the metabolism of neu-rohormones of the hypothalamo-pituitary axis.

Complementary distributions of NEP and NEP2 mRNAswere observed in a large number of brain areas with, forinstance the former being highly expressed in the striatum inwhich NEP2 transcripts were almost undetectable. In con-trast, NEP2 was highly expressed in numerous thalamic, hy-pothalamic and brainstem nuclei from which NEP was ab-sent. Since both peptidases are able to cleave the sameneuropeptides, this pattern may suggest a complementaryrole in their peptide inactivation functions in the CNS. Finally,ECE-1 mRNAs were generally observed in neuronal popula-tions known to express the pre-proendothelin-1 gene, con-firming the function of the metallopeptidase in endothelin-1generation. © 2003 IBRO. Published by Elsevier Science Ltd.All rights reserved.

Key words: M13 sub-family, enkephalinase, peptide-inactivation, mRNA, in situ hybridisation, endothelin.

A restricted number of peptidases have yet been trulyidentified as physiologically involved in the metabolism ofendogenous neuronal or hormonal peptides, despite theirpotentially critical role in the biosynthesis or inactivation ofthese extracellular messenger molecules. Amongst these,neprilysin (NEP, enkephalinase, EC 3.4.24.11), for whichthe term neuropeptidase was coined (Schwartz, 1983), isthe best known in terms of molecular properties, localisa-tion, pharmacology and function (see Turner et al., 1996,2001; Roques et al., 1993 for review). This ectopeptidasewas first isolated from kidney brush-border membranes(Kerr and Kenny, 1974) and subsequently identified inbrain as an enzyme involved in the termination of enkepha-linergic signals (Malfroy et al., 1978; Roques et al., 1980;Patey et al., 1981). However, the substrate specificity ofthe enzyme determined in vitro is rather large, as it isequally able to cleave and thusly inactivate tachykinins(Matsas et al., 1983), neurotensin (Kitabgi et al., 1992),natriuretic peptides (Kenny and Stephenson, 1988;Schwartz et al., 1990) and cholecystokinin-octapeptide(CCK-8, Matsas et al., 1984). In the latter case, however,NEP was shown not to be involved in the inactivation ofendogenous CCK-8 (Zuzel et al., 1985), whilst anotherpeptidase was demonstrated to play such role (Rose et al.,1996), presumably in relationship to its appropriate local-isation in the vicinity of CCK-releasing terminals (Facchi-netti et al., 1999). Furthermore, a detailed immunohisto-chemical mapping of NEP in rat brain, in comparison withthat of alleged neuropeptide substrates, has brought tolight several discrepancies (Pollard et al., 1989), therebyillustrating the idea that i) topographical and not only bio-chemical factors have to be taken into account to clarify thefunctional roles of a given peptidase and ii) other pepti-dases probably remain to be characterised as candidatesin the metabolism of neuronal and hormonal peptides.

Recently, a novel metallopeptidase of the M13 sub-family was cloned, which displays a high degree of func-tional similarities with NEP, thusly justifying its designationas neprilysin 2 (NEP2) (Ouimet et al., 2000; Rose et al.,2002) or NEP-like peptidase-1 (Ghaddar et al., 2000).These metalloproteases are type II membrane-bound gly-coproteins usually expressed at the cell surface whereNEP, for instance, exerts its peptide-inactivating function.NEP2 is unique in that it is coded for by a single gene withmultiple splice variants (Ouimet et al., 2000), the two mostabundant isoforms resulting in either a membrane-boundor secreted protein which are both synthesised as inactive

*Corresponding author. Tel: �33-01-4078-9280; fax: �33-01-4580-7293.E-mail address: ouimet@broca.inserm.fr (T. Ouimet).Abbreviations: CCK, cholecystokinin; E, embryonic day; ECE-1, endo-thelin-converting enzyme 1; ET-1, endothelin-1; GnRH, gonadotropin-releasing hormone; NEP, neprilysin; NEP2, neprilysin 2; PBS, phos-phate-buffered saline.

Neuroscience 118 (2003) 627–639

0306-4522/03$30.00�0.00 © 2003 IBRO. Published by Elsevier Science Ltd. All rights reserved.doi:10.1016/S0306-4522(02)01002-3

627

forms requiring migration out of the endoplasmic reticulumin order to become active (Rose et al., 2002). RecombinantNEP2 isoforms were shown to cleave the same neuropep-tides as NEP, sometimes with closely similar specificityconstants and inhibitory patterns (Rose et al., 2002), sug-gesting that NEP2 may, in fact, possess some functionspreviously attributed to NEP. Northern blot analysis andpreliminary in situ hybridisation studies have disclosed thatthe NEP and NEP2 genes display distinct expression pat-terns (Ouimet et al., 2000). These considerations haveobvious therapeutic implications, as recently shown withthe introduction of NEP inhibitors in gastroenterologic(Baumer et al., 1992; Salazar-Lindo et al., 2000) and car-diovascular fields (Bralet and Schwartz, 2001).

In the present work, we have established a detaileddistribution of NEP2 mRNAs in the developing and adultrat and compared it with that of NEP which has alreadybeen described, albeit not in a detailed manner (Gaudouxet al., 1993; Wilcox et al., 1989). We have also comparedit to the expression pattern of endothelin-converting en-zyme 1 (ECE-1, 3.4.24.71, Xu et al., 1994), another relatedmember of the M13 sub-family presumably responsible forthe generation of endothelin-1 (ET-1) by cleavage of itsinactive precursor “Big-ET1,” present in a discrete popula-tion of central neurons.

EXPERIMENTAL PROCEDURES

Probe synthesis

The single-stranded complementary RNA (cRNA) probes for NEP,ECE-1 and NEP2 have already been described in Ouimet et al.(2000). The enolase probe consisted of subcloned PCR fragmentof 300 base pairs (nucleotide 1302–1602, AC M11931, Sakimuraet al., 1985) into pGEM-4Z (Promega, France). The cRNA probeswere synthesised using a Riboprobe kit (Promega, France) with[33P]UTP or a mixture of digoxigenin-UTP and non-labelled UTPaccording to a standard protocol (Diaz et al., 1995).

Tissue processing

All experiments were performed on Wistar rats (Iffa Credo,France). Rats were mated on embryonic day (E) 0 and the follow-ing day was considered as the first E (E1). On days E14, E16,E18, E20 and E22, four dams were anaesthetised with pentobar-bital and embryos were surgically removed. Five adult male ratsweighing 180–200 g were killed by decapitation; brains and pe-ripheral organs were surgically removed. Organs and embryoswere immediately frozen by immersion in liquid monochlorodiflu-oromethane and kept at �80 °C until use. Cryostat sections (10–12 �m) were thaw-mounted on RNase-free slides (Superfrostplus). Sections were fixed for 40 min at 4 °C in 4% paraformalde-hyde dissolved in 0.1-M phosphate-buffered saline pH 7.4 (PBS)and rinsed twice for 5 min in 0.1-M PBS at 4 °C followed by 5 minin PBS at room temperature. They were dehydrated through agraded series of ethanol baths (30%, 60%, 95% for 30 s each, and100% for 2 min), dried under a stream of cold air and stored at�80 °C until in situ hybridisation experiments. All experimentalprocedures met the guidelines of the European Community’s di-rective 86/609/EEC. The number of killed animals was kept to aminimum and every care was taken to reduce their suffering.

In situ hybridisation histochemistry

The day of the experiment, slides were thawed and treated withproteinase K (1 �g/ml) for 10 min at 37 °C, rinsed for 1–2 min in

diethylpyrocarbonate-treated distilled water and then in 0.1-M tri-ethanolamine–HCl buffer, pH8, at room temperature, followed bya treatment of 10 min in 0.25% acetic anhydride in triethanolaminebuffer. The sections were rinsed twice for 3 min in 2� SSC (1�SSC�150-mM NaCl, 15-mM sodium citrate) and dehydrated asdescribed above. Cryostat sections were incubated with either aNEP, NEP2 or ECE-1 33P-labelled cRNA antisense or senseprobe according to a standard protocol (Diaz et al., 1995). Simul-taneous detection of NEP2 and enolase mRNA within the samesection was performed using a cocktail of digoxigenin- and 33P-labelled cRNA probes. The labelled probes were denatured byheating at 70 °C, quickly cooled on slushy ice and added to thehybridisation buffer consisting of 50% deionised formamide, 10%dextran sulphate, 1� Denhardt’s solution, 2� SSC, 0.1% sodiumpyrophosphate, 100 �g/ml yeast tRNA and 100 �g/ml denaturedsalmon-sperm DNA. Each section was covered with 50 �l ofhybridisation buffer containing a final radiolabelled probe concen-tration of 1�106 cpm or with 20 ng of the digoxigenin-labelledprobe mixed with 1�106 cpm radiolabelled probe. Sections werecovered with Nescofilm (Roth, France) and incubated overnight at58 °C in a moist chamber.

The following day, nescofilms were removed by rinsing in 2XSSC at room temperature. Sections were treated at 37 °C for 40min with RNase A (200 �g in 500 �l of 10-mM Tris–HCl containing0.5-M NaCl). Slides were washed twice in 2� SSC (20 min each)at room temperature, once in 0.5� SSC (30 min) at 55 °C andthen in 0.1� SSC (30 min) at 60 °C, followed by a final wash in0.1� SSC for 20 min at room temperature.

For autoradiographic detection, films (Hyperfilm �-max, Am-ersham Biosciences, France) were apposed to sections previ-ously dehydrated through a graded series of ethanol baths con-taining 300-mM ammonium acetate and then dried with a streamof cold air. Autoradiograms were developed after 20, 16 and 26days, for NEP, ECE-1 and NEP2 respectively. In the case ofautoradiography on emulsion, slides were dipped in LM-1 (Amer-sham Biosciences, France) melted at 43 °C, air-dried for at least3 h and stored in the dark, in plastic slide boxes at 4 °C, for 31, 24and 40 days for NEP, ECE-1 and NEP2 respectively. The slideswere developed in D-19 developer, fixed in 30% sodium thiosul-fate and after rinsing in distilled water for 1 h, they were counter-stained with Mayer’s haemalun-eosin, dehydrated, coverslippedwith Permount and observed with a photomicroscope (Zeiss).

Simultaneous detection of digoxigenin-labelled enolase andradioactive cRNA NEP2 probes within the same section wasaccomplished by incubating the slides in 2% normal sheep serumand 0.3% Triton X-100 in buffer A (100-mM Tris–HCl pH 7.5;150-mM NaCl) for 30 min at room temperature, before overnightincubation at 4 °C with an anti-digoxigenin antibody (conjugated toalkaline phosphatase, Roche Biochemicals, France) diluted 1:100in buffer A (supplemented with 1% normal sheep serum and 0.3%Triton X-100). Sections were washed three times in buffer A for 10min and then washed in buffer B (100-mM Tris–HCl; 100-mMNaCl; 50-mM MgCl2) for 10 min. The conjugated antibody wasvisualised by incubating the sections in a chromogen solution(buffer B containing 340 �g/ml nitroblue tetrazolium chloride,175 �g/ml bromo-4-chloro-3 indolyl phosphate and 240 �g/mllevamisole). After washing in buffer B, each slide was coveredwith 500 �l of the chromogen solution and placed in humidifiedplastic boxes. After completion of the reaction (1–3 h at roomtemperature and in the dark), sections were washed in buffer C(10-mM Tris–HCl, pH 8; 1-mM EDTA) for 30 min. They wererinsed in distilled water for 30 min, briefly in 70% ethanol and thendried with a stream of cool air for 1 h. Subsequently, the slideswere dipped in Ilford K5 photographic emulsion diluted 1:1 withdistilled water, air dried for 3 h and stored in the dark at 4 °C for 40days. The emulsion was developed in D19 developer, rinsed indeionised water, fixed in 30% sodium thiosulfate, rinsed again indeionised water for 1 h and the slides were coverslipped using a

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639628

water-soluble mounting medium. Single- or double-labelled cellswere visualised with bright- and dark-field illumination using aZeiss photomicroscope. In every experiment, control slides wereincubated with the corresponding sense probes, which yielded nosignal and are, therefore, not presented.

RESULTS

The study of the localisation of NEP2 transcripts in the ratwas performed by in situ hybridisation on whole-body sec-tions during development and on sections of adult tissues.Since no NEP2 signal was observed in tissues of theperiphery other than in the testis, which has been previ-ously described (Ouimet et al., 2000), we focussed onresults regarding the CNS in the adult.

Comparative distributions of NEP2, NEP and ECE-1mRNAs during embryonic development

A limited expression of NEP2 gene was observed through-out development as reflected by the results observed ateither E14 or E22, which contrasted with the widespreadexpression of the NEP and ECE-1 genes (Fig. 1). At alldevelopmental stages, NEP2 mRNA signals (faint) wereonly detected in brain, nasal epithelium (Fig. 1) and retina.At E14, NEP2 transcripts in brain were restricted to thehypothalamic differentiating field (Fig. 1A, B).

NEP gene expression in CNS at E14 was restricted tothe mesencephalic flexure, more specifically to the teg-mental neuroepithelium and to the neuroepithelium of theisthmus in the rhombencephalic vesicle (Fig. 1C). HighNEP expression was also apparent in a number of periph-eral tissues, e.g. mesenchymatous condensations fromwhich bones of the skull roof originate, vertebrae or hind-limb anlage, as well as in vibrissae or heart (Fig. 1C).

ECE-1 transcripts at E14 (Fig. 1D) were detected in thewhole neuroepithelium with a high density of labellingaround cerebral vesicles and in endothelial cells of all largeblood vessels. In the coelomic cavity, all endocardial cellswere densely labelled and hepatocytes expressed ECE-1at a higher level than in all other tissues.

At E22, NEP2 mRNAs (Fig. 1E) were only detected inthe head, namely in several sensory organs, such as retina(in ganglion cell and neuroblastic layers, data not shown)and nasal epithelium (in olfactory receptor neurons). Afaint labelling was detected throughout the CNS with thehighest level in brainstem (not shown). Other organs suchas spinal ganglia (Fig. 1H) remained unlabelled. In con-trast, NEP mRNAs displayed a broad distribution at E22(Fig. 1F) which was already apparent from E16. In brain,the olfactory bulbs, cerebral cortex and striatal areas werelabelled. Several sensory organs such as parts of the nasalepithelium, the vibrissae and the auditory ducts exhibitedhigh levels of NEP expression. The bones of the maxillaand the mandible were homogeneously labelled. In theregion of the neck, the cervical brown fat pads were lightlylabelled and a strong labelling appeared in the salivaryglands. In the trunk, the lowest labelling was found in theheart. In contrast, the lungs as well as the kidneys, theadrenal gland and the terminal portions of the intestineexhibited high levels of expression. The epiphyses of the

vertebrae were surrounded by a thin rim of labelling. Inspinal ganglia (Fig. 1I), the signal was restricted to largeneurons.

ECE-1 transcripts at E22 (Fig. 1G) were evenly ex-pressed throughout the head and the body, according to apattern already apparent from E16 (as for NEP). The largeblood vessels were prominently labelled. In the brain, sig-nals were observed in the olfactory bulbs, cerebral cortex,thalamus and hippocampus. In the head, a strong signalwas found in the nasal epithelium, vibrissae, skin andtongue epithelium. At the level of the neck, the cervicalbrown fat pads, salivary glands and thymus also exhibitedan intense labelling. In the thoracic and abdominal cavitiesseveral structures displayed various degrees of expres-sion: a faint labelling was detected in kidney, heart, oe-sophagus and stomach, whereas the adrenals werestrongly labelled with the highest expression localised inthe lung, intestine and liver. All neuronal cells of the spinalganglia also exhibited an intense labelling (Fig. 1J).

Comparative distributions of NEP2, NEP and ECE-1in adult rat CNS

Whereas the NEP and ECE-1 gene transcripts were ob-served in a large number of adult rat tissues, NEP2 mR-NAs displayed a restricted distribution (Figs. 2, 3, 4, 5, 6and Table 1). NEP2 expression was observed in brain,pituitary, spinal cord (Figs. 2–5), retina (Fig. 6) and testis(not shown). The NEP2 gene transcripts in brain weredistributed in a markedly heterogeneous manner, display-ing a somewhat increasing gradient according to a rostro-caudal axis as illustrated on a sagittal section (Fig. 2). Thisgene was mainly expressed in neurons, as evidenced byits co-localisation with enolase, a specific marker for neu-ronal cells (Sakimura et al., 1985; Fig. 3F). However,NEP2 mRNAs (like NEP and ECE-1 mRNAs) were alsopresent in choroid plexus and ependymal cells (Fig. 4), aswell as in the anterior and intermediate lobes of the pitu-itary (Figs. 2–4).

In most cases, gene transcripts of the three metal-lopeptidases displayed a clearly distinct distributionamongst brain areas, with limited overlap (Table 1). Whilstthe neocortex was (slightly) labelled with the three probesin a similar laminar fashion (with laminae V and VI beingthe most clearly labelled), the cingulate, retrosplenial,perirhinal and piriform cortices were unequalled for theirlabelling with the NEP2 probe (Figs. 2, 3A and B, 4).Whereas the hippocampus expressed NEP2, NEP andECE-1 in the same neuronal layers (namely the pyramidalcell layer) of CA1, CA2 and CA3 (Fig. 4D, G, J; Fig. 4E, H,K; Fig. 4F, I, L respectively) the olfactory bulb exclusivelycontained high levels of NEP and ECE-1 gene transcripts.

In the basal forebrain a distinct but almost complemen-tary distribution of the NEP2 and NEP signals was ob-served, with the former being absent from the caudateputamen and nucleus accumbens, two regions that displaythe highest NEP expression in brain (Figs. 4A, B respec-tively). The reverse situation was observed in other areas,e.g. globus pallidus, subthalamic nucleus and ventral pal-lidum. In the thalamus, where the ECE-1 gene is abundant

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 629

Fig. 1. Comparative mRNA expression patterns of NEP2 (A, B, E), NEP (C, F) and ECE-1 (D, G) during embryonic development. Whole sections ofrat embryos obtained at E14 (A–D) and E22 (E–G) were used for in situ hybridisation experiments. In (A) the E14 embryo selectively expresses NEP2mRNA in the hypothalamus differentiating field which is shown in enlarged form in (B). Comparative cellular localisation of NEP2 (H), NEP (I) andECE-1 (J) mRNAs in spinal ganglia, during prenatal development (E22).

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639630

Fig. 2. Expression of NEP2 mRNAs amongst adult rat brain regions revealed by in situ hybridisation of a sagittal section prepared at L�0.90 mm.

Abbreviations used in the figures

3 oculomotor nucleusA amygdaloid nucleiAcb accumbens nucleusAD anterior thalamus nucleusAH anterior hypothalamus areaAPit anterior lobe pituitaryAPT anterior pretectal nucleusArc arcuate hypothalamus nucleusCA1-3 fields CA1-3 of Ammon’s hornCG central grey matterChP choroid plexusCPu caudate putamenCx cortexDG dentate gyrusDH dorsal hornDk nucleus of DarkschewitschDM dorsomedial hypothalamus nucleiDMTg dorsomedial tegmental areaDPGi dorsal paragigantocellular nucleusDpMe deep mesencephalic nucleusGCL ganglion cell layerGi gigantocellular reticular nucleusHDB nucleus of horizontal limb of diagonal bandHip hippocampusIC inferior colliculiINL inner nuclear layerInt interposed cerebellar nucleii.p. interpeduncular nucleiIPit intermediate lobe pituitaryIPL inner plexiform layerLat lateral cerebellar nucleiLM lateral mammillary nucleus

MA3 medial accessory oculomotor nucleusMD mediodorsal thalamus nucleiMdV medullary reticular nucleus ventralME median eminenceMed medial cerebellar nucleusMM medial medullary nucleiMT medial terminal nucleus of tractOB olfactory bulbONL outer nuclear layerOPL outer plexiform layerOS photoreceptor outer segmentPir piriform cortexPn pontine nucleusR red nucleusRt reticular thalamic nucleusRtTg reticulotegmental nucleus of the ponss.c. superior colliculiSNC substantia nigra, compactSNR substantia nigra, reticularSOC superior olivary complexSol nucleus solitary tractSP5O spinal trigeminal nucleus, oral partSuM supramammillary nucleusTu olfactory tubercleTz nucleus trapezoid bodyVCA ventral cochlear nucleus anteriorVe vestibular nucleiVH ventral hornVMH ventromedial hypothalamus nucleusVP ventral pallidumZI zona incerta

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 631

Fig. 3. Expression of NEP2 mRNA amongst regions of adult rat CNS revealed by in situ hybridisation of frontal sections. Sections were prepared atIA�4.20 mm (A), IA�2.96 mm (B), IA �0.68 mm (C) and cervical 5 (D). In (E) enlargement of the red nucleus (R) showing that hybridisation signalsare located on large size neurons. In (F), enlargement of the trapezoid nucleus (Tz) showing co-expression of enolase mRNA (detected with adigoxigenin-labelled probe) and NEP2 mRNA (detected with a 33P-labelled probe). All NEP2-expressing neurons also expressed enolase.

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639632

Fig. 4. Comparative mRNA expression patterns of NEP2 (A, D, G, J), NEP (B, E, H, K) and ECE-1 (C, F, I, L) revealed by in situ hybridisation on frontalsections of rat brain. Sections were prepared at IA�10.60 mm (A–C), IA�6.70 mm (D–F), IA�4.00 mm (G–I) and IA�3.40 mm (J–L).

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 633

in most nuclei whereas the NEP gene is not expressed, theNEP2 gene was distinctly observed in the mediodorsal,anterodorsal and paraventricular nuclei. In the zona in-certa, a uniquely high signal level was detected for NEP2(Fig. 4D).

The hypothalamus and mesencephalon displayed thebroadest NEP2 distribution with nuclei expressing NEP2 ata moderate to high level. The NEP and ECE-1 signals,however, were detected only in a few nuclei (Fig. 4D, G, J;4E, H, K; 4F, I, L respectively). The brainstem also con-tained many nuclei with a high signal level for NEP2 (Figs.

2, 3C) and devoid of NEP or ECE-1 expression, e.g. thesuperior olivary complex or nucleus of the trapezoid body.In the cerebellum, whereas the NEP2 gene was selectivelyexpressed in interposed, lateral and medial nuclei (Fig. 2),transcripts for the other two metallopeptidases were abun-dantly co-expressed in the Purkinje cell layer.

Finally, in the spinal cord (cervical part), NEP2 washighly expressed in motor neurons and small scatteredcells in the ventral horn, whereas the NEP gene wasalmost exclusively localised to the superficial laminae ofthe dorsal horn and the ECE-1 gene moderately and rather

Fig. 5. Comparative distributions of NEP2 (A), NEP (B) and ECE-1 (C) mRNAs by in situ hybridisation on sections of adult rat cervical spinal cord.

Fig. 6. Comparative distributions and cellular localisation of NEP2 (A, D), NEP (B, E) and ECE (C, F) mRNAs in adult rat retina.

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639634

uniformly expressed (Fig. 5A, B, C respectively). The pitu-itary displayed a high and uniform level of expression ofthe three metallopeptidases in the intermediate lobe,whereas scattered cells of the anterior lobe were labelledand no distinct signal observed in the posterior lobe. In theretina (Fig. 6), NEP2 gene transcripts were detected in fewganglion cells and in numerous cells of the inner nuclearlayer (INL) (Fig. 6D). In contrast, NEP mRNAs were con-centrated in pigmental epithelium cells, and a faint signalwas observed over the INL and the outer nuclear layer(Fig. 6E), whilst ECE-1 mRNAs were expressed evenlythroughout the retina (Fig. 6F).

DISCUSSION

The major finding of the present study lies in the predom-inant expression of the recently identified peptidase NEP2in selected sub-populations of neurons in the CNS,throughout development and onto adulthood. This obser-vation, taken together with its ability to cleave neuropep-tides (Rose et al., 2002), suggests that it belongs to theclass of “neuropeptidases.” Although this term was firstproposed to characterise the function of NEP (Schwartz,1983), it seems better suited to NEP2, which is almostexclusively expressed in brain (with the exception of tes-tis), whereas NEP is, in addition, found in a variety ofperipheral tissues, e.g. kidney, intestine and adrenals asinitially detected by the characterisation of its catalyticactivity (Llorens-Cortes and Schwartz, 1981) and con-firmed by immunohistochemical (Ronco et al., 1988) andhere, by mRNA in situ hybridisation studies (Fig. 1).

The ontogeny of the three metallopeptidases, NEP2,NEP and ECE-1 differ substantially, inasmuch as NEP2expression seems restricted to developing and differenti-ating neurons whilst the NEP and ECE-1 genes are widelyexpressed early on in the CNS and periphery. This latterdistribution could reflect the ontogeny of their substrates,be it enkephalins for NEP, and “Big-ET-1” for ECE-1, aswell as the alleged roles of these peptides during devel-opment (Brand et al., 1998; Zagon et al., 1999). The re-stricted and late expression of NEP2 may, on the otherhand, point to a more specific function of this peptidase,aimed toward the metabolism of neuropeptides rather thanpeptides involved in growth.

The NEP2 neuronal localisation was initially suggestedby the highly heterogeneous and sometimes laminar dis-tribution of the signal on autoradiographic films (withoutlabelling of white matter) and confirmed by the high densityof silver grains over selected neurons (e.g. lateral mam-millary, red, pontine, trapezoid, oculomotor nuclei) de-tected using emulsion autoradiography. Finally, the NEP2gene transcripts were shown to co-localise with enolase, aspecific neuronal cell marker (Sakimura et al., 1985; Fig.3F).

Nevertheless, extra-neuronal expression of NEP2 wasdetected in selected cell populations of the pituitary,ependymal layer of the brain ventricles and the choroidplexuses. In the pituitary, the intermediate lobe stronglyexpresses the three studied metallopeptidases whose

functions therein appears unknown since they are not ableto cleave messenger peptides found in this area: for in-stance, products of the POMC gene (�-MSH, �-endorphinor ACTH) are not cleaved by either recombinant NEP2 orNEP (Rose et al., 2002). In contrast, the high expression ofboth NEP and NEP2 in scattered cells of the anterior lobecould correspond to a role of these peptidases in theinactivation of gonadoliberins, e.g. gonadotropin-releasinghormone (GnRH) identified as a substrate for these twopeptidases (Rose et al., 2002). Gonadotropin-producingcells in pig adenohypophysis were previously found toexpress NEP (Barnes and Kenny, 1988), but the specificityof the antibodies raised against NEP2 have not yet allowedthe assessment of its expression therein. It will be inter-esting in the future to characterise the peptidergic nature ofthe cells expressing NEP2 within this tissue. Also notewor-thy is the fact that NEP2 expression was high in the variousnuclei of the hypothalamus and in the median eminence,possibly implying a role in the metabolism of neurohor-monal peptides in the hypothalamo-pituitary axis.

In the choroid plexuses and ependymal cells, the pres-ence of high NEP activity, immunoreactivity, and mRNAlevels (Matsas et al., 1985; Pollard et al., 1989; Gaudoux etal., 1993) has been linked to the production of a solubleform of the enzyme in the cerebrospinal fluid (Spillantini etal., 1990) and a barrier function protecting the brain fromhigh levels of circulating peptides in the blood. This mayalso be a function of NEP2 which is co-expressed in thesecells and for which an isoform generated by alternativesplicing has been found to lead to a secreted, enzymati-cally active peptidase (Rose et al., 2002). This secretedisoform has been identified in the testis (Ouimet et al.,2000) but could be expressed in other non-neuronal cells.

The overall distribution of NEP2 in CNS areas wasclearly distinct from that of NEP and ECE-1 (Table 1).Strikingly, in a large number of areas, NEP and NEP2 genetranscripts displayed complementary distributions. Thiswas namely the case in the basal ganglia in which NEP2gene expression is almost undetectable whereas NEP ishighly expressed in caudate putamen and nucleus accum-bens. Indeed, these areas are the most highly endowedwith NEP (enkephalinase) catalytic activity (Malfroy et al.,1979; Llorens-Cortes and Schwartz, 1981; Back andGorenstein, 1989) and immunoreactivity (Marcel et al.,1990; Pollard et al., 1987) corresponding to expression inmedium spiny neurons and all along a striatonigral path-way (Marcel et al., 1990). Such a localisation is alleged toaccount for the physiological role of the enzyme in theinactivation of endogenous enkephalins and tachykinins.Interestingly, neurons in the substantia nigra pars reticu-lata, upon which projects the striatonigral pathway, do notexpress NEP, but contain high levels of NEP2 mRNAs.Since both NEP and NEP2 are able to cleave enkephalinsand tachykinins with comparable efficiency (Rose et al.,2002), this suggests that the complementary expression ofthe two peptidases in some areas reflects their comple-mentary function in the inactivation of the same neuropep-tides. The role of NEP in the inactivation of endogenousenkephalins in striatal slices (Patey et al., 1981) and of

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 635

endogenous substance P in nigral slices (Mauborgne etal., 1987) would be consistent with this view. Since this role

has been proposed following studies using inhibitors whichhave recently been shown not to discriminate NEP from

Table 1. Compared distributions of neprilysin (NEP)2, NEP and endothelin-converting enzyme 1 (ECE-1) mRNAs in adult brain areas

Brain structures NEP2 NEP ECE-1

I—Cerebral cortex and hippocampusNeocortex (Cx) 1–2� 0–2� 0–1�

Cingulate, agranular insular and orbitofrontal cortex 1� 0 0Retrosplenial and perirhinal cortex 1-2� 0 0Piriform cortex (Pir) 2� 0 0Hippocampic pyramidal cell layer (Hip, CA1, CA2, CA3) 2� 1� 3-4�

Dentate gyrus granular layer 1� 2� 4�

Dentate gyrus polymorph layer 1� 0II—Olfactory system

Anterior olfactory nucleus (OB) 1� 0 0Glomerular layer of the olfactory bulb 0 4� 0Granular layer of the accessory bulb 0 0 4�

III—Basal forebrainMedial septal nucleus 1–2� 0 0Nucleus of the diagonal band (HDB) 1–2� 0 3�

Nucleus accumbens (Acb) 0–1� 3� 0Caudate putamen (CPu) 0 4� 0Olfactory tubercle (Tu) 1� 4� 0Globus pallidus 1� 0 0Islands of Calleja 1� 0 0Substantia innominata 1� 0 4�

Subthalamic nucleus 3� 0 0Ventral pallidum (VP) 2� 0 0

IV—Amygdaloid complexMedial, central, lateral and basolateral nucleus (A) 1� 0 0Bed nucleus of the stria terminalis 1� 0 0

V—Thalamus, habenula and zona incertaMediodorsal thalamus nucleus (MD) 4� 0 3�

Anterodorsal thalamus nucleus (AD) 3� 0 3�

Anteroventral thalamus nucleus 0 0 3�

Anteromedial thalamus nucleus 0 0 3�

Paraventricular thalamus nucleus 3� 0 3�

Paratenial thalamus nucleus 0 0 3�

Paracentral thalamus nucleus 0 0 3�

Thalamic reticular nucleus (Rt) 1� 0 4�

Habenular nucleus 1–2� 4� 0Zona incerta (ZI) 4� 0 0

VI—HypothalamusPeriventricular nucleus 1–2� 0 0Medial preoptic area 1� 0 0Medial preoptic nucleus 1–2� 0 0Suprachiasmatic nucleus 3–4� 0 4�

Paraventricular nucleus 2–3� 0 4�

Anterior hypothalamic area (AH) 1� 0 0Anterior area: anterior, central and posterior parts 1� 0 0Lateral hypothalamic area 1� 0 0Supraoptic nucleus 0 0 4�

Arcuate nucleus (Arc) 2–3� 2� 4�

Median eminence (ME) 2� 0 3�

Tuber cinereum area 1–2� 0 0Ventromedial nucleus (VMH) 2–3� 0 0Dorsomedial nucleus (DM) 2–3� 0 0Posterior hypothalamic area 1� 0 0Premammillary nucleus 2–3� 0� 0Medial mammillary nucleus (MM) 1–2� 4� 0Lateral mammillary nucleus (LM) 4� 0 0Supramammillary nucleus (SuM) 1� 4� 0

Regional levels of expression of NEP2, NEP and ECE-1 were rated as: (0) no labelling, (1�) low, (2�) moderate, (3�) high and (4�) very high.

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639636

NEP2 activity in a clear-cut manner (Rose et al., 2002), itwould be of interest to re-assess the participation of NEP2versus that of NEP in tachykinin metabolism in areas suchas the substantia nigra where the former enzyme is hereshown to be expressed (Barnes et al., 1993; Pollard et al.,1989; Matsas et al., 1984).

Complementary distributions of NEP and NEP2 can beobserved in a number of other areas, e.g. in most thalamic,hypothalamic and brainstem nuclei in which NEP2 mRNAsare generally expressed in the absence of NEP. This is, forinstance, the case in the locus coeruleus or raphe nucleiwhere the effects of exogenous opioids upon noradrener-gic and serotoninergic neuronal activity is reproduced by“enkephalinase” (i.e. NEP/NEP2) inhibitors in a naloxone-reversible manner (Lecomte et al., 1986; Llorens-Cortesand Schwartz, 1984). This raises the possibility that thebeneficial effects of these drugs on the opiate withdrawalsyndrome in rodents (Glimcher et al., 1984; Haffmans etal., 1987) and addicted human patients (Hartmann et al.,1991), which apparently involves these monoaminergicsystems (Nestler and Aghajanian, 1997; Gold et al., 1978),

may rely upon NEP2 rather than NEP inhibition. This isparticularly relevant when in vivo pharmacological studieswere performed using inhibitors such as phosphoramidon,equipotent on the two peptidases (Rose et al., 2002).

This complementary distribution of NEP2 and NEP canalso be observed in the retina, in which NEP2, expressedin some ganglion cells as well as in the INL (Fig. 6), isbetter poised to inactivate peptides expressed therein thanNEP. Indeed, NEP2 is either co-expressed or localised inthe proximity of cells expressing tachykinins, [Met5]-en-kephalin or GnRH within this tissue (Oyamada et al., 1999;Kondoh et al., 1996; Kolb et al., 1995; Isayama et al., 1996;Fukuda et al., 1982), which is not the case for NEP.

In some areas, however, the expression patterns of thetwo peptidases clearly overlapped, and co-expressioncould even be observed at the cellular level. In agreement,both are expressed in the same laminae of the neocortexor in neurons of the red nucleus (Fig. 4J and K respective-ly). In superficial layers of the dorsal horn of the spinalcord, where NEP/NEP2 inhibitors could exert their antino-ciceptive activity (Roques et al., 1980), NEP2 is expressed

Table 1. (Continued)

Brain structures NEP2 NEP ECE-1

VII—MesencephalonAnterior pretectal nucleus (APT) 3� 2� 0Inferior and superior colliculi (IC, SC) 2–3� 2–3� 0Central gray (CG) 3� 0 0Deep mesencephalic nucleus (DpMe) 3� 0 0Principal oculomotor nucleus (3) 4� 0 0Substantia nigra, compact part (SNC) 1� 0 4�

Substantia nigra, lateral part 2� 0 0Substantia nigra, reticular part (SNR) 3� 0 0Interpeduncular nucleus (IP) 1–2� 4� 0Red nucleus (R) 4� 2� 0Medial terminal nucleus of accessory tract (MT) 2� 2� 0Nucleus of Darkschewitsch (Dk) 4� 0 0

VIII—BrainstemLocus coeruleus 2� 0 4�

Dorsal tegmental nucleus (DMTg) 3–4� 4� 0Superior olivary complex (SOC) 3� 0 0Nucleus of the trapezoid body (Tz) 4� 0 0Pontine nucleus (Pn) 3� 4� 0Vestibular nucleus (Ve) 3–4� 2� 0Raphe nucleus 2� 0 3�

Medullary reticular nucleus (Gi, MdV, DPGi) 3–4� 0 0Spinal tract nucleus (SP5O) 4� 0 0Cochlear nucleus (VCA) 3� 0 0Solitary tract nucleus (Sol) 2� 0 0

IX—CerebellumInterposed, lateral and medial cerebellar nucleus

(Int,Lat,Med)4� 0 0

Granular layer of cerebellar cortex 0 0 2�

Purkinje cell layer 0 4� 4�

X—OthersChoroid plexuses (ChP) 2� 4� 1�

Anterior lobe of pituitary (APit) 2� 1� 4�

Intermediate lobe of pituitary (IPit) 4� 3� 3�

Posterior lobe of pituitary 0 0 1�

Spinal cord, dorsal horn (DH) 1� 3� 2�

Spinal cord, ventral horn (VH) 3� 1� 2�

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 637

in scattered neurons which are likely to also express NEP(abundantly present therein), whereas NEP2 is abundantlyexpressed in the ventral horn (Fig. 5A and B respectively).

The original description of ECE-1 gene expression inrat CNS (Ouimet et al., 2000) is here confirmed and over-all, the expression pattern of ECE-1 is a more restrictedone than that of NEP2 or NEP (Fig. 4). This observation isin accordance with its apparently restricted enzymaticspecificity and physiological function, which consists in theproduction of mature ET-1 (Xu et al., 1994). It is notewor-thy that the expression of ECE-1 in the CNS confirms therole of this peptide as a neurotransmitter, and that itsdistribution seems restricted to neurons that express theET-1 precursor (Sluck et al., 1999; Kurokawa et al., 2000).

In the cerebral cortex, hippocampus and olfactory sys-tem, the observed ECE-1 expression was in agreementwith the reported expression of the ECE-1 protein (Sluck etal., 1999). The highest and broadest expression of theECE-1 transcript was found in the thalamic and hypotha-lamic nuclei, e.g. in the suprachiasmic, paraventricular andsupraoptic nuclei, the arcuate nucleus and median emi-nence, as well as in the anterior lobe of the pituitary,localisations which may confer a neurohormonal functionto ET-1. In other regions of the CNS, the substantia nigra,locus coeruleus and the Purkinje cell layer of the cerebel-lum also contain high levels of ECE-1 transcript. Finally,the expression of ECE-1 in the spinal cord (Fig. 5) is alsodemonstrated, a localisation which was only known inhuman tissue (Giaid et al., 1989), as well as its localisationto all layers of the retina (Fig. 6), once again underlying theimportance of the endothelin system in the eye (Ripodas etal., 2001).

Recently, a novel function of NEP and other metal-lopeptidases of the M13 sub-family has been brought tolight, as NEP and ECE-1 have been shown to cleave theamyloid peptides A�1-40 and A�1-42 (see Carson andTurner, 2002 for review). Although these peptides are mostefficiently degraded by NEP, the data presented here sug-gest that NEP2 and ECE-1 may be better poised to do soas they are more abundantly expressed in areas relevantto Alzheimer’s disease pathology, i.e. in the cortex for theformer and hippocampus for the latter.

In conclusion, the recent characterisation of the sub-strate specificity of NEP2 suggested that it may be in-volved in the inactivation of neuropeptides or hormonespreviously known as NEP substrates (Rose et al., 2002).The results presented here show that this may well be thecase, and that the respective roles of these two peptidasesis probably dictated by their different localisation in theCNS, whereas the localisation of ECE-1 is in agreementwith its alleged role in ET-1 biosynthesis. Finally, theirrespective potential involvement in �-amyloid peptideclearance will need to be reassessed in light of their local-isations, though in vivo experiments will require specificinhibitors for each of the candidates involved.

REFERENCES

Back SA, Gorenstein C (1989) Histochemical visualization of neutralendopeptidase-24.11 (enkephalinase) activity in rat brain: cellular

localization and codistribution with enkephalins in the globus palli-dus. J Neurosci 9:4439–4455.

Barnes K, Kenny J (1988) Endopeptidase 24.11 in the adenohypoph-ysis of the pig is localized in the gonadotropic cells. Peptides9:55–63.

Barnes K, Turner AJ, Kenny AJ (1993) An immunoelectron micro-scopic study of pig substancia nigra shows co-localization of endo-peptidase-24.11 with substance P. Neuroscience 53:1073–1082.

Baumer P, Danquechin-Dorval E, Bertrand J, Vetel JM, Schwartz JC,Lecomte JM (1992) Effects of acetorphan, an enkephalinase inhib-itor, on experimental and acute diarrhea. Gut 33:753–758.

Bralet J, Schwartz JC (2001) Vasopeptidase inhibitors: an emergingclass of cardiovascular drugs. Trends Pharmacol Sci 22:106–109.

Brand M, Le Moullec JM, Corvol P, Gasc JM (1998) Ontogeny ofendothelin-1 and -3, their receptors, and endothelin convertingenzyme-1 in the early human embryo. J Clin Invest 101:549–559.

Carson JA, Turner AJ (2002) �-Amyloid catabolism: roles for neprilysin(NEP) and other metallopeptidases? J Neurochem 81:1–8.

Diaz J, Levesque D, Lammers CH, Griffon N, Martres MP, SchwartzJC, Sokoloff P (1995) Phenotypical characterization of neuronsexpressing the dopamine D3 receptor in the rat brain. Neuro-science 65:731–745.

Facchinetti P, Rose C, Rostaing P, Triller A, Schwartz JC (1999)Immunolocalization of tripeptidyl peptidase II, a cholecystokinin-inactivating enzyme, in rat brain. Neuroscience 88:1225–1240.

Fukuda M, Ishimoto I, Shiosaka S, Kuwayama Y, Shimizu Y, InagakiS, Sakanaka M, Takagi H, Senba E, Takatsuki K, Umegaki K,Tohyama M (1982) Localization of LH-RH immunoreactivity in theavian retina. Curr Eye Res 2:71–74.

Ghaddar G, Ruchon AF, Carpentier M, Marcinkiewicz M, Seidah NG,Crine P, Desgroseillers L, Boileau G (2000) Molecular cloning andbiochemical characterization of a new mouse testis soluble-zinc-metallopeptidase of the neprilysin family. Biochem J 347:419–429.

Gaudoux F, Boileau G, Crine P (1993) Localization of neprilysin (EC3.4.24.11) mRNA in rat brain by in situ hybridization. J NeurosciRes 34:426–433.

Giaid A, Gibson SJ, Ibrahim BN, Legon S, Bloom SR, Yanagisawa M,Masaki T, Varndell IM, Polak JM (1989) Endothelin 1, an endothe-lium-derived peptide, is expressed in neurons of the human spinalcord and dorsal root ganglia. Proc Natl Acad Sci USA 86:7634–7638.

Glimcher PW, Giovino AA, Margolin DH, Hoebel BG (1984) Endoge-nous opiate reward induced by an enkephalinase inhibitor, thior-phan, injected into the ventral midbrain. Behav Neurosci 98:262–268.

Gold MS, Redmond DG Jr, Kleber HD (1978) Clonidine blocks acuteopiate withdrawal symptoms. Lancet 2:599–601.

Haffmans J, Walsum MV, Van Amsterdam JC, Dzojic MR (1987)Phelorphan, an inhibitor of enzymes involved in the biodegradationof enkephalins, affected the withdrawal symptoms in chronic mor-phine-dependant rats. Neuroscience 22:233–236.

Hartmann F, Poirier MF, Bourdel MC, Loo H, Lecomte JM, SchwartzJC (1991) Comparison of acetorphan with clonidine for opiatewithdrawal symptoms. Am J Psychiatry 148:627–629.

Isayama T, McLaughlin PJ, Zagon IS (1996) Ontogeny of preproen-kephalin mRNA expression in the rat retina. Vis Neurosci 13:695–704.

Kerr MA, Kenny AJ (1974) The purification and specificity of a neutralendopeptidase from kidney brush border. Biochem J 137:477–488.

Kenny AJ, Stephenson SL (1988) Role of endopeptidase-24.11 in theinactivation of atrial natriuretic peptide. FEBS Lett 232:1–8.

Kitabgi P, Dubuc I, Nouel D, Costentin J, Cuber JC, Fulcrand H, DoulutS, Rodriguez M, Martinez J (1992) Effects of thiorphan, bestatinand a novel metallopeptidase inhibitor JMV 390-1 on the recoveryof neurotensin and neuromedin N released from mouse hypothal-amus. Neurosci Lett 42:200–204.

Kolb H, Fernandez E, Ammermuller J, Cuenca N (1995) Substance P:

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639638

a neurotransmitter of amacrine and ganglion cells in the vertebrateretina. Histol Histopathol 10:947–968.

Kondoh A, Houtani T, Ueyama T, Baba K, Ikeda M, Yamagishi K, MikiH, Uyama M, Nakanishi S, Sugimoto T (1996) In situ hybridizationof substance P receptor in the rat retina. Exp Brain Res 112:181–186.

Kurokawa K, Yamada H, Liu Y, Kudo M (2000) Immunohistochemicaldistribution of the endothelin-converting enzyme-1 in the rat hypo-thalamo-pituitary axis. Neurosci Lett 284:81–84.

Lecomte JM, Costentin J, Vlaiculescu A, Chaillet P, Marcais-ColladoH, Llorens-Cortes C, Leboyer M, Schwartz JC (1986) Pharmaco-logical properties of acetorphan, a parenterally active “enkephali-nase” inhibitor. Pharmacol Exp Ther 237:937–944.

Llorens-Cortes C, Schwartz JC (1981) Enkephalinase activity in ratperipheral organs. Eur J Pharmacol 69:113–116.

Llorens-Cortes C, Schwartz JC (1984) Changes in turnover of cerebralmonoamines following inhibition of enkephalin metabolism by thi-orphan and bestatin. Eur J Pharmacol 104:369–374.

Malfroy B, Swerts JP, Guyon A, Roques BP, Schwartz JC (1978) High-affinity enkephalin-degrading peptidase in brain is increased aftermorphine. Nature 276:523–526.

Malfroy B, Swerts JP, Llorens C, Schwartz JC (1979) Regional distri-bution of a high-affinity enkephalin-degrading peptidase (‘enkeph-alinase’) and effects of lesions suggest localization in the vicinity ofopiate receptors in brain. Neurosci Lett 11:329–334.

Marcel D, Pollard H, Verroust P, Schwartz JC, Beaudet A (1990)Electron microscopic localization of immunoreactive enkephalinase(EC 3.4.24.11) in the neostriatum of the rat. J Neurosci 10:2804–2817.

Matsas R, Fulcher IS, Kenny AJ, Turner AJ (1983) Substance P and[Leu]enkephalin are hydrolyzed by an enzyme in pig caudate syn-aptic membranes that is identical with the endopeptidase of kidneymicrovilli. Proc Natl Acad Sci USA 80:3111–3115.

Matsas R, Turner AJ, Kenny AJ (1984) Endopeptidase-24.11 andaminopeptidase activity in brain synaptic membranes are jointlyresponsible for the hydrolysis of cholecystokinin octapeptide (CCK-8). FEBS Lett 175:124–128.

Matsas R, Rattray M, Kenny A, Turner AJ (1985) The metabolism ofneuropeptides: endopeptidase 24.11 in human synaptic membranepreparations hydrolyses Substance P. Biochem J 228:487–492.

Mauborgne A, Bourgoin S, Benoliel JJ, Hirsh H, Berthier JL, Hamon M,Cesselin F (1987) Enkephalinase is involved in the degradation ofendogenous Substance P released from slices of rat substantianigra. J Pharmacol Exp Ther 243:674–680.

Nestler EJ, Aghajanian GK (1997) Molecular and cellular basis ofaddiction. Science 278:58–63.

Ouimet T, Facchinetti P, Rose C, Bonhomme MC, Gros C, SchwartzJC (2000) Neprilysin II: a putative novel metalloprotease and itsisoforms in CNS and testis. Biochem Biophys Res Commun 271:565–570.

Oyamada H, Takatsuji K, Senba E, Mantyh PW, Tohyama M (1999)Postnatal development of NK1, NK2, and NK3 neurokinin receptorsexpression in the rat retina. Dev Brain Res 117:59–70.

Patey G, De La Baume S, Schwartz JC, Gros C, Roques BP, Fournie-Zaluski MC, Soroca-Lucas E (1981) Selective protection of methi-onine enkephalin release from brain slices by enkephalinase inhi-bition. Science 212:1153–1155.

Pollard H, Llorens-Cortes C, Couraud JY, Ronco P, Verroust P,Schwartz JC (1987) Enkephalinase (EC 3.4.24.11) is highly local-ized to a striatonigral pathway in rat brain. Neurosc Lett 77:267–271.

Pollard H, Bouthenet ML, Moreau J, Souil E, Verroust P, Ronco P,Schwartz JC (1989) Detailed immunoautoradiographic mapping ofenkephalinase (EC 3.4.24.11) in rat central nervous system: com-

parison with enkephalins and Substance P. Neuroscience 30:339–376.

Ripodas A, de Juan JA, Roldan-Pallares M, Bernal R, Moya J, Chao M,Lopez A, Fernandez-Cruz A, Fernandez-Durango R (2001) Local-isation of endothelin-1 mRNA expression and immunoreactivity inthe retina and optic nerve from human and porcine eye: evidencefor endothelin-1 expression in astrocytes. Brain Res 912:137–143.

Ronco P, Pollard H, Galleran M, Delauche M, Schwartz JC, VerroustP (1988) Distribution of enkephalinase (membrane metalloen-dopeptidase, EC 3.4.24.11) in rat organs: detection using a mono-clonal antibody. Lab Invest 58:210–217.

Roques BP, Fournie-Zaluski MC, Soroca E, Lecomte JM, Malfroy B,Llorens-Cortes C, Schwartz JC (1980) The enkephalinase inhibitorthiorphan shows antinociceptive activity in mice. Nature 288:286–288.

Roques BP, Noble F, Dauge V, Fournie-Zaluski MC, Beaumont A(1993) Neutral endopeptidase 24.11: structure, inhibition, and ex-perimental and clinical pharmacology. Pharmacol Rev 45:87–146.

Rose C, Vargas F, Facchinetti P, Bourgeat P, Bambal RB, Bishop PB,Chan SM, Moore AN, Ganellin CR, Schwartz JC (1996) Character-ization and inhibition of a cholecystokinin-inactivating serine pepti-dase. Nature 380:403–409.

Rose C, Voisin S, Gros C, Schwartz JC, Ouimet T (2002) Cell-specificactivity of neprilysin 2 isoforms and enzymic specificity comparedwith neprilysin. Biochem J 363:697–705.

Sakimura K, Kushiya E, Obinata M, Odani S, Takahashi Y (1985)Molecular cloning and the nucleotide sequence of cDNA for neu-ron-specific enolase messenger RNA of rat brain. Proc Natl AcadSci USA 82:7453–7457.

Salazar-Lindo E, Santisteban-Ponce J, Chea-Woo E, Gutierrez M(2000) Racecadotril in the treatment of acute watery diarrhea inchildren. N Engl J Med 343:463–467.

Schwartz JC (1983) Metabolism of enkephalins and the inactivatingneuropeptidase concept. Trends Neurosci 6:15–18.

Schwartz JC, Gros C, Lecomte JM, Bralet J (1990) Enkephalinase (EC3.4.24.11) inhibitors: protection of endogenous ANF against inac-tivation and potential therapeutic applications. Life Sci 47:1279–1287.

Sluck JM, Lin RCS, Ktolik LI, Jeng AY, Lehmann JC (1999) Endothelinconverting enzyme-1-, endothelin-1-, and endothelin-3-like immu-noreactivity in the rat brain. Neuroscience 91:1483–1497.

Spillantini MG, Sicuteri F, Salmon S, Malfroy B (1990) Characteriza-tion of endopeptidase 3.4.24.11 (“enkephalinase”) activity in humanplasma and cerebrospinal fluid. Biochem Pharmacol 39:1353–1356.

Turner AJ, Isaac RE, Coates D (2001) The neprilysin (NEP) family ofzinc metalloendopeptidases: genomics and function. Bioessays 23:261–269.

Turner AJ, Murphy LJ, Medeiros MS, Barnes K (1996) Endopeptidase-24.11 (neprilysin) and relatives: twenty years on. Adv Exp Med Biol389:141–148.

Wilcox JN, Pollard H, Moreau J, Schwartz JC, Malfroy B (1989)Localization of enkephalinase mRNA in rat brain by in situhybridization: comparison with immunohistochemical localization ofthe protein. Neuropeptides 14:77–83.

Xu D, Emoto N, Giaid A, Slaughter C, Kaw S, DeWit D, Yanagisawa M(1994) A membrane-bound metalloprotease that catalyses the pro-teolitic activation of big-endothelin 1. Cell 78:473–485.

Zagon IS, Wu Y, McLaughlin PJ (1999) Opioid growth factor and organdevelopment in rat and human embryos. Brain Res 839:313–322.

Zuzel KA, Rose C, Schwartz JC (1985) Assessment of the role of“enkephalinase” in cholecystokinin inactivation. Neuroscience15:149–158.

(Accepted 25 November 2002)

P. Facchinetti et al. / Neuroscience 118 (2003) 627–639 639