Neuronal expression and interaction with the synaptic protein …material.bccn.uni-freiburg.de/publications-bccn/2006/Gerke06_466.pdf · In the adult brain, Neph1 and Neph2 are predom-

Neuronal Expression and Interactionwith the Synaptic Protein CASK Suggest

a Role for Neph1 and Neph2 inSynaptogenesis

PETER GERKE,1 THOMAS BENZING,1 MARTIN HOHNE,1 ANDREAS KISPERT,3

MICHAEL FROTSCHER,2 GERD WALZ,1AND OLIVER KRETZ2*

1Renal Division, University of Freiburg, D-79104 Freiburg, Germany2Institute of Anatomy and Cell Biology, University of Freiburg,

D-79104 Freiburg, Germany3Institute for Molecular Biology, Hannover Medical School, D-30625 Hannover, Germany

ABSTRACTFormation, differentiation, and plasticity of synapses require interactions between pre- and

postsynaptic partners. Recently, it was shown that the transmembrane immunoglobulin super-family protein SYG-1 is required for providing synaptic specificity in C. elegans. However, it isunclear whether the mammalian orthologs of SYG-1 are also involved in local cell interactions todetermine specificity during synapse formation. We used in situ hybridization, immunohisto-chemistry, and immunogold electron microscopy to study the temporal and spatial expression ofNeph1 and Neph2 in the developing and adult mouse brain. Both proteins show similar patternswith neuronal expression starting around embryonic days 12 and 11, respectively. Expression isstrongest in areas of high migratory activity. In the adult brain, Neph1 and Neph2 are predom-inantly seen in the olfactory nerve layer and the glomerular layer of the olfactory bulb, in thehippocampus, and in Purkinje cells of the cerebellum. At the ultrastructural level, Neph1 andNeph2 are detectable within the dendritic shafts of pyramidal neurons. To a lesser extent, thereis also synaptic localization of Neph1 within the stratum pyramidale of the hippocampal CA1 andCA3 region on both pre- and postsynaptic sites. Here it colocalizes with the synaptic scaffoldercalmodulin-associated serin/threonin kinase (CASK), and both Neph1 and Neph2 interact withthe PDZ domain of CASK via their cytoplasmic tail. Our results show that Neph proteins areexpressed in the developing nervous system of mammals and suggest that these proteins mayhave a conserved function in synapse formation or neurogenesis. J. Comp. Neurol. 498:466–475,2006. © 2006 Wiley-Liss, Inc.

Indexing terms: olfactory bulb; hippocampus; synapse formation; axonal guidance; kirre; duf

Members of the immunoglobulin (Ig) superfamily of celladhesion molecules are required for a variety of cellularprocesses in the developing and adult brain including cellrecognition, migration, axonal guidance, and synapse for-mation (Rougon and Hobert, 2003). These processes aremediated through homophilic or heterophilic interactionswith other Ig proteins frequently leading to intracellularsignaling events (Benzing, 2004).

Within the Ig superfamily, the Neph proteins comprisea group of three type I membrane proteins with five ex-tracellular Ig domains that are highly conserved throughevolution (Sellin et al., 2003; Gerke et al., 2005). Althoughmost strongly expressed in the central nervous system(CNS), Neph1 and Neph2 are best studied in the kidneywhere they participate in the formation of a specialized

epithelial cell contact called the slit diaphragm (Khosh-noodi et al., 2003; Liu et al., 2003; Gerke et al., 2005).Neph proteins interact with nephrin, another Ig super-

Grant sponsor: Deutsche Forschungsgemeinschaft; Grant number: DFGBE2212, SFB 505; Grant sponsor: Else-Kroner-Fresenius Foundation.

The first two authors contributed equally to this study.*Correspondence to: Oliver Kretz, M.D., Anatomy and Cell Biology,

University Freiburg, D-79104 Freiburg, Germany.E-mail: [email protected] 27 December 2005; Revised 14 March 2006; Accepted 28 April

2006DOI 10.1002/cne.21064Published online in Wiley InterScience (www.interscience.wiley.com).

THE JOURNAL OF COMPARATIVE NEUROLOGY 498:466–475 (2006)

© 2006 WILEY-LISS, INC.

family membrane protein with expression in brain, kid-ney, and pancreas (Putaala et al., 2001; Gerke et al.,2003). Little is known about neuronal functions of Nephproteins in mammals. Neph1 and Neph2 are associatedwith lipid rafts via interactions with the stomatin familymember podocin (Sellin et al., 2003). Interactions with thescaffolder zonula occludens-1 (ZO-1) induce tyrosine phos-phorylation of their cytoplasmic domains leading to theactivation of canonical signaling cascades in renal cells(Huber et al., 2003).

It is of interest that a pivotal role in neuronal develop-ment has been attributed to the Caenorhabditis elegansNeph ortholog SYG-1 (Shen et al., 2004). During synapseformation, the nephrin-like protein SYG-2 interacts withSYG-1 expressed on presynaptic axons and directs SYG-1accumulation and synapse formation to adjacent regionsof the axon. Mutants of either SYG-1 or SYG-2 causedefects in synaptic specificity with ectopic synapse forma-tion on inappropriate targets. Similarly, loss of the Dro-sophila melanogaster Neph ortholog Irregular chiasmC-roughest (IrreC-rst) is associated with severe neuronaldefects, including misguided axonal projections of sensoryorgans and disrupted patterning of the eye (Schneider etal., 1995; Venugopala Reddy et al., 1999).

Neph1-deficient mice die few days after birth of severenephrotic syndrome (Donoviel et al., 2001). Therefore,structural or functional postnatal brain defects cannot beexcluded. In this study, we use in situ hybridization, im-munohistochemistry, and immunogold electron micros-copy to localize Neph1 and Neph2 in fetal and adult mouseCNS and analyze the temporal and spatial expressionpatterns. Furthermore, we demonstrate interactions withthe synaptic protein calmodulin-associated serin/threoninkinase (CASK), suggesting a role for Nephs in synapseformation or maintenance.

MATERIALS AND METHODS

Animals

C57Bl6 mice were maintained under controlled condi-tions, and water and food were available ad libitum. Beforethe preparation of CNS tissues for immunohistochemistryand immuno gold labeling, young adult mice (3.5 months old)were deeply anesthetized and transcardially perfused byusing 4% peraformaldehyde (PFA) in 0.1 m phosphate buffer(PB) (pH 7.4; light microscopy) or 4% PFA � 0.1% glutaral-dehyde (GA) in 0.1 m PB (electron microscopy) as fixatives.All experiments were performed in accordance with institu-tional guidelines for animal welfare.

In situ hybridization

We carried out in situ hybridization analysis of wholemouse embryos and embryonic brain sections (embryonicdays 10.5–16.5) using a digoxigenin-labeled antisense ri-boprobe derived from the coding regions of Neph1 andNeph2 complementary DNA (cDNA), respectively (proto-cols available on request).

Plasmids and antibodies

Truncations were generated by polymerase chain re-action (PCR) and standard cloning procedures andverified by automated sequencing. The domain struc-tures of NEPH1 and Neph2 were predicted by usingSMART (http://smart.embl-heidelberg.de). F.Neph1-cyt

and F.Neph2-cyt contain the entire intracellular do-mains of mouse Neph1 and Neph2, respectively, fusedwith an N-terminal flag-tag. mYFP.CASK479 –578 con-tains the PDZ domain of mouse CASK fused with yellowfluorescent protein.

The polyclonal antisera directed against the ectodo-mains of Neph1 and Neph2, respectively, were recentlydescribed including immunohistochemistry on mouse re-nal tissue with negative isotype controls (Sellin et al.,2003; Gerke et al., 2005). Anti-Neph1 antiserum was gen-erated by immunization of a rabbit with purified bacterialrecombinant maltose-binding fusion protein of Neph1amino acids (aa) 65–234. The antiserum was affinity pu-rified against bacterial recombinant glutathione-S-transferase fusion protein of Neph1 (aa) 65–234. Anti-Neph2 antiserum was generated by immunization of arabbit with the peptide MAKDKFRRMNEGQVY (Neph2aa 34–48). The antiserum was affinity-purified againstthe same peptide (Eurogentec, Herstal, Belgium). West-ern blot analysis showed that the antiserum specificallyrecognizes NEPH1 and shows no cross-reactivity (Sellin etal., 2003). The anti-NEPH2 antisera were shown to recog-nize a single band at approximately 95 kD. The anti-Nephantiserum showed no cross-reactivity with Neph1 or neph-rin (Gerke et al., 2005). Monoclonal anti-CASK antiserumwas obtained commercially from upstate (Waltham, MA).The antibody is directed against aa 318–415 of humanCASK and specifically recognizes human, mouse, and ratCASK. Crude mouse brain cytosolic fractions were usedfor specificity controls. Protein G-sepharose and proteinA-sepharose were obtained from Amersham PharmaciaBiotech (Freiburg, Germany). M2 anti-FLAG beads wereobtained from Sigma-Aldrich (Taufkirchen, Germany),

Immunohistochemistry

After the brain was removed, frontal sections of 50-�mthickness were cut by using a vibratome. Sections werewashed in 0.1 M PB and then incubated with the followingprimary antibodies overnight at 4°C: polyclonal anti-Neph1 (1:40,), polyclonal anti-Neph2 (1:50), and mouseanti-CASK (1:100, Waltham, MA). After washing in 0.1 MPB, sections were incubated in secondary antibodies cou-pled to cy2 or cy3 (1:100, Dianova, Hamburg, Germany)overnight at 4°C. After the sections were rinsed in 0.1 MPB for 1 hour, they were mounted in Mowiol. Immunoflu-orescent double labeling was performed by sequential in-cubation with primary antibodies against CASK andNeph1 or Neph2, respectively.

Electron microscopy

Adult mice were anesthetized with sodium pentobarbi-tal and transcardially perfused by using 4% paraformal-dehyde and 0.1% glutaraldehyde in 0.1 PB. Kidneys andbrains were removed and postfixed in the same fixative(overnight at 4°C). Tissues were washed in phosphate-buttered saline (PBS), and then frontal sections (50 �m)were cut on a vibratome and cryoprotected in a solutioncontaining 25% sucrose and 10% glycerol in 50 mM PBS.The sections were freeze-thawed and incubated in block-ing solution containing 2% normal goat serum in 50 mMTris-buffered saline for 1 hour, followed by incubationwith an anti-Neph1 antiserum (1:20 for 48 hours at 4°C).Positive controls were performed by using adult mousekidney sections; negative controls of brain and kidneywere performed by using the same staining procedure

The Journal of Comparative Neurology. DOI 10.1002/cne

467NEPH1 AND NEPH2 IN THE MOUSE CNS

except primary antibody incubation. After the sectionswere washed, the sections were incubated with 1.4-nmgold-coupled goat anti-rabbit secondary antibody (1:100,Nanogold; Nanoprobes, Stony Brook, NY) for immunogoldreaction. Immunogold labeling was then enhanced withHQ silver kit (Nanoprobes). After the sections weretreated with OsO4, the sections were stained with uranylacetate, dehydrated, and flat-embedded in epoxy resin(Durcupan ACM, Fluka; Sigma-Aldrich, Gillingham, UK).Ultrathin sections were cut and analyzed in a Philipps CM100 electron microscope.

Coimmunoprecipitation

HEK 293T cells were transiently transfected by using thecalcium phosphate method. After the cells were incubed for24 hours, they were washed twice and lysed in a 1% TritonX-100 lysis buffer (20 mM Tris-HCl, pH 7.5; 1% TritonX-100; 50 mM NaCl; 50 mM NaF; 15 mM Na4P2O7; 0.1 mMEDTA; 2 mM Na3VO4; and protease inhibitors). After cen-trifugation at 15,000 � g (15 minutes at 4°C) and ultracen-trifugation at 100,000 � g (30 minutes at 4°C), cell lysatescontaining equal amounts of total protein were preclearedwith protein G-sepharose and then incubated for 1 hour at4°C with M2 anti-FLAG beads. The beads were washedextensively with lysis buffer, and bound proteins were re-solved by 10% sodium dodecyl sulfate-polyacrylamide gelelectrophoresis (SDS-PAGE). For endogenous coimmunopre-cipitations, tissues were freshly prepared and perfused insitu with ice-cold PBS before lysis. Mouse brains were ho-mogenized by using a glass potter, cleared by centrifugation,and solubilized in lysis buffer supplemented with 20 mMCHAPS and 3 mM ATP. After centrifugation at 15,000 � g(15 minutes at 4°C) and ultracentrifugation at 100,000 � g(30 minutes at 4°C), cell lysates containing equal amounts oftotal protein were precleared with protein A-sepharose andthen incubated for 1 hour at 4°C with the appropriate anti-body, followed by incubation with 30 �l of proteinA-sepharose beads for 3 hours. The beads were washed ex-tensively with lysis buffer, and bound proteins were resolvedby 10% SDS-PAGE.

RESULTS

Localization of Neph1 in the adultmouse brain

The distribution of Neph1 protein in the adult mousebrain was analyzed by immunohistochemistry. Strong ex-pression was detected in the olfactory nerve layer and theglomerular layer of the olfactory bulb (Fig. 1A). Only asubset of the glomeruli (�5%) was positive for Neph1,showing immunoreactivity in the axons of olfactory recep-tors cells (Fig. 1B). In the hippocampus, Neph1 immuno-reactivity was found in the CA1–3 regions, in the innermolecular layer of the dentate gyrus, and in the mossyfiber projection (Fig. 1C). Within the CA1 region, Neph1was detected in a dotted staining pattern around the py-ramidal cells of the stratum pyramidale (Fig. 1E) as wellas in the stratum radiatum and lacunosum moleculare(Fig. 1D). In the latter, labeling of the dendrites of CA1pyramidal cells was observed. The CA3 region displayed adotted staining surrounding the pyramidal cells and alabeling of the mossy fiber projection (Fig. 1F). Within thecerebellum, the somata and dendritic trees of Purkinjecells were Neph1 immunopositive (Fig. 1G). In mouse

kidney sections serving as positive controls, anti-Neph1labeled glomerular podocytes as expected (Fig. 1H).

Localization of Neph2 in the adultmouse brain

As observed for Neph1, Neph2-immunoreactivity wasdetected in the olfactory nerve layer and the glomerularlayer of the olfactory bulb (Fig. 2A). Again, only a subset of5% of the glomeruli was labeled. Analysis of consecutiveserial sections stained for either Neph1 or Neph2 showedno overlap of both labelings in the glomeruli. Neph2 stain-ing was seen in the axons of olfactory receptor cells withinthe glomeruli (Fig. 2B). In the hippocampus, Neph2 im-munoreactivity was found in the CA1–3 regions, in thedentate gyrus, and in the mossy fiber projection (Fig. 2C).Within the CA1 region of the hippocampus, there wasdotted Neph2 expression around the pyramidal cells in thestratum pyramidale as well as in the stratum radiatumand lacunosum moleculare, labeling the dendrites of CA1pyramidal cells. In addition, some immunopositive inter-neurons were found in the stratum pyramidale (Fig. 2D).In the dentate gyrus, Neph2 was seen in a dotted stainingpattern surrounding the granule cells, in the inner molec-ular layer, and in some hilar neurons (Fig. 2E). Again,kidney sections served as positive controls showing a glo-merular podocyte staining pattern for Neph2 (Fig. 2F).

Ultrastructural localization of Neph1 inneurons of the hippocampus

By using silver-enhanced preembedding immunogoldlabeling, Neph1 was detected at synapses of the stratumradiatum of the mouse hippocampal CA1 region (Fig. 3).Neph1 immunoreactivity localized to the pre- and on thepostsynaptic sites. Only occasionally, Neph1 was detectedin direct neighborhood to the synaptic cleft or the postsyn-aptic density, respectively (Fig. 3A–D). In the stratumradiatum of the CA1 region, Neph1 was most frequently(�70% of the gold grains) detected within the dendriticshafts of pyramidal neurons (Fig. 3E). Positive control forNeph1 immunogold labeling was performed on adultmouse kidney sections. Neph1 immunoreactivity was de-tected at the podocyte slit diaphragm of renal glomeruli(Fig. 3F). Both in mouse kidney and hippocampus-negative controls, lacking the first antibody showed noimmunogold labeling.

Colocalization of Neph1 and Neph2with CASK

CASK is a synapse-associated scaffolder with mostlypostsynaptic but in a few cases (e.g., the mossy fiber but-tons) also presynaptic expression within the CNS. LikeNeph1 and Neph2, it localizes to lipid rafts and interactswith ZO-1 and the Ig superfamily protein nephrin in ratglomeruli. Therefore, we performed double labeling forNeph1 and Neph2 with CASK on mouse hippocampussections to look for matching neuronal expression pat-terns. As shown in Figure 4, CASK labels the mossy fiberprojections (Fig. 4A, green). Neph1 immunoreactivity isalso detected within the region of mossy fiber terminalsand CA3 pyramidal cells (Fig. 4A, red). High-power mag-nification reveales colocalization of the two proteins inmossy fiber terminals (Fig. 4B, yellow). In addition, wefound CASK and Neph1 to be colocalized within the den-dritic shafts of CA1 pyramidal cells (data not shown).


468 P. GERKE ET AL.

Fig. 1. Localization of Neph1 protein in the adult mouse brain.A: Neph1 was detected in the olfactory nerve layer and the glomerularlayer of the olfactory bulb. Please note that only a subset of glomeruliis labeled. B: Higher magnification from (A) showing Neph1-immunopositive axon terminals of olfactory receptor cells within aglomerulum. C: In the hippocampus, Neph1 immunoreactivity wasfound in the CA1–3 region, in the inner molecular layer of the dentategyrus, and in the mossy fiber projection. D,E: Higher magnificationsof the CA1 region of the hippocampus showing dotted staining aroundthe pyramidal cells in the stratum pyramidale (E) as well as in the

stratum radiatum and lacunosum moleculare (D) labeling the den-drites (arrows) of CA1 pyramidal cells. F: Higher magnification of theCA3 region showing a dotted staining surrounding the pyramidal cellsand a labeling of the mossy fiber projection. G: Within the cerebellum,the somata and dendritic tree of Purkinje cells were Neph1 immu-nopositive. H: Positive control for Neph1 immunoreaction was per-formed on adult kidney sections. A typical staining outlining thepodocytes within the renal glomerula was observed. Scale bars � 300�m in C; 120 �m in A; 60 �m in D,F; 30 �m in B,E,G,H.


Fig. 2. Localization of Neph2 protein in the adult mouse brain.A: As observed for Neph1, Neph2 was detected in the olfactory nervelayer and the glomerular layer of the olfactory bulb. Again, only asubset of glomeruli was labeled. B: Higher magnification of (A) show-ing Neph2-immunopositive axon terminals of olfactory receptor cellswithin the glomerula. C: In the hippocampus, Neph2 immunoreactiv-ity was found in the CA1–3 regions, in the dentate gyrus, and in themossy fiber projection. D: Higher magnification of the CA1 region ofthe hippocampus showing dotted staining around the pyramidal cellsin the stratum pyramidale as well as in the stratum radiatum and

lacunosum moleculare labeling the dendrites (arrows) of CA1 pyra-midal cells. In addition, a few immunopositive interneurons werefound in the stratum pyramidale (arrowhead). E: Higher magnifica-tion of the dentate gyrus showing a dotted staining surrounding thegranule cells, labeling of the inner molecular layer and of some hilarneurons (arrowhead). F: Positive control for Neph2 immunoreactionwas performed on adult kidney sections. A typical staining outliningthe podocytes within the renal glomerula was observed. Scale bars �300 �m in C; 120 �m in A; 60 �m in B,D–F.


470 P. GERKE ET AL.

Fig. 3. Ultrastructural localization of Neph1 using silver-enhanced preembedding immunogold labeling. A–D: Examples of syn-aptic localization of Neph1 within the stratum radiatum of the hip-pocampal CA1 region. Note that Neph1 is localized on the pre-(arrows) and on the postsynaptic (arrowheads) site. Only in few casesNeph1 was detected in direct neighborhood to the synaptic cleft or thepostsynaptic density, respectively. E: Within the stratum radiatum of

the CA1 region, Neph1 was most frequently detected within thedendritic shafts of pyramidal neurons (arrows). F: Positive control forNeph1 immunogold labeling was performed on adult kidney sections.Neph1 (arrows) was detected in podocyte foot processes (FP) formingthe slit diaphragm within renal glomerula. Scale bars � 200 nm in E;100 nm in A–D,F.



Fig. 4. Double labeling for Neph1 (A,B) and Neph2 (C–E) withCASK in the adult mouse hippocampus. A: CASK labeling of themossy fiber projection (green) and dotted Neph1 staining within theregion of mossy fiber terminals and CA3 pyramidal cells (red).B: Higher magnification from (A) showing colocalization of CASK(green) and Neph1 (red) within the region of mossy fiber terminals(yellow). Within the CA1 region of the hippocampus, Neph2 (C, red)and CASK (D, green) are colocalized in the somata and in the den-drites of CA1 pyramidal neurons (E, yellow). F: Interaction of thecytoplasmatic domians of Neph proteins with the PDZ domain ofCASK in overexpressing HEK 293T cells. The PDZ domain of CASK

(mYFP.CASK479–578) was precipitated by the cytoplasmic domains ofNeph1 (F.Neph1-cyt) and Neph2 (F.Neph2-cyt), respectively, but notby an equally tagged control protein (F.EPS15L). G,H: Interactionbetween Neph2 and CASK in vivo. Lysates prepared from mousekidneys (F) or brains (G) were incubated with anti-Neph2, anti-nephrin, and anti-HA antiserum or rabbit IgG and subsequentlyprecipitated with protein A. Immobilized CASK was detected by West-ern blot analysis. Nephrin (positive control) and Neph2 but not HA orrabbit IgG immobilize CASK, suggesting that Neph2 and CASK in-teract in vivo. Scale bars � 60 �m in C–E; 30 �m in A; 15 �m in B.

As observed for Neph1 in the CA1 region of the hip-pocampus Neph2 (Fig. 4C, red) and CASK (Fig. 4D, green)colocalize in the somata and in the dendrites of CA1 py-ramidal neurons (Fig. 4E, yellow).

Neph2 interacts with CASK

In overexpressing HEK 293T cells, both Neph1 andNeph2 interact with CASK (Fig. 4F). The PDZ domain ofCASK (mYFP.CASK479-578) is precipitated by the cyto-plasmic domains of Neph1 (F.Neph1-cyt) and Neph2(F.Neph2-cyt), respectively, but not by an equally taggedcontrol protein (F.EPS15L). In lysates prepared frommouse kidneys Neph2, but not anti-HA antiserum, precip-itates CASK (Fig. 4G). Antinephrin antiserum served aspositive control. Similarly, Neph2 immobilizes CASK inmouse brain lysates, but rabbit IgG does not (Fig. 4H).

Expression of Neph1 messenger RNA(mRNA) in the developing mouse brain

The distribution of Neph1 mRNA in the developingmouse brain was determined by in situ hybridization.Neph1 expression starts around embryonic day (E)12 inthe developing forebrain and at the base of the thirdventricle. The strongest signal at that time is observed inthe developing posterior lobe of the pituitary (Fig. 5A). AtE16, Neph1 is detected in various brain regions includingthe developing cortex (Fig. 5B), the hippocampus (Fig.5C), and the midbrain. At that time point, a strong signalis also visible in dorsal root ganglia and a weaker signal inthe spinal cord (Fig. 5D). There is strong labeling of thedeveloping kidney serving as a positive control.

Expression of Neph2 mRNA in thedeveloping mouse brain

The distribution of Neph2 mRNA in the developingmouse brain was determined by in situ hybridization.Neph2 expression starts around E11 in various brain re-gions. The strongest signals are detectable in the develop-ing diencephalon, rhombencephalon, and in the spinalcord (Fig. 5E). At E16, there is strong labeling of thedeveloping midbrain and forebrain (Fig. 5F), particularlyin the developing cortex (Fig. 5G). At the same time point,strong Neph2 signals are detectable in the spinal cord andin dorsal root ganglia (Fig. 5H).

DISCUSSION

Proper CNS development requires neuronal migrationand axonal outgrowth. These processes are guided by var-ious attractive and repellent molecules including Ig super-family proteins (Rougon and Hobert, 2003). Moreover, Igmolecules have been reported to play a major role insynapse formation both during development and in syn-aptic plasticity in the adult brain (Schachner, 1997; Clan-dinin and Zipursky, 2002). The Neph subgroup of the Igsuperfamily consists of three members, Neph1–3, whichare highly conserved through evolution. Neph proteins areknown to interact with nephrin at the renal podocyte footprocesses, a structure displaying high plasticity. This in-teraction seems to be crucial for the formation of the renalslit diaphragm (Gerke et al., 2003; Liu et al., 2003; Gerkeet al., 2005). However, previous studies also revealed highexpression levels of Neph1–3 and nephrin in the mousebrain (Putaala et al., 2001; Sellin et al., 2003; Gerke et al.,2005; Tamura et al., 2005).

In our study, we investigated the mRNA expressionpatterns of Neph1 and Neph2 in the mouse CNS duringdevelopment. In addition, Neph1 and Neph2 protein ex-pression was studied at the light- and the ultrastructurallevel in the adult mouse brain. We found Neph1 andNeph2 to be highly expressed in various brain regionsduring mouse CNS development. mRNA expression wasparticularly high in regions of ongoing neuronal migra-tion. This widespread distribution of Neph1 and Neph2mRNA during development does only partly overlap (e.g.,in the hippocampus) with the protein expression patternin the adult brain. Here, protein expression of Neph1 andNeph2 showed a more restricted pattern. The strongestlabeling was found in the hippocampus, the cerebellum,and in glomeruli of the olfactory bulb. The latter regionsare known to display ongoing adult neurogenesis as wellas high levels of synaptic plasticity.

The ultrastructural localization of Neph1 as detected byimmunogold labeling was somewhat surprising. The anti-serum directed to the ectodomain of the protein detectedNeph1 not only at cell membranes but also within den-drites and axons. We speculate that there is continuoustransport of Neph1 along cell processes to specific plasmamembrane regions. The observation that membrane-anchored Neph1 is not predominant might be due to cleav-age of the ectodomain, a process referred to as shedding.Shedding is a common mechanism in the Ig superfamily toinduce local release of physiologically active ectodomains,signal termination, or rearrangement of cell contacts (Ar-ribas and Merlos-Suarez, 2003). In fact, shedding hasalready been documented for Neph2 and nephrin (Dou-blier et al., 2003; Gerke et al., 2005).

Neph proteins in submammalian species have been im-plicated in axonal guidance, cell recognition, and synapseformation. The Drosophila Neph ortholog IrreC-rst is ex-pressed in the larval eye-antenna disc, imaginal disc, andouter optic anlagen. Accordingly, the loss of function mu-tant displays severe projection errors of visual axons inthe first and second optic chiasms as well as defects in eyepatterning (Schneider et al., 1995; Reiter et al., 1996). InCaenorhabditis elegans, the Neph homolog SYG-1 is tran-siently expressed on presynaptic axons during synapseformation. Its accumulation is essential to drive synapseformation from the motor neuron onto target cells. Thisaccumulation is mediated by heterophilic interactionswith a nephrin ortholog (SYG-2) (Shen et al., 2004). Theobserved expression pattern for Neph1 and Neph2 in asubset of glomeruli of the olfactory bulb, in particular,points to a possible role of these proteins in axon guidance.Scattered olfactory receptor cells (ORCs) in the olfactoryepithelium synapse precisely with the same few glomeruliin the olfactory bulb. Although the mechanisms underly-ing this convergence are only partly understood, other Igsuperfamily members (e.g., Robo and L1) have been re-ported to play a role in olfactory axon guidance (Hivert etal., 2002; Knafo et al., 2005). Thus, the expression pat-terns of Neph1 and Neph2 suggest a possible role of theseproteins in glomeruli specific synapse formation of ORCs.

Although we were not able to show a clear, exclusiveassociation of Neph1 with the pre- or postsynaptic mem-branes in the adult mouse hippocampus by electron micros-copy, some of our findings support a role for mammalianNeph proteins in synapse formation. The temporal and spa-tial expression levels of both Neph proteins were highest inareas with a high degree of synaptic plasticity. Furthermore,



Fig. 5. In situ hybridization for Neph1 and Neph2 in the develop-ing mouse brain. A: Neph1 expression starts around day E12 in thedeveloping forebrain and at the base of the third ventricle. The stron-gest signal at that time is observed in the developing posterior lobe ofthe pituitary (arrowhead). B: At day E16, Neph1 is detected in variousbrain regions including the developing cortex, hippocampus (C) andmidbrain. At that time point, a strong signal is also visible in dorsalroot ganglia (D, arrowhead) and a weaker signal in the spinal cord.

E: Neph2 expression starts around day E11 in various brain regions.The strongest signals are detectable in the developing diencephalon,rhombencephalon, and in the spinal cord. F: At day E16, a stronglabeling in the developing midbrain and forebrain, especially in thedeveloping cortex (G), is visible. At the same time point, strong Neph2signals are detectable in the spinal cord and in dorsal root ganglia(arrowhead, H). Scale bars � 1 mm in A,B,E,F; 300 �m in D,H; 120�m in C,G.


Neph1 and Neph2 colocalized with the scaffolder CASK ataxonal and dendritic projections. Because all Neph proteinscontain a conserved C-terminal PDZK-1 binding site, wetested for interactions with the PDK-domain containingCASK. Indeed, Neph2 precipitated CASK from brain andrenal tissues. Furthermore, our overexpression experimentswith the cytoplasmic tails of both Neph1 and Neph2 revealedspecific interactions with PDZ domain of CASK.

CASK is a multidomain scaffolding protein of themembrane-associated guanylate kinase (MAGUK) family.In the nervous system, it binds to neurexin, a transmem-brane protein localized in the presynaptic membrane(Hata et al., 1996), as well as to Mint1 and Veli3, whichhave been associated with synaptic vesicle exocytosis(Leonoudakis et al., 2004; Zhang et al., 2001). It is ofinterest that the Drosophila ortholog of Mint1 also has arole in signal transduction during synapse formation anddirectly interacts with IrreC-rst (Ashley et al., 2005;Vishnu et al., 2006). CASK is also a cytoplasmic-bindingpartner for the synaptic cell surface heparan sulfate pro-teoglycans syndecan-2 and -3, and its subcellular distri-bution shifts from a primarily axonal distribution in thefirst 2 postnatal weeks to a somatodendritic distributionin adult brain (Hsueh and Sheng, 1999). Therefore, CASKis believed to play an important role in recruiting synapticproteins into functional complexes.

Considering the widespread distribution of Neph1 inthe developing CNS, it is surprising that the Neph1knockout mouse does not display a neuronal phenotype.We presume redundant functions of Neph1, Neph2, andpossibly Neph3, which is supported by the great overlap inexpression patterns. Similar findings have been reportedfor IrreC-rst and kirre (Strunkelnberg et al., 2001).

Taken together, our results point at Nephs as mediatorsof axonal guidance and synapse formation in certain areasof the CNS. Further studies will be required to address theprecise function of Neph proteins in neuronal communi-cation and the role of the Neph-nephrin interactions forneuronal development.

ACKNOWLEDGMENTS

The authors thank Barbara Joch, Petra Stunz, andMarianne Petry for skillful technical assistance and mem-bers of the Benzing and Walz laboratories for helpfuldiscussions.

LITERATURE CITED

Arribas J, Merlos-Suarez A. 2003. Shedding of plasma membrane proteins.Curr Top Dev Biol 54:125–144.

Ashley J, Packard M, Ataman B, Budnik V. 2005. Fasciclin II signals newsynapse formation through amyloid precursor protein and the scaffold-ing protein dX11/Mint. J Neurosci 25:5943–5955.

Benzing T. 2004. Signaling at the slit diaphragm. J Am Soc Nephrol15:1382–1391.

Clandinin TR, Zipursky SL. 2002. Making connections in the fly visualsystem. Neuron 35:827–841.

Donoviel DB, Freed DD, Vogel H, Potter DG, Hawkins E, Barrish JP, MathurBN, Turner CA, Geske R, Montgomery CA, Starbuck M, Brandt M, GuptaA, Ramirez-Solis R, Zambrowicz BP, Powell DR. 2001. Proteinuria andperinatal lethality in mice lacking NEPH1, a novel protein with homologyto NEPHRIN. Mol Cell Biol 21:4829–4836.

Doublier S, Salvidio G, Lupia E, Ruotsalainen V, Verzola D, Deferrari G,Camussi G. 2003. Nephrin expression is reduced in human diabeticnephropathy: evidence for a distinct role for glycated albumin andangiotensin II. Diabetes 52:1023–1030.

Gerke P, Huber TB, Sellin L, Benzing T, Walz G. 2003. Homodimerizationand heterodimerization of the glomerular podocyte proteins nephrinand NEPH1. J Am Soc Nephrol 14:918–926.

Gerke P, Sellin L, Kretz O, Petraschka D, Zentgraf H, Benzing T, Walz G.2005. NEPH2 is located at the glomerular slit diaphragm, interactswith nephrin and is cleaved from podocytes by metalloproteinases.J Am Soc Nephrol 16:1693–1702.

Hata Y, Butz S, Sudhof TC. 1996. CASK: a novel dlg/PSD95 homolog withan N-terminal calmodulin-dependent protein kinase domain identifiedby interaction with neurexins. J Neurosci 16:2488–2494.

Hivert B, Liu Z, Chuang CY, Doherty P, Sundaresan V. 2002. Robo1 andRobo2 are homophilic binding molecules that promote axonal growth.Mol Cell Neurosci 21:534–545.

Hsueh YP, Sheng M. 1999. Regulated expression and subcellular localiza-tion of syndecan heparan sulfate proteoglycans and the syndecan-binding protein CASK/LIN-2 during rat brain development. J Neurosci19:7415–7425.

Huber TB, Schmidts M, Gerke P, Schermer B, Zahn A, Hartleben B, SellinL, Walz G, Benzing T. 2003. The carboxyl terminus of Neph familymembers binds to the PDZ domain protein zonula occludens-1. J BiolChem 278:13417–13421.

Khoshnoodi J, Sigmundsson K, Ofverstedt LG, Skoglund U, Obrink B,Wartiovaara J, Tryggvason K. 2003. Nephrin promotes cell-cell adhe-sion through homophilic interactions. Am J Pathol 163:2337–2346.

Knafo S, Barkai E, Libersat F, Sandi C, Venero C. 2005. Dynamics ofolfactory learning-induced up-regulation of L1 in the piriform cortexand hippocampus. Eur J Neurosci 21:581–586.

Leonoudakis D, Conti LR, Radeke CM, McGuire LM, Vandenberg CA.2004. A multiprotein trafficking complex composed of SAP97, CASK,Veli, and Mint1 is associated with inward rectifier Kir2 potassiumchannels. J Biol Chem 279:19051–19063.

Liu G, Kaw B, Kurfis J, Rahmanuddin S, Kanwar YS, Chugh SS. 2003.Neph1 and nephrin interaction in the slit diaphragm is an importantdeterminant of glomerular permeability. J Clin Invest 112:209–221.

Putaala H, Soininen R, Kilpelainen P, Wartiovaara J, Tryggvason K. 2001.The murine nephrin gene is specifically expressed in kidney, brain andpancreas: inactivation of the gene leads to massive proteinuria andneonatal death. Hum Mol Genet 10:1–8.

Reiter C, Schimansky T, Nie Z, Fischbach KF. 1996. Reorganization ofmembrane contacts prior to apoptosis in the Drosophila retina: the roleof the IrreC-rst protein. Development 122:1931–1940.

Rougon G, Hobert O. 2003. New insights into the diversity and function ofneuronal immunoglobulin superfamily molecules. Annu Rev Neurosci26:207–238.

Schachner M. 1997. Neural recognition molecules and synaptic plasticity.Curr Opin Cell Biol 9:627–634.

Schneider T, Reiter C, Eule E, Bader B, Lichte B, Nie Z, Schimansky T,Ramos RG, Fischbach KF. 1995. Restricted expression of the irreC-rstprotein is required for normal axonal projections of columnar visualneurons. Neuron 15:259–271.

Sellin L, Huber TB, Gerke P, Quack I, Pavenstadt H, Walz G. 2003.NEPH1 defines a novel family of podocin interacting proteins. FASEBJ 17:115–117.

Shen K, Fetter RD, Bargmann CI. 2004. Synaptic specificity is generatedby the synaptic guidepost protein SYG-2 and its receptor, SYG-1. Cell116:869–881.

Strunkelnberg M, Bonengel B, Moda LM, Hertenstein A, de Couet HG,Ramos RG, Fischbach KF. 2001. rst and its paralogue kirre act redun-dantly during embryonic muscle development in Drosophila. Develop-ment 128:4229–4239.

Tamura S, Morikawa Y, Hisaoka T, Ueno H, Kitamura T, Senba E. 2005.Expression of mKirre, a mammalian homolog of Drosophila kirre, inthe developing and adult mouse brain. Neuroscience 133:615–624.

Venugopala Reddy G, Reiter C, Shanbhag S, Fischbach KF, Rodrigues V.1999. Irregular chiasm-C-roughest, a member of the immunoglobulinsuperfamily, affects sense organ spacing on the Drosophila antenna byinfluencing the positioning of founder cells on the disc ectoderm. DevGenes Evol 209:581–591.

Vishnu S, Hertenstein A, Betschinger J, Knoblich JA, Gert de Couet H,Fischbach KF. 2006. The adaptor protein X11Lalpha/Dmint1 interactswith the PDZ-binding domain of the cell recognition protein Rst inDrosophila. Dev Biol 289:296–307.

Zhang Y, Luan Z, Liu A, Hu G. 2001. The scaffolding protein CASKmediates the interaction between rabphilin3a and beta-neurexins.FEBS Lett 497:99–102.



Documents

Neuronal expression and interaction with the synaptic protein …material.bccn.uni-freiburg.de/publications-bccn/2006/Gerke06_466.pdf · In the adult brain, Neph1 and Neph2 are predom-