FEBS Letters 581 (2007) 4727–4733

Mover is a novel vertebrate-specific presynaptic protein withdifferential distribution at subsets of CNS synapses

Thomas Kremer, Christian Kempf, Nina Wittenmayer, Ralph Nawrotzki, Thomas Kuner,Joachim Kirsch, Thomas Dresbach*

Institute for Anatomy and Cell Biology, University of Heidelberg, Im Neuenheimer Feld 307, D-69120 Heidelberg, Germany

Received 23 July 2007; revised 7 August 2007; accepted 29 August 2007

Available online 6 September 2007

Edited by Jesus Avila

Abstract Presynaptic nerve terminals contain scaffolding pro-teins that orchestrate neurotransmitter release at active zones.Here we describe mover, a yet unknown non-transmembrane pro-tein that is targeted to presynaptic terminals when overexpressedin cultured neurons. Confocal immunomicroscopy revealed thatmover colocalizes with presynaptic markers in the calyx of Held.In the hippocampus, mover localizes to mossy fibre terminals, butis absent from inhibitory nerve terminals. By contrast, moverlocalizes to inhibitory terminals throughout the cerebellar cortex.Our results suggest that mover may act in concert with generallyexpressed scaffolding proteins in distinct sets of presynaptic ter-minals.� 2007 Federation of European Biochemical Societies. Pub-lished by Elsevier B.V. All rights reserved.

Keywords: Presynaptic; Hippocampus; Cerebellum;Calyx of Held

1. Introduction

Synapses of the central nervous system are specialized sites

of cell–cell contact. The presynaptic site is designed to allow

for highly regulated, exocytotic neurotransmitter release from

synaptic vesicles (SVs) [1]. The majority of proteins involved in

neurotransmitter release are evolutionarily conserved among

vertebrates and invertebrates and include two types of protein

complexes, which are crucially involved in neurotransmitter

release: first, SNARE-complexes, built from the SV-transmem-

brane protein VAMP and the plasmamembrane proteins syn-

taxin-1 and SNAP25 which mediate exocytotic fusion of SVs

with the plasmamembrane [2]. Second, the cytomatrix of active

zones (CAZ) which is thought to represent a supramolecular

complex serving several functions, including formation of a

scaffold for the recruitment of additional proteins by organiz-

ing presynaptic protein–protein interactions, regulation of

SNARE-complexes, and spatial restriction of SV fusion to

active zones [3–5].

The expression of two related multidomain CAZ-proteins,

namely bassoon and piccolo/aczonin, is restricted to verte-

*Corresponding author. Present address: Department of Medical CellBiology, University of Heidelberg, Im Neuenheimer Feld 307, D-69120Heidelberg, Germany. Fax: +49 6221 544952.E-mail address: thomas.dresbach@urz.uni-heidelberg.de (T. Dresbach).

0014-5793/$32.00 � 2007 Federation of European Biochemical Societies. Pu

doi:10.1016/j.febslet.2007.08.070

brates and therefore may add vertebrate-specific features to

presynaptic nerve terminals [5]. To identify novel vertebrate

specific presynaptic proteins, we performed a yeast-2-hybrid

screen using one of the cognate vertebrate specific CAZ-com-

ponents, namely bassoon, as bait.

Here, we describe the tissue distribution and subcellular

localization of a putative bassoon binding partner, which we

call mover alluding to its association with mossy fibre terminals

and its exclusive expression in vertebrates. We show that

mover is a presynaptic molecule with differential distribution

at distinct sets of excitatory and inhibitory synapses.

2. Materials and methods

2.1. AntibodiesPrimary antibodies: anti-myc (9E10; Santa Cruz), anti-MAP-2, anti-

synapsin, anti-synaptophysin (Sigma), anti-bassoon (Stressgen), anti-synaptotagmin 1, anti-VGAT (Synaptic Systems). Rabbit polyclonalantibodies were raised against mover fused to glutathione-S-transfer-ase (GST). Mover-specific antibodies were purified using two affinitycolumns (Sterogene).

2.2. Yeast-2-hybrid screenScreening was performed using the L40 yeast strain harbouring

HIS3 and b-gal as reporter gene. Nucleotide sequences encoding aminoacids 3263–3938 of rat bassoon were subcloned into the lexA fusionvector pHyblexZeo and used to screen an adult mouse brain cDNAlibrary constructed in pPC86 vector containing the GAL4 activationdomain vector (Invitrogen). Approximately 2 · 107 clones of a mousecDNA library were screened. Positive clones of the initial screen wereisolated, sequenced and retransformed to validate their ability to bindto the respective bassoon fragment.

2.3. ConstructsThe rat mover cDNA was cloned from rat brain total RNA by

RT-PCR with oligonucleotides based on the 5 0- and 3 0-terminal mousesequences, and subcloned into CMV-promotor based mammalianexpression vectors (Clontech). The myc-sequence was fused to theC-terminus of full-length rat mover. Further constructs were synapto-physin-GFP and PSD-95-GFP [6,7].

2.4. Cell culturePrimary cultures of rat hippocampal neurons were prepared, main-

tained and transfected using the calcium phosphate method on DIV 3or 4 as described [8].

2.5. Immunocytochemistry and fluorescence imagingCultured neurons were fixed in 4% paraformaldehyde, processed and

viewed as described [8]. Adult rats were perfused with phosphatebuffered saline (PBS) containing 4% paraformaldehyde, 2% polyvinyl-pyrrolidone, and 0.01% glutaraldehyde. The fixed tissue was kept in

blished by Elsevier B.V. All rights reserved.

4728 T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733

fixative for 1 h at 4 �C and cryoprotected by incubation in PBS withincreasing sucrose concentrations, frozen in embedding medium (Tis-sue Tek) on dry ice and sectioned at 20 lm on a cryostat. Confocalimages were acquired using a Leica TSC SP confocal microscope.

2.6. Subcellular fractionation of adult rat brainFractionation was performed as described [9,10]. The individual

fractions were designated as follows: Hom, homogenate; P1, nuclearpellet; S1, supernatant after synaptosome sedimentation; P2, crudesynaptosomal pellet; P3, light membrane pellet; S3, cytosolic fraction;LP2, crude synaptic vesicle fraction; LS2, cytosolic synaptosomal frac-tion; SPM, synaptic plasma membranes.

2.7. Synaptotagmin uptake assayThis assay identifies functional neurotransmitter release sites by

stimulation-dependent uptake of an antibody that binds to the lume-nal domain of the synaptic vesicle transmembrane protein synapto-tagmin-1 [11]. Hippocampal neurons transfected with a myc-taggedmover construct were incubated for 90 s with anti-synaptotagmin-1antibody in depolarisation buffer (44 mM NaCl, 90 mM KCl,2 mM CaCl2, 1 mM MgCl2, 20 mM HEPES pH 7.4, 30 mM glu-cose) followed by a 10 min incubation with anti-synaptotagmin-1antibody in culture medium at 37 �C. After washing and fixation,the internalized antibodies were detected using anti-mouse secondaryantibodies.

Fig. 1. Sequence alignment of rat (r-mover) with mouse (m-mover), humanComputational analysis predicts no known functional domains or conse(Phosphothreonine).

2.8. Northern blottingA premade multiple tissue Northern blot (BD) was hybridized with

32P-labelled mover cDNA for 1 h at 68 �C, washed three times with2·SSC, 0.05% SDS, followed by two washes of 1·SSC, 0.1% SDS at50 �C. To check for equal loading, the blot was stripped and re-probedwith radioactively labelled human b-actin cDNA.

3. Results

The most abundant prey in the yeast-2-hybrid screen was a

cDNA encoding a 266 amino acid protein, which we propose

to name mover. Murine mover cDNA is found in the NCBI

database as RIKEN cDNA 1200015A19 gene encoding a hypo-

thetical protein (GenBank accession number NP_080664).

Mover homologues exist in vertebrates, but not in invertebrates.

Because rat mover sequences were not deposited, we used

RT-PCR to clone a rat mover cDNA, which encodes a protein

with 99% amino acid sequence identity to murine mover

(Fig. 1). Sequence comparison of several homologues revealed

a divergent N-terminal region followed by evolutionary

conserved regions. Mover has no predicted transmembrane

domains or putative membrane anchor sites (Fig. 1).

(h-mover), zebrafish (z-mover) and Xenopus laevis (x-mover) mover.rved motifs. Mover has a predicted phosphorylation site at aa13

Fig. 2. Mover expression in adult rat tissue. (A) Northern Blot analysis. An actin-cDNA probe was used as a loading control. Marker lane: 2.4 kb.(B) Mover protein expression (10 lg protein per lane). In skeletal muscle, a weak 40 kD band is detectable (asterisk).

Fig. 3. Subcellular fractions of rat brain homogenate analyzed byWestern Blot using affinity purified mover antibodies. Homogenate(Hom); nuclei (P1); supernatant after synaptosome sedimentation (S1);crude synaptosomal pellet (P2); light membrane pellet (P3); cytosolicfraction (S3); cytosolic synaptosomal fraction (LS2); crude synapticvesicle fraction (LP2); synaptic plasma membranes (SPM).

T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733 4729

Northern blot analysis revealed the presence of mover

mRNA in brain, and also in heart, liver, kidney and testis

(Fig. 2A). We generated an antiserum against mover, which

detected a single band in Western blots of tissue homogenates

and lysates prepared from cultured neurons (Fig. 2B and data

not shown). Protein levels of mover are highest in the brain.

Mover is also expressed in testis, while there is relatively weak

expression in heart, spleen and liver (Fig. 2B). Upon subcellu-

lar fractionation of adult rat brain homogenate, mover be-

haved similar to the SV-marker protein synaptophysin being

abundant in the LP2 fraction, which is enriched for synaptic

vesicles. Mover was not enriched in the SPM-fraction contain-

ing synaptic plasma membranes as well as pre- and postsynap-

tic scaffolding proteins such as PSD-95 (Fig. 3). Moreover,

mover levels were lower in cytosolic protein fractions than in

all other fractions. These results suggest that mover may be

a protein of presynaptic nerve terminals.

To further test this hypothesis, we performed immunofluo-

rescence localization studies. In cultured rat hippocampal neu-

rons, mover was diffusely distributed in the somatodendritic

region but enriched in puncta, which co-localized with synap-

tophysin (Fig. 4A). Synaptotagmin antibody-uptake assays

identified these puncta as functional neurotransmitter release

sites (Fig. 4B). Moreover, like the synaptic vesicle protein syn-

aptophysin but unlike the CAZ protein bassoon, mover could

be extracted by Triton X-100 treatment of live cultured neu-

rons Fig. S3 [8].

We next generated recombinant versions of mover fused to

N- or C-terminal tags, including myc, GFP and mCherry.

When co-expressed in HEK293-cells recombinant mover and

bassoon constructs co-aggregated. In co-immunoprecipita-

tion attempts recombinant mover turned out to be largely

insoluble, however (data not shown). This observation pre-

cluded additional co-precipitation attempts.

Fig. 4. Synaptic localization of mover in cultured hippocampal neurons. (A) Neurons were fixed and double-immunolabelled for mover andsynaptophysin (SyPh). (B) Neurons were fixed after stimulation in the presence of anti-synaptotagmin antibodies (to load active synapses with theantibody) and immunolabelled for mover. Bar, 30 lm.

4730 T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733

To test the targeting behavior of recombinant mover in a

neuronal environment, we expressed myc-tagged mover in cul-

tured neurons and analyzed its trafficking in individual cells. In

neurons cultured for more than 9 days, myc-tagged mover was

concentrated in synapsin containing puncta at various dis-

tances from the soma (Fig. 5A). Upon co-expression, mover-

myc and synaptophysin-GFP displayed identical distribution

patterns outside the somatodendritic region and complete

overlapping of mover-myc, synaptophysin-GFP and endoge-

nous synapsin immunoreactivities (Fig. 5B). By contrast, when

co-expressed with the postsynaptic marker construct PSD-95-

GFP, divergent distributions were observed. Interestingly, no

enrichment of mover-myc and PSD-95 puncta was observed,

even in cells expressing high levels of mover-myc, where a uni-

form distribution of mover immunoreactivity (IR) in the soma-

dendritic compartment was observed (Fig. 5C). From these

Fig. 5. Presynaptic targeting of mover-myc in cultured hippocampalneurons. Neurons were fixed and immunolabelled with anti-mycantibody. (A) Punctate mover-myc immunoreactivity colocalizes withendogenous synapsin. (B) Both mover-myc and synaptophysin-GFPco-localize with endogenous synapsin immunoreactivity in a double-transfected neuron. (C) Mover-myc immunoreactivity and PSD-95-GFP do not co-localize in a double-transfected neuron (arrows). Bars,30 lm.

T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733 4731

data we conclude that recombinant mover is preferentially tar-

geted to presynaptic sites.

To investigate whether mover is localized to synapses in the

brain, we immunostained brain regions harboring well-charac-

terized types of presynaptic nerve terminals, including the ca-

lyx of Held in the auditory brain stem, mossy fibre boutons

in the hippocampus and mossy fibre terminals in the cerebel-

lum. In the calyx of Held, a giant synaptic terminal that forms

around the soma of principal cells in the medial nucleus of the

trapezoid body (MNTB), mover immunoreactivity (IR) is en-

riched in the presynaptic compartment and showed a punctate

staining that co-localized with synaptophysin IR (Fig. 6A) but

was absent from the soma of the postsynaptic cell.

In the hippocampal CA3 region, mover IR was enriched in

the stratum lucidum, the region harboring mossy fibre termi-

nals and not in the stratum pyramidale, harboring the cell

bodies of pyramidal neurons (Fig. 6B–D). Higher magnifica-

tion revealed punctate mover IR along MAP-2 stained

dendrites (Fig. 6B), which co-localized with that of synapto-

physin (Fig. 6C) and VGLUT1 (data not shown). Mover IR

did not colocalize with VGAT suggesting its absence from

inhibitory synapses in the CA3 region (Fig. 6D). By contrast,

in the cerebellum VGAT-positive puncta co-localized with

mover IR (Fig. 6E). VGAT positive presynaptic terminals of

basket and stellate cells and of recurrent Purkinje cells axons

providing inhibitory input to the molecular layer and Purkinje

cell layer co-localized extensively with mover IR. Likewise, in

the granular layer, VGAT co-localized with mover at synaptic

glomeruli (Fig. 6E), harbouring inhibitory presynaptic termi-

nals of Golgi cells [12], but not with VGLUT1 (data not

shown). Thus, mover appears to be absent from inhibitory

nerve terminals in the hippocampal CA3-region, whereas it is

present at inhibitory terminals in all layers of the cerebellar

cortex.

4. Discussion

We identified a hitherto unknown protein, which is specifi-

cally expressed in vertebrates and found in many tissues, pre-

dominantly brain. Subcellular fractionation of brain tissue

suggests that the novel protein, which we call mover, is en-

riched in the synaptic vesicle fraction. This notion is further

corroborated by targeting and colocalisation studies. Mover

IR gives a synaptic staining pattern at several well-character-

ized synapses of the brain. The observation that the recombi-

nant protein has presynaptic, but not postsynaptic targeting

capacity strongly suggests that the staining observed in the

brain reflects presynaptic localization. This is supported fur-

ther by the colocalization of mover with presynaptic markers

in the calyx of Held and the absence of mover IR in the soma

of cognate postsynaptic cells.

Although no transmembrane region or putative sites for

membrane anchoring are predicted from the primary structure,

mover was enriched in the synaptosomal fraction upon bio-

chemical fractionation and localized to synapses both in neu-

ronal cultures and the brain. These data are consistent with

the assumption that mover binds to synaptic organelles such

as SVs.

Furthermore, mover bound to a C-terminal portion of bas-

soon in yeast, co-localized with bassoon in neurons and the

Fig. 6. Differential localization of mover to subsets of synapses in the brain. A–E show an overview for the double-staining for mover and therespective markers in different areas of the brain. (A) Calyx of Held. (B–D) CA3 region of rat hippocampus. (E) Cerebellar cortex. Double-stainingfor mover and MAP2 (B), mover and synaptophysin (A, C) or mover and VGAT (D, E). SR, stratum radiatum, SL, stratum lucidum, SP, stratumpyramidale. GL, granular layer, PJ, Purkinje cell layer, ML, molecular layer. A 0–E 0 show corresponding merged images and the respective singlechannels at higher magnification (2.5·). The images shown in lane B 0 and C 0 are derived from the corresponding field. Bars, 30 lm.

4732 T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733

brain, and co-aggregated with recombinant bassoon constructs

in HEK293-cells. These data are consistent with the assump-

tion that mover can bind to bassoon in vivo. As recombinant

mover is insoluble, it was not possible to test this assumption

by co-immunoprecipitation, however. The sensitivity of mover

IR to Triton X-100 extraction of live cultures indicates that

mover is not an integral component of the CAZ-network,

but may rather be a vertebrate-specific CAZ-associated pro-

tein. This would be consistent with the fact that mover is pres-

ent but not concentrated in the synaptic plasma membrane

fraction, which is enriched for CAZ-proteins [9].

Surprisingly, mover is differentially localized to subsets of

synapses. Whereas in CA3 it is found at excitatory synapses,

in cerebellar cortex mover is found at inhibitory terminals.

Interestingly, several pan-neuronally expressed presynaptic

proteins seem to have distinct roles at excitatory and inhibitory

synapses, respectively. For example, knock-outs of the genes

for synapsin or RIM affect excitatory and inhibitory transmis-

sion in different ways, although these proteins are present at

both types of synapses [13,14]. Similarly, bassoon is expressed

at excitatory and inhibitory synapses throughout the brain, but

deletions in the bassoon gene cause silencing of only a fraction

of synapses, as well as an epileptic phenotype, suggesting an

imbalance between excitation and inhibition in these animals

[15]. Mover could account for such effects by functioning as

an effector for CAZ-proteins in specific sets of synapses.

Acknowledgements: We are grateful to Karin Gorgas and Gabi Kra-mer. This work was supported by the DFG Grants Ki 339/10-1 (to.J.K.) and DR373/3-2 (to T.D.).

T. Kremer et al. / FEBS Letters 581 (2007) 4727–4733 4733

Appendix A. Supplementary data

Supplementary data associated with this article can be

found, in the online version, at doi:10.1016/j.febslet.2007.08.

070.

References

[1] Sudhof, T.C. (2004) The synaptic vesicle cycle. Annu. Rev.Neurosci. 27, 509–547.

[2] Jahn, R. and Scheller, R.H. (2006) SNAREs – engines formembrane fusion. Nat. Rev. Mol. Cell Biol. 7, 631–643.

[3] Dresbach, T., Qualmann, B., Kessels, M.M., Garner, C.C. andGundelfinger, E.D. (2001) The presynaptic cytomatrix of brainsynapses. Cell Mol. Life Sci. 58, 94–116.

[4] Rosenmund, C., Rettig, J. and Brose, N. (2003) Molecularmechanisms of active zone function. Curr. Opin. Neurobiol. 13,509–519.

[5] Schoch, S. and Gundelfinger, E.D. (2006) Molecular organizationof the presynaptic active zone. Cell Tissue Res. 326, 379–391.

[6] Kaether, C., Skehel, P. and Dotti, C.G. (2000) Axonal membraneproteins are transported in distinct carriers: a two-color videomicroscopy study in cultured hippocampal neurons. Mol. Biol.Cell 11, 1213–1224.

[7] Bresler, T., Ramati, Y., Zamorano, P.L., Zhai, R., Garner, C.C.and Ziv, N.E. (2001) The dynamics of SAP90/PSD-95 recruitmentto new synaptic junctions. Mol. Cell Neurosci. 18, 149–167.

[8] Dresbach, T., Hempelmann, A., Spilker, C., tom Dieck, S.,Altrock, W.D., Zuschratter, W., Garner, C.C. and Gundelfinger,E.D. (2003) Functional regions of the presynaptic cytomatrix

protein bassoon: significance for synaptic targeting and cytoma-trix anchoring. Mol. Cell Neurosci. 23, 279–291.

[9] Kalla, S., Stern, M., Basu, J., Varoqueaux, F., Reim, K.,Rosenmund, C., Ziv, N.E. and Brose, N. (2006) Moleculardynamics of a presynaptic active zone protein studied in Munc13-1-enhanced yellow fluorescent protein knock-in mutant mice. J.Neurosci. 26, 13054–13066.

[10] Huttner, W.B., Schiebler, W., Greengard, P. and De Camilli, P.(1983) Synapsin I (protein I), a nerve terminal-specific phospho-protein. III. Its association with synaptic vesicles studied in ahighly purified synaptic vesicle preparation. J. Cell Biol. 96, 1374–1388.

[11] Kraszewski, K., Mundigl, O., Daniell, L., Verderio, C., Matteoli,M. and De Camilli, P. (1995) Synaptic vesicle dynamics in livingcultured hippocampal neurons visualized with CY3-conjugatedantibodies directed against the lumenal domain of synaptotagmin.J. Neurosci. 15, 4328–4342.

[12] Chaudhry, F.A., Reimer, R.J., Bellocchio, E.E., Danbolt, N.C.,Osen, K.K., Edwards, R.H. and Storm-Mathisen, J. (1998) Thevesicular GABA transporter, VGAT, localizes to synaptic vesiclesin sets of glycinergic as well as GABAergic neurons. J. Neurosci.18, 9733–9750.

[13] Gitler, D., Takagishi, Y., Feng, J., Ren, Y., Rodriguiz, R.M.,Wetsel, W.C., Greengard, P. and Augustine, G.J. (2004) Differentpresynaptic roles of synapsins at excitatory and inhibitorysynapses. J. Neurosci. 24, 11368–11380.

[14] Schoch, S. et al. (2002) RIM1alpha forms a protein scaffold forregulating neurotransmitter release at the active zone. Nature 415,321–326.

[15] Altrock, W.D. et al. (2003) Functional inactivation of a fractionof excitatory synapses in mice deficient for the active zone proteinbassoon. Neuron 37, 787–800.