The Journal of Neuroscience, July 1995, 15(7): 4906-4917 Cellular Localization of Synaptotagmin I, II, and Ill mRNAs in the Central Nervous System and Pituitary and Adrenal Glands of the Rat B. Marquhze,’ J. A. Boudier,’ M. Mizuta: N. Inagaki: S. Seine,* and M. Seagar’ 1INSERM U 374, lnstitut Jean Roche, Faculte de Medecine-Nord, 13916 Marseille Cedex 20, France and *Division of Molecular Medicine, Center for Biomedical Science, Chiba University School of Medicine, Chuo-ku, Chiba 260, Japan Three isoforms of synaptotagmin, a synaptic vesicle pro- Structural, biochemical, and physiological data are consistent tein involved in neurotransmitter release, have been char- with a crucial role of synaptotagminin neurotransmitterrelease acterized in the rat, although functional differences be- (reviewed by DeBello et al., 1993). It is one of the major mem- tween these isoforms have not been reported. In situ hy- brane componentsof synaptic vesicles (Matthew et al., 1981; bridization was used to define the localization of synapto- Fournier and Trifaro, 1988) and is exclusively expressedin neu- tagmin I, II, and Ill transcripts in the rat CNS and pituitary rons and neuroendocrine cells. Synaptotagmin binds Ca*+ at and adrenal glands. Each of the three synaptotagmin genes concentrationsin the lo-100 PM range in a phospholipid-de- has a unique expression pattern. The synaptotagmin Ill pendent manner (Brose et al., 1992) and has been postulatedto gene is expressed in most neurons, but transcripts are be the Caz+ sensorwhich inducesexocytosis. much less abundant than the products of the synaptotag- Synaptotagmin interacts in vitro with multimolecular com- min I and II genes. A majority of neurons in the forebrain plexes which include syntaxin/HPC-1 and the N-type Ca2+chan- expressed both synaptotagmin I and Ill mRNAs while syn- nel (Bennett et al., 1992b; LCv&queet al., 1992, 1994; Yoshida aptotagmin II gene expression was confined to subsets of et al., 1992), SNAP 25, synaptobrevin/VAMP (Sijllner et al., neurons in layers IV-VI of the cerebral cortex, in the den- 1993), the a-latrotoxin receptor, and other related proteins of the tate granule cell region, the hilus, and the CAl-CA3 areas neurexin family (Petrenko et al., 1991; Hata et al., 1993; of the hippocampus. In the cerebellum, all three transcripts O’Connor et al., 1993). These complexes are thought to be in- were visualized in the granule cell layer. Furthermore, syn- volved in synaptic vesicle docking at presynaptic active zones aptotagmin I probes revealed striking differences between in close vicinity to the site of Ca2+entry. distinct populations of neurons, as in addition to moderate Furthermore, synaptotagmin may prevent assembly of the labeling of granule cells, much more prominent hybridiza- constitutive core machinery, formed by the associationof N-eth- tion signals were detected on scattered cell bodies likely ylmaleimide-sensitive fusion protein (NSF) with soluble NSF to be Golgi interneurons. In the most caudal part of the attachment proteins (SNAPS), and its putative receptor: the brain, synaptotagmin II transcripts were abundant and VAMP, syntaxin, and SNAP 25 complex (SGllner et al., 1993). were coexpressed with synaptotagmin Ill mRNAs. This pat- Also, synaptotagmin appears to play a role in endocytosis tern was found in putative motoneurons of the spinal cord, through its interaction with the clathrin adaptor AP-2 (Zhang et suggesting that the two isoforms might be involved in exo- al., 1994). cytosis at the neuromuscular junction. Only synaptotagmin Drosophila mutants lacking synaptotagmin display a severe I mRNAs were detected in the anterior and intermediate reduction of evoked neurotransmitterrelease and an increasein pituitary and in adrenal medullary cells. These data reveal the frequency of spontaneousquanta1 events at the neuromus- an unexpectedly subtle segregation of the expression of cular junction (Littleton et al., 1993; DiAntonio et al., 1994). In synaptotagmin genes and the existence of multiple com- the giant synapseof the squid, injection of peptides from syn- binations of synaptotagmin isoforms which may provide aptotagmin into presynaptic terminals (Bommert et al., 1993) diversity in the regulation of neurosecretion. inhibits neurotransmission. [Key wordsr synaptotagmin, neurotransmitter release, Synaptotagmin is expressedas variant isoforms encoded by mRNA distribution, CNS, pituitary gland, adrenal gland] different genesin the rat [synaptotagmin I (Perin et al., 1990); synaptotagminII (Geppert et al., 1991); synaptotagminIII (Mi- Received Oct. 27, 1994; revised Jan. 10, 1995; accepted Jan. 13, 1995. zuta et al., 1994)]. The overall identity of amino acid sequence We thank C. lborra for her expert technical assistance, M. Grino for sug- amongthem is relatively low; however, the region corresponding gestions and precious help with tissue sections, and E. Jover and M. Takahashi for discussion and advice. We are grateful to J.-L. Boudier, E Couraud, and A. to the carboxyl terminal C, repeats is well conserved. Synap- Zamorra for their comments on the manuscriot. totagmin I and II are more closely related to each other than to This research was supported by the InstiLt National de la SantC et de la Recherche MBdicale (INSERM), France, and by the Scientific Research Grants synaptotagminIII. No functional differenceshave been reported of the Ministry of Education, Science and C&re, Japan. between these isoforms. Correspondence should be addressed to B. Marqukze, INSERM U 374, In- Variations in sequencebetween synaptotagminscould lead to stitut Jean Roche, FacultC de MCdecine-Nerd, Boulevard Pierre Dramard, 13916 Marseille Cedex 20, France. differences in docking and fusion providing fine tuning of reg- Copyright 0 1995 Society for Neuroscience 0270.6474/95/154906-12$05.00/O ulated vesicular trafficking required for development and neu- The Journal of Neuroscience, July 1995, 15(7) 4907 Table 1. S-3’ sequences of oligodeoxyribonucleotide probes used for in situ hybridization Probes Sequence Synaptotagmin I TgIa Complementary to bases 1800-l 843 TACTGGCTAAAGAGCACTATGTGGGCAGA- TGCAGAAAGGCTTCG TgIb Complementary to bases 2527-2571 TGAAGCTATGCTAGATGCAGTGGTAGGAA- CGCATTGGCTCCTGTT TgIs Bases 1800-1843 CGAAGCCTTTCTGCATCTGCCCACATAGTG- CTCTTTAGCCAGTA Synaptotagmin II TgIIa Complementary to bases 1923-1966 TITCGCAAGGACTATGAGAGCTTCTGGCCT- CTGACCACTTAAGC TgIIb Complementary to bases 2447-2491 AG-ITGTGAGGAGCTCTGCAATGTCTAGCTT- GTCACTGTCCACCAA TgIIs Bases 1923-1966 GCTTAAGTGGTCAGAGGCCAGAAGCTCTC- ATAGTCCTTGCGAAA Synaptotagmin III TgIIIa Complementary to bases 1809-1853 TTCTCTGACAATCCTTTGCCGCCCTTGGTA- AAGCTGCTTAGAGTC TgIIIb Complementary to bases 1853-1896 GTCCAATCCCAGGCCTAGACCAGACCCTC- ACTCTGAATTCTCTT TgIIIs Bases 1809-1853 GACTCTAAGCAGCTTTACCAAGGGCGGCA- AAGGATTGTCAGAGAA rotransmission. In order to analyze the distribution of the dif- Results ferent isoforms, we have carried out an in situ hybridization Specijcity of probes study of the distribution of synaptotagmin mRNAs in the rat Probes to each of the three rat synaptotagmins produced unique CNS and in the adrenal and pituitary glands. patterns of hybridization in the rat brain (Figs. l-5), spinal cord Materials and Methods (Figs. 6, 7), and adrenal and pituitary glands (Fig. 8). Two probes, complementary to different regions of the same cDNA, Oligonucleotideprobes. Oligonucleotide probes [44&45 base pairs (bp)] were tested for each synaptotagmin. In each case, both members of unique sequence were synthesized (CIML, France) and purified by ethanol precipitation. The probe sequences derived from the rat syn- of the pair produced identical patterns of hybridization. No sig- aptotagmin I (Perin et al., 1990), synaptotagmin II (Geppert et al., nal could be detected with the sense probes TgIs, TgIIs, and 1991), and synaptotagmin III (Mizuta et al., 1994) cDNAs are compiled TgIIIs when used in the same conditions at equivalent specific in Table 1. Sense probes TgIs, TgIIs, and TgIIIs exactly complementary activity (data not shown). Also, no labeling was observed on to TgIa, TgIIa, and TgIIIa antisense probes, respectively, were used as controls. sections hybridized with radiolabeled TgIa and TgIIa in the pres- Probes were 3’-end labeled with 5’-[a - ?S]-dATP (> 1000 Ci/mmol; ence of an excess of the unlabeled oligonucleotide (data not Amersham) to similar specific activities in the range of 2 X 108-1.4 X shown). 10’ dpm/pg using terminal deoxynucleotidyl transferase (Boehringer Mannheim) with a 3O:l molar ratio of dATP:oligonucleotide. Unincor- porated nucleotides were removed by a spin column procedure using In situ hybridization Sephadex G-25 (Pharmacia). In situ hybridization.Nonperfused rat brains were removed and fro- Macroscopic transcript distributions on sagittal and coronal sec- zen on dry ice. Cryostat sections (12 pm) were cut, mounted onto poly- tions of brain (Fig. l), spinal cord (Fig. 6), and adrenal and L-lysine-coated slides, and dried at room temperature. Sections were fixed in 4% paraformaldehyde, rinsed in phosphate-buffered saline pituitary glands (Fig. 8) are shown. Each probe produced re- (PBS), and dehydrated into 95% ethanol for storage at 4”C, or dried gionally distinct hybridization signals. Regions of white matter and stored at -80°C until required. Radiolabeled probe was dissolved such as the corpus callosum were devoid of synaptotagmin in hybridization buffer [50% (v/v) formamide, 4X saline sodium citrate mRNAs. Cellular localizations of synaptotagmin transcripts (SSC: 0.15 M NaCl, 0.015