Volume 292, n u m b e r 1,2, 111-I 14 FEBS 10331 © 1991 Federation of European Biochemical Societies 00145793/91/$3.50 ADONIS 001457939101019D
November 1991
Cloning of a c D N A that encodes an invertebrate glutamate receptor subunit Michael L. H u t t o n , R o b e r t J. Harvey*, Eric A. B a r n a r d and M a r k G. Darlison* M R C Molecular Neurobiology Unit, M R C Centre, Hills Road, Cambridge CB2 2Qtt, UK Received 1 A u g u s t 1991 A full-length eDNA which encodes a putative glutamate receptor polypeptide was isolated from the pond snail Lymnaea stagnalis, using a short stretch of exonic sequence and two variants of the polymerase chain reaction. In this first comparison of invertebrate and vertebrate glutamate receptor sequences, the mature molluscan polypeptide, which comprises 898 amino acids and has a predicted M~ of 100 913, displays between 37% and 46% amino-acid identity to the rat ionotropie glutamate receptor subunits, GIuRI to GIuR6.
eDNA cloning; Gene cloning; Glutamate receptor; Invertebrate receptor; Lymnaea stagnalis; Polymerase chain reaction
1. I N T R O D U C T I O N In vertebrates, the excitatory post-synaptic effects of glutamate are mediated by ligand-gated cation channels (ionotropic receptors), the receptor types of which can be distinguished by the agonists N-methyl-D-asparrate (NMDA), c~-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA), and kainate (KA); glutamate can also modulate the release of intracellular Ca 2+ through binding to a metabotropic receptor (reviewed in [1]). In contrast, in invertebrates, L-glutamate can act both as an ex.citatory and an inhibitory neurotransmitter. For example, the effect of L-glutamate, when applied to moll ascan neurones, is to induce either a fast depolarizing cationic current, a slow hyperpolarizing K* current, or a fast hyperpolarizing C1- current [2]. Recently, molecular biological studies have led to the isolation of 6 different vertebrate non-NMDA ionotropic glutamate receptor subunit cDNAs [3-8]. When these cDNAs were expressed singly either in Xenopus oocytes [3,5-8] or in mammalian cells [4], homo-oligo-
Abbreviations: AMPA, =-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; eDNA, complementary DNA; KA, kainate; NMDA, N-methyl-D-aspartate; PCR, polymerase chain reaction. *Present address: Institut fttr Zellbiochemie und klinische Neurobiologie, Universitats-Krankenhaus Eppcndorf, Universitat Hamburg, Martinistrasse 52, 2000 Hamburg 20, Germany. Reprint address." M.L. Hutton, MRC Molecular Neurobiology Unit, MRC Centre, Hills Road, Cambridge CB2 2QH, UK. Correspondence address." M G . Darlison, lnstitut ftlr Zellbi0chemie und Neurobiologie, Universitats-Krankenhaus Eppendorf, Universit'at Hamburg, Martinistrasse 52, 2000 Hamburg 20, Germany.
Published by Elsevier Science Pt¢blishers B. V.
meric receptors were formed that were AMPA-selective (GIuRI to GluR4), KA-selective (GIuR6) or responsive to L-glutamate alone (GIuR5). No sequence has yet been reported for any invertebrate glutamate receptor. We describe here the isolation o f genomie and eDNA sequences that encode a putative glutamate receptor polypeptide from the fresh-water snail Lymnaea stagna-
lis. 2. MATERIALS A N D METHODS 2.1. Isolation of a molluscan ghttamate receptor genomic clone 5 x 10s clones of an amplified L. stagnalis genomic library, constructed in ~.EMBL3, were screened with an ,'~700 bp polymerase chain reaction (PCR)-derived eDNA fragment that encodes the first three putative membrane-spanning domains (TMI to TM3) of the rat GIuRI subunit [3]. Hybridization was in 6 x SSC (1 x SSC = 0.15 M NaCI, 0.015 M sodium citrate, pH 7), 0.5% (w/v) SDS, 5 x Denhardt's solution and 100 btg/ml yeast tRNA, at 50°C for 48 h; the final wash conditions were 1 x SSC, 0.1% (w/v) SDS at 55°C for 30 rain. The library filters were then stripped and hybridized, under identical conditions, with an ,'-,1.8 kb EcoRl eDNA fragment that encodes part of a locust putative glutamate receptor polypeptide (unpublished results). Only one clone hybridized to both probes; a 166 bp Sau3A~ fragment from this was subsequently subcloned and sequenced. 2.2. Isolation of molluscan fidl-length gh~tamate receptor cDNAs Full-length cDNAs were generated using two variants of the PCR. The 3' end of the eDNA was amplified (see Fig. 1) using nested primers [9]. For this, first-strand eDNA was synthesized from L. stagrlalis egg mass poly(A)+ RNA using oligonucleotide RoRidTi7: 5'-ATCGATGGTCGACGCATGCGGATCCAAAGCTTGAATTCGAGCTC'fl 1 l-l-l-l-I 1-1 1 ~ - 3 ' . First.stage PCR was then performed using primers Ro: 5"-ATCGATGGTCGACGCATGCGGA. TCC-Y, which eorrespond~ to part of RoR~dT~7, and LGLI: 5'TGTCGGGAGAATTCGTCGACTCAGTCTGGTGGTTCTTCACGCTCA-Y, which is based on exonic sequence from the 166 bp Sau3A~ fragment. Reactions (50 gl) contained: 50 mM KCI, I0 mM Tris-HCI (pH 9 at 25°C), i.5 mM MgCI~, 200/.tM each dNTP, 0.1% (v/v) Triton X.100, 0.01% (w/v) gelatin, 1 nmol ofT4 gene 32 protein (United States Biochemical), "-,20 ng of first-strand eDNA, 25 pmol
111
V o l u m e 292, nutnber t,2
FEBS LE'I'I'ERS
ofeach primer, and 2.5 units of T~iermusaquaticus DNA polymerase (Promega). Amplification was for 40 cycles of 94°C for I rain (denaturation) and 72°C for 3 rain (annealing and extension), with a final e~:tension at 72°C for 10 rain. For second-stage PCR, primers R~: 5'-GGATCCAAAGC'I~GAA'VrCGAGCTC-3". which also corresponds to part of RoRidTt7, and LGL2: 5"-GCCGCGTTGTCGA. CTCTAGAGCGGATG'I-FGACGCCCATAG ACTCG-3". which again is based on exonic sequence from the Sau3A= genomic fragment, were used. Reaction conditions were as described above except that 2gtl of the first-stage PCR products was substituted for the first-strand eDNA. The 5' end of the eDNA was amplified directly [10] from a L. stagnalis nervous system eDNA library (Stratagene), constructed in A.gtl0. Essentially, PCR amplification was performed on DNA from •" ~ 1 0 7 pfu using primers whose sequences were based on those of the vector arms (2A: 5"-C'I-FGAGCTCAAGTTCAGCCTGGTTAAGTCCAAGCT-3' or ).B: 5'-AGAGGAAGCTTATGAGTATTTCTTCCAGGGTAAAAA-3'; see Fig. 1), in combination with LGL3: 5'-
CTGGCCAGGAATFCTGTCGACTCTATGGGCGTCAACATCCGCTCCA-3'. which was designed using exonic sequence from the Sa,~3At genomic fragment. Reaction conditions were as described for 3'-end eDNA amplification. Since truncated cDNAs are amplified from the library preferenti:llly, an additional primer. LGL4: 5'-
GACTTITI'FGAATTCGTCGACAATCATTTGACCGTGACCAGTCTrG-3', was used with either 2A or 2B to isolate further 5" sequences. The sequence of LGL4 was based on that of partial cDNAs amplified from the library as described above. Full-length cDNAs were amplified using primers LGL5: 5"- ATACTGTGGGATCCCCGTGTCAGATGGACACCTGTGT-3" and LGL6: 5"-TGTTATCACTCGAGATGAGCTGTGTCTTGGCAAGAGCT-3' (see Fig. I) the sequences of which were based on parts of the 5'- and 3'-untranslated regions that were predicted from partial eDNA clones. The reaction contained 25 pmol ofeach primer and ~20 ng first-strand eDNA. synthesized from L. stag~lalis egg mass poly(A) + RNA using R,,R,dT~7. PCR was for 40 cycles of 94°C for 1 rain (denaturation), 65°C for I rain (annealing) and 72°C for4 rain (extension), followed by a final extension at 72°C for 10rain. Specific 5'- and 3'-eDNA products were cloned into MI3 vectors, and full-length products were cloned into pBluescript KS+ (Stratagene), taking advantage of restriction endonuclease recognition sites incorporated into the PCR primers, and sequenced.
3. RESULTS A N D DISCUSSION DNA sequence analysis of the 166 bp Sau3A, fragment from a genomic clone that was isolated revealed the presence of an open reading frame that encodes a portion of a polypeptide exhibiting strong similarity (86%) to the sequence of TM3 and to part of the proposed intracellular loop of the rat GIuR1 sequence [3]. Northern blot analysis of Lymnaean poly(A) ÷ RNA (from either egg mass, dissected adult nervous systems, or adult muscle tissue) using this genomic fragment as a probe failed to detect the molluscan transcript, indicating its extremely low abundance. The exonic sequence was, therefore, used to amplify several corresponding full-length cDNAs by PCR (Fig. 1). Sequence analysis of one of these (plGIuR.PCR4) revealed an open reading frame that predicts a mature polypeptide of 898 amino acids, which has a calculated Mr of 100 913, and a 19 amino-acid signal peptide [I 1]. Hydropathic analysis.J12] of this sequence (Fig. 2) clearly reveals the presence of three strongty-hydrophobic regions (each ,~20 amino acids in length), the locations of which corre112
N o v e m b e r 1991 Deduced exonlc sequerlee
Gene 5'
~p
TMI-TM3 ZA or ~B
AAAAA 3'
TM4
mRNA
LGL3
LGL4
~ A or ~B
r~
LGLI O
RO r~
LGL2 LGL5 5' 0
LGL6 ~ 3' F u l l - l e n g t h eDNA
Fig. I. Diagram showing the PCR strategy used to isolate a full-length eDNA. The relationship of the exonic sequence of the 166 bp Sau3At genomic fragment (hatched box) to that of the m R N A is indicated. The relative locations of the signal peptide (SP) and the four putative membrane-spanning domains (TMI to TM4: [4]), that are encoded by the mRNA, are shown. The details of the cloning are fully described in section 2.
spend to those of TM I, TM3 and TM4 (as defined by Kein~inen et al. [4]) of vertebrate ionotropic glutamate receptor subunits. Several other less-pronounced hydrophobic segments are also present. In the absence of biochemical data on this invertebrate subunit, it is difficult to draw definite conclusions about the exact number and locations of membrane-spanning domains; however, for simplicity, we assume that the topology of this polypeptide is similar to that proposed for vertebrate glutamate receptor subunits [3,4]. The mature Lymnaean polypeptide displays significant identity to rat [3-6,8] glutamate receptor subunits (GluRI, 43%: GIuR2, 46%; GIuR3, 45%; GIuR4, 43%; GluR5, 37% and GIuR6, 39%); in particular, it exhibits pronounced similarity to vertebrate sequences around TM3 (Fig. 3). The molluscan subunit has 13 sites for the possible addition of N-linked sugars, 12 o f which are in the proposed N-terminal extracellular domain [4]; the 40
Hydrophobtc
. • 2O _= _u
J¢
e
-@
~
r~
-20 Hydrophlll~
~
,=, i
i
-40 100
200
300
400
500
600
700
800
Amino-acid reslduos
Fig. 2. Hydropathy plot of the mature molluscan polypcptide. The algorithm of Kyte and D~olittle [12] with a window size of 17 residues was used. Areas above the horizontal line are hydrophobic and those below the line are hydrophilic. Filled bars and open bars mark the relative locations of membrane-spanning domains, TMI to TM4, as proposed by Seeburg and colleagues [4] and by Heinemann and colleagues [3], respectively, for rat glutamate receptor subunits.
Voluroe 292, number 1,2 Lym GIuRI GIuR5 GIuR6
-i~ -18 -30 -31
November 1991 NDTCVFPLVVLWIsMRITs MPYIFAFFCTGFLGAVVG MERSTVLIQPGLWTRDTSWTLLYFLCYILP MKIISPVLSNLVFSRSIKVLLCLLWIGYSQG
+ + 1 TLDEVPIGGIFDSRSVQALT AFRH~IHMFNRAYSHVYRYKLKNDTTILDVTDSFAVSMALCMMLS~GDLAIFGVSNASS i ANFPNNIQIGGLFPNQQSQ~A AFRFALSQLTEPP KLLPQIDIVN~SDTFEMTYRFCSQFSKGVYAIFGFYE~RT 1 QTS~QVLRIGGIF~TV~N~PVNV~LAFKFAVTSINRNRTLM~NTTLTYDIQRI~LFDSF~ASRRA~DQLALGVAALFG~S~SS~ 1 TTHVLRFGGIFEYV~SGPMGAEELAFRFAVMTINRNRTLL~NTTLT¥DTQKI~L~DSFEA~KKACDQLSLGVAAIFG~S~ ** • ** • , • , • • **
Lym Glu|(l GIuR5 GIuR6
Lym GIuRI GIuR5 GIuR6
FEBS LE'I'rERS
80 76 86 84
+ + LATIQSYTDTFNVPFVTISMAQNNSH NGSYQIYMRPMYIN ALVDVIVHYRWEKVAFYYDSDEGLVRLQQLFQATMKYDKMI VMMLTSFCGALHVCFITPSFPVDTSN Q FVLQLRPELQE ALISIIDHYKWQTFVYIYDADRGLSVLQRVLDTAAEKNWQV VSAVQSIcNAL~VPHIQTRWK~PSVD~RDLFYINLYpDYAAI~R~`VLDLV~YYNWKTVTVVY~DSTGLIRLQELIKApSRYN ANAVQSICNALGV•HIQTRWKHQVSDNKDSFYVSLY•DFSS•SRAILDLVQFFKWKTVTwYDDSTGLIRLQELIKAP•B•• • • • • • • ** **
Lym GIuRI GIuR5 GIuR6
+ + ÷ 161 I S I D T K R I T S V ~ N G ~ H M L K ~ L H L M D P E M E H R V L L D V R T D K A E Q I ~ L K V M N D S K I N N A K F H F L L G D L G M L E I N T T H F K I G G V N I T G 155 T A V N I LTTTEEGYRML FQDLEKKKERLWVDCES~RLNAILGQIVKLEK NGIGYHYILANLGFMDIDLNKFK£SGRNVTG 168 I K Z K I R Q L P P A N K D A K ~ L LKEMKKSKEFYVIFDCSH~TAAEILKQILFMGMMTEY YNYFFTTLDLFALDLELYRYSGVNMTG 166 L R L K I R Q L P A D T K D A K P L LMEMKRGKEFHVIFDCSHEMAAGILKQALAMGMMT£Y YHYIFTTLDLFALDVEPYRYSGVNMTG
Lym GIuRI GIUR5 GIuR6
~46 ~35 ~50 ~48
Lym GIURI GIuR5 GIuR6
4+' + 331 G H G Q M I L E E M K K V E F E G V T G H V A F N E Q G H R K D F T L G V Y N V A M T R G T A K I G Y W N ~ R E G K LHAH N P R L F Q N N S S D M N R T R I V T T 314 G Q G I D I Q R A L Q Q V R F E G L T G N V Q F N E KGRRTNYTLHVIEM KHDGIRKIGYWNEDDKF VPAATDAQAGGDNSSVQNRTYIVTT 325 R L G P R F M N L I K ~ A R W D G L T G R I T F N K T D G L R K D F D L D I I S L KEEGTKKIGIWNSNSGLNMTDGNRDRSNNITDSLANRTLIVTT 323 R F G T R F M S L I K E A H W E G L T G R I T F N K T N G L R T D F D L D V I S L KEEGLEKIGTWDFASGLNMTESQKGK~ANITDSLSNRSLIVTT * * ** ** * * * * *we * * ** ****
+ FQLVD~FNSTSELFISTW~SLDPVYW~GAGTNHVNY~AALAADSVRLFK~AFG~ILQKD~NFLRRSRSGTAGKSM~TDDS~KT FQLVNYTDTIPARIMQQWRTSDSRDHTRVDWK~PKYTSALTYDGVKVMAEAFQSLRRQRIDISRRGNAGD CLANPAV~W FRLLNIDNPHVSSIIEKWSMERLQAPPRPZTGLL DGMMTTEAALMYDAVYMVAIASHRASQLTVSSLQ C HRMKPW FRILNTENTQVSSZ~EKWSMERLQAPPMPDSGLL DGFMTTDAALMYDAVHVVSVAVQQFPQMTVSSLQ C NRNKPW
Lym 413 GIuRI 396 GIuR5 409 GIuR6 407
+ IIKEPYVMVNNVIRDGKPLVGNEPVEGFCZDLTKAVAEKVGFDFVIQFVKDGSYGSVLSNG TWDGIVGELIRHEADMAIAPFTI ILEDPYVMLK KNANQF~GNDRYEGYCVELAAEIAKNVGYSYRLZIVSDGKYGARDPDTKAWNGMVGELVYGRADVAVA~LTI IL~EPYVMYR KSDKPLYGNDRFEGYCLDLLK~LSNILGFLYDVKLVPDGKYGAQNDKG EWNGMVKELIDHRADLAVAPLTI ILEEPYVLFK KSDKPLYGNDRFEGYCIDLLRELSTILGFTYEIRLVEDGKYGAQDDVNGQWNGMVRELIDMKADLAVAPLAI
Lym GIuRI GIuR5 GIuR6
497 478 490 ~89
TADRSRVIDFTKPFMSLGISIMIKRFQPAGKHFFSFMEPLSSEIWMCIVFAYIGVSVVLFLVSRFSPNEWM LSEAHHSYIAN TLVRZ~VIDFSKPFMSLGISIMI.KK~QKSK~GVFSFLD~LAY~I~CIVFAYIGVsVVLFLV~RFSPYEWHSEEFKEGRDQTTSD TYVREKVIDFSK~FMTLGISILYRKpNGTNPGVFSFLN~LSPDIWMYVLLACLGV~CVLFVIARFT~YEWYNPHPcNPDSDwEN T~VREK~IDF~KPFMTLGI~ILYRKPNGTNPGVFSFLNPLS~DI~4X-qL~c~L-G~ic9-L~VIARFSPY~WYN~H~NPD~DVVEN
Lym GIuRI GIuR5 GIuR6
579 DFSISNSLWFSLGAFMQQGCDI~PRSM~GRIVGS~WWFFTLIIISSYTANLAAFLTVE~ML~IDSAEDLARQTEIQYGTIM 563 Q ~ N ~ F G I F N S L W F S L G A F M Q Q G C D I S P R s L S G R I V G G V W W F F T L I I I S S Y T A N L A A F ~ T V E R M V S P I E S A E D L A K Q T E I A Y G T L E 575 N~TLLN~FWFGVGALMQQGsELM~KALSTRIVGGIWWFFTLIIISSYTANLAAFLTVERMESPID~ADDLAKQTKIEYGAVR 574 NFT~LNSI~qVGALMB0~$.ELMPKALSTRIIIGG~I~F~LI~ ~Y_T~NLA~FLTVERMESPIDSADDLAKQTKIEYGAVE
o 661 S G S T M A F F K N ~ Q F Q T Y Q R M W A Y M T S A Q ~ S V F V K T H E E G I Q R V R Q S N G K Y A Y L T ~ S ~ T I D Y V S N R K P C D T L K V G S N L N S D G F G I G T Lym GIuRI 648 A G S T K E F F R R ~ K I A V F E K M W T Y M K S A E P S V F V L T T E E G M I R V R K T K G K Y A Y L L £ S T M N E Y I E Q R K P C D T M K V G G N L D S K G Y G I A T GIuR5 657 D G S T M T F F K K S K I S T Y E K M W A F M S S R Q Q S A L V K N S D E G I Q R V L T T D YALLMESTSIEYVTQF, N C N L T Q I G G L I D S K G Y G V G T YAFLNESTTIEFVTQRN CNLTQIGGLIDSKGYGVGT GIuR6 656 D G A T M T F F K K S K I S T Y D K M W A F M S S R R Q S V L V K S N E E G I Q R V L T S D
Lym GIuRI GIuR5 GIuR6
746 P V G S D L R D K L N F A V L E L R E N G D L A K W E K I W F D R G E C P Q H S S N K E G A Q S A L T L A N V A G I F Y I L I G G L V V A V L S A A F E F L Y K S K M D 733 PKGSALRNPVNLAVLKLNEQGLLDKLKNKWWYDKGECGTGGGDSKDKTSALSLSNVAGVFYILIGGLGL/hMLVALIEFCYKSRSE 739 P I G S P Y R D K I T I A I L Q L Q E E G K L H M M K E K W W R G N G CPEEDSKE ASALGVENIGGIFIVLAAGLVLSVFVAIGEFLYKSRKN 738 P M G S P Y R D K I T I A I L Q L Q E E G K L H M M K E K W W R G N G CPEEESKE ASALGVQNIGGTFIVL~AGLVLSVFVA~GEFLYKSKKN
Lym GIuRI GIuR5 GIuR6
830 818 820 819
+ SRKSR MSFGSALRTKARLSFKG HIDSEQKTTGNGTRRRPHNSVTYTYTGPTN~GG SHAFEDSNTHTEV SKRMKGFCLIPQQSINEAIRTSTLPRNSGAGASGGGGSGENGRVVSQDFPKSMQSIPSMSHSSGMPLGATGL N D V E Q C L S F N A I M E E L G I S L K N QKKLKI(KSRTKGKSS FTSILTCHQRRTQRKETVA AQLEK RSFCSAMVEELRMSLKC QRRLKHKPQ PQLL
Fig. 3. Alignment of the amino-acid sequence of the molluscan polypeplide with those of three rat glutamate receptor subunits, The sequences o|" GIuRI (AMPA-selective: [3]), GIuRS-2 (one form of GluR5 that is responsive to L-glutamate only; [6]). GIuR6 (KA-selective; [8]) and that of the Lymnaean polypeptide (Lyre; encoded by plGIuR.PCR4) were aligned with the aid of tile computer program M U LTA LIGN [I 3], Positions at which the same amino acid is found in all four sequences are denoted by asterisks. Residues are numbered from the mature N-terminus ofeach polypeptide: signal pep|ides [11] are denoted by negative numbering, Numbers refer to the left-most amino acid on e;Lch line, Putative membrane-spanning domains (TMI to TM4; [4]) are underlined, Sites in the Lymnaean sequence for the possible addition of N-linked sugars are indicated by crosses, and protein kinase C and calmodulin-dependent protein kinas¢ type I! consensus recognition sites [14], in the proposed inlracellular domain (between TM3 and TM4), are marked by open and filled circles, respectively, The sequences of several Lymnaean. PCR.generatcd. full.length cDNAs have been given the EMBL accession number X60086.
113
Volume 292, number 1,2
FEBS LETTERS
other is found in the C-terminal 79 amino acids which occur after TM4, a region that has also been assigned an extracellular location in vertebrates [3,4]. In view of its predicted size, hydropathic profile and strong sequence similarity to vertebrate ionotropic glutamate receptor subunits, we conclude that this polypeptide is a subunit of a molluscan glutamate-gated ion-channel receptor. To test whether the Lymnaean subunit is capable of forming a functional homo-oligomeric ion channel, R N A was transcribed in vitro from a full-length cDNA (plGluR.PCR4) and injected into Xenopus oocytes. The sensitivity of the oocytes to the application of different glutamate receptor agonists was then investigated for a period of up to 15 days. No reproducible responses could be recorded in a large number of oocytes (n>100) tested (N.S. Bhandal and P.N.R. Usherwood, unpublished). Negative results were also obtained upon injection of oocytes with RNA from three other PCR-generated full-length cDNAs, which encode polypeptides that differ from each other, and from that encoded by plGluR.PCR4, by only a few amino acids. The electrophysiological results suggest either that the expression of the molluscan R N A in Xenopusoocytes is poor or, perhaps more likely, that the Lymnaean glutamate receptor subunit described here requires at least one additional subunit to form a fully-functional receptor complex. In this context, it is noteworthy that two vertebrate homo-oligomeric glutamate receptors (GIuR2 and GIuR5) show only very small responses to agonists when their corresponding cDNAs are expressed singly in Xenopus oocytes [5-7] and that coexpression studies [5,7] indicate that vertebrate glutamate receptors exist in vivo as hetero-oligomers.
114
N o v e m b e r 1991
Acknowledgements: We thank Dr Emo Vreugdenhil (Vrije Universiteit, Amsterdam) for his help in the initial stages of this work and for the L. stagnaiis nervous system cDNA library, Michael D. Squire (Cambridge) for help with DNA sequence analysis, and Dr Fergus Earley (ICI, Jealott's Hill) for his interest in this work. M.L.H. is supported by an MRC Partnership Award with ICI. and R.J.H. was supported by an SERC Research Studentship.
REFERENCES [l] Monaghan, D.T., Bridges, R.J. and Cotman, C.W. (1989) Annu. Rev. Pharmacol. Toxicol. 29, 365--402. [2] Yarowsky, P.J. and Carpenter, D.C. (1976) Science 192, 807-809. [3] Hollmann, M., O'Shea-Greenfield, A., Rogers, S.W. and Heinemann, S. (1989) Nature 342, 643-648. [4] Keintinen, K., Wisden, W., Sommer, B., Werner, P., Herb, A., Verdoorn, T.A., Sakmann, B. and Seeburg, P.H. (1990) Science 249, 556-560. [5] Boulter, J., Hollmann, M., O'Shea-Greenfield, A., Hartley, M., Deneds, E., Maron, C. and Heinemann, S. (1990) Science 249, 1033-1037. [6] Bettler, B., Boulter, J., Hermans-Borgmeyer, I., O'Shea-Greenfield, A., Deneris, E.S., Moll, C., Borgmeyer, U., Hollmann, M. and Heinemann, S. (1990) Neuron 5, 583-595. [7] Sakimura, K., Bujo, H., Kushiya, E., Araki, K., Yamazaki, M., Yamazaki, M., Meguro, H., Warashina, A., Numa, S. and Mishina, M. 0990) FEBS Lett. 272, 73-80. [8] Egebjerg, J., Bettler, B., Hermans-Borgmeyer, I. and Heinemann. S. (1991) Nature 351,745-748. [9] Frohman, M.A. and Martin, G.R. 0989) Technique l, 165-170. liO] Jansen, R., Kalousek, F., Fenton, W.A., Rosenberg, L.E. and Ledley, F.D. 0989) Genomics 4, 198-205. [11] yon Heijne, G. (1986) Nucleic Acids Res. 14, 4683-4690. [12] Kyte, J. and Doolittle, R.F. 0982) J. Mol. Biol. 157, 105-132. [13] Barton, G.J. and Sternberg, M.J.E. (1987) J. MoL Biol. 198, 327-337. [14] Kemp, B.E. and Pearson, R.B. (1990) Trends Biochem. Sci. 15, 342-346.
November 1991
Cloning of a c D N A that encodes an invertebrate glutamate receptor subunit Michael L. H u t t o n , R o b e r t J. Harvey*, Eric A. B a r n a r d and M a r k G. Darlison* M R C Molecular Neurobiology Unit, M R C Centre, Hills Road, Cambridge CB2 2Qtt, UK Received 1 A u g u s t 1991 A full-length eDNA which encodes a putative glutamate receptor polypeptide was isolated from the pond snail Lymnaea stagnalis, using a short stretch of exonic sequence and two variants of the polymerase chain reaction. In this first comparison of invertebrate and vertebrate glutamate receptor sequences, the mature molluscan polypeptide, which comprises 898 amino acids and has a predicted M~ of 100 913, displays between 37% and 46% amino-acid identity to the rat ionotropie glutamate receptor subunits, GIuRI to GIuR6.
eDNA cloning; Gene cloning; Glutamate receptor; Invertebrate receptor; Lymnaea stagnalis; Polymerase chain reaction
1. I N T R O D U C T I O N In vertebrates, the excitatory post-synaptic effects of glutamate are mediated by ligand-gated cation channels (ionotropic receptors), the receptor types of which can be distinguished by the agonists N-methyl-D-asparrate (NMDA), c~-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA), and kainate (KA); glutamate can also modulate the release of intracellular Ca 2+ through binding to a metabotropic receptor (reviewed in [1]). In contrast, in invertebrates, L-glutamate can act both as an ex.citatory and an inhibitory neurotransmitter. For example, the effect of L-glutamate, when applied to moll ascan neurones, is to induce either a fast depolarizing cationic current, a slow hyperpolarizing K* current, or a fast hyperpolarizing C1- current [2]. Recently, molecular biological studies have led to the isolation of 6 different vertebrate non-NMDA ionotropic glutamate receptor subunit cDNAs [3-8]. When these cDNAs were expressed singly either in Xenopus oocytes [3,5-8] or in mammalian cells [4], homo-oligo-
Abbreviations: AMPA, =-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; eDNA, complementary DNA; KA, kainate; NMDA, N-methyl-D-aspartate; PCR, polymerase chain reaction. *Present address: Institut fttr Zellbiochemie und klinische Neurobiologie, Universitats-Krankenhaus Eppcndorf, Universitat Hamburg, Martinistrasse 52, 2000 Hamburg 20, Germany. Reprint address." M.L. Hutton, MRC Molecular Neurobiology Unit, MRC Centre, Hills Road, Cambridge CB2 2QH, UK. Correspondence address." M G . Darlison, lnstitut ftlr Zellbi0chemie und Neurobiologie, Universitats-Krankenhaus Eppendorf, Universit'at Hamburg, Martinistrasse 52, 2000 Hamburg 20, Germany.
Published by Elsevier Science Pt¢blishers B. V.
meric receptors were formed that were AMPA-selective (GIuRI to GluR4), KA-selective (GIuR6) or responsive to L-glutamate alone (GIuR5). No sequence has yet been reported for any invertebrate glutamate receptor. We describe here the isolation o f genomie and eDNA sequences that encode a putative glutamate receptor polypeptide from the fresh-water snail Lymnaea stagna-
lis. 2. MATERIALS A N D METHODS 2.1. Isolation of a molluscan ghttamate receptor genomic clone 5 x 10s clones of an amplified L. stagnalis genomic library, constructed in ~.EMBL3, were screened with an ,'~700 bp polymerase chain reaction (PCR)-derived eDNA fragment that encodes the first three putative membrane-spanning domains (TMI to TM3) of the rat GIuRI subunit [3]. Hybridization was in 6 x SSC (1 x SSC = 0.15 M NaCI, 0.015 M sodium citrate, pH 7), 0.5% (w/v) SDS, 5 x Denhardt's solution and 100 btg/ml yeast tRNA, at 50°C for 48 h; the final wash conditions were 1 x SSC, 0.1% (w/v) SDS at 55°C for 30 rain. The library filters were then stripped and hybridized, under identical conditions, with an ,'-,1.8 kb EcoRl eDNA fragment that encodes part of a locust putative glutamate receptor polypeptide (unpublished results). Only one clone hybridized to both probes; a 166 bp Sau3A~ fragment from this was subsequently subcloned and sequenced. 2.2. Isolation of molluscan fidl-length gh~tamate receptor cDNAs Full-length cDNAs were generated using two variants of the PCR. The 3' end of the eDNA was amplified (see Fig. 1) using nested primers [9]. For this, first-strand eDNA was synthesized from L. stagrlalis egg mass poly(A)+ RNA using oligonucleotide RoRidTi7: 5'-ATCGATGGTCGACGCATGCGGATCCAAAGCTTGAATTCGAGCTC'fl 1 l-l-l-l-I 1-1 1 ~ - 3 ' . First.stage PCR was then performed using primers Ro: 5"-ATCGATGGTCGACGCATGCGGA. TCC-Y, which eorrespond~ to part of RoR~dT~7, and LGLI: 5'TGTCGGGAGAATTCGTCGACTCAGTCTGGTGGTTCTTCACGCTCA-Y, which is based on exonic sequence from the 166 bp Sau3A~ fragment. Reactions (50 gl) contained: 50 mM KCI, I0 mM Tris-HCI (pH 9 at 25°C), i.5 mM MgCI~, 200/.tM each dNTP, 0.1% (v/v) Triton X.100, 0.01% (w/v) gelatin, 1 nmol ofT4 gene 32 protein (United States Biochemical), "-,20 ng of first-strand eDNA, 25 pmol
111
V o l u m e 292, nutnber t,2
FEBS LE'I'I'ERS
ofeach primer, and 2.5 units of T~iermusaquaticus DNA polymerase (Promega). Amplification was for 40 cycles of 94°C for I rain (denaturation) and 72°C for 3 rain (annealing and extension), with a final e~:tension at 72°C for 10 rain. For second-stage PCR, primers R~: 5'-GGATCCAAAGC'I~GAA'VrCGAGCTC-3". which also corresponds to part of RoRidTt7, and LGL2: 5"-GCCGCGTTGTCGA. CTCTAGAGCGGATG'I-FGACGCCCATAG ACTCG-3". which again is based on exonic sequence from the Sau3A= genomic fragment, were used. Reaction conditions were as described above except that 2gtl of the first-stage PCR products was substituted for the first-strand eDNA. The 5' end of the eDNA was amplified directly [10] from a L. stagnalis nervous system eDNA library (Stratagene), constructed in A.gtl0. Essentially, PCR amplification was performed on DNA from •" ~ 1 0 7 pfu using primers whose sequences were based on those of the vector arms (2A: 5"-C'I-FGAGCTCAAGTTCAGCCTGGTTAAGTCCAAGCT-3' or ).B: 5'-AGAGGAAGCTTATGAGTATTTCTTCCAGGGTAAAAA-3'; see Fig. 1), in combination with LGL3: 5'-
CTGGCCAGGAATFCTGTCGACTCTATGGGCGTCAACATCCGCTCCA-3'. which was designed using exonic sequence from the Sa,~3At genomic fragment. Reaction conditions were as described for 3'-end eDNA amplification. Since truncated cDNAs are amplified from the library preferenti:llly, an additional primer. LGL4: 5'-
GACTTITI'FGAATTCGTCGACAATCATTTGACCGTGACCAGTCTrG-3', was used with either 2A or 2B to isolate further 5" sequences. The sequence of LGL4 was based on that of partial cDNAs amplified from the library as described above. Full-length cDNAs were amplified using primers LGL5: 5"- ATACTGTGGGATCCCCGTGTCAGATGGACACCTGTGT-3" and LGL6: 5"-TGTTATCACTCGAGATGAGCTGTGTCTTGGCAAGAGCT-3' (see Fig. I) the sequences of which were based on parts of the 5'- and 3'-untranslated regions that were predicted from partial eDNA clones. The reaction contained 25 pmol ofeach primer and ~20 ng first-strand eDNA. synthesized from L. stag~lalis egg mass poly(A) + RNA using R,,R,dT~7. PCR was for 40 cycles of 94°C for 1 rain (denaturation), 65°C for I rain (annealing) and 72°C for4 rain (extension), followed by a final extension at 72°C for 10rain. Specific 5'- and 3'-eDNA products were cloned into MI3 vectors, and full-length products were cloned into pBluescript KS+ (Stratagene), taking advantage of restriction endonuclease recognition sites incorporated into the PCR primers, and sequenced.
3. RESULTS A N D DISCUSSION DNA sequence analysis of the 166 bp Sau3A, fragment from a genomic clone that was isolated revealed the presence of an open reading frame that encodes a portion of a polypeptide exhibiting strong similarity (86%) to the sequence of TM3 and to part of the proposed intracellular loop of the rat GIuR1 sequence [3]. Northern blot analysis of Lymnaean poly(A) ÷ RNA (from either egg mass, dissected adult nervous systems, or adult muscle tissue) using this genomic fragment as a probe failed to detect the molluscan transcript, indicating its extremely low abundance. The exonic sequence was, therefore, used to amplify several corresponding full-length cDNAs by PCR (Fig. 1). Sequence analysis of one of these (plGIuR.PCR4) revealed an open reading frame that predicts a mature polypeptide of 898 amino acids, which has a calculated Mr of 100 913, and a 19 amino-acid signal peptide [I 1]. Hydropathic analysis.J12] of this sequence (Fig. 2) clearly reveals the presence of three strongty-hydrophobic regions (each ,~20 amino acids in length), the locations of which corre112
N o v e m b e r 1991 Deduced exonlc sequerlee
Gene 5'
~p
TMI-TM3 ZA or ~B
AAAAA 3'
TM4
mRNA
LGL3
LGL4
~ A or ~B
r~
LGLI O
RO r~
LGL2 LGL5 5' 0
LGL6 ~ 3' F u l l - l e n g t h eDNA
Fig. I. Diagram showing the PCR strategy used to isolate a full-length eDNA. The relationship of the exonic sequence of the 166 bp Sau3At genomic fragment (hatched box) to that of the m R N A is indicated. The relative locations of the signal peptide (SP) and the four putative membrane-spanning domains (TMI to TM4: [4]), that are encoded by the mRNA, are shown. The details of the cloning are fully described in section 2.
spend to those of TM I, TM3 and TM4 (as defined by Kein~inen et al. [4]) of vertebrate ionotropic glutamate receptor subunits. Several other less-pronounced hydrophobic segments are also present. In the absence of biochemical data on this invertebrate subunit, it is difficult to draw definite conclusions about the exact number and locations of membrane-spanning domains; however, for simplicity, we assume that the topology of this polypeptide is similar to that proposed for vertebrate glutamate receptor subunits [3,4]. The mature Lymnaean polypeptide displays significant identity to rat [3-6,8] glutamate receptor subunits (GluRI, 43%: GIuR2, 46%; GIuR3, 45%; GIuR4, 43%; GluR5, 37% and GIuR6, 39%); in particular, it exhibits pronounced similarity to vertebrate sequences around TM3 (Fig. 3). The molluscan subunit has 13 sites for the possible addition of N-linked sugars, 12 o f which are in the proposed N-terminal extracellular domain [4]; the 40
Hydrophobtc
. • 2O _= _u
J¢
e
-@
~
r~
-20 Hydrophlll~
~
,=, i
i
-40 100
200
300
400
500
600
700
800
Amino-acid reslduos
Fig. 2. Hydropathy plot of the mature molluscan polypcptide. The algorithm of Kyte and D~olittle [12] with a window size of 17 residues was used. Areas above the horizontal line are hydrophobic and those below the line are hydrophilic. Filled bars and open bars mark the relative locations of membrane-spanning domains, TMI to TM4, as proposed by Seeburg and colleagues [4] and by Heinemann and colleagues [3], respectively, for rat glutamate receptor subunits.
Voluroe 292, number 1,2 Lym GIuRI GIuR5 GIuR6
-i~ -18 -30 -31
November 1991 NDTCVFPLVVLWIsMRITs MPYIFAFFCTGFLGAVVG MERSTVLIQPGLWTRDTSWTLLYFLCYILP MKIISPVLSNLVFSRSIKVLLCLLWIGYSQG
+ + 1 TLDEVPIGGIFDSRSVQALT AFRH~IHMFNRAYSHVYRYKLKNDTTILDVTDSFAVSMALCMMLS~GDLAIFGVSNASS i ANFPNNIQIGGLFPNQQSQ~A AFRFALSQLTEPP KLLPQIDIVN~SDTFEMTYRFCSQFSKGVYAIFGFYE~RT 1 QTS~QVLRIGGIF~TV~N~PVNV~LAFKFAVTSINRNRTLM~NTTLTYDIQRI~LFDSF~ASRRA~DQLALGVAALFG~S~SS~ 1 TTHVLRFGGIFEYV~SGPMGAEELAFRFAVMTINRNRTLL~NTTLT¥DTQKI~L~DSFEA~KKACDQLSLGVAAIFG~S~ ** • ** • , • , • • **
Lym Glu|(l GIuR5 GIuR6
Lym GIuRI GIuR5 GIuR6
FEBS LE'I'rERS
80 76 86 84
+ + LATIQSYTDTFNVPFVTISMAQNNSH NGSYQIYMRPMYIN ALVDVIVHYRWEKVAFYYDSDEGLVRLQQLFQATMKYDKMI VMMLTSFCGALHVCFITPSFPVDTSN Q FVLQLRPELQE ALISIIDHYKWQTFVYIYDADRGLSVLQRVLDTAAEKNWQV VSAVQSIcNAL~VPHIQTRWK~PSVD~RDLFYINLYpDYAAI~R~`VLDLV~YYNWKTVTVVY~DSTGLIRLQELIKApSRYN ANAVQSICNALGV•HIQTRWKHQVSDNKDSFYVSLY•DFSS•SRAILDLVQFFKWKTVTwYDDSTGLIRLQELIKAP•B•• • • • • • • ** **
Lym GIuRI GIuR5 GIuR6
+ + ÷ 161 I S I D T K R I T S V ~ N G ~ H M L K ~ L H L M D P E M E H R V L L D V R T D K A E Q I ~ L K V M N D S K I N N A K F H F L L G D L G M L E I N T T H F K I G G V N I T G 155 T A V N I LTTTEEGYRML FQDLEKKKERLWVDCES~RLNAILGQIVKLEK NGIGYHYILANLGFMDIDLNKFK£SGRNVTG 168 I K Z K I R Q L P P A N K D A K ~ L LKEMKKSKEFYVIFDCSH~TAAEILKQILFMGMMTEY YNYFFTTLDLFALDLELYRYSGVNMTG 166 L R L K I R Q L P A D T K D A K P L LMEMKRGKEFHVIFDCSHEMAAGILKQALAMGMMT£Y YHYIFTTLDLFALDVEPYRYSGVNMTG
Lym GIuRI GIUR5 GIuR6
~46 ~35 ~50 ~48
Lym GIURI GIuR5 GIuR6
4+' + 331 G H G Q M I L E E M K K V E F E G V T G H V A F N E Q G H R K D F T L G V Y N V A M T R G T A K I G Y W N ~ R E G K LHAH N P R L F Q N N S S D M N R T R I V T T 314 G Q G I D I Q R A L Q Q V R F E G L T G N V Q F N E KGRRTNYTLHVIEM KHDGIRKIGYWNEDDKF VPAATDAQAGGDNSSVQNRTYIVTT 325 R L G P R F M N L I K ~ A R W D G L T G R I T F N K T D G L R K D F D L D I I S L KEEGTKKIGIWNSNSGLNMTDGNRDRSNNITDSLANRTLIVTT 323 R F G T R F M S L I K E A H W E G L T G R I T F N K T N G L R T D F D L D V I S L KEEGLEKIGTWDFASGLNMTESQKGK~ANITDSLSNRSLIVTT * * ** ** * * * * *we * * ** ****
+ FQLVD~FNSTSELFISTW~SLDPVYW~GAGTNHVNY~AALAADSVRLFK~AFG~ILQKD~NFLRRSRSGTAGKSM~TDDS~KT FQLVNYTDTIPARIMQQWRTSDSRDHTRVDWK~PKYTSALTYDGVKVMAEAFQSLRRQRIDISRRGNAGD CLANPAV~W FRLLNIDNPHVSSIIEKWSMERLQAPPRPZTGLL DGMMTTEAALMYDAVYMVAIASHRASQLTVSSLQ C HRMKPW FRILNTENTQVSSZ~EKWSMERLQAPPMPDSGLL DGFMTTDAALMYDAVHVVSVAVQQFPQMTVSSLQ C NRNKPW
Lym 413 GIuRI 396 GIuR5 409 GIuR6 407
+ IIKEPYVMVNNVIRDGKPLVGNEPVEGFCZDLTKAVAEKVGFDFVIQFVKDGSYGSVLSNG TWDGIVGELIRHEADMAIAPFTI ILEDPYVMLK KNANQF~GNDRYEGYCVELAAEIAKNVGYSYRLZIVSDGKYGARDPDTKAWNGMVGELVYGRADVAVA~LTI IL~EPYVMYR KSDKPLYGNDRFEGYCLDLLK~LSNILGFLYDVKLVPDGKYGAQNDKG EWNGMVKELIDHRADLAVAPLTI ILEEPYVLFK KSDKPLYGNDRFEGYCIDLLRELSTILGFTYEIRLVEDGKYGAQDDVNGQWNGMVRELIDMKADLAVAPLAI
Lym GIuRI GIuR5 GIuR6
497 478 490 ~89
TADRSRVIDFTKPFMSLGISIMIKRFQPAGKHFFSFMEPLSSEIWMCIVFAYIGVSVVLFLVSRFSPNEWM LSEAHHSYIAN TLVRZ~VIDFSKPFMSLGISIMI.KK~QKSK~GVFSFLD~LAY~I~CIVFAYIGVsVVLFLV~RFSPYEWHSEEFKEGRDQTTSD TYVREKVIDFSK~FMTLGISILYRKpNGTNPGVFSFLN~LSPDIWMYVLLACLGV~CVLFVIARFT~YEWYNPHPcNPDSDwEN T~VREK~IDF~KPFMTLGI~ILYRKPNGTNPGVFSFLNPLS~DI~4X-qL~c~L-G~ic9-L~VIARFSPY~WYN~H~NPD~DVVEN
Lym GIuRI GIuR5 GIuR6
579 DFSISNSLWFSLGAFMQQGCDI~PRSM~GRIVGS~WWFFTLIIISSYTANLAAFLTVE~ML~IDSAEDLARQTEIQYGTIM 563 Q ~ N ~ F G I F N S L W F S L G A F M Q Q G C D I S P R s L S G R I V G G V W W F F T L I I I S S Y T A N L A A F ~ T V E R M V S P I E S A E D L A K Q T E I A Y G T L E 575 N~TLLN~FWFGVGALMQQGsELM~KALSTRIVGGIWWFFTLIIISSYTANLAAFLTVERMESPID~ADDLAKQTKIEYGAVR 574 NFT~LNSI~qVGALMB0~$.ELMPKALSTRIIIGG~I~F~LI~ ~Y_T~NLA~FLTVERMESPIDSADDLAKQTKIEYGAVE
o 661 S G S T M A F F K N ~ Q F Q T Y Q R M W A Y M T S A Q ~ S V F V K T H E E G I Q R V R Q S N G K Y A Y L T ~ S ~ T I D Y V S N R K P C D T L K V G S N L N S D G F G I G T Lym GIuRI 648 A G S T K E F F R R ~ K I A V F E K M W T Y M K S A E P S V F V L T T E E G M I R V R K T K G K Y A Y L L £ S T M N E Y I E Q R K P C D T M K V G G N L D S K G Y G I A T GIuR5 657 D G S T M T F F K K S K I S T Y E K M W A F M S S R Q Q S A L V K N S D E G I Q R V L T T D YALLMESTSIEYVTQF, N C N L T Q I G G L I D S K G Y G V G T YAFLNESTTIEFVTQRN CNLTQIGGLIDSKGYGVGT GIuR6 656 D G A T M T F F K K S K I S T Y D K M W A F M S S R R Q S V L V K S N E E G I Q R V L T S D
Lym GIuRI GIuR5 GIuR6
746 P V G S D L R D K L N F A V L E L R E N G D L A K W E K I W F D R G E C P Q H S S N K E G A Q S A L T L A N V A G I F Y I L I G G L V V A V L S A A F E F L Y K S K M D 733 PKGSALRNPVNLAVLKLNEQGLLDKLKNKWWYDKGECGTGGGDSKDKTSALSLSNVAGVFYILIGGLGL/hMLVALIEFCYKSRSE 739 P I G S P Y R D K I T I A I L Q L Q E E G K L H M M K E K W W R G N G CPEEDSKE ASALGVENIGGIFIVLAAGLVLSVFVAIGEFLYKSRKN 738 P M G S P Y R D K I T I A I L Q L Q E E G K L H M M K E K W W R G N G CPEEESKE ASALGVQNIGGTFIVL~AGLVLSVFVA~GEFLYKSKKN
Lym GIuRI GIuR5 GIuR6
830 818 820 819
+ SRKSR MSFGSALRTKARLSFKG HIDSEQKTTGNGTRRRPHNSVTYTYTGPTN~GG SHAFEDSNTHTEV SKRMKGFCLIPQQSINEAIRTSTLPRNSGAGASGGGGSGENGRVVSQDFPKSMQSIPSMSHSSGMPLGATGL N D V E Q C L S F N A I M E E L G I S L K N QKKLKI(KSRTKGKSS FTSILTCHQRRTQRKETVA AQLEK RSFCSAMVEELRMSLKC QRRLKHKPQ PQLL
Fig. 3. Alignment of the amino-acid sequence of the molluscan polypeplide with those of three rat glutamate receptor subunits, The sequences o|" GIuRI (AMPA-selective: [3]), GIuRS-2 (one form of GluR5 that is responsive to L-glutamate only; [6]). GIuR6 (KA-selective; [8]) and that of the Lymnaean polypeptide (Lyre; encoded by plGIuR.PCR4) were aligned with the aid of tile computer program M U LTA LIGN [I 3], Positions at which the same amino acid is found in all four sequences are denoted by asterisks. Residues are numbered from the mature N-terminus ofeach polypeptide: signal pep|ides [11] are denoted by negative numbering, Numbers refer to the left-most amino acid on e;Lch line, Putative membrane-spanning domains (TMI to TM4; [4]) are underlined, Sites in the Lymnaean sequence for the possible addition of N-linked sugars are indicated by crosses, and protein kinase C and calmodulin-dependent protein kinas¢ type I! consensus recognition sites [14], in the proposed inlracellular domain (between TM3 and TM4), are marked by open and filled circles, respectively, The sequences of several Lymnaean. PCR.generatcd. full.length cDNAs have been given the EMBL accession number X60086.
113
Volume 292, number 1,2
FEBS LETTERS
other is found in the C-terminal 79 amino acids which occur after TM4, a region that has also been assigned an extracellular location in vertebrates [3,4]. In view of its predicted size, hydropathic profile and strong sequence similarity to vertebrate ionotropic glutamate receptor subunits, we conclude that this polypeptide is a subunit of a molluscan glutamate-gated ion-channel receptor. To test whether the Lymnaean subunit is capable of forming a functional homo-oligomeric ion channel, R N A was transcribed in vitro from a full-length cDNA (plGluR.PCR4) and injected into Xenopus oocytes. The sensitivity of the oocytes to the application of different glutamate receptor agonists was then investigated for a period of up to 15 days. No reproducible responses could be recorded in a large number of oocytes (n>100) tested (N.S. Bhandal and P.N.R. Usherwood, unpublished). Negative results were also obtained upon injection of oocytes with RNA from three other PCR-generated full-length cDNAs, which encode polypeptides that differ from each other, and from that encoded by plGluR.PCR4, by only a few amino acids. The electrophysiological results suggest either that the expression of the molluscan R N A in Xenopusoocytes is poor or, perhaps more likely, that the Lymnaean glutamate receptor subunit described here requires at least one additional subunit to form a fully-functional receptor complex. In this context, it is noteworthy that two vertebrate homo-oligomeric glutamate receptors (GIuR2 and GIuR5) show only very small responses to agonists when their corresponding cDNAs are expressed singly in Xenopus oocytes [5-7] and that coexpression studies [5,7] indicate that vertebrate glutamate receptors exist in vivo as hetero-oligomers.
114
N o v e m b e r 1991
Acknowledgements: We thank Dr Emo Vreugdenhil (Vrije Universiteit, Amsterdam) for his help in the initial stages of this work and for the L. stagnaiis nervous system cDNA library, Michael D. Squire (Cambridge) for help with DNA sequence analysis, and Dr Fergus Earley (ICI, Jealott's Hill) for his interest in this work. M.L.H. is supported by an MRC Partnership Award with ICI. and R.J.H. was supported by an SERC Research Studentship.
REFERENCES [l] Monaghan, D.T., Bridges, R.J. and Cotman, C.W. (1989) Annu. Rev. Pharmacol. Toxicol. 29, 365--402. [2] Yarowsky, P.J. and Carpenter, D.C. (1976) Science 192, 807-809. [3] Hollmann, M., O'Shea-Greenfield, A., Rogers, S.W. and Heinemann, S. (1989) Nature 342, 643-648. [4] Keintinen, K., Wisden, W., Sommer, B., Werner, P., Herb, A., Verdoorn, T.A., Sakmann, B. and Seeburg, P.H. (1990) Science 249, 556-560. [5] Boulter, J., Hollmann, M., O'Shea-Greenfield, A., Hartley, M., Deneds, E., Maron, C. and Heinemann, S. (1990) Science 249, 1033-1037. [6] Bettler, B., Boulter, J., Hermans-Borgmeyer, I., O'Shea-Greenfield, A., Deneris, E.S., Moll, C., Borgmeyer, U., Hollmann, M. and Heinemann, S. (1990) Neuron 5, 583-595. [7] Sakimura, K., Bujo, H., Kushiya, E., Araki, K., Yamazaki, M., Yamazaki, M., Meguro, H., Warashina, A., Numa, S. and Mishina, M. 0990) FEBS Lett. 272, 73-80. [8] Egebjerg, J., Bettler, B., Hermans-Borgmeyer, I. and Heinemann. S. (1991) Nature 351,745-748. [9] Frohman, M.A. and Martin, G.R. 0989) Technique l, 165-170. liO] Jansen, R., Kalousek, F., Fenton, W.A., Rosenberg, L.E. and Ledley, F.D. 0989) Genomics 4, 198-205. [11] yon Heijne, G. (1986) Nucleic Acids Res. 14, 4683-4690. [12] Kyte, J. and Doolittle, R.F. 0982) J. Mol. Biol. 157, 105-132. [13] Barton, G.J. and Sternberg, M.J.E. (1987) J. MoL Biol. 198, 327-337. [14] Kemp, B.E. and Pearson, R.B. (1990) Trends Biochem. Sci. 15, 342-346.
