A receptor guanylyl cyclase expressed specifically in ... - Europe PMC

Proc. Natl. Acad. Sci. USA Vol. 92, pp. 3571-3575, April 1995 Neurobiology

A receptor guanylyl cyclase expressed specifically in olfactory sensory neurons (cGMP/signal transduction/smell/molecular cloning/in situ hybridization)

HANS-JURGEN FULLE*t, ROBERT VASSARt, DAVID C. FOSTERt, RUEY-BING YANGt, RICHARD AXELt, AND DAVID L. GARBERS*t§ *Howard Hughes Medical Institute and tDepartment of Pharmacology, University of Texas Southwestern Medical Center, Dallas, TX 75235-9050; and

*Department of Biochemistry and Molecular Biophysics and Howard Hughes Medical Institute, College of Physicians and Surgeons, Columbia University, New York, NY 10032 Contributed by Richard Axel, January 11, 1995

ABSTRACT We have cloned an additional member (GC-D) of the membrane receptor guanylyl cyclase [GTP pyrophosphate-lyase (cyclizing), EC 4.6.1.2] family that is specifically expressed in a subpopulation of olfactory sensory neurons. The extracellular, putative ligand-binding domain of the olfactory cyclase is similar in primary structure to two guanylyl cyclases expressed in the retina but diverges considerably from other known guanylyl cyclases. The expression of GC-D RNA is restricted to a small, randomly dispersed population of neurons that is within a single topographic zone in the olfactory neuroepithelium and resembles the pattern of the more diverse seven-transmembrane-domain odorant receptors. These observations suggest that GC-D may function directly in odor recognition or in modulating the sensitivity of a subpopulation of sensory neurons to specific odors.

MATERIALS AND METHODS Cloning and Sequencing of GC-D. Degenerate oligodeoxynucleotides were designed-5'-TGT GGA CIG CIC CIG ARC WNY T-3' and 5'-TAI GCR TCI CCI AYI GTY TCI ACY TTR TA-3' (Midland Certified Reagent, Midland, TX), where R = G or A, W = A or T, Y = T or C, and N = G, A, T, or C-based on conserved amino acid sequences in the kinase-like and cyclase catalytic domain, respectively, of membrane guanylyl cyclases (10, 13, 14). A rat olfactory cDNA library (provided by R. R. Reed, The Johns Hopkins University, Baltimore) was used as a template for PCR. Amplified products were subcloned into M13mpl9 (Boehringer Mannheim) and sequenced by the dideoxy chain-termination method (18) and Sequenase (United States Biochemical). A PCR product with a novel guanylyl cyclase sequence was 32P-labeled and used for further screening of the rat olfactory cDNA library by plaque hybridization (washes at 63°C in 0.15 M NaCl/0.015 M sodium citrate, pH 7/0.1% SDS for 1 hr each; ref. 19). Six overlapping, partial-length cDNA clones were isolated that corresponded to the PCR product. Both strands of cDNA clones G7 and Gl, covering nucleotides 1-3167 and 1362-3645 of rat GC-D cDNA, respectively, were sequenced completely by using sequence-specific oligodeoxynucleotides. The 5' 2521-nucleotide EcoRI/HindIII fragment of G7 and the 3' 1124-nucleotide HindIII/EcoRV fragment of Gl were used to generate a chimeric full-length construct (GC-D) in the mammalian expression vector pCMV5 (20). Nucleic acid and deduced amino acid sequences were analyzed with DNASTAR software (DNAstar, Madison, WI). Transient Expression and Assay of Guanylyl Cyclase Activity. Transient transfections of COS cells with GC-D were performed by the DEAE-dextran method (21), and guanylyl cyclase activity (given as mean ± SEM of three independent experiments) was measured as described (22). In Situ Hybridization. In situ hybridization experiments were performed as described (23). Specific RNA probes were generated from a construct in pBluescript II KS (Stratagene) that contained the 1638-nucleotide 5' EcoRI/BamHI cDNA fragment of clone G7.

The vertebrate olfactory system can recognize and discriminate a large number of odorants of diverse molecular structure. The initial event in odor discrimination involves the association of odorous ligands with GTP-binding protein (G protein)coupled, seven-transmembrane-domain receptors in olfactory sensory neurons (1-3). Exposure of olfactory neurons to several odorants then leads to the rapid stimulation of adenylyl cyclase and increases in cAMP (4-7), which, in turn, directly activate a cyclic nucleotide-gated cation channel responsive to both cAMP and cGMP, leading to changes in membrane potential (8, 9). However, the role of cGMP in olfactory signal transduction remains obscure. Guanylyl cyclase [GTP pyrophosphate-Jyase (cyclizing), EC 4.6.1.2] membrane receptors appear to compose a relatively small family of proteins with a similar topographic domain structure; a single membrane-spanning domain separates an extracellular ligand-binding domain from two intracellular signature domains, a protein kinase-like and cyclase catalytic domain (10-17). Guanylyl cyclases A and B (GC-A and GC-B) are members of the natriuretic peptide receptor family (1013), and guanylyl cyclase C (GC-C) encodes a heat-stable enterotoxin/guanylin peptide receptor (14, 15). Recently, we found two membrane guanylyl cyclases in the eye (GC-E and GC-F), which remain orphan receptors (17). We have now cloned and sequenced a membrane receptor guanylyl cyclase (GC-D) specifically expressed in a subpopulation of olfactory sensory neurons. Although only a single member of the GC-D family of receptor cyclases has been identified in olfactory epithelium, the similarities in the patterns of expression of this orphan receptor guanylyl cyclase and members of the more diverse seven-transmembrane-domain odorant receptor family suggest that this guanylyl cyclase also functions in the response of sensory neurons to specific odors.

RESULTS We used PCR to identify receptor guanylyl cyclases expressed in olfactory epithelium. Degenerate primers from conserved motifs within the protein kinase-like and cyclase catalytic domains of receptor guanylyl cyclases (10, 13, 14) were designed to amplify homologous sequences from a rat olfactory Abbreviations: GC-A to -F, guanylyl cyclases A-F; OMP, olfactory marker protein. §To whom reprint requests should be addressed. 1The sequence reported in this paper has been deposited in the GenBank data base (accession no. L37203).

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

3571

3572

Neurobiology: Fiille et at

Proc. Natl. Acad. Sci. USA 92 (1995)

cDNA library. The DNA sequences of the subcloned PCR products identified the natriuretic peptide receptor GC-A and GC-B (10, 13) as well as one sequence (GC-D) that was homologous yet distinct from the known guanylyl cyclases (10-16). This PCR fragment was used as a hybridization probe to isolate six overlapping cDNA clones that together gave a 3330-nucleotide reading frame that could encode a 1110amino acid protein with significant sequence and structural homology with receptor guanylyl cyclases (Fig. 1 Upper). Cleavage of a predicted N-terminal hydrophobic signal peptide (residues Met-1 to Ala-66) would result in a mature protein of 1044 amino acids (calculated molecular mass, 114,725 Da) with a proposed topography and domain structure similar to other receptor guanylyl cyclases (10-17, 24-26). A single membrane-spanning domain (residues Pro-476 to Ile500) separates a large, 409-amino acid extracellular domain MAGLQQGCHP

51

WGVLLWADSL 101 NQDASILLGS

151 AAALLAQGWG 201 AIVSSHQDIW

251 301 351 401 451 501 551 601 651 701 751 801 851 901 951 1001 1051 1101

EGQDWTAPHW SLPAWARETF OLDFKI LPTG

VTTAQQDjTA

SQRIRIGKG PGGSPIHDA

SCWFPPNTLC

HLGLQQLRIL T1KASLLLIL

RGPHRI LLTP RSVPAFLEHT NVEf1L FV R I

LESHCSFRPQ

PAPEEI LWTA

GPEQVSPLFG

SPQSV IQGST

DIKMREMRHE

GLTVRHLRTV

&DTPHADTF

VAHRNTVAAF

EjXjLIPRYC

ILFGDTVNTAS K I4E

SSISVFSVVF MATQLAVDRV

IRVNPGY*P VMRHFGARL

FRAHGLPIGL ITSLGPGEKG QTVLLRCARE DAVLTISLDT HALNHSETHG TILSGAHLGN TNGEGSQLVP THILDVSTQQ IRGVQPLGSL LTLTITCVLA QELTFLORTP SRRRPHVDSG NVALYQGEij WLIKFEAGTA Hk gSEDL GPEVSAMVLj HFPZR 3 L WGPGKATFKG RKVASPPPLC ILVSPDQGP INQGKKTSVA ISMLRMLEK GT1EPEYFD SHVYKVETII| A0DVPI RVFA jlPYPYR IKTGFCRSLP TPLSIQPGDP

ATEVCKOLHS LPY1jTLLjL KMRGGAAANL HIRALDVAJ: VQPLGTAVHF LVGGFLAYFI SESRSVVDGG PDLRPSSLSL IRNEDLRLB IT HSYAEF

OLRY~ [LDZGPRGP EVLTI)PPYC SWGLSAEEMI REEA~)DRS LDOIYTQFKS REELELDJR KTERJ4SQI4 SPIHAUM GFTTISALjE 01 E5GF LffrMJVWLj EIANE]EIJ SYAGNDMRf] DV@pTELR

DPIFAQALPS

LDMADVLLS

KSLFSWACk PEGGGALVPT MHSALLGGLE DDDGPLQEAY TIYDAVIgLA RRLPQYVI LD

VHGLKIVVLC PYRNRSYLVL

KTCRALPGPR TLGjLGPWDC

EGLTDGRLIfF

SRSHAlrAT

RVSRNMVQA

DVFSLGFJLM LECIQO4LJ SQSqG3[ QVTIYFSDVI

MMA S

PLSLDEGYKI WODHINQEIR

TGFAKLARVC

GC-D GC-E GC-F GC-A GC-B GC-C FIG. 1. Predicted amino acid sequence of GC-D and similarity to other membrane guanylyl cyclases from rat. (Upper) A protein 1110 amino acids long is predicted from a 3330-nucleotide open reading frame. Numbering is from the putative initiation methionine of the deduced amino acid sequence (single-letter code). The complete cDNA nucleotide sequence of rat GC-D includes a 94-nucleotide 5' and a 221-nucleotide 3' untranslated region. The initiation codon is preceded immediately upstream by a Kozak consensus initiation sequence (24) and further upstream by termination codons in all reading frames. Hydropathic analysis of the deduced amino acid sequence predicts an N-terminal hydrophobic signal peptide with a cleavage site after residue Ala-66 and a single transmembrane domain (underlined in bold) (25, 26). Residues invariant in rat GC-A (10), GC-B (13), GC-C (14), and GC-D are boxed. In the extracellular domain, four conserved cysteine residues are marked with triangles, and two potential N-linked glycosylation sites are underlined. (Lower) Similarity of the deduced GC-D amino acid sequence with GC-A, GC-B, GC-C, and GC-E and GC-F (17) shown by using the DNAstar program MEGALIGN (clustal algorithm with PAM250 weight table). Branch order is a function of structural similarity, and branch length reflects sequence identity.

(residues Arg-67 to Gln-477) with nine cysteine residues and two potential N-linked glycosylation sites from a 610-amino acid intracellular domain that can be subdivided into a protein kinase-like domain (residues Val-546 to Arg-808) followed by a cyclase catalytic domain closer to the carboxyl terminus (residues Met-836 to Gly-1071). GC-D is 40-45% identical at the amino acid level with the known peptide receptor guanylyl cyclases from rat (GC-A, GC-B, and GC-C) within the intracellular domain, and this identity falls to 16-21% within the extracellular domain (10, 13, 14). GC-D is more similar to GC-E and GC-F (17), being 58-61% and 40-42% identical within the intracellular and extracellular domains, respectively. Thus, GC-D, GC-E, and GC-F may define a new subfamily of membrane guanylyl cyclases (Fig. 1 Lower). The relative positions of some amino acids in the extracellular domains, in particular cysteines, are remarkably conserved. Residues Cys-121 and Cys-149 of GC-D correspond to cysteines in the atrial natriuretic peptide clearance receptor that form an intramolecular disulfide bond (27,28), and the juxtamembrane residues Cys-462 and Cys-470, conserved in receptor guanylyl cyclases except GC-C, may be involved in receptor oligomerization (10-17, 27, 28). Within the protein kinase-like domain, GC-D shows amino acid identity at 21 of the 33 residues that characterize the catalytic domain of protein kinases (29). It lacks, however, the consensus Gly-Xaa-Gly-Xaa-Xaa-Gly nucleotide-binding motif (subdomain I) (30) as well as an invariant aspartate residue (replaced with arginine, residue Arg-676) in subdomain VI of protein kinases. Therefore, GC-D more closely resembles domain JH2 of JAK protein tyrosine kinases in which asparagine replaces aspartate. This JH2 domain has not been reported to possess protein kinase activity and its function remains unknown (30). Intrinsic guanylyl cyclase activity of GC-D was examined by subcloning the GC-D cDNA into the mammalian expression vector pCMV5 (20), followed by transient expression in COS cells (21). Forty-eight hours after transfection, guanylyl cyclase activity (40.4 ± 6.8 pmol of cGMP per min per mg of protein), assayed in the presence of Triton X-100 and Mn2+ to elicit ligand-independent enzyme activity, was detectable in membrane but not cytosolic fractions (22). COS cells transfected with vector only did not show significant guanylyl cyclase activity. The incubation of GC-D-transfected cells with known guanylyl cyclase-activating compounds (rat natriuretic peptides, bacterial heat-stable enterotoxin, or sodium nitroprusside) did not result in increases in cGMP (data not shown). Thus, GC-D encodes a membrane-associated protein with guanylyl cyclase activity that does not respond to ligands known to activate other guanylyl cyclase receptors. We next performed whole-mount in situ hybridization to examine the pattern of expression of GC-D in the olfactory epithelium (23). In situ hybridization with a digoxigeninlabeled GC-D probe revealed staining of individual olfactory sensory neurons (Fig. 2 A and B). GC-D expression was restricted to a broad but circumscribed zone of cells in the center of each of the four olfactory turbinates (Fig. 2A). This spatial pattern was bilaterally symmetric; similar patterns were observed in the right and left turbinates of a single nose. Moreover, this pattern was reproducible in multiple individuals and was similar in neonatal and adult animals (data not

shown). In situ hybridization with several odorant receptor probes similarly revealed restricted expression to one of four topographically segregated zones within the epithelium (23, 31). The spatial patterns of expression of two seven-transmembrane-domain receptors, 17 and F6 (1), are compared with the GC-D profile in Fig. 2. GC-D and F6 are both expressed in neurons restricted to a more central zone (Fig. 2 A and C), whereas 17 is expressed in a distinct, more ventral zone within the epithelium (Fig. 2D). Although cells expressing GC-D or

Neurobiology: Fiille et aL

Proc. Natl Acad Sci USA 92 (1995)

3573

FIG. 2. Localization of GC-D RNA in rat olfactory epithelium. Intact olfactory turbinates of neonatal rats (postnatal day 8) were subjected to whole-mount in situ hybridization (23). The following digoxigenin-labeled antisense RNA probes were used: GC-D (A), odorant receptor F6 (C), odorant receptor 17 (D), and olfactory marker protein (OMP) (E). Hybridization with sense GC-D RNA probe is shown in F. A section through a whole-mount turbinate annealed with antisense GC-D RNA probe (A) is shown in B. Olfactory neurons expressing GC-D RNA (A) are restricted to a central zone of sensory epithelium that largely overlaps with the pattern of odorant receptor F6 expression (C), although the GC-D pattern extends further ventrally. Note that GC-D-expressing neurons are randomly distributed within the central zone in a fashion similar to cells expressing odorant receptors. Sections of whole mounts reveal labeling of individual olfactory neurons with the GC-D probe (B). Labeling with the OMP RNA probe defines the extent of the olfactory epithelium (E). No signal is apparent over the neuroepithelium after hybridization with a sense GC-D RNA probe (F). (Bar in F = 1 mm in A and C-F and 65 ,um in B).

F6 largely overlap within the central zone, the domain of GC-D-expressing neurons extends more ventrally than the domain of F6-expressing neurons. As a control, hybridization was performed with a probe for OMP RNA, which is expressed in all mature olfactory neurons (32) and delineates the boundaries of the olfactory neuroepithelium (Fig. 2E). Negative controls with a GC-D sense RNA probe reveal no hybridization signal (Fig. 2F). We also performed in situ hybridizations on serial sections of the olfactory epithelium that allow the visualization of the septal neurons and the more lateral aspects of the turbinates not detectable when using whole-mount procedures. Adja-

cent serial sections through the entire epithelium were hybridized with 35S-labeled probes for GC-D, the odorant receptor F6, or OMP (Fig. 3). Consistent with the wholemount studies, only a subpopulation of cells expressed GC-D, and these cells fall within a defined central zone of the neuroepithelium (Fig. 3 A and B). The zone of GC-Dexpressing cells largely overlapped with the domain of F6-expressing cells (Fig. 3 C and D). These data demonstrate that GC-D is expressed in a small subpopulation of neurons within a broad yet circumscribed zone of the olfactory epithelium. Within this topographic zone, however, neurons expressing GC-D appeared to be randomly distributed rather

3574

Neurobiology: Fiille et at

Proc. NatL Acad Sci. USA 92 (1995)

FIG. 3. In situ hybridization of coronal turbinate sections with GC-D- and F6-specific 35S-labeled probes. (A-D) Serial coronal sections of turbinates from a single neonatal rat (postnatal day 8) were hybridized with 35S-labeled antisense RNA probes for either GC-D (A and B) or odorant receptor F6 (C and D) RNAs. (E and F) As a control, other neighboring sections were hybridized with 35S-labeled OMP probe. After autoradiography, sections were photographed at low magnification under dark-field illumination. (A, C, and E) Set of serial sections taken from a position in the middle of the turbinate system along the anterior-posterior axis. (B, D, and F) Set of serial sections from the posterior recesses of the nasal cavity. The punctate grain clusters over the olfactory epithelium represent hybridization of 35S-labeled probes to individual sensory neurons. Note that the central zone of GC-D expression (A and B) overlaps with, but is somewhat broader than, the zone of F6 expression (C and D). (Bar in F = 1 mm for A-F).

than spatially localized. Thus, the pattern of GC-D expression closely resembles the pattern of seven-transmembranedomain receptor expression in the olfactory neuroepithelium

analysis of olfactory RNA failed to detect GC-D RNA, presumably because expression is restricted to a small subpopulation of neurons.

(23, 31). We next asked whether GC-D expression could be detected in tissues other than olfactory epithelium. GC-D RNA was not detectable by Northern blot hybridization or reverse transcriptase-PCR (RT-PCR) with RNA from various rat tissues including eye, pineal gland, heart, kidney, liver, lung, skeletal muscle, small intestine, spleen, and testis (data not shown). In contrast, GC-D RNA was readily detectable after RT-PCR amplification of olfactory epithelium RNA. Northern blot

DISCUSSION We have identified an orphan receptor, a membrane guanylyl cyclase specifically expressed in a subpopulation of olfactory sensory neurons. Comparison of the sequence of GC-D with natriuretic peptide receptors strongly suggests that GC-D is indeed a membrane receptor. The cysteines that form disulfide bonds critical for the binding of natriuretic peptides with the

Neurobiology: Fiille et al natriuretic peptide clearance receptor (27, 28) are conserved in GC-D, suggesting the conservation of a structure within the extracellular domain that facilitates ligand binding. What is the role of this receptor in olfactory sensory neurons? The expression of GC-D RNA in the olfactory epithelium is restricted to a small, randomly dispersed subset of neurons within a topographic zone, resembling the pattern of expression of the diverse seven-transmembrane-domain odorant receptors (23, 31). Recent attempts to clone additional guanylyl cyclases homologous to GC-D from olfactory cDNA have failed, suggesting that the family of olfactory-specific membrane guanylyl cyclases may be small compared with the seven-transmembrane-domain odorant receptor family (1). Therefore, GC-D may represent a member of a different class of odorant receptors of limited diversity capable of interacting with a very restricted group of odors. If GC-D encodes an odorant receptor, the olfactory-specific cation channel responsive to both cAMP and cGMP may represent the cGMPresponsive molecular target (8, 9). Alternatively, GC-D may be coordinately expressed with one or a subset of seven-transmembrane-domain receptors and may serve as a neuromodulatory receptor regulating the response to odors. Finally, GC-D may not play a role in odor recognition, but rather this receptor may regulate events essential for the differentiation or function of a subpopulation of olfactory neurons. We thank J. Craig, C. K. Green, and D. E. Miller for technical assistance in DNA sequencing and cell culture and Raquel Sitcheran for help with in situ hybridizations. We thank Dr. R. R. Reed for providing a rat olfactory cDNA library and Dr. D. Russell for providing pCMV5. We acknowledge Dr. A. Beuve, Dr. C. Craft, and Dr. J. E. Schultz for valuable suggestions and discussions. We also thank Phyllis J. Kisloff for assistance in preparing the manuscript. This work was supported by Grant I-1233 from the Robert A. Welch Foundation and by National Institutes of Health Grant HD10254 to D.L.G. H.-J.F. was a recipient of a fellowship from the Deutsche Forschungsgemeinschaft (Fu-233/1-1). R.V. is supported by a fellowship from the Helen Hay Whitney Foundation. 1. Buck, L. & Axel, R. (1991) Cell 65, 175-187. 2. Ngai, J., Dowling, M. M., Buck, L., Axel, R. & Chess, A. (1993) Cell 72, 657-666. 3. Raming, K., Krieger, J., Strotmann, J., Boekhoff, I., Kubick, S., Baumstark, C. & Breer, H. (1993) Nature (London) 361,353-356. 4. Pace, U., Hanski, E., Salomon, Y. & Lancet, D. (1985) Nature (London) 316, 255-258. 5. Sklar, P. B., Anholt, R. R. H. & Snyder, S. H. (1986) J. Bio. Chem. 261, 15538-15543.

Proc. Natl. Acad. Sci USA 92 (1995)

3575

6. Breer, H., Boekhoff, I. & Tareilus, E. (1990) Nature (London) 345, 65-68. 7. Boekhoff, I., Tareilus, E., Strotmann, J. & Breer, H. (1990) EMBO J. 9, 2453-2458. 8. Nakamura, T. & Gold, G. H. (1987) Nature (London) 325, 442-444. 9. Dhallan, R. S., Yau, K.-W., Schrader, K. A. & Reed, R. R. (1990) Nature (London) 347, 184-187. 10. Chinkers, M., Garbers, D. L., Chang, M.-S., Lowe, D. G., Chin, H., Goeddel, D. V. & Schulz, S. (1989) Nature (London) 338, 78-83. 11. Lowe, D. G., Chang, M.-S., Hellmiss, R., Chen, E., Singh, S., Garbers, D. L. & Goeddel, D. V. (1989) EMBO J. 8, 1377-1384. 12. Chang, M.-S., Lowe, D. G., Lewis, M., Hellmiss, R., Chen, E. & Goeddel, D. V. (1989) Nature (London) 341, 68-72. 13. Schulz, S., Singh, S., Bellet, R. A., Singh, G., Tubb, D. J., Chin, H. & Garbers, D. L. (1989) Cell 58, 1155-1162. 14. Schulz, S., Green, C. K., Yuen, P. S. T. & Garbers, D. L. (1990) Cell 63, 941-948. 15. de Sauvage, F. J., Camerato, T. R. & Goeddel, D. V. (1991) J. Biol. Chem. 266, 17912-17918. 16. Shyjan, A. W., de Sauvage, F. J., Gillet, N. A., Goeddel, D. V. & Lowe, D. G. (1992) Neuron 9, 727-737. 17. Yang, R.-B., Foster, D. C., Garbers, D. L. & Fulle, H.-J. (1995) Proc. Natl. Acad. Sci. USA 92, 602-606. 18. Sanger, F., Nicklen, S. & Coulsen, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. 19. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY), 2nd Ed. 20. Andersson, S., Davis, D. L., Dahlback, H., Jornvall, H. & Russell, D. W. (1989) J. Biol. Chem. 264, 8222-8229. 21. Cullen, B. R. (1987) Methods Enzymol. 152, 684-704. 22. Domino, S. E., Tubb, D. J. & Garbers, D. L. (1991) Methods Enzymol. 195, 345-355. 23. Vassar, R., Ngai, J. & Axel, R. (1993) Cell 74, 309-318. 24. Kozak, M. (1989) J. Cell Biol. 108, 229-241. 25. Kyte, J. & Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132. 26. von Heijne, G. (1983) Eur. J. Biochem. 133, 17-21. 27. Itakura, M., Iwashina, M., Mizuno, T., Ito, T., Hagiwara, H. & Hirose, S. (1994) J. Biol. Chem. 269, 8314-8318. 28. Stults, J. T., O'Connell, L., Garcia, C., Wong, S., Engel, A. M., Garbers, D. L. & Lowe, D. G. (1994) Biochemistry 33, 1137211381. 29. Hanks, S. K., Quinn, A. M. & Hunter, T. (1988) Science 241, 42-52. 30. Ziemiecki, A., Harpur, A. G. & Wilks, A. F. (1994) Trends Cell Biol. 4, 207-212. 31. Ressler, K. J., Sullivan, S. L. & Buck, L. B. (1993) Cell 73, 597-609. 32. Rogers, K. E., Dasgupta, P., Gubler, U., Grillo, M., Khew-

Goodall, Y. S. & Margolis, F. L. (1987) Proc. Natl. Acad. Sci. USA 84, 1704-1708.