Characterization of a Phosphoinositide-mediated Odor Transduction Pathway Reveals Plasma Membrane Localization of an Inositol 1,4,5-Trisphosphate Receptor in Lobster Olfactory Receptor Neurons*

The role of phosphoinositide signaling in olfactory transduction is still being resolved. Compelling functional evidence for the transduction of odor signals via phosphoinositide pathways in olfactory transduction comes from invertebrate olfactory systems, in particular lobster olfactory receptor neurons. We now provide molecular evidence for two components of the phosphoinositide signaling pathway in lobster olfactory receptor neurons, a G protein α subunit of the Gq family and an inositol 1,4,5-trisphosphate-gated channel or an inositol 1,4,5-trisphosphate (IP3) receptor. Both proteins localize to the site of olfactory transduction, the outer dendrite of the olfactory receptor neurons. Furthermore, the IP3 receptor localizes to membranes in the ciliary transduction compartment of these cells at both the light microscopic and electron microscopic levels. Given the absence of intracellular organelles in the sub-micron diameter olfactory cilia, this finding indicates that the IP3receptor is associated with the plasma membrane and provides the first definitive evidence for plasma membrane localization of an IP3R in neurons. The association of the IP3receptor with the plasma membrane may be a novel mechanism for regulating intracellular cations in restricted cellular compartments of neurons.

The role of phosphoinositide signaling in olfactory transduction is still being resolved. Compelling functional evidence for the transduction of odor signals via phosphoinositide pathways in olfactory transduction comes from invertebrate olfactory systems, in particular lobster olfactory receptor neurons. We now provide molecular evidence for two components of the phosphoinositide signaling pathway in lobster olfactory receptor neurons, a G protein ␣ subunit of the G q family and an inositol 1,4,5-trisphosphate-gated channel or an inositol 1,4,5-trisphosphate (IP 3 ) receptor. Both proteins localize to the site of olfactory transduction, the outer dendrite of the olfactory receptor neurons. Furthermore, the IP 3 receptor localizes to membranes in the ciliary transduction compartment of these cells at both the light microscopic and electron microscopic levels. Given the absence of intracellular organelles in the sub-micron diameter olfactory cilia, this finding indicates that the IP 3 receptor is associated with the plasma membrane and provides the first definitive evidence for plasma membrane localization of an IP 3 R in neurons. The association of the IP 3 receptor with the plasma membrane may be a novel mechanism for regulating intracellular cations in restricted cellular compartments of neurons.
Ca 2ϩ plays a central role in many physiological processes and regulates a plethora of ion channels, enzymes, and structural proteins. A pervasive mechanism for mobilizing Ca 2ϩ is the direct gating of a receptor ion channel (IP 3 R) 1 in the endoplasmic reticulum (ER) by the ubiquitous signaling molecule, inositol 1,4,5-trisphosphate (IP 3 ), thereby permitting release of the ion from ER stores (1,2). IP 3 -induced calcium release has been implicated in such diverse cellular processes as oogenesis (3), T-cell receptor signaling (4), and long term depression (5).
The enzyme phospholipase C liberates IP 3 , along with diacylglycerol, from the membrane phospholipid phosphoinositide 4,5-bisphosphate. The phosphoinositide (PI) signaling pathway can be regulated by both intrinsic receptors and by ligandactivated seven transmembrane-domain external receptors that in turn activate both the ␣ and ␤␥ subunits of heterotrimeric G proteins.
The role of PI signaling is still being resolved in olfactory transduction (6 -11). Compelling functional evidence for PI signaling in olfactory transduction comes from invertebrate olfactory systems, in particular spiny lobster olfactory receptor neurons (ORNs). Several lines of evidence support the hypothesis that PI signaling mediates excitation in spiny lobster ORNs. Intracellular dialysis of IP 3 mimics the odor-induced inward current in cultured lobster ORNs (12). Odors elevate IP 3 in biochemical preparations of the outer dendrites of lobster ORNs, the site of olfactory transduction in these cells (13). IP 3 activates unitary currents in cell-free inside-out patches of outer dendritic membrane (14). Antisera against PI pathwayspecific G␣ q/11 proteins block the excitatory odor response in these cells (15). Finally, PIs regulate the activity of a sodiumactivated channel that has been implicated in amplifying the transduction current in lobster ORNs (16,17). G q and phospholipase C proteins have been molecularly identified from the olfactory organ of clawed lobster (18,19). In an effort to characterize further the molecular substrate for PI signaling in lobster ORNs, we isolated cDNAs from spiny lobster olfactory organ that encode a protein conserved with IP 3 Rs, in addition to the G␣ q/11 protein. Consistent with a role in olfactory transduction in this animal, both messages are expressed in neural tissue, including the olfactory organ, and both proteins are localized to ORN dendrites. Antisera directed against a unique region of the lobster IP 3 R localize the protein to the outer dendritic membrane of lobster ORNs. Furthermore, our studies avoid a major complication that has confounded attempts to localize unequivocally IP 3 Rs to the plasma membrane in neurons where ER and other membranous compartments are closely apposed to the plasma membrane; intracellular membranous compartments, including ER, are absent in the sub-micron diameter outer dendrites of lobster ORNs (20) (Fig. 1). Therefore, our findings provide the first definitive evidence for plasma membrane association of an IP 3 R. Association of the IP 3 R with the plasma membrane may be a novel mechanism for regulating intracellular ions within restricted cellular compartments of neurons.

EXPERIMENTAL PROCEDURES
RNA Extraction-Olfactory organs (lateral antennular filament, 100 per isolation) were harvested into a dry ice/ethanol bath from freshly caught specimens of the spiny lobster, Panulirus argus, and stored at Ϫ80°C until used. The organs were homogenized by mortar and pestle and then by Polytron, and total RNA was extracted with guanidinium thiocyanate followed by centrifugation on a cesium chloride cushion.
Cloning of IP 3 R-Degenerate oligonucleotides were designed against regions of conserved amino acid and nucleotide sequence corresponding to the final putative transmembrane region (TTCTTCATIGTCAT(C/ T)ATCAT(C/T)GT, sense) and the carboxyl tail (TA(A/G)TGCCACAT-GTTGTG(C/T)TC, antisense) of vertebrate and Drosophila IP 3 Rs. Total olfactory organ RNA (10 g) was reverse-transcribed with SuperScript II reverse transcriptase (Life Technologies, Inc.) using the antisense primer. The resulting cDNA served as template for polymerase chain reaction (PCR) amplification. The PCR cycling profile was as follows: 94°C for 5 min, 60°C for 2 min, and 72°C for 3 min ϫ 1 cycle; 94°C for 1 min, 59°C (Ϫ1°C/cycle) for 1 min, and 72°C for 2 min ϫ 9 cycles; 94°C for 1 min, 52°C for 1 min, and 72°C for 2 min ϫ 30 cycles. The resulting 203-bp product (78/79-1) was fully sequenced. The cDNA was extended into the 3Ј-untranslated region by 3Ј-rapid amplification of cDNA ends (RACE (21)). A unidirectional cDNA minilibrary was constructed in the bacteriophage gt22A vector system (Life Technologies, Inc.) from olfactory organ poly(A) ϩ RNA. cDNA ligated into the phage arms was reverse-transcribed from olfactory organ poly(A) ϩ RNA with the antisense degenerate primer. The library was screened by plaque hybridization at high stringency using clone 78/79-1 as probe. Several identical clones were isolated. Three more overlapping clones were also isolated from the library by PCR. 5Ј-RACE (Life Technologies, Inc. (21)) was used to isolate the 5Ј end of the cDNA, extending the contig into the 5Ј-untranslated region. Sequencing of cDNAs was done by standard chain termination methods (22) or by the Taq DyeDeoxy Terminator and Dyeprimer Cycle sequencing protocols (Applied Biosystems) at the University of Florida's Interdisciplinary Center for Biotechnology DNA Sequencing Core Laboratory. cDNA sequences were translated, assembled, and analyzed with GeneRunner (Hastings Software, Inc.) and BioImage DNA Sequence Film Reader software (BioImage). To control for errors that may have resulted from the actions of the DNA polymerases, the coding region of the IP 3 R cDNA was re-amplified in duplicate, independent RT-PCRs; the consensus sequence is reported.
Cloning of G␣ q/11 -Degenerate oligonucleotide primers were designed against a common G␣ sequence (amino acids KWIHCFE, sense primer (AA(AG)TGGAT(ATC)CA(CT)TG (TC)TT(TC)GA)) and to the common carboxyl tail of G␣ q and G␣ 11 (amino acids KEYNLV, antisense primer ((ACTG)ACIA(AG)(AG)TT(AG)TA(TC)TC(TC)TT, where I is inosine). Total RNA was reverse-transcribed, as described, using the antisense primer. The PCR cycling profile was as follows: 94°C for 5 min, 50°C for 2 min, and 72°C for 1.5 min ϫ 1 cycle; 94°C for 30 s, 49°C (Ϫ1°C/cycle) for 30 s, and 72°C for 1.5 min ϫ 9 cycles; 94°C for 30 s, 40°C for 30 s, and 72°C for 1.5 min ϫ 30 cycles. The PCR product was diluted 1:1000 and served as template for a second, identical PCR. The resulting 435-bp product was gel-purified, ligated into the plasmid pGem-T, and transformed into JM109 Escherichia coli for subcloning. After colonies were screened by PCR for inserts of appropriate size, individual clones were selected for sequencing by standard chain termination methods. One clone (RTG-4) was fully sequenced. The 3Ј and 5Ј ends of the cDNA were isolated by 3Ј-and 5Ј-RACE. Again, two independent clones, which spanned the entire contig from the 5Ј-to the 3Ј-untranslated regions, were amplified in duplicate RT-PCRs, sequenced, and a consensus sequence translated for analysis.
Ribonuclease Protection Assay (RPA) and Northern Blotting-The RPA was done according to the RPA II kit (Ambion) protocols. RNA samples (1.5 and 3 g of olfactory organ poly(A) ϩ RNA and 5 and 10 g of brain total RNA, along with yeast RNA controls) were hybridized in solution to a 32 P-labeled antisense riboprobe transcribed from clone 78/79-1. The RNA samples were then subjected to RNase digestion (RNase A and T1) and separated on an 8 M urea, 5% acrylamide gel. For the Northern, 100 g of total RNA from olfactory organ and brain (IP 3 R) or 150 g from olfactory organ (G q ) were denatured with 15% glyoxal FIG. 1. The spiny lobster olfactory sensilla. a, photograph of one section of the olfactory organ (antennule) of the spiny lobster showing the regularly arrayed hair-like olfactory sensilla (aesthetascs) that comprise the organ. The lumen of the olfactory organ contains the somata of an estimated 320 bipolar primary olfactory receptor neurons associated with each sensillum. Each receptor cell sends an inner dendrite into the sensillum. The inner dendrite subsequently branches into an estimated 25 outer dendritic segments. Outer dendrites are restricted to the distal aesthetascs (such as those indicated by the white box). b, electron micrograph of a cross-section of one sensillum taken about one-third of the length from the tip showing that the outer dendritic segments of the primary receptor neurons are the only cellular components in the distal 2/3 of the sensillum (dark spot is a fixation artifact). C, higher magnification electron micrograph showing that each outer dendritic segment consists only of 1-3 microtubules surrounded by plasma membrane. b and c, from Ref. 20, with permission. and run on 1.1% sodium phosphate/agarose gels. Gels were blotted to MagnaPlus membrane (Micron Separations) and hybridized with the 78/79-1 riboprobe at 65°C (IP 3 R) or with a riboprobe transcribed from clone RTG-4 (G q ) at 60°C (both in 50% formamide, 5ϫ Denhardt's, 1% SDS, 5ϫ SSPE, 100 g/ml salmon sperm DNA), and washed twice for 15 min with 0.2ϫ SSC, 1% SDS at 65°C.
Generation of Antisera 23341-Rabbits were inoculated with the peptide CEDDALSKPKKPPAPK (corresponding to lobster IP 3 R amino acids 1170 -1185) conjugated to bovine serum albumin. Inoculations, enzyme-linked immunosorbent assay screenings, and affinity purifications to the antigenic peptide were performed by Coast Scientific (San Diego, CA).
Immunochemistry-For Western blots, the olfactory sensilla (aesthetascs) of 20 olfactory organs were shaved onto a metal block cooled by liquid nitrogen. The organs minus the sensilla were cut in small segments into a tube cooled by liquid nitrogen. Both samples were homogenized by mortar and pestle, transferred to liquid nitrogen-cooled tubes, and resuspended in 3 ml of protease inhibitor mixture (50 mM MOPS, pH 7.5, 200 mM NaCl, 2.5 mM MgCl 2 , 10 mM EGTA, 1 mM dithiothreitol, 0.05% sodium cholate, 0.1 mM phenylmethylsulfonyl fluoride, 4 g/ml leupeptin, 0.1 mM benzethonium, 0.1 mg/ml bacitracin, 5 g/ml pepstatin, and 0.1 mM benzamidine). Samples were sonicated, centrifuged for 15 min at 500 rpm (4°C), the supernatant removed, and the pellet resuspended in 2 ml of inhibitor mixture. The sonication, centrifugation, and resuspension procedure was repeated until the supernatant was clear. The pellets were resuspended, and the samples were centrifuged at 20,000 rpm for 40 min (4°C). Pellets were resuspended in inhibitor mixture, with aliquots reserved for protein quantification. Samples were separated on a 3-12.5% SDS-polyacrylamide gel electrophoresis gradient gel (12.3 g of protein/lane). The gel was blotted to polyvinylidene difluoride membrane (Millipore), and the membrane was blocked with 5% bovine serum albumin, 2% normal goat serum in Tris-buffered saline with Tween (2 h, 37°C), then incubated with primary antisera (1:1000) or preabsorbed antisera overnight in the cold. Labeled proteins were visualized with a horseradish peroxidaselabeled secondary antibody (Roche Molecular Biochemicals) and a chemiluminescent substrate (NEN Life Science Products). Preabsorption of antisera was done with a 10-fold excess of antigenic peptide (w/w) overnight.
Sections were cut with a diamond knife on an RMC MT-6000-XL ultramicrotome, collected on Formvar-coated nickel grids, and processed by standard immunogold protocol. After the grids were blocked with 1% dry milk in phosphate-buffered saline (PBS, 0.145 M NaCl, 0.004 M KH 2 PO 4 , 0.006 M Na 2 HPO 4 , pH 7.2) for 15 min, they were floated overnight at 4°C on primary antibody (or primary antibody pre-absorbed with the antigenic peptide (control condition)), diluted 1:10 in PBS, rinsed with high salt Tris/Tween buffer (HST, 0.5 M NaCl, 0.02 M Tris/HCl, 0.1% Tween 20, pH 7.2) (2 ϫ 10 min) followed by PBS (2ϫ for 10 min), and reacted for 1 h at 22°C with secondary antibody carrying a 12-nm gold label. After again rinsing with HST (2ϫ for 10 min) and PBS (2ϫ for 10 min), grids were floated on Trump's fixative (40) for 10 min followed by gentle washing with deionized water. Finally, sections were lightly post-stained with 0.5% aqueous uranyl acetate (1 min) and Reynolds lead citrate (30 s), washed with deionized water, and examined with a Zeiss EM-10CA transmission electron microscope.
In pilot trials the density of gold labeling was found to be optimal for tissues fixed in 1 or 2% glutaraldehyde for 1 h and using a primary antibody dilution of 1:10. Longer fixation time (i.e. 2 h) or higher antibody dilution (i.e. 1:100) resulted in considerably lower labeling densities.

RESULTS
Isolation of an IP 3 R cDNA from Olfactory Organ-Degenerate oligonucleotides reflecting conserved amino acid residues of vertebrate and Drosophila IP 3 Rs were used in a reverse tran-scription-polymerase chain reaction (RT-PCR) against lobster olfactory organ total RNA. A 203-bp cDNA fragment similar to known IP 3 Rs was amplified. By using a combination of 3Ј-RACE, 5Ј-RACE, and screening of an olfactory organ-specific IP 3 R cDNA minilibrary, the entire coding region and partial 5Јand 3Ј-untranslated regions were isolated. The contiguous cDNA contains an 8349-bp open reading frame encoding a 2783-amino acid protein of 320,000 predicted molecular weight (Fig. 2). The initiating methionine is proposed at the first methionine in the open reading frame; a consensus translation initiation sequence (23) is present at this point. The protein contains several residues predicted to be subject to post-translational modification as follows: one consensus site for extracellular N-glycosylation (Asn 2310 ), three possible sites for phosphorylation by protein kinase A (Ser 1001 , Ser 1051 , and Ser 2509 ), and one putative site for tyrosine phosphorylation (Tyr 1832 ), as well as numerous putative sites for phosphorylation by casein kinase II and protein kinase C. All sites require experimental confirmation of their relevance to the native protein. Unlike many vertebrate IP 3 Rs, there is no consensus sequence for ATP binding, consistent with the insensitivity of native lobster olfactory IP 3 Rs to ATP (12), although there is a related NAD/ FAD consensus binding sequence (Gly 2020 -Gly 2026 ) (24).
The deduced amino acid sequence for the full-length clone is similar to other IP 3 Rs and shows equivalent similarity to type 1 and type 2 IP 3 Rs but is less similar to type 3 IP 3 Rs ( Fig. 3; Drosophila, 59% identity; rat (type 1), 57%; rat (type 2), 55%; and rat (type 3), 40%). Like other IP 3 Rs, it is clearly only distantly related to ryanodine receptors, with the greatest similarity seen in the channel domain. The lobster IP 3 R shows no similarity to the Drosophila plasma membrane-localized calcium-selective channel TRP (25) or related proteins. The entire lobster IP 3 R sequence seems to maintain the distribution of conserved and variable regions of IP 3 Rs (Fig. 3), except for one portion that is completely absent in the other IP 3 Rs identified to date, amino acids Thr 1159 -Leu 1186 comprise a lysine-rich, hydrophilic stretch of the receptor (Fig. 3). Comparison of this stretch using a variety of search algorithms yielded no significant matches with any other sequences.
Isolation of a G q cDNA from Olfactory Organ-A 435-bp cDNA product was amplified with RT-PCR and degenerate oligonucleotide primers; this product was subcloned, sequenced, and conceptually translated. A comparison of the deduced amino acid sequence with sequences in the Gen-Bank TM data base revealed that this product was similar to members of the G q family of G proteins. This product was extended to the 5Ј-untranslated region by 5Ј-RACE and to the 3Ј-untranslated region by 3Ј-RACE.
The assembled full-length clone has an open reading frame of 1059 bp coding for 353 amino acids (Fig. 4). The predicted protein has a calculated molecular mass of 41.5 kDa. The deduced amino acid sequence for the full-length clone shows a high degree of identity to other known G q proteins (Fig. 4, Drosophila, 84%; Limulus, 85%; and mouse, 82%) and is less similar to other G␣ types (e.g. Drosophila G␣ o , 51% (Gen-Bank TM accession number P16378)). The spiny lobster protein is 99% identical to the clawed lobster G␣ q/11 (GenBank TM accession number P91950). The position of the initiating methionine was selected for several reasons. It is the first methionine in the open reading frame. There is a well conserved consensus initiation sequence (23). Initiating transcription at this position results in a protein of identical length to most G q proteins (Fig. 4).
Sequence analysis of the lobster G␣ q/11 shows it to contain several motifs characteristic of other G q proteins (Fig. 4 (26 -28)) as follows: putative sites for palmitoylation, two N-termi-nal cysteines (Cys 3 , Cys 4 ); an absence of N-terminal myristoylation sites (although putative myristoylation sites do exist at several sites in the middle of the protein); a putative cholera toxin ADP-ribosylation site (Arg 177 ) but no cysteine at the C-terminal subject to pertussis toxin ADP-ribosylation; and the G 40 TGES "GAG box" sequence that is present in the GTPbinding domain of other G q proteins.
Expression in Neural Tissues-A ribonuclease protection assay (RPA) using a 203-bp lobster IP 3 R probe highly conserved with other IP 3 Rs indicates that the lobster IP 3 R is expressed in olfactory organ and brain, although at much higher levels in brain (Fig. 5A). RT-PCR from olfactory organ, brain, muscle, hepatopancreas, and antennal gland indicates that a sequence containing the unique hydrophilic region (Thr 1159 -Leu 1186 ) as well as flanking sequence is expressed in olfactory organ, brain, and muscle but not in the other tissues (Fig. 5B). Northern blot analysis of brain RNA shows a single band, greater than 10 kb, when probed with the same conserved probe used for the RPA (Fig. 5C). A Northern blot of olfactory organ total RNA was also probed with a riboprobe transcribed from the G␣ q/11 clone RTG-4 (nucleotides 1049 -1483 of the complete clone) at high stringency. A single band of 4.6 kb is seen (Fig. 6).

Localization of the IP 3 R to the Plasma Membrane of Olfactory
Receptor Neurons-A polyclonal antibody (antibody 23341) was generated against a synthetic peptide (Glu 1170 -Lys 1185 ; Fig. 3) contained within the unique hydrophilic region and affinity purified against the antigenic peptide (several commercially available antibodies to IP 3 Rs were unable to recognize the lobster protein, data not shown). The antisera recognize a single protein band much greater than 220 kDa on Western blots of dendritic membrane protein obtained by scraping the olfactory sensilla from the olfactory organ (Fig. 7). This band is not detectable in a membrane protein preparation of the organ minus the sensilla and is abolished by preabsorption with the antigenic peptide. The presence of the high molecular weight band in the sensilla-only lane indicates that this protein is enriched in, if not specific to, the sensilla.
The antisera (antibody 23341) were also immunoreactive with the cut tips of the sensilla in situ, and this immunolabeling could be abolished by preabsorption of the antisera with the antigenic peptide (data not shown). These results were corroborated by immunochemistry followed by electron microscopy (Fig. 8, A and B). Grü nert and Ache (20) previously showed that only the plasma membrane and microtubules of the outer den- dritic segments are found in the tips of the olfactory sensilla. The immunogold labeling of putative IP 3 R protein observed in the present study appears to be associated with both of these ultrastructural features (Fig. 8A). Evidence for association with the plasma membrane, in particular, derives from the observation that a major portion of the gold label in most sections tends to be distributed at the perimeter of the outer dendritic segments. Polyclonal antisera C-19 (Santa Cruz Biotechnology) that recognize G␣ q/11 localize the antigen to the cut tips of dendritic cilia in situ but do not label the tissue upon FIG. 3. Schematic of the lobster IP 3 R (GenBank TM accession number AAC61691). The three domains of IP 3 Rs (ligand binding, modulatory/transducing, and channel) are indicated. Also shown are consensus sites for PKA-dependent phosphorylation (three), tyrosine kinase-dependent phosphorylation (one) and N-glycosylation (one), as well as the six proposed transmembrane regions. The unique hydrophilic sequence (Thr 1158 -Leu 1186 ) is expanded, with the antigenic peptide in bold. Percent amino acid identities, as compared with the lobster IP 3 R, are shown for the Drosophila IP 3 R (GenBank TM accession number A43360) and the rat type 1 (A36579), type 2 (S17796), and type 3 (A46719) IP 3 Rs, in blocks of 125 amino acids.
FIG. 4. Amino acid sequence of the spiny lobster G␣ q/11 . Amino acid identities, as compared with the lobster G␣ q/11 , are shown for the Drosophila G␣ q/11 (Gen-Bank TM accession number P23625), Limulus G␣ q/11 (AAB48510), and the mouse G␣ q (P21279) proteins. The spiny lobster G␣ q/11 contains several motifs indicative of this G␣ family, including N-terminal cysteines (*) that may be sites for palmitoylation, a putative cholera toxin ADP-ribosylation site (q), and a GAG box (double underline).
preabsorption with the antigenic peptide (data not shown). These results are consistent with Western blot data reported earlier (15). DISCUSSION We have shown that the spiny lobster olfactory organ expresses two genes that are critical for PI signaling, a G␣ q/11 subunit and an IP 3 R. We confirmed that the encoded proteins are in the appropriate cells and the appropriate cellular compartment within those cells to play a role in olfactory transduc-tion. ORNs in another species of lobster that express the gene for the G␣ q/11 subunit (18) also express a gene for a phospholipase C␤, a protein that associates with the G protein in response to odors (19). Lobster ORNs, therefore, appear to express the essential molecular elements of the PI signaling pathway.
The G␣ q/11 subunit we cloned has the sequence features characteristic of the G q family (26 -28). These include a putative cholera toxin ADP-ribosylation site, N-terminal cysteines that are putative substrates for palmitoylation, and a "GAG box" sequence in the GTP-binding domain. The high degree of sequence identity between the spiny lobster G␣ q/11 and homo-  2 and 4). Lanes 1 and 2 were incubated with a 1:500 dilution of the primary antisera 23341, and lanes 3 and 4 were incubated with antisera preabsorbed with the antigenic peptide. A single band greater than 220 kDa is observed in the sensilla (arrow) but not in the remaining organ. Preabsorption abolishes the immunoreactivity. Nonspecific immunoreactivity seen at the origin of the gel in both the sensilla lanes is likely due to the retention of cuticular pigment at the stacker/separator gel interface.
logues found in other species is expected in this conserved family of G protein ␣ subunits. It is also consistent with the finding that antisera against mammalian G␣ q/11 proteins block the excitatory odor response in these cells (15).
The primary structure of the cloned IP 3 R reveals several putative functional motifs that are consistent with IP 3 Rs characterized electrophysiologically in cultured lobster ORNs. The cloned IP 3 R does not contain a complete ATP-binding motif, although it does contain a related NAD/FAD-binding motif (Fig. 2). Unlike mammalian IP 3 Rs, ATP does not modulate lobster IP 3 Rs (12). ATP insensitivity may be a feature of lobster calcium-release channels, as lobster ryanodine receptors are similarly insensitive to ATP (29). The cloned IP 3 R also contains putative sites for phosphorylation by protein kinase A and by protein kinase C. The single-channel open probabilities of lobster olfactory IP 3 Rs are sensitive to phosphorylation by both of these kinases. 2 Lobster ORNs appear to express two functionally different types of IP 3 Rs based on their conductance and kinetic properties (12,14). It is unclear whether these two IP 3 Rs, differentiated on the basis of electrophysiological properties, represent the products of different genes, alternative splicing of transcripts from the same gene, different states of posttranslational modification, or are heteromultimers with different subunit stoichiometry. Screening of olfactory organ cDNA libraries, olfactory organ and brain RNA, and genomic DNA yielded only the single cDNA reported in this study, and Northern blot analysis of brain RNA showed only a single band (Fig. 5C), but the existence of one or more additional isoforms cannot be excluded. Interestingly, at least two alternative transcripts of a single IP 3 R gene have been observed in Drosophila (30).
Ultrastructural evidence presented here and in an earlier study (20) shows that the outer dendritic compartment lacks intracellular membranous organelles. As the same fixation methods are sufficient to preserve mitochondria, ER, and related structure in the inner dendrite (20), we assume that these structures would have been preserved in the outer dendrite if present. Immunogold label associated with the microtubules of the outer dendritic segments (Fig. 8A) may reflect transport of the IP 3 R protein from the Golgi complex to the plasma membrane via these cytoskeletal elements (31)(32)(33). In the absence of possible contamination from ER and other intracellular membranes, our results indicate that the lobster olfactory IP 3 R is associated with the plasma membrane in this cellular compartment. The high homology of the lobster olfactory IP 3 R with mammalian IP 3 Rs argues that it is an integral membrane protein. Our data do not allow us to conclude that the receptor is integral in the plasma membrane per se, since we cannot exclude the possibility of submembranous cisternae or caveolae, which have been shown to contain IP 3 R-like proteins in non-neuronal cells (34). It is unknown whether the lobster IP 3 R contains a signal sequence that directs the receptor to the plasma membrane. An intriguing possibility is that the unique stretch of amino acids (Thr 1159 -Leu 1186 ) might constitute such a signal, but whether these amino acids target this protein to the plasma membrane, or rather are involved in some other function perhaps specific to lobster IP 3 Rs, remains to be determined.
Although IP 3 Rs typically are not associated with the plasma membrane in neurons, recent evidence localizes type I IP 3 Rs to the plasma membrane of the outer segments of mammalian retinal cone cells (35). Earlier, IP 3 Rs were implicated in the plasma membrane of vertebrate olfactory cilia (36) and presynaptic nerve terminals (37). Recruiting an IP 3 R to the plasma membrane may be related to the anatomy of the neuronal compartment. In Drosophila photoreceptors, for example, depletion of internal IP 3 -sensitive Ca 2ϩ stores leads to the activation of TRP channels on the nearby plasma membrane (38). In lobster olfactory receptors, the extremely thin (0.1 m diameter) and long (700 -800 m) outer dendrite (20) presumably would impede diffusion of either IP 3 or Ca 2ϩ between the site of olfactory transduction and the ER at the distal end of the inner dendrite (39). Thus, IP 3 Rs associated with the plasma membrane could regulate the levels of cations in spatially constrained cellular compartments, as do intracellular IP 3 Rs in larger cellular compartments. 2 D. A. Fadool and B. W. Ache, unpublished observations. FIG. 8. The lobster IP 3 R protein is localized to membranes in ORN outer dendrites. Electron micrographs of the outer dendritic segments of lobster ORNs, cut longitudinally, illustrating immunogold labeling of putative IP 3 Rs recognized by primary antisera antibody 23341. A, representative distribution of 12-nm gold label associated with the outer dendritic segments. Although the light fixation required to retain antigenicity poorly preserves tissue structure, gold label is clearly associated with the outer dendritic segments and appears to be localized to the plasma membrane (e.g. white arrows) as well as microtubules (e.g. black arrowhead). B, outer dendritic segments showing the virtual absence of gold label when the primary antibody is preincubated with the antigenic peptide. Scale bars equal 0.2 m.