Putative odour receptors localize in cilia of olfactory ... - Springer Link

52 downloads 111 Views 1MB Size Report
displayed the most intense labelling with both antibodies. Consequently, the ...... Mickie L. Weiss is thanked for copy-editing near-final drafts. The work was ...
Journal of Neurocytology 26, 691–706 (1997)

Erratum paper

Putative odour receptors localize in cilia of olfactory receptor cells in rat and mouse: a freeze-substitution ultrastructural study B E R T P H. M . M E N C O 1* , A N N E M . C U N N I N G H A M 2, P A N K A J Q A S B A 3, N I N A L E V Y 3 and R A N D A L L R . R E E D 3 1

Department of Neurobiology and Physiology, O.T. Hogan Hall, Northwestern University, Evanston, IL 60208-3520, USA Neurobiology Division, Garvan Institute of Medical Research, 384 Victoria Street, Darlinghurst, NSW, 2010 Australia 3 Molecular Biology and Genetics, Howard Hughes Medical Institute, Johns Hopkins University School of Medicine, 725 N. Wolfe Street, Baltimore, MD 21205, USA 2

Received 17 October 1996; revised 15 January 1997; accepted 17 January 1997

Summary Two different polyclonal antibodies were raised to synthetic peptides corresponding to distinct putative odour receptors of rat and mouse. Both antibodies selectively labelled olfactory cilia as seen with cryofixation and immunogold ultrastructural procedures. Regions of the olfactory organ where label was detected were consistent with those found at LM levels. Immunopositive cells were rare; only up to about 0.4% of these receptor cells were labelled. Despite chemical, species, and topographic differences both antibodies behaved identically in their ultrastructural labelling patterns. For both antibodies, labelling was very specific for olfactory cilia; both bound amply to the thick proximal and the thinner and long distal parts of the cilia. Dendritic knobs showed little labelling if any. Dendritic receptor cell structures below the knobs, supporting cell structures, and respiratory cilia did not immunolabel. There were no obvious differences in morphology between labelled and unlabelled receptor cells and their cilia. Labelling could be followed up to a distance of about 15 lm from the knobs along the distal parts of the cilia. When labelled cells were observed, this signal was detectable in two, sometimes three, sections taken through these cells while being consistently absent in neighbouring cells. This pattern argues strongly for the specificity of the labelling. In conclusion, very few receptor cells labelled with the antibodies to putative odour receptors. Additionally the olfactory cilia, the cellular regions that first encounter odour molecules and that are thought to transduce the odorous signal, displayed the most intense labelling with both antibodies. Consequently, the results showed these cilia as having many copies of the putative receptors. Finally, similar patterns of subcellular labelling were displayed in two different species, despite the use of different antibodies. Thus, this study provides compelling evidence that the heptahelical putative odour receptors localize in the olfactory cilia.

Introduction Vertebrate olfactory receptor cells are specialized neurons that have numerous long tapering cilia. There is accumulating evidence that these cilia, which line the interface between the external odorous environment and the luminal surface of the olfactory epithelium, contain the olfactory signal-transduction apparatus (reviews: Breer, 1994; Mori & Yoshihara, 1995; Breer et al., 1996; Buck, 1996; Shepherd et al, 1996; Sullivan & Dryer, 1996). Ultrastructural studies provided part of that proof (Menco, 1994; 1997). However, up to now the ultrastructural localization of the membrane-associated odour receptors remained enigmatic, * To whom correspondence should be addressed.

0300–4864/97 ( 1997 Chapman and Hall

as suitable probes to these receptors had not been available. Buck and Axel (1991) identified a large multigene family encoding a series of novel proteins in the olfactory epithelium that are predicted to share structural similarities to the light receptor rhodopsin, namely seven transmembrane domains and a G-protein association site. These authors proposed that each protein is a different odour receptor. Compelling in situ hybridization (Ngai et al., 1993; Ressler et al., 1993; Vassar et al., 1993; Strotmann et al., 1994; Byrd et al., 1996; Leibovici et al., 1996), and LM immunohistochemical studies (Koshimoto et al., 1994; Krieger

692

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

et al., 1994), and additionally also physiological (Raming et al., 1993; Kiefer et al., 1996), genetic (Sengupta et al., 1996) and theoretical (Singer et al., 1996; Turin, 1996) studies, suggest that these proteins are indeed odour receptors. Similar receptors subserve taste. One of these, a monosodium glutamate receptor, has recently been characterized (Chaudhari et al., 1996; see also Finger et al., 1996). The sequencing of the heptahelical odour receptor genes made it possible to raise receptor-specific antibodies (Harrington et al., 1997; Cunningham et al., in preparation) that allowed ultrastructural localization of the putative receptor proteins with post-embedding immunocytochemistry on cryofixed tissues. Some of the data were presented earlier in abstract form (Menco et al., 1997). Materials and methods Animal selection Adult Sprague-Dawley rats for both light and electron microscopy were obtained from Harlan (Indianapolis, IN). Several strains of mice (Jackson, Bar Harbor, MN) were used for LM. Data on one of these strains, C57BL6/J, are presented here. Swiss Webster mice (Harlan) were used for electron microscopy. All procedures described below were performed in accordance with Federal and NIH animal use guidelines, using approved animal protocols.

Light microscopy: tissue preparation Adult rats and mice were deeply anaesthetized with 85 mg kg~1 sodium-pentobarbital, injected intra-peritoneally (i.p.), and transcardially perfused with freshly prepared 4% paraformaldehyde in phosphate-buffered saline (PBS) or Bouin’s fixative (Sigma, St. Louis, MO). Rat olfactory tissue was removed and post-fixed in fresh fixative for 2 h at 4° C, and immersed in 70% ethanol. The tissue was further dehydrated through ethanol, infiltrated with Histosol (National Diagnostics, Atlanta, GA) and paraffin, and then embedded in paraffin; 6 lm sections were collected on chromalum subbed slides. Mouse olfactory tissue was treated similarly through the postfixation step and was then embedded in Tissue-Tek' OCT cryo-embedding substance (Miles, Elkart, Indiana). Cryostat sections, 12 lm thick, cut with a Microm Cryostat (Walldorf, Germany; Carl Zeiss, USA), were mounted on subbed slides.

Electron microscopy: tissue preparation Olfactory epithelia were collected from ten rats and ten mice; half of these were asphyxiated with CO and used for 2 freeze-substitution without further chemical fixation. The others were deeply anaesthetized with 85 mg kg~1 sodiumpentobarbital, injected i.p., and transcardially perfused. Following these procedures the nasal septa, covered with olfactory epithelium, were excised for further processing.

Slam-freezing of unfixed specimens and plunge-freezing of fixed cryoprotected specimens Posterior (closest to the cribriform plate) and anterior regions (closer to the respiratory area) of the nasal septa of the

groups used for freeze-substitution without further chemical fixation were rapidly frozen by dropping them on a liquid nitrogen-cooled copper block (Gentleman Jim Quick-Freeze System: Energy Beam Sciences, Agawam, MA; Phillips & Boyne, 1984). The other animals were perfusion fixed with 4% paraformaldehyde in 0.1 M Sorensen phosphate buffer (pH 7.2)]0.15 mM CaCl2. The tissues were further fixed by overnight immersion (4° C in 4% paraformaldehyde in 0.1 M bicarbonate buffer ]0.15 mM CaCl2, pH 9.4), washed [Sorensen buffer supplemented with 4% sucrose and 0.15 mM CaCl (Berod et al., 1981; Eldred et al., 1983; Finger 2 & Bo¨ ttger, 1993)], quenched for free aldehydes [1% Naborohydride (NaBH4) in washing buffer for 30 min at room temperature], and cryoprotected [5% glycerol]10% sucrose in washing buffer: 30 min; 10% glycerol]15% sucrose: 1 h; and 10% glycerol]20% sucrose: 24 h (Van Lookeren Campagne et al., 1991; Menco, 1995b)]. Posterior and anterior regions of the nasal septa were separately frozen by dropping them in liquid propane using the Gentleman Jim Quick-Freezing System. Freeze-substitution Unfixed specimens were freeze substituted in dry acetone/0.1% uranyl acetate (UAc), fixed specimens were freeze substituted in methanol/0.1% UAc. Freeze substitution, infiltration, and low-temperature embedding were carried out in a CS Auto Cryo-Substitution System (LEOReichert Instruments, Vienna, Austria). Infiltrated specimens were embedded in Lowicryl K11M while the temperature rose slowly from [60° C to ambient (Phillips & Bridgman, 1991; Menco et al., 1994; Menco, 1995a, b).

Antibodies Synthetic peptides were generated to the predicted protein sequences of the receptors D3 (rat) and M4 (mouse) based on cDNA sequences of these novel putative odour receptors (Cunningham et al., in preparation; R. R. Reed, unpublished). One antibody was made to a synthetic peptide (RDIKGAMERIFCKRKIQLNL) corresponding to the C-terminus of odour receptor D3 by coupling the peptide to an equal weight of bovine serum albumin (BSA) (final concentration of 500 lg ml~1) with 0.1% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.0). After 1 h at 20° C, the reaction was stopped with one half volume of 1 M glycine in 0.01 M Na-phosphate buffer/0.15 M NaCl (pH 7.2) and dialysed against PBS overnight at 4° C. An antibody to a peptide corresponding to the mouse odour receptor M4 (KELKNAIIKSFHRNVCQQSIA) was generated in an identical fashion. The antisera were affinity purified by linking the specific peptides to Affi-Gel 10 Support (Bio-Rad, Melville, NY) according to the manufacturer’s instructions. Crude serum was passed over a peptide affinity column, and the matrix was washed with 20 bed volumes of 0.01 M Tris (pH 7.5) followed by 20 bed volumes of 0.01 M Tris, 0.5 M NaCl (pH 7.5). Antibodies were eluted at low pH with 0.01 M glycine (pH 3.0) and rapidly neutralized with 1 M Tris (pH 8.0). Material eluted from the column was stabilized with BSA (100 lg ml~1). Antibody aliquots were stored at [80° C.

Localization of putative odour receptors Assessing antibody specificity Recombinant baculovirus expressing either D3 or one of two other novel rat odour receptors (designated D4 and L5), as well as wild type virus, were used to infect Sf9 insect cells (e.g., Raming et al., 1993). Whole cell lysates were prepared, fractionated on denaturing SDS-polyacrylamide gels, transferred to Immobilon (Millipore, Bedford, MA), and incubated with the D3 receptor antibody at a dilution of 1 : 200. Signal was visualized with Renaissance chemiluminescence (NEN Life Sciences, Boston, MA). Rainbow Marker molecular weight standards were obtained from Amersham Life Sciences Inc. (Arlington Heights, IL). In experiments with the M4 antisera, HEK-293 (Human Embryonic Kidney) cells were transfected with the M4 cDNA under the control of the CMV (Cytomegalovirus) promotor by standard methods (Dhallan et al., 1990), fixed in cold ([20° C) methanol and allowed to air dry before staining. Immunolabelling at an antibody dilution of 1 : 200 was carried out as described below for light microscopy.

Light microscopy: immunocytochemistry Paraffin sections were dewaxed and cryosections were rehydrated in PBS for 5 min. Both were preincubated in 10% normal goat serum (NGS, Vector, Burlingame, CA) in PBS at room temperature for 20 min. Next, they were incubated with the affinity purified primary antibodies at proper dilutions (D3 antiserum : 1 : 50; M4 antiserum : 1 : 250—1 : 500) in PBS, 2% NGS, for 30 min—4 h at room temperature. Immunoreactivity was visualized using the rabbit Vectastain ABC Elite Kit (Vector). Colour was developed with diaminobenzidine as the chromogenic agent.

Electron microscopy: immunocytochemistry For each experiment 2–8 serial sections, at 150–200 nm thick, were cut with a LEO-Reichert Ultramicrotome S. Immunocytochemistry was carried out as described earlier (Menco et al., 1994; Menco, 1995b). In short : Tris-buffered saline, 0.01 M (pH 8.0) supplemented with 0.5 M NaCl and 0.1% acetylated BSA (AcBSA, Aurion, Electron Microscopy Sciences, Fort Washington, PA; Leunissen, 1990), was used for blocking, incubations, dilutions, and most washings. Sections were immersed in the blocking solution for 3–4 h at room temperature and incubated with primary antibodies at appropriate dilutions (D3 antiserum: 1 : 20; M4 antiserum: 1 : 100) overnight at 4° C. Antibody binding was visualized with goat-anti-rabbit IgG-10 nm gold [GAR, OD520 (optical density)+0.1; Aurion] as the secondary probe to which the sections were exposed for about 4 h at room temperature. The buffer was supplemented with 0.1% Tween 20 for the 1 min jet-washings that followed the gold incubation. Next, the grids were jet-washed with distilled water for 1 min. Gold-conjugates were diluted with the same buffer in which they were supplied, supplemented with 0.1% AcBSA.

693 at the light microscopic level, it led to no stained cells. Antibodies to signal-transduction proteins explored in earlier studies, in particular those to the G-protein subunits, Golfa and Gsa , and to Type III adenylyl cyclase (AC) and cyclic nucleotide gated channels were used here as positive controls. All of these worked at least as well in the mouse (Menco, 1997) as in the rat (Menco et al., 1994, 1995). Anterior septal regions contained, besides olfactory epithelium, respiratory epithelium that included non-sensory kinocilia and microvilli, which served as control tissue. A most important control and extension of the experiment was that, whenever possible, it was attempted to retrieve the same labelled cell in serial sections on the same or consecutive grids.

Electron microscopy Grids containing the immunolabelled sections were stained with 0.5% methanolic UAc, formvar and/or carbon-coated on the tissue containing side, and examined at 120 kV in a JEOL 100 CX electron microscope. Formvar applied to the sections after all staining procedures reduced destruction of immunolabelled sections in the electron microscope. Considering the sparsity of labelled cells, Formvar reinforcement was important in these studies. Despite precautions, only in rare instances was it possible to retrieve all of the sections on the grids during the experimental procedures. Tissue blocks and sections were stored in an evacuated desiccator.

Identification of the same cells in serial sections As the morphology of the receptor cells was more or less the same all over the olfactory epithelium, several cues were required to identify a particular cell in serial sections. After having encountered portions of a labelled cell approximate corresponding regions in neighbouring sections were matched at low magnifications (100–200]). In case the examined area contained a special landmark, in particular infrequent microvillous cells – a heterogeneous population of sparsely occurring cells that differ distinctly from supporting cells, but which do have microvilli (Menco, 1992, for review) – this landmark was used as a topographic identifier in each section. In case no such landmark was nearby, low magnification maps, 2000–5000], were generated and common features in successive sections were identified, such as the localization of nuclei. When aligned according to the position of these features, it was determined whether the localization of the labelled receptor cells in the successive sections matched. In the rat we were more successful with the first method, whereas in the mouse we resorted more often to the second method. The resin often dissociated from its border area with the epithelial surface, exactly the area of interest, but in neighbouring sections the resin tended to remain attached in corresponding areas. This allowed observation of the same region in serial sections.

Statistical analyses Controls Absorption controls were carried at LM level only. Peptide (100 lg ml~1) was incubated with the appropriate antibody (1 : 10) overnight at 4° C. Specific staining was abolished by preincubation with the specific peptide. Because of the sparsity of labelled cells, omitting primary antibodies was not a useful control in the ultrastructural experiments, although

Densities of gold grains overlaying olfactory cilia and within surrounding areas of about 50 nm were determined using a grid divided in cm2 squares. Grain densities in surrounding areas, respiratory cilia, and Lowicryl K11M resin devoid of tissue served as comparisons. One-way Anovas were performed using the Statview 4.0 packet on a Macintosh microcomputer.

694

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

Results Antibody assessment The ability of the antibodies to D3 and M4 to recognize the corresponding putative odour receptor proteins has been assessed in Sf9 insect cell and HEK-293 cell expression systems. In immunoblotting experiments, the antiserum to D3 specifically recognized recombinant D3 protein and did not cross-react with the two other recombinant putative rat odour receptors (Fig. 1A). The molecular weight of the protein is 25–30 kDa, which is consistent with the anomalous fast mobility seen for putative odour receptors and other seven transmembrane receptors (Buck & Axel, 1991; Sung et al., 1991; Krieger et al., 1994). The antibody to M4 did not work by immunoblotting. However, this antibody

recognized the M4 putative odour receptor protein by immunocytochemistry in cells transfected with the M4 cDNA under the control of the CMV promotor (Fig. 1B). The same antiserum failed to react with untransfected cells (Fig. 1C) and with cells expressing other odour receptors under the control of the same viral promotor. Light microscopy The staining observed at the LM level for the antibodies to the D3 and M4 putative odour receptors in rat and mouse are shown in Figs 2 and 3. The antibody to the rat receptor labelled dendritic knobs and discrete patches of the surrounding ciliary layers (Fig. 2) in a zonally restricted fashion. The LM distribution of the D3 receptor in the anterior region of the rat olfactory

Fig. 1. Immunoblot and immunocytochemical analyses of recombinant D3 and M4 receptor proteins. (A) Lysates prepared from uninfected insect Sf9 cells, wild type virus (wt) infected Sf9 cells, and Sf9 cells infected with recombinant baculovirus expressing D3, D4, or L5 putative odour receptors. Protein molecular weight markers are indicated in kDa. Only the D3 blot was immunopositive. Monomer molecular weight is about 25–30 kDa. The fact that there are several bands in this region suggests some breakdown of the protein. The additional bands near 45 kDa likely represent aggregates of the D3 monomeric protein as has been reported for rhodopsin (Sung et al., 1991). Antibody dilution: 1 : 200. (B) A fraction of HEK-293 cells express the M4 putative odour receptor when transiently transfected with a CMV promotor/M4 coding region construct. Antibody dilution: 1 : 200; chromophore: diaminobenzidine. (C) Only background staining is observed in untransfected HEK-293 cells.

Localization of putative odour receptors

695

Fig. 2. Light micrograph (paraffin section) of the labelling pattern of antibodies to the C-terminal peptide of rat putative odour receptor D3 in anterior regions of rat olfactory epithelium. Knobs of two dendrites are clearly seen labelled (arrows). Antibody dilution: 1 : 50. Scale bar \100 lm. Fig. 3. Light micrograph (cryostat) of the labelling pattern of antiserum to the C-terminal peptide of mouse putative odour receptor M4 in mid-posterior regions of the mouse olfactory epithelium. Besides knob structures (arrows), the antibody also labelled dendrites and cell somata (asterisks). Antibody dilution: 1 : 250. Scale bar \ 100 lm. Table 1. A semiquantitative evaluation of the fraction of receptor cell dendritic knobs cells that immunolabelled with antibodies to putative odour receptors.

1. 2. 3. 4. 5. 6. 7. 8. 9.

Total number of grids examined Total number of positive cells seen Approximate section length Estimate: percentage sections undamaged and visible Estimate: total section length examined (no. 1] no. 3] no. 4) Estimate: number of immunopositive cells cm~1 (no. 2: no. 5) Estimate: number of immunopositive cells cm~2 (no. 62) ! Total numbers of knobs cm~2 Percentage of knobs labelled (no. 7: no. 8)

Rat

Mouse

20 15 0.1 cm 30% 0.6 cm 25 625

54 12 0.1 cm 30% 1.6 cm 7.5 56 106 –7—106

0.06–0.4%

0.006–0.04%

! From Menco (1983) and Menco and Jackson (1997).

epithelium resembles that of the highly homologous rat OR5 receptor determined by in situ hybridization (Strotmann et al., 1994, 1995; Harrington et al., 1997; Cunningham et al., in preparation). Several mouse strains were used for the LM examination of the distribution of the M4 putative odour receptor. All gave indistinguishable patterns of labelling. Moreover, molecular biological analysis of the M4 receptor expression revealed that a gene with its transcribed amino acid sequence identical to that of Mus musculus M4 putative odour receptor exists in wild mice (Mus spretus). As was true for the D3 antiserum in the rat, the antiserum directed against the mouse M4 receptor labelled a small number of cells. These were located more posteriorally and were further away from the respiratory epithelium than the antibody to D3 in the equivalent rat epithelial areas in a restricted region of the epithelium (&Z2, see Sullivan et al., 1996; R. R. Reed, unpublished). At a higher magnification, the dendritic process leading to the luminal surface and the dendritic knob can be seen clearly labelled (Fig. 3).

Electron microscopy Specimens collected from animals that were not chemically fixed before rapid freezing and freeze-substitution gave inconsistent results, with some labelling that was hard to distinguish from background. However, inclusion of paraformaldehyde fixation and controlled cryoprotection steps before cryofixation improved specificity. Cilia of only a few cells labelled distinctly, while background was low (Figs 4 through 13; especially Fig. 9 and its higher magnification Fig. 10; see also Table 1 and Table 2). Labelled cells were identified by examining numerous sections and all of the exposed epithelial surfaces in these sections, i.e., receptor cell by receptor cell whenever possible; thousands of cells were examined. In both rodent species only very few immunopositive cells were found, no more than two or three labelled cells per section, but most often only one. Usually not all of a section was retained, and some portions were obscured by grid bars, so numbers of labelled cells may actually have been somewhat higher.

696

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

Localization of putative odour receptors Most of the data were collected on septa of three individual animals of each species. In the rat a total of 15 labelled cells were encountered in 20 grids examined in four different experiments (Table 1). Five of these could be followed through two or three sections. The labelled cells were present somewhat anterior in the nasal septum in sections that often also included respiratory epithelium. More posterior regions without respiratory epithelium were examined in eight sections in three different experiments. In the mouse 12 positive cells were encountered on 54 grids examined in eight different experiments. Four immunopositive cells could be followed through two or three sections (Table 1). The positive cells were all found in mid-posterior parts of the nasal septum, but not very close to the cribriform plate. Anterior septal regions, that included respiratory epithelium, were examined in eight sections in three different experiments. Labelled areas were consistent with LM observations. Figures 4 through 8 show labelled receptor cells in the rat, while Figs 9 through 13 show such cells in the mouse. Figures 4–12 all reflect sets of serial sections. Proximal (Figs 4–12, see also inset Fig. 6) and distal parts of the cilia bound the antibodies (Figs 6 through 8, 10, and, in particular, Fig. 13). The cilia labelled amply (Figs 4–13; Table 2), suggesting that these cilia contain many copies of the individual putative odour receptors. Labelling is roughly homogeneously distributed over these cilia (see in particular Fig. 13; the quantitative evaluations of Table 2 are based on all of the cilia, proximal as well as distal parts.). All cilia seen associated with a labelled dendritic knob immunoreacted (Figs 4–12). In both, rat and mouse, immunogold labelling was detectable in olfactory cilia in two, sometimes three, serial sections taken through these rare cells, while being absent in other areas. The results differed somewhat from those seen with LM. Fewer labelled cells were seen with EM than with LM, and no labelled structures were seen apart from cilia and, to some extent, dendritic knobs (Figs 4 and 5; and, in particular, Figs 6 and 10). There was no clear distinction between the very basal necklace regions and other parts of the proximal regions of the cilia regarding labelling pattern. This suggests that some

697 receptor proteins may have been included in the necklace regions. Ultrastructurally, labelling near the apex of the knob, e.g., in Fig. 4, may have been caused by labelling at and near these bases of the cilia, or emerging cilia. Such basal areas are conceivably included in the rather thick sections. Because of the sparseness of labelled cells and virtual absence of background (Table 2), labelled insular portions of individual cilia present at a somewhat remote distance from the knobs, but within reach of the estimated length of cilia (about 50 lm; Seifert, 1970), likely pertained to one and the same labelled knob. Thus, labelling could be followed along the distal parts of the cilia. Small patches of labelled distal cilium segments, similar to those marked with arrowheads in Figs 6 and 8, were seen at a distance of 10–15 lm on either side of the knob of Figs 9 and 10. Hence, the labelled cilia of a single knob could be identified with a fair degree of certainty over 20–30% of their total length. There were no obvious differences in morphology between labelled and unlabelled olfactory receptor cells and their cilia (e.g. Figs 9 and 10). Controls

Figure 1 shows that the D3 antibody specifically recognized recombinant D3 receptor protein by immunoblotting and not two other, unrelated, putative odour receptor proteins (Fig. 1A). The M4 antibody labelled recombinant M4 protein by immunocytochemistry (Fig. 1B and C). Furthermore, the characteristic zonal pattern of staining as well as the small number of strongly staining cells of typical morphology within the epithelium supported the specificity of the antibody. For both antibodies, patterns of immunolabelling in the olfactory epithelia were consistent with those seen with in situ hybridization using receptorspecific probes. Immunoabsorption of the M4 antisera with the M4 peptide followed by immunohistochemistry, conducted only at LM levels, abolished immunoreactivity (R. R. Reed, unpublished). The number of labelled cells was too sparse to make this a valid control in the ultrastructural experiments. However, several other controls were conducted.

Fig. 4. Proximal parts of rat olfactory cilia (arrow), including the necklace regions at their bases, and to some extent, also the apex of a dendritic knob immunolabelled with antibodies to the C-terminal peptide of rat putative odour receptor D3 in paraformaldehyde fixed, cryoprotected, freeze-substituted rat olfactory epithelium. Goat-anti-rabbit IgG, conjugated to 10 nm colloidal gold, was used as a secondary probe. The section was taken from the anterior region of the olfactory epithelium. Ciliary structures of the same cell labelled in three serial sections. Cilia of no other receptor cells in these sections are immunolabelled. Background was negligible. Neighbouring supporting cell microvilli (asterisk) also did not label. The same was true for microvilli of a neighbouring microvillous cell used as a landmark to retrieve the same cells in successive sections (outside field of micrograph). Scale bar \ 1 lm. Fig. 5. Ciliary labelling (arrow) as in Fig. 4, but in a nearby section. Neighbouring supporting cell microvilli (asterisk) did not label. Background labelling was negligible. Scale bar \ 1 lm.

698

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

Localization of putative odour receptors Only olfactory cilia labelled with the receptor antibodies. Cilia and microvilli of other cells, including microvilli of neighbouring supporting cells (Fig. 4 through 10), sparse microvillous cells (not shown; Menco, 1992, for references), and cilia and microvilli of respiratory epithelia, did not label (see also Table 2). The fields where labelled cells were seen with EM were consistent with those seen with LM. Labelled patches detected in epithelial areas that were not supposed to label according to LM investigations never matched up in adjacent sections and were attributed to non-specific labelling. Positive controls needed to assess the stability of the immunogold label and patterns of labelling were regularly conducted with antibodies studied before in olfactory epithelial tissues. For these studies, antibodies to Golfa , Gsa , Type III AC, and cyclic nucleotide gated channels were used. The overall pattern of labelling with these antibodies was quite different from those to the putative odour receptors (Menco et al., 1992, 1994, 1995; Menco, 1997). However, for individual cells the labelled regions included those that immunoreacted with the antibodies to the receptor proteins. Antibodies to signalling proteins activated by the odour-receptor interaction bound particularly well to the long and thin distal parts of the cilia and labelled cilia of all or

699 most receptor cells all over the olfactory epithelial surface, in anterior as well as posterior regions. In contrast, antibodies to the putative receptors bound equally well to the short and thick proximal parts of the cilia and immunoreacted only with cilia of very few receptor cells in specified epithelial regions. Also, ultrastructurally, antibodies to most downstream signalling proteins worked equally well in fixed and unfixed tissues (exception: antibodies to cyclic nucleotide gated channels only worked well in unfixed tissues), whereas those to the receptors required some chemical fixation of the labelled tissues. Discussion Qualitative evaluations This study showed unequivocally that antibodies to putative odour receptors label olfactory cilia. For both antibodies, and in both rat and mouse, labelling was highly specific for these cellular regions. In both species numbers of labelled cells were low and, at a subcellular level, both antibodies displayed similar patterns and equally abundant intensities of labelling (Table 2). In labelled cells, both antibodies immunoreacted with all of the cilia over their entire

Table 2. Densities of colloidal gold particles marking the ultrastructural localization of putative odour receptors.

Olfactory cilia Tissue surround Respiratory cilia Resin

Rat

(%)

Mouse

(%)

82.7^55.2 (n \ 53, SE \ 7.6)! 3.1^3.8 (n \ 49, SE \ 0.5) 5.5^4.6 (n \ 4, SE \ 2.3) 0.9^2.0 (n \ 51, SE \ 0.3)

100 4 7 1

108.7^86.7 (n \ 47, SE \ 12.6) 1.1^2.6 (n \ 39, SE \ 0.4) 0.7^1.4 (n \ 6, SE \ 0.6) 0.8^2.0 (n \ 45; SE \ 0.3)

100 1 1 1

! Densities of gold grains (expressed per lm2 ) overlaying olfactory cilia and within an area of about 50 nm surrounding these cilia. Means^SDs, n \ number of areas counted, SE \ standard error. Values in the first row differ significantly (p(0.0001, significance level 5%) from those in other rows, and this is true for both rat and mouse. Second columns: values recalculated as a percentage of the most intensely labelled structure, olfactory cilia. Numbers of cilia presented in this table are higher than numbers of dendritic knobs in Table 1, because several cilia of the same knobs are included in the counts.

Fig. 6. Another example showing proximal (straight arrow, area shown at a higher magnification in the inset) and distal parts (curved arrow) of olfactory cilia immunolabelled with antibodies to rat putative odour receptor D3. Experimental conditions were the same as in Fig. 4. This section too was taken from the anterior region of the olfactory epithelium. Ciliary structures of the same cell labelled in three serial sections. Neighbouring supporting cell microvilli (open arrow) did not label. Microvilli of a nearby microvillous cell used as a landmark to retrieve the same cells in adjacent sections (outside field of micrograph) also did not label. Background labelling was negligible. Scale bar \ 1 lm. Fig. 7. The same section as Fig. 6, but a neighbouring region. The straight arrow marks the same points in Figs 6 and 7. Distal parts of cilia that most likely pertained to the same labelled receptor cell whose knob is depicted in Fig. 6, did immunoreact (arrowhead). Cilia of neighbouring cells (asterisk) did not label, and this was true for nearly all knobs in the section, apart for one or two others (see also Fig. 9). Supporting cell microvilli (open arrow) did not label. Scale bar \ 1 lm. Fig. 8. The corresponding region of Fig. 7, but in an adjacent section. The straight arrow marks the same approximate area in Figs 7 and 8, but with a different labelled cilium of the same knob. The same patch of distal parts of cilia seen labelled in Fig. 7 immunoreacted (arrowhead). Cilia of the same neighbouring cell seen unlabelled in Fig. 7 (asterisk) were also unlabelled here. Scale bar \ 1 lm.

700

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

Localization of putative odour receptors course examined. Thus, the results of this study strongly support the localization of these proteins to the sensory apparatus of the olfactory receptor neurons. A comparison with the ultrastructural localization of other signalling proteins For individual cilia the intensity of labelling with the antibodies to the putative odour receptors is at least as high as that seen with antibodies to proteins involved in the downstream signalling cascade (Menco et al., 1994; Menco, 1997). However, the labelling patterns of individual cells differed somewhat from that found with anti-bodies to proteins in these cascades. Whereas antibodies to Golfa , Gsa , Type III AC, and cyclic nucleotide gated channels immunoreacted especially well with the distal parts of the olfactory cilia (Asanuma & Nomura, 1991; Menco et al., 1992, 1994, 1995; Menco, 1997), those to the putative odour receptors immunoreacted equally well with proximal and distal parts of these cilia. Semiquantitative evaluations Table 1 presents a rough estimate of the densities of labelled cells in both species, rat and mouse. Several assumptions and corrections had to be made. (1) In some instances the sections contained both sides of the septum. (2) Often large areas of the sections were damaged during the preparation procedures. (3) Some of the labelled cells were likely missed, especially if labelled areas reflecting such cells were small. (4) Portions of other cells were not exposed to the antibodies because they were enclosed by the resin, and therefore not present at the resin’s surface. [A cross-section of a cilium taper measures about 100 nm (Menco, 1983) and the sections were 150–200 nm thick]. (5) The EM sections were about 0.15 cm long, with 0.1 cm visible, e.g., not obscured by grid bars, folded, or covered by other sections (Line 3, Table 1). Considering points 1–4 above, about 30% of the visible area in each section may have been exposed to the antibodies (Line 4, Table 1). Table 1 shows that 0.006–0.4% of all receptor cell knobs reacted with the antibodies, values within range of those obtained with LM (see first part of Results). However, it should be taken into account that the labelled cells are not homogeneously distributed (Ressler et al., 1993, 1994; Vassar et al., 1993; Strotmann et al., 1994), and in focal areas numbers of labelled cells

701 are considerably higher, 0.1–1.3% (Ressler et al., 1993; Strotmann et al., 1994). Taking further into account that the latter estimate was obtained on cells and that only about 50% of the receptor cells expresses dendritic knobs at the epithelial surface (Paternostro & Meisami, 1991), a value of 0.05–0.65% in focal areas was still given. This range in focal areas is 1.6–108 times as high as our estimate on all areas examined. The knobs have numerous cilia (average: about 17, irrespective of nasal area; maximum: about 30) and these may spread as far as 50 lm (Seifert, 1970; Menco, 1983). Linearly, the cilia of a single cell cover about 0.1 mm (50 lm in each direction). Thus, despite a sparse distribution for each individual putative odour receptor, the special shape of the cilia of the receptor cells implies that the long cilia cover a considerable part of the epithelial surface. Ten cells can extend their cilia over 1 mm. This, combined with the fact that the cells have many copies of each receptor in their cilia (Figs 4–13; Table 2), may mean that there is no need for a model that requires that each cell expresses more than one or a few different receptor genes. This corroborates observations and calculations based on other methods that suggest that only one gene is expressed per cell (Chess et al., 1994; Lancet, 1994; Leibovici et al., 1996). Therefore, it is unlikely that a given sensory cell in the rat has both D3 and OR5 putative receptors. Labelling intensity should be considered relative rather than absolute. However, that it was possible to discern cellular structures that labelled distinctly above background means that, at the level of the single receptor cell, labelling was abundant. This implies a presence of multiple identical receptors per labelled cell. Considering some technical aspects there are, likely, more rather than fewer identical receptor copies per labelled cell as reflected in the density of gold particles in Table 2. Each gold particle could have bound to more receptors, and/or the 10 nm gold particles could not access all receptors. Actual receptor densities could even be around ten times higher than the density values given in Table 2 (see Phillips & Bridgman, 1991; Bittermann et al., 1992; Menco, 1995b). For that reason the values of Table 2 merely serve as a guide for minimum numbers of putative odour receptors in cilia of labelled olfactory receptor cells.

Fig. 9. Only cilia of the dendritic knob of the olfactory receptor cell marked with an asterisk immunolabelled with antibodies to the C-terminal peptide of putative mouse odour receptor M4 (barely visible in this micrograph but see Fig. 10 for higher magnification). Cilia of neighbouring receptor cells did not label. Labelling was also not seen elsewhere in the area shown here; this includes receptor cell and supporting cell structures. The tissue was paraformaldehyde fixed, cryoprotected, and freeze-substituted. Goat-anti-rabbit IgG, conjugated to 10 nm colloidal gold, was used as the secondary probe. The section was taken from a more posterior region of the olfactory epithelium than those shown for rat in Figs 4–8. Scale bar \ 1 lm.

702

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

Localization of putative odour receptors Putative odour receptors and freeze-fracture particles of olfactory cilia Earlier calculations suggested that there may be as many as 106 freeze-fracture particles per ciliated receptor cell knob and its complement of cilia (Menco, 1980a; see also Takagi, 1989). Although the density of gold-labelled particles is lower than that of the freezefracture particles (Menco, 1983, 1994, 1995b) density values are quite high, certainly if one takes into account the fact that the estimates of Table 2 are too low rather than too high (see previous paragraph). Considering this and the more widespread distribution of odour receptor proteins than downstream signalling proteins in individual receptor cells, may mean that a substantial portion of the ciliary freeze-fracture particles represents odour receptor proteins. (Besides odour receptors, the freeze-fracture particles are bound to also include other transmembrane signalling proteins, such as ion-channels and adenylyl cyclase; see also Menco et al., 1992.) Earlier calculations suggested that the odour–odour receptor affinity is quite low (Menco, 1980a; Lancet, 1986), but the special morphology of the receptor cell cilia and the abundance of putative odour receptors per cell (Figs 4–13, Table 2; see also Menco 1980a, for freeze-fracture particle densities) increases chances that incoming odour molecules can interact with these receptors. The presence of putative odour receptor proteins in necklace regions near the base of olfactory cilia (Menco, 1980b, 1995b) suggests that at least some of the freeze-fracture particles inside these necklaces represent odour receptor proteins. Indeed, high-resolution freeze-fracturing indicated some particles in between the necklace strands that have a different morphology (Menco, 1994, 1995b).

703 Identification of putative odour receptor proteins with light and electron microscopy Cellular structures that were seen labelled with EM were also seen labelled with LM, and, topographically, labelling included the same epithelial regions. There were some differences, however. Whereas ultrastructurally labelling was especially restricted to the cilia, this was not true with LM, where the antibodies also labelled other regions of the receptor cells, including somata and axons (Fig. 3; Harrington et al., 1997; Cunningham et al., in preparation). These differences in labelling patterns may be ascribed to differences in preparation methods and antigen accessibility. Most gold particles did not exactly line the membranes (exception Figs 11 & 12, where membrane planes appear to be perpendicular to the plane of sectioning). This lack of congruence is, in part, inherent to the techniques used here. Possible explanations for this were given elsewhere (Menco et al., 1992). Briefly, not all of the tissue is exposed to section surfaces in postembedding experiments. Moreover, the plane of sectioning might not be optimal, and some displacement may have occurred during experimental procedures. Also, the antigens assessed here are present at the inside of the membranes. Nevertheless, the gold grains overlaid labelled structures or were positioned very close to the ciliary membranes, within a range of about 50 nm. Other antibodies to putative odour receptors With antibodies to different putative odour receptors that gave results at the level of the light microscope (Koshimoto et al., 1992, 1994; Krieger et al., 1994) one of us (BPhMM) using the same techniques was

Fig. 10. Higher magnification of the immunolabelled dendritic knob of Fig. 9. Proximal and distal parts of mouse olfactory cilia (snake-shaped arrows) immunolabelled with the antibodies to mouse putative odour receptor M4. The knob itself did not immunolabel, but labelling includes necklace regions at the base of the cilia. Cilia of neighbouring dendritic knobs did not label; this was true for proximal (large arrows) and distal parts of olfactory cilia (arrowheads). Supporting cell microvilli (asterisk) also lacked immunoreactivity. Background was negligible. Labelled cilia of the same cell could be followed in two other sections. Only one other labelled cell was found in these sections. The pattern of labelling is remarkably similar to that seen for the D3 antibody in the rat (Figs 6–8). Scale bar \ 1 lm. Fig. 11. Another example of proximal parts of mouse olfactory cilia that immunoreacted (asterisk) with antibodies to the M4 putative odour receptor. Experimental conditions were the same as those described in Fig. 9. The knob did not immunolabel, but labelling includes necklace regions at the base of the cilia. Cilia of a neighbouring knob below the labelled one did not immunoreact. The labelled knob could be followed through three sections. No other labelled cells were found in this section. Background was negligible. Scale bar \ 1 lm. Fig. 12. The same labelled knob (asterisk) as the one seen in Fig. 11, but in a neighbouring section. No other labelled cells were found in this section, and this includes the same unlabelled neighbouring ciliated knob shown in Fig. 11. Background was negligible. Scale bar \ 1 lm. Fig. 13. A 5 lm long segment of a distal region of an olfactory cilium that immunoreacted with antibodies to the M4 putative odour receptor. Experimental conditions are described in Fig. 9. Immunoreactivity was virtually absent in the rest of the section. Scale bar \ 1 lm.

704

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

not able to observe distinct labelling of cilia at an ultrastructural level. Possibly, C-terminal regions of putative odour receptors were less susceptible to differences in preservational and experimental conditions of light and electron microscopy than other regions of the receptor proteins, and antibodies to these C-termini may therefore have been particularly well suited for the ultra-structural studies. Conclusion This study showed unequivocally that putative odour receptors localize in olfactory cilia. Thus it provided strong evidence for a localization of the first protein involved in the odour-receptor interaction in these cilia. Activation of this protein initiates a downstream signalling cascade that leads to an electrical impulse. Earlier ultrastructural studies showed that, at least for the AC signalling cascade, all subsequently activated proteins also localize in these cilia (Asanuma & Nomura, 1991; Menco et al., 1992, 1994, 1995; Menco, 1997). Considering recent evidence obtained in transgenic mice that lacked cyclic nucleotide-gated channels, the role of signalling events that involve IP3 is somewhat obscure in vertebrates. These studies using transgenic mice implied that the AC pathway seems to be involved in all odour signalling (Dhallan et al., 1990; Brunet et al., 1996; Firestein,

1996; Nakamura et al., 1996). Though more diffuse, components of the IP3 pathway are also found in the olfactory cilia (Cunningham et al., 1993; Kalinoski et al., 1994; Menco et al., 1994; DellaCorte et al., 1996). Therefore, olfactory cilia are truly very specialized organelles. They contain all elements needed to transform the odour-receptor interaction into an electrical signal, resembling in several respects the modified cilia that form vertebrate retinal photoreceptor outer segments. Acknowledgments Drs Heinz Breer (University of Stuttgart-Hohenheim, Stuttgart, Germany) and Kensaku Mori (Osaka Biosciences Institute, Osaka, Japan) and collaborators are thanked for giving us a chance to assess some of their antibodies to putative odour receptors at an ultrastructural level. The exquisite technical assistance of Maya P. Yankova was greatly appreciated. The critical comments of the referees and the editor of this journal were very helpful in making this a much better paper. Mickie L. Weiss is thanked for copy-editing near-final drafts. The work was supported by NIH-NIDCD (RO1 DC02491, BPhMM), the Garnett Passe & Rodney Williams Memorial Foundation (AMC), and the Howard Hughes Medical Institute (RRR).

References AS A N U MA , N. & N O MU RA , H. (1991) Cytochemical local-

ization of adenylate cyclase activity in rat olfactory cells. Histochemical Journal 23, 83–90. BE RO D, A ., HAR TMA N, B. K . & PUJ OL , J. F. (1981) Importance of fixation in immunocytochemistry: use of formaldehyde solutions at variable pH for the localization of tyrosine hydroxylase. Journal of Histochemistry and Cytochemistry 29, 844–50. B IT T E RMA N N , A . G. , K N OL L , G. , N E¨ M E T H , A . & PL A T T NE R, H . (1992) Quantitative immuno-gold

labeling and ultrastructural preservation after cryofixation (combined with different freeze-substitution and embedding protocols) and after chemical fixation and cryosectioning. Histochemistry 97, 421–9. BRE E R, H. , (ed.) (1994) Biology of the Olfactory System. Seminars in Cell Biology 5 (1). B RE E R, H . , W A N N E R, I. & ST RO TMA N N, J. (1996) Molecular genetics of mammalian olfaction. Behavioral Genetics 26, 209–19. BRU N E T, L. J . , G OLD, G. H. & N GA I, J. (1996) General anosmia caused by a targeted disruption of the mouse cyclic nucleotide gated cation channel. Neuron 17, 681–93. BUCK , L . B. (1996) Information coding in the vertebrate olfactory system. Annual Reviews in Neurosciences 19, 517–44. BUC K , L . & A XEL , R. (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65, 157–67. BYRD, C. A ., JO NE S, J. T ., QU A TT RO, J. M. , RO GER S, M . E ., B RU N JE S, P. C . & VOGT , R . G . (1996) Ontogeny of

odorant receptor gene expression in zebrafish, Danio rerio. Journal of Neurobiology 29, 445–58. CH AU DH A RI , N . , YA N G, H . , L A MP. , C . , D E L AY , E . , CAR TF OR D, C ., T HAN , T . & ROPE R , S. (1996)

The taste of monosodium glutamate: membrane receptors in taste buds. Journal of Neuroscience 16, 3817–26. CH E SS, A ., SI MO N , I ., C E D AR , H . & AX E L , R. (1994) Allelic inactivation regulates olfactory receptor gene expression. Cell 78, 823–34. CUN N IN G H AM , A . M. , R YU G O, D. K . , SH A RP, A. H . , RE E D, R. R. , SN YDE R, S . H. & RO NN E TT , G. V . (1993) Neuronal inositol 1,4,5-trisphosphate

receptor localized to the plasma membrane of olfactory cilia. Neuroscience 57, 339–52. D EL LA COR T E , C. , RE S T R E P O, D . , ME N CO , B . P H . M. , AN DR E IN I , I. & K A L IN OSK I, D. L . (1996) Gaq /Ga11 :

immunolocalization in the neuroepithelium of the rat (Rattus rattus) and the channel catfish (Ictalurus punctatus). Neuroscience 74, 261–73. DH ALL A N, R . S. , YA N , K . -W. , SCHR ADER , K . A . & R E E D, R. R. (1990) Primary structure and functional

expression of a cyclic nucleotide-activated channel from olfactory neurons. Nature 347, 184–7. E L DRE D, W. D. , Z UC K E R, C . , K AR T E N , H . J . & YA ZU L L A , S. (1983) Comparison of fixation and

penetration enhancement techniques for use in ultrastructural immunocytochemistry. Journal of Histochemistry and Cytochemistry 2, 285–92.

Localization of putative odour receptors

705

FI NGE R, T . E. & B O® T TGE R , B. (1993) Peripheral peptider-

ME N C O, B . P H . M . (1983) The ultrastructure of olfactory

gic fibers of the trigeminal nerve in the olfactory bulb of the rat. Journal of Comparative Neurology 334, 117–24.

1,4,5-trisphosphate receptors in the olfactory neuroepithelium of the rat and channel catfish. Chemical Senses 19, 493, abstract 137.

and nasal respiratory epithelium surfaces. In Nasal Tumors in Animals and Man, Vol. 1, Anatomy, Physiology and Epidemiology (edited by R E ZN IK , G. & ST IN SO N, S. F.) pp. 45–102. Boca Raton: CRC Press Inc. ME N C O, B . P H . M. (1992) Lectins bind differentially to cilia and microvilli of major and minor cell populations in olfactory and nasal respiratory epithelia. Microscopy Research and Technique 23, 181–99. ME N C O, B . P H . M. (1994) Ultrastructural aspects of olfactory transduction and perireceptor events. Seminars in Cell Biology 5, 11–24. ME N C O, B . P H . M. (1995a) Scanning- and transmission electron microscopy of olfaction. In Experimental Cell Biology of Taste and Olfaction. Current Techniques and Protocols (edited by SPI E L M AN , A . I . & B RA N D, J. G.) pp. 115–25. Boca Raton, FL: CRC Press. ME N C O, B . P H . M. (1995b) Freeze-fracture, deep-etch, and freeze-substitution studies of olfactory epithelia, with special emphasis on immunocytochemical variables. Microscopy Research and Technique 32, 337–56. ME N C O, B . P H . M. (1997) Ultrastructural aspects of olfactory signaling. Chemical Senses 22, 295–311.

KI E FE R, H. , KRIEG E R, J . , OLS Z EWS K I, J . D. , V ON HE IJ NE , S., PRES TW IC H , G . D. & B REE R, H . (1996)

ME N C O, B . P H . M. , B R UC H , R . C ., DA U , B . & DA N HO , W. (1992) Ultrastructural localization of olfactory trans-

Expression of olfactory receptor in Escherichia coli: purification, and ligand binding. Biochemistry 35, 16077–84.

duction components: the G protein subunit Golfa and type III adenylyl cyclase. Neuron 8, 441–53.

KO SH IMOT O, H. , KA T OH , K. , YO SH IH A RA , Y. & M OR I , K. (1992) Distribution of putative odour receptor

ME N C O, B . P H . M ., C U NN IN GH A M, A. M ., Q A S B A, P. , L E VY , N . & RE E D, R. R. (1997) Putative odor receptors

FI NGE R, T. E . , BR YA N T, B . P. , KA LI N OSK I, D. L . , TE ETE R, J. H . , B O® T TGER , B . , G R OS VEN OR , M. , C AGA N , R. H . & BRA ND, J. C . (1996) Differential

localization of putative amino acid receptors in taste buds of the channel catfish, Ictalurus punctatus. Journal of Comparative Neurology 373, 129–38. FI RE ST EIN , S. (1996) Scentsational ion channels. Neuron 17, 803–6. HAR RI NGTO N, C. , BU C KL A N D, M. , LE V Y, N ., REE D, R . & C U N N IN GH A M, A. (1997) Odorant receptor pro-

teins: expression in olfactory axons and olfactory bulb glomeruli supports a role in axonal guidance and/or target recognition. Chemical Senses 22, 181–182, abstract 2. KA L I NOSK I, D. L . , DEL LA C OR T E , C . , ME NC O, B. P H . M. & RE S T R E P O , D. (1994) Localization of inositol

proteins 521–3.

in

olfactory epithelium. NeuroReport

3,

KO SH IMOT O, H . , K A TOH , K. , YOS HI HA RA , Y. , NE MOT O, Y. & MO RI, K . (1994) Immunohistochemical

demonstration of embryonic expression of an odor receptor protein and its zonal distribution in the rat olfactory epithelium. Neuroscience Letters 169, 73–6.

localize in cilia of olfactory receptor cells in rat and mouse. Chemical Senses 22, 182, abstract 3. ME N C O, B . P H . M . & JAC K S ON , J. E . (1997) A banded topography in the developing rat’s olfactory epithelial surface. Journal of Comparative Neurology, in press. ME N C O, B . P H . M ., MA TSU ZA K I, O. , B A K IN , R. E . , RO N NE TT , G . V . , S TR OT M AN N , J . & B R E E R, H .

KRIE G E R, J . , SC H LE IC H E R, S . , STRO TMA NN , J. , W ANNE R, I . , BOE K H OFF , I . , R AM I NG , K . , D E GE U S , P . & B R EE R, H . (1994) Probing olfactory receptors with

(1995) Ultrastructural localization of putative odor receptors and cyclic nucleotide-gated channels in rat olfactory epithelia. Chemical Senses 20, 742, abstract 191.

sequence-specific antibodies. European Journal of Biochemistry 219, 829–35. LANC E T , D. (1986) Vertebrate olfactory reception. Annual Reviews in Neurosciences 9, 329–55. LANC E T , D. (1994) Exclusive receptors. Nature 372, 321–2.

ME N C O, B . P H . M. , TE KU L A, F . D. , FA RB MA N, A. I . & DAN H O, W . (1994) Developmental expression of G-

LE IB OV IC I , M. , L A PO IN T E, F ., ALE TT A , P. & AY E R-L E L IE© V RE, C. (1996) Avian olfactory receptors:

differentiation of olfactory neurons under normal and experimental conditions. Developmental Biology 175, 118–31. LE UN I SSEN , J . L . M. (1990) Background suppression using Aurion BSA-C and/or Tween-20'. Aurion Newsletter 1. ME N C O, B. P H . M. (1980a) Qualitative and quantitative freeze-fracture studies on olfactory and nasal respiratory epithelial surfaces of frog, ox, rat, and dog. II. Cell apices, cilia, and microvilli. Cell and Tissue Research 211, 5–30. ME N C O, B . P H . M. (1980b) Qualitative and quantitative freeze-fracture studies on olfactory and nasal respiratory epithelial surfaces of frog, ox, rat, and dog. IV. Ciliogenesis and ciliary necklaces (including highvoltage observations). Cell and Tissue Research 212, 1–16.

proteins and adenylyl cyclase in peripheral olfactory systems. Light microscopic and freeze-substitution electron microscopic immunocytochemistry. Journal of Neurocytology 23, 708–27. MO RI, K. & Y OSH IH A RA , Y. (1995) Molecular recognition and olfactory processing in the mammalian olfactory system. Progress in Neurobiology 45, 585–619. NAK AM UR A, T. , L E E , H . - H . , K O B AY ASH I , H . & SA T OH , T . -O . (1996) Gated conductances in native

and reconstituted membranes from frog olfactory cilia. Biophysical Journal 70, 813–17. NGA I, J., C H E S S, A. , DO WL IN G, M. M. , N E C L E S, N . , MA C AG NO, E . R. & A XE L , R. (1993) Coding of olfac-

tory information: topography of odorant receptor expression in the catfish olfactory epithelium. Cell 72, 667–80. PA TE RNO ST RO , M . A . & ME IS A MI, E . (1991) Lack of thyroid hormones but not their excess affect the maturation of olfactory receptor neurons: a quantitative morphology study in the postnatal rat. International Journal of Developmental Neuroscience 9, 439–52.

706

M E N C O , C U N N I N G H A M , Q A S B A , L E V Y and R E E D

PH IL L IPS, G. W. & BRIDGM AN, P. C . (1991) Immuno-

electron microscopy of acetylcholine receptors and 43 kD protein after rapid freezing, freeze-substitution, and low temperature embedding in Lowicryl K11M. Journal of Histochemistry and Cytochemistry 39, 625–34. PH IL L IPS, T. E . & B OYN E, A. F . (1984) Liquid-nitrogen bound quick freezing: experiences with bounce-free delivery of cholinergic nerve terminals to a metal surface. Journal of Electron Microscopy Technique 1, 9–29. RA MIN G, K . , K RIE GE R , J . , ST R OT M AN N, J. , B OE KHOFF, I. , K UB ICK , S. , B A UM S T AR K, C . & BR EE R , H .

(1993) Cloning and expression of odorant receptors. Nature 361, 353–6. RE SSL E R, K . J . , SULL IV A N , S. L . & B U C K , L . B . (1993) A zonal organization of odorant receptor gene expression in the olfactory epithelium. Cell 73, 597–609. RE SSL E R, K . J . , SULL IV A N , S. L . & B U C K , L . B . (1994) A molecular dissection of spatial patterning in the olfactory system. Current Opinion in Neurobiology 4, 588–96. SEI FE RT, K . (1970) Die Ultrastruktur des Riechepithels beim Makrosmatiker. Eine elektronenmikroskopische Untersuchung. In Normale und pathologische Anatomie, Heft 21 (edited by B AR GMA N N , W . & DOE RR , E.). Stuttgart: Georg Thieme Verlag. SE N GU PT A , P. , C H O U, J. H . & B A RG M AN N, C . I . (1996) odr-10 Encodes a seven transmembrane domain olfactory receptor required for responses to the odorant diacetyl. Cell 84, 899–909. SHE PH E RD, G. M ., SI NG E R, M. S. & G RE E R, C . A . (1996) Olfactory receptors: a large gene family with broad affinities and multiple functions. The Neuroscientist 2, 262–71. SIN GER, M. S. , WE ISI N G E R-L E W IN , Y ., L A NC E T, D. & SH E P H E RD, G. M. (1996) Positive selection moments

identifying potential functional residues in human olfactory receptors. Receptors and Channels 4, 141–8.

STRO TMA N N , J. , WA N NE R, I. , H E L F RI C H, T. , B E C K , A. , M E I N KE N , C ., K U B IC K , S. & B RE E R , H . (1994)

Olfactory neurones expressing distinct odorant receptor subtypes are spatially segregated in the nasal neuroepithelium. Cell and Tissue Research 276, 429–38. STRO TMA N N , J. , W A NN E R, I. , H E L F RIC H , T ., & B RE E R , H . (1995) Receptor expression in olfactory

neurons during rat development: in situ hybridization studies. European Journal of Neuroscience 7, 492–500. SUL L IV A N , S. L . & DR YE R, L . (1996) Information processing in mammalian olfactory system. Journal of Neurobiology 30, 20–36. SULL IV A N , S. L . , A DA MS ON , M . C. , RE SSL ER, K . J. , KO ZA K , C . A. & B U C K , L . B . (1996) The chromo-

somal distribution of mouse odorant receptor genes. Proceedings of the National Academy of Sciences (USA) 93, 884–8. SUNG , C .-H . , SC H NE IDE R, B . G ., A GA RWA L , N ., P APE RM AST E R , D. S. & N ATH A N S, J. (1991) Functional

heterogeneity of mutant rhodopsins responsible for autosomal dominant retinitis pigmentosa. Proceedings of the National Academy of Sciences (USA) 88, 8840–4. TAK AG I, S. F. (1989) Human Olfaction. Tokyo: University of Tokyo Press. TUR I N , L. (1996) Spectroscopic mechanism for primary olfactory reception. Chemical Senses 21, 773–91. VA N L OO K E RE N CA MPA GN E , M. B . , OE STRE IC H E R , B ., VA N DE R KRIF T, T . P. , GISPE N, W . H. & VE RKL EIJ, A. J. (1991) Freeze-substitution and Lowicryl

HM20 embedding of fixed rat brain: suitability for immunogold ultrastructural localization of neural antigens. Journal of Histochemistry and Cytochemistry 39, 1267–79. VA SSA R, R ., N GA I, J . & A XE L, R . (1993) Spatial segregation of odorant receptor expression in the mammalian olfactory epithelium. Cell 74, 309–18.

.