JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ryba, N. J. P.
Right arrow Articles by Tirindelli, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ryba, N. J. P.
Right arrow Articles by Tirindelli, R.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Volume 270, Number 12, Issue of March 24, 1995 pp. 6757-6767
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
A Novel GTP-binding Protein -Subunit, G8, Is Expressed during Neurogenesis in the Olfactory and Vomeronasal Neuroepithelia (*)

(Received for publication, August 26, 1994; and in revised form, January 17, 1995)

Nicholas J. P. Ryba (1), (§),  Roberto Tirindelli (2)

From the  (1)Laboratory of Immunology, NIDR, National Institutes of Health, Bethesda Maryland 20892 and (2)Fisiologia Umana, Universita di Parma, Via Gramsci 14, 43100 Parma, Italy

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES

ABSTRACT

A novel heterotrimeric G-protein -subunit has been cloned, and its function has been confirmed by expression and purification. This -subunit is only detected in the olfactory epithelium, the vomeronasal epithelium and, to a lesser extent, the olfactory bulb. It is absent from all other tissues studied including the nasal respiratory epithelium. During development, expression of G8 in the olfactory epithelium parallels neurogenesis, peaking shortly after birth and declining in the adult. In situ hybridization studies localize expression of this novel -subunit to the sensory neurons; hybridization is strongest in the region of the epithelium that contains immature neurons. Unlike proteins that are expressed only in mature olfactory neurons (e.g. olfactory marker protein or Golfalpha), expression of G8 in the olfactory epithelium is relatively unaffected by olfactory bulbectomy. In the vomeronasal epithelium expression of G8 is also highest in the developing neurons. Taken together, these findings are consistent with a very specific role for G8 in the development and turnover of olfactory and vomeronasal neurons.

INTRODUCTION

Heterotrimeric G-proteins (^1)are central to a wide variety of receptor-effector coupling pathways(1, 2) . It was believed that the alpha-subunit of these proteins was the only critical determinant of G-protein receptor and G-protein effector interaction. However, it is now becoming clear that the diverse beta-subunits (2) also have distinct roles. One of the first examples of this was the mating response pathway in yeast where molecular genetic experiments demonstrated that the beta-subunits are responsible for signaling and that the alpha-subunit has an inhibitory role(3) . Interactions between different Gbetas and specific G-protein-linked receptors have been shown to vary in vitro(4) . Recently, in vivo coupling of specific receptors to effectors has also been shown to be determined by the nature of the beta-subunit (5) and the -subunit (6) . A number of different effector enzymes are influenced by beta-subunits, for example different subtypes of PLCbeta (7, 8) and adenylate cyclase (9) differ in sensitivity to beta-subunits. Another role that beta-subunits appear to play is in desensitization of G-protein-linked receptor pathways by recruitment of the G-protein receptor kinase, beta-adrenergic receptor kinase-1, to the membrane(10, 11) , and possibly of other more diverse proteins involved in signal transduction(12) . The diversity of beta- and -subunits also seems able to influence the interaction between effectors and beta-subunit (13, 14) . The role of heterotrimeric G-proteins in control of cell fate and development has been documented in several organisms. For example, G-proteins mediate cell-cycle arrest in haploid Saccharomyces cerevisiae(3) and the growth and development of Dictyostelium discoidium(15) . In multicellular organisms, G-proteins have also been implicated as mediators of development; for example the developmental mutant of Drosophila, concertina, results from a defect in the gene for a G-protein alpha-subunit(16) .

The understanding of olfactory signal transduction has advanced rapidly over the last few years. Proteins that appear to have a role in G-protein-mediated coupling of olfactory receptors to cAMP-controlled ion flux have been cloned from olfactory epithelium and have been shown to be highly enriched in the sensory cilia(17, 18, 19) . More recently, much attention has focused on what appears to be a very large family of G-protein-linked olfactory receptors(20, 21, 22) . However, no coupling between these receptors and the cAMP second messenger pathway has been shown yet. Other signaling pathways have also been suggested to have a role in olfactory signal transduction. These pathways include pertussis toxin-sensitive G-protein-dependent stimulation of PLC(23) , a pathway which may involve stimulation of PLC-beta2 by G-protein beta-subunits in many cells(7, 8) . Therefore, we were interested in the diversity of G-protein -subunits in the olfactory epithelium.

Among vertebrate neurons, the olfactory and vomeronasal neurons are unique in that they turnover throughout life. Olfactory neurons are replaced through differentiation of the basal cells of the olfactory epithelium(24, 25) . The vomeronasal organ possesses a neuroepithelium like that of the olfactory epithelium. However, the role of this organ appears to be in perception of stimuli related to social and/or reproductive behavior in many species(26, 27) . As in the olfactory epithelium, the neural receptor cells of the vomeronasal organ appear to turnover throughout life with the principal regions of neurogenesis being at the junctions between sensory and nonsensory epithelia(28, 29) . In studying the diversity of G-protein subunits that may have roles in olfaction, we have characterized a novel -subunit which was expressed specifically in neural cells in the olfactory and vomeronasal epithelia. The expression of this G-protein -subunit was not limited to mature neurons as is the case for proteins believed to be involved in olfactory signal transduction, but was highest in developing neurons, suggesting a signaling role for a novel heterotrimeric G-protein in neurogenesis in these tissues.

EXPERIMENTAL PROCEDURES

Oligonucleotides

Oligonucleotides were synthesized (40 nmol scale) using an Applied Biosystems Inc. model-392. They were analyzed by acrylamide gel electrophoresis and were used without purification: oligo 1, GTIGA(A/G)CA(A/G)CTIAAGATIGA(A/G)G; oligo 2, CTTCTTITCICGGAAIGG(A/G)TT; oligo 3, CGTTTCGCAGAAAGCCAATAGCTC; oligo 4, TCTAGATCTTTTTTTTTTTTTTTTTT; oligo 5, GCCTCAGCGATCTTGGC; oligo 6, TTGGCCATGTTGTTGGACATGGCT; oligo 7, GGAGGATTCATTATTGCAGG; oligo 8, GGGAGGATCCAACTATCTTGGGG; oligo 9, CCGCCAGGATCCCAGCCCTGAGC.

Cloning of the -Subunit

The methods used were essentially as described in standard molecular biology texts(30) . Partially degenerate oligonucleotides (oligo 1 and oligo 2) corresponded to conserved regions of G-protein -subunits (see Fig. 2). Template for PCR amplification was derived by in vivo whole library excision (31) of an olfactory epithelium cDNA-library in -Zap II (Stratagene). PCR amplification was carried out using 20 ng of template and 100 pmol of each primer in a Perkin Elmer 9600 thermocycler (95 °C, 270 s; then 30 cycles, 95 °C, 20 s; 50 °C, 30 s; 72 °C, 30 s; followed by 72 °C, 600 s). The single detectable PCR product (150 bp) was purified by gel electrophoresis and was ligated to pBluescript (Stratagene) for sequencing using Sequenase II (United States Biochemical Corp.).

Figure 2: Comparison of the predicted protein sequence of G8 to previously reported Gs. The sequence of the novel G was aligned with the sequences of several mammalian and one insect (D-1) subtypes of G subunit(53, 54, 55, 56, 57, 58, 59) . Residues in other Gs identical to those in G8 are indicated by a solid line above the sequence, residues that are similar by a colon above the sequence. Gaps introduced into the sequences to optimize the alignment are represented by periods. The sequences of G4 and G-S1 are incomplete. The regions of sequence of G2 used for design of the partially degenerate primers, oligo 1 and 2, are underlined.


An oligonucleotide specific for the new -subunit (oligo 3) was designed, 5`-end labeled and was used to screen approximately 150,000 plaques of an olfactory cDNA library in gt10 (hybridization 45 °C: 5 times NET (0.15 M NaCl, 15 mM Tris-HCl, pH 8.3, 1 mM EDTA), 5 times Denhardt's, 100 µg/ml yeast tRNA, 0.25% SDS; stringency washes, 40 °C, 0.5 times SSC (0.15 M NaCl, 15 mM sodium citrate pH 7.0). Several hybridizing plaques were purified, and the cDNA inserts were excised and subcloned into pBluescript for analysis. None of these initial isolates represented a full-length transcript of the -subunit, therefore 5`-RACE was carried out. Oligo 3 was used to prime first strand cDNA synthesis using total olfactory RNA as a template (10 µg of RNA, 10 pmol of oligo 3). Residual nucleotides and primer were removed by two rounds of dilution and concentration using ultrafiltration (Centricon 100). The cDNA was tailed using terminal transferase and dATP. PCR amplification was carried out using oligo 4 and oligo 5 (95 °C, 180 s; three cycles 95 °C, 20 s; a linear ramp lasting 90 s from 45 to 72 °C; 72 °C, 600 s; then 35 cycles 95 °C, 15 s, 50 °C, 30 s, 72 °C, 60 s, followed by 72 °C, 600 s). The single 150-bp product was gel purified and cloned into pBluescript for sequence analysis.

Oligo 6 was synthesized on the basis of the sequence of the 5`-RACE product and was used to screen 100,000 clones of the olfactory -Zap II library (hybridization 65 °C: 5 times NET, 5 times Denhardt's, 100 µg/ml yeast tRNA, 0.25% SDS; stringency washes: 60 °C, 0.5 times SSC). A single hybridizing plaque was isolated. Using the in vivo excision protocol, the insert was obtained in pBluescript. The sequence of this clone was determined for both strands using Sequenase II. To check whether the 5`-non-coding region of this clone, not contained within the 5`-RACE product, was an alternative (longer) form of G, PCR was carried out using oligo 7 and oligo 3. Total olfactory epithelium RNA (10 µg) was used as a template for cDNA synthesis using an oligo(dT) primer. PCR amplification was carried out with 150 pmol of each primer and 10 ng of cDNA (95 °C, 270 s; 30 cycles 95 °C, 30 s; 55 °C, 30 s; 72 °C, 30 s; followed by 72 °C, 600 s).

Expression of G8 in Sf9 Cells

Insect larval Sf9 cells were grown in serum-free medium (Sf9-II, Life Technologies, Inc.), and standard techniques were used for construction and propagation of recombinant baculovirus (32) . The PCR product obtained using oligos 8 and 9 was treated with BamHI and ligated with pBacpak (Clonetech) linearized with BamHI to yield pB. The sequence of the construct pB was confirmed across the cloning junction and the full-length of the G8 insert. Recombinant baculovirus was obtained by homologous recombination following cotransfection of Sf9 cells with pB and Bacpak-1 (Clonetech) DNA. Virus was plaque purified, was confirmed to contain the G8 insert by Southern analysis, and was amplified in suspension culture. Recombinant viruses encoding the G-protein subunits Gbeta1 and G2 (kindly provided by Dr. J. Northup, National Institute of Mental Health, NIH) were also amplified. For expression studies, suspension cultures of Sf9 cells were infected at an multiplicity of infection of 1:1 with recombinant baculovirus encoding Gbeta1 and either G8 or G2.

Purification of Recombinant Gbeta

10^9Sf9 cells (1 liter) were infected with recombinant baculoviruses and were grown in shaker culture. 64 h after infection cells were pelleted, washed two times with PBS, and were lysed in 50 ml of lysis buffer (50 mM HEPES, pH 7.5, 5 mM EDTA, 1 mM DTT, 100 µM aminoethyl-benzenesulfonyl fluoride, AEBSF) by homogenization using a glass-Teflon homogenizer on ice. Cell debris was removed by centrifugation (5 min, 5000 times g), was washed with a further 30 ml of lysis buffer, and centrifuged a second time. Supernatants were combined and membranes pelleted by centrifugation (250,000 times g, 60 min). Membranes were washed by resuspension in 50 ml of EED (EED, 10 mM HEPES, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol) containing 100 µM AEBSF followed by centrifugation as above. Membrane protein was solubilized by homogenization in 30 ml of 1% cholate in EED. Insoluble material was removed by centrifugation (250,000 times g, 30 min), and the supernatant was applied to a 10-ml DEAE-Sephacel (Pharmacia) column, pre-equilibrated with EED, 1% cholate, 25 mM NaCl. The column was developed by washing with 3 volumes, 25 mM NaCl in EED, 1% cholate, followed by a 40-ml linear gradient of 25-400 mM NaCl in this buffer. beta-Heterodimers eluted as a broad peak from 100 to 300 mM NaCl. This peak was combined, diluted 10 times with EED, 250 mM NaCl, and applied to a 10-ml phenyl-Sepharose column, pre-equilibrated with 250 mM NaCl in EED, 0.15% cholate. The column was washed with 10 ml of equilibration buffer, and beta-heterodimers were eluted by application of a 40-ml gradient of 0.15-2.5% cholate. beta-Heterodimers eluted as a peak centered at 1% cholate were concentrated to 2 ml by ultrafiltration (Amicon, YM-30 membrane) and were subjected to gel filtration over Sephacryl HR-100 equilibrated with 250 mM NaCl in EED, 1% cholate. beta-Heterodimers eluted as a single sharp peak with an estimated size of 55 kDa.

Functional Studies of beta-Heterodimers

The initial rate of activation of transducin by a limiting concentration of rhodopsin (30 nM) was studied as a function of beta-heterodimer concentration by monitoring GTP--S binding as has been previously described(4) . Gbeta in 1% cholate was mixed with rhodopsin and 1 µM GTP--S (200 pmol/µCi) on ice, and the reaction was started by addition of transducin (diluting cholate to 0.1%) and warming rapidly to 30 °C. After 10 min the reaction was terminated by 50-fold dilution with ice-cold buffer. Protein-bound GTP--S was separated from free GTP--S by filtration through nitrocellulose and was determined by liquid scintillation counting.

Northern Analysis

Total RNA and A-RNA were isolated from a number of tissues and from the olfactory epithelium of bulbectomized rats (6 days post-operation). RNA was denatured in the presence of glyoxal and was size fractionated using 1.2% agarose gel electrophoresis. RNA was transferred to Nytran membranes. The full coding sequence of G8 was amplified using the polymerase chain reaction. 10 ng of the pBluescript full-length clone was used as template for amplification with 150 pmol of oligo 8 and oligo 9 (95 °C, 270 s; then 25 cycles 95 °C, 20 s; 55 °C, 20 s; 72 °C, 30 s; followed by 72 °C, 600 s). The product was gel purified and used for generation of probes by random priming. Hybridization was in 5 times SSC containing 5 times Denhardt's solution, 100 µg/ml yeast tRNA, 100 µg/ml sheared salmon sperm DNA, 0.5% SDS. High stringency washing was for 20 min in 0.2 times SSC at 55 °C.

RNase Protection Assays

The PCR product, obtained using oligo 8 and oligo 9, was treated with BamHI and was ligated into pBluescript. PCR was also used to clone parts of the coding regions of beta-actin and Golfalpha into pBluescript. Plasmids were linearized using HindIII and antisense cRNA for Golfalpha, beta-actin and the novel -subunit were transcribed using T7-RNA polymerase. For control experiments and quantitation, sense cRNA for the -subunit was also synthesized. cRNA was radiolabeled by incorporation of [alpha-P]GTP. After RNA synthesis, template DNA was degraded by treatment with 20 units of DNase (30 min, 37 °C). Unincorporated radionucleotide was separated from cRNA using Sephadex G-50 spin-columns.

RNA from a variety of tissues (10 µg) was denatured (10 min, 85 °C) in 30 µl of hybridization buffer (80% formamide, 0.4 M NaCl, 40 mM PIPES, pH 6.7, 1 mM EDTA) containing antisense RNA for the -subunit, Golfalpha, and actin. Sense cRNA for the -subunit was mixed with 10 µg of yeast tRNA and was denatured with the same mixed antisense probe. Denatured RNA was cooled rapidly to 45 °C and was incubated for 16 h to allow hybridization. Hybridized samples were cooled to room temperature and treated with RNase-T1 (300 µl, 2 µg/ml; 60 min), followed by proteinase K (0.6 mg/ml; 30 min, 37 °C). RNA was denatured, fractionated on 6% acrylamide sequencing gels containing 8 M urea, and protected RNA was visualized using autoradiography.

Olfactory Bulbectomy

Wistar rats were anesthetized, and the olfactory bulbs were aspirated with a glass pipette. The surgical site was packed with sterile gel-foam, and the skin was sutured. The animals were allowed to recover, and the olfactory epithelium was isolated 6-days post-operatively.

In Situ Hybridization

The product of the PCR, using oligo 8 and oligo 9, was cut with BamHI and was ligated into BamHI, BglII cut pSP72 (Promega) to generate both sense and antisense orientations when linearized with BamHI. This procedure minimized probe sequence from the vector. Template for each orientation was transcribed in the presence of digoxigenin-UTP (Boehringer Mannheim) according to the manufacturer's protocol or using [alpha-S]UTP.

S-Labeled Probes

Olfactory turbinates were dissected from an adult rat and fixed in 4% paraformaldehyde for 6 h at 4 °C. Tissue was embedded in paraffin, and 5-µm sections were cut and mounted on silanized slides, heated at 45 °C overnight, and stored at 4 °C. Paraffin was removed with xylene, sections were rehydrated using an ethanol series, and postfixed for (4% paraformaldehyde, 20 min). Preparation for hybridization was as described (32) and included incubation in 0.2 M HCl (5 min), proteinase K digestion (20 mg/ml; 10 min), post-fixation (4% paraformaldehyde; 5 min), treatment with iodoacetamide (0.37 g/400 ml) and N-ethylmaleimide (0.25 g/400 ml) for 30 min at 45 °C, reaction with acetic anhydride (0.5% in 0.1 M triethanolamine-HCl, pH 8.0; 2 times 10 min) and dehydration using a graded series of ethanol. Sections were hybridized with 0.4 ng/µl probe in 50% formamide, 10% dextran sulfate, 4 times SSC, 10 mM DTT, 1 times Denhardt's solution, 500 µg/ml each of salmon sperm DNA and yeast tRNA under silanized coverslips for 16 h at 50 °C in a humid chamber. Washing was 15 min in 2 times SSC at 50 °C; 20 min in 50% formamide, 2 times SSC, 20 mM DTT at 65 °C; 2 times 10 min TEN (10 mM Tris-HCl, pH 7.5, 5 mM EDTA, 0.5 M NaCl) at 37 °C; 30 min in TEN containing 20 µg/ml RNase A; 10 min in TEN at 37 °C; 2 times 15 min in 2 times SSC at 65 °C; 2 times 15 min in 0.1 times SSC at 65 °C. Sections were dehydrated in an ethanol series containing 0.3 M ammonium acetate, were covered with NTB-2 emulsion (Kodak) and were exposed for 3 weeks at 4 °C. After development sections were stained with 0.1% toluidine blue, were dehydrated, cleared with xylene, and mounted with Permount.

Digoxigenin-labeled Probes

Rats (Wistar males) anesthetized with sodium penthobarbital were perfused intracardiacally with PBS; this was followed by 100 ml of Bouins fixative or 4% paraformaldeyde in PBS. After perfusion, the olfactory turbinates were dissected, post-fixed at 4 °C for 4 h, and decalcified in 250 mM EDTA, pH 8, overnight. Tissue was then cryoprotected in 400 mM sucrose in PBS, included in Tissue Tek, and rapidly frozen in liquid nitrogen-cooled pentane. Tissue sections, cut at a nominal 16 µm, were mounted on poly-L-lysine-coated slides and desiccated at 45 °C overnight before storage at 4 °C. Sections were rehydrated in an ethanol series (100, 95, 85, and 70%), 1 times SSC and water. Sections that were fixed with paraformaldeyde were treated with protease (5 µg/ml proteinase K for 15 min at 37 °C) followed by post-fixation in 4% paraformaldeyde in PBS for 5 min. Acetylation was carried out using 0.25% acetic anhydride in 100 mM triethanolamine, pH 8, for 15 min. Slides were rinsed in 1 times SSC and dehydrated in stages from 70 to 100% ethanol. The hybridization mix contained: 50% formamide, 4 times SSC, 10% dextran sulfate, 1 times Denhardt's solution, 1 mg/ml yeast tRNA, 100 µg/ml denatured salmon sperm DNA, and 2-5 µg/ml riboprobe (40 µl/section covered with parafilm). Hybridization was carried out at 57 °C overnight in a humid chamber. After hybridization sections were soaked in 2 times SSC to remove parafilm and were washed three times (15 min each in 2 times SSC). Sections were then washed: 50% formamide, 2 times SSC at 55 °C, 30 min followed by 2 times SSC, 37 °C, 10 min; treated with 50 µg/ml RNase A at 37 °C for 30 min in 2 times SSC; and washed in 0.5 times SSC, 58 °C, 30 min. Visualization of the hybridized riboprobe was according to the manufacturer's protocol using an antidigoxigenin antibody conjugated to alkaline phosphatase (1: 500).

RESULTS

Cloning of a Novel G from the Olfactory Epithelium

A PCR product of 150 bp was obtained from olfactory epithelium cDNA primed with partially degenerate primers to conserved regions of Gs. The PCR product was cloned into pBluescript and the sequence of six constructs determined. The sequences of four of these were very closely related to the sequence of bovine G3, whereas the sequences of the other two were distinct from but related to the sequences of this region of all known G-protein -subunits. Restriction analysis of 36 other subclones from the PCR product indicated that all were either G3 or the novel G (data not shown). Therefore, to facilitate isolation of a specific clone of the novel -subunit an oligonucleotide based on the sequence of the most variable region of the new -subunit cDNA was synthesized. Using this oligonucleotide as a probe, four hybridizing plaques were identified in 150,000 plaques of an olfactory epithelium library. The longest of these four clones appeared to encode almost the entire sequence of the novel -subunit, G8, but on the basis of homology with other known G-protein -subunits, still lacked the 5` starting ATG (Fig. 1). The sequence determined for the other three clones was contained within that of this clone.

Figure 1: The cDNA sequence of the novel G-protein -subunit, G8. The nucleotide sequence of the full-length clone of the novel G-protein -subunit was determined from both strands. The amino acid sequence predicted for the protein is shown above the nucleotide sequence. The start site for the longest clone obtained from screening of the gt10 library is indicated by double underlining. The positions of the primers used in the cloning and generation of the full coding sequence for expression and generation of probes are indicated by underlining. The product of 5`-RACE was identical to the sequence shown except that the 5`-residue, indicated by an arrow, was not A but G. The isoprenylation motif at the C terminus of the protein and the consensus Olf1-binding site are also highlighted.


An oligonucleotide was synthesized on the basis of the sequence of 5`-RACE (5` to the coding region) and was used to rescreen the olfactory cDNA library. A single full-length clone of the novel -subunit (containing the sequence determined by 5`-RACE and that of all partial clones) was isolated from 100,000 plaques. The sequence of this clone (determined for both strands) is shown in Fig. 1. The predicted protein sequence of G8 shares features present in other Gs, most notably, the 3`-isoprenylation site (-CaaX) and the small size (7 kDa). At the amino acid level, the predicted sequence identity to known mammalian Gs is in the range 25-70%. The most similar known G-subunit is G2. An alignment of the novel G and other subtypes from several species is shown in Fig. 2. The full-length clone isolated from the library contained sequence that extended beyond the 5` end of the RACE product. An appropriate sized product was obtained when olfactory cDNA was used as a template for PCR amplification with a primer within this region of 5`-extended sequence and a second in the coding sequence of the G (data not shown).

Heterologous Expression of G8

G8 was expressed in insect-larval cells infected with baculovirus containing the G8 coding sequence under the control of the polyhedrin promoter. Expression of G8 could be detected by [S]methionine incorporation in cells infected only with virus expressing this protein (data not shown). Mixing of membranes of cells expressing G8 with membranes or soluble protein extracts from cells expressing Gbeta1 did not result in the formation of beta-heterodimers capable of stimulating the activation of transducin by rhodopsin. However, when cells were coinfected with two viruses one expressing G8 and the other Gbeta1, a cholate extract of cell membranes contained functional beta-heterodimers. Purification of the Gbeta18 to >90% purity was achieved by a combination of ion-exchange, hydrophobic interaction, and gel filtration chromatography (Fig. 3A). A yield of 15 nmol/10^9 cells of purified Gbeta18 was achieved.

Figure 3: Heterologous expression of G8. G8 and G2 were expressed as heterodimers with Gbeta1 in insect larval cells (Sf9). A, purified heterodimers were analyzed by SDS-PAGE, 16% Tricine gel (60) stained with Coomassie Blue. Lane 1, Gbeta18; lane 2, Gbeta12; M molecular weight standards. B, the efficacy of beta-heterodimers at stimulating the rhodopsin-dependent activation of transducin: the concentration dependence of the initial rate of beta-dependent stimulation of rhodopsin-mediated activation of transducin was determined using a 10-min standard reaction. This contained 30 nM regenerated rhodopsin, 0.2 µM transducin, 2 µM GTP--S, 0.1% cholate, and either Gbeta18 or Gbeta12 (indicated in the inset) at concentrations shown. Activation of transducin was determined by GTP--S binding. Full activation of transducin was achieved by 1 h of incubation with 500 nM brain Gbeta and 2 µM rhodopsin and corresponded to the calculated saturation. Data points are the mean ± standard deviation of triplicates; curves are single-site fits with half-maximal values of 120 nM for Gbeta18 and 40 nM for Gbeta12. C, time course of Gbeta-dependent stimulation of rhodopsin-mediated activation of transducin: standard reactions containing 90 nM Gbeta18 (three independent expressions and purifications indicated by filled or open triangles) or 30 nM Gbeta12 (two independent expressions and purifications indicated by filled or open circles) were incubated for 90 min. Aliquots were removed at indicated times and transducin activation measured by GTP--S binding. Curves are simple exponential fits of the means obtained for the three preparations of Gbeta18 and the two preparations of Gbeta12. For comparison the activation of 0.2 µM transducin by 30 nM rhodopsin in the absence of beta-heterodimers is shown.


The efficacy of purified Gbeta18 at stimulating rhodopsin-mediated activation of transducin was compared with that of Gbeta12 expressed and purified under identical conditions. Both beta-heterodimers had a profound effect on the initial rate of GTP--S binding to transducin (Fig. 3B). For the conditions shown, 30 nM rhodopsin and 0.2 µM transducin, the half-maximal rate of binding was achieved at a Gbeta18 concentration of 120 nM, whereas this rate was reached when 40 nM Gbeta12 was added. In time course experiments, the effectiveness of three separate preparations of Gbeta18 was compared with that of two preparations of Gbeta12 (Fig. 3C). In all cases 90 nM Gbeta18 resulted in very similar stimulation of GTP--S binding by transducin to that produced by 30 nM Gbeta12.

Tissue Distribution of Expression of G8

The tissue distribution of the novel G was determined by Northern analysis. A probe made from the entire coding region of the new G (see Fig. 1for details) detected a major transcript in RNA from the olfactory epithelium of 600 nucleotides when hybridized and washed at high stringency (Fig. 4A). Under these conditions no hybridization was detected to RNA from several other tissues; no cross-hybridization of the probe to G8 was detected to the most closely related G2 which is present at high abundance in brain. Staining of ribosomal RNA (not shown) demonstrated that similar amounts of RNA were loaded for the different tissues. Two larger transcripts were also present in the olfactory epithelium (Fig. 4B). The difference in length of the 5`-RACE product and the full-length clone that was isolated suggests that there may be transcripts with different lengths of 5`-non-coding sequence. Olfactory bulbectomy was carried out to diminish the expression of olfactory neuron markers in RNA derived from the olfactory epithelium (Fig. 4B). This treatment had a substantial effect on the level of expression of Golfalpha and OMP, no effect on the level of actin expression, and relatively little influence on the expression of the novel G.

Figure 4: Tissue distribution of G8: Northern analysis. Northern analysis of the tissue distribution of G8 indicated that its expression was specific to the olfactory epithelium and was not found in other tissues. A, 20 µg of total RNA from brain (1), heart (2), intestine (3), kidney (4), liver (5), lung (6), olfactory epithelium (7), and testis (8) was probed for transcripts that hybridized at high stringency with cDNA probes to the full coding sequence of G8. B, 2 µg of A-RNA from brain (1), olfactory epithelium (2), olfactory epithelium of bulbectomized rats (3), and testis (4) was probed under identical conditions to A; the blot was stripped and reprobed for beta-actin (center panel) and was stripped and hybridized with a mixed Golfalpha and OMP probe (lower panel; both cDNAs used to generate probes were of similar length but the relative specific activity of the Golfalpha probe was 10 times higher than that of OMP).


RNase protection was used to make a more detailed quantitation of the expression of G8 in the olfactory epithelium, in other nasal epithelia, and in a wide variety of other adult rat tissues (Fig. 5). G8 mRNA was detected at comparable levels in the olfactory and vomeronasal epithelia; a much lower level was present in the olfactory bulb, and no G8 mRNA was detected in any other tissue. This pattern of expression was entirely consistent with that detected using Northern analysis (Fig. 4A). Based on protection of antisense G8 cRNA by known amounts of sense cRNA, G8 was estimated to be expressed at a level of about 10^5 copies/µg total RNA (2 molecules in 10^5 of mRNA). Preliminary experiments to investigate the relative amounts of G8, Golfalpha, and beta-actin expression in the olfactory epithelium indicated that in the adult rat a ratio of about 1:100:1000, G8/Golfalpha/beta-actin was present. Therefore, in order to carry out RNase protection assays shown in Fig. 5, the specific activities of the Golfalpha and beta-actin probes were reduced 100 and 1000-fold, respectively, by increasing the concentration of GTP in the labeling reaction. The relative specific activities of actin/Golfalpha/G8 probes used in the RNase protection assays shown in Fig. 6were 1:20:1000. The relatively even protection of cRNA probes for actin and G8 indicated that in the olfactory epithelium of adult rats the expression of actin is about three orders of magnitude higher than that of G8; quantitation of Golfalpha and G8 is less precise but it appears that about 10-fold higher levels of Golfalpha are expressed than of G8 ( Fig. 5and Fig. 6). The adult vomeronasal epithelium contained a similar ratio of G/actin RNA to that observed in the olfactory epithelium but, unlike in the olfactory epithelium, no expression of Golfalpha was detected. The olfactory bulb RNA contained 10-fold less G RNA relative to actin RNA. G8 RNA was not detected in several regions of the brain, nor was it present in a number of other tissues including the nasal respiratory epithelium. However, in contrast to G8, which was only expressed in the olfactory tract, considerable expression of Golfalpha could be detected in the brain (Fig. 6A).

Figure 5: Tissue distribution of G8: RNase protection. 10 µg of total RNA isolated from a number of different adult rat tissues was analyzed by RNase protection. The relative specific activities of probes for G8, Golfalpha, and actin were 1000:100:1, respectively. Protected fragments were the full coding sequence (0.26 kb) for G8 and were from the coding sequence of Golfalpha (0.45 kb) and actin (0.4 kb). A, cerebellum (1), brainstem (2), mid-brain (3), frontal lobe (4), olfactory bulb (5), eye-cup (6), olfactory epithelium (7), respiratory epithelium (8), vomeronasal organ (9) heart (10), intestine (11), kidney (12), liver (13), lung (14), muscle (15), spleen (16), testis (17), tongue (18), tRNA (19 and 20). Too little RNA from respiratory epithelium relative to that from other tissues was used in A, therefore the assay was repeated with new preparations of RNA in B: respiratory epithelium (1), vomeronasal epithelium (2), olfactory epithelium (3), olfactory bulb (4), frontal lobe (5), tRNA (6). The positions of protected bands and the undigested G8 probe are marked.


Figure 6: Developmental changes in G8 expression. 10 µg of total RNA from total brain or from olfactory epithelium were analyzed by RNase protection at different times before and after birth. The relative specific activities of probes for G8, Golfalpha, and actin were 1000:20:1, respectively. Protected fragments were the full coding sequence (0.26 kb) for G8 and were from the coding sequence of Golfalpha (0.45 kb) and actin (0.4 kb). A, comparison of expression of Golfalpha and G8 during development of brain and olfactory epithelium: brain E19.5 (1), brain P6.5 (2), brain adult (3), tRNA (4), olfactory E19.5 (5), olfactory P6.5 (6), olfactory P13.5 (7), olfactory adult (8), tRNA (9), vomernasal adult (10). B, expression of G8 in the olfactory epithelium was analyzed in more detail during early development P13.5 (1), P9.5 (2), P6.5 (3), P4.5 (4), P2.5 (5), P0.5 (6), E21.5 (7), E19.5 (8), whole head E13.5 (10), and tRNA (11). The positions of protected bands and the undigested G8 probe are marked.


Expression of G8 during Development of the Olfactory Epithelium

RNase protection was also used to analyze the expression of G8 on a number of days during development (E for embryonic and P for post-natal) both in the olfactory epithelium and in the brain (Fig. 6A). No expression above background was detected in brain RNA from E19.5 to adult, nor was expression detected in the whole head at E13.5 (Fig. 6B). In the olfactory epithelium, expression of G8 was detected at E19.5 and the level of expression increased to a maximum by P13.5. In adult rat olfactory epithelium, a lower level of G8 expression was detected than in the tissue from immature animals. In contrast, the level of expression of Golfalpha in the olfactory epithelium also increased over the first 2 weeks of life, but still higher levels were detected in the adult. The post-natal rise in Golfalpha expression appeared to be more dramatic but later than that of G8. Expression of Golfalpha detected in the brain also increased from birth to adult, but the change in expression was less pronounced than in the olfactory epithelium.

In Situ

Localization of G8 Expression in the Olfactory and Vomeronasal Epithelia-The cellular localization of the mRNA for G8 was determined by in situ hybridization of 5-µm sections of an adult rat olfactory epithelium with S-labeled probes to the full coding sequence of G8 (Fig. 7). At low magnification (Fig. 7, A-C), a layer of antisense-specific hybridization could be detected throughout the epithelium using dark-field optics. Higher magnification (Fig. 7, D and E) revealed that this layer of hybridization was localized at the basolateral side of the olfactory neural cell layer. At high magnification (Fig. 7, F and G), bright-field microscopy revealed antisense specific clusters of silver grains superimposed on the cell bodies of some cells in the region of the epithelium made up of olfactory neurons and developing neurons. Many olfactory neurons within this region appeared negative. No hybridization was detected to the sustentacular cells with nuclei and cell bodies at the apical surface, to the basal cells (at the basolateral surface), or to glandular cells (staining purple) toward the base of the epithelium.

Figure 7: Neural cell-specific localization of G8 in the olfactory epithelium. S-Labeled cRNA probes for the full coding sequence of G8 were hybridized with adult rat olfactory epithelium and were washed at high stringency; following autoradiography, tissue was lightly stained using toluidine blue. A, low magnification bright-field of a region of the olfactory epithelium hybridized with G8 sense cRNA shown using dark-field illumination (B); C, dark-field of an adjacent section hybridized with antisense cRNA; the boxed region of C is shown at higher magnification in bright-field (D); marked by arrows is the zone of hybridization clearly seen in dark-field (E); F, bright-field high magnification of a region of epithelium showing different cell layers hybridized with sense G8 cRNA; G, bright-field of an adjacent section hybridized with antisense cRNA; two representative clusters of silver grains are arrowed; the positions of the cell-bodies of the sustentacular cells (s) at the apical surface of the epithelium, the basal cells at the basolateral surface (b), and the olfactory (o) neurons are indicated. Bar = 36 µm.


A consistent antisense-specific hybridization of G8 to a subpopulation of olfactory neurons mostly with nuclei toward the base of the neural cell layer was also detected in thicker (16 µm) sections with digoxigenin-labeled probes (Fig. 8, A and B). The localization of G8 was studied after olfactory bulbectomy because significant expression of G8 was still detected (Fig. 4). The major histologic consequence of olfactory bulbectomy was a marked thinning of the olfactory epithelium resulting from loss of mature neurons. The expression of G8 appeared relatively unaffected by bulbectomy (Fig. 8C). Hybridization was not detected in the apical region of the epithelium that is made up of sustentacular cells but was strong toward the basolateral surface. In bulbectomized animals, the region containing G8 mRNA is made up of immature olfactory neurons not killed by bulbectomy and also by neurons that started to develop after bulbectomy.

Figure 8: Comparison of the cellular localization of G8 expression in the olfactory and vomeronasal epithelia. Digoxigenin-labeled cRNA probes for the full coding sequence of G8 were hybridized (and washed at high stringency) to sections of normal olfactory epithelium (A and B), epithelium isolated 7 days after bulbectomy (C), and the vomeronasal organ (D-G). A, C, D, and F were probed with G8 antisense cRNA; B, E, and G with G8 sense cRNA. A-C were lightly stained with eosin and were photographed using normal illumination. In D and E, regions of unstained sections of the vomeronasal epithelium were photographed using Nomarski optics. F and G show low magnification cross-section through the whole vomeronasal organ, under normal illumination; arrows indicate the primary regions of neurogenesis. High magnification (A-E), bar = 50 µm; low magnification (F) and (G), bar = 200 µm.


In the vomeronasal epithelium, the expression of G8 appeared somewhat higher than in the olfactory epithelium (e.g. compare Fig. 8, A and D). Antisense-specific hybrization of the digoxigenin-labeled probe to the full coding sequence of G8 in the perinuclear region of cells located throughout the sensory vomeronasal epithelium was detected (Fig. 8, D and E). This perinuclear distribution was consistent with the clustering of silver grains observed in the olfactory epithelium (Fig. 7G). At low magnification a clear gradation in the distribution of G8 mRNA was detected, with the highest expression being at the boundaries between the sensory and non-sensory regions of the epithelium (arrowed in Fig. 8F). No specific hybridization was detected in the convex, non-sensory (respiratory) epithelium (Fig. 8, F and G). Thus, the cellular distribution of mRNA for G8 indicates that the G-protein that contains this -subunit probably has a very similar role in the olfactory and vomeronasal epithelia. No hybridization of G8 was detected to sections of the olfactory bulb (data not shown) even though this tissue contained low levels of G8 mRNA (Fig. 5A).

DISCUSSION

We investigated whether novel G-protein -subunits might play a role in olfaction, and as a result identified and cloned a new -subunit, G8, (^2)from olfactory cDNA. The sequence of G8 is typical of known G-protein -subunits and contains a conserved site for C-terminal isoprenylation(34) . The -CaaL motif found at the C terminus of G8 is predicted to result in geranyl-geranylation of G8 in vivo with concomitant localization of beta8-heterodimers to the cell membrane(35, 36) . The membrane localization and the similarity of the chromatographic properties of baculovirus expressed Gbeta18 with other beta-heterodimers (13, 37) suggest that recombinant G8 was also isoprenylated.

Functional assay places Gbeta18 as intermediate between Gbeta12, brain, or placental beta and the retinal beta-heterodimer at stimulating transducin activation by rhodopsin(4, 37) . The 3-fold difference in activity of Gbeta18 and Gbeta12 is relatively large considering their sequence similarity particularly when compared with results from other in vitro studies of defined beta-heterodimers(13, 14, 38) . However, the specificity of beta-heterodimers that has been reported in vivo(5, 6) has still to be explained in view of the redundancy that is consistently observed in vitro.

In adult rats there is a very specific expression of G8 in the olfactory and vomeronasal epithelia, with a trace of mRNA in the olfactory bulb ( Fig. 4and Fig. 5). The low levels of G8 RNA in the olfactory bulb may be associated with small amounts of mRNA transported along the axon as has been observed for other neural mRNAs(39) . Of particular significance is the lack of expression of G8 in the nasal respiratory mucosa. The respiratory epithelium is continuous with the olfactory epithelium but distinct from the olfactory epithelium in that it does not contain sensory neurons. The absence of G8 mRNA in brain and other tissues studied indicates that G8 is more restricted to the olfactory system than the G-protein alpha-subunit Golfalpha(40) , the olfactory adenylate cylase(41) , or apparently some of the olfactory receptors(42) . Therefore, at least in the adult, the signal transduction pathway, in which G8 functions, appears to be olfactory-specific.

In situ hybridization was used to examine the localization of G8 expression in the olfactory and vomeronasal epithelia. Nonspecific hybridization to other related proteins might influence interpretation of results. Therefore, to minimize this potential problem, very high stringency wash conditions and RNaseA digestion were used following hybridization. The most similar nucleotide sequence to G8 is that of G2; their full coding sequences are only 62% identical (with no long regions of continuous identity). Comparison of the results of Northern analysis (Fig. 4) and RNase protection (Fig. 5) indicates that cross-hybridization of G8 probes to G2 (or other sequences) is unlikely, under stringent conditions. We were unable to detect hybridization of G8 probes in the olfactory bulb by in situ hybridization, indicating that the in situ technique does not, in itself, result in reduced specificity. Moreover, the localization of G8 expression that we detected in the olfactory epithelium (both with S- and digoxigenin-labeled probes) was discrete, only a subpopulation of the neural cells showed significant hybridization (see below). All non-neuronal cells were negative. This expression pattern was consistent with that found in the vomeronasal epithelium and also with the developmental and bulbectomy data (see below). Therefore, it is very likely that the in situ hybridization that we detect reflects the true expression of G8.

In the olfactory epithelium, G8 expression is localized to olfactory neural cells ( Fig. 7and Fig. 8). However, the expression is not evenly distributed throughout the neural cell layer of the epithelium, but is concentrated toward the base of this layer ( Fig. 7and Fig. 8). Olfactory neurons develop by differentiation of basal cells in the olfactory epithelium(24) . As neural development proceeds, there is a migration of the cell body of the neuronal precursors toward the epithelial surface. Therefore it appears that G8 is predominantly expressed in immature neurons. To study this in more detail and to investigate whether G8 has a more general role in neurogenesis, we examined the expression of G8 during the development of the olfactory neuroepithelium and the brain. The first olfactory neurons begin to appear at about E14 in rats(43) . However, the largest rise in the number of sensory cells occurs in the period shortly after birth(44) . In the rat brain considerable neurogenesis occurs over a similar time period. No expression of G8 above background was detected in the brain indicating that it is unlikely that G8 is a molecule essential to neurogenesis in general. However, in the olfactory epithelium the marked rise in the expression of G8 shortly post-natum parallels the high rate of neurogenesis shortly after birth. In adult rats, the number of mature neurons is higher but the rate of neurogenesis lower than at P13.5; a corresponding decrease in expression of G8 was observed. Golfalpha expression appears to lag behind that of G8. Also, in contrast to G8 expression, that of Golfalpha corresponds with the number of mature neurons; Golfalpha mRNA was also detectable in brain confirming previous reports(40) .

Further evidence that G8 is predominantly expressed in developing neurons comes from olfactory bulbectomy studies. After olfactory bulbectomy (which results in the degeneration of mature olfactory neurons (45) and the loss of mRNA species specifically expressed in these cells of the olfactory epithelium), there is only a small change in the level of G8 expression in comparison to that seen with other olfactory markers (Fig. 4). This means that the majority of the expression of G8 is not in mature olfactory neurons. However, the in situ hybridization studies are not consistent with expression of G8 in several other cell types that make up the epithelium: the perinuclear region and the majority of the cytoplasm of the sustentacular, basal, and glandular cells are devoid of G8 mRNA (Fig. 7). Nevertheless, after bulbectomy hybridization toward the base of the epithelium is still observed (Fig. 8). Thus most G8 mRNA is expressed early in the maturation of olfactory neurons before the markers of functional sensory neurons, e.g. OMP or Golfalpha are found.

In the adult vomeronasal epithelium, the level of expression of G8 relative to actin is at least as high as it is in the olfactory epithelium ( Fig. 5and Fig. 6). The preparation of the vomeronasal epithelium used for isolation of RNA was relatively crude, and it is likely that the actual purity of vomeronasal neuroepithelium used for RNA was substantially lower than that of the olfactory epithelium. This means that the level of G8 expression in the vomeronasal epithelium is as much as 2-5-fold that found in the adult olfactory epithelium. A similar difference in intensity of signal was also noted in in situ hybridization studies (Fig. 8). At low magnification it is clear that the highest level of G8 expression is localized to the regions at the boundaries between sensory and non-sensory epithelia. These areas of maximum G8 expression are the primary regions of neurogenesis in rodent vomeronasal epithelia(28, 29) . The rest of the epithelium is a pseudostratified epithelium similar in organization to, but much thicker than the olfactory epithelium. Here the distribution of G8 mRNA resembled that of the olfactory epithelium (Fig. 8, A and D). Therefore, both in the olfactory and in the vomeronasal epithelia, G8 appears to play a role in signal transduction during neural development.

The level of expression of G8 in adult rat olfactory epithelium is considerably lower (at least 10-fold) than that of Golfalpha ( Fig. 5and 6). Taken together with the localization of Golfalpha to fully developed neurons that are lost after olfactory bulbectomy (Fig. 4), and the absence of Golfalpha from the vomeronasal epithelium (Fig. 6), it does not appear that the majority of G8 and Golfalpha are associated in a specific heterotrimeric G-protein. We cannot rule out that some G8 does occur in such a protein, but it is much more likely that this G-protein subunit is associated with one or more of the other G-protein alpha-subunits that are expressed in the olfactory epithelium(46) . Thus it is unlikely that G8 is a factor that directly mediates coupling of olfactory receptors to cAMP production in olfactory cells. It will be important to determine whether G3 is the principal G-protein -subunit associated with Golfalpha, or whether other -subunits that were not amplified by PCR participate in olfactory signal transduction. It is also of considerable interest that we did not detect expression of Golfalpha in the vomeronasal epithelium ( Fig. 5and Fig. 6), whereas its expression in whole brain was clear. We are currently investigating whether other olfactory components involved in the cAMP signaling pathway are present in the vomeronasal epithelium.

Two proteins have been reported which appear to have a relatively similar distribution in the olfactory epithelium to G8. The first is a tyrosine phosphatase (NE-3) that also is found in the adult brain, and to a lower level, in several other tissues(47) . Like G-proteins, protein tyrosine phosphatases have been implicated in Drosophila development: both corkscrew a protein tyrosine phosphatase(48) , and concertina a G-protein subunit (16) are maternal genes required for normal embryonic development. Similarly, both G-proteins and protein tyrosine phosphatases are involved in Dictyostelium development(15, 49) . The other protein with a distribution in the olfactory epithelium similar to G8 is an olfactory-specific transcription factor Olf1, which appears to be expressed both in developing and mature olfactory neurons(50) . An alternatively spliced form of this protein appears to function as a transcription factor in pre-B-cells(51) . The expression of Olf1 is highest in rat olfactory epithelium at times shortly after birth and declines in the adult (52) in a manner similar to that seen for G8. It is interesting to note that the full-length clone of G8 isolated from the cDNA library (but not the 5`-RACE product) contains a strong consensus sequence for Olf1 binding. This sequence binds Olf1 in vitro, (^3)and may play a role in directing the expression of G8 to the olfactory tract.

In summary, we have characterized the expression of a novel G-protein -subunit that appears to be expressed early in the maturation of olfactory and vomeronasal neurons. This protein, that is clearly involved in signal transduction, is most similar to G2 in structure, but heterodimers of Gbeta1 and these two different Gs have quantitatively distinct functional properties. G8 is expressed in a fashion that suggests it plays a role in olfactory neurogenesis. Neurogenesis involves considerable changes in gene expression, extension of a dendrite to the surface of the epithelium, and an axon to the olfactory bulb. Control of all these processes, essential for development and maintenance of a functional olfactory system, must require controlled response to many signals. Thus the spatial and temporal pattern of expression of G8 suggest that this G-protein subunit is required for olfactory and vomeronasal neurons to respond to signals that modulate their growth and/or development.

FOOTNOTES

*
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank(TM)/EMBL Data Bank with accession number(s) L35921[GenBank].

§
To whom correspondence should be addressed: Bldg. 10, Rm. 1A09, NIDR, NIH, Bethesda, MD 20892. Tel.: 301-402-2401; Fax: 301-480-8328.