The Olfactory Signaling Molecules Biology Essay

Abstract

Background

Olfactory signaling molecules (such as ORs, Gaolf and AC3) have been reported outside of the olfactory system sporadically. However, little is known about the systematical analysis of olfactory components throughout many parts of the body.

Principal Findings

Olfactory sense is mediated by specialized olfactory receptor neurons (ORNs) in the nose. Olfactory signaling molecules, such as G protein a subunit (Gaolf) and Adenylyl Cyclase 3 (AC3), have recently been reported to be found outside of the olfactory system, suggesting that the olfactory sense may play a role in other tissues. However, ectopic expressions of olfactory signaling molecules (Gaolf, AC3, OMP, ORs) and their functional roles still remain to be elucidated. This study demonstrates the presence of olfactory signaling molecules in other tissues by systemically using a RT-PCR, double-antibody immunoprecipitation-immunodetection analysis, and western blotting. Immunohistochemical analysis was utilized to find OMP positive cells in non-olfactory tissues. Unexpectedly, the result showed that OMP positive cells were found in some tissues. Gene expression of olfactory receptors was also observed in OMP positive tissue and expressed in several tissues in common. More importantly OR protein expressed in OMP positive cell of thymus.

Conclusions

To investigate olfactory-like signaling molecules in non-olfactory tissues, we used OMP antibody, and selected several OMP positive tissues. And we observed OR gene expression in OMP positive tissues, moreover OR protein expression with OMP in same cell of tissue. These results indicated that the apporoach for ectopic expression of olfactory signaling molecules in non-olfactory tissues is more effective than that of ectopic expression of OR and many ORs and olfactory signaling molecules (OMP, Golf, and AC3) are expressed in various non-olfactory tissues extensively. Finally, these observations suggested that the chemical sensing by those molecules is operated in non-chemosensory tissues.

Introduction

Animals have evolved chemosensory systems that are able to recognize and discriminate thousands of environmental stimuli. Detection and discrimination of various odorants by olfactory sensory neurons (OSNs) critically depends on a large family of G protein-coupled olfactory receptors (ORs). Biopolar OSNs possess a thin axon and a thick dendrite with dendrite knob, including 10-20 cilia, which contain ORs and several signaling molecules. ORs are members of the superfamily of G protein-coupled receptors (GPCR) expressed in the olfactory epithelium, where they detect small and volatile molecules. ORs are the largest receptor family in mammals comprising 3%–5% of all genes [1] and Buck and Axel first identified mammalian ORs two decades ago [2]. In the mouse genome, there are ~1000 genes encoding different types of odorant receptors [2-4]. In OSNs, olfactory signal transduction begins with the activation of OR in the ciliary membrane. Odorants binds to ORs, which activate olfactory specific G protein (Gaolf) and this leads to the activation of adenylyl cyclase type 3 (AC3) [5]. The cyclase catalyzes the production of cAMP that causes opening a cyclic nucleotide–gated (CNG) channels [6,7]. This nonselective cation channel leads to an increase in intracellular Ca2+ concentration [8], which depolarize the cell membrane. Increased Ca2+ in OSN induces a depolarizing Cl− current, which amplify the sensory responses for membrane depolarization [9-11]. Ca2+ also involves termination and adaptation of sensory response by negatively regulating the activities of several signaling components, which leads to reduced sensitivity to repeated odor exposure [12]. Ca2+ accumulation and removal is very important to not only the olfactory sensitivity but also the rates of activation, termination, and adaptation of the olfactory signaling pathway. Thus, proper Ca2+ regulation should be critical for sensing olfactory stimuli. Because olfactory cilia do not contain internal organelles [13], the extracellular milieu is the only source and sink for Ca2+ influx and efflux. Therefore, in the ciliary membrane, Ca2+ homeostasis is believed to be achieved by plasma membrane Ca2+ transporters, including an Na+⁄Ca2+ exchanger (NCX), a potassium-dependent Na+⁄Ca2+ exchanger (NCKX4) [14-16] and a plasma membrane Ca2+-ATPase (PMCA) [17-19]. During this Ca2+ clearance, olfactory marker protein (OMP) regulates NCX activity and allows rapid Ca2+ extrusion from OSN knobs through interaction with Bex protein and calmodulin [15]. OMP is abundant, small, conserved protein [20] expressed exclusively mature olfactory neurons in the main olfactory epithelium, the vomeronasal organ, the septal organ and the Gruenberg ganglion [21-26]. Although OMP is widely used as a marker for OSNs, its function has remained largely elusive. Some results suggested that in OMP knockout mice, electrophysiological responses to odor molecules are impaired [27,28] and detection threshold for odors is increased by 50–100-fold [29,30], demonstrating an ability to respond to odor stimuli and a modulatory function in olfactory signal transduction. In addition, the OMP−/− mice shows the deficits in odor detection threshold and odor quality discrimination, implicating a novel mechanism for fast and robust response termination to ensure the temporal resolution of the odor stimulus [31].

Among several molecules of the olfactory signaling pathway, OMP, Golf, AC3, and ORs have known to be olfactory specific molecules. Microarray analyses suggest olfactory signaling molecules including ORs are expressed in many non-olfactory tissues [32-37], however, the hypothesis that ectopically expressed olfactory signaling molecules carry additional functions in non-olfactory tissues remains largely unsupported. The importance of the ectopic expressions of olfactory specific molecules is raised since the physiological function of the sperm OR was characterized [38], but still unknown in other cases. Ectopic expression of ORs has been widely founded in non-olfactory tissues. Testis is first case isolated olfactory-like proteins among non-chemosensory tissues [39], and characterized their physiological function, chemotaxis [38]. And, during myogenesis and muscle regeneration, changes of multiple ORs’ expression were observed and their ligand concentration affected myocytes migration through canonical olfactory signal transduction pathway [40]. Besides testis and muscle, several OR gene expression was reported in tongue, heart, spleen, pancreas, brain, prostate, muscle, blood, embryo, liver, lung, kidney, GI tract, and placenta without knowledge of their physiological function (review). Some studies suggest the existence and function of olfactory signaling molecules in some non-olfactory tissues. OMP is detected in taste buds of circumvallate papillae, indicating perception of taste stimuli [41]. And Golf is identified in testis, spleen, lung, heart, placenta, kidney [42-47], and also pancreatic islet [48], suggesting relation with insulin status, cell survival, and regeneration of the insulin-secreting beta-cells during development and diabetes [49]. In the kidney, the presence of the olfactory signaling molecule (OR/Golf/AC3) is identified, indicating the regulation of renin secretion and glomerular filtration rate [47]. In addition, ORs are present in human enterochromaffin cells, suggesting regulation of serotonin release [50]. Furthermore, a number of ORs are also expressed by regenerating muscle, which may suggest a larger role for ORs in tissue repair [40].

As mentioned above, olfactory signaling molecules (such as ORs, Golf and AC3) have been reported outside of the olfactory system, suggesting that the olfactory sense may play a role in the non-olfactory tissues. Little is known about the systematical analysis of olfactory components throughout the whole body since many studies focused on ectopic OR expression. The number of the OR genes is greater than 1,000 in the mouse, and the homology between ORs is 40-90%. For this reason, there are not enough high quality specific OR antibody and appropriate ligands of specific ORs. Therefore in the present study, we selected olfactory signaling molecules (OMP, AC3, and Golf) as targets to investigate ectopic olfactory sensing, and we hypothesized that it will be a substitute to figure out the ectopic OR expression in non-chemosensory tissues. This study demonstrates that olfactory signaling molecules are present in the non-olfactory tissues by systemically using microarray database, RT-PCR, western blot, and immunohistochemical analysis. The gene expressions of OMP, Golf, AC3 were detected in several tissues using RT-PCR and western blot analysis. And using immunohistochemical analysis using OMP antibody, olfactory signaling molecules are present only in the restricted region in some tissues, which is unexpected. And we observed that olfactory receptor was co-expressed with OMP in the same cell. These results suggest that olfactory signaling molecules are appeared in various tissues extensively, and the chemical sensing through the molecules may be operated in non-olfactory tissues.

Results

Since OE specific OR expression was found about 20 years ago by Buck and Axel [2], some of the OR genes have been found in the non-olfactory tissue and the research to finger out their function in other tissues have continued until now. Although there are reports about ectopic expression of the ORs through the microarray analysis, except for a few cases, almost cases are still unknown in the function of the ORs. This is result from the lack of available specific OR antibody due to similarity between the ORs and the small number of appropriate ligand corresponds to specific OR for the functional study and the problem about expression of OR in heterologous system. These factors act to as limited factors to study ectopic expression of ORs and their functions. As mentioned previous paper [51], to overcome limitation of the OR as a target to search olfactory sensing in non-olfactory tissues, we try to find olfactory sensing in non-olfactory tissues using olfactory signaling molecules including OMP, authentic protein in olfactory.

Identification of tissues in refined microarrays

In the first step, we examined the GeneAtlas2 microarrays for our systematic re-analysis of the expression of mouse olfactory specific genes, and we found the gene expression of OMP, Golf, and AC3 is noticeable in several tissues. OMP is noticeable in skeletal muscle, pancreas, spleen, and so on, Golf noticeable in pancreas, spleen, thyroid, and so on, and AC3 noticeable in thymus, stomach, heart, and so on. We selected 13 tissues as an interesting group of tissues for further analysis.

OMP, AC3, and Gaolf are expressed in various non-olfactory tissues.

To investigate expression of olfactory signaling molecules in selected 13 tissues, we tested the mRNA levels of olfactory signaling molecules in non-olfactory tissues. Surprisingly, RT-PCR analysis indicated that OMP is expressed in almost every tissue, although the amount of expression was different from each other (Fig 1A). The result show that the expression level of OMP is very strong in OE, and is very weak in pancreas and testis. And OMP expression in bladder, stomach, duodenum, spleen, thymus, and thyroid was moderately lower than that in OE (Fig 1A and B). In the case of AC3, its expression level is showed nearly similar amount in almost tissues except OE. In contrast, expression pattern of Golf is more restrict than OMP. Golf signal is rarely observed in liver, pancreas, spleen, and kidney, and is weak in skeletal muscle, bladder, stomach, duodenum, thyroid, and lung. These results show that mRNA of olfactory signaling molecules was expressed in almost tissues, extensively. With this, we tested protein expression of olfactory signaling molecules. First, we tested OMP protein expression using western blot, but we could not observe OMP specific band in other tissues except OE. Because other tissues can be expected to be very low expression of OMP, tissue extracts were prepared for immunoprecipitation. The 20mg of tissue extract (100ug of OE and 3mg of thyroid) were used for immunoprecipitation of OMP using goat anti-OMP IgG, and then performed immunodetection using rabbit anti-OMP IgG (from Prof. Margolis at Maryland Univ.), The result, amazingly, showed OMP signal in almost tissues except kidney (Fig 2A). Although the large amount of tissues extract was used in the experiment, this result indicate that OMP protein, which known as olfactory marker protein and authentic protein in olfactory system expressed in non-olfactory tissues. Protein expression of AC3 is observed in almost tissues similar to RT-PCR result, and is confirmed peptide blocking experiment in a few tissues (Fig 2B), representatively. Golf is typically detected as a band 42-47kDa in OE and, according to previous reports, Golf protein was founded in mouse and rat kidney [47], striatum [52], rat placenta [46]. Our result showed Golf protein expression in skeletal muscle, pancreas, duodenum, testis, and heart (Fig 2C). Especially, very strong band of Golf in skeletal muscle and heart was observed and demonstrated using blocking peptide. These result suggest that olfactory signaling molecules are expressed in several tissues through RT-PCR and western blot analysis, systematically.

olfactory signaling molecules are present only in the restricted region in some tissues.

The RT-PCR and western blot analysis showed expression of olfactory signaling molecule in non-olfactory tissue. Especially, OMP expression was detected in non-olfactory tissues, extensively, although the amount of expression is relatively very low. With these results, we speculated about two possibilities for the OMP expression in several tissues. One is that OMP expression is very low in entire tissue and the other is that OMP expression is restricted in particular region or cell in the tissue. To verify this possibility, histological approach was performed using antibody for OMP known as the authentic protein in olfactory system. The result demonstrates that OMP signal was observed in restricted cell in a few tissues (Fig 4). OMP signal was detected in bladder, testis, stomach, thymus, skeletal muscle, thyroid, lung, and heart, and this signal is limited to particular cell. In addition, we checked whether AC3 and Golf are co-localized with OMP. The result shows three types of expression pattern. First, AC3 and Golf expression were not observed in OMP positive cells of testis and thymus. Second, OMP, AC3, and Golf, all are co-localized in one type cell in heart and skeletal muscle. Third, OMP is co-expressed with AC3 or Golf in lung, stomach, thyroid, and bladder. This result is inconsistent with our hypothesis, which is AC3 and Golf may be co-expressed with OMP in the same cells. However, expression pattern of AC3 and Golf was varied in OMP positive cells, suggesting that, for the olfactory sensing, OMP positive cell in non-olfactory tissues may not used AC3 and Golf known as canonical signaling molecules in olfactory system but used specialized subtypes of adenylyl cyclase or G protein in own cell.

Identification of candidate OR genes for six tissues

OR is starting protein in olfactory signaling pathway and whether or not OR is expressed in cell is very important issue for searching chemical sensing in non-olfactory tissues. Because of the problems to target the OR, we used OMP antibody, and selected OMP positive tissues, and then investigated OR expression in the selected tissues. To selected OR in each OMP positive tissue, we systematically re-analyses the expression of mouse OR genes using the GeneAtlas2 microarrays database, and we identified only valid ORs from the microarray data. As a result, we found 491 valid ORs. We found a small number (e.g., 15) of ORs from validated ORs were relatively highly expressed in six tissues we selected using IHC analysis. Those set of ORs could be regarded as good candidates for the next step, that is, observing OR expression.

Olfactory receptors are expressed in OMP positive tissues

In order to determine experimentally 15 OR (Table ) selected through array data in each tissue, RT-PCR was performed. The result shows the expression of about 2-3 OR in each tissue and a large number of OR expression was observed in testis (Fig 4). A few of the ORs (olfr181 and olfr1386) commonly expressed in various tissues and the others expressed in tissue-specific. These RT-PCR products were sequenced and found that the OR sequences in non-olfactory tissues was indistinguishable from already published sequences from the OE. Furthermore, these results suggest that the approach to set olfactory signaling molecules rather than OR as a target is working properly to figure out ectopic olfactory sensing.

Olfactory receptors is co-localized with OMP

These results so far confirm that OR actually expressed in OMP positive tissues. Lastly, we carried out the experiment to prove that particular OR expressed with OMP in the same cell of each tissue. At this time, we are faced with a big problem. Sufficient high quality OR-specific antibodies to identify particular OR do not exist because the number of ORs is more than 1,000 in the mouse and the homology between ORs is 40-90%. This has been acting with a great difficulty to study ORs and acts as a limiting factor in this study. Therefore, we tested OR expression using commercial antibody in OMP positive tissues owing to impossibility to confirm the expression of OR corresponding to RT-PCR result. The result showed that olfr1386, olfr1496, olfr544 expressed in OMP positive cells of thymus, suggesting the possibility for sensing extracellular chemical through OR in non-olfactory tissues.

Discussion

In present study, we reported OMP, AC3, Golf, OR expression in non-olfactory tissues. And we suggest that olfactory-like signal pathway may function in non-olfactory tissues through OR activaton. Through previous studies, ectopic expression of OR has been reported in various tissues (review 참고). However, for that function, besides muscle [40], testis [38], enterochromaffin cell [50], and prostate cancer [53], there is no progress to speak of. Because the study about ectopic expression and function of ORs as a target is difficult the limiting factors. In this study, to overcome the limiting factor of experiments utilizing the OR as a target, we set the OMP with olfactory signaling molecules as targets, and then carried out screening about ectopic olfactory sensing in non-olfactory tissues. The distribution of olfactory signaling molecules in many non-olfactory tissues was confirmed by the results of RT-PCR and WB. By this time, OMP expression was reported in only hypothalamus [54] and taste buds of circumvallate papillae [41] except olfactory system. However, amazingly RT-PCR result showed that OMP mRNA detected in many tissues and immunoprecipitation-immunodetection method showed that OMP protein expressed in several tissues although a large amount of tissue extracts were used. It has more meaning that olfactory marker protein is expressed in other tissues. Because, it can enhance the possibility that OMP expressing tissues can be expressed OR. On the other hands, AC3 mRNA appeared in entire tissues and protein also expressed in all tested tissues. AC3 expression has already been reported in various other studies. AC3 mRNA level of testis [55,56], kidney [47], rat and mouse ovary [57] have been identified, AC3 expression has been reported in spontaneously diabetic Goto-Kakizaki (GK) rat pancreatic cell [58]. In addition, through northern blot, AC3 identified in brain, spleen, lung, kidney, testis, and skeletal muscle [59]. Also, AC3 was confirmed in thyroid using Northern blotting, Western blotting and RT-PCR experiments [60]. In particular, in rat and mouse kidney, 55kDa and 90kDa bands of AC3 was detected and glomerular filtration rate (GER) was reduced in AC3-/- mouse, suggesting that olfactory signal molecules (AC3 and Golf) affected renal function in the kidney [47]. The reports about Golf expression was insufficient than that of AC3. Originally, Golf was discovered in olfactory epithelium [43] and brain striatum [52,61,62], after that, Golf mRNA was detected in human insulinoma cell line, testis, retina, liver [45], HEK293T cell, SV40-immortalized human collecting duct (HCD) cell line, human prostatic cancer cells (LNCaP) [63], rat brain [61] and protein expression was identified in retina, lung, spleen [45], mouse and rat kidney [47], and rat placenta [46]. In addition, Golf expression was confirmed in rat heart using Northern blot and western blot and was restricted in cardiomyocytes through IHC [42]. On the other hands, in this study, Golf expression of RT-PCR was most strong in the OE and low expression in other tissues. However, surprisingly, the result from western blot showed that Golf expression was higher in skeletal muscle and heart than OE. In addition, 40~45kDa of Golf band was detected in pancreas, duodenum, testis and this band was completely blocked by its antigenic peptide, suggesting that the band is Golf specific band. As such, Golf expression has already been reported was not matched in this experiment. This is seems to the differences between experimental conditions. We used the same amount of tissue extracts under the same condition instead of under the optimal condition in a particular tissue. As identified in the above results, mRNA and protein of OMP, AC3, Golf were observed in non-olfactory tissues, widely. This result raises the possibility that the olfactory-like signaling pathway can operate in each tissue. To investigate which cell population used olfactory-like signaling pathway, we screened each tissue using OMP antibody. OR expression is most important starting point for activation of olfactory-like signaling pathway. But, there are many problem as mentioned above, we used OMP antibody to search the tissue operating olfactory-like signaling pathway as a new alternative instead of OR. The result shows OMP positive cells in testis, heart, stomach, bladder, thymus, lung, skeletal muscle and thyroid. Until now, OMP positive cells have not been identified except in OE, brain, and taste bud. However, surprisingly, IHC result shows existence of unidentified OMP positive cells in non-olfactory tissues. In addition, the expression of AC3 and Golf, canonical olfactory signal transduction molecule was determined with OMP in the tissues. As a result, there are cells that are expressed OMP, AC3, and Golf all together in some tissues and that are expressed OMP/AC3 or OMP/Golf and that is expressed only OMP without AC3 and Golf. This result is different from those expected in early experiment. At first, we hypothesized that AC3 and Golf were co-localized with OMP in particular cell and they are used signal transduction molecules like olfactory system, but, IHC result indicates that this expectation digressed. This data provides that OMP positive cells in non-olfactory tissues may be not used AC3 and Golf as signaling molecules like olfactory system but specialized, different type of adenylyl cyclase and G proteins for the chemical sensing. Above all, whether OR is expressed in OMP positive tissues is critical. Because the detection of OR expression in OMP positive tissue is correspond to initial purpose of this study to find olfactory sensing in non-olfactory tissues using olfactory signaling molecules as a target to find ectopic OR expression. To find OR expression in OMP positive tissue, we selected 15 OR list in each tissue using array database and performed RT-PCR to confirm OR expression in that OMP positive tissues. The RT-PCR results show that OR mRNA actually expressed in OMP positive tissue. This finding suggests that this alternative method that approach to find expression of olfactory signaling molecules in non-olfactory tissues may be more efficient rather than that to find ectopic expression of ORs. Finally, we should determine whether OR is expressed with OMP in same cell, where we reached the limit that the antibody to determine the OR protein expression is not enough. In previous studies, ISH method was used to identify OR expression in the tissues because of lack of available, high quality OR specific antibody. Therefore, in this experiment, we used commercially available antibodies to determine OR expression with OMP. The results show that olfr1386, 544, 1496 were expressed with OMP in same cells of thymus. In other words, this means that OMP expression may be replaced OR expression. Eventually, this result suggested that the approach to find olfactory signaling molecule instead of OR for ectopic olfactory sensing is effective and reasonable. Also, ectopic OMP expression can be a counterplan to escape limitation of ectopic OR expression and a proper approach to search tissues operating ectopic olfactory signaling pathway, systematically.

In conclusion, present study demonstrate that mRNA and protein of OMP, AC3, Golf were expressed in non-olfactory tissues, extensively, and OMP was expressed in particular cell population in some tissues with OR. Our data suggest that OR expression can be seen indirectly through OMP expression and chemical sensing through OR operated in several tissues as well as olfactory system, suggesting operation olfactory sensing in several tissues. In future study, it will be important to identify OMP and OR positive cell to better understand what activates these signaling and what physiological functions they support.

MATERIALS AND METHODS

Identification of tissues in refined microarrays

To find target tissues that are expected to express olfactory signal molecules, the GeneAtlas2 microarrays database was used [33,34]. We found the gene expression of OMP, Golf, and AC3 is noticeable in several tissues of the above refined microarrays. OMP is noticeable in skeletal muscle, pancreas, spleen, and so on, Golf noticeable in pancreas, spleen, thyroid, and so on, and AC3 noticeable in thymus, stomach, heart, and so on. We selected 13 tissues as an interesting group of tissues for further analysis. For those 13 candidate tissues, we performed the extensive biological experiments such as RT-PCR, western blot, and immunohistochemistry.

Isolation of Total RNA and RT–PCR

Total RNA of C57BL/6 (male, 7 week) mouse tissues were isolated by using MagNA lyser (Roche Molecular Diagnostics GmbH, Penzburg, Germany) with Trizol reagent (Invitrogen, Carlsebad, CA, USA). Briefly, tissue was dissolved in 1 ml of Trizol, and 0.2 ml of chloroform (Sigma Aldrich, St Louis, MO) was added. Samples were mixed thoroughly and centrifuged at 13,000 rpm for 15 min. The upper aqueous phase was transferred, and the RNA was precipitated by the addition of 0.45 ml of isopropanol (Sigma Aldrich, St Louis, MO) followed by centrifugation at 13,000 rpm for 10 min. The precipitate was washed with ice cold 70% ethanol (Sigma Aldrich, St Louis, MO), and the final pellet was resuspended in RNAse free water. Total RNA (2 ug) was reverse transcribed using cDNA synthesis kit (Takara Shuzo, Otsu, Japan) (In the case of (-) RT, 1 ul of water was added instead of 1 ul PrimeScript RTase). The PCR reaction mixture consisted of first-strand cDNA template, PCR master mix (Takara Shuzo, Otsu, Japan) and primer sets (Table ). Each PCR cycle consisted of denaturation at 94°C for 1 min, annealing at 55°C or 60.4°C (b-actin) for 1 min, and elongation at 72°C for 1 min. 35 cycles were performed with a T3000 thermocycler (Biometra, Göttingen, Germany).

Western blotting

Total tissue extracts (70ug/lane for AC3 or 100ug/lane for Golf) was isolated from C57BL/6 mouse tissues by using MagNA lyser with T-PER buffer (Pierce Biotechnology, Rockford, IL, USA). The protein concentrations were determined by the Bradford procedure (Bio-Rad, Laboratories, Hercules, CA, USA). The tissue extracts were separated in 7% or 10% SDS–polyacrylamide gel electrophoresis (PAGE) and transferred to nitrocellulose (NC) membrane (Whatman, Maidstone, Kent, UK) for western blot analysis. The nitrocellulose membrane was rinsed in TBST (25 mM Tris–HCl, 2.7 mM KCl, 137 mM NaCl, pH 7.4, and 0.1% Tween-20) and blocked for 1 h in 5% non-fat dry milk in TBST. The membrane was incubated overnight at 4°C with using the rabbit anti-AC3 antibody (1:500, SantaCruz Biotechnology, Santa Cruz, CA, USA) or rabbit anti-Gaolf antibody (1:2,000, SantaCruz Biotechnology), followed by a secondary HRP conjugated donkey anti-rabbit IgG (1:100,000, Jackson Immuno-Research, Westgrove, PA, USA). In control experiments, antibodies were preincubated with an excess of competing peptide. Immunoreactive proteins were detected using the ECL chemiluminescent detection system according to the manufacturer’s directions (Supersignal-West Pico, Pierce, Rockford, IL, USA)

Immunoprecipitation

Total tissue extracts (20mg except Thyroid-3mg and OE-100ug) of each tissue were incubated with goat anti-OMP IgG (Wako, Richmond, VA, USA) in a final volume of 500 ul at 4°C for 2 h and protein G sepharoseTM 4 fast flow bead (20 ul, GE healthcare, Uppsala, Sweden) were added and incubated at 4°C overnight with gentle mixing. The beads were washed 5 times with lysis buffer and the drained beads were then incubated at 95°C in SDS sample buffer and electrophoresed on 10–20% SDS–PAGE and blotted to nitrocellulose as described above. Rabbit anti-OMP antibodies were used at a dilution of 1: 20,000 for western analyses, respectively. Bands were visualized via chemiluminescence as above.

Immunohistochemistry

Mice were anesthetized with ? mg/kg zoletil and perfused transcardially with 20 ml of ice cold PBS followed by 30 ml of 4% freshly prepared phosphate buffered paraformaldehyde (Sigma Aldrich, St Louis, MO). Tissue was dissected and post-fixed for 2 h in cold fixative and cryoprotected overnight in 30% sucrose at 4°C. Tissues were embedded in OCT compound (Tissue Tek, Sakura Finetek, Torrance, CA, USA) and snap-frozen in a dry ice. Coronal cryostat sections of each tissue were attached to Superfrost-plus microscope slides (Matsunami, Tokyo, Japan), dried at 37°C for 30 min and stored at -80°C until needed. Tissue slides were treated with 4% normal horse serum (NHS, Jackson Immuno-Research, Westgrove, PA, USA) and 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 1 h to block nonspecific antibody binding in a humidity chamber. The slides were incubated overnight at 4°C with using the goat anti-OMP antibody (1:10,000), rabbit anti-AC3 antibody (1:10), and rabbit anti-Golf antibody (1:10) antibodies diluted with PBS containing 4% NHS and 0.1% Triton X-100. For enzymatic detection of the bound primary antibodies, Cy3-conjugated anti-goat IgG (1:2,000) and Dylight488-conjugated anti-rabbit IgG (1:1,000, Jackson Immuno-Research, Westgrove, PA, USA) were used in PBS containing 0.1% Triton X-100 for 1 hour at RT in the dark. After rinsing, the slides were coverslipped with VECTASHIELD containing DAPI (Vector Laboratories, Burlingame, CA, USA) and immunohistochemical images were taken on a confocal laser scanning biological microscope LSM700 (Zeiss, Oberkochen , USA).

Identification of valid OR genes in GeneAtlas2 microarrays

We examined the GeneAtlas2 microarrays for our systematic re-analysis of the expression of mouse OR genes. As the first step of the analysis, we identified only valid ORs from the microarray data. Even though a large number of ORs of from Olfr1 to Olfr1570 are mentioned in several studies [32,64], only a subset of them is known as "valid" ORs. We performed the following steps for each number from 1 to 1,570, and identified a set of valid ORs.

Search "Olfr<number>" in the Nucleotide database of the GenBank site (http://www.ncbi.nlm.nih.gov/genbank).

In the search result, click the link where its accession number starting with "NM_", which means the nucleotides of mRNA.

Identify the current OR as the valid one if the comment section of the result page starts with VALIDATED, and at the same time, the title of the result page does not include "partial" or "pseudo".

Identification of candidate OR genes for six tissues

We found some a small number (e.g., 15) of ORs (Table 3) from 352 valid ORs were relatively highly expressed in six tissues we selected above. Those set of ORs could be regarded as good candidates for the next step, that is, observing OR signals through biological experiments like RT-PCR. We could easily make 352 valid ORs ordered by their intensity values for each tissue. However, it might not be reasonable to sort ORs only by the intensity values because microarrays are well-known to be "noisy" [65], and furthermore, there are another raking factor that we should consider. For example, if a certain OR shows the highest intensity in heart among all tissues, it is highly likely that the OR is really expressed in heart even though the OR has a relatively low intensity compared to the intensities of other ORs for heart. Here, we propose a new ranking method that uses a combination of two kinds of ranking : one is ranking in a tissue, and the other is ranking in an OR.

Let M be the refined microarray of m ORs and n tissues (e.g., m = 352 and n = 78 in the refined microarray), and Mij the specific expression of the i OR and the j tissue where 1 ≤ i ≤ m, and 1 ≤ j ≤ n. Let also p(Mj, Mij) be the ranking position of Mij in the set Mj = , and similarly, p(Mi, Mij) be the ranking position of Mij in the set Mi = . Then, we define the total ranking function tj(i) for a given OR i and a given tissue j as follows :

tj(i) = p(qj, p(Mj, Mij) + p(Mi, Mij))

where qj = {p(Mj, M1j)+p(M1, M1j), p(Mj, M2j)+p(M2, M2j), …, p(Mj, Mmj)+p(Mm, Mmj)}.

For example, olfr1339 has the 37-th largest intensity out of 352 intensities for lung, but it has the first largest intensity out of 78 intensities for olfr1339. The sum of two ranks is 38, which is the fifth smallest rank value (i.e., highest rank) out of 352 rank values for lung, and so tlung(olfr1339) is 5. We evaluated the rank positions of 352 valid ORs for each of the six tissues and selected top-15 ORs for each tissue, which is summarized in Table 2.

In order to check OR signals for the six tissues via fewer number of RT-PCR experiments, it would be desirable to use a single list of top-15 ORs for all six tissues instead of the six lists. For evaluating such a single list, we adopt the framework of merging multiple ranked lists into a single ranked list and selecting top-k from the single list, which is a widely used framework in database area [66-68]. We define the ranking function tsix(i) for a given OR i and for the six tissues as follows :

tsix(i) = p(qsix, )

where sixts = {heart, bladder, stomach, lung, testis, thymus}, and

qsix = {, , …, }.

According to tsix(i), we re-evaluated the rank positions of 352 valid ORs for the all six tissues and selected top-15 ORs, which is summarized in the farthest column to the right in Table 3.

Table 3. Analysis of OR genes for six tissues.

Tissue

Rank

Bladder

Heart

Lung

Stomach

Testis

Thymus

skeletal muscle

thyroid

1

Olfr1339

Olfr521

Olfr1411

Olfr1039

Olfr112

Olfr521

Olfr325

Olfr1196

2

Olfr1133

Olfr1339

Olfr250

Olfr1411

Olfr1133

Olfr39

Olfr883

Olfr190

3

Olfr325

Olfr190

Olfr393

Olfr1206

Olfr181

Olfr1339

Olfr190

Olfr325

4

Olfr521

Olfr883

Olfr521

Olfr883

Olfr190

Olfr1217

Olfr875

Olfr693

5

Olfr190

Olfr66

Olfr1339

Olfr39

Olfr1126

Olfr874

Olfr1145

Olfr1217

6

Olfr1145

Olfr1219

Olfr143

Olfr631

Olfr1145

Olfr1219

Olfr181

Olfr181

7

Olfr1028

Olfr1518

Olfr978

Olfr1196

Olfr883

Olfr190

Olfr1028

Olfr883

8

Olfr250

Olfr181

Olfr325

Olfr1145

Olfr1042

Olfr250

Olfr1168

Olfr1028

9

Olfr66

Olfr1133

Olfr92

Olfr1386

Olfr325

Olfr378

Olfr1196

Olfr1219

10

Olfr1219

Olfr325

Olfr1217

Olfr325

Olfr393

Olfr1028

Olfr1217

Olfr1145

11

Olfr968

Olfr1145

Olfr39

Olfr66

Olfr66

Olfr1042

Olfr291

Olfr1042

12

Olfr1168

Olfr1168

Olfr883

Olfr1143

Olfr968

Olfr1386

Olfr44

Olfr66

13

Olfr181

Olfr1126

Olfr1386

Olfr190

Olfr1028

Olfr1145

Olfr1042

Olfr393

14

Olfr44

Olfr291

Olfr1219

Olfr181

Olfr1386

Olfr1270

Olfr1143

Olfr1133

15

Olfr1196

Olfr1392

Olfr1039

Olfr968

Olfr44

Olfr1411

Olfr66

Olfr288

Distribution analysis of the selected OR genes

Before RT-PCR experiments, we verified the selected ORs were sufficiently differentially expressed from non-selected ORs. Since there are unrelated groups of ORs, we performed unpaired t-tests, especially unequal-variance t-tests. The sizes of the two groups are n1 = 25 for selected ORs and n2 = 329 for non-selected ORs. The p-values of each t-test for each tissue are 4 summarized in Table 4. Since all p-values are sufficiently small, we could reject the null hypotheses and say that it is highly likely that the signals of some of selected ORs would be observed in RT-PCR experiments.

Table 4. p-Value of unpaired t-test for six tissues (n1=25, n2=329).

Tissue

p-value

Bladder

5.6873E-04

Heart

2.3284E-05

Lung

6.6385E-05

Stomach

3.0081E-05

Testis

1.9455E-05

Thymus

2.0151E-05

Table 1. Primers and Annealing Temperatures Used for RT-PCR Experiments

Primer

Sequence

Product Size (bp)

Ta (°C)

Ref

OMP (sense)

5'-AAGCTGCAGTTCGATCACTG-3'

682

55

this study

OMP (antisense)

5'-TGTTCCTGTCCAGTCTCAGTCT-3'

 

 

[27]

AC3 (sense)

5'-TTGGCAGGCTTTCTTTGTCT-3'

462

55

[47]

AC3 (antisense)

5'-TCTGCAAACAGGATGCTGAC-3'

 

 

Golf (sense)

5'-TACCAGCTGATCGACTGTGC-3'

682

55

Golf (antisense)

5'-TGGCATACTCCGGGAAATAG-3'

 

 

Table . Primers and Annealing Temperatures Used for OR RT-PCR Experiments

Primer

Sequence

Product Size (bp)

Ta (°C)

Olfr1339-S

55

 

Olfr1339-AS

Olfr1133-S

Olfr1133-AS

Olfr325-S

Olfr325-AS

Olfr521-S

Olfr521-AS

Olfr190-S

Olfr190-AS

Olfr1145-S

Olfr1145-AS

Olfr1028-S

Olfr1028-AS

Olfr250-s

Olfr250-AS

Olfr66-S

Olfr66-AS

Olfr1219-S

Olfr1219-AS

Olfr968-S

Olfr968-AS

Olfr1168-S

Olfr1168-AS

Olfr181-S

Olfr181-AS

Olfr44-S

Olfr44-AS

Olfr1196-S

Olfr1196-AS

Olfr883-S

Olfr883-AS

Olfr1518-S

Olfr1518-AS

Olfr1126-S

Olfr1126-AS

Olfr291-S

Olfr291-AS

Olfr1392-S

Olfr1392-AS

Olfr1411-S

Olfr1411-AS

Olfr393-S

Olfr393-AS

Olfr143-S

Olfr143-AS

Olfr978-S

Olfr978-AS

Olfr92-S

Olfr92-AS

Olfr1217-S

Olfr1217-AS

Olfr39-S

Olfr39-AS

Olfr1386-S

Olfr1386-AS

Olfr1039-S

Olfr1039-AS

Olfr1206-S

Olfr1206-AS

Olfr631-S

Olfr631-AS

Olfr1143-S

Olfr1143-AS

Olfr112-S

Olfr112-AS

Olfr1126-S

Olfr1126-AS

Olfr1042-S

Olfr1042-AS

Olfr874-S

Olfr874-AS

Olfr1270-S

Olfr1270-AS

Olfr875-S

Olfr875-AS

Olfr693-S

Olfr693-AS

Olfr288-S

Olfr288-AS

ACKNOWLEDGMENTS

AUTHOR CONTRIBUTIONS

FUNDING