Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Mar 3.
Published in final edited form as: Curr Biol. 2016 Mar 31;26(8):1083–1090. doi: 10.1016/j.cub.2016.02.039

An Expression Refinement Process Ensures Singular Odorant Receptor Gene Choice

Ishmail Abdus-Saboor 1,3, Mohammed J Al Nufal 1,4, Maha V Agha 1, Marion Ruinart de Brimont 2, Alexander Fleischmann 2, Benjamin M Shykind 1,5,*
PMCID: PMC12952689  NIHMSID: NIHMS2143403  PMID: 27040780

SUMMARY

Odorant receptor (OR) gene choice in mammals is a paradigmatic example of monogenic and monoallelic transcriptional selection, in which each olfactory sensory neuron (OSN) chooses to express one OR allele from over 1,000 encoded in the genome [1–3]. This process, critical for generation of the circuit from nose to brain [4–6], is thought to occur in two steps: a slow initial phase that randomly activates a single OR allele, followed by a rapid feedback that halts subsequent expression [7–14]. Inherent in this model is a finite failure rate wherein multiple OR alleles may be activated prior to feedback suppression [15, 16]. Confronted with more than one receptor, the neuron would need to activate a refinement mechanism to eliminate multigenic OR expression and resolve unique neuronal identity [16], critical to the generation of the circuit from nose to olfactory bulb. Here we used a genetic approach in mice to reveal a new facet of OR regulation that corrects adventitious activation of multiple OR alleles, restoring monogenic OR expression and unique neuronal identity. Using the tetM71tg model system, in which the M71 OR is expressed in >95% of mature OSNs and potently suppresses the expression of the endogenous OR repertoire [10], we provide clear evidence of a post-selection refinement (PSR) process that winnows down the number of ORs. We further demonstrate that PSR efficiency is linked to OR expression level, suggesting an underlying competitive process and shedding light on OR gene switching and the fundamental mechanism of singular OR choice.

In Brief

Odorant receptor (OR) choice is monogenic and monoallelic, giving olfactory neurons a unique identity. Heterochromatin limits initial activation of OR alleles and is important for singular choice. Abdus-Saboor et al. uncover a post-selection refinement (PSR) mechanism that restores singular OR expression if multiple alleles are initially activated.

RESULTS

We previously generated a transgenic mouse line, tetM71tg, that expresses the M71 odorant receptor (OR), along with marker protein tau-lacZ, under the control of the tetracycline-dependent transactivator [10]. Expression of tetM71tg leads to pervasive expression of M71 in the olfactory epithelium and efficient suppression of the endogenous OR repertoire when driven by tTa supplied by a modified allele of the OMP (olfactory marker protein) gene (OMP-IRES-tTa, [17]). This phenomenon was interpreted as consistent with the feedback model of OR regulation, which would predict that ectopic expression of functional receptor in a neuron, prior to endogenous choice, would halt activation of additional OR alleles [12–14]. Prior studies have suggested that OR onset may precede that of OMP [18, 19]; nonetheless, robust suppression of endogenous choice is observed in the tetM71tg line [10]. Suggestively, however, in genetic experiments designed to force simultaneous expression of both alleles of a modified OR gene, we observed a strong skew toward monoallelic expression and a paucity of biallelic expression in older olfactory sensory neurons (OSNs) [9]. These data introduce a caveat into the interpretation of the tetM71tg-mediated suppression of endogenous OR choice, and we have therefore reexamined this phenomenon.

Onset of OR Expression and Suppression by tetM71tg

The expression of five OR genes, along with tetM71tg, was examined in the olfactory epithelia of tetM71tg and wild-type mice by immunohistochemistry and RNA in situ hybridization (Figure 1). Compared to wild-type controls, the endogenous OR alleles are quantitatively suppressed by tetM71tg expression as previously reported (Figures 1A–1J, data not shown, and [10]). Neurons still expressing endogenous receptor, however, are observed to be displaced on average to a more basal position (Figure 1K) of the neuroepithelium that is comprised predominantly of the younger, OMP-negative subpopulation of neurons [20]. If the population of OSNs still expressing OR were the result of inefficient suppression by tetM71tg, we would expect to see the escapees randomly distributed throughout the height of the epithelium. However, the qualitative change in location suggests that neurons in the tetM71tg line may rather be expressing endogenous OR only as they transit through the OMP-negative developmental stage, shutting it off after the onset of OMP, when they begin to express tTa (from the OMP-IRES-tTa allele) and thus activate tetM71tg. To further characterize the relative timing of OR activation with respect to OMP, we examined the onset of OR and OMP expression. Using two-color RNA in situ hybridization, we confirm and extend previous findings [18, 19] that the onset of OR expression in the olfactory epithelium may precede that of the OMP gene and demonstrate that the proportion of OMP+ OR+ neurons increases with age (Figure S1). These experiments demonstrate a progression in which OR onset precedes that of OMP and suggest that each OSN goes through a single-positive (OR+OMP−) stage prior to becoming double-positive (OR+OMP+), consistent with recently published studies [21–23]. In the context of the tetM71tg line, these data suggest a scenario in which a subset of neurons, or perhaps all, may express both M71 and an endogenous OR prior to shutoff of the endogenous receptor allele.

Figure 1. Suppression and Spatially Restricted Expression of Endogenous ORs by tetM71tg.

Figure 1.

Open in a new tab

(A and B) Coronal sections of olfactory epithelia of control (A) and tetM71tg (B) animals subjected to two-color fluorescent RNA in situ hybridization with riboprobes directed against RNA of odorant receptor B2 (red) and marker protein lacZ (green), with nuclei counterstained by DAPI (blue), visualized by confocal microscopy.

(C and D) Coronal sections of olfactory epithelia of control (C) and tetM71tg (D) animals subjected to immunohistochemistry with antisera directed against odorant receptor M50 (red) and marker protein lacZ (green), with nuclei counterstained by DAPI (blue), visualized by confocal microscopy.

(E and F) Coronal sections of olfactory epithelia of control (E) and tetM71tg (F) animals subjected to immunohistochemistry with antisera directed against odorant receptor MOR28 (red) and marker protein lacZ (green), with nuclei counterstained by DAPI (blue), visualized by confocal microscopy.

(G and H) Coronal sections of olfactory epithelia of control (G) and tetM71tg (H) animals subjected to two-color fluorescent RNA in situ hybridization with riboprobes directed against RNA of odorant receptor P2 (red) and marker protein lacZ (green), with nuclei counterstained by DAPI (blue), visualized by confocal microscopy.

(I and J) Coronal sections of olfactory epithelia of control (I) and tetM71tg (J) animals subjected to two-color fluorescent RNA in situ hybridization with riboprobes directed against odorant receptor I7 (red) and marker protein lacZ (green), with nuclei counterstained by DAPI (blue), visualized by confocal microscopy.

(K) Distribution of relative position of neurons expressing endogenous OR in the olfactory epithelia of wild-type (black circles) and tetM71tg (green circles) animals as determined from RNA in situ and immunohistochemical analyses in (A)–(J). Mean relative position of all data points per OR, normalized to the height of the epithelium, is shown (red bars). p values from a two-tailed unpaired t test comparing relative OR position in wild-type and tetM71tg are statistically significant as indicated.

See also Figures S1 and S2.

Lineage-Marking Demonstrates Post-selection Shutdown

To directly determine whether tetM71tg-induced suppression of the endogenous OR repertoire occurs after initial receptor choice, we used a genetically encoded lineage-marking strategy [12] with a modified allele of the P2 odorant receptor gene, P2-cre (J.A. Gogos, personal communication). The P2-cre allele produces an mRNA encoding both P2 OR and cre recombinase, allowing translation of both, via an internal ribosome entry site (IRES). When crossed to the lineage-marking allele Rosa-loxP-STOP-loxP-CFP [24], selection of the P2-cre allele activates irreversible expression of the CFP marker protein, even if P2-cre expression is subsequently extinguished (Figure 2Ai). We may thus determine whether tetM71tg-induced suppression of endogenous OR occurs pre (Figure 2Aii) or post (Figure 2Aiii) selection of OR by scoring for the presence or absence of cre and CFP expression. Consistent with prior observations [12, 25], the P2-cre allele undergoes switching at a rate of ~20% (Figure 2B versus 2C; quantitated in Figure 2H). In the context of pervasive expression of tetM71tg, the P2-cre allele, as expected, is efficiently suppressed (Figure 2B versus 2E; quantitated in Figure 2H). Strikingly, however, the number of neurons expressing CFP from the cre-activated lineage-marking allele demonstrates a history of wild-type selection frequency of P2-cre (Figure 2C versus 2F; quantitated in Figure 2H). These data represent prima facie evidence that the P2-cre allele is initially expressed at wild-type frequencies, prior to being shut down in the tetM71tg line. Consistent with a transient double-positive stage, we observe the presence of double-positive OSNs, expressing both OR and tetM71tg (Figure S2 and [10]). These data further suggest that OSNs in the tetM71tg line express the endogenous OR repertoire as they transit through an OMP-negative stage and are briefly double-positive (endogenous OR+/tetM71Tg+) prior to efficient shutdown of the endogenous receptor.

Figure 2. Permanent Lineage Marking Reveals Post-selection Shutdown of a Functional OR by tetM71tg.

Figure 2.

Open in a new tab

(A) Diagram of lineage-marking strategy and possible experimental outcomes—analysis of timing of tetM71tg suppression of endogenous OR using a cre recombinase-expressing allele of odorant receptor P2 and the lineage-marking allele Rosa loxP-STOPloxP-CFP. (i) In the control genotype, P2-cre allele selection activates expression of CFP to generate cre+ CFP+ cells. (ii and iii) In the experimental setup, tetM71tg (activated by OMP-IRES-tTa) suppresses endogenous OR expression. If suppression occurred before OR choice (ii), the P2-cre allele would not be expressed. Epithelia would be cre−CFP−. If endogenous ORs are suppressed by tetM71tg after initial expression (iii), epithelia would be cre−CFP+.

(B–D) Immunohistochemical analysis of coronal sections through epithelia of control animals bearing P2-cre and the Rosa loxP-STOP-loxP-CFP lineage-marking allele reveal expression of the P2-cre (B) by antiserum directed against cre (red), and activation of the Rosa loxP-STOP-loxP-CFP allele by cre recombinase (C) with antiserum against CFP (green). Signals from (B) and (C) are merged in (D). Nuclei are counterstained with Toto-3 (blue).

(E–G) Immunohistochemical analysis of coronal sections through epithelia of experimental animals bearing P2-cre and Rosa loxP-STOP-loxP-CFP lineage-marking alleles, and expressing the M71 Tg. Suppression of the P2-cre allele is revealed by antiserum directed against cre (red) (E), and activation of Rosa loxP-STOP-loxP-CFP expression by prior P2-cre expression is demonstrated by antiserum against CFP (green) (F). Signals from (E) and (F) are merged in (G). Nuclei are counterstained with Toto-3 (blue).

(H) Expression data from (B)–(G) (number of positive cells per section) plotted with corresponding two-tailed unpaired t test p values. Expression of tetM71tg suppresses the P2-cre allele (magenta circles, p < 0.0001). Analyses of control (−) and tetM71tg-expressing (+) animals reveal no difference in activation of Rosa loxP-STOP-loxP-CFP in control and tetM71tg-expressing animals (green circles, p = 0.747). A statistically significant level of switching of the P2-cre allele, in the absence of tetM71tg expression, is observed (p = 0.0423).

Suppression of the OR Transcriptome in the tetM71tg Line

The observed post-selection shutdown phenomenon could be due to the simple control of gene expression level. In this scenario, the neuron would not be able to distinguish OR protein produced by an endogenous OR locus from that of the tetM71tg. Unable to regulate the tTa-driven tetM71tg, the neuron may extinguish expression from the endogenous locus to achieve appropriate levels of OR protein. To explore this possibility, we took advantage of the conditional regulation of tetM71tg afforded by the tet/tTa system and administered doxycycline [26] to tetM71tg animals to determine the effect of diminished M71 levels on endogenous OR expression (Figure 3A). In the absence of doxycycline, expression of M71 protein is robust and can be detected by immunostaining with antisera against lacZ or M71 (Figures 3D–3F; quantitated in Figures 3B and 3C). However, after administration of doxycycline for 4 and then 8 days, tetM71tg is efficiently shut off and M71 protein dissipates (Figures 3G–3L; quantitated in Figures 3B and 3C). Throughout the doxycycline feeding regimen and the ensuing diminution of M71 protein, the expression of the endogenous ORs remains suppressed (Figures 3M–3R), indicating that tetM71tg-mediated suppression likely represents a permanent state of the neuron, and not a temporary downregulation of endogenous OR. The doxycycline-fed tetM71tg animals, in which endogenous OR expression fails to be re-initiated after M71 levels wane, likely represent a state in which the olfactory neurons do not express any OR.

Figure 3. Timed Administration of Doxycycline Extinguishes tetM71tg Expression but Does Not Restore Endogenous OR Repertoire.

Figure 3.

Open in a new tab

(A) Diagram depicting strategy used in staged doxycycline (dox) administration experiments. In the absence of dox, tTa activates the tetM71tg allele (on), whereas dox administration shuts off expression (off).

(B and C) Staged administration of dox to tetM71tg animals from 0 to 8 days suppresses lacZ expression (B) and M71 expression (C) from the tetM71tg, as determined by immunohistochemical analyses (D–L) quantified by ImageJ software.

(D–F) Pervasive high-level tetM71tg expression revealed by two-color immunohistochemistry of coronal sections of tetM71tg olfactory epithelia at P30 with antiserum directed against lacZ (D, green) and M71 (E, red). (F) shows merged image. Nuclei are counterstained with Toto-3 (blue).

(G–L) Suppression of tetM71tg expression after 4 (G–I) and 8 (J–L) days of dox administration revealed by two-color immunohistochemistry (IHC) of coronal sections of tetM71tg olfactory epithelia of adult animals subjected to antiserum directed against lacZ (green) and M71 (red). Nuclei are counterstained with Toto-3 (blue).

(M and N) Expression of M50 and M71 in coronal sections of control (no dox) olfactory epithelia revealed by IHC with antisera directed against M50 (M, red) and M71 (N, red). Nuclei are counterstained with Toto-3 (blue).

(O–R) Suppression of endogenous ORs persists after dox treatment. Suppression of M50 and M71 persists after 4 (O and P) and 8 (Q and R) days of dox administration, revealed by IHC of coronal sections of M71-Tg olfactory epithelia of adult animals subjected to antiserum directed against M50 (O and Q, red) and M71 (P and R, red). Nuclei are counterstained with Toto-3 (blue).

Means ± SD are shown.

To gain insight into the nature of tetM71tg-mediated OR suppression of the endogenous OR repertoire, we analyzed the transcriptome of the olfactory epithelia of tetM71tg mice by RNA sequencing (RNA-seq). As expected, M71 is massively overexpressed in the tetM71tg olfactory epithelium (Figure S3). The expression of the majority of ORs was greatly reduced in tetM71tg olfactory epithelia compared to wild-type controls (Figure S3). Suggestively, the RNA-seq data revealed ranks of suppression among the endogenous ORs, with receptors transcribed at a high level, such as MOR28 [27], less efficiently downregulated than more modestly expressed ones like P2 (Figure S3). These data suggest the possibility that receptor expression levels play a role in the susceptibility of tetM71tg-mediated shutdown. One mechanism to account for such expression-level dependence may involve inter-allele competition between ORs, with more highly transcribed genes more likely to remain expressed.

Pervasive, Low-Level OR Expression Does Not Suppress

To examine the effect that the level of receptor expression has on suppression, we derived a mouse line with pervasive, low-level OR expression. Using homologous recombination in mouse ES cells, we modified the OMP locus by inserting the coding region of MOR28 linked to an IRES element into the 3′ non-coding region of the OMP gene, to create the OiR line (Figure 4A). Thus, all neurons that express this engineered OMP allele will synthesize a bicistronic mRNA allowing the translation of both OMP and MOR28 receptor proteins. Importantly, the onset of MOR28 receptor expression from the OiR allele mirrors that of the M71 receptor in the tetM71tg mouse, as both use OMP gene transcription to drive receptor expression. Coronal sections through the olfactory epithelia of OiR heterozygous mice were subject to immunohistochemistry with antiserum directed against the MOR28 receptor and reveal pervasive, low-level expression of MOR28 across the neuroepithelium (Figures 4B and 4C). The expression level of MOR28 from its endogenous locus, which is comparable to the level of M71 expressed from tetM71tg, is on average 3-fold higher than MOR28 expression from OiR, as revealed by quantitative immunohistochemical analyses (Figures 4D–4F and data not shown). In marked contrast to the consequence of tetM71tg expression, the endogenous OR repertoire is unaltered in the OiR line, with comparable frequencies of expression of M71 and M50 in OiR and control neuroepithelia (Figures 4G–4L). These experiments demonstrate that the level of expression of a pervasively driven OR likely dictates the efficiency with which post-selection shutdown may occur. Together, these data suggest a model in which a competitive relationship between OR alleles could mediate post-selection shutdown, consistent with the observation of differential efficiency of suppression of endogenous OR alleles in the RNA-seq analysis of the tetM71tg line (Figure S3).

Figure 4. Pervasive Low-Level OR Expression Does Not Suppress Endogenous OR Expression.

Figure 4.

Open in a new tab

(A) Targeted generation of OiR allele at the OMP locus. (i) The OiR targeting construct transcribes a bicistronic mRNA allowing translation of both the OMP and MOR28 proteins via an internal ribosome entry site (IRES). (ii) The unmodified OMP genomic locus. (iii) Homologous recombination in mouse ES cells followed by self-excision of the ACN selection cassette yields the OiR allele.

(B) Coronal section through the olfactory epithelium of an OiR animal reveals pervasive expression of the OiR allele in the OMP+ layer of the main olfactory epithelium at postnatal day 5, detected by immunohistochemistry with antiserum directed against MOR28 receptor protein (red) and nuclei counterstained with Topro-3 (blue).

(C) High-power magnification of a region of the field shown in (B).

(D and E) Coronal section through the MOR28-expressing zone of the olfactory epithelium of an OiR animal subjected to anti-MOR28 IHC reveals sparse endogenous MOR28+ cells (bright red) and MOR28+ cells expressed from OiR (weak red) (D). Nuclei were counterstained with Toto-3 (blue) in (E).

(F) Graph of relative MOR28 expression levels assayed by immunohistochemistry with anti-MOR28 antibody and quantified by ImageJ software (see Supplemental Experimental Procedures for details) reveals much lower levels of MOR28 generated by the OiR allele as compared to the endogenous MOR28 gene. Comparison was made between 100 endogenous MOR28+ cells (black circles) and 100 OMP-IRES-MOR28+ cells (green circles). A two-tailed unpaired t test performed was performed (p < 0.0001).

(G–L) Pervasive low-level expression of MOR28 from the OiR allele does not suppress endogenous OR expression.

(G and H) Coronal sections of adult control (G) and OiR (H) olfactory epithelia subjected to two-color immunohistochemistry with antiserum directed against the odorant receptors M71 (green) and MOR28 (red). Nuclei were counterstained with Toto-3 (blue).

(I) Numbers of M71+ cells in control (wt) versus OiR olfactory epithelia are not statistically different (n > 50 each for genotype).

(J and K) Coronal sections of P60 control (J) and OiR (K) olfactory epithelia subjected to two-color immunohistochemistry with antiserum directed against the odorant receptors M50 (green) and MOR28 (red, with cells expressing the endogenous MOR28 allele visible in J). Nuclei were counterstained with Toto-3 (blue).

(L) Numbers of M50+ cells in control (wt) and OiR olfactory epithelia are not statistically different (n > 50 each for genotype).

Means ± SD are shown. See also Figures S3 and S4.

DISCUSSION

Cleaning Up After Feedback

Recently, the Lomvardas lab has revealed components of OR feedback. Induction of the unfolded protein response (UPR) by functional OR protein leads to activation of adenylyl cyclase III and downregulation of Lsd-1, halting further derepression, and thus activation, of additional ORs [28, 29]. This model requires sparse initial OR activation, where one allele is derepressed and expressed. The process we describe here may represent a “failsafe” mechanism when the UPR-mediated pathway cannot generate a single outcome. The timing of the UPR-mediated pathway differs from that of post-selection refinement (PSR), which functions after the onset of OMP (tetM71tg is driven by OMP-IRES-tTa). Furthermore, OMP is expressed as Lsd-1 wanes [8], and conditional knockout studies of Lsd-1 reveal no late (OMP+) role [8]. OMP-driven tetM71tg is not likely to reactivate initial selection, or feedback, as no aberrant expression or reactivation of Lsd-1 is observed in the tetM71tg line (data not shown). Logic and timing differences between onset of feedback and PSR suggest a hierarchy wherein the UPR pathway mediates early-phase OR regulation and PSR refines it when multiple ORs are activated. However, given the uncertainty in timing of transition from initiation to maintenance, PSR may utilize or reactivate elements of the UPR-mediated process.

Prior Evidence of PSR and the Phenomenon of Switching

Previous studies have suggested the existence of PSR [9]. In experiments to force biallelic OR expression using tet/tTa, we observed high-frequency activation of the tet-modified locus. However, we observed just 3% biallelic expression, not the 25% expected [9]. Significantly, allelic inclusion was restricted to the younger stratum of the olfactory epithelium [9]. These findings suggest that biallelic OR expression is not stably maintained. PSR could provide a mechanism to filter out the adventitious activations, expected at some finite level, in stochastic choice [16]. Importantly, PSR may explain “wild-type switching” observed in lineage-marking experiments [12]. Interpreted as serial OR expression, wild-type switching is better seen as parallel expression: multiple ORs transcribed early, with one remaining after PSR [16]. This interpretation fits well with recent RNA sequencing experiments [21–23]. Work describing coexpression of OR alleles supports PSR with allelic inclusion declining with age [30], although other, exploratory results describe rare, stable allelic inclusion [31]. Consistent with competition underlying PSR, wild-type switching rates are generally inversely proportional to OR expression levels: higher per-cell expression correlates with lower switching (Figure 2; [12]). Furthermore, frequently expressed ORs are also most abundantly expressed per cell [27], consistent with a model where highly transcribed ORs frequently prevail.

The kinetic model of OR choice posits sparse initial activation to ensure that a single OR is expressed when feedback repression takes effect, and lineage-marking studies show restricted OR transcription in the epithelium [12]. The epigenetic state of OR chromatin provides a biochemical underpinning for inefficient activation and the role of repression [7], with H3K9 methylation likely underlying the kinetic restriction. Lyons et al. have ablated generation of H3K9-methylated OR chromatin, with genetic knockouts of G9a and GLP methyltransferases [32]. Without H3K9 methylation, pervasive, not sparse, activation of ORs would be expected. Without H3K9 methylation, a less diverse OR repertoire is observed, with a few OR genes predominating [32]. Strikingly, however, wholesale violation of the one-OR-per-neuron rule is not observed [32]. This phenomenology is analogous to the above-described forced allelic inclusion experiments, where inclusion is underrepresented [9], further supporting the existence of PSR and its role in singular OR expression.

The Monoclonal Nose Mouse Revisited

The tetM71tg line was originally employed to test the feedback model, which predicted that pervasive, early expression of OR would suppress endogenous selection and “monoclonalize” the olfactory epithelium [10]. The efficient suppression observed was interpreted as consistent with the feedback model. Our experiments emend this interpretation. While efficiently suppressing endogenous choice, tetM71tg expression occurs after onset of endogenous ORs, suppressing them via PSR; it is through this aperture that the interpretations in Fleischmann et al. [10] should be viewed.

Competition and PSR

The correlation of OR transcriptional level with the likelihood of suppressing contemporaneously expressed alleles suggests an autonomous, competitive mechanism underlying PSR. What mechanism might effect such competition? Compellingly, OR loci cluster in subnuclear domains decorating chromocenters, which are critical for initiating and maintaining receptor expression [33–35]. One model of PSR envisions eviction and silencing of endogenous alleles by tetM71tg. However, expressed tetM71tg does not associate with elements of these foci (Figure S4), including the pericentromeric heterochromatin or the H element [34, 35]. It remains possible that tetM71tg expression disrupts these interactions indirectly or transiently translocates itself to the compartment. Many examples of monoallelic expression involve regulatory non-coding RNAs (ncRNA; [36]). Similarly, an ncRNA derived from OR sequences may effect PSR. Dosage of putative ncRNA regulators may explain the threshold level of expression required for PSR: low level (OiR) does not suppress. This model would explain the inverse relationship between OR expression level and stability, in which highly expressed ORs switch less often (are more likely to outcompete). An ncRNA in PSR could be vestigial from other competitive, random, monoallelic choices like X chromosome inactivation [36].

We have previously shown that the OR promoter is the substrate for tetM71tg-induced suppression [9]. The tetM71tg allele suppresses endogenous OR without being shut down, as it contains an exogenous promoter [26]. While tetM71tg contains an OR-coding sequence, we see no evidence for the assertion [37] that it affects any aspect of OR regulation. Furthermore, the ability to express MOR28 under OMP control (OiR) directly belies an OR-coding suppressive effect [37].

Summary and the Role of PSR

OR regulation generates >2,000 transcriptional outcomes, endowing an equal number of OSN identities. This extreme selectivity results from a slow initial phase, when individual OR alleles are infrequently activated, followed by a feedback stage halting the process and preserving singular choice (reviewed in [11, 16, 28]). Mathematical modeling has determined parameters for activation and feedback that ensure a high probability of singular expression [15]. These analyses also defined a failure rate, when activation proceeds too quickly, or feedback proceeds too slowly, resulting in neurons expressing multiple ORs. OSNs are unlikely to use feedback suppression to restore singular OR expression once more than one allele is activated. We have revealed a post-selection refinement (PSR) mechanism, which restores singular OR expression and unique neuronal identity.

How large a role could PSR play in OR regulation? The feedback suppression elicited after receptor expression is robust, and OR pseudogenes are unable to evoke it, as they cannot activate UPR [28]. We may estimate the fraction of PSR-mediated selection outcomes by examining the rate of wild-type switching, which is ~10% for highly transcribed MOR28 [12], and roughly twice that for P2 (this study), which is expressed at a lower per-cell level. OR alleles expressed at even lower per-cell levels, and found less frequently in the epithelium, may demonstrate even higher rates of wild-type switching and higher dependence on PSR.

Supplementary Material

S1

SUPPLEMENTAL INFORMATION

Supplemental Information includes four figures and Supplemental Experimental Procedures and can be found with this article online at http://dx.doi.org/10.1016/j.cub.2016.02.039.

Highlights.

  • A genetic approach in mice reveals a new facet of odorant receptor (OR) regulation

  • Adventitious expression of multiple ORs activates post-selection refinement (PSR)

  • PSR works via the shutdown of all but one OR and may involve allelic competition

  • PSR ensures that mature olfactory neurons express a single OR allele

ACKNOWLEDGMENTS

We wish to thank the Joseph Gogos (Columbia University) for the P2-IRES-Cre mouse strain and Monica Mendelsohn, Adriana Nemes, Yonghua Sun, and Richard Axel (Columbia University) for generation of mouse lines. We thank Gina Gacutan for editorial assistance. We also thank Gilad Barnea (Brown University) for antisera to MOR28 and M71. This work was supported by Qatar National Research Fund NPRP grant 4-1033-3-279, the Biomedical Research Program (BMRP) at Weill Cornell Medical College in Qatar, and the Microscopy Core (both funded by the Qatar Foundation). This work was also supported by a Marie Curie International Reintegration grant (IRG 276869) and the “Amorçage de jeunes équipes” program (AJE201106) of the Fondation pour la Recherche Médical (to A.F.). All procedures were conducted according to animal protocols approved by Institutional Animal Care and Use Committee of the Weill Cornell Medical College and NIH guidelines.

REFERENCES

  • 1.Buck L, and Axel R (1991). A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65, 175–187. [DOI] [PubMed] [Google Scholar]
  • 2.Chess A, Simon I, Cedar H, and Axel R (1994). Allelic inactivation regulates olfactory receptor gene expression. Cell 78, 823–834. [DOI] [PubMed] [Google Scholar]
  • 3.Malnic B, Hirono J, Sato T, and Buck LB (1999). Combinatorial receptor codes for odors. Cell 96, 713–723. [DOI] [PubMed] [Google Scholar]
  • 4.Wang F, Nemes A, Mendelsohn M, and Axel R (1998). Odorant receptors govern the formation of a precise topographic map. Cell 93, 47–60. [DOI] [PubMed] [Google Scholar]
  • 5.Ressler KJ, Sullivan SL, and Buck LB (1994). Information coding in the olfactory system: evidence for a stereotyped and highly organized epitope map in the olfactory bulb. Cell 79, 1245–1255. [DOI] [PubMed] [Google Scholar]
  • 6.Vassar R, Chao SK, Sitcheran R, Nuñez JM, Vosshall LB, and Axel R (1994). Topographic organization of sensory projections to the olfactory bulb. Cell 79, 981–991. [DOI] [PubMed] [Google Scholar]
  • 7.Magklara A, Yen A, Colquitt BM, Clowney EJ, Allen W, Markenscoff-Papadimitriou E, Evans ZA, Kheradpour P, Mountoufaris G, Carey C, et al. (2011). An epigenetic signature for monoallelic olfactory receptor expression. Cell 145, 555–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Lyons DB, Allen WE, Goh T, Tsai L, Barnea G, and Lomvardas S (2013). An epigenetic trap stabilizes singular olfactory receptor expression. Cell 154, 325–336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Fleischmann A, Abdus-Saboor I, Sayed A, and Shykind B (2013). Functional interrogation of an odorant receptor locus reveals multiple axes of transcriptional regulation. PLoS Biol. 11, e1001568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Fleischmann A, Shykind BM, Sosulski DL, Franks KM, Glinka ME, Mei DF, Sun Y, Kirkland J, Mendelsohn M, Albers MW, and Axel R (2008). Mice with a “monoclonal nose”: perturbations in an olfactory map impair odor discrimination. Neuron 60, 1068–1081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Shykind BM (2005). Regulation of odorant receptors: one allele at a time. Hum. Mol. Genet. 14, R33–R39. [DOI] [PubMed] [Google Scholar]
  • 12.Shykind BM, Rohani SC, O’Donnell S, Nemes A, Mendelsohn M, Sun Y, Axel R, and Barnea G (2004). Gene switching and the stability of odorant receptor gene choice. Cell 117, 801–815. [DOI] [PubMed] [Google Scholar]
  • 13.Serizawa S, Miyamichi K, Nakatani H, Suzuki M, Saito M, Yoshihara Y, and Sakano H (2003). Negative feedback regulation ensures the one receptor-one olfactory neuron rule in mouse. Science 302, 2088–2094. [DOI] [PubMed] [Google Scholar]
  • 14.Lewcock JW, and Reed RR (2004). A feedback mechanism regulates monoallelic odorant receptor expression. Proc. Natl. Acad. Sci. USA 101, 1069–1074. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Tan L, Zong C, and Xie XS (2013). Rare event of histone demethylation can initiate singular gene expression of olfactory receptors. Proc. Natl. Acad. Sci. USA 110, 21148–21152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Abdus-Saboor I, Fleischmann A, and Shykind B (2014). Setting limits: maintaining order in a large gene family. Transcription 5, e28978. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Yu CR, Power J, Barnea G, O’Donnell S, Brown HEV, Osborne J, Axel R, and Gogos JA (2004). Spontaneous neural activity is required for the establishment and maintenance of the olfactory sensory map. Neuron 42, 553–566. [DOI] [PubMed] [Google Scholar]
  • 18.Iwema CL, and Schwob JE (2003). Odorant receptor expression as a function of neuronal maturity in the adult rodent olfactory system. J. Comp. Neurol. 459, 209–222. [DOI] [PubMed] [Google Scholar]
  • 19.Rodriguez-Gil DJ, Bartel DL, Jaspers AW, Mobley AS, Imamura F, and Greer CA (2015). Odorant receptors regulate the final glomerular coalescence of olfactory sensory neuron axons. Proc. Natl. Acad. Sci. USA 112, 5821–5826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Farbman AI, and Margolis FL (1980). Olfactory marker protein during ontogeny: immunohistochemical localization. Dev. Biol. 74, 205–215. [DOI] [PubMed] [Google Scholar]
  • 21.Tan L, Li Q, and Xie XS (2015). Olfactory sensory neurons transiently express multiple olfactory receptors during development. Mol. Syst. Biol. 11, 844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Hanchate NK, Kondoh K, Lu Z, Kuang D, Ye X, Qiu X, Pachter L, Trapnell C, and Buck LB (2015). Single-cell transcriptomics reveals receptor transformations during olfactory neurogenesis. Science 350, 1251–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Saraiva LR, Ibarra-Soria X, Khan M, Omura M, Scialdone A, Mombaerts P, Marioni JC, and Logan DW (2015). Hierarchical deconstruction of mouse olfactory sensory neurons: from whole mucosa to single-cell RNA-seq. Sci. Rep. 5, 18178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Srinivas S, Watanabe T, Lin CS, William CM, Tanabe Y, Jessell TM, and Costantini F (2001). Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus. BMC Dev. Biol. 1, 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Bader A, Bautze V, Haid D, Breer H, and Strotmann J (2010). Gene switching and odor induced activity shape expression of the OR37 family of olfactory receptor genes. Eur. J. Neurosci. 32, 1813–1824. [DOI] [PubMed] [Google Scholar]
  • 26.Gossen M, Freundlieb S, Bender G, Müller G, Hillen W, and Bujard H (1995). Transcriptional activation by tetracyclines in mammalian cells. Science 268, 1766–1769. [DOI] [PubMed] [Google Scholar]
  • 27.Young JM, Shykind BM, Lane RP, Tonnes-Priddy L, Ross JA, Walker M, Williams EM, and Trask BJ (2003). Odorant receptor expressed sequence tags demonstrate olfactory expression of over 400 genes, extensive alternate splicing and unequal expression levels. Genome Biol. 4, R71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Dalton RP, Lyons DB, and Lomvardas S (2013). Co-opting the unfolded protein response to elicit olfactory receptor feedback. Cell 155, 321–332. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Magklara A, and Lomvardas S (2013). Stochastic gene expression in mammals: lessons from olfaction. Trends Cell Biol. 23, 449–456. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Tian H, and Ma M (2008). Activity plays a role in eliminating olfactory sensory neurons expressing multiple odorant receptors in the mouse septal organ. Mol. Cell. Neurosci. 38, 484–488. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Rawson NE, Eberwine J, Dotson R, Jackson J, Ulrich P, and Restrepo D (2000). Expression of mRNAs encoding for two different olfactory receptors in a subset of olfactory receptor neurons. J. Neurochem. 75, 185–195. [DOI] [PubMed] [Google Scholar]
  • 32.Lyons DB, Magklara A, Goh T, Sampath SC, Schaefer A, Schotta G, and Lomvardas S (2014). Heterochromatin-mediated gene silencing facilitates the diversification of olfactory neurons. Cell Rep. 9, 884–892. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Lomvardas S, Barnea G, Pisapia DJ, Mendelsohn M, Kirkland J, and Axel R (2006). Interchromosomal interactions and olfactory receptor choice. Cell 126, 403–413. [DOI] [PubMed] [Google Scholar]
  • 34.Clowney EJ, LeGros MA, Mosley CP, Clowney FG, Markenskoff-Papadimitriou EC, Myllys M, Barnea G, Larabell CA, and Lomvardas S (2012). Nuclear aggregation of olfactory receptor genes governs their monogenic expression. Cell 151, 724–737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Markenscoff-Papadimitriou E, Allen WE, Colquitt BM, Goh T, Murphy KK, Monahan K, Mosley CP, Ahituv N, and Lomvardas S (2014). Enhancer interaction networks as a means for singular olfactory receptor expression. Cell 159, 543–557. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Lee JT (2009). Lessons from X-chromosome inactivation: long ncRNA as guides and tethers to the epigenome. Genes Dev. 23, 1831–1842. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Nguyen MQ, Zhou Z, Marks CA, Ryba NJP, and Belluscio L (2007). Prominent roles for odorant receptor coding sequences in allelic exclusion. Cell 131, 1009–1017. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1
NIHMS2143403-supplement-S1.pdf (1.4MB, pdf)

RESOURCES