In mice, discrete odors can selectively promote the neurogenesis of sensory neuron subtypes that they stimulate

Abstract

In mammals, olfactory sensory neurons (OSNs) are born throughout life, ostensibly solely to replace neurons lost via turnover or injury. This assumption follows from the hypothesis that olfactory neurogenesis is stochastic with respect to neuron subtype, as defined by the single odorant receptor that each neural precursor stochastically chooses out of hundreds of possibilities. This assumption is challenged, however, by recent findings that the birthrates of a fraction of OSN subtypes are selectively reduced by olfactory deprivation. These findings raise questions about how, and why, olfactory stimuli are required to accelerate the neurogenesis rates of some subtypes, including whether the stimuli are specific (e.g., discrete odorants) or generic (e.g., broadly activating odors or mechanical stimuli). Based on previous findings that the exposure of mice to sex-specific odors can increase the representations of subtypes responsive to those odors, we hypothesized that the neurogenic stimuli comprise discrete odorants that selectively stimulate OSNs of the same subtypes whose birthrates are accelerated. In support of this, we have found, using scRNA-seq and subtype-specific OSN birthdating, that exposure to male and exogenous musk odors can accelerate the birthrates of subtypes responsive to those odors. These findings reveal that certain odor experiences can selectively 'amplify' specific OSN subtypes and suggest that persistent OSN neurogenesis serves, in part, an adaptive function.

Figure 1.

scRNA-seq analyses of the open and closed sides of whole OEs from UNO-treated adolescent male mice reveal greater quantities of newborn OSNs of musk-responsive subtypes on the open side of the OE relative to the closed. A, B. Experimental timeline (A) and conditions (B) used to generate scRNA-seq datasets for assessing the effects of UNO on quantities of newborn OSNs of specific subtypes in the OEs of adolescent male mice. Datasets were generated from the open and closed side of the OEs of mice that were UNO-treated at PD 14, weaned sex-separated at PD 21, and sacrificed at PD 28 . C. A UMAP representation of all cells within the open (left) and closed (right) side datasets corresponding to OE sample 2 (OE 2). Cells within the OSN lineage are represented by colored dots, as defined in the legend: horizontal basal cells (HBC; red), globose basal cells (GBC; orange), immediate neuronal precursor 1 cells (INP1; pink), immediate neuronal precursor 2 cells (INP2; purple) immediate neuronal precursor 3 cells (INP3; yellow), immature OSNs (iOSN; green), and mature OSNs (mOSNs; blue). D. UMAP representation of cells within the OSN lineage of the open (left) and closed (right) side datasets of OE 2. Immature (Gap43+) OSNs of the 5 known musk-responsive subtypes are represented by colored dots, as indicated in the legend. E. Quantification of individual (lines) and mean (bars) percentages of the OSN population represented by immature OSNs of musk-responsive subtypes (left) or randomly chosen zone 2/3 subtypes (right) within the open and closed side datasets. Quantifications represent the averages of subtype-specific iOSN quantities obtained from scRNA-seq datasets corresponding to the OEs from 2 different mice (OE 1 and OE 2). Error bars: SEM. See also Figure 1–figure supplements 1 and 2.

Figure 2.

EdU birthdating analyses confirm that UNO-treated adolescent male mice exhibit greater quantities of newborn OSNs of specific musk-responsive subtypes on the open side of the OE relative to the closed. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of UNO on quantities of newborn OSNs of specific subtypes in adolescent male mice. Mice were UNO-treated at PD 14, weaned sex-separated at PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C, E–H. Left: Representative images of OE sections from UNO-treated adolescent male mice that were exposed, at the time of EdU labeling, to male littermates (♂ → ♂), with newborn OSNs (OR+/EdU+) indicated by white arrows. Right: Quantifications of newborn OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated male mice reveal significant open-side biases in quantities of newborn OSNs of musk-responsive subtypes Olfr235 (C), Olfr1440 (E), and Olfr1431 (F), but not the musk-responsive subtype Olfr1437 (G) or the SBT-responsive subtype Olfr912 (H). D. Enlarged, split-channel view of boxed region in (C), with Olfr235+/EdU+ OSNs indicated by white arrows. Scale bars: 150 μm. Each line represents a distinct mouse (n = 4–10 mice [≥ 5 sections/mouse] per OSN subtype). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P > 0.05; ratio paired two-tailed t-test. Error bars: SEM. See also Figure 2–figure supplement 1.

Figure 3.

Exposure of mice to male mouse odors causes selective increases in quantities of newborn OSNs of the musk-responsive subtype Olfr235. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of exposure to male mouse odors on quantities of newborn OSNs of specific subtypes. Male and female mice were either UNO-treated or untreated (non-occluded) at PD 14, weaned sex-separated or sex-combined at PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C, D. Representative images of OE sections from UNO-treated female mice that were exposed, at the time of EdU labeling, to either female (♀ → ♀) (C) or male (♀ → ♂) (D) littermates, with newborn Olfr235 OSNs (OR+/EdU+) indicated by white arrows. E, F. Quantifications of newborn (E) and total (F) Olfr235 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated mice reveal significant open-side biases in quantities of newborn Olfr235 OSNs in male or female mice exposed to male littermates (♂ → ♂ or ♀ → ♂), but not in females exposed to only female littermates (♀ → ♀) (E-left), and greater UNO effect sizes for quantities of newborn (E-right) and total (F) Olfr235 OSNs within the OEs of mice that were exposed to male littermates. G-left, middle. Representative images (middle) corresponding to the region outlined in the schematic (left, red box) of OE sections from non-occluded mice that were exposed to male littermates (♂ → ♂, ♀ → ♂) or to female littermates (♀ → ♀) at the time of EdU labeling, with Olfr235+/EdU+ OSNs indicated by white arrows. G-right, H. Quantifications of newborn Olfr235 (G-right) and Olfr912 (H) OSNs in tissue sections spanning the anterior-posterior lengths of OEs reveal significantly greater quantities of newborn Olfr235 OSNs in non-occluded mice exposed, at the time of EdU labeling, to male littermates (♂ → ♂, ♀ → ♂) compared to those exposed only to female littermates (♀ → ♀), while quantities of newborn Olfr912 OSNs showed no significant differences between mice exposed to male littermates and those exposed only to females. Scale bars: 150 μm (C, D), 50 μm (G). Each line or circle represents a distinct mouse (n = 4–10 mice [≥ 5 sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P> 0.05; ratio paired two-tailed t-test (E, left); one-way ANOVA test, FDR-adjusted (E, right, F-H). Error bars: SEM. Data in panel E for newborn OSN quantities in ♂ → ♂ OE samples correspond to Figure 2. See also Figure 3–figure supplements 1–3.

Figure 4.

Exposure of mice to musk odors causes selective increases in quantities of newborn OSNs of musk-responsive subtypes. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of exposure to musk (muscone, ambretone) or non-musk (SBT, IAA) odorants on quantities of newborn OSNs of specific subtypes. Female mice were either UNO-treated or untreated (non-occluded) at PD 14, either exposed or unexposed to an exogenous musk or non-musk odorant starting at PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C. Left: Representative image of an OE section from a UNO-treated female mouse that was exposed, at the time of EdU labeling, to 0.1% muscone, with newborn Olfr235 OSNs (OR+/EdU+) indicated by white arrows. Middle, right: Quantifications of newborn Olfr235 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated female mice reveal significant open-side biases (middle) and significantly greater UNO effect sizes (right) in mice exposed to 0.1% or 1% (but not 10%) muscone compared to 0%, with a maximum effect size observed at a concentration of 0.1%. D–F. Left, middle: Representative images (middle) corresponding to the regions outlined in schematics (left, red box) of OE sections from non-occluded female mice that were exposed just to female littermates (♀ → ♀) or also to 0.1% muscone or 0.1% ambretone, with newborn Olfr235 (D), Olfr1440 (E), and Olfr1431 (F) OSNs (OR+/EdU+) indicated by white arrows. Right: Quantifications of newborn Olfr235 (D), Olfr1440 (E), and Olfr1431 (F) OSNs in tissue sections spanning the anterior-posterior lengths of OEs reveal that, compared to females exposed just to female littermates (♀ → ♀), significantly greater quantities of newborn OSNs of all 3 subtypes were observed in females also exposed to muscone (musc; 0.1%) or ambretone (amb; 0.1%), but not the non-musk odorants SBT (0.1%) or IAA (0.1%). Scale bars: 150 μm (C), 50 μm (D–F). Each line or circle represents a distinct mouse (n = 5–10 mice [≥ 5 OE sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P> 0.05; ratio paired two-tailed t-test (C, middle); one-way ANOVA test, FDR-adjusted (C, right, D–F). Error bars: SEM. Data for ♀ → 0% muscone samples in panel C correspond to Figure 3E. Image and data for ♀ → ♀ samples (D) correspond to Figure 3G. See also Figure 4–figure supplements 1–4.

Figure 5.

Musk exposure increases quantities of newborn OSNs of musk-responsive subtypes in adulthood. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of exposure to musk odors on quantities of newborn OSNs of specific subtypes in non-occluded adult mice. Mice were either exposed or unexposed to 0.1% muscone starting at PD 21, EdU-labeled at PD 56–58, and sacrificed at PD 65. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C–F. Left: Representative images corresponding to the region outlined in the schematic in panel B (top), of OEs from non-occluded adult female mice that were either exposed to just female littermates (♀ → ♀) or also to muscone (♀ → 0.1% muscone), with newborn Olfr235 (C), Olfr1440 (D), Olfr1431 (E), and Olfr912 (F) OSNs (OR+/EdU+) indicated by white arrows. Right: Quantifications of newborn OSNs within tissue sections spanning the anterior-posterior lengths of OE from non-occluded adult female mice reveal greater quantities of newborn OSNs of all 3 musk-responsive subtypes (Olfr235, Olfr1440, and Olfr1431) in mice exposed to muscone compared to just female littermates (♀ → ♀), while quantities of newborn OSNs of the SBT-responsive subtype Olfr912 were relatively unaffected. Scale bars: 50 μm. Each circle represents a distinct mouse (n = 5–7 mice [≥ 5 OE sections/mouse] per OSN subtype and condition). *P < 0.05; n.s. P> 0.05; unpaired two-tailed t-test. Error bars: SEM.

Figure 6.

Stimulation-dependent increases in quantities of newborn OSNs of musk-responsive subtypes are stable over time following birth, consistent with a mechanism involving altered neurogenesis. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the time-dependence of the effects of exposure to male or musk odors on quantities of newborn OSNs of specific subtypes. Mice were either UNO-treated or untreated (non-occluded) at PD 14, either exposed or unexposed to muscone starting at PD 21, EdU-labeled at PD 28, and sacrificed at either PD 32 (4 d post-EdU) or PD 35 (7 d post-EdU). OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C. Representative image of an OE section from a UNO-treated male mouse that was exposed, at the time of EdU labeling, to male littermates (♂ → ♂) and sacrificed 4 d post-EdU, with newborn Olfr235 OSNs (OR+/EdU+) indicated by white arrows. D, E. Quantifications of newborn Olfr235 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated male mice that were exposed to male littermates (♂ → ♂) (D) or female mice that were exposed to 0.1% muscone (♀ → 0.1% muscone) (E) and sacrificed 4 or 7 d post-EdU. Under both conditions, significant open-side biases in quantities of newborn Olfr235 OSNs were observed at both timepoints (left), with no statistically significant differences in UNO effect sizes observed over time (right). F. Representative images (right) corresponding to the region outlined in schematic (left, red box) of OE sections from non-occluded female mice that were exposed, at the time of EdU labeling, to either just their female littermates (♀ → ♀) or also to muscone (♀ → 0.1% muscone) and sacrificed 4 d post-EdU, with newborn Olfr235 OSNs (OR+/EdU+) indicated by white arrows. G. Quantifications of newborn Olfr235 (top), Olfr1440 (middle), and Olfr1431 (bottom) OSNs within tissue sections spanning the anterior-posterior lengths of OEs from non-occluded female mice that exposed to either just their female littermates (♀ → ♀) or also to muscone (♀ → 0.1% muscone) and sacrificed 4 or 7 d post-EdU. Greater quantities of newborn OSNs of all 3 subtypes were observed at both timepoints in muscone-exposed mice compared to controls, with no statistically significant interaction between time and newborn OSN quantities observed. Scale bars: 150 μm (C), 50 μm (F). Each circle represents a distinct mouse (n = 4–10 mice [≥ 5 OE sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; ratio paired two-tailed t-test (D, E left); unpaired two-tailed t-test (D, E right); two-way ANOVA test (factors: muscone exposure, days post-EdU) (G). Error bars: SEM. Data for 7 d post-EdU samples correspond to Figures 2–4. See also Figure 6–figure supplement 1.

Figure 7.

Model for how specific odors selectively increase quantities of newborn OSNs of subtypes that are stimulated by them. A fraction of OSN subtypes (e.g., Olfr235), upon stimulation by discrete odors (e.g., musks), undergo accelerated rates of neurogenesis. Most subtypes (e.g., Olfr912) do not exhibit altered rates of neurogenesis upon stimulation by odors that stimulate them (e.g., SBT). One hypothetical mechanism involves selective signaling from odor-stimulated mature OSNs of specific subtypes to neural progenitors.

Appendix 1–figure 1.

Identification of OSN subtypes that are candidates for undergoing sex-specific- and/or musk odor-accelerated neurogenesis. A. In a previous study, the OE transcript profiles of female and male mice that were housed either sex-separated or sex-combined from weaning (PD 21) until 6 months of age were profiled and compared via bulk RNA-seq ^,. B. OSN subtypes previously identified as responsive to sex-specific odors and/or musk-like odors. SBT, 2-sec-butyl-4,5-dihydrothiazole; MTMT, (methylthio)methanethiol; †, ^,; ‡, ; #, ; §, ; ¶, . Error bars: 95% confidence intervals.

Figure 1–figure supplement 1.

scRNA-seq analyses of the open and closed sides of whole OEs from UNO-treated adolescent male mice reveal greater quantities of newborn OSNs of musk-responsive subtypes on the open side of the OE relative to the closed. A. Experimental conditions used to generate scRNA-seq datasets for assessing the effects of UNO on quantities of newborn OSNs of specific subtypes in the OEs of adolescent male mice. B. t-distributed stochastic neighbor embedding (t-SNE) representations of all cells within scRNA-seq datasets corresponding to the open (left) and closed (right) sides of a male mouse OE (OE 1; generated in a previous study ), with Omp (mOSNs; top) and Gap43 (iOSNS; bottom) expression shown. OE 1 datasets were generated as outlined in Figure 1 (UNO-treated at PD 14, weaned sex-separated at P21, and sacrificed at PD 28). C. Enlarged view of the region of the t-SNE plot outlined in panel B, with OSNs of known musk-responsive subtypes indicated by colored dots, and immature (Gap43+) OSNs indicated by black arrows. D, E. Quantifications of immature (Gap43+) OSNs of musk-responsive subtypes as percentages of all OSNs (D) or all cells (E) within the OE 1 and OE 2 (generated in this study) datasets (lines), with mean values for the two datasets indicated (bars). In the OE 1 dataset, OSNs represent 943 of 4501 (21%) and 1573 of 7990 (20%) of all cells on the closed and open sides, respectively. In the OE 2 dataset, OSNs represent 8271 of 11919 (66%) and 9649 of 13101 (71%) of all cells on the closed and open sides, respectively.

Figure 1–figure supplement 2.

Individual OSNs of musk-responsive subtypes on the open and closed sides of OEs from UNO-treated male mice exhibit typical OR expression within cells of the OSN lineage. A-F. Cellular OR transcript levels (normalized to all transcripts within each cell) within INP3 (horizontal dark yellow bars), iOSN (horizontal bright green bars), and mOSN (horizontal dark blue bars) cells that contain transcripts encoding the musk responsive ORs Olfr235 (A, B) and Olfr1440 (C, D), and the randomly selected OR Olfr701 (Or2ag2b) from scRNA-seq datasets corresponding to 2 different male mice (OE 1, generated in a previous study , and OE 2, this study). As observed in previous studies ^,–, INP3 cells exhibit low levels of polygenic OR transcripts, while iOSNs and mOSNs exhibit high levels of monogenic OR transcripts. Moreover, OSN lineage cells of the same subtype and stage of maturity exhibit similar OR transcript levels and monogenicity within the open (light yellow bars) and closed sides (gray bars) of the OEs. Note: Olfr235 transcripts were not identified within INP3 cells of either dataset.

Figure 2–figure supplement 1.

EdU birthdating analyses show that UNO-treated adolescent male mice do not exhibit significantly different quantities of newborn OSNs of control subtype Olfr1463 on the open side of the OE relative to the closed. OE tissue samples used for assessing the effects of UNO on quantities of newborn OSNs of subtype Olfr1463 were generated as outlined in Figure 2. OEs were sectioned and analyzed using Olfr1463-specific RNA-FISH and EdU staining. Left: Representative image of OE sections from a UNO-treated adolescent male mouse that was exposed, at the time of EdU labeling, to male littermates (♂ → ♂), with newborn OSNs (OR+/EdU+) indicated by white arrows. Right: Quantifications of newborn OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated male mice reveal no significant open-side bias in quantities of newborn OSNs of subtype Olfr1463. Scale bar: 150 μm. Each line represents a distinct mouse (n = 4 mice [≥ 5 sections/mouse]). n.s. P > 0.05; ratio paired two-tailed t-test. Error bars: SEM.

Figure 2–figure supplement 2.

Comparison of normalization methods for assessing the effects of exposure to male odors on quantities of newborn Olfr235 OSNs in the OEs of UNO-treated male mice. A. Representative image corresponding to an EdU- and DAPI-stained OE section from an adolescent male mouse that was generated as outlined in Figure 2. EdU+ nuclei (top) and DAPI+ regions (bottom) are outlined in white, with medium (middle), and high magnification (right) views corresponding to the boxed regions within the preceding images. Scale bars: 150 μm. B. Experimental conditions used to generate OE tissues from UNO-treated adolescent male mice. C. Quantifications of newborn Olfr235 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs. OE sections were analyzed using Olfr235-specific RNA-FISH and EdU staining and newborn Olfr235 OSN quantities were normalized by half-section, number of EdU+ cells, or DAPI+ area. Each of the normalization methods yielded open-side biases of similar magnitude in quantities of newborn Olfr235 OSNs (left), with no significant differences in corresponding UNO effect sizes observed (right). Each line or circle represents a distinct mouse (n = 7–10 mice [≥ 5 sections/mouse] per condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P> 0.05; ratio paired two-tailed t-test (C, left); one-way ANOVA test, FDR-adjusted (C, right). Error bars: SEM. Data for newborn OSN quantities normalized by half-section correspond to Figure 2.

Figure 3–figure supplement 1.

Effects of exposure to male odors on differences in quantities of newborn and total OSNs of musk-responsive and control subtypes on the open and closed sides of the OEs of UNO-treated mice. A. Experimental conditions used for tissue generation. OE tissue samples were generated according to the experimental timeline outlined in Figure 3, sectioned, and analyzed using OR-specific RNA-FISH and EdU staining. B. Quantifications of total (OR+) Olfr235 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated mice reveal significant open-side biases in males and in females exposed to male littermates (♂ → ♂ or ♀ → ♂) but not in females exposed to just female littermates (♀ → ♀). C-F. Quantifications of Olfr1440 (C, D) and Olfr1431 (E, F) OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated mice reveal significant open-side biases (left) in quantities of newborn (OR+/EdU+; C, E) and total (D, F) OSNs of both subtypes under all odor conditions (♂ → ♂, ♀ → ♂, ♀ → ♀), with no significant differences in UNO effect sizes (right) observed. G, H. Quantifications of Olfr912 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated mice reveal no significant open-side biases in quantities of newborn OSNs (G), but significant closed-side biases in total OSNs of this subtype (H) under all 3 conditions. Observed effects of UNO on total Olfr912 OSNs may be attributable to an increased rate of cell death for mature OSNs of this subtype due to overstimulation in the presence of male odors ^,, to which all mice were exposed during at least part of the UNO period. I, J. Quantifications of Olfr1463 OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs from UNO-treated mice reveal no significant open-side biases in quantities of newborn (I) or total (J) OSNs of this subtype under any of the conditions. Each line or circle represents a distinct mouse (n = 4–10 mice [≥ 5 sections/mouse] per OSN subtype and condition). ****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P > 0.05; ratio paired two-tailed t-test (B, C-J, left); one-way ANOVA test, FDR-adjusted (C-J, right). Error bars: SEM. Data for newborn OSN quantities in ♂ → ♂ samples (C, E, G, I) correspond to Figure 2 and Figure 2-figure supplement 1.

Figure 3–figure supplement 2.

Comparison of normalization methods for assessing the effects of exposure to male odors on quantities of newborn Olfr235 OSNs in the OEs of non-occluded mice. A. Experimental conditions used to generate tissues. B. Quantifications of newborn Olfr235 OSNs within tissue sections spanning the anterior-posterior lengths of the OEs of non-occluded adolescent male and female mice that were weaned sex-separated. Newborn Olfr235 OSN quantities were normalized by half-section, number of EdU+ cells, or DAPI+ area. Each of the 3 normalization methods yielded differences of similar magnitude in quantities of newborn Olfr235 OSNs in non-occluded males exposed to male littermates (♂ → ♂) compared to females exposed to female littermates (♀ → ♀), with no significant differences observed in the corresponding odor exposure effect sizes (right; calculated from the ratios of normalized newborn Olfr235 OSNs within individual sex-separated male mice relative to the mean of normalized newborn Olfr235 OSNs within sex-separated females). Each line or circle represents a distinct mouse (n = 7–10 mice [≥ 5 sections/mouse] per condition). **P < 0.01; *P < 0.05; n.s. P> 0.05; ratio paired two-tailed t-test (B, left); one-way ANOVA test, FDR-adjusted (B, right). Error bars: SEM. Data for newborn OSN quantities normalized by half-section (B) correspond to Figure 3G.

Figure 3–figure supplement 3.

UNO-induced open-side biases in quantities of newborn Olfr1431 OSNs are increased by exposure to adult mice. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of exposure to adult mouse odors on quantities of newborn OSNs of specific subtypes. Male mice were UNO-treated at PD 14, weaned either sex-separated (♂ → ♂) or kept with parents (♂ → adult ♀ + ♂) at PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C, H. Representative images of OE sections from UNO-treated adolescent male mice that were exposed, at the time of EdU labeling, to their parents, with newborn Olfr1431 (C) or Olfr912 OSNs (OR+/EdU+) indicated by white arrows. D-G, I. Quantifications of newborn Olfr1431 (D), Olfr235 (E), Olfr1440 (F), Olfr1437 (G), and Olfr912 (I) OSNs on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs of UNO-treated mice that were exposed, at the time of EdU labeling, to just their male littermates (♂ → ♂) or also their parents (♂ → adult ♀ + ♂). Greater open-side biases (left) and corresponding UNO effect sizes (right) were observed for quantities of newborn OSNs of subtype Olfr1431 (D) in the OEs of UNO-treated male mice that were exposed to their parents compared to just littermates, while other subtypes tested were not significantly affected. Scale bars: 150 μm. Each line or circle represents a distinct mouse (n = 3–10 mice [≥ 5 sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P> 0.05; n.t. not tested (n < 4); ratio paired two-tailed t-test (D-G, I, left); unpaired two-tailed t-test (D-G, I, right). Error bars: SEM. Data for ♂ → ♂ samples correspond to Figure 2.

Figure 4–figure supplement 1.

Exposure of UNO-treated female mice to muscone causes concentration-dependent differences in quantities of newborn OSNs of musk responsive subtypes on the open and closed sides of the OE. A–D. Left: Experimental conditions used to generate OE tissue samples for assessing the effects of exposure to muscone on quantities of newborn OSNs of specific subtypes. OE tissue samples were generated according to the experimental timeline outlined in Figure 4, sectioned, and analyzed using OR-specific RNA-FISH and EdU staining. Middle, right: Quantifications of newborn (OR+/EdU+) OSNs of the musk-responsive subtypes Olfr1440 (A) and Olfr1431 (B), and the control subtypes Olfr912 (C) and Olfr1463 (D) on the open and closed sides of tissue sections spanning the anterior-posterior lengths of the OEs of UNO-treated female mice that were exposed, at the time of EdU labeling, to 0, 0.1, 1, or 10% muscone. Each line or circle represents a distinct mouse (n = 4–8 mice [≥ 5 sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P> 0.05; ratio paired two-tailed t-test (middle); unpaired two-tailed t-test (right). Error bars: SEM. Data for ♀ → 0% muscone samples correspond to Figure 3–figure supplement 1.

Figure 4–figure supplement 2.

Muscone exposure induces concentration-dependent differences in quantities of total OSNs of musk responsive subtypes on the open and closed sides of the OEs of UNO-treated mice. A–E. Left: Experimental conditions used to generate OE tissue samples for assessing the effects of exposure to muscone on quantities of total OSNs of specific subtypes. OE tissue samples were generated according to the experimental timeline outlined in Figure 4, sectioned, and analyzed using OR-specific RNA-FISH and EdU staining. Middle, right: Quantifications of total (OR+) OSNs of the musk-responsive OSN subtypes Olfr235 (A), Olfr1440 (B) and Olfr1431 (C), and the control subtypes Olfr912 (D) and Olfr1463 (E) on the open and closed sides of tissue sections spanning the anterior-posterior lengths of the OEs of UNO-treated female mice that were exposed, at the time of EdU labeling, to 0, 0.1, 1, or 10% muscone. Note that the reduced quantities of total OSNs of subtype Olfr912 may be attributable to a reduced rate of survival for mature Olfr912 OSNs in the presence of male odors ^,, to which mice were exposed prior to weaning. Each line or circle represents a distinct mouse (n = 4–8 mice [≥ 5 sections/mouse] per OSN subtype and condition). **P < 0.01; *P < 0.05; n.s. P> 0.05; two-tailed paired t-test (middle); two-tailed unpaired t-test (right). Error bars: SEM.

Figure 4–figure supplement 3.

Comparison of normalization methods for assessing the effects of exposure to a musk odor on subtype-specific newborn OSN quantities in the OEs of non-occluded mice. A. Experimental conditions used to generate OE tissue samples. Samples were generated according to the experimental timeline outlined in Figure 4, sectioned, and analyzed using OR-specific RNA-FISH and EdU staining. B. Quantifications of newborn Olfr235 OSNs (OR+/EdU+) within tissue sections spanning the anterior-posterior lengths of the OEs of non-occluded female mice that were either exposed to 0.1% muscone (♀ → muscone) or unexposed to an exogenous odorant (♀ → ♀) starting from PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OE sections were analyzed using Olfr235-specific RNA-FISH and EdU staining and newborn Olfr235 OSN quantities were normalized by half-section, number of EdU+ cells, or DAPI+ area. Each of the normalization methods yielded significant differences in quantities of newborn Olfr235 OSNs in muscone-exposed compared to unexposed mice (B-left), with no significant differences observed in the corresponding odor exposure effect sizes (B-right; calculated from the ratios of normalized newborn Olfr235 OSNs within individual odor-exposed mice relative to the mean of normalized newborn Olfr235 OSNs within control mice). Each circle represents a distinct mouse (n = 5–10 mice [≥ 5 OE sections/mouse] per condition). ***P < 0.001; n.s. P> 0.05; unpaired two-tailed t-test (B-left); one-way ANOVA test, FDR-adjusted (B-right). Error bars: SEM. Data for newborn OSN quantities normalized by half-section correspond to Figure 4D.

Figure 4–figure supplement 4.

Effects of exposure of mice to musk and non-musk odors on quantities of newborn OSNs of non-musk-responsive subtypes, including those previously found to undergo stimulation-dependent changes in OSN birthrate. A, B. Experimental timeline (A) and conditions (B) used to generate OE tissue samples for assessing the effects of exposure to musk (muscone, ambretone) or non-musk (SBT) odors on quantities of newborn OSNs of specific subtypes. Non-occluded female mice were weaned sex-separated and either exposed or unexposed to an exogenous musk or non-musk odorant starting at PD 21, EdU-labeled at PD 28, and sacrificed at PD 35. OEs were sectioned and analyzed using OR-specific RNA-FISH and EdU staining. C. Left: Quantifications of newborn OSNs of the SBT-responsive subtype Olfr912 within tissue sections spanning the anterior-posterior lengths of OEs reveal no significant differences in non-occluded females that were exposed, at the time of EdU labeling, to a musk odorant (♀ → muscone; ♀ → ambretone) or to SBT (♀ → SBT) compared to those exposed to just female littermates (♀ → ♀). Middle, right: Quantifications of newborn OSNs of two subtypes previously found to undergo stimulation-dependent changes in birthrates, Olfr827 (middle) and Olfr1325 (right) , reveal no significant differences in non-occluded females that were exposed to a musk odorant compared to those exposed to just female littermates. Each circle represents a distinct mouse (n = 3–4 mice [≥ 5 OE sections/mouse] per OSN subtype and condition). n.s. P> 0.05; one-way ANOVA test, FDR-adjusted (C-left); unpaired two-tailed t-test (C-middle, right). Error bars: SEM. Data for newborn Olfr912 OSNs in ♀ → ♀ samples correspond to Figure 3H.

Figure 6–figure supplement 1.

Stimulation-dependent increases in quantities of newborn OSNs of musk-responsive subtypes are stable over time following neurogenesis, consistent with a mechanism involving altered birthrate. OE tissue samples were generated according to the experimental timeline and conditions outlined in Figure 6, sectioned, and analyzed using OR-specific RNA-FISH and EdU staining. A, B, D–G. Quantifications of newborn Olfr1440 (A), Olfr1431 (B), Olfr912 (D, F), Olfr1463 (E), and Olfr235 (G) OSNs (OR+/EdU+) on the open and closed sides of tissue sections spanning the anterior-posterior lengths of OEs of UNO-treated males exposed to male littermates (♂ → ♂) (A, B, D, E), females exposed to muscone (♀ → 0.1% muscone) (F), or females exposed to female littermates (♀ → ♀) (G). Under all conditions, similar open-side biases in quantities of newborn OSNs were observed at both 4 and 7 d post-EdU, with no significant differences in UNO effect sizes observed over time. C. Representative image of OE sections from a UNO-treated female mouse that was exposed to muscone (♀ → 0.1% muscone) and sacrificed 4 d post-EdU, with newborn Olfr235 OSNs (OR+/EdU+) indicated by white arrows. Greater numbers of newborn Olfr235 OSNs were observed on the open side compared to the closed side at this timepoint. Scale bar: 150 μm. H. Quantifications of newborn OSNs of the SBT-responsive subtype Olfr912 within OEs of non-occluded female mice that were either exposed just to female littermates (♀ → ♀) or also to muscone (♀ → 0.1% muscone) and sacrificed 4 or 7 d post-EdU. No significant changes in newborn OSN quantity differences were observed between 4 and 7 d post-EdU. Each line or circle represents a distinct mouse (n = 2–10 mice [≥ 5 OE sections/mouse] per OSN subtype and condition). ***P < 0.001; **P < 0.01; *P < 0.05; n.s. P > 0.05; n.t. not tested (n < 4); ratio paired two-tailed t-test (A, B, D–G-left); unpaired two-tailed t-test (A, B, D–G-right); two sample ANOVA - fixed-test, using F distribution (right-tailed; H). Error bars: SEM. Data for 7 d post-EdU samples correspond to Figures 2–4.

Appendix 2–figure 1.

Gas chromatography – mass spectrometry (GC-MS) analyses of mouse preputial gland extracts for molecules from with structural similarity to known musk odorants. A, B. GC-MS signals from male mouse preputial gland extracts (A) and commercially available samples of 4 known musk compounds (B). Analyses of preputial glands revealed signals (Unknowns 1–3) with predicted potential matches to the indicated odorants based on spectral similarities (A). Experimental differences in retention times between unknowns and standards are indicated. C. Left: Region of the extracted ion chromatograph (m/z 236) of a preputial gland extract, with the signal corresponding to Unknown 1 highlighted in red. Right: Mass spectra corresponding to Unknown 1 (red) and a predicted match, 8-cyclohexadecenone (blue), a musk compound that was previously found to activate Olfr235 and Olfr1440 . The retention times of Unknown 1 and 8-cyclohexadecenone differ by 0.06 and 0.11 minutes (possibly corresponding to the cis and trans isomers) (A), indicating potential structural similarity. D. Left: Region of the extracted ion chromatograph (m/z 236) of a preputial gland extract, with the signal corresponding to Unknown 2 highlighted in red. Right: Mass spectra corresponding to Unknown 2 (red) and a predicted match, cycloheptadecanol (blue). Confirmation of a match will require comparison of the observed retention time of Unknown 2 (12.27 minutes) with that of a cycloheptadecanol standard (not determined). Although Unknown 3 exhibited a predicted match to cyclopentadecanol, a musk compound previously found to activate Olfr235 and Olfr1440 , the observed retention time difference of 4.48 minutes (A) indicates substantial structural dissimilarity.

Appendix 4–figure 1.

Quantile-quantile (QQ) plots for comparison of actual data to a theoretically normal distribution. A. Analysis of quantities of newborn OSNs on the open and closed sides of the OEs of UNO-treated mice. B. Analysis of quantities of total OSNs on the open and closed sides of the OEs of UNO-treated mice. C. Analysis of quantities of newborn OSNs within the OEs of non-occluded mice. D. Analysis of UNO effect sizes of newborn and total OSNs on the open and closed sides of the OEs of UNO-treated mice. In each plot, predicted values (assuming sampling from a Gaussian distribution) are plotted on the vertical axis, while actual values are plotted on the horizontal axis. Points generally follow the line of identity, indicating that the data reflect a Gaussian (normal) distribution.