Age-Dependent Activation and Neuronal Differentiation of Lgr5+ Basal Cells in Injured Olfactory Epithelium via Notch Signaling Pathway

Aging is an important factor affecting function of smell, leading to the degeneration of mature olfactory sensory neurons and inducing the occurrence of smell loss. The mammalian olfactory epithelium (OE) can regenerate when subjected to chemical assaults. However, this capacity is not limitless. Inactivation of globose basal cells and failure to generate sensory neurons are the main obstacles to prevent the OE regeneration. Here, we found the significant attenuation in mature sensory neuronal generation and apparent transcriptional alternation in the OE from aged mice compared with young ones. The recruitment of leucine-rich repeat-containing G-protein coupled receptor 5 (Lgr5)-positive cells in injured OE was weakened in aged mice, and more Lgr5+ cells remained quiescence in aged OE postinjury. Lineage-traced progenies from Lgr5+ cells were significantly fewer in the OE with aging. Moreover, Notch activation enhanced the neuronal regeneration in aged OE, making the regenerative capacity of aged OE comparable with that of young animals after injury. The growth and morphology of three-dimensional (3D)-cultured organoids from the OE of young and aged mice varied and was modulated by small molecules regulating the Notch signaling pathway. Thus, we concluded that activation of Lgr5+ cells in injured OE was age dependent and Notch activation could enhance the capacity of neuronal generation from Lgr5+ cells in aged OE after injury.


INTRODUCTION
The mammalian olfactory epithelium (OE) is a neuroepithelial structure that has self-renewal capacity throughout life. Horizontal basal cells (HBCs) and globose basal cells (GBCs) are two types of stem/progenitor cell populations responsible for the regeneration of injured OE (Jang et al., 2003;Chen et al., 2004;Schnittke et al., 2015;Schwob et al., 2017).
When subjected to chemical assaults, dormant stem cells (mainly HBCs) are recruited to generate GBCs and then differentiate into olfactory sensory neurons, sustentacular cells, and other cellular subtypes constituting the OE (Leung et al., 2007). Acute injury drives proliferating cells to differentiate into immature neurons expressing Tuj1 and GAP43, and then grow into mature neurons expressing olfactory marker protein (OMP) and PGP9.5. However, the neuro-regenerative capacity of basal cells in the OE is not limitless and aging is a principal element causing the sensory neuronal death and inactivation of GBCs (Child et al., 2018). Studies on both human and laboratory animals such as mice have shown age-related morphology and functional changes in the olfactory nerve epithelium (Mobley et al., 2014). Age-associated decline in the neuro-regenerative capacity of the injured OE is related with the weakness in proliferative activity (Suzukawa et al., 2011). Spontaneous lesions occur in the neuroepithelium and Bowman's glands in mouse olfactory mucosa with progression of aging (Kondo et al., 2009). Several biomarkers have been identified as hallmarks of the aged olfactory mucosa, such as low expression of extracellular matrix genes (Ueha et al., 2018b) and ApoE deficiency (Zhang et al., 2018). With the increase of age, the recovery rate slows down and respiratory metaplasia may occur (Suzukawa et al., 2011;Child et al., 2018). Strategies against age-related neuronal degradation have been reported, including insulin-like growth factor 1 (IGF-1) administration (Ueha et al., 2018a), intranasal treatment with fibroblast growth factor-2 (FGF-2; Fukuda et al., 2018), activation of inositol trisphosphate receptor type 3 (IP3R3), and the neuroproliferative factor neuropeptide Y (NPY) signaling (Jia and Hegg, 2015).
The Notch signaling pathway plays important roles in the entire organism development process and in maintaining tissue self-renewal (Artavanis-Tsakonas and Muskavitch, 2010). Activation of Notch signaling transduction leads to proteolysis of the Notch receptor, allowing intracellular domain of Notch (NICD) to enter the nucleus, and subsequently binds to DNA-binding proteins and assembles transcription complexes that activate corresponding downstream target genes (Kopan and Ilagan, 2009). Notch expression is present in the neurogenesis of the peripheral olfactory system (Doi et al., 2004), and various Notch subtypes play different roles in determining neuronal and glial lineage during development of the OE (Carson et al., 2006). Notch signaling is necessary to direct the neuronal differentiation in the OE regeneration , and Notch1 is required to maintain dormancy of reserve HBCs . Our previous work indicated that Notch1 activation enhanced the proliferation in leucine-rich repeat-containing G-protein coupled receptor 5 (Lgr5)-positive progenitor cells and regulated the generation of mature sensory neurons in the OE (Dai et al., 2018). Thus, it is of interest to elucidate whether the effect of aging on the OE regeneration will be counteracted via Notch signaling activation.
In this study, we elucidated the transcriptional change in the OE from young and aged mice. Aging attenuated the recruitment of Lgr5+ cells in injured OE. Previous work reported that Lgr5 marked GBCs in the OE (Chen et al., 2014) and Lgr5+/Notch1+ cells participated in the OE regeneration postinjury (Dai et al., 2018). Based on these findings, here, we showed that aging altered the stemness of Lgr5+ cells in injured OE, keeping more Lgr5+ cells in dormancy. In aged OE, the capacity of neuronal generation from Lgr5+ cells were weakened, while Notch activation mimicking by NICD overexpression offset aging-induced intervention on neuronal regeneration in injured OE. In vitro-cultured OE organoids from aged mice showed more apparent cystic morphology compared with those from young mice, which was altered by chemical treatment regulating the Notch signaling. Thus, these data supported the conclusion that aging was a critical factor determining the participation of Lgr5+ cells in OE regeneration, and the adverse effect of aging on neuronal differentiation was counteracted by Notch activation. This study supplemented new evidence to elucidate the regulatory role of Notch signaling in aging and OE regeneration.

Animals
Lgr5-EGFP-IRES-CreERT2 (#008875, harboring a ''knock-in'' allele that abolishes Lgr5 gene function and expresses EGFP and CreERT2 fusion protein from the Lgr5 promoter/enhancer elements), Rosa26-TdTomato (#007914, a Cre reporter strain with a loxP-flanked STOP cassette prevents transcription of the downstream red fluorescent protein), and RosaNICD (#008159, containing a sequence encoding an intracellular portion of the mouse Notch1 gene with a loxP-flanked STOP cassette, leading to elevating intracellular signal transduction and mimicking Notch signaling pathway activation by Cre reporter) mice were purchased from the Jackson Laboratory. Mice were housed with open access to food and water at the Experimental Animal Center of Eye, Ear, Nose, and Throat Hospital, Fudan University (Shanghai, China). The animals were housed in a temperature-controlled environment (23 ± 2 • C) with a light/dark cycle of 12 h. All experiments were approved by the Shanghai Medical Experimental Animal Administrative Committee (Permit Number: 2009-0082). Both male and female mice were used in this study, and the data were grouped together because no sex difference was evident. All animals were maintained at C57BL/6J background.

Chemical Treatment
All the reagents were purchased from Sigma Aldrich unless specified. The final concentration of chemicals used in organoid culture was: 5 µM LY411575 (#SML0506) and 40 ng/ml Jagged 1 (aa 1-1067, #SRP8012). LY411575 was dissolved in sterilefiltered DMSO (#D2650) while Jagged 1 was dissolved in sterile phosphate-buffered saline (PBS) to make stock solution. Chemicals were diluted by 1,000 times into the cultures. Organoids were treated with chemicals on Day 3 after passaging and incubated with chemicals for 4 days.

Lineage Tracing and Notch Activation Mimicking
For lineage tracing and Notch activation mimicking in injured OE, 220 µg/g (body weight) tamoxifen (#T5648) was injected subcutaneously into LT and LTN mice for respective three consecutive days just prior to and after methimazole administration (the scheme is shown as Figure 6A). In the control group, mice were received with same doses of sunflower seed oil (#S5007), served as a solvent to tamoxifen. For lineage tracing in LT mice with different ages including 1-, 3-, 12-, and 24-month old, three doses of tamoxifen (1 dose/day) was injected prior to the methimazole treatment. The scheme for lineage tracing was shown as Figure 5A. For lineage tracing, administration of tamoxifen induced the generation of Cre, removed the STOP cassette, and induced the expression of TdTomato. Thus, all TdTomato+ cells were Lgr5+ cells or progenies derived from Lgr5+ cells. For Notch activation mimicking, Cre induced by tamoxifen administration led to the expression of NICD and then activated Notch signaling pathway.

RNA Sequence
The protocol for RNA sequence was described previously (Ren et al., 2020). RNA was prepared from mouse olfactory mucosa using TRIZOL reagent. RNA-Seq analysis was performed by Majorbio Corp. (Shanghai, China). Briefly, sequencing reads were mapped to the mouse genome using HISAT2. Transcriptome from RNA-seq reads was reconstructed by StringTie, and expression differences were evaluated using DESeq2. Pearson's coefficient was calculated to determine the correlation among different groups. The clustering analysis of the global gene expression pattern in different samples was carried out using K-means clustering algorithm by RSEM software. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed at http://www. geneontology.org/ and http://www.genome.jp/kegg/. I-Sanger (www.i-sanger.com) was used to analyze all the sequence data.

Statistical Analysis
Each experiment was repeated in triplicate independently. Quantitative data were expressed as mean ± SEM. The evaluators involved in assessing organoid size and cell counting were blinded to the experimental design including age and genotypes of the animals from which the specimens were derived. Organoid counting for the ratio of cystic and filled appearance were based on the images captured under bright field. The number of positively stained cells (including OMP+, Tuj1+, Krt19+, Krt14+, Sox2+, Lgr5-EGFP+, Lgr5+/Sox2+, Lgr5+/Ki67+, Lgr5+/Mash1+, Lgr5+/ICAM1+, TdTomato+, PGP9.5+, TdTomato+ (TdT+)/PGP9.5+, and TdT+/Sox2+ cells), organoid size, and OE thickness were measured by Image J Software. All the cell counting data were derived from the number of cells on six OE sections per mice. We biologically triplicated each experiment, thus 18 sections from three animals were used for each cell counting analysis. Each slide had OE sections from anterior to posterior zones. On each OE section, six different regions (three in dorsal and three in ventral) were selected to measure the cell density. These selected regions were consistent among all groups. The density of positively stained cells was defined as the number of positively stained cells per 100 µm or 0.5 mm OE. The density of TdT+ cells in Figure 5 was defined as the number of TdT+ cells per OE section. The OE thickness was averaged from five different places on each OE section. Two-way ANOVA with Sidak's multiple comparisons test and Tukey's multiple comparisons test was used to determine the statistical difference among multiple groups, while comparison between two groups was performed by unpaired t-test using GraphPad Prism 6.0 software. p values < 0.05 were considered statistically significant.

Aging Attenuates Sensory Neuronal Generation in the OE
To elucidate the effect of aging on the OE, we performed immunostaining against various biomarkers for cellular subtypes constituting the OE in wild-type C57BL/6J mice at 3-and 25-month old. The intensity of staining against mature sensory neuronal markers OMP was weaker in the OE of aged animals compared with the young mice (Figures 1A,C,D, differential OMP+ staining signals were noted by dash lines in Figure 1A). The density of OMP+ cells was decreased by 28.0 ± 4.4% in aged OE compared with the young one ( Figure 1M, p = 0.0072, n = 3 mice). The intensity of staining against immature neuronal marker TUJ1 did not show apparent difference (Figures 1A,C,D), while the density of TUJ1+ cells in aged OE was still significantly lowered by 18.3 ± 4.9% than the OE of young mice ( Figure 1M, p = 0.0213, n = 3 mice). Besides, the intensity of PGP9.5+ staining signals in the OE of young mice was more intensive than the old animals ( Figures 1B,E,F, PGP9.5+ staining signals were labeled by asterisks in Figure 1B). By contrast, staining against Krt19, a biomarker for columnar ciliated respiratory epithelial cells, was more obvious in the OE of aged animals compared with the young mice ( Figure 1B, noted by dash lines, Figures 1G,H, noted by arrows). The density of Krt19+ cells in the OE from aged mice was significantly increased by 2.3 ± 0.3-fold, compared with the young OE ( Figure 1M, p < 0.001). However, we did not observe the apparent change in apical Sox2+ staining (marker for sustentacular cells, Figures 1I,J) or in Krt14+ staining (marker for HBCs, Figures 1K,L). The density of apical Sox2+ sustentacular cells or Krt14+ basal cells was not significantly changed in the OE between aged and young animals ( Figure 1M, p = 0.3769 for Krt14+ cells and p = 0.6724 for apical Sox2+ cells, n = 3). The OE thickness of aged mice was not significantly different from the young animals ( Figure 1N, p = 0.3151). Thus, we concluded that aging led to the attenuation in mature sensory neuronal generation in the OE.

Transcriptional Alteration in Aged OE
To elucidate the panoramic effect of aging on the OE, we performed transcriptional analysis between the OE from aged and young animals at 3-and 25-month old, respectively. Totally, 1,342 upregulated and 1,005 downregulated genes were found in aged OE. By KEGG enrichment analysis, we found five upregulated and 265 downregulated genes associated with olfactory transduction in the OE of aged mice (Figure 2A). The heatmap showed that the expression level of HBC markers Krt14, Krt5, ICAM1, and respiratory epithelial cell marker Krt19 was increased in aged OE, while level of GBC markers NeuroD1, NeuroG1, Ascl1, neuronal marker OMP, Tuj1, and GAP43, as well as sustentacular cell marker Cyp2g1 was decreased (Figures 2B,C). KEGG pathway analysis found that genes involved in inflammation pathway such as NF-kappa B, PI3K-Akt, TNFα, JAK/STAT, and MAPK signaling were upregulated in aged OE ( Figure 2D). Genes participating in critical signal pathways such as Wnt, Notch, and Hippo were differentially expressed between the OE from young and aged mice ( Figure 2D). Thus, we provided a transcriptional landscape showing the alteration in the OE with aging.

Aging Leads to the Attenuation in Recruitment of Lgr5+ Cells in Injured OE
Lgr5 marks GBCs and Lgr5+ GBCs are recruited in injured OE (Chen et al., 2014;Dai et al., 2018). Since GBC markers such as Ascl1, NeuroD1, and NeuroG1 at transcriptional level were decreased in aged OE (Figures 2B,C), we determined whether aging weakened the recruitment of Lgr5+ GBCs in lesioned OE. Lgr5-EGFP-CreERT2 mice at 1-and 12-month old were injected with methimazole, and the density of Lgr5-EGFP+ cells was measured postlesion. The recruitment of Lgr5-EGFP+ cells in injured OE was not significantly different at Day 3 postinjury between 1-and 12-month-old mice (Figures 3A,B) but significantly retarded at Day 5 in aged OE (Figures 3C,D). Statistically, the density of Lgr5-EGFP+ cells was significantly increased in injured OE of both 1-and 12-month-old mice at Day 3 postlesion compared with the unlesioned controls ( Figure 3M, p < 0.001 by unpaired t-test, n = 3). However, the density of Lgr5-EGFP+ cells was still drastically enhanced at Day 5 in young animals (p < 0.001 compared with Day 3, n = 3) while the density was reduced in the OE of 12-month-old mice at Day 5 compared with the density at Day 3 postinjury and not significantly changed in contrast to the saline control (Figure 3M). At Days 10, 17, and 32 postinjury, the density of Lgr5-EGFP+ cells at basal cell layer in the OE of both 1-and 12-month-old mice continued to decrease in contrast to the density at Day 5 and returned to the threshold level similar as the density of Lgr5-EGFP+ cells in the saline control (Figures 3E-M). Compared with the OE of young mice, the density of Lgr5-EGFP+ cells in aged OE at Days 5 and 32 postinjury was significantly decreased (Figure 3M, p < 0.001 at Day 5 and p < 0.05 at Day 32 using two-way ANOVA with Sidak's multiple comparisons test). Surprisingly, the Lgr5-EGFP+ cells appeared in the nonbasal cell layer during the  Figures 3E -J ). The density of nonbasal Lgr5-EGFP+ cells was significantly increased at Days 10, 17, and 31 postinjury in the OE of 12-month-old mice, compared with the density in unlesioned controls ( Figure 3N, p < 0.001 by unpaired t-test, n = 3), while the density of Lgr5+ nonbasal cells was dramatically increased at Days 17 and 31 postinjury in the OE of young animals ( Figure 3N, p < 0.001, n = 3). Compared with the young mice, the density of nonbasal Lgr5-EGFP+ cells in the aged OE was significantly increased at Days 10 and 17 postinjury (Figure 3N, p < 0.01 at Day 10 and p < 0.001 at Day 17 using two-way ANOVA with Sidak's multiple comparisons test). Collectively, these data demonstrated that aging weakened the recruitment of Lgr5+ basal cells in injured OE.
of Lgr5+/Sox2+ basal cells (Figures 4A,B), Lgr5+/Ki67+ proliferative cells (Figures 4E,F), and Lgr5+/Mash1+ GBCs (Figures 4I,J) in the OE of 1-month-old mice. By contrast, Lgr5+/Ki67+ and Lgr5+/Mash1+ cells were only recruited at Day 3 in injured OE of 12-month-old mice, but their densities was significantly decreased compared with the counterparts in young mice at Day 3 (Figures 4E,G,I 4H,L,Q). However, Lgr5+/Sox2+ cells were apparently activated at Day 3 and still present at Day 5 postinjury in the OE of 12-month-old mice (Figures 4C,D). In 1-month-old mice, Lgr5+ cells were not co-immunostained with ICAM1, a biomarker for HBCs, at Days 3 and 5 postlesion (Figures 4M,N, squares were magnified as Figures 4M ,N ). By comparison, Lgr5+/ICAM1+ cells were found in the OE of 12-month-old mice at Days 3 and 5 postinjury (Figures 4O,P, labeled by arrowheads in magnified squares as Figures 4O ,P ). The density of Lgr5+/ICAM1+ cells was significantly increased in injured OE of 12-month-old mice at Days 3 and 5, compared with the density in the OE of 1-monthold mice (Figure 4Q, p < 0.001 by two-way ANOVA with Tukey's multiple comparisons test). This showed a portion of Lgr5+ cells remained in dormancy in injured OE of aged mice. Collectively, we concluded that aging led to the alteration in stemness of injury-recruited Lgr5+ cells in the OE, keeping some Lgr5+ cells in dormancy after OE injury.

Aging Weakens Progeny Generation From Lgr5+ Cells in Injured OE
Since Lgr5+ cells in the OE from young and aged animals were differentially recruited postinjury, we then determined whether the progeny generation from Lgr5+ cells in injured OE was affected by aging. Lineage tracing of Lgr5+ cells in LT mice through tamoxifen administration ( Figure 5A) showed the TdT+ cell bundles (progenies from Lgr5+ cells) were obvious in the OE from mice at 1-and 3-month old (Figures 5B,C, noted by asterisks). In mice at 12-month old, single TdT+ cells were scattered throughout the OE while the cell bundles were seldom observed ( Figure 5D). However, fewer TdT+ cells were present in injured OE of LT mice at 24-month old ( Figure 5E). Statistical analysis indicated that the density of TdT+ cells (number of TdT+ cells per half OE section) was significantly lowered by 77.7 ± 14.4 and 75.3 ± 4.4% in injured OE of 12-and 24month-old mice compared with that in the OE of 1-month-old animals ( Figure 5F, p = 0.0014 and p = 0.0002 by unpaired t-test). Similarly, the density of TdT+ cells in the OE of 12- and 24-month-old mice was decreased by 73.8 ± 16.9 and 70.9 ± 5.1% compared with the density in mice at 3-month old ( Figure 5F, p = 0.0055 and p = 0.0002). Furthermore, we observed the apparent TdT+/PGP9.5+ cells in the OE of young mice at Day 30 postinjury (Figure 5G, noted by arrowheads), while significantly fewer TdT+/PGP9.5+ cells were found in the OE of 12-month-old mice at Day 30 postinjury ( Figure 5H). Statistically, the density of TdT+/PGP9.5+ cells in LT mice at 12-month old was decreased by 51.6 ± 6.6% compared with 1-month-old mice ( Figure 5K, p = 0.0025). Meanwhile, TdT+/Sox2+ cells were present in the OE of both 1-and 12-month old at Day 30 and did not show the apparent difference ( Figures 5I-K, p = 0.8356, noted by arrowheads). Therefore, aging dampened neuronal differentiation of Lgr5+ cells in injured OE.

Notch Activation Strengthens Neuronal Regeneration in Aged OE
To determine whether the Notch signaling pathway could differentially regulate the neuronal regeneration in injured OE of young and aged mice, LTN mice at 1-and 12-month old were injected with tamoxifen to induce the overexpression of NICD ( Figure 6A). With Notch activation mimicking, we found the significant increase in TdT+ cell density by 31.8 ± 6.5% in the OE of 12-month-old LTN mice compared with the density in 12month-old LT mice (Figures 6C,E,N, p = 0.0081 by unpaired t-test). However, the TdT+ cell density was not significantly different between the OE of 1-month-old LT and LTN mice (Figures 6B,D,N, p = 0.8864). Furthermore, the TdT+ cell density in the OE of LTN mice at 12-month old was comparable with that in the OE of LT mice at 1-month old ( Figure 6N, p = 0.4518), suggesting Notch activation enhanced the progeny generation from Lgr5+ cells in the OE of aged mice to the similar level as in the young animals. Staining against PGP9.5 was denser in the OE of LTN mice at both 1-and 12-month old, compared with the counterparts in LT mice (Figures 6B-E). The density of PGP9.5+ cells was higher in LTN mice at both 1-and 12-month old by 36.8 ± 7.4 and 26.7 ± 5.9%, compared with the counterparts in LT mice ( Figure 6O, p < 0.05 by two-way ANOVA with Sidak's multiple comparisons test). By contrast, the density of apical Sox2+-supporting cells did not show a significant difference in the OE of LT and LTN mice at either 1or 12-month old ( Figure 6O, p > 0.1). The TdT+/PGP9.5+ cell density was significantly increased by 60.8 ± 14.5% (p < 0.05) and 125.3 ± 25.0% (p < 0.01) in the OE of young and aged LTN mice compared with the density in LT mice (Figures 6F-I

Notch Regulates Aging-Induced Morphological Alteration in OE Organoids
We then determined whether Notch signaling could regulate the aging-induced variation in OE organoid growth in vitro.
The organoid size was not significantly different at Day 3 postculture but drastically increased in organoids from aged mice (19-month-old) at Day 7 (Figures 7A-E, p < 0.001 by two-way ANOVA with Sidak's multiple comparisons test).
Compared with the organoids from young mice (3-monthold), percentage of organoids from aged mice with ''filled'' morphology was decreased by 23.7 ± 2.8%, while the ratio of cystic organoids was increased by 42.1 ± 2.8% (Figures 7A-D,F, p < 0.001). Then, we elucidated whether regulation of Notch signaling by small molecule chemicals in vitro could affect the organoid size and morphology. In the presence of Notch activator Jagged1, the size of organoids from aged mice was increased by 51.9 ± 5.1% compared with the organoids derived from young OE (Figures 7H,K   of Notch signaling also changed the organoid morphology. LY411575 treatment led to significant increase in ratio of filled organoids from aged OE by 52.1 ± 6.5% compared with the DMSO-treated organoids (Figures 7J,L,N, p < 0.001). This was even higher than the ratio of filled organoids in untreated and LY411575-treated cultures from young mice by 16.0 ± 5.0 and 14.1 ± 4.9% (Figures 7G,I,L,N, p < 0.001). However, no apparent morphological change was observed in LY411575-treated organoids from young OE compared with the DMSO-treated control ( Figure 7N). Collectively, these data demonstrated that regulation of Notch signaling by chemical stimulation affected aging-induced alteration in organoid size and morphology.

DISCUSSION
In this study, we reported the cellular and transcriptional alteration between young and aged OE. Aging weakened the recruitment of Lgr5+ cells, altered the stemness, and attenuated generation of neuronal progenies from Lgr5+ cells in the OE postinjury. Notch activation could recover the neuronal differentiation in aged OE, making it comparable with the OE of young animals. Therefore, this work provided the evidence to support the correlation between Notch signaling activation and alleviation in aging-induced incapacity of neuronal regeneration in injured OE, potentially providing a new concept in the therapy against aging-triggered anosmia. Lgr5+ cells are active globose basal cells and are recruited in injured OE (Chen et al., 2014;Dai et al., 2018). However, it is still not clear whether dormant Lgr5+ cells are present in the OE, and this may lead to the discrepancy in the recruitment of Lgr5+ cells when OE is exposed to toxic reagent. Previous reports showed that Lgr5+ label-retaining cells were dormant stem cells in intestine (Li et al., 2016) and Lgr5+ intestinal stem cells were in G1 phase characterized by dormant periods (Carroll et al., 2018), showing the presence of Lgr5+ quiescent stem cells. In mammary gland, the deeply quiescent Lgr5+ mammary stem cells resided within the proximal region (Fu et al., 2017). These cells remained FIGURE 7 | Chemical cocktail regulated organoid growth and morphology via Notch signaling pathway. (A-D) Images showing the morphological difference in organoids derived from the OE of young (3-month-old) and aged (19-month-old) mice at Days 3 and 7 after in vitro culture. (E) Statistical analysis on size of organoids from young and aged mice at Days 3 and 7. At Days 3 and 7, 391 and 501, 1,113, and 604 organoids from aged and young OE were measured. (F) Statistical analysis on the ratio of filled and cystic organoids from young and aged mice; 825 and 934 organoids from young and aged OE were counted. (G-L) Images showing organoids from young and aged animals in the presence of DMSO (G,J), Notch activator Jagged 1 (H,K), and Notch inhibitor LY411575 (I,L). (M) Statistical analysis on size of organoids treated with DMSO, Jagged 1, or LY411575; 518 and 415, 463 and 505, and 359 and 379 organoids from young and aged OE were measured when treated with DMSO, Jagged 1, or LY411575. (N) Statistical analysis on ratio of cystic and filled organoids from young and aged mice, treated with DMSO, Jagged 1, or LY411575; 973 and 1, 040, 518 and 717, and 683 and 300 organoids from young and aged OE were counted in the presence of DMSO, Jagged 1, and LY411575. The statistical significance was determined by two-way ANOVA with Sidak's multiple comparisons test. F (1,814) = 4.533, p = 0.0335 in (E); F (1,56) = 209.9, p < 0.0001 in (F); F (2,1,493) = 147.3, p < 0.0001 in (M). The statistical significance among different stimulation groups in (M) was determined by unpaired t-test. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001. Scale bar: 250 µm. active in embryonic mammary primordia before switching to a quiescent state after birth and can re-enter the cell cycling under ovarian hormone stimulation (Fu et al., 2017). This resembles the situation of Lgr5+ cells in the OE, which were active in early postnatal stage, remained in dormancy in adults, and were activated by chemical lesion (Chen et al., 2014;Dai et al., 2018). In aged OE, the recruitment of Lgr5+ cells was attenuated compared with that in young OE (Figure 3). This raises the possibility that more Lgr5+ cells in aged OE postinjury are still in quiescence. This was supported by the observation that Lgr5+/ICAM1+ cells were present in aged OE at Days 3 and 5 postlesion while it was seldom found in young OE (Figure 4). Thus, driving the quiescent stem cell to enter the cell cycle is applicable to alleviate aging-induced smell loss.