Construction of in vitro models derived from distinct regions of mouse nasal mucosa

Summary

The respiratory epithelium (RE) and olfactory epithelium (OE) of the nasal mucosa are structurally similar yet functionally distinct: RE primarily facilitates air conditioning and host defense, while OE is responsible for odor detection and signal transmission. Although organoids have been applied in previous studies of both epithelia, there remains considerable room for refinement and further development. Here, we isolated primary cells from the respiratory and olfactory regions of the mouse nasal mucosa and established corresponding organoids using both three-dimensional (3D) and air-liquid interface (ALI) culture methods. The resulting in vitro models exhibited distinct lineage preferences and characteristics mirroring their tissue of origin. Furthermore, we demonstrated that IL-17a significantly inhibits ciliogenesis under the RE-ALI model, an effect that could be partially rescued by IL-17a inhibition.

Graphical abstract

Highlights

• •Establishment of 3D and ALI organoids from murine RE and OE tissues
• •Cross-induction reveals differential lineage plasticity between RE and OE organoids
• •IL-17a impairs ciliary differentiation in RE organoid under ALI culture

Biological sciences; Cell biology

Introduction

The nasal cavity comprises two structurally similar yet functionally distinct regions: the respiratory region and the olfactory region. The respiratory epithelium (RE) consists of ciliated cells, brush cells, goblet cells, secretory cells, and basal cells, forming the first line of defense against external damage and carrying out essential functions such as immune defense, mucociliary clearance, and maintenance of gas exchange.^1,2,3 The olfactory epithelium (OE) is a pseudostratified columnar epithelium composed of globose basal cells (GBCs), horizontal basal cells (HBCs), olfactory sensory neurons (OSNs), microvilli cells (MVCs), and sustentacular cells (SUSs). It is responsible for detecting odorants and transmitting olfactory signals to the central nervous system (Figure 1A).^4,5

Figure 1.

Structural organization and cellular composition of the olfactory and respiratory regions in nasal mucosa

(A) Architecture and cellular composition of the human olfactory epithelium (OE) and respiratory epithelium (RE).

(B) Schematic diagram delineating the murine OE and RE regions, based on HE staining and immunostaining for olfactory sensory neurons (OSNs) markers OMP and Tuj1 (top: coronal views; bottom: horizontal views).

(C) Representative immunostaining images of murine OE markers (OMP, Tuj1) and RE markers (β-Tubulin Ⅳ, MUC5AC).

(D) Relative mRNA expression levels of representative marker genes analyzed by RT-qPCR.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm (except for B, 500 μm). Data are presented as mean ± SD; ns: non-significant, ∗p < 0.05, ∗∗p < 0.01.

In human, OE is primarily distributed over the superior nasal concha, middle nasal concha, and upper portion of the nasal septum. RE is interspersed within these regions, making it difficult to completely distinguish or biopsy the mucosa during surgical procedures.⁶ In mice, OE is predominantly located in the middle and posterior regions of the nasal septum.⁷ This region is characterized by the presence of OSNs that are immunopositively for specific markers: olfactory marker protein (OMP) for mature OSNs (mOSNs) and Tuj1 for immature OSNs (iOSNs).⁸ The remaining mucosa areas, which are negative for both OMP and Tuj1, are largely considered to be RE. Within the RE, the ciliary marker β-Tubulin IV is distributed across the apical surface, indicating the presence of ciliated cells. Scattered among these are a small number of goblet cells, as marked by MUC5AC (Figures 1B and 1C).⁹ The immunostaining images of coronal and horizontal sections in Figure 1B clearly localize the OE to the posteromedial region of the mouse nasal mucosa. Figure 1C shows representative, high-magnification details of both the OE and RE regions. In OE, densely packed OMP⁺ and Tuj1⁺ OSNs at different differentiation stages were observed, with mOSNs localized closer to the apical surface than iOSNs. In RE, dense β-Tubulin IV⁺ cilia were distributed along the apical surface, interspersed with MUC5AC⁺ goblet cells. This clear spatial separation in mice facilitates comparative studies of RE and OE functions.

Functional olfactory organoids derived from both human and mouse OE have been successfully established in 2021, and have demonstrated electrophysiological responses to odorant stimuli.¹⁰ Additionally, air-liquid interface (ALI) models generated from human nasal epithelium or nasal polyp have been shown to effectively recapitulate the in vivo mucosal environment.¹¹ These ALI models were widely used in studying the pathogenesis of nasal diseases such as chronic rhinosinusitis (CRS),¹² allergic rhinitis (AR),¹³ as well as mechanisms of viral infection.¹⁴

While these methodologies provide valuable in vitro platforms for rhinological research, substantial room for improvement remains. In the present study, building upon the established protocols, we expanded primary cells isolated from distinct regions of mouse nasal mucosa and subsequently directed the differentiation via two distinct culture systems: three-dimensional (3D) organoids and the ALI models. The resulting constructs were characterized through morphological analysis and identification of key cellular markers. Our work refines in vitro models of region-specific nasal mucosa and provides more reproducible, stable, and accessible tools for future investigations in rhinology.

Results

Expansion and characterization of primary mouse nasal mucosal cells from RE and OE regions

To ensure the absence of cross-contamination between the RE and OE tissues during dissection and thereby guarantee the reliability of subsequent experiments, we performed quantitative reverse-transcription PCR (RT-qPCR) analysis using representative marker genes for each region. The results showed that OSNs marker genes (Omp and Gap43) were highly expressed in OE tissue, while respiratory epithelial marker Krt19 and ciliary markers (Foxj1 and Tubb4a) exhibited significantly reduced expression in OE relative to RE (Figure 1D). These findings confirmed the validity and feasibility of our tissue isolation protocol. However, as the relative mRNA expression levels of RE markers remained detectable in the harvested OE tissues (Figure 1D), we reasoned that the OE preparations may contain a minor contamination of RE cells. To minimize the potential impact of such contamination on subsequent cell-fate analyses, particularly in cross-induction assays, all comparative groups were derived exclusively from the identical batch of primary cells. Primary single cells isolated from the RE or OE region of nasal mucosa were cultured in growth medium to allow stem cell expansion. Within the initial days of culture, a small number of spindle-shaped adherent cells appeared in both groups, followed by a phase of accelerated growth (Figures 2A and 2B). Co-staining for β-Tubulin Ⅳ, also known as a cytoskeletal marker and the tight junction protein ZO-1⁹ confirmed the formation of intercellular tight junctions. In selected fields, cells undergoing mitosis were observed (enlarged views of boxed area), providing further evidence of active proliferation (Figure 2C). Immunostaining for the basal cell markers (KRT5 and TRP63)¹⁵ indicated that the majority of expanded cells were basal cells with differentiation potential (Figure 2D, left). Furthermore, nearly all expanded cells expressed the epithelial marker E-cadherin, while a subset was positive for the proliferation marker Ki67, confirming active proliferation of the harvested epithelial cells (Figure 2D, right). Statistical analysis showed no significant difference in cell counts between the two groups, indicating that basal cells derived from RE and OE exhibited comparable growth during the proliferation phase (Figure 2E). However, RT-qPCR analysis revealed that the relative mRNA expression levels of Krt19 and Krt5 were moderately lower in OE group than in RE group, while Tubb4a expression showed no significant difference (Figure 2F). We attribute this variance to the distinct cellular origins of the cultures: RE-derived cultures originate from the RE, which naturally exhibits higher Krt19 expression. Conversely, OE-derived cultures originate from the olfactory region, where the basal cells include HBCs (normally quiescent) and GBCs, thereby constituting a more heterogeneous subpopulation. This cellular complexity likely accounts for the attenuated Krt5 expression observed in the OE group.

Figure 2.

Culture and characterization of primary cells from RE and OE regions

(A) Schematic depicting the outward radial expansion of primary cells from a single cell origin.

(B) Phase-contrast images of RE- and OE-derived cells after 5 days of culture.

(C) Immunostaining for β-Tubulin Ⅳ and ZO-1 in expanded RE- and OE-derived cultures, the rightmost image shows enlarged views of the boxed areas, revealing cell undergoing mitosis marked by β-Tubulin Ⅳ.

(D) Immunostaining for KRT5, TRP63, E-cadherin, and Ki67 in expanded RE- and OE-derived cultures.

(E) Quantification of specific cell numbers per field of view in RE- and OE-derived cultures.

(F) Relative mRNA expression levels of basal epithelial markers in RE- and OE-derived cultures analyzed by RT-qPCR.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm. Data are presented as mean ± SD; ns: non-significant, ∗p < 0.05, ∗∗p < 0.01.

Formation and phenotypic characterization of RE- and OE-derived 3D organoids

Following 5 days of culture, basal cells derived from the RE and OE reached approximately 80% confluency. These cells were then detached and re-suspended in the growth medium containing 8% Matrigel to initiate 3D culture (Figure 3A). The RE group was cultured using the commercial airway differentiation medium system (combined with PneumaCult Ex Plus medium for expansion medium and PneumaCult-ALI medium for differentiation medium purchased from STEMCELL Technologies, named SC), whereas the OE group was cultured using the homemade olfactory differentiation culture system (supplemented with Noggin, Wnt3a, and R-spondin, named NWR). Within approximately 3 days, cell colonies became visible under the microscopy. The medium was subsequently switched to a differentiation formulation. By day 14, the colonies had gradually acquired a spherical morphology and continued to increase in size (Figure 3B). Notably, RE-derived colonies predominantly formed cyst-like structures and expanded more rapidly, similar to the hollow or cyst-like architecture of the airway in vivo. In contrast, OE-derived colonies remained largely solid and were significantly smaller in volume (Figures 3C and 3D).

Figure 3.

Construction and characterization of 3D organoids derived from mouse RE and OE regions

(A) Schematic diagram illustrating the 3D differentiation process of RE- and OE-derived cells.

(B) Phase-contest images of RE- and OE-3D organoids on days 3, 7, and 14 of culture.

(C) Quantification of organoid size from RE and OE regions during differentiation.

(D) Pie chart illustrating the distribution of 3D organoid morphologies in RE and OE groups, with cystic structures predominant in RE group (92.1%) and solid structures predominant in OE group (85.6%).

(E) Immunostaining images showing the expression of specific markers in RE- and OE-3D organoids.

(F) Statistical analysis of the proportion of KRT19⁺ cells per organoid between RE- and OE-organoids.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm. Data are presented as mean ± SD, ∗∗∗∗p < 0.0001.

At day 14, 3D organoids were harvested for phenotypic analysis. In RE-3D organoids, immunostaining confirmed that nearly all cells expressed the epithelial marker E-cadherin. β-Tubulin IV localized to the central vesicular regions, while MUC5AC⁺ staining was evident within sheet-like luminal secretions. Moreover, a large proportion of cells expressed KRT19, a key respiratory epithelial marker,¹⁶ collectively indicating that these 3D organoids effectively recapitulated RE features in vitro (Figure 3E, top). In OE-3D organoids, immunostaining for OMP and Tuj1 verified their capacity to differentiate into OSNs. Concurrently, ICAM1⁺ and KRT5⁺ basal cells were predominantly localized to the colony periphery (Figure 3E, bottom).⁸ Consistent with their distinct differentiation pathways, KRT19 expression was significantly lower in OE-3D organoids than in RE group (Figure 3F).

In summary, although RE- and OE-derived basal cells showed no remarkable discernible differences during the two-dimensional expansion phase, the 3D organoids generated through subsequent differentiation exhibited clear distinctions in morphology, cellular composition, and lineage potential.

Establishment and characterization of ALI models from RE- and OE-derived cells

Given the analogous physiological architecture of the olfactory and respiratory epithelia, both being directly exposed to the external environment, we established parallel ALI models using mouse RE- and OE-derived basal cells.¹⁷ Cells were seeded into the upper chamber of Transwell inserts and cultured in the growth medium (RE- and OE-derived cells were cultured in the same expansion medium consistent with the expansion stage of primary cells) for 3–5 days until reaching full confluence. The apical medium was then removed, and the basolateral medium was replaced with either commercial airway differentiation medium (for RE) or homemade olfactory differentiation medium (for OE) to initiate differentiation (Figure 4A).

Figure 4.

Establishment of ALI models from RE and OE regions

(A) Schematic diagram illustrating the procedure for generating RE- and OE-ALI cultures, along with structural representation of differentiated cells.

(B) Phase-contrast images of RE- and OE-ALI on day 21 of differentiation.

(C) Magnified views of the regions enclosed by dashed boxes in the RE- and OE-ALI shown in (B), respectively.

(D) HE staining and corresponding high-resolution bright-field images of RE- and OE-ALI cultures on day 21 of differentiation.

(E) Immunostaining of key cell types in longitudinal sections of RE- and OE-ALI cultures.

(F) Statistical analysis of KRT19⁺ cells between RE- and OE-ALI cultures.

(G) Whole-mount immunostaining of key cell types in RE- and OE-ALI cultures.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm. Data are presented as mean ± SD, ∗∗∗∗p < 0.0001.

After about 3 days of culturing, dense cellular monolayer was observed in both RE- and OE-ALI cultures. The apical medium was aspirated, and the basolateral medium was replaced with differentiation medium. By approximately day 12 of differentiation, vigorous, coordinated ciliary beating was clearly visible in the RE-ALI cultures under microscopy (Video S1). Culturing was maintained until day 21, at which point the cell layers were washed and fixed for subsequent analysis.¹⁷ On day 21, the RE-ALI cultures exhibited densely distributed round cells, whereas the OE-ALI cultures predominantly comprised compactly arranged, spindle-shaped cells (Figures 4B and 4C. Figure 4C represents a magnified view of the area enclosed by the dashed box in Figure 4B). HE staining and phase-contrast imaging revealed that the RE-ALI cultures developed a pseudostratified epithelium with visible apical cilia (Figure 4D, left). In contrast, OE-ALI cultures showed densely packed cells without apparent ciliary movement (Figure 4D, right). Immunostaining for β-Tubulin IV and FOXJ1 in RE-ALI further confirmed the presence of ciliated cells in the apical layer (Figures 4E and 4G).

We next performed immunostaining on both RE-ALI and OE-ALI cultures, using either cross-sections (Figure 4E, side views) or whole-mount preparations (Figure 4G, apical top views) to compare their molecular profiles. KRT5⁺ cells were widely distributed in both models. In contrast, KRT19 was abundantly expressed throughout the RE-ALI cultures but only sparsely detected in OE-ALI models (Figures 4E–4G), confirming the respiratory identity of the RE-ALI model.

Immunostaining for β-Tubulin IV further revealed distinct morphological features. In RE-ALI cultures, cilia were clustered and orderly arranged on the apical surface, consistent with the HE staining observations (Figure 4D).¹⁸ Wide-field imaging showed concentrated β-Tubulin IV signals at the apical layer, confirming ciliated cell differentiation and successful establishment of the RE-ALI model (Figure 4E).

In contrast, OE-ALI cultures exhibited fine filamentous β-Tubulin IV⁺ signals distributed among cells (Figures 4E and 4G), aligning with previously reported ultrastructural features of olfactory cilia.¹⁹ Whole-mount staining of β-Tubulin IV further illustrated the cytoskeletal organization of differentiated cells in the OE-ALI model. Moreover, Tuj1⁺ cells exhibiting neuronal morphology, which were nearly absent in RE-ALI model. The distinct Tuj1 signal and characteristic morphology indicated successful neuronal differentiation, a key criterion for validating the olfactory epithelial model (Figure 4G).

Collectively, these results underscore the distinct cellular composition, differentiation outcomes, and morphological organization between RE-ALI and OE-ALI, reflecting their respective tissue origins and functions. These well-characterized models provide essential tools for studying region-specific physiology and mechanisms in vitro.

Cross-induction reveals differential lineage plasticity between RE- and OE-derived 3D organoids

In the previous section on 3D organoids construction, OE-3D organoids were stably cultured in the homemade olfactory differentiation system (named NWR), while RE-3D organoids were maintained in the commercial airway differentiation medium (named SC). To investigate the predominant influence of microenvironment signals on cell fate determination, we conducted a cross-induction experiment. During the differentiation phase of the 3D organoids, we reciprocally exchanged the established OE- and RE-specific differentiation medium to test whether extrinsic signaling cures could override intrinsic differentiation programs. The results revealed a striking difference in lineage plasticity. OE-3D organoids cultured in SC medium exhibited a marked shift in phenotype, characterized by increased size compared to those grown in the original NWR system (Figures 5A and 5B). Immunostaining results further indicated significant downregulation of Tuj1⁺, OMP⁺, and concomitant upregulation of KRT19⁺, acetylated-tubulin⁺ (Ace-Tubulin⁺, a ciliary/cytoskeletal marker similar to β-Tubulin IV²⁰) cells in OE-3D organoids under SC culture conditions (Figures 5C–5G). Conversely, RE-3D organoids largely maintained their lineage identity in either SC or NWR system, showing no substantial change in cellular composition or overall size (Figures 5A–5G).

Figure 5.

Differential responses of OE- and RE-3D organoids to cross-induction with SC and NWR system

(A) Phase-contrast images of OE- and RE-3D organoids cultured in SC or NWR system for 7 days.

(B) Quantitative analysis of organoid area across the four groups.

(C) Representative immunostaining images of OE- and RE-3D organoids cultured in SC or NWR system, stained for KRT19, Tuj1, OMP, and Ace-Tubulin.

(D) Percentage of KRT19⁺ cells per organoid in the four groups.

(E) Quantification of Tuj1⁺ cells per organoid in the four groups.

(F) Percentage of OMP⁺ cells per organoid in the four groups.

(G) Percentage of Ace-Tubulin⁺ cells per organoid in the four groups.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm. Data are presented as mean ± SD, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Together, these findings demonstrate that the differentiation trajectory of OE is substantially more sensitive to exogenous signaling factors than that of RE. Specifically, the neuronal differentiation program in OE can be more easily suppressed or redirected by signals promoting respiratory epithelial fate, enabling the tissue to dynamically respond to and integrate changing environmental stimuli.

IL-17a inhibits ciliary differentiation in nasal RE under ALI culture

Having demonstrated the high plasticity of OE in response to altered differentiation signals, we next investigated whether a specific pathologic cytokine could disrupt the differentiation program of nasal epithelium.

IL-17a is a key inflammatory cytokine implicated in CRS, known to disrupt the nasal mucosal barrier and promote nasal inflammation.²¹ Our previous study demonstrated that IL-17a can affect the fate of HBCs in OE, thereby exacerbating olfactory dysfunction induced by inflammatory aging.²² Given its clinical association with mucociliary impairment, we hypothesized that IL-17a might directly inhibit ciliary differentiation in RE. To test this, IL-17a was applied during the differentiation phase of the RE-ALI model.

To contextualize the effect of IL-17a and distinguish it from a general blockade of differentiation, we included a control group treated with CHIR99021, a small-molecule agonist of the canonical Wnt/β-catenin pathway. This pathway is crucial for maintaining airway epithelial integrity and progenitor states,¹⁷ and our previous work in OE organoid confirmed the essential role of Wnt signaling during proliferation.¹⁰ Therefore, we reasoned that sustained Wnt activation via CHIR99021 during ALI differentiation could serve as a positive control for differentiation inhibition by maintaining cells in a progenitor-like state, thereby preventing terminal ciliogenesis.

As expected, both HE histological and immunostaining (β-Tubulin IV and Ace-Tubulin) showed that CHIR99021 treatment reduced ciliary differentiation, confirming that exogenous activation of stem cell maintenance signals effectively blocks the terminal differentiation in this system. Notably, IL-17a produced a similar of ciliary-loss phenotype. Crucially, addition of an IL-17a neutralizing antibody (NAb) restored cilia formation, confirming the specific and direct inhibitory role of IL-17a (Figures 6A–6D). Immunostaining for MUC5AC revealed that IL-17a enhanced the differentiation of secretory cells, whereas neutralization of IL-17a markedly decreased the population of MUC5AC⁺ goblet cells (Figures 6C and 6D). Furthermore, RT-qPCR analysis of the mRNA expression levels of the cilia-related genes Foxj1 and Tubb4a, as well as Muc5ac, further corroborated the impact of IL-17a on ciliary and goblet cell differentiation (Figure 6E).

Figure 6.

IL-17a inhibits ciliary differentiation in RE-ALI model

(A) HE staining of longitudinal sections from RE-ALI cultures treated with IL-17a, IL-17a + Nab, or CHIR99021.

(B) Proportion of ciliary region length per 500 μm ALI cultures.

(C) Immunostaining of β-Tubulin IV, MUC5AC, and Ace-tubulin in RE-ALI cultures under the indicated treatment conditions.

(D) Quantification of β-Tubulin IV⁺ clusters, MUC5AC⁺ cells or Ace-Tubulin⁺ clusters per field of view.

(E) Relative mRNA expression levels of Foxj1, Tubb4a, or Muc5ac in RE-ALI cultures analyzed by RT-qPCR.

Nuclei were counterstained with Hoechst (blue). Scale bars, 50 μm. Data are presented as mean ± SD, ns: non-significant, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Upon binding to its receptor, IL-17 activates downstream signaling cascades, including NF-κB and mitogen-activated protein kinase (MAPK) pathways (p38 MAPK and MEK/ERK), which in turn induce the expression of pro-inflammatory cytokines and drive the inflammatory response.^23,24 To further elucidate the mechanisms by which IL-17a impairs ciliary differentiation in RE-ALI cultures, we supplemented IL-17a with specific inhibitors targeting the NF-κB, p38 MAPK, and MEK/ERK pathways (IL-17a+NF-κBi, IL-17a+p38i, and IL-17a+MEKi). After 21 days of culture, all three inhibitor-treated groups exhibited increased number of ciliary clusters compared with the IL-17a group. Among these inhibitor-treated groups, RT-qPCR revealed that IL-17a+p38i group exhibited slightly stronger effects on both the restoration of ciliary differentiation and the suppression of goblet cell differentiation compared with the other two groups, a finding that was further corroborated by immunostaining (Figures 6C–6E and S1A–S1C). These results demonstrate that IL-17a potently inhibits ciliogenesis in vitro. Our findings not only identify IL-17a as a direct inhibitor of respiratory epithelial differentiation but also validate the utility of the ALI model for mechanistically dissecting how specific inflammatory cytokines contribute to epithelial dysfunction in nasal diseases. The parallel inhibition observed with CHIR99021 suggests that IL-17a may disrupt a critical checkpoint in the differentiation cascade, potentially intersecting with or mimicking pathways that govern progenitor cell fate.

Discussion

External stimuli or internal physiological changes can compromise the barrier function of the nasal mucosa or lead to olfactory dysfunction, as indicated in recent studies.^25,26 However, the specific mechanisms underlying these effects, as well as strategies to promote structural and functional recovery in distinct regions of nasal mucosal after injury, remain critical unresolved questions. Organoids have emerged as pivotal tools for elucidating disease pathogenesis, evaluating drug efficacy, and investigating tissue regeneration, particularly in the repair of the OE.²⁷ Nevertheless, obtaining patient-derived tissues presents substantial challenges for many researchers, and certain genetic modifications are difficult to recapitulate in human subjects, potentially limiting further investigations.

In our previous olfactory organoid culture system, single cells were obtained directly from OE and cultured in 3D model.¹⁰ It was reported that respiratory cells can be collected from nasal mucosa and cultured under ALI conditions.²⁸ To enable a direct comparison between OE- and RE-derived organoids, the present study first adopted a stem cell expansion phase. During this stage, both OE and RE progenitors exhibited similar proliferative capacities without obvious morphological differences, although the relative mRNA levels of Krt19 and Krt5 were downregulated in OE group compared to RE group. Subsequently, differentiation was induced under either 3D or ALI culture conditions.

In 3D cultures, RE-derived organoids formed expanding cystic structures expressing characteristic markers (E-cadherin, KRT19, and MUC5AC), while OE-derived organoids remained as compact solid aggregates, capable of generating OSNs (OMP⁺, Tuj1⁺) and retaining basal cells (KRT5⁺, ICAM1⁺) at their periphery. Unlike KRT5, which is specifically expressed in basal cells, ICAM1 exhibits a broader expression pattern in nasal epithelial cells. In addition to basal cells, Icam1 is also highly expressed in other cell types, including ciliated cells, endothelial cells, and immune cells. Consequently, the localizations of these two markers differ within 3D organoids.^29,30

The significantly lower proportion of KRT19 expression in OE-3D organoids further underscores their divergent differentiation trajectory. KRT19 is a type 1 intermediate filament protein, expressing across a broad range of epithelial tissues systemically.³¹ Although KRT19 is widely recognized as a relatively specific marker for nasal RE, its expression is not strictly confined to the respiratory region, and it also detectable in injured OE, a phenomenon known as respiratory epithelial metaplasia.²² Consistent with previous reports, the OSNs positive region within the OE of aged mice with inflamm-aging was progressively replaced by KRT19⁺ RE. Concomitantly, vesicular 3D organoids derived from OE showed a marked increase in both proportion and KRT19⁺ cell percentage. As the expansion of KRT19⁺ RE reflects neuronal loss, a hallmark of respiratory metaplasia, these findings collectively suggest a further decline in olfactory function.^16,22

Under ALI culture conditions, RE-derived cells consistently differentiated into a pseudostratified epithelium with motile cilia (β-Tubulin IV⁺), while OE-derived cells efficiently generated Tuj1⁺ iOSNs. Clear ciliary patterning became evident in RE-ALI cultures by around day 12, and the ALI architecture remained intact through day 21.

Our results revealed a striking divergence in differentiation plasticity between OE- and RE-derived 3D organoids. Specifically, upon exposure to altered external signals, such as cross-induction with lineage-inappropriate differentiation media, OE-3D organoids demonstrated higher phenotypic plasticity, readily shifting toward a respiratory epithelial-like phenotype. In contrast, RE-3D organoids largely maintained their lineage identity, their response to external perturbations was primarily manifested as impaired ciliary differentiation or structural abnormalities rather than transdifferentiation.

The adult OE retains a considerable degree of neuroregenerative capacity, and its basal stem cells are bipotent, capable of giving rise to either OSNs or respiratory-like epithelial cells depending on microenvironmental cues.⁵ For example, inflamm-aging has been shown to induce metaplastic changes in the OE of aged mice.²² In this study, the upregulation of KRT19, Ace-Tubulin and downregulation of OMP, Tuj1 in OE-3D organoids cultured under original commercial airway differentiate medium (SC) indicate that their differentiation trajectory can be reprogrammed by exogenous respiratory signals, reflecting the adaptive plasticity of OE in response to environmental changes. Conversely, the RE is primarily responsible for mucosal barrier function and mucociliary clearance under physiological conditions, and its stem cells are more committed to maintaining epithelial homeostasis and repair rather than undergoing cross-lineage conversion.³² Thus, under pathological stimulation, RE-3D organoids mainly exhibit differentiation arrest, such as inhibition of ciliogenesis, rather than lineage switching.¹² This differential plasticity carries important pathophysiological implications. In inflammatory diseases such as CRS, elevated local cytokines (such as IL-17a) may directly impair respiratory epithelial function by suppressing ciliogenesis.² Meanwhile, under certain injury or inflammatory conditions, the OE may undergo respiratory metaplasia due to dysregulated microenvironmental signals, leading to olfactory dysfunction.³³ The OE and RE organoids established in this study provide a robust and physiologically relevant platform to simulate these pathological processes in vitro, dissect underlying signaling pathways, and screen potential therapeutic strategies. Future studies can employ OE- and RE-derived 3D organoids or ALI models to further investigate the molecular mechanisms underlying respiratory epithelial metaplasia in OE, with the goal of reversing or slowing the metaplastic process and restoring olfactory function impaired by disease.

Using RE-ALI models, we demonstrated that IL-17a markedly suppresses ciliary differentiation while increasing secretory cell abundance. Given the elevated IL-17a levels reported in nasal mucosa or polyps of CRS patients and aged individuals,^22,34 our results indicate that upregulated IL-17a impairs ciliary differentiation, promotes abnormal secretions, reduces pathogen clearance, disrupts epithelial barrier function, and worsens clinical outcomes. Conversely, inhibiting IL-17a secretion or its downstream signaling pathways can restore ciliary differentiation and alleviates symptoms, providing critical insights for clinical therapy. Our present data indicate that blockade of either the NF-κB or the MAPK pathway downstream of IL-17 can partially rescue ciliary regeneration. Nevertheless, the relative contribution of each pathway to this process remains unclear, and the precise molecular mechanisms involved warrant further exploration. Therefore, it is necessary continuing to address these unresolved issues using the ALI culture system developed in this study.

Through the establishment of 3D and ALI organoids derived from mouse RE and OE, as well as studies on their differentiation plasticity and the modulation of ciliary differentiation in the RE-ALI model, we have provided multiple highly feasible and easily accessible in vitro models for future research. By incorporating organoids derived from distinct anatomical regions, we can better compare the responses and underlying mechanisms of different nasal regions (RE and OE) to external stimuli. Future studies will include mouse tracheal organoids to investigate disease processes across the entire airway.³⁵

Limitations of the study

Due to the lack of specific markers to distinguish basal cells of the RE and OE, we relied solely on anatomical localization for region isolation, making minor contamination unavoidable. However, characterization of terminal differentiated cell types confirmed that this did not fundamentally affect the results. Nevertheless, although the present study distinguished the OE and RE regions based on morphological and anatomical landmarks, and subsequent experiments have confirmed the feasibility of this approach, the identification of reliable live-sort markers capable of distinguishing OE- and RE-derived basal cells remains a fundamental challenge. Notably, due to the technical challenges in completely separating OE from RE during tissue harvest, we cannot rule out the possibility that a portion of KRT19⁺ signals detected in OE-derived 3D organoids resulted from the proliferation or differentiation of contaminating RE cells. While the use of single-batch primary cells ensured internal consistency across our comparative assays, this limitation suggests that the observed RE marker expression reflects a potential for lineage transition rather than definitive fate conversion. Further studies employing lineage-tracing models will be required to definitively dissect the cell-of-origin. Moreover, while certain markers have been validated in mice to effectively discriminate between the RE and OE regions, such as KRT19, their translational validity in humans has yet to be established.³⁶ Therefore, future studies should place greater emphasis on addressing this issue and refining the relevant research efforts. Furthermore, due to interspecies differences, mouse-derived organoids cannot fully replace patient-derived organoids, and drug candidates identified in mice require validation in human-based studies and clinical trials. This study aimed to establish simple, rapid, and highly reproducible in vitro models for basic research; translation to clinical applications must ultimately rely on patient-based studies and rigorous clinical trials to validate efficacy and safety.

Resource availability

Lead contact

Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Wenwen Ren (wenwenren@smmu.edu.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •This study did not generate standardized datasets. All data reported in this paper are presented in the main figures and supplemental materials, and raw data supporting the findings are available from the lead contact upon request.
• •This study does not report original code.
• •Any additional information required to re-analyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

This work was supported by National Natural Science Foundation of China grants (U23A20439 to H.L., 82271136 and 82671458 to W.R.). We thank the staff members of the Integrated Laser Microscopy System at the National Facility for Protein Science in Shanghai (NFPS), Zhangjiang Lab, China for providing technical support and assistance in data collection and analysis. The graphical abstract was created in BioRender.

Author contributions

W.R., J.W., H.Z., and Y.Z. conceived and designed the work; Y.X. and Z.X. conducted the experiments; C.X. acquired and analyzed the data; and H.L. wrote the manuscript. All authors read and approved the manuscript.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit anti-OMP	Abcam	Cat# ab183947; RRID: AB_2858281
Mouse anti-beta Ⅲ Tubulin (Tuj1)	Abcam	Cat# ab78078; RRID: AB_225675
Rabbit anti-beta Tubulin (β-Tubulin)	AiFang biological	Cat# AF300018
Mouse anti-MUC5AC	Abcam	Cat# ab3649; RRID: AB_2146844
Goat anti-E-cadherin	R&D	Cat# AF748；RRID:AB_355568
Rabbit anti-Ki67	Abcam	Cat# ab16667; RRID: AB_302459
Rabbit anti-KRT5	Abcam	Cat# ab52635; RRID: AB_869890
Mouse anti-TRP63	Abcam	Cat# ab735; RRID: AB_305870
Mouse anti-ZO-1	Invitrogen	Cat# 33-9100; RRID: AB_87181
Rabbit anti-KRT19	Abcam	Cat# ab52625; RRID: AB_2281020
Goat anti-ICAM1	R&D	Cat# AF796; RRID:AB_2248703
Mouse anti-Ace-Tubulin	Proteintech	Cat# 66200; RRID: AB_2722562
Hoechst 33258	Solarbio	Cat# C0021
Donkey Anti-Rabbit, Alexa Fluor 488	Abcam	Cat# ab150073; RRID: AB_2636877
Donkey Anti-Rabbit, Alexa Fluor 568	Abcam	Cat# ab175470; RRID: AB_2783823
Donkey Anti-Mouse, Alexa Fluor 568	Abcam	Cat# ab175472; RRID: AB_2636996
Donkey Anti-Goat, Alexa Fluor 488	Abcam	Cat# ab150129; RRID: AB_2687506
Donkey Anti-Goat, Alexa Fluor 647	Abcam	Cat# ab150135; RRID: AB_2687955
Chemicals, peptides, or recombinant proteins
Recombinant Mouse IL-17A	Novoprotein	Cat# CX14
IL-17A Monoclonal Antibody	ThermoFisher	Cat# 16-7173-85; RRID: AB_763585
CHIR99021	Tocris Bioscience	Cat# 4423
BAY 11–7082 (NK-κB inhibitor)	MedChemExpress	Cat# HY-13453
Adezmapimod (p38 MAPK inhibitor)	MedChemExpress	Cat# HY-10256
PD98059	MedChemExpress	Cat# HY-12028
PneumaCult™ Ex Plus medium	Stemcell Technologies	Cat# 05040
PneumaCult™-ALI medium	Stemcell Technologies	Cat# 05001
DMEM/F12 + GlutaMAX medium	Gibco	Cat# 10565-018
HEPES (1M)	Gibco	Cat# 15630-080
Penicillin-Streptomycin (P/S)	Gibco	Cat# 15140-122
R-spondin-1	R&D system	Cat# 4645-RS
Noggin	Peprotech	Cat# 250-38
Wnt 3a	R&D	Cat# 5036-WN-010
Y-27632	Sigma-Aldrich	Cat# Y0503
Murine EGF	Peprotech	Cat# 315-09-100
N-2 Supplement	Gibco	Cat# 17502048
B-27 Supplement	Gibco	Cat# 17504044
EZBioscience® Fast Tissue RNA Purification Kit	EZBiosence	Cat# EZB-RN5
HiScript® III All-in-one RT SuperMix Perfect for qPCR	Vazyme	Cat# R333
ChamQ Universal SYBR qPCR Master Mix	Vazyme	Cat# Q711
Experimental models: Organisms/strains
Wild-type adult C57/BL6 mice	SPF Biotechnology Co., Ltd	N/A
Oligonucleotides
GAPDH (forward: 5′- GAGTCCACTG GCGTCTTCAC -3′; reverse: 5′-ATCTTGAGGCTGTTGTC ATACTTCT-3′).	Sangon Biotech	N/A
Krt5 (forward: 5′-GGCACCAAGACCATAAAGCAG-3′; reverse: 5′- GAACCCCGTCCA GCTGTCTA-3′).	Sangon Biotech	N/A
Krt19 (forward: 5′-GGGGGTTC AGTACGCATTGG-3′; reverse: 5′-GAGGACGAGGTCACGAAGC-3′).	Sangon Biotech	N/A
Tubb4a (forward: 5′-CTCCCGTAAGTCCCCGTA-3′; reverse: 5′- TCTCTCACCCCAGGCAA-3′).	Sangon Biotech	N/A
Gap43 (forward: 5′- CTGCTACTAC CGATGCAGCC -3′; reverse: 5′- ATCTTCAGCCTTGGAGGACG -3′).	Sangon Biotech	N/A
Muc5ac (forward: 5′- GGACTTCAATATCCAGCTACGC -3′; reverse: 5′- CAGCTCAA CAACTAGGCCATC -3′).	Sangon Biotech	N/A
Foxj1 (forward: 5′- ATCTCAGC ACCCAGGCA -3′; reverse: 5′- GTCCCCAGTTCCCAGTCTT -3′).	Sangon Biotech	N/A
Software and algorithms
Prism 10	GraghPad	https://www.graphpad.com/
LAS X Office	Leica Microsystems	RRID:SCR_013673
Olympus OlyVIA software	Olympus Life Science	RRID:SCR_016167

Experimental model and study participants details

Animals

To eliminate the influence of gender, species or age, all mice used in our experiment were wild-type C57BL/6 male mice, aged 6–8 weeks, purchased from SPF Biotechnology Co., Ltd (Beijing, China). Animals were housed under specific pathogen-free (SPF) conditions at 22 ± 2 °C with a 12h light/dark cycle and free access to food and water. Mice were randomly allocated to experimental groups. The animal handling procedures were approved by the Ethics Committee of Laboratory Animals in Naval Medical University (Permit Number: SYXK2022-0011).

Method details

Primary cells isolation and expansion

Mice were deeply anesthetized and then euthanized by cervical dislocation. The head of each mouse was excised, and excess facial muscles and connective tissues were removed. The skull was cut open along the midsagittal suture, and the entire brain, including the olfactory bulbs, was carefully removed. The nasal roof was then reflected forward along the midsagittal suture, and the nasal septum together with the attached mucosa was harvested.

The respiratory epithelium (RE) mucosa was originated from the area of the nasal septum adjacent to the tip of the nose, while the olfactory epithelium (OE) mucosa was originated from the posteromedial region of the nasal septum. The mucosa was soaked in sterile DPBS and minced into small fragments with scissors. The mucosa fragments were digested in 0.25% trypsin-EDTA with 40 U/mL DNase I for 15 min. Single cell suspensions from both RE and OE mucosa were filtered through a 40 μm nylon mesh, washed 1–2 times with DPBS and cultured in PneumaCult Ex Plus medium (Stemcell Technologies, #05040). The primary cells were expanded to the sufficient number for nest steps.

Culturation of RE- and OE-derived 3D organoids (cell colonies)

The primary cells were digested in 0.25% trypsin-EDTA and washed 1-2times with DPBS to completely remove the trypsin-EDTA. The single cells were cultures in the growth medium containing 8% Matrigel and dispensed into ultra-low-attached 24-well plate (Corning, #3473). The cells were maintained for 2–3 days until the colonies were formed. After that, the growth medium was replaced to the differentiation medium and maintained to the end of culturation (until day 14).

The growth medium for RE-3D organoids is PneumaCult Ex Plus medium (Stemcell Technologies, #05040), for OE-3D organoids is a housemade medium based on DMEM/F12 medium supplemented with GlutaMAX (1%), HEPES (1 mM), P/S (1%), Wnt 3a (50 ng/mL), Noggin (100 ng/mL), R-spondin-1 (200 ng/mL), Y-27632 (10 μm/mL), EGF (50 ng/mL), N2 (1%), B27(2%).^10,37 The differentiation medium for RE-3D organoids is PneumaCult-ALI medium (Stemcell Technologies, #05001), and for OE- 3D organoids is OE growth medium without Y-27632.

For the subsequent 3D organoid experiments, the commercial stem cell airway differentiation medium system (combined with PneumaCult Ex Plus medium and PneumaCult-ALI medium) was named SC, and the housemade OE expansion/differentiation medium system was named NWR.

Culturation of air-liquid interface (ALI) models

The primary cells were seeded onto the collagen type-Ⅰ (Corning, #354236) coated upper chamber of Transwell for 3–5 days till the cells grew all over the bottom. Then the apical medium was completed discarded and replaced the basolateral medium with differentiation medium. The apical-layer cells were washer twice a week with DPBS. The cilia motility was observed under the microscope after day 12 of culture, and the differentiation period was extended to 21 days as previously described.¹⁷ The growth medium of expansion stage (both RE-ALI and OE-ALI) was PneumaCult Ex Plus medium. The differentiation medium for RE-ALI and OE-ALI were described in the part of the culturation of cell 3D organoids (cell colonies).

To assess the effect of IL-17a, the recombinant mouse IL-17A (40 ng/mL), IL-17A Monoclonal Antibody (100 ng/mL), the Wnt signaling pathway activator CHIR99021(3 μM), the NF-κB inhibitor (10 μM), p38 MAPK inhibitor (10 μM) and MEK/ERK inhibitor (10 μM) were added to the differentiation medium at the beginning of the differentiation stage, and continued culturing for 21 days.

Real-time quantitative PCR (RT-qPCR)

Total RNA was extracted from the cell samples using the EZBioscience® Fast Tissue RNA Purification Kit (EZBioscience, # EZB-RN5) according to the manufacturer’s instructions. Reverse transcription was performed with a reverse transcription kit (Vazyme, #R333) to obtain cDNA. Quantitative PCR was performed using the ChamQ Universal SYBR qPCR Master Mix (Vazyme, #Q711) following the manufacturer’s protocol. GAPDH was used as the internal control, and the relative expression levels of target mRNAs were calculated using the 2^−ΔΔCt method. All reactions were conducted in triplicate. Primers for RT-qPCR were synthesized by Sangon Biotech, and their sequences are listed in key resources table.

Paraffin section preparation

The cell colonies or ALI models were collected at the day 21 of culturation. For cell colonies, the Matrigel was completely removed and fixed with 4% paraformaldehyde (PFA) overnight. The next day, the PFA was removed and the cell colonies were resuspended in pre-warmed 3% agarose. After the agarose solidified, the mixtures were subjected to gradient dehydration and clearing and embedded in paraffin for further sectioning procedures. For the ALI models, the chambers were performed overnight using 4% PFA, followed by gradient dehydration. The cell layers were then embedded in paraffin for subsequent procedures.

Immunostaining

For paraffin-embedded sections, deparaffinization with a gradient series of solvents and antigen retrieval were performed prior to staining, whereas for whole cell colonies or ALI models, the procedure was commenced directly with blocking. All samples were blocked by incubation with 5% bovine serum albumin (BSA) containing 0.5–1% Triton X-100 for 1 h at room temperature. Primary antibodies were then applied and incubated overnight at 4 °C. Following three washes with PBS containing 1% Tween 20 (PBST), the corresponding secondary antibodies were applied and incubated for 2 h at room temperature. After another three PBST washes, nuclei were stained with Hoechst 33258 for 10 min at room temperature. Finally, the samples were mounted for imaging. Fluorescent images were acquired using a Leica SP8 confocal microscope (LAS X Office software) or an Olympus SpinSR10 super-resolution microscope (Olympus OlyVIA software). Given that β-Tubulin IV serves as both a general cytoskeletal marker and a specific marker for cilia in RE, its positive signals were quantified differently across figures. In Figure 2, where it labeled the cytoskeleton of primary cells, positivity was counted based on the number of labeled cells (% of β-Tubulin IV⁺ cells). However, in Figure 6, where it identified clustered ciliary structures, positivity was quantified by counting discrete clusters of the signal (% of β-Tubulin IV⁺ clusters) per field of view, the counting method for β-Tubulin IV⁺ signals described in Figure 6 is also applicable to the quantification of Ace-Tubulin⁺ signals.

Quantification and statistical analysis

Statistical analysis

Graphpad Prism version 10.0 was used for the statistical analysis. For normally distributed data, we used the unpaired Student’s t test (two-sided) to compare the differences between two groups and the analysis of variance (ANOVA) with Tukey post hoc test or Sidak’s multiple comparison test (one-sided) to adjust multiple comparisons. For non-normally distributed data, the two-tailed Mann-Whitney U test was used for comparisons between two groups. Each experiment was independently performed at least three times, with similar results across replicates. The icon of p value: ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001. Each experiment was independently performed at least three times, with similar results across replicates.