Chronic rhinosinusitis (CRS) is defined as persistent sinonasal inflammation lasting more than 12 weeks and accompanied by symptoms such as nasal obstruction and nasal discharge [1]. CRS has traditionally been divided into two phenotypes: CRS with nasal polyps (CRSwNP) and CRS without nasal polyps (CRSsNP). However, with the advent of endotype-based classification systems that reflect underlying disease mechanisms, CRS is now categorized as type 2 CRS or non-type 2 CRS, each demonstrating distinct clinical manifestations and treatment outcomes [2,3].

Type 2 CRS is characterized by enhanced type 2 inflammation mediated by type 2 innate lymphoid cells (ILC2s) and T helper 2 (Th2) cells, leading to marked eosinophilic infiltration in local tissues. In addition, other innate immune cells, including neutrophils and macrophages, play important roles in type 2 CRS pathophysiology [4,5]. Notably, the depletion of eosinophils alone has not produced substantial clinical improvement, underscoring the heterogeneous nature of CRS pathogenesis [6]. Clinically, one of the most important aspects of type 2 CRS is its strong association with comorbid conditions such as asthma and nonsteroidal anti-inflammatory drug–exacerbated respiratory disease, both of which are linked to severe type 2 inflammation and poorer prognosis [7]. Immunologically and histomorphologically, type 2 CRS shares pathogenic mechanisms with asthma. Therefore, management of comorbid conditions has become a central consideration in treating severe type 2 CRS. This “one airway, one disease” concept has contributed significantly to the development and effectiveness of biologic therapies [8].

Innate or innate-like T cells, also known as nonconventional T cells, acquire a memory phenotype in the thymus and are restricted by nonclassical major histocompatibility complex (MHC) molecules such as CD1d and MR1 [9,10]. These cells are distinct from conventional αβ T cells and natural killer (NK) cells, and their defining characteristics are summarized in Table 1. Invariant natural killer T (iNKT) and mucosal-associated invariant T (MAIT) cells are the prototypical innate T cells in both mice and humans, while certain γδ T cell subsets in mice share similar properties [11,12]. MAIT and iNKT cells express semi-invariant T cell receptors (TCRs) that recognize distinct classes of antigens (Table 2): MAIT cells respond to the vitamin B2 metabolite 5-(2-oxopropylideneamino)-6-D-ribitylaminouracil (5-OP-RU), produced by commensal microbes and presented by MR1, whereas iNKT cells recognize lipid antigens presented by CD1d. Both cell types characteristically express the transcription factor promyelocytic leukemia zinc finger protein (PLZF), which drives their innate-like functional programming. In the periphery, these cells preferentially localize to the liver, lung, gut, and mesenteric lymph nodes, where they can be rapidly activated within hours of antigen recognition [13]. Moreover, they may also be activated independently of TCR signaling by cytokines such as interleukin (IL)-12 plus IL-18 or IL-23 plus IL-1β, functioning as immunological bridges that link innate responses in peripheral tissues to adaptive immunity in lymphoid organs [14].

The antigen-presenting molecules MR1 and CD1d are highly conserved across species, and therefore the antigen specificity of MAIT and iNKT cells is largely shared between mice and humans. Nevertheless, innate T cells also exhibit notable interspecies differences, posing challenges when translating findings from mouse models to human biology. For instance, iNKT cells are relatively abundant in mice, with frequencies reaching 10%–30% of T cells in the liver [15]. In contrast, they are much less frequent in humans, typically constituting less than 0.5% of peripheral blood T cells [16]. MAIT cells display the opposite pattern: they are rare in mice, accounting for less than 1% of peripheral T cells, but comparatively enriched in humans, where they generally represent about 1%–10% of circulating T cells [17].

Interspecies differences are further emphasized by the functional diversification of innate T cells. In mice, mature MAIT cells and iNKT cells are divided into two distinct populations. MAIT1 and NKT1 cells (PLZF^low T-bet^⁺) predominantly produce interferon (IFN)-γ, whereas MAIT17 and NKT17 cells (PLZFintermediate [PLZF^int] RORγt^⁺) mainly secrete IL-17 [18]. Human MAIT and iNKT cells, however, coexpress RORγt and T-bet, and the majority produce IFN-γ and TNF upon stimulation, while IL-17 secretion remains limited [12]. Consequently, human MAIT cell subsets are more commonly classified according to chemokine receptor expression profiles and cytokine production rather than transcription factor dominance. MAIT1 cells are characterized by a CXCR3⁺ CCR4⁻ phenotype and IFN-γ secretion, whereas MAIT17 cells display a CXCR3⁻ CCR4⁻ CCR6⁺ profile and preferentially produce IL-17A [19].

γδ T cells can be subdivided into γδT1 and γδT17 subsets based on transcription factor expression or cytokine profiles, and they are alternatively classified according to their TCR chain usage (Fig. 1) [20-25]. In mice, γδ T cells are primarily defined by their Vγ chain usage (Vγ1–Vγ7), which determines their developmental timing, tissue distribution, and effector functions [26]. For example, Vγ5^⁺ T cells are enriched in the skin epithelium, whereas Vγ7^⁺ cells preferentially localize to the intestinal epithelium. Vγ5^⁺ and Vγ6^⁺ subsets arise during fetal development, while other subsets are largely generated after birth. IL-17-producing γδ T cells are restricted to the Vγ4 and Vγ6 subsets, whereas Vγ1^⁺ and Vγ4^⁺ γδ T cells, which are abundant in secondary lymphoid organs, predominantly produce IFN-γ.

Unlike in mice, human γδ T cells are classified mainly according to their Vδ chain usage [27]. Vδ1^⁺ γδ T cells are enriched in tissues such as the skin, liver, gut epithelium, and spleen but constitute only a minor fraction of circulating γδ T cells; functionally, they exhibit adaptive-like properties. By contrast, Vδ2^⁺ γδ T cells, which typically pair with the Vγ9 chain, arise early during fetal development and predominate in peripheral blood, displaying innate-like functional features. Vδ3^⁺ γδ T cells represent a minor subset enriched primarily in the liver and gut epithelium [28]. These fundamental distinctions in classification and functional bias highlight the challenges of extrapolating findings from mouse γδ T cell studies to human biology.

Beyond species-specific and functional differences, methodological advances have refined how innate T cell subsets are detected and characterized. Traditionally, γδ T cells were identified using antibodies recognizing shared epitopes on the TCRγδ heterodimer, while invariant T cell subsets such as iNKT and MAIT cells were identified based on their semi-invariant TCR chains. For instance, in humans, iNKT cells were classically defined as Vα24-Jα18^⁺ CD3^⁺ T cells, whereas MAIT cells were identified as Vα7.2^⁺ CD161^high CD3^⁺ T cells. More recently, the development of PBS57-loaded CD1d tetramers for iNKT cells and 5-OP-RU–loaded MR1 tetramers for MAIT cells has enabled more precise and reliable identification of these populations [29-31].

Despite the accumulating evidence implicating innate T cells in mucosal immunity, their roles in chronic airway inflammation remain incompletely understood. Recent advances in single-cell analysis and immunophenotyping have allowed more detailed characterization of these cells in human tissues, revealing subset-specific functions that may differ between the upper and lower airways. Elucidating how these cells contribute to the pathogenesis of CRS and asthma is essential for identifying novel therapeutic targets that extend beyond conventional type 2 immunity.

In this review, we summarize recent advances in understanding the roles of innate T cells in the immunopathology of CRS and asthma. We compare their phenotypic and functional characteristics across disease endotypes, examine the discrepancies between murine and human studies, and discuss how local inflammatory environments shape their polarization. We also highlight methodological differences in their identification, which may partly account for the inconsistencies reported in the literature.

Recent studies have highlighted a potential role for MAIT cells in the pathogenesis of CRS [32,33]. Rha et al. [32] were the first to report the presence of MAIT cells in the sinonasal mucosa. These cells expressed tissue-residency markers such as CD69, CD103, and CD49a, indicating their tissue-resident phenotype. Although the proportion of MAIT cells among CD3^⁺ T cells was similar between control and CRS mucosa, those from patients with eosinophilic CRSwNP (E-NP) exhibited increased CD38 expression and IL-17A production. The frequency of CD38^⁺ MAIT cells positively correlated with that of IL-17–producing MAIT cells in E-NP patients, whereas this relationship was not observed in patients with non-eosinophilic CRSwNP (NE-NP) or CRSsNP. Furthermore, the frequencies of total MAIT cells, CD38^⁺ MAIT cells, and IL-17–secreting MAIT cells were all positively associated with disease severity, as measured by the Lund-Mackay computed tomography score, in E-NP patients.

Ye et al. [33] further investigated the role of MAIT17 cells in patients with CRSwNP. Their study subclassified CRSwNP into four endotypes: paucigranulocytic, eosinophilic, neutrophilic, and mixed granulocytic. Sinonasal MAIT cell frequencies, particularly those of IL-17–producing subsets, were elevated in the neutrophilic and mixed granulocytic endotypes compared to controls, whereas no such increase was observed in the eosinophilic endotype. In the neutrophilic endotype, MAIT cell frequency correlated with both greater neutrophil infiltration and higher disease severity. Consistent with these findings, MAIT cells isolated from neutrophilic polyps promoted neutrophil migration, survival, and activation, supporting their proinflammatory role in this disease subtype.

Interestingly, the findings of Rha et al. [32] and Ye et al. [33] diverged regarding the association between IL-17–producing MAIT cells and CRSwNP endotypes. While Rha et al. [32] reported a positive correlation between these cells and disease severity in E-NP, Ye et al. [33] observed a similar relationship in neutrophilic CRSwNP. In addition, a single-cell RNA sequencing study by Wang et al. [34] revealed reduced MAIT cell numbers in E-NP patients compared with healthy controls. These discrepancies may be partly attributable to methodological differences in MAIT cell identification and variations in the criteria used to define disease endotypes. Rha et al. [32] detected MAIT cells using 5-OP-RU–loaded MR1 tetramers, Ye et al. used antibodies against TCR Vα7.2 and CD161 [33], and Wang et al. [34] defined MAIT cell clusters based on gene expression profiles without canonical TCR sequence information. Notably, only 38.7% of TCR Vα7.2⁺ CD161⁺ CD3⁺ T cells in sinonasal tissue were positive for MR1 tetramers, compared to 96% in peripheral blood [32]. Given that MR1 tetramer binding represents the gold standard for MAIT cell identification, these findings underscore the need for further validation to reconcile the observed inconsistencies.

MAIT cells are a potential source of cytokines such as IFN-γ, IL-17, and IL-4 induced 1 (IL-4I1), and their role in asthma remains unclear [35]. In MR1-deficient mice, acute Alternaria inhalation induced airway inflammation accompanied by increased ILC2 activation, an effect attributed to the absence of MAIT cell–derived IL-4I1, an immunoregulatory enzyme that suppresses ILC2 activation [36]. Consistent with these findings, subsequent studies demonstrated that the absence of MAIT cells exacerbated ovalbumin (OVA)-induced asthma [37], while transgenic mice with an increased MAIT cell abundance were resistant to Alternaria-induced airway inflammation, which was associated with IFN-γ-mediated suppression of ILC2s [38]. In humans, MAIT cell frequencies were found to be reduced in both peripheral blood and lung tissue of asthma patients [39]. Moreover, higher peripheral MAIT cell frequencies at one year of age were associated with a lower risk of developing asthma by age 7, potentially reflecting a Th1-skewed immune tendency [40]. Notably, MAIT cell frequencies were decreased in asthma patients and showed a stronger association with the neutrophilic subtype than the eosinophilic subtype [41]. Collectively, these findings support a protective role for MAIT cells in asthma pathogenesis in both mice and humans.

In contrast to reports suggesting a protective function, Lezmi et al. [42-44] classified MAIT cells into IFN-γ– and IL-17–secreting subsets by flow cytometry and observed an increase in circulating MAIT17 cells in pediatric patients who experienced asthma exacerbations, with their frequency positively correlating with disease severity. These studies analyzed peripheral blood mononuclear cells (PBMCs) and bronchoalveolar lavage fluid (BALF) from children with severe asthma. Although the overall MAIT cell frequency did not differ significantly between the two compartments, the proportion of IL-17–producing MAIT cells was higher in BALF and correlated with severe exacerbations in the previous year [42]. Supporting these findings, Wen et al. [45] further stratified pediatric patients into MAIT17-high and MAIT17-low groups, showing that the former experienced more severe exacerbations. They also proposed a link between the MAIT17 phenotype and the neutrophilic asthma endotype. In line with this, the therapeutic efficacy of monoclonal antibodies such as mepolizumab and reslizumab, which neutralize IL-5 or deplete eosinophils, was predicted by lower MAIT cell counts in patients with severe asthma [37]. Although these studies were limited by the absence of healthy controls and relatively small sample sizes, they collectively suggest that the functional composition and anatomical distribution of MAIT cells may act as confounding variables in assessing their role in allergic asthma pathogenesis.

A recent study demonstrated that MAIT17 cells can convert into MAIT1-like populations during bacterial infection, indicating substantial phenotypic plasticity among MAIT cells [46]. This finding suggests that MAIT cells, influenced by the local cytokine milieu (such as IL-12, IL-18, and IL-23) [47], may adapt their responses in settings such as allergic airway inflammation.

Altogether, these studies indicate that the role of MAIT cells in asthma cannot be defined in absolute terms but depends on their functional polarization and the surrounding immune context. Rather than being inherently protective or pathogenic, MAIT cells exhibit significant plasticity, with their effects shaped by factors including cytokine environment, microbial signals, disease endotype, and therapeutic exposure [46]. This context-dependent nature likely explains the conflicting findings across studies and underscores the need for mechanistic and stratified clinical investigations to clarify their contribution to asthma pathogenesis.

iNKT cells have recently emerged as potential contributors to the inflammatory milieu of CRSwNP [48,49]. Their frequency was higher in nasal polyp tissue than in peripheral blood [48]. An increased frequency of iNKT cells was observed both locally and systemically in CRSwNP patients compared with healthy controls [49]. However, the proportion of HLA-DR^⁺ PD-1^⁺ iNKT cells was selectively elevated in nasal polyps but not in peripheral blood. Notably, this population was increased in neutrophilic CRSwNP and reduced in E-NP, showing an inverse correlation with eosinophil infiltration in the sinonasal mucosa [48,49]. Moreover, NKT2 and NKT17 subsets were preferentially expanded in eosinophilic and neutrophilic nasal polyps, respectively, suggesting that distinct inflammatory microenvironments shape the functional polarization of iNKT cells [49].

In a BALB/c mouse model of asthma induced by OVA immunization, Jα18 knockout mice failed to develop airway hyperresponsiveness (AHR), which was restored by adoptive transfer of iNKT cells, demonstrating the essential role of iNKT cells in murine asthma pathogenesis [50,51]. The spontaneous development of asthma in T-box transcription factor TBX21 (T-bet)–deficient mice was initially attributed to an imbalance between Th1 and Th2 CD4^⁺ T cells [52]. However, Tbx21 and Cd1d double knockout mice also failed to develop AHR, further confirming that iNKT cells are required for asthma induction [53]. In T-bet–deficient mice, iNKT cell progenitors reciprocally differentiated into NKT2 cells, which constitutively produced IL-4 and established an allergy-prone phenotype [54]. Depletion of iNKT cells in a house dust mite–induced asthma model attenuated AHR without affecting pulmonary inflammation or mast cell degranulation [55]. Collectively, these findings strongly support a central role for iNKT cells in the development of asthma in mice.

In humans, the frequency of iNKT cells in PBMCs from asthma patients was reduced [56], whereas the iNKT/NK cell ratio was elevated [57], suggesting migration of peripheral iNKT cells into the lungs during asthma. Although these studies did not employ CD1d tetramers to identify iNKT cells, a separate study using CD1d tetramers reported consistent results, showing that BALF from children with asthma contained a higher frequency of iNKT cells [58]. Akbari et al. [59] further reported that up to 60% of CD4^⁺ T cells in BALF from patients with moderate-to-severe asthma were iNKT cells. However, subsequent studies failed to replicate these findings, raising concerns about potential nonspecific staining or inappropriate flow cytometry gating strategies [60,61]. Later, the Umetsu group demonstrated substantial variability in iNKT cell frequencies depending on disease severity [62]. To date, the causal relationship between iNKT cells and asthma in humans remains to be elucidated [63].

The role of γδ T cells in CRS has also been investigated [64-66]. The density of γδ T cells was significantly higher in E-NP than in other groups [64]. Quantitative reverse transcription PCR (qRT-PCR) analysis of four functional TCR Vγ chains revealed that the relative expression of Vγ1 (TRGV2, TRGV3, TRGV4, TRGV5, TRGV8) was elevated in E-NP, whereas expression of Vγ2 (TRGV9), Vγ3 (TRGV10), and Vγ4 (TRGV11) did not differ significantly among groups. Furthermore, the expression levels of the TCR γ chain positively correlated with tissue eosinophil counts, recurrence rate, and Lund–Kennedy endoscopic score, but not with peripheral blood eosinophil levels.

The activity of γδ T cells, particularly Vγ1^⁺ subsets, may be linked to effector molecules such as eosinophil cationic protein (ECP), which contributes to epithelial injury, and matrix metalloproteinase-7 (MMP-7), a key mediator of tissue remodeling in chronic inflammatory conditions [65]. The mRNA expression levels of ECP and MMP-7 were significantly higher in CRSwNP than in CRSsNP or control tissues and were positively correlated with the expression of the TCR Vγ1 chain. Tissue protein levels of both ECP and MMP-7 were also higher in E-NP than in NE-NP. Additionally, the relative expression of Th2 cytokines, including IL-4, IL-5, and IL-13, was positively correlated with the expression of the TCR Vγ1 chain [66]. In a murine model of E-NP, administration of an anti-TCR Vγ1 antibody reduced eosinophilic infiltration [66]. Collectively, these findings support a role for sinonasal Vγ1⁺ γδ T cells in the pathogenesis of E-NP.

In a BALB/c mouse model of OVA–induced allergic asthma, γδ T cells were shown to play a critical role in disease initiation. Mice lacking γδ T cells exhibited attenuated OVA-specific immunoglobulin (Ig)E and IgG1 responses, along with reduced IL-5 levels and pulmonary eosinophilia [67,68]. These findings were further supported in C57BL/6 mice using asthma models induced by either OVA [69] or Blomia tropicalis [70]. In contrast, Lahn et al. [71] reported that γδ T cells suppressed AHR independently of αβ T cells in OVA-sensitized C57BL/6 mice. Similarly, in a rhinovirus-induced asthma model, γδ T cells were found to exert a regulatory function [72].

Upon asthma induction, γδ T cells predominantly produced Th2-type cytokines [73], particularly through Vγ1⁺ subsets [74]. In the OVA-induced asthma model, lung-infiltrating γδ T cells expressed high levels of IL-17, whereas few IFN-γ– or IL-4–producing γδ T cells were detected [75], possibly due to dysregulated apoptosis [76]. The expansion of Tγδ17 cells was associated with the use of adjuvants and enhanced neutrophil recruitment, which in turn suppressed AHR [77].

The functional subsets of γδ T cells are determined by their Vγ chain usage, and their composition is strongly influenced by the microbiota [78,79]. However, it is now recognized that the expression of specific combinations of transcription factors, rather than TCR usage alone, determines their functional lineages [12]. Therefore, discrepancies across studies may stem from inter-facility differences in microbial composition, which impact γδ T cell subset distribution.

Environmental factors such as air pollutants are also implicated in modulating γδ T cell responses. Air pollutants, including particulate matter and ozone, are known to exacerbate asthma, accounting for approximately 20% of asthma-related emergency room visits [77]. Particulate matter has been shown to induce IL-17 production from γδ T cells, MAIT cells, and iNKT cells in the lungs, thereby promoting neutrophilic inflammation [79]. Ozone exposure in obese mice promoted IL-13–producing suppression of tumorigenicity 2–positive (ST2^⁺) γδ T cells in the lungs via IL-33 signaling, leading to AHR [80].

In humans, γδ T cell frequencies were elevated in BALF from asthma patients [81], whereas their frequencies were decreased in peripheral blood [82]. Interestingly, the proportions of IL-4–producing γδ T cells were increased in both PBMCs and induced sputum [83,84]. Another study reported that although γδ T cell frequencies in BALF from patients with stable asthma were not elevated, these cells produced high levels of Th2 cytokines following in vivo allergen challenge [73]. A subsequent meta-analysis confirmed a significant reduction in γδ T cell frequencies in PBMCs from asthma patients [77]. Unlike findings from murine models, however, recruitment of γδ17 cells has not been observed in human peripheral blood or induced sputum [39,83].

In CRSwNP samples from human patients, innate T cells are increased and/or activated, although minor discrepancies between studies remain (Fig. 2). In murine asthma models, iNKT cells play a critical role, as mice lacking iNKT cells fail to develop AHR. However, the roles of MAIT and γδ T cells in these models remain controversial. In human asthma, the role of iNKT cells is still debated, whereas MAIT17 and Tγδ2 cells have been implicated in disease pathogenesis (Fig. 3). To resolve these inconsistencies, future studies with larger sample sizes and better-controlled experimental designs are needed.

Although MAIT17, NKT2, and Vγ1^⁺ γδ T cells are correlated with airway inflammation and disease severity, specific tools to control their abundance or localization are not yet available. Developing such approaches could yield clinically meaningful biomarkers and therapeutic targets that extend beyond established type 2 immune pathways.