Introduction
Multiple sclerosis (MS) is a central nervous system disorder driven by autoimmune mechanisms, in which inflammatory activity results in demyelination and subsequent axonal impairment ( 1 ). It is one of the most common causes of non-traumatic disability among individuals aged 18–40 years, affecting an estimated 2.8 million people worldwide ( 2 ). Immune imbalance is a major contributor to MS pathogenesis, characterized by dysregulated expansion and infiltration of T and B lymphocytes ( 3 ). Patients during the remission phase are generally managed with disease-modifying therapies (DMTs). DMTs modulate immune cell activity, suppress lymphocyte-mediated responses, reduce relapse rates, and slow short-term disability progression ( 4 ). Some DMTs may also contribute to improved neurological function in MS ( 5 ). However, whether they ameliorate olfactory dysfunction remains unclear.
Olfactory function refers to the ability of an organism to detect and identify odors through the olfactory system ( 6 ). Patients with MS may experience olfactory dysfunction of varying severity, and recent evidence suggests that approximately 27.2% have olfactory impairment ( 7 ). However, the mechanisms underlying olfactory dysfunction in MS remain unclear. Abnormalities may occur at multiple sites along the olfactory pathway, and both olfactory bulb atrophy and neuroinflammation have both been reported in patients with MS ( 8 ). The nasal cavity is the first site of entry for odorant molecules, and the local nasal microbiota plays an important role in the proper development of the olfactory epithelium ( 9 ). Previous studies have reported a reduced abundance of Corynebacterium in patients with postviral olfactory dysfunction, suggesting that nasal microbiota dysbiosis may be associated with impaired olfactory function ( 10 ). The nasal microbiota is closely linked to olfactory function, and whether olfactory dysfunction in MS is associated with alterations in the nasal microbiota remains to be elucidated.
The nasal microbiota is located on the nasal epithelium, a key component of the body’s first line of immune defense ( 11 ). Host immune status and the nasal microbiotainfluence each other ( 12 ). The nasal microbiota contributes to maintaining the balance among Th1, Th2, and regulatory T cells (Tregs) ( 13 ). Colonization of the nasal cavity with Lactobacillus murinus and Streptococcus pyogenes may induce proinflammatory cytokine expression and modify the composition and abundance of circulating lymphocyte subsets ( 14 ). In MS, the gut microbiota has been implicated in shaping circulating lymphocyte profiles. DMT treatment has also been reported to be associated with alterations in the gut microbiota, suggesting potential bidirectional interactions between DMTs and gut microbial communities ( 15 ).
This study aimed to investigate the effects of DMTs on olfactory function in MS and to explore the underlying mechanisms, with a focus on circulating lymphocyte subsets, nasal microbiota, and their interactions.
Materials and methods
Participants
From February 2021 to March 2023, individuals diagnosed with multiple sclerosis (MS) were enrolled at the Affiliated Hospital of Guizhou Medical University. Diagnosis was based on the 2017 revised McDonald criteria, and disability was evaluated using the Expanded Disability Status Scale (EDSS) ( 16 , 17 ). Eligibility criteria included an age range of 18–69 years and an EDSS score of 0–6.5. Baseline clinical characteristics, including sex, age, and and body mass index (BMI), were collected at enrollment.
Exclusion criteria were recent MS relapse (<30 days), comorbid neurological, psychiatric, nasal, or systemic disorders that could affect olfaction, head trauma or MRI-detected structural abnormalities, a history of upper respiratory tract infection within the preceding 3 weeks, prior radiotherapy or chemotherapy, substance dependence, menstruation or pregnancy, prolonged use (>1 month) of drugs that impair olfaction, an EDSS score exceeding 6.5, and an MMSE score below 24.
Ethical approval for this study was granted by the Clinical Ethics Committee of the Affiliated Hospital of Guizhou Medical University (approval Nos. 2021 Lun 495 and 2023 Lun 897). The study was performed in accordance with the principles outlined in the Declaration of Helsinki. All participants provided written informed consent prior to inclusion.
The dataset used in this study partly overlaps with a previously published cohort ( 18 ). However, the present analysis focused specifically on the effects of DMT exposure, which were not a primary focus of the previous publication.
Olfactory assessment methods
Olfactory performance was measured using the University of Pennsylvania Smell Identification Test (UPSIT), a validated method for quantitative assessment ( 19 , 20 ). The validated simplified Chinese 40-item version with microencapsulated odorant strips was used ( 20 ). The UPSIT score was defined as the number of correctly identified odorants (0–40), and age- and sex-adjusted normative criteria were applied to classify the severity of olfactory dysfunction. All tests were administered by two trained investigators ( 20 ). Two trained researchers carried out the testing for all subjects and performed the statistical analyses.
Biospecimen collection
Nasal samples were collected using sterile swabs (LONGSEE, Guangzhou, China), which were inserted approximately 2–3 cm into the middle nasal passage and gently rotated for 5–10 s. After collection, each swab was transferred into 1 mL of sterile saline preservative and stored at −80°C for subsequent analysis.
Sequence analysis
Genomic DNA was extracted from human nasal swabs using the QIAamp DNA Mini Kit (QIAGEN). The V4 region of the 16S rRNA gene was amplified with primers 515F and 806R using Premix Taq (TaKaRa) under the following conditions: 94°C for 5 min; 30 cycles of 94°C 30 s, 52°C 30 s, 72°C 30 s; and 72°C for 10 min. Amplicons were purified and sequenced on the Illumina platform (PE250).
Raw reads were quality-trimmed using fastp (v0.14.1; sliding window: -W 4, -M 20) and primer-trimmed using cutadapt. Paired-end reads were merged with usearch v10 (-fastq_mergepairs; min overlap 16 bp, max mismatch 5 bp). OTUs were clustered at 97% identity using UPARSE, and chimeras, singletons, and sequences assigned to mitochondria or unclassified at the kingdom level were filtered out. Taxonomic classification was performed with SINTAX against the SILVA 138 database at a confidence threshold of 0.8; unclassified taxa at the genus level were labeled at the family level with an “unclassified” suffix.
Functional profiles were predicted using PICRUSt2, with annotations mapped to the KEGG Orthology (KO) and COG databases.
Percentages and absolute counts of lymphocyte subsets
Peripheral venous blood (2 mL) was collected from MS patients into heparinized tubes. For flow cytometry, 100 μL of whole blood was incubated for 30 min in the dark with fluorochrome-conjugated monoclonal antibodies against CD3, CD4, CD8, CD19, and CD16/CD56, including CD3-FITC, CD16/CD56-PE, CD19-APC, CD4-PerCP-Cy5.5, and CD8-BV421 (BD Biosciences, USA). Phosphate-buffered saline was obtained from Solarbio (Beijing, China), and all flow cytometry tubes and instruments were from BD Biosciences (USA).
Statistical analysis
Statistical analyses were performed using SPSS 24.0, GraphPad Prism 8.0.2, and R 3.5.1. Continuous variables were compared using unpaired two-tailed t-tests, and categorical variables were analyzed with the chi-square test or Fisher’s exact test, as appropriate. Pearson and Spearman correlation analyses were applied to assess relationships between clinical variables, lymphocyte subsets, and UPSIT scores. A two-sided p value < 0.05 was considered statistically significant.
For nasal microbiota analysis, α-diversity was evaluated using richness, Chao1, Shannon, and Simpson indices, and Beta diversity was assessed using Bray–Curtis dissimilarity and visualized by PCoA. Differences in community composition among groups were tested by PERMANOVA based on Bray–Curtis distance matrices using the Adonis R 2 . Taxonomic composition was summarized at the phylum and family levels, with taxa of relative abundance < 0.5% and unclassified taxa grouped as “Others”. Differential taxa were identified using LEfSe, and correlations between nasal microbiota and lymphocytes were visualized using heatmaps.
Results
Baseline characteristics
The baseline features of this cohort are outlined here and have been described in detail in our earlier report ( 18 ). The study included 30 individuals with MS and 30 healthy controls (HCs) matched for age and sex. No significant differences were found between MS and HC groups in sex distribution, age, BMI, or MMSE score ( 18 ). MS patients were stratified into untreated (MS(DMTs−), n = 13) and treated [MS(DMTs+), n = 17] groups. A significant difference in sex distribution was observed between subgroups ( p = 0.01); however, other baseline demographic and clinical variables were comparable ( Table 1 ). A subset of participants, comprising 14 healthy controls, 10 MS(DMTs+), and 3 MS(DMTs−) patients, underwent nasal microbiota analysis. The remaining participants were not included because of incomplete sample collection/insufficient sample volume/failure to meet predefined sample quality-control criteria/participant noncompliance ( Figure 1A ). In this study, the V3–V4 region of the 16S rRNA gene was sequenced in 27 nasal samples, generating a total of 3,349,902 raw paired-end reads. After quality filtering, sequence merging, and chimera removal, 3,136,801 high-quality clean tags were retained for downstream analyses. The mean number of clean tags per sample was 115,646 in the HC group, 116,518 in the MS (DMT+) group, and 117,526 in the MS (DMT−) group. The comparable sequencing depths across the three groups indicate that the sequencing data were sufficient for microbial composition and diversity analyses ( Supplementary Table 1 ).
Table 1
Characteristics of MS patients with DMTs and without DMT .
a Data are presented as mean ± standard error of the mean (SEM) or b median (interquartile range, Q1–Q3), as appropriate. * p < 0.05 indicates statistical significance. # Comparisons between groups were performed using the independent-samples t-test for normally distributed variables. & The Mann–Whitney U test ( Z value) was used for non-normally distributed variables. † The chi-square ( χ² ) test was applied for categorical variables. MS, multiple sclerosis; BMI, body mass index; EDSS, Expanded Disability Status Scale; DMTs, Disease Modifying Therapies; UPSIT, University of Pennsylvania Smell Identification Test.
| Characteristics | MS (DMTs-) n=13 | MS (DMTs+) n=17 | Statistical values | p -V value |
| Gender b (Females/Males) | 7/6 | 16/1 | 6.679 † | 0.01 * |
| Age b | 31.00 (28.00-35.50) | 34.00 (29.00-43.50) | -1.175 & | 0.240 |
| BMI b | 22.27 (19.30-24.28) | 20.96 (19.78-22.74) | -0.314 & | 0.754 |
| Age at onset b | 28.00 (20.50-30.50) | 30.00 (23.00-37.00) | -0.943 & | 0.346 |
| Disease duration a | 5.74 ± 1.3 | 4.96 ± 1.03 | 0.503 # | 0.619 |
| Number of relapses a | 2.92 ± 1.61 | 2.65 ± 1.62 | 0.465 # | 0.646 |
| EDSS Score b | 2.00 (1.25-3.25) | 2.00 (1.75-2.75) | -0.149 & | 0.881 |
| Type of DMTs | ||||
| Teriflunomide | 11 | |||
| Siponimod | 4 | |||
| Dimethyl fumarate | 1 | |||
| Rituximab Injection | 1 | |||
| UPSIT Score a | 26.31 ± 0.59 | 29.06 ± 0.82 | -2.574 # | 0.016 * |
Olfactory function
Olfactory function was assessed using UPSIT. The mean UPSIT scores in individuals with MS were significantly reduced compared to those in HCs [27.87 ± 3.10 vs. 29.93 ± 3.84, p = 0.027, 95% CI: (−3.887 to −0.247)] ( 18 ). There was no significant difference between MS (DMTs+) and HCs ( p = 0.437; Figure 1B ), whereas MS(DMTs−) showed significantly lower UPSIT scores than HCs [29.93 ± 3.84 vs. 26.31 ± 0.59, p < 0.001,95% CI: (−5.484 to −1.767), Figure 1C ]. Among patients with MS, those in the MS (DMTs−) group had significantly lower UPSIT scores compared with the MS (DMTs+) group [26.31 ± 0.59 vs. 29.06 ± 0.82, p = 0.016, 95% CI: (0.5614 to 4.941) Table 1 ; Figure 1C ].
Lymphocyte subsets
Lymphocyte subset counts and proportions were compared between MS(DMTs−) and MS(DMTs+). Absolute counts of total T cells [ p = 0.001, 95% CI: (307.009 to 1169.162)], CD8 + T cells [ p = 0.002, 95% CI: (96.554 to 395.743)], CD4 + T cells [ p = 0.001, 95% CI: (214.720 to 807.496)], and B cells ( p < 0.001) were significantly lower in MS(DMTs+) than MS(DMTs−) ( Table 2 ; Figures 2A–C, E ). Absolute NK-cell counts were also lower in MS(DMTs+) [ p = 0.049, 95% CI: (-244.7 to -0.2109) ( Table 2 ; Figure 2F )]. There was no difference in the CD4/CD8 ratio among the groups ( p = 0.739; Table 2 ; Figure 2D ). In contrast, the combined proportion of T+B+NK% was significantly lower in MS(DMTs+) than in MS(DMTs-) ( p = 0.018; Table 2 ). NK-cell proportions tended to be higher in MS(DMTs+), though not significantly ( p = 0.098; Figures 2G, H ). MS (DMTs+) patients exhibited lower proportions of total T cells, CD8 + T cells, CD4 + T cells, and B cells, but these differences were not significant ( Figures 2G, H ).
Table 2
Lymphocyte subgroup test results of MS patients.
a Data are presented as mean ± standard error of the mean (SEM) or b median (interquartile range, Q1–Q3), as appropriate. * p < 0.05, ** p < 0.01, *** p < 0.001 indicate statistical significance. † The Mann–Whitney U test ( Z value) was used for non-normally distributed variables. All other continuous variables were analyzed using the independent-samples t -test. MS, multiple sclerosis; DMTs, disease-modifying therapies.
| Characteristics | MS (DMTs-) n=13 | MS (DMTs+) n=17 | Statistical values | p -V value |
| T Cells a | 1661.62 ± 173.85 | 923.53 ± 127.48 | 3.507 | 0.001 ** |
| CD8 + T Cells a | 596.39 ± 52.43 | 350.23 ± 49.68 | 3.371 | 0.002 ** |
| CD4 + T Cells a | 976.46 ± 138.57 | 465.35 ± 69.74 | 4.940 | 0.001 ** |
| NK Cells a | 332.39 ± 53.35 | 209.94 ± 21.90 | 2.123 † | 0.049 * |
| B Cells b | 241.00 (187.00-380.00) | 138.00 (45.00- 188.00) | -3.327 † | <0.001 *** |
| CD4/8 a | 1.72 ± 0.23 | 1.60 ± 0.26 | 0.336 | 0.739 |
| T+B+NK % b | 99.84(99.75-99.93) | 99.73 (99.50-99.81) | -2.349 † | 0.018 * |
| T Cells % a | 72.89 ± 1.88 | 68.55 ± 3.25 | 1.064 | 0.296 |
| CD8 + TCells % a | 28.15 ± 2.99 | 26.99 ± 2.25 | 0.314 | 0.756 |
| CD4 + T Cells % a | 41.31 ± 2.57 | 34.49 ± 3.24 | 1.573 | 0.127 |
| NK Cells % a | 14.34 ± 1.88 | 20.16 ± 2.73 | -1.712 | 0.091 |
| B Cells % a | 12.64 ± 1.29 | 10.12 ± 1.34 | 1.318 | 0.198 |
UPSIT and lymphocyte subsets
Correlation analysis was used to examine the relationship between olfactory performance and lymphocyte subsets in MS. In the overall MS group, UPSIT scores were not significantly correlated with NK-cell absolute counts ( r = -0.164, p = 0.386; Figure 2I ) or in the MS(DMTs−) subgroup ( r = -0.051, p = 0.868; Figure 2J ). By comparison, the MS (DMTs+) subgroup showed a significant inverse correlation ( r = −0.527, p = 0.029; Figure 2K ). There were no significant correlations between UPSIT scores and the absolute numbers of total T cells, CD8 + T cells, CD4 + T cells, or B cells ( Supplementary Figure 1 ).
Nasal microbiota
α- and β-diversity
For α-diversity, Chao1 richness ( p = 0.753), Shannon index( p = 0.104), and Simpson index( p = 0.145) were not significant in HCs and overall MS 17 .Chao1 richness did not differ significantly among HCs vs. MS(DMTs+); HCs vs. MS(DMTs−); and MS(DMTs+) vs. MS(DMTs−) (all p >0.05) ( Figure 3A ). Conversely, healthy controls (HCs) showed a significantly higher Shannon index compared with the MS (DMTs−) group ( p = 0.038), whereas not significant in the HCs vs. MS(DMTs+), and MS(DMTs+) vs. MS(DMTs−),(all p >0.05); ( Figure 3B ). The Simpson index was significantly lower in HCs than in MS(DMTs−) ( p = 0.023), but not in the other comparisons (HCs vs. MS(DMTs+), MS(DMTs+) vs. MS(DMTs−), all p >0.05); ( Figure 3C ).
For β-diversity, PERMANOVA confirmed no significant difference in HCs vs. MS ( R² = 0.0316, p = 0.658; Figure 3D ). After stratification by DMT exposure, no significant differences were detected among all the following groups: HCs vs. MS(DMTs+), HCs vs. MS(DMTs−), MS(DMTs+) vs. MS(DMTs−) (all p > 0.05) ( Figure 3D ). Bray-Curtis-based PCoA showed no significant separation between HCs, MS(DMTs+)and MS(DMTs−) ( Figure 3E ) (all p >0.05).
Differential abundance
At the phylum level, we presented the top 20 abundant taxa, comparing HC vs. MS ( Figures 4A left), MS(DMT+) vs. MS(DMT−) ( Figures 4A middle), and HC vs. MS(DMT−) ( Figures 4A right). Differential analysis revealed Proteobacteria was more abundant in MS(DMTs−) than HCs ( p = 0.004, 95% CI [-0.1708 to -0.040], two-sided unpaired Welch’s t test; Figure 4B ).
At the family level, we presented the top 20 abundant taxa, comparing HC vs. MS ( Figures 4C left), MS(DMT+) vs. MS(DMT−) ( Figures 4C middle), and HC vs. MS(DMT−) ( Figures 4C right). Statistical analysis using the Wilcoxon rank-sum test showed Moraxellaceae ( p = 0.037, 95% CI [-0.055 to -0.002]) and Enterobacteriaceae ( p = 0.033, 95% CI [-0.028 to -0.001]) were significantly less abundant in MS(DMTs−) than HCs ( Figure 4D ). Comparisons among the three groups at the phylum and family levels are presented in Supplementary Figure 2 . OTU profiles in HCs and MS(DMT−) were similar overall and clustered together ( Figure 4E ), while MS(DMT+) formed a separate branch. Several OTUs (e.g., Corynebacterium, unclassified Enterobacteriaceae ) were enriched in untreated MS patients but decreased after DMT treatment. With comparable class- and genus-level community structures ( Figure 4F ).
Differential taxa and predicted functions
LEfSe analysis identified differential taxa across groups. In the HC vs. MS(DMTs+) comparison, 11 taxa were significantly different (LDA score > 2.0, p < 0.05), with 9 ( Acetoanaerobium sp., Bacillus sp. TX3, Flavihumibacter cheonanensis , Leptotrichia sp. PG10, Marinococcus sp. GSP32, Nocardioides sp. S2_84, Paenibacillus sp. YJ01, Selenomonas lacticifex , Sphingobacterium thalpophilum )enriched in MS(DMTs+) and 2( Azospirillum sp. CRRI_31, Sphingomonas sp. 2D232011) in HCs (all p < 0.05) ( Figure 5A ). In HC vs. MS(DMTs−), 16( Bacillus sp. TX3, Chromobacterium alkanivorans , Chryseobacterium sp. LW_NC3, Jaaginema geminatum SAG 1459-8, Lachnospiraceae bacterium 610, Lagierella massiliensis , Marinococcus sp. GSP32, Microlunatus lacustris , Ndongobacter massiliensis , Nevskia ramosa , Pleurocapsa sp. HA4230, Polyangium sp. CC_YHH645, Sphingobacterium thalpophilum , uncultured Actinomyces sp., uncultured Hymenobacter group bacterium, uncultured actinobacterium) different taxa were detected, most enriched in MS(DMTs−), indicating more pronounced microbial changes in untreated patients (all p < 0.05) ( Figure 5A ). The MS(DMTs+) vs. MS(DMTs−) comparison identified 4 different taxa( Bacteroidetes bacterium Oral Taxon 274, Chryseobacterium sp. LW_NC3, Lagierella massiliensis , Nesterenkonia sp. MCCC 1A10687), all enriched in MS(DMTs−), suggesting partial modulation of microbial composition by DMT treatment (all p < 0.05) ( Figure 5A ).
Functional prediction based on KEGG showed that, compared with HCs the MS(DMTs−) had a significantly higher relative abundance of the valine, leucine and isoleucine biosynthesis pathway ( p = 0.026; Figure 5B ). The other pathways showed no statistically significant differences. Functional prediction based on COG databases ( Supplementary Figure 3 ).
Associations between lymphocyte subsets and the nasal microbiota
Spearman correlation heatmaps were used to assess relationships between nasal microbial genera and peripheral immune cell subsets in MS and MS(DMTs+) patients ( Figures 5C, D ). Color intensity reflects the strength of correlations (red indicating positive and blue indicating negative), while statistically significant associations are marked (* p < 0.05; ** p < 0.01). Campylobacter, Anaerococcus, Peptoniphilus, Prevotella , and Finegoldia exhibited positive correlations with NK cell count(all p <0.05). These taxa were uniquely associated with NK cells in MS (DMT+) compared to the MS group analyzed earlier. No significant associations were observed in the MS(DMTs−) subgroup, which may reflect the limited sample size of the untreated group ( Supplementary Figure 4 ).
Discussion
This study provides evidence, for the first time, that DMTs may be associated with improvements in olfactory function in patients with MS, accompanied by coordinated changes in immune status and the nasal microbiota. These findings extend the recognised immunomodulatory effects of DMTs and suggest a microbiota-related regulatory pathway involved in olfactory recovery.
MS patients receiving DMT scored higher on the UPSIT compared to untreated patients, suggesting that olfactory function could serve as a sensitive indicator for monitoring treatment effects. This also helps explain the wide variation in the prevalence of olfactory dysfunction observed in previous studies, with reported rates ranging from 27.2% to 93% ( 7 , 21 ). While methodological and clinical differences contribute to this, our findings suggest DMT treatment could be an overlooked confounding factor. Future assessments of olfactory dysfunction in MS should consider DMT use, treatment type, and duration.
Previous studies suggest that changes in olfactory function in MS may reflect inflammatory status and serve as potential biomarkers of treatment response ( 22 ). The olfactory improvement associated with DMTs may involve multiple mechanisms. DMTs could reduce excessive immune activation and pro-inflammatory mediators ( 23 ), thereby alleviating neuroinflammatory damage in the olfactory pathways, although this mechanism requires further validation. DMTs may indirectly shape the nasal microbiome through immune modulation. Consistently, untreated patients showed significant changes in microbial diversity and composition compared with healthy controls. Compared with healthy controls, untreated MS patients exhibited a significantly decreased Shannon index and an increased Simpson index. The Shannon index comprehensively evaluates alpha diversity by integrating species richness and community evenness; its reduction in untreated MS patients indicates a diminished pool of microbial species and less evenly distributed taxa within the nasal cavity. In contrast, the Simpson index is more sensitive to dominant microbial populations, and its elevation reflects the over-proliferation of a small number of dominant strains and imbalanced community distribution. These concordant shifts in the two diversity indices jointly demonstrate reduced richness and skewed dominance of the nasal microbiota in untreated MS patients. The elevated levels of Proteobacteria may indicate immune activation and ongoing inflammation. Proteobacteria has been shown to significantly increase in nasal inflammatory diseases such as chronic rhinosinusitis ( 24 ). The reduction of Moraxellaceae and Enterobacteriaceae may indicate depletion of commensal bacteria that sustain physiological nasal mucosal homeostasis under healthy conditions. Moraxellaceae constitute core commensal taxa of the upper respiratory tract, and their abundance exhibits bidirectional correlations with local airway inflammation and mucosal immune status: physiological colonization maintains immune balance, whereas excessive proliferation drives inflammatory lesions. A recent large-scale nasal microbiome genome-wide association study demonstrated that host genetic variation and multiple host-related and lifestyle factors contribute to variation in nasal microbial community composition ( 25 ). In addition, Moraxella catarrhalis has been shown to induce nasal airway epithelial damage and increase the expression of the pro-inflammatory cytokines IL-8 and IL-33, supporting a potential role for this species in contributing to mucosal inflammatory responses ( 26 ). Alterations in Enterobacteriaceae are frequently reported as a hallmark of nasal microbial dysbiosis ( 27 ). These two bacterial families are tightly associated with the immune microenvironment of nasal mucosa. In our study, these differences were no longer observed after DMT treatment, suggesting that DMT may reshape the immune microenvironment by restoring nasal microbial homeostasis. Immune dysregulation may contribute to dysbiosis of the upper respiratory microbiota, whereas a well-regulated mucosal immune environment helps maintain microbial homeostasis ( 28 ). Conversely, a balanced microbial community may help preserve olfactory epithelial homeostasis and support its repair and regeneration through microbe–host interactions, including bioactive metabolites and regulation of local inflammation ( 29 ). Functional prediction analyses revealed metabolic adaptation in untreated MS patients, as evidenced by the enrichment of branched-chain amino acid (BCAA) biosynthesis pathways. As BCAAs can influence both microbial community structure and Th17 cell differentiation via mTOR signaling, this finding may indicate a shift toward a pro-inflammatory metabolic profile ( 30 ). Consistent with the above metabolic and microbial findings, our prior research has identified two key features in patients with multiple sclerosis: significant shifts in nasal commensal composition, as well as a broken correlation between olfactory function (quantified by UPSIT scores) and superior turbinate morphological volume. Together, these earlier observations indicate that disrupted crosstalk between nasal mucosa and resident microbiota is maybe one of critical driver of MS-related olfactory loss ( 18 ). Building on this evidence, we hypothesize that DMTs exert reciprocal regulatory impacts on both nasal mucosal immune homeostasis and upper respiratory microbial communities. DMT treatment was associated with reductions in T cells, CD4 + T cells, CD8 + T cells, B cells and NK cells, indicating broad immunomodulatory effects. However, the included DMTs may differ in their effects on NK cells. Most DMT-treated participants received teriflunomide, which primarily inhibits activated T- and B-cell proliferation and is generally considered to have a relatively limited direct effect on NK cells. Siponimod may reduce circulating NK-cell counts, whereas dimethyl fumarate and rituximab are also thought to exert comparatively limited direct effects on NK cells. Therefore, the observed NK-cell alterations may not be fully attributable to DMT exposure alone. Associations between multiple nasal taxa and T cells, CD8 + T cells, B cells and NK cells, observed in both the overall MS cohort and the DMT-treated subgroup, further suggest an interaction between peripheral immune status and the nasal microenvironment. Notably, in the DMT-treated group, NK cell counts were inversely associated with UPSIT scores, while Campylobacter, Anaerococcus, Peptoniphilus, Prevotella and Finegoldia showed positive associations with NK cells, supporting a potential role for NK cell-related immune status in linking specific nasal taxa to olfactory function. Previous studies offer partial support for these observations, with Campylobacter linked to poorer olfactory thresholds ( 31 ) and Anaerococcus enriched in anosmic individuals ( 32 ). In contrast, evidence for Peptoniphilus remains limited, and Finegoldia has not been directly associated with olfactory scores, although its pro-inflammatory properties suggest possible indirect relevance. Although direct evidence that NK cells mediate olfactory epithelial injury is lacking, activated NK cells may induce mucosal epithelial damage through granzyme B release and perforin-dependent cytotoxicity ( 33 ). These findings suggest that NK cells may represent a potential immunological link between DMT treatment, nasal microbial changes and olfactory dysfunction, although the mechanisms involved and the roles of distinct NK cell subsets remain to be clarified. Nevertheless, because the individual DMT subgroups were small and unevenly distributed, particularly for siponimod, dimethyl fumarate and rituximab, treatment-specific effects on NK cells could not be evaluated, and residual confounding by DMT exposure cannot be excluded.
Several limitations of this study should be acknowledged. The sample sizes of several subgroups were small. The untreated MS group included only three patients; therefore, findings related to this subgroup should be interpreted cautiously and regarded as exploratory. The dimethyl fumarate and rituximab subgroups each included only one patient, precluding reliable drug-specific conclusions and highlighting the need for validation in larger, adequately powered cohorts. Although available immunological evidence suggests that the DMTs included in this study have limited direct effects on NK cells, supporting our interpretation that nasal immune alterations may be driven primarily by microbial remodelling rather than direct drug-induced immunosuppression, this hypothesis requires confirmation in larger studies with more balanced representation across DMT subgroups.
In summary, DMT treatment may be associated with improved olfactory function in MS, accompanied by changes in peripheral immunity and the nasal microbiota. The inverse correlation between NK cell counts and UPSIT scores, along with their associations with specific nasal taxa, suggests a potential link among immune modulation, microbial balance, and olfactory function. Further studies are needed to clarify the underlying mechanisms.
Acknowledgments
We sincerely thank all participants for their active involvement and commitment to this clinical study. We also appreciate the medical writing support provided by Zhang Yang (Department of Neurology, Affiliated Hospital of Guizhou Medical University, Guizhou Medical University, Guiyang, China).
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Natural Science Foundation of China (Grant Nos. 82371330 and 82460251), the Key Advantageous Discipline Construction Project of the Guizhou Provincial Health Commission (2023), the Guizhou Provincial Clinical Medical Research Center Construction Project for Neurological Diseases [No. LCZX(2025)003], the 2025 Guizhou Provincial Merit-Based Innovation and Entrepreneurship Funding Program for Overseas Returnees [No. (2025)17], Guizhou Provincial Science and Technology Projects [Qiankehe Foundation-ZK (2023) General 393], the Medical Research Union Fund for High-Quality Health Development of Guizhou Province (2024GZYXK YJJXM0103), the Major and Refractory Disease Clinical Collaboration Pilot Project of Integrated Traditional Chinese and Western Medicine for Cerebral Infarction, and the Doctoral Research Initiation Fund of the Affiliated Hospital of Guizhou Medical University (No. gyfybsky-2022-45).
Footnotes
Edited by: Valentyn Oksenych , University of Bergen, Norway
Reviewed by: Varun Suroliya , Artemis Hospitals, India
Shuqi Li , Harvard Medical School, United States
Data availability statement
The raw 16S rRNA gene amplicon sequencing data generated in this study have been deposited in the National Microbiology Data Center (NMDC) under accession number, NMDC10020881. ( https://nmdc.cn/resource/genomics/project/detail/NMDC10020881 ).
Ethics statement
The studies involving humans were approved by Clinical Ethics Committee of the Affiliated Hospital of Guizhou Medical University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
ZD: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. ZY: Formal analysis, Funding acquisition, Writing – review & editing. LC: Funding acquisition, Resources, Writing – review & editing. ZG: Funding acquisition, Investigation, Project administration, Supervision, Visualization, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. Artificial intelligence-assisted tools were used solely for language polishing and the preparation of the graphical abstract. Figure 1A flowchart was generated by AI, then manually revised and refined for professionalism by the authors. All scientific content, data interpretation, conclusions, and final manuscript revisions were performed and verified by the authors. The authors take full responsibility for the accuracy and integrity of the manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1889149/full#supplementary-material
References
1. Guan S, Zhu H, Zhang M, Shi F-D, Yan B. Cell death in multiple sclerosis. Cell Death Diff. (2025) 33:433–46. doi: 10.1038/s41418-025-01576-7
2. Jakimovski D, Bittner S, Zivadinov R, Morrow SA, Benedict RH, Zipp F, et al. Multiple sclerosis. Lancet (Lond England). (2023) 403:183–202. doi: 10.1016/S0140-6736(23)01473-3
3. Ochs J, Nissimov N, Torke S, Freier M, Grondey K, Koch J, et al. Proinflammatory CD20+ T cells contribute to CNS-directed autoimmunity. Sci Transl Med. (2022) 14:eabi4632. doi: 10.1126/scitranslmed.abi4632
4. McGinley MP, Goldschmidt CH, Rae-Grant AD. Diagnosis and treatment of multiple sclerosis: A review. JAMA. (2021) 325:765–79. doi: 10.1001/jama.2020.26858
5. Zhong M, van der Walt A, Campagna MP, Stankovich J, Butzkueven H, Jokubaitis V. The pharmacogenetics of rituximab: Potential implications for anti-CD20 therapies in multiple sclerosis. Neurother J Am Soc For Exp Neurother. (2020) 17:1768–84. doi: 10.1007/s13311-020-00950-2
6. Bushdid C, Magnasco MO, Vosshall LB, Keller A. Humans can discriminate more than 1 trillion olfactory stimuli. Sci (New York NY). (2014) 343:1370–2. doi: 10.1126/science.1249168
7. Mirmosayyeb O, Ebrahimi N, Barzegar M, Afshari-Safavi A, Bagherieh S, Shaygannejad V. Olfactory dysfunction in patients with multiple sclerosis; A systematic review and meta-analysis. PloS One. (2022) 17:e0266492. doi: 10.1371/journal.pone.0266492
8. Shin T, Kim J, Ahn M, Moon C. Olfactory dysfunction in CNS neuroimmunological disorders: A review. Mol Neurobiol. (2018) 56:3714–21. doi: 10.1007/s12035-018-1341-0
9. Fatuzzo I, Niccolini GF, Zoccali F, Cavalcanti L, Bellizzi MG, Riccardi G, et al. Neurons, nose, and neurodegenerative diseases: Olfactory function and cognitive impairment. Int J Mol Sci. (2023) 24. doi: 10.3390/ijms24032117
10. Biswas K, Ramakrishnan VR, Hollemann E, Lorenz K, Wagner Mackenzie B, Frank DN, et al. Bacterial communities in the nasal passage of postviral olfactory dysfunction patients. Int Forum Allergy Rhinol. (2023) 13:1962–5. doi: 10.1002/alr.23149
11. Ridley C, Thornton DJ. Mucins: The frontline defence of the lung. Biochem Soc Trans. (2018) 46:1099–106. doi: 10.1042/BST20170402
12. Bonillo-Lopez L, Carmona-Vicente N, Tarrés-Freixas F, Kochanowski K, Martínez J, Pérez M, et al. Porcine nasal organoids to model interactions between the swine nasal microbiota and the host. Microbiome. (2025) 13(1):131. doi: 10.1186/s40168-025-02088-9. bioRxiv (Cold Spring Harbor Laboratory). (2024) 13(1):131. doi: 10.1101/2024.08.13.606910
13. Panpan Z, Jinli H, Qiuhong L, Bo D, Juan Z, Hui S, et al. Changes in respiratory tract and gut microbiota in AR mice and their relationship with Th1/Th2/Treg. Microb Pathogen. (2024) 195:106881. doi: 10.1016/j.micpath.2024.106881
14. Costalonga M, Cleary PP, Fischer LA, Zhao Z. Intranasal bacteria induce Th1 but not Treg or Th2. Mucosal Immunol. (2008) 2:85–95. doi: 10.1038/mi.2008.67
15. Pilotto S, Zoledziewska M, Fenu G, Cocco E, Lorefice L. Disease-modifying therapy for multiple sclerosis: Implications for gut microbiota. Multiple Sclerosis Related Disord. (2023) 73:104671. doi: 10.1016/j.msard.2023.104671
16. Hartung H-P, Graf J, Aktas O, Mares J, Barnett MH. Diagnosis of multiple sclerosis: Revisions of the McDonald criteria 2017 - continuity and change. Curr Opin Neurol. (2019) 32:327–37. doi: 10.1097/WCO.0000000000000699
17. Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. (1983) 33(11):1444–52.
18. Gao Z, Danzhen Z, Li Y, Zhu J, Chu L, Yang Z. Exploring nasal structural-microbial interactions in multiple sclerosis-associated olfactory impairment. Ann Clin Transl Neurol. (2025) 12:2546–57. doi: 10.1002/acn3.70181
19. Doty RL, Shaman P, Dann M. Development of the University of Pennsylvania Smell Identification Test: a standardized microencapsulated test of olfactory function. Physiol Behav. (1984) 32:489–502. doi: 10.1016/0031-9384(84)90269-5
20. Devanand DP, Liu X, Cohen H, Budrow J, Schupf N, Manly J, et al. Long-term test-retest reliability of the UPSIT in cognitively intact older adults. Chem Senses. (2019) 44:365–9. doi: 10.1093/chemse/bjz025
21. Printza A, Boziki M, Valsamidis C, Bakirtzis C, Constantinidis J, Grigoriadis N, et al. Smell as a disease marker in multiple sclerosis. J Clin Med. (2022) 11. doi: 10.3390/jcm11175215
22. Berek K, Hegen H, Auer M, Barket R, Di Pauli F, Hocher J, et al. Odour discrimination and identification as a biomarker of long-term disability worsening in multiple sclerosis. Multiple Sclerosis (Houndmills Basingstoke England). (2023) 30:55–62. doi: 10.1177/13524585231201093
23. Pardo G, Jones DE. The sequence of disease-modifying therapies in relapsing multiple sclerosis: Safety and immunologic considerations. J Neurol. (2017) 264:2351–74. doi: 10.1007/s00415-017-8594-9
24. Shi P, Wei H, Liu X, Dong S, He S, Zeng Y, et al. The nasal bacteria microbiome comparison among fungal ball sinusitis, chronic sinusitis with polyps. Indian J Microbiol. (2023) 63:120–8. doi: 10.1007/s12088-023-01062-z
25. Liu X, Tong X, Zou L, Ju Y, Liu M, Han M, et al. A genome-wide association study reveals the relationship between human genetic variation and the nasal microbiome. Commun Biol. (2024) 7:139. doi: 10.1038/s42003-024-05822-5
26. McCauley K, Durack J, Valladares R, Fadrosh DW, Lin DL, Calatroni A, et al. Distinct nasal airway bacterial microbiotas differentially relate to exacerbation in pediatric patients with asthma. J Allergy Clin Immunol. (2019) 144:1187–97. doi: 10.1016/j.jaci.2019.05.035
27. Zhang Y, Yuan F, Liu Z, Huang X, Hong J, Chang F, et al. Rare constituents of the nasal microbiome contribute to the acute exacerbation of chronic rhinosinusitis. Inflammation Res Off J Eur Histamine Res Soc [et al]. (2025) 74:14. doi: 10.1007/s00011-025-01995-9
28. Özçam M, Lynch SV. The gut-airway microbiome axis in health and respiratory diseases. Nat Rev Microbiol. (2024) 22:492–506. doi: 10.1038/s41579-024-01048-8
29. Lazarini F, Roze E, Lannuzel A, Lledo P-M. The microbiome-nose-brain axis in health and disease. Trends Neurosci. (2022) 45:718–21. doi: 10.1016/j.tins.2022.08.003
30. Kang YJ, Song W, Lee SJ, Choi SA, Chae S, Yoon BR, et al. Inhibition of BCAT1-mediated cytosolic leucine metabolism regulates Th17 responses via the mTORC1-HIF1α pathway. Exp Mol Med. (2024) 56:1776–90. doi: 10.1038/s12276-024-01286-z
31. Koskinen K, Reichert JL, Hoier S, Schachenreiter J, Duller S, Moissl-Eichinger C, et al. The nasal microbiome mirrors and potentially shapes olfactory function. Sci Rep. (2018) 8:1296. doi: 10.1038/s41598-018-19438-3
32. Biswas K, Wagner Mackenzie B, Ballauf C, Draf J, Douglas RG, Hummel T. Loss of bacterial diversity in the sinuses is associated with lower smell discrimination scores. Sci Rep. (2020) 10:16422. doi: 10.1038/s41598-020-73396-3
33. Lim YS, Lee AG, Jiang X, Scott JM, Cofie A, Kumar S, et al. NK cell-derived extracellular granzyme B drives epithelial ulceration during HSV-2 genital infection. Cell Rep. (2023) 42:112410. doi: 10.1016/j.celrep.2023.112410
Associated Data
Supplementary Materials
Data Availability Statement
The raw 16S rRNA gene amplicon sequencing data generated in this study have been deposited in the National Microbiology Data Center (NMDC) under accession number, NMDC10020881. ( https://nmdc.cn/resource/genomics/project/detail/NMDC10020881 ).