Tropical urban environments reveal a strong association of CD45RBloCD161+ Th2 subset to allergic rhinitis To the Editor:Allergic airway diseases such as allergic rhinitis (AR) affects more than 400 million individuals worldwide and afflicts substantial health and economic morbidity. [1] AR is strongly associated with a type 2 response, characterized by the cytokines IL-5, IL-4 and IL-13. However, the key drivers behind AR immunopathogenesis remained to be elucidated. This study aims to identify critical pathogenic cell populations associated with AR using the Singapore System Immunology Cohort (SSIC) [2] and a clinician-diagnosed paediatric cohort with active AR manifestation (Supplementary Table 1 ). In both cohorts, the eosinophilic nature of AR was confirmed by higher blood eosinophil numbers (Supplementary Figure 1 ).Whole blood gene expression analysis revealed a total of 23 probes representing 20 unique genes were associated with AR in the SSIC (Table 1A ). To account for ethnicity and environmental influences we validated our findings in BAMSE population-based cohort comprising of Swedish adolescents. Table 1B shows 11 DEGs which was also associated with AR, confirming the transferability of our findings to other populations. For the top DEGs that reached nominal significance in the SSIC we performed an Ingenuity Pathway Analysis (IPA). Supplementary Table 2 revealed important pathways related to hypersensitivity and inflammation and also functional enrichment for eosinophils, basophils and mast cells. In particular, functional activation of Th2 was highlighted as a key pathway for AR pathogenesis. As CRTH2 was reported to be expressed by cell types involved in the eosinophilic response, [3] an unsupervised cluster analysis was performed on the CRTH2+ subset in PBMC of individuals from SSIC (Figure 1 A and B ) to determine CRTH2+ subsets associated with AR. We found that CD161+Th2 subsets in particularly to be strongly associated with AR (Figure 1C and D ) (Supplementary Figure 2 ). Further characterization found that the marker CD45RB to be significantly downregulated on CD161+Th2 cells of AR individuals (Figure 1E ). Low CD45RB expression on T cells is indicative of a mature phenotype. Interestingly, significantly higher circulatory plasma IL-5 levels (Figure 1F) . Furthermore we could also demonstrate AR individuals produced significantly higher IL-5 in an in vitro PMA-stimulation assay (Figure 1G ).While we noted a small population of IL-5 secreting conventional CD161-Th2 (cTh2), IL-5 secretion was significantly elevated in CD161+Th2 cells (Figure 1H ). Strikingly, IL-5 was found to be predominantly secreted by CD45RBlo subset in both cTh2 and CD161+Th2 (Figure 1H and I ). There was also a significant increase in the IL-5 producing CD45RBloCD161+Th2 population from the AR individuals (Figure 1I ). These findings confirm CD45RBloCD161+Th2 as the main producers of IL-5.We further validated our findings in a second paediatric cohort with clinically diagnosed active AR manifestations. To further refine CD161+Th2 subset that is associated with AR, we performed unsupervised PhenoGraph and UMAP clustering on CD161+Th2 (Supplementary Figure 3A and B ). Amongst the UMAP clusters, “cluster 3” was found to be significantly associated with active AR (Supplementary Figure 3C and D ). Deep characterization reveals “cluster 3” to be an IL-5 secreting CD45RBlopopulation, confirming our earlier observation (Supplementary Figure 3E ). Furthermore, this cluster appeared to be a highly differentiated population of mature CD161+Th2 cells with an activated phenotype secreting IL-2, IL-3, IL-4, IL-9 and IL-13 concomitantly (Supplementary Figure 3E and F ). Thus, the severity of eosinophilic airway allergies such as AR seems to be driven by an activated terminally differentiated CD161+Th2 subset that is able to secrete a complex set of inflammatory cytokines.The presence of CD45RBloCD161+Th2 population in both cohorts shows the persistence and pertinence of this population in the pathogenesis of AR. Both cohorts described in this study were collected in Singapore, whereby majority of the individuals are sensitized against HDM. HDM is a perennial allergen in tropical nations such as Singapore, thus T cells in atopic individuals undergo constant stimulation. This could explain the strong association observed between CD45RB expression on CD161+Th2 cells and atopy markers despite the fact that not all subjects demonstrated active AR symptoms during the collection of SSIC cohort. Taken together, our current study unifies the markers previously reported for allergic-specific Th2 subsets and provides clarity for the pathogenic Th2 subset previously reported in different allergic diseases.[4-6] Neutralizing the CD45RBloCD161+Th2 subset should disrupt the allergic response pathway, thus providing a target for lasting therapeutic interventions. Moreover, these cells may also be leveraged as a biomarker for the effectiveness of immunotherapy as well as a potential biomarker of public health surveillance of allergic individuals.

Background: Allergic rhinitis (AR) is strongly associated with a type 2 response, characterized by the cytokines IL-5, IL-4 and IL-13. Several studies have implicated ILC2 and TH2A (CD161+ TH2) but it is not yet entirely clear which subsets are driving the common allergic reactions underlying AR. The objective of this study aims to identify critical pathogenic cell populations associated with AR and to determine their phenotype and functional contribution to disease progression. Methods: We identified integral allergic-specific cell types by transcriptomic sequencing. Whole blood, PBMCs and plasma from a cross-sectional cohort of 216 individuals were analysed by 9-colour flow cytometry and ultra-sensitive cytokine bead arrays using unsupervised clustering algorithms. Clinically active AR cases were further analysed by functional mass cytometry to define phenotype and cytokine secretion (IL-2, IL-3, IL-4, IL-5, IL-9, IL13, IL-17A and IL-22) as well as the expression of the hematopoietic prostaglandin D synthetase (HPGDS). Results: The unbiased analysis revealed that atopy and AR manifestation corelated only with eosinophils, plasma IL-5 and CD161+ TH2 cells. In-depth characterization further revealed that the CD45RB lo CD161+ TH2 subset were most closely associated with severity. This subset is able to concomitantly secrete multiple cytokines including IL-5, IL-13 and IL-4 and has been previously reported to be associated with other eosinophilic allergies. Conclusion: CD45RB lo CD161+ TH2 have key roles in driving the allergic response in AR. Neutralizing the CD45RB lo CD161+ TH2 subset should disrupt the allergic response pathway, thus providing a target for lasting therapeutic interventions.

Transcriptome changes during peanut oral immunotherapy and omalizumab treatmentTo the Editor,Peanut allergy is a common food allergy and the main cause of anaphylaxis among children1. In recent years, oral immunotherapy has emerged as a promising treatment for children with different IgE-mediated food allergies, although safety issues must be considered2. The main aim of immunotherapy is to induce tolerance or desensitization to an allergen which otherwise causes an allergic reaction. For oral immunotherapy this means ingesting the allergen in a controlled manner with gradually increasing dosages. Specifically, peanut oral immunotherapy (pOIT) is able to induce tolerance/desensitization3. While the pathogenesis of food allergy in general is relatively well-studied4, mechanisms of OIT-induced tolerance are not well understood. Omalizumab (anti-IgE) used as treatment for severe allergic asthma and other IgE-driven allergies, can facilitate OIT initiation5, however, little is known about the involved mechanisms, including possible changes at the transcriptional level. We therefore investigated transcriptional changes in whole blood using RNA-sequencing profiles during omalizumab treatment and pOIT in participants from the FASTX (Food Allergen Suppression Therapy with Xolair ®) study previously described in detail elsewhere5.In brief, peanut-allergic adolescents (n=23 of whom 17 completed the study, age 12-18 years) were started on omalizumab (baseline) and treated for at least 8 weeks before starting pOIT (pOIT start) while on omalizumab. The peanut-dose was gradually increased during the 8 weeks until reaching a maintenance dose. Guided by a basophil activation test (BAT/CD-sens)6 after 8 weeks on the maintenance dose, participants decreased the omalizumab dose by 50% (maintenance) and continued to decrease the omalizumab dose if pOIT was tolerated. Eleven patients were able to tolerate pOIT without omalizumab protection for >8 weeks and then passed an open peanut food challenge (final); 6 patients could not discontinue omalizumab, but blood samples were obtained for analysis after 2-3 years of omalizumab treatment (final); 6 patients dropped out of the study. RNA-sequencing was performed on whole blood at baseline, pOIT start, maintenance and final time-points using the NovaSeq 6000 platform. DESeq2 was used for differential expression analysis of the omalizumab effect and a linear mixed-effect model for analyses during pOIT in combination with omalizumab (pOIT+O) after adjustment for treatment outcome and cell type. A complete description of the treatment protocol and method is given in Appendix S1.General characteristics of the study participants at baseline can be found in Table S1 . To elucidate if omalizumab treatment alone induced alterations in peripheral blood gene expression, we investigated the two first timepoints, baseline and pOIT start, however no significant differences were observed (Figure S1 ). In the longitudinal analysis (pOIT start to final), 680 genes associated with pOIT+O at nominal p <0.005 (Table S2 ). The Gene Ontology (GO) biological process of the up- and down-regulation of these 680 genes are presented in Figure 1A,B . Upregulation of 337 genes were linked to GO terms “protein regulation and modification”, while “neutrophil degranulation, immune response, phagocytosis, and metabolic process” were among the top terms for the downregulated 343 genes. Out of the 680 genes, 16 were differentially expressed at false discovery rate (FDR) adjusted p<0.05 (Table 1, Figure S2 ). The three genes with the largest negative and positive coefficients, respectively, are displayed in Figure 1C,D ; downregulation of ASGR2 ,GPBAR1 and HM13, and upregulation of USP44 ,ICOS and CDKN2AIP . Finally, we evaluated the enrichment of 680 pOIT+O-associated genes, relative to peripheral blood gene expression associated with acute peanut allergic reactions in a recently published clinical study by Watson et al using the same p-value cut-off (p<0.005)7. Out of our 680 significant genes, 108 genes overlapped with the differentially expressed genes in Watson et al7, mostly with opposite direction, Penrichment = 0.0095 (Figure 2 ).Our results demonstrate that omalizumab treatment alone does not induce alterations in whole blood gene expression in patients with severe food allergy. This is not surprising given that these patients were unexposed to peanut allergen at the time of blood sampling, and any concomitant asthma, rhinitis or eczema were well controlled. However, the longitudinal analysis during pOIT+O identified up- and downregulation of several immune-related genes. CD278/ICOS (Inducible T-cell costimulatory) is expressed on activated T-cells and appears to play a role in directing effector T-cell differentiation and responses during inflammatory conditions8. ICOS-expression on T regulatory cells and T follicular helper cells may be involved in the allergic disease mechanism9. In the pathway analyses, we observed significant enrichment for several GO biological process terms related to T-cell function and immune responses. Notably, we have previously described alterations in T-cell polyclonal in vitroactivation during pOIT +O in the FASTX study10. Comparing our findings with data described by Watson et al7 , suggests that pOIT+O may alter the expression level of many genes that were found activated during an acute peanut allergy reaction.The main limitations of this study are lack of any control subjects and small sample size. Moreover, further studies are needed to evaluate the long-term biological effect of pOIT+O.In conclusion, omalizumab treatment alone does not alter the transcriptional signature in peripheral blood of peanut allergic patients, but during pOIT+O, several immune-related signatures were observed. These results may provide insights into mechanisms of allergen tolerance.