Multiomics approaches disclose very-early molecular and cellular switches during insect-venom allergen-specific immunotherapy

doi:10.21203/rs.3.rs-3917969/v1

Download PDF

Article

Multiomics approaches disclose very-early molecular and cellular switches during insect-venom allergen-specific immunotherapy

https://doi.org/10.21203/rs.3.rs-3917969/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 26 Nov, 2024

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Allergen-specific immunotherapy (AIT) induces immune tolerance, showing the highest success rate (>95%) for insect venom while a much lower chance for pollen allergy. However, the molecular switches leading to successful durable tolerance restoration remain elusive. Here we applied a multilayer-omics approach to reveal dynamic peripheral immune landscapes during AIT-initiation phase in venom allergy patients (VAP) versus pollen-allergic and healthy controls. Already at baseline, VAP exhibited altered abundances of several cell types, including antigen-presenting cells (APC) and hybrid types, especially pDC-mDC hybrids. At 8-24h following AIT launch in VAP, we identified a uniform AIT-elicited pulse of late-transitional/IL-10-producing B cells, IL-6 signaling within Th2 cells and non-inflammatory serum-IL-6 levels. Sequential induction of activation and survival protein markers also immediately occurred. A disequilibrium between serum IL-6 and APC in VAP baseline was restored at day seven following AIT launch. Our longitudinal analysis discovers molecular switches during initiation-phase insect-venom AIT that secure long-term outcomes.

Health sciences/Diseases/Immunological disorders

Biological sciences/Immunology/Immunotherapy

Immune tolerance

Allergen-specific immunotherapy (AIT)

Type 2 (T2) disease

Longitudinal cohort

Insect-venom allergy

IL-6 signaling

Hybrid immune cell subsets

Systems immunology

User-friendly Interactive Resource

Repeated administration of relevant specific allergens, also known as allergen-specific immunotherapy (AIT), to patients with allergy is the only curative option available. The process of AIT mimics the frequent exposures to microbial factors in pre-school children with the traditional or rural farm lifestyle that secures the normal maturation process of the immune system, which, however, is most likely delayed by the modern westernized lifestyle in many patients with allergy (von Mutius and Smits 2020). Therefore, one could regard AIT as a re-training process of the disturbed immune system. Notably, although AIT has been applied to many types of allergic diseases, the rates of clinical cure with a complete control of the allergic reaction in anaphylactic patients (Muller 2010) in case of re-exposures were the highest (> 95%) after the accomplishment of insect-venom AIT (Calderon et al. 2007, Sturm et al. 2018). These facts make insect-venom AIT a prime candidate as a clinical model to investigate the immune factors underpinning the very-high long-term curative rates.

AIT should be performed for at least three years to ensure a maximum immune tolerance-inducing effect (Tabar et al. 2011, Shamji et al. 2022). Therefore, most of the studies have focused on the investigation of specific immune changes that occur later during or after the completion of long-lasting AIT. It has been well documented that AIT induces a decrease in soluble and cellular allergy effectors. This has been reflected by a reduction in IgE levels (Michils et al. 2000, Galli et al. 2008), activation and degranulation of mast cells and basophils (Maintz et al. 2009, Akdis and Akdis 2014) as well as frequencies of allergen-specific Th2 cells (Wambre et al. 2017, Rudulier et al. 2019) and type 2 innate lymphoid cells (ILCs) (Golebski et al. 2021). In the meantime, the allergic-response dampening factors or cell types, such as IgE-blocking antibodies (Abs), i.e., allergen-specific IgG4 (Aalberse et al. 2009) and the immune regulatory subsets with suppressive functions [i.e., regulatory CD4 + T cells (Treg) and B cells (Breg)] are enhanced by AIT (Francis et al. 2003, van de Veen et al. 2013, Akdis and Akdis 2014, Morita et al. 2015, Stanic et al. 2015, Scadding et al. 2017, Varona et al. 2019).

Nevertheless, since AIT takes at least three years, requiring a long-term compliance of patients and being extremely resource-demanding, it is invaluable to investigate whether and which robust early-window cellular or molecular predictive events could aid in the decision of initiating and continuing AIT (Eguiluz-Gracia et al. 2018) as early as possible. In order to identify the early molecular and cellular switches from pathological type-2 (T2) immunity (Pulendran and Artis 2012, Nakayama et al. 2017) to long-term immunological tolerance induced by insect-venom AIT, we compared dynamic immune responses between VAP and pollen allergy patients (PAP) already within the first hours and days following AIT launch, as AIT for the latter exhibits a much lower likelihood of durable success. Considering the effects of diurnal fluctuation (i.e., circadian rhythm) in immune cell circulation and function (Scheiermann et al. 2018), here we stringently controlled for the exact blood sampling hours at each visit of each participant. To this end, our unbiased integrative analysis of the early dynamic immune landscape provides a unique resource by deeply characterizing ~ 200 million of peripheral immune cells and ~ 200 samples of purified Th2 cells collected in the initiation phase (mainly build-up phase including baseline) of AIT in a real-world, prospective and longitudinal cohort. Such a patient-based, ex-vivo big-data resource will allow to generate novel testable cellular and molecular hypotheses for identifying early predictive biomarkers for long-term AIT outcomes and targets for enhancing AIT efficacy in T2 diseases, beyond insect venom anaphylaxis.

Revealing disturbed CD4 + and CD8 + T-cell subsets during the initiation phase of AIT

Here we administered AIT in two groups of patients allergic to insect venom or pollen. For 18 VAP, we implemented an ultra-rush (within eight hours) up-dosing protocol while we administered a conventional (within six weeks) up-dosing AIT in 16 PAP (Fig. 1A, Methods). Although the exact sampling periods differed between VAP and PAP, our sampling schemes enabled us to cover comparable phases during the build-up of AIT until the targeted cumulative maintenance dose was achieved for both types of AIT. We performed time-series sampling of peripheral blood during the first seven days, including baseline before AIT, in VAP. As for PAP, we extended our sampling period to two weeks (2W), six weeks (6W) and 12 weeks (12W). To benchmark with the natural immune fluctuations, we also recruited 10 healthy controls (HC), who did not receive any AIT or other immunotherapy, and sampled seven times within 12 weeks to match various time points of both patient groups. For the samples collected at each time point, we performed a multi-layer cellular and molecular omics (multi-omics) analysis, including single-cell mass cytometry (CyTOF) for deep immunophenotyping (Davis et al. 2017, Delhalle et al. 2018, Krieg et al. 2018, Mrdjen et al. 2018, Bonaguro et al. 2022, Capelle et al. 2023, Nuñez et al. 2023) of 77 immune subsets, sorting-enabled Th2-cell-type-specific whole-genome RNA-sequencing (RNA-seq), kinome analysis of ~ 350 kinases, multiplexing analysis of 10 serological cytokines and conventional 17-parameter whole-blood-count analysis (Fig. 1B). Our multi-omics analysis was accompanied by a detailed clinical workup in all the patients (Table S1). For details of various immunological datasets from each participant at each time point, please visit our interactive resource website that allows investigators to examine all the datasets and figures in an interactive and interlinked manner (Fig. 1C, Methods).

We first integrated and compared the datasets of single-cell CyTOF analysis (Fig. 2A) (for the gating strategy and marker information, refer to Figure S1, Table S2 and S3), serological cytokine multiplexing assays and whole-blood-count analysis between VAP and HC at various matched time points. Already at baseline, but also at the three following very-early time points [8h, 24h and day 7 (D7)], VAP demonstrated general immunological features that were different from HC in the principal component analysis (PCA) (Fig. 2B, Figure S2A). Meanwhile, the treatment-resulting peripheral immune responses among VAP were higher than the natural fluctuation of immune cells or soluble factors in HC over the same period (Figure S2B).

We next asked which specific immune subsets exhibited differential frequencies between VAP and HC at various matched time points. To identify the changed subsets in an unbiased approach, we performed volcano plot analysis (Fig. 2C) on all the 77 immune cell subsets defined by the combinations of various markers (Figure S1, Table S2 and S3 as well as the i3Dare website). At least 10 different immune subsets were significantly increased or decreased in frequency among total living singlets (refer to Methods) in VAP vs. HC, although not all of them showed a significant difference at each time point (Fig. 2C). Importantly, Th2 cells (CD45RO⁺CD45RA^¯CXCR3^¯CCR6^¯CCR4⁺, for the full gating strategy, please refer to Figure S1) showed a substantial increase in the frequency in VAP vs. HC at all time points. As allergy is known to be a predominantly IgE-driven Th2-mediated pathology (Robinson et al. 1992, Platts-Mills et al. 2001, Larché et al. 2006, Palm et al. 2012), this observation was expected. Simultaneously, the pathogenic allergen-specific Th2 cells, characterized as CRTH2⁺ Th2 cells (Hirai et al. 2001, Luster and Tager 2004, Wambre et al. 2017), were also substantially enhanced in VAP (Fig. 2C, 3A, 3B). A dominant response of total and pathogenic Th2 cells in VAP revealed by our supervised CyTOF analysis was further consolidated by an unsupervised clustering analysis (Kratochvíl et al. 2020) (Fig. 3C). In contrast to the scenario in VAP, total CD4 + Th2 cells were not significantly changed in PAP vs. HC at any of the tested time points during the first 12 weeks (Figure S2C). The general immune features of PAP started to show a trend to be different from that of HC only at 12W (Figure S2D). Indeed, PCA showed that at both matched time points (baseline and 24h), VAP and PAP exhibited distinguishable peripheral immune fingerprints (Figure S2E, S2F). These data reveal that upregulated total and pathogenic CD4 + Th2 cells characterize the immune signatures at baseline and during the AIT initiation phase in VAP, but surprisingly not in PAP, although recent environmental allergen/antigen exposure was absent in both groups (no recent stinging event in VAP; outside pollen season in PAP).

Notably, although CD8 + T cells have been rarely described in patient cohorts or animal studies in the allergy field, except for a few cases (Aguilar-Pimentel et al. 2010, Yu et al. 2019), the frequency of CCR4⁺CXCR3^¯ type-2 cytotoxic CD8 + T cells (CD8 + Tc2, Fig. 3D, 3E) (Inaoki et al. 2003) among living cells was also increased in VAP vs. HC at each time point already at baseline while the frequency of CCR4^¯CXCR3⁺ type-1 CD8 + T cells (CD8 + Tc1) was modestly but significantly decreased (Figure S3A). CD8 + Tc2-like cells, referred to as the direct parent gate of CD8 + Tc2 (Figure S1), were also among the two significantly increased immune subsets in PAP vs. HC at several time points (Figure S2C, S3B). Remarkably, the frequency of naïve CD8 + regulatory T cells (Tregs) (CD8⁺CD25^highCD127^lowCD45RA⁺) was significantly decreased in VAP vs. HC at both baseline and D7 (Fig. 2C, Figure S3C). The reduction in naïve and total CD8 + Tregs (CD8⁺CD25^highCD127^low) was also noticed among PAP vs. HC (Figure S2C, S3C, S3D). These findings demonstrated that the frequency of type-2 CD8 + T cells among peripheral immune cells dominates the immune profile in VAP while the fraction of naïve CD8 + Tregs declines in VAP relative to HC at baseline and AIT initiation phase. In short, our data underscore an unrecognized link between several CD8 + T-cell subsets and insect-venom allergy pathogenesis and possibly the long-term immune-tolerance induced by AIT.

Although CD4 + Tregs (CD4⁺CD25^highCD127^low) have been reported to play an important role in various allergic T2 diseases (Palomares 2013), we observed no significant difference in CD4 + Treg frequency between VAP and HC at any of our analyzed time points (Fig. 2C). Interestingly, the frequency of CD45RO⁺ CD4 + Tregs (memory Tregs) was significantly although modestly decreased in PAP vs. HC at three of five time points (Figure S2C, S4A). Remarkably, a negative correlation between pathogenic Th2 (i.e., CRTH2 + Th2 cells) and CD4 + Tregs was observed in HC, whereas a positive correlation between these two types of cells was detected in patients, especially in PAP (Figure S2G). This correlation shift indicates that CD4 + Tregs can successfully suppress pathogenic Th2 cells in HC, which became impossible in both patient groups, as it has been shown that Th2 responses escape Treg control due to different antigen specificities in respiratory allergic diseases (Bacher et al. 2016). In agreement with that scenario, in peanut-allergy patients, more evidence for a multi-functional Th2 response than a Treg deficiency was also documented (Chiang et al. 2018). The data in our cohort show that although the fraction of CD4 + Tregs does not differ in VAP vs HC before and during the AIT initiation phase, the equilibrium between CD4 + Tregs and Th2 is notably dysregulated in VAP.

Multiple aberrant hybrid subsets characterize the initiation phase of AIT

Among the decreased subsets in VAP vs. HC, the Th1-Th17 hybrid cells (CD45RO⁺CD45RA^¯CXCR3⁺CCR6⁺), co-expressing CXCR3/CD183 and CCR6/CD196 (Sundrud and Trivigno 2013), unexpectedly remained down-regulated in the volcano plot analysis at each time point (Fig. 2C, 3F, 3G). Notably, the frequency of Th1-Th17 hybrid cells was also declined in PAP vs. HC at all the five matched time points (Figure S2C). Th1-Th17 hybrid cells have been characterized by an enhanced and persistent anti-tumor function relative to classical Th1 or Th17 cells (Annunziato et al. 2012, Chatterjee et al. 2018). Th1-Th17 hybrids are also associated with late-forming and low-burden granulomas in tuberculosis control (Gideon et al. 2022). So far, no role for Th1-Th17 hybrid cells has been attributed in the control of Type 2 immune diseases. The substantial decrease of these hybrid cells was also independently confirmed in the unsupervised analysis (Fig. 3H). Even more surprisingly, the frequency of the CD8 + counterparts (CD8 + Tc1-Tc17 hybrids, CD45RO⁺CD45RA^¯CXCR3⁺CCR6⁺) was also decreased in VAP vs. HC at all the time points (Fig. 3I, J). This also held true in PAP at three of five time points (Figure S2C). Our results reveal a surprising dysregulation of both CD4 + and CD8 + type 1-type 17 hybrid cell types in both patient groups during the AIT initiation phase, already starting at baseline.

As expected (Sokol et al. 2009), the frequency of basophils was significantly enhanced in VAP vs. HC at several time points, from baseline on (Fig. 2C, Figure S4B). In PAP vs. HC at several time points, basophil-like cells, defined as the direct parent gate of basophils (Figure S1, S2C, S4C), were significantly increased together with CD8 + Tc2-like cells. Among increased immune subsets in VAP but not PAP relative to HC, classical monocytes (cMono, CD3^¯CD56^¯CD19^¯CD20^¯HLA-DR⁺CD14⁺CD16^¯) and total HLA-DR⁺ antigen-presenting cells (APC, for the gating strategy refer to Figure S1) were significantly enhanced in frequency (Fig. 2C, 3K, Figure S4D). Completely unexpected, a less-characterized immune subset, the so-called pDC-mDC hybrids, which co-express CD123 and CD11c, were also significantly higher in VAP at both baseline and 8h following AIT (Fig. 2C, 3L). This change was not statistically significant in PAP (Figure S2C). The frequency of this under-explored DC hybrid subset is elevated in patients with acute myeloid leukemia with the duplication of the oncogene FLT3 (Rickmann et al. 2011), although FLT3 and its ligands are known to play critical roles in DC development (Karsunky et al. 2003, Waskow et al. 2008). Our data show an unexpected abundance of cMono and one hybrid APC subset (i.e., pDC-mDC hybrids) during the AIT initiation phase starting at baseline in VAP.

Another decreased subset were ILC2 (Moro et al. 2010), which were diminished in VAP vs. HC at each time point (Fig. 2C, Figure S4E) as well as in PAP at several time points (Figure S2C). In line with our ILC2 data, the frequency of IL-10⁺KLRG1⁺ ILC2 is lower in patients with grass pollen allergy (Golebski et al. 2021). Our data indicate that ILC2 might play a critical role related to immune responses not only in PAP, but also in VAP. In addition to ILC2, another bridging subset between innate and adaptive immunity (Carding and Egan 2002, Papadopoulou et al. 2020), namely γδ + T cells, were decreased or showed a trend to be decreased (p = 0.0504 or p = 0.0553) in VAP vs. HC at all the time points (Fig. 2C, Figure S4F). Such a drop was not statistically significant in PAP (Figure S2C, 4F). In summary, our results firmly demonstrate that dysregulation of several bridging or plastic immune cell types, such as innate-adaptive bridging immune subsets (i.e., ILC2 and γδ + T cells) and hybrid plastic immune cell types (i.e., Th1-Th17 hybrids, Tc1-Tc17 hybrids and pDC-mDC hybrids), characterizes immune profiles during the AIT initiation phase already at baseline in VAP.

Immediate pulse of Breg accompanied with temporary contraction of allergen-specific Th2 in VAP following AIT launch

We next analyzed which specific immune subsets were changed in VAP immediately following AIT onset vs. baseline. No significant cellular change was observed at any time points except for 8h and 24h following AIT launch. The only subset significantly upregulated at 8h vs. baseline was late transitional B cells as defined by the combination of CD24 and CD38 (Palanichamy et al. 2009) (Fig. 4A, B). Before AIT, the frequency of late transitional B cells (CD24⁺⁺CD38⁺⁺) was lower in VAP vs. HC (Fig. 4B). Since the ‘regulatory’ protective role of transitional B cells in kidney allograft rejection has been reported (Shabir et al. 2015, Oleinika et al. 2019), the enhanced frequency of transitional B cells in our study might contribute to rapid tolerance induction following AIT. As human regulatory B cells (Bregs) are enriched in CD24^highCD38^high transitional B cells (Blair et al. 2010), we further analyzed IL-10-producing Bregs in our cohort. Notably, following in-vitro CpG stimulation (Fig. 4C), CD19⁺CD3^¯ B cells from VAP at 8h indeed showed a significantly higher fraction of IL-10⁺TNF-α^¯ cells than that at baseline (Fig. 4D, E), indicating an immediate boost of peripheral Bregs in VAP following AIT launch. The ratios between IL-10 + TNF-α¯ and IL-10¯TNF-α + cells, indicative of the suppressive capacity of B cells (Cherukuri et al. 2014), were also increased at 8h vs. baseline in VAP (Fig. 4D). As expected, no significant change in Bregs was observed in HC samples after 8h compared to baseline (Fig. 4D). At the same time, we noticed an expression pulse of the key activation marker CD25 among memory B cells, total CD45RO⁺ memory CD4 + T cells, CD4 + Th1 and Th2 cells immediately at 8h vs. baseline among VAP following AIT (Fig. 4F). A significant increase in the expression of the key functional marker HLA-DR was noticed in antibody-secreting cells, i.e., plasma cells (Fig. 4F). At 8h, a transient increase of CD127 among CD8 + Tc1 was observed (Fig. 4F). It has been well-established that CD127 (IL-7R) plays a non-redundant role in the survival of memory CD8 + T cells (Schluns and Lefrançois 2003). The temporary induction of CD25 expression at 8h was followed by the wave of the key costimulatory receptors, e.g., upregulation of CD27 among memory CD4 + T cells, CD4 + Th1, CD4 + Th2 cells and CD8 + Tregs within VAP at 24h (Fig. 4G). Importantly, the crucial antigen-presenting molecule HLA-DR was significantly but only transiently upregulated in non-classical monocytes (ncMono, CD3^¯CD56^¯CD19^¯CD20^¯HLA-DR⁺CD14^¯CD16⁺) in VAP at 24h following AIT launch (Fig. 4G). These results demonstrate that not only the temporarily-increased frequency of late transitional and IL-10-producing B cells, but also the waves of key activation/survival markers in several subsets of T-cells, B-cells and ncMono during the build-up time window might contribute to the long-term antigen tolerance achieved through AIT in VAP.

The finding of a temporary activation of several key extracellular markers of B cells and CD8 + T cells in VAP prompted us to ask whether the downstream signaling pathways are also affected during the early course of AIT. Encouragingly, following the unbiased kinome measurements of ~ 350 kinases in peripheral blood mononuclear cells (PBMC) depleted of CD4 + T cells, several PKC (protein kinase C) isoforms, especially PKC-α and PKC-β, which play a critical role in regulating T and B-cell activation and survival, were upregulated at 8h vs. baseline in VAP (Lim et al. 2015) (Fig. 4H, Figure S5). Several CDKs (cyclin-dependent kinases), vital for cell-division regulation and actin cytoskeleton organization, also showed increased activity, indicative of cellular activation, active migration and cell-cell interaction. Several essential Ca2+/calmodulin-dependent protein kinases, including AMPKa1, a well-known regulator of metabolic homeostasis and an anti-inflammatory gene (Jeon 2016), were also among top-ranked activated kinases, signifying the general activation of peripheral immune cells immediately following AIT launch. Although our kinome analysis was performed without prior sorting of each immune subset that would otherwise allow to pinpoint the exact cellular origin of the observed signaling cascades, it supports the overall notion that AIT immediately boosts activation of several types of peripheral immune cells in VAP during the early up-dosing phase.

The only decreased subset at 8h vs. baseline among VAP were pDC (plasmacytoid dendritic cells, HLA-DR⁺CD11c^¯CD123⁺, Fig. 4A). Notably, such a decrease was also observed among HC in the samples taken in the afternoon (4pm) as compared to those sampled eight hours earlier in the morning of the same day (8am, Fig. 4I, please also refer to Methods). No change in pDC was observed at other later time points vs. baseline among HC when we sampled strictly at the identical early morning hour for each HC as at baseline (Fig. 4J, Figure S2H). Although another study has reported a marked decrease of pDC at 52h (2pm on day 3) after AIT start (8am on day 1) in patients with venom allergy (Dreschler et al. 2011), our results strongly suggest that the natural circadian rhythm of the peripheral distribution (Scheiermann et al. 2018) and trafficking of immune cells is the likely cause for changes in pDC frequency of peripheral blood amongst both VAP and HC.

We also observed a temporary decrease in the frequency of CD27^¯CD45RA^¯ CD4 + effector memory T cells (CD4 + TEM) at 24h vs. baseline among VAP (Fig. 4K). As summarized elsewhere (Wambre 2015), CD27^¯ T cells are allergen-specific Th2 cells and confined to allergic patients. Our results showed that CD4 + TEM were transiently declined early on at 24h among VAP following the launch of ultra-rush AIT, which might also contribute to the well-known long-term beneficial outcomes in VAP. This data was in line with an increased CD27 MFI measurement in Th2 cells (Fig. 4G). Concordantly, mRNA levels of another antigen-specific marker, CD40LG, were significantly decreased in sorted Th2 cells at 8h vs. baseline in VAP (Fig. 4K, Figure S6). In contrast, we did not find any significant changes of any analyzed cell type in PAP at any of the later sampling time points vs. baseline (Fig. 4L, Figure S2I). The contraction in allergen-specific T cells might be an immediate consequence of the observed temporary enhancement of suppressive Bregs at 8h in VAP.

Non-inflammatory transient induction of IL-6 signaling in Th2 cells of VAP at 8h following AIT launch

So far, we have observed a temporary and early increase of several functional markers in T cells, B cells and APC. Following our unbiased CyTOF analysis, we indeed found Th2 as one of the most dysregulated immune subsets in VAP versus HC in the first week following AIT launch (Fig. 2C). Furthermore, as Th2 cells are well-known effectors of allergic responses that need to be modified through AIT, we decided to comprehensively analyze the underlying molecular changes in Th2 cells of allergic patients and performed a genome-scale RNA-seq analysis in sorted purified Th2 cells (Figure S6A-F and Table S4). Cell-type specific RNA-seq analysis, showing a certain degree of granularity between bulk and single-cell RNA-seq methods, can reveal highly-accurate pseudo-single transcriptomic profiles for a given purified cell type, especially for a rare subpopulation, e.g., Th2 cells. Similar to our CyTOF results (Fig. 3A), the frequency of sorted ex-vivo Th2 among living CD4 + T cells was indeed much higher in VAP vs. HC (Fig. 5A) and the frequency of sorted Th2 cells was highly correlated with the frequency/percentage of Th2 cells in the CyTOF analysis (Fig. 5B). Our RNA-seq analysis did not show any culprit-allergen bias (Figure S6D, E). All the marker genes specific for Th2 but not Th1, Th17 and Tregs were highly expressed among 10 randomly-selected RNA-seq samples (Figure S6F). Interestingly, our RNA data also exhibited a diurnal distribution pattern of several key circadian regulatory genes in sorted Th2 cells. The expression level of the key circadian rhythm genes, such as NR1D1 (also known as REV-ERBα) was uniformly increased in Th2 cells at 8h vs. baseline among almost all VAP following AIT (p-value = 4.0E-4, Figure S7A). Among HC, another circadian regulatory gene, PER3, changed at 8h vs. baseline (p-value = 0.0429, Figure S7B). Fluctuation in transcript levels of key circadian genes within Th2 cells and in frequency of circulating pDC (Fig. 4I) together demonstrate the importance of considering natural circadian rhythm during sampling for clinical investigations related to dynamic immune responses.

Unexpectedly, the number of differentially-expressed genes (DEGs) immediately peaked in VAP at 8h after AIT onset, followed by a decline to a very-low level thereafter (Figure S7C). We then assessed whether those DEGs (Fig. 5C) are significantly enriched in any molecular pathways or processes at 8h vs. baseline. Surprisingly, we found the IL-6 signaling pathway was ranked on top (Fig. 5D). We further asked which specific genes were affected in VAP at 8h following AIT. Through the unbiased volcano analysis, we found three genes, namely SOCS3, S1PR1 and BCL3, were substantially upregulated, thus being displayed as “outliers” (Fig. 5E) in the upper right corner of the volcano plot. Suppressor of cytokine signaling 3 (SOCS3) is a central regulatory component of the IL-6 signaling pathway (Croker et al. 2003). Patients with genetic defects in IL-6 signaling (Chen et al. 2021), in particular those with homozygous mutations in IL-6Rα mainly present with primary atopy such as elevated IgE, eczema and eosinophilia (Spencer et al. 2019). Our observation showed a very-early-stage temporarily, but uniformly increased SOCS3 transcript response among VAP but not HC (Fig. 5F), which therefore might contribute to the long-term high-curative outcomes of venom AIT. In line with the notion of enhanced IL-6 signaling after AIT launch in VAP, mRNA levels of JAK1 and STAT3 were also significantly upregulated at 8h vs. baseline in Th2 cells of VAP following AIT start (Figure S7D-E).

Our unbiased analysis identified S1PR1 (sphingosine-1-phosphate receptor 1) as another significantly-upregulated gene in sorted Th2 cells at 8h vs. baseline (Fig. 5E, G). S1PR1 plays a critical role in regulating lymphocyte trafficking between blood and lymphoid organs. Although inhibiting S1PR1 slows egress of lymphocytes from lymph nodes and reduces the number of circulating lymphocytes, no obvious increase of Th2 cells was observed at 8h vs. baseline in VAP (Fig. 2C, 3A). This might be due to the compensatory upregulation of the chemokine receptor CCR7 (Figure S7F), which supports the back migration of lymphocytes from blood to lymphoid organs (Aoki et al. 2016).

Our transcriptome analysis also pinpointed a substantial increase (~ fourfold, Fig. 5H) in mRNA levels of BCL3, a member of the non-canonical NFKB pathway, in Th2 cells among VAP at 8h vs. 0h. IL-6 induces BCL3 expression in murine CD4 + T cells in a STAT3 (Figure S7E)-dependent manner (Meguro et al. 2015). Concurrently, the kinase IKKα that critically activates the non-canonical NFKB pathway (Pelzer and Thome 2011) was also enhanced in PBMC depleted of CD4 + T cells at 8h vs. baseline in VAP immediately following AIT launch (Fig. 4H). In line with the notion of several temporarily-activated immune subsets as indicated by various layers of our multi-omics analysis, a T-cell survival- and homeostasis-regulatory gene, SLC29A1 (Equilibrative Nucleoside Transporter 3, ENT3) (Wei et al. 2018), was also located in the upper right segment of the volcano plot (Fig. 5E, Figure S7G). We further made use of our genome-scale ex-vivo time-series RNA-seq dataset to investigate whether any other pathways were enhanced during the AIT process by slightly reducing the fold-change thresholds while only confining to those genes showing a continuous dynamic pattern (Figure S7H). Following this additional analysis, the detected 29 significantly-increased genes were predominantly enriched in the pathways regulating mitotic cell cycle, inhibiting apoptosis, regulating MAPK signaling or JAK-STAT signaling (Figure S7H-M). Interestingly, the unusual kinases (PIM2 and PIM3), known to be induced by JAK-STAT signaling (Fox et al. 2005), were already upregulated in Th2 cells at 8h vs. baseline (Figure S7J). Concordantly, as the activity of these unusual kinases is mainly regulated by expression levels but not via post-translational modifications (Fox et al. 2005), our unbiased kinome analysis also revealed an increase in kinase activity levels of PIM2 and PIM3 (Fig. 4H), although only non-CD4 + immune cell samples were available and analyzed. The other genes regulating MAPK and JAK-STAT signaling, such as MYC, MAPKAPK3, DUSP7 and OSM [another IL-6 family cytokine (Hunter and Jones 2015)] showed upregulated transcripts in Th2 cells one day after AIT launch (Figure S7J, S7L). The genes tubulin tyrosine ligase like 12 (TTLL12, Figure S7K), regulating the mitotic phase, and immediate early response 3 (IER3, Figure S7M), reported to protect T cells from apoptosis (Mittal et al. 2006), were also upregulated in Th2 cells one day later. Another MAPK signaling molecule, fibroblast growth factor 18 (FGF18), was increased in Th2 cells one week post AIT launch (Figure S7M). Together, our unbiased Th2-specific transcriptomic analysis pinpoints several critical components in IL-6 signaling and the related pathways (e.g., JAK-STAT) within 8-24h immediately following AIT launch as key molecular switches towards long-term immunological-tolerance reinstallation in VAP.

Having discovered the enhanced IL-6 signaling pathway in Th2 cells on the transcript level, we next asked whether serological levels of the ligand (i.e., IL-6) are also changed. Highly interestingly, among all the 10 analyzed Th1 (TNF-α, IFN-γ, IL-12), Th2 (IL-4, IL-13) and other inflammation-related cytokines (Fig. 5I), IL-6 was the only one showing an uniform pattern of temporary upregulation immediately following AIT treatment among VAP (Fig. 5I, J). Although the average increase was almost five- or four-fold at 8h or 24h relative to baseline, respectively, the absolute IL-6 concentration still remained in the near-homeostatic non-inflammatory range (Tsigos et al. 1997) (Fig. 5J). Notably, the ROC (receiver operating characteristic) analysis showed that the non-inflammatory absolute IL-6 levels alone, already at baseline (Figure S8A), especially at 8h post AIT launch (Figure S8B), presented an outstanding diagnostic power [the area under curve (AUC) up to 0.9941] to distinguish VAP and HC (Figure S8A-D). Since the main producers of IL-6 are APC and endothelial cells (Mauer et al. 2015), APC-derived IL-6 plays a critical role in restricting Th2 differentiation (Mayer et al. 2014) and several types of APC were more abundant among VAP vs. HC (Fig. 2C), we investigated a potential correlation between serological IL-6 levels and specific circulating APC subsets or total APC. IL-6 levels were positively correlated with the frequency of total APC only in HC but not in PAP and VAP at baseline (Fig. 5K). In line with this notion, circulating IL-6 levels were highly correlated with the frequency of intermediate monocytes (iMono, CD3^¯CD56^¯CD19^¯CD20^¯HLA-DR⁺CD14^dimCD16^dim), classical monocytes (cMono) and total APC among HC, but not in VAP or PAP (Figure S8E-G), when samples from different time points were grouped together. Interestingly, the increased frequency of total APC at D7 in VAP relative to baseline was significantly correlated with the augmented IL-6 levels at 24h post AIT (Fig. 5L). These results indicate that the existing equilibrium between IL-6 and total APC or APC subsets in HC was disturbed in both groups of allergic patients. Importantly, our data show that AIT can restore this pathological shift early on in the first days after AIT launch in VAP.

In contrast to the fast response in VAP, PAP displayed a substantial increase in the number of DEGs in Th2 cells only after the completion of the up-dosing phase, i.e., at 12W following AIT launch (Figure S7C), which is also consistent with the slow dynamic evolution as suggested by our CyTOF analysis in PAP (Figure S2D). The top-ranked genes with increased expression (p < = 5e-4, fold change > = 1.4) were significantly enriched in the pathways regulating mitochondrial functions and tRNA processing (Figure S7N), indicative of enhanced metabolic and translational activities, all preparing for cell activation and proliferation. The differential response dynamics on the molecular levels represented one of the key differences between VAP and PAP during the very-early phase of AIT, which may play an important role for regulating long-term superior AIT outcomes in VAP.

Insect-venom AIT results in the highest (~ 95%) clinical cure rates among all AITs used for the treatment of T2 diseases. Identifying the early-stage molecular or cellular switches of insect-venom AIT could play a two-fold role: i), supporting the early decision-making whether to continue the multi-year AIT procedure or not; ii), guiding a better design of other AITs in consideration of intervening with those unrecognized early molecular and cellular switches. Here we report the first comprehensive analysis of the immunological response landscape during the first week of ultra-rush AIT in VAP. On the cellular level, we observed an immediate and temporary upregulation in the frequency of late transitional and regulatory B cells at 8h vs. baseline following AIT launch in VAP that was accompanied by a transient decrease in the frequency of allergen-specific CD4 + T cells at 24h. On the functional level, we observed a transient upregulation of several essential activation or survival markers in the first 8 to 24 hours, such as CD25, CD27 and CD127 among several subsets of CD4 + T cells, CD8 + T cells and memory B cells. Notably, we also noticed a transient increase in the expression levels of the crucial antigen-presenting functional marker HLA-DR among APC subsets at 24h following AIT launch in VAP. In line with the immediate activation status in T and B cells, a kinome analysis provided further unbiased support for a general pattern of activation in various non-CD4 + immune subsets among PBMC at 8h in VAP. Meanwhile, we observed an immediate AIT-induced surge of transcripts of several key components of the IL-6 signaling pathway in sorted Th2 cells at 8h in VAP only. The transient upsurge of cellular IL-6 signaling was accompanied by the temporary enhancement of MAPK and JAK-STAT signaling in Th2 cells as well as serological IL-6 levels. Notably, we observed a restoration of the balance between the levels of circulating IL-6 and the frequency of APC in the initiation phase immediately following AIT launch, which was disrupted before AIT start in VAP. In contrast to VAP, we could only observe a clear transcriptional response in Th2 cells of PAP at a much later stage (i.e., only at 12 weeks following AIT launch). With our current datasets, we cannot distinguish whether the observed differential dynamic responses between VAP and PAP were attributable to different AIT allergens or due to different up-dosing schemes employed in the initiation phase, which warrants further investigation. However, the combinations of early immune responses on various layers, e.g., by showing an enhanced regulatory B-cell profile, non-inflammatory IL-6 signaling in Th2 cells, activation/survival signaling in various immune subsets and a rebalance between serum IL-6 levels and APC, although being transient, likely impact on the well-known long-term beneficial outcome of AIT in VAP. Moreover, from a methodological point of view, our study that was based on real-world patient recruitment also provided unexpected evidence for the need to consider the circadian rhythm during the design of immune monitoring studies, as DC have been recently shown to functionally control circadian anti-tumor responses (Wang et al. 2022).

Our discovery that IL-6 signaling (e.g., SOCS3) is the major upregulated pathway in ex vivo-purified Th2 lymphocytes during the earliest time window of insect-venom AIT (8-24h) is further supported by a uniformly detectable peak of increased serum IL-6 cytokine levels in VAP only. However, the resultant serum IL-6 levels following venom AIT launch still remained strictly within a near-homeostasis, non-inflammatory range of IL-6 in human blood (peak range in insect venom AIT at 8h: 1.5–8.5 pg/mL; normal median healthy human plasma level: 0.75 pg/mL). Pharmacokinetic studies performed in human healthy volunteers defined a maximum physiological plasma concentration of 50 pg/mL for IL-6 (Tsigos et al. 1997). Although an increased IL-6 level might indicate an ongoing inflammatory response or acute infectious disease (Hunter and Jones 2015), IL-6 also plays an important role in regulating cellular homeostatic functions and in repressing Th2 immune responses (Mayer et al. 2014, Chen et al. 2021). In line with our data, in-vitro cellular IL-6 binding studies have recently shown that only few IL-6 molecules are sufficient to activate IL-6 signaling (Hansen 2020). An acute, but controlled increase in circulating IL-6 levels, as observed in our work, might transiently trigger IL-6 signaling in relevant immune subsets, such as Th2 cells, and initiate a protective tolerance response very early during AIT in VAP. In short, similar to transition states in chemistry (Laidler and King 1983), here we revealed an AIT-induced transition-state-like immune landscape (IL-6-related data and several other related observations) that is uniformly present in all VAP and might thus be a critical intermediate switch to subsequently trigger a sustainable healthy immune state. In supporting our notion that the observed very-early transition-state-like responses might have clinical implications, it has been long shown that > 88% of AIT-initiated bee venom patients already tolerate re-stings within one week after reaching the maintenance dose (Goldberg and Confino-Cohen 2010).

Meanwhile, our deep immunophenotyping analysis identified a general fingerprint distinguishing VAP from HC already at baseline. This fingerprint was largely maintained in the first week during AIT. Surprisingly, our observations indicated that reduced CD8 + but not CD4 + Tregs, especially naïve CD8 + Tregs in both allergic patient groups, might also critically contribute to the pathogenesis of allergic diseases. At the same time, an unexpected strong synergy between CD4 + and CD8 + T cell responses was observed, i.e., both CD4 + Th2 and type-2 CD8 + T (CD8 + Tc2) cells were enhanced only in VAP relative to controls. Notably, the frequency of both CD4 + Th1-Th17 and CD8 + Tc1-Tc17 hybrid cells was diminished in both patient groups. Thus, we discovered previously unrecognized abnormal proportions of several hybrid cell types including CD4 + Th1-Th17 and CD8 + Tc1-Tc17 hybrids in both groups of allergic patients while the APC hybrid subsets, i.e., pDC-mDC hybrid cells were only upregulated in VAP. Our data firmly indicate an unexpected crucial role of several poly-functional hybrid immune subsets with higher plasticity in insect-venom allergic diseases. Our findings complemented another type of plastic Th17 cells that can acquire Th2-cytokine production capacity, although reported in a murine asthma model (Tortola et al. 2020, Lozano-Ojalvo et al. 2021). Concordantly, our observations in ILC2 and γδ + T cells supported a particular involvement of immune cell types that play a bridging function between innate and adaptive immunity.

Here we analyzed ~ 220 blood samples over time, which represents the first systemic effort to characterize the initiation-phase immune dynamic molecular and cellular atlas in allergic patients following AIT launch. Our multi-omics analysis identified uniform, nevertheless yet unrecognized, cellular and molecular switches during the up-dosing phase (i.e., first week) of AIT in VAP. As this clinical AIT generates exceptionally successful long-term curative outcomes, the identified underlying early mechanisms could also be generated to the improvements of AIT for other types of allergic diseases or T2 diseases still with lower curative rates. Moreover, optimizing the sampling time (that excludes interference by circadian rhythm) and other pre-analytical procedures in our pipeline ensures an exceptional quality of the overall datasets, which can be used as a best practice example for immunological clinical studies to reduce data heterogeneity (Bedard et al. 2013, Cho and Feldman 2015), even in a small-scale study design as demonstrated here. The generated rich molecular and cellular resource will provide a vigorous toolbox for gaining a systemic and dynamic comprehension of very-early immune responses during the AIT initiation phase, towards long-term curative antigen-specific immune tolerance. Mapping to our interactive reference datasets, investigators can monitor the early immune responses of individual patients following the AIT launch. Although T2 diseases also involve many types of non-immune cells (Gieseck et al. 2018), our resource will facilitate the deciphering process of type-2 homeostatic immune regulation and the complex disequilibrium of immunity in various T2-associated diseases.

Cohort design.

The study was approved by the Luxembourg National Research Ethics Committee (CNER) (SYS-T-Act: No. 201509/11) and registered at ClinicalTrials.gov (NCT02931955). Informed consent was obtained from each participant prior to sampling and personal data collecting, which were in full compliance with the Luxembourg National Commission for Data Protection (CNPD, before the implementation of GDPR in Europe). We obtained notification on 11th Sep, 2015 and modification approval on 17th Jul, 2017 from CNPD.

The study participants with moderate-to-severe allergic rhinitis (pollen allergy patients, PAP, n = 16) and insect sting reactions (venom allergy patients, VAP, n = 18) eligible for AIT in accordance with the current international guidelines were recruited at a tertiary center (Service National d’Immuno-Allergology of Centre Hospitalier de Luxembourg) between August 2016 and February 2018. The healthy controls (HC, n = 10) were recruited at the Clinical and Epidemiological Investigation Center of the Luxembourg Institute of Health (CIEC-LIH) between Jan and Apr 2018. The study was designed as a real-world, prospective, exploratory, data-driven trial for identification of novel potential biomarkers and molecular switches in the early time window of AIT in the outpatient clinical setting. The inclusion criteria were the confirmed clinical diagnosis of allergy (skin prick test positivity and elevated sIgE titers to plant pollen or insect venom allergens, respectively), clinical indication for launch of AIT for the patients and absence of known allergies for healthy volunteers. The exclusion criteria were age less than 18 years, pregnancy and lactation; overt asthma or chronic obstructive pulmonary disease, concomitant autoimmune disorders, chronic diseases in exacerbation or not adequately controlled, history of hematological malignancies or solid tumors, treatment with systemic steroids, immunomodulatory agents or biologics, acute and exacerbation of chronic infections, traumas and surgeries in six months prior to the study enrollment. No restrictions were introduced regarding the number of sensitizing allergens, duration of the allergic disease, the time elapsed since the last field sting and treatment with mono- versus polyvalent allergen immunotherapy products. In VAP, an ultra-rush AIT protocol was administered to reach the maintenance dose within 8 hours with 14 injections (Fig. 1A). For PAP, a conventional up-dosing AIT was applied to reach the maintenance dose within 6 weeks with a weekly injection scheme. Therefore, although the exact sampling periods are different between the two groups of patients, our sampling schemes allowed us to cover comparable periods (essentially build-up phase) until the beginning of the maintenance stage for both types of AIT.

While the age of the PAP was comparable to that of HC (median, ∼37 and ∼35 years, respectively, Table S1), VAP were older (median, ∼49 years). While both patient groups included male and female participants at almost 1:1 ratio, the HC group was predominantly female (9 of 10). Five of 16 PAP were sensitized to major birch pollen allergen (Bet v1), two of 16 exhibited elevated sIgE and clinical reactivity to timothy grass pollen allergens (Phl p1 & Phl p5) and eight were allergic to the pollen of both species. One PAP was timothy grass pollen-allergic, but the sIgE results were not available. Most VAP (14 of 18) presented with a history of systemic reactions to wasp (yellow jacket; Vespula spp.) stings, being sensitized to Ves v1 (one of 18), Ves v5 (eight of 18) or both allergens (five of 18). Three of four bee VAP showed elevated sIgE levels to both Api m1 and Api m5, another VAP being mono-sensitized to Api m1. Two VAP are hobby beekeepers.

On the first day of the ultra-rush protocol, most of the VAP received their first injection at 8:00 AM with a gradually increasing dose in µg (i.e., starting from 0.003 to 0.006, 0.015, 0.030, 0.060, 0.150, 0.30, 0.60, 1.50, 3.00, 6.00, 15.00, 30.00, ending with 55.00) at 15-min intervals. While the baseline (0h) blood sample was drawn at 8:00 AM, the second blood sample (8h) was taken at 16:00. There were a few cases, where the first injection was given at 8:15 AM and then the afternoon sampling was delayed to 16:15. Biological samples were collected by trained study nurses, transported at room temperature (RT) within maximum two hours to the bio-specimen laboratory of the Integrated Biobank of Luxembourg (IBBL), where they were either used freshly in the experiments on the same day, or cryopreserved in the IBBL biorepository for further analysis later on.

Isolation and cryopreservation of PBMCs.

PBMCs (peripheral blood mononuclear cells) were obtained from fresh whole blood by density gradient immediately followed by CD4 T-cell isolation and Th2 sorting or cryopreservation. For PBMC isolation and deep immunophenotyping analysis, we followed a similar procedure as we described in another clinical study (Capelle et al. 2022). Briefly, we collected up to 50 ml of blood per patient at each time point. 40ml were collected in BD ACD tubes for PBMC analyses and 10ml in BD EDTA tubes for whole-blood-count and cytokine measurements. We first added 13 ml of Ficoll Paque Plus (GE17-1440-02, Merck) at the bottom of the Falcon tubes. Blood from ACD tubes was mixed and split into two 50-ml Falcon tubes that were diluted with up to a total volume of 35ml with DPBS (14190144, Thermo Fisher Scientific). Diluted blood was added carefully on the top of Ficoll-containing Falcon tubes. The Falcon tubes were then centrifuged at 400g at room temperature (RT) for 30 min. The PBMC layer was collected, diluted with DPBS plus 2% heat-inactivated FBS (fetal bovine serum, 10500-064, Thermo Fisher Scientific) and centrifuged twice at 300g for 10 min at RT. Purified PBMCs were either cryopreserved using 90% heat-inactivated FBS plus 10%DMSO (D2650, Sigma Aldrich) freezing medium and stored in liquid nitrogen or used freshly to isolate CD4 + T cells as described below. The plasma layer was also collected during the PBMC isolation process. Whole-blood-count data were measured by ABX Micros CRP 200 (Horiba, Axonlab) and plasma were isolated from fresh blood collected in 10 ml BD EDTA tubes.

Single-cell mass cytometry (CyTOF) measurement, analysis and visualization.

Peripheral blood mononuclear cells (PBMCs) were isolated and stored in liquid nitrogen until the day of the staining when the cryovials were thawed using warm DMEM 4.5 g/L Glucose with L-Glutamine (BE12-604F, Lonza) containing 5% of FBS (10500-064, Thermo Fischer Scientific) and 2 mM EDTA (15575-038, Invitrogen). Cells were then washed with PBS (17-516F, Lonza). Prior to the metal-conjugated antibody staining and for assessing their viability, cells were re-suspended at a concentration of 1 x 10⁷ /1 mL and were incubated for 5 min at RT with Cell-ID Cisplatin (201064, Fluidigm) at a final concentration of 1 µM. The incubation was stopped by adding the staining buffer (PBS, 5%FBS, 2 mM EDTA; of note, EDTA was not used in other scenarios unless specified). Surface staining was performed by adding a cocktail of pre- or in-house-conjugated antibodies (Table S2) for 30 min at RT; excess antibodies were removed by washing (400g, RT, 10 min). It is important to mention that each tube was then split into half for introducing intense washing steps since the samples had initially shown strong signals of iodine (most probably coming from the Ficoll). As a last step, samples were incubated with Ir-Intercalator (201192B, Fluidigm), diluted in MaxPar Fix&Perm (201067, Fluidigm) at a final concentration of 50 nM, and rested at 4^oC until the day of the acquisition. Prior to the acquisition, cells were washed twice with PBS and with de-ionized water two times. Centrifugation conditions after fixation, on the acquisition day, were 800g, for 10 min at 4^oC. Cells were re-suspended at 5 x 10⁵ per ml in 1:10 calibration beads (EQ Four Element Calibration Beads, 201078, Fluidigm), diluted with de-ionized water and the samples were analyzed with the Helios™ mass cytometer (Fluidigm) at a flow rate of 0.030 mL per min. Generated .fcs files were normalized with the HELIOS acquisition software (v 7.0.5189) by using EQ beads as a standard. Unless otherwise specified, the indicated frequency of the given subset was always relative to the total living singlets (for the full gating strategy, please refer to Figure S1). Statistical analysis of supervised CyTOF analysis was done in Qluocore Omics Explore v 3.8 (1). The non-paired two-tailed Mann-Whitney test was applied to compare the immune subsets between different groups at the given matched time point while paired two-tailed Mann-Whitney test was used to compare the AIT response or natural immune fluctuations between a later time point and baseline from the same patients or controls. The volcano plot was visualized in Graphpad Prism v9.0. Of note, due to technical staining issues, the number of samples used for TCRγδ analysis (n = 149) was lower than that for the other immune subsets (n = 199) (for details regarding TCRγδ staining on which sample, participant and time point, please refer to our i3Dare website).

Although our CyTOF data was essentially analyzed by the supervised manual gating as described above, we also independently validated our discoveries of several selected subsets of interest (only Fig. 3C, 3H) by our unsupervised clustering and visualization methods. The unsupervised clustering analysis was performed using the pre-gated living singlets. Pre-gating of identifying live singlets was done in FlowJo v10. Lineage markers used for clustering have been arcsin transformed using a co-factor of 5. Unsupervised clustering has been performed using GigaSOM (https://github.com/LCSB-BioCore/GigaSOM.jl). GigaSOM is a FlowSOM-style clustering and dimension reduction method, but adapted for our huge-scale data sets with ~ 200 million of cells using the high performance computing system in the Julia programming language (v 1.4). Of note, the analysis and visualization was performed without down-sampling. High resolution clustering settings were 32x32 SOM-grid, 40 rounds (epochs) and exponential radius decay. The 1024 clusters were annotated to different immune cell subsets according to the marker combinations similar to that established in the manual gating strategy (Figure S1). The plot (Fig. 2A, 3C, 3H) was visualized using EmbedSOM v2 (https://github.com/exaexa/EmbedSOM). The representation of a two-dimensional embedding of each single cell has been performed using the default settings of EmbedSOM.

Measurement of circulating cytokine levels by the MSD assay.

Ten analysts were either cytokines or chemokines (IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13 and TNF-α) measured in plasma samples (isolated from EDTA tubes) from patients and healthy controls at all time points using a multiplexing assay [The V-Plex Proinflammatory Panel 1 (hu) assay from MSD, kit catalog K15049D-1]. The samples were undiluted and measured in duplicates. The assay was performed according to the manufacturer’s instructions. Data were recorded and analyzed on a MESO QuickPlex SQ 120 instrument (software version LSR_4_0_12).

Analysis of intracellular cytokines in B cells following CpG stimulation.

Detection of IL-10 production in human B cells followed the procedures described elsewhere (Menon et al. 2021). Nine samples were selected from those collected at baseline and 8h per participant group (either VAP or HC). The reason why only nine samples per group were selected was that we had to analyze both Bregs and kinome simultaneously to save precious clinical samples, where the number of sample spots was a limiting factor in the available chips for the kinome analysis. We selected VAP samples based on the top ranking of the plasma IL-6 level fold-changes between 8h vs baseline while randomly selecting nine out of ten HC samples per time point. Following the thawing of bio-banked cryopreserved PBMC depleted of CD4 T cells [refer to the section ‘Isolation and cryopreservation of PBMCs’], we re-suspended those cells in RPMI complete media (RPMI-1640 supplemented with 100 U/ml Penicillin, 100 µg/ml Streptomycin, 2 mM L-Glutamine and 10% heat-inactivated FBS). 8.5 x 10⁵ cells were stimulated by 1 µM of CpG (ODN2395, Invivogen) for 72h in 200 µl of RPMI complete media with a U-bottom 96-well plate. In the last 4.5h before staining, 50 ng/ml PMA, 250 ng/ml Ionomycin plus 1 µl of BD Golgiplug were added together. The cells were then washed (200g, 10 min, 4^oC) and blocked with Fc blocker (BD) for 5 min at 4^oC. The cells were again washed and stained with the surface marker antibody mastermix (CD3 BUV737, 741822, BD; CD19 PercP-Cy5.5, 45-0199-42, Invitrogen; LIVE/DEAD Fixable Near-IR Dead Cell Stain, L10119, Thermo Fisher Scientific) in BD Brilliant stain buffer at 4^oC for 30 min. The cells were subsequently stained with selected cytokine antibodies (IL-10-APC, 506807, Biolegend; TNF-a-BV650, 502938, Biolegend) using the Fixation/Permeablization kit (554714, BD) following the manufacture’s recommendations. The cells were finally re-suspended in the staining buffer (2% FBS in Ca²⁺/Mg²⁺ free PBS) and acquired in a BD LSRFortessa™ (v8.0.1). The Breg stimulation results were analyzed in Kaluza software (v2.1, Beckman).

Ex-vivo human kinome analysis using the PamGene array.

The same nine samples of the patient groups (refer to the rationale of selecting these nine samples above) processed for Breg analysis were also utilized to asses kinase activity using PamGene’s (PamGene International B.V., HH's-Hertogenbosch, the Netherlands) phosphotyrosine kinase (PTK) and serine-threonine kinase (STK) assay. Following the thawing of bio-banked PBMCs depleted of CD4 T cells [refer to the section ‘Isolation and cryopreservation of PBMCs’], cells were directly (without any in-vitro treatment) pelleted (800g, 5 min, 4^oC) before resuspension in 1ml ice-cold PBS and pelleted (1000g, 10 min, 4^oC). Pellets were lysed with 45 µl of M-PER (78501, Thermo Fisher Scientific) supplemented with Halt Phosphatase Inhibitor Cocktail (78428, Thermo Fisher Scientific) and Halt Protease Inhibitor Cocktail, EDTA free (78437, Thermo Fisher Scientific) for 15 min on ice. The lysate was clarified (10,000g, 15 min, 4^oC) before preparing 10 µl aliquots, each snap frozen in liquid nitrogen before storage at -80^oC. One aliquot was used for protein quantification by Bradford.

Adopting a balanced design, a separate aliquot of cell lysate was assessed on the PTK chip (86402, PamGene) and the STK chip (87102, PamGene) using the manufacturer's protocol. From each chip, a series of images describing the phosphorylation of peptides (196 for PTK and 144 for STK) corresponding to ~ 350 native kinases in the applied lysate was obtained. Using the instrument manufacturer’s BioNavigator software (v6.3.67.0), data was curated and normalized by chip using COMBAT. Upstream kinase analysis, a multiple hypothesis testing approach, identifies responsible kinases for the changes in peptide phosphorylation observed. R v4.1.1 was used to compile the results and to display them using an adaptation of the CORAL tool (https://github.com/dphansti/CORAL).

Principal component analysis (PCA) and integrative analysis.

Based on the ‘corr’ function in MATLAB 2020a, the correlation coefficient matrixes within or between different datasets were calculated (e.g., within different immune subsets, between immune subsets and cytokines). Since the datasets often do not follow a normal distribution, the ‘Spearman’ correlation coefficient was employed for the correlation analysis described in this section. The continuous range of correlation coefficients is (-1, + 1). The P-values following the two-tailed test were also directly generated by the 'corr' function. When the P-value is less than 0.05, we considered that the corresponding correlation coefficient was significantly different from zero.

The principal component analysis (PCA) was performed on the default setting of the ‘pca’ function in MATLAB 2020a. Considering different types of datasets, different groups of participants and different sampling time points, we divided the analysis into two categories. One type was to analyze the sample distribution between different participant groups using different combinations of datasets at a fixed sampling time point; the second type was to analyze time evolution responses within a fixed participant group. Before the PCA analysis, all the data were centered by ‘zscore’ function so that each variable had a mean value of 0 and a standard deviation of 1. In order to compare the changes of the overall immune features of each patient following immunotherapy (or each healthy control following the sampling period), we calculated the average distance among pairs of the adjacent two time points of each participant in the two-dimensional PCA plots. The larger the distance is, the more the overall peripheral immune system of the given subject has changed over the sampling period either caused by immunotherapy and/or natural fluctuation. Of note, each plot in the time-slice PCA figure (Fig. 2B and Figure S2D) is independent and different PCA plots were draw together only for better visualization.

Th2-cell-type-specific RNA-seq analysis.

Sorting Th2 cells and cell-type-specific mRNA sequencing

Following the PBMC isolation from fresh blood, CD4 + T cells were first enriched with human CD4 microbeads (130-045-101, Miltenyl Biotec) following the manufacture’s recommendations while the remaining fractions of PBMC were immediately cryopreserved as described above. The CD4 + T cells were then immediately shipped at 4°C to our sorting flow-cytometry facility. The CD4 + T cells were first stained in Brilliant Stain Buffer (563794, BD) using all the Abs with specified dilution factors and configuration settings as defined in Table S4. Samples were incubated for 30 min at 4°C in the dark (and briefly vortexed once after 15 min). After two washes at 350g, 10 min, 4°C, the cells were re-suspended at a concentration of 10 million per ml and aseptically sorted (20 PSI, 100 µm nozzle, purity mode) in a BD Aria III sorter (for the FACS gating strategy, please refer to Figure S6A). After sorting, cells were centrifuged at 350g, 10 min, 4°C, pelleted and lysed in 700 µl of the Qiazol buffer (Qiagen). The samples were transferred into 1.5 ml Eppendorf tubes and immediately frozen at -80°C until RNA extraction. RNA was extracted using RNeasy Micro kit (Qiagen) following the manufacture’s recommendations. RNA-seq was performed at the EMBL Genomics Core Facility (Heidelberg, Germany) using single-read 75bp high output by NextSeq 500 with ~ 50 Million reads/sample. The NEBNext Ultra II RNA Library Prep Kit was used and the rRNA deletion step was included.

Th2-cell-specific RNA-seq pre-processing and quality control

Quality control of the raw sequencing data was conducted using the FastQC software (v0.11.7-Java-1.8.0_162) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Reads were mapped to the reference human genome 38 (http://hgdownload.soe.ucsc.edu/goldenPath/hg38/bigZips/) and summarized to genes using the R Bioconductor Rsubread package (v1.32.4). One sample was excluded for further analysis because CD127 Ab was forgotten to add before sorting.

Analysis of potential confounding factors

Clinical and demographic factors that are significantly different between groups can have a confounding effect on subsequent analyses if no correction for these factors is applied. The following statistical tests were conducted to identify potential confounding factors. For categorical variables, a Fisher’s exact test was performed, whereas for normally distributed numerical variables having equal variances, one-way ANOVA was applied. In case of a normally distributed variable with unequal variances between the groups, Welch one-way test was used. For non-normally distributed values, the Kruskal-Walls rank sum test was employed. Homogeneity of variances was determined using Levene’s test, while normality was tested using quantile-quantile (Q-Q) plots.

Filtering, normalization, voom transformation and quality control

Genes that consistently had zero or very low counts were first removed using the Bioconductor edgeR package (v3.24.3). Trimmed Mean of M-values (TMM) normalization was performed on the filtered data with the edgeR package. The normalized data were voom-transformed to enable differential expression analysis with the Bioconductor limma package (v3.38.3). Principal component analysis (PCA) was performed on the voom-transformed data, and a score plot was presented where each batch was represented by a different colored symbol. In case of no batch effect, the batches do not cluster in the PCA plot. Furthermore, a principal coordinate analysis (PCoA) was conducted with the R Stats package and the score plot was inspected to check for potential outliers.

Differential expression and enrichment analysis

Differentially-expressed genes (DEGs) between allergic patients and controls, and between AIT-treated samples and baseline were determined by applying an empirical Bayes moderated t-test on the voom-transformed data, implemented in the R Bioconductor limma package (v3.38.3). Nominal p-values were corrected for multiple hypothesis testing by computing false discovery rates (FDR) according to the Benjamini-Hochberg procedure and using an FDR significance threshold of 0.05. Entrez gene IDs were converted to gene symbols with the R Bioconductor org.Hs.eg.db package (v3.7.0). DEGs were visualized in heatmaps and volcano plots using the R Bioconductor ComplexHeatmap (v1.20.0) and R ggplot2 (v3.2.0) packages, respectively.

Enriched differentially-regulated pathways and GO (Gene Ontology) biological processes were determined by over-representation analysis in ConsensusPathDB (release 34, http://cpdb.molgen.mpg.de/ ) using all measured genes as background list. Pathways were selected from Reactome, KEGG, Wikipathways, Biocarta and PID. Pathways and GO biological processes with Benjamini-Hochberg corrected FDR ≤ 0.05 were considered significant.

Interactive and interlinked resource (‘i3Dare’) sharing

To enhance the re-usability of our valuable clinical immunology resource, we employed the business intelligence tool Tableau to integrate and share various datasets in a website (Hedin et al. 2021), allowing investigators to scrutinize all our results in an user-friendly, interactive and interlinked manner. In the manuscript, we could not show the dynamic response patterns of each measured immune subset following AIT. However, in our resource website, for each measured immune subset, independent of significance levels and change folds, following the click, one can visualize and explore its dynamic response pattern following the launch of AIT in different patient groups. Importantly, through our interactive website, each subset in each volcano plot can be related to its dynamic response pattern over time. It is also possible to compare the dynamic responses between different immune subsets, which is not realizable for any static figures. Moreover, one can also explore immune responses of each individual patient at different time points following the launch of AIT, or benchmark the measurements of clinical samples with our reference datasets generated from HC or VAP or PAP.

Data Availability:

We deposited all the raw and processed Th2-specific RNA-seq dataset in the GEO database with the accession number GSE240503 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE240503; to review, please input the bold token sfenmuqsljilxul). Our large-scale interactive interlinked immunological Data resource (‘i3Dare’) allows investigators to more effectively explore or reuse our immunological datasets (https://review-tableau.lih.lu/; to review, please use the username: LIH-SYSTACT-Reviewer; password: G6iqQg04BXNQ2IkmXz2o; be sure to activate the full-screen mode in the top right corner for better navigation).

Acknowledgements:

We first would like to acknowledge all the anonymous participants in the SYSTACT cohort. We appreciate all the support of research nurses especially the coordinator Jean-Yves Ferrand at CIEC of LIH and clinical nurses at Central Hospital of Luxembourg. We also thank the processing and biorepository teams at the Integrated Biobank of Luxembourg (IBBL). We thank the LIH IT service unit to support and secure our resource sharing. This work was initially supported by Luxembourg Personalized Medicine Consortium (PMC, 2015), followed by Luxembourg National Research Fund (FNR) PRIDE programs (11012546/NEXTIMMUNE and 10907093/CriTiCS) and FNR AFR program (PHD-2015-1/9989160) and EAACI long-term fellowship (2016, S.F.N.B.). P.W. has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (grant agreement No. 863664) and the Luxembourg National Research Fund (FNR) under the grant CORE/15/BM/10404093. D.B. is supported by the FNR-CORE program (C18/BM/12691266 and C21/BM/15796788). A.P. is supported by the Action Lions “Vaincre le Cancer”. Some of the icons in our Figures were created from the BioRender pre-drawn library.

Author Contributions:

D.P., S.F.N.B., F.C.M. and C.L.W. collected clinical information and samples. D.P., S.F.N.B., F.H., A.B., A.D., C.M.C, N.Z., A.P., W.A., M.K., S.C., S.D., I.W., G.M., D.R. and O.D. contributed to experimental procedures. C.G. and L.G. designed and supervised sorting. F.H. performed supervised CyTOF analysis. E.G. performed RNA-Seq analysis. O.H. performed unsupervised CyTOF analysis. G.M. and O.D. performed Breg analysis. J.L. performed kinome analysis. X.H. performed integrative computational analysis. P.A. contributed to large-scale data management. F.H. and A.C. contributed to interactive resource sharing. R.S., J.Z., M.H.S., D.B., F.B., P.W., A.C. and J.G. provided supervision into the various subprojects. M.M. and F.H. coordinated the clinical work. F.Q.H. and M.O. designed and oversaw the entire project. F.Q.H and M.O. wrote and revised the manuscript.

Declaration of Interests:

The authors F.Q.H. and M.O. have filed an EP patent application on some of the results generated from this work. The remaining coauthors declare no competing interests.

Aalberse, R. C., Stapel, S. O., Schuurman, J. and Rispens, T. (2009). "Immunoglobulin G4: an odd antibody." Clin Exp Allergy 39(4): 469-477. DOI: 10.1111/j.1365-2222.2009.03207.x.
Aguilar-Pimentel, J. A., Alessandrini, F., Huster, K. M., Jakob, T., Schulz, H., Behrendt, H., Ring, J., Angelis, M. H. d., Busch, D. H., Mempel, M. and Ollert, M. (2010). "Specific CD8 T Cells in IgE-mediated Allergy Correlate with Allergen Dose and Allergic Phenotype." American Journal of Respiratory and Critical Care Medicine 181(1): 7-16. DOI: 10.1164/rccm.200902-0190OC.
Akdis, M. and Akdis, C. A. (2014). "Mechanisms of allergen-specific immunotherapy: multiple suppressor factors at work in immune tolerance to allergens." J Allergy Clin Immunol 133(3): 621-631. DOI: 10.1016/j.jaci.2013.12.1088.
Annunziato, F., Cosmi, L., Liotta, F., Maggi, E. and Romagnani, S. (2012). "Defining the human T helper 17 cell phenotype." Trends in Immunology 33(10): 505-512. DOI: https://doi.org/10.1016/j.it.2012.05.004.
Aoki, M., Aoki, H., Ramanathan, R., Hait, N. C. and Takabe, K. (2016). "Sphingosine-1-Phosphate Signaling in Immune Cells and Inflammation: Roles and Therapeutic Potential." Mediators Inflamm 2016: 8606878. DOI: 10.1155/2016/8606878.
Bacher, P., Heinrich, F., Stervbo, U., Nienen, M., Vahldieck, M., Iwert, C., Vogt, K., Kollet, J., Babel, N., Sawitzki, B., Schwarz, C., Bereswill, S., Heimesaat, M. M., Heine, G., Gadermaier, G., Asam, C., Assenmacher, M., Kniemeyer, O., Brakhage, A. A., Ferreira, F., Wallner, M., Worm, M. and Scheffold, A. (2016). "Regulatory T Cell Specificity Directs Tolerance versus Allergy against Aeroantigens in Humans." Cell 167(4): 1067-1078.e1016. DOI: https://doi.org/10.1016/j.cell.2016.09.050.
Bedard, P. L., Hansen, A. R., Ratain, M. J. and Siu, L. L. (2013). "Tumour heterogeneity in the clinic." Nature 501(7467): 355-364. DOI: 10.1038/nature12627.
Blair, P. A., Noreña, L. Y., Flores-Borja, F., Rawlings, D. J., Isenberg, D. A., Ehrenstein, M. R. and Mauri, C. (2010). "CD19+CD24hiCD38hi B Cells Exhibit Regulatory Capacity in Healthy Individuals but Are Functionally Impaired in Systemic Lupus Erythematosus Patients." Immunity 32(1): 129-140. DOI: https://doi.org/10.1016/j.immuni.2009.11.009.
Bonaguro, L., Schulte-Schrepping, J., Ulas, T., Aschenbrenner, A. C., Beyer, M. and Schultze, J. L. (2022). "A guide to systems-level immunomics." Nat Immunol 23(10): 1412-1423. DOI: 10.1038/s41590-022-01309-9.
Calderon, M. A., Alves, B., Jacobson, M., Hurwitz, B., Sheikh, A. and Durham, S. (2007). "Allergen injection immunotherapy for seasonal allergic rhinitis." Cochrane Database Syst Rev(1): CD001936. DOI: 10.1002/14651858.CD001936.pub2.
Capelle, C. M., Ciré, S., Domingues, O., Ernens, I., Hedin, F., Fischer, A., Snoeck, C. J., Ammerlaan, W., Konstantinou, M., Grzyb, K., Skupin, A., Carty, C. L., Hilger, C., Gilson, G., Celebic, A., Wilmes, P., Del Sol, A., Kaplan, I. M., Betsou, F., Abdelrahman, T., Cosma, A., Vaillant, M., Fagherazzi, G., Ollert, M. and Hefeng, F. Q. (2022). "Combinatorial analysis reveals highly coordinated early-stage immune reactions that predict later antiviral immunity in mild COVID-19 patients." Cell Reports Medicine 3(4): 100600. DOI: https://doi.org/10.1016/j.xcrm.2022.100600.
Capelle, C. M., Ciré, S., Hedin, F., Hansen, M., Pavelka, L., Grzyb, K., Kyriakis, D., Hunewald, O., Konstantinou, M., Revets, D., Tslaf, V., Marques, T. M., Gomes, C. P. C., Baron, A., Domingues, O., Gomez, M., Zeng, N., Betsou, F., May, P., Skupin, A., Cosma, A., Balling, R., Krüger, R., Ollert, M. and Hefeng, F. Q. (2023). "Early-to-mid stage idiopathic Parkinson’s disease shows enhanced cytotoxicity and differentiation in CD8 T-cells in females." Nature Communications 14(1): 7461. DOI: 10.1038/s41467-023-43053-0.
Carding, S. R. and Egan, P. J. (2002). "γδ T cells: functional plasticity and heterogeneity." Nature Reviews Immunology 2(5): 336-345. DOI: 10.1038/nri797.
Chatterjee, S., Daenthanasanmak, A., Chakraborty, P., Wyatt, M. W., Dhar, P., Selvam, S. P., Fu, J., Zhang, J., Nguyen, H., Kang, I., Toth, K., Al-Homrani, M., Husain, M., Beeson, G., Ball, L., Helke, K., Husain, S., Garrett-Mayer, E., Hardiman, G., Mehrotra, M., Nishimura, M. I., Beeson, C. C., Bupp, M. G., Wu, J., Ogretmen, B., Paulos, C. M., Rathmell, J., Yu, X. Z. and Mehrotra, S. (2018). "CD38-NAD(+)Axis Regulates Immunotherapeutic Anti-Tumor T Cell Response." Cell Metab 27(1): 85-100 e108. DOI: 10.1016/j.cmet.2017.10.006.
Chen, Y.-H., Spencer, S., Laurence, A., Thaventhiran, J. E. D. and Uhlig, H. H. (2021). "Inborn errors of IL-6 family cytokine responses." Current Opinion in Immunology 72: 135-145. DOI: https://doi.org/10.1016/j.coi.2021.04.007.
Cherukuri, A., Rothstein, D. M., Clark, B., Carter, C. R., Davison, A., Hernandez-Fuentes, M., Hewitt, E., Salama, A. D. and Baker, R. J. (2014). "Immunologic Human Renal Allograft Injury Associates with an Altered IL-10/TNF-<em>α</em> Expression Ratio in Regulatory B Cells." Journal of the American Society of Nephrology 25(7): 1575-1585. DOI: 10.1681/asn.2013080837.
Chiang, D., Chen, X., Jones, S. M., Wood, R. A., Sicherer, S. H., Burks, A. W., Leung, D. Y. M., Agashe, C., Grishin, A., Dawson, P., Davidson, W. F., Newman, L., Sebra, R., Merad, M., Sampson, H. A., Losic, B. and Berin, M. C. (2018). "Single-cell profiling of peanut-responsive T cells in patients with peanut allergy reveals heterogeneous effector T(H)2 subsets." J Allergy Clin Immunol 141(6): 2107-2120. DOI: 10.1016/j.jaci.2017.11.060.
Cho, J. H. and Feldman, M. (2015). "Heterogeneity of autoimmune diseases: pathophysiologic insights from genetics and implications for new therapies." Nature Medicine 21(7): 730-738. DOI: 10.1038/nm.3897.
Croker, B. A., Krebs, D. L., Zhang, J.-G., Wormald, S., Willson, T. A., Stanley, E. G., Robb, L., Greenhalgh, C. J., Förster, I., Clausen, B. E., Nicola, N. A., Metcalf, D., Hilton, D. J., Roberts, A. W. and Alexander, W. S. (2003). "SOCS3 negatively regulates IL-6 signaling in vivo." Nature Immunology 4(6): 540-545. DOI: 10.1038/ni931.
Davis, M. M., Tato, C. M. and Furman, D. (2017). "Systems immunology: just getting started." Nat Immunol 18(7): 725-732. DOI: 10.1038/ni.3768.
Delhalle, S., Bode, S. F. N., Balling, R., Ollert, M. and He, F. Q. (2018). "A roadmap towards personalized immunology." NPJ Syst Biol Appl 4: 9. DOI: 10.1038/s41540-017-0045-9.
Dreschler, K., Bratke, K., Petermann, S., Bier, A., Thamm, P., Kuepper, M., Virchow, J. C. and Lommatzsch, M. (2011). "Impact of immunotherapy on blood dendritic cells in patients with Hymenoptera venom allergy." Journal of Allergy and Clinical Immunology 127(2): 487-494.e483. DOI: https://doi.org/10.1016/j.jaci.2010.12.003.
Eguiluz-Gracia, I., Tay, T. R., Hew, M., Escribese, M. M., Barber, D., O'Hehir, R. E. and Torres, M. J. (2018). "Recent developments and highlights in biomarkers in allergic diseases and asthma." Allergy 73(12): 2290-2305. DOI: https://doi.org/10.1111/all.13628.
Fox, C. J., Hammerman, P. S. and Thompson, C. B. (2005). "Fuel feeds function: energy metabolism and the T-cell response." Nature Reviews Immunology 5(11): 844-852. DOI: 10.1038/nri1710.
Francis, J. N., Till, S. J. and Durham, S. R. (2003). "Induction of IL-10+CD4+CD25+ T cells by grass pollen immunotherapy." J Allergy Clin Immunol 111(6): 1255-1261. DOI: 10.1067/mai.2003.1570.
Galli, S. J., Tsai, M. and Piliponsky, A. M. (2008). "The development of allergic inflammation." Nature 454(7203): 445-454. DOI: 10.1038/nature07204.
Gideon, H. P., Hughes, T. K., Tzouanas, C. N., Wadsworth, M. H., Tu, A. A., Gierahn, T. M., Peters, J. M., Hopkins, F. F., Wei, J.-R., Kummerlowe, C., Grant, N. L., Nargan, K., Phuah, J. Y., Borish, H. J., Maiello, P., White, A. G., Winchell, C. G., Nyquist, S. K., Ganchua, S. K. C., Myers, A., Patel, K. V., Ameel, C. L., Cochran, C. T., Ibrahim, S., Tomko, J. A., Frye, L. J., Rosenberg, J. M., Shih, A., Chao, M., Klein, E., Scanga, C. A., Ordovas-Montanes, J., Berger, B., Mattila, J. T., Madansein, R., Love, J. C., Lin, P. L., Leslie, A., Behar, S. M., Bryson, B., Flynn, J. L., Fortune, S. M. and Shalek, A. K. (2022). "Multimodal profiling of lung granulomas in macaques reveals cellular correlates of tuberculosis control." Immunity 55(5): 827-846.e810. DOI: https://doi.org/10.1016/j.immuni.2022.04.004.
Gieseck, R. L., Wilson, M. S. and Wynn, T. A. (2018). "Type 2 immunity in tissue repair and fibrosis." Nature Reviews Immunology 18(1): 62-76. DOI: 10.1038/nri.2017.90.
Goldberg, A. and Confino-Cohen, R. (2010). "Bee venom immunotherapy – how early is it effective?" Allergy 65(3): 391-395. DOI: https://doi.org/10.1111/j.1398-9995.2009.02198.x.
Golebski, K., Layhadi, J. A., Sahiner, U., Steveling-Klein, E. H., Lenormand, M. M., Li, R. C. Y., Bal, S. M., Heesters, B. A., Vilà-Nadal, G., Hunewald, O., Montamat, G., He, F. Q., Ollert, M., Fedina, O., Lao-Araya, M., Vijverberg, S. J. H., Maitland-van der Zee, A.-H., van Drunen, C. M., Fokkens, W. J., Durham, S. R., Spits, H. and Shamji, M. H. (2021). "Induction of IL-10-producing type 2 innate lymphoid cells by allergen immunotherapy is associated with clinical response." Immunity 54(2): 291-307.e297. DOI: https://doi.org/10.1016/j.immuni.2020.12.013.
Hansen, M. B. (2020). "Interleukin-6 signaling requires only few IL-6 molecules: Relation to physiological concentrations of extracellular IL-6." Immun Inflamm Dis 8(2): 170-180. DOI: 10.1002/iid3.292.
Hedin, F., Konstantinou, M. and Cosma, A. (2021). "Data integration and visualization techniques for post-cytometric analysis of complex datasets." Cytometry Part A 99(9): 930-938. DOI: https://doi.org/10.1002/cyto.a.24359.
Hirai, H., Tanaka, K., Yoshie, O., Ogawa, K., Kenmotsu, K., Takamori, Y., Ichimasa, M., Sugamura, K., Nakamura, M., Takano, S. and Nagata, K. (2001). "Prostaglandin D2 Selectively Induces Chemotaxis in T Helper Type 2 Cells, Eosinophils, and Basophils via Seven-Transmembrane Receptor Crth2." Journal of Experimental Medicine 193(2): 255-262. DOI: 10.1084/jem.193.2.255.
Hunter, C. A. and Jones, S. A. (2015). "IL-6 as a keystone cytokine in health and disease." Nature Immunology 16(5): 448-457. DOI: 10.1038/ni.3153.
Inaoki, M., Sato, S., Shirasaki, F., Mukaida, N. and Takehara, K. (2003). "The Frequency of Type 2 CD8+ T Cells Is Increased in Peripheral Blood from Patients with Psoriasis Vulgaris." Journal of Clinical Immunology 23(4): 269-278. DOI: 10.1023/A:1024584732614.
Jeon, S.-M. (2016). "Regulation and function of AMPK in physiology and diseases." Experimental & Molecular Medicine 48(7): e245-e245. DOI: 10.1038/emm.2016.81.
Karsunky, H., Merad, M., Cozzio, A., Weissman, I. L. and Manz, M. G. (2003). "Flt3 ligand regulates dendritic cell development from Flt3+ lymphoid and myeloid-committed progenitors to Flt3+ dendritic cells in vivo." J Exp Med 198(2): 305-313. DOI: 10.1084/jem.20030323.
Kratochvíl, M., Hunewald, O., Heirendt, L., Verissimo, V., Vondrášek, J., Satagopam, V. P., Schneider, R., Trefois, C. and Ollert, M. (2020). "GigaSOM.jl: High-performance clustering and visualization of huge cytometry datasets." GigaScience 9(11): giaa127. DOI: 10.1093/gigascience/giaa127.
Krieg, C., Nowicka, M., Guglietta, S., Schindler, S., Hartmann, F. J., Weber, L. M., Dummer, R., Robinson, M. D., Levesque, M. P. and Becher, B. (2018). "High-dimensional single-cell analysis predicts response to anti-PD-1 immunotherapy." Nature Medicine 24(2): 144-153. DOI: 10.1038/nm.4466.
Laidler, K. J. and King, M. C. (1983). "Development of transition-state theory." The Journal of Physical Chemistry 87(15): 2657-2664. DOI: 10.1021/j100238a002.
Larché, M., Akdis, C. A. and Valenta, R. (2006). "Immunological mechanisms of allergen-specific immunotherapy." Nature Reviews Immunology 6(10): 761-771. DOI: 10.1038/nri1934.
Lim, P. S., Sutton, C. R. and Rao, S. (2015). "Protein kinase C in the immune system: from signalling to chromatin regulation." Immunology 146(4): 508-522. DOI: https://doi.org/10.1111/imm.12510.
Lozano-Ojalvo, D., Tyler, S. R. and Berin, M. C. (2021). "Is the plasticity of the Th17 subset a key source of allergenic Th2 responses?" Allergy 76(10): 3238-3240. DOI: 10.1111/all.14888.
Luster, A. D. and Tager, A. M. (2004). "T-cell trafficking in asthma: lipid mediators grease the way." Nature Reviews Immunology 4(9): 711-724. DOI: 10.1038/nri1438.
Maintz, L., Bussmann, C., Bieber, T. and Novak, N. (2009). "Contribution of histamine metabolism to tachyphylaxis during the buildup phase of rush immunotherapy." J Allergy Clin Immunol 123(3): 701-703. DOI: 10.1016/j.jaci.2008.11.009.
Mauer, J., Denson, J. L. and Bruning, J. C. (2015). "Versatile functions for IL-6 in metabolism and cancer." Trends Immunol 36(2): 92-101. DOI: 10.1016/j.it.2014.12.008.
Mayer, A., Debuisson, D., Denanglaire, S., Eddahri, F., Fievez, L., Hercor, M., Triffaux, E., Moser, M., Bureau, F., Leo, O. and Andris, F. (2014). "Antigen presenting cell-derived IL-6 restricts Th2-cell differentiation." Eur J Immunol 44(11): 3252-3262. DOI: 10.1002/eji.201444646.
Meguro, K., Suzuki, K., Hosokawa, J., Sanayama, Y., Tanaka, S., Furuta, S., Ikeda, K., Takatori, H., Suto, A., Sakamoto, A., Ohara, O. and Nakajima, H. (2015). "Role of Bcl-3 in the development of follicular helper T cells and in the pathogenesis of rheumatoid arthritis." Arthritis Rheumatol 67(10): 2651-2660. DOI: 10.1002/art.39266.
Menon, M., Bradford, H. F., Haljasmagi, L., Vanker, M., Peterson, P., Wincup, C., Abida, R., Gonzalez, R. F., Bondet, V., Duffy, D., Isenberg, D. A., Kisand, K. and Mauri, C. (2021). "Inactive disease in lupus patients is linked to autoantibodies to type-I interferons that normalize blood IFNα and B cell subsets." medRxiv: 2021.2004.2007.21255049. DOI: 10.1101/2021.04.07.21255049.
Michils, A., Baldassarre, S., Ledent, C., Mairesse, M., Gossart, B. and Duchateau, J. (2000). "Early effect of ultrarush venom immunotherapy on the IgG antibody response." Allergy 55(5): 455-462. DOI: 10.1034/j.1398-9995.2000.00412.x.
Mittal, A., Papa, S., Franzoso, G. and Sen, R. (2006). "NF-κB-Dependent Regulation of the Timing of Activation-Induced Cell Death of T Lymphocytes1." The Journal of Immunology 176(4): 2183-2189. DOI: 10.4049/jimmunol.176.4.2183.
Morita, H., Arae, K., Unno, H., Miyauchi, K., Toyama, S., Nambu, A., Oboki, K., Ohno, T., Motomura, K., Matsuda, A., Yamaguchi, S., Narushima, S., Kajiwara, N., Iikura, M., Suto, H., McKenzie, A. N., Takahashi, T., Karasuyama, H., Okumura, K., Azuma, M., Moro, K., Akdis, C. A., Galli, S. J., Koyasu, S., Kubo, M., Sudo, K., Saito, H., Matsumoto, K. and Nakae, S. (2015). "An Interleukin-33-Mast Cell-Interleukin-2 Axis Suppresses Papain-Induced Allergic Inflammation by Promoting Regulatory T Cell Numbers." Immunity 43(1): 175-186. DOI: 10.1016/j.immuni.2015.06.021.
Moro, K., Yamada, T., Tanabe, M., Takeuchi, T., Ikawa, T., Kawamoto, H., Furusawa, J.-i., Ohtani, M., Fujii, H. and Koyasu, S. (2010). "Innate production of TH2 cytokines by adipose tissue-associated c-Kit+Sca-1+ lymphoid cells." Nature 463(7280): 540-544. DOI: 10.1038/nature08636.
Mrdjen, D., Pavlovic, A., Hartmann, F. J., Schreiner, B., Utz, S. G., Leung, B. P., Lelios, I., Heppner, F. L., Kipnis, J., Merkler, D., Greter, M. and Becher, B. (2018). "High-Dimensional Single-Cell Mapping of Central Nervous System Immune Cells Reveals Distinct Myeloid Subsets in Health, Aging, and Disease." Immunity 48(2): 380-395 e386. DOI: 10.1016/j.immuni.2018.01.011.
Muller, U. R. (2010). "Insect venoms." Chem Immunol Allergy 95: 141-156. DOI: 10.1159/000315948.
Nakayama, T., Hirahara, K., Onodera, A., Endo, Y., Hosokawa, H., Shinoda, K., Tumes, D. J. and Okamoto, Y. (2017). "Th2 Cells in Health and Disease." Annu Rev Immunol 35: 53-84. DOI: 10.1146/annurev-immunol-051116-052350.
Nuñez, N. G., Berner, F., Friebel, E., Unger, S., Wyss, N., Gomez, J. M., Purde, M.-T., Niederer, R., Porsch, M., Lichtensteiger, C., Kramer, R., Erdmann, M., Schmitt, C., Heinzerling, L., Abdou, M.-T., Karbach, J., Schadendorf, D., Zimmer, L., Ugurel, S., Klümper, N., Hölzel, M., Power, L., Kreutmair, S., Capone, M., Madonna, G., Cevhertas, L., Heider, A., Amaral, T., Hasan Ali, O., Bomze, D., Dimitriou, F., Diem, S., Ascierto, P. A., Dummer, R., Jäger, E., Driessen, C., Levesque, M. P., van de Veen, W., Joerger, M., Früh, M., Becher, B. and Flatz, L. (2023). "Immune signatures predict development of autoimmune toxicity in patients with cancer treated with immune checkpoint inhibitors." Med 4(2): 113-129.e117. DOI: https://doi.org/10.1016/j.medj.2022.12.007.
Oleinika, K., Mauri, C. and Salama, A. D. (2019). "Effector and regulatory B cells in immune-mediated kidney disease." Nat Rev Nephrol 15(1): 11-26. DOI: 10.1038/s41581-018-0074-7.
Palanichamy, A., Barnard, J., Zheng, B., Owen, T., Quach, T., Wei, C., Looney, R. J., Sanz, I. and Anolik, J. H. (2009). "Novel Human Transitional B Cell Populations Revealed by B Cell Depletion Therapy." The Journal of Immunology 182(10): 5982-5993. DOI: 10.4049/jimmunol.0801859.
Palm, N. W., Rosenstein, R. K. and Medzhitov, R. (2012). "Allergic host defences." Nature 484(7395): 465-472. DOI: 10.1038/nature11047.
Palomares, O. (2013). "The role of regulatory T cells in IgE-mediated food allergy." J Investig Allergol Clin Immunol 23(6): 371-382; quiz 372 p preceding 382.
Papadopoulou, M., Sanchez Sanchez, G. and Vermijlen, D. (2020). "Innate and adaptive γδ T cells: How, when, and why." Immunological Reviews 298(1): 99-116. DOI: https://doi.org/10.1111/imr.12926.
Pelzer, C. and Thome, M. (2011). "IKKalpha takes control of canonical NF-kappaB activation." Nat Immunol 12(9): 815-816. DOI: 10.1038/ni.2082.
Platts-Mills, T., Vaughan, J., Squillace, S., Woodfolk, J. and Sporik, R. (2001). "Sensitisation, asthma, and a modified Th2 response in children exposed to cat allergen: a population-based cross-sectional study." The Lancet 357(9258): 752-756. DOI: https://doi.org/10.1016/S0140-6736(00)04168-4.
Pulendran, B. and Artis, D. (2012). "New Paradigms in Type 2 Immunity." Science 337(6093): 431-435. DOI: doi:10.1126/science.1221064.
Rickmann, M., Krauter, J., Stamer, K., Heuser, M., Salguero, G., Mischak-Weissinger, E., Ganser, A. and Stripecke, R. (2011). "Elevated frequencies of leukemic myeloid and plasmacytoid dendritic cells in acute myeloid leukemia with the FLT3 internal tandem duplication." Ann Hematol 90(9): 1047-1058. DOI: 10.1007/s00277-011-1231-2.
Robinson, D. S., Hamid, Q., Ying, S., Tsicopoulos, A., Barkans, J., Bentley, A. M., Corrigan, C., Durham, S. R. and Kay, A. B. (1992). "Predominant TH2-like bronchoalveolar T-lymphocyte population in atopic asthma." N Engl J Med 326(5): 298-304. DOI: 10.1056/NEJM199201303260504.
Rudulier, C. D., Tonti, E., James, E., Kwok, W. W. and Larche, M. (2019). "Modulation of CRTh2 expression on allergen-specific T cells following peptide immunotherapy." Allergy 74(11): 2157-2166. DOI: 10.1111/all.13867.
Scadding, G. W., Calderon, M. A., Shamji, M. H., Eifan, A. O., Penagos, M., Dumitru, F., Sever, M. L., Bahnson, H. T., Lawson, K., Harris, K. M., Plough, A. G., Panza, J. L., Qin, T., Lim, N., Tchao, N. K., Togias, A., Durham, S. R. and Immune Tolerance Network, G. S. T. (2017). "Effect of 2 Years of Treatment With Sublingual Grass Pollen Immunotherapy on Nasal Response to Allergen Challenge at 3 Years Among Patients With Moderate to Severe Seasonal Allergic Rhinitis: The GRASS Randomized Clinical Trial." JAMA 317(6): 615-625. DOI: 10.1001/jama.2016.21040.
Scheiermann, C., Gibbs, J., Ince, L. and Loudon, A. (2018). "Clocking in to immunity." Nature Reviews Immunology 18(7): 423-437. DOI: 10.1038/s41577-018-0008-4.
Schluns, K. S. and Lefrançois, L. (2003). "Cytokine control of memory T-cell development and survival." Nature Reviews Immunology 3(4): 269-279. DOI: 10.1038/nri1052.
Shabir, S., Girdlestone, J., Briggs, D., Kaul, B., Smith, H., Daga, S., Chand, S., Jham, S., Navarrete, C., Harper, L., Ball, S. and Borrows, R. (2015). "Transitional B lymphocytes are associated with protection from kidney allograft rejection: a prospective study." Am J Transplant 15(5): 1384-1391. DOI: 10.1111/ajt.13122.
Shamji, M. H., Sharif, H., Layhadi, J. A., Zhu, R., Kishore, U. and Renz, H. (2022). "Diverse immune mechanisms of allergen immunotherapy for allergic rhinitis with and without asthma." Journal of Allergy and Clinical Immunology 149(3): 791-801. DOI: 10.1016/j.jaci.2022.01.016.
Sokol, C. L., Chu, N.-Q., Yu, S., Nish, S. A., Laufer, T. M. and Medzhitov, R. (2009). "Basophils function as antigen-presenting cells for an allergen-induced T helper type 2 response." Nature Immunology 10(7): 713-720. DOI: 10.1038/ni.1738.
Spencer, S., Köstel Bal, S., Egner, W., Lango Allen, H., Raza, S. I., Ma, C. A., Gürel, M., Zhang, Y., Sun, G., Sabroe, R. A., Greene, D., Rae, W., Shahin, T., Kania, K., Ardy, R. C., Thian, M., Staples, E., Pecchia-Bekkum, A., Worrall, W. P. M., Stephens, J., Brown, M., Tuna, S., York, M., Shackley, F., Kerrin, D., Sargur, R., Condliffe, A., Tipu, H. N., Kuehn, H. S., Rosenzweig, S. D., Turro, E., Tavaré, S., Thrasher, A. J., Jodrell, D. I., Smith, K. G. C., Boztug, K., Milner, J. D. and Thaventhiran, J. E. D. (2019). "Loss of the interleukin-6 receptor causes immunodeficiency, atopy, and abnormal inflammatory responses." Journal of Experimental Medicine 216(9): 1986-1998. DOI: 10.1084/jem.20190344.
Stanic, B., van de Veen, W., Wirz, O. F., Ruckert, B., Morita, H., Sollner, S., Akdis, C. A. and Akdis, M. (2015). "IL-10-overexpressing B cells regulate innate and adaptive immune responses." J Allergy Clin Immunol 135(3): 771-780 e778. DOI: 10.1016/j.jaci.2014.07.041.
Sturm, G. J., Varga, E.-M., Roberts, G., Mosbech, H., Bilò, M. B., Akdis, C. A., Antolín-Amérigo, D., Cichocka-Jarosz, E., Gawlik, R., Jakob, T., Kosnik, M., Lange, J., Mingomataj, E., Mitsias, D. I., Ollert, M., Oude Elberink, J. N. G., Pfaar, O., Pitsios, C., Pravettoni, V., Ruëff, F., Sin, B. A., Agache, I., Angier, E., Arasi, S., Calderón, M. A., Fernandez-Rivas, M., Halken, S., Jutel, M., Lau, S., Pajno, G. B., van Ree, R., Ryan, D., Spranger, O., van Wijk, R. G., Dhami, S., Zaman, H., Sheikh, A. and Muraro, A. (2018). "EAACI guidelines on allergen immunotherapy: Hymenoptera venom allergy." Allergy 73(4): 744-764. DOI: https://doi.org/10.1111/all.13262.
Sundrud, M. S. and Trivigno, C. (2013). "Identity crisis of Th17 cells: Many forms, many functions, many questions." Seminars in Immunology 25(4): 263-272. DOI: https://doi.org/10.1016/j.smim.2013.10.021.
Tabar, A. I., Arroabarren, E., Echechipia, S., Garcia, B. E., Martin, S. and Alvarez-Puebla, M. J. (2011). "Three years of specific immunotherapy may be sufficient in house dust mite respiratory allergy." J Allergy Clin Immunol 127(1): 57-63, 63 e51-53. DOI: 10.1016/j.jaci.2010.10.025.
Tortola, L., Jacobs, A., Pohlmeier, L., Obermair, F. J., Ampenberger, F., Bodenmiller, B. and Kopf, M. (2020). "High-Dimensional T Helper Cell Profiling Reveals a Broad Diversity of Stably Committed Effector States and Uncovers Interlineage Relationships." Immunity 53(3): 597-613 e596. DOI: 10.1016/j.immuni.2020.07.001.
Tsigos, C., Papanicolaou, D. A., Kyrou, I., Defensor, R., Mitsiadis, C. S. and Chrousos, G. P. (1997). "Dose-Dependent Effects of Recombinant Human Interleukin-6 on Glucose Regulation." The Journal of Clinical Endocrinology & Metabolism 82(12): 4167-4170. DOI: 10.1210/jcem.82.12.4422.
van de Veen, W., Stanic, B., Yaman, G., Wawrzyniak, M., Sollner, S., Akdis, D. G., Ruckert, B., Akdis, C. A. and Akdis, M. (2013). "IgG4 production is confined to human IL-10-producing regulatory B cells that suppress antigen-specific immune responses." J Allergy Clin Immunol 131(4): 1204-1212. DOI: 10.1016/j.jaci.2013.01.014.
Varona, R., Ramos, T., Escribese, M. M., Jimeno, L., Galan, A., Wurtzen, P. A., Vega, F., Marin, A., Martin, S., Carrera, A. C., Blanco, C. and Barber, D. (2019). "Persistent regulatory T-cell response 2 years after 3 years of grass tablet SLIT: Links to reduced eosinophil counts, sIgE levels, and clinical benefit." Allergy 74(2): 349-360. DOI: 10.1111/all.13553.
von Mutius, E. and Smits, H. H. (2020). "Primary prevention of asthma: from risk and protective factors to targeted strategies for prevention." Lancet 396(10254): 854-866. DOI: 10.1016/S0140-6736(20)31861-4.
Wambre, E. (2015). "Effect of allergen-specific immunotherapy on CD4+ T cells." Current Opinion in Allergy and Clinical Immunology 15(6): 581-587. DOI: 10.1097/aci.0000000000000216.
Wambre, E., Bajzik, V., DeLong, J. H., O'Brien, K., Nguyen, Q. A., Speake, C., Gersuk, V. H., DeBerg, H. A., Whalen, E., Ni, C., Farrington, M., Jeong, D., Robinson, D., Linsley, P. S., Vickery, B. P. and Kwok, W. W. (2017). "A phenotypically and functionally distinct human TH2 cell subpopulation is associated with allergic disorders." Sci Transl Med 9(401). DOI: 10.1126/scitranslmed.aam9171.
Wang, C., Barnoud, C., Cenerenti, M., Sun, M., Caffa, I., Kizil, B., Bill, R., Liu, Y., Pick, R., Garnier, L., Gkountidi, O. A., Ince, L. M., Holtkamp, S., Fournier, N., Michielin, O., Speiser, D. E., Hugues, S., Nencioni, A., Pittet, M. J., Jandus, C. and Scheiermann, C. (2022). "Dendritic cells direct circadian anti-tumor immune responses." Nature. DOI: 10.1038/s41586-022-05605-0.
Waskow, C., Liu, K., Darrasse-Jeze, G., Guermonprez, P., Ginhoux, F., Merad, M., Shengelia, T., Yao, K. and Nussenzweig, M. (2008). "The receptor tyrosine kinase Flt3 is required for dendritic cell development in peripheral lymphoid tissues." Nat Immunol 9(6): 676-683. DOI: 10.1038/ni.1615.
Wei, C.-W., Lee, C.-Y., Lee, D.-J., Chu, C.-F., Wang, J.-C., Wang, T.-C., Jane, W.-N., Chang, Z.-F., Leu, C.-M., Dzhagalov, I. L. and Hsu, C.-L. (2018). "Equilibrative Nucleoside Transporter 3 Regulates T Cell Homeostasis by Coordinating Lysosomal Function with Nucleoside Availability." Cell Reports 23(8): 2330-2341. DOI: 10.1016/j.celrep.2018.04.077.
Yu, W., Zhou, X., Dunham, D., Lyu, S. C., Manohar, M., Zhang, W., Zhao, F., Davis, M. M. and Nadeau, K. (2019). "Allergen-specific CD8(+) T cells in peanut-allergic individuals." J Allergy Clin Immunol 143(5): 1948-1952. DOI: 10.1016/j.jaci.2019.01.011.

Yes there is potential Competing Interest. The authors F.Q.H. and M.O. have filed an EP patent application on some of the results generated from this work. The remaining coauthors declare no competing interests.

SupplementalFiguresandTables.docx

Download PDF

Journal Publication

published 26 Nov, 2024

Read the published version in Nature Communications →

Version 1

posted

You are reading this latest preprint version

Multiomics approaches disclose very-early molecular and cellular switches during insect-venom allergen-specific immunotherapy

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Revealing disturbed CD4 + and CD8 + T-cell subsets during the initiation phase of AIT

Multiple aberrant hybrid subsets characterize the initiation phase of AIT

Discussion

Methods

Sorting Th2 cells and cell-type-specific mRNA sequencing

Th2-cell-specific RNA-seq pre-processing and quality control

Analysis of potential confounding factors

Filtering, normalization, voom transformation and quality control

Differential expression and enrichment analysis

Interactive and interlinked resource (‘i3Dare’) sharing

Declarations

Data Availability:

Acknowledgements:

Author Contributions:

Declaration of Interests:

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1