Food allergy associated immune remodelling during OIT
Sequencing bulk cell populations such as whole blood or peripheral blood mononuclear cells (PBMC) to identify biomarkers of allergy or treatment success is an attractive endeavour due to the relative ease of collecting this type of specimen in clinical settings. Blood is routinely collected in these settings and a marker measured directly with little to no need for cell culture would be practical for ease of implementation into clinical practice. This “baseline predicts outcome” approach has been successfully applied to predict the immune response to vaccination using systems biology methods14. However, one important limitation to this approach is lack of resolution since allergen reactive cells make up only a tiny fraction of the overall whole blood or PBMC population and do not persist for very long in the peripheral circulation15. A second challenge is that the cellular composition of these fractions can be highly variable between individuals and across different age groups, limiting the likelihood of identifying a universal biomarker 16. Despite these challenges, transcriptional studies have identified immune changes in whole blood and PBMC that are linked to peanut or egg allergy, acute allergic reactions, and OIT treatment effects.
In egg OIT, downregulation of genes involved in TREM1 signaling, IL-6 and IL-17 were identified in PBMC after eight months of treatment17. Changes were linked to regulation of innate immunity and inflammation: Toll-like receptors (TLRs), interferons (IFNs), chemokines, cytokines, and adhesion molecules; and regulation of T cell genes including Th17 cells, regulatory T cells, and Th2 cells. Comparing outcome groups, expression profiles were largely overlapping between groups, however there was some evidence of delayed immunological remodelling in the group who did not reach the target dose (n=23; partially desensitised) versus those who reached the target maintenance dose (n=22). Characterised by persistence of Th2 responses and lack of upstream regulator engagement in the partially desensitised group. Interpretation of this study in the context of lasting outcomes following OIT was limited as outcomes were defined by the ability of participants to reach the target dose of 1g following 8 months of build-up. Because there is no longitudinal assessment of clinical outcome following treatment it is not possible to delineate whether molecular changes were linked to remission or desensitisation without remission.
Several differentially regulated genes have been reported following OIT with omalizumab in whole blood18. Of interest, 108 of the 680 differentially expressed genes (DEG) associated with peanut OIT + omalizumab overlapped with DEG linked to acute allergic reactions to peanut in another study19. The common DEG between the two studies primarily had opposing direction of effect, indicating that treatment effects were in part mediated through modulation of downstream events that occur during an acute allergic reaction. There were no consistent transcriptional changes linked to omalizumab alone prior to peanut OIT. However, the cohort was a mix of those who managed to tolerate the maintenance dose (n=11), and those who failed to tolerate the maintenance dose (n=6). Transcriptional changes were not analysed based on clinical outcomes. Instead, the analysis was adjusted to account for changes associated with specific outcomes. Therefore, the detected immune changes are not linked to treatment success but rather to treatment independent of outcome18. Comparing DEG between the egg and peanut OIT transcriptional studies, there was little overlap between the early transcriptional signals detected in the egg study (3 months into OIT) compared to the genes associated with omalizumab + peanut OIT (measured after 2-3 years on treatment). Of the genes that did overlap between the two studies, the majority were not detected with the same direction of effect. This finding may be due to the considerably distant windows at which DEG was measured in each study, each possibly capturing distinct phases of immunological activity. Alternatively, differences may be due to other variable factors between the two populations, such as differences in treatment regime, genetic background of individuals, age or other demographic factors. Genes that were consistent between the two studies were not immune genes but rather genes associated with cellular processes such as transport proteins, protein synthesis, mitochondrial function.
Although there has been interest in the potential role of T follicular helper 2-like cells in allergy pathogenesis20, a study applying single-cell sequencing reported that OIT did not modify this population21. The T follicular helper 2-like cell population was, however, positively associated with peanut sIgE production in the same OIT population. Another study identified that changes in gamma delta Tregs (γδTregs) may play a role in immunological mechanisms of OIT. Reduced numbers, comparable to those of health controls were observed by week 6 of OIT22. However, data are only available within a limited window of time and no food challenges were performed to determine clinical outcomes.
B and T cell receptor profiling may also provide further mechanistic insight. A novel method using gene expression profiling was recently developed and applied to study how OIT modifies the B cell repertoire in n=17 peanut allergic individuals before and after treatment23. Enrichment of selected immunoglobulin heavy chain alpha and gamma clones was identified following OIT. TCR beta-chain profiling showed that peanut allergy was associated with a pool of convergent activated clones, with 17% of the sequences shared between individuals. The frequency of these clones in the effector cell, but not regulatory cell compartment was higher in those with more reactive phenotypes24. Application of methods to study allergen-specific B and T cell receptors, alongside gene expression profiling, in larger cohorts may reveal insights into selective clonal anergy or other patterns linked with clinical outcomes following OIT.
Key findings from studies applying techniques to study genes within co-expression networks have identified genes that are predicted to be drivers of the allergic response. By examining the transcriptional profile of whole blood samples (n=19) across a time course (baseline, 2hr, 4hr) following food challenge, researchers were able to identify six genes LTB4R, PADI4, IL1R2, PPP1R3D, KLHL2, and ECHDC3 as having causal roles during peanut reactions19. Using similar methods to examine whole blood collected from children with (n=23) or without (n=7) nut allergy, another group identified three gene drivers of nut allergy; IFIH1, DRAM1 and ZNF512B. A module enriched for type I interferon genes, with IFIH1 as the key driver, was most positively associated with nut allergy25.