Evolution of IgE responses to multiple allergen components throughout childhood

Background There is a paucity of information about longitudinal patterns of IgE responses to allergenic proteins (components) from multiple sources. Objectives This study sought to investigate temporal patterns of component-specific IgE responses from infancy to adolescence, and their relationship with allergic diseases. Methods In a population-based birth cohort, we measured IgE to 112 components at 6 follow-ups during childhood. We used a Bayesian method to discover cross-sectional sensitization patterns and their longitudinal trajectories, and we related these patterns to asthma and rhinitis in adolescence. Results We identified 1 sensitization cluster at age 1, 3 at age 3, 4 at ages 5 and 8, 5 at age 11, and 6 at age 16 years. “Broad” cluster was the only cluster present at every follow-up, comprising components from multiple sources. “Dust mite” cluster formed at age 3 years and remained unchanged to adolescence. At age 3 years, a single-component “Grass” cluster emerged, which at age 5 years absorbed additional grass components and Fel d 1 to form the “Grass/cat” cluster. Two new clusters formed at age 11 years: “Cat” cluster and “PR-10/profilin” (which divided at age 16 years into “PR-10” and “Profilin”). The strongest contemporaneous associate of asthma at age 16 years was sensitization to dust mite cluster (odds ratio: 2.6; 95% CI: 1.2-6.1; P < .05), but the strongest early life predictor of subsequent asthma was sensitization to grass/cat cluster (odds ratio: 3.5; 95% CI: 1.6-7.4; P < .01). Conclusions We describe the architecture of the evolution of IgE responses to multiple allergen components throughout childhood, which may facilitate development of better diagnostic and prognostic biomarkers for allergic diseases.

Allergic sensitization is a risk factor for asthma and rhinitis 1-3 , but the strength of this association 78 is inconsistent 4,5 . A patient is typically deemed to be sensitized based on a positive skin prick 79 test (SPT) or a blood test measuring specific IgE to a range of common inhalant and food 80 allergens 6,7 . However, both these tests can be positive without the patient having any 81 symptoms 8 , and neither positive SPT nor IgE can confirm the expression of symptoms upon 82 allergen exposure 8,9 . This is partly because the natural sources which are used to prepare the 83 whole-allergen extracts for skin or blood testing contain multiple allergenic proteins 84 (components), with each component potentially containing multiple epitopes for binding IgE 10 . 85 There is increasing evidence that sensitization to some, but not all of these proteins is important Manchester Asthma and Allergy Study is an unselected birth cohort; participants were recruited 121 prenatally and followed prospectively 24,25 . The study was approved by the Research Ethics 122 Committee; parents gave written informed consent. Participants attended review clinics at ages 123 1, 3, 5, 8, 11 and 16 years. Validated questionnaires were interviewer-administered to collect 124 information on parentally-reported symptoms, physician-diagnosed diseases and treatments 125 received. Blood samples were collected from participants who gave assent. 126 Detection and annotation of component-specific IgE antibodies 127 We measured IgE to 112 components from 51 sources using ImmunoCAP ISAC (Thermo respectively 34,35 . Once the optimal number of clusters, K, was inferred at each age, the cluster 158 membership was inferred conditional on that value. 159 Associations with clinical outcomes: CRD data from ages 1 and 3 years were sparse; we 160 therefore evaluated the association between component clusters at ages 5 and 16 years with 161 asthma and rhinitis at age 16 years. Children who did not respond to any active component 162 were a priori assigned to a "Non-sensitized" group. A child was classed as being sensitized to a 163 component cluster if he/she responded to at least 1 component within the cluster. We examined 164 the association between sensitization to component clusters and clinical outcomes (asthma, 165 wheeze and rhinitis) through logistic regression analyses (univariable and multiple); results are 166 reported as odds ratios (OR) with 95% confidence intervals (CI).   Table   178 S2; note, one or two children had positive IgE to some of these components, and for 3 179 components (Asp f 1, Bla g 5, Hev b 5) there was no positive response in any subject at any 180 age. Inactive components at each age are listed in Table S3. Table S4 shows 24 components which dropped out (not necessarily permanently), and number 182 of children who were sensitized to these components. Figure S2 shows detailed longitudinal 183 response profiles of each component that ever becomes inactive after first becoming active, for 184 each child who has ever responded; for 12 components, we linked their drop-out to the  Table 2 shows the number of component clusters inferred at each time point, and their posterior 189 probabilities determined using Bayesian inference. The optimal solution identified one 190 sensitization cluster at age one, 3 at age three, 4 at ages five and eight, 5 at age 11, and 6 at 191 M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT 9 age 16 years. The posterior probabilities for the most probable number of clusters were at least 192 0.87 for the first five time points, and remained above 0.70 at age 16 years. Tables S5-S10 list 193 components in each cluster at each time point. 194 We qualitatively labelled clusters at each age based on the profile of allergen components to 195 which sensitization occurred. Figure 1 shows the number of active components contained within 196 each cluster for each time point (red), how many components were inactive (blue), and how 197 many components were shared between clusters at adjacent time points.

198
The "Broad" cluster comprising of components originating from multiple sources was the only 199 cluster identified at every time point. Components forming this cluster differed at different ages; 200   Table S11 shows 24 components which were only ever assigned to the "Broad" cluster.

201
From age three onwards, the "HDM" cluster formed and remained unchanged by age 16, 202 consisting of four mite components (Der p 1-2, Der f 1-2). Also at age three, the "Grass" cluster 203 emerged, consisting of a single component (Phl p 1; Table S6). This cluster absorbed an 204 additional 3 grass components, as well as cat component Fel d 1 to form the "Grass/cat" cluster 205 at age five (Table S7). The membership of this cluster remained unchanged at age eight, 206 although Fel d 1 assignment probability was reduced from >0.95 at age five to 0.70 (Table S8).

207
A further cluster that was shared across ages five and eight was the "Alternaria" cluster, 208 comprising of only Alt a 1. At age 11, this component was reabsorbed by the "Broad" cluster, the 209 only component to do so throughout this flow (Figure 1).

210
Two new clusters formed at age 11 years: the "Cat" cluster (comprising of Fel d 1) and the "PR-211 10/profilin" cluster (Table S9). The latter was composed solely of components which have 212 moved from the "Broad" cluster at age 8. Additional grass components were absorbed from the 213 "Broad" into the "Grass" cluster at age 11 years (Phl p 2 and Phl p 6). This cluster divided at age 214 16 into two: "PR-10" and "Profilin" (Table S10); other clusters remained unchanged at age 16.  (Table S4) were assigned only to the "Broad" 219 cluster. Components from all other clusters remained active once they first became so.  Table   222 S12. For children who were sensitized to at least one cluster at age 5, the most common 223 response (n=42) was to the "Grass" cluster only. The confusion matrix in Table S13 displays the 224 number of children who shared sensitization to the clusters at ages 5 and 16, for 255 children 225 who had CRD data at both follow-ups. Of 62 children who were sensitized to "Broad" cluster at 226 age 5, 53 went on to respond to "Grass" cluster at age 16, with 51 remaining sensitized to the 227 "Broad" cluster as well.  respectively). When we extended the time frame to investigate the relationship between cluster 243 sensitization at age 5 years and clinical outcomes at age 16 ( Figure 3b, Table S14), there was 244 no significant association between asthma and sensitization to "Broad" and "HDM" clusters, and 245 the strongest risk of subsequent asthma was conferred by sensitization to the "Grass/cat" and  We describe the architecture of the evolution of IgE responses to multiple allergen components 253 throughout childhood, taking into account responses to more than 100 allergenic molecules. By 254 applying novel machine learning techniques to CRD sensitization data from infancy to 255 adolescence among children from a population-based birth cohort, we identified latent structure 256 in the diversification of the IgE responses during childhood (Figures 1 and 2). Our  While children were frequently sensitized to more than one cluster, sensitization to distinct 261 clusters was associated with different clinical presentations, indicating that some sensitization 262 patterns pose greater risk for the development of specific clinical symptoms than others.

263
One of the limitations of our study includes the lack of potentially important components which 264 are not included on the ISAC chip, such as those from HDM and fungi (e.g. ISAC has 6/109 265 fungal allergens identified in IUIS). This may be one of the reasons why the "Alternaria" cluster 266 contained only one component (Alt a 1). Of note, sensitization to this small cluster at age 5 267 years conferred a strong risk for asthma in later life. This is also of relevance to the "HDM" 268 cluster, which was the only cluster to remain unchanged once it had formed at age three, with  We acknowledge that the number of sensitized children in early life was small (only 10/226 at 275 age 1 year), and we cannot exclude the possibility that this may have introduced bias in our analyses. However, we believe that presenting data at all ages is important to ascertain the life-277 course perspective. 278 We were unable to determine the effect of partial or complete sensitization to each cluster, and 279 the relative importance of sensitization to specific "lead" component(s) compared the number of 280 components within each cluster. This question will need to be addressed in future studies. We 281 also acknowledge that our study population comes from a specific geographical area, and that  may also indicate that the sensitization process has started earlier. Our data extend the 299 relatively broad concepts of "polysensitization" and "early sensitization" to demonstrate that for a 300 more precise ascertainment of future and current risk of allergic diseases, we need accurate information about the specific patterns of sensitization to unique sets of allergenic molecule, as 302 well as the timing of onset of sensitization. 303 Our results suggest that the timing of onset of specific sensitization patterns may be a key 304 indicator of future risk, and that apparently similar cross-sectional profiles of component-specific 305 IgE responses may have different clinical associations depending on the age at which they 306 emerge. This expands upon our previous study in which we used a limited number of Timothy 307 grass and HDM components, which described two grass pollen IgE trajectories ("Late onset" 308 and "Early onset") 19 . Although the progression of IgE component responses over time was 309 identical in the two trajectories, following the sequence of Phl p 1/5Phl p 2/4/6Phl p 7/11/12, 310 their clinical associations were different. The "Early onset" trajectory (in which Phl p 1/5 IgE 311 responses emerged in preschool age) was associated with asthma and multimorbidity, while the 312 "Late onset" trajectory (in which the same component-specific IgE responses were first 313 observed in the school-age) was associated with rhinitis 19 . At the time when we conducted 314 previous analyses, limitations including computing power and available methodologies 315 precluded us from investigating the developmental pathways across all 112 components. In the 316 current study, a more complex structure emerged. This is highlighted by the emergence of 317 "Grass/cat" cluster at age 5 years, in which allergenic proteins from diverse sources, and with a 318 fundamentally different function, clustered together. Although it may appear counterintuitive that 319 Fel d 1 should be in the same cluster as the Timothy and Bermuda grass components, the 320 assignment probability for the cat component belonging to this cluster was very high (0.97). The 321 response to this cluster was strongly associated with asthma at age 16 years (3.5-fold increase

Screening & Recruitment
All pregnant women were screened for eligibility at antenatal visits (8 th -10 th week of pregnancy).
Of the 1499 couples who met the inclusion criteria (<10 weeks of pregnancy, maternal age >18 years), 288 declined to take part and 27 were lost to follow-up between recruitment and birth of a child. A total of 1184 participants had some evaluable data.

Follow-up
Children have been followed prospectively, and attended review clinics at ages 1, 3, 5, 8, 11 and 16 years. At age 1 year, only children with either both atopic parents, or no atopic parents were invited to attend for clinical follow up. At all other time points for all other measures all children were invited to participate.

Statistical grouping of allergen components
We assumed that there exist clusters of components to which subjects have similar IgE responses (i.e., either being sensitized or not to most of the components within the same cluster). At each age, we inferred component clusters by clustering the data through Bayesian The observed likelihood for a binary data matrix x under the BMM model is given by: where is between 0 and 1 and represents the frequency of sensitisation to component j for subjects in cluster k and represents the weight of cluster k which is the prior probability that a subject belongs to that cluster.

Associations with clinical outcomes
The relationships between a subject's responses to the BMM's cluster output and their disease outcomes were assessed using univariable and multiple logistic regression analyses (adjusting for sensitization to each of the component clusters, and the sex of the child). In addition to frequentist intervals, we calculated Bayesian posterior credible regions. Table S1. Number of children with component-resolved diagnostics data and proportion of those with at least one positive allergen component response at each follow-up Table S2. The list of 26 components which were labelled inactive at all 6 time points Table S3. Components labelled inactive for ages a) 1, b) 3, c) 5, d) 8, e) 11, and f) 16 years. Table S4. Subject response totals for each of the allergen components which "dropped-out" (i.e. become inactive after first being active). Italics indicates when that component is "Active"; bold if "inactive" but 1 or 2 subjects have positively responded to that component at that time point; otherwise if "inactive" and no subjects have positively responded at that time.       Table S11. Components which were only ever assigned to the "Broad" cluster.  Table S13. Confusion matrix for the reduced response frequencies of the 255 children that had ISAC data for both ages 5 and 16. The clusters from age 5: the left, the clusters from age 16: the top. Note that rows and columns do not sum to the totals, as responses to the clusters are not mutually exclusive. Note the relatively small proportion of children that have reduced responses to each of the clusters at each of these ages (but particularly at age 5), acting as a main source for a wide range in confidence intervals for associations with clinical outcomes.

RESULTS
Table S14. C-statistic reported for each of the multivariate logistic regression models applied to both age 5 and age 16's cluster response data, with relation to rhinitis and asthma-related clinical outcomes at age 16. b) The remaining 12 components, all of whose drop-outs can be explained by the subject loss to follow-up. Figure S3. Odds ratios and 95% CIs from univariable logistic regression for asthma and rhinitis at age 16, based on subjects' reduced responses to component clusters at (a) age 16; (b) age 5. Bayesian posterior credible regions were also computed, and agreed closely with the 95% CIs shown.  Perennially inactive components M A N U S C R I P T A C C E P T E D ACCEPTED MANUSCRIPT Table S3c Age 5