Keywords: Allergic asthma, Atopic dermatitis, Immune
disorders, Transcriptomic profiling, Type I hypersensitivity
Introduction
Allergy of type I hypersensitivity is a severe condition experienced by
about 150 million people in Europe1. Type I
hypersensitivity can be manifested as atopic dermatitis, rhinitis,
conjunctivitis, rhinoconjuctivitis or allergic asthma. These
manifestations very often appear together2. In
addition, they usually occur gradually from atopic dermatitis to
allergic asthma, collectively known as the atopic
march3, 4. All of these clinical signs, even if they
are distinct, come from a single intrinsic process: IgE mediated (type
I) hypersensitivity. In type I hypersensitivity, T helper type 2 (Th2)
cells play a key role in the patient response to antigens by activating
immunoglobulin (Ig) E sensitization3, 5.
The type I hypersensitivity reaction occurs when a mucosal surface (i.
e. respiratory or gastrointestinal tract) or the skin is exposed to an
antigen, which causes an allergy (allergen)6, 7. The
immune response to the antigen begins with the sensitization
phase7. During this phase, the antigen is captured by
antigen-presenting cells (APCs), and subsequently presented to naive T
cells7, 8. This interaction leads to the activation of
naive T cells, which then secrete interleukin-4 (IL-4), which together
with IL-4 produced by basophils and mast cells, initiate their
differentiation into Th2 cells7, 9..Additionally, differentiated Th2 cells further secrete IL-4 and IL-13
which stimulate B cell activation, leading to class-switching and
production of IgE antibodies by activated B cells7, 8,
10. Consequently, the binding of IgE antibodies to high-affinity IgE
receptors (FcεRI) on mast cells and basophils cause their sensitization,
and this leads to the subsequent effector phase. The effector phase of
type I hypersensitivity starts with cross-linking of the antigen and IgE
antibodies bound to FcεRI on mast cells and
basophils10, 11. The cross-linking triggers the
activation of mast cells and basophils leading to their degranulation.
During degranulation, inflammatory mediators in granules, such as
histamine, leukotrienes, and cytokines, are rapidly released, which
induce characteristic hypersensitivity symptoms, such as itching,
sneezing, nasal congestion, wheezing, dermatitis, or gastrointestinal
symptoms7, 8, 11, 12. When re-exposure to the allergen
occurs, the type I hypersensitivity response is
immediate8.
Although the general progression of type I hypersensitivity reaction is
well known, the underlying mechanisms at the transcript level remains
understudied. To understand the complex transcriptomic mechanisms
underscoring patients with type I hypersensitivity symptoms, we
investigated the blood transcriptome of adult participants from Central
European Longitudinal Study of Parents and Children: Young Adults
(CELSPAC: YA) cohort study. We performed differential gene expression
analysis (DGEA) and compared transcriptomic profiles of participants
with allergic reaction to population without any of allergic triggers or
symptoms. By analysing differences in gene expression and identifying
key biomarkers, we aim to better understand how these conditions develop
and potentially predict their future progression. The analysis of the
blood transcriptome not only serves as a non-invasive and accessible
source of information, but also offers a comprehensive view of the
systemic changes occurring in response to antigens. Such knowledge can
pave the way for more targeted and effective therapeutic interventions,
leading to improved patient outcomes and ultimately, better management
of type I hypersensitivity.
Materials and methods
Study population and sample collection
The data analysed in this study is derived from the CELSPAC: YA cohort
study (ethical approval no. ELSPAC/EK/2/2019), a follow-up study of
European Longitudinal Study of Pregnancy and Childhood (ELSPAC-CZ)
performed in Czech Republic (for more detailed description of ELSPAC-CZ,
see Piler et al. (2017)13). Participants underwent
blood sampling and completed comprehensive questionnaires. The
questionnaire data contain information about participants such as their
age, BMI, education, smoking status, alcohol consumption, and health
outcome. The health outcome information includes information on
immune-related diseases, such as atopic dermatitis, asthma, allergies to
food, drugs, insects and pollen, dust, or mite, contact dermatitis,
celiac disease, psoriasis, or lupus. The detailed information of
CELSPAC: YA cohort is provided in our recent study by Rudzanova et al.
(2023)14.
All participants with non-allergic triggers or symptoms (e.g. contact
dermatitis) were excluded from the analysis to avoid coincidence with
inflammation other than type I hypersensitivity. To analyse the
underlying mechanisms of type I hypersensitivity, we formed 3 test
groups: 1) subjects with allergy: All participants who answered “yes”
to the question “Have you ever been diagnosed with an allergy (pollen,
dust mite, food, insects), asthma or atopic dermatitis by a doctor? ”
in a questionnaire; 2) Subjects with allergy without asthma and atopic
dermatitis: All whose answered “yes” to the questionnaire question
“Have you ever been diagnosed by a doctor with an allergy (pollen, dust
mite, food, insect)” and simultaneously answered “no” to the question
“Have you ever been diagnosed by a doctor with asthma or atopic
dermatitis”. 3) Subjects with allergic asthma and atopic dermatitis:
All those who answered “yes” to the question “Have you ever been
diagnosed by a doctor with asthma and atopic dermatitis at the same
time?”. The questionnaires were completed under the supervision of
health professionals.
The collected whole blood samples (9 mL) were immediately centrifuged,
and the buffy coat fraction (i.e., white blood cell fraction) was
separated by Ficoll-Paque to isolate peripheral blood mononuclear cells
(PBMC). The PBMC fraction was suspended in RNAprotect Cell Reagent and
frozen (-80 °C) in 300 μL aliquots containing ~13
million cells until use for analysis. RNA was then extracted from PBMC
in RNAprotect Cell Reagent with a Zymo Research: Quick-RNA Whole Blood
(R1201) extraction kit according to manufacturer´s instructions. Quality
parameters like concentration, purity (Nanodrop, Thermo Fisher
Scientific) and integrity (Agilent 5200 Fragment Analyzer system) of
extracted RNA were determined. For the library preparation and
sequencing, 1 μg of high-quality RNA per sample was used. The mean RNA
Integrity Number (RIN) for samples was 9.0 (min − max: 7.3 – 10.0).
Library preparation and Sequencing
Genome-wide analysis of gene expression was conducted using the Next
Generation Sequencing (NGS) platform with the QuantSeq library
preparation step. cDNA libraries for each sample (RNA) were generated
from 1 μg of total RNA using the Quantseq 3′ mRNA-Seq Library Prep kit
for Illumina (Lexogen) following the manufacturer’s instructions.
QuantSeq generates strand-specific NGS libraries close to the 3′ end of
poly-A RNA15. Standard external barcodes were ligated
to allow for multiplex sequencing. After PCR amplification, the
libraries were size-selected with Agencourt AMPure XP magnetic beads
(Beckman Coulter). Libraries were quantified by Qubit (Life
Technologies), and their size (∼250 bp) was determined using an Agilent
2100 Bioanalyzer. Libraries were sequenced (Illumina NovaSeq platform)
and quality checked (110 bp single read) to obtain min. 20–25 million
reads per sample. Following this, NGS data were demultiplexed.
Data preprocessing
After sample demultiplexing, UMI tags (6 bp) were removed from all reads
using umi_tools (1.0.0). Afterwards, TruSeq adapters and low quality
3’-ends of reads (treshold: 13) were trimmed using bbmap (38.42). Short
reads (length < 30 bp) and reads with low quality (treshold:
24) were filtered out. Samples were mapped against reference human
genome (Genome Reference Consortium Human Build 38) using STAR (2.7.7a)
and deduplicated using umi_tools (1.0.0). Transcript features were
summarized using htseq-count (0.11.1). During data pre-processing,
quality of samples was continuously controlled using fastqc (0.11.5),
qualimap (11_12-16) and multiqc (1.8, 1.12).
Data processing and statistical analysis were performed using R
programming software (version 4.2.2)16. Only samples
with allergies to food, insect, pollen, dust, or mite were included in
the analysis; and participants without any other immune-related diseases
(i.e., celiac disease, psoriasis, or lupus) were assigned as a control
group. Samples with any missing data were filtered out. Genes with at
least 5 CPM (counts per million) in at least 20% of samples were kept
in the data set. The data was normalized using TMM (trimmed mean of M
values) normalization and transformed to continuous log2 scale using
limma-voom17, 18. Batch effects were accounted for
using principal component analysis (PCA) plots and by correlation of
principal components with potential confounders. Surrogate variable
analysis was performed to the data and the first 10 surrogate variables
were used to adjust the unknown cell blood
composition19, 20, 21.
Statistical analysis
Differentially expressed genes were identified using limma lmFit model
and p-values were corrected for multiple testing using
Benjamini-Hochberg false discovery rate (FDR)22. The
model was adjusted for biological, socio-economic, and technical
covariates (sex, age, BMI, education, smoking status, alcohol
consumption and library preparation batch). Genes were annotated using
GeneCard22, DAVID and GO databases23, 24, 25.
Functional analysis, namely Gene Set Enrichment Analysis (GSEA) and
Sub-Network Enrichment Analysis (SNEA) was performed using Pathway
Studio (version 12.0)26.
Results
Study population
The final dataset included 218 participants from CELSPAC: YA cohort
study between 20–37 years old with an average BMI of 23.41. Overall,
114 participants had at least one allergic manifestation and 104
participants did not suffer from any of allergic triggers or symptoms.
The detailed description of the population is presented in Table 1.
Table 1 : Description of study population participants (subset
derived from CELSPAC: YA cohort).