Bronchial epithelial gene expression does not differ with rhinovirus infection

Key takeaways:

• Most treatments target type 2 asthma even though 30% of children have non-type 2 asthma.
• Children with non-type 2 asthma had lower FeNO and mean peripheral eosinophil counts.

SAN DIEGO — Bronchial epithelial gene expression did not differ between type 2 and non-type 2 pediatric asthma before or after rhinovirus infection, according to a presentation at the American Thoracic Society International Conference.

Also, there was no correlation between clinical type 2 phenotypes and epithelial expression of periostin and SERPINB2 in an ex vivo epithelial system, Patricia dela Cruz, MD, PhD, pediatric resident, Debley Lab, Center for Respiratory Biology & Therapeutics, Seattle Children’s Research Institute, University of Washington, said during her presentation.

“The understanding of asthma, of course, has shifted from this singular disease to one with multiple phenotypes and molecular endotypes,” dela Cruz said.

Clinical type 2 asthma is characterized by elevated FeNO and peripheral eosinophilia, she said. Also, brushings from adults with type 2 asthma show increases in the three-gene mean of CLCA1, periostin and SERPIN2B.

“Interestingly, despite the fact that almost 30% of children and 50% of adults have non-type 2 asthma, most of the available asthma pharmaceutics target the more common type 2 endotypes, leading to this therapeutic gap,” dela Cruz said.

To elucidate the mechanisms behind underlying type 2 asthma and guide personalized therapies, dela Cruz said that she and her colleagues infected primary bronchial epithelial cells from 37 children with type 2 and non-type 2 asthma in an organotypic system with rhinovirus — “which is just a very common trigger of viral asthma exacerbation,” dela Cruz said — to compare their responses.

Cells were cultured in an air-liquid interface for 21 days, with RNA harvesting at 2, 4, 7 and 10 days after infection. The researchers analyzed epithelial gene expression via bulk transcriptomics.

As expected, dela Cruz said, the children with type 2 asthma (n = 16; 31% girls; mean age, 12.6 years) had elevated FeNO (mean, 27.3 ppb) and peripheral eosinophils (mean, 481 cells/µL) compared with the children with non-type 2 asthma (n = 21; 48% girls; mean age, 11 years), who had a mean FeNO of 10.3 ppb and a mean peripheral eosinophil count of 133 cells/µL.

The type 2 group also had higher serum IgE (368 IU/mL vs. 139 IU/mL) and a higher proportion of patients with two or more aeroallergen sensitivities confirmed by allergen-specific IgE (69% vs. 38%), which dela Cruz said also was expected.

“When we looked, dichotomizing by these clinical definitions of T2, we don’t actually find any significant differentially expressed genes, both pre-infection and post-viral infection,” dela Cruz said. “So, with this, we shifted more to a molecular characterization.”

The researchers then analyzed CLCA1, periostin and SERPINB2 before and during infection.

“We had very low to no expression of CLCA1, which we anticipated was because of the purely epithelial model without the immune cells and subsequent IL-13 stimulation,” dela Cruz said.

Surprisingly, she continued, periostin and SERPINB2 did not correlate with the clinical parameters for type 2 asthma, and their expression based on clinical parameters was essentially identical in the type 2 and non-type 2 groups.

“Since our clinical parameters didn’t actually result in differentially expressed genes, we began stratifying using a ‘two-gene-mean,’ rather than a ‘three-gene-mean,’ leaving out CLCA1, because of the low expression within our model,” dela Cruz said.

The researchers split the cohort into a group with high periostin and SERPINB2 (n = 9) and a group with low periostin and SERPINB2 (n = 10). The 18 children with intermediate levels of expression of these genes were excluded for a cleaner picture, dela Cruz said.

“When we do this using this molecular definition, we see a very large amount of genes expressed differentially at baseline,” dela Cruz said.

Groups included 367 genes elevated in molecular non-type 2 asthma and 729 genes elevated in molecular type 2 asthma. Specifically, 80 gene modules showed differential expression and kinetics in rhinovirus-infected epithelium from children with non-type 2 asthma.

HLA-DPA1 and HLA-DRB1, which dela Cruz said have been implicated in increased asthma risk, particularly in the setting of specific environmental allergens, saw increased expression among children with non-type 2 asthma.

Non-type 2 asthma also was associated with higher expression of CD74, which dela Cruz called a chaperone for these class 2 molecules, indicating that the epithelium may consist of non-professional antigen presenting cells.

Further, the non-type 2 asthma group had increased expression of heat shock protein A5, which dela Cruz said is involved in endoplasmic reticulum (ER) stress and unfolded protein response.

“In the literature, adults with neutrophilic and eosinophilic asthma have dysregulated unfolded protein responses in the setting of ER stress,” dela Cruz said.

DNMT1, which plays a role in DNA methylation during the epithelial-mesenchymal transition and could allude to changes in airway epithelial remodeling in the context of asthma, also saw increased expression in non-type 2 asthma, dela Cruz continued.

Overall, dela Cruz said that this comparison of pediatric donors with clinically defined type 2 and non-type 2 asthma showed no differences in bronchial epithelial gene expression before or after infection.

Next, clinical type 2 asthma phenotypes did not correlate with the epithelial expression of periostin and SERPINB2 in the study’s ex vivo epithelial system.

“And three, rhinovirus-infected bronchial epithelium from these pediatric donors with molecular non-T2 asthma demonstrated increased expression of a novel 80-gene module,” dela Cruz said, “notable for heat shock proteins, HLA-associated genes and genes associated with DNA methylation.”

Looking ahead, dela Cruz said that she and her colleagues will assess the epigenetic state of T cells via an allergy-specific DNA methylation assay. The researchers also will look at genetic variants to see how they affect the results of this model and correlate with clinical outcomes later, dela Cruz continued.