The Relationship Between Gene Expression Pattern and the Severity of Asthma


 Background Because of heterogeneity, complexity of diagnosis and diversity of pathogenesis, the incidence and mortality of asthma are increased seriously. A structured and specific approach to assess and treat asthma may help clinicians. Results A total of 838 common genes were found from quietness to exacerbation then to recovery of asthma. PPI network analysis identified 7 modules, and found 7 hub genes with high degree. Then, we verified the expression of hub genes in patients by quantitative real-time polymerase chain reaction (qRT-PCR). Enrichment analysis and gene set enrichment analysis (GSEA) showed that exacerbation related genes were significantly related to immune and inflammatory response. Transcriptional regulators factor STAT1 had a significant regulatory effect on exacerbation related genes. Conclusions These results indicated that seven hub genes were potential biomarkers and targets of asthma exacerbation. they were involved in the development of asthma through immune inflammatory signaling pathway. The results of this study not only provide a new research direction and theoretical basis for the exacerbation mechanism of asthma, but also provide a new target for clinical treatment.


Introduction
Asthma is a chronic in ammatory airway disease was characterized by heterogeneity, but there is no clear pathophysiological mechanism (Perlikos, et al. 2016). It has variable airway obstruction and bronchial hyperresponsiveness (Carr and Kraft 2017). Clinically, asthma patients will repeatedly appear wheezing, coughing, chest tightness and shortness of breath (Mims 2015). Due to the pathological unclear mechanism, the incidence and mortality of asthma are increasing in many parts of the world, making it a global health problem (Lemanske and Busse 1997). It is estimated that asthma will affect 334 million people in the world ). The prevalence of asthma varies widely from 0. 2-21.0% in adults and from 2.8-37.6% in children aged 6 to 7 years(Aaron, et al. 2018 ;Papi, et al. 2018).
Asthma is the result of complex gene environment interaction (Papi, et al. 2018). Therefore, the disease mechanism is complicated and diverse. Theoretically, many cytokines and chemokines may be related to the endogenous "neutrophil rich" and "eosinophil rich" of asthma (Wenzel 2012). More and more evidence showed that there were two broad in ammatory phenotypes in the airway of asthmatic patients: eosinophilic in ammation and other neutrophil dominated in ammation (Gibson, et al. 2001). Recent studies had highlighted the increased response of Th1 / IFN-γ and Th17 / IL-17 in asthma, which was often associated with neutrophil in ltration in the airway (Raundhal, et al. 2015;Ray, et al. 2016).
In the investigation, extensive clinical evaluation and testing were often impractical (Kirenga, et al. 2019). Accurate diagnosis of asthma is important because it will contribute to enhancing the incidence and mortality of this disease. Despite drug therapy recommended by national and international asthma guidelines, asthma control is often poor (Chung, et al. 2014; National Asthma and Prevention 2007). The evaluation and diagnosis of asthma should rst have a comprehensive history and physical examination (Winders, et al. 2019). Other parts include lung function, medical level, eosinophil count and allergy test (Boulet, et al. 2019;Bousquet, et al. 2010;Singh, et al. 2019).
Asthma is sometimes di cult to treat, resulting in a disproportionate cost to the medical system (Jones, et al. 2018). The objectives of asthma treatment include prevention of recurrent exacerbations (Fuhlbrigge, et al. 2012). However, gave the heterogeneity of the disease, a "one size ts all" treatment strategy seems no longer suitable for effectively pursuing these goals. At present, a variety of biomarkers can identify the characteristics of high endogenous in ammation in adults and children with asthma, and targeted treatment research has been carried out in (Galeone, et al. 2018;Uwaezuoke, et al. 2018). IL-17 pathway may be the target of severe asthma disease control (Al-Ramli, et al. 2009;Suzuki, et al. 2016). Therefore, it is extremely important to determine the corresponding biomarkers according to the exacerbation and recovery of asthma. In this study, we will identify biomarkers and potential molecular mechanisms of asthma exacerbation based on the expression patterns of genes during exacerbation and recovery of asthma patients.

Results
Systemic molecules in the process of asthma attack and recovery To identify which factors were involved in the process of asthmatic patients, we analyzed the transcriptome data of GSE19301 data. Compared with the quiet group, Toll like receptor and NOD like receptor signaling pathway in the exacerbation group were activated (Fig. 1A). Compared with the followup group, NOD like receptor signaling pathway and apoptosis in the exacerbation group were activated ( Fig. 1B). The results showed that there were the same regulatory mechanism and different molecular changes in the process of asthma from quiet to exacerbation to follow-up. Therefore, we compared the gene expression of three groups. There were 2194 differentially expressed genes (DEGs) between exacerbation and quiet groups (Fig. 1C). There were 1376 DEGs between exacerbation and follow-up groups (Fig. 1D). These results indicated that some genes and signaling pathways played crucial roles in the process of asthma from tranquility to exacerbation and then to tranquility.

Identifying asthma-associated PPI modules from DEGs
To further explore the interactions of differentially expressed genes in the process of exacerbation to recovery, we mapped the common genes in the two groups of DEGs into PPI network. 838 common genes were mapped into PPI network. Next, the module searching procedure identi ed 7 modules ( Figure S1).
The distribution of the genes with the highest or lowest expression level of exacerbation group in the module compared with quiet or follow-up was shown by heatmap ( Fig. 2A). Up regulated genes of exacerbation and quiet were all in M2, and the down regulated genes were all in M5, respectively (Fig. 2B).
Similarly, the up regulated genes of exacerbation and follow-up were all in M2, and the down regulated genes were all in M5 (Fig. 2C). Therefore, M2 genes may promote asthma, while M5 genes may inhibit asthma.
On the other hand, hub gene of each module by the degree in the module was screened (Table 2).
Depending on the expression of hub genes in three groups of asthma, ENO1, OAS3 and TLR4 were the highest expressed in exacerbation, while HSP90AB1, HSPA9, PTCD3 and UBA52 were the lowest expressed in exacerbation (Fig. 2D). Therefore, they are considered to be potential genes for positive or reverse regulation of asthma exacerbation. Importantly, we veri ed the expression of these key genes during the exacerbation of asthma by qPCR (Fig. 2E). In addition, the results of correlation analysis also proved that there was a signi cant negative correlation between high or low expression genes in exacerbation (Fig. 2F). These results suggested that key genes can regulate the exacerbation of asthma, which may be used as a biomarker of asthma exacerbation for further study. Biological function and signal pathway involved in the process of asthma To identify the biological functions and signaling pathways involved in the progression of asthma, module genes were enriched and analyzed. A total of 1741 BP, 113 CC, 232 MF and 65 KEGG signaling pathways were enriched by the module genes. Among them, genes mainly enriched in activation of the innate immune response and other immune in ammatory functions (Fig. 3A, B, C). Also enriched in Toll like receptor signaling pathway, regulation of cellular amine metallic process signal pathways (Fig. 3D).
In the up-regulated or down-regulated pathway, we found that NOD − like receptor signaling pathway was activated in exacerbation group, but decreased in quiet and follow-up (Fig. 3E, F). ENO1, OAS3 and TLR4 were mainly enriched in coenzyme metallic process, nucleotidyltransferase activity, regulation of in ammatory response and others. HSP90AB1, HSPA9, PTCD3 and UBA52 were mainly involved in response to interleukin-4, response to interleukin-12, positive regulation of NF-kappaB transcription factor activity and others. In addition, the functional similarity among PTCD3, OAS3 and ENO1 is high (Fig. 3G).

Transcription factors (TF) regulating dysfunctional molecules
Transcription regulation played an important role in the progress of diseases. Since we had observed that the module genes were associated with the exacerbation and recovery of asthma. We obtained 76 transcription factor regulatory module genes (Fig. 4A). The results of correlation analysis identi ed the transcription factors STAT1 was highly correlated with the module genes, including XAF1, PSMB9, IFIT3 and IRF7 (Fig. 4B). Therefore, we believed that STAT1 was an important regulatory role in the pathogenesis of asthma.

Discussion
The pathogenesis of asthma is not completely clear (Daya, et al. 2019). At present, the diagnosis and treatment of asthma is also challenging. Therefore, this study is to explore the gene expression patterns in the process of asthma. In order to identify biomarkers that can diagnose the exacerbation of asthma and the mechanism of exacerbation with potential therapeutic value.
The differentially expressed genes were thought to represent the development of asthma (Pascoe, et al. 2017). Interestingly, we found that the top six genes up regulated from quiet to exacerbation were the top six genes down regulated in follow up group. Among them, the expression of SIGLEC-1 in lung tissue was speci c to alveolar macrophages, and the single nucleotide polymorphisms in SIGLEC-1 was related to the severity of asthma (Souza de Lima, et al. 2017). CCL2 expression was enhanced in allergic asthma induced by OVA sensitized mice (Jiang, et al. 2019). The expression of genes in asthma patients with exacerbation ran counter to that in patients with recovery, they may be indication of exacerbation, as well as potential therapeutic target.
The genes in each module may represent a mechanism of asthma. Enrichment analysis showed that genes in M2 were mainly related to activation of innate immune response, Toll like receptor signaling pathway. Asthma was often considered as the result of abnormal activation of Th2 to environmentally sound allergens (Thiriou, et al. 2017). Toll like receptor signaling pathway was activated in patients in the exacerbation group. In fact, the complex interaction between environmental allergens and TLRs results in In the hub genes, TLR4 was reported to be signi cantly associated with the occurrence and development of asthma (Hwang, et al. 2019). Heat shock protein (HSP) played a major role in promoting in ammation and anti-in ammatory response (Hallenbeck and Kaplan 1987). Heat shock factor 1 (HSF 1) could attenuate airway hyperresponsiveness and airway in ammation in asthmatic mice (Wang, et al. 2017). In general, the highly expressed hub gene was mainly enriched in the in ammatory response. There was an inseparable relationship between in ammation and asthma (Denlinger, et  In this study, we also predicted the regulator of the module gene of asthma. STAT1, a transcription factor with a high correlation with module genes. Studies had shown that STAT1 was associated with asthma exacerbation (Gomez, et al. 2018;Su, et al. 2016). We believed that STAT1 could be involved in the exacerbation and recovery of asthma by regulating target genes.
The pathophysiology of asthma was complicated. There was no single cytokine related to the whole pathogenesis of asthma. Despite the best guidelines for treatment, and regardless of the severity of the underlying disease, asthma patients will experience exacerbation, which is caused by aggravation of the existing in ammatory process and the loss of disease control (Castillo, et al. 2017). Our results showed that many in ammatory cytokines were involved in the innate and adaptive immunity of asthma. In particular, hub genes we identi ed may serve as a biomarker of asthma exacerbation and a potential target for treatment.

Conclusions
PPI network was used to analyze the key genes of asthma exacerbation, and 7 hub genes were identi ed to be related to asthma exacerbation. Molecular mechanism analysis showed that these genes participate in the development of asthma through the signal pathway of immune in ammation.
Transcription regulators predicted that STAT1 had a signi cant regulatory effect on exacerbation related genes.

Datasets
We collected the dataset GSE19301 from GEO database. which includes quiet (394 samples), exacerbation (166 samples) and follow-up (125 samples) (Bjornsdottir, et al. 2011). Series matrix les and clinical information tables were downloaded from the GEO website.

Identi cation of differentially expressed genes
The "sva" R package was used to conduct batch normalization of the expression data from the three different datasets. Then, the normalized expression matrix of quiet, exacerbation and follow-up differences were analyzed by "limma" R package respectively. The differentially expressed genes were obtained by setting threshold p < 0.05.

Constructing protein protein interaction (PPI) network and identify modules
Download the protein interaction data from the protein interaction database (String). Interaction data containing only differential genes was screened, and the score > 500 was set. For the analysis of the recognition module, we rst import the selected protein interaction into the software of Cytoscape. We used the ClusterONE plug-in for module mining. Then, network analysis of gene subnets in the module was carried out, and the degree of nodes was obtained. The higher the degree of a node, the more important it is, the more likely it is to represent the characteristics of the module.

Patients
Whole blood samples were collected from patients with asthma, including 3 quiet, 3 exacerbation and 3 follow-up, respectively. The human study protocol, consent form, and consent procedure were approved by the medical ethic committee of Xinjiang Medical University. All patients were informed and gave their written consent to have their blood for research purpose. This protocol complied with the Code of Ethics of the World Medical Association (Declaration of Helsinki) and in agreement with ethical guidelines set by China law.

Quantitative real-time polymerase chain reaction
To isolate RNA, 200 µL human blood samples were collected from patients and mix well with 1 ml of Trizol reagent (Invitrogen, USA), according to the manufacturer's instructions. After quanti cation, 0.5 µg of RNA was reversed transcribed with SuperScriptIII (Invitrogen, USA). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed on Bio-Rad CFX96 with SYBR Master Mix (Invitrogen, USA). All samples were normalized to the signal generated from GAPDH. Sequences of primers used in this study are listed in Table 1. The experiments were repeated at least 3 times, and results were analyzed with analysis of variance. Data are shown as fold change (2 − ΔΔCt). Statistical analysis was performed using SPSS software and student's t test. P < 0.05 was considered statistically signi cant. Enrichment and gene-set enrichment analysis (GSEA) To clarify the possible biological roles of these genes in PPI networks, the cluster Pro ler R package was used to create KEGG enrichment paths, results and plots. P < 0.05 indicated that KEGG pathway had signi cant difference.
We performed gene set enrichment analysis using the application available at the Broad Institute Gene Set Enrichment Analysis website. We tested the KEGG pathways gene sets. Activation or inhibition of KEGG was evaluated by Gene Set Variation Analysis (GSVA) algorithm.

Prediction of regulatory transcription factors for modular genes
To explore the pivotal regulators of module genes in the exacerbation and recovery of asthma, all human transcription factor target data was used in TRRUST V2 database as a background set for pivot analysis (Daya, et

Consent for publication
All authors agree to publish. Figure 1 Gene expression and signal transduction from an asthma attack to recovery. A. KEGG signaling pathway activated during asthma exacerbation. B. KEGG signaling pathway inhibited during the course of asthma exacerbation to recovery. C. Compared with patients with asthma at quiet, the differentially expressed genes in the patients with exacerbation. D. Compared with the follow up of asthmatic patients, the differential expression of genes in patients with exacerbation. Red node represents up regulated gene, blue node represents down regulated gene.

Figure 2
Key genes recognition of maladjusted molecules in the process of asthma exacerbation and recovery. A. The expression of the intersection genes for two groups in modules was demonstrated by thermogram.
The red node represents the genes up regulated in two groups, and the blue node represents the genes down regulated in two groups. In the module, the up regulation or down-regulation genes of exacerbation compared with quiet (B) and the up regulation or down regulation genes of exacerbation compared with follow-up (C). D. The expression of each module's hub gene in quiet, exacerbation and follow up groups in dataset. E. The expression of hub genes in blood of quiet, exacerbation and follow up patients. *P <0.05, **P<0.01 compared to exacerbation. F. The correlation between module hub genes and between genes and phenotypes.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.