Although respiratory infections are largely preventable causes of illness and death, they remain the leading cause of death from infectious diseases worldwide, ranking fifth among all total causes of death (GBD 2015 LRI Collaborators, 2017). Influenza A virus (IAV), measles virus, rubella virus, and mumps virus infect a wide range of hosts, are highly transmissible and are prone to a mutation that can cause pandemics in a short period of time, placing an enormous burden and pressure on public health systems (Marshall and Plotkin, 2019; Harrington et al., 2021; Iacobucci, 2022; Raghunathan and Orenstein, 2022). The incidence of IAV, measles, mumps, and rubella has decreased drastically following the implementation of vaccination programs. However, the segmented genome of IAV allows for easy recombination and antigenic shift, resulting in novel antigens (Bouvier and Palese, 2008). Thus, the ability of IAV to rapidly evolve can result in highly pathogenic viral strains. Measles and rubella remain endemic in many countries, leading to the importation of cases and occasional local transmission within China, and the resurgence of mumps has recently been reported, with outbreaks still occurring and challenges remaining while controlling these diseases (Kauffmann et al., 2021). There is considerable variation in the severity of respiratory infections caused by viral infections. There are major determinants of this variability, such as intrinsic virus pathogenicity, acquired host factors (e.g., immunity and coexisting conditions), and innate host susceptibility. While viral genetic determinants of respiratory disease severity and host immunity have been well studied, host genetic determinants have been much less explored (Kennedy et al., 2014; Voigt et al., 2018; Sabikunnahar et al., 2022).

Genome-wide association studies (GWAS) are effective methods for understanding the genetic basis of many complex traits in common human diseases (Hayes, 2013). In particular, these have proven to be well-suited for identifying common single-nucleotide polymorphism (SNP) variants with moderate to large effects on phenotypes (Gorlova et al., 2022). However, the specific biological mechanisms and functional consequences of many of the genetic variants identified by the GWAS remain unclear, especially their role in disease severity; that is, GWAS methods may miss small effect–trait associations. Gene expression is an intermediate phenotype between genetic variation and an underlying disease predisposition trait (Albert and Kruglyak, 2015). Many genetic variants affect complex traits by regulating gene expressions. Unfortunately, large-scale expression–trait associations are hampered by sample availability and cost, as well as intrinsic factors and small effects. Therefore, to address these issues, transcriptome-wide association studies (TWASs) were developed, which integrate gene expression into large-scale GWAS (Derks and Gamazon, 2020). Through extensive simulation of existing GWAS data, TWASs have identified candidate genes associated with mental disorders (Thériault et al., 2018), calcified aortic stenosis (Liu et al., 2020), pancreatic cancer (Lin et al., 2017), and inflammatory biological age (Zhang et al., 2022). However, for many complex traits, biologically relevant tissues are unknown. Most existing studies identify gene–trait associations on the basis of a single tissue, whereby the significant gene effects that are identified get inflated. In addition, it has been shown that eQTLs with larger effects tend to regulate gene expressions in multiple tissues. The TWAS analysis of multiple tissues improves the accuracy of the results.

There have been some studies exploring the association of respiratory infectious diseases with some human tissues. Pulmonary inflammation and airway epithelial injury are hallmarks of human pulmonary infectious diseases. The association of HLA alleles with respiratory infectious diseases in lung cells has been confirmed by Zhang et al. (2022). Moreover, Xiao et al. (2023) found that interleukin-11 (IL-11) as a fibrotic factor may be associated with respiratory diseases leading to respiratory failure or even death. These potential connections deserve our attention.

In this study, we used TWASs to explore genes associated with IAV, measles, rubella, and mumps. Given that respiratory infectious diseases are associated with multiple tissues, we used eQTL reference panels for three tissues, namely, whole blood, lung, and transformed fibroblast cells, and pooled the GWAS data for IAV, measles, rubella, and mumps to identify tissue-specific susceptibility genes. This study lays the foundation for further understanding the pathogenesis of these four viruses.

In this study, we used tissue-specific TWAS to explore the association of genetically predicted gene expression with IAV, measles, rubella, and mumps and identified 19, 14, 15, and 13 genes in three tissues as potential IAV-, measles-, mumps-, and rubella- related genes, respectively. In IAV, including three in whole blood (ULK4, AC010132.11, and SURF1), four in the lung tissue (NIPAL2, TRAP1, TAF1C, and AC000078.5), nine in the transformed fibroblast cells (RP4-639F20.1, RMDN2, ATP1B3, SRSF12, RP11-477D19.2, TFB1M, XXyac-YX65C7_A.2, TAF1C, and PCGF2), and one expressed in three tissues as significantly associated with IAV (BNIP1). In measles, including six in whole blood (SOAT1, COLGALT2, AC021860.1, HCG11, METTL21B, and MRPL10), six in the lung tissue (GSTM4, PAQR6, RP11-617D20.1, SNX8, METTL21B, and ANKRD27), and two in the transformed fibroblast cells (CBWD2, and TSFM). In mumps, including six in whole blood (MTOR, LAMC1, TRIM38, U91328.21, POLR2J, and SCRN2), six in the lung tissue (Smpd4, UBN1, CNTROB, SCRN2, HOXB-AS1, and SLC14A1), and three in the transformed fibroblast cells (AC007566.10, AC093668.2, and CPD). In rubella, including five in whole blood (JAGN1, RRP12, RP11-452K12.7, CASP7, and AP3S2), four in the lung tissue (IL17RC, FAM86HP, AMACR, and RRP12), and four in the transformed fibroblast cells (PPP2R1B, C11orf1, DLAT, and TMEM117). The above 59 genes have never been clearly reported to be associated with IAV, measles, mumps, and rubella.

TNF receptor-associated protein-1 (TRAP1), a member of the mitochondria-specific Hsp90 family, is located in the mitochondrial matrix, mitochondrial endometrium, and the intermembrane space (Cechetto and Gupta, 2000; Pridgeon et al., 2007). However, among the few real TRAP1 clients described, two are subunits of the electron transport chain complex (ETC), a component of the complex II succinate dehydrogenase subunit A/B (SDHA/B) (Sanchez-Martin et al., 2020; Sanchez-Martin et al., 2021), and complex IV cytochromes c oxidase subunit 2 (COXII) (Xiang et al., 2016; Xiang et al., 2019). The II/SDH complex is a protein complex containing iron and sulfur groups whose function is to transfer electrons from succinate to the coenzyme Q10-ubiquinone (III complex) (Bezawork-Geleta et al., 2017). TRAP1 leaves the SDH in a partially unfolded state. Inhibition of TRAP1 can release active SDH and increase its activity (Agarwal et al., 2019; Hu et al., 2020). In addition, SDH activity (Masgras et al., 2017; Hu et al., 2020) and oxygen consumption (Masgras et al., 2017) were negatively correlated with TRAP1 expression, suggesting that TRAP1 promotes the Warburg effect (Guzzo et al., 2014). It should be noted that SDH also oxidizes succinic acid to fumarate, thus integrating the TCA cycle, suggesting a broad impact of TRAP1 on mitochondrial metabolism (Guzzo et al., 2014; Hsu et al., 2016). It can be assumed that the abnormal expression of the TRAP1 gene may contribute to the development of IAV by affecting mitochondrial metabolic processes.

GSTM4 belongs to the mu class of glutathione S-transferase (GST). Compared to other GSTMs (Comstock et al., 1993), GSTM4 exhibits greater consistency in amino acid sequence but has significant differences in physicochemical properties and tissue distribution. GSTM4 does not exhibit activity comparable to standard TPS substrates, and its specific substrate has not been identified. It is well known that the GST gene is highly polymorphic to adapt to the growing number of exotic compounds (Hayes and Strange, 2000). This polymorphism can alter individual susceptibility to the disease and response to therapeutic agents. For example, the T2517C polymorphism in GSTM4 has been shown to be associated with an increased risk of lung cancer (Liloglou et al., 2002). The mechanism underlying this association is not yet clear. Currently, there are few reports of the relationship between the GSTM4 gene and the disease. Efforts may be needed to verify its role in infectious respiratory diseases.

In addition to the aforementioned genes, we have identified other candidate genes (SOAT1, HCG11, PAQR6, and AP3S2) in different tissues, of which AP3S2 has been reported to be associated with the development of type 2 diabetes (Kazakova et al., 2017). Other genes have also been implicated in cancer. For example, SOAT1 has been linked to the development of stomach cancer (Zhu et al., 2021), and pancreatic cancer (Oni et al., 2020); HCG11 may influence the development of nasopharyngeal carcinoma (Zheng et al., 2022); the PAQR6 gene has also been associated with several cancers (Yang et al., 2021). Although there is little research on these genes and respiratory infectious diseases, the potential links between these genes deserve further attention in future studies.

In summary, our results list some genes for respiratory infections that have not been studied before. Based on the previous knowledge of these genes, this means that they may be involved in host infection through transcriptional processes of susceptible genes and RNA degradation processes. This study is helpful to deepen the understanding of the pathogenesis of respiratory infectious diseases.