Chronic obstructive pulmonary disease (COPD) is the most common pulmonary disease characterized by small airway lesions and chronic respiratory symptoms and continuous airflow restriction [1]. Its common pathological features include bronchiolitis and destruction of lung parenchyma (emphysema) [1]. COPD usually progresses slowly, often leading to the neglect of early treatment and prevention [2]. Thus, when the disease is brought to the patients’ attention, it has usually progressed to the third or fourth stage and may be accompanied by complications [3]. It is estimated that 4.5 million deaths yearly will be attributed to COPD by 2030 [4], and it is on track to become the third leading cause of death worldwide [5]. Due to the prolonged course and the high prevalence of the disease, COPD will also place a heavy burden on the global healthcare economy.

COPD is considered to be a process of chronic hypoxia and chronic systemic inflammation with complex pathogenesis. It is reported that environmental and genetic factors affect the risk of COPD jointly [6, 7]. It is well known that smoking is a recognized risk factor for COPD, still, this proportion in long-term smokers is only 10%–20%, and there is a considerable proportion of non-smoker people diagnosed with COPD, which indicates the importance of genetic factors in COPD [3]. With the development of genome-wide association analysis (GWAS), Pillai and colleagues first found that the single nucleotide polymorphisms (SNPs) of CHRNA3/CHRNA5/IREB2 located on chromosome 15q25 were associated with COPD significantly [8]. Since then, GWAS at home and abroad have reported lots of genes influencing susceptibility to COPD successively [9, 10, 11]. However, there is still largely unknown in the genetic etiology of COPD.

Erythropoietin (EPO) stimulated by Hypoxia is a recognized hematopoietic cytokine that can control the oxygen-carrying capacity of blood and enhance ventilatory function under hypoxic conditions [12, 13, 14]. EPO was observed to play anti-inflammatory, anti-apoptosis, and anti-oxidant/fibrosis roles in the lung by interacting with inflammation, apoptosis, oxidative stress, or fibrosis-related pathways [15, 16, 17, 18, 19, 20]. Some studies indicated that the expression of EPO was correlated with the severity of COPD positively, while correlated with the pulmonary function index FEV${}_1$ negatively [21, 22, 23]. In addition, it has been pointed out that an increase in EPO levels can be observed in exacerbations of COPD [24, 25]. Based on the evidence above, we hypothesized that EPO may be involved in the occurrence and development of COPD.

Therefore, we conducted a case-control study enrolled 1095 COPD patients and 1144 healthy controls in several hospitals in Guangdong to evaluate the association between EPO polymorphisms (rs1617640 A$>$C, rs507392 A$>$G, rs564449 G$>$T) and COPD susceptibility. 872 participants were recruited from a prevalence study performed in Gansu Province to verify the effect of EPO polymorphisms on lung function. Using the method developed by Gail and Bruzzi, we fitted a 10-year absolute risk model for southern Chinese.

To reveal the role of EPO polymorphisms on COPD, we investigated the relationship between three SNPs of EPO (rs1617640 A$>$C, rs507392 A$>$G, and rs564449 G$>$T) and the risk of COPD in southern Chinese in this study. We found that the C allele of EPO rs1617640 A$>$C was associated with a reduced risk of COPD and an increase in FEV${}_1$ and FVC in Chinese significantly. Meanwhile, compared with EPO rs1617640 AA genotype, AC and CC genotypes retained protective effects in the strata of non-smokers, avoiding coal and biofuels (p 0.05).

EPO, a pleiotropic factor, can affect the occurrence and progression of many diseases [34, 35]. EPO was reported that expressed in the lung and responded to Hypoxia by promoting erythropoiesis and neovascularization [13, 36, 37, 38]. In addition, in acute lung injury, EPO could activate the proliferation of pulmonary endothelial cells and reduce the infiltration of inflammatory cells, which plays a key role in controlling pulmonary ventilation. Some studies have shown that the inflammatory response and hypoxemia caused by COPD change the expression of EPO. For example, COPD patients have up-regulated EPO expression due to increased inflammatory factors and decreased erythrocytes in the body [39, 40]. Thus, an increase in hemoglobin concentration by upregulation of EPO’s expression can correct hypoxemia seen in COPD patients commonly [22, 41, 42]. Existing researches indicate that EPO may participate in the development of COPD. However, no study has explored whether EPO polymorphisms modify the susceptibility to COPD.

We finally retained three SNPs (rs1617640 A$>$C, rs507392 A$>$G, and rs564449 G$>$T) to conduct the research. Online web tools (NCBI, Ensembl and SNPinfo Web Server) were used to find the location and functional information of SNPs. Rs1617640 A$>$C is a transcription factor located in upstream of the EPO transcription start site (Chr 7:100719675), as well as impact the potential transcription factors such as AP1\STAT5\STAT3\GATA-2 and so on. In addition, it has been shown to be strongly association with a variety of diseases including complications of diabetic, hepatitis C, myelodysplasia, erythropoietic and so on in the present studies [43, 44, 45, 46]. Rs507392 A$>$G, located at Chr 7:100722313, is an intron variant of the EPO, considering the current studies reported it associated with diabetic retinopathy [47, 48], so we decided to keep the SNP for further analysis after consulting the relevant experts. Rs564449 G$>$T is a miRNA binding site located in 3′-UTR of the EPO (Chr 7:100723515) and no research has reported that it is associated with any disease.

In this study, we first discovered that the variation of EPO rs1617640 A$>$C was associated with COPD risk, and compared with AA genotype, patients with AC/CC genotypes had a lower incidence of COPD. Unfortunately, no association was found between the other two SNPs (rs507392 A$>$G, and rs564449 G$>$T) and the risk of COPD. Rs1617640 A$>$C is located in the upstream promoter region of EPO. The promoter SNP can cause abnormal gene expression by affecting the binding of transcription factors to templates [49]. In the present studies, EPO rs1617640 polymorphism was shown to be associated with erythropoietin expression and level [43], CC genotype can promote the expression of erythropoietin [45], which leads to the increasing of Hb level, hematocrit and erythrocyte count [50]. In other words, EPO rs1617640 polymorphism can cause abnormal expression of EPO and affect the hematopoietic function of the body [50]. Moreover, we found a correlation between the EPO rs1617640 A$>$C genetic variant and the level of pre-drug lung function indexes in both Guangdong and Gannan. Similar to our results, Wang L found that serum EPO levels in COPD patients were correlated with FEV${}_1$% significantly and negatively through a clinical study (p 0.05) [21]. EPO levels may, to some extent, reflect the condition and prognosis of COPD patients. Therefore, we concluded that rs1617640 might affect the risk of COPD by influencing the binding of transcription factors to DNA fragments and regulating the expression level of target genes.

Our clinical cohort found that the incidence of COPD and the mortality without COPD differed between men and women, both indicators were higher in males than in females in all age strata (see Supplementary Table 3). This is consistent with the findings of Garcia Rodriguez in the incidence of COPD [51]. Hence, we decided to fit a 10-year absolute risk model for COPD in southern Chinese by sex. We found the factors retained in the models of males and females were slightly different (see Table 3), which may be related to the reduction of sample size after stratification. At the same time, the differences in the social division of labor and lifestyles between men and women may also lead to differences in risk factors [52, 53]. In the model for the female, the population attributable risk was 0.262 (0.128–0.396), indicating that additional risk factors need to be considered to predict the risk of COPD accurately [54].

There are some limitations in this study. First, this was a case-control study, and there was a certain degree of recall bias when obtaining environmental exposure information. Secondly, the participants in this study were mainly Han Chinese, which may restrict the extrapolation of conclusions to other populations. Moreover, this study clarified the association between EPO polymorphisms and COPD using epidemiological methods, which needed further verification by cellular and molecular experiments. Finally, this study did not conduct whole genome sequencing and analysis, it’s a potential limitation of the study.

In conclusion, this study reported for the first time that the EPO rs1617640 A$>$C polymorphism was associated with COPD susceptibility, and the C allele was associated with better lung function in southern Chinese. On this basis, the study suggests that EPO rs1617640 A$>$C can be used to predict the 10-year absolute risk for COPD among southern Chinese as a genetic marker when combined with environmental factors.

COPD, chronic obstructive pulmonary disease; EPO, Erythropoietin; GWAS, genome-wide association analysis; SNPs, the single nucleotide polymorphisms; FEV${}_1$, forced expiratory volume in 1 second; FVC, forced vital capacity; MAFs, the minor allele frequencies; PCR, TaqMan real-time polymerase chain reaction.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

YW and ZL conceived and designed the study with support from XW and JL as well as wrote the manuscript; XZ conducted the experiment; YW, ZL, AL and XZ preformed the statistical analysis of the results; YP, CM, CC, CX, DH, YD, and XZ participated in sample collection for the case-control study and prevalence study, as well as the electronic medical record entry. XW and JL were responsible for the design and quality control of the whole research and the guidance of the methodology. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work to take public responsibility for appropriate portions of the content and agreed to be accountable for all aspects of the work in ensuring that questions related to its accuracy or integrity.

This study was performed under the Declaration of Helsinki and approbated by the institutional review boards of Xi’an Jiaotong University Health Science Center (approval: XJTU 2016-411) and Guangzhou Medical University (approval: GZMC2007-07-0676).

In addition, we explained the purpose of the study to all participants at the time of the study and obtained their signed informed consent.

Thanks to the Institute of Public Health, Guangzhou Medical University. Thanks to the researchers of this laboratory for their guidance on experimental technology.

This study was supported by the following grants: National Key Research and Development Project of China 2017YFC0907202 (Xinhua Wang); National Natural Science Foundation of China 81872694, 82173609 (Jiachun Lu); Local Innovative and Research Teams Project of Guangdong Pearl River Talents Program 2017BT01S155 (Jiachun Lu); Guangdong Medical Science and Technology Research Fund Project A2020324 (Cuiyi Chen); National Natural Science Foundation of China (81460123) and Guangxi Natural Foundation (2018GXNSFAA281187) (Yibin Deng); National Natural Science Foundation of China (82260889) (Xuhui Zhang); the Science and Technology Project of Guangzhou (No.202103000073).

The authors declare no conflict of interest.