Introduction

Cystic fibrosis (CF) (OMIM: 219700) has a high clinical variability, even in patients with similar genotype of the cystic fibrosis transmembrane regulator gene (CFTR) (Cutting 2010). The variability occurs due to environmental factors (Collaco et al. 2010), including medical practices and socioeconomic status (Simmonds et al. 2010), to modifier genes (Cutting 2010; Simmonds et al. 2010; Stanke et al. 2011), and to classes of CFTR mutations (Marson et al. 2015). Numerous modifier genes are involved in immunity, inflammation, and controlling infection, and they area also important factors for clinical variability in CF (Davies et al. 2005; Slieker et al. 2005). Some of these modifier genes were studied by our research group as well as other international teams (Lima et al. 2012; Furgeri et al. 2012; Marson et al. 2012a, b, 2013a, b, c, 2014a, b; Rezende et al. 2013; Coutinho et al. 2014; Gallati 2014; Bombieri et al. 2015; Corvol et al. 2015; Brennan and Schrijver 2016).

Admittedly, interleukin 8 (IL8) is a pro-inflammatory cytokine and is predominantly produced by monocytes, macrophages, smooth muscle cells and endothelial cells. The mainly role of IL8 is the initiation and increase of the response to inflammatory processes caused by pathogens, promoting the activation, migration, adhesion and phagocytosis of neutrophils cells from peripheral blood to tissues. IL8 also has chemotactic activity against T cells and basophils (Jobe and Ikegami 1998; Harada et al. 1994; Papoff 2000; Velloso et al. 2005; Jundi and Greene 2015). Inflammation, a final common pathway that causes lung injury, plays an important role in increasing vascular permeability, contributing to interstitial, alveolar, and airway edema. One important aspect in CF lung disease is a neutrophil disease with a crucial role of IL8 protein. In this context, the role of IL8 in the complex inflammatory response in CF should be pointed out, especially in lung diseases (Reeves et al. 2010), the studies of which highlight the relationship between IL8 gene polymorphisms [rs4073 (-251T/A), rs2227306 (+781C/T) and rs2227307 (+396T/G)], clinical symptoms, and signs in CF and other lung diseases (Heinzmann et al. 2004; Hillian et al. 2008; Corvol et al. 2008; Scarel-Caminaga et al. 2011), such as the modifier gene.

The IL8 gene is located on chromosome 4 in the region q13.3, spanning a region of approximately 2.75 Mb, with 4 exons described. It has a transcript of 1705 bps and expresses a protein of 99 amino acid residues. Several mutations have been described in the IL8 gene, but only rs4073, rs2227306 and rs2227307 polymorphisms were correlated with the clinical severity of CF. Despite that IL8 plays a key role in the pathophysiology of inflammation of the airways of CF patients caused by a deficiency or absence of the CFTR protein to the best of our knowledge, there are few studies of variants in IL8 gene in CF. We searched in the scientific literature the main articles published to choose the polymorphisms that were evaluated in our data (Hillian et al. 2008; Corvol et al. 2008). Therefore, this study aimed to evaluate the influence of IL8 gene polymorphisms (rs4073, rs2227307 and rs2227306) on CF, in association with clinical severity markers in patients with one or two identified mutations in the CFTR gene belonging to Class I, II and/or III, as well as patients with no identified mutation in this gene, or belonging to Class IV, V or VI.

Method

Patients

A cross-sectional study was conducted with 186 CF patients, selected in two university referral centers (Campinas/São Paulo/Brazil and Ribeirão Preto/São Paulo/Brazil) for CF treatment, during the period from 2013 to 2015. All CF patients were invited to participate in the study. The CF patients were consistently followed at each participating center (4–6 routine visits per year). The CF patients inclusion was performed by clinical data and laboratorial diagnosis (sweat test and/or CFTR mutation screening). The exclusion of CF patients was performed in the absence of genetic material for analysis of IL8 polymorphisms and/or absence of clinical variables assessed in the last year. The study was approved by the Research Ethics Committee of the School of Medical Sciences of the University of Campinas, São Paulo, Brazil (#528/2008). All participants were informed of the study and signed an informed consent document. For patients under the age of 18, the informed consent document was signed by the minor’s parent or guardian. The study followed the recommendations of the Declaration of Helsinki.

The diagnosis of CF was confirmed by the presence of two altered concentrations of sodium and chloride in sweat (chloride level higher than 60 mEq/L). In addition, no patient underwent initial immunoreactive trypsinogen (IRT) measurement. In a group of 91 patients, there were two mutations in the CFTR gene belonging to Class I, II and/or III (associated with greater disease severity, due to the absence or non-functionality of the CFTR protein) (Marson et al. 2015). A group of 60 patients did not show identified mutation of the CFTR gene, or had two mutations belonging to Class IV, V or VI. A group of 35 patients had a mutation in the CFTR gene belonging to Class I, II or III, a non-identified mutation, or belonging to Class IV, V or VI (Table 1).

Table 1 Distribution of patients with cystic fibrosis considering the genotype for mutations in the CFTR gene and classes of identified mutations
Full size table

Clinical variables

The following variables were analyzed: clinical scores [Shwachman–Kulczycki (≤65 and >65), Kanga (≤17.5 and >17.5), and Bhalla (≤8 and >8)] (Santos et al. 2004); body mass index (BMI) for patients older than 18 years, using the formula BMI = weight/(height)2; the WHO (World Health Organization) Anthro program version 3.0.1 (2006) was used for children under 5 years of age and the WHO Anthro Plus version 1.0.2 (2007) was used for 5-year-old children to 18-year-old teenagers; patient age (≤143 and >143 months) and age at diagnosis (two groups according to the altered concentrations of sodium and chloride at the end of the second test: ≤20 and >20 months); first clinical symptom (general: ≤3 and >3 months; pulmonary symptoms: ≤6–>6 months; digestive symptoms:≤3 and >3 months); period until the first colonization with Pseudomonas aeruginosa (≤31 and >31 months); microorganisms identified in the routine sputum culture (mucoid and non-mucoid P. aeruginosa, Achromobacter xylosoxidans, Burkolderia cepacia and Staphylococcus aureus); transcutaneous arterial hemoglobin oxygen saturation (SaO2: ≤96 and >96 %); spirometry and comorbidities (nasal polyps, osteoporosis, meconium ileus, diabetes mellitus and pancreatic insufficiency).

Clinical severity scores were assessed by two trained clinicians using double-blind trials; in case of disagreement, a third clinician evaluated the scores. The Shwachman–Kulczycki (SK) scoring system compares the clinical manifestations of patients, detects treatment effects, and helps to determine criteria for the diagnosis of clinical severity. In this context, the system assesses four major domains: general activity, nutrition, radiographic findings of the thorax and physical examination. The score for each domain is evaluated according to five to 25 points. The lower values are associated with greater severity of CF (Santos et al. 2004).

The Bhalla score is a tomographic scoring system, which assesses pulmonary involvement in CF, determines therapeutic effects, and helps to screen patients for lung transplantation. The score shows low variation among examiners, high reproducibility, and correlation with pulmonary function values. It includes nine categories, with three points each and a maximum of 25 points. Lower values are associated with the severity of CF (Santos et al. 2004).

The Kanga scoring system detects CF pulmonary exacerbation, predicts the improvement or worsening of respiratory function, evaluates therapeutic effects with little variation among examiners and high correlation with the results of pulmonary function [mainly for forced vital capacity (FVC) and forced expiratory volume in 1 s of FVC (FEV1)]. The score identifies day-to-day clinical changes and includes analysis of five common symptoms (cough, fluid secretion, decreased appetite, shortness of breath and fatigue) and five physical signs (temperature, weight, respiratory rate, wheezing and breathing sounds). Each criterion varies from one to five points. The highest values are associated with severity of CF (Santos et al. 2004).

Spirometry was performed in patients over seven years of age, using the CPFS/D spirometer (MedGraphics, Saint Paul, Minnesota, USA). Data were recorded by PF BREEZE software version 3.8B for Windows 95/98/NT (American Thoracic Society 2012) and assessed in percentage predicted value: FVC %, FEV1 %, FEV1/FVC, forced expiratory flow at 25 % of FVC (FEF25 %), forced expiratory flow at 50 % of FVC (FEF50 %), forced expiratory flow at 75 % of FVC (FEF75 %), forced expiratory flow between 25 and 75 % of FVC (FEF25–75 %) and expiratory reserve volume (ERV). Spirometry data are shown in percentage of the predicted value according to the Polgar and Promadhat (1971), Pereira et al. (2007) and Duarte et al. (2007) equations.

DNA extraction and genotyping

Genomic DNA was extracted from peripheral blood samples using standard phenol–chloroform method and quantified by spectrophotometer GE NanoVue™ (GE Healthcare Biosciences, Pittsburgh, USA). In this study, the sample final concentration was set at 50 ng/µL.

Mutations of CFTR were analyzed by polymerase chain reaction (PCR) (F508del) followed by enzymatic digestion (G542X, R1162X, R553X, G551D and N1303K)—[PCR/restriction fragment length polymorphism (RFLP)]. Other mutations in the CFTR gene could be identified by sequencing or use of the SALSA MLPA method (Multiplex Ligation-dependent Probe Amplification) Kit P091-C1 CFTR-MRC-Holland S4X, 2183A>G, 1717-G>A, I618T with MegaBace1000® (GE Healthcare Biosciences, Pittsburgh, USA) and ABI 3500 (Applied Biosystems—Thermo Fisher Scientific, São Paulo/SP, Brazil) (Bonadia et al. 2014).

Considering the CFTR genotypes, patients were divided into three groups: (1) with two identified mutations belonging to Class I, II and/or III; (2) with one identified mutation belonging to Class I, II and/or III; (3) no identified mutation belonging to Class I, II and/or III. Other mutations identified in the CFTR gene belonging to Class IV, V and/or VI were not included in the statistical analysis.

IL8 gene polymorphisms were analyzed by PCR/RFLP. For the rs4073 polymorphism, the primers 5′-CCA TCA TGA TAG CAT CTG TA-3′ and 5′-CCA CAA TTT GGT GAA TTA TTA A-3′ as well as the AseI restriction enzyme (Heinzmann et al. 2004) were used; for the rs2227306 polymorphism, primers 5′-CTC TAA CTC TTT ATA TAG GAA TT-3′ and 5′-GAT TGA TTT TAT CAA CAG GCA-3′, and EcoRI restriction enzyme (Heinzmann et al. 2004); for the rs2227307 polymorphism, primers 5′-TAA AGG TTT GAT CAA TAT AGA-3′ and 5′-CTT CCT TCT AAT TCCA ATA TG-3′, and ScrFI restriction enzyme (Scarel-Caminaga et al. 2011). The products of the enzymatic restriction were submitted to electrophoresis on a 12 % polyacrylamide gel, or 4 % agarose gel (Heinzmann et al. 2004; Scarel-Caminaga et al. 2011), and stained with Red® gel.

Statistical analysis

Statistical analyses were performed using the Statistical Package for Social Sciences software (SPSS) version 22.0 (SPSS Inc., Chicago, IL, USA). The GPower software version 3.0.3.1 (Faul et al. 2007) was used to calculate the sample power, considering the genotype of the analyzed polymorphisms and adopting power for value above 80 %.

Several tests have been performed. (1) Mann–Whitney and Kruskal–Wallis were used for data with numerical distribution. In case of significance between groups for the Kruskal–Wallis test, identification of the difference between groups was performed by MedCalc® software for Windows, version 16.1 (MedCalc® Software, Ostend, Belgium). (2) Chi-square test (χ 2) and Fisher’s exact test were used for data with categorical distribution, with subsequent analysis by odds ratio (OR). Confidence interval was established considering the Fisher’s exact test as parameter for the test. The calculation of the OR was conducted with OpenEpi software version 3.03a (Dean et al. 2004). For analysis of the Hardy–Weinberg equilibrium (HWE), the Online Encyclopedia software for Genetic Epidemiology Studies (OEGE) (Rodriguez et al. 2009) was used.

The data regarding clinical severity, with high standard deviation, were evaluated by the median, including: (1) patient’s age; (2) time to diagnosis; (3) time to onset of symptoms (general, respiratory and digestive symptoms); (4) time to first colonization with isolated Pseudomonas aeruginosa; (5) SaO2.

For the identification of mutations in the CFTR gene, the patients were analyzed based on two contexts: (1) all CF patients, regardless of mutations of the gene (N = 186 patients); (2) patients with two mutations belonging to Class I, II and/or III (N = 91 patients). As for the polymorphisms, four analysis models were adopted: (1) co-dominant; (2) recessive; (3) dominant; (4) over-dominant, applied in association with clinical variables. For all analyses, alpha value was set at 0.05.

The false discovery rate (FDR) test was applied to correct the multiple test comparison (Online Resource 1, 2, 3 and 4). FDR is an approach to the multiple comparisons problem. Instead of controlling the chance of any false, FDR controls the expected proportion of false positives among supra-threshold voxels. An FDR threshold is determined from the observed p value distribution, and hence is adaptive to the amount of signal in the data (Benjamini 2010). The p value (P) and p value corrected (P c) are shown. The linkage disequilibrium analysis was performed in Haploview software version 4.2 (Barrett et al. 2005).

Results

The descriptive analysis of clinical data and severity markers of CF patients is presented in Table 2. Table 3 shows genotype and allele frequencies of the IL8 gene polymorphisms and their respective HWE calculations, indicating absence of equilibrium in rs2227306 and rs4073 (P < 0.05). There was prevalence of the heterozygous genotype for all polymorphisms (rs4073 = 42.7 %; rs2227306 = 38.6 %; rs2227307 = 43.3 %), with emphasis on the homozygous variant genotype for the rs2227307 polymorphism (TT = 39.3 %), as well as the variant allele (T = 0.61).

Table 2 Descriptive analysis of clinical and laboratory markers in patients with cystic fibrosis
Full size table
Table 3 Distribution of genotypes and alleles of IL8 polymorphisms (rs4073, rs2227306 and rs2227307) in patients with cystic fibrosis
Full size table

Online resource Tables 1, 2 and 3 show genotype comparisons of IL8 gene polymorphisms for recessive, dominant, co-dominant and over-dominant models, considering pulmonary, clinical and demographic markers. Variables with statistical significance (P < 0.05; Pc < 0.05) are presented in Table 4. An association with SaO2 (≤96 and >96 %), presence of mucoid Pseudomonas aeruginosa (MPA) and age (≤143 and >143 months) could be observed. There was an association of rs4073 (AT genotype), higher values of SaO2 (>96 %; OR = 0.459; 95 % CI = 0.228–0.914) and rs2227306 (wild-type homozygous CC genotype) with reduced SaO2 values (≤96 %; OR = 2.412; 95 % CI = 1.165–5.127). The rs2227307 polymorphism was associated with increased SaO2 values (GT genotype: OR = 0.390; 95 % CI = 0.190–0.790) and reduced SaO2 values [homozygous variant (TT) genotype: OR = 3.669; 95 % CI = 1.734–8.054].

Table 4 Association of clinical markers and age with IL8 gene polymorphisms in patients with cystic fibrosis, regardless of CFTR classification and with two mutations (2IM)
Full size table

For MPA, there was an association with wild-type homozygous genotype of rs2227306 (CC: OR = 2.065; 95 % CI = 1.05–4.098) as well as the homozygous variant of rs2227307 (TT: OR = 2.037; 95 % CI = 1.029–4.068).

Regarding age, the age group younger or equal to 143 months was associated with the heterozygous genotype of rs2227307 (GT) in patients with two mutations of the CFTR gene (OR = 3.199; 95 % CI = 0.106–0.770); whereas the homozygous variant (TT) genotype was associated with the age group older than 143 months (OR = 0.290; 95 % CI = 0.106–0.770). The same occurred for the homozygous variant (TT) genotype of rs2227307, regardless of the CFTR genotype (OR = 0.358; 95 % CI = 0.175–0.718).

The IL8 gene polymorphisms were also associated with their respective measures of pulmonary function markers (SaO2, FVC, FEV1, FEV1/FVC, FEF25 %, FEF50 %, FEF75 %, FEF25-75 %, FEFmax and ERV) and clinical severity (according to Bhalla, Kanga and Shwachman–Kulczycki scoring systems). The mean, standard deviation and median values are shown in Table 5. The rs2227307 polymorphism, represented by the homozygous variant (TT) genotype prevailed in association with the above-mentioned variables in CF patients, regardless of mutations in the CFTR gene. They also showed significant reduction in their respective values (except for increased values of Kanga score), compared with the GG and/or GT genotypes (P < 0.05; P c < 0.05). This demonstrated an association with the severity of the disease. For the Bhalla score, there was prevalence of lower values in the presence of wild-type allele (G, P = 0.009; P c = 0.018), demonstrating a protective character. On the other hand, in patients with CF and two mutations in the CFTR gene, the rs4073 and rs2227306 polymorphisms of IL8 prevailed, represented by their respective homozygous variant genotype (TT for both), mainly related to lower spirometric parameters (P < 0.05; P c < 0.05), both demonstrating an association with CF severity. Table 5 shows further comparative analysis among analysis models that also showed significance (P < 0.05; P c < 0.05).

Table 5 Association of clinical markers with IL8 gene polymorphisms in patients with cystic fibrosis, regardless of CFTR classification, and with two mutations (2IM)
Full size table

Additional statistical analysis considering the pancreatic insufficiency status as clinical marker of CF severity is shown in the Online Resource 4. The haplotype distribution for the IL8 gene polymorphisms is presented in the Online Resource 5.

Discussion

IL8 plays an important role for the initiation and increase of response to inflammatory processes, promoting the activation and migration of neutrophils, acting in the complex inflammatory response in CF. This study showed an association of different IL8 gene polymorphisms with clinical parameters related to severity of CF, especially with lung disease markers, in agreement with other authors (Hillian et al. 2008). The protein of IL8 is involved in the activation and migration of neutrophils, acting as pro-inflammatory mediator of initiation and amplification of acute and chronic responses. This emphasizes the impact of rs2227307 on changes of the spirometry.

The multiple factors involved in CF and few of the studies with IL8 variants in diseases with pulmonary phenotypic expression make the discussion of this study results rather challenging. This research adds significant information about genetic factors involved in the severity and variability of CF, which has also been reported in other diseases with pulmonary phenotypic expression: bronchial asthma (Heinzmann et al. 2004; Cheong et al. 2006; Puthothu et al. 2006; Klaassen et al. 2014), acute lung injury (O’Mahony et al. 2012), bronchiolitis (Hull et al. 2000, 2001; Puthothu et al. 2006), chronic obstructive pulmonary disease (Matheson et al. 2006), recurrent wheezing (Esposito et al. 2014a, b), and juvenile idiopathic arthritis (Heinzmann et al. 2004). Furthermore, the association of IL8 variants with other diseases should be pointed out, namely lupus nephritis (Rovin et al. 2002), periodontitis (Scarel-Caminaga et al. 2011), breast cancer (Wang et al. 2014), lung diseases (Gao et al. 2014), prostate diseases (Chen et al. 2015), ovarian diseases (Koensgen et al. 2015), and sepsis in premature neonates (Esposito et al. 2014b).

In this study, two groups of patients were considered for the analysis of IL8 gene polymorphisms, according to the mutation in the CFTR gene: (1) with two mutations belonging to Class I, II and/or III; (2) no identified mutation in CFTR or belonging to Class IV, V or VI. Group (1) includes patients with mutations associated with greater severity (Marson et al. 2015). In this group, there were changes in spirometry parameters related to rs4073 and rs2227306 polymorphisms, represented by their respective homozygous variant genotypes, suggesting worsening of clinical symptoms. In Group (2), rs2227307 polymorphism stood out, whose homozygous variant genotype is related to changes in a larger number of spirometric parameters, in relation to other polymorphisms studied. This may potentiate severity of the disease.

There are reports on the association of IL8 gene polymorphism, represented by rs2227307, with severity of lung disease in CF, involving F508del-CFTR homozygous patients (Hillian et al. 2008). On the other hand, that study showed the same results for rs4073 and rs2227306 polymorphisms, but only in male patients with other CFTR genotypes and suggested the differential expression of IL8 variant allele after a promoter luciferase assay. In addition, they found the strongest effect among males. In this context, gender, CFTR genotype and modifier genes modulate the presence and severity of lung diseases (Hillian et al. 2008).

However, in genetic-related diseases with phenotypic pulmonary expression, IL8 gene polymorphisms may act differently. Interestingly, T-allele of rs4073 (variant) was associated with greater risk of asthma; whereas A-allele (wild-type), with bronchiolitis (Hull et al. 2000, 2001). In cases of asthma, the rs2227306 polymorphism alone was associated with: Group 1—asthma versus controls, and Group 2—asthma versus patients with rhinovirus infections and in haplotype with the rs4073 polymorphism, considering both groups (Puthothu et al. 2006). The study conducted by Heinzmann et al. (2004) associated four IL8 gene polymorphisms with asthma, juvenile idiopathic arthritis and allergy (atopy), with emphasis on patients with asthma, compared with the control group and patients with juvenile idiopathic arthritis. These findings suggest that IL8 gene polymorphisms have different effects on the dynamics of pulmonary inflammation, and they are likely to be associated with multiple genes, which may influence the inflammatory response. Furthermore, this study showed that IL8 gene polymorphisms also act differently on the expression of spirometric parameters in a same disease. This increases the severity of lung disease, according to the type of mutation in the CFTR gene, which determines CF.

In this study, the rs4073 and rs2227306 polymorphisms, represented by their homozygous variant alleles, showed an association with fewer changes in spirometry parameters, particularly in group (2). In another study, involving acute lung injury, rs4073 was associated with the duration of mechanical ventilation (O’Mahony et al. 2012). Moreover, in patients with recurrent wheezing, rs4073 showed higher risk and severity of the disease (Esposito et al. 2014a). However, other studies of lung diseases do not confirm an association with the rs4073 polymorphism (Matheson et al. 2006; Klaassen et al. 2014). This study pointed out the rs2227307 polymorphism, particularly in group (2), which is associated with a greater number of changes in spirometry, and stands out as a possible severity marker in CF.

Regarding the HWE, two polymorphisms (rs4073 and rs2227306) were not in balance. We must remember that the HWE assumes an ideal population, without the interference of evolutionary factors. However, in genes as those involved in immunity, inflammation, and infection control, the HWE imbalance may appear secondarily associated with the selection mechanisms that favored a particular allele that can bring a more effective response (Huttley and Wilson 2000). The disequilibrium does not invalidate the association study since the groups are part of the same population (Huttley and Wilson 2000).

Further studies with different populations involving molecular markers may contribute to clarify the mechanisms related to CF and the complex pulmonary dynamics, including the antagonist data about IL8 and genetic variants, as well as the IL8 expression associated with the genetic polymorphisms.

Conclusion

IL8 gene polymorphisms (rs4073, rs2227306, and especially rs2227307) are associated with CF markers of severity, mainly related to lung diseases, a major factor of morbidity and mortality. Even with monogenic autosomal recessive inheritance, CF has high clinical variability, and the IL8 and other inflammatory molecules seem to potentiate this process.