PLoS ONE plos plosone PLOS ONE 1932-6203 Public Library of Science San Francisco, CA USA 10.1371/journal.pone.0195015 PONE-D-17-29874 Research Article Medicine and health sciencesVascular medicineCoronary heart diseaseMedicine and health sciencesCardiologyCoronary heart diseaseBiology and life sciencesPhysiologyImmune physiologyAntibodiesAutoantibodiesMedicine and health sciencesPhysiologyImmune physiologyAntibodiesAutoantibodiesBiology and life sciencesImmunologyImmune system proteinsAntibodiesAutoantibodiesMedicine and health sciencesImmunologyImmune system proteinsAntibodiesAutoantibodiesBiology and life sciencesBiochemistryProteinsImmune system proteinsAntibodiesAutoantibodiesMedicine and health sciencesVascular medicineAtherosclerosisMedicine and health sciencesVascular medicineBlood pressureHypertensionMedicine and health sciencesEndocrinologyEndocrine disordersDiabetes mellitusMedicine and health sciencesMetabolic disordersDiabetes mellitusBiology and life sciencesBiochemistryBiomarkersMedicine and health sciencesPulmonologyApneaSleep apneaMedicine and health sciencesNeurologySleep disordersSleep apneaMedicine and health sciencesNeurologySleep disorders Circulating autoantibodies against neuroblastoma suppressor of tumorigenicity 1 (NBL1): A potential biomarker for coronary artery disease in patients with obstructive sleep apnea NBL1-Abs: Potential biomarker for CAD in patients with OSA Matsumura Takuma Data curation Formal analysis Investigation Methodology Writing – original draft 1 http://orcid.org/0000-0001-9019-4053 Terada Jiro Conceptualization Formal analysis Funding acquisition Investigation Methodology Project administration Writing – original draft 1 * Kinoshita Taku Data curation Investigation 1 Sakurai Yoriko Data curation Investigation 1 Yahaba Misuzu Data curation Investigation 1 Tsushima Kenji Data curation Investigation Supervision 1 Sakao Seiichiro Data curation Investigation Supervision 1 Nagashima Kengo Formal analysis Methodology Validation 2 Ozaki Toshinori Methodology Supervision 3 Kobayashi Yoshio Data curation Supervision 4 Hiwasa Takaki Conceptualization Data curation Formal analysis Investigation Methodology Resources Supervision 5 Tatsumi Koichiro Conceptualization Project administration Resources Supervision Validation Writing – original draft 1 Department of Respirology, Graduate School of Medicine, Chiba University, Chiba, Chiba, Japan Department of Global Clinical Research, Graduate School of Medicine, Chiba University, Chiba, Chiba, Japan Laboratory of DNA Damage Signaling, Chiba Cancer Center Research Institute, Chiba, Chiba, Japan Department of Cardiovascular Medicine, Graduate School of Medicine, Chiba University, Chiba, Chiba, Japan Department of Biochemistry and Genetics, Graduate School of Medicine, Chiba University, Chiba, Chiba, Japan Taniyama Yoshiaki Editor Osaka University Graduate School of Medicine, JAPAN

The authors have declared that no competing interests exist.

 * E-mail: jirotera@chiba-u.jp 29 3 2018 2018 13 3 e0195015 12 8 2017 15 3 2018 2018 Matsumura et al This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Objective

Although severe obstructive sleep apnea (OSA) is an important risk factor for atherosclerosis-related diseases including coronary artery disease (CAD), there is no reliable biomarker of CAD risks in patients with OSA. This study aimed to test our hypothesis that circulating autoantibodies against neuroblastoma suppressor of tumorigenicity 1 (NBL1-Abs) are associated with the prevalence of CAD in patients with OSA.

 Methods

Eighty-two adults diagnosed with OSA by polysomnography, 96 patients with a diagnosis of acute coronary syndrome (ACS) and 64 healthy volunteers (HVs) were consecutively enrolled. Serum samples were collected from patients with OSA at diagnostic polysomnography and from patients with ACS at disease onset. Serum NBL1-Ab level was measured by amplified luminescence proximity homogeneous assay and its association with clinical variables related to atherosclerosis was evaluated.

 Results

NBL1-Ab level was significantly elevated in patients with both OSA and ACS compared with HVs. Subgroup analyses showed that NBL1-Ab level was markedly higher in patients with severe OSA and OSA patients with a history of CAD. Weak associations were observed between NBL1-Ab level and apnea-hypopnea index, age, mean SpO2 and arousal index, whereas significantly higher NBL1-Ab levels were observed in OSA patients with a history of CAD than in those without a history of CAD. Sensitivity analysis using a logistic regression model also demonstrated that increased NBL1-Ab levels were associated with the previous history of CAD in patients with OSA.

 Conclusions

Elevated NBL1-Ab levels may be associated with the prevalence of CAD in patients with OSA, which needs to be confirmed further.

 Japan Society for the Promotion of Science (JSPS) KAKENHI 16K09528 http://orcid.org/0000-0001-9019-4053 Terada Jiro Grants-in-Aid of the Japan Science and Technology Agency (JST) 14657335 Hiwasa Takaki This study was supported in-part by a grant to The Intractable Respiratory Diseases and Pulmonary Hypertension Research Group, the Ministry of Health, Labor and Welfare, Japan (http://www.mhlw.go.jp/english/index.html); the Japan Agency for Medical Research and Development (AMED, https://www.amed.go.jp), Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant number 16K09528, https://www.jsps.go.jp/index.html) and Grants-in-Aid of the Japan Science and Technology Agency (JST) in Japan (Exploratory Research No. 14657335, http://www.jst.go.jp/index.html). There was no additional external funding received for this study. These funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript. Data Availability All relevant data are within the paper and its Supporting Information files. Introduction

Patients with obstructive sleep apnea (OSA) experience repetitive upper airway obstruction during sleep, which causes frequent sleep fragmentation and nocturnal intermittent hypoxia (IH). Sleep fragmentation and IH promote oxidative stress by accelerating the production of reactive oxygen species. They also enhance inflammation of systemic, vascular and adipose tissue by elevating levels of inflammatory cytokines and cause upregulation of blood pressure with increased sympathetic activity [1, 2]. These OSA-related reactions induce endothelial dysfunction, leading to the development of atherosclerosis and eventually to coronary artery disease (CAD), strokes and atherosclerosis-related deaths, especially in patients with severe OSA [3–5].

OSA treatment with continuous positive airway pressure (CPAP) therapy can reduce the impact of inflammation, blood lipid values and elevated blood pressure in patients with OSA [6–8]. Despite its benefits, adherence to CPAP therapy has been reported to range from 29% to 85%, which is not ideal [9]. Therefore, biomarkers of patients with OSA who require more careful and intensive treatment (i.e. greater adherence to CPAP therapy) are needed. Several markers of inflammation and endothelial dysfunction have been proposed as candidate markers for OSA patients with atherosclerosis-related diseases (e.g. soluble tumour necrosis factor receptor, tumour necrosis factor-beta, interleukin-6 and soluble intercellular cell adhesion molecule-1). However, there is no consensus on the optimal biomarker of CAD risks in OSA. In clinical practice, although the ankle-brachial pressure index, brachial-ankle pulse wave velocity (baPWV), cardio-ankle vascular index, carotid intima-media thickness (IMT) and funduscopic examination have been used to evaluate arterial atherosclerosis, predicting the development of fatal acute coronary syndrome (ACS) is difficult.

Recently, circulating (i.e. serum) autoantibodies against atherosclerosis-specific antigens have been considered as novel biomarkers for evaluating atherosclerosis-related diseases. We have previously described the presence of the autoantibodies recognized by IgG antibodies in the sera of patients with atherosclerosis-related diseases such as myocardial infarction, stroke and diabetes [10, 11]. Among them, we have reported circulating autoantibodies against coatomer protein complex subunit epsilon, which is a potential biomarker of cardiovascular and cerebrovascular risk in patients with OSA [12]. Additionally, we have focused on autoantibodies against members of the bone morphogenetic protein (BMP) family because BMP may be associated with atherosclerosis and cardiovascular disease (CVD) [13]. BMP signalling is activated by BMP1, which degrades BMP antagonist chordin [14], resulting in atherosclerosis. The serum level of autoantibodies against BMP1 has been reported as a potential marker of CVD, ischemic stroke and transient ischemic attack [10, 15]. Other studies suggest that BMP2 and BMP4 are linked to the development of atherosclerosis [13], and BMP antagonist neuroblastoma suppressor of tumorigenicity 1 (NBL1, accession number X66872.1) specifically antagonizes BMP2, BMP4 and BMP14 [16, 17]. Thus, it is conceivable that there exists a functional relationship between NBL1 and atherosclerosis.

In this study, we evaluated serum autoantibodies against NBL1 (NBL1-Ab) level in patients with OSA and ACS, and tested the hypothesis that NBL1-Abs are increased in OSA patients with a history of CAD.

 Material and methods Ethical approval

All study procedures involving human participants were approved by the ethical standards of the Ethical Review Board of the Chiba University, Graduate School of Medicine (approval number 973). The study conformed to the ethical principles of 1964 Helsinki declaration and subsequent amendments. All study participants provided written informed consent.

 Study subjects

Eighty-two Japanese adults, 56 men and 26 women with a median age of 59.0 years, who were diagnosed with OSA by polysomnography (PSG) in our hospital between June 2012 and January 2014, were studied. Ninety-six patients with ACS, 81 men and 15 women, with a median age of 67.0 years, who were diagnosed in our hospital, were recruited at the disease onset between April 2011 and March 2014. ACS was defined as acute myocardial infarction (AMI) or unstable angina pectoris (UAP). Sixty-four healthy volunteers (HVs), 38 men and 26 women, with a median age of 42.5 years and no history of CAD or OSA were control subjects. The history of CAD was defined as a previous clinical history of myocardial infarction or angina pectoris. All HVs were enrolled at Chiba University. OSA and ACS patients with autoimmune diseases and ACS patients with a history of sleep-related breathing disorders according to the International Classification of Sleep Disorders-Third Edition criteria were also excluded.

 Clinical assessment

Atherosclerosis risk factors, including age, sex, body mass index (BMI), smoking status, hypertension, diabetes, hyperlipidemia, CAD and stroke, were collected from patient clinical records. Patients were divided into three groups by smoking status (i.e. never smoked, ex-smoker and current smoker). Hypertension was defined as a history of systolic blood pressure > 140 mmHg, diastolic blood pressure > 90 mmHg or use of antihypertensive agents, and diabetes was defined as the use of antidiabetes therapy or a history of diabetes. Hyperlipidemia was defined as a history of total cholesterol > 220 mg/dL, triglycerides > 150 mg/dL or use of lipid-lowering agents. Stroke was defined as a history of cerebral infarction or cerebral haemorrhage.

PSG was scored according to the 2007 American Academy of Sleep Medicine alternative criteria. Apnea was defined as a reduction of nasal airflow to < 10% of the baseline for ≥ 10 s. Hypopnea was defined as a reduction in the nasal airflow signal amplitude of ≥ 50% for ≥10 s in association with either a ≥ 3% oxygen desaturation or electroencephalographic arousal. OSA was defined as an apnea-hypopnea index (AHI) of five or more events per hour combined with predominantly obstructive respiratory events. OSA severity was classified according to AHI values as mild, 5–15; moderate, > 15–30 or severe, > 30.

 Blood sampling/purification and analysis

Blood samples were collected from patients with OSA on PSG evaluation and from HVs during their medical checkup. Blood was collected from patients with ACS at the diagnosis of AMI or UAP during emergency admission for percutaneous coronary intervention or coronary artery bypass grafting. Sera were centrifuged at 3000 g for 10 min at room temperature and the supernatant was stored at −80°C until use. Repeated thawing and freezing of the samples was avoided.

A full-length NBL1 cDNA was used to construct the expression vector for glutathione-S-transferase (GST)-tagged NBL1 protein. The cDNA sequence between positions 267 and 731 was inserted into Eco RI/Xho I site of pGEX-2T (GE Healthcare Life Sciences, Pittsburgh, PA, USA), which produces the truncated NBL1 protein (from amino acid residue 25 to 178) lacking its potential signal peptide sequence. GST-NBL1 protein was purified as described previously [10, 15].

Amplified luminescence proximity homogeneous assay (AlphaLISA) was performed in 384-well microtiter plates (white opaque ProxiPlate™, PerkinElmer, Waltham, MA, USA) containing 2.5 μL of 1/100-diluted sera and 2.5 μL of GST or a GST-fusion protein (10 μg/mL) in AlphaLISA buffer (25 mM HEPES, pH 7.4, 0.1% casein, 0.5% Triton X-100, 1 mg/mL dextran-500 and 0.05% Proclin-300). The resulting reaction mixture was incubated at room temperature for 6–8 h. Anti-human IgG-conjugated acceptor beads (2.5 μL of 40 μg/mL) and glutathione-conjugated donor beads (2.5 μL of 40 μg/mL) were added, and the samples were incubated for an additional 14 days at room temperature in the dark. Chemiluminescence was read on an EnSpire Alpha microplate reader (PerkinElmer) as previously described [11, 15]. Specific reactions were calculated by subtracting the Alpha values of GST control samples from those of samples containing GST-fusion proteins.

 Statistical analysis

For the baseline variables, summary statistics were constructed using medians and interquartile ranges for numerical data and frequencies, respectively, and proportions for categorical data. Patient characteristics were compared using the Fisher’s exact test for categorical data and the Mann—Whitney U or Kruskal—Wallis test for numerical data, as appropriate. NBL1-Ab levels in patients with OSA, patients with ACS and HVs were compared using the Kruskal—Wallis test. A post hoc analysis was performed for all pairwise comparisons using the Steel—Dwass method. Subgroup analyses were performed for NBL1-Ab levels by OSA severity (mild, moderate and severe) and by the presence or absence of CAD history. The subgroup analyses were conducted in the same manner as the pooled analysis. Correlations were evaluated using the Spearman’s correlation analysis, and the Fisher’s exact test was used to determine the significance of differences in the proportions of groups. The cut-off value of NBL1-Ab level for the prevalence of CAD among all patients with OSA was determined to maximize the sum of sensitivity and specificity by receiver operating characteristic (ROC) curve analysis. Univariate and multivariate logistic regression analyses were used to identify the set of variables that could classify patients according to CAD status. Nine covariates were included in the models, including age (years), obesity (BMI ≥ 25 kg/m2), smoking (current or ex-smoker), hypertension, diabetes, hyperlipidemia, stroke, severe OSA (AHI > 30) and elevated NBL1-Ab. All tests were two-tailed, and statistical significance was defined as a p value < 0.05. All statistical analyses were performed using JMP Pro 12.2.0 software (SAS Institute Inc., Cary, NC, USA). (Whole raw data sheet is provided in S1 Table)

 Results Patient characteristics

Clinical characteristics of patients with OSA, patients with ACS and HVs are shown in Table 1. Sixty-four HVs, 82 patients with OSA (11 mild, 17 moderate and 54 severe OSA) and 96 patients with ACS (72 AMI and 24 UAP) were enrolled in this study. As compared to HVs, patients with OSA were significantly older and had a higher BMI index. The patients with ACS were significantly older and included more male subjects than HVs. A history of atherosclerosis-related diseases (e.g. hypertension or CAD) was more frequently observed in patients with OSA and ACS than in HVs. As shown in Fig 1A, a significantly higher serum NBL1-Ab level was observed in patients with OSA (p = 0.034) and patients with ACS (p = 0.003) than in HVs. There were no significant differences in the NBL1-Ab levels in patients with OSA and ACS (p = 0.864). Subgroup analysis found that only patients with severe OSA and a history of CAD had significantly higher NBL1-Ab levels than HVs had (p = 0.010, Fig 1B and p = 0.008, Fig 1C). In patients with mild-to-moderate severity OSA (Fig 1B) and those with no history of CAD (Fig 1C), the NBL1-Ab level was similar to that in HVs.

10.1371/journal.pone.0195015.t001 Baseline characteristics of subjects.
HV (n = 64) OSA (n = 82) ACS (n = 96)
Age (years) 42.5 (35.3–55.8) 59.0 (49.8–66.5)*** 67.0 (60.0–73.0)***
Male sex 38 (59.4%) 56 (68.3%) 81 (84.4%)***
BMI (kg/m2) 23.1 (20.6–25.5) 25.9 (23.9–29.4)*** 23.4 (21.3–25.2)
OSA severity
Mild 11 (13.4%)
Moderate 17 (20.7%)
Severe 54 (65.9%)
AHI (events/h) 36.7 (22.6–50.4)
Mean SpO2 (%) 94.0 (93.0–96.0)
Lowest SpO2 (%) 78.0 (69.0–83.0)
Arousal index (events/h) 37.3 (22.2–50.3)
Smoking status (n = 60)
Never smoked 42 (70.0%) 42 (51.2%) 30 (31.3%)
Ex-smoker 10 (16.7%) 34 (41.5%) 38 (39.6%)
Current smoker 8 (13.3%) 6 (7.3%)** 28 (29.2%)***
Hypertension 8 (12.5%) 30 (36.6%)** 45 (46.9%)***
Diabetes 1 (1.6%) 17 (20.7%)** 17 (17.7%)**
Hyperlipidemia 2 (3.1%) 22 (26.8%)* 14 (14.6%)*
Previous history of CAD 0 (0.0%) 10 (12.2%)** 14 (14.6%)***
Stroke 0 (0.0%) 5 (6.1%) 5 (5.2%)

Data are medians (interquartile range) for numerical data and n (%) for categorical data.

ACS, acute coronary syndrome; AHI, apnea-hypopnea index; BMI, body mass index; CAD, coronary artery disease; HV, healthy volunteer; OSA, obstructive sleep apnea.

*p < 0.05 versus HV,

**p < 0.01 versus HV,

***p < 0.001 versus HV.

 10.1371/journal.pone.0195015.g001 Association between NBL1-Ab level and disease.

Pooled OSA group data were compared with those of patients with ACS and HVs. Significant differences were observed among the three groups using the Kruskal—Wallis test (p < 0.004). Post hoc analysis using the Steel—Dwass test revealed significant differences between both patients with OSA and ACS and the HVs (Fig 1A). Patients with OSA were stratified by OSA severity (Fig 1B) or by having a history of CAD (Fig 1C) and then compared with patients with ACS and HVs. Significant differences were observed in both comparisons using the Kruskal—Wallis test (p < 0.003 and p < 0.001). Post hoc analysis using the Steel—Dwass test revealed significant differences between both patients with severe OSA and OSA patients with CAD versus HVs. Horizontal lines represent medians, boxes represent the 25th and 75th percentiles, whiskers represent the 10th and 90th percentiles and dots represent outliers. ACS, acute coronary syndrome; CAD, coronary artery disease; HV, healthy volunteer; NBL1-Abs, autoantibodies against NBL1; OSA, obstructive sleep apnea; w/o, without.

 Association of NBL1-Ab level with clinical parameters of patients with OSA

The relationships of NBL1-Ab level and clinical parameters in patients with OSA are shown in Fig 2. An association was observed between NBL1-Ab levels and AHI (ρ = 0.30, p = 0.006, Fig 2D), indicating a weak but significant dose response for NBL1-Ab levels in OSA. In addition, a very weak to weak association was observed between NBL1-Ab level and age (ρ = 0.13, p = 0.042, Fig 2A), mean SpO2 (ρ = −0.28, p = 0.010, Fig 2E) and arousal index (ρ = 0.28, p = 0.012, Fig 2F), whereas a significantly higher NBL1-Ab level was observed in patients with CAD than in those without CAD (p = 0.015, Fig 2K). No significant associations were found between NBL1-Ab level and sex, BMI, smoking status, hypertension, diabetes, hyperlipidemia or stroke. However, regarding absolute values of serum lipid levels, a weak association was observed between NBL1-Ab level and low-density lipoprotein (LDL) (ρ = 0.24, p = 0.034) and total cholesterol (ρ = 0.28, p = 0.014), whereas no significant associations were found between NBL1-Ab level and high-density lipoprotein and triglyceride level (data not shown).

10.1371/journal.pone.0195015.g002 Association of NBL1-Ab with clinical parameters in patients with OSA.

Correlations between NBL1-Abs and age (A), gender (B), BMI (C), AHI (D), mean SpO2 (E), arousal index (F), smoking status (G), hypertension (H), diabetes (I), hyperlipidemia (J), CAD (K), and stroke (L) were evaluated. Weak associations were observed between NBL1-Ab levels and age, AHI, mean SpO2 and arousal index. OSA patients with CAD had significantly higher NBL1-Ab levels than OSA patients without CAD. Spearman’s correlation analysis (A, C—F), Mann—Whitney U test (B, H-L) and Kruskal—Wallis test (G) were used. Horizontal lines represent medians, boxes represent the 25th and 75th percentiles, whiskers represent the 10th and 90th percentiles and dots represent outliers. AHI, apnea-hypopnea index; BMI, body mass index; CAD, coronary artery disease; NBL1-Abs, autoantibodies against NBL1; OSA, obstructive sleep apnea.

We performed a sensitivity analysis to identify the strength of the association of clinical parameters including NBL1-Ab level with the prevalence of CAD using univariate and multivariate logistic regression models as shown in Table 2. For patients with OSA, an optimal cut-off value of NBL1-Ab for predicting the prevalence of CAD history was determined to be 27512 by ROC curve analysis, with a sensitivity of 70.0% and a specificity of 77.8% (Fig 3). The area under the curve was 0.739 (95% confidence interval [CI]: 0.560–0.863). Using the NBL1-Abs cut-off value, univariate logistic regression revealed an association of elevated NBL1-Ab level with the prevalence of CAD (odds ratio [OR]: 8.17, 95% CI: 2.03–41.4, p = 0.003). Multivariate logistic regression analysis, which included variables with p < 0.10 in the univariate analysis (severe OSA and elevated NBL1-Ab levels), demonstrated an independent association of only elevated NBL1-Ab level with CAD (OR: 6.75, 95% CI: 1.63–34.9, p = 0.008). We also performed additional multivariate logistic regression analysis with three covariates; two same covariates in the multivariate analysis above and age, which was significantly associated with NBL1-Ab levels in Fig 2. The result did not change, and only elevated NBL1-Ab level was found to be associated with CAD (data not shown).

10.1371/journal.pone.0195015.t002 Logistic regression of the prevalence of CAD in patients with OSA (n = 82; no. of events = 10).
Univariate Analysis Multivariate Analysis
OR 95% CI p value OR 95% CI p value
Age(per year) 1.04 0.97–1.11 0.26
Obesity(BMI ≥ 25, kg/m2) 0.80 0.21–3.36 0.75
Smoking(Current or Ex-smoker) 2.76 0.71–13.6 0.15
Hypertension 0.71 0.14–2.81 0.64
Diabetes 3.03 0.69–12.2 0.13
Hyperlipidemia 1.20 0.24–4.79 0.81
Stroke 5.75 0.68–40.2 0.10
Severe OSA(AHI > 30) 5.40 0.94–102.3 0.06 3.63 0.56–71.1 0.19
Elevated NBL1-Ab levels(≥ 27512)† 8.17 2.03–41.4 0.003 6.75 1.63–34.9 0.008

†NBL1-Ab cut-off was 27512 based on ROC curve analysis.

AHI, apnea-hypopnea index; BMI, body mass index; CAD, coronary artery disease; CI, confidence interval; NBL1-Abs, autoantibodies against NBL1; OSA, obstructive sleep apnea; OR, odds ratio; ROC, receiver operating characteristic.

 10.1371/journal.pone.0195015.g003 ROC curve analysis for the predictive value of NBL1-Abs for CAD among all OSA patients.

The area under the curve was 0.739 (95% CI: 0.560–0.863). When the NBL1-Ab cut-off was 27512, the sensitivity of NBL1-Abs for predicting CAD was 70.0% and the specificity was 77.8% (n = 82; no. of events = 10). CAD, coronary artery disease; CI, confidence interval; NBL1-Abs, autoantibodies against NBL1; OSA, obstructive sleep apnea; ROC, receiver operating characteristic.

 Discussion

The major finding of this study was that NBL1-Ab level was significantly elevated in patients with both OSA and ACS compared with HVs. Subgroup analyses showed that NBL1-Ab level was significantly higher in patients with severe OSA and OSA patients with a past history of CAD. Weak associations were observed between NBL1-Ab level and age, AHI, mean SpO2 and arousal index, whereas a significantly increased NBL1-Ab level was observed in OSA patients with CAD compared with those without CAD.

Regarding circulating autoantibodies for patients with atherosclerosis-related diseases, some antibodies have been previously reported as markers, including autoantibodies against phospholipids in patients with ACS [18], apolipoprotein A-1 in patients with increased atherosclerotic plaque vulnerability [19] and oxidized LDL (ox-LDL), which is associated with plaque formation and coronary risk, in some patients with systemic lupus erythematosus [20]. We have previously identified several atherosclerosis antigens serologically using recombinant cDNA expression cloning, and found a close association of the levels of antibodies against several autoantigens (e.g. RPA2, ATP2B4 and BMP1) with atherosclerosis-related diseases and risk factors of atherosclerosis-related diseases such as stroke, hypertension and smoking habits [10, 15]. Based on these observations, we hypothesized that certain antibody markers have association with past CAD in patients with OSA, which was confirmed in this present study.

NBL1, a member of the family of differential screening-selected gene aberrative in neuroblastoma (DAN) [21], is associated with not only neuroblastoma but also with remodelling of vessels by inhibiting pulmonary arterial smooth muscle cell proliferation [22]. One of the main roles of DAN family members including NBL1 is to act as specific antagonists of BMPs. BMPs have initially been found to induce bone formation and later identified as a multifunctional cytokine associated with anaemia, progressive bone formation and various cancers [23]. Among BMP family members, BMP2 and BMP4 are mediators of endothelial inflammation in response to shear stress, oxidative stress and pro-inflammatory cytokines [24, 25]. Clinically, serum BMP2 has been correlated with plaque burden and coronary artery calcification in patients with type 2 diabetes [26]. BMP4 has been shown to stimulate vascular endothelial growth factor synthesis [27]. Infusion of BMP4 induced vascular nicotinamide adenine dinucleotide phosphate oxidases and impaired vasodilation, leading to hypertension [28]. As NBL1 is a specific antagonist of BMP2 and BMP4, both of which promote atherosclerosis, NBL1 is capable to inhibit atherosclerosis. As NBL1 would be elevated concomitantly with BMP in atherosclerotic lesions, its autoantibodies (NBL1-Abs) may increase subsequently. Indeed, ox-LDL stimulation induced an increase in BMP2 in aortic valve interstitial cells, which corresponds with the association of LDL and NBL1-Abs in this study [29]. Therefore, NBL1-Ab may have the potential to be an optional marker of atherosclerosis in addition to other conventional investigations. However, a prospective study of patient groups without overlap is required to confirm the utility of NBL1-Ab, because ACS patients with OSA were excluded only by patient histories in this study, and thus, some patients with ACS may have had undiagnosed OSA.

Considering the significance of elevated NBL1-Abs in this study, a comparison between NBL1-Abs in OSA and ACS should be performed. Similar to all patients with ACS, patients with AMI without a past history of CAD (i.e. first episode of AMI) also had higher NBL1-Ab levels than HVs (p = 0.002). Because it generally takes several weeks for the production of IgG antibodies in response to antigen exposure and because sera were collected from patients when AMI was diagnosed, NBL1-Abs might be produced before the onset of AMI. Additionally, NBL1-Ab levels were significantly higher in OSA patients with a history of CAD than in those without a history of CAD. Accordingly, it appears that NBL1-Abs might have potential to be used as a biomarker of a past history of CAD and also as a predictive biomarker of CAD in patients with OSA. Among 72 OSA patients without a history of CAD, 16 patients in the present study showed elevated NBL1-Ab levels (>cut-off value, data are not shown). From these results, it is indicated that the 16 patients may require more clinical attention (e.g. increased adherence to CPAP therapy). Regarding the practical utility for evaluation of atherosclerosis in patients with OSA, coronary artery calcification, carotid IMT, baPWV and flow-mediated dilation are often used [30, 31]. However, the autoantibody assay described in this study is non-invasive and more convenient (only requires a serum sample) than the existing tests. Therefore, monitoring circulating NBL1-Ab levels might be useful for evaluating the risk of lethal events in patients with OSA; however, prospective cohort studies are needed to confirm its effectiveness as a marker for atherosclerosis risk.

This study has several limitations. First, ACS patients with sleep-related breathing disorders were excluded only by evidence obtained from patient histories not by PSG or home sleep tests. Similarly, the history of OSA in HVs were excluded by asking if they had any kind of diseases or medical examination history and not by conducting specific tests (e.g. PSG or home sleep tests) or by questioning clinical symptoms of OSA (e.g. snoring, daytime sleepiness). Consequently, some patients with ACS and HVs may have had sleep-related breathing disorders. Second, the number of OSA patients with CAD was relatively small. It is possible that the NBL1-Ab cut-off for the prevalence of CAD in all patients with OSA was inaccurate. AUC for predicting the prevalence of CAD history was 0.738, indicating that the ROC curve analysis had only fair utility. In addition, all patients with CAD were male. Therefore, we could not include gender as a covariate in the logistic regression analysis. Third, we did not evaluate atherosclerosis by conventional physiological tests. Finally, the analyses were not adjusted for potential confounding factors between patients and control HVs (e.g. age, BMI and history of atherosclerosis-related diseases).

In conclusion, patients with OSA had higher circulating NBL1-Ab levels than HVs, as was the case for patients with ACS. Serum NBL1-Ab levels were significantly increased in OSA patients with a history of CAD and severe disease. Further study is warranted to confirm whether elevated NBL1-Ab levels indicate increased risk of future CAD in patients with OSA.

 Supporting information Whole raw data sheet.

(XLSX)

The authors thank Dr. Shigeru Sakiyama (Japanese Red Cross Chiba Blood Center) and Dr. Hiroki Nagase (Chiba Cancer Center Research Institute) for their valuable suggestion.

 References Dewan NA, Nieto FJ, Somers VK. Intermittent hypoxemia and OSA: implications for comorbidities. Chest. 2015;147(1):266–74. Epub 2015/01/07. doi: 10.1378/chest.14-0500 25560865. Lee MY, Wang Y, Mak JC, Ip MS. Intermittent hypoxia induces NF-kappaB-dependent endothelial activation via adipocyte-derived mediators. Am J Physiol Cell Physiol. 2016;310(6):C446–55. Epub 2016/01/08. doi: 10.1152/ajpcell.00240.2015 26739492. Yaggi HK, Concato J, Kernan WN, Lichtman JH, Brass LM, Mohsenin V. Obstructive sleep apnea as a risk factor for stroke and death. N Engl J Med. 2005;353(19):2034–41. Epub 2005/11/12. doi: 10.1056/NEJMoa043104 16282178. Marin JM, Carrizo SJ, Vicente E, Agusti AG. Long-term cardiovascular outcomes in men with obstructive sleep apnoea-hypopnoea with or without treatment with continuous positive airway pressure: an observational study. Lancet. 2005;365(9464):1046–53. Epub 2005/03/23. 15781100. Young T, Finn L, Peppard PE, Szklo-Coxe M, Austin D, Nieto FJ, et al. Sleep disordered breathing and mortality: eighteen-year follow-up of the Wisconsin sleep cohort. Sleep. 2008;31(8):1071–8. Epub 2008/08/22. 18714778. Xie X, Pan L, Ren D, Du C, Guo Y. Effects of continuous positive airway pressure therapy on systemic inflammation in obstructive sleep apnea: a meta-analysis. Sleep Med. 2013;14(11):1139–50. Epub 2013/09/24. doi: 10.1016/j.sleep.2013.07.006 24054505. Jullian-Desayes I, Joyeux-Faure M, Tamisier R, Launois S, Borel AL, Levy P, et al. Impact of obstructive sleep apnea treatment by continuous positive airway pressure on cardiometabolic biomarkers: a systematic review from sham CPAP randomized controlled trials. Sleep Med Rev. 2015;21:23–38. Epub 2014/09/16. doi: 10.1016/j.smrv.2014.07.004 25220580. Guo J, Sun Y, Xue LJ, Huang ZY, Wang YS, Zhang L, et al. Effect of CPAP therapy on cardiovascular events and mortality in patients with obstructive sleep apnea: a meta-analysis. Sleep Breath. 2016;20(3):965–74. Epub 2016/02/14. doi: 10.1007/s11325-016-1319-y 26873722. Wozniak DR, Lasserson TJ, Smith I. Educational, supportive and behavioural interventions to improve usage of continuous positive airway pressure machines in adults with obstructive sleep apnoea. Cochrane Database Syst Rev. 2014;1:Cd007736. Epub 2014/01/09. doi: 10.1002/14651858.CD007736.pub2 24399660. Machida T, Kubota M, Kobayashi E, Iwadate Y, Saeki N, Yamaura A, et al. Identification of stroke-associated-antigens via screening of recombinant proteins from the human expression cDNA library (SEREX). J Transl Med. 2015;13:71. Epub 2015/04/19. doi: 10.1186/s12967-015-0393-4 25890248. Goto K, Sugiyama T, Matsumura R, Zhang X-M, Kimura R, Taira A, et al. Identification of cerebral infarction-specific antibody markers from autoantibodies detected in patients with systemic lupus erythematosus. J Mol Biomark Diagn. 2015;6(2). Matsumura T, Terada J, Kinoshita T, Sakurai Y, Yahaba M, Ema R, et al. Circulating Anti-Coatomer Protein Complex Subunit Epsilon (COPE) Autoantibodies as a Potential Biomarker for Cardiovascular and Cerebrovascular Events in Patients with Obstructive Sleep Apnea. J Clin Sleep Med. 2017;13(3):393–400. Epub 2016/12/08. doi: 10.5664/jcsm.6488 27923433. Morrell NW, Bloch DB, ten Dijke P, Goumans MJ, Hata A, Smith J, et al. Targeting BMP signalling in cardiovascular disease and anaemia. Nat Rev Cardiol. 2016;13(2):106–20. Epub 2015/10/16. doi: 10.1038/nrcardio.2015.156 26461965. Garrigue-Antar L, Francois V, Kadler KE. Deletion of epidermal growth factor-like domains converts mammalian tolloid into a chordinase and effective procollagen C-proteinase. J Biol Chem. 2004;279(48):49835–41. Epub 2004/09/24. doi: 10.1074/jbc.M408134200 15381708. Hiwasa T, Machida T, Zhang X-M, Kimura R, Wang H, Iwase K, et al. Elevated levels of autoantibodies against ATP2B4 and BMP-1 in sera of patients with atherosclerosis-related diseases. Immunome Res. 2015;11(2). doi: 10.4172/1745-7580.1000097 Hung WT, Wu FJ, Wang CJ, Luo CW. DAN (NBL1) specifically antagonizes BMP2 and BMP4 and modulates the actions of GDF9, BMP2, and BMP4 in the rat ovary. Biol Reprod. 2012;86(5):158, 1–9. Epub 2012/02/24. doi: 10.1095/biolreprod.111.096172 22357543. Dionne MS, Skarnes WC, Harland RM. Mutation and analysis of Dan, the founding member of the Dan family of transforming growth factor beta antagonists. Mol Cell Biol. 2001;21(2):636–43. Epub 2001/01/03. doi: 10.1128/MCB.21.2.636-643.2001 11134349. Veres K, Lakos G, Kerenyi A, Szekanecz Z, Szegedi G, Shoenfeld Y, et al. Antiphospholipid antibodies in acute coronary syndrome. Lupus. 2004;13(6):423–7. Epub 2004/08/12. doi: 10.1191/0961203304lu1011oa 15303568. Montecucco F, Vuilleumier N, Pagano S, Lenglet S, Bertolotto M, Braunersreuther V, et al. Anti-Apolipoprotein A-1 auto-antibodies are active mediators of atherosclerotic plaque vulnerability. Eur Heart J. 2011;32(4):412–21. Epub 2011/01/13. doi: 10.1093/eurheartj/ehq521 21224292. Fesmire J, Wolfson-Reichlin M, Reichlin M. Effects of autoimmune antibodies anti-lipoprotein lipase, anti-low density lipoprotein, and anti-oxidized low density lipoprotein on lipid metabolism and atherosclerosis in systemic lupus erythematosus. Revista Brasileira de Reumatologia. 2010;50(5):539–51. Epub 2010/12/03. 21125190. Ozaki T, Sakiyama S. Tumor-suppressive activity of N03 gene product in v-src-transformed rat 3Y1 fibroblasts. Cancer Res. 1994;54(3):646–8. Epub 1994/02/01. 8306325. Cui C, Zhang H, Guo LN, Zhang X, Meng L, Pan X, et al. Inhibitory effect of NBL1 on PDGF-BB-induced human PASMC proliferation through blockade of PDGFbeta-p38MAPK pathway. Biosci Rep. 2016;36(4):e00374. Epub 2016/07/31. doi: 10.1042/BSR20160199 27474499. Hopkins CR. Inhibitors of the bone morphogenetic protein (BMP) signaling pathway: a patent review (2008–2015). Expert Opin Ther Pat. 2016;26(10):1115–28. Epub 2016/08/02. doi: 10.1080/13543776.2016.1217330 27476794. Csiszar A, Smith KE, Koller A, Kaley G, Edwards JG, Ungvari Z. Regulation of bone morphogenetic protein-2 expression in endothelial cells: role of nuclear factor-kappaB activation by tumor necrosis factor-alpha, H2O2, and high intravascular pressure. Circulation. 2005;111(18):2364–72. Epub 2005/04/27. doi: 10.1161/01.CIR.0000164201.40634.1D 15851600. Sorescu GP, Song H, Tressel SL, Hwang J, Dikalov S, Smith DA, et al. Bone morphogenic protein 4 produced in endothelial cells by oscillatory shear stress induces monocyte adhesion by stimulating reactive oxygen species production from a nox1-based NADPH oxidase. Circ Res. 2004;95(8):773–9. Epub 2004/09/25. doi: 10.1161/01.RES.0000145728.22878.45 15388638. Zhang M, Sara JD, Wang FL, Liu LP, Su LX, Zhe J, et al. Increased plasma BMP-2 levels are associated with atherosclerosis burden and coronary calcification in type 2 diabetic patients. Cardiovasc Diabetol. 2015;14:64. Epub 2015/05/25. doi: 10.1186/s12933-015-0214-3 26003174. Kozawa O, Matsuno H, Uematsu T. Involvement of p70 S6 kinase in bone morphogenetic protein signaling: vascular endothelial growth factor synthesis by bone morphogenetic protein-4 in osteoblasts. J Cell Biochem. 2001;81(3):430–6. Epub 2001/03/20. 11255225. Miriyala S, Gongora Nieto MC, Mingone C, Smith D, Dikalov S, Harrison DG, et al. Bone morphogenic protein-4 induces hypertension in mice: role of noggin, vascular NADPH oxidases, and impaired vasorelaxation. Circulation. 2006;113(24):2818–25. Epub 2006/06/14. doi: 10.1161/CIRCULATIONAHA.106.611822 16769910. Nadlonek NA, Lee JH, Weyant MJ, Meng X, Fullerton DA. ox-LDL induces PiT-1 expression in human aortic valve interstitial cells. J Surg Res. 2013;184(1):6–9. Epub 2013/07/16. doi: 10.1016/j.jss.2013.05.001 23849774. Ali SS, Oni ET, Warraich HJ, Blaha MJ, Blumenthal RS, Karim A, et al. Systematic review on noninvasive assessment of subclinical cardiovascular disease in obstructive sleep apnea: new kid on the block! Sleep Med Rev. 2014;18(5):379–91. Epub 2014/03/22. doi: 10.1016/j.smrv.2014.01.004 24650521. Drager LF, Bortolotto LA, Lorenzi MC, Figueiredo AC, Krieger EM, Lorenzi-Filho G. Early signs of atherosclerosis in obstructive sleep apnea. Am J Respir Crit Care Med. 2005;172(5):613–8. doi: 10.1164/rccm.200503-340OC 15901608.