Circulating Anti-Sorting Nexins 16 Antibodies as an Emerging Biomarker of Coronary Artery Disease in Patients with Obstructive Sleep Apnea

Biomarkers are not available for monitoring the onset and progression of coronary artery disease (CAD) in patients with obstructive sleep apnea (OSA), a major risk factor for arteriosclerotic cardiovascular diseases. This study aimed to test for correlation between circulating anti-Sorting Nexins 16 antibody (SNX16-Ab) levels, CAD history and clinical parameters of patients with OSA. Sixty-four healthy donors, 82 adults with OSA, and 96 with acute coronary syndrome (ACS) were studied. Serum samples were collected at diagnostic polysomnography in the OSA group or at the disease onset in the ACS group. Serum SNX16-Ab levels were measured by amplified luminescence proximity homogeneous assay (AlphaLISA), and correlation between SNX16-Ab levels and clinical parameters was analyzed. SNX16-Ab levels and apnea-hypopnea index (AHI) were weakly correlated. Additionally, logistic regression analyses of OSA group identified that elevated SNX16-Ab level associated with the history of CAD. Circulating SNX16-Ab could increase during CAD pathogenesis in patients with OSA. Further prospective studies are required to prove the predictive potential of SNX16-Ab level in CAD onset of patients with OSA.


Introduction
Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder characterized by recurrent upper airway obstruction, intermittent hypoxia, frequent sleep fragmentation, and intra-thoracic pressure swings during sleep. These nocturnal respiratory disturbances potentially impact the hormone-metabolic, hemodynamic, oxidative, and immuno-inflammatory mechanisms.

Clinical Data
Clinical data including age, sex, BMI, hypertension, diabetes, dyslipidemia, smoking status, CAD, and stroke were collected from clinical records. Hypertension, diabetes, and dyslipidemia were defined as the history of those diagnoses or the use of drugs for those diseases. People were divided into three groups based on their smoking history: never smokers, ex-smokers, and current smokers. PSG admission was scored based on the 2007 American Academy of Sleep Medicine alternative criteria [20]. OSA was diagnosed according to the third edition of the International Classification of Sleep Disorders. OSA severity was classified based on AHI values; mild, 5-15; moderate, 15-30, and severe, > 30.

Blood Sample Collection and Experimental Method
Serum samples were collected from OSA group during PSG admission, HA group during medical checkup, and ACS at diagnosis of UAP or AMI during admission for coronary artery bypass graft or percutaneous coronary intervention. The samples were centrifuged at 3000 g for 10 min at room temperature and stored at −80 • C. A full-length SNX-16 cDNA was expressed using an expression vector pGEX-4T-3 for the glutathione-S-transferase (GST) -tagged SNX-16 protein. The product of the gene was purified as previously described [18,21,22]. AlphaLISA (PerkinElmer, Waltham, MA, USA) was conducted in 384-well microtiter plates (PerkinElmer) containing 2.5 µL of GST or a GST-fusion protein (10 µg/mL) in AlphaLISA buffer (25 mM HEPES, pH 7.4, 0.05% proclin 300, 1 mg/mL dextran 500, 0.1% casein, and 0.5% Triton X-100) and 2.5 µL of 1/100-diluted sera. The mixture was incubated at room temperature for 1 to 14 days. 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 at room temperature in the dark for 14 days. The chemical emission was measured with an EnSpire Alpha microplate reader (PerkinElmer, Waltham, MA, USA) as described previously [21,22]. The target antibody level was measured by subtracting the alpha value of GST control sample from the alpha value of the sample containing GST fusion protein.

Statistical Analysis
All statistical analyses were performed with JMP Pro 12.2.0 software (SAS Institute Inc., Cary, NC, USA). The significance of differences in baseline characteristics between groups was analyzed using the Kruskal-Wallis test for categorical data and Mann-Whitney U or Kruskal-Wallis test for numerical data. The significance of differences among HA, mild OSA, moderate OSA, and severe OSA group was analyzed using the Steel-Dwass test as a post hoc analysis. Subgroup analyses were performed for SNX16-Ab levels by OSA severity or the history of CAD. Correlation of SNX16-Ab level and clinical data of OSA group was evaluated using Spearman correlation analysis. The SNX16-Ab level cut-off value for the history of CAD in OSA group was calculated by ROC curve analysis to maximize the sum of specificity and sensitivity. Multivariate and univariate logistic regression analysis was used to identify variables that could predict patients with the history of CAD. Statistical significance was defined as a p value < 0.05, and all tests were two-sided.

Ethics Approval and Consent to Participate
All experiments were performed with the approval of the Animal Experiment Ethics Committee of the Murayama Medical Center and in accordance with the Guiding Principles for the Care and Use of Animals of the Physiological Society of Japan.

Availability of Data and Material
The datasets generated during the current study are available from the corresponding author on reasonable request.

Clinical Characteristics
Characteristics of the healthy adults (HA), OSA, and acute coronary syndrome (ACS) group are shown in Table 1. The OSA and ACS group were significantly older than the HA group. ACS group included more male patients than HA group. The history of CAD, diabetes, dyslipidemia, and hypertension was more frequently observed in the ACS and OSA group than in the HA group.

Difference in SNX16-Ab Level for Each Group
After screening the serum of patients with OSA for multiple candidates of autoantigens recognized by IgG antibodies using protein arrays, we selected and identified SNX16-Abs that were elevated. As shown in Figure 1A, the serum levels of SNX16-Ab in the OSA and ACS group were significantly higher than those in HA group. SNX16-Ab levels between the OSA group and ACS group were not significantly different. (p = 0.9314). SNX16-Ab levels in patients with severe OSA or the history of CAD were significantly higher than those in HA group (p < 0.001, Figure 1B,C). The SNX16-Ab levels in OSA group with no history of CAD or with mild-to-moderate severity were similar to those in HA group. The level of SNX16-Ab, measured by AlphaLISA, was compared among the three groups; OSA, ACS, and HA group. (a) Kruskal-Wallis test revealed significant differences among the three groups (p < 0.001). Steel-Dwass test revealed significant differences between patients with ACS or OSA versus HA group. (b) Significant differences were observed between patients with severe OSA group and HA group and (c) between patients with OSA with CAD versus HA group. Horizontal lines represent medians. After performing the Kruskal-Wallis test, the Steel-Dwass test was performed on all pairs; p values of Steel-Dwass test were described. Since the number of groups to be analyzed was different, the p values were different even between the same groups. ACS: acute coronary syndrome; CAD: coronary artery disease; HA: healthy adults; OSA: obstructive sleep apnea; SNX16-Abs: antibodies against SNX16; w/o: without.

Correlation of SNX16-Ab Level and Clinical Data of OSA Group
The association between SNX16-Ab levels and the clinical data of the OSA group and ACS group are shown in Figure 2 and supplementary Figure S1. Weak but significant correlation was observed between SNX16-Ab level and apnea-hypopnea index (ρ = 0.32, p = 0.003, Figure 2c). A very weak significant association was observed between SNX16-Ab level and body mass index (BMI) (ρ = 0.24, p = 0.003, Figure 2b), mean SpO2 (ρ = −0.29, p = 0.008, Figure 2d) and arousal index (ρ = 0.28, p = 0.011, Figure 2e). A significantly elevated SNX16-Ab levels were found in patients with CAD than in those without CAD (p = 0.008, Figure 2k). No significant association was observed between SNX16-Ab levels and age, sex, smoking status, history of diabetes, hypertension, dyslipidemia, and stroke. However, univariate analysis of association between the SNX16-Ab level and each factor in ACS group suggests smoking status as an influencing factor (supplementary Figure S1). Multiple regression analysis suggests correlation between smoking status and SNX16-Ab level in ACS group (p = 0.036), not in OSA group (p = 0.19). Kruskal-Wallis test revealed significant differences among the three groups (p < 0.001). Steel-Dwass test revealed significant differences between patients with ACS or OSA versus HA group. (B) Significant differences were observed between patients with severe OSA group and HA group and (C) between patients with OSA with CAD versus HA group. Horizontal lines represent medians. After performing the Kruskal-Wallis test, the Steel-Dwass test was performed on all pairs; p values of Steel-Dwass test were described. Since the number of groups to be analyzed was different, the p values were different even between the same groups. ACS: acute coronary syndrome; CAD: coronary artery disease; HA: healthy adults; OSA: obstructive sleep apnea; SNX16-Abs: antibodies against SNX16; w/o: without.

Correlation of SNX16-Ab Level and Clinical Data of OSA Group
The association between SNX16-Ab levels and the clinical data of the OSA group and ACS group are shown in Figure 2 and Supplementary Figure S1. Weak but significant correlation was observed between SNX16-Ab level and apnea-hypopnea index (ρ = 0.32, p = 0.003, Figure 2c). A very weak significant association was observed between SNX16-Ab level and body mass index (BMI) (ρ = 0.24, p = 0.003, Figure 2b), mean SpO 2 (ρ = −0.29, p = 0.008, Figure 2d) and arousal index (ρ = 0.28, p = 0.011, Figure 2e). A significantly elevated SNX16-Ab levels were found in patients with CAD than in those without CAD (p = 0.008, Figure 2k). No significant association was observed between SNX16-Ab levels and age, sex, smoking status, history of diabetes, hypertension, dyslipidemia, and stroke. However, univariate analysis of association between the SNX16-Ab level and each factor in ACS group suggests smoking status as an influencing factor (Supplementary Figure S1). Multiple regression analysis suggests correlation between smoking status and SNX16-Ab level in ACS group (p = 0.036), not in OSA group (p = 0.19).
We used univariate and multivariate logistic regression models to clarify the strength of the correlation between the history of CAD and clinical parameters in the OSA group, as shown in Table 2. An optimal cutoff value of SNX16-Ab for predicting the presence of the history of CAD in the OSA group was 59,735 by receiver operating characteristic (ROC) curve analysis, with a sensitivity of 71.4%, a specificity of 74.6%, area under the curve of 0.712 (Figure 3). The results of applying the SNX16-Ab cut-off value in each group are as shown in Supplementary Table S1. The univariate logistic regression clarified a correlation between the history of CAD and elevated SNX16-Ab level using the SNX16-Ab cut-off value [odds ratio (OR): 8.87, 95% confidence interval (CI): 2.19-45.1, p = 0.002]. The multivariate logistic regression analysis using parameters with p < 0.10 by univariate analysis (elevated SNX16-Ab level and severe OSA) revealed significant correlation between elevated SNX16-Ab level and the history of CAD (OR: 8.61, 95% CI: 2.07-45.0, p = 0.003). We used univariate and multivariate logistic regression models to clarify the strength of the correlation between the history of CAD and clinical parameters in the OSA group, as shown in Table  2. An optimal cutoff value of SNX16-Ab for predicting the presence of the history of CAD in the OSA group was 59,735 by receiver operating characteristic (ROC) curve analysis, with a sensitivity of 71.4%, a specificity of 74.6%, area under the curve of 0.712 (Figure 3). The results of applying the SNX16-Ab cut-off value in each group are as shown in supplementary Table S1. The univariate logistic regression clarified a correlation between the history of CAD and elevated SNX16-Ab level using the SNX16-Ab cut-off value [odds ratio (OR): 8.87, 95% confidence interval (CI): 2.19-45.1, p = 0.002]. The multivariate logistic regression analysis using parameters with p < 0.10 by univariate analysis (elevated SNX16-Ab level and severe OSA) revealed significant correlation between elevated SNX16-Ab level and the history of CAD (OR: 8.61, 95% CI: 2.07-45.0, p = 0.003).

Discussion
This study presents two major findings of circulating SNX16-Ab in patients with OSA. First, serum levels of SNX16-Ab are significantly higher in OSA and ACS group compared to those in HA (control) group. Second, subgroup analyses show especially elevated serum levels of SNX16-Ab in severe OSA group and OSA with the history of CAD group. Thus, our results suggest that serum SNX16-Ab may increase in connection with CAD pathogenesis in patients with preexisting OSA.
The sorting nexins (SNXs), a family of cytoplasmic and membrane-associated proteins, are responsible for sorting membrane proteins [23]. Even though a clear evidence between SNXs-CAD pathogenesis-OSA is lacking, potential mechanisms might be considered. For instance, epidermal growth factor receptor (EGFR), one of other sorting membrane proteins, is expressed in endothelial cells, vascular smooth muscle cells, cardiomyocytes and macrophages [24], and causes downstream activation of transcription factors such as nuclear factor-κB. It also stimulates pro-inflammatory gene transcription in macrophages, playing a role in foam cell transformation, cellular dysfunction, and proliferation of vascular smooth muscle cells [25]. On the other hand, SNX16 is localized in the early

Discussion
This study presents two major findings of circulating SNX16-Ab in patients with OSA. First, serum levels of SNX16-Ab are significantly higher in OSA and ACS group compared to those in HA (control) group. Second, subgroup analyses show especially elevated serum levels of SNX16-Ab in severe OSA group and OSA with the history of CAD group. Thus, our results suggest that serum SNX16-Ab may increase in connection with CAD pathogenesis in patients with preexisting OSA.
The sorting nexins (SNXs), a family of cytoplasmic and membrane-associated proteins, are responsible for sorting membrane proteins [23]. Even though a clear evidence between SNXs-CAD pathogenesis-OSA is lacking, potential mechanisms might be considered. For instance, epidermal growth factor receptor (EGFR), one of other sorting membrane proteins, is expressed in endothelial cells, vascular smooth muscle cells, cardiomyocytes and macrophages [24], and causes downstream activation of transcription factors such as nuclear factor-κB. It also stimulates pro-inflammatory gene transcription in macrophages, playing a role in foam cell transformation, cellular dysfunction, and proliferation of vascular smooth muscle cells [25]. On the other hand, SNX16 is localized in the early and recycling endosomes via its Phox domain, and it may negatively regulate EGFR-mediated signaling by potentiating the degradation of EGFR [26]. In addition, SNX16 overexpression in late endosomes is reported to cause cholesterol accumulation. Generally, low-density lipoproteins (LDL)-derived cholesterol taken up via the LDL receptor is transported to the cell membrane, endoplasmic reticulum, and mitochondria by endocytosis. However, SNX16 overexpression in late endosomes inhibit cholesterol transport, accumulate cholesterol, and may cause arteriosclerosis. [27] Preceding cardiovascular disease or stroke, atheromatous plaques are partially ruptured, leaking accumulated SNX16 into the blood. Subsequently, autoantibody against SNX16 may be produced by such repetitive SNX16 leakage. In this study, elevated SNX-16 Ab correlated with the history of CAD and OSA parameters such as frequent arousal, intermittent hypoxia (i.e., AHI), and obesity (i.e., BMI). However, since these results and the current literature do not reveal a direct molecular or clinical link between SNX16 and arteriosclerosis in patients with OSA, this question remains to be explored.
In regard to autoantibody production, it generally takes several weeks to produce IgG antibodies after antigen exposure. In this study, the sera were collected from the ACS group at the onset of acute Diagnostics 2020, 10, 71 8 of 11 myocardial infarction (AMI) diagnosis. Therefore, SNX16-Ab might have been produced several weeks before the onset of AMI. This suggests that SNX16-Ab level could be useful in predicting the onset of AMI.
Recently, several potential biomarkers for evaluating arteriosclerosis have been reported. Oxidized low-density lipoproteins (oxLDL) are reported to be related to the progression of arteriosclerosis [28]. Circulating anti-oxLDL antibodies and oxLDL may be optimal biomarkers for cardiovascular disease risk, however, their true significance remains uncertain [29]. Although increased levels of autoantibody against heat shock protein (Hsp) family are reported in patients with arteriosclerotic diseases [30,31], their ability to predict the occurrence of arteriosclerotic diseases remains unclear. Sun et al. [32] reported that higher circulating ESM-1, previously named endothelial cell-specific molecule-1, correlated with the presence of CAD in patients with OSA. Furthermore, several studies demonstrated that the level of placental growth factor (PIGF), as a biomarker of ACS, was significantly higher in patients with OSA [33]. Thus, although a single marker might not be sufficient as a predictive biomarker of ACS in patients with OSA, a combination of several biomarkers including SNX-16 Ab may have a predictive potential in future clinical practice.
This study includes several limitations. First, the existence of OSA in patients with ACS or in HA group was excluded only by medical history and not by diagnostic testing such as polysomnography (PSG). Subclinical OSA was not completely excluded in the ACS or HA group. Additionally, although the analysis in the OSA group has denied the correlation between age, sex, smoking status, hypertension, diabetes, dyslipidemia and SNX16-Ab level (see Figure 2), potential confounding factors in between the OSA, ACS and HA group were not completely excluded. Second, the SNX16-Ab cutoff for the history of CAD in the OSA group might not be accurate, since the number of populations with OSA and CAD was small. Third, sex was not included as a covariate in the logistic regression analysis, because all patients with CAD in the OSA group were male. Forth, the severity of atherosclerosis was not evaluated by conventional physiological tests. Fifth, potential confounding factors between the patients and HAs were not adjusted. Finally, only Japanese patients participated in this study. Further studies are needed in other ethnic groups.
In summary, our subjects with OSA had higher SNX16-Ab levels than healthy subjects. SNX16-Ab levels were significantly elevated in the OSA group with the history of CAD or severe OSA compared to HA group. Based on our results, we suggest that SNX16-Ab could increase in connection with the pathogenesis of CAD in patients with OSA. However, additional prospective studies are warranted to prove the predictive potential of SNX16-Ab level for the onset of CAD in the OSA group. The results of this study might evoke further studies to research biomarkers involved in the devastating progression of cardiovascular diseases in OSA.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2075-4418/10/2/71/s1, Figure S1: Association between serum SNX16-Ab level and clinical data in ACS group. (a) Association between SNX16-Ab level and smoking status, (b) Hypertension, (c) Diabetes, and (d) Dyslipidemia were analyzed. Mann-Whitney U test (a-d) were used. Horizontal lines represent medians. SNX16-Ab: antibodies against SNX16. Table S1: SNX16-Ab level of each group. Funding: 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 Development (AMED, https://www.amed.go.jp), Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant number 19K08596, 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.