Macrophage Mannose Receptor CD206 Predicts Prognosis in Community-acquired Pneumonia

CD206, a mannose receptor, is mainly expressed on the surface of alternatively activated macrophages where it acts as a pattern recognition receptor and plays a role in innate and adaptive immunity. This study investigated serum soluble CD206 (sCD206) levels in community-acquired pneumonia (CAP) and examined their clinical significance. sCD206 concentrations were measured in the sera of two independent cohorts with CAP (127 and 125 patients, respectively) and 42 controls. The expression of CD206 in the lung from autopsied cases was also examined. Patients with CAP showed significantly elevated sCD206 levels than did the controls (p < 0.0001). Notably, fatal CAP patients had more than two-fold higher sCD206 concentrations than survivors in both cohorts (p < 0.0001). Serum sCD206 concentrations were associated with Pneumonia Severity Index (PSI) and CURB-65 values. Importantly, even fatal CAP patients classified as PSI I-IV, CURB65 0–2 or age <75 years had comparatively higher levels of sCD206 than those classified as PSI V, CURB-65 3–5 or age ≥75 years. Immunohistochemically, the infiltration of CD206+ macrophages was found in the lungs of fatal cases. Elevated levels of sCD206 are associated with CAP prognosis, suggesting sCD206 might be a potential biomarker to predict severity for CAP.

Serum concentrations of sCD206 in Cohort1. The serum concentrations of sCD206 in CAP patients in Cohort1 are presented in Fig. 1 Correlations between sCD206 and disease severity or clinical parameters. We next evaluated serum sCD206 levels according to CAP severity using PSI and CURB-65 scales. As shown in Fig. 2A,B, serum concentrations of sCD206 were increased according to severity of the CAP index. Additionally, correlation analyses showed that sCD206 was positively correlated with PSI score and CUB-65 score, as well as BUN and CRP levels, but was negatively associated with PaO 2 /FiO 2 ratio (P/F ratio) and albumin levels (Supplement Table 2).
sCD206 levels in patients and non-survivors according to disease severity and age. Although a significantly higher mortality rate was found for CAP patients classified into PSI V (p = 0.001), CURB-65 3-5 (p = 0.001) and age ≥75 years groups, more than 40% of the non-surviving patients were classified as PSI < V (n = 10, 45.5%), CURB-65 < 3-5 (n = 11, 50%), and age <75 years (n = 10, 45.5%  www.nature.com/scientificreports www.nature.com/scientificreports/ Univariate and multivariate analyses of prognostic factors. We next calculated the optimal cut-off values for sCD206 to predict death using ROC analysis. Based on the ROC analysis, the area under the ROC curve (AUC) was 0.856 for sCD206 (Fig. 4), which was superior to that for CRP, PSI, or CURB65 score (0.674, 0.770, and 0.747, respectively). The AUC for PCT was 0.603 for 78 patient samples. Using the optimal cut-off for serum sCD206 levels (1413 ng/ml), sensitivity and specificity values obtained were 86.4% and 79.1%, respectively, with a likelihood ratio of 4.13. We then assessed the prognosis of patients with CAP based on this cut-off value, using the Kaplan-Meier method and log-rank test. The high-sCD206 (n = 67) group showed significantly lower survival rates than the low-sCD206 (n = 185) group (p < 0.0001; Fig. 5A). Although age-related scoring systems such as PSI and CURB65 are widely validated and confirmed 12,13 , certain situations have been associated with concerns about validity. For example, there are potential risks of overestimating severity in elderly patients and  www.nature.com/scientificreports www.nature.com/scientificreports/ misclassifying younger patients with severe disease into lower classes [14][15][16] . Therefore, we performed a subgroup analysis of sCD206 according to patient age and confirmed that this cut-off worked, regardless of patient age (Fig. 5B,C). Finally, to determine the prognostic value of sCD206 with regard to outcome, we performed logistic regression analyses (Table 3). By univariate analysis and age/sex-adjusted multivariate analyses, sCD206, as well as PSI, CURB-65, and CRP, was found to be a significant predictor of death.

Expression of CD206 in the lung biopsies of patients with CAP.
To investigate the involvement of M2 macrophages and sources of sCD206 in patients with severe CAP, we assessed CD206 expression in autopsied-lung tissues from fatal CAP patients by immunohistochemistry. An autopsied lung specimen from a 75-year-old male with fatal CAP showed dense infiltration of neutrophils and CD206 + macrophages with fibrinous exudates in the alveolar space (Fig. 6). Similarly, the accumulation of CD206-positive macrophages was found in the lungs of the remaining cases (Supplement Figs. 1, 2). Meanwhile, only a few CD206-positive macrophages were found in the alveolar spaces in the resected lung sections from patient with early lung cancer (Supplement Fig. 3).

Discussion
In the present study, we measured the serum sCD206 levels in patients with CAP and evaluated their clinical implications using two independent cohorts. Compared with control subjects, we found that patients with CAP demonstrated a significant increase in serum sCD206 that is associated with disease severity and prognosis. The accumulation of CD206 + macrophages in the alveolar and interstitial spaces was also found in the lungs of autopsied cases. Importantly, we found that sCD206 is a potential surrogate marker to predict CAP outcome.
Macrophages are abundant cells in the lungs and have pivotal roles in first-line defense and the maintenance of homeostasis 8,17 . On the surface of macrophages, CD206 serves as a PPR and functions in pathogen recognition and internalization by binding and scavenging unwanted glycoprotein and glycolipids 10,11 . Notably, CD206 itself can bind polysaccharides from Streptococcus pneumoniae and lipopolysaccharide from Klebsiella pneumoniae subsequently promoting recognition, phagocytosis, and destruction of the bacterial cell 10 . Additionally, although detailed mechanisms underlying the proteolytic cleavage of CD206 are not fully understood, the shedding of CD206 from the membrane is enhanced upon the recognition of several pathogens 18 . The present study showed elevated sCD206 concentrations in accordance with CAP severity and the accumulation of CD206-positive macrophages in the lungs of fatal patients with CAP. These results suggested that increased sCD206 levels represent the activation of CD206 + macrophages during this disease and might partly result from macrophage-pathogen interactions.
A dichotomous approach to macrophage polarization is essential to understand the immune response. M1 macrophages are essential for anti-microbicidal and anti-tumor responses, whereas M2 macrophages have a central role in tissue repair and the resolution of inflammation 19,20 . CD206 is considered a marker of the M2 phenotype [6][7][8][9] . Although the precise role of M2 macrophages in infectious disease with regard to complex M1/ M2 polarization remains unknown, increased expression of CD206 was previously reported in monocytes from patients with sepsis 21,22 . Additionally, the deletion of CD206 + macrophages exacerbates lung injury in endotoxemic mice 23 and during Cryptococcus infection 24 . CD206 −/− mice have been also reported to show increased allergic airway inflammation together with an elevated Th2/Th17 response 25 . This suggests a protective role for M2 macrophages and sCD206 by inhibiting excessive inflammation.
It is well established that PSI and CURB65 predict mortality of CAP patients with high sensitivity and specificity. Meanwhile, it has been reported that these indicators might not be accurate for elderly patients with CAP. As atypical clinical presentations are usually found in elderly patients with CAP, the cut-off values for the scoring system could prove unreliable in these patients [14][15][16] . Although comorbidities are more frequent and the relative risk of mortality is higher in elderly patients, there is also a risk of the misclassification of younger patients with severe disease into lower classes. We previously reported that modified PSI together with performance status can predict mortality more accurately than conventional PSI in patients with CAP ≥80 years of age 15 . Therefore, there is a need for age-independent surrogate markers to assess the severity of CAP. In this study, we demonstrated that sCD206 has the potential to predict mortality in patients with CAP, with a higher AUC value than that obtained   www.nature.com/scientificreports www.nature.com/scientificreports/ for PSI. Additionally, sCD206 values were consistent in both age groups assessed, specifically, <75 years and ≥75 years, suggesting that this approach could complement PSI or CURB65 assessments.
To date, in line with our study on CD206, the utility of macrophage-related markers such as scavenger receptor CD163 and the immunomodulatory molecule IDO have been vigorously investigated in clinical settings with a variety of diseases, as a surrogate marker to estimate disease severity and/or prognosis 4,5,[26][27][28][29][30] . In addition, elevated levels of sCD206 in patients with interstitial lung disease 31 , critical illness 32 , and pneumococcal bacteremia 33 were also reported. Therefore, elevated levels of these molecules including sCD206 are not disease-specific but might also comprehensively represent a whole-body immune response.
The present study had several limitations. First of all, although we examined two independent cohorts of CAP patients, the sample size was still inadequate and univariate and multivariate analyses were not performed with each cohort separately, thus a further validation study is necessary. Therefore, the absolute significance of sCD206 in CAP could not be determined. Second, the results clearly indicate that increased sCD206 is observed in patients with CAP; however, the origin of sCD206-and CD206-expressing monocytes was not investigated. In addition, the role of increased sCD206 levels in the etiology of infectious diseases is not completely elucidated.

Conclusion
In conclusion, the present study demonstrated elevated levels of serum sCD206 in patients with CAP, which is related to disease severity and clinical outcome. Furthermore, the sCD206 was useful for predicting outcome even in younger patients with CAP as well as elderly those with CAP. Our results suggested that sCD206 could be utilized as a novel prognostic marker for CAP and will help for better understanding the role of macrophage polarization in infectious disease.

Methods
Subjects. This prospective study was conducted using two cohorts of patients who were administered for the treatment of CAP in Hamamatsu, Japan. A cohort of 127 consecutive CAP patients admitted to Seirei-Mikatahara Hospital and Hamamatsu University School of Medicine between January 2007 and December 2010, and a cohort of 125 consecutive CAP patients hospitalized at Hamamatsu University School of Medicine between March 2013 and February 2018, were enrolled in this study. The former and latter cohorts were evaluated as a discovery cohort and validation cohort, respectively. Pneumonia was diagnosed according to previously published international guidelines 34,35 . CAP was defined as pneumonia that did not fulfill the criteria of hospital-or nursing and health care-acquired pneumonia in patients with symptoms of acute-onset lower respiratory tract infection, www.nature.com/scientificreports www.nature.com/scientificreports/ who demonstrated new infiltration on a chest radiograph. All patients were stratified into risk classes using the validated prediction rule, calculated according to the PSI and CURB-65 calculator (covering confusion, urea nitrogen, respiratory rate, and blood pressure, ≥65 years of age) 12,13 .
This study also included sera from 42 age and sex-matched subjects (30 men and 12 women, mean age of 73 years) who visited Hamamatsu University Hospital for health checks, as a control group. None of the control subjects had pulmonary infectious disease, as assessed by chest radiographs. This study was approved by the ethics committees of Hamamatsu University School of Medicine, Seirei-Mikatahara Hospital , and was carried out in accordance with approved guidelines. Written informed consent was obtained from all subjects in accordance with institutional guidelines. The study was registered in the University Hospital Medical Information Network in Japan (http://www.umin.ac.jp/. UMIN000003400 and UMIN000019472).
Sample collection. Blood samples were drawn at the time of admission before the start of empirical treatment. Serum samples were frozen at −80 °C until analysis; routine laboratory examinations such as blood cell counts and biochemical analyses were subsequently performed. The serum concentration of sCD206 was determined using an enzyme-linked immunosorbent assay (ELISA) kit (Ray Biotech, Norcross, GA, USA).
Immunohistochemistry. Lung biopsy specimens were obtained from autopsy cases, in which the patient died of severe pneumonia, and resected lung from a patient with early lung cancer was also examined. Tissues were fixed in 10% formalin and embedded in paraffin. Deparaffinized sections (5-μm-thick) were immersed in epitope retrieval solution (Target Retrieval Solution S1700; Dako North America, Inc., Carpinteria, CA, USA) and preheated at 120 °C for 10 min. After blocking endogenous peroxidase activity with 3% H 2 O 2 for 15 min, slides were incubated overnight with a mouse anti-human CD206 monoclonal antibody (15 µg/ml; R&D Systems, Minneapolis, MN, USA) or IgG2b at 4 °C. Subsequently, sections were incubated with visualization reagent (ChemMate Envision kit; Dako Japan, Inc., Tokyo, Japan) for 30 min, followed by counterstaining with hematoxylin.
Statistical analysis. Discrete variables are expressed as counts (percentage), and continuous variables are expressed as the median [range] unless otherwise specified. The Mann-Whitney test was used for continuous variables. Categorical data were compared between groups by using the Fisher's exact test for independence. Correlations between sCD206 and clinical parameters were analyzed by the Spearman's rank correlation method. Overall survival time was measured from the date of CAP diagnosis. To examine the ability of sCD206 levels to predict mortality in patients with CAP, combined cohort data were analyzed. The areas under the receiver operating characteristic (ROC) curve were used to evaluate the ability of this marker to predict mortality. The optimal cut-off value of sCD206 in the combined cohort that ensured the best combinations of sensitivity and specificity was obtained. Cumulative survival probabilities were estimated by the Kaplan-Meier method with optimal cut-off values. The Log-rank test was used to compare survival among patients. Univariate and multivariate analyses were performed with Cox proportional hazards regression analysis with combined cohort subjects to predict mortality. Among the statistically significant covariates in the univariate analyses, several were excluded because of potential confounders and statistical limitations. Statistical analyses were performed using GraphPad Prism Version 8 (GraphPad Software, San Diego, CA, USA) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was considered at a P-value of 0.05.

Data availability
The data are available from the authors upon reasonable request.