Forty sarcoidosis patients treated at Toho University Omori Medical Center, and 30 age- and sex-matched healthy controls were recruited for evaluation of eHSP90α. The study was approved by the Ethics Committee of Toho University School of Medicine (protocol number A22080). All study subjects provided written informed consent for participation.

Diagnosis was based on the American Thoracic Society/European Respiratory Society/World Association for Sarcoidosis and Other Granulomatous Disorders statement on sarcoidosis (11). Briefly, patients with histological findings of noncaseating epithelioid granulomas from tissue specimens with relevant clinical and radiologic findings were diagnosed as having sarcoidosis. ACE, sIL-2R, and lysozyme levels were measured in the clinical laboratory. Radiologic staging of lung lesions was determined based on chest radiography and computed tomography (12).

Serum was collected at the time patients were enrolled in the study. Patients newly diagnosed with sarcoidosis during the study period underwent bronchoalveolar lavage with a fiberoptic bronchoscope. A total of 50 mL of saline was administered three times to the right medial lobe or left lingular lobe, and BALF was collected after each instillation. Collected serum and BALF were centrifuged at 3000 rpm for 10 min, and the resulting supernatants were aliquoted and frozen at −80°C until analysis.

The concentration of eHSP90α in patient serum and BALF and in the supernatant of medium was determined using a human HSP90α ELISA kit (Enzo Life Sciences, NY, USA) according to the manufacturer’s instructions.

Formalin-fixed, paraffin-embedded human lung tissues from sarcoidosis and control lungs were obtained from the Biobank for Interstitial Lung Diseases at St Joseph’s Healthcare in Hamilton, Ontario, Canada. All work conducted using human tissues was approved by the Hamilton Integrated Research Ethics Board (11–3,559 and 13,523-C). Affected lung lesions of sarcoidosis patients and non-tumor areas of lung tissues from control subjects were selected and placed in a tissue microarray (TMA) block using TMA Master II (3DHISTECH Ltd., Hungary).

TMA slides were stained with hematoxylin and eosin (H&E), anti-CD68 antibody (Agilent Dako M0876, CA, USA), and anti-HSP90α antibody (Novus Biologicals NBP1-77685, ON, Canada). High-definition images were acquired using an Olympus VS120 Slide Scanner.

THP-1 cells were purchased from the American Type Culture Collection (ATCC#TIB-202). THP-1 cells and monocytes were differentiated into macrophages by treatment with phorbol myristate acetate (Millipore Sigma, ON, Canada) at 10 ng/mL for 48 h.

Human monocyte-derived macrophages (MDMs) were generated from three healthy donors. Peripheral blood mononuclear cells (PBMCs) were isolated from peripheral blood and purified by density-gradient centrifugation using BD vacutainer mononuclear cell preparation tubes (Becton Dickinson and Company, NJ, USA). CD14-positive cells were magnetically isolated from PBMCs using an Easy sep Human CD14-positive selection kit (STEMCELL Technologies, BC, Canada) according to the manufacturer’s protocol. CD14-positive cells were differentiated into unpolarized macrophages by treatment with 20 ng/mL macrophage colony-stimulating factor (Peprotech, QC, Canada) for 6 days.

Macrophages were cultured in RPMI-1640 medium supplemented with 2 mM L-glutamine, 1% penicillin/streptomycin, and 10% fetal bovine serum. Macrophages were polarized toward the M1 phenotype by treatment with 100 ng/mL lipopolysaccharide (Peprotech, QC, Canada) and 20 ng/mL recombinant human interferon (IFN)-γ (Peprotech QC, Canada). The supernatant was collected at each time point and analyzed for eHSP90α concentration.

Data are presented as mean ± SD. The Student’s t test or χ² test were used to compare differences between two groups. Receiver operating characteristic (ROC) curve analysis was conducted to determine the optimal cut-off value of serum eHSP90α. Analyses of correlations between two groups were performed using Pearson’s correlation test. A p value of <0.05 was considered to indicate statistical significance. All statistical analyses were carried out using GraphPad Prism, version 8 (MDF Co., Ltd., CA, USA).

We examined the expression of HSP90α by immunohistochemistry on TMA slide generated from 34 sarcoidosis lung cores and 8 control lung cores. Control lung tissues displayed strong expression of HSP90α by lung macrophages and some alveolar epithelial cells (Figure 1).

Figure 2 shows representative TMA core sections of sarcoidosis lung tissues. Lung core regions containing granulomas were selected and isolated from lung biopsy specimens of 12 sarcoidosis patients under instruction of a lung pathologist (Supplementary Figure S1). Multiple epithelioid granulomas were observed upon H&E staining (Figure 2A). Granulomas showed overall highly positive signals for CD68, a pan macrophages marker, and HSP90α in the center region of granulomas (Figure 2A). Magnified images revealed that HSP90α expression was prominent in the central epithelial cells and multinucleated giant cells of granulomas but scarce in the interstitial cells and inflammatory lymphocytes surrounding the granuloma (Figure 2B).

Based on the finding that HSP90α was strongly expressed in sarcoid granulomas, we hypothesized that eHSP90α may be upregulated in the circulation and at local sites of disease. To test this hypothesis, we determined the eHSP90α levels in serum and BALF of sarcoidosis patients. Forty sarcoidosis patients and 30 healthy sex- and age-matched individuals were included in the study. Baseline characteristics of the patients and healthy controls are described in Table 1.

Serum eHSP90α levels were significantly higher in sarcoidosis patients (18,300 ± 8,100 vs. 6,988 ± 3,930 pg./mL, p < 0.0001) (Figure 3A). ROC curve was generated to determine the diagnostic value of serum eHSP90α (Figure 3B). A cut-off value of 11,088 pg./mL allowed discrimination of sarcoidosis patients and healthy controls with 86.7% sensitivity and 82.5% specificity (area under the curve [AUC] 0.9196, p < 0.0001).

To determine whether serum eHSP90α reflects sarcoidosis disease activity, we examined the associations between eHSP90α level and other disease markers. Serum eHSP90α level in sarcoidosis patients was significantly correlated with several biomarkers of sarcoidosis, including ACE, sIL-2R, and lysozyme (Figures 4A–C).

When examining the eHSP90α associations in the chest stages of the disease (12), the serum eHSP90α level was found to be higher in patients at more advanced chest stages compared with patients at lower chest stages (14,035 ± 5,697 vs. 20,128 ± 8,366 pg./mL, p = 0.031) (Figure 4D). These data suggest that serum eHSP90α corresponds to sarcoidosis disease activity and severity.

The level of eHSP90α in BALF was determined as a measure of eHSP90α in the lung. HSP90α was detectable by ELISA in the BALF of sarcoidosis patients (35,595 ± 38,816 pg./mL, n = 14), and the HSP90α level in BALF was higher in severe patients compared with patients at lower chest stages (16,928 ± 7,173 vs. 45,966 ± 45,649 pg./mL, p = 0.042) (Figure 5A).

To examine the association between serum and lung HSP90α levels, we analyzed the correlation between serum and BALF HSP90α concentrations in the same patients (Figure 5B) and found a strong correlation of BALF and serum HSP90α (r = 0.6974, p = 0.008).

Based on the results of staining of tissues from healthy and sarcoidosis lungs, we hypothesized that macrophages are a source of eHSP90α in the circulation and lungs. To examine the capacity of macrophages to produce eHSP90α, we differentiated THP1 monocytes into macrophages and polarized them into the M1 phenotype. Measurement of eHSP90α level in the supernatant demonstrated that macrophages can constitutively secrete HSP90α to the extracellular space, and cytokine stimulation can further promote HSP90α production (Figure 6A). Similar to cell lines, human MDMs can also produce HSP90α at steady state, and M1 macrophages are capable of secreting greater amounts of HSP90α (Figure 6B).