Serum Levels of N- and C-ERC/Mesothelin and Clinicopathological Factors in Mesothelioma Patients and Those without Mesothelioma

Objectives ERC/mesothelin is a glycosylphosphatidylinositol (GPI)-anchor protein expressed in mesothelioma. A precursor protein is cleaved by proteases and an N-terminal fragment (N-ERC) is extracellularly secreted. A remaining C-terminal fragment (C-ERC) is tethered on cellular membranes by the GPI-anchor, but C-ERC is also released after cleavage by proteases. We and other groups reported that serum N-/C-ERC levels are associated with stages of mesothelioma and suggested the possibility of their usefulness as diagnostic markers. However, the N-ERC level is also influenced by renal functions that are not directly associated with conditions of mesothelioma. It is not known whether other clinical factors influence serum N-/C-ERC values. Furthermore, their relationship to the amount of ERC/Mesothelin in mesothelioma is not yet validated. The objective of this study is to clarify the relationship of serum N-/C-ERC levels and the status of mesothelioma and several clinical factors. Materials and Methods We analyzed relations of serum N-/C-ERC levels and ages, gender and other clinical factors in 522 patients without mesothelioma and examined their relation to the amount of ERC/Mesothelin in mesothelioma tissues in 13 mesothelioma cases. Results Serum N-ERC levels were influenced by renal functions. On the contrary, those of C-ERC were not influenced by any clinical factors examined in this study and were significantly correlated with the amount of ERC/Mesothelin in mesothelioma. Conclusion Although both markers are good indicators of treatment-responses in individual patients with mesothelioma, only C-ERC reflected the amount of ERC/Mesothelin in mesothelioma among multiple patients, possibly because N-ERC was influenced by renal functions.


Introduction
Mesothelioma is an aggressive malignant disease arising from mesothelial cells that cover the surface of pleural, pericardial and peritoneal cavities, and is commonly associated with asbestos exposure 1) .It is intractable to conventional therapies.Even in the latest clinical trials using immune checkpoint inhibitors such as of nivolumab or pembrolizumab, the progression free survival was approximately 6 months and the overall survival was approximately 18 months 2,3) .Unsatisfactory effects of these treatments are partly associated with the difficulty in early diagnosis of mesothelioma.
Expressed in Renal Carcinoma (ERC) was originally isolated from renal carcinoma cells of Eker rat that hereditarily develops renal carcinoma 4) , and ERC is a homologue of human Mesothelin (MSLN) 5) .ERC/Mesothelin is a glycosylphosphatidylinositol (GPI)-anchor protein that is expressed on surface of normal mesothelium, epithelioid-type mesothelioma 6) or epithelioid components of biphasic mesothelioma.A 71-kDa ERC/Mesothelin-precursor protein is cleaved by proteases and a 31-kDa N-terminal fragment (N-ERC), that is identical to megakaryocyte potentiation factor (MPF) 7) , is extracellularly secreted.A remaining 40-kDa C-terminal frag-ment (C-ERC) is tethered on cellular membranes by the GPI-anchor, but C-ERC is also released after incomplete cleavage by other proteases, and C-ERC partially remains on cellular membranes 8) (Figure 1, Supplementary Figure 1).
Asbestos/mesothelioma outpatient clinic was established in Juntendo University hospital in 2005, to screen the asbestos-exposed laborers and their family members for the early diagnosis of mesothelioma [9][10] . Te patients took the regular check of chest x-ray or blood test including N-ERC and C-ERC.
We and others previously reported that both N-ERC and C-ERC serum levels are increased in mesothelioma patients and suggested that they can be useful for the early diagnosis of mesothelioma, or indicators of the effectiveness of chemotherapy or surgical treatments [11][12][13][14][15][16][17][18][19][20] . Hwever, the relationship between their serum levels and the amount of ERC in mesothelioma tissue is not yet validated.Furthermore, Shiomi et al., reported that the serum N-ERC level is increased in patients with renal failure 21) , and their findings suggested that it may be influenced by the clinical conditions that are not directly related to the status of mesothelioma.In this study, at first, we tried to clarify the relationship between the amount of ERC in mesothelioma tissue and the serum levels of N-or C-ERC.As a result, the serum levels of C-ERC were positively correlated to the amount of ERC in mesothelioma, but those of N-ERC were not.Secondly, because N-ERC is reported to be affected by the renal function, we examined whether the serum levels of N- or C-ERC are influenced by the clinical factors that are not directly associated with the mesothelioma.As a result, the serum levels of N-ERC were influenced by the renal function, as reported previously, but those of C-ERC was not.The C-ERC levels were not influenced by any of the clinical conditions examined in this study, other than those related to mesothelioma.Thirdly, we compared the clinical factors between the patients whose N-ERC was higher than C-ERC (N-higher group) and those whose C-ERC was higher than N-ERC (C-higher group).As a result, N-higher group included more women than men (p<0.05),although the reason for this phenomenon is to be clarified in the future.
Our study gave us a caution that we must be careful in the interpretation of serum C-and N-ERC values as the marker of mesothelioma among different patients.Both markers are, however, still valuable to monitor the status of mesothelioma in individual patients.

Patients
Serum samples and biopsy or surgically resected specimens of mesothelioma were obtained from 42 mesothelioma patients and 522 outpatients who visited Asbestos/mesothelioma clinic in Juntendo University between 2005 and 2019.The patient characteristics in this study are shown in Table 1.None of 522 outpatients showed clinical evidence of mesothelioma.All procedures were performed in accordance with the Ethics Committee at Juntendo University School of Medicine (approval number: H05-0014) and with the Declaration of Helsinki.Written informed consent for participation in this study was obtained from all patients.As for the clinical course of mesothelioma patients during the chemotherapy, the clinical data was retrospectively obtained, and the effectiveness of treatment was evaluated by guidelines for response evaluation criteria in solid tumor (RECIST) 22) .

Immunohistochemistry of ERC in mesothelioma
To evaluate the expression status of ERC in mesothelioma tissue, we performed immunohistochemistry of mesothelioma tissue by using mouse monoclonal anti-C-ERC antibody 22A31 (#10357, Immuno-Biological Laboratories (IBL), Fujioka, Gunma, Japan) as a primary antibody.Tissue sections with 4-μm thickness were prepared from formalin fixed, paraffin embedded specimens of mesothelioma obtained by surgical resection or biopsy.After deparaffinization, the tissues sections were heated in 10mM citrate buffer (pH 6.0) for antigen retrieval and treated with 3% hydrogen peroxide.They were blocked with 5% normal goat serum and incubated with 2μg/mL 22A31 in Tris-buffered saline with 0.1% Tween 20 (TBS-T) at room temperature for 180 minutes.After washing with PBS-T, the specimens were incubated with EnVision+ System-HRP labeled polymer conjugated to goat anti-mouse immunoglobulins (DAKO K4001, Agilent Pathology Solutions, Santa Clara, CA, USA), at room temperature for 60 minutes.Finally, the slides were incubated with 3,3-diaminobenzidine (DAB) at room temperature for 3 min.Two pathologists observed findings and evaluated the percentage of ERC-positive area that was expressed as ERC-positive rate (%).

Enzyme-linked immunosorbent assay (ELISA) of serum N-ERC or C-ERC
Serum levels of N-ERC and C-ERC were determined by sandwich ELISA systems described previously 12,23) with some modifications.Briefly, as for detection of N-ERC, two anti-N-ERC antibodies, monoclonal antibody (MoAb) 7E7 11) and horseradish peroxidase (HRP)-conjugated MoAb 16K16 12) were used as capture-and detection-antibodies respectively.As for detection of C-ERC, two anti-C-ERC antibodies, polyclonal antibody (PoAb) anti-C- ERC6 23) and HRP-conjugated PoAb anti-C-ERC 23) were used as capture-and detection-antibodies respectively.Dilution buffer for serum and detection antibody was 1% bovine serum albumin (BSA) in phosphate-buffered saline with 0.05% Tween 20 (PBS-T).Washing buffer was PBS-T.Diluted serum 100μL was loaded on each well coated with a capture-antibody and incubated at 37°C for 1 hour.After washing, 100μL of detection-antibody solution was added and incubated at 4°C for 30 minutes.After washing, for colorization, 100μL of tetramethylbenzidine solution (#19903, Immuno-Biological Laboratory, Fujioka, Gunma, Japan) was added and incubated at room temperature for 30 min in the dark place.Color development was stopped by 100μL of 1 N H 2 SO 4 .Absorbance of the solution at 450nm was measured in an ELISA reader (E-Max, Molecular Devices, Sunnyvale, CA, USA).The concentration of N-or C-ERC was determined by the standard curve derived from purified N-or C-ERC proteins 12,23) .Because of the outlying values of C-ERC in cases 28 and 294, and of N-ERC in a case 220 shown in Table 2, these three cases are removed from the further analyses.

Definition of N-higher and C-higher patients
As shown in Tables 2 and 3, some patients showed N-ERC higher than C-ERC, and the others showed the opposite tendency.We defined N-higher patients whose N-ERC values were more than twice higher than C-ERC, and similarly we defined C-higher patients showing C-ERC more than twice higher than N-ERC.In Tables 2 and 3, C-higher patients are shown in dark gray, N-higher ones are shown in white, and the others are shown in light gray backgrounds.Then we examined the relationship of N-or C-higher status and the clinical parameters.

Quantification of the ERC-positive volume of mesothelioma
We tried to clarify the relationships between the amount of ERC expressed in mesothelioma and the serum levels of N-or C-ERC.As an indicator of the amount of ERC protein in mesothelioma, we defined the ERC-positive volume (mL), as shown in the following formula.
ERC-positive volume (mL) = mesothelioma volume (mL) x ERC-positive rate (%) Mesothelioma volume was calculated in 13 patients whose image data of mesothelioma were available by CT or macroscopic pictures of surgical specimens.These 13 cases included 11 epithelioid-and 2 biphasic-types.By using ImageJ software 24 25) , we set region of interest (ROI) by drawing the circumference of mesothelioma on CT axial image of a 5-mm thick slice and calculated the area of ROI.The area (mm 2 ) was multiplied by thickness (5 mm) to induce the volume (mL) of mesothelioma in each slice, and they were summed up to create tumor volume (mL) of whole mesothelioma.In cases in which the macroscopic pictures of surgical specimens were available, ROI was drawn around mesothelioma in each slice.Volume of whole mesothelioma was calculated by a same way as in the cases with CT images.Then we calculated ERCpositive volume (mL) by multiplying mesothelioma volume and ERC-positive rate (%) based on the results of immunohistochemistry in mesothelioma.

Statistical analysis
All data were analyzed with SAS version 9.4 (SAS Institute, Cary, NC, USA).To compare serum N-/C-ERC levels between mesothelioma patients and those without mesothelioma, t-test was used.Pearson's correlation coefficient test was used to examine the correlation of serum N-ERC levels and C-ERC levels, serum N-/C-ERC levels and the ERC-positive volume in mesothelioma, and serum N-/C-ERC and clinical parameters.We considered that two factors are positively correlated when r > 0.40 and p < 0.05.Further, serum levels of N-/C-ERC between male and female, and clinical parameters between N-higher and C-higher patients were analyzed by t-test.The relationship between gender and N-/C-higher status was analyzed by chi-square test.A value of p < 0.05 was considered statistically significant.

Results
As expected, serum levels of N-and C-ERC were higher in mesothelioma patients than in those without mesothelioma (N-ERC; 19.0 ± 24.  2A) and those without mesothelioma (Figure 2B), with correlation coefficients 0.71 and 0.54 in mesothelioma patients and those without mesothelioma, respectively.
In 13 mesothelioma patients whose image data of the lesion was available, we checked the correlation between the serum levels of N-or C-ERC and the amount of ERC in mesothelioma that is expressed as the ERC-positive volume (mL).Serum levels of C-ERC correlated positively to the ERC-positive volume in mesothelioma (Figure 3B), but those of N-ERC did not (Figure 3A).Their raw data is shown in Supplementary Table 1.
In patients without mesothelioma, we examined the relationship between the serum levels of N-or C-ERC and the clinical parameters such as age, gender and status relating to anemia, inflammation, liver damages, kidney function, nutrition, platelets count, and diabetes, as described in Materials and Methods.As a result, serum creatinine levels, as the marker of renal function, correlated positively to serum levels of N-ERC (Figure 4A), but not to those of C-ERC (Figure 4B).All the other clinical parameters examined in this study did not have any correlation to either of C-or N-ERC (Supplementary Figure 2).
Serum levels of N-/C-ERC in 522 patients without mesothelioma and 42 mesothelioma patients are listed in Tables 2 and 3, respectively.In some patients C-ERC was higher than N-ERC, and in the others N-ERC was higher than C-ERC.As described in        Materials and Methods, we defined C-higher (C > N x 2), N-higher (N > C x 2) and the other patients, and they are shown in dark-gray, white, and lightgray backgrounds respectively in Tables 2 and 3.
In 522 patients without mesothelioma, the numbers of C-higher and N-higher patients were 43 and 184, respectively.Between these two groups, we compared age, gender and the other clinical parameters.Results showed that no relationship was detected between N-or C-higher states and all clinical factors (Table 4) except for gender.More females than males were included in N-higher group with statistical significance (Table 5).
Figure 5 shows the changes of serum N-or C-ERC in 2 mesothelioma patients who received chemotherapies.Response rates to therapies as indicated by RECIST 16) system are also indicated.
Both N-and C-ERC were shown to be reliable markers to reflect the tumor burdens in these two cases, as reported previously by other groups 14 16) .

Discussion
In mesothelioma tissues, ERC is localized on cellular membranes (Supplementary Figure 1), and it is detected by anti-C-ERC antibodies.On the contrary, anti-N-ERC antibodies 11,12,26) , that recognize the internal amino acid sequences of N-ERC and detect N-ERC in the extracellular fluid, cannot make specific signals in any cellular components of mesothelioma (data not shown).These findings suggest that almost all N-ERC molecules are released into the extracellular spaces, and that C-ERC are partially released but some of them are remaining on the cellular membrane.Therefore,    the amount of ERC in mesothelioma is evaluated by IHC using anti-C-ERC antibodies.Creaney et al. 16) previously reported that serum levels of C-ERC correlated with volume of mesothelioma, and their work supports our result shown in Figure 3B.They, however, just measured volume of mesothelioma and they did not count the expressional state of ERC in mesothelioma.We compared the serum levels of N-or C-ERC and three parameters related to ERC in mesothelioma tissue:  2, serum C-ERC level more significantly associated with ERC-positive volume (ml), than ERC-positive area (%).In our study, the significant relationship was not observed between serum C-ERC and tumor volume, as reported by Creaney et al. 14) possibly because of small number of cases (n=13) in our study.Serum level of N-ERC did not have any significant relationships with these three parameters.
Shiomi et al. 21) previously reported that serum levels of N-ERC are increased in the patients with renal failure.Therefore, in this study, we examined whether N-or C-ERC is influenced by renal functions and by the other clinical factors such as age, gender, inflammatory status, nutritional condition, anemia, liver function, platelets numbers, or diabetes.As shown in Figure 4, serum creatinine level positively correlated with N-ERC, but not with C-ERC, and this result is compatible with the report by Shiomi et al 21) .Both of N-and C-ERC did not have any correlation to the other clinical factors examined in this study (Supplementary Figure 2).N-ERC is identical to megakaryocyte potentiation factor (MPF) 7) that has the activity to simulate megakaryocyte development in murine system 27) , although the similar activity has not been reported in human.We studied the relationship between N-ERC and platelets numbers, and we did not find the correlation between them.Shiomi et al. 12) also reported that serum levels of N-ERC is influenced by age.In our data, the elder patients tended to have higher levels of N-ERC, but the correlation was not significant (Supplementary Figure 2).
Multiple studies reported that serum levels of N-ERC 11-14, 17-18, 20) and C-ERC [15][16]19) reflect the tumor burden of mesothelioma. Afterthe effective chemotherapy or surgical treatments, the serum levels of N-ERC 14) and C-ERC 16) decrease, and similar results are shown in Figure 5A.These findings indicated that both N-and C-ERC are good markers to monitor the status of mesothelioma in the clinical course of the same patients.Our present data (Figure 3) showed that, among different patients, C-ERC reflected the amount of ERC in mesothelioma more accurately than N-ERC, partly because N-ERC was influenced by renal function that is not directly associated with condition of mesothelioma.
It was puzzling for us why some patients showed that N-ERC was higher than C-ERC, and the others showed the opposite tendency.We tried to identify the causes of these phenomenon, but we were not able to find them, except for that N-higher patients included more female than male with statistical significances (Table 5).Renal clearance rate is generally higher in men than in women 28) .If N-ERC is excreted through renal routes, the higher clearance activity of men can explain our finding that N-higher group included more female.This explanation is compatible with the data that serum levels of N-ERC showed a tendency to be higher in female than in male, although there was no statistical significance (Supplementary Figure 2).In a patient shown in Figure 5A, C-ERC was always higher than N-ERC, and the relationship was opposite in a patient in Figure 5B, and these data suggested the possibility that N-higher or C-higher states were determined by some factors intrinsic to each patient.Recently, sex differences in carcinogenesis are being discussed 29) , and some hormonal condition may influence N-higher or C-higher states, although they are to be elucidated in future.
In conclusion, our study gave us a caution that we must be careful in the interpretation of serum N-and C-ERC values as the marker of mesothelioma among different patients, because N-ERC was influenced by renal functions that are not directly associated with conditions of mesothelioma.These markers are, however, still very valuable to monitor the status of mesothelioma in the same patients.

Figure 1
Figure 1 Structures of a 71-kD precursor ERC protein and two cleaved products, a 31-kDa N-ERC and a 40 kDa C-ERC.N-ERC and C-ERC are cleaved by proteases and released to the extracellular space.C-ERC, however, partially remains on the cellular membrane.

C
-higher patients (C > N x 2) are shown in dark gray, N-higher ones (N > C x 2) are shown in white, and the others are shown in light gray.

Figure 2
Figure 2 Relationship between serum N-and C-ERC in 42 mesothelioma patients (A) and 519 outpatients without mesothelioma in asbestos/mesothelioma clinic (B).Serum levels of N-ERC were positively correlated to those of C-ERC both in mesothelioma patients (A) and those without mesothelioma (B).

Figure 3
Figure 3 Relationship between ERC-positive volume (mL) in mesothelioma and serum levels of N-ERC (A) and C-ERC (B).Serum levels of C-ERC correlated to ERC-positive volume (mL) in mesothelioma (B), but that of N-ERC did not (A).ERC-positive volume (ml) was calculated by multiplying tumor volumes (mL) and ERC-positive area (%) in mesothelioma.n=13.

Figure 4
Figure 4 Relationship between serum levels of creatinine and N-ERC (A) or C-ERC (B) in 504 patients without mesothelioma.Serum levels of creatinine was positively correlated to N-ERC.

Figure 5
Figure 5 Changes of the serum levels of N-and C-ERC in two mesothelioma patients.Both of N-and C-ERC reflected the tumor burden of mesothelioma in individual patients.(A) The case showing C-ERC consistently higher than N-ERC, (B) The case showing N-ERC consistently higher than C-ERC.PR; partial response, PD; progressive disease, SD; stable disease, CDDP; Cisplatin, PEM; Pemetrexed sodium hydrate, CPT; Camptothecin (Irinotecan hydrochloride hydrate), VNR; Vinorelbine detartrate, GEM; Gemcitabine hydrochloride.
(A) tumor volume (ml), or (B) ERC-positive rate (%), or (C) ERC-positive volume [(C) = (A) x (B)].All three parameters are defined in Materials and Methods.As shown in Supplementary Table

0005 Supplementary Table 2
ERC-positive rate (%) 0.74 0.0039 (C) ERC-positive volume (ml) 0.69 0.Relationships between N-ERC or C-ERC levels in serum and three indicators (A), (B), (C) in mesothelioma (C) = (A) x (B).Serum C-ERC is more significantly associated with ERC-positive volume (C) than ERC-positive rate (B) or Tumor volume (A).Serum N-ERC is not significantly associated with any of three indicators.

Table 2
Serum levels [ng/mL] of N-ERC (N) and C-ERC (C) in 522 patients without mesothelioma

Table 2 (continued)
C-higher patients (C > N x 2) are shown in dark gray, N-higher ones (N > C x 2) are shown in white, and the others are shown in light gray background.N/D, not determined (case 169).Note that Cases 28, 220 and 294 are removed from the further analysis because of outlying values.

Table 3
Serum levels [ng/mL] of N-ERC (N) and C-ERC (C) in 42 mesothelioma patients

Table 4
Clinical factors and N-or C-ERC higher status in patients without mesothelioma