Introduction

Cadherin-16 (CDH16) is a calcium-dependent, membrane bound cell-adhesion protein with a role in the formation of tubular epithelial structures in only a few organs. In the kidney, CDH16 promotes the formation of renal tubuli1 and shows persistent high-level expression in the adult kidney. Accordingly, CDH16 has also been named kidney specific cadherin (ksp-cadherin)1,2. However, CDH16 also plays a role in the development of thyroid follicles, and it is expressed in all follicular cells of the adult thyroid gland3.

RNA expression data suggest that CDH16 expression in normal tissues may be limited to the kidney, the thyroid and only few other tissues1,4,5,6,7,8,9. In the kidney and the thyroid, reduced expression of CDH16 has been linked to the development of cancer4,9. Only a small number of studies have used immunohistochemistry to analyze CDH16 expression in cancer and these were limited to renal cell carcinomas (RCC). CDH16 protein expression has been described to occur in 0–30% of clear cell RCC10,11,12,13,14,15, 0–29% of papillary RCC10,11,12,13,16, 5–100% of chromophobe RCC10,11,12,13,14,15,17, and in 0–95% of oncocytomas of the kidney10,11,12,13,14,17. Data from publicly available RNA databases suggest that CDH16 expression can—less commonly—also be found in other tumor entities including cervical, endometrial, and ovarian cancers4,9,18,19,20.

Given the predilection of CDH16 RNA expression to the kidney and the thyroid, CDH16 antibodies may be useful for the distinction of renal or thyroidal neoplasms from other cancers. However, immunohistochemical analyses of CDH16 expression are so far lacking for most tumor entities. To assess the diagnostic utility of immunohistochemical CDH16 expression analysis, the protein was evaluated in more than 15,800 tumor tissue samples from 152 different tumor types and subtypes as well as in 76 non-neoplastic tissue categories by immunohistochemistry (IHC) in a tissue microarray format in this study.

Results

Technical issues

A total of 13,424 (88.1%) of 15,584 tumor samples and more than 540 normal samples were interpretable in our TMA and large section analysis. Non-interpretable samples demonstrated lack of unequivocal tumor cells or absence of tissue in the respective TMA spots.

CDH16 in normal tissues

CDH16 immunostaining was predominantly seen in the kidney, thyroid and the epididymis. In the kidney, CDH16 immunostaining was stronger in proximal tubuli and in collecting ducts than in distal tubuli. The staining pattern was membranous (predominantly basolateral) and also cytoplasmic. In the thyroid, a strong membranous CDH16 staining occurred in follicular cells. In the epididymis, a predominantly membranous but also cytoplasmic staining was preferably seen in epithelial cells of the cauda while staining was absent or markedly weaker in the caput. A small fraction of epithelial cells, often arranged in nests or groups, showed a moderate to strong CDH16 staining in seminal vesicles. In some (but not all) analyzed samples, a focal weak to moderate membranous and cytoplasmic staining of individual cells, groups of cells or individual glands was seen in gallbladder epithelium, endometrium, and in the fallopian tube. Large section analyses also identified a strong CDH16 staining in Wolffian (mesonephric) duct remnants of the fallopian tube and in scattered cells, small groups of cells or of a limited number of glands in endocervical epithelium. Representative images of normal tissues are shown in Fig. 1. All these findings were obtained by using the monoclonal rabbit recombinant antibody MSVA-516R and the monoclonal rabbit antibody EPR13090, although EPR13090 resulted in a markedly less favorable signal to noise ratio. A cytoplasmic staining of gastric glands and of adrenocortical cells was only seen by EPR13090, but not by MSVA-516R, and was thus considered an antibody specific cross-reactivity. A comparison of antibody staining is shown in Supplementary Fig. 1.

Figure 1
figure 1

CDH16 immunostaining of normal tissues. In the kidney, CDH16 staining is predominantly basolateral and stronger in distal tubuli and collecting ducts than in proximal tubuli (A). In follicular cells of the thyroid (B) and epithelial cells of the cauda epididymis (C), a diffuse strong membranous staining is seen. A membranous staining of individual cells, groups of cells, or of individual glands can be seen in the endometrium (D), seminal vesicles (E), the fallopian tube (F), gallbladder epithelium (G), and the endocervix (H).

Full size image

CDH16 in cancer

CDH16 immunostaining was detectable in 1074 (8.0%) of the 13,424 analyzable tumors, including 476 (3.5%) with weak, 309 (2.3%) with moderate, and 289 (2.2%) with strong immunostaining. Overall, 40 (26.3%) of 152 tumor categories showed detectable CDH16 expression and 19 (12.5%) tumor categories included at least one case with strong positivity (Table 1). Representative images of CDH16 positive tumors are shown in Fig. 2. The highest rate of positive staining was found in renal cell carcinomas and follicular neoplasms of the thyroid, followed by several tumor entities of the female genital tract and of neuroendocrine neoplasms. CDH16 positivity was particularly frequent in nephrogenic adenomas (100%), oncocytomas (98%), chromophobe (97%), clear cell (85%), and papillary (76%) renal cell carcinomas (RCCs), follicular adenomas (86%) and follicular carcinomas (60%) of the thyroid, clear cell (56%), mucinous (36%), and endometroid (16%) carcinoma and carcinosarcoma (18%), of the ovary, adenocarcinoma of the cervix uteri (40%), serous (33%), clear cell (33%), and endometroid carcinoma (18%) of the endometrium as well as in various subtypes of neuroendocrine neoplasms (4–26%). In thyroid tissues, there was a significant decrease of CDH16 positivity from normal thyroid (8 of 8 positive, 100%) to follicular adenomas (81 of 94, 86.2%), follicular carcinomas (40 of 67, 59.7%) and papillary carcinomas (14 of 212, 6.6%; (p < 0.0001). In renal cell tumors, CDH16 positivity was significantly more frequent in oncocytoma (104 of 106 positive, 98%) and chromophobe cancers (64 of 66, 97%) than in clear cell (384 of 452, 85%) and papillary renal cell carcinomas (105 of 138, 76%; p < 0.0001). A CDH16 positivity was also seen in up to 8% of cases in 19 additional tumor categories but the staining was only weak to moderate in the vast majority of these cases. An additional large section analysis revealed a CDH16 positivity in all 12 nephrogenic adenomas of the urinary bladder (9 strong, 3 moderate) and absence of CDH16 staining in 3 peritoneal and 2 mesotheliomas of the tunica albuginea. A ranking order of tumor categories according to their rate of CDH16 positive and strongly positive cases is given in Fig. 3.

Table 1 CDH16 immunostaining in human tumors.
Full size table
Figure 2
figure 2

CDH16 immunostaining in cancer. The panels show a predominantly membranous CDH16 immunostaining of variable intensity in samples from a chromophobe (A) and a clear cell renal cell carcinoma (B), a clear cell carcinoma of the ovary (C), an adenocarcinoma of the cervix uteri (D), a nephrogenic adenoma (E), and a follicular adenoma of the thyroid (F). Samples from CDH16 negative papillary carcinomas of the thyroid are depicted in (G) and—adjacent to CDH16 positive normal thyroid follicles—in (H).

Full size image
Figure 3
figure 3

Ranking order of CDH16 immunostaining in tumors. Both the frequency of positive cases (blue dots) and the frequency of strongly positive cases (orange dots) are shown.

Full size image

CDH16 vs. Thyreoglobulin (TG) expression

The relationship between CDH16 expression and TG expression is shown in Supplementary Table 1. As TG expression was strictly limited to normal and neoplastic thyroidal epithelial cells21, dual positivity was commonly seen in benign thyroidal tissues, while the combination “CDH16 negative/TG positive” was strongly linked to thyroidal neoplasms and positivity for CDH16 alone was only seen in non-thyroidal neoplasms.

Discussion

Our successful analysis of 13,424 tumors from 150 entities identified CDH16 expression in 40 of 152 analyzed tumor categories and enabled a ranking of tumor types according to their CDH16 positivity rate. The most commonly CDH16 positive cancers included renal cell carcinomas, nephrogenic adenomas, and follicular neoplasms of the thyroid, followed by tumors of the female genital tract and various categories of neuroendocrine tumors. This is largely consistent with RNA expression data from The Cancer Genome Atlas Research Network (https://www.cancer.gov/tcga), suggesting overwhelmingly high rates and levels of CDH16 expression in renal cell carcinomas and—less frequently and at lower levels—in carcinomas of the thyroid, endometrium, ovary, and the uterine cervix. These data suggest three diagnostic applications of CDH16 immunohistochemistry including a) diagnosing papillary thyroid cancer, b) the distinction of renal cell carcinomas from other tumors in case of metastases with unknown primary tumor or in renal masses where a urothelial carcinoma or a metastatic tumor remain diagnostic options, and c) the diagnosis of nephrogenic adenomas.

The histologic diagnosis of papillary carcinoma of the thyroid is less complex than of follicular carcinoma but many papillary neoplasms of the thyroid remain diagnostically challenging22,23,24,25. This especially applies to the more than 41% of papillary carcinomas that show a pure follicular growth pattern26. Lloyd et al. reported a concordance rate of only 39% between 10 expert pathologists for the follicular variant of papillary carcinomas27. Difficulties are even higher in cytology where the sensitivity is reported to be 40%-96% for recognizing papillary carcinomas28,29,30,31,32,33,34,35,36,37,38,39,40,41. Considering the unequivocal and strong CDH16 staining in all normal thyroid samples, as well as the CDH16 expression loss in more than 90% of our papillary carcinomas, CDH16 loss appears to constitute a strong argument in favor of a papillary carcinoma. The high rate of papillary thyroid cancers lacking CDH16 staining is consistent with data from Li et al.4, describing markedly lower CDH16 RNA levels in papillary carcinomas as compared to normal tissues in 505 patients from the TCGA dataset and in 16 own cases. The functional role of CDH16 expression loss in thyroid neoplasms is unclear. Koumarianou et al.3 found a role of CDH16 for the formation of follicular structures which are, however, at least partially retained in many papillary carcinomas. It is of note that Cali et al.9 also described a reduced CDH16 expression in follicular carcinomas. A possible role of reduced CDH16 expression for a subset of follicular thyroidal neoplasms is consistent with the 13.8% adenomas and the 40.3% follicular carcinomas with CDH16 negativity in this study. In a previous study on a subset of these tumors we had identified thyroglobulin (TG) as a highly specific and sensitive marker for follicular and papillary thyroid cancer which is, however, unable to distinguish benign from malignant throidal tissue21. A particular diagnostic value of CDH16 may exist for thyroidal tissue detected in cervical lymph nodes where benign thyroid inclusions (expected to be TG and CDH16 positive) and metastases (TG positive, often CDH16 negative) must be considered.

The high rate of CDH16 positive kidney tumors in our cohort fits with data from existing RNA databases (https://www.cancer.gov/tcga). In analogy to the higher staining intensity in distal than in proximal tubuli of the normal kidney, the CDH16 staining was generally more intense in oncocytomas and chromophobe carcinomas—both derived from distal tubuli—than in papillary and clear cell carcinomas arising from proximal tubuli10,12,13. All 8 previous studies analyzing CDH16 by immunohistochemistry in tumors were all limited to renal cell carcinomas10,11,12,13,14,15,16,17. They described CDH16 positivity in 0% to 95% of 6–41 analyzed oncocytomas 10,11,12,13,14,17, 5.6% to 100% of 7–36 chromophobe RCCs10,11,12,13,14,15,17, 0% to 29% of 14–46 papillary RCCs10,11,12,13,16, and 0% to 30% of 15–102 clear cell RCCs10,11,12,13,14,15. The rather high rate of CDH16 positive clear cell (85%) and papillary (76%) RCCs in our study as compared to earlier data appears to be due to a higher sensitivity of our IHC approach which may not have negatively affected its specificity based on the virtual absence of non-specific staining in normal tissues. Although CDH16 expression differences between renal cancer subtypes are statistically significant, our data do not suggest a relevant practical utility of CDH16 immunohistochemistry for subtype distinction at the selected experimental conditions. This is also because of the striking utility of CD117 for this distinction42,43. We previously found a significant link between low CDH16 expression and unfavorable tumor phenotype and poor prognosis in clear cell RCC which may argue for a functional role of CDH16 expression loss in the progression of these tumors44. The high sensitivity of our assay may also be responsible for the detection of a significant CDH16 staining in 12 of 12 nephrogenic adenomas while Ortiz-Rey et al.45 had described CDH16 positivity in only 9 of 12 cases. Whether some of the novel oncocytic and molecularly defined RCC subtypes (eosinophilic vacuolated tumour, low-grade oncocytic tumour and TFE3-rearranged, TFEB-altered, ELOC (formerly TCEB1)-mutated, fumarate hydratase-deficient, succinate dehydrogenase-deficient, ALK-rearranged renal cell carcinomas and SMARCB1-deficient renal medullary carcinoma)—which were not distinguished in our historic tumor collection—may be particularly linked to CDH16 negativity needs to be determined in further studies.

Our data suggest that CDH16 immunohistochemistry may be useful for the distinction of renal cell carcinomas from other neoplasms. Although CDH16 is not specific for renal cell carcinomas and can also be seen in gynecological, neuroendocrine and several other tumors, it is noteworthy that CDH16 expression is mostly weak and not involving all cells in these extrarenal neoplasms. Completely renal specific antibodies have so far not been discovered. Immunohistochemical markers that are most commonly used for the distinction of renal cell carcinomas include PAX8 and CAIX46,47,48. However, PAX8 is abundantly expressed in gynecological tumors49,50,51 and thyroid cancers52,53,54, and can be found in various other tumors as well55,56,57,58. CAIX lacks expression in chromophobe kidney cancer59 and can also be expressed at high levels in various extrarenal tumors60,61,62,63,64,65. Studies are now needed to determine to what extent the additional use of CDH16 will improve the diagnostic precision of panels applied for establishing a renal cell tumor origin. The same applies for a potential diagnostic utility of CDH16 IHC in endocervical adenocarcinomas. Given the conspicuously high rate of CDH16 positive cervical adenocarcinomas as compared to the paucity of CDH16 positive cells in normal endocervical epithelium, significant CDH16 positivity may argue for malignancy at this location. The strong CDH16 positivity in mesonephric duct remnants must be considered, however, as these do regularly also occur in the uterine cervix66.

Considering the large scale of our study, our assay was extensively validated by comparing our IHC findings in normal tissues with data obtained by another independent anti-CDH16 antibody and RNA data derived from three different publicly accessible databases5,6,7,8. To ensure that the widest possible range of proteins would be tested for a possible cross-reactivity, 76 different normal tissues categories were included in this analysis. Validity of our assay was supported by the detection of significant CDH16 immunostaining in all organs with documented CDH16 RNA expression (thyroid, kidney, epididymis, seminal vesicles, and the fallopian tube). Additional CDH16 staining in gallbladder epithelium, the uterine cervix, endometrium glands, or mesonephric remnants, for which CDH16 RNA expression had not been described, were confirmed by the independent second antibody (Abcam EPR13090). In these organs, the CDH16 positive cells constitute such small fraction of the total number of cells that CDH16 RNA may not be present at detectable quantities in usual tissue samples.

Our data provide a comprehensive overview on CDH16 expression in normal and neoplastic human tissues. These findings suggest that—as a part of a panel—CDH16 immunohistochemistry might assist the identification of papillary thyroid cancer, the distinction of renal cell carcinomas from other neoplasms in cases of uncertain tumor origin, and the diagnosis of a nephrogenic adenoma.

Material and methods

Tissue microarrays (TMAs)

The normal tissue TMA was composed of 8 samples from 8 different donors for each of 76 different normal tissue types (608 samples on one slide). The cancer TMAs contained a total of 15,873 primary tumors from 150 tumor types and subtypes. The composition of both normal and tumor TMAs is described in detail in the results section. All samples were from the archives of the Institutes of Pathology, University Hospital of Hamburg, Germany, the Institute of Pathology, Clinical Center Osnabrueck, Germany, and Department of Pathology, Academic Hospital Fuerth, Germany. Tissues were fixed in 4% buffered formalin and then embedded in paraffin. The TMA manufacturing process was described earlier in detail67,68. In brief, one tissue spot (diameter: 0.6 mm) was transmitted from a tumor containing donor block in an empty recipient paraffin block. The use of archived remnants of diagnostic tissues for manufacturing of TMAs and their analysis for research purposes as well as patient data analysis has been approved by local laws (HmbKHG, §12) and by the local ethics committee (Ethics commission Hamburg, WF-049/09). All work has been carried out in compliance with the Helsinki Declaration. For data confirmation and extension, large section analyses were also executed on 10 cases each of normal thyroid, endocervix, fallopian tube, and gallbladder, 12 nephrogenic adenomas of the urinary bladder, as well as on 3 peritoneal and 2 mesotheliomas of the tunica albuginea. Data on thyroglobulin (TG) immunostaining were available for a subset of 8643 of our tumors from a previous study21.

Immunohistochemistry (IHC)

Freshly prepared TMA sections were immunostained on one day in one experiment. Slides were deparaffinized with xylol, rehydrated through a graded alcohol series and exposed to heat-induced antigen retrieval for 5 min in an autoclave at 121 °C in pH 7.8 DakoTarget Retrieval Solution™ (Agilent, CA, USA; #S2367). Endogenous peroxidase activity was blocked with Dako Peroxidase Blocking Solution™ (Agilent, CA, USA; #52023) for 10 min. Primary antibody specific against CDH16 protein (Recombinant monoclonal rabbit, MSVA-516R, MS Validated Antibodies, Hamburg, Germany) was applied at 37 °C for 60 min at a dilution of 1:150. For the purpose of antibody validation, the normal tissue TMA was also analyzed by the monoclonal rabbit CDH16 antibody [EPR13090] (Abcam; Cambridge, United Kingdom, ab214092) at a dilution of 1:300 and an otherwise identical protocol. Bound antibody was visualized using the EnVision Kit™ (Agilent, CA, USA; #K5007) according to the manufacturer’s directions. The sections were counterstained with haemalaun. For normal tissues, the staining intensity of positive cells was semi-quantitively recorded (+ , +  + , +  + +). For tumor tissues, the percentage of CDH16 positive tumor cells was estimated and the staining intensity was semi-quantitatively recorded (0, 1 + , 2 + , 3 +). For statistical analyses, the staining results were categorized into four groups as follows: Negative: no staining at all, weak staining: staining intensity of 1 + in ≤ 70% or staining intensity of 2 + in ≤ 30% of tumor cells, moderate staining: staining intensity of 1 + in > 70%, staining intensity of 2 + in > 30% but in ≤ 70% or staining intensity of 3 + in ≤ 30% of tumor cells, strong staining: staining intensity of 2 + in > 70% or staining intensity of 3 + in > 30% of tumor cells.

Statistics

Statistical calculations were performed with JMP 14 software (SAS Institute Inc., NC, USA). Contingency tables and the chi2-test were performed to search for associations between CDH16 expression and tumor phenotype. A p-value ≤ 0.05 was considered significant.

Ethics declarations

The study was approved by the Ethics commission Hamburg (WF-049/09) and conducted in accordance with the Declaration of Helsinki. Informed consent has not been collected specifically for the patient samples included in this study. Usage of routinely archived formalin fixed leftover patient tissue samples for research purposes by the attending physician is approved by local laws and does not require written consent (HmbKHG, §12,1).