Non‐invasive urine markers for the differentiation between RCCs and oncocytoma

Abstract Background Recently, our group showed that Vim3 is overexpressed in tissue samples of renal oncocytomas and Mxi‐2 in clear cell renal carcinoma (ccRCC). The mechanism leading to the truncation of both proteins is known and involves with two miRs, both detectable in urine. Since the analysis of miRs is time‐consuming, our aim was to identify the truncated proteins in urine instead. Furthermore, urine samples from small renal masses (SRMs) (n = 45, <4 cm) were analyzed to get a pre‐surgical differentiation of the cancer subtypes. Methods Urines were accessed from the urological biobank (n = 350). Proteins were isolated from urine samples, and Western blots were performed. Each sample was analyzed with ELISA for the expression of Vim3 and Mxi‐2. A lateral flow assay was established. For the detection of SRMs, the miRs were isolated and qRT‐PCR was performed. Results A significant increase of Vim3 in urines from patients with oncocytoma (n = 20) was detectable with ELISA compared to all other subtypes of RCCs (chromophobe (n = 50), papillary (n = 40), ccRCC (n = 200), and controls (n = 40) (***p < 0.0001)). Mxi‐2 was predominantly overexpressed in ccRCCs (***p < 0.0001). Lateral flow assay of Vim3 and Mxi‐2 shows two bands in the case of oncocytoma and ccRCC indicating the specificity of this test. For SRMs, an overexpression of miR‐15a/Mxi2 was detectable in urine samples from ccRCC and chromoRCC patients. In contrast to that, miR‐498/Vim3 were predominantly overexpressed in oncocytoma patients. Conclusion Both proteins (Vim3 and Mxi‐2) were detectable in patients’ urines and can be used for the non‐invasive differentiation of kidney cancers.

a protein with an unique C-terminal ending. 1 Under normal conditions, full-length vimentin is responsible for and important in cell shape, structure, and anchorage. 2 Since Vim3 misses the C-terminal ending of the full-length vimentin, a normal cell structure cannot be established, due to the missing tetramer formation of vimentin in these cells. As a result, more cell organelles are capable of an increased movement in the cytoplasm and a larger number of mitochondria can be present in the cell. 2 An overexpression of mitochondria exists in benign renal oncocytoma; nevertheless, these mitochondria are without function. Increased Vim3 levels were detectable only in oncocytoma tissue samples compared to all other kidney tumor entities. 3 However, due to this finding and the wellknown correlation between the miR-498 and Vim3, our aim was to detect Vim3 in patients' urine, since this is a non-invasive method.
Former studies could demonstrate that miR-15a is overexpressed in urine samples from patients with RCC 4 and decreases to a normal miR-15 level after tumor removal. miR-15a is responsible for the truncation of MAPK p38α and also results in a C-terminal truncated protein with a unique C-terminal ending, called Mxi-2. 1 An increased expression of Mxi-2 has already been shown to be detectable in urine samples in former studies. This led us to question whether a non-invasive method to differentiate benign renal tumors and kidney cancer could be performed by analyzing patients' urine for both miRs, namely miR-498 and miR-15a, in addition to the much cheaper and faster ELISA analysis of the target proteins (Vim3 and Mxi-2). Furthermore, we collected urine samples from patients suffering from small kidney cancers (<4 cm) for pre-surgical classification of the cancer.  The group of small kidney cancers (n = 10) with a size of <4 cm and a regression rate <10% consists of n = 45 patients including controls (n = 20) from healthy donors. The regression rate <10% was of importance for over a regression about 10% PKCα increase again which decrease the expression of miR-15a and therefore the Mxi-2 expression as well. 4

| Western blot analysis
Western blot analysis was performed from 50 µl urine samples.

No. of cycles
Rev. 5'-GAAATAAAATGCTTACCCCTCAG−3' Finally, the plates were washed 3× with PBS. TMB was used according to the manufacturer's protocol, and all reactions were stopped after exactly 10 min with stopping solution. For ELISA analysis, the FLUOstar Omega reader was used.

| Lateral flow assay
The lateral flow assay was performed as exemplified in Figure 4. HF 180 was used as analytic membrane. Antibodies were incubated overnight at room temperature and in the dark. As a conjugation pad, the GDFX membrane was used and incubated with the corresponding antibody for 3 h in a dry environment at 37°C. Finally, the lateral assay was set up and the urine sample (50 µl) was incubated for 10 min. For the analysis, fluorescent secondary antibody was used. All lateral flow assays were done in triplicates and analyzed with INTAS Chemostar.

| Total RNA isolation from urine
For miR isolation from patients' urine, 500 µl of urine was used and added to the QIAzol reagent, mixed, and further used according

| Statistical analysis
The GraphPad Prism 5 (San Diego, California, USA) program was used for statistical analysis. Analysis of variance (ANOVA) was performed, and the significant differences were calculated (*p < 0.05, **p < 0.01, and ***p < 0.001). All differences without stars were not of statistically significant. To analyze the target genes at the protein level, we performed a Western blot from sedimented urine samples. Here, it was possible to detect increased protein levels in both target sequences, namely Vim3 and Mxi-2 in the corresponding kidney tumors, with significant differences (*p < 0.05, **p < 0.01, and ***p < 0.001) ( Figure 1).

| RE SULTS
Since it was possible to detect the protein levels in the Western blot, we performed the ELISA method as well. Figure 2 shows the results from ELISA test. Significantly increased urine levels of Vim3 were found in patients suffering from oncocytoma as compared to patients with RCC and the control group (***p < 0.0001) (sensitivity of 90.2% and specificity of 82.4%).

F I G U R E 2 ELISA results of urine samples with different kidney tumor entities. Significant differences for Vim3 and Mxi-2 expression
were detectable (*p < 0.05, **p < 0.01, and ***p < 0.001) In contrast, Mxi-2 urine levels were significantly increased in patients with ccRCCs as compared to patients with an oncocytoma and the control group (p < 0.001).
For the faster analysis of urines, a lateral flow assay for the detection of Vim3 was also designed. Figure 3 illustrates the lateral flow assay that was performed for the detection of Vim3 in the patients' urine.

| DISCUSS ION
There is a great demand for non-invasive diagnostic tools of renal masses in the clinical routine. About 3-7% of renal tumors in adults are classified as benign renal oncocytoma. 8 The frequency of benign histology in small renal masses (SRMS), that is, <4 cm in diameter, is 20-30%.. 9 Although most of SRMS undergo percutaneous biopsy prior to any procedure, such as nephron sparing surgery or active surveillance, 12-20% of biopsies are inconclusive. 10 Table 2. As mentioned earlier, both miRs are responsible for the production of the C-terminal truncated proteins Vim3 and Mxi-2, so we can assume that we managed to transfer the former findings of increased miR levels to the protein level.
As protein diagnostics are significantly faster and more economical than miR tools, our procedure presents a significant step toward biomarker-supported diagnosis of renal masses. This is especially true in the realm of clinical routine and financing. Beyond this, the pre-surgical differentiation between benign and malignant kidney tumors based on the cross-sectional imaging fails to diagnose benign masses in more than 20% of all small lesions (<4 cm). 12 Therefore, a non-invasive test for the differentiation between benign and malignant RCCs is of great importance, especially with a view toward preventing the overtreatment of patients. Nevertheless, two other urine markers are also reported to be upregulated in urine samples, that is, aquaporin 1 and perilipin 2. These two markers can differentiate the clear cell and papillary RCC subtypes with a high sensitivity and specificity 13 ; however, to date no marker for the differentiation between benign and malignant kidney tumors is available.
Using the highly specific, non-invasive urine markers, Vim3 and Mxi-2, we would be able to fill the gap and reduce the number of overtreated patients as well as the number of unnecessary surgeries.

F I G U R E 3
Exemplification of the performed lateral flow assay (upper part). Lower part demonstrates the Vim3 signal in oncocytoma urine sample and the corresponding negative control with RCC urine Furthermore, even the detection as well as pre-surgical differentiation in small kidney cancers is possible with the measurement of the two predicted miRs as well as the proteins. MvB wrote the article with input from all authors.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets analyzed during the current study are available from the corresponding author on reasonable request.