Altered glutamyl-aminopeptidase activity and expression in renal neoplasms

We analysed renal tissue in a series from patients with CCRCC, PRCC, ChRCC and RO. Hospital Ethics Committee approval was obtained a priori. Fresh tissue samples were obtained from surgical specimens from renal tumour patients. Tumour and normal (surrounding uninvolved tissue) areas were obtained in all cases. For activity studies, tissue samples were stored at −80°C until the enzyme assays were performed. In addition, selected tissue samples were formalin-fixed and paraffin-embedded for histopathological studies. The 2004 WHO histological classification of adult renal cell tumours [2] and the 2002 TNM Edition for tumour staging [19] were used for pathological diagnosis. In addition, Furhman’s method [20] was applied for grading CCRCC subtype. Clinical follow up of CCRCC patients was closed by December 31, 2012, and a total of 14 patients died of disease at that time. Mean follow up was 65.2 months (range, 8–130 months).

Although GAP has been classically reported as a cell-surface peptidase [16], several studies have also demonstrated GAP activity in soluble fractions of different tissues and cells [21‐23]. We therefore analysed GAP activity in both membrane-bound and soluble fractions of kidney tissues.

Explanted tissue samples were homogenised in 10 mM Tris–HCl buffer at pH 7.4, for 30 seconds at 800 rpm using a Heidolph PZR 50 Selecta homogeniser, and ultracentrifuged in a Centrikon T-2070 Kontron Instruments apparatus at 100,000 g for 35 min. The resulting supernatants were used to measure cytosolic (soluble) enzyme activities and protein concentrations. To avoid contamination with soluble enzymes, the resulting pellets were washed three times by suspension in 10 mM Tris–HCl buffer at pH 7.4. Pellets were then homogenised in 10 mM Tris–HCl buffer at pH 7.4, and centrifuged at low speed (1500 g) for 3 min to purify the samples. The supernatants thus obtained were used to determine membrane-bound enzyme activities and protein concentrations. All the aforementioned steps were carried out at 4°C.

The enzyme activity of tumour and non-tumour tissue samples from 50 patients with CCRCC (40 men, 10 women; mean age: 63 years), 10 patients with ChRCC (5 men, 5 women; mean age: 64 years) and 8 patients with RO (6 men, 2 women; mean age: 67 years) was analysed by spectrofluorometric methods. GAP activity was measured by a modified version of Tobe et al.’s method [24], using Glu-β-naphthylamide (0.125 mM) as a substrate. The assay is based on the fluorescence of β-naphthylamine generated from the substrate by GAP. The components of the assay mixture (total volume 2 mL) included the following: 50 mM of Tris–HCl buffer (pH 7.4) and 0.15 mg/mL of bovine serum albumin. The reaction was initiated by adding 30 μL of sample to 1 mL of the assay mixture. This was incubated at 37°C for 30 min and the reaction was discontinued by the addition of 1 mL of 0.1 M sodium acetate buffer (pH 4.2). The excitation and emission wavelengths were 345 and 412 nm, respectively. Blanks were used to determine background fluorescence. Relative fluorescence was converted into picomoles of product using a standard curve constructed with increasing concentrations of β-naphthylamine.

Protein concentration was measured in triplicate by the Bradford method [25], using BSA (1 mg/mL) as the calibrator. Results were recorded as units of peptidase (UP) per milligram of protein. One unit is equivalent to the release of one mole of beta-naphthylamine per minute. Fluorogenic assays were linear with respect to hydrolysis time and protein content.

Formalin-fixed and paraffin-embedded tumour tissue from 12 CCRCC (9 men and 3 women; mean age: 62 years); 6 PRCC (5 men and 1 woman; mean age: 64 years), 4 ChRCC (2 men, 2 women; mean age: 67 years) and 4 RO (all men, mean age: 63 years) were immunostained with a rabbit polyclonal antibody specific for glutamyl-aminopeptidase (Anti BP-1, Abcam plc, Cambridge, UK, working diluton 1:250). The immunostaining process was performed following routine methods in an automatic immunostainer (Dako Autostainer Plus). In short, endogenous peroxidase activity was blocked by incubating the slides in 3% hydrogen peroxide in absolute methanol for 10 minutes. Antigen retrieval was carried out in citrate buffer (10 mM, pH = 6) for 15 minutes at 100°C in a microwave oven. The primary antibody was applied for 1 hour at room temperature. A subsequent reaction was performed with secondary antibodies and biotin-free HRP enzyme labelled polymer of the EnVision-Flex detection system (Dako, Carpinteria, CA). Nonspecific IgG was used as a negative control. A positive reaction was visualized with diaminobenzydine solution followed by counterstaining with haematoxylin. Two independent observers analysed the immunohistochemical slides separately assigning staining intensity according to a semiquantitative scale [(negative (−), mild (+), moderate (++), and intense (+++)]. Minor disagreements were reconciled under a multihead microscope.

Quantitative RT-PCR for detecting ENPEP mRNA was performed to assess the transcription levels of this enzyme. The total RNA of tumour tissue samples from 30 CCRCC patients (18 male, 12 female; mean age: 62 years) was isolated following the standard protocol previously described [26].

First-strand cDNA was synthesized from 25 μg of total RNA from each human sample using Moloney murine leukemia virus reverse transcriptase and random hexamers according to the manufacturer’s instructions (first-strand cDNA Synthesis Kit, Amersham Biosciences, Essex, UK). The resulting cDNA samples were amplified by PCR with specific oligonucleotide primer pairs designed with the analysis software Primer 3 [27]. Based on previous experiments on human renal cell carcinoma [28] and other human tissues [29, 30], TATA box binding protein (TBP), peptidylprolyl isomerase A (PPIA), β-actin (ACTB) and succinate dehydrogenase complex subunit A (SDHA) were chosen as endogenous reference genes. The sequences of the primers used to amplify ENPEP and the four housekeeping genes are shown in Table 1. All primers were synthesized and purified by Sigma-Genosys (Cambridge, UK).

Expression of the target and housekeeping genes was quantified in all cDNAs by real-time PCR using the iCycler iQ real-time detection apparatus (BioRad Laboratories, Hercules, CA, USA). Dilutions of the cDNA template were prepared with each tissue and amplified in triplicate using SensiMix Plus SYBR + FLUORESCEIN (Quantace Ltd., London, UK). Three negative controls (with no template, no reverse transcriptase and no RNA in the reverse transcriptase reaction) were also included in each plate to detect any possible contamination. After a hot start (10 min at 94°C), the parameters used for PCR amplification were: 10 s at 94°C, 20 s at 60°C and 30 s at 72°C, for 50 cycles.

Real-time PCR data were expressed as the fold change of the target gene expression relative to the geometric mean (g.m.) mRNA expression of the housekeeping genes in each sample, as described by Vandesompele et al. [31]. The fold change in gene expression was calculated by the formula: 2 T – ∆ ∆ C, where C_T is the threshold cycle, as calculated by the iCycler software, ∆C_T = (C_Ttarget gene–C_Tg.m.reference genes) and ∆∆C_T = (∆C_T test sample–∆C_T control sample).

CCRCCs with different Fuhrman’s gade and stage (low and high) were always measured in the same analytical run to exclude inter-run variations..

Data obtained in the GAP activity assays across the different tumour types and in stratified CCRCC are reported in Figures 1 and 2.

Figure 1 shows GAP activity measured in tumour and non-tumour tissue (normal) of CCRCC, ChRCC and RO patients. Activity was recorded as pmol of product/min/mg protein (UP/mg protein) and is presented as mean ± SE. As shown in Figure 1A, when compared with non-tumour tissues, membrane-bound GAP activity decreased significantly in CCRCC (two-fold) (Mann–Whitney test, P < 0.001), and drastically in ChRCC (16-fold, P < 0.01) and RO (34-fold, P < 0.001). The soluble GAP activity in renal tumours (Figure 1B) also decreased significantly in all tumour types analysed when compared with normal tissue samples (P < 0.001).

Table 2 represents GAP activity in the different grades and stages of CCRCC group (Low grade: G1-G2, n = 24 vs. High grade: G3-G4, n = 26; Low stage: T1-T2, n = 34 vs. High stage: T3-T4, n = 16). Activity was recorded as pmol of product/min/mg protein (UP/mg protein) and is presented as mean ± SE.

The correlation study demonstrated that membrane-bound GAP activity is positively correlated with tumour size in CCRCC samples (Spearman’s rank test, P < 0.05) (Figure 2).

Finally, Kaplan-Meier curves revealed that GAP activity was not correlated with patients’ five-year survival (log-rank test P > 0.05) (Figure 3A and B).

Figure 4 shows the immunohistochemical results in non-tumour and tumour tissue. The pattern of staining of each tumour subtype coincided fully in every case. GAP immunostaining was strongly positive in proximal convoluted tubules in non-tumour renal tissue and slightly positive in distal tubules. CCRCC showed diffuse and mild membrane immunostaining. PRCC also showed positive immunoreaction located in cytoplasmic membranes, but the pattern of staining was focal. Both ChRCC and RO shared the same immunohistochemical expression with mild and diffuse cytoplasmic staining.

Table 3 shows the ENPEP mRNA levels in CCRCCs with different Fuhrman’s grade [G1/2 (low) vs. G3/4 (high)] and stage [pT1/2 (organ confined) vs. pT3/4 (non-organ confined)]. ENPEP mRNA levels were higher in low grade CCRCCs than in high grade tumours (Table 3a), however, this result was not statistically significant (Mann–Whitney test p = 0.322). Organ confined CCRCCs showed similar ENPEP mRNA levels to non-organ confined tumours (Table 3b) (p = 0.856).