Relationships between Chromosome 7 Gain, MET Gene Copy Number Increase and MET Protein Overexpression in Chinese Papillary Renal Cell Carcinoma Patients.

To investigate the relationships between Chromosome 7 gain, mesenchymal-epithelial transition factor (MET) gene copy number increase and MET protein overexpression in Chinese patients with papillary renal cell carcinoma (PRCC), immunohistochemistry (IHC), immunofluorescence (IF) and fluorescence in situ hybridization (FISH) were performed on 98 formalin-fixed, paraffin-embedded (FFPE) PRCC samples. Correlations between MET gene copy number increase, Chromosome 7 gain and MET protein overexpression were analyzed statistically. A highly significant correlation was observed between the percentage of tumor cells with MET gene copy number ≥3 and CEP7 copy number ≥3 (R2 = 0.90, p<0.001) across two subtypes of PRCC. In addition, the percentage of tumor cells with MET gene copy number ≥3 was found to increase along with increases in MET IHC score. This correlation was further confirmed in those PRCC tumor cells with average MET gene copy number >5 using combined IF and FISH methodology. Overall, this study provides evidence that Chromosome 7 gain drives MET gene copy number increase in PRCC tumors, and appears to subsequently lead to an increase in MET protein overexpression in these tumor cells. This supports MET activation as a potential therapeutic target in sporadic PRCC.

Introduction

Papillary renal cell carcinoma (PRCC) is the second most common subtype of renal cell carcinoma (RCC) and accounts for 10% ~ 15% of all RCC in the West, with clear cell renal cell carcinoma (CRCC) accounting for 80% of all RCC [1, 2]. Previous studies by Delahunt and Eble have divided PRCC into two morphologically different subtypes [3]. Type 1 PRCC is characterized by papillae covered by cells with scanty cytoplasms arranged in a single layer on the papillary basement membrane, while Type 2 PRCC is characterized by cells with eosinphilic cytoplasms and pseudostratified nuclei on papillary cores. Besides the morphological differences, Type 2 PRCC is usually more aggressive and presents a higher nuclear grade than Type 1 PRCC [3]. Unlike CRCC, where targeted therapy against vascular endothelial growth factor (VEGF) has dramatically improved the outcome of patients [4], VEGF-targeted agents show poor efficacy in PRCC. Up to now, no specific systematic therapy is available for metastatic PRCC [1].

Mesenchymal-epithelial transition factor (MET) protein functions as a transmembrane tyrosine kinase receptor [5]. When bound to its only known ligand, hepatocyte growth factor (HGF), MET protein activates downstream signaling pathways which promote cell proliferation, migration, invasion, angiogenesis and prevent cells from apoptosing [5]. It has been shown that germline mutations in MET lead to the development of hereditary Type 1 PRCC [6-8], sparking interest in the development of MET inhibitors to treat PRCC patients. Savolitinib, a MET inhibitor, was reported to induce tumor regressions in PRCC patient-derived xenograft models [9], and a phase II clinical trial to evaluate its efficacy in PRCC patients was recently launched (ClinicalTrials.gov. Maryland: the U.S. National Institutes of Health, Inc.; NCT02127710 [updated 2015 May 17]. Available from: https://clinicaltrials.gov/ct2/show/NCT02127710. Accessed May 26, 2015.). In sporadic PRCC patients, MET gene mutation [7], gene copy number alteration [10] and MET protein overexpression [10-13] were also observed. Recently, a study by Albiges et al reported MET gene copy number increases accompanied with high MET mRNA expression in a large cohort of 220 French PRCC patients [10]. Meanwhile, chromosome 7, where the MET gene resides, frequently exhibits trisomy in PRCC [14-21], also indicative of the occurrence of MET gene copy number increase in PRCC. Furthermore, tumors from PRCC patients carrying MET gene mutations commonly show trisomy 7 with non-random duplicated mutant MET genes and one wildtype MET gene [20].

Nevertheless, the etiology of sporadic PRCC is still largely unknown especially in Asian patients, possibly due to the lower prevalence of the disease in Asia [22]. Thus, our study aimed to investigate the association of Chromosome 7 gain, MET gene copy number variation and MET protein expression level in PRCC tumor tissues from a cohort of Chinese patients.

Materials and Methods

Patients

Tumor samples were collected from 98 PRCC patients who underwent surgery between 2010 and 2013 at Ren Ji Hospital, Shanghai, China. Prior written informed consent was obtained from all patients and the study protocol was approved by the ethics committee at Ren Ji hospital. Adjuvant chemotherapy was administered to 6 patients, while 46 patients did not receive chemotherapy. Chemotherapy status for the rest of the 65 patients was not available. Survival data was only available for 54 patients and therefore overall survival was not included in the data analysis due to the low follow-up response rate (55.1%). Histological subtypes (Type 1 and Type 2) were determined after review of tumor sections by two independent pathologists. This study was approved by Renji Hospital Institutional Review Board.

Fluorescence in situ hybridization (FISH)

Dual-color FISH assay was performed as previously described [23] on tissue microarray (TMA) at 4 μm thickness. The MET probe was derived by labeling BAC (CTD-2270N20, Invitrogen) DNA with SpectrumRed-dUTP (Cat #02N34-050, Enzo Biochem). CEP7-SpectrumAqua probe, specifically targeting the centromere of chromosome 7, was purchased from Vysis (32–111007, Vysis) and used as an internal control. FISH signals were observed using a fluorescence microscope equipped with the appropriate filters to allow visualization of the intense red/aqua signals and the blue counterstained nuclei. MET gene and CEP7 copy number was scored on 100 tumor cells for each case. Cases with MET gene average copy number ≥6 were defined as MET amplification. The percentage of tumor cells with MET gene copy number increase (≥3) (%MET) and CEP7 copy number increase (≥3) (%CEP7) was further calculated to statistically assess the correlation between MET and chromosome 7 copy number variation by Spearman’s correlation analysis.

Immunohistochemistry (IHC)

IHC analysis was performed on a 4 μm thick TMA. MET immunostaining was performed using CONFIRM anti-total c-MET (SP44) rabbit monoclonal antibody (790–4430, Ventana Medical Systems) on an automatic immunostainer (Discovery XT, Ventana Medical Systems) following standard staining protocol from Ventana Medical Systems. Ki67 IHC was performed as follows: slides were dried at 37°C overnight and then baked at 56°C for 30min, then deparaffinized in xylene and rehydrated through a graded series of ethanol concentrations. Antigen retrieval was performed in PTlink using high pH target retrieval solution at 97°C, 15 minutes (Target Retrieval Solution, pH 9, K8004, DAKO). Slides were covered with primary antibody solution (M7240, DAKO) and incubated at room temperature for 60 minutes, then washed twice in TBS-T and incubated with the EnVision+ system-HRP labelled polymer anti-rabbit secondary antibody (K4003, DAKO) for 30 minutes. Following two additional washes in TBS-T, slides were visualized using DAB substrate-chromagen (K3468, DAKO). All slides were independently evaluated and results were agreed by two pathologists who were blinded to patient information.

Combined immunofluorescence (IF) and FISH assays

4 μm thick PRCC sections were firstly de-waxed and pretreated in high pH antigen retrieval solution (Tris-EDTA, pH = 9.0; K8004, DAKO) at 95°C for 30 minutes. Sections were then treated with protease 3 (760–2020, Ventana Medical Systems) for 6 minutes and then incubated in CONFIRM anti-total c-MET (SP44) rabbit monoclonal antibody for 1 hour. Goat anti-rabbit IgG FITC (5ug/ml, ab6717, Abcam) was added and sections were incubated for 30 minutes at room temperature. After post-IF fixation in neutrally-buffered formalin for 10 minutes, sections were dehydrated in ethanol series and air-dried. Probes for MET gene and CEP7 were applied and slides were co-denatured at 79°C for 6 minutes, and then incubated at 37°C for more than 30 hours. After overnight incubation, slides were washed with 0.3% NP40/2 x SSC (pH 7.0–7.5) at 72°C for 2 minutes, and then in 2 x SSC at room temperature for 2 minutes. Slides were finally dehydrated in ethanol series, air-dried, and mounted with mounting medium with DAPI (Cat #H-1200, Vector Labs). Slides were analyzed using fluorescent microscopy.

Statistical analysis

Statistical analysis was performed using R statistical software, and p>0.05 was defined as statistically significant.

Results

Patient characteristics

Detailed patient demographics and tumor characteristics are listed in Table 1. PRCC was more prevalent within males in this cohort (3:1), with the two subtypes almost evenly distributed (Type 1, 45.9% versus Type 2, 54.1%). The percentages of tumor grades were Grade 1, 8.2%; Grade 2, 48.0%; Grade 3, 32.6%; Grade 4, 11.2%. The percentages of clinical stages were Stage 1, 67.3%; Stage 2, 9.2%; Stage 3, 11.2%; Stage 4, 9.2%; Unknown 3.1%. Following independent review by two pathologists, all 98 samples were classified as PRCC (Fig 1). Consistent with previous studies [3], Type 2 PRCC was found to be correlated with higher tumor grades (p<0.001) and clinical stages (p = 0.01), when compared with Type 1 PRCC.

Table 1

Patient Demographics and Tumor Characteristics (N = 98). Factors	Number	%
Age, years
≤ Median (58)	45	45.9
>Median (58)	45	45.9
Unknown	8	8.2
Gender
Male	71	72.4
Female	27	27.6
Histological type
Type 1	45	45.9
Type 2	53	54.1
Tumor grade
Grade 1	8	8.2
Grade 2	47	48.0
Grade 3	32	32.6
Grade 4	11	11.2
T
1a	47	48.0
1b	21	21.4
2	11	11.2
3	9	9.2
4	7	7.1
Unknown	3	3.1
N
0	89	90.8
1	6	6.1
Unknown	3	3.1
M
0	95	96.9
1	0	0
Unknown	3	3.1
Clinical stage
1	66	67.3
2	9	9.2
3	11	11.2
4	9	9.2
Unknown	3	3.1

Fig 1

Representative images showing typical morphology in Type 1 (A) ad Type 2 (B) PRCC respectively.

Type 1 PRCC was characterized by papillae covered by cells with scanty cytoplasms which were arranged in a single layer on the papillary basement membrane, while Type 2 PRCC was characterized by cells with eosinphilic cytoplasms and pseudostratified nuclei on papillary cores.

Biomarker evaluation

For biomarker evaluation, samples which failed quality control (no tumor tissue found or failed assays) were excluded. MET protein expression was analyzed in 94 PRCC samples using IHC staining. The MET protein overexpression rate, as defined by IHC 2+ and 3+, was 73.4% (69/94). MET gene and chromosome 7 copy numbers were analyzed in 91 PRCC samples using FISH. Tumor cells with MET gene copy number ≥3 were observed in 90.1% (82/91) patients, while tumor cells with CEP7 copy number ≥3 were observed in 92.3% (84/91) patients. The percentage of tumor cells with MET or CEP7 copy number ≥3 was variable among patients. One case (1/91) of MET gene amplification (as defined by gene copy number ≥6), was observed in our cohort. Ki67 index, defined as the number of Ki67 positive cells per 1000 tumor cells, was evaluated in 98 PRCC samples. The overall Ki67 index was 0.65±1.72. Comparisons between Type 1 and Type 2 PRCC revealed similar biomarker prevalences (Table 2). All biomarkers showed no correlation with clinical parameters, with the exception that the Ki67 index showed a higher trend in male patients than females (p<0.001).

Table 2

Biomarker details in Type 1 and Type 2 PRCC.

PRCC subtype	MET protein expression (by IHC)		MET gene copy number (by FISH)		CEP7 copy number (by FISH)		Ki67 index (by IHC)
	low expression (IHC 0/1+)	high expression (IHC 2/3+)	disomy	with copy number ≥3	disomy	with copy number ≥3	Mean±SD
Type 1	16.3% (7/43)	83.7% (36/43)	7% (3/42)	93% (39/42)	5% (2/42)	95% (40/42)	0.66±1.53
Type 2	35.3% (18/51)	64.7% (33/51)	12% (6/49)	84% (43/49)	10% (5/49)	90% (44/49)	0.71±1.88
All	26.6% (25/94)	73.4% (69/94)	9.9% (9/91)	90.1% (82/91)	7.7% (7/91)	92.3% (84/91)	0.65±1.72

Relationships between MET gene copy number increase and Chromosome 7 gain

FISH assays were performed to determine the MET gene copy number and CEP7 copy number in 91 PRCC samples. After evaluation of MET and CEP7 copy numbers in the tumor cells, a significant correlation was found between the percentage of tumor cells with MET gene copy number increase (≥3) (%MET) and the percentage of tumor cells with CEP7 gene copy number increase (≥3) (%CEP7) (R2 = 0.90, p<0.001, Spearman’s correlation analysis, Fig 2). This indicates that copy number gains in Chromosome 7 are likely the cause of MET gene copy number increase in PRCC tumors. Such a correlation was observed in both PRCC tumor subtypes (R2 = 0.85 & 0.91, Type 1 and Type 2, respectively).

Fig 2

Correlation between percentage of tumor cells with MET gene copy number ≥3 and percentage of tumor cells with CEP7 gene copy number ≥3 in PRCC (N = 91) patients.

As shown in Fig 2, the %MET was significantly higher in Type 1 PRCC samples, as compared to Type 2 (p = 0.01). No significant relationship was found between %MET and clinical stage (p = 0.82), indicating that the increase in the percentage of tumor cells with MET gene copy number increase may not be directly related to the severity of the disease or tumor metastasis. After stratification of the tumor subtypes, no significant relationship was found between %MET and tumor grades (p = 0.89, Type 1; p = 0.73, Type 2) or clinical stages (p = 0.82, Type 1; p = 0.54, Type 2). Similar results were observed in %CEP7 (data not shown).

Relationships between MET gene copy number increase and MET protein overexpression

A positive trend was observed between %MET and increasing IHC score (Fig 3), indicating a possible correlation between MET gene copy number increase and protein overexpression in PRCC tumor cells. Further studies using IF combined with FISH showed that MET protein overexpression (IF 3+) was colocalized in tumor tissues with average MET gene copy number >5 (PRCC-009, 052, 097; Fig 4). In contrast, in tumor tissue where MET IF showed negative staining, the average MET gene copy number = 2 (Fig 4). This result suggests a correlation between MET gene copy number increase and MET protein overexpression in the same tumor cells.

Fig 3

Distribution of percentage of tumor cells with MET gene copy number ≥3 among different MET IHC scores in PRCC (N = 88) patients.

Fig 4

Correlation between MET protein overexpression and MET gene copy number increase.

The upper row shows representative images of MET protein overexpression analyzed by both IHC (A) and IF (B) staining, correlating with MET gene copy number increase (average copy number >5) (B). The lower rows show representative images of no MET protein expression shown by both IHC (C) and IF (D) staining, correlating with MET disomy (average copy number = 2). For the fluorescent images, green signals represent IF staining which indicated by the triangle, red signals represent MET gene copy number which indicated by the arrows, aqua signals represent CEP7 copy number and nuclei are counterstained as blue.

Discussion

PRCC is the second most common type of renal carcinoma, and is associated with poor prognosis in some studies, especially in its metastatic form [24]. Hereditary Type 1 PRCC has been linked to a genetic mutation in the MET gene [6, 25]. In addition, in a previous phase II clinical trial, Foretinib (targeting VEGF/MET) showed a high (50%) response rate in PRCC patients with germline MET mutations [26]. However, the pathogenetic background and etiology of sporadic PRCC is still largely unknown. A recent study investigating 220 western patients with PRCC found a significant correlation between MET gene copy number and MET mRNA expression across the two subtypes of PRCC [10].

In this study, we used FISH and IHC techniques to investigate MET gene copy number variation and protein expression in 98 Chinese PRCC patients. Instead of calculating average gene copy number as previously reported, we looked into the detailed percentage of tumor cells with MET gene copy number ≥3 or with CEP7 gene copy number ≥3. In contrast to previous studies of whole tumor tissues using array CGH analysis, our detailed FISH scoring based on the percentage of tumor cells with MET gene copy number ≥3 should help to better evaluate gene copy number variations. Our results showed a highly significant correlation between %MET gene copy number increase (≥3) and %CEP7 gene copy number increase (≥3) in PRCC tumor cells. Only a single case of MET gene amplification (≥6) was found in our PRCC sample set. These results suggest a role for Chromosome 7 trisomy or polysomy during the pathogenesis of sporadic PRCC. Chromosome 7 gain (trisomy 7) has also been observed in hereditary PRCC, and causes duplication of a mutated MET gene [20]. However, in other tumor types such as lung cancer and gastric cancer, MET gene amplification, without Chromosome 7 gain, was found to be a potent activator of MET pathway signaling [27-30], underscoring the complexity of MET pathway activation as an oncogenic driver across different tumor types.

Furthermore, we also found an increasing trend in the percentage of tumor cells with MET gene copy number ≥3 together with higher MET IHC score, suggestive of a relationship between MET gene copy number increase and MET protein overexpression. This confirms previous findings observed between MET gene copy number variation and MET mRNA expression [10]. Finally, our findings from the combined IF/FISH studies provide direct evidence that MET protein overexpression is observed in the same PRCC tumor cells with a high MET gene copy number increase (>5), but not in those tumor cells with normal MET gene copy number.

Conclusions

Herein, we have studied the relationships between Chromosome 7 gain, MET gene variation and MET protein expression across the two subtypes of PRCC in a Chinese patient cohort. Our results clearly demonstrate that MET gene copy number increases in PRCC tumors are highly associated with increases in the copy number of Chromosome 7, which differs from previous observations in lung and gastric cancer. Such increases in MET gene copy number likely lead to increases in MET protein overexpression in PRCC tumor cells, especially in those with a high increase in the average MET gene copy number (>5). This study lends additional support to the concept of therapeutically targeting MET in sporadic PRCC patients.