B-MYB—p53-related relevant regulator for the progression of clear cell renal cell carcinoma

Abstract

Purpose

To investigate the mRNA expression of B-MYB and MDM2 together with their p53 relatedness in clear cell renal cell carcinoma (ccRCC).

Methods

Genes were screened for their mRNA expression from 529 patients in a publicly available ccRCC cohort (TCGA). A cohort of 101 patients with ccRCC served as validation by qRT-PCR mRNA tissue expression analysis.

Results

Expression: B-MYB expression was significantly higher in high-grade tumours (p < 0.0001 and p = 0.048) and in advanced stages (p = 0.005 and p = 0.037) in both cohorts.

Correlation: p53-B-MYB as well as MDM2-B-MYB showed significant correlations in local and low-grade ccRCCs, but not in high grade tumours or advanced stages (r < 0.3 and/or p > 0.05).

Survival: Multivariable Cox regression of the TCGA cohort revealed B-MYB upregulation and low MDM2 expression as predictors for an impaired overall survival (OS) (HR 1.97; p = 0.0003; HR 2.94, p < 0.0001) and progression-free survival (PFS) (HR 2.86; p = 0.0005; HR 1.58, p = 0.046). In the validation cohort, the results were confirmed for OS by univariable, but not multivariable regression: high B-MYB expression (HR = 3.05, p = 0.035) and low MDM2 expression (HR 3.81, p value 0.036).

Conclusion

In ccRCC patients with high-grade tumours and advanced stages, high B-MYB expression is common and is associated with poorer OS and PFS. These patients show a loss of their physiological B-MYB–p53 network correlation, suggesting an additional, alternative regulatory, oncogenic mechanism. Assuming further characterization of its signalling pathways, B-MYB could be a potential therapy target for ccRCC.

Electronic supplementary material

The online version of this article (10.1007/s00432-020-03392-7) contains supplementary material, which is available to authorized users.

Introduction

Renal cell carcinoma (RCC) represents a heterogenous group of tumours, all originating from the tubular epithelium of the kidney (Hsieh et al. 2017). The current WHO classification divides RCC into over 15 different histopathological and molecular RCC subtypes (Moch et al. 2016), where clear cell renal cell carcinoma (ccRCC) forms the largest subgroup including approximately 75% of cases. Key advances in genetics and molecular biology offer improved access to genome-based risk and treatment stratifications of RCC (Hsieh et al. 2018). Based on an increasing understanding of the transcriptomic signatures of RCC, mRNA clusters have been identified that are specifically expressed in the different histological RCC subtypes and thus form a cornerstone of personalized cancer management (Linehan and Ricketts 2019; Network 2013; Syed et al. 2020).

Genetic mutations of the well-known tumour suppressor gene, p53, have been studied extensively (Mantovani et al. 2019; Pitolli et al. 2019; Vousden and Lane 2007). They are key features of many cancer entities and are often associated with poor patient outcomes (Vieler and Sanyal 2018). Interestingly, in ccRCC, p53-mediated cell cycle regulation is often inactivated, but the p53 mutation rate is comparatively low (about 2–9%) (Mitchell et al. 2018). Thus, the presence of a wild-type p53 alone cannot be equated with a functional p53 network (Dalgliesh et al. 2010; Gurova et al. 2004; Sato et al. 2013). This brings upstream mechanisms of p53 repression, such as transcript dysregulation of p53 regulators around murine double minute 2 (MDM2), into focus (Warburton et al. 2005).

In cellular balance, high p53 expression via a feedback loop leads to upregulation of MDM2 (Wu et al. 1993). MDM2 in turn downregulates p53 activity through p53 degradation and thereby inhibiting p53-dependent transcriptional activity (Wade et al. 2013). This ensures a systematic flow of cell cycle arrest, apoptosis, autophagy, and other relevant cellular processes (Beckerman and Prives 2010).

At this point, the transcription factor B-MYB (formerly MYBL2) comes into play. B-MYB is a positive p53 regulator that acts by destabilizing the p53–MDM2 complex. Thus, it enhances the effect of serine-threonine kinase receptor-associated protein (STRAP) by promoting p53 translocation to the nucleus and increasing its stability (Seong et al. 2011). p53 activation in turn causes a complex formation of MYB–MuvB (MMB) to DREAM via hypophosphorylation of p130, which results in suppression of B-MYB (Quaas et al. 2012).

Dysregulation of the p53 upstream proteins MDM2 and B-MYB has already been associated with different tumour entities. For example, B-MYB has been shown to have an oncogenic function in colorectal cancer (Ren et al. 2015), to be crucial for the progression of human hepatocellular cancer (Calvisi et al. 2011) and has been proven in Ewing sarcoma to influence as a downstream player the inter-tumoral heterogeneity and thus tumour growth, drug response and patient survival (Musa et al. 2019). But little is known on their importance in renal cancer (Musa et al. 2017). To dissect the upstream regulation of p53, we analysed its components, their correlations, and their influence on clinical outcome in ccRCC using publicly available in silico data and qRT-PCR validation of a ccRCC tissue cohort.

Material and methods

In silico data analysis

An in silico data analysis was performed using publicly available normalized and log2-transformed RNA-Seq (HiSeq) data from The Cancer Genome Atlas (TCGA) Kidney Renal Clear Cell Carcinoma (KIRC) data set. Access and download of the expression data were provided via the Xena browser (https://xenabrowser.net/). Clinical–pathological information was supplemented via the cBioPortal for cancer genomics (https://www.cbioportal.org/). The data set was downloaded on February 12, 2019, checked for accuracy, and patients with missing clinical or expression data were excluded. Data from 529 patients were included in the final analysis (detailed cohort information is given in Table 1). The association of p53, MDM2, and B-MYB gene expression was analysed with respect to age, gender, tumour stage, and tumour grade according to the 7th edition of the AJCC (American Joint Committee on Cancer) guidelines. The comparison of tumour stage was done in groups, contrasting local ccRCCs (T1-3 and N0, M0) to locally or distantly advanced RCCs (all T4, N1, M1). Overall survival (OS) and progression-free survival (PFS) were defined as end points.

Table 1.

Cohort data

	TCGA cohort	Validation cohort
	n = 529	n = 101
Median follow-up (IQR)	42.7 (20–64)	85.5 (63–102)
Sex
Male, n (%)	343 (64.8)	72 (71.3)
Age
Median (IQR)	61 (52–70)	63 (53–70)
T-stage
pT1, n (%)	270 (51.0)	52 (51.5)
pT2, n (%)	69 (13.0)	14 (13.8)
pT3, n (%)	179 (33.9)	33 (32.7)
pT4, n (%)	11 (2.1)	2 (2.0)
Grading
G1, n (%)	14 (2.7)	14 (14.0)
G2, n (%)	226 (43.4)	76 (76.0)
G3, n (%)	206 (39.5)	10 (10.0)
G4, n (%)	75 (14.4)	0 (0.0)
N-stage
N + , n (%)	16 (6.3)	4 (4.0)
M-stage
M + , n (%)	78 (15.7)	4 (4.0)

Detailed perioperative, clinicopathological data of the TCGA and validation cohorts

Patients

Validation of in silico data analysis was performed on a separate cohort of ccRCCs. Tissue from patients who underwent radical or partial nephrectomy between February 2008 and July 2011 at the Department of Urology and Urosurgery at the University Medical Centre Mannheim was preserved. All included patients consented to the use of their biomaterials and a positive ethics vote was updated on May 28, 2015 by the Ethics Committee II at the University of Heidelberg (2015-549N-MA). Following standard protocols, we fixed tissue with formaldehyde and embedded it in paraffin (FFPE) before obtaining microscopic sections and storing them at room temperature. FFPE sections were hematoxylin–eosin stained and examined by a trained pathologist. Tumour staging and grading were classified according to the 7th edition of TNM Classification of Malignant Tumours by the Union for International Cancer Control (UICC) (Sobin L.H. 2011). Non-clear cell renal cell carcinoma or samples with high fibrosis or necrosis were excluded from further analysis. Only sections with a tumour fraction > 10% were included.

The following clinicopathological parameters were recorded: age, sex, tumour stage, tumour grade, lymph node, and metastasis stage. Follow-up data were actualized. Patients with a follow-up ≤ 60 days were excluded from survival analyses. Ultimately, 101 patients were included in the validation cohort. Detailed clinicopathological data are listed in Table 1.

RNA isolation

Total RNA was manually isolated from the FFPE sections as previously described (Kriegmair et al. 2016, 2018). RNA was purified according to the manufacturer's instructions using the XTRAKT FFPE Kit from Stratifyer (Cologne, Germany). First, 10 μm FFPE sections were deparaffinized in the lysis buffer, followed by tissue digestion with Proteinase K. For isolation of the nucleic acids, tissue lysates were mixed and incubated with germanium-coated magnetic particles in a special buffer system. Samples were washed three times by magnetic separation and careful removal of the supernatant. Total RNA was solubilized from the beads by addition of 100 μl elution buffer and incubation at 70 °C, before putting the samples immediately on ice. RNA quality and quantity were checked on a Nanodrop spectrophotometer, and only samples with a 260/280-ratio > 1.8 were further analysed. Purified RNA was stored at − 80 °C until further use. 5 μl total RNA from FFPE sections was used per 20 μl cDNA synthesis, which was carried out by sequence-specific reverse transcription with Super Script III reverse transcriptase (Thermo Fisher Scientific, Waltham, USA).

Quantitative real-time PCR

For validation of the MDM2, B-MYB, and p53 TCGA gene expression profiles, an FFPE tissue-based ccRCC cohort with n = 101 patients was analysed. qRT-PCR gene expression assays were carried out using TaqMan Fast Advanced Mastermix (Thermo Fisher Scientific, Waltham, USA). Gene-specific primers and probes are given in the Online Resource (Table 5). Each assay included a negative control (no template) and positive control (cDNA prepared from cell line 769-P). The cell line 769-P (RRID: CVCL_1050) was derived from a primary renal clear cell carcinoma (human, female Caucasian, 63 years old). The expression levels of the established housekeeping gene calmodulin 2 (Kriegmair et al. 2016) and the target genes were determined in duplicates on a StepOne Plus qRT-PCR system (Thermo Fisher Scientific, Waltham, USA). In detail, the samples were first heated to 60 °C for 20 s. DNA amplification was monitored over 40 cycles of 3 s at 95 °C and 30 s at 60 °C and cycle threshold (Ct) values were recorded. The relative expression of the target genes was calculated using the common 40-ΔCT method: 40-((Ct-target gene)-(Ct-housekeeping gene)).

Statistical analysis

The statistical analysis and graphics design of both data sets were performed with the software JMP® (Version 14.0 from SAS Institute Inc., Cary, NC, USA). For descriptive data analysis, the following parameters were calculated: mean value, median with interquartile range (IQR), and standard deviations. Bivariable group comparisons were performed using the nonparametric Mann–Whitney U test. Correlation analyses were performed using the Pearson method. Correlation coefficients (r) and p values were calculated for all aspects with their subgroup analysis. Survival time analyses were calculated for OS and PFS using uni- and multivariable Cox regression. As we excluded patients with an initial M + status from the PFS calculation and due to the overall low recurrence rate in the validation cohort, we were unable to record a sufficiently high number of events for a reliable and valid recurrence analysis and confirm the in silico data. Survival analyses were visualized as Kaplan–Meier curves. Survival times were compared with the log-rank test. For all analyses a significance threshold of < 0.05 was applied.

Results

Study cohorts

Detailed clinicopathological data of both study cohorts are summarized in Table 1. The in silico analysis included data from 529 patients with ccRCC. Median patient age was 61 years (IQR 52–70) and 64.8% (n = 343) were male. The available median follow-up was 42.7 (IQR 20.3–64.4) months. Overall, half of the tumours (51.0%, n = 258) were classified as T1. At the time of surgery, 6.3% (n = 16) of the patients had lymph node metastases and 15.7% (n = 78) had distant metastases. 101 patients with ccRCC were included in the validation cohort. Here, median age was 63 years (IQR 53–70) and 71.3% (n = 72) were male. Overall, the median follow-up was at 85.5 (IQR 63–102) months. While 51.5% (n = 52) of patients had T1 tumours and 4% (n = 4) lymph node metastases, the proportion of patients with distant metastases at the time of surgery was 4% (n = 4).

mRNA expression of p53, MDM2, and B-MYB in ccRCC tumour specimens

In the TCGA cohort, B-MYB showed a significantly higher expression in advanced tumour stages (p = 0.005) and high-grade tumours (p < 0.0001, Fig. 1). For the tumour suppressor gene, p53, similar mRNA expression levels were observed when stratified for tumour stage or grade (p > 0.05). Regarding tumour grade, no significant differences were found for MDM2 expression (p = 0.737), whereas MDM2 showed lower expression levels in tissue derived from advanced tumour stages (p = 0.005).

Fig. 1.

B-MYB expression profile. Relations between the relative gene expression of B-MYB and clinicopathological parameters. Compared to local (pT1-3, N0,M0) ccRCC, we observed a significantly higher expression in advanced ccRCCs (T4, all N1, all M1). p values: TCGA-cohort p = 0.005, validation cohort p = 0.037. We also observed a higher expression of MYBL2 in high-grade tumours (G3 and G4) compared to low-grade tumours (G1 and G2). p values: TCGA-cohort p < 0.0001, validation cohort p = 0.048. *Indicates significant differences between the two groups

In the validation cohort, we observed a significantly higher B-MYB expression in advanced (p = 0.037) and high-grade (p = 0.048) ccRCC tumours (Fig. 1). Advanced tumour stage was accompanied by reduced p53 expression (p = 0.04), while no p53 variance was observed for tumour grade (p > 0.05). For MDM2, no differences in mRNA expression were found with regard to tumour stage or grade (both p > 0.05).

Correlation of mRNA gene expression

We performed correlation analyses to examine the interaction of the different network regulators. In the TCGA cohort, we found a significant positive correlation of p53 expression with MDM2 (r = 0.55, p value < 0.0001) and B-MYB expression (r = 0.31, p value < 0.0001). Subgroup analysis revealed the p53–MDM2 correlation to be independent of tumour stage and grading (r range = 0.41–0.59, all p ≤ 0.0001). In contrast, p53–B-MYB correlation was not significant in the advanced tumour stages subgroup (local r = 0.35, p < 0.0001 vs. advanced r = 0.16, p = 0.15) and showed a below relevant level for the subgroup of high tumour grading (low r = 0.43, p = < 0.0001 vs. high r = 0.2, p = 0.0007). A detailed list of the correlation parameters is provided in Table 2 with an example shown in Fig. 2.

Table 2.

Correlation analysis between the different gene candidates of the p53 network

	TCGA Cohort			Validation cohort
	Correlation index	95% Confidence interval	p value	Correlation index	95% Confidence interval	p value
p53 and MDM2
Local RCC	0.58*	0.51–0.64	< 0.0001	0.78*	0.67–0.85	< 0.0001
Advanced RCC	0.41*	0.21–0.58	0.0001	0.65*	0.17–0.88	0.01
Low grading	0.59*	0.50–0.66	< 0.0001	0.76*	0.65–0.83	< 0.0001
High grading	0.53*	0.43–0.61	< 0.0001	0.76*	0.19–0.95	0.02
p53 and B-MYB
Local RCC	0.35*	0.27–0.43	< 0.0001	0.36*	0.11–0.56	0.005
Advanced RCC	0.16	− 0.06–0.37	0.15	0.49	− 0.08–0.82	0.09
Low grading	0.43*	0.32–0.53	< 0.0001	0.30*	0.06–0.51	0.02
High grading	0.20	0.08–0.31	0.0007	0.30	− 0.51–0.83	0.45
MDM2 and B-MYB
Local RCC	0.20	0.10–0.29	< 0.0001	0.35*	0.11–0.56	0.01
Advanced RCC	− 0.06	− 0.28–0.16	0.59	0.29	− 0.31–0.72	0.34
Low grading	0.27	0.15–0.39	< 0.0001	0.37*	0.13–0.56	0.002
High grading	0.03	− 0.09–0.14	0.67	0.04	− 0.69–0.72	0.93

The subgroup analyses show a loss of gene expression correlation of p53 and B-MYB as well as of MDM2 with B-MYB in advanced RCC (T4, all N1, all M1) and high grading RCC (G3 + G4)

Fig. 2.

Correlation plots (TCGA) of (a) p53 and B-MYB and (b) MDM2 and B-MYB to illustrate the gene expression correlation loss with advancing tumour stage (T4, all N1, all M1) and grading (G3 + G4)

In the validation cohort, we observed the same significant correlation pattern for p53–B-MYB and MDM2–B-MYB in localized tumour stage (r = 0.36, p = 0.005 and r = 0.35, p = 0.005) and low tumour grade (r = 0.30, p = 0.015 and r = 0.37, p = 0.002), whereas in advanced tumour stages and high-grade tumours, the correlation decreased and became statistically insignificant (r < 0.3 and/or p > 0.5) (Table 2).

Survival and recurrence analysis

Median follow-up of the TCGA cohort was 42.7 months (IQR 20.3–64.4) and median follow-up of the validation cohort was 85.5 months (IQR 63–102). To exclude perioperative complication bias, patients with follow-up data < 60 days were excluded from survival analysis. In the TCGA cohort, 32.9% (n = 166) died and 21.5% (n = 78) suffered from recurrence. The validation cohort includes 18.4% (n = 18) deaths and 21.8% (n = 19) with cancer progression.

Univariable Cox regression analyses of the TCGA data showed that high B-MYB (HR 2.6, p < 0.0001), high p53 (HR 1.93, p = 0.004), and low MDM2 expression (HR 1.84, p = 0.006), as well as age (HR 1.03/unit, p < 0.0001), high tumour stage (HR = 4.37, p = < 0.0001), and high tumour grade (HR = 2.6, p < 0.0001) had a significant, negative impact on the OS. Multivariable regression analysis confirmed the negative impact on OS for high B-MYB and low MDM2 expression, age, high tumour stage, and tumour grade. A visual representation by Kaplan–Meier survival curves is shown in Fig. 3.

Fig. 3.

Kaplan–Meier plots of the TCGA cohort. Visualizing the impact of the different gene expressions on the overall survival. Blue bars indicate low gene expression; red bars indicate high gene expression. Log-rank values are given in addition to the related Cox regression analysis results from Table 3. *Indicates significant differences between the two groups

These findings were confirmed by univariable Cox regression of the validation cohort for high B-MYB level (HR = 3.05, p = 0.035), low MDM2 level (HR = 3.81, p = 0.036), age (HR = 1.05/unit, p = 0.043), high tumour stage (HR = 11.53, p = < 0.0001), and high tumour grade (HR = 5.93, p = 0.004) (Fig. 4). p53 expression did not have a relevant impact on the OS, as multivariable regression identified only advanced tumour stage as significant (p < 0.0001). Full survival data are listed in Table 3.

Fig. 4.

Kaplan–Meier plots of the validation cohort. Expression data verifies the impact of MDM2 and B-MYB on the OS. Blue bars indicate low gene expression; red bars indicate high gene expression. Log-rank values are given in addition to the related Cox regression analysis results from Table 3. *indicates significant log-rank values in survival analysis

Table 3.

Overall survival analysis of uni- and multivariable Cox regression of the different gene expressions

Relative gene expression	TCGA cohort						Validation cohort
	Univariable analysis			Multivariable analysis			Univariable analysis			Multivariable analysis
	Hazard ratio	95% CI	p value	Hazard ratio	95% CI	p value	Hazard ratio	95% CI	p value	Hazard ratio	95% CI	p value
Age (per regressor unit)	1.03	1.0–1.0	< 0.0001*	1.03	1.0–1.1	< 0.0001*	1.05	1.0–1.1	0.043*	1.02	1.0–1.1	0.455
Gender (male vs. female)	0.94	0.7–1.3	0.698	–	–	–	0.96	0.4–2.7	0.939	–	–	–
Stage (advanced vs. local)	4.37	3.2–6.0	< 0.0001*	3.62	2.6–5.1	< 0.0001*	11.53	4.6–29.5	< 0.0001*	14.26	4.2–57.3	< 0.0001*
Grade. (high vs. low)	2.60	1.9–3.7	< 0.0001*	1.57	1.8–2.3	0.017*	5.93	1.9–16.0	0.004*	2.24	0.4–10.2	0.323
B-MYB (high vs. low)	2.60	1.9–3.6	< 0.0001*	1.97	1.4–2.8	0.0003*	3.05	1.1–9.8	0.035*	1.67	0.5–6.3	0.433
p53 (high vs. low)	1.93	1.2–2.9	0.004*	1.34	0.8–2.1	0.212	0.57	0.2–1.5	0.240	–	–	–
MDM2 (low vs. high)	1.84	1.2–2.7	0.006*	2.94	1.9–4.5	< 0.0001*	3.81	1.1–24.1	0.036*	2.70	0.7–18.2	0.177

Significant values are marked with an * and the significant values of the multivariable analysis are additionally printed in bold

A regression analysis for PFS in the TCGA data showed no significant influence of p53 on PFS (HR = 2.22, p = 0.079). In accordance with our results on OS, high B-MYB (HR = 2.86, p = 0.0005) and low MDM2 expression (HR = 1.58, p = 0.046) as well as age (HR = 1.02/unit, p = 0.020) and high tumour grade (HR = 0.91, p = 0.010) showed a significant influence on PFS in uni- and multivariable Cox regression analysis (Online Resource; Table 4, Fig. 4).

Discussion

B-MYB is a well-known cell cycle controlling transcription factor (Ness 1996) that is regulated by the DREAM complex (dimerization partner, RB-like, E2F and multi-vulval class B) (Sadasivam and DeCaprio 2013). In its active form, B-MYB influences a variety of regulatory genes (e.g. c-Myc, Bcl-2, cyclin) (Musa et al. 2017) and therefore has high oncogenic potential for typical cancer-specific properties such as uninhibited replication potential, low apoptosis rate, and metastasis. B-MYB has previously been associated with cancer progression in a number of tumour entities (Chen and Chen 2018; Gu et al. 2020; Li et al. 2020a) including urological tumours (Li et al. 2019, 2020b). In renal cancer, however, the relevance of B-MYB is poorly defined (Sun et al. 2020). Our group is the first to investigate B-MYB expression and its association to clinical parameters in a tissue cohort in ccRCC.

By classifying patients according to local tumour growth, lymphogenic and distant metastasis, or high-grade tumours, we could show that B-MYB levels increase in patients with advanced ccRCC. In addition, high B-MYB expression in ccRCC tumour tissue is associated with shorter OS and earlier recurrence. Accordingly, B-MYB was also identified as a negative prognosis marker for OS in glioblastoma. siRNA knockdown experiments were performed to functionally demonstrate that B-MYB develops its oncogenic potential through cell cycle regulation via a downstream mechanism of the Akt/FoxM1 signalling pathway (Zhang et al. 2017), a signalling axis known to be implicated in ccRCC development (Liu et al. 2019; Yang et al. 2018). Cell cycle-related pro-oncogenic effects of B-MYB were also shown in non-small lung cancer via the G1–S phase transition (Fan et al. 2018) and in oesophageal squamous-cell carcinoma via an increase of cell cycle transition to the G2/M phase (Qin et al. 2019). In addition to its involvement in cell cycle progression, B-MYB also regulates cell survival by an anti-apoptotic effect (Lang et al. 2005), thereby shortening survival in colorectal carcinoma (Ren et al. 2015). The unanimous preclinical data from various types of cancer strengthens the prognostic capacity of B-MYB and shifts it into focus for new therapeutic approaches.

The fact that in our data the B-MYB expression, but not the p53 or MDM2 expression, shows a valid and significant expression difference between local/advanced and low-/high-grade ccRCCs with simultaneous loss of B-MYB/p53 and B-MYB/MDM2 correlation of B-MYB in advanced and high-grade ccRCCs, suggests that B-MYB upregulation masks an additional, MDM2-p53-independent regulatory mechanism that triggers one of the aforementioned oncogenic cascades. Deregulation of the DREAM complex, hr20q13 amplification or post-transcriptional modification are known deregulation mechanisms of B-MYB in other cancers types (Musa et al. 2017).

Many cancers are characterized by a loss or mutation of p53 (Pietsch et al. 2006; Vogelstein et al. 2000); hence, the restoration of physiological p53 function is the basis for many proposed therapeutic, targeted strategies (Bykov et al. 2018; Lamont et al. 2000). Apart from its mainly oncogenic effects, some findings suggest that B-MYB also takes over functions that antagonize survival via an upstream control of the p53 network (Musa et al. 2017). In line with numerous preliminary studies, we could show that p53 expression in ccRCC is only moderately influenced by clinical pathological parameters and that a rather low p53 and a high MDM2 expression have a beneficial effect on OS. In such a constellation, a non-functional tumour suppressor gene, p53, is present in most tumour types (Pfister and Prives 2017). For ccRCC, however, it has been repeatedly shown that the occurrence of p53 mutations is low (Sato et al. 2013). But even p53 wild-type tumours can show a loss of function, depending on other aspects of the tumour, and suggests a functional limitation by other p53 regulators, such as B-MYB (Toledo and Wahl 2006).

Seong et al. showed that B-MYB mediates p53-effected apoptosis and cell cycle arrest via a direct STRAP interaction (Seong et al. 2011). The significant, observed loss of coherent p53–B-MYB expression in higher ccRCC stages and gradings could partly explain the reduced apoptosis rate and prolonged cell survival of cancer cells during ccRCC progression. Intriguingly, we found the same loss for correlated MDM2–B-MYB expression. Physiologically, the p53 network is kept in equilibrium, inter alia, by MDM2-directed degradation via the ubiquitin–proteasome pathway (Momand et al. 2000). Both proteins form a nuclear complex that is transported into the cytoplasm where p53 degradation takes place (Freedman and Levine 1998; Honda et al. 1997). B-MYB in turn promotes p53 nuclear translocation and enhances its stability by nuclear accumulation. We showed that in ccRCC, the MDM2–p53 correlation is unaltered, whereas B-MYB correlation with both molecules is simultaneously phased out. Thus, advanced ccRCC may be molecularly characterized by decreased p53 activation via B-MYB due to a lack of gene interaction. This might serve as an explanation for a non-functioning upstream p53 network in ccRCC. In conclusion, both high p53 and B-MYB expression and low MDM2 expression have a negative effect on the ccRCC OS.

In a restricted respect, it must be mentioned that many of the p53 regulatory mechanisms occur particularly at the protein level. Therefore, our transcription data should be extended by further proteome analyses to draw a definitive functional conclusion (Nag et al. 2013; Seong et al. 2011).

Another limiting factor is the slight differences in significance between the different survival time analyses, despite their similar trends. This indicates that our study is underpowered and has more an explorative character. The results are valuable as they provide the basis for sample size and effect estimation for further well-designed studies.

In summary, B-MYB is a newly identified and promising oncogene for ccRCC with many cellular regulation and pro-survival functions beneficial to cancer cells. During ccRCC progression, B-MYB loses elements of its anti-survival function via the p53 network. Decreased apoptosis and prolonged cell survival could lead to increased resistance of cancer cells, but a p53-independent, upregulatory oncogenic mechanism behind B-MYB is also conceivable. Further studies are necessary to elucidate the specific functional mechanism. A detailed understanding of the molecular mechanism of B-MYB correlation loss with the p53 network will provide a promising approach for new target therapies or tumour markers.

Conclusion

In this study, B-MYB is the only investigated gene of the p53 network that is upregulated in patients with advanced ccRCC and associated with impaired oncological outcome.

In advanced ccRCC, it shows a loss of its physiological correlation to p53 and MDM2, which implies that B-MYB overexpression in this context may be more strongly mediated by a mechanism that is independent of MDM2–p53 signalling. This mechanism should be the subject of further molecular cancer research, as B-MYB offers potential as a biomarker and eventually as a therapeutic target in individualized cancer medicine in the ccRCC.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

AJCC

American Joint Committee on Cancer

ccRCC

Clear cell renal cell carcinoma

FFPE

Formaldehyde and embedded in paraffin

HR

Hazard ratio

IQR

Interquartile range

OS

Overall survival

PFS

Progression-free survival

RCC

Renal cell carcinoma

STRAP

Serine-threonine kinase receptor-associated protein

UICC

Union for International Cancer Control

TCGA

The Cancer Genome Atlas

Author contributions

MN: guarantor of integrity of the entire study, statistical analysis, manuscript preparation; KM: experimental studies/data analysis; KN and AS experimental guidance; ZVP and SP pathological validation; PE, JM, and FW: manuscript editing; MCK: study concepts and design, guarantor of integrity of the entire study, manuscript editing.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The authors confirm that the data supporting the findings of this study are available within the article and its Online Resource.

Code availability

Not applicable.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no competing interests.

Ethical approval

The authors declare that no competing financial interests exist. This study was performed in adherence to the Declaration of Helsinki, and all patients gave approval to participate in this study.

Consent to participate

All participants have given their consent to participate in the study freely and after being informed.