Introduction

Vestibular schwannomas (VSs) are benign, slowly growing encapsulated neoplasms originating from the Schwann cell sheath surrounding the vestibular branch of cranial nerve VIII. Regarding VSs, their frequency of occurrence is rather high in comparison with schwannomas at other regions in the head and neck. Although they are benign, their location inside the internal auditory canal and growth into the cerebellopontine angle make management difficult, and patients experience significant morbidity and mortality. The great majority of VSs are sporadic and unilateral schwannomas, which usually develop between ages 40 and 60; they also occur as bilateral lesions in the context of neurofibromatosis 2 (NF2), a highly penetrant, autosomal dominant disorder. The growth rates of VSs are highly variable [1, 2], but generally increase with decreasing patient's age at onset of symptoms [3, 4]. The younger age correlates with larger tumour size [3,4,5]. Mutations in the NF2 tumour suppressor gene, which is located on chromosome 22, band q11-13.1, are the cause of neurofibromatosis 2, and occur in the majority of sporadic VSs [3, 6,7,8]. Agnihotri et al. have identified additional point mutations in [9]. have identified additional point mutations in sporadic schwannomas, including ARID1, DDR1, TAB3, ALPK2, CAST, TSC1 and TSC2.

Despite the huge diversity in the genes implicated in tumourigenesis, TP53 stands out as a key tumour suppressor and the p53 transcription factor (encoded by TP53 gene) serves a master regulator of various signaling pathways involved in this process. Phosphorylation of p53 at Ser 392 is suggested to be critical for its function [10]. A report indicated that Ser 392 phosphorylation was the more frequent modification among other possible phosphorylation sites of p53 in human tumour tissues and tumour-derived cell lines [11]. Sequence analysis of TP53 gene in various tumours revealed point mutations in the coding region in majority of the cases. Point mutations of TP53 have been identified in brain tumours including astrocytomas [12] and gliomas [13], but very rare in schwannomas [14,15,16], with one exception, where a point mutation was found in a malignantly transformed vestibular schwannoma [17]. The previous literature concerning schwannomas has used direct DNA sequencing, which detects point mutations including small intraexonic deletions/insertions; however, this technique is not able to detect deletions/duplications encompassing ≥ 1 TP53-exons. These lesions have been reported to be associated with several conditions, such as Li-Fraumeni syndrome [18], Breast Cancer [19] and transitional-cell carcinoma [20]. Monoh et al. Analyzed loss of heterozygosity (LOH) at the coding region of TP53 and demonstrated no LOH in 13 informative cases [15]; however, a conflicting result has been reported in another research group, which indicated that LOH occurred at the first intron of TP53 locus in about half of the informative cases [21]. To the best of our knowledge, no study has been made so far regarding the exonic copy number changes of TP53 in sporadic schwannomas.

The objectives of this study were to use Multiplex ligation-dependent probe amplification (MLPA), a PCR-based semiquantitative gene dosage assay, to characterize the pattern and frequency of copy-number changes of TP53 in patients with sporadic vestibular schwannomas and to examine the association between the TP53 mutations, phosphorylation status of p53 protein and clinical phenotypes.

Materials and methods

Study population

A total of 44 patients who presented with a unilateral vestibular schwannoma were treated surgically in our department between March 2018 and September 2018. The duration between the onset of the first symptoms and preoperative clinic visits was 2.14 ± 2.32 years. Tumour volume of a vestibular schwannoma by magnetic resonance imaging (MRI) was calculated according to the following method [22]. Maximum tumour diameter was measured in three dimensions (transverse and longitudinal on the axial images, and vertical on the coronal images) using fine calipers. Measurements were defined as x, y, and z, respectively. Tumour volume was xyz/2. Tumour size was evaluated by both maximum tumour diameter and tumour volume. All patients had at least one MRI examination before surgery (Table S1, Supporting Information). For patients with MRI scans at least 6-month intervals, their tumour growth rates were calculated on the basis of the measured changes in the maximum axial diameter in millimeters per year. Informed written consent was obtained from all patients donating tissue. Five cases of vestibular nerves from vestibular neurectomy for Meniere’s disease were included as controls. This study was conducted in compliance with institutional policy to protect patients’ private information, and was approved by the Institutional Review Board of the ethics committee of Shanghai Jiao Tong University.

Polymerase chain reaction (PCR) and sanger sequencing

Sanger sequencing was conducted to detect microlesions in the genes. DNA extraction from the tumour (fresh and paraffin fixed) specimens was performed using the TIANamp Genomic DNA Kit (Tiangen Biotech, Beijing, China). The whole coding sequence and exon–intron boundaries of the genes were amplified by the polymerase chain reaction (PCR) using standard methods. The sequence data were analyzed using the Sequencer 4.9 software (Genecode, MI, USA) and compared with the sequences of NF2 (NM_016418) and TP53 (NM_016418) in GeneBank. Mutations were described according to the standard nomenclature for DNA sequence changes reported by the Human Genome Variation Society (HVGS).

Multiplex ligation-dependent probe amplification analysis (MLPA) analysis

To identify exonic alterations in the TP53 locus, we used a commercial MLPA kit for analysis (SALSA P056 TP53; MRC-Holland, Amsterdam, The Netherlands). The probemix contains probes for each of the exons of TP53. The MLPA PCR products of test samples were separated on a 3500 Genetic Analyzer (Applied Biosystems, CA, USA), and peak heights for each PCR product were compared to a normalized average of 3 independent nerves to determine the dosage quotient (DQ) for each individual exon. A range of 0.7 < DQ < 1.2 was considered normal, and 0.4 < DQ < 0.6 was considered to show a heterozygous deletion. A subset of probes located on the regions of 17q 13 were also included as positive controls. Eleven reference probes were also used to detect different autosomal chromosomal regions.

Immunoblotting analysis

The tumour tissues were ultrasonicated in Cell Lysis Buffer (Beyotime, Shanghai China) containing a protease inhibitor (phenylmethanesulfonyl fluoride). Immunoblotting analyses were performed with antibodies specific for total p53 (# P6874, Sigma-Aldrich), phospho-p53-Ser 392 (#9281, Cell Signaling) and cyclinD1 (#2922, Cell Signaling). The β-actin antibody (#AA128, Beyotime) was used to ensure equal loading of total protein. The protein bands were detected using a chemiluminescence HRP substrate (Millipore). The band densities were quantified using Imager Lab Software (Bio-Rad). The band intensities of target proteins (p53, phospho-p53 and cyclinD1) in tumour specimens were expressed relative to those in the nerve controls to obtain relative values.

Cell culture and immunofluorescence

Human Schwann cells (HSCs) were purchased from ScienCell Research Laboratories (catalog no., 1700), and cultured in Schwann Cell Medium (ScienCell, Cat. No.1701). For primary cultures, schwannomas were cut to 1 mm3 in size and digested with 0.25% trypsin as previously described [3]. Cells were then collected and re-suspended in the Schwann Cell Medium. Schwannoma cells were plated on glass slides, fixed with 4% paraformaldehyde, and then permeabilized in 0.3% Triton X-100. The glass slides were blocked with bovine serum albumin before the incubation with an anti-phospho-p53 (pSer392) antibody (# ab134190, Abcam). An Alexa-594–conjugated anti-rabbit antibody was used as the secondary antibody. The Alexa Fluor® 488 Phalloidin (# A12379, Thermo) was used to stain the F-actin cytoskeleton in schwannoma cells. Finally, the sections were nuclear counterstained with 4, 6-diamidino-2-phenylindole (DAPI, Beyotime, China).

Statistical analysis

Results are expressed as mean ± SD ranges. Normally distribution and homogeneity of variance tests were performed on the data using SPSS statistics software (SPSS, Chicago, IL). The statistical analyses of clinical/ genetic parameters and expression/phosphorylation profiles were assessed using Chi-square test, two-tailed t test, bivariate correlations (Pearson’s for continuous variables), and multivariate linear regression. P values of less than 0.05 were considered to be statistically significant.

Results

Patients

The cohort of patients consisted of 16 males (36.4%) and 28 females (63.6%) with a median age at diagnosis of 47.8 years (range from 11 to 69 years). The most frequent symptoms were hearing impairment (31 patients, 70.5%) and tinnitus (23 patients, 52.3%). Details regarding the clinical characteristics and the results of Sanger sequencing in all tumours analyzed were given in Table 1. Mutations throughout the whole coding sequence of the NF2 gene were detected in 30 (68.2%) out of 44 cases. The majority of these mutations were truncating mutations, which were distributed along almost the entire NF2 sequence. As expected, no point mutations of TP53 were observed in schwannomas by Sanger sequencing.

Table 1 Clinical and genetic information of patients with sporadic vestibular schwannomas
Full size table

Age-dependent multi-exon deletion mutations of the TP53 gene in sporadic VSs

The exonic copy number changes of TP53 in the tumour were investigated by MLPA analysis. As shown in Figs. 1 and 2, the heterozygous exonic deletion mutations of TP53 were identified in 29/44 tumour samples (65.7%). Of these deletions, nine were single exon and twenty were multi-exons. The frequency of mutation in a specific exon was determined using the methylation values, defined as the number of tumour mutated /the number of tumour tested. These mutations were more likely to occur in exons 9, 1, 6, with respective frequencies of 50.0%, 45.5%, and 45.5%. Exons 2, 4 and 11 had lower mutation frequencies (27. 2%, 25.0%, and 2.3%, respectively), and none of the mutations were found in exons 7 and 10.

Fig. 1
figure 1

MLPA analyses of copy number changes of TP53 in schwannomas. In the graph, the y-coordinate represented the Dosage Quotient (DQ) and 0.4 < DQ < 0.6 was considered to have a heterozygous deletion. MLPA showed one tumour (# T363) with normal TP53 alleles and one tumour (# T351) with heterozygous deletions of exon 1, 6, and 9 at the TP53 locus

Full size image
Fig. 2
figure 2

Clinical and genetic characteristics of patients with sporadic VSs aged from 21 to 65 years. The clinical parameters including the age at diagnosis, gender and tumour size were present in the upper panel. Point mutations in the NF2 gene were demonstrated in the majority of tumours, whereas no point mutations of TP53 were found by direct sequencing. The mutation index (MI) for each tumour (bottom) and the mutation frequencies (indicated as ‘mutation values’; right) for each exon of TP53 were shown

Full size image

An overview regarding clinical and genetic characteristics of mutated (n = 29) and non-mutated (n = 15) patients is shown in Table 2. There was no predilection for sex, tumour side or tumour type of the lesion in both groups. Notably, we found that patients with mutant TP53 had an earlier age at diagnosis (mean 25.4 ± 9.5 vs. 55.2 ± 12.1 years, P = 0.001) compared to patients with no TP53 mutations, and mutation carriers were more likely to suffer from larger tumours (maximum diameter, mean 26.8 ± 7.4 vs. 22.4 ± 8.9 mm, P = 0.120; volume, mean 7.0 ± 5.6 vs. 5.6 ± 6.0 cm2, P = 0.460). The mutation index (MI), defined as the number of TP53-exons mutated/ the number of TP53-exons tested in a tumour, showed a negative correlation with the patient's age (Fig. 3a; Spearman’s rho = -0.35, P = 0.02). Moreover, there were a significant positive correlation between MI and both maximum tumour diameter (Fig. 3b; Spearman’s rho = 0.42, P = 0.002) and tumour volume (Fig. 3c; Spearman’s rho = 0.34, P = 0.024). Younger the age of patients or large tumor size is known to be associated with higher growth rate of VSs [3,4,5]. The growth rates of three TP53-mutated patients (T351, T360 and T380; marked in red; Table S1) was 3.7 ± 3.2 mm/ year, while the growth rates of three non-mutated patients (T365, T422 and T406; marked in green) was 0.63 ± 1.1 mm/ year. These patients underwent MRI studies at least 6-month intervals. Although the sample size was very small, the data showed that TP53-mutated tumours grow faster than non-mutated counterparts (p = 0.087). Thus, exonic deletion mutations of TP53 may play a role in the acceleration of tumour growth, thus contributing to an early age at symptom in patients with sporadic VSs.

Table 2 Clinical characteristics in patients with or without exonic deletion mutations of TP53
Full size table
Fig. 3
figure 3

Correlation between the mutation index of TP53 and both tumour size and patient's age. The mutation index showed a significant correlation with the patient's age (a) and tumour size, which was measured by the maximal tumour diameter (b) and tumour volume (c)

Full size image

Phosphorylation patterns of p53 in correlation with mutational status of TP53 in schwannomas

Function of the wild-type p53 protein is known to be dependent on the conformation and phosphorylation of its serine residues, Ser 392 in particular [10]. High levels of phospho-Ser392-p53 have been observed in tumours with mutant p53 [23]. Next, we investigated if TP53 mutations influenced Ser 392 phosphorylation of p53 in schwannomas. As shown in Fig. 4a–c, the levels of total/phosphorylated p53 and its downstream cyclin D1 varied between the 26 cases of tumour samples. The band intensities of target proteins in tumours were expressed relative to those in the nerve controls to obtain relative grey scale values. As indicated in Fig. 4d, the phosphorylation levels of p53 were changed in positive correlation to its total protein levels (r2 = 0.6, P < 0.001), suggesting that p53 was frequently phosphorylated at Ser 392 in schwannomas. There was a highly negative correlation between the patient's age and levels of both total and phosphorylated p53 (r2 = 0.21, P = 0.018; r2 = 0.27, P = 0.006; respectively). The mutation index (MI) and patient's age were also correlated (r2 = 0.29, P = 0.043). Thus, an age-dependent phosphorylation pattern of p53 in correlation with its mutation status was suggested in schwannomas. No statistically significant correlation was observed between the level of cyclin D1 and patient's age (Spearman’s rho = 0.08, P = 0.71). As shown in Fig. 4a–c, four cases (T364, 384, 406 and 402) exhibited no marked levels of total/ phosphorylated p53; these patients were above 50 years of age and 3 out (75%) of them had no TP53 mutations. By contrast, an appreciable amount of phospho-Ser392-p53 was observed in five mutated tumours (T381, 378, 424, 385 and 408), most of which were from young patients with an average age of 33.8 years. Altogether, these findings indicated that Ser 392 phosphorylation of p53 in correlation with its mutation status was more likely to occur in young patients.

Fig. 4
figure 4

Expression and phosphorylation status of p53 protein in schwannomas with and without TP53 mutations. a–c The expression levels of cyclinD1 and both total and phosphorylated p53 were investigated by immunoblotting analysis. d The band intensities of target proteins in tumours were expressed relative to those in the nerve controls to obtain relative grey scale values. The correlations of the patient's age with both relative grey scale values and mutation index of TP53 were analyzed

Full size image

Nuclear accumulation of p53 protein associated with its Ser 392 phosphorylation in mutated tumours

It was reported that the modification of Ser 392 could stabilize p53 by blocking the nuclear export of p53, thus inhibiting proteasome-mediated degradation via MDM2 [24]. To characterize the subcellular localization of phospho-Ser392-p53 in schwannomas, we performed fluorescence analyses of cultures from one mutated tumour (#T381) with hyperphosphorylated p53, one non-mutated tumour (#T406) with underphosphorylated p53 and human Schwann cells (HSCs). The phospho-Ser392-p53 was predominantly stained in the nucleus of HSCs (Fig. 5, upper). The diffuse nuclear and cytoplasmic staining of phospho-Ser392-p53 was observed in the schwannoma cultures # T406 (Fig. 5, center). Additionally, this finding was consistent with a previous study showing a diffuse pattern of p53 staining with a p53-specific antibody in schwannomas [25]. By contrast, schwannoma cells from tumour # T381 exhibited strong localization of phospho-Ser392-p53 to the nucleus (Fig. 5, lower). Thus, we demonstrated an enhanced stability of phospho-Ser392-p53, which was characterized by nuclear accumulation in schwannoma cultures with p53 mutations.

Fig. 5
figure 5

The subcellular localization of Ser 392 phosphorylated p53 in schwannoma cultures and human Schwann cells (HSCs). Phospho-Ser392-p53 was predominantly stained in the nucleus of HSCs (upper). The diffuse nuclear and cytoplasmic staining of phospho-Ser392-p53 was observed in the schwannoma cultures # T406 (center). By contrast, schwannoma cells from tumour # T381 exhibited strong localization of phospho-Ser392-p53 to the nucleus (lower)

Full size image

Discussion

Sequence analysis have revealed that the point mutations of TP53 are the most frequent alterations in the transitional stage during which a tumour progresses; these mutations are distributed in all coding exons of TP53 [26]. Our results confirmed the previous reports on the absence of point mutations at TP53 locus in schwannomas; instead exonic deletion mutations were noted in 64% of tumours. Unlike the malignancies where point mutations of TP53 frequently occur in exons 5 through 8 [26], the exonic deletion mutations in schwannomas were demonstrated to be clustered within exons 9, 6 and 1. There are several reports on the role of TP53 alterations in relation to the clinical/pathological features of tumours. Niederacher et al. correlated LOH [27]. correlated LOH of TP53 with younger patients and a higher histological grading in sporadic endometrial cancer. NF2-TP53-double mutant mice exhibited an increased predisposal to neoplasms compared with each of the single mutant mice [28]. In a cohort of meningiomas, whose tumourigenesis was also related to NF2 mutation, TP53 polymorphism increased the risk of tumour progression [29]. Consistent with these studies, we found that schwannomas with exonic deletion mutations of TP53 exhibited rapid growing behaviors including larger tumour size and younger patient's age. It has been reported that the wild-type p53 functions optimally when it binds to DNA as a tetramer; the wild type and mutant p53 can form hetero-tetramers of attenuated activity, via their tetramerization domains [30, 31]. In this mechanism, the severer phenotype of TP53-mutated schwannomas might be explained by the fact that the mutant p53 could disturb a functional tetramer and was therefore able to override wild-type p53 tumor suppressor activity.

Phosphorylation of p53 at Ser 392 can activate the specific DNA binding function of p53, presumably by stabilizing its tetramer [10]. Ser 392 phosphorylation of mutant p53 protein might promote its dominant negative effects through hetero-oligomerization, thereby contributing to the aggressive behaviors of tumours [32]. An increased percentage of the Ser 392 phosphorylated form of p53 was found only in histological high-grade gliomas (grades III and IV). On the contrary, a lower percentage of the Ser 392 phosphorylated form of p53 was present in histological low-grade gliomas, meningiomas and other brain tumour types [33]. In present study we suggested that hyperphosphorylated p53 appeared to occur more often in mutated schwannomas in young patients. Down-regulation of p53 protein is commonly seen in schwannoma tissues and in vitro cultures [25, 34]; these studies showed that p53 was abnormally sequestered in the cytoplasm of p53-deficient schwannomas. It has been suggested that merlin neutralizes the inhibitory effect of MDM2 on p53 [35], and, merlin loss by NF2 mutation confers an enhanced nuclear accumulation of MDM2 and thereby induces nuclear export of p53 for degradation in schwannomas [25]. We show that p53 deficiency along with its hypophosphorylation at Ser 392 is frequently observed in schwannomas with no TP53 mutation. In addition, some studies have reported a subset of schwannomas characterized by p53 overexpression [14, 21, 25]. This current study suggests that this deregulation may be independent of merlin loss and linked to the occurrence of exonic deletion mutation of TP53. It is possible that the deficiency of wild-type p53 due to merlin loss by NF2 mutations may play a role on the early biological changes during schwannnoma tumnourigenesis. The exonic deletion mutation of TP53 may occur as a second genetic hit, leading to the phosphorylation and stabilization of p53 protein with attenuated tumour-suppression activity, thus promoting tumour growth and resulting in early clinical symptoms.

Using several lines of investigation, therefore, our studies propose a connection between the mutation pattern of TP53 and phosphorylation status of p53 and clinical characteristics in human sporadic VSs. These studies deserve further investigations to uncover mechanisms underlying the biological effect of TP53 mutation characterized by Ser 392 phosphorylation of p53 on the growth of schwannoma. Unlike genetic mutations, phosphorylation is a reversible process, and thus countering the effects of Ser 392 phosphorylation of p53 protein could be a novel target for therapeutic purposes to reduce the growth of schwannomas.