Introduction

Retinoblastoma (Rb; OMIM 180200), the most common primary intraocular malignancy in children, has a worldwide incidence rate of 1:15,000—20,000 live births [1]. About 95% of Rb cases are diagnosed before the age of five [2]. In developed countries, due to effective modern treatment, this disease is associated with a lower mortality rate and improved outcomes. In contrast, in developing countries, due to lack of prompt diagnosis and effective treatment, Rb patients have a 40–70% mortality rate and an increased risk of enucleation [3]. In addition, Rb survivors have higher risks of secondary tumors [4], including bone and soft tissue sarcomas, melanomas, leukemia, lung cancer, uterine leiomyosarcoma, and radiotherapy-related central nervous system tumors [5]. Secondary tumors adversely affect the quality of life and survival rate of Rb patients.

Mutations in RB1, the first described tumor suppressor gene located on chromosome 13q14.2, have been associated with heritable Rb [6, 7]. Mutated RB1 gene may result in a loss of function of the retinoblastoma protein (pRB). Moreover, pRB knockdown was shown to promote tumor cell proliferation and repressed E2F activities [8]. Based on Knudson’s two-hit hypothesis, Rb development is mainly caused by mutations of both RB1 alleles [9]. Individuals with RB1 germline mutations are genetically predisposed to Rb [3]. Moreover, the types and frequencies of mutated RB1 in children developing Rb vary geographically. Therefore, identification of RB1 (Gene ID:5925; NM_000321) is key in genetic Rb testing and screening [10]. Genetic testing of the RB1 gene is inevitable for early diagnosis in heritable Rb to provide timely therapies and improve prognosis [11]. However, in China, limited data are available on RB1 mutations for prompt screening of all RB1 mutations, avoiding false negatives in genetic testing of Rb patients, and informing the provision of effective therapies.

In this study, we characterized the spectrum of RB1 mutations in children with Rb in China. Furthermore, the data generated in this study will form the basis for effective genetic counseling of Rb patients and their families, thereby improving Rb diagnosis and prognosis.

Material and methods

Patients

This study included 114 patients with retinoblastoma from Southern China. Clinical diagnosis of retinoblastoma was based on the clinical features and confirmed by an expert pathologist. All patients with Rb were diagnosed and treated according to the latest China guidelines for the diagnosis and treatment of retinoblastoma. Clinical data of all patients were collected from the hospital information system, including age, gender, tumor laterality, age at diagnosis, and associated symptoms (Table 1). All participants gave their written informed consent. The study was approved by the Ethics Committee of The Fifth Affiliated Hospital, Sun Yat-sen University, and was conducted in accordance with the principles of the 1975 Declaration of Helsinki.

Table 1 Clinical characteristics of the Rb patients
Full size table

Mutation detection

PCR

Genomic DNA was extracted from blood samples of patients firstly diagnosed with Rb using the iPure blood genomic DNA kit (IGEbio, CZ313-S). The extracted DNA was quantified using Nanodrop 2000 (Nanodrop, Wilmington, DE, USA). The primers used for detecting RB1 mutations (shown in Supplementary Table S1) were designed using Primer Premier 3.0 software for 26 translating exons and the adjacent intron–exon regions of the RB1 gene, except exon 15. The PCR had a total volume of 20 µl, including 10 µl HiFipfu polymerase (IGEbio, P300-S), 7 µl nuclease-free water, 1 µl forward and reverse primers (10 uM) each, and 2 µl DNA template. A BIO-RAD T100 TM Thermal cycler (Bio-Rad, Gladesville, Australia) and a touchdown PCR protocol were used for amplification. The PCR conditions were as follows: initial denaturation at 94 °C for 5 min followed by 10 cycles of denaturation at 94 °C for 30 s, annealing at 65 °C for 30 s, with the temperature of the heating block decreased by 1 °C per cycle, and then extension at 72 °C for 40 s. This was then followed by 25 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 40 s and final extension at 72 °C for 3 min.

Sanger sequencing and sequence analysis

The PCR products were visualized by 1.5% agarose gel electrophoresis. The target band was excised, after which the DNA was purified and sequenced with ABI3730XL (Applied Biosystems, Foster City, CA, USA) using the BigDye method. Nucleotide sequences of the RB1 gene (GenBank L11910.1) were used as the reference. A Sequence Scanner, version 1.0 (Applied Biosystems, Streetsville, Ontario, Canada), and chromas (2.6.5) were used to align each exon’s sequence with the reference sequence to annotate mutations. Variants were described based on the latest nomenclature for the description of sequence variants (Human Genome Variation database: http://www.HGVS.org/varnomen). Additional information on RB1 gene mutations was obtained from Clinvar (https://www.ncbi.nlm.nih.gov/clinvar/), Human Gene Mutation Database (HGMD: http://www.hgmd.cf.ac.uk/ac/), LOVD (http://www.lovd.nl/3.0/home), gnomAD (http://gnomad.broadinstitute.org/), ensemble, and dbSNP. pRB protein structures were retrieved from UniProt (http://www.uniprot.org/uniprot, P06400). Open reading frame (ORF) and frameshift mutation predictions were performed using the NCBI ORF finder (https://www.ncbi.nlm.nih.gov/orffinder). The pathogenicity of variants was interpreted according to ACMG guidelines.

Statistical analysis

Student’s t test was performed to compare differences between continuous variables. Categorical variables were compared using Chi-square or Fisher’s exact tests. p ≤ 0.05 was considered statistically significant. Data analysis was performed using the SPSS software, version 25.0.

Results

Clinical characteristics

A total of 114 Chinese patients with Rb were recruited for this study, including 68 males and 46 females (59.6% vs 40.4%). A summary of the cases and their clinical characteristics are shown in Table 1. Fifty-five patients were bilaterally affected, while fifty-nine patients were unilaterally affected (48.2% versus 51.8%). In detail, unilateral cases were observed in 40 males and 19 females, while bilateral cases were observed in 28 males and 27 females (p = 0.0859). In unilateral cases, distributions of left and right ocular lesions were relatively similar (30 versus 29, respectively). The age at diagnosis ranged from 1 to 64 months old, with an average age of 14 months. Moreover, 91 patients (79.8%) had Rb onset before the age of 2 years, while 23 patients (20.2%) had Rb onset after 2 years of age. Most of the bilaterally affected patients (53/55, 96.4%) were diagnosed before 2 years. Mean age at diagnosis for patients with bilateral Rb was lower than for unilateral cases (11 ± 7 months versus 20 ± 14, p < 0.0001). This finding is consistent with those of previous studies [3, 12, 13]. Leukocoria was the most common initial symptom in Rb cases, accounting for 65.8% of all cases, followed by squint (21.9%). Several patients presented with other symptoms, including redness, accompanied by tearing, vision diminution, calcification, and white spots.

Mutations

Novel mutations

A total of five novel mutations were identified in our Rb cases. Sequencing analysis of the five novel variants is presented in Fig. 1. There were four frameshift mutations and one splice site mutation consisting of c.180_187del, c.528del, c.2035_2039del, c.2299_2300del, and c.1050-2A > T. All mutations were identified in bilaterally affected Rb patients with a mean age at diagnosis of 13.6 months. The splice site mutation, c.1050-2A > T, presenting at 3’ splice site of intron 10, altered the sequence from AG to TG and disrupted the canonical splice site. Intriguingly, the four novel frameshift mutations were predicted to result in a truncated RB protein arising from a premature stop codon in the reading frame and the loss of coding exons. Furthermore, three of the four frameshift mutations resulted in premature termination of amino acid translation in the conserved pRB pocket region, encoding transcriptional regulation and forming the repressor motif [14]. As shown in Fig. 2, c.180_187del (p.Cys61Ilefs*46) and c.528del (p.Gln176Hisfs*10) mutations occurred in exon2 and exon5, causing a frameshift variant and a premature stop codon at amino acid 106 and 185, and truncating 822 and 743 amino acids, respectively, at the C terminus of the RB1 protein. In addition, c.2035_2039del (p.Ile679Leufs*11) was located at domain B of the pRB pocket, resulting in premature translational termination at position 689, subsequently truncating 239 amino acids of the predicted protein. The other novel frameshift mutation, c.2299_2300del (p.Asn767Tyrfs*27), generated a frameshift and premature termination of the open reading frame.

Fig. 1
figure 1

Sequencing chromatograms illustrating novel mutations in the patients. The left panels show the mutations identified in Rb patients, while the right panels show the wildtype RB1 sequences. Arrow indicates the mutated site. A c.180_187del heterozygous mutation in patient Rb12. B c.528del heterozygous mutation in patient Rb119. C c.2035_2039del heterozygous mutation in patient Rb47. D c.1050-2A > T heterozygous mutation in patient Rb7. E c.2299_2300del heterozygous mutation in patient Rb81

Full size image
Fig. 2
figure 2

Effects of the four novel frameshift mutations on the RB1 protein. c.180_187del (p.Cys61Ilefs*46) mutations caused a frameshift change and a premature stop codon at amino acid 106, truncating 822 amino acids at the C terminus of the RB1 protein. c.528del (p.Gln176Hisfs*10) mutations caused a frameshift change and a premature stop codon at amino acid 185, truncating 743 amino acids at the C terminus of the RB1 protein. c.2035_2039del (p.Ile679Leufs*11) mutations caused a frameshift change and a premature stop codon at amino acid 689, truncating 239 amino acids at the C terminus of the RB1 protein. c.2299_2300del (p.Asn767Tyrfs*27) mutations caused a frameshift change and a premature stop codon at amino acid 793, truncating 135 amino acids at the C terminus of the RB1 protein

Full size image

Other mutations

Twenty-five other previously reported RB1 mutations we found in this study are shown in Table 2. The most common mutational type was nonsense mutations (fourteen of twenty-five; 56.0%), followed by frameshift (three of twenty-five; 12.0%), splice mutations (three of twenty-five; 12.0%), and missense mutation (three of twenty-five, 12.0%).

Table 2 A summary of RB1 mutations in the 30 Rb patients
Full size table

As shown in Table 3, in a total of 30 variants, the most commonly observed RB1 gene alteration was nonsense mutation, accounting for 46.7% of the detected mutations, followed by frameshift mutations (26.6%), splicing alterations (16.7%), and missense mutations (10.0%). Mutations located at exons 8, 18, 20 exhibited the highest mutation frequencies and were, respectively, found in 3 patients. Mutations at exons 1, 13, 14, 17, 22, and intron 21 were, respectively, found in 2 patients. There were no mutations in exons 6, 7, 9, 11, 16, 22, and 24–27. Furthermore, 50.0% (15/30) of the detected mutations were located at regions encoding two conserved domains, A (residue 373–573) and B (residue 646–765), of the pRB pocket containing five nonsense, four splicing, four frameshift, and two missense variants. Specifically, 4/5 splice site mutations and 2/3 missense mutations were located at the pRB pocket region. The mutation rate in domain A located at exons 12–18 was 26.7% (8/30), while the mutation rate in domain B located at exons 19–23 was 23.3% (7/30). In addition, 21/30 mutations caused a premature stop codon on RNA transcripts resulting in a truncated protein.

Table 3 Distributions of mutations in RB1 by type
Full size table

Of the 14 cases with nonsense mutations, nine (64.3%) patients had C to T transitions in CGA codons in exons 8, 10, 14, 18, with several of these mutations affecting more than two unrelated patients. For instance, the c.1735C > T (p.Arg579*) mutation in exon 18 was detected in three unrelated patients, while the c.763C > T (p.Arg255*) mutation in exon8 was detected in two unrelated patients. Additionally, c.2242G > T(p.Glu748*) was detected in the proband’s mother, who did not suffer from Rb, indicating that this mutation was maternally inherited. Four of the eight reported frameshift mutations were novel mutations that had never been previously reported or characterized in mutated RB1. Moreover, four of the five splicing mutations occurred by substitution of G to A/T. Cases with splice site mutation were diagnosed at a younger age, ranging from 1 to 8 months.

Genotype–phenotype correlations

Sequencing analysis was performed for the 26 exons and adjacent intronic regions, except exon 15. DNA analysis revealed 27 different causative RB1 mutations in 30 patients (26.3%), which were identified in 4/59 (6.8%) patients with unilateral Rb and 26/55 (47.3%) patients with bilateral Rb (p < 0.0001) (Fig. 3A). The mutation rate in the bilaterally affected patients was significantly higher than in the unilaterally affected patients, consistent with previous findings [15]. Mean age at diagnosis for patients with mutated RB1 was 13 ± 11 months and 16 ± 12 months for patients without mutated RB1. Differences in age at diagnosis between the two groups were not significant (p = 0.1807) (Fig. 3B).

Fig. 3
figure 3

Mutation detection rate and mean age at diagnosis in retinoblastoma patients. A The left pie chart shows an overall mutation detection rate of 26.3% (30/114) in all Rb cases. The middle pie chart shows the mutation rate of 47.3% (26/30) in bilateral Rb cases. The right panel shows a mutation detection rate of 6.8% (4/59) in unilateral Rb cases. B Age at diagnosis of probands with positive (n = 30) or negative (n = 84) RB1 mutation test results

Full size image

Discussion

In this study, we identified RB1 gene mutations in 26.3% of the 114 recruited Chinese Rb patients. This included five novel mutations that had never been previously reported. Bilaterally affected Rb patients exhibited higher mutation rates and an earlier age at diagnosis than unilaterally affected patients.

Most frameshift or nonsense mutations in inherited diseases or cancers result in premature termination codons (PTCs) [16]. In this study, nonsense and frameshift mutations were the major mechanisms of RB1 inactivation, accounting for 73.3% of all RB1 mutations. This finding is consistent with previously reported rates of 50–80.32% [4, 12, 17,18,19]. Furthermore, 50.0% of all RB1 mutations lie in the A/B “pocket” domain of pRB, which can bind E2F transcription factors and promote protein–protein interactions [20]. This mutation rate was comparable to previously reported rates of 40–78.95% [4, 12, 17, 21, 22]. In addition, subtle mutations may result in partial inactivation of pRB, decreasing protein stability. Therefore, variants occurring in or adjacent to known conserved protein domains have a great influence on disease onset [23].

Premature stop codons resulting in loss-of-function (LOF) alleles are strongly associated with inactivation of the RB protein [24]. Accordingly, frameshift mutations with premature translation termination might contribute to the onset of diseases by degrading RB1 transcripts, resulting in the complete loss of pRb. In this study, four of the five novel RB1 variants were frameshift mutations and were reported in patients with bilateral lesions. According to the ACMG standards and guidelines [25], all novel RB1 frameshift variants were pathogenic. This finding supports the postulate that frameshift mutations are highly associated with retinoblastoma disease. Studies should evaluate the specific impacts of these novel mutations on protein functions.

A marginally lower degree of RB1 germline mutation rate was reported in this study, compared to studies done in other countries [17,18,19, 26, 27]. Our findings are consistent with several studies conducted in China, with 21–27.1% of total RB1 mutation rates detected using Sanger sequencing alone [7, 28, 29]. Therefore, we hypothesized that geographical variations or race accounted for these differences. The heterogeneity in mutation detection rates can also be explained by the different techniques, such as MLPA and next-generation sequencing [30], used to detect mutations. In addition, the fact that we missed sequencing analysis of the promoter region and exon 15 due to the poly (A) and poly (T) sequence on both sides could also explain this difference.

Retinoblastoma is a highly complex disease. Single-nucleotide variants, large rearrangements, and promoter hypermethylation of the RB1 gene have a role in its development [31, 32]. Furthermore, low-level mosaic and deep intronic variants have been identified [33]. Besides, other genes may also be attributed to the Rb phenotype. For instance, MED4, a synthetic lethal target in tumors, explains the low penetrance of retinoblastoma [34]. In addition, amplification of the MYCN oncogene contributes to retinoblastoma [35]. Besides, retinoblastoma has also been associated with MDM2, the first modifier gene so far identified in retinoblastoma [36]. Further, the PIN1 gene that alters pRB phosphorylation is also important for Rb development [37]. Despite extensive research, the exact pathomechanisms of Rb remain elusive and should be investigated further. Given the relatively high negative rate of RB1 mutations in this study, we will perform whole-exome sequencing (WES) in samples obtained from patients without RB1 mutations to explore other possible disease-causing genes and variants. We postulate that our findings will elucidate the pathogenesis of Rb.

The spectrum of RB1 mutations in this study will facilitate the development of better-targeted therapies and inform on optimal disease management as well as effective treatment of Rb patients [38]. Prenatal diagnosis and family screening should be performed to identify individuals with a genetic predisposition to Rb [39].