Received 4 July 2007; accepted revised manuscript 28 September 2007.
Pathogenic Validation of Unique Germline Intronic Variants of RB1 in Retinoblastoma Patients Using Minigenes Angelo Gámez-Pozo1, Itziar Palacios1, Milica Kontic2, Ibis Menéndez3, Isabel Camino1, Purificación García-Miguel4, José Abelairas5, Ángel Pestaña1, and Javier Alonso1*
1OncoLab, Departamento de Biología Molecular y Celular del Cáncer, Instituto de Investigaciones Biomédicas “A. Sols”, CSIC-UAM, C/ Arturo Duperier 4, 28029 Madrid, Spain; 2School of Medicine, Institute of Biology and Human Genetics, Belgrade, Serbia; 3Instituto Nacional de Oncología y Radiobiología, 10400 La Habana, Cuba; 4Unidad de Hemato-Oncología Pediátrica, Hospital La Paz, 28046 Madrid, Spain; 5Unidad de Ofptalmología Pediátrica, Hospital La Paz, 28046 Madrid, Spain
*Correspondence to: Javier Alonso, OncoLab, Departamento de Biología Molecular y Celular del Cáncer, Instituto de Investigaciones Biomédicas “A. Sols”, CSIC-UAM, C/ Arturo Duperier 4, 28029 Madrid, Spain; Tel.: 34 91 585 4418; Fax: 34 91 585 4401; E-mail: [email protected] Grant sponsor: Fondo de Investigación Sanitaria (Grant number PI150128) and Ministerio de Educación y Ciencia (Grant number SAF2005-00946). Angelo Gámez-Pozo is a predoctoral fellow from the Ministerio de Educación y Ciencia, Itziar Palacios was supported by a contract from the Comunidad de Madrid and FGUAM, and Milica Kontic and Ibis Menéndez were supported with UICC grants. Communicated by Daniel F. Schorderet
Retinoblastoma (MIM# 180200), is a rare embryonic neoplasm of retinal origin, which has a relative incidence of 3% among all paediatric tumours and a frequency estimated between 1:28,000 and 1:15,000 live-born. Most of the clinical phenotypes can be explained by the double mutational inactivation of the retinoblastoma susceptibility gene (Cavenee, et al., 1983), the prototype tumour suppressor gene that controls cell cycle progression (Kaelin, 1999). In the hereditary form of the disease, a germ line mutation is transmitted as a highly penetrant autonomic dominant trait, resulting in a 45% risk in the offspring of patients with hereditary retinoblastoma, while the second inactivating mutation occurs in retinal cell precursors (Knudson, 1971). Most of these patients have bilateral retinoblastoma and a mean age at diagnosis of 12 months. In the non-hereditary form of the disease, both inactivating events occur during somatic development of retinal cells and this results in the relatively late onset of a single tumour in one eye (Draper, et al., 1992). However, nearly 15% of the unilaterally affected patients have germ line RB1 mutations with the associated 45% risk in their offspring. Considering these situations together, the hereditary form represents nearly 50% of all the retinoblastoma patients.
The development of sensitive and reliable genetic tests to detect RB1 mutations has greatly improved the identification of carriers, facilitating accurate genetic counselling. Attempts by several groups to define the mutations that cause the inactivation of the RB1 gene in retinoblastoma have led to the identification of more than 600 distinct somatic or germ line RB1 mutations, and more than 400 recurrences that have been logged in two locus-specific public databases: the Retinoblastoma Gene Mutation database (RBGMdb, http://www.es.embnet.org/Services/MolBio/rbgmdb/idx.html) (Valverde, et al., 2005); and the Retinoblastoma Genetics Website (RBGW, http://rb1-lsdb.d-lohmann.de/) (Lohmann, 1999). Splicing mutations in RB1 are moderately recurrent (19%), doubling the proportion of splice-site mutations found in the Human Gene Mutation Database (HGMD, http://www.hgmd.org) (Stenson, et al., 2003). Most of these RB1 mutations occur at the invariant GT and AG dinucleotides and cause exon skipping or alternative splicing of the pre-mRNA, resulting in defective and oncogenic pRb. For the remainder of the mutations affecting the degenerate consensus sequences flanking the invariant dinucleotides or the branch site sequence, the consequences for the pre-mRNA splicing are not so obvious and require additional biological assessment (Baralle and Baralle, 2005). Recent evidence suggests that many sequence variants, including silent exon mutations and intronic SNPs, might influence the accuracy of splicing and they should thus be routinely evaluated to assess their possible consequences on pre-mRNA processing (Cartegni, et al., 2002). One approach is to test the effects of the mutation on splicing using minigenes, in which the relevant genomic segment is cloned into a plasmid between an upstream ubiquitous transcriptional promoter and a downstream polyA signal for mRNA 3’ end termination (Cooper, 2005).
In the course of a seven year long program of RB1 mutational screening in retinoblastoma patients (Alonso, et al., 2005; Alonso, et al., 2001a; Kontic, et al., 2006), we have encountered several unique intronic variants whose oncogenic significance have now been assessed using minigenes and in silico tools. Our results demonstrate that minigenes accurately define the oncogenic character of intronic variants and distinguish them from non-pathogenic polymorphisms. This strategy is especially relevant when RNA from lymphocytes (or whatever other source) is not available for analysis, or when aberrant RNA can not be observed as a consequence of nonsense mediated decay (Lejeune and Maquat, 2005).
MATERIALS AND METHODS
Subjects and exon-by-exon RB1 mutational screening Intronic variants were selected from mutational screens carried out on more than 200 retinoblastoma patients
from Spain, Cuba, Colombia and Serbia (Alonso, et al., 2005; Alonso, et al., 2001a; Kontic, et al., 2006). RB1 mutational screening was performed as described previously (Alonso, et al., 2001a) and the nucleotide positions were numbered on the basis of the cDNA (GenBank accession number L41870.1) and genomic (L11910.1) reference sequences. When lymphocyte RNA was available, the presence of aberrant mRNA products was ascertained by RT-PCR analysis as described previously (Alonso, et al., 2001a).
Minigene construction
Minigenes were designed to include the exon of RB1 that was close to the intronic variant and both flanking exons. In order to include most of the element regulating splicing, the minigene was designed to include more than
Pathogenic Validation of Intronic Variants in Retinoblastoma 3
200 nucleotides of the upstream and downstream intronic sequences of each exon included in the minigene (Cooper, 2005). In addition, the size of the genomic fragments containing the three exons and their adjacent intronic sequences was restricted to less than 1500 nucleotides. Since most of the RB1 introns are longer than 1000 nucleotides, the genomic fragment to be cloned was usually comprised of 2 or 3 genomic fragments. These DNA fragments were obtained by PCR using normal or patient DNA as the template. The sequences of the primers used to construct the minigenes are available upon request. Primers contained restriction sites at their 5’ end to facilitate cloning. PCRs were performed in a reaction mixture of 25 µl containing 100 ng of genomic DNA, 200 µM of each dNTP, 2 mM MgCl2, 0.1 µM of each primer, 1x PCR buffer and 1 U of Taq DNA polymerase (Biotools, Madrid, Spain). The PCR reactions involved a single denaturation step followed by 5 cycles of: denaturation at 94ºC for 60 seconds; annealing at 54ºC for 30 seconds and extension at 72ºC for 60 seconds, and then 35 cycles of: denaturation at 94ºC for 30 seconds, annealing at 54ºC for 30 seconds and extension at 72ºC for 60 seconds, followed by a final extension step of 7 minutes at 72ºC. Amplicons were visualized by agarose gel electrophoresis and cloned into the pGEM-T (Promega, Madison, WI). Finally, DNA fragments containing the exon (or exons) and its adjacent intronic sequences were subcloned into the expression vector pCIneo (Promega). All cloned genomic DNA fragments and minigene constructs were fully sequenced to confirm their identity.
Minigene expression assay
A total of 3.5 x 105 293 cells were plated in six-well plates and grown in DMEM supplemented with 10% foetal calf serum. After 24 hours, the cells were transfected using FuGene6 (Roche Applied Science, Penzberg, Germany) at a 3:1 ratio FuGene6:DNA. Forty eight hours later, total RNA was isolated with TRI-reagent (Sigma-Aldrich, St. Louis, MI) and minigene expression was analyzed by RT-PCR. First-strand cDNA was synthesized from 1 µg of total RNA in a 20 µl reaction containing RT Buffer, 200 µM of each dNTP, 10 mM DTT, 2.5 µM random hexamers, 200 units of MMLV reverse transcriptase (Invitrogen, Paisley, UK) and 40 units of RNasin (Promega). The reaction was incubated at 42º C for 30 minutes and then terminated by incubating at 95º C for 5 minutes. The PCR reactions were performed as described above using 3 µl of the RT reaction as template. The sequences of the primers used to detect minigene expression were: pCI-F, 5’-GTTCAATTACAGCTCTTAAG-3’; and T3-R, 5’-AATTAACCCTCACTAAAGGG-3’, which hybridize to both sides of the cloned minigene in the pCIneo expression vector. Expression of the TATA-binding protein (TBP) was used to control that similar amounts of RNA were analyzed in each sample. The specific primers for TBP were: TBP-F, 5’- CAGCCGTTCAGCAGTCAAC -3’ and TBP-R, 5’- TGTTGGTGGGTGAGCACAAG -3’. The RT-PCR products were visualized by agarose gel electrophoresis, purified and sequenced to confirm their identity.
In silico analysis The Genscan program (http://genes.mit.edu/GENSCAN.html) was used to predict the location and exon-intron
structure of wildtype and mutant genomic sequences (Burge and Karlin, 1997). Genscan assign meaningful probabilities to each predicted exon, providing a useful quantitative guide to the likelihood that a given exon is correct. This probabilty depends on global as well as local sequence properties, e.g., it depends on how well the exon fits with neighboring exons. A probabilty > 0.75 means that there is a high probability that the predicted exon is exactly correct.
Analysis of intron sequences for putative branch points and calculation of the strength of acceptor splice sites was performed with the ASD (Alternative Splicing Database) workbench (http://www.ebi.ac.uk/asd-srv/wb.cgi). In this tool, intron sequences are compared with a weight matrix constructed with a collection of sequences present in the ASD in order to compute a score. Perfect match with the consensus weight matrix will compute a score of 10 (Clark and Thanaraj, 2002; Stamm, et al., 2006).
RESULTS
The aim of this study was to determine the oncogenic character, using minigene approaches, of unique RB1 intronic variants of unclear pathogenic significance, and for which no previous functional data was available. In order to validate the minigene assays, we first investigated four bona fide mutations affecting invariable dinucleotides at splice sites (and thus expected to produce exon skipping), as well as a frequent neutral polymorphism.
Analysis of bona fide splicing mutations Figure 1A shows the effect of the c.1499-1G>C mutation on splicing using a minigene construct. A minigene
comprising RB1 exons 15, 16, 17 and 18 was constructed and transfected into 293 cells. Agarose gel electrophoresis showed that the RT-PCR product obtained from cells transfected with the mutant minigene was smaller than that obtained from cells transfected with the wild type minigene. Sequencing of this RT-PCR product demonstrated that the mutated minigene produced only mature mRNAs in which exon 17 was completely skipped, as expected for a splice mutation affecting one of the two invariable dinucleotides at the 3’ splice site. We analyzed the effect of this mutation on lymphocyte RNA from the patient (Fig. 1B). RT-PCR with primers flanking exon 17 produced a unique band that corresponded to the size and sequence of the normally processed mRNA. Interestingly, no RT-PCR products corresponding to mRNA with exon 17 skipping were observed, indicating that the mutant RNA was extraordinarily unstable, probably as a consequence of nonsense mRNA decay.
We analyzed three additional splice mutations affecting invariable dinucleotides at 5’ splice sites: c.607+1G>T, c.607+2T>G and c.2520+1G>A (Table 1). In each of these, the minigene assays accurately identified the complete skipping of the affected exon. For c.607+1G>T and c.607+2T>G total lymphocyte RNA from patients was also available for RT-PCR analysis. In contrast to the c.1499-1G>C mutant, mRNA lacking exon 6 was clearly detected in total lymphocyte mRNA isolated from these patients, although different levels of aberrant mRNA were observed in relation to the wildtype depending on the mutation (Table 1).
Table 1. Summary of results obtained in bona fide splicing mutations affecting invariable dinucleotides and the c.1390-14A>T neutral polymorphism
Variant (1) ID proband Minigene assay RT-PCR on lymphocyte RNA
RB0143 exon 17 skipping mutant mRNA not detected (nonsense
mRNA decay)
premature stop codon
c.2520+1G>A (exon 24/intron 24)
RB1663 exon 24 skipping no available premature stop codon
Polymorphisms
c.1390-14A/T (intron 14/exon 15)
- normal splicing no available none
(1) Reference sequences for variant nomenclature are L41870.1 (cDNA) and L11910.1 (genomic DNA). c.1 corresponds to the A of the ATG start codon in the reference sequence L41870.1. The intron/exon boundary where the variant is localized is also indicated.
Analysis of the neutral polymorphism c.1390-14A>T The intronic variant c.1390-14A>T is a polymorphism relatively frequent in the Spanish population (Alonso, et
al., 2001b). In order to confirm the true neutral nature of this intronic variant, we constructed a minigene comprising of exons 14, 15 and 16. Cells transfected with either the A or T alleles produced normally processed mRNAs with no evidence of either exon 15 skipping or alternative splicing (Fig. 1C, Table 1). These results confirm the neutral character of this intronic variant in accordance with the population studies.
Pathogenic Validation of Intronic Variants in Retinoblastoma 5
A B
proba
nd
mw
800 -
Exon 15 Exon 18Exon 17
c.1499-1G>C
Exon 16
wildtype
Exon 17 / Exon 16
Mutant: Exon 17 skipping
Exon 18 / Exon 16wild
type
mutant
FALTA EL GEL
mw
Exon 15Exon 14
c.1390-14A>T
Exon 16
allele A and T
Exon 14 / Exon 15 Exon 15 / Exon 16
allele
A
allele
T
mw
300 -200 -
CA B
proba
nd
mw
800 -
proba
nd
mw
800 -
Exon 15 Exon 18Exon 17
c.1499-1G>C
Exon 16
wildtype
Exon 17 / Exon 16
Mutant: Exon 17 skipping
Exon 18 / Exon 16wild
type
mutant
FALTA EL GEL
mw
Exon 15 Exon 18Exon 17
c.1499-1G>C
Exon 16
wildtype
Exon 17 / Exon 16
Mutant: Exon 17 skipping
Exon 18 / Exon 16wild
type
mutant
FALTA EL GEL
mwwildtyp
e
mutant
FALTA EL GEL
mw
Exon 15Exon 14
c.1390-14A>T
Exon 16Exon 15Exon 14
c.1390-14A>T
Exon 16
allele A and T
Exon 14 / Exon 15 Exon 15 / Exon 16
allele A and T
Exon 14 / Exon 15 Exon 15 / Exon 16
allele
A
allele
T
mw
300 -200 -
allele
A
allele
T
mw
300 -200 -
C
Figure 1. RNA splicing alterations caused by the bona fide splicing mutation c.1499-1G>C and the neutral polymorphism c.1390-14A>T. A: Schematic representation of the minigene used to analyse the effect of the c.1499-1G>C mutation on mRNA splicing. Arrows indicate the localization of the primers used to construct the minigene. The minigenes (wildtype and mutant) were transfected into 293 cells and the total RNA was analyzed by RT-PCR. The mutant minigenes yielded a unique band smaller to that obtained from the wildtype minigene. Sequencing of this RT-PCR product demonstrated that it corresponded to a mature mRNA in which exon 17 was entirely skipped. mw, DNA molecular weight marker. B: RT-PCR analysis with primers flanking exon 17 of lymphocyte RNA isolated from the proband identified a unique RT-PCR product that corresponded to the correctly spliced mRNA. No additional bands corresponding to the mutant mRNA, which should be approximately 200 base pairs smaller that wildtype mRNA, could de detected. C: Schematic representation of the minigene used to analyse the effect of the c.1390-14A>T polymorphism on splicing. The minigenes (A and T alleles) were transfected into 293 cells and the total RNA was analyzed by RT-PCR. Both alleles yielded a correctly spliced mRNA (arrow), without any sign of exon skipping or alternative splicing. The bands larger than the fully mature mRNA that can be observed in the figure represent pre-mRNA intermediates that were not completely processed, as confirmed by sequencing.
Analysis of unique intronic variants affecting degenerate splicing sequences As shown above, minigene constructs accurately identified the oncogenic character of bona fide splicing
mutation and distinguish them from non-pathogenic polymorphisms. Next, we analyzed by minigene assays RB1 variants affecting less conserved splice sites sequences. The variants studied represent unique RB1 alterations for which no previous clinical and functional data were available.
Figure 2 shows the results obtained with the intronic variant c.719-9C>G. This mutation was observed in a sporadic bilateral late onset retinoblastoma patient (Alonso, et al., 2001a) and have not been found among the more than 500 different mutations logged in the retinoblastoma gene mutation public databases (RBGMdb and RBGW). A wildtype and a mutant minigene containing exons 7, 8 and 9 and their flanking intronic sequences were constructed to study the effect of this variant on mRNA splicing. The expression of the wildtype minigene in 293 cells yielded one unique RT-PCR fragment that corresponded to a processed mRNA containing the three correctly spliced exons (data not shown). By contrast, expression of the mutant minigene yielded two different RT-PCR fragments (Fig. 2A). One of them was of a similar size to that observed for the wildtype minigene but corresponded to an alternatively spliced exon, including eight additional nucleotides from intron 7. The smaller RT-PCR fragment observed upon expression of the mutant minigene corresponded to a processed mRNA in which exon 8 was entirely skipped. Thus, expression of the mutant minigene produced two different mRNAs. RT-PCR analysis of the lymphocyte RNA isolated from the patient showed two RT-PCR products: one with a higher mobility corresponding to the skipping of exon 8 and another one that displayed a similar size to the RT-PCR product generated from the RNA obtained from the normal parents (Fig. 2B). Direct sequencing of this RT-PCR product showed a double sequence indicating that two different species co-migrated on the agarose gel. To confirm the identity of both species and to quantify the proportion of each of them, we cloned these RT-PCR products into pGEM-T and analyzed 70 independent clones by PCR-SSCP. Two migration patterns were clearly identified by PCR-SSCP, which corresponded to the normal spliced mRNA and to the alternatively spliced exon 8 containing
eight additional nucleotides from intron 7. Approximately 20% of the clones analyzed displayed this mutant pattern. Assuming that RB1 is expressed from both alleles with similar efficiency and that the correctly spliced mRNA is only attributable to the wildtype allele, thereby representing 50% of overall RB1 mRNA, then 37% of the transcripts will have the skipped exon 8 and 13% will contain the alternatively spliced exon 8 (Fig. 2B).
Exon 7 / ins 8 nt / Exon 8Exon 7 / ins 8 nt / Exon 8
mw
Alternative splicing
Exon 7 / Exon 9
Exon 8 skipping
Exon 7 / Exon 9Exon 7 / Exon 9
Exon 8 skipping
400 -300 -
Figure 2. RNA splicing alterations caused by the intronic variant c.719-9C>G. Abbreviations and symbols are as in figure 1. A: Schematic representation of the minigene used to analyse the effect of the c.719-9C>G mutation on mRNA splicing. Upon transfection in 293 cells, the mutant minigene yielded two bands that corresponded to an alternatively spliced mRNA with 8 additional nucleotides between exons 7 and 8 and a mutant mRNA in which the entire exon 8 was skipped. The wildtype minigene only yielded correctly spliced mRNA. B: RT-PCR analysis with primers flanking exon 8 of lymphocyte RNA isolated from the proband identified two bands on agarose gel electrophoresis. The smaller band corresponded to mRNA lacking exon 8 and the larger band, with a double sequence pattern, was cloned and individuals clones analyzed by PCR-SSCP. Two different migration patterns were observed, corresponding to the normal spliced mRNA and the alternatively spliced mRNA including eight additional nucleotides, respectively. The proportion of the different mRNA species in the leukocyte mRNA from the patient are also shown. C: A fragment of the intervening (lower case letters) and coding sequences (capital case letters) of the wildtype and mutant alleles are shown. The mutated nucleotide is indicated with a circle. The 3’ splice site scores of the normal splice site and the new cryptic splice site created by the mutation as identified by the ASD (Alternative Splicing Database) workbench are indicated. D: Schematic representation of mRNA splicing showing the skipping of exon 8 and the new alternatively spliced exon 8.1, indicating the new stop codons generated.
Genscan analysis of the genomic fragment predicted the existence of an alternative exon (Ex 8.1, P=0.839), in
which the 3’ splice site was shifted to the new AG dinucleotide created by the C>G transversion. However, Genscan not recognized the existence of the wildtype exon 8 in the mutated DNA, with a P value significantly lower (P=0.039) that the computed for this exon in the normal sequence (P=0.933). According to ASD (Alternative Splicing Database) workbench analysis, the 3’ splice site of the new alternatively spliced exon has a moderate score of 2.82 (perfect match with the consensus matrix is scored as 10), while 3’splice site scores of the wildtype exon 8 in the normal and mutant sequences were 7.67 and 6.16 respectively (Fig. 2C).
Pathogenic Validation of Intronic Variants in Retinoblastoma 7
The functional consequences of the c.719-9C>G variant are showed in Figure 2D and Table 2. This variant gave rise to either skipping of exon 8 or alternative splicing that result in a frameshift, producing the formation of new stop codons at positions 261 (S240fsX261) and 266 (I240fsX266) respectively, and resulting in an inactive protein lacking the known functional domains of pRB ().
The intronic variant c.2326-8T>A was observed in three affected members of a Cuban retinoblastoma family across three consecutive generations (Alonso, et al., 2005). The expression of the disease in this family was very heterogeneous and while the proband developed bilateral retinoblastoma at the age of one month, the daughter developed unilateral unifocal retinoblastoma at the age of eleven months, with no signs of additional retinoblastoma tumours three years later. No precedents for such a variant were found amongst the published RB1 mutations logged in the RBGMdb and the RBGW. In order to study the effect of this mutation on mRNA splicing, we constructed a minigene containing exon 22, the complete intron 22, and exons 23 and 24 linked by a fragment of intron 23. As shown in figure 3A, one RT-PCR product was observed in cells transfected with the mutant minigene, of a size undistinguishable from the RT-PCR product observed in cells transfected with the wildtype minigene on agarose electrophoresis. However, sequencing of RT-PCR products revealed that the mRNA produced from the mutated minigene corresponded to a processed mRNA containing six additional nucleotides at the exon 22-23 boundary. In silico approaches using Genscan failed to predict alternative splicing or exon skipping (P values for exon 23 in normal and mutant sequences were 0.944 and 0.936 respectively). However, the ASD tool predicted an alternative acceptor splice site six nucleotides upstream of the normal splice site, with a score of 6.1 (Fig. 3B). The usage of this cryptic site as demonstrated by the minigene assay result in a premature stop codon (X777) that will lead to the loss of the entire C terminal region containing the Cdk phosphorylation sites, essential for the inactivation of pRb in the G1 to S phase transition (Fig. 3C, Table 2). Unfortunately, no constitutive lymphocyte RNA was available from any of these patients to analyze splicing activity in vivo.
A
C
Ex 22 Ex 23 Ex 24
Wildtype
Ex 22 Ex 23.1 Ex 24
Mutant TAG
Exon 22 Exon 24Exon 23
c.2326-8T>A
wildtyp
e
mutant Exon 23 / ins 6 nt / Exon 22Exon 23 / Exon 22
wildtype Mutant: Alternative splicing
mw
400 -300 -
B
ttttttgtttt t gctctagCCCCCT
3’ splice site score
3’ splice site score
ttttttgtttt a gCTCTAGCCCCCT
7.25
5.146.10
Wildtype
Mutant
A
C
Ex 22 Ex 23 Ex 24
Wildtype
Ex 22 Ex 23.1 Ex 24
Mutant TAG
C
Ex 22 Ex 23 Ex 24
Wildtype
Ex 22 Ex 23.1 Ex 24
Mutant TAG
Exon 22 Exon 24Exon 23
c.2326-8T>A
wildtyp
e
mutant Exon 23 / ins 6 nt / Exon 22Exon 23 / Exon 22
wildtype Mutant: Alternative splicing
mw
400 -300 -
Exon 22 Exon 24Exon 23
c.2326-8T>A
Exon 22 Exon 24Exon 23
c.2326-8T>A
wildtyp
e
mutant Exon 23 / ins 6 nt / Exon 22Exon 23 / Exon 22
wildtype Mutant: Alternative splicing
mw
400 -300 -
wildtyp
e
mutant Exon 23 / ins 6 nt / Exon 22Exon 23 / ins 6 nt / Exon 22Exon 23 / Exon 22
wildtype Mutant: Alternative splicing
mw
400 -300 -
B
ttttttgtttt t gctctagCCCCCT
3’ splice site score
3’ splice site score
ttttttgtttt a gCTCTAGCCCCCT
7.25
5.146.10
Wildtype
Mutant
B
ttttttgtttt t gctctagCCCCCT
3’ splice site score
3’ splice site score
ttttttgtttt a gCTCTAGCCCCCT
7.25
5.146.10
Wildtype
Mutant
Figure 3. RNA splicing alterations caused by the intronic variant c.2326-8T>A. Abbreviations and symbols are as in figure 1. A: Schematic representation of the minigene used to analyse the effect of the c.2326-8T>A mutation on mRNA splicing. Upon transfection, the mutant minigene yielded a band with a similar size to that obtained from the wildtype minigene. Sequencing of this RT-PCR product demonstrated that it corresponded to an alternatively spliced mRNA including six additional nucleotides between exons 22 and 23. B: A fragment of the intervening and coding sequences of the wildtype and mutant alleles are shown, indicating the 5’ splice site scores and the new cryptic splice site identified by in silico analysis. C: Schematic representation of mRNA splicing showing the new alternative spliced exon 23.1 created in the mutant, and the premature stop codon generated.
The intronic variant c.501-15T>G was observed in the constitutional DNA of a unilaterally affected Serbian patient and his unaffected father (Kontic, et al., 2006). This rare variant has not been reported in other mutational studies of retinoblastoma patients. Since constitutive RNA from the patient was not available, we performed a minigene assay to confirm the effect of this mutation on splicing (Fig. 4A). The RT-PCR product obtained from cells transfected with the mutated minigene was smaller than that obtained from cells transfected with the wildtype (Fig. 4A). Indeed, sequencing of this RT-PCR product demonstrated that the mutated minigene produced only mature mRNAs in which exon 5 had been completely skipped. In silico analysis using the Genscan tool showed that the RB1 exon 5 has a P value of only 0.081 in the wildtype sequence. Even so, exon 5 was not identified at all when the mutant sequence was analyzed. This variant does not affect the consensus value of the splice acceptor signal localized in intron 4, but it destroys, according to ASD tool, a branch point located sixteen nucleotides upstream of exon 5 (Fig. 4B). Skipping of exon 5 does not affect the open reading frame of the protein but rather results in the loss inframe of eleven amino acids (Fig. 4C, Table 2).
A
C
Ex 4 Ex 5 Ex 6
Wildtype
Ex 4 Ex 6
Mutant
Exon 4 Exon 6Exon 5
c.501-15T>G
Exon 4 / Exon 5
wildtype Mutant: Exon 5 skipping
Exon 4 / Exon 6wild
type
mutant
mw
600 -
B
aaaaagtcataa t gtttttcttttcagGACAAG
Branch point score 3.66
aaaaagtcataa g gtttttcttttcagGACAAG
Branch point score not found
Wildtype
Mutant
A
C
Ex 4 Ex 5 Ex 6
Wildtype
Ex 4 Ex 6
Mutant
C
Ex 4 Ex 5 Ex 6
Wildtype
Ex 4 Ex 6
Mutant
Exon 4 Exon 6Exon 5
c.501-15T>G
Exon 4 / Exon 5
wildtype Mutant: Exon 5 skipping
Exon 4 / Exon 6wild
type
mutant
mw
600 -
Exon 4 Exon 6Exon 5
c.501-15T>G
Exon 4 Exon 6Exon 5
c.501-15T>G
Exon 4 / Exon 5
wildtype Mutant: Exon 5 skipping
Exon 4 / Exon 6wild
type
mutant
mw
600 -
Exon 4 / Exon 5
wildtype
Exon 4 / Exon 5
wildtype Mutant: Exon 5 skipping
Exon 4 / Exon 6
Mutant: Exon 5 skipping
Exon 4 / Exon 6wild
type
mutant
mw wildtyp
e
mutant
mw
600 -
B
aaaaagtcataa t gtttttcttttcagGACAAG
Branch point score 3.66
aaaaagtcataa g gtttttcttttcagGACAAG
Branch point score not found
Wildtype
Mutant
B
aaaaagtcataa t gtttttcttttcagGACAAG
Branch point score 3.66
aaaaagtcataa g gtttttcttttcagGACAAG
Branch point score not found
Wildtype
Mutant
Figure 4. RNA splicing alterations caused by the intronic variant c.501-15T>G. Abbreviations and symbols are as in figure 1. A: Schematic representation of the minigene used to analyse the effect of the c.501-15T>G mutation on mRNA splicing. Upon transfection, the mutant minigene yielded a unique band smaller than that obtained from the wildtype minigene. Sequencing of this RT-PCR product demonstrated that it corresponded to a mature mRNA in which exon 5 was skipped. B: A fragment of the intervening and coding sequences of the wildtype and mutant alleles are shown indicating the branch point site score. According to the in silico analysis, the branch point was destroyed in the mutant allele. C: Schematic representation of mRNA splicing showing the skipping of exon 5. Exon 5 is deleted inframe in the mutant mRNA and thus, no premature stop codons are generated.
Finally, we analyzed the intronic variant c.1961-12T>C found in a sporadic bilateral retinoblastoma patient and
his unaffected mother (Alonso, et al., 2005). In order to ascertain the possible pathogenic nature of this mutation, a minigene comprising RB1 exons 19, 20 and 21 was constructed and assayed in 293 cells. The analysis of the RNA from the cells expressing the normal or mutated minigene gave identical results (data not shown), indicating that this mutant did not alter normal pre-mRNA processing. We also performed RT-PCR analysis of constitutive RNA obtained from the patient and again no aberrant mRNA splicing could be documented (data not shown). In concordance with these results, in silico analysis did not identify changes on the splice site/branch point scores values (Table 2).
Pathogenic Validation of Intronic Variants in Retinoblastoma 9
Table 2. Summary of results obtained in unique RB1 intronic variants of uncertain pathogenic nature
RNA-based studies in silico predictions (2)Consequence
on the protein
Variant (1) ID proband
Minigene assay
RT-PCR on Lymphocyte
RNA Genscan ASD
c.719-9C>G (intron 7/exon 8)
RB0233 exon skipping & alternative
splicing
exon skipping & alternative
splicing
exon skipping
alternative splicing
new cryptic splice site
premature stop codon
c.2326-8T>A (intron 22/exon 23)
CU49 alternative splicing no available none new crytic
splice site premature stop codon
c.501-15T>G (intron 4/exon 5)
SR08 exon skipping no available exon skipping
branch point destruction
exon 5 deletion in
frame
c.1961-12T>C (intron 19/exon 20)
RB1053 normal splicing normal splicing none none none
(1) Reference sequences for variant nomenclature are L41870.1 (cDNA) and L11910.1 (genomic DNA). c.1 corresponds to the A of the ATG start codon in the reference sequence L41870.1. The intron/exon boundary where the variant is localized is also indicated. (2) Genscan, http://genes.mit.edu/GENSCAN.html; ASD (Alternative Splicing Database) workbench, http://ebi.ac.uk/asd-srv/wb.cgi.
DISCUSSION
Precise identification of the pathogenic character of any mutation in the RB1 gene is a prerequisite to provide accurate genetic counselling to patients at risk of developing retinoblastoma. Nonsense or frameshift RB1 mutations resulting in premature stop codons within exons 1 – 25 are considered oncogenic and are frequently associated with bilateral retinoblastoma of high penetrance and expressivity (Lohmann and Gallie, 2004). This group of bona fide oncogenic mutations also includes mutations affecting the invariant dinucleotides at splice sites, which produce aberrant splicing and result in frameshifts or inframe deletions of complete exons. However, the pathogenic effect of intronic variants affecting less conserved splice signals is unclear and they require further study to characterize their oncogenic nature. Although the frequency of these mutations is low in comparison with bona fide mutations (approximately 5% of all RB1 mutations), their impact on genetic counselling is great, since they are usually associated with low penetrance phenotypes and unaffected carriers with a high risk of transmission of the mutation to their offspring. In addition, the majority of these variants have only been described once and thus, no additional functional or clinical data can aid in the assessment of their pathogenicity. Accordingly, we have used minigene constructions to identify aberrant RNA splicing in several unique RB1 germline intronic variants of unclear oncogenic significance.
In order to establish the accuracy of the minigene assay, we first analyzed several bona fide splice mutations affecting invariable dinucleotides at the splice site and a frequent neutral polymorphism. In all these cases, minigene assays exactly reflected the expected character of these variants. Minigene assays were then used to analyze RB1 intronic variants affecting to less conserved splice sequences. Using this approach, we have been able to demonstrate the oncogenic nature of the intronic variants c.501-15T>G, c.719-9C>G and c.2326-8T>A. By contrast, the intronic variant c.1961-12T>C) was categorized by the minigene assay as a very infrequent neutral
polymorphism. Excepting only the Genscan analysis on the c.2326-8T>A mutation, the in silico results were in close accord with those obtained in vitro.
Minigene studies have been used to define the pathogenic character of intronic variants in several genetic diseases (Baralle, et al., 2003; Bergmann, et al., 2006; Di Leo, et al., 2004; Duponchel, et al., 2006; Raponi, et al., 2006; Rickard and Wilson, 2003; Tran, et al., 2006; Zeniou, et al., 2004). However, to our knowledge this is the first report describing the use of minigene constructs to study the oncogenic character of intronic RB1 mutations detected during mutational screening. Minigene studies appeared to be especially useful when no lymphocyte RNA was available to analyze the splicing aberrations, as for two out of the three pathogenic mutations studied here. In addition, the minigene approach was also helpful when mutant mRNA can not be detected in RNA samples as a consequence of nonsense mediated decay (Lejeune and Maquat, 2005), as observed for the c.1499-1G>C mutation affecting an invariable dinucleotide at the 3’ splice site.
Our results also emphasize the complex regulation of splicing, as seen by the consequences on RNA splicing of two very similar mutations: c.719-9C>G and c.2326-8T>A. Both mutations created new cryptic splice sites favouring the appearance of an alternatively spliced exon that included some nucleotides of the upstream intron. However, only the former mutation also induced the complete skipping of the affected exon, which was favoured in detriment to the alternatively spliced form in vivo. The reasons for this different behaviour remains unknown but will be probably related to the existence or formation of binding sites for regulatory splicing proteins, as it has been described for an intronic polymorphism in the KLF6 gene (Narla, et al., 2005).
In silico analysis of the c.501-15T>G variant showed the lost of the branch point consensus sequence, which was in concordance with the skipping of exon 5 observed in the minigene assay. These branch point mutations are rare in human disease and have not been previously described in the RB1 gene. Skipping of exon 5 does not affect the open reading frame of the protein and only results in the loss of eleven amino acids. This small peptide does not include any of the known phosphorylation sites in the amino-terminal region of human retinoblastoma protein (Zarkowska and Mittnacht, 1997) and hence, the functional significance of this region is not known. The resulting inframe skipping of exon 5 and the presence of the mutation in the unaffected father suggests a low-penetrance retinoblastoma phenotype. Similar phenotypes has been observed in other cases in which RB1 germline mutations also produce inframe exon skipping (Dryja, et al., 1993; Lefevre, et al., 2002; Lohmann and Gallie, 2004; Sanchez-Sanchez, et al., 2005; Sanchez, et al., 2000; Taylor, et al., 2007).
From a technical point of view, the minigene constructs presented here are time-consuming, in the majority of cases requiring PCR amplification and cloning from two or three genomic fragments to produce minigenes of a length suitable for subsequent cloning and analysis. Minigene construction could be especially tedious in the case of genes with multiple introns, some of which are extremely long, as in the RB1 gene. In order to make the method more straightforward we are designing an expression vector to facilitate the study of intronic variants, in the context of two pre-cloned flanking exons. This type of expression vector should facilitate the analysis of rare intronic variants in the RB1 and other genes by reducing the time and the number of steps to construct the minigenes.
In summary, using minigene constructs we have been able to demonstrate the occurrence of altered splicing and the resulting oncogenic nature of three unique intronic RB1 mutations (c.501-15T>G, c.719-9C>G, c.2326-8T>A). On the whole, our results show the utility of this approach to analyze pre-mRNA processing, especially in cases associated with variable retinoblastoma expressivity and in which clinical and genetic antecedents are not available.
ACKNOWLEDGMENTS
Parts of this work were presented at the 19th meeting of the European Association for Cancer Research (Budapest, 2006) where was conceded the Pezcoller-EACR Poster Awards. We thank Dr. Mark Shefton for support in proofreading the manuscript.
REFERENCES
Alonso J, Frayle H, Menendez I, Lopez A, Garcia-Miguel P, Abelairas J, Sarret E, Vendrell MT, Navajas A, Artigas M, Indiano JM, Carbone A, Torrenteras C, Palacios I, Pestaña A. 2005. Identification of 26 new constitutional RB1 gene mutations in Spanish, Colombian, and Cuban retinoblastoma patients. Hum Mutat 25:99.
Pathogenic Validation of Intronic Variants in Retinoblastoma 11
Alonso J, Garcia-Miguel P, Abelairas J, Mendiola M, Sarret E, Vendrell MT, Navajas A, Pestaña A. 2001a. Spectrum of germline RB1 gene mutations in Spanish retinoblastoma patients: Phenotypic and molecular epidemiological implications. Hum Mutat 17:412-422.
Alonso J, Moreno C, Lopez A, Mendiola M, Garcia-Miguel P, Abelairas J, Sarret E, Vendrell MT, Navajas A, Pestaña A. 2001b. Five novel single nucleotide polymorphisms of the RB1 gene (g.5625T>C, g.70169T>G, g.76875A>T, g.78026delA, and g.150072T>C) in retinoblastoma patients. Hum Mutat 17:437.
Baralle D, Baralle M. 2005. Splicing in action: assessing disease causing sequence changes. J Med Genet 42:737-748.
Baralle M, Baralle D, De Conti L, Mattocks C, Whittaker J, Knezevich A, Ffrench-Constant C, Baralle FE. 2003. Identification of a mutation that perturbs NF1 agene splicing using genomic DNA samples and a minigene assay. J Med Genet 40:220-222.
Bergmann C, Frank V, Kupper F, Schmidt C, Senderek J, Zerres K. 2006. Functional analysis of PKHD1 splicing in autosomal recessive polycystic kidney disease. J Hum Genet 51:788-793.
Burge C, Karlin S. 1997. Prediction of complete gene structures in human genomic DNA. J Mol Biol 268:78-94.
Cartegni L, Chew SL, Krainer AR. 2002. Listening to silence and understanding nonsense: exonic mutations that affect splicing. Nat Rev Genet 3:285-298.
Cavenee WK, Dryja TP, Phillips RA, Benedict WF, Godbout R, Gallie BL, Murphree AL, Strong LC, White RL. 1983. Expression of recessive alleles by chromosomal mechanisms in retinoblastoma. Nature 305:779-784.
Clark F, Thanaraj TA. 2002. Categorization and characterization of transcript-confirmed constitutively and alternatively spliced introns and exons from human. Hum Mol Genet 11:451-464.
Cooper TA. 2005. Use of minigene systems to dissect alternative splicing elements. Methods 37:331-340.
Di Leo E, Panico F, Tarugi P, Battisti C, Federico A, Calandra S. 2004. A point mutation in the lariat branch point of intron 6 of NPC1 as the cause of abnormal pre-mRNA splicing in Niemann-Pick type C disease. Hum Mutat 24:440.
Draper GJ, Sanders BM, Brownbill PA, Hawkins MM. 1992. Patterns of risk of hereditary retinoblastoma and applications to genetic counselling. Br J Cancer 66:211-219.
Dryja TP, Rapaport J, McGee TL, Nork TM, Schwartz TL. 1993. Molecular etiology of low-penetrance retinoblastoma in two pedigrees. Am J Hum Genet 52:1122-1128.
Duponchel C, Djenouhat K, Fremeaux-Bacchi V, Monnier N, Drouet C, Tosi M. 2006. Functional analysis of splicing mutations and of an exon 2 polymorphic variant of SERPING1/C1NH. Hum Mutat 27:295-296.
Kaelin WG, Jr. 1999. Functions of the retinoblastoma protein. Bioessays 21:950-958.
Knudson AG, Jr. 1971. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 68:820-823.
Kontic M, Palacios I, Gamez A, Camino I, Latkovic Z, Rasic D, Krstic V, Bunjevacki V, Alonso J, Pestaña A. 2006. New RB1 oncogenic mutations and intronic polymorphisms in Serbian retinoblastoma patients: genetic counseling implications. J Hum Genet 51:909-913.
Lefevre SH, Chauveinc L, Stoppa-Lyonnet D, Michon J, Lumbroso L, Berthet P, Frappaz D, Dutrillaux B, Chevillard S, Malfoy B. 2002. A T to C mutation in the polypyrimidine tract of the exon 9 splicing site of the RB1 gene responsible for low penetrance hereditary retinoblastoma. J Med Genet 39:E21.
Lejeune F, Maquat LE. 2005. Mechanistic links between nonsense-mediated mRNA decay and pre-mRNA splicing in mammalian cells. Curr Opin Cell Biol 17:309-315.
Lohmann DR. 1999. RB1 gene mutations in retinoblastoma. Hum Mutat 14:283-288.
Lohmann DR, Gallie BL. 2004. Retinoblastoma: revisiting the model prototype of inherited cancer. Am J Med Genet C Semin Med Genet 129:23-28.
Narla G, Difeo A, Reeves HL, Schaid DJ, Hirshfeld J, Hod E, Katz A, Isaacs WB, Hebbring S, Komiya A, McDonnell SK, Wiley KE, Jacobsen SJ, Isaacs SD, Walsh PC, Zheng SL, Chang BL, Friedrichsen DM, Stanford JL, Ostrander EA,
12 Gámez-Pozo et al.
Chinnaiyan AM, Rubin MA, Xu J, Thibodeau SN, Friedman SL, Martignetti JA. 2005. A germline DNA polymorphism enhances alternative splicing of the KLF6 tumor suppressor gene and is associated with increased prostate cancer risk. Cancer Res 65:1213-1222.
Raponi M, Upadhyaya M, Baralle D. 2006. Functional splicing assay shows a pathogenic intronic mutation in neurofibromatosis type 1 (NF1) due to intronic sequence exonization. Hum Mutat 27:294-295.
Rickard SJ, Wilson LC. 2003. Analysis of GNAS1 and overlapping transcripts identifies the parental origin of mutations in patients with sporadic Albright hereditary osteodystrophy and reveals a model system in which to observe the effects of splicing mutations on translated and untranslated messenger RNA. Am J Hum Genet 72:961-974.
Sanchez-Sanchez F, Kruetzfeldt M, Najera C, Mittnacht S. 2005. A novel constitutional mutation affecting splicing of retinoblastoma tumor suppressor gene intron 23 causes partial loss of pRB activity. Hum Mutat 25:223.
Sanchez F, Mateu E, Beneyto M, Najera C, Prieto F. 2000. A constitutional homozygous mutation in the RB1 gene in a patient with unilateral retinoblastoma. J Med Genet 37:615-620.
Stamm S, Riethoven JJ, Le Texier V, Gopalakrishnan C, Kumanduri V, Tang Y, Barbosa-Morais NL, Thanaraj TA. 2006. ASD: a bioinformatics resource on alternative splicing. Nucleic Acids Res 34:D46-55.
Stenson PD, Ball EV, Mort M, Phillips AD, Shiel JA, Thomas NS, Abeysinghe S, Krawczak M, Cooper DN. 2003. Human Gene Mutation Database (HGMD): 2003 update. Hum Mutat 21:577-581.
Taylor M, Dehainault C, Desjardins L, Doz F, Levy C, Sastre X, Couturier J, Stoppa-Lyonnet D, Houdayer C, Gauthier-Villars M. 2007. Genotype-phenotype correlations in hereditary familial retinoblastoma. Hum Mutat 28:284-293.
Tran VK, Takeshima Y, Zhang Z, Yagi M, Nishiyama A, Habara Y, Matsuo M. 2006. Splicing analysis disclosed a determinant single nucleotide for exon skipping caused by a novel intraexonic four-nucleotide deletion in the dystrophin gene. J Med Genet 43:924-930.
Valverde JR, Alonso J, Palacios I, Pestaña A. 2005. RB1 gene mutation up-date, a meta-analysis based on 932 reported mutations available in a searchable database. BMC Genet 6:53.
Zarkowska T, Mittnacht S. 1997. Differential phosphorylation of the retinoblastoma protein by G1/S cyclin-dependent kinases. J Biol Chem 272:12738-12746.
Zeniou M, Gattoni R, Hanauer A, Stevenin J. 2004. Delineation of the mechanisms of aberrant splicing caused by two unusual intronic mutations in the RSK2 gene involved in Coffin-Lowry syndrome. Nucleic Acids Res 32:1214-1223.
