| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Molecular Diagnostics and Genetics |
1 Genomic Unit for the Diagnosis of Human Pathologies, San Raffaele Scientific Institute, Milan, Italy.
2 Molecular Genetics Unit, Clinical Pathology Department, Azienda Ospedaliera OIRM-S.Anna, Turin, Italy.
3 Department of Pediatrics, University of Turin, Turin, Italy.
4 Neurosurgery Unit, Ospedale Meyer, Firenze, Italy.
5 Diagnostica e Ricerca San Raffaele SpA, Milan, Italy.
6 Università Vita-Salute San Raffaele, Milan, Italy.
aAddress correspondence to this author at: Genomic Unit for the Diagnosis of Human Pathologies, San Raffaele Scientific Institute, Via Olgettina 60, 20132 Milano, Italia. Fax 39-02-26434351; e-mail cremonesi.laura{at}hsr.it.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: To identify 10 common mutations in the genes for fibroblast growth factor receptors 2 and 3 (FGFR2 and FGFR3), we developed a microelectronic microchip assay that exploited the PCR multiplexing format and coupled it with serial addressing and probe hybridization on the same pad. For the molecular characterization of patients who tested negative in the microchip screening, we also developed conditions for denaturing HPLC (DHPLC) analysis of the most mutated regions of FGFR2 and FGFR3 and the entire coding region of the TWIST1 gene.
Results: In our cohort of 159 patients with various craniosynostosis syndromes, mutations were found in 100% of patients with Apert syndrome, 83.3% with Pfeiffer syndrome, 72.7% with Crouzon syndrome, 50.0% with Saethre-Chotzen syndrome, 27.7% with plagiocephaly, 31.8% with brachicephaly, 20% of complex cases, and 6.9% of mixed cases. No mutations were found in syndromic cases.
Conclusions: The combined microchip-DHPLC strategy allows rapid and specific molecular diagnosis of craniosynostosis and is an effective tool for the medical and surgical management of these common congenital anomalies in a newborn or an infant with a developmental defect of the cranial vault.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Genetic studies have helped in elucidating the molecular bases of this complex and heterogeneous group of developmental disorders, thanks to the identification in a number of syndromes (e.g., Apert, Crouzon, Pfeiffer, and Saethre-Chotzen craniosynostosis) of mutations in fibroblast growth factor receptor (FGFR) 1 genes and the TWIST1, MSX2, EFNB1, and ALPL genes (1)(2). Most of the mutations have been found in FGFR2 and FGFR3. Such studies have also led to the classification of novel syndromes diagnosed at the molecular level, such as Muenke syndrome, which is related to the FGFR3 P250R mutation (3)(4). Moreover, evidence suggests that a molecular diagnosis could predict outcome and help in clinical management and long-term follow-up. Recent data showed that the frequency of repeated transcranial surgery to reduce intracranial pressure was significantly increased in a cohort of patients heterozygous for the FGFR3 P250R mutant, indicating that this molecular lesion may represent a negative prognostic factor. Conversely, no significant link was found between perceived disease severity and reoperation rate (3).
Molecular diagnosis of craniosynostosis, via either direct sequencing (5) or mutational scanning with denaturing HPLC (DHPLC), had been described only for the FGFR2 gene (6); however, a DHPLC procedure was recently developed for detecting mutations in FGFR3, but it has been applied only to patients affected by achondroplasia and thanatophoric dysplasia (7).
To standardize and simplify the procedures for diagnosing craniosynostoses, we first set up DHPLC conditions for scanning a selected panel of gene regions where the majority of mutations have been found. Findings obtained in this effort led to the identification of a number of molecular defects in a cohort of patients in our department who were affected by various craniosynostoses. Some of these mutations were more frequent than others. These results prompted us to develop a microelectronic microchip, an advanced method of directly identifying mutations for fast and automated diagnosis of molecular defects in our patient population.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
for details). In 23 cases, we also analyzed samples from the patient’s relatives for a total of 40 individuals. Written informed consent was obtained from all patients and/or their relatives in accordance with the standards of the local ethics committee.
|
dna extraction
Genomic DNA was isolated and purified from 150 µL of peripheral blood on an ABI PrismTM 6100 Nucleic Acid Prep Station (Applied Biosystems) according to the manufacturer’s instructions.
dhplc analysis
We established DHPLC conditions for the molecular scanning of DNA regions where mutations had been previously found in FGFR2, FGFR3, and TWIST1. Most FGFR2 and FGFR3 mutations occur in sequences encoding the 3rd extracellular immunoglobulin-like domain (FGFR2 exons IIIa and IIIc; FGFR3 exon 7) (4)(8). We used the WAVE system (Transgenomic) for DHPLC.
The amplification primers for FGFR2 exons IIIa and IIIc, FGFR3 exon 7, and the entire TWIST1 coding region have already been described (9)(10). The PCR was performed by amplifying 50–100 ng of genomic DNA in a 50-µL volume. To screen for homozygous changes, we added an equal quantity of previously sequenced wild-type amplicon to the PCR of the patient sample before heteroduplex formation. The PCR conditions for exons 1, 2, and 5 were as follows: 95 °C for 10 min followed by 16 cycles of 94 °C for 20 s, 60 °C for 20 s, and 72 °C for 30 s; 16 cycles of 94 °C for 20 s, 52 °C for 20 s, and 72 °C for 30 s; and 72 °C for 10 min. For exons 3 and 4, the PCR conditions were as follows: 95 °C for 10 min followed by 16 cycles of 94 °C for 20 s, 60 °C for 20 s, and 72 °C for 30 s; 20 cycles of 94 °C for 20 s, 52 °C for 20 s, and 72 °C for 30 s; and 72 °C for 10 min. To obtain heteroduplexes, we denatured the samples for 5 min at 95 °C and cooled them to 65 °C for 30 min. The DHPLC temperatures used were 61 °C for FGFR2 exon IIIa, 59.1 °C for FGFR2 exon IIIc, 66.2 °C for FGFR3 exon 7, 64.5 °C for TWIST1 3', and 67.5 °C for TWIST1 5'.
sequence analysis
Samples that displayed altered DHPLC elution profiles were sequenced directly. We sequenced PCR products from both ends with the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems) and then purified the PCR products with Centri-Sep columns (Princeton Separations). The PCR products were run on an ABI Prism 3100 DNA sequencer and analyzed with Factura and Sequence Navigator software (Applied Biosystems).
pcr conditions for microchip analysis
To optimize assay conditions, we used 20 DNA control samples carrying the following 10 mutations: S252W (755C>G), P253R (758C>G), F276V (826T>G), C278F (833G>T), C342R (1024T>C), C342Y (1025G>A), C342S (1025G>C), A344P (1030G>C), and S347C (1040C>G) mutations in the FGFR2 gene and the P250R (749C>G) mutation in the FGFR3 gene.
FGFR2 exons IIIa and IIIc were coamplified in a multiplex reaction, whereas FGFR3 exon 7 and the 3' TWIST1 region were each amplified with optimized primer sets. One of the primers in each set was biotinylated at the 5' end (see Table 1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol53/issue10).
For both the single and duplex formats, 100 ng of genomic DNA was amplified in a 50-µL volume containing 200 µmol/L of each deoxynucleoside triphosphate, 1.5 mmol/L MgCl2, 15 pmol of each primer, 1.5 U AmpliTaq Gold (Applied Biosystems), and 1x PCR Buffer II (Applied Biosystems). PCR reactions were carried out in a PCR Express thermal cycler (Hybaid) with an initial denaturation step at 95 °C for 10 min followed by 35 cycles of denaturation at 95 °C for 30 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s. The PCR was terminated with a final extension at 72 °C for 5 min. Each PCR product was purified and desalted with the MultiScreen Separation System (Millipore). We evaluated the amplified DNA by gel electrophoresis with a 100-bp molecular weight marker ladder and checked with a Nanogen conductivity meter to ensure that conductivity values were <100 µsiemens. Finally, the biotinylated amplicon was transferred to a 96-well plate in a 50-mmol/L histidine buffer for the loading step.
probe design
The design of the microchip probe requires sets of fluorescently labeled reagents, including wild-type and mutant reporters and a nonlabeled specific stabilizer oligonucleotide. Stabilizer oligonucleotides are necessary to promote opening of the secondary structures of DNA templates and for the base-stacking format.
To analyze the 10 mutations in the FGFR genes, we designed 8 sets of primers (see Table 2 in the online Data Supplement) with the assistance of programs available for free on the Internet [DNA folding serverhttp://frontend.bioinfo.rpi.edu/zukerm/cgi-bin/rna-index.cgi; OligoAnalyzer 3.0 (Integrated DNA Technologies),http://www.idtdna.com]. We adjusted base length within each pair of wild-type and mutant probes to obtain more similar melting behaviors.
Seven of the sequence variants were located either at the 5' terminus (P250R, S252W, F276V, C342R, S347C) or at the 3' terminus (A344P, P253R) of the probe, because a base-stacking interaction with an adjacent stabilizer oligonucleotide permitted a more stable probe-to-template interaction in this format. For the remaining 3 variants (C342S, C342Y, C278F), we designed a dot-blot format with the base variation located within the probes. The fluorescent cyanine dyes Cy3 and Cy5 were used to label wild-type and mutant reporters, respectively, at the 5' end. Alternatively, we labeled the reporters at the 3' end in cases in which secondary structures forming at the 5' end of the target strand could interfere with the hybridization process. Because the labeling efficiencies of Cy3 and Cy5 at the 3' end were low, we labeled wild-type and mutant reporters with 6-carboxytetramethylrhodamine and BODIPY 650/665, respectively (see Table 2 in the online Data Supplement). Oligonucleotide purification was necessary for optimal results.
microchip analysis
For microchip analysis, we used the NMW 1000 NanoChipTM Molecular Biology Workstation (Nanogen) (11)(12)(13)(14)(15)(16), which uses microelectronics to enable the active movement and concentration of charged molecules to designated test sites on a cartridge formatted with 100 microelectrodes. Samples were electronically placed on the chip via a loader and electrophoresed to the selected pads by means of positive-bias direct current. After sodium hydroxide denaturation, the cartridge was hybridized with stabilizers and reporters specific for each mutation. The temperature was then specifically increased to obtain optimal differentiation of matched and mismatched probes. We detected hybridization by automated fluorescence scanning, and the system automatically analyzed the data with dedicated software that provides patient genotyping. Detailed information on the Nanogen technology and protocols has already been reported (16)(17).
Serial addressing of the PCR products and serial probe hybridization were performed on the same pads. Before each addressing and before subsequent hybridization steps, we chemically stripped the pad with 0.1 mol/L NaOH for 3 min to detach all previously hybridized probes.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
summarizes the results of DHPLC mutational analysis (see Fig. 1
for examples of DHPLC elution profiles) and direct sequencing for samples that displayed an altered elution profile. Forty-seven (29.5%) of the 159 patients with a diagnosis of some form of craniosynostosis carried mutations, and 20 different mutations were identified (Table 1
). For the 23 patients who had mutations and for whom relatives were also evaluated, the patient’s mutation was also present in the mother in 8 cases and the father in 2 cases. The relatives of the other patients had wild-type genes, indicating a de novo origin of the mutations in these patients.
|
microchip analysis
We initially optimized the analytical protocols for detecting each mutation under stringent thermal conditions to ensure that we had established specific hybridization conditions. This preliminary version of the assay was blindly validated with 80 identified DNA samples from wild-type or mutant control individuals, and the results were completely concordant. The final version of the craniosynostosis assay exploited the PCR multiplexing format coupled with serial addressing of the PCR products and serial probe hybridization on the same pads to detect mutations in the FGFR genes (17).
FGFR3 exon 7 (where the P250R mutation is located) could not be included in the multiplexing PCR format because of a high sequence similarity with FGFR2 exon IIIa (location of the P253R mutation). Interestingly, the P250R and P253R mutants are both produced by a C>G change within a stretch of 20 nucleotides that differ in the 2 sequences by only 3 bp. Hence, when the 2 amplicons were simultaneously addressed on the same pad and probed with the P250R mutant reporter, the reporter cross-hybridized with the P253R mutant template but not vice versa. This result might be explained by different stabilities for the 2 mismatched hybrids. Fig. 2
shows that the mismatched hybrid P253R probe stabilizer/P250R template contains 3 mismatches, 2 within the probe and 1 at the critical stabilizer position involved in the base-stacking interaction with the probe. These particular mismatches produced no cross-hybridization. Conversely, the opposite situation (mismatched hybrid P250R probe stabilizer/P253R template) contains 3 mismatches (2 within the stabilizer and 1 within the probe), none of which involve the critical base-stacking interaction (Fig. 2
). This situation leads to cross-hybridization. This problem was overcome by 1st addressing and probing the FGFR3 exon 7 amplicon for the P250R mutation; the duplex PCR for FGFR2 exons IIIa and IIIc was subsequently addressed on the same pad and analyzed for mutations S252W, P253R, F276V, C278F, A344P, S347C, C342Y, C342S, and C342R through serial probe hybridization with all the different sets of wild-type and mutant probes.
|
Under these analytical conditions, we were able to analyze all mutations and still obtain reliable identification until the last serial hybridization step, with matched-mismatched fluorescence ratios of >5 (range, 10:1 to >10 000:1) (Fig. 3
; see Table 3 in the online Data Supplement). Unambiguous differentiation of wild-type homozygous samples and heterozygous mutant samples was obtained in all cases; no cross-hybridization was observed, even if most mutations involved the same or adjacent nucleotides.
|
Validation of the final multiplexed/multiprobed craniosynostosis biochip assay was performed blindly with new aliquots of the same panel of 80 control samples. Results obtained from the analysis of all mutations showed 100% concordance with those obtained by other methods.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
FGFR2 mutations are associated with the common Apert, Crouzon, and Pfeiffer craniosynostosis syndromes (5)(6). As expected, we found that all of the patients with suspected Apert syndrome in our cohort carried the 2 FGFR2 gene mutations specific for this disease (i.e., P252W and P253R), thereby confirming the clinical diagnoses. Molecular lesions in FGFR2 have been found in 72.7% of Crouzon syndrome patients and in 83.3% of Pfeiffer syndrome patients.
Notably, the craniosynostosis associated with the P250R FGFR3 mutant has previously been referred to as Muenke syndrome (4)(18)(19). Nevertheless, because of the high heterogeneity of the phenotypes, the diseases of the patients in our cohort with this molecular defect had been classified on the basis of their clinical presentations. Consequently, the P250R FGFR3 mutant and other TWIST1 gene defects were associated with the Saethre-Chotzen, brachicephaly, and plagiocephaly syndromes (5)(18)(19).
The FGFR3 P250R mutation and TWIST1 mutations have been identified in 50% of Saethre-Chotzen cases and in 27.7% of isolated plagiocephaly cases and 31.8% of brachicephaly cases. TWIST1 and FGFR2 mutations have also been found in a minority of complex cases (20%) and mixed cases (6.9%). No mutations have been found in patients with a complex syndromic phenotype. Notably, genetic analysis has been useful for refining the final clinical diagnosis in the 1st months of life, when the clinical phenotype often is not yet fully manifested.
Mutations have thus far been identified with DHPLC and/or direct sequencing of DNA samples from affected individuals. Reliable identification of mutations in the heterozygous state requires sequencing of both the forward and reverse strands of the DNA template.
As an alternative innovative and efficient testing strategy for diagnosing craniosynostosis, we propose the integration of a microchip approach for rapidly and directly identifying the most frequent known mutations [according to our findings and the published literature (6)] with the use of DHPLC to screen for rarer and new mutations in patients for whom previous analyses have revealed no mutations.
The microelectronic microchip approach has already been shown to be particularly suitable for the design of assays tailored to any local situation. This flexibility is facilitated by the capability of integrating several multiplexing formats, including multiplexing the PCR reaction, multiple addressing of several amplicons to the same pad, and serially hybridizing the same pad with several probe sets. An interesting problem we faced in the present work was analyzing mutations in highly homologous DNA regions, such as in FGFR3 exon 7 and FGFR2 exon IIIa, where the P250R and P253R variants are located, respectively. This tricky situation both precluded coaddressing these 2 amplicons to the same pad because of the cross-hybridization of the P253R probe stabilizer with the P250R template and prevented multiple amplifications.
To overcome this problem, it was sufficient first to address and probe the critical amplicon (FGFR3 exon 7) with the wild-type and mutant reporter set for the P250R mutation and subsequently to address the duplex PCR of FGFR2 exons IIIa and IIIc to allow serial probe hybridization with all probe sets on the same pad. This approach further underscores the high flexibility of the microelectronic system, and it may be used in any analysis of highly homologous regions.
In the present work, we used a traditional format, with the fluorophores linked directly to specific probes. An alternative design that uses a universal format based on reporters containing a sequence-specific discriminator with a tail that hybridizes to either a wild-type or mutant universal Cy3- or Cy5-labeled probe would further decrease the cost of the technique (20)(21)(22).
The optimized final version of the craniosynostosis microchip allows the identification of FGFR2 and FGFR3 mutations on a single pad per patient. This pad can produce reliable genotyping results until the end of the analysis. This format allows automated genotyping of 100 samples for all 10 mutations in 3–4 h at a cost of approximately $8 US per sample, including PCR reagents.
In contrast to the microchip approach, DHPLC is an indirect procedure that identifies sequence alterations, but the method still requires sequencing to precisely characterize the nature and position of the molecular defect. The DHPLC approach, however, does allow reliable molecular screening for other defects in patients who have tested negative for the mutants in the microchip screening.
The microchip approach identified 37 (78.7%) of the 47 FGFR2 and FGFR3 mutations in our retrospective study of a cohort of patients affected by a variety of craniosynostosis syndromes and avoided the time and expense of the DHPLC scanning and sequencing procedures that would otherwise have been required for these 37 patients. We thus propose an integrated strategy that uses the microchip approach for a preliminary fast and automated screening for the predominant known mutations, followed by DHPLC scanning in patients who test negative in the microchip analysis (Fig. 4
).
|
An alternative approach would be to directly sequence all samples for these 3 FGFR2 and FGFR3 exons in both the forward and reverse directions to screen for the presence of a heterozygous mutation. Nevertheless, we still consider our integrated strategy the cheapest and the fastest approach for diagnosing the craniosynostosis syndromes, even if it is probably not the simplest method.
A rapid and specific molecular diagnosis of craniosynostosis in a newborn or an infant with a developmental defect of the cranial vault is an effective tool for the medical and surgical management of these common congenital anomalies. Moreover, the identification of mutations in genes of interest would also have obvious implications in genetic counseling regarding any future offspring. With the combined microchip-DHPLC strategy, the correct molecular diagnosis is achieved in few days and with lower cost.
|
Acknowledgments |
|---|
Financial disclosures: None declared.
|
Footnotes |
|---|
3 Current address: CPG-Microfluidic Division Lab-On-Chip R&D Team, STMicroelectronics, Catania, Italy.
1 Human genes: FGFR2, fibroblast growth factor receptor 2 (bacteria-expressed kinase, keratinocyte growth factor receptor, craniofacial dysostosis 1, Crouzon syndrome, Pfeiffer syndrome, Jackson-Weiss syndrome); FGFR3, fibroblast growth factor receptor 3 (achondroplasia, thanatophoric dwarfism); TWIST1, twist homolog 1 (acrocephalosyndactyly 3; Saethre-Chotzen syndrome) (Drosophila); MSX2, msh homeobox 2; EFNB1, ephrin-B1; ALPL, alkaline phosphatase, liver/bone/kidney.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and gray (
) indicate hybridization with the mutant (mut) and wild-type probe, respectively. Case numbers are indicated (e.g., CS48).