| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Human Genetics Division, School of Medicine, Southampton University Hospital, Southampton, UK;2 Department of Physiology, Federal University of Para, Belem-Para, Brazil;3 School of Biological Sciences, University of Southampton, Southampton, UK;
aaddress correspondence to this author at: Human Genetics Division, Duthie Building (Mp808), School of Medicine, Southampton University Hospital, Tremona Road, Southampton SO16 6YD, UK; fax 44-(0)23-80794264, e-mail inmd{at}soton.ac.uk
Various physical mutation-scanning methods have been developed to avoid unnecessary resequencing of long stretches of DNA (1)(2)(3)(4)(5)(6). Protein-based mutation-scanning techniques include enzymatic digestion [reviewed in Ref. (7)], protein binding to a DNA duplex, and direct analyses of the in vivo or in vitro gene product. One such enzyme is T4 endonuclease VII (endoVII), the product of gene 49 of bacteriophage T4 (8). Radiolabel replacement with fluorescent tags has facilitated automated analysis (9). EndoVII recognizes heteroduplex structural distortions, nicking 2–6 bp 3' to the distortion, with efficiency dependent on sequence context (10) and mismatch type(11). Perfectly matched DNA undergoes some background digestion, which produces a highly reproducible pattern (12). Mutation detection sensitivity obtained with endoVII digestion was found to be similar to that for denaturing HPLC and direct sequencing (13).
Microplate array diagonal gel electrophoresis (MADGE) (14) provides an open-faced 96-well gel format for polyacrylamide gels. Recently, nondenaturing 192-, 384-, and 768-well formats of MADGE for high-throughput checking of PCR and post-PCR reactions (15) have been developed. We have combined, in proof-of-principle experiments, the mismatch digestion properties of endoVII with the high-throughput capabilities of MADGE and a newly developed denaturing MADGE format to create a simple mutation-scanning technique that can screen 
1000 PCR samples during a single 35-min electrophoretic run.
Plasmid pRB210 (T4 endonuclease VII in pET11a) was a kind gift from Professor B. Kemper (Institute for Genetics, University of Cologne, Germany). The PCR primers used to amplify the endoVII gene from pRB210 were as follows: forward, 5'-GCGCCATATGATGTTATTGAC-3'; reverse, 5'-CAGCGGATCCTCATTTTAAACT-3'. After trimming was performed with BamHI and NdeI (New England Biolabs), pETendoVII was generated by ligation into pET15b (Novagen). Expressed N-terminal His-tagged endoVII was then purified by affinity chromatography.
We used a single colony from pETendoVII-transfected BL21 (DE3) Gold cells (Stratagene) to inoculate a 1-L Luria broth culture containing 100 µg/L carbenicillin. After overnight culture at 30 °C, an identical fresh 500-mL passage was made, and at mid-log phase of growth (absorbance at 600 nm, 0.6–0.8), protein expression was induced by 1 mmol/L isopropyl-ß-D-thiogalactopyranoside. Cells were harvested after 2 h by centrifugation at 5000g for 10 min, and then lysed by sonication (10 cycles of 30 s on and 30 s off at a probe amplitude of 10–15 µm in a MSE Soniprep 150). Cell debris and intact cells were removed by centrifugation at 10 000g for 40 min. All steps were carried out at 4 °C. The cell lysate was passed through a Schleicher & Schuell 0.2 µm single-use filter.
EndoVII was purified by use of 1-mL HiTrap columns in conjunction with the 
KTATM FPLCTM chromatography system (Amersham Bioscience), according to the manufacturer’s instructions. Protein purity was assessed by sodium dodecyl sulfate gel electrophoresis (Fig. 1 in the Data Supplement that accompanies the online version of this Technical Brief at http://www.clinchem.org/content/vol51/issue6/), enzyme activity was confirmed (without His-tag removal) with digests of synthetic heteroduplex substrates (data not shown), and protein quantification was by Bradford assay. Storage was in 50 mmol/L Tris-HCl (pH 8) with 1 mmol/L dithiothreitol and 500 mL/L glycerol at –80 °C.
All primers were from MWG-Biotech. Exon 3 from wild-type LDLR (GenBank accession no. Nm_000527) was PCR-amplified using primers LDLR-F (5'-GCCTCAGTGGGTCTTTCCTT-3') and LDLR-R (5'-CCAGGACTCAGATAGGCTCAA-3'), respectively, with 6-carboxyfluorescein (FAM) and hexachloro-6-carboxyfluorescein (HEX) 5' end labels for probe generation or without end labels for generating amplicon from genomic DNAs for testing. Jumpstart Taq polymerase (Sigma-Aldrich) was used to ensure the highest quality probe generation, but the thermal and ionic conditions for probe and test sample amplifications were otherwise identical and essentially as given by Whittall et al. (16). Probe PCR parallel reactions from microplate wells were pooled and purified with Wizard PCR prep reagents (Promega). The same 220-bp PCR amplicon of LDLR exon 3 was generated (with unlabeled primers) from samples from 330 unrelated familial hypercholesterolemic individuals previously mutation scanned by single-strand conformation polymorphism (SSCP) analysis (16) and by meltMADGE (17). Six previously defined heterozygotes, c.259T>G (p.W66G), c.266G>A (p.C68Y), c.269A>G (p.D69G), c.301G>A (p.E80K), c.301delG (p.E80fs), and c.313 + 1G>A (splice site), were examined. Initially samples known to contain 1 of the 6 mutations were used to test endoVII digestion and were analyzed by capillary electrophoresis on an ABI-310 instrument. Subsequently, 330 amplicons were screened blind with denaturing MADGE (below) as the analytical platform. All protocols were developed by M.J. Smith and were validated by independent use by G. Pante-de-Sousa and X. Chen.
To form the heteroduplexes, we mixed 2.5 µL of purified fluorescently labeled probe (representing an equivalent volume of original PCR) and 5.5 µL of unpurified test PCR amplicon, heated the mixture to 95 °C, and allowed it to cool to reform duplex DNA. For endoVII digestion, we used a 10-µL reaction volume containing 8 µL of probe/test mixture and 2 µL of 5x endoVII reaction mixture [250 mmol/L K2HPO4 (pH 6.5), 25 mmol/L MgCl2, 5 mmol/L dithiothreitol, and 0.1 g/L endoVII]. Phosphate ions have been shown to improve the efficiency of endoVII (18). Digestions were for 20 min at 37 °C.
EndoVII reaction mixture (2.5 µL) was mixed with 12 µL of deionized formamide, denatured at 95 °C for 5 min, and then chilled on ice before capillary electrophoresis (Applied Biosystems 310 Genetic Analyzer).
For endoVII-MADGE, the reaction was terminated by addition of 3 µL of loading dye (10 mmol/L NaOH, 50 mmol/L EDTA, 800 mL/L formamide, 2.5 g/L bromphenol blue, and 2.5 g/L xylene cyanole FF). Samples were denatured by heating at 95 °C for 5 min and placed on ice until gel loading.
EndoVII digestion fragments were resolved on a 10% polyacrylamide denaturing MADGE gel containing 7 mol/L urea and 1x Tris-borate-EDTA buffer [90 mmol/L Tris-HCl (pH 8.3), 90 mmol/L boric acid, 2 mmol/L EDTA]. After sample loading, the gel was covered by a second glass plate. This plate–gel–plate sandwich was secured by rubber bands, and silicon rubber tubing was inserted along the long edge of the sandwich to prevent electrophoretic edge artifacts. The assembly was placed in a purpose-built 2-L gel tank (19) (with capacity for 10 gels) containing 1x Tris-borate-EDTA buffer at 65 °C for electrophoresis at 10 V/cm for 35min. EndoVII-MADGE gels were scanned and analyzed with either a FluorImagerTM 595 or a Typhoon Trio+ (Molecular Dynamics, Amersham Biosciences) and ImageQuant fragment analysis software (Molecular Dynamics).
LDLR mutations c.259T>G, c.301delG, c.301G>A, and c.313 + 1G>A generated a strong digest fragment for at least 1 probe strand, whereas c.266G>A generated a lower yield of digest fragment on 1 of the probe strands. c.269A>G displayed cleavage of the A·C heteroduplex when the label was on the C strand (mutant as probe). A typical example of the digestion pattern of the LDLR mutants can be seen in Fig. 2 of the online Data Supplement. The extra peaks observed corresponded to expected digest fragment sizes. These same products were trialed under various conditions in denaturing MADGE gels followed by fluoroimaging: the protocol described above was efficient.
A typical 96-well endoVII-MADGE gel from blind scanning of 330 familial hypercholesterolemic individuals is shown in Fig. 1
(also shown, with dual label, in Fig. 3 of the online Data Supplement). Previous mutation scanning of this sample set had identified 47 heterozygous individuals with 1 of the 6 mutations: c.259T>G, c.266G>A, c.269A>G, c.301G>A, c.301delG, or c.313 + 1G>A (Table 1
). When we used only wild-type probe, endoVII-MADGE identified 51 samples containing additional digest fragments; 46 of these corresponded to the previously identified mutations covering 5 of the 6 known LDLR mutations (c.259T>G, c.266G>A, c.301G>A, c.301delG, and c.313 + 1G>A). The c.269A>G mutation remained undetected (see above). Of the 5 additional samples, 3 displayed digestion patterns matching those for positively identified known LDLR mutations: 1 with the pattern for c.259T>G and 2 with the pattern for c.301G>A. The remaining 2 samples displayed unique digest patterns that did not correspond to digest patterns for the 5 known mutations (Fig. 4A in the online Data Supplement). One digest pattern was similar to that for c.313 + 1G>A, but c.313 + 1G>A was characterized by a strong digestion fragment, whereas the unidentified mutation produced a significantly weaker fragment (Fig. 4B in the online Data Supplement). Sequencing showed that the sample was heterozygous for the base change c.311G>T. The second sample produced a digestion fragment close to the undigested amplicon. Sequencing showed a 2-base deletion, c.196_197delGT. c.311G>T has been reported previously (www.ucl.ac.uk/fh/genebook.html), whereas c.196_197delGT appears to be a novel mutation. Of the 7 mutations detected, 2 displayed detectable mismatch-specific digestion patterns in both the sense and antisense strands, c.196_197delGT and c.259T>G, whereas the remainder were identified by digestion of 1 strand.
|
|
This study suggests the feasibility of combining the mismatch digestion properties of endoVII with the high-throughput capabilities of MADGE to create a simple high-throughput mutation-scanning method. We found that the reduced resolution and increased relative background associated with short-track electrophoresis did not decrease the rate of mutation detection. EndoVII-MADGE also identified 2 previously unrecognized mutations in the sample set. EndoVII-MADGE consistently compared favorably with SSCP analysis of the same region (Table 1
) in many heterozygotes, detecting 7 of 8 different sequence variations (8 of 8 when test samples were end labeled). This approach could potentially add to strategies for the investigation of unknown mutations at the population level.
Acknowledgments
M.J. Smith was the recipient of a University of Southampton Faculty of Health Medicine and Life Science cross-school PhD studentship. We thank Professor Borries Kemper for clone pRB210. This work was also supported by the UK Department of Health, National Genetics Reference Laboratory (Wessex), and HOPE.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
