Materials and methods

POLYMERASE CHAIN REACTION

The same PCR protocol was initially used in order to amplify all PCR products. Red Hot Taq DNA polymerase and buffers were supplied by Advanced Biotechnologies, dNTPs were purchased from Pharmacia, and primers were made by Genosys-Sigma. PCR was performed using 50 ng of DNA in a volume of 25 μl and the final concentration of reagents was 1 × buffer (20 mmol/l (NH₄)₂SO₄, 75 mmol/l Tris-HCl, pH 8.8, at 25°C, 0.01% (v/v) Tween 20, 1.5 mmol/l MgCl₂), dNTPs (0.2 mmol/l each), 0.2 μmol/l of each primer, and 1.25 units Red Hot Taq polymerase. Amplification was carried out with the use of a Touchdown thermal cycler (Hybaid). Conditions for each PCR product were optimised by varying the annealing temperature and the Mg²⁺ concentration in the PCR reaction. A typical PCR programme (a) comprised the following steps: 94°C for three minutes (one cycle), 94°C for one minute, variable annealing temperature for one minute, 72°C for one minute (35 cycles), 72°C for 10 minutes (one cycle). For several amplifications, a touchdown PCR programme with variable annealing temperatures was used. In that case, the annealing temperature ranged between 70-60°C (b), 70-58°C (c), 65-53°C (d), or 60-50°C (e) depending on the melting properties of the primers. A typical touchdown PCR programme consisted of the following steps: 94°C for three minutes. 94°C for one minute, annealing started at 70°C and decreased by 1° per cycle, 72°C for one minute (11 cycles), 94°C for one minute, 60°C for one minute, 72°C for one minute (30 cycles), 72°C for 10 minutes (one cycle). Finally, in some cases, amplification conditions given by Splawski et al 19 were chosen in order to amplify certain KCNQ1 andKCNE1 fragments. These were (f) 94°C for three minutes (one cycle), 94°C for 10 seconds, 58°C for 20 seconds, 72°C for 20 seconds (30 cycles), 72°C for five minutes (one cycle) and (g) 94°C for three minutes (one cycle), 94°C for 10 seconds, 64°C for 20 seconds, 72°C for 20 seconds (five cycles), 94°C for 10 seconds, 62°C for 20 seconds, 72°C for 20 seconds (30 cycles), and 72°C for five minutes (one cycle).

Products that did not amplify satisfactorily were subjected to a different PCR method using HotStarTaq DNA polymerase and following the manufacturer's protocol (Qiagen). In some cases, a PCR additive, Q solution (also supplied by Qiagen), was included in PCR reactions to facilitate the amplification of difficult templates by modifying the melting behaviour of DNA. Amplification conditions, primer sequences, and sizes of amplified products are given in tables 1-4.

KCNQ1

All primers were taken from Splawski et al 19 except 9R (5′ GAC ACA GGC TGT ACC AAG CCA 3′), 13F (5′ CAC TGC CTG CAC TTT GAG CC 3′), 13R (5′ GTG AGG AGA AGG GGG TGG TT 3′) and 15R (5′ GCA GGA GCT TCA CGT TCA CA 3′). PCR conditions are given in table 1.

HERG

Primers were taken from Splawski et al 19 and Itoh et al.23 The details and PCR conditions are shown in table 2.

SCN5A

Primers first published by Wang et al 22 were initially used to amplify all 28 exons of the SCN5A gene. All PCR products were then purified and sequenced. New primers were designed for PCR products with lengths unsuitable for SSCP analysis (approximately >350 bp). Primers and PCR conditions for the amplification of theSCN5A gene are given in table 3.

KCNE1 ANDKCNE2

Primers for amplification of the coding region (single exon) of the KCNE1 andKCNE2 genes were taken from Splawskiet al 19 and Abbottet al,16 respectively (table4).

SINGLE STRAND CONFORMATIONAL POLYMORPHISM ANALYSIS

SSCP analysis was performed with the use of a Protean Cell system (Biorad) at two different temperatures (room temperature and 4°C). Polyacrylamide gels (20 cm) were made using 2 × stock SequaGel MD solution (National Diagnostics) following the manufacturer's protocol. PCR products (2 μl) were resolved overnight on SequaGel MD gels at different concentrations (0.4-0.8×) depending on their size. Wherever possible more than one PCR fragment were multiplexed in the same lane. After each run gels were silver stained using standard methods.

SEQUENCING

PCR products were sequenced using BigDye Terminator chemistry on a 377 ABI PRISM DNA sequencer (PE Biosystems) according to standard protocols.

Results

Using the PCR conditions detailed in tables 1-4, PCR consistently yielded clean products for all LQT fragments. ForKCNQ1, 17 PCR fragments were produced as exon 1 was divided into two separate amplicons. ForHERG, 19 PCR fragments were amplified as exons 4, 6, 7, and 12 were split into smaller amplicons. Four new primers were designed for KCNQ1, and existing sequences were used for HERG. However, in the former case, 10 primer sets needed changes to suggested PCR protocols, while for HERG this was necessary for all amplicons, and primers from two sources19 23 had to be used to obtain optimal results. In the case ofSCN5A, the whole coding region was amplified with the use of 34 PCR fragments. Exon 1 was excluded as it corresponds to the 5′ UTR part of the gene and similarly part of exon 28 was not amplified as it is the 3′ UTR. Initially, all PCR fragments amplified using the existing SCN5Aprimers22 were sequenced in order to determine their base composition. This was considered necessary as (1) many fragments were of unsuitable size for SSCP analysis and (2) the intronic sequence surrounding all SCN5Aexons was completely unknown with the exception of the splice junctions which were determined previously.22 Twelve new primers yielding smaller PCR amplicons were designed for exons 2, 7, 8, 14, 16, 23, 25, and 28 while exons 12, 16, 17, and 28 were subdivided into several PCR products because of their large size. ForKCNE1 and KCNE2, existing primer sets were used for the amplification of the whole coding region of these genes which is contained in a single exon.16 19 However, we reduced the number of PCR fragments needed from three to two. ForKCNE1, the two amplified fragments correspond to fragments 1 and 2 (264 and 231 bp respectively) reported by Splawski et al. 19 These completely cover the coding region of this gene and therefore there was no need for a third fragment. With regard toKCNE2, we combined the forward primer from fragment 1 and the reverse primer from fragment 2 quoted by Abbottet al 16 to create a new PCR amplicon of 293 bp. A second PCR amplicon was produced with the primers which Abbott et al 16 used for fragment 3 in their study (213 bp).

SSCP analysis using the methods described earlier produced clear bands for all PCR amplicons. In many cases, depending on their size, PCR products of two or more exons were loaded onto the same lane and resolved simultaneously on an SSCP gel, for example,KCNQ1 exons 2, 11, and 15 orSCN5A exons 11, 12, and 15.

A number of samples from sudden death cases and controls (healthy unrelated subjects) were used to evaluate the efficiency of our methods. SSCP analysis of over 60 samples forKCNQ1, KCNE1, andKCNE2 identified six polymorphisms in the first two genes. These were 1086 C>A V308V, 1638 G>A S546S, and 1986 C>T Y642Y in KCNQ1 and 84 G>A S29S, 112 G>A S38G, and 253 G>A D85N in KCNE1. No base changes were identified inKCNE2.

Discussion

In this study we produced a robust and reproducible set of primers for PCR/SSCP analysis of all five genes involved in the long QT syndrome. We also defined the exact PCR conditions for a successful amplification of every PCR fragment, which allows effective mutational analysis of the five known LQT genes. In particular, forKCNQ1 and HERG we based our analysis on primers published in two previous studies, Splawski et al 19 and Itohet al,23 respectively. However, PCR conditions for a number of fragments had to be optimised, as in our experience some PCR products amplified poorly under the original conditions. In addition, new primers were needed forKCNQ1 exons 9, 13, and 15. Based on the details given in table 1, all KCNQ1 exons can now be amplified without the use of formamide. Using our PCR methods, mutation screening of HERG can be achieved by amplifying only 19 fragments compared to 2019and 24.23 This was made possible by recombining existing primers in a more effective way and then optimising PCR conditions for each fragment. For example, HERG exon 6 can be amplified in two fragments instead of three as used by Itohet al 23 and there is no need for nested primers for exons 1 and 11 of the same gene.19

Genomic organisation and sequence of theSCN5A gene were first reported by Wanget al. 22 In that study a total of 41 primer pairs were used to determine the sequence of 28 exons by direct sequencing. Several studies have since used those primers for mutational analysis of some SCN5A exons by SSCP.25-27 However, a number of PCR fragments are too large for effective SSCP analysis and contain long intronic sequences of unknown composition. Using the original set of primers, we sequenced all SCN5A exons and then produced several new PCR fragments suitable for SSCP analysis. For amplifyingKCNE1 and KCNE2fragments, no changes were made to the published PCR methods as they were found to work well. However, we recombined the existing primers in a way that permits amplification of the entire coding region of these genes in two fragments instead of the three used in previous studies.16 19

PCR products were analysed on SSCP gels by combining more than one fragment in each lane. In that way a large number of fragments can be screened on the same gel in various combinations depending on their electrophoretic mobility, thus reducing the number of gels needed for mutational screening. Evaluation of our system using DNA samples from sudden death cases and controls detected four known polymorphisms and two novel changes in KCNQ1 andKCNE1. In the first gene we found two SNPs at amino acid positions 546 and 64228 and a novel one at position 308. On the other hand, in KCNE1 we report a novel SNP (84 G>A S29S) and two known polymorphisms at positions 38 and 85.29 30 Owing to the nature of these two novel changes, we speculate that they may be benign polymorphisms even though they were not observed in 40 controls analysed. These findings provide good evidence that the methods described in this study, considering their limitations, can be effective for screening the known LQT genes for mutations.