A novel procedure for simple and efficient genotyping of single nucleotide polymorphisms by using the Zn²⁺–cyclen complex

Abstract

The analysis of single nucleotide polymorphisms (SNPs) is increasingly utilized in the study of various genetic determinants. Here, we introduce a simple, rapid, low-cost and accurate procedure for the detection of SNPs by polyacrylamide gel electrophoresis (PAGE) with a novel additive, the Zn²⁺– cyclen complex (cyclen = 1,4,7,10-tetraazacyclododecane). The method is based on the difference in mobility of mutant DNA (in the same length) in PAGE, which is due to Zn²⁺–cyclen binding to thymine bases accompanying a total charge decrease and a local conformation change of target DNA. Various nucleotide substitutions (e.g. AT to GC) in DNA fragments (up to 150 bp) can be visualized with ethidium bromide staining. Furthermore, heteroduplex and homoduplex DNAs are clearly separated as different bands in the gel. We demonstrate the analysis of single- and multiple-nucleotide substitutions in a voltage-dependent sodium channel gene by using this novel procedure (Zn²⁺–cyclen–PAGE).

INTRODUCTION

Single nucleotide polymorphisms (SNPs) in the human genome are important markers for establishing genetic linkage and genetic diseases (1–4). In order to make large-scale genotyping feasible, various SNP detection methods have been developed (5–6). However, simpler, faster, and more economical procedures are required for the analysis of genetic diseases. The most commonly used procedure for determining single-base mutations is as follows: (i) amplification of a target DNA sequence by PCR; (ii) scanning a mutation of the PCR products by using a single-strand conformation polymorphism (7), enzymatic or chemical cleavage of mismatched base pairs (8–13), conformation-sensitive gel electrophoresis (14) and denaturing gradient gel electrophoresis (15); (iii) confirmation of the sequence of mutant DNA by various DNA sequence analyses. However, the procedures so far developed require complicated processes, special apparatus, expensive reagents and/or a skillful analyst.

In 1993, Shionoya et al. (16) reported that Zn²⁺–cyclen (cyclen = 1,4,7,10-tetraazacyclododecane) selectively and reversibly binds to an imide-containing nucleobase, deoxythymidine (dT), in aqueous solution with a dissociation constant K_d = [free dT][free Zn²⁺–cyclen] / [dT^––Zn²⁺– cyclen] = 0.8 mM at pH 7.4 (Fig. 1). In the resulting 1:1 complex, the nucleobase is an imide-deprotonated species (dT^–) that binds with the Zn²⁺ ion, where the total charge of the dT molecule increases from 0 to +1. Recently, we reported that Zn²⁺–cyclen derivatives selectively bind dT-rich regions and change the local conformation in double-stranded DNA (e.g. a bulbous structure, as shown in Fig. 2), as proven by nuclease footprinting experiments and gel mobility shift assays (17–19). The dissociation of A–T hydrogen bonds is promoted by Zn²⁺–cyclen, as observed by lowering the melting temperature (T_m) with an increase in the concentration of Zn²⁺–cyclen (17,20,21). We have extended such T-recognizing property of Zn²⁺–cyclen in the polyacrylamide gel electrophoretic separation of various DNA fragments. We selected 18 mutants of a skeletal muscle voltage-dependent sodium channel α-subunit (Na_v1.4) gene as DNA samples (see DNA fragments 1–19 in Table 1). These mutants were constructed in a study on the relationship between the structure and function of the channel (22–25). Several of the mutations are related to the loss of binding activity of a sodium channel activator, Grayanotoxin (a diterpenoid extracted from the family of Ericaceae). We here describe a simple, rapid, low-cost and accurate polyacrylamide gel electrophoresis (PAGE) method for analyzing PCR products by using Zn²⁺–cyclen as a novel additive (Zn²⁺–cyclen–PAGE). As the first practical example, we demonstrate the analysis of SNPs in the sodium channel mutants.

Figure 1.

Equilibrium for Zn²⁺–cyclen binding to deoxythymidine (dT).

Figure 2.

A proposed mechanism of Zn²⁺–cyclen binding to double-helical DNA. (A) Native DNA and (B) Zn²⁺–cyclen-bound DNA.

Table 1. Sequences of target region in the double-stranded DNA, 1–19.

Samplea	Sequence (5′–3′)b
1 (wild-type)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcctggcc··
2 (V422K·V423K·I424K)	··ttcaagaaaaaaatcttcctgggctccttctacctcatcaatctgatcctggcc··
3 (V422K)	··ttcaaggtcatcatcttcctgggctccttctacctcatcaatctgatcctggcc··
4 (V423K)	··ttcgtgaaaatcatcttcctgggctccttctacctcatcaatctgatcctggcc··
5 (I424K)	··ttcgtggtcaaaatcttcctgggctccttctacctcatcaatctgatcctggcc··
6 (V422M)	··ttcatggtcatcatcttcctgggctccttctacctcatcaatctgatcctggcc··
7 (V423L)	··ttcgtgttaatcatcttcctgggctccttctacctcatcaatctgatcctggcc··
8 (I424V)	··ttcgtggtcgtcatcttcctgggctccttctacctcatcaatctgatcctggcc··
9 (I433K)c	··ttcgtggtcatcatcttcctgggctccttctacctcaaaaatctgatcctggcc··
10 (I433A)c	··ttcgtggtcatcatcttcctgggctccttctacctcgccaatctgatcctggcc··
11 (I433V)c	··ttcgtggtcatcatcttcctgggctccttctacctcgtcaatctgatcctggcc··
12 (N434K)c	··ttcgtggtcatcatcttcctgggctccttctacctcatcaaactgatcctggcc··
13 (N434A)c	··ttcgtggtcatcatcttcctgggctccttctacctcatcgctctgatcctggcc··
14 (L437K)c	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcaaggcc··
15 (L437A)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcgcggcc··
16 (L437L-1)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatccttgcc··
17 (L437M)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcatggcc··
18 (L437L-2)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcctcgcc··
19 (L437V)	··ttcgtggtcatcatcttcctgggctccttctacctcatcaatctgatcgtggcc··

Bold characters show mutated bases.

^aMutant names in parentheses are referenced by the original amino acid followed by its number and introduced amino acid.

^bThis region corresponds to nucleotides 1261–1314 of the Na_v1.4 gene described by Trimmer et al. (27).

^cThe mutations are related to the lowering of binding activity of a sodium channel activator, Grayanotoxin, described by Ishii et al. (22) and Kimura et al. (23,24).

MATERIALS AND METHODS

Single-stranded oligonucleotides

All oligonucleotides (20–28) and PCR primers were obtained commercially (Espec Oligo Service, Japan). The PCR primers for 300 bp (pUC 19 nucleotides 1731–2030) were 5′-ATT AAG CAT TGG TAA CTG TC-3′ and 5′-AGT TAC CTT CGG AAA AAG AG-3′, and those for 336 bp (pUC 19 nucleotides 481–816) were 5′-GCG GTA TTT CAC ACC GCA TA-3′ and 5′-AAT GTA TTT AGA AAA ATA AA-3′ (26).

Target double-stranded DNA molecules

The mutations onto the cDNA coding the rat skeletal muscle voltage-dependent sodium channel α-subunit (Na_v1.4) (27) were introduced by site-directed mutagenesis using a PCR strategy as described before (22–25,28). The wild-type and mutated DNA fragments were cloned into a pGEM-T Easy Vector (Promega, USA), and the inserted regions were then confirmed with restriction mapping and sequencing entirely using an ABI PRISM™ 310 Genetic Analyzer (Applied Biosystems, USA). The sequences of target regions for wild-type (1) and mutated DNA (2–19) are listed in Table 1.

The target regions were amplified by PCR using the forward/reverse primers 5′-CAA TTG TGG GAG CCC TGA TCC-3′/5′-CGT ACG CCA TGG CCA CCA CG-3′, which produced 621 bp fragments from nucleotides 713–1333 of the Na_v1.4 gene. Each 50 µl of PCR solution contained 50 ng of pGEM-T Easy Vector DNA in which a wild-type or a mutated fragment was inserted, forward and reverse primers (0.6 µM), dNTPs (each at 65 µM) and 2.5 U Thermoprime plus (Advanced Biotechnologies, UK). After initial denaturation at 95°C for 3 min, amplification was carried out for 30 cycles of 1 min denaturation at 95°C, 1 min annealing at 60°C and 1 min extension at 72°C. PCR products were purified by precipitation in ethanol. PCR products from Na_v1.4 mutants were digested with restriction endonucleases to generate short fragments, which have appropriate length for the electrophoretic analysis.

Zn²⁺–cyclen–PAGE

Electrophoresis was performed at 200 V for 150 min at room temperature, in a 1-mm-thick, 9-cm-wide and 9-cm-long gel prepared with an appropriate concentration of polyacrylamide (30:1 ratio of acrylamide to N,N′-methylenebisacrylamide), 5 mM Zn²⁺–cyclen, 90 mM Tris and 90 mM borate, on a standard PAGE apparatus (ATTO, Japan; model AE-6500). DNA samples were dissolved in 3 µl of a loading buffer containing 60% (v/v) glycerol, 18 mM Tris/18 mM borate, 0.05% (w/v) bromophenol blue and 0.05% (w/v) xylene cyanol. The anode buffer was 90 mM Tris/90 mM borate, and the cathode buffer was 90 mM Tris/90 mM borate/5 mM Zn²⁺–cyclen. At the end of the run, the gels were stained in an aqueous solution of ethidium bromide (10 µg/ml). Zn²⁺–cyclen was prepared as dinitrate salt (colorless prisms) by a similar method to that reported for Zn²⁺–cyclen diperchlorate salt (16). Anal. Calc. for C₈H₂₀N₆O₆Zn (Zn²⁺–cyclen·2NO₃^–): C, 26.57; H, 5.57; N, 23.24. Found: C, 26.61; H, 5.60; N, 23.28. ¹H NMR (500 MHz, in D₂O): δ 2.80 (8H, m), 2.94 (8H, m). ¹³C NMR (125 MHz, in D₂O): δ 46.6. IR: 3177, 2918, 1483, 1444, 1384 (NO₃^–), 1279, 1092, 1010, 993, 806 cm^–1. The dissociation constant for the HPO₄^2–-bound Zn²⁺–cyclen complex was determined with 1 mM Zn²⁺–cyclen and 1 mM Na₂HPO₄ by potentiometric pH titration (29) to be 0.10 ± 0.02 mM (= [HPO₃^2–][Zn²⁺–cyclen] / [HPO₃^2––Zn²⁺–cyclen]) at 35°C with I = 0.10 (NaNO₃) in aqueous solution.

RESULTS AND DISCUSSION

Effect of Zn²⁺–cyclen on the electrophoretic migration of single-stranded DNA

In order to evaluate the binding effect of Zn²⁺–cyclen to single-stranded DNA fragments, we conducted PAGE in the presence of 5 mM Zn²⁺–cyclen (Zn²⁺–cyclen–PAGE) with 20mer (20–24) and 50mer (25–28) oligonucleotides, as shown in Table 2. Those DNA fragments are different in T-content, which should influence the binding number of Zn²⁺–cyclen [5 mM Zn²⁺–cyclen is enough for its thymine-binding (16)]. In the absence of Zn²⁺–cyclen, same-length oligonucleotides were observed to have slightly different mobilities (see Fig. 3A for 20mer and Fig. 4A for 50mer fragments). The small differences are possibly due to variation in the single-stranded conformation. On the other hand, in the presence of 5 mM Zn²⁺–cyclen, the electrophoretic migration was retarded as the number of thymine bases increased (see Fig. 3B for 20mer and Fig. 4B for 50mer fragments). The migration distances of 20mer fragments 20–24 are inversely proportional to the T-contents [i.e. 22 (2T) > 24 (3T) > 21 (4T) > 20 (5T) > 23 (7T)]. Among the 50mer samples, the most T-rich fragment, 26 (20T), showed the largest retardation compared to the other oligonucleotides, 25, 27 and 28 (10T). The retardation of T-rich oligonucleotides by Zn²⁺–cyclen could be explained by the decrease in the negative charge of oligonucleotides (i.e. phosphodiester polyanions) and the increase in the molecular size (see a proposed structure of Zn²⁺–cyclen binding DNA in Fig. 2).

Table 2. Sequences of single-stranded oligonucleotides, 20mer 20–24 and 50mer 25–28.

Sample	Sequence (5′–3′)	T	A	C	G
20	aattgctgcagtaatacgac	5	7	4	4
21	agttaccttcggaaaaagag	4	8	3	5
22	gcgtcagaccccgtagaaaa	2	7	6	5
23	attaagcattggtaactgtc	7	6	3	4
24	aagtgcggcgacgatagtca	3	6	4	7
25	agtcagtcagtcagtcagtcaaaaaaaaaaagtcagtcagtcagtcagtc	10	20	10	10
26	tcagtcagtcagtcagtcagtttttttttttcagtcagtcagtcagtcag	20	10	10	10
27	tcagtcagtcagtcagtcagcccccccccctcagtcagtcagtcagtcag	10	10	20	10
28	agtcagtcagtcagtcagtcggggggggggagtcagtcagtcagtcagtc	10	10	10	20

Thymine bases are underlined.

Figure 3.

Electrophoresis of 20mer oligonucleotides 20–24 with 20% (w/v) polyacrylamide gel in the absence (A) and presence (B) of 5 mM Zn²⁺–cyclen.

Figure 4.

Electrophoresis of 50mer oligonucleotides 25–28 with 12% (w/v) polyacrylamide gel in the absence (A) and presence (B) of 5 mM Zn²⁺–cyclen.

Effect of Zn²⁺–cyclen on electrophoretic migration of double-stranded DNA

Zn²⁺–cyclen–PAGE experiments with various concentrations of Zn²⁺–cyclen and two double-stranded DNA molecules, 300 bp DNA (AT content of 61.7%) and 336 bp DNA (AT content of 48.5%) in the plasmid pUC19 (26), were conducted. In the absence of Zn²⁺–cyclen, both DNA fragments moved in proportion to the molecular weight (see the DNA bands in Fig. 5, left), where the 300 bp DNA showed faster migration. The difference of the migration distances of those DNA fragments decreased with the increase in the concentration of Zn²⁺–cyclen, and the migration distances of those DNAs were then reversed above 2 mM Zn²⁺–cyclen (Fig. 5). We used 5 mM Zn²⁺–cyclen for a general Zn²⁺–cyclen–PAGE experiment, as shown below. These facts imply that Zn²⁺–cyclen prefers the AT-rich double-stranded DNA and that the electrophoretic migration is retarded as the thymine content increases. This result clearly explains our previous finding that Zn²⁺–cyclen changes the local conformation in double-stranded DNA (Fig. 2B) (17–19).

Figure 5.

Relationship between [Zn²⁺–cyclen] and migration distance of 300 (AT-content of 61.7%) and 336 bp (AT-content of 48.5%) DNA fragments in 4% (w/v) polyacrylamide gel. Electrophoresis results in the absence and presence of 5 mM Zn²⁺–cyclen gel are shown on the left and right, respectively.

Analysis of single- and multiple-base substitutions by Zn²⁺–cyclen–PAGE

To examine the sensitivity of Zn²⁺–cyclen–PAGE for the detection of DNA mutations, 19 PCR products of a wild-type DNA, 1, and various base-substituted mutant DNAs, 2–19, of the voltage-dependent sodium channel α-subunit (Na_v1.4) gene (Table 1) were prepared. The series of mutants consist of substitutions of a single base (6, 8, 11, 12 and 16–19), two bases (3, 5, 7, 9, 10 and 13–15), three bases (4) and seven bases (2). The PCR-amplified 621 bp products were digested with restriction endonuclease AluI, which produces a fragment of a mutated region (99 bp) and seven other fragments (177, 158, 63, 61, 35, 15 and 13 bp). In the absence of Zn²⁺–cyclen, all 99 bp fragments for 1–19 showed almost the same migration distance (Fig. 6A). In contrast, by using Zn²⁺– cyclen–PAGE, the 99 bp fragment bands varied widely (Fig. 6B). The other fragments stayed at individual positions in both electrophoresis experiments. With the increase in the number of T bases (2–7, 9, 14, 16 and 17), the migration distance decreased in the Zn²⁺–cyclen gel (Fig. 6B). Especially, mutations forming poly-AT alignments resulted in a significant decrease of the migration distance (see 2, 4 and 7). On the contrary, the migration distances of 8, 10, 11, 13 and 15 (A, T→G, C) increased (Fig. 6B). The mobility shift was, however, not proportional to the T-content, as shown for 5 and 6 (+1 AT pair) or 10 and 13 (–2 AT pairs). Each mutant DNA has a characteristic conformation at the mutated region in the double-stranded DNA, which would reflect the binding affinity to Zn²⁺–cyclen. In fact, the mutants with unchanged T-content, 18 (G to C) and 19 (C to G), showed a small decrease in the migration distances compared to the wild-type 1. Further determination of various mutant DNAs by using the Zn²⁺–cyclen–PAGE would give a rational explanation for the relationship between the DNA sequence and local conformation.

Figure 6.

Electrophoresis of the AluI digested PCR products from wild-type 1 and its mutants 2–19 of the voltage-dependent sodium channel α-subunit (Na_v1.4) gene in 12% (w/v) polyacrylamide gel in the absence (A) and presence (B) of 5 mM Zn²⁺–cyclen. Fragment sizes are shown on the right.

In order to determine the appropriate length of a DNA fragment for Zn²⁺–cyclen–PAGE, the PCR-amplified 621 bp products were also digested with restricted endonucleases MspI and XmnI to prepare longer fragments of 151 and 120 bp, respectively. Although differences in mobility for the samples were observed in both cases, they were smaller than those with AluI. Thus, a DNA fragment length of up to 150 bp should be suitable for this analysis.

Application of Zn²⁺–cyclen–PAGE to heterozygosity analysis

Finally, we applied Zn²⁺–cyclen–PAGE to a more accurate SNP detection procedure by a combination of the well known heterozygosity screening technique. PCR using a 1:1 mixture of a wild-type and a mutant gene as a template gave the complementary DNA (two homoduplexes) and a single or more mismatched DNA (two heteroduplexes). The PCR products were digested with AluI and then subjected to Zn²⁺–cyclen–PAGE. Figure 7A shows the case of single-nucleotide substitutions 6, 8, 11, 12 and 16–19. One or two upper bands corresponding to the two heteroduplexes are clearly observed. Even the substitutions A to T (i.e. 12) and G to C (i.e. 18 and 19) showed clear band(s) at different position(s) from those for homoduplexes. In the case of multiple-base substitutions 2–5, 7, 9, 10 and 13–15, heteroduplexes are more clearly detectable as upper shifted bands (Fig. 7B–D). The migration distances of all heteroduplexes are shorter than those of corresponding homoduplexes. These results indicate that a DNA mismatch promotes Zn²⁺–cyclen binding to the thymine base around the mismatch site, resulting in a relatively large conformation change (e.g. a large bubble in the double-helical DNA), which enables visualization of all mutations as different DNA bands.

Figure 7.

Electrophoresis of heteroduplexes and homoduplexes of the AluI-digested PCR products (99 bp) prepared with wild-type 1 and its mutants 2–19 of the voltage-dependent sodium channel α-subunit (Na_v1.4) gene in 12% (w/v) polyacrylamide gel in the presence of 5 mM Zn²⁺–cyclen: (A) single-base-, (B) two-base-, (C) four-base- and (D) seven-base-substituted mutants.

CONCLUSIONS

We introduced the novel procedure ‘Zn²⁺–cyclen–PAGE’ for a simple, rapid, low-cost and accurate analysis of DNA mutation. The Zn²⁺–cyclen–PAGE is based on the principle that the binding of Zn²⁺–cyclen to the thymine base changes the local DNA conformation, resulting in different electrophoretic mobility of a mutant DNA. Combination of a PCR technique for heterozygosity screening and Zn²⁺– cyclen–PAGE enables more accurate detection of single nucleotide mutations even for the less detectable substitutions AT to TA and GC to CG. Since the Zn²⁺–cyclen–PAGE procedure requires a general electrophoretic system and only one additive Zn²⁺–cyclen, it would be a very useful tool for various SNP analyses in an ordinary laboratory. Furthermore, the DNA-binding Zn²⁺–cyclen can be easily dissociated by adding a pH 7 phosphate buffer (20 mM) (K_d value of 0.10 mM for HPO₃^2––Zn²⁺–cyclen) or decreasing the gel pH to ∼4, these two methods being simple post-treatments for subsequent sequence analysis, such as mass spectroscopy. It is worthwhile to consider using Zn²⁺–cyclen–PAGE in the medical field for the screening and genotyping of various disease-causing mutations.