Clive R. Pullinger, Arghavan K. Shahidi, Andrea L. Verhagen andJohn P. Kane*‡

Cardiovascular Research Institute, *Department of Medicine,‡Department of Biochemistry and Biophysics, University ofCalifornia, San Francisco, California

IntroductionIn a number of fields, techniques are required for searchingcandidate genes for mutations that underlie hereditary diseases.We have used the denaturing gradient gel electrophoresismethod (DGGE)1 using the Bio-Rad DCode system to identifya number of new, as well as known, mutations and polymor-phisms in the genes for apolipoprotein A-I (apoA-I) andapolipoprotein B (apoB).

Materials And MethodsGenomic DNA was prepared from patients with dyslipidemia.Segments of the apoA-I and apoB genes were amplified fromgenomic DNA using a temperature cycler (Ericomp Inc., SanDiego, CA). Both oligonucleotide primers had GC clamps.2

Following restriction endonuclease digestion, the samples weresubjected to DGGE.

The amplification reactions were performed in 50 mM Tris-HCl,pH 9 (at 25 °C), 20 mM (NH4)2SO4, 1.5 mM MgCl2, 200 µM ofeach dNTP, 1 unit of Hot Tub™ polymerase (Amersham LifeScience Inc., Cleveland, OH), 15 pmol of each primer and a minimum of 200 ng of DNA in a total volume of 50 µl. After

initial denaturation at 96 °C for 1 minute, PCR* was carried out fora total of 33 cycles at an annealing temperature of 62 °C (apoA-I)or 58 °C (apoB), in each case for 30 s. The denaturing and elongation steps were 96 °C for 30 s and 72 °C for 60 s. A finalelongation step of 8 minutes at 72 °C was then carried out. Thetemperature cycler was programmed to denature the amplifica-tion products at 96 °C for 8 minutes and to then perform slowcooling to 40 °C. Without this last step, heteroduplex DNA maynot be visible on the DGGE gels.

In the case of apoA-I, a 420 bp region at the 5' end of the gene,containing the core promoter and the first exon, was amplifiedusing two GC clamped primers such that the final PCR productwas 532 bp in length. Similarly, 465 bp of the LDL receptor binding domain of apoB was amplified with a final size of 591 bp. In each case, 10 µl of the PCR product was digested in avolume of 30 µl with either AvaII (apoA-I) or EcoRI (apoB).Figure 1 shows the melting profile graphs for the 5' and 3' fragments generated after digestion of the apoB PCR product.3

These graphs were produced using Bio-Rad MacMelt™ softwareand were used to predict the percentage range of the denaturinggradient to be used in the gels. Multiple aliquots of the digestswere loaded onto 7.5% acrylamide (38.5:1) DGGE gels at hourlyintervals. These travel gels were run in the DCode system appa-ratus at a constant temperature of 56 °C and 200 V for a total of5 hours to determine the optimal running time. This was foundto be 3 h for the apoA-I fragments and 4h for apoB. The gels werestained for 30 minutes in SYBR® Green and photographed underepi-illumination at 254 nm.4 A 35–65% denaturing gradient was used for apoA-I and a 30-50% gradient for apoB.

US/EG Bulletin 2171

Detection of Apolipoprotein Gene Variantsby Denaturing Gradient Gel Electrophoresis

Using the DCode™ System

Fig. 1. Melting profiles of a portion of the receptor binding domain of the apolipoprotein B gene amplified using GC clamped primers.The graphs were generated using MacMelt software. The positions of five single-base substitution mutations are shown.

0 50 100 150 200 250 0 50 100 150 200 250 300 350Base pairs

120 °C

112 °C

104 °C

96 °C

88 °C

80 °C

72 °C

64 °C

56 °C

5' EcoRI fragment 3' EcoRI fragment

GC clamp

GC clamp

R3480P R3500Q

R3500QT3552TR3531C

Bulletin 2171 6/19/98 1:40 PM Page 1

Molecular Bioscience Group

Bio-Rad Laboratories Main Office, 2000 Alfred Nobel Drive, Hercules, California 94547, Ph. (510) 741-1000, Fx. (510) 741-5800Also in: Reagents Park, Australia, Ph. 02-9914-2800, Fx. 02-9914-2888 Wien, Austria, Ph. (1) 877 89 01, Fx. (1) 876 56 29 Nazareth, Belgium, Ph. 09-385 55 11, Fx. 09-385 65 54Mississauga, Canada, Ph. (905) 712-2771, Fx. (905) 712-2990 Beijing, China, Ph. (01) 2046622, Fx. (01) 2051876 Copenhagen, Denmark, Ph. 39 17 9947, Fx. 39 27 1698Espoo, Finland, Ph. 90 804 2200, Fx. 90 804 1100 Ivry sur Seine Cedex, France, Ph. (1) 49 60 68 34, Fx. (1) 46 71 24 67 München, Germany, Ph. 089 31884-0, Fx. 089 31884-100New Delhi, India, Ph. 91-11-461-0103, Fx. 91-11-461-0765 Milano, Italy, Ph. 02-21609.1, Fx. 02-21609.399 Tokyo, Japan, Ph. 03-5811-6270 Fx. 03-5811-6272 Veenendaal, The Netherlands, Ph. 0318-540666, Fx. 0318-542216 Auckland, New Zealand, Ph. 09-443 3099, Fx. 09-443 3097 Kowloon, Hong Kong, Ph. 7893300, Fx. 7891257 Singapore, Ph. (65) 272-9877, Fx. (65) 273-4835 Solna, Sweden, Ph. 46 (0) 8 735 83 00, Fx 46 (0) 735 54 60 Madrid, Spain, Ph. (91) 661 70 85, Fx. (91) 661 96 98 Glattbrugg, Switzerland, Ph. 01/809 55 55, Fx. 01/809 55 00 Hemel Hempstead, United Kingdom, Ph. 0800 181134, Fx. 01442 259118

SIG 020996 Printed in USA

ResultsWhile no new mutations were found in the apoA-I promoterregion, we were able to detect, using DGGE, a common polymorphism at nucleotide -75 (Figure 2), which had beenpreviously shown to be associated with differences in plasmahigh-density lipoprotein (HDL) levels.

Three mutations in the LDL receptor binding domain of theapoB gene that cause the disorder familial ligand-defectiveapoB (FDB)5, 6, 7 were detected by DGGE (Figure 3, lanes 3, 4and 5). In addition, a rare missense mutation at codon 3480(Figure 3, lane 1) and a new silent mutation (Figure 3, lane 2)were detected.

DiscussionDetection of naturally mutant alleles of candidate genes and astudy of their consequent effects can be expected to provideimportant clues to the precise biochemical roles of these geneproducts. For example, new causes of atherogenic dyslipidemiawill be revealed in the search for mutations in proteins involvedin lipid metabolism. We have shown that the DGGE approachis an excellent method to detect these mutations.

This work was supported by National Institutes of HealthGrants HL14237 (Arteriosclerosis Specialized Center ofResearch), HL50782, HL50779, by a gift from Donald andSusan Schleicher and a grant from the Joseph DrownFoundation.

References1. Myers, R. M., Maniatis, T. and Lerman, L. S., Methods in

Enzymology, 155, 501–527 (1987).

2. Myers, R. M., Fischer, S. G., Lerman, L. S. and Maniatis, T.,Nuc. Acids Res., 13, 3131–3145 (1985).

3. Lerman, L. S. and Silverstein, K., Methods in Enzymology,155, 482–501 (1987).

4. Karlsen, F., Steen, H. B. and Nesland, J. M., J. Virol. Methods,55, 153–156 (1995).

5. Soria, L. F., et al. Proc. Natl. Acad. Sci. USA, 86, 587–591(1989).

6. Pullinger, C. R., et al. J. Clin. Invest., 95, 1225–1234 (1995).

7. Gaffney, D., et al. Arterioscler. Thromb. Vasc. Biol., 15,1025–1029 (1995).

* The Polymerase Chain Reaction (PCR) process is covered by patentsowned by Hoffmann-LaRoche. Use of the PCR process requires a license.

Bio-Rad Laboratories

Bulletin 2171 US/EG REV A 97-1037 0497

Fig. 2. DGGE gel of the apolipoprotein A-I gene promoter showing an MspI polymorphism. Lanes 1 and 4 are homozygousfor a C, and lanes 2 and 6 are homozygous for an A, at position -75.Lanes 3 and 5 are heterozygous subjects with the low-melting heteroduplexes clearly visible.

1 2 3 4 5 6

MspI +/+ -/- +/- +/+ +/- -/-

Fig. 3. DGGE gel of part of the receptor binding domain of the apolipoprotein B gene showing five separate single-basesubstitutions. Four of the mutations, lanes 2 through 5, are on the5’ fragment (following EcoRI digestion) and, in each case, this bandshows splitting. The fifth mutation (R3480P, lane 1) is on the 3’fragment and is only detectable because of the presence of heteroduplex bands.

1 2 3 4 5 6

Nucleotide–75:

A ➝C ➝

Heteroduplex{➝Bands

–5’ fragment

3’ fragment:

} Heteroduplexes

} Homoduplexes

3’ fragment

5’ fragment

R34

80P

T355

2T

R35

00Q

R35

00W

R35

31C

No

rmal

➝

Bulletin 2171 6/19/98 1:40 PM Page end1