- Email: [email protected]
Deletions of the low density lipoprotein receptor gene underlying familial hypercholesterolaemia: screening by polymerase chain reaction using pooled DNA and blood samples
Molecular and Cellular Probes (1997) 11, 65–70
Deletions of the low density lipoprotein receptor gene underlying familial hypercholesterolaemia: screening by polymerase chain reaction using pooled DNA and blood samples Alpo F. Vuorio,1,3∗ Lars Paulin,1 Hannu Turtola2 and Kimmo Kontula3 1
Institute of Biotechnology, P.O. Box 56, University of Helsinki, FIN-00014, Finland, Central Hospital of North Karelia, FIN-80210 Joensuu, Finland and 3Department of Medicine, University of Helsinki, FIN-00290 Helsinki, Finland
2
(Received 4 October 1996, Accepted 31 October 1996) We evaluated the feasibility of methods based on the polymerase chain reaction (PCR) and nonautomated or automated gel electrophoresis to detect clinically important DNA deletions in pooled DNA and blood samples. Two common low density lipoprotein (LDL) receptor mutations causing familial hypercholesterolaemia (FH) in the Finnish population were easily identified in pools corresponding to 20 individuals. One of these mutations (FH-North Karelia) deletes seven nucleotides from exon 6 of the LDL receptor gene. PCR amplification of DNA samples from the heterozygous patients with the FH-North Karelia gene results in the formation of DNA heteroduplexes, which markedly improves mutation detection. These studies show the applicability of semi-automated PCR techniques in the screening of DNA deletions and demonstrate the clinical diagnostic usefulness of heteroduplex formation. 1997 Academic Press Limited
KEYWORDS: diagnostic, PCR, heteroduplex, automated sequencer, fluorescein.
INTRODUCTION Familial hypercholesterolaemia (FH) is characterized by lifelong elevation of serum low density lipoprotein (LDL) levels, tendon xanthomatosis and premature atherosclerosis.1 FH is caused by mutations of the LDL receptor gene, and in its heterozygous form is one of the most common inherited diseases.1 A large variety of mutant LDL receptor genes have been reported, ranging from single-nucleotide substitutions to small and large deletions.2 In several inbred populations, one to three founder mutations have been shown to account for the majority of genes causing FH.2 In Finland, two mutant genes make up about two-thirds of the FH mutations. The FH-Helsinki
allele (FH-Hki) is a deletion of approximately 9·5 kb extending from intron 15 to exon 183 while the FHNorth Karelia allele (FH-NK) deletes 7 nucleotides from exon 6.4 DNA diagnosis appears to have definite value in an unequivocal and early diagnosis of heterozygous FH.5 In attempts to screen relatively rare but treatable genetic diseases in a population, methods based on sample pooling may be of great benefit. The present study was undertaken in order to evaluate the usefulness of a tailor-made duplex polymerase chain reaction (PCR) assay6 to detect heterozygous FH patients in pooled DNA and whole-blood samples.
∗ Author to whom all correspondence should be addressed at: Department of Medicine, University of Helsinki, FIN-00290 Helsinki, Finland.
0890–8508/97/010065+06 $25.00/0/ll960078
1997 Academic Press Limited
A. F. Vuorio et al.
66 FH-NK
FH-Hki
93 bp
159 bp
6
16
P1
P2
17
P3
18
P4
100 bp
P5
209 bp
Fig. 1. Schematic illustration of the two deletions of the LDL receptor gene and the principle of the duplex PCR assay. The FH-NK mutation deletes 7 bp from exon 6 and the FH-Hki mutation deletes approximately 9500 bp, extending from intron 15 to exon 18. When a normal LDL receptor allele is amplified, fragments of 100 and 209 bp are formed. In DNA samples from heterozygous FH-NK and heterozygous FH-Hki patients additional bands with sizes of 93 and 159 bp, respectively, are generated.
MATERIALS AND METHODS DNA was isolated from venous blood samples according to Vandenplats et al.7 Blood samples were obtained from known heterozygous FH patients with the FH-Hki allele or the FH-NK allele, from a previously recognized compound heterozygous patient possessing both of these deletions,6 and from healthy controls. DNA concentration was measured spectrophotometrically at 260 nm. The five oligonucleotide primers6 P1 (5′-GCATCACCCTGGACAAAGTC-3′), P2 (5′-GCAAGCCGCCTGCACCGAGACTCAC-3′), P3 (5′-AACAGTTCTTGCCCTCTTTG-3′), P4 (5′-CTGAGACACCCGGTTACCTT-3′) and P5 (5′-ATCCCAACACACACGACAGA3′) were synthesized on an Applied Biosystems model 392 DNA synthesizer by standard phosphoramidite chemistry. Primers P1 and P5 were labelled with fluorescein. The principle of the duplex PCR assay has been described previously,6 locations of the primers as well as the FH-Hki and FH-NK deletions are schematically illustrated in Fig. 1. In pooling experiments, either DNA samples from individual subjects were used, or DNA to be analysed was prepared from mixed whole-blood samples. For each amplification, 150 ng of DNA was used. Conditions for PCR were essentially as described previously.6 The amplified DNA fragments (20 ll) were analysed by electrophoresis on non-denaturing 12% vertical polyacrylamide gel in 0·09 Tris-borate, 0·09 boric acid, 0·002 EDTA (TBE buffer), pH 7·5, at 250 V for 5 h. After electrophoresis, the amplified DNA fragments were visualized by ethidium bromide staining. The molecular weight marker (100 Base-pair Ladder, Pharmacia Biotech) was used to determine the molecular sizes of the amplified fragments. The
amplified DNA fragments (amount adjusted to be within detection scale) were also analysed with an automated DNA Sequencer (A.L.F., Pharmacia LKB Biotechnology AB) employing 5% denaturing polyacrylamide gels. The relative amount of PCR product obtained in each sample was determined with the aid of the A.L.F. DNA Fragment Manager V1.0 program (Pharmacia LKB, Molecular Biology System Division, Uppsala, Sweden).
RESULTS PCR amplification of DNA samples from healthy controls resulted in the formation of DNA fragments of 209 and 100 bp in size (Fig. 1). In the case of a normal LDL receptor allele, primers P3 and P5 are separated by a distance of about 9500 bp and thus do not result in a visible PCR product, whereas the same primers generate a fragment of 159 bp from the mutant FH-Hki gene (Fig. 1). The combination of primers P4 and P5 serve to identify an intact exon 18 (Fig. 1). Primers P1 and P2 generate either a 100bp fragment (normal allele) or a 93-bp fragment (FHNK gene) (Fig. 1). Our previous experience showed that another band corresponding to a migration rate of about 230 bp is generated from heterozygous FHNK patients, apparently due to a heteroduplex formation by annealing of the two (normal and mutant) different PCR products.6 Firstly, DNA from healthy controls and an FH-NK patient were pooled, and analyses using duplex PCR and polyacrylamide gel electrophoresis were carried out. When the ratio of quantities of DNA from an FHNK invidual to control DNA was 1/5, the fragments derived from the normal allele (209 and 100 bp), the somewhat larger heteroduplex band and the 93-bp
Screening for LDL deletions (a)
1
2
3
4
5
6
7
8
9 10 11 12
(b)
200
200
100
100
1
2
67 3
4
5
6
7
8
9
10 11 12
Fig. 2. Analysis of (a) DNA pools from healthy subjects and a heterozygous FH-NK patient, and (b) pools from healthy subjects and a heterozygous FH-Hki patient using PCR and non-denaturing polyacrylamide gel electrophoresis. The amount of DNA in each amplification reaction was 150 ng. Molecular sizes (in bp) are indicated on the left. In the following, the dilutions of the mutant DNA sample with normal DNA samples are indicated as ratios mutant/normal. Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/10; lane 4, pool 2/10; lane 5, pool 3/10; lane 6, pool 4/10; lane 7, pool 5/10; lane 8, pool 6/10; lane 9, pool 7/10; lane 10, pool 8/10; lane 11, pool 9/10; and lane 12, an unmixed sample from a heterozygous FH-NK (a) or FH-Hki (b) patient.
fragment from the deleted allele were visible (Fig. 2a). However, when the DNA ratio was diminished to 1/10, only the heteroduplex band was seen, the 93-bp band from the deleted allele being undetectable (Fig. 2a). To analyse pools of DNA from an FH-NK patient and control subjects by fluorescence detection, we went to further dilutions so that the ratios between FH-NK and normal DNA reached values of 1/15 and 1/20. Even at these dilutions the deleted allele was still visible (Fig. 3b). Because the gel in this case was denaturing, no heteroduplex band was formed. Comparison of the peak area percentages corresponding to the deleted allele (93 bp) and the normal allele (100 bp) is illustrated in Fig. 3b. Secondly, we pooled DNA from healthy controls with that derived from a heterozygous FH-Hki patient. The presence of mutant DNA is revealed by the formation of a 159-bp band, in addition to the 209bp and 100-bp bands generated from the normal allele (Fig. 2b). When the DNA concentration ratio (FH-Hki patient/healthy control) was 1/10, the fragment produced by the FH-Hki allele was still seen on polyacrylamide gel analysis. When the analysis of the FH-Hki allele was carried out using an automated sequencer, the deleted allele could be easily detected even at additional dilutions of 1/15 and 1/20 (Fig. 3c). Peak area percentages of the bands corresponding to the deleted allele (159 bp) compared to the normal allele (209 bp) are shown in Fig. 3c. Thirdly, a similar analysis was carried out using DNA from a compound heterozygous patient having both the FH-Hki and FH-NK mutations. The two mutant alleles were thus shown to be effectively detected, even when derived from a single DNA sample (Fig. 3a). For routine screening purposes, the use of pooled whole-blood samples may offer advantages over the use of pooled DNA samples. Therefore, we performed
additional experiments by diluting aliquots of EDTAanticoagulated blood from heterozygous FH patients with increasing amounts of blood from healthy controls, isolated DNA from these pools, and repeated the assays (polyacrylamide gel electrophoresis followed by ethidium bromide staining) described above. When blood from an FH-NK patient was diluted to 1/10, both the heteroduplex band and the fragment from the deleted allele (93 bp) were visible, but when the dilution was extended to 1/15, only the heteroduplex band was visible (Fig. 4a). Similar analysis of the sample with the FH-Hki gene revealed that this mutation was likewise detectable at blood dilutions of 1/10–1/15 (Fig. 4b).
DISCUSSION Screening for DNA sequence variants associated with, or causally related to, human disease is achieving increasing importance in genetic–epidemiological studies. Thus far most of the screening methods have been designed for detection of point mutations. These methods include ligase-mediated analysis,8 ‘mini-sequencing’ or primer-guided nucleotide incorporation assay,9 and direct fluorescence analysis by hybridization with oligonucleotide arrays on glass supports.10 The technique described in this paper illustrates the feasibility of detection of small and large deletions in population-based blood and DNA samples. A mutation analysis carried out using an automatic sequencer is suitable for screening purposes, avoids the use of radioactive reagents, and allows the detection of small and large DNA deletions in pools of at least 20 DNA samples. The heteroduplex DNA band generated by amplification of a portion of exon 6 from DNA samples from heterozygous FH-NK patients is of particular
A. F. Vuorio et al.
68 (a)
1/2 1
2
3
110
120
130
4
140
150
160
170
180
(min) (b)
1/5 2
(c)
1/5 2
1
4 3
4 1/10
1/10
1/15
1/15
1/20
1/20
110 120 130 140 150 160 170 180 Calculated peak area FH-NK/healthy (%) 178/2269 = 7.8 45/1195 = 3.8 36/1031 = 3.5 156/5437 = 2.9
Expected (%) 10 5 3.3 2.5
120
130
140
150
Calculated peak area FH-Hki/healthy (%) 1198/2839 = 42 435/2694 = 16 539/3780 = 14 223/2582 = 8.6
160
170
180
Expected (%) 10 5 3.3 2.5
Fig. 3. Analysis of (a) DNA pool from a compound heterozygous patient possessing both the FH-Hki and FH-NK mutations, (b) DNA pools from healthy subjects and a heterozygous FH-NK patient, and (c) DNA pools from healthy subjects and a heterozygous FH-Hki patient using PCR and denaturing polyacrylamide gel electrophoresis in an automated DNA sequencer. Pool ratios (mutant/normal) are indicated (1/2, 1/5, 1/10, 1/15 and 1/20). Peak designation: 1=93-bp fragment from an FH-NK carrier; 2=100-bp fragment from the normal allele; 3=159-bp fragment from an FH-Hki carrier; and 4=209-bp fragment from the normal allele. Peak areas are counted with the aid of the A.L.F. DNA fragment manager V1.0 program. Expected peak area ratio is calculated to be the pool ratio/2 (because patients are heterozygous).
interest. It seems to improve the sensitivity of our assay, partly because it is larger in size and therefore is able to bind ethidium bromide more efficiently. Upon increasing dilutions with normal DNA, samples containing the FH-NK allele appear to ‘lose’ relatively quickly the 93-bp PCR product (Fig. 1), apparently because the relative overdose of the normal 100-bp PCR product interacts with it, leading to heteroduplex
formation. In fact, this is the basis whereby our technique can be applied for semi-quantitative estimation of the ratio between the mutant and intact DNA in the pooled samples: when only the heteroduplex band is visualized, the mutant/normal DNA concentration ratio must be equal to or less than 1/10. It has been shown earlier that heteroduplex formation may be useful in detecting a 4-bp insertion-mutation of in-
Screening for LDL deletions (a)
1
2
3
4
5
6
7
8
(b)
200
200
100
100
69 1
2
3
4
5
Fig. 4. Analysis of (a) whole-blood pools from healthy subjects and a heterozygous FH-NK patient, and (b) wholeblood pools from healthy subjects and a heterozygous FH-Hki patient using PCR and non-denaturing polyacrylamide gel electrophoresis. Molecular sizes (in bp) are indicated on the left. In the following, the dilutions of the mutant blood sample with normal blood samples are indicated as ratios mutant/normal. (a) Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/15; lane 4, pool 1/10; lane 5, pool 1/5; lane 6, pool 1/3; lane 7, pool 1/2; and lane 8, an unmixed sample from a heterozygous FH-NK patient. (b) Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/15; lane 4, pool 1/10; and lane 5, an unmixed sample from a heterozygous FH-Hki patient.
dividual samples.11 Heteroduplex formation has also been useful in detecting sequence variation in populations12 or between two species.13 When analysing pools with an automated DNA sequencer and calculating peak area ratios (mutated/ healthy) (Fig. 3), it was possible to obtain semiquantitative estimation of the amount of mutated DNA in a pool. In the case of the FH-NK mutation (Fig. 3b) the estimation was near the expected, but this was not so in the FH-Hki mutation (Fig. 3c). This may be due to bigger size difference between peaks 3 and 4 compared to peaks 1 and 2 (Fig. 3). Shorter PCR products are amplified more effectively. If only detection of the mutation is needed (with an automated DNA sequencer) it is possible to overload the gel. However, quantitative estimation is not then possible, because no peak area can be obtained, as seen in Fig. 3c (peak 2). In conclusion, our study shows that PCR techniques combined with semi-automated DNA detection instruments can be applied to the detection of small and large DNA deletions in pooled DNA or blood samples. By virtue of heteroduplex formation in samples from heterozygous patients, small DNA deletions may occasionally be particularly easily detected.
ACKNOWLEDGEMENTS The expert technical assistance of Ms Kaija Kettunen and Ms Pa¨ivi Laamanen is gratefully acknowledged. This work was supported by grants from the Finnish Academy of Sciences, the Sigrid Juselius Foundation, the University of Helsinki, the Paulo Foundation, the Finnish Heart Foundation, the Orion Corporation Research Foundation, Finnish
Cultural Foundation and the Finnish Medical Society Duodecim.
REFERENCES 1. Goldstein, J. L. & Brown, M. S. (1989). Familial hypercholesterolemia. In The Metabolic Basis of Inherited Disease. (Scriver, C. S., Beaudet, A. L., Sly, W. S. & Valle, D., eds) Pp. 1215–50. New York, NY: McGraw-Hill. 2. Hobbs, H. H., Brown, M. S. & Goldstein, J. L. (1992). Molecular genetics of the LDL receptor gene in familial hypercholesterolemia. Human Mutation 1, 445–66. 3. Aalto-Seta¨la¨, K., Helve, E., Kovanen, P. T. & Kontula, K. (1989). Finnish type of low density lipoprotein receptor mutation (FH-Helsinki) deletes exons encoding the carboxy-terminal part of the receptor and creates an internalization-defective phenotype. Journal of Clinical Investigations 84, 499–505. 4. Koivisto, U-M., Turtola, H., Aalto-Seta¨la¨, K. et al. (1992). The familial hypercholesterolemia (FH)-North Karelia mutation of the low density lipoprotein receptor gene deletes seven nucleotides of exon 6 and is a common cause of FH in Finland. Journal of Clinical Investigations 90, 219–28. 5. Koivisto, P. V., Koivisto, U-M., Miettinen, T. A. & Kontula, K. (1992). Diagnosis of heterozygous familial hypercholesterolemia. DNA analysis complements clinical examination and analysis of serum lipid levels. Arteriosclerosis and Thrombosis 12, 584–92. 6. Koivisto, U-M., Ha¨ma¨la¨inen, L., Taskinen, M-R., Kettunen, K. & Kontula, K. (1993). Prevalence of familial hypercholesterolemia among young North Karelian patients with coronary heart disease: a study based on diagnosis by polymerase chain reaction. Journal of Lipid Research 34, 269–77. 7. Vandenplas, S., Wiid, I., Brebner, O. K. et al. (1984). Blot hybridisation of genomic DNA. Journal of Medical Genetics 21, 164–72.
70
A. F. Vuorio et al.
8. Samiotaki, M., Kwiatkowski, M., Parik, J. & Landegren, U. (1994). Dual-color detection of DNA sequence variants by ligase-mediated analysis. Genomics 20, 238–42. 9. Syva¨nen, A-C., Aalto-Seta¨la¨, K., Harju, L., Kontula, K. & So¨derlund, H. (1990). A primer-guided nucleotide incorporation assay in the genotyping apolipoprotein E. Genomics 8, 684–92. 10. Guo, Z., Guilfoyle, A., Thiel, A. J., Wang, R. & Smith, L. M. (1994). Direct fluorescence analysis of genetic polymorphisms by hybridization with oligonucleotide arrays on glass supports. Nucleic Acids Research 22, 5456–65.
11. Triggs-Raine, B. & Gravel, R. A. (1990). Diagnostic heteroduplexes: simple detection of carriers 4-bp insertion mutation in Tay-sachs disease (letter). American Journal of Human Genetics 46, 183–4. 12. Ruano, G., Deinard, A. S., Tishkoff, S. & Kidd, K. K. (1994). Detection of DNA sequence variation via deliberate heteroduplex formation from genomic DNAs amplified en masse in ‘population tubes’. PCR Methods and Applications 3, 225–31. 13. Ruano, G. & Kidd, K. K. (1992). Modeling of heteroduplex formation during PCR from mixtures of DNA templates. PCR Methods and Applications 2, 112–16.
Deletions of the low density lipoprotein receptor gene underlying familial hypercholesterolaemia: screening by polymerase chain reaction using pooled DNA and blood samples Alpo F. Vuorio,1,3∗ Lars Paulin,1 Hannu Turtola2 and Kimmo Kontula3 1
Institute of Biotechnology, P.O. Box 56, University of Helsinki, FIN-00014, Finland, Central Hospital of North Karelia, FIN-80210 Joensuu, Finland and 3Department of Medicine, University of Helsinki, FIN-00290 Helsinki, Finland
2
(Received 4 October 1996, Accepted 31 October 1996) We evaluated the feasibility of methods based on the polymerase chain reaction (PCR) and nonautomated or automated gel electrophoresis to detect clinically important DNA deletions in pooled DNA and blood samples. Two common low density lipoprotein (LDL) receptor mutations causing familial hypercholesterolaemia (FH) in the Finnish population were easily identified in pools corresponding to 20 individuals. One of these mutations (FH-North Karelia) deletes seven nucleotides from exon 6 of the LDL receptor gene. PCR amplification of DNA samples from the heterozygous patients with the FH-North Karelia gene results in the formation of DNA heteroduplexes, which markedly improves mutation detection. These studies show the applicability of semi-automated PCR techniques in the screening of DNA deletions and demonstrate the clinical diagnostic usefulness of heteroduplex formation. 1997 Academic Press Limited
KEYWORDS: diagnostic, PCR, heteroduplex, automated sequencer, fluorescein.
INTRODUCTION Familial hypercholesterolaemia (FH) is characterized by lifelong elevation of serum low density lipoprotein (LDL) levels, tendon xanthomatosis and premature atherosclerosis.1 FH is caused by mutations of the LDL receptor gene, and in its heterozygous form is one of the most common inherited diseases.1 A large variety of mutant LDL receptor genes have been reported, ranging from single-nucleotide substitutions to small and large deletions.2 In several inbred populations, one to three founder mutations have been shown to account for the majority of genes causing FH.2 In Finland, two mutant genes make up about two-thirds of the FH mutations. The FH-Helsinki
allele (FH-Hki) is a deletion of approximately 9·5 kb extending from intron 15 to exon 183 while the FHNorth Karelia allele (FH-NK) deletes 7 nucleotides from exon 6.4 DNA diagnosis appears to have definite value in an unequivocal and early diagnosis of heterozygous FH.5 In attempts to screen relatively rare but treatable genetic diseases in a population, methods based on sample pooling may be of great benefit. The present study was undertaken in order to evaluate the usefulness of a tailor-made duplex polymerase chain reaction (PCR) assay6 to detect heterozygous FH patients in pooled DNA and whole-blood samples.
∗ Author to whom all correspondence should be addressed at: Department of Medicine, University of Helsinki, FIN-00290 Helsinki, Finland.
0890–8508/97/010065+06 $25.00/0/ll960078
1997 Academic Press Limited
A. F. Vuorio et al.
66 FH-NK
FH-Hki
93 bp
159 bp
6
16
P1
P2
17
P3
18
P4
100 bp
P5
209 bp
Fig. 1. Schematic illustration of the two deletions of the LDL receptor gene and the principle of the duplex PCR assay. The FH-NK mutation deletes 7 bp from exon 6 and the FH-Hki mutation deletes approximately 9500 bp, extending from intron 15 to exon 18. When a normal LDL receptor allele is amplified, fragments of 100 and 209 bp are formed. In DNA samples from heterozygous FH-NK and heterozygous FH-Hki patients additional bands with sizes of 93 and 159 bp, respectively, are generated.
MATERIALS AND METHODS DNA was isolated from venous blood samples according to Vandenplats et al.7 Blood samples were obtained from known heterozygous FH patients with the FH-Hki allele or the FH-NK allele, from a previously recognized compound heterozygous patient possessing both of these deletions,6 and from healthy controls. DNA concentration was measured spectrophotometrically at 260 nm. The five oligonucleotide primers6 P1 (5′-GCATCACCCTGGACAAAGTC-3′), P2 (5′-GCAAGCCGCCTGCACCGAGACTCAC-3′), P3 (5′-AACAGTTCTTGCCCTCTTTG-3′), P4 (5′-CTGAGACACCCGGTTACCTT-3′) and P5 (5′-ATCCCAACACACACGACAGA3′) were synthesized on an Applied Biosystems model 392 DNA synthesizer by standard phosphoramidite chemistry. Primers P1 and P5 were labelled with fluorescein. The principle of the duplex PCR assay has been described previously,6 locations of the primers as well as the FH-Hki and FH-NK deletions are schematically illustrated in Fig. 1. In pooling experiments, either DNA samples from individual subjects were used, or DNA to be analysed was prepared from mixed whole-blood samples. For each amplification, 150 ng of DNA was used. Conditions for PCR were essentially as described previously.6 The amplified DNA fragments (20 ll) were analysed by electrophoresis on non-denaturing 12% vertical polyacrylamide gel in 0·09 Tris-borate, 0·09 boric acid, 0·002 EDTA (TBE buffer), pH 7·5, at 250 V for 5 h. After electrophoresis, the amplified DNA fragments were visualized by ethidium bromide staining. The molecular weight marker (100 Base-pair Ladder, Pharmacia Biotech) was used to determine the molecular sizes of the amplified fragments. The
amplified DNA fragments (amount adjusted to be within detection scale) were also analysed with an automated DNA Sequencer (A.L.F., Pharmacia LKB Biotechnology AB) employing 5% denaturing polyacrylamide gels. The relative amount of PCR product obtained in each sample was determined with the aid of the A.L.F. DNA Fragment Manager V1.0 program (Pharmacia LKB, Molecular Biology System Division, Uppsala, Sweden).
RESULTS PCR amplification of DNA samples from healthy controls resulted in the formation of DNA fragments of 209 and 100 bp in size (Fig. 1). In the case of a normal LDL receptor allele, primers P3 and P5 are separated by a distance of about 9500 bp and thus do not result in a visible PCR product, whereas the same primers generate a fragment of 159 bp from the mutant FH-Hki gene (Fig. 1). The combination of primers P4 and P5 serve to identify an intact exon 18 (Fig. 1). Primers P1 and P2 generate either a 100bp fragment (normal allele) or a 93-bp fragment (FHNK gene) (Fig. 1). Our previous experience showed that another band corresponding to a migration rate of about 230 bp is generated from heterozygous FHNK patients, apparently due to a heteroduplex formation by annealing of the two (normal and mutant) different PCR products.6 Firstly, DNA from healthy controls and an FH-NK patient were pooled, and analyses using duplex PCR and polyacrylamide gel electrophoresis were carried out. When the ratio of quantities of DNA from an FHNK invidual to control DNA was 1/5, the fragments derived from the normal allele (209 and 100 bp), the somewhat larger heteroduplex band and the 93-bp
Screening for LDL deletions (a)
1
2
3
4
5
6
7
8
9 10 11 12
(b)
200
200
100
100
1
2
67 3
4
5
6
7
8
9
10 11 12
Fig. 2. Analysis of (a) DNA pools from healthy subjects and a heterozygous FH-NK patient, and (b) pools from healthy subjects and a heterozygous FH-Hki patient using PCR and non-denaturing polyacrylamide gel electrophoresis. The amount of DNA in each amplification reaction was 150 ng. Molecular sizes (in bp) are indicated on the left. In the following, the dilutions of the mutant DNA sample with normal DNA samples are indicated as ratios mutant/normal. Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/10; lane 4, pool 2/10; lane 5, pool 3/10; lane 6, pool 4/10; lane 7, pool 5/10; lane 8, pool 6/10; lane 9, pool 7/10; lane 10, pool 8/10; lane 11, pool 9/10; and lane 12, an unmixed sample from a heterozygous FH-NK (a) or FH-Hki (b) patient.
fragment from the deleted allele were visible (Fig. 2a). However, when the DNA ratio was diminished to 1/10, only the heteroduplex band was seen, the 93-bp band from the deleted allele being undetectable (Fig. 2a). To analyse pools of DNA from an FH-NK patient and control subjects by fluorescence detection, we went to further dilutions so that the ratios between FH-NK and normal DNA reached values of 1/15 and 1/20. Even at these dilutions the deleted allele was still visible (Fig. 3b). Because the gel in this case was denaturing, no heteroduplex band was formed. Comparison of the peak area percentages corresponding to the deleted allele (93 bp) and the normal allele (100 bp) is illustrated in Fig. 3b. Secondly, we pooled DNA from healthy controls with that derived from a heterozygous FH-Hki patient. The presence of mutant DNA is revealed by the formation of a 159-bp band, in addition to the 209bp and 100-bp bands generated from the normal allele (Fig. 2b). When the DNA concentration ratio (FH-Hki patient/healthy control) was 1/10, the fragment produced by the FH-Hki allele was still seen on polyacrylamide gel analysis. When the analysis of the FH-Hki allele was carried out using an automated sequencer, the deleted allele could be easily detected even at additional dilutions of 1/15 and 1/20 (Fig. 3c). Peak area percentages of the bands corresponding to the deleted allele (159 bp) compared to the normal allele (209 bp) are shown in Fig. 3c. Thirdly, a similar analysis was carried out using DNA from a compound heterozygous patient having both the FH-Hki and FH-NK mutations. The two mutant alleles were thus shown to be effectively detected, even when derived from a single DNA sample (Fig. 3a). For routine screening purposes, the use of pooled whole-blood samples may offer advantages over the use of pooled DNA samples. Therefore, we performed
additional experiments by diluting aliquots of EDTAanticoagulated blood from heterozygous FH patients with increasing amounts of blood from healthy controls, isolated DNA from these pools, and repeated the assays (polyacrylamide gel electrophoresis followed by ethidium bromide staining) described above. When blood from an FH-NK patient was diluted to 1/10, both the heteroduplex band and the fragment from the deleted allele (93 bp) were visible, but when the dilution was extended to 1/15, only the heteroduplex band was visible (Fig. 4a). Similar analysis of the sample with the FH-Hki gene revealed that this mutation was likewise detectable at blood dilutions of 1/10–1/15 (Fig. 4b).
DISCUSSION Screening for DNA sequence variants associated with, or causally related to, human disease is achieving increasing importance in genetic–epidemiological studies. Thus far most of the screening methods have been designed for detection of point mutations. These methods include ligase-mediated analysis,8 ‘mini-sequencing’ or primer-guided nucleotide incorporation assay,9 and direct fluorescence analysis by hybridization with oligonucleotide arrays on glass supports.10 The technique described in this paper illustrates the feasibility of detection of small and large deletions in population-based blood and DNA samples. A mutation analysis carried out using an automatic sequencer is suitable for screening purposes, avoids the use of radioactive reagents, and allows the detection of small and large DNA deletions in pools of at least 20 DNA samples. The heteroduplex DNA band generated by amplification of a portion of exon 6 from DNA samples from heterozygous FH-NK patients is of particular
A. F. Vuorio et al.
68 (a)
1/2 1
2
3
110
120
130
4
140
150
160
170
180
(min) (b)
1/5 2
(c)
1/5 2
1
4 3
4 1/10
1/10
1/15
1/15
1/20
1/20
110 120 130 140 150 160 170 180 Calculated peak area FH-NK/healthy (%) 178/2269 = 7.8 45/1195 = 3.8 36/1031 = 3.5 156/5437 = 2.9
Expected (%) 10 5 3.3 2.5
120
130
140
150
Calculated peak area FH-Hki/healthy (%) 1198/2839 = 42 435/2694 = 16 539/3780 = 14 223/2582 = 8.6
160
170
180
Expected (%) 10 5 3.3 2.5
Fig. 3. Analysis of (a) DNA pool from a compound heterozygous patient possessing both the FH-Hki and FH-NK mutations, (b) DNA pools from healthy subjects and a heterozygous FH-NK patient, and (c) DNA pools from healthy subjects and a heterozygous FH-Hki patient using PCR and denaturing polyacrylamide gel electrophoresis in an automated DNA sequencer. Pool ratios (mutant/normal) are indicated (1/2, 1/5, 1/10, 1/15 and 1/20). Peak designation: 1=93-bp fragment from an FH-NK carrier; 2=100-bp fragment from the normal allele; 3=159-bp fragment from an FH-Hki carrier; and 4=209-bp fragment from the normal allele. Peak areas are counted with the aid of the A.L.F. DNA fragment manager V1.0 program. Expected peak area ratio is calculated to be the pool ratio/2 (because patients are heterozygous).
interest. It seems to improve the sensitivity of our assay, partly because it is larger in size and therefore is able to bind ethidium bromide more efficiently. Upon increasing dilutions with normal DNA, samples containing the FH-NK allele appear to ‘lose’ relatively quickly the 93-bp PCR product (Fig. 1), apparently because the relative overdose of the normal 100-bp PCR product interacts with it, leading to heteroduplex
formation. In fact, this is the basis whereby our technique can be applied for semi-quantitative estimation of the ratio between the mutant and intact DNA in the pooled samples: when only the heteroduplex band is visualized, the mutant/normal DNA concentration ratio must be equal to or less than 1/10. It has been shown earlier that heteroduplex formation may be useful in detecting a 4-bp insertion-mutation of in-
Screening for LDL deletions (a)
1
2
3
4
5
6
7
8
(b)
200
200
100
100
69 1
2
3
4
5
Fig. 4. Analysis of (a) whole-blood pools from healthy subjects and a heterozygous FH-NK patient, and (b) wholeblood pools from healthy subjects and a heterozygous FH-Hki patient using PCR and non-denaturing polyacrylamide gel electrophoresis. Molecular sizes (in bp) are indicated on the left. In the following, the dilutions of the mutant blood sample with normal blood samples are indicated as ratios mutant/normal. (a) Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/15; lane 4, pool 1/10; lane 5, pool 1/5; lane 6, pool 1/3; lane 7, pool 1/2; and lane 8, an unmixed sample from a heterozygous FH-NK patient. (b) Lane 1, DNA molecular size marker; lane 2, healthy control; lane 3, pool 1/15; lane 4, pool 1/10; and lane 5, an unmixed sample from a heterozygous FH-Hki patient.
dividual samples.11 Heteroduplex formation has also been useful in detecting sequence variation in populations12 or between two species.13 When analysing pools with an automated DNA sequencer and calculating peak area ratios (mutated/ healthy) (Fig. 3), it was possible to obtain semiquantitative estimation of the amount of mutated DNA in a pool. In the case of the FH-NK mutation (Fig. 3b) the estimation was near the expected, but this was not so in the FH-Hki mutation (Fig. 3c). This may be due to bigger size difference between peaks 3 and 4 compared to peaks 1 and 2 (Fig. 3). Shorter PCR products are amplified more effectively. If only detection of the mutation is needed (with an automated DNA sequencer) it is possible to overload the gel. However, quantitative estimation is not then possible, because no peak area can be obtained, as seen in Fig. 3c (peak 2). In conclusion, our study shows that PCR techniques combined with semi-automated DNA detection instruments can be applied to the detection of small and large DNA deletions in pooled DNA or blood samples. By virtue of heteroduplex formation in samples from heterozygous patients, small DNA deletions may occasionally be particularly easily detected.
ACKNOWLEDGEMENTS The expert technical assistance of Ms Kaija Kettunen and Ms Pa¨ivi Laamanen is gratefully acknowledged. This work was supported by grants from the Finnish Academy of Sciences, the Sigrid Juselius Foundation, the University of Helsinki, the Paulo Foundation, the Finnish Heart Foundation, the Orion Corporation Research Foundation, Finnish
Cultural Foundation and the Finnish Medical Society Duodecim.
REFERENCES 1. Goldstein, J. L. & Brown, M. S. (1989). Familial hypercholesterolemia. In The Metabolic Basis of Inherited Disease. (Scriver, C. S., Beaudet, A. L., Sly, W. S. & Valle, D., eds) Pp. 1215–50. New York, NY: McGraw-Hill. 2. Hobbs, H. H., Brown, M. S. & Goldstein, J. L. (1992). Molecular genetics of the LDL receptor gene in familial hypercholesterolemia. Human Mutation 1, 445–66. 3. Aalto-Seta¨la¨, K., Helve, E., Kovanen, P. T. & Kontula, K. (1989). Finnish type of low density lipoprotein receptor mutation (FH-Helsinki) deletes exons encoding the carboxy-terminal part of the receptor and creates an internalization-defective phenotype. Journal of Clinical Investigations 84, 499–505. 4. Koivisto, U-M., Turtola, H., Aalto-Seta¨la¨, K. et al. (1992). The familial hypercholesterolemia (FH)-North Karelia mutation of the low density lipoprotein receptor gene deletes seven nucleotides of exon 6 and is a common cause of FH in Finland. Journal of Clinical Investigations 90, 219–28. 5. Koivisto, P. V., Koivisto, U-M., Miettinen, T. A. & Kontula, K. (1992). Diagnosis of heterozygous familial hypercholesterolemia. DNA analysis complements clinical examination and analysis of serum lipid levels. Arteriosclerosis and Thrombosis 12, 584–92. 6. Koivisto, U-M., Ha¨ma¨la¨inen, L., Taskinen, M-R., Kettunen, K. & Kontula, K. (1993). Prevalence of familial hypercholesterolemia among young North Karelian patients with coronary heart disease: a study based on diagnosis by polymerase chain reaction. Journal of Lipid Research 34, 269–77. 7. Vandenplas, S., Wiid, I., Brebner, O. K. et al. (1984). Blot hybridisation of genomic DNA. Journal of Medical Genetics 21, 164–72.
70
A. F. Vuorio et al.
8. Samiotaki, M., Kwiatkowski, M., Parik, J. & Landegren, U. (1994). Dual-color detection of DNA sequence variants by ligase-mediated analysis. Genomics 20, 238–42. 9. Syva¨nen, A-C., Aalto-Seta¨la¨, K., Harju, L., Kontula, K. & So¨derlund, H. (1990). A primer-guided nucleotide incorporation assay in the genotyping apolipoprotein E. Genomics 8, 684–92. 10. Guo, Z., Guilfoyle, A., Thiel, A. J., Wang, R. & Smith, L. M. (1994). Direct fluorescence analysis of genetic polymorphisms by hybridization with oligonucleotide arrays on glass supports. Nucleic Acids Research 22, 5456–65.
11. Triggs-Raine, B. & Gravel, R. A. (1990). Diagnostic heteroduplexes: simple detection of carriers 4-bp insertion mutation in Tay-sachs disease (letter). American Journal of Human Genetics 46, 183–4. 12. Ruano, G., Deinard, A. S., Tishkoff, S. & Kidd, K. K. (1994). Detection of DNA sequence variation via deliberate heteroduplex formation from genomic DNAs amplified en masse in ‘population tubes’. PCR Methods and Applications 3, 225–31. 13. Ruano, G. & Kidd, K. K. (1992). Modeling of heteroduplex formation during PCR from mixtures of DNA templates. PCR Methods and Applications 2, 112–16.