Use of Polymerase Chain Reaction to Detect

for the treatment of patients with AIDS and related diseases. Clin Pharrnacol Ther 1987;41:407-12. 3. Blum RM, Liao SH, Good SS, et al. Pharmacokinetics and bioavailability of zidovudine in humans. Am J Med 1988;85:189-94. 4. Richman DD, Fischl MA, Grieco MH, et al. The toxicity of azidothymidinc (AZT) in the treatment of patients with AIDS and AIDS-related complex:a double-blind, placebo-controlled trial. N Engl J Med 1987;317:192-7. 5. Good SS, Reynolds DJ, de Miranda P. Simultaneous quantillcation of zidovudine and its glucuromde in serum by high-performance liquid chromatography. J Chromatogr 1988;431:123-33. 6. Unadkat JD, Crosby SB, Wang JP, et al. Simple and rapid high-performance liquidchromatographic assay for zidovudine (azidothymidine) in plasma and urine. J Chromatogr 1988;

7. HedayaMA, Sawchuk RJ. A sensitive liquid-chromatographic method for determination of 3’-azido-3’-deoxythymidine (AZT) in plasma and urine. Clin Chem 1988;348:1565-8. 8. Quinn RP, Tadepalli SM, Orban BS. Radioimmunoassay for azidothymidine using a high-specific activity tritiated antigen. J Ixnmunoasaay 1989;l0:177-89. 9. Tadepalli SM, Quinn RP. Determination of zidovudine concentration in serum by enzyme-linked immunosorbent assay and by time-resolved fluoroimmunoassay. J AIDS 1990;3:19-27. 10. Granich GG, Eveland MR Krogstad N. Fluorescencepolarization immunoassay for zidovudine. Antimicrob Agents Chemother 1989;33:1275-9. 11. McDougal JS, Martin LS, Cort SP, et al. Thermal inactivation of the acquired iznmunodeflciency syndrome virus, human Tlymphotropic virus-ifi/lymphadenopathy associated virus, with special reference to antthemophilic factor. J Clin Invest

1985;76:875-7.

430:420-3.

CLIN. CHEM. 36/6, 900-903 (1990)

Use of Polymerase Chain Reaction to Detect Heterozygous Familial Hypercholesterolemia Maurl KeinAnen,1’ Jukka-P.kka

OJaIa,2E.ro Halve,2 Katrilna Aalto.SetII&L4

We used a modification of the polymerase chain reaction (PCR), invoMng two pairs of oligonucleotide primers, to detect a mutation in the low-density lipoprotein (LDL) receptor gene, commonly occurring among patients with familial hypercholesterolemia (FH) In Finland. This mutation, called FH-Helsinkl, involves a large (about 9500 base pairs, bp)

deletion In the LDL receptor gene extending from Intron 15 to exon 18. For the PCR, one pair of primers was designed to cover both sides of the deletion in its immediate vicinity. In the presence of the deletion, the primers were brought close enough to each other toallow the amplification and electrophoretic detection of a 300-bp amplification product. In the absence of the deletion, no amplification occurred and this band accordingly was not visible in the gel. To render the Interpretation of the results unequivocal, we designed a second pair of oligonucleotide primers. This pair of primers allowed another amplification product (158 bp) to appear in samples containinga normal exon 17, i.e., in DNA spedmens from healthy subjects and FH heterozygotes with or without the FH-Helsinki deletion. The technique is easy to perform, avoids the use of radioactive reagents, and is applicable to the detection of any extensive DNA deletion. AdditIonal K.yphrase.: gene probes receptors electrophoresis,agarose

low-densityilpoprotein

.

A specific deletion receptor

in the low-density gene, designated FH-Helsinki,

lipoprotein (LDL)7 is found in at least

‘Department of Medical Genetics,2 Third Department of Medicine, and 3lnstitute of Biotechnology, University of Helsinki, Helsinki, Finland. 4United Laboratories, Helsinki, Finland. 5Wihuri Research Institute, Helsinki, Finland. 6Medix Clinical Laboratory, Nihtisillankuja 1, 02630 Espoo, Finland (address for correspondence). ‘Nonstandard abbreviations: LDL, low-density lipoprotein; FH, fmi1iaI hypereholeaterolemia; and PCR, polymerase chain reaction. Received February 23, 1990; accepted March 27, 1990. 900 CLINICALCHEMISTRY, Vol.36,No.6,1990

Kimmo Kontula,M and Petrl 1. Kovanen5#{176}

one-third of Finnish patients heterozygous for familial hypercholesterolemia (Fl!) (1, 2). The 9.5-kb (kb = kilobase pairs) deletion leads to loss of exons 16 and 17 and part of exon 18. The mutation appears to give rise to a phenotype in which internalization of LDL particles by the cells is defective (2). The high prevalenceof the deletion justifies the development of a convenient assay for detecting the mutation by using DNA techniques. The deletion can be identifiedby analysis of genomic DNA, with use of a conventional Southern blotting technique with radioactive probes (1). Here we present another method for detecting the FHHelsinki deletion, by using the polymerase chain reaction (PCR) (3). PCR allows rapid visual detection of DNA deletions in ethidium bromide-stained agarose or polyacrylamide gels (4, 5). Small deletions, from a few to several hundreds of base pairs (bp), can be identified in this way by means of the length differences between the amplifled DNA fragments from controls and patients. However, longer deletions are likewise amenable to PCR detection. Thus, if the amplification primers are chosen to represent DNA stretches flanking the two ends of the deletion, the deletion brings the primers close enough to allow DNA amplification; in the absence of the deletion, this technique fails to show any bands, the upper limit for PCR amplification being in the range of 2000 bp (6). To avoid the problem of a false-negative result, we have designed two pairs of amplification primers and used a “duplex” PCR reaction (5, 7) to be able to verify with certainty the presence or absence of the FH-Helsinki mutation. Materials

and Methods

The samples for PCR were obtained from nine patients heterozygous for FH and from normal controls. The diagnosis of FH was based on typical clinical features, including a total serum cholesterol concentration >9.5 mmolfL, the presence of tendon xanthomas, and a positive family history (at least one first-degree family member with hypercholesterolemia), and was confirmed by establishing a

functional defect in the LDL receptor activity in cultured skin fibroblasts from the patients (2). Template DNA was prepared either by extracting genomic DNA by a standard technique involvingdigestion with proteinase K, phenol extraction, and ethanol precipitation (8), or by obtaining DNA from lysed leukocytes. Cell lysis was performed by incubating about 100 000 leukocytes in 100 zL of distilled water at 95#{176}C for 5 mm (6). Before cell lysis, erythrocytes were removed by sedimenting them with the aid of dextran (median molecular mass 500 kDa; Pharmacia, Uppsala, Sweden), 6 g/L, for 30-50 miii at room temperature. About 500 ng of extracted DNA or DNA from 10 pL of cell lysate was used as a template DNA for PCR. Similar results were obtained with DNA samples prepared with the two different methods. The oligonucleotide primers from both sides of the deletion were designed according to the published LDL receptor cDNA sequence (9); the sense primer (a) was derived from exon 15 and the antisense primer (a’) from exon 18 (Figure 1). “Control primers” (b) and (b’) were derived from exon 17 (Figure 1). The PCR reaction was performed according to the method described by Saiki et al. (3). The PCR mixture (100 L) contained the template DNA and, per liter, 0.6 iniol of each oligonucleotide primer and 200 iimol each of dATP, dCTP, dGTP, and dTl”P in “Faq” buffer(containing, per liter, 1.5 mmol of MgCl2, 50 mmol of KC1, 10 mmol of Tris, pH 8.3, and 0.1 g of gelatin). After an initial denaturation step for 10 miii at 95#{176}C, 1 U of the “Faq” polymerase enzyme (Cetus Corp.,Berkeley,CA) was added, the mixture overlayered with 100 L of mineral oil, and the amplification reaction performed as follows: 1.5 miii at

NORMAL a

b-158bp--b’

a’

H 9.5kb deletion FH-HELSINKI

a -SOObp- a’

95 #{176}C for denaturation, 1.2 mm at 56#{176}C for primer annealing, and 2 mm at 72 #{176}C for extension. The last (30th) extension reaction at 72#{176}C was prolonged to 10 miii. Amplification cycles were carried out by using a programmable heater blocker (Techne, Cambridge Ltd., Cambridge, U.K.). After amplification, we applied 20 L of the amplification product to a 2% agarose gel containing ethidium bromide and ran it for 1.5 h at 100 V. For Southern blot analysis, 10-pg DNA samples were digested with the restriction enzyme Pvu II, electrophoresed in 0.6% agarose, transferred to nitrocellulose filters, and hybridized with an LDL receptor cDNA probe covering the exons 11 to 17(1). These conditions reveal a restriction fragment length polymorphism in normal subjects (the presence of 16-kb or 14-kb fragments, or both), whereas patients with the FH-Helsunki mutation display a unique 11-kb fragment (1). Results Four of the nine FH patients had the FH-Helsinki deletion, as determined by the standard Southern blotting technique, and five had another, as yet unidentified, mutation of the LDL receptor gene (Figure 2A). The FH-Helsinki deletion brings the primers (a) and (a’) as close as 250 bp to each other, resulting in an amplified fragment spanning 300 bp (two 25-bp primers and 250 bp between them). These 300-bp fragments were visible in ethidium bromidestained agarose gels whenever the FH-Helsinki mutation was present (Figure 2B, lanes 2, 4, 5, and 7). In the absence of the FH-Helsinki mutation, i.e., in subjects without FH and in FH patients with LDL recep-

tor gene mutations other than the FH-Helsinki type, the distance between the primers (a) and (a’) is about 9.5 kb. This distance shouldbetoo long to permit an amplification of the spanning region between (a) and (a’). Data for five FH patients without the FH-Helsinki mutation are illustrated in Figure 2B (lanes 1,3,6,8, and 9). As expected,no fragments in the 300-bp size class were visible in these lanes. For these patients, the lack of any amplified bands would have made the interpretation of results equivocal. To ensure that the amplification reaction operates correctly even in samples from FH patients without the FH-Helsinki mutation and in samples from normal individuals, we added to the reaction mixtures primers derived from exon 17, (b) and (b’) (Figure 1), capable of yielding an amplification product of 158 bp. Generation of this 158-bp fragment was demonstrated in all samples, whether positive or negative for the FH-Helsinki mutation (Figure 2B, lanes 1-9), including those from healthy controls (data not shown). DiscussIon

--

5’-GAGATAGTGACAATGTCTCACCAAG-3’ 1.a’ 5’-TACAACTCTGAACTGAGAAAGTGCA-3’ Ja

lb

5’-TGCTCCTCGTCTTCCTTTGCCTGGG-3’

tb’

5’-CGAGGGGTAGCTGTAGCCGTCCTGG-3’

FIg. 1. Schematic presentationof a section of the normal WL receptor geneand Its deleted FH-Helsinklcounterpart Primers (a) (5’-GAGATAGTGACAATGTCTCACCAAG-3’) and (a’) (5’TACAACTCTGAACTGAGAAAGTGCA-3’) were used to detect the deleted allele. Primers (b) (5’-TGCTCCTCGTCTTCC1TTGCCTGGG-3’) and (b’) (5’-

CGAGGGGTAGCTGTAGCCGTCCTGG-3’)were used to generatethe amplification of DNAfrom axon 17, to identifythe normal allele. Note that the

exoMntron relations are not drawn Inscale

The FOR is very useful for detecting mutations of genes with known DNA sequences. Identification of single-base changes usually requires hybridization of amplified and filter-immobilized DNA with allele-specific oligonucleotide probes (10-12) or direct sequencing of the amplification product (13-15). If a single nucleotide change creates or eliminRtes a restriction enzyme cleavage site, direct analysis of the mutation becomesfeasible by electrophoretic analysis of the enzyme-digested DNA-amplificate covering this restriction site (16). These types of techniques have been used to set up DNA diagnostic methods for two mutations of the LDL receptor gene very prevalent in South African Africaners (17).

CLINICALCHEMISTRY,Vol. 36, No.6, 1990 901

A.

1314

15

II

I

1u1

16 17

18

I

P

1

II

II

P

I

FH-HELSINKI probe

V

0

#{149}.#{149}/

I

-

I

.

-

___ --I -w#{149} *4w. ,#{248},.. #{149}

1kb

-

-

.

.

11 kb

12345

We thank Dr. D. W. Russell for providing us with the LDL receptor cDNA. The expert technical assistance of Ms. Mira Vuorinen, Meeri Lappalainen, and Maija Kettunen is gratefully acknowledged. This work was supported by grants from the Finnish Academy of Sciences, The University of Helsinki, The Paavo Nurmi Foundation, The Finnish Heart Foundation, and The Sigrid Juselius Foundation.

6789

B.

...300 bp -

158 bp

St

1

2

In summary, our data indicate that a genetic disease characterized by a constant large DNA deletion may be diagnosed by a very convenient “duplex” PCR technique basedon two pairs of oligonucleotideprimers. This method is cheap, rapid, and easy to perform and circumvents the use of radioisotopes.

#{149}.4

V

3 4 5 6 7 8 9

FIg. 2. (A) Southern blot analysis ofthe FH-Helsinkl mutation DNAsamples were digested with Pvv 11(P),electrophoresed In0.6% agarose gel, transferred to nitrocellulose filter, and hybridized with an LDL receptor cDNA probe.The schemeshows the relevant portions of the normal and mutant (FH-HeIsInkl) allele. Lanes, 2, 4, 5, and 7: samples from FH-Helsinki patients with the typIcal11-kb Pvu IIfragment.Lanes 1,3,6.8, and 9: samples fromother FH patients. Lanes 6 and 9: samples from subjects showing the common Pvu IIpolymorphism(P)

(

to closelyspaced intradeletional DNA stretches, would identify the normal allele. The presence or absence of the FH.Helsinki mutation was correctly identified in every case examined, as compared with the results of the standard Southern blot technique. The principle ofthe technique described here shouldalso be applicableto the diagnosis ofFH amongFrench Canadians, in whom about two-thirds of the FH patients are characterized by a large (>10 kb) deletion of the LDL receptor gene (18). of primers, corresponding

NORMAL 10 1112

PCRproductsseparatedin a 2% agarosegel

Lane St 4X174/Hae III molecular size markers; other lanes correspond to those In Fig. 2A The presence of the 300-bpfragmentpositivelyidentifiesthe FH-HelsInki mutation

In patients Jzomozygous for an autosomally inherited disease, large DNA deletions can be detected by demonstrating the absence of an amplification product if the technique involves the use of amplification primers corresponding to portions of the deleted sequence (5). This principle, however, cannot be applied to the detection of similar deletions in heterozygous patients, because the normal allele would serve as a template for the generation of this amplificate. Small deletions can be detected by using the PCR technique and primers flanking the deletion at

both its ends;in these conditionsthe mutant allele generates a DNA ampliflcate of smaller size than that produced by the normal allele (4). The latter technique would be unsuitable for the identification of the extensive FH-Helsinki deletion. We circumvented these problemsby using a “duplex” FOR technique, in which one set of primers, corresponding to the immediate flanking regions of the deletion, would identify the mutant allele, and the other set 902 CLINICALCHEMISTRY,Vol.36, No.6, 1990

References 1. Aalto-SetAla K, Gylling H, Miettinen T, Kontula K. Identification of a deletion in the LDL receptor gene. A Finnish type of mutation. FEBS Lett 1988;230:31-4. 2. Aalto-SetAlA K, Helve E, Kovanen PT, Kontula K. Finnish type of low density lipoprotein gene mutation (FH-Helsinki) deletes exons encoding the carboxy-terminal part of the receptor and creates an internalization-defective phenotype. J Clin Invest 1989;84:499-505. 3. Saiki RK, Gelfand DH, Stoffel S, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNApolymerase. Science1988;239:487-91. 4. Winichagoon P, Kownkon J, Yenchitsomanus P, Thonglairoam V, Siritanaratkul N, Fucharoen S. Detectionof $-thalassemia and hemoglobin E genes in Thai by a DNA amplificationtechnique. Hum Genet 1989;82:389-90. 5. Chamberlain JS, Gibbs RA, Ranier JE, Nguyen PN, Caskey CT. Deletion screening of the Duchenne muscular dystrophy locus via multiplex DNA amplification. Nucleic Acids Res 1988;16:11141-56. 6. Kogan SC, Doherty M, Gitschier J. An improved method for prenatal diagnosis of genetic diseases by analysis of amplified DNA sequences. Application to hemophilia A. N Engl J Med 1987;317:985-90. 7. Boehm CD. Use of polymerase chain reaction for diagnosis of inherited disorders. Clin Chem 1989;35:1843-.8. 8. Kunkel LM, Smith KD, Boyer SH, et al. Analysis of human Y-chromosome-specific reiterated DNA in chromosome variants. Proc Natl Aced Sci USA 1977;74:1245-9. 9. Yamamoto T, Davis CG, Brown MS, et al. The human LDL receptor a cysteine-rich protein with multiple Alu sequences in its mRNA. Cell 1984;39:27-38. 10. Gregersen N, Winter V, Peterson KB, et al. Detection of point mutations in amplified single copy genes by biotin-labelled oligonucleotides: diagnosis of variants of alpha-1-antitrypsin. Clin Chim Acta 1989;182:151-64. 11. Janssen JWG, Steenvoorden ACM, Lyons J, et al. RAS gene mutations in acute and chronic myelocytic leukemias, chronic myeloprolilbrative disorders, and myelodysplastic syndromes. Proc Natl Aced Sci USA 1987;84:9228-32. 12. Weisgraber KH, Newhouse YM, Mahley RW. Apolipoprotein E genotyping using the polymerase chain reaction and allelespecific oligonucleotide probes. BiochemBiophysRes Commun

1988;157:1212-7. 13. Bar-Eli M, Ahuja H, Gonzalez-Cadavid N, Foti A, Cline MJ. Analysis of N-RAS exon-1mutations in myelodysplastic syndromes by polymerase chain reaction and directsequencing. Blood

1989;73:281-3. 14. Them SL, Hesketh C, Brown JM, Anstey AV, Weatherall DJ. Molecular characterization of a high A2p-thalassemia by direct sequencing of single-strand enriched amplified genomic DNA. Blood 1989;73:924-30. 15. SyvanenA-C,Aalto-SetfiIAK, Kontula K, SoderlundH. Direct sequencing of afilnity-captured amplified human DNA: application to the detection of apolipoprotein E polymorphisms. FEBS Lett 1989;258:71-4. 16. Kulozik AE, Lyons J, Kohne E, Bartram CR, Kleihauer E.

Rapid and non-radioactiveprenatal diagnosis of $-thalassemia and sickle cell disease: application of the polymerase chain reac(PCR).Br J Haematol 1988;70:455-8. 17. LeitersdorfE, Van Der Westhuyzen DR, Goetzee GA, Hobbs HH. Two common low density lipoprotein receptor gene mutations cause familial hypercholesterolemia in Africaners. J Clin Invest 1989;84:954-61. 18. Hobbs HH, Brown MS, Russell DW, Davignon J, Goldstein JL. Deletion in the gene for the low-density-lipoprotein receptor in a majority of French Canadians with familial hypercholesterolemia. N Engl J Med 1987;317:734-7. tion

CLIN.CHEM.36/6,903-905(1990)

Convenient Screening for Hemoglobin Variants by Using the Diamat HPLC System Thomas Delahunty Hemolysatesfrom 102 differentpatientspreviouslyassessed for the presence of hemoglobinvariantsby celluloseacetate electrophoresiswere reanalyzed with the Diamat, a microprocessor-controlled step-gradient HPLC technique designed to determine glycohemoglobin (Hgb A10). This pool contained 81 abnormal specimens, each with one of seven different variant abnormalities.In all cases the HPLC technique correctly detected the presence or absence of a variant. The common S trait gave a large peak immediately after Hgb A, and the SS and SC variants gave clearly abnormal patterns,caused in part by the absence of Hgb A. The pattern from EE variants was distinguished by the appearance of the glycatedfractionHgb E1, which is eluted at 4.7 instead of 4.2 mm, whereas the AE pattern had two glycated peaks. Hemoglobins F, “fast”, S, and C were discernedby the presenceof a major peak at 3, 5.2, 6.5, and 7.5 mm, respectively.The findingssuggestthat the Diamat is a convenienttool to detect and help diagnose variants in a screening pool. AddItIonal Keyphrases: chromatography, cation-exchange glycohemoglobin

been reported

by Bio-Rad personnel

(8).

MaterIals and Methods Hemolysates of washed erythrocytes (EDTA-treated specimens) were electrophoresed according to instructions by Helena Laboratories (Beaumont, TX 77704). This cellulose acetate method (Super Z 5442) is capable of detecting

the presenceof the Hgb variants S, F, C, E, or “fast,” but does not distinguish between C and E. To differentiate C and E, citrate agar electrophoresis (9) was used (Titan 1V, Helena; and Paragon Acid Hb, Beckman, Brea, CA 92621). All specimens were also subjected to HPLC in the Diamat system (Bio-Rad, Hercules, CA 94547). This instrument is designed to automatically detect the relative abundance of glycated Hgb (% of total Hgb) in the blood of diabetics, with a throughput of seven samples per hour. After a simple dilution, samples are loaded in a refrigerated 48-well holder and analysis is initiated, with no further technologistintervention required. Sample compo-

nents are separatedby step-gradient elution of phosphate

The clinical benefits of a rapid screening method for (Hgb) variants in infants are well recognized (e.g., 1). However, the popular electrophoretic techniques involving cellulose acetate or agarose (2) are relatively labor-intensive and often difficult to interpret. Isoelectric focusing is the method of choice for mass screening (3), but requires special training and equipment. Although solubility, open-column, and immunoassay methods directed at specific variants can also be used (4,5), there is a need for a convenient method that will detect the presence of any variant in a screening pool. Earlier workers have shown that Hgb variants can be resolved by HPLC, but the equipment they used is not readily available in clinical laboratories (6). I describe here how a commercially available technique for detecting glycated Hgb A10 (7) can be hemoglobin

Laboratory Services, Division Road, SanJose, CA 95133.

used to detect and help diagnose 11gbvariants in a clinical setting by using two different operating modes. The quantification of Hgb A2 by a modified Diamat system has also

of Metpath

Inc., 967 Mabury

ReceivedFebruary 13, 1989;acceptedApril 5, 1990.

buffers matched to a specific cation-exchangecolumn, and detectedby measuring the absorbanceofthe effluent at 415 and 690 nm. To achieve precise retention times (Re), one line-tunes the R1 of the Bio-Rad A10 standard by vernier adjustment. Data reduction and chromatogram production are controlled by a microprocessor. Although the analysis routinely produces a chromatogram with a 10% scale (standard mode), for variant confirmation the scale can be converted to 100% by a simple program that extends the analysis time and allows the major peaks to be clearly delineated. The R and relative areas of all peaks are always reported, regardless of the scale. Results

Standard Scale (10%) Figure IA depicts a typical HPLC chromatogram for glycated hemoglobin obtained when healthy patients’ blood (n = 21) was analyzed on the Diamat. Only minor peaks were eluted before the A1 peak (R = 4.2 mm); the major peak was that for nonglycated HgbA(A0, = 6.1 min). By CLINICALCHEMISTRY,Vol.36, No.6, 1990 903