A Novel Method for Detecting Point Mutations or Polymorphisms and Its Application to Population Screening for Carriers of Phenylketonuria

A Novel Method for Detecting Point Mutations or Polymorphisms and Its Application to Population Screening for Carriers of Phenylketonuria STEVE S. SOMMER, M.D., Ph.D., JOSLYN D. CASSADY, B.A., Department of Biochemistry and Molecular Biology; JANET L. SOBELL, M.S., Department of Health Sciences Research; CYNTHIA D. K. BOTTEMA, Ph.D., Department of Biochemistry and Molecular Biology We describe a method termed PCR (polymerase chain reaction) amplification of spe­ cific alleles (PASA), a generally applicable technique for detection of point mutations or polymorphisms. The ease and technical simplicity of PASA will make genetic analyses more accessible to the general medical community. In addition, PASA shows promise for population screening because the technique is rapid, highly reproduc­ ible, inexpensive, nonisotopic, and amenable to automation. PASA is a modification of PCR that depends on the synthesis of a PCR oligonucleotide primer that precisely matches with one of the alleles but mismatches with the other. When the mismatch occurs near the 3' end of the PCR primer, amplification is inefficient. Therefore, preferential amplification of the perfectly matched allele is obtained. We demon­ strate the applicability of PASA by performing carrier detection in the family of a patient with phenylketonuria (PKU) and by screening a population of unrelated subjects for the presence of the two mutations most commonly associated with PKU. Multiple persons were screened simultaneously for the mutant alleles because a mu­ tation could be detected in the presence of at least a 40-fold excess of the normal allele. The two PKU mutations could be detected concurrently by using a mixture of only three PCR primers, an indication that simultaneous screening of multiple mutations can be done even if three or more mutations are closely clustered. In addition to the detection of mutations, PASA can be used to detect polymorphic alleles rapidly and to distinguish pseudogenes or repetitive sequences that differ by as little as one base.

The ability to screen populations for carriers of genetic disease accurately, inexpensively, and rapidly would provide the opportunity for widespread genetic counseling. Ultimately, the incidence of disease may be dramatically reduced, One of the few successful examples of protein-

based carrier screening is Tay-Sachs disease (GM2 gangliosidosis type B), which is caused by a deficiency in ß-hexosaminidase activity. Be­ cause noncarrier and carrier levels of enzymatic activity do not overlap, genetic status can be unequivocally assigned. 1 Screening for TaySachs disease has substantially reduced the incidence of this disease in Ashkenazi Jews. 2

T^r--—r~j , i ■ . , n—, n A ,,-f.oQ This study was supported in part by Grant CA 15083i5Fi.3fromtheNationallnstitutesofHealth,PublicHealth Unfortunately, measurements of the amount of Service. a metabolite or the activity of a protein seldom . . xxi-wooo π ^ x provide the discrimination needed for detection AJJ „ . Address reprint requests to Dr. Ö. S. Sommer, Department of Biochemistry and Molecular Biology, Mayo Clinic, Roch- ο ί carriers in the general population. In conMayoClinProc 1361trast, DNA-based methods are likely to be more ester, MN 55905.64:1361-1372,1989

1362 PCR AMPLIFICATION OF SPECIFIC ALLELES

broadly applicable for such population screen­ ing. Phenylketonuria (PKU) is one disease that is potentially amenable to DNA-based screening. Classic PKU maps to chromosome 12. It is an autosomal recessive disease that affects 1 in 10,000 newborn Caucasians of northern Euro­ pean descent. The disease is the result of a deficiency in hepatic phenylalanine hydroxylase activity (PAH), which causes a primary increase in serum phenylalanine concentration and sev­ eral secondary changes in the level of com­ pounds derived from aromatic amino acids.3 If PKU is untreated in infancy, severe mental retardation ensues. Although treatment with a low phenylalanine diet can prevent mental re­ tardation, the disease has not been rendered benign. Patients with PKU still encounter prob­ lems, including (1) failure to reach full intellec­ tual potential because of incomplete compliance with the stringent dietary therapy, 4 (2) a high frequency of birth defects in children of affected females,5 and (3) an increased incidence of be­ havioral problems. 46 Standard methods of detection of carriers of PKU are impractical for population screening.7 The most accurate method is to assay for enzy­ matic activity in liver tissue, but the hazards of liver biopsy preclude routine clinical use. Phe­ nylalanine challenge is the best alternative, but the test is cumbersome. Finally, measurement of the ratio of phenylalanine to tyrosine under "semifasting" conditions is the simplest of the tests, but the required precision of the measure­ ments, the expense, and the rate of false-positive results preclude screening. Consequently, vir­ tually all persons who discover that they are carriers do so only after an affected child has been born into the family. Subsequent to the cloning of PAH comple­ mentary DNA,8 DNA-based testing became an option. In the Danish population, 90% of the PKU alleles are associated with four polymor­ phic patterns known as haplotypes.9 The muta­ tions in haplotypes 2 and 3 represent 20% and 40% of the PKU alleles, respectively. In haplotype 2, a C (cytidine) is mutated to T (thymidine) in exon 12. This mutation results in a

Mayo Clin Proc, November 1989, Vol 64

protein product in which tryptophan (Trp) sub­ stitutes for the arginine (Arg) that is normally present at amino acid 408.10 The mutation in haplotype 3 is a G (guanosine) mutated to an A (adenosine), which disrupts the function of the intron 12 donor splice junction. 11 The mutant alleles associated with haplotypes 2 and 3 are also prevalent in the United States population.12 When the mutations in haplotypes 1 and 4 are defined, most PKU carriers of northern Euro­ pean descent (approximately 4 million persons in the United States alone) will be directly diagnosable by DNA methods. The current methods that can detect such point mutations include the following: (1) direct DNA sequencing,13 (2) denaturing gradient gel electrophoresis,14 (3) polymerase chain reaction (PCR; see glossary [Table 1]) followed by allelespecific oligonucleotide hybridization,15 (4) allele-specific DNA ligation,16 and (5) ribonuclease or chemical cleavage of mismatched heteroduplexes.1718 In their current form, however, these techniques are unlikely to find widespread application in population screening because they lack the requisite speed, technical ease, and cost-effectiveness. In an effort to provide a suitable means of screening a large number of persons, we have developed PCR amplification of specific alleles (PASA), a method that uses differential amplification for rapid and reliable distinction between alleles that differ at a single base pair. PASA is an adaptation of PCR,19 which is a method of amplifying a segment of DNA. PCR uses a pair of oligonucleotides (referred to as PCR primers) specific to the region of interest to achieve amplification by the multiple repetition of three steps: denaturation of the DNA, hy­ bridization of the oligonucleotide primers to the segment to be amplified, and DNA synthesis from the primers by a DNA polymerase. After 30 to 40 cycles of PCR, a 1 to 10 million-fold ampli­ fication can routinely be achieved. The applica­ tion of PCR to research and to medical practice has been facilitated by the automation of the most laborious aspects of the technique.20 The abundance of material produced by PCR has facilitated but not replaced the aforemen-

Mayo Clin Proc, November 1989, Vol 64

PCR AMPLIFICATION OF SPECIFIC ALLELES 1363

Table 1.—Glossary of T e r m s Term

Definition

5' and 3' ends

These numbers refer to the ribose carbons of adjacent nucleotides t h a t are connected by phosphodiester bonds during the polymerization of DNA or RNA. Nucleic acid synthesis in vivo is always in the 5' to 3' direction. Thus, the 5' end is the beginning of the nucleic acid, and any extension of an oligonucleotide by DNA polymerase occurs from the 3' end of the oligonucleotide. In duplex DNA, the two strands are oriented in the opposite direction (see Fig. 1)

Allele

One of two or more alternative forms of a DNA sequence. For example, most persons have two functional (normal) alleles of the phenylalanine hydroxylase gene, whereas carriers of phenylketonuria have one functional allele and one m u t a n t allele

Haplotype

The pattern of polymorphism on a particular chromosome. For example, if three restriction fragment-length polymorphisms (RFLPs) in the phenylalanine hydroxylase gene are examined and each RFLP has two alleles, a given chromosome will have one of 2 3 possible haplotypes

Oligonucleotide

A short segment of single-stranded DNA t h a t is typically less t h a n 50 nucleotides. Automation of the chemical synthesis of oligonucleotides has made any desired oligonucleotide sequence readily available for application in methods such as polymerase chain reaction

Polymerase chain reaction (PCR)

A method t h a t can routinely amplify a segment of DNA by more t h a n 1 million-fold. PCR consists of multiple repetitions of three steps (see text)

PCR primer

Oligonucleotides t h a t confer specificity to PCR by providing an initiation point for DNA synthesis. For each reaction, a pair of PCR primers are hybridized to the region of DNA to be amplified. The hybridized oligonucleotides allow (that is, prime) DNA poly­ merase to initiate DNA synthesis by extension from the 3' end of the oligonucleotide

RFLP

Abbreviation for restriction fragment-length polymorphism. A difference in the sizes of DNA fragments produced by the action of a restriction endonuclease because of a nucleotide change t h a t results in the gain or the loss of a restriction site

Transition

A sequence change t h a t occurs when one pyrimidine (cytidine [C] or thymidine [T]) is changed to the other, or when one purine (guanosine [G] or adenosine [A]) is changed to the other. In contrast, transversions change a purine to a pyrimidine or a pyrimidine to a purine. The phenylketonuria mutations discussed in this article are examples of transitions because one mutation substitutes an A for a G and the other mutation substitutes a T for a C

tioned detection methods. In contrast, PASA PKU or a family history of PKU. An important modifies the PCR so that the detection of poly­ advantage of DNA-based testing is that the morphisms or point mutations requires only the results are unaffected by factors such as age, sex, amplification reaction. The utility of PASA is and state of health. demonstrated by performing carrier detection in Polymerase Chain Reaction.—Genomic a family with PKU and by screening for PKU DNA (250 ng) was added to 25 μΐ of 50 mM carriers in a population without a family history potassium chloride, 10 mM Tris HC1 (pH 8.3), of PKU. 1.5 to 2.5 mM magnesium chloride, 0.01% (wt/ vol) gelatin, 200 mM of each deoxyribonucleotide, and 1 μΜ of each primer. After 10 minutes at 94°C, one-half unit ofTaq polymerase (PerkinMETHODS 21 Elmer Cetus) was added, and, typically, 35 cycles For population screening, DNA was extracted from peripheral blood samples of persons of of PCR were performed (denaturation: 1 minute northern European descent who did not have at 94°C; annealing: 2 minutes at 50°C; and

1364 PCR AMPLIFICATION OF SPECIFIC ALLELES

elongation: 3 minutes at 72°C) with the PerkinElmer Cetus automated thermal cycler (a modi­ fication of the protocol of Saiki and associates19). For each pair of PCR primers, the allele-specific PCR primer was designed to have an estimated melting temperature of 44°C under standard conditions (1 M sodium chloride).22 The primer that does not anneal to the polymorphic site was designed to have a melting temperature of 48°C. Oligonucleotide concentrations were initially 1 μΜ each, but subsequently it was observed that decreasing the total oligonucleotide concentra­ tion to the range of 0.05 to 0.25 μΜ increased the level of amplification. Thirty to 40 cycles of PCR were performed, and little observable difference in results was noted. Consequently, 35 cycles were used routinely. The PCR amplification products were electrophoresed through 2.5% agarose gel and visualized by staining with ethidium bromide.23 Direct Sequencing.—Sequencing was per­ formed by genomic amplification with tran­ script sequencing (GAWTS), as previously de­ scribed.24·25 In brief, the region to be sequenced is amplified by PCR with an oligonucleotide that has an appended T7 phage promoter sequence. An aliquot of the PCR reaction is transcribed in vitro by T7 RNA polymerase to produce an abun­ dance of single-stranded RNA. The RNA tran­ script is then used for Sanger dideoxy sequenc­ ing with reverse transcriptase. Oligonucleotides for Use With PASA and Direct Sequencing.—Figure 1 shows the five oligonucleotides that were synthesized comple­ mentary to the PAH gene. Oligonucleotides I and II are specific for Trp408 and the intron 12 splice junction mutations, respectively. When PCR was performed with oligonucleotides I and III, a 1.4-kb amplified segment was detected when the mutant allele was the input genomic DNA. Spurious amplification products of other sizes were also detected in these reactions. In contrast, when a normal (functional) allele was the input DNA, no specific amplification was detected. Spurious segments, however, were present, and these served as a positive control by showing that the amplification reaction had been technically successful.

Mayo Clin Proc, November 1989, Vol 64

When PCR was performed with oligonucleo­ tides II and III, a 1.3-kb segment was detected when the mutant allele was the input genomic DNA, but, it was not seen when the normal allele was the input DNA. The T7 phage pro­ moter sequence appended to oligonucleotide III is not necessary for PASA, but its presence allows oligonucleotide III to be used for the con­ firmation of the results of PASA by the direct se­ quencing technique of GAWTS. For GAWTS, oligonucleotides III and IV were used for the PCR; then oligonucleotide V was used as the primer for the reverse transcriptase-mediated sequencing of the RNA transcript. RESULTS Mismatches Near the 3 'End of a PCR Primer Dramatically Inhibit Amplification.—Todetermine whether single-base mismatches could reproducibly and dramatically interfere with amplification of DNA by PCR, we synthesized a series of oligonucleotides with mismatches to the target DNA located at different positions. For these experiments, two sequence variants of the factor DX gene were used. In one of the sequence variants, an A substituted for a T, and in the other, a T substituted for a C. The data indicated that a mismatch at the 3' nucleotide or a mismatch at the penultimate 3' nucleotide of a PCR primer resulted in inefficient amplifica­ tion. When the products of the PCR were electro­ phoresed on agarose gel, an amplified segment was seen when the input DNA contained the allele that matched the PCR primers, but no segment could be detected when the allele mis­ matched the primer (Fig. 2). The details of the analysis of the factor IK sequence variants will be published elsewhere. These results strongly suggested that muta­ tions causing PKU could be detected by PCR amplification of specific alleles. To perform PASA for the detection of the PAH mutations associated with haplotypes 2 and 3, we synthe­ sized oligonucleotides I, II, and III (see Fig. 1 and Methods). Oligonucleotide I precisely matched the mutant Trp408 allele (haplotype 2 mutation), but it mismatched the normal Arg408 allele at the penultimate 3' nucleotide. Oligonucleotide II

PCR AMPLIFICATION OF SPECIFIC ALLELES 1365

Mayo Clin Proc, November 1989, Vol 64

i 5' * 3' GCCACAATACCTTG UÖ ,408 T (arg^ - trp) IV v 5' 1422 1432 3' t 1455 ...tgtggttttggtcttagGAACTTTGCTGCCACAATACCTCGGCCCTTCTCA 70 bp.

--.

Intron 11

-Exon 12

X

1τ 5' * 3' GCTGATTCCATTAACAA a (defective splice donor) 1522 t GCTGATTCCATTAACAgtaagt ca. 1200 bp. . ,

•

Exon 12--X

■Intron 12--

1538 tttgtagGTGA...58 bp. X-

1601 1626 TCTGTCAGCTGTGAATCTGTTGATGGAGATCCACCTAT. 3' ACACTTAGACAACTACC III

AGAGGGATATCACTCAGCATAATCCATGG 5' T7 phage promoter sequence

Exon 13 Fig. 1. Location of oligonucleotides I through V in phenylalanine hydroxylase gene. Exonic sequence is in upper case letters; intronic sequence is in lower case letters. Each oligonucleotide sequence is underlined, and its 5' to 3' direction is indicated. Oligonucleotides I and II precisely match the two mutant phenylketonuria alleles and mismatch the normal allele at the penultimate and the 3' nucleotides, respectively (designated by *). Oligonucleo­ tide III is located in exon 13, and it is oriented opposite to oligonucleotides I and II. If the mutant allele that results in tryptophan at position 408 is present, polymerase chain reaction (PCR) with oligonucleotides I and III will produce a 1.4-kb segment that spans from base 1432 in exon 12 to base 1626 in exon 13 (plus the 29 base T7 phage promoter sequence). Most of the segment is intron 12, which is 1.2 kb. If the mutant allele that disrupts the intron 12 donor splice junction is present, PCR with oligonucleotides II and III will produce a 1.3-kb segment that spans from base 1522 to base 1626. For verification of the results of PCR amplification of specific alleles by direct sequencing, PCR is performed with oligonucleotides III and IV. T7 phage promoter sequence appended to oligonucleotide III allows an RNA transcript to be made and then sequenced with reverse transcriptase by using oligonucleotide V as a sequencing primer (see Methods). Two nucleotides (indicated by "..") overlap between the 3' end of oligonucleotide IV and the 5' end of oligonucleotide V. The numbering system is that of Kwok and associates.8

precisely matched the intron 12 splice junction defect (haplotype 3 mutation), but it mismatched the normal splice junction at the 3' nucleotide. When oligonucleotide III was used with either oligonucleotide I or II, PCR amplified the mu­

tant but not the normal alleles (see subsequent material). Carrier Testing.—PASA was used for car­ rier testing in a family in which one member (family member 6) was affected with PKU (Fig.

1366 PCR AMPLIFICATION OF SPECIFIC ALLELES

Mayo Clin Proc, November 1989, Vol 64

B Chromosomal DNA mismatches oligonucleotide b near 3' end

Chromosomal DNA is precisely complementary to both oligo­ nucleotides a and b

A..

W' C'

V I 35 cycles of PCR >1 million-fold amplification

I

A

w-

•*v

I 35 cycles of PCR Inefficient amplification

Gel electrophoresis

Fig. 2. Diagram of polymerase chain reaction (PCR) amplification of specific alleles (PASA). The two antiparallel strands of chromosomal DNA are indicated by W, the Watson strand, and C, the Crick strand. The 5' to 3' directions are indicated by the half-arrows. Strands have been melted apart by high temperature, and PCR oligonucleotide primers, a and b, have been annealed. In this diagram, elongation with DNA polymerase, the third step of PCR, is under way. This process is represented by dashed lines that originate from the 3' end of the oligonucleotides. If the oligonucleotides are precisely complementary to the chromosomal DNA (panel A), elongation initiates efficiently and results in four strands in the region where two were initially. If there is a mismatch in chromosomal DNA (X in panel B), elongation cannot be initiated efficiently for oligonucleotide b. After 35 cycles, a 1 to 10 million-fold amplification occurs in panel A, whereas much less amplification occurs in panel B. When an aliquot of the material is electrophoresed, an abundance of amplified sequence of appropriate size can be detected in panel A by staining with ethidium bromide. In contrast, no segment is detected in panel B. Thus, a mutation or polymorphism can be detected by using an appropriately designed oligonucleotide that promotes efficient elongation from a mutant chromosome and inefficient elongation from a normal chromosome.

3 A). DNA was extracted, and PCR was per­ formed with oligonucleotides I and III and with oligonucleotides II and III (Fig. 3 B). One of the nonfunctional genes in the affected family member was due to the Trp 408 mutation (lane 6 with oligonucleotides I and III), and the other was due to the splice junction defect (lane 6 with oligonucleotides II and III). The additional bands seen with oligonucleotides I and II represent spurious amplification products. This finding is useful because it serves as a convenient positive

control for determining that the reaction was successfully performed. The mother, brother, and one half-brother carry the Trp 408 mutation, whereas the father carries the splice junction mutation. The other half-brother is a noncarrier. These results were confirmed by direct sequencing with GAWTS. Population Screening.—For population screening, it is useful to determine whether the mutant alleles can be detected in the presence of DNA from multiple normal subjects. To that

Mayo Clin Proc, November 1989, Vol 64

PCR AMPLIFICATION OP SPECIFIC ALLELES 1367

Fig. 3. Phenylketonuria (PKU) carrier testing with polymerase chain reaction amplification of specific alleles (PASA). PASA was performed, as described in the text (see Methods), with oligonucleotides specific for the mutation at Trp408 (oligonucleotides I and III) or the mutation at the intron 12 splice junction (oligonucleotides II and III). A, Pedigree of family of proband with PKU (family member 6). B, PASA for two PKU mutations. The oligonucleotide concentrations were 0.25 μΜ when I and III were used and 0.05 μΜ when II and III were used. S = standards: 250 ng 0X174 Haelll restriction fragments; lanes 1 through 6 correspond to family members 1 through 6 in pedigree. Arrow indicates size of expected amplified segment.

end, 250 ng of DNA from a person with two functional PAH genes (that is, a noncarrier) was mixed with decreasing concentrations of DNA from family member 6, the PKU proband (Fig. 4). A signal was consistently detected from as little as 12.5 ng of DNA from the proband. Thus, the Trp408 mutation was detected in the presence of a 40-fold excess of the normal gene. Similar sensitivity was noted in detecting the splice junction mutation when oligonucleotides II and III were used (Fig. 4). In an effort to remain well within the sensitiv­ ity of detection, screening was performed on groups of four persons with 60 ng of DNA per person. DNA from 300 unrelated subjects (600 chromosomes containing the PAH gene) with no known family history of PKU was screened in 75 groups with oligonucleotides I and III. For each group, PCR was repeated in the presence of 60 ng of genomic DNA from the PKU proband, in order to verify that the mutation would have been detected had it been present (Fig. 5 A). No carriers of the Trp408 mutation were found. DNA from 50 subjects (100 chromosomes) was also screened with oligonucleotides II and III. One group of four contained a carrier of the intron 12 splice junction mutation (Fig. 5 B). The specific carrier, N8, was identified by subse­ quently performing PCR allele-specific amplifi­

cations on DNA from each subject (Fig. 5 C). Sequence analysis by GAWTS confirmed that N8 was a carrier (Fig. 6). Simultaneous Screening for Clustered Mutations.—Because population screening would be most efficient if both mutations could be detected simultaneously, both oligonucleo­ tides I and II were included with oligonucleotide III during PCR. By modifying the oligonucleo­ tide concentrations, the Trp 408 and the splice junction mutations could be amplified simulta­ neously with similar efficiencies (Fig. 7). This result occurred despite the fact that the se­ quence of one of the amplified segments was a subset of the other (Fig. 1). In a general sense, these data suggest that multiple clustered mutations in a gene should not pose a problem for analysis by PASA. DISCUSSION The technique of PCR amplification of specific alleles can discriminate against a particular allele because the 3' end of one of the PCR oligonucleotide primers mismatches with that allele. Consequently, this allele is not amplified efficiently, whereas the other allele is amplified well. PASA was used to perform carrier testing in a family in which the proband had one copy of each of the two common point mutations that

1368 PCR AMPLIFICATION OF SPECIFIC ALLELES

Mayo Clin Proc, November 1989, Vol 64

Fig. 4. Detection of phenylketonuria (PKU) mutations in the presence of an excess of normal alleles. Polymerase chain reaction amplification of specific alleles was performed, as described in the text (see Methods), with decreasing concentrations of DNA from patient with PKU (family member 6 in Fig. 3 A) in the presence of 250 ng of DNA from a normal person (E102F). Oligonucleotide concentrations were 0.25 μΜ for oligonucleotides I and III (A) and 0.10 μΜ for oligonucleotides II and III (B). In lanes 1 through 6, oligonucleotides I and III were used for screening, and in lanes 7 through 12, oligonucleotides II and III were used. S = standards: 250 ng 0X174 ifaelll restriction fragments. Lanes l a n d 7: family member 6 (250 ng). Lanes 2 and 8: E102F(250 ng). Lanes 3 and 9: family member 6/E102F (250 ng/ 250 ng). Lanes 4 and 10: family member 6/E102F (62 ng/250 ng). Lanes 5 and 11: family member 6/ E102F (25 ng/250 ng). Lanes 6 and 12: family member 6/E102F (12.5 ng/250 ng). Arrow indicates size of expected amplified segment.

cause PKU. PASA was also used to screen a population of northern European descent for these two mutations. The first mutation, the C -> T transition at amino acid 408, was not de­ tected among 600 screened chromosomes. This result is consistent with the reported population frequency of approximately 1 per 500 chromo­ somes.9 In screening 100 chromosomes for the second PKU mutation at the intron 12 splice junction, one carrier was identified. The carrier status was verified by direct sequencing. This muta­ tion has been reported to occur in approximately 1 of 200 chromosomes,11 a prevalence consistent with that in the current study. The PKU alleles described herein and the factor IX alleles used in our initial experiments with PASA represent a total of three transitions and one transversion. Few difficulties in the optimization of PASA were encountered. Spe­

cific amplification was obtained over a range of stringencies if the oligonucleotides had the allele-specific base near the 3' end. In occasional DNA samples, amplification did not occur under standard conditions. Apparently this result was due to the contamination of genomic DNA with ethylenediaminetetraacetic acid (EDTA),21 in­ asmuch as higher concentrations of Mg++ re­ solved these problems. PASA is a rapid general method for detecting mutations and polymorphisms (including re­ striction fragment-length polymorphisms). In addition, the simplicity, accuracy, and reproducibility of PASA help make it a promising method of population screening. Many persons can be screened concomitantly, and once the DNA is available, the cost of supplies and labor is relatively low (approximately $3.00/PASA reaction). Partial automation already exists, and complete automation is feasible by combin-

Mayo Clin Proc, November 1989, Vol 64

(A) I & III Spike

PCR AMPLIFICATION OF SPECIFIC ALLELES 1369

(B) II & III

+ + +

(Oll&lll

+ + +

Group 3

8123123 8123123

Sabcd

Fig. 5. Screening a population for phenylketonuria (PKU) carriers with polymerase chain reaction amplification of specific alleles (PASA), which was performed as described in the text (see Methods). A, Screening for the Trp408 mutation in exon 12. PASA was performed on groups of four persons (60 ng of DNA each) with oligonucleotides I and III at concentrations of 0.25 μΜ each (lanes indicated by "spike -"). As a positive control, PASA was repeated with each group after addition of 60 ng of DNA from patient with PKU, family member 6 (lanes indicated by "spike +"). S = standards: 250 ng 0X174 Haelll restriction fragments. Lane 1: group 1. Lane 2: group 2. Lane 3: group 3. B, Screening for the intron 12 splice junction mutation. PASA was performed on groups of four persons (60 ng of DNA each) with ("spike +") and without ("spike -") DNA from family member 6 (60 ng). Concentrations of oligonucleo­ tides II and III were 0.05 μΜ each. S = standards: 250 ng 0X174 Haelll restriction fragments. Lane 1: group 1. Lane 2: group 2. Lane 3: group 3. C, Identifying the PKU carrier in group 3. With use of oligonucleotides II and III at 0.05 μΜ each, DNA (250 ng) from each person in group 3 was screened by PASA. S = standards: 250 ng 0X174 Haelll restriction fragments. Lane a: N5. Laneb: N6. Lanec: N7. Lane d: N8. Arrow indicates size of expected amplified segment.

ing PASA with currently available robotic26·27 and nucleic acid detection28 technology. A bat­ tery of analytic strategies for detecting nucleic acid can be applied to eliminate the need for agarose gel electrophoresis. Two of many pos­ sible approaches are depicted in Figure 8 and described in the accompanying legend. PCR followed by hybridization with either isotopically or nonisotopically labeled allelespecific oligonucleotides15 is an alternative to PASA for detecting single base changes. For population screening, however, PASA has sev­ eral advantages: (1) the test is qualitative rather than quantitative, (2) 40 or more chromosomes

can be screened concurrently, (3) multiple dis­ tinct alleles can be analyzed concomitantly in one lane of a gel, and (4) automation may be easier to achieve because fewer steps are in­ volved. We have shown that spurious bands can serve as an internal control for the effectiveness of the PCR and that two PKU mutations can be de­ tected simultaneously. Other investigators have shown that six simultaneous PCR reactions can be performed.30 Thus, it is likely that, once defined, the six or so mutations that account for the overwhelming majority of the PKU alleles in the US Caucasian population potentially could

1370 PCR AMPLIFICATION OF SPECIFIC ALLELES

A.

#3

B. N8

A T G C

A T G C

G A gap

A

''•ffc^

^^^

WL·

T '<■» A ΛΒ A

Mayo Clin Proc, November 1989, Vol 64

*A/G A

· ^BP

1

| _

" "

Fig. 6. Sequencing (genomic amplification with transcript sequencing) of phenylketonuria intron 12 splice junction mutation, performed as described in the text (see Methods). The oligonucleotides used for the polymerase chain reaction amplification step were III and IV (see Fig. 1). The nested (internal) sequencing primer was V. A, Family member 3 (see Fig. 3), a noncarrier. B, N8 (see Fig. 5 C), a carrier. Asterisk indicates point of mutation in intron 12 splice junction. A = adenosine; C = cytidine; G = guanosine; T = thymidine.

be screened with one internally controlled PASA reaction. If 25 persons are screened per tube, half of the tubes, on the average, will contain DNA from a carrier. By consecutively subdivid­ ing the sample, 11 PASA reactions will suffice to identify one carrier. By also combining the tubes that are positive for different mutations, 12 reactions, on the average, can detect two or more carriers. Therefore, 22 million successful PASA reactions at approximately $3.00 per reaction could possibly be sufficient to screen the US population. Automation and economies of scale are likely to reduce the cost further. The cost of PASA would not be limiting because the collec­ tion of blood and extraction of DNA and the subsequent counseling of tested persons would be far more expensive. Once DNA has been collected, however, subsequent screening for other genetic diseases would entail only the incremental costs of PASA reactions and patient counseling. In addition to identifying persons at high risk for having a child with a genetic dis­ ease, it should eventually become possible to screen subjects for genetic susceptibility to cer­ tain environmental agents of disease and for

Fig. 7. Simultaneous detection of the two phenylketonuria (PKU) mutations with polymerase chain reaction amplifi­ cation of specific alleles, which was performed as described in the text (see Methods) with oligonucleotide III and oligonucleotide I or II (or both). Oligonucleotide concentra­ tions were 0.10 μΜ each. S = standards: 250 ng 0X174 Haelll restriction fragments. Lane 1: oligonucleotides I and II, DNA from E102F (noncarrier, see Fig. 4). Lane 2: oligonucleotide I, DNA from patient with PKU (carrier of both mutations, family member 6 in Fig. 3 A). Lane 3: oligonucleotide II, DNA from family member 6. Lane 4: ol­ igonucleotides I and II, DNA from family member 6. Lane 5: oligonucleotides I and II, DNA from carrier of the intron 12 splice junction mutation (family member 2 in Fig. 3 A). Lane 6: oligonucleotides I and II, DNA from carrier of the Trp408 mutation (family member 1 in Fig. 3 A).

genetic predisposition to such common diseases as cancer and diabetes. Such profound informa­ tion would considerably improve the ability of each person to assess the risks and benefits of life-style and career decisions. ADDED NOTE Recently, two other groups have reported that mutations in the globin and oCj-antitrypsin genes could be detected specifically by performing PCR with oligonucleotides that mismatch the normal sequence at their 3' end.31·32 ACKNOWLEDGMENT We thank John S. Kovach, M.D., for his support and encouragement, Robert O. Fisch, M.D., for providing blood samples, and Mary Ann C. Johnson for patiently typing the many revisions of this manuscript.

Mayo Clin Proc, November 1989, Vol 64

PCR AMPLIFICATION OF SPECIFIC ALLELES 1371

B Chromosomal DNA mismatches oligonucleotide b near 3' end

Chromosomal DNA is pre­ cisely complementary to both oiigonucleotides a and b; b is labeled(*) 35 cycles of PCR with excess *b "

^

—

Inefficient amplification

;

"b^

Specific amplification Incubate with immobilized oligo­ nucleotide c Wash

Detect presence or absence of label

Fig. 8. Detection of specific amplification by polymerase chain reaction amplification of specific alleles (PASA) without gel electrophoresis. As in Figure 1, the oiigonucleotides for PASA are a and b. Oligonucleotide b is labeled at its 5' end (*) so that it can be directly detected by color, fluorescence, or chemiluminescence or indirectly detected by binding complexes such as the multitude of systems that depend on the interaction of biotin and avidin.28 With an excess of oligonucleotide b relative to a, singlestranded DNA will be made in addition to double-stranded DNA.13 In the general case, amplification also will produce nonspecific as well as specific amplification (panel A). The allele that mismatches with oligonucleotide b will be inefficiently amplified, while amplification of any nonspecific region will still occur (panel B). The amplification products are incubated in microtiter dishes with an attached oligonucleotide (c), which is complementary to an internal region of the specifically amplified segment. After washing, only the specific segments will be retained by hybridization (paneLA). An even simpler system could be achieved if a homogeneous (nonseparation) system were devised by labeling oiigonu­ cleotides b or c (or both) so that hybridization would be required for signal emission (for example, see Kingsbury and Falkow29). In this manner, amplification and detection could be performed in one reaction tube.

REFERENCES 1. Ben-Yoseph Y, Reid JE, Shapiro B, Nadler HL: Diag­ nosis and carrier detection of Tay-Sachs disease: direct determination of hexosaminidase A using 4methylumbelliferyl derivatives of ß-Af-acetylglucosamine-6-sulfate and ß-iV-acetylgalactosamine-6sulfate. Am J Hum Genet 37:733-740, 1985 2. O'Brien JS: The gangliosidoses. In The Metabolic Basis of Inherited Disease. Fifth edition. Edited by JB Stanbury, JB Wyngaarden, DS Fredrickson, JL

Goldstein, MS Brown. New York, McGraw-Hill Book Company, 1983, pp 945-969 Blau K: Phenylalanine hydroxylase deficiency: bio­ chemical, physiological and clinical aspects of phenylketonuria and related phenylalaninemias. In Aro­ matic Amino Acid Hydroxylases and Mental Disease. Edited by MBH Youdim. New York, John Wiley & Sons, 1979, pp 79-139 Holtzman NA, Kronmal RA, van Doorninck W, Azen C, Koch R: Effect of age at loss of dietary control on

1372 PCR AMPLIFICATION OF SPECIFIC ALLELES

5. 6.

7.

8.

9.

10.

11.

12.

13.

14. 15.

16. 17.

intellectual performance and behavior of children with phenylketonuria. N Engl J Med 314:593-598, 1986 Scriver CR, Clow CL: Phenylketonuria and other phenylalanine hydroxylation mutants in man. Annu Rev Genet 14:179-202, 1980 Realmuto GM, Garfinkel BD, Tuchman M, Tsai MY, Chang P-N, Fisch RO, Shapiro S: Psychiatric diagno­ sis and behavioral characteristics of phenylketonuric children. J Nerv Ment Dis 174:536-540, 1986 Scriver CR, Kaufman S, Woo SLC: The hyperphenylalaninemias. In The Metabolic Basis of Inherited Disease. Sixth edition. Voll. Edited by CR Scriver, AL Beaudet, WS Sly, D Valle. New York, McGraw-Hill Information Services Company, 1989, pp 495-546 Kwok SCM, Ledley FD, DiLella AG, Robson KJH, Woo SLC: Nucleotide sequence of a full-length com­ plementary DNA clone and amino acid sequence of human phenylalanine hydroxylase. Biochemistry 24:556-561, 1985 Chakraborty R, Lidsky AS, Daiger SP, Güttier F, Sullivan S, Dilella AG, Woo SLC: Polymorphic DNA haplotypes at the human phenylalanine hydroxylase locus and their relationship with phenylketonuria. Hum Genet 76:40-46, 1987 DiLella AG, Marvit J, Brayton K, Woo SLC: An amino-acid substitution involved in phenylketonuria is in linkage disequilibrium with DNA haplotype 2. Nature 327:333-336, 1987 DiLella AG, Marvit J, Lidsky AS, Güttier F, Woo SLC: Tight linkage between a splicing mutation and a specific DNA haplotype in phenylketonuria. Nature 322:799-803, 1986 Moore SD, Huang WM, Koch R, Snyderman S, Woo SLC: Molecular population genetics of phenylketon­ uria in the United States and potential for carrier screening (abstract). Am J Hum Genet 43 (Suppl):A90, 1988 Gyllensten UB, Erlich HA: Generation of singlestranded DNA by the polymerase chain reaction and its application to direct sequencing of the HLA-DQA locus. Proc Natl Acad Sei U S A 85:7652-7656,1988 Myers RM, Lumelsky N, Lerman LS, Maniatis T: Detection of single base substitutions in total genomic DNA. Nature 313:495-498, 1985 DiLella AG, Huang W-M, Woo SLC: Screening for phenylketonuria mutations by DNA amplification with the polymerase chain reaction. Lancet 1:497499, 1988 Landegren U, Kaiser R, Sanders J, Hood L: A ligasemediated gene detection technique. Science 241:10771080, 1988 Myers RM, Larin Z, Maniatis T: Detection of single base substitutions by ribonuclease cleavage at mis­ matches in RNA:DNA duplexes. Science 230:12421246, 1985

Mayo Clin Proc, November 1989, Vol 64

18. Cotton RGH, Rodrigues NR, Campbell RD: Reactiv­ ity of cytosine and thymine in single-base-pair mis­ matches with hydroxylamine and osmium tetroxide and its application to the study of mutations. Proc Natl Acad Sei U S A 85:4397-4401, 1988 19. Saiki RK, Bugawan TL, Horn GT, Mullis KB, Erlich HA: Analysis of enzymatically amplified ß-globin and HLA-DQa DNA with allele-specific oligonucleotide probes. Nature 324:163-166, 1986 20. Saiki RK, Gelfand DH, Stoffel S, Scharf SJ, Higuchi R, Horn GT, Mullis KB, Erlich HA: Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-491, 1988 21. Gustafson S, Proper JA, Bowie EJW, Sommer SS: Parameters affecting the yield of DNA from human blood. Anal Biochem 165:294-299, 1987 22. Bonner TI, Brenner DJ, Neufeld BR, Britten RJ: Reduction in the rate of DNA reassociation by se­ quence divergence. J Mol Biol 81:123-135, 1973 23. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, New York, Cold Spring Harbor Laboratory, 1982 24. Stoflet ES, Koeberl DD, Sarkar G, Sommer SS: Genomic amplification with transcript sequencing. Science 239:491-494, 1988 25. Sarkar G, Sommer SS: Access to a messenger RNA sequence or its protein product is not limited by tissue or species specificity. Science 244:331-334, 1989 26. Wilson RK, Yuen AS, Clark SM, Spence C, Arakelian P, Hood LE: Automation of dideoxynucleotide DNA sequencing reactions using a robotic workstation. Biotechniques 6:776-787, 1988 27. KnobelochDW, Hildebrand CE, Moyzis RK, Longmire JL, Sirotkin KM, Beugelsdijk TJ: Robotics in the human genome project. Biotechnology 5:1284-1287, 1987 28. Matthews JA, Kricka LJ: Analytical strategies for the use of DNA probes. Anal Biochem 169:1-25,1988 29. Kingsbury DT, Falkow S (eds): Rapid Detection and Identification of Infectious Agents. Orlando, Florida, Academic Press, 1985, pp 245-256 30. Chamberlain JS, Gibbs RA, Ranier JE, Nguyen PN, Caskey CT: Deletion screening of the Duchenne muscular dystrophy locus via multiplex DNA ampli­ fication. Nucleic Acids Res 16:11141-11156,1988 31. Wu DY, Ugozzoli L, Pal BK, Wallace RB: Allelespecific enzymatic amplification of ß-globin genomic DNA for diagnosis of sickle cell anemia. Proc Natl Acad Sei U S A 86:2757-2760, 1989 32. Newton CR, Graham A, Heptinstall LE, Powell SJ, Summers C, Kalsheker N, Smith JC, Markham AF: Analysis of any point mutation in DNA: the amplifi­ cation refractory mutation system (ARMS). Nucleic Acids Res 17:2503-2516, 1989