Laser desorption mass spectrometry for point mutation detection

Genetic Analysis: Biomolecular Engineering

ELSEVIER

NENETIC ALYSIS BlomolecularEngineering

13 (1996) 87-94

Laser desorption mass spectrometry for point mutation detection ] N.I. Taranenko a, K.J. Matteson b, C.N. Chung a, Y.F. Zhu a, L.Y. Chang c, S.L. Allman a, L. Haff ~, S.A. Martin ~, C.H. Chen a'* ~Health Sciences Research Division, Photophysics Group, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6378, USA bUniversity of Tennessee Medical Center, Knoxville, TN, USA ~Chinese Academia Siniea Tiapei, Taiwan dPerSeptive Biosystems, hzc., Framingham, MA, USA Received 16 April 1996; revised 10 July 1996; accepted 14 July 1996

Abstract

A point mutation can be associated with the pathogenesis of inherited or acquired diseases. Laser desorption mass spectrometry coupled with allele specific pclymerase chain reaction (PCR) was first used for point mutation detection. G551D is one of several mutations of the cystic fibrosis transmembrane conductance regulator (CFTR) gene present in 1-3% of the mutant CFTR alleles in most European population:s. In this work, two different approaches were pursued to detect G551D point mutation in the cystic fibrosis gene. The strategy is 1:o amplify the desired region of DNA template by PCR using two primers that overlap one base at the site of the point mutation and which vary in size. If the two primers based on the normal sequence match the target DNA sequence, a normal PCR product will be produced. However, if the alternately sized primers that match the mutant sequence recognize the target DNA, an abnormal PCR product will be produced. Thus, the mass spectrometer can be used to identify patients that are homozygous normal, heterozygous for a mutation or homozygous abnormal at a mutation site. Another approach to identify similar mutations is the use of sequence specific restriction enzymes which respond to changes in the DNA sequence. Mass spectrometry is used to detect the length of the restriction fragments generated by digestion of a PCR generated target fragment.

I. Introduction

Since the successful demonstration of matrix-assisted laser desorption/ionization ( M A L D I ) for detecting large proteins by Hillenkamp and his co-workers [1,2], laser desorption mass spectrometry has been broadly used for large biomolecule detection. However, the detection of oligonucleotides has been limited to small D N A fragments due to the weak glycosidic bonds and

* Corresponding author. Tel: + 1-423-5745894; Fax: + 1-4235762115. ' The sumbitted manuscript has been authorized by a contractor of the US Government under contract No. DE-AC05-96OR22464. Accordingly, the US Government retains a non-exclusive, royalty-free license to publish or reproduce the published form of this contribution, or allow others to do so, for US Government purposes.

P l l S1050-3862(95)00154-4

lack of adequate matrices. Recently, Wu et al. [3] discovered that 3-hydroxypicolinic acid (3-HPA) is a useful matrix for M A L D I of oligonucleotides and succeeded in detecting oliognucleotides of single-stranded (ss) D N A of 67 bases in size. Tang et al. [4] also have used 3-HPA as a matrix and detected oligonucleotides of 150 bases in size by increasing the desorbed ion energy up to 45 keV. Chen and his co-workers [5,6] found picolinic acid and 3-aminopicolinic acid as new effective matrices for oligonucleotides as well, and succeeded in detecting 190 and 246 base pairs doublestranded (ds) D N A fragments. They also found that only single-stranded ions can be detected when a dsD N A fragment was used as a target [7]. With mixed matrixes, D N A fragments up to 500 base pairs (bp) were successfully detected by laser desorption mass spectrometry [8]. Liu et al. [9] have used nitrocellulose

N.I. Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87-94

88

as a substrate and detected PCR products as large as 426 bp with MALDI. Although somewhat larger D N A fragments can be successfully detected by this method, the resolution for large oligonucleotides by mass spectrometry is usually poor. Recently, Vestal et al. [10] developed a new delayed ion extraction method and obtained mass resolution (M/AM) of near 1000 for 25 mer oligonucleotides. Reilly and his co-workers [11] used a similar method and obtained high resolution for proteins and small D N A fragments. Brown and Lennon [12], as well as Whittal and Li [13], also reported success in the use of delayed pulse ion extraction [14,15] to significantly improve the resolution of M A L D I of proteins. It had been hoped that M A L D I would be an excellent tool for the high volume D N A sequencing required by the Human Genome Project. While that hope may yet be realized, the M A L D I technique has not been able to increase the resolution of larger D N A fragments to meet the needs of sequencing. Pieles [16] has described a special case of sequencing of small oligonucleotides using a combination of endonucleases and MALDI. Although laser desorption mass spectrometry still cannot be used for routine D N A sequencing, it can be applied to fast D N A analysis in disease diagnosis. With the invention of polymerase chain reaction (PCR), a selected region of genomic D N A can be amplified for analysis. Chen and his co-workers [17] recently demonstrated the detection of the most common mutation (A F508 deletion) for the cystic fibrosis gene by laser desorption mass spectrometry from 30 patient samples. Lubman and his co-workers [9] have

0.070

0.060 00-

0.050

o= "~0.040

38

~ 0.050 n.-

23

0.020

0.010

380.000

~o ....

60~oo

.

.

.

.

11doo ....

16()oo''

21(~o6''

281 100

Moss in Doltons

Fig. 1. Negative-ion mass spectrum of 38 bp P C R product (template was 60 bp with two primers overlapping at one base; their lengths were 23 and 16 mer). Total a m o u n t of D N A was about 2.5 pmol. The matrix was 1 /~1 of 0.3 M 3-HPA. Laser wavelength was 355 n m and laser fluezlce was 50 mJ/cm 2.

0.08

0.07 0.08 0.05 0

5, 0.04 0.03 0.02 0.01 25

0.00 -0.01 -0.02

1ODd

' ' 8o'od ' ' io6o6 ' ' i46o6 ' ' io~o6 ' ' 22~oo ' '26

O0

kloss in Doltons

Fig. 2. Negative-ion mass spectrum after PCR. No PCR product was produced. Template was 60 bp; the two primers of 23 and 16 mer were not complementary to the normal template. Total a m o u n t of D N A loaded was about 2.5 pmol. The matrix was 1 pl of 0.3 M 3-HPA. Laser wavelength was 355 n m and laser fluence was 67 mJ/cm 2.

demonstrated the feasibility of using M A L D I to detect other mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene as well. A point mutation in a single base pair can be associated with the pathogenesis of inherited or acquired diseases. The G551D mutation in the C F T R gene involves a G--*A mutation at nucleotide 1784, which results in a Gly--*Asp substitution at amino acid 551 [18]. G551D is one of the several mutations of the C F T R gene present in 1-3% of the mutant C F T R alleles in most European populations [19]. In this work, we present two different approaches to the detection of the G551D point mutation in the cystic fibrosis gene. The strategy is to amplify the desired region of D N A template by PCR using two primers of different sizes which overlap one single base at the site of a point mutation. If the two primers based on the normal sequence match the target D N A sequence a normal PCR product will be produced. However, if the alternate primers that match the mutant hybridize, a PCR product specific for mutant sequencing will be produced. Thus the mass spectrometer can be used to identify patients that are homozygous normal, heterozygous for a mutation or homozygous abnormal at a mutation site. Another approach to identify similar mutations is the use of sequence specific restriction enzymes which respond to changes in the D N A sequence such as that described previously by Lubman and coworkers [3]. Mass spectrometry is used to detect the length of D N A segments produced by the restriction enzyme digestion of a D N A segment from PCR.

N.I. Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87 94

89

Table 1 19,

a.)

N~NRNr

b.)

NfNRHT

c.)

NfNRM r

d.)

MfMRN r

e.)

MfMRH T

f.)

MfMRMr

g.)

NfNR + MfM R + N r

h.)

N~N R + MfM R + H T

i.)

N~NR + MfMR + Mr

19',

19''

46

37

28

I

base pair

2. Experimental 2.1. Genomic D N A isolation

Genomic DNA was isolated from anonymous patient samples or cultured cell line (GMl1274, Human Genetic Mutant Cell Repository, Camden, N J) using a rapid preparation method [20,21]. The first PCR reaction amplified sequences in the exon 11 region of

the CFTR gene and produced a DNA product of 114 bp from normal and heterozygous samples. A second mutation specific PCR was performed using the product of the first PCR reaction. The normal and mutant primer sets that overlap at the site of the mutation were reacted with this target using the standard conditions and an annealing temperature of 52°C.

Table 2 Sequences of two sets of oligonucleotide primers were designed to directly detect the G551D genotype without the use of restriction enzyme digestion Normal

5000

19 19'

Mutant

"Forward N f ( 1 9 b p ) 5'-GAATCACACTGAGTGGAGG-3' Reverse N R ( 1 9 b p ) 5'-CAAATTCTTGCTCGTTGAC-3'

Forward M r ( 1 9 b p ) 5'-GAATCACACTGAGTGGAGA-3' Reverse M R ( 2 8 b p ) 5'-CTTGGTAAACAAATTCTTGCTCGTTGAT3' 5'-GGAGAAGGTGGAATCACACTGAGTGGAGGTCAAC_ GAGCAAGAATTTGTTTAGCAAG-3'

Sequence of the C F T R D N A in the area of the 551 mutation. The sequence G G T is mutated to GAT, for G551D mutation.

3000

"~ r~

2000

1oo0

0 ,o'oo ' ,o;o

8ooo'

' ioooo

,5ooo

' 1,ooo

~6ooo

M/Z Fig. 3. Negative-ion mass spectrum of P C R product of 37 bp (normal template and two normal primers 19 mer each). The total D N A was 2 pmol. The matrix was trihydroxyacetophenone (THAP). Laser wavelength was 337 n m and laser fluence was 30 mJ/cm 2.

90

N.I. Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87 94 2.3. Restriction enzyme analysis

19o0 19

1700

Restriction endonuclease Hinc II (USB, Inc.) recognizes the sequence in double-stranded C F T R D N A at the position of the G551D mutation and cleaves it as follows:

1500

;:3 '~

19'

1300 1100 900

e~

~

700

50o

37

300 I O0 1O0 =

4000

,

t

,

i

6000

,

t

,

t

~

,

8000

t

,

t

10000

,

t

,

12000

h

i

14000

M/Z Fig. 4. Negative-ion mass spectrum of PCR product in 37 bp (heterozygous template and two normal primers 19 mer each). The total amount was about 2 pmol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 nm, and laser fluence was 45 mJ/cm2.

5'...GTY~RAC...3'

where:R = A or G and

CARTYTG...5'

Y = T or C

.The G551D mutation changes the 5' of most G to an A, thereby destroying the recognition site for the enzyme. Restriction enzyme digestion was performed with 25 ~1 of the first PCR product in a total volume of 30/~1 containing 8 units of Hinc II at 37°C overnight. 2.4. D N A purification and sample preparation for M S

2.2. Amplification o f target sequences Amplification of selected genomic D N A targets, or ssDNA targets was performed by P C R using a Perkin Elmer D N A Thermal Cycler (Norwalk, CT, USA) using generally applicable conditions and buffers. A synthetic ssDNA representing a portion of the p53 gene was used as a target for PCR using primers that overlapped at a potential mutation site. D N A from patients or cell lines that were heterozygous for the G551D mutation in the C F T R gene or homozygous normal was used as a template in a P C R reaction using conditions and primers that were reported previously [17]. All synthetic D N A including primers and s s D N A (used as templates) were purchased from Oligos Etc. without further purification.

2000 1800

19

1600 1400 "~

1200

""

1000

N

6~

800

2

400

For M A L D I MS analysis, D N A amplified by PCR needs to be purified to remove enzyme and buffer salts remaining in the solution. Although one of the most common purification methods is phenol/chloroform extraction, this procedure is laborious and time-consuming. Therefore, in this work it was substituted with an easier method, by a commercially available purification kit (QIA quick P C R purification kit) from Q I A G E N , Inc., Chatworth, CA, USA. The final product then was precipitated in alcohol, centrifuged, dried under vacuum, and dissolved in 6 10 pl of H20. The sample for MS was prepared by mixing 1 pl of aqueous analyte solution with 1 /~1 of matrix solution and then 1 /~1 of this mixture was spotted on a stainless steel plate and dried under forced nitrogen gas at ambient temperature. The dried sample was loaded into the M A L D I - T O F mass spectrometer immediately without further exposure to the atmosphere. Two different matrices were used in this work. One of them was a mixture of 0.3 M 3-hydroxypicolinic acid plus 0.5 M picolinic acid plus 0.3 M a m m o n i u m fluorate (molar ratio 9:1:1). Another matrix was a mixture of 0.2 M 2,3,4-trihydroxyacetophenone, 0.2 M 2,4,6-trihydroxyacetophenone, and 0.3 M a m m o n i u m citrate dibasic. The matrix materials were purchased from Aldrich and used without further purification. 2.5. Mass spectrometry measurements

200

4000

6000

8000

10000

12000

14000

16000

M/Z Fig. 5. Negative-ion mass spectrum of PCR product (no PCR product was produced; normal template and two mutant primers 19 and 28 mer in length). The total amount was about 2 pmol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 nm and laser fluence was 50 mJ/cm2.

Some of the experiments in this work were conducted with a home-made linear time-of-flight mass spectrometer. A N d : Y A G laser with third harmonic generation was used for laser desorption and ionization. The details of this instrument have been described previously [4-8]. Other experimental results are from a Voyager II instrument from PerSeptive Biosystems, Inc. A N2 laser

N.I. Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87 94

91

500

400

19 300

.,fi

J/

c~

200

¢

100

r.,q

28

0

i

-100

,

I

t

000

~

,

7000

I

9000

I

,

I

L

I 1000

,

L

,

~

I

13000

,

I

I

| 5000

17000

M/Z Fig. 6. Negative-ion mass spectrum of PCR product in 46 bp (heterozygous template and two m u t a n t primers 19 and 28 met in length). The total a m o u n t of D N A was about 2 p~aaol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 n m and laser fluence was 50 mJ/cm 2.

was used for laser ablation of samples. The sample plate can hold as many as 100 samples. The ion energy is typically fixed at 28125 eV. The laser energy was measured by an external power meter (GenTec). A central guiding wire was used to deflect small matrix ions to prevent the possible saturation of the detector. The vacuum of the time-oJF-flight chamber was typically at 1 × 10 - 7 Torr. The ions were detected by a microchannel plate. Electronic signals were then digitized by a digital scope (Tektronix 520), which was controlled by a PC computer.

3. Results and discussion

3. I. Detection by single base overlap of two primers The goal of this approach is to conduct PCR using two primers with a single base overlap. The expected PCR products should be equal to the sum of both primers minus one. On the other hand, little PCR product will be produced if the template cannot be annealed to the chosen primers. Our first attempt to use this method involved a synthetic ssDNA as a

5000

18o0 ~

,9,

:r

]

141111

i=0

1000

4o

.6

4000

19'

3000

• ,...I

b~

600

o

400

r~

28

o

I--4

20110

28

~

J 000

2OO

0 ,

4000

i

,

i

6000

,

i

,

i

8000

,rl

,

f

,

10000

i

,

r

12000

,

i

,

i

14000

,

I

,

I

16000

M/Z Fig. 7. Negative-ion mass spectrum of PCR product of 37 bp (normal template and four primers: two normal 19 mer each, and two m u t a n t 19 and 28 mer). The total a m o u n t of D N A was about 2 pmol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 nm, laser fluence was 50 mJ/cm 2.

0

q , 4000

i

,

i , 6000

37 i

,

i , 8000

~

,

i , 10000

i

, i , 12000

46 ~

,

~ , | 4000

i

,

i 16000

M/Z Fig. 8. Negative-ion mass spectrum of PCR product in 37 bp and 46 bp (heterozygous template and four primers: two normal primers, 19 mer each, and two m u t a n t primers, 19 and 28 mer). The total a m o u n t of D N A was 2 pmol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 nm, laser fluence was 50 m J / c m 2.

N.I. Taranenko et aL / Genetic Analysis: Biomolecular Engineering 13 (1996) 87-94

92

amplified DNA product should be 38 bp. The experimental result shown in Fig. 1 clearly indicates the production of a 38 bp DNA PCR product. The length of the template used was 60 mer ssDNA. For a mutant template, PCR products are not expected. The sequence of the mutant template and the primers used were

300

-~

114

200

ee~

Primer 1 (23 mer)

1oo

2

5'-CAGCTGTGGGTTGATTCCACACC-3' i

i

20000

25000

. . . . . . . . . .

•

30000

.........

L _

35000

~

p

40000

,

i

45000

M/Z Fig. 9. Negative-ion mass spectrum of C F T R gene on exon 11, 114 bp in length. The total a m o u n t of D N A was about 2.5 pmol. The matrix was 3-HPA/PA (9:1). Laser wavelength was 337 nm, laser fluence was 45 mJ/cm 2.

template. The chosen template was a simulation of a fragment of the p53 gene, which is of critical importance in several cancer syndromes. The sequence of the template and the primers used are Primer 1 (23 mer) 5'-CAGCTGTGGGTTGATTCCACACC-3' Template (60 bp): 5'-CAGCTGTGGGTTGATTCCACACCCCCGCCCGGCACCCGCGTCCGCGCCATGGCCATCTAC3'3'-GGGGCGGGCCGTGGGC-5' Primer 2 (16 mer) When the second primer was changed to one with the 3' end base not complementary to the normal template, no PCR products were observed by laser desorption mass spectrometry. Upstream and downstream primers are overlapped at the base which may be the site of a point mutation. If the template is normal, PCR will proceed. The size of

5O0 .~

=~

54 400

60

300 "~

200 I

O0 0

11000

i

i

16000

21000

~

i

,

26000

i

31000

,

i

36000

,

Template (60 bp):

50000

i

41000

M/Z Fig. 10. Negative-ion mass spectrum of D N A ( C F T R gene on exon 11) after digestion with Hinc II. The total a m o u n t was about 2.5 pmol. The laser wavelength was 337 nm, laser fluence was 50 mJ/cm 2.

5'-CAGCTGTGGGTTGATTCCACACACCCGCCCGGCACCCGCGTCCGCGCCATGGCCATCTAC-3' 3'-GGGGCGGGCCGTGGGC-5' Primer 2 (16 mer) Results are shown in Fig. 2. Encouraged by these results, we decided to try this approach to measure a point mutation in the CFTR gene. Table 1 illustrates the different experimental schemes for the point mutation detection. In Table 1, N f , N R , M r , M R , NT, HT, and MT represent normal forward primer, reverse normal primer, mutant forward primer, reverse mutant primer, normal template, heterozygous template, and mutant template, respectively. The two sets of oligonucleotide primers designed to select the amplified G551D mutation are shown in Table 2. The oligos were complementary to 19 bases on either side of the mutation site. One of the mutation specific primers had a non-complementary 5' tail to increase the size of the resultant product. Using the 114 bp target DNA fragment produced from normal or heterozygous genomic DNA, the secondary set of PCR primers complementary to the normal template should be able to differentiate a template with normal or mutant sequences. When only normal sequences are present, the normal primer pair should produce the expected 37 bp PCR product while the mutant primers should not produce any significant product under these conditions. Similarly, when template with both normal and mutant sequences are present, the expected 37 and 46 bp PCR products should be produced. A mass spectrum for PCR product with normal forward and normal backward primers is shown in Fig. 3. The two peaks for 19 mer are due to the difference in molecular weights of these two primers. A PCR product of 37 bp in size is also observed. Due to the poor resolution, it is difficult to distinguish amplification products from the possible dimer ions from primers. A DNA target specific product can be shown to be responsible for this signal as shown by the absence of a peak when PCR conditions are applied to a blank reaction containing all reagents except target

N.L Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87 94

DNA. In addition, the presence of detectable product was confirmed in polyacrylamide gel electrophoresis. Experimental results for PCR products with two normal primers and heterozygous template are shown in Fig. 4. PCR products of 137 bp in length are observed in the mass spectrum. However, the signal is somewhat smaller than the one observed in Fig. 3, since only the normal chromosome can be used to produce the PCR product. Primer ions and double-charged ions are also observed. Fig. 5 shows the mass spectrum of PCR reaction with two mutant primers and a normal template. No PCR products were observed. Fig. 6 shows the mass spectrum with two mutant primers and a heterozygous template. PCR products of 46 bp in size were observed. The lengths of the primers used are 19 mer and 28 mer, respectively. In order to do disease diagnosis, it is more convenient to have all four primers in ti~e sample for PCR reaction. Fig. 7 shows the result of mass spectrum after PCR reaction with all four primers and a normal template. PCR product with a signal at 37 bp was observed. A similar experiment was conducted using a heterozygous template, both 37 bp and 46 bp were observed (see Fig. 8). The above results suggest that this approach can be successf~ally used to do point mutation diagnosis for diseases. 3.2. Detection with restriction enzyme digestion at point mutation site

The length of PCR product generated from exon 11 of the CFTR gene in genomic DNA was 114 bp. A mass spectrum of this product is shown in Fig. 9, suggesting that laser desorption mass spectrometry can be used to measure PCR products without further purification. The 114 bp PCR product was digested with Hinc II according to the man~afacturer's instructions. The digested product was separated on a 10% polyacrylamide gel, and stained with ethidium bromide to visualize the DNA fragments. The normal allele is digested into two fragments of 60 and 54 bp, while the mutant allele does not have a Hinc II restriction site and remains 114 bp in length. A template DNA was produced as described above and analyzed before and after digestion with Hinc II by MALDI. The 114 bp fragment and the 60 and 54 bp fragments were detected in the digested heterozygote template DNA, while only the 60 and 54 bp products were detected in the normal samples. Experimental results are shown in Fig. 10 for the heterozygous patient. For a homozygous normal sample, signals corresponding to 60 and 54 bp were observed. In conclusion, we have demonstrated two different successful methods to use MALDI for CF patient sample detection. With the progress of mass spec-

93

trometry and its high potential for fast DNA analysis, MALDI has the potential to become an important clinical tool for molecular diagnosis in the near future.

Acknowledgements The authors would like to thank K.L. Lee and K. Tang for valuable discussions. Research sponsored by the Director's Fund at Oak Ridge National Laboratory and the Office of Health and Environmental Research, US Department of Energy under contract DE-AC05-96OR22464 with Lockheed Martin Energy Research Corporation. This research is also partially supported by the United States Department of Energy, Office of Energy Research, Laboratory Technology Applications Division with PerSeptive Biosystems, Inc. K. J. Matteson wishes to acknowledge the support from the Tennessee Department of Mental Health and Mental Retardation, the Tennessee Department of Health Environment, and the Tennessee Department of Human Services. Preparation of the manuscript by Darlene Holt is also acknowledged.

References [1] Karas M, Hillenkamp F. Anal Chem 1988; 60: 2299-2301. [2] Karas M, Backmann D, Bahr U, Hillenkamp F. Int J Mass Spectrom Ion Processes 1989; 92: 231. [3] Wu KJ, Steding A, Becker CH. Rapid Comm Mass Spectrom 1993; 7: 142-146. [4] Tang K, Allman SL, Chen CH. Rapid Comm Mass Spectrom 1993; 7: 943-948. [5] Tang K, Taranenko NI, Allman SL, Chen CH, Chang LY, Jacobson KB. Rapid Comm Mass Spectrom 1994; 8: 673-677. [6] Taranenko NI, Tang K, Allman SL, Chang LY, Chen CH. Rapid Comm Mass Spectrom 1994; 8: 1001-1006. [7] Tang K, Allman SL, Chen CH, Chang LY, Schell M. Rapid Comm Mass Spectrom 1994; 8: 183-186. [8] Tang K, Taranenko NI, Allman SL, Chang LY, Chen CH. Rapid Comm Mass Spectrom 1994; 8: 727-730. [9] Liu YH, Bai J, Zhu Y, Liang X, Siemieniak D, Venta P J, Lubman DM. Rapid Comm Mass Spectrom 1995; 9: 735-743. [10] Vestal ML, Juhasz P, Martin SA. Rapid Comm Mass Spectrom 1995; 9: 1044-1050. [11] Christian NP, Colby SM, Giver L, Houston CT, Arnold RJ, Ellington AD, Reilly JP. Rapid Comm Mass Spectrom 1995; 9: 1061-1065. [12] Brown RS, Lennon JJ. Anal Chem 1995; 67:1998 2003. [13] Whittal RM, Li L. Anal Chem 1995; 67: 1950-1954. [14] Wiley WC, McLaren IH. Rev Sci Instrum 1995; 26:1150 1157. [15] Browder JA, Miller RL, Thomas WA, Sanzone G. Int J Mass Spectrom Ion Phys 1981; 37:99 108. [16] Pieles U, Zurcher W, Schar M, Moser HE. Nucleic Acids Res 1993; 21:3191 3196. [17] Chang LY, Tang K, Schell M, Ringelberg C, Matteson KJ, Allman SL, Chen CH. Rapid Comm Mass Spectrom 1995; 9: 772-774. [18] Richards B, Skaletsky J, Shuber AP, Balfour R. Multiplex PCR amplification from the CFTR gene using DNA prepared from buccal brushes/swabs. Hum Mol Genet 1993; 2 : 1 5 9 163.

94

N.L Taranenko et al. / Genetic Analysis: Biomolecular Engineering 13 (1996) 87-94

[19] Cutting GR, Kasch LM, Rosenstein BJ, Zielenski J, Antonarakis SE, Tsui L-C, Kazazian HH. A cluster of cystic fibrosis mutation in the nucleotide binding fold of the cystic fibrosis conductance regulator protein. Nature 1990; 346: 366-369. [20] Ng ISL, Pace R, Richard MV, Kobyashi K, Kerem BS, Tsui

L-C, Beaudet AL. Methods for the analysis of multiple cystic fibrosis mutations. Hum Genet 1991; 87: 613-617. [21] Cystic Fibrosis Genetic Analysis Consortium, Population variation of common cystic fibrosis mutation. Hum Mutat 1994; 4: 167-177.