Discrimination among the transcripts of the allelic human β-globin genes βA, βS and βC using oligodeoxynucleotide hybridization probes

Gene, 43 (1986) 23-28

23

Elsevier GENE

1579

Discrimination among the transcripts of the allelic human /3-globin genes /J*, fi” and PC using oligodeoxynucleotide hybridization probes (Recombinant DNA; point mutations; human genetic disease; Northern blotting)

G. Nozari a, S. Rahbar a and R. Bruce Wallace b* u Department of Hematology, City of Hope Medical Center, Duarte, CA 91010, Tel. (818)359-8111, and b Department of Molecular Genetics, Beckman Research Institute of the City of Hope, Duarte, CA 91010 (U.S.A.) Tel. (818)357-9711 (Received

December

(Accepted

February

27th, 1985) 2nd, 1986)

SUMMARY

Three nonadecadeoxynucleotides complementary to the sense strand of the normal human fl-globin gene, j?*, and to the two allelic genes fis and /F were synthesized. The ps and p globin genes both differ from the PA gene by a single nucleotide substitution in the sequence coding for codon 6. The oligodeoxynucleotides are complementary to the genes in the region of the mutations and are therefore allele-specific. When radiolabeled and used as hybridization probes, the oligodeoxynucleotides are found to hybridize specifically to the mRNA transcribed from each allele.

INTRODUCTION

The ability to discriminate between two genes that differ by only a single bp has led to important applications, particularly in the diagnosis of certain human genetic diseases as well as to the screening for mutations created by site-directed mutagenesis

* To whom

correspondence

and

reprint

requests

should

be

addressed. Abbreviations:

aa, amino acid(s); AA, homozygous

gene; AC, heterozygous zygous

for /IA and /?s globin

dimethylsulfoxide;

Na, .citrate, phosphate,

genes;

SDS,

sodium

for fls globin gene; pH

bp, base pair(s);

DTT, dithiothreitol;

oligodeoxyribonucleotide; homozygous

7-8;

for /IA globin

for jI” and /Ic globin genes; AS, hetero-

SSPE,

1 mM EDTA,

DMSO,

nt, nucleotide(s); dodecyl

sulfate;

oligo, SS,

SSC, 0.15 M NaCI, 0.015 M

0.18 M NaCl,

10 mM sodium

pH 7.0.

0378-l 119/X6/$03.50 0 1986 Elsevier

Science Publishers

B.V. (Biomedical

(Itakura et al., 1984). The most general approach to this discrimination has been through the use of specific synthetic oligo hybridization probes. Oligo : DNA duplexes containing single bp mismatches are significantly less stable than those which are perfectly base-paired (Gillam et al., 1975; Wallace et al., 1979; 1981; Dodgson and Wells, 1977). Under appropriate conditions oligo probes will only hybridize to their cognate sequence and not to a sequence containing one of more non-complementary nt (Wallace et al., 1979; 1981; Conner et al., 1983). Thus, oligo hybridization can be used to determine the sequence of a short region of a DNA molecule. Until recently it has not been possible to discriminate between RNA molecules which differ by only a single nt (see Winter et al., 1985). Although oligo probes have been used to discriminate between highly related RNAs (see Miyada et al., 1984 for Division)

24

example), never have single nt differences been probed. It has been shown that G: T mismatches in oligo: DNA duplexes are less destabilizing than other mismatches (Gillam et al., 1975; Agarwal et al., 1981; Kidd et al., 1983). Similarly, G:U mismatches in RNA : RNA duplexes are more stable than other mismatches (Uhlenbeck et al., 1971). To optimally discriminate two genes which differ by a single transition mutation, oligo probes can be synthesized such that they form an A : C mismatch with the DNA of the non-complementary allele. Due to the fact that RNA is single-stranded, it is not always possible to avoid an oligo probe forming a G : T or G : U mismatch with a non-complementary RNA when hybridizing oligo probes to two RNA molecules which differ by a single nt. To discriminate the two RNAs on the basis of oligo hybridization one must attempt to optimize the destabilizing effect of the mismatch formed. In this paper we provide evidence that oligodeoxyribonucleotide probes can be used to distinguish between two RNA sequences that differ by a single nt.

(b) Gel electrophoresis and blotting RNA was denatured by dissolving in 6 M glyoxal and 50 y0 dimethyl sulfoxide and heating at 50’ C for 60 min (Maniatis et al., 1982). The denatured RNA was subjected to electrophoresis in a 1.5% agarose gel using 0.01 M sodium phosphate pH 7 buffer. Electrophoresis was at 3.5 V/cm for 3 h. RNA was transferred to GeneScreen (New England Nuclear) by capillary blotting using 20 x SSC buffer. The blots were dried at room temperature and baked in vacua at 80°C for 2 h. (c) Probe labeling The three 19-m oligos HP19A’, H/?19S’ and HB19C’ (see Table I) were synthesized on a Systec Microsyn 1450 automated DNA synthesizer. The oligos were radiolabeled at their 5’-ends using ( F~~P]ATP and T4 polynucleotide kinase and purified by electrophoresis on a 19.35% acrylamide, 0.65% bisacrylamide, 7 M urea gel as described recently (Studencki and Wallace, 1984; Miyada et al., 1984). (d) Hybridization

MATERIALS

AND METHODS

(a) RNA preparation RNA was prepared from 20 ml of heparinized blood obtained from various individuals of differing B-globin genotypes as described (Kan and Goosens, 1981). Red cells were lysed by resuspending the washed cell pellet in 0.144 M NH&l, 3 mM DTT followed by the addition of 0.1 vol. of 0.01 M NH,HC03. To the lysate is added 0.1 vol. of 1.5 M sucrose, 0.5 M KC1 and 0.5% SDS and the solution was centrifuged at 3000 x g for 20 min at 4°C. The ribonucleoprotein was recovered from the supernatant by precipitation at pH 5 using 10% acetic acid followed by centrifugation at 3000 x g for 20 min at 4°C. The pellet was dissolved in 0.1 M Tris . HCl pH 9, 0.1 M NaCl, 1 mM EDTA and 0.1% SDS and extracted first with phenol-chloroform and then with chloroform-isoamyl alcohol followed by ethanol precipitation.

The blots were prehybridized in 10 x Denhardt’s solution (0.2% bovine serum albumin, 0.2% polyvinylpyrrolidone and 0.2% Ficoll) and 0.1% SDS for 1 h at 60°C. The blots were then washed in 2 x SSC and hybridized with the labeled oligo (1 x lo6 cpm/ml) in 5 x Denhardt’s, 5 x SSPE and 0.1% SDS for 3 h at 60°C (58°C in the case of H/319C’). For the hybridizations where unlabeled competitor oligo is included, it is added in a lo-fold molar excess over the labeled one. After hybridization the filters were washed three times with 6 x SSC at room temperature for 15 min followed by one wash with 6 x SSC at 57°C for one min. The filter is then exposed to Kodak XAR X-ray tihn with two DuPont Quantum III intensifier screens at -70°C for 0.5-3 h.

25

RESULTS

AND DISCUSSION

(a) Oligo probe design The sequences of the oligos used in this study are given in Table I. The position and length of the sequences is based on several considerations: (1) The length of 19 nt has been previously shown to give a probe that recognizes a unique sequence in the human genome (Conner et al., 1983). (2) The mismatches are centrally located to optimize thermal destabilization. (3) All sequences are anti-sense and are thus complementary to the mRNA. (4) Each oligo is complementary to one allele of the B-globin gene and forms either one or two mismatches with the other alleles. Hj19A’ is specific for the PA allele, Hj19S for the bs allele and H/?19C’ for the 8” allele as described previously (Studencki et al., 1985). For simplicity of discussion, an oligo which forms a duplex with no mismatches will be called c-oligo (complementary oligo) and one capable of forming duplexes with one or more mismatches will be termed non-c-oligo. (b) Effect of G:T mismatch

In preliminary experiments it was possible to discriminate the /I* transcript from the /I” transcript using the two probes Hj19A’ and Hb19S’ and the hybridization conditions described previously (Conner et al., 1983) for discriminating the two genes (not shown). Similarily, it was possible to discriminate /?” mRNA from PA mRNA using the Hp19A’ probe which forms an A: C mismatch with the p transcript. It was, however, much more difficult to discriminate /I* mRNA from /?” mRNA using the HP19C’ probe which forms a G:T mismatch with the fl* transcript (Fig. 1A). Wallace et al. (1981) have previously shown that hybridization of a c-oligo to its target sequence is unaffected by the presence of a molar excess of unlabeled non-c-oligo. Therefore, to decrease the hybridization of non-c-oligos forming a G:T mismatch to the noncomplementary target sequence, hybridizations were attempted using labeled c-oligo in the presence of a lo-fold molar excess of unlabeled non-c-oligo. The presence of the competitor oligo should effectively suppress any hybridization of the

26

with [32P]Hfi19C’ in the absence (Fig. 1A) or presence (Fig. 1B) of a IO-fold molar excess of unlabeled H,819A’. In both cases the HJ319C’ probe hybridizes strongly with the ~-~ob~ mRNA present in the AC RNA. In the absence of the competitor however, there was a low level of binding of the Hj09C’ probe to the normal @-globin mRNA in the AA sample. This residual hybridization could only be reduced by long high-criteria washes (65°C in 6 x SSC) which also resulted in the loss of signal from the AC lane. In the presence of the non-c-oligo competitor (Fig. IB), there is essentially no binding of the Hgl9C’ oligo to the PA-globin mRNA. (c) Specific hybridization of aIIeie~peci~cprobes to /I*, /Is and PC mRNA

Fig. 1. Effect of competitor oligo on the hybridization specificity of ohgo probes to &lobin mRNA. RNA was prepared from blood cells of individuals which are either homozygous for the normal /I-globin gene (AA) or heterozygous for the normal and p-globin genes (AC). The RNA was denatured, subjected to electrophoresis on an agarose gel and transferred to a GeneScreen filter. The filter was hybridized with HB19C’[32P]probe in the absence (A) or presence (B) of a IO-fold molar excess of unlabeled Hfl19A’. The hybridizing band has the same mobility as commercially available rabbit globin mRNA (Bethesda Research Laboratories).

labeled oligo to its noncomplement~ target sequence. Fig. 1 shows the effect of having the competitor oligo present during hybridization. mRNA was isolated from blood cells of individuals which were either homozygous for the normal p-globin gene (AA) or heterozygous for the normal and 8” allele (AC). The RNA was glyoxylated in the presence of DMSO, subjected to electrophoresis on an agarose gel and transferred to GeneScreen. Duplicate Northern blots containing the two RNAs were hybridized

RNA was isolated from the blood cells of individuals with the following @globin genotypes: AA, AC, AS and SS. The total RNA was denatured, subjected to electrophoresis and blotted onto a GeneScreen filter as described for Fig. 1. The blot cont~n~g the four RNA samples was hybridized first with HB19A’ [32P]probe in the absence of unlabeled non-c-oligo (Fig. 2A). After the hyb~dized probe was removed, the filter was rehybridized with H/319S’ [32P]probe in the presence of unlabeled Hp19A’ oligo (Fig. 2B). The hybridized probe was once more removed and the filter was again rehybridized with Hp19C’ [32P]probe in the presence of unlabeled Hfll9A’ oligo (Fig. 2C). As can be seen, each probe only hybridized to RNA samples containing the homologous fl-globin mRNA and not to samples cont~ing non-homologous allelic transcripts. Thus, the allele-specific oligo probe can ~~bi~ously astonish among the tr~sc~pts of allelic genes both in the case where the genes differ by a single transversion mutation as for 8” vs. fls (compare the SS, AS and AA lanes in Fig. 2A and B) as well as in the case where the genes differ by a single tr~sition mutation as for 8” vs. p” (compare the AC and AA lanes in Fig. 2A and C). Under appropriate conditions oligos will only form duplexes with complementary DNA sequences when perfect base pairing is possible. All nonWatson/Crick bp have a destabilizing effect. Thus, oligo probes are capable of ~sc~inating between alleles which differ by as little as a single bp. This allelic hybridization specificity was first demon-

21

Obviously, discriminate

it is highly between

desirable

to be able to

RNAs which differ by as little

as a single nt. In this paper

we show that this is

indeed possible, even in the case where the probe forms a single G : T mismatch with the noncomplementary transcript. This is accomplished by having an excess of unlabeled non-c-oligo present during the hybridization

with 32P-labeled

oligo blocks hybridization quence

to the non-c

hybridization mentary

c-oligo.

The non-c-

of the labeled c-oligo sesequence

of the labeled

without

affecting

c-oligo to the comple-

sequence.

Use of the hybridization conditions described here should enable one to examine the transcription of

c

Fig. 2. Discrimination

among the transcripts

of the allelic globin

genes PA, /?” and /?c by oligo hybridization. from blood cells of individuals

with the genotypes

and AA. The RNA was denatured, as described

RNA was isolated

in Fig. 1. The filter was hybridized

follows: A. Hybridized

with HB19A’ [32P]probe

unlabeled

B. Hybridized

non-c-oligo

the presence

of unlabeled

C. Hybridized unlabeled removed

with

HB19A’ between

and blotted sequentially

as

in the absence of

with Hb19S’[32P]probe

Hb19A’

as the non-c-oligo

HB19C’[32P]probe as the non-c-oligo.

hybridizations

SS, AS, AC

electrophoresed

in

the

presence

Hybridized

by washing

in and

probe

of was

the blot in 1 x SSC

at 50°C for 15 min then in 0.1 x SSC at 50°C for 15 min and finally in 0.01 x SSC at 68” for 30 min. The filter was autoradiographed

overnight

to check

that

hybridized

probe

was com-

highly related genes present within the same cell, either alleles differing by one or more bp [e.g., the alleles of the H-ras (Reddy et al., 1982; Tabin et al., 1982) or N-rus genes (Bos et al., 1985)) or nonallelic, but highly homologous genes (e.g. MHC class I genes (Miyada et al., 1984)]. It should also be possible to quantitate such transcripts through the use of an appropriate internal control. The competition hybridization technique should also be useful in oligo hybridization to DNA sequences which differ by only a single nt. Recently, Winter et al. (1985) have described the detection of single nt differences between different ras oncogene transcripts. The method is based on the ability of pancreatic RNase to cleave RNA heteroduplexes at the bp mismatches. Not all mismatches are likely to be amenable to such analysis, especially G: U mismatches. Oligo hybridization may represents the only technique to permit the discrimination between any two RNAs differing by as little as a single nt.

pletely removed.

strated in experiments designed to determine carriers of the genetic disorder, sickle-cell anemia (Conner et al., 1983). 19-bp oligos were used to discriminate the wild-type /?-globin allele (PA) from the mutant p” allele, which is responsible for sickle-cell anemia in the homozygous state. Even though these alleles differ by a single bp, oligo hybridization unambiguously detected each gene when present in either homozygous or hemizygous states (Conner et al., 1983). Similar discrimination has been demonstrated for normal and mutant genes in several other genetic diseases (Kidd et al., 1984; Orkin et al., 1983; Pirastu et al., 1983; Studencki et al., 1985).

ACKNOWLEDGEMENTS

This work was supported by a grant 119.5145.0000 from the Sickle Cell Disease Research Foundation to the City of Hope National Medical Center. We are grateful to Dr. Ram Chillar for providing us with the blood samples.

28

REFERENCES

Orkin,

S.H., Markham,

A.F. and Kazazian,

tion of the common Agarwal,

K.L., Brunstedt,

for detection

and characterization

oligonucleotide

D., Marshall,

at codon

C.J., Verlaan-de

Nature

M.,

and Wallace,

R.B.: Detection

hybridization

in human

with

acute myeloid

K., Teplitz,

of sickle cell Bs-globin oligonucleotides.

R.L.

allele by

Proc.

Nat].

J.B. and Wells, R.D.: Synthesis

behavior

of oligomer-polymer

lengths

of mismatched

Biochemistry

melting

containing

defined

complexes

dA:dG

and

dG:dG

K. and

Nucl.

Res.

2 (1975)

K., Rossi, J.J. and Wallace,

synthetic

R.B.: Synthesis

ohgos. Annu. Rev. Biochem.

Kan, Y.Y. and Goosens, hemoglobin

M.: DNA analysis

disorders.

Methods

and use of

in the diagnosis

Enzymol.

76

of

(198 1)

M.S., Wallace,

Prenatal analysis

diagnosis

R.B., Itakura,

of a,-antitrypsin

of the mutation

K. and Woo, deficiency

by

site in the gene. New Engl.

V.J., Wallace,

Antitrypsin mutation Maniatis,

T., Fritsch,

Wallace,

by direct

S.L.C.:

analysis

a,-

of the

J.: Molecular

Cloning.

Laboratory,

Cold

Studencki,

Tabin,

R.B.:

Evidence

ofsubgene

B.J., Impraim,

genes

using

A.A., Mclaughlin-Taylor, that

polymorphism

complex sequenes.

bladder

M.: A of trans-

carcinoma

A.B. and Wallace,

CC., Teplitz, R.L. and

among the human PA, ps and

allele

Am. J. Hum. Genet.

specific

oligo

hybridization

37 (1985) 42-51.

R.B.: Allele specific hybridization

of very high specific

activity:

discrimi-

PA and /?s-globin genes. DNA 3 (1984)

C.J., Bradley,

E. in the

may be generated

Proc. Nat]. Acad. Sci.

S.M., Bargmann,

A.G., Scolnick,

Chang, E.H.: Mechanism

C.I., Weinberg,

R.A.,

E.M., Dhar, R., Lowy, D.R. and

of activation

of a human oncogene.

300 (1982) 143-149.

Uhlenbeck,

O.C., Martin,

F.H. and Doty, P.: Self-complemen-

tary oligoribonucleotides: Wallace,

effects ofhehx defects and guanylic

acid base pairs. J. Mol. Biol. 57 (197 1) 217-229.

R.B.,

Johnson,

M.J.,

Hirose,

T.,

Miyake,

Kawashima,

E.H. and Itakura,

K.: The use of synthetic

nucleotides

as hybridization

probes,

of mixed sequence

T., oligo-

II. Hybridization

of

to rabbit /l-globin DNA.

Nucl. Acids Res. 9 (1981) 879-894. Wallace,

R.B., Shaffer, J., Murphy,

nucleotides mismatch. Winter,

NY 1982. C., Reyes,

by the T24 human

R.B.: Discrimination

probes.

E. and Barbacid,

for the acquisition

300 (1982) 149-l 52.

A.B., Conner,

Wallace,

and Itakura,

304 (1983) 230-304.

Cold Spring Harbor

major histocompatibility

by the assortment

K. and Woo,

E.F. and Sambrook,

Manual.

C.G., Klofelt,

murine

detection

in the gene. Nature

Spring Harbor, and

R.B., Itakura,

deficiency

A Laboratory Miyada,

Nature

oligonucleotides

J. Med. 3 10 (1984) 639-642. Kidd,

Studencki,

R.K., Santos,

is responsible

properties

oncogene.

acid-cytidylic

Kidd, V.J., Golbus, direct

mutation

Nature

53 (1984) 323-356.

805-817. S.L.C.:

E.P., Reynolds,

point

Papageorge,

625-634. Itakura,

Detection

in DNA. New Eng. J. Med.

7-15.

M.: The base-pairing Acids

B.J., Teplitz, R.L. and

of)?-thalassemia.

mutation

nation of the human

Smith,

of cellulose-pdT9.

diagnosis

309 (1983) 284-287.

using oligo probes

nucleotides.

16 (1977) 2367-2373.

S., Waterman,

specificity

and thermal

gene with

J. Clin. Invest. 71 (1983) 775-779.

R.B.: Prenatal

of a single nucleotide

H.H.: Direct detec-

/I-thalassemia

M., Kan, Y.W., Cao, A., Conner,

Wallace,

p-globin

Sci. USA 80 (1983) 278-282.

Dodgson,

Pirastu,

DNA probes.

forming

C., Itakura,

synthetic

synthetic

Reddy,

substitutions

315 (1985) 726-730.

B.J., Reyes, A.A., Morin,

Gillam,

Vries,

A.C.M.: Amino-acid

13 of the N-ras oncogene

leukaemia.

Acad.

using an

C&H., Van der Eb, A.J., Van Boom, J.H., Janssen,

J.W.G. and Steenvoorden,

Conner,

of an mRNA

probe. J. Biol. Chem. 256 (1981) 1023-1028.

Bos, J.L., Toksoz, Veeneman,

J. and Noyes, B.E.: A general method

mediterranean

K.: Hybridization to $X174

DNA:

R.F., Bonner, of synthetic

J., Hirose,

T.

oligodeoxyribo-

the effect of single base pair

Nucl. Acids Res. 6 (1979) 3543-3557.

E., Yamamoto,

method

to detect

scribed

genes:

F., Almoguera, and characterize

amplification

and

C. and Perucho, point mutations overexpression

M.: A in tranof the

mutant c-Ki-rus allele in human tumor cells. Proc. Nat]. Acad. Sci. USA 82 (1985) 7575-7579.

USA 82 (1984) 2890-2894. Communicated

by A.D. Riggs.