DNA

Volume 3, Number 6, 1984

Mary Ann Liebert, Inc., Publishers

LABORATORY METHODS
Oligonucleotide-Directed Mutagenesis: A Simple Method
Using Two Oligonucleotide Primers and a Single-Stranded
DNA Template

MARK J. ZOLLER* and MICHAEL SMITHt

ABSTRACT
This paper presents a simple and efficient method for oligonucleotide-directed mutagenesis using vectors de-
rived from single-stranded phage. This modification of our previously published procedure (Zoller and
Smith, 1982) features the use of two primers, one of which is a standard M13 sequencing primer and the other
is the mutagenic oligonucleotide. Both primers are simultaneously annealed to single-stranded template
DNA, extended by DNA polymerase I (large fragment), and ligated together to form a mutant wild-type
gapped heteroduplex. Escherichia coli is transformed directly with this DNA; the isolation of covalently
closed circular DNA as in our previous report is not necessary. Mutants are identified by plaque lift hybridiza-
tion using the mutagenic oligonucleotide as a probe. As an example of the method, a heptadecanucleotide was
used to create a T —
G transversion in the MA Ta gene of Saccharomyces cerevisiae cloned into the vector
M13mp5. The efficiency of mutagenesis was approximately 50%. Production of the desired mutation was
verified by DNA sequencing. The same procedure has been used without modification to create insertions of
restriction sites as well as specific deletions of 500 bases.

INTRODUCTION 1982; Adams et al., 1983; Caruthers et al., 1983; Frank et

al., 1983; Matthes et al., 1984) along with an increasing ac-
of cloned DNA has become to DNA synthesizers have made gene synthesis a rea-
/Ndard toolmutagenesis
vitro

of exist
a stan-
in the functional analysis of nucleic acids and
which can broadly
be
cess
sonable method to produce specific mutations. The gene is
constructed using segments of oligonucleotides that are li-
proteins. A number approaches
grouped into random (Abarzua and Marians, 1984; Hef- gated together. By this procedure, an entire gene can be
fron et al., 1978; Matteucci and Heyneker, 1983; Shortle et synthesized that contains a desired change. In addition, a
al., 1981, 1982; Stone et al., 1984) and site-directed meth- simple construct can be produced that allows for the re-
ods (Wallace et al., 1981; Zoller and Smith, 1982, 1983; Lo placement of segments by a number of mutated 'oligonu-
et al., 1984). The repertoire of procedures to mutagenize a cleotide cartridges' positioned between restriction endonu-
fragment of DNA randomly continues to grow each year. clease sites (Matteucci and Heyneker, 1983; Lo et al.,
These random techniques are used to identify the location 1984).
and boundaries of a particular function. Once an important Oligonucleotide-directed mutagenesis is similar in princi-
region has been identified, site-directed mutagenesis can be ple to mutagenesis by gene synthesis in that an oligonucleo-
employed to determine the role of specific nucleotides (or tide that bears the desired mutant sequence is inserted into a
amino acids). The construction of cloned DNA bearing a cloned gene. However, oligonucleotide-directed mutagene-
specified alteration is best accomplished by either gene syn- sis differs from gene synthesis in the manner in which the
thesis or by oligonucleotide-directed mutagenesis. cloning is accomplished. In this case, an oligonucleotide
Both of these strategies have been used to create DNA consisting of the mutant sequence is hybridized to its com-
with a desired sequence. The availability of reliable proce- plementary sequence in a clone of wild-type DNA, thereby
dures for manual oligonucleotide synthesis (Gait et al., forming a mutant wild-type heteroduplex. The oligonucleo-

•Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 11724.

'Department of Biochemistry, Faculty of Medicine, University of British Columbia, Vancouver, V67 1W5, Canada.

479
480 ZOLLER AND SMITH

tide serves as a primer for in vitro enzymatic DNA synthesis Biolabs. [a-32P]dATP and [7-32P]rATP were purchased
of regions that are to remain genotypically wild type. A from Amersham. Unlabeled deoxyribonucleic acids, rATP,
double-stranded heteroduplex is formed, which is subse- and dideoxyribonucleic acids were obtained from P-L Bio-
quently segregated in vivo into separate mutant and wild- chemicals and supplied in solid form. Nitrocellulose filters
type clones. The two can be distinguished by a number of (BA85) were purchased from Schleicher and Schuell. All
screening procedures. other materials were of standard quality for molecular bio-
The basic strategy of oligonucleotide-directed mutagene- logical work. The heptadecanucleotide (5'-CCGCAACAG-
sis was developed using the single-stranded phage 0X174 GAAAATTT-3') used for mutagenesis was synthesized by a
(Hutchison et al., 1978). The technique has been success- manual phosphite-triester method similar to that of Adams
fully applied to genes cloned into either plasmid (Wallace et et al. (1983). The M13 sequencing primer used as the "sec-
al., 1981; Lewis et al., 1983; Oostra et al., 1983; Itakura et ond primer" (5-CCCAGTCACGACGTT-3') and the hexa-
al., 1984) or phage vectors (Wasylyk et al., 1980; Montell et decanucleotide (5'-TAAACGTATGAGATCT-3') used to
al., 1982; Winter et al., 1982; Zoller and Smith, 1982, 1983; sequence the mutation were also synthesized manually ac-
Norris et al., 1983). In an earlier report (Zoller and Smith, cording to Adams et al. (1983). Purification of oligonucleo-
1982), we described an efficient procedure for oligonucleo- tides following synthesis was carried out using a 20% poly-
tide-directed mutagenesis using M13-derived vectors. The acrylamide-7 M urea sequencing gel (40 x 20 x 0.05 cm) in
major features of this method were the use of single- TBE buffer. The desired product was eluted from the gel by
stranded cloned DNA as template, the purification of in a crush-soak method (Maxam and Gilbert, 1980) using 0.5
vitro synthesized closed covalently circular DNA by alka- M ammonium acetate. Isolation of the oligonucleotide was
line sucrose gradient centrifugation, and the use of the accomplished by either ethanol precipitation or by C-18
mutagenic oligonucleotide to screen for mutants by hybrid- Sep-pak chromatography (Lo et al., 1984).
ization. We report here a variation of our original method
that is simpler, faster, and of equal efficiency. It is based on Solutions
the use of two primers, one of which is the mutagenic oligo-
nucleotide and the other is a standard sequencing primer. The following buffers and solutions were used: 10x ki-
This method obviates the need to isolate closed covalently nase buffer: 1 MTris-HCl, 0.1 MMgCl2, 0.1 Mdithiothrei-
circular DNA in order to obtain mutants in high yield, and tol (DTT) pH 8.3. 10x Buffer A: 0.2 MTris-HCl, 0.1 M
retains the convenience of the M13 system for template iso- MgCl2, 0.5 MNaCl, 0.01 M DTT pH 7.5. 10 x Buffer B:
0.2 MTris-HCl, 0.1 MMgCl2, 0.1 M DTT pH 7.5. 2 itiM
lation, sequencing, and screening. Variations of two-primer
dNTPs: 2 mM dCTP, 2 mM dTTP, 2 mM dATP, 2 mM
mutagenesis are described in the accompanying paper by dGTP. 20 x SSC: 3 MNaCl, 0.3 M sodium citrate, 10 mM
Schold et al. and in a recent paper by Norris et al. (1983).
EDTA pH 7.2. 100 x Denhardt's: 2% bovine serium al-
bumin, 2% polyvinyl pyrolidone, 2% Ficoll. Prehybridiza-
tion solution I: 3 ml of 20 x SSC, 1 ml 100 x Denhardt's,
MATERIALS AND METHODS 0.2 ml 10% NaDodSC, 5.8 ml H20. Prehybridization so-
lution II: 3 ml of 20 x SSC, 1 ml 100 x Denhardt's, 6 ml
Strains H20. Probe solution: 10 ml prehybridization solution II,
E. coli JM101 (Alacpro, SupE, thil, F', proAE*, loci*, 2-5 x lO6 cpm oligonucleotide (1-2 ml). TBE: 50mMTris,
lacZ AM15 ¡Yc7A36) was used for all work with M13 and 50mMborate, 1 mMEDTApH 8.3. TE: lOmMTris-HCl,
was the generous gift of J. Messing (University of Minne- 1 mM EDTA pH 8.0.
sota). A streak of JM101 was maintained at 4°C for ap-
proximately 1 month on a plate of supplemented M9 + Bl. Media
Overnight cultures were grown in liquid M9 (supplemented) The following media were used: YT: 8 g tryptone, 5 g
+ Bl (Miller, 1972). yeast extract, 5 g NaCl per liter. YT plates: YT media and
15 g agar per liter. YT top agar: YT media and 8 g agar per
Vectors liter. 2YT: 16 g tryptone, 10 g yeast extract, 5 g NaCl per
liter. M9 supplemented + Bl: 1 x M9 (Miller, 1972) plus
For these studies, a 4.2-kb Hind III fragment containing
the MA 7a gene from S. cerevisiae was cloned into the vec-
0.2% glucose, 1 mMMgS04, 0.0001% Bl, 0.4% casamino
acid, 0.1 mMCaClj.
tor M13mp5 as described by Messing (1983). A detailed de-
scription of the construction of this clone is presented in
Zoller and Smith (1982). The same mutagenic procedures Transformation of E. coli JM101
have been utilized with other Messing vectors as well as the Competent cells are prepared on the day of use by
pEMBL vectors (Dente et al., 1983). diluting 0.5 ml overnight JM101 into 50 ml of YT media.
Cells are grown at 37°C to A60o 0.5, harvested, and re-
=

Materials suspended in 25 ml of ice-cold 0.05 M CaCl2. Cells are left

on ice 30 min, harvested, and resuspended in 5 ml of ice-
E. coli DNA polymerase (large fragment) and T4 DNA cold 0.05 M CaCl2. Glycerol can be added to 15% final
ligase were obtained from Bethesda Research laboratories. concentration and the cells can be stored at -70°C without
T4 polynucleotide kinase was obtained from New England significant loss of transformation efficiency.
OLIGONUCLEOTIDE-DIRECTED MUTAGENESIS 481

Next, 200 /il of CaCl2-treated cells are aliquoted into Fal- are pelleted and M13RF is isolated by alkaline lysis plasmid
con 2059 tubes and placed on ice. Appropriate amounts preparation method (Maniatis et al., 1982). The DNA is re-
(under 10 /d) of DNA are added to the cells which were kept suspended in 10 ml of TE (50:1), digested with 0.5 mg of
on ice for 30 min. The cells are placed at 42°C for 2 min, RNase for 15 min at 23CC, then subjected to cesium chlo-
after which 100 ¡A of overnight JM101 and 2.5 ml of YT top ride ethidium bromide isopycnic centrifugation using a
agar are added and the entire mixture poured onto a YT Beckman VTÍ50 rotor for 24 hr at 45,000 rpm and 18°C. A
plate. Freshly prepared competent cells can be plated with- single fluorescent band of DNA is usually obtained. Typi-
out the addition of the 100 ¡A of overnight culture. cal yields range from 100 to 200 /ig/500 ml preparation. The
success of this procedure depends on the proper multiplicity
of infection (at least 1010 phage/ml culture). The phage
Preparation of single-stranded-DNA stock can be prepared a day ahead, titered, then used the
(A) Miniprep: (for sequencing or dot blots). Single- next day to prepare RF.
stranded DNA (SSDNA) is prepared as described by Sänger
et al. (1980). A single plaque is cored with a 100-/J glass mi-
cropipet and added to 1.5 ml of 2YTto which 15 u\ of over- 5' phosphorylation of the mutagenic
night JM101 has also been added. Growth is carried out in oligonucleotide
a Falcon 2059 tube with vigorous shaking at 37°C for 5-6
hr. The cells are transferred to a 1.5-ml Eppendorf tube and
(A) For mutagenesis: The mutagenic oligonucleotide
are pelleted by a 5-min spin in a microcentrifuge. The
(200 pmol) is mixed with 2 ¡A of 10 x kinase buffer, 1 /d of
10 mMrATP, and 4 units of T4 polynucleotide kinase in a
supernatant is transferred to a fresh 1.5-ml Eppendorf total volume of 20 ¡A. The reaction is carried out in a 1.5-ml
tube. Then 50 ¡A of supernatant is removed and stored as a
phage stock at 4°C; 200 ¡A of 2.5 M NaCl/20% PEG 8000 Eppendorf tube incubated at 37°C for 1 hr and is termi-
nated by heating at 65°C for 10 min.
are added to the remainder of the phage suspension, the
tube is mixed, and incubated at room temperature for 15
(B) For hybridization or primer extension: The muta-
min. The phage are pelleted by centrifugation for 5 min at genic oligonucleotide (20 pmol) is phosphorylated as in (A)
with 50 /iCi [7-32P]ATP (3000 Ci/mmol) as the sole source
room temperature and the supernatant is removed by as-
of ATP.
piration. At this point, a small white pellet can be observed.
The phage are resuspended in 100 /A of TE (10:1), extracted
with 50/il of phenol, 50 ¡A of phenol/chloroform (1:1), and
Separation of labeled oligonucleotide from
finally 2 x 500 ¡A of ether or chloroform. Then 10 /A of 3 M unincorporated label
sodium acetate (pH 5.5) and 250 ¡A of ethanol are added
and the tube is placed at -70°C for 15 min. The DNA is Labeled oligonucleotide is removed from unincorporated
pelleted by a 5-min centrifugation at 4°C, washed with 0.5 ATP by DE-52, Sephadex G-50, or C-18 Sep-pak chroma-
ml of 70% ethanol, dried in vacuo, and resuspended in 50 tography as described below. For preparation of hybridiza-
¡A of TE (10:1). Typical yields range from 2-5 /ig. tion probes, the DE-52 method is used. For preparation of
(B) Maxiprep: (for template). A phage stock from a sin- a salt-free oligonucleotide solution for primer extension,
gle plaque grown in 1.5 ml of 2YT is prepared as described the C-18 Sep-pak method is used. The effectiveness of la-
in (A). The cells are spun out and the phage containing beling can be assessed by running an aliquot of the kinase
supernatant is poured into a fresh 1.5-ml tube. For each ml reaction on a 20% sequencing gel. Alternatively, an aliquot
of 2YT are added 10 /il overnight JM101 and 10 /il of phage can be chromatographed on Whatman DE-81 paper using
stock. For example, to prepare SSDNA from a 50-ml cul- 0.3 M ammonium formate (pH 8.0). Under these condi-
ture, 0.5 ml of overnight JM101 and 0.5 ml of phage stock tions, the oligonucleotide remains at the origin and ATP
are added to 50 ml of 2YT, incubated at 37°C for 5-6 hr, migrates with an Re of approximately 0.7.
after which the cells are harvested by centrifugation and the (A) DE-52 batch elution: 0.25 ml of DE-52 equilibrated
supernatant is treated as described in (A) with the volumes in TE (10:1) is packed into a 1-ml pipetman tip or a dis-
scaled up proportionally. posable column. The sample is diluted by addition of 200 ¡A
of TE (10:1), applied to the column, and washed with 5 x 1
ml of TE (10:1). When using the pipetman tip, the washes
Preparation of M13-RF can be forced through by attachment of the pipetman to the
First, 5 ml of phage stock are prepared as described in (B) end. Unincorporated ATP is eluted by 5 x 1 ml washes
above. Then 50 /il of overnight JM101 and 50 /d of phage with 0.2 MNaCl in TE (10:1). The labeled oligonucleotide
from a 1-ml stock are added to 5 ml of 2YT. (Alternatively, is eluted with 3 x 1 ml washes with 1 MNaCl in TE (10:1).
one plaque can be added in place of phage stock.) The cul- (B) Sephadex G-50 chromatography: Sephadex G-50
ture is incubated at 37°C for 5-6 hr after which the cells are (superfine) is prepared according to the manufacturer's di-
harvested and the supernatant is retained. Two hours after rections, then equilibrated in either 10 mM triethylammo-
the phage culture is started, 2.5 ml of JM101 overnight are nium bicarbonate (pH 8.0) or 10 mM ammonium bicarbo-
added to 500 ml of M9 (supplemented) and incubated at nate (pH 8.0). The column is prepared in a 10-ml plastic
37°C with shaking. This culture is grown to A600 = 0.7-0.8 syringe (bed volume is 10 ml). Buffer is removed from the
then inoculated with 5 x 1012 phage (1010 phage/ml). top of the Sephadex and the sample is applied (total vol-
Growth at 37°C is continued for 2 hr after which the cells ume 200-500 /il). The sample is allowed to enter the bed,
482 ZOLLER AND SMITH

then 2 ml of buffer is applied carefully and 1-ml fractions bridization temperature is 23 °C.) The probe solution is re-
are collected. (A characteristic of this column is that once moved and stored at -20°C. The filter is washed at room
the buffer reaches the top of the Sephadex, the flow essen- temperature with 3 x 100 ml of 6 x SSC, covered with
tially stops until more buffer is added.) Further 1-ml ali- Saran wrap, and autoradiographed for 6-12 hr at -70°C
quots of buffer are added and more fractions are collected. using Kodak XAR-5 film and a Dupont screen. Subsequent
Each fraction is monitored with a Geiger counter. The la- washes at higher temperatures are carried out by placing the
beled oligonucleotide elutes reproducibly between fractions filter into 100 ml of preheated 6 x SSC for 2 min at the de-
4-7, and the sample is dried on a Savant Speed-vac. The sired temperature. Autoradiography is carried out as above
dried sample can be resuspended in water and evaporated for 6-12 hr, depending on the strength of the hybridization
again. signal.
(C) C-18 Sep-pak chromatography: A Sep-pak (Waters)
is prepared by applying 10-ml acetonitrile (Baker, HPLC
Dot blot hybridization
grade) to the matrix using a 10-ml polypropylene syringe,
and the organic solution is washed out with 10-20 ml of Single-stranded DNA (2 /il out of 50 /A from a miniprep)
water. The oligonucleotide sample is diluted with 1-2 ml of is spotted onto a dry sheet of nitrocellulose placed over a
TE (10:1) and applied through the syringe to the Sep-pak. piece of Whatman 3MM. The filter is baked in vacuo for 1
The Sep-pak is washed sequentially with 10 ml of 25 mM hr at 60-80°C, then wetted with 10 ml of prehybridization
ammonium bicarbonate (pH 8.0), 10 ml of 25 mM ammo- solution II for 15 min at room temperature in a petri dish.
nium bicarbonate/5% acetonitrile, two 10-ml aliquots of This is removed and 10 ml of probe solution are added. Hy-
5% acetonitrile/95% H20. The oligonucleotide is then bridization is carried out at 37CC for 1 hr. All washes are
eluted with 3 x 1 ml 30% acetonitrite/70% H20. The sam- carried out as described above. Autoradiography is con-
ples evaporated to dryness on a Speed-vac. This is a
are ducted for 1 hr.
modification of a procedure described by Lo et al. (1984).
Plaque purification
Two-primer oligonucleotide-directed mutagenesis Individual plaques are transferred with a sterile tooth-
(A) Annealing: Single-stranded template DNA (0.5 pick or needle to 100 u\ of sterile TE (10:1). This is mixed
pmol) is mixed with 5' phosphorylated mutagenic oligonu- well, diluted 10" and 105 with TE (10:1), and 10 /d of each
cleotide (10 pmol), nonphosphorylated M13 sequencing dilution are plated with 100 /il of overnight JM101 and 2.5
primer (10 pmol), and 1 /d of 10 x Buffer A in a total vol- ml of YT top agar. Single-stranded DNA is prepared from
ume of 10 /tl. The mixture is heated at 55°C for 5 min, then six plaques of each putative mutant.
placed at room temperature (23°C) for 5 min. During the
annealing reaction, the enzyme/nucleotide solution is pre- DNA sequencing
pared by addition of the following components: 1 ¡A of
lOx Buffer B, 4/dof 2mMdNTPs, 1 /dof lOmMrATP, 3 Chain-terminator DNA sequencing is carried out accord-
units of T4 DNA ligase, 2 units of E. coli DNA pol I flarge ing to Sanger et al. (1980). Single-stranded DNA (5 ¡A out
fragment), and H20 to 10 ¡A. This is kept on ice until use. of 50 /il from a miniprep) is mixed with the sequencing pri-
Shi and Fersht (1984) have suggested that increased fidelity mer (0.2 pmol) and 1 /d 10 x Buffer A in a total volume of
of DNA synthesis is obtained by reducing the final concen- 10 /il. The mixture is heated at 55°C for 5 min, then placed
tration of dNTPs to 20 /iM; we have not investigated this. at room temperature for 5 min. Next, 1 ¡A of [a-32P]dATP
(B) Extension and ligation: After 5 min at room tempera- (600-800 Ci/mmol, 10 /iCi//d) and 0.5 /d (2 units) of DNA
ture, 10 /J of the enzyme/nucleotide solution are added to polymerase I (large fragment) are added to the annealed
the annealed DNA. The contents are mixed, then the tube is DNA. Then 2 /d of this mixture are added to 2 ¡A of each of
placed at 15°C for 6-12 hr. (We generally use 8 hr.) the four dideoxy stop solutions in a 1.5-ml Eppendorf tube
(C) Transformation: The DNA sample is diluted 20 x, and the reaction is initiated by a quick spin in a micro fuge.
100 x, and 500 x with TE (10:1). Then 1 ¡A and 5 /il of each After 15 min at room temperature, 2 /tl of 1 mM dNTPs are
dilution is used to transform JM101 as described above. added to each of the tubes. The reaction is terminated after
15 min by addition of 4 ¡A of a formamide/dye/EDTA so-
lution. The tubes are heated at 90°C for 3 min with the caps
Plaque hybridization off; 1 fA from each tube is electrophoresed on an 8% poly-
A plate with approximately 100-200 plaques is chosen acrylamide-8 M urea gel for 2 hr at 32 watts constant
and placed at 4°C for 15 min. A dry piece of nitrocellulose power. Following electrophoresis, the gel is dried and auto-
is placed onto the plate for 5 min. The filter is carefully re- radiographed with Kodak XRP film overnight.
moved, dried for 5 min, then baked in vacuo at 60-80°C
for 1 hr. The filter is placed in a boilable bag, 10 ml of pre-
hybridization solution I are added, and the sealed bag is in-
Preliminary tests for specific priming by the
cubated at 65°C for 1 hr. Following this, the prehybridiza- mutagenic oligonucleotide
tion solution is removed, and 10 ml of probe solution are First, 20 pmol of mutagenic oligonucleotide are kinased
added. Hybridization is carried out at 37°C for 8-16 hr. asdescribed above. The unincorporated [7-32P]ATP is re-
(For shorter oligomers or AT-rich oligonucleotides, the hy- moved by either G-50 or Sep-pak chromatography. The la-
OLIGONUCLEOTIDE-DIRECTED MUTAGENESIS 483

Mutagenic
5 CCGCAACAG AAAATTT 3 Oligonucleotide
3'-AAGTCGAAAGGCGTTGTC TTTTAAAATATTT -5' Single-stranded
A Template DNA
FIG. 1. Mutagenic oligonucleotide annealed to single-stranded template DNA.

beled oligonucleotide is dried in a Speed-Vac, then resus- quencing procedure described above. The oligonucleotide
pended in 20 /il of water. Single-stranded template DNA does not have to be kinased. The ratio of primer to tem-
(0.5 pmol) is mixed with 1 ¡A 10 x Buffer A, 1 /d unlabeled plate should be adjusted to 20:1 (or the ratio that will be
mutagenic oligonucleotide (8 pmol), and 2 /d kinased oligo- used with the mutagenesis reaction).
nucleotide (2 pmol) in a total volume of 10 ¡A. The mixture
is heated at 55°C for 5 min, then placed at room tempera-
ture for 5 min. During annealing, a mixture of DNA poly-
merase I (\aige fragment), buffer, and dNTPs is made as RESULTS
described above for mutagenesis, except that DNA ligase
and the Ml3 sequencing primer are omitted. Of this mix- This paper presents a simplified protocol of our original
ture, 10 ¡A are added to the primer-template mixture, exten- method for oligonucleotide-directed mutagenesis using vec-
sion is done at 15°C for 2 hr, and the reaction is stopped by tors derived from M13. A single T —
G transversion in the
heating to 65°C for 10 min. After cooling, 10 units of the yeast M>1 Ta gene was constructed to illustrate the proce-
desired restriction endonuclease are added and digestion is dure. The rationale on which this experiment was based will
conducted for 1-2 hr. The extension reaction conditions are be discussed elsewhere in context with other mutations in
compatible with enzymes that are active in 25 mMNaCl, 20 this region.
mM Tris, 10 mM MgCl2, 5 mM DTT. For enzymes that cut
under other conditions, the solution can be either diluted to
decrease NaCl or NaCl can be added. The endonuclease re-
action is stopped by addition of an equal volume of form- Experimental rationale
amide/dye/EDTA solution. The mixture is heated at 90°C Figure 1 shows the DNA sequence of the template in the
for 5 min, then placed on ice until use. An aliquot (5 /d) is region of the desired mutation hybridized with the muta-
electrophoresed on a sequencing gel alongside the reactions genic oligonucleotide. The heptadecanucleotide is perfectly
from dideoxynucleotide sequencing using the mutagenic matched with the template except for a single G-A mis-
oligomer (see below). match that encodes the desired change. Before conducting
the mutagenesis reaction, the two preliminary tests, primer
DNA sequencing using a mutagenic oligonucleotide extension and chain terminator sequencing using the muta-
The mutagenic oligonucleotide is substituted for the se- genic oligonucleotide as primer, were carried out. The basis
quencing primer in the standard dideoxynucleotide se- for these two experiments is summarized in Fig. 2.

PRELIMINARY TESTS
PRIMER EXTENSION

CHAIN TERMINATOR SEQUENCING

C T A G

FIG. 2. Two preliminary tests used to demonstrate specific priming of the mutagenic oligonucleotide at the desired site.
484 ZOLLER AND SMITH

Preliminary tests coelectrophoresed with the chain terminator sequencing re-

actions. The 100-base fragment comigrated with the second
The purpose of these preliminary tests is to determine the
G in the sequence GGCC (Hae III) at the expected site.
optimum conditions to get priming by the mutagenic oligo-
nucleotide from the desired site. The primer-to-template
ratio can be altered as well as the temperature of annealing
Scheme for two-primer mutagenesis
and extension. Generally, we use a 10- to 30-fold molar ex-
cess of primer over template, the annealing is done at 55°C The scheme for two-primer mutagenesis is shown in Fig.
for 5 min, and extension is carried out at 15°C. For this 3. A universal sequencing primer and 5' phosphorylated
particular experiment, we used a 20:1 primer-to-template mutagenic oligonucleotide are annealed to single-stranded
ratio. The results (not shown) indicated that the mutagenic template DNA. Extension of each primer is carried out
oligonucleotide hybridized to the desired site. Primer exten- using E. coli DNA polymerase I (large fragment). The
sion followed by Hae III digestion yielded a fragment 100 newly synthesized strand from the sequencing primer is li-
bases long that represented greater than 90% of the ex- gated to the 5' end of the mutagenic oligonucleotide by T4
tended products observed on the autoradiogram. In the sec- DNA ligase. This forms a gapped heteroduplex bearing a
ond test, use of the mutagenic oligonucleotide as a sequenc- mismatch in the middle. Covalently closed circular mole-
ing primer, the sequencing pattern was unique and clear, cules are not formed since the sequencing primer is not
and it showed the region immediately downstream from the phosphorylated. Extension and ligation are carried out for
desired priming site. The primer extension products were 6-12 hr at 15°C. The DNA is diluted, then used to trans-

HO' Sequencing
Primer
I MATQ/MI3 1
P- Mutagenic
Oligonucleotide
Anneal
(55°C/5min)
\1/

Extend/Ligate (l5°C/6-l2 hrs)

Klenow, DNA Ligase
^ dNTPs, rATP

Transform
E. coli

Screen plaques
for mutants
FIG. 3. Scheme for two-primer mutagenesis.
OLIGONUCLEOTIDE-DIRECTED MUTAGENESIS 485

form E. coli. The resulting plaques are screened for mu- (Fig. 4 and Zoller and Smith, 1982) or by plaque hybridiza-
tants by filter hybridization using the mutagenic oligonu- tion where 100-200 plaques can be screened at once. It is
cleotide as a probe. generally easier to do a plaque lift to obtain a number of
putative mutants than to grow and screen individual
plaques. Plaque lifts also facilitate screening several differ-
Screening for mutants ent mutagenesis experiments at the same time. Figure 5
The basis for detection of mutant DNA is shown in Fig. shows autoradiograms of plaque lift hybridization follow-
4. Radioactively labeled mutagenic oligonucleotide is hy- ing low (23°C) and high (62CC) temperature washes. A
bridized to a filter onto which mutant and wild-type phage number of the hybridization signals present at 23°C were
DNA are bound. The principle of this procedure is that the lost after the 62°C wash. These are presumably wild-type,
mutagenic oligonucleotide will form a more stable duplex whereas the plaques that still have a signal after the 62°C
with a mutant clone having a perfect match than with a wash are putative mutants. On this basis, the efficiency of
wild-type clone bearing a mismatch. Hybridization is car- mutagenesis was 50%.
ried out under conditions of low stringency where both mu- Two putative mutants and one wild-type from this plate
tant and wild-type DNA hybridize with the oligonucleotide were plaque-purified. Single-stranded DNA was prepared
(Fig. 4A). Upon increasing the temperature at which the fil- from six individual plaques from each original clone. Ali-
ter is washed, the mutagenic oligonucleotide remains hy- quots were spotted onto nitrocellulose and hybridized with
bridized to mutant DNA and dissociates from wild-type the labeled mutagenic oligonucleotide (dot blot hybridiza-
DNA (Fig. 4B). At a certain temperature, mutant DNA can tion). All isolates from the suspected mutants hybridized
be easily discriminated from wild-type DNA (Fig. 4C). The after both the 23°C and 62°C wash. In contrast, the DNA
temperature at which mutant and wild-type DNA are dis- from the suspected wild-type only hybridized after the 23°C
criminated depend on the particular mutation and the spe- wash (data not shown).
cific sequence of the mutagenic oligonucleotide. For a It is significant to note that, theoretically, each original
17-mer with one or two mismatches, the wash temperature plaque should contain a 50:50 mixture of wild-type and mu-
is increased in 10° intervals. tant plage. In practice, we typically find that a suspected
Screening can be carried out by dot blot hybridization of mutant plaque contains between 80-100% mutants by anal-
phage DNA prepared from individually isolated plaques ysis of plaque-purified clones. This may reflect a strand

mismatch
DNA bound to
nitrocellulose

••••••
• •••••

probe remains bound to mutants

FIG. 4. Screening for mutants using the mutagenic oligonucleotide as a hybridization probe. Example of 18 clones + 1
WT clone assayed by dot blot. Hybridization was conducted at 37°C, washes were carried out at 23°C, 50CC, and 62°C.
486 ZOLLER AND SMITH

t« •

• * • »
••
**
*
• ••
.1 *• t
t *
• * II 1

23° 62°
FIG. 5. Screening for mutants by plaque lift hybridization using the mutagenic oligonucleotide as a probe.

bias at an early stage of DNA replication or correction of DISCUSSION

the mismatch prior to DNA replication.
DNA from one of the suspected mutants and one wild- We have presented an improved method for oligonucleo-
type were sequenced in the region of the desired mutattion tide-directed mutagenesis using M13-derived vectors. This
using an oligonucleotide that primed 65 bases away. As procedure entails fewer steps than our previous method and
shown in Fig. 6, the correct T —
G transversion was con- results in an equally high efficiency of mutagenesis. Two
structed in the putative mutant. No other base change was primers are used, one of which is the mutagenic oligonu-
observed in this region. For functional analysis of this mu- cleotide and the other is an oligonucleotide that primes
tant, M13 RF was prepared, and the MA T insert was iso- some distance away. Each primer is extended on the same
lated and cloned into a yeast vector (see Discussion). single-stranded template by E. coli DNA polymerase I
Wild type Mutant

C T A G

' ;
m

FIG. 6. DNA sequence of mutant and wild-type clones in the region of the mutation. A sequencing primer was used that
primed 65 bases away from the desired mutation.
OLIGONUCLEOTIDE-DIRECTED MUTAGENESIS 487

(large fragment). The "upstream" primer extends toward A useful feature of the Norris et al. procedure is that the
the 5' end of the mutagenic oligonucleotide, and upon liga- mutant can be placed in a vector suitable for subsequent
tion to it forms a gapped heteroduplex. Wild-type and mu- biological experiments. This is generally not the case with
tant phage are propagated and segregated in vivo by trans- M13-derived vectors, from which the mutant insert must be
fecting E. coli. Screening for mutant phage is accomplished isolated and recloned. However, it is possible to minimize
using the mutagenic oligonucleotide as a hybridization this problem by constructing vectors ready for biological
probe. The basis for this screen is the differential thermal use that also contain the origin from M13 or a similar sin-

stability of a matched versus mismatched duplex. As an ex- gle-stranded phage. Two such derivatives have been devel-
ample, a 17-mer was used to create a T G transversion in
—
oped by Dente et al. (1983) and Zagursky and Berman
the MAT* gene of S. cerevisiae. The procedure, from pri- (1984). The pEMBL vectors of Dente et al. have been suc-
mer extension to sequenced mutant, took about 5 days with cessfully used in conjunction with the protocol presented
less than 3 hr of work per day. The efficiency of mutagene- here (D. Goodin, UBC, Vancouver, personal communica-
sis was approximately 50%. tion).
There are several differences between this procedure and Currently there are a number of relatively simple and re-
our previously published protocol (Zoller and Smith, 1982, liable procedures for oligonucleotide-directed mutagenesis
1983). The extension and ligation reaction has been short- (Wallace et al., 1981 and accompanying paper; Zoller and
ened from 12-20 hr to 6-12 hr. We generally extend and Smith, 1982 and this paper; Lewis et al., 1983; Norris et al.,
ligate for 8 hr. Following this, the DNA is used directly for 1983; Oostra et al., 1983). The choice of which method to
transfection; the isolation of closed covalently circular use may depend simply on the familiarity of an individual
DNA by alkaline sucrose gradient centrifugation has been with the particular vector system and enzymes utilized with
omitted. The formation of covalently closed double- the method.
stranded molecules cannot occur since the "upstream" pri-
mer was not 5' phosphorylated. The use of the upstream

primer is to place the mismatch within the interior of the ex- ACKNOWLEDGMENTS
tended fragment, reminiscent of marker rescue experiments
with 0X174 (Hutchison and Edgell, 1971). We thank Louisa Dalessandro for typing this manu-
The position of the upstream primer may be important,
script, Debbie Balke for art work, and Dave Greene for
but we have not investigated this aspect fully. In this experi-
photography. Additional thanks go to Gary Pielak for
ment, the upstream primer was an M13 sequencing primer
that extends from a position 2.9 kb upstream of the muta-
helpful comments during development of this procedure
and to Tom Atkinson for help with manual DNA synthesis.
genic oligomer. Oligonucleotides other than the universal We would also like to acknowledge Peter Seeburg and Jo
sequencing primer have been used which prime 1 kb from Messing for suggesting the use of two primers. Lastly, we
the mutagenic oligomer (M. Zoller, unpublished). In addi- would like to thank the students of the Cold Spring Harbor
tion, a successful experiment has been completed in which Laboratory Advanced Cloning Course (1983 and 1984),
the upstream primer was only 200 bases away (W. Herr, who demonstrated the versatility of this protocol. This
Cold Spring Harbor Laboratory, personal communica- work was supported by grants to M.S. from the MRC of
tion). However, in this case, the efficiency was around 1 %. Canada. M.S. is a career investigator of the MRC of
This may be a reflection of the proximity of the two Canada.
oligonucleotides.
The features of the mutagenic oligonucleotide for use in
this procedure are the same as described earlier (Zoller and
Smith, 1982). Generally, in making small changes of 1-3 REFERENCES
bases, 17-20-mers are used. The same procedure has been
successfully used to make large deletions (0.5-1 kb) as well ABARZUA, P. and MARIANS, K.J. (1984). Enzymatic tech-
as insertions of restriction endonuclease sites. Deletions niques for the isolation of random single-base substitutions in
vitro at high frequency. Proc. Nati. Acad. Sei. USA 81, 2030-
and restriction site insertions are generally made using 30-
2034.
40-mers. Prior to conducting the experiment, it is impor-
ADAMS, S.P., KAVKA, K.S., WYKES, E.J., HOLDER, S.B.,
tant to carry out the preliminary tests to demonstrate that and GALLUPI, G.R. (1983). Hindered dialkylamino nucleo-
the mutagenic oligonucleotide primes specifically. side phosphite reagents in the synthesis of two DNA 51-mers. J.
The procedure described here differs from that of Norris Am. Chem. Soc. 105, 661-663.
et al. (1983) in the steps that follow the formation of the CARUTHERS, M.H., BEAUCAGE, S.L., BECKER, C., EFCA-
gapped heteroduplex. In their procedure, single-stranded VITCH, J.W., FISHER, E.F., GALLUPPI, G., GOLDMAN,
circular DNA serves as a template for two-primer exten- R., DEHASETH, P., MATTEUCCI, M., McBRIDE, L., and
sion. Following ligation, the insert, a fully double-stranded STABINSKY, Y. (1983). Deoxyoligonucleotide synthesis via
the phosphoramidite method. In Gene Amplification and Anal-
wild-type mutant heteroduplex, is isolated and inserted into ysis, vol. 3. T.S. Papas, M. Rosenberg, and J.G. Chirikjian,
a plasmid-based vector. Screening and sequencing are
Eds. pp. 1-25. Elsevier, New York.
carried out on the double-stranded plasmid. The efficiency
DENTE, L., CESARENI, G., and CÓRTESE, R. (1983).
of mutagenesis was equally high compared with that pre- pEMBL: A new family of single stranded plasmids. Nucleic
sented here. However, it is clear from our studies that mu- Acids Res. 11, 1645-1655.
tants can be obtained with high frequency using gapped FRANK, R., HEIKENS, W., HEISTER-MOUTSIS, G., and
molecules. BLOCKER, H. (1983). A new general approach for the simulta-
488 ZOLLER AND SMITH

neous chemical synthesis of large numbers of oligonucleotides: and SMITH, A.E. (1983). Transforming activity of polyoma vi-
Segmental solid supports. Nucleic Acids Res. 11, 4365-4377. rus middle-T antigen probed by site-directed mutagenesis. Na-
GAIT, M.J., MATHES, H.W.D., SINGH, M., SPROAT, B.S., ture 304, 456-459.
and TITMAS, R.C. (1982). Rapid synthesis of oligodeoxyribo- SANGER, F., COULSON, A.R., BARRELL, B.G., SMITH,
nucleotides VII. Solid phase synthesis of oligodeoxyribonucleo- A.J.H., and ROE, B.A. (1980). Cloning in single-stranded bac-
tides by a continuous flow phosphotriester method on a kiesel- teriophage as an aid to rapid DNA sequencing. J. Mol. Biol.
guhr-polyamide support. Nucleic Acids Res. 10, 6243-6254. 143, 161-178.
HEFFRON, F., SO, M., and MCCARTHY, B.J. (1978). In vitro SHI, J.-P. and FERSHT, A.R. (1984). Fidelity of DNA replica-
mutagenesis of a circular DNA molecule using synthetic restric- tion under conditions used for oligonucleotide-directed muta-
tion sites. Proc. Nati. Acad. Sei. USA 75, 6012-6016. genesis. J. Mol. Biol. 177, 268-278.
HUTCHISON, C.A. and EDGELL, M.H. (1971). Genetic assay SCHOLD, M., COLOMBERO, A., REYES, A.A., and WAL-
for small fragments of bacteriophage 0X174 deoxyribonucleic LACE, R.B. (1984). Oligonucleotide-directed mutagenesis
acid. J. Virol. 8, 181-189. using plasmid DNA templates and two primers. DNA 3, 469-
HUTCHISON, C.A., PHILLIPS, S., EDGELL, M.H., GIL- 477.
LAM, S., JAHNKE, P., and SMITH, M. (1978). Mutagenesis SHORTLE, D., DIMAIO, D., and NATHANS, D. (1981). Di-
at a specific position in a DNA sequence. J. Biol. Chem. 253, rected mutagenesis. Ann. Rev. Genet. 15, 265-294.
6551-6560. SHORTLE, D., GRISAFI, P., BENKOVIC, S.J., and BOT-
ITAKURA, K., ROSSI, J.J. and WALLACE, R.B. (1984). Syn- STEIN, D. (1982). Gap misrepair mutagenesis: Efficient site-di-
thesis and use of oligonucleotides. Ann. Rev. Biochem. 53, rected induction of transition, transversion, and frameshift mu-
323-356. tations in vitro. Proc. Nati. Acad. Sei. USA 79, 1588-1592.
LEWIS, E.D., CHEN, S., KUMAR, A., BLANCK, G., POL- STONE, J.C., ATKINSON, T., SMITH, M., and PAWSON, T.
LACK, R.E., and MANLEY, J.L. (1983). A frameshift muta- (1984). Identification of functional regions in the transforming
tion affecting the carboxyl terminus of the simian virus 40 large protein in Fujinami sarcoma virus by in-phase insertion muta-
tumor antigen results in a replication- and transformation-de- genesis. Cell 37, 549-558.
fective virus. Proc. Nati. Acad. Sei. USA 80, 7065-7069. WALLACE, R.B., SCHOLD, M., JOHNSON, M.J., DEMBEK,
LO, K.M., JONES, S.S., HACKETT, N.R., and KHORANA, P., and ITAKURA, K. (1981). Oligonucleotide directed muta-
H.G. (1984). Specific amino acid substitutions in bacterioopsin: genesis of the human 0-globin gene: A general method for pro-
Replacement of a restriction fragment in the structural gene by ducing specific point mutations in cloned DNA.
synthetic DNA fragments containing altered codons. Proc. WASYLYK, B., DERBYSHIRE, R., GUY, A., MOLKO, D.,
Nati. Acad. Sei. USA 81, 2285-2289. ROGET, A., TEOULE, R., and CHAMBÓN, P. (1980). Spe-
MANIATIS, T., FRITSCH, E.F., and SAMBROOK, J. (1982). cific in vitro transcription of conalbumin gene is drastically de-
Molecular Cloning: A Laboratory Manual. (Cold Spring Har- creased by single-point mutation in T-A-T-A homology se-
bor Laboratory, Cold Spring Harbor, New York). quences. Proc. Nati. Acad. Sei. USA 77, 7024-7028.
MATTEUCCI, M.D., and HEYNEKER, H.L. (1983). Targeted WINTER, G., FERSHT, A.R., WILKINSON, A.J., ZOLLER,
random mutagenesis: The use of ambiguously synthesized M.J., and SMITH, M. (1982). Redesigning enzyme structure by
oligonucleotidesto mutagenize sequences immediately 5' of an site-directed mutagenesis: Tyrosyl tRNA synthetase and ATP
ATG initiation codon. Nucleic Acids Res. 10, 3113-3121. binding. Nature 299, 756-758.
MATTHES, H.W.D., ZENKE, W.M., GRUNDSTROM, T., ZAGURSKY, R.J. and BERMAN, M.L. (1984). Cloning vectors
STAUB, A., WINTZERITH, M., and CHAMBÓN, P. (1984). that yield high levels of single-stranded DNA for rapid sequenc-
Simultaneous rapid chemical synthesis of over one hundred ing. Gene 27, 183-191.
oligonucleotides on a microscale. EMBO J. 3, 801-805. ZOLLER, M.J. and SMITH, M. (1982). Oligonucleotide-directed
MAXAM, A.M. and GILBERT, W. (1980). Sequencing end-la- mutagenesis using M13-derived vectors: An efficient and gen-
beled DNA with base-specific chemical cleavages. Methods eral procedure for the production of point mutations in any
Enzymol 65, 499-560. fragment of DNA. Nucleic Acids Res. 10, 6487-6500.
MESSING, J. (1983). New M13 vectors for cloning. Methods En- ZOLLER, M.J. and SMITH, M. (1983). Oligonucleotide-directed
zymol. 101, 20-78. mutagenesis of DNA fragments cloned into M13-derived vec-
MILLER, J.H. (1972). Experiments in Molecular Genetics. (Cold tors. Methods Enzymol. 100, 468-500.
Spring Harbor Laboratory, Cold Spring Harbor, New York).
MONTELL, C, FISHER, E.F., CARUTHERS, M.H., and Address reprint requests to:
BERK, A.J. (1982). Resolving the functions of overlapping Dr. Mark Zoller
genes by site-specific mutagenesis at a mRNA splice site. Nature Cold Spring Harbor Laboratory
295, 380-384. P.O. Box 100
NORRIS, K., NORRIS, F., CHRISTIANSEN, L., and FUL, N. Cold Spring Harbor, NY 11724
(1983). Efficient site-directed mutagenesis by simultaneous use
of two primers. Nucleic Acids Res. 11, 5103-5112.
OOSTRA, B.A., HARVEY, R., ELY, B.K., MARKHAM, A.F., Received for publication September 9, 1984.