Positive selection of recombinant plasmids based on the EcoK restriction activity of Escherichia coli K-12

Gene, 150(1994)197-198 0 1994 Elsevier Science B.V. All rights reserved.

197

037%1119/94/$07.00

GENE 08306

Positive selection of recombinant plasmids based on the EcoK restriction activity of Escherichia coli K-12” (DNA methylation; DNA modification; N6-methyladenine; recombinant DNA; cloning vector)

Olivier De Backer, Patrick Chomez and Etienne De Plaen Ludwig Institute,for

Cancer Research,

Received by R.E. Yasbin: 28 March

Brussels Branch, B-1200 Brussels. Belgium 1994; Revised/Accepted:

7 June/l5

June 1994; Received at publishers:

28 July 1994

SUMMARY

We have constructed a pTZ19R-derived vector which allows efficient positive selection of recombinant plasmids. The system uses the EcoK restriction activity of Escherichia coli K-12 to select against non-recombinant plasmids. The vector contains an EcoK site which, if deleted or disrupted by ligating a DNA fragment, yields recombinant plasmids that are no longer susceptible to EcoK restriction when transformed into a restriction-proficient E. coli host.

Escherichia coli K-12 has a type-1 restrictionmodification (R-M) system encoded by the hsdRMS genes (Glover, 1970). The EcoK restriction enzyme encoded by these genes recognizes the hyphenated sequence S-AACNNNNNNG’T’GC-3’ and cleaves DNA at random sites outside of this sequence (Kan et al., 1979). The same EcoK enzyme also protects the resident bacterial DNA from the restriction activity by methylating one adenosine residue (corresponding to the positions of the asterisks) in each strand of the recognition sequence. Most E. coli strains used in gene cloning are r, derivatives of K-12. Some strains are defective in restriction but can still methylate EcoK sites (rim:) while others are deficient in both restriction and methylation (r;m;). However, some commonly used strains, like JMlOl, show a wild-type rK+mKfphenotype, and cleave DNA containCorrespondence Research,

to: Dr. 0.

74 avenue

De Backer,

Hippocrate,

Ludwig

Institute

B-1200 Brussels,

Belgium.

for Cancer Tel. (32-2)

764-7424; Fax (32-2) 762-9405; e-mail: [email protected] *On request, the authors will supply detailed experimental evidence for the conclusions reached in this Brief Note. Abbreviations: Ap, ampicillin; BGal, f3-galactosidase; b-D-thiogalactopyranoside; MCS, multiple cloning

IPTG. isopropylsite; mk. indicates

EcoK modification phenotype; p, plasmid; ‘, resistance/resistant; rk, indicates EcoK restriction phenotype; XGal, 5-bromo-4-chloro-3indolyl-B-D-galactopyranoside. SSDZ 0378-1119(94)00517-6

ing non-methylated EcoK sites; we reasoned that these could be used to select plasmids where an unique EcoK site has been disrupted or removed as a consequence of the insertion of a DNA fragment. Most commonly used plasmid vectors contain a single EcoK site in the coding region of the ApR gene. As a first step, we have analyzed the influence of this site on the ability of plasmid pTZ19R (Pharmacia) to transform the ri strain JMlOl. To do this, the single EcoK site (5’-AACCCACTCGTGC) of pTZ19R was mutated to f-ACCCCACTCGTGC by in vitro site-directed mutagenesis. This A-C transversion affects the third and silent residue in the Gly43 codon of the ApR gene. Nonmethylated DNA of the mutated plasmid (pTZ19R.K-) and of pTZ19R were extracted from the rim; E. coli strain RR1 and used to transform JMlOl (rK+)or DHScl (r;) by electroporation. As shown in Table I, pTZ19R transformed JMlOl about 7%fold less efficiently than does pTZ19R.K-. However DHScr was found to be transformed at the same rate by the two plasmids (Table I). The low transformation rate of JMlOl by pTZ19R was therefore attributed to the EcoK restriction activity of this strain. A similar EcoK bias was observed in transformation experiments using the CaCl, procedure (data not shown). These experiments suggest that the disruption or

198 TABLE

I

Transformation Transformed

efficiencies

of E. coli strains

strain

JMlOl

Transforming

(rK+)and DH5a

(ri) by various

pTZ19Ra

pTZl9R.K(0)

Ap’ colonies

% Lac-

(r$ m:)

2900+400

DH5a

(r; mi)

88000+_28000

per ng of DNA

=

pTZl9RBS.K”

from three transformation

experiments,

kc1857 phage DNA. Percent Ap’ colonies.

of lac-

standard

deviations

by electroporation

colonies

228000+31000

4300 * 400

65000+22000

55000+11000

recombinant

plasmids

are indicated

replacement of an unique EcoK site by ligation of an insert should allow the positive selection of recombinant plasmids in a rK+rnK+ host. We therefore designed vector pTZ19RBS.K. In this vector, the BarnHI-Sac1 S-GATCCCCGGGTACCGAGCT fragment in the MCS of pTZ19R.K- is replaced by the synthetic sequence 5’-GATCGAGCACGGATCCmGGAGCT that contains an EcoK site (underlined) with a BumHI site as the central hexamer (italics). This replacement changes a ProArg-Val-Pro motif into Arg-Arg-Ala-Arg-Ile-Arg-Trp in the 1acZ gene product but conserves the ability of the a peptide to carry out PGal complementation, allowing the blue-white selection of transformed bacteria on XGal/IPTG indicator plates. Non-methylated pTZ19RBS.K was shown to transform JMlOl at about the same low efficiency as pTZ19R (Table I). To test the efficiency of pTZ19RBS.K as a positive selection vector, non-methylated DNA of this plasmid was cut with BumHI and ligated to xc1857 phage DNA digested with the same enzyme. The ligation mixtures were used to electroporate JMlOl (rg) or DHSa (r;), and ApR transformants were selected. As expected, a greater

mixb

% Lac-

colonies

55 (43/78) 2.5 (31124)

pTZlYR.Kor pTZl9RBS.K. The number of EcoK sites efficiencies (Ap’ colonies per ng of DNA) were calculated

by +.

with 20 ng of pTZl9RBS.K with disrupted

Ligation

(1) in MCS

a JMlOl and DH5a were transformed by electroporation with 100 pg of plasmid pTZl9R, in each plasmid is shown in parentheses (followed by its location). Average transformation

b JMlOl and DHSa were transformed

DNAs

DNA

( 1) in ApR gene

JMlOl

non-modified

cut with BumHI and ligated

EcoK site in MCS were deduced

with 20 ng of a BamHI

from the number

of Lac-

digest of

colonies

among

proportion of recombinant colonies were obtained with JMlOl (43 recombinant colonies among 78 analyzed) than with DH5a (3 recombinant colonies among 124) (Table I). In conclusion, the use of pTZ19RBS.K in combination with a Resf host strain enables a positive selection for recombinant plasmids, provided the insert does not contain non-methylated EcoK sites. We believe that the EcoK counterselection of non-recombinant plasmids will be applicable as a general method, since the introduction of an EcoK site at an appropriate location in a plasmid should convert it easily into a positive-selection vector.

REFERENCES Glover,

S.W.:

Functional

analysis

of host-specificity

Escherichia coli. Genet. Res. Camb. 15 (1970) 237-250. Kan, N.C., Lautenberger, J.A., Edgell, M.H. and Hutchinson

mutants

in

III, C.A.:

The nucleotide sequence recognized by the Escherichia coli K-12 restriction and modification enzymes J. Mol. Biol. 130 (1979) 191-209.