Simple Cloning via Direct Transformation of PCR Product (DNA Multimer) to Escherichia coli and Bacillus subtilis

Abstract

We developed a general restriction enzyme-free and ligase-free method for subcloning up to three DNA fragments into any location of a plasmid. The DNA multimer generated by prolonged overlap extension PCR was directly transformed in Escherichia coli [e.g., TOP10, DH5α, JM109, and BL21(DE3)] and Bacillus subtilis for obtaining chimeric plasmids.

TEXT

The limited choices of restriction enzymes, relatively low efficiencies in digestion and ligation, and possible self-ligation of the digested plasmid may result in difficulties in constructing chimeric plasmids. Recently, several companies have developed recombinase-based technologies, such as the Invitrogen Gateway cloning technology, Clontech In-Fusion, BioCat Cold-Fusion, and Red/ET Recombination, but these rely heavily on specialized kits containing vectors, enzymes, or hosts (4, 7, 10, 11, 14, 15). Several overlap extension PCR-based methods were developed for subcloning. However, RF cloning (9) and overlap extension PCR cloning (2) require DpnI to digest the vector template. Additionally, the maximum inserted DNA length is ∼6.7 kb (2). Another technology, called “Quick Assemble,” has low positive cloning efficiencies, ∼33% (16).

We developed a sequence-independent “simple cloning” method without the need for restriction and ligation enzymes. The protocol includes three steps (Fig. 1): (i) linear DNA fragments (i.e., inserted DNA fragment and vector backbone), both of which contained 3′ and 5′ 40- to 50-bp overlapping termini, are generated by high-fidelity PCR with the New England BioLabs (NEB) Phusion polymerase (Ipswich, MA); (ii) the DNA multimer is generated based on these DNA templates by prolonged overlap extension PCR (POE-PCR) with Phusion polymerase; and (iii) the POE-PCR products (DNA multimer) are transformed into competent Escherichia coli or Bacillus subtilis strains directly, yielding the desired chimeric plasmid.

Fig 1.

The scheme of simple cloning. First, two 3′ and 5′ overlapped insertion and vector fragments are generated by regular PCR. Second, a DNA multimer is formed by POE-PCR without primers and with a prolonged extension time. Third, E. coli or B. subtilis strains cleaved the transformed DNA multimer into a circular plasmid, yielding the desired chimeric plasmid.

A 1.3-kb insertion fragment (Cherry-cbm17) encoding a cherry fluorescent protein and a family 17 carbohydrate-binding module from Clostridium cellulovorans cellulase 5A (1) was subcloned into a 3.6-kb pET20b vector backbone, yielding a 4.9-kb plasmid, pET20b-cherry-cbm17, where the fusion protein was controlled by a T7 promoter. A linear vector backbone was amplified by using the forward primer VF (5′TAGCCTGGACAATATCAAATTTACCCTCGAGCACCACCACCACCACCACT3′) and the reverse primer VR (5′TATCCTCCTCGCCCTTGCTCACCATATGTATATCTCCTTCTTAAAGTTAA3′). VF and VR contain the last 25 bp of the 3′ terminus of the insertion sequence (underlined) and the first 25 bp of the 5′ terminus of the vector sequence (bold). Similarly, the insertion fragment was amplified by primers IF (5′TTAACTTTAAGAAGGAGATATACATATGGTGAGCAAGGGCGAGGAGGAT3′) and IR (5′AGTGGTGGTGGTGGTGGTGCTCGAGGGTAAATTTGATATTGTCCAGGCTA3′). IF and IR have the reverse complementary sequences of VR and VF, respectively. The standard extension time (SET) in PCR was calculated based on the amplified fragment length divided by 3 kb/min for Phusion polymerase at 72°C. Two linearized DNA fragments were purified with a Zymo DNA Clean and Concentrator kit (Irvine, CA) (Fig. 2A and C).

Fig 2.

Optimization of extension time in POE-PCR. (A) PCR products generated by the overlap extension PCR at different extension times from 0.3 to 2.5 SET. Lane M, 1-kb DNA ladder from NEB; lane V, vector backbone generated by PCR; lane I, inserted DNA generated by PCR. (B) Transformation efficiency of DNA multimers as a function of extension time. (C) Analysis of a 0.8% agarose gel for the case in which a 1.3-kb DNA was subcloned into 3.6-kb plasmid. Lane 1, PCR linearized vector; lane 2, PCR-linearized insertion; lane 3, DNA multimer generated by modified overlap extension PCR; lane 4, PCR products digested with two restriction enzymes; lane 5, resulting plasmid from a randomly selected E. coli colony; lane 6, resulting plasmid digested with two restriction enzymes; and lane M, a 1-kb DNA ladder from NEB. (D) E. coli BL21(DE3) transformants containing pET20b-cherry-cbm17, expressing red fluorescent protein. (E. coli TOP10, DH5α, and JM109 are preferred to BL21 for regular subcloning.)

In POE-PCR, each tube contained 0.2 mM deoxynucleoside triphosphate (dNTP), 2 ng μl⁻¹ purified insertion DNA fragment, equimolar purified vector backbone, and 0.04 U μl⁻¹ Phusion polymerase without the addition of primers, where the insertion and vector fragments were concomitantly used as primers and templates (Fig. 1). The POE-PCR was conducted as follows: denaturation at 98°C for 30 s; 25 to 30 cycles of denaturation at 98°C for 10 s, annealing at 60°C for 10 s, and extension at 72°C at a rate of 2 kb/min based on the length of the desired chimeric vector. When the extension time was shortened (e.g., 30% of the SET), the PCR products smeared in a 0.8% agarose gel (Fig. 2A). When the SET was extended (0.9-fold or longer), large DNA multimers of repeated insertions and vector backbones in tandem were formed, which cannot move in the gel (Fig. 2A). Although the highest transformation efficiency was obtained at a 0.9 SET (Fig. 2B), it was recommended that the extension time for POE-PCR be 1.3 SET (i.e., 2 kb/min), because of (i) a distinct formation of DNA multimers and (ii) an acceptable transformation efficiency.

Five μl of the POE-PCR product (i.e., approximately one μg DNA multimer) was mixed with 100 μl of competent E. coli BL21 cells or other cells. [Commonly used E. coli strains are capable of cleaving assimilated DNA multimers into the circular plasmid (3, 5)]. Through a standard chemical transformation protocol (6), nearly all colonies appeared red in the LB petri dish (Fig. 2D), suggesting that more than 99% of the transformants contained the plasmid expressing red fluorescent protein. The plasmid randomly isolated from the dish was digested by two restriction enzymes, yielding two fragments, as expected (Fig. 2C, lane 6). The plasmid sequence was further verified by DNA sequencing.

The POE-PCR product (DNA multimer) can be transformed in other commonly used chemically competent and electrocompetent E. coli hosts, such as JM109, DH5α, and Top10, as well as B. subtilis hosts. Transformation efficiencies of DNA multimers were 3.3 × 10⁴/μg in commercial DH5α competent cells, approximately five orders of magnitude lower than the intact circular plasmid (Table 1). When home-prepared (low-transformation-efficiency) E. coli strains BL21, JM109, and Top10 were used, the transformation efficiencies were 43 to 420/μg of DNA multimer. However, such efficiencies were sufficient for subcloning. Much higher transformation efficiencies were obtained with B. subtilis (Table 1), because of its preference for assimilating DNA multimers into circular plasmid (12, 13).

Table 1.

Efficiency of transformation of DNA multimers to different hosts

Strain	Source	Transformation efficiency (/μg)a		Source or reference
		Multimer	Plasmid
E. coli
DH5α	Commercial (Invitrogen)	3.3 × 104	2.7 × 109
	Home-prepared	2.8 × 102	4.5 × 107
BL21(DE3)	Home-prepared	4.2 × 102	6.7 × 106	Invitrogen, Carlsbad, CA
JM109	Home-prepared	4.3 × 101	5.1 × 105
Top10	Home-prepared	3.5 × 102	2.0 × 107	Invitrogen, Carlsbad, CA
B. subtilis SCK6	Home-prepared	1.0 × 107	1.0 × 104	12, 13

^aAll the E. coli cells are chemically competent.

We highly recommend using Phusion polymerase in both PCR amplifications due to its high fidelity and high speed. DNA multimers could be obtained even when the overlap length between two templates was shortened to 20 bp (data not shown), as in regular overlap extension PCR (8). However, to ensure a positive result, 40- to 50-bp overlap lengths are recommended. The concentration of each DNA template in POE-PCR is recommended to be ca. 2 ng/μl or higher, with a 1:1 molar ratio. The extension time of POE-PCR is calculated based on a rate of 2 kb/min for Phusion polymerase to ensure the formation of DNA multimers, in contrast to the use of shorter extension times described elsewhere (2, 9, 16).

By using this method, more than 100 plasmids containing one DNA insertion fragment ranging in length from 0.2 to 10 kb were constructed, and the largest plasmid tested was 11 kb (data not shown). For one DNA fragment insertion cloning, the success rate was 95%. Simple cloning did not work when the DNA assembly by overlap PCR failed. Simple cloning was able to assemble up to four fragments containing overlap regions at both 5′ and 3′ termini in tandem, yielding the desired plasmids (data not shown). In conclusion, simple cloning is rapid, efficient, inexpensive, and flexible.