Vibration and glycerol-mediated plasmid DNA transformation for Escherichia coli


Correspondence: Abolfazl Barzegari, Research Center for Pharmaceutical Nanotechnology, Tabriz University of Medical Sciences, Tabriz 5165665811, Iran. Tel.: +98 411 3367914; fax: +98 411 3367929; e-mail:


Escherichia coli transformation is an essential step in many molecular biology experiments. Despite earlier advances in the field, many studies including shotgun cloning still require more efficient transformation protocols. Chemical transformation has been the most popular method, in which competent cells are transformed following a brief period of heat shock. Here, we report a novel protocol with higher efficiency, in which competent E. coli cells (treated with CaCl2) grown in media containing glycerol experience a gentle vibration. Three E. coli strains DH5α, Jm107 and BL21 (DE3) and three plasmids pGEM-T, pET-28a and pCAMBIA with different sizes (3000, 5369 and 8428 bp, respectively) were used to test the protocol. The results indicated a significant increase in number of transformed colonies compared with heat-shock method. Our findings also demonstrated the favourable impacts of glycerol on transformation of E. coli.


Escherichia coli is extensively used in molecular biology as a desired micro-organism for various cloning experiments. Invention of efficient methods for introducing DNA vectors into this bacterium is of special importance. Successful cloning of DNA fragments from different sources (e.g. in cDNA or genomic library construction) with minimum loss of the input DNA throughout the transformation process is crucially important, especially when large numbers of colonies are needed or when the source DNA is limited (Sheng et al., 1995). To date, several methods including chemical procedures (Inoue et al., 1990), high-voltage electroporation (Dower et al., 1988) and biolistic gun (Smith et al., 1992) have been exploited for transformation of E. coli. Microwaves and ultrasounds have also been utilized to improve the transformation efficiency (Song et al., 2007; Fregel et al., 2008). Owing to their accessibility, simplicity and quickness, chemical methods have attracted plenty of interest and are continuously being modified to improve efficiency (Norgard et al., 1978; Dagert & Ehrlich, 1979).

Basically, introduction of exogenous DNA into E. coli was first demonstrated by Mandel & Higa (1970). They showed that bacteria that are treated with ice-cold solutions of CaCl2 and then briefly heated to 37 °C or 42 °C could be transfected with bacteriophage λ DNA. Subsequently, a number of modifications were suggested to improve the existing protocol. These attempts consisted of changes in chemicals and solutions used for preparation of competent cells (Sambrook & Russell, 2006), optimizing the appropriate duration and temperature of heat-shock treatment (Singh et al., 2010), testing the transformability of E. coli at different phases of growth (Brown et al., 1979) and finding the proper optical density (OD) for preparation of competent cells (Tang et al., 1994).

In this study, we aimed to improve previous chemical methods utilizing a gentle vibration process. Three plasmid vectors (pGEM-T, pET-28a and pCAMBIA) with different sizes (3000, 5369 and 8428 bp, respectively) were transformed by vibration of three E. coli strains, namely DH5α, Jm107 and BL21 (DE3) grown with and without glycerol. Finally, the efficiency of this method was compared with that of heat-shock method in which an ice-cold mixture of competent cells (grown in the presence and absence of glycerol) and plasmids were exposed to a short period of heat shock (42 °C for 90 s) (Cohen et al., 1972).

Materials and methods


Escherichia coli strains DH5α, Jm107 and BL21 (DE3) were obtained from Life Technologies, Fermentas and Novagen, respectively. pGEM–T, pET-28a and pCAMBIA 3300 plasmids were purchased from Promega (#A3600), Novagen (#69864-3) and Cambia Labs, respectively. Luria–Bertani (LB) broth and LB agar were both from Sigma-Aldrich Co. (#L3022 and #L2897). Autoclaved CMC solution was made of 80 mM CaCl2(2H2O) (Sigma-Aldrich, #C2536) and 20 mM MgCl2 (Sigma-Aldrich, #M8266). Glycerol (should be autoclaved), ampicillin (Ap) and kanamycin (Km) were all obtained from Sigma-Aldrich (corresponding to #G5516, #A9393 and #K1377 catalogue numbers).

Transformation protocol

Single colonies of E. coli strains included in this study were separately inoculated in 10 mL of LB broth medium and were left overnight at 37 °C with moderate shaking (250 r.p.m.). A falcon tube containing culture medium (without bacteria) was considered as control. The aforementioned cultures in a volume of 200 μL were transferred into 10 mL of LB broth containing 2.5% glycerol inside 50-mL falcon tubes and incubated at 37 °C for 3 h with moderate shaking (250 r.p.m.). OD of the cultures was regularly monitored. The culture tubes were incubated on ice when the OD600 nm became 0.9 (Tang et al., 1994). The bacterial cultures were transferred into 2-mL microtubes and centrifuged at 1600 g for 5 min at 4 °C. The supernatant was discarded, and the tubes were stood in an inverted position on a sterile Whatman paper to remove the last drops of media. Pellets were gently resuspended in 250 μL ice-cold CMC solution and stored on ice for 30 min. The cells were recovered by centrifugation at 1000 g for 3 min at 4 °C. The supernatant was discarded, then 250 μL of CMC/glycerol mixture (70 : 30 ratio) was added, and pellets were gently resuspended. 1 × 109 copy numbers of pCAMBIA3300, pET-28a and pGEM-T plasmids in uncut form were added in a volume not exceeding 5% of that of the competent cells (Hanahan, 1983). Nanogram weights for pCAMBIA3300, pET-28a and pGEM-T vectors to include 1 × 109 copies were 9.1, 5.8 and 3.25, respectively. Contents of tubes were gently mixed and incubated on ice for 30 min. Ice-cold LB broth (20 μL) was added to tubes while they were being kept on ice (In this protocol, SOC medium could be used instead of LB). Tubes were immediately vibrated at 37 °C using a Heidolph Reax Top test tube vibrating shaker for 10 min. 500 μL of prewarmed 37 °C LB broth was added to each tube. Tubes were incubated at 37 °C shaking incubator for 30 min with moderate shaking at 250 r.p.m. Cells were collected by centrifugation at 1600 g for 3 min at room temperature. The liquid phase was discarded under aseptic conditions, and 50 μL of 37 °C LB broth was added into each tube; then, the bacterial pellets were resuspended completely (Km resistance was considered as selectable marker for pET-28a and pCAMBIA. Am resistance was used for screening of bacteria transformed with pGEM-T vector). Final concentrations of Am and Km in the reaction tubes were adjusted to 50 and 30 μg mL−1, respectively. Contents of these tubes were transferred on screening plates containing appropriate antibiotics. Each mixture was spread on a single plate with a sterilized bent glass rod. The plates were incubated at 37 °C for 12–16 h in an inverted position, and subsequently, plates were collected to count the colonies. All experiments were carried out in triplicate.

Technical hints

The current protocol should be carried out under aseptic condition. Furthermore, during the transformation process, starvation of bacteria has a negative impact on their viability. It should also be noted that if the competent cells are not being used immediately, they should be stored in −70 °C until needed. An important consideration in preparation of competent cells is choosing a suitable OD to harvest the bacteria. The OD600 nm must be < 0.9–1. The highest transformation efficiencies have been obtained at two separate points in the growth curve of E. coli: I; in early-to-mid-log phase (OD600 nm = 0.4) (Hanahan, 1983) and II; in late-log phase (OD600 nm = 0.95) (Tang et al., 1994). The optimum concentration of glycerol used in media was 2.5%. In this concentration, transformation efficiency and OD were both acceptable (data not shown).

Statistical analysis

In order to compare vibration and glycerol-mediated method with the common heat-shock procedure (Cohen et al., 1972), vibration and heat-shock methods were both applied in media with and without glycerol and the four different processes were compared. Data were analysed using spss 14 according to completely randomized design. Analysis of variance was performed within each plasmid. Means were compared by the Duncan test at alpha value of (< 0.05).

Results and discussion

The common heat-shock method has an acceptable efficiency for routine cloning experiments. However, the vibration and glycerol-mediated transformation method demonstrated higher efficiency than heat-shock method. As shown in Table 1, vibration leads to a significant increase in the number of transformed colonies either in presence or absence of glycerol. Adding glycerol to the culture media improves the transformation efficiency both in vibration and heat-shock methods.

Table 1. Comparison of transformation efficiency among heat-shock and vibration methods with and without glycerol
Transformation methodE. coli strainPlasmidEfficiency (mean log)
Heat shock without glycerol DH5α pGEM 6.279 ± 0.02
pET-28a 6.176 ± 0.02
pCAMBIA 5.255 ± 0.02
BL21(DE3) pGEM 6.255 ± 0.02
pET-28a 6.176 ± 0.02
pCAMBIA 5.255 ± 0.02
Jm107 pGEM 6.230 ± 0.02
pET-28a 6.204 ± 0.02
pCAMBIA 5.146 ± 0.02
Vibration without glycerol DH5α pGEM 7.114 ± 0.02
pET-28a 7.079 ± 0.02
pCAMBIA 6.204 ± 0.02
BL21(DE3) pGEM 7.146 ± 0.02
pET-28a 7.079 ± 0.02
pCAMBIA 6.146 ± 0.02
Jm107 pGEM 7.114 ± 0.02
pET-28a 7.079 ± 0.02
pCAMBIA 6.079 ± 0.02
Heat shock & glycerol DH5α pGEM 6.580 ± 0.02
pET-28a 6.505 ± 0.02
pCAMBIA 5.462 ± 0.02
BL21(DE3) pGEM 6.531 ± 0.02
pET-28a 6.477 ± 0.02
pCAMBIA 5.431 ± 0.02
Jm107 pGEM 6.580 ± 0.02
pET-28a 6.477 ± 0.02
pCAMBIA 5.505 ± 0.02
Vibration & glycerol DH5α pGEM 7.380 ± 0.02
pET-28a 7.301 ± 0.02
pCAMBIA 6.477 ± 0.02
BL21(DE3) pGEM 7.342 ± 0.02
pET-28a 7.362 ± 0.02
pCAMBIA 6.462 ± 0.02
Jm107 pGEM 7.398 ± 0.02
pET-28a 7.398 ± 0.02
pCAMBIA 6.415 ± 0.02

Although high number of transformants is not necessary in all cloning experiments, construction of genomic or cDNA libraries still demands highly efficient transformation protocols. In fact, missing fragments during genome sequencing projects can lower the accuracy of techniques such as shotgun cloning, subsequently affecting the precision of the entire results.

In comparison with common heat-shock method (Table 1), all other three methods (vibration without glycerol, vibration and glycerol as well as heat shock with glycerol) significantly improved the transformation efficiency of the three tested plasmids in the three bacterial strains (< 0.05). Vibration and glycerol proved to be the most efficient transformation method for every plasmid (< 0.05). Furthermore, all methods were significantly different from each other in terms of transformation efficiency (< 0.05). The pattern of transformation efficiency (mean log) of tested transformation methods for pGEM in E. coli strains DH5α, BL21 and Jm107 has been shown in Fig. 1 (All other tested plasmids follow the same pattern as pGEM).

Figure 1.

The transformation efficiency (mean log) of tested transformation methods for pGEM in Escherichia coli strains DH5α, BL21 and Jm107. (All other tested plasmids follow the same pattern as pGEM).

Glycerol may affect the transformation of plasmids in two stages throughout the protocol: I, when glycerol is added into culture media and II, when glycerol is added to the competent cells. The effects of different chemicals such as saponins or cholates on permeability of the bacterial membranes and transformation have been previously studied (Ravnikar et al., 2009). Stuy & Walter (1986) had also reported the influence of glycerol on the plasmid uptake of bacteria. They suggested that glycerol might disrupt the organization of the membrane proteins and thus increase the plasmid uptake through the membranes. Besides, according to Kobayashi et al. (1986), effect of glycerol on E. coli membrane structure is certain. The limited synthesis of phosphatidylethanolamine in the temperature-sensitive E. coli mutants results in abnormal composition of membrane phospholipids (Kobayashi et al., 1986). They showed that the intracellular pool of glycerol 3-phosphate is a limiting factor for acidic phospholipid synthesis in the mutants. This could lead to facilitated transfer of plasmids through the less stable membranes. Furthermore, in Bacillus lichenisformis, the addition of glycerol influences the synthesis of cell membrane phospholipids, which play important roles in substrate absorption and metabolite secretion (Du et al., 2005). Glycerol addition leads to increase in phosphatidylglycerol content in the membrane. The addition of glycerol also results in an obvious decrease in C16:1 and C18:1 fatty acids (which leads to increased permeability of cell membrane) and increase in C10:0 and C12:0 fatty acids (which may improve the fluidity of cell membrane) (Du et al., 2005). A similar phenomenon in E. coli may explain the higher efficiency of transformation with increase in the permeability and fluidity of the membrane with addition of glycerol.

A number of other physical techniques have been developed for transformation of E. coli competent cells. A sliding-friction method has been described by Yoshida et al. (2007) for transferring plasmid vectors. In this method, a mechanical force makes plasmid–nanocarrier complex penetrate the bacterial coverings. In our method, vibration of a mixture of competent cells, glycerol and plasmid vectors possibly results in a similar outcome. Utilizing lower amounts of kinetic energy in vibration technique compared with sliding-friction prevents the destruction of competent cells. Furthermore, glycerol contributes to the viscosity and prevents severe shaking and shear stress, which is believed to crush competent cells during vibration. Hypothetically, vibration may cause more disruption in structure and functioning of glycerol-treated membranes. This could result in plasmid penetration in cell enclosure. There is one other possibility that has been previously described for ultrasound-mediated transfer of plasmid DNA. Song et al. (2007) have proposed that transmitted energy form ultrasound source could result in temporary porosity in the cell membrane, which enables the plasmids to enter through the pores.

In this study, a novel vibration and glycerol-mediated transformation method was devised. This new technique was shown to transfect three plasmids of varying sizes into three different strains of Ecoli with higher efficiency than that of heat-shock method. Besides, the favourable effect of glycerol on transformation in vibration and heat-shock methods was demonstrated. Further research into the mechanism by which vibration and glycerol enhance the transformation of E. coli is encouraged.


Authors would like to acknowledge Dr Abbas Rafat for his valuable contribution in data analysis. The authors declare no competing interests.

Authors' contribution

D.S., A.A.S. and H.Z contributed equally to this work.