Amplifiable DNA from Gram-negative and Gram-positive bacteria by a low strength pulsed electric field method

Abstract

An efficient electric field-based procedure for cell disruption and DNA isolation is described. Isoosmotic suspensions of Gram-negative and Gram-positive bacteria were treated with pulsed electric fields of <60 V/cm. Pulses had an exponential decay waveform with a time constant of 3.4 µs. DNA yield was linearly dependent on time or pulse number, with several thousand pulses needed. Electrochemical side-effects and electrophoresis were minimal. The lysates contained non-fragmented DNA which was readily amplifiable by PCR. As the method was not limited to samples of high specific resistance, it should be applicable to physiological fluids and be useful for genomic and DNA diagnostic applications.

INTRODUCTION

Cell disintegration is the first and one of the most critical steps in the extraction of intracellular compounds (1,2). Traditionally, mechanical, physical, chemical and biological disruption techniques have been used to disintegrate cells (3). The choice of the appropriate disintegration method or their combination depends on the cell and tissue type, the kind of intracellular compound to be extracted, their susceptibility to the treatment and the desired degree of purity. Integrity of the intracellular compounds to be analyzed may be another important criterium. Speed of the procedure as well as other considerations including sample number, i.e. the degree of parallelization, and sample scale are additional determinants.

For genomic applications and in nucleic acid-based diagnostics, isolation of DNA is of particular interest with aspects of parallelization or arraying and automation increasingly governing the favored methodical approaches. Combined automated processes encompassing tissue homogenization, cell disruption and nucleic acid isolation are of increasing interest (4–6).

Disintegration methods should ideally be universal with respect to the type and property of sample used. DNA yield is of prime importance and the isolated DNA should be of high integrity. Purity is of minor interest, because several excellent DNA purification systems are available. Stimulated by the soaring interest in genomic analyses, automated high sample throughput has become a particular important aspect. Thus, standardization, parallelization and prevention of cross contamination have become guidelines when a cell disintegration method is to be implemented. Unfortunately, most presently used methods are either toxic, expensive, time consuming and laborious, or limited to certain applications (7).

Of the methods available, chemical or biochemical cell disintegration methods are best suited for automated applications in DNA analysis. They are especially suited for routine applications such as the analysis of DNA from blood-borne viruses or for leucocyte DNA analysis. However, chemical lysis is restricted to certain sample types and properties, thus lacking universality. For example, standard procedures do not cover the disintegration of Gram-positive bacteria or fungi, cells for which specialized protocols need to be established (8–10).

Electric methods have a great potential in combining cell disruption and DNA isolation steps without the need for extensive chemical procedures to be included. Moreover, electric sample preparation bears the potential to be combined with down-stream steps such as electrically-enhanced hybridization procedures or electric detection (11–13). In this context, one approach has been described that includes direct disruption of Gram-negative bacteria on an electric sample preparation array and subsequent optical detection of the released nucleic acids (4). Unfortunately, this and other related recent approaches are currently limited to the application of cell suspensions of low ionic strength and high specific resistance (4,14) requiring further preparative steps including the massive dilution of physiological fluids. Other electric methods are based on the use of high electric field strengths, thus being disadvantageous with respect to aspects of laboratory safety and routine use (15–17).

Here we set out to devise a procedure to efficiently disintegrate bacteria by electric field pulses of low to moderate strength. The method was not to be limited to solutions of low ionic strength, was ideally to be applicable to solutions of physiological osmotic strength and should allow for the subsequent isolation and analysis of elected target DNAs.

MATERIALS AND METHODS

Materials

The high-voltage generator, HVG 30-2/HII, was purchased from Eltex (Weil am Rhein, Germany). The reaction chamber (RC) and the pulse generator (PG) were developed and manufactured with the kind assistance of A. Scherrmann (Fraunhofer IPA, Stuttgart, Germany). Polycarbonate was chosen as non-conducting material for the cell disintegration chamber. The aluminium or stainless steel electrodes had an area (A) of ∼0.64 cm² and a spacing (d) of 0.4 cm. The cell disintegration chamber was thermostated by a Julabo F 10 thermostat from Julabo Labortechnik GmbH (Seelbach, Germany). The PG was built as outlined previously (18) to prevent electrochemical and electrophoretic effects. It was equipped with a capacitance (C) of ∼1.67 nF. The time constant (τ) which describes the exponential decay waveform of the electric field strength according to:

E_t = E₀ e ^(–t/τ)

was ∼3.4 µs. E₀ is the initial field strength and t the time. The time constant is the product of the capacitance and the overall resistance (R). The overall resistance is the sum of the resistance of the PG (R_PG; ∼1 kΩ) plus the junction resistance of the RC (R_J; ∼1 kΩ) and the RC itself containing the cell suspension [R_RC ∼40 Ω and τ = C (R_PG + R_J + R_RC)]. A Tektronix 2430A oscilloscope from Tektronix Inc. (Köln, Germany) was used to determine the initial voltage U₀. The initial field strength was calculated as follows:

E₀ = (U₀/d) (1/(R_J/R_RC + 1)) F_Osc

with F_Osc being the conversion factor of the oscilloscope. To modulate the frequency applied, a frequence generator type 166 from Wavetek GmbH (Ismaning, Germany) was used.

Micrococcus luteus was kindly provided by I. Trick from the Fraunhofer IGB. Escherichia coli DH5α containing the pET11b plasmid carrying a human macrophage migration inhibitory factor (MIF) insert (phuMIF) has been described (19).

The SYBR Green I DNA quantification reagent and the AIDA 2.0 software were from Raytest Isotopenmeßgeräte GmbH (Straubenhardt, Germany). The NucleoSpin C+T kit for the isolation of genomic DNA was from Macherey-Nagel GmbH and Co. KG (Düren, Germany) and the QiaPrep kit was from Qiagen GmbH (Hilden, Germany). DNA molecular weight markers and other molecular biology reagents were from Roche Diagnostics (Mannheim, Germany), Gibco BRL Life Technologies GmbH (Karlsruhe, Germany) or Qiagen. Primers used were from Gibco. Primers 5′-GCT TTA CCT CAA GGT CGA TAC C-3′ and 5′-GGA GGT GTT CAC CAT GTA TCT G-3′ covered a 348 bp huMIF cDNA sequence (19) and primers 5′-GGT GGG CTG CTT TAA ATA TAT TAC C-3′ and 5′-AGT TGC TTA AAG AAG CAG AAA CAG A-3′ were used to amplify a 90 bp region of the E.coli lactose operon (20). The BioPrint system from LTF-Labortechnik (Wasserburg, Germany) was used for gel documentation and DNA quantification (21). Miscellaneous chemicals, salts and media were from Merck KGaA (Darmstadt, Germany), Sigma-Aldrich Chemie GmbH (Deisenhofen; Steinheim, Germany), Carl Roth GmbH and Co. (Karlsruhe, Germany) and Difco Laboratories GmbH (Augsburg, Germany) and were of the highest analytical grade available.

Bacterial cultures

Micrococcus luteus and E.coli bacteria were grown overnight at 37°C on DSM 53 and Luria-Bertani (LB) agar plates containing 100 µg/ml ampicillin, respectively (22). One liter DSM 53 media contained 10 g caseinpeptone tryptically digested, 5 g bacto yeast extract, 5 g glucose, 5 g sodium chloride and 15 g bacto agar. The pH value was adjusted to 7.4. Single colonies were grown in the corresponding liquid medium for 8 h at 37°C on a shaker at ∼200 r.p.m. Aliquots of these pre-cultures were grown overnight under the same conditions until the stationary phase was reached (23). The cells were pelleted at 600 g for 15 min at 4°C, washed twice with phosphate-buffered saline, pH 7.4, containing 3 mM EDTA (PBS/EDTA) (22), and the cells resuspended in the same buffer. EDTA was added to prevent DNA non-specific adsorption to the cells (24) and to minimize DNA degradation. To determine cell concentrations, appropriate dilutions of E.coli and M.luteus suspensions were grown overnight on LB (22) or plate count agar plates (25), respectively. Cell suspensions contained about 3 × 10⁹ c.f.u./ml. As an additional control for the concentration and constitution of the cell suspensions, the genomic and plasmid DNA yields were determined by using the NucleoSpin C+T and Qiaprep kits, respectively. To isolate DNA from M.luteus, cells were additionally incubated with lysozyme according to the manufacturer’s instructions.

Analysis of degree of cell disintegration

Following cell disintegration, suspensions were centrifuged at 4°C for 5 min at 16 000 g. The supernatant, termed herein the lysate, was then used for subsequent spectrophotometric DNA and protein analyses following described methods (21,26). Negative controls were not subjected to the electric treatment procedure but were otherwise treated identically with the electrically-lysed samples. Values from treated samples are expressed as net values with the control values already subtracted. Lysates were first scanned spectrophotometrically from 320 to 220 nm, protein concentrations determined by Bradford analysis in microplates (27,28), and DNA concentrations measured as described previously by us (21).

Polymerase chain reaction (PCR) analysis

Amplifications were performed with the GeneAmp PCR system 9700 from Perkin Elmer Applied Biosystems GmbH (Weiterstadt, Germany) at maximal ramp time. Analysis of the plasmid, phuMIF, was performed as described previously (29). For PCR amplification of a 90 bp region from the E.coli lac operon (20), a total reaction volume of 25 µl was used. Each reaction contained 1× reaction buffer, 200 µM of each desoxyribnonucleoside triphosphate, 0.2 µM of each primer, 8 mM magnesium chloride and 0.5 U Taq polymerase. To 20 µl reaction mixture, 5 µl lysate or phenol–chloroform purified DNA (22) were added. PCR was performed according to a simplified hot start protocol. Briefly, the reaction tubes were directly heated to 94°C and the temperature kept at 94°C for 5 min. Cycles were: 1 min at 94°C, 1 min at 65°C and 1 min at 72°C. The final extension was 10 min at 72°C.

Escherichia coli disintegration and λDNA fragmentation by sonication

Sonication was performed with a Branson cub horn sonifier (G. Heinemann Ultraschall- und Labortechnik, Schwäbisch Gmünd, Germany) with constant duty cycle at a temperature of ∼0°C. Escherichia coli suspensions diluted in PBS/EDTA were disintegrated with an acoustic energy output of ∼20%. For DNA fragmentation studies, a 35% output was employed (21).

RESULTS

Cell disintegration by low field strength: dependence on the number of electric pulses

The manipulation of cells by electric field-based methods has generally been performed in solutions of low ionic strength, thus requiring special compositions of electrolytes (30). Initial experiments using standard electric conditions in combination with solutions of low ionic strength indicated that under such conditions cell disintegration was possible (data not shown) (15,17). However, these experiments made it clear that a time-consuming and laborious adaptation of the cell suspension’s specific resistance was found to be required to achieve the high resistance, necessary for the application of high electric field strengths in the range of several kV/cm.

We thus investigated novel methods involving the disintegration of cells by pulsed electric fields in physiological buffers, i.e. in PBS/EDTA, with specific resistances of ∼60 Ωcm at 25°C. First, the initial electric field strength was varied (Fig. 1). Under the conditions chosen (see below), the maximally applied initial electric field strength was limited to ∼60 V/cm. Analysis of the solutes released by protein and DNA determinations showed that >30 V/cm were necessary to obtain significant release of cellular contents. We observed a linear dependence of cell disintegration on the initial field strength above this threshold value. Interestingly, different threshold values were noted when the various read-outs, i.e. absorption at 260 nm, protein concentration and DNA concentration, were compared (Fig. 1).

Figure 1.

Effect of the applied initial electric field strength. Escherichia coli suspensions were treated with 18 000 electric field pulses with a frequency of 5 Hz at 25°C and detected as indicated.

Significant cell disintegration as illustrated in Figure 1 was only observed when several thousand electric field pulses were applied (Fig. 2A and B). In fact, only above a value of about 5000 electric field pulses significant disintegration occurred. Previous methods involving electric field methods for cell disruption have generally relied on the use of much lower numbers of pulses, suggesting that the pulse number may be the critical parameter for efficient cell disintegration to occur. Dependence of cell disintegration on the number of pulses appeared hyperbolic and similar curves were obtained independent of whether protein or DNA was analyzed. A linear increase of intracellular compounds released was obtained up to 25 000 pulses, with a maximum quickly reached when the number of electric field pulses was increased further.

Figure 2.

Dependence of the degree of cell disintegration on the number of electric field pulses and/or treatment time. An initial electric field strength of 60 V/cm at 25°C was applied. (A, B, and D) Escherichia coli suspensions were treated with pulses of an initial electric field strength of 60 V/cm and frequencies between 0.5 and 50 Hz were applied. (C) Escherichia coli suspensions were subjected to different treatment times with a constant number of electric field pulses of ~10 000 (filled circles); for comparison, the last three values from (B) are shown (filled squares). The dependence of the absorption at 260 nm and the protein concentration were similar (data not shown). (D) Release of intracellular compounds as monitored by measuring the absorbance at 260 nm. Plot of the percentage of retained intracellular compounds using an exponential equation (1 – φ) = e^–kt. Inset of (D), semi-logarithmic presentation. Data represent the mean ± SEM of three to five determinations.

As the number of electric field pulses is the product of treatment time and frequency, the latter parameters had to be taken into account. Applying a constant number of electric field pulses, an optimal treatment time with a maximum of DNA released at ∼30 min was noted (Fig. 2C). A similar maximum value was measured when the released concentration of protein or the absorption at 260 nm were determined (data not shown). Of note, if both the treatment time and the number of electric field pulses were increased, the increase in DNA release continued for at least 2 h (Fig. 2C). It is likely, therefore, that the decrease in DNA yield was not due to DNA degradation instigated by the disintegration method applied but was rather due to cellular degradation processes.

Thus, we next sought to describe the dependence of cell disintegration on the various influences such as the number of electric field pulses applied, the treatment time used, and the frequency. No influence of frequency was observed (data not shown). However, when the release of intracellular compounds into the extracellular solution as measured by the absorbance at 260 nm was expressed as a fraction (φ) of the maximally released content in intracellular compounds (A_{260 nm}: 6 ± 0.5), a reasonable dependence was given by:

(1 – φ) = e^–kt

where t is expressed in min and is given by the product of the number of electric field pulses and the treatment time, and k is the DNA release constant in min^–1 (Fig. 2D) (31,32). Using this equation, electric cell disintegration could be described by a first-order kinetic model that depended both on treatment time and the number of electric field pulses within a certain range of a given initial electric field strength.

Enhancing effect of temperature

The temperature has been shown to affect electric effects, for example the breakdown voltage in electroporation can be decreased by raising the temperature (33). To achieve the breakdown voltage, a critical field strength has to be overcome. To electroporate E.coli, several kV/cm are necessary compared to the 30 V/cm found as the threshold value of the initial electric field strength applied in our study. Despite the use of considerably lower electric field strength, i.e. a subcritical initial field strength 20 V/cm was used, temperature-mediated effects were observed. The concentration of DNA released markedly increased when temperatures >50°C were applied (Fig. 3B). Measures of the absorption at 260 nm and the protein concentration both showed maximum curves with peak values seen at ∼50°C (Fig. 3A and C). This indicated that protein denaturation effects superimposed on the release of temperature-stable DNA molecules.

Figure 3.

Effect of temperature on electric cell disintegration. Escherichia coli suspensions were subjected to a subcritical initial electric field strength of ~20 V/cm for 6 min and 18 000 electric field pulses were applied. The release of intracellular components was monitored by the absorbance at 260 nm (A), the DNA concentration (B) and the protein concentration (C).

The specific resistance of the cell suspension rose by 80 ± 3 mΩcm/°C between 10 and 65°C. Thus, the time constant and the initial electric field strength at 10 and 65°C only differed by 0.1 and 6.5%, respectively. This demonstrated that the observed temperature dependence was not due to changes in the time constant or the initial electric field strength. It rather appears likely that the threshold value of the initial electric field strength is decreased by increasing temperature.

We also considered that the temperature, caused by the energy input of the large number of electric field pulses, could potentially have adverse effects. However, measurements showed that even without cooling of the RC, the maximal temperature increase was only ∼5°C due to the excellent heat conductivity of the RC. Nevertheless, the RC was thermostated in all experiments and increases in temperature were not detected (data not shown).

Electrochemical side effects

Although a system devised to prevent electrochemical and electrophoretic effects (18) was used, we observed electrophoresis after the application of 18 000 electric field pulses and a treatment time of 60 min, as detected by migration of bromophenolblue [0.25% (w/v)] in a 1% agarose gel poured into the RC (data not shown). Compared to the cell suspensions measured previously, the electric current was only half as high. It appeared therefore that electrochemical and electrophoretic effects would occur in the cell suspensions as well. The observation of minor but detectable electrode fouling supported this assumption. In addition, deposition of protein and DNA was observed. However, by rinsing the RC with PBS/EDTA solution, the deposited material could be fully recovered (Table 1). Hence, the electrophoretic effects seen were both minor and reversible, indicating that they did not interfere with cell disintegration and the release and isolation of intracellular compounds such as protein or DNA.

Table 1. Recovery of depositioned protein and DNA from the electrodes.

Concentration [µg/ml]	Lysate	1. Rinse	2. Rinse
Protein	15.5 ± 0.5	0.8 ± 0.6	–0.2 ± 0.9
DNA	1.34 ± 0.18	0.09 ± 0.09	0.004 ± 0.093

In a typical experiment, depositioned molecules could be readily eluted by rinsing the reaction chamber with buffer. In the experiment, 90 000 electric field pulses of ∼10 V/cm were applied over 30 min at room temperature.

Electric field-based method results in DNA of high integrity

Electrochemical reactions may lead to the formation of free radicals that may then act to destroy the purine or pyrimidine bases or the nucleic acid backbones (34,35). The application of electric fields may also have other adverse effects on cellular components.

To examine the integrity of DNA obtained by our method, we performed PCR and checked whether the released DNA was fragmented. Figure 4A shows that high molecular weight genomic DNA was obtained by the electric cell disintegration procedure. Identical results were seen for crude lysates and phenol chloroform-purified DNA. By contrast, other physical cell disintegration methods such as ultrasonication (Fig. 4C) or high pressure homogenization (36) result in fragmentation of DNA. Both the release and degradation of intracellular compounds by ultrasonication follow a first-order kinetic model (31). Thus, using such methods, the recovery of an intracellular compound has an optimum.

Figure 4.

Integrity of DNA released by the pulsed electric field method and comparison to sonicated DNA. (A) Electrophoretic analysis of genomic DNA from E.coli following electric field treatment: lane 1, 1 kb extension DNA marker; lane 2, crude lysate of an E.coli suspension treated with 360 000 electric field pulses at 60 V/cm and 25°C for 120 min. (B) PCR analysis of a phuMIF plasmid released from E.coli by the electric field method: lane 1, DNA molecular weight marker V; lane 2, PCR controls without template; lanes 3 and 4, incubation controls without electric field treatment at the beginning and the end of the experiment; lanes 5–12, application of 90, 300, 1500, 4500, 9000, 13 500 and 18 000 electric field pulses at 60 V/cm, a frequency of 5 Hz, and a temperature of 25°C. (C) Electrophoretic analysis of λDNA following a 30 min sonication treatment: lane 1, DNA molecular weight marker V; lane 2, fragmented λDNA. (D) DNA fluorescence of E.coli lysates following sonication treatment. Relative fluorescence units are comparable to the DNA concentration in µg/ml.

If the release of DNA is monitored by SYBR Green I fluorescence, fragmentation of high molecular weight DNA can be followed by a decrease in fluorescence (21). Such a behaviour was obtained when E.coli bacteria were disintegrated in a cup horn sonifier (Fig. 4D), but not when the electric cell disintegration method was employed (Fig. 2B).

PCR amplification of a 348 bp MIF cDNA sequence revealed that also plasmid DNA of a size of 6003 bp was efficiently recovered by the electric disintegration method (Fig. 4B). Importantly, this finding also demonstrated that treatment with the low electric field strength readily allowed for DNA amplification, thus indicating that the method could be combined with DNA diagnostic applications. The PCR analysis also confirmed that high numbers of electric field pulses did not lead to DNA fragmentation (Fig. 4B). On the contrary, and consistent with the data obtained in Figure 2B, the PCR product was found to increase in band intensity up to 18 000 pulses applied. It is implied by the data that electrochemical reactions, while occurring, did not result in destruction of the released plasmid in a way to prevent its amplification.

Amplification analysis was next applied to the 90 bp lac operon sequence of the E.coli genome. This DNA region was also readily amplifiable even after the application of 360 000 electric field pulses at 60 V/cm and a frequency of 50 Hz. Again, both phenol chloroform-purified samples or crude lysates were suitable for amplification (data not shown).

Disintegration of M.luteus and efficiency of the electric field method

The successful disintegration of the Gram-negative bacterium E.coli by electric field pulses led to the question, whether other organisms could be lysed as well. Thus we next investigated the Gram-positive bacterium M.luteus. Bacteria were exposed to 13 000 electric field pulses at 60 V/cm for 30 min at ∼55°C. This resulted in the release of 88 ± 14 µg/ml of protein and 5.9 ± 1.2 µg/ml of DNA, which was consistent with the values obtained for E.coli.

To further evaluate the efficiency of the electric cell disintegration procedure, DNA yields were compared with those obtained following lysis by the NucleoSpin C+T kit. For E.coli, the yield obtained by the chemical method was about 2.4-times higher than that of the electric cell disintegration procedure. Of note, the electric disintegration method was 2.1-times more efficient compared to the commercial kit when the Gram-positive bacteria were examined (Fig. 5).

Figure 5.

Efficiency of the electric cell disintegration procedure and comparison with standard chemical lysis procedures. The NucleoSpin C+T kit was used for chemical sample preparation according to the recommended protocols for Gram-negative** and Gram-positive bacteria*. Micrococcus luteus was disintegrated by ~13 000 electric field pulses at 60 V/cm for 30 min (55°C). For disintegration of E.coli, 18 000 electric field pulses with an initial electric field strength of ∼20 V/cm were applied for 6 min at around 70°C.

DISCUSSION

Electric field pulses can be applied for electroporation, gene transfer, electrofusion, electrorotation, dielectrophoresis and the killing of cells (30,37–43). Generally, extremely high electric field strengths and solutions of high specific resistance are necessary to perform these electric manipulations. Under the conditions of electroporation or the like, cells can be lysed (4,37), and low molecular weight compounds such as metabolites, soluble proteins, plasmids and others may be released (16,17,44,45).

We have shown that the disintegration of E.coli and M.luteus in a near-physiological solution of low specific resistance can be achieved by the application of electric field pulses of low strength. Various parameters were evaluated and the different cellular compounds released analyzed. Released soluble intracellular compounds, in particular soluble proteins and test plasmids, were monitored primarily to examine the effects of the electric pulses on membrane integrity, but also to evaluate total cellular disintegration. Completion of the latter aspect was confirmed by the release of high molecular weight genomic DNA. To precisely quantify cell disintegration, DNA concentrations were measured directly in the obtained E.coli lysates by SYBR Green I fluorescence (21). To prevent interference of the measurement with plasmid DNA, which may be released before complete cell disintegration occurs (16), a low copy plasmid was used. The plasmid employed, the pET11b vector carrying a phuMIF insert, contributed by an apparent value of ∼14 ± 8% to the total content of genomic DNA.

Variance of the initial electric field strength revealed different threshold values for the release of the more soluble intracellular compounds, i.e. soluble proteins, versus genomic DNA. Intracellular products were liberated at lower electric field strengths, probably by perforated or partly damaged cells. Complete cell disintegration required higher electric field strengths. As the system’s maximum initial electric field strength was limited to 60 V/cm, a further dependence could not be investigated in the model used.

Due to the application of electric field pulses of low strength, relatively large numbers of pulses were necessary to disintegrate the cells. The obtained correlation between cell disintegration and the number of electric field pulses appeared to follow a biphasic manner. The relatively high scattering of the values was not only a consequence of slight variances in cell concentrations, but was likely to also be the result of the different treatment times. The frequency studied over a range of 5–50 Hz did not appear to have an appreciable effect. Further studies will be necessary to determine the detailed interaction of the initial electric field strength and number of pulses.

The current study did not fully elucidate the mechanism of cell disintegration at low electric field strengths. One possible hint for the mechanism involved could be the observation that temperature increase supports cell disintegration. Similar effects have been observed by others when the temperature-mediated decrease of the breakdown voltage was investigated (33). Together with the obtained threshold values for the initial electric field strength, a mechanism comparable to that of electroporation, especially the secondary effects, could be plausible (46–48). Such a phenomenon would then be described by the release of soluble low molecular weight compounds through relatively small pores and the release of high molecular weight compounds through large pores or as a consequence of complete cell disintegration (49). This notion is supported further by the finding that stationary and very small pore sizes at subcritical electric field strengths play a role in electroporation (50). On the other hand, several thousand electric field pulses were necessary to disintegrate cells as compared to the low numbers of pulses applied in electroporation or for the killing of bacteria (37,51).

As a PBS-based saline solution of isoosmotic strength was used, osmolytic effects (37,52) can be neglected in our procedure. Furthermore, no sparc discharges were observed; thus shock waves comparable to ultrasound waves are unlikely to account for the obtained efficient cell disintegration (53). If this mechanism was responsible for cell disintegration, released DNA would have been fragmented (31,36,54).

Electrochemical and electrophoretic effects may have contributed to cell disintegration (18,51). However, the traces of DNA and protein that were found to be depositioned on the electrodes could easily be rinsed off. Also, the released DNA was not affected by electrochemical reactions as best demonstrated by the achieved integrity of the high molecular weight genomic DNA. Most importantly, the released DNA could be readily amplified by PCR, indicating that the DNA should be suitable for a range of molecular biology detection techniques that involve but should not be limited to hybridization-type diagnostic assays.

Of note, as PBS/EDTA solutions were used throughout the study, it should be possible to apply the electric field method to various other solutions with relatively low specific resistance. This implies that the established procedure should be suitable for sample preparation purposes of a large variety of biological fluids. Moreover, a combination of this novel procedure with chemical methods offers to apply the method to issues of isolating RNA and other sensitive intracellular compounds. Alternatively, a combined electric field-chemical method could simply serve to markedly accelerate the cell disintegration step for a variety of applications.

The electric disintegration of very small tissue samples, like needle biopsy material, may be possible by electromechanical deformation (48). However, for larger tissue samples a combination of electric disintegration and mechanical homogenization procedures will be necessary.

Variations in the time constant, the pulse form and the frequency carry the potential of further improving the electric cell disintegration procedure and, based on our results, appear worthwhile of being studied in greater detail (55,56). The universality of the method will have to be further verified by examining the disintegration other organisms such as fungi, plant cells or others.