Ligation mediated PCR performed at low denaturation temperatures—PCR melting profiles

Abstract

We show that using low denaturation temperatures (80–88°C) during ligation mediated PCR (LM PCR) of bacterial DNA leads to the amplification of limited sets of the less stable DNA fragments. A set of electrophoretic patterns of such fragments obtained at different denaturation temperatures forms the PCR melting profile (PCR MP). A single pattern obtained for a given temperature and a set of patterns arising after application of several denaturation temperatures (PCR MP) are very specific for the given bacterial genome and may be used for strain characterisation and differentiation. The method may also be used for amplification and isolation of the less stable DNA fragments in a genome.

INTRODUCTION

Melting temperatures of genomic DNA fragments obtained by digestion of restriction nuclease depends on their GC content and length. It is known that intervals of temperatures corresponding to the full transition from double- to single-stranded structure for a restriction DNA fragment may vary from ∼0.5°C to several degrees depending on GC content distribution along the DNA fragment (1). From DNA sequencing projects it is known that bacterial genomes show substantial genomic heterogeneity in GC content (2,3). It means that after digestion with restriction nuclease of bacterial genomic DNA, fragments, which arise as a result of the digestion, have different thermal stabilities. This feature may be used to obtain sets of electrophoretic patterns of DNA fragments amplified during ligation mediated PCR (LM PCR) (4) performed at various denaturation temperatures—PCR melting profiles (PCR MPs). PCR MPs are characteristic for a given genome and restriction nuclease applied for digestion. During PCR, denaturation temperatures (T_d) around 94–95°C are applied with the aim to achieve full separation of DNA strands of all DNA fragments present in the amplified sample. Under such conditions, during LM PCR, all DNA fragments in the sample should be amplified. Lowering of T_d applied during LM PCR should decrease the number of amplified fragments because only single-stranded DNA molecules may serve as a template during LM PCR. In this work, we show that using low T_d during LM PCR leads to limited and specific amplification of a small number of the less stable DNA fragments. The electrophoretic patterns of DNA fragments obtained after such amplifications are characteristic for the bacterial strain taken for DNA isolation. The set of electrophoretic patterns (PCR MP) obtained at different T_d values for a given bacterial strain is a very specific fingerprint which may be used for strain characterisation and differentiation because the order of appearance of amplified DNA fragments in subsequent increasing denaturation temperatures (T_d) is stable for a given genomic DNA. The outline of the proposed method is shown in Figure 1. The application of thermal stability of the DNA restriction fragments as the selecting factor for the creation of limited DNA representations seems to be a completely new approach to the target template selection in restriction site polymorphism based PCR fingerprinting methods. Techniques such as AFLP (5) and other LM PCR based methods (6,7) rely on the mutual distribution of two or more relatively short DNA sequences within the distance enabling efficient PCR amplification. In contrast to the above mentioned methods, PCR MP does not rely exclusively on the distribution of two short DNA sequences within the PCR amplification range, because from the pool of DNA fragments of similar length only the less stable ones are amplified. For this reason, PCR MP allows application of single, relatively frequent cutter restriction nuclease without defining any portion of the DNA sequence adjacent to a restriction site.

Figure 1.

Principle of the method. The thick black line represents a melting curve of a genomic DNA. Vertical and horizontal axes represent the percentage of ssDNA and denaturation temperature (T_d), respectively. Range a: sequences potentially amplifiable by PCR are not present—all DNA fragments remain double-stranded and primers cannot bind. Range b: some DNA particles are single-stranded (grey shaded) and primers can bind—PCR amplification occurs. Range c: majority of DNA particles are ssDNA thus are amplifiable, and at the end of the range all DNA fragments are ssDNA—and are amplifiable. The working range is from the right part of range a to the left part of range c and is limited by the lowest T_d where a PCR occurs and the T_d where the number of PCR products reaches the border of the resolution capacity of polyacrylamide gels, respectively. PCR primers are of equal length as terminal complementary sequences within every single strand and compared to ligated oligonucleotides are extended by a protruding end sequence and the remaining part of the restriction site on their 5′ ends. Terminal complementary sequences are created by filling in ssDNA termini obtained by ligation of oligonucleotides to protruding 5′ ends of restriction fragments.

MATERIALS AND METHODS

Bacterial strains

The following bacterial strains were used: Escherichia coli: K-12 MG 1655 (3), OM12, OM72, OM16, Klebsiella oxytoca OM 36, 22 and Klebsiella pneumoniae OM 42 and OM 86. Strains designated OM were isolated from patients in one hospital and all of them bear the INT37CZD integron (accession number AF282595).

Converting entire genome to amplicons and melting selective amplification

From 3 ml overnight cultures of bacterial strains, 1.5 ml was centrifuged at 1200 g for 2 min and DNA was isolated from bacterial pellets using a protocol originally applied to DNA isolation from blood, urine and cerebrospinal fluid (8), the version with diatom suspension, only the last step of the procedure was modified: 150 µl of deionised water was added to elute DNA after the isolation procedure; 1 µl of DNA solution was electrophoresed on 0.8% (w/v) agarose gels with TAE buffer and subsequently stained in 0.5 mg/l ethidium bromide solution for 10 min, DNA concentration was estimated by comparison to reference DNA. The DNA concentration was from ∼100 ng/µl to several hundred ng/µl in distinct samples. Genomic DNA was digested with HindIII (10 U/µl). Digestion reactions were performed under uniform conditions; 500 ng of DNA was dissolved in 30 µl buffer H (50 mM Tris–HCl, pH 7.5, 10 mM MgCl₂, 100 mM KAc, 1 mM DTE; Roche Molecular Biochemicals), 1 µl of the endonuclease was added, after 2 h another 1 µl of the endonuclease was added and the reaction was continued for 2 h at 37°C. After digestion, the volume was adjusted to 100 µl with TE, 100 µl of phenol pH 7.8 (9) was added, samples were vortex mixed and centrifuged for 5 min at 12 000 g. After centrifugation, 100 µl of aqueous phase was transferred into new Eppendorf tubes and equal volumes of chloroform–isoamyl alcohol (24:1) were added. Samples were vortex mixed and centrifuged for 5 min at 12 000 g. Chloroform–isoamyl alcohol extraction was repeated and the aqueous phase was transferred into new tubes. Subsequently, 15 µl of 3 M sodium acetate and 3 µl of glycogene 20 mg/ml were added, samples were mixed and 300 µl of 96% ethanol was added. The samples were mixed again and incubated for 10 min on dry ice and centrifuged for 10 min at 12 000 g. The pellets were washed with 200 µl of 75% ethanol for 5 min, then centrifuged for 3 min at 12 000 g. Dry pellets of digested DNA were dissolved in ligation mix comprising: two oligonucleotides forming an adaptor, POWIE 5′-CTCACTCTCACCAACGTCGAC-3′ and HINLIG 5′-AGCTGTCGACGTTGG-3′, 20 pmol each, in total volume of 20 µl ligation buffer (66 mM Tris–HCl, pH 8.5, 6.6 mM MgCl₂, 10 mM DTT, 66 mM ATP; Amersham Pharmacia Biotech). The mixture was heated in a water bath for 2 min at 56°C and cooled for 10 min at room temperature. Subsequently, 1 µl of T4 DNA ligase 1 U/µl was added and the samples were incubated overnight at 16°C.

The PCR mixture contained: 50 pmol of POWAGCTT primer 5′-CTCACTCTCACCAACGTCGACAGCTT-3′, 100 µmol each dNTPs and 1 µl of ligation reaction products in a total volume of 50 µl of PCR buffer (50 mM KCl, 20 mM Tris pH 8.9, 2.5 mM MgCl₂, 0.01% gelatine). In an Eppendorf thermal cycler (Mastercycler® 5330) PCRs were performed as follows: 2 min at 72°C to release unligated oligonucleotides HINLIG, subsequently 1 µl 2.5 U/µl of Taq DNA polymerase (10) was added and mixed with a pipette tip. The reaction mix was incubated for an additional 5 min at 72°C to fill in the single-stranded ends and create amplicons, followed by initial denaturation at (80 + x)°C for 1 min and 21 cycles of denaturation at (80 + x)°C for 40 s, annealing and elongation at 72°C for 2 min (where x is a digit between 0 and 8 and is constant for a single PCR reaction). After the last cycle, samples were incubated for 5 min at 72°C. PCRs were performed in 500 µl tubes and the reaction mix was covered with mineral oil. Primer annealing temperatures were calculated with the Premier Primer on-line applet and those temperatures were used as the starting point for the annealing temperature optimisation. Alternatively, PCR reactions were performed in a gradient thermal cycler (MJ Research PTC200 DNA Engine Gradient) with a gradient range of 84–88°C in 200 µl thin walled PCR tubes with PCR mixtures as described above. Cycler settings: sample type was set to tubes and sample volume to 75 µl (50 µl reaction mixture and 25 µl oil) and the heating lid was disabled. PCR reactions were performed as follows: 2 min at 72°C, subsequently 1 µl 2.5 U/µl of Taq DNA polymerase was added and mixed with a pipette tip. The reaction mix was incubated at 72°C for an additional 5 min followed by initial denaturation at a gradient across the thermal block of 84–88°C for 1 min 30 s and 21 cycles of denaturation at a gradient across the thermal block of 84–88°C for 1 min, annealing and elongation at 72°C for 2 min 15 s and 5 min at 72°C after the last cycle. Identical experiments were performed at gradients of 85–87°C and 88–95°C. Reproducibility was tested in independent runs of the cyclers. By independent runs, we mean two completely independent experiments starting from DNA isolation. The runs were performed in the second and fourth rows of the gradient thermal cycler block. In the Eppendorf Mastercycler® 5330, all the runs were performed in the same position of one thermal block.

Gel electrophoresis of PCR products

PCR products, 8 µl out of 50 µl, were electrophoresed on 6% polyacrylamide gels (60:1 acrylamide/N,N-methylene bis-acrylamide) with TAE buffer, stained in ethidium bromide 0.5 mg/l aqueous solution for 10–15 min. Images of the gels were made using a White/Ultraviolet Transilluminator (UVP).

RESULTS AND DISCUSSION

PCR MP for E.coli strain K-12 MG1655 is shown in Figure 2. The number of DNA fragments amplified during LM PCR depends on T_d used in the denaturation step of the amplification reaction. The comparison of the results of two experiments shows that there are small differences between the independent runs. The reason for this may be fluctuations of temperature and salt concentration during independent experiments. Analysis of many patterns like that shown in Figure 3 leads to the conclusion that the most stable results for E.coli strains are obtained when T_d is in the range 87–88°C. The resolution power of PCR MP patterns for bacterial strain differentiation is illustrated in Figures 3 and 4. In Figure 3, complete PCR MPs for two E.coli strains are compared. Differences between the strains are visible for every T_d applied to obtain the profiles. In Figure 4, it is seen that a very efficient strain differentiation may also be carried out at a single T_d (88°C). In this case, different strains of E.coli and Klebsiella were compared.

Figure 2.

PCR MPs of E.coli K-12 MG1655 DNA. PCR products from two independent experiments (as defined in Materials and Methods) were loaded to adjacent wells. Pairs of adjacent lanes are described with T_d applied in PCR. Lane M—DNA molecular weight marker; fragment lengths in bp are indicated. (a) Eppendorf Mastercycler 5330 and (b) PTC-200 MJ Research cycler were used for amplification. In the case of the Eppendorf cycler, 18 independent experiments were carried out. In the case of (b), a T_d gradient of 84–88°C was used in two independent experiments.

Figure 3.

Comparison of PCR MPs of two E.coli strains. (A) Clinical strain OM12 and (B) K-12 MG1655. T_d gradients of 84–88°C were performed using the PTC-200 MJ Research cycler. Lanes M, DNA molecular weight marker; fragment lengths in bp are indicated.

Figure 4.

Comparison of electrophoretic patterns obtained after LM PCR for different E.coli and Klebsiella strains carried out at a single T_d = 88°C. Two independent experiments were performed using the PTC-200 MJ Research cycler. Lanes M, DNA molecular weight marker; fragment lengths in bp are indicated; the remaining lanes are described with OM series strain numbers as follows: K.pneumoniae (42, 86), K.oxytoca (36, 22), E.coli clinical strains (16, 72, 12) and E.coli K-12 MG1655 (MG).

Optimal gradient range on a gradient cycler and reproducibility of the results

The optimal gradient range should be established experimentally for the strains subjected to analysis and using a particular thermal cycler. We consider the gradient range 84–88°C to be optimal for differentiation of the analysed Enterobacteriaceae strains, because it ensures good reproducibility and satisfactory differentiation efficiency. A steady increase in the number of amplified DNA fragments, which is dependent on T_d increase, is observed (Figs 2b and 5a). Surprisingly, a relatively high rate of reproducibility is obtained for the T_d interval equal to 2°C between 85 and 87°C (Fig. 5b) despite the fact that the differences between T_d in adjacent tubes are smaller than the accuracy of the cycler which is supposed to be ±0.4°C. PCR MPs with a T_d upper limit equal to 95°C are not very useful for bacteria fingerprinting because of high background and lower reproducibility. The upper temperature gradient range limit, where reproducible results are obtained, may depend on the particular cycler properties. The results obtained at T_d 88.0°C with the first PTC-200 gradient thermal cycler tested (Fig. 2b) differ by two DNA bands from the results obtained at T_d 88.0°C with the second cycler (Fig. 5a). However, those two missing bands occurred at a T_d gradient of 88–95°C at T_d 88.6°C in the experiment performed with the second cycler (Fig. 6a). The shortest DNA fragment amplified in PCR conducted with the second MJ Research PTC-200 gradient cycler at T_d 85.8°C (Fig. 5a), occurs at T_d 85.2°C in PCR MP performed at the same T_d gradient (84–88°C) on the first of the tested cyclers (Fig. 2b). Variability of the results obtained with two separate PTC-200 cyclers probably resulted from a shift in calibration by at least 0.6°C.

Figure 5.

Comparison of PCR MPs of E.coli MG 1655 performed at a temperature gradient of 84–88°C (a) and 85–87°C (b). PCR products from two independent experiments (as defined in Materials and Methods) were loaded to adjacent wells. Pairs of adjacent lanes are described with T_d applied. Instability occurs only at 87.0°C at a temperature gradient of 84–88°C (a), and in the gradient range 85–87°C at temperatures: 85.6, 86.2, 87.0°C (b). Results were obtained on another MJ Research PTC-200 gradient cycler unit than those presented in Figures 2b and 3B. Some DNA bands are missing on (a) lanes 88.0°C compared to Figures 2b and 3B, lanes 88.0°C; however, they occur when higher T_d values are applied (Fig. 6a, lane 88.6°C).

Figure 6.

Comparison of PCR MPs of E.coli MG1655 performed at a temperature gradient of 88–95°C in two independent experiments.

On the other hand, for some DNA fragments, the transition temperature from dsDNA to ssDNA may be as low as 0.5°C (1). The temperature of the thermal block wells may differ by 0.4°C from the nominal temperature. For this reason, some DNA fragments do not melt in every cycle of PCR. For example in PCR MP performed at a gradient of 84–88°C, with the first PTC-200 cycler, at T_d 87.8°C one unstable band occurred which did not occur at T_d 87.4°C but was amplified at 88.0°C (Fig. 2B). The most probable causes of this phenomenon were temperature fluctuations within the cycler accuracy range. For the reasons presented above, some instability of the results should be expected at particular T_d values. T_d values where high rates of reproducibility are obtained must be determined experimentally and it is improbable to establish a gradient where the reproducibility of the results would be always complete at all T_d values. The second and fourth rows of the thermal block were tested and the results obtained at defined T_d values in independent runs did not depend on the thermal block row used. It must be emphasized that the order of appearance of DNA fragments along with an increase in T_d values is always the same and does not depend on the thermal cycler used (Figs 2b and 5a). Thus the order of appearance of DNA bands in PCR performed in subsequent increasing temperatures is invariable for a given genomic DNA and should be treated as the most important criteria for the interpretation of results. Reproducibility defined as complete repetitiveness of the results at any T_d is not achieved, as described above, but reproducibility of the results defined as the order of appearance of DNA bands is complete for both cyclers tested.

We have not tested how polymerases from different sources influence the results.

Stability of the results for an old type cycler

We obtained the best reproducibility in particular T_d values applied when independent experiments were performed in the same position of the thermal block (Fig. 2a) and more variable results were obtained at individual T_d values when distant wells were used (results not shown). Measurements performed in Eppendorf instruments revealed that temperatures in distant wells within the same thermal block can differ by about 1°C at 72 and 96°C. It should be emphasised that the Eppendorf Mastercycler® 5330 is an old type cycler, with an accuracy range of +0.8/–1.2 at 94°C and ±0.6 at 72°C. However, DNA fragments which were not amplified repetitively at a particular T_d, were amplified in the T_d higher by 1°C (Fig. 2a). Thus, the accuracy of the thermal cycler (T_d fluctuations) was the most probable source of variable results in some T_d values. Nevertheless, the order of appearance of the DNA bands along with the increase in T_d is stable, which enables interpretation of the results obtained even on an old type cycler with relatively low accuracy.

CONCLUSIONS

The proposed method allows specific gradual amplification of the genomic DNA differing in the thermal stability starting from the less stable DNA fragments amplified at lower T_d values to more stable ones amplified at higher T_d values.

The PCR MPs are very specific fingerprints for bacterial strains and may be used for strain differentiation.

PCR MPs may be obtained very conveniently with PCR gradient cyclers with the possibility of T_d programming across a heating block.

The order of appearance of DNA fragments at subsequent, increasing T_d values obviously depends on the investigated DNA properties and is characteristic of a given genomic DNA; furthermore, the order does not depend on the cycler used.

PCR primers should be longer (extended on their 5′ ends by a protruding end sequence and the remaining part of the restriction site) compared to ligated oligonucleotides, in order to overcome the PCR suppression phenomenon (11).