Department of Microbial Genetics, University of Tübingen, Germany.
Listeria monocytogenes EGD lacking penicillin-binding protein 5 (PBP5) produces a thicker cell wall
Article first published online: 9 JAN 2006
FEMS Microbiology Letters
Volume 251, Issue 2, pages 281–288, October 2005
How to Cite
Korsak, D., Vollmer, W. and Markiewicz, Z. (2005), Listeria monocytogenes EGD lacking penicillin-binding protein 5 (PBP5) produces a thicker cell wall. FEMS Microbiology Letters, 251: 281–288. doi: 10.1016/j.femsle.2005.08.009
Edited by Dr. R.A. Bonomo
- Issue published online: 9 JAN 2006
- Article first published online: 9 JAN 2006
- Received 11 February 2005, Revised 8 June 2005, Accepted 9 August 2005
- Listeria monocytogenes EGD;
- Penicillin-binding proteins;
We report on the cloning of the structural gene for penicillin-binding protein 5 (PBP5), lmo2754. We also describe the enzymatic activity of PBP5 and characterize a mutant lacking this activity. Purified PBP5 has dd-carboxypeptidase activity, removing the terminal d-alanine residue from murein pentapeptide side chains. It shows higher activity against low molecular weight monomeric pentapeptide substrates compared to dimeric pentapeptide compound. Similarly, PBP5 preferentially cleaves monomeric pentapeptides present in high-molecular weight murein sacculi. A Listeria monocytogenes mutant lacking functional PBP5 was constructed. Cells of the mutant are viable, showing that the protein is dispensable for growth, but grow slower and have thickened cell walls.
Listeria monocytogenes is an ubiquitous gram-positive opportunistic pathogen that may cause very serious food-borne infections in humans, which are particularly severe for newborns and immunocompromised individuals [1,2].
In general, L. monocytogenes is susceptible to a wide range of antibiotics. However, some strains are resistant to amdinocillin and methicillin and L. monocytogenes in general has a relatively high natural resistance to monobactams and some cephalosporins [3,4]. The treatment of choice for L. monocytogenes infections is an aminopenicillin or imipenem alone, or in combination with an aminoglycoside [5,6]. There are very few reports on the emergence in the clinical setting of L. monocytogenes strains resistant to β-lactams used for treatment [7,8]. Independent studies with laboratory mutants resistant to imipenem or dicloxacillin [9–11] showed that the primary lethal target of β-lactams in L. monocytogenes is PBP3 and that resistance is due to the reduced affinity of this protein to the β-lactam.
However, neither PBP3 nor any of the four other membrane-bound PBPs of L. monocytogenes[12–14] have been studied in detail. Moreover, the action of β-lactam antibiotics against L. monocytogenes, at least in laboratory conditions, is often bacteriostatic – loss of viability without cell lysis occurs after prolonged incubation with the drug. In view of the above, we have launched an effort to characterize the individual PBPs as well as other putative PBPs coded for by the L. monocytogenes genome. Herein, we report on the cloning of the structural gene for PBP5, lmo2754, characterization of the activity of its product and properties of a mutant lacking the enzyme.
2Materials and methods
2.1Strains, plasmids, and growth conditions
Listeria monocytogenes EGD was grown at 37 °C in Tryptic Soy Broth (TSB) medium (Difco Laboratories). Cells transformed with plasmid derived from pAUL-A were grown at 30 °C in TSB medium supplemented with 3 μg erythromycin ml−1. Integration of the plasmid was achieved by growing transformed strains at 42 °C, a temperature restrictive for the replication of plasmid pAUL-A. Escherichia coli XL-1-Blue (Stratagene) was grown aerobically at 37 °C in LB medium. E. coli XL-1-Blue transformed with plasmid derived from pBAD/Myc-HisB was grown in LB medium supplemented with 50 μg ampicillin ml−1 to maintain the plasmid. If necessary, glucose (0.2%) was added to inhibit expression from pBAD. E. coli XL-1-Blue cells transformed with a derivative of pAUL-A, which carries an erythromycin resistance gene, were grown in LB medium containing 300 μg erythromycin ml−1, as described elsewhere. E. coli D456 was grown in LB medium with 30 μg kanamycin ml−1 and 25 μg spectinomycin ml−1.
2.2DNA manipulations and PCR
PCR was performed using MJ Research PTC-200 (Biozym, Germany) machine using Powerscript or PanScript polymerase from Pan Systems, Germany. Standard techniques were used for manipulation of plasmid and genomic DNA. E. coli was transformed using modified CaCl2 procedure. L. monocytogenes was transformed as described [18,19]. Restriction endonucleases and other DNA-modifying enzymes were purchased from commercial sources and used according to the recommendations of the manufacturers.
2.3Construction of insertion mutant
Primers LPWdk 5′-CCGGTAATAACCAGTTCCAG-3′ and PPW 5′-AGACGCTGACAAGCTAATGG-3′ were used to amplify an internal 564 bp fragment of lmo2754 from genomic DNA of L. monocytogenes. The amplicon was cloned into the vector pAUL-A, which was prepared by digestion with Sma I. After chemical transformation (CaCl2) of the ligation mixture into E. coli cells, the recombinants were identified as Lac− colonies. The resulting plasmid pAUL-A::lmo2754 was transformed into wild-type L. monocytogenes.
2.4Construction of an inducible expression system
Cloning and expression experiments were performed with E. coli XLI-Blue and harbouring the vector pBAD/Myc-HisB (Invitrogen). Gene lmo2754 was amplified from genomic DNA of L. monocytogenes EGD. Appropriate restriction sites (Nco I and Eco R1, underlined) were inserted by using the following primers: PBP5-up 5′-GTAACCATGGCAGCATCCCTTGCAGTAAG-3′ and PBP5-downStop 5′-CCCGAATTCTTAATTAAACCACCCTTTTACACC-3′. The PCR product was digested with Nco I and Eco R1 and ligated into pBAD/Myc-HisB treated with the same enzymes. The resulting plasmid pDW5 was transformed into E. coli XL1-Blue.
2.5Preparation of cell extracts and purification of PBP5
For the purification of PBP5, E. coli XLI-Blue harbouring pDW5 was grown at 37 °C in 1800 ml of LB medium. Overexpression of the gene encoding PBP5 was induced by the addition of 0.1%l-arabinose when an absorbance at 578 nm (A578) of 0.5–0.6 was reached. Growth was continued for another 90 min. Cells were cooled down and harvested by centrifugation (7000g, 10 min, 4 °C). All steps were performed at 4 °C. The pellet was resuspended in 10 mM Tris–maleate buffer, pH 6.8, 10 mM MgCl2. After adding 10 μg DNase ml−1 and phenylmethanelsulfonyl fluoride (1 mM), the cells were broken in a French pressure cell (American Instrument Company) at 82,800 kPa. Cell envelopes were collected by centrifugation 100,000g, 1 h. The pellet was resuspended in 10 mM Tris–maleate, pH 6.8, 10 mM MgCl2, containing 1 M NaCl and stirred for 2 h. Membranes were spun down (see above) and proteins were extracted by resuspending the membranes in Triton X-100 extraction buffer (10 mM Tris–maleate, pH 6.8, 10 mM MgCl2, 150 mM NaCl, 2% Triton X-100) and stirring for 12 h. Unsolubilized membrane debris was removed by centrifugation (100,000g, 1 h). The Triton X-100 extract was dialyzed against 10 mM Tris–maleate, pH 4.7, 10 mM MgCl2, 50 mM NaCl, 0.02% NaN3 (buffer I). Based on the properties of the protein as established by computer analysis, a column containing 10 ml of CM-Sepharose (Amersham Pharmacia Biotech) was equilibrated under low pressure chromatography conditions with buffer I containing 1% Triton X-100. The sample (20 ml of Triton X-100 extract, above) was applied at a flow rate of 24 ml h−1. The column was washed with 60 ml of buffer I containing 0.2% Triton X-100. The proteins were eluted with linear salt gradient from buffer I to 1 M NaCl in the same buffer. Fractions 6–9 (3.7 ml each) containing PBP5 were combined and dialyzed against 10 mM Tris–maleate, pH 6.8, 10 mM MgCl2, 500 mM NaCl, 0.02 % NaN3. The yield was 5.6 mg of purified PBP5 present in a pool of four fractions.
2.6Penicillin-binding assay with BioAmp
Fractions containing pure PBP were pooled and dialyzed against 10 mM Tris–maleate buffer, pH 6.8, containing 10 mM MgCl2 and 500 mM NaCl. Two microlitres of this pool of purified PBP5 (0.5 mg ml−1) were incubated with 4 μl biotynylated ampicillin (BioAmp) prepared by us as described, at 37 °C for 30 min, and 1 μl of 10 mg penicillin G ml−1 was added. The samples were incubated at 37 °C for 15 min, then the reaction was stopped by addition of an equal volume of sample buffer and boiling for 5 min. The proteins were separated by SDS–10% polyacrylamide gel electrophoresis (PAGE), followed by transfer to fluorotrans membrane 0.2 μm (Pall Europe Limited). PBPs were detected by the ECL method using streptavidin-horseradish peroxidase (Amersham Biosciences). In experiments with a mutant of L. monocytogenes lacking functional PBP5, the absence of the protein was shown by autoradiography after electrophoresis and autoradiography as described.
The soluble precursor UDP-MurN Ac-l-Ala-d-Glu-m-DAP-d-Ala-d-Ala (UDP-MurN Ac-PP) was isolated from Bacillus cereus as described and used as a substrate in the dd-carboxypeptidase assay. Ten microlitres of purified PBP5 solution (0.5 mg ml−1) were incubated with 5 μl of substrate (2.5 mM) in a total volume of 100 μl in sodium acetate buffer, pH 5.2, at 37 °C for 0.5 h. The reaction was stopped by the addition of 5 μl of 20% phosphoric acid and boiling the sample for 5 min. Precipitated material was removed by centrifugation. Prior to chromatography, 20 μl of 0.05% trifluoroacetic acid (TFA) was added to the supernatant. The samples were applied to a 125 × 4.6 mm column prepacked with 5 μm ODS Hypersil (octadecyl silica), (Bischoff, Germany) and elution was at room temperature with a flow rate of 1 ml min−1, using a linear gradient of 1.2–12% acetonitrile in 0.05% (TFA) in H2O, that was built up in 20 min. This was followed by washing the column with 60% acetonitrile solution in 0.05% TFA acid for 7.5 min. The eluent was monitored in UV at 205 nm. As tested with standard compounds, UDP-MurN Ac-pentapeptide eluted at 22.9 min and was separated from UDP-MurN Ac-tetrapeptide (retention time 21.4 min) (data not shown).
2.8pH dependence of PBP5 activity
Ten microlitres of purified PBP5 (0.5 mg ml−1) were incubated at 37 °C for 0.5 h with 5 μl of the substrate (2.5 mM) and 85 μl of the following buffers: (A) 100 mM Tris–HCl, pH 8.0, 0.05% Triton X-100; (B) 10 mM Tris–maleate, pH 6.0, 10 mM MgCl2, 50 mM NaCl, 0.05% Triton X-100; (C) 10 mM NaAc, pH 5.2, 50 mM NaCl, 0.05% Triton X-100; (D) 10 mM NaAc, pH 4.5, 50 mM NaCl, 0.05% Triton X-100; (E) 10 mM NaAc, pH 4.0, 50 mM NaCl, 0.05% Triton X-100; (F) 10 mM NaAc, pH 3.5, 50 mM NaCl, 0.05% Triton X-100; (G) 10 mM NaAc, pH 3.0, 50 mM NaCl, 0.05% Triton X-100. The reaction was terminated as described above.
2.9Inhibition of dd-carboxypeptidase activity by benzylpenicillin
Ten microlitres of purified protein PBP5 (0.5 mg ml−1) was incubated at 37 °C with 2 μl of penicillin G (10 mg ml−1) in buffer containing 10 mM sodium acetate, pH 5.2, 50 mM NaCl, 0.05% Triton X-100 for 10 min. Five microlitres of substrate UDP-MurN Ac-PP (2.5 mM) was added and the sample was incubated for 0.5 h at 37 °C.
2.10Activity of PBP5 on E. coli muropeptides and murein sacculi
Strain D456 was used because its murein contains a high percentage of pentapeptide side chains, the likely substrate for dd-carboxypeptidases. When an A578 of 0.6 was reached, the D456 culture was rapidly cooled to 4 °C, and the cells were harvested by centrifugation (7000g, 10 min, 4 °C). Murein sacculi were isolated as described previously. For the preparation of muropeptides, 200 μl of murein suspension was digested overnight with 2 μg Cellosyl (Hoechst) ml−1 in 20 mM phosphate buffer, pH 4.8. The reaction was stopped by boiling for 5 min. The sample was divided into two aliquots. One was reduced with sodium borohydride and the other (100 μl) was incubated for 3 h at 37 °C with 5 μl of PBP5 (0.5 mg ml−1). The reaction was terminated by boiling. The sample was reduced with sodium borohydride and fractionated by reversed-phase HPLC. In a second experiment, the high-molecular weight murein from E. coli D456 was tested as a substrate for PBP5. For this, 200 μl of a suspension of murein sacculi (2 mg ml−1) were incubated for 3 h at 37 °C with 20 μl of the purified protein (0.5 mg ml−1). The incubation was terminated by the addition of SDS (final concentration 1%) and boiling. SDS was removed by washing the sacculi with water and ultracentrifugation (90,000g, 20 min). The sacculi were incubated overnight with Cellosyl as described above. After reduction with sodium borohydride the resulting muropeptides were fractionated by reversed-phase HPLC.
2.11Preparation of samples for transmission electron microscopy
Samples were prepared from cells of wild-type L. monocytogenes and mutant DA01 in logarithmic phase of culture growth at 37 °C. The cells were spun down and resuspended in 1 ml solution of fixative (1.5% solution of glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2) and fixed for 12 h at room temperature (RT). The cells were spun down again, washed free from fixative and resuspended in 0.1 M cacodylate buffer and incubated for 30 min at RT. This step was repeated 4 times prior to fixing the cells in OsO4 solution for 2 h. OsO4 was removed by spinning down the cells and re-suspending them in cacodylate buffer. After the final spin the cells were left for 2 h in 0.1 M cacodylate buffer. The cell suspension was dehydrated using a graded series of increasing ethanol solutions: 70% (2 × 10 min), 80% (1 × 10 min), 96% (1 × 10 min), absolute ethanol (2 × 10 min), followed by acetone (2 × 5 min). The dehydrated material was embedded in Epon 812. Thin sections were prepared from the fixed material using Tesla microtome. Sections with diameter from 20–30 to 60–80 nm were stained with uranyl acetate and lead citrate. The samples were viewed under Leo 1912AB transmission microscope.
Analysis of the complete genome sequence of L. monocytogenes EGD (GenBank AL591824) allowed us to identify a DNA segment with similarity to genes coding for low molecular weight (lmw) PBPs or dd-carboxypeptidases from other bacterial species. Gene lmo2754 encodes a complete protein of 48,000 Da. PBP5 has two predicted start codons, GTG and ATG that are preceded by a theoretical ribosome binding site. Therefore, it is possible that the cell produces two variants of PBP5 that differ in their N-terminus and that are not resolved by SDS–PAGE. The gene cloned by us begins with the methionine and our sequence analysis concerns the mature form of the protein without the signal sequence.
3.1Cloning of gene lmo2754 into pBAD/Myc-HisB
An inducible expression system to overproduce PBP5 for purification was constructed on the basis of pBAD/Myc-HisB vector. Gene lmo2754 was amplified from genomic DNA of L. monocytogenes EGD by PCR using the oligonucleotides: PBP5-up and PBP5-downStop as primers. By inserting the lmo2754 gene into the Nco I and Eco R1 restriction sites of pBAD/Myc-HisB, native PBP5 without a hexahistidine tag was produced. The final construct was named pDW5.
3.2Purification of PBP5
The addition of 0.1% arabinose to the growth medium induced the high level production of PBP5 in E. coli XL1-Blue pDW5. The protein was present in all cellular fractions, but was the most abundant protein in the membrane fraction. PBP5 could be released from the membrane pellet by extraction with 1 M NaCl. However, the protein had a high tendency to precipitate if the salt extract was dialyzed. On the other hand, PBP5 was stable and did not precipitate if solubilized from the membranes by 2% Triton X-100. PBP5 was purified from the crude Triton X-100 extract of membranes by anion exchange chromatography using CM-Sepharose column. Proteins were eluted with a NaCl gradient from 50 mM to 1 M. PBP5 eluted at a NaCl concentration of 0.25–0.30 M. The established purification method yielded 5.6 mg of pure enzyme from 4.8 g cells (wet weight) (Fig. 1).
3.3Characterization of PBP5
As expected for a penicillin-binding protein, the isolated PBP5 bound covalently biotinylated ampicillin (BioAmp), indicating that the purified enzyme had an active conformation (data not presented).
The soluble murein precursor UDP-MurN Ac-pentapeptide (UDP-MurN Ac-PP) was used to test for a possible dd-carboxypeptidase activity of PBP5. UDP-MurN Ac-PP and the product of a dd-carboxypeptidase, UDP-MurN Ac-tetrapeptide (UDP-MurN Ac-TP) were separated by HPLC. Indeed, PBP5 was found to cleave the peptide bond between the subterminal and terminal d-alanine moieties (positions 4 and 5) of the pentapeptide side chain of the substrate and converted the pentapeptide to a tetrapeptide. No such cleavage occurred in the absence of PBP5.
3.3.3pH dependence of the activity of PBP5
The pH-dependence of the activity of PBP5 against UDP-MurN Ac-PP was determined from pH 3.0 to 8.0. Maximum activity of the enzyme was observed at pH 4.5.
3.3.4Inhibition of dd-carboxypeptidase activity by penicillin G
In order to determine whether the dd-carboxypeptidase activity of PBP5 is inhibited by penicillin, the enzyme was pre-incubated with the antibiotic before being added to the substrate UDP-MurN Ac-PP. As expected, the dd-carboxypeptidase activity of PBP5 was completely abolished by penicillin G (Fig. 2).
3.3.5Activity of PBP5 against E. coli D456 murein and muropeptides
E. coli D456 lacks the three lmw PBPs with dd-carboxypeptidase activity. The murein of the strain has a greater content of pentapeptide-containing muropeptides, which occur in two major forms: monomer – disaccharidepentapeptide (Penta) with retention time about 39 min (under the conditions used), and dimer – bis-disaccharidenonapeptide (TetraPenta), 73.8 min (Fig. 3(a)). Murein from E. coli strain D456 was digested and analyzed by HPLC as described in Section 2. The percent participation of pentapeptide muropeptides in the total muropeptide pool was calculated. We found that the digestion of murein with purified PBP5, followed by Cellosyl, results in a reduction in the amount of Penta monomer by 32.3% and of dimer TetraPenta by 3.4% compared to the amounts of these muropeptides present in murein degraded with Cellosyl alone (Fig. 3(b)). In turn when murein was first digested with Cellosyl and the obtained muropeptide pool was treated with PBP5, complete disappearance of the Penta monomer and reduction of the PentaTetra dimer by 18.4%, was observed (Fig. 3(c)).This result indicated that the preferred substrates for PBP5 are pentapeptide monomers.
3.4Characterization of mutant lacking functional lmo2754 mutant
Integration of plasmid pAUL-A::lmo2754 into L. monocytogenes gene lmo2754 was achieved as described above. The PBP profile of the mutant proved the absence of PBP5 (Fig. 4). Inactivation of the structural gene for PBP5 was not lethal for the cells, but the generation time of the mutant at 37 °C in TSB was 59 min, compared to 49 min for the parental strain. Gross morphology of the cells as observed by scanning electron microscopy also did not reveal any dramatic changes, such as the formation of filaments, etc., and the average cell size of the mutant in log phase culture was no different than that of the wild-type (data not shown). However, examination of micrographs of ultrathin sections of wild-type and mutant cells indicated slightly thicker cell wall in the case of the mutant (Fig. 5). Measurements involving 150 sections each of both cell types revealed that the thickness of the cell wall of mutant DA01 was 31.99 ± 1.1 nm (P < 0.05), compared to 27.35 ± 1.27 nm (P < 0.05) for wild-type cells. These data were confirmed by parallel measurements of dry weight of crude cell wall preparations from the similar number of wild-type and mutant cells. The MICs of the β-lactam antibiotics ampicillin and imipenem, and also of vancomycin and nisin, other antibiotics that interfere with the biosynthesis of murein were exactly the same for both the mutant and wild-type.
Although L. monocytogenes causes diseases treatable with β-lactam antibiotics, the mode of action of such antibiotics as ampicillin against this bacterium (death without lysis) has not been elucidated. Very little is also known about the cellular targets of ampicillin, that is the PBPs of this bacterium visualized using labeled antibiotic. Of the five PBPs of L. monocytogenes, only PBP3 has received some attention and was described as the primary lethal target for β-lactam antibiotics. The protein has also been found to play a role in the acquisition of dicloxacillin resistance by L. monocytogenes but was not investigated further. To learn more about the action of beta-lactams against L. monocytogenes we have initiated a thorough investigation of the PBPs of this bacterium, beginning with the lmw PBP5 described in this paper. The results of a study on PBP4 are in preparation. We have shown that PBP5 is a dd-carboxypeptidase that is associated with the membrane fraction of L. monocytogenes, with highest activity in vitro at pH 4.5. The enzyme preferentially degrades low-molecular-weight substrates, as demonstrated in experiments involving murein rich in pentapeptides isolated from genetically engineered E. coli D456 lacking PBP4, PBP5 and PBP6 and UDP-MurN Ac-l-Ala-d-Glu-l-Lys-d-Ala-d-Ala. These observations may point to the role of PBP5 of L. monocytogenes in the maturation and turnover of the murein sacculus. dd-carboxypeptidases are considered dispensable enzymes, i.e., their inactivation is not lethal for a bacterial cell. On the other hand, they have been shown to play an important role in determining the shape of the murein sacculus and thus that of the bacterial cell though otherwise the in vivo functions of these enzymes are poorly understood. Both overproduction of dd-carboxypeptidase by bacteria, as shown for E. coli and depriving a cell of this activity by a mutational event as demonstrated for Streptococcus pneumoniae can result in altered morphology. In the latter case the cells grew in the form of irregular, enlarged spheres with uneven distribution of septa. Extensive changes in the muropeptide composition of the murein of the mutant cells was also described. More recently, the dd-carboxypeptidase was shown to be involved in the division process of S. pneumoniae. The situation is more complex in the case of E. coli mutants lacking dd-carboxypeptidase since this bacterium has four low-molecular weight PBPs with similar dd-carboxypeptidase activity. Loss of PBP5 has been shown to result in morphologically aberrant E. coli cells with altered diameters and loss of uniform outer contour [28,29] as well as a strong branching phenotype.
The mutant of L. monocytogenes lacking PBP5 constructed by us grew slightly slower than the wild-type. No differences in the gross morphology of the cells, that is presence of filaments or other irregular shapes was observed, and the size of the mutant cells in log phase culture was no different from the wild-type. However, detailed scrutiny of thin sections using transmission electron microscopy (TEM) revealed somewhat thicker cell walls in the case of the mutant. There is a possibility that the activity of L. monocytogenes PBP5 is compensated for in the mutant by another protein with similar activity, since a cursory search of the L. monocytogenes genome has revealed another putative dd-carboxypeptidase beside PBP5. This possibility is currently being investigated. The situation could be similar to that in Bacillus subtilis, in which a dd-carboxypeptidase is required for normal cell morphology during stationary phase growth and has been demonstrated to be instrumental in cleavage of non-cross-linked pentapeptides in the vegetative cell wall and involved in lateral wall growth. Inactivation of the enzyme is not lethal for the cells but genome sequence analysis of the B. subtilis genome has revealed other lmw PBPs that may take over its function.
- 1997) The pathogenicity of Listeria monocytogenes: a public health perspective. Rev. Med. Bacteriol. 8, 1–14.(
- Pollock, S.S., Pollock, T.M., Harrison, M.J. Ampicillin resistant Listeria monocytogenes meningitis. Arch. Neurol. 43, (1986) 106
- Rapp, M.F., Pershadingh, H.A., Long, J.W., Pickens, J.M. Ampicillin resistant Listeria monocytogenes meningitis in a previously healthy 14 year old athlete. Arch. Neurol. 41, (1984) 1304
- 2003) Ampicillin resistance in Listeria monocytogenes acquired as a result of transposon mutagenesis. Acta Microbiol. Pol. 52, 131–142., , (
- Edwards, D.H., Donachie, D.W. Construction of a triple deletion mutant of penicillin-binding proteins 4, 5 and 6 in Escherichia coli. DePedro, M.A., Holtje, J.V., Loffelhardt, W., Eds. Bacterial Growth and Lysis: Metabolism and Structure of the Bacterial Sacculus. (1993) Plenum Press, London
- Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.