Choosing a Bacterial Strain for your Cloning Application

Escherichia coli is one of the most popular hosts for overexpression of cloned proteins because of its fast and efficient growth, and easy cultivation and induction. The most widely used E. coli strains in expression of cloned proteins are BL21 and its derivatives. This is mainly because BL21 strains are deficient in a cytoplasmic protease, Lon, and an outer membrane protease, OmpT. Both of these mutations (lon and ompT) allow intracellular accumulation of heterologous proteins (i.e., proteins derived from species or cell types different from the host) at a high rate while minimizing protein degradation during purification [1].

The host cell’s endogenous RNA polymerases express genes on the transformed plasmid, including the cloned gene, which may be driven by a promoter like the lac or tac. However these promoters are relatively weak and thus not ideal for strong expression of the cloned gene. Therefore, BL21 strains had been modified by adding bacteriophage T7 promoter to enhance expression of cloned recombinant proteins [2].

In the T7 expression system, the BL21 strain is engineered to contain a λDE3 lysogen to express the T7 RNA polymerase under the control of a mutant lac promoter, lacUV5. The lacUV5 promoter is positively regulated by IPTG (isopropylthio-β-galactoside), which is commonly used to induce transcription by T7 RNA polymerase inside the cells. As such, genes to be expressed can be cloned downstream of the T7 RNA polymerase promoter on the vector and induced from DE3 via IPTG (Figure 1A). This modified BL21 strain is commonly known as BL21(DE3).

Figure 1. Modified BL21 strains for the T7 expression system.

To enhance strong RNA expression from the T7 promoter, the BL21(DE3) strain had been genetically modified for increased mRNA stability. One example is a mutation of the rne gene, which encodes an essential endonuclease, RNase E. RNase E is the main E. coli RNA degradosome component; its function is mRNA degradation and RNA transcript processing. RNase E cuts single-stranded RNA rich in AU bases, thereby helping to factually mature mRNA and rRNA. The rne131 mutation results in a truncated RNase E that lacks the C-terminal domain required for mRNA degradation [3-5]. Thus, the BL21(DE3) strain with rne131 increases stability of mRNA transcripts, in turn increasing expression of proteins (Figures 1B, 2).

Figure 2. Protein expression from T7 expression vectors in BL21 strains with or without the rne131 mutation. (U = uninduced, I = induced with IPTG).

The T7 promoter system of the BL21(DE3) strain results in higher basal expression levels of heterologous genes than in the original BL21 strain. This may become problematic if the cloned protein is toxic to the host cell. A controllable promoter such as lacUV5 allows the cells to be cultured to a suitable density before expression of the cloned gene is induced. However, the promoter may be “leaky,” meaning there may be a considerable level of gene expression without induction. To reduce the basal expression level, some BL21(D3) strains are designed to carry a pLys plasmid (commonly pLysS or pLysE; Table 1) that expresses T7 lysozyme. T7 lysozyme inhibits T7 RNA polymerase activity, thereby reducing basal expression of the toxic gene from the vector (Figure 1C) [6,7]. The inhibition by T7 lysozyme is overcome upon induction with IPTG. For tighter regulation of protein expression, BL21 strains containing lacI^q can be used. lacI^q strains overproduce the lac repressor, ensuring that genes are not expressed unless specifically induced (see Figure 3 for review of the lac operon).

Figure 3. The lac operon. When the endogenous level of lactose is low, the lac repressor (expressed by lacI) is able to bind to the lacO sequence and inhibit transcription of the lac genes (lacZ, lacY, and lacA) from the lac promoter (lacP). If lactose (more precisely, a metabolite called allolactose) or its analog IPTG is present, these bind to the lac repressor and change its conformation. This prevents the lac repressor binding to lacO, thereby derepressing expression of the operon.

As an alternative to lacI^q, some strains carry an arabinose-inducible promoter, araBAD, upstream of the gene encoding T7 RNA polymerase, in their genome. The araBAD promoter offers high-level protein expression with one of the tightest regulation mechanisms available for a T7 expression system, making it suitable for toxic protein expression (Figures 4, 5).

Figure 4. Protein expression from lacUV5 promoter (with basal regulation by LysS) vs. araBAD promoter. (A) The araBAD promoter offers lower basal (uninduced) and higher induced levels of T7 RNA polymerase–mediated protein expression than does the lacUV5 promoter with pLysS. (B) Expression of a toxic protein, elastin-like peptide-TEV protease fusion (ELP-TEV), was enhanced using the araBAD promoter.

When the gene encoding the T7 RNA polymerase is under the control of the araBADpromoter, expression of protein from the recombinant vector can be regulated by L-arabinose and glucose [9,10].

• To induce expression of the cloned protein, L-arabinose is added to the growth medium to express the T7 RNA polymerase, which then binds to the T7 promoter of the recombinant vector (Figure 5A). The L-arabinose concentration may be varied to modulate protein expression. IPTG is also required for induction if the vector contains the lacI gene.
• To repress basal expression from the araBADpromoter, glucose can be added to the growth medium, although basal expression is generally low in the absence of glucose. Glucose lowers cAMP levels, which in turn reduces the binding of cAMP activator protein (CAP) to the DNA, further decreasing transcriptional activation (Figure 5B).

Therefore, by varying the concentrations of L-arabinose and glucose, expression of the cloned protein can be modulated appropriately.

Figure 5. (A) A modified BL21 strain that carries a chromosomal insertion of a cassette containing the T7 RNA polymerase (T7 RNAP) gene in the araB locus, allowing regulated expression of T7 RNA polymerase from the araBAD promoter. (B) AraC is a transcriptional regulator that forms a complex with L-arabinose. In the absence of L-arabinose, the C-terminal domains of the AraC dimer bind to the araO₂ and araI₁ sites. For transcriptional activation, L-arabinose recognizes the N-terminal domains of the AraC dimer and switches AraC binding from araO₂ to araI₂, allowing transcription to begin. Contact of AraC with the araI₁ and araI₂ sites is also stimulated by the cAMP activator protein (CAP)-cAMP complex binding to the DNA when the glucose level is low [9,10].

Top

The majority of laboratory E. coli competent cells are derived from the wild type strains K-12 and B, which possess the EcoKI and EcoBI restriction-modification systems, respectively. EcoKI recognizes the sequence 5′ AAC(N)₆GTGC 3′, and EcoBI recognizes the sequence 5′ TGA(N)₈TGCT 3′ [10,11]. Their Type I restriction-modification systems are encoded by hsdR, hsdM,and hsdS genes to form a multi-unit enzyme with restriction, modification, and specificity activities (hence the abbreviation RMS) [12]. When their recognition site is unmodified (i.e., unmethylated), the enzyme cleaves (“restricts”) the sequence. When only one strand of the recognition sequence is methylated (i.e., hemimethylated), the enzyme methylates the unmodified strand and prevents restriction.

Therefore, the competent cell’s genotype with respect to hsdRMS must be considered when propagating unmethylated (foreign) DNA containing EcoKI and EcoBI sites, such as PCR inserts.

• Host hsdR mutants (phenotype: r^–, m⁺) can modify but not restrict DNA.
• Host hsdM and/or hsdS mutants (phenotype: r^–, m^–) are defective for both DNA modification and restriction because the affected gene products form a complex that modifies DNA.
• Plasmids derived from host hsdRMS mutants (phenotype: r^–, m^–) are unmethylated at EcoKI and EcoBI sites. When transformed into wild type hsdRMS strains, these plasmids may be cleaved. To prevent their cleavage, plasmids should be methylated first by propagating in an hsdR host.

Some restriction enzymes are known to be methylation-sensitive: they are unable to cleave methylated DNA sequences. The dam and dcm genes encode two of the most common methylases found in E. coli, DNA adenine methylase and DNA cytosine methylase, respectively. One general practice to keep DNA unmethylated is to use competent cells deficient in dam and dcm for transformation. The dam methylase modifies the adenine in the sequence 5′ GATC 3′ [13] while the dcm methylase methylates the second cytosine in the sequence 5′ CC(A/T)GG 3′ [14]. Therefore, DNA containing restriction sites that include the dam and dcm sequences should be propagated in adam^–/ dcm^–strain. While providing a unique strategy for controlling cleavage, dam^–/dcm^–strains should not be used in routine transformation because undesirable mutations may be introduced into the propagated DNA.

For cloning of eukaryotic genomic DNA, strains of competent cells that carry mutations in the methylation-dependent restriction systems (MDRS) should be selected for transformation. Genomic DNA from animals, plants, and lower eukaryotes contain methylated cytosines that are involved in epigenetic regulation of gene expression in many cellular processes [15,16]. DNA sequences containing these methylated cytosines can become the targets of methylation-dependent endonucleases that are part of the McrA, McrBC, and Mrr E. coli restriction systems (Table 2).

Therefore, an E. coli strain that carries mutated mcrA, mcrBC, and mrr is critical for propagation of DNA of eukaryotic origin that may contain methylated cytosines and/or adenines. It also ensures better representation for construction of genomic libraries of eukaryotic origin.

Top

Certain DNA constructs are prone to DNA recombination and therefore are unstable in the bacterial cells after transformation. Examples of such plasmid constructs include retroviral and lentiviral vectors, which contain sequences with long terminal repeats (LTRs), inverted repeats, and tandem repeats.

Competent cells with recA1 and/or recA13 mutations are widely employed for stable propagation of constructs with retroviral and lentiviral vector backbones. Mutations of the host recA gene render the enzyme involved in DNA repair inactive, thereby reducing recombination and allowing maintenance of unstable constructs. The recA mutation also prevents homologous recombination between the host chromosome and vector sequences. Nevertheless, recA deficiency alone is not sufficient to prevent deletion and recombination of DNA sequences with repeats, so competent cells engineered to maintain the stability of DNA repeats must be considered (Figure 6) [23-26].

Figure 6. Stability of a plasmid with direct repeat sequences. (A) Three strains of chemically competent cells were transformed with pH30, a 7.3 kb plasmid that contains ~100 repeats of a 32-bp sequence. Plasmid DNA recovered from randomly selected colonies of the three strains was analyzed for stability by agarose gel electrophoresis. (B) pH30, electroporated into Invitrogen Stbl4 cells and isolated from 5 separate transformations, was analyzed for stability by agarose gel electrophoresis.

Stability of DNA constructs may be further enhanced by performing the following:

• Incubate the cells at 30°C for 90 minutes for recovery after transformation.
• After plating cells on LB agar plates, grow colonies at 30°C.
• Use transformant colonies from plates no older than 4 days post-transformation.
• Inoculate transformant colonies into Terrific Broth (TB medium) at 30°C for starter and larger cultures for plasmid isolation.
• Harvest cultures for DNA isolation during or before the early stationary phase of cell growth.

Top

Large plasmids (>10 kb) are known to pose issues in bacterial transformation, since transformation efficiency tends to decrease as plasmid size increases [28,29]. As illustrated in Figure 7, recommendations to improve the success of transformation of large plasmids include:

1. Choose competent cells with higher transformation efficiencies (>1 x 10⁹ CFU/μg) for transformation with 10–30 kb plasmids.
2. Consider transformation by electroporation instead of heat shock to yield higher transformation efficiencies.
3. For DNA constructs >30 kb (e.g., cosmids), select competent cells specifically tested to handle such sizes.
4. Consider competent cells with the deoR mutation, which reportedly improves maintenance of large DNA constructs [30,31].

Figure 7. Transformation efficiencies with large plasmids up to 200 kb. (A) Chemically competent (C) or electrocompetent (E) cells of different transformation efficiencies were transformed with plasmids ranging from 8 kb to 28 kb, and efficiencies of DNA uptake were measured (n = 4). (B) Large plasmids of 60 kb, 100 kb, and 200 kb were transformed into two strains of electrocompetent cells with high transformation efficiencies, and efficiencies of DNA uptake were measured (n = 4). (C) Large plasmids of 60–200 kb were transformed into three strains of electrocompetent cells at with transformation efficiency of 1 x 10¹⁰ CFU/μg for determination of DNA uptake (n = 6).

Top

One of the methods to screen colonies carrying desired plasmids with an insert is positive selection. In this approach, the vector harbors a lethal gene within the multiple cloning site (MCS). Successful cloning of the insert into the vector disrupts the expression of the lethal gene and permits colony formation.

The ccdB gene, derived from the F′ episome, is one of the lethal genes used for positive selection and serves as a selectable marker in many cloning vectors [32,33]. The CcdB protein is a toxin that interferes with the rejoining step of DNA gyrase, leaving the host chromosome fragmented and inhibiting cell growth. In order to propagate cloning vectors with an active ccdB gene, host cells must be resistant to the toxic effects of CcdB.

Most CcdB-resistant strains carry gyrA462, a mutation in DNA gyrase that suppresses double-stranded DNA breakage mediated by the CcdB protein [34]. Note that bacterial strains carrying an F′ episome possess the ccdA gene, which negatively regulates ccdB [35]. However, use of F′ competent cells to propagate ccdB vectors is not recommended, since endogenous expression of short-lived CcdA protein may not be able to negate CcdB’s lethality. Conversely, when using a ccdB vector for positive selection, F^– strains should be used for transformation to avoid ccdA expression and prevent survival of colonies without the cloned DNA. 

Top

Some plasmids contain a specific origin of replication called R6Kγ. In order to replicate these plasmids in bacteria, competent cells must express the protein pi (π) to bind to the R6Kγ sequence and initiate replication of the plasmids. The protein π is encoded by the gene pir.

Two types of competent cells with the pir gene are commonly available: pir⁺ (wild type) and pir-116 (mutant) [36].

• pir⁺ cells maintain the R6Kγ plasmid at about 15 copies per cell, which is good for maintenance of the transforming construct and expression of a toxic cloned gene.
• pir-116 cells maintain the plasmid at about 250 copies per cell, which is preferable for cloning and purification of the transforming construct.

Certain R6Kγ plasmid cloning systems, such as the Univector system (Figure 8) [37], employ Cre-lox site-specific recombination. The Cre recombinase protein, when supplied, binds to loxP sites on the Univectors. This effects recombination of the donor and acceptor plasmids, creating a recombinant version that carries the gene of interest [38]. 

Figure 8. The Univector cloning system. The loxP sites on the donor and acceptor plasmids recombine in the presence of Cre recombinase to produce a recombinant plasmid.

Top

Single-stranded DNA (ssDNA) is useful for dideoxy DNA sequencing, preparation of strand-specific probes, in vitro mutagenesis, and construction of subtracted cDNA libraries. A popular cloning-based approach for generating ssDNA involves taking advantage of the replication cycle of the filamentous bacteriophage M13 (Figure 9) [39-45]. For the M13 bacteriophage to attach to and enter bacterial cells, the E. coli strain must express the F pilus, denoted by the genetic marker F′ or F⁺.

Figure 9. M13 phage life cycle. (1) The M13 phage infects the bacterial cell via attachment to the F pilus, and injects its circular ssDNA genome. (2) The host’s cellular machinery uses the ssDNA (the + strand) as a template to synthesize a complementary strand, resulting in a dsDNA circle called the replicative form, or RF, which then multiplies. (3) The (+) strand of RF is nicked and extended using the (–) strand as a template, in a process called rolling cycle replication. (4) After one round of extension, the (+) strand is nicked again and recircularized to form the ssDNA genome of the progeny phage. (5) The progeny genome is packaged into a phage particle after a series of processes and is released from the host.

To generate ssDNA, the dsDNA of interest may be subcloned into (1) an M13 vector, or (2) a phagemid (Figures 10, 11).

1. M13 vectors are based on the double-stranded form of the M13 bacteriophage genome (also known as the replicative form, or RF), modified with a vector multiple cloning site (MCS) to enable insertion of the gene of interest [46,47]. After transforming the cells, the M13 vectors multiply as phage intermediates and eventually form ssDNAs. The ssDNAs are packaged into M13 phage particles, as in the M13 replication cycle (Figure 9). Hence, the ssDNA can be purified from phage particles in the culture medium while the dsDNA remains inside the cells. However, cloning of DNA inserts >2 kb into M13 vectors tends to be less stable and poses challenges.
2. A phagemid provides an alternative approach for propagating large DNA sequences. Phagemids consist of plasmid sequences along with an MCS and replication origin (rep) but also carry a filamentous phage’s origin of replication (f1 ori) (Figure 11A). After transformation, F′ competent cells propagate and maintain phagemids with cloned DNA as plasmids. Only upon infection with a helper phage (e.g., M13KO7) do the cells generate ssDNA phage particles from the phagemid, because the helper phage carries necessary M13 packaging genes. Helper phages themselves are inefficient in self-packaging due to the replacement of f1 ori with a plasmid origin of replication (e.g., p15A ori,
   Figure 11B) [48].

Figure 11. Phagemid and the helper phage M13KO7.

As such, any E. coli strain with an F´ episome can be used to generate ssDNA via the M13 replication pathway. However, contaminating bacterial and double-stranded plasmid/phagemid DNA can interfere with preparation of high-quality ssDNA from these strains. Therefore, it is desirable to use strains that are endA⁺ (in addition to F′) to remove dsDNA during ssDNA purification (Figure 12).

Figure 12. Single-stranded DNA purified from strains containing the F´ episome. Clones carrying cDNA inserts of 0.3, 0.6, 0.9, 1.5, 2 kb, and no insert were analyzed on an agarose gel. Strain 1, which expresses wild type endA, shows the least dsDNA contamination.

Top

Protein expression in insect cells provides certain advantages, such as posttranslational modifications similar to those of mammalian cells and ease of scaling up. Baculoviruses, which infect insect cells, are used in one approach to introduce recombinant genes into insect cells [49]. One method to generate a baculovirus is by transfecting insect cells with a bacmid, which is a shuttle plasmid between E. coli and insect cells, carrying the gene of interest [50]. A recombinant bacmid may be constructed in E. coli by transforming competent cells harboring a parent bacmid and a helper transposon plasmid with a donor plasmid specially designed for carrying the gene of interest (Figure 13). 

Figure 13. Creation of a recombinant bacmid. In this system, the transforming donor plasmid carries the gene of interest and the site-specific transposon Tn7, which recognizes the lacZ-mini-attTn7 fusion sequence on the bacmid of the competent cell. Transposition from the plasmid to the bacmid occurs in the presence of the transposition proteins (transABCD) provided by a helper plasmid. Successful transposition disrupts the lacZ gene on the bacmid, resulting in white bacterial colonies.

Top

Recombinant DNA is routinely introduced into plant cells with ease and simplicity by an Agrobacterium-mediated process [51]. In this approach, Agrobacterium tumefaciens cells are transformed with plant DNA that has been cloned into a tumor-inducing plasmid, or Ti plasmid, modified for gene transfer into plant cells [52].

The transformation of plant DNA constructs into electrocompetent A. tumefaciens with a disarmed Ti plasmid offers some advantages:

• A disarmed Ti plasmid, also known as a vir helper plasmid, carries only essential components for DNA transfer [53]. In a wild type Ti plasmid, most of the T-DNA, which is transferred to the plant cells, is fairly large (~25 kb), does not contain useful restriction sites for cloning, and has been found to be nonessential for gene transfer (Figure 14A). The T-DNA region is eliminated and only the essential sequences (such as vir, ori, LB, RB) are retained to create a disarmed Ti plasmid that is easier to manipulate (Figure 14B). pAL4404 is a commonly used disarmed Ti plasmid.
• The desired sequence can be cloned into a binary vector like pBI121, which can replicate in both A. tumefaciens and E. coli [54]. The binary vector contains a smaller T-DNA with the border sequences LB and RB, along with a cloning site and a selectable marker (Figure 14C). The desired gene may be cloned into the binary vector between LB and RB, and propagated in E. coli; then A. tumefaciens can be transformed with a disarmed Ti plasmid for eventual gene transfer.
• Electroporation of A. tumefaciens offers higher transformation efficiency with fewer steps and less time while eliminating the risk of E. coli contamination [55].

Figure 14. Ti plasmid and the binary vector system.

For the best transformation efficiency of A. tumefaciens LBA4404 cells, YM (yeast-mannitol) broth should be used for recovery following transformation (Table 3) [56].

Electroporated cells were cultured at 30°C for 3 hours in different expression media for recovery and plated on either a YM or LB plate.

Top

Transformation efficiency and host genotype are especially important for DNA library construction in order to create a representative library covering genes of varying abundance and sizes. In the preparation of a library, saturating amounts of pooled DNA inserts (gDNA or cDNA fragments) that have been ligated into vectors are used to transform competent cells. Saturating amounts of DNA maximize the library size with representative DNA from a minimum number of transformation reactions, but lower the efficiency of DNA uptake by the cells. Therefore, when selecting competent cells for library construction, the highest transformation efficiency available and electrocompetency are recommended to obtain the greatest number of transformants from a single transformation.

In addition to transformation efficiency, the genotype of the selected competent cells are an important aspect in library preparation, and the following genetic markers should be noted (Table 4).

In summary, molecular cloning experiments often require competent cells with certain properties for transformation. To save time and effort, it is crucial to select competent cells that are appropriately designed for the desired applications.

1. Samuelson JC (2011) Recent developments in difficult protein expression: a guide to E. coli strains, promoters, and relevant host mutations. Methods Mol Biol 705:195–209. 
2. Studier FW, Rosenberg AH, Dunn JJ (1990) Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol 185:60–89. 
3. Kido M, Yamanaka K, Mitani T et al. (1996) RNase E polypeptides lacking a carboxyl-terminal half suppress a mukB mutation in Escherichia coli. J Bacteriol 178(3):3917–3925. 
4. Grunberg-Manago, M (1999) Messenger RNA stability and its role in control of gene expression in bacteria and phages. Annu Rev Genet 33:193–227. 
5. Lopez PJ, Marchand I, Joyce SA et al. (1999) The C-terminal half of RNase E, which organizes the Escherichia coli degradosome, participates in mRNA degradation but not rRNA processing in vivo. Mol Microbiol 33(1):188–199. 
6. Moffatt BA, Studier FW (1987) T7 lysozyme inhibits transcription by T7 RNA polymerase. Cell 49(2):221–227. 
7. Studier FW (1991) Use of bacteriophage T7 lysozyme to improve an inducible T7 expression system. J Mol Biol 219(1):37–44. 
8. Miyada CG, Stoltzfus L, Wilcox G (1984) Regulation of the araC gene of Escherichia coli: catabolite repression, autoregulation, and effect on araBAD expression. Proc Natl Acad Sci U S A 81(13):4120–4124. 
9. Lee N, Francklyn C, Hamilton EP (1987) Arabinose-induced binding of AraC protein to araI2 activates the araBAD operon promoter. Proc Natl Acad Sci U S A 84(24):8814–8818. 
10. Lautenberger JA, Kan NC, Lackey D (1978) Recognition site of Escherichia coli B restriction enzyme on phi XsB1 and simian virus 40 DNAs: an interrupted sequence. Proc Natl Acad Sci U S A 75(5):2271–2275. 
11. Kan NC, Lautenberger JA, Edgell MH (1979) The nucleotide sequence recognized by the Escherichia coli K12 restriction and modification enzymes. J Mol Biol 130(2):191–209. 
12. Murray NE (2000) Type I restriction systems: sophisticated molecular machines (a legacy of Bertani and Weigle). Microbiol Mol Biol Rev 64(2):412–434. 
13. Geier GE, Modrich P (1979) Recognition sequence of the dam methylase of Escherichia coli K12 and mode of cleavage of Dpn I endonuclease. J Biol Chem 254(4):1408–1413. 
14. Buryanov YI, Bogdarina IG, Bayev AA (1978) Site specificity and chromatographic properties of E. coli K12 and EcoRII DNA-cytosine methylases. FEBS Lett 88(2):251–254. 
15. Costello JF, Plass C (2001) Methylation matters. J Med Genet 38(5):285–303. 
16. Adams RL (1995) Eukaryotic DNA methyltransferases--structure and function. Bioessays 17:139–145. 
17. Raleigh EA, Wilson G (1986) Escherichia coli K-12 restricts DNA containing 5-methylcytosine. Proc Natl Acad Sci U S A 83(23):9070–9074. 
18. Raleigh EA, Benner J, Bloom F et al. (1991) Nomenclature relating to restriction of modified DNA in Escherichia coli. J Bacteriol 173(8):2707–2709. 
19. Dila D, Sutherland E, Moran L (1990) Genetic and sequence organization of the mcrBC locus of Escherichia coli K-12. J Bacteriol 172(9):4888–4900. 
20. Stewart FJ, Raleigh EA (1998) Dependence of McrBC cleavage on distance between recognition elements. Biol Chem 379(4–5):611–616. 
21. Kelleher JE, Raleigh EA (1991) A novel activity in Escherichia coli K-12 that directs restriction of DNA modified at CG dinucleotides. J Bacteriol 173(16):5220–5223. 
22. Waite-Rees PA, Keating CJ, Moran LS et al. (1991) Characterization and expression of the Escherichia coli Mrr restriction system. J Bacteriol 173(16):5207–5219. 
23. Lovett ST, Drapkin PT, Sutera VA Jr (1993) A sister-strand exchange mechanism for recA-independent deletion of repeated DNA sequences in Escherichia coli. Genetics 135(3):631–642. 
24. Trinh T, Jessee J, Bloom FR et al. (1994) Stbl2: An Escherichia coli strain for the stable propagation of retroviral clones and direct repeat sequences. Focus 16(3):78–80.
25. Bi X, Liu LF (1996) A replicational model for DNA recombination between direct repeats. J Mol Biol 256(5):849–958. 
26. Schmidt BJ, Bloom FR. (1999) ElectroMAX Stbl4 cells for stable maintenance of repeat sequences. Focus 21(2):52–53.
27. Leonardo ED, Sedivy JM (1990) A new vector for cloning large eukaryotic DNA segments in Escherichia coli. Biotechnology (N Y) 8(9):841–844. 
28. Leonardo ED, Sedivy JM (1990) A new vector for cloning large eukaryotic DNA segments in Escherichia coli. Biotechnology (N Y) 8(9):841–844. 
29. Siguret V, Ribba AS, Chérel G (1994) Effect of plasmid size on transformation efficiency by electroporation of Escherichia coli DH5 alpha. Biotechniques 16(3):422–426. 
30. Hanahan D (1989) US Patent 4,851,348. 
31. Durfee T, Nelson R, Baldwin S et al. (2008) The complete genome sequence of Escherichia coli DH10B: insights into the biology of a laboratory workhorse. J Bacteriol 190(7):2597–2606. 
32. Bernard P, Gabant P, Bahassi EM et al. (1994) Positive-selection vectors using the F plasmid ccdB killer gene. Gene 148(1):71–74. 
33. Bernard P (1996) Positive selection of recombinant DNA by CcdB. Biotechniques 21(2):320–323. 
34. Bernard P, Couturier M (1992) Cell killing by the F plasmid CcdB protein involves poisoning of DNA-topoisomerase II complexes. J Mol Biol 226(3):735–745. 
35. Bernard P, Kézdy KE, Van Melderen L et al. (1993) The F plasmid CcdB protein induces efficient ATP-dependent DNA cleavage by gyrase. J Mol Biol 234(3):534–541. 
36. Metcalf WW, Jiang W, Wanner BL (1994) Use of the rep technique for allele replacement to construct new Escherichia coli hosts for maintenance of R6K gamma origin plasmids at different copy numbers. Gene 138(1-2):1-7. 
37. Marsischky G, LaBaer J (2004) Many paths to many clones: a comparative look at high-throughput cloning methods. Genome Res (10B):2020–2028. 
38. Liu Q, Li MZ, Liu D, Elledge SJ (2000) Rapid construction of recombinant DNA by the univector plasmid-fusion system. Methods Enzymol 328:530–549. 
39. Zagursky RJ, Berman ML (1984) Cloning vectors that yield high levels of single-stranded DNA for rapid DNA sequencing. Gene 27(2):183–191. 
40. Nakagami S, Matsunaga H, Miyoshi K et al. (1991) Nonradioactive detection of nucleic acid by the universal probe system. Anal Biochem 192(1):11–16. 
41. Carter P (1987) Improved oligonucleotide-directed mutagenesis using M13 vectors Methods Enzymol 154:382–403.
42. McClary JA, Witney F, Geisselsoder J (1989) Efficient site-directed in vitro mutagenesis using phagemid vectors. Biotechniques 7(3):282–289. 
43. Duguid JR, Rohwer RG, Seed B (1988) Isolation of cDNAs of scrapie-modulated RNAs by subtractive hybridization of a cDNA library. Proc Natl Acad Sci U S A 85(15):5738–5742. 
44. Rubenstein JL, Brice AE, Ciaranell RD et al. (1990) Subtractive hybridization system using single-stranded phagemids with directional inserts. Nucleic Acids Res 18(16):4833–4842. 
45. Klickstein LB (2001) Production of a subtracted cDNA library. Curr Protoc Mol Biol Chapter 25: Unit 25B.1.  
46. Yanisch-Perron C, Vieira J, Messing J (1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33(1):103–119. 
47. Messing J (1991) Cloning in M13 phage or how to use biology at its best. Gene 100:3–12. 
48. Vieira J, Messing J (1987) Production of single-stranded plasmid DNA. Methods Enzymol 153:3–11. 
49. Kost TA, Condreay JP, Jarvis DL (2005) Baculovirus as versatile vectors for protein expression in insect and mammalian cells. Nat Biotechnol 23(5):567–575. 
50. Luckow VA, Lee SC, Barry GF (1993) Efficient generation of infectious recombinant baculoviruses by site-specific transposon-mediated insertion of foreign genes into a baculovirus genome propagated in Escherichia coli. J Virol 67(8):4566-79. 
51. Hoekema A, Roelvink PW, Hooykaas PJ et al. (1984) Delivery of T-DNA from the Agrobacterium tumefaciens chromosome into plant cells. EMBO J 3(11):2485–2490. 
52. Lee LY, Gelvin SB (2008) T-DNA binary vectors and systems. Plant Physiol 146(2):325–332. 
53. Hoekema A, van Haaren MJ, Fellinger AJ (1985) Non-oncogenic plant vectors for use in the agrobacterium binary system. Plant Mol Biol 5(2):85–89. 
54. Bevan M (1984) Binary Agrobacterium vectors for plant transformation. Nucleic Acids Res 12(22):8711–8721. 
55. Shen WJ, Forde BG (1989) Efficient transformation of Agrobacterium spp. by high voltage electroporation. Nucleic Acids Res 17(20):8385. 
56. Lin JJ (1994) Optimization of the transformation efficiency of Agrobacterium tumefaciens cells using electroporation. Plant Sci 101(1):11–15. 
57. Jespers LS, Messens JH, De Keyser A (1995) Surface expression and ligand-based selection of cDNAs fused to filamentous phage gene VI. Biotechnology (N Y) 13(4):378–382. 
58. Laird-Offringa IA, Belasco JG (1996) In vitro genetic analysis of RNA-binding proteins using phage display libraries. Methods Enzymol 267:149-68. 
59. Oh MY, Joo HY, Hur BU (2007) Enhancing phage display of antibody fragments using gIII-amber suppression. Gene 386(1-2):81–89. 
60. Clarke L, Carbon J (1976) A colony bank containing synthetic Col El hybrid plasmids representative of the entire E. coli genome. Cell 9(1):91–99. 
61. William JG (1981) The Preparation and Screening of a cDNA Clone Bank. In: Williamson R (editor), Genetic Engineering. New York: Academic Press. p 6. 
62. Toole JJ, Knopf JL, Wozney JM (1984) Molecular cloning of a cDNA encoding human antihaemophilic factor. Nature 312(5992):342–347. 
63. Wood WI, Capon DJ, Simonsen CC et al. (1984) Expression of active human factor VIII from recombinant DNA clones. Nature 312(5992):330–337. 
64. Donahue RA., Bloom FR (1998) Large-volume transformation with high-efficiency chemically competent cells. Focus 20(2):54–56.
65. Hanahan D, Jessee J, Bloom FR (1991) Plasmid transformation of Escherichia coli and other bacteria. Methods Enzymol 204:63–113.