Appl Microbiol Biotechnol DOI 10.

1007/s00253-012-3987-2

METHODS AND PROTOCOLS

An efficient transformation method for Bacillus subtilis DB104

Ljubica Vojcic & Dragana Despotovic & Ronny Martinez & Karl-Heinz Maurer & Ulrich Schwaneberg

Received: 18 December 2011 / Revised: 15 February 2012 / Accepted: 17 February 2012 # Springer-Verlag 2012

Abstract Bacillus subtilis strains are used for extracellular expression of enzymes (i.e., proteases, lipases, and cellulases) which are often engineered by directed evolution for industrial applications. B. subtilis DB104 represents an attractive directed evolution host since it has a low proteolytic activity and efficient secretion. B. subtilis DB104 is hampered like many other Bacillus strains by insufficient transformation efficiencies (103 transformants/g DNA). After investigating five physical and chemical transformation protocols, a novel natural competent transformation protocol was established for B. subtilis DB104 by optimizing growth conditions and histidine concentration during competence development, implementing an additional incubation step in the competence development phase and a recovery step during the transformation procedure. In addition, the influence of the amount and size of the transformed plasmid DNA on transformation efficiency was investigated. The natural competence protocol is easy in handling and allows for the first time to generate large libraries (1.5105

transformants/g plasmid DNA) in B. subtilis DB104 without requiring microgram amounts of DNA. Keywords B. subtilis DB104 . Directed evolution . Natural competence . Transformation protocol

Introduction Directed evolution is a powerful algorithm to improve enzyme properties in iterative cycles of diversity generation and screening. A typical directed evolution experiment comprises three major steps: (1) diversity generation, (2) screening to identify improved mutants out of a large pool of variants, and (3) isolating the gene encoding for the improved protein variant (Tee and Schwaneberg 2007). The first step in a directed evolution campaign includes the generation of a DNA mutant library and its subsequent transformation into the host organism. Advances in screening technologies like flow cytometry-based screening systems (Aharoni et al. 2005; Prodanovic et al. 2011; Tu et al. 2011) allow a throughput of up to 108 variants so that transformation efficiencies are increasingly becoming the limiting step in directed evolution experiments. B. subtilis is used as host for production of secretory proteins especially proteases and lipases which have a significant market value (Westers et al. 2004). Low transformation efficiencies are a main challenge when using Bacillus strains in directed evolution campaigns. In order to ensure an efficient and secreted expression of the target enzyme in a Bacillus host, it is essential to perform the directed evolution directly in the Bacillus production strain. B. subtilis DB104 is one of the most used strains for the production of industrially important extracellular enzymes, especially subtilisin proteases. B. subtilis DB104 is a derivative of B. subtilis 168 Marburg strain (Kawamura and Doi 1984), generated by lesions in the genes

Electronic supplementary material The online version of this article (doi:10.1007/s00253-012-3987-2) contains supplementary material, which is available to authorized users. L. Vojcic : D. Despotovic : R. Martinez : U. Schwaneberg (*) Lehrstuhl fr Biotechnologie, RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany e-mail: u.schwaneberg@biotec.rwth-aachen.de K.-H. Maurer International Research Laundry & Home Care, Biotechnology, Henkel AG & Co. KGaA, 40191 Dsseldorf, Germany K.-H. Maurer AB Enzymes GmbH, Feldbergstrae 78, 64293 Darmstadt, Germany

Appl Microbiol Biotechnol

coding for the alkaline protease (aprA3) and the neutral protease (nprE18). As a consequence, B. subtilis DB104 shows less than 4% of the extracellular protease activity of B. subtilis 168 Marburg strain, which is a key advantage when proteases are expressed heterologously. B. subtilis DB104 requires histidine as an essential amino acid which is usually supplemented in the growth medium (Kawamura and Doi 1984). Transformation of Bacillus strains can be accomplished by natural competence or artificial methods. Natural competence for DNA uptake is in Gram-positive and Gram-negative bacteria a physiologically and genetically determined trait in response to environmental stress (Hamoen et al. 2003). The development of competence in B. subtilis is, in part, dictated by nutritional conditions and growth stage. In addition, the development of natural competence is strain specific due to divergent structure of the quorum sensing components which control development of natural competence (Tran et al. 2000). Transformation based on natural competence allows DNA uptake from various sources, for instance phage DNA, plasmid DNA, and chromosomal DNA. DNA uptake occurs by a common pathway through the following stages: binding, fragmentation, uptake, as well as additionally integration and replication/or mismatch repair in the case of chromosomal DNA (Dubnau 1991). Reported transformation efficiency for B. subtilis 168 strain reached 106 transformants/g chromosomal DNA (Anagnostopoulos and Spizizen 1961). Artificial transformation methods employ, in contrast to the abovementioned natural competence methods, physical and chemical treatments such as electroporation or addition of polyethylene glycol or mannitol (Brigidi et al. 1990; Chang and Cohen 1979) (Table 1). The technically least demanding method, reported as simple and rapid method (Table 1, no. 1), allows transformation on solid media by overlaying plated Bacillus cells with chromosomal or plasmid DNA (23 g). This agar plate-based transformation method is described to yield approximately 100200 transformants/g DNA (Hauser and Karamata 1994). An advancement of the agar plate-based transformation method was achieved by adding DNA in protoplast lysates (Table 1, no. 2), yielding a transformation efficiency of 2.3103/g chromosomal DNA (Akamatsu and Taguchi 2001); this transformation efficiency decreased

100 times when plasmid DNA was transformed. For transformation of plasmid DNA, an electroporation method (Table 1, no. 3) with a transformation efficiency of 104/g DNA was developed (Brigidi et al. 1990). Electroporation combined with a previous treatment of the cells with glycine (Table 1, no. 4) increased efficiency to 1.7106 transformants/g plasmid DNA. Preparation and transformation of B. subtilis protoplasts (Table 1, no. 5) in presence of polyethylene glycol yielded up to 1.4 107 transformants/g plasmid DNA (Chang and Cohen 1979). All methods summarized in Table 1 were investigated, yielding, in case of B. subtilis DB104, up to 5 10 3 transformants/ g plasmid DNA (Table 1, no. 5). The transformation efficiency represents a bottleneck in the directed evolution of proteases when B. subtilis DB104 is used as an expression host. In this report, a highly efficient transformation method based on development of natural competence in B. subtilis DB104 was optimized, resulting in an efficiency of 1.5105 transformants/g plasmid DNA. The developed protocol is based on transformation protocol for B. subtilis 168 strain employing chromosomal DNA (Anagnostopoulos and Spizizen 1961).

Materials and methods Chemicals All chemicals were purchased from AppliChem (Darmstadt, Germany) or Carl Roth GmbH (Karlsruhe, Germany) except casamino acids (ForMediumTM, Norfolk, UK). Materials The cell culture was cultivated in a Certomat RM shaker (Sartorius Stedim Biotech GmbH, Goettingen, Germany). The amount of DNA in the experiments was quantified by using a NanoDrop photometer (ND-1000; NanoDrop Technologies, Wilmington, DE, USA). Plasmid isolation kit was purchased from Qiagen (Hilden, Germany). Optical density of the cell culture at 600 nm (OD600) was

Table 1 Summary of reported artificial methods for transformation of Bacillus species No. Transformation method Bacillus strain Transformants/ g DNA 100200 2.3103 1104 1.69106 4107 Remarks 23 g of DNA needed Optimized for chromosomal DNA Low survival rate due to applied voltage Survival rate due to applied voltage (216%) 23 days recovery time

1 2 3 4 5

Solid media Solid media in protoplast lysates Electroporation of intact Bacillus cells Electroporation with glycine treatment Protoplasts

B. subtilis 168/W23 B. subtilis AYG2 B. subtilis PB1424 B. pseudofirmus OF4 Derivatives of B. subtilis Marburg 168 strain

Appl Microbiol Biotechnol

measured using a BioPhotometer plus photometer (Eppendorf AG, Hamburg, Germany). Bacterial strain and plasmids

subtilis DB104 cells were competent for approximately 1 h. Histidine concentration of 200 g/ml (final concentration) was used in final protocol. Transformation procedure

The bacterial strain used in this study was B. subtilis DB104 (Kawamura and Doi 1984). Plasmids used for transformation were as follows: (1) pC194 (2,910 bp; cat. no. 4393; DSMZ, Braunschweig, Germany); (2) pHY300PLK, a BacillusEscherichia coli shuttle vector (4,870 bp; Takara Bio Inc., Shiga, Japan); (3) pHY300Car, derivative of pHY300PLK vector containing a subtilisin Carlsberg (6,221 bp); and (4) pHCMC04 BacillusE. coli shuttle vector (8,089 bp) (Nguyen et al. 2005). The final concentrations of antibiotics in agar plates were 15 g/ml tetracycline for cells containing pHY300PLK/pHY300Car plasmids, and 10 g/ml or 5 g/ml chloramphenicol for cells containing the pC194 or the pHCMC04 plasmid. Protocol Media composition Starvation medium 1 (SM1) contains 0.2% ammonium sulfate, 1.4% dipotassium hydrogen phosphate, 0.6% potassium dihydrogen phosphate, 0.07% sodium citrate, 0.5% glucose, 0.02% magnesium sulfate heptahydrate, 0.2% yeast extract, and 0.025% casamino acids. All the components were mixed together and autoclaved. Starvation medium 2 (SM2) is less rich in nutrients and includes 0.2% ammonium sulfate, 1.4% dipotassium hydrogen phosphate, 0.6% potassium dihydrogen phosphate, 0.07% sodium citrate, 0.5% glucose, 0.08% magnesium sulfate heptahydrate, 0.1% yeast extract, 0.01% casamino acids, and 0.05% calcium chloride. All components of SM2 medium were mixed and autoclaved together. Histidine solution was sterilized by filtration using 0.2-m filters (PuradiscTM 25 mm, cat. no. 67802502; GE Healthcare, Munich, Germany). Preparation of competent cells B. subtilis DB104 cells were spread on a LB agar plate without antibiotics (37 C; for 9 h). Antibiotic-free SM1 media was inoculated by transferring a single colony and subsequent cultivation (37 C; 250 rpm for 1416 h). Overnight culture was diluted in SM1 medium by adjusting optical density at 600 nm (OD600) to 0.5 (approximately 2.9107 cells/ml) in a volume of 10 ml and incubated (37 C; 200 rpm for 3 h). The cell culture volume was doubled by addition of SM2 medium and subsequently supplemented with varied histidine concentrations (final concentration 0, 10, 50, 200, 500, and 1,000 g/ml; total volume 20 ml) and incubated (37 C; 300 rpm for 2 h). After the treatment, B.

Five hundred microliters of the competent cells was mixed with varied amounts of plasmid DNA (pHY300Car2, 5, 10, 20, 40, 100, 250, 500, 1,000 ng) and incubated (37 C; 200 rpm for 30 min). In order to recover the cells, 300 l of fresh LB medium was supplemented and competent B. subtilis DB104 were additionally incubated (37 C; 200 rpm for 30 min). Two hundred microliters of B. subtilis DB104 cell suspension was subsequently spread on LB agar plates with selective antibiotic (tetracycline, final concentration 15 g/ml).

Results An efficient protocol for B. subtilis DB104 strain for transforming plasmid DNA was developed which represents an extension of earlier work by Anagnostopoulos and Spizizen (1961). Table 2 summarizes the optimization steps during competence development and transformation procedures. The protocol optimization comprises steps such as omitting centrifugation, adapting incubation times, and optimization of the cell growth, transformation times, and histidine concentration. Subsequently, the influence of plasmid DNA amount and size on transformation efficiency were investigated. Optimization of the growth time of B. subtilis DB104 B. subtilis develops its natural competence at the end of exponential growth phase with the expression of the comK gene (Hamoen et al. 2003). In order to determine the time required for the cells to enter into a stationary phase, the growth of B. subtilis DB104 in SM1 and SM2 medium was monitored. Overnight culture (37 C; 200 rpm for 1416 h) of B. subtilis DB104 in SM1 medium was diluted with SM1 medium until OD600 reached a value of 0.5. B. subtilis DB104 cells were grown (37 C; 200 rpm) and OD600 was monitored every 30 min. Fig. 1 shows the growth curve of B. subtilis DB104 in SM1 medium (closed square) reaching a stationary phase after 3 h. At the end of exponential phase (3 h of growth time), the total cell culture volume was diluted 1:1 with SM2 medium and supplemented with varied histidine concentrations (final concentration 0, 10, 50, 200, 500, and 1,000 g/ml). Upon dilution with SM2 medium (SM1/SM2, open squares), the end of exponential phase was reached after 2 h of growth and different histidine concentrations show no detectable effect on growth of B.

Appl Microbiol Biotechnol Table 2 Comparison of the developed plasmid transformation protocol for B. subtilis DB104 to the transformation protocol of Spizizen for B. subtilis 168 strain (Anagnostopoulos and Spizizen 1961) Spizizens protocol Organism: B. subtilis 168 Overnight culture on an LB agar plate 108 cells/ml inoculums cultivated in SM1 medium containing essential tryptophan (50 g/ml) Time of incubation: 4 h Centrifugation of the cells 1:10 dilution in SM2 medium containing tryptophan (5 g/ml) No additional incubation step in SM2 medium Chromosomal DNA added to the cell culture followed by incubation for 90 min (37 C; rpm not defined, final volume 1 ml) No recovery step in LB medium Plated on agar plates with selective antibiotic Optimized protocol for B. subtilis DB104 Organism: B. subtilis DB104 Overnight culture in SM1 liquid medium Dilution in SM1 medium (no essential histidine amino acids in medium) to reach 107 cells/ml (OD600 ~0.5) Time of incubation: 3 h No centrifugation step 1:1 dilution in SM2 medium containing histidine (200 g/ml) Incubation for additional 2 h (37 C; 300 rpm without DNA; final volume 20 ml) Plasmid DNA added to the cell culture followed by incubation for 30 min (37 C; 200 rpm; final volume 0.8 ml) Recovery of the cells with 300 l fresh LB medium for 30 min (37 C; 200 rpm) Plated on agar plates with selective antibiotic

subtilis DB104. After optimization, B. subtilis DB104 cells reach a state of natural competence after a cultivation time of 3 h in SM1 and of 2 h in SM1/SM2 medium. Influence of histidine concentration on transformation efficiency To investigate the dependence of different histidine concentrations on transformation efficiency of B. subtilis DB104, the concentration range of histidine as supplement to SM1/ SM2 medium was varied (final concentration 0, 10, 50, 200, 500, and 1,000 g/ml). Fig. 2 shows that concentrations of histidine up to 500 g/ml influenced the transformation efficiency of B. subtilis DB104 with the range varying from 8.0104 to 1.5105. At higher concentrations of histidine
Fig. 1 Growth curve of B. subtilis DB104 in SM1 medium (filled squares) and SM1/SM2 medium (open squares). Optical density (OD600) monitored at 600 nm in function of time. The cells were grown (37 C, 200 rpm) and 1ml aliquots were taken for OD600 measurements every 30 min. In order to ensure that OD600 reading is within the linear range (0.10.5) of accuracy of photometer, the cell suspension was diluted prior to the measurement

(1,000 g/ml), the competence of B. subtilis DB104 decreased to 4.5104. In the final protocol, 200 g/ml histidine was used. Dependence of different shuttle vectors on the transformation efficiency of B. subtilis DB104 cells Three different shuttle expression vectors for Bacillus (pC194, pHY300PLK, and pHCMC04) were used for transformation of natural competent B. subtilis DB104 in order to investigate the dependence of vector size and antibiotic resistance on the transformation efficiency. The used plasmids ranged from 2.9 to 8.1 kb [pC194 (2.9 kb), pHY300PLK (4.9 kb), pHY300Car (6.2 kb), pHCMC04 (8.1 kb)]. The highest number of transformants was obtained

Appl Microbiol Biotechnol

Fig. 2 Transformation efficiency (transformants per microgram of plasmid DNA 0 cfu/g) at varied concentrations of histidine supplemented (0, 10, 50, 200, 500, and 1,000 g/ml) to SM1/SM2 medium. Error bars represent standard deviation from the mean value between three triplicate experiments

using pHY300Car (1.5 10 5 transformants/ g DNA) (Table 3). This result indicates that the size of the used plasmid does not correlate with the transformation efficiency. In case of the antibiotics (chloramphenicol and tetracycline), 5- to 10-fold higher transformation efficiency could be obtained using plasmids bearing the tetracycline resistance cassette. Dependence of amount of plasmid DNA on the transformation efficiency of B. subtilis DB104 cells Transformation efficiencies of B. subtilis DB104 cells were determined as a function of DNA amount (pHY300Car2, 5, 10, 20, 40, 100, 250, 500, and 1,000 ng). Results presented in Fig. 3 indicate the correlation between transformation efficiency and DNA amount (range of 21,000 ng). Results also showed a linear relationship between the DNA amount used for transformation and the total number of transformants indicating single-hit kinetics (see Electronic Supplementary Material Fig. 1). An optimum was found between 5 and 10 ng of plasmid DNA. At plasmid amounts higher than 40 ng, the transformation efficiency declines continuously from 7104 to 4103 transformants/g plasmid DNA.

Fig. 3 Transformation efficiency (transformants per microgram of DNA 0 cfu/ g) in function of amount of DNA (ng). B. subtilis DB104 cells were transformed with different amounts of plasmid DNA (pHY300Car2, 5, 10, 20, 40, 100, 250, 500, and 1,000 ng). Error bars represent standard deviation from the mean value between triplicate experiments

Discussion Advances in screening technologies with throughputs up to 108 variants (Prodanovic et al. 2011; Tu et al. 2011) enable novel directed evolution strategies for instance using high mutational loads. In order to match throughput screening, efficient transformation protocols are necessary to generate a sufficient number of variants, especially in non-E. coli hosts. The use of Bacillus strains as protein expression hosts is especially important for industrially used hydrolases (proteases and/or lipases) which have to be secreted into media for high level production. B. subtilis DB104 represents an attractive directed evolution host due to its low proteolytic activity and efficient secretion. Transformation protocols for Bacillus strains shown in Table 1 were investigated for B. subtilis DB104, yielding a transformation efficiency of up to 5103 variants/g plasmid DNA using protoplast method (Table 1, no. 5). The protoplast method required after transformation a 3-day recovery phase and resulted in a growth of high number of B. subtilis DB104 cells without plasmid DNA despite of antibiotic use. The obtained transformation efficiency and requirement of high DNA concentrations (microgram range; Table 1, nos. 1, 2, and 5) made the physical and chemical transformation methods not suitable for directed evolution in B. subtilis DB104. Based on Spizizens protocol for transformation of chromosomal DNA by natural competence into the Bacillus 168 strain, a novel transformation protocol for B. subtilis DB104 was developed to transform plasmid DNA. After various optimizations [growth phase, histidine/DNA concentration, shuttle vectors (size, resistance/four constructs)], the highest transformation efficiency of B. subtilis DB104 was obtained by addition of plasmid

Table 3 Influence of plasmid resistance (CmR/TcR) and size on the transformation efficiency of B. subtilis DB104 Plasmid Size (kb) Phenotypea Transformants/ g plasmid DNA 1.3104 1.2105 1.5105 2.2104

pC194 pHY300PLK pHY300Car pHCMC04

a

2.9 4.9 6.2 8.1

CmR TcR TcR CmR

CmR chloramphenicol resistance, TcR tetracycline resistance

Appl Microbiol Biotechnol

DNA at the end of the exponential growth phase with efficiencies up to 1.5105 transformants/g DNA. The obtained efficiency is two orders of magnitude higher compared to that obtained after transformation of B. subtilis DB104 with any of the methods summarized in Table 1. The obtained result is in concordance with a study of Dubnau (1991) reporting that natural competence is developed at the end of exponential phase. The robustness of the transformation method was determined by investigating whether supplementing essential histidine affects the growth rate and transformation efficiency of B. subtilis DB104. In addition, the influence of the employed plasmid system and the amount of transformed plasmid DNA on transformation efficiency was determined. Varied histidine concentrations showed no effect on growth curve of B. subtilis DB104 but influenced transformation efficiency (Fig. 2). A histidine concentration of 200 g/ml was finally selected for B. subtilis DB104 transformation. Table 3 shows that the transformation efficiency of B. subtilis DB104 does not depend on the plasmid size. For instance, the empty vector pHY300PLK (4.9 kb) and harboring a Carlsberg protease gene (pHY300Car; 6.2 kb) have a similar transformation efficiency (Table 3; 1.2 vs. 1.5105 variants/g plasmid DNA). Interestingly, the transformation efficiency for the tetracycline resistant shuttle vectors (pHY300PLK, pHY300Car) is 5- to 10-fold higher compared to the pC194 and pHCMC04 vectors which use chloramphenicol for selection. Independence of transformation efficiency from the vector size can be explained by differences in DNA uptake between natural competent and physically treated (voltage) Bacillus cells (Brigidi et al. 1990). In case of natural competence uptake, the doublestranded DNA is first digested to single-stranded DNA by NucA nuclease (Hamoen et al. 2003), taken up through pilin-like structures at the Bacillus surface, and transported as single-stranded DNA across the membrane, complementary strand synthesis, and nick repair (Dubnau 1999; Kidane et al. 2009). This reconstruction could introduce deletions or repeats into the transformed DNA, which is undesired in most molecular biology work. Approaches to avoid DNA damage include the uptake of plasmid DNA by competent cells harboring a partially homologous resident helper plasmid to rescue the incoming DNA (Haima et al. 1990). The use of helper plasmid increases the probability of correct reconstruction of the input plasmid and potentially increases the transformation efficiency. To asses if the uptake mechanism used by Bacillus competent cells in this work was prone to damage the transformed plasmid DNA, a physical analysis of the plasmid isolated from 28 transformants was performed to investigate possible deletions and/or mutations and to test the suitability of the transformation method for directed evolution. The restriction pattern (two restriction enzymes, six fragments)

of the transformed recovered plasmid and sequencing analysis of the promoter-gene insert showed no changes compared to the input transformed plasmid (see Electronic Supplementary Material Fig. 2). The obtained results indicate that the transformed plasmid remains intact after transformation. The natural competence transformation protocols allow in contrast to many other artificial transformation protocols (Table 1, nos. 1, 2, and 5) to transform efficiently low amounts of plasmid DNA. An optimum was found between 5 and 10 ng of plasmid DNA (Fig. 3) and can likely be attributed to the number of specific receptors (~50) (Dubnau 1999) on the Bacillus surface which can be regarded as limiting factor for the natural uptake of high amounts of plasmid DNA. In summary, a transformation protocol was developed for B. subtilis DB104 as an expression host in directed protease evolution experiments. The natural competence protocol allows for the first time to generate large libraries (~105 transformants/g plasmid DNA) in B. subtilis DB104 and is furthermore robust, fast, and simple in handling. We hope that the current protocol can be extended to other Bacillus strains and will be used to encourage researchers to perform directed evolution experiments using Bacillus species as expression hosts.

Acknowledgments This work was supported by the German Government through the Bundesministerium fr Bildung and Forschung (Bioindustrie-2021, FKZ0315250) and Henkel AG & Co. KGaA.

References
Aharoni A, Griffiths AD, Tawfik DS (2005) High-throughput screens and selections of enzyme-encoding genes. Curr Opin Chem Biol 9:210216 Akamatsu T, Taguchi H (2001) A simple and rapid method for intraand interspecific transformation of Bacillus subtilis on solid media by DNA in protoplast lysates. Biosci Biotechnol Biochem 65:446448 Anagnostopoulos C, Spizizen J (1961) Requirements for transformation in Bacillus subtilis. J Bacteriol 81:741746 Brigidi P, De Rossi E, Bertarini ML, Riccardi G, Matteuzzi D (1990) Genetic transformation of intact cells of Bacillus subtilis by electroporation. FEMS Microbiol Lett 55:135138 Chang S, Cohen SN (1979) High frequency transformation of Bacillus subtilis protoplasts by plasmid DNA. Mol Gen Genet 168:111 115 Dubnau D (1991) Genetic competence in Bacillus subtilis. Microbiol Rev 55:395424 Dubnau D (1999) DNA uptake in bacteria. Annu Rev Microbiol 53:217244 Haima P, Bron S, Venema G (1990) Novel plasmid marker rescue transformation system for molecular cloning in Bacillus subtilis enabling direct selection of recombinants. Mol Gen Genet 223: 185191

Appl Microbiol Biotechnol Hamoen LW, Venema G, Kuipers OP (2003) Controlling competence in Bacillus subtilis: shared use of regulators. Microbiology 149:9 17 Hauser PM, Karamata D (1994) A rapid and simple method for Bacillus subtilis transformation on solid media. Microbiology 140:16131617 Kawamura F, Doi RH (1984) Construction of a Bacillus subtilis double mutant deficient in extracellular alkaline and neutral proteases. J Bacteriol 160:442444 Kidane D, Carrasco B, Manfredi C, Rothmaier K, Ayora S, Tadesse S, Alonso JC, Graumann PL (2009) Evidence for different pathways during horizontal gene transfer in competent Bacillus subtilis cells. PLoS Genet 5:e1000630 Nguyen HD, Nguyen QA, Ferreira RC, Ferreira LCS, Tran LT, Schumann W (2005) Construction of plasmid-based expression vectors for Bacillus subtilis exhibiting full structural stability. Plasmid 54:241248 Prodanovic R, Ostafe R, Scacioc A, Schwaneberg U (2011) Ultrahighthroughput screening system for directed glucose oxidase evolution in yeast cells. Comb Chem High Throughput Screen 14:55 60 Tee KL, Schwaneberg U (2007) Directed evolution of oxygenases: screening systems, success stories and challenges. Comb Chem High Throughput Screen 10:197217 Tran LS, Nagai T, Itoh Y (2000) Divergent structure of the ComQXPA quorum-sensing components: molecular basis of strain-specific communication mechanism in Bacillus subtilis. Mol Microbiol 37:11591171 Tu R, Martinez R, Prodanovic R, Klein M, Schwaneberg U (2011) A flow cytometry-based screening system for directed evolution of proteases. J Biomol Screen 16:285294 Westers L, Westers H, Quax WJ (2004) Bacillus subtilis as cell factory for pharmaceutical proteins: a biotechnological approach to optimize the host organism. Biochim Biophys Acta 1694:299310