Enzyme synthesis and export is an energy-dependent event 3. In order to select a suitable host cell for BLA overproduction, following genetic and physiological criteria were applied: 1) the ability of the strain to sporulate should be poor in order to extend the duration of BLA production; 2) the strain should form little or no lichenysin in order to reduce consumption of ATP and the amino acid pool; 3) the strain should not clump during cultivation to maintain efficiency in a production bioreactor; 4) the strain should grow well on either low-cost fermentation medium containing a high substrate concentration appropriate for an industrial process; 5) the strain should contain no native plasmids but be sensitive to kanamycin or tetracyclin to facilitate further genetic manipulation and 6) the strain should be amenable to transformation to enable genetic modification.

Following a selection procedure described above, based on these criteria, a candidate strain designated as CBB0302 was selected out of a total of 526 B. licheniformis isolates (Figure 1). The strain produced catalase, amylase and protease, utilized citrate, propionate and nitrate, grew in 7% NaCl, and at 50°C but not at 60°C, typical of B. licheniformis strains. Notably, fewer than 0.5% of the cells in culture formed spores and no cell clumping was found after 72 h cultivation. Furthermore the strain did not produce lichenysin-like pigments and harboured no native plasmids. The cell shape of the strain CBB0302 was identical to that of B. licheniformis CICIM B30306, an industrial BLA-producing strain, when both strains were cultivated in LB medium for 10-12 h at 45°C and 220 rpm but the average cell volume was 30-40% less (Figure 2).

The transformation efficiency of the natural B. licheniformis cells is poor and strains routinely need long periods with difficulty to finally obtain a desired transformant [19; our own unpublished results]. This is mainly due to the existence of two type I restriction modification systems (RMS) in B. licheniformis 2021. Single as well as double knock-outs of the RMS resulted in strains being readily transformable with plasmids isolated from Bacilli. Introduction of shuttle plasmids isolated from Escherichia coli is routinely possible when the double mutant B. licheniformis MW3 (ΔhsdR1, ΔhsdR2) was used in transformation experiments 19.

A strain CBBD302 was developed by deletion of a type I RMS locus in strain CBB0302 by using homolog-mediated recombination according to the method described by Waschkau et al 19. The growth and secretion of extracellular enzymes by this strain were unaffected by the deletion (data not shown). The transformation efficiency with a shuttle plasmid pHY-300PLK isolated from E. coli was significantly improved and 42 cfu/μg DNA in strain CBBD302 but only 5 cfu/μg DNA in strain CBB0302 were attained.

In order to test BLA production in B. licheniformis CBBD302, a recombinant plasmid pHY-amyL was constructed. A 1.6 kb fragment containing B. licheniformis B0204 amyL coding for the mature BLA peptide was recovered from pET28a-amyL_NEW by PCR, inserted into the EcoRI and SmaI sites of pHY-WZX and functionally tested in E. coli (Figure 3), yielded hybrid plasmid pHY-amyL. Subsequently, pHY-amyL was transferred into B. licheniformis CBBD302, yielding B. licheniformis CBBD302 (pHY-amyL). The BLA production was carried out in LB supplemented with 40 g/l lactose in shake-flasks (Figure 4). B. licheniformis CBBD302, B. licheniformis B30306, B. licheniformis B0204 and B. licheniformis CBBD302 (pHY-amyL) produced 0.1, 0.7, 0.9 and 2.6 mg BLA per ml growth medium, respectively. Strain CBBD302 carrying pHY-amyL produced a 26-fold improvement in BLA production compared to the parent strain CBBD302 and about three times compared to the B. licheniformis B0204. Although the growth rate was slower than its parent CBBD302, strain CBBD302 (pHY-amyL) grew significantly faster than industrial strains B0204 and B30306 did (Figure 4).

Bacterial extracellular enzyme production is a complex process, in which the efficiencies of transcription and translation of the enzyme-encoding genes as well as protein translocation define the enzyme concentration in the growth medium and are under control of the bacterial host. A B. licheniformis strain has been found to have a specific capacity maximum for protein synthesis and secretion by the introduction of different copy numbers of amyL in B. licheniformis B0204 3. Evidently the capacity could be improved since B. licheniformis CBB0302 with physiological properties of less than 0.5% sporulation, no production of lichenysin-like pigments, reduced nutrient requirement for growth, smaller cell volume, identical cell shape and no cell clumping during propagation gave an increased BLA yield.

α-Amylase production is subjected to catabolite repression by glucose and other sugars, similar to most inducible enzymes. Therefore, the use of glucose in the production of α-amylase in certain cases is problematic 2223. A number of other substrates such as lactose 24 have also been used for the production of α-amylase. The effect of the addition of different carbon sources to the fermentation medium on BLA production was investigated and the results are summarized in Figure 5. BLA secretion by strain CBBD302 carrying pHY-amyL varied between carbon sources and the lactose gave the highest BLA concentration and glucose, starch and corn cob hydrolysate supported a 70~80% BLA production level compared to lactose (Figure 5). These results indicate that many various carbon sources can be used for BLA production and BLA synthesis in CBBD302 is not subject to catabolite repression.

Yeast extract alone or in conjunction with other nitrogen sources such as bactopeptone and ammonium sulfate has been used for the production of α-amylase from Bacillus sp. 25. In this study, the effect of various nitrogen sources in the fermentation medium on BLA production was investigated. When soybean meal, fish meal or (NH₄)₂SO₄ was added as nitrogen source for BLA production to the fermentation medium, lower respective BLA production levels of 95.8, 28.6 and 25.3% were obtained relative to cottonseed meal. Much higher BLA production levels were obtained with cottonseed meal and/or soybean meal supplemented 0.01 mol/l ammonium sulphate as nitrogen source (respectively 132 and 140% of cottonseed level).

The maximum production of BLA by B. licheniformis CBBD302(pHY-amyL) of 17.6 mg/ml was obtained when the strain was cultivated in the fermentation medium consisting of 40 g/l lactose, 25 g/l soybean meal, 20 g/l cottonseed meal, 30 g/l corn-steep liquor and 0.01 mol/l ammonium sulfate in a pH-controlled 15 l bioreactor fermentations (Figure 6). The BLA production was boosted by the 40 g/l lactose when fed at the 54^th hour.

Plasmid pHY300PLK 26 was used to determine transformation efficiency and plasmid pHY-WZX 27 was used as an expression vector in B. licheniformis. E. coli XL1 was used as a host cell for functional identification of the recombinant plasmid. Molecular biology methods were described previously 28. A 1.6 kb fragment containing amyL coding for the mature peptide as well as its 180 bp downstream sequence was amplified with primers F2-EcoRI (5'-CGGAATTCCTTAATGGGACGCTGATGC-3') and R1-SmaI (5'-TACCCGGGTACATCAGATAACGTTGCC-3') using pET28a-amyL_NEW 29 as template. The amplified product was purified and digested with EcoRI and SmaI and subsequently cloned into the same sites of pHY-WZX to yield recombinant pHY-amyL. B. licheniformis isolates as well as B0204 and B030306 were purchased from CICIM-CU http://cicim-cu.jiangnan.edu.cn. CBBD302 was developed from a native B. licheniformis isolate CBB0302, in which the restriction modification system locus was deleted according to the method described by Waschkau et al 19. B. licheniformis CBBD302(pHY-amyL) was CBBD302 harboring pHY-amyL by using electroporation 30.

B. licheniformis strains were recovered from a culture collection held at -70°C and single colonies were picked onto LB plate and cultivated at 45°C for up to 72 h. Colonies with no or reduced lichenysin formation were picked onto LB plate supplemented with 5 μg/ml kanamycin or 10 μg/ml tetracycline and cultivated for 48 h. Colonies that failed to grow were examined for the existence of the native plasmids by plasmid extraction and agarose electrophoresis as described previously 28. Strains carrying no native plasmids and sensitive to kanamycin and tetracycline were cultivated in LB containing 2% glucose at 45°C for up to 72 h and their degree of sporulation and cells dispersion were checked by light microscopy. Strains with sporaulation rate less than 5% after 72 h and with no cell clumping were inoculated onto LB supplemented 30% glucose or 30% starch and cultivated at 45°C for 24 h. The rapidly growing strains were selected and cultivated at 45°C and 24 h on a nutrient limited medium consisting of 0.02% peptone, 0.01% yeast extract, 1% NaCl and 0.01% glucose. Those strains that grew well on the nutrient limited medium were examined for their biological properties including the major secreted enzymes, genetic transformation efficiency as well as plasmid stability as described by Zhuge & Wang 31. The cells were examined with a Quanta-200 scanning electron microscope (FEI, Netherland).

E. coli XL1 was cultivated at 37°C in LB medium. As required, 100 μg/ml ampicillin and/or 25 μg/ml kanamycin were added to the medium. B. licheniformis B0204, B030306 and CBBD302 were cultivated at 42°C in LB medium. For the shake-flask fermentation evaluation, B. licheniformis strains were grown in 500 ml Erlenmeyer flasks containing 50 ml LB supplemented with 40 g/l lactose at 42°C and 220 rpm. For optimization studies the fermentation medium consisted of 30 g/l corn-steep liquor, 30 g/l nitrogen source (cottonseed meal, soybean meal and fish meal, and former two combination with 0.01 mol/l (NH₄)₂SO₄ and 40 g/l) carbon source (lactose, glucose, starch or corn cob hydrolysate) at pH 6.0 was used. For bioreactor cultivation studies, a Biostat (B. Braun, Melsungen, Germany) with a 15 l working volume was used. The bioreactor was inoculated with 5% (v/v) broth and cultivated at 42°C with a controlled pH of 6.0.

For shake flask cultivation, samples for cell density and BLA activity were taken at regular intervals. The optical density (OD_{600 nm}) was measured in triplicate with an Ultrospec 3100 pro spectrophotometer (Amersham Pharmacia, UK). SDS-PAGE was performed using a Mini Protean 3 apparatus (Bio-Rad, USA). Proteins were stained by Coomassie Brilliant Blue G250. Directly after sampling, BLA activity was measured spectrophotometrically (Ultrospec 3100 pro, Amersham Pharmacia, UK) as described previously 32. One unit was defined as the amount of enzyme that hydrolyzes 1 mg water soluble corn starch per minute at 70°C and pH 6.0. The amount of BLA protein (g/ml) in the medium was calculated using BLA specific activity parameter of 1 mg BLA is equal to 996 U.