Cell-associated degradation affects the yield of secreted engineered and heterologous proteins in the Bacillus subtilis expression system

Christina L. Jensena,1, Keith Stephensonb,2, Steen T. Jørgensen1 and Colin Harwood2

Bacterial Gene Technology, Novo Nordisk A/S, DK-2880 Bagsværd, Denmark1
Department of Microbiology and Immunology, The Medical School, University of Newcastle upon Tyne, Framlington Place, Newcastle upon Tyne NE2 4HH, UK2

Author for correspondence: Steen T. Jørgensen. Tel: +45 44 44 88 88. Fax: +45 44 42 73 03. e-mail: stjq{at}novo.dk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
A series of chimeric {alpha}-amylase genes derived from amyL, which encodes the liquefying {alpha}-amylase from Bacillus licheniformis, were constructed in vitro using gene splicing techniques. The gene constructs were cloned in Bacillus subtilis, where their ability to direct the synthesis and secretion of active {alpha}-amylase was determined. Detectable {alpha}-amylase activity was observed for some, but not all, of the chimeric proteins. Studies on the secretion of wild-type AmyL and its chimeric derivatives revealed that, whilst these proteins were stable in the extracellular milieu, all were subject to some degree of degradation during secretion. The chimeric enzymes were degraded to a greater extent than the native enzyme. These findings suggest that cell-associated proteolysis is a significant problem affecting the use of B. subtilis as host bacterium for the production of heterologous proteins.

Keywords: Bacillus subtilis, chimeric proteins, protein secretion, proteolysis, industrial enzymes

Abbreviations: SOE, splicing by overlap extension

a Present address: Carlsberg Research Laboratory, Gamle Carlsberg vej 10, DK-2500, Copenhagen, Denmark.

b Present address: Department of Molecular and Experimental Medicine, The Scripps Research Institute, 10550 N Torrey Pines Road, La Jolla, CA 92037, USA.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacillus subtilis and the closely related bacilli Bacillus amyloliquefaciens and Bacillus licheniformis are capable of secreting enzymes into the culture supernatant at concentrations in excess of 10 g l-1 (Ferrari et al., 1993 ), and these species are currently used in the large-scale manufacture of industrial enzymes. However, attempts to use B. subtilis, for which an extensive knowledge of physiology and molecular biology exists, to secrete heterologous proteins at commercially significant concentrations have, with few exceptions, been disappointing (Harwood, 1992 ).

Today’s industrial enzymes are frequently subjected to protein engineering to obtain products which exhibit improved characteristics with respect to specific industrial applications. Whilst such approaches may result in enzymes with significantly improved enzymic performance under application conditions, the introduced modifications are not always beneficial with regard to production and secretion from the intended host strain. It is not uncommon for the development of new products with commercial potential to be abandoned due to adverse production economics. One factor contributing to low yields is proteolytic degradation of secreted enzymes by extracellular proteases produced by the host. Strains in which genes encoding extracellular proteases have been deleted have been successful in reducing the proteolysis of certain products in the culture medium (Sloma et al., 1989 , 1991 ; Wu et al., 1991 ), although the integrity of such strains can be compromised (Stephenson et al., 1999 ). An alternative approach has been to identify protease-labile sites in the product and to modify these sites by protein engineering (Boguslawski et al., 1996 ). However, these approaches may only improve the yield of products which are already efficiently secreted.

Conceptually, the Bacillus secretory pathway (Simonen & Palva, 1993 ) can be divided into three functional stages: (i) early stages involving the synthesis of secretory pre-proteins, their interaction (if any) with chaperones and binding to the secretory translocase; (ii) translocation across the cytoplasmic membrane via the translocase; (iii) late stages, including removal of the signal peptide, release from the translocase, folding on the trans side of the cytoplasmic membrane and passage through the cell wall.

A rate-limiting step in the secretion of certain proteins from B. subtilis is release from the cytoplasmic membrane and concomitant folding into the native conformation (Petit-Glatron et al., 1993 ; Chambert et al., 1995 ; Bolhuis et al., 1999a ). The folding step is modulated by ions such as Ca2+, Fe3+ and H+ at the outer surface of the cytoplasmic membrane (Chambert et al., 1995 ; Leloup et al., 1997 ) which promote the rapid folding of the secretory proteins. In addition, folding may be assisted by PrsA, a lipoprotein located at the outer surface of the cytoplasmic membrane (Kontinen & Sarvas, 1993 ). Although its precise role in secretion is not clear, it is thought to act as an extracellular chaperone, assisting the folding of a subset of secretory proteins as they emerge from the translocase on the trans side of the cytoplasmic membrane (Kontinen et al., 1991 ; Kontinen & Sarvas, 1993 ; Jacobs et al., 1993 ). In addition to these extrinsic factors, the efficiency of secretory protein folding is also dependent on the characteristics of the individual secretory proteins.

To facilitate studies on the influence of secretory protein charge on passage through the negatively charged cell wall of B. subtilis, we have constructed a series of chimeric {alpha}-amylases with different isoelectric points (pI). AmyL, the {alpha}-amylase of B. licheniformis (Ortlepp et al., 1983 ; Laoide et al., 1989 ) is efficiently expressed and secreted in B. subtilis, and we took advantage of the observations that functional chimeric proteins could be constructed using {alpha}-amylases from B. licheniformis (AmyL), B. amyloliquefaciens (AmyQ) and Bacillus stearothermophilus (AmyS) (Gray et al., 1986 ; Diderichsen et al., 1987 ; Jørgensen et al., 1990 ) to construct genes encoding chimeric {alpha}-amylases with modified net charge. This was achieved by replacing segments of the amyL gene with corresponding segments from the amyQ and amyS genes. This paper reports the design and construction of genes encoding chimeric {alpha}-amylases with altered pI values, and the enzymic characterization and production in terms of transcription and yield of their protein products.

Pulse–chase and immunoprecipitation techniques were used to follow the secretion of wild-type and chimeric enzymes from B. subtilis, as well as the secretion of wild-type AmyL from B. licheniformis. These studies revealed a significant cell-associated degradation of amylase synthesized in B. subtilis, a degradation which was not observed when wild-type AmyL was produced in B. licheniformis.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
Escherichia coli SJ2 (Diderichsen et al., 1990 ) and E. coli XL-1 Blue (Stratagene) were grown at 37 °C in TY medium (Diderichsen et al., 1990 ). B. subtilis DN1885 (amyE amyR2), a derivative of B. subtilis 168 strain RUB200 (Yoneda et al., 1979 ), was grown at 37 °C in TY medium containing 0·4% (w/v) glucose. B. licheniformis ATCC 9789 was obtained from the American Type Culture Collection, Manassas, VA, USA.

Strains of B. subtilis DN1885 harbouring plasmids were maintained on LB agar plates (Maniatis et al., 1982 ) supplemented with 0·4% glucose and 6 µg chloramphenicol ml-1. Strains of E. coli SJ2 harbouring plasmids were maintained on LB agar plates supplemented with 100 µg ampicillin ml-1. To compare the yields of {alpha}-amylase released from B. subtilis, the strains were also grown in a complex liquid medium (Jørgensen et al., 1991b ) for 7 d at 37 °C with shaking. For pulse–chase experiments, the B. subtilis strains were grown in Spizizen’s minimal medium (Spizizen, 1958 ) modified by the replacement of glucose with 1% (w/v) ribose, a non-catabolite-repressing carbon source, and 1% (w/v) xylose to induce the synthesis of {alpha}-amylase.

The plasmids used throughout the work are given in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. Plasmids used in this work

 
Construction of plasmids.
pSX63, containing a xylose-inducible amyL gene, was used as the basis for the development of an expression system for the production of chimeric proteins. To facilitate the cloning of the engineered {alpha}-amylase genes as PstI–HindIII fragments, an unwanted PstI site between the cat and xylR gene was removed in three stages: (i) pSX63 was digested with ClaI and the large fragment purified, religated and transformed into B. subtilis DN1885, resulting in plasmid pKS200; (ii) pKS200 was digested at the now unique PstI site and the overhanging ends trimmed with S1 nuclease to form blunt ends. The blunt ends were ligated prior to transforming into B. subtilis DN1885, resulting in plasmid pCJ72; (iii) finally, the EcoRI–HindIII amyL fragment from pSX63 was cloned into the EcoRI and HindIII restriction sites in pCJ72 in B. subtilis DN1885 to produce pCJ92 (Fig. 1a).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 1. Schematic diagrams of the expression vectors pCJ92 (a) and pKS301 (b). The large EcoRI–BglII DNA fragment of pCJ92 encodes the elements required for the inducible expression of the various {alpha}-amylase genes. The smaller EcoRI–BglII fragment of pCJ92 contains the element necessary for the replication of the plasmid in B. subtilis. The shuttle vector pKS301 was generated from pCJ92 by introducing a fragment containing the origin of replication and the ß-lactamase gene from pUC19 into the unique BglII restriction site.

 
To generate an E. coli/B. subtilis shuttle vector, the SspI to SapI (positions 2501 to 683) fragment of pUC19, containing the origin of replication and ß-lactamase gene (Yanisch-Perron et al., 1985 ), was amplified by PCR using oligonucleotide primers with BglII restriction sites incorporated into their 5' ends. The PCR product was digested with BglII and cloned into the unique BglII site of pCJ92. The resulting plasmid, pKS301, was able to replicate in both E. coli and B. subtilis (Fig. 1b).

DNA manipulation.
Plasmid DNA was prepared from strains of E. coli and B. subtilis by an alkaline-phenol extraction method (Kieser, 1984 ). Enzymes for restriction, ligation, end-filling and exonuclease treatments were used as recommended by the supplier (New England Biolabs). DNA fragments were purified from agarose gels using either Spin-X centrifuge tube filters (Costar) or a QIAquick Gel Extraction kit (Qiagen), as recommended by the manufacturers.

Transformation.
Competent cells of B. subtilis were prepared and transformed with plasmid DNA by the procedure of Yasbin et al. (1975) and transformants selected on LB agar plates containing chloramphenicol (6 µg ml-1).

E. coli cells were electro-transformed in a Bio-Rad Gene Pulser according to the manufacturer’s instructions. Transformants were selected on plates of LB agar supplemented with 100 µg ampicillin ml-1.

PCR.
Oligonucleotide primers used for the construction of genes encoding chimeric {alpha}-amylases were synthesized on an Applied Biosystems DNA synthesizer and purified before use.

Taq DNA polymerase (SuperTaq, HT Biotechnology) was used for PCR amplifications according to the supplier’s instructions. The reaction cycles were as follows: 1x(2 min at 94 °C); pause for the addition of 0·5 µl SuperTaq DNA polymerase; 15x(30 s at 94 °C, 30 s at 55 °C, 30 s at 72 °C); 1x(5 min at 72 °C).

Splicing by overlap extension (SOE; Horton et al., 1989 ), a PCR-based in vitro method, was used for fusing specific fragments of DNA without the need for restriction sites. Briefly, four oligonucleotide primers were synthesized; primers at the fusion junction included non-annealing extensions at their 5' ends which were complementary to each other, whilst the flanking primers were complementary to their respective template DNA. The DNA regions were amplified in separate PCR reactions and the products mixed. Since one strand of each of these PCR products had complementary sequences in their 3' ends (i.e. at the fusion junction), they served as primers in the first step of the SOE reaction. In the second step, the fused product was amplified using the flanking primers. The spliced fragments were purified from agarose gels using Spin-X columns. The number of SOE steps was dependent on the number of fragments required for the construction of the final spliced product.

The following primers were used in the construction of amyLQS50 chimeras (sequence 5'–3'): 50/1, GGGGGAATTCTTGCTGCCTCATTCTGCAGCAGCG; 50/2, GTACCCTACGTCGCTGCGGCTTGTTCCCTTATATGCCGGGGG; 50/3, ACAAGCCGCAGCGACGTAGGGTACGGTGCTTACGACCTTTATGATTTAG; 50/4, GGGGGGAGCTCTTCAACCGCGGTTACATCTTCGG; 50/5, GGGGGGAGCTCTGTAACCGCGGTTGAAGTCGATC; 50/6, CCAGCGCCACTTAAAGCTGGAGTAGGTGTTGCCCCGCCCGGGAAAATGAAAATGTGTCCAGG; 50/7, CCCGGGCGGGGCAACACCTACTCCAGCTTTAAGTGGCGCTGGTACCATTTTGACGGAACCGATTG; 50/8, GGGGGGGATCCTGGTCATAATCGATGTCGGCATAC; 50/9, GGGGGGGATCCATGCCGACATCGATTATGACCATC; 50/10, GTAAATAGCGGCTTGCCAGTTTTTTCCCTGACATGATTAAG; 50/11, TTAATCATGTCAGGGAAAAAACTGGCAAGCCGCTATTTAC; 50/12, CTCATATCATAGCCGCCTCCTGATTTGGAAGCGGTATAAAATTTG; 50/13, TTTATACCGCTTCCAAATCAGGAGGCGGCTATGATATGAG; 50/14, CCATGTTTGGACAGTCGACTCAAGC; 50/15, GGGGCAATCGCTTGAGTCGACTGTCCAAAC; 50/16, TTCCTTTGGCGATGTCCCTTTCGTCCCGTACATATCCC; 50/17, GGGACATCGCCAAAGGAAATTCCCTCACTGAAACACAAAATTG; 50/18 GGGGGGCATGCCTGTGCTCCGTACGCGTACTGTTTTCTG; 50/19, GGGGGGGATCCCGAGAAAACAGTACGCGTACGGAGCAC; 50/20, TGTTGTTTGCCAACGTACATTCGCTTTGCCCCACCGGGTC; 50/21, GACCCGGTGGGGCAAAGCGAATGTACGTTGGCAAACAACAC; 50/22, CCGTCCTCTCTGCTCTTCTAAGGCCATGCCACCAACCGTG; 50/23, CACGGTTGGTGGCATGGCCTTAGAAGAGCAGAGAGGACGG; 50/24, GACAGCAAGCTTTTTCAATTCATCCG.

Chimeric block 1 was constructed using primers 50/1–50/4, and wild-type block 1 using 50/1+50/4. Chimeric block 2 was constructed using primers 50/5–50/8, and wild-type block 2 using 50/5+50/8. Chimeric block 3 was constructed using primers 50/9–50/14, and wild-type block 3 using 50/9+50/14. Chimeric block 4 was constructed using primers 50/15–50/18, and wild-type block 4 using 50/15+50/18. Chimeric block 5 was constructed using primers 50/19–50/24, and wild-type block 5 using 50/19+50/24.

The following primers were used in the construction of amyLQS55 chimeras (sequence 5'–3'):

55/1, AAACGGAGCTCCGCCCGATTGCTGACG; 55/2, GCATCATTCTGCAATCGGGTCCATAACGTGCCATCATCCGGCATGTACCATTCAAAATACTGC; 55/3, GACCCGATTGCAGAATGATGCGGAACATTTATCGGATATCGGAATCACTGCCGTCTGGATCCC; 55/4, GAATGGCTTGAAGATATTGTCCTTTTGTGCCGTACTTTGTCCG; 55/5, CGGTAGAGCTCGCCGTCTGGATCCCCCC; 55/6, CGGACAAAGTACGGCACAAAAGGACAATATCTTCAAGCCATTCAAGCC; 55/7, GATTTGATATTCCTCCGAGATTTCTTGGTTGCGGTCG; 55/8, GACCGCAACCAAGAAATCTCGGAGGAATATCAAATCAAAGCGTGGACGGATTTTCGTTTTCCGGGGGCCGGCAGC; 55/9, CTTAAAGATCTGCTGCGACATCAGGATGG; 55/10, CGGCAGCACATACAGCG; 55/11, CATTAACGTGCGCATATCATAGCCGCCTCCCTGTGTCG; 55/12, CGACACAGGGAGGCGGCTATGATATGCGCACGTTAATGACC; 55/13, GGACAGTCGACTCAAGCGATTGCCCCGGCTGTGTGTCATGATTATCAACGAAGG; 55/14, CCATGATACACAGCCCGG; 55/15, CGATTTTGCTTTTCAGCGAAGGAATTTCGCGCTGGG; 55/16, CCCAGCGCGAAATTCCTTCGCTGAAAAGCAAAATCG; 55/17, GGCATTTTTCAGGCCGGCGTACATCCATTTGCTTCC; 55/18, GGAAGCAAATGGATGTACGCCGGCCTGAAAAATGCC; 55/19, CGTCCTCTCTGCTCTTCTATTTCTGAACATAAATGGAGACG; 55/20, CCATTTATGTTCAGAAATAGAAGAGCAGAGAGGACG; 55/21, GACAGCAAGCTTTTTCAATTCATCCG.

Chimeric block 1 was constructed using primers 55/1–55/4, and wild-type block 1 using 55/1+55/4. Chimeric block 2 was constructed using primers 55/4–55/9, and wild-type block 2 using 55/5+55/9. Chimeric block 3 was constructed using primers 55/10–55/13, and wild-type block 3 using 55/10+55/13. Chimeric block 4 was constructed using primers 55/14–55/21, and wild-type block 4 using 55/14+55/21.

DNA sequencing.
DNA sequences were determined by the dideoxy chain termination method (Sanger et al., 1977 ) using the Sequenase 2.0 kit (United States Biochemicals) and [35S]dATP{gamma}S [>1000 Ci mmol-1 (37 TBq mmol-1); DuPont NEN].

{alpha}-Amylase assay.
{alpha}-Amylase activity was determined using the Phadebas Amylase Test kit (Pharmacia Upjohn), using the protocol provided with the assay kit and purified AmyL as a standard. One unit of activity was defined as the amount of enzyme catalysing the hydrolysis of 1 µmol glycoside linkage min-1 at 37 °C.

Dot-blot hybridization of {alpha}-amylase-specific mRNA.
Cultures of B. subtilis were grown in 30 ml Spizizen’s minimal medium with 1% ribose as the carbon source and 1% xylose to induce the synthesis of {alpha}-amylase. Samples were removed for the determination of growth and {alpha}-amylase activity, and for the isolation of total RNA.

RNA was extracted from the cells of 1 ml culture samples using the RNeasy kit (Qiagen) according to the manufacturer’s protocol. RNA samples were stored at -75 °C in the presence of RNase inhibitor (RNasin; Promega). Contaminating DNA was removed with RNase-free DNase I (Boehringer Mannheim). RNA was quantified by spectrometry (GeneQuant II; Pharmacia Upjohn). For each sample, 3 µg total RNA was dot-blotted onto a positively charged nylon membrane (Boehringer Mannheim) using standard protocols (Sambrook et al., 1989 ).

The DNA probes, corresponding to common regions in the 5' ends of amyL, amyLQS50 and amyLQS55, were amplified from amyL by PCR. Probe 1, used to probe RNA from strains DN1885 xylR::pKS401 (expressing AmyL) and DN1885 xylR::pKS405B (expressing AmyLQS50-5), was amplified using oligonucleotide primers LQS50/27 (5'-GGT GCT TAC GAC CTT TAT G-3') and LQS50/28 (5'-AAA TGA AAA TGT GTC CAG GC-3'). Probe 2, used to probe RNA from strains DN1885 xylR::pKS401 (expressing AmyL), DN1885 xylR::pCJ201 (expressing AmyLQS55-3) and DN1885 xylR::pCJ272 (expressing AmyLQS55-6), was amplified using oligonucleotide primers KS15 (5'-ACT GCA ATT GGA CGG TTT CCG-3') and KS16 (5'-CTG TGT CGA TGC AGC ATG G-3').

The probes were radiolabelled with [{alpha}-32P]dATP (Amersham) using the Rediprime kit (Amersham) and were hybridized to immobilized RNA as described by Sambrook et al. (1989) . The amount of probe bound to total RNA samples was quantified by phosphor-imaging (Phosphor-Imager; Molecular Dynamics) and analysed using the ImageQuant software (version 3.22; Molecular Dynamics).

Pulse–chase labelling and immunoprecipitation.
Cultures of B. subtilis growing exponentially to an optical density at 660 nm of approximately 0·5 were pulse-labelled with L-[35S]methionine as described previously (Jensen, 1997 ; Stephenson & Harwood, 1998 ) and samples removed at the following time intervals: 0, 0·5, 1, 2, 3, 4, 5, 10 and 15 min. Labelled {alpha}-amylases in whole-culture samples and released into the growth medium [obtained by filtration through 0·45 µm PVDF filters (Whatman)] were immunoprecipitated and subjected to SDS-PAGE and fluorography. The relative amounts of {alpha}-amylase present at each time interval were determined by phosphor-imaging (Phosphor-Imager; Molecular Dynamics) using the associated software (ImageQuant version 3.22; Molecular Dynamics).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Computer-aided design of chimeric {alpha}-amylases with altered net charge
Chimeric proteins with pI values higher (AmyLQS50) and lower (AmyLQS55) than that of AmyL were designed in silico (PC/GENE version 5.16; Genofit) by swopping specific segments of the mature portion of AmyL with the corresponding segments from AmyQ or AmyS (Palva, 1982 ; Takkinen et al., 1983 ; Gray et al., 1986 ; Jørgensen et al., 1991a ). The segments were chosen so that each swop led to an increase or decrease in the pI, as calculated by the ChargPro programme in the PC/GENE software package. For example, the carboxy-terminal extension of AmyS includes a large number of positively charged amino acid residues so the inclusion of this ‘tail’ in the chimeric {alpha}-amylase AmyLQS50 increases its predicted pI.

The chimeric proteins were designed prior to the resolution of the crystal structure of AmyL (Machius et al., 1995 ), but secondary structure predictions were used to ensure that the boundaries of swopped segments were not internal to ß-strands or {alpha}-helices in an attempt to minimize perturbations to the structure of the protein.

A strategy was developed that involved a relatively small number of segment swops to generate variants of AmyL, AmyLQS50 and AmyLQS55, with predicted pI values of 6·0, 8·1 and 4·7, respectively (Fig. 2).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 2. Structural organization of the chimeric {alpha}-amylase genes. (a) The genes encoding the chimeric {alpha}-amylases, which were designed by substituting regions of AmyL ({blacksquare}) with corresponding regions from AmyQ () or AmyS (). (b) The genes encoding the chimeric {alpha}-amylases were divided into a number of DNA blocks using the unique restriction sites within the amyL gene. Distances, in nucleotides, were calculated in relation to the first codon in each gene. Asterisks indicate restriction sites created by silent mutation during the construction of the DNA blocks. amyLQS50-5 is composed of wild-type block 1 (amyL 79-213), and chimeric blocks 2, 3, 4 and 5. amyLQS55-3 is composed of wild-type blocks 1 (amyL 79-213), 2 (amyL 214-561) and 4 (amyL 1096-1798), and chimeric block 3. amyLQS55-6 is composed of wild-type block 2 (amyL 214-561), and chimeric blocks 1, 3 and 4.

 
Construction of genes encoding chimeric {alpha}-amylases
The native amyL gene has an unique PstI site located within the signal peptide coding region, and an unique HindIII site downstream (3') of its transcription terminator. Consequently, the chimeric genes were designed as in-frame PstI–HindIII DNA fragments, and the AmyL signal peptide was used to direct the secretion of all the chimeric {alpha}-amylases and the reconstructed AmyL.

The chimeric genes were subdivided into blocks bordered by unique restriction sites already present in amyL (Fig. 2) and synthesized using a PCR-based gene splicing method (Horton et al., 1989 ). As well as naturally occurring unique restriction sites, the gene constructions were facilitated by additional sites introduced as silent mutations: a MluI site in the case of amyLQS50 and a BamHI site in the case of amyLQS55. amyLQS50 and amyLQS55 were constructed from five and four blocks, respectively, with regions of amyL being replaced by corresponding regions of either amyQ or amyS (Fig. 2). In addition to the chimeric DNA blocks, equivalent wild-type amyL blocks were also constructed. This allowed individual chimeric blocks to be replaced with the corresponding wild-type blocks as and when required.

Individual DNA blocks were cloned into pUC19 using the unique restriction sites generated at their ends. A number of independently isolated pUC19 clones were sequenced to detect errors generated during the PCR and to ensure the fidelity of the splicing reactions. amyLQS55 block 2 included a base substitution that changed the expected amino acid residue at position 115 from alanine to glycine, whilst amyLQS55 block 4 had two base substitutions changing the expected residues at positions 403 (from serine to alanine) and 485 (from glycine to serine). These changes were accepted since they were unlikely to affect the overall charge of the respective amylases.

Assembly of chimeric {alpha}-amylase genes
The individual DNA blocks were assembled stepwise in pUC19 to generate PstI–HindIII gene fragments encoding the mature protein of the chimeric {alpha}-amylases. As illustrated in Fig. 2, a number of intermediate genes were assembled using combinations of chimeric and wild-type DNA blocks.

The assembled amyLQS55 gene fragments were cloned into plasmid pCJ92 (Fig. 1a) using B. subtilis DN1885 as the host, whilst amyLQS50 gene fragments were cloned into an E. coli/B. subtilis shuttle vector plasmid, pKS301 (Fig. 1b), using E. coli XL-1 Blue as an intermediate host prior to transfer to B. subtilis DN1885. In both cases, the assembled PstI–HindIII gene fragments replaced the pre-existing wild-type amyL gene. The presence of a MluI or BamHI restriction site within the cloned {alpha}-amylase genes was taken as evidence that the wild-type amyL gene had been replaced.

The {alpha}-amylase genes were cloned downstream of the promoter/operator (P1O) from the B. subtilis xyl operon, which enabled {alpha}-amylase production to be induced in the presence of xylose. In the absence of xylose, repression is mediated by the XylR repressor, which is also encoded by the expression plasmids pCJ92 and pKS301 (Fig. 1).

The transformants were screened on starch plates for the secretion of a functional {alpha}-amylase by staining with iodine vapour. Neither of the strains encoding the full chimeric {alpha}-amylases AmyLQS50 and AmyLQS55 showed detectable {alpha}-amylase activities. Data on intermediate chimeric {alpha}-amylases indicated that the presence of block 1 in amyLQS50 and block 2 in amyLQS55 prevented the production of detectable amounts of {alpha}-amylase activity. Constructs in which these two blocks were replaced with wild-type blocks showed {alpha}-amylase activities that could be detected on the starch plates, although there were marked differences in the sizes of their zones of hydrolysis compared to that for the strain encoding AmyL (data not shown).

The plasmid-encoded expression systems (Fig. 1) were based on the replicon of the Staphylococcus aureus plasmid pUB110, which can be relatively unstable in B. subtilis (Bron & Luxen, 1985 ; Bron, 1991 ). A more stable expression system was obtained by integrating the {alpha}-amylase genes into the chromosome of B. subtilis. The integration vectors were constructed by cloning EcoRI–BglII fragments of pCJ92 and pKS301 (Fig. 1), containing {alpha}-amylase, xylR and cat genes, into pUC19. The resulting plasmids, which lack an origin of replication that is functional in B. subtilis, were transformed into B. subtilis DN1885 and chloramphenicol resistance was used to select for cells containing integrated plasmids. Integration occurred by single, homologous recombination between the plasmid and chromosomally encoded copies of the xylR gene, resulting in the tandem duplication of this gene. The integration events were verified by Southern blot hybridization (data not shown).

Production of chimeric {alpha}-amylases in B. subtilis
B. subtilis DN1885 strains containing integrated chimeric constructs were grown in shake flasks for 7 d at 37 °C in a complex medium containing 0·2% xylose and 6 µg chloramphenicol ml-1 (conditions chosen to mimic industrial fed-batch fermentations), and the accumulation of {alpha}-amylase in the culture medium was determined by Western blotting and by enzyme activity assays. The data confirmed that strains containing the full chimeric constructs encoding AmyLQS50 and AmyLQS55 produced no detectable amylase activity and that no amylase protein could be detected in the culture supernatant by Western blotting. The active variant with highest pI, AmyLQS50-5, was present in the fermentation medium at a concentration that was about half that of AmyL, whereas the active variant with lowest pI, AmyLQS55-6, was present at a concentration that was less than 1% that of AmyL (Table 2). Western blot analysis (data not shown) confirmed the results obtained from the enzyme activity assays: reduced amounts of chimeric {alpha}-amylases accumulated in the culture supernatants as compared to AmyL.


View this table:
[in this window]
[in a new window]
 
Table 2. Amount of accumulated {alpha}-amylase secreted by B. subtilis DN1885 containing wild-type or chimeric {alpha}-amylase genes integrated into the chromosome

 
Following fermentation, stability of the chimeric enzymes in the spent culture medium was determined. The chimeric amylases were as stable as AmyL during incubation for more than 140 h, retaining more than 85% of their initial activity over this time. This indicates that the observed differences in accumulated {alpha}-amylase were not due to increased susceptibilities of the folded forms of these enzymes to the extracellular proteases produced by B. subtilis. Selected chimeric amylases were purified and, relative to AmyL, no major differences were observed with respect to thermostability and pH and temperature profiles, whilst the measured pI values showed the desired differences, being 5·0, 7·0 and 10·0 for enzymes AmyLQS55-6, AmyL and AmyLQS50-5, respectively (Stephenson et al., 2000 ).

The production of the chimeric {alpha}-amylases were also tested under other conditions, for example during growth in minimal medium and TY medium (data not shown). B. subtilis DN1885 strains encoding the chimeric {alpha}-amylases grew at identical growth rates as compared to the wild-type strain. However, when the yields of secreted {alpha}-amylases were compared, the strains encoding the chimeric {alpha}-amylases produced a significantly reduced level of {alpha}-amylase activity. This was observed for the production of {alpha}-amylase in both exponential and stationary phases.

Transcription of genes encoding AmyL and chimeric enzymes
Quantitative dot-blot hydridization was used to determine whether the observed variations in the amount of secreted chimeric enzymes were due to differences in the transcriptional activities of their genes. Although all the genes were transcribed from the same promoter system, differences in mRNA levels might have arisen from the unwitting incorporation of secondary structures or RNase processing sites during chimeric gene construction, or from variations in gene copy number. Cultures of B. subtilis DN1885, expressing AmyL or representative high and low pI {alpha}-amylases from chromosomally located expression cassettes, were sampled in exponential and stationary phases to determine {alpha}-amylase activities and to quantify {alpha}-amylase-specific mRNA. Separate probes were designed for genes encoding high and low pI {alpha}-amylases, using regions that were conserved in both amyL and the constructed genes.

The amount of {alpha}-amylase activity detected in the culture medium varied from <60 (the minimum detection limit) to >1400 enzyme units per culture OD unit whilst the relative amount of mRNA remained relatively constant, varying by no more than about twofold (Fig. 3). In the case of the strain encoding AmyLQS55-6 (Fig. 3a), it was never possible to detect any {alpha}-amylase activity, despite the fact that the gene was transcribed at a level similar to that of amyL. These data confirm that the reduced levels of {alpha}-amylase activity observed for the chimeric enzymes were due to post-transcriptional events.



View larger version (24K):
[in this window]
[in a new window]
 
Fig. 3. Transcriptional activity of amyL and chimeric {alpha}-amylase genes during growth. (a–c) amyL (c) and low pI variants amylQS55-6 (a) and amyLQS55-3 (b) hybridized with probe 2. (d–e) amyL (e) and high pI variant amyLQS50-5 (d) hybridized with probe 1. {square}, Growth; {lozenge}, {alpha}-amylase specific activity, normalized to the OD600 of the culture; bars, relative amounts of {alpha}-amylase-specific mRNA.

 
Pulse–chase analysis of wild-type and chimeric {alpha}-amylase secretion
The secretion of the chimeric {alpha}-amylases was studied using a combination of pulse–chase and immunoprecipitation techniques. Strains of B. subtilis with chromosomally integrated {alpha}-amylase genes were used since they provided an expression system that was stable in the absence of antibiotic selection.

The precursor of AmyL was processed rapidly into the mature form (Fig. 4a, b); immediately after the addition of the chase (T0 min) >70% of the labelled AmyL was already in the mature form. The amount of labelled mature AmyL peaked at about 1 min after the addition of the chase and then declined until, at about 6 min, it reached a constant level at around 25% of the total amylase synthesized during the pulse (Fig. 4a, b). This indicates that a significant fraction of newly synthesized AmyL was degraded during or shortly after translocation across the cytoplasmic membrane, and possibly before release into the culture medium. AmyL was detected in the culture medium within 1 min of the addition of the chase and reached a constant level after about 5 min (Fig. 4a, b). AmyL appeared to be stable once it had reached the culture medium.



View larger version (36K):
[in this window]
[in a new window]
 
Fig. 4. Secretion of wild-type and chimeric {alpha}-amylases. (a, c, e, g) Autoradiographs from representative pulse–chase experiments. ‘Whole culture’ indicates {alpha}-amylase immunoprecipitated from whole-culture samples. ‘Medium’ indicates {alpha}-amylase immunoprecipitated from culture medium only. (b, d, f, h) Quantification by phosphor-imaging of {alpha}-amylases at intervals following the chase, expressed as relative amounts. {square}, Whole culture; {lozenge}, medium. (a, b) B. subtilis expressing wild-type AmyL; (c, d) B. subtilis expressing AmyLQS55-3; (e, f) B. subtilis expressing AmyLQS50-5; (g, h) B. licheniformis expressing AmyL.

 
For the low pI chimeric {alpha}-amylase, AmyLQS55-3, degradation of the newly synthesized protein was much more pronounced than was observed for AmyL (Fig. 4c, d). Approximately 90% of the labelled {alpha}-amylase was rapidly degraded. As with AmyL, AmyLQS55-3 was detected in the culture supernatant after about 2 min of chase. Similar observations were made for the high pI chimeric {alpha}-amylase, AmyLQS50-5; newly synthesized AmyLQS50-5 was degraded to a similar extent as compared with the low pI {alpha}-amylase, AmyLQS55-3 (Fig. 4e, f).

The degradation of wild-type AmyL in B. subtilis might reflect the fact that this protein is not native to this strain. We therefore studied its secretion in B. licheniformis strain ATCC 9789 (Fig. 4g, h) and observed that, whilst the AmyL precursor was rapidly processed, the amount of newly synthesized {alpha}-amylase remained constant throughout the 15 min chase. This implies that newly synthesized AmyL is stable in its native host, B. licheniformis, and that the observed degradation in B. subtilis might reflect the fact that production of AmyL in B. subtilis takes place in a heterologous expression and secretion system.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Gene splicing techniques and the related genes from B. amyloliquefaciens, B. licheniformis and B. stearothermophilus encoding liquefying {alpha}-amylases were used to construct a series of genes encoding {alpha}-amylases with pI values that were either higher or lower than the pI of the wild-type B. licheniformis amylase, AmyL. Whilst no activity was detected from either of the full chimeric constructs, in each case the replacement of one of the constituent blocks with a wild-type block led to the production of active enzymes. The purification and characterization of these enzymes is described elsewhere (Stephenson et al., 2000 ).

Chimeric and wild-type {alpha}-amylases were expressed in B. subtilis from constructs in which the transcriptional and translational regulatory regions, including the signal peptide coding sequence, were identical. Additionally, the amounts of {alpha}-amylase mRNA synthesized from the genes were comparable. Despite this, large differences were observed in terms of amylase activity and amylase protein accumulating in growth medium from strains expressing different chimeric enzymes. These differences were observed irrespective of the particular growth medium and growth stage investigated. The observed differences could not be ascribed to differences in susceptibility of the fully folded, secreted amylases to the B. subtilis extracellular proteases as both chimeric amylases and wild-type AmyL were fully stable during incubation in spent growth medium at 37 °C for at least 7 d. However, pulse–chase experiments showed that, even in the case of AmyL, substantial amounts of the newly synthesized enzymes were degraded within a few minutes of their synthesis. Because the fully folded proteins were stable in growth medium, it is likely that the observed degradation occurs in a cell-associated location, possibly before or during post-translocational folding into their native conformations. When the native B. licheniformis host was used, newly synthesized AmyL was stable. We therefore propose that the instability of AmyL and its chimeric derivatives in B. subtilis reflects different degrees of mismatch between their folding behaviour and this bacterium’s systems for screening the authenticity and/or fidelity of secreted proteins. The likely role of these screening systems is to prevent blockage of the presecretory protein translocases and/or the formation of protein aggregates within the matrix of the cell wall.

Subtilisin–alkaline phosphatase fusion proteins have been shown to be degraded extensively in a prsA3 mutant of B. subtilis (Jacobs et al., 1993 ), presumably because the limited amount of the folding effector PrsA reduces the rate at which the fusion proteins are folded. Furthermore, increased amounts of PrsA have a positive effect on the accumulation of AmyL when the amyL gene is expressed at high rates from a multicopy plasmid. The increased AmyL accumulation coincides with a strong reduction in the extent to which newly synthesized AmyL is degraded in the PrsA-overexpressing strain, indicating that co- or post-translocational folding is an important factor for efficient secretion (Jensen, 1997 ). The increased cell-associated degradation observed for the chimeric enzymes is therefore likely to be a consequence of changes in their folding kinetics (Stephenson et al., 1998 ).

An intriguing observation, consistent with the notion that folding kinetics is an important parameter in determining secretory protein yield, was the finding that production of one of the chimeric {alpha}-amylases was significantly affected by growth temperature. The production of AmyLQS55-6, which at 37 °C was produced at about 1% of the level of AmyL, was increased 50-fold when the growth temperature was lowered to 26 °C (Jensen, 1997 ); at this temperature, AmyL production decreased twofold.

The E. coli outer-membrane protein A (OmpA) has also been shown to be susceptible to proteolytic degradation during secretion from both B. subtilis and Staphylococcus carnosus (Meens et al., 1997 ). There were strong indications that the degradation of OmpA was caused by membrane- and/or cell-wall-associated protease(s). Whilst problems might have been expected when attempting to secrete an outer-membrane protein from a Gram-negative bacterium in a Gram-positive bacterium, our observations that a protein from a closely related species is also susceptible to cell-associated degradation are more surprising, and point to an efficient mechanism for detecting and removing non-native protein structures. We have recently shown that CWBP52, a cell-wall-associated serine protease, is responsible for only 15% of the observed cell-associated degradation of AmyL (Stephenson & Harwood, 1998 ; Bolhuis et al., 1999b ). This and other wall- or membrane-associated proteases may act to prevent defective or foreign proteins from blocking secretory translocases, a potentially lethal event.

An implication of these studies is that, even for apparently robust Bacillus enzymes such as AmyL, cell-associated degradation is a factor limiting the yield of secretory proteins and is an important parameter that needs to be addressed when Bacillus strains are being optimized for high-level production systems. Our work also demonstrates that engineering changes in native proteins, for example to optimize their enzymic properties for a specific industrial application, could have secondary effects that reduce their suitability for secretion into the culture medium. Therefore, when protein engineering is used to modify the specific characteristics of a enzyme of commercial interest, it should be borne in mind that the introduced modifications have to be compatible with the secretion system of the production host. Even small modifications within an enzyme can dramatically affect the yield of secreted product.


   ACKNOWLEDGEMENTS
 
Antibodies raised in rabbits against AmyL were kindly supplied by S. A. Hansen of Novo Nordisk A/S. The work was partly supported by grants from the EC Biotech programme (contract nos BIO2-CT93-0254 and BIO4-CT96-0097) and the UK Biotechnology and Biological Sciences Research Council. The figures are reproduced with the permission of Novo Nordisk A/S.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Boguslawski, G., Grosch, J., Schultz, J. W. & Subramanian, S. (1996). Proteases of altered stability to autolytic degradation. US Patent 5543302.

Bolhuis, A., Tjalsma, H., Smith, H. E., de Jong, A., Meima, R., Venema, G., Bron, S. & van Dijl, J.-M. (1999a). Evaluation of bottlenecks in the late stages of protein secretion in Bacillus subtilis. Appl Environ Microbiol 65, 2934-2941.[Abstract/Free Full Text]

Bolhuis, A., Tjalsma, H., Stephenson, K., Harwood, C. R., Venema, G., Bron, S. & van Dijl, J.-M. (1999b). Different mechanisms for thermal inactivation of Bacillus subtilis signal peptidase. J Biol Chem 274, 15865-15868.[Abstract/Free Full Text]

Bron, S. (1991). Plasmids. In Molecular Biological Methods for Bacillus , pp. 146-147. Edited by C. R. Harwood & S. M. Cutting. Chichester:Wiley.

Bron, S. & Luxen, E. (1985). Segregational instability of pUB110-derived recombinant plasmids in Bacillus subtilis. Plasmid 14, 235-244.[Medline]

Chambert, R., Haddaoui, E. A. & Petit-Glatron, M.-F. (1995). Bacillus subtilis levansucrase: the efficiency of the second stage of secretion is modulated by external effectors assisting folding. Microbiology 141, 997-1005.

Diderichsen, B., Outtrup, H., Schuein, M. & Norman, B. E. (1987). Chimeric enzymes. European Patent Application EP-87305781.

Diderichsen, B., Wedsted, U., Hedegaard, L., Jensen, B. R. & Sjøholm, C. (1990). Cloning of aldB, which encodes {alpha}-acetolactate decarboxylase, an exoenzyme from Bacillus brevis. J Bacteriol 172, 4315-4321.[Medline]

Ferrari, E., Jarnagin, A. S. & Schmidt, B. F. (1993). Commercial production of extracellular enzymes. In Bacillus subtilis and Other Gram-positive Bacteria: Biochemistry, Physiology, and Molecular Genetics , pp. 917-937. Edited by A. L. Sonenshein, J. A. Hoch & R. Losick. Washington:American Society for Microbiology.

Gray, G. L., Mainzer, S. E., Rey, M. W., Lamsa, M. H., Kindle, K. L., Carmona, C. & Requadt, C. (1986). Structural genes encoding the thermophilic {alpha}-amylase of Bacillus stearothermophilus and Bacillus licheniformis. J Bacteriol 166, 635-643.[Medline]

Harwood, C. R. (1992). Bacillus subtilis and its relatives: molecular biological and industrial workhorses. Trends Biotechnol 10, 247-256.[Medline]

Horton, R. M., Hunt, H. D., Ho, S. N., Pullen, J. K. & Pease, L. R. (1989). Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77, 61-68.[Medline]

Jacobs, M., Andersen, J. B., Kontinen, V. & Sarvas, S. (1993). Bacillus subtilis PrsA is required in vivo as an extracytoplasmic chaperone for secretion of active enzymes synthesised with or without pro-sequence. Mol Microbiol 8, 957-966.[Medline]

Jensen, C. L. (1997). Secretion of chimeric {alpha}-amylases from Bacillus subtilis. PhD thesis, Technical University of Denmark (DTU).

Jørgensen, P. L., Hansen, C. K., Poulsen, G. B. & Diderichsen, B. (1990). In vivo genetic engineering: homologous recombination as a tool for plasmid construction. Gene 96, 37-41.[Medline]

Jørgensen, P. L., Poulsen, G. B. & Diderichsen, B. (1991a). Cloning of a chromosomal {alpha}-amylase gene from Bacillus stearothermophilus. FEMS Lett 77, 271-276.

Jørgensen, S., Skov, K. W. & Diderichsen, B. (1991b). Cloning, sequence, and expression of a lipase gene from Pseudomonas cepacia: lipase production in heterologous hosts requires two Pseudomonasgenes. J Bacteriol 173, 559-567.[Medline]

Kieser, T. (1984). Factors affecting the isolation of ccc DNA from Streptomyces lividans and Escherichia coli. Plasmid 12, 19-36.[Medline]

Kontinen, V. P. & Sarvas, M. (1993). The PrsA lipoprotein is essential for protein secretion in Bacillus subtilis and sets a limit for high-level secretion. Mol Microbiol 8, 727-737.[Medline]

Kontinen, V. P., Saris, P. & Sarvas, M. (1991). A gene (prsA) of Bacillus subtilis involved in a novel, late stage of protein export. Mol Microbiol 5, 1273-1283.[Medline]

Laoide, B. M., Chambliss, G. H. & McConnell, D. J. (1989). Bacillus licheniformis {alpha}-amylase gene, amyL, is subject to promoter-independent catabolite repression in Bacillus subtilis. J Bacteriol 171, 2435-2442.[Medline]

Leloup, L., Haddaoui, E. A., Chambert, R. & Petit-Glatron, M.-F. (1997). Characterization of the rate-limiting step of the secretion of Bacillus subtilis {alpha}-amylase overproduced during the exponential phase of growth. Microbiology 143, 3295-3303.[Abstract]

Machius, M., Wiegand, G. & Huber, R. (1995). Crystal structure of calcium-depleted Bacillus licheniformis {alpha}-amylase at 2·2  resolution. J Mol Biol 246, 545-559.[Medline]

Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982). Molecular Cloning: a Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Meens, J., Herbort, M., Klein, M. & Freudl, R. (1997). Use of the pre-pro part of Staphylococcus hyicus lipase as a carrier for secretion of Escherichia coli outer membrane protein A (OmpA) prevents proteolytic degradation of OmpA by cell-associated protease(s) in two different Gram-positive bacteria. Appl Environ Microbiol 63, 2814-2820.[Abstract]

Ortlepp, S. A., Ollington, J. F. & McConnell, D. J. (1983). Molecular cloning in Bacillus subtilis of a Bacillus licheniformis gene encoding a thermostable {alpha}-amylase. Gene 23, 267-276.[Medline]

Palva, I. (1982). Molecular cloning of an {alpha}-amylase gene from Bacillus amyloliquefaciens and its expression in Bacillus subtilis. Gene 10, 81-87.

Petit-Glatron, M.-F., Graycar, L., Munz, A. & Chambert, R. (1993). The contribution of the cell wall to a transmembrane calcium gradient could play a key role in Bacillus subtilis protein secretion. Mol Microbiol 9, 1097-1106.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74, 5463-5467.[Abstract]

Simonen, M. & Palva, I. (1993). Protein secretion in Bacillus species. Microbiol Rev 57, 109-137.[Abstract]

Sloma, A., Rufo, G. A., Rudolph, C. F., Sullivan, B. J., Theriault, K. A. & Pero, J. (1989). Cloning and deletion of the genes for three minor extracellular proteases of Bacillus subtilis. In Genetics and Biotechnology of Bacilli , pp. 295-302. Edited by M. M. Zukowski, A. T. Ganesan & J. A. Hoch. London & New York:Academic Press.

Sloma, A., Rufo, G. A., Theriault, K. A., Dwyer, M., Wilson, S. W. & Pero, J. (1991). Cloning and characterization of the gene for an additional extracellular serine protease of Bacillus subtilis. J Bacteriol 173, 6889-6895.[Medline]

Spizizen, J. (1958). Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc Natl Acad Sci U S A 44, 1072-1078.

Stephenson, K. & Harwood, C. R. (1998). The influence of a cell-wall-associated protease on the production of {alpha}-amylase by Bacillus subtilis. Appl Environ Microbiol 64, 2875-2881.[Abstract/Free Full Text]

Stephenson, K., Carter, N. M., Harwood, C. R., Petit-Glatron, M.-F. & Chambert, R. (1998). The influence of protein folding on the secretion of {alpha}-amylase from Bacillus subtilis. FEBS Lett 430, 385-389.[Medline]

Stephenson, K., Bron, S. & Harwood, C. R. (1999). Cellular lysis in Bacillus subtilis: the effect of multiple extracellular protease deficiencies. Lett Appl Microbiol 29, 141-145.

Stephenson, K., Jensen, C., Jørgensen, S. T., Lakey, J. H. & Harwood, C. R. (2000). Characterisation and properties of chimeric {alpha}-amylases produced in Bacillus subtilis. Biochem J (in press).

Takkinen, K., Pettersson, R. F., Kalkkinen, N., Palva, I., Söderlund, H. & Kääriäinen, L. (1983). Amino acid sequence of {alpha}-amylase from Bacillus amyloliquefaciens deduced from the nucleotide sequence of the cloned gene. J Biol Chem 258, 1007-1013.[Abstract/Free Full Text]

Wu, X.-C., Lee, W., Tran, L. & Wong, S.-L. (1991). Engineering a Bacillus subtilis expression-secretion system with a strain deficient in six extracellular proteases. J Bacteriol 173, 4952-4958.[Medline]

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, 103-119.[Medline]

Yasbin, R. E., Wilson, G. A. & Young, F. E. (1975). Transformation and transfection in lysogenic strains of Bacillus subtilis: evidence for selective induction of prophage in competent cells. J Bacteriol 121, 296-304.[Medline]

Yoneda, Y., Graham, S. & Young, F. E. (1979). Cloning of foreign gene for {alpha}-amylase in Bacillus subtilis AHU1031. Eur J Biochem 141, 83-89.[Abstract]

Received 28 February 2000; revised 12 June 2000; accepted 30 June 2000.