1
Bacillus and the Story of Protein Secretion and Production

Giulia Barbieri1, Anthony Calabria2, Gopal Chotani2, and Eugenio Ferrari1

1Dipartimento de Genetica Molecolare Batterica, Universidad di Pavia, Pavia PV, Italy

2DuPont Industrial Biosciences, Palo Alto, CA, USA

1.1 Bacillus as a Production Host: Introduction and Historical Account

Contrary to logical thinking, the use of enzymes in daily activities may actually predate the development of modern agricultural societies. Nomad populations of hunters and gatherers exploited rennin produced by the stomach of ruminants for the cheese-making process; while the development of fermentation processes for alcohol can be traced back to more than 7000?years (McGovern et al. 2004; Alba-Lois and Segal-Kischinevzky 2010). However, it is only in the nineteenth-century that enzymes were identified as responsible factors for century old processes such as leather tanning and conversion of starch to sugar (Payen and Persoz 1833).

At the beginning of the twentieth-century, thanks to the work of Otto Rohm, enzymes started playing a wider role in industrial processes as well as in household applications (Wallerstein 1939; Maurer 2010). Two US patents were granted on the use of enzymes for the conversion of starch to sugar; one filed by Schultz et al. (1939) describing the use of a Bacillus mesentericus "extract," and the other by Dale and Langlois (1940) claiming the use of fungal saccharifying enzymes.

In the mid-1950s, microbial enzymes started being used extensively in several applications. Large-scale enzyme preparations, obtained via microbial fermentation thus prominently entered the industrial world (Underkofler et al. 1958). The 1960s saw the dawn of Bacillus as a production workhorse. Toward the end of the decade, Bacillus-derived proteases took hold as essential components of laundry detergents (Roald and De Tieme 1969). At about the same time, high temperature-resistant amylases, useful in the saccharification process, were identified in Bacillus licheniformis and Bacillus amyloliquefaciens. At first, due to the insufficient genetic characterization, the strains used in large-scale fermentation were isolated via a labor-intensive and time-consuming approach of mutagenesis and screening. Most likely, several thousand mutants were tested for improved production characteristics, such as relief of catabolite repression, antibiotic resistance (most likely mutation in one or more ribosomal components), and sporulation deficiency (Ingle and Boyer 1976). The choice of sporulation mutations is particularly important since it allows extending production time in fermentors and, due to poor survival of nonsporulating cells in the environment, precludes isolation of production strains by competitors. The advent of genetic engineering allowed making rapid targeted changes in enzymes and accelerated construction of ad hoc production strains starting from laboratory strains, allowing budding industrial biotechnology companies, such as Genencor, to introduce the first detergent alkaline protease produced by a recombinant microorganism in 1984 (E. Ferrari, unpublished).

For the reasons mentioned above, and for their ease of growth in large-scale submerged fermentation, members of the genus Bacillus play a very important role in the manufacture of a number of industrially important products. While the use of Bacilli has been explored for the synthesis of pharmaceutical products, their most important commercial role is in the production of industrial enzymes (Aehle 2007). It is estimated that in the current greater than $4-billion industrial enzyme market, Bacilli produce about 50% of the enzymes (G. Nedwin, personal communication). These products are employed in a variety of important commercial applications such as laundry, dishwashing, starch-derived ethanol and sweeteners, baking, animal feed, textile, and leather (for review see Aehle 2007).

Several traits make the genus Bacillus attractive for protein production especially since B. subtilis has a long history of safe use. Bacillus natto, a very close relative of the laboratory strain B. subtilis, has been used to obtain natto, a staple of Japanese cuisine from soybean fermentation, for over a thousand years (Nishito et al. 2010). Furthermore, what makes the use of Bacillus for the production of industrial enzymes particularly attractive is its ability to secrete proteins in the culture fluid. This is a necessary feature to keep the cost of the enzymes low, an essential aspect for this class of product. In fact, in most cases, the cost of enzyme production has to be below the $500?kg-1 mark, hence the necessity to have low recovery-associated costs. Over the years, a number of tools have been developed to ease and speed up Bacillus genetic manipulation. The availability of the sequenced genomes of both B. subtilis (Kunst et al. 1997) and B. licheniformis (Rey et al. 2004; Veith et al. 2004) has allowed studies aimed at better understanding their behavior during growth and production (Buescher et al. 2012; Nicolas et al. 2012). Moreover, the well-characterized fermentation, its relatively short time, and the possibility to use cheap feedstock add to the appeal of using these bacteria for the production of industrial enzymes.

This chapter is divided in two main sections: the first section focuses on the genetic tools and strategies useful for the efficient cloning and expression of proteins in Bacillus, while the second section provides an up-to-date status on fermentation and recovery of heterologous enzymes from Bacillus.

1.2 The Building of a Production Strain: Genetic Tools for B. subtilis Manipulation

Numerous genetic manipulation techniques of B. subtilis laboratory strains have been established over the years. These tools have helped in refining the genetic and biochemical characterization of this microbe. Hence, even if B. subtilis had never played a major role in the historical development of the genus Bacillus as an industrial workhorse, the availability of new or genetically engineered enzymes with new properties, and the need to express them at very high levels, has placed this laboratory microbe as a frontrunner in the expression of industrial enzymes. In fact, the only tool available to carry out the needed manipulations in the traditional industrial strains, namely B. licheniformis and B. amyloliquefaciens, was and is to a large extent protoplast transformation. But this approach is very time-consuming and not always reliable. Recently, however, in some instances, it has become possible to develop competence in B. licheniformis strains using the comK induction system described below (Diaz-Torres et al. 2003; Hoffmann et al. 2010). Given that the tools and ways to transform B. subtilis are simple, the building of a B. subtilis production strain for a secreted protein is relatively straightforward when the transcriptional and translational determinants for the synthesis of the protein of interest as well as a signal sequence to direct its secretion are available.

Some of the genetic techniques and tools currently available for building a B. subtilis production strain are briefly outlined in the next section.

1.2.1 Promoters

There are a number of promoters that can be used to direct transcription of any target protein. One of the best-characterized B. subtilis promoters is aprE, which is responsible for the transcription of the alkaline protease. It is yet difficult to explain why a promoter responsible for the expression of one of several scavenging enzymes is so complexly regulated (Ferrari et al. 1993). The transcription of aprE is controlled by at least two different repressors, AbrB and ScoC, and by a pleiotropic transcriptional activator, DegU (Henner et al. 1988), which can boost the transcription of the aprE mRNA by about 100-fold. The presence of both AbrB and ScoC assure that AprE is not synthesized before the transition phase, e.g. before the culture enters the stationary phase. One advantage of using the aprE promoter for heterologous expression is the presence of a transcriptional leader sequence responsible for extending the half-life of its mRNA to about 25-30?minutes (Hambraeus et al. 2002). The mRNA stability is transferred to most genes hooked to this transcriptional leader, allowing robust expression.

Another widely used promoter is the amylase promoter, in its different versions, amyE, amyQ, and amyL, which are derived from B. subtilis, B. amyloliquefaciens, and B. licheniformis, respectively. The amylase promoter, albeit not under strict sporulation control, is temporally regulated and its transcription is turned on at the end of the vegetative growth, just before the cells enter the stationary phase. This is most likely due to the control exerted by catabolite repression in both B. subtilis and B. licheniformis (Nicholson et al. 1987; Laoide et al. 1989).

Two other promoters worth mentioning are the sacB and sacC...