Novozymes develops enzyme products using a large array of different technologies. We screen en-zymes of all enzyme classes from natural diversity and we generate in vitro diversity by e.g. site di-rected mutagenises or shuffling of homologous genes. When an enzyme has been screened or de-signed for an application, a production strain is developed.
If it is an enzyme selected from natural diversity, the production strain could simply be the native mi-croorganism that produces this enzyme. But mostly the product yields are very limited and microor-ganism produces unwanted site activities. Sometimes the microorganism just cannot grow in modern production facilities.
At Novozymes we solve these problems by the use of gene technology. We design and construct su-perior recombinant production strains based on our experience from fermentation technology.
A good recombinant production strain is characterized by several different properties.
The starting point for the strain construction, the host strain, has to be versatile. Novozymes capability to produce the huge diversity of enzymes that we develop means that our host strains must be able to produce an array of different enzymes.
It has to be an absolutely safe strain to use, meaning that it is non pathogenic towards humans, ani-mals and plants and that it has no adverse effects on the environment.
It must be very efficient to secrete proteins. The fermentation yield of the enzyme product must be very high to secure an economically and environmentally sustainable production process. Also the product purity and quality of the final product is highly dependant on the fermentation yield.
The strain should not produce unwanted site acti vities such as proteases that can negatively influence the stability of the enzyme product or just other enzymes that hamper the product purity.
It must also have growth characteristics suitable for growth in a stainless steel fermentor. The strain must grow submersed to an appropriate biomass at an appropriate temperature and the morphology of the culture must allow for efficient stirring and aeration.
Host strain development.
When Novozymes first faced the challenge of producing recombinant enzymes, the company was al-ready a very experienced enzyme producer. So we had a very solid basis for choosing microorganisms that could be used as host strains for recombinant production from the criteria listed in the previous paragraph.
Novozymes has developed both bacterial and fungal expression systems and we continue to improve these systems.
Novozymes has developed a number of different bacterial hosts that have a well documented safe use for over 30 years in production. The hosts are derived from species like Bacillus subtilis, Bacillus licheniformis and Bacillus clausii. The genus Bacillus is known to have a very high potential for secret-ing enzymes into the medium in a pure form (Zukowski, 1992) and is an ideal organism for enzyme production at large scale. Tools for genetic engineering in our bacillus hosts have been developed, patented and implemented for industrial application over the years at Novozymes that has given us a leading position in this field. Significant improvements of the wildtype host strains have been intro-duced by either classical mutagenesis or specific changes imposed by use of recombinant techniques. These alterations cover e.g. disruption of genes coding for unwanted enzyme activities like amylases and proteases and genes that are responsible for initiation of sporulation.
Novozymes has developed fungal host cells from type strains of Aspergillus oryzae, Aspergillus niger and Fusarium venenatum. Aspergillus oryzae and Aspergillus niger both have an extended history of safe use (Barbesgaard, 1992) and are both generally regarded as safe (GRAS). The strain of Fusa-rium venenatum that we are using is used for single cell protein for human consumption in the United Kingdom (Royer, 1995).
Aspergillus niger and Aspergillus oryzae both secrete substantial amounts of amylases and amylogly-cosidase and are thus very efficient enzyme producers. The extended history of safe use and the abil-ity to produce huge amounts of extracellular protein encouraged Novozymes to develop Aspergillus oryzae as our first fungal recombinant expression system.
The first A. oryzae production strains developed by Novozymes were made by transforming an A. oryzae type strain with a DNA construct that directed efficient production of product gene (Christensen, 1988). These strains were a substantial improvement compared to existing technology, but they only met the characteristics of a good recombinant production strain in part.
When producing an enzyme product the amylases and amyloglycosidases of the host strain are un-wanted side activities. Some of the genes encoding these enzymes have been deleted by gene disruption from our A. oryzae and A. niger host strains. The protease levels are also very high in these strains. In addition to being unwanted protein contaminations in the product, proteases can also de-grade the enzyme products we wish to produce. So individual protease genes and protease regulators have been disrupted.
Also a number of unwanted metabolites have been removed. For example Aspergillus niger produce rather high am ounts of oxalic acid. Oxalic acid forms precipitates with calcium that causes problems in product recovery and formulation. Also the oxalic acid production is a waste of the carbohydrate used in the medium. Finally presence of oxalic acid in the final product cannot be tolerated. Novozymes has solved that problem by disrupting a gene crucial for oxalic acid formation so that our current A. niger host strain does not produce oxalic acid (Pedersen, 2000).
Expression vector improvement.
In addition to an optimal host strain an optimal expression vector is required for an optimal production strain. At Novozymes our expression vectors are plasmids or fragments of plasmids that are integrated into the genome of the host organisms in one copy or in multiple copies. The genetic stability of genomic integrated vectors has proven to be superior to that of episomal plasmids.
In principle an expression vector holds a selection marker, an expression cassette and for bacterial vectors a replication function which determines the copy number of the vector in a given host. The expression cassette consists of a strong promoter, the product gene coding for the enzyme of choice and a transcriptional terminator (figure). The efficiency of the expression cassette that to a large extent determined by the amount of functional messenger RNA (mRNA) is a result of a complex interplay between the promoter efficiency, the structural gene and the host strain. The nature of the product gene determines the choice of host, promoter and the vector used for chromosomal integration.
The promoter of choice in Aspergillus is an amylase promoter. This was used even in the very first recombinant A. oryzae production strain that was launched in 1988. Since then we have gained de-tailed knowledge about the transcription factors acting on this promoter and its regulation (Petersen, 1999). This knowledge has enabled us to isolate variants of the promoter that are even stronger pro-moters. For many Bacilli production strains amylase promoters and selected strong variants thereof are used for production of many different enzyme products.
Selection of high yielding transformants
Since the expression vector integrates in the genome at various copy numbers the expression levels of the product of different transformants are quite different.
In our fungal systems the expression vectors integrate by ectopic multicopy integration. This means that different transformants have the vector integrated at different loci in the genome and at different copy numbers.
Some loci in the genome are hotspots for expression whereas others are relatively silent in terms of expression. So the site of integration is important. Even though no strict correlation between copynumber and expression exists, investigations on strains having the expression plasmid integrated in the same locus but at different copynumbers have shown, that higher copynumbers results in higher expression levels. So a population of transformants shows a quite large diversity in expression of the product gene.
In order to select the transformants with the highest expression levels among a population of trans-formants, the transformants are grown in different systems and the product yields are assayed.
Growth in microtiter wells can be used to deselect the poor transformants and to some extend to pinpoint the high expressing transformants. Fermentation in shakeflasks is another way of doing th is. By microtiter and shakeflask fermentation a small number of transformants is selected. These are then fermented in lab tank fermentors. The transformants are compared in respect to fermentation yields but also other parameters such as morphology are considered.
Selection of mutants
When an optimal transformant is selected as described in the previous section the fermentation yield can be increased by classical mutagenises of the transformant. In this process Novozymes use the classical microbial skills that the company completely relied on before Gene Technology was introduced.
The selected transformant is subjected to chemical mutagenises, UV or radioactive irradiation. Mutated progeny is screened for yield increase very much in the same way as the transformants were originally screened.
Barbesgaard, P et. al. (1992) Appl Microbiol Biotechnol, 36: 569 -572.
Christensen, T et. al. (1988) BioTechnology, 6, 1419 - 1422.
Pedersen, H et. al. (2000) Metab. Eng., 2: 34 -41
Petersen, K. L. et al. (1999) Mol Gen Genet, 262: 668-676.
Royer, J et. al (1995) Bio/Technology, 13:1479-1483
Zukowski, M. M. (1992) Biology of Bacilli: Applications to Industry. Butterworth-Heinemann, Boston: 311-337