As most small-molecule drugs modify the function of proteins involved in disease processes, proteins are essential to the drug discovery process. Furthermore, because proteins comprise a large fraction of all biologically active molecules and cover the entire spectrum of biological activities, they represent a rich source of potential drug candidates. In order to fully realize the pharmaceutical uses of proteins, expression systems are needed that can support drug discovery, as well as development and manufacturing. Currently, a variety of systems is available to perform these steps individually, although not comprehensively. Accordingly, some such systems are limited in their abilities to discover and express genes from a large portion of the protein “universe;” others are impractical for commercial production because of restrictive conditions necessary for growth and cultivation of the biological systems, or the inability of those systems to produce proteins economically on a commercial scale. For example, bacteria are unable to process introns or to glycosylate proteins. Accordingly, they are unable to express many genes from eukaryotic genomic DNA sources. Further, many of these systems are not amenable to current automation of high-throughput screening (HTS). Mammalian cells are difficult to cultivate and produce relatively low levels of expressed proteins. While filamentous fungi are easier to cultivate and produce larger amounts of protein, they are, for the most part, unsuitable for growth in a high-throughput environment. Cultures can be very viscous, and cultivation can be complicated by the formation of surface mats and aerial sporulation.

Taking these issues into account, drug discovery is challenged to develop a protein expression system capable of producing a sufficient supply of diverse proteins to support the industry’s growing needs. To this end, Dyadic International Inc. (Jupiter, Florida, USA), in collaboration with TNO Voeding (Zeist, the Netherlands), is pioneering a system that allows gene discovery, gene expression, product development and manufacturing in a single biological host: Chrysosporium lucknowense (C1).

Proteins in Drug Discovery
Because proteins mediate biological processes, pharmaceutical agents commonly target them. HTS, in which large groups of potential agents are screened against a protein target to identify those members that are active against the target, is considered the gold-standard technique for discovering hits. After HTS is performed, the hits then are evaluated for potential drug properties and eventually could become new chemical entities. The number of protein targets introduced into HTS is expected to increase dramatically as the genomic and proteomic sciences identify new disease-related genes and the proteins that they produce.

Proteins themselves also are important sources of new drugs. Seven percent of drug sales currently are for protein-based therapeutics, and this percentage is expected to rise to 13 percent in the next three years (1). Protein drugs, generally referred to as biologicals, often are developed from endogenous proteins that have demonstrated therapeutically relevant activities. In some cases, the endogenous protein — or a modified version of it — becomes the drug; in others, proteins that are related to the endogenous protein and share some of its activities are the starting point for drug development. Alternatively, therapeutically active proteins can be discovered by screening protein libraries from various sources, which represent a largely untapped source of biological diversity.

To satisfy the protein expression needs for the various stages of drug discovery and development, an expression system for post-genomic drug discovery must possess certain characteristics. These include the ability to: 1) express most of the members of the protein “universe,” particularly eukaryotic proteins; 2) express proteins in a biologically active form; 3) produce proteins in a rapid and inexpensive fashion; 4) interface with HTS automation seamlessly; 5) produce sufficient quantities of protein to support the confirmation of hits and the many other activities involved in transforming a hit into a drug; and 6) scale up to commercial drug production. It is with these needs in mind that the C1 system was developed.

System Overview
The C1 gene discovery engine utilizes Zymark Hopkinton, Massachusetts, USA) automation systems for the preparation and assay of gene expression libraries in C. lucknowense. The requisite genes can be from various sources — genomic or cDNA from individual organisms, DNA isolated from environmental samples or individual gene variant sets generated in the laboratory. Gene libraries are transferred to C1 and individual clones are transferred to the wells of microtiter plates using an Allegro™ system. The Allegro system was chosen because its ultra high-throughput screening capability and storage capacity allow the creation of large numbers of libraries, each containing many clones. After incubation of the cultures for growth and gene expression, cultures are assayed, using a Staccato™ system, for proteins of therapeutic interest. The Staccato is an integrated automation system that can adapt to a wide variety of assay types. Once the relevant proteins are discovered, expression is further optimized in C1 and a fermentation process is developed, allowing large-scale manufacturing of the product of interest.

The C1 fungal host has differentiating advantages, making it uniquely capable of functioning as an integrated drug discovery/manufacturing system. As a eukaryote, C1 can faithfully express genes from virtually any source. C1 can perform post-translational modifications that are characteristic of many eukaryotic proteins (for example, glycosylation, with no evidence of the hyperglycosylation described for yeast), and unlike bacteria, it is capable of splicing introns. This allows access to portions of the biodiverse gene pool not accessible to other screening systems. Because greater than 90% of the Earth’s species are eukaryotic by some estimates, the C1 system will be able to discover genes and gene products that bacterial and yeast systems cannot. As an organism that was developed originally for the production of extracellular enzymes, C1 is capable of producing proteins inexpensively at high yields and in large volume using simple media. However, the truly unique advantage of C1 over other fungal expression systems is its physical compatibility with automated liquid handling systems. Critical features of C1 are summarized in Table I, with comparison to existing biopharmaceutical production systems (2–4).

C1 has a culture morphology, unique among fungi, that confers the ability not only to grow in up to 384-well microtiter dishes but also for those microvolume cultures to be robotically transferred from well to well for culture replication. The morphology of C1 leads to the formation of individual mycelial fragments, referred to as transferable elements, or propagules. The formation of these propagules results in nonviscous growth in cultures, allowing adequate aeration in microtiter wells. In addition, formation of propagules lacks pelleting, surface matting and aerial sporulation, which, in turn, eliminates canula tip clogging in robotic liquid handling systems. Additionally, the lack of aerial sporulation prevents well-to-well cross-contamination. Of some 15 species of fungi tested by Dyadic and TNO Voeding, none exhibited the combination of efficient propagule formation as well as lack of tip clogging that C1 did.

The morphology and physiology of the C1 strain also confers advantages in the area of large scale manufacturing of protein products. The fragmented nature of the culture promotes high-level production and secretion of proteins. The lower viscosity of the culture also allows the use of fermentation parameters — for example high feed and aeration rates — which would otherwise be unmanageable. The ability of the strain to grow and produce protein at a neutral pH level allows for the production of pH-labile proteins. Other commercially viable yeast and fungal protein production systems typically operate in the acidic pH range.

Conclusions and Prospects
The C1 gene expression system currently is in late-stage development, with commercial launch anticipated for mid-2003. About 15 genes, including two human genes, have been overexpressed in C1. Several examples of homologous proteins have been overproduced to multiple gram per liter levels. In two cases, this has led to enzyme products for the textile industry that have improved performance over those products that could be generated by traditional strain selection methods. One of these products is currently in commercial production and the other is in late stage development. Expression analysis of seven genes, representing the major classes of pharmaceutically relevant proteins, currently is in progress. In addition to determining the yields of these proteins, their biological activity and post-translational modifications are being examined to confirm the utility of producing pharmaceuticals in this fungal system.

Other uses of the C1 system for drug discovery and development are in advanced stages of validation. Current activities involve the discovery of useful proteins from libraries of expressed fungal genes. Upon completion of this work, these studies will be expanded to the preparation and screening of human cDNA libraries for useful proteins. Currently there are about 500 known druggable targets in the human genome, but various estimates put the total number of potential targets at 5000–10,000 (5). The functions of most of these have not yet been established; in fact, of the 40,000 or so genes comprising the human genome, the functions of about half are yet to be determined (6). By allowing the expression of sufficient quantities of proteins from these genes for the establishment of their functions, the C1 system will make these proteins available for drug discovery uses.

With the current capacity shortfall in biopharmaceutical production, a robust system that will produce large amounts of protein economically can ensure that a drug, once discovered, can be produced in sufficient quantities and at a viable cost for commercialization. Protein expression technologies with qualitatively better characteristics, such as the C1 system described, will be especially valuable if the number of biopharmaceutical candidates emerging from the laboratory continues to grow as anticipated.

References

1. S. Aldridge, Bioengineering News (22)14, 1:90–91 (2002).
2. R.K. Bretthauer and F.J. Castellino, Biotechnol. Appl. Biochem. 30, 193–200 (1999).
3. F.J. Wiebel et al., ATLA 25, 625–639 (1999).
4. J. Mullins, “Use of expression systems in research and by the pharmaceutical industry.” Molecular Biology Notebook Online.
5. A.L. Hopkins and C.R. Groom, Nat. Rev. Drug Discov. 1(9), 727–730 (2002).
6. J.C. Venter et al., Science 291, 1304–1351 (2001).PG

Richard P. Burlingame* is executive director of R&D at Dyadic International Inc., in Jupiter, Florida, USA; Michael R. Kozlowski is CEO of Fifth Day Therapeutics in Poway, California, USA; and Brian G. Lightbody is vice president of strategic marketing and business development at Zymark Corporation in Hopkinton, Massachusetts, USA. Richard Burlingame can be reached at 561-743-8333 or at rburlingame@dyadic-group.com.

*To whom all correspondence should be addressed.