Protein Production by Biotechnology

In this review we consider the possible use of Bacillus subtilis as a host for the production of heterologous proteins. There are several potential advantages to be gained from the use of this organism, particularly its efficient secretion of proteins into the growth medium. Although it is unlikely to become the first choice host for the production of certain types of protein, for example potentially therapeutic mammalian proteins, which often undergo specific and necessary post-translation modifications, there are certainly some important groups of proteins, for example industrial enzymes, for which the Bacilli are already heavily used. The market for such enzymes is likely to grow very rapidly as recombinant DNA methods allow for the production of proteins from diverse and often little characterised microorganisms. Exploitation of these natural products will depend upon the development of versatile and robust expression systems, and there is growing evidence that Bacillus will be a useful system. On the other hand, Bacillus has certain undesirable properties as a host, such as the elaboration of proteases, which can cause loss of product by degradation. If such problems can be overcome, and in this review we assess the technical difficulties involved, then B. subtilis may turn out to be the host of choice for the production of many of the new industrial enzymes.

Filamentous fungi have been of both academic and commercial interest for many years. The last few years have seen the techniques of genetic manipulation extended to a wide range of filamentous fungi. Initially, work was carried out with the genetically well characterised fungi such as Aspergillus nidulans and Neurospora crassa, but there has been a growing interest in application of these techniques to commercially important fungi. Antibiotics (Penicillium chrysogenum, Cephalosporium acremonium), organic acids (Aspergillus niger, Aspergillus terreus) and enzymes (A. niger, A. oryzae, A. awamori, Trichoderma reesei) are the major economically important products from fungi, and application of molecular techniques is seen as a way of either improving current processes or using the fungi as hosts for production of heterologous proteins. Fungi already used for pharmaceutical or food products are seen to have a commercial advantage for heterologous protein production in that they enjoy the GRAS (Generally Regarded As Safe) status from the US Food and Drug Administration (FDA).

Man and other animals survive the relentless assault of a large array of microbial pathogens because they have evolved mechanisms that enable them not only to recover from initial attacks and purge themselves of the offender, but also to remember sufficient information about the molecular morphology of an individual aggressor that they are able to repulse further intrusions very quickly and effectively. Through this combination of exposure to a pathogen and clearance and recollection of it, the animal becomes immune to that particular pathogen.

Many extracellular proteins consist of numerous distinct but repeated domains. Figure 1 illustrates how relatively few domains, or modules, can be combined to make a wide variety of proteins associated with blood-clotting, fibrinolysis, complement and the extracellular matrix.^1,2 The modules are often small disulphide bonded units of about 50 amino acids. They include: epidermal growth factor-like units (G); the ‘Kringle’ unit (K); fibronectin ‘fingers’ (F1, F2 and F3); the short consensus repeat seen in the C4 binding protein and other proteins of complement (C);³ a module which appears in thrombospondin and properdin (T);⁴ one seen in the LDL receptor (A)⁵ and modules with homology to the EF hands in calmodulin (E).⁶ There is a close relationship between these domains and the exon/intron structure of their genes. Exon duplication and shuffling was probably involved in the evolution of these mosaic proteins. It is significant that nearly all of these modules have phase 1 exon boundaries,⁷ such that exons may be duplicated, inserted or deleted without changing the reading frame for the rest of the protein. The modules have acquired diverse roles in their subsequent evolution. They appear to form autonomously folding structural units since disulphides are usually within rather than between modules. This is supported by the structures known so far.

Human serum albumin (HSA) is the largest single protein component of plasma¹ where its role is to maintain normal osmolarity and to act as a carrier for numerous small molecules (including nutrients and metabolites) many of which would otherwise have low solubility or be poorly tolerated in free solution. Compounds which it is capable of binding include fatty acids, bilirubin and numerous drugs.² Unlike many recombinant proteins currently being considered by the biotechnology industry, albumin is already sold and used in large quantities.³ It is prepared by fractionation of donated blood and is used in the treatment of patients requiring fluid replacement.⁴ It is used particularly in the treatment of burn victims, those suffering from traumatic shock and some special groups of surgical patients. However its use is much affected by the custom and practice of particular countries and the preference of individual doctors.

The ability to express cloned genes in mammalian cell lines has proved to be a powerful technology. It is now possible to produce a rare protein of some scientific or therapeutic interest in sufficient quantity to be able to study it and/or produce it commercially. With the genes coding for interesting proteins cloned and expressable, the future holds the possibility for genetically modifying natural proteins to design and produce new proteins as required. The best example, to date, of a cloned gene expressed in mammalian cells to produce a human therapeutic agent is tissue plasminogen activator (tPA). Interestingly, there are already underway projects to engineer natural tPA protein to create new ‘second-generation’ tPA proteins with altered properties such as longer half-life and increased substrate specificity¹ (see Chapter 8, this volume).

In 1801, 3 years after he had introduced vaccination,¹ Edward Jenner made his famous prediction, ‘… that the annihilation of the smallpox, the most dreaded scourge of the human species, must be the final result of this practice’. One hundred and seventy six years later this prophecy was fulfilled.² This triumph of preventive medicine was achieved by using the antigenically related orthopoxvirus, vaccinia, for immuno-prophylaxis against variola virus the causative agent of smallpox. It might also have been predicted that once smallpox had been eradicated interest in poxviruses would diminish. Paradoxically research on vaccinia and other poxviruses is now more intense than ever. A major factor contributing to this increased interest is the development of techniques that permit the expression of foreign proteins from recombinant poxviruses. These expression vectors have a variety of applications which include potential as new live vaccines in veterinary or human medicine. The construction and application of recombinant vaccinia viruses is the topic of this chapter.

Acute myocardial infarction (AMI) is a major cause of death in the Western World (Table 1). One of the main precipitating events in AMI is the formation of an occluding thrombus in a coronary artery. The thrombus prevents blood flow to the myocardium thus reducing oxygen supply which in turn leads to an, often fatal, infarction. It is now widely accepted that early removal of the blood clot, by dissolution of the fibrin network which holds it together, is of major benefit in reducing morbidity and mortality.¹ A number of fibrinolytic proteins are now in use in the clinic. One of these is tissue-type plasminogen activator (t-PA).² t-PA is a serine protease; however, it does not dissolve fibrin directly, instead it activates plasminogen, a plasma protein, by cleaving the arg₅₆₁-val₅₆₂ peptide bond thus revealing the active enzyme, plasmin. Plasmin in turn dissolves fibrin and lyses the thrombus. The primary structure of t-PA was only established after cloning the mRNA as cDNA (Fig. 1).^3,4 The 527 amino acid protein, like many serine proteases, consists of two chains (A and B) linked by a protease-susceptible peptide bond. The B chain carries the catalytic centre, but somewhat unusually for a serine protease, t-PA appears to be active in both single and two-chain forms. The A chain is believed to be divided into four structural domains (the finger, the growth factor and two kringles).

Factor VIII (FVIII) is an essential cofactor involved in blood clotting. It functions in the intrinsic pathway of coagulation in the step where the activated form of factor IX (FIXa) activates in its turn factor X to FXa. Lack or malfunction of FVIII results in the chromosome X-linked bleeding disorder haemophilia A. In blood, FVIII circulates as a complex with von Willebrand factor (VWF) which serves as carrier molecule stabilising FVIII procoagulant activity and may target FVIII to damaged sites. At present, haemophiliacs are treated by replacement therapy with plasma derived products which pose well documented problems associated with low purity and risk of contamination by viral agents such as those causing hepatitis A, B, non-A non-B and AIDS.

The strongest evidence for the existence of an important control in the flanking region of the globin gene domain was provided by the analysis of human γβ-thalassaemias.^1,2 Patients with heterozygous Dutch γβ-thalassaemia have a deletion that removes 100 kb of DNA, leaving the β-globin gene and the promoter and enhancer regions intact. However, it abolishes expression of the deleted chromosome and leaves the gene in an inactive chromatin configuration.^3–5 The wild-type allele on the other chromosome is expressed at normal levels, indicating that there is no shortage of trans-acting factors. This suggests a cis effect on β-globin gene transcription, which could be caused by a loss of positive acting elements or by the juxtaposition of the intact β-globin gene and sequences that remain in an inactive chromatin configuration in erythroid cells. The first indication that positive acting sites may be involved in activation of the β-globin domain came with the observation of erythroid specific DNasel hypersensitive sites that map 6–18 kb upstream from the ε-globin gene (Fig. 1).^6–8

Proteins can be made by a variety of routes for applications ranging from industrial enzymes to human therapeutics. Recently interest in this subject has increased greatly since the ability to manipulate DNA and to transfer it into different host cells has meant that virtually unlimited amounts of individual proteins can now be made.

In many respects this chapter can be seen as a statement of the obvious — mankind has been obtaining tons of important protein products from plants for industrial uses for generations. What genetic engineering can do is to add a new source of variation to the natural variation that is already exploited in agriculture. In this review we will outline the recent progress made in the understanding of gene expression, gene product targetting, gene isolation and tissue culture that will be important for the development of plants and plant tissue culture as sources of novel proteins.

Animal cell technology has been playing an important role in the production of therapeutic products for more than 30 years. Following the introduction of polio vaccine in 1954 a number of other vaccines have been produced in animal cell culture, including measles, mumps and rubella. The technology has also been applied to the production of veterinary products such as foot-and-mouth disease virus vaccine. This particular vaccine, made in baby hamster kidney (BHK) cells, still represents one of the largest animal cell processes with millions of litres of culture fluid a year being processed.1 During the 1970s animal cell culture was developed for the manufacture of some natural cell products which are now used clinically; such as urokinase and interferon ß. The early human therapeutic products were manufactured in ‘normal’ cells, typically human diploid fibroblasts. These cell types have limited utility for applications requiring very large scale manufacture because they have a finite lifespan, are anchorage dependent and tend to require complex (and therefore expensive) culture media. In recent years attention has turned to the use of immortal cell lines such as myelomas and Chinese Hamster Ovary (CHO) cells which have an infinite lifespan, usually grow in suspension and can be cultured in less complex media than ‘normal’ cells.

To be able to design a chemical compound that would interact with a characterized receptor on a cell in a specific and predictable way has been the stated goal of pharmacology from very early on. Such rational drug design would in fact be a great achievement. However, we can expect that many new specific drugs will also be discovered by testing of compounds as currently practised. What we propose to do in this chapter is to examine how recent developments in molecular biology have provided several important techniques which can be used to study the interactions of drugs with specific receptor subtypes and how these same systems can be used to study the structure of specific receptors. Hopefully this will lead to the goal of rational drug design as well as provide new and more specific screening procedures for new drugs.

The advances made during the past decade in recombinant DNA science has led to the ability to identify, clone, express and produce on a large scale many proteins. As more becomes known about the pharmacology of proteins and polypeptides, it is becoming increasingly apparent that a variety of new peptide and protein drugs can be expected to be used clinically in the coming decades. Polypeptides and proteins proposed for therapy usually have regulatory or homeostatic functions. They include both endogenous polypeptides and proteins and their (heterologous) derivatives. This latter class of molecules may be produced after sitedirected mutagenesis or gene fusion, or by proteolysis or protein aggregation, and/or conjugation with (other) biologically active effector functions.

Many of the recombinant DNA systems of current interest for production of therapeutic proteins have been covered in preceding chapters. Progress in recent years has been rapid; genes for the first proteins of clinical interest were cloned and expressed in Escherichia coli in the late 1970s and in less than 10 years nine materials have entered the market including insulin, growth hormone, interferon-alpha, tissue plasminogen activator and erythropoietin. This progress is remarkable when one considers the current time scales for discovery and development of chemical therapeutic agents. Although there are several very promising new production systems for therapeutic proteins, work over several years has produced some robust and generally useful systems. For simple proteins, E. coli expression has the advantage of high yield. For large, complex proteins such as plasminogen activator and Factor VIII, which have extensive secondary and tertiary modifications, mammalian cell production is likely to be essential. These systems can often be applied now in predictable ways and the accumulated knowledge to date would seem to be a reasonable basis for assessing accomplishments and making some extrapolations.