BIOTECHNOLOGY peptides The unique characteristics of (AMPs) – together with an improved understanding of their universal nature – has prompted renewed interest in the development of this group of antimicrobial agents.

Hans-Henrik Kristensen and Debbie Yaver, Novozymes A/S

ntimicrobial peptides (AMPs) are a recently of the molecular ‘target’ – the antimicrobial effect is Adiscovered group of antimicrobial agents. essentially receptor-independent. ey are simple peptides, but are widely distrib- Novozymes A/S has a strong heritage in the dis- uted in animals and plants, and show activity covery and development of peptidic compounds, against a broad range of pathogens. ey have a using technologies that share common ground number of characteristics that make them interest- with companies in the biotechnology and phar- ing candidates for pharmaceutical development; maceutical fields – such as genetic manipulation, notably, they are fast-acting, microbicidal (rather recombinant expression, high throughput screen- than microbiostatic) and are associated with little ing and protein design. Although Novozymes observed resistance development – a key property traditionally has used these technologies in its in an age of multi-resistant , as represented core businesses – industrial enzyme production by MRSA. – the company is now looking to apply these Most AMPs are cationic and amphipathic – fea- competencies to biopharmaceutical discovery and tures that promote interaction with the negatively development. One lead area is anti-infectives, charged bacterial and fungal membranes. ey with particular progress being made in the devel- work primarily by compromising the membrane of opment of AMPs. the target organism. When analysed at the molecu- At Novozymes, our focus has been on both the AMPs are gene- lar level, several different mechanisms of membrane development of a solid technology platform around encoded; this disruption have been shown to exist. However, for AMPs and the utilisation of this platform for iden- is an essential most AMPs, the overall outcome is membrane dis- tifying and developing lead molecules. By applying feature that has ruption and/or cell lysis. our unique and proven ability to modify peptides, allowed the many Selectivity for microbial membranes is medi- we believe that our AMP discovery platform holds biotechnology ated by membrane composition, membrane charge, great potential. tools originally trans-membrane potential and lipid polarity. e developed for outer leaflet of microbial membranes is populated AMP Technology Platform enzymes to be with negatively charged phospholipids, whereas AMPs are gene-encoded; this is an essential feature applied to AMPs. the outer leaflet of plant and animal membranes that has allowed the many biotechnology tools is composed primarily of neutral lipids. In addi- originally developed for enzymes to be applied to tion to the differences in polarity, the specific types AMPs. At Novozymes, focus has been mainly on and ratios of phospholipids in microbes differ from three separate areas: those of higher organisms, allowing for discrimina- tion between cell types. • A Discovery Platform for identifying new, potent A key feature of AMPs is the inability of bacteria and structurally diverse AMPs, to become resistant. It has proven extremely diffi- • A Directed Evolution Platform for tailoring spe- cult to induce resistance to AMPs in sensitive target cific AMPs towards specific clinical indications, organisms. is is a reflection of the unique nature and

76 Innovations in Pharmaceutical Technology Innovations in Pharmaceutical Technology 77 BIOTECHNOLOGY

Figure 1. Schematic outline of Transposon-Assisted Signal Trapping (TAST).

• A Recombinant Production Platform for securing have – when combined with directed evolution and high industrial-scale recombinant production. throughput screening (HTS) – proven valuable tools to overcome these limitations. Discovery Platform Novozymes has a large microbial strain collection with more than 25,000 catalogued Directed Evolution Platform and High Throughput and well-characterised fungi and bacteria. Both Screening e AMPs found in nature have been traditional and genomic approaches are used to evolutionarily optimised to function in specific physi- identify new AMPs from the strain collection. In the ological or ecological niches – such as tears, saliva, the traditional approach, we screen bacterial and fungal mucosal barrier or the phagosomes of macrophages – culture supernatants for peptide-based , in response to specific microbial challenges. However, and then isolate and characterise the peptides. After AMPs have not been optimised by evolution to func- characterisation, reverse genetics is used to clone and tion optimally in complex therapeutic conditions such express the corresponding gene. In one of our power- as burn wounds, chronic ulcers or cystic fibrosis. To ful genomic approaches, Transposon-Assisted-Signal- overcome these limitations, a number of proprietary Trapping (TAST WO 0177315) has been used with technologies for generating and screening molecular great success (Figure 1). is technology allows us diversity through HTS have been developed. ese to selectively identify and sequence genes that, in the screening systems – with the capacity to analyse mil- natural host, are secreted. As most or all AMPs would lions of different AMPs – have been designed to avoid While it has be expected to be secreted from the host to limit niche- or, in some special cases, directly take advantage of the proven fairly easy competition and aid in host defences, this approach obvious dilemma of expressing potent and broadly to design AMPs focuses and reduces the effort from sequencing whole active antimicrobials in microorganisms. with a basal level genomes or large cDNA libraries down to only the e platform technologies include cis-acting screen- of antimicrobial much smaller subset of secreted cDNAs. ing systems such as the Suicide Expression System (SES activity, it has A number of structurally different AMPs – WO 0073433) (Figure 2), where the antimicrobial proven more – including one of our lead AMP classes, Plectasin effect is directed towards the producing cell. In this difficult to optimise (WO 03044049) – have been identified using TAST. screening system, potent AMPs can be selected on the other features e design principles of many AMP classes are basis of their ability to inhibit or kill the host itself. e such as potency, relatively simple when compared with larger and more most severely inhibited host cells, expressing the most antimicrobial structurally complex enzymes. Correspondingly, rational potent AMP, can be identified and the amino acid spectrum, approaches have been used to design novel AMPs. Half sequence of the specific AMP determined by DNA haemolytic activity, of our lead series are novel biosynthetic molecules, which sequencing of the corresponding gene. In yet other toxicity and are not naturally-occurring. While it has proven fairly trans-acting screening systems (for example, Trans- systemic stability. easy to design AMPs with a basal level of antimicrobial Active-Peptide-system – patent pending), the AMPs activity, it has proven more difficult to optimise other are secreted and interact with target cells different from features such as potency, antimicrobial spectrum, haemo- the producing cells. is more traditional microtiter- lytic activity, toxicity and systemic stability. Structural or agar plate-compatible format allows the screening of modelling and various other computational analyses large libraries against a range of desired pathogens, or

78 Innovations in Pharmaceutical Technology Innovations in Pharmaceutical Technology 79 BIOTECHNOLOGY

Figure 2. Schematic flow-chart of Suicide Expression System (SES).

We have discovered the optimisation of other functionalities such as serum require different expression hosts and host-cell protec- that different stability or haemolysis. Finally, various computational tion strategies. Accordingly, we have categorised our structural classes Quantitative Structure Activity Relationship (QSAR) various AMPs according to structure and bioactivity, of AMPs require algorithms are used to predict antimicrobial efficacy, and have constructed complementary expression sys- different expression and hence limit the sequence space to be screened by tems for most of the lead classes. hosts and host-cell traditional methods. protection strategies. AMP Portfolio Accordingly, we Recombinant Production Platform Being the Currently, Novozymes has more than ten different have categorised world-leader in the production of industrial enzymes, classes of AMPs (see Figure 3). ese AMPs range in our various Novozymes has large recombinant production capa- size from 15 to more than 50 amino acids, and they AMPs according bilities. More than 500,000 metric tonnes of enzymes display a variety of antimicrobial activities. Some lead to structure and are produced each year at production plants located series (such as the NZ1000-, NZ5000- and NZ10000- bioactivity, and in Denmark, Switzerland, the US, Brazil and China. series) are broadly active against micro-organisms have constructed Recently, a fully operational cGMP production facility including Gram-positive and Gram-negative bacteria, complementary in Sweden was acquired in order to handle contract fungi and yeasts, while others are strictly antifungal expression systems production of pharmaceutical peptides. (NZ3000) or antibacterial with specificity for either for most of the lead To achieve the highest possible production yields, Gram-positive or Gram-negative organisms (for exam- classes. both fungal and bacterial strains have been optimised ple, NZ2000 and NZ4000). e major pre-clinical for the production of proteins and peptides. We have focus has been on two backbone-series: Novispirin discovered that different structural classes of AMPs and Plectasin.

Figure 3. Development status of Novozymes’ AMPs.

80 Innovations in Pharmaceutical Technology Innovations in Pharmaceutical Technology 81 BIOTECHNOLOGY

Figure 4. 3D-structure of Plectasin. Courtesy of Svend Plectasin an hour after microbial challenge eliminated Ludvigsen, Novo Nordisk. all bacteria in a mouse peritonitis model. is indicates that Plectasin, which can be recombinantly expressed on a commercial scale, is a promising candidate for further development as a novel therapeutic peptide against Gram-positive bacterial infections.

Perspective A few AMPs A few AMPs – bacitracin, polymyxins and gramicidins – – bacitracin, have been used for decades as efficient topical . polymyxins and However, the pharmaceutical exploitation of gene-encoded gramicidins – have AMPs has been slow so far. ere are several reasons for been used for Novispirins (NZ1000-series) e NZ1000-series (US this including a lack of oral availability and systemic decades as efficient 6492328), recently developed by Robert Lehrer and co- potential for peptides, high production costs and a lack of topical antibiotics. workers at UCLA, are small alpha-helical peptides with serious commitment from ‘big pharma’ – the latter prob- However, the broad antimicrobial activity against both Gram-positive ably being due to a reliance on traditional approaches pharmaceutical and Gram-negative bacteria, as well as certain yeasts and and a hesitance to consider lead research candidates that exploitation of gene- fungi. Despite this broad activity, very little haemolysis act on cell membranes. is has been compounded by encoded AMPs has or toxicity has been observed. A lead peptide from this the emergence of several small biotechnology companies been slow so far. series, NZ1001, has been tested in several topical skin in this field, developing a single lead against clinical and pulmonary animal models. For instance, in a wound indications with ill-defined clinical end-points, and with model, a fast and dramatic 4-log decline in the number limited financial resources. of bacteria was observed. is in vivo bactericidal activity Nevertheless, the emergence of multi-resistant against clinical isolates of both P. aeruginosa and S. aureus pathogenic bacteria and a corresponding demand for was exerted in a matter of minutes. In a rat pulmonary antibiotics with novel modes of action, combined with an infection model, NZ1001 significantly reduced the improved understanding of the universal nature of AMPs, number of bacteria and macro- as well as microscopic have prompted renewed interest in this unique group of pulmonary pathologies when compared with placebo. antimicrobial agents. At this point, more than 14 compa- ese pre-clinical observations indicate that the nies have – or have had – an active clinical development Novispirins are potent alternatives to current programme for AMPs; hopefully, one or more of these therapies directed against wounds, acne and other topical companies will succeed and secure further pharmaceutical infections. e potency, antimicrobial spectrum, lack of confidence in these fascinating compounds. spontaneous resistance and speed of bacterial elimination makes this class of antibiotics an ideal candidate for fur- Hans-Henrik Kristensen ther clinical development. received his MSc and PhD from the University of Odense Plectasin (NZ2000-series) e second lead series, the (Denmark), focusing on vari- Plectasins or NZ2000-series, has been isolated from the ous aspects of transcriptional saprophytic ascomycete fungus, Pseudoplectania nigrella. regulation in E. coli. In Plectasin consists of 40 amino acids and belongs to the 1997, for his postdoctoral Defensin-family of AMPs. Sequence analysis and homol- training, he joined the labo- ogy alignment indicates that the molecule has homology ratory of Professor Gary K Schoolnik, Chief of Infectious to the arthropod and mollusc defensin family of AMPs. Diseases at Stanford University School of Medicine (USA). e characteristic Cys-pattern is conserved, as well as sev- Dr Kristensen has been with Novozymes A/S for six years, eral other regions in the peptide, and the 3-dimensional and is currently responsible for the research and discovery structure resembles that of other defensins (Figure 4). efforts on antimicrobial peptides. Plectasin is selective in its antimicrobial spectrum and displays potent bactericidal activity against Gram-positive Debbie Yaver received a bacteria, with MICs and MBCs as low as 0.4 µg/ml PhD in microbiology from against clinical isolates of S. pneumonia. Analysis of the the University of California killing kinetics revealed that Plectasin eliminates >99.9% at Davis (USA). She has of the bacteria in less than 3 hours, with no haemolytic been with Novozymes for activity detected at concentrations up to 400 µg/ml. e 12 years. Research in her Plectasin cDNA has been cloned, and the fully matured department includes expres- and correctly folded peptide has been successfully sion of fungal enzymes, expressed in several industrial expression systems. In sys- development of fungal expression hosts, and discovery temic sepsis and pulmonary animal models, the potency and development of antimicrobial compounds. Dr and efficacy of Plectasin were comparable with those of Yaver is currently leading the preclinical development conventional antibiotics. For example, one IV dose of of a novel antimicrobial peptide.

82 Innovations in Pharmaceutical Technology Innovations in Pharmaceutical Technology 83