The Garden of Antimicrobial Delights

If we are to continue with the current practices in the treatment of infectious disease, new and reliable sources of novel antimicrobials are badly needed for all types of microbial diseases. Where will they come from? And who will produce them? I take issue with the oft-repeated assertion that it is “hard” to find new compounds with the required therapeutic activity from nature, the assumption being that Big Pharma has exhaustively screened all of the available sources of bioactive compounds and that slim pickings remain. Emerging technological advances make it increasingly clear that this stance is indefensible.

If we assume that upwards of 100,000 different strains of bacteria and fungi may have been screened for therapeutic agents by the industry,² this represents but a fraction of the population of the microbial kingdoms (a gram of soil may contain a thousand different bacterial species). There are likely billions of antibiotic-producing strains in the biosphere, each with the potential to produce 10 times this number in terms of bioactive small molecules with drug potential. This information comes from two sources. First, more than 90 percent of microbes cannot be grown under laboratory conditions. This includes all possible environmental sources: marine, terrestrial, endosymbionts, and so on. Second, extensive genome sequencing over the past 10 years has revealed that one of the most productive bacterial sources of antibiotics and other therapeutic agents, the streptomycetes, possess the genetic capacity to make as many as 30 different bioactive small molecules.³ The so-called cryptic biosynthetic pathways are not expressed (or are poorly expressed) under traditional fermentation conditions, and their products have been missed because of inadequate and insensitive screening processes.

This treasure trove of bioactive compounds will not yield its secrets and benefits easily. New methodology must be developed to grow fastidious microbes in the laboratory and creative approaches to the cloning and expression of identified, but product-unknown, biosynthetic pathways must be developed; new expression hosts and pathway expression processes are essential for success. Another major hurdle is the isolation, purification, and structure determination of the compounds produced by the microbial hosts; traditional natural product chemistry is almost a dark art, with very few skilled practitioners worldwide. However, new methods of cell-based screening, compound isolation, and structure determination are coming on the scene; increasingly sensitive instrumentation may resolve other aspects of the discovery problem. There is an excellent model to be found from the revolution in DNA sequencing—it took several billions of dollars to obtain a draft nucleotide sequence of the human genome in 2000; 10 years later we are speaking of $1000 genomes (or less!).⁴ More important, for low-molecular-weight bioactives, microbial genome sequences will soon be priced at $10 a shot and completed in a day. Antibiotic discovery will start with the complete nucleotide sequence of a biosynthetic pathway for an unknown compound, reversing the traditional approach.

Of course, the identified pathways still have to be converted into products, but prototypes for this type of engineering are already established; even Escherichia coli is considered a good candidate as expression host.⁵ In addition, the possibility of creating new types of molecules by making hybrid biosynthetic pathways is a field of investigation that will develop in parallel with these approaches.⁶ Engineering the amino acid components of nonribosomal peptides is yet another method that shows promise for the production of nonnatural molecules in vivo.⁷

To capitalize on the increasing availability of new classes of small-molecule therapeutics, comparable advances in high-throughput screening and testing of their diverse bioactivities using cell-based assays will be essential; one can predict more extensive use of reliable insect and worm models of human disease.⁸ Imagine screening thousands of microbial compounds against hundreds of disease syndromes in microtiter plates! Many of the necessary procedures are already in place but they need to be made amenable to high-throughput analysis.

Also needed is a more comprehensive understanding of natural microbial communities and their interactions with their hosts, especially in humans. Studies of the nature of human and animal microbiomes are revealing extensive roles of commensal microbes in the health and maintenance of their complex hosts. Advances in this field will surely reveal unanticipated drug targets for a variety of afflictions.

All of these technological advances—and more, such as systems biology and metabolic network mapping—will contribute to the means of producing newer generations of antimicrobial agents. But who will be responsible for the implementation of these revolutionary developments in the process of drug discovery and development? Big Pharma is too hidebound by the tradition of overmanagement, and “small” biotech is too dependent on venture capital. What is needed is a new drug discovery model—perhaps university research combined with government grant funding. A number of universities around the world have initiated preclinical, preindustry discovery groups with the goal of bridging the freedom and originality of academic research with the hard facts of drug development. This should be the way of the future!

This essay is adapted from F1000 Biology Reports 2010, 2:26 (http://f1000biology.com/reports). See the original paper for a deeper discussion of the proposed technologies, complete reference listings, and links to their faculty evaluations.

Julian Davies is head of faculty of Microbiology at F1000. He is a Fellow of the Royal Societies of the United Kingdom and Canada. He is past president of the American Society of Microbiology and the International Union of Microbiological Societies.