By Kevin She
3 Feb 2016
Humans have been using plants and plant parts since antiquity to effectively treat a large number of illnesses and conditions; the effectiveness of these natural medicines is based on the small molecules that are produced and accumulate in the material. As traditional and folk medicine gave way to modern chemistry, it was increasingly recognized that in many cases the effectiveness of these natural products are based on a single or a few specific small molecules.
In the modern era, generally, two major routes of new drug discovery were pursued; modification or screening. Introducing modifications to existing drugs and seeing if a more useful compound was created has been a mainstay of “new” drug discovery. Another approach is through rational design which uses modeling to predict which modifications could achieve specific effects such as increased/decreased binding affinity to its target or adding side chains to improve permeability. Indeed, a vast constellation of drugs have been derived from the products of opium as well as cocaine. Many modern cancer chemotherapeutic drugs were natural products and derivatives (Wilson & Danishefsky 2006) or are based on modifications of those existing drugs through both trial and error as well as by rational design (Patani & LaVoie 1996).
Today, partially because of perceived exhaustion of “low hanging fruit” (for both biologically simple diseases with “drugable” targets as well as useful small molecules) and the current stable of existing drugs developed over the past century, new drug discovery and development is vastly more complex than in the past. Additionally, the development of antibody- and other protein- based biologics appears to be spearheading current drug research programs in lieu of small molecules.
Regulatory oversight and approval is also profoundly stricter than in the 19th century and can be an additional economic hurdle when developing new drugs. Developing a new drug from an original idea to the launch of an approved product can take 10 – 15 years and cost from hundreds of millions of dollars to in excess of a billion. Much of the cost comes from failures, which accounts for about 75 percent of total research and development costs. While failures are disappointing, they still contribute to further understanding of the disease that is targeted (Corr & Williams 2009).
The first step to one approach of modern drug development begins in the laboratory where basic scientific research is conducted to investigate the biological origins of a disease. Usually the research is conducted at academic institutions and funded by governmental research grants. The hopes of this class of research are to identify targets that may be accessible to a known drug molecule. For example, salicin, salicylic acid, and acetylsalicylic acid are cyclooxygenase (COX) inhibitors. Humans possess two variants, COX-1 and COX-2 with different distribution in the body with COX-1 being abundant in the digestive tract. These salicin-based drugs bind and inactivate both variants, but affects COX-1 more strongly and the binding is irreversible. This is one of the causes of the side effects that include gastric irritation, bleeding, and diarrhea. To develop a non-steroidal anti-inflammatory drug that is superior to aspirin, the drug target would be COX-2 and ideally, from a theoretical standpoint, the drug would only inhibit COX-2 and not COX-1 and that the binding (and thus inhibition) would not be permanent.
Subsequently, further research must be conducted to determine whether the target and the approach is valid (target validation). In the above example, extensive research was performed in vitro and in animal models to demonstrate, among many other things, that COX-2 inhibition is sufficient to treat the condition, that inhibition of only COX-2 avoids the side effects seen with aspirin, and that selective COX-2 inhibition doesn’t lead to other side effects.
Following target validation is hit identification. Typically, this is conducted through screening libraries of small molecules. Libraries are collections of various small molecules whose structure and composition are known, and these collections may be of 1) naturally occurring small molecules, 2) a series of synthetic molecules of various classes, or 3) a combination.
Historically, libraries were composed of whatever small molecules had been collected and characterized. As the field of medicinal chemistry advanced, libraries evolved into collections of small molecules with similar chemical characteristics that are predicted to interact with certain types of proteins in a certain way. For example, in the early 1960’s it was known that there were excitatory receptors in the central nervous system that were responsible for instructing neurons to fire and that the amino acid glutamate was probably the endogenous (naturally occurring in the body) signal for these receptors. It was also known that there were at least several different types of receptors that all respond to glutamate, but in different ways. In order to try to identify specific glutamate receptors in the brain, about a hundred small molecules with a similar structure to glutamate were synthesized and screened (Curtis & Watkins 1962).
The actual screening – assay development – can be done in vast variety of ways but typically starts with biochemical studies with purified or synthetic target protein and typically progresses to cell-based assays to determine whether the same interaction occurs in living cells. In the above example, compounds were individually tested to see their individual effects on live neuronal cells, these experiments having preceded the development of recombinant proteins. A few different compounds were found to have strong activity and subsequently it was shown that N-methyl-D-aspartate specifically activated one kind of receptor, which was subsequently named after the synthetic molecule (NMDA receptors). Similar experiments would identify other families of excitatory neurotransmitter receptors such as the AMPA- and kainate- type receptors.
In successful drug discovery programs there are typically multiple hit candidates. Candidate molecules that pass this stage are considered leads. The next step is hit optimization which is to determine which hit to pursue and whether chemical modifications to each of those hits can improve their performance. After extensive laboratory and animal testing, a pre-clinical candidate molecule may have been found. Within industry, each project may have screened 200,000 to 1,000,000 distinct compounds to identify about one hundred leads, and hopefully results in one or two pre-clinical candidates. Of these, only one in ten pre-clinical candidates enters clinical study. At the clinical study stage, again, only one in ten candidates end up reaching market (Hughes &al., 2011).
Using the above example, several COX-2 specific inhibitors have been developed such as celecoxib, rofecoxib, and meloxicam.
The path of biologics development and regulatory oversight is even more complicated than the above sketch of the development of small molecule therapeutics. However, they are potentially much more flexible, specific, and when designed well can overcome general toxicity issues associated with small molecules and are gaining both market share and clinical utility. Discussion on biologics and the development of biologic pharmaceuticals are outside the scope of this article.
Although natural product based therapeutics development has been declining in the face of synthetics and biologics, interest is being renewed in natural products. Synthetic small molecule libraries are often limited to small molecules that can be easily synthesized from precursors; due to the physics of conventional chemical synthesis, only so many different “shapes” of molecules can be synthesized with a high enough efficiency to be worthwhile. However, enzymes can catalyze many chemical reactions that chemical synthesis cannot perform at all or are hugely inefficient.
Actinobacteria (commonly referred to as actinomycetes), in particular Actinomyces and Streptomyces, are a class of bacteria that live primarily in soil but are also found in water bodies (streams, ponds, seawater, &c.). These environments are nutritionally, physically, and biologically complex. Actinomycetes are among the most numerous and ubiquitous bacteria in these environments and are crucial in the carbon recycling of the remains of other organisms which are rich in biopolymers like lignocelluloses and chitin which very few other organisms have the means (enzymes) to break down.
Figure 1 Left: some wild actinomycetes colonies grown on solid agar media, Right: laboratory E. coli colonies grown on solid agar media.
Actinomycetes, when grown in culture, have very unique morphologies (appearance) compared to other bacteria such as E. coli. Note the variety of colours and the”shadow” around some of the colonies. The colours are a legacy of the large variety of small molecules that actinomycetes produce. The shadows are the accumulation of small molecules that have been excreted into the agar media by the actinomycete colony as they grow.
One of the most interesting things about actinomycetes is that over 10% of its relatively large genome has evolved to actively recombine to produce novel enzymes and proteins that manipulate small molecules; these include many enzymes that break down environmental macromolecules for nutrients and energy as well as mechanisms for producing a vast array of small molecules including those with antibiotic activity.
Indeed, about 80% of all antibiotics for human use come from actinomycetes (Watve &al., 2001) and include streptomycin, chloramphenicol, neomycin, tetracycline, erythromycin, vancomycin, kanamycin, and daptomycin among many others. Two anti-rejection drugs, FK506 (Tacrolimus) and rapamycin (Sirolimus), were discovered in actinomycetes and revolutionized organ transplantation. Numerous other drugs also have their origins in these soil dwelling bacteria.
Bently &al., 2002
Figure 2 Graphical representation of the Streptomyces coelicolor genome.
The regions outlined in dark blue are the “core” regions and regions outlined in light blue are the “arm” regions where extensive recombination occurs. Depending on the lifecycle stage, the Strepomyces genome can exist in a circular or linear state.
One of the evolutionary legacies behind the success of the actinomycetes was the development of “non-ribosomal polypeptide synthetases” (NRPS). Normally, proteins and peptides are made in ribosomes – messenger RNA is threaded through a ribosome and amino acids are linked together based on the sequence of the messenger RNA. NRPS, on the other hand, don’t rely on messenger RNA to string together amino acids into proteins. Instead, they are very large proteins with many different subunit enzymes. Each subunit enzyme is able to catalyze a specific chemical reaction. In functional NRPSs, amino acids are strung together one by one according to the sequence of the subunit enzymes. Perhaps the most striking thing about NRPS is that they can incorporate amino acids that other organisms can’t use to make proteins. The ability to incorporate these exotic amino acids, sometimes in positions that ribosomes are unable to catalyze, vastly increases the diversity of potential shapes of small molecules that can be generated. Furthermore, some NRPS subunits are also able to add or modify lipids and sugars to the budding small molecules to even further increase the diversity of shapes that can be made.
During the development of daptomycin (Cubicin), efforts were undertaken to perform hit optimization on the molecule through synthetic chemistry. The exotic nature of the molecule, a lipopeptide (small protein-like molecule with lipid modifications), was assessed to be impractical – if not impossible – to synthesize through synthetic chemistry and that the complex chemical structure was not amenable to rational design via modification through conventional chemistry. Instead, efforts were then directed to manipulate the gene that makes the NRPS that makes daptomycin by replacing parts of that gene that codes for individual enzyme subunits with those found in other NRPS through molecular biology.
Figure 3 Skeletal formula for daptomycin.
Daptomycin, one of the antibiotics of last resort, is produced by a non-ribosomal polypeptide synthetase (NRPS) by stringing together amino acids, including rare non-proteinogenic amino acids, and lipids to create a lipopeptide with a structure that is very different than secondary metabolites (small molecules) produced in other species.
Curiously, although NRPS derived small molecules contain amino acids in their makeup and are classified as peptides (or lipopeptides or glycopeptides if they contain lipids or sugars), these molecules are treated as “New Chemical Entities,” like small molecule drugs, by regulatory bodies and are not considered “biologics.” Incidentally, vaccines – which contain viral proteins or peptides – are regulated by the FDA in the same category as blood and biologics.
The astounding diversity of small molecules that can be produced by the various biosynthetic pathways in actinomycetes far outstrips the molecular diversity readily achievable by synthetic chemistry and represents an as yet only sparsely explored frontier for drug discovery and development. Although many actinomycetes are cultivable in laboratory conditions, the majority of actinomycetes remain uncharacterized and unexamined. However, new methods have been developed to grow actinomycetes in their native environments and an increasing number of “missing ingredients” that allow exotic actinomycetes to grow in the lab are being identified (Lewis &al., 2010).
Bently, SD., Cater, KF., Cerdeno-Tarraga, AM., Challis, GL., Thomson, NR., James, KD., Harris, DE., Quail, MA., Kieser, H., Harper, D., Bateman, A., Brown, S., Chandra, G., Chen, CW., Collins, M., Cronin, A., Fraser, A., Goble, A., Hildago, J., Hornsby, T., Howath, S., Huang, CH., Rajandream, MA., Rutherford, K., Rutter, S., Seeger, K., Saunders, D., Sharp, S., Sqaures, R., Squares, S., Taylor, K., Warren, T., Wietzorrek, A., Woodward, J., Barrell, BG., Parkhill, J., & Hopwood, DA. 2002 Complete genome sequence of the model actinomycete Strepomyces coelicolor A3(2). Nature 417: 141-47
Corr, P. & Williams, D. 2009 Conflict of interest in medical research, education, and practice. Lo, B ., Field, MJ. National Academies Press (US). Washington (DC).
Curtis, DR. & Watkins, JC. 1962 Acidic amino acids with strong excitatory actions on mammalian neurons. J. Physiol. 166: 1-14
Hughes, JP.,, Rees, S., Kalindjian, SB., & Philpott, KL. 2011 Principles of early drug discovery. Br. J. Pharmacol. 162(6): 1239-49
Lewis, K., Epstein, S., D’Onofio, A., & Ling, LL. 2010 Uncultured microorganisms as a source of secondary metabolites. J. Antibiot. 63(8): 478-76
Patani, GA. & LaVoie, EJ. 1996 Bioisosterism: A rational approach in drug design. Chem. Rev. 96: 3147-76
Watve, MG., Tickoo, R., Jog, MM., Bhole, BD. 2001 How many antibiotics are produced by the genus Streptomyces? Arch. Microbiol. 176: 386-90
Williams, RM. & Danishefsky, SJ. 2006 Small molecule natural products in the discovery of therapeutic agents: the synthesis connection. J. Org. Chem. 71(22): 8329-51