How to Make a Small Molecule Drug

William Ripka, one of the pioneers of structure-based drug design, highlighted the conceptual simplicity of drug design. “Conceptually, the idea of constructing a molecule that represents a geometric and electrostatic fit to a well-defined enzyme site would not necessarily appear to be a difficult problem.”

Although there have been significant advances in our abilities to design drugs, the challenge of constructing a molecule that is a geometric and electrostatic fit for the drug-target pocket in question remains a significant challenge still to this day.

“Conceptually, the idea of constructing a molecule that represents a geometric and electrostatic fit to a well-defined enzyme site would not necessarily appear to be a difficult problem.”

— William C. Ripka, 1987

The discovery of a small molecule drug can be separated into four aspects:

1. design & synthesize
2. selectivity & stability
3. pharmacology & adme
4. treatment & tolerability

design & synthesize

Medicinal chemists typically limit themselves to only seven atoms (H, C, N, O, S, F and Cl) when designing molecules. However, even seven atoms leads to, for all intents and purposes, an infinite number of possible molecules. The combination of these atoms into molecular scaffolds and functional groups which achieve Ripka’s geometric and electrostatic fit is the first key challenge of molecular design. Algorithms which predict molecular properties are now central to the design process, but despite these advances molecules still need to be synthesized and physically tested.

Drug design is an interative process driven by trial and error. This means that rapid synthetic access to designed molecules is a key aspect in order to achieve optimization goals in an acceptable time frame. Importantly, specific synthetic schemes need to be devised in order to perform late-stage variation of chemical functionality in important positions on the molecular scaffold.

selectivity & stability

Most contemporary drug discovery programs usually target one disease causing protein or a small number of structurally similar proteins. The flip side of this is to avoid off-targets – we call this selectivity. There are around 10,000 off-targets in the form of the expressed proteins from our 20,000 genes. The potential to interact with undesired areas of the proteome gets worse. Most of these 10,000 proteins have multiple domains for which there can be multiple druggable pockets. This adds up to a lot of off-target binding pockets which ideally need to avoided. Selectivity is a challenging parameter to quantify in drug discovery and can be a cause of significant late-stage attrition due to unwanted toxicological side effects.

The stability of a drug substance is arguably not as significant a challenge as selectivity but can raise its head early and later in the drug discovery process. Small quantities of drug candidates are synthesized for testing in in vitro assays at the early stage of drug discovery. The material is typically stored in the solvent dimethylsulfoxide (DMSO) at low temperature where DMSO is frozen and undergoes freeze-though cycles each time material is needed . A molecular structure needs to be stable enough to withstand this process otherwise spurious results occur overtime which can be arduous to trouble-shoot if chemical stability is not monitored.

Once a final drug candidate is identified a salt form with the best properties for development needs to be identified. Here scalability of synthesis and stability of the drug substance become of paramount importance. A salt form which allows ease of purification in the absence of chromatography and high stability under various conditions which simulate shelf-life stability is was it desired.

pharmacology & adme

Binding tightly and highly selectively to a given target is a challenging and pivotal milestone on the path towards a drug. Which pocket (eg orthosteric or allosteric) a drug binds to and how it fills that pocket determines the functional perturbation of the target. This step from binding to function is an equally important early step. These aspects are first studied in the in vitro setting using biophysical, biochemical and cellular assay systems for quantification. Significant complexity is added when a molecule progresses to in vivo: drug concentrations vary over time and cells are now placed in the 3-dimensional context of their respective tissues.

It is of course crucial that the drug reaches the compartment in the body where the disease cells resides and that it does so in concentrations so as the sufficiently modulate the target for the disease in question. This is where ADME, absorption, distribution, metabolism and elimination comes in. We (and other organisms) have evolved to avoid absorbing poisons thanks to membrane barriers and the so-called efflux transporter proteins expressed in them. Many drugs and drug candidates look like poisons (from an evolutionary point of view), and their oral absorption is limited by these evolutionary protection systems. The gastro-intenstinal tract represents a major hurdle for medicinal chemists which needs to be overcome. Sufficient drug also needs to be dissolved in the GI-tract as only the drug in solution is capable of passing through membranes. Once a drug is absorbed it starts to distribute, (typically via the blood stream) to tissues in the body while at the same time the drug gets broken down by metabolic enzymes which occurs largely in the liver. Ultimately, the drug is excreted largely via the faeces and urine.

treatment & tolerability

Dosis sola facit venenum – only the dose makes the poison, credited to Paracelsus highlights the central importance of the dose required for disease treatment. The dose required for disease treatment is the intersection of a molecules ADME and pharmacological properties. A therapeutic window is required between the dose of drug required for disease treatment and the dose of the drug which elicits unacceptable side-effect. Toxicological side-effects can be on-target (from the disease causing protein itself) or off-target (the other 10,000 proteins mentioned above). Last, but most importantly, patient selection can make the difference in clinical studies. Historically in cancer but increasingly in other therapeutic areas, patients who are expected to benefit most from a given drug can be selected using genetic markers or biomarkers.

Design & synthesize a molecule with sufficient stability & selectivity to reach, interact and perturb the function of the target such that a disease is treated with acceptable tolerability.