Drug development is in it for the long game. The discovery phase alone can take around ten years before the potential medicine is approved for clinical trials in humans.
But have you ever wondered just how we discover these drugs? Perhaps with the recent rollout of new COVID-19 vaccines, you’ve thought: where do these medicines even come from?
Modern medicines have come from a variety of places. Penicillin was discovered by chance in a mouldy Petri dish, whereas willow bark has a widespread historical use that was later mass-produced as aspirin for the modern era. However, now that many countries have an ageing population, we can’t rely on chance encounters to discover new drugs. Researchers in the pharmaceutical industry use more streamlined processes to find potential new treatments.
Think of drug discovery like a gruelling job interview (The Apprentice for molecules, perhaps), where thousands of candidates jump through multiple stages with increasing levels of scrutiny. On top of that, each molecule only has a 0.1% chance of succeeding in the next stage. Sounds like one difficult interview!
Let’s dive into the world of drug discovery:
Bullseye!—The Targeted Approach
Before any drugs are involved, researchers look at the pathophysiology of a disease to understand how best to approach treatment. An important example is the ongoing research into the COVID-19 virus.
Once researchers understand a disease pathway, they have to find a specific target to attack. A drug molecule has to act on something, somewhere within the body.
A target is usually a gene or protein, like an ion channel or enzyme. Good targets should be accessible for drug molecules to bind to, which leads to a therapeutic effect. This is the “job” that our candidates are interviewing for, so we ask: what’s the role our drug molecules have to perform for?
Cyclooxygenase-2, or COX-2, is an enzyme that is part of the pain and inflammation pathway. Inhibition of COX-2 causes a decrease in inflammation, which is how ibuprofen and other anti-inflammatory drugs work.
The two main steps to a targeted approach are identification and validation, which can happen alongside each other.
1. Target identification
We can use different techniques to find the best target for a given disease.
Genetic engineering, or mapping the genome to find a protein, can help figure out which genes cause a change in certain phenotypes. Once we identify a gene to be part of the disease pathway, we have to see if we can control the resultant protein with a drug, otherwise, there’s no point in trying to attack it.
A phenotype is an observable or physical trait, like hair colour or height.
We can then screen large compound libraries to see if any small molecules bind to the proposed target. There’s also the virtual approach, in which we model the target protein on specialised computer software.
These techniques are best used in combination, but some researchers may favour one over the other.
2. Target validation
Validation means to confirm that this particular gene or protein is involved in the disease pathway and that altering this target will have a biological effect.
By understanding the relationship between a specific target and the relevant disease, we can predict any potential side effects that may occur from modulating a target.
The target needs to be “druggable”, which means it should have a high affinity to binding with a drug molecule. If the molecule can’t access the target, then it’s pointless to pursue it.
The affinity of a drug refers to how tightly a drug binds to its target, due to the forces between various chemical groups in both structures.
To find out if a target is likely to be druggable, we can look to other drugs that are already on the market and see what their targets are because we know they can be modified to produce a clinical effect. We can also look at the target’s chemical structure using computer modelling, and predict what sort of molecules will have an affinity for it.
After we’ve put up the job advert, we can start to look at the applications we’ve received. Well, 5,000-10,000 applications to be exact (it’s a tough market).
The Hit Discovery Process
A ‘hit’ is a molecule that exhibits a therapeutic effect on our biological target.
A fast way to find multiple hits is via high throughput screening (HTS), where we test molecules from compound libraries in an automated process. The benefit of this method is screening a very large number of molecules in a short space of time.
HTS involves assays (analytical procedures). An assay allows us to test whether any of these thousands of molecules will bind to the target and cause a reaction, which is measured.
HTS uses microplates, square plates with many small wells, almost like mini test tubes.
Once a selection of hits has been identified, they will be further confirmed with different types of assays. This can be a time-consuming process: one assay can take several weeks from start to completion!
Several hits are identified at this stage and we’ll assess each one for its potency and affinity. This involves collecting data on dose-response curves, structural-activity relationship (SAR) between the target and hit, and basic pharmacokinetic (PK) data.
Now we’ve narrowed it down to about 10-20 interview candidates—great! We just have to ask each of them more specific questions about their abilities and hope that one of them will truly shine.
Hit to Lead (H2L) Phase
The H2L phase optimises the hits we’ve discovered by further investigating PK and SAR studies. This will refine each molecule’s potency and efficacy because they need to meet a level of biological activity that brings about a therapeutic effect.
We measure this activity level in several ways:
- Selectivity—non-selective drugs, or ‘dirty drugs’, affect many different receptors, causing all sorts of side effects and reactions. We want the opposite: for the drug to be selective for the target above other receptors.
- Solubility—our bodies are famously fluid-filled, so the molecule needs to stay stable in that fluid.
- Permeability—to get access through the lipid bilayer cell membrane isn’t easy. Some drugs are small enough to slip through, but others may need to utilise a channel protein.
- Metabolic stability—the drug needs to avoid being broken down before it meets its target.
- Low Cytochrome P450 inhibition
Cytochrome P450 (CYP) are a group of enzymes usually found in the liver that metabolise many drugs. If a drug molecule has a high inhibition of CYP450, it’s less likely to be a successful candidate as it might disrupt the metabolism of other drugs and cause accumulation. On the other hand, being an inducer of CYP450 means other drugs will lose their therapeutic effect.
This is a tricky stage as optimising one aspect of a molecule can have an impact on another characteristic. For example, increasing the drug’s permeability of the cell membrane might involve increasing its lipophilicity (ability to dissolve in lipids). However, this may change the hydrophilicity (ability to dissolve in water) of the drug in the intracellular fluid, which poses a problem for its solubility.
Once they’ve answered all their questions, we know which ones are more likely to succeed and which ones just won’t make the cut. Now there’s a final step in the process left.
Lead Optimisation—A Final Hurdle Before Clinical Trials
Lead optimisation requires the research team to collate enough data to prove to regulatory bodies that these molecules are ideal candidates in clinical trials, especially their safety and efficacy profile. Yep, you’ve got the pattern by now: this means another round of optimising and refining our candidate molecules, usually following the same principles as we’ve mentioned before.
As the process of optimisation tends to repeat itself throughout the whole drug discovery phase, a lot of these stages will overlap. If you can spot an issue in the hit discovery phase, it makes sense to resolve it sooner rather than later.
Drug Discovery—Costly but Crucial
We made it! But this is not the end of the journey for these molecules as they face an even bigger challenge: clinical trials.
Drug discovery is incredibly costly, but the probability of discovering a new drug by chance is slim. Though it can take over £400 million just to reach clinical trials, it’s far more likely that we can hopefully discover a new medicine that changes and improves lives by following these methods.