Many medications come from nature. An animal, a bark, or a fungus can hide inside a chemical molecule that can affect our body. Sometimes a compound that acts as a poison in high concentrations can have an antitumor or antibiotic effect in the right concentration. For example, some fungi secrete antibiotic substances that we have not hesitated to isolate for our personal use, such as penicillin.
But what if we want a completely new medicine? Something that does not exist in nature and we cannot be inspired by any known molecule. In this case, we must go into pharmaceutical chemistry and its strategies to make a drug from scratch.
Aiming at the target
To start, we need a target. This is what the protein we want to affect is called through the binding of our molecule. Depending on the target, we will be looking for one type of medication or another. For example, if our molecule manages to bind to any protein involved in the growth and proliferation of a pathogenic bacterium and inhibit its functioning, we will have a possible antibiotic.
This possible candidate drug, then, will have to go through a series of clinical and safety trials. First it will be tested through cell cultures, then in experimental animals and finally humans. An arduous process in which the majority of candidates are discarded. It is estimated that approximately for every ten thousand candidates, only one will be effective enough and with few side effects to be worthwhile. For that reason, the more new candidates we can find, the more likely we are to obtain some new hidden drug at a glance.
A protein consists of a long chain or several chains of amino acids, twisted in space forming a three-dimensional structure. Not all parts of this structure will be equally important: in some joining a small molecule will not have any effect, but in others we will cause an interference in the normal functioning of the protein. These points are called active regions, and they are usually hollows that have a given size and load. For example, if a protein works through the union with oxygen, it will have an active region with the perfect way to join that molecule, a negative template that fits like a glove.
If we really want to affect the normal functioning of the target protein, we must look for molecules that fit into those same gaps, blocking, competing or displacing the compound that must bind in the usual way.
If we have the protein well studied, then we can know the exact shape of that hole and we can even model it on the computer. In this way we can simulate what the union of different candidate molecules would be like without any experiment. Although these candidates are finally tested in real life and sometimes bring surprises, the fact of being able to experiment at low cost makes it the favorite technique for many laboratories and pharmaceutical companies.
But let’s get in the worst case: we don’t know well the form of the target protein and its active regions. The reasons may be several. The protein may have some mobile structure that changes easily. Or, basic science may not have had the resources or the time to discover its form. For these cases, a strategy called “discovery directed through fragments” is used. The idea, a priori, is simple: If we don’t know which molecules work against the target, let’s try them all.
Try and failure
Although it may seem crazy since there are an infinite number of combinations of molecules in the world, we can give it a try. The truth is that, if you look, many molecules have a similar shape to each other. If we look to fit in the hollow of the active region, we can try small fragments, see those that fit better and cause an effect. Then we play to modify them and grow them little by little.
In this way, the initial small fragment works like a Christmas tree, to which we are adding modifications. We can exchange a hydrogen atom for a lithium atom, which is a little bigger. Or add a small aromatic ring of carbon atoms that gives it greater stability in exchange for a larger molecule size at its left end. With each change it is checked if it has had any effect on the target, always seeking to maximize it.
This strategy is long and requires many repetitions of trial and error, but it is still cheaper and faster than randomly testing molecules. Gradually you get a candidate molecule that manages to affect the target with good success, and in fact, the intermediate steps are also used as possible candidates.
This system has been used for the past ten years and with quite good success. Several antitumor agents in the market have been created in this way, placing proteins that are especially present in tumor cells as the target.
To do this technique, laboratories use an already established mixture of organic molecules, which includes between 100 and 1,000 molecules that are very different from each other. All of them have a small size but a great variety in their composition and shape, ensuring maximizing the chances of finding a candidate molecule … but there is a problem: most of these molecules are flat.
This preference for the flat comes from the organic chemistry itself. Organic molecules are mainly formed by chains of bonds between carbon and hydrogen, and although some have the ability to rotate and rotate like a snake, they usually grow in a two-dimensional manner. It is not easy to see organic molecules in the shape of a double pyramid with two floors.
And this lack of three-dimensionality makes us lack a chemical space to discover. The active regions of the target that we want to attack really have three dimensions, and there may be three-dimensional molecules that allow us to act as scaffolds for drugs never seen before.
A team of American and French scientists has worked as a team to address this problem. Since pure organic chemistry does not favor these types of structures, they have synthesized a collection of metalorganic molecules, which differ by having some atom of a metal, such as cobalt or nickel. Some of these atoms allow more bonds than carbon and we can easily build three-dimensional molecules, some sandwich-shaped, others pyramid-shaped. And so up to seventy-one molecules, all very different from each other.
To prove their potential, they tested their collection by looking for candidate drugs against three specific targets: a protein that confers resistance to antibiotics in bacteria, a protein complex that replicates viral RNA in influenza A and a protein that appears increased in some tumors and favors its growth. In all three cases, candidate molecules for new treatments were found, at the same time proving their effectiveness in creating antibacterial, antiviral and antitumor drugs respectively.
With some luck and patience, some of these new fragments will act as scaffolding for some totally new and never seen medication. If nature does not give us new ideas, we will have to use our ingenuity and create new forms.
DON’T KEEP IT UP:
- This article focuses on the first step in obtaining candidates for new medications. Then all these candidates must pass thorough efficacy and safety tests to be able to come to light.
- Although some medications have a natural origin, the dose is controlled in the drugs and any impurities are cleaned to improve their effectiveness. Obtaining the same medication from its original source may be ineffective at best or even dangerous.
- Morrison, Christine N., et al. “Expanding Medicinal Chemistry into 3D Space: Metallofragments as 3D Scaffolds for Fragment-Based Drug Discovery.” Chemical science, (2020)
- Scott, Duncan E., et al. “Fragment-Based Approaches in Drug Discovery and Chemical Biology.” Biochemistry, vol. 51, no. 25, pp. 4990–5003 (2012)