Drug discovery for coronary artery disease using an optimized dual-target software “Boltchem”.

INTRODUCTION

Coronary artery disease(CAD) or coronary heart disease(CHD) is a leading cause of death in many countries owing to the increased sedentary lifestyle and the unhealthy diet choices of today’s technology-driven world.

It is a condition in which the supply of blood to the heart is obstructed or blocked by the deposition of fatty acids in the artery wall(endothelium), mainly the coronary arteries supplying blood to the heart tissue. This condition can occur either when there is a clot(emboli) formed somewhere else in the body and lodges itself in the heart's coronary arteries and decreases the blood supply to the heart directly causing stroke or by forming a progressive plaque made up of fatty acids engulfing macrophages.

Coronary artery disease leads to ischemic heart disease where heart tissue doesn’t get enough oxygen to function and loses the ability to pump the blood causing heart attack.

SYMPTOMS

Symptoms for this condition are often misunderstood as an indigestion problem. however, it’s still better to get screened than regret later. Coronary artery disease initially is not symptomatic but as the disease progresses with the occlusion of the artery, symptoms begin to roll out affecting the capabilities of the heart.

The variety of symptoms are:

Chest pain(angina)

Fatigue

Chest tightness

Chest burning

Pain in arms and shoulders

Chest squeezing

Unexplainable sweating

Dizziness

Shortness of breath

PATHOPHYSIOLOGY OF THE DISEASE

This plaque initially doesn’t produce any symptoms, the symptoms arise only after the progression of the plaque to moderate to severe or when an emboli lodges itself in the already plaque artery and occlude it fully blocking the entire blood flow to the heart tissue. The symptoms begin to show at this stage as more platelets aggregate and decrease the blood flow and on the other hand, increases the myocardial oxygen demand.

Sometimes this plaque ruptures causing haemorrhage and blocks the artery entirely causing acute myocardial ischemia which may lead the patient to further 4 conditions.

1. Acute myocardial infarction(Heart attack) is the term given to the death of the cells or the tissue whose arterial blood supply was blocked by the plaque. Due to the lack of oxygen heart loses its ability to pump the blood-producing symptoms like pain in the chest and shoulder, this can also occur when the patient is at rest.
2. Angina pectoris is a symptom involving chest squeezing and tightness which may occur not only while walking or exercise but may occur at rest and while sleeping too (Variant angina). these symptoms are usually severe in form.
3. Arrhythmias are irregular heartbeats occurring due to the irregular blood flow to the heart, this alters the heart’s ability to pump the blood to the brain and other vital organs and produces symptoms like fatigue, dizziness, shortness of breath, feeling light-headed.
4. Sudden cardiac death, is different from a heart attack. In this the electrical activity in the heart malfunctions and makes the heart beat faster causing ventricular fibrillation and the supply to the brain drops drastically making the patient lose consciousness.

Death can occur in minutes of the onset of symptoms. Giving CPR(cardiopulmonary resuscitation) or using AED(automated external defibrillator) to the patient can increase their chances of survival.

DRUG DISCOVERY

Where a discovery of a drug for efficiently treating or benefitting a disease takes about years until it reaches the market and the patient population, using artificial intelligence will help in the reduction of this significant year gap of discovery to reach the patient population.

Here, using Boltchem-an optimised dual-target small molecule generating artificial intelligence to produce a novel drug for a reduction in coronary artery disease.

TARGET IDENTIFICATION:

Coronary artery disease is a complication frequently occurring in the population having diabetes, this may be due to the metabolic dysfunction and insulin resistance involved in diabetes.

The dual targets selected for the drug discovery were SIRT1 and PPARG.

SIRT1(Silent mating type Information Regulation Two 1) is a gene that is involved in a variety of functions like glucose homeostasis, lipid or fat homeostasis, increasing insulin sensitivity this gene is involved in managing the energy consumption of the body. SIRT’s are 7different gene’s divided into 4 different classes. Among all the types SIRT1 is the most studied gene and studies also found that it has protective mechanisms for protecting the body from oxidative stress and preventing the endothelium of the blood vessels.

The second target PPARG( Peroxisome Proliferator-Activated Receptor Gamma)gene is an important gene in fatty acid metabolism. In inflammation, activating PPARG can protect against atherosclerotic plaque balancing the metabolic abnormalities usually found in diabetes mellitus through activating Tcell in fatty tissue. The PPARG gene also has the potential to reduce cardiovascular risk in diabetes patients.

As both the targets are involved in the fat homeostasis, targeting this pair would provide an added benefit to the chronic inflammatory condition i.e diabetes which poses a high-risk factor for coronary artery disease.

LEAD IDENTIFICATION

Property prediction:

SIRT1 had 960 hits and PPARG had 6840 hits on bindingdb and 3,000 tested compounds for SIRT1 and 1,047,500 tested compounds for PPARG on the PubChem database and all the ligands were downloaded for property prediction. This ligand data was run for the property prediction experiment for property prediction with the IC50 threshold values summarised from the literature of the molecules acting on the same targets as the chosen one’s. Next, for the validation of the predicted property. we used the data belonging to the drugs which were already on the market. The IC50 value of the already existing drug was converted into a log scale as the experiment predicts the property in the log scale and then compared for validation.

Substructure generation:

After the validation of the predicted property. The molecules whose property was predicted and validated is sent for the substructure generation experiment. The substructures for both the targets were generated by specific threshold values and for specific atom sizes. Boltchem’s substructure generator produced approx. 900 substructures for SIRT1 and 1600 for PPARG.

Merge Substructure:

Both target substructures were fed to the merge substructure experiment of the BoltChem platform to produce merged substructures for lead molecule generation which took about a day or two. Over 1200 merged substructures were produced in this experiment. The merged molecules would provide a base for the lead which will be generated in few days rather than months to produce the lead for the disease.

Generator:

The generator experiment of the BoltChem platform utilises the already fed data points of SIRT1 and PPARG and generates the lead molecules in just 2 hours or so and these leads undergo novelty to test for their uniqueness and qualitative drug-likeness (QED) to produce a potential lead for experimentation and comparison with the already available drug molecules.