Functional groups and what they do

About twenty-five functional groups account for the overwhelming majority of marketed drugs. This module is where a drawing stops being a drawing and starts predicting behaviour: for each group, whether it donates or accepts a hydrogen bond, what charge it carries at pH 7.4, how much polarity it adds, and what the body will do to it.

Whenever you meet a functional group, ask the same four questions. They are the ones that determine whether a molecule reaches its target and what it does when it gets there.

• Can it donate a hydrogen bond? It needs a hydrogen on nitrogen or oxygen. A C-H does not count.
• Can it accept one? It needs an available lone pair — one that has not been spent on a pi system.
• What charge does it carry at pH 7.4? Not the charge as drawn: the charge in plasma. Module 10 makes this quantitative.
• What will metabolism do to it? Some groups are cleared in minutes and some survive indefinitely.

The table below is the one you should build for yourself. Fill it in from what you already know, check it, and keep it.

Three pairs of groups get mixed up constantly, and each confusion leads somewhere expensive.

This is the big one. An amine nitrogen has an available lone pair: it is basic, it is protonated at pH 7.4, and it carries a positive charge. An amide nitrogen has given its lone pair to the carbonyl, so it is not basic, it is neutral at every physiological pH, and it is a hydrogen bond donor but a poor acceptor.

Two structures that look almost identical on paper therefore behave completely differently in the body. Get this wrong and you will predict the wrong charge, the wrong solubility and the wrong binding mode.

An ester has a carbonyl; an ether is just an oxygen between two carbons. The ester is hydrolysed by esterases, often within minutes of reaching plasma. The ether is not, and it is one of the more metabolically boring groups in medicinal chemistry.

That difference is exploited deliberately. An ester is the standard prodrug handle: hide a carboxylic acid as its ester, get the neutral molecule across the gut wall, and let plasma esterases unmask the acid once it is inside. Enalapril is the textbook case — it is the ethyl ester of enalaprilat, and enalaprilat itself is barely absorbed.

Both have a sulfur carrying two oxygens. The sulfonamide has a nitrogen on the sulfur; the sulfone has two carbons. The nitrogen makes the difference: an N-H sulfonamide is weakly acidic, with a pKa around 10, so it is partly ionised in plasma and it is a hydrogen bond donor. A sulfone is neither.

Now the point of all this. Three groups have consequences large enough to shape whole programmes.

A carboxylic acid has a pKa near 4, so at pH 7.4 it is more than 99.9% ionised. An anion does not cross a lipid membrane by passive diffusion. That does not make acids useless — plenty of marketed drugs are acids, and many reach their targets on transporters — but it does mean an acid is a deliberate choice with a cost, never a free addition.

It also means an acid is excellent at one thing: making a salt bridge to an arginine or a lysine in a binding site. Ibuprofen's acid finds an arginine, and you will watch it do so in module 16.

Where an ester would be cleared in minutes, an amide survives. Procaine to procainamide is the classic swap: the same molecule with an ester replaced by an amide, and a duration of action changed from minutes to hours.

A phenol is rapidly glucuronidated or sulfated by the liver. That is why so many phenol-containing drugs have poor oral bioavailability, and why a medicinal chemist looking at a phenol immediately asks whether it is doing essential binding work or whether it can be capped or replaced.

Atorvastatin is a good test: it has four of the groups you have just tabulated, and each is there for a reason.