The Binding Site and Protein-Ligand Docking

How does a drug find its target?

Lecture 3 — Shape, Chemistry and Binding Energy

"Ibuprofen is non-polar and hydrophobic. The COX-2 binding site is also largely hydrophobic. Why does this chemical match matter? What would happen if you put a highly charged, polar drug into a hydrophobic pocket?"

The binding site is where the drug physically docks

A binding site is a cavity or groove on the protein surface where a small molecule settles and interacts with surrounding amino acids.

The drug does not interact with the whole protein — only with the 10–20 amino acids that line this pocket.

Three things must match:
1. Shape — the drug must fit
2. Size — not too large, not too small
3. Chemistry — matching polarity and charge

Connect to Project 4: You will identify every amino acid within 5 Å of Ibuprofen inside COX-2 — these residues form the binding site.

Rule 1 — The drug must fit the shape of the pocket

Situation	Result
Drug too large	Cannot enter the pocket — steric clash
Drug too small	Fits but makes few contacts — weak binding
Wrong shape, right size	Some contacts made but poor fit — moderate binding
Matches pocket shape precisely	Maximum contacts — strong binding ✓

                    Induced fit: A real binding site is not a rigid lock — the protein can flex slightly to accommodate a drug that is close to the right shape.
                

Concrete numbers: The COX-2 binding channel is 6–8 Å wide. Ibuprofen is ~8 Å across its longest axis — a near-perfect fit. This geometric match is part of why it binds strongly.

Rule 2 — Drug VS Pocket matched chemistry

Binding site character	Key residues	Drug should have
Hydrophobic pocket	Phe, Val, Ile, Leu, Met	Non-polar groups, aromatic rings
Polar region	Ser, Thr, Asn, Gln	H-bond donors or acceptors (-OH, -NH₂)
Positively charged region	Arg, Lys, His	Negatively charged group on drug
Negatively charged region	Asp, Glu	Positively charged group on drug

                    Analogy: 
 Shape complementarity = finding a key that fits the lock. 
Chemical complementarity = making sure the key is made of the right material.
 A key made of butter might fit the lock perfectly in shape — but it will fail to turn it.
                

COX-2 — a real binding site example

COX-2 converts arachidonic acid into prostaglandins (pain and inflammation). Blocking its active site prevents this.

Drug	Brand name	Character
Ibuprofen	Advil	Small, non-polar, flexible
Celecoxib	Celebrex	Larger, selective COX-2 inhibitor, sulfonamide group
Meloxicam	Mobic	Intermediate size, thiazine ring system

                    All three are approved COX-2 inhibitors — they all bind and they all work. But they bind with different strengths, at different positions in the channel, making different contacts. Project 3 will quantify those differences.
                

Protein-ligand docking:
predicting how a drug fits

                    Docking predicts where a drug sits in the binding site (pose) and how tightly it binds (binding energy).
                

What we do in Projects 3 & 4

The drug is already inside the protein — co-crystallised in the PDB structure. We are scoring a known, experimentally confirmed pose, not predicting a new one. Simpler and more reliable.

Real-world power

Screen thousands of candidates digitally in hours. Only synthesise and test the top scorers in the lab. Reduces the >90% failure rate of untargeted drug development.

Binding energy —
the number that ranks drugs

Binding Energy =
Complex Score − (Protein Score + Ligand Score)

Component	Score
Complex (COX-2 + Ibuprofen)	−1,820 REU
Protein alone (COX-2)	−1,768 REU
Ibuprofen alone	−25 REU
Binding energy	−1,820 − (−1,768 + −25) = −27 REU

                    That −27 REU is the energy gained by bringing drug and protein together — all the van der Waals contacts, hydrogen bonds, and hydrophobic interactions that form when Ibuprofen settles into the COX-2 pocket.
                

Interpreting binding energy values

Negative (e.g. −27 REU) — Drug genuinely binds Energy is released when the complex forms. The more negative, the tighter the binding. This is what you want for a drug candidate.

Near zero (e.g. −2 REU) — Drug barely binds Almost no energy is gained from the interaction. Unlikely to be therapeutically useful at normal doses.

Positive (e.g. +8 REU) — Drug does not bind Binding is energetically unfavourable — the complex is less stable than the separated components.

Important: Binding energy is one factor — not the only factor. Bioavailability, selectivity, and chemical stability also matter. Docking tells you about binding; the rest requires lab and clinical testing.

Structure quality affects every docking result

If the binding site geometry in your crystal structure is wrong by even 0.3 Å, your calculated binding energy will be wrong too.

Source of error	Effect on docking
Low resolution (> 2.5 Å)	Side chain positions uncertain — binding site shape approximate
Missing residues	Parts of the binding site absent — contacts not calculated
Crystal contacts	Packing forces in the crystal may distort the binding site slightly

                    Why Project 2 comes before Project 3: In Project 2 you rank five Lysozyme structures by quality. In real drug discovery, choosing the highest-quality starting structure always happens before any docking. 
Garbage in, garbage out.
                

The contact shell: which residues touch the drug?

                    A contact residue = any amino acid with at least one atom within 5.0 Å of any atom of the drug.
                

Distance	Interaction type
< 2.0 Å	Clash — atoms overlapping (problem)
2.0–3.5 Å	Covalent bond or strong hydrogen bond
3.5–5.0 Å	Van der Waals contact — atoms touching ✓
5.0–8.0 Å	Close neighbourhood — weak or indirect
> 8.0 Å	Too far to interact meaningfully

A binding site may have 200 residues nearby, but only 12–15 are actually touching the drug. Those are the ones that matter for binding energy, drug design, and resistance mutations — and the target of Project 4.

Summary & Preparation

A binding site is a cavity on the protein surface — shape and chemistry both determine whether binding occurs.
Shape complementarity — the drug must physically fit the pocket geometry.
Chemical complementarity — hydrophobic drug regions face hydrophobic residues; polar faces polar; charged faces oppositely charged.
Binding energy = Complex score − (Protein score + Ligand score). More negative = tighter binding.
Structure quality matters — a poor crystal structure gives unreliable docking results.
Contact residues are within 5.0 Å of the drug — typically 12–15 residues form the functional binding site.

                    Preparation for Lecture 4:

                    "In Project 5 we will mutate every binding site residue to Alanine and measure how much each mutation weakens drug binding. Think about this: if you removed the side chain of a hydrophobic residue lining the binding site, what would happen to the drug's binding energy? Why?"