Mutations, Hot Spots and Alanine Scanning

Residues are essential for drug binding

Lecture 4 — From Binding Site to Design Target

"If you removed the side chain of a hydrophobic residue lining the binding site, what happens to the binding energy?"

Mutation Types

Level	What changes	Example
DNA level	One nucleotide base is substituted	C → T in codon
Protein level	One amino acid is replaced by another	Serine → Glycine at position 130

Synonymous (silent)

DNA changes but the amino acid stays the same — due to codon redundancy. No effect on the protein. Invisible to Rosetta.

Non-synonymous (missense) ← what we study

DNA change results in a different amino acid. May or may not affect function depending on where the change is and what the new amino acid is.

Nonsense

DNA change creates a stop codon — the protein is truncated. Usually catastrophic for function.

Wild type vs mutant

Wild type

The original, naturally occurring protein sequence — the version found in nature before any change. In our simulations, wild type = the protein as it appears in the PDB structure.

Mutant

Any version where one or more amino acids have been changed. In drug resistance, the mutant is what the bacterium evolves. In protein engineering, it is what the scientist designs.

                    Mutation naming convention:  V102A means Valine (V) at position 102 → Alanine (A).

                    Format:  [original AA][position][new AA]

Notation	Meaning	Context
S130G	Serine at 130 → Glycine	Resistance mutation in β-Lactamase (Project 6)
V102A	Valine at 102 → Alanine	Alanine scan probe (Project 5)
K73R	Lysine at 73 → Arginine	Resistance mutation in β-Lactamase (Project 6)

One amino acid change can make a drug fail

1. Removes a key contact

If the mutated residue was forming a hydrogen bond or van der Waals contact with the drug, removing it weakens binding directly. The drug still enters the pocket but holds on less tightly.

2. Changes the pocket shape

Some substitutions alter the binding site geometry — making it slightly larger, smaller, or differently shaped. The drug may no longer fit as well even if it still enters.

3. Changes the pocket chemistry

Replacing a charged residue (Arginine) with a neutral one (Serine) removes an electrostatic interaction. A drug designed to interact with that charge loses a significant portion of its binding energy.

Real example (Project 6): β-Lactamase S130G — Serine 130 normally forms a hydrogen bond with clavulanic acid. Mutating to Glycine removes that bond. The inhibitor fails. One letter change in a 263-residue protein renders a drug clinically ineffective.

Why Alanine? The logic of the probe

Property of Alanine	Why it matters
Smallest side chain with a carbon (-CH₃)	Removes original chemistry without removing the backbone
Non-polar and chemically inert	Does not introduce new interactions — truly erases the original contribution
Does not disrupt the backbone	Unlike Glycine, maintains correct backbone geometry
Naturally present in all proteins	The protein already knows how to fold with Alanine — minimally disruptive

                    Why not Glycine? Glycine has no side chain at all, but its unusual backbone flexibility can distort local structure — making it hard to tell whether any binding change is due to losing the side chain or due to backbone distortion. Alanine avoids this problem entirely.
                

Mutating Valine → Alanine trims a four-carbon branching arm down to a single methyl group. The backbone is untouched. But all van der Waals contacts that Valine was making with the drug — gone.

ΔΔG measures the energetic cost of each mutation

                    Step 1: ΔG_wt = binding energy of wild-type protein + drug

                    Step 2: ΔG_mut = binding energy of mutant protein + drug

                    Step 3: ΔΔG = ΔG_mut − ΔG_wt

ΔΔG value	Meaning	Classification
> +2.0 REU	Mutation greatly weakens binding	Hot spot — critical residue
+1.0 to +2.0 REU	Mutation moderately weakens binding	Moderate contributor
−0.5 to +1.0 REU	Mutation has little effect	Dispensable residue
< −0.5 REU	Mutation strengthens binding	Rare (removing side chain helps)

                    Example: Wild-type binding = −27 REU. After mutating Phe381 → Ala, binding = −19 REU.

                    ΔΔG = −19 − (−27) = +8 REU → Phe381 is a hot spot. The drug was heavily relying on that phenylalanine ring.

Hot spots and dispensable residues

Hot spot — ΔΔG ≥ +2.0 REU

The drug depends on this residue heavily. Mutating it away causes a large loss in binding affinity. These are your primary design targets.

Moderate contributor — ΔΔG +1.0 to +2.0 REU

The residue contributes meaningfully but is not essential. Worth engaging but not as critical as hot spots.

Dispensable — ΔΔG < +1.0 REU

Present in the binding site but contributes little to binding. The drug barely notices if it is mutated away.

                    The 80/20 rule of binding: In a typical binding site of 12–15 residues, only 2–4 are true hot spots. Roughly 80% of the binding energy comes from 20% of the contact residues. Identifying that 20% is the whole point of alanine scanning.
                

SAR connects structure to activity

Structure-Activity Relationship (SAR) is the systematic study of how changes in a drug's molecular structure affect its biological activity. It is the central framework of medicinal chemistry.

Alanine scan identifies hot spot residues

↓

Hot spot chemical class tells you what interaction to target

↓

Drug is modified to maximise that interaction

Hot spot residue	Chemical class	SAR implication
Phenylalanine (Phe)	Hydrophobic aromatic	Add aromatic ring to drug for π-π stacking
Arginine (Arg)	Positively charged	Add negatively charged group to drug
Serine (Ser)	Polar, H-bond donor	Add H-bond acceptor to drug (carbonyl, ether)

Alanine scanning in the real world

Experimental scanning

Physically synthesise each mutant protein, purify it, and measure binding affinity using SPR or ITC. Each mutant takes days to weeks and costs £5,000–£20,000.

Computational scanning

Each mutation takes seconds. An entire binding site of 15 residues scanned in minutes. An approximation — but accurate enough to prioritise which mutants to test in the lab.

Computational scan — all 15 residues (hours)

↓

Identify top 3–4 predicted hot spots

↓

Synthesise and test only those mutants in the lab

↓

Confirm or refine → use hot spots to guide drug optimisation

Alanine scanning appears three times in your pipeline

Project	What is scanned	What is measured	Question answered
Project 1	Any residue in Lysozyme	Change in protein stability	Which residues keep the protein folded?
Project 5	Binding site residues in COX-2	Change in drug binding (ΔΔG)	Which residues are hot spots for Ibuprofen?
Project 6	Known resistance positions in β-Lactamase	Change in inhibitor binding (ΔΔG)	How much does each clinical mutation weaken the drug?

                    Project 5 vs Project 6: In Project 5 we always mutate to Alanine — a neutral probe asking "what could hurt?" In Project 6 we mutate to the actual amino acid bacteria have evolved — asking "what has actually hurt in patients?"
                

Summary & Preparation

A point mutation changes one amino acid — it can strengthen, weaken, or have no effect on drug binding depending on which residue is affected.
Wild type is the natural protein sequence; mutant is any version with at least one amino acid change.
Alanine is used as the probe because it removes the side chain without disrupting the backbone — it truly erases one residue's contribution.
ΔΔG = ΔG_mut − ΔG_wt — positive ΔΔG means the mutation weakened binding; that residue was contributing.
Hot spots are residues with ΔΔG ≥ +2.0 REU — the drug depends on these most heavily.
SAR connects hot spot identity to drug modification decisions — hot spots tell you where to improve the drug.

                    Preparation for Lecture 5:

                    "If a bacterium mutates one of the hot spot residues your drug relies on heavily, what happens to the drug's effectiveness? And what would you do?"