🧬 Lecture Summary — Solving the Phase Problem with Heavy Atoms (SIR & MIR)

🌊 1. Fourier Transform — The Bridge Between Diffraction Data and Structure

A core idea in crystallography is that we can move back and forth between diffraction space and real space using the Fourier transform.

What do we measure experimentally?

From the diffraction experiment we obtain:

• Many reflections labeled by h, k, l
• Each reflection has:
  • Intensity (I) → related to structure factor amplitude
  • Phase angle (α) → not measured directly

Since intensity ∝ |F|², we can calculate the structure factor amplitude (|F|) from the detector signal.

🧠 Key Concept

• If we know amplitudes + phases → we can reconstruct electron density
• If we know electron density → we can Fourier transform back → obtain reflections

This means:

🔁 Diffraction pattern ↔ Electron density map No information is lost if both amplitude and phase are known.

This is why crystallography is powerful — the structure is encoded in the diffraction data.

❗ 2. The Phase Problem — The Central Challenge

The experiment only gives intensities, not phases.

But phases are extremely important.

Famous demonstration

If you combine:

• Amplitudes from one image
• Phases from another

👉 The resulting image looks like the one providing the phases.

This shows:

🎯 Phases determine the final electron density much more strongly than amplitudes.

Without phases → no structure.

📜 3. Historical Insight — Proteins Were Once Thought to Be “Glue”

Before protein crystallography:

• Scientists thought proteins were random aggregates.
• Then hemoglobin crystals were shown to diffract X-rays.
• This proved proteins are highly ordered structures.

Max Perutz later worked ~15 years and discovered:

💡 Introducing heavy atoms could help solve the phase problem.

This became a revolutionary idea.

⚡ 4. Why Heavy Atoms Help

Protein atoms (C, N, O, H, S) scatter weakly.

Also:

• ~99% of X-rays pass straight through crystal
• Diffraction spots are weak
• Interference cancels much scattering

But heavy atoms:

• Have many electrons
• Scatter strongly
• Can constructively or destructively interfere

Result:

👉 Heavy atoms change reflection intensities measurably.

This change contains phase information.

➕ 5. Structure Factor Vector Addition (Argand Diagram)

We describe scattering as vectors:

• Fp → protein alone
• Fh → heavy atom contribution
• FpH → combined scattering

Vector relation:

F_ = F_p + F_h

This geometric representation is essential for solving phases.

🧪 6. Native vs Derivative Dataset

• Native dataset → protein crystal alone
• Derivative dataset → heavy atom soaked into crystal

Comparing diffraction images shows:

• Some reflections increase
• Some decrease
• Some disappear

These measurable differences allow phase determination.

📐 7. Harker Construction — Phase Determination by Geometry

Used in Single Isomorphous Replacement (SIR).

Steps:

1. Draw circle with radius = |Fp|
2. Draw vector −Fh (heavy atom known)
3. Draw second circle centered at tip of −Fh with radius = |FpH|
4. Circle intersections = possible protein phases

Result:

🎯 Two possible phase solutions → Phase ambiguity

This happens for every reflection → enormous combinatorial uncertainty.

⭐ 8. “Best Phase” Strategy

Instead of choosing one solution:

• Take average phase between the two
• Scale amplitude based on phase uncertainty

Interpretation:

• If solutions are close → high confidence → strong weight
• If far apart → low confidence → weak weight

Surprisingly:

😄 This approximation works well enough to build electron density maps.

🧩 9. Multiple Isomorphous Replacement (MIR) — Removing Ambiguity

Use multiple heavy atom derivatives.

Each derivative gives:

• New Harker circles
• New phase intersection candidates

Only one intersection is common → correct phase.

Thus:

✅ Using ≥2 derivatives can resolve phase ambiguity.

In practice:

• Often need several derivatives
• Heavy atoms must bind at different sites

⚙️ 10. How to Introduce Heavy Atoms

Many experimental strategies exist:

🧬 Intrinsic labeling

• Sulfur in cysteine/methionine (weak signal)
• Selenomethionine substitution (very common)

🔄 Ion substitution

• Replace Ca²⁺ with lanthanides
• Replace Zn²⁺ with Hg²⁺

🧪 Chemical soaking

• Mercurials bind cysteines covalently
• Platinum, gold coordinate surface sites
• Uranyl / lanthanides bind electrostatically

💨 Gas pressurization

• Xenon / krypton insertion

🧷 Heavy-atom ligands

• Brominated ATP
• Iodinated peptides

These approaches aim to give:

• Known heavy atom position
• Strong scattering signal
• Minimal crystal damage

🧮 11. Final Summary — Workflow for Solving the Phase Problem

To determine protein structure:

1. Measure diffraction intensities → get |F|
2. Collect native + derivative datasets
3. Determine heavy atom positions
4. Calculate heavy atom scattering
5. Use Harker construction → estimate phases
6. Use MIR → resolve ambiguity
7. Fourier transform → obtain electron density map

Remaining challenge becomes:

• Refining heavy atom parameters:
  • Position (x,y,z)
  • Occupancy
  • Temperature factor

Once phases are known → structure solution becomes feasible.

🧠 Big Picture Understanding

This lecture explains how crystallographers historically solved the biggest obstacle in X-ray crystallography.

The key ideas:

• Diffraction gives amplitudes, not phases
• Heavy atoms perturb intensities
• Geometric vector methods recover phase information
• Multiple derivatives increase accuracy

This strategy enabled the first protein structures (like hemoglobin) and remains foundational for modern phasing methods.