🎬 Lecture 4 – Video 1

🧬 Protein Structure Determination by NMR (Part 2 – Overview)

This lecture introduces the second part of protein NMR spectroscopy, focusing on how we determine protein structures using NMR — and how it compares to X-ray crystallography .

Let’s break everything down clearly and systematically.

🧩 From Resonance Assignment to Structure

You’ve already learned how to:

1. Prepare the sample
2. Assign resonances

Now we move into the core structural part:

➡️ Collect structural constraints ➡️ Calculate the structure ➡️ Validate and present it properly

This lecture focuses mainly on:

🔎 What structural information can we extract from NMR? 🧮 How do we use it to calculate structures?

🧠 What Structural Information Does NMR Provide?

There are three main categories of structural constraints:

1️⃣ Distance Information (The Most Important!)

Distance is the primary source of structural information in NMR .

📏 Where do distances come from?

• NOEs (Nuclear Overhauser Effects)
• Hydrogen bonds
• Paramagnetic relaxation enhancement (PRE)

All of these give distances between atoms.

Why distances are powerful:

• Short distances (close in sequence) → 🧬 Secondary structure
• Long distances (far apart in sequence) → 🏗️ Tertiary and quaternary structure

So distances give you:

✔ Local information ✔ Global information

They are the backbone of structure determination.

2️⃣ Dihedral Angle Information

We also obtain:

• Scalar couplings
• Chemical shifts

These provide information about:

➡️ Backbone dihedral angles (φ and ψ)

This mainly tells us about:

🧬 Secondary structure

Think back to the Ramachandran plot:

• α-helices → specific φ/ψ values
• β-sheets → different φ/ψ values

But important:

⚠ Dihedral angles alone cannot define the overall fold.

They provide local information only.

3️⃣ Orientation Information

This comes from:

🧭 Residual Dipolar Couplings (RDCs)

RDCs give:

➡️ Orientation of bond vectors relative to the laboratory frame

This is powerful because:

• All vectors are measured in the same lab coordinate system
• This gives indirect global structural information

So RDCs give:

🌍 Global information 🏗️ Tertiary / quaternary structure insight

🧩 Structure Determination = Solving a Puzzle

The lecturer compares structure determination to solving a puzzle :

You must collect:

• 🧩 Many distances
• 🧩 Many dihedral angles
• 🧩 Orientation constraints

Then assemble them computationally.

But there are problems:

❌ Missing pieces

• Signal overlap
• Invisible signals
• Incomplete data

❌ Extra pieces

• Impurities
• Signals that don’t belong to your protein

So structure calculation is an imperfect, constraint-based optimization problem.

📚 What Will the Next Lectures Cover?

The series is divided into:

1️⃣ Distances (NOEs, etc.) 2️⃣ Dihedral angles & orientations 3️⃣ Structure calculation + validation

⚖ NMR vs X-ray Crystallography

Now the lecture compares the two major structure techniques.

🧪 Working Conditions

It’s often argued solution is more physiological.

However:

• NMR samples are highly concentrated
• Structures determined by both methods are usually very similar

So in practice, structures do not differ dramatically.

✅ Advantages of NMR

• No need for crystals
• Can study proteins that refuse to crystallize
• Easy to:
  • Add ligands
  • Change pH
  • Change temperature
  • Perform titrations
  • Measure binding constants
• Can study:
  • Folding
  • Dynamics
  • pKa values
  • Flexibility
  • Catalytic mechanisms

NMR gives more biophysical richness.

❌ Disadvantages of NMR

🚫 Size limitation

Approx. 30–40 kDa upper limit

Most NMR structures are:

• < 16 kDa
• Few above 25 kDa

Meanwhile:

• X-ray handles larger proteins
• Cryo-EM handles very large complexes

⏳ Time-consuming

• Data recording
• Data evaluation
• Structure calculation

🧪 Requires isotope labeling

NMR:

• Requires stable isotopes (¹³C, ¹⁵N)

X-ray:

• Natural abundance protein is fine

💰 Expensive instrumentation

Both are expensive, but:

• NMR labs often own spectrometers
• X-ray users typically book synchrotron beamtime

📊 Statistics (PDB Distribution)

At the time referenced:

• 87% X-ray
• 12% NMR
• 1% Cryo-EM

Cryo-EM is expected to grow.

NMR structures:

• Mostly proteins
• Some nucleic acids

🔬 Workflow Comparison

🧬 NMR Workflow

1. Express & purify protein
2. Isotope labeling
3. Sample optimization
4. Record data
5. Evaluate data
6. Structure calculation

Critical time-consuming step: ➡ Data acquisition & evaluation

Once sample is good → you're mostly safe.

💎 X-ray Workflow

1. Express & purify protein
2. Crystallization
3. Heavy atom derivative (for phasing)
4. Data collection
5. Structure calculation

Critical bottleneck: ➡ Crystallization

Once you have a good crystal → you're safe.

🧠 When Should You Choose NMR?

Use NMR when:

• You cannot crystallize the protein
• You want more than just structure:
  • Ligand binding
  • Folding pathways
  • Dynamics
  • pKa determination
  • Mechanistic insight
• Or if you are fascinated by NMR 😉

🎯 Core Takeaways

🏗️ Structural Constraints in NMR

🧩 Structure Determination = Constraint Optimization

• Collect incomplete experimental constraints
• Solve computationally
• Validate carefully

⚖ NMR vs X-ray

🔜 What Comes Next?

The next three lectures will cover:

1️⃣ Distance constraints in depth 2️⃣ Dihedral angle & orientation constraints 3️⃣ Structure calculation and validation