Structural biology Study Guide

📖 Core Concepts Structural biology – study of 3‑D shapes of macromolecules (proteins, nucleic acids, membranes) and how shape determines function. Primary structure – linear sequence of amino acids (proteins) or nucleotides (DNA/RNA). Secondary structure – local regular motifs (α‑helix, β‑sheet). Tertiary structure – overall 3‑D fold of a single macromolecule. Quaternary structure – spatial arrangement of multiple subunits into a functional complex. Experimental techniques – X‑ray crystallography, NMR spectroscopy, cryogenic EM, electron crystallography, complementary physical methods (mass spec, light scattering). Computational methods – molecular dynamics (MD), deep‑learning protein prediction (AlphaFold), hybrid modeling, bioinformatics motif analysis. --- 📌 Must Remember X‑ray crystallography → atomic‑resolution structures; requires well‑ordered crystals. NMR → works in solution/solid‑state; provides dynamics; limited to smaller proteins (<30 kDa for solution NMR). Cryo‑EM → single‑particle imaging of flash‑frozen samples; near‑atomic resolution for large complexes (>150 kDa). AlphaFold → deep‑learning predicts 3‑D coordinates directly from sequence with high accuracy (2020s). Macromolecular function depends on correct fold; mis‑folding → loss of function or disease. Light‑scattering (SLS/SEC‑MALS) → measures size, shape, aggregation in solution, not atomic detail. --- 🔄 Key Processes X‑ray Crystallography Workflow Purify macromolecule → grow crystals. Mount crystal; expose to X‑ray beam. Record diffraction pattern (spots = Bragg reflections). Apply Fourier transform to convert intensities → electron density map. Build atomic model into map; refine against data. Cryo‑EM Single‑Particle Pipeline Flash‑freeze purified sample on EM grid (vitreous ice). Collect thousands of 2‑D projection images with electron beam. Align and classify particles → 2‑D class averages. Reconstruct 3‑D density map via iterative algorithms. Fit atomic model; refine. Molecular Dynamics Simulation Steps Prepare structure (add hydrogens, solvate, neutralize). Choose force field; assign parameters. Energy‑minimize to relieve bad contacts. Equilibrate (NVT → NPT ensembles). Run production trajectory; analyze motions, RMSD, interactions. AlphaFold Prediction Procedure Input amino‑acid sequence. Retrieve multiple‑sequence alignment (MSA) & templates. Deep‑learning model predicts inter‑residue distances & angles. Build 3‑D coordinates; output confidence scores (pLDDT). --- 🔍 Key Comparisons X‑ray Crystallography vs Cryo‑EM Crystals required vs vitrified particles. Typically higher resolution for small proteins (≤ 3 Å) vs Cryo‑EM excels for large complexes. Solution NMR vs Solid‑state NMR Solution NMR → high‑resolution for small soluble proteins, provides dynamics. Solid‑state NMR → works for membrane proteins, fibrils, and larger assemblies. Primary vs Secondary vs Tertiary vs Quaternary Primary = sequence; Secondary = local patterns; Tertiary = full 3‑D fold; Quaternary = assembly of multiple subunits. AlphaFold vs Experimental Determination AlphaFold: prediction only, no direct experimental validation; confidence varies. Experimental: provides empirical electron density / spectra; validates actual conformation. --- ⚠️ Common Misunderstandings “Primary structure determines function directly.” Function emerges from the folded (tertiary/quaternary) structure, not the linear sequence alone. “AlphaFold gives a perfect structure for any protein.” Prediction confidence can be low for disordered regions, novel folds, or complexes. “Cryo‑EM always yields atomic resolution.” Resolution depends on particle size, symmetry, data quality; many maps are 3–5 Å. “Mass spectrometry provides atomic coordinates.” MS gives mass and subunit composition, not spatial arrangement. --- 🧠 Mental Models / Intuition “Fit‑the‑Puzzle” – Think of each technique as a puzzle piece: X‑ray gives a high‑resolution picture of a static crystal; NMR shows pieces moving in solution; Cryo‑EM reveals the whole assembled machine in near‑native state. “Sequence → Fold → Function” – Visualize a protein as a string (sequence) that folds (tertiary/quaternary) into a functional shape; perturb any step and the downstream effect changes. --- 🚩 Exceptions & Edge Cases Membrane proteins – Often resist crystal growth; electron crystallography or Cryo‑EM (single particles) are preferred. Very large complexes (> 1 MDa) – Cryo‑EM is usually the only feasible high‑resolution method. Intrinsically disordered regions – Poorly resolved by X‑ray/NMR; AlphaFold confidence low; may require complementary biophysical probes (SAXS, CD). --- 📍 When to Use Which | Situation | Best Technique(s) | Reason | |-----------|-------------------|--------| | Small soluble protein (< 30 kDa) with crystals | X‑ray crystallography or solution NMR | Crystallization often feasible; NMR gives dynamics. | | Large multi‑subunit complex (> 150 kDa) | Cryo‑EM | No need for crystals; works with heterogeneous samples. | | Membrane protein, microcrystals only | Electron crystallography / Micro‑ED | Electron beam penetrates tiny crystals; complements X‑ray. | | Need rapid structural insight, no experimental data | AlphaFold prediction | Fast, high‑confidence for many folds; verify later. | | Study overall shape/aggregation in solution | Light scattering (SLS/MALS) or SAXS | Provides size/shape without atomic detail. | | Identify secondary‑structure content | Circular dichroism (CD) | Measures differential absorption of circularly polarized light. | --- 👀 Patterns to Recognize Diffraction spots forming concentric rings → well‑ordered crystal, high resolution potential. Broad, weak NMR peaks → large molecular weight or aggregation. Cryo‑EM 2‑D class averages showing distinct views → good particle orientation distribution, higher chance of high‑resolution map. High pLDDT scores (> 90) in AlphaFold → reliable region; low scores → flexible/disordered. Linear increase in scattering intensity with angle → larger, elongated particles (SAXS pattern). --- 🗂️ Exam Traps Confusing “primary structure” with “primary function.” – Remember primary structure is just the sequence. Choosing Cryo‑EM for a 10 kDa protein – Typically not feasible; X‑ray or NMR are better. Assuming a high‑confidence AlphaFold model guarantees the biologically relevant conformation. – It may miss ligand‑induced changes or alternative states. Interpreting a mass‑spectrometry peak as a full‑length protein. – Peaks reflect mass, not whether the protein is folded or part of a complex. Reading a CD spectrum and concluding exact secondary‑structure percentages. – CD gives approximate content; confirm with high‑resolution methods. ---