Molecular Forces and Protein Folding — fun, detailed walkthrough (skipping the boxed “Special topic” sections)

1) Big picture: what proteins are and why folding matters 🧬

• Proteins are unbranched polymers of L-amino acids whose sequence is encoded by genes. Each amino acid contributes a repeating backbone (–NH–CαH–CO–) plus a side chain (size/shape/charge varies).
• Despite being “just a chain,” proteins often adopt a unique, stable 3D structure because internal interactions (within the chain, plus with water) strongly bias certain conformations.
• Many proteins contain secondary structure (α-helices, β-sheets). Some proteins do not have a single stable fold, ranging from equilibria between conformations to fully unstructured states.

2) Covalent modifications: permanent-ish structural changes 🔗

Beyond the normal amino acids + peptide bonds, proteins often include extra covalent features:

• Disulfide bonds (Cys–Cys): common within a chain or connecting chains (insulin, immunoglobulins). They can strongly constrain folding and stability.
• Phosphorylation (Ser/Thr/Tyr): reversible covalent addition that changes electronic structure and supports signaling cascades.
• There are hundreds of known covalent modifications; each can create new functional properties.

3) Electrostatic forces: charges, dipoles, and “screening” ⚡

Core rule: electrostatic interaction energy depends on charges, distance, and the dielectric constant of the medium (water screens strongly).

Key ideas:

• Ion–ion interactions are long-range compared to other noncovalent forces.
• Even without net charge, molecules can interact via:
  • Ion–dipole
  • Dipole–dipole
  • Induced dipole / dispersion interactions (requires polarizability)
• Water’s dielectric constant (~80) strongly reduces charge–charge interactions compared to vacuum, but fixed ion pairs (e.g., Lys–Asp) can still stabilize proteins—even on the surface.

How electrostatics can unfold proteins:

• Low ionic strength → less shielding by counter-ions → charge interactions become stronger → repulsion can destabilize folding and promote denaturation/aggregation.
• Extreme pH changes protonation states:
  • Low pH protonates carboxylates → breaks salt bridges and can increase net positive charge → repulsion → unfolding.
  • High pH deprotonates amino groups → net negative charge rises → repulsion → unfolding.

4) Van der Waals interactions: “weak but everywhere” 🤝

• These are collected non-ionic electrostatic interactions often described by a Lennard–Jones potential: attractive at moderate distances, strongly repulsive when atoms get too close (orbital overlap).
• Important point: vdW forces are short-range but become powerful in a well-packed protein core because there are so many contacts.

5) Hydrogen bonding: directional glue (but context matters) 🧷

• A hydrogen bond forms when an electronegative atom approaches a hydrogen that’s already covalently bound to an electronegative atom. It’s shorter than a simple electrostatic interaction.
• Peptide bonds have strong dipoles (O partially negative, N partially positive), supporting hydrogen bonding patterns and secondary structure.
• Two key experimental facts:
  1. Hydrogen bonds are abundant in native proteins (networks; most peptide bonds are hydrogen bonded).
  2. Breaking internal H-bonds without water is very unfavorable (large stability in dry environments like seeds/spores).

Important nuance: In water, the “benefit” of an internal H-bond competes with H-bonding to water, so isolating its net contribution is difficult.

Water as a “denaturant-ish” solvent:

• Water is unusual: open hydrogen-bonded structure and easily perturbed by solutes (Hofmeister effects are mentioned as relevant elsewhere in the course).

6) The hydrophobic effect: folding’s main driver 🌊➡️🫧

• Nonpolar groups (hydrocarbon side chains) interact poorly with water, so they tend to be buried in the protein interior—like oil separating from water.
• A hydrophobicity scale ranks residues by their tendency to prefer nonpolar environments (e.g., Trp/Ile/Phe/Leu high; Lys/Arg very low).

Origin (in this chapter’s framing):

• Water forms dynamic H-bond networks that are favorable while remaining disordered.
• Near hydrophobic surfaces, water molecules become more ordered to maintain H-bonds, which is entropically unfavorable.
• Folding reduces exposed hydrophobic area → fewer ordered water molecules → entropy increases → favorable free energy change.

7) Thermodynamics of nonpolar interactions: ΔG, ΔH, ΔS in water 📉

When putting a hydrophobe into water:

• Entropy decreases (water ordering) → unfavorable
• Enthalpy decreases somewhat (favorable water–solute interactions) → partially compensatory
• Net ΔG > 0, so solubility is low (though not zero).

When two hydrophobes associate:

• Exposed hydrophobic surface area decreases → fewer ordered waters → ΔS increases
• Fewer water–surface favorable contacts → ΔH increases
• Net ΔG decreases, so association is favored.

8) Motions in proteins: dynamics contribute stability 🕺

Proteins move across huge timescales (ps to years): vibrations, side-chain rotations, domain motions, allosteric transitions, folding/unfolding, complex dissociation, etc. Thermodynamic punchline in this text:

• More dynamics → higher entropy → lower free energy → can stabilize (flexible loops can contribute noticeably).

9) The protein folding problem: Levinthal paradox 🧩

Levinthal’s argument: random search across conformations is impossibly slow.

• Even with only 2 conformations per residue, a 100-residue chain has 2^100 possibilities (~10^30).
• At ~10^-13 s per conformational change, exploring everything would take ~10^9 years.
• Yet many proteins fold in seconds/minutes → folding is not a blind search; it uses efficient routes.

10) Folded ⇌ unfolded transitions: two-state unfolding curves 📈

Experimentally, unfolding can be driven by:

• Chemical denaturants (urea, guanidinium chloride)
• Temperature
• pH changes

Many proteins show sigmoidal “melting/unfolding curves,” often interpreted as a two-state equilibrium:

• unfolded ⇌ folded Within the transition region, you can estimate fractions of folded/unfolded from spectroscopic signals (CD, fluorescence, etc.).

Kinetics connection (two-state case):

• Folding/unfolding rates: kf and ku
• Equilibrium constant: K = folded/unfolded = kf/ku (often observed; supports two-state).

Complications to two-state behavior mentioned:

• Disulfide bond formation (especially wrong pairings)
• Proline cis/trans isomerization
• Proteolytic processing producing multiple disulfide-linked chains (often not reversible).

11) Thermodynamics of folding: ΔG = ΔH − TΔS ⚖️

Define folding free energy:

• ΔG°folding = G°folded − G°unfolded Stable folding means ΔG°folding is negative.

Magnitude reality-check:

• Typical ΔG°folding is only about –20 to –60 kJ/mol, comparable to “only a few hydrogen bonds,” meaning stability is a small difference between large opposing terms.

Thermodynamic message:

• Folding stability is a delicate balance:
  • Favorable enthalpy (many interactions in the folded state)
  • Unfavorable entropy (loss of chain conformational freedom)
  • Plus big solvent contributions (especially hydrophobic effect).

12) Calorimetry and DSC: watching unfolding via heat capacity 🔥

Differential scanning calorimetry (DSC) measures heat capacity CP as temperature changes.

• Unfolding shows up as a peak (large heat absorption).
• Tm is where CP is maximal (midpoint of unfolding).
• The baseline difference between folded and unfolded CP gives ΔCP.

Interpretations emphasized:

• Unfolded proteins have larger CP partly because exposing hydrophobics increases water ordering effects.
• Integrating the peak area gives ΔHcal (calorimetric enthalpy).
• Comparing ΔHcal with ΔHvH (van’t Hoff enthalpy from equilibrium analysis) helps assess:
  • Two-state cooperativity (often ΔHcal ≈ ΔHvH for small single-domain proteins)
  • Multi-domain behavior (domains unfolding independently can change ratios).

13) Energetics breakdown: where enthalpy/entropy come from 🧠

This chapter lays out qualitative contributors:

Unfolded state (in water):

• Many favorable interactions of polar/ionized groups with water (enthalpically favorable)
• High chain conformational freedom (entropy favorable)
• But water around hydrophobics is restricted (entropy unfavorable).

Folded state:

• Many intramolecular favorable interactions (H-bonds, packing, salt bridges)
• Chain entropy decreases (unfavorable)
• Hydrophobic burial releases ordered waters → water entropy increases (favorable).

A key table-like message:

• No single term dominates; folding free energy is a net of many contributions.

16) What DSC instruments actually do (practical view) 🧪

• Two cells: sample (protein solution) and reference (buffer). Heat both at the same rate.
• The instrument measures extra power needed to keep temperatures equal → gives excess heat capacity vs temperature (thermogram).
• Thermogram regions:
  1. pre-transition baseline
  2. unfolding peak
  3. post-transition baseline

Model-free vs model-based:

• ΔHcal: model-free (area under peak, baseline-corrected).
• ΔHvH: model-dependent (assumes a model, often two-state), derived from the temperature dependence of K extracted from the thermogram.
• Pitfalls: aggregation and irreversibility can distort peaks and mislead ΔH estimates.

17) Ligands and stability: Tm shifts with binding 🎯

Core principle (Le Chatelier applied to folding):

• If ligand binds preferentially to the folded/native state, it stabilizes folding → Tm increases.
• If ligand binds preferentially to the unfolded state, it destabilizes folding → Tm decreases.

The chapter writes this with equilibria like:

• N + L ⇌ NL and N ⇌ U, showing ligand binding changes the effective unfolding equilibrium constant and thereby shifts Tm.

18) Denaturants vs osmolytes: opposite effects 🧴🛡️

Denaturants (chaotropes): urea, guanidinium chloride

• At high concentrations (6–10 M) they unfold proteins to random-coil-like states.
• They work differently than SDS (explicitly emphasized).
• Mechanistic idea emphasized here:
  • Denaturants increase solubility of both polar and nonpolar side chains, correlating with accessible surface area.
  • They effectively make it more favorable for protein surfaces—especially the huge buried interior surface—to be solvated → unfolding.

Free-energy picture:

• Denaturant stabilizes both folded and unfolded states, but stabilizes unfolded more (more surface exposed → stronger effect), so ΔG°folding becomes less negative and can cross 0 at a midpoint concentration.
• Often observed: linear dependence of ΔG°folding on denaturant concentration with slope m.

Osmolytes: TMAO, sarcosine, sucrose, proline

• Protect proteins against stress by stabilizing the folded state.
• Mechanistic picture here: osmolytes make peptide–osmolyte interactions less favorable; the unfolded state is penalized more (more exposure), increasing folded stability.

20) Folding/unfolding kinetics and chevron plots ⛰️

Typical experiment design:

• Start unfolded at high denaturant → rapidly dilute to low denaturant → watch refolding (often first-order kinetics).
• For unfolding: mix folded protein into strong denaturant → watch unfolding (also first-order).

For a two-state system:

• Observed rate: kobs = kf + ku In refolding conditions usually ku ≪ kf, so kobs ≈ kf; in unfolding conditions often kf ≪ ku, so kobs ≈ ku.

Empirical linear relationships used:

• log kf = log kf(water) + mfden (mf negative)
• log ku = log ku(water) + muden (mu positive)

Plotting log(rate) vs den gives a V-shape (the “chevron plot”), with intersection at Cm where kf = ku.

21) Mutational studies and the transition state: Φ-values 🧷

Protein engineering (point mutations, insertions/deletions) is used to probe:

• stability changes
• folding/unfolding rate changes
• which residues are essential vs merely stabilizing

Key quantities:

• Transition state free energies for folding/unfolding (relative to unfolded or native).
• The Φ-value measures how much a mutation destabilizes the transition state relative to how much it destabilizes the folded state:
  • Φ ≈ 0: residue not “native-like” in transition state
  • Φ ≈ 1: residue interactions in transition state similar to folded state

This enables mapping of what parts of the structure are already formed at the transition state.

23) What the transition state is like (conceptually) 🧠

• The transition state typically contains native hydrophobic long-range interactions that stabilize weak secondary structure.
• It often resembles a distorted folded structure, with a more “folded-like” nucleus and increasing distortion away from it.
• The chapter stresses ensembles:
  • unfolded: very diverse ensemble
  • folded: tight ensemble
  • transition state: intermediate ensemble, often closer to folded than unfolded

24) Molten globule: compact + secondary structure, but loose core 🫠

A “molten globule” is a partially structured intermediate often seen under mildly denaturing conditions. Core characteristics listed:

1. relatively compact (only ~10–30% larger than native)
2. substantial secondary structure (far-UV CD remains)
3. weak/absent tight tertiary packing (near-UV CD largely disappears)
4. increased flexibility (rapidly interconverting conformations)

25) Folding funnels: many routes down the landscape 🌀

The chapter explains why 1D reaction-coordinate diagrams are misleading for protein folding:

• unfolded state has many conformations; there isn’t a single path
• real energy landscapes have “blind alleys” and nonproductive intermediates

So the modern cartoon is a folding funnel:

• wide at top (many unfolded conformations)
• narrowing as trajectories converge
• key barrier region corresponds to the transition state (“saddle point”)

Quick memory hooks ✅

• ΔG = ΔH − TΔS: folding is a small net difference between big competing terms.
• Hydrophobic effect = water entropy story: bury nonpolar surface → release ordered waters → folding becomes favorable.
• Two-state unfolding: sigmoidal curves; Tm is midpoint; DSC peak area → ΔHcal.
• Denaturants vs osmolytes: denaturants favor solvation/exposure; osmolytes favor burial/folding.