Skip to content

Protein Structure, Conformation and Stability

Amino acids, the peptide bond, Ramachandran plot, helices and sheets, motifs, domains and folds, folding and protein stability for CSIR NET Life Sciences.

  • 20 min read
  • Updated
On this page · 17 sections
  1. Prerequisites
  2. The Hook: Hard Spheres in Madras
  3. Amino Acids: the Building Blocks
  4. The Peptide Bond
  5. The Ramachandran Plot
  6. Secondary Structure
  7. Supersecondary Structure (Motifs), Domains and Folds
  8. Tertiary and Quaternary Structure
  9. Protein Folding
  10. Stability of Proteins
  11. Memory Tricks
  12. Don’t Confuse
  13. Common Mistakes
  14. Practice
  15. Quick Revision Table
  16. Key Takeaways
  17. What’s Next

Prerequisites

The Hook: Hard Spheres in Madras

How CSIR NET asks it. Part B: φ and ψ of a helix or sheet, which residue occupies the left-handed region, helix geometry (3.6 residues, 5.4 Å), motif–example pairs (HTH, zinc finger, EF-hand, Rossmann fold). Part C: Anfinsen-type experiments, reading a CD spectrum or a melting curve, two-state ΔG calculations, and statements about what stabilises a thermophilic protein.


Amino Acids: the Building Blocks

All 20 standard amino acids except glycine are chiral at Cα; proteins use L-amino acids. Lehninger groups them by side-chain (R group) polarity at pH 7:

GroupAmino acidsNotes
Non-polar, aliphaticGly, Ala, Pro, Val, Leu, Ile, MetGly is achiral and flexible; Pro is a cyclic imino acid that restricts φ
AromaticPhe, Tyr, TrpTyr and Trp absorb at 280 nm; Trp is the main fluorophore
Polar, unchargedSer, Thr, Cys, Asn, GlnCys forms disulfides; Ser/Thr/Tyr are phosphorylated; Asn is N-glycosylated
Positively chargedLys, Arg, HisHis (pKa about 6) is the common acid–base catalyst near neutral pH
Negatively chargedAsp, GluGlu is γ-carboxylated in vitamin K-dependent proteins

Selenocysteine (encoded by UGA with a special insertion element) and pyrrolysine (UAG in some archaea and bacteria) are the 21st and 22nd genetically encoded amino acids. Hydroxyproline and hydroxylysine (collagen), γ-carboxyglutamate and phospho-amino acids are post-translational modifications.

The Peptide Bond

The peptide (amide) bond links the α-carboxyl of one residue to the α-amino of the next with loss of water. Pauling and Corey showed from crystal structures that:

  • The C–N bond (about 1.32 Å) is shorter than a C–N single bond (about 1.49 Å): resonance between C=O and the N lone pair gives it partial double-bond character.
  • So the six atoms Cα–C(=O)–N(H)–Cα lie in one plane (the peptide plane), and there is no free rotation about C–N.
  • The peptide bond is almost always trans (ω ≈ 180°): the two Cα atoms are on opposite sides, minimising clash. cis occurs far more often for X–Pro bonds, because both forms clash similarly there. cis–trans isomerisation of X–Pro bonds is slow and is catalysed by peptidyl-prolyl isomerases.
  • The backbone is therefore a series of rigid planes joined at Cα, with rotation allowed only about N–Cα (φ, phi) and Cα–C (ψ, psi).

The Ramachandran Plot

The Ramachandran plot is a map of ψ (y-axis) against φ (x-axis), each from −180° to +180°, showing which combinations are sterically allowed (no clash of backbone and Cβ atoms using hard-sphere van der Waals radii).

ConformationφψRegion of plot
Right-handed α-helixabout −57°about −47°Lower left quadrant
Antiparallel β-sheetabout −139°about +135°Upper left quadrant
Parallel β-sheetabout −119°about +113°Upper left quadrant
Collagen triple helix (each strand)about −51°about +153°Upper left, near polyproline II
Left-handed α-helixabout +57°about +47°Upper right quadrant; mostly Gly (also Asn, Asp)

(Values as tabulated in Lehninger.)

graph TD
    P["Ramachandran plot
phi on x-axis, psi on y-axis"] --> Q2["Upper left: phi negative, psi positive
beta strands, polyproline II, collagen"] P --> Q3["Lower left: phi negative, psi negative
right-handed alpha-helix, 3-10 helix"] P --> Q1["Upper right: phi positive, psi positive
left-handed helix, mostly glycine"] P --> Q4["Lower right: phi positive, psi negative
rare, glycine only"]

Key points:

  • Glycine, with only H as its side chain, has a much larger allowed area, including positive φ, and its map is symmetric about the origin. Gly is common in tight turns and the left-handed region.
  • Proline has its side chain bonded back to the backbone N, so its φ is fixed near −60°; it also lacks an N–H donor. It breaks α-helices (except in the first turn) and is common in turns and polyproline helices.
  • β-branched residues (Val, Ile, Thr) are restricted and favour β-strands.
  • The plot uses only backbone sterics; it says nothing about energy or side-chain packing. Residues of a high-quality structure fall overwhelmingly in the allowed regions.

Secondary Structure

The α-helix

Proposed by Pauling, Corey and Branson (1951).

  • 3.6 residues per turn, rise 1.5 Å per residue, pitch 5.4 Å; each residue is rotated 100° about the axis.
  • The C=O of residue i H-bonds to the N–H of residue i + 4. Each H-bond closes a ring of 13 atoms, so the α-helix is also called the 3.6₁₃ helix. Nearly all backbone C=O and N–H are bonded, except the first four N–H and the last four C=O.
  • Essentially all α-helices in proteins are right-handed (for L-amino acids the left-handed form has more steric strain).
  • Side chains point outward. Residues i, i + 3, i + 4 and i + 7 lie on the same face: amphipathic helices have a hydrophobic stripe on one side (seen on a helical wheel).
  • Helix dipole: the aligned peptide dipoles give a partial positive charge at the N-terminus and partial negative at the C-terminus. Negatively charged residues near the N-terminus and positively charged ones near the C-terminus stabilise it, and phosphate groups of ligands often bind at helix N-termini.
  • Helix formers: Ala (highest propensity), Leu, Met, Glu, Gln, Lys. Breakers: Pro and Gly. Adjacent like charges or bulky β-branched residues at i and i + 3/4 destabilise.

Related helices: the 3₁₀ helix (i → i + 3 H-bonds, tighter, often at helix ends) and the π-helix (i → i + 5, wider, short segments).

β-strands and β-sheets

  • Extended backbone, about 3.3–3.5 Å rise per residue; side chains alternate above and below the sheet.
  • H-bonds form between strands, not within one strand.
  • Antiparallel sheets have more linear H-bonds and are slightly more stable; parallel sheets need a crossover connection (usually an α-helix) between strands.
  • Sheets are usually twisted, not flat. A β-bulge is an irregularity where an extra residue disrupts pairing.

Turns and loops

  • β-turn: four residues; the C=O of residue i H-bonds to the N–H of i + 3, reversing chain direction. Gly (flexible) and Pro (fixed φ, cis-capable) are common. Types I and II differ in the orientation of the central peptide; type II often has Gly at i + 2.
  • Ω-loops and other irregular loops, usually on the surface, form many binding and catalytic sites.

Fibrous proteins: secondary structure as the whole protein

ProteinStructureKey composition
α-Keratin (hair, nails)Right-handed α-helices twisted into a left-handed coiled coilDisulfide cross-links (hard keratin is Cys-rich)
CollagenThree left-handed polyproline-like chains (about 3 residues per turn) wound into a right-handed triple helixRepeating Gly–X–Y; X often Pro, Y often 4-hydroxyproline; Gly every third residue fits the crowded centre
Silk fibroinAntiparallel β-sheetsRich in Ala and Gly; small side chains allow tight stacking

The collagen triple helix was proposed by G. N. Ramachandran and G. Kartha (1954–55), another Madras contribution. Vitamin C deficiency blocks prolyl hydroxylation, weakening collagen (scurvy). A Gly substitution in the triple helix causes osteogenesis imperfecta.

Supersecondary Structure (Motifs), Domains and Folds

Motif (supersecondary structure): a small, recurring arrangement of a few secondary-structure elements. Domain: a compact, independently folding part of a polypeptide, often with its own function; large proteins have several. Fold: the overall arrangement and connectivity of secondary structures in a domain.

Motif / foldDescriptionExample(s)
β-α-βTwo parallel strands joined by a helix; the crossover is almost always right-handedBuilding block of α/β proteins
β-hairpinTwo antiparallel strands joined by a tight turnMany β-proteins
Greek keyFour antiparallel strands, the fourth folded back next to the firstImmunoglobulin-like and crystallin domains
Helix-turn-helix (HTH)Two helices at an angle; the recognition helix sits in the DNA major grooveλ repressor, Cro, lac repressor, homeodomain
Zinc finger (C₂H₂)ββα unit with Zn²⁺ held by two Cys and two HisTFIIIA, Zif268
Leucine zipperAmphipathic helix with Leu at every seventh position (d of the heptad); dimerises as a coiled coilGCN4, Fos–Jun
Helix-loop-helixTwo helices joined by a loop; dimerises (bHLH binds DNA)MyoD, Myc
EF-handHelix–loop–helix that binds Ca²⁺ in the loopCalmodulin, troponin C
Rossmann foldTwo βαβαβ units forming a parallel sheet; binds NAD⁺/FADLactate and other dehydrogenases
TIM barrelEight parallel β-strands forming a barrel, surrounded by eight α-helices, (βα)₈Triose phosphate isomerase, many enzymes
Four-helix bundleFour antiparallel helices packed togetherCytochrome b562, ferritin
β-barrel / β-sandwichSheet closed into a barrel, or two sheets packed face to facePorins (barrel), Ig fold (sandwich)

Structural classification (SCOP, CATH) groups domains into classes: all-α, all-β, α/β (alternating α and β, mainly parallel sheets) and α+β (α and β segregated, mainly antiparallel sheets). Common protein-interaction domains: SH2 binds phosphotyrosine, SH3 binds Pro-rich sequences, PH binds phosphoinositides.

Tertiary and Quaternary Structure

Tertiary structure is the full 3D arrangement of one chain. General features of globular proteins:

  • A hydrophobic core of non-polar side chains, tightly packed (packing density similar to organic crystals); polar and charged residues on the surface.
  • Buried backbone C=O and N–H are almost always H-bonded (in helices and sheets), because an unpaired buried polar group is costly.
  • Disulfide bonds are common in extracellular proteins; the reducing cytosol keeps cysteines free.
  • Some proteins or regions are intrinsically disordered, folding only when they bind a partner.

Quaternary structure is the arrangement of subunits. Haemoglobin is an α₂β₂ tetramer (a dimer of αβ protomers). Oligomers usually have cyclic (Cₙ) or dihedral (Dₙ) symmetry. Quaternary structure enables cooperativity and allostery (see enzyme regulation).

graph LR
    A["Primary
sequence"] --> B["Secondary
helix, sheet, turn"] B --> C["Supersecondary
motifs"] C --> D["Tertiary
domains, fold"] D --> E["Quaternary
subunit assembly"]

Protein Folding

Anfinsen’s experiment (ribonuclease A). RNase A (124 residues, 4 disulfide bonds) was denatured in 8 M urea + β-mercaptoethanol, losing all activity.

  • Removing urea and β-mercaptoethanol together (dialysis), then allowing air oxidation, regained nearly full activity: the sequence contains all the information needed to fold.
  • Oxidising in the presence of urea gave “scrambled” RNase with about 1% activity: 8 Cys can pair into 105 possible sets of 4 disulfides (7 × 5 × 3 × 1), and only one is native, so random pairing gives about 1/105.
  • Adding a trace of β-mercaptoethanol to scrambled RNase (no urea) let disulfides shuffle, and activity returned: the native structure is the thermodynamically most stable (the thermodynamic hypothesis). Anfinsen shared the 1972 Nobel Prize in Chemistry.

Levinthal’s paradox. A random search through all backbone conformations would take longer than the age of the universe, yet proteins fold in microseconds to seconds. Folding is therefore not random: it proceeds through a funnel-shaped energy landscape, with local structure forming quickly, hydrophobic collapse to a molten globule (native-like secondary structure, loosely packed side chains) and then final packing.

Helpers in the cell (they speed folding and prevent aggregation but do not supply structural information):

  • Hsp70 binds exposed hydrophobic segments of nascent chains (ATP-dependent cycles).
  • Chaperonins: GroEL (two stacked heptameric rings) with the GroES cap, ATP-driven; the eukaryotic analogue is TRiC/CCT. They provide an isolated folding chamber.
  • Protein disulfide isomerase (PDI) shuffles disulfides; peptidyl-prolyl cis–trans isomerase speeds X–Pro isomerisation.

Misfolding: amyloid fibrils have a cross-β structure (strands perpendicular to the fibre axis); prion disease involves conversion of α-rich PrPᶜ into β-rich PrPˢᶜ that templates further conversion.

Stability of Proteins

Marginal stability

The native state is only marginally more stable than the unfolded state: the free-energy difference in typical proteins under physiological conditions is quoted (Lehninger) in the range of about 20–65 kJ/mol, the small difference between large opposing terms.

Favours foldingOpposes folding
Hydrophobic effect (largest)Loss of conformational entropy of the chain
van der Waals packing in the coreBurial of polar groups not perfectly H-bonded
H-bonds and ion pairs (net contribution smaller, as they replace H-bonds with water)
Disulfides (they lower the entropy of the unfolded state)

The two-state model

Many small proteins unfold cooperatively, with only N and U populated:

$$K_U = \frac{[\mathrm{U}]}{[\mathrm{N}]} = \frac{f_U}{1 - f_U}, \qquad \Delta G_U = -RT\ln K_U = \Delta H_U - T\Delta S_U$$
  • At the melting temperature Tm, [U] = [N], K = 1, ΔG = 0, so \(\Delta S_U = \Delta H_U/T_m\).
  • Unfolding exposes non-polar groups, so it has a positive heat-capacity change (ΔCp). As a result ΔG_U(T) is curved, with maximum stability at an intermediate temperature: proteins can also unfold on cooling (cold denaturation).

What denatures a protein

AgentHow it acts
HeatIncreases TΔS of unfolding
Extreme pHChanges ionisation, creating like-charge repulsion
Urea, guanidinium chloride (about 6–8 M)Solubilise non-polar groups and the backbone; weaken the hydrophobic effect
Detergents (SDS)Bind hydrophobic regions and coat the chain with charge
Organic solventsWeaken the hydrophobic effect
β-Mercaptoethanol, DTTReduce disulfides (they do not denature by themselves if the protein is otherwise stable)

Salts follow the Hofmeister series: kosmotropes (sulfate, phosphate, ammonium) stabilise proteins and salt them out (ammonium sulfate fractionation); chaotropes (thiocyanate, guanidinium) destabilise.

Thermophilic proteins are typically stabilised by more ion pairs (networks of salt bridges on the surface), shorter or Pro-rich loops, tighter core packing, and higher oligomeric states, rather than by any single special interaction.

Measuring structure and stability

  • Far-UV circular dichroism (CD): α-helix gives negative bands at about 222 and 208 nm and a positive band near 192 nm; β-sheet a negative band near 218 nm; random coil a negative band near 200 nm. Loss of the 222 nm signal tracks unfolding.
  • Trp fluorescence: buried Trp emits at shorter wavelength; on unfolding the emission maximum red-shifts (towards about 350 nm) as Trp meets water.
  • Differential scanning calorimetry: heat absorbed during melting gives Tm and ΔH directly.
  • Chemical denaturation curves (CD or fluorescence vs [urea]) extrapolated to zero denaturant give ΔG_U in water.

The nucleic-acid side of “stability” (Tm, GC content, salt, hyperchromicity) is on the nucleic acids page.


Memory Tricks

Don’t Confuse

ThisvsThat
φ: rotation about N–Cαvsψ: rotation about Cα–C; ω: the peptide bond itself (fixed near 180°)
Motif: small recurring pattern, may not fold alonevsDomain: independently folding unit
α-keratin: α-helices, left-handed coiled coilvsCollagen: left-handed chains, right-handed triple helix
Chaperones: prevent aggregation, assist foldingvsSequence: carries the information that specifies the fold (Anfinsen)
Denaturation: loss of 3D structure, peptide bonds intactvsHydrolysis: peptide bonds broken
Molten globule: native secondary structure, loose packingvsNative state: unique tight packing
Disulfide stabilises by destabilising the unfolded state (entropy)vsHydrophobic effect stabilises the folded state (water entropy)

Common Mistakes


Practice

All questions below are original practice questions written in CSIR NET style. They are not previous-year questions.

Level 1: Foundation

  1. Between which atoms does the α-helix H-bond run?
  2. Which amino acid is most often found with positive φ on a Ramachandran plot?
  3. What did scrambled ribonuclease show about the native disulfide pattern?
Answers
  1. From the C=O of residue i to the N–H of residue i + 4.
  2. Glycine.
  3. That the native disulfide set is one of 105 possible ones and is reached only when the chain can fold, because it is the thermodynamically most stable structure.

Level 2: Part B Style

Q1 (practice). A 36-residue segment forms a single continuous α-helix. Its length and number of turns are approximately:

  • (a) 36 Å, 6 turns
  • (b) 54 Å, 10 turns
  • (c) 122 Å, 10 turns
  • (d) 54 Å, 13 turns
Answer and explanation

(b) 54 Å, 10 turns. 36 × 1.5 Å = 54 Å; 36/3.6 = 10 turns. As a β-strand the same 36 residues would stretch to roughly 120 Å, which is option (c)’s distractor.

Q2 (practice). In a leucine zipper, leucine residues occur every seventh position because:

  • (a) leucine is the only residue that forms β-sheets
  • (b) with 3.6 residues per turn, positions i and i + 7 lie on nearly the same face of the helix, creating a hydrophobic stripe for coiled-coil dimerisation
  • (c) leucine coordinates Zn²⁺
  • (d) every seventh residue must be a helix breaker
Answer and explanation

(b). Seven residues span almost exactly two turns (7 × 100° = 700°, 20° short of 720°), so heptad positions a and d form a stripe that winds slowly around the helix; two such helices wrap around each other in a left-handed coiled coil.

Q3 (practice). A far-UV CD spectrum of a purified protein shows strong negative bands at 208 nm and 222 nm. In 6 M guanidinium chloride the spectrum changes to a single negative band near 200 nm. The best interpretation is:

  • (a) the protein is mainly β-sheet and becomes α-helical in denaturant
  • (b) the protein is substantially α-helical and unfolds to random coil in denaturant
  • (c) the protein is disordered in both conditions
  • (d) the change reflects reduction of disulfides only
Answer and explanation

(b). Double minima at 208 and 222 nm are the α-helix signature; a single minimum near 200 nm is random coil. Guanidinium chloride unfolds proteins. Reduction of disulfides alone would not be expected to remove helical CD.

Q4 (practice). Which motif–protein pair is incorrect?

  • (a) EF-hand – calmodulin
  • (b) Rossmann fold – NAD⁺-binding dehydrogenases
  • (c) Helix-turn-helix – λ repressor
  • (d) TIM barrel – collagen
Answer and explanation

(d). The TIM barrel, (βα)₈, is named after triose phosphate isomerase and is common among enzymes. Collagen is a fibrous triple helix.

Level 3: Part C Style

Q1 (practice). A small protein unfolds by a two-state mechanism. At 310 K, 20% of molecules are unfolded. Its Tm is 330 K and ΔH of unfolding (assume temperature-independent, ΔCp = 0) is 400 kJ/mol.

(i) Calculate ΔG_U at 310 K from the population data. (ii) Calculate ΔS_U and the ΔG_U at 310 K predicted from ΔH and ΔS. (iii) Are the two values consistent? What might explain a difference?

Answer and worked solution
  1. \(K_U = 0.2/0.8 = 0.25\). \(\Delta G_U = -RT\ln K = -(8.314)(310)\ln 0.25 = +3.57\) kJ/mol.
  2. At Tm, ΔG = 0, so \(\Delta S_U = \Delta H/T_m = 400/330 = 1.21\) kJ mol⁻¹ K⁻¹. At 310 K: \(\Delta G_U = 400 - 310(1.212) = +24.2\) kJ/mol.
  3. Not consistent. +24.2 kJ/mol would make the unfolded fraction at 310 K tiny (\(K \approx e^{-24\,200/(8.314\times310)} \approx 8\times10^{-5}\)). The large observed unfolded fraction means the assumptions fail: the protein may not be two-state (an intermediate is populated), ΔCp is not zero (making ΔG(T) curved), or the “unfolded” signal includes a partially folded state. This kind of data-consistency reasoning is typical of Part C.

Q2 (practice). An experimenter repeats Anfinsen’s ribonuclease experiment with four variations. Predict the activity recovered in each.

(A) Denature in 8 M urea + β-mercaptoethanol; dialyse away both; air-oxidise. (B) Denature as in A; oxidise in the presence of 8 M urea; then remove urea. (C) Take the product of B; add a trace of β-mercaptoethanol without urea; incubate. (D) Denature as in A; remove urea but keep a high concentration of β-mercaptoethanol permanently.

Which combination is correct?

  • (a) A: about 100%; B: about 1%; C: high; D: low or none (disulfides cannot form)
  • (b) A: about 1%; B: about 100%; C: low; D: high
  • (c) A: about 100%; B: about 100%; C: high; D: high
  • (d) A: about 50%; B: about 50%; C: about 50%; D: about 1%
Answer and worked solution

(a).

  • (A) Folding happens before disulfides lock in, so the native pairing forms: near-full activity.
  • (B) Random pairing of 8 Cys: 7 × 5 × 3 × 1 = 105 possibilities, about 1/105 ≈ 1% native: scrambled.
  • (C) A trace of reductant allows disulfide exchange; the chain moves to the most stable (native) pairing: activity largely restored.
  • (D) With excess reductant, the four disulfides needed for a stable active enzyme cannot form, so little activity is expected under these conditions.

Q3 (practice). A thermophilic enzyme and its mesophilic homologue have 45% sequence identity and superimposable folds. Which statements are likely to be correct?

(A) The thermophilic enzyme probably has more surface salt bridges. (B) The thermophilic enzyme must contain more disulfides, because only covalent bonds survive high temperature. (C) Introducing a disulfide that links two residues far apart in sequence can stabilise a protein by lowering the entropy of its unfolded state. (D) The thermophilic enzyme has a larger Ramachandran-allowed area for each residue.

  • (a) A and C
  • (b) A and B
  • (c) B and D
  • (d) C and D
Answer and explanation

(a) A and C. Thermostability usually comes from many small effects, prominently ion-pair networks, tighter packing and shorter loops (A). An engineered disulfide reduces the conformational entropy gained on unfolding, stabilising the native state (C). (B) is wrong: many thermophilic proteins are intracellular with few disulfides. (D) is meaningless: the allowed φ/ψ area is set by backbone sterics, not by the organism.


Quick Revision Table

ItemFact
Peptide bondPlanar, trans, partial double bond; C–N about 1.32 Å
φ / ψ / ωN–Cα / Cα–C / peptide bond (about 180°)
α-helix (φ, ψ)about −57°, −47°
Antiparallel / parallel βabout −139°, +135° / −119°, +113°
Left-handed helixφ positive; mostly Gly
α-helix geometry3.6 residues/turn, 1.5 Å/residue, 5.4 Å pitch, i → i+4, 3.6₁₃
3₁₀ helix / π-helixi → i+3 / i → i+5
β-turn4 residues, i → i+3; Gly, Pro
CollagenGly–X–Y, Pro/Hyp; Ramachandran & Kartha
Coiled coilHeptad, Leu at d (zipper)
FoldsRossmann (NAD), TIM barrel (βα)₈, Ig β-sandwich
AnfinsenRNase A, 8 M urea + β-ME; 105 disulfide sets; ~1% scrambled
StabilityMarginal; hydrophobic effect dominant
Two-stateK = f_U/(1 − f_U); at Tm ΔG = 0, ΔS = ΔH/Tm
CDα: −222, −208 nm; β: −218 nm; coil: −200 nm
Trp fluorescenceRed-shift on unfolding

Key Takeaways

What’s Next