BMB170a_2011_LECTURE - BMB/Bi/Ch170a Proteins Lecture 4 Oct 6 •  More on protein stability •  Protein Folding

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: BMB/Bi/Ch170a Proteins Lecture 4, Oct 6 •  More on protein stability •  Protein Folding Problem •  Chaperones Problem sets due next Tuesday! Hydrophobic effect thermodynamics of the transfer of nonpolar groups from a nonpolar solvent to water liquid phase hexane water hexane at 20˚ C thermodynamic parameters for hydrocarbons are: ΔG > 0 (unfavorable) ΔH = 0 to < 0 (slightly favorable) - TΔS > 0 (very unfavorable – cage effect of water) ΔCp > 0 (large and posi\ve) ΔG = ΔH − TΔS ∂ΔH = ΔCP ∂T ∂ΔS ΔCP = ∂T T ∂ΔG = −ΔS ∂T Kauzmann Adv Prot Chem (1959) 14:1 Hydrophobicity scale slope ~0.1 kJ/mol/Å2 Nozaki & Tanford JBC (1971) 246:2211 Measure ∆G for amino acids rela\ve to glycine ∆Gtransfer = γ ∆A (inverse to the plot) ∆A = change in accessible surface area γ ~ 0.1 kJ/mol/Å2 (surface tension of water ~ 0.4 kJ/mol/Å2 ) Chothia Nature (1974) 248:338 Molecular interpreta\on of the hydophobic effect (complicated) benzene è༎ water •  It involves the interac\ons between non- polar and water molecules. •  at low T, water has restricted orienta\ons, but these can form good H bonds •  at high T, water has more orienta\ons, but with weaker or fewer H bond Dill Biochem (1990) 29:7133 Temperature dependence of the hydrophobic effect ∂ΔG = −ΔS > 0 ∂T ⇒ hydrophobic effect becomes stronger with T at 20˚C, 20˚C € 140˚C ΔG = ΔH − T ΔS ∂ΔH = ΔCP ∂T ∂ΔS ΔCP = ∂T T ∂ΔG = −ΔS ∂T Creighton Curr. Opin. Struct. Biol. 1, 5 (1991) Electrosta\c interac\ons in proteins •  Electrosta\c energy (U) U = 1389 € z1z2 kJ/mole (r in Å) εr –  Dilectric constant (εeff) – dependent on solvent –  2 opposite charges, r = 4 Å in water (ε=80) U = - 4.3 kJ/mol –  ε decreases with increasing T •  experimentally, εeff ~ 40- 80 –  Rees JMB (1980) 141:323 –  Lee et al Prot Sci (2002) 11:1004 •  from protein structures: numerical solu\ons to finite difference –  Poisson- Boltzmann equa\on, such as DELPHI –  Honig & Nicholls Science (1995)268:1144 Are favorable electrosta\c interac\ons (ion pairs) stabilizing? •  Brian Maohews “There is now ample evidence that electrosta\c interac\ons between largely solvent exposed amino acids contribute liole to protein stability” •  Experimental and theore\cal indicate they may not be favorable –  Bond formed vs. two desolvated polar groups •  ΔH = 0 to slightly less than 0 –  Fixed residues (- ΔS) •  Mul\ple networks –  Addi\onal bond only half the cost –  Only one - ΔS and desolvated group Goldman Structure (1995)3:1277 P. furiosus formaldehyde ferredoxin oxidoreductase, Hu et al JMB (1999) 286:899 How stabilizing is a salt bridge (if at all?)? CPK Ubiqui\n K11- E34 salt bridge interac\on Electrosta/cs Double Mutant Cycles (DMC) (~0) Double Mutant (~0) Wildtype Folded •  •  ∆G unfolding for PXY Unfolded Unfolded Folded Coupling energy ∆∆Gint defined for state B rela\ve to state A: ∆∆Gint = ∆GPXY→PY - ∆GPX→P (corrects for interac\ons between X and P=Y) ≅ ∆GPXY + ∆GP - ∆GPX - ∆GPY (from thermodynamic cycles) ∆∆Gint = 0 means no change in coupling energy from B to A, –  not the same as no coupling energy Horovitz Folding&Design 1,R121 (1996) Double mutant analysis of ion pair strength K11/E34 E11/K34 •  Same coupling energy no maoer how you measure it •  Ion pairs are small contribu\on (vdw interac\ons ~4 kJ/mol) Makhatadze et al JMB (2003)327:1135 Packing, cavi\es and protein stability Small to large subs/tu/ons ΔΔG Kcal/mole Large to small subs/tu/ons ΔΔGstrain ~ −0.6 kJ mol-1 Å-3 ΔV Liu et al JMB (2000)295:127 ΔΔG ~ ΔΔGLeu →Ala - ΔΔGcavity ΔV ΔΔG ~ 0.1 kJ mol-1Å-3 ΔV Eriksson et al Science (1992) 255:178 € € for comparison: “internal pressure” of protein # ∂E & α 3 % ( = T − P ~ 6000 atm ~ 0.4 kJ/mole/Å $ ∂V 'T β Repacking of lambda repressor core •  core repacking variants •  can tolerate volume varia\ons of several methylene groups •  ~same as <∆V2>1/2 calculated from compressibili\es ΔV 2 1 2 = kB TVβ ~ 50 Å 3 for a typical protein ~ 3 methylene groups Lim & Sauer Nature (1989) 339:31 Effects of crosslinks protein stability •  stabiliza\on of N through incorpora\on of crosslinks •  ideally, ∆G˚ is increased through entropic destabiliza\on of the unfolded D state (fewer states accessible in presence of crosslink) •  Disulfides s s s s •  circularly permuted proteins •  Metals •  ~half of known domains are permuted variants. Jung & Lee Prot. Sci. 10, 1881 (2001) Engineering of disulfide bonds red ox Lindorfer & Becktel in Curr. Research in Protein Chemistry: J.J. Villafranca, ed. (1990) stability curve for ox, red Cys3/97 T4 lysozyme Matsumura et al Nature (1989) 342:291 Other Considera\ons •  Stabilizing forces –  Hydrogen bonds (coopera\vity) –  Hydrophobic effect •  Ideal temperature •  Stability related to surface tension strain –  vdw interac\ons at 4 Å ~ .5 kJ/mole •  steep repulsive barrier; however, strongest when \ghtly packed, falls off quickly 1/r6 –  Salt bridges •  Not a big effect, beoer when addi\ve –  –  –  –  Long range electrosta\cs crosslinks (disulfides and metals) entropic - depends on length of closed loops φ,ψ stabiliza\on (use non- natural amino acids) •  D- ala for Gly in Lα region (JACS 176, 13194 (2004)) •  Solvent effects (surface tension) –  This discussion has really concerned water –  Precipitants   Increase in surface tension lowers solubility •  Bind protein surface •  Denaturants (GdCl vs GdSO4) –  Stabilizing addi\ves (increase polar environment) Protein Stability “The problem of protein structure is twofold: the first is that of the form and proper\es of the protein molecule, and the second that of its internal structure. Owing to the extreme instability of the protein molecule, only the gentlest physical methods can be used” J D Bernal “The Structure of Proteins” Nature (1939) Two state model for protein unfolding: Na\ve ↔ Denatured ∆G˚ = GD - GN If ∆G˚ > 0, N is more stable than D For small, single domain proteins, this model provides a good first approxima\on for the thermodynamics of protein stability. “Theore\cal” approach •  ∆G˚ reflects balance between –  GN, interac\ons stabilizing N (destabilizing D) –  GD, interac\ons stabilizing D (or destabilizing N) ∆G˚ = GD - GN = ΣD stabiliza\on - ΣN stabiliza\on –  N stabilized by: hydrogen bonds, salt bridges, van der waals interac\ons, hydrophobic effect Σ= few ± few kcal/mole/residue –  D stabilized by configura\onal entropy penalty for exposure of buried polar groups, burial of exposed apolar groups Σ= few ± few kcal/mole/residue •  for a small protein of ~ 100 residues, “predict” ∆G˚ ~ 0 ± 100s kcal/mole Experimental approach •  A two state model N → D K = (D)/(N), ∆G˚ = - RTln K = ∆H˚ - T∆S˚ •  Although K is the equilibrium “constant”, it is constant only for a specific set of condi\ons (T, pH, εo, (ligands), (denaturants), etc. etc.) •  “How stable is a protein” or “is this protein more stable than another” –  unique answers only when the experimental condi\ons are defined! •  Kine\c analyses of stability may not reflect the thermodynamics of protein stability Basic experimental approach Measure varia\on in K as a func\on of an environmental parameter (typically T or denaturant), by monitoring some property (spectroscopic, hydrogen exchange, thermodynamic) that differs between N and D. cN c D AN − A(T ) K (T ) = = c N A(T ) − AD cD ΔG˚(T ) = − RT ln[K (T )] Absorbance AD A AN Na\ve T Denatured € Hermans & Scheraga JACS (1961) 83:3283 Anfinsen and Protein folding •  Chris\an Anfinsen •  Used ribonuclease to show that the amino acid sequence determined fold •  Denatured ribonuclease could be refolded to ac\ve form •  “Thermodynamic hypothesis” –  Anfinsen’s Dogma •  1972 Nobel Prize in Chemistry •  Ul\mate basis of protein folding Need for different folding rates? •  To get across intracellular membranes, proteins must be unfolded or can’t be translocated •  Consider some of the membranes proteins have to cross –  Inner membrane of E. coli –  Mitochondrial/chloroplast membrane –  Endoplasmic reticulum •  Folding fastest not always best Maltose binding protein (MBP) Needs to remain unfolded for transport •  E. coli periplasmic space protein •  If MBP folds in cytoplasm, it can’t be transported across inner membrane –  Found altered signal sequence that reduced its rate of transport •  MBP accumulated in the cytoplasm and folded •  Once folded, couldn’t be transported •  Randall & Hardy Cell (1986) 46: 921 –  2nd amino acid substitution within mature protein restored export •  Reduced folding rate •  Look for mutants with slow folding rates by screening for intragenic suppressors of signal sequence mutations? •  Cover et al J Bacteriol (1987) 169:174 Transport to Golgi requires FOLDED protein •  Flu virus haemagglu\nin –  only transported to Golgi if it is folded –  requires stable trimeric form of protein •  Mutant forms of protein accumulate as monomers inside ER and are UNFOLDED Gething et al Cell (1986) 46:939 Folding in vitro •  Remove protein from denaturant –  –  –  –  Collapses into compact shape within msecs Buries hydrophobic residues 2˚ structures form Native 3˚ structures form (takes seconds to minutes) Hartl “Molecular chaperones in cellular protein folding” Nature (1996) 381: 571- 80 Folding in vivo •  Part of nascent chain hidden inside ribosome. •  Time to fold a 40 kD polypeptide chain: ~15 sec in E. coli; ~2-3 minutes in eukaryotic cytosol. Hartl “Molecular chaperones in cellular protein folding” Nature (1996) 381: 571- 80 In vitro studies of protein (un)folding D N •  Small, single- domain proteins likely fold same way in vivo •  Folding mechanisms likely apply to domains of large proteins •  Relevant to unfolding (transfer across membrane; proteasome degrada\on) •  Misfolding (amyloidosis; prions) Fersht & Daggeo (2002) Cell 108:573- 82 What characterizes the N state during folding? D N Anfinsen’s thermodynamic hypothesis: N is the state of lowest free energy “The studies on the renatura\on of fully denatured ribonuclease required many suppor\ng inves\ga\ons (9, 10, 11, 12) to establish, finally, the generality which we have occasionally called (13) the “thermodynamic hypothesis”. This hypothesis states that the three- dimensional structure of a na\ve protein in its normal physiological milieu (solvent, pH, ionic strength, presence of other components such as metal ions or prosthe\c groups, temperature, etc.) is the one in which the Gibbs free energy of the whole system is lowest; that is, that the na\ve conforma\on is determined by the totality of interatomic interac\ons and hence by the amino acid sequence, in a given environment. In terms of natural selecCon through the “design” of macromolecules during evolu\on, this idea emphasized the fact that a protein molecule only makes stable, structural sense when it exists under condi\ons similar to those for which it was selected - the so- called physiological state.” Anfinsen Nobel Lecture (1972) How is this state found kine\cally? •  Levinthal Paradox –  Assume a random search of all possible conforma\ons –  100 residues @ 10 conforma\ons/residue = 10100 conforma\ons (if all conforma\ons accessible) –  Sampled at 10- 13 sec = 1087 sec = ~1080 years –  proteins fold in ~10- 6 to 10+2 sec •  Folding pathways must exist Levinthal J Chim Phys (1968) 65:44- 45 Possible folding mechanisms Fersht “Structure and Mechanism in Protein Science” Implica\ons for folding pathways •  Proteins cannot fold up sequen\ally (β- strand order) vs 1 2 3 1 3 2 •  Proteins don’t fold like a “ball of yarn” •  Residues near in sequence tend to be near in space (Domains) •  N- and C- termini tend to be close together •  Strong tendency against knots •  Different sequences can have the same fold •  Pentapep\de and higher sequences can have different folds ...
View Full Document

This note was uploaded on 01/03/2012 for the course BI 170a taught by Professor List during the Fall '09 term at Caltech.

Ask a homework question - tutors are online