Enzyme catalysis and protein engineering Flashcards
(27 cards)
Saturation Kinetics and the Michaelis-Menten equation
Kcat and Km
Catalysis meaning
Catalysis= transition state stabilisation: making and breaking bonds is charge costly, especially charge accumulation, so catalytic groups often offset charge development in the TS and hence compounds mimicking TS predicted to bind active site tighter than TS, and are common drugs, e.g., phosphoamidate (tetrahedral) binds acyl transfer enzymes more strongly than TS, replacing C=O group at which peptide cleavage happens. Tetrahedral TS binding also explains stereospecificity of enzymes w/planar initial substrates. Binding also contributes directly to catalysis.
Catalytic toolkit: Entropic effect, nucleophilic catalysis and general acid/base
- Entropic effect: bringing substrates together increases effective molarity, intramolecular reaction orders of magnitude faster than intermolecular. Size of approximation effect gauged by comparing intra+inter-molecular rates. More constrained binding-> greater entropy loss, which can be made up for using binding energy.
- Nucleophilic catalysis: inert substrate-> more reactive enzyme-bound species. E.g., in chymotrypsin peptide hydrolysis, amide bond turned into ester (rapidly hydrolysed) by Nu catalyst serine hydroxyl. Advantage over direct attack by water as alcohols are better nucleophiles/more reactive intermediate and covalent linkage to the substrate gives the enzyme a more rigid/direct handle on substrate (pre-orientation of substrate and nucleophile). Residues involved include Ser, Cys, Lys, Tyr, Glu+ Asp
- General acid: transfer of a proton to the TS to produce better leaving group. His, Lys
- General base: deprotonation of a substrate can make it more reactive (active site has H+ acceptor). Asp, Glu, His, Lys
Catalytic toolkit: metal ions and hydrophobic effect
- Metal ion: can deprotonate nucleophiles (like general base) to make them more reactive. Metal-coordinated water has a lower pKa so significant concentrations of active OH- are generated. Metal ions can also offset -ve charge (as Lewis acids) to stabilize a leaving group. Metals sometime AKA “superprotons” as exist independently of pH. Ser, Cys, His, Asp and Glu can coordinate metals.
- Hydrophobic effects- active sites can be as hydrophobic as organic solvents, leading to: pKa of active site groups perturbed (carboxylic acid pKas go up, amine base pKas go down, destabilising their charged forms); charge-charge interactions reinforced; substrate reactivity potentially altered by desolvation (organic solvents can mimic this effect)
Studying mechanisms of catalysis
Can be stablished by structural info (Xray, etc), biochemical reaction rates, mutagenesis/protein engineering (find essential aas).
Serine proteases include alpha-chymotrypsin, as well as trypsin. Elastase thrombin, plasmin (dissolved blood clots), subtilisin, complement proteases.
Alpha-chymotrypsin synth from inactive precursor chymotrypsinogen, activation by proteolysis by trypsin (cleave arg15/Ile16 bond), then chymotrypsin -> excision of 2 dipeptides (Ser14-Arg15+ Thr147-Asn148). Zymogen residue numbering retained in enzyme.
Kinetic characterisation of enzymes and biphasic kinetics
: challenge w/ various substrates+ ID right one by turnover, quantify enzyme ability w/ Michaelis-Menten kinetics. For alpha-chymotrypsin, specificity broad. Analysis of steady state data gives idea of how active site looks-> larger, aromatic sidechains better, so evolved for substrates w/ aromatic sidechains.
Biphasic kinetics: found by adding large amount of enzyme to ester substrate 4-nitrophenyl acetate-> burst of 4-nitrophenol. Burst of product indicates fast formation of intermediate (good lg)+ slow turnover intermediate (hydrolysis of EP complex)-> free enzyme (i.e., the reaction has 2 steps). Acyl group attached to Ser195. Burst kinetics indicate nucleophilic catalysis of Ser. 2 phases= fast (single turnover/ pre-steady state kinetics- ms to a few s, followed prior to assuming equilibria that form basis of MM kinetics)+ slower (RDS) steady state linear phase. Can be measured w/ stopped flow instrument: rapidly mis E+S, stop reaction by opposing flow within few ms, monitor reaction progress. Important when investigating existence, rate of formation+ breakdown of intermediate. Possible to observe similar pattern with strong product inhibition.
Covalent chemical modification
probe ID+ role of aas in active site. If specific mod to residue-> proportional loss in catalytic activity, residue must have role in catalysis. Caveats: must show reagent is specific by direct characterisation of modified protein; bulk introduced by mod may disturb/block active site even if residue not essential for mech; covalent modification only proves a residue is in/near active site, doesn’t evidence its function/ degree of involvement. For chymotrypsin, mods w/ affinity labels succeeded in IDing 2 key active site residues by treatment w/ covalent (suicide) inhibitors
* Mod w/ TPCK (inhibitor)- mimics substrate. Chymotrypsin prefers cutting after L-Phe, so aromatic ring in TPCK directs it into the active site. Reactive chloroketone group reacts w/ Nu groups. This reagent labelled His57.
* Mod w/ DIPF: oxygen-specific reagent, specifically labelled Ser195. Ser normally protonated @ physio pH, but Ser 195 in active site env where arrangement of functional groups makes it reactive/Nu. Modified protein hydrolysed, chromatographically analysed-> show which aa modified.
Crystal structures and alanine scanning. Catalytic effect
reveal spatial arrangement of catalytic triad including His and Ser ID’d by cov mod+ Asp102 apparently not accessibly to cov mod. Arrangement= charge relay system- helps position Ser 195 for Nu attack+ tune its reactivity. Also shows oxyanion hole formed by backbone NH H bonds- 3 main chain NH groups precisely placed, but provide little/no stabilisation of enzyme-substrate complex- form strong H bonds as electron density increases in TS. Good evidence for specific TS stabilization as mech for rate increase.
Alanine scanning: validates role of each aa and their contribution to catalysis. e.g., subtilisin (Ser protease)- catalytic triad replaced w/Ala to probe role of potential catalytic residues ID’d after seq comparisons- if mutation lowers rate, mutated residue has catalytic role. Ser195Ala and His57Ala mutants had 106x lower kcat and same Km, highlighting their importance in catalytic triad. Ser/His/Asp-> 3ala still has acceleration 10k fold- evidence for catalytic contribution from oxyanion hole.
Catalytic effect in both steps of reaction, during formation of tetrahedral intermediates in both acylation of Ser195+ hydrolysis of acyl-enzyme by bound water. Proposed mech from this evidence: His57 acts as general base, accepts proton from Ser195, which then Nu attacks peptide bond. Asp102 binds His57 to modulate its position and reactivity.
Alternative solutions to enzymatic acyl transfer
Other solutions for enzymatic acyl transfer: second acyl transfer architecture w/ different functionality in carboxypeptidase A. Direct attack of metal-bound water, no intermediate. Catalytic base carbolylate, rather than His. Hydrolytic enzymes all use water, but positioning it tricky- resolve by attaching as ligand to metal (here, Zn), increasing chemical reactivity/reducing pKa to favour OH- formation. Fine-tuning of reactivity by metal and #/type/geometry of other ligands.
Steady-state kinetics and pH-rate profile
Steady-state kinetics-Usual MM equation derivation assumes constant [ES] after initial turnovers (steady state). Experimental pH rate profiles show some catalytic groups on sidechains are ionisable, exist in different ionic form at different pHs- this may contribute to catalysis, and proposed mechs should be consistent with observed pH dependence. For uncatalyzed peptide hydrolysis in water, rate fastest at extremes of pH: at low pH, protons help stabilise charge development on leaving group+ carbonyl oxygen; at high pH, more OH (more reactive than H2O). For catalysed reaction kcat/Km (rate at low [S]) with an uncharged peptide substrate, reaction fastest at neutral pH (graph complementary shape to rate of uncatalyzed reaction). Rate is a combination of 2 factors: Overall pH rate profile for kcat (pH on x axis) is a sigmoid, i.e., pH dependence of ES complex (chemistry works best at high pH), as substrate fully bound under saturating conditions. pH dependence for 1/Km (where binding is strongest) i.e. pH dependence arising from ionisations in free E and S is inverse sigmoid (binding tighter at low pH). At high pH, asp-194 blocks specificity pocket. At neutral pH, protonated Ile forms ion pair with asp carboxylate, opening pocket and allowing hydrophobic sidechains of substrates to bind. Asp102 C=O holds imidazole on His57 in place so it can accept proton from Ser195 when the OH group attacks the substrate. pH-rate profiles add info on functional roles of residues, verify mech model.
Conservation of catalytic motif and binding pocket in Ser proteases
Conservation of catalytic motif and binding pocket in Ser proteases: Catalytic triad present in all serine proteases. Aa seq can vary but 3D structure identical (convergent evolution), shown by: structural alignments of chymotrypsin, elastase+ trypsin backbones- highest structural similarity in triad; possibility of implementing catalytic triad in different fold- e.g., Subtilisin has different secondary structure+ overall fold to chymotrypsin, but identical positioning of critical groups (in space, not in seq)
Comparative analysis of binding pocket structure
Comparative analysis of binding pocket structure: trypsin and elastase specificity due to binding pocket geometry, so sidechain immediately N-terminal to the peptide bond being broken; pocket lined w/ hydrophobic residues, accounting for specificity. Chymotrypsin lined w/ 2Gly+Ser, can accommodate larger sidechains; elastase lined w/ ser, val, thr so fits smaller sidechains. Further changes to side chain shape can occur- when chymotrypsinogen activated by proteolysis, new salt bridge forms between Ile16+Asp194-> movements in residues 189-192 form binding pocket; also, main-chain NH of Gly193 re-oriented to correct position.
Protein engineering
Protein engineering: alteration of specificity attempted by increasing a sidechain’s bulk (Gly226->Ala) in specificity pocket (disfavour Arg bonding)- specificity (measures by ratio of kcat/Km for arg cleavage vs Lys cleavage) for Arg dropped over 10x, Lys became favoured.
Replacement of Asp (D) 102 w/Asn (N) (same shape but uncharged), expressed in rats-> wt enzyme purified from this was identical to natural rat enzyme; mutant had Km for specific substrates only 2x that of wild type, but kcat 2x10-4 that of wt, showing that the Asp residue was important for activity, through not essential. X-ray structure showed that Asp acts as H-bond donor to His57. NB Asp102 inaccessible to chemical modification so function unknown before this replacement study, but knew it may be important as present in all Ser proteases.
Immunoglobins
Antigen binding is a combined property of 2 light variable (VL) and heavy variable (VH) chains through interactions w/ complementarity determining regions- 3 hypervariable segments of total ~100aa. Treatment w/ proteinase papain liberates 2 monovalent antigen-binding fragments (Fab) by cleavage within the mobile region. The residual constant part is the Fc region, w/ no binding affinity to antigen. Since Fab binding properties= those of whole antibody, they’ve been used for molecular approaches to antibody engineering- easier w/ smaller proteins. Binding pockets are flat+ extensive for protein antigens, cleft-like for small molecule antigens.
Discrete globular domains containing an identical fold: aa seq+ X-ray structure of IgG from human myeloma-> pronounced domain structure. Basic motif= immunoglobin fold- 2 sheets of antiparallel beta strands, bridged by disulphide bond, ~110 aa. Found in C and V domains of Igs, T cell receptors, proteins of major histocompatibility complex.
How immunoglobins bind antigens and X-ray cryst evidence
How immunoglobins bind antigens: 3 hypervariable regions (loop out from Ig fold as seen on X-ray) on each of heavy and light V regions- CDRs. Loops from heavy+ light chains form antigen combining site. V and C domains of light chain folded into 2 separate globular units, w/ 4-stranded beta sheet in each domain, +CDRs at one end of the elongated molecule. In the Fab fragment (+ intact IgG), domains associate pairwise so Vh interacts Vl and Ch1 with Cl, so CDR on both V domains close together-> form binding site.
X-rays show CDRs of both H+L chains interact antigen similarly to those that stabilize E-S binding (charge neutralisation, vdW+ short-range interactions). Monoclonal antibodies-> easier to get homogeneous materials for Xray. E.g., IgGs and Fabs of those raised against hen lysozyme found to recognise discontinuous epitope on lysozyme residues 18-27+ 116-129. No evidence for conformational change on binding. Large contact surface area, with all CDRs (~17 aas) involved. Contact like that between 2 subunits of oligomeric enzyme w/ no H2O, many H-bonds, hydrophobic interactions and vdWs. Glutamine 121 of lysozyme goes fairly deep into antibody combining site- where this is changed to His, mutant no longer bound by antibody (reflects instability of a “buried” +ve charge).
Other cases have evidence for antigen-induced domain rearrangements, e.g., in complex of Fab fragment w/ flu enzyme neuraminidase, structure indicates 3A movements of V domains of both H and L chains upon binding enzyme+ small mvmts in antigen. Complex of Fab+HIV-1 peptide show some/ largest changes. Variability makes modelling attempts to predict CDR loop conformations less soundly based.
Human antibody engineering
Human antibody engineering- application of clinically significant mouse monoclonal antibodies problematic due to human immune response to mouse protein. Structural info about modularity of IgGs hels protein engineering circumvent this. Attepmts to humanise a mouse anti-lymphocyte monoclonal (CAMPATH-1): grafting mouse V domains onto human IgGs didn’t work; grafting mouse CDRs onto framework of human V domains-> nearly-human monoclonal now being used against cancers (lymphomas) resistant to other treatments, as well as rheumatoid arthritis.
Phage display and protein design
Phage display: modify genes for capsid proteins of filamentous bacteriophage (M13) by inserting extra DNA for aa residues appearing as a peptide/protein insert close to N terminus of native protein displayed on surface of virion. Can use minor coat protein p3 in a few copies at 1 end of virion (1 display at the end) or major coat protein p8 (2700 copies, bulk of tubular capsid)- multiple copies displayed, can look for multiple binding sites.
Create libraries by displaying random seqs. Select phages displaying peptide able to bind desired ligand by “panning” (ligand immobilised on surface, mixed w/ library, wash, elute+ amplify binding phage), re-screen. After 3-4 rounds, individual phage characterised- good method to obtain lead compounds for drug design/ selecting proteins w/ desired new properties.
Phase display of antibodies: finding which hybridomas produced a suitable antibody involves cumbersome screening. Phage display doesn’t rely on characterising individual clones by screening. Instead, select survivors by panning- more efficient, so good tool for antibody generation. Display either single-chain variable region fragments or Fab fragments- genes derived from human or animal/ constructed in vitro w/ synthetic V-gene repertoires. Antibodies against multiple targets can be joined by engineering- multivalency of constructs raised against multiple epitopes makes them bind more efficiently (“avidity”).
Antibody phage display can lead to commercial products, e.g., HUMIRA= antibody targeting tumour necrosis factor alpha (TNF- secreted mostly by monocytes/macrophages w/ effects on lipid metabolism, coagulation, insulin resistance, endothelial function). Medicine to alleviate pain in rheumatoid arthritis+ other diseases.
Alternative protein scaffolds for protein engineering
Alternative protein scaffolds increasingly explored- can be converted by directed evolution (e.g., by phage display) into binding proteins w/ KD values similar to antibodies but w/ different properties like stability, disulfide bonds, domain composition, interaction surface, immunogenicity, etc. Could be future alternative to antibody-derived protein binders if have useful properties, lower cost, longer shelf life and less restrictions (e.g., intellectual property restrictions).
Designer antibody enzymes and alternative approach via existing enzyme modification
Designer antibody enzymes: combines monoclonal antibody technology w/ expectation of enzyme tightly binding TS. Raise monoclonals against putative analogues of TS, test chemical reaction. E.g., hapten TS analogue used to raise antibodies for ester hydrolysis- antibodies raised against tetrahedral phosphorus species matching TS in charge and geometry were catalytic (up to 106x rate acceleration). Possible applications in treating drug addictions, but activity too low to be useful. So far, catalytic antibodies fail to match natural enzyme efficiency: many reactions involve 2+ TSs; enzymes also position functional groups that can move during a reaction+ recruit cofactors- chance of generating arrays capable of acting synergistically infinitesimal; binding rather than catalysis drives selection; Ig fold may place intrinsic limitations on catalytic process; TS analogues are at best poor mimics.
Alternative approach: adapt existing enzymes. Start w/ DNA library; randomise gene by error-prone PCR, clone and grow transformed cells; cell lysis, screen proteins/lysates under desired conditions, pick colonies, isolate DNA, return to randomisation step and repeat. E.g., B subtilis nitrobenzyl esterase used in industry, has low activity at higher temperatures- procedure above done w/ screen under high temperatures. Several cycles-> esterase with 100-fold higher activity and 17 degree higher denaturation (better thermostability)- this due to new salt bridges and H bonds+ better packing of secondary structure elements in loose, unstructured regions. Most mutations not in/near active site, so couldn’t predict improvement- combinatorial approach complements design. Could also apply this technique to adapt enzyme to different pH/solvent/substrate etc
For directed evolution approach, improvement gradual+ only small fraction of sequence space sampled- could never screen all possibilities- makes finding new activities difficult.
T4 bacteriophage lysozyme protein engineering
Greater conformational entropy= factor favouring unfolded state. Folded state stabilised by hydrophobic effect- weak interactions, only stabilise when many interactions cooperate+ only contribute intramolecularly. Typical total free energy small (most globular protein free energy of denaturation 20-60kJ/mol- equivalent of a couple H bonds). Comparison w/ thermostable enzymes from thermophilic bacteria yields no simple rules to increase heat stability.
T4 lysozyme has 164 aa, 2 domains, no S-S bridges. Xray straightforward so many mutants directly characterised. Each aa mutated systematically-> 85% mutants active, over ½ positions tolerated any substitution, insertions tolerated if amphipathic pattern followed. Replacing bulky hydrophobic groups in core w/ Ala yielded some unstable (void produced), some stable (repacking fills space) structures. Conclude that many residues don’t contribute much to stability.
Adding disulfide bridges helped increase stability: look at structures, measure S-S distances, find similar distances in query structure and introduce S-S bonds. Three new bridges introduced+ Tm of each mutant recorded for both oxidised S-S and reduced SH forms-> S-S forms more stable, increase in Tm additive, reduced forms were less stable than wild type (so new S-S bridges also involved loss of some favourable interactions).
Stabilising dipoles of alpha-helices improves stability: T4 lysozyme already had -ve residues near N terminus in 7/11 helices. Introduce -ve residues to 2 more helices-> more stable, effects additive, Xray shows electrostatic effect as new residue not in bonding distance to main-chain amide groups. (double mutant Tm 4o higher).
Compare and contract Design and Library approaches to protein engineering
Protein folding and stability: Anfinsen’s experiments and Levinthal’s paradox
Anfinsen’s experiments on bovine pancreatic ribonuclease showed:
* S-S bonds reduced+ treat w/ 6M urea-> lose all activity, structure near to random coil
* Re-oxidise in absence of urea-> regain all original activity
* Re-oxidise in urea, then remove urea-> activity v low (statistically expected). If catalytic amount of reducing agent added, rapidly revert to native structure w/ correct S-S bridges
Conclusion: all info for folding is in aa seq. shown by convergence to native structure from many starting conformations. Native structure= lowest free energy (likely global free energy minimum), though folding not random search for all conformations: Levinthal’s paradox: assuming even very rapid sampling of each conformation, would take 1048 years to test each conformation of a 150 aa seq: there is a folding funnel with guiding intermediates.
Protein folding: folding funnel and enzyme catalysis
Folding funnel: unfolded->entropic effects (rapid)-> molten globule ( helices and beta strands start to form- final 2o structures present, reducing #conformations to sample)- local E minimum; energetic effects slower, go through TS (E max)-> folded state (global E min).
Enzymes catalysing slow steps of folding: protein disulphide isomerase (shuffles S-S links- essential gene in yeast); prolyl peptide isomerases (cis-trans isomerisation of many Xaa-Pro bonds, lower activation barrier by ~67kJ/mol)- identical to immunophilins (intracellular targets of potent immunosuppressant drugs)
