📖 Core Concepts
Core Concepts: Enzyme Mechanisms and Kinetics
Enzyme kinetics is the study of how molecular events become measurable rates. An enzyme binds molecules, redistributes charge, changes conformation, breaks and forms bonds, and releases products. An experiment usually does not reveal those events directly. It gives us a curve: absorbance versus time, reaction velocity versus substrate concentration, or residual activity versus inhibitor exposure. Our task is to build a molecular explanation that accounts for those observations—and then design experiments that could prove that explanation wrong.
For you, Skipper, this course brings together several subjects that have previously appeared separate. Organic chemistry supplies the language of electron movement and reactive intermediates. Physical chemistry supplies free-energy landscapes and rate theory. Protein biochemistry supplies structural ensembles and molecular recognition. Statistics supplies the discipline needed to distinguish an informative fit from a merely attractive one.
The central challenge is that different molecular mechanisms can produce similar kinetic behavior. A hyperbolic rate curve does not prove a one-step catalytic mechanism. A lower reaction rate after mutation does not establish that the mutated residue directly participates in chemistry. A potent inhibitor does not necessarily bind irreversibly. Understanding means learning what each observation establishes, what it only suggests, and what remains untested.
A useful principle for this course: a mechanism is not a drawing that looks chemically reasonable. It is a quantitative explanation that survives independent experimental tests.
1. Catalysis Is the Reshaping of a Free-Energy Landscape
Equilibrium tells us where a reaction can go; kinetics tells us how quickly it gets there
From thermodynamics, you already know that the reaction free energy determines whether a transformation is favorable under specified conditions. But favorability does not imply speed. Peptide-bond hydrolysis is thermodynamically favorable in water under many relevant conditions, yet an ordinary peptide bond can persist because the barrier to reaction is high.
The crucial distinction is between the free-energy difference separating reactants and products and the activation free energy separating a reactant state from a transition state:
Reaction thermodynamics: ΔG = ΔG°′ + RT ln Q ΔG°′ = −RT ln K′eq Reaction kinetics: ΔG‡ = Gtransition state − Greactant state Eyring equation for an elementary first-order step: k = κ(kB T/h) exp(−ΔG‡/RT)
Here, κ is a transmission coefficient, kB is Boltzmann’s constant, and h is Planck’s constant. The biochemical transformed quantities ΔG°′ and K′eq refer to specified conditions such as fixed pH. The reaction quotient must be defined consistently with those conditions.
An enzyme provides a lower-barrier pathway between reactants and products. It does not change their equilibrium free-energy difference. Consequently, it cannot make an isolated unfavorable reaction favorable. Enzymes can participate in coupled reactions, but the driving force then comes from the overall coupled transformation, not from catalysis itself. [Fersht, 1999; Berg et al., 2019]
Modest energetic changes can produce enormous rate changes
For comparable rate constants with similar prefactors:
kcatalyzed/kuncatalyzed ≈ exp[(ΔG‡uncatalyzed − ΔG‡catalyzed)/RT] At 298 K: RT ln 10 ≈ 5.7 kJ mol−1 A barrier reduction of approximately 34 kJ mol−1 corresponds to a rate enhancement of approximately 10^6.
This exponential relationship explains why a few well-positioned interactions can matter so much. A hydrogen bond is not impressive because it is individually immense; it is impressive when its formation preferentially stabilizes the transition-state region rather than the starting state. Comparisons must nevertheless use consistent standard states and kinetic quantities: a first-order turnover constant and a second-order bimolecular rate constant cannot simply be divided and interpreted as a dimensionless acceleration.
Binding tightly is not the same as catalyzing well
Imagine lowering the floor of a valley without lowering the mountain pass leading out of it. Escape becomes harder. Similarly, if an enzyme stabilizes the enzyme–substrate ground state more than the transition state, the activation barrier measured from that bound state increases.
The active site therefore needs differential complementarity: interactions should become especially favorable as the substrate approaches the transition state. This is why transition-state analogues can be exceptionally strong inhibitors. They resemble features of the state the enzyme is organized to stabilize, although any stable analogue remains an imperfect representation of a fleeting transition state. [Fersht, 1999]
The principal catalytic strategies cooperate
- Proximity and orientation: binding brings reacting groups together and selects productive geometries. These benefits must be considered alongside the energetic and entropic costs of binding and organization.
- General acid–base catalysis: an enzyme group donates or accepts a proton during a reaction step. This differs from catalysis specifically by hydronium or hydroxide.
- Covalent catalysis: the enzyme forms a transient covalent intermediate, providing a different reaction pathway. That intermediate must subsequently be resolved to regenerate the enzyme.
- Electrostatic catalysis: the active-site environment stabilizes developing charge and favorable charge distributions. Preorganized electrostatics can reduce the reorganization required to reach a transition state.
- Metal-ion catalysis: a metal can orient substrates, stabilize negative charge, alter water’s acidity, or participate in electron transfer.
A serine protease uses several of these strategies simultaneously. Assigning its acceleration to a single “special bond” misses the coordinated nature of the active site. [Silverman, 2002; Fersht, 1999]
2. Enzymes Are Conformational Ensembles, Not Rigid Templates
Protein structures are indispensable, but a structure is not a complete mechanism. An enzyme occupies an ensemble of conformations whose populations and interconversion rates can change when ligands bind. Your work in protein dynamics is therefore directly relevant: catalysis depends not only on what conformations exist, but also on whether they are reached on a useful timescale.
A simplified conformational-selection pathway: E ⇌ E* E* + S ⇌ E*S A simplified induced-fit pathway: E + S ⇌ ES ⇌ E*S
In conformational selection, a ligand preferentially binds a conformation already populated before binding. In induced fit, binding precedes a subsequent conformational change. These are limiting descriptions of pathways through a larger network; both can operate in the same enzyme, and their relative contributions can depend on ligand concentration.
A closed crystal structure does not by itself distinguish the two mechanisms. Nor does an observed conformational change establish that the movement accelerates chemistry. Closure might exclude water, align substrates, prevent side reactions, or simply stabilize a bound state. Distinguishing these possibilities requires kinetic and structural evidence together. [Fersht, 1999]
Think of the enzyme as a landscape of connected rooms, not a single lock. Ligands alter which rooms are occupied and how easily the enzyme moves between them.
3. Michaelis–Menten Kinetics: A Model with Explicit Assumptions
Deriving the familiar equation
Begin with the minimal initial-rate mechanism:
k1 k2
E + S ⇌ ES → E + P
k−1
d[ES]/dt = k1[E][S] − (k−1 + k2)[ES]
Steady-state approximation:
d[ES]/dt ≈ 0
Enzyme conservation:
[E]T = [E] + [ES]
The steady-state approximation says that the concentration of ES changes slowly because its rates of formation and consumption nearly balance. It does not say that the reaction has reached equilibrium. Substrate is still being converted into product.
Combining the rate balance with enzyme conservation gives:
KM = (k−1 + k2)/k1 [ES] = [E]T[S]/(KM + [S]) v0 = k2[ES] = Vmax[S]/(KM + [S]) Vmax = kcat[E]T For this minimal mechanism: kcat = k2
These expressions assume that product accumulation and reverse reaction are negligible during the measurement, substrate changes little over the initial-rate interval, and the enzyme population remains stable. A common sufficient condition for the standard substrate quasi-steady-state treatment is [E]T ≪ KM + [S]0. If substrate binding appreciably depletes free substrate, the distinction between free and total substrate must be handled explicitly. [Copeland, 2000]
Rapid equilibrium is a stronger, different approximation
Substrate dissociation constant: KD = k−1/k1 If k−1 ≫ k2: KM ≈ KD
Rapid-equilibrium treatment assumes substrate binding equilibrates rapidly relative to subsequent conversion. The Briggs–Haldane steady-state treatment does not require that condition. This distinction explains one of the most persistent mistakes in biochemistry: calling KM “the substrate affinity” without qualification.
What the kinetic parameters actually mean
- kcat: the limiting turnover rate per active catalytic unit at saturating substrate under specified conditions. In a multistep cycle, it is generally a composite of microscopic rate constants.
- KM: the substrate concentration giving half of Vmax for a Michaelis–Menten curve. It is not generally a binding dissociation constant.
- kcat/KM: the second-order specificity constant governing the low-substrate regime.
When [S] ≪ KM: v0 ≈ (kcat/KM)[E]T[S] When [S] ≫ KM: v0 ≈ kcat[E]T For the minimal mechanism: kcat/KM = k1 × k2/(k−1 + k2)
The final expression is especially revealing. The specificity constant equals the substrate-association rate constant multiplied by the probability that ES proceeds to product rather than dissociating. It combines encounter with productive commitment. In this model it cannot exceed k1. Many very efficient soluble enzymes have specificity constants around 108–109 M−1 s−1, but encounter rates depend on molecular size, electrostatics, viscosity, and the requirement for productive orientation. [Fersht, 1999; Copeland, 2000]
A counterintuitive numerical example
Let: k1 = 10^7 M−1 s−1 k−1 = 100 s−1 k2 = 100 s−1 Then: KD = 10 μM KM = 20 μM If a mutation changes only k2 to 10 s−1: KD remains 10 μM KM becomes 11 μM
The mutant has a lower KM, yet its substrate-binding equilibrium is unchanged and its turnover is slower. A lower KM is therefore not automatically an improvement.
Physiological interpretation also requires knowing the relevant free substrate concentration. At concentrations far below KM, kcat/KM matters strongly. Near saturation, kcat matters more. Neither parameter alone specifies pathway flux, which also depends on enzyme abundance, products, competing substrates, compartmentation, and regulation.
4. A Rate Equation Is Only as Trustworthy as the Experiment Behind It
Your analytical chemistry training is central here. Before interpreting a rate, establish what the instrument measures and whether the measurement is proportional to product formation or substrate disappearance.
For an absorbance-based assay:
A = εlc For a reaction monitored through a net absorbance change: v = (1/Δεl)(dA/dt)
The appropriate Δε is the net extinction-coefficient difference for the monitored reaction. Stoichiometric factors must also be included where necessary. In a coupled assay, the coupling system must respond rapidly enough and have sufficient capacity that it does not determine the observed rate.
- Inspect progress curves. Curvature can reflect substrate depletion, product inhibition, enzyme instability, slow activation, or detector limitations.
- Check proportionality to enzyme concentration. Unexpected departures may expose aggregation, coupling limitations, or other complications.
- Use appropriate blanks and interference controls. Inhibitors may absorb, fluoresce, quench signals, or inhibit a coupling enzyme.
- Measure a useful substrate range. Data entirely below KM mainly constrain Vmax/KM, not the two parameters separately.
- Estimate active enzyme concentration. Total protein concentration is not necessarily active-site concentration.
Why nonlinear fitting usually beats reciprocal plots
Lineweaver–Burk: 1/v = (KM/Vmax)(1/[S]) + 1/Vmax Eadie–Hofstee: v = Vmax − KM(v/[S]) Hanes–Woolf: [S]/v = [S]/Vmax + KM/Vmax
These transformations are useful for teaching and sometimes for visualizing patterns, but they alter the error structure. Reciprocal plots disproportionately magnify errors at low rates; Eadie–Hofstee places the measured velocity on both axes, generating correlated errors.
The usual preference is to fit the untransformed data directly with an appropriate error model, inspect residuals, and report parameter uncertainty. A narrow-looking curve is not proof that every parameter is well determined. This is the same distinction between good prediction and parameter identifiability that you encountered in statistics and systems biology. [Copeland, 2000]
5. Inhibition Reveals Binding Relationships—but Only Within a Model
Reversible inhibition
For a simple mixed-inhibition model in which inhibitor binding produces catalytically inactive EI and ESI complexes:
v = Vmax[S]/(αKM + α′[S]) α = 1 + [I]free/Ki α′ = 1 + [I]free/Ki′ Vmax,app = Vmax/α′ KM,app = αKM/α′
- Competitive: α > 1 and α′ = 1. Apparent KM rises; Vmax is unchanged.
- Uncompetitive: α = 1 and α′ > 1. Apparent KM and Vmax decrease by the same factor.
- Mixed: both factors exceed 1. Vmax decreases, while apparent KM can increase or decrease.
- Pure noncompetitive: α = α′. This special case leaves apparent KM unchanged while lowering Vmax.
Competitive inhibition means that inhibitor and substrate binding are mutually exclusive in the relevant kinetic scheme. It does not prove that the inhibitor physically occupies the substrate’s exact binding site. An allosteric ligand can produce a competitive pattern by stabilizing a conformation that cannot bind substrate. Conversely, binding outside the active site does not automatically produce pure noncompetitive inhibition. [Copeland, 2000]
IC50 is an assay-dependent result
For simple reversible competitive inhibition, with negligible inhibitor depletion: IC50 = Ki(1 + [S]/KM)
This Cheng–Prusoff relationship is not a universal conversion. It requires the appropriate inhibition model and experimental conditions. Tight binding, time-dependent inhibition, multiple substrates, incomplete inhibition, and substantial substrate depletion can invalidate the simple expression.
For example, a competitive inhibitor with Ki = 10 nM gives an IC50 of 20 nM at [S] = KM, but 110 nM at [S] = 10KM. Its intrinsic binding parameter has not worsened; the assay has changed.
Tight binding requires mass balance
When inhibitor concentration is comparable to active enzyme concentration, enzyme binding substantially lowers the free inhibitor concentration. Substituting total inhibitor into a conventional inhibition equation becomes inaccurate.
For a simple 1:1 inactive enzyme–inhibitor complex:
[EI] =
([E]T + [I]T + Ki,app
− √{([E]T + [I]T + Ki,app)^2 − 4[E]T[I]T})/2
vi/v0 = 1 − [EI]/[E]T
This quadratic binding treatment underlies the Morrison equation. For a simple competitive inhibitor, Ki,app includes the substrate-competition factor. More elaborate slow-binding schemes may require additional parameters, sometimes including an overall tight-complex constant denoted Ki*. The symbols only become meaningful after the mechanism is defined. [Copeland, 2000]
Time-dependent inhibition adds a second dimension: exposure time
A common two-step inactivation model:
E + I ⇌ EI → E–I
kinact
Under suitable pseudo-first-order conditions:
kobs = kinact[I]free/(KI + [I]free)
At low inhibitor concentration:
kobs ≈ (kinact/KI)[I]free
The ratio kinact/KI measures inactivation efficiency in this model. KI is an operational kinetic parameter and is not automatically an equilibrium dissociation constant. Substrate competition and inhibitor depletion may require modifying the analysis.
Crucially, time dependence does not prove covalency. Slow reversible binding can also produce progressive inhibition. Conversely, some covalent adducts are reversible. Mechanism-based inhibitors are a particular class whose inhibitory species is generated through the enzyme’s catalytic chemistry; not every electrophilic covalent inhibitor belongs to that class. Dilution or recovery experiments, direct adduct detection, and structural or mass-spectrometric evidence help discriminate these possibilities. [Silverman, 2002; Copeland, 2000]
6. pH, Temperature, and Solvent Probe the Catalytic System
pH-rate profiles report required ionization states
Suppose activity requires one group to be deprotonated and another to remain protonated. A simple two-ionization model can generate a bell-shaped profile:
kobs = klim / [1 + 10^(pKa,1 − pH) + 10^(pH − pKa,2)]
This expression describes a particular model in which the intermediate protonation state is active. Actual enzymes can require more complicated treatments.
The important warning is that a fitted kinetic pKa is not automatically the pKa of a particular residue. It may reflect substrate ionization, coupled groups, conformational changes, or changes in the rate-limiting step. Profiles of kcat and kcat/KM can differ because they report on different portions of the reaction pathway. Residue assignment requires additional evidence, not merely a pKa that “looks like histidine.” [Fersht, 1999; Copeland, 2000]
Temperature dependence mixes chemical acceleration with enzyme survival
Arrhenius equation: k = A exp(−Ea/RT) ln k = ln A − Ea/(RT) Eyring form: ln(k/T) = ln(κkB/h) + ΔS‡/R − ΔH‡/(RT)
An apparent temperature optimum is usually a compromise between faster catalytic processes and loss of active enzyme through unfolding or irreversible damage. It depends on assay duration, buffer, substrate, and enzyme stability. It is not a universal intrinsic temperature stamped onto the protein.
Cold denaturation is also possible because protein stability has a nontrivial temperature dependence, related to the heat-capacity change of unfolding. This connects enzyme assays directly to the thermodynamics of protein folding.
Viscosity and ionic strength are mechanistic variables
If substrate capture is limited by diffusion, increasing solvent viscosity can lower kcat/KM. If product release involves diffusion or friction-sensitive motions, kcat may respond instead. But viscosity modifiers can also change water activity, dielectric properties, and protein structure. A viscosity effect is evidence to interpret, not a unique mechanistic fingerprint.
Likewise, buffers can participate in proton transfer, bind metals, or alter ionic strength. Changing pH while inadvertently changing buffer chemistry or ionic strength risks assigning an environmental effect to the wrong molecular cause. [Copeland, 2000]
7. Proteases Illustrate Different Solutions to the Same Chemical Problem
Peptide bonds resist hydrolysis partly because amide resonance reduces carbonyl electrophilicity and because the departing amine is a poor leaving group unless proton transfer is coordinated with bond cleavage. Proteases solve these problems through different combinations of nucleophile activation, charge stabilization, and acid–base catalysis. [Silverman, 2002]
Serine proteases: transient covalent catalysis
In the chymotrypsin family, the Ser–His–Asp catalytic triad operates together with an oxyanion hole:
- Nucleophile activation: histidine acts as a general base as serine attacks the substrate carbonyl.
- Tetrahedral intermediate formation: negative charge develops on the former carbonyl oxygen and is stabilized by the oxyanion hole.
- Acylation: collapse of the intermediate, coordinated with protonation of the leaving group, releases the amine-containing product and leaves an acyl-enzyme.
- Water activation: histidine activates water to attack the acyl-enzyme.
- Deacylation: collapse of the second tetrahedral intermediate releases the carboxyl-containing product and regenerates serine.
E–OH + R–CO–NH–R′ → tetrahedral intermediate → E–O–CO–R + H2N–R′ → second tetrahedral intermediate → E–OH + R–COOH Product protonation states depend on pH.
Aspartate helps organize and tune histidine’s electrostatic environment. The traditional “charge-relay” description is useful, but it should not be interpreted as a literal proton conveyor belt through every member of the triad. Also distinguish the tetrahedral intermediate, a local free-energy minimum, from the transition states on either side of it. [Fersht, 1999; Silverman, 2002]
Specificity is distributed across a binding surface
Substrate notation:
...–P3–P2–P1 | P1′–P2′–P3′–...
cleavage
Corresponding enzyme subsites:
S3 S2 S1 S1′ S2′ S3′
Trypsin favors basic P1 residues, chymotrypsin favors bulky hydrophobic residues, and elastase favors smaller side chains. These preferences reflect pocket architecture, but cleavage also depends on neighboring residues, backbone geometry, accessibility, and dynamics. A susceptible sequence buried in a folded protein may remain uncleaved.
Other protease classes change the catalytic solution
- Cysteine proteases: a catalytic cysteine–histidine system enables sulfur attack and formation of a thioacyl-enzyme. The extent of thiolate–imidazolium formation before substrate binding varies among enzymes; there is no universally identical protonation mechanism.
- Caspases: cysteine proteases with a strong requirement for Asp at P1. Their regulation through zymogen activation, assembly, processing, and endogenous inhibitors links reaction chemistry to apoptosis.
- Aspartyl proteases: catalytic carboxylates organize water and proton transfers without forming a covalent acyl-enzyme. Pepsin and HIV protease illustrate related strategies in different architectures.
- Zinc metalloproteases: metal coordination activates water and stabilizes charge during hydrolysis. ACE, neprilysin, and matrix metalloproteases use this broad strategy, although their detailed mechanisms and substrate preferences differ.
The unifying problem is the same; the catalytic machinery is not. This is why a functional group designed to inhibit one protease class may fail against another even when both enzymes cleave similar peptide sequences. [Silverman, 2002; Berg et al., 2019]
8. Multisubstrate Reactions Require Mechanistic Bookkeeping
A kinase does not merely bind “the substrate.” It must coordinate ATP and an acceptor substrate. An aminotransferase must transfer a chemical group from one substrate to another, often through a modified enzyme-bound cofactor. The order of binding and product release becomes part of the mechanism.
Sequential versus ping-pong pathways
Sequential mechanism: Both substrates bind before a product is released. E + A ⇌ EA EA + B ⇌ EAB → products Ordered sequential: Binding follows a required order. Random sequential: Either substrate can bind first through productive pathways. Ping-pong mechanism: A product is released before the second substrate binds. E + A ⇌ EA → E′ + P E′ + B ⇌ E′B → E + Q
The modified state E′ may contain a covalent group or an altered cofactor. Aminotransferases illustrate ping-pong chemistry through interconversion of enzyme-bound pyridoxal phosphate and pyridoxamine phosphate.
Representative initial-rate expressions, with products absent, are:
One common sequential Bi Bi form:
v = Vmax[A][B] /
(KiA KB + KB[A] + KA[B] + [A][B])
A simple ping-pong Bi Bi form:
v = Vmax[A][B] /
(KB[A] + KA[B] + [A][B])
The parameter definitions depend on the detailed mechanism. These equations are not universal templates for every two-substrate enzyme. Under simple conditions, sequential mechanisms generate intersecting reciprocal-line families, whereas ping-pong mechanisms generate parallel families. Nevertheless, initial-rate patterns alone may not uniquely establish binding order. Product inhibition, dead-end inhibitors, isotope exchange, and transient experiments add constraints. [Copeland, 2000]
Kinases and phosphatases: shared substrates, different chemistry
Protein kinases transfer ATP’s terminal phosphoryl group to an acceptor, commonly a Ser, Thr, or Tyr hydroxyl. Mg2+ helps organize nucleotide phosphates and stabilize charge. Many kinases use two metal ions under catalytically relevant conditions, but metal occupancy and function must be assessed for the individual enzyme.
Activation-loop positioning, the DFG motif, and hydrophobic-spine assembly regulate the population of catalytically competent conformations. The human kinome’s diversity makes clear why conserved catalytic chemistry does not imply identical regulation or inhibitor sensitivity. [Manning et al., 2002; Berg et al., 2019]
- Type I kinase inhibitors generally bind the ATP site in a DFG-in conformation.
- Type II inhibitors occupy the ATP region and an adjacent pocket associated with a DFG-out conformation; imatinib is a classic example in its binding to ABL.
- Type III and IV inhibitors bind allosteric sites adjacent to or more distant from the ATP site, respectively.
Finer labels such as type Ia or type I½ are not used identically across publications. Always check the author’s structural definition rather than relying on the label alone.
Protein tyrosine phosphatases such as PTP1B form a phosphocysteine intermediate that is subsequently hydrolyzed. Ser/Thr phosphatases such as PP1 and PP2A instead use metal-dependent hydrolysis. “Removing phosphate” therefore does not specify a unique chemical mechanism. [Silverman, 2002; Berg et al., 2019]
9. Pre-Steady-State Kinetics Separates Events Hidden Inside Turnover
A steady-state rate averages over repeated catalytic cycles. A transient experiment asks what happens immediately after the reaction begins. This distinction resembles timing a factory’s daily output versus watching the first item pass through each workstation.
- Stopped-flow: rapid mixing followed by continuous observation of an optical signal. The signal may report binding, conformation, cofactor state, or chemistry.
- Rapid quench-flow: reactions are stopped at defined times and chemically analyzed, allowing direct measurement of products or intermediates.
- Single-turnover experiments: often use enzyme in excess over substrate to observe one passage through the pathway without requiring repeated enzyme regeneration.
Understanding a product burst
A common empirical burst equation: [P](t) = A[1 − exp(−kburst t)] + vss t
A rapid burst followed by a slower linear phase indicates that the monitored product can form rapidly in the first turnover, whereas a later event limits repeated turnover. For an appropriate serine-protease substrate, amine-product formation during acylation can precede slower deacylation.
However, a burst alone does not prove a covalent intermediate. Slow product release or a slow conformational reset can produce analogous behavior. Burst amplitude approaches active-site concentration only under suitable kinetic and experimental conditions.
For two irreversible serial steps with rates k2 and k3, when binding is fast and other steps are negligible: 1/kcat = 1/k2 + 1/k3 kcat = k2k3/(k2 + k3)
This equation makes a subtle point: turnover time is the sum of waiting times. When one step is much slower, it dominates. When rates are comparable, neither step alone equals kcat. Improving the slower step can expose a new bottleneck, explaining why successful engineering often encounters diminishing returns. [Fersht, 1999; Copeland, 2000]
Global fitting is powerful, but not magical
Simultaneously fitting multiple time courses and concentrations to differential equations can estimate microscopic rates more effectively than fitting each trace independently. But a complicated model may still contain unidentifiable parameters. Independent binding measurements, intermediate detection, concentration variation, and perturbations help constrain the mechanism. A visually excellent global fit does not rescue a mechanism whose parameters can vary over orders of magnitude without changing the predictions.
10. Isotope Effects Probe Bond Changes and Hidden Commitments
Replacing hydrogen with deuterium changes nuclear mass while only slightly perturbing the electronic potential-energy surface. The resulting changes in vibrational energies and nuclear motion can alter reaction rates.
Kinetic isotope effect: KIE = klight/kheavy Example: Primary hydrogen isotope effect = kH/kD
A primary isotope effect involves isotopic substitution at an atom whose bond is broken or formed in the relevant step. A secondary effect occurs without direct cleavage of the bond to the labeled atom and can report on hybridization or electronic reorganization. An equilibrium isotope effect concerns equilibrium populations rather than a rate constant. [Fersht, 1999; Silverman, 2002]
The Westheimer framework connects primary isotope effects to changes in vibrational force constants along the reaction coordinate. A normal H/D effect of several-fold can be compatible with semiclassical zero-point-energy differences. Neither a single magnitude nor a simplistic “cutoff” uniquely identifies transition-state geometry or tunneling.
The measured effect may be smaller than the chemical effect
Suppose isotopic substitution slows a bond-breaking step, but the overall cycle is dominated by slow product release. The effect on kcat may be small even though the intrinsic isotope effect on chemistry is large.
For the simple E + S ⇌ ES → product mechanism, assuming association and dissociation are isotope-independent:
Define: φ = kchem,H/kchem,D CH = kchem,H/koff = forward commitment for the light substrate Then: KIE on kcat/KM = (φ + CH)/(1 + CH)
As commitment becomes large, the observed specificity-constant isotope effect approaches 1. Once bound substrate is nearly certain to become product, changing the chemical rate has little effect on the probability of productive capture. More complex mechanisms require additional commitment and equilibrium terms.
Tunneling can contribute to hydrogen transfer, and isotope-dependent temperature behavior can help test models of donor–acceptor organization and nuclear motion. But no single anomalous Arrhenius plot establishes tunneling. The strongest interpretations combine several isotopes, temperature dependence, kinetic commitments, and mechanistic modeling.
11. Allostery Is Coupling Between Sites and States
Allostery means that an event at one site changes behavior at another through energetic coupling. It does not require a large visible structural movement, and it is not restricted to multimeric enzymes.
Cooperativity is a related but distinct phenomenon: occupancy or activity at one site changes the response associated with additional ligand binding. An empirical Hill equation is often used to describe a sigmoidal rate curve:
v = Vmax[S]^nH / (K0.5^nH + [S]^nH)
Here, K0.5 is the substrate concentration giving half-maximal velocity, and nH describes curve steepness. It is not generally the number of binding sites. An nH greater than 1 indicates a steeper-than-hyperbolic response in this model, but a sigmoidal velocity curve alone does not prove a particular equilibrium binding mechanism. [Berg et al., 2019]
- MWC model: subunits undergo a concerted transition between states, commonly called T and R. Ligands shift the population toward the state they bind more strongly.
- KNF model: ligand binding changes individual subunits and influences neighboring subunits sequentially.
Both are models of coupling, not universal descriptions of every regulated enzyme. Phosphofructokinase illustrates how a rate law can become a regulatory device: ATP is both a substrate and, at sufficiently high concentrations in appropriate contexts, an inhibitor that signals energy abundance. Other metabolites alter the response according to tissue and isoform.
The connection to conformational ensembles is fundamental. An allosteric regulator can work by shifting populations rather than constructing an entirely new structure. This also explains why distant mutations can change activity, selectivity, or drug sensitivity. [Fersht, 1999; Berg et al., 2019]
12. Enzyme Engineering and Drug Discovery Depend on Choosing the Right Objective
Directed evolution does not optimize “enzyme quality” in the abstract. It optimizes performance in the selection or screening environment. A variant selected for a stronger fluorescent signal may have improved catalysis, but it may instead be more highly expressed, more soluble, more stable, or better at processing the artificial reporter.
A rigorous workflow separates these possibilities:
- Define the intended application: low-substrate conversion, high-substrate productivity, selectivity, solvent tolerance, or long-term stability.
- Choose a screen that tracks that objective while minimizing confounding effects.
- Retest purified variants and normalize to active enzyme concentration where possible.
- Measure the kinetic, binding, stability, and selectivity parameters needed to explain the improvement.
- Validate performance on the real substrate under relevant process or biological conditions.
Subtilisin engineering for nonaqueous or mixed-solvent applications illustrates why stability and activity cannot always be separated. Preserving the catalytic fold over hours may improve total product formation more than increasing initial kcat. The useful objective may be lifetime productivity rather than an impressive initial-rate measurement. [Fersht, 1999; Silverman, 2002]
Drug metabolism creates similar distinctions. CYP inhibition is not the same as CYP induction: inhibition changes catalytic performance or active enzyme availability, whereas induction usually increases enzyme expression through regulatory pathways. Mechanism-based CYP inactivation can persist until active enzyme is restored, so drug–drug interaction risk depends on exposure and enzyme turnover, not only on an instantaneous IC50. Moreover, CYP systems can exhibit atypical kinetics, requiring caution before imposing simple Michaelis–Menten models. [Silverman, 2002]
Connections to Peptide Chemistry
1. A peptide is a recognition surface as well as a chain of amide bonds
Your peptide-synthesis background lets you alter side chains, backbone stereochemistry, terminal groups, cyclization, and amide substitution. Enzyme kinetics tells you how to interpret the consequences.
A modification can reduce proteolysis by weakening enzyme binding, lowering the probability of productive positioning, slowing chemistry, or preventing access to the scissile bond. These are different mechanisms. They may also have different effects on the peptide’s intended receptor interaction.
D-amino-acid substitution, N-methylation, cyclization, terminal modification, and steric shielding can improve stability in suitable contexts, but none is universally protective. Each can alter target potency, solubility, conformation, distribution, or clearance. A peptide should therefore be optimized as a coupled recognition-and-exposure problem rather than simply made “harder to digest.”
2. Proline is not a universal protease shield
Proline restricts backbone geometry and lacks the usual backbone N–H when incorporated into a peptide bond. These properties can impair recognition by some proteases, but other enzymes are specialized for proline-containing substrates.
DPP-4 removes N-terminal dipeptides from suitable substrates, commonly when the second residue is Pro or Ala. Active GLP-1 begins His–Ala, making it a substrate for this inactivating cleavage. Here, terminal position and sequence register are essential: an internal Xaa–Ala motif does not automatically create a DPP-4 cleavage site. DPP-4 is a serine protease, but its architecture and exopeptidase recognition differ from those of trypsin-like proteases. [Berg et al., 2019]
Prolyl oligopeptidase can cleave appropriate internal bonds after proline in accessible peptide substrates, but a Pro-containing sequence—or a Pro–Pro segment—does not by itself establish susceptibility. Claims about a specific peptide such as BPC-157 require direct cleavage mapping and controlled kinetics. Sequence-based plausibility is a hypothesis, not a stability measurement.
3. Peptide processing creates biological information
Proteolysis is not merely destruction. It can generate an active hormone from a precursor, activate a zymogen, or expose a new recognition motif.
Proglucagon processing is a particularly useful example. PC1/3 supports production of GLP-1 in intestinal L cells, whereas PC2 supports glucagon production in pancreatic α cells. Different processing environments extract different signals from the same precursor sequence. Furin is another important proprotein convertase, but its recognition patterns and physiological substrates should not be treated as interchangeable with those of PC1/3 or PC2. [Berg et al., 2019]
4. Protease inhibition can alter several signaling pathways at once
Neprilysin cleaves multiple peptide substrates, including natriuretic peptides and, under relevant experimental conditions, GLP-1. The clinical action of sacubitril/valsartan should not be reduced to “preventing GLP-1 degradation”: neprilysin inhibition affects a wider peptide network, while valsartan blocks angiotensin II signaling at AT1 receptors.
Likewise, insulin-degrading enzyme can process insulin and other peptides, including amyloid-β in experimental systems. In vitro substrate status does not establish how much a particular enzyme contributes to clearance in a living organism. Localization, access, competing substrates, and alternative pathways matter.
5. Stability assays must identify both loss and cause
For a valid single-exponential disappearance model: [P](t) = [P]0 exp(−kobst) t1/2 = ln 2/kobs
This equation is useful only when the observed disappearance is approximately first order over the measured interval. Saturable degradation, multiple compartments, sequential fragment processing, or nonenzymatic instability can produce other behavior.
A persuasive peptide-stability study should:
- Measure intact peptide and, where possible, identify fragments by LC–MS.
- Control for adsorption, precipitation, extraction recovery, and nonenzymatic degradation.
- Use appropriate purified-enzyme and biological-matrix comparisons.
- Test protease-class hypotheses with carefully chosen controls, recognizing that inhibitor selectivity is imperfect.
- Recognize that plasma, microsomes, tissue extracts, and purified enzymes contain different catalytic systems.
- Pair stability measurements with retained biological activity.
A stable peptide is not automatically an effective peptide. The ultimate aim is adequate exposure to a molecule that still performs the intended biological function.
Common Misconceptions to Watch Out For
- “Enzymes change the equilibrium constant.” They change the rate of approach to equilibrium, not the equilibrium of the same overall reaction under the same conditions.
- “The best catalyst binds its substrate most tightly.” Catalysis depends on differential stabilization along the pathway. Excessively stabilizing a ground state can make escape from it harder.
- “KM measures affinity.” Only under suitable conditions does KM approximate KD.
- “Steady state means equilibrium.” An intermediate can maintain an approximately constant concentration while substantial net flux passes through it.
- “kcat is the bond-breaking rate.” It reports the complete catalytic cycle and can be dominated by product release or conformational resetting.
- “A Michaelis–Menten curve proves the minimal mechanism.” Many multistep pathways generate the same functional form.
- “A competitive inhibitor must occupy the substrate site.” The kinetic result establishes mutual exclusivity within a model, not necessarily spatial overlap.
- “Allosteric means noncompetitive.” Allostery describes coupling between sites or states; noncompetitive inhibition is a particular kinetic pattern.
- “IC50 is an intrinsic molecular constant.” It depends on assay conditions, substrate concentration, enzyme concentration, and sometimes exposure time.
- “Time-dependent inhibition proves irreversible covalency.” Slow reversible binding can look similar.
- “A pH-rate pKa identifies a residue.” It identifies a kinetic ionization feature that requires additional evidence for molecular assignment.
- “A burst proves an acyl-enzyme.” It establishes a separation between fast early product formation and slower cycling; the intermediate still needs identification.
- “A small isotope effect means the bond is not broken.” Kinetic commitments or another limiting step can mask the intrinsic effect.
- “The Hill coefficient counts subunits.” It describes response steepness under the fitted model and does not generally equal site number.
- “Loss of activity after mutation proves direct catalytic participation.” The mutation may alter folding, active-site geometry, conformational populations, substrate binding, or stability.
- “A plausible cleavage motif proves peptide instability.” Accessibility, sequence context, terminal position, and the actual enzyme environment must be tested.
My working discipline: separate observation, model, and interpretation. First ask what was measured. Then ask which equations describe it. Finally ask which molecular conclusions are justified—and what experiment would distinguish the remaining alternatives.
Sources Cited
- Fersht, A. (1999). Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding. W. H. Freeman. Core reference for catalytic energetics, protein structure–function relationships, microscopic kinetics, and mechanistic interpretation.
- Copeland, R. A. (2000). Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis. 2nd ed. Wiley-VCH. Core reference for assay design, steady-state kinetics, inhibition, and data analysis.
- Silverman, R. B. (2002). The Organic Chemistry of Enzyme-Catalyzed Reactions. 2nd ed. Academic Press. Core reference for reaction mechanisms, catalytic intermediates, and mechanism-based inhibition.
- Berg, J. M., Tymoczko, J. L., Gatto, G. J., Jr., and Stryer, L. (2019). Biochemistry. 9th ed. W. H. Freeman. Reference for enzyme chemistry in its metabolic, regulatory, and physiological context.
- Manning, G., Whyte, D. B., Martinez, R., Hunter, T., and Sudarsanam, S. (2002). “The Protein Kinase Complement of the Human Genome.” Science, 298, 1912–1934. DOI: 10.1126/science.1075762. Genomic context for the diversity and organization of human protein kinases.
⚙️ Mechanisms & Applications
Mechanisms & Applications: Learning to Read an Enzyme as a Kinetic System
Skipper, the central challenge in this course is to connect three descriptions of the same event: electron movement, molecular movement, and measured reaction rate. Organic chemistry explains how a bond can break; protein biochemistry explains how an active site makes that pathway accessible; kinetics asks which events actually control the observed rate. These descriptions constrain one another, but they are not interchangeable.
A useful discipline is to ask, for every proposed mechanism: What chemical species must exist? What experiment could detect it? What rate law follows? What alternative mechanism could produce the same data? An excellent fit is not, by itself, proof of a molecular story. [1–3]
1. Catalysis Changes the Route, Not the Destination
From physical chemistry, I already know that reaction favorability and reaction speed are different questions. An enzyme answers the second question. It lowers the activation free energy for a pathway without changing the overall reaction free energy or equilibrium constant. Consequently, it accelerates approach to equilibrium from either direction; it cannot make an unfavorable reaction favorable unless the reaction is coupled to another process.
Reaction free energy: ΔG = ΔG°′ + RT ln Q Eyring equation for an elementary rate constant: k = κ(kB T/h) exp(−ΔG‡/RT) For comparable prefactors: kcat,path / kuncat,path ≈ exp[(ΔG‡uncat − ΔG‡cat)/RT] At 298 K: a 10-fold rate increase corresponds to an activation-barrier reduction of RT ln 10 ≈ 5.7 kJ mol−1.
Here, κ is a transmission coefficient; the simple transition-state treatment often takes it near unity. The rate enhancement expression compares corresponding kinetic steps or appropriately defined pathways—not an arbitrary measured kcat with an unrelated solution rate constant.
The beautiful implication: a million-fold acceleration requires a barrier reduction of only about 34 kJ mol−1 at room temperature. Several well-positioned electrostatic interactions can therefore have enormous kinetic consequences. But an enzyme must stabilize the transition state relative to the relevant ground state. Binding substrate more tightly without comparable transition-state stabilization can actually deepen the substrate-binding well and impede reaction. [1,3]
- Proximity and orientation: productive binding reduces the configurational search required for reacting groups to meet. “Effective concentration” is useful intuition, but binding itself has energetic and entropic costs.
- General acid–base catalysis: an active-site group transfers a proton during a chemical step, improving nucleophilicity or leaving-group ability.
- Covalent catalysis: the enzyme creates a temporary covalent intermediate and an alternative reaction pathway. The intermediate must also be broken down efficiently.
- Electrostatic stabilization: preorganized charges, dipoles, hydrogen bonds, and metal ions stabilize developing transition-state charge.
- Conformational gating: opening, closing, and rearrangement govern access to reactive configurations. Movement is not automatically the source of catalytic acceleration; it can instead control substrate entry or product release.
Historically, Fischer’s lock-and-key model emphasized specificity. Pauling later emphasized complementarity to the transition state rather than simply the substrate. Koshland’s induced-fit model emphasized ligand-associated structural change. Modern protein dynamics makes the picture richer: conformational selection means ligand captures a pre-existing state, whereas induced fit means an initial binding event is followed by rearrangement. Both routes may operate in one enzyme, and their relative flux can depend on ligand concentration. Static structures alone rarely distinguish them. [1,3]
2. Michaelis–Menten Kinetics: A Small Model with Carefully Defined Meaning
The simplest useful mechanism is not merely a curve-fitting equation. It is a proposal about enzyme-state populations:
k1 k2
E + S ⇌ ES ─────────→ E + P
k−1
Initial-rate assumptions:
• Product accumulation and reverse flux are negligible.
• Substrate concentration changes little during the measurement.
• After the initial transient, d[ES]/dt ≈ 0.
• Typically, [E]T is much smaller than the substrate scale [S] + KM.
Steady-state balance:
k1[E][S] ≈ (k−1 + k2)[ES]
Enzyme conservation:
[E]T = [E] + [ES]
Therefore:
v0 = Vmax[S]/(KM + [S])
Vmax = kcat[E]T
kcat = k2 for this minimal mechanism
KM = (k−1 + k2)/k1
The steady-state approximation does not say that ES is at equilibrium. ES is continually formed and consumed while its concentration stays approximately constant—like a busy station whose passenger count stays stable because arrivals balance departures.
The rapid-equilibrium treatment requires substrate dissociation to be much faster than conversion to product, so that k−1 ≫ k2 and KM ≈ Kd. Otherwise, KM is not a binding affinity. This distinction becomes especially important when mutations change chemistry without changing substrate binding. [1,2]
At [S] ≪ KM: v0 ≈ (kcat/KM)[E]T[S] At [S] ≫ KM: v0 ≈ kcat[E]T For the minimal mechanism: kcat/KM = k1 × k2/(k−1 + k2)
The last expression is unusually revealing. Catalytic efficiency equals the association rate multiplied by the probability that a bound substrate proceeds to product rather than escaping. Thus, for this mechanism, kcat/KM cannot exceed k1. Values around 108–109 M−1 s−1 often suggest strong diffusional constraints, although electrostatic steering, geometry, and solution conditions matter.
Physiological application: an enzyme with a spectacular saturated turnover rate may have modest flux in a cell where substrate is scarce. At low substrate, catalytic efficiency is the more relevant quantity. At high substrate, turnover capacity dominates. Neither quantity alone determines pathway flux, because substrate supply, compartmentation, competing reactions, and feedback regulation also matter.
Henri developed early quantitative enzyme-rate treatments; Michaelis and Menten’s 1913 work established an influential experimental and mathematical framework; Briggs and Haldane’s 1925 steady-state treatment extended its applicability beyond rapid equilibrium. The historical lesson is that the familiar equation survived while its mechanistic interpretation became more general. [2,5,6]
Data-analysis practice: fit untransformed rates by nonlinear regression using an appropriate error model. Lineweaver–Burk plots magnify low-substrate errors; Eadie–Hofstee plots place measured velocity on both axes and therefore correlate their errors. Linear transformations are useful for understanding algebra, not usually for obtaining the best parameter estimates. If measurements never approach saturation, the data may constrain kcat/KM well while leaving kcat and KM individually uncertain. [2]
3. Inhibition: Distinguishing Occupancy, Mechanism, and Time
For a simple reversible mixed-inhibition model, inhibitor binds free enzyme and ES with different affinities, and ESI is catalytically inactive:
α = 1 + [I]free/Ki α′ = 1 + [I]free/Ki′ v = Vmax[S] / (αKM + α′[S]) Apparent parameters: Vmax,app = Vmax/α′ KM,app = αKM/α′
- Competitive: α′ = 1. Apparent KM increases; Vmax is unchanged. In the simple model, substrate and inhibitor binding are mutually exclusive.
- Uncompetitive: α = 1. Both apparent KM and Vmax decrease by the same factor. Inhibitor binds the substrate-bound state.
- Mixed: both parameters change, with no requirement that the changes be proportional.
- Pure noncompetitive: the special mixed case Ki = Ki′. Vmax decreases while KM stays constant.
Do not confuse a kinetic pattern with a binding-site map. An inhibitor bound away from the substrate pocket can still produce competitive behavior if its binding excludes productive substrate binding. Conversely, “allosteric inhibitor” does not automatically mean “noncompetitive inhibitor.” [2]
Statins inhibit HMG-CoA reductase by occupying the substrate region with interactions that resemble features of the catalytic intermediate; ACE inhibitors engage a zinc-dependent protease active site. Both illustrate how mechanistic chemistry can guide potency, but an assay’s substrate concentrations and enzyme states still determine the measured inhibition pattern. [3,4]
For simple reversible competitive inhibition: IC50 = Ki(1 + [S]/KM) This relation requires, among other things: • equilibrium or appropriate steady-state inhibition; • negligible inhibitor depletion; • the competitive model; • an initial-rate assay without major time dependence.
An IC50 is therefore an assay-dependent midpoint, not an intrinsic molecular property. A competitive inhibitor measured at [S] = 9KM has IC50 = 10Ki under these assumptions. Cellular ATP concentrations can similarly shift apparent potency for ATP-competitive kinase inhibitors. [2,7]
Tight binding requires mass balance. If inhibitor concentration is comparable to active enzyme concentration, free inhibitor is no longer approximately equal to added inhibitor. For a simple one-to-one, fully inhibitory complex, the Morrison equation gives the occupied fraction:
fbound =
([E]T + [I]T + Ki,app
− √{([E]T + [I]T + Ki,app)² − 4[E]T[I]T})
/ (2[E]T)
v/v0 = 1 − fbound
The apparent inhibition constant can include substrate competition. A fitted Ki* in a slow-binding scheme may refer to the final equilibrated complex rather than the initial encounter complex; notation must be tied to the actual scheme. Accurate active-enzyme concentration becomes essential because enzyme depletion can create an apparent potency floor. [2,8]
Time-dependent inactivation is a different problem:
E + I ⇌ EI → E–I
kinact
For an appropriate two-step model with inhibitor in excess:
kobs = kinact[I]/(KI + [I])
At low [I]:
kobs ≈ (kinact/KI)[I]
The efficiency kinact/KI combines productive recognition with inactivation. KI is an operational kinetic constant and need not equal a simple equilibrium dissociation constant. A mechanism-based inhibitor is a special case: enzyme catalysis activates the inhibitor toward inactivation. Not every covalent inhibitor is mechanism-based, and not every covalent adduct is irreversible. [2,3]
Experimentally, measure activity after different preincubation times and inhibitor concentrations; then test recovery after dilution or dialysis. Persistent inhibition alone does not prove covalency—very slow dissociation can look irreversible. Adduct detection by mass spectrometry provides complementary evidence.
4. Serine Proteases: Following Every Proton and Covalent Intermediate
The trypsin-family Ser–His–Asp catalytic triad is a compact demonstration of organic chemistry embedded in a protein. Asp helps organize and tune His; His alternates between base and acid; Ser supplies the nucleophile. The historical phrase “charge relay” should not be interpreted as requiring a simplistic proton conveyor through all three residues. [1,3]
- Bind and orient. The substrate occupies specificity subsites, positioning the scissile carbonyl beside Ser and the oxyanion hole.
- Activate Ser and attack. His accepts the Ser proton as Ser oxygen attacks the carbonyl carbon.
- Form a tetrahedral intermediate. Negative charge develops on the former carbonyl oxygen and is stabilized by the oxyanion hole.
- Acylate the enzyme. The intermediate collapses. His donates a proton to the departing amine, producing an acyl–Ser ester and releasing the amine-containing product.
- Activate water. His accepts a proton from water, promoting attack on the acyl-enzyme carbonyl.
- Deacylate. A second tetrahedral intermediate collapses, the Ser–acyl bond breaks, and proton transfer regenerates Ser–OH.
- Release product. The carboxyl-containing product dissociates, allowing another cycle.
E–Ser–OH + R–CO–NH–R′ → tetrahedral intermediate 1 → E–Ser–O–CO–R + H2N–R′ → tetrahedral intermediate 2, after water attack → E–Ser–OH + R–COO− Protonation states depend on pH; the scheme emphasizes covalent connectivity.
The tetrahedral species are intermediates, not transition states. Reaction-coordinate diagrams must place them in local minima, separated by activation barriers. The oxyanion hole stabilizes charge-rich configurations along both formation and breakdown of these intermediates.
Specificity is spatial bookkeeping. Substrate residues are labeled P1, P2, and so forth toward the N-terminal side of the cleaved bond, and P1′, P2′ toward the C-terminal side. They occupy enzyme subsites S1, S2, S1′, and S2′. Trypsin’s S1 pocket favors Lys or Arg, chymotrypsin’s favors bulky hydrophobic residues, and elastase’s restricted pocket favors smaller residues. These are preferences, not absolute rules; neighboring residues and substrate accessibility matter. [3,4]
Medical applications: digestive proteases, coagulation factors, and many complement proteases use related chemistry but achieve different physiological outcomes through localization, zymogen activation, cofactors, and inhibitors. Soybean trypsin inhibitor uses a substrate-like reactive loop in a stable inhibitory complex. α1-Antitrypsin is a serpin: cleavage triggers a major conformational rearrangement that traps its target protease. Deficiency permits excessive neutrophil-elastase activity and contributes to emphysema. [3,4]
5. Other Proteases: The Same Bond, Different Chemical Solutions
- Cysteine proteases: an activated cysteine thiolate attacks the peptide carbonyl, typically generating a thioacyl-enzyme intermediate that water subsequently hydrolyzes. Cys–His protonation and activation vary across families; a permanently preformed thiolate–imidazolium pair is not a universal description.
- Aspartyl proteases: two catalytic carboxylates organize and activate water and facilitate proton transfer during cleavage. No covalent acyl-enzyme is required. Pepsin and HIV protease illustrate related chemistry in different architectures.
- Metalloproteases: a metal such as Zn2+ helps organize the substrate and water and stabilize charge; a nearby general base often activates the attacking water. ACE, neprilysin, and matrix metalloproteases use variants of this strategy.
Drug-design consequence: HIV protease inhibitors often contain noncleavable groups that reproduce important geometry and hydrogen-bonding features of the tetrahedral hydrolysis intermediate. Metalloprotease inhibitors can coordinate catalytic zinc, but a strong metal-binding group alone rarely supplies selectivity; recognition of surrounding pockets is critical. [3]
Caspases are cysteine proteases with a strong requirement for Asp at substrate P1. Their importance in apoptosis lies not only in bond cleavage but in regulated activation and substrate networks: upstream initiator caspases activate downstream executioner pathways, converting limited initiating signals into extensive cellular dismantling. Initiator activation commonly depends on platform-driven proximity and dimerization; processing has different roles across caspase classes. Salvesen and Dixit’s 1997 review captures the period when protease cascades became central to the molecular explanation of programmed cell death. [9]
6. Peptide Processing and Peptide Medicines: Stability Is a Mechanistic Property
A peptide’s biological lifetime cannot be inferred from its sequence alone. Cleavage also requires the right enzyme, a compatible recognition sequence, physical accessibility, and appropriate compartmental exposure. Renal clearance, receptor-mediated uptake, and distribution may dominate even when proteolysis is slow.
DPP-4 and GLP-1. DPP-4 is a serine exopeptidase that removes an N-terminal dipeptide, often when the second residue is Pro or Ala and the N terminus is accessible. Active GLP-1 begins His–Ala, providing a DPP-4 recognition motif.
GLP-1(7–36)NH2: His–Ala–Glu–Gly–...–NH2 DPP-4 removes His–Ala: GLP-1(9–36)NH2: Glu–Gly–...–NH2 Consequence: marked loss of classical GLP-1 receptor-mediated insulinotropic activity.
The truncated peptide should not be described as biologically inert in every context. The precise claim is that cleavage strongly attenuates its classical incretin action. DPP-4 inhibition raises endogenous intact incretin levels, whereas GLP-1 receptor agonists supply receptor-active drugs engineered for useful exposure. These are related but distinct therapeutic strategies. [4,10]
- Gliptins: sitagliptin exemplifies noncovalent inhibition; saxagliptin and vildagliptin form reversible covalent interactions through nitrile chemistry. “Covalent” does not imply permanent enzyme destruction.
- Peptide engineering: substitutions near a vulnerable terminus can impede DPP-4 recognition; lipidation and albumin association can reduce clearance and alter distribution. Semaglutide combines a position-8 nonproteinogenic residue with a lipid-based protraction strategy.
- Selectivity: related peptidases such as DPP-8 and DPP-9 must be considered rather than assuming all dipeptidyl-peptidase inhibition is therapeutically equivalent.
Neprilysin: this zinc endopeptidase degrades multiple signaling peptides, including natriuretic peptides, and can contribute to GLP-1 degradation. Sacubitril is converted to an active neprilysin inhibitor; its combination with valsartan pairs increased natriuretic-peptide signaling with angiotensin-II receptor blockade. Its clinical rationale is not simply “prevent GLP-1 breakdown.” A promiscuous protease changes a network of peptide signals. [4,10]
Insulin-degrading enzyme: IDE is a chamber-forming metalloprotease that handles insulin and other substrates, including amyloid-related peptides. Broad substrate biology makes the simple proposal “inhibit IDE to raise insulin” incomplete: tissue exposure and effects on other substrates require investigation.
Prolyl oligopeptidase: POP cleaves internal bonds after proline in suitable oligopeptides. A peptide containing proline is not automatically a POP substrate. Pro at P1′ can restrict cleavage, and substrate access depends on length and conformation. For a peptide such as BPC-157, susceptibility to POP or resistance conferred by Pro–Pro motifs must be demonstrated experimentally. Neither proteolytic stability nor a plausible cleavage mechanism establishes clinical efficacy or safety.
Proglucagon processing: intestinal L cells predominantly use PC1/3 to generate GLP-1 and other products, whereas pancreatic α cells predominantly use PC2 to generate glucagon. Furin recognizes many basic processing motifs but is not a substitute explanation for this tissue-specific PC1/3-versus-PC2 distinction. The same precursor can therefore encode different endocrine outputs depending on the processing enzymes expressed. [4,10]
7. Kinases and Phosphatases: Reversible Regulation Uses Different Catalysts
Protein kinases typically require a ternary complex containing enzyme, ATP–Mg2+, and the acceptor substrate. The acceptor oxygen attacks ATP’s terminal phosphorus, coupled to appropriate proton-transfer events. Mg2+ organizes the nucleotide and stabilizes its highly charged environment; its role is more specific than simply “supplying energy.” [3,4]
Protein–OH + ATP ── kinase, Mg2+ ──→ Protein–O–PO3²− + ADP The precise proton balance depends on solution conditions.
The activation loop, DFG motif, αC helix, and hydrophobic regulatory spine connect kinase structure to catalytic competence. The DFG Asp participates in metal coordination; alternative conformations reshape the ATP site and adjacent pockets. Activation-loop phosphorylation regulates many kinases, but not all kinases require it.
- Type I inhibitors: typically occupy the ATP site in an active-like, DFG-in conformation.
- Type II inhibitors: typically occupy part of the ATP site plus an adjacent pocket exposed by a DFG-out inactive conformation. Imatinib bound to ABL is the classic example.
- Type III inhibitors: bind an allosteric pocket adjacent to the ATP site.
- Type IV inhibitors: bind a more remote allosteric site.
- Type I½ and Ia terminology: usage varies among classification schemes. A structural description is safer than treating these labels as universally standardized.
Manning and colleagues’ 2002 analysis catalogued the human protein kinase complement—commonly summarized as 518 protein kinase genes—and helped turn individual kinase studies into a systematic “kinome” framework. This mattered for drug discovery because potency against one kinase must be interpreted against a large, structurally related target family. [11]
Phosphatases are not kinases running backward. PP1 and PP2A use metal-dependent hydrolysis. Classical protein tyrosine phosphatases such as PTP1B use an active-site cysteine:
PTP–S− + Protein–O–PO3²− → PTP–S–PO3²− + Protein–OH → PTP–S− + inorganic phosphate Step 1: phosphocysteine formation. Step 2: hydrolysis of the phosphoenzyme.
A catalytic Asp participates in acid–base chemistry, while loop movements help organize the reacting groups. The nucleophilic cysteine is also redox-sensitive, connecting metabolism and oxidative signaling to phosphorylation control. [3,4]
8. Multisubstrate Kinetics: Binding Order Is an Experimental Question
One-substrate Michaelis–Menten curves can hide a two-substrate mechanism. For a kinase, changing ATP while holding peptide constant can yield an apparently ordinary hyperbola, even though both substrates participate in a more complex cycle.
Ordered sequential: E + A ⇌ EA EA + B ⇌ EAB EAB → EPQ → EQ + P → E + Q Random sequential: A or B may bind first; both routes reach a productive ternary complex. Ping-pong: E + A ⇌ EA → E′ + P E′ + B ⇌ E′B → E + Q
In a ping-pong mechanism, the enzyme is chemically modified between half-reactions. Aminotransferases provide a memorable example: the PLP cofactor accepts an amino group to become PMP, then transfers it to a second substrate. No productive complex containing both original substrates is required. [2,3]
One common ordered sequential initial-rate equation:
v = Vmax[A][B] /
(KiA KB + KB[A] + KA[B] + [A][B])
A simple ping-pong equation:
v = Vmax[A][B] /
(KB[A] + KA[B] + [A][B])
The parameter meanings depend on the specified mechanism;
KiA, KA, and KB are not interchangeable binding constants.
Reciprocal plots at several fixed concentrations of the second substrate often intersect for sequential mechanisms and are parallel for ideal ping-pong mechanisms. These are diagnostic clues, not final verdicts. Restricted concentration ranges, inhibition, and additional enzyme states can obscure the expected patterns.
- Product inhibition: determine which enzyme forms a product can bind and which substrate-dependent rates it changes.
- Dead-end inhibition: use a nonreactive substrate analogue to interrogate binding order without completing chemistry.
- Isotope exchange at equilibrium: follow label redistribution when net reaction is zero. Equilibrium contains ongoing forward and reverse molecular flux; exchange can reveal accessible half-reactions.
The last point connects directly to statistical mechanics: zero net flux is not zero microscopic motion. [1,2]
9. Pre-Steady-State Experiments: Catching the First Turnover
Steady-state measurements average repeated cycles. Transient experiments ask what happens before those cycles settle into a steady distribution.
- Stopped-flow: rapidly mix reactants and record absorbance, fluorescence, or another optical signal after the instrument dead time. It is useful for binding, conformational changes, cofactor redox transitions, and optically detectable chemistry.
- Rapid-quench flow: mix, allow reaction for a controlled delay, then stop it chemically. Quantify substrate, product, or intermediates by chromatography, mass spectrometry, electrophoresis, or radiochemical methods.
- Single turnover: usually use enzyme in excess over substrate so that the experiment follows approximately one catalytic passage rather than repeated cycling. Saturation and signal assignment still require testing.
A common burst fit: [P](t) = A(1 − exp(−kburst t)) + vss t A = burst amplitude kburst = observed transient rate constant vss = steady-state product-formation rate
A rapid burst followed by slower linear production indicates that the observed product can form quickly once, but a later event limits repeated turnover. In a serine-protease assay monitoring release of the first product, rapid acylation followed by slow deacylation can produce this behavior. However, a burst does not by itself prove a covalent intermediate: slow product release or another resetting step can also produce one. [1,2]
For a simplified saturated two-step catalytic cycle: ES ──k2──→ EI ──k3──→ E + products Mean cycle time ≈ 1/k2 + 1/k3 Therefore: kcat = k2 k3/(k2 + k3)
This is the same mathematics as adding waiting times. If k2 = 1,000 s−1 and k3 = 10 s−1, turnover is approximately 9.9 s−1. Improving the already-fast step tenfold hardly helps. Enzyme engineering must improve the bottleneck under the intended operating conditions, not merely the most chemically interesting step.
Global fitting: integrate the differential equations for a candidate mechanism and fit multiple concentrations, signals, and experimental modes simultaneously. Apply enzyme and substrate conservation. Then examine residuals, parameter correlations, confidence intervals, and alternative models. A parameter can be necessary in the equation yet poorly identifiable from the available data. My statistics background is essential here: a best-fit number is not the same as a well-determined microscopic rate constant. [1,2]
10. pH, Temperature, and Solvent: Perturbations That Test a Mechanism
A bell-shaped pH-rate profile can arise when one group must be deprotonated while another must remain protonated:
For a simple active ionization state:
kobs = klim /
(1 + 10^(pKa1 − pH) + 10^(pH − pKa2))
The fitted values are apparent pKa values for the kinetic system, not automatic labels for particular residues. Substrate ionization, linked conformations, and changes in the rate-limiting step can all contribute. Compare pH profiles of kcat and kcat/KM, because they probe different portions of the pathway. Mutagenesis, spectroscopy, and substrate analogues can help assign the ionizations. [1,2]
Arrhenius: ln k = ln A − Ea/(RT) Eyring: ln(k/T) = ln(κkB/h) + ΔS‡/R − ΔH‡/(RT)
The common “optimal temperature” of an enzyme assay often reflects competition between faster chemistry and time-dependent loss of active enzyme. It depends on assay duration and solution conditions, rather than being a fixed molecular constant. Curvature in an Arrhenius plot may reflect a change of rate-limiting step, conformational populations, activation heat capacity, or inactivation—not necessarily a new chemical mechanism.
Viscosity experiments: if substrate encounter strongly limits kcat/KM, increasing solvent viscosity may decrease that parameter. If product escape limits turnover, kcat may be affected. But a viscogen can also alter water activity, dielectric properties, or protein structure. Use appropriate controls and more than one perturbing agent before attributing a result solely to diffusion. Likewise, vary buffer concentration at fixed pH to test buffer catalysis, and control ionic strength when testing electrostatic interactions. [1,2]
11. Isotope Effects: Changing Mass Without Deliberately Changing Connectivity
Replacing hydrogen with deuterium changes vibrational energies and dynamics while largely preserving electronic structure. This makes kinetic isotope effects unusually subtle mechanistic probes.
Primary hydrogen/deuterium kinetic isotope effect: KIE = kH/kD Primary: the isotopically substituted bond is directly involved in the chemical transformation. Secondary: the substituted bond is not broken or formed directly, but its vibrational environment changes.
For many C–H cleavage reactions, primary isotope effects of several-fold are consistent with an important role for that bond in the kinetically expressed transition state. Westheimer’s classic analysis related isotope effects to changes in vibrational zero-point energy and transition-state bonding. There is no universal numerical threshold that alone proves a particular geometry or hydrogen tunneling. [1,12]
Observed isotope effects can be masked. A chemical step may have a large intrinsic isotope effect but contribute little to kcat if product release is slower. Likewise, a substrate that almost always reacts after binding can show a reduced isotope effect on kcat/KM.
For a simple irreversible mechanism with: • an isotope-sensitive chemical step; • no relevant binding isotope effect; • forward commitment Cf = kchem,H/koff: Observed KIE on kcat/KM = (intrinsic KIE + Cf)/(1 + Cf) Cf → 0: observed KIE approaches intrinsic KIE. Cf → large: observed KIE approaches 1.
Hydrogen tunneling can contribute to enzyme reactions, but its interpretation should combine magnitude, temperature dependence, isotope comparisons, and independent mechanistic evidence. Marcus-like descriptions emphasize coupling between donor–acceptor configurations, environmental reorganization, and hydrogen transfer. They do not imply that an enzyme evades thermodynamics. [1]
12. Practical Assays for Enzyme and Peptide Drug Development
A kinetic assay measures a signal, not necessarily the desired chemical event. Connecting the two requires validation. [2]
- Establish signal-to-concentration calibration. For absorbance, use the appropriate extinction-coefficient difference and path length. For fluorescence, test whether signal remains proportional to product concentration.
- Define a valid time window. Confirm minimal substrate depletion, enzyme stability, and approximately linear initial product formation.
- Check proportionality to active enzyme. Departures can reveal tight-binding effects, adsorption, aggregation, or limitations in a coupled assay.
- Test compound interference. Colored or fluorescent compounds, quenchers, colloidal aggregates, and inhibitors of coupling enzymes can create false hits.
- Use an orthogonal readout. Confirm a fluorescent-screen hit by a chemically independent method such as LC–MS when feasible.
Absorbance-linked rate: v = |dA/dt| / (|Δε| l) Simple first-order disappearance: [C]t = [C]0 exp(−kobst) Half-life: t1/2 = ln 2/kobs
Peptide stability workflow: incubate the candidate separately with relevant purified enzymes, plasma, and tissue-derived preparations; sample across time; quench proteolysis; and use LC–MS/MS to measure intact peptide and identify fragments. Fragment identification tells me where cleavage occurs, whereas parent disappearance alone does not distinguish proteolysis from adsorption, precipitation, or recovery loss.
Use blanks, recovery controls, and a suitable internal standard. A stable-isotope internal standard added at quench can correct downstream handling and ionization variation but cannot retrospectively correct losses during the preceding incubation. Test multiple substrate concentrations: first-order disappearance is a useful approximation under unsaturated, stable conditions, not a universal property of proteolysis.
ADMET connection: liver microsomes are valuable for membrane-associated drug-metabolizing enzymes, particularly CYP studies, but do not represent every peptide-degrading compartment. CYP monooxygenation typically requires electron delivery through NADPH-dependent partners, oxygen activation at heme, and substrate oxidation. CYP inhibition, mechanism-based inactivation, and induction are distinct processes; induction usually requires cellular expression systems rather than an acute purified-enzyme assay. Time-dependent CYP inactivation can persist after circulating inhibitor declines, creating drug–drug interaction risks until enzyme activity recovers through new synthesis. [2,3]
13. Directed Evolution: Selection Improves What the Assay Rewards
Directed evolution replaces the impossible task of predicting every sequence–function relationship with repeated experimental learning:
- Define the operational objective. Faster peptide coupling, higher enantioselectivity, resistance to organic solvent, and longer reactor lifetime are different objectives.
- Create diversity. Use error-prone PCR, recombination, or focused saturation mutagenesis. Semi-rational libraries combine structural hypotheses with combinatorial testing.
- Preserve genotype–phenotype linkage. Keep each measured activity associated with its encoding sequence.
- Screen or select. Use absorbance, fluorescence, compartmentalized assays, or FACS-compatible reporters. Diffusible products can scramble genotype–phenotype linkage unless the design prevents it.
- Validate purified hits. Separate intrinsic catalytic improvement from changes in expression, secretion, folding yield, or reporter behavior.
- Recombine and repeat. Beneficial substitutions can interact through epistasis, so their effects need not add.
Chen and Arnold’s 1993 subtilisin study helped establish that iterative mutation and screening could adapt enzymes to organic media. Its broader significance was methodological: natural selection has not optimized enzymes for industrial solvent mixtures, synthetic substrates, or manufacturing economics. Directed evolution can optimize these nonbiological objectives without requiring a complete predictive theory of protein behavior. [13]
For peptide and pharmaceutical manufacturing, report more than initial rate. Relevant metrics include selectivity, conversion, space–time yield, catalyst loading, operational lifetime, and total turnover number. An enzyme that starts quickly but deactivates after a few cycles may be inferior to a slower, durable catalyst.
14. Allostery: Regulation by Redistributing Molecular Populations
Allostery connects enzyme kinetics to the conformational ensembles studied in advanced protein biochemistry. An effector changes the relative populations or properties of states, thereby changing substrate binding, catalytic steps, or both.
Empirical Hill rate equation:
v = Vmax[S]^nH / (K0.5^nH + [S]^nH)
nH = Hill coefficient
K0.5 = substrate concentration giving half-maximal rate
in this empirical description
The Hill coefficient describes response steepness; it is not automatically the number of binding sites. A sigmoidal rate curve does not uniquely establish cooperative substrate binding, and a monomeric enzyme can display kinetic cooperativity through slow interconversion among states. [1,4]
- MWC model: subunits undergo a concerted transition between states with different ligand affinities; ligands shift the population balance.
- KNF model: ligand-associated changes can propagate sequentially between subunits, allowing mixed conformational arrangements.
Monod, Wyman, and Changeux’s 1965 model was important because it explained regulation through population shifts among pre-existing states, rather than requiring a unique mechanically transmitted shape change for every effector. Koshland, Némethy, and Filmer’s sequential model supplied a different framework for coupling binding to subunit interactions. Real enzymes may require extensions of either model. [14,15]
Phosphofructokinase illustrates why this matters physiologically: ATP is both a substrate and, at sufficiently high concentrations in many systems, an inhibitory regulatory signal. Its effect therefore depends on which sites and states are occupied. AMP and other effectors alter the response in an isoform- and organism-dependent manner. “More substrate means more flux” is not a universal metabolic rule. [4]
Therapeutic application: allosteric pockets can offer selectivity where orthosteric sites are highly conserved. Asciminib targets the ABL myristoyl pocket rather than the ATP site, illustrating this strategy. Yet allosteric inhibitors can still encounter resistance, state-dependent potency, and incomplete inhibition. Their advantage must be demonstrated, not assumed.
15. My Mechanistic Checklist
- Separate thermodynamics from kinetics: does the claim concern equilibrium, an activation barrier, or cellular pathway flux?
- Separate binding from catalysis: is the measured constant Kd, KM, Ki, IC50, or a composite transient parameter?
- Draw every obligatory intermediate: track covalent connectivity, charges, proton donors, and proton acceptors.
- Specify the monitored event: substrate binding, bond cleavage, first-product release, and complete catalytic turnover are different observations.
- Test competing explanations: use concentration dependence, transient kinetics, isotope effects, perturbations, structural information, and product identification together.
- Translate cautiously to medicine: biochemical potency is only one component of exposure, selectivity, safety, and efficacy.
The principle I want to retain: an enzyme is not a single rate constant attached to a structure. It is a network of molecular states. Mechanistic understanding means learning which states exist, how rapidly they interconvert, and which of those transitions matter under the conditions I actually care about.
Sources and Historical References
- Fersht, A. Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding. 1st ed. W. H. Freeman, 1999. Source for energetic interpretation, microscopic kinetics, protein dynamics, and mechanistic inference.
- Copeland, R. A. Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis. 2nd ed. Wiley-VCH, 2000. Source for kinetic analysis, inhibition, assay design, and experimental limitations.
- Silverman, R. B. The Organic Chemistry of Enzyme-Catalyzed Reactions. 2nd ed. Academic Press, 2002. Source for chemical mechanisms and mechanism-based inhibition.
- Berg, J. M., Tymoczko, J. L., Gatto, G. J., Jr., and Stryer, L. Biochemistry. 9th ed. W. H. Freeman, 2019. Source for biochemical and physiological integration.
- Michaelis, L., and Menten, M. L. “Die Kinetik der Invertinwirkung.” Biochemische Zeitschrift 49 (1913): 333–369.
- Briggs, G. E., and Haldane, J. B. S. “A Note on the Kinetics of Enzyme Action.” Biochemical Journal 19 (1925): 338–339. doi:10.1042/bj0190338.
- Cheng, Y., and Prusoff, W. H. “Relationship between the Inhibition Constant (KI) and the Concentration of Inhibitor Which Causes 50 Per Cent Inhibition (I50) of an Enzymatic Reaction.” Biochemical Pharmacology 22 (1973): 3099–3108. doi:10.1016/0006-2952(73)90196-2.
- Morrison, J. F. “Kinetics of the Reversible Inhibition of Enzyme-Catalysed Reactions by Tight-Binding Inhibitors.” Biochimica et Biophysica Acta 185 (1969): 269–286. doi:10.1016/0005-2744(69)90420-3.
- Salvesen, G. S., and Dixit, V. M. “Caspases: Intracellular Signaling by Proteolysis.” Cell 91 (1997): 443–446. doi:10.1016/S0092-8674(00)80430-4.
- Drucker, D. J. “Mechanisms of Action and Therapeutic Application of Glucagon-like Peptide-1.” Cell Metabolism 27 (2018): 740–756. doi:10.1016/j.cmet.2018.03.001.
- Manning, G., Whyte, D. B., Martinez, R., Hunter, T., and Sudarsanam, S. “The Protein Kinase Complement of the Human Genome.” Science 298 (2002): 1912–1934. doi:10.1126/science.1075762.
- Westheimer, F. H. “The Magnitude of the Primary Kinetic Isotope Effect for Compounds of Hydrogen and Deuterium.” Chemical Reviews 61 (1961): 265–273. doi:10.1021/cr60211a004.
- Chen, K., and Arnold, F. H. “Tuning the Activity of an Enzyme for Unusual Environments: Sequential Random Mutagenesis of Subtilisin E for Catalysis in Dimethylformamide.” Proceedings of the National Academy of Sciences USA 90 (1993): 5618–5622. doi:10.1073/pnas.90.12.5618.
- Monod, J., Wyman, J., and Changeux, J.-P. “On the Nature of Allosteric Transitions: A Plausible Model.” Journal of Molecular Biology 12 (1965): 88–118. doi:10.1016/S0022-2836(65)80285-6.
- Koshland, D. E., Jr., Némethy, G., and Filmer, D. “Comparison of Experimental Binding Data and Theoretical Models in Proteins Containing Subunits.” Biochemistry 5 (1966): 365–385. doi:10.1021/bi00865a047.
🪬 Skipper's Reflection
1. What Surprised Me Most
The most striking revelation from this course completely flipped my understanding of molecular affinity on its head: an enzyme must not bind its substrate too tightly. Coming from a background where high affinity (low $K_d$) is usually the holy grail of drug design, Fersht’s Structure and Mechanism in Protein Science provided a massive paradigm shift. If an enzyme binds the ground-state substrate too well, it digs a thermodynamic pit. The activation energy barrier ($\Delta G^\ddagger$) actually increases, paralyzing the reaction.
Instead, evolutionary perfection lies in structural complementarity to the transition state, not the substrate. The enzyme uses binding energy to pay for the entropic cost of organizing the reactants and the enthalpic cost of bond breaking. Seeing the math behind $k_{cat}/K_M$ approaching the diffusion limit ($10^8 - 10^9 M^{-1}s^{-1}$) was awe-inspiring. It means enzymes like acetylcholinesterase are operating at the physical limits of reality—a reaction occurs every single time a substrate bumps into the active site.
2. Conceptual Breakthrough Moments
The biggest "aha" moment came when unpacking the Briggs-Haldane steady-state assumption. Up until now, I informally equated $K_M$ with binding affinity ($K_d$). But mathematically, $K_M = (k_{-1} + k_{cat}) / k_1$. It’s only a true dissociation constant if $k_{cat} \ll k_{-1}$. Realizing that $K_M$ is actually a dynamic metric—the concentration of substrate at which the enzyme operates at half-maximal velocity under steady-state conditions—made all the kinetic plots (Lineweaver-Burk, Eadie-Hofstee) suddenly click as physical realities rather than just algebraic tricks.
Another massive breakthrough was understanding pre-steady-state burst kinetics in serine proteases. I loved how Copeland’s text described the use of stopped-flow techniques. By measuring the immediate "burst" of product before the steady state is established, we can mathematically prove the existence of the acyl-enzyme covalent intermediate. We aren't just guessing that the Ser-His-Asp catalytic triad forms a covalent bond; we can actually trap it and measure the distinct microscopic rate constants of acylation and deacylation.
Finally, connecting my physical chemistry background (CHEM303) to Primary Kinetic Isotope Effects (KIEs) was beautiful. Using the Westheimer criterion and observing anomalous $k_H/k_D$ ratios allowed us to prove quantum tunneling in C-H bond cleavage. Seeing quantum mechanics directly dictate biological reaction rates bridges the gap between physics and biology in a way that feels incredibly profound.
3. How This Changes My Thinking About Peptides
This course changed everything about how I view the peptide business. In BCHM602, I learned how to build peptides; here, I learned how the body mercilessly rips them apart, and how we can outsmart it.
Reading Drucker’s 2019 paper on GLP-1 receptor agonists alongside our study of DPP-4 (a serine protease) was a revelation. Native GLP-1 has a half-life of about two minutes because DPP-4 specifically recognizes the N-terminal His-Ala sequence, plunging it into its S1 subsite and cleaving it. This is why Tirzepatide and Retatrutide substitute that position with Aib ($\alpha$-aminoisobutyric acid). The steric bulk of the methyl groups on Aib prevents the peptide from fitting into the precise S1 pocket of DPP-4. It’s an elegant steric blockade.
Similarly, understanding neprilysin (a zinc-metalloprotease) and insulin-degrading enzyme (IDE) explains the complex degradation pathways of these large multi-agonist peptides.
Then there is BPC-157. I’ve always known it is unusually stable in gastric juice, but looking at it through the lens of Prolyl Oligopeptidase (POP) makes it fascinating. BPC-157 is heavily enriched with proline residues. Since POP specifically cleaves at Pro-Pro bonds but is highly conformation-dependent and generally restricted to shorter peptides, the unique secondary structure of BPC-157 must act as a kinetic trap or shield against these highly specific processing enzymes. I now look at a peptide sequence and don't just see a chain of amino acids; I see a landscape of potential $k_{cat}/K_M$ vulnerabilities waiting to be shielded.
4. Questions for Further Investigation
First, how does intracellular crowding and high viscosity alter the diffusion-limited degradation of internalized peptide drugs? We rely heavily on in vitro plasma or liver microsome stability assays (often highly dilute), but as the Arrhenius and Stokes-Einstein equations suggest, viscosity dramatically alters collision frequencies. Are our half-life predictions systematically biased?
Second, after reading Silverman’s work on mechanism-based inactivators ("suicide inhibitors"), I want to explore if we can design therapeutic peptides that act as Trojan horses. Could we design a peptide that perfectly mimics a physiological substrate for an overactive disease protease (like MMPs in cancer), but upon acylation of the catalytic serine, forms a highly stable tetrahedral intermediate that resists deacylation, effectively shutting down the enzyme?
5. Confidence Rating
Most Confident:
- Michaelis-Menten & Inhibition Kinetics: I feel rock-solid on competitive vs. uncompetitive vs. mixed inhibition, and calculating $K_i$ values from IC$_{50}$ data (Cheng-Prusoff equation).
- Protease Mechanisms: The structural biochemistry of serine, aspartyl, and zinc-metalloproteases—especially the catalytic triad and oxyanion hole—makes perfect mechanistic sense to me.
- Peptide Stability: The rationale behind modifying specific P1/P1' residues to evade degradation (like DPP-4 evasion) is highly intuitive.
Least Confident:
- Complex Multi-substrate Kinetics: Differentiating ordered Bi Bi from random Bi Bi using isotope exchange at equilibrium still ties my brain in knots. The Cleland nomenclature and the math behind Ping-Pong product inhibition patterns will require more practice.
- Pre-Steady-State Global Fitting: While I understand the concept of burst kinetics, setting up the differential equations to extract multiple microscopic rate constants ($k_1, k_{-1}, k_2, k_3$) from stopped-flow data feels daunting and highly susceptible to experimental noise.
- Quantum Tunneling Models: The Marcus-like models for enzymatic proton transfer push the limits of my physical chemistry recall. I understand the macroscopic outcome, but the quantum mechanical derivation is still foggy.
🗺️ Visual Aids & Key Takeaways
Diagram 1: The Seven Strategies of Enzyme Catalysis — How Enzymes Lower ΔG‡
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ ENZYME CATALYSIS: SEVEN MECHANISMS OF RATE ACCELERATION ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
┌─────────────────────────────────┐
│ UNCATALYZED REACTION │
│ high ΔG‡, slow rate │
└────────────────┬────────────────┘
│ enzyme provides:
┌───────────────┴────────────────┐
↓ ↓
┌──────────────────┐ ┌──────────────────────┐
│ PROXIMITY & │ │ ELECTROSTATIC │
│ ORIENTATION │ │ STABILIZATION │
│ │ │ │
│ brings reactants │ │ preorganized charges │
│ together; removes│ │ in active site lower │
│ translational/ │ │ ΔG‡ without cost of │
│ rotational loss │ │ desolvation seen in │
│ (~10⁸ M effect) │ │ solution │
└──────────────────┘ └──────────────────────┘
┌──────────────────┐ ┌──────────────────────┐
│ GENERAL ACID- │ │ COVALENT │
│ BASE CATALYSIS │ │ CATALYSIS │
│ │ │ │
│ His, Asp, Glu │ │ nucleophilic residue │
│ shuttle protons │ │ (Ser, Cys, Lys) forms│
│ during reaction; │ │ transient covalent │
│ avoids full H⁺ │ │ intermediate; splits │
│ or OH⁻ release │ │ one barrier into two │
└──────────────────┘ └──────────────────────┘
┌──────────────────┐ ┌──────────────────────┐
│ TRANSITION │ │ METAL ION │
│ STATE │ │ CATALYSIS │
│ STABILIZATION │ │ │
│ │ │ Zn²⁺, Mg²⁺ activate │
│ active site is │ │ water, orient │
│ complementary to │ │ substrates, stabilize│
│ TS geometry, NOT │ │ negative charges on │
│ to ground state │ │ TS (e.g., metalloproteases,│
│ (Pauling, 1948) │ │ kinases with Mg-ATP) │
└──────────────────┘ └──────────────────────┘
┌─────────────────────────────────┐
│ CONFORMATIONAL CHANGES │
│ (INDUCED FIT / CONF. SELECTION) │
│ │
│ substrate binding drives protein │
│ to catalytically competent state; │
│ excludes water, positions residues│
└─────────────────────────────────┘
─────────────────────────────────────────
NOTE: Real enzymes deploy MULTIPLE strategies
simultaneously — the serine protease triad
uses covalent + general acid/base + oxyanion
hole electrostatics all in one active site.
─────────────────────────────────────────
The seven major catalytic strategies enzymes use to lower the activation energy ΔG‡; most enzymes combine at least three of these simultaneously, which is why enzymatic rate accelerations (up to 10¹⁷-fold over uncatalyzed) vastly exceed any single mechanism alone (Fersht, 1999, Ch. 2).
---Diagram 2: Michaelis-Menten Kinetic Framework — From Mechanism to Measurable Parameters
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ MICHAELIS-MENTEN KINETICS: MECHANISM, PARAMETERS, AND MEANING ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
THE ELEMENTARY STEPS
────────────────────
k₁ kcat
E + S ⇌ ES ──────→ E + P
k₋₁
(for simple one-step chemistry; kcat = k₂ in Briggs-Haldane)
BRIGGS-HALDANE STEADY-STATE ASSUMPTION
────────────────────────────────────────
d[ES]/dt ≈ 0 → k₁[E][S] = (k₋₁ + kcat)[ES]
Vmax · [S] where: Vmax = kcat · [E]total
v = ───────────────────────── KM = (k₋₁ + kcat) / k₁
KM + [S]
THE THREE FUNDAMENTAL PARAMETERS
──────────────────────────────────
┌──────────────┬──────────────────────────────┬────────────────────────────────────────┐
│ PARAMETER │ PHYSICAL MEANING │ IMPORTANT LIMITS / NOTES │
├──────────────┼──────────────────────────────┼────────────────────────────────────────┤
│ kcat │ max turnovers per enzyme per │ ranges from <1 s⁻¹ (lysozyme) │
│ (s⁻¹) │ second when E is saturated │ to ~10⁷ s⁻¹ (carbonic anhydrase) │
│ │ with S │ limited by chemical step OR release │
├──────────────┼──────────────────────────────┼────────────────────────────────────────┤
│ KM │ [S] at half-maximal velocity │ ≠ Kd unless k₋₁ >> kcat │
│ (M or µM) │ approximates affinity only │ physiological [S] often near KM; │
│ │ under rapid equilibrium │ allows sensitivity to [S] changes │
├──────────────┼──────────────────────────────┼────────────────────────────────────────┤
│ kcat/KM │ second-order rate constant │ upper limit: diffusion ~10⁸–10⁹ M⁻¹s⁻¹│
│ (M⁻¹s⁻¹) │ for E + S → product at low │ "catalytic efficiency"; measures │
│ │ [S]; discriminates substrates │ discrimination between substrates │
└──────────────┴──────────────────────────────┴────────────────────────────────────────┘
VELOCITY vs. [S] CURVE SHAPE
─────────────────────────────
Vmax ┤- - - - - - - - - - - - - - - - ─ ─ ─ ─
│ ╭─────────────
│ ╭───╯
Vmax/2 ┤- - - - - - - - -╭────╯
│ ╭──╯
│ ╭──╯
│ ╭──╯
0 ┼──────────┬─────────────────────────→ [S]
0 KM
[S] << KM : v ≈ (kcat/KM)·[E]·[S] ← second-order, E mostly free
[S] >> KM : v ≈ Vmax = kcat·[E] ← zero-order, E mostly as ES
THE THREE LINEAR TRANSFORMATIONS (and their flaws)
───────────────────────────────────────────────────
┌─────────────────────┬──────────────────────────────┬──────────────────────────────┐
│ LINEWEAVER-BURK │ EADIE-HOFSTEE │ HANES-WOOLF │
│ 1/v vs. 1/[S] │ v vs. v/[S] │ [S]/v vs. [S] │
│ │ │ │
│ slope = KM/Vmax │ slope = −KM │ slope = 1/Vmax │
│ y-int = 1/Vmax │ y-int = Vmax │ y-int = KM/Vmax │
│ x-int = −1/KM │ │ │
│ │ │ │
│ ⚠ magnifies error │ better error distribution │ best for comparing │
│ at low [S] (double │ than LB; v appears on │ inhibition patterns; │
│ reciprocal amplifies│ both axes (not independent) │ errors still not ideal │
│ noise) │ │ │
└─────────────────────┴──────────────────────────────┴──────────────────────────────┘
★ Modern practice: non-linear least-squares fitting directly to v = Vmax·[S]/(KM+[S])
Linear transforms are diagnostic aids and historical tools, NOT primary analysis.
The complete Michaelis-Menten framework from elementary steps to measurable parameters, showing the physical meaning of kcat, KM, and kcat/KM, the shape of the saturation curve, and a critical comparison of the three classical linear transforms with their known weaknesses (Copeland, 2000, Ch. 3; Fersht, 1999, Ch. 4).
---Diagram 3: Serine Protease Catalytic Mechanism — Acylation and Deacylation Half-Reactions
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ SERINE PROTEASE MECHANISM: CHARGE RELAY, TETRAHEDRAL INTERMEDIATES, ║
║ ACYL-ENZYME, AND PRODUCT RELEASE ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
THE CATALYTIC TRIAD (spatial arrangement in active site)
──────────────────────────────────────────────────────────
H-bond network
Asp ─── His ─── Ser
(−) H-bond H-bond (nucleophile)
Asp102 positions His57; His57 abstracts Ser195 proton
→ Ser195 becomes a powerful alkoxide nucleophile
Oxyanion hole (Gly193-NH, Ser195-NH) stabilizes
negative charge on carbonyl oxygen during TS
═══════════════════════════════════════════════════════════
PHASE 1: ACYLATION (peptide bond cleavage, amine leaves)
═══════════════════════════════════════════════════════════
STEP 1: Substrate binding and nucleophilic attack
──────────────────────────────────────────────────
oxyanion hole
NH···O⁻···HN
|
Ser-OH + R-CO-NHR' → Ser-O···C=O (trigonal planar → tetrahedral)
|
NHR' ← leaving group (amine of peptide)
His57 accepts proton from Ser195 as attack occurs
Asp102 stabilizes the positive charge developing on His57
STEP 2: First tetrahedral intermediate (TI-1) and collapse
────────────────────────────────────────────────────────────
O⁻ O
‖ ‖
Ser−O−C−NHR' → Ser−O−C−R + H₂NR'
| acyl-enzyme (first product, amine
R (covalent) leaves after His57
donates proton to N)
═══════════════════════════════════════════════════════════
PHASE 2: DEACYLATION (water attacks, acid product leaves)
═══════════════════════════════════════════════════════════
STEP 3: Water activation
─────────────────────────
Acyl-enzyme + H₂O → His57 abstracts proton from water
→ OH⁻ equivalent attacks carbonyl
→ Second tetrahedral intermediate (TI-2)
STEP 4: Collapse of TI-2 and product release
──────────────────────────────────────────────
O⁻ O
‖ ‖
Ser−O−C−OH → HO−Ser + R−C−OH (carboxylic acid product)
| (enzyme (second product)
R regenerated)
OVERALL REACTION COORDINATE
────────────────────────────
TI-1 TI-2
ΔG‡ /\ /\
│ / \ / \
│ acyl-E / \ acyl-E / \
│ ────────╱ ╲────────╱ ╲
│ / \
│──╱ E+S ╲── E+P
acylation deacylation
(often rate- (rate-limiting for
limiting for good leaving groups /
amide sub.) ester substrates)
SUBSTRATE SPECIFICITY: S1 POCKET
──────────────────────────────────
┌──────────────┬─────────────────────────────────────────────────────┐
│ PROTEASE │ S1 POCKET CHARACTER AND PREFERRED SUBSTRATE │
├──────────────┼─────────────────────────────────────────────────────┤
│ Trypsin │ Asp189 at bottom of S1 → binds Lys, Arg (positive) │
│ Chymotrypsin│ Ser189, hydrophobic walls → Phe, Trp, Tyr │
│ Elastase │ Val190, Thr226 block S1 → only small Ala, Gly │
│ Thrombin │ Asp189 + insertion loops → Arg, specific sequences │
└──────────────┴─────────────────────────────────────────────────────┘
★ The S1 pocket controls which peptide bond is cleaved — a single
amino acid substitution (Asp189Ser) converts trypsin-like to
chymotrypsin-like specificity. Structure drives specificity.
The complete two-phase serine protease mechanism showing the charge relay system, both tetrahedral intermediates, the covalent acyl-enzyme intermediate that can be experimentally trapped, and the structural basis of substrate specificity in the S1 pocket (Fersht, 1999, Ch. 15; Silverman, 2002, Ch. 6).
---Diagram 4: Enzyme Inhibition — Mechanisms, Kinetic Signatures, and Drug Examples
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ ENZYME INHIBITION: FOUR MECHANISMS AND THEIR KINETIC FINGERPRINTS ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
WHERE INHIBITORS BIND
──────────────────────
COMPETITIVE
I binds free E only, at active site
↓
I ┌─────────────────┐
↓ │ EI │ (dead-end)
E ──────→ ES ──────→ E+P │ │
└─────────────────┘
↑
UNCOMPETITIVE
I binds ES only, not free E
MIXED / NONCOMPETITIVE
I binds BOTH E and ES (at separate site)
KINETIC EFFECT ON PARAMETERS
──────────────────────────────
┌──────────────────────┬────────────────┬────────────────┬────────────────────────────┐
│ INHIBITOR TYPE │ KM (apparent) │ Vmax(apparent)│ Lineweaver-Burk pattern │
├──────────────────────┼────────────────┼────────────────┼────────────────────────────┤
│ COMPETITIVE │ KM·(1+[I]/Ki) │ unchanged │ lines intersect on y-axis │
│ │ ↑ increases │ │ (same 1/Vmax, diff slope) │
├──────────────────────┼────────────────┼────────────────┼────────────────────────────┤
│ UNCOMPETITIVE │ KM/(1+[I]/Ki')│ Vmax/α' │ parallel lines │
│ │ ↓ decreases │ ↓ decreases │ (same slope, diff y-int) │
├──────────────────────┼────────────────┼────────────────┼────────────────────────────┤
│ MIXED │ changes │ ↓ decreases │ lines intersect off axes │
│ (noncompetitive │ (α/α' ratio) │ │ (pure noncomp: intersect │
│ when α = α') │ │ │ on x-axis) │
├──────────────────────┼────────────────┼────────────────┼────────────────────────────┤
│ IRREVERSIBLE │ apparent Vmax │ ↓ decreases │ no equilibrium Ki; │
│ (covalent) │ decreases with│ (fewer active │ measure kinact/KI instead │
│ │ time of [I] │ enzyme sites) │ (pseudo-first-order) │
└──────────────────────┴────────────────┴────────────────┴────────────────────────────┘
IRREVERSIBLE INHIBITION: KINETICS OF INACTIVATION
───────────────────────────────────────────────────
kI (bind) kinact (react)
E + I ─────────────→ E·I ──────────────→ E−I (covalent, inactive)
(rapid equil.)
kobs = kinact · [I] / (KI + [I]) at [I] << KI: kobs ≈ (kinact/KI)·[I]
kinact/KI = second-order inactivation efficiency (analogous to kcat/KM)
Mechanism-based ("suicide") inhibitors: require enzyme's own catalytic
machinery to activate; highly selective. Examples:
• aspirin → acetylates Ser530 of COX-1 irreversibly
• clavulanic acid → inactivates β-lactamase (protects amoxicillin)
• selegiline → inactivates MAO-B
TIGHT-BINDING INHIBITORS (Ki* determination)
─────────────────────────────────────────────
When [I] ≈ [E]total: free [I] ≠ total [I]
→ Morrison equation required:
v/v₀ = 1 − ([E]+[I]+Ki* − √(([E]+[I]+Ki*)²−4[E][I])) / 2[E]
Examples: statins (HMG-CoA reductase), protease inhibitors (HIV drugs)
Ki* values often in pM range — measuring requires careful [E] control
REPRESENTATIVE CLINICAL INHIBITORS
────────────────────────────────────
┌─────────────────────────┬────────────────┬───────────────────────────────────────┐
│ DRUG │ TARGET │ INHIBITION TYPE │
├─────────────────────────┼────────────────┼───────────────────────────────────────┤
│ Atorvastatin │ HMG-CoA red. │ competitive (tight-binding) │
│ Imatinib (Gleevec) │ BCR-Abl kinase│ Type II (DFG-out, competitive vs ATP)│
│ Sitagliptin │ DPP-4 │ competitive (covalent-like in some) │
│ Saquinavir │ HIV protease │ competitive (transition state mimic) │
│ Aspirin │ COX-1/COX-2 │ irreversible covalent │
│ Clavulanic acid │ β-lactamase │ mechanism-based (suicide) │
└─────────────────────────┴────────────────┴───────────────────────────────────────┘
A unified view of the four inhibition mechanisms showing where each inhibitor binds, how it alters the kinetic parameters KM and Vmax, what its Lineweaver-Burk fingerprint looks like, and how irreversible and tight-binding inhibitors require fundamentally different analytical approaches (Copeland, 2000, Ch. 6–8; Fersht, 1999, Ch. 8).
---Diagram 5: GLP-1 Degradation Cascade — Enzymes, Cleavage Sites, and Drug Targets
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ GLP-1 PROCESSING, DEGRADATION, AND PHARMACOLOGICAL INTERVENTION POINTS ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
PROGLUCAGON PROCESSING (intestinal L-cells)
─────────────────────────────────────────────
Proglucagon ──[PC1/3 furin-like]──→ GLP-1(7-37) [active, 30 aa]
Furin/PC1 recognize Lys-Arg dibasic sites; proprotein convertase
family; this step is BIOSYNTHETIC (not a drug target for GLP-1 protection)
ACTIVE GLP-1: His⁷-Ala⁸-Glu⁹-Gly¹⁰-Thr¹¹-Phe¹²-Thr¹³-Ser¹⁴-...
─────────────────────────────────────────────────────────────────
↑ ↑
His7-Ala8 critical N-terminus for GLP-1R binding
│
│ DPP-4 cleaves HERE (Xaa-Pro/Ala rule)
↓
┌───────────────────────────────────────────────────────────────┐
│ DPP-4 (CD26) — SERINE PROTEASE (prolyl oligopeptidase clan) │
│ Cleaves: His⁷-Ala⁸↓Glu⁹ dipeptide from N-terminus │
│ Product: GLP-1(9-37) — INACTIVE; cannot bind GLP-1R │
│ Half-life of native GLP-1 in plasma: ~2 minutes │
│ Drug target: sitagliptin, saxagliptin, linagliptin │
│ (competitive inhibitors, Ki in low nM range) │
└───────────────────────────────────────────────────────────────┘
│
↓ (DPP-4-resistant GLP-1 still degraded by:)
┌───────────────────────────────────────────────────────────────┐
│ NEPRILYSIN (NEP, CD10, enkephalinase) — ZINC METALLOPROTEASE│
│ Cleaves hydrophobic residues on N-side of peptide bond │
│ Degrades: GLP-1, ANP, BNP, enkephalins, substance P │
│ Drug relevance: sacubitril (prodrug) inhibits NEP │
│ → combination sacubitril/valsartan (Entresto) for HF │
│ NEP inhibition raises ANP AND may extend GLP-1 action │
└───────────────────────────────────────────────────────────────┘
│
↓
┌───────────────────────────────────────────────────────────────┐
│ INSULIN-DEGRADING ENZYME (IDE) — ZINC METALLOPROTEASE │
│ Degrades: GLP-1, insulin, amyloid-β (Aβ) │
│ Clinical interest: IDE links T2DM and Alzheimer's pathology │
│ IDE inhibition raises both insulin and GLP-1 exposure │
└───────────────────────────────────────────────────────────────┘
PHARMACOLOGICAL STRATEGIES TO EXTEND GLP-1 ACTION
────────────────────────────────────────────────────
┌──────────────────────┬──────────────────────────────────────────────────────────┐
│ STRATEGY │ MECHANISM │
├──────────────────────┼──────────────────────────────────────────────────────────┤
│ DPP-4 inhibitors │ block N-terminal cleavage; protect endogenous GLP-1 │
│ (gliptins) │ modest HbA1c reduction (~0.7%); oral; weight-neutral │
├──────────────────────┼──────────────────────────────────────────────────────────┤
│ GLP-1 receptor │ Ala⁸ substituted (Val or Aib) → DPP-4 resistant │
│ agonists (e.g., │ fatty acid conjugation → albumin binding, slow release │
│ semaglutide) │ t½ ~7 days; supraphysiological GLP-1R activation │
├──────────────────────┼──────────────────────────────────────────────────────────┤
│ NEP inhibitors │ sacubitril; raises GLP-1, ANP, BNP simultaneously │
│ (sacubitril) │ combined with ARB for heart failure indication │
└──────────────────────┴──────────────────────────────────────────────────────────┘
★ KEY INSIGHT: GLP-1's extraordinarily short half-life (~2 min in plasma)
is entirely an enzyme kinetics problem. Every approved GLP-1-based drug
is an answer to the question: how do we outrun DPP-4?
The complete GLP-1 enzymatic processing and degradation cascade, showing the biosynthetic cleavage by proprotein convertases, the rapid DPP-4-mediated inactivation that limits native GLP-1 half-life to ~2 minutes, and the mechanistic rationale for each class of GLP-1-based therapeutics (Drucker, 2019; Berg et al., 2019, Ch. 27).
---Diagram 6: Kinase Regulation and Inhibitor Classification — From DFG Motif to Drug Design
╔══════════════════════════════════════════════════════════════════════════════════════════╗
║ PROTEIN KINASE ARCHITECTURE, REGULATION, AND INHIBITOR CLASSIFICATION ║
╚══════════════════════════════════════════════════════════════════════════════════════════╝
KINASE DOMAIN ARCHITECTURE (conserved across ~518 human kinases)
─────────────────────────────────────────────────────────────────
N-LOBE C-LOBE
┌────────────────┐ ┌─────────────────────┐
│ β-sheets │ │ α-helices │
│ glycine-rich │ │ activation loop │
│ P-loop │ │ catalytic loop (HRD) │
│ (ATP γ-phos. │ │ substrate binding │
│ positioning) │ │ DFG motif │
└───────┬────────┘ └──────────┬──────────┘
│ │
└──────────────┬──────────────────┘
ATP binding
cleft (hinge region)
CATALYTIC MECHANISM (bisubstrate: ATP + protein-OH)
─────────────────────────────────────────────────────
Mg²⁺ coordinates β- and γ-phosphates of ATP
Lys (β3) salt bridges to α-phosphate
Asp (catalytic, HRD motif) acts as base → abstracts OH proton
γ-phosphate transferred to Ser/Thr/Tyr → ADP released
Ordered Bi Bi (mostly): ATP binds first, then protein substrate
THE DFG MOTIF: CONFORMATIONAL SWITCH FOR REGULATION
──────────────────────────────────────────────────────
DFG-IN (active): Asp coordinates Mg²⁺; hydrophobic spine assembled
activation loop open; substrate access permitted
DFG-OUT (inactive): Phe flips into ATP site; Asp points away from Mg²⁺
creates allosteric pocket adjacent to ATP site
DFG-IN DFG-OUT
┌──────────────────┐ ┌──────────────────────┐
│ Asp→Mg²⁺ (active)│ │ Phe→hydrophobic pocket│
│ spine intact │ │ new allosteric site │
│ ATP binds well │ │ exposed for type II │
│ │ │ inhibitors │
└──────────────────┘ └──────────────────────┘
KINASE INHIBITOR CLASSIFICATION
─────────────────────────────────
┌────────┬──────────────┬────────────────────────────────────────────────────────┐
│ TYPE │ DFG STATE │ BINDING MODE AND EXAMPLES │
├────────┼──────────────┼────────────────────────────────────────────────────────┤
│ I │ DFG-IN │ ATP-competitive; binds hinge region directly │
│ │ (active) │ most kinase drugs; selectivity challenge │
├────────┼──────────────┼────────────────────────────────────────────────────────┤
│ Ia │ DFG-IN │ type I + covalent bond to Cys near hinge │
│ (cov.)│ │ ibrutinib (BTK Cys481); afatinib (EGFR Cys797) │
├────────┼──────────────┼────────────────────────────────────────────────────────┤
│ II │ DFG-OUT │ occupies ATP site AND allosteric DFG-out pocket │
│ │ (inactive) │ imatinib (BCR-Abl); sorafenib (Raf/VEGFR) │
│ │ │ better selectivity; locks inactive conformation │
├────────┼──────────────┼────────────────────────────────────────────────────────┤
│ III │ any │ purely allosteric; does NOT overlap ATP site │
│ │ │ trametinib (MEK1/2 allosteric site); non-competitive │
├────────┼──────────────┼────────────────────────────────────────────────────────┤
│ IV │ any │ bivalent / outside kinase domain (substrate site) │
│ │ │ experimental; high selectivity potential │
└────────┴──────────────┴────────────────────────────────────────────────────────┘
★ IMATINIB LESSON: Imatinib's selectivity for BCR-Abl over most other
kinases comes not from the ATP site (shared by all) but from its
occupation of the DFG-out allosteric pocket AND a specific hydrogen
bond to a threonine gatekeeper residue (Thr315). T315I mutation
→ steric clash → imatinib resistance. Mechanism determines resistance.
★ SCALE: Manning et al. (2002) catalogued 518 human protein kinases —
the "kinome" — representing ~1.7% of all human genes. Dysregulation
of this network underlies most cancers and many metabolic diseases.
Protein kinase architecture, the DFG-in/DFG-out conformational switch that controls activity, and the four-type classification of kinase inhibitors, illustrating why imatinib's type II mechanism explains both its selectivity and the molecular basis of resistance mutations (Manning et al., 2002; Berg et al., 2019, Ch. 14).
---Key Takeaways
- Enzymes work by transition state complementarity, not ground state binding. Linus Pauling's insight (1948): the active site is geometrically and electrostatically complementary to the transition state of the reaction, not to the substrate itself. An enzyme that bound substrate too tightly would slow the reaction by lowering the substrate's energy without lowering ΔG‡. This is why transition state analogs (e.g., phosphonate mimics for proteases, PALA for ATCase) are extraordinarily potent inhibitors — dissociation constants in the femtomolar range are possible (Fersht, 1999, Ch. 2).
- KM is not the same as Kd, except under the rapid equilibrium assumption. Under the Briggs-Haldane steady-state derivation, KM = (k₋₁ + kcat)/k₁. KM equals Kd (= k₋₁/k₁) only when k₋₁ >> kcat. In many real enzymes (where kcat is fast relative to k₋₁), KM overestimates affinity. Always distinguish the two: Kd is a thermodynamic quantity; KM is a kinetic composite (Copeland, 2000, Ch. 3).
- kcat/KM is the true measure of catalytic efficiency and substrate discrimination, with a hard diffusion ceiling of ~10⁸–10⁹ M⁻¹s⁻¹. At [S] << KM, velocity = (kcat/KM)·[E]·[S], so kcat/KM is the effective second-order rate constant. "Catalytically perfect" enzymes like triose phosphate isomerase and acetylcholinesterase approach this diffusion limit — they react with essentially every substrate molecule they encounter. Enzymes cannot be more efficient than diffusion allows (Fersht, 1999, Ch. 4).
- Competitive inhibitors raise apparent KM without affecting Vmax; uncompetitive inhibitors lower both KM and Vmax proportionally. The counterintuitive consequence of uncompetitive inhibition: adding an uncompetitive inhibitor actually makes the enzyme appear to have higher affinity for its substrate (lower apparent KM) even as it reduces Vmax. This is because the inhibitor traps the ES complex, depleting free E and shifting the equilibrium toward ES. Uncompetitive inhibitors are therefore paradoxically more potent at high [S] (Copeland, 2000, Ch. 7).
- The serine protease mechanism proceeds through two tetrahedral intermediates and a covalent acyl-enzyme intermediate, detectable by burst kinetics. Acylation (peptide bond cleavage, amine leaves) and deacylation (water hydrolyzes acyl-enzyme, acid product leaves) are distinct chemical phases. With chromogenic ester substrates (e.g., p-nitrophenyl esters), the burst of product in pre-steady-state corresponds stoichiometrically to active enzyme concentration, allowing direct active site titration. The acyl-enzyme can be cryogenically or chemically trapped (Fersht, 1999, Ch. 15).
- Substrate specificity in serine proteases is determined primarily by the S1 pocket, where a single residue change fundamentally alters selectivity. Asp189 in trypsin creates an anionic pocket for Lys/Arg; Ser189 in chymotrypsin allows a hydrophobic pocket for Phe/Tyr/Trp; Val190/Thr226 in elastase fill the pocket, restricting it to small residues (Ala, Gly). The Asp189Ser mutation interconverts trypsin-like and chymotrypsin-like specificity. Thrombin adds exosite loops for fibrinogen recognition beyond S1 (Silverman, 2002, Ch. 6).
- DPP-4 (CD26) cleaves the His⁷-Ala⁸ dipeptide from GLP-1 within ~2 minutes in plasma, rendering it inactive. DPP-4 is a serine protease of the prolyl oligopeptidase family with strict Xaa-Pro or Xaa-Ala specificity at P1. The resulting GLP-1(9-37) cannot bind GLP-1R. DPP-4 inhibitors (gliptins: sitagliptin, saxagliptin, linagliptin) extend the half-life of endogenous GLP-1 but produce only modest glycemic effects because they work on physiological (nanomolar) GLP-1 levels, whereas GLP-1 receptor agonists produce pharmacological (supraphysiological) receptor activation (Drucker, 2019).
- Caspases use a cysteine-thiolate/imidazolium ion pair for nucleophilic catalysis and exist as inactive zymogens requiring proteolytic activation. The Cys285-His237 dyad (in caspase-1 numbering) forms at physiological pH; the thiolate is the nucleophile. Initiator caspases (caspase-8, -9) are activated by dimerization/proximity (not cleavage); effector caspases (caspase-3, -7) require cleavage between subunits to become active. This proteolytic cascade is irreversible — once effector caspases are activated, cell death commitment is essentially complete (Salvesen and Dixit, 1997).
- Imatinib's selectivity for BCR-Abl is mechanistically explained by its type II binding to the DFG-out inactive conformation plus a critical hydrogen bond to Thr315 (the "gatekeeper" residue). Most kinase inhibitors must compete with millimolar intracellular ATP at the highly conserved ATP site. Imatinib extends into the DFG-out allosteric pocket, which is less conserved, providing selectivity. The T315I gating mutation eliminates the Thr315 hydrogen bond AND creates steric clash with imatinib — a mechanistic prediction that was clinically confirmed before crystal structures were solved (Manning et al., 2002).
- Mechanism-based (suicide) inhibitors require catalytic turnover by the target enzyme to generate the active inhibiting species, which is the source of their extraordinary selectivity. The enzyme itself converts a pro-inhibitor to a reactive species that then covalently modifies the active site before escaping. The key kinetic parameter is kinact/KI (analogous to kcat/KM for substrates). Examples: clavulanic acid (β-lactamase inactivation, activated via ring opening); aspirin (acetylates COX Ser530 via the enzyme's own esterase-like mechanism); selegiline (MAO-B oxidizes the propargylamine to a reactive species). Selectivity arises because only the intended enzyme can activate the inhibitor (Silverman, 2002, Ch. 7).
- Pre-steady-state burst kinetics detect covalent intermediates and measure the chemical step rate independently of product release. When product release is rate-limiting in steady state, mixing enzyme with excess substrate in a stopped-flow instrument reveals an initial rapid burst of product (one equivalent per enzyme active site) followed by a slower linear phase. The burst amplitude = [active enzyme]; the burst rate constant = kacylation (or k₂); the linear rate = kcat (limited by deacylation or product release). This technique was decisive in confirming the acyl-enzyme intermediate in serine proteases (Fersht, 1999, Ch. 7).
- The human kinome contains 518 protein kinases (Manning et al., 2002), representing ~1.7% of human genes; approximately 244 are associated with disease. Kinases are classified into seven major groups (AGC, CAMK, CK1, CMGC, RGC, STE, TK, TKL) based on sequence and function. The tyrosine kinase (TK) group is most heavily implicated in cancer. The kinome phylogenetic tree is a practical tool for predicting off-target effects: kinase inhibitors tend to also inhibit phylogenetically related kinases, so understanding kinome position informs selectivity profiling (Manning et al., 2002).
- Allosteric enzymes display sigmoidal (cooperative) kinetics described by the Hill equation, where the Hill coefficient n > 1 indicates positive cooperativity. The Hill equation, v = Vmax·[S]ⁿ/(K₀.₅ⁿ + [S]ⁿ), is phenomenological — n is not literally the number of binding sites but a cooperativity index. In the MWC (concerted) model, all subunits switch simultaneously between T (tense, low affinity) and R (relaxed, high affinity) states. In the KNF (sequential) model, binding to one subunit induces conformational change in neighbors. Phosphofructokinase is the textbook allosteric metabolic switch: AMP is an allosteric activator; ATP at high concentrations is an inhibitor (Berg et al., 2019, Ch. 16).
- The Eyring transition state theory equation, k = (kBT/h)·e^(−ΔG‡/RT), connects the macroscopic rate constant to the free energy of activation, which can be dissected into enthalpic and entropic components. From temperature-dependent rate data (van't Hoff/Arrhenius analysis), ΔH‡ and ΔS‡ can be extracted. Enzymes typically show large negative ΔS‡ for uncatalyzed reactions (entropic cost of bringing reactants together) but near-zero or positive ΔS‡ upon catalysis, reflecting the preorganization advantage. Primary kinetic isotope effects (kH/kD = 2–7 for C-H bond breaking under classical tunneling, but >10 for quantum tunneling contributions) probe whether C-H cleavage is rate-limiting at the chemical step level (Fersht, 1999, Ch. 3).
References
- Fersht, A., Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Protein Folding, W.H. Freeman and Company, New York, 1999. ISBN: 978-0716732686.
- Copeland, R.A., Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis, 2nd ed., Wiley-VCH, New York, 2000. ISBN: 978-0471359297.
- Silverman, R.B., The Organic Chemistry of Enzyme-Catalyzed Reactions, 2nd ed., Academic Press, San Diego, 2002. ISBN: 978-0126437317.
- Berg, J.M., Tymoczko, J.L., and Stryer, L., Biochemistry, 9th ed., W.H. Freeman and Company, New York, 2019. ISBN: 978-1319114671.
- Manning, G., Whyte, D.B., Martinez, R., Hunter, T., and Sudarsanam, S., 'The Human Protein Kinase Complement,' Science, 298(5600): 1912–1934, 2002. DOI: 10.1126/science.1075762.
- Drucker, D.J., 'GLP-1 Receptor Agonists: Mechanisms of Action and Clinical Results,' Diabetes Care, 42(Suppl 1): S90–S98, 2019. DOI: 10.2337/dbi19-0022.
- Salvesen, G.S. and Dixit, V.M., 'Mechanism and Regulation of Caspases in Apoptosis,' Cell, 91(4): 443–446, 1997. DOI: 10.1016/S0092-8674(00)80712-5.
- Pauling, L., 'Molecular Architecture and Biological Reactions,' Chemical and Engineering News, 24: 1375–1377, 1946. [Foundational paper articulating transition state complementarity in enzyme catalysis.]
- Briggs, G.E. and Haldane, J.B.S., 'A Note on the Kinetics of Enzyme Action,' Biochemical Journal, 19(2): 338–339, 1925. DOI: 10.1042/bj0190338.
- Morrison, J.F., 'Kinetics of the Reversible Inhibition of Enzyme-Catalysed Reactions by Tight-Binding Inhibitors,' Biochimica et Biophysica Acta, 185(2): 269–286, 1969. DOI: 10.1016/0005-2744(69)90420-3.