How encounters became biochemical rate laws

The experiments and mathematical ideas behind Lecture 2: what was observed, what a model added, and which distinctions a modern cellular calculation must preserve.

CCBS · EST 5053 · Fall 2026 · Lecture 2 lineage

The quantities in a biochemical rate law came from different experimental questions. A microscope records particle positions. An enzyme assay records product formation. An equilibrium binding experiment records occupied sites. A cellular perturbation records changing pools and fluxes. Connecting these measurements is powerful because they constrain different parts of a mechanism.

The core combines these constraints into an argument about cellular rates. The exposition derives the calculations and their conditions. This essay follows their origins, grouped by the questions they answer. A modern formula may appear in several historical programmes; its appearance does not imply that all its present interpretations were available at once.

1A cellular number carries its measurement conditions

The modern quantitative-biology programme made the provenance of a number part of its meaning.

A copy number becomes a concentration only after choosing a volume. The conversion $c=n/(N_AV)$ is elementary, but its biological inputs vary with growth conditions and measurement method. A reference bacterium of 1 fL makes one molecule about 1.7 nM. Treating that reference as the volume of every bacterium changes the inferred occupancy and kinetic regime before any reaction has been modeled.

Volkmer and Heinemann measured condition-dependent E. coli volumes to make such conversions more reliable. Their reported means ranged from 1.5 fL in one stationary-phase condition to 4.4 fL in LB. Milo's later reassessment of protein counts compared independent estimates of protein mass concentration and mean molecular mass, showing why a familiar total can deserve recalculation.¹²

Metabolite surveys supplied a complementary census. Bennett and colleagues measured absolute metabolite concentrations and compared them with enzyme parameters. The comparison connects a chemical pool with possible active-site occupancy, but requires distinguishing free concentration, total extracted material and the kinetic meaning of $K_M$.³ These measurements motivate the core's concentration ledger; they do not supply every enzyme abundance, association rate or pathway flux needed later.

Recover the volume, condition and averaging convention behind a biological number before transferring it to a new calculation.

2Brown's 1827 observations tested a biological explanation

The moving objects were suspended particles visible to a microscope, not the water molecules responsible for their agitation.

Brown began with particles contained in pollen. His 1828 account describes observations made the previous summer, including the irregular motion of small particles released from pollen. The original wording calls some small bodies “molecules,” but that historical usage must not be read as identification of the molecular constituents of water.⁴

Changing the material tested whether the motion depended on life. Brown examined dried plant specimens and then inorganic material. Powdered window glass and minerals also yielded suspended moving particles. His account even identifies a fragment of the Sphinx among the granite specimens. These comparisons undermined explanations tied specifically to living pollen; they did not yet provide a quantitative molecular theory of the motion.

Pearle and colleagues later revisited what Brown could have seen, combining historical reading with modern microscopy and calculations. Their displacement estimates distinguish a small particle from a much larger pollen grain.⁵ A displacement calculated from Stokes–Einstein theory is a prediction to compare with an observation. It should not be cited as a measurement Brown made, or reused as independent experimental calibration of the same theory.

Distinguish the observed particle, the historical interpretation and a later theoretical reconstruction when using Brownian motion as evidence.

3Stokes and Einstein connected drag to fluctuations

A deterministic resistance law and a statistical displacement law became two views of the same thermal process.

Stokes's problem concerned fluid resistance. His 1851 paper investigated how internal fluid friction affects pendulum motion. Within that much longer analysis, he treated slow uniform motion of a sphere and obtained the drag now written $F=6\pi\eta aU$. Here F is the drag force, U the sphere speed, a its radius and $\eta$ the dynamic viscosity. The result depends on the continuum fluid equations, negligible inertia and the surface boundary condition; its coefficient is more specific than dimensional reasoning alone.⁶

Einstein's 1905 question concerned the observable consequences of molecular thermal motion. He related diffusion of suspended particles to their mobility by balancing diffusion against a force-driven drift in equilibrium. Combining that relation with the sphere's drag gives $D=k_BT/(6\pi\eta a)$, where $k_B$ is Boltzmann’s constant and T is absolute temperature. He also derived a displacement distribution and a one-coordinate mean-square displacement proportional to time.⁷

The first relation translates a diffusion coefficient into a statistic of trajectories. The second connects that coefficient to temperature, viscosity and radius. Together they allow macroscopic observations of suspended particles to constrain molecular-scale constants. Neither relation says a particle follows a straight trajectory of length $\sqrt{2Dt}$, or that this RMS displacement is a mean time to find a particular target.

Use mean-square displacement to connect a trajectory statistic with mobility, while retaining the mechanical assumptions behind the drag coefficient.

4Mass action developed across equilibrium, kinetics and network theory

The modern framework joins related ideas whose historical formulations were not identical.

Guldberg and Waage's 1864 study sought quantitative laws of chemical affinity. Their experimental discussion used reacting amounts and equilibrium relations, with powers and coefficients whose formulation differs from a modern elementary-reaction law. Reading it as though it already contained every present-day stoichiometric rate rule erases the problem they were trying to solve.⁸

Van 't Hoff's 1884 Études de dynamique chimique developed the analysis of chemical change in time, including reaction order and the relation between opposing reactions and equilibrium.⁹ For today's reader, the useful distinction is between a net chemical equation, a measured concentration dependence, and a mechanism whose elementary events explain that dependence. The exposition derives a bimolecular rate from pair counting to make the extra mechanistic assumption visible.

Later network mathematics formalized rate laws at another level. Horn and Jackson's 1972 treatment studied general mass-action systems, including formal reaction structures beyond closed, mass-conserving elementary chemistry.¹⁰ A mathematical theorem applies to the equations and assumptions it states. Our use of straight labeled arrows for elementary steps and bare squiggly arrows for composite processes is a teaching convention about chemical resolution, not a historical claim about the authors' notation or a substitute for checking a theorem's hypotheses.

Read a mass-action claim at its stated level: equilibrium relation, empirical rate dependence, elementary mechanism or mathematical network model.

5Smoluchowski turned colloidal coagulation into a capture problem

The absorbing sphere describes transport to a specified boundary before it describes an enzyme.

The original setting was the aggregation of colloidal particles. Smoluchowski formulated diffusion toward a particle that captures an approaching partner. His paper imposes vanishing concentration at the capture radius, a dilute surrounding suspension and a specified far-field concentration. The archived publisher volume is dated 1918, and the paper records submission in 1916; the work is also commonly cited as 1917 in later literature.¹¹

For an absorbing radius a and number concentration $c_\infty$, the long-time spherical solution gives an arrival flux $J=4\pi Da c_\infty$. The original treatment also contains the transient correction proportional to $a/\sqrt{\pi Dt}$. Neglecting it requires a time long compared with $a^2/D$, as well as conditions under which surrounding depletion does not invalidate the dilute-pair picture.

Relative motion connects the model to a pair of molecules. Independent diffusion gives $D=D_A+D_B$, and touching spheres have $a=a_A+a_B$. The resulting coefficient has volume-per-time units. Converting it into M⁻¹s⁻¹ makes it dimensionally comparable with a bimolecular association coefficient or a specificity constant. The conversion does not make physical contact, formed complex and product the same event.

Identify the capture boundary and transient assumptions before interpreting an encounter coefficient as a biochemical rate.

6Michaelis and Menten repaired the experiment as well as the model

Their 1913 invertase study addressed what an optical measurement was actually reporting.

Michaelis and Menten explicitly built on Henri's earlier work. Their paper opens by crediting his treatment of enzyme kinetics and identifying two experimental issues requiring renewed investigation: hydrogen-ion concentration and sugar mutarotation. They controlled the solution with an acetate mixture and worked at a specified temperature.¹² The historical advance was a combined programme of measurement control and mechanistic interpretation, rather than the isolated invention of a familiar hyperbola.

Optical rotation could change after the enzyme had made product. Invertase converts sucrose into glucose and fructose. Newly produced glucose changes its anomeric composition over time, changing optical rotation independently of additional sucrose cleavage. The experiment therefore took timed samples, stopped invertase action with alkali, allowed rapid sugar equilibration, and then read the polarization angle. The signal had to be separated from this additional chemical relaxation before it could reliably report enzymatic progress.

Initial rates and full time courses answered complementary questions. Early measurements reduced the influence of accumulating products. Separate tests examined inhibition by glucose and fructose. Their integrated treatment then included product binding to account for progress over a wider interval. Johnson and Goody's 2011 translation and reanalysis make these parts of the original programme accessible and compare the historical analysis with modern numerical fitting.¹³

Ask whether the assay signal reports the desired chemical event directly, or also contains a later transformation of the product.

7Briggs and Haldane changed the meaning of the saturation constant

The same rate curve can follow from a stationary complex balance without rapid binding equilibrium.

The 1913 treatment used an equilibrium relation for substrate binding. With free substrate S, complex $C_{ES}$ and enzyme total $q_E$, that approximation gives $C_{ES}=q_ES/(K_d+S)$. Product formation proportional to complex then produces a saturating rate. Under this rapid-equilibrium interpretation, the half-saturation concentration is approximately the binding dissociation constant.

Briggs and Haldane's two-page 1925 paper retained catalytic loss from the complex. Let the microscopic association coefficient be $k_{\mathrm{on}}$, the dissociation rate $k_{\mathrm{off}}$ and the catalytic conversion rate $k_{\mathrm{cat}}$. In modern notation, the balance for the minimal mechanism is

Balancing formation against both exits gives $K_M=(k_{\mathrm{off}}+k_{\mathrm{cat}})/k_{\mathrm{on}}$ and the same functional form for product flux v, $v=k_{\mathrm{cat}}q_ES/(K_M+S)$. Their paper explicitly recognized that kinetic observations of this form need not determine the relative sizes of dissociation and decomposition.¹⁴

The change matters biologically: $K_M$ remains the substrate concentration giving half of the limiting product rate in this model, but is no longer generally an equilibrium affinity. If catalysis is appreciable compared with dissociation, product formation itself contributes to the removal that binding must replenish.

Interpret a fitted saturation constant through the mechanism and approximation that produced it before calling it a dissociation constant.

8The specificity constant needs enzyme normalization

The non-saturating slope links the historical assay to an effective bimolecular rate law.

The raw product-rate slope includes how much enzyme was added. At low free substrate, $v\approx(V_{\max}/K_M)S$. The coefficient $V_{\max}/K_M$ has units s⁻¹. Dividing it by enzyme concentration gives the per-enzyme slope $k_{\mathrm{cat}}/K_M$, with units M⁻¹s⁻¹.

Michaelis and Menten's integrated analysis estimated a constant corresponding to the enzyme-containing quantity $q_Ek_{\mathrm{cat}}/K_M$. As Johnson and Goody explain, the enzyme's molecular concentration was not known, so the 1913 data could not yield the modern per-enzyme specificity constant.¹³ This is exactly the normalization needed to compare enzymes at different abundances.

The slope describes a composite conversion that behaves like mass action. When $S\ll K_M$, the overall reaction $S\rightsquigarrow P$ has the effective law $v\approx(k_{\mathrm{cat}}/K_M)q_ES$. For the minimal mechanism,

The fraction is the probability that an already formed complex makes product before dissociating. This gives a physical explanation for the effective coefficient: complex formation multiplied by the probability of a productive exit. Bar-Even and colleagues' 2015 discussion explicitly uses this relation to examine futile encounters.¹⁵ The coefficient equals neither the association rate nor the catalytic rate in general.

Normalize the non-saturating product-rate slope by enzyme concentration to obtain the coefficient governing effective bimolecular conversion.

9A partially reactive surface changes the boundary condition

Later diffusion theories made imperfect capture part of the model.

Collins and Kimball separated arrival from reaction at contact. Their 1949 analysis replaced perfect absorption by a boundary with finite surface reactivity. In modern notation, the steady radial condition is $Dc'(a)=\kappa c(a)$, where $\kappa$ has length-per-time units. Solving the exterior diffusion problem gives

The two terms express resistance associated with transport and with finite reactivity at the surface. This is a spatial boundary model; it is not automatically identical to the elementary enzyme's binding-and-exit scheme.¹⁶ Comparing the two requires deciding which physical event defines the bound state.

Berg and Purcell asked a related question about receptor geometry. Their 1977 chemoreception analysis considered small absorbing patches on an otherwise reflecting cell surface. Dispersed patches can capture much more effectively than their fractional area alone suggests because nearby diffusing molecules can revisit the surface.¹⁷ The extension works through this result and the penalty for clustering equal absorbing area into one patch.

Choose a boundary model that distinguishes perfect absorption, finite surface reactivity and spatially restricted capture.

10Orientation makes a molecular encounter more specific

Repeated close approaches change the cost of requiring an aligned binding patch.

Proteins can touch without forming the specified complex. Northrup and Erickson's 1992 Brownian-dynamics calculations compared association under different orientational requirements. Whole-surface capture and contact between aligned reactive patches produced different association rates.¹⁸ The calculation showed why a geometrical collision coefficient is an informative reference but may greatly exceed complex formation.

A failed orientation does not necessarily end the encounter. Partners that remain close can rotate and approach again before diffusing far apart. Multiplying the ideal collision coefficient by the fraction of favorable orientations at one independently sampled contact can therefore exaggerate the orientation penalty. The relative motion and the criterion for complete escape belong in the calculation.

This historical refinement prevents two opposite mistakes in the core's rate comparison. A small active site need not reduce capture exactly in proportion to its area, and a large rate deficit below an absorbing-sphere reference need not mean a formed enzyme–substrate complex usually dissociates. Some of the deficit can precede formation of that complex.

Distinguish a geometrical touch, a repeated close encounter and a formed binding complex when assigning a success probability.

11Fast association tests the geometry of the search

Electrostatics and DNA-mediated search change the assumptions behind an uncharged-sphere reference.

Electrostatic guidance depends on the arrangement of charges. Getzoff and colleagues engineered faster superoxide dismutase variants by changing residues around the approach to the active site. Their 1992 study connected altered reaction rates with electrostatic guidance and a local hydrogen-bonding network.¹⁹ Neutralizing selected negative charges and replacing them by positive ones were not equivalent improvements. A single net-charge count cannot replace the structural and kinetic evidence.

A DNA-binding protein can use more than one route to a target. Binding away from the specific site, moving along DNA, dissociating and rebinding can combine with three-dimensional arrival. Halford and Marko reviewed how such mechanisms should be tested and how frequently very high association rates had been generalized beyond the supporting data.²⁰ A dependence on DNA length is informative, but it does not alone distinguish continuous sliding from other mechanisms involving nonspecific association and rebinding.

An apparent rate above one computed diffusion reference is consequently a reason to revisit its geometry, interactions, units and event definition. It is not evidence that diffusion has ceased to matter. Salt dependence also belongs to a specified molecular pair and condition; it does not supply one universal physiological reduction factor.

Use an unusually fast rate to test the reference model's geometry and interactions before treating its numerical value as a universal physical ceiling.

12Albery and Knowles evaluated the whole free-energy profile

Their catalytic-perfection question included substrate concentration and the constraints of a reversible cycle.

Albery and Knowles asked how much a catalyst could still improve. Their 1976 analysis used the measured and constrained free-energy profile of triosephosphate isomerase. They separated uniform stabilization of bound states, changes in relative stability among intermediates, and lowering barriers for elementary chemical steps.²¹ These changes do different things to the time an enzyme spends in each part of its cycle.

The relevant performance depended on physiological concentration. Their efficiency measure included all the kinetic steps and the substrate concentration, with a correction for opposing flux near equilibrium. At a chosen substrate standard of 40 µM, their comparison gave efficiency about 0.6 for the enzyme and about $2.5\times10^{-11}$ for a simple carboxylate catalyst. The enzyme's value depended on the association reference used; the paper discusses that dependence explicitly.

The later 1977 account explains why the limiting diffusion step can be associated with the less stable substrate and why both association and product separation matter in a reversible reaction.²² Near-diffusion specificity is particularly informative below saturation. At higher substrate concentrations, bound-state waiting and release can change what limits output. Thus “catalytic perfection” in this programme cannot be reduced to a context-free threshold on $k_{\mathrm{cat}}/K_M$.

Judge catalytic improvement at specified substrate and product concentrations using the rate-limiting parts of the complete cycle.

13A broad survey made moderate efficiency the empirical comparison

The survey describes distributions of parameters, which must be kept distinct from a matched microscopic census.

Bar-Even and colleagues assembled enzyme kinetic measurements across many reactions. Their 2011 analysis reported a broad distribution with typical turnover around 10 s⁻¹ and median specificity of order $10^5\,\mathrm{M^{-1}s^{-1}}$.²³ The units are essential: $10^5$ is an effective bimolecular coefficient, not a dimensionless measure of excellence and not a turnover frequency.

The comparison with an ideal encounter reference around $10^9\text{--}10^{10}\,\mathrm{M^{-1}s^{-1}}$ says that product formation per enzyme and per free substrate is often far below ideal whole-surface arrival. It does not by itself allocate the difference among orientational access, complex formation, nonproductive exits and later chemistry. A survey median also need not describe any one enzyme's full parameter combination.

The mechanistic identity narrows what additional measurement is needed. In the minimal scheme, $p_{\mathrm{product}}=(k_{\mathrm{cat}}/K_M)/k_{\mathrm{on}}$. Bar-Even and colleagues' 2015 discussion develops the role of futile encounters and enzyme dynamics.¹⁵ Applying that identity to a specific enzyme requires its actual complex-forming association rate. Substituting an ideal sphere coefficient adds a substantive assumption.

The 2015 paper defines its encounter state more broadly than a chemically resolved substrate complex. Its coarse-grained association coefficient refers to a diffusion-controlled initial encounter, and its downstream rates collect multiple processes. Its inference about futile encounters belongs to that definition. Inferring rapid equilibrium of a particular active-site complex additionally requires resolving how that encounter state connects to the complex.

Likewise, independently sampling marginal distributions of enzyme abundance, substrate concentration and $K_M$ can explore a hypothetical population, but cannot establish the percentage of measured enzyme–substrate pairs satisfying a QSSA criterion. That percentage requires matched measurements and an appropriate experimental setting.

Use a survey to set a scale, then obtain matched microscopic and concentration measurements before explaining an individual enzyme or a population fraction.

14Uncatalyzed chemistry supplies another denominator

A catalyst's acceleration over background and its proximity to diffusion measure different achievements.

Wolfenden and Snider compared enzyme-catalyzed reactions with their uncatalyzed counterparts. Some uncatalyzed reactions are so slow at ordinary temperature that their rates require measurements at higher temperatures followed by an explicitly modeled extrapolation.²⁴ The resulting comparisons reveal how large a chemical barrier an enzyme can overcome, while carrying the uncertainty of the uncatalyzed benchmark.

An enzyme can achieve enormous acceleration relative to an extremely slow uncatalyzed reaction and still have a modest product rate compared with molecular encounters. Conversely, a high turnover number alone does not reveal the size of its chemical rate enhancement. The substrates, reaction, standard state and rate units must match the denominator being used.

Transition-state stabilization makes that distinction physical. The thermodynamic-cycle argument compares preferential binding of the transition state with binding of the substrate ground state. Under its rate-theory assumptions, $k_{\mathrm{chem}}/k_{\mathrm{non}}=K_d^S/K_d^{\ddagger}$. This compares chemical barriers. A measured multi-step turnover can instead be limited by product release, and its value should not automatically replace the resolved chemical rate in the relation.

Choose the denominator that matches the question: uncatalyzed chemistry for acceleration, or an applicable encounter coefficient for capture-to-product performance.

15Binding-energy measurements are increments with a reference state

Mutation, confinement and ligand-size studies measure different aspects of molecular recognition.

Fersht and colleagues used protein engineering to perturb particular interactions. Their 1985 tyrosyl-tRNA synthetase study compared kinetic specificity after deleting potential hydrogen-bonding groups. Typical good neutral interactions in their examples contributed 0.5–1.5 kcal/mol, while two interactions with charged substrate groups gave larger effects, about 3.5–4.5 kcal/mol.²⁵ The authors explicitly describe the inferred quantities as incremental transition-state binding energies.

The observable was the change in $k_{\mathrm{cat}}/K_M$, converted to an energy through a logarithmic ratio. It was not a direct measurement of every mutant's equilibrium dissociation constant. This distinction matters when using the paper to motivate the core's contact-count model: the measurements demonstrate variable net interaction energies in water, while the illustrative equilibrium-affinity slope remains an assumption.

Binding also restricts the partners' motion. Finkelstein and Janin's 1989 analysis estimated translational and rotational entropy costs and showed their dependence on residual freedom.²⁶ A confinement estimate provides an offset in standard binding free energy; it does not establish the energy gained by each interface contact.

Kuntz and colleagues' 1999 survey instead compared binding strength with ligand heavy-atom count and described an upper envelope that tended to level off for larger ligands.²⁷ Atom count, deleted interaction and constrained configuration volume are distinct measurements. Together they constrain plausible recognition models without supplying one universal linear rule for affinity.

Identify the experimental perturbation and reference state before translating an energy increment into a claim about affinity or specificity.

16Antibody affinity linked recognition to biological timing

Equilibrium strength became meaningful alongside capture, internalization and antigen presentation.

Foote and Eisen's 1995 discussion examined kinetic and affinity constraints during immune responses. It considered why affinity maturation could encounter practical limits associated with association, dissociation and the time available for antigen handling, and why multivalent viral antigens complicate comparisons with small monovalent ligands.²⁸ These are biological and assay-dependent constraints, rather than a universal thermodynamic upper bound on binding.

The identity $K_d=k_{\mathrm{off}}/k_{\mathrm{on}}$ makes the needed information explicit. At $K_d=0.1\,\mathrm{nM}$, the illustrative association range $10^5\text{--}10^6\,\mathrm{M^{-1}s^{-1}}$ implies dissociation-only residence of roughly 28–3 h. A value of affinity without its association rate cannot select a residence time from that range.

Multivalent attachment adds states and opportunities to rebind. A molecule can lose one contact while remaining tethered by another. The rate of complete escape therefore differs from the rate of breaking a single contact. A very persistent multivalent complex need not imply an equally small monovalent dissociation constant, and the concentration or rate normalized per antigen particle can differ from one normalized per epitope.

Interpret affinity in the context of absolute kinetic rates, assay normalization and the number of binding contacts that must break for escape.

17In-cell diffusion challenged a universal solvent correction

A protein's mobility in cytoplasm is a measurement of that probe under those conditions.

Elowitz and colleagues measured fluorescent-protein mobility in living E. coli. Their 1999 study reported GFP diffusion of about $7.7\pm2.5\,\mu\mathrm m^2/\mathrm s$ in the cell, compared with a much larger aqueous value used in the study.²⁹ This provides a concrete comparison for the water calculation, but does not determine a universal viscosity multiplier for every intracellular species.

Konopka and colleagues subsequently examined crowding and confinement effects on in-cell protein diffusion.³⁰ Size, geometry, binding and physiological state can affect measured mobility. A small metabolite and a protein complex need not experience a single common reduction relative to water.

The model decision also depends on distance and consumption. A diffusion coefficient becomes an exploration clock through a specified length scale. Comparing that clock with pool replacement helps assess a well-mixed approximation; localized consumption or supply requires a spatial balance. The core's reaction–diffusion example makes this dependence visible through the length $\sqrt{D/k}$, where k is a specified first-order removal coefficient.

Use the diffusion measurement appropriate to the molecule and condition, then compare transport with the local process that could create a gradient.

18Timescale separation became an approximation to justify

A short complex transient, a short bound episode and rapid binding equilibrium are different statements.

The Briggs–Haldane balance needs a regime in which substrate changes slowly enough. Segel and Slemrod's 1989 perturbation analysis developed the standard closed-reaction QSSA as a mathematical approximation with an initial transient and a subsequent slower phase. For an assay begun with substrate $S_0$ and free enzyme total $q_E$, the familiar sufficient small parameter is $q_E/(K_M+S_0)$.³¹

Gunawardena's later analysis places enzyme timescale separation in a broader model-reduction setting.³² The condition is attached to equations, initial conditions and an observable. It should not be exported without examination to an open, driven system, a comparable-enzyme-total regime, or an experiment that resolves the very transient the reduction removes.

Two mechanisms can have the same full stationary product curve. With $k_{\mathrm{cat}}=10\,\mathrm s^{-1}$ and $K_M=100\,\mu\mathrm M$, one mechanism can have rapid association and frequent dissociation; another can form complexes more slowly and almost always make product once bound. Their common specificity $10^5\,\mathrm{M^{-1}s^{-1}}$ does not distinguish them. The exposition calculates both microscopic examples and shows that standard QSSA can hold even when rapid binding equilibrium does not.

Justify a reduction for its experiment and observable; a stationary rate curve alone does not identify the microscopic timescale hierarchy.

19Cellular timing surveys compare different kinds of clocks

A catalogue of times is useful when each entry states what begins and ends the measurement.

Shamir and colleagues assembled representative cellular timescales rather than measuring one universal clock. Their 2016 snapshot brings together transport, conformational change, synthesis and degradation examples. Its ligand-induced conformational example is 1 ms; the 0.1 µs entry concerns passage through a channel.³³ Exchanging those labels changes the argument even though both numbers are “fast.”

Pool replacement combines a concentration with a throughput. Bennett's concentration measurements can supply the pool, but an ATP or PEP replacement time also requires a matched or explicitly assumed flux. At steady state, a short pool-to-flux ratio means rapid replacement despite constant concentration. It is a depletion time only under additional assumptions about the loss of supply.

Enzyme abundance has its own response law. Bernstein and colleagues' mRNA decay measurements describe transcript half-lives, not a compulsory wait before any new protein appears.³⁴ For a stable enzyme in constant exponential growth, dilution gives $\dot q_E=\beta-\mu q_E$. After a synthesis step, half of the final concentration change takes one doubling time. A large induction can double the initial concentration sooner.

The related ribosome budget separates serial production from exponential population growth. Dai and colleagues' elongation measurements and Scott and colleagues' growth-law analysis constrain translation and allocation.³⁵³⁶ The time for one ribosome to produce one ribosome's protein complement is not the ideal doubling time of an exponentially expanding ribosome population; the latter contains the factor $\ln2$.

Compare times only after naming their event, normalization and dynamical meaning.

20Feedback experiments identified a fast regulatory route

The case for regulation came from specific biochemical perturbations, beyond a general need for rapid response.

End-product inhibition gave a concrete pathway-level mechanism. Umbarger's 1956 report examined inhibition in isoleucine biosynthesis. Yates and Pardee's study that year examined feedback in pyrimidine synthesis.³⁷³⁸ These experiments connected a downstream product with an earlier enzymatic activity, providing a route for pathway output to affect its own production.

Gerhart and Pardee examined the relation between catalytic and regulatory functions. Their 1962 work on aspartate transcarbamylase showed that treatments affecting regulatory behavior could leave catalytic activity, supporting a distinction between the functions.³⁹ Such evidence is more specific than the statement that binding is often faster than protein synthesis.

Monod, Wyman and Changeux proposed a structural model in 1965. Their model assumed equivalent subunits, concerted transitions preserving symmetry, and ligand affinities that differ between conformational states. Ligand binding can then redistribute the population of states and couple occupancy at different sites.⁴⁰ Those postulates explain a class of cooperative and regulatory behavior; they are not consequences of the timescale ledger alone, nor requirements that every allosteric protein must satisfy.

Use timing to motivate a regulatory hypothesis, then seek perturbations that identify its molecular interaction and coupling mechanism.

21Modern metabolic studies separated candidate mechanisms

Rapid perturbations and comparisons across steady states constrain different aspects of metabolic response.

Link and colleagues combined time-resolved measurements with kinetic modeling. Their 2013 study used metabolic responses to identify candidate allosteric protein–metabolite interactions and test their effect on enzyme activity.⁴¹ Substrate and product changes already alter rates through ordinary kinetic laws, so the inference of a regulatory interaction requires comparing the observed response with those alternatives.

Hackett and colleagues compared 25 steady-state yeast cultures. Their 2016 analysis combined enzyme abundances, metabolites and fluxes to assess how much variation could be attributed to different inputs. In that analysis, metabolites accounted for about twice as much flux variation as enzyme abundance.⁴² This is a result about those conditions and models; it is not a universal percentage of control, and it is not a measurement of a transient response delay.

The core's non-saturating interpretation supplies a useful baseline for both questions. At fixed enzyme and $S\ll K_M$, a proportional substrate change produces an approximately proportional flux change. As the enzyme saturates, that local substrate sensitivity falls. An additional rapid response can motivate a regulatory mechanism, but its identification still requires the relevant metabolite, state or interaction to be tested.

Match the inference to the experiment: transient data constrain response dynamics, while steady-state comparisons constrain variation across maintained conditions.

22The synthesis is a method for choosing the next measurement

The historical programmes jointly explain why one rate constant cannot answer every question about a cell.

Each programme supplied a different link in the calculation. Particle tracking connected random displacement to transport. Mass-action reasoning connected resolved events to concentration-dependent rates. Enzyme assays connected substrate concentration to product formation. Binding and energetic studies distinguished equilibrium occupancy from catalytic performance. Cellular measurements added volume, supply, turnover, growth and feedback.

The quantitative-biology teaching programme exemplified by Phillips and Milo and the BioNumbers database makes these links reusable by preserving units, scale and provenance.⁴³⁴⁴ Lecture 2 applies that practice to a particular chain of decisions: when a pool can be treated as mixed, when an intermediate can be eliminated, and which coefficient predicts flux in the concentration range of interest.

The unresolved quantity tells you what to measure next. To predict non-saturating product formation, measure specificity and enzyme abundance. To explain why specificity is below ideal encounter, measure complex formation and the exits from the complex. To infer affinity, obtain an equilibrium or microscopic-rate measurement. To predict cellular response, add the relevant pool balances, spatial inputs and regulation.

Choose the next experiment by identifying which link between count, encounter, binding, product and cellular response remains unconstrained.

†References

Original papers, verified translations and the measurement studies discussed in the text. Historical notation has been translated into the course's notation where equations are compared.