Binding, recognition and structure-function relationships

A binding contact does not specify all of recognition

Molecular recognition involves interactions that favor particular associations under specified conditions. A compatible arrangement of groups can help explain a binding hypothesis, but a drawing alone supplies neither a measured affinity nor the full functional response. Solvation, protonation, conformational populations and other partners can affect association.

Begin with the hypothetical reaction P+L ⇌ PL, where one protein site binds one ligand and no other states are included. In the ideal concentration convention, define K_d=c_P c_L/c_PL. This concentration parameter has concentration units. A smaller value favors more binding at the same maintained free ligand concentration within this model.

The actual molecular system might have multiple sites, competing ligands or coupled conformations. These are changes to the model, not details that can automatically be ignored because a curve looks saturating.

Finite totals require a finite binding balance

Invent protein total P_T=5.00 μmol L⁻¹, ligand total L_T=6.00 μmol L⁻¹ and K_d=2.00 μmol L⁻¹. Let x be complex concentration. The balances give free concentrations c_P=P_T−x and c_L=L_T−x, so K_d=(P_T−x)(L_T−x)/x. All concentrations in the numerical equation use the same unit.

With numerical concentrations in μmol L⁻¹, rearrangement gives x²−13x+30=0. Its roots are 3 and 10. Only x=3.00 μmol L⁻¹ satisfies 0≤x≤min(P_T,L_T). Free protein is 2.00 and free ligand 3.00 μmol L⁻¹, and protein occupancy is x/P_T=0.600.

Using the familiar free-ligand hyperbola with L_T instead of c_L would give 6/(2+6)=0.750. That ignores ligand depletion in this finite inventory. The square comparison markers in the plot show precisely that substitution error, while the circle markers and dashed line retain the physical quadratic balance. When depletion is negligible a justified approximation may be available; it is not valid merely by changing the label “total” to “free”.

Free reservoirs lead to a different normalization

For a different stipulated equilibrium, expose one exclusive site to maintained free A and B concentrations of 10.0 μmol L⁻¹ each. Assign K_d,A=5.00 and K_d,B=50.0 μmol L⁻¹. The site has three states: unoccupied, A-bound or B-bound. Double occupancy is excluded.

Relative weights are 1, c_A/K_d,A=2.00 and c_B/K_d,B=0.200. Dividing by their sum 3.20 gives fractions 0.3125 unoccupied, 0.6250 A-bound and 0.0625 B-bound. They add to one. The A-bound/B-bound ratio is ten for these equal maintained free concentrations.

Free reservoir concentrations are essential. If A and B totals were finite and depleted, their free concentrations would need coupled balances rather than this direct substitution. The calculation describes neither irreversible capture nor universal biological specificity; changing partners or conditions can change the comparison.

Compare compatible standards using a dimensionless ratio

For the ideal 1:1 convention with common standard concentration c°, dimensionless association constant K_assoc°=c°/K_d. Standard binding free energy is ΔG_bind°=−RT ln K_assoc°=RT ln(K_d/c°). The dimensional K_d itself is not placed directly inside a logarithm.

Comparing A and B under the same standard cancels c°: ΔG_bind,A°−ΔG_bind,B°=RT ln(K_d,A/K_d,B). Use assigned K_d values 5.00 and 50.0 μmol L⁻¹ at 298.15 K. Their dimensionless ratio is 0.100, giving −5708.00950 J mol⁻¹, or about −5.708 kJ mol⁻¹.

This standard comparison makes A binding more favorable in the compatible assigned model. It is not the actual free-energy change in a cell, a kinetic barrier or a functional efficacy difference. Different coupled protonation, solvent or standards would require explicit treatment rather than an automatic comparison of reported numbers.

A dissociation time is not an equilibrium constant

Take an independent hypothetical no-rebinding first-order dissociation model with k_off=0.0200 s⁻¹. If S(t) denotes fraction still bound after time t, dS/dt=−k_off S and S(0)=1 give S(t)=exp(−k_off t). The mean exponential residence time is τ=1/k_off, while the half-time is t_½=ln 2/k_off.

The mean is 50.0 s and half-time 34.6573590 s. At 100 s, surviving bound fraction is exp(−2)=0.135335283. Mean and half-time differ because they describe different properties of the same exponential distribution. Neither implies that every individual complex dissociates at that exact time.

Rebinding and additional states can change an observed dissociation record. The assigned k_off alone does not establish equilibrium affinity; association information and a valid kinetic scheme would also be required. A slow loss of binding does not by itself demonstrate a cellular function.

Test the association and the proposed function separately

Structure can suggest why a partner fits or which groups participate in recognition. A binding measurement then needs evidence about molecular identities, active concentrations, conditions and alternatives. Surface attachment, labels, multivalency and other measurement features can change what an apparent binding parameter means.

A functional claim asks a further question: does the association produce the proposed response in the actual system? Appropriate controls and complementary observations are needed. The finite balance, exclusive-state normalization, standard affinity comparison and no-rebinding lifetime model here are separate tools for asking that question precisely; their invented numbers are not biological performance measurements.

Worked examples

Try it yourself

1. For ideal 1:1 binding, P_T=4.00, L_T=3.00 and K_d=2.00 μmol L⁻¹. Find complex, free ligand and protein occupancy.

   Show the answer and reasoning

   Complex 1.62771868, free ligand 1.37228132 μmol L⁻¹; occupancy 0.406929669.

   The physical root is (9−√33)/2 in the numerical μmol L⁻¹ convention. Using total ligand in the hyperbola would incorrectly give 0.600.

2. An exclusive site has maintained free A=5.00 and B=25.0 μmol L⁻¹, with K_d,A=5.00 and K_d,B=50.0 μmol L⁻¹. Find the three state fractions.

   Show the answer and reasoning

   Unoccupied 0.400, A-bound 0.400 and B-bound 0.200.

   Normalize weights 1, 1 and 0.500 by 2.50. Finite totals would need a different coupled balance.

3. At 298.15 K under the same compatible standard, change K_d,A to 20.0 and K_d,B to 5.00 μmol L⁻¹. Find ΔG_bind,A°−ΔG_bind,B° and identify the favored standard binding.

   Show the answer and reasoning

   +3436.56415 J mol⁻¹; B has the more favorable standard binding.

   RT ln(20/5) is positive. The smaller K_d,B means a more negative binding standard free energy in this model.

4. For first-order k_off=0.0400 s⁻¹ without rebinding, find mean residence, half-time and bound survival after 25.0 s.

   Show the answer and reasoning

   Mean 25.0 s; half-time 17.3286795 s; survival exp(−1)=0.367879441.

   Use the three different definitions. The dissociation rate alone does not determine equilibrium affinity or a biological response.

What to carry forward

• Free ligand concentration and finite ligand total are not interchangeable.
• Reject roots that violate the actual inventories.
• Exclusive state fractions require the specified free reservoirs.
• Standard affinity comparisons require compatible standards and dimensionless ratios.
• Residence, half-time, equilibrium affinity and function answer different questions.