Applications
Biochemical Applications >> Aggregate and Promiscuous Binding
Small drug-like compounds that do not fit the classical 1:1 binding inhibition behavior have been described and may act through formation of compound aggregates. Determining whether these aggregates cause promiscuous inhibition or more desirable target-specific inhibition is critical for prioritizing compound progression through the drug development pipeline. The BIND® platform is a label-free detection technology that produces highly quantitative data with kinetic readout and is uniquely positioned for characterizing the time course, stoichiometry, specificity and mechanism of action of small molecule inhibitors.

Application Features
  • Low to Ultra-high Throughput Formats
  • Rapid Assay Development with Plug-and-play Chemistry
  • Equilibrium Binding Measurements Quantify Stoichiometry and Affinity
  • Short time-courses Provides Information on Binding Specificity
  • Identification of Mechanistically Unacceptable Compounds in Primary Screening
Small Molecule Aggregation
Small drug-like compounds (<600Da) that do not fit the classical 1:1 stoichiometric target association model have been identified. Particular attention has focused on a behavior that is associated with formation of small compound aggregates (size 30-400 nm) that can interact with protein surfaces and thereby inactivate targets. Some compounds displaying this behavior inhibit a wide range of different proteins, and hence have been termed “promiscuous”, though other compounds can show remarkable specificity and potency. Such behavior can lead to the initial progression of compounds with undesirable properties or conversely result in the omission of weaker but more desirable binders. Measurements based on effects of function have been used to distinguish such behavior from classical inhibitors, and also aggregation itself can be used.
Stoichiometry
For most targets with Kd in the range 0.1->1000 µM, the association rate constant for a drug-like small molecule binding to a target will be upwards of 106 M-1s-1, and often 107 or higher. At 106 M-1s-1, with 10 µM compound, the on-rate would be 10 s-1, corresponding to a half-time of association of just 60 msec. Thus, binding is complete within 1 sec, and well before the first read after addition. Any slower binding is indicative of either non-specific binding, or a rate-determining structural change in the target or ligand. Figure 1 depicts binding of various small molecules at a single concentration measured over time using a label-free BIND assay. Upon analysis, several categories of binders can be identified with different binding stoichiometry.

Time Course Analysis of Binding

Figure 1- Mechanistic insight from time courses of binding. This figure shows examples of BIND time courses with mechanistically acceptable binding profiles and others that have varying levels of non-specific interaction. The plots shown represent examples of different categories (A,B,C,D) of compound binding responses commonly seen during a single concentration screen. The insert provides a closer look at the binding time courses for compounds in categories C and D.
Binding Characterization
Generally, well-behaved compounds produce a dose-response curve that shows saturation with a stoichiometry consistent with 1 molecule of ligand bound per binding site and a Kd compatible with solution measurement. Higher apparent stoichiometries suggest an element of non-specific binding.

Dose-response Curve Analysis

Dose-response assays can be run using 96-, 384-, or 1536-well BIND Biosensors and the data fitted to appropriate 1:1 binding molecules to derive Kd and stoichiometry. Moreover, appropriate models can be used which allow the discrimination of a saturable, specific binding component from any non-specific responses (Figure 2). Such data can be obtained by titrations run on target-coated biosensor plates and do not require reference or control surfaces.
Figure 2 - Use of dose-response curves to give Kd , stoichiometry and detect non-specific binding. A). Example of a compound which shows only specific, saturable binding. The solid line is the fit to a 1:1 binding model with a Kd of 1.2 µM and a stoichiometry close to 1:1 (1:1 stoichiometry is 26 pm). B Example of a compound which binds specifically at low concentrations around its Kd, but shows additional non-specific binding at and above its Kd. The full-dose response curve does not show saturation and goes above 1:1 stoichiometry (1:1 stoichiometry is about 70 pm). The solid line shows how such data can readily be fitted to a model based on 1:1 binding combined with a linear function to represent a non-specific component to yield the Kd (4 µM) and stoichiometry of the specific binding component. The dashed line shows the predicted binding curve for specific binding with 1:1 stoichiometry.
Binding Specificity
When a standard binding compound that occupies a specific site is available, or is discovered during the screening process, BIND provides a very effective means of distinguishing specific site binders from non-specific responses.

Promiscuous Binders

Promiscuous binder can be identified by comparing binding of test compounds to target-coated BIND biosensor to uncoated reference biosensors. Binding specificity of both agonist and antagonist can be characterized using such an assay (Figure 3).
Figure 3 – Single concentration of known agonists, antagonists or test samples were incubated with BIND biosensors either coated with target protein or bare biosensors. Promiscuous binders show a high level of binding to both target-coated and uncoated biosensors.

Specificity Assessed Using Site-specific Competitors

BIND assays using a pair of target-coated biosensors and site-specific competitors enables examination of binding specificity. The known site binder is added to one of the biosensors at sufficient concentration as to prevent binding by test compounds at that site. The set of test compounds is then added to both plates. The difference in response between the two plates to the test compounds is a direct measurement of specific-site binding (see Figure 4)
Figure 4 - The ATPase domain target was immobilized in all biosensor wells and saturating amounts of a reversible nM specific site-binder was added to half of the plate. Test compounds were added in duplicate to each half of the plate and binding of test compounds to the target (blocked or unblocked) was measured on BIND. A plot of the signal on blocked vs. unblocked is useful to determine same site binding.