How a rapid method for detecting non-specific small molecule aggregation can facilitate compound screening
Robert P. Collins, Mid-Atlantic Regional Manager of Wyatt Technology reports:
Introduction
High-throughput screening (HTS) is one of the key methods for the generation of new compound leads in the medicinal chemistry industry, widely used for the identification of novel small molecules that modulate enzyme or protein activity. However, high-throughput screening often cannot distinguish between desirable inhibitors, those that bind specifically and stoichiometrically, and undesirable compounds, those that non-specifically interfere with the targeted activity.
There are several explanations for the mechanisms underlying the undesirable, promiscuous inhibitors interfering with the desired activity: non-stoichiometric binding; protein unfolding; and/or inactivating of proteins, among others. The non-stoichiometric binding compounds in particular are problematic as “false positives”; compounds which appear to inhibit the target yet produce confusing results in ensuing tests and functional biological assays, and ultimately fail as drug candidates.
If undetected early in the development cycle, these promiscuous, non-stoichiometric inhibitors can be a serious problem in the HTS of large libraries of compounds, representing costly false leads in the drug development cycle. The implications of undetected false positives in the competitive sphere of industrial drug discovery means measures need to be taken to identify non-stoichiometric binding inhibitors: the reliable identification of such compounds and their elimination from HTS and hit lists early in the development cycle would save considerable time and resources during hit and lead optimization.
Facilitating screening hits selection
Within drug discovery, the concern surrounding small drug-like compounds that do not fit the classical 1:1 binding inhibition behaviour is the possibility that they act through the formation of compound aggregates. Determining whether compounds aggregate and lead to promiscuous non-stoichiometric inhibition is critical for prioritising the progression of certain compounds through the product development process. If these promiscuous inhibitors are not identified quickly enough, their behaviour can lead to the initial progression of compounds with undesirable properties, weakening the desired binders and hindering potential drug discoveries.
According to one widely accepted model (McGovern et al, J. Med. Chem. 2003) promiscuous inhibition is connected to the formation of micelle-like aggregates of the compound, with the aggregates typically ranging in size from tens to hundreds of nanometers. Most drugs are not promiscuous, even at high concentrations. Nevertheless, at high enough concentrations (20-400 microMolar, or lower), some drugs aggregate and act promiscuously, suggesting that aggregation may be present among small molecules at micromolar concentrations, at least in biochemical buffers.
A method to detect the formation of sub-micron particles from drug candidates would greatly facilitate screening hits selection. Dynamic light scattering is one technique for detecting sub-micron particles, helpful for identifying promiscuous non-stoichiometric inhibitors, and in a high throughput format, is particularly important in helping progress the discovery of lead compounds.
Dynamic light scattering
Dynamic light scattering (DLS) is an established technique for determining the size of submicron particles, such as proteins and other biomolecules, liposomes, nanoparticles and compound aggregates. Widely used in biochemistry, biotechnology and pharmaceutical development, DLS is applicable wherever the size and size distributions of macromolecules and nanoparticles need to be measured quickly and easily, with little sample volume or material.
DLS is a non-invasive technology that measures samples directly, without perturbation, offering fast and accurate results in seconds. The technique determines size distributions without fractionation or dilution, and provides an estimate of the heterogeneity or polydispersity of the sample, in addition to the hydrodynamic radii.
Using an automated DLS instrument to identify compounds that exhibit a promiscuous non-stoichiometric trait, like aggregation, could accelerate productivity in an efficient and effective way. Low-volume, automated DLS measurements can be performed directly in standard 96, 384 or 1536 well microtiter plates. Beyond applications in HTS, a plate based instrument reveals new horizons for the method of dynamic light scattering. Some of the other high-throughput screening applications where DLS is advantageous include quantifying the type and degree of aggregation in hundreds of buffer conditions, important in formulation and crystallization studies; pre-formulation of select biotherapeutic candidates to evaluate developability; assessment of buffer excipients and pH for colloidal stability and thermal (conformational) stability; and analysing thermal stability and chemical denaturation, differentiating pure unfolding from aggregation.
DLS in action
As previously described, a rapid method of detecting compound behaviour when screening libraries of compounds, would greatly facilitate screening hits selection. In order to develop such a method, researchers at biopharmaceutical companies can utilise an automated dynamic light scattering instrument (for instance the DynaPro DLS Plate Reader, Wyatt Technology) to examine the light scattering properties of a dilution series of known aggregating inhibitors.
The DLS Plate Reader instrument is ideally suited for companies who require a high throughput technology to examine the light scattering behaviour of compounds, such as known promiscuous inhibitors. The application example below highlights how one such company employed the DLS Plate Reader in order to aid them in their research study.
Method
The research team created a plate with a three-fold compound dilution series in duplicate along with two DSMO (non-compound) controls. A 96-well format plate was selected with a sample volume of 100 μl.
Each well contained 99 μl of 50 mM potassium phosphate buffer, pH7 (filtered through a 0.2 μm filter) and 1 μl of either DSMO or the small molecule dilution series. The plate was subjected to a brief centrifugation prior to reading in order to remove any bubbles.
Due to the heterogeneous nature of the samples, the researchers used short acquisition times with multiple reads per well to allow for averaging of the population.
Results
The DLS instrument helped the researchers identify several trends in compound behaviour versus the concentration of the compound. In general, the total intensity in scattering signal detected by the DLS instrument decreased with decreasing compound concentration (Figure 1). For aggregating compounds, the decrease was dramatic, whereas for non-aggregating compounds the trend was subtle. For the DMSO controls, there was no trend in the signal from well to well.
Figure 1: Figure 1 Plate layout and representative results. Lanes A-H are 3-fold serial dilutions of compounds starting at 100mM (row A). Lanes 1,12 DMSO controls; 2,3 Quercetin: 4, 5 Miconazole; 6,7 Clolrimazole, 8,9 Rottlerin; and 10, 11 I4PTH. Measurement of R, hydrodynamic radius in nm. Colors black, out of range; blue, fully saturated signal, graded to red, minimum signal; white, no sample/undetectable scattering signal.
When examining the amplitude of the autocorrelation function, the raw data acquired by DLS, an increase in the amplitude with increasing concentration of small molecules was observed. As the concentration of a compound decreased below a threshold, the total intensity and the amplitude of the autocorrelation reverted to buffer-like behaviour (Figure 2), indicating a lack of particles or aggregates.
Figure 2: Figure 2 Comparison of Buffer to Rottlerin. A) Light scattering intensity autocorrelation for buffer with DMSO alone. B) Light scattering intensity autocorrelation for the dilution series of Rottlerin. The amplitude of the signal decreases with decreasing compound concentration (inset plot).
An increase in normalised intensity was correlated with increasing small molecule concentration (Figure 3), with the intensity increasing exponentially upon formation of large particles or aggregates above a critical concentration.
Figure 3: Correlation of Normalized Intensity and promiscuous inhibitor concentration. The scattering intensity was significantly larger for aggregate forming compounds.
Conclusion
The research team successfully utilised automated DLS instrumentation to examine the concentration dependent light scattering behaviour of compounds, and distinguish the light scattering behaviour of known promiscuous inhibitors and non-aggregation forming controls. As a result, the DLS method can now be extended to a wider range of compounds and to study the effect of enzymes on aggregate formation.
Automated DLS can help to increase productivity when sizing biomolecules and other nanoparticles compared to conventional systems. DLS provides data that are a true measure of sample repeatability and reproducibly, which is crucial for high-throughput applications such as screening compounds as well as for estimating populations of aggregates for small molecules, proteins or liposomes.
To learn more about Wyatt Technology, please visit www.wyatt.com or e-mail info@wyatt.com
