Let there be light

Dynamic light scattering is a widely used technique for measuring the size of molecules in solution, but it can do much besides. Here, we take a look at the protein scientists’ best friend in detail

FROM the late 1980s commercial Dynamic Light Scattering (DLS) detectors started being used in commercial biophysical laboratories throughout the world. Earlier examples of their use had been reported but various technical restrictions had limited their use and application. These instruments enabled size measurements of small volumes of proteins and other biomaterials in solutions to be made, with small sample volumes and in relatively low concentration ranges.

The ability to make these measurements in a relatively straightforward way, or in any case with complete recovery of the sample, soon made these instruments very popular. This is particularly true in protein studies where DLS instruments were used to screen proteins to predict if the protein was likely to crystallise before such a lengthy procedure was carried out. Other issues of stability, aggregation, complex formation and conformation in bio-molecular research were also studied using these instruments.

Early versions of these instruments were beset with problems of reliability, reproducibility, contamination, poor software and a lack of understanding of the instrument limitations, but as the instrument optics, lasers and software improved so did the reliance put on this technique. This has resulted in the situation where the industry leading DLS instruments can now be found in over 80% of the worlds biophysical characterisation laboratories.

Figure 1: Results path with %Pd being polydispersity reading

Dynamic light scattering works by measuring the intensity of light scattered by the molecules in the sample as a function of time. When light is scattered by a molecule or particle some of the incident light is scattered. If the molecule was stationary then the amount of light scattered would be a constant, however as all molecules in solution diffuse with Brownian motion in relation to the detector there will be interference (positive or negative) which causes a change in this intensity. The faster the particles diffuse, the faster the intensity will change (if the light was bright enough this would be seen as a twinkling effect). The speed of these changes is therefore directly related to the motion of the molecule. The diffusion of the molecules is essentially controlled by the following factors; temperature – the higher the temperature the faster the molecules will move. Viscosity of the solvent – the more viscous the solvent the slower the molecules move. The size of the molecules – the bigger the molecules, the slower they move.

If the temperature and solvent are constant and known, the variation in the intensity of the scattered light is directly related to the size of the molecule. This number is referred to as the Hydrodynamic radius, Rh. 
This random pattern of light intensity changes into an Rh measurement by comparing the changes as a function of time. Thus two plots of the same data are essentially overlaid but with a small time delay between them. The correlation between the first and second plot is then noted. This correlation is then calculated for ever increasing time periods until no correlation is present (i.e. the changes in intensity are random in respect to each other). The faster the molecules are moving, the quicker the correlation becomes zero.
Once the correlation function has been plotted (normally on a log scale time base) this is converted into a size measurement by drawing in all the other factors that affect the signal such as the temperature, solvent viscosity, laser wavelength as well as the Rh. 

This process becomes even more complex when more than one size material is present. This results in a correlation function that is essentially a mix of the functions from each species present but strongly biased towards the bigger species (the larger the molecule the greater the scatter). It is possible to derive the individual correlation functions but there are an almost infinite number of ways of arriving at the final function unless some assumptions are made for the distribution of material. The mathematical impasse is overcome by assuming that either the distribution is monomodal (i.e. only one species is present) or that if more than one species is present the distribution of each of these obeys certain limits. Good DLS software will allow both of these results to be displayed. The theoretical limit of this technique is that species of less than 40% difference in size cannot be resolved, however in practice a factor of 2-3 times can be required. If full resolution is necessary the sample must be physically separated by centrifugation, filtration, HPLC or Field Flow Fractionation techniques.
The hydrodynamic volume is the sphere defined by the molecule rotating in all directions plus the hydration layer, modified by how easy it is to pass the solvent through that volume. It is actually a measure of how easy it is to move the molecule through the effluent. Examples of the use of DLS include:

Protein Crystallisation
Dynamic Light Scattering has a highly diverse range of applications but the use that most often occurs in publications is for protein screening prior to crystallisation. This application involves simply making a reading of the protein in the relevant buffer and observing if any aggregates are present and/or if the main monomer peak is polydisperse (indicating that more than one species is present but not resolved). Figure 1 shows the results path with %Pd being the polydispersity reading. These readings should be an integral part of the results displayed by the software.
A similar experiment can be carried out for samples prior to NMR analysis. The presence of aggregates can affect the NMR signal and cause problems in peak assignment. If the sample is shown to be complex before NMR analysis, the sample can either be further purified or the knowledge of the presence of these species can be used to assist assignment and quantification.

Figure 2: "Spectral view" of the DynaPro plate reader system. This allows the user to allocate ranges for each sample and the spectral view will show which of these meet the specification range

Micelles and Liposomes
The recent interest in Micelles and Liposomes for carrying other materials (especially drug delivery systems) has made the study of these materials highly relevant to many research areas. The nature of the particles formed (size, range of sizes etc) can obviously be studied but more fundamental characteristics are also open to analysis.

Automation
The rate limiting step in most DLS applications is the cleaning and filling of the cuvette that the sample goes in. If dust or air bubbles appear in the solution these swamp the signal from any polymer molecules present and ruin the result. In order to speed up the analysis or to examine a matrix of samples in one session it is possible to use an industry standard well plate (96 or 384 wells) instead of the normal quartz cuvette.
The draw back of this system is that the sensitivity decreases (by about 10 times as the optical system is a mass produced plastic plate rather than a custom built quartz cuvette) and the sample volume increases from around 15μl to 45-100μl depending on the well size. However the advantages of speed, convenience and reduced operator input make this well worthwhile.
Figure 2 shows the “Spectral View” of the DynaPro plate reader system. This allows the user to allocate ranges for each sample and the spectral view will show which of these meet the specification and where they lie within the range and also those samples that fall outside the specified range. This has obvious advantages in scanning hundreds of samples with a simple colour-coded view to screen the results. The individual readings are still available should a more in depth analysis be required.
Additional automation has been achieved with the use of liquid handling robots. These prepare the plates and then place it on the DLS plate reader for subsequent analysis.

Dynamic light scattering is a widely applied technique for measuring the effective size of molecules in solution. Using mathematical algorithms it can also be used to determine the presence of a mixture of species without physical separation. Molar masses can also be estimated from the particle sizes measured but the accuracy of these results is always subject to the model used.
The applications of the technique are wide ranging but include many of the fastest growing areas of research – proteins, biomaterials, liposomes and micelles as well as the established uses for particle sizing and aggregation detection. Batch mode measurements can be made with as little as 10µl of sample at a concentration of 0.1mg/ml (less for a larger molecule or particle). Alternatively the experiment can be automated with the use of well plates to hold the samples.

Christoph Johann (right) is Managing Director of Wyatt Technology Europe GmbH, in Germany. Kevin Jackson (left) runs Wyatt Technology UK.