Optimizing SEC for Biologics Analysis

Mar 01, 2018
Volume 36, Issue 3, pg 218

Parameters such as pore size, column dimensions, temperature, flow rate, and mobile phase are important to consider when developing robust size-exclusion chromatography (SEC) methods because many of these can impact the level of aggregation detected.

Choosing the optimum pore size for your molecules of interest is very important. You need to ensure the protein monomers and dimers are physically able to permeate the pore structure to obtain a separation. One rule of thumb is that the pore size of the column should be three times the diameter of the molecules of interest. If the pore size is too small, the protein molecules will be excluded from the pores and will be eluted in the void volume of the column, which will result in inaccurate quantitative data. Conversely, if the pore size is too large all of the proteins will be able to fully permeate the particles and there will be very little separation. Because the choice of pore size influences the resolution obtained when using SEC, testing a range of pore sizes to match this to the analytes is worthwhile.

Column internal diameter affects the flow rate and injection volume. There are two common column internal diameters in SEC: 4.6 and 7.8 mm. Since the separation mechanism is purely based on diffusion into and out of the pores of the stationary phase, the greatest separation comes from having larger column sizes. Using a slow flow rate allows the molecules sufficient time to diffuse into and out of the static pool of mobile phase contained within the pore structure. The normal flow rate for a 7.8-mm i.d. SEC column is 1.0 mL/min. This translates to 0.35 mL/min when using a smaller 4.6-mm i.d. column. The internal diameter difference also means the amount of sample injected on a 4.6-mm i.d. column can be reduced by a similar amount (~33%), which is useful if you have a limited amount of sample available. It is important to recognize that operating 4.6-mm i.d. columns at 0.35 mL/min can lead to differences in performance related to the extracolumn dead volume in the system; peaks can become broader if long capillaries with wide bores are used to connect the injection valve to the column or the column to the detector. Longer columns provide more resolution, but require longer run times. Shorter columns produce shorter run times, greatly increasing throughput (for even faster separations, use higher flow rates). Since separation relies on the available pore volume, using longer columns or multiple columns in series increases the available pore volume and therefore increases resolution. Going from a 30-cm column to a 15-cm column means the run time can be cut in half. As long as you still have the required amount of resolution, using a shorter column can greatly improve sample throughput. Sample throughput may be particularly important to you if you are screening multiple samples during early development phases, or taking regular measurements from a fermentation.

Temperature is sometimes overlooked in simple approaches such as the isocratic methods used in SEC. Methods often state the temperature simply as ambient. However, it is highly desirable to use a column oven if you are looking at ensuring good reproducibility. In a laboratory environment where the ambient temperature could change more than 10 °C during the course of the day or night, you will see a noticeable impact. This difference in temperature will change the viscosity of the mobile phase significantly, which in turn will change the column operating pressure, and the diffusion process into and out of the pore structure will also change. The temptation is to increase the temperature—higher temperatures will mean significantly lower viscosity, much lower operating pressures, and much faster diffusion, giving sharper peaks and better resolution. However, if the temperature is too high, more aggregation is likely to result—precipitation of the sample before analysis may even occur because of exposure to excessive temperature.

Chromatographers who work with reversed-phase separations are used to operating at high flow rates and achieving optimum plate counts for small molecules (for example, 1.1–1.2 mL/min on a 300 mm × 7.8 mm column). However, when you start to look at the column efficiency for larger molecules, such as proteins, the optimum flow rate is much lower (in this case, 0.6 mL/min compared to 1.2 mL/min for the previous example). This flow rate difference of course means that the run times will be considerably different.

Mobile-phase selection can have a noticeable effect on some proteins, with differences in ionic strength, pH, and buffer composition resulting in changes in resolution, selectivity, and peak shape. It is therefore essential to consider what effect even minor changes in buffer composition may have to demonstrate method robustness and method optimization. Particular care needs to be taken with detergents and other denaturants because they can cause proteins to unfold and become larger in solution, or can bind to such an extent that molecular weight and size in solution increase dramatically, which leads to shorter retention times. Different columns might behave differently too; most are silica-based and it is common to see undesirable interactions occurring at low ionic strength. However, as you increase ionic strength you may also begin to see other effects; hydrophobic interactions may start to occur as you move toward conditions that begin to look like hydrophobic interaction chromatography.

lorem ipsum