Bioanalysis Technology Forum

Bioanalysts are faced with a variety of separations challenges, including how to select appropriate stationary phases and problems with sample preparation and cleanup. Participants in this Tech Forum are Maureen Joseph and Linda L. Lloyd of Agilent Technologies; Patrik Appelblad, Ling Bei, and Dave Lentz of EMD Millipore; and Jason P. Weisenseel of PerkinElmer.

What impact have sub-2-µm and core–shell particles made in bioanalysis?

Joseph and Lloyd: Driven down the total analysis time for biocolumns – for protein characterization we are now seeing separations in 1–2 min where previously we could have been looking at 20–30 min. With large biomolecules that have slow diffusion coefficients the reduction in interparticle distance is most dramatic for reduction in band spreading. For peptide separations, the core–shell columns and the sub-2-µm columns have added choices for high peak capacity separations in far less time — separations that were 1–2 h are now 20–30 min with greater sensitivity. For separations of analytes in biological matrices, the sub-2-µm and core–shell particles have provided dramatic improvements in efficiency. With that has come the ability to resolve more components in one analysis with these complex samples and to do it with greater speed and sensitivity. This has helped improve sensitivity in liquid chromatography–mass spectrometry (LC–MS) and LC–MS-MS analyses. The core–shell particles are particularly rugged for the analyses in biological matrices. These columns are designed to be more resistant to plugging than the sub-2-µm columns, due to the standard frit used in these columns.

Appelblad, Bei, and Lentz: Groundbreaking novel column technologies will always inspire new ideas and pump energy into the daily routines of chromatographers. The new generation of ultrahigh-performance sub-2-µm, core–shell, and also monolith (non-particulate) products is certainly no exception. We’re getting closer to the top of the mountain as each year passes. The more daring chemists, the early adopters (too often the too-small minority in the relatively conservative high performance liquid chromatography [HPLC] community), will race to embrace them and be rewarded with new capabilities that eventually send “traditional” methods down memory lane. But owning a 21st-century column is sometimes — unfortunately — like buying a Ferrari that goes 200 mph when you live in a 25-mph town.

These column innovations will continue to slowly gather their followings as users try to fit these state-of-the-art quantum-leap improvements into the existing limitations of all the ancillary processes that support the columns as the heart of the overall separations process. Avoiding these roadblocks requires matching the benefits as well as the disadvantages of each new technology to your current bioanalysis resources as well as your goals.

In the world of bioanalysis, there are very specific needs and specific roadblocks to adopting specific latest-and-greatest column technologies. For example, there is often a need to separate metabolites and the parent compound simultaneously. So peak capacity will probably play an important role. The more plates in the column, the better chance to get good results — let’s use smaller particles and our problems will be over. And let’s hope our system can accommodate that. Even if it can, however, too much resolution is not only a waste of costly chromatographic resources but can even be counterproductive and sometimes just plain wrong.

Since bioanalysis often involves dirty, complicated samples, more up-front sample prep is typically needed. This is compounded by the fact that achieving adequate resolution primarily through smaller particle size also requires other physical changes in the column such as smaller inlet frits, which also puts even more demands on sample prep thoroughness and can dramatically shorten column lifetime. For these reasons, many chemists prefer to stay with 3–5 µm material anyway to avoid the need of more extensive sample preparation, additional sample filtration, and even HPLC system replacement. So, often the best approach is to rethink everything — challenge all assumptions. For example, chromatographers can now get ultrahigh performance without ultrahigh pressure. And reversed-phase particulate column technology, after decades of dominance, has not been the only game in town for a long time.

I’m sure almost all of us would agree that the right chromatographic selectivity is a key component for quantitative bioanalysis. Besides the ubiquitous reversed-phase leader C18, hydrophilic interaction liquid chromatography (HILIC) has been gaining popularity dramatically in the past 10 years. As the fastest growing area in chromatography, HILIC provides excellent selectivity to molecules with small or negative logP, which is complementary to reversed phase. Many bioanalysis endogenous materials are polar, and are a much better fit for HILIC. Using high organic ratios in HILIC mobile phases also increases sensitivity, especially in mass spectrometry (MS) detection. Validated methods for quantitative bioanalysis using zwitterionic HILIC ligands have been set up as routine tests in clinical labs, breaking the old paradigm that polar packings and reproducibility are not very good companions. Again, it’s all about challenging those old assumptions.

Weisenseel: The major impact of sub-2-µm and core–shell columns on bioanalysis is reduction in analysis time and improvement in throughput. These columns allow you to cut run times by 75% or more when compared to 3- or 5-µm particle size columns. Additionally, the high efficiency of these columns can often improve the separation of analytes from interferents that can cause ion suppression in LC–MS applications.

The major advantage of both sub-2-µm and core–shell particles is the high column efficiency and reduced plate heights that result in narrow peaks, which allow much shorter run times. Unfortunately, the sub-2-µm ultrahigh-pressure liquid chromatography (UHPLC) columns have very high back pressure when compared to 3- and 5-µm particle size columns. These back pressures often exceed the pressure limits (400 bar or 6000 psi) of most analytical HPLC systems, making it necessary to invest in high-pressure, low-dispersion UHPLC systems. These systems have pumps capable of pressures in the range of 15,000–20,000 psi (1035–1375 bar) and have low dead volumes to minimize peak dispersion.

Core–shell columns have efficiencies often equivalent to sub-2-µm UHPLC columns at about 50% of the back pressure. In many cases this allows one to use these columns on a traditional analytical HPLC system, minimizing the need to purchase new systems. The only drawback of most analytical HPLC systems is that the dead volume of the system is often not optimized for use with these high efficiency columns, so there is excessive band broadening. This can be easily overcome by replacing the current tubing in the system with low-volume, small-i.d. tubing and replacing flow cells with low-volume, high-efficiency flow cells. Additionally, core–shell columns are generally more rugged than sub-2-µm columns so that column lifetimes are significantly improved.

If these high-efficiency particle formats have improved throughput, where is the bottleneck now? What improvements are needed in separation, sample cleanup, and detection?

Joseph and Lloyd: For biomolecules this has to be in the sample preparation and data interpretation. The quality of the separation is good now, but additional gains can be made in resolution by considering 2-D separations with high-speed, high-resolution sub-2-µm and core-shell column formats. Sample cleanup will benefit from more on-line methods, and these and 2-D separations will benefit from greater simplicity in setting up these configurations.

Appelblad, Bei, and Lentz: Certainly, more efficient columns can increase throughput, but they tend to shift the bottleneck to other areas-especially sample prep. One can certainly continue to use traditional particulate technology but with smaller sizes to achieve the higher efficiency and throughput desired. But then one needs to use narrower filters and more rugged sample preparation protocols to make sure a higher proportion of the matrix is removed before injection of samples or there is a big risk to only get a limited amount of injection per column. This will all together add to the cost per injection. Certainly in labs with a very high sample volume, sample preparation can be dealt with by using robot handling, 96- or 384-well plates, but in smaller labs it means more work on sample prep, more spending on consumables, and more efficient sample preparation units needed.

Higher throughput certainly takes its toll on data-mining activities. The higher the number of samples per day, the more data generated per time unit, and the more work is needed to do data analysis per working day. The good news is it is much easier to do this off-line and not allow it to be a rate-limiting step.

One always needs to balance throughput increases against the total analysis economy. The more unit operations added, the higher the cost per injection. But at the same time, manual one-at-a-time syringe-type filtrations will add significantly to the total expense of the assay even though they answer a lot of needs in bioanalysis. This is counterproductive to the overall goals of the method. Automation is often not considered for smaller labs with 10–100 samples per day due to the variety and volume of samples preventing enough justification for a mass process. Yet much-smaller-scale vacuum-driven bench units that handle even less than a dozen simultaneous samples will still dramatically cut down overall manual prep time. Added benefits are higher yields from lower hold-up volume and less waste created. Again, the lesson is that assumptions can be dangerous.

Weisenseel: The bottleneck is now the time and manpower required for sample preparation. Sample preparation techniques such as precipitation, desalting, liquid-liquid extraction, solid-phase extraction (SPE) and filtration are time consuming and labor intensive and often the major bottleneck in high-throughput labs. Automation is the way to minimize this bottleneck. There are now may sample preparation systems available. These systems allow automation of tasks like desalting, SPE, and filtration, to name a few. This not only significantly improves throughput but also improves sample to sample reproducibility by taking many steps of sample manipulation out of the hands of lab technicians.

SPE is one of the fastest growing techniques in the area of sample preparation. There are many solid-phase chemistries on a variety of supports, including silica and polymers, available to accommodate the clean up and concentration a wide variety of analytes from different samples matrices. SPE media are available in many formats including packed syringe barrels, cartridges, and 96-well plates that can be utilized by most of the automated sample preparation platforms currently available.

Many, if not all, commercial formulations of protein therapeutic drugs contain multiple charge variants that are difficult to characterize and even harder to remove. What improvements are needed in this field?

Joseph and Lloyd: Ion-exchange chromatography (IEC) — methods that can be used with on-line MS for identification of the charge variants. More work is needed to make sub-2-µm ion-exchange columns true UHPLC options. But that will only help with the characterization of charge variants.

Appelblad, Bei, and Lentz: The typical chromatographic modes for protein separation in native form are size-exclusion chromatography (SEC) and ion-exchange chromatography (IEC). While SEC is the workhorse for separating and quantifying soluble protein aggregation, IEC is the choice to analyze the charged heterogeneity of proteins. Although the particle sizes and dimensions of the newly developed ion-exchange columns are optimized for highest throughput or for highest efficiency, a typical ion-exchange separation of charged variants and pegylated proteins will still take about 30 min or longer. HILIC also has been used for protein glycosylation, which is suitable for complete characterization of glycan structures. HILIC also fits well within multidimensional LC methods due to its orthogonality with reversed phase, which opens up very interesting areas to explore. With the introduction of the so-called biosimilars in Europe, the demand for highly efficient analysis methods with SEC, IEC, and HILIC is increasing. Continuous development of new stationary phases in biochromatography, which will provide higher productivity and throughput for large biopolymers analysis, is needed.