Trends in Sample Preparation for Chromatography
Derek Wachtel, scientist in the DMPK department at Ironwood Pharmaceuticals, and Mingliang Bao, PhD, senior scientist at Labstat International ULC, talk to contributing editor Tanuja Koppal, PhD, regarding various issues they face with sample prep in their laboratories. They both stress that sample prep is very important and a necessary step in any analysis and with newer technologies making it easier and faster to accomplish, there should be no reason to ignore or overlook it.
Q: What kind of samples do you routinely test in your lab?
WACHTEL: I work in a midsize pharmaceutical company, and we mostly analyze plasma samples. However, we also test urine samples and tissues such as the liver, heart, brain, lung, and kidney. I am in the DMPK group, where we study the absorption, distribution, metabolism, and excretion properties of our molecules. We have about 10 scientists in DMPK, and they are mostly chemists or biologists by training.
DR. BAO: We work as a third-party contract testing laboratory. Our lab is specifically focused on tobacco products and tobacco smoke analysis, and we are one of the biggest labs in the world that provides this type of analysis. We test all forms of tobacco products, including cigarette smoke (mainstream and sidestream tobacco smoke), whole tobacco, and smokeless tobacco. We have nearly 200 people working in two labs; one lab is focused on chemistry and the other on toxicology. In the chemistry lab, where I work, we have about 120 people, including technicians, analysts, and scientists. The sample generation and preparation for the tobacco smoke analysis is unique. Tobacco smoke is generated by a smoking machine under ISO, Canadian Intense, or client-specified smoking conditions. Target compounds in tobacco smoke are either trapped in impinges with organic solvents or collected on glass wool fiber filter pads. We then use different sample prep tools to extract the compounds from the cigarette smoke for various instrumental analyses.
Q: What types of compounds are you looking to analyze, and what techniques do you use?
WACHTEL: We typically analyze peptides or small molecules in various biological matrices using LC-MS/MS for quantitation or time-of-flight (TOF) technology when elucidating metabolic pathways. Routine sample extraction is performed using protein precipitation and solidphase extraction (SPE) when we’re dealing with a particularly complex matrix with endogenous components that can cause high background noise or ionization suppression of our molecules. We also pay particular attention to developing a robust chromatographic method to achieve adequate separation from endogenous interferences.
DR. BAO: The compounds that we analyze are mostly organic, but we also look at inorganic compounds such as nitrates, nitrites, ammonia, and heavy metals. We typically use chromatography-gas chromatography (GC) and liquid chromatography (LC), and both highperformance liquid chromatography (HPLC) and ultrahigh-performance LC (UHPLC) for sample analysis. For trace-level analysis of compounds, we use techniques like LC-MS/MS that uses mass spectrometry (MS) for detection.
Nearly 4,000 compounds have been identified in cigarette smoke, so these samples are very complicated. The smoke samples are also dirtier than most other types of environmental samples. Hence, they require a lot of pretreatment, cleanup, and concentration before analysis.
Q: Can you talk about some of the sample prep steps that you use on your complex samples?
WACHTEL: Quantification of large peptides in plasma using LC-MS/MS can be very challenging, as large peptides exhibit multiple charging and wide isotopic distribution patterns and they tend to show poor fragmentation efficiency due to the formation of numerous low abundant product ions. The biological matrix is also very complex and gives a high background noise. We overcame the poor fragmentation efficiencies by using single ion monitoring (SIM), without fragmenting the parent molecule. To overcome the reduced selectivity and higher background resulting from the use of SIM, a differential mobility separation (DMS) device was incorporated to add an additional level of selectivity to the compound of interest from isobaric species and co-eluting interferences found in the plasma. In this case we were able to minimize the background noise using DMS, rather than performing a more extensive sample extraction procedure to achieve the desired lower limit of quantitation.
Q: What types of compounds are you looking to analyze?
DR. BAO: The compounds that we analyze are mostly organics, such as nicotine, tobacco-specific nitrosamines (TSNA), polycyclic aromatic hydrocarbons (PAHs), aromatic amines, phenolic compounds, carbonyls, volatiles (1,3-butadiene, isoprene, acrylonitrile, benzene, toluene), and semivolatiles (pyridine, quinoline, styrene, acetamide, acrylamide). We also look at inorganic compounds such as carbon monoxide (CO), oxides of nitrogen (NOx), ammonia, and heavy metals. Most of the organic compounds that we analyze are in the trace-level, microgram, or nanogram amounts. Hence, the tobacco or tobacco smoke samples usually require a lot of pretreatments, such as derivatization, solidphase extraction (SPE) cleanup, and solid-phase microextraction, to improve sensitivity and to reduce sample matrix effects.
Q: What trends or improvements have you seen in sample prep?
WACHTEL: I recently developed an LCMS/ MS method for quantifying an endogenous, extremely polar small molecule in a particular tissue matrix. However, when I tried to use the same method with the same tissue matrix at a later date, results were not reproducible and I found myself reevaluating the original method. It was discovered that using a simple protein precipitation extraction was not sufficiently cleaning up the samples, resulting in reduced column life. Incorporation of a phospholipid removal plate seemed to alleviate the problem.
In terms of trends, I see more use of plate-based methods, such as these phospholipid removal plates, that improve the column and detector life by producing cleaner samples for MS analysis. The other trend I see is the use of supported liquid extraction (SLE), which is a simplified liquid-liquid extraction performed on a plate that has a sorbent to separate aqueous and organic phases. People generally prefer high-throughput sample preparations, and plates are easier to use than individual tubes, they produce clean samples, and they handle the same typical sample volumes.
DR. BAO: I have been working for nearly 15 years in this lab and remember doing a number of different types of liquid-liquid extractions and Soxhlet extractions for tobacco analysis that were manual and very time-consuming. Sometimes one technician would end up analyzing only three to five samples a day using these techniques. Over time we have improved most of our methods by switching to automated SPEs to significantly improve the work efficiency and reduce the time for analysis.
Fifteen years ago most of our mass spectrometers were ion-trap MS because of its relatively low cost, the ease of utilization, and better sensitivity as compared with quadrupole MS. Now we use quadrupole MS, as the new quadrupole MS can offer 10-20 times better sensitivity, which allows us to use smaller volumes of sample and automate the extraction procedures. It is very important to our lab to be able to reduce costs and time for analysis and keep it competitive.
For the analysis of trace-level organic compounds in complicated sample matrices, such as tobacco smoke and environmental and food samples, the test methods should be highly sensitive and selective. Most of the time the sensitivity and selectivity of the analytical instruments (GC-MS or LCMS/ MS) cannot be easily changed or improved. Improvement of sensitivity and selectivity of the test method can be more easily obtained by using SPE with selective sorbents such as molecular imprinting polymers (MIPs), immunosorbents, and mixed-mode sorbents.
Q: What would your advice be to other lab managers involved in sample prep?
WACHTEL: Sample prep is a very important step, and just because a sample looks clean does not mean that it actually is. Good sample prep can increase signal-to-noise, thereby resulting in a lower limit of quantitation and keeping columns and instruments performing for a longer period.
DR. BAO: Sample prep is one of the most important parts of analysis for the labs that are looking to reduce cost and time of analysis. Sample preparation is usually the most timeconsuming step in any analytical procedure. One efficient way to reduce the time of analysis is to automate the sample preparation procedures.
To switch from manual SPE to automated SPE, modifications on the SPE procedure or SPE column might be necessary. For example, we used a test method developed more than 15 years ago that required the use of two different SPE columns connected together. For a long time, the SPE procedure was manually run, as our automated SPE system could not handle two columns. We wanted to design a single SPE column with two adsorbents that would fit our automated SPE system. That idea was discussed with our vendors to see if they could custom make a single SPE column with two different adsorbents. Ten years ago an SPE column with multiphases was not very common, but we have succeeded in finding a vendor that agreed to make that multiphase SPE column for us. Since we switched to automated SPE, the costs of material (SPE column and solvents) and the cost of and time for sample preparation have been reduced more than 60 percent, and the reproducibility of data has also significantly improved.
Derek Wachtel obtained his master’s degree in chemistry from Northeastern University and his bachelor’s degree in biology, with a minor in business, from the University of New Hampshire. During the first eight years of his career, he worked for Nova Biomedical’s Research and Development department to assist in the advancement of biosensor technology. His interest in working with mass spectrometry encouraged him to make the jump to the pharmaceutical field. His first opportunity in the pharmaceutical field was given to him by Enanta Pharmaceuticals, working for their Drug Metabolism and Pharmacokinetics (DMPK) department. He currently serves as a scientist for the DMPK department at Ironwood Pharmaceuticals, and he has been with them for over six years. With nine years of experience in the pharmaceutical field, he has acquired a diverse skill set in mass spectrometry and DMPK.
Dr. Mingliang Bao received a B.S. and an M.S. in environmental chemistry at Nankai University, China, in 1988 and 1991, respectively, and a PhD in analytical chemistry at Florence University, Italy, in 1997. He worked as an assistant professor at Nankai University from 1991 to 1993. From 1997 to 1999, he was a research chemist at the chemical-biological laboratory of the Water Works of Florence in Florence, Italy. At Water Works of Florence, his primary responsibility was to determine the occurrence, distribution, seasonal variations, and fate of some organic micropollutants, including pesticides, biogenic taste- and odorcausing compounds, phenolic compounds, and ozonation by-products in surface and drinking water. Since 2000, Dr. Bao has held an appointment as senior scientist in Labstat, the largest independent third-party tobacco testing company in the world.