Q: What changes have you seen in the MS field in recent times?
A: The biggest change is the fact that previously, most MS experts came from the organic chemistry groups. Now MS crosses every discipline, and there are many more MS experts in each of these fields. On the instrumentation side the biggest change has been in sensitivity. It seems to be following Moore’s Law [a computing term that states that the processing power for computers will double every two years], where sensitivity is now down to sub-pico and femto molar ranges. This is amazing, as it now allows us to do imaging and work with low sample volumes and still get good data. Miniaturization is another change, where mass spectrometers are getting smaller, with higher throughput.
In terms of coupling techniques, the one that the community has embraced is ion mobility as a separation technique for high-throughput MS (IMS-MS). With multiple reaction monitoring (MRM) and ion mobility, you can get a huge amount of structural data very quickly, in the same timeframe where previously you could collect only one spectrum. With ion mobility you can get separations in milliseconds, so it’s incredibly fast. Back in the day when you had separation techniques in the front end of mass spectrometers, you had a certain time window to get all this information on an eluting analyte. Now, with ion mobility and MRM, you can get gigabytes of data from peaks that are eluting in 10 or 15 seconds.
Q: What about changes in the MS data itself?
A: With MS imaging, it’s almost like a separation technique, where you are looking at an ion chromatogram per shot as opposed to when the analyte elutes. If you are going to collect MS-MS data on every shot, then a single MS imaging file, like the one you see published in our recent papers, can be up to 12 gigabytes. We have terabyte backup drives to store all this data. If there is a problem during the run or there are interruptions, we have to delete the file, since it takes up a lot of space to store it. As a part of our defense contracts, our lab is a closed system, so we cannot store anything in an external cloud.
Q: Which of the projects that you are currently working on most excites you?
A: In situ MALDI, where you don’t have to prep the sample anymore, is very exciting. But more interesting is a recently published paper where we did true single-cell analysis using nano [manipulation] in MS. We extracted and looked at the lipodome of one organelle in a cell, and we could do that reproducibly, which is truly amazing. We could do that because of the MS technology moving to femto molar sensitivity, which gave us a lot of information using a very small amount of sample.
Q: Can you shed some light on nano manipulation and how you see it impacting applications in the future?
A: With the amount of sample you are dealing with on a single-cell level, separation is almost impossible. You have to rely on creating very specific chemistry within the tip (pulled capillary tip) of your extraction so that you are looking at a specific compound class that can then be charged. In our case, that was lipids, and so we used chloroform with ammonium acetate (a lipid solvent that you would typically use with electrospray MS applications). If you don’t fine-tune the chemistry, then you will extract all the proteins, sugars, and everything else that is present in that cell. So we are using the same principles that we used for extractions with separation funnels, except now we do it on a single tip. We go to a particular location in a cell that we are interested in to extract the right compound using the right chemistry. So it’s not a shotgun approach where you extract and analyze everything. We have to be very exact about what we want to see and how we are going to get it.
Q: What innovations in MS instrumentation have made this possible?
A: The innovations in nanospray have made it possible for us take a 1um capillary extraction tip and turn it into a nanospray tip. That was a huge leap forward from electrospray to nanospray MS. The MS ionization efficiency has also become very high in the new designs of instruments, which allows us to do what we do. We have used this nanospray technique for drug and residue extractions from fingerprints in forensic applications. Again, you have to tune the solvent chemistry to look for specific explosives or drugs, but if that chemistry exists, then you can use this technique for any type of extraction.
Mass spectrometers are also getting cheaper and smaller in footprint, making them portable for field use. You can now get a commercially available 12-pound mass spectrometer. It’s not as comprehensive as most benchtop mass spectrometers, but it’s truly field-portable. Hence, a lot of environmental and forensics groups are finding MS instruments that are perfect for their use and are also affordable.
Q: Where do you see the next breakthroughs in MS?
A: Being able to do high-throughput large mass ranges and having high resolution over the entire mass range will be the next major leap. The other breakthrough will be in informatics—taking that 12 gigabytes of imaging data and being able to process it properly. The 12 gigabytes of data take only four hours to collect, but it takes my students a week to process [them]. Database and informatics innovations that will help people use algorithms to assess data based on what they are looking for will become very important. It can be done today, but it takes a very long time.
Q: Do you think people using MS should receive specialized training?
A: In all honesty, sometimes people get overconfident about the instruments because all vendors are making them “plugand- play,” with recipe-driven analysis. We just did an experiment for the U.S. Drug Enforcement Administration (DEA) that showed some of the underlying problems. We took three different mass spectrometers and looked at various nuances that can be controlled, such as trapping time, ionization current, buffer gas, [and] size of collision cells, and we found that even if the instruments were run under the same set of conditions, they didn’t give the same results. For a person using MS as a research tool it may be fine, but for the DEA this is a huge problem when it comes to compound identification and quality assurance (QA). Hence, I think it’s important for every lab manager to get some fundamental training on MS and not just rely on the instrument settings.
Q: Are there problems that have surfaced because of the recent innovations in MS?
A: MS has now become so sensitive and accurate that it can pick up any small amount of impurity or changes in reagents due to lot-to-lot variations. Lab managers and QA professionals need to be very aware of this. We have to routinely standardize our assays and protocols because we sometimes get a new peak from a solvent or from polymers leaching in from the centrifuge tubes. In lipid analysis we have to be extremely careful because the solvents we use can extract oils and polymers from everywhere, which is why we only use glass. With direct inject for MALDI imaging, everything comes up from the sample itself. We have to be exact in what we are looking for and then do a lot of MS-MS or even MS-MS-MS analysis to determine it is what we say it is.
Q: Any pitfalls lab managers should look to avoid?
A: One thing we have noticed on the separations front is that everyone is coming out with new gas chromatography (GC) and liquid chromatography (LC) columns, and because of the improved sensitivity of MS we are now picking up different bleeds. Not all columns are the same, and hence, we cannot blindly pick up any two columns and expect them to perform the same. The column consistency does not match the sensitivity of the mass spectrometer. So lab managers should look to buy the same columns and run them under the same conditions. This makes our standard operating procedures (SOPs) longer, but we have to include all the nuances to be able to stand behind our data.
Q: What resources do you turn to for advice and help?
A: The first place we turn to is another successful lab, [to] get their SOPs and build our protocols based on their consistent results. That’s better than going to the literature, where you sometimes cannot reproduce what has been done. You could also turn to your instrument vendor and seek help from their application scientists. They are paid to help customers solve their problems, and it’s very likely that they have seen this problem before and can help you troubleshoot.