The Simple Beauty of TLC-MALDI-MS

I’ve always found chemistry a fascinating field, and I recognized early on that it’s a prerequisite for many other studies, for example, in biology and medicine; after all, biological molecules stem from chemistry. In my opinion, without a detailed knowledge of chemical aspects, it’s very difficult to understand physiological features. Nevertheless, today I find myself conducting research in a medical physics/biophysics department, which in some ways emphasizes the need for multidisciplinary approaches to solve complex problems.

In my early days, I had no knowledge of mass spectrometry. I was more fascinated by nuclear magnetic resonance (NMR) spectroscopy (which we still rely heavily upon today, as you will see). In fact, my diploma thesis focused on the application of NMR spectroscopy in inorganic chemistry. And my PhD continued my work with NMR but familiarized me with its clinical application (my thesis was: “NMR investigations of synovial fluids and contributions to the modelling of cartilage destruction during rheumatoid arthritis”).

My introduction to mass spectrometry was somewhat accidental. During my postdoctoral research, I was involved in a much bigger project that aimed to use both NMR and MALDI MS. At first, we attempted to analyze proteins released from cells and then moved onto carbohydrates – both efforts were unsuccessful, to be honest. But when we looked at lipids (specifically, phospholipids), we found great success. And so, the application of MALDI-time-of-flight (TOF)-MS to the study of lipids and phospholipids has been my major research focus for around 15 years.

We are particularly interested in modifying lipids using oxidation reactions to understand what is possible from a chemical point of view; even with a very simple lipid, the introduction of reactive oxygen species (ROS) can result in many different species. Our ultimate goal is to understand what that means in the context of an organism, among other things.

TLC: tried and tested – and simple

I first used thin-layer chromatography (TLC) many years ago in organic chemistry to monitor the number of products after our numerous chemical reactions. Even nowadays, I believe students of chemistry use TLC in this way. And I suspect many people may consider TLC just as a simple separation technique for chemistry students.

In fact, it is TLC’s simplicity – and its low barriers to entry – that make it so attractive in combination with mass spectrometry. In mass spectrometry, you are typically faced with the problem of ion suppression when it comes to complex mixtures – and that means you don’t detect all analytes with the same sensitivity. Therefore, some form of separation is desirable if not essential. In lipid analysis, sensitivity in mass spectrometry is determined by the lipid headgroup (that is to say, its charge) and we quickly ran into the effects of ion suppression in complex lipid mixtures. We were not specialists in chromatography, so you might say we chose the simplest option, but in fact, TLC is actually very well established in lipid separations and seemed like the obvious choice.

For us, the main advantage of TLC is the ability to gain high quality separations with unsophisticated equipment (and with little expertise). In fact, all you need is a TLC plate, a sample syringe, and a developing chamber. And yet, despite its simplicity, it allows many samples to be analyzed in parallel, increasing throughput. Suspicious samples are also no problem – I am sure you would not want to inject some of our samples into your HPLC system... Indeed, the ability to load large sample volumes or highly contaminated samples can be very advantageous in our work.

Ten years ago, we used to separate our phospholipid classes using TLC, and then scratch the corresponding spots off the plate for re-elution ahead of MS analysis. It actually worked well, but you can probably imagine that with large numbers of samples, the process was somewhat tedious. At the same time, researchers were using MALDI-MS to analyze biological tissues, and the parallels between TLC plates and slices of tissue became apparent. Could we not detect our samples directly on the TLC plate? Yes, we could – and we are still doing that today, though there have been improvements along the way.

TLC-MALDI-MS 2.0

Advances in commercial instrumentation are certainly making the combination of TLC-MALDI-MS even more attractive (for example, Bruker Daltronics has produced a TLC-plate adapter than can be inserted directly into the system and software that facilitates analysis). But quieter innovations have been occurring in the actual TLC plates themselves.

Dedicated MS-grade TLC plates that have been optimized for use in MALDI-MS are now commercially available. Such TLC plates use a thinner layer of extremely pure silica (standard plates typically use 200 µm layers, but the MALDI-grade plates use 100 µm silica layers), and we have shown that stationary phase thickness determines the quality of MALDI-MS spectra for lipids (1), which is illustrated in Figure 1. The matrix background is significantly reduced with thinner layers, and though we do not fully understand the reason, I believe the advantage is conferred by an analyte concentration effect and the fact that the UV laser of MALDI-MS does not penetrate deeply into the surface. In any case, if we look at a diluted extract from cells (as we so often do), we will likely see very poor spectral quality with standard plates, but remarkably good spectra with optimized plates.

TLC-MALDI-MS in action

In terms of our own research, we are applying TLC-MALDI-MS in a number of different research areas. For example, we are currently looking at the effect of lipid composition on spermatozoa samples with a view to discovering lipid biomarkers of fertility. Of course, this could have a clinical impact, but it’s also of interest in the world of animal breeding, where artificial insemination could benefit from a prediction of sperm quality.

Another big focus area for us is in understanding the role of lipids in the mechanisms of obesity. In particular, we are interested in learning how different compositions of fat in the diet correspond to uptake of fatty acids in fat tissues or other organs. The third field, which I mentioned earlier, is the area of lipid oxidation.

Oxidative stress is well-known terminology in biology, but it is difficult to assess the contributing factors to lipid oxidation in physiology. We began by trying to understand which oxidizing species lead to which products in an isolated system, but we have also started researching the physiological aspects. There are many different enzymes (lipases and phospholipases) and one of our questions is: are oxidized lipids/phospholipids metabolized to the same extent as their native equivalents. Here, we are applying a combination of different mass spectrometric methods, as well as NMR.

In all of these cases, we might begin with a mixture of lipids from a cellular extract (for example, neutrophils or spermatozoa) and use NMR (in particular, phosphorous-31 NMR) to gain quantitative information about the lipid classes. Why NMR and not mass spectrometry? Well, if you are typically running the same samples (for example, blood samples in clinical chemistry), you may already have an idea about the concentrations of expected phospholipids and can therefore use dedicated deuterated standards to obtain quantitative information. However, we deal with many different kinds of samples and cannot easily make such predictions, which makes NMR a very useful analytical technique as it avoids additional time-consuming experimentation but gives reliable quantitative information. However, NMR does not provide detailed information about the fatty acid composition, which is where the use of TLC-MALDI-MS comes to the fore. We can run the same samples and investigate the MS spectra to complete our investigations.

Conversely, with TLC-MALDI-MS it is difficult to gain quantitative information about the compounds separated on the plate. Why? Because the distribution of analytes is not also homogenous across a spot, so the mass spectra produced are dependent on the position of the laser irradiation zone (which is significantly smaller than the TLC spot to offer increased resolution in MS imaging). One possible solution may be the option to adjust the laser spot size depending on the application – but that is a question for the instrument vendors...

What I think our projects emphasize is the fact that to answer complex questions, the use of complementary techniques can be hugely beneficial – and the simplicity and utility of TLC-MALDI-MS fits perfectly in our research. Finally, a book entitled “Planar Chromatography - Mass Spectrometry” (2) will be at available at the end of this year - and really emphasizes the relevance of this scientific field.