Quantitative Methods for Multielemental Analysis in Low Volume Biofluids

In your paper (1), you state that single-cell inductively coupled plasma mass spectrometry (SC-ICP-MS) is an emerging methodology that holds promise to enable multielemental analysis of low-volume samples in the future. Can you explain why you believe this?

In my view, SC-ICP-MS has become a hot topic in the field of elemental analysis, and rightly so: Scientist working in the field of bioanalytics applied this approach to look into the smallest units of life. As a matter of fact, most biochemical processes take place in cells. Elemental bioanalysis in cells makes it possible to obtain deep insights into cellular processes and the biological status of specific cells. In comparison, the analysis of bioliquids mainly provides insights on a systemic level of the corresponding organism. Over the last decades, many scientists applied ICP-MS to study the elemental composition in all kinds of biofluids and tissues. Due to the latest technical developments it is now possible to achieve ever lower detection limits. Thus, it is now possible to examine individual cells and their elemental composition—at least partially. This may open a completely new field of application for ICP-MS and there are many scientists who are about to prove this.

Why is ICP-MS generally considered being the method of choice for multielemental analysis as compared to other ICP techniques (AES or OES) or more traditional atomic absorption (F-AAS or GF-AAS)?

ICP-MS is unarguably the gold standard for elemental and especially multielemental analysis. This is mainly due to the fact that this technique is extremely versatile compared to ICP-OES/AES and atomic absorption spectroscopy. Most AAS devices apply element-specific radiation sources (hollow cathode lamps) that enable the detection of a single element at the same time. Although there are some commercially available AAS instruments that allow the detection of 10-20 elements simultaneously, sensitivity limitations hamper the analysis of most trace and ultratrace elements. ICP-OES allows multielemental analysis, is relatively easy to setup, and well-suited for high-throughput analysis—however, spectral interferences can be challenging. ICP-MS allows high-throughput, multielemental analysis with the lowest-possible detection limits. It offers a linear detection range of up to 10 orders of magnitude, enables isotopic analysis (isotope dilution analysis for quantification, control of spectral interferences), and it is possible to analyze any type of sample material (gases, liquids, solids).Furthermore, high-resolution separation techniques such as HPLC, GC, CE etc. can be hyphenated to the ICP-MS. Overall, this makes ICP-MS the most powerful and versatile technique for elemental analysis.

What motivated you to focus on this specific application?

Several times, we have encountered situations, where limited sample volume was available. For example, limited volume of biofluids (blood serum or plasma) might be available in human sample cohorts for interdisciplinary clinical studies. In such cases, the principal investigator often selects the panel of analytical tests according to the required sample volume. Sample volume restrictions can also arise from the donor itself, even if there is no competing analysis. This is the case for animal models, especially for studies with mice where only a few µL of blood serum or plasma are available per animal. Since most of the published multielemental profiling methodologies require relatively large sample volumes, our aim was to provide a tool that overcomes this challenge. For this reason, in 2016 we have developed an ICP-MS based method, which allows the analysis of 29 elements in a 150 µL blood serum sample. However, we soon realized that there is a need for multielement analysis in even smaller sample volumes. This is for example the case when only small amounts of serum and sample are available (for example, in precious clinical cohorts or preclinical studies with rodents). We accepted the challenge and found that it is possible to detect 20 to 30 elements in 3–10 µL of biological fluids. These positive results have motivated us to optimize and validate the method. Finally, we have successfully applied the method in a preclinical model. The method turned out to be very easy to use, it does not require expensive or complicated changes in the experimental setup, and, above all, it has shown very good analytical performance. We then decided to publish the results of our study as well as the experimental setup in order to encourage other readers to perform multi-element analyses in the smallest sample volumes in a straightforward way.

What are some cancer-specific multielemental profiles that you have identified? What type of data analysis techniques were used to determine these profiles? Are there a few predominant elements that seem to be most often present in the mice cancer cells?

In the mouse cancer modes under evaluation, the presence of the tumor influences the serum concentration levels of various elements. Some elements, such as sulfur, calcium, manganese, zinc, strontium and molybdenum, were not affected significantly by the presence of the tumor. On the other hand, elements such as phosphorous and copper were increased in the serum samples of tumor bearing mice, independently of the tumor type. Iron, on the other hand, was depleted. These results seem to indicate that the presence of tumor systemically influences the concentration levels of some serum elements, independently from the tumor type. On the contrary, elements such as vanadium, chromium, bromine, rubidium, cesium and cobalt show trends for tumor-specific concentration changes. Furthermore, our results show that different immune T cells (CD4+ and CD8+) have specific mineral profiles reflecting their specific immune function. Moreover, the presence of a tumor seems to have an impact in the mineral contents of T cells. Indeed, we observed elemental modulations within CD8 cells that could be associated with presence of cancer, such as higher levels of magnesium and manganese and lower levels of iron, cobalt and rubidium. These findings might indicate an increased metabolic need of immune cells for Mn and Mg in the presence of a tumor.

What are the biggest challenges that you encounter in this analysis method? Are there any limitations to using this technique? What options or alternatives are available to overcome these challenges?

While the analysis of small amounts of biofluids is quite straightforward (centrifugation to remove particles and a simple dilution step), the analysis of cells is more challenging. In fact, one hurdle we faced during method development was to remove the cell sorting buffer (containing a high concentration of various elements of interest) from the cells. As large volumes of this buffer are required for a single cell-sorting experiment (roughly 10 liters in our case), it was not possible to prepare a buffer solution based on reagents which are suitable for trace elemental analysis. For a washing step, we applied filters or considered osmosis for buffer removal, but none of these sample preparation strategies proved to be suitable. Finally, applying a mild centrifugation step to form a cell pellet, followed by removing the supernatant with a pipette, provided satisfying results. However, the removal of a few microliters from an Eppendorf tube, and at the same time maintaining the integrity of a tiny cell pellet, remains the Achilles heel of the method and requires some finesse and experience.

With such small samples, is contamination an issue, and how have you overcome this potential problem?

Contamination is always a challenge in ultratrace elemental analysis. Acid-washing, working in a clean environment and under a fume hood are the minimum requirements to obtain high quality results. However, contamination is not more challenging in this method, as compared to most other ICP-MS based methods dealing with trace and ultratrace analysis. This is mainly due to the fact that a relatively moderate sample dilution of (1:10) and few sample preparation steps are applied. Contamination could arise, for example, from plastic materials. We mitigate this risk by applying an acid-washing step up front. Another, often underestimated source of contaminants are chemicals and reagents applied during sample preparation. They often contain elements (for example, iron) that enter the sample in this way and thus mix with the analytes. Due to the inherent properties of elemental mass spectrometry, it is then no longer possible to distinguish easily between endogenous and exogenous elements. This can lead to an overestimation of the element concentration. In our study we only apply buffer solutions and nitric acid during sample preparation. Thus, the overall contamination risk is relatively low.

Can you briefly summarize the results that you have achieved so far?

We have developed a quantitative analytical strategy that enables the analysis of 20 biologically relevant elements in low-volume samples (10 µL) of serum. By applying an optimized workflow, the method is also applicable for the quantification of 12 elements in a low number of sorted cells (250,000 cells). In our paper, we give detailed descriptions of the experimental setup. This enables scientist to apply the approach to study biological-relevant elements (major trace and ultratrace elements) in very small sample volumes. Applied to tumor cells, we could demonstrate that the presence of certain cancer types influences differentially the elemental composition of serum concentrations. Furthermore, we could show that different immune T cells (CD4+ and CD8+) have specific mineral profiles reflecting their specific immune function. Last but not least, the presence of a tumor effects the mineral composition of T cells.

Is there anything that you have done in your study that is different from what other researchers have done as they have examined the same problem?

Most of the ICP-MS-based approaches aiming to detect major trace and ultratrace elements in cells are challenged by the media or solvent in which the cells are transported into the plasma ionization source. If the cells are present in Milli-Q water, the “elemental background” is very low. This facilitates the detection of very small quantities of cellular elements. However, cells are prone to lyse in Milli-Q-water. This makes the cellular membranes permeable for ions and small molecules, which are eventually leaking out of the cell and changing the elemental composition of the cell. Fixation strategies, aiming to stabilize cellular membranes, can prevent lysis. Unfortunately, fixation methods tend to increase permeability of the cell membranes and consequently alter the endogenous elemental composition of the cell. Moreover, aggregation of cells is often seen after fixation and needs to be monitored. In fact, the presence of cells in the monodisperse state has to be confirmed by complementary analysis. Many of the reported SC-ICP-MS methods are overcoming this issue by using yeast cells. The use of yeast cells has two advantages: first, the cells are relatively big, meaning higher absolute quantity of elements per cell. Second, these cells are extremely robust compared to immune cells. Re-suspended in Milli-Q water, yeast cells can stay alive for a couple of hours without undergoing lysis. This timespan allows conducting SC-ICP-MS experiments with intact cells while maintaining an extremely low elemental background of the solvent. Suspensions of cells in buffer solutions, on the other hand, allow keeping the cells alive and preserve the integrity of the cell. However, such buffer solutions contain “contaminants”; for example, major or trace elements which are present in the raw materials. This makes it difficult to distinguish between “cellular” elements and elements from the buffer solution. We were able to overcome this limitation by using a sample preparation strategy that includes a washing step with a suitable buffer solution (containing a relatively low elemental background).

What are your next steps in this work? Do you know how well the discoveries of multielemental profiles in mice cancer cells may extrapolate to human cancer cells?

I am now working in a company which is rather focused on elemental analysis is food and pharmaceutical products. However, my former colleagues are continuing our research work and other scientist are working on different approaches to study trace- and ultratrace elements in cells. I am confident that rather sooner than later we will see new findings in this field, both in pre-clinical and clinical studies.

Tobias Konz

Tobias Konz is currently leading the Elemental Analysis group at UFAG Laboratorien (in Sursee, Switzerland), specializing in quantitative analysis of elements in food and pharmaceutical products (GMP and ISO environment). He studied biomedical chemistry at the Johannes-Gutenberg University (2000-2007), and completed his diploma thesis and PhD in Analytical Chemistry in the research group of Alfredo Sanz-Medel, under the supervision of Maria Montes-Bayón (University of Oviedo, Spain). In 2014, he started a professional career as ICP-MS specialist at the Nestlé Institute of Health Sciences (Lausanne, Switzerland) in the Nestlé Research organization. The main focus of his work has been to develop quantitative methods for multielemental analysis in low volume samples (biofluids) in a GLP and ISO regulated environment.

Reference

(1) T. Konz, C. Monnard, M. Rincon Restrepo, J. Laval, F. Sizzano, M. Girotra, G. Dammone, A. Palini, G. Coukos, S. Rezzi, J.-P. Godin, and N. Vannini, Anal. Chem. 92(13), 8750–8758 (2020). https://dx.doi.org/10.1021/acs.analchem.9b05643​