Using Synchrotron XRF to Map Trace Metals in Biological Systems

Metals and metalloids, while essential to living organisms, can, in high concentrations, be toxic. An understanding of how these metals and metalloids are accumulated and transported within plants and animals is possible with the use of synchrotron X-ray fluorescence (SXRF) microtomography. The technique is used in the imaging of major and trace element distributions within natural materials with high spatial resolution. Stefan Vogt of the Argonne National Laboratory in Lemont, Illinois, has been exploring the use of SXRF to detect metal content in biological and other systems. He recently discussed the various challenges, applications, and advantages associated with this technique.

In a recent paper, you and your team discussed the use of SXRF microtomography for 3D imaging of transition metals in zebrafish embryos (1).  What other techniques have been used for quantifying trace metals in biological systems? How does synchrotron X-ray fluorescence microscopy (SXFM) compare with those techniques with respect to determining the metal distribution in cells, tissues, and whole organisms? 

There are a number of techniques that can be used to map metal content in various biological (and other) systems, each with its own advantages and disadvantages. Visible light microscopy, in combination with metal-sensitive fluorescent dyes, is great for looking at live cells, but it is limited in that it only “sees” ions (as opposed to total metal content), absolute quantification tends to be difficult, there may be artifacts due to the dye itself, and one needs a dye that is sensitive to the metal of interest (and only this metal) in the first place. It does also have the advantage of accessibility-most labs do have light microscopes. 

Electron microscopy (EM)-either analytical EM, where electrons are used to excite characteristic X-rays, or energy-loss spectroscopy-can also be used to visualize metal content. These instruments tend to provide very high spatial resolution, but only work effectively on very thin samples (depending on the system, 100 nm and less). Radiation damage is an issue, and (relative) sensitivity tends to be less than using SXFM (also known as X-ray fluorescence microscopy [XFM]), but with the advantage of being able to detect individual atoms on appropriate samples. Instead of using electrons for excitation of characteristic X-rays, one can also use protons (proton-induced X-ray emission [PIXE]). Similar to SXRF and analytical EM, one tends to visualize 10 or more elements at a time, thicker samples than with EM can be probed, but at a resolution of micrometers.

SXRF uses a focused X-ray beam to excite characteristic XRF, typically detecting 10 or more elements at a time. At the moment, a routine high spatial resolution on the order of 100–200 nm can be achieved, with the best instruments getting down to 30 nm and below. By scanning the incident X-ray energy across an absorption edge, one can also determine the chemical state of the metals of interest, and thereby “map” oxidation states across a sample. It does have the disadvantage that access to these instruments tends to be limited, since only synchrotrons provide high-brightness X-rays required for this kind of instruments. It should be mentioned that laboratory-based micro-XRF has been steadily improving and now can reach the ~10-µm length scale quite well. Synchrotrons typically provide a large fraction of the instrument time to the public free of charge using a competitive proposal system.

Some of the main advantages of SXRF include the ability to investigate comparatively “large” systems (for example, millimeters in size) at still a very reasonable spatial resolution, as well as the comparatively easy quantification of the results. A table that summarizes the various above mentioned techniques can be found in the Journal of Eukaryotic Microbiology (2).

Can you describe some of the challenges in using this technique?

Sample preparation is probably one of the main challenges (as with pretty much any other microscopic technique). In particular, one needs to preserve both specimen structure, as well as chemical integrity. A lot of the chemical fixation techniques developed for microscopy are great for structural preservation, but then immediately have an impact on elemental distribution. For example, fixation with 4% paraformaldehyde permeabilizes cell membranes and leads to the redistribution of diffusible ions. In a healthy cell, for example, intracellular potassium tends to be high, but calcium very low-after chemical fixation the potassium leaches out and calcium enters, leading to a reversal of their relative abundance. Depending on the system to be investigated, this situation may be perfectly fine, but one should keep it in mind. In general, cryogenic preservation is probably an ideal sample preparation (rapidly freezing the sample to liquid nitrogen temperature so that the water– buffer solution does not crystallize, but becomes vitreous), but it involves somewhat of a learning curve and requires appropriate instrumentation for both sample preparation and subsequent investigation.

What are some other important applications for SXRF in the life sciences that are currently being investigated?

Most of the research carried out at our facility, (the Advanced Photon Source, which is a U.S Department of Energy Office of Science user facility at Argonne National Laboratory), is collaborative-“General Users” receive beam time based on proposals, and we work with those users to help make experiments happen. Consequently, we have a fairly large range of applications that are being investigated. They include the role of metals in natural systems such as Zn in fertility (3); metals that are artificially introduced, either as contaminants or as supplements, such as Cr in dietary supplements (4); and properties of wood as a construction material (5).

Another paper mentioned the development of TiO₂-DNA-nanocomposites that could enter cells and perform relevant functions in vivo and in situ with regard to a potential nanodevice that could be used for gene therapy (6). How is X-ray fluorescence microscopy currently being used in this research?

The lab of Gayle Woloschak at Northwestern University has been spearheading that work with us (7). There are numerous different aspects to that work, but it involves the creation of inorganic functional nanoparticles that can target specific cells (cancer). The reason SXRF is brought to bear on these questions is that it can specifically visualize the nanoparticles even in comparatively thick specimens, can determine whether they localize to specific target regions (such as the nucleus), whether the nanoparticles remain bound to other components (such as optical dyes for visualization with visible light microscopy, or Fe₃O₄ or Gd as a contrast agent for MRI). SXRF is also very useful to not only determine intracellular localization, but also to map comparatively large (centimeters) tissue areas to determine overall localization-for example, if one wishes to develop a compound that targets specific tissue regions, one does want to also show that it essentially only localizes to those regions, and not anywhere else.

What are your next steps in this area of research?

There are a number of areas we are working on. In the near term, we are continuing to push our instrumentation, with the goal to make it faster. In particular, we are working to make SXRF tomography a routine tool that is as easy to use as 2D elemental mapping. Correlated with this effort, we are developing computational tools, such as advanced reconstruction algorithms for tomography, and semiautomated methods for analysis in general. Since the datasets acquired are fairly large, as well as complex, at some point it becomes difficult to adequately analyze the data we can acquire. While it is fairly straightforward to interpret one or two gray-scale images of a sample, if one has 10 different images of the sample, each showing the distribution of different elements, for dozens of different samples and different conditions, it becomes increasingly difficult to extract the relevant differences. Advanced computational methods can really make a difference.

What we are most excited about, though, is the prospect of an upgrade to our facility, the proposed Advanced Photon Source Upgrade. Recent developments in the synchrotron community have demonstrated an approach that can make these facilities 100x brighter than today-for X-ray microprobes, this translates into essentially 100x faster and better instruments. At the Advanced Photon Source, we expect to be able to carry out experiments where we can visualize metal content in soft materials (such as biological samples) down to a spatial resolution of 10 nm, in 3D, with sensitivities on the order of a handful of atoms in thick (tens of micrometers) samples, something completely impossible today.

How much of a revolution is a factor of 100? Consider that the difference in top speed between the world’s fastest human and a fighter jet is only a factor of 50. In a number of experiments today, we can only scratch the surface of what we actually wanted to do. In the zebrafish embryo example that you mentioned above (2), the actual spatial resolution of the metal maps is only on the order of 5 µm; our main limitation was the total scan time required (~3 days). With an upgraded Advanced Photon Source, we anticipate that we would be able to map metal content in such a system at something like 100–200 nm in perhaps an hour or so, due to improvements to the X-ray source, optics, detectors, and computational advances. The upgraded source will enable us to study not just a handful of such samples, but truly compare various mutations, different environmental influences, and so on.

References

• D. Bourassa, S. Gleber, S. Vogt, H. Yi, F. Will, H. Richter, C. Shin, and C. Fahrni, Metallomics 6(9), 1567–1760 (2014).

• B.S. Twining, S.B. Baines, S. Vogt, and M.D. de Jonge, J. Eukaryot. Microbiol.55(3), 151–162, (2008).

• E.L. Que, R. Bleher, F.E. Duncan, B.Y. Kong, S.C. Gleber, S. Vogt, S. Chen, S.A. Garwin, A.R. Bayer, V.P. Dravid, T.K. Woodruff, and T.V. O'Halloran, Nature Chemistry7(2), 130–139, (2015).

• L.E. Wu, A. Levina, H.H. Harris, Z.H. Cai, B. Lai, S. Vogt, D.E. James, and P.A. Lay, Angew. Chem., Int. Ed. 55(5), 1742–1745, (2016).

• J.E. Jakes, C.G. Hunt, D.J. Yelle, L. Lorenz, K. Hirth, S. C. Gleber, S. Vogt, W. Grigsby, and C. R. Frihart, ACS Appl Mater Interfaces7(12), 6584–6589 (2015).

• S. Vogt and A. Lanzirotti, Synchrotron Radiat News 26(2), 32–38 (2013).

• Y. Yuan, S. Chen, T. Paunesku, S.C. Gleber, W.C. Liu, C.B. Doty, R. Mak, J.J. Deng, Q.L. Jin, B. Lai, K. Brister, C. Flachenecker, C. Jacobsen, S. Vogt, and G.E. Woloschak, ACS Nano7(12), 10502–10517 (2013).