How Live Chemical Imaging can Help Biologists Find Rare Events

Introduction

Researchers can often spend hours on a microscope trying to find the perfect, representative area of a sample to collect data from. In many cases, this is done in advance of setting up regions of interest for additional acquisition, for example, queueing up multiple sites for electron tomography. All of this in the hope that one of these areas contains the elusive region or rare event that is of interest, such as a dividing cell in a tissue, or the best orientation of subcellular structures. Finding these relatively rare events is a continuous challenge in biological electron microscopy.

The introduction of the Unity BEX imaging detector, which uses Backscattered Electrons and X-ray simultaneously to deliver rapid high-definition colour images embedded with elemental data as you navigate around your sample, will revolutionise this in the future. However, for those who can't yet access the BEX technique, there is still a lot that can be done to streamline the process, particularly with AZtec live chemical imaging.

Signals and Labels

Labelling and selective staining to highlight a rare event can greatly speed up the localisation of regions of interest (ROI) and the imaging process. This may also require some form of correlative light and electron microscopy. Alternatively, a rapid elemental analysis in the electron microscope using energy dispersive X-ray spectrometry (EDS) can significantly speed up localising ROIs, particularly when searching for discrete native elements or a label with distinctive composition, such as quantum dots.

Figure 1. Macrophages that have been exposed to patient-derived cobalt wear particles (sample courtesy of Zhidao Xia, Swansea University). The blue is the carbon signal, and the pink is the cobalt signal. This layered EDS map took 45 minutes to acquire. Not all macrophages contained the cobalt nanoparticle clusters

Standard EDS Workflows

The standard workflow with EDS is to use the electron signal to localise a region within the sample and to subsequently collect compositional data. High-quality EDS maps can often take time to acquire (Figure 1), particularly from life science samples imaged at low kV, and where the X-ray emission is relatively low. Several regions can be queued up and the analysis itself left to run in an automated fashion. This is particularly useful when acquiring large area maps, a process that captures many individual tiles at higher magnifications and montages them for the final result. In this way, large sample areas can be covered whilst maintaining high image resolution, however, this can take a significant amount of time.

Figure 2 shows an example of the task of trying to localise clusters of cobalt nanoparticles in macrophages. EDS analysis demonstrated that the BSE signal alone was not enough to identify cells containing aggregates of the particles. Identification of the particles would often take at least 10–15 minutes per site using the standard EDS workflow approach (and often much longer when working with high image resolutions), and that is if the correct region is located in the first place. Using point and ID or spot analysis on EM images alone did not speed this up significantly, as once again the BSE signal was not a good indicator of whether there was cobalt in the cell (Figure 2), meaning multiple points need be analysed.

Figure 2. A BSE EM image of the same macrophages shown in Figure 1. Several areas of similar contrast were selected for point and ID analysis (obtaining a spectrum from a small region that covers the potential nanoparticle clusters). Regions 1 and 2 appeared to be slightly beneath cell membranes, 3 and 4 on the surface of the sample. The spectra from each of these 4 regions shows that the intensity of the BSE signal alone was not able to predict composition or the location of cobalt nanoparticles.

Live Chemical Imaging

A method that completely changes the standard EDS workflow is live chemical imaging (LCI). This is where both EM images and EDS maps are acquired and overlayed while moving around the sample, providing a live update on sample composition. This works best on samples with few elements and a significant X-ray emission. Biological specimens typically have a lot of different elements in low concentrations and a few elements that dominate (for example, carbon and oxygen). X-ray emission is also relatively low for most biological samples. So, how useful is live chemical imaging for biologists?

Take the example of a rare event or the search for a specific sample. The entire specimen could be imaged and mapped in the search for a region of interest, possibly spending days acquiring data. A CLEM approach could be used that likewise might take a significant amount of time and typically involves complex sample preparation. Or live chemical imaging can be used to speed up the search within the electron microscope and focus time on acquiring the data needed.

Using LCI, the search for cobalt-containing cells can be sped up significantly, simply by moving around the sample and pausing where potential ROIs were found to check for the presence of cobalt nanoparticle clusters.

The short sequence shown in Figure 3 plays in real time. As it shows, a few seconds pause is enough to identify where clusters are; this is much faster than going through the process of capturing a high resolution SEM image, without even considering the time to perform EDS analysis as well. Data can be exported straight from LCI or, once a ROI is found using LCI, data can be acquired using a more traditional workflow.

Figure 3. A video of live chemical imaging to find cobalt nanoparticles in macrophages. The sample was courtesy of Zhidao Xia (Swansea University). The sequence is in real time and shows navigation around the sample, pausing to acquire some additional data and confirm which bright spots on the cells contain cobalt (note that not many do), and spot analysis on a few regions to verify composition. The analysis takes less than a minute, a much faster process than standard EDS workflows.

Conclusion

While not as potent as BEX, LCI is a powerful tool for electron microscopists and can significantly speed up any investigation by combining ultrastructural information from the electron images with compositional information from EDS.