A comprehensive understanding of cellular function on the microscopic level in a whole cell or organism is the ultimate goal in system biology. Where system biology methods isolate individual molecular clusters for further investigations [1], microscopy offers the potential to directly "see" the complexes [2] ideally in their natural environment [3].

The structure researcher's dream - in order to study a complex morphological context in its native functional state - is an imaging tool, which allows viewing in vivo from a millimeter (mm) scale and continuously zooming in to a few nanometer (nm) at different tissue levels to the supra-molecular wiring in a cell. Due to physical limitation of the excitation wavelength, such a device is hard to imagine because it would have to cover wavelengths from sub-nanometer (electrons, x-rays) to hundreds of nanometers (photons, ultrasound). Thus, the structure-researchers face a methodological problem if they want to study a tissue or cell on a large resolution and information level.
This problem has been partially solved by establishing preparation techniques, which allow investigating the same tissue biopsy or cells with different complementary imaging techniques and hence circumvent the physical limitation by using a multimodal imaging strategy on optimally preserved samples [4, 5] as briefly summarized below. In general all these approaches are bridging LM findings with EM investigation - by using fluorescence labels in LM for selection of a region of interest (ROI) and re-investigating exactly these ROIs in TEM or SEM - generally called correlative light and electron microscopy (CLEM) approach to merge structural and functional information at multiple resolution levels into one common view (Advantages see tab.1).
Different labs have worked on establishing CLEM methods for investigations on one and the same ROI over the last years [6, 7].
All these methods have in common to be very elaborative and only well established in leading labs.

Missing Tools

If we want that correlative light and electron microscopy (CLEM) reaches a broad community in life-science, it is not only essential that methods and protocols are proofed to work and be established in leading structure research labs, but become easily available for many labs.

One of the existing bottle necks to spread these techniques are easy to use sample handling tools and communication software between the different microscope types to re-find and overlap identical regions of interest, without spending hours to re-find identical areas.
We present here a first attempt in CLEM that uses an easily transferable sample carrier between light and scanning electron microscope and a software that controls the necessary functions of both microscope systems. This package of hardware and software for CLEM is now commercially available from the Carl Zeiss company (see also page 19). In addition we discuss potential applications with some first step-in protocols to start to explore these new opportunities.
The combination LM-SEM is beneficial because HRSEM has reached an imaging power, which can replace basic TEM ultra structure work and easily extends into large scale 3D serial sectioning by array-tomography [8] to correlate large CLSM data with serial section SEM data or by LM - FIB/SEM [9, 10]

Possible Workflows in CLEM

The different kinds of investigation are summarized in (fig. 1&2) according to surface or to interior related co-localization of structures on cells and tissues, or in cells and tissues, respectively. We distinguish two ways of sample immobilization - chemical fixation by cross-linking with chemical agents (cream-colored) and by rapid freezing (cryo-fixation - light blue colored). The individual major preparation steps are interrupted by colored arrow heads representing potential microscopy time-points in the work-flow of the sample preparation (yellow for RT LM/SEM and blue for cryo cLM/SEM imaging). Double-headed arrows represent levels where the samples can be transferred for- and backwards for re-investigation of the identical area. It is obvious that correlation with in-vivo data depends on the time and precision the sample preparation takes for EM compatibility. In addition it also highlights that at different process-steps it is possible to take control images (filled arrowheads) depending on the kind of investigation one is heading for. This may be of special interest when establishing new fluorescence probes.

LM-SEM Examples:

Since most bio-labs have access to fluorescence LM and dedicated SEM we just want to show with two examples that finding the same areas in cell culture work or tissue work is straight forward and gives additional information. The first example (fig. 3) is a chemically fixed and extracted cell culture as used for anti-body staining - with the corresponding SEM image of the identical area after freezing, dehydration and freeze-drying out of EtOH and 2nm W coating for cryo-SEM investigation (similar results can be obtained by water free critical point drying).
Another example (fig. 4) shows a high pressure frozen and embedded tissue sample. It is a serial section of root cells from mung beans that are partially infected with nitrogen binding bacteria. The same area can reproducibly be imaged in LM and SEM at various magnifications.
For starting experiments and first steps in LM/SEM applications please find supplementary material under the following link (http://www.emez.ethz.ch/education/clem) e.g. preparation protocols, which help you to get first results, list of fluorochromes which can be used and learn more about the details to perform CLEM experiments.