ID-8-IN-3280 Large-scale 3D reconstructions making use of multimodal correlative light and electron microscopy markers and analytical SEM imaging – fantasy or a new paradigm?

While 3D reconstructions at highest resolution will undoubtedly remain the domain of TEM imaging modalities, more and more interest has been generated in SEM based techniques. This was mainly driven by the need of ultrastructural information of really large volumes such as whole cells and cellular networks in tissue (for a comprehensive review on the variations of Array Tomography [1] based applications see [2]), which via TEM could only be obtained by effusive and tedious work.

An additional facet of SEM based methods are new perspectives of imaging with low energy electrons: Low voltage TEM has proven to reduce beam damage of carbon materials and to allow analytical imaging of e.g. fluorescent polymers or other π-electron systems (cf polymer/fullerene imaging at 60keV in [3]), which uses electron energy loss spectroscopy in the optical / near optical region as readout. Because of beam damage at 60keV we did not succeed to apply low-loss TEM-EELS to typical fluorescent dyes used as markers in biological samples, even though more irradiation-resistant QuantumDots - used as multi-color LM/TEM correlative markers - can be distinguished by (S)TEM-EELS and could potentially facilitate correlative “multi-color” light and electron microscopy in biology [4]. It may thus be an attractive alternative to investigate analytical imaging possibilities at really low electron energies where less ionization damage of the sample might be expected. This will reduce the mean free path beyond usability of TEM and thus novel SEM approaches to analytical surface imaging are needed.

On the way to such a scenario we have established a hierarchical Array Tomography workflow [5], which allows super-resolution LM and SEM imaging of fluorescently labeled serial thin sections of tissue material. A typical 3D reconstruction with a super-resolution fluorescent LM modality is shown in fig. 1, which illustrates the distribution of labeled acetylcholine receptors on the post-synaptic membrane of a neuromuscular junction (mouse diaphragm). However, such samples would now need to be imaged via low-loss SEM-EELS. At present such an SEM-modality is not yet established. As a first step in this direction fig. 2 shows a comparison of a typical SE image and an energy-filtered BSE equivalent recorded on a novel high-efficiency detector. Currently we analyze how usable EEL spectra can be extracted from such images.

[1] Micheva and Smith (2007) Neuron 55, 25, [2] Wacker and Schröder (2013) J Microscopy 252, 93, [3] Pfannmöller et al. (2011) Nano Lett. 11, 3099, [4] cf abstract at IMC2014 by Pfannmöller et al. and Pfannmöller et al. in: Proceedings Microscopy & Microanalysis (2011), [5] cf abstracts at IMC2014 by Wacker et al. and Röder et al.