Each dendrite integrates the activity of a large number of synapses that determines the output of the parent cell. It is known that the computational algorithm of synaptic integration depends on a wealth of spatio-temporal parameters of pre- and postsynaptic structures (Goetz et al.
2021). However, most neuron models rely on assumptions even when basic parameters such as density of excitatory vs. inhibitory synapses should be taken into account due to scarcity of essential structural and functional parameters. As a consequence, point neuron models prevail and are obliged to employ overt simplification for calculating output activity (for review, see Tzilivaki et al.
2019). From a functional point of view, the many synapses acting on the dendrites of a neuron are assumed not only to form an integrative system but also to interact with each other (Hu and Vervaeke
2018; Poirazi and Papoutsi
2020). In this process, precision in terms of timing and voltage parameters is likely to be significant and often goes beyond the ability of neuronal models which use estimations and averaging due largely to lack of precise data. The structural constraints of synaptic connections have been traditionally studied using transmission electron microscopy (TEM). A clear advantage of TEM is that the visualized structures can be paralleled with functional properties such as asymmetric (type I) versus symmetric (type II) synapses being presumed excitatory versus inhibitory role, respectively (Colonier
1968; Gray
1959). TEM is commonly used for disclosing synaptic input–output connectivity characteristics of identified structures, for example, labeled boutons, dendrites and cell bodies. Typically, single ultrathin sections or a short section series is examined. While such an approach could unravel invaluable synaptic information for the population of structures, the precise spatial relationship between pre- and postsynaptic components could not be tackled for the entire length of dendritic processes. From this perspective, distal dendrites have long been in the focus of interest since little is known about their role and contribution in shaping the output activity of neurons. In this regard, synapse type and distribution on the dendritic surface represent important parameters that determine dendritic signal interaction and propagation (Shepherd et al.
1985). Recent advancements in volume electron microscopic applications (FIB-SEM, SBEM, for rev., see Briggmann and Bock
2012) allow examination of large tissue samples as compared with traditional TEM. SBEM is particularly suited to image tissue volume in the range of mm that matches the spatial scale of most dendrites. On the other hand, a major challenge remains how to exploit the benefit of SBEM on neurochemically identified structures. Neurochemical approaches commonly compromise membrane integrity which, however, is an essential requirement for electron microscopy (EM) including SBEM. Recent attempts tried to solve the above issue. Using the fluorescent signal of the labeled structures, the sections were subjected to confocal- and light microscopy (LM), after fixation, followed by collecting serial EM images with the help of fiducial landmarks (Maclachlan et al.
2018). Another study used the mirror technique that employed adjoining sections with cut cell bodies on the mirror surfaces. In one section, immunohistochemistry was carried out to visualize the neurochemical marker and in the other section, the complement of the cell bodies and the emerging dendrites were reconstructed using serial section TEM (Talapka et al.
2021). While both reports managed to preserve high-quality ultrastructure, the alignment process required the use of light microscopic sections which have a limited thickness, typically, not more than 100 µm. The spatial constraint of the above methods does not favor the full reconstruction of dendrites which often run several hundreds of µm or longer. The method presented here is an adaption of the mirror technique (Talapka et al.
2021) which offers 3D-electron microscopic reconstruction of structures in a large tissue volume, for example, the synaptome of entire dendrites. It is based on the identification of neuronal cell bodies in the surface of thick tissue blocks using confocal reflectance imaging instead of LM of sections. This method opens new vistas to carry out combined light- and electron microscopic analysis on neurochemically identified structures for providing quantitative measures of synapse organization of specific neuron types.