Today’s scientists wish to understand the biological processes at a molecular level. Fluorescent microscopy, allowing the labeling of biomolecules with high specificity by the use of antibodies or by tagging with fluorescent proteins, appeared as a great tool for such an aim. However, the spatial resolution of fluorescent microscopes is limited by the wavelength of light at around ~220 nm (Figure 1). While this resolution is sufficient to visualize cellular organelles, it is not sufficient to resolve the molecular organization of these complexes. In parallel with the development of the optical approaches, multiple tools have been invented to overcome this obstacle. For example, electron microscopy (EM) allows ultrastructural-level analyses of biological structures, and the detection of specific molecules can also be accomplished by immunogold labeling. However, achieving high-density immunogold labeling is challenging, and it is far from being efficient. Scanning probe and nearfield optical scanning microscopy also offer outstanding resolution, but, unfortunately, they are largely surface-bound and not able to image the intact cellular interior, and hence their popularity is limited. In the end, the main goal remained the same: to go under the diffraction limit by using visible light. The breakthrough came from revisiting the principles of fluorescence microscopy, and that is what we call fluorescence nanoscopy today (also often referred to as super-resolution microscopy (SRM)). By bridging the gap between conventional light microscopy and structural biology, the founders of these new concepts were awarded the Nobel Prize in Chemistry in 2014. In this review, we summarize the principles of the main nanoscopic approaches, and their impact on understanding the molecular organization of several highly conserved cellular machineries by using Drosophila models.

As it is outlined above, despite the rather different rational designs, each SRM technique is suitable to break the resolution limit. Nevertheless, in large part owing to the differences in their technical concept, each method has its own benefits and drawbacks, which are summarized in Table 1, together with the main technical features. The choice of method in the case of a biological sample of interest must take into account these differences, and ideally, it is advised to compare these approaches experimentally. For example, when we decided to study the sarcomeric distribution of a set of actin regulatory proteins, the Drosophila Titin ortholog (Kettin) was used as a ‘resolution’ control. Based on immunogold EM images, the 16th immunoglobulin domain of Kettin (Ig16) is present in two stripes at the edges of the Z-disk separated by ~100 nm ^[1]. Due to their close proximity, the individual ‘lines’ cannot be resolved as distinct structures with confocal laser scanning microscopes (Figure 2). However, SIM or STED microscopy already provides significant improvements, while STORM is able to resolve the sarcomeric structures with a high resolution, which can be further improved by structure averaging (Figure 2). Thus, Z-disk localization of Kettin not only perfectly demonstrated the resolving power of the different nanoscopic approaches but also provided us with a useful batch of information on the available techniques, which was key to our subsequent studies on sarcomere organization. This example will be reviewed, in detail, in the following sections together with several other Drosophila models highlighting how resolution improvement can lead to major discoveries in various fields of life sciences.