Abstracts of the 3rd European Super-Resolution User-Club Meeting

Alberto Diaspro, Italian Institute of Technology, Genoa, Italy

It is well established that for fluoresence, the most popular imaging mode in optical microscopy, the diffraction barrier no longer provides an unsurpassable limitation for resolution and localization accuracy. Furthermore, the terms "super-resolution" and the earlier coined "optical nanoscopy", have been implemented in real far field optical microscopes, and are now available for everyone to use without extreme complexity. Since the new actor for getting a resolution improvement is the fluorescent molecule being used and its photophysics is sometimes sample-related, at the Istituto Italiano di Tecnologia (IIT) we decided to attack the problem from different sides.

For this reason, we implemented and adapted a number of different optical fluorescence set-ups towards super-resolution. In considering super-resolution approaches, we also considered their implementation for imaging thick (>50 um)
biological specimens. Within the framework of stochastic read-out methods based on single molecule localization, we implemented individual molecule localization (IML) within a selective plane illumination microscopy (SPIM) architecture in order to address 3D super-resolution imaging.

An application related to mammalian cellular spheroids will be presented. Recently, the very same system has been endowed of the possibility of performing IML-SPIM using two-photon photo-activation. Moreover, an optical solution will be presented to overcome the SPIM request of inserting the sample in a cylindrical sample holder. On the side of the targeted read-out methods like STED, we introduced two-photon excitation including the possibility of using a single wavelength (SW) both for two-photon excitation and STED depletion by implementing a SW-2PE-STED microscope. Since STED provides immediate access to the resolution improvement without requiring computational tools, we expanded the method to direct writing lithography and multimodal super resolution microscopy. Direct writing lithography at super resolution will be outlined demonstrating the possibility of exploiting the STED-like mechanism. On the other side, in terms of multimodal super resolution microscopy, we will show the coupling of STED and Atomic Force Microscopy. This allows us to more precisely and specifically address AFM and to use it in an active way by tip-particle interactions. So far, a variety of architectures will be outlined in regard to specific applications demanding nano-scale investigations.

This aspect is particularly in tune with the multidisciplinary IIT scientific environment: from neuroscience to drug discovery and delivery, from smart materials to computational imaging and more.

Ilaria Testa, Max Planck Institute for Biophysical Chemistry, Department of NanoBiophotonics, Göttingen, Germany

Lens-based fluorescence microscopy, which has long been limited in resolution to >200 nanometer by diffraction, is rapidly evolving into a nanoscale imaging technique. Here, we show that emergent RESOLFT fluorescence microscopy enables fast and continuous imaging of sensitive, nanosized features in living brain tissue. Using low intensity illumination to switch photochromic fluorescent proteins reversibly between a fluorescent ON-state and a non-fluorescent OFF-state, we obtained more than a 3-fold increase in all three spatial dimensions over that of confocal microscopy. Dendritic spines located 10–50 μm deep inside living organotypic hippocampal brain slices were recorded for hours without signs of degradation.

Using a fast-switching fluorescent protein increased the imaging speed 50-fold over reported RESOLFT schemes, which in turn enabled us to record spontaneous and stimulated changes of dendritic actin filaments and spine morphology occurring on time scales from seconds to hours.

Silvio Rizzoli, European Neuroscience Institute, Göttingen, Germany

Material exchange between cellular compartments relies on the formation of cargo-loaded vesicles from donor compartments and their subsequent delivery to acceptor compartments. These processes involve a plethora of protein cofactors whose functions are well known and are conserved between different pathways. Nevertheless, the quantitative organization of cofactors is not yet available in any membrane trafficking pathway. We analyzed here synaptic vesicle recycling, the pathway that maintains neuronal transmission. We measured the absolute copy numbers and estimated the locations for 60 proteins that sum to more than 40 % of the protein in this pathway. Together with measurements of synaptic morphology, these data allowed us to generate a model of the synapse that also takes into account the protein structures. Finally, we also estimated the abundance of a further 1,140 proteins. Synaptic vesicle recycling was dominated by proteins involved in cargo delivery. Cofactors responsible for the recovery of vesicle components from the acceptor compartment (plasma membrane) were less abundant and likely represent bottlenecks for synaptic function.

Giuseppe Vicidomini, Italian Institute of Technology, Genoa, Italy

Stimulated emission depletion (STED) microscopy reaches sub-diffraction spatial resolution by reducing the spatial extend of the volume from which fluorescence signal is effectively collected. To this end, it employs a regular excitation beam co-aligned with a doughnut-shaped beam. The role of the second beam (usually called the STED beam) is to transiently quench (by stimulated emission) all the fluorophores in the focal excitation region except those located in the close proximity of the doughnut center, where the STED beam intensity is nearly zero. Only when the STED beam is intense enough, can the region that fluoresces reach sizes far below the diffraction limit. Theoretically, STED microscopy can approach "infinite" spatial resolution, but at the cost of very high STED beam intensities. Practically, possible photodamage and phototoxic effects limit the amount of STED beam intensities that can be focused on the sample, and thereby the ultimate resolution of a STED microscope.

Recently, it has been demonstrated that by using time-gated detection it is possible to significantly reduce the intensity demand of the STED beam without losing in spatial resolution. Reduction of STED beam intensity not only has great impact for live-cell imaging, but triggers also the possibility to efficiently implement STED microscopy with STED beam operating in continuous-mode (CW) and thereby to substantially reduce cost and complexity of a STED microscope. This implementation is usually called gCW-STED microscope.

We first analyze the fundamental principles and limits of the gCW-STED technique. Gated CW-STED microscopy is in essence limited (only) by the reduction of the signal that is associated with gating. Reduction in signal translates into a reduction of signal-to-noise (SNR) and -background (SBR) ratio of the images, which could limit the effective resolution of the system.

Thereby, we propose different hardware- and software-based approaches able to improve the SNR and SBR of the gCW-STED images. We finally show that the synergy of these approaches leads to a versatile gCW-STED implementation able to reach spatial resolution below 50 nm with dose of STED beam’s light below 100 mW.

Ralf Jacob, Philipps-Universität Marburg, Germany

The plasma membrane of epithelial cells is divided into separate membrane compartments, the apical and the basolateral domain. This polarity is maintained by intracellular machinery that directs newly synthesized material into the correct target membrane [1]. Moreover, glycoproteins and lipids can enter distinct pathways for apical or basolateral delivery. To discover these pathways and unravel the complex network of post Golgi endosomal apical trafficking a combination of biochemical and microscopic approaches was used.
Targeting of galectin-3, an apical sorting receptor [2], was visualized by epifluorescence-, TIRF- and GSD-microscopy. Here, we could show that the lectin cycles between the plasma membrane and endosomal organelles.

Intracellular trafficking is maintained by microtubules as intracellular transport tracks. Posttranslational modification of these tracks is required for correct delivery to the apical plasma membrane. Highly resolved GSD-imaging depicts the distribution of these modifications along single microtubules [3].

Eric Hosy, CNRS, University of Bordeaux, France

The majority of synapses in the central nervous system use glutamate as a neurotransmitter, and the strength of synaptic transmission is proportional to the number of glutamate receptors (AMPA type) present under the synaptic glutamate release site. Many studies have reported modification of AMPA receptor quantity, organization or composition in response to various physiological stimuli which underlie synaptic maturation and plasticity, memory, disease, etc. However, available optical tools have not led to a precise description of the basic organization of receptors due to the limited pointing accuracy of the optical microscopy. The emergence of super-resolution techniques has broken this limitation barrier, allowing us to understand the organization of the AMPA receptors, and the variation of mobility as a function of its localization inside the synapse.

Here we used different super-resolution techniques (GS-DIM, STED, PALM and U-PAINT) to study extensively the organisation and the mobility of AMPAR inside the synapse and we discovered that AMPA receptors are not randomly distributed inside the synapse or even the PSD, but structured in nanodomains of about 80nm. Such distribution allows for the maintenance of the high fidelity of the synaptic response. In parallel, perturbation of one of the main scaffold protein of the PSD, PSD95, affect the dynamic organization of AMPAR and the synaptic currents in the same range.

Colin Sheppard, Italian Institute of Technology, Genoa, Italy

Different methods for achieving super-resolution are reviewed. First we discuss different definitions of resolution. In particular, we say what is often called resolution, but is not really resolution. The fundamental principles behind super-resolution are analyzed, and different schemes classified. We define three different classes that are sometimes called super-resolution, one of these being true, unrestricted, super-resolution. Basically, according to Lukosz, it is not resolution that is limited, but rather the capacity of an optical instrument to transmit information. According to this principle, resolution can be improved by trading off some other property of the object, or by using some a priori information about the object.

Hell has explained that true super-resolution is based on the property of switching. However, others have maintained that this is unnecessarily restricted, and that nonlinearity is the requirement, of which switching is only a special case. Hell has also made the point that STED is a far-field technique. While this may be true in some sense, it seems that fluorescence in general could be regarded as a near-field process in itself, i.e. the fluorescent molecule itself is interacting with light in the near-field. Perhaps this is just semantics, or maybe this is a principle that could give insight into which schemes could not form the basis of super-resolution.