Implementation of Multi-Color Super-Resolution Microscopy into a High-Throughput Platform for Quantitative Imaging of Cellular Structures

Abstract

In this work, single-molecule super-resolution microscopy with photoswitchable fluorophores was integrated into a multi-component microscopy platform. This platform enabled combined high-throughput widefield/confocal microscopy, in order to first validate relevant target cells, and subsequent sub-diffractional imaging of selected cells via direct stochastic optical reconstruction microscopy (dSTORM). In a first project, a single-molecule sensitive fluorescence microscope was set up, and equipped for single-molecule localization microscopy. As suitable photoswitchable fluorophores for dual-colour imaging, Alexa Fluor 647/ Cy5 and Alexa Fluor 532 were identified. Software solutions were developed to analyze the raw data of a localization-based experiment and to reconstruct super-resolution images. Strategies to correct for chromatic aberration between two spectrally separate channels were developed by using fiducial markers and a registration software. The achieved localization precision was determined to be sigma(SMLM) = 12.5 nm in the image plane. The chromatic aberration could be corrected with a precision of 10 nm. In order to integrate the super-resolution imaging into the microscopy platform, a work-flow was developed that allowed transferring the sample between different microscope setups and relocating identified targets into the field of view. This procedure is based either on external reference markers on the respective cellarray/coverslip, or builds on using a stitched overview image with low resolution. Individual cells were relocated into the field of view with a precision of less than 4 µm. For validation, widefield low-resolution images of a cis- and a trans-Golgi marker (GalT and Gm130) in cells were recorded and served to select individual cells for further analysis by confocal and super-resolution imaging. Images of representative control cells as well as cells treated with nocodazol (depolymerizes tubul network) and brefeldin A (relocates Golgi into the endosplasmic reticulum (ER)) were analyzed for colocalization using intensity-based colocalization analysis (ICA). The results indicated that nocodazol treatment fragmentizes the Golgi complex by the depolymerization of the tubul network. Contrarily, brefeldin A leads to a relocalization of the Golgi into the ER. The colocalization of GalT and GM130, calculated from confocal images, decreased from 9.6% (control cells) over 7.7% (nocodazol treatment) to 5.5% (brefeldin A treatment). This observation is in accordance with reports in the literature. The same decrease of colocalization was calculated from the super-resolution images (5.5%, 4.2% and 1.6% respectively). Next to conventional colocalization analysis, single-molecule localization microscopy data can directly be analyzed using the coordinates of individual fluorophores (coordinate-based colocalization (CBC)). This analysis showed that after treatment with nocodazole, the colocalization of GalT and GM130 was not affected compared to the control cell. The Golgi complex disassembled into smaller structures, but the internal membrane structure remained intact. The treatment with brefeldin A degraded the whole Golgi and membrane assembly and therefore changed the colocalization pattern of the cis- and trans-Golgi marker. In a second project, newly developed and sequence-specific oligonucleotide probes for messenger ribonulceic acid (mRNA) detection (synthesized in the group of professor Mokhir, Erlangen) were evaluated at the single-molecule level. This mRNA probe is synthesized to be intrinsically photoactivatable and specific due to a dual-strand design. The probe consists of a first oligonucleotide strand, which is conjugated to a fluorophore and linked to a quencher by a cleavable linker (-CHS=SHC-). The second oligonucleotide strand is conjugated to a photosensitizer, which upon irradiation with light catalytically produces singlet oxygen. Singlet oxygen cleaves the linker between the quencher and the fluorophore, releases the quencher and restores the fluorescence signal. Single-molecule experiments were performed to validate the functioning of a probe that was designed to specifically bind actin-mRNA. Here, the fluorophore was conjugated to the first oligonucleotide strand by the cleavable linker. Illumination with light leads to the production of singlet oxygen, subsequent cleavage of the linker and release of the fluorophore. The experiments showed that the cleavage reaction is highly specific and is not triggered if no activation light is applied, nor the strand of the photosensitizer includes a mismatching sequence to the target sequence. If the fluorophore was conjugated to the first oligonucleotide strand without a cleavable linker, the singlet oxygen for cleavage did not induce a release of the fluorophore.