Evaluation of Swin Transformer and knowledge transfer for denoising of super-resolution structured illumination microscopy data.

BACKGROUND: Convolutional neural network (CNN)-based methods have shown excellent performance in denoising and reconstruction of super-resolved structured illumination microscopy (SR-SIM) data. Therefore, CNN-based architectures have been the focus of existing studies. However, Swin Transformer, an alternative and recently proposed deep learning-based image restoration architecture, has not been fully investigated for denoising SR-SIM images. Furthermore, it has not been fully explored how well transfer learning strategies work for denoising SR-SIM images with different noise characteristics and recorded cell structures for these different types of deep learning-based methods. Currently, the scarcity of publicly available SR-SIM datasets limits the exploration of the performance and generalization capabilities of deep learning methods. RESULTS: In this work, we present SwinT-fairSIM, a novel method based on the Swin Transformer for restoring SR-SIM images with a low signal-to-noise ratio. The experimental results show that SwinT-fairSIM outperforms previous CNN-based denoising methods. Furthermore, as a second contribution, two types of transfer learning-namely, direct transfer and fine-tuning-were benchmarked in combination with SwinT-fairSIM and CNN-based methods for denoising SR-SIM data. Direct transfer did not prove to be a viable strategy, but fine-tuning produced results comparable to conventional training from scratch while saving computational time and potentially reducing the amount of training data required. As a third contribution, we publish four datasets of raw SIM images and already reconstructed SR-SIM images. These datasets cover two different types of cell structures, tubulin filaments and vesicle structures. Different noise levels are available for the tubulin filaments. CONCLUSION: The SwinT-fairSIM method is well suited for denoising SR-SIM images. By fine-tuning, already trained models can be easily adapted to different noise characteristics and cell structures. Furthermore, the provided datasets are structured in a way that the research community can readily use them for research on denoising, super-resolution, and transfer learning strategies.

Synthesis of the New Cyanine-Labeled Bacterial Lipooligosaccharides for Intracellular Imaging and in Vitro Microscopy Studies.

Endotoxin (lipooligosaccharide, LOS, and lipopolysaccharide, LPS) is the major molecular component of Gram-negative bacteria outer membrane, and very potent pro-inflammatory substance. Visualizing and tracking the distribution of the circulating endotoxin is one of the fundamental approaches to understand the molecular aspects of infection with subsequent inflammatory and immune responses, LPS also being a key player in the molecular dialogue between microbiota and host. While fluorescently labeled LPS has previously been used to track its subcellular localization and colocalization with TLR4 receptor and downstream effectors, our knowledge on lipopolysaccharide (LOS) localization and cellular activity remains almost unexplored. In this study, LOS was labeled with a novel fluorophore, Cy7N, featuring a large Stokes-shifted emission in the deep-red spectrum resulting in lower light scattering and better imaging contrast. The LOS-Cy7N chemical identity was determined by mass spectrometry, and immunoreactivity of the conjugate was evaluated. Interestingly, its application to microscopic imaging showed a faster cell internalization compared to LPS-Alexa488, despite that it is also CD14-dependent and undergoes the same endocytic pathway as LPS toward lysosomal detoxification. Our results suggest the use of the new infrared fluorophore Cy7N for cell imaging of labeled LOS by confocal fluorescence microscopy, and propose that LOS is imported in the cells by mechanisms different from those responsible for LPS uptake.

The expression of nestin delineates skeletal muscle differentiation in the developing rat esophagus.

The muscularis externa of the developing rodent esophagus is initially composed of smooth muscle, and later replaced by skeletal muscle in a craniocaudal progression. There is growing evidence of distinct developmental origins for esophageal smooth and skeletal muscles. However, the identification of skeletal muscle progenitor cells is controversial, and the detailed cell lineage of their descendants remains elusive. In the current study, we carried out multiple labeling immunofluorescence microscopy of nestin and muscle type-specific markers to characterize the dynamic process of rat esophageal myogenesis. The results showed that nestin was transiently expressed in immature esophageal smooth muscle cells in early developing stages. After nestin was downregulated in smooth muscle cells, a distinct population of nestin-positive cells emerged as skeletal muscle precursors. They were mitotically active, and subsequently co-expressed MyoD, followed by the embryonic and later the fast type of skeletal muscle myosin heavy chain. Thus, the cell lineage of esophageal skeletal muscle differentiation was established by an immunotyping approach, which revealed that skeletal myocytes arise from a distinct lineage rather than through transdifferentiation of smooth muscle cells during rat esophageal myogenesis.