FRaeppli, a multispectral imaging toolbox for cell tracing and dense tissue analysis in zebrafish

Visualizing cell shapes, interactions and lineages of differentiating cells is instrumental for understanding organ development and repair. Across species, strategies for stochastic multicolour labelling have greatly facilitated tracking cells in in vivo and mapping neuronal connectivity. Nevertheless, integrating multi-fluorophore information into the context of developing tissues in zebrafish is challenging given their cytoplasmic localization and spectral incompatibility with commonly used fluorescent markers. Here, we developed FRaeppli (Fish-Raeppli) expressing bright membrane- or nuclear-targeted fluorescent proteins for efficient cell shape analysis and tracking. High spatiotemporal activation flexibility is provided by the Gal4/UAS system together with Cre/lox and/or PhiC31integrase. The distinct spectra of the FRaeppli fluorescent proteins allow simultaneous imaging with GFP and infrared subcellular reporters or tissue landmarks. By tailoring hyperspectral protocols for time-efficient acquisition, we demonstrate FRaeppli s suitability for live imaging of complex internal organs, like the liver. Combining FRaeppli with polarity markers revealed previously unknown canalicular topologies between differentiating hepatocytes, reminiscent of the mammalian liver, suggesting shared developmental mechanisms. The multispectral FRaeppli toolbox thus enables the comprehensive analysis of intricate cellular morphologies, topologies and tissue lineages at single-cell resolution in zebrafish.

Ex vivo Fluorescence Confocal Microscopy: Chances and Changes in the Analysis of Breast Tissue

Background: Rapid histologic diagnosis of frozen sections is essential for a variety of surgical procedures. Frozen sections however, require specialized lab equipment, are prone to freezing artifacts and are not applicable to all types of tissue. Adipose tissue is especially difficult to process in frozen sections. Although these limitations are well known, no alternative method for microscopic tissue analysis that might replace frozen sections could be established. Our objective was to evaluate whether tissue imaging based on ex vivo fluorescent confocal microscopy (FCM) is applicable for rapid microscopic assessment of breast tumors specimens with abundant adipose tissue.Methods: We evaluated 17 tissue samples from mastectomy specimens, rich in adipose tissue, submitted to the department of pathology at the Medical University of Vienna. We conducted our study on the FCM VivaScope® 2500M-G4 (Mavig GmbH, Munich, Germany; Caliber I.D.; Rochester NY, USA).Results: When comparing FCM to frozen sections, we found a very similar overall processing time for FCM images and frozen sections respectively. Image quality was mostly superior to frozen sections (especially for adipose tissue and nuclear detail) but inferior to H&E stained FFPE sections. Limitations of the technology were uneven coloring, invisibility of ink applied for marking tissue margins and distortion artifacts if too much pressure is applied to the tissue. Conclusion: FCM has the potential to expand the application and usefulness of rapid tissue analysis as speed is comparable and quality exceeds that of frozen sections especially in tissues rich in adipose cells such as breast specimen.

Recovery of Altered Diabetic Myofibroblast Heterogeneity and Gene Expression Associated with CD301b+ Macrophages

Diabetic wound healing is associated with impaired function and reduced numbers of myofibroblasts, a heterogeneous cell population with varying capacities to promote repair. To determine how diabetes alters myofibroblast composition, we performed flow cytometry and spatial tissue analysis of myofibroblast subsets throughout the healing process in diabetic (db/db) and control (db/+) mouse skin. We observed reduced numbers of profibrotic SCA1+; CD34+; CD26+ myofibroblasts in diabetic wounds five days after injury, with decreased expression of fibrosis-associated genes compared to myofibroblasts from db/+ mouse wounds. While the abundance of myofibroblasts remained reduced in db/db mouse wounds compared to controls, the altered myofibroblast heterogeneity and gene expression in diabetic mice was improved seven days after injury. The natural correction of myofibroblast composition and gene expression in db/db wound beds temporally corresponds with a macrophage phenotypic switch. Correlation analysis from individual wound beds revealed that wound healing in control mice is associated with CD206+ macrophages, while the rescued myofibroblast phenotypes in diabetic wounds are correlated with increased CD301b+ macrophage numbers. These data demonstrate how diabetes impacts specific subsets of myofibroblasts and indicate that signaling capable of rescuing impaired diabetic wound healing could be different from signals that regulate wound healing under nonpathological conditions.

Spike protein multiorgan tropism suppressed by antibodies targeting SARS-CoV-2

While there is SARS-CoV-2 multiorgan tropism in severely infected COVID-19 patients, it’s unclear if this occurs in healthy young individuals. In addition, for antibodies that target the spike protein (SP), it’s unclear if these reduce SARS-CoV-2/SP multiorgan tropism equally. We used fluorescently labeled SP-NIRF to study viral behavior, using an in vivo dynamic imaging system and ex in vivo tissue analysis, in young mice. We found a SP body-wide biodistribution followed by a slow regional elimination, except for the liver, which showed an accumulation. SP uptake was highest for the lungs, and this was followed by kidney, heart and liver, but, unlike the choroid plexus, it was not detected in the brain parenchyma or CSF. Thus, the brain vascular barriers were effective in restricting the entry of SP into brain parenchyma in young healthy mice. While both anti-ACE2 and anti-SP antibodies suppressed SP biodistribution and organ uptake, anti-SP antibody was more effective. By extension, our data support the efficacy of these antibodies on SARS-CoV-2 multiorgan tropism, which could determine COVID-19 organ-specific outcomes.

Sinonasal teratocarcinosarcoma: a case report

<p class="abstract">Sinonasal teratocarcinosarcoma (TCS) is a very rare malignant neoplasm of sinonasal tract with intermixed teratomatous, carcinomatous and sarcomatous elements. While the diagnosis is largely based on tissue analysis and immunohistochemistry, the mode of management demands further study. Surgical resection with or without chemotherapy and radiation therapy is currently the most accepted treatment regimen. Locally aggressive, while also associated with metastatic lesions, SNTCS is not easily resectable owing to its location and possible intracranial extension. Due to its aggressive nature over one-third of TCS tend to recur leading to treatment failure with a mean survival time of 1.9 years. Possible differentials include squamous cell carcinoma, olfactory neuroblastoma, adenocarcinoma, malignant mixed tumor of salivary gland type, undifferentiated carcinoma, malignant craniopharyngioma, mucoepidermoid carcinoma, transitional carcinoma of Schneiderian type and adenosquamous carcinoma. In this report, we present a case of TCS in a 55 years old male patient who presented to us with complaints of hyposmia, blurring of vision, diplopia and epiphora.</p>

Generative modeling of histology tissue reduces human annotation effort for segmentation model development.

Segmentation of histology tissue whole side images is an important step for tissue analysis. Given enough annotated training data modern neural networks are capable accurate reproducible segmentation, however, the annotation of training datasets is time consuming. Techniques such as human in the loop annotation attempt to reduce this annotation burden, but still require a large amount of initial annotation. Semi-supervised learning, a technique which leverages both labeled and unlabeled data to learn features has shown promise for easing the burden of annotation. Towards this goal, we employ a recently published semi-supervised method: datasetGAN for the segmentation of glomeruli from renal biopsy images. We compare the performance of models trained using datasetGAN and traditional annotation and show that datasetGAN significantly reduces the amount of annotation required to develop a highly performing segmentation model. We also explore the usefulness of using datasetGAN for transfer learning and find that this greatly enhances the performance when a limited number of whole slide images are used for training.