The keratins are the typical intermediate filament proteins of epithelia, showing an outstanding degree of molecular diversity. Heteropolymeric filaments are formed by pairing of type I and type II molecules. In humans 54 functional keratin genes exist. They are expressed in highly specific patterns related to the epithelial type and stage of cellular differentiation. About half of all keratins—including numerous keratins characterized only recently—are restricted to the various compartments of hair follicles. As part of the epithelial cytoskeleton, keratins are important for the mechanical stability and integrity of epithelial cells and tissues. Moreover, some keratins also have regulatory functions and are involved in intracellular signaling pathways, e.g. protection from stress, wound healing, and apoptosis. Applying the new consensus nomenclature, this article summarizes, for all human keratins, their cell type and tissue distribution and their functional significance in relation to transgenic mouse models and human hereditary keratin diseases. Furthermore, since keratins also exhibit characteristic expression patterns in human tumors, several of them (notably K5, K7, K8/K18, K19, and K20) have great importance in immunohistochemical tumor diagnosis of carcinomas, in particular of unclear metastases and in precise classification and subtyping. Future research might open further fields of clinical application for this remarkable protein family.
Abstract Cardiac benign metastatic leiomyoma (BML) is a rare cardiac tumor that is usually asymptomatic, frequently misdiagnosed and may result in serious complications, including embolization, heart failure and death. This review highlights the importance of considering cardiac BML in the differential diagnosis of cardiac masses, especially in women with a history of uterine leiomyomas. This review summarizes the current knowledge about cardiac BML, including its demographics, clinical presentation, etio-pathogenesis, diagnosis, and management. The authors discuss the challenges associated with diagnosing cardiac BML and emphasize the importance of a thorough history, physical examination, and imaging studies. They also review the different treatment options for cardiac BML, including surgical resection and role of medical and surgical castration. Early diagnosis and management of cardiac BML is crucial to prevent complications. This review provides valuable insights for clinicians who may encounter this rare condition. By raising awareness of cardiac BML and its management strategies, this review can improve patient care and outcomes.
Diseases of the circulatory (Cardiovascular) system, Neoplasms. Tumors. Oncology. Including cancer and carcinogens
Background. Neurotransmitter adenosine and B-group vitamins have neuroprotective, remyelinizing and anti-neuroinflammatory properties. Despite the studies of these molecules for decades, the molecular mechanisms of their synergistic effect on neuroinflammation processes are unexplored and not systematized.Objective: to establish the molecular mechanisms of synergism of adenosine, thiamine, niacin and cyanocobalamin in counteracting the pathology of diabetic polyneuropathy (DPN).Material and methods. The molecular mechanisms of action of adenosine, thiamine (vitamin B1), niacin (vitamin PP) and cyanocobalamin (vitamin B12) in the pathophysiology of DPN were determined using functional analysis of genomic and proteomic databases.Results. The analysis of 20,180 annotated proteins of the human proteome identified 504 vitamin-PP-dependent, 22 vitamin-B1-dependent, 24 vitamin-B12-dependent and 50 adenosine-dependent proteins. The proteins of the human proteome were detected, the activity or levels of which are important for reducing neuroinflammation, remyelination, neurogenesis, biosynthesis of neuronal adenosine triphosphate, myelin homeostasis, neuroplasticity, neutralization of homocysteine, regeneration of nerve fibers and maintaining the endothelium of the microvascular bed.Conclusion. The discovered molecular mechanisms of synergism of the studied molecules are of fundamental importance for comprehension of the processes of neuroinflammation regulation and remyelination to prevent diabetic polyneuropathy and other neurodegenerative diseases.
Therapeutics. Pharmacology, Economics as a science
Ruben T. Lucassen, Tijn van de Luijtgaarden, Sander P. J. Moonemans
et al.
Vision-language models in pathology enable multimodal case retrieval and automated report generation. Many of the models developed so far, however, have been trained on pathology reports that include information which cannot be inferred from paired whole slide images (e.g., patient history), potentially leading to hallucinated sentences in generated reports. To this end, we investigate how the selection of information from pathology reports for vision-language modeling affects the quality of the multimodal representations and generated reports. More concretely, we compare a model trained on full reports against a model trained on preprocessed reports that only include sentences describing the cell and tissue appearances based on the H&E-stained slides. For the experiments, we built upon the BLIP-2 framework and used a cutaneous melanocytic lesion dataset of 42,433 H&E-stained whole slide images and 19,636 corresponding pathology reports. Model performance was assessed using image-to-text and text-to-image retrieval, as well as qualitative evaluation of the generated reports by an expert pathologist. Our results demonstrate that text preprocessing prevents hallucination in report generation. Despite the improvement in the quality of the generated reports, training the vision-language model on full reports showed better cross-modal retrieval performance.
Histopathology image analysis is fundamental to digital pathology, with hematoxylin and eosin (H&E) staining as the gold standard for diagnostic and prognostic assessments. While H&E imaging effectively highlights cellular and tissue structures, it lacks sensitivity to birefringence and tissue anisotropy, which are crucial for assessing collagen organization, fiber alignment, and microstructural alterations--key indicators of tumor progression, fibrosis, and other pathological conditions. To bridge this gap, we construct a polarization imaging system and curate a new dataset of over 13,000 paired Polar-H&E images. Visualizations of polarization properties reveal distinctive optical signatures in pathological tissues, underscoring its diagnostic value. Building on this dataset, we propose PolarHE, a dual-modality fusion framework that integrates H&E with polarization imaging, leveraging the latter ability to enhance tissue characterization. Our approach employs a feature decomposition strategy to disentangle common and modality specific features, ensuring effective multimodal representation learning. Through comprehensive validation, our approach significantly outperforms previous methods, achieving an accuracy of 86.70% on the Chaoyang dataset and 89.06% on the MHIST dataset. These results demonstrate that polarization imaging is a powerful and underutilized modality in computational pathology, enriching feature representation and improving diagnostic accuracy. PolarHE establishes a promising direction for multimodal learning, paving the way for more interpretable and generalizable pathology models.