ABT-888 (Veliparib) In hematoxylin and eosin H
In hematoxylin and eosin (H&E)-stained sections MDBs appear as irregular eosinophilic ABT-888 (Veliparib) of variable size. They have to be distinguished from condensed cytoplasmic areas and can be specifically identified by immunostaining since they contain keratins, ubiquitin, sequestosome 1/p62 (p62) and other stress proteins [, , ]. For detection and identification of MDBs immunohistochemistry is superior to H&E staining due to higher sensitivity and specificity , and their presence may be underestimated in conventionally stained sections . Longitudinal studies revealed MDBs as indicators of poor prognosis and predictors of disease progression [, , , ]. Hepatocyte necrosis, apoptosis, bilirubinostasis, and ductular reaction may be present [2,3]. The predominantly lobular inflammation is usually of mixed type including mononuclear cells, polymorphonuclear leukocytes, clustered Kupffer cells and occasional microgranulomas. Hepatocytes in zone 3 are often surrounded by a rim of collagen (pericellular fibrosis; Fig. 1B) or are replaced by fibrous tissue (central sclerosis), eventually with obliteration of the central vein. Pericellular fibrosis appears to be more pronounced in the presence of MDBs [18,19]. In the absence of active disease this fibrosis pattern suggests prior episodes of SH. With disease progression incomplete fibrous septa arise and bridging septal fibrosis and cirrhosis may follow .
In their morphologic appearance ASH and NASH closely resemble each other but in most instances the lesions in NASH are less severe than in ASH. For example, in NASH central sclerosis and veno-occlusive lesions are very rare, MDBs are less distinct and clusters of neutrophils are much less common than in ASH [2,3,10]. Hence, intense neutrophil granulocyte infiltration points to alcoholic etiology . In this situation, about 25% of MDB-containing hepatocytes are associated with, or penetrated by, neutrophils (termed “satellitosis”; Fig. 1, Fig. 2).
Of note, immunohistochemistry revealed a correlation of MDBs and satellitosis with TNF-α and IL-1 expression . Pediatric NASH may differ from the adult form by showing either panacinar or periportal steatosis, increased portal mononuclear inflammation and fibrosis, few polymorphonuclear leukocytes, minimal or no hepatocyte ballooning or MDBs, particularly in younger children [2,10,23].
Relevance of animal models in human NAFLD/NASH research A variety of interacting genetic and environmental factors determine human NAFLD/NASH. To obtain firm insights into the pathogenesis of the disease it is important to reproduce pathologic situations under standardized conditions in vitro, in isolated perfused organs and/or in whole animals . Considerable efforts have been made to generate mouse models, since mice share many physiological, anatomical and metabolic features with humans and are particularly suitable for genetic manipulations. As evident from this and other reviews, the ideal animal model, that resembles human NAFLD/NASH in clinical manifestations, etiology, pathophysiology and pathomorphology, does not exist. Since experimental systems reproduce only certain aspects of human disease it is necessary to focus on decisive features and interpret them appropriately [31,36]. Many discoveries in NAFLD/NASH research are based on dietary models or leptin-deficient (ob/ob) or leptin-resistant (db/db) rodents [32,34]. Feeding fat-enriched diets with or without additives produces features closely resembling human metabolic syndrome and its pathophysiologic consequences. The MCD model differs in its clinical consequences from human disease since the mice lose weight and lack IR. Although the leptin-related models imitate human metabolic syndrome, leptin or leptin receptor mutations do not prevail in NASH patients and leptin levels only poorly correlate with progression of simple steatosis to NASH . Striking is the lack of clear-cut MDB formation in most models, which together with hepatocyte ballooning is regarded as diagnostically and prognostically important morphologic feature in human disease. Although ballooning and MDBs are mentioned in some reports they were usually not adequately characterized. To unequivocally identify MDBs, light microscopy should be complemented by immunohistochemistry using antibodies to keratins and/or p62 and/or ubiquitin. Consequently, only the chronic DDC and GF intoxication and the K18 deficiency models, although etiologically different from human NASH, resemble the entire morphologic spectrum of human NAFLD/NASH [15,16,120]. Since the keratin IF cytoskeleton is evolutionary highly conserved and keratins as major components of MDBs are in addition to their mechanical functions involved in a variety of vital cellular pathways , it can be expected that mechanisms leading to MDB formation, irrespective of their cause, are particularly relevant to the understanding of essential pathogenic principles in NASH.