Background

Histology is the use of immunodetection-based reagents to visualize cellular and extracellular compartments of tissue, usually via thin-slice light or fluorescent microscopic analysis. Histologic analysis is routinely applied daily in clinical and research settings and is considered essential for visually assessing the structure and pathology of the tissue under study. Advances in cell-based immunohistochemical and immunofluorescent detection protocols have led to multi-faceted progress in most biological fields (Kamimoto et al., 2020; Ramachandran et al., 2020; Tan et al., 2020; Taylor, 2020; Guilliams et al., 2022). Historically, light microscopy has been utilized by pathologists to determine disease etiology as well as determining the extent or degree of damage. More recently, there has been a need, more so in research settings, to provide quantitation of cellular infiltrates, proliferation, and cell-specific expression of selected protein markers, to define the cellular response to damage as well as to identify pathways of repair (Ben-Moshe and Itzkovitz, 2019; Saviano et al., 2020; Hu et al., 2022). The typical method of quantifying cells or fibrosis has been to randomly select regions of interest (ROI) and manually count cells or notable puncta (Abercrombie, 1946; Endo et al., 2002; Baratta et al., 2009; Vogel et al., 2015; Baghdasaryan et al., 2016; Husain et al., 2018). There are no standards for this approach but, by definition, there are inadvertent opportunities for significant regional cherry-picking, and as it does not take into account the entire tissue area of the slide, this method may not provide accurate quantitation (Baker, 2016; Murphy and Aguinis, 2022). Moreover, staining variation between and within slides, small sample sizes (< 1,000 detections/sample), extensive use of laboratory time, and no agreed-upon method all limit the utility and applicability of these methods between experiments and laboratories (Gurcan et al., 2009; Alturkistani et al., 2016; Bogen, 2019).

To address these issues, we focus on this over-arching problem using a specific model organ, the liver. The liver serves as the human body’s biofilter and can be highly reactive in response to disease, genetic predisposition, or external challenges (Nagy et al., 2020). The main cell types in the liver include parenchymal hepatocytes, liver sinusoidal endothelial cells, Kupffer cells/macrophages, cholangiocytes, and endothelium (Heymann and Tacke, 2016). Biomarkers of liver-related injury predominately present through the recruitment of F4/80+ macrophages, an increase in the extracellular production of collagen, and the rapid expansion of cholangiocytes (Fabris et al., 2019).

In this report, we present our experience with an automated whole-slide approach to quantify cells and fibrosis in standard liver slide histology termed LiverQuant. LiverQuant is a new customizable method that uses scripts written in QuPath to automate the detection of macrophages, fibrosis, and cholangiocytes in whole-slide scanned images of liver tissue. QuPath is a free, open-source program for analyzing pathology, and can be used by anyone with an efficient computer (Bankhead et al., 2017). Recently, LiverQuant methodology was used to quantitate various cell types and fibrosis in a novel genetic-based mouse model of biliary atresia, the liver conditional knockout of Pkd1l1 (Hellen et al., 2023).

While this protocol is intended for the detection of specific cell and extracellular markers within the liver, its principles and code can be optimized for use in other organs, within any species, and can be accomplished using either Mac or Windows-based computers.