Three-Dimensional Histopathology of the Human Liver

Abstract

The liver performs hundreds of functions and is considered second only to the brain in structural and functional complexity. One critical spatial aspect of liver architecture is the division of the primary functional unit (the lobule) into "zones" where cells perform different functions (functional zonation). Lobular zones also contain different types of vascular and biliary structures (structural zonation). The homeostasis of this intricate architecture is critical for liver function and health, and its disruption is pathologic. Visual examination of tissue under the microscope (histopathology) is the gold standard for diagnosing many liver diseases, including the severe scar-tissue buildup known as cirrhosis, examined here. The liver's architecture at cell-level resolution has been elucidated largely via two-dimensional (2D) traditional histology, the interrogation of thin slices of tissue under a microscope. 2D histology has provided centuries' worth of information but is not without limitations. For one, sectioning tissue into thin slices can physically alter the examined structures. Beyond that, the 2D nature of histology can misrepresent actual biological structures: 1) three-dimensionally (3D) convoluted structures often appear as multiple distinct objects in a 2D cross-section, 2) cross-sections may not accurately capture the distribution of objects within a tissue volume, and 3) 2D slices may simply not contain rare or sparse objects. To overcome these limitations, groups have recently begun imaging thick liver samples via 3D immunostaining and tissue clearing, which render the tissue optically clear and permit deeper imaging than traditional methods. The rodent liver was the subject of most of this work; the few 3D human liver studies have imaged less than one lobule to a depth equivalent to approximately 20 traditional tissue sections. Here, we have surpassed existing 3D human liver imaging studies by imaging a depth equivalent to approximately 100 traditional tissue sections, roughly 200 times the previously captured tissue volume. We have developed a 3D imaging and analysis pipeline to define structures across multiple human liver lobules empirically. First, we optimized immunostaining for diverse hepatic cell types and structures, customized tissue clearing to work robustly in the liver, and established a large-field, micrometer-resolution 3D confocal imaging workflow. Next, we developed a pipeline for image processing and 3D analysis & morphometry of branching and non-branching structures. Utilizing this, we uncover the dysregulation of the structurally zonated vascular and biliary networks of the human liver within cirrhosis. We find a significant expansion of periportal structures with significant regression of pericentral structures. These findings suggest a decentralization phenotype within cirrhosis of multiple etiologies, with possible functional consequences for blood drainage and nutrient and drug metabolism. Together, these efforts provide methods for staining, clearing, reconstructing, and analyzing structures within cell-dense tissues like the liver; unveil never-before-seen-in-3D human liver structures across multiple size scales; and reveal volumetric dysregulations of structural and functional zonation in cirrhosis.