3D reconstruction of human liver tissue at cellular resolution

Wesley B Fabyan; Chelsea L Fortin; Dorice L Goune; Heidi L Kenerson; Susana P Simmonds Bohorquez; Jonathan T C Liu; Matthew M Yeh; Rotonya M Carr; Raymond S W Yeung; Kelly R Stevens

doi:10.1126/sciadv.adz2299

. 2026 Feb 18;12(8):eadz2299. doi: 10.1126/sciadv.adz2299

3D reconstruction of human liver tissue at cellular resolution

Wesley B Fabyan ^1,^2,^†, Chelsea L Fortin ^1,^2,^3,^†, Dorice L Goune ^1,², Heidi L Kenerson ⁴, Susana P Simmonds Bohorquez ^1,², Jonathan T C Liu ^1,^3,⁵, Matthew M Yeh ³, Rotonya M Carr ⁶, Raymond S W Yeung ⁴, Kelly R Stevens ^1,^2,^3,^7,^*

^¹Department of Bioengineering, University of Washington, Seattle, WA 98195, USA.

^²Institute for Stem Cells and Regenerative Medicine, University of Washington, Seattle, WA 98195, USA.

^³Department of Laboratory Medicine and Pathology, University of Washington, Seattle, WA 98195, USA.

^⁴Department of Surgery, University of Washington, Seattle, WA 98195, USA.

^⁵Department of Mechanical Engineering, University of Washington, Seattle, WA 98195, USA.

^⁶Department of Medicine, Division of Gastroenterology, University of Washington, Seattle, WA 98195, USA.

^⁷Brotman Baty Institute for Precision Medicine, Seattle, WA 98195, USA.

Corresponding author. Email: ksteve@uw.edu

^†

These authors contributed equally to this work.

Roles

Wesley B Fabyan: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing - original draft, Writing - review & editing

Chelsea L Fortin: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Validation, Visualization, Writing - original draft, Writing - review & editing

Dorice L Goune: Data curation, Formal analysis, Resources, Validation, Visualization, Writing - review & editing

Heidi L Kenerson: Methodology, Resources, Writing - review & editing

Susana P Simmonds Bohorquez: Investigation, Validation, Visualization, Writing - review & editing

Jonathan T C Liu: Methodology, Resources, Supervision, Writing - review & editing

Matthew M Yeh: Conceptualization, Data curation, Investigation, Methodology, Visualization, Writing - original draft, Writing - review & editing

Rotonya M Carr: Conceptualization, Formal analysis, Writing - review & editing

Raymond S W Yeung: Resources, Writing - review & editing

Kelly R Stevens: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Visualization, Writing - review & editing

PMCID: PMC12915621 PMID: 41706846

Abstract

The liver contains an intricate microstructure that is critical for proper liver function. Architectural disruption of this spatial structure is pathologic. Unfortunately, two-dimensional (2D) histopathology—the gold standard for pathological understanding of many liver diseases—can misrepresent or leave gaps in our understanding of complex 3D structural features. Here, we used immunostaining, tissue clearing, microscopy, and computational software to create 3D multilobular reconstructions of both nonfibrotic and cirrhotic human liver tissue. We found that spatial architecture in human cirrhotic liver samples with varying etiologies had sinusoid zonal dysregulation, reduction in glutamine synthetase–expressing pericentral hepatocytes, regression of central vein networks, disruption of hepatic arterial networks, and fragmentation of biliary networks, which together suggest a pro-portalization/decentralization phenotype in cirrhotic tissue. Further implementation of 3D pathological analyses may provide a deeper understanding of cirrhotic pathobiology and inspire treatments for liver disease.

3D human liver reconstructions reveal architectural hallmarks of cirrhosis and offer structural insights for tissue engineering.

INTRODUCTION

Microscopy stands as an indispensable tool in biological sciences, offering researchers the ability to delve into the intricate world of cells and tissues with exquisite detail. Since its inception in the 17th century, microscopy has undergone numerous revolutionary advances propelled by innovations in optics, imaging technology, chemistries, and computational analyses (1). These innovations have led to vastly improved spatial resolution, imaging depth, and experimental throughput (1). Recently, such advances have equipped scientists with an enhanced imaging toolkit to overcome limitations of microscopic imaging of two-dimensional (2D) sections, which sliced convoluted structures into multiple objects, inaccurately captured object distribution, and routinely missed sparse objects (2). This toolkit in tissue clearing, labeling, and imaging has enabled 3D volumetric imaging of tissues at cellular resolution, better elucidating tissue structure and its disease-associated alterations. Because tissue structure and function are intimately linked in biological tissues, such 3D analyses may provide a deeper understanding of functional abnormalities and inspire treatments for disease.

The liver is an exemplary organ where its 3D microarchitecture is integral to its proper function, and disruption to its structural architecture is pathologic (3). Healthy liver tissue consists of structural units called lobules, which are thought to be polygonal in shape and tessellate throughout the liver. At the lobule vertices are the portal tracts, which contain portal veins, hepatic arterioles, and bile ducts among other structures, and at the lobule center is the central vein (4). Blood enters the liver through the portal tracts at the vertices of the lobule, percolates inward through the parenchymal space via the liver sinusoids, and exits through the central vein at the center of the lobule (4, 5). The lobule’s cellular organization is critical for its functional and structural zonation wherein cells are spatially divided into concentric zones that perform unique metabolic functions and/or exhibit unique structural features (6).

Liver cirrhosis—an advanced stage of liver disease marked by the increase in fibrotic scar tissue—may cause the liver’s intricately organized microvasculature to become massively rearranged and results in functional deficiencies. Cirrhosis typically originates from consistent exposure to damaging agents, such as alcohol [decompensated alcohol–associated cirrhosis (DAC)], fatty acids (metabolic dysfunction–associated steatotic liver disease), hepatitis B or C viruses, and/or drug abuse or misuse (7). In some cases, the root cause that leads to cirrhosis may be unknown, such as in primary sclerosing cholangitis (PSC), which is a rare disease that results in cirrhosis characterized by bile duct inflammation and scarring (8). The cirrhotic etiology may dictate the presentation of abnormal vascular and/or biliary architecture in a patient’s cirrhotic liver (9–12). Understanding how changes in liver architecture correlate with cirrhotic disease states and impaired liver function may provide avenues for understanding and potentially treating cirrhosis.

Despite the potential power of 3D imaging for shedding light on liver structure and pathology, most prior 3D imaging of the liver has been performed in rodent liver tissue (13–21) or in humans on the scale substantively less than a single liver lobule (22, 23). For example, the largest prior human liver 3D imaging study with cellular resolution imaged from a portal vein to a central vein along the portal-central axis (~500 μm by 100 μm by 100 μm) (23). To date, no published study has visualized nonfibrotic (i.e., normal) or cirrhotic human liver tissue across multiple lobules with cellular resolution, which is key for understanding how the 3D lobular architecture of human liver tissue remodels during cirrhosis and results in functional deficiencies.

Here, we developed an accessible LiverMap pipeline to capture cellularly resolved human liver tissue on the mesoscale (4000 μm by 4000 μm by 500 μm), enabling visualization of several adjacent human lobules in 3D with cellular resolution over a depth of 500 μm. 3D comparative analyses of nonfibrotic control and cirrhotic tissues indicated that cirrhosis correlates with the dysregulation of sinusoid expression patterns, reduction in glutamine synthetase (GS)–expressing pericentral hepatocytes, regression of central vein networks, disruption of hepatic arterial networks, and fragmentation of biliary networks. Together, these observations suggest a pro-portalization/decentralization phenotype in correlation with human liver cirrhosis. The LiverMap pipeline thus offers an accessible method to gain a more holistic, empirical view of 3D human liver architecture and its cirrhosis-induced adaptive remodeling.

RESULTS

LiverMap pipeline for 3D visualization of human liver tissue

To democratize the ability to visualize human liver tissue across several lobules with cellular resolution, we developed the LiverMap pipeline for 3D staining, clearing, reconstructing, and analyzing structures in the human liver. Toward this end, human nonfibrotic (i.e., “normal liver”) control or cirrhotic liver samples were excised from either the livers of patients undergoing oncologic partial hepatectomy or from the explanted livers of patients receiving liver transplantation, respectively (Fig. 1A). Nonfibrotic control samples were obtained from resected nondiseased human liver tissue located at least 2 cm away from any present tumor, a distance at which the tissue is typically considered free from field effect (fig. S1) (24, 25). For all human liver specimens used in this work, an adjacent slice was also sectioned and stained via traditional 2D histology methods and assigned an ISHAK score to denote the degree of fibrosis by a blinded liver pathologist (fig. S2). Next, each fixed human liver sample was punched into 6-mm cores and sliced to a thickness of 2 mm (“liver slices”), a size expected to contain several human liver lobules (Fig. 1A). Samples were then immunostained with various antibodies to identify hepatocytes, liver sinusoidal endothelial cells (LSECs), portal and central veins, hepatic arterioles, bile ducts, and/or bile canaliculi (Table 1 and fig. S3). Stained samples were optically cleared using the Ce3D protocol for lipid-rich, wholemount tissue (Fig. 1A) (26) and then tile imaged with a confocal laser scanning microscope (figs. S4 and S5). Last, commercially available software was used to computationally reconstruct immunofluorescent 3D human liver images via thresholding, segmentation, and binary interpretation for visualization and quantification (Fig. 1, B and C, and figs. S6 to S9).

Fig. 1. — (A) Human liver tissue was collected during surgical resection of nonfibrotic control liver (harvest, left) and cirrhotic liver (harvest, right). Next, biopsy punches were taken of the excised liver tissue and sliced into 2-mm-thick sections before being stained with fluorescent molecules or antibodies and optically cleared via Ce3D for 3D imaging. (B and C) After cleared tissues were imaged in 3D, immunofluorescent signals were used to computationally create object surfaces using Imaris 9.7 software (B) and vessel networks using Vesselucida software (C) for visualization and quantification. (A) Created in BioRender. Stevens, K. (2025) https://BioRender.com/oqeue4m.

Table 1. Antibodies and small molecules.

Target	Antibody/small molecule	Vendor	Dilution
Bile ducts	Anti-Cytokeratin 19 (EP1580Y)	Abcam, #AB52625; RRID: AB_2281020	1:100
GS	Anti-Glutamine Synthetase (GS-6)	Sigma-Aldrich, #MAB302 RRID: AB_2110656	1:100
CYP2E1	Anti-Cytochrome P450 2E1	Abcam, #AB28146 RRID: AB_2089985	1:100
Zone 2/3 LSECs	Anti-Human Fc gamma RII/CD32b	R&D, #AF1330 RRID: AB_354737	1:100
Smooth muscle cells	Anti-alpha Smooth Muscle Actin (1A4)	Sigma-Aldrich, #A21202 RRID: AB_476701	1:500
Endothelial cells	Anti-CD31/PECAM-1 (JC/70A)–DyLight 550	Novus, #NB600-562R	1:400
Bile canaliculi	Anti-MRP2 [EPR10997(2)]	Abcam, #AB187644	1:100
Nuclei	Hoechst 33342	Thermo Fisher Scientific, #H3570	1:400

Open in a new tab

Multiscale visualization of architectural features of human liver tissue

We leveraged the LiverMap pipeline to expand upon prior human liver tissue 3D imaging volumes to map how cellular and structural features were arranged across several length scales in 3D space. To do this, we first visualized how multiple lobules tessellate across a volume of human liver (Fig. 2, A, left, and B). We identified the center of each polygonal liver lobule, the central vein that transports processed blood out of the liver (Fig. 2A, middle), by immunostaining the hepatocytes immediately surrounding the central vein using GS (Fig. 2B) and manually tracing the central vein lumen through the 3D image stack. The polygonal vertices at the peripheral edge of each lobule were identified by immunostaining the bile ducts in portal tracts via cytokeratin 19 (CK19; Fig. 2B) (27, 28). We were able to visualize and reconstruct portions of 17 different liver lobules across a volumetric space of 4 mm by 4 mm by 500 μm (Fig. 2B). Using these 3D reconstructions, we found the median average human liver lobule polygonal radius to be 595 μm (n = 17 lobules; range, 428 to 717 μm). To validate this measurement, we independently quantified lobule radii from Masson trichrome (MT)–stained sections of distinct nonfibrotic samples (n = 6) and found an average polygonal radius of 592 ± 87 μm, which is in close agreement with the measurements from our 3D reconstructions and consistent with prior human liver imaging studies (22, 29–31).

We next sought to map the volumetric topology of mesoscale transport networks in the portal tracts and lobules. To do this, we stained, imaged, and reconstructed portal veins that bring partly deoxygenated but nutrient-rich blood to the liver, hepatic arterioles that bring oxygenated blood into the liver, and bile ducts that carry the hepatic by-product bile out of the liver to the gallbladder and the small intestine [smooth muscle actin (SMA), CD31, and CK19, respectively; Fig. 2C]. SMA also stains central veins and thus demarcated the center of lobules (Fig. 2C, SMA marked by asterisks). The resultant 3D visualizations demonstrated how bile ducts (CK19, cyan) and hepatic arterioles (CD31, orange) extend, wrap, encircle, and branch alongside large interlobular portal veins (SMA, purple) (Fig. 2C and movie S1). These reconstructions volumetrically displayed how several portal tracts together encase a liver lobule (Fig. 2C, left, lobule center denoted by asterisk-labeled central vein).

As a next step, we sought to further visualize transport networks within a given lobule (Fig. 2A, middle). We first used CD32b to visualize vascular sinusoids, which are specialized capillaries lined with highly permeable endothelial cells (i.e., LSECs) that facilitate macromolecular exchange between the blood and hepatocytes. 2D histological studies have previously shown this marker is highly expressed in Zone 2/3 human LSECs in the midlobular and pericentral regions of the lobule and lowly expressed in Zone 1 human LSECs in the portal regions of the lobule (Fig. 2A, middle) (32). To visualize the sinusoids, we computationally created binary reconstructions of immunofluorescent CD32b signal to visualize all cells expressing CD32b regardless of expression level (Fig. 2, B and D, magenta). We then isolated a single lobule via Imaris clipping planes using portal features as lobular boundaries. Variation in LSEC CD32b expression was further visualized by pseudocoloring the reconstructions based on immunofluorescent signal intensity (Fig. 2D, bottom), which showed roughly a radial gradient pattern with LSECs most strongly expressing CD32b at the central region of the lobule.

We further sought to visualize not only microscale vascular transport networks within the lobule but also the bile canalicular networks, which are 1- to 2-μm-wide interconnected spaces between adjacent hepatocytes that collect and carry by-product bile salts produced by hepatocytes to the bile ductules (33). To do this, we immunostained separate samples for multidrug resistance–associated protein (MRP2), a transporter located on the canalicular (apical) surface of hepatocytes, and CD32b (34). The resultant volumetric images of CD32b⁺ LSECs and MRP2⁺ bile canaliculi depicted the complexity of how these two micro–transport networks interweave in 3D (Fig. 2E). These surfaces also demarcated hepatocyte location as these cells have surfaces that face both the sinusoids (Fig. 2E, magenta) and bile canaliculi (Fig. 2E, green). Together, the LiverMap pipeline enabled capture and 3D visualization of cellularly resolved tissue volumes that are ~1600-fold larger than prior studies for multiscale reconstruction of 3D human liver structure.

Dysregulation of human hepatocyte and sinusoidal CD32b spatial expression in cirrhosis

Zonation of the liver lobule is essential for healthy liver function as it allows the numerous tasks of the liver to be delegated across cell subpopulations (35). Most prior investigations of liver zonation have focused primarily on hepatocytes, such as defining hepatocyte periportal functions and pericentral functions. Recent research suggests that LSECs are also spatially zonated across the lobule (32). However, these studies relied on 2D traditional methods and characterized only LSECs in normal liver tissue. The extent to which LSECs are volumetrically zonated in both normal and cirrhotic tissue remains unclear.

To address this knowledge gap, we first sought to benchmark how LSEC CD32b spatial expression correlates with hepatocyte zonation (35). To do this, we used the LiverMap pipeline to visualize the extent to which LSEC expression of CD32b correlates spatially with hepatocyte expression of zonation markers cytochrome P450 2E1 and GS (CYP2E1 and GS; Fig. 3). We found that expression levels of CD32b in LSECs and CYP2E1 in hepatocytes in nonfibrotic human liver tissue closely correlated spatially across both multilobular line profiles and 3D signal intensity plots, exhibiting gradient patterns that emanated from the center of the lobules (Fig. 3, A to C, and movie S2). Conversely, GS was confined to the innermost layers of hepatocytes surrounding the central vein, as shown previously by us and others (36), and thus overlapped spatially with only the most centrally located CD32b⁺ LSECs (Figs. 2 and 3, D and E).

Fig. 3. — (A) En face 3D MIP of nonfibrotic control tissue with CD32b (magenta) staining Zone 2/3 LSECs strongly and CYP2E1 (teal) staining metabolically active pericentral hepatocytes. Asterisks denote a large vessel running through the sample. (B) Spatial line profiles of the CD32b signal intensity (magenta) and CYP2E1 signal intensity (teal) across horizontal lines denoted in (A). (C) Angled 3D signal intensity plots showing the pixel intensity on the z axis corresponding to each pixel on the xy plane of the three regions denoted in (A). (D) Left: Nonfibrotic control sample with CD32b surfaces pseudocolored based on signal intensity. CD32b-high regions (orange) are typically in the central and midlobule zones, and CD32b-low regions (blue) are portal. Approximate lobule border drawn in white with inset showing an angled view of the isolated lobule from control tissue with a manually traced central vein. Right: Cirrhotic sample with CD32b surfaces pseudocolored based on signal intensity. CD32b-high regions are less regularly patterned than seen in the nonfibrotic control sample. (E) Top: En face MIP of entire nonfibrotic control (left) and cirrhotic (right) biopsy punches labeled for CK19⁺ bile ducts (cyan), GS⁺ pericentral hepatocytes (yellow), and CD32b⁺ Zone 2/3 LSECs (magenta). Bottom: Angled 3D signal intensity plots showing the merged image pixel intensity (z axis) corresponding to each pixel of the 2D image (xy plane) for CK19, GS, and CD32b stains. (F) Statistical comparison of the GS⁺ hepatocyte volume for nonfibrotic control (n = 6) and cirrhotic (n = 3) tissues using a two-tailed Mann-Whitney test (*P < 0.05; P value = 0.0238).

As liver microvasculature can become disrupted during cirrhosis, we next sought to visualize spatial expression patterns of CD32b⁺ LSECs across several lobules in both nonfibrotic and cirrhotic tissue. We stained six nonfibrotic and three cirrhotic liver slices to identify sinusoidal endothelial cells, bile ducts, and pericentral hepatocytes (CD32b, CK19, and GS, respectively; fig. S4) (37). We then created intensity-based pseudocolored plots of CD32b immunofluorescent signal intensity across the multilobular tissue volume (Fig. 3D and fig. S10; blue color indicates low CD32b expression, and orange color indicates high CD32b expression). Furthermore, we also generated 3D surface plots of CD32b, CK19, and GS signal intensity from maximum intensity projection (MIP) images, in which the z axis represents pixel intensity across the xy plane from all three channels (Fig. 3E and fig. S11). Figure S11 presents single-channel signal intensity surface plots side-by-side with merged image signal intensity surface plots (fig. S11). In nonfibrotic control human liver tissue, CD32b was most highly expressed at the center of each liver lobule, nearest to GS-expressing hepatocytes, and displayed roughly radial expression gradients emanating from the pericentral region of each lobule and abating at the lobule edges, demarcated by CK19⁺ bile ducts (Fig. 3, D and E, left, and figs. S10A and 11A). Each roughly radial gradient of CD32b expression associated with each given liver lobule appeared to be interspersed regularly across nonfibrotic control human liver tissue. Conversely, in the cirrhotic tissue, this regular pattern of radial CD32b expression gradients was disrupted, with some but not all cells retaining CD32b expression (Fig. 3, D and E, right, and figs. S10B and 11B). Furthermore, CD32b expression appeared to be interlaced with CK19 expression, suggesting that spatial dysregulation of zonated CD32b expression may accompany lobular vasculobiliary architectural changes in the cirrhotic liver tissue. Last, similarly to that shown by others in 2D sections (36), we found a near absence of volumetric GS expression in the cirrhotic samples (Fig. 3F and fig. S9), which may imply that CD32b dysregulation could be linked to disruption of hepatic zonation patterns and/or hepatocellular function.

In summary, the LiverMap pipeline enables visualization of volumetric functional and structural sinusoidal and hepatocyte zonation across several human liver lobules in 3D, and through leveraging this pipeline, we found that cirrhosis correlates with the dysregulation of sinusoidal CD32b spatial expression and a reduction in GS-expressing pericentral hepatocytes.

Alterations to liver vascular networks in cirrhotic tissue

Functional zonation is largely associated with the vasculobiliary structural hallmarks of the lobules, such as the central veins from which zonation patterns typically emanate. We and others have reasoned that disruption in zonation patterns, such as the reduction we observed in GS-expressing hepatocytes, may stem from disrupted central vein network architecture (10, 38). Therefore, we next set out to further map mesoscale 3D vasculobiliary network architectures in both nonfibrotic and cirrhotic human liver tissue.

To do this, we first reconstructed central vein, hepatic arterial, and biliary networks within six nonfibrotic control tissues and three cirrhotic tissues (Fig. 4, A and B). Central veins were delineated by semimanual tracing then parameterized and quantified in image processing software (fig. S12) (39, 40). Parameterization and quantification of traced central vein networks demonstrated that nonfibrotic liver tissue had a higher number of central veins, a greater total network volume, and more network endings than cirrhotic tissue (Fig. 4C). Although all cirrhotic tissue samples exhibited a reduction in central veins compared to nonfibrotic tissue, each cirrhotic sample had distinct architectural arrangements of mesoscale hepatic arterial networks (Fig. 4B). Specifically, we observed an expansion of the CD31⁺ vessels demarcating hepatic arterioles and possibly dysregulated sinusoids, which may become CD31⁺ after sinusoidal defenestration or capillarization (41, 42), in the samples from patients with DAC and PSC (Fig. 4B, top and bottom, respectively) but not in the sample from the patient with hepatitis B viral infection (Fig. 4B, middle).

Fig. 4. — (A) Top: En face 3D reconstructions of identifiable central veins (CVs; gray), CK19⁺ bile ducts (cyan), and CD31⁺ microvessels (orange) across representative nonfibrotic control human liver tissue. Bottom: Depiction of the vasculature (gray and orange) isolated from the nonfibrotic control 3D model above. (B) Left: En face 3D models of identifiable central veins (gray), CK19⁺ bile ducts (cyan), and CD31⁺ microvessels (orange) across three cirrhotic human liver tissue sections with different etiologies. Right: 3D models of identifiable central veins (gray) visually isolated from adjacent cirrhotic samples. (C) Morphometric measurements describing each sample’s central vein network in the nonfibrotic control (n = 6) and cirrhotic (n = 3) groups. Data are represented by violin plots and were statistically analyzed via a two-tailed Mann-Whitney test (*P < 0.05; n.s., not significant).

The different presentations of vascular architectures within the cirrhotic group may be attributed to samples having different cirrhotic etiologies, disease chronicity, and/or disease severity (fig. S2). Together, these volumetric findings indicated and allowed the visualization of central vein network regression and hepatic arterial network disruption in cirrhotic human liver tissue.

Architectural changes of the biliary network in cirrhosis

Bile ducts carry hepatic bile out of the liver to the gallbladder and the small intestine, which helps facilitate the excretion of hepatic metabolites and intestinal absorption of lipids and fat-soluble vitamins (43). In liver cirrhosis, biliary networks can undergo extensive remodeling and fail to transport bile out of the liver, thereby causing a buildup of bile in the liver and prompting bile leakage into the bloodstream—a condition called cholestasis (44). Our initial vasculobiliary network tracings suggested architectural dysregulation of the CK19 biliary networks from patients with cirrhosis (Fig. 4B, cyan). We thus lastly sought to further investigate bile duct architecture in both nonfibrotic control and cirrhotic human liver tissue.

To do this, we created Imaris surface renderings of CK19 immunostains within six nonfibrotic samples and three cirrhotic samples (Fig. 5 and fig. S9). In the nonfibrotic group, we observed biliary structures that extended and branched through tissue volumes, resembling network architectures (Fig. 5A and fig. S9). However, in the cirrhotic group, each cirrhotic sample presented a unique CK19 biliary network architecture, which may be a result of the cirrhotic etiology or disease severity within each sample (Fig. 5B and fig. S9). We observed ductular expansion primarily within the thick fibrotic bands in the DAC sample, discrete clusters of CK19⁺ cellular structures throughout the hepatitis B cirrhosis sample, and a combination of long bile ducts within thin fibrotic bands and bile duct segments within the bulk of the tissue in the PSC sample (Fig. 5B and figs. S2 and S9).

For the cirrhotic samples, we were particularly interested in 3D morphological analysis of a pathological bile duct phenotype called ductular reaction (DR), which resembles an expansion of small bile ductules and may be caused by the proliferation of cholangiocytes or hepatic progenitor cells or by transdifferentiation of hepatocytes (11). DR is not fully understood, but it is believed that the cell type of origin for DR may depend on the cirrhotic etiology (11, 12). Typically, 2D histology is used for the evaluation of DR; however, 2D analysis often cannot elucidate the extent to which such 3D structures are interconnected within the tissue, especially when structures are lumenized. As examples, we point to the morphological similarity of CK19⁺ features in several 5-μm-thick slices across two different human liver specimens (fig. S13). Upon 2D inspection, the two samples appeared to have similar presentation of CK19⁺ features (fig. S13, left). However, using 3D volumetric imaging, we observed that one sample had CK19⁺ bile ducts forming interconnected networks (fig. S13, top right), and the other sample had CK19⁺ features representing distinct, nonconnected smaller structures (fig. S13, bottom right). To our knowledge, DR architecture has not yet been quantifiably measured across several human liver lobules in 3D.

Toward this end, we calculated the volume of each discrete bile ductular structure in every nonfibrotic control and cirrhotic sample (figs. S8 and S9) and generated exponential histograms of CK19⁺ ductular volumes for each sample (Fig. 5C). This method provides a previously unavailable quantitative approach to describe differences in ductular fragmentation between different diseased samples, a metric that is not possible to accurately derive from 2D images (Fig. 5C and fig. S13). Here, the three samples with different cirrhotic etiologies had histograms of markedly different shapes (Fig. 5C), which reflected their distinctive ductular architectures visualized in 3D (Fig. 5B). The frequency counts for nonfibrotic or cirrhotic samples were then grouped and plotted as average frequency distributions (Fig. 5D, top). The resultant analyses showed that, overall, cirrhotic samples, on average, had both a higher quantity of the smallest (<10³ and <10⁴ μm³) CK19 structures (Fig. 5D, top) and a higher total number of CK19 structures (Fig. 5D, bottom), compared to nonfibrotic samples. Together, our method thus provided a means to quantitatively describe ductular remodeling and fragmentation in human liver cirrhosis, enabling us to quantitatively visualize the differing degrees of ductular fragmentation regarding each cirrhotic sample as well as conclude that, on average, cirrhotic samples had more fragmented ductular networks than those in the nonfibrotic human liver control samples.

DISCUSSION

In this work, we used immunostaining, tissue clearing, 3D imaging, and computational modeling to develop an accessible LiverMap pipeline (Fig. 1). The resultant 3D images had a depth equivalent to ~100 traditional tissue sections and a volume roughly 1600 times larger than the previously captured human liver volume at cellular resolution (Fig. 2 and figs. S4 and S5). These efforts unveiled never-before-seen-in-3D human liver structures across multiple size scales and visualized volumetric dysregulations of vasculobiliary architectures in cirrhosis.

Although we did find shared vasculobiliary architectural changes across all cirrhotic samples compared to nonfibrotic tissue (Figs. 4C and 5D), we also identified distinct vasculobiliary architectural features across cirrhotic samples (Figs. 4B and 5, B and C). Further studies are thus needed to parse out key differences in vasculobiliary presentation across individuals with cirrhosis. Although numerous factors could contribute to architectural variation across individuals with cirrhosis, two possibilities are the degree of cirrhosis, which was similar across our cirrhotic samples as per ISHAK scoring (fig. S2C), and cirrhotic etiology (9–12), which importantly was different for each cirrhotic sample here. Notably, the LiverMap pipeline now provides a tool to perform such future studies. It could also be deployed to shed light on how architectural changes evolve over disease progression.

Through leveraging the LiverMap pipeline, we observed a spatial dysregulation of CD32b⁺ LSECs in cirrhotic tissue in comparison to nonfibrotic control tissue (Fig. 3), which could imply a disruption in critical Zone 2/3 LSECs roles such as functions related to blood-derived leukocyte adhesion and antigen clearance in cirrhotic tissue (32). In interpreting these findings, it is important to note that altered CD32b expression patterns could reflect multiple processes, including alteration to global LSEC zonation patterns, loss or redistribution of LSEC subpopulations, shifts in endothelial cell identity, or even inflammatory remodeling. Additional endothelial markers in future studies could provide deeper insight into global LSEC zonation and its remodeling in disease. Furthermore, we found a significant reduction in GS-expressing pericentral hepatocytes in cirrhotic tissue when compared to nonfibrotic tissue (Fig. 3F). As GS plays an important enzymatic role in nitrogen metabolism by catalyzing glutamate and ammonia to form glutamine, this observation may imply decreased ammonia metabolism for patients with cirrhosis, thereby indicating a higher risk of hyperammonemia or a buildup of toxic ammonia in the blood as is commonly observed in patients with cirrhosis and liver failure (45, 46).

In addition, our 3D analyses revealed a significant decrease in the presence of central veins within cirrhotic tissue in comparison to nonfibrotic tissue (Fig. 4). Because the central vein serves as the primary transporter of blood out of the lobule, fewer central veins may result in dysfunctional blood drainage from the liver and potentially contribute to portal vein hypertension, a clinically observed condition within patients with cirrhosis (9, 10, 47). Notably, central vein identification in thick tissue is complicated by depth-related signal loss and vessel heterogeneity, which limit utility of automated approaches (39, 40). In this study, semimanual tracing with autofluorescence guidance provided the most reliable delineation (fig. S12). Emerging computational strategies such as hybrid filters, machine learning, and deep learning could increase throughput and thus make possible exciting future studies with additional samples (48–51).

Last, we provided a previously unreported quantitative framework to describe CK19 biliary network fragmentation in nonfibrotic and cirrhotic human tissues. We found that the presentation of CK19 biliary networks varied substantially between cirrhotic etiologies (Fig. 5 and fig. S9), which was reflected by each cirrhotic sample having CK19 volumetric histograms with markedly different distributions (Fig. 5C). We also found that, on average, cirrhotic samples had both a higher quantity of the smallest CK19 structures (Fig. 5D, top) and a higher total number of CK19 structures (Fig. 5D, bottom) than nonfibrotic samples. Together, these statistics indicated that the average cirrhotic bile ductular network was more fragmented than the average nonfibrotic control bile ductular network, which may imply abnormal biliary pressure and flow in cirrhotic tissue.

In summary, this work introduced the LiverMap pipeline and presented the first 3D visualization and volumetric characterization of sparse and convoluted structures across human liver tissue. Although the image sets acquired here represent a substantial increase in imaging volume compared to prior studies, the 500-μm imaging depth here remains insufficient to capture the entire depth of a human liver lobule. Future studies that extend imaging depths beyond 1 mm will be required to achieve complete human lobular reconstructions. Despite these limitations, the LiverMap pipeline nonetheless enabled the examination of nonfibrotic and cirrhotic human liver tissue to identify functional and architectural abnormalities associated with cirrhosis. 3D comparative analyses of nonfibrotic control and cirrhotic tissues indicated that cirrhosis correlates with the dysregulation of sinusoid expression patterns, reduction in GS-expressing pericentral hepatocytes, regression of central vein networks, disruption of hepatic arterial networks, and fragmentation of biliary networks. Together, these observations suggest a pro-portalization/decentralization phenotype in correlation with human liver cirrhosis. As liver structure is closely related to its function, further implementation of the LiverMap pipeline may enhance our 3D pathological understanding of architectural remodeling and functional deficiencies during cirrhosis and inspire treatments for liver disease.

METHODS

Human liver samples

Patients undergoing surgical resection at the University of Washington Medical Center consented under an IRB-approved protocol (#00001852, Liver and Abdominal Tumor Biorepository) for tissue donation and access to deidentified demographic and clinical data. Nonfibrotic control samples were obtained from partial hepatectomy (n = 6), where each sample was dissected at least 2 cm away from any present tumor to ensure the sample was free from field effect (fig. S1) (24, 25). Cirrhotic liver samples were collected from patients undergoing transplant (n = 3). Parenchymal tissue sections (~2 cm by 2 cm by 2 cm) not needed for pathologic evaluation were fixed in 10% formalin, biopsy punched into 6-mm cores (Integra Miltex), and sliced with a vibratome (Leica VT 1200S) to a thickness of 2 mm. Tissue samples were fixed for 48 to 72 hours at 4°C before being transferred to phosphate-buffered saline (PBS) at 4°C until further use. Pieces of each tissue sample were sectioned and stained with hematoxylin and eosin (H&E) and MT to evaluate tissue structure (H&E) and fibrosis content (MT) using conventional methods (fig. S2). A liver pathologist then performed a blinded ISHAK grading of the H&E and MT images for each patient sample to confirm the nonfibrotic and cirrhotic tissue groupings (fig. S2C). The histological, demographic, pathologic, and clinical details of all samples in this work are summarized in fig. S2.

LiverMap immunostaining and tissue clearing

For immunostaining, tissue samples were stained with antibodies or small molecules to identify bile ducts (CK19), pericentral hepatocytes (GS and CYP2E1), Zone 2/3 LSECs (CD32b), portal and central veins (SMA), hepatic arterioles (CD31), bile canaliculi (MRP2), and/or cell nuclei (Hoechst) (Table 1 and fig. S3). For tissue clearing, the Ce3D protocol for lipid-rich, wholemount tissue was used (26).

Briefly, fixed human liver tissue samples were washed in fresh PBS for 1 hour and then blocked in Ce3D alternative blocking buffer (1x BD Perm/Wash buffer + 1% bovine serum albumin + 1% normal donkey serum + 0.3% Triton X-100 in PBS) overnight at 37°C with gentle shaking. Next, samples were incubated in primary antibodies diluted in Ce3D alternative blocking buffer with 5% dimethyl sulfoxide (DMSO) for 24 hours at 37°C with gentle shaking (Table 1). If primary antibodies were conjugated to a fluorophore, samples were protected from light from this step on. Samples were then washed with Ce3D wash buffer (0.2% Triton X-100 + 0.5% 1-thioglycerol in PBS) for 6 to 8 hours at room temperature with gentle shaking, changing the buffer every 2 hours. Washed samples were then incubated in fluorophore-conjugated secondary antibodies diluted 1:400 in Ce3D alternative blocking buffer with 5% DMSO overnight at 37°C with gentle shaking and were protected from light from this step on. Samples were next washed in Ce3D wash buffer for 6 to 8 hours at room temperature with gentle shaking, changing the buffer every 2 hours. Last, samples were transferred to Ce3D clearing solution with Hoechst 33342 (1:400 dilution) for overnight incubation at 37°C with gentle shaking, allowing for water-clearing solution exchange within the tissue. After 12 hours, the samples were transferred to fresh Ce3D clearing solution and imaged. For long-term storage, samples were placed in multiwell plates, and plates were covered in parafilm and aluminum foil and stored at 4°C.

Technical validation of Ce3D clearing effectiveness is demonstrated qualitatively by the cleared tissue shown over graph paper in Fig. 1A, consistent with standard practices in the tissue clearing field (52). Quantitative assessment of clearing efficiency and antibody compatibility has been comprehensively established in the original Ce3D methodology papers (26, 53), which reported robust tissue transparency, preservation of morphology, and multiplex immunostaining compatibility. Our antibody panel and staining results have been provided in fig. S3, supporting the reliability of the clearing and labeling used in this study.

Confocal imaging and image processing

Human liver samples were submerged in Ce3D clearing solution within glass-bottom dishes and covered with coverslips to keep the samples flat before being imaged on a confocal laser scanning microscope (Leica TCS SP8) with a 10x 0.4–numerical aperture dry objective (Leica). Hoechst, Alexa Fluor 488, Alexa Fluor 555 or DyLight 550, and Alexa Fluor 647 were excited with 405, 488, 552, and 638 laser lines, respectively, and detected with hybrid detectors (HyD). Z-stacks were acquired through Leica Application Suite X (LAS X) software with the bottom of the stack set ~5 μm below the start of the visible signal and with the top of the stack set at the end of the visible signal, which equated to ~500 μm of signal depth in our samples. A Z-compensation was then set within LAS X to increase laser power and gain throughout the stack as necessary. To capture the 6-mm biopsy punches, tiles of 4 by 4 image stacks were stitched using the Mosaic Merge function within LAS X. Stitched confocal Z-stack images were then visualized and processed using Imaris 9.7 (Bitplane). First, Z-stack images were converted from .lif file format to .ims using Imaris File Converter 9.7 (Bitplane). Images were then processed with the following steps: (i) flip channels in the z direction for all channels; (ii) normalize layers of all channels; (iii) background subtraction (filter width: 10 μm) of all channels; (iv) normalize layers of all channels; (v) rotate channels counterclockwise for all channels about the x axis, normalize layers for all channels, and rotate channels clockwise for all channels about the x axis; and (vi) rotate channels counterclockwise for all channels about the y axis, normalize layers for all channels, and rotate channels clockwise for all channels about the y axis. All immunofluorescent images from this study after stitching and processing are shown in figs. S4 and S5.

Computational reconstructions for visualization and quantification

Imaris 9.7 (Bitplane) software was used to generate surface renderings of the immunofluorescent signal from each antibody in all samples for enhanced visualization and feature volume quantification, similarly to the other published literature (54–57). The following parameters were used for the Imaris Surface Creation Wizard: surface grain size = 4.01 μm; background subtraction—diameter of the largest sphere = 15.0 μm; thresholding for background subtraction was manually adjusted for each sample to help account for signal variability; voxel or volume filter was applied to exclude small surfaces created from background signal. Any background signal or tissue autofluorescence that was captured in the surface reconstructions was manually identified and removed. Central vein visualization: For visualization purposes, central veins were identified via positive GS staining or positive SMA staining outside of portal tracts, and central vein surfaces were created by manually tracing lumens throughout the tissue using autofluorescence (Figs. 2 and 3 and fig. S12). Lobule measurements: Lobule boundaries were identified and isolated via recognition of portal tracts and the use of Imaris clipping planes. The Imaris measurement tool was then used to measure the distances between each central vein and all its nearest surrounding portal tracts for every full or partial lobule (n = 17) identified in the representative sample (nonfibrotic #6). Each full or partial lobule’s measurements were then averaged together to get an average lobule radius for each full or partial lobule. Average lobule radii were then put in rank order, and the median lobule radius was identified and reported. To validate this measurement, lobule radii (n = 52 measurements) were independently quantified from MT-stained sections of nonfibrotic liver samples (n = 6) using the Fiji software measurement tool (58). Immunofluorescent pseudocoloring: For intensity-based pseudocoloring of the CD32b signal, the Split Touching Objects option was enabled and a custom color gradient with an intensity mean range of 3000 units was used to visualize reconstructed sinusoid CD32b-expressing segments throughout the tissue sample. These Imaris surfaces aided in CD32b immunofluorescent signal visualization (Figs. 2D and 3D). GS volumetric measurements: Each sample was processed to normalize the total image volume to 4000 μm by 4000 μm by 300 μm (fig. S8). For each sample’s normalized tissue volume, GS immunofluorescent signal was rendered into Imaris surfaces, and the total surface volume for each sample was calculated via the Imaris statistics module (Fig. 3F).

Line profile analysis and 3D surface plots

2D MIP images of 500-μm-thick z-stack images were exported from Imaris 9.7 software as .tiff files. The 2D MIP images were then opened in Fiji software (58). Line profile analysis (Fig. 3, A and B) was performed in Fiji with the following pipeline: (i) convert the stack to 8-bit, (ii) apply a 15-μm Gaussian blur to help smooth the image and reduce noise, (iii) apply the Equalize Histogram function to help normalize pixel intensities across images, and (iv) use Fiji’s Plot Profile analysis tool for equally spaced horizontal lines throughout the sample. 3D signal intensity plots (Fig. 3, C and E, and fig. S11) were also generated in Fiji by selecting a rectangular region of interest and using Fiji’s 3D Surface Plot analysis tool with the following parameters: grid size: 512; smoothing: 6.0; perspective: 0.22; lighting: 0.2; scale: 1.73; z-scale: 0.52.

Central vein morphometrics

Vesselucida (MBF Bioscience) was used for the reconstruction and morphometric analysis of central veins (Fig. 4 and fig. S12). Before performing morphometric analysis within Vesselucida software, images were resized to 25% using the Resize tool within LAS X software. This provided sufficient image resolution for analysis while reducing the file size to meet program and computing limitations. After resizing, a semimanual process was used to trace branching structures in Vesselucida, similarly to the methodology of prior works (39, 40). First, the automatic tracing module was used to identify the microstructures with diameters of ~15 μm, such as CD31⁺ microvessels and small bile ductules. Then, larger vessels or ducts were manually traced, and these were manually connected to smaller automatically traced vessels. Central veins were manually traced as these features were all too large for the automatic tracing in Vesselucida to identify reliably as automated algorithms were limited by signal loss and diameter heterogeneity. From these tracings, Vesselucida generated morphometric data describing the architecture of traced networks.

Ductular fragmentation quantification

The volume and quantity of each discrete CK19⁺ structure in every nonfibrotic control and cirrhotic sample were calculated and recorded via the Imaris statistics module (figs. S8 and S9). Exponential histograms of CK19⁺ surface volumes (bin width: log₁₀ of surface volume in cubic micrometers) for each nonfibrotic control sample and cirrhotic sample were then created (Fig. 5C). The frequency counts of the resulting exponential histograms from each sample were then grouped with their respective nonfibrotic or cirrhotic classification and plotted as average frequency distributions (Fig. 5D, top), and frequency counts of CK19⁺ surface volumes within each bin from the nonfibrotic group and cirrhotic group were statistically compared via a two-way analysis of variance (ANOVA) (Fig. 5D, top). In addition, the average total number of CK19⁺ surfaces for the nonfibrotic group and cirrhotic group were plotted and compared via an unpaired t test (Fig. 5D, bottom).

Statistics

Statistical comparisons were performed using a two-tailed Mann-Whitney test, a two-way ANOVA with a Bonferroni multiple comparisons test, or a two-tailed, unpaired t test. Sample sizes for each dataset and relevant statistics are described in figure legends. All P values < 0.05 were considered statistically significant.

Acknowledgments

We thank D. Hailey from the UW Garvey Imaging Core for assistance with imaging. We thank all patient tissue donors for contributions to science. We acknowledge that the experiments described in this text were conducted on the land of the past, present, and future Coast Salish peoples, the land that touches the shared waters of all tribes and bands within the Suquamish, Tulalip, and Muckleshoot nations; we honor the land itself and these Coast Salish tribes. We further acknowledge that papers authored by women and scholars from underrepresented racial and ethnic groups are systematically undercited and have made every attempt to reference relevant papers in a manner that is equitable in terms of gender and racial representation.

Funding:

This research was funded by the NIH NIDDK R01DK128551 (K.R.S.), the NIH Environmental Pathology and Toxicology Training Grant T32 ES007032 (W.B.F.), the NCATS Translational Research Training Program TL1 TR002318 and the NIGMS Molecular Medicine Training Grant T32 GM095421 (C.L.F.), the NSF Graduate Research Fellowship Program DGE-2140004 and the Howard Hughes Medical Institute Gilliam Fellows Program GT16560 (S.P.S.), the NIH NIAAA R01AA026302 (R.M.C.), and the Advanced Research Projects Agency for Health (ARPA-H) under award number D25AC00460-00 (K.R.S.). Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funders.

Author contributions:

Conceptualization: W.B.F., C.L.F., R.M.C., M.M.Y., and K.R.S. Data curation: W.B.F., C.L.F., D.L.G., and M.M.Y. Formal analysis: W.B.F., C.L.F., D.L.G., and R.M.C. Funding acquisition: C.L.F. and K.R.S. Investigation: W.B.F., C.L.F., S.B., and M.M.Y. Methodology: W.B.F., C.L.F., H.L.K., J.T.C.L., M.M.Y., and K.R.S. Project administration: C.L.F. and K.R.S. Resources: W.B.F., C.L.F., D.L.G., H.L.K., J.T.C.L., R.S.W.Y., and K.R.S. Software: W.B.F. and C.L.F. Supervision: J.T.C.L. and K.R.S. Validation: W.B.F., C.L.F., D.L.G., and S.B. Visualization: W.B.F., C.L.F., D.L.G., S.B., M.M.Y., and K.R.S. Writing—original draft: W.B.F., C.L.F., and M.M.Y. Writing—review and editing: W.B.F., C.L.F., D.L.G., H.L.K., S.B., J.T.C.L., R.M.C., M.M.Y., R.S.W.Y., and K.R.S.

Competing interests:

The authors declare that they have no competing interests.

Data, code, and materials availability:

All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials. Liver map source data has been deposited in Zenodo at https://zenodo.org/records/18479511. This paper did not generate new materials.

Supplementary Materials

The PDF file includes:

Figs. S1 to S13

Legends for movies S1 and S2

sciadv.adz2299_sm.pdf^{(3.6MB, pdf)}

Other Supplementary Material for this manuscript includes the following:

Movies S1 and S2

sciadv.adz2299_movies_s1_and_s2.zip^{(48.3MB, zip)}

REFERENCES

1.Bullen A., Microscopic imaging techniques for drug discovery. Nat. Rev. Drug Discov. 7, 54–67 (2008). [DOI] [PubMed] [Google Scholar]
2.Liu J. T. C., Glaser A. K., Bera K., True L. D., Reder N. P., Eliceiri K. W., Madabhushi A., Harnessing non-destructive 3D pathology. Nat. Biomed. Eng. 5, 203–218 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Hano H., Takasaki S., Three-dimensional observations on the alterations of lobular architecture in chronic hepatitis with special reference to its angioarchitecture for a better understanding of the formal pathogenesis of liver cirrhosis. Virchows Arch. 443, 655–663 (2003). [DOI] [PubMed] [Google Scholar]
4.Dingle A. M., Yap K. K., Gerrand Y.-W., Taylor C. J., Keramidaris E., Lokmic Z., Kong A. M., Peters H. L., Morrison W. A., Mitchell G. M., Characterization of isolated liver sinusoidal endothelial cells for liver bioengineering. Angiogenesis 21, 581–597 (2018). [DOI] [PubMed] [Google Scholar]
5.Lorente S., Hautefeuille M., Sanchez-Cedillo A., The liver, a functionalized vascular structure. Sci. Rep. 10, 16194 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.K. K. Yap, G. M. Mitchell, “Chapter one—Recapitulating the liver niche in vitro,” in Advances in Stem Cells and Their Niches (Elsevier, 2022), pp. 1–55. [Google Scholar]
7.Johns Hopkins Medicine, Chronic Liver Disease/Cirrhosis (2023); https://hopkinsmedicine.org/health/conditions-and-diseases/chronic-liver-disease-cirrhosis.
8.C. Bowlus, Primary Sclerosing Cholangitis (NIH–National Institute of Diabetes and Digestive and Kidney Diseases, 2022); https://niddk.nih.gov/health-information/liver-disease/primary-sclerosing-cholangitis.
9.McConnell M., Kostallari E., Ibrahim S., Iwakiri Y., The evolving role of liver sinusoidal endothelial cells in liver health and disease. Hepatology 78, 649–669 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Schuppan D., Afdhal N., Liver cirrhosis. Lancet 371, 838–851 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Sato K., Marzioni M., Meng F., Francis H., Glaser S., Alpini G., Ductular reaction in liver diseases: Pathological mechanisms and translational significances. Hepatology 69, 420–430 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.S. Khan, R. Saxena, Liver & intrahepatic bile ducts—Ductular reaction (2022); https://pathologyoutlines.com/topic/liverductularreaction.html.
13.Hoehme S., Brulport M., Bauer A., Bedawy E., Schormann W., Hermes M., Puppe V., Gebhardt R., Zellmer S., Schwarz M., Bockamp E., Timmel T., Hengstler J., Drasdo D., Prediction and validation of cell alignment along microvessels as order principle to restore tissue architecture in liver regeneration. Proc. Natl. Acad. Sci. U.S.A. 107, 10371–10376 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Hammad S., Hoehme S., Friebel A., von Recklinghausen I., Othman A., Begher-Tibbe B., Reif R., Godoy P., Johann T., Vartak A., Golka K., Bucur P., Vibert E., Marchan R., Christ B., Dooley S., Meyer C., Ilkavets I., Dahmen U., Dirsch O., Böttger J., Gebhardt R., Drasdo D., Hengstler J. G., Protocols for staining of bile canalicular and sinusoidal networks of human, mouse and pig livers, three-dimensional reconstruction and quantification of tissue microarchitecture by image processing and analysis. Arch. Toxicol. 88, 1161–1183 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Morales-Navarrete H., Segovia-Miranda F., Klukowski P., Meyer K., Nonaka H., Marsico G., Chernykh M., Kalaidzidis A., Zerial M., Kalaidzidis Y., A versatile pipeline for the multi-scale digital reconstruction and quantitative analysis of 3D tissue architecture. Elife 4, e11214 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Vartak N., Damle-Vartak A., Richter B., Dirsch O., Dahmen U., Hammad S., Hengstler J., Cholestasis-induced adaptive remodeling of interlobular bile ducts. Hepatology 63, 951–964 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.H. Morales-Navarrete, H. Nonaka, F. Segovia-Miranda, M. Zerial, Y. Kalaidzidis, “Automatic recognition and characterization of different non-parenchymal cells in liver tissue,” in Proceedings of the IEEE 13th International Symposium on Biomedical Imaging (ISBI) (IEEE, 2016), pp. 536–540. [Google Scholar]
18.Peeters G., Debbaut C., Laleman W., Monbaliu D., Vander Elst I., Detrez J., Vandecasteele T., De Schryver T., Van Hoorebeke L., Favere K., Verbeke J., Segers P., Cornillie P., De Vos W., A multilevel framework to reconstruct anatomical 3D models of the hepatic vasculature in rat livers. J. Anat. 230, 471–483 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Meyer K., Ostrenko O., Bourantas G., Morales-Navarrete H., Porat-Shliom N., Segovia-Miranda F., Nonaka H., Ghaemi A., Verbavatz J., Brusch L., Sbalzarini I., Kalaidzidis Y., Weigert R., Zerial M., A predictive 3D multi-scale model of biliary fluid dynamics in the liver lobule. Cell Syst. 4, 277–290.e9 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Peeters G., Debbaut C., Friebel A., Cornillie P., De Vos W., Favere K., Vander Elst I., Vandecasteele T., Johann T., Van Hoorebeke L., Monbaliu D., Drasdo D., Hoehme S., Laleman W., Segers P., Quantitative analysis of hepatic macro- and microvascular alterations during cirrhogenesis in the rat. J. Anat. 232, 485–496 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Morales-Navarrete H., Nonaka H., Scholich A., Segovia-Miranda F., de Back W., Meyer K., Bogorad R., Koteliansky V., Brusch L., Kalaidzidis Y., Jülicher F., Friedrich B. M., Zerial M., Liquid-crystal organization of liver tissue. Elife 8, e44860 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Segovia-Miranda F., Morales-Navarrete H., Kücken M., Moser V., Seifert S., Repnik U., Rost F., Brosch M., Hendricks A., Hinz S., Röcken C., Lütjohann D., Kalaidzidis Y., Schafmayer C., Brusch L., Hampe J., Zerial M., Three-dimensional spatially resolved geometrical and functional models of human liver tissue reveal new aspects of NAFLD progression. Nat. Med. 25, 1885–1893 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.O. Strauss, N. Björkström, “Sample preparation of optically cleared liver tissue to identify liver macrophages using 3D microscopy” in Kupffer Cells: Methods in Molecular Biology, M. Aouadi, V. Azzimato, Eds. (Humana, 2020), vol. 2164, pp. 55–63. [DOI] [PubMed] [Google Scholar]
24.Chai H., Brown R., Field effect in cancer—An update. Ann. Clin. Lab. Sci. 39, 331–337 (2009). [PubMed] [Google Scholar]
25.Hoshida Y., Villanueva A., Llovet J. M., Molecular profiling to predict hepatocellular carcinoma outcome. Expert Rev. Gastroenterol. Hepatol. 3, 101–103 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Li W., Germain R. N., Gerner M. Y., High-dimensional cell-level analysis of tissues with Ce3D multiplex volume imaging. Nat. Protoc. 14, 1708–1733 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Adori C., Daraio T., Kuiper R., Barde S., Horvathova L., Yoshitake T., Ihnatko R., Valladolid-Acebes I., Vercruysse P., Wellendorf A. M., Gramignoli R., Bozoky B., Kehr J., Theodorsson E., Cancelas J. A., Mravec B., Jorns C., Ellis E., Mulder J., Uhlén M., Bark C., Hökfelt T., Disorganization and degeneration of liver sympathetic innervations in nonalcoholic fatty liver disease revealed by 3D imaging. Sci. Adv. 7, eabg5733 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Tanaka M., Iwakiri Y., The hepatic lymphatic vascular system: Structure, function, markers, and lymphangiogenesis. Cell. Mol. Gastroenterol. Hepatol. 2, 733–749 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Hall A., Covelli C., Manuguerra R., Luong T. V., Buzzetti E., Tsochatzis E., Pinzani M., Dhillon A. P., Transaminase abnormalities and adaptations of the liver lobule manifest at specific cut-offs of steatosis. Sci. Rep. 7, 40977 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.E. Kuntz, H.-D. Kuntz, “Morphology of the liver” in Hepatology: Textbook and Atlas (Springer Medizin Verlag, 2008), pp. 16–32. [Google Scholar]
31.J. M. Crawford, C. Lui, “Liver and biliary tract” in Robbins and Cotran’s Pathologic Basis of Disease (Saunders Elsevier, ed. 8, 2010). [Google Scholar]
32.Strauss O., Phillips A., Ruggiero K. R., Bartlett A., Dunbar P. R., Immunofluorescence identifies distinct subsets of endothelial cells in the human liver. Sci. Rep. 7, 44356 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.R. McCuskey, “Chapter 1—Anatomy of the liver” in Zakim and Boyer’s Hepatology (Elsevier Inc., ed. 6, 2012), pp. 3–19. [Google Scholar]
34.Keppler D., König J., Büchler M., The canalicular multidrug resistance protein, cMRP/MRP2, a novel conjugate export pump expressed in the apical membrane of hepatocytes. Adv. Enzyme Regul. 37, 321–333 (1997). [DOI] [PubMed] [Google Scholar]
35.Gebhardt R., Metabolic zonation of the liver: Regulation and implications for liver function. Pharmacol. Ther. 53, 275–354 (1992). [DOI] [PubMed] [Google Scholar]
36.Fleming K. E., Wanless I. R., Glutamine synthetase expression in activated hepatocyte progenitor cells and loss of hepatocellular expression in congestion and cirrhosis. Liver Int. 33, 525–534 (2013). [DOI] [PubMed] [Google Scholar]
37.Mak K., Png C., The hepatic central vein: Structure, fibrosis, and role in liver biology. Anat. Rec. 303, 1747–1767 (2020). [DOI] [PubMed] [Google Scholar]
38.Ferrell L., Liver pathology: Cirrhosis, hepatitis, and primary liver tumors—Update and diagnostic problems. Mod. Pathol. 13, 679–704 (2000). [DOI] [PubMed] [Google Scholar]
39.Alese O., Wilson A., Tessier K., Loavenbruck A., Dedeker C., Lassig A., Characterization of cutaneous microvasculature using 3D imaging: A feasibility study in a cohort of head and neck surgery patients with attention to smoking status. JPRAS Open 44, 493–505 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Ventura-Antunes L., Nackenoff A., Romero-Fernandez W., Wang Y., Bosworth A. M., Prusky A., Wang E., Carvajal-Tapia C., Shostak A., Harmsen H., Mobley B., Maldonado J., Womble N., Solopova E., Snider J. C., Merryman W. D., Lippmann E. S., Schrag M., Arteriolar degeneration and stiffness in cerebral amyloid angiopathy are linked to Aβ deposition and lysyl oxidase. Alzheimers Dement. 21, e70254 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Lalor P., Lai W., Curbishley S., Shetty S., Adams D., Human hepatic sinusoidal endothelial cells can be distinguished by expression of phenotypic markers related to their specialised functions in vivo. World J. Gastroenterol. 12, 5429–5439 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Schaffner H., Popper H., Capillarization of the sinusoids. Gastroenterology 44, 339–342 (1963). [PubMed] [Google Scholar]
43.H. Pitt, A. Nakeeb, N. Espat, “Chapter 7—Bile secretion and pathophysiology of biliary tract obstruction” in Blumgart’s Surgery of the Liver, Pancreas and Biliary Tract (W. B. Saunders, ed. 5, 2012), pp. 113–122.e1. [Google Scholar]
44.Sasaki M., Ikeda H., Haga H., Manabe T., Nakanuma Y., Frequent cellular senescence in small bile ducts in primary biliary cirrhosis: A possible role in bile duct loss. J. Pathol. 205, 451–459 (2005). [DOI] [PubMed] [Google Scholar]
45.Zhou Y., Eid T., Hassel B., Danbolt N., Novel aspects of glutamine synthetase in ammonia homeostasis. Neurochem. Int. 140, 104809 (2020). [DOI] [PubMed] [Google Scholar]
46.Qvartskhava N., Lang P. A., Görg B., Pozdeev V. I., Ortiz M. P., Lang K. S., Bidmon H. J., Lang E., Leibrock C. B., Herebian D., Bode J. G., Lang F., Häussinger D., Hyperammonemia in gene-targeted mice lacking functional hepatic glutamine synthetase. Proc. Natl. Acad. Sci. U.S.A. 112, 5521–5526 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Henriksen J., Møller S., Liver cirrhosis and arterial hypertension. World J. Gastroenterol. 12, 678–685 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Song A. H., Williams M., Williamson D. F. K., Chow S. S. L., Jaume G., Gao G., Zhang A., Chen B., Baras A. S., Serafin R., Colling R., Downes M. R., Farré X., Humphrey P., Verrill C., True L. D., Parwani A. V., Liu J. T. C., Mahmood F., Analysis of 3D pathology samples using weakly supervised AI. Cell 187, 2502–2520 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Yousefi S., Liu T., Wang R., Segmentation and quantification of blood vessels for OCT-based micro-angiograms using hybrid shape/intensity compounding. Microvasc. Res. 97, 37–46 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Glaser A. K., Reder N. P., Chen Y., Yin C., Wei L., Kang S., Barner L. A., Xie W., McCarty E. F., Mao C., Halpern A. R., Stoltzfus C. R., Daniels J. S., Gerner M. Y., Nicovich P. R., Vaughan J. C., True L. D., Liu J. T. C., Multi-immersion open-top light-sheet microscope for high-throughput imaging of cleared tissues. Nat. Commun. 10, 2781 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Liu J. T., Chow S. S., Colling R., Downes M. R., Farré X., Humphrey P., Janowczyk A., Mirtti T., Verrill C., Zlobec I., True L. D., Engineering the future of 3D pathology. J. Pathol. Clin. Res. 10, e347 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Richardson D. S., Guan W., Matsumoto K., Pan C., Chung K., Ertürk A., Ueda H. R., Lichtman J. W., Tissue clearing. Nat. Rev. Methods Primers 1, 84 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Li W., Germain R. N., Gerner M. Y., Multiplex, quantitative cellular analysis in large tissue volumes with clearing-enhanced 3D microscopy (C_e3D). Proc. Natl. Acad. Sci. U.S.A. 114, E7321–E7330 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Kruzich E., Phadke R. A., Brack A., Stroumbakis D., Infante O., Cruz-Martín A., A pipeline for STED super-resolution imaging and Imaris analysis of nanoscale synapse organization in mouse cortical brain slices. STAR Protoc. 4, 102707 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Bakar Z. H. A., Bellier J.-P., Ngah W. Z. W., Yanagisawa D., Mukaisho K.-I., Tooyama I., Optimization of 3D immunofluorescence analysis and visualization using Imaris and MeshLab. Cells 12, 218 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Gautier M. K., Ginsberg S. D., A method for quantification of vesicular compartments within cells using 3D reconstructed confocal z-stacks: Comparison of ImageJ and Imaris to count early endosomes within basal forebrain cholinergic neurons. J. Neurosci. Methods 350, 109038 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Zhu J., Liu X., Liu Z., Deng Y., Xu J., Liu K., Zhang R., Meng X., Fei P., Yu T., Zhu D., SOLID: Minimizing tissue distortion for brain-wide profiling of diverse architectures. Nat. Commun. 15, 8303 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Schindelin J., Arganda-Carreras I., Frise E., Kaynig V., Longair M., Pietzsch T., Preibisch S., Rueden C., Saalfeld S., Schmid B., Tinevez J.-Y., White D. J., Hartenstein V., Eliceiri K., Tomancak P., Cardona A., Fiji: An open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figs. S1 to S13

Legends for movies S1 and S2

sciadv.adz2299_sm.pdf^{(3.6MB, pdf)}

Movies S1 and S2

sciadv.adz2299_movies_s1_and_s2.zip^{(48.3MB, zip)}

Data Availability Statement

[R1] 1.Bullen A., Microscopic imaging techniques for drug discovery. Nat. Rev. Drug Discov. 7, 54–67 (2008). [DOI] [PubMed] [Google Scholar]

[R2] 2.Liu J. T. C., Glaser A. K., Bera K., True L. D., Reder N. P., Eliceiri K. W., Madabhushi A., Harnessing non-destructive 3D pathology. Nat. Biomed. Eng. 5, 203–218 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] 3.Hano H., Takasaki S., Three-dimensional observations on the alterations of lobular architecture in chronic hepatitis with special reference to its angioarchitecture for a better understanding of the formal pathogenesis of liver cirrhosis. Virchows Arch. 443, 655–663 (2003). [DOI] [PubMed] [Google Scholar]

[R4] 4.Dingle A. M., Yap K. K., Gerrand Y.-W., Taylor C. J., Keramidaris E., Lokmic Z., Kong A. M., Peters H. L., Morrison W. A., Mitchell G. M., Characterization of isolated liver sinusoidal endothelial cells for liver bioengineering. Angiogenesis 21, 581–597 (2018). [DOI] [PubMed] [Google Scholar]

[R5] 5.Lorente S., Hautefeuille M., Sanchez-Cedillo A., The liver, a functionalized vascular structure. Sci. Rep. 10, 16194 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.K. K. Yap, G. M. Mitchell, “Chapter one—Recapitulating the liver niche in vitro,” in Advances in Stem Cells and Their Niches (Elsevier, 2022), pp. 1–55. [Google Scholar]

[R7] 7.Johns Hopkins Medicine, Chronic Liver Disease/Cirrhosis (2023); https://hopkinsmedicine.org/health/conditions-and-diseases/chronic-liver-disease-cirrhosis.

[R8] 8.C. Bowlus, Primary Sclerosing Cholangitis (NIH–National Institute of Diabetes and Digestive and Kidney Diseases, 2022); https://niddk.nih.gov/health-information/liver-disease/primary-sclerosing-cholangitis.

[R9] 9.McConnell M., Kostallari E., Ibrahim S., Iwakiri Y., The evolving role of liver sinusoidal endothelial cells in liver health and disease. Hepatology 78, 649–669 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Schuppan D., Afdhal N., Liver cirrhosis. Lancet 371, 838–851 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] 11.Sato K., Marzioni M., Meng F., Francis H., Glaser S., Alpini G., Ductular reaction in liver diseases: Pathological mechanisms and translational significances. Hepatology 69, 420–430 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] 12.S. Khan, R. Saxena, Liver & intrahepatic bile ducts—Ductular reaction (2022); https://pathologyoutlines.com/topic/liverductularreaction.html.

[R13] 13.Hoehme S., Brulport M., Bauer A., Bedawy E., Schormann W., Hermes M., Puppe V., Gebhardt R., Zellmer S., Schwarz M., Bockamp E., Timmel T., Hengstler J., Drasdo D., Prediction and validation of cell alignment along microvessels as order principle to restore tissue architecture in liver regeneration. Proc. Natl. Acad. Sci. U.S.A. 107, 10371–10376 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Hammad S., Hoehme S., Friebel A., von Recklinghausen I., Othman A., Begher-Tibbe B., Reif R., Godoy P., Johann T., Vartak A., Golka K., Bucur P., Vibert E., Marchan R., Christ B., Dooley S., Meyer C., Ilkavets I., Dahmen U., Dirsch O., Böttger J., Gebhardt R., Drasdo D., Hengstler J. G., Protocols for staining of bile canalicular and sinusoidal networks of human, mouse and pig livers, three-dimensional reconstruction and quantification of tissue microarchitecture by image processing and analysis. Arch. Toxicol. 88, 1161–1183 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Morales-Navarrete H., Segovia-Miranda F., Klukowski P., Meyer K., Nonaka H., Marsico G., Chernykh M., Kalaidzidis A., Zerial M., Kalaidzidis Y., A versatile pipeline for the multi-scale digital reconstruction and quantitative analysis of 3D tissue architecture. Elife 4, e11214 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Vartak N., Damle-Vartak A., Richter B., Dirsch O., Dahmen U., Hammad S., Hengstler J., Cholestasis-induced adaptive remodeling of interlobular bile ducts. Hepatology 63, 951–964 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.H. Morales-Navarrete, H. Nonaka, F. Segovia-Miranda, M. Zerial, Y. Kalaidzidis, “Automatic recognition and characterization of different non-parenchymal cells in liver tissue,” in Proceedings of the IEEE 13th International Symposium on Biomedical Imaging (ISBI) (IEEE, 2016), pp. 536–540. [Google Scholar]

[R18] 18.Peeters G., Debbaut C., Laleman W., Monbaliu D., Vander Elst I., Detrez J., Vandecasteele T., De Schryver T., Van Hoorebeke L., Favere K., Verbeke J., Segers P., Cornillie P., De Vos W., A multilevel framework to reconstruct anatomical 3D models of the hepatic vasculature in rat livers. J. Anat. 230, 471–483 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Meyer K., Ostrenko O., Bourantas G., Morales-Navarrete H., Porat-Shliom N., Segovia-Miranda F., Nonaka H., Ghaemi A., Verbavatz J., Brusch L., Sbalzarini I., Kalaidzidis Y., Weigert R., Zerial M., A predictive 3D multi-scale model of biliary fluid dynamics in the liver lobule. Cell Syst. 4, 277–290.e9 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Peeters G., Debbaut C., Friebel A., Cornillie P., De Vos W., Favere K., Vander Elst I., Vandecasteele T., Johann T., Van Hoorebeke L., Monbaliu D., Drasdo D., Hoehme S., Laleman W., Segers P., Quantitative analysis of hepatic macro- and microvascular alterations during cirrhogenesis in the rat. J. Anat. 232, 485–496 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Morales-Navarrete H., Nonaka H., Scholich A., Segovia-Miranda F., de Back W., Meyer K., Bogorad R., Koteliansky V., Brusch L., Kalaidzidis Y., Jülicher F., Friedrich B. M., Zerial M., Liquid-crystal organization of liver tissue. Elife 8, e44860 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] 22.Segovia-Miranda F., Morales-Navarrete H., Kücken M., Moser V., Seifert S., Repnik U., Rost F., Brosch M., Hendricks A., Hinz S., Röcken C., Lütjohann D., Kalaidzidis Y., Schafmayer C., Brusch L., Hampe J., Zerial M., Three-dimensional spatially resolved geometrical and functional models of human liver tissue reveal new aspects of NAFLD progression. Nat. Med. 25, 1885–1893 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.O. Strauss, N. Björkström, “Sample preparation of optically cleared liver tissue to identify liver macrophages using 3D microscopy” in Kupffer Cells: Methods in Molecular Biology, M. Aouadi, V. Azzimato, Eds. (Humana, 2020), vol. 2164, pp. 55–63. [DOI] [PubMed] [Google Scholar]

[R24] 24.Chai H., Brown R., Field effect in cancer—An update. Ann. Clin. Lab. Sci. 39, 331–337 (2009). [PubMed] [Google Scholar]

[R25] 25.Hoshida Y., Villanueva A., Llovet J. M., Molecular profiling to predict hepatocellular carcinoma outcome. Expert Rev. Gastroenterol. Hepatol. 3, 101–103 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Li W., Germain R. N., Gerner M. Y., High-dimensional cell-level analysis of tissues with Ce3D multiplex volume imaging. Nat. Protoc. 14, 1708–1733 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Adori C., Daraio T., Kuiper R., Barde S., Horvathova L., Yoshitake T., Ihnatko R., Valladolid-Acebes I., Vercruysse P., Wellendorf A. M., Gramignoli R., Bozoky B., Kehr J., Theodorsson E., Cancelas J. A., Mravec B., Jorns C., Ellis E., Mulder J., Uhlén M., Bark C., Hökfelt T., Disorganization and degeneration of liver sympathetic innervations in nonalcoholic fatty liver disease revealed by 3D imaging. Sci. Adv. 7, eabg5733 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Tanaka M., Iwakiri Y., The hepatic lymphatic vascular system: Structure, function, markers, and lymphangiogenesis. Cell. Mol. Gastroenterol. Hepatol. 2, 733–749 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Hall A., Covelli C., Manuguerra R., Luong T. V., Buzzetti E., Tsochatzis E., Pinzani M., Dhillon A. P., Transaminase abnormalities and adaptations of the liver lobule manifest at specific cut-offs of steatosis. Sci. Rep. 7, 40977 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.E. Kuntz, H.-D. Kuntz, “Morphology of the liver” in Hepatology: Textbook and Atlas (Springer Medizin Verlag, 2008), pp. 16–32. [Google Scholar]

[R31] 31.J. M. Crawford, C. Lui, “Liver and biliary tract” in Robbins and Cotran’s Pathologic Basis of Disease (Saunders Elsevier, ed. 8, 2010). [Google Scholar]

[R32] 32.Strauss O., Phillips A., Ruggiero K. R., Bartlett A., Dunbar P. R., Immunofluorescence identifies distinct subsets of endothelial cells in the human liver. Sci. Rep. 7, 44356 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.R. McCuskey, “Chapter 1—Anatomy of the liver” in Zakim and Boyer’s Hepatology (Elsevier Inc., ed. 6, 2012), pp. 3–19. [Google Scholar]

[R34] 34.Keppler D., König J., Büchler M., The canalicular multidrug resistance protein, cMRP/MRP2, a novel conjugate export pump expressed in the apical membrane of hepatocytes. Adv. Enzyme Regul. 37, 321–333 (1997). [DOI] [PubMed] [Google Scholar]

[R35] 35.Gebhardt R., Metabolic zonation of the liver: Regulation and implications for liver function. Pharmacol. Ther. 53, 275–354 (1992). [DOI] [PubMed] [Google Scholar]

[R36] 36.Fleming K. E., Wanless I. R., Glutamine synthetase expression in activated hepatocyte progenitor cells and loss of hepatocellular expression in congestion and cirrhosis. Liver Int. 33, 525–534 (2013). [DOI] [PubMed] [Google Scholar]

[R37] 37.Mak K., Png C., The hepatic central vein: Structure, fibrosis, and role in liver biology. Anat. Rec. 303, 1747–1767 (2020). [DOI] [PubMed] [Google Scholar]

[R38] 38.Ferrell L., Liver pathology: Cirrhosis, hepatitis, and primary liver tumors—Update and diagnostic problems. Mod. Pathol. 13, 679–704 (2000). [DOI] [PubMed] [Google Scholar]

[R39] 39.Alese O., Wilson A., Tessier K., Loavenbruck A., Dedeker C., Lassig A., Characterization of cutaneous microvasculature using 3D imaging: A feasibility study in a cohort of head and neck surgery patients with attention to smoking status. JPRAS Open 44, 493–505 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] 40.Ventura-Antunes L., Nackenoff A., Romero-Fernandez W., Wang Y., Bosworth A. M., Prusky A., Wang E., Carvajal-Tapia C., Shostak A., Harmsen H., Mobley B., Maldonado J., Womble N., Solopova E., Snider J. C., Merryman W. D., Lippmann E. S., Schrag M., Arteriolar degeneration and stiffness in cerebral amyloid angiopathy are linked to Aβ deposition and lysyl oxidase. Alzheimers Dement. 21, e70254 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Lalor P., Lai W., Curbishley S., Shetty S., Adams D., Human hepatic sinusoidal endothelial cells can be distinguished by expression of phenotypic markers related to their specialised functions in vivo. World J. Gastroenterol. 12, 5429–5439 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Schaffner H., Popper H., Capillarization of the sinusoids. Gastroenterology 44, 339–342 (1963). [PubMed] [Google Scholar]

[R43] 43.H. Pitt, A. Nakeeb, N. Espat, “Chapter 7—Bile secretion and pathophysiology of biliary tract obstruction” in Blumgart’s Surgery of the Liver, Pancreas and Biliary Tract (W. B. Saunders, ed. 5, 2012), pp. 113–122.e1. [Google Scholar]

[R44] 44.Sasaki M., Ikeda H., Haga H., Manabe T., Nakanuma Y., Frequent cellular senescence in small bile ducts in primary biliary cirrhosis: A possible role in bile duct loss. J. Pathol. 205, 451–459 (2005). [DOI] [PubMed] [Google Scholar]

[R45] 45.Zhou Y., Eid T., Hassel B., Danbolt N., Novel aspects of glutamine synthetase in ammonia homeostasis. Neurochem. Int. 140, 104809 (2020). [DOI] [PubMed] [Google Scholar]

[R46] 46.Qvartskhava N., Lang P. A., Görg B., Pozdeev V. I., Ortiz M. P., Lang K. S., Bidmon H. J., Lang E., Leibrock C. B., Herebian D., Bode J. G., Lang F., Häussinger D., Hyperammonemia in gene-targeted mice lacking functional hepatic glutamine synthetase. Proc. Natl. Acad. Sci. U.S.A. 112, 5521–5526 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Henriksen J., Møller S., Liver cirrhosis and arterial hypertension. World J. Gastroenterol. 12, 678–685 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Song A. H., Williams M., Williamson D. F. K., Chow S. S. L., Jaume G., Gao G., Zhang A., Chen B., Baras A. S., Serafin R., Colling R., Downes M. R., Farré X., Humphrey P., Verrill C., True L. D., Parwani A. V., Liu J. T. C., Mahmood F., Analysis of 3D pathology samples using weakly supervised AI. Cell 187, 2502–2520 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Yousefi S., Liu T., Wang R., Segmentation and quantification of blood vessels for OCT-based micro-angiograms using hybrid shape/intensity compounding. Microvasc. Res. 97, 37–46 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.Glaser A. K., Reder N. P., Chen Y., Yin C., Wei L., Kang S., Barner L. A., Xie W., McCarty E. F., Mao C., Halpern A. R., Stoltzfus C. R., Daniels J. S., Gerner M. Y., Nicovich P. R., Vaughan J. C., True L. D., Liu J. T. C., Multi-immersion open-top light-sheet microscope for high-throughput imaging of cleared tissues. Nat. Commun. 10, 2781 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] 51.Liu J. T., Chow S. S., Colling R., Downes M. R., Farré X., Humphrey P., Janowczyk A., Mirtti T., Verrill C., Zlobec I., True L. D., Engineering the future of 3D pathology. J. Pathol. Clin. Res. 10, e347 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] 52.Richardson D. S., Guan W., Matsumoto K., Pan C., Chung K., Ertürk A., Ueda H. R., Lichtman J. W., Tissue clearing. Nat. Rev. Methods Primers 1, 84 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] 53.Li W., Germain R. N., Gerner M. Y., Multiplex, quantitative cellular analysis in large tissue volumes with clearing-enhanced 3D microscopy (C_e3D). Proc. Natl. Acad. Sci. U.S.A. 114, E7321–E7330 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R54] 54.Kruzich E., Phadke R. A., Brack A., Stroumbakis D., Infante O., Cruz-Martín A., A pipeline for STED super-resolution imaging and Imaris analysis of nanoscale synapse organization in mouse cortical brain slices. STAR Protoc. 4, 102707 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] 55.Bakar Z. H. A., Bellier J.-P., Ngah W. Z. W., Yanagisawa D., Mukaisho K.-I., Tooyama I., Optimization of 3D immunofluorescence analysis and visualization using Imaris and MeshLab. Cells 12, 218 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R56] 56.Gautier M. K., Ginsberg S. D., A method for quantification of vesicular compartments within cells using 3D reconstructed confocal z-stacks: Comparison of ImageJ and Imaris to count early endosomes within basal forebrain cholinergic neurons. J. Neurosci. Methods 350, 109038 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R57] 57.Zhu J., Liu X., Liu Z., Deng Y., Xu J., Liu K., Zhang R., Meng X., Fei P., Yu T., Zhu D., SOLID: Minimizing tissue distortion for brain-wide profiling of diverse architectures. Nat. Commun. 15, 8303 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R58] 58.Schindelin J., Arganda-Carreras I., Frise E., Kaynig V., Longair M., Pietzsch T., Preibisch S., Rueden C., Saalfeld S., Schmid B., Tinevez J.-Y., White D. J., Hartenstein V., Eliceiri K., Tomancak P., Cardona A., Fiji: An open-source platform for biological-image analysis. Nat. Methods 9, 676–682 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

3D reconstruction of human liver tissue at cellular resolution

Wesley B Fabyan

Chelsea L Fortin

Dorice L Goune

Heidi L Kenerson

Susana P Simmonds Bohorquez

Jonathan T C Liu

Matthew M Yeh

Rotonya M Carr

Raymond S W Yeung

Kelly R Stevens

Roles

Abstract

INTRODUCTION

RESULTS

LiverMap pipeline for 3D visualization of human liver tissue

Fig. 1. Schematic of the LiverMap pipeline.

Table 1. Antibodies and small molecules.

Multiscale visualization of architectural features of human liver tissue

Fig. 2. Multilobular reconstructions of zonated and architectural features.

Dysregulation of human hepatocyte and sinusoidal CD32b spatial expression in cirrhosis

Fig. 3. CD32b+ LSECs associate with CYP2E1+ hepatocytes, and 3D modeling reveals dysregulation of human sinusoid zonation in cirrhosis.

Alterations to liver vascular networks in cirrhotic tissue

Fig. 4. 3D reconstructions reveal a decentralization of blood vessels in cirrhosis.

Architectural changes of the biliary network in cirrhosis

Fig. 5. Bile duct networks consist of more architectural fragmentation in cirrhosis.

DISCUSSION

METHODS

Human liver samples

LiverMap immunostaining and tissue clearing

Confocal imaging and image processing

Computational reconstructions for visualization and quantification

Line profile analysis and 3D surface plots

Central vein morphometrics

Ductular fragmentation quantification

Statistics

Acknowledgments

Funding:

Author contributions:

Competing interests:

Data, code, and materials availability:

Supplementary Materials

The PDF file includes:

Other Supplementary Material for this manuscript includes the following:

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases