H2O2-Enhanced CUBIC Clearing Coupled with Multicolor Nanoprobes for High-Resolution Mapping of Liver Vasculature.

(A) Comparison of processing times for four different liver clearing protocols: conventional CUBIC (25% urea, red), high-concentration urea (40% urea, yellow), oxidation treatment alone (25% urea + 4.5% H2O2, blue), and the optimized Liver-CUBIC (40% urea + 4.5% H2O2, green). The optimized Liver-CUBIC protocol significantly reduced clearing time (n = 6 per group). (B) Schematic illustration of mouse liver lobes, defining left lateral lobe, left medial lobe, right medial lobe, right lateral lobe, caudate lobe, and quadrate lobe according to reference [19] . (C) Brightfield images of whole mouse livers (8–9 weeks old, male) after processing with the four clearing protocols. Grid size: 1.6 mm × 1.6 mm. Scale bar: 8 mm. (D) Quantitative analysis of tissue volume changes following each clearing protocol (n = 6 per group). No statistically significant difference was observed between the 25% urea group and the 40% urea + H2O2 group. Statistical test: unpaired two-tailed t-test. PT: pre-treatment. (E) Transmission spectra (400–900 nm) of 1 mm-thick mouse liver samples after each clearing protocol (n = 5 per group). (F) Schematic diagram of the four-channel ductal-vascular labeling strategy: yellow MCNPs injected via the left ventricle to label hepatic arteries, green MCNPs via retrograde common bile duct injection to label biliary ducts, pink MCNPs via the portal vein trunk to label portal veins, and black MCNPs via the inferior vena cava to label hepatic veins. (G) Three-dimensional fine structures of the biliary tree in the central and peripheral regions of the mouse liver, labeled with MCNP-Green. Images are shown at magnifications of 100×, 200×, and 400×. The rightmost panel presents a high-magnification view of the area outlined by the blue box. Scale bars: 400 μm, 200 μm, 100 μm, and 20 μm. The arrow indicates the terminal ductal structures with a polygonal shape. (H) Three-dimensional reconstruction of the portal venous system labeled with MCNP-Pink, showing branching features in central and peripheral zones. Images captured at 100×, 200×, and 400× magnifications; the rightmost image is a higher-magnification view of the blue-boxed region. Scale bars: 400 μm, 200 μm, 100 μm, and 20 μm. (I) Three-dimensional structural details of the central vein labeled with MCNP-Black. Images captured at 100×, 300×, and 400× magnifications; the rightmost image is a higher-magnification view of the blue-boxed region. Scale bars: 400 μm, 200 μm, 100 μm, and 20 μm. (J) Three-dimensional structural details of the hepatic artery labeled with MCNP-Pink. Images captured at 100×, 200×, and 400× magnifications; the rightmost image is a higher-magnification view of the blue-boxed region. Scale bars: 400 μm, 200 μm, 100 μm, and 20 μm.

The Periportal Lamellar Complex (PLC) Serves as a Low-Permeability Gateway Bridging Portal Veins to Hepatic Lobules

(A) Scanning electron microscopy revealed that the Metal Compound Nanoparticles (MCNP) dyes consisted of aggregates of particles approximately 100 nm in length. (B) Dual-channel vascular typing labeling scheme (Dual-channel NanoFluor™). hepatic arteries labeled with pink fluorescent nanoparticlesportal veins labeled with green fluorescent nanoparticles. (C) Extended depth-of-field imaging (tissue thickness: 200 μm) reveals fine three-dimensional structural details of the hepatic artery and portal vein. The right panel shows magnified lateral and top views of the orange-boxed region, highlighting dense terminal branches and the close spatial proximity between the two vascular systems. Green arrows indicate portal veins, and pink arrows indicate hepatic arteries. (D) Magnified image showing periodic alignment of PLC structures (orange arrows) along the adventitial layer of the portal vein trunk (green) in the direction of vascular flow. Orange dashed lines delineate the boundaries of a classical hepatic lobule. Scale bar: 200 μm. (E) Three-dimensional confocal imaging of hepatic arteries (pink dye, excitation/emission: 561/640 nm) and portal veins (green dye, excitation/emission: 495/519 nm). The right panel presents a high-resolution view with hepatic arteries rendered in blue and the portal vein surface filled in white; arrows indicate points of interaction between PLC structures and hepatic arteries. Scale bar: 200 μm. (F) High-magnification confocal imaging further depicting micro-branches of portal veins and hepatic arteries, with terminal branches intertwining in a coiled distribution. Scale bar: 20 μm. (G, H) Single-channel confocal images showing the distribution of MCNP-Green-labeled portal veins (G) and MCNP- Pink-labeled hepatic arteries (H) within PLC regions. White arrows indicate the paths of fluorescence intensity profile measurements, with arrowheads denoting the direction of line scans. Scale bar: 50 μm. (I) Schematic illustration of fluorescence intensity profile measurements across PLC structures. The midpoint of the portal vein (green) intensity profile corresponds to the junction between the PLC and the outer wall of the portal vein, while the midpoint of the hepatic artery (pink) intensity profile aligns with the terminal edge of the PLC adjacent to liver sinusoids. Arrows indicate scan directions. (G) Fluorescence intensity profile plots. The X-axis represents the scan distance (0–400 μm), and the Y-axis represents fluorescence intensity. Both portal vein (green) and hepatic artery (pink) signals showed significant increases within the PLC region. The regions were defined as follows: 0–100 μm, portal vein region; 100–250 μm, PLC region; 250–400 μm, liver sinusoid region (n = 5 per group). (K) Analytical workflow for characterizing PLC structures: the primary portal vein trunk carrying PLC structures was selected as the reference axis for hepatic lobule boundaries, combined with its two adjacent secondary branches to define a classical hexagonal lobule computational unit. The diameter of the primary portal vein trunk (dPV) and the area of PLC structures were quantified using the extended depth-of-field imaging system. (L) Distribution of PLC areas along primary portal vein trunks with diameters ranging from 63.45 to 321.42 μm. Each value represents the PLC area associated with a portal vein of corresponding diameter (n = 19).

Spatial Juxtaposition of the Periportal Lamellar Complex with Canals of Hering at the Portal Venous Interface

(A) MCNP-Pink labeling of portal veins combined with three-dimensional DAB immunohistochemistry for CK19 (brown) to visualize bile duct epithelial cells. Top and lateral views highlight the PLC–bile duct interaction sites (cyan arrows). Scale bar: 50 μm. (B) MCNP-Green labeling of bile ducts combined with three-dimensional DAB immunohistochemistry for CK19 (brown). Scale bar: 40 μm. (C) MCNP-Green labeling of bile ducts combined with three-dimensional TSA immunofluorescence for CK19 (red), displaying detailed structures at the interface between dye-labeled ducts and immunostained bile duct terminals. The arrows indicate the terminal structures of the bile ducts. Scale bar: 10 μm. (D) Three-dimensional TSA multiplex immunofluorescence staining for ZO-1 (red, marking bile canaliculi networks), HNF4α (gray, marking hepatocyte nuclei), and CK19 (green, marking bile ducts). Scale bar: 20 μm. The right panel illustrates the spatial relationship between ZO-1-labeled bile canaliculi and CK19-labeled bile duct terminals, with arrows indicating the terminal positions of bile ducts. (E) Extended depth-of-field imaging of the whole liver showing high-pressure perfusion of green fluorescent nanoparticles into the bile duct, yellow nanoparticles labeling the portal vein. Red circles indicate sites of green dye leakage localized to the PLC regions. (F) High-pressure retrograde perfusion of red fluorescent nanoparticles into the bile duct, combined with three-dimensional TSA immunofluorescence for CK19 (green). Arrows indicate sites of dye leakage at the PLC region. Scale bar: 50 μm. (G) Schematic diagram illustrating the leakage sites of bile duct-perfused dye following high-pressure injection. Green dashed boxes represent the positions of free bile duct terminal epithelial cells at leakage sites.

Single-Cell Transcriptomics Identifies CD34⁺Sca-1⁺ as a Novel Endothelial Signature of the Periportal Lamellar Complex with Hematopoietic Niche Potential

(A) UMAP projection of single-cell transcriptomes of liver endothelial cells from normal adult mice reveals 10 distinct cellular clusters. Each dot represents one cell. (B) Heatmap showing expression profiles of representative genes across the 10 clusters, including hepatic stellate cells, T cells, macrophages, hepatocytes, cholangiocytes, prtal vein ECs, periportal LSECs, midzonal LSECs, pericentral LSECs, and central vein ECs . Marker genes were selected based on average cluster expression. (C) Representative cluster-specific markers for five major liver endothelial subpopulations. (D) Spatial schematic illustrating the anatomical position of the Periportal Lamellar Complex (PLC), located exclusively between the portal vein ECs and periportal LSECs. (E) Venn diagram showing overlapping top 20 highly expressed genes between prtal vein ECs and periportal LSECs. (F) Multiplex immunofluorescence images showing CD31 (yellow), Sca-1 (green), and CD34 (red) expression in distinct hepatic endothelial zones, including prtal vein ECs, midzonal LSECs, and central vein ECs. The PLC structure is demarcated by dashed white arrows. Scale bar: 40 μm. (G) Gating strategy and classification of CD34⁺Sca-1⁺ double-positive versus double-negative endothelial subpopulations. (H) Volcano plot illustrating differentially expressed hematopoietic-associated genes in CD34⁺Sca-1⁺ cells. The x-axis shows log₂ fold-change (log₂FC); the y-axis shows −log₁₀ adjusted P value. Significantly upregulated genes are located in the upper right quadrant. (I) Gene Ontology enrichment analysis of genes upregulated in CD34⁺Sca-1⁺ ECs, highlighting functional categories such as stem cell development, differentiation, and somatic stem cell maintenance. (J) The schematic illustration depicts the spatial localization of Sca-1, a marker for mesenchymal or hematopoietic stem cells, and CD34⁺Sca-1⁺CD31⁺ endothelial cells as the trunk of the Periportal Lamellar Complex (PLC).

CD34⁺Sca-1⁺ Endothelium in the Periportal Lamellar Complex is potential Regulator of Spatial Patterning of Intrahepatic Bile Duct Branching during cirrhosis.

(A) Volcano plot of differentially expressed bile duct-related genes in CD34⁺Sca-1⁺ double-positive cells. The x-axis represents log2 fold change (log2FC), and the y-axis represents -log10 adjusted P value (-log10(P-adjust)). Genes significantly upregulated are located in the upper right quadrant. (B) Functional enrichment analysis of upregulated genes in CD34⁺Sca-1⁺ cells, showing enriched categories such as epithelial morphogenesis and branching morphogenesis of epithelial tubes, represented by -log10(P value). (C) Visualization of portal vein labeled with MCNP-Pink and bile duct epithelial cells stained for CK19 (brown) by 3D DAB immunohistochemistry in control and CCl₄ 6-week fibrotic mice. Scale bar: 200 μm. (D) Quantification of the distance that bile duct termini extend from portal vein surfaces into hepatic parenchyma in control and CCl₄ 6-week mice, presented as mean ± SD (control n=20, CCl₄ 6-week n=18). (E) Multiplex immunofluorescence showing expression and spatial distribution of CD31 (yellow), Sca-1 (green), and CK19 (red) in control and fibrotic models (CCl₄ 3-week and 6-week). White arrows in CK19 single-channel magnified images indicate bile duct termini. Scale bar: 50 μm. (F) Quantitative measurement of bile duct termini extension distances along PLC structures into hepatic parenchyma in control, early (CCl₄ 3-week), and late (CCl₄ 6-week) fibrosis mice. Data represent mean ± SD (n=5 per group). Statistical significance determined by one-way ANOVA with Tukey’s multiple comparisons test; *P < 0.05, **P < 0.01, ****P < 0.0001. (G) Volcano plot of differentially expressed bile duct-related genes in CD34⁺Sca-1⁺ cells from fibrotic livers compared to controls. Axes as in (A). Significantly upregulated genes are in the upper right quadrant. (H) Schematic illustration showing spatial localization of CK19 and CD31 within the PLC structure.

CD34⁺Sca-1⁺ Endothelium in the Periportal Lamellar Complex is a Potential Neurovascular Niche Regulating Hepatic Autonomic Nerve Patterning in Cirrhosis

(A) Differential expression analysis of nerve-related genes in CD34⁺Sca-1⁺ endothelial cells. The x-axis indicates log2 fold change (log2FC), and the y-axis represents –log10 adjusted P value (–log10(P-adjust)). Significantly upregulated genes are located in the upper right quadrant. (B) Gene ontology (GO) enrichment analysis of upregulated genes in CD34⁺Sca-1⁺ cells, showing enrichment in functional categories such as semaphorin-plexin mediated axon guidance, regulation of neuronal projection regeneration, and modulation of postsynaptic neurotransmitter receptor endocytosis. Enrichment significance is indicated by –log10(P-value). (C) Multiplex immunofluorescence staining showing tyrosine hydroxylase (TH, green) labeling sympathetic nerves and CD31 (yellow) labeling portal vein endothelium. Scale bar: 20 μm. (D) Distribution of CD31 (yellow) and TH (green) expression in control and CCl₄-induced liver fibrosis models at week 3 and week 6, visualized by multiplex immunofluorescence. In the green channel images, the white arrows marks the terminal location of TH-positive sympathetic nerve endings. Scale bar: 50 μm. (E) Quantification of the distance from sympathetic nerve endings to the hepatic parenchyma along PLC structures in control, early fibrosis (CCl₄-3 weeks), and advanced fibrosis (CCl₄-6 weeks) mice. Data are presented as mean ± standard deviation (Mean ± SD), n=5 per group. Statistical analysis was performed using one-way ANOVA followed by Tukey’s multiple comparison test; *P < 0.05, **P < 0.01, ****P < 0.0001. (F) Differential expression analysis of nerve-related genes between CD34⁺Sca-1⁺ cells from cirrhotic and control livers. The x-axis indicates log2 fold change (log2FC), and the y-axis represents –log10 adjusted P value (–log10(P- adjust)). Significantly upregulated genes are located in the upper right quadrant. (G) Schematic diagram illustrating the spatial localization of TH-positive sympathetic nerves and CD31-positive endothelial cells within PLC structures.

Establishment of a method for simultaneous three-dimensional visualization of the mouse hepatic vascular system, related to Figure 1.

(A) Key steps of the optimized clearing protocol applied to the left liver lobe: untreated liver lobe (far left), following cardiac perfusion (second from left), after pretreatment (second from right), and after final clearing (far right, using 40% urea + 4.5% H2O2). Grid size: 1.6 mm × 1.6 mm. Scale bar: 8 mm. (B) Bright-field images of 1 mm-thick liver samples before and after treatment with four different clearing protocols for 20 minutes. Grid size: 1.6 mm × 1.6 mm. Scale bar: 8 mm. (C) Bright-field images of 200 μm-thick liver sections before and after treatment with four different clearing protocols for 5 minutes. (D) Scanning electron microscopy revealed that the metallic compound nanoparticle (MCNP) dyes consisted of aggregates of particles approximately 100 nm in length. After vascular labeling in liver tissue, these dye-pigment complexes adhered densely and uniformly to the vessel walls, forming well-defined labeled structures. (E–F) Whole-lobe bright-field images of the left liver lobe after bile duct labeling with green dye followed by clearing using the optimized protocol (40% urea + H2O2). In (E), grid size: 1.6 mm × 1.6 mm. In (F), scale bar: 1 mm. (G) Three-dimensional reconstructions of the hepatic vascular system: MCNP-Yellow labels the portal vein, MCNP- Pink labels the hepatic artery, MCNP-Green labels the bile duct, and MCNP-Black labels the central vein. Scale bar: 100 μm.

Analysis of PLC Distribution Along Portal Veins by Length, Diameter, and Area, related to Figure 2.

(A) Distribution of the number of periportal lamellar complex (PLC) structures along portal vein trunks with lengths ranging from 269.85 μm to 1513.67 μm. Each value represents the number of PLCs embedded along portal veins of a given length. (B) Distribution of the number of PLC structures along portal vein trunks with diameters ranging from 63.45 μm to 321.42 μm. Each value represents the number of PLCs embedded along portal veins of a given diameter. (C) Distribution of the total PLC area along portal vein trunks with lengths ranging from 269.85 μm to 1513.67 μm. Each value represents the cumulative area of PLC structures associated with portal veins of a given length.

Interactions Between PLC Structures and Terminal Biliary Tree in the Mouse Liver, related to Figure 3.

(A) Portal veins labeled with pink metallic nanoparticle dye combined with three-dimensional DAB immunohistochemistry (CK19⁺, brown) marking biliary epithelial cells. A larger view shows the accompanying distribution of bile ducts along portal veins. (B) Portal veins labeled with pink nanoparticle dye and bile ducts labeled with green nanoparticle dye, combined with three-dimensional DAB immunohistochemistry (CK19⁺, brown). The boxed region highlights a PLC structure adjacent to terminal bile duct branches. (C) Whole-lobe overview of the liver captured by extended-depth-of-focus imaging, showing high-pressure perfusion of green dye for bile ducts and yellow dye for portal veins, visualizing the spatial relationship between biliary and vascular systems.

Single-Cell Transcriptomic and Immunofluorescence Analysis of Endothelial Cell Subpopulations in the Mouse Liver, related to Figure 4.

(A) UMAP dimensionality reduction analysis showing five distinct endothelial cell clusters isolated from normal mouse liver after excluding non-endothelial cells. Each dot represents an individual cell. (B) Quantification of the number of cells in the five endothelial subpopulations: Central vein ECs, Pericentral LSECs, Periportal LSECs, Midzonal LSECs, and Portal vein ECs. The results showed that Central vein ECs were the most abundant (669 cells), while Portal vein ECs were the least (146 cells). This reflects the spatial quantitative differences among liver endothelial subpopulations based on single-cell RNA sequencing clustering. (C) UMAP-based heatmaps displaying the expression patterns of CD34 (left) and Sca-1 (right) across different liver endothelial subpopulations. Both markers were predominantly expressed in Portal vein ECs, Periportal LSECs, and Midzonal LSECs regions. The color gradient indicates expression levels, with deeper red representing higher expression. (D) Immunofluorescence staining of CD34, Sca-1, CD36, and PDGFRB in mouse liver tissue sections. CD34 and Sca- 1 were primarily localized around portal veins (PV) and adjacent vascular regions. CD36 was predominantly expressed in periportal liver sinusoidal endothelial cells, while PDGFRB showed widespread distribution in multiple intrahepatic endothelial cell populations. CV: central vein; PV: portal vein. Scale bar: 200 μm.

Distribution of CD31⁺, Sca-1⁺, and CK19⁺ Cells and Portal Vein–Associated Bile Duct Morphology in Control and Fibrotic Mouse Livers, related to Figure 5.

(A) Multiplex immunofluorescence staining showing the distribution of CD31 (yellow), Sca-1 (green), and CK19 (red) in the liver of control mice. Scale bar: 50 μm. (B) Higher-magnification view of multiplex immunofluorescence staining showing CD31 (yellow), Sca-1 (green), and CK19 (red) distribution in control mouse livers. Scale bar: 20 μm. (C) Quantification of the terminal bile duct area on the surface of portal veins with different diameters in control mice. Data are presented as mean ± standard deviation (Mean ± SD). Number of portal veins: control = 20, CCl₄-6 weeks = 18. (D) Quantification of CD34⁺ Sca-1⁺ endothelial cell subpopulations in control and fibrotic (CCl₄-induced) mouse livers.

Distribution of TH⁺ Sympathetic Nerve Fibers in Mouse and Human Liver, and Their Spatial Association with Hepatic Vessels, related to Figure 6.

(A) Immunofluorescence staining showing the distribution of tyrosine hydroxylase (TH) in liver sections from both mice and humans. Scale bar: 200 μm. (B) Multiplex immunofluorescence staining in control mouse livers showing the spatial distribution of CD31 (yellow), TH (green), and α-SMA (red). Scale bar: 50 μm. (C) Multiplex immunofluorescence staining in control mouse livers showing the distribution of CD31 (yellow), TH (green), and DAPI (red). Scale bar: 20 μm.