Intravital imaging reveals inflammation as a dominant pathophysiology of age-related hepatovascular changes

Abstract

Aging is the most significant risk factor for the majority of chronic diseases, including liver disease. The cellular, molecular, and pathophysiological mechanisms that promote age-induced hepatovascular changes are unknown due to our inability to visualize changes in liver pathophysiology in live mice over time. We performed quantitative liver intravital microscopy (qLIM) in live C57BL/6J mice to investigate the impact of aging on the hepatovascular system over a 24-mo period. qLIM revealed that age-related hepatic alterations include reduced liver sinusoidal blood flow, increased sinusoidal vessel diameter, and loss of small hepatic vessels. The ductular cell structure deteriorates with age, along with altered expression of hepatic junctional proteins. Furthermore, qLIM imaging revealed increased inflammation in the aged liver, which was linked to increased expression of proinflammatory macrophages, hepatic neutrophils, liver sinusoidal endothelial cells, senescent cells, and procoagulants. Finally, we detected elevated NF-κB pathway activity in aged livers. Overall, these findings emphasize the importance of inflammation in age-related hepatic vasculo-epithelial alterations and highlight the utility of qLIM in studying age-related effects in organ pathophysiology.

INTRODUCTION

The majority of people over the age of 65 have at least two chronic degenerative diseases (1). These chronic diseases among elderly individuals consume an increasing portion of our health-care costs while robbing people of their independence and quality of life. Creating therapies that target the primary risk factor for all of these diseases is a promising but difficult solution. Aging is the major risk factor for most chronic diseases, including diseases of the liver (2). Chronic liver disease (CLD) is one such chronic degenerative disease and is the 12th leading cause of death in the United States (3, 4). CLD results in over 70,000 deaths every year in the United States (3), and the current treatment is limited to supportive therapy. The cellular, molecular, and biophysical mechanisms that contribute to the progression of CLD are incompletely understood. Identifying the common molecular pathways that promote age-related liver injury would allow for the development of more effective therapies to prevent or halt the progression of CLDs.

Previous studies using experimental mouse models and clinical samples have shown that liver aging is associated with impaired proliferative (5) and metabolic functions (6), dysregulated cellular senescence (7), DNA damage (8), and apoptosis (9, 10). Similarly, studies on experimental models have shown changes in the hepatic sinusoidal environment (11) and adjacent biliary epithelial cells (12). However, the cellular and molecular basis that underlies the physiological manifestation of liver aging has remained elusive due to our inability to visualize liver pathophysiology in mice in real time in vivo. In the current study, we used quantitative liver intravital microscopy (qLIM) (13, 14) in live mouse livers to systematically compare the effects of aging using live C57BL/6J mice over a 24-mo period. Our findings are the first in vivo real-time studies to our knowledge in aged mouse liver and show that age-related hepatic alterations include decreased liver sinusoidal blood flow, increased sinusoidal vascular diameter, and loss of small hepatic vessels. We also show that the ductular cell structure changes with age, affecting hepatic junctional protein expression. Remarkably, qLIM imaging revealed increased inflammation in aged livers, linked to the enrichment of hepatic proinflammatory macrophages, neutrophils, senescent cells, procoagulants, and liver sinusoidal endothelial cells. Finally, aged livers had increased NF-κB pathway activity. Overall, these findings suggest that inflammation is the central pathophysiology associated with liver aging and highlights the importance of qLIM in studying age-related organ effects and molecular pathways.

METHODS

Animals

C57 BL/6J (3–24 mo old) mice were housed in a specific pathogen-free animal facility at the University of Pittsburgh. All animal experiments were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh. Livers from age- and sex-matched mice were used as indicated in results. Atleast three mice were assessed at all given time points. Both male and female mice were used for this study. All animal experiments and procedures were performed according to the guidelines of the NIH. The experiments were approved by university Institutional Animal Care and Use Committee.

Surgical Preparation and qLIM Imaging

Mice were anesthetized with an intraperitoneal (ip) injection of 100 mg·kg⁻¹ of body weight ketamine HCl (100 mg·mL⁻¹; Henry Shein Animal Health; Dublin, OH) and 20 mg·kg⁻¹ of body weight xylazine (20 mg·mL⁻¹; LLOYD Laboratories; Shenandoah, IA) and repositioned in the supine position. The right lobe of the liver was exposed through removal of the overlying skin and fat. Details of the surgical method are described in Ref. 15. Intravascular fluorescent dyes included 200 μg of Texas red (TXR) dextran, 100 μg of carboxyflurescein (CF), 100 μg of Violet 450 (V450) conjugated rat anti-mouse Ly6G mAb (clone DX5), 100 μg of AF488-anti CD31 Ab, 100 μg of AF488-Anti p21CIP1 antibody (Supplemental Table S2; all Supplemental material is available at https://doi.org/10.6084/m9.figshare.17203601.v1). TXR dextran/(MW 70,000) was used to visualize the blood flow through the liver sinusoids whereas CF (molecular weight 377) (16) was used to visualize uptake of the dye from blood to hepatocytes and then from the hepatocyte to the bile-canaliculi. Neutrophils were marked by violet 450 (V450) conjugated rat anti-mouse Ly6G mAb, liver sinusoidal endothelial cells were marked by AF488-anti CD31 Ab, and senescent cells were marked by AF488-Anti p21CIP1Ab. Microscopy was performed using a Nikon MPE multi-photon excitation microscope. This study was done in collaboration with CBI, University of Pittsburgh.

Image Analysis

Movies were processed using Nikon’s NIS Elements (Nikon Elements 3.10). A median filter with a kernel size of 3 was applied over each video frame to improve signal-to-noise ratio. Signal contrast in each channel of a multicolor image was further enhanced by adjusting the maxima and minima of the intensity histogram of that channel.

Measurement of hepatic blood flow.

To measure hepatic sinusoidal blood flow we tracked the movement of RBC (Red circular structure inside the hepatic sinusoidal vessel) as shown in Fig. 2A and Supplemental Videos S1–S3. In Fig. 2A, the “t” represents the specific time points in the video. The box outlines an individual RBC that is being tracked. The dashed line represents distance covered by the individual RBC within a specific time interval.

Measurement of sinusoidal vessel diameter.

To measure the hepatic sinusoidal vessel diameter, atleast 15 measurements of the hepatic sinusoidal vessel diameter for each image were made and averaged. Each data point represents average of 15 data points.

Quantification of total number of blood vessels/field of view.

To quantify the total number of hepatic sinusoidal vessels, total number of vessels were counted in each field of view (FOV) and repeated with atleast 10 FOV in each age group examined.

Measurement of CF intensity in hepatic bile ducts.

CF intensity was quantified using fluorescent intensity measurement by NIL software. Atleast six images were used/group to get average bile duct intensity.

Histology, Immunohistochemistry, and Immunofluorescence

Tissue sections (4–6 μm) were stained with hematoxylin and eosin (H&E). Immunohistochemistry (IHC) on paraffin-embedded sections was performed on livers as described elsewhere (15).

mRNA Isolation and Real-Time Polymerase Chain Reaction

mRNA was isolated and purified from livers of different age groups of C57/Bl6J mice (n = 3/group). mRNA was isolated using Trizol (Invitrogen). DNase treatment was performed on the RNA, and equal microgram amounts of RNA from each sample were used to create individual cDNA samples using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen). Each real-time PCR reaction used cDNA, 1× Power SYBR-Green PCR Master Mix (Applied Biosystems), and the appropriate primers. n = 3 samples were tested in duplicate for each condition using the Applied Biosystems StepOnePlus Real-Time PCR System and StepOne v2.1 software. The comparative ΔΔCT method was used for analysis of the data and all data are presented normalized to wild type (WT) at baseline. GAPDH and 18S expression was used as the internal control as described previously (13). Error bar represents standard deviation. Sequences of primers used in this study are available in Supplemental Table S1.

Western Blot

Approximately 20 mg of liver was homogenized in a buffer containing 25 mM Tris-HCl (pH 7.6), 150 mM NaCl, 1% Nonidet P-40, 1% sodium deoxycholate, and 0.1% SDS (RIPA buffer) supplemented with protease inhibitor. Protein assays were carried out using the bicinchoninic acid protein assay. SDS-PAGE analysis was used to resolve equal amounts of protein (10–50 g) on Tris-HCl precast gels (Bio-Rad) using the Mini-PROTEAN 3 Electrophoresis Module Assembly (Bio-Rad). The proteins were transferred to polyvinylidene difluoride membranes, then immunoblotted, and visualized using enhanced chemiluminescence. Immunoblotting was performed as described elsewhere (13). Briefly, membranes were washed five times for 5 min each in Tris-buffered saline-Tween 20 (TBST) before being probed with horseradish peroxidase (HRP)-conjugated secondary antibodies (1:5,000 diluted in TBST; Santa Cruz Biotechnology) for 1.5 h at room temperature. Membranes were washed three times for 10 min each in TBST and visualized using the Enhanced Chemiluminescence System (GE Healthcare). Supplemental Table S2 contains information on primary antibodies.

Statistical Analysis

All comparisons between two groups were deemed statistically significant by unpaired two-tailed Student’s t test if P < 0.05(*) or P < 0.001 (**). When more than two groups were compared, statistical analysis was performed with Prism version 7.0a (GraphPad Software) using one-way and two-way ANOVA with Bonferroni correction.

RESULTS

Vacuum-Enabled Liver Window for the Visualization of Real-Time Hepatic Blood and Bile Flow, Inflammation, and Senescence in C57Bl6 Mice

To enable qLIM imaging of the liver microcirculation in live mice, we utilized a modified liver window device (17). The device applies a gentle vacuum (negative pressure) to immobilize the right lobe of the mouse liver against a glass coverslip. We used WT C57BL/6 mice to establish the protocol for intravital liver imaging as described by Pradhan-Sundd (14). Upon intravascular administration of Triton X-dextran (TXR-dextran), our approach enabled visualization of the hepatic microcirculation. Real-time videos of the microcirculation show robust perfusion as the TXR dextran flows through the vasculature and around RBCs, which appear as dark spots in the capillaries and hepatic sinusoidal structure. The dynamic movement of hepatic blood flow and robust perfusion confirm that the liver window does not impede hepatic blood flow. Similarly, to visualize bile rafficking, we used carboxyfluorescein diacetate (CFDA). Recently, we have shown that CFDA can be used to analyze bile trafficking in different models of acute and chronic liver injury (13, 14). After being internalized by hepatocytes, CFDA is hydrolyzed by esterase into fluorogenic CF, which emits green fluorescence at 517 nm. In mice, the uptake of CFDA and hydrolysis into CF occurs within 1 min after injection. CF can thus be used as a surrogate marker for bile transfer. Alexa Fluor (AF) 488-conjugated anti-CD31, Alexa Fluor (AF) 405-conjugated anti-ly6g, FITC-conjugated F4/80, and AF488-conjugated anti-P-21 Abs were used to visualize hepatic sinusoidal endothelial cells, neutrophils, macrophages, and senescent cells, respectively. Following the selection of fluorescent dyes, we used qLIM to image the vascular and epithelial changes in C57BL/6J mice up to 2 years of age (Fig. 1A). The specific time points examined were 3, 7, 14, 21, and 24 mo. We also evaluated liver enzymes (ALT and AST) in the blood and the liver weight-to-body weight ratio in these age groups to evaluate potential liver injuries. As shown in Supplemental Fig. S1, A and B, we did not observe significant misexpression of serum ALT and AST levels in any age group examined. However, we found a moderate but significant increase in liver weight/body weight (LW/BW) with age (Supplemental Fig. S1C). Immunohistochemical analysis of the liver across different age groups also appeared comparable (Supplemental Fig. S1D). Taken together, we established a real-time imaging scheme to visualize age-induced hepatic pathophysiologic changes, which is otherwise impossible, using established biochemical and immunohistochemical analyses.

Figure 1.

Experimental design of real-time intravital imaging of age-induced hepatovascular changes. A: schematic diagram of quantitative liver intravital microscopy (qLIM) of mice across different age groups. B: experimental scheme: quantitative liver intravital imaging was used in 3-, 7-, 14-, 21-, and 24-mo-old mice to access the sinusoidal blood flow, hepatic vessel diameter, ductular structure, hepatic inflammation, and liver senescence. Scale bar 50 µm.

qLIM Identifies Reduced Hepatic Blood Flow and Increased Sinusoidal Vessel Diameter in Aged Mice

Aging is associated with vascular alterations and impaired blood flow (18, 19). We visualized the effect of aging on hepatic blood flow by analyzing the movement of RBCs within the hepatic microcirculation in live C57BL/6 WT mice using TXR-dextran. As shown in Fig. 2A, RBCs are present in large numbers in a mouse liver and are easy to visualize as round, shadowed objects as they travel through hepatic sinusoids. Real-time videos demonstrated that RBC flow is continuous and stable in a mouse liver. Thus, we measured the movement of RBCs across different frames as a surrogate for blood flow analysis. Briefly, a single RBC was marked, its movement was tracked per frame, and the distance covered per frame was quantified to identify the velocity per frame per second (Fig. 2A).

Figure 2.

Quantitative liver intravital microscopy (qLIM) identifies reduced hepatic blood flow in aged mice. A: analysis of hepatic blood flow by tracking RBC movement/frame within the hepatic microcirculation in 3–24-mo-old live C57BL/6J WT mice using Texas red (TXR)-dextran. B: quantification of hepatic blood flow as seen in 3–24-mo-old mice liver show significant reduction in blood flow at both 21- and 24-mo-old liver. **P > 0.05. Scale bar 50 µm.

Remarkably, as the mice aged, we noted a significant reduction in blood flow. As shown in Fig. 2A, compared with those in 3- and 7-mo-old mice, the hepatic blood flow in 14-mo-old mice was significantly reduced (Fig. 2B). Moreover, compared with 14-mo-old mice, we found significant reductions in blood flow in both 21- and 24-mo-old livers. As shown in Fig. 2A and quantified in Fig. 2B, both 21- and 24-mo-old mice showed significant reductions in hepatic blood flow. These observations are supported by previous studies, which have shown evidence of reduced blood flow within the hepatic sinusoids of aged livers (18–20).

As blood flow is closely related to sinusoidal vessel diameter, we next hypothesized that aging can cause changes in hepatic blood vessel diameter. Conceptually, the hepatic vessel diameter can be visualized as in Fig. 3A. As shown in Fig. 3, B and C, the hepatic blood vessel diameter was comparable in 3-, 7-, and 14-mo-old livers. Interestingly, as the mice aged, the hepatic sinusoidal diameter increased rapidly, resulting in broadening of the hepatic sinusoidal structure (Fig. 3, B and C). Sinusoidal vessels in the hepatic vasculature can be of varying lengths, both large and small, as shown in Fig. 3D. This phenotype was more prevalent in large vessels (<10 µm in diameter). Small vessels did not show a significant increase in vessel diameter (Fig. 3B).

Figure 3.

Quantitative liver intravital microscopy (qLIM) identifies increased sinusoidal vessel diameter in aged mice. A: experimental design of imaging and quantification of hepatic vessel diameter. B: qLIM imaging reveals broadening of the hepatic blood vessel diameter at 24 mo of age as compared with 3 mo old mice. C: quantification of hepatic vessel diameter across different age groups. D: schematic of large and small hepatic vessels in the liver. E: qLIM imaging reveals that the number of small vessels per field of view (FOV) were significantly reduced in aged mouse liver (21–24 mo old) as compared with 3–14 months old liver. F: quantification of small vessels per field of view (FOV) across different age groups. *P < 0.05. **P < 0.005. Scale bar 50 µm.

Moreover, when examined, the number of small vessels per FOV was significantly reduced in aged mouse livers compared with 3- to 14-mo-old livers. As shown in Fig. 3E and quantified in Fig. 3F, 3- to 14-mo-old mouse livers showed 10–20 small vessels per FOV, which was reduced to 6–8 small vessels in 21- to 24-mo-old livers (Fig. 3F). Thus, qLIM imaging was able to identify age-specific hepatic blood vessel changes, including blood flow, vessel diameter, and changes in vessel number, with age.

Impaired Bile Trafficking is Associated with Aging in the Murine Liver

Bile synthesis and transport are two of the most important functions of the liver (21). Bile, upon its synthesis in hepatocytes, is modified downstream by the bile duct epithelium and is then trafficked to bile canaliculi (22) (the movement of bile is shown in Fig. 4A). To examine the effect of aging on bile trafficking, we next compared CF trafficking in different age groups of mice. CFs trafficked from hepatocytes to bile canaliculi immediately after injection in the livers of 3-mo-old mice (Fig. 4B). CF staining completely localized to the bile canaliculi, indicative of real-time transport and an intact blood biliary barrier (BBlB) (Fig. 4B). When compared, the localization of CFs into bile canaliculi appeared comparable in 3- to 14-mo-old mice (Fig. 4B). However, in 21- to 24-mo-old mice, the overall CF intensity was significantly reduced, and many of the canaliculi were devoid of CF staining (Fig. 4B). As shown in Fig. 4B (quantified in Fig. 4C), the fluorescence intensity of CFs was not uniformly present in aged mice starting from 14 mo onward. Finally, the 21- to 24-month-old mouse livers showed significant diminishments in CF intensity in bile canaliculi/FOV. Altogether, the real-time analysis of CF suggested that both bile canalicular structure and function diminish with age, impacting overall bile trafficking.

Figure 4.

Impaired bile trafficking is associated with aging in the murine liver. A: schematic showing the organization of hepatic vasculoepithelial system showing the position of ductular cells in the liver. B: intravital images of 3-, 7-, 14-, 21-, and 24-mo-old liver mice liver using carboxyflurescein (CF) (green) and Texas red (TXR)-dextran (red). C: quantification of CF intensity across different age groups shows significant reduction in CF intensity across bile duct in 21–24 mo-old mice. D: heat map showing the RT-PCR data of Junctional markers associated with blood biliary barrier in young and old mice. E: heat map showing the RT-PCR data of epithelial to mesenchymal transition (EMT) markers associated with blood biliary barrier in young and old mice. **P = 0.05. Scale bar 50 µm.

Next, we determined whether aging can also lead to misexpression of blood biliary barrier (BBlB)-related genes and impaired polarization of hepatic epithelial cells. Recently, we established a role for the tight junction (TJ) proteins claudins and occludin and the adherence junction (AJ) proteins β-catenin and e-cadherin in the maintenance of BBlB integrity (14). Remarkably, we found that the mRNA levels of plaque-forming claudins (3 and 5) (23) responsible for the maintenance of transepithelial resistance were significantly reduced, while the expression levels of pore-forming claudins (2 and 15) known to decrease transepithelial resistance were significantly increased (Fig. 4C, Supplemental Fig. S3A). Moreover, among other junctional proteins associated with BBlB and hepatocyte polarity, β-catenin, E-cadherin, and occludin mRNA levels were upregulated in the livers of aged mice (24 mo old) (Fig. 4D, Supplemental Fig. S3B). This finding was further supported by Western blot analyses, which revealed a significant reduction in claudin-3 protein and increase in occludin and e-cadherin expression in the livers of aged mice (Fig. 4D, Supplemental S3C).

Previously, we showed that loss of hepatic junction proteins is associated with loss of epithelial identity and transition to a more mesenchymal characteristic (24), also described as epithelial to mesenchymal transition (EMT). We hypothesized that the loss of junctional markers is associated with EMT in the aged liver. Remarkably, the livers of aged mice showed significant upregulation of mesenchymal markers, such as vimentin and fibroblast-specific protein-1 (Fig. 4E). Moreover, the mRNA expression levels of EMT inducers, such as snail, slug, TGF-β, and twist, showed significant increases in aged livers compared with young livers (Fig. 4E, Supplemental Fig. S2C). Altogether, the localization and expression of key molecules that convey epithelial identity were aberrant in aged livers. Collectively, our data suggest that the progression of aging in the liver is associated with impaired bile trafficking and misexpression of EMT and hepatic junctional markers.

Aging is Associated with Increased Inflammation and Augmented NF-κB Signaling in the Liver

Chronic age-related diseases are increasingly being linked to increased inflammation (1, 25, 26). We hypothesized that the age-induced hepatic changes are caused by inflammation. We performed qLIM to confirm the presence of exacerbated inflammation in the aged liver. The vascular dye Texas red (TXR)-dextran (red) and FITC-conjugated F4/80 antibody (green) were intravenously administered to visualize blood flow and hepatic macrophages, respectively. Remarkably, we found significant increases in AF-488-anti F4/80 antibody in the livers of 21-mo-old mice, which were not seen in the livers of 3-mo-old mice (Fig. 5, A and D), suggesting enrichment of hepatic Kupffer cells and enhanced inflammation. Interestingly, we also found a strong enrichment of AF488-anti F4/80 antibody in the small veins of the liver upon aging (Fig. 5A). Chronic inflammation has been associated with activation in liver sinusoidal endothelial cells (27, 28). Upon real-time analysis of liver sinusoidal endothelial cells (LSECs) using an AF488-conjugated anti-CD31 antibody, we found a significant increase in LSECs in the livers of 21-mo-old mice, which was not observed in the livers of young (3 mo old) mice and was suggestive of activation of LSECs upon aging (Fig. 5, B and E). Interestingly, we did not observe a very significant enrichment of neutrophil-specific Ly6g in the hepatic microvasculature upon aging (Fig. 5, C and F).

Figure 5.

Aging is associated with increased inflammation in the liver. A: schematic diagram of quantitative liver intravital microscopy (qLIM) imaging of mice using macrophage marker (AF488-F4/80Ab) and Texas-red dextran. Intravital images of young (3 mo old) and old mice (24 mo old) mouse liver post administration of AF488-F4/80 Ab (green) and Texas-red dextran (red). B: schematic diagram of qLIM imaging of mice using liver sinusoidal endothelial cell (LSEC) marker (AF488-CD31 Ab) and Texas-red dextran. Intravital images of young (3 mo old) and old (24 mo old) mouse liver post administration of AF488-CD31 Ab (green) and Texas-red dextran (red). C: schematic diagram of qLIM imaging of mice using neutrophil marker (AF405-Ly6G Ab) and Texas-red dextran. Intravital images of young (3 mo old) and old mice (24 mo old) mice liver post administration of AF405-Ly6G Ab (violet) and Texas-red dextran (red). D: quantification of F4/80 positive cells/field of view (FOV) in young (3 mo old) and old mice (24 mo old). E: quantification of CD31 positive cells/FOV in young (3 mo old) and old (24 mo old) mouse liver. F: quantification of Ly6G positive cells/FOV in young (3 mo old) and old mice (24 mo old) mouse liver. *P < 0.05. Scale bar 50 µm.

To confirm the enhanced inflammation in aged livers, we next evaluated molecular markers of inflammation. Western blot analysis of markers associated with liver inflammation (F4/80, CD45) confirmed the increased expression in aged liver (Fig. 6A). Given that aged mice show a significant increase in hepatic macrophages, we hypothesized that aging in the liver leads to activation of a proinflammatory phenotype. Interestingly, when examined, qRT-PCR analysis showed significant activation of the M1 group of macrophages and the loss of M2-specific macrophages in aged livers at baseline (Fig. 6B, Supplemental Fig. S3A). Similarly, the expression levels of LSEC markers such as vascular cell adhesion molecule 1 (VCAM1), endothelial NOs (eNOs), MAF, and endostatin were significantly upregulated in the aged liver (Fig. 6C, Supplemental Fig. S3B). In support of the inflammatory milieu associated with aged liver, we examined the expression of hepatic stellate cell (HSC) activation, which accompanies inflammation in models of chronic liver diseases (29). As expected, markers of HSCs, including collagen (col)1a2, col2a1, transforming growth factor β (TGFβ), TIMP metalloprotease inhibitor 1 (TIMP1), and alpha smooth muscle cells actin (αSMA), showed significant levels of enrichment in aged liver (Fig. 6D). Consistent with persistent inflammation, qRT-PCR analysis also confirmed increased cellular adhesion by activation of cell adhesion components in aged liver (Fig. 6E).

Figure 6.

Aging is associated with increased inflammation in the liver. A: Western blot analysis of macrophage markers F4/80 and CD45 in young (3 mo old) and old (24 mo old) mice. B: heat map showing the RT-PCR data of M1 and M2 macrophage markers in young (3 mo old) and old (24 mo old) mice. C: heat map showing the RT-PCR data of cell adhesion molecules in young (3 mo old) and old (24 mo old) mice. D: heat map showing the RT-PCR data of liver sinusoidal endothelial cell (LSEC) markers in young (3 mo old) and old (24 mo old) mice. E: heat map showing the RT-PCR data of hepatic stellate cell (HSC) markers in young (3 mo old) and old (24 mo old) mice. F: heat map showing the RT-PCR data of coagulation markers in young (3 mo old) and old (24 mo old) mice. G: Western blot analysis of shows increased expression of von Willebrand factor (VWF) in old (24 mo old) liver. H: heat map showing the RT-PCR data of NF-κB pathway components in young (3 mo old) and old (24 mo old) mice. I: Western blot analysis of shows increased expression of NFκB (P65) in old (24 mo old) liver.

Finally, we examined the effect of aging on the expression of coagulant proteins in the liver. As illustrated in Fig. 6, F and G, procoagulant factor VIII (FVIII), von Willebrand factor (vWF), and α and β fibrinogen showed strong enrichment, suggestive of procoagulant activity in aged liver (Fig. 6, F and G).

Given that the livers of aged mice show persistent hepatic inflammation, we hypothesized the potential activation of the nuclear factor kappa B (NF-κB) pathway in the livers of aged mice (30, 31). An evaluation of NF-κB and target genes, including NF-κB1, NF-κB2, v-rel avian reticuloendotheliosis viral oncogene homolog A (RelA), v-rel avian reticuloendotheliosis viral oncogene homolog B (RelB), intercellular adhesion molecule 1 (ICAM1), and VCAM1, by qRT-PCR revealed generalized activation in aged mouse livers (Fig. 6H, Supplemental Fig. S3F). This was further validated by Western blotting using antibodies against phospho-P65, which showed a significant upregulation in the livers of aged mice compared with young mouse livers (Fig. 6I). Thus, these data suggest that NF-κB and its downstream targets are upregulated in aged livers.

The mRNA expression levels of the proinflammatory cytokines TNFα, IL-1β, and IL-6, which are known to activate NF-κB signaling, were also significantly upregulated (Fig. 6J, Supplemental Fig. S3G) in aged mouse livers. In summary, these data show that persistent inflammation is associated with liver aging, causing significant changes in hepatic vasculoepithelial changes, including increased cellular adhesion, enrichment of the proinflammatory macrophage population, activation of hepatic stellate cells, and endothelial cells and procoagulant activity.

Real-Time Imaging of Cellular Senescence Shows p21CIP1 Accumulation in Aged Mouse Liver

Finally, as aging is closely associated with cellular senescence (32), we analyzed senescence in aged liver. We used qLIM to visualize markers of cellular senescence in young and old mouse livers. Texas red (TXR)-dextran (red) and AF488-anti P-21 antibody (green) were intravenously administered to visualize hepatic blood flow and senescent cells, respectively (Fig. 7A). In young livers, control mouse livers showed normal blood flow and mild expression of AF488-anti P-21. However, we found significant increases in AF488-p21CIP1 in aged mice from 14 mo onward, suggesting increased senescence upon aging (data no shown). Twenty-one-month-old liver RT-PCR and Western blot analysis also showed a significant upregulation of p21CIP1 expression in aged liver compared with young liver (Fig. 7, C and D). Interestingly, other regulators of cellular senescence, p16^INK4a, HMGA1, and gH2AX, were not significantly upregulated by RT-PCR compared with the control (Fig. 7C). Thus, aging is associated with p21CIP1-dependent cellular senescence in murine liver.

Figure 7.

Real-time visualization of cellular senescence in aged mouse liver. A: schematic diagram of quantitative liver intravital microscopy (qLIM) imaging of mice using senescence marker (AF488-p21CIP1Ab) and Texas-red dextran. Intravital images of young (3 mo old) and old (24 mo old) mice liver post administration of AF488-p21CIP1 Ab (green) and Texas-red dextran (red). B: quantification of p21CIP1 positive cells/field of view (FOV) in young (3 mo old) and old (24 mo old) mice. C: qRT-PCR analysis reveals the expression level of senescence markers in young (3 mo old) and old (24 mo old) mice. Scale bar 50 µm. D: Western blot for p21CIP1 confirms significant upregulation in the liver of aged (24 mo old) mice. E: schematic diagram depicting hepatovascular changes in aged liver compared with control young mouse liver as seen using intravital imaging.

DISCUSSION

Several previous studies have performed immunohistochemistry of fixed liver tissue and biochemical analyses of liver enzymes to characterize age-induced liver injury in mice (18, 19, 33, 34); however, the cellular, molecular, and biophysical events that promote age-induced changes in the liver remain largely unknown. Using real-time intravital imaging, we found that aging mice had reduced hepatic blood flow, increased sinusoidal diameter, aberrant bile trafficking, procoagulant activity, enhanced cellular adhesion, low-grade persistent chronic inflammation, and cytokine activation in the liver. Mechanistically, we show that NF-κB activation is linked to liver aging. This study serves as the first step in establishing real-time cellular events that promote age-induced hepatic changes and provides an experimental strategy that can be used to answer questions pertaining to the cellular and molecular mechanisms of age-induced chronic liver injury in response to an inflammatory stimulus in future studies.

The vascular alterations in aged livers included reduced blood flow, increased sinusoidal vessel diameter, and loss of small hepatic vessels. Persistent inflammation in the hepatic-sinusoidal milieu may promote adhesion of sinusoidal components, resulting in shear stress-induced vessel wall widening and decreased total blood flow. Similarly, in biliary epithelial cells, we found progressive loss of CF-positive ductular structures and overall reduced intensity of signaling upon aging. Interestingly, progressive loss of bile ductular structure is associated with various cholestatic liver diseases in which patients have to undergo liver transplants due to cholestasis and inflammation-induced liver failure (35–37). The changes seen in biliary cells of aged mice could also be predominantly inflammation driven. It would be interesting to see if aged liver can show cholestatic disease (such as primary biliary cholangitis or primary sclerosing cholangitis) like phenotypes and if blocking inflammation can lead to amelioration of these phenotypes.

Analysis of junctional and epithelial-mesenchymal transition (EMT) markers indicated age-related disruption of hepatic junctions and EMT. EMT is seen in several cholestatic liver disorders (38, 39); however, the molecular pathways linking polarity loss to cholestasis and increased inflammation are not well understood. Determining these links would enable the creation of innovative cholestatic treatments that stabilize EMT or hepatic junctions. In addition, it would be fascinating to see if forced re-expression of junctional proteins could repair bile canaliculi abnormalities in the elderly mouse liver.

Previous research has linked aging to cellular inflammation (6, 40, 41). We show that chronic inflammation in the elderly liver activates NF-κB and its target genes. Apoptosis, fibrosis, ductular injury, vascular malformation, and inflammation are all controlled by NF-kB signaling. Our data imply that inhibiting NF-κB may be a treatment option for age-induced chronic liver injury. Several NF-κB inhibitors have shown anticancer and antinecrotic activities (42) in the liver, which should be studied further.

Our findings are associated with a few limitations that warrant further investigation in future studies. First, our study did not examine the effects of liver zonation on age-induced vascular and nonvascular changes. Second, the current study strongly establishes the presence of inflammation as one of the potential driving factors of age-induced vascular changes; however, the cause-and-effect relationship between these pathological events remains to be elucidated. Notwithstanding these limitations, the current study is the first to identify the real-time events associated with hepatic vascular and nonvascular changes in aged mice. An improved understanding of these processes could potentially benefit the development of new therapies and biomarkers to treat or prevent age-induced chronic liver injury. We anticipate that qLIM will open up new avenues for the investigation of age-induced chronic liver injury in vivo.

SUPPLEMENTAL DATA

Supplemental Videos S1–S3, Supplemental Tables S1 and S2, Supplemental Figs. S1–S3: https://doi.org/10.6084/m9.figshare.17203601.v1.

GRANTS

This work was supported by a predoctoral grant from the American Heart Association (to R. Vats), a postdoctoral grant from the American Heart Association (to T. W. Kaminski), and by NIDDK grant DK125617 to (T. Pradhan-Sundd).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

T.P-S. conceived and designed research; R.V., Z.L., E-M.J., and R.K.D. performed experiments; R.V., Z.L., E-M.J., T.W.K., and T.P-S. analyzed data; T.W.K., S.W., and T.P-S. interpreted results of experiments; Z.L., E-M.J., and T.P-S. prepared figures; T.P-S. drafted manuscript; T.P-S. edited and revised manuscript; R.V., Z.L., E-M.J., R.K.D., T.W.K., S.W., and T.P-S. approved final version of manuscript.