Intraocular liver spheroids for non-invasive high-resolution in vivo monitoring of liver cell function

**Background:** Longitudinal monitoring of liver function in vivo is limited by the lack of high-resolution, non-invasive imaging techniques. Current in vitro 3D liver models lack physiological cues, vascularization, and innervation, while intravital imaging of the liver is often terminal. The anterior chamber of the eye (ACE) offers a transparent window for non-invasive imaging of engrafted tissues. This study aimed to establish and characterize an ACE-based platform for longitudinal, high-resolution in vivo imaging of liver spheroid function.

**Methods:** Mouse liver spheroids were generated from primary hepatocyte-enriched liver cells and transplanted into the ACE of recipient mice. Engraftment, vascularization, and innervation were assessed using in vivo confocal microscopy after intravenous injection of fluorescent lectin and labeled red blood cells, and ex vivo immunofluorescence staining. Transcriptomic analysis (RNA-seq) of explanted spheroids at 2 months post-transplantation was performed and compared to in situ liver and naïve iris. Hepatocyte functionality was evaluated in vivo by injecting fluorescent probes: CMFDA for bile acid transport, pHrodo Red-LDL for lipoprotein uptake, and SF44 for lipid droplets. Cell cycle activity was monitored using spheroids from FUCCI transgenic mice. The platform's ability to report on diet-induced steatosis was tested by feeding mice a high-fat-high-fructose diet (HFHFrD) for 12 weeks and comparing lipid accumulation in ACE-spheroids and in situ liver via Nile Red staining and TEM.

**Key Results:** Transplanted liver spheroids successfully engrafted on the iris, becoming vascularized with a capillary network (mean diameter 3.8 ± 0.9 µm) and receiving sympathetic innervation, but not parasympathetic. RNA-seq revealed that explanted spheroids retained hepatocyte-specific gene expression signatures and expressed genes related to bile production, lipid metabolism, and zonation (e.g., Glul, Cyp2f2). In vivo imaging demonstrated functional bile canaliculi (CMFDA excretion), LDL uptake (pHrodo Red-LDL accumulation), and cell cycle progression (FUCCI spheroids showed a boost in S/G2/M cells at 2 days post-transplantation, returning to low levels by 1 week). After 12 weeks of HFHFrD, ACE-spheroids accumulated lipid droplets in amounts and sizes similar to those in the in situ liver, as quantified by Nile Red staining and visualized in vivo with SF44.

**Clinical Implications:** This ACE-based imaging platform provides a novel, non-invasive method for longitudinal, single-cell resolution monitoring of liver cell function in living animals. It overcomes key limitations of current in vitro and in vivo models by providing a physiological environment with vascularization and innervation. The platform can be used to study liver physiology, disease progression (e.g., steatosis), and drug effects in real-time. The demonstration of human liver spheroid engraftment in immunodeficient mice suggests potential for translational applications, including preclinical drug testing and personalized medicine using patient-derived cells. However, the authors note that ACE-spheroids lack portal vein input and full lobular architecture, so they should be considered a reporter system rather than a complete liver model.