Figure 1.
Detection of hepatic microvesicular steatosis with ORO histology and CARS microscopy.
Representative ORO histology images (left column), CARS images (middle column), and surface plots of CARS images (right column) for wildtype mice with no steatosis (upper row), UPase-TG mice with mild microvesicular steatosis (middle row), and wildtype mice fed with 400 mg/kg/day fenofibrate for 5 days with severe microvesicular steatosis (lower row). Three mice, thus 3 livers, per animal group were used for direct comparison of ORO histology and CARS imaging analysis of microvesicular steatosis. Surface plots were performed with ImageJ software.
Figure 2.
Quantitative analysis of liver lipid level and composition with biochemical assay and multimodal CARS microscopy.
(A) Triglyceride concentration in milligram per gram liver tissue measured with a commercial triglyceride quantitation kit. (B) Liver lipid level determined with CARS signal intensity and normalized to 1 for wildtype mice and respectively for other animal groups. Error bars in A and B are standard deviations across 9 mice analyzed per animal group. (C) Representative Raman signatures of single liver lipid droplets of wildtype (light green), UPase-TG (blue), and wildtype+fenofibrate mice (red). (D) Lipid unsaturation (I1660/I1445) of liver lipid droplets as a function of animal group. Error bars are standard deviations across 60 lipid droplets analyzed per animal group (9 mice per group).
Figure 3.
ImageJ-assisted quantitative analysis of hepatic microvesicular steatosis imaged with CARS microscopy.
Columns (from left to right) describe the analysis sequence of a routine ImageJ-assisted analysis.
Figure 4.
Liver lipid droplet number and size of mice groups.
Liver lipid droplet number (A) and size (area) (B) as a function of mice groups determined with ImageJ-assisted analysis. Error bars are standard deviations of 9 liver volumes analyzed per mouse and 9 mice analyzed per animal group, or 81 volumes analyzed per animal group. The xyz dimensions of each analysis volume are 167 µm x 167 µm x 25 µm.
Figure 5.
Visualization and characterization of lipid-rich non-parenchymal cells (NPCs) with CARS and TPF imaging and spontaneous Raman microspectrometry analysis.
(A) CARS imaging of a lipid-rich NPC (arrow) in the interstitial space of hepatocytes. (B) CARS and TPF imaging of lipid-rich and autofluorescent lipid-rich NPCs. (C) Comparison of representative Raman spectra of hepatocytes (blue) and lipid-rich NPCs (red). Note the distinctive Raman peak at 1620 cm−1 for lipid-rich NPCs.
Figure 6.
Primary hepatocyte cultures for in vitro studies.
(A) CARS images and surface plots of representative primary hepatocytes at 24 hours post-isolation. Quantitative analysis performed on CARS images for (B) lipid level within individual hepatocyte, (C) lipid droplets per hepatocyte, and (D) lipid droplet areas. Error bars are standard deviations across 60 hepatocytes evaluated per animal group.
Figure 7.
Mitochondrial toxins induced microvesicular steatosis in cultured primary hepatocytes.
(A) CARS images and surface plots of representative hepatocytes treated with mitochondrial toxins for 24 hours. Quantitative analysis performed on CARS images for (B) lipid level within individual hepatocyte, (C) lipid droplets per hepatocyte, and (D) lipid droplet areas. Error bars are standard deviations across 60 hepatocytes evaluated per animal group.