Fig 1.
Computed tomography images (coronal multiplanar reconstruction views) of bearded dragons livers with different histologic classes based on an established grading system [30].
A) no-lipid, 48 Hounsfield Units (HU). B) Mild, 7HU. C) Severe, -36HU.
Fig 2.
Box plot with superimposed dot plot of hepatic computed tomography hounsfield units (HU) across two simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) based on an established grading system in bearded dragons [30] evaluated by a Mann-Whitney U test.
The dotted red line represents a cut-off value of 21 HU to separate the classes of no-lipid to mild and moderate to severe as determined on receiver operating characteristic curve analysis.
Fig 3.
Scatter plot of bearded dragon liver lipid content (%) from histological samples measured by digital analysis as a function of hepatic density (Hounsfield units) measured on computed tomography and analyzed using linear regression.
The blue line represents the regression line and the shading the 95% confidence interval.
Fig 4.
Box plot with superimposed dot plot of hepatic to fat pad echogenicity ratio on ultrasound across two simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) based on an established grading system in bearded dragons [30] evaluated by a t-test.
Fig 5.
Box plot with superimposed dot plot of beta-hydroxybutyric acid (umol/L) across two simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) in bearded dragons based on an established grading system [30] evaluated by a t-test.
The dotted red line represents a cut-off value of 272 umol/L determined on receiver operating characteristic analysis.
Fig 6.
Score plot of sparse partial least square- discrimination analysis (sPLS-DA) analysis of metabolomics, lipidomics, and biochemistry data between the 2 first principal components in bearded dragons with 2 simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) in bearded dragons based on an established grading system [30].
The right panel shows the loadings plot for the first component showing the most important variables for classification.
Fig 7.
Heatmap showing clustering of metabolites and biochemical analytes between simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) in bearded dragons based on an established grading system [30].
A clustering dendrogram is also present on the left; the different bearded dragons are on the x-axis and the analytes on the y-axis. Only the 75 most important analytes based on their t-test p-values are displayed. It should be noted that most of these analytes did not show significant differences between histologic classes on univariate analysis with a false discovery rate of 0.05. Color coding represents fold changes on normalized plasma concentrations with an increasing depth of orange representing increase in analyte concentration and an increasing depth of blue representing decrease in analyte concentration.
Fig 8.
Side-by-side box plots with superimposed dot plots of selected plasma variables (cholesterol [mmol/L], non-HDL [mol/L], NEFA [mmol/L], total ceramides [umol/L], carnitine [umol/L], and cholines [umol/l]) across two simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe) in bearded dragons based on an established grading system [30] evaluated by serial t-tests.
These variables were not statistically significant after correcting for false discovery, but interesting trends can be observed.
Fig 9.
Side-by-side box plots with superimposed dot plots of selected variables (CT [HU], BHBA [umol/L], succinic acid [umol/l], triglycerides [mmol/L]) across two simplified classes of hepatic lipid changes (no-lipid to mild and moderate to severe), based on an established grading system [30], in bearded dragons with and without (control) gemfibrozil treatment evaluated by t-tests.
The red boxplots represent baseline values and the teal represents values after a 2-month treatment trial with gemfibrozil. The dotted red line on the CT HU side-by-side box plot represents a cut-off value of 21 HU to separate the classes of no-lipid to mild and moderate to severe as determined on receiver operating characteristic analysis.There was a significant effect of gemfibrozil on succinic acid. While succinic acid significantly increased over time in the control group, it did not with gemfibrozil treatment. The remainder of the variables were not statistically significant but did show interesting trends.