Fig 1.
Gross morphology and histological appearance of Glmpgt/gt mouse liver.
A) Representative images of liver and spleen from 4 and 10 months old male Glmpgt/gt and WT mice. The Glmpgt/gt mice had a tuberous liver appearance and splenomegaly, consistent with previous reports [37, 39]. B-D) Paraffin-embedded liver sections stained with B) haematoxylin and eosin (arrows point to tissue injury and infiltration of inflammatory cells in the Glmpgt/gt liver sections), C) Picro Sirius Red collagen stain, and D) an antibody to ICAM-1. Positive ICAM-1 staining is seen as brown pigment in sinusoids (arrows) and veins in the Glmpgt/gt liver sections. Scale bars in B-D: 50 μM.
Fig 2.
FITC-FSA uptake and collagen content in the liver from Glmpgt/gt versus WT mice.
A) Uptake of FITC-FSA in liver of Glmpgt/gt mice (n = 16) and WT mice (n = 14), presented as % FITC-positive area in liver sections. Organs were harvested 10 min after intravenous injection of FITC-FSA (dose: 2 μg/g body weight). Age of young groups: Glmpgt/gt, 4 months; WT, 3–6 months. Old groups (both genotypes): 9–10 months. Each dot represents one animal, median values are presented as horizontal lines, and upper and lower lines represent the interquartile range. Mann Whitney U test compared results from Glmpgt/gt and age- and gender-matched WT mice. Young males: *p-value < 0.01; young females: not statistically significant; old males: *p-value < 0.01. B) Collagen content (% Picro Sirius Red stained area) in the liver of Glmpgt/gt mice (n = 16) and WT mice (n = 14), quantified by image analysis as described in Methods. Each dot represents one animal, median values are presented as horizontal lines, and upper and lower lines represent interquartile range. Mann Whitney U test: young males, *p-value < 0.01; young females, not statistically significant; old males, *p-value < 0.01. C) Distribution of FITC-FSA (bright green fluorescence, white arrows), and collagen (red stain) in parallel sections of liver samples from 4 months old WT and Glmpgt/gt mice. In the WT liver, FITC-fluorescence is seen as small bright green dots in the endothelial cells of most sinusoids. At the same time the parallel Picro Sirius Red stained section shows only a low amount of collagen in the sinusoids. In the Glmpgt/gt mouse liver uptake of FITC-FSA was low or absent in areas with collagen accumulation. Scale bars: 20 μm.
Fig 3.
Stabilin-1 and stabilin-2 expression in liver from Glmpgt/gt and WT mice.
A-B) Quantitative PCR analysis of Stab1 (A) and Stab2 (B) expression in liver tissue from 4 months (“young”) and 9 months (“old”) WT and Glmpgt/gt male mice (young: n = 4 per group; old: n = 3 per group). Error bars represent standard deviation. Results were not significantly different with Mann Whitney U test or the Kruskal-Wallis test. C) Quantitative image analysis of stabilin-1-stained liver sections from WT and Glmpgt/gt mice, presented as % positively stained tissue area. Groups: Young WT, 3–6 months (n = 5); young Glmpgt/gt, 4 months (n = 5); old WT, 9–10 months (n = 4); and old Glmpgt/gt, 9–10 months (n = 4). Each dot represents one animal, the median value for each group is presented as a horizontal line, and the upper and lower lines represent the interquartile range. Statistical analysis showed no significant differences between groups (One-way non-parametric ANOVA on ranks/Kruskal-Wallis test). D-E) Distribution pattern of stabilin-1 (red fluorescence) in liver of D) 4 months old, and E) 9–10 months old Glmpgt/gt and WT mice injected intravenously with FITC-FSA (2 μg/g body weight, 10 min monitoring time). Scale bars: 20 μμm. In the WT mice (both age groups) stabilin-1 was widely distributed in the sinusoids, highly colocalizing with FITC-FSA (arrows in D-E). At the same time, a few stabilin-1 positive, FITC-negative cells were observed in the portal tract (arrowhead in D, WT row). In the Glmpgt/gt mice, stabilin-1 staining was seen in FITC-positive cells in the liver sinusoids (arrows in D-E) and FITC-negative inflammatory cell aggregates (arrowheads in D-E). F) Co-localisation of stabilin-1 (red fluorescence) with the macrophage marker VSIG4 [51] (light blue fluorescence, arrowheads) in Glmpgt/gt liver. Arrows point to positive stabilin-1 staining of sinusoidal endothelial cells, which are VSIG4 negative [20]. Scale bar: 20 μm.
Fig 4.
Mannose receptor expression in liver from Glmpgt/gt and WT mice.
A-B) Mannose receptor expression in liver sinusoids of A) 4 months old, and B) 9 months old WT and Glmpgt/gt male mice. Positive mannose receptor staining (red fluorescence, arrows) was seen in the sinusoids, in the same cells that had endocytosed FITC-FSA (green fluorescence). Nuclei are stained with DAPI (blue). Scale bars: 20 μm. C) Quantitative image analysis of mannose receptor staining (% positive area) in liver sections from WT and Glmpgt/gt mice. Groups: Young WT, 3–6 months (n = 5); young Glmpgt/gt, 4 months (n = 5); old WT, 9–10 months (n = 4); and old Glmpgt/gt, 9–10 months (n = 5). Each dot represents one animal, the median value for each group is presented as a horizontal line, and upper and lower lines represent the interquartile range. One-way non-parametric ANOVA on ranks (Kruskal-Wallis test) showed no significant differences between groups. D) Mcr1 mRNA expression (qPCR) in liver tissue from 4 months (“young”) and 9 months (“old”) WT and Glmpgt/gt male mice (young: n = 4 per group; old: n = 3 per group). Error bars represent standard deviation. Results were not significant in non-parametric tests. E-F) Western blots showing mannose receptor expression in whole liver lysates from E) 4 WT mice and 3 Glmpgt/gt mice, aged 4 months, and F) 4 WT mice and 3 Glmpgt/gt mice, aged 9 months, all male. LSEC: Mouse liver sinusoidal endothelial cell lysates C57Bl/6JRj, WT). Protein loaded per lane: Liver lysates, 25 μg; LSEC, 5 μg. Beta-actin loading control was performed on the bottom part of the blots.
Fig 5.
FcγRIIb expression in liver from Glmpgt/gt and WT mice.
A-B) FcγRII expression in liver sinusoids of A) 4 months old, and B) 9–10 months old Glmpgt/gt and WT male mice. In WT livers, positive immunostaining (red fluorescence) was seen along the sinusoids in the same cells that had taken up FITC-FSA (green fluorescence), while the expression was low or absent in Glmpgt/gt livers, including cells that had taken up FITC-FSA (arrows). Nuclei in A) are stained with DAPI (blue). Scale bars: 20 μm. C) Quantitative image analysis of FcγRII staining (% positive area) in liver sections from WT and Glmpgt/gt mice. Groups: Young WT, 3–6 months (n = 5); young Glmpgt/gt, 4 months (n = 5); old WT, 9–10 months (n = 4), and old Glmpgt/gt, 9–10 months (n = 6). Each dot represents one animal. Medians are presented as horizontal lines, and upper and lower lines represent interquartile range. *p-value < 0.05, **p-value < 0.01, One-way non-parametric ANOVA on ranks (Kruskal-Wallis test). D) Fcgr2b expression (qPCR) in liver tissue from 4 months (“young”) and 9 months (“old”) WT and Glmpgt/gt male mice (young: n = 4 per group; old: n = 3 per group). *p-value < 0.05 (Mann Whitney U test). Error bars represent standard deviation. E, F) Western blots showing FcγRII expression in whole liver lysates from E) 4 WT mice and 3 Glmpgt/gt mice, aged 4 months, and F) 4 WT mice and 3 Glmpgt/gt mice, aged 9 months, all male. LSEC: Mouse liver sinusoidal endothelial cell lysates (C57Bl/6JRj, WT). Protein loaded per lane: Liver lysates, 25 μg; LSEC, 5 μg. Beta-actin loading control was performed on the stripped blots.
Fig 6.
Scanning EM of liver samples from 4 months old WT and Glmpgt/gt mice.
A) Scanning EM image of a liver sample from WT mice. A1 and A2 show high magnification images of two typical sinusoids, labelled 1 and 2 in the overview image in A. B-D) Scanning EM of liver samples from Glmpgt/gt mice, including overview images and high magnification images from the areas indicated with numbered arrows in the overviews. The LSECs of Glmpgt/gt were generally well fenestrated, with fenestrae arranged in sieve plates (circles in A1, B1, C1, D2). Large gaps (Ga) were observed in LSECs close to or within areas with hepatocyte damage and infiltration of immune cells (C1, D1-2). D3 shows a non-fenestrated capillary; these were only observed in areas with hepatocyte destruction and infiltration of leukocytes (dashed ellipse in the overview image in D. Abbreviations in A-D: LSEC, liver sinusoidal endothelial cell; HC, hepatocyte; SD, space of Disse; Ga, gap in LSEC; Coll, collagen; Cap, non-fenestrated capillary; L, leukocyte.
Fig 7.
Scanning EM of liver samples from 10 months old Glmpgt/gt mice, and frequency of fenestrae and gaps in LSECs of Glmpgt/gt and WT mice in situ.
A-C) Scanning EM images of representative sinusoids in the liver of 10 months old Glmpgt/gt mice. D) Insert in image C, showing part of a highly fenestrated LSEC overlaying a collagen bundle. White circles in A-B show fenestrae arranged in sieve plates, while arrows in D point to single fenestrae in sieve plates. Abbreviations in A-D: LSEC, liver sinusoidal endothelial cell; HC, hepatocyte; RBC, red blood cell; Ga, gap in LSEC; Coll, collagen. E) Frequency of fenestrae (i.e. number of open holes 30–400 nm in diameter, per μm2), and F) frequency of gaps (open holes > 400 nm in diameter, per μμm2) in LSECs were measured on scanning EM images of liver samples from 4 young WT mice (age: 4–6 months), 4 young Glmpgt/gt mice (age: 4 months), and 4 old Glmpgt/gt mice (age: 9–10 months). The images were captured at 20.000 x magnification, and 206 images (11–29 images/liver) were analysed as described in Methods. Each dot represents the average value for one mouse, i.e. the value included in the statistical analysis (Kruskal-Wallis test), the median value for each group is presented as a horizontal line, and the upper and lower lines represent the interquartile range. *p-value < 0.05.
Fig 8.
Uptake of FITC-FSA in sinusoids of bone marrow and spleen.
Distribution of FITC-FSA uptake in sinusoids of the spleen (A) and bone marrow (C) 10 min after intravenous injection of ligand in Glmpgt/gt and WT mice. Specific FITC-fluorescence is seen as small, bright green dots (arrows) in the sinusoidal lining cells. Scale bars: 20 μm. B) FITC-FSA uptake in the spleen, presented as % area covered by FITC-fluorescence in tissue sections. Groups: Young WT, 3–6 months (n = 5); old WT, 9–10 months (n = 4); young Glmpgt/gt, 4 months (n = 6); and old Glmpgt/gt, 9–10 months (n = 4). Medians are presented as horizontal lines, and the upper and lower lines represent the interquartile range. Significantly different in old Glmpgt/gt vs WT, * p-value <0.05, One-way non-parametric ANOVA on ranks (Kruskal-Wallis test). D) Images of bone marrow sections from WT and Glmpgt/gt mice, stained with haematoxylin and eosin. S, sinusoid. Scale bars: 50 μm.