Formation of fenestrae in murine liver sinusoids depends on plasmalemma vesicle-associated protein and is required for lipoprotein passage - PubMed (original) (raw)

Formation of fenestrae in murine liver sinusoids depends on plasmalemma vesicle-associated protein and is required for lipoprotein passage

Leonie Herrnberger et al. PLoS One. 2014.

Abstract

Liver sinusoidal endothelial cells (LSEC) are characterized by the presence of fenestrations that are not bridged by a diaphragm. The molecular mechanisms that control the formation of the fenestrations are largely unclear. Here we report that mice, which are deficient in plasmalemma vesicle-associated protein (PLVAP), develop a distinct phenotype that is caused by the lack of sinusoidal fenestrations. Fenestrations with a diaphragm were not observed in mouse LSEC at three weeks of age, but were present during embryonic life starting from embryonic day 12.5. PLVAP was expressed in LSEC of wild-type mice, but not in that of Plvap-deficient littermates. Plvap(-/-) LSEC showed a pronounced and highly significant reduction in the number of fenestrations, a finding, which was seen both by transmission and scanning electron microscopy. The lack of fenestrations was associated with an impaired passage of macromolecules such as FITC-dextran and quantum dot nanoparticles from the sinusoidal lumen into Disse's space. Plvap-deficient mice suffered from a pronounced hyperlipoproteinemia as evidenced by milky plasma and the presence of lipid granules that occluded kidney and liver capillaries. By NMR spectroscopy of plasma, the nature of hyperlipoproteinemia was identified as massive accumulation of chylomicron remnants. Plasma levels of low density lipoproteins (LDL) were also significantly increased as were those of cholesterol and triglycerides. In contrast, plasma levels of high density lipoproteins (HDL), albumin and total protein were reduced. At around three weeks of life, Plvap-deficient livers developed extensive multivesicular steatosis, steatohepatitis, and fibrosis. PLVAP is critically required for the formation of fenestrations in LSEC. Lack of fenestrations caused by PLVAP deficiency substantially impairs the passage of chylomicron remnants between liver sinusoids and hepatocytes, and finally leads to liver damage.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. PLVAP in the liver.

A, β-galactosidase stained sections through the livers of a Plvap -/- animal and its wild-type littermate at E15.5. Staining is seen in the sinusoidal endothelium of the _Plvap_-deficient embryo (white arrows), but not in the wild-type. Staining in the liver is patchy and weaker than the homogenous staining of capillaries in the small intestine (In, black arrows) or skin (Sk, arrowheads). Lower panels show magnification of the staining in liver sinusoids (white arrows). B, Immunolabeling for PLVAP (red) shows staining in the sinusoidal endothelium of a wild-type embryo at E16.5 (boxed areas in the left panel). Stronger immunostaining for PLVAP is seen in capillaries of the small intestine (white arrow) and the pancreas (black arrow). No staining is seen in the _Plvap_-deficient littermate (right panel). Middle panel shows magnification of the boxed panel. C, Liver staining for PLVAP (red) at P11 shows positive immunoreactivity in wild-type sinusoidal endothelium, but not in the Plvap -/- littermate. Nuclear DNA is labeled with DAPI (blue). D and E, TEM. Fenestrations bridged by 5 - 7 nm thick diaphragms (black arrows) are seen in wild-type sinusoidal endothelium at E12.5 (D). Caveolae of wild-type sinusoidal endothelium at P20 are not bridged by stomatal diaphragms (black arrow, E).

Figure 2. Plvap expression in the liver.

A, RT-PCR analysis from lung and liver tissues. A 901 bp fragment illustrates the mRNA expression of Plvap in the liver (+). cDNA from lung tissues served as positive control. RNA that was not reverse transcribed into cDNA (−) and a non-template reaction served as negative controls. No specific band was amplified from cDNA of Plvap -/- lung and liver. B, Real-time RT-PCR for Plvap with RNA from wild-type liver (n = 4) and lung (n = 3) at P17 to P18. GAPDH was used as a reference gene. C, Western blot analysis for PLVAP in proteins isolated from lungs of wild-type, Plvap+/- and Plvap-/- animals at P25.

Figure 3. Ultrastructure of _Plvap_-deficient sinusoids.

A, TEM of a 3-week-old wild-type animal shows numerous fenestrations in sinusoidal endothelial cells (black arrows). In contrast, fenestrations are very rare in the Plvap -/- littermate (black arrow). Lower panels show magnification. While in the wild-type animal Disse's space (Di) is open, it is compressed and filled with fine granular electron-dense material in the _Plvap_-deficient littermate (white arrows). He, hepatocyte, Lu, sinusoidal lumen. B, SEM of a 3-week-old wild-type animal shows the arrangement of fenestrations (arrows) in sieve plates. In contrast, sinusoidal endothelial cells of a _Plvap_-deficient littermate show only few fenestrations (arrows) which form individual openings. C, Quantitative analysis of fenestrations per µm endothelial length in sinusoids of wild-type and Plvap -/- animals. N = 5, mean ± SEM, ***p = 0.00069.

Figure 4. Diminished permeability of liver sinusoids in _Plvap_-deficient mice.

Neither by immunohistochemistry with antibodies against CD31 (A) nor by light microscopy of 1 µm semi-thin sections (B, Richardson's stain) obvious differences are detected with regards to the overall orientation and the density of liver sinusoids between 3-week-old Plvap -/- mice and wild-type littermates. Sinusoids of _Plvap_-deficient mice show a higher number of macrophages in their lumen (white arrows) and focal areas with accumulations of mononuclear cells in Disse's space (black arrows). Lower panels in B show higher magnifications. C, After perfusion of a wild-type animal with FITC-dextran, a strong FITC-signal (green) throughout the liver is detected. Immunolabeling with CD31 (red) suggests that FITC-dextran molecules have accumulated in the space of Disse. In contrast, in the _Plvap_-deficient littermate, the signal for FITC-dextran is much weaker and barely detectable. Nuclear DNA is labeled with DAPI (blue).

Figure 5. Hyperlipoproteinemia in _Plvap_-deficient mice.

A, Plasma of a P10 _Plvap_-deficient pup looks milky (arrows), while that from a wild-type littermate is clear. No obvious differences in hematocrit readings are observed between a Plvap -/- pup and its wild-type littermate (right-handed panel). B, The blood in the mesenterial or kidney vessels of a 3-week-old _Plvap_-deficient animal is milky (white arrows), while that of wild-type animals is red. Boxed area displays a magnification of mesenteric capillaries that are cut open to show milky blood (black arrow). Kidneys and livers of wild-type animals are of reddish color (black arrows), while those of _Plvap_-deficient littermates are pale.

Figure 6. Lipoproteins in capillaries and hepatocytes of _Plvap_-deficient animals.

A, PPD-labeled granular structures fill the lumen of kidney capillaries in a 3-week-old Plvap -/- mouse (arrows, boxed area, counterstain: Richardson's stain), while only erythrocytes are seen in the wild-type littermate (arrow, boxed area). B, The granular structures (black arrows) in the lumen of _Plvap_-deficient capillaries have diameters between 50 to 500 nm and are homogenously electron-dense. The particles are surrounded by extracellular fibrils with the ultrastructural characteristics of fibrin (white arrow) (TEM, lower panel shows higher magnification). C, PPD-labeled granular structures are seen in the lumen of liver sinusoids in a Plvap -/- mouse but not in a wild-type littermate (white arrows). The cytoplasm of the vast majority of Plvap -/- hepatocytes is densely filled with PPD-labeled granules (black arrows) which are rare in hepatocytes of wild-type littermates (black arrows). D, By TEM, numerous liposomes (black arrows) are seen in the _Plvap_-deficient liver indicating a pronounced microvascular steatosis. Hepatocytes of the wild-type animal contain large aggregates of glycogen (asterisks), which are absent in hepatocytes of the _Plvap_-deficient littermates. The lumen of Plvap -/- liver sinusoids is occluded by fibrillar fibrin (white arrows) suggestive of thrombus formation.

Figure 7. NMR spectroscopy of wild-type and _Plvap_-deficient plasma.

A, Part of 600 MHz 1H NMR spectra of wild-type and Plvap -/- mice at different gradient strengths displaying the (CH3) signals from the methyl groups of the lipids of lipoproteins with maximum gradient strength of 0.535 T/m ±0.01 T/m determined with a water reference sample, the different colors represent different gradient strengths. A strong decrease of the signal intensity corresponds to a small hydrodynamic radius. B, Average difference (black bars) of the relative particle numbers ΔNrel in the lipoprotein subclasses L15 to L122 between wild-type and Plvap -/- mice. The red bar indicates the signal intensity difference of the triplet signal visible in the Plvap -/- mice shown in A.

Figure 8. Liver fibrosis and necrosis in _Plvap_-deficient mice.

A, The surface of a wild-type liver at three weeks of age is smooth while that of a Plvap -/- mouse shows scars (white arrows). B, H&E stained paraffin section through the liver of a P17 Plvap -/- animal shows areas with microvesicular and macrovesicular steatosis (white arrow), hepatocyte ballooning (black arrow), and infiltrates of inflammatory cells. C, H&E stained paraffin section through the liver of a P18 Plvap -/- animal shows a focal necrotic area (white arrows). Similar areas are not seen in the wild-type littermate. D, Numerous TUNEL-positive nuclei (white arrows) accumulate in a focal necrotic area of the P18 Plvap -/- animal shown in C. In contrast, only individual TUNEL-positive cells are rarely seen in the wild-type littermate (black arrows). E, Immunostaining for IBA-1 (green) at P18 shows a higher number of macrophages (white arrows) in the Plvap -/- liver when compared to the wild-type littermate. F, G, Sirius red staining in the liver of an P17 Plvap -/- mouse indicates focal fibrosis (F, white arrows) and diffuse fibrosis around sinusoids (G, white arrows). In the wild-type littermate, positive staining for sirius red is only seen around larger hepatic vessels. H, Intense staining for PAS is seen in the cytoplasm of wild-type hepatocytes. In contrast, hepatocytes are not reactive for PAS in the liver of a Plvap -/- littermate with focal areas of positive extracellular PAS-labeling (black arrows). I, α-smooth muscle actin-positive cells accumulate in the liver parenchyma of a Plvap -/- mouse (white arrows), but are only seen in the walls of blood vessels in the wild-type littermate. J, A similar area as in G shows immunoreactivity for fibronectin in the liver of a Plvap -/- mouse (white arrows). Staining for fibronectin is essentially absent in the liver of the wild-type littermate. Nuclear DNA is labeled with DAPI (blue). K. Plasma analysis of wild-type (n = 13) and Plvap -/- mice (n = 9) (mean ± SD).

Figure 9. Accumulation of collagen in livers of _Plvap_-deficient mice.

By TEM of a _Plvap-_deficient animal at three weeks of age, bundles of collagen fibrils with the typical 67 nm periodicity are seen in Disse's space (arrows). No fibrillar extracellular matrix is present in Disse's space of the wild-type littermate. Lower panels show higher magnification.

References

1. Braet F, Wisse E (2002) Structural and functional aspects of liver sinusoidal endothelial cell fenestrae: a review. Comp Hepatol 1:1. - PMC - PubMed
1. McCuskey RS (2008) The hepatic microvascular system in health and its response to toxicants. Anat Rec (Hoboken) 291:661–671. - PubMed
1. Aird WC (2007) Phenotypic heterogeneity of the endothelium: II. Representative vascular beds. Circ Res 100:174–190. - PubMed
1. Wisse E, Braet F, Luo D, De Zanger R, Jans D, et al. (1996) Structure and function of sinusoidal lining cells in the liver. Toxicol Pathol 24:100–111. - PubMed
1. Wisse E, Jacobs F, Topal B, Frederik P, De Geest B (2008) The size of endothelial fenestrae in human liver sinusoids: implications for hepatocyte-directed gene transfer. Gene Ther 15:1193–1199. - PubMed

Formation of fenestrae in murine liver sinusoids depends on plasmalemma vesicle-associated protein and is required for lipoprotein passage - PubMed (original) (raw)