Cavin1 Deficiency Causes Disorder of Hepatic Glycogen Metabolism and Neonatal Death by Impacting Fenestrations in Liver Sinusoidal Endothelial Cells

Abstract It has been reported that Cavin1 deficiency causes lipodystrophy in both humans and mice by affecting lipid metabolism. The ablation of Cavin1 in rodents also causes a significant deviation from Mendelian ratio at weaning in a background‐dependent manner, suggesting the presence of undiscovered functions of Cavin1. In the current study, the results show that Cavin1 deficiency causes neonatal death in C57BL/6J mice by dampening the storage and mobilization of glycogen in the liver, which leads to lethal neonatal hypoglycemia. Further investigation by electron microscopy reveals that Cavin1 deficiency impairs the fenestration in liver sinusoidal endothelial cells (LSECs) and impacts the permeability of endothelial barrier in the liver. Mechanistically, Cavin1 deficiency inhibits the RhoA‐Rho‐associated protein kinase 2‐LIM domain kinase‐Cofilin signaling pathway and suppresses the dynamics of the cytoskeleton, and eventually causes the reduction of fenestrae in LSECs. In addition, the defect of fenestration in LSECs caused by Cavin1 deficiency can be rescued by treatment with the F‐actin depolymerization reagent latrunculin A. In summary, the current study reveals a novel function of Cavin1 on fenestrae formation in LSECs and liver glycogen metabolism, which provide an explanation for the neonatal death of Cavin1 null mice and a potential mechanism for metabolic disorders in patients with Cavin1 mutation.


Experimental Section
Mice and treatments:

Measurements of blood glucose and serum glucagon:
Blood was collected from tail tip of pups at 9 hours after birth or fetus at 24 hours before birth, and used to measure blood glucose with a blood glucose meter (SureStep Plus, Johnson). For serum glucagon measurement, blood from heart of pups or fetus was incubated at room temperature for 20 min and centrifuged at 626 g for 20 min. Serum was collected and used to detect serum glucagon level glucagon ELISA kit (Jiancheng, H183).

Histological analysis and immunofluorescence staining:
Liver was fixed in Carnoy's solution containing 60% methanol, 30% chloroform, 10% glacial acetic acid at 4 °C overnight. Subsequently, the samples were hydrated with 100%, 95% and for 10 minutes at room temperature. All images were taken by a laser-scanning confocal microscopy (Leica, Sp8). The area fraction of blood vessels per field in liver were quantitatively analyzed using Image J software (NIH Image, Bethesda, MD, USA).

Western blot:
Total protein of tissues was extracted using the buffer containing 1% sodium deoxycholate, 1%

Transmission and scanning electron microscopy (TEM, SEM):
The liver was cut into small pieces with a razor blade and fixed in 2.5% glutaraldehyde and then in 1% osmium. Dehydration was carried out by successive incubation with 50%, 75%, 80%, 95% and 100% ethanol. For TEM, the sample was further dehydrated with acetone twice and embedded in resin. The thin sections were stained with uranyl acetate and lead citrate for TEM observation. To explore the fenestrae of liver sinusoidal endothelial cells (LSECs) by SEM in vitro, primary LSECs were isolated and cultured as previously described [1] . For SEM, the samples after critical point drying were ion-sputtered using a carbon coater (Leica microscope system). The number of fenestrae per μm 2 in liver sinusoidal endothelial cells was quantitatively analyzed by Image J software (NIH Image, Bethesda, MD, USA).

Histamine or nicotine treatment:
8-week-old mice were fed for 12 hours, then intraperitoneally injected with histamine (40 mg/kg body weight) or nicotine (1.967 mg/kg body weight). Subsequently, these mice were fasted for 6 hours and sacrificed for detecting the level of liver glycogen. For glycogen storage experiments, 8 weeks old wild type C57BL/6J mice were fasted overnight and then injected histamine (40 mg/kg body weight) or nicotine (1.967 mg/kg body weight) or PBS control 1 hour prior to re-feeding. 6 hours after the re-feeding, the mice were sacrificed the hepatic glycogen in the liver were analyzed.

The induction of fenestrae formation by latrunculin A treatment:
The formation of fenestrae in the LSECs was induced by latrunculin A as previously reported [2] .
Briefly, isolated LSECs were cultured in RIPA1640 medium supplemented with 10% fetal bovine serum and 2 μg/ml vascular endothelial growth factor A (VEGFA). At 70% confluence, 250 ng/ml latrunculin A was added to RIPA1640 medium for 3 hours to induce fenestrae.
Glycogen synthase activity assay:

Glycogen phosphorylase activity assay
Liver was collected and weighed, then homogenized in 1 ml lysis buffer. Subsequently, glycogen phosphorylase activity was determined using glycogen phosphorylase activity assay kit (Solarbio, MS3601) according to the manufacturer's instructions. And the value of OD340 was measured every 5 minutes for 10 minutes. The relative glycogen phosphorylase activity was calculated and normalized by tissue weight.

Glucose uptake assay:
Mice were weighed and intraperitoneally injected with 2 g/kg body weight of 50% D-glucose which mixed 3 H-labeled glucose (PerkinElmer, Germany) at the volume ratio of 10:1. After 1 hour, hepatic glycogen was extracted by 10% trichloroacetic acid (TCA) solution and the radioactivity was detected by liquid scintillation counter (Beckman Coulter, Chaska, USA).

RhoA activity assay:
RhoA activity assay was performed using RhoA activity assay kit (NewEast, 80601) according to the manufacturer's instructions. Briefly, fresh liver was homogenized in lysis buffer. The lysates were centrifuged at 11800 g for 10 min at 4 °C. Equal amounts of protein was diluted to 1 ml of lysis buffer on ice, and incubated with anti-RhoA-GTP antibodies or normal IgG for 2 hours at 4 °C. Subsequently, the solution were incubated with protein A/G agarose beads for 1.5 hour at 4 ℃, and centrifuged at 3913 g for 1 min at 4 °C. The beads were rinsed by lysis buffer for three times, and eluted with Laemmeli's SDS buffer.

Insulin and glucagon signaling pathway assay:
Mice were fasted for 4 hours and then i.p. injected 10U/g bodyweight insulin or 10mg/kg bodyweight glucagon. 15 minutes after the injection, the mice were sacrificed and the livers wre collected for pAKT western blot and cAMP assay.

Endothelial permeability assay:
Mice were anesthetized and placed on the stage of the upright microscope. Open the abdominal cavity, find the portal vein, and intubate with the indwelling needle. Cut the IVC above the liver to make a gap. And clip the inferior vena cava to the lower part of the liver, so that the portal vein blood flow can only flow unidirectionally into the vein of the heart and out of the gap.
Then, infuse 2ml 0.5mg/ml dextran-FITC at the same rate. At this time, the fluorescent signal on the liver surface is imagined with a 20X water lens. Wash the dextran-FITC by infusing 10ml of PBS + 2mM EGTA solution at the same speed. After a few seconds, the dextran-FITC in the blood vessel will be washed away. At this time, the fluorescent signal were imagined again. The collected images were analyzed for fluorescence density using Image Pro Plus.
Single-cell RNA sequencing data analysis: Single cell data was obtained from (GSE129516). Seurat package were used for standardizing the data and for identification of highly variable features. Then, data was rescaled according to Primary hepatocyte culture and treatment: glucose DMEM with 10% FBS and antibiotics. Once attached with the plate, the cells were serum starved for 2 hours before treated with 100nM glucagon or 100nM insulin. After 6 hours treatment, the glycogens in the cells were measured and analyzed.

Statistical analysis:
The data were analyzed with GraphPad Prism 5.0 (GraphPad Software Inc, LaJolla, CA, USA) and expressed as the mean ± SD. Mann Whitney test was used for statistical comparisons between two groups. One-way analysis of variance (ANOVA) followed by post hoc Newman-Keuls test was used for statistical comparisons between multiple groups. Two-way ANOVA was used for grouped analysis. P values of < 0.05 were considered statistically significant.