Loss of endothelial ZEB2 in mice attenuates steatosis early during metabolic dysfunction-associated steatotic liver disease

Dheedene, Wouter; Verhulst, Stefaan; Demuynck, Louise; Callewaert, Bram; de Haan, Willeke; Vinckier, Stefan; Van Wauwe, Jore; Vandervoort, Petra; Lox, Marleen; Stroobants, Mathias; Lavend’homme, Renaud; van IJcken, Wilfred F. J.; Jones, Elizabeth A. V.; Zwijsen, An; Jacquemin, Marc; van Grunsven, Leo A.; Martinod, Kimberly; Huylebroeck, Danny; Mulugeta, Eskeatnaf; Luttun, Aernout

doi:10.1038/s41598-025-05881-6

Download PDF

Article
Open access
Published: 02 July 2025

Loss of endothelial ZEB2 in mice attenuates steatosis early during metabolic dysfunction-associated steatotic liver disease

Scientific Reports volume 15, Article number: 23434 (2025) Cite this article

1881 Accesses
1 Citations
Metrics details

Subjects

Abstract

Metabolic dysfunction-associated steatotic liver disease (MASLD), formerly known as non-alcoholic fatty liver disease, features liver sinusoidal endothelial cell (LSEC) alterations with ill-defined driving factors. Zinc-Finger E-Box-binding Homeobox (ZEB)2 in LSECs preserves their specialized features, prevents capillarization and protects against liver fibrosis. To investigate a potential protective role against steatosis, the initial MASLD stage, we fed EC-specific Zeb2 knockout (EC^Zeb2KO) mice a western-type diet (WD). In healthy and steatotic wild-type livers, Zeb2 was ubiquitously and similarly expressed across blood-vascular EC types. LSEC RNA sequencing revealed ZEB2 deficiency-triggered expression changes greatly overlapping with those evoked by WD-feeding. Endothelial ZEB2-loss and WD-feeding interacted to boost capillarization and fat metabolism, shown by increased expression of continuous EC markers and peroxisome proliferator-activated receptor (PPAR)α signaling components, respectively. Altered communication among LSECs after combined endothelial ZEB2-loss and WD-exposure revealed similar functional repercussions. Endothelial ZEB2-loss eventually corrected WD-induced liver hypo-vascularization while ameliorating hepatic damage and steatosis. Thus, endothelial ZEB2-loss amplifies WD-induced LSEC fat metabolism and capillarization, while decreasing steatosis, in part through altered LSEC-LSEC communication. The disease-promoting role of endothelial ZEB2 in early MASLD as opposed to its protective role in fibrosis underscores a context-dependent effect in liver disease.

Semaphorin-3A regulates liver sinusoidal endothelial cell porosity and promotes hepatic steatosis

Article Open access 14 June 2024

ACBP/DBI neutralization for the experimental treatment of fatty liver disease

Article Open access 16 November 2024

Deep proteome profiling of metabolic dysfunction-associated steatotic liver disease

Article Open access 03 March 2025

Introduction

Liver sinusoidal endothelial cells (LSECs) have special features adapted to the main liver functions, encompassing glucose and lipid storage/release, drug/toxin/pathogen clearance, and coagulation factor production^1,2,3,4. To support their central function in bi-directional exchange between blood and liver parenchyma, they have fenestrae organized in sieve plates, lack an organized basement membrane, and express scavenger receptors^1,5,6. This specialization is co-installed by LSEC-enriched transcription factors (TFs)^{2,7,8,9,10,11,12}. During liver disease, LSECs lose sinusoidal characteristics and acquire a capillary EC phenotype, a process called capillarization¹³. Recently, we showed that TF Zinc-finger E-Box-binding Homeobox (ZEB)2 is important for LSEC specialization, since its loss results in capillarization and increased liver fibrosis^2,7.

The most frequent form of chronic liver disease (CLD) is metabolic dysfunction-associated steatotic liver disease (MASLD; formerly known as non-alcoholic fatty liver disease), considered to be the hepatic manifestation of the metabolic syndrome^14,15,16. MASLD represents a clinical spectrum, from steatosis to metabolic dysfunction–associated steatohepatitis (MASH), often combined with advanced fibrosis. Its initial stage, i.e., steatosis, encompasses excess hepatic triglyceride storage. LSEC-engagement during MASLD is poorly understood, but interesting – often contrasting – hypotheses exist¹⁷. First, Hammoutene and Rautou hypothesized that LSEC capillarization hampers trans-endothelial passive lipid transport, thereby releasing the brake on lipogenesis in hepatocytes and increasing liver fat content³. Indeed, lack of fenestrae, a capillarization hallmark, resulted in steatosis in mice deficient for plasmalemma vesicle-associated protein¹⁸. Mice with endothelial loss of Semaphorin 3a or Group IVa phospholipase A2 (encoded by Pla2g4a) showed attenuated loss of fenestrae in LSECs along with decreased steatosis^19,20. Yet, loss of fenestrae has not always been observed in mouse models of MASLD²¹ and information in patients is scarce, with one study suggesting that loss of fenestrae would occur more prominently in early pre-MASH stages²². Second, single-cell RNA sequencing (scRNAseq) revealed that among organ ECs, capillary liver ECs were the second-most responsive to western-type diet (WD), after adipose tissue ECs, and the most responsive to diet reversal, suggesting a prominent role for LSECs during MASLD²³. Third, while proliferative and angiogenic liver ECs have been described during MASLD^23,24,25their impact on total liver vascularization and disease progression is unclear. On the other hand, dysfunctional ECs have been documented early during MASLD, before the onset of inflammation²⁶. Finally, because peroxisome proliferator-activated receptor (PPAR)α elimination in hepatocytes aggravates steatosis²⁷, PPAR(α) agonists were introduced as MASLD treatment²⁸. PPAR signaling has been implicated in functional preservation in the systemic and hepatic vasculature^29,30. However, it remains unknown how (much) these agonists affect LSEC behavior, since it was recently shown that PPAR(α) signaling is active in LSECs at baseline already and also upon lipid challenge^1,23,31,32. Additionally, the benefits of modifying LSEC behavior for MASLD treatment need to be explored.

Given our observations that endothelial ZEB2 protects against LSEC capillarization and hepatic fibrosis⁷we wondered whether ZEB2 would also have a protective role during steatosis. Therefore, we exposed mice to WD and investigated the consequences of EC-specific Zeb2 knockout for LSEC capillarization and steatosis.

Results

Zeb2 is ubiquitously expressed in endothelium of healthy and fatty livers

Our previous profiling analysis put forward ZEB2 as a TF ubiquitously expressed across the liver blood-vascular system and highly enriched in liver compared to heart and brain ECs^2,7. For a more complete view on endothelial Zeb2 expression, we expanded our analysis with extra-hepatic vascular beds from other metabolically demanding organs or organs with sinusoidal ECs and analyzed its expression in steatotic livers from patients with obesity or mice exposed to WD. Quantitative (q) real-time (RT)-PCR on ECs sorted from different organs of tamoxifen-inducible EC reporter mice newly generated for the current study by inter-crossing Cdh5-Cre^ERT2 mice with mice carrying the RCE reporter (EC^GFP)³³ (Supplementary Fig.S1) and data-mining of the scRNAseq dataset from Kalucka et al.³⁴ confirmed that liver endothelium is the highest Zeb2 expresser among other sinusoidal ECs or ECs from metabolically demanding organs (Fig. 1a,b). Analysis of our WD-fed mice as well as scRNAseq datasets on mouse and human livers from Guilliams et al.³⁵ and Remmerie et al.³⁶ confirmed that Zeb2 expression was ubiquitously detected in healthy liver endothelium, while WD-exposure/disease did not notably alter its expression (Fig. 1c; Supplementary Fig.S1). Altogether, endothelial Zeb2 expression in the liver vasculature was ubiquitous and not influenced by (early) MASLD.

EC ^Zeb2KO amplifies WD-induced LSEC capillarization early during MASLD

We previously showed that endothelial ZEB2 preserves specialized features in LSECs after a fibrotic challenge⁷. To study the effect of endothelial ZEB2-loss on capillarization during MASLD, we used our EC-specific Zeb2 knockout mice (EC^Zeb2KO) mice in which Zeb2 was specifically⁷ and to a comparable extent genetically inactivated in endothelium from healthy and steatotic mouse livers upon tamoxifen-treatment (Supplementary Fig.S1).

To induce MASLD, we exposed mice to WD, upon which WT mice (i.e., either tamoxifen-treated Cre-negative littermates (‘EC^WT ‘) or EC^GFP reporter mice; Supplementary Fig.S1; Supplementary NoteS1) progressively gained weight and developed steatosis from 4 weeks (w) onwards (Supplementary Fig.S1 + S2). After 24w of WD, steatosis was highly prominent while inflammation and parenchymal fibrosis were limited, suggesting that mice were still in the early MASLD stage, mainly featuring steatosis (Supplementary Fig.S2). Indeed, initial WD-induced increases in Ptprc and Adgre1 total liver mRNA levels were small and not reflected by the number of CD45⁺ inflammatory cells in situ (Supplementary Fig.S2). Inflammation and fibrosis parameters were not significantly altered by endothelial ZEB2 loss (Supplementary Fig.S2). Thus, our model was most suitable to study the role of endothelial ZEB2 in steatosis.

Capillarization signs were already manifest in this early stage, as evident from RNAseq on eGFP⁺ LSECs from EC^Zeb2KO and EC^GFP livers fed a standard diet (SD) or WD for 4w (Fig. 2a). Relative expression levels of landmark genes of continuous ECs and other liver cell-types demonstrated the purity and sinusoidal identity of the sorted LSEC populations (Supplementary Fig.S3). One sample from the EC^GFP/WD condition was omitted in subsequent analyses based on insufficient quality (Supplementary Fig.S3). All 12 mice from the EC^Zeb2KO/WD and EC^Zeb2KO/SD conditions showed successful exon 7 deletion (Supplementary Fig.S3).

Plotting the expression of a selected gene-set in a heatmap revealed an overall decreased LSEC signature gene expression upon endothelial ZEB2-loss, and WD-exposure alone had a partially similar effect. Simultaneously, EC^Zeb2KO increased continuous EC marker expression in LSECs, including Vwf. Endothelial ZEB2-loss and WD-exposure together boosted these effects (Fig. 2a). Accordingly, at 8w of diet, VWF antigen plasma levels were significantly increased by the combined challenge (105% vs. 58% of reference plasma pool levels; P < 0.05) compared to SD/EC^WT. Immunofluorescence staining for collagen-type IV revealed increased basement membrane formation triggered by WD, which was amplified upon combined challenge of EC^Zeb2KO and WD-feeding (Fig. 2b-f). QRT-PCR on whole-liver RNA revealed an increase in capillarization by the combined challenge, with gradual increase of capillarization marker Cd34 mRNA³⁷ from 8w of diet onwards (Fig. 2g). Transmission electron microscopy (TEM) at 8w revealed that fenestrae were still observed in all challenged conditions (Fig. 2h-k). Quantitative assessment of fenestrae revealed a downward trend in their number and size, resulting in a lower porosity upon endothelial ZEB2 loss and/or WD-exposure (Fig. 2l). Altogether, endothelial ZEB2-loss and WD-exposure cooperated in causing increased sinusoidal capillarization.

EC ^Zeb2KO and WD-exposure both induce LSEC proliferation early during MASLD

To gain more unbiased insight into changes caused by EC^Zeb2KO and/or WD-feeding, we performed three pairwise analyses revealing differentially expressed genes (DEGs; Fig. 3a; Supplementary Fig.S4). First, EC^Zeb2KO had a more dramatic and uniform effect on LSEC expression profiles than WD-exposure. Indeed, the EC^Zeb2KO vs. EC^GFP condition (comparison 1) had twice the number of DEGs and a clearer segregation on clustering heatmaps and principal component analysis plots compared to the WD vs. SD condition (comparison 2; Fig. 3a; Supplementary Figs.S3 + S4; Table 1). As expected from its predominant role as transcriptional repressor³⁸a majority of DEGs (65.5%) was upregulated by EC^Zeb2KO (Fig. 3a; Supplementary Fig.S4; Table 1). Similarly, WD-exposure mostly caused gene expression upregulation (58.7% of DEGs; Fig. 3a; Supplementary Fig.S4; Table 1). Remarkably, from the 523 DEGs elicited by WD-feeding in EC^GFP mice, a considerable number (41%) were also affected by EC^Zeb2KO upon SD-feeding (Fig. 3b; Supplementary Table 1).

Second, functional annotation on the combined up- and downregulated DEGs within comparison 1 revealed that the dominating biological processes/pathways altered by EC^Zeb2KO were related to proliferation (‘cell division’, ‘RHO-GTPase’ or ‘PLK1_pathway’; Fig. 3c; Supplementary Table 1). As expected by the large overlap in DEGs with the effect of EC^Zeb2KOthe dominating biological processes/pathways represented by DEGs caused by WD-exposure (comparison 2) were similarly related to proliferation (Fig. 3c; Supplementary Table 1). Accordingly, biological processes/pathways emerging from functional annotation analysis on genes commonly altered by EC^Zeb2KO and WD-exposure were equally related to proliferation (Fig. 3c; Supplementary Table 1). Separate functional annotation on the up- or downregulated DEGs revealed that the common proliferation-related terms emerged from the upregulated DEG sets, while the downregulated DEGs revealed some differences between genotype and diet effects, the former affecting ‘leukocyte migration’ and the latter ‘angiogenesis regulation’ (Supplementary Fig.S5 + Table 1). Thus, Zeb2 inactivation in ECs and WD-exposure both triggered proliferative gene expression in LSECs. The net effect of proliferation-related gene expression changes was an increase in proliferation confirmed in situ by the increased KI67⁺ cell-fraction within the eGFP⁺ LSEC population compared to the unchallenged condition. Altogether, EC^Zeb2KO and WD-feeding both provoked a proliferative response in LSECs, however, there was no additive effect of the combined challenge (Fig. 3d-h).

EC ^Zeb2KO and WD-exposure interact in affecting endothelial lipid metabolism

After we showed that the combination of EC^Zeb2KO and WD-feeding boosted gene expression changes reflecting LSEC capillarization, we investigated the interaction between both challenges in a more unbiased way by performing a whole-genome pairwise comparison (comparison 3), i.e. the combined challenge vs. no challenge. The resulting DEG list contained 1,640 genes, 70.5% of which were upregulated upon combined challenge (Fig. 3a; Supplementary Fig.S4; Table 1). Notably, overlap between the compiled DEG list from the single challenges with that of the combined challenge was only ∼45% (737 common DEGs; Fig. 4a; Supplementary Table 1). In accordance with the common effect of each challenge on cell proliferation (Fig. 3b-h; Supplementary Fig.S5), the dominating annotated biological processes and pathways of the 737 common up- and downregulated DEGs were related to ’cell cycle’ and ‘PLK1_pathway/RHO GTPase’, respectively (Fig. 4a; Supplementary Table 1). On the other hand, EC^Zeb2KO and WD-feeding interacted to induce a substantial amount of unique expression changes (895 unique DEGs; Fig. 4a; Supplementary Table 1). Functional annotation on the combined up- and downregulated unique DEGs revealed biological processes and pathways involving ‘(lipid) metabolism’ and ‘PPAR_signaling_pathway’ (Fig. 4a; Supplementary Table 1). Separate functional annotation on the up- or downregulated DEGs revealed that the common proliferation-related terms and the unique metabolism/PPAR-related terms emerged from the upregulated DEG sets, while the downregulated DEGs were too few to retrieve a sufficient number of significant terms (Supplementary Fig.S6 + Table 1). The combined challenge substantially extended the common effect on ‘PPAR signaling’. From the 26 DEGs related to ‘PPAR signaling’, 17 were uniquely induced by the combined challenge (Fig. 4b; Supplementary Table 1). The 26 common and unique genes together widely represented biological processes known to be regulated by PPAR signaling, including lipogenesis, lipid/fatty acid (FA) transport, FA oxidation, ketogenesis and cholesterol metabolism (Fig. 4c)³⁹. Accordingly, FA binding protein (FABP)4 protein expression was upregulated in LSECs of WD-exposed livers, most notably those simultaneously lacking endothelial ZEB2-expression (Fig. 4d-h). Altogether, EC^Zeb2KO and WD-challenge significantly interacted in boosting LSEC lipid metabolism.

EC ^Zeb2KO and WD-exposure interact in altering LSEC-LSEC communication during MASLD

LSEC communication with other liver cell-types is an important determinant of MASLD development⁴⁰. Therefore, by overlap with a compiled list of 2,356 genes encoding highly likely secreted proteins⁴¹we first identified within our DEG lists ~ 13% genes encoding proteins that potentially interact with neighboring cells (Fig. 5a; Supplementary Table 2). WD-exposure and ZEB2-loss in ECs triggered changes in the expression of these genes, with significant interaction between both challenges, similar to the overall gene expression changes (Figs. 3a and 5a). Indeed, 93 genes were commonly regulated by at least one of the challenges and the combined challenge, while about 100 genes were uniquely regulated by the combined challenge (Supplementary Table 2). Thus, like for the whole gene-set, EC^Zeb2KO and WD-challenge significantly interacted in determining expression alterations of genes encoding secreted products.

To identify the preferential target cell-type(s) of LSEC-secreted products during MASLD, we used Multi-NicheNet to predict communication activity using available scRNAseq data of mice exposed to SD or WD (Supplementary Fig.S7)³⁵. We considered both WD-induced gain and loss in communication by looking at up- or downregulated ligands in LSECs, respectively, for which the predicted receptors are expressed in the receiver cell. Collectively, WD-exposure had a substantial effect on LSEC-LSEC and LSEC-Kupffer cell communication compared to LSEC communication with hepatocytes or hepatic stellate cells. This was evident from a high number of lost or gained ligand-receptor interactions and resulting altered target genes (Supplementary Fig.S7).

We next zoomed in on LSEC-LSEC communication and used NicheNet on our own dataset to filter out those ligands that likely give rise to functional communication activity based on corresponding receptor and altered target gene expression in receiving LSECs (Supplementary Fig.S8). Upon Zeb2 genetic inactivation, WD-exposure, or the combination of both, several ligands were predicted to give rise to altered target expression in LSECs (Fig. 5b). As above, we evaluated gain or loss in communication by screening for ligands up- or downregulated, respectively, by the challenge. Again, there was substantial overlap in predicted active ligands between single and combined challenges (Fig. 5b ‘common’; Supplementary Fig.S8). However, 11 out of 20 ligand alterations in communication were uniquely invoked by the combined challenge (Fig. 5b). Eighteen out of 20 ligands were upregulated by the combined challenge and gave rise to a gain in communication (Fig. 5c). Unique and common ligands together were linked to 83 significantly altered target genes (Fig. 5c; Supplementary Fig.S8). Fifteen targets, including PPARα signaling pathway members Ppara and Apoc3 and vasoconstrictor endothelin-1 (Edn1) were triggered by minimally 5 ligands.

Functional annotation of all 20 ligands and 83 associated predicted targets revealed biological processes/pathways, like terms related to extracellular matrix (ECM) and metabolism/PPAR (Fig. 5d; Supplementary Table 2). When splitting up the annotation between gained and lost communications, the metabolism/PPAR terms were represented in the gained communications, while on the lost communications side other terms emerged, including ‘ECM/collagen synthesis/degradation’, ‘leukocyte migration’ and ‘erythrocyte aggregation’ (Supplementary Fig.S9 + Table 2). Altogether, this suggests that some processes, such as metabolism and PPAR(α) signaling, identified earlier within our RNAseq dataset, were caused at least partially by altered LSEC-LSEC communication.

EC ^Zeb2KO rescues WD-induced liver damage, steatosis and hypo-vascularization

We subsequently wondered whether changes in endothelial expression and LSEC-LSEC communication triggered by EC^Zeb2KO and WD-exposure had repercussions for steatosis progression and liver vascularization. In accordance with previous observations⁴², WD-exposure significantly increased cholesterol and downregulated triglycerides in plasma (Supplementary Fig.S10). While WD-feeding invoked hepatic triglyceride accumulation in both genotypes, 24w of WD-feeding caused a significantly higher increase in WT livers (Fig. 6a). In addition, WD-induced body weight was significantly lower in mice lacking endothelial ZEB2 by 24w (Fig. 6b). Furthermore, liver parenchyma integrity was compromized after 24w of WD-exposure, as evident by increased plasma liver enzymes (Fig. 6c,d). In accordance with the hepatic triglyceride content, liver damage and hepatocyte hypertrophy were also attenuated in livers lacking endothelial ZEB2 (Fig. 6c–e; Supplementary Fig.S10).

While it has been hypothesized that capillarization triggers a compensatory increase in hepatic lipogenesis³the increased capillarization upon endothelial ZEB2-loss did not provoke more hepatic lipid production. Indeed, 4w of WD similarly induced whole liver expression of lipogenesis regulator Srebf1. At 24w, WD-induced levels of genes encoding lipogenesis enzymes Fasn and Acaca were not significantly affected by ZEB2-loss (Supplementary Fig.S10).

To evaluate in a representative and global manner whether and how ZEB-2 loss and/or WD-exposure affect liver vascularization, we used Q-VAT (Supplementary Fig.S11)⁴³. While WD-exposure did not significantly alter the capillary/hepatocyte ratio, vascular branch density per parenchymal area was significantly decreased at 24w of WD in accordance with the increased hepatocyte size in WT livers (Fig. 6e,f). This hypo-vascularization suggests that increased proliferation upon 4w of WD-exposure was not sufficient to compensate for the expanding liver (Figs. 3d-h and 6e,f). In contrast, in EC^Zeb2KO livers, vascular branch density per parenchymal area was partially restored, again in accordance with the attenuated WD-induced hepatocyte hypertrophy (Fig. 6e,f). Altogether, endothelial ZEB2-loss ameliorated WD-induced liver steatosis, damage and hypo-vascularization.

Discussion

MASLD is becoming the most prevalent CLD due to the ongoing obesity epidemic. Since fibrosis is the etiology-independent result of CLD, most studies, including those related to MASLD, have focused on understanding the mechanisms that orchestrate excessive collagen deposition⁴⁴. Nevertheless, before the challenge evokes a fibrotic response, there are other connected pathogenic mechanisms. Understanding initial events is important not only for early therapeutic intervention, but also to diagnose MASLD before reaching advanced stages. In MASLD, excessive lipid accumulation initiates hepatocyte damage.

Here, we identified endothelial ZEB2 as a multi-factorial instigator of steatosis, thereby down-tuning PPAR(α) signaling, partly through altered communication among LSECs (Fig. 7). Because endothelial ZEB2-loss also hampered the WD-induced body weight increase, we cannot exclude an effect of endothelial ZEB2 deficiency on whole body metabolism and/or calorie intake. Furthermore, since ECs from certain extrahepatic lipid-handling organs (such as WAT) also express low ZEB2 levels, we cannot rule out indirect effects on hepatic lipid handling. Although WD-induced adipocyte hypertrophy occurred to a similar extent in both genotypes (Supplementary Fig.S12), excluding an indirect role for ZEB2 in adipose tissue ECs would require a more in-depth assessment of adipose tissue (metabolic) function. The second-highest expression level for endothelial ZEB2 was found in the lung where its loss may potentially affect lung function (Fig. 1a). A recent study found an inverse correlation between lung function and the risk of developing MASLD⁴⁵. Nevertheless, to exclude the involvement of extrahepatic endothelial ZEB2, an approach where ZEB2 is specifically deleted in liver ECs is required. Despite endothelial ZEB2’s multi-factorial effect on steatosis, short- or long-term WD-exposure did not notably alter Zeb2 expression in liver ECs. In contrast, GATA4, another TF enriched in LSECs, did not affect hepatocyte triglyceride content in a MASLD model⁴⁶. Although endothelial ZEB2 counteracted capillarization during both steatosis and fibrosis⁷it was only protective against fibrosis, partly through altered communication with hepatic stellate cells. Thus, TFs involved in LSEC differentiation have different effects on the development of steatosis, and ZEB2 has a challenge-type dependent role in liver disease.

Capillarization has been causally linked to fibrosis in several liver disease models, in part through allowing hepatic stellate cell activation⁴⁴. A causal role for LSEC capillarization in MASLD has also been suggested in studies showing that it occurs before the onset of steatosis⁴² as well as in studies where changes in LSEC fenestration were linked to MASLD pathogenesis upon genetic interventions in LSECs^19,20,47. Moreover, while Bondareva and colleagues observed ∼500 DEGs in LSECs after 3 months of WD-exposure, we found a similar amount of DEGs already after 4w, with largely similar functional annotation, again suggesting a role for LSECs in early MASLD²³. Hammoutene and Rautou hypothesized that LSEC capillarization causes steatosis by hampering passive lipid transport and therefore inducing endogenous lipogenesis in hepatocytes³. Here, the WD-induced increase in capillarization indeed induced lipogenesis regulator Srebf1. However, the strongest capillarization occurred in the absence of endothelial ZEB2, a condition with decreased steatosis compared to WT. This suggests that additional phenomena determine the degree of steatosis induced by WD-feeding. We therefore propose to fine-tune the abovementioned hypothesis by stating that capillarization only partly determines the amount of steatosis. In our mouse model, we saw signs of capillarization through a prominent gene signature shift and basement membrane deposition while TEM revealed that fenestrae tended to be smaller and less numerous after 8w of endothelial ZEB2 loss and/or WD-exposure. One of the genes that based on the abovementioned genetic intervention studies in LSECs was proposed to be linked with decreased LSEC fenestration²⁰i.e., Pla2g4a, was significantly upregulated upon the combined challenge with WD and EC^Zeb2KO (Supplementary Table 1). As the number of mice available for TEM analysis was limited, in order to obtain a more definitive picture on LSEC fenestration in our mouse model, complementary scanning electron microscopic analysis, including at later stages, is needed. We hypothesize that passive diffusion of lipids through the fenestrated LSECs may be complemented with active transport mechanisms and that the latter were amplified by endothelial ZEB2-loss (Fig. 7). In support, we found that endothelial ZEB2-loss in WD-exposed mice prominently activated PPARα-signaling, regulating lipid transport and metabolism, in LSECs. This implies that LSECs might have a ‘buffering’ role that protects against lipotoxicity in the parenchyma by actively taking up and metabolizing lipids. Such a role has been previously suggested for non-fenestrated ECs^48,49,50. LSECs may be especially proficient in this task by expression of many receptors active in lipid transport^12,51,52. Additional functional support for altered sinusoidal lipid transport and metabolism upon endothelial ZEB2-loss requires additional testing.

The majority of hepatocytic lipids in patients with MASLD originate from peripheral FAs^53,54 and LSECs are capable of storing and degrading lipids following exposure to free FAs^55,56. This Ppar(α)-driven role in lipid homeostasis in LSECs also implies that Ppar(α) agonists that have been proposed for MASLD treatment because of their role in hepatocytes³⁹may offer additional benefit through their action on liver endothelium. In addition to its effect on EC metabolism, PPAR signaling has anti-inflammatory effects, also desired during treatment of MASH³⁹. Like in other studies⁵⁷the degree of inflammation and fibrosis was limited in our model, precluding us to evaluate whether the increased PPAR signaling upon endothelial ZEB2-loss protected against transition to MASH and fibrosis.

One of the prominent common responses of LSECs to ZEB2-loss and WD-exposure was cell proliferation (Fig. 7). Others, who interpret this response as a possible compensatory mechanism because of liver expansion, also reported proliferation of LSECs as response to WD^23,36. While the increased proliferation tended to increase vascular branch density after 24w in mice on a SD, simultaneous hepatocyte hypertrophy in mice fed a WD overruled this increase, resulting in a significant decrease in vascular branch density per parenchymal area. While the number of vascular branches per hepatocyte was unaltered upon WD exposure, the hepatocyte hypertrophy likely requires an increased ratio to cope with higher energy and oxygen demands. Hence, during steatosis, the hepatic parenchyma is hypo-vascularized, potentially co-contributing to hepatocyte damage and representing a hypoxic trigger for angiogenesis seen during transition to MASH⁵⁸. Whereas angiogenesis may drive the progression of MASLD from the transition to MASH onwards²⁴, it may not directly promote steatosis. Indeed, blocking angiogenesis by antagonizing Angiopoietin-2 was shown to reduce hepatocyte ballooning and fibrosis, but did not affect steatosis⁵⁹. In our previous studies, a more short-term ZEB2-loss in ECs caused expansion of the liver vasculature in the absence of increased proliferation, reflecting intussusceptive angiogenesis⁷. This indicates that genetic inactivation of endothelial Zeb2 causes both proliferative and non-proliferative forms of angiogenesis. Altogether, angiogenesis is stage-dependently involved in MASLD progression.

We previously showed that endothelial ZEB2-loss both causes expression changes in LSECs as well as non-LSECs, mainly hepatic stellate cells⁷. Also in the brain and the heart, ZEB2-loss in neurons and cardiomyocytes, respectively, caused secondary expression changes in other cell-types^60,61. Here, endothelial ZEB2-loss had a significant impact on communication among LSECs (Fig. 7). Interestingly, while PPARα signaling genes were among the most notable ones upregulated by combined WD and Zeb2 knockout challenges, PPARα itself was also one of the important targets of LSEC-LSEC communication. Another intriguing ligand and target of LSEC-LSEC communication was vasoconstrictor Edn1, which was upregulated in response to 12 predicted ligand-receptor interactions (Fig. 5c; Supplementary Fig.S8). The most likely receptor on LSECs predicted for communication with LSECs through Edn1 was endothelin receptor B (Ednrb; Supplementary Fig.S8), whose interaction in human umbilical vein ECs was shown to promote proliferation⁶². Increases in Endothelin-1 during steatosis have also been reported in a rat model, where its vasoconstrictive effect – most likely through interaction with its endothelin receptor A on hepatic stellate cells – was responsible for increased intrahepatic vascular resistance causing impaired blood flow^26,63. The most prominent predicted target of the LSEC-LSEC communication was proliferating cell nuclear antigen (PCNA) clamp-associated factor (Pclaf; Supplementary Fig.S8), which has been associated with cell cycle progression in multiple tumors, including neuroblastoma, through its interaction with PCNA^64,65. Hence, the increased LSEC proliferation may also be the result of altered LSEC-LSEC communication. Altogether, endothelial ZEB2-loss has an etiology-dependent effect on cell-cell communication in the liver.

Currently, weight loss is the main pragmatic and therapeutic option for patients with MASLD, but life-long compliance has proven to be problematic, invoking the need for drug development. Recently, a first anti-MASH drug (resmetirom) has been approved by the U.S. Food and Drug Administration and several phase 2/3 trials are ongoing (clinicaltrials.gov), including with aspirin and PPAR agonists⁶⁶. While hepatocytes have been mostly the target of such drugs, LSECs are another highly appealing target given their multi-factorial involvement from early stages of MASLD onwards. To uncover which pathways are critical for the observed phenotypes, EC-specific expression modulation of individual factors, like PPARα, would be required. In addition, we conclude that endothelial ZEB2 has a stage-dependent effect on liver disease, which will necessitate customized approaches when considering ZEB2 as a target for therapy.

Methods

Mice and MASLD model

Mouse experiments were approved by the KU Leuven Animal Ethics Committee (Ethics Committee Dossier 169/2014, 208/2017 and 121/2019) and were performed in accordance with the Committee’s guidelines and those from directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes and the ARRIVE guidelines. Mice were housed in filter top cages with wood bedding and cocoons as enrichment under standard conditions with 10/14-hour light/dark cycles. Mice had ad libitum access to SD (Rat/Mouse – Maintenance, Ssniff) or WD (high-fat/high sucrose/high cholesterol; TD.88137, Ssniff; the full composition of the diet is given in Supplementary Table S3) and water.

All interventions were done during the light cycle. Male mice (known to be more susceptible to diet-induced steatosis⁶⁷) were used that were 5 weeks of age at the start of the experiment. Three different mouse lines were used: 1° tamoxifen-inducible EC-specific Zeb2 knockout mice (EC^Zeb2KO) generated by inter-crossing the tamoxifen-inducible Cdh5-Cre^ERT2 driver line⁶⁸ with Zeb2^fl/fl mice (on a 129 Sv/CD1 background) carrying a Zeb2 exon 7 flanked by loxP sites^7,69; 2° tamoxifen-treated Cre-negative littermates of EC^Zeb2KO mice (EC^WT), used as WT control for phenotypic analysis⁷; 3° tamoxifen-inducible EC reporter mice newly generated for the current study by inter-crossing Cdh5-Cre^ERT2 mice with mice carrying the RCE reporter (EC^GFP)³³used as Cre-controls for RNA profiling of sorted eGFP⁺ liver ECs, for comparative analysis of Zeb2 expression in different organs and for part of the phenotypic analysis (Supplementary Fig.S1 + NoteS1). To induce efficient Cre-mediated recombination and Zeb2 deletion in ECs, mice were intraperitoneally (i.p.) injected with 1 mg tamoxifen (Sigma) in 100 µL sunflower oil (Sigma) for 5 consecutive days (Supplementary Fig.S1). Because the sorted EC population used for expression profiling mainly (> 99%) represents LSECs, we consider that expression changes detected upon Cre-mediated Zeb2 deletion largely reflect those in LSECs. We therefore, where appropriate, labeled the EC population in the methods, results and discussion sections and corresponding figures as ‘LSECs’ rather than ‘ECs’. Three weeks after the first tamoxifen injection, mice were randomized to SD or WD and were fed this diet for a period of 4w, 8w or 24w (Supplementary Fig.S1). Blood was drawn via the heart using a heparin-coated syringe and mice were euthanized by exsanguination under sodium pentobarbital (Dolethal^®; 66.7 µg/g i.p.) anesthesia. Anesthesia depth was checked by toe pinch and, if necessary, mice received another i.p. injection with 22.2 µg/g sodium pentobarbital.

For immunohistological, weight and/or plasma analyses, mice were euthanized after 4w, 8w or 24w of diet under anesthesia as described above. Livers were subsequently perfusion-fixed with zinc-formalin (Sigma), followed by overnight immersion fixation with zinc-formalin. The left lateral lobe was processed for paraffin sectioning, the left medial lobe was kept overnight in 25% Tris-buffered saline (TBS)-sucrose, before embedding in TissueTek. Liver and WAT paraffin blocks were sectioned at 7 and 8 μm thickness, respectively. Liver cryosections were sectioned at 10 μm thickness. For cell isolation and gene profiling, weight, plasma and part of the immunohistological analyses, mice were euthanized under anesthesia (as described above) after 4w of diet and after a 4-hour starvation period, to reduce biological variation. The right medial and lateral lobe were dissected and immersion-fixed overnight with zinc-formalin before processing for paraffin- and cryo-sectioning as described above. The caudate lobe was isolated, snap-frozen and stored at -80 °C until further use for liver triglyceride determination and RNA isolation. For liver triglyceride determination, a lipid extraction was performed as described in de Haan et al.⁷⁰ followed by applying a colorimetric triglyceride detection kit (MAK266, Sigma) according to the manufacturer’s instructions.

Cell isolation and gene expression analysis

To study tissue-specific endothelial Zeb2 expression, eGFP⁺ ECs were isolated from EC^GFP mouse livers, hearts, brains, visceral WAT, muscles, spleens, bone marrow and intestines and comparative gene expression was performed by qRT-PCR. To isolate LSECs, EC^Zeb2KO and EC^GFP (both Cre-positive) mice were perfused via the portal vein with 2 mL digestion buffer (1.2 U/mL dispase in Gey’s buffer). The liver was excised and incubated at 37 °C for 10 min followed by 10 min at room temperature in the same digestion buffer (see Supplementary Table 4). The suspension was shaken manually every 5 min, filtered through a 70 μm strainer and centrifuged for 5 min at 50 g to separate the hepatocytes from the non-parenchymal cell fraction. Hepatocytes were immediately snap-frozen after centrifugation and stored. To isolate LSECs from the non-parenchymal cell fraction, the hepatocyte-free suspension was used for Percoll density gradient centrifugation. After pelleting the cells, red blood cells were removed from the non-parenchymal cell fraction by incubation with erythrocyte lysis buffer (0.15 M NH₄Cl; 10 mM KHCO₃; 0.1 mM EDTA) for 5 min, followed by washing in phosphate-buffered saline (PBS). The resulting cell pellet was resuspended in fluorescence-activated cell sorting (FACS) buffer (2 mM EDTA; 25 mM HEPES pH7; 0.2% bovine serum albumin (BSA; in PBS) containing 7-aminoactinomycin D (7-AAD) as live/dead marker and filtered through a 40 μm mesh before cell sorting of 7AAD⁻/UV⁻/eGFP⁺ LSECs on a FACS Aria-III sorter.

To isolate other tissue-specific ECs, tissues were dissected, minced and digested in a customized digestion buffer (Supplementary Table 4). Heart, skeletal muscle, lung, kidney, adipose and spleen tissue digests were filtered through a 70 μm mesh in PBS containing 2% fetal bovine serum (FBS). After pelleting the cells for 7 min at 600 g, erythrocyte lysis was performed for 2 min before washing with PBS. Next, cells were resuspended in FACS buffer and filtered in a FACS tube. To isolate bone marrow ECs, bones were flushed with 10 mL Hank’s balanced salt solution using a 23–26 G syringe. Then, the cell suspension was processed like the tissue digests described above. To isolate brain ECs, the tissue homogenate was spun down in 2% FBS at 600 g for 7 min before resuspending the cells in DMEM with 20% BSA. After centrifuging at 600 g for 10 min, the pellet was resuspended in 0.05% trypsin at 37 °C for 2 min. Then PBS containing 2% FBS was added and the suspension was filtered through a 70 μm cell strainer. After pelleting the cells for 7 min at 600 g, erythrocyte lysis was performed for 2 min before washing with PBS. Next, cells were resuspended in FACS buffer and filtered through a 40 μm mesh before sorting. FACS data were analyzed with FACS DIVA software.

RNA was isolated from sorted ECs using Reliaprep (Promega) for RNA sequencing or qRT-PCR purposes. For RNA sequencing, RNA samples from EC^GFP and EC^Zeb2KO mice (n = 6 per diet and genotype) were quality-controled and prepared according to the Smart-seq2 method⁷¹. In brief, polyA + RNA was reverse transcribed using an oligo(dT) primer. Template switching by reverse transcriptase was achieved using an LNA-containing TSO oligo. The reverse transcribed cDNA was pre-amplified with primers for 18 cycles followed by clean-up. Tagmentation was performed on 500 pg of the pre-amplified cDNA with Tn5 followed by gap repair. The tagmented library was extended with Illumina adaptor sequences by PCR for 14 cycles and purified. The resulting sequencing library was measured on a Bioanalyzer and equimolar loaded onto a flow cell and sequenced according to the Illumina TruSeq v3 protocol on the HiSeq2500 with a single read 50 bp and dual 9 bp indices. Illumina adapter sequences and poly-A stretches were trimmed from the reads⁷². The remaining sequences were aligned to the mouse GRCm38 reference sequence using HISAT2 (version 2.1.0)⁷³. Transcript abundance level (transcript count) was generated using HTSeq (version 0.9.1) based on the ENSEMBL 84 gene annotation⁷⁴. The transcript counts were further processed using R software environment for statistical computing and graphics (version 3.4.0). Data normalization, removal of batch effect and other variance was performed using EDASeq R and RUVSeq packages^75,76. Differential expression analysis was performed using edgeR (R package)⁷⁷applying the negative binomial general linear model (GLM) approach. DEGs with false discovery rate (FDR) < 0.05, Benjamini-Hochberg multiple testing correction and expression level in WT samples of > 1 counts per million (CPM) were retained and used for further processing (i.e., principal component analysis, cluster analysis, Volcano plots, heatmaps, Gene Enrichment analysis by Toppfun (https://toppgene.cchmc.org/enrichment.jsp) and ligand-target prediction analysis by NicheNet (using an open source R implementation; https://github.com/saeyslab/nichenetr)⁷⁸. The full code of the analysis is shown in a Supplementary NoteS2. To visualize overlap between DEG lists from each comparison, circos plots were generated using Metascape⁷⁹. To analyze changes in predicted interactions between LSEC-derived ligands and targets in LSECs, V-Set And Immunoglobulin Domain Containing (Vsig)4⁺ Kupffer cells, hepatic stellate cells and hepatocytes from mice challenged with 24/36 weeks of WD, a Multi-NicheNet pipeline was applied on the scRNAseq data generated by Guilliams et al.³⁵ (GSE192742). DEGs were determined based on a log(2)fold-change threshold of (-)0.25 and a P-value of less than 0.05. To integrate the CD45⁺ (including the Vsig4⁺ Kupffer cells) and CD45⁻ datasets, batch correction was performed in Multi-NicheNet, and selection of receptors relied solely on expression values and the top 100 interactions were extracted from the analysis. To obtain RNA from whole liver tissue, the liver was homogenized in TRIZol and RNA was isolated according to manufacturer’s instructions. For quantitative qRT-PCR, cDNA was made using the GoScript reverse transcription system (Promega) according to the manufacturer’s protocol. QRT-PCR was performed with the ABI system using Sybr green (ABI). Gapdh was used as reference gene after validation of its stability across the tested conditions through comparison with two other housekeeping genes (Hprt1 and Ywhaz), as previously described⁷; for primer sequences, see Supplementary Table 4).

Histological and Immunofluorescence staining and imaging

To assess liver fibrosis, paraffin sections were stained with Sirius red to reveal collagen. Hepatocyte sizes were determined by dividing tissue area by the manually counted number of hepatocytes (at least 800 per sample) on images of liver cryosections either stained with Rhodamine-Phalloidine (R415, ThermoFisher) to mark hepatocyte cell borders and Hoechst to label nuclei or with H&E. Hepatocyte sizes were corrected for polyploidy, as previously described^80,81. Adipocyte sizes were measured by manually selecting a representative amount of cells on H&E-stained paraffin sections visualized by using a 40x objective on a brightfield microscope. Cell sizes and the amount of collagen were quantified using Image J. To evaluate inflammation throughout the MASLD model, paraffin sections were stained for CD45 and CD45-positive cells were counted in pericentral and periportal areas. To characterize LSEC capillarization and PPARα signaling target expression, immunofluorescence stainings were performed for collagen-type IV and FABP4, respectively. To assess LSEC proliferation, paraffin sections were double-stained for KI67 and eGFP. Antibodies were tested using a negative control (no primary antibody) and WT healthy mouse liver sections as positive control where we compare the expression pattern with stainings found on the manufacturers’ websites and in literature; for antibodies and concentrations see Supplementary Table 4. Where necessary, amplification was performed using a Cy3- or fluorescein-tyramide kit according to the manufacturer’s instructions (Perkin Elmer). Microscopy images were analyzed in Image J. To take brightfield or regular fluorescence microscopic images, a Leica upright brightfield or an upright Zeiss fluorescence microscope equipped with Axiovision imaging software were used, respectively. For confocal microscopy, a Zeiss confocal microscope was used (see Figure legends for more details).

Vascular area analysis by quantitative vascular analysis tool

To determine branch density of the liver blood vasculature, paraffin sections were stained for ENDOGLIN. Sections were integrally imaged at high spatial resolution (i.e., 0.32175 μm/px) using a Nikon NiE-Marzhauser Slide Express 2 equipped with a Hamamatsu Orca Flash 4.0 camera (VIB-CBD). The vascular network of the entire sections was quantified in a tile-wise manner using the in-house developed Quantitative Vascular Analysis Tool (Q-VAT; Supplementary Fig.S11)⁴³. The ENDOGLIN-stained fluorescent slide scanner images were pre-processed using the Q-VAT masking tool to generate a tissue mask for normalization and a binary vascular mask from the original immuno-stained images. These masks were partitioned into tiles with the same size as the original acquisition tiles and the masks were validated by overlaying them with the original immunofluorescence image. The following input parameters were used in the Q-VAT masking tool: Calibration, 0.32175 μm/px; Radius of biggest object, 16 μm; Particle size lower range, 1,000 μm²; Radius for median filtering, 15 μm and Thresholding method according to Huang et al.⁸². Automated analysis and quantification of the segmented vasculature was subsequently performed on each tile using Q-VAT with the following input parameters: Calibration, 0.32175 μm/px; Vascular compartment separation threshold, 10 μm; Close label radius, 3 μm; Prune ends threshold, 5 μm. The vessel diameter and branch density were calculated for each acquisition tile of the entire cross-sectional liver area. Tiles with aberrant values were excluded using a stationary cut-off, and outliers were removed using Robust regression and Outlier removal (ROUT) method with a ROUT coefficient Q = 1%. After outlier removal, the average vessel diameter and branch density were calculated for the entire cross-sectional area of the liver. Group-wise comparison was performed using two-way ANOVA with Tukey post-hoc test for multiple comparison. The Q-VAT tool can be downloaded, together with a more detailed user guide on GitHub (https://github.com/bramcal/Q-VAT.git).

Published single-cell RNA sequencing dataset data-mining and re-analyses

The online portal from Kalucka et al.³⁴ (https://endotheliomics.shinyapps.io/ec_atlas/; GSE99235 ref⁸³; ArrayExpress: E-MTAB-8077) was used to evaluate inter-organ heterogeneity of adult mouse endothelial Zeb2 expression. The online portal linked to Guilliams et al.³⁵ (https://www.livercellatlas.org/datasets_diet.php; GSE192742) was used to evaluate heterogeneity in Zeb2 expression in livers from lean and obese mice. ZEB2 expression in healthy livers and livers from patients with MASLD was studied using the online portal from Guilliams et al.³⁵ (https://www.livercellatlas.org/datasets_diet.php; GSE192742). The online portal described in Remmerie et al.³⁶ (https://www.livercellatlas.org/datasets_diet.php; GSE156059) was used to evaluate Zeb2 expression in liver cell clusters from mice challenged with SD or WD for 24w or 36w.

TEM

TEM was performed to document the fenestration size and frequency, and calculate porosity of the liver sinusoids in vivo. Hereto, mice were perfusion-fixed with 2.5% glutaraldehyde in 0.05 M sodium-cacodylate buffer, pH 7.3. Liver slices were dissected, rinsed with 0.05 M Na-cacodylate buffer and post-fixed in 2% osmium tetroxide (OsO₄) in sodium-cacodylate buffer at 4 °C for 24 h. Samples were dehydrated in a graded acetone series and embedded in Agar 100 Resin (Agar Scientific, Stansted, UK). Semi-thin (± 1 mm) sections were cut with a Reichert Jung Ultracut E microtome and stained with 0.1% thionin-0.1% methylene blue. The ultra-thin (± 70 nm) sections, on copper grids, were stained with uranyl acetate and lead citrate. Pictures were recorded on a JEOL JEM1400 (JEOL Europe BV, Zaventem, Belgium) transmission electron microscope (VIB BioImaging Core, platform Leuven). In order to be representative, analysis of fenestration involved a sufficient number of pictures, length of EC lining, and number of fenestrae (Supplementary Table 5).

Plasma analyses

To evaluate liver function and systemic metabolic parameters, blood was drawn via the right heart using a heparin-coated 26G needle on a 1 mL syringe and stored as plasma at -80 °C after centrifuging at 3,000 rpm for 10 min. The plasma was sent to the UZ Leuven routine testing laboratory for determination of alanine transferase (ALT), aspartate transferase (AST), triglycerides and total cholesterol. VWF: Ag levels were determined on plasma on 0.37% sodium citrate as described in Brill et al.⁸⁴. As a reference, a plasma pool of 20 untreated mice from our Cdh5-Cre; Zeb2^fl/fl mouse colony was generated.

Statistics

Graphpad Prism 9 was used for all statistical analyses. Quantitative data are expressed as mean ± standard error of the mean (s.e.m). Normal distribution of the data (or the residuals in case of ANOVA) was tested in Graphpad Prism 9 after performing outlier detection. If not normally distributed, data were log-transformed before performing parametric statistical analysis. Student’s t-test was used to compare 2 groups. One-way or two-way ANOVA with Tukey post-hoc test was used to compare > 2 groups. P < 0.05 was considered statistically significant. Clear trends (0.05 < P < 0.2) are also indicated on graphs. In case of 4 group comparisons, statistical significance for the comparison between EC^Zeb2KO on SD versus EC^WT or EC^GFP on WD are systematically not indicated in the graphs as these comparisons were not considered as biologically relevant. Statistical analyses for the RNAseq data are described above.

Data availability

The RNA sequencing datasets generated for this study are available in the NCBI GEO repository (https://www.ncbi.nlm.nih.gov/geo/), under series number GSE224595. Any additional information required to reanalyze the data generated for and reported in this paper is available upon request from the corresponding author. All data was analyzed with standard programs and packages as indicated in the Methods.

References

Aird, W. C. Phenotypic heterogeneity of the endothelium: II. Representative vascular beds. Circ. Res. 100, 174–190. https://doi.org/10.1161/01.RES.0000255690.03436.ae (2007).
Article CAS PubMed Google Scholar
de Haan, W. et al. Unraveling the transcriptional determinants of liver sinusoidal endothelial cell specialization. Am. J. Physiol. Gastrointest. Liver Physiol. 318, G803–G815. https://doi.org/10.1152/ajpgi.00215.2019 (2020).
Article CAS PubMed PubMed Central Google Scholar
Hammoutene, A. & Rautou, P. E. Role of liver sinusoidal endothelial cells in non-alcoholic fatty liver disease. J. Hepatol. 70, 1278–1291. https://doi.org/10.1016/j.jhep.2019.02.012 (2019).
Article CAS PubMed Google Scholar
Shahani, T. et al. Human liver sinusoidal endothelial cells but not hepatocytes contain factor VIII. J. Thromb. Haemost. 12, 36–42. https://doi.org/10.1111/jth.12412 (2014).
Article CAS PubMed Google Scholar
Sorensen, K. K. et al. The scavenger endothelial cell: a new player in homeostasis and immunity. Am. J. Physiol. Regul. Integr. Comp. Physiol. 303, R1217–1230. https://doi.org/10.1152/ajpregu.00686.2011 (2012).
Article CAS PubMed Google Scholar
Sorensen, K. K., Simon-Santamaria, J., McCuskey, R. S. & Smedsrod, B. Liver sinusoidal endothelial cells. Compr. Physiol. 5, 1751–1774. https://doi.org/10.1002/cphy.c140078 (2015).
Article PubMed Google Scholar
de Haan, W. et al. Endothelial Zeb2 preserves the hepatic angioarchitecture and protects against liver fibrosis. Cardiovasc. Res. 118, 1262–1275. https://doi.org/10.1093/cvr/cvab148 (2022).
Article CAS PubMed Google Scholar
De Smedt, J. et al. PU.1 drives specification of pluripotent stem cell-derived endothelial cells to LSEC-like cells. Cell. Death Dis. 12, 84. https://doi.org/10.1038/s41419-020-03356-2 (2021).
Article CAS PubMed PubMed Central Google Scholar
Geraud, C. et al. GATA4-dependent organ-specific endothelial differentiation controls liver development and embryonic hematopoiesis. J. Clin. Invest. 127, 1099–1114. https://doi.org/10.1172/JCI90086 (2017).
Article PubMed PubMed Central Google Scholar
Gomez-Salinero, J. M. et al. Specification of fetal liver endothelial progenitors to functional zonated adult sinusoids requires c-Maf induction. Cell Stem Cell 29, 593–609 e597. https://doi.org/10.1016/j.stem.2022.03.002 (2022).
Nolan, D. J. et al. Molecular signatures of tissue-specific microvascular endothelial cell heterogeneity in organ maintenance and regeneration. Dev. Cell. 26, 204–219. https://doi.org/10.1016/j.devcel.2013.06.017 (2013).
Article CAS PubMed Google Scholar
Koch, P. S., Lee, K. H., Goerdt, S. & Augustin, H. G. Angiodiversity and organotypic functions of sinusoidal endothelial cells. Angiogenesis 24, 289–310. https://doi.org/10.1007/s10456-021-09780-y (2021).
Article CAS PubMed PubMed Central Google Scholar
DeLeve, L. D. Liver sinusoidal endothelial cells in hepatic fibrosis. Hepatology 61, 1740–1746. https://doi.org/10.1002/hep.27376 (2015).
Article CAS PubMed Google Scholar
Gofton, C., Upendran, Y., Zheng, M. H. & George, J. M. A. F. L. D. How is it different from NAFLD? Clin. Mol. Hepatol. 29, S17–S31. https://doi.org/10.3350/cmh.2022.0367 (2023).
Article PubMed Google Scholar
Estes, C., Razavi, H., Loomba, R., Younossi, Z. & Sanyal, A. J. Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology 67, 123–133. https://doi.org/10.1002/hep.29466 (2018).
Article CAS PubMed Google Scholar
Younossi, Z. M. et al. Global epidemiology of nonalcoholic fatty liver disease-Meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 64, 73–84. https://doi.org/10.1002/hep.28431 (2016).
Article PubMed Google Scholar
Furuta, K., Guo, Q., Hirsova, P. & Ibrahim, S. H. Emerging roles of liver sinusoidal endothelial cells in nonalcoholic steatohepatitis. Biology (Basel). 9 https://doi.org/10.3390/biology9110395 (2020).
Herrnberger, L. et al. Formation of fenestrae in murine liver sinusoids depends on plasmalemma vesicle-associated protein and is required for lipoprotein passage. PLoS One. 9, e115005. https://doi.org/10.1371/journal.pone.0115005 (2014).
Article CAS PubMed PubMed Central Google Scholar
Eberhard, D. et al. Semaphorin-3A regulates liver sinusoidal endothelial cell porosity and promotes hepatic steatosis. Nat. Cardiovasc. Res. 3, 734–753. https://doi.org/10.1038/s44161-024-00487-z (2024).
Article CAS PubMed PubMed Central Google Scholar
Kawashita, E., Ozaki, T., Ishihara, K., Kashiwada, C. & Akiba, S. Endothelial group IVA phospholipase A(2) promotes hepatic fibrosis with sinusoidal capillarization in the early stage of non-alcoholic steatohepatitis in mice. Life Sci. 294, 120355. https://doi.org/10.1016/j.lfs.2022.120355 (2022).
Article CAS PubMed Google Scholar
Kus, E. et al. LSEC fenestrae are preserved despite Pro-inflammatory phenotype of liver sinusoidal endothelial cells in mice on high fat diet. Front. Physiol. 10, 6. https://doi.org/10.3389/fphys.2019.00006 (2019).
Article PubMed PubMed Central Google Scholar
Verhaegh, P. et al. Electron microscopic observations in perfusion-fixed human non-alcoholic fatty liver disease biopsies. Pathology 53, 220–228. https://doi.org/10.1016/j.pathol.2020.07.018 (2021).
Article CAS PubMed Google Scholar
Bondareva, O. et al. Single-cell profiling of vascular endothelial cells reveals progressive organ-specific vulnerabilities during obesity. Nat. Metab. 4, 1591–1610. https://doi.org/10.1038/s42255-022-00674-x (2022).
Article CAS PubMed PubMed Central Google Scholar
Lefere, S., Devisscher, L. & Geerts, A. Angiogenesis in the progression of non-alcoholic fatty liver disease. Acta Gastroenterol. Belg. 83, 301–307 (2020).
CAS PubMed Google Scholar
Verhulst, S. et al. Gene signatures detect damaged liver sinusoidal endothelial cells in chronic liver diseases. Front. Med. (Lausanne). 8, 750044. https://doi.org/10.3389/fmed.2021.750044 (2021).
Article PubMed Google Scholar
Francque, S. et al. Increased intrahepatic resistance in severe steatosis: endothelial dysfunction, vasoconstrictor overproduction and altered microvascular architecture. Lab. Invest. 92, 1428–1439. https://doi.org/10.1038/labinvest.2012.103 (2012).
Article CAS PubMed Google Scholar
Wang, Y., Nakajima, T., Gonzalez, F. J. & Tanaka, N. PPARs as metabolic regulators in the liver: lessons from Liver-Specific PPAR-Null mice. Int. J. Mol. Sci. 21 https://doi.org/10.3390/ijms21062061 (2020).
Liss, K. H. & Finck, B. N. PPARs and nonalcoholic fatty liver disease. Biochimie 136, 65–74. https://doi.org/10.1016/j.biochi.2016.11.009 (2017).
Article CAS PubMed Google Scholar
Guixe-Muntet, S. et al. Review article: vascular effects of PPARs in the context of NASH. Aliment. Pharmacol. Ther. 56, 209–223. https://doi.org/10.1111/apt.17046 (2022).
Article CAS PubMed PubMed Central Google Scholar
Rodriguez-Vilarrupla, A. et al. PPARalpha activation improves endothelial dysfunction and reduces fibrosis and portal pressure in cirrhotic rats. J. Hepatol. 56, 1033–1039. https://doi.org/10.1016/j.jhep.2011.12.008 (2012).
Article CAS PubMed Google Scholar
Tegge, A. N. et al. Transcriptomic analysis of hepatic cells in multicellular organotypic liver models. Sci. Rep. 8, 11306. https://doi.org/10.1038/s41598-018-29455-x (2018).
Article CAS PubMed PubMed Central Google Scholar
McMahan, R. H., Porsche, C. E., Edwards, M. G. & Rosen, H. R. Free fatty acids differentially downregulate chemokines in liver sinusoidal endothelial cells: insights into Non-Alcoholic fatty liver disease. PLoS One. 11, e0159217. https://doi.org/10.1371/journal.pone.0159217 (2016).
Article CAS PubMed PubMed Central Google Scholar
Sousa, V. H., Miyoshi, G., Hjerling-Leffler, J., Karayannis, T. & Fishell, G. Characterization of Nkx6-2-derived neocortical interneuron lineages. Cereb Cortex 19(Suppl 1), i1–10. https://doi.org/10.1093/cercor/bhp038 (2009).
Kalucka, J. et al. Single-Cell Transcriptome Atlas of Murine Endothelial Cells. Cell 180, 764–779 e720. https://doi.org/10.1016/j.cell.2020.01.015 (2020).
Guilliams, M. et al. Spatial proteogenomics reveals distinct and evolutionarily conserved hepatic macrophage niches. Cell 185, 379–396. https://doi.org/10.1016/j.cell.2021.12.018 (2022).
Remmerie, A. et al. Osteopontin Expression Identifies a Subset of Recruited Macrophages Distinct from Kupffer Cells in the Fatty Liver. Immunity 53, 641–657. https://doi.org/10.1016/j.immuni.2020.08.004 (2020).
Park, Y. N. et al. Neoangiogenesis and sinusoidal capillarization in dysplastic nodules of the liver. Am. J. Surg. Pathol. 22, 656–662. https://doi.org/10.1097/00000478-199806000-00002 (1998).
Article CAS PubMed Google Scholar
Birkhoff, J. C., Huylebroeck, D. & Conidi, A. ZEB2, the Mowat-Wilson syndrome transcription factor: confirmations, novel functions, and continuing surprises. Genes (Basel). 12 https://doi.org/10.3390/genes12071037 (2021).
Bougarne, N. et al. Molecular actions of PPARalpha in lipid metabolism and inflammation. Endocr. Rev. 39, 760–802. https://doi.org/10.1210/er.2018-00064 (2018).
Article PubMed Google Scholar
Gracia-Sancho, J., Caparros, E., Fernandez-Iglesias, A. & Frances, R. Role of liver sinusoidal endothelial cells in liver diseases. Nat. Rev. Gastroenterol. Hepatol. 18, 411–431. https://doi.org/10.1038/s41575-020-00411-3 (2021).
Article PubMed Google Scholar
Meinken, J., Walker, G., Cooper, C. R. & Min, X. J. MetazSecKB: the human and animal secretome and subcellular proteome knowledgebase. Database (Oxford) (2015). https://doi.org/10.1093/database/bav077 (2015).
Miyao, M. et al. Pivotal role of liver sinusoidal endothelial cells in NAFLD/NASH progression. Lab. Invest. 95, 1130–1144. https://doi.org/10.1038/labinvest.2015.95 (2015).
Article CAS PubMed Google Scholar
Callewaert, B., Gsell, W., Himmelreich, U. & Jones, E. A. V. Q-VAT: qunatitative vascular analysis tool. Front Cardiovasc. Med (2023).
Lafoz, E., Ruart, M., Anton, A., Oncins, A. & Hernandez-Gea, V. The endothelium as a driver of liver fibrosis and regeneration. Cells 9 https://doi.org/10.3390/cells9040929 (2020).
Zhou, C. et al. Inverse association between lung function and nonalcoholic fatty liver disease: an observational and Mendelian randomization study. Nutr. Metab. Cardiovasc. Dis. 34, 2705–2712. https://doi.org/10.1016/j.numecd.2024.07.007 (2024).
Article CAS PubMed Google Scholar
Winkler, M. et al. Endothelial GATA4 controls liver fibrosis and regeneration by preventing a pathogenic switch in angiocrine signaling. J. Hepatol. https://doi.org/10.1016/j.jhep.2020.08.033 (2020).
Article PubMed Google Scholar
Guo, Q. et al. Liver sinusoidal endothelial cell expressed vascular cell adhesion molecule 1 promotes liver fibrosis. Front. Immunol. 13, 983255. https://doi.org/10.3389/fimmu.2022.983255 (2022).
Article CAS PubMed PubMed Central Google Scholar
Coppiello, G. et al. Meox2/Tcf15 heterodimers program the heart capillary endothelium for cardiac fatty acid uptake. Circulation 131, 815–826. https://doi.org/10.1161/CIRCULATIONAHA.114.013721 (2015).
Article CAS PubMed Google Scholar
Ibrahim, A. & Arany, Z. Does endothelium buffer fat?? Circ. Res. 120, 1219–1221. https://doi.org/10.1161/CIRCRESAHA.117.310865 (2017).
Article CAS PubMed Google Scholar
Pi, X., Xie, L. & Patterson, C. Emerging roles of vascular endothelium in metabolic homeostasis. Circ. Res. 123, 477–494. https://doi.org/10.1161/CIRCRESAHA.118.313237 (2018).
Article CAS PubMed PubMed Central Google Scholar
Oie, C. I. et al. Rat liver sinusoidal endothelial cells (LSECs) express functional low density lipoprotein receptor-related protein-1 (LRP-1). J. Hepatol. 55, 1346–1352. https://doi.org/10.1016/j.jhep.2011.03.013 (2011).
Article CAS PubMed Google Scholar
Wilkinson, A. L., Qurashi, M. & Shetty, S. The role of sinusoidal endothelial cells in the Axis of inflammation and Cancer within the liver. Front. Physiol. 11, 990. https://doi.org/10.3389/fphys.2020.00990 (2020).
Article PubMed PubMed Central Google Scholar
Donnelly, K. L. et al. Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J. Clin. Invest. 115, 1343–1351. https://doi.org/10.1172/JCI23621 (2005).
Article CAS PubMed PubMed Central Google Scholar
Alves-Bezerra, M. & Cohen, D. E. Triglyceride metabolism in the liver. Compr. Physiol. 8, 1–8. https://doi.org/10.1002/cphy.c170012 (2017).
Article PubMed PubMed Central Google Scholar
Matuszyk, E. et al. Differential response of liver sinusoidal endothelial cells and hepatocytes to oleic and palmitic acid revealed by Raman and CARS imaging. Biochim. Biophys. Acta Mol. Basis Dis. 1866, 165763. https://doi.org/10.1016/j.bbadis.2020.165763 (2020).
Article CAS PubMed Google Scholar
Szafraniec, E. et al. Raman spectroscopy-based insight into lipid droplets presence and contents in liver sinusoidal endothelial cells and hepatocytes. J. Biophotonics. 12, e201800290. https://doi.org/10.1002/jbio.201800290 (2019).
Article CAS PubMed Google Scholar
Farrell, G. et al. Mouse models of nonalcoholic steatohepatitis: toward optimization of their relevance to human nonalcoholic steatohepatitis. Hepatology 69, 2241–2257. https://doi.org/10.1002/hep.30333 (2019).
Article PubMed Google Scholar
van der Graaff, D., Kwanten, W. J. & Francque, S. M. The potential role of vascular alterations and subsequent impaired liver blood flow and hepatic hypoxia in the pathophysiology of non-alcoholic steatohepatitis. Med. Hypotheses. 122, 188–197. https://doi.org/10.1016/j.mehy.2018.11.014 (2019).
Article PubMed Google Scholar
Lefere, S. et al. Angiopoietin-2 promotes pathological angiogenesis and is a therapeutic target in murine nonalcoholic fatty liver disease. Hepatology 69, 1087–1104. https://doi.org/10.1002/hep.30294 (2019).
Article CAS PubMed Google Scholar
Gladka, M. M. et al. Cardiomyocytes stimulate angiogenesis after ischemic injury in a ZEB2-dependent manner. Nat. Commun. 12, 84. https://doi.org/10.1038/s41467-020-20361-3 (2021).
Article CAS PubMed PubMed Central Google Scholar
Seuntjens, E. et al. Sip1 regulates sequential fate decisions by feedback signaling from postmitotic neurons to progenitors. Nat. Neurosci. 12, 1373–1380. https://doi.org/10.1038/nn.2409 (2009).
Article CAS PubMed Google Scholar
Dong, F. et al. Endothelin-1 enhances oxidative stress, cell proliferation and reduces apoptosis in human umbilical vein endothelial cells: role of ETB receptor, NADPH oxidase and caveolin-1. Br. J. Pharmacol. 145, 323–333. https://doi.org/10.1038/sj.bjp.0706193 (2005).
Article CAS PubMed PubMed Central Google Scholar
van der Graaff, D. et al. Vasoconstrictor antagonism improves functional and structural vascular alterations and liver damage in rats with early NAFLD. JHEP Rep. 4, 100412. https://doi.org/10.1016/j.jhepr.2021.100412 (2022).
Article PubMed Google Scholar
Liu, L. J., Liao, J. M. & Zhu, F. Proliferating cell nuclear antigen clamp associated factor, a potential proto-oncogene with increased expression in malignant Gastrointestinal tumors. World J. Gastrointest. Oncol. 13, 1425–1439. https://doi.org/10.4251/wjgo.v13.i10.1425 (2021).
Article PubMed PubMed Central Google Scholar
Liu, X. et al. PCLAF promotes neuroblastoma G1/S cell cycle progression via the E2F1/PTTG1 axis. Cell. Death Dis. 13, 178. https://doi.org/10.1038/s41419-022-04635-w (2022).
Article CAS PubMed PubMed Central Google Scholar
Keam, S. J. & Resmetirom First Approval Drugs 84, 729–735. https://doi.org/10.1007/s40265-024-02045-0 (2024).
Article CAS PubMed Google Scholar
Ganz, M., Csak, T. & Szabo, G. High fat diet feeding results in gender specific steatohepatitis and inflammasome activation. World J. Gastroenterol. 20, 8525–8534. https://doi.org/10.3748/wjg.v20.i26.8525 (2014).
Article CAS PubMed PubMed Central Google Scholar
Wang, Y. et al. Ephrin-B2 controls VEGF-induced angiogenesis and lymphangiogenesis. Nature 465, 483–486. https://doi.org/10.1038/nature09002 (2010).
Article CAS PubMed Google Scholar
Higashi, Y. et al. Generation of the floxed allele of the SIP1 (Smad-interacting protein 1) gene for Cre-mediated conditional knockout in the mouse. Genesis 32, 82–84. https://doi.org/10.1002/gene.10048 (2002).
Article CAS PubMed Google Scholar
de Haan, W., Bhattacharjee, A., Ruddle, P., Kang, M. H. & Hayden, M. R. ABCA1 in adipocytes regulates adipose tissue lipid content, glucose tolerance, and insulin sensitivity. J. Lipid Res. 55, 516–523. https://doi.org/10.1194/jlr.M045294 (2014).
Article CAS PubMed PubMed Central Google Scholar
Picelli, S. et al. Smart-seq2 for sensitive full-length transcriptome profiling in single cells. Nat. Methods. 10, 1096–1098. https://doi.org/10.1038/nmeth.2639 (2013).
Article CAS PubMed Google Scholar
Bolger, A. M., Lohse, M. & Usadel, B. Trimmomatic: a flexible trimmer for illumina sequence data. Bioinformatics 30, 2114–2120. https://doi.org/10.1093/bioinformatics/btu170 (2014).
Article CAS PubMed PubMed Central Google Scholar
Kim, D., Langmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods. 12, 357–360. https://doi.org/10.1038/nmeth.3317 (2015).
Article CAS PubMed PubMed Central Google Scholar
Anders, S., Pyl, P. T. & Huber, W. HTSeq–a Python framework to work with high-throughput sequencing data. Bioinformatics 31, 166–169. https://doi.org/10.1093/bioinformatics/btu638 (2015).
Article CAS PubMed Google Scholar
Risso, D., Ngai, J., Speed, T. P. & Dudoit, S. Normalization of RNA-seq data using factor analysis of control genes or samples. Nat. Biotechnol. 32, 896–902. https://doi.org/10.1038/nbt.2931 (2014).
Article CAS PubMed PubMed Central Google Scholar
Risso, D., Schwartz, K., Sherlock, G. & Dudoit, S. GC-content normalization for RNA-Seq data. BMC Bioinform. 12, 480. https://doi.org/10.1186/1471-2105-12-480 (2011).
Article CAS Google Scholar
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. EdgeR: a bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140. https://doi.org/10.1093/bioinformatics/btp616 (2010).
Article CAS PubMed Google Scholar
Browaeys, R., Saelens, W. & Saeys, Y. NicheNet: modeling intercellular communication by linking ligands to target genes. Nat. Methods. 17, 159–162. https://doi.org/10.1038/s41592-019-0667-5 (2020).
Article CAS PubMed Google Scholar
Zhou, Y. et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat. Commun. 10, 1523. https://doi.org/10.1038/s41467-019-09234-6 (2019).
Article CAS PubMed PubMed Central Google Scholar
Celton-Morizur, S., Merlen, G., Couton, D., Margall-Ducos, G. & Desdouets, C. The insulin/akt pathway controls a specific cell division program that leads to generation of binucleated tetraploid liver cells in rodents. J. Clin. Invest. 119, 1880–1887. https://doi.org/10.1172/jci38677 (2009).
Article CAS PubMed PubMed Central Google Scholar
Gentric, G. et al. Oxidative stress promotes pathologic polyploidization in nonalcoholic fatty liver disease. J. Clin. Invest. 125, 981–992. https://doi.org/10.1172/JCI73957 (2015).
Article PubMed PubMed Central Google Scholar
Huang, L. K. & Wang, M. J. Image thresholding by minimizing the measures of fuzziness. Pattern Recogn. 28, 41–51. https://doi.org/10.1016/0031-3203(94)E (1995).
Article Google Scholar
He, L. et al. Single-cell RNA sequencing of mouse brain and lung vascular and vessel-associated cell types. Sci. Data. 5, 180160. https://doi.org/10.1038/sdata.2018.160 (2018).
Article CAS PubMed PubMed Central Google Scholar
Brill, A. et al. Von Willebrand factor-mediated platelet adhesion is critical for deep vein thrombosis in mouse models. Blood 117, 1400–1407. https://doi.org/10.1182/blood-2010-05-287623 (2011).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

We acknowledge the VIB Bioimaging Core & VIB-KU Leuven Center for Brain and Disease Research (VIB-CBD), where the immunohistochemical samples were imaged and TEM was performed. We would like to thank them for training, support and access to their instrumentation. We acknowledge Dr. S. Gordts, Dr. S. Kersten (Wageningen University) and David Carbonez for helpful discussions and M. Wallays and A. Bruffaerts for technical assistance. This work was supported by: internal KU Leuven funding [C12/16/023 to W.dH. and A.Z, C14/19/095 to A.L., A.Z. E.A.V.J. and D.H.], a Program Financing grant [PF/10/014] to A.L. and M.J.; a European Research Council Grant [FP7-StG-IMAGINED203291] to A.L. and a H2020 848109-CRUCIAL grant to E.A.V.J., a Cosmetics Europe/European Commission FP7 Grant [FP7-Health-HemiBio266777] to A.L. and L.A.vG.; an Interuniversity Attraction Poles Grant [IUAP/P7/07] to A.L., A.Z. and D.H.; two Fonds voor Wetenschappelijk Onderzoek [FWO] grants [WOG001420N to A.L., E.A.V.J. and A.Z.; G.0A3116 to D.H.]; a ZONMW OffRoad program [2018/23115|ZONMW] to E.M.; a Marie Sklodowska-Curie Actions post-doctoral fellowship [H2020-MSCA-IF-REZONABLE658666] to W.dH; a FWO junior post-doctoral fellowship [1243121N] to S.Ve; two pre-doctoral FWO fellowships [1157318 N to W.D., 1S25817N to J.V.W.]; Erasmus University Medical Center departmental funds [to D.H.] and BIG project funding [including D.H and E.M.], a collaborative extra support between Erasmus University Rotterdam and its University Medical Center for promoting fundamental research at the Theme of Biomedical Sciences, and hence the Department of Cell Biology. A.L. has drawn the schematic images.

Author information

Authors and Affiliations

Department of Cardiovascular Sciences, Center for Molecular and Vascular Biology, Endothelial Cell Biology Unit, KU Leuven, Campus Gasthuisberg, Onderwijs & Navorsing 1, Herestraat 49, box 911, Leuven, 3000, Belgium
Wouter Dheedene, Louise Demuynck, Bram Callewaert, Willeke de Haan, Jore Van Wauwe, Petra Vandervoort, Marleen Lox, Mathias Stroobants, Renaud Lavend’homme, Elizabeth A. V. Jones, An Zwijsen, Marc Jacquemin, Kimberly Martinod & Aernout Luttun
Liver Cell Biology research group, Vrije Universiteit Brussel, Brussels, Belgium
Stefaan Verhulst & Leo A. van Grunsven
Department of Oncology, Laboratory of Angiogenesis and Vascular Metabolism, KU Leuven, Leuven, Belgium
Stefan Vinckier
Laboratory of Angiogenesis and Vascular Metabolism, Center for Cancer Biology, VIB, Leuven, Belgium
Stefan Vinckier
Center for Biomics-Genomics, Erasmus University Medical Center, Rotterdam, The Netherlands
Wilfred F. J. van IJcken
Department of Cell Biology, Erasmus University Medical Center, Rotterdam, The Netherlands
Wilfred F. J. van IJcken, Danny Huylebroeck & Eskeatnaf Mulugeta
CARIM, Maastricht University, Maastricht, The Netherlands
Elizabeth A. V. Jones
Department of Development and Regeneration, KU Leuven, Leuven, Belgium
Danny Huylebroeck
Department of Internal Medicine, Erasmus University Medical Center, Rotterdam, The Netherlands
Eskeatnaf Mulugeta

Authors

Wouter Dheedene
View author publications
Search author on:PubMed Google Scholar
Stefaan Verhulst
View author publications
Search author on:PubMed Google Scholar
Louise Demuynck
View author publications
Search author on:PubMed Google Scholar
Bram Callewaert
View author publications
Search author on:PubMed Google Scholar
Willeke de Haan
View author publications
Search author on:PubMed Google Scholar
Stefan Vinckier
View author publications
Search author on:PubMed Google Scholar
Jore Van Wauwe
View author publications
Search author on:PubMed Google Scholar
Petra Vandervoort
View author publications
Search author on:PubMed Google Scholar
Marleen Lox
View author publications
Search author on:PubMed Google Scholar
Mathias Stroobants
View author publications
Search author on:PubMed Google Scholar
Renaud Lavend’homme
View author publications
Search author on:PubMed Google Scholar
Wilfred F. J. van IJcken
View author publications
Search author on:PubMed Google Scholar
Elizabeth A. V. Jones
View author publications
Search author on:PubMed Google Scholar
An Zwijsen
View author publications
Search author on:PubMed Google Scholar
Marc Jacquemin
View author publications
Search author on:PubMed Google Scholar
Leo A. van Grunsven
View author publications
Search author on:PubMed Google Scholar
Kimberly Martinod
View author publications
Search author on:PubMed Google Scholar
Danny Huylebroeck
View author publications
Search author on:PubMed Google Scholar
Eskeatnaf Mulugeta
View author publications
Search author on:PubMed Google Scholar
Aernout Luttun
View author publications
Search author on:PubMed Google Scholar

Contributions

W.D. designed the study, performed experiments, analyzed data and wrote the manuscript; W.dH. edited the manuscript and provided critical intellectual input; L.D. performed experiments and analyzed data; S.Vi., B.C, J.V.W., P.V., M.L., M.S., and R.L. performed experiments. W.F.J.vIJ. coordinated RNAseq experiments; S.Ve. performed (Multi-)NicheNet analysis and data-mining of existing scRNAseq datasets; E.M. performed RNAseq data analysis; E.A.V.J., A.Z., M.J., K.M., and L.A.vG provided resources and/or critical intellectual input; D.H. provided transgenic mouse models, and methodological and critical intellectual input; A.L. designed the study and edited the manuscript and figures. All authors reviewed and approved the manuscript.

Corresponding author

Correspondence to Aernout Luttun.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Dheedene, W., Verhulst, S., Demuynck, L. et al. Loss of endothelial ZEB2 in mice attenuates steatosis early during metabolic dysfunction-associated steatotic liver disease. Sci Rep 15, 23434 (2025). https://doi.org/10.1038/s41598-025-05881-6

Download citation

Received: 08 October 2024
Accepted: 03 June 2025
Published: 02 July 2025
DOI: https://doi.org/10.1038/s41598-025-05881-6