Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 10 (1), 4688

Embryonic Mesothelial-Derived Hepatic Lineage of Quiescent and Heterogenous Scar-Orchestrating Cells Defined but Suppressed by WT1

Affiliations

Embryonic Mesothelial-Derived Hepatic Lineage of Quiescent and Heterogenous Scar-Orchestrating Cells Defined but Suppressed by WT1

Timothy James Kendall et al. Nat Commun.

Abstract

Activated hepatic stellate cells (aHSCs) orchestrate scarring during liver injury, with putative quiescent precursor mesodermal derivation. Here we use lineage-tracing from development, through adult homoeostasis, to fibrosis, to define morphologically and transcriptionally discreet subpopulations of aHSCs by expression of WT1, a transcription factor controlling morphological transitions in organogenesis and adult homoeostasis. Two distinct populations of aHSCs express WT1 after injury, and both re-engage a transcriptional signature reflecting embryonic mesothelial origin of their discreet quiescent adult precursor. WT1-deletion enhances fibrogenesis after injury, through upregulated Wnt-signalling and modulation of genes central to matrix persistence in aHSCs, and augmentation of myofibroblastic transition. The mesothelial-derived lineage demonstrates punctuated phenotypic plasticity through bidirectional mesothelial-mesenchymal transitions. Our findings demonstrate functional heterogeneity of adult scar-orchestrating cells that can be whole-life traced back through specific quiescent adult precursors to differential origin in development, and define WT1 as a paradoxical regulator of aHSCs induced by injury but suppressing scarring.

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
Functionally distinct subpopulations of qHSCs derive from alternative embryological origins. a Lineage labelling of hepatic mesothelial cells and their progeny was induced at E10.5 in WT1CreERT2/+;Ai14 animals, and qHSCs isolated or livers examined in adulthood. b Adult WT1CreERT2/+;Ai14 livers, labelled at E10.5, were stained for lineage label (RFP) and qHSC markers (desmin and PDGFRβ); 68.4 ± 3.46% qHSCs were lineage label positive (mean ± s.e.m., n = 3, scale bars 100 μm). c The presence of the lineage label was the major component describing variation in differentially expressed genes between lineage-labelled and unlabelled populations of qHSCs isolated from adult WT1CreERT2/+;Ai14 livers (n = 3), labelled at E10.5. d GO term analysis of differentially expressed genes between lineage-labelled and unlabelled qHSCs showed distinct functional profiles depending on origin
Fig. 2
Fig. 2
Adult injury induces a population of WT1-positive cells by activation of qHSCs that derive from the embryonic hepatic mesothelium. a Lineage labelling of hepatic mesothelium-derived cells was induced at E10.5 in a further cohort of WT1CreERT2/+;Ai14 animals and fibrosis induced by iterative injury with CCl4. b Lineage-labelled cells after injury were confirmed to be aHSCs by confocal microscopy demonstrating colocalization of the lineage label (RFP—lilac) with the HSC marker GFAP (yellow). Activated lineage-labelled cells were quantified after staining of injured livers for WT1 (green), lineage label (RFP—lilac), and αSMA (yellow) to demonstrate a population of WT1-positive activated cells originating from mesothelial-derived qHSCs. c 78.1 ± 5.3% of WT1-positive cells were lineage labelled, compared with 20.5 ± 3.7% of WT1-negative cells (Welch two-sample t-test t(7.144) = −7.9601, *p = 8.437 × 10–5, n = 5 animals, data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5× interquartile range (whiskers). Scale bars 100 μm
Fig. 3
Fig. 3
Subpopulations of aHSCs defined by WT1 expression are distributed differentially within hepatic scars and show differences in cell cycling. a WT1-positive cells are present in murine liver injured by CCl4 injection, colocalizing with markers of aHSCs (rows show colocalized staining with PDGFRβ, GFAP, desmin, and αSMA). Scale bars 100 μm. b The nuclear position of WT1-positive and WT1-negative aHSCs, and hepatic vein profile were marked in each image of liver from animals chronically injured with CCl4 to allow quantification of the fibrospatial distribution of aHSCs (c). Scale bars 100 μm. d The distances from nuclei to central vein were calculated for each biological replicate, demonstrating different distributions within hepatic scars of WT1-positive and WT1-negative aHSCs (representative density distribution plot, bootstrapped Kolmogorov–Smirnov test p < 0.00001). e The mean distances of WT1-positive and WT1-negative aHSCs to the central vein, on a per animal basis, were calculated to show that WT1-positive cells were significantly closer to central vein profiles (*, Welch two-sample t-test, t(9.6195) = 8.4046, p = 0.0000098, n = 6). Data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5× interquartile range (whiskers). f Livers injured with chronic CCl4 injection were stained for WT1, Ki67, and αSMA. Ki67 immunopositivity was significantly greater for WT1-negative compared with WT1-positive aHSCs (Welch two-sample t-test, p < 0.001, 20.3 ± 0.9% versus 1.9 ± 0.7%, 10 pericentral fields/animal, n = 3). Scale bars 100 μm
Fig. 4
Fig. 4
WT1 expression defines morphologically distinct subpopulations of aHSCs. a Quiescent HSCs from WT1GFP/+ animals were activated by 7 days of culture on plastic and examined by flow cytometry, demonstrating three distinct subpopulations based on GFP expression (wild-type cells left panel, WT1GFP/+ cells right panel, representative plots with population percentages). b All three subpopulations were profibrogenic but the WT1-intermediate population demonstrated enhanced fibrogenic gene expression (one-way ANOVA (Acta1 F(2,12) = 11.72, p = 0.0015; Timp1 F(2,12) = 11, p = 0.0019; Col1a1 F(2,12) = 4.496, p = 0.0349; post-hoc Tukey, *< 0.05, ** < 0.01, n = 5, n,s p > 0.05). Data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5× interquartile range (whiskers). c WT1-high aHSCs are round in profile, whilst WT1-intermediate and WT1-negative cells are myofibroblast-shaped. Scale bars 100 μm. d Representative population distribution of ‘circularity’ from a single cell preparation. e The circularity of subpopulations of aHSCs based on WT1 status was significantly different; post-hoc Tukey testing indicated that the circularity of WT1-high cells was significantly different to that of WT1-intermediate cells (*, one-way ANOVA F(2,21) = 4.997, p = 0.0168, n = 8). Data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5× interquartile range (whiskers). f Morphologically biphasic WT1-positive cells in unsorted aHSC cultures of 4-hydroxytamoxifen-treated WT1CreERT2/+;Ai14 cells, subsequently fixed and stained for RFP and WT1 (g, scale bars 200 μm (upper row), 50 μm (middle and lower rows), representative of three biological replicates)
Fig. 5
Fig. 5
Subpopulations of aHSCs defined by WT1 expression in chronic fibrotic injury in vivo are transcriptionally distinct. a WT1GFP/+ animals were iteratively injured with CCl4 to induce HSC activation and fibrosis prior to isolation of subpopulations defined by WT1(GFP) expression by flow cytometry (n = 3 animals). b Subpopulations of WT1(GFP)-high and WT1(GFP)-intermediate were present within the isolated HSC fraction from injured WT1GFP/+ animals (right panel) but not injured wild-type animals (left, representative plots with population percentages). c A multi-dimensional scaling plot of gene expression after microarray analysis of WT1-positive in vivo activated and WT1-negative in vitro aHSCs demonstrated separation based on WT1(GFP) expression. d Over and underexpressed probes between subpopulations of aHSCs determined by fitting of a linear model. e Hierarchical clustering independently separated subpopulations into groups defined by WT1 expression
Fig. 6
Fig. 6
The transcriptional profiles of aHSC populations defined by WT1 expression are classically activated but distinct, with enhanced scar-related processes in WT1-intermediate cells. a GO term analysis of differentially expressed genes between subpopulations of aHSCs defined by WT1 expression demonstrate significant differences in profiles between all subpopulations, with WT1-high and WT1-intermediate populations showing fewer differences. b Specific mapping of differentially expressed genes to GO terms for indicative scarring responses demonstrate the enhancement in WT1-intermediate aHSCs compared with WT1-high aHSCs. c The transcriptome, determined by RNAseq, of aHSCs isolated from PDGFRβCre;WT1GFP/+ animals (n = 6) with liver fibrosis induced by iterative injury with CCl4 was compared with that of quiescent lineage-label positive HSCs from WT1CreERT2/+;Ai14 animals (n = 3) induced at E10.5. Gene ontology terms mapped to differentially expressed genes for WT1-high cells demonstrate engagement of cellular processes associated with the HSC activation paradigm
Fig. 7
Fig. 7
The hepatic mesothelium in development gives rise to a lineage of aHSCs that re-engages a mesothelial gene signature, and in adulthood is a limited de novo source of subcapsular WT1-defined aHSCs. a Only mesothelial-derived WT1-positive lineages of aHSCs re-engage a mesothelial gene signature after fibrotic injury. Data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5x interquartile range (whiskers). b Lineage labelling of hepatic mesothelium was induced in adult WT1CreERT2/+;Ai14 animals before iterative injury with CCl4, and c livers stained for WT1 (green) and lineage-label (lilac). Scale bars 100 μm
Fig. 8
Fig. 8
Loss of WT1 expression by PDGFRβ-expressing HSCs and myofibroblasts causes an enhanced fibrotic response to chronic injury without a change in myofibroblast number. a Chronic fibrosis was induced by iterative injury with CCl4 in animals in which WT1 had been constitutively deleted, or control animals. b The number of WT1-positive aHSCs after chronic injury in PDGFRβCre;WT1−/fl animals is significantly reduced compared with wild-type animals (295.8 ± 59.4 vs 829.8 ± 116.3 WT1-positive cells/10 pericentral fields, n = 6; Welch two-sample t-test, t(7.4397) = −4.0896, *p = 0.00407). Example pericentral fields from the livers of injured wild-type and WT1-deleted animals stained for WT1 and αSMA are shown. Scale bars 100 μm. c There is an enhanced fibrotic response to chronic injury in WT1-deleted PDGFRβCre;WT1−/fl animals (whole-slide PSR quantification of fibrotic matrix, 8.1 ± 0.6% vs 4.8 ± 0.5%, n = 6 (wild-type), n = 11 (PDGFRβCre;WT1−/fl); Welch two-sample t-test, t(13.765) = −4.194, *p = 0.0009388). Example PSR-stained sections of whole lobes of the livers of injured wild-type and WT1-deleted animals are shown. Scale bars 1 mm. d The enhanced fibrotic response to chronic injury in PDGFRβCre;WT1−/fl animals is not associated with increased numbers of aHSCs compared with wild-type animals (1007.8 ± 273.8 vs 927.0 ± 186.5 αSMA-positive cells/10 pericentral fields; Welch two-sample t-test, t(8.819) = −0.244), p = 0.8128). Data are represented as individual points with median (centre line), first and third quartiles (lower and upper box limits), 1.5× interquartile range (whiskers). e The use of PDGFRβCre;WT1GFP/fl animals allows continued GFP reporting to demonstrate the persistence of WT1-deleted aHSCs after pericentral chronic injury. Scale bars 100 μm
Fig. 9
Fig. 9
Loss of WT1 expression by aHSCs upregulates extracellular matrix and other profibrotic transcriptional pathways and permits transition to a myofibroblast morphology. a Genes mapping to GO terms for ‘extracellular matrix’, morphological and developmental transitions, non-canonical Wnt signalling and responses to growth factors are upregulated following WT1 loss in PDGFRβCre;WT1−/fl animals after iterative injury with CCl4 in vivo. Genes mapping to immune and inflammatory responses and DNA replication are downregulated. b The population density distributions of circularity of activated RFP-positive cells under the condition of WT1 deletion was significantly different from that of control RFP(WT1)-positive cells (Bootstrapped two-sample Kolmogorov–Smirnov test, p = 0.002) during activation of qHSCs isolated from WT1CreERT2/fl;Ai14 animals in vitro, demonstrating a shift from round to myofibroblast-shaped aHSCs. Representative of three biological replicates. Scale bars 100 μm
Fig. 10
Fig. 10
WT1 is a paradoxical negative regulator of fibrogenesis in distinct subpopulations of aHSCs defined by WT1 expression. WT1-positive subpopulations in liver injury are generated by activation of specific progenitors that derive from the embryonic hepatic mesothelium in early development. Developmental origin determines the functional categorisation of qHSCs; those derived from the mesothelium demonstrate engagement of vascular regulatory pathways, whilst those from non-mesothelial sources have immunomodulatory transcriptional profiles

Similar articles

See all similar articles

References

    1. Mederacke I, et al. Fate tracing reveals hepatic stellate cells as dominant contributors to liver fibrosis independent of its aetiology. Nat. Commun. 2013;4:2823. doi: 10.1038/ncomms3823. - DOI - PMC - PubMed
    1. Coulouarn C, Clément B. Stellate cells and the development of liver cancer: therapeutic potential of targeting the stroma. J. Hepatol. 2014;60:1306–1309. doi: 10.1016/j.jhep.2014.02.003. - DOI - PubMed
    1. Friedman SL. Seminars in medicine of the Beth Israel Hospital, Boston. The cellular basis of hepatic fibrosis. Mechanisms and treatment strategies. N. Engl. J. Med. 1993;328:1828–1835. doi: 10.1056/NEJM199304223281620. - DOI - PubMed
    1. Semela D, et al. PDGF signaling through ephrin-B2 regulates hepatic vascular structure and function. Gastroenterology. 2008;135:671–679. doi: 10.1053/j.gastro.2008.04.010. - DOI - PMC - PubMed
    1. Weiskirchen R, Tacke F. Cellular and molecular functions of hepatic stellate cells in inflammatory responses and liver immunology. Hepatobiliary Surg. Nutr. 2014;3:344–363. - PMC - PubMed

Publication types

Feedback