記住我
Research ArticleAngiogenesisVascular biology
Open Access | 
10.1172/jci.insight.177819
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Arribas, V. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Onetti, Y. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Ramiro-Pareta, M. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Villacampa, P. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Beck, H. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Alberola, M. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by
Esteve-Codina, A.
in:
JCI
|
PubMed
|
Google Scholar
|
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Merkel, A. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by
Sperandio, M.
in:
JCI
|
PubMed
|
Google Scholar
|
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Martínez-Estrada, O. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Schmid, B. in: JCI | PubMed | Google Scholar
1Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona and Bellvitge Biomedical Research Institute (IDIBELL), L’Hospitalet del Llobregat, Spain.
2Celltec-UB, Department of Cell Biology, Physiology, and Immunology, Faculty of Biology, and
3Institute of Biomedicine (IBUB), University of Barcelona, Barcelona, Spain.
4Walter Brendel Centre of Experimental Medicine, Biomedical Center, Institute of Cardiovascular Physiology and Pathophysiology, Ludwig-Maximilians-University Munich, Planegg-Martinsried, Germany.
5CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and Technology, Barcelona, Spain.
6Josep Carreras Leukemia Research Institute (IJC), Barcelona, Spain.
7German Center for Neurodegenerative Diseases (DZNE), Munich, Germany.
Address correspondence to: Eloi Montañez Miralles, Department of Physiological Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Feixa Llarga s/n, 08907 L’Hospitalet del Llobregat, Spain. Phone: 34.934031970; Email: emontanez@ub.edu.
Find articles by Montanez, E. in: JCI | PubMed | Google Scholar
Published February 1, 2024 - More info
Published in Volume 9, Issue 5 on March 8, 2024
AbstractTAR DNA-binding protein 43 (TDP-43) is a DNA/RNA-binding protein that regulates gene expression, and its malfunction in neurons has been causally associated with multiple neurodegenerative disorders. Although progress has been made in understanding the functions of TDP-43 in neurons, little is known about its roles in endothelial cells (ECs), angiogenesis, and vascular function. Using inducible EC-specific TDP-43–KO mice, we showed that TDP-43 is required for sprouting angiogenesis, vascular barrier integrity, and blood vessel stability. Postnatal EC-specific deletion of TDP-43 led to retinal hypovascularization due to defects in vessel sprouting associated with reduced EC proliferation and migration. In mature blood vessels, loss of TDP-43 disrupted the blood-brain barrier and triggered vascular degeneration. These vascular defects were associated with an inflammatory response in the CNS with activation of microglia and astrocytes. Mechanistically, deletion of TDP-43 disrupted the fibronectin matrix around sprouting vessels and reduced β-catenin signaling in ECs. Together, our results indicate that TDP-43 is essential for the formation of a stable and mature vasculature.
Graphical Abstract
Introduction
The function of the central nervous system (CNS) depends on the formation and integrity of a complex vascular network that ensures adequate supply of oxygen and nutrients (1). Developing CNS is vascularized through angiogenic invasion of blood vessels from vascular networks outside the CNS (2). Maturing CNS blood vessels establish highly selective semipermeable cellular membranes called blood-brain barrier (BBB), blood–spinal cord barrier (BSCB), and blood-retina barrier (BRB) that control the flow of molecules and cells from the systemic circulation to the CNS parenchyma (3–5). These blood-CNS barriers are essential for maintaining the internal milieu of the CNS necessary for proper neuronal function and for protecting the CNS parenchyma from toxins and pathogens present in the circulation. Defects in the integrity of these vascular barriers are associated with inflammatory and immune responses that can trigger neuronal loss and underlie the ontogeny or progression of several neurological disorders, such as stroke, diabetic retinopathy, and neurodegenerative diseases (6, 7). The functional unit underlying blood-CNS barriers is the neurovascular unit, which consists of a monolayer of endothelial cells (ECs), sealed by tight junctions (TJs) and resting on a basement membrane, surrounded by pericytes and astrocytes (8). CNS vascularization and blood-CNS barrier development are controlled by multiple signaling systems, including integrin-mediated cell-matrix adhesion signaling and Wnt/β-catenin signaling (9–12). Upon binding to the extracellular matrix (ECM), integrins recruit adaptor and signaling proteins to their cytoplasmic domains to form focal adhesions (FAs), through which they relay signals into the cells. Integrin-mediated signals promote EC proliferation and migration during vessel sprouting and the stabilization of the nascent vasculature (9–11). On the other hand, binding of Wnt to the Frizzled/Lrp receptor complex on ECs stabilizes β-catenin in the cytoplasm, promoting its translocation to the nucleus, where it activates the transcription of target genes involved in vascular growth and barrier formation (12).
TAR DNA-binding protein 43 (TDP-43) is a highly and ubiquitously expressed DNA/RNA-binding protein that shuttles between the nucleus and the cytoplasm to regulate different aspects of RNA metabolism including transcription, splicing, stabilization, and transport (13). Loss of TDP-43 results in incorrect splicing of pre-mRNAs, often disrupting their translation and promoting nonsense-mediated decay (14). Cytoplasmic aggregation of TDP-43, accompanied by its nuclear clearance, is a common pathological hallmark of several neurodegenerative diseases, including amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), Alzheimer’s disease, and limbic-predominant age-related TDP-43 encephalopathy (LATE) (15–17). Cytoplasmatic TDP-43 aggregation is thought to compromise its function in the nucleus, ultimately leading to neurodegeneration. The identification of missense mutations in the TDP-43 encoding gene (Tardbp) in patients with ALS, mechanistically linked neurodegeneration to TDP-43 function (18–20). Although neurons and glia cells have been the main focus of research on TDP-43–associated pathologies, cytoplasmic aggregation, and nuclear clearance of TDP-43 has also been detected in the CNS vasculature of ALS and patients with FTD (21), suggesting that TDP-43 function in CNS vascular cells of these patients may also be affected. Interestingly, LATE, one of the most prevalent TDP-43 proteinopathy, has been linked to cerebral small vessel pathology (22). However, little is known about the role of TDP-43 in ECs and vascular homeostasis.
TDP-43 binds and regulates the metabolism of nearly 30% of the entire transcriptome in many different cell types and organ systems (23, 24); consequently, constitutive deletion of the murine Tardbp gene leads to periimplantation lethality, long before the vasculature is formed (25–27). Similarly, constitutive loss of TDP-43 in zebrafish is embryonic lethal (28). The lethality of TDP-43–deficient zebrafish embryos is associated with abnormal blood vessel patterning, neuronal defects, and muscle degeneration (28). The vascular malformation phenotype is very severe and is the first visible phenotype in TDP-43–deficient embryos, suggesting that TDP-43 is critical for EC function and proper vessel formation. Consistent with these findings, we have recently reported that TDP-43 controls intersegmental vessel growth in zebrafish embryos (29). However, the role of endothelial TDP-43 in postnatal CNS angiogenesis and blood-CNS barrier function has not been studied so far and remains poorly understood.
In this study, using EC-specific TDP-43–KO mice, human and mouse TDP-43–deficient ECs, and RNA-Seq analysis, we demonstrate that TDP-43 is essential for retinal sprouting angiogenesis, vascular network remodeling, blood-CNS barrier integrity, and CNS blood vessel stability. Consequently, mice lacking endothelial TDP-43 exhibit multiple hemorrhages and vascular degeneration in the brain and spinal cord. These vascular defects are associated with an inflammatory response with activation of microglia and astrocytes. At the cellular level, we show that TDP-43 is essential for EC proliferation and survival, as well as for cell-ECM adhesion and cell-cell junction stability. Mechanistically, we show that TDP-43 regulates fibronectin (FN) matrix assembly during sprouting angiogenesis and β-catenin signaling in ECs.
Overall, our results uncover novel roles for TDP-43 in the growth and stabilization of CNS vessels, and they identify endothelial–TDP-43 as a potential contributing factor to the vascular disorders and inflammation observed in patients diagnosed with TDP-43–associated diseases.
ResultsTDP-43 function in ECs is critical for retinal angiogenesis. In healthy cells, TDP-43 localizes predominantly in the nucleus, where it controls mRNA processing (13). To determine whether TDP-43 is expressed in ECs of developing vessels, we analyzed TDP-43 protein expression pattern in retinas of P7 and P16 mice and in brain and liver sections of P16 mice, and we found that it is expressed in ECs in vivo and localizes mainly in the nucleus (Figure 1A and Supplemental Figure 1, A–C; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.177819DS1). The analysis also showed that TDP-43 was evenly expressed throughout the entire vasculature, with clear expression in ECs of capillaries, arteries, and veins in the remodeling vascular plexus as well as in sprouting ECs at the vessel growth front (Supplemental Figure 1, A and B). Immunostaining of TDP-43 in cultured human umbilical vein ECs (HUVECs) showed that human ECs also express TDP-43 mainly in the nucleus (Supplemental Figure 1D). To uncover the function of TDP-43 in ECs, we crossed TDP-43fl/fl mice (30) with tamoxifen-inducible Cdh5CreERT2 mice (31). Tamoxifen administration in newborn or 8-week-old mice resulted in sudden death of TDP-43fl/fl;Cdh5CreERT2 (hereafter referred to as TDP-43iΔEC), but not of TDP-43fl/fl (control) mice, mice between 11 and 22 days after the first tamoxifen injection (Supplemental Figure 1E). These results indicate that endothelial TDP-43 is critical for vascular function.
Figure 1Postnatal endothelial cell–specific TDP-43 inactivation induces sprouting defects in the retina. (A) Confocal high-magnification image of P7 WT mouse retinas stained for the endothelial membrane marker isolectin B4 (IB4) (green), TDP-43 (red), and cell nuclei labeled with Hoechst (blue). Dotted red lines highlight the endothelial border of the blood vessels. White rectangle indicate the magnified region of interest (ROI). Scale bar: 20 μm. (B) Low-magnification overview images of whole-mount P7 control and TDP-43iΔEC retinas stained for IB4 to visualize vascularization and vascular outgrowth of the primary vascular plexus. Note reduction of vessel radial outgrowth in TDP-43iΔEC retinas compared with control retinas. Scale bar: 500 μm. (C) Confocal high-magnification images of control and TDP-43iΔEC P7 retinas stained for IB4 (majenta) and the endothelial cell (EC) nuclear marker ERG (green). Note the clustering of ECs at the vessel growth front of TDP-43iΔEC retinas. Arteries (A) and veins (V) are indicated. Scale bar: 200 μm. (D) Confocal high-magnification images of the angiogenic growth front of P7 control and TDP-43iΔEC retinas stained for IB4. Red dots indicate endothelial sprouts. (E) Quantification of vascular parameters in P7 control and TDP-43iΔEC retinas as indicated. Data are shown as mean ± SEM. ***P < 0.001. Mann-Whitney U test. (F) Confocal high-magnification images of control and TDP-43iΔEC P7 retinas stained for IB4 (green), the cell proliferation marker Ki67 (red) and ERG (blue). Scale bar: 50 μm. (G) Quantification of the relative ECs proliferation ratio. Data are shown as mean ± SEM. *P < 0.05. Mann-Whitney U test.
Next, to uncover the role of endothelial TDP-43 in sprouting angiogenesis in vivo, we analyzed postnatal retinal vascularization. During the first postnatal week, a primary vascular plexus grows within the retinal ganglion layer (32). Following a FN template, blood vessels expand from the optic stalk toward the retinal periphery and establish a superficial vascular plexus around P8. Therefore, we induced TDP-43 gene deletion in EC in newborns mice via tamoxifen administration between P1 and P3, and we analyzed retinal vascularization at P7. Western blot analysis of the lung lysates from P6 TDP-43fl/fl and TDP-43iΔEC mice showed downregulation of TDP-43 expression when compared with lysates from Cre-negative control littermates (Supplemental Figure 2A). TDP-43 staining of P7 retinas confirmed a strong decrease in TDP-43 expression in ECs, but not in surrounding non-ECs, of TDP-43iΔEC mice (Supplemental Figure 2B). Isolectin B4 labeling of control and TDP-43iΔEC retinas showed delayed radial vessel outgrowth, decreased vessel branching, and reduced endothelial sprouting at the vessel growth front in TDP-43iΔEC retinas compared with control retinas (Figure 1, B–E, and Supplemental Figure 2C). Immunostaining of ERG, a specific nuclear marker of ECs, showed a decreased number of ECs in TDP-43iΔEC retinas compared with control retinas and irregular distribution of ECs at the angiogenic front of TDP-43iΔEC retinas (Figure 1, C and E). EC proliferation, assessed by Ki67 staining, decreased with the loss of TDP-43 (Figure 1, F and G). Together, these observations indicate that TDP-43 function in the ECs is required for endothelial sprouting and proliferation, and they argue for a role of TDP-43 in EC migration.
Impaired endothelial FN matrix assembly in TDP-43iΔEC retinas. During retinal vascular development, ECs assemble a fibrillar matrix of FN that promotes vessel sprouting and branching (33). FN staining of P7 control and TDP-43iΔEC retinas showed severe morphological alterations in the FN matrix around the sprouting vessels of TDP-43iΔEC mice (Figure 2A). Whereas in control retinas, sprouting vessels were associated with a dense FN matrix, organized as a fibrillar meshwork with thin FN fibrils running longitudinally along the vessel axis, sprouting vessels of TDP-43iΔEC retinas were associated with a disorganized FN matrix, with sparse fibrils and multiple irregular aggregates of FN attached to the blood vessels (Figure 2A). During retinal angiogenesis, FN is mainly expressed by ECs and astrocytes (33, 34). To address whether FN matrix defects around TDP-43iΔEC vessels were consequence of impaired assembly of EC-derived FN, we isolated ECs from TDP-43fl/fl and TDP-43iΔEC mice, treated them with 4-hydroxytamoxifen (4-OHT), and assessed FN matrix assembly. Forty-eight hours after of 4-OHT treatment, TDP-43iΔEC cells showed decreased TDP-43 protein levels compared with control cells (Figure 2B). Similar to the situation in TDP-43iΔEC retinas, FN staining of control and TDP-43–depleted ECs showed disturbed FN matrix assembly in the absence of TDP-43 (Figure 2C). While control cells assembled a dense meshwork of thin FN fibrils, TDP-43–depleted cells assembled a disorganized and sparse FN matrix with thicker fibrils and irregular aggregates (Figure 2C). In addition, staining for paxillin, a marker of focal complexes and FAs, showed that FN matrix defects in TDP-43–deficient ECs correlated with reduced FA formation (Figure 2D). Next, to determine whether TDP-43 also regulates cell-matrix adhesion in human ECs, we knocked down TDP-43 protein expression in cultured HUVECs (Figure 2E). While control ECs showed multiple focal complexes along the membrane edges in the vicinity of FA, TDP-43–depleted ECs showed few focal complexes, and paxillin was detected mainly in aberrant matrix adhesion structures located at the cell periphery (Figure 2F). Morphological analysis showed reduced cell spreading of TDP-43–depleted ECs compared with control cells (Figure 2G). Taken together, these results indicate that deletion of TDP-43 impairs endothelial FN matrix assembly during sprouting angiogenesis and impaired EC-ECM adhesion.
Figure 2TDP-43 is required for FN matrix assembly and cell-matrix adhesion. (A) Confocal high-magnification images of P7 control and TDP-43iΔEC retinas stained for IB4 (green) and fibronectin (FN, red). White rectangles indicate magnified regions. Note the FN fibrils (arrows) associated to control vessels and the presence of irregular FN aggregates (arrowheads) in TDP-43iΔEC vessels. A indicates magnified region of image. Scale bar: 20 μm. (B) Western blot analysis of TDP-43 in total lysates of TDP-43fl/fl and TDP-43fl/fl;Cadh5Cre mouse lungs ECs (mLECs) 48 hours after 4-hydroxytamoxifen (4-OHT) treatment. Tubulin was used as a loading control. B indicates magnified region of image. (C) Confocal high-magnification images of TDP-43fl/fl and TDP-43fl/fl;Cadh5Cre mLECS stained for EC marker VE-cadherin (VEcad, green), FN (red), and Hoechst (nuclei, blue). Note the abnormal FN aggregates (arrowheads) in TDP-43fl/fl;Cadh5Cre mLECs. Scale bar: 10 μm. (D) Confocal high-magnification images of TDP-43fl/fl and TDP-43fl/fl;Cadh5Cre mLECS stained for VEcad (green), the focal adhesion maker Paxillin (Pax, red), and Hoechst (nuclei, blue). Scale bar: 10 μm. (E) Western blot analysis of TDP-43 in total lysates of human umbilical vein ECs (HUVECs) transfected with either control (scramble, Scr) or TDP-43 small interfering RNAs (siRNAs). GAPDH was using as a loading control. Graph shows the quantification of the mean TDP-43 protein expression levels corrected for background and normalized to expression in Scr and TDP-43 siRNA–transfected HUVECs. Data are shown as mean ± SEM. ***P < 0.001. One-way ANOVA. (F) Confocal high-magnification images of control and TDP-43–depleted HUVECs stained for Pax (green), F-actin (red), and TDP-43 (blue). Scale bar: 20 μm. (G) Quantification of the ratio between the longest side and the shortest side (cellular elongation index) of the ECs in control and TDP-43–depleted HUVECs. Data are shown as mean ± SEM. ***P < 0.001. One-way ANOVA.
Endothelial TDP-43 is required for vascular remodeling and EC ingression into the deep retina. After the formation of the superficial vascular plexus, endothelial sprouts descend vertically through the inner retina to form 2 additional capillary layers, the deep and intermediate vascular plexus (35). The deep vascular plexus develops in the outer plexiform layer (OPL) and reaches the periphery of the retina around P12, while the intermediate vascular plexus forms in the inner plexiform layer (IPL) between P12 and P15 (35). To further characterize the role of TDP-43 in retinal vessel growth, we induced the deletion of TDP-43 in ECs from P5 to P7 in TDP-43iΔEC mice and analyzed retinal vascularization at P16. The analysis revealed severe defects in deep retina vascularization in TDP-43iΔEC mice (Figure 3, A–C). TDP-43iΔEC retinas showed large avascular zones at the periphery of the OPL and numerous round aggregates of ECs in the vicinity of the superficial vascular plexus in the nerve fiber layer (NFL) (Figure 3, A and B, and Supplemental Figure 3A). In addition, vertical sprouts in the superficial vascular plexus, as well as vessel density and branching in the OPL, were reduced in TDP-43iΔEC retinas compared with control retinas (Figure 3C). From P9 onward and in parallel to the formation of vascular layers in the deep retina, the superficial vascular plexus undergoes extensive remodeling, eventually establishing an efficient and mature hierarchical vascular network (35). Superficial vascular plexus remodeling involves regression of excessive vessels, pruning of side branches, and reduction of vessel diameter (11). Quantification analysis showed increased vessel branching and capillary vessel diameter in the superficial vascular plexus of TDP-43iΔEC retinas compared with control retinas (Figure 3C). Regressing ECs leave empty basal membrane sleeves rich in collagen-IV (36). Staining of collagen-IV revealed abundant basement membrane segments devoid of ECs in TDP-43iΔEC retinas but not in control retinas, arguing that vessel remodeling is also perturbed in TDP-43iΔEC mice (Figure 3D and Supplemental Figure 3B). Together, these findings establish that endothelial TDP-43 is indispensable for endothelial sprouting into the deep retina and for the formation of a stable and mature vascular network.
Figure 3TDP-43 is indispensable for vessel remodeling and vascular plexus formation in the deeper retina. (A) Confocal high-magnification images of P16 control and TDP-43iΔEC retinas stained for IB4. Dotted line highlights the avascular area in the deep vascular plexus of TDP-43iΔEC retinas. Scale bar: 200 μm. (B) Optical sections of Z-stacked confocal high-magnification images were divided to illustrate the vascular plexus in the nerve fiber layer (NFL), inner plexiform layer (IPL), and outer plexiform layer (OPL). Note the bulging endothelial structures associated with the dilated vessels in TDP-43iΔEC retinas. Scale bar: 20 μm. (C) Quantification of vascular parameters in the control and TDP-43iΔEC P16 retinas as indicated. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001. Mann-Whitney U test. (D) Confocal high-magnification images of control and TDP-43iΔEC P16 retinas stained for IB4 (green) and collagen-IV (red). White squares indicate the magnified ROIs. Note the abundance of empty collagen-IV sleeves (arrowheads) in TDP-43iΔEC retinas. Arteries (A) and vein (V) are indicated. Scale bar: 50 μm.
Endothelial deletion of TDP-43 impairs cell-cell junction morphology and BRB integrity. VE-cadherin staining of P16 control and TDP-43iΔEC retinas revealed several morphological and junctional alterations in the vasculature of TDP-43iΔEC mice (Figure 4A). In vessels of control retinas, ECs were elongated and showed a spindle-like morphology with continuous, straight adherens junctions (AJs) running longitudinally along the vessel axis (Figure 4A). In contrast, ECs from TDP-43iΔEC vessels showed a less elongated round morphology, and their AJs appeared irregular with a zig-zagged distribution of VE-cadherin (Figure 4A). In addition, whereas ECs from control capillaries showed linear and thin AJs, ECs from TDP-43iΔEC capillaries showed tortuous junctions with diffuse and discontinuous VE-cadherin stain (Figure 4A). Shear stress regulates AJ morphology (37). To determine whether the altered AJ morphology observed in TDP-43iΔEC vessels was independent of shear stress signals, we analyzed AJ morphology in mouse and human EC cultures under static conditions. Similar to the situation in TDP-43iΔEC mice, VE-cadherin staining of control and TDP-43–depleted ECs showed altered AJ morphology in TDP-43–depleted ECs (Figure 4, B and C). In contrast to control cells, in which a high fraction of VE-cadherin stain was distributed in a linear, continuous pattern along cell-cell borders, junctional VE-cadherin signals in TDP-43–depleted ECs were distributed in discontinuous finger-like structures oriented perpendicular to cell-cell borders (Figure 4, B and C). Quantitative analysis showed decreased levels of linear AJs in TDP-43–depleted ECs compared with control ECs (Figure 4D). Linear AJs, also called stable AJs, are associated with thin actin filaments running parallel to cell membranes, whereas finger-like AJs are connected to thick stress fibers (37). F-actin staining of control and TDP-43–depleted ECs showed that knockdown of TDP-43 also induces changes in the actin cytoskeleton, which were characterized by the formation of a prominent cortical F-actin belt and an increase in short radial actin bundles oriented perpendicular to interrupted AJs (Figure 4C and Supplemental Figure 4A). Similar F-actin defects were observed in ECs of TDP-43iΔEC retinas (Supplemental Figure 4B). Together, these results indicate that TDP-43 regulate AJ stability independently of shear stress signals.
Figure 4Loss of TDP-43 disrupts endothelial adherens junctions. (A) Confocal high-magnification images of P16 control and TDP-43iΔEC retinas stained for IB4 (green) and VE-cadherin (VEcad; red). White squares indicate the magnified ROIs. Arrows highlight vessel segments with diffuse punctuated VEcad stain. Scale bar: 40 μm. (B) Confocal high-magnification images of TDP-43fl/fl and TDP-43fl/fl;Cadh5Cre mLECS stained for VEcad (red) and Hoechst (nuclei, blue). Scale bar: 10 μm. (C) Confocal high-magnification images of HUVECs transfected with Scr or TDP-43 siRNAs stained for F-actin (green), VEcad (red), and TDP-43 (blue). White rectangles indicate the magnified ROIs. Scale bar: 40 μm. (D) Quantification of percentages of continuous (stable) and discontinuous (remodeling) adherens junctions in control and TDP-43–depleted HUVECs. Data are shown as mean ± SEM. *P < 0.05, **P < 0.01. One-way ANOVA.
CNS ECs form TJs to establish the blood-CNS barrier (3). The distribution of claudin-5, a structural component of the endothelial TJs, was also altered in TDP-43iΔEC retinal vessels (Figure 5A). Whereas in control vessels, claudin-5 decorated cell-cell junctions, in TDP-43iΔEC vessels, claudin-5 was found in punctate cytoplasmic clusters, and its junctional distribution was diffuse and discontinuous (Figure 5A). Morphological defects in the cell-cell junctions of the TDP-43iΔEC vessels were accompanied by elevated staining signal of the EC fenestration component Plvap (Figure 5, B and C), a widely used marker of blood-CNS barrier disruption (11, 38). Consistent with the cell-cell junction defects, freshly isolated TDP-43iΔEC retinas showed extensive hemorrhages (Figure 5D). Immunostaining for the RBC marker Ter119 confirmed RBC extravasation in TDP-43iΔEC retinas but not in control retinas (Figure 5D). These findings indicated that deletion of endothelial TDP-43 impairs AJ stability and TJ morphology, compromises BRB function, and causes hemorrhages.
留言 (0)