Dual function of perivascular fibroblasts in vascular stabilization in zebrafish
Authors:
Arsheen M. Rajan aff001; Roger C. Ma aff001; Katrinka M. Kocha aff001; Dan J. Zhang aff002; Peng Huang aff001
Authors place of work:
Department of Biochemistry and Molecular Biology, Cumming School of Medicine, Alberta Children’s Hospital Research Institute, University of Calgary, Calgary, Alberta, Canada
aff001; Department of Biological Sciences, University of Calgary, Calgary, Alberta, Canada
aff002
Published in the journal:
Dual function of perivascular fibroblasts in vascular stabilization in zebrafish. PLoS Genet 16(10): e1008800. doi:10.1371/journal.pgen.1008800
Category:
Research Article
doi:
https://doi.org/10.1371/journal.pgen.1008800
Summary
Blood vessels are vital to sustain life in all vertebrates. While it is known that mural cells (pericytes and smooth muscle cells) regulate vascular integrity, the contribution of other cell types to vascular stabilization has been largely unexplored. Using zebrafish, we identified sclerotome-derived perivascular fibroblasts as a novel population of blood vessel associated cells. In contrast to pericytes, perivascular fibroblasts emerge early during development, express the extracellular matrix (ECM) genes col1a2 and col5a1, and display distinct morphology and distribution. Time-lapse imaging reveals that perivascular fibroblasts serve as pericyte precursors. Genetic ablation of perivascular fibroblasts markedly reduces collagen deposition around endothelial cells, resulting in dysmorphic blood vessels with variable diameters. Strikingly, col5a1 mutants show spontaneous hemorrhage, and the penetrance of the phenotype is strongly enhanced by the additional loss of col1a2. Together, our work reveals dual roles of perivascular fibroblasts in vascular stabilization where they establish the ECM around nascent vessels and function as pericyte progenitors.
Keywords:
Blood vessels – Cell differentiation – Collagens – Fibroblasts – Hemorrhage – Heterozygosity – Pericytes – Zebrafish
Introduction
The vascular system is crucial to the survival of vertebrates. Blood vessels must rapidly expand and contract in response to systemic cues, but also be sturdy to withstand the stress of blood flow. To maintain their integrity, blood vessels are supported by a highly specialized perivascular architecture comprised of blood vessel associated cells and the surrounding extracellular matrix (ECM) [1, 2]. Compromised vascular integrity can result in devastating human diseases, such as aneurysms, vascular malformations, and hemorrhagic strokes [3–6]. However, how blood vessels are stabilized by different perivascular components is still poorly understood.
The prevailing model is that vascular stability is maintained at three different levels: endothelial cells, mural cells and the surrounding ECM [2]. First, blood vessels are lined by endothelial cells. Adherens and tight junctions between endothelial cells provide the primary barrier to passage of fluids, cells, and macromolecules between blood and tissues [7]. Second, vascular smooth muscle cells (vSMCs) and pericytes, collectively known as mural cells, closely interact with and stabilize the endothelium [8]. vSMCs form a continuous protective sheath around large diameter blood vessels, whereas pericytes are solitary cells partially covering smaller blood vessels such as capillaries. Defects in mural cell specification or recruitment result in widespread vascular leakage and early lethality in mice [9–13]. Lastly, the vascular ECM provides structural support for blood vessels [1, 14]. Mutations in ECM proteins, such as collagens and laminins, lead to early death due to blood vessel rupture in mice [15–18]. In humans, defects in fibril forming collagens (type I, III and V), as well as molecules involved in collagen modification and processing, have been associated with Ehlers-Danlos syndrome (EDS) [19]. EDS patients are characterized by a significant vascular fragility that often leads to spontaneous rupture of blood vessel walls. Thus, work in both animal models and human patients shows the importance of collagen in maintaining vascular stability. However, one unresolved question is how newly formed blood vessels are stabilized in the early time window before the differentiation of mural cells.
vSMCs and pericytes are the most well-studied blood vessel associated cells and are classically defined by the expression of alpha smooth muscle actin (ACTA2) and platelet-derived growth factor receptor b (PDGFRb), respectively [20, 21]. Lineage tracing in mice and chick-quail chimeras demonstrate a heterogeneous developmental origin of mural cells [20–23]. For example, vSMCs covering the trunk aorta originate from either the neural crest or the sclerotome of the somite, depending on the axial position. Other than mural cells, several types of perivascular cells have been identified, including adventitial cells [24], fibroblasts [25], immune cells [26], and astrocytes [27]. However, whether these different cell types play roles in vascular stabilization is unknown.
Recent single cell transcriptional profiling studies in mice have revealed the presence of perivascular fibroblast-like cells in the adult mouse brain that are not labeled by classical mural cell markers [28–31]. In contrast to mural cells, perivascular fibroblast-like cells adhere loosely to blood vessels and show robust expression of many ECM components, such as collagens. Previous studies have described similar cell populations in the mouse central nervous system that contribute to fibrotic scar formation after injury [32–35]. However, where these perivascular fibroblast-like cells originate from during development and how they regulate blood vessel development and stabilization have not been explored.
Zebrafish is a powerful model to study human cardiovascular diseases [36–38]. The organization, development and function of the vasculature, including mural cells, are highly conserved between zebrafish and mammals [36, 39–41]. In the zebrafish trunk, intersegmental vessels (ISVs) sprout from the dorsal aorta (DA) at around 24 hpf (hours post fertilization) and fully establish the metameric pattern by 36 hpf [42]. Interestingly, pericytes, the primary mural cells along the ISVs, only emerge at 60 hpf [43]. This raises the question how nascent blood vessels are stabilized prior to the differentiation of pericytes. In our work, we describe a novel population of collagen-expressing perivascular fibroblasts that become closely associated with ISVs soon after vessel formation. Perivascular fibroblasts originate from the sclerotome and are distinct from pericytes in their morphology, distribution, and marker expression. Using a combination of in vivo live imaging, genetic ablation and CRISPR mutant analysis, we demonstrate that perivascular fibroblasts play dual roles in stabilizing nascent ISVs by regulating the vascular ECM and later functioning as pericyte progenitors. Together, our work provides important insights into the development of perivascular support structures and suggests a molecular and cellular basis of Ehlers-Danlos syndrome.
Results
Perivascular fibroblasts express several collagen genes
We previously developed a sclerotome-specific transgenic line nkx3.1:Gal4; UAS:Nitroreductase-mCherry (nkx3.1NTR-mCherry, similar designations are used for all Gal4/UAS transgenic lines in this paper) [44]. This reporter labels the initial sclerotome domains as well as their descendants due to the perdurance of the mCherry protein. Examination of the nkx3.1NTR-mCherry line in combination with an endothelial reporter, kdrl:EGFP, revealed that a population of mCherry+ cells was closely associated with intersegmental vessels (ISVs) at 48 hpf (Fig 1A). This result suggests that the sclerotome contributes to a population of perivascular cells well before the appearance of pericytes at 60 hpf [43]. Interestingly, nkx3.1NTR-mCherry-expressing perivascular cells at 52 hpf were also labeled by a pan-fibroblast reporter, col1a2:GFP [44] (Fig 1B). On average, 37% of mCherry+ perivascular cells were positive for col1a2:GFP (S1A Fig), which likely reflects the mosaic nature of the col1a2:GFP line. Consistent with this result, fluorescent in situ hybridization showed that ISV-associated perivascular cells co-expressed several ECM genes, including col1a2 (collagen 1a2) and col5a1 (collagen 5a1) at 48 hpf (Figs 1C and S1B). These collagen-expressing perivascular cells are reminiscent of perivascular fibroblast-like cells recently identified in the adult mouse brain by single cell RNA sequencing [28–31]. We therefore refer to this population of collagen-expressing ISV-associated cells as ‘perivascular fibroblasts’.
Perivascular fibroblasts originate from the sclerotome
The expression of nkx3.1NTR-mCherry in perivascular fibroblasts suggests that the sclerotome is the embryonic source of these cells. We have previously shown that the sclerotome in zebrafish has a bipartite organization with a ventral domain and a smaller dorsal domain [44]. The ventral sclerotome domain further gives rise to a population of notochord associated cells. At 24 hpf, the sclerotome consists of three compartments: the dorsal sclerotome domain, the ventral sclerotome domain, and notochord associated cells derived from the ventral domain (Fig 2A). To determine the contribution of these three compartments to perivascular fibroblasts, we performed confocal time-lapse imaging in nkx3.1NTR-mCherry; kdrl:EGFP embryos. The nkx3.1NTR-mCherry reporter labels sclerotome progenitors and their progeny, while the endothelial specific kdrl:EGFP line allows us to visualize the ISVs. Perivascular fibroblasts can be distinguished from tenocytes based on their perivascular locations as well as the lack of long tree-like processes extending into the intersomitic space, a characteristic unique to tenocytes [44]. The trunk region of embryos (somite 12–18) was imaged laterally at 6–9 minute intervals starting at 25 hpf, immediately after the emergence of ISV sprouts (Fig 2B and S1 Video). At 30 hpf, most ISVs became visibly lumenized. During ISV sprouting and lumenization, mCherry+ cells can be observed associated with the kdrl:EGFP-labeled endothelium (Figs 2B and S2A; S1 Video). By the end of the movie at 49.5 hpf, mCherry+ perivascular fibroblasts ‘decorated’ the entire ISVs along the dorsal-ventral (D-V) axis. Interestingly, in the absence of ISVs, numerous mCherry+ sclerotome derived cells can still populate the interstitial space between muscles, the spinal cord and the notochord (S2B Fig), suggesting that the migration and proliferation of sclerotome progenitors and their descendants are independent of the vasculature. By retrospective cell tracing, we showed that perivascular fibroblasts can be traced to all three compartments of the sclerotome. Of 170 perivascular fibroblasts traced, 28 cells originated from the dorsal sclerotome domain (16%), 112 from the ventral sclerotome domain (66%), and 30 from notochord associated cells (18%). Since notochord associated cells are derived from the ventral sclerotome domain, this result suggests that the ventral sclerotome is the main contributor of perivascular fibroblasts (142/170 cells, 84%).
To better compare the contribution of sclerotome progenitors from different compartments, we subdivided perivascular fibroblasts into two domains based on their final D-V positions along the ISVs. The position of the horizontal myoseptum was used as a landmark to define perivascular fibroblasts at dorsal or ventral ISVs (Fig 2A). The 170 perivascular fibroblasts can be traced back to 122 sclerotome progenitors. These sclerotome progenitors typically underwent 1–2 cell divisions over the 24-hour period to give rise to multiple daughter cells, at least one of which became associated with the neighboring ISV. We categorized sclerotome progenitors into three groups based on the final positions of perivascular fibroblasts they generated: 1) only at the dorsal ISV, 2) only at the ventral ISV, or 3) at both dorsal and ventral ISV (Fig 2C). Sclerotome progenitors in the dorsal domain gave rise to exclusively perivascular fibroblasts along dorsal ISVs (20/20 cells, 100%) (Fig 2C). Similarly, the majority of sclerotome progenitors around the notochord (i.e., sclerotome derived notochord associated cells) (21/23 cells, 91%) contributed to only perivascular fibroblasts at the dorsal ISV (Fig 2C). By contrast, 48% of sclerotome progenitors in the ventral domain (38/79 cells) gave rise to only perivascular fibroblasts at the ventral ISV, 35% of ventral sclerotome progenitors (28/79 cells) generated only dorsal perivascular fibroblasts, while the remaining 16% of ventral progenitors (13/79 cells) contributed to perivascular fibroblasts at both dorsal and ventral ISV regions (Fig 2C). Similar to the contribution of tendon fibroblasts (tenocytes) [44], the dorsal sclerotome generates only dorsally positioned perivascular fibroblasts, whereas the ventral sclerotome contributes to perivascular fibroblasts along the entire D-V axis of ISVs (the combined bar in Fig 2C). Moreover, perivascular fibroblasts along a given ISV were always derived from the sclerotome of the same overlying somite (170/170 cells, 100%). Together, our results demonstrate that perivascular fibroblasts originate from all three sclerotome compartments in a stereotypic manner.
Perivascular fibroblasts are distinct from pericytes
Mural cells, including pericytes and vSMCs, are known blood vessel associated support cells. We next asked whether perivascular fibroblasts represent a different perivascular cell population from mural cells. In the zebrafish trunk, pericytes are associated with ISVs, while vSMCs are localized to larger vessels such as the dorsal aorta [40, 41, 43]. Platelet derived growth factor (PDGF) signaling is known to play an important role in pericyte recruitment and the platelet derived growth factor receptor-beta (pdgfrb) is a well-established pericyte marker [43, 45, 46]. We utilized the pdgfrb:GFP [47] and pdgfrb:Gal4FF; UAS:NTR-mCherry (referred to as pdgfrbNTR-mCherry) [48] lines to label pericytes. Interestingly, while both pdgfrb:GFP and pdgfrbNTR-mCherry lines labeled pericytes (GFPhighmCherry+ expression) at 4 dpf, pdgfrb:GFP but not pdgfrbNTR-mCherry also marked perivascular fibroblasts (GFPlowmCherry- expression) at 2 and 4 dpf (days post fertilization) (S3A and S3B Fig). Fully differentiated pericytes can be reliably identified based on their elevated pdgfrb:GFP expression (2.5 fold) compared to perivascular fibroblasts (S3C Fig). The expression of pdgfrb:GFP in sclerotome descendants is consistent with the previous report that the sclerotome expresses pdgfrb [49], suggesting that pdgfrb:GFP is a more sensitive reporter than the pdgfrbNTR-mCherry line. Indeed, pdgfrbNTR-mCherry labeled about 62% of pdgfrb:GFP-positive pericytes at 4 dpf (S3D Fig).
Using the two pdgfrb lines for pericytes as well as the nkx3.1NTR-mCherry and col1a2:GFP reporters for perivascular fibroblasts, we found that perivascular fibroblasts and pericytes represent distinct cell populations. First, they showed different developmental timing and cell numbers. Consistent with our time-lapse movies, nkx3.1NTR-mCherry-positive perivascular fibroblasts were present along ISVs at 2 dpf (Fig 3A). They were equally distributed along arterial and venous ISVs, and remained largely constant in number (ranging from 9.7 to 10.6 per ISV) from 2 to 4 dpf (Fig 3A). By contrast, pericytes marked by high level of pdgfrb:GFP expression did not appear on ISVs until 3 dpf. They were more concentrated along arterial ISVs than venous ISVs, and increased in cell number to 2.2 per aISV and 1.0 per vISV by 4 dpf (Fig 3A), similar to what has been previously described [43]. Second, perivascular fibroblasts and pericytes showed distinct marker expression (Fig 3B). Perivascular fibroblasts labeled by col1a2:GFP did not express the pdgfrbNTR-mCherry reporter at 4 dpf. Conversely, pdgfrbNTR-mCherry-positive pericytes at 4 dpf showed minimal col1a2:GFP expression. Further, while perivascular fibroblasts appeared globular, pericytes were more elongated with long cellular processes (Fig 3B). Perivascular fibroblasts were also more laterally located along ISVs compared to pericytes, with pericyte processes often sandwiched between the neighboring perivascular fibroblasts and the endothelium (Fig 3B). To further examine the morphological differences between perivascular fibroblasts and pericytes, we performed high resolution confocal microscopy at the single cell resolution at 3 dpf. pdgfrbNTR-mCherry-positive pericytes showed the typical pericyte morphology: small and flat cell bodies with several elongated cellular processes that extended longitudinally along and tightly wrapped around the ISV, labeled by kdrl:EGFP (Fig 3C). Since perivascular fibroblasts are more abundant along ISVs compared to pericytes, we took advantage of a mosaic col1a2:Gal4; UAS:Kaede line (referred to as col1a2Kaede) [50] to sparsely label single perivascular fibroblasts for easier visualization. Unlike the tight association between pericyte and the endothelium (10/10, Fig 3C), 36% of perivascular fibroblasts (5/14) showed visible gaps between the cell body and the underlying endothelium (Fig 3D), suggesting a looser association. In contrast to pericytes, perivascular fibroblasts had a globular morphology with short processes that wrapped diametrically around the associated ISV in an ‘awkward hug’ (Fig 3D). Quantification of cellular morphologies revealed key differences between these two cell types (Fig 3E–3I). In contrast to perivascular fibroblasts, pericytes displayed a significantly larger aspect ratio (Fig 3F), suggesting that they were more elongated along the long axis of the ISVs. Moreover, perivascular fibroblasts possessed more but shorter cellular processes compared to pericytes (Fig 3G–3I). Taken together, our results suggest that perivascular fibroblasts and pericytes represent two distinct populations of perivascular cells with unique marker expression, morphology, distribution and developmental timing.
A sub-population of perivascular fibroblasts functions as pericyte progenitors
While occupying the similar perivascular space, perivascular fibroblasts appear along ISVs at least one day before pericytes (Fig 3A). We hypothesized that perivascular fibroblasts act as progenitors for ISV pericytes. To test this possibility, we conducted time-lapse imaging of pdgfrbNTR-mCherry; col1a2:GFP embryos from 54 to 73 hpf during which pericytes started to appear on ISVs. Interestingly, a subset of col1a2:GFP-positive perivascular fibroblasts gradually lost GFP expression, switched on the pericyte reporter pdgfrbNTR-mCherry, and developed long cellular processes characteristic of pericytes at this stage (Fig 4A and S2 Video). Indeed, of 15 newly formed pericytes traced in our time-lapse movies, 60% were derived from col1a2:GFP-expressing perivascular fibroblasts (9/15 cells). This result suggests that a sub-population of perivascular fibroblasts can differentiate into pericytes. Since col1a2 transgenic lines such as col1a2:GFP are mosaic (S1A and S3E Fig) [50], the number above likely under-estimates the proportion of pericytes derived from perivascular fibroblasts. To further test our model, we performed time-lapse experiments in nkx3.1NTR-mCherry; pdgfrb:GFP embryos. Similar to movies in pdgfrbNTR-mCherry; col1a2:GFP embryos, cell tracing in nkx3.1NTR-mCherry; pdgfrb:GFP embryos showed that about 8% of mCherry+GFPlow perivascular fibroblasts (100/1208 cells) at 54 hpf slowly upregulated GFP expression, and extended pericyte-like cellular processes wrapping around the ISV (S3 Video). Of 125 newly formed GFPhigh pericytes, 90% of them (113/125 cells) were derived from mCherry+ cells, including perivascular fibroblasts residing along the ISV (101/113 cells), cells migrating from near the dorsal aorta (5/113 cells), as well as some cells with unclear origin due to incomplete imaging (7/113 cells). This result confirms that a sub-population of perivascular fibroblasts functions as pericyte precursors. Since perivascular fibroblasts originate from the sclerotome, we conclude that most pericytes along ISVs are also derived from the sclerotome.
To further test if perivascular fibroblasts serve as pericyte precursors, we performed genetic ablation of perivascular fibroblasts using the col1a2:Gal4; UAS:NTR-mCherry line (referred to as col1a2NTR-mCherry). Nitroreductase (NTR) converts metronidazole (MTZ), an otherwise harmless prodrug into a cytotoxic compound, resulting in the death of NTR-expressing cells [51]. To ablate perivascular fibroblasts prior to pericyte differentiation, col1a2NTR-mCherry; pdgfrb:GFP embryos were treated with MTZ from 25 to 54 hpf, after which the drug was washed off and embryos were imaged for pericytes at 4 dpf (Fig 4B). In contrast to water-treated controls, we observed a complete loss of perivascular fibroblasts in MTZ-treated embryos, with only residual mCherry+ debris visible in the trunk (Fig 4C). Interestingly, we observed a substantial loss of pericytes on both arterial ISVs (aISVs) and venous ISVs (vISVs) in MTZ-treated embryos compared to water-treated controls (Fig 4D). This result is consistent with our model that perivascular fibroblasts function as pericyte precursors on both arteries and veins as depletion of perivascular fibroblasts leads to universal decrease in pericyte number on all ISVs. Together, our work suggests that the sclerotome generates perivascular fibroblasts, some of which further differentiate into pericytes along the ISVs.
Loss of perivascular fibroblasts results in dysmorphic ISVs
Our work suggests that perivascular fibroblasts function as progenitors to generate pericytes to support the vasculature. However, only a small fraction of perivascular fibroblasts actually differentiates into pericytes, raising question about the function of the remaining perivascular fibroblasts. Moreover, the emergence of perivascular fibroblasts occurs concurrently with ISV development (Figs 2B and S2A; S1 Video), well before pericyte differentiation at 60 hpf [43]. These observations raise the possibility that perivascular fibroblasts play an early role in stabilizing nascent blood vessels prior to pericyte formation. To test this idea, we examined the impact of early ablation of perivascular fibroblasts on ISV development using the nitroreductase-based system. nkx3.1NTR-mCherry; kdrl:EGFP embryos were incubated in water or MTZ for a 24 hour period starting at 38 hpf when ISV formation was complete and blood flow had commenced in ISVs (Fig 5A). ISV morphology was examined after the drug treatment at 62 hpf. Water-treated control embryos showed stereotypical ISVs with many mCherry+ perivascular fibroblasts (Fig 5B). MTZ treatment resulted in a complete loss of perivascular fibroblasts, with only some mCherry+ cell debris remaining (Fig 5B). Compared to control ISVs, most ISVs in ablated embryos showed visible distortions in their diameters, shrinking and dilating dramatically at different points along their length (Fig 5B). To quantify this variability, the diameter of each ISV was measured at four equidistant points along its length and standard deviation from the mean diameter was graphed as a readout for ISV diameter variability (Fig 5C). Indeed, ISVs in the absence of perivascular fibroblasts were significantly more variable than those of water-treated controls (Fig 5C). Examination of larger blood vessels in the trunk revealed that MTZ treated embryos also showed significantly reduced mean diameter for both the dorsal aorta (DA) and posterior cardinal vein (PCV), with substantial increase in PCV diameter variability (S4 Fig). These results suggest that sclerotome derived cells play a crucial role in maintaining the trunk vasculature during the early stage. Interestingly, similar ablation experiments using col1a2NTR-mCherry; kdrl:EGFP during later stages between 4 and 5 dpf did not significantly alter the ISV morphology (S5A–S5C Fig). Together, our results suggest that perivascular fibroblasts play an early role in the stabilization of nascent blood vessels but are dispensable for maintaining ISV morphology at later stages.
Perivascular fibroblasts deposit collagens around ISVs
Our work suggests that perivascular fibroblasts are crucial for the stabilization of nascent ISVs. Since perivascular fibroblasts show high-level expression of collagen genes, including col1a2 and col5a1, we predicted that perivascular fibroblasts are the main source of the vascular ECM around newly formed ISVs. To test this model, we generated a UAS:Col1a2-GFP construct to express GFP-fused Col1a2 in live embryos. A similar Col1a2-GFP reporter has been previously used to visualize collagen deposition during skin development and repair in zebrafish [52]. nkx3.1:Gal4; UAS:NTR-mCherry embryos injected with the UAS:Col1a2-GFP plasmid showed many mCherry+GFP+ perivascular fibroblasts surrounding ISVs at 56 hpf (Fig 5D). Indeed, numerous thin GFP+ ‘strings’, likely corresponding to collagen fibers, wrapped around the ISV with the intensity increasing from 56 to 72 hpf, suggesting continuous Col1a2-GFP protein deposition (Fig 5D). To determine whether perivascular fibroblasts directly contribute to collagens around ISVs, we injected nkx3.1NTR-mCherry embryos with a low dose of the UAS:Col1a2-GFP plasmid to achieve mosaic labeling of perivascular fibroblasts. Imaging of individually labeled mCherry+GFP+ perivascular fibroblasts showed local Col1a2-GFP deposition around the ISV within 1–2 cell diameters from the Col1a2-GFP-expressing perivascular fibroblast (Fig 5E). Our results suggest that perivascular fibroblasts contribute to collagens in the ECM surrounding nascent ISVs that likely contribute to vessel stabilization.
To visualize the dynamics of collagen deposition, we combined cell ablation experiments with a stable UAS:Col1a2-GFP transgenic line. nkx3.1:Gal4; UAS:NTR-mCherry; UAS:Col1a2-GFP (nkx3.1NTR-mCherry;Col1a2-GFP) embryos were incubated in water or MTZ from 38 to 62 hpf (Fig 5F). Embryos were imaged individually before and after the water/MTZ treatment to compare collagen deposition around the same ISVs (Fig 5G). Water-treated control embryos showed an average of 27% increase in Col1a2-GFP protein deposition around ISVs (Fig 5G and 5H). By contrast, complete ablation of perivascular fibroblasts with MTZ treatment resulted in a 21% decrease in Col1a2-GFP signal (Fig 5G and 5H). This result suggests that perivascular fibroblasts play an active role in contributing to collagens in the vascular ECM during early stages of ISV development. To determine whether perivascular fibroblasts continue to deposit collagens at later stages, we performed similar cell ablation experiments in col1a2NTR-mCherry;Col1a2-GFP embryos from 4 to 5 dpf (S5D Fig). Water-treated controls showed 8% increase in Col1a2-GFP level around ISVs, whereas collagen deposition in MTZ-treated embryos remained largely unchanged (S5E and S5F Fig). In summary, these results show that the vascular collagen network undergoes active remodeling during initial development, but becomes more stable at later stages, highlighting the early requirement of perivascular fibroblasts.
collagen mutants show severe ISV hemorrhage
Our work suggests that perivascular fibroblasts secrete collagens and stabilize nascent ISVs. We next asked whether Col1a2 and Col5a1, which are expressed in perivascular fibroblasts (Figs 1C and S1B), are required to maintain the integrity of nascent blood vessels. Using CRISPR/Cas9 editing [53], we generated mutants for col1a2 (col1a2ca108, 1-bp deletion) and col5a1 (col5a1ca109, 4-bp deletion) (S6A and S6B Fig). Both alleles resulted in frame shifts in the corresponding coding sequences leading to premature stops (S6C and S6D Fig). For simplicity, we designate wild-type, heterozygous or homozygous fish as +/+, +/-, and -/-, respectively. For example, heterozygous carriers of col1a2ca108 are referred to as col1a2+/-. Whole mount in situ hybridization showed that mutant mRNAs for both genes underwent nonsense-mediated decay (Fig 6A), suggesting that both col1a2ca108 and col5a1ca109 are null alleles. Crosses of heterozygous col1a2+/- parents gave rise to progeny segregated in roughly Mendelian ratios at all stages examined (S7A Fig). Although homozygous col1a2-/- mutants were viable as adults, they were substantially smaller than their wild-type siblings and often showed severe spine curvature (S7B Fig). This phenotype is reminiscent of those described in other type I collagen mutants [54, 55]. By contrast, from crosses of heterozygous col5a1+/- parents, homozygous col5a1-/- mutants were under-represented at 11 dpf, and completely absent by 14 dpf (S7C Fig). These results suggest that Col5a1 is a more critical component of the ECM compared to Col1a2, and it is required for the survival of the fish to adulthood. Strikingly, a small percentage of embryos from crosses of heterozygous col5a1+/- parents (2/88: 2%) showed spontaneous hemorrhage in the trunk region at 2 dpf and were genotyped as homozygous col5a1-/- (Fig 6B), suggesting compromised vascular integrity in the absence of Col5a1.
To further characterize the hemorrhage phenotype in collagen mutants, we developed a phenotype screening procedure (Fig 6C). In this assay, we analyzed both single mutants as well as double mutants to test genetic interactions between col1a2 and col5a1. In addition, we tested whether the increase in physical stress exacerbates the hemorrhage phenotype by raising embryos in a high-viscosity medium (0.6% methylcellulose, MC) [56]. Briefly, heterozygous adults harboring either one or both of the collagen mutations (col1a2+/-, col5a1+/-, or col1a2+/-; col5a1+/-) were intercrossed to generate wild type, heterozygous, and homozygous sibling embryos for different combinations (Fig 6C). The resulting embryos were then incubated in either water or the viscous methylcellulose solution from 48 to 80 hpf, during which embryos were imaged and screened for hemorrhage in the trunk. Embryos were then genotyped to correlate the genotype with the phenotype. Similar to their wild-type col1a2+/+ siblings (water: 0/14; MC: 0/27) and heterozygous col1a2+/- siblings (water: 0/39; MC: 0/38), homozygous col1a2-/- mutants did not show any trunk hemorrhage in either water (0/22) or the methylcellulose solution (0/21). This result suggests that loss of col1a2 alone does not compromise vascular stability even with increased physical stress. By contrast, while neither wild-type col5a1+/+ embryos (0/11) nor heterozygous col5a1+/- siblings (0/39) showed any hemorrhage under normal conditions, 12% of col5a1-/- embryos (3/25) displayed obvious hemorrhages in the trunk at 2 dpf, which was worsened substantially by physical stress (17/26, 65%). Analysis of different allele combinations of double mutants revealed several key characteristics of the hemorrhage phenotype (Fig 6D). First, the loss of both col5a1 wild-type alleles was required to cause the trunk hemorrhage phenotype. Second, even though loss of col1a2 alone in col1a2+/- or col1a2-/- embryos was not sufficient to cause hemorrhage, loss of one or two col1a2 wild-type alleles progressively increased the penetrance of the hemorrhage phenotype in col5a1-/- mutants, from 23% in col1a2+/-; col5a1-/- embryos (5/22) to 58% in col1a2-/-; col5a1-/- embryos (7/12) (Fig 6D). Lastly, increased physical stress by incubating embryos in the viscous methylcellulose solution substantially exacerbated both the penetrance and the severity of the hemorrhage phenotype (Figs 6D, 6E and S7D). It is interesting to note that stress-induced blood vessel rupture is typical of the vascular phenotypes associated with classical Ehlers-Danlos syndrome in humans. Together, our mutant analysis suggests that Col5a1 and Col1a2 function redundantly to maintain blood vessel integrity with Col5a1 being a more critical component.
The distribution of hemorrhagic foci is consistent with defects in ISVs in the trunk. To directly examine the ISV morphology in collagen mutants, we crossed col5a1+/-; kdrl:EGFP adults with col1a2+/-; col5a1+/- double heterozygous fish. Resulting embryos were incubated in the methylcellulose solution and their ISVs were analyzed at 53 hpf. In contrast to wild-type siblings, some col5a1-/- mutants showed severe ISV deformity and breakage, accompanied by hemorrhage (Fig 6F). It is likely that the hemorrhage phenotype is caused by physical forces exerted on fragile and deformed ISVs during muscle contractions. Quantifications of the diameter variability of intact ISVs as described earlier showed that both col1a2+/+; col5a1-/- and col1a2+/-; col5a1-/- embryos displayed significantly more variable ISVs than wild-type siblings (Fig 6G). Consistent with a stronger hemorrhage phenotype, ISVs in col1a2+/-; col5a1-/- embryos were significantly more variable than those in col1a2+/+; col5a1-/- single mutants. Together, our work suggests that the deposition of Col5a1 and Col1a2 between 1–2 dpf, likely by perivascular fibroblasts, is crucial for the stabilization of nascent ISVs.
Discussion
In this study, we characterize the origin and function of perivascular fibroblasts, a novel population of blood vessel associated cells in zebrafish. Our work provides three main conclusions. First, perivascular fibroblasts originate from the sclerotome and are distinct from pericytes. Second, a subset of perivascular fibroblasts functions as pericyte progenitors. Third, cell ablation and mutant analysis reveal that perivascular fibroblasts stabilize nascent blood vessels through deposition of collagens, including Col1a2 and Col5a1. Together, we propose a dual role of perivascular fibroblasts in vascular stabilization: they first establish the ECM around nascent blood vessels for initial stabilization and then function as pericyte progenitors to generate pericytes that further contribute to vessel integrity (Fig 7). Our work provides new insights into the molecular and cellular mechanisms underlying vascular stabilization and vascular diseases, such as Ehlers-Danlos syndrome.
Perivascular fibroblasts are a unique population of perivascular cells
Pericytes are well studied blood vessel support cells associated with intersegmental vessels in the zebrafish trunk. Four lines of evidence suggest that collagen-expressing perivascular fibroblasts represent a unique population of perivascular cells, distinct from mature and fully differentiated pericytes. First, these two cell populations show distinct marker expression, with fully differentiated pericytes labeled by high level of pdgfrb expression, whereas perivascular fibroblasts marked by the expression of collagen genes, such as col1a2 and col5a1. Second, perivascular fibroblasts become associated with ISVs as early as 31 hpf, whereas pericytes do not differentiate until at least 60 hpf [43]. Third, perivascular fibroblasts are much more abundant than pericytes, and are associated with both arterial and venous ISVs in similar numbers unlike pericytes. Finally, perivascular fibroblasts are positioned abluminal to pericytes and appear more globular with numerous shorter processes that extend around the endothelium. Based on the differences in marker expression, developmental timing, cell number, and cell morphology, we propose that perivascular fibroblasts represent a novel perivascular cell population, distinct from pericytes. Intriguingly, several single cell RNA sequencing studies have revealed a similar perivascular ‘fibroblast-like’ cell population in the adult murine brain [28–31]. Murine fibroblast-like cells show robust expression of ECM genes, including collagen genes col5a1 and col1a2, but much less mural cell markers such as pdgfrb [31, 57]. Furthermore, similar to zebrafish, murine perivascular fibroblast-like cells are loosely adhered to the blood vessels, abluminal to mural cells and the basement membrane [31]. Combined with our work, these studies suggest that perivascular fibroblasts represent a unique and evolutionarily conserved population of blood vessel associated cells across vertebrates.
The sclerotome is the embryonic origin of perivascular fibroblasts and pericytes
The sclerotome is the embryonic compartment that contributes to the axial skeleton of vertebrates. Our previous and current studies demonstrate that the sclerotome also gives rise to several populations of tissue support cells, including tenocytes (tendon fibroblasts) [44] and perivascular fibroblasts. The zebrafish sclerotome has a unique bipartite organization with a larger ventral domain and a smaller dorsal domain. Using time-lapse imaging and cell tracing, we have built a spatial map of origin for perivascular fibroblasts. Similar to tenocytes, different sclerotome domains contribute to perivascular fibroblasts in a stereotypic spatial pattern, with the ventral sclerotome domain being the major contributor (84%). The ventral sclerotome domain generates perivascular fibroblasts along the entire length of the ISV, whereas the dorsal sclerotome domain contributes only locally to perivascular fibroblasts associated with the dorsal half of the ISV.
Since perivascular fibroblasts also function as pericyte precursors (see below), our results suggest that most pericytes associated with ISVs also originate from the sclerotome. Consistent with our results, vSMCs around the dorsal aorta have been traced back to cells derived from the sclerotome in zebrafish [49]. In addition, mural cells in the zebrafish trunk vessels have been shown to be dependent on the mesoderm but not the neural crest lineage [58]. Thus, blood vessel support cells in the zebrafish trunk, including perivascular fibroblasts and mural cells, likely originate from the sclerotome. Combined with previous lineage analysis in mouse and chick [59, 60], these data suggest a model where mural cells in the brain are neural crest derived, whereas mural cells in the trunk are sclerotome derived.
Perivascular fibroblasts function as pericyte progenitors
Although perivascular fibroblast-like cells have been identified in the mouse brain by single cell RNA sequencing, their biological functions are still unclear. In our study, in vivo time-lapse imaging reveals that the majority of newly born pericytes on ISVs at 3 dpf can be traced back to perivascular fibroblasts at 2 dpf. In a complementary approach, we show that genetic ablation of perivascular fibroblasts results in a general reduction in the number of pericytes on both arterial and venous ISVs. Together, these results indicate that some perivascular fibroblasts serve as pericyte precursors, contributing to most, if not all, pericytes associated with ISVs in the trunk. Incomplete labeling or ablation likely reflects the mosaic nature of the col1a2 transgenic lines [50].
It is interesting to note that there are about 10 perivascular fibroblasts associated with each ISV, however, only about 8% of these cells differentiate into pericytes. It thus raises the question of how these perivascular fibroblasts are ‘selected’ as pericyte progenitors. There are two possibilities. In the first scenario, all perivascular fibroblasts have the equal potential to differentiate into pericytes, but extrinsic cues determine which cell to switch on the pericyte fate. Notch signaling has been implicated in the regulation of pericyte formation in both mouse and zebrafish [58, 61–64]. In particular, recent work in zebrafish has shown that the activation of Notch signaling in naïve mesenchymal cells (likely corresponding to perivascular fibroblasts in our study) is crucial to induce pericyte differentiation around ISVs [58]. Therefore, active Notch signaling might bias some perivascular fibroblasts towards the pericyte lineage. Alternatively, perivascular fibroblasts might represent a heterogenous population of cells, containing a small fraction of pericyte progenitors. Supporting this idea, recent scRNA sequencing studies have shown that perivascular fibroblast-like cells in the mouse brain can be further divided into several subtypes based on unique gene expression signatures [30, 31].
Perivascular fibroblasts function to stabilize nascent blood vessels
Although mural cells are known to stabilize mature blood vessels, it is not clear how newly formed blood vessels are supported prior to the differentiation of mural cells. Our work provides strong evidence that perivascular fibroblasts function to stabilize nascent blood vessels by depositing collagens in the vascular ECM. Time-lapse imaging shows that the emergence of collagen-expressing perivascular fibroblasts along ISVs occurs concurrently with ISV development, at least one day before the differentiation of the first pericytes. The live Col1a2-GFP reporter reveals that perivascular fibroblasts contribute to a network of collagen fibers around nascent ISVs as early as 38 hpf, which continues to expand until at least 3 dpf. Consistent with this observation, genetic ablation of perivascular fibroblasts from 38 to 62 hpf prior to pericyte differentiation largely prevents the expansion of the collagen network around ISVs, leading to aberrant ISV morphology. These findings suggest that perivascular fibroblasts contribute to the collagen network to stabilize newly formed blood vessels before pericyte differentiation. Consistent with this model, loss of col5a1 results in dysmorphic ISVs with spontaneous hemorrhage in the trunk, which is exacerbated by the additional loss of col1a2 or increasing physical stress. Crucially, we find that these ISV phenotypes can be observed as early as 2 dpf, well before the emergence of first pericytes on ISVs. Consistent with our work, recent study has described a similar population of col22a1+ perivascular fibroblast-like cells associated with the cranial vasculature in zebrafish [65]. Null mutations in col22a1, a minor fibril-associated collagen, result in compromised cranial vessel integrity in both embryos and adults following cardiovascular stress [65], highlighting the importance of collagen and perivascular fibroblasts in vascular stabilization.
Intriguingly, early ablation of perivascular fibroblasts rarely results in the severe hemorrhage phenotype observed in col5a1-/- mutants (under either normal or stressed conditions). Using the live Col1a2-GFP reporter, we find that low level of Col1a2-GFP deposition is already visible around ISVs at 38 hpf prior to MTZ treatment, some of which persists even after ablation at 62 hpf (Fig 5G). It is therefore possible that these residual collagen fibers provide sufficient support to newly formed ISVs and prevent severe hemorrhage seen in collagen mutants. It is also interesting to note that Col1a2-GFP level decreases substantially upon early but not late ablation of perivascular fibroblasts, indicating a higher turnover rate of collagen fibers during early stages of vascular development. This is supported by our observation that late ablation of perivascular fibroblasts from 4 to 5 dpf minimally impacts ISV morphology. Although our work points to perivascular fibroblasts as the main contributor of the vascular ECM, it remains possible that other cell type(s) in the vicinity, such as tenocytes, might contribute to collagen deposition around the ISVs. Development of specific genetic driver lines will allow us to directly test this possibility in the future.
Zebrafish model of Ehlers-Danlos syndrome
Collagen is the most abundant protein in the human body and is a major component of the ECM [66]. Fibrillar collagens, including collagen I and collagen V, constitute the most common sub-group within the collagen family which often co-localize in tissues such as the dermis, tendons and ligaments [66]. Mutations in fibrillar collagens or collagen modification enzymes have been implicated in a number of connective tissue diseases, including Ehlers-Danlos syndrome (EDS). The classical subtype of EDS, caused by mutations in collagen V, and less frequently collagen I, is characterized by skin hyperextensibility, joint hypermobility, and blood vessel fragility [19, 67]. Classical EDS patients experience easy bruising and excessive hemorrhaging under normal exertion. By contrast, vascular phenotypes are more prominent in the vascular subtype of EDS, resulting from mutations in collagen III, with most patients experiencing severe spontaneous rupture of large arteries [19]. Interestingly, the zebrafish genome does not contain a collagen III gene, which is likely lost during evolution as in many other teleost fish species [68]. Characterization of zebrafish col1a2 and col5a1 mutants reveals several key features of vascular EDS. Embryos lacking functional col5a1 genes develop spontaneous hemorrhages in the trunk under normal physiological conditions, reminiscent of easy bruising in patients with vascular EDS. One hallmark of vascular EDS is that physical stress often leads to more severe symptoms [69]. Similarly, both the penetrance and severity of the hemorrhage phenotype in col5a1-/- mutants are strongly enhanced by the increase of physical stress. Thus, our results suggest that the lack of collagen III in zebrafish is likely compensated by other fibrillar collagens (types I and V). Our col1a2 and col5a1 mutants represent the first zebrafish model of vascular EDS that we know of.
Analysis of collagen mutants also provides new insights on vascular EDS. First, Col5a1 is a more critical component of the vascular ECM compared to Col1a2. Lack of col5a1 function leads to ISV hemorrhages, whereas loss of col1a2 alone does not result in any obvious vascular defects even with increased physical stress. Intriguingly, collagen I is the most abundant component in collagen fibrils (> 90%), while collagen V constitutes only a minor fraction (< 5%) [70]. Why do col5a1-/- mutants show more severe phenotypes than col1a2-/- mutants? Previous work has shown that Col5a1 plays an important role in proper nucleation and diameter regulation of collagen fibrils [71]. The loss of Col5a1 might result in defects in the assembly of collagen I into fibrils, leading to a more severe phenotype. Second, our mutant analysis reveals genetic interactions between col1a2 and col5a1. Loss of wild-type col1a2 alleles in the col5a1-/- background substantially enhances both the penetrance and the severity of the hemorrhage phenotype in a dosage dependent manner. This result suggests that Col5a1 and Col1a2 function redundantly to maintain vascular integrity. Lastly, as discussed above, our work provides strong evidence that defects in perivascular fibroblasts are the underlying cellular basis for the vascular phenotypes of EDS patients. Perivascular fibroblasts thus represent the cellular targets for potential therapeutic interventions.
In summary, our work identifies perivascular fibroblasts as a novel perivascular cell population with dual function in vascular stabilization: they regulate the ECM of nascent blood vessels and also function as pericyte progenitors. Interestingly, many perivascular fibroblasts remain associated with mature blood vessels even after pericyte differentiation. This raises several questions about the function of these cells at later stages. Do perivascular fibroblasts continue to regulate the ECM around the mature vasculature? Do they function as resident stem cells to replace aging/damaged pericytes? Previous work has suggested that tissue injury in the mouse brain and spinal cord often triggers a fibrotic response by blood vessel associated cells [25, 32–34]. Attenuation of the fibrotic response has been shown to limit scar formation and improve axon regeneration post spinal cord injury in mice, suggesting that perivascular stromal cells might be an important therapeutic target [72]. It is plausible that zebrafish perivascular fibroblasts also play an active role in tissue injury repair, an exciting question for future investigations.
Materials and methods
Zebrafish strains
Zebrafish strains were maintained according to standard protocols. Animal research was conducted in accordance with current guidelines of the Canadian Council on Animal Care. All protocols were approved by the Animal Care Committee at the University of Calgary (#AC17-0128). The following transgenic strains were utilized in this study: TgBAC(col1a2:Gal4)ca102 [44, 50], TgBAC(col1a2:GFP)ca103 [44], Tg(kdrl:EGFP)la116 [73], Tg(kdrl:mCherry)ci5 [74], TgBAC(nkx3.1:Gal4)ca101 [44], TgBAC(pdgfrb:Gal4FF)ca42 [48], TgBAC(pdgfrb:GFP)ca41 [47], Tg(UAS:Col1a2-GFP)ca111, Tg(UAS:Kaede)s1999t [75], and Tg(UAS:NTR-mCherry)c264 [75]. The mosaic col1a2:Gal4; UAS:Kaede line was maintained by growing embryos with more mosaic Kaede expression. The col1a2ca108 and col5a1ca109 mutant lines were maintained as heterozygotes, and homozygous embryos were generated by intercrossing heterozygous carriers.
Generation of CRISPR mutants
The col1a2ca108 and col5a1ca109 mutant lines were generated using the CRISPR/Cas9 system as previously described [53]. Briefly, target sites were identified using the web program CHOPCHOP (http://chopchop.cbu.uib.no) [76]. sgRNA target sequences are 5’-GGGGGTTCCATTTGATCCAG-3’ (col1a2) and 5’-GGCTCCAGCAGATCATCCAG-3’ (col5a1). To assemble DNA templates for sgRNA transcription, gene-specific oligonucleotides containing the T7 promoter sequence (5’-TAATACGACTCACTATA-3’), the 20 base target site, and a complementary sequence were annealed to a constant oligonucleotide encoding the reverse-complement of the tracrRNA tail. sgRNAs were generated by in vitro transcription using the Megascript kit (Ambion). Cas9 mRNA was transcribed from linearized pCS2-Cas9 plasmid using the mMachine SP6 kit (Ambion). To generate mutants, one-cell stage wild-type embryos were injected with a mix containing the appropriate sgRNA at 20 ng/μl and Cas9 mRNA at 200 ng/μl. Injected fish were raised to adulthood and crossed to generate F1 embryos. T7 Endonuclease I assay (NEB) was then used to identify the presence of indel mutations in the targeted region of F1 fish. A col1a2 allele containing 1 bp deletion (col1a2ca108) and a col5a1 allele with 4 bp deletion (col5a1ca109) were identified (S6 Fig). For genotyping, small regions around the mutation sites were specifically amplified by PCR, which was followed by allele-specific restriction enzyme analysis. For col1a2ca108, the primers used were 5’-TTTTAAAGACTCACATTTGCCTT-3’ (forward) and 5’-CTCCGGGCTAGCTTTATATTTCGATT-3’ (reverse). For col5a1ca109, the primers used were 5’-ACTCTTGTTTGCTGTGCAGGT-3’ (forward) and 5’-CTCACCGGTATTGGCCGTGTT-3’ (reverse). Following PCR, the products were digested with a restriction enzyme which cuts only the wild type or the mutant allele. The enzymes used were BclI and BstCI (NEB) for col1a2ca108 and col5a1ca109, respectively. The resulting band sizes were analyzed using gel electrophoresis. For col1a2ca108, the BclI digest resulted in a single band of 473 bp in wild-type embryos, and two bands of 400 bp and 73 bp in homozygous col1a2-/- embryos. For col5a1ca109, the BstCI digest resulted in multiple bands at 89, 44, 30, 18 and 9 bp in wild-type embryos, and bands of 107, 44, 30 and 9 bp in homozygous col5a1-/- embryos.
Plasmid injection
To visualize collagen distribution, we generated a UAS:Col1a2-GFP construct based on a previously published pME-Col1a2-GFP plasmid [52]. To generate the stable transgenic line, UAS:Col1a2-GFP (40 ng/μl) was co-injected with tol2 transposase mRNA (40 ng/μl) into col1a2:Gal4 embryos at the one-cell stage with 1 nl per embryo. Injected embryos were raised to adulthood and founders were identified based on GFP expression in F1 embryos. The UAS:Col1a2-GFP line was subsequently propagated in the absence of the Gal4 driver to prevent silencing of the transgene. For more mosaic labeling, UAS:Col1a2-GFP was injected at 10 ng/μl with tol2 transposase mRNA. Injected embryos were screened for GFP expression at appropriate stages for imaging.
Morpholino injection
To inhibit ISV formation, morpholino oligonucleotides (Gene Tools, LLC) targeting the etsrp gene (etsrpMO: 5’-CACTGAGTCCTTATTTCACTATATC-3’) [77] were injected at 0.8 mM into one-cell stage embryos with 1.5 nl per embryo. Injected embryos were imaged at appropriate stages for the distribution of sclerotome derived cells.
In situ hybridization and immunohistochemistry
Whole-mount in situ hybridization and antibody staining were performed according to standard protocols. col1a2 and col5a1 antisense probes were used in this study. Double fluorescent in situ hybridization was performed using digoxigenin (DIG) and dinitrophenyl (DNP) labeled probes. For antibody labeling, the rabbit polyclonal antibody to GFP (1:500, MBL) was used. For fluorescent detection of antibody labeling, appropriate Alexa Fluor-conjugated secondary antibodies were used (1:500, Thermo Fisher).
Time-lapse imaging
Imaging was performed using the Olympus FV1200 confocal microscope as previously described [44]. Embryos older than 24 hpf were incubated in fish water with 1-phenyl 2-thiourea (PTU) to prevent pigmentation. Fish were anesthetized in 0.4% tricaine and mounted in 0.6% low melting point agarose. Z-stack images of the region of interest were then collected using the 20x objective at appropriate time intervals (5–15 mins) for up to 24 hours. Movies and images were processed using the Olympus Fluoview software and the Fiji software [78]. Cell tracing was performed manually using Fiji.
Cell ablation
To ablate perivascular fibroblasts, nkx3.1NTR-mCherry or col1a2NTR-mCherry embryos at desired developmental stages were treated with water (control group) or 5 mM metronidazole (MTZ, experimental group) for 24–48 hours. In some experiments, mCherry- siblings treated with 5 mM MTZ in the same time windows were used as additional controls. Embryos were subsequently washed in fish water and imaged to confirm successful ablation of mCherry+ cells.
Mechanical stress assay
To test the effect of physical stress on the vascular phenotype of collagen mutants, we adopted a mechanical overloading assay by raising embryos in a high-viscosity medium (0.6% methylcellulose) [56]. Heterozygotes for single mutants (col1a2+/- or col5a1+/-) or double mutants (col1a2+/-; col5a1+/-) were intercrossed to obtain sibling embryos with different allele combinations. Resulting embryos were incubated in water (control group) or 0.6% methylcellulose (MC, experimental group) at 48 hpf. Fish were screened for hemorrhage in the trunk region approximately every 3 hours from 48 to 80 hpf and subsequently genotyped for col1a2 and col5a1.
Quantification of single cell analysis
All measurements of single cell morphology were taken using the ‘line selection’ tool in Fiji. Cell length measurements were taken along the long axis of the cell, parallel to the neighboring ISV as shown in Fig 3E. Cell height measurements were taken at the midpoint of the cell body, orthogonal to the length measurement. Aspect ratio was calculated as cell length divided by cell height. Process lengths were measured by tracing along single processes using the ‘segmented line’ option under the line selection tool.
Quantification of blood vessel diameter variability
To quantify blood vessel diameters, we used the endothelial specific kdrl:EGFP reporter to fluorescently label intersegmental vessels. Embryos were imaged laterally in the trunk region (somite 12 to 18) and 8–10 ISVs were quantified per embryo. Using the ‘line selection’ tool in Fiji, four diameter measurements were taken for each ISV along its length and measurements were averaged to obtain mean vessel diameter. Standard deviation from the mean vessel diameter was plotted as a readout of ISV diameter variability. For dorsal aorta (DA) and posterior cardinal vein (PCV) measurements, diameter measurements were taken at 6–10 equidistant points along their length in the mid-trunk region with one measurement per somite. Mean vessel diameter and standard deviation from the mean were similarly graphed as ISV measurements.
Quantification of fluorescence intensity
To compare pdgfrb:GFP expression between pericytes and perivascular fibroblasts, GFP intensity measurements were taken using the ‘oval selection’ tool in Fiji. Circular regions of interest (ROIs) of 5.56 μm2 area were drawn within individual cells and ‘mean gray value’ was measured. Five pericyte and perivascular fibroblast pairs were measured per embryo. For Col1a2-GFP measurements, a rectangular area of 500 by 160 pixels (393 μm x 126 μm) was drawn using the ‘rectangular selection’ tool in Fiji. The mean fluorescence intensity of the same region was measured before and after the water/MTZ treatment (GFPbefore and GFPafter, respectively). The percentage change in Col1a2-GFP deposition was calculated using the following formula: (GFPafter—GFPbefore) / GFPbefore x 100%.
Statistical analysis
All graphs and statistical analysis were generated using the GraphPad PRISM software. Data were plotted with mean ± SEM indicated. Significance was calculated by performing using the non-parametric Mann-Whitney U test with two-tailed p values: p > 0.05 (ns, not significant), p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) and p < 0.0001 (****).
Supporting information
S1 Fig [a]
Characterization of collagen expression in perivascular fibroblasts.
S2 Fig [a]
Concurrent development of perivascular fibroblasts and the vasculature.
S3 Fig [a]
Characterization of transgenic reporters.
S4 Fig [mtz]
Effect of early perivascular fibroblast ablation on major trunk vessels.
S5 Fig [a]
Late ablation of perivascular fibroblasts does not alter ISV morphology and minimally impacts collagen deposition.
S6 Fig [a]
Generation of collagen mutants.
S7 Fig [a]
Characterization of collagen mutants.
S1 Video [mp4]
Perivascular fibroblasts originate from the sclerotome.
S2 Video [mp4]
Perivascular fibroblasts function as pericyte progenitors.
S3 Video [mp4]
Perivascular fibroblasts function as pericyte progenitors.
S1 Table [xlsx]
This table contains all the numerical data presented in this manuscript.
Zdroje
1. Xu J, Shi G-P. Vascular wall extracellular matrix proteins and vascular diseases. Biochim Biophys Acta. 2014;(11):2106–19. doi: 10.1016/j.bbadis.2014.07.008 25045854
2. Murakami M, Simons M. Regulation of vascular integrity. Journal of Molecular Medicine. NIH Public Access; 2009; 87(6):571–82. doi: 10.1007/s00109-009-0463-2 19337719
3. Michel JB, Martin-Ventura JL, Egido J, Sakalihasan N, Treska V, Lindholt J, et al. Novel aspects of the pathogenesis of aneurysms of the abdominal aorta in humans. Cardiovasc Res. 2011;90(1):18–27. doi: 10.1093/cvr/cvq337 21037321
4. Sakalihasan N, Michel JB, Katsargyris A, Kuivaniemi H, Defraigne JO, Nchimi A, et al. Abdominal aortic aneurysms. Nature Reviews Disease Primers. 2018;4:34. doi: 10.1038/s41572-018-0030-7 30337540
5. Qureshi AI, Mendelow AD, Hanley DF. Intracerebral haemorrhage. The Lancet. 2009;373:1632–44. doi: 10.1016/S0140-6736(09)60371-8 19427958
6. Brisman JL, Song JK, Newell DW. Cerebral aneurysms. The New England Journal of Medicine. 2006;355(9):928–39. doi: 10.1056/NEJMra052760 16943405
7. Dejana E, Tournier-Lasserve E, Weinstein BM. The Control of Vascular Integrity by Endothelial Cell Junctions: Molecular Basis and Pathological Implications. Dev Cell. 2009 Feb 17;16(2):209–21. doi: 10.1016/j.devcel.2009.01.004 19217423
8. Gaengel K, Genové G, Armulik A, Betsholtz C. Endothelial-mural cell signaling in vascular development and angiogenesis. Arterioscler Thromb Vasc Biol. 2009 May 1;29(5):630–8. doi: 10.1161/ATVBAHA.107.161521 19164813
9. Levéen P, Pekny M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C. Mice deficient for PDGF B show renal, cardiovascular, and hematological abnormalities. Genes Dev. 1994;8(16):1875–87. doi: 10.1101/gad.8.16.1875 7958863
10. Lindahl P, Johansson BR, Levéen P, Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science. 1997 Jul 11;277(5323):242–5. doi: 10.1126/science.277.5323.242 9211853
11. Soriano P. Abnormal kidney development and hematological disorders in PDGF β- receptor mutant mice. Genes Dev. 1994;8(16):1888–96. doi: 10.1101/gad.8.16.1888 7958864
12. Hellström M, Gerhardt H, Kalén M, Li X, Eriksson U, Wolburg H, et al. Lack of pericytes leads to endothelial hyperplasia and abnormal vascular morphogenesis. J Cell Biol. 2001 Feb 5;152(3):543–53. doi: 10.1083/jcb.153.3.543 11331305
13. Bjarnegård M, Enge M, Norlin J, Gustafsdottir S, Fredriksson S, Abramsson A, et al. Endothelium-specific ablation of PDGFB leads to pericyte loss and glomerular, cardiac and placental abnormalities. Development. 2004 Apr;131(8):1847–57. doi: 10.1242/dev.01080 15084468
14. Rhodes JM, Simons M. The extracellular matrix and blood vessel formation: Not just a scaffold: Angiogenesis Review Series. J Cell Mol Med. 2007 Mar;11(2):176–205. doi: 10.1111/j.1582-4934.2007.00031.x 17488472
15. Löhler J, Timpl R, Jaenisch R. Embryonic lethal mutation in mouse collagen I gene causes rupture of blood vessels and is associated with erythropoietic and mesenchymal cell death. Cell. 1984 Sep 1;38(2):597–607. doi: 10.1016/0092-8674(84)90514-2 6467375
16. Liu X, Wu H, Byrne M, Krane S, Jaenisch R. Type III collagen is crucial for collagen I fibrillogenesis and for normal cardiovascular development. Proc Natl Acad Sci. 1997 Mar 4;94(5):1852–6. doi: 10.1073/pnas.94.5.1852 9050868
17. Pöschl E, Schlötzer-Schrehardt U, Brachvogel B, Saito K, Ninomiya Y, Mayer U. Collagen IV is essential for basement membrane stability but dispensable for initiation of its assembly during early development. Development. 2004 Apr;131(7):1619–28. doi: 10.1242/dev.01037 14998921
18. Thyboll J, Kortesmaa J, Cao R, Soininen R, Wang L, Iivanainen A, et al. Deletion of the Laminin 4 Chain Leads to Impaired Microvessel Maturation. Mol Cell Biol. 2002 Feb 15;22(4):1194–202. doi: 10.1128/mcb.22.4.1194-1202.2002 11809810
19. Malfait F. Vascular aspects of the Ehlers-Danlos Syndromes. Matrix Biol. 2018;71–72:380–395. doi: 10.1016/j.matbio.2018.04.013 29709596
20. Yamazaki T, Mukouyama Y. Tissue Specific Origin, Development, and Pathological Perspectives of Pericytes. Front Cardiovasc Med. 2018;5:78. doi: 10.3389/fcvm.2018.00078 29998128
21. Holm A, Heumann T, Augustin HG. Microvascular Mural Cell Organotypic Heterogeneity and Functional Plasticity. Trends in Cell Biology. 2018;28(4): 302–316. doi: 10.1016/j.tcb.2017.12.002 29307447
22. Armulik A, Genové G, Betsholtz C. Pericytes: Developmental, Physiological, and Pathological Perspectives, Problems, and Promises. Dev Cell. 2011;21(2):193–215. doi: 10.1016/j.devcel.2011.07.001 21839917
23. Majesky MW. Developmental basis of vascular smooth muscle diversity. Arterioscler Thromb Vasc Biol. 2007 Jun;27(6):1248–58. doi: 10.1161/ATVBAHA.107.141069 17379839
24. Crisan M, Corselli M, Chen WCW, Péault B. Perivascular cells for regenerative medicine. J Cell Mol Med. 2012 Dec;16(12):2851–60. doi: 10.1111/j.1582-4934.2012.01617.x 22882758
25. Soderblom C, Luo X, Blumenthal E, Bray E, Lyapichev K, Ramos J, et al. Perivascular fibroblasts form the fibrotic scar after contusive spinal cord injury. J Neurosci. 2013;33(34):13882–7. doi: 10.1523/JNEUROSCI.2524-13.2013 23966707
26. Guillemin GJ, Brew BJ. Microglia, macrophages, perivascular macrophages, and pericytes: a review of function and identification. J Leukoc Biol. 2004 Mar 1;75(3):388–97. doi: 10.1189/jlb.0303114 14612429
27. Macvicar BA, Newman EA. Astrocyte regulation of blood flow in the brain. Cold Spring Harb Perspect Biol. 2015;7(5):1–15.
28. Marques S, Zeisel A, Codeluppi S, Van Bruggen D, Falcão AM, Xiao L, et al. Oligodendrocyte heterogeneity in the mouse juvenile and adult central nervous system. Science (80-). 2016 Jun 10;352(6291):1326–9. doi: 10.1126/science.aaf6463 27284195
29. Zeisel A, Hochgerner H, Lönnerberg P, Johnsson A, Memic F, van der Zwan J, et al. Molecular Architecture of the Mouse Nervous System. Cell. 2018 Aug 9;174(4):999–1014.e22. doi: 10.1016/j.cell.2018.06.021 30096314
30. Saunders A, Macosko EZ, Wysoker A, Goldman M, Krienen FM, de Rivera H, et al. Molecular Diversity and Specializations among the Cells of the Adult Mouse Brain. Cell. 2018 Aug 9;174(4):1015–1030.e16. doi: 10.1016/j.cell.2018.07.028 30096299
31. Vanlandewijck M, He L, Mäe MA, Andrae J, Ando K, Del Gaudio F, et al. A molecular atlas of cell types and zonation in the brain vasculature. Nature. 2018 Feb 14;554(7693):475–80. doi: 10.1038/nature25739 29443965
32. Kelly KK, MacPherson AM, Grewal H, Strnad F, Jones JW, Yu J, et al. Col1a1+ perivascular cells in the brain are a source of retinoic acid following stroke. BMC Neurosci. 2016;17(1):49. doi: 10.1186/s12868-016-0284-5 27422020
33. Fernández-Klett F, Potas JR, Hilpert D, Blazej K, Radke J, Huck J, et al. Early loss of pericytes and perivascular stromal cell-induced scar formation after stroke. J Cereb Blood Flow Metab. 2013 Mar;33(3):428–39. doi: 10.1038/jcbfm.2012.187 23250106
34. Göritz C, Dias DO, Tomilin N, Barbacid M, Shupliakov O, Frisén J. A pericyte origin of spinal cord scar tissue. Science. 2011 Jul 8;333(6039):238–42. doi: 10.1126/science.1203165 21737741
35. Makihara N, Arimura K, Ago T, Tachibana M, Nishimura A, Nakamura K, et al. Involvement of platelet-derived growth factor receptor β in fibrosis through extracellular matrix protein production after ischemic stroke. Exp Neurol. 2015 Feb 1;264:127–34. doi: 10.1016/j.expneurol.2014.12.007 25510317
36. Gore A V., Monzo K, Cha YR, Pan W, Weinstein BM. Vascular development in the zebrafish. Cold Spring Harb Perspect Med. 2012 May 1;2(5):a006684. doi: 10.1101/cshperspect.a006684 22553495
37. Howe K, Clark MD, Torroja CF, Torrance J, Berthelot C, Muffato M, et al. The zebrafish reference genome sequence and its relationship to the human genome. Nature. 2013 Apr 25;496(7446):498–503. doi: 10.1038/nature12111 23594743
38. Chico TJA, Ingham PW, Crossman DC. Modeling Cardiovascular Disease in the Zebrafish. Trends Cardiovasc Med. 2008 May;18(4):150–5. doi: 10.1016/j.tcm.2008.04.002 18555188
39. Isogai S, Horiguchi M, Weinstein BM. The vascular anatomy of the developing zebrafish: An atlas of embryonic and early larval development. Dev Biol. 2001 Feb 15;230(2):278–301. doi: 10.1006/dbio.2000.9995 11161578
40. Santoro MM, Pesce G, Stainier DY. Characterization of vascular mural cells during zebrafish development. Mech Dev. 2009 Aug;126(8–9):638–49. doi: 10.1016/j.mod.2009.06.1080 19539756
41. Bahrami N, Childs SJ. Pericyte biology in zebrafish. In: Advances in Experimental Medicine and Biology. Springer New York LLC; 2018. p. 33–51.
42. Isogai S, Lawson ND, Torrealday S, Horiguchi M, Weinstein BM. Angiogenic network formation in the developing vertebrate trunk. Development. 2003 Nov 1;130(21):5281–90. doi: 10.1242/dev.00733 12954720
43. Ando K, Fukuhara S, Izumi N, Nakajima H, Fukui H, Kelsh RN, et al. Clarification of mural cell coverage of vascular endothelial cells by live imaging of zebrafish. Development. 2016 Apr 15;143(8):1328–39. doi: 10.1242/dev.132654 26952986
44. Ma RC, Jacobs CT, Sharma P, Kocha KM, Huang P. Stereotypic generation of axial tenocytes from bipartite sclerotome domains in zebrafish. Moens C, editor. PLOS Genet. 2018 Nov 2;14(11):e1007775. doi: 10.1371/journal.pgen.1007775 30388110
45. Olson LE, Soriano P. PDGFRβ signaling regulates mural cell plasticity and inhibits fat development. Dev Cell. 2011 Jun 14;20(6):815–26. doi: 10.1016/j.devcel.2011.04.019 21664579
46. Hellström M, Kalen M, Lindahl P, Abramsson A, Betsholtz C. Role of PDGF-B and PDGFR-β in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development. 1999;(126):3047–55. 10375497
47. Bahrami N, Childs SJ. Development of vascular regulation in the zebrafish embryo. Dev. 2020 May 1;147: dev183061. doi: 10.1242/dev.183061 32423977
48. Whitesell TR, Chrystal PW, Ryu JR, Munsie N, Grosse A, French CR, et al. foxc1 is required for embryonic head vascular smooth muscle differentiation in zebrafish. Dev Biol. 2019 Sep 1;453(1):34–47. doi: 10.1016/j.ydbio.2019.06.005 31199900
49. Stratman AN, Pezoa SA, Farrelly OM, Castranova D, Dye LE, Butler MG, et al. Interactions between mural cells and endothelial cells stabilize the developing zebrafish dorsal aorta. Dev. 2017 Jan 1;144(1):115–27. doi: 10.1242/dev.143131 27913637
50. Sharma P, Ruel TD, Kocha KM, Liao S, Huang P. Single cell dynamics of embryonic muscle progenitor cells in zebrafish. Dev. 2019 Jul 1;146(14):dev178400. doi: 10.1242/dev.178400 31253635
51. Curado S, Stainier DYR, Anderson RM. Nitroreductase-mediated cell/tissue ablation in zebrafish: A spatially and temporally controlled ablation method with applications in developmental and regeneration studies. Nat Protoc. 2008 May;3(6):948–54. doi: 10.1038/nprot.2008.58 18536643
52. Morris JL, Cross SJ, Lu Y, Kadler KE, Lu Y, Dallas SL, et al. Live imaging of collagen deposition during skin development and repair in a collagen I—GFP fusion transgenic zebrafish line. Dev Biol. 2018;441(1):4–11. doi: 10.1016/j.ydbio.2018.06.001 29883658
53. Gagnon JA, Valen E, Thyme SB, Huang P, Ahkmetova L, Pauli A, et al. Efficient Mutagenesis by Cas9 Protein-Mediated Oligonucleotide Insertion and Large-Scale Assessment of Single-Guide RNAs. PLoS One. 2014 May 29;9(5): e98186. doi: 10.1371/journal.pone.0098186 24873830
54. Fisher S, Jagadeeswaran P, Halpern ME. Radiographic analysis of zebrafish skeletal defects. Dev Biol. 2003 Dec 1;264(1):64–76. doi: 10.1016/s0012-1606(03)00399-3 14623232
55. Gistelinck C, Kwon RY, Malfait F, Symoens S, Harris MP, Henke K, et al. Zebrafish type I collagen mutants faithfully recapitulate human type I collagenopathies. Proc Natl Acad Sci. 2018 Aug 21;115(34):E8037–46. doi: 10.1073/pnas.1722200115 30082390
56. Hall TE, Bryson-Richardson RJ, Berger S, Jacoby AS, Cole NJ, Hollway GE, et al. The zebrafish candyfloss mutant implicates extracellular matrix adhesion failure in laminin α2-deficient congenital muscular dystrophy. Proc Natl Acad Sci. 2007 Apr 24;104(17):7092–7. doi: 10.1073/pnas.0700942104 17438294
57. He L, Vanlandewijck M, Mäe MA, Andrae J, Ando K, Gaudio F Del, et al. Single-cell RNA sequencing of mouse brain and lung vascular and vessel-associated cell types. Sci Data. 2018 Aug 21;5:180160. doi: 10.1038/sdata.2018.160 30129931
58. Ando K, Wang W, Peng D, Chiba A, Lagendijk AK, Barske L, et al. Peri-arterial specification of vascular mural cells from naïve mesenchyme requires notch signaling. Dev. 2019;146: dev165589. doi: 10.1242/dev.165589 30642834
59. Etchevers HC, Vincent C, Le Douarin NM, Couly GF. The cephalic neural crest provides pericytes and smooth muscle cells to all blood vessels of the face and forebrain. Development. 2001;128(7):1059–68. 11245571
60. Bergwerff M, Verberne ME, DeRuiter MC, Poelmann RE, Gittenberger-de Groot AC. Neural crest cell contribution to the developing circulatory system implications for vascular morphology? Circ Res. 1998 Feb 9;82(2):221–31. doi: 10.1161/01.res.82.2.221 9468193
61. Wang Y, Pan L, Moens CB., Appel B. Notch3 establishes brain vascular integrity by regulating pericyte number. Dev. 2014;141: 307–17. doi: 10.1242/dev.096107 24306108
62. Domenga V, Fardoux P, Lacombe P, Monet M, Maciazek J, Krebs LT, et al. Notch3 is required for arterial identity and maturation of vascular smooth muscle cells. Genes Dev. 2004 Nov 15;18(22):2730–5. doi: 10.1101/gad.308904 15545631
63. Liu H, Zhang W, Kennard S, Caldwell RB, Lilly B. Notch3 is critical for proper angiogenesis and mural cell investment. Circ Res. 2010 Oct 1;107(7):860–70. doi: 10.1161/CIRCRESAHA.110.218271 20689064
64. Kofler NM, Cuervo H, Uh MK, Murtomäki A, Kitajewski J. Combined deficiency of Notch1 and Notch3 causes pericyte dysfunction, models CADASIL, and results in arteriovenous malformations. Sci Rep. 2015;5(June):1–13.
65. Ton Q V, Leino D, Mowery SA, Bredemeier NO, Lafontant PJ, Lubert A, et al. Collagen COL22A1 maintains vascular stability and mutations in COL22A1 are potentially associated with intracranial aneurysms. Disease Models and Mechanisms. 2018;11:dmm033654. doi: 10.1242/dmm.033654 30541770
66. Ricard-Blum S. The Collagen Family. Cold Spring Harb Perspect Biol. 2011;3(1):1–19. doi: 10.1101/cshperspect.a004978 21421911
67. Mak KM, Png CYM, Lee DJ. Type V Collagen in Health, Disease, and Fibrosis. Anat Rec. 2016;299(5):613–29.
68. Bretaud S, Nauroy P, Malbouyres M, Ruggiero F. Fishing for collagen function: About development, regeneration and disease. Semin Cell Dev Biol. 2019 May 1;89:100–8. doi: 10.1016/j.semcdb.2018.10.002 30312775
69. De Paepe A, Malfait F. Bleeding and bruising in patients with Ehlers-Danlos syndrome and other collagen vascular disorders. Br J Haematol; 2004;127:491–500. doi: 10.1111/j.1365-2141.2004.05220.x 15566352
70. Aszódi A, Legate KR, Nakchbandi I, Fässler R. What Mouse Mutants Teach Us About Extracellular Matrix Function. Annu Rev Cell Dev Biol. 2006 Nov;22(1):591–621. doi: 10.1146/annurev.cellbio.22.010305.104258 16824013
71. Wenstrup RJ, Florer JB, Brunskill EW, Bell SM, Chervoneva I, Birk DE. Type V collagen controls the initiation of collagen fibril assembly. J Biol Chem. 2004 Dec 17;279(51): 53331–7. doi: 10.1074/jbc.M409622200 15383546
72. Dias DO, Kim H, Holl D, Werne Solnestam B, Lundeberg J, Carlén M, et al. Reducing Pericyte-Derived Scarring Promotes Recovery after Spinal Cord Injury. Cell. 2018;173(1):153–165.e22. doi: 10.1016/j.cell.2018.02.004 29502968
73. Choi J, Dong L, Ahn J, Dao D, Hammerschmidt M, Chen JN. FoxH1 negatively modulates flk1 gene expression and vascular formation in zebrafish. Dev Biol. 2007 Apr 15;304(2):735–44. doi: 10.1016/j.ydbio.2007.01.023 17306248
74. Proulx K, Lu A, Sumanas S. Cranial vasculature in zebrafish forms by angioblast cluster-derived angiogenesis. Dev Biol. 2010 Dec 1;348(1):34–46. doi: 10.1016/j.ydbio.2010.08.036 20832394
75. Davison JM, Akitake CM, Goll MG, Rhee JM, Gosse N, Baier H, et al. Transactivation from Gal4-VP16 transgenic insertions for tissue-specific cell labeling and ablation in zebrafish. Dev Biol. 2007 Apr 15;304(2):811–24. doi: 10.1016/j.ydbio.2007.01.033 17335798
76. Montague TG, Cruz JM, Gagnon JA, Church GM, Valen E. CHOPCHOP: A CRISPR/Cas9 and TALEN web tool for genome editing. Nucleic Acids Res. 2014 Jul 1;42:W401–7. doi: 10.1093/nar/gku410 24861617
77. Sumanas S, Lin S. Ets1-related protein is a key regulator of vasculogenesis in zebrafish. PLoS Biol. 2006;4(1): e10. doi: 10.1371/journal.pbio.0040010 16336046
78. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: An open-source platform for biological-image analysis. Nature Methods. 2012;9: 676–82. doi: 10.1038/nmeth.2019 22743772
Článek vyšel v časopise
PLOS Genetics
2020 Číslo 10
- Duševní onemocnění se u teenagerů přenáší v rámci školní třídy
- Prof. Libor Grubhoffer: Nová varianta mpox je příležitostí pokročit v poznání původce varioly
- Co obsahují nové pokyny WHO pro dozor nad antibiotickou rezistencí?
- Naděje pro diabetiky: transplantace beta-buněk jim jednou nahradí inzulin
- Ukažte mi, jak kašlete, a já vám řeknu, co vám je
Nejčtenější v tomto čísle
- Evaluation of both exonic and intronic variants for effects on RNA splicing allows for accurate assessment of the effectiveness of precision therapies
- RNA-directed DNA Methylation
- The DNA methylome of human sperm is distinct from blood with little evidence for tissue-consistent obesity associations
- Correction: Molecular predictors of brain metastasis-related microRNAs in lung adenocarcinoma