A non-canonical role for desmoglein-2 in endothelial cells: implications for neoangiogenesis
- 2.6k Downloads
Desmogleins (DSG) are a family of cadherin adhesion proteins that were first identified in desmosomes and provide cardiomyocytes and epithelial cells with the junctional stability to tolerate mechanical stress. However, one member of this family, DSG2, is emerging as a protein with additional biological functions on a broader range of cells. Here we reveal that DSG2 is expressed by non-desmosome-forming human endothelial progenitor cells as well as their mature counterparts [endothelial cells (ECs)] in human tissue from healthy individuals and cancer patients. Analysis of normal blood and bone marrow showed that DSG2 is also expressed by CD34+CD45dim hematopoietic progenitor cells. An inability to detect other desmosomal components, i.e., DSG1, DSG3 and desmocollin (DSC)2/3, on these cells supports a solitary role for DSG2 outside of desmosomes. Functionally, we show that CD34+CD45dimDSG2+ progenitor cells are multi-potent and pro-angiogenic in vitro. Using a ‘knockout-first’ approach, we generated a Dsg2 loss-of-function strain of mice (Dsg2 lo/lo) and observed that, in response to reduced levels of Dsg2: (i) CD31+ ECs in the pancreas are hypertrophic and exhibit altered morphology, (ii) bone marrow-derived endothelial colony formation is impaired, (iii) ex vivo vascular sprouting from aortic rings is reduced, and (iv) vessel formation in vitro and in vivo is attenuated. Finally, knockdown of DSG2 in a human bone marrow EC line reveals a reduction in an in vitro angiogenesis assay as well as relocalisation of actin and VE-cadherin away from the cell junctions, reduced cell–cell adhesion and increased invasive properties by these cells. In summary, we have identified DSG2 expression in distinct progenitor cell subpopulations and show that, independent from its classical function as a component of desmosomes, this cadherin also plays a critical role in the vasculature.
KeywordsDESMOGLEIN-2 Endothelial progenitor cells Endothelium Hematopoietic stem and progenitor cells Mouse models of neoangiogenesis
Desmoglein-2 (DSG2), a type 1 transmembrane protein belonging to the cadherin family, is an adhesion molecule with well-described functions in the formation of cell–cell adhesion complexes called desmosomes. The desmosome is a specialised junction that provides intercellular links giving tensile strength to tissues that experience mechanical stress, such as in the myocardium, skin and gastrointestinal mucosa [1, 2, 3]. These complex, transversely symmetrical structures are composed of the transmembrane spanning cadherins, which interact via homotypic and/or heterotypic combinations to form the adhesive interface between cells, together with the intracellular proteins plakoglobin, desmoplakin and plakophilin which link the junction to the intermediate filament cytoskeleton. In humans, four desmoglein isoforms (DSG1-4) and three desmocollin isoforms (DSC1-3) have been identified, which are temporo-spatially expressed . DSG2 is expressed in all desmosome-bearing tissues, including cardiac muscle, whereas distribution of other isoforms is confined primarily to stratified epithelial tissue [1, 2, 3]. In keeping with its role in tissue integrity, loss of DSG2 function in humans is directly linked to arrhythmogenic right ventricular cardiomyopathy (ARVC), an autosomal recessive disease underpinned by myocyte apoptosis and fibrous degeneration of the myocardium [5, 6].
In the context of tissue remodelling, the identification of endothelial progenitor cells (EPCs) has significant clinical implications for the repair processes of wound healing , limb ischaemia  and myocardial ischaemia [9, 10]. Reduced numbers of circulating EPCs are commonly observed in pathological conditions such as diabetes and cardiovascular disease [11, 12, 13, 14]. Despite ongoing controversies regarding the nature and role of EPCs in vasculogenesis [12, 15, 16, 17, 18], the ability of human EPCs to rescue diminished blood flow in a number of animal models [19, 20] provided a rationale for clinical trials of EPC infusion therapy in humans. Pre-clinical studies found infusion of CD34+ and CD133+ EPCs to be safe and beneficial, but the effects of EPCs in humans were less robust . Nevertheless, since the discovery of EPCs , significant steps forward have been taken to better define and functionally characterise these cells, and in the light of the enormous therapeutic potential of modulating vasculogenesis in disease, efforts continue to better understand their origin, plasticity and role in the vasculature.
In this study, we reveal that DSG2 expression and function extend beyond myocardial cells and the epithelium to hematopoietic stem and progenitor cells (HSPCs), EPCs and some mature endothelium but not mature hematopoietic cells. Through functional studies, we demonstrate that DSG2+ EPCs are pro-angiogenic, and the generation of a Dsg2 loss-of-function strain of mice has revealed that this surface-expressed cadherin regulates EC morphology and is important for vascular sprouting and colony formation ex vivo, as well as vessel formation in vivo. Moreover, we demonstrate that despite ECs being a non-desmosome-forming cell, reduction of DSG2 on these cells significantly impacts on their cell–cell adhesive capacity which is likely via reduced DSG2–DSG2 homotypic interactions. Taken together, we provide novel insights into an underappreciated role for DSG2 in hematopoietic cells and the vasculature.
Materials and methods
The collection of primary human umbilical vein endothelial cells (HUVEC), mononuclear cells (MNC) from buffy coats or freshly collected peripheral blood, human mesenchymal stromal cells, gingival and periodontal stem cells, healthy donor peripheral blood, bone marrow, normal tissue as well as cancerous tissue was approved by the Human Research Ethics Committee of the Royal Adelaide Hospital (RAH), Adelaide, South Australia. The collection of primary human umbilical cord blood (UCB) was approved by the Human Research Ethics Committee of the Children, Youth and Women’s Health Service (CYWHS), North Adelaide, South Australia. Animal experiments were approved by the Animal Ethics Committees of SA Pathology and the Peter MacCallum Cancer Centre (protocol E526) and conformed to the guidelines established by the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’.
Isolation and culture of UCB CD133+ non-adherent endothelial forming cells (naEFCs)
UCB (20–130 ml) was obtained from healthy pregnant women undergoing elective caesarean section and collected into MacoPharma cord blood collection bags (MSC1201DU; MacoPharma, Mouvaux, France). CD133+ cells were isolated prior to naEFC cell culture using published methods .
Peripheral blood MNCs, human umbilical vein endothelial cells (HUVEC), bone marrow endothelial cells (BMEC) and normal human bone marrow
Peripheral blood from healthy individuals was collected in lithium heparin coated Vacuette tubes (Greiner Bio-One, Kremsmuenster, Austria) or was provided as buffy coats from the Australian Red Cross Blood Service. For most experiments, MNCs were isolated using Lymphoprep. However, for analysis of VEGFR2+ EPCs, whole blood was subjected to erythrocyte lysis using PharmLyse (BD, Franklin Lakes, NJ, USA) followed by depletion of mature leucocytes using the Lineage Cell Depletion kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s instructions. Primary HUVEC were extracted from human umbilical veins by collagenase digestion and cultured in HUVE medium as previously described [23, 24] and were used for no more than two passages. Human bone marrow endothelial cells (TrHBMEC) were a kind gift from B Weksler (Cornell University Medical College, NY, USA) [25, 26] and hereafter labelled as BMEC. Normal human bone marrow samples were pre-filtered through a 70-µm nylon filter (BD Falcon) to remove debris and then subject to red blood cell lysis using PharmLyse (BD) according to the manufacturer’s instructions prior to flow cytometric staining and analysis.
Induced pluripotent, dental pulp and mesenchymal stem cells
Bone marrow-derived human mesenchymal stem cells (Merck Millipore, NSW, Aust.) were cultured as per manufacturers’ instructions. Induced pluripotent stem (iPS) cells were generated and confirmed for pluripotency as previously described . Similarly, dental pulp stem cells (DPSCs) were isolated from dental pulp tissue and enzymatically digested as per previous instructions .
Flow cytometric analysis of cell surface protein expression
Staining was performed in complete RF-10 medium (RPMI + 10 % FCS) with cells (<1 × 106) blocked with 1 mg/ml human IgG, incubated with unconjugated mAbs (Online resource Supp. Table 1), washed and incubated with secondary antibody. After washing, 5 μl of normal mouse serum was added and conjugated mAbs were then added to the cells. After washing, cells were fixed in 1 % formaldehyde, 20 g/l glucose, 5 mM sodium azide in PBS (‘FACS-Fix’), prior to analysis using either an Accuri C6 Cytometer (BD) or Gallios Flow Cytometer (Beckman Coulter, Brea, CA, USA). Further analysis was performed using FCS Express 4 Flow Cytometry: Research Edition (De Novo Software, CA, USA).
DSG2+ cell enrichment by flow cytometric sorting and expansion
DSG2+ cells were isolated from UCB or fresh bone marrow (BM)-derived MNCs using a FACSAria II (BD). MNCs were co-stained with anti-DSG2, anti-CD34-PerCP-Cy5.5 and anti-CD45-FITC. Within the CD34+CD45dim population, the DSG2+ and DSG2− cells were identified and sorted by FACS, with the sorting gate position based on the fluorescence-minus-one (FMO) control. Immediately following isolation, a small portion of the sorted cells was re-analysed on the FACSAriaTM II (BD) to assess purity. Sorted DSG2+ and DSG2− cells were cultured in EPC expansion CellGro media (CellGenix, Freiburg, Germany) as previously described .
Detection of human DSG2 mRNA using quantitative polymerase chain reaction (qPCR)
Quantification of mRNA levels was carried out using qPCR. Primers were designed for human DSG2 (F-5′-CCATTGCGGAAAGGGCGCCA-3′, R-5′-GGCAGAAATGATGGCACCACCTTGT-3′) using Primer Blast (NIH, MD, USA) and purchased from GeneWorks (Hindmarsh, SA, Aust.). Quantitative PCR amplification was performed using QuantiTect™ SYBR Green master mix (QIAGEN, Hilden, Germany) on a Rotor-Gene thermocycler (Corbett Research, Mortlake, NSW, Aus.) with reaction parameters: 15 min at 95 °C, then cycling of 10 s 95 °C, 20 s 55 °C and 30 s 72 °C; for 45 cycles followed by a melt phase. Resultant data were analysed using Rotor-Gene Analysis Software version 6 (Corbett Research). Relative gene expression levels were calculated using standard curves generated by serial dilutions of cDNAs normalised to the human house-keeping gene cyclophilin A (CycA) (F-5′GGCAAATGCTGGACCCAACACAAA-3′, R-5′CTAGGCATGGGAGGGAACAAGGAA3′).
Hematopoietic colony formation assay
Freshly isolated DSG2+ naEFCs (±expansion) (~1800 DSG2+/− sorted and expanded cells/well or ~500 freshly sorted cells/well) were plated in methylcellulose and cultured with growth factors as previously described  with hematopoietic colonies scored after 14 days.
Generation of the Dsg2 ‘loss-of-function’ mouse strain
The pipeline for the design and construction of the Dsg2 ‘loss-of-function’ mouse strain was based on the publication by Skarnes et al. . Briefly, the CSD (Children’s Hospital Oakland Research Institute/Wellcome Trust Sanger Institute/University of California at Davis) approach incorporates a promoter-driven targeting cassette for the generation of a ‘knockout-first allele’ wherein loxP and FRT sites are placed in introns of genes of interest . More specifically for this study, this ‘gene trap’ approach utilised a targeted insertion of a FRT-flanked lacZ-neomycin cassette into intron 1 of the murine Dsg2 locus of in embryonic stem (ES) cells (i.e., clone EPD0156_5_E04; EuMMCR, Neuherberg, Germany). The Dsg2tma1 strain of mice were generated via ES cell microinjection into Balb/c blastocysts to generate chimeric mice which were then confirmed to be heterozygous for insertion of the lacZ-neomycin construct prior to being inter-bred bred for germ line transmission using C57Bl/6 N mice at the Australian Phenomics Network (APN, Melbourne, Vic. Aust.). These mice yielded fertile wildtype (WT), heterozygous and homozygous offspring at the expected Mendelian ratios (1:2:1), with litter sizes similar to WT C57BL/6N controls.
Detection of murine Dsg2: genomic DNA and mRNA
Genomic DNA was extracted from mouse-tail tips using a REDExtract-N-Amp Tissue PCR Kit (Sigma-Aldrich, St. Louis, MO, USA) prior to genotype confirmation via PCR with reaction parameters as follows: 2 min at 94 °C, then cycling of 30 s at 94 °C, 30 s at 61 °C and 45 s at 72 °C; for 35 cycles followed by 4 °C. The following primers were used to determine WT, heterozygous and homozygous Dsg2tma1 genotypes; P1: 5′CTGCTGCAACTTGATGTCC3′, P2: 5′GAAAGGCGGTTCTGTAATTCC3′ and P3: 5′CAACGGGTTCTTCTGTTAGTCC3′.
Total RNA was extracted from hearts and skin harvested from C57Bl/6 wildtype (WT), Dsg2tma1 heterozygous and homozygous mice. Hearts were homogenised in Trizol prior to chloroform extraction. The aqueous layer containing the RNA was then isolated and 70 % ethanol added before completing the extraction method using an RNEasy Mini Kit (QIAGEN). 0.5–1 μg of total RNA was converted into first-strand cDNA using Superscript III reverse transcriptase (Life Technologies, Carlsbad, CA, USA) following the manufacturer’s instructions.
Dsg2 gene knockdown was validated by qPCR using primers designed to span an intron/exon border between exons 14 and 15 (F-5′AACGAAGCCGTAAGGACAAG3′ R-5′GCCGCTTTCTCTGTGAAGTA3′) via Primer Blast, and purchased from GeneWorks. qPCR amplification was performed using QuantiTect™ SYBR Green master mix (QIAGEN) on a Rotor-Gene thermocycler (Corbett Research) with reaction parameters: 15 min at 95 °C, then cycling of 10 s 95 °C, 20 s 55 °C and 30 s 72 °C; for 45 cycles followed by a melt phase. Data obtained were analysed using Rotor-Gene Analysis Software version 6 (Corbett Research). Relative gene expression levels were calculated by normalising to the mouse house-keeping gene hprt (F-5′CCCAGCGTCGTGATTAGCG3′) (R- 5′GCA CACAGAGGGCCACAATG3′) using geNorm software .
Immunohistochemistry and immunofluorescence
Mouse hearts were fixed in PBS containing 4 % formaldehyde, washed and embedded in paraffin; 4-μm-thick tissue sections were cut and stained with Mayer’s hematoxylin and eosin staining (H&E) (Dako Agilent Technologies, Glostrup, Denmark).
Tissue microarray (TMA) slides were purchased from BioChain (Newark, CA, USA). Slides were dewaxed and rehydrated prior to pressure cooker antigen retrieval in 10 mM citrate buffer (pH 6) for 20 min. After cooling to 30 °C, the sections were incubated for 60 min at room temperature with primary mAb directed against DSG2 (1:50 dilution; clone 3G132; Abcam) or an isotype-matched negative control. The polymer system ADVANCE HRP (Dako Australia Pty Ltd, Vic., Australia) employing DAB as the detection system was used for detection of primary mAb binding. Counterstaining was performed using Mayer’s hematoxylin.
Immunofluorescence microscopy was performed on paraffin-embedded human cervical sections with heat mediated antigen retrieval in 10 mM citrate buffer (pH 6.5) incubated overnight at 4 °C with a primary antibody against DSG2 (2 μg/ml, clone 6D8, Invitrogen, Carlsbad, CA, USA) or CD31 (2 μg/ml, pAb, Bethyl Laboratories, Montgomery, TX, USA). Sections were washed and secondary antibodies (goat anti-mouse Alexa fluor-594 (1:500, Abcam, Cambridge, UK) or goat anti-rabbit Alexa Fluor-488 (1:500, Abcam) applied for 1 h at RT prior to washing and mounting in ProLong Gold with DAPI (Life Technologies). Sections were analysed using an Olympus CV1000 spinning disk microscope and integrated software.
For whole mount staining of adult mouse ear skin, ears were fixed in 4 % paraformaldehyde overnight at 4 °C. Skin was then peeled away from cartilage and incubated in blocking solution (PBS containing 1 % BSA and 0.3 % Triton X-100) overnight at 4 °C. Following blocking, ear skin samples were incubated with rat anti-CD31 (BioLegend, San Diego, CA, USA), rabbit anti-LYVE-1 (Angiobio, Del Mar, CA, USA) and Cy3-conjugated anti-SMA (Sigma-Aldrich) diluted in blocking solution overnight at 4 °C and were then washed extensively in PBS-0.1 % Triton X-100 (6 × 1 h) at 4 °C. Ear skin samples were then incubated with Alexa Fluor®-conjugated secondary antibodies (Invitrogen) overnight at 4 °C and washed as outlined above. Samples were mounted using DAPI FLUOROMOUNT™ G (ProSciTech). Images were captured using a Zeiss LSM 700 confocal microscope (Zeiss Laboratories, Jena, Germany) and a Zeiss inverted Axio Observer Z1 fluorescence microscope.
For immunofluorescence staining of mouse pancreases, mice were humanely killed via cervical dislocation and the whole pancreas was removed. Pancreases were washed in cold phosphate buffered saline (PBS; Life Technologies) and fixed in 4 % paraformaldehyde (Chem Cruz, Santa Cruz Biotechnology Inc., Dallas, TX, USA). They were then embedded in OCT (TissueTek, Sakura Finetek, Torrance, CA, USA), frozen using dry ice and cut into 5-μm-thin sections using a cryostat (CM1520, Leica Biosystems, North Ryde, NSW, Aust.). Frozen sections were probed with primary antibodies: rat anti-mouse CD31 (Biolegend), Cy-3-conjugated anti-SMA (Sigma-Aldrich) or isotype control of rat IgG (Abcam) followed by goat anti-rat Alexa Fluor-488 or -594 (Life Technologies). DAPI Pro-long gold anti-fade mounting medium (Life Technologies) was used for nuclear staining and mounting with images captured using a Zeiss LSM710 multiphoton or Olympus IX71 (Olympus Corporation, Tokyo, Japan) fluorescence microscope.
Immunofluorescence of siRNA-KD BMEC cells was undertaken by fixing cells in either 2 % paraformaldehyde or cold 1:1 acetone:methanol (for DSG2) 72 h post-transfection. After 5 % goat serum blocking, cell-containing coverslips were labelled overnight with the following mouse monoclonal antibodies: VE-cadherin (1:50; clone F-8-Santa Cruz Biotechnology) and DSG2 (1:50; clone 6D8-Thermo Fisher). Primary antibody binding was detected using 1:500 goat anti-mouse IgG conjugated to AlexaFluor 488. To label F-actin filaments, cells were incubated with 1:1000 Phalloidin-Rhodamine (Thermo Fisher) for an hour. Coverslips were mounted on glass slides in Prolong Gold Antifade Reagent with DAPI (Thermo Fisher). Images were captured using Olympus CV1000 with a 40× objective, and analysis of VE-cadherin staining was undertaken using an ImageJ macro. Staining was split into respective channels, i.e., blue (nuclei) and green (VE-cadherin). This macro-determined the total positive areas and number of nuclei per image. Results were presented as the total fluorescence area over nuclei count. To characterise cell phenotypes of DSG2-KD cells, phalloidin staining was analysed in a randomised and blinded manner. Cells were categorised and counted as either cortical actin (circumferential) or stress fibre (mild to extensive actin network) and the percentage was determined and averaged for each siRNA and control. In addition, Image J was used to determine the localisation of VE-cadherin protein via pixel intensity as previously described .
Matrigel tube formation assay: in vitro
In vitro angiogenesis assay of HUVEC and DSG2+ sorted cells was performed using a Matrigel matrix as previously described . Briefly, cells were seeded together on a layer of Matrigel (In vitro Technologies, Noble Park, Vic., Aust.) at a cell density of 1.7 × 104 HUVEC, with or without 1.0 × 104 DSG2+ sorted cells/well, in duplicate. Similarly, the BMEC cells were seeded at 1.5 × 104 cells per well in triplicate. Images covering the entire well were captured after 5–6 h using an inverted imaging microscope (EVOS XL, Life technologies), merged using Adobe Photoshop and tube-like structures manually counted using Image J software (1.47v, NIH, Bethesda, MD, USA).
Matrigel plug assay: in vivo
Female Dsg2 +/+ (wildtype, WT) and Dsg2 lo/lo mice were used between 6 and 8 weeks of age and housed in specific pathogen-free conditions in the SA Pathology Animal Care Facility with Matrigel plug assays conducted as previously described . Briefly, 500 µl of Matrigel containing 2 µg/ml basic fibroblast growth factor (R&D Systems, Minneapolis, MN, USA) and 50 units/ml heparin (Sigma-Aldrich) were injected subcutaneously into both flanks of 6–8 week-old mice. Mice were humanely killed by cervical dislocation on day 7 post-injection and Matrigel plugs removed, washed in PBS, fixed in 4 % paraformaldehyde (VWR International, Radnor, PA, USA) overnight, embedded in paraffin, cut into 8 µm sections on a microtome (RM 2235 Leica, Solms, Germany) and placed onto slides (Polysine®, Menzel-Gläser, Thermo Scientific). The sections were stained for CD31+ ECs (goat anti-mouse/human CD31, Santa Cruz, Dallas, TX) with biotinylated rabbit anti-goat secondary (Abcam), followed by Vectastain elite ABC regent (Vector Labs, Burlingame, CA, USA) for DAB staining as per manufacturer’s instructions and hematoxylin counterstaining. Negative controls included isotype-matched irrelevant antibodies. Images of sections were collected using a Hamamatsu Nanozoomer slide scanner with erythrocyte-containing CD31+ vessels counted manually and section areas quantified using the NDPview software (Hamamatsu Photonics, Hamamatsu, Japan).
Mouse endothelial cell colony-forming assay
Using published methods, whole BM cells were flushed from femurs and tibiae of Dsg2 +/+ and Dsg2 lo/lo mice and adherent EC colony units were defined as a cell mass composed of a central cord of round cells with elongated spindle-shaped cells sprouting at the periphery .
Aortic ring assay
The aortic ring assay was performed on thoracic aorta-derived rings from 12- to 14 week-old Dsg2 +/+ or Dsg2 lo/lo mice as described  where aortic rings were observed over a period of 6 days without or with VEGF (30 ng/ml) added at day three. Aortic sprouts were stained with rat anti-CD31 (BioLegend) and goat anti-rat Alexa Fluor-594 (Life Technologies) to confirm endothelialisation, captured by a Zeiss LSM 700 confocal microscope (Zeiss Laboratories, 10× obj) and quantitated using phase-contrast microscopy via an inverted IX70 microscope 4x/0.13NA obj, an S15 F view camera and Analysis Life Sciences software (Olympus). Number of rings quantitated per group was 3–4.
B16 melanoma mouse model
To evaluate CD31+ vessels in tumours, 1.4 × 104 B16 mouse melanoma cells were injected into the right flanks of age-matched, syngeneic (C57Bl/6 N) Dsg2 +/+ (WT) and Dsg2 lo/lo mice and resultant tumours harvested at endpoint. Sections were cut from FFPE-stored tumour tissues and labelled with anti-CD31 antibody (BD Biosciences). Secondary detection was performed using biotinylated anti-rat-Ig, streptavidin-conjugated horse-radish peroxidase and 3-amino-9-ethylcarbazole chromogen. Stained sections were scanned using an Olympus VS120 scanner and CD31+ blood vessels were counted in each field to obtain an average number of CD31 + blood vessels per field.
DSG2 gene knockdown using siRNA
‘Trilencer-27’ siRNA duplexes targeting human DSG2 were purchased from Origene Technologies (Rockville, MD, USA). Transfection was performed using Lipofectamine RNAiMax (Thermo Fisher) according to the manufacturer’s protocol, using a final concentration of 10 nM for both DSG2-targeting and control siRNAs. On the basis of time course experiments, BMEC cells were used in experiments 72 h after knockdown where DSG2 knockdown efficiency was routinely assessed by flow cytometry.
Cell viability was assessed by incubating the cells with annexin V-FITC (1:50 in 1x binding buffer; BD Bioscience) for 15 min at RT prior to incubation with 7-AAD (BD Biosciences) for 5 min at 4 °C. Stained cells were then resuspended in FACS-Fix and analysed by flow cytometry as described above.
EC cell adhesion assay
BMEC (±DSG2-targeting or control siRNA) transfected 48 h prior were cultured at 1 × 106 cells/35 mm × 10 mm dish (Corning, Sigma-Aldrich) to form a confluent monolayer. Twenty-four hours later, 0.5 × 106 of 72 h-transfected BMEC (±DSG2-targeting or control siRNA) were labelled with 0.5 µM Calcein-AM (eBiosciences, San Diego, CA, USA) and incubated on top of the confluent monolayer for 15 min after which time the non-adherent cells were washed with Hanks Balanced Salt Solution (HBSS, Sigma-Aldrich) using a parallel plate flow chamber system (GlycoTech Corporation, Gaithersburg, MD, USA) with a flow rate of 0.156 ml/min for 2 min. Fluorescently labelled adherent cells were imaged using an inverted epifluorescence microscope Olympus IX71 and a 4×/0.10 objective. Adherent cells were quantified using ImageJ in a blinded and randomised manner.
Inverse invasion assay
Adapted from Hennigan et al. , 100 µl of growth factor-reduced Matrigel diluted 1:1 in cold PBS was added into an 8.0-µm-pore-sized Transwell (Corning Incorporated) and allowed to set. Transwells were then inverted, and 3000 siRNA-transfected (at 24 h) BMEC were seeded onto the underside of the membrane. Four hours later, the unbound cells were rinsed and Transwells immersed right-way up in serum-free HUVE media; 25 ng/ml VEGF in 10 % serum was added to the upper chamber as chemoattractant, and cells allowed to migrate upward into the Matrigel for 96 h. Transwells were paraformaldehyde fixed, RNAse-treated (100 µg/ml, Affymetrix USB) and stained with 0.05 mg/ml of Propidium Iodide (Thermo Fisher), and all steps were carried out for 30 min with two PBS washes between each. Transwells were imaged at fixed intervals (6.68 μm) starting at the membrane and in a direction towards the chemoattractant using z-stack setting of Zeiss LSM 700 confocal microscope with a 10× objective. Cells from 3 fields of view per slice of distance travelled were quantified using ImageJ and then averaged.
Results were expressed as mean ± standard error of the mean (SEM) from at least 3 experiments. Unless otherwise stated, an unpaired Student’s t test, 1- or 2-way ANOVA for multiple comparisons, was performed to determine statistical significance between groups with p values <0.05 considered significant.
Human endothelial progenitor cells (EPCs) express desmoglein (DSG)-2
DSG2 expression in the hematopoietic lineage is restricted to subsets of stem and progenitor cells
Investigation of myeloid lineages revealed DSG2 expression on common myeloid progenitors (CMP; CD34+CD45dimCD117+CD33+CD13+) and on pro-myelocytes, in which CD34 is no longer expressed (CD34−CD45dimCD117+CD33+CD13+) (Fig. 3b). DSG2 was also observed on ~40 % of pro-erythrocytes (CD34+CD45dimCD117+CD71+CD235a−) and B cell lineage progenitors, including pro-B cells (CD34+CD45dimCD22+CD19−CD10−CD20−) (Fig. 3c, d).
For each of these lineages, DSG2 expression progressively decreased as a function of differentiation and was absent on all analysed mature cell types in the BM (Fig. 3b–d). This is in keeping with our observations of mature leucocyte populations in peripheral blood being negative for DSG2 (CD3+ T cells, CD14+ monocytes, CD16+ NK cells and CD19+ B cells) (Online resources Supp. Figure 2). Interrogation of publically available microarray data using the HemaExplorer tool  revealed near perfect concordance with our analysis at the gene expression level. Interestingly, further investigation revealed that DSG2 gene expression correlates tightly with expression of PROM1, which encodes the CD133 progenitor marker (R 2 = 0.92) as well as expression of CD34 (R 2 = 0.60).
Using flow cytometric analysis, we also examined expression of DSG2 on other stem cell populations and observed that DSG2 was strongly and uniformly expressed on three independently generated lines of induced pluripotent stem (iPS) cells (Fig. 3e). Interestingly, the parental fibroblast lines from which the iPS cells were derived completely lacked DSG2 expression (Online resources Supp. Figure 3). Also shown in Fig. 3e, three independently generated lines of dental pulp stem cells (DPSCs) and repeated measures of a MSC-like line demonstrated that DSG2 was not expressed at detectable levels by these progenitor cell types. Thus, DSG2 is expressed by pluripotent stem cells but not by stem cells with restricted mesenchymal differentiation potential.
DSG2+ progenitor cells expand in vitro and exhibit multi-potent differentiation potential
We next conducted cell fate analyses of the expanded DSG2+ versus DSG2− cell populations using methylcellulose hematopoietic colony formation assays. DSG2+ cells cultured with GM-CSF, SCF, IL-3 and EPO for 14 days produced various colony types, including bi-potent progenitors of the erythrocyte and myeloid lineages (CFU-GEMM) and erythrocyte lineage (BFU-E) and granulocyte/monocyte lineage (CFU-GM, CFU-G and CFU-M) progenitors (Fig. 4c). Notably, the DSG2+ cells generated substantially more BFU-E and CFU-G colonies compared to DSG2− cells (Fig. 4c). Similarly, freshly isolated, unexpanded DSG2+ cells purified from fresh human adult BM gave rise to significantly more BFU-E, CFU-G and CFU-M colonies compared with freshly isolated, unexpanded DSG2− cells (Fig. 4d). These findings suggest that CD34+CD45dimDSG2+ hematopoietic progenitor cells have multi-lineage differentiation potential and enhanced colony formation capacity compared to their DSG2− counterparts.
We next examined the capacity of the DSG2+ sorted cells to promote endothelial tube-like formation using Matrigel [22, 29]. HUVEC were seeded either alone or after mixing with expanded DSG2+ cells, and the formation of capillary-like networks was monitored over a 6 h period. Quantitation of vessel formation identified a significant increase in the number of tube-like structures from HUVEC in the presence of DSG2+ cells, suggesting that DSG2+ cells are pro-angiogenic (Fig. 4e). As expected from their immature phenotype, DSG2+ cells were unable to form tubes in the absence of HUVEC (data not shown).
Heterogeneous expression of DSG2 in mature ECs in vivo
In silico analysis of publically available microarray data revealed that Dsg2 is also heterogeneously expressed by the mouse vasculature. In a recent study, Lee and colleagues generated whole genome expression data from lymphoid tissue-derived ECs isolated from either high endothelial venules (HEV) or capillaries (GSE58056) . Our interrogation of this data using GEO2R revealed significantly elevated levels of Dsg2 in the HEV-derived ECs compared to capillary-derived ECs from the lymphoid tissue (Fig. 5c). No difference in expression was observed for other desmosomal cadherins (Dsg1&3 and Dsc1-3) (Fig. 5c).
Generation of Dsg2 lo/lo mice
Upon examination of Dsg2 gene expression in these mice, we observed a significant reduction in Dsg2 mRNA in the heart (where Dsg2 is strongly expressed ) in heterozygous and homozygous mice compared to WT mice (Fig. 6b), indicating perturbed Dsg2 transcription with this ‘knockout-first gene trap’ approach. Similar results were observed in the skin (data not shown). This observation of reduced Dsg2 expression in the heart prompted us to examine for cardiac fibrosis, a previously reported phenotype of Dsg2-knockout mice [5, 43] and humans with DSG2 mutations [3, 6]. In dissected hearts of the homozygous mutant mice, fibrotic lesions were frequently seen in the enlarged hearts (Fig. 6c, arrow). Fibrotic lesions were confirmed by histological analysis using a hematoxylin/eosin stain with an arrow identifying a fibrotic lesion (Fig. 6d). Notably, this pathology was less evident in the heterozygous mice (Fig. 6c, d). We thus concluded that we had generated a ‘hypomorphic’ Dsg2 allele (termed Dsg2 lo) which resulted in normal viability and fertility and is in contrast to the embryonic lethality observed in the Dsg2-knockout mice detailed in Eshkind et al. . Based on these observations, we reasoned that our Dsg2 line would provide a useful tool for evaluating the in vivo consequences of deficiency.
A role for DSG2 in neoangiogenesis
Knockdown of DSG2 in human ECs suggests a role in EC barrier integrity, adhesion and migration
This study reveals that, in addition to epithelial cells and cardiomyocytes, DSG2 is also expressed on sub-populations of human blood and BM-derived progenitor cells. We also provide multiple lines of evidence that DSG2 is expressed by cells of the endothelial lineage and that it contributes to neoangiogenesis. Our findings provide critical new information that DSG2 is much more than a simple desmosomal adhesion molecule.
In their canonical role, the Ca2+-binding transmembrane glycoprotein desmosomal cadherins DSG1-4 and DSC1-3 associate to form homotypic and heterotypic interactions with adjacent cells, resulting in specialised junctions designed to withstand mechanical stress in the myocardium and stratified epithelium [3, 4]. Here we have uncovered an additional role for DSG2, in non-desmosome-forming progenitor cells and endothelial cells. We first identified DSG2 as a highly expressed gene and cell surface protein on human EPCs. Further investigation revealed that DSG2 is also expressed on the closely related hematopoietic progenitor cell lineage, being detectable on early hematopoietic progenitors, common myeloid progenitors, pro-erythroblasts and pro-B-cells, but not on the more differentiated progenitors or mature leucocyte subsets. Our observation of a tight correlation between DSG2, CD133 and CD34 genes suggests that DSG2 may play an important, as yet unidentified, role in symmetric cell division and progenitor renewal. CD34 is the classical marker of hematopoietic progenitors and contributes to cell adhesion and hematopoietic differentiation . BM-derived CD34+ cells reconstitute hematopoietic cells  and CD34 is also expressed by EPCs and mature ECs [22, 51]. CD133 is also a marker of hematopoietic stem and progenitor cells (HSPCs) and is restricted to ~20–60 % of the CD34+ cells isolated from adult BM, UCB or peripheral blood , and the CD133+CD34+ fraction of HSPCs has a high capacity for expansion . We demonstrate that DSG2 is expressed almost ubiquitously by CD133+ cells and at highest levels by the CD38−CD90+ subset. Thus, our study identifies DSG2 as a new cell surface marker for the most primitive and proliferative of HSPCs. In data also shown here, we observed that DSG2 is expressed by induced pluripotent (iPS) cells, which supports our contention of this cadherin being a marker of progenitor cells and extends a previous publication of DSG2 being one of the early activated genes in the reprogramming of fibroblasts to iPS cells . Based on these findings, we propose that DSG2 is a new, and potentially useful, marker of primitive HSPCs and we speculate that it may influence their proliferation and survival. One intriguing possibility is that DSG2 on the surface of HSPCs undergoes homotypic interactions with DSG2 expressed by ECs within the BM, thereby contributing an adhesive and/or signalling function within this established stem cell niche .
The studies presented here also provide clear evidence that DSG2 is expressed uniformly by EPCs and by some mature ECs and that it regulates neoangiogenesis. First, we showed that DSG2 is expressed by an enriched population of non-adherent EPCs as well as freshly isolated CD34+CD45dimVEGFR2+ EPCs from blood. This finding is supported by the recent work of Patsch et al.  who converted human pluripotent stem cells into functional ECs and investigated their transcriptional gene signature. Our in silico analysis of this publically available gene expression array confirms that human ES cells express DSG2 and reveals that this gene remains detectable, albeit to a lesser extent, in the transformed ECs (data not shown). Further interrogation of the endothelial lineage revealed that, while HUVEC are uniformly negative for DSG2, the human BM EC line (TrHBMEC) contains a distinct subpopulation that is DSG2 positive. This is consistent with our observations in normal human and cancerous tissues where, by immunohistochemistry and immunofluorescence, we detected DSG2 on some but not all blood vessels. Notably, we were able to detect DSG2 on both arterioles and venules and this is supported by our in silico analysis of the gene expression array conducted by Aranguren et al.  wherein we were able to detect DSG2 in both arterial and venous ECs (data not shown). Interestingly, expression of DSG2 within this cohort also varied across samples which supports our observation of differential expression of this cadherin by vascular ECs. Heterogeneity was also observed amongst mouse ECs when we performed an in silico analysis of a recent publication by Eugene Butcher’s laboratory, wherein a gene expression analysis was conducted on ECs isolated from mouse lymphoid organ HEVs and capillaries . In support of our contention that certain ECs can express DSG2, Giusti et al. have also reported DSG2 expression by skin microvascular ECs . It is tempting to speculate that DSG2 is expressed by newly formed vasculature, but the molecular basis underlying differential expression of DSG2 on ECs is yet to be elucidated and is part of ongoing studies within our laboratory.
A role for DSG2 in the vasculature is supported by our ‘loss-of-function’ Dsg2 lo/lo mice which show that Dsg2 promotes BM-derived EPC colony formation, ex vivo vascular sprouting and formation of vessel-like structures in vitro and in vivo. These Dsg2 lo/lo mice were generated using a ‘knockout-first allele’ strategy which ablates gene function by insertion of a cassette into an intron of the intact gene and produces a knockout at the RNA processing level. This strain produced viable and fertile homozygous mice which displayed reduced, but not completely absent, levels of Dsg2; thus, our mice were termed Dsg2 ‘hypomorphs’. The Dsg2 lo/lo mice described herein are comparable to those generated by Krusche et al.  who obtained largely fertile Dsg2-knockout mice but are distinct from the Dsg2-knockout mice published by Eshkind et al. , which were embryonic lethal at the time of blastocyst implantation. Importantly, the cardiomyopathy observed in the Krusche et al.  mice, characterised by fibrotic lesions in the heart, was also evident in our strain of Dsg2 lo/lo mice and is in keeping with the human clinical condition associated with DSG2 ‘loss-of-function’, arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC) [5, 6]. Fundamental differences between these three Dsg2 ‘loss-of-function’ mice warrant further explanation; Eshkind et al. substituted exons 7 and 8 of the Dsg2 gene with a cassette conferring lacZ expression and neomycin resistance in C57BL/6 and 129/Sv mouse strains, while Krusche et al.  utilised a targeting construct for inducible deletion of exons 4–6 in C57BL/6J mice. Our Dsg2 lo/lo mice differ again in that construct insertion is characterised by an in-frame intronic insertion between exons 1 and 2 on the C57Bl/6N background. It is possible that the positioning in intron 1 disrupts an enhancer region, resulting in ‘loss-of-function’, as has been documented for the epithelial cadherin (E-cadherin). To note, Eshkind et al.  demonstrated that the lethality of Dsg2 depletion is influenced by genetic background of the mice, with 129/Sv mice displaying a higher lethality than the C57BL/6 strain.
Historically, studies of DSG2 function have focused on its role in desmosomes, with important clinical implications for understanding the pathogenesis of cardiac disease with increasing reports of dominantly inherited frameshift, splice site, nonsense and missense mutations of DSG2 (mostly in the extracellular domain) in sudden death from ARVC as well as sudden unexplained death [60, 61, 62]. A recent study by Kant et al. showed that mice with cardiomyocyte-specific Dsg2 ablation lack desmosome-like structures in the heart and that other desmosomal cadherins, e.g., Dsc2, are unable to fully compensate for the loss of Dsg2 . However, with the identification of DSG2 on several normal and cancerous cell types that completely lack desmosomes [44, 64], there is increasing evidence for pleiotropic functions of this desmosomal cadherin. For example, Nava and colleagues demonstrated in intestinal epithelial cells that proteolytic cleavage of an intracellular fragment of DSG2 caused an increase in caspase 3 activation, cleavage of poly(ADP-ribose) polymerase (PARP) and cell apoptosis . Brennan et al. also observed that overexpression of DSG2 in suprabasal keratinocytes induced hyperproliferation, resistance to anoikis and enhanced carcinogenesis . Reports also document ‘non-junction-bound’ DSG2 on epithelial cells and cancer cells where it provides additional biological roles including regulation of apoptosis , proliferation [66, 67] and vasculogenic mimicry in melanoma . Whether such effects are primary (a direct consequence of the change in DSG2 expression) or secondary (a response to changes in cell adhesion) is not certain. However, the ProteinPredict program of ExPASy indicates many potential protein kinase C (PKC)-target motifs exist in DSG2, including 13 in the cytoplasmic domain , and epidermal growth factor treatment of A431 epithelial cells induces tyrosine phosphorylation of DSG2 . Moreover, the aforementioned Brennan et al. study of suprabasal expression of DSG2 in keratinocytes activated multiple pathways, including PI3K/Akt, MAPK/ERK, STAT3 and NFκB . Thus, multiple lines of evidence suggest that DSG2 may contribute to intracellular signalling events in addition to its established role in cell adhesion.
In addition to the well-recognised link between DSG2 mutation and cardiomyopathy in humans, a second clinical manifestation linked to DSG2 loss-of-function has been identified in patients with systemic sclerosis (SSc, scleroderma). These patients exhibit insufficient angiogenesis in connective tissues, which is associated with a reduction in DSG2 expression in microvascular ECs . This observation supports our contention that DSG2 regulates EC function and angiogenesis. The precise role for DSG2 in systemic sclerosis and angiogenesis is still unclear, with recent evidence suggesting impaired actin assembly and correlation with integrin β8 complex formation . Interestingly, our observations contradict this study as we demonstrate that that loss of DSG2 by ECs induces stress fibre formation rather than impairing it. An important point of difference between these two studies is that we examined actin assembly in a confluent monolayer of ECs, while their coverage of ECs is clearly subconfluent, likely impacting on overall response and outcome. Herein we also show that loss of DSG2 by ECs interrupts VE-cadherin patterning within adherens junctions and culminates in a phenotype with heightened invasion through an extracellular matrix. Our adhesion assays suggest, for the first time, that DSG2 associates to form homotypic interactions between adjacent ECs, which form junctions designed to support vascular development. This is supported by documentation of DSG2–DSG2 interactions within the desmosome complex occurring via the ‘cell adhesion recognition’ (CAR) site. This conserved YAL sequence within the adhesive interface is located within the first of four extracellular repeat domains at the N-terminus of DSG2 and 10-mer peptides targeting this site have proven effective in their inhibition of epithelial cell interactions .
This study has revealed, for the first time, DSG2 expression in distinct progenitor cell subpopulations and shows that, independent from its canonical role as a component of desmosomes, this cadherin also plays a critical role in the vasculature. Because of this work, continued investigation of DSG2-mediated biology in EPCs and ECs, including the intercellular adhesive events, will undoubtedly provide further insight into the pathogenesis of vascular disease and the potential of manipulation of DSG2 for therapeutic use.
This work and CSB were supported by a Heart Foundation Fellowship (CR 10A 4983) and an NHMRC Project grant to CSB and MS (GNT1022150). LME was supported by a Florey Fellowship from the Royal Adelaide Hospital Research Foundation, and MS was supported by a Pfizer Australia Senior Research Fellowship and a veski Innovation Fellowship. We thank A/Prof. Ian Lewis for the collection of normal human bone marrow; Joy Mundy for assistance with cell staining, Drs J Pippal and S Appleby for their assistance with gene expression analysis and validation; S Escarbe for preparing the naEFCs; and the staff and consenting donors at Women’s and Children’s Hospital and Burnside Memorial Hospital for the donation and collection of the umbilical cords and UCB.
Compliance with ethical standards
Conflict of interest
The authors indicate no potential conflicts of interest.
- 5.Pilichou K, Remme CA, Basso C, Campian ME, Rizzo S, Barnett P, Scicluna BP, Bauce B, van den Hoff MJ, de Bakker JM, Tan HL, Valente M, Nava A, Wilde AA, Moorman AF, Thiene G, Bezzina CR (2009) Myocyte necrosis underlies progressive myocardial dystrophy in mouse dsg2-related arrhythmogenic right ventricular cardiomyopathy. J Exp Med 206(8):1787–1802CrossRefPubMedPubMedCentralGoogle Scholar
- 6.Rasmussen TB, Palmfeldt J, Nissen PH, Magnoni R, Dalager S, Jensen UB, Kim WY, Heickendorff L, Molgaard H, Jensen HK, Baandrup UT, Bross P, Mogensen J (2013) Mutated desmoglein-2 proteins are incorporated into desmosomes and exhibit dominant-negative effects in arrhythmogenic right ventricular cardiomyopathy. Hum Mutat 34(5):697–705CrossRefPubMedGoogle Scholar
- 7.Cho HJ, Kim HS, Lee MM, Kim DH, Yang HJ, Hur J, Hwang KK, Oh S, Choi YJ, Chae IH, Oh BH, Choi YS, Walsh K, Park YB (2003) Mobilized endothelial progenitor cells by granulocyte-macrophage colony-stimulating factor accelerate reendothelialization and reduce vascular inflammation after intravascular radiation. Circulation 108(23):2918–2925CrossRefPubMedGoogle Scholar
- 10.Kocher AA, Schuster MD, Szabolcs MJ, Takuma S, Burkhoff D, Wang J, Homma S, Edwards NM, Itescu S (2001) Neovascularization of ischemic myocardium by human bone-marrow-derived angioblasts prevents cardiomyocyte apoptosis, reduces remodeling and improves cardiac function. Nat Med 7(4):430–436CrossRefPubMedGoogle Scholar
- 14.Loomans CJ, van Haperen R, Duijs JM, Verseyden C, de Crom R, Leenen PJ, Drexhage HA, de Boer HC, de Koning EJ, Rabelink TJ, Staal FJ, van Zonneveld AJ (2009) Differentiation of bone marrow-derived endothelial progenitor cells is shifted into a proinflammatory phenotype by hyperglycemia. Mol Med 15(5–6):152–159PubMedPubMedCentralGoogle Scholar
- 19.Asahara T, Masuda H, Takahashi T, Kalka C, Pastore C, Silver M, Kearne M, Magner M, Isner JM (1999) Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. CircRes 85(3):221–228Google Scholar
- 22.Appleby SL, Cockshell MP, Pippal JB, Thompson EJ, Barrett JM, Tooley K, Sen S, Sun WY, Grose R, Nicholson I, Levina V, Cooke I, Talbo G, Lopez AF, Bonder CS (2012) Characterization of a distinct population of circulating human non-adherent endothelial forming cells and their recruitment via intercellular adhesion molecule-3. PLoS ONE 7(11):e46996CrossRefPubMedPubMedCentralGoogle Scholar
- 30.Skarnes WC, Rosen B, West AP, Koutsourakis M, Bushell W, Iyer V, Mujica AO, Thomas M, Harrow J, Cox T, Jackson D, Severin J, Biggs P, Fu J, Nefedov M, de Jong PJ, Stewart AF, Bradley A (2011) A conditional knockout resource for the genome-wide study of mouse gene function. Nature 474(7351):337–342CrossRefPubMedPubMedCentralGoogle Scholar
- 31.Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A, Speleman F (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3(7):RESEARCH0034Google Scholar
- 36.Botta R, Gao E, Stassi G, Bonci D, Pelosi E, Zwas D, Patti M, Colonna L, Baiocchi M, Coppola S, Ma X, Condorelli G, Peschle C (2004) Heart infarct in NOD-SCID mice: therapeutic vasculogenesis by transplantation of human CD34+ cells and low dose CD34+ KDR+ cells. FASEB J 18(12):1392–1394PubMedGoogle Scholar
- 38.Masuda H, Alev C, Akimaru H, Ito R, Shizuno T, Kobori M, Horii M, Ishihara T, Isobe K, Isozaki M, Itoh J, Itoh Y, Okada Y, McIntyre BA, Kato S, Asahara T (2011) Methodological development of a clonogenic assay to determine endothelial progenitor cell potential. Circ Res 109(1):20–37CrossRefPubMedGoogle Scholar
- 45.Schweitzer KM, Vicart P, Delouis C, Paulin D, Drager AM, Langenhuijsen MM, Weksler BB (1997) Characterization of a newly established human bone marrow endothelial cell line: distinct adhesive properties for hematopoietic progenitors compared with human umbilical vein endothelial cells. Lab Invest 76(1):25–36PubMedGoogle Scholar
- 53.Goussetis E, Theodosaki M, Paterakis G, Peristeri J, Petropoulos D, Kitra V, Papassarandis C, Graphakos S (2000) A functional hierarchy among the CD34+ hematopoietic cells based on in vitro proliferative and differentiative potential of AC133+ CD34(bright) and AC133(dim/)-CD34+ human cord blood cells. J Hematother Stem Cell Res 9(6):827–840CrossRefPubMedGoogle Scholar
- 56.Patsch C, Challet-Meylan L, Thoma EC, Urich E, Heckel T, O’Sullivan JF, Grainger SJ, Kapp FG, Sun L, Christensen K, Xia Y, Florido MH, He W, Pan W, Prummer M, Warren CR, Jakob-Roetne R, Certa U, Jagasia R, Freskgard PO, Adatto I, Kling D, Huang P, Zon LI, Chaikof EL, Gerszten RE, Graf M, Iacone R, Cowan CA (2015) Generation of vascular endothelial and smooth muscle cells from human pluripotent stem cells. Nat Cell Biol 17(8):994–1003CrossRefPubMedPubMedCentralGoogle Scholar
- 57.Aranguren XL, Agirre X, Beerens M, Coppiello G, Uriz M, Vandersmissen I, Benkheil M, Panadero J, Aguado N, Pascual-Montano A, Segura V, Prosper F, Luttun A (2013) Unraveling a novel transcription factor code determining the human arterial-specific endothelial cell signature. Blood 122(24):3982–3992CrossRefPubMedGoogle Scholar
- 58.Giusti B, Fibbi G, Margheri F, Serrati S, Rossi L, Poggi F, Lapini I, Magi A, Del Rosso A, Cinelli M, Guiducci S, Kahaleh B, Bazzichi L, Bombardieri S, Matucci-Cerinic M, Gensini GF, Del Rosso M, Abbate R (2006) A model of anti-angiogenesis: differential transcriptosome profiling of microvascular endothelial cells from diffuse systemic sclerosis patients. Arthritis Res Ther 8(4):R115CrossRefPubMedPubMedCentralGoogle Scholar
- 60.Gehmlich K, Asimaki A, Cahill TJ, Ehler E, Syrris P, Zachara E, Re F, Avella A, Monserrat L, Saffitz JE, McKenna WJ (2010) Novel missense mutations in exon 15 of desmoglein-2: role of the intracellular cadherin segment in arrhythmogenic right ventricular cardiomyopathy? Heart Rhythm 7(10):1446–1453CrossRefPubMedPubMedCentralGoogle Scholar
- 61.Gehmlich K, Syrris P, Reimann M, Asimaki A, Ehler E, Evans A, Quarta G, Pantazis A, Saffitz JE, McKenna WJ (2012) Molecular changes in the heart of a severe case of arrhythmogenic right ventricular cardiomyopathy caused by a desmoglein-2 null allele. Cardiovasc Pathol 21(4):275–282CrossRefPubMedGoogle Scholar
- 68.Tan LY, Mintoff C, Johan MZ, Ebert BW, Fedele C, Zhang YF, Szeto P, Sheppard KE, McArthur GA, Foster-Smith E, Ruszkiewicz A, Brown MP, Bonder CS, Shackleton M, Ebert LM (2016) Desmoglein 2 promotes vasculogenic mimicry in melanoma and is associated with poor clinical outcome. Oncotarget (in press)Google Scholar
- 71.Giusti B, Margheri F, Rossi L, Lapini I, Magi A, Serrati S, Chilla A, Laurenzana A, Magnelli L, Calorini L, Bianchini F, Fibbi G, Abbate R, Del Rosso M (2013) Desmoglein-2-integrin Beta-8 interaction regulates actin assembly in endothelial cells: deregulation in systemic sclerosis. PLoS ONE 8(7):e68117CrossRefPubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.