First published online 9 June 2004
doi: 10.1242/jcs.01132
Journal of Cell Science 117, 3129-3140 (2004)
Published by The Company of Biologists 2004
Desmoplakin is required for microvascular tube formation in culture
Xuan Zhou1,
August Stuart2,
Luis E. Dettin1,
Gisela Rodriguez1,
Bonnie Hoel1 and
G. Ian Gallicano1,*
1 Department of Cell Biology, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, 3900 Reservoir Road NW, Washington, DC 20007, USA
2 Interdisciplinary Tumor Biology Program, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, 3900 Reservoir Road NW, Washington, DC 20007, USA
View larger version (79K):
[in a new window]
|
Fig. 1. Localization of DP within capillaries was analyzed by immunofluorescence. (A) Dermal capillary (arrows) in rat skin is recognized by antibodies to VE-Cad (green) and DP (red). Arrowheads point to DP staining within desmosomes that are involved in adhering cells within hair follicles including inner and outer root sheath cells. (B) High magnification of area outlined in A shows that DP (red; arrows) is localized directly adjacent to VE-Cad (green; arrowheads). (C) Confocal view of an E9 day mouse yolk sac shows vimentin (green) associated with DP (red; arrows). D-F are cross-sectional, high-magnification views of the yolk sac that show the close association in F of vimentin (green) with DP (red) surrounding the capillary lumen (lu). Nuclei are stained with DAPI in each figure. Bar in A=100 µm; 10 µm in B; 50 µm in C.
|
|
View larger version (101K):
[in a new window]
|
Fig. 2. DP protein localization was characterized in HMEC-1 and C166 cells. (A-C) HMEC-1 cells grow as monolayers in tissue culture dishes; however,  5 hours after plating onto matrigel they form microvascular tubes (B,C). Inset in C is a toludine blue, semithin section of a microvascular tube grown on matrigel in culture. The lumen (arrows) in these tubes is usually found embedded in the matrigel. In monolayers VE-Cad is localized to cell membranes (D; arrow), while DP (E) appears to be cytoplasmic. (F,G) Once plated onto matrigel, VE-Cad (F) and DP (G) colocalize with one another (H). I and J show morphologically and immunohistochemically the localization pattern of VE-Cad (I; arrowheads) and DP (J; arrows) in microvascular tubes. (K) DP also is localized to cell-cell contacts in C166 microvascular tubes. Note: the microvascular tube shown in F-H is morphologically similar to those found in I-K. Bars, 100 µm (A,B); 50 µm (C and inset); 25 µm (D,E), 20 µm (F-H); 40 µm (I-K).
|
|
View larger version (38K):
[in a new window]
|
Fig. 3. RT-PCR of C166 and HMEC-1 cells grown on matrigel in growth factor containing EGM-2 medium shows DP gene expression at 12 and 24 hours in C166 cells. No DP was detected in C166 cells grown in monolayers in DMEM medium (without growth factors). Tubulin primers were used on the same time 0, C166 mRNA used in lane 1 to show presence of mRNA. +, with reverse transcriptase; –, no reverse transcriptase.
|
|
View larger version (71K):
[in a new window]
|
Fig. 4. To determine if siRNA functioned in HMEC-1 cells, a CMV-GFP gene and its corresponding siRNA were simultaneously transfected into a monolayer of HMEC-1 cells, followed by incubation for 24 hours. A subset of these HDME-1 cells (A,B) were then placed onto matrigel and analyzed for tube formation 12-24 hours later (C,D). GFP siRNA successfully knocked down translation of GFP. Note: GFP siRNA did not affect tube formation. (E-H) As a control, GFP was expressed in HMEC-1 cells without the addition of GFP siRNA. In contrast to cells transfected with CMV-GFP and siRNA, cells without GFP siRNA were found expressing GFP both in monolayers (F) and in their corresponding tubes (H). Bar in A, for A-D, 20 µm; bar in E, for E-H, 20 µm.
|
|
View larger version (137K):
[in a new window]
|
Fig. 5. Microvascular tube formation was compared in the absence and presence of DP siRNA. (A,E) HDME-1 cells in a monolayer do not form tubes. The absence (A) or presence (E) of DP siRNA did not affect the monolayer. (B-D) Cells placed onto a thin layer of matrigel form tubes after 5-8 hours, which persist up to 10 days in culture. Addition of 250 ng of DP siRNA significantly compromised tube formation (E1-4; F-H). E1-4 are four frames taken from a five hour video showing migration patterns of DP siRNA treated cells after plating onto matrigel. Four to five cells within the white bracket in E-1 are joined by a migrating cell (arrowhead E1-4) but do form a tube. The arrow in E1-4 points to another cell migrating towards a stationary cell. Adhesion between cells appears to be compromised up to 72 hours post siRNA treatment (H). Bars, 50 µm (A,B,E,F); 30 µm (C,G); 15 µm (D,H).
|
|
View larger version (14K):
[in a new window]
|
Fig. 10. Tube number and lengths were quantified. (A) Tubes were counted at increasing time points in a one cm2 area from at least two wells per experiment. Each DP siRNA experiment was performed four times on both C166 cells and HMEC-1 cells. (B) Tube lengths were calculated over a 144 hour period.
|
|
View larger version (110K):
[in a new window]
|
Fig. 7. HMEC-1 cells treated with DP siRNA eventually form tubes. (A) By 144 hours after cells were transfected with DP siRNA they proceeded to form microvascular tubes. (B) High-magnification view of a microvascular tube from the same set of cells in Fig. 5H that had been treated with DP siRNA shows cells adhered well to one another. Bar, 100 µm (A); 50 µm (B).
|
|
View larger version (148K):
[in a new window]
|
Fig. 8. Endothelial cells treated with DP siRNA are able to send out cellular projections. (A) Wild-type HMEC-1 cells send out filopodia (arrow) within 2-3 hours of plating onto matrigel. (B) Arrowhead and arrow point to filopodia from adjacent cells coming in close contact with each other. The arrowhead points to a filopodia extending from the middle cell, while the top arrow points to a filopodia extending from the top cell towards the middle cell filopodia. (C) Low-magnification view of cells treated with DP siRNA. Arrows point to long filopodia. (D) Higher magnification of cells in C show at least two filopodia. Arrows point to a filopodia extending from the lower left cell, while arrowheads point to a filopodia extending from the upper right cell. Bar, 10 µm (A, B, D); 25 µm (C).
|
|
View larger version (174K):
[in a new window]
|
Fig. 9. Electron microscopy of HMEC-1 cells grown in culture with DP siRNA show distinct differences in cell-cell adhesion when compared with control cells. (A) Untreated, wild-type (w.t.) cells plated for 5 hours on matrigel showed distinct microvilli extending from each cell (arrows). (B) Arrows point to microvilli between cells treated with DPsiRNA. Note, cells come in close contact with each other. Nu, nucleus. (C) High-magnification view of two juxtaposed cells 5 hours after plating onto matrigel shows two microvilli (from A) sliding across each other. Arrows show direction of progress for each microvilli. Arrowheads point to distinct amorphous material adhering microvilli to each other. (D) DPsiRNA does not appear to compromise microvilli sliding or their adhesion to one another. An amorphous material similar to that seen in untreated cells (C) adhering two microvilli to each other was clearly visible (arrowheads). (E) Later stages of cell adhesion are evident 24 hours after plating onto matrigel. Areas of adhesion between untreated cells commonly spanned >3 µm (arrows), while areas of adhesion between DPsiRNA-treated cells rarely spanned beyond 500 nm (F). While some intercellular spaces were evident between untreated cells (asterisk in E), morphologically similar spaces were commonplace between DPsiRNA-treated cells (asterisks in F). (G-H) High-magnification views of complexus adherens junctions in untreated cells 24 hours after plating onto matrigel. Arrows point to filaments with diameters of 10-12 nm (similar to vimentin) coming into close contact with electron dense plaques (arrowheads in G and H). Bar in G = 100 nm in H.
|
|
View larger version (82K):
[in a new window]
|
Fig. 11. DP gene expression in C166 cells was measured by RT-PCR. (A,B) The DMEM+serum medium normally used to grow C166 cells in monolayers was replaced with EGM-2 medium containing growth factors. Monolayers were placed back into the incubator for specified times. RT-PCR of these monolayers demonstrated that DP gene expression was first detected at 1.5 hours of incubation in both 70% (A) and 90% (B) confluent layers. Tubulin primers were used on the same samples as those used for DP to show the presence of mRNA in all samples. No RT controls also were performed on the same mRNA samples.
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2004
