1. Plakoglobin Is Required for Effective Intermediate Filament Anchorage to Desmosomes 
Devrim Acehan, Christopher Petzold, Iwona Gumper, David D. Sabatini, Eliane J. Müller, Pamela Cowin, David L. Stokes 
Journal of Investigative Dermatology 
Volume 128, Issue 11, Pages (November 2008)
DOI: /jid
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

2. Figure 1
Characterization of cell cultures. (a) PCR amplified distinct bands for DNA from wild type and PG-knockout alleles. As expected, DNA from heterozygous cells (PG+/−) produced both bands. Size markers are shown in lanes marked “M”. (b) Immunoblotting verified expression of PG in wild type, but not in knockout cells. Control experiments involved blotting the same protein extracts with antibodies to either β-catenin or E-cadherin.
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

3. Figure 2
Immunofluorescence images from PG+/+ (top row) and PG−/− (bottom row) cell cultures. The primary antibody target is indicated at the top margin. (a, b) As expected, PG is only present in PG+/+ cells and is concentrated at the cell boundaries. (c, d) Plakophilin-1 staining is similar for both cell types. (e, f) Desmoplakin is concentrated at cell boundaries with a less punctate appearance and greater cytoplasmic distribution in PG−/− cells. (g, h) Keratin-14 labeling is similar in both cell types, although the transcell network visible in PG−/− cells was a relatively rare occurrence. (i, j) E-cadherin served as a control and, as expected, produces a similar labeling pattern in both cell types. (k, l) β-Catenin is concentrated along the cell surface in both cell types, but has a more punctate appearance in PG−/− cells. Bar=20μm.
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

4. Figure 3
Broken desmosomes from PG−/− cells. (a,b) Ruptured cells are frequently observed when PG−/− cells are scraped from culture dishes during the EM preparation. Given the consequent mechanical stress, PG-deficient desmosomes frequently break off from one cell, leaving a fragment of the cell membrane. This observation suggests that the intercellular bond provided by the cadherin molecules is intact but that the lack of intermediate filament attachments produces a vulnerability in cell adhesion that ultimately leads to cell rupture. Bars=100nm.
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

5. Figure 4
Projection images of desmosomes. (a) Newborn mouse skin. (b) Primary culture of keratinocytes from newborn mouse skin. (c) PG+/− cell culture. (d–f) PG+/+ cell culture. (g–i) PG−/− cell culture. Note that magnifications are variable, but that all scale bars correspond to 100nm. The outer dense plaque (ODP, large arrowheads) and inner dense plaque (IDP, small arrowheads) are labeled in panel (f) and indicated in panels (d), (e), and (i). Intermediate filaments are indicated by arrows in panel (d). The greater section thickness used for tomography and the retention of soluble cytoplasmic components in freeze-substituted samples cause features to appear less distinct relative to projection images in some previous publications. Similarly, the appearance of cadherins in the intercellular region depends on the precise angle of view, with a better defined midline produced when the section is cut precisely normal to the membrane plane. Taking this effect into account, the intercellular cadherin network appears similar in all these cells, but the PG−/− desmosomes are characterized by a narrower, less dense cytoplasmic plaque with severely reduced connections to the intermediate filament network.
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

6. Figure 5
Tomographic imaging of desmosomes. (a–c) PG+/+. (d–f) PG−/−. Grayscale images correspond to 7-Å thick slices through the tomographic volume and color has been applied to objects representing the three-dimensional segmentation of selected components of the desmosome. The outer dense plaque (ODP) and inner dense plaque (IDP) are indicated in panel (a) and appear to be substantially more developed in wild type desmosomes compared to the knockout. Cell membranes are shown in cyan, desmoplakin in yellow and intermediate filaments in blue. In knockout desmosomes, ribosomes (red) and microtubules (purple) are seen more frequently in the region of the IDP. Close-up views (c, f) show that extracellular cadherin interactions are comparable in the two cell types, although the corresponding densities are not as well preserved as in epidermal tissue (for example, Figures 3a and 7). Animated representations of these tomographic reconstructions are shown in Movies S1 and S2.
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

7. Figure 6
Immunogold localization of desmosomal proteins. Cell type is indicated along the left-hand margin, whereas antibody targets are indicated at the top and bottom. Lower frequency gold labels are marked with arrowheads. (a–c) Three different PG antibodies resulted in robust labeling in PG+/+ desmosomes, but did not label PG−/− desmosomes at all (e–g). (d, h) β-Catenin was absent from PG+/+ desmosomes, but present in PG−/− desmosomes. (i, m) Plakophilin-1 labeling was observed for both cells types, although levels were considerably lower for the knockout cells. (j, k, n, o) Two different antibodies targeted the N- and C-terminal domains of desmoplakin. Although both localized to the cytoplasmic plaque, the N-terminus was significantly closer to the membrane than the C-terminus. Also, fewer desmoplakin antibodies were observed in knockout cells. (l) Histograms illustrate the distribution of antibodies for the two ends of desmoplakin. Student's t-test indicated a significant difference between N- and C-terminal locations in both cell lines (P<0.01) and also a significant difference between C-terminal locations in PG+/+ and PG−/− cells (P<0.02).
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions

8. Figure 7
Segmentation of the inner dense plaque of a desmosome from wild-type mouse epidermis. (a) Overview of this desmosome, which has relatively low density of cadherins and intracellular plaque proteins, making segmentation easier. Segmented area is framed; the inner dense plaque (IDP) and outer dense plaque (ODP) are indicated. (b–e) Sections from lower, middle and upper portions of the tomogram framed in panel (a). (f) Segmentation of the densities for the membrane (cyan), cadherins (purple and pink), desmoplakin (yellow and orange), and intermediate filaments (blue).
Journal of Investigative Dermatology  , DOI: ( /jid )
Copyright © 2008 The Society for Investigative Dermatology, Inc Terms and Conditions