Advertisement

Large-Scale Electron Microscopy Maps of Patient Skin and Mucosa Provide Insight into Pathogenesis of Blistering Diseases

  • Ena Sokol
    Affiliations
    Department of Cell Biology, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands

    Department of Dermatology, Center for Blistering Diseases, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Duco Kramer
    Affiliations
    Department of Dermatology, Center for Blistering Diseases, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Gilles F.H. Diercks
    Affiliations
    Department of Dermatology, Center for Blistering Diseases, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Jeroen Kuipers
    Affiliations
    Department of Cell Biology, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Marcel F. Jonkman
    Affiliations
    Department of Dermatology, Center for Blistering Diseases, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Hendri H. Pas
    Affiliations
    Department of Dermatology, Center for Blistering Diseases, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
  • Ben N.G. Giepmans
    Correspondence
    Department of Cell Biology, University Medical Centre Groningen, University of Groningen, A. Deusinglaan 1, Groningen 9713 AV, The Netherlands
    Affiliations
    Department of Cell Biology, University Medical Centre Groningen, University of Groningen, Groningen, The Netherlands
    Search for articles by this author
      Large-scale electron microscopy (“nanotomy”) allows straight forward ultrastructural examination of tissue, cells, organelles, and macromolecules in a single data set. Such data set equals thousands of conventional electron microscopy images and is freely accessible (www.nanotomy.org). The software allows zooming in and out of the image from total overview to nanometer scale resolution in a ‘Google Earth’ approach. We studied the life-threatening human autoimmune blistering disease pemphigus, using nanotomy. The pathomechanism of cell–cell separation (acantholysis) that underlies the blistering is poorly understood. Ultrastructural examination of pemphigus tissue revealed previously unreported findings: (i) the presence of double-membrane structures between cells in all pemphigus types; (ii) the absence of desmosomes around spontaneous blisters in pemphigus foliaceus (PF); (iii) lower level blistering in PF when force induced; and (iv) intercellular widening at non-acantholytic cell layers. Thus, nanotomy delivers open-source electron microscopic maps of patient tissue, which can be analyzed for additional anomalies from any computer by experts from different fields.

      Abbreviations

      Dsg1
      desmoglein 1
      Dsg3
      desmoglein 3
      EM
      electron microscopy
      mcPV
      mucocutaneous pemphigus vulgaris
      mdPV
      mucosal dominant pemphigus vulgaris
      PF
      pemphigus foliaceus
      PV
      pemphigus vulgaris
      SSSS
      staphylococcal scalded skin syndrome

      Introduction

      Nanotomy (nano-anatomy), based on large-scale electron microscopy (EM), is a technique that, like Google Earth, allows users to examine tissue from a complete overview to in depth analysis at macromolecular resolution (
      • Kuwajima M.
      • Mendenhall J.M.
      • Lindsey L.F.
      • et al.
      Automated transmission-mode scanning electron microscopy (tSEM) for large volume analysis at nanoscale resolution.
      ;
      • Ravelli R.B.
      • Kalicharan R.D.
      • Avramut M.D.
      • et al.
      Destruction of tissue, cells and organelles in type 1 diabetic rats presented at macromolecular resolution.
      ). Using this technique, large areas of tissue are scanned and presented online. A major advantage of nanotomy over conventional EM is unbiased data acquisition, presentation, and sharing at both high resolution (macromolecule scale), as well as in the context of tissue. Here, we applied nanotomy to study human disease and investigated the skin and mucosal epithelium from patients suffering from the life-threatening autoimmune blistering disease pemphigus.
      In pemphigus, the architecture of the epidermal and/or mucosal epithelia is disrupted by loss of cell–cell adhesion (acantholysis;
      • Stanley J.R.
      • Amagai M.
      Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome.
      ). In healthy epidermis and stratified mucosal epithelium, the cells are organized into multiple layers where specialized cell–cell contacts, desmosomes, interconnect their intermediate filament cytoskeletons, thereby providing tissue strength (
      • Simpson C.L.
      • Patel D.M.
      • Green K.J.
      Deconstructing the skin: Cytoarchitectural determinants of epidermal morphogenesis.
      ). Desmosomes are built from transmembrane cadherins (desmogleins (Dsgs) and desmocollins) and from associated cytoplasmic armadillo proteins and plakins (
      • Garrod D.
      • Chidgey M.
      Desmosome structure, composition and function.
      ;
      • Delva E.
      • Tucker D.K.
      • Kowalczyk A.P.
      The desmosome.
      ;
      • Kowalczyk A.P.
      • Green K.J.
      Structure, function and regulation of desmosomes.
      ). In pemphigus, the autoantibodies are directed to Dsg1 and Dsg3, the major Dsgs of stratified epithelia (
      • Amagai M.
      • Stanley J.R.
      Desmoglein as a target in skin disease and beyond.
      ). These are differentially expressed over the epithelial layers within the epidermis, Dsg1 being present in all cell layers and Dsg3 strongly expressed in the basal layer, but absent in the superficial layers. In contrast, in mucosal epithelium, Dsg3 is present in all layers, but Dsg1 is absent in the basal layer (
      • Mahoney M.G.
      • Hu Y.
      • Brennan D.
      • et al.
      Delineation of diversified desmoglein distribution in stratified squamous epithelia: Implications in diseases.
      ). Pemphigus is divided into two main clinical forms, pemphigus vulgaris (PV), affecting mucosa (mucosal dominant, mdPV) and sometimes skin (mucocutaneous, mcPV), and pemphigus foliaceus (PF) that only affects skin. Blistering is suprabasal in case of PV and subcorneal in case of PF, which is explained by the differential expression of Dsg1 and Dsg3 over the cell layers and the different autoantibody profiles: anti-Dsg3 in case of mdPV, anti-Dsg3 and anti-Dsg1 in case of mcPV, and anti-Dsg1 in case of PF (
      • Mahoney M.G.
      • Wang Z.
      • Rothenberger K.
      • et al.
      Explanations for the clinical and microscopic localization of lesions in pemphigus foliaceus and vulgaris.
      ). In PF and in mcPV––where Dsg1 is affected––the skin becomes vulnerable prior to blistering: by rubbing seemingly healthy skin a blister can be evoked, the so-called Nikolsky sign used in diagnosis.
      The mechanism by which the antibodies evoke acantholysis is still poorly understood. Various contradicting theories have been proposed over the last years, and a predominant recent hypothesis is that the major devastating effect of antibodies is weakening and final collapse or ‘melting’ of desmosomes (
      • Aoyama Y.
      • Nagai M.
      • Kitajima Y.
      Binding of pemphigus vulgaris IgG to antigens in desmosome core domains excludes immune complexes rather than directly splitting desmosomes.
      ;
      • Oktarina D.A.
      • van der Wier G.
      • Diercks G.F.
      • et al.
      IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
      ). Analysis of the fate of cells and desmosomes in the different pemphigus forms at all cellular layers in both the skin and mucosa will help better to understand the underlying mechanism of blistering. Therefore, we implemented nanotomy to pemphigus patient skin and mucosa. Here, we not only share our gigabyte data sets for ongoing analysis by other experts but provide hints of what may be underlying the basic mechanism of pemphigus.

      Results

      Nanotomy allows zooming in and out into the tissue

      The nanotomy data sets of healthy and patient skin and mucosa was made by scanning EM and subsequent stitching (Table 1). The data sets equal tens of thousands traditional high magnification EM images and are open source available for scientists and health-care professionals. We examined these data sets at the level of tissue, cells, organelles, and other ultrastructural macromolecular complexes, with special attention to desmosomes and anomalies in the patient samples (www.nanotomy.org). The power of analyzing over multiple scales is illustrated for normal human skin (Figure 1).
      Table 1Patient and biopsy characteristics
      DiagnosisELISA titers Dsg1/Dsg3BiopsyAbbreviation
      1ControlNAControl skinNA
      2ControlNAControl skinNA
      3Pemphigus foliaceus>150/0Nikolsky positive non-lesional skin*N+PF skin
      4Pemphigus foliaceus110/2Lesional skinPF skin
      5Mucocutaneous pemphigus vulgars80/100Lesional skinmcPV skin
      6Mucosal-dominant pemphigus vulgaris16/>150Non-lesional skinmdPV skin
      7Staphylococcal scaled skin syndromeNALesional skinSSSS skin
      8ControlNAControl mucosaNA
      9Pemphigus foliaceus>150/0Non-lesional mucosaPF mucosa
      10Mucosal-dominant pemphigus vulgaris16/>150Lesional mucosamdPV mucosa
      Abbreviations: Dsg, desmoglein; NA, not applicable.
      Full data of all results are present through www.nanotomy.org
      Biopsies from patients with pemphigus foliaceus, mucosal pemphigus vulgaris, mucocutaneous pemphigus vulgaris, and the staphylococcus scaled skin syndrome were included in this study. ELISA index values of anti-Dsg1 and anti-Dsg3 antibodies were measured at the time of taking the biopsy from the patient. Biopsies were considered non-lesional if taken from an area of seemingly healthy skin where no blister could be induced, or Nikolsky positive if taken from an area where a blister could be induced by rubbing, and lesional if a blister was present.
      Figure thumbnail gr1
      Figure 1Healthy human skin at different zoom levels. (a) Healthy human skin, including part of the dermis and all epidermal layers. The dermo–epidermal border is presented by the striped line. (b) Magnified region (red square in a) showing keratinocytes in the epidermal layers: stratum basale (red), stratum spinosum (purple), stratum granulosum (green), and stratum corneum (blue). (c) Magnified region (purple square in b) showing multiple desmosomes (asterix). (d) Magnified region (yellow square in c) showing outer dense plaque (ODP), inner dense plaque (IDP), and extracellular core domain (ECD) of a desmosome. Red bar=1 μm for all images. Note that we are sliding across scales. For interpretation purposes, the height of the images is noted at the right.

      Histological overview of pemphigus tissue

      Histological changes in main pemphigus types were first addressed at the lowest zoom level in both the skin and mucosa (Figures 2 and 3). The different cell layers in the epidermis and stratified epithelium of the buccal mucosa are readily identified by the distinct morphology of the cells and were pseudocolored: basal layer (stratum basale; red), spinous layer (stratum spinosum; purple), granular layer in the skin or intermediate layer in mucosa (also known as stratum granulosum (skin) or stratum intermedium (mucosa); green), and corneal layer/ superficial layer (stratum corneum in the skin or stratum superficiale in mucosa; blue; Figures 2 and 3. left panels). The regions of intercellular space widening (pink) and acantholysis (yellow) are also indicated (Figures 2 and 3, right panels). In both PF skin biopsies, acantholysis is present but in different cell layers. The blister from Nikolsky-positive PF skin is present in the spinous layer (Figure 2b), whereas the blister in lesional skin is located higher in the granular layer (Figure 2c). In both biopsies, massive intercellular widening is recognizable in the lower cell layers. In non-lesional mdPV skin, no histological signs of pathology can be observed (Figure 2d), whereas in lesional mcPV skin acantholysis and intercellular space widening beneath the blister are present (Figure 2e).
      Figure thumbnail gr2
      Figure 2Intercellular space widening and acantholysis in pemphigus and SSSS patient skin. (a) Healthy human skin, (b) Nikolsky-positive PF skin, (c) lesional PF skin, (d) mdPV skin, (e) lesional mcPV skin, and (f) lesional SSSS skin. Epidermal layers: stratum basale (red), stratum spinosum (purple), stratum granulosum (green), stratum corneum (blue), blister cavity (yellow), intercellular space widening (pink). Bar=50 μm. mcPV, mucocutaneous pemphigus vulgaris; mdPV, mucosal dominant pemphigus vulgaris; PF, pemphigus foliaceus; SSSS, staphylococcal scalded skin syndrome.
      Figure thumbnail gr3
      Figure 3Intercellular space widening and acantholysis in pemphigus patient mucosa. (a) Healthy human mucosa, (b) non-lesional PF mucosa, and (c) lesional mdPV mucosa. Layers of the epithelium of the mucosa: stratum basale (red); stratum spinosum (purple); stratum intermedium (green); stratum superficiale (blue); loss of cell–cell adhesion (yellow); intercellular space widening (pink). Bar=50 μm. mcPV, mucocutaneous pemphigus vulgaris; mdPV, mucosal dominant pemphigus vulgaris; PF, pemphigus foliaceus.
      We also included a lesional skin biopsy of a patient with the staphylococcal scalded skin syndrome (SSSS) that clinically and histologically mimics PF. In SSSS skin, exfoliative toxins of Staphylococcus aureus cleave Dsg1, which leads to superficial acantholysis (
      • Hanakawa Y.
      • Schechter N.M.
      • Lin C.
      • et al.
      Molecular mechanisms of blister formation in bullous impetigo and staphylococcal scalded skin syndrome.
      ;
      • Aalfs A.S.
      • Oktarina D.A.
      • Diercks G.F.
      • et al.
      Staphylococcal scalded skin syndrome: loss of desmoglein 1 in patient skin.
      ). In our tissue sample, acantholysis was complete below the corneal layer (absent from specimen) and extended into the spinous layer (Figure 2f). Blisters on mucosal membranes are a hallmark characteristic for PV. Figure 3c shows lesional mdPV mucosa with acantholysis above the basal cell layer and widening of the intercellular spaces. Although acantholysis is absent in PF mucosa, intercellular widening is present (Figure 3b). In the control mucosa, minimum intercellular space widening is present in the lower cell layers (Figure 3a). Taken together, localization of acantholysis is as expected in all lesional samples, except in Nikolsky-positive PF skin. Furthermore, intercellular space widening is not an exclusive precursor of acantholysis but also happens in layers that never become acantholytic.

      Double-membrane structures

      To address ultrastructural changes within their anatomical context, we further analyzed pemphigus with emphasis on cell–cell contact areas, as desmosomes are targeted by the autoantibodies. A common ultrastructural characteristic of all lesional pemphigus tissues is the presence of areas where plasma membranes of neighboring cells align in close proximity of ~40 nm, which gives them the appearance of peculiar double-membrane structures (Figure 4a–d). We measured stretches of up to 6 μm long. The double-membrane structures appear as interdigitation-like connections (Figure 4a) or as protrusions of one cell into another cell (Figure 4b). Close to both types of such structures, circular structures with the same double-membrane appearance are present (Figure 4d). Using nanotomy we mapped all such structures (yellow dots; Figure 4e–j). Control tissues and SSSS do not contain these double-membrane structures (see online data sets 1,2,7, and 8). In Nikolsky-positive PF skin (Figure 4e) and in lesional PF skin (Figure 4f), the majority of these structures are in the lower layers of the epidermis. Double-membrane structures are also abundant in lesional mcPV skin (Figure 4h) and mdPV mucosa (Figure 4j), whereas in the non-lesional clinically healthy mdPV skin (Figure 4g) and non-lesional PF mucosa (Figure 4i) only a few are present. Close examination of double-membrane structures reveals that they lack normal desmosomes, despite the presence of electron dense structures on one of both membranes or at opposite sides on both membranes. (Figure 4b; black arrows). These electron dense structures, 60–90 nm wide, may represent newly assembled half desmosomes or dismantling desmosomes. In the lesional biopsy, the double-membrane structures had many small invaginations into the cytoplasm, likely representing active endocytic processes (Figure 4c; red arrows).
      Figure thumbnail gr4
      Figure 4Double-membrane structures in pemphigus tissue skin and mucosa. (a) Double-membrane interdigitations in Nikolsky-positive PF skin (e, red square); (b) double-membrane protrusion from mcPV skin (h, pink square); (c) double-membrane structure from lesional PF skin (f, green square); (d) circular double-membrane structures from lesional mdPV mucosa (j, blue square). Black arrows in b indicate electron dense structures. Red arrows in c indicate membrane invaginations. (e) N+PF, (f) PF, (g) mdPV, (h) mcPV skin, (i) PF, and (j) mdPV mucosa. Yellow dots indicate localization of double-membrane structures including interdigitation-like structures, protrusions, and circular structures. Red, green, purple, and blue squares indicate localization of images (ad). Bar in ad = 0.5 μm, bar in ej 50 μm. mcPV, mucocutaneous pemphigus vulgaris; mdPV, mucosal dominant pemphigus vulgaris; PF, pemphigus foliaceus.

      Desmosomes

      In PF and mcPV skin and in mdPV mucosa an overall reduction in size and number of desmosomes is apparent (see data online). In lesional PF skin, desmosomes are completely absent in the cells surrounding the blister cavity. Keratin filaments in these cells do not reach the plasma membrane and remain perinuclear (Figure 5b). Despite the absence of desmosomes, cells remain in close connection without visible intercellular spaces (Figure 5b). In contrast, in Nikolsky-positive PF skin, keratinocytes surrounding the blister cavity contain half desmosomes that are connected to keratin filaments and sometimes appear as small extrusions from the plasma membrane (Figure 5a). Their average size is 178±74 nm and likely represent either newly synthesized half desmosomes or dismantling desmosomes. Similar structures are also found on cells round the blister cavity in mcPV skin (Figure 5c) and in mdPV mucosa (Figure 5d). The average size of half desmosomes in mcPV skin is 174±69 nm and in mcPV mucosa 178±74 nm. In the areas of intercellular widening, pulled out desmosomes are present on cellular extensions that are still connected to the cell. In mcPV epidermis, some desmosomes seem to be ruptured from the keratin filaments (“torn-off desmosomes”, Figure 5e orange arrow).
      Figure thumbnail gr5
      Figure 5Distinct ultrastructure of desmosomes in blister cavities and widened intercellular spaces. (a) Nikolsky-positive PF skin, (b) lesional PF skin, (c) lesional mcPV skin and (d) lesional mdPV mucosa, and (e) area in mcPV skin. Yellow squares present regions of enlarged images on the right side. Black arrows indicate half desmosomes; black arrow heads: no presence of desmosomes; orange arrow: torn-off or pulled out desmosome; and orange stars: keratin filaments that are not retracted from the cell membrane. Notice no signs of nuclear fragmentation. Bar on the left panel: 15 μm; bar on the right panel 0.5 μm. ACANTH, acantholytic cavity; IC, intercellular space; mcPV, mucocutaneous pemphigus vulgaris; mdPV, mucosal dominant pemphigus vulgaris; PF, pemphigus foliaceus. * from the same data set.

      Other anomalies

      We further note that all lesional tissues contain enlarged mitochondria, nuclei lack any signs of apoptosis (Figure 5), and the dermis in the skin and the connective tissue in mucosa are rich in the number of mast cells, lymphocytes, and plasma cells (data not shown in figures, see www.nanotomy.org).

      Discussion

      Here we employ nanotomy to study human disease. Although large-scale EM has been pioneered on animal models, ultrastructural information on human tissue in scientific publications is typically limited to few selected images. Nanotomy data sets, however, give an unbiased in depth view of the complete tissue section and reveal every detail. The image data set is enormous but can be straightforward analyzed as the data are transmitted by quickly streaming from an open source. The data sharing further allows to expand the number of patients, which, in contrast to the data content, is still limited at this stage. Thus, interdisciplinary scientists can check and further explore the original images by browser with simple zoom and pan function or even produce data from their own patients in a similar manner.
      Central in this study are human pemphigus tissue samples. Despite much research, the pathomechanism of acantholysis is still unresolved. Over the years, a variety of pathomechanisms have been proposed including (i) coating by IgG of the Dsg transadhesive binding domains such that two opposing half desmosomes cannot connect anymore (steric hindrance hypothesis;
      • Amagai M.
      • Karpati S.
      • Prussick R.
      • et al.
      Autoantibodies against the amino-terminal cadherin-like binding domain of pemphigus vulgaris antigen are pathogenic.
      ), (ii) impeding of desmosome formation by depletion of the newly synthesized Dsg leading to “melting” away of desmosomes (non-assembly depletion hypothesis;
      • Aoyama Y.
      • Nagai M.
      • Kitajima Y.
      Binding of pemphigus vulgaris IgG to antigens in desmosome core domains excludes immune complexes rather than directly splitting desmosomes.
      ;
      • Oktarina D.A.
      • van der Wier G.
      • Diercks G.F.
      • et al.
      IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
      ), (iii) interaction with signal transduction pathways that lead to destruction of desmosomes (cell signaling hypothesis;
      • Müller E.J.
      • Williamson L.
      • Kolly C.
      • et al.
      Outside-in signaling through integrins and cadherins: a central mechanism to control epidermal growth and differentiation?.
      ), (iv) tearing off of desmosomes from the keratin filaments (basal cell shrinkage hypothesis;
      • Bystryn J.C.
      • Grando S.A.
      A novel explanation for acantholysis in pemphigus vulgaris: the basal cell shrinkage hypothesis.
      ), and (v) apoptosis (
      • Frusic-Zlotkin M.
      • Raichenberg D.
      • Wang X.
      Apoptotic mechanism in pemphigus autoimmunoglobulins-induced acantholysis–possible involvement of the EGF receptor.
      ).
      Our data confirm the recently observed absence of apoptosis (
      • Janse I.C.
      • van der Wier G.
      • Jonkman M.F.
      • et al.
      No evidence of apoptotic cells in pemphigus acantholysis.
      ) as no retraction of pseudopods, pyknosis, karyorrhexis, plasma membrane blebbing, or engulfment by phagocytes was seen.
      In all pemphigus tissue, except mdPV skin where Dsg1 is unaffected, intercellular interdesmosomal widening, especially in the lower layers, is apparent. Intercellular widening is defined as >0.5 μm gap between two neighboring cells, which are still in contact. Until recently, this widening was considered to be the first evidence of acantholysis; however, it is also present in basal and suprabasal layers of the skin and mucosa of PF patients where acantholysis is absent (
      • Guedes A.C.
      • Rotta O.
      • Leite H.V.
      • Leite V.H.
      Ultrastructural aspects of mucosas in endemic pemphigus foliaceus.
      ;
      • Oktarina D.A.
      • van der Wier G.
      • Diercks G.F.
      • et al.
      IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
      ;
      • van der Wier G.
      • Jonkman M.F.
      • Pas H.H.
      • et al.
      Ultrastructure of acantholysis in pemphigus foliaceus re-examined from the current perspective.
      ). Thus, non-junctional Dsg1 apparently is, in a yet-to-be uncovered way, important for interdesmosomal cohesion as loss is associated with widening. In PF, non-junctional Dsg1 is lost when removed into clusters (
      • Oktarina D.A.
      • van der Wier G.
      • Diercks G.F.
      • et al.
      IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
      ), and loss through haploinsufficiency of Dsg1 as in palmoplantar keratoderma also leads to intercellular widening (
      • Bergman R.
      • Hershkovitz D.
      • Fuchs D.
      • et al.
      Disadhesion of epidermal keratinocytes: a histologic clue to palmoplantar keratodermas caused by DSG1 mutations.
      ). Also, in our SSSS biopsy where Dsg1 is cleaved by exfoliative toxins widening is present. Taken together, intercellular space widening is not characteristic for pemphigus only and is here associated with the dysfunction of Dsg1.
      All pemphigus biopsies, but not healthy tissue nor SSSS tissue, typically contain “double membrane structures”. These are more abundant in lesional than in non-lesional tissue. The structures are characterized by two membranes approximately 40 nm apart with highly irregular shapes: interdigitations, protrusions, or circular structures. Separate shapes were found close together in areas that could be up to 3 μm wide. As our sections are only 70 nm thick, these are most likely part of a larger structure: the different shapes likely result from the EM cutting angle. Such structures have been reported before as so-called curvicircular structures in the skin, but not mucosa, of PF patients (
      • Tada J.
      • Hashimoto K.
      Curvicircular intracytoplasmic membranous structures in keratinocytes of pemphigus foliaceus.
      ). We, however, find that the double-membrane structures are not restricted to PF skin as they are also abundant in mcPV skin, mdPV mucosa, and to a lesser extent in mdPV skin and PF mucosa. The distribution of the structures reminds of the distribution of the IgG/Dsg clusters observed before that is believed to be formed through cross-linking of Dsg by the autoimmune IgG (
      • Oktarina D.A.
      • van der Wier G.
      • Diercks G.F.
      • et al.
      IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
      ). Our follow-up studies will address whether the double-membrane structures indeed represent IgG clusters.
      Keratin filament retraction from the cell membrane has been suggested as a potential cause of acantholysis in PV (
      • Bystryn J.C.
      • Grando S.A.
      A novel explanation for acantholysis in pemphigus vulgaris: the basal cell shrinkage hypothesis.
      ), but both in lesional mcPV skin and lesional mdPV mucosa signs of keratin retraction are absent. We did find keratin retraction in our lesional PF skin only in cells that had lost their desmosomes, suggesting that desmosome disappearance rather than keratin retraction is the cause of acantholysis.The blister in normal looking Nikolsky-positive PF skin is located in the spinous layer, whereas the lesional PF blister is found in the granular layer. Nikolsky-positive PF skin is loaded with IgG and vulnerable to mechanical pressure as blisters are easily evoked by rubbing. In such skin, desmosomes shrink in size and number in the basal and spinous layer but are less affected in the granular layer (
      • van der Wier G.
      • Pas H.H.
      • Kramer D.
      • et al.
      Smaller desmosomes are seen in the skin of pemphigus patients with anti-desmoglein 1 antibodies but not in patients with anti-desmoglein 3 antibodies.
      ). Thereby, the spinous layer may become the “locus minorus resistentiae” where some pressure as during punch biopsing may induce cell–cell separation (“scissor Nikolsky”). The cells surrounding the blister cavity still contained desmosomes by which they connected to other cells. On the site facing the blister cavity, however, half desmosomes are present that may be either split desmosomes or newly synthesized half desmosomes that seek an opposite partner (
      • Demlehner M.P.
      • Schafer S.
      • Grund C.
      • et al.
      Continual assembly of half-desmosomal structures in the absence of cell contacts and their frustrated endocytosis: a coordinated sisyphus cycle.
      ). Taken together, the spontaneous PF blister occurred in an area with complete absence of desmosomal adhesion, whereas the Nikolsky-induced blister occurs in an area of weakened desmosomal adhesion.
      In lesional mcPV skin and, to a lesser extent, in mdPV mucosa, keratinocytes still are connected to other cells with small desmosomes but contain half desmosomes at the sites where they face the blister cavity, as observed previously (
      • Diercks G.F.
      • Pas H.H.
      • Jonkman M.F.
      The ultrastructure of acantholysis in pemphigus vulgaris.
      ;
      • Wang W.
      • Amagai M.
      • Ishiko A.
      Desmosome splitting is a primary ultrastructural change in the acantholysis of pemphigus.
      ). In mcPV skin also some “torn-off” or pulled out desmosomes are present along numerous floating elements that look little membrane enclosed elongated objects. Given the 70 nm thickness of our sections, these are likely caused by the cutting angle and are not individual objects but cell protrusions pulled out of the cell body. The pulled out desmosomes therefore probably are still functional connections between two cells. To assess whether desmosomes are really torn-off or only pulled out between cells will require further three-dimensional-EM analysis.
      As for the underlying mechanism of acantholysis, our lesional PF biopsy shows that keratinocytes in the basal and spinous layer contain less and smaller desmosomes, whereas in the granular layer all desmosomes are lost. This best fits the non-assembly depletion hypothesis as does the reduction in size and number of the desmosomes in the other biopsies. However, seen the rudimentary half desmosomes at the cell membranes facing the blister cavity, we cannot exclude the possibility that more than one mechanism is involved in the pathogenesis of pemphigus.
      The connection between blister localization and desmosome structure through layers, thus ranging from the ~mm to the ~nm scale within a single data set, has been made possible by nanotomy. Given the unbiased approach and raw data sharing, nanotomy may become the standard for EM. It allows easy analyzing vast ultrastructural images simply from behind a desktop. Illustrative is the ease by which we quantified the unique double-membranes structures in our data sets. The presence of these double-membrane structures in pemphigus tissue had remained unnoticed until 1996 and thereafter was ignored for a long time in pemphigus research as their significance was unclear (
      • Tada J.
      • Hashimoto K.
      Curvicircular intracytoplasmic membranous structures in keratinocytes of pemphigus foliaceus.
      ). Here, we find how these structures are especially abundant in all lesional pemphigus tissue, which suggests that they are an important clue to the pathogenesis of pemphigus. By making our data sets open source highly informative data become available for an interdisciplinary research audience, ranging from medical doctors to cell biologists. Nanotomy therefore will greatly accelerate research on understanding pathomechanisms of human disease, as anyone can join the exploration of diseased tissue.

      Materials and Methods

      Human tissue

      The skin or buccal mucosa biopsies from patients or controls were from the University Medical Centre Groningen (Table 1). For this study, institutional approval and written informed patient consents were obtained. The control mucosa was from redundant material in our EM archive. Pemphigus was diagnosed on the basis of clinical appearance and confirmed by histology, positive immunofluorescence, and the ELISA for anti-Dsg1 and/or anti-Dsg3 antibodies. We also included a biopsy of a patient with the SSSS.

      Sample preparation for EM

      Two mm skin or oral buccal mucosa biopsies were fixed in 2% glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.4 washed in 0.1 M sodium cacodylate buffer and postfixed in 1% osmiumtetroxide and 1.5% potassiumferrocyanide. Samples were dehydrated, embedded in epon, and sectioned. Ultrathin sections (70 nm) were positioned on single slot holders A600 Nickel supported by Formvar and contrasted with 2% uranylacetate (methanol) followed by Reynolds lead citrate. Acquisition was performed using a Zeiss supra 55 EM (Oberkochen, Germany) with ATLAS software developed by Fibics (Ottawa, Ontario, Canada).

      Data acquisition, stitching, and annotations

      Samples were recorded at 2.5 nm pixel size. Scans were stitched, and raw data sets were rendered as HTML files using ATLAS VE viewer software. For illustrations, the data sets were exported with ATLAS VE viewer as TIFF files downscaled to 10 nm per pixel. Pseudocoloring of epithelial layers, blister cavity, widening, and double-membrane structures was performed using Adobe Photoshop CS5.1 (San Jose, CA, USA). In all figures, data sets are orientated such that the epidermis/superficial layers are facing up. Note that, because of acquisition reasons, HTML files are orientated differently. The size of all half desmosomes on plasma membranes of cells that face and make the walls of blister cavities was measured from the data sets online.
      We thank Eugene Berezikov (UMCG) for providing help with the server and Gerda van der Wier (UMCG) who supported us with patient material. For these studies were provided by the Netherlands Organization for Scientific Research (ZonMW91111006), Jan Kornelis de Cock Funding (DeCock2013-55), and Erasmus Mundus action two program—JoinEUsee.

      REFERENCES

        • Aalfs A.S.
        • Oktarina D.A.
        • Diercks G.F.
        • et al.
        Staphylococcal scalded skin syndrome: loss of desmoglein 1 in patient skin.
        Eur J Dermatol. 2010; 20: 451-456
        • Amagai M.
        • Karpati S.
        • Prussick R.
        • et al.
        Autoantibodies against the amino-terminal cadherin-like binding domain of pemphigus vulgaris antigen are pathogenic.
        J Clin Invest. 1992; 90: 919-926
        • Amagai M.
        • Stanley J.R.
        Desmoglein as a target in skin disease and beyond.
        J Invest Dermatol. 2006; 132: 776-784
        • Aoyama Y.
        • Nagai M.
        • Kitajima Y.
        Binding of pemphigus vulgaris IgG to antigens in desmosome core domains excludes immune complexes rather than directly splitting desmosomes.
        Br J Dermatol. 2010; 162: 1049-1055
        • Bergman R.
        • Hershkovitz D.
        • Fuchs D.
        • et al.
        Disadhesion of epidermal keratinocytes: a histologic clue to palmoplantar keratodermas caused by DSG1 mutations.
        J Am Acad Dermatol. 2010; 62: 107-113
        • Bystryn J.C.
        • Grando S.A.
        A novel explanation for acantholysis in pemphigus vulgaris: the basal cell shrinkage hypothesis.
        J Am Acad Dermatol. 2006; 54: 513-516
        • Delva E.
        • Tucker D.K.
        • Kowalczyk A.P.
        The desmosome.
        Cold Spring Harb Perspect Biol. 2009; 1: a002543
        • Demlehner M.P.
        • Schafer S.
        • Grund C.
        • et al.
        Continual assembly of half-desmosomal structures in the absence of cell contacts and their frustrated endocytosis: a coordinated sisyphus cycle.
        J Cell Biol. 1995; 131: 745-760
        • Diercks G.F.
        • Pas H.H.
        • Jonkman M.F.
        The ultrastructure of acantholysis in pemphigus vulgaris.
        Br J Dermatol. 2009; 160: 460-461
        • Frusic-Zlotkin M.
        • Raichenberg D.
        • Wang X.
        Apoptotic mechanism in pemphigus autoimmunoglobulins-induced acantholysis–possible involvement of the EGF receptor.
        Autoimmunity. 2006; 39: 563-575
        • Garrod D.
        • Chidgey M.
        Desmosome structure, composition and function.
        Biochim Biophys Acta. 2008; 1778: 572-587
        • Guedes A.C.
        • Rotta O.
        • Leite H.V.
        • Leite V.H.
        Ultrastructural aspects of mucosas in endemic pemphigus foliaceus.
        Arch Dermatol. 2002; 138: 949-954
        • Hanakawa Y.
        • Schechter N.M.
        • Lin C.
        • et al.
        Molecular mechanisms of blister formation in bullous impetigo and staphylococcal scalded skin syndrome.
        J Clin Invest. 2002; 110: 53-60
        • Janse I.C.
        • van der Wier G.
        • Jonkman M.F.
        • et al.
        No evidence of apoptotic cells in pemphigus acantholysis.
        J Invest Dermatol. 2014; 134: 2039-2041
        • Kowalczyk A.P.
        • Green K.J.
        Structure, function and regulation of desmosomes.
        Prog Mol Biol Transl Sci. 2013; 116: 95-118
        • Kuwajima M.
        • Mendenhall J.M.
        • Lindsey L.F.
        • et al.
        Automated transmission-mode scanning electron microscopy (tSEM) for large volume analysis at nanoscale resolution.
        PLos One. 2013; 8: e59573
        • Mahoney M.G.
        • Hu Y.
        • Brennan D.
        • et al.
        Delineation of diversified desmoglein distribution in stratified squamous epithelia: Implications in diseases.
        Exp Dermatol. 2006; 15: 101-109
        • Mahoney M.G.
        • Wang Z.
        • Rothenberger K.
        • et al.
        Explanations for the clinical and microscopic localization of lesions in pemphigus foliaceus and vulgaris.
        J Clin Invest. 1999; 103: 461-468
        • Müller E.J.
        • Williamson L.
        • Kolly C.
        • et al.
        Outside-in signaling through integrins and cadherins: a central mechanism to control epidermal growth and differentiation?.
        J Invest Dermatol. 2008; 128: 501-516
        • Oktarina D.A.
        • van der Wier G.
        • Diercks G.F.
        • et al.
        IgG-induced clustering of desmogleins 1 and 3 in skin of patients with pemphigus fits with the desmoglein nonassembly depletion hypothesis.
        Br J Dermatol. 2011; 165: 552-562
        • Ravelli R.B.
        • Kalicharan R.D.
        • Avramut M.D.
        • et al.
        Destruction of tissue, cells and organelles in type 1 diabetic rats presented at macromolecular resolution.
        Sci Rep. 2013; 3: 1804
        • Simpson C.L.
        • Patel D.M.
        • Green K.J.
        Deconstructing the skin: Cytoarchitectural determinants of epidermal morphogenesis.
        Nat Rev Mol Cell Biol. 2011; 12: 565-580
        • Stanley J.R.
        • Amagai M.
        Pemphigus, bullous impetigo, and the staphylococcal scalded-skin syndrome.
        N Engl J Med. 2006; 355: 1800-1810
        • Tada J.
        • Hashimoto K.
        Curvicircular intracytoplasmic membranous structures in keratinocytes of pemphigus foliaceus.
        J Cutan Pathol. 1996; 23: 511-517
        • van der Wier G.
        • Jonkman M.F.
        • Pas H.H.
        • et al.
        Ultrastructure of acantholysis in pemphigus foliaceus re-examined from the current perspective.
        Br J Dermatol. 2012; 167: 1265-1271
        • van der Wier G.
        • Pas H.H.
        • Kramer D.
        • et al.
        Smaller desmosomes are seen in the skin of pemphigus patients with anti-desmoglein 1 antibodies but not in patients with anti-desmoglein 3 antibodies.
        J Invest Dermatol. 2014; 134: 2287-2290
        • Wang W.
        • Amagai M.
        • Ishiko A.
        Desmosome splitting is a primary ultrastructural change in the acantholysis of pemphigus.
        J Dermatol Sci. 2009; 54: 59-61