Alterations in Fibroblast α1β1 Integrin Collagen Receptor Expression in Keloids and Hypertrophic Scars  Greg Szulgit  Journal of Investigative Dermatology  Volume 118, Issue 3, Pages 409-415 (March 2002) DOI: 10.1046/j.0022-202x.2001.01680.x Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 1 Removing epithelial and endothelial cells. (A) HaCaT cells are used as a positive control for α6 staining. The left histogram shows the 2° stain alone whereas the right histogram shows positive staining with the anti-α6 IgG. (B) Cells from a typical keloid, triple staining for α1, α2, α6. The α6-positive region is marked on the histogram and, whereas the α1-α2 dot plot shows the α6-positive events in red, the majority of the α6-positive cells (which are most likely to be endothelial) are a subset of the α1 low-intermediate, α2/α5 high population (population 2 in subsequent figures). Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 2 Immunofluorescent staining of tissues (see Table I). All samples were examined and photographed through a triple wavelength filter cube. Magnifications refer to objective compounded by ocular magnifications. (A) Normal skin sample. Note the autofluorescence of abundant dermal elastic fibers. Left panel: α1-A488, α2-PE, 20x. Faint α1 and α2 staining is seen in the papillary dermis. α2 is abundant in the basal epidermis. Right lower panel: α1-A488, 40x. α1 stains the medium of a deep dermal artery. Right upper panel: α5-PE, 20x. α5 is diffusely distributed throughout the reticular dermis. (B) Keloid, α1-PE, DAPI, 10x. α1 is abundant and distributed throughout the papillary and reticular dermis. (C) Keloid, α1-PE, procollagen-A488, 20x. Note the colocalization of these markers in a cellular region of the keloid. (D) Keloid, α5-PE, procollagen-A488, 20x. α5 and procollagen staining are similarly colocalized. (E) Keloid, α1-PE, smooth muscle actin FITC, 20x. This keloid was one of a minority with palisades of α1-positive, smooth-muscle-actin-positive myofibroblasts surrounding dermal vessels. The majority of specimens were smooth muscle actin negative consistent withEhrlich et al (1994). (F) Same specimen as (E), α1-PE, procollagen-A488, DAPI, 40x. At the center of the palisade cells express high amounts of α1 and collagen, and may be myofibroblasts, or possibly also endothelial cells as previously observed byPeltonen et al (1991). Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 3 Density plots for a typical keloid specimen stained singly or in combination, with the integrin antibodies or the isotype controls. “Iso” indicates isotype control antibody. Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 4 Integrin expression in normal skin and keloids. (A) A gallery of FACS profiles showing the variation between several normal and keloid specimens. First and second columns are α1 versus α2, and third and fourth columns are α1 versus α5. First and third columns are normal skin samples; second and fourth are keloids. Note the marked increase in population 1 (the α1 high α2/5 low population) in the keloids. A pattern similar to the keloids was seen in hypertrophic scars and radiation ulcers. (B) An example of a keloid specimen costained with anti-α2-FITC and anti-α5-PE. α2 and α5 appear to be coexpressed. This costain was performed on two other specimens with similar results. (C) Enlargement of a panel from (A) showing definitions of populations 1 (high α1, low α2/α5) and 2 (high α2/5, low α1). Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 5 Summary of high integrin α1 expression in different tissue types. Bars and errors indicate the mean and SEM (SEM is not included for radiation ulcers due to small sample size). (A) The proportion of highly α1 events (population 1) is expressed as a ratio of cells in population 1 to cells in populations 1 + 2, calculated from the α1 versus α2 dot plots of each specimen. Both keloids and hypertrophic scars have a significant increase in the size of population 1 relative to normals (p < 0.05) (n = 8 for normal skin, 11 for keloids, 5 for hypertrophic scars, and 2 for radiation ulcers). (B) Intensity of α1 expression in population 1. Fluorescence intensity is significantly higher for keloids and hypertrophic scars than for normal skin samples (p < 0.05) (some early specimens were not used in this calculation because they were processed at incomparable amplification settings; n = 3 for normal skin, 7 for keloids, 4 for hypertrophic scars, and 2 for radiation ulcers). Radiation ulcers not included in statistical tests for significance due to small sample size. Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions

Figure 6 FACS profiles showing the effects of culturing directly harvested fibroblasts on plastic. (A) Successive passages show the progressive changes of α1 and α5 in the populations of a keloid, eventually resulting in a monomorphic population. (B) Sample of normal skin and a keloid. Normals, keloids, and hypertrophic scars (not shown) look similar after passage four, suggesting a “default” integrin expression pattern that arises from culturing. Journal of Investigative Dermatology 2002 118, 409-415DOI: (10.1046/j.0022-202x.2001.01680.x) Copyright © 2002 The Society for Investigative Dermatology, Inc Terms and Conditions