The samples were mounted in embedding medium (Tissue-Tek OCT; Miles, Elkhart, IN), rapidly frozen in propane chilled with liquid nitrogen (−160°C), and stored at −80°C until use. Serial cross sections, 5 μm thick, were processed for immunohistochemistry with a previously characterized polyclonal antibody (Ab) specific for integrin α11 chain.
11 In the adult normal and diseased corneas, monoclonal Ab 1A4 against α-SMA (α-smooth muscle actin)
28 was used to detect myofibroblasts in especially scarred regions of the corneas.
29–31 In vitro α11β1 integrin mediates cell adhesion to collagens I and IV (preferably collagen I).
13 To investigate whether the staining pattern for integrin α11 chain correlates with the presence of collagens I and IV, a polyclonal Ab specific for collagen I (Ab D9-R349; Southern Biotechnology Associates, Inc., Birmingham, AL) and a monoclonal Ab (Ab CIV22; Dako, Glostrup, Denmark) specific for the α1-α2 chain-containing collagen IV trimer
18 were used. As type V collagen molecules are buried within the collagen type I fibrils and are considered to be a limiting factor for collagen fibril growth, a monoclonal Ab (Ab 36382) specific for collagen V (Abcam, Cambridge, UK) was also used to reveal possible correlation to the staining pattern of integrin α11 chain. To visualize breaks in the epithelial BM, the monoclonal Ab 4C7
32,33 against laminin α5 chain was used. The bound Abs were visualized by using a standard indirect fluorescence technique, with the secondary Ab conjugated with a fluorochrome (Cy3; Jackson ImmunoResearch Laboratories, West Grove, PA, or Alexa 488; Molecular Probes, Leiden, The Netherlands). Nuclei were detected with DAPI staining (Vectashield; Vector Laboratory, Burlingame, CA). Control sections from which the primary Ab was omitted were completely negative. Control sections prepared with normal rabbit serum instead of the primary polyclonal Ab revealed the level of unspecific staining and served as reference for the evaluation of the level of specific staining in the other sections. Although the staining level observed in corneal epithelium was higher than that of background, this is unspecific as epithelium is known to be “sticky.” Furthermore, similar staining of skin epithelium was observed but no corresponding labeling was seen with in situ hybridization (Popova SN, Gullberg D, unpublished data, 2008).