Statistics from Altmetric.com
The αEβ7-integrin (CD103) is expressed almost uniquely by T cells of the mucosal immune system, where it is upregulated on activated cells by the action of transforming growth factor β. The only known ligand for this integrin is E-cadherin,1 which is expressed by all epithelial cells, where it constitutes a homotypic adhesion system necessary for tight junction formation. A role for interaction between the αEβ7-integrin and E-cadherin in the localisation of intraepithelial T cells is supported by the reduction in numbers of mucosal T cells seen in CD103 deficient mice.2
The role of CD103+ T cells remains unclear. However, the potential of these cells to bind specifically to the epithelium is consistent with a capacity to mediate damage localised to this tissue. Indeed, CD8+CD103+ T cells have been shown to kill epithelial targets in vitro,3 and have been implicated in disease processes such as tubular destruction during renal allograft rejection.4 The potential for modulation of experimental colitis by the administration of antibodies directed at CD103 provides evidence that these cells might also act as effectors during this disease.5 Given the increasing severity of ulcerative colitis from the proximal to distal colon, it is perhaps reasonable to propose the existence of a similar gradient in the number of potential T cell effectors within the epithelium of the normal colon.
In our study we performed a survey of the linear distribution of cells expressing the CD3, CD8, and CD103 phenotypic markers within the normal human colon. Pinch biopsies were collected from the ascending, transverse, descending, and sigmoid colon and the rectum of patients attending clinic for routine diagnosis. Frozen sections were analysed from eight patients who were considered normal after routine histological evaluation. Endogenous peroxidase was blocked and the sections were stained with appropriate monoclonal antibodies (CD3, clone T3-4B5; CD8, clone CD8/144B; and CD103, clone BerAct8). In each case an isotype matched control antibody was also applied to demonstrate the specificity of the staining process. The labelled sections were visualised using a streptavidin–biotin–peroxidase kit. After counterstaining with Mayer’s haematoxylin, the number of CD3, CD8, and CD103 positive cells was counted in each crypt cross section, and the mean number of each cell type in each crypt was calculated. Image analysis was used to demonstrate that the crypt cross sectional area did not vary between different sites within the colon.
Figure 1 shows the typical distribution of CD103+ T cells within the normal colon; it is apparent that many, but not all, of these cells are present within the epithelium. Figure 2 presents a summary of the numerical data derived from each of the eight normal patients. In the case of each phenotypic marker, the data are reproducible between individuals and show a significant decrease in the number of cells in each crypt from ascending colon to the rectum (CD3, p < 0.005; CD8, p < 0.02; CD103, p < 0.03).
Our data show clearly and for the first time a linear decrease from the normal ascending colon to the rectum in the number of cells expressing the CD3, CD8, and CD103 phenotypic markers. Although it appears paradoxical that this gradient runs contrary to that which may be expected if CD103+ T cells are, indeed, the effectors responsible for tissue damage in ulcerative colitis, it is tempting to speculate that this gradient of T cell distribution has some impact on the potential for immune reactivity within the gut.
This work was supported by the National Association for Colitis and Crohn’s Disease.