Article Text

Download PDFPDF

A ticket to the gut for thymic T cells
  1. W Falk
  1. Correspondence to:
    Dr W Falk
    Department of Internal Medicine I, University of Regensburg, D-93042 Regensburg, Germany; werner.falk{at}

Statistics from

The intestinal epithelium is colonised by, among other cells, normal T helper and cytotoxic T cells which obtain their antigen experience in the gut associated lymphatics after migration from the thymus. New evidence suggests a CD8 positive T cell that directly migrates to the intestinal epithelium increasing the role of the thymus in shaping intestinal immunity

A normal splenic T cell would be abhorred should it get close to the intestinal epithelium. Instead of its usual rather sterile environment, it would be bathed in an ocean of usually non-pathogenic microbial and food derived antigens never seen during its passage through the educating thymus. It would get ready to become very excited. However, although the number of T cells in the mouse small intestine is approximately 50% of those residing in all lymphoid organs,1,2 there is peace. This peace is the result of a balanced presence of activated protecting T cells and activated controlling T cells.

The gut represents a unique dilemma for the immune system. It has the largest surface of the body harbouring numerous species of microorganisms. It is thus challenged with the task to be a barrier to the outside world and also to be the organ for most effective absorptive and secretory functions. To be able to accomplish this task, mammals have evolved a special highly regulated protective immune system comprising T cells which are on constant alert, as is evident from their activated phenotype. Broadly, the gut harbours three categories of T cells. Those in the gut associated lymphoid tissues are very similar to T cells found in peripheral lymphoid organs. Another population of T cells is scattered between other different cell types in the lamina propria. Their education follows the standard path of meeting their antigens presented by antigen presenting cells in the mesenterial lymph nodes and trafficking back to the lamina propria.3

The third category is a very peculiar one, intraepithelial lymphocytes (IEL). They sit on the luminal side of the basement membrane, singly scattered between and protected by intestinal epithelial cells, and are presumably the first T cells to encounter antigens. IEL comprise T cells carrying the αβ T cell receptor (TCR) or the γδTCR, with their ratio depending on the species. Further subdivision may be made between conventional T cells characterised by a αβTCR and coexpression of CD4 or CD8αβ and non-conventional T cells which express either αβTCRs or γδTCRs but no CD4 or CD8αβ (double negative (DN)).4 Such cells are residents of the intestinal epithelium. A proportion of these cells express the CD8αα homodimer which, however, can also be detected on double positive conventional IEL together with CD4 or CD8αβ. Non-conventional T cells are rarely found in tissues other than the intestinal epithelial layer and they are also characterised by a surprisingly high degree of autoreactivity. It was reported, using transgenic models, that CD8αα αβTCR IEL can be found in animals that efficiently delete their CD8αβ cells in the thymus by negative selection.2 This and other findings have led to the search for extrathymic sites allowing differentiation into gut specific T cells. In mice, small aggregations of haematopoietic cells were found in the lamina propria close to crypts which were named cryptopatches (CP).5 In elegant studies it was shown however that CP are most likely not the sites for extrathymic sources of IEL and that most, if not all, gut IEL originate from the thymus. Some recombinase activating gene activity was found in the absence of a functional thymus6,7 indicating that, under special conditions, extrathymic maturation of T cells can still occur in the mesenterial lymph nodes or Peyer’s patches which is heavily skewed to generation of γδTCR T cells.

These γδ TCR T cells do not recognise peptide antigens and leave the thymus at a DN stage to later colonise predominantly epithelial surfaces. They can be rapidly activated by some non-peptide antigens and self stress antigens, such as the human MICA and MICB molecules.8,9 Their capacity to secrete growth factors like KGF has also led to the assumption that they might be involved in repair of damaged tissue.10 Another fascinating feature was recently shown. Human Vδ2 expressing cells through activation were able to present peptide antigens to naïve αβ TCR T cells, thus bridging innate and acquired immunity.11 Whether a similar function is also true for IEL γδ TCR T cells remains to be shown.

Non-conventional αβTCR CD8αα T cells are likely to be a consequence of “agonist” selection by self antigens in the thymus (reviewed and discussed by Cheroutre12). It is proposed that these T cells interact very strongly with self antigens at a DN stage and so acquire the activated phenotype. These cells can then express the mucosal integrin αEβ7 and in this way be guided to the intestinal epithelium where the binding partner E-cadherin is abundantly expressed. Such cells do not damage the epithelium13 but, in contrast, have a protective capacity, as was shown in the CD45RBhigh transfer model of colitis.14

This leaves us with the conventional αβTCR T cells which normally undergo positive and negative selection due to intermediate strength interactions with self antigens presented by self MHC molecules in the thymus. They leave the thymus as either CD4 or CD8αβ mature T cells to follow the well trodden path of being activated by professional antigen presenting cells to become effector and memory T cells which then settle in defined locations. There is evidence that the more chancy moment of antigen encounter in the intestinal lymphoid structures imprints the cells to specifically home to the intestinal epithelium.15,16 This is the current thinking, but in this issue of Gut, Annunziato and colleagues17 challenge this view (see page 961). The thymus seems more sophisticated than just pouring out mostly conventional single positive T cells. The authors found a subset of mature thymocytes which in many respects closely resemble a subpopulation of conventional IEL, thus suggesting that the thymus produces these cells specifically targeted to the intestinal epithelium. Although their specific local function is still unknown, these cells may not be so “conventional”.

A hallmark of all IEL is their constitutively activated phenotype. Accordingly, the thymus cell population identified by Annunziato and colleagues17 produces perforin and granzyme A, qualifying them as potential cytotoxic cells. They also produce cytokines such as interleukin 2 and interferon γ on polyclonal stimulation which support inflammation, and they express activation markers such as CD69 and CD45RO. All of this makes them look different from naïve mature T cells that await activation in peripheral lymphoid structures.

The current literature supports the notion that the gut TCR repertoire of conventional T cells is shaped by repeated encounter with gut specific antigens in the gut associated lymphoid structures, including Peyer’s patches and mesenterial lymph nodes.3 This progressive selection leads to the typical oligoclonality normally found for gut associated T cells. The Vβ repertoire of this new subpopulation of T cells did not reveal any enrichment, ruling out the standard activation pathway. Support for this notion stems from the observation that these cells still have comparatively long telomers. Repeated stimulation results in frequent cell divisions which reduces the length of telomers.18 One would expect long telomers in a cell just leaving the thymus but not in antigen experienced cells, as usually found in the gut. Yet this population of conventional gut T cells shows long telomers, supporting the conclusion that they just recently colonised the epithelium without having undergone multiple divisions.

How are the cells lured to the intestinal epithelial layer? The thymus microenvironment seems to provide a ticket and a guide. T cells are characterised by expression of the chemokine receptor CXCR3 and the presence of CD103, representing the αE of the αEβ7 integrin. This is a sticky ticket because epithelial cells display E-cadherin and this interaction was shown to be essential for IEL localisation.19,20 Attraction is also provided by epithelial cells which can produce large amounts of two of the three ligands for CXCR3, CXCL9, and CXCL10.21 Support for the functional importance of CXCR3 is given by the authors by the use of CXCR3 deficient mice which show clear underrepresentation of CD8αβ T cells only in the gut, compared with wild-type mice, which was normalised within 24 hours after adoptive transfer of wild-type cells. In addition, it was shown recently that the chemokine receptor CCR7 is needed for T cell exit from peripheral tissue to enter afferent lymph.22 These cells specifically lack CCR7 so that they may not be trapped in lymphatic tissue. However, this equipment also means that cells are meant to stay in the intestinal epithelium.

There is no information in the report about a specific function of these interesting cells or about the signals in the thymus that are required to generate them. Provided the surface markers do not change, they will not be able to meet antigens in the lymphatics of the gut as the conventional cells are supposed to do. But there are other options. Firstly, their nearest neighbours, the intestinal epithelial cells, can perform efficient antigen presentation and produce cytokines and chemokines.23,24 Also, some dendritic cells, which can send their protrusions between epithelial cells to sample the intestinal lumen, are likely candidates.25 Due to their polyclonal TCR, this subpopulation of T cells might be the firstline of defence to quickly combat antigens not normally present in the gut. Unveiling the role these cells play in intestinal homeostasis and disorders should be the next step.

The intestinal epithelium is colonised by, among other cells, normal T helper and cytotoxic T cells which obtain their antigen experience in the gut associated lymphatics after migration from the thymus. New evidence suggests a CD8 positive T cell that directly migrates to the intestinal epithelium increasing the role of the thymus in shaping intestinal immunity


View Abstract


  • Conflict of interest: None declared.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles