A growing body of evidence indicates that the division between TCM and Teff/Tem can be complex and fluid. For example, some human cells with a TCM phenotype express not only LN homing receptors but also traffic molecules for peripheral tissues, and such cells have been detected in extralymphoid organs (Campbell et al. 2001). In addition, at least in certain mouse models, some TCM-like cells exhibit immediate effector activity comparable to that of Teff (Debes et al. 2002; Unsoeld et al. 2002; Wherry et al. 2003). Given these observations, there are no universally accepted criteria for unambiguous distinction between these memory T cell subsets, other than their differential capacity to migrate to PLN.
The molecular mechanisms driving the homing of effector/memory lymphocytes to the small intestine and the skin have been partially elucidated. Although gut DC and RA seem to be necessary and sufficient to imprint T cell with small bowel tropism under steady-state conditions, it will be interesting to explore whether there are additional mechanisms to generate gut-homing lymphocytes under inflammatory conditions. Moreover, T cells seem to require different signals to migrate into the large bowel (Kunkel and Butcher 2002; Mora et al. 2003), and it will be important to determine the specific multistep cascade for colon homing.
Is there truly "homing imprinting" for tissues other than the gut? Recent evidence suggests that the induction of skin homing (at least selectin ligands and CCR4) is a default pathway in the absence of gut-derived signals (Mora et al. 2005). It will be interesting to determine whether other skin-associated receptors, such as CCR10 and CCR8, require skin-specific molecular signals. In addition, it will be important to explore other proposed tissue-specific lymphocyte imprinting pathways, such as those for the lung (Ainslie et al. 2002; Xu et al. 2003), liver (Geissmann et al. 2005; Sato et al. 2005), bone marrow (Mazo et al. 2005), and central nervous system (Calzascia et al. 2005).
Similarly, although ASC can exhibit gut tropism, which depends on the site of antigen encounter (Kunkel and Butcher 2003), it is presently unknown whether and how B cells get imprinted. Are tissue-specific DC and/or RA involved in this process (Fig. 3B)? What mechanisms coordinate the expression of a4p7 and CCR10 and/or CCR9 in ASC? How is B cell imprinting correlated with switching to IgA versus IgG production?
Most studies on lymphocyte migration to the gut have focused on steady-state, noninflamed conditions. However, recent evidence suggests that other adhesion molecules, different from the canonical gut-specific homing receptors, may be invoked under inflammatory conditions (Rivera-Nieves et al. 2005). To devise optimal therapeutic strategies, it will be important to dissect the relative contribution of tissue-specific versus non-tissue-specific traffic molecules in different inflammatory settings.
Because gut-associated DC play an active role in programming T cells to express gut-homing molecules (Iwata et al. 2004; Mora and von Andrian 2004), it will be important to understand how DC themselves are "educated" to acquire this tissue-specific imprinting potential. It is likely that "DC education" happens in the periphery after bone marrow-derived DC precursors have established a tissue residence, but the alternative, a priori specialization of precursors in the bone marrow, has not been ruled out. Wherever DC imprinting specialization may have its origin, the signals involved in this process are unknown. Possible candidates include the gut microflora, signals through toll-like receptors, the influence of epithelial and/or other stromal cells, intrinsic differentiation programs triggered during hematopoiesis, or a mixture of these diverse factors. The elucidation of this mechanism(s) may have implications for the design or improvement of DC-based vaccines targeting mucosal immune responses, such as in HIV and cancer.
Acknowledgements Supported by NIH Grants HL-62524, HL-54936, HL-56949, and AI-061663.
Was this article helpful?