LTi (lymphoid tissue inducer cells) are cells that appear in primitive lymphoid organs (lymphoid analagen) that produce lymphotoxin (LTα1β1) and tumor necrosis factor (TNF) and thus stimulate the mesenchymal cell production of chemokines and adhesion molecules essential for lymphoid organogenesis (Mebius, 2003
). These cells are innate cells in that they do not develop in the thymus and lack antigen-specific receptors; instead they respond to cytokines induced by innate stimuli or, as discussed below, via NK cell-type receptor ligands. LTi cells express retinoic acid receptor-related orphan receptor (RORγt or RORC) transcription factor as well as interleukin (IL)-7Rα (CD127), and, indeed, these components are necessary for the generation of LTi cells as well as the lymphoid structures they induce, such as the Peyer's patches (Eberl et al., 2004
). Although LTi cells are readily found in fetal tissues, they are relatively rare in adult tissues, perhaps reflecting the reduced need of adult tissue to form new lymphoid structures. However, in the intestine where the latter need may be higher, one finds larger numbers of LTi, particularly associated with intestinal cryptopatches and isolated lymphoid follicles (Vivier et al., 2009
; Colonna, 2009
). Recently, another subset of RORγt-expressing LTi cells that bear NK cell receptors (such as NKp46) have been identified (Colonna, 2009
). These cells differ from conventional NK cells in that they express little or no NK1.1 or perforin. In addition, these cells do not produce IFN-γ upon stimulation with interleukin-12 (IL-12) but do produce IL-22 in response to IL-23. The advent of this subset has prompted speculation about the relation of LTi cells to NK cells. In one theory, these cells arise from a common precursor, which then gives rise to either NK cells or NKp46+
LTi cells; in another theory, these cells have separate lineages (Vivier et al., 2009
). In this issue of Immunity
, Vonarbourg et al. (2010)
define the relation of NK cells to NK receptor+
LTi that do and do not express RORγt; in doing so, they more clearly define the role of RORγt in LTi function.
In their initial studies, Vonarbourg et al. (2010)
show that in Rag2–/–Il2rg–/–
recipient mice (lacking all lymphocytes) repleted with purified NK cells or GFP-marked NKp46-
LTi cells (which express sLTα1β1 and low perforin), the transferred NK cells remain RORγt–
, whereas most tranferred LTi cells upregulate NKp46. In the small intestine of the recipient mouse, RORγt expression in the transferred RORγt+
cells is maintained, whereas in the colon some cells lose RORγt. An essentially similar result was obtained with transfer of peripheral lymph node cells. The results of these transfer studies conducted in lymphopenic mice were then confirmed in intact mice by studies of fate-mapped mice in which cells that have expressed RORγt at some point in their life are engineered to permanently express YFP. Here, it was found that all NKp46+
cells were RORγt fate-mapped cells and most small intestinal RORγt fate-mapped cells express RORγt whereas only 25% of colon or spleen cells do so. These transfer and fate mapping studies thus settle the question of the origin of NKp46+
LTi (NKR-LTi cells): these cells arise from NKp46-
LTi precursors and then lose RORγt expression in certain lymphoid environments (). Finally, it should be noted that loss of RORγt associated with lower expression of CCR6, CD127, and sLTα1β1 and increased expession of perforin and granzyme B. In other words such loss causes diminished LTi function and increased cytotoxic (NK-like) function.
Changes in Intenstinal LTi Populations during Maturation
In further studies, Vonarbourg et.al. (2010)
show that the loss of RORγt expression in the NKR-LTi population is determined by environmental factors. In particular, they demonstrate that the gut microbiota support persistance of RORγt expression and germ-free mice lack RORγt+
cells. Such microbial influence is probably acting through the induction of cytokines. Thus, whereas IL-15 and IL-12 favor downregulation of RORγt, IL-7 prevents downregulation, and IL-7R deficient mice lack RORγt-LTi cells. In contrast,TSLP, a cytokine that shares a receptor chain with IL-7, has no effect on downregulation.
In a final but very important series of studies, Vonarbourg et.al. (2010)
address the effect of RORγt expression in NKR-LTi cells on surface marker expression and on cell function. Whereas RORγt+
NKR-LTi cells express high amounts of sLTα1β1 and IL-7R, RORγt–
NKR-LTi cells express perforin and granzyme. In addition, RORγt expression correlates with IL-23R expression and is highest in CD4+
NKR-LTi cells. The significance of this lies in the fact that IL-23R expression leads to IL-23-dependent NKR-LTi production of IL-22. Cells that express RORγt are poorly responsive to IL-12 but gain responsiveness to this cytokine as RORγt expression wanes; thus, RORγt acts as a reostat for the ability to respond to this cytokine. Overall, LTi cells that retain RORγt expression maintain their potential to act as inducer cells and through the production of IL-22 maintain a capacity to sustain epithelial cell integrity whereas those that lose RORγt expression tend to function as NK-like cells in that they produce IFN-γ and are capable of cytotoxic function. It should be noted, however, that CD4–
NKR-LTi cells that have lost RORγt expression during their sojourn in the colon retain IL-23R expression (as opposed to those in the peripheral lymph nodes) and these cells alone among LTi cells or NK cells produce IFN-γ in response to IL-23 stimulation (). These cells are potentially important in the light of a concluding series of studies showing that they are the subset of LTi cells that has previously been linked to the induction anti-CD40-induced colitis in Rag2–/–
mice (Buonocore et al., 2010
The relationship between RORγt expression and IL-22 production induced by IL-23 in LTi cells shown in these studies and previously appears to be different from that in Th17 cells. Recall that in the latter situation, RORγt expression is induced by the cytokines TGF-β and IL-6 and such induction leads to IL-23R expression followed by differentiation and/or expansion of Th17 cells by IL-23; thus, IL-23 is only indirectly involved in Th17 cell IL-17 and IL-22 cytokine induction (McGeachy et al., 2009
). In contrast, LTi cells already express RORγt yet do not produce IL-22 in the absence of IL-23, indicating that in this case IL-23 has an inductive role and that IL-23 signaling in LTi cells is different from that in Th17 cells. Along related lines, although NKp46–
LTi cells produce IL-17, NK46+
LTi do not (Cella, et al., 2009
). Because RORγt is so intimately associated with IL-17 production, it seems likely that the lack of IL-17 production by the latter cells is not due to an inductive defect but rather to suppression by factors in the intestinal tissues. Whatever mechanism is involved, this sudden loss of IL-17 producing capacity may be fortuitous in that it renders the NKR-LTR less proinflammatory.
These differences between Th17 and LTi cells are reflected in the different role of these cells in mucosal immune function. LTi cells have a distinct and unique role as tissue inducer cells, and although their function is most apparent early in life when lymphoid structures are initially forming, they also function in this manner during adult life in response to infection in peripheral tissue and most of the time in the GI tract, which is constantly responding to new gut antigens. In addition, there is now considerable evidence that LTi cell production of IL-22 is an important host defense element, particular in relation to infectious insults involving the intestinal epithelium, such as Citrobacter rodentium
infection (Vivier, et al., 2009
; Colonna, 2009
). This can be attributed to the fact that IL-22 been shown to preserve the integrity of the epithelium and to induce epithelial cell production of factors that are cytotoxic for potential bacterial invaders (Cella et al., 2009
). Th17 cells, on the other hand, are more clearly proinflammatory cells which, via the production of IL-17, cause the influx of other inflammatory cells as well as secondary proinflammatory cytokines, including those generated by the inflammasome. So it appears that although LTi cells are geared to prevent or avoid inflammation, Th17 cells are geared to cause inflammation.
The above discussion of the roles of LTi cells and Th17 cells (or indeed Th1 cells) brings us to the important question raised in these studies of whether LTi cells participate in causing inflammation in inflammatory bowel disease (Crohn's disease; CD). This possibility was posed initially by studies already alluded to showing that LTi cells producing IL-17 and IFN-γ are responsible for colitis induced in (lymphopenic) Rag2–/–
mice with Helicobacter hepaticus
infection or by LTi cells producing IFN-γ in Rag2–/–
mice with anti-CD40-induced colitis (Buonocore, et al., 2010
) . As mentioned above, in the studies by Vonarbourg et al. (2010)
, the LTi cells causing anti-CD40 colitis were identified as cells that have lost RORγt expression and that respond to IL-23 with IFN-γ production. However, in studies of patients with Crohn's disease addressing whether LTi cells are involved in disease pathogenesis, it was shown that cells bearing an NK marker (NKp44) and expressing RORγt and CD127 are actually decreased in the lamina propria, and this cell population is not a source of increased IFN-γ production. In contrast, cells bearing NK markers (including NKp46 and CD56) and lacking the LTi marker CD127 are increased in CD and do produce increased amounts of IFN-γ in this disease. These cells, however, appear to be NK cells not derived from LTi cells (Takayama et al., 2010
). One must therefore conclude that although various kinds of LTi cells can mediate colitis in a lymphopenic host, they do not do so in humans with inflammatory bowel disease.