Previous
studies in mice identified a population of immature NK
cells with phenotypic and functional features typical of regulatory cells (Ebata et al., 2006). In addition, these cells
were abundant at the feto–maternal interface where they
may play a substantial role in the maintenance of pregnancy (Lin et al., 2009). On the other hand, studies in
humans revealed the presence of Tregs both in PB and
in decidua during the early phases of normal pregnancies
(Heikkinen et al., 2004; Saito et al., 2010). In some cases,
spontaneous abortions have been associated with low Treg
cell proportions in PB (Sasaki et al., 2004; Saito, 2010).
Remarkably, in tumors, it has been shown that Treg cells
generate a tolerant microenvironment as a result of the
interaction with other cell types, including myelomonocytic cells (Rabinovich et al., 2007). A similar mechanism
could take also place in decidual tissues during early pregnancy (Barrientos et al., 2009). In this context, however, it
was unclear how Tregs could be induced. A possible explanation was that Tregs could be induced by indoleamine
2,3-dioxygenase (IDO) and/or other suppressive mechanisms, possibly involving the particular CD14
+
myeloid
cells present in the decidua (Munn et al., 1998; Fallarino
et al., 2003; Terness et al., 2007).
Notably, previous studies in mice revealed that IDO
is expressed by myeloid cells in decidua and that IDO
inhibitors can cause miscarriages, suggesting a key role
in the maintenance of pregnancy and in the induction
of Tregs. Remarkably, dCD14
+
cells display phenotypic
characteristics typical of polarized M2 macrophages or
of immature dendritic cells (DCs) (Bulmer and Johnson,
1984; Blois et al., 2007; Laskarin et al., 2008). Histochemical analysis revealed that dNK cells can reside in
close association with dCD14
+
cells, a condition that could
result in functional cross-talk between the two cell types
(Kammerer et al., 2003). Indeed, we showed that the
interaction between dNK cells and dCD14
+
cells promotes
the induction of Tregs characterized by the CD25
bright
and FOXP3
+
phenotype. Upon interaction with dCD14
+
cells, dNK cells release IFN. In turn, IFN induces the
expression of IDO in dCD14
+
cells. The IDO metabolite
l-kynurenine, does not affect IFN production by dNK
cells (thus differing from PB NK cells). Other mechanisms are involved in Treg induction by dCD14
+
cells,
including the release of TGF and the expression of CTLA-4, which interacts with its ligand expressed by T cells
(Vacca et al., 2010). Thus, the cross-talk between decidual NK and CD14
+
cells may initiate a process leading
to Treg induction and immunosuppression. Remarkably,
previous studies reported a marked reduction of dNK
cells in cases of spontaneous abortion during the first
trimester. Taken together, these studies support the notion
that, in decidual tissues, NK cells might play an important role in modulating maternal immune responses and
in tolerance induction (Karimi and Arck, 2010). Moreover, they provide new insight into the immunoregulatory
events occurring at the maternal–fetal interface resulting
from cellular interactions among dCD14
+
, dNK, and Treg
cells.