T-cell costimulation in graftversus- host disease and graft-versus-leukemia effect - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 207

are now implicated in negative regulation. Likewise, those negative regula-
tors such as PD-1 and CTLA4 for Teffs are now considered a basis for Treg
function [138] . In this chapter, we focus on studies from many laborato-
ries, including ours, which demonstrated that although CD28 and CTLA4
are expressed on both regulatory and effector T cells, the B7/CD28/CTLA4
pathways play a central role in the biology of Tregs. Thus, a strategy that
targets costimulatory signals for Teff suppression should also be considered
for its effect on Tregs in the induction of transplant tolerance.

The first evidence indicating the role of CD28 in the homeostasis of Tregs was
observed in the autoimmune-prone NOD mouse [139] . Whereas mice defi-
cient for CD28 or B7 generally display defective immune responses, NOD
mice deficient for CD28 or B7 surprisingly develop diabetes at 8-10 weeks of
age. Exacerbation of disease was accompanied by a dramatic decrease in the
percentage of CD4 + CD25 + Tregs in those deficient NOD mice, and the infu-
sion of Tregs from WT mice could control diabetes in NOD-CD28KO mice.
These studies clearly indicate that CD28-B7/CD28 interaction is critical for
maintaining normal levels of Tregs in the periphery and consequently for the
prevention of autoimmune diseases. The critical role of B7/CD28 interac-
tions for Tregs has been observed in murine models of GVHD, in which CD28
costimulatory signals were shown to be indispensable for the generation of
Tregs that suppressed GVHD-induced allogeneic lymphocyte infusion [140] .

210

Impaired homeostasis of Tregs in CD28-deficient mice resulted from
defects in Treg development in thymus and Treg maintenance in the periph-
ery [141] . The potential effects of CD28 Treg homeostasis are likely to be
on survival, proliferation, and IL-2 production. Although CD28 mediates
Bcl-xL upregulation through PI3-kinase signaling that promotes survival
of Teffs [142] , the same activity seems not necessary for CD28-mediated
FOXP3 stabilization and Treg development [143] , suggesting that Bcl-xL is
not involved in the CD28 control of the Treg population. Tregs can vigor-
ously proliferate, particularly in vivo [144,145] , and blocking CD28 com-
pletely prevents the spontaneous proliferation of Tregs [141] . Furthermore,
CD28 also affects the tissue distribution of Tregs that influences tolerance
versus immunity [146] . A third major mechanism by which CD28 regulates
Treg homeostasis could be an indirect influence of IL-2 and IL-2 signal-
ing, given that IL-2 and its receptor are critically important for the main-
tenance of the Treg population in the periphery [147,148] . However, the
functions of CD28 and IL-2 in Treg homeostasis only partially overlap, and
an additional role of CD28 directly supporting Treg homeostasis remains to
be further defined. Application of CD28 blockade in the control of GVHD
has been aimed at suppressing Teffs, but more attention should be paid
to its effects on Tregs as well in allogeneic HCT. In fact, Beyersdorf et al.
[29] showed that engagement of CD28 by an agonistic conventional anti-
CD28 mAb protected GVHD by increasing the frequency of Tregs among
total CD4 + T cells. The same group also demonstrated that in vivo pretreat-
ment of donor mice or short-term in vitro culture of donor lymph node
cells with a superagonistic anti-CD28 mAb efficiently protected recipient
mice from acute GVHD [34] . The protection strongly relied on the presence
of activated Tregs in the donor T-cell inoculum. Because of the critical role
of CD28 on Tregs, caution must be exercised in applying the CD28 blockade
in combination with Treg therapy in GVHD.