Nature Immunology 2, 757 - 758 (2001)
© Nature America, Inc.
Autoimmunity Correspondence
Brigitta Stockinger, Thomas Barthlott
& George Kassiotis
Division of Molecular Immunology,
The National Institute for Medical Research Mill Hill, London NW7 1AA,
UK. (bstocki@nimr.mrc.ac.uk)
Despite the burial of the archetypal suppressor T cell concept prevalent in the 1970s and 1980s (1), the theory of active immune down-regulation as the mechanism by which peripheral tolerance is maintained has strong support (2). The appeal of immune regulation may be that, in contrast to central tolerance induction by deletion of T cells in the thymus, immune regulation offers the chance for therapeutic intervention. Given the importance of maintaining tolerance to self, redundancy in the underlying mechanisms should be expected, although recently a functionally specialized subpopulation of regulatory CD4+ T cells (TR cells), whose existence has been proposed on the basis of a variety of experimental systems, has gained in prominence (3).
Although it may be tempting to propose that a unique subpopulation of T cells is responsible for all of peripheral tolerance, it is dangerous to extrapolate from the highly divergent experimental systems used (3). Most of the characteristics and proposed mode of action of TR cells were defined in experimental systems that depended on a state of lymphopenia to observe autoimmunity in the first place. We now put forward an alternative view of T cell regulation based on the importance of homeostatic control mechanisms that operate in a normal immune system in vivo; this view refers specifically to experimental systems that use lymphopenic mice.
The development of autoimmune disease
is frequently associated with either congenital (4) or experimentally (5-7)
induced lymphopenia. Abnormal T cell homeostasis was initially considered
to be a possible cause for the association between lymphopenia and autoimmunity
(8). However, the finding that reconstitution of lymphopenic mice with
certain T cell subsets prevented autoimmunity (9, 10) prompted the conclusion
that a population of suppressor or regulatory T cells is dedicated to the
suppression of potentially autoreactive T cells (11). Cytokines such as
TGF-a or IL-10 have
been invoked and cell surface molecules such as CD25 and CTLA-4 were assigned
as being characteristic of TR cells. Details of the phenotype and proposed
mode of action of CD25+CD4+ TR cells can be found in the Review by Powrie
in this issue of Nature Immunology (3).
A healthy functional immune system
depends on the homeostatic maintenance of a diverse repertoire of naïve
and activated (or memory) cells that exist in distinct niches, compete
for survival with recent thymic emigrants and compete for space with those
clones that have an activated phenotype (12, 13). The term "space" is currently
not well defined, but access to MHC proteins, as well as cytokines and
chemokines that promote survival, are important features (14, 15). Most
of the T cells found after transfers into lymphopenic hosts have the phenotype
of activated or memory T cells, which suggests that proliferation to fill
space is a hallmark of preactivated T cells. However, more recent experiments
with T cells from various TCR-transgenic strains indicate that different
degrees of proliferation are also evident in T cells that have a naïve
phenotype at the time of transfer.
Although interpretation of the expansion
potential of T cells from individual transgenic strains in isolation would
give a skewed picture, their collective behavior reflects the situation
in the polyclonal repertoire. Thus, some T cells do not expand at all when
transferred into lymphopenic hosts. These are represented by H-Y TCR–transgenic
CD8+ T cells or by OT-II TCR–transgenic CD4+ T cells (16). Others reveal
extensive proliferation, and even differentiation into effector cells,
in the absence of nominal antigen (such as CD8+ T cells from the OT-1,
P14 or 2C strains (17-19)). CD4+ T cells from the A1 or A18 TCR–transgenic
strains undergo only few divisions upon transfer, whereas CD4+ T cells
from the AND strain proliferate extensively (20).
All T cell clones with extensive
proliferative capacity tend to up-regulate activation markers such as CD44,
although some studies suggested that their phenotype was not identical
to that of bona fide effector cells (17, 21). Thus, the polyclonal repertoire
contains a subset of T cells that have enormous expansion potential, and,
given their conversion to cells with an activated phenotype, they probably
occupy the memory niche, competing for space with other activated or memory
cells. The factors determining the initial expansion potential are not
clearly defined, but responsiveness to IL-7 is definitely important because
transfer of sorted naïve polyclonal T cells into IL-7-/- mice abrogates
all proliferation15.
TR cells, as they are described,
express characteristic markers of activated T cells. Both CD25 and CTLA-4
are transiently expressed and such cells would be in direct competition
for space with activated and memory T cells. In addition, as CD25+CD4+
TR cells do not produce IL-2 and are absent in IL-2–deficient mice (22,
23), their expansion in normal mice would likely be compromised. The expansion
potential of the CD45RBlo (activated and memory cell) population in vitro
is limited. However, in vivo transfer of CD45RBlo cells alone causes a
17-fold expansion over the original input, compared with a 40-fold expansion
of the CD45RBhi population when injected alone (24).
Coinjection of allotypically distinct
subsets showed that the "regulatory" subset curtailed expansion of the
subset that was originally CD45RBhi, especially in the gut, but did not
prevent their activation, as indicated by its conversion to the CD45RBlo
phenotype (25). Thus, TR cells may be in charge of controlling the magnitude
of immune responses, irrespective of whether they are directed toward self
or foreign (for example, gut bacteria) antigens. This may simply be a side
effect of their own expansion in response to antigen and their need for
externally supplied IL-2 (26, 27).
The crucial difference between a
healthy "regulated" immune system and a destabilized system prone to autoimmune
disease may therefore mainly lie with the clone size that certain T cell
specificities achieve. Release of cytokines (such as IL-10 and TGF-) alone
may well be contributory, but, in principle, homeostatic competition for
space and cytokines in the activated and memory niche may be sufficient
to prevent the uncontrolled expansion of autoreactive T cells.
Thus, any T cell clone with an activated
phenotype and reactive to nonself antigens—provided they are not deposited
in sensitive peripheral organs—could prevent autoimmune disease. In addition,
it is conceivable that even certain naïve T cells—such as TCR-transgenic
AND, OT-1 or P14 T cells—would qualify as TR cells. This would presumably
involve competition for IL-7, which controls homeostatic proliferation
of naïve cells and, to some extent, of memory cells (15). Once expanded
or converted these cells would occupy the compartment for activated and
memory T cells, thus restricting the expansion of the autoreactive CD4+
T cell population.
The concerted effort of maintaining
a healthy balance of T cells in the periphery probably involves many antigenically
diverse, activated T cells as well as cytokines and chemokines. We propose
that there are at least two ways to arrive at immune regulation. One possibility
is that the action of cells in their effector phase (CD45RBloCD25+CTLA-4+),
irrespective of TCR specificity, might curtail the expansion potential
of any naïve cell with possible autoreactive function via competition
for IL-2 and access to antigen-presenting cells. Alternatively, IL-7–driven
homeostatic expansion of high-avidity naïve cells in a lymphopenic
environment, followed by at least partial activation, could out-compete
the expansion of other naïve T cells. As long as the competing naïve
population does not consist of autoreactive cells, they might provide protection
against pathological immune responses or autoimmunity.
It might be useful to study the regulatory
potential of a variety of T cells with defined antigenic specificity to
find out what the mechanisms underlying differential homeostatic competition
are. Although this might not lead us to a specialized subset of TR cells,
it might, nevertheless, help in defining protocols aimed at establishing
or maintaining peripheral T cell tolerance.
References
© 2001 Nature Publishing
Group