PerspectiveImmunology

Selection for Survival?

See allHide authors and affiliations

Science  27 Jun 1997:
Vol. 276, Issue 5321, pp. 2000-2001
DOI: 10.1126/science.276.5321.2000

An organism's immune system must satisfy two contradictory requirements. On the one hand, it has to respond to a broad range of antigenic challenges from foreign substances. On the other hand, it should react to particular antigens quickly and efficiently, especially those pathogens that it encounters repeatedly. The immune system solves this dilemma by creating two compartments, one to meet each of these needs. The primary lymphoid organs (thymus and bone marrow) produce millions of lymphocytes displaying a vast diversity of antigen-specific receptors that form the basis of the broad immune response; these cells are exported to the peripheral organs (blood, spleen, and lymph nodes) where they form a “naïve” pool capable of recognizing most antigens. Upon antigenic stimulation, the relevant lymphocytes become activated, proliferate, and eventually dispense with the antigen. The second compartment is formed when some of the antigen-stimulated lymphocytes become diverted to a “memory” pool, composed of cells that can respond to antigen more immediately than those in the naïve pool. The memory cells differ from the naïve cells in their surface marker profiles, homing properties, signaling requirements, and cytokine secretion pattern (1).

What factors control the maintenance of these two pools, both of which are constantly subject to dynamic influences? The importance of this question is underscored by recent findings that HIV patients show perturbations in the homeostasis of lymphocyte populations (2). The difficulty in addressing this question arises from the labyrinthine complexity of the immune system. Now, by exploiting mouse strains engineered to have simplified immune systems, Tanchot et al. (3, page 2057) have begun to define the factors that drive the dynamics of the naïve and memory lymphocyte populations.

Tanchot et al. focus on the ligands required by the naïve and memory CD8+ T cells for proliferation or survival in a mouse, clearly key control points for maintenance of the two pools. Classically, CD8+ T cells become activated and proliferate when stimulated by a specific peptide antigen bound to a particular major histocompatibility complex (MHC) class I molecule expressed at the cell surface. Even in the absence of stimulation by specific antigen, naïve and at least some memory CD8+ cells can survive for long periods in a normal mouse (4). Tanchot et al. used the strategy illustrated in the figure to define the precise ligand or ligands that allow the proliferation and long-term survival of these cells. To simplify the problem, they generated monoclonal CD8+ T cell populations that expressed only one type of T cell receptor (TCR); the receptor is specific for the antigen HY (found only in males) presented by an MHC molecule called H-2Db. The authors were sure these cells were all identical because they were derived from a transgenic mouse strain that could not rearrange its endogenous TCR genes. The naïve population was isolated directly from female mice engineered to express these uniform receptors on all of their CD8+ T cells. The memory population was generated when some of the naïve cells were transferred into T cell-deficient mice that were male-female chimeras and so contained the stimulus for the naïve cells, HY antigen. The cells appropriately exhibited the patterns of surface marker expression and the cytokine secretion profiles expected of naïve and memory CD8+ lymphocytes. To test what molecules these cells require for proliferation and survival, Tanchot et al. compared the cells' fate after their transfer into irradiated hosts that differed in their expression of the potential ligands. In addition to control animals containing the correct antigen (HY) and correct MHC class I molecules (Db), cells were injected into animals without the antigen, animals with only an incorrect MHC class I molecule (Kb instead of Db), or animals without any class I molecules at all.

Necessities of life.

T lymphocytes that effect immune responses (naïve cells in blue) require continuous exposure to their correct MHC ligand for maintenance (and also antigen for proliferation), whereas memory cells that have previously responded to antigen (red) are less fussy and are maintained even in the presence of the wrong MHC ligand.

The naïve and memory CD8+ T cell populations turned out to have quite different requirements for both proliferation and survival. As expected, both proliferated vigorously when they encountered the HY antigen and the correct MHC class I molecule, Db. Yet memory, but not naïve, cells could also proliferate, although with reduced vigor, in the absence of antigen or of Db, or even in the complete absence of class I molecules. Similar to previous reports (4), both the naïve and memory populations could survive for a prolonged period without intentional stimulation by HY antigen, but they did require some kind of “tickling” by MHC class I molecules or else they disappeared by 2 weeks after transfer. Naïve cells needed the correct class I molecule, Db, whereas memory cells again seemed less particular, surviving in mice that did not express Db but did display other (incorrect) class I molecules, although they disappeared rapidly in animals devoid of all class I molecules.

Because these results were obtained (of necessity) with a heavily engineered system, and, like findings with all such systems, might harbor unexpected artifacts, they need to be confirmed by independent strategies. This is important because the results of Tanchot et al. have several potentially interesting implications. First, the different requirements of naïve and memory cells for both proliferation and survival underline the distinctiveness of these two compartments. The memory CD8+ population was clearly the less fastidious. Perhaps most striking was its ability to expand in the absence of specific antigen and to proliferate to a significant extent even in the absence of all MHC class I molecules. This observation recalls the recent finding that interferon-α can stimulate the proliferation of memory but not naïve CD8+ cells in the absence of intentional antigenic stimulation (5). It is now imperative to determine how many such factors help to maintain the memory pool; in what contexts they are produced; how memory—but not naïve—cells stay attuned to them; and how they influence the mobilization of an antigen-specific memory response.

The fact that naïve CD8+ T cells require contact with the correct MHC class I molecule in order to survive for prolonged periods in the periphery hints at a process akin to positive selection in the thymus. Positive selection favors export of a useful T cell repertoire by promoting the survival and differentiation of only those thymocytes that can productively interact with the MHC molecules expressed on thymic stromal cells. That such a peripheral selection process might be generally occurring is supported by two recent reports, relying on nontransgenic systems, that CD4+ T cells survive for much longer periods in the periphery if they make contact with MHC class II molecules (6, 7). Some of the questions immediately raised are whether the repertoire of T cell specificities undergoes any significant further shaping as a result of peripheral selection, which peripheral cells need to express MHC molecules in order to enhance survival, and precisely what receptor interactions and internal signals are involved.

These new results preview a story that will demand our close attention as it unfolds.

References

View Abstract

Navigate This Article