Investigating T-cell memory

Abstract

Memory is a long-recognized, crucial and poorly understood property of adaptive immunity. Jacob and Baltimore¹ have designed an elegant genetic approach to marking memory T cells and their precursors irreversibly and have obtained intriguing results^1,2. I suggest that the particular enzyme (human placental alkaline phosphatase) chosen for irreversible marking in postnatal life contributes to certain features of this system and perturbs the homeostatic mechanisms of T-cell activation and death.

Main

Jacob and Baltimore used Cre-recombinase-mediated genetic recombination, driven by the granzyme-B promoter, to obtain irreversible and selective surface expression of placental alkaline phosphatase (PLAP; driven by the CD2 promoter) in T cells activated by antigen in adult life. In principle, this system should result in the irreversible marking of activated T cells and enhance the visualization, purification and fate mapping of activated T cells and memory T cells. As such, it should represent an invaluable tool for analysing T-cell memory, a subject central to the pathophysiology of immunity and vaccine development².

However, this genetic marking approach yielded some unexpected and unexplained results^1,2. For instance, in unimmunized mice, none of the T cells expressing high levels of CD44, which have presumably responded to environmental antigens, was positive for the PLAP marker; upon exposure to antigen, only a fraction of activated T cells became PLAP-positive. During a response to lymphocytic choriomeningitis virus (LCMV), only about 10% of LCMV-specific T cells became PLAP-positive.

Different forms of alkaline phosphatase have been described³. Human PLAP shows only limited homology to the mouse isoforms, sharing 55% identity with mouse PLAP for example⁴. Antibodies against human PLAP have thus been generated easily in various animal species, including mice³, so expression of human PLAP in T cells activated in adult life may result in a cellular and/or humoral anti-human-PLAP immune response.

An anti-human-PLAP response could explain some of the unusual features of the results obtained from irreversibly marking activated T cells with this human enzyme. For instance, selective pressure may favour weakly expressing T-cell clones. In general, the activation of immune responses directed against activated T cells may perturb the system in an unphysiological way.

These considerations caution against the overinterpretation of results derived from the indelible marking of activated T cells with human PLAP. To be informative, the same genetic approach should take advantage of non-immunogenic tracers.