Signaling cascades: Escape from kinetic proofreading

The conversion of an event or signal outside the plasma membrane of the cell to one inside the membrane, called signal transduction, often involves an extended reaction cascade. Usually, the process must be remarkably sensitive and selective; however, the relative inactivity of the cell in the absence of a signal cannot be sacrificed to obtain sensitivity. Many cell responses are mediated by receptor proteins and the extended reaction cascade that follows stimulation of the receptor includes not only many proteins, but often multiple steps for the activation of each protein. It is therefore desirable to find generalized traits of these systems that do not require an understanding of each microscopic parameter associated with each of the components. One general formulation of the reaction cascade has been termed kinetic proofreading. Two papers in this issue of PNAS, Liu et al. (1) and Hlavacek et al. (2), examine the characteristics of a reaction in rat basophilic leukemia cells, which have been shown previously to have behavior consistent with a kinetic proofreading scheme. The experimental study by Lui et al. presents evidence for a response in these cells that does not follow the kinetic proofreading scheme, whereas the second paper explores the theoretical underpinnings that allow a response to escape from kinetic proofreading.

The selection and development of T cells seems to be a process whereby either too much or too little signal leads to cell death with only intermediate strength signals causing further differentiation or maturation.

Historically (circa 1974), kinetic proofreading schemes were explored to help understand the remarkable accuracy of DNA replication and protein synthesis (3–8), using reactions that otherwise seemed to waste a lot of energy. This apparent waste of energy instead may be a reflection of a mechanism that ensures the fidelity of the process. Twenty years later, at a time when T cell biologists were wrestling with the finer aspects of signaling through the T cell antigen receptor (TCR), it became apparent that a kinetic proofreading scheme might explain some important general properties of this system as well (9). The selection and development of T cells seems to be a process whereby either too much signal or too little signal leads to cell death with only intermediate strength signals causing further differentiation/maturation. Perhaps more importantly, with simple signaling schemes, it was difficult to understand how the whole cell could differentiate among many weak signals (exposure to self-antigens for example) vs. very few strong signals (foreign antigens). Kinetic proofreading afforded one explanation. By viewing the signaling cascade as a series of reversible sequential steps, with only a late step being relevant for activating the cell's response, it could be seen that an antigen that displayed weak affinity—generally characterized as having rapid dissociation—would simply not activate enough of the terminal component of the cascade to influence the cell response (see Fig. 1). In effect, kinetic parameters determine the outcome rather than ligand affinity. For example, in calculations by McKeithan (9), two antigens that differed only 10-fold in the amount of peptide-MHC bound to TCR could result in a 10,000-fold difference in the “activity” of the terminal step of a cascade that was only 4 steps long. Because many of these intermediate steps in TCR activation involve energy-requiring phosphorylation reactions (with reversal-requiring phosphatase activities), it can be seen that from one standpoint, this scheme uses a lot more energy than one might suppose would be necessary. However, it may be this characteristic that confers on the T cell its ability to properly discriminate among antigens. Thus, kinetic proofreading provides interesting and maybe necessary attributes to this cellular response. The specific details of the reaction sequence are not yet needed to understand these salient features. Notably, a variety of experimental studies have supported this theoretical model (e.g., refs. 10–12).

Figure 1.

A kinetic proofreading scheme. The initiating signal, a ligand binding to a receptor, for example, induces a reaction cascade. The number of steps (N) between the initiating signal (S₀) and a signaling element that is the major determinant (S_N) of subsequent events is a variable controlling discrimination. Strictly speaking, each step (S_i) in the cascade need not be a distinct protein. For instance, several steps may simply be sequential phosphorylations of the same protein with only the terminal multiply phosphorylated state being fully active. Escape may entail generation of a messenger, a species independent of the complex, as a consequence of the activity of some intermediate step.

T cell receptors are members of the immunoreceptor class of receptors, generally characterized as requiring associated kinases to properly express their functions. In 1998, the authors of the current studies extended some of the kinetic proofreading ideas to the IgE receptor system (13), also a member of this family of receptors. There are important similarities and differences between the IgE receptor “system” and the TCR “system.” But, like all immunoreceptors, indeed, most receptors, there is an assemblage of proteins into a complex that regulates activation of the cell. Like the TCR, a very similar signaling cascade follows aggregation of the high-affinity IgE receptor (FcɛRI), with many of the same or nearly the same components involved. FcɛRI confers to cells that express this receptor—the ability to bind IgE antibody and therefore participate in biological reactions where IgE is present. There are a diverse range of reactions, some of which regulate the secretion of mediators from mast cells or basophils, and some of which may mediate antigen presentation (monocytes or dendritic cells). Like the TCR system, a given IgE molecule—bound to its respective FcɛRI on the plasma membrane—may bind a range of closely related antigens. In a 1998 study, Torigoe et al. (13) demonstrated that weakly binding antigens initiate only strong activity of the earliest components of the signaling cascade, whereas components further downstream were only weakly activated. Tightly binding antigens, on the other hand, generated strong activation of downstream elements as well. Likewise, mediator secretion from the rat basophilic leukemia cells used for these studies occurred only after stimulation with the high-affinity antigen. These results supported a kinetic proofreading scheme.

In the paper by Liu et al. (1), RBL cells again were used to examine responses stimulated through the IgE receptor. Like many cells involved in inflammatory reactions, basophils and mast cells secrete several classes of mediators. Secretion involving degranulation occurs rapidly, whereas de novo protein synthesis (e.g., cytokines and chemokines) often occurs more slowly. In experiments analogous to the studies in 1998, the authors show that low- and high-affinity antigens result in similar phosphorylation of the earliest signaling components (e.g., the beta subunit of FcɛRI), whereas only the high-affinity antigen results in significant hexosaminidase release (hexosaminidase being a component of the RBL cell granules). In contrast, generation of a chemokine mRNA (and its protein) was essentially equivalent for the two antigens. These results suggested that not all responses to signaling cascades, which otherwise seemed to involve kinetic proofreading, followed the expected traits of such a system. In the theoretical treatment of this model system explored by Hlavacek et al. (2), the authors provide insights into how it is possible to have discordant results like this in a system that nevertheless has properties consistent with kinetic proofreading. The authors offer several possible routes for responses to escape the intrinsic kinetic proofreading of the initial signaling events. For example, generation of a messenger species that is not part of the signaling complex allows events that depend on the messenger to escape proofreading. It is noted also that if some of the intermediate species in the kinetic proofreading scheme are the initiators of other cascades, then these species may be populated better by a low-affinity ligand than one with higher affinity and therefore may drive a downstream process better than the high-affinity ligand.

Evidence for kinetic proofreading now can be found for a wide range of biological processes. However, the evidence for this behavior in receptor-mediated signaling cascades is still limited, and it is not known whether it is restricted to early signaling or occurs in other compartments of the cell at other times during a cellular response. Greater enumeration of the instances of kinetic proofreading and the possible escape mechanisms may allow for a more generalized view of the behavior of the large number of reactions that involve complex protein assemblies. The current studies also indicate a need for tracking down the precise nature of the escape mechanism that allows the RBL cell to be selectively insensitive to the affinity of a ligand. In a manner analogous to predicting rough protein structure from primary sequence information or binding partners on the basis of protein domains (e.g., SH2), the signal-processing behavior of a protein assemblage—whether it displays kinetic proofreading or has escape mechanisms—may become predictable.