PerspectiveCell Biology

A Scaffold Makes the Switch

See allHide authors and affiliations

Science Signaling  21 Oct 2008:
Vol. 1, Issue 42, pp. pe46
DOI: 10.1126/scisignal.142pe46


Protein kinase cascades are a reoccurring feature of signal transduction pathways. Recent investigations have focused on how kinase-scaffolding proteins help to convert a graded stimulus into a switch-like or binary response. New findings reveal that the graded-to-binary conversion can be turned on or off, depending on the location of the scaffold within the cell.

Mitogen-activated protein kinases (MAPKs) are well known to regulate cell growth and differentiation and are conserved across all eukaryotic species. They play an important role in the mating cycle of yeast Saccharomyces cerevisiae. In this example, haploid cells secrete pheromones that promote cell-cell fusion. This signal-response system has provided important insights into parallel systems in humans. Indeed, no signaling pathway has been as thoroughly dissected using genetic, biochemical, and (in more recent years) computational methods (1).

Most of the pheromone signaling components were identified through the isolation of mating-deficient (or sterile) gene mutations. Further genetic, biochemical, and molecular biological analysis revealed the order of events and established many basic principles of signaling by G proteins and MAPKs and their regulation (1). The pheromone receptor acts through a G protein, which in turn activates a protein kinase cascade that includes the MAPKs Fus3 and Kss1, a MAPK kinase (Ste7), a MAPK kinase kinase (Ste11), and a MAPK kinase kinase kinase (Ste20). Another component required for mating is Ste5. However, the role of Ste5 has until recently remained elusive. Ste5 is a member of the “scaffold” family of proteins, which have the ability to assemble multiple distinct binding partners and promote their mutual interactions. For instance, Ste5 binds to the G protein βγ subunits, Ste11, Ste7, and Fus3 (but not Kss1) (2). Further, Ste5 is absolutely required for activation of Fus3, although it does not preclude simultaneous activation of Kss1 (3). A long-standing question has been how scaffolded kinases differ from unscaffolded kinases.

To put the latest findings in perspective, recall that most extracellular signals increase or decrease in a graded fashion. At the receptor level, pathway activation is determined by the fraction of hormone-occupied receptors, and they in turn transmit a signal in proportion to that stimulus. Signaling in yeast is no exception. A graded increase in mating pheromone produces a graded response at the level of G protein activation (4) and transcriptional output (5). Other processes are less nuanced, however; for example, pheromone stimulation also leads to cell division arrest. Such an all-or-none (or binary) response is an extreme form of “ultrasensitivity,” a term defined as any response with a Hill coefficient greater than one (6). Ultrasensitivity and binary behavior have been widely observed in pathways involving MAPKs (6).

Thus, a single pathway can simultaneously produce a response that is either graded or binary, depending on the output or response measured. Scaffold proteins are likely mediators of the graded-to-binary conversion. One way this could arise is through the assembly of multistep enzyme cascades, which have the capacity to promote positively cooperative binding interactions (6). Scaffolds are also commonly believed to confer a processive kinetic mechanism, allowing each step of the pathway to proceed efficiently and thereby preserve the graded input-output relationship. In the case of MAPK activation, this entails two phosphorylation events in rapid succession, leading to full activation of the enzyme. However, the Ste5 scaffold slows MAPK activation and, thus, may actually impose a distributive kinetic mechanism in which two separate events are required for activation (7). Experimental evidence for a distributive phosphorylation mechanism comes from classical studies on the MAPKs that promote oocyte maturation in Xenopus (6, 8).

In many ways, yeast is an ideal system to determine whether a scaffold confers a graded or ultrasensitive response. This is because the same pheromone stimulus activates two MAPKs, only one of which is scaffolded. The scaffolded (Fus3) and unscaffolded (Kss1) signals then reconverge on a common transcription factor. Initial investigations of the two responses were not informative with regard to understanding how binary or graded responses are achieved, however. Transcriptional induction is commonly monitored using a pheromone-responsive promoter fused to green fluorescent protein, and the induction of fluorescence appears graded even in cells lacking either Fus3 or Kss1 (5, 9). When MAPK activity is monitored directly, however, a different picture emerges. Using antibodies that recognize the fully phosphorylated and activated forms of each kinase, Fus3 exhibits ultrasensitivity, whereas Kss1 does not (7). Moreover, a Ste5 mutant with impaired binding to Fus3 largely eliminates ultrasensitivity (7, 10, 11). Thus, it appears that scaffolds promote ultrasensitivity, but this difference is no longer evident when transcriptional induction is being measured.

How does ultrasensitivity at one step become graded at another? One explanation is the differences in time scale for the two measurements. Experiments that detect the activated forms of Fus3 or Kss1 exhibit a response within minutes of the stimulus, whereas the transcription reporter requires 30 min or more to manifest itself. This lag is due to the time needed to complete transcription, translation, and maturation of the fluorescent protein reporter. Given the steps involved, a MAPK could easily be activated in an ultrasensitive manner, yet produce a graded response at the level of reporter protein expression.

Now work from the Pryciak group reveals that ultrasensitivity is dependent on the subcellular distribution of the MAPK scaffold (9). Previous work from the same group demonstrated that Ste5 is recruited to the plasma membrane upon pheromone stimulation (12). This membrane recruitment occurs through a lipid-binding domain and interaction with the activated form of the G protein (12, 13). These investigators have now demonstrated that Ste5 recruitment converts an ultrasensitive response to a graded response (Fig. 1) and that the difference is evident even using the transcription-reporter assay. Given the limitations of the method noted above, the findings are nothing short of dramatic.

Fig. 1

A graded pheromone stimulus will confer a response that is either graded or switch-like, depending on whether the kinase scaffold protein is located to the plasma membrane or in the cytosol.

For their analysis, Takahashi and Pryciak (9) needed to compare activation of Ste5 both on and off the plasma membrane. But membrane recruitment and G protein activation go hand in hand. Thus, the authors employed some clever strategies to bypass the receptor but nevertheless allow dose-dependent pathway activation. First, they exploited a well-known (but poorly understood) phenomenon, wherein mutants lacking the MAPK Hog1 exhibit inappropriate activation of the mating pathway when subjected to an osmotic stress (14). Under these circumstances, a stress stimulus yields an ultrasensitive transcription-reporter response. However, the osmotic stress pathway normally exhibits ultrasensitivity, and so the graded-to-binary switch may occur earlier in that particular pathway (15). The authors, therefore, used a second approach by which they could activate the mating pathway at different steps, all downstream of the receptor, in a dose-dependent manner. Activation was achieved through overexpression of the G protein βγ subunits (to recruit Ste5 to the plasma membrane), by physically tethering Ste5 to the plasma membrane, or through overexpression of a constitutively active Ste11 mutant. In each case, induction of the activated mutant was achieved with a drug-inducible gene promoter. By this approach, the authors showed that Gβγ overexpression or the tethered form of Ste5 yields a graded response, whereas activation by Ste11 produces a dose-response curve that is clearly ultrasensitive. In short, it appears that the response is inherently ultrasensitive but is graded when Ste5 is restricted to the plasma membrane.

Given that Ste5 is recruited to the plasma membrane by the activated G protein, these findings imply that ultrasensitivity is a G protein–regulated phenomenon. A further implication of the work is that there may be functionally distinct pools of the MAPK: One is a pool that signals in a graded fashion at the plasma membrane, and another signals in a binary fashion within the cell.

Why bother? In yeast, the transition from graded-to-binary signaling is likely to be important during the transition from rapid cell division to pheromone-induced cell division arrest. Between these two developmental extremes is a third developmental phase whereby cells divide more slowly and also become elongated toward and later bud in the direction of a weak pheromone gradient (such as might be produced by a distant mating partner) (16, 17). This so-called chemotropic growth response occurs at doses that produce a binary transcription response (18) and requires a scaffolded MAPK (7). It also requires proper cycling of G protein activation and inactivation, given that the chemotropic growth and binary transcription are absent if G protein activity is prolonged (through loss of a GTPase-accelerating protein) (17, 18).

Thus, it seems that G proteins and scaffolds each promote switch-like behaviors, ranging from kinase activation to cell cycle arrest. Further, it is perhaps most useful to think of stimulus-response pathways as a cascade of switches, working together to produce the broad spectrum of cellular behaviors that we observe. One thing is absolutely certain: A full understanding of signal-response systems will require a detailed blueprint of the architecture that underlies the scaffold.


View Abstract

Navigate This Article