Decoding Ca²⁺ signals

Abstract

Receptor-stimulated Ca²⁺ signals come in several flavors. The Ca²⁺ signals can be decoded linearly or by integration of the response. How the duration of the signal conveyed by cytosolic Ca²⁺ concentration ([Ca²⁺]_i) changes is regulated is not well understood. Liu et al. (Liu, Q., S.A. Walker, D. Gao, J.A. Taylor, Y.-F. Dai, R.S. Arkell, M.D. Bootman, H.L. Roderick, P.J. Cullen, and P.J. Lockyer. 2005. J. Cell Biol. 170:183–190) now report an example of decoding based on the differential regulation of Ras function by two Ca²⁺-sensitive Ras inhibitors: Ca²⁺-promoted Ras activator (CAPRI), which extends the duration of the effect of Ca²⁺ on Ras activity, and Ras GTPase activating-like protein (RASAL), which functions as a linear decoder of the Ca²⁺ signal.

The Ca²⁺ signal generated by physiological agonist concentrations is most often in the form of Ca²⁺ oscillations that initiate at discrete cellular sites and can either remain confined to a particular cellular microdomain (such as the apical pole of secretory cells) or propagate to all cellular domains in the form of repetitive Ca²⁺ waves (Petersen, 2004). The amplitude and frequency of the Ca²⁺ oscillations and speed of the Ca²⁺ waves are regulated by the intensity of the stimulus (Berridge et al., 2000). Furthermore, the pattern of the Ca²⁺ oscillations and waves are receptor specific in the same cell (Kiselyov et al., 2003). These intricate Ca²⁺ signals regulate cellular activities as diverse as vision and neurotransmitter release on a micro- or millisecond time scale and gene regulation on the scale of hours or days (Berridge et al., 2003).

How is the same Ca²⁺ signal decoded to regulate multiple functions with different spatial and temporal characteristics? A simple form of decoding is an exact translation of the Ca²⁺ signal to a physiological response. An example is direct binding of Ca²⁺ to effector proteins, such as the Ca²⁺-activated Cl⁻ and K⁺ channels in epithelia (Melvin et al., 2005). Another example is the uptake by the mitochondria of Ca²⁺ released from the ER to regulate enzymes involved in mitochondrial energy metabolism (Hajnoczky et al., 1995). A more complex form of decoding is achieved by differential Ca²⁺ sensitivity of the sensors. An example is the family of synaptotagmins, which display a range of apparent affinities for Ca²⁺ to confer specific Ca²⁺ dependence to exocytotic events (Sudhof, 2004). The most intricate form of decoding requires translating the information in the frequency and amplitude of Ca²⁺ oscillations into a physiological response, such as the activation of the NFAT, OAP, and NF-κB transcription factors (Dolmetsch et al., 1998; Li et al., 1998; Tomida et al., 2003). The molecular mechanism responsible for this form of decoding is not known.

Although we intuitively assume that decoding of complex Ca²⁺ signals involves activation of multiple Ca²⁺ sensors with different spatial and temporal characteristics that converge on the same pathway, to date there has been no clear example of this. The study by Liu et al. (2005) shows the selective response of two Ras GTPase-activating proteins (GAPs), Ca²⁺-promoted Ras activator (CAPRI), and Ras GTPase activating-like protein (RASAL) to complex Ca²⁺ signals.

Ras proteins are binary switches that cycle between the GTP-active and GDP-inactive forms and regulate many signaling pathways, including the MAPK cascade, to regulate cell proliferation and differentiation (Downward, 2003). The intensity and duration of Ras activation is determined by the relative activity of the guanine nucleotide exchange factors (GEFs) that catalyze the exchange of GDP for GTP to activate Ras and the GAPs that catalyze the intrinsic Ras GTPase and inactivate Ras. Ras is activated by tyrosine kinase-linked receptors through recruitment and activation of Ca²⁺-independent GEFs and GAPs (Takai et al., 2001), and by Ca²⁺-mobilizing receptors, including G protein–coupled receptors, that recruit and activate Ca²⁺-dependent GEFs and GAPs (Cullen and Lockyer, 2002). Previous work identified CAPRI and RASAL as Ca²⁺-dependent Ras GAPs (Lockyer et al., 2001). CAPRI and RASAL are members of the GAP1 family that also includes GAP1^IP4BP and GAP1^m (Cullen, 1998). As illustrated in Fig. 1 A for RASAL and CAPRI, the family is typified by four conserved structural domains; tandem C2 domains; a GAP-related domain (GRD); a pleckstrin homology (PH) domain, and a Bruton's tyrosine kinase motif (Btk). The C2 domains of GAP1^IP4BP and GAP1^m do not mediate Ca²⁺-dependent plasma membrane (PM) translocation (Lockyer et al., 1997); however, the C2 domains are essential for the Ca²⁺-dependent translocation of CAPRI and RASAL.

Figure 1.

Structural motif prediction and alignment of PH domains. Structural motif prediction and alignment of PH domains was done using the 3D-PSSM Web Server V 2.6.0 at http://www.sbg.bio.ic.ac.uk/~3dpssm. The structure of BTK PH domain was taken from Hyvonen and Saraste (1997).

In an earlier work, Walker et al. (2004) showed that RASAL faithfully translates the changes in [Ca²⁺]_i to PM translocation events and inactivation of Ras. Intense receptor stimulation resulted in a single [Ca²⁺]_i transient and a matching transient translocation of RASAL to the PM, whereas weak receptor stimulation generated Ca²⁺ oscillations and a synchronized oscillatory translocation of RASAL. Translocation required RASAL's tandem C2 domains—a finding confirmed in Liu et al. (2005). Hence, the response of RASAL to [Ca²⁺]_i indicates that RASAL is a linear decoder of [Ca²⁺]_i signals.

Although the translocation of CAPRI to the PM is also dependent on its tandem C2 domains, Liu et al. (2005) report that CAPRI does not behave as a linear decoder of [Ca²⁺]_i changes. The initial [Ca²⁺]_i increase triggered by intense or weak stimulation caused C2 domain–dependent translocation of CAPRI to the PM. CAPRI remained at the PM despite the reduction of [Ca²⁺]_i to a plateau or maintained Ca²⁺ oscillations. Switching between RASAL and CAPRI C2 domains converted CAPRI to a linear decoder of Ca²⁺ oscillations and conferred persistent retention of RASAL at the PM. Importantly, Liu et al. (2005) found that the removal of external Ca²⁺ to prevent Ca²⁺ influx by store-operated channels converted CAPRI translocation from persistent to transient, and removal of the stimulus rapidly retrieved CAPRI from the PM. The dependence of CAPRI's retention at the PM on Ca²⁺ influx indicates that CAPRI is a sensor of Ca²⁺ influx. Ca²⁺ influx may maintain high Ca²⁺ levels next to the PM. The requirement for persistent agonist stimulation may reflect dependence on receptor-mediated lipid metabolism and binding of the PH domain of CAPRI to phosphatidylinositides because mutation of the conserved tryptophan of the PH domain of CAPRI (W664A) resulted in a transient, rather than persistent, translocation of CAPRI to the PM. Sequence analysis suggests that the PH domains of CAPRI and RASAL are most similar to that of Bruton's tyrosine kinase (Hyvonen and Saraste, 1997). The predicted CAPRI PH domain is more intact than the RASAL PH domain (Fig. 1 B) and it is functional, as evident from the effect of the W664A mutation. The RASAL PH domain lacks important structural motifs, which may render it nonfunctional. Indeed, Liu et al. (2005) show that PM retention segregates with CAPRI's PH domain.

The cooperation between the tandem C2 domains and the PH domain makes CAPRI an integrator of the Ca²⁺ signal. As a consequence, the duration of CAPRI's retention at the PM, and therefore its action as a GAP, is markedly extended. Hence, the study of Liu et al. (2005) provides an example for a mechanism by which the timing and duration of the Ca²⁺ signal can be differentially decoded. CAPRI and RASAL are both deactivators of Ras, and their combined action can be used to decode the complex forms of the receptor-evoked Ca²⁺ oscillations and waves.

The findings of Liu et al. (2005) raise several questions. For example, do the cells use the properties of the two GAPs to regulate physiological functions of different duration? Recent work showed that activation of Ras and the ERK/MAPK cascade is sensitive to the frequency of Ca²⁺ oscillations (Kupzig et al., 2005). What is the role of CAPRI and/or RASAL in this form of regulation? How do the tandem C2 domains of CAPRI mediate [Ca²⁺]_i-dependent translocation while the PH domain mediates Ca²⁺ influx-dependent retention? Decoding of Ca²⁺ signals by CAPRI and RASAL is only one side of the equation. Two families of Ca²⁺-dependent GEFs have been identified: Ras-GRF1 and Ras-GRF2, and the Ras-GRP/CalDAG GEFs (Cullen and Lockyer, 2002). At present, however, it is not known how these GEFs decode Ca²⁺ signals. Further insight into this process is needed before we can solve the puzzle of how a complex pathway like the Ras/MAPK pathway decodes agonist-evoked Ca²⁺ signaling.