PKABased Sensors

Tsien and co-workers (Adams et al. 1991) devised the first application of the FRET technology for measuring cAMP concentration in living cells. The sensor they generated, termed FlCRhR, is based on PKA, the R and catalytic C subunits of which were labeled with rhodamine and fluorescein, respectively. In the absence of cAMP, R and C subunits interact to form the holoenzyme R2C2 and the two fluorophores come close enough for FRET to occur. When cAMP binds to the R subunits, the C subunits dissociate, thereby disrupting FRET. Although this approach allowed, for the first time, imaging of cAMP fluctuations in a living sample, it is affected by major limitations and technical difficulties, such as the necessity to microinject a large amount of the probe (|M concentrations), the aggregation of the labeled subunits and the non-specific interactions of labeled subunits with cellular structures (Goaillard et al. 2001).

An evolution of the FlCRhR prototype led to the generation of a sensor (R-CFP/ C-YFP) in which the C and R subunits of PKA are fused to the yellow (YFP) and cyan (CFP) mutants of GFP, respectively (Lissandron et al. 2005; Zaccolo et al. 2000) (Fig. 1).

cAMP binding to the R subunit generates a conformational change that reduces the affinity of the R subunit for the C subunit, leading to dissociation of the holotetramer. At the low cAMP concentration of a resting cell, most of the R and C subunits are associated, and the fused GFPs are close and in the correct orientation

Fig. 1 PKA-based sensor. (a) Schematic representation of the mechanism of action of the PKA-based FRET sensor R-CFP/C-YFP. (b) Schematic structure of the R-CFP and C-YFP components of the PKA-based sensor (left panel) and distribution of the R-CFP subunit in rat neonatal cardiac myocytes (rightpanel). The localization of the probe in parallel striations was shown to depend on anchoring of R-CFP to endogenous AKAPs (Zaccolo and Pozzan 2002). The distribution of the C-YFP subunit (not shown) overlays the distribution of the R-CFP subunit. (c) Kinetics of FRET changes recorded in rat neonatal cardiac myocytes overexpressing the R-CFP/C-YFP sensor (left panel) or the RR230K-CFP/C-YFP sensor (rightpanel). R, PKA regulatory subunit; C, catalytic PKA subunit; DD, dimerization/docking domain; IS, inhibitory sequence; domain A, cAMP binding domain A; domain B, cAMP binding domain B; Rol, rolipram; IBMX, 3-isobutyl-1-methylxan-thine; NE, norepinephrine

Fig. 1 PKA-based sensor. (a) Schematic representation of the mechanism of action of the PKA-based FRET sensor R-CFP/C-YFP. (b) Schematic structure of the R-CFP and C-YFP components of the PKA-based sensor (left panel) and distribution of the R-CFP subunit in rat neonatal cardiac myocytes (rightpanel). The localization of the probe in parallel striations was shown to depend on anchoring of R-CFP to endogenous AKAPs (Zaccolo and Pozzan 2002). The distribution of the C-YFP subunit (not shown) overlays the distribution of the R-CFP subunit. (c) Kinetics of FRET changes recorded in rat neonatal cardiac myocytes overexpressing the R-CFP/C-YFP sensor (left panel) or the RR230K-CFP/C-YFP sensor (rightpanel). R, PKA regulatory subunit; C, catalytic PKA subunit; DD, dimerization/docking domain; IS, inhibitory sequence; domain A, cAMP binding domain A; domain B, cAMP binding domain B; Rol, rolipram; IBMX, 3-isobutyl-1-methylxan-thine; NE, norepinephrine for FRET to occur. When cAMP levels increase and the two subunits dissociate, CFP and YFP disengage, and FRET is no longer possible. When FRET occurs and CFP is excited at its proper wavelength (430 nm), part of its excited-state energy is transferred to YFP, which can emit at its own wavelength (545 nm). When

FRET is abolished upon cAMP binding, only CFP emission (480 nm) can be detected upon excitation of CFP. FRET changes can be measured as changes in the emission spectrum of the probe (emission YFP/emission CFP) upon illumination at a wavelength that excites selectively the donor CFP (430 nm). Being entirely genetically encoded, the R-CFP/C-YFP sensor can be easily introduced into cells by transfection or infection (Warrier et al. 2005), therefore extending the application of this methodology to most cell types.

The R-CFP/C-YFP sensor shows an EC50 for cAMP of about 0.3 |M (Table 1), and its catalytic activity is comparable to wild-type PKA (Mongillo et al. 2004).

One specific advantage of the PKA-based sensors for cAMP is that they report fluctuations of the second messenger specifically in the compartments in which PKA resides. Thus, being the R component of the probe the isoform type II, the overexpressed R-CFP subunit binds, via its DD domain, to those AKAPs that are present in the cell under study, thereby localizing the sensor in that specific compartment. Overexpression of the R-CFP/C-YFP sensor in cardiac myocytes provides a clear example of such localization (Fig. 1b).

Using the R-CFP/C-YFP sensor, it was possible to visualize, for the first time, microdomains of cAMP in cardiac cells in correspondence with signaling units organized by AKAPs (Zaccolo and Pozzan 2002). In addition, using the R-CFP/C-YFP sensor, insight into the role of PDEs in shaping the intracellular gradients of cAMP in cardiac myocytes has been gained, leading to the conclusion that different PDEs are selectively engaged in degrading cAMP pools generated by the activation of specific G-protein coupled receptors (Mongillo et al. 2004, 2006).

In specific experimental set-ups, the affinity of the PKA-based sensor may result in being too high, and changes in cAMP concentration may not be detected

Table 1 Design and sensitivity of different probes for real-time detection of cAMP. Approximate

EC50 for each sensor is indicated.

Table 1 Design and sensitivity of different probes for real-time detection of cAMP. Approximate

EC50 for each sensor is indicated.

Sensor

Design

EC50

References

FlCRhR

Tetrameric PKA

90 nM

Adams et al. (1991)

R-CFP/C-YFP

Tetrameric PKA

0.3 |lM

Mongillo et al. (2004)

Rr23OK-cfP/c"yfP

Mutant tetrameric PKA

31.3 |iM

Mongillo et al. (2004)

Wild-type CNGA2

CNG channel

36 |iM

Rich et al. (2001)

A61-90C460W/E583M

Mutant CNG channel

15 |iM

Rich et al. (2001)

H30

Mutant Epac 1

12.5 |iMa

Terrin et al. (2006)

mpH30

Mutant Epac 1

20 |iMa

Terrin et al. (2006)

nlsH30

Mutant Epac 1

17.5 |iMa

Terrin et al. (2006)

Epac 1-camps

cAMP-binding domain

2.4 |iM

Nikolaev et al.

from Epac 1

(2004)

Epac 2-camps

cAMP-binding domain B

0.9 |iM

Nikolaev et al.

from Epac 2

(2004)

PKA-camps

cAMP-binding domain B

1.9 |iM

Nikolaev et al.

from PKA

(2004)

HCN 2-camps

cAMP-binding domain

5.9 |iM

Nikolaev et al.

from HCN 2

(2006)

aValues determined in living cells. All other values are determined in vitro aValues determined in living cells. All other values are determined in vitro accurately because of probe saturation. To overcome this limitation, the R230K mutation was introduced in the R subunit, thus generating a probe (RR230K-CFP/ C-YFP) with a two-order of magnitude lower affinity for cAMP (Table 1). Figure 1c compares the cAMP signal detected by either the R-CFP/C-YFP or the RR230K-CFP/C-YFP sensor in cardiac myocytes in which the concentration of cAMP was raised by different stimuli. The R-CFP/C-YFP sensor cannot discriminate the cAMP response to norepinephrine (NE) from the amount of cAMP released upon selective PDE4 inhibition with rolipram (Rol) or non-selective inhibition of all PDEs with 3-Isobutyl-1-methylxanthine (IBMX). This is clearly the result of probe saturation as the RR230K-CFP/C-YFP sensor can easily report the peak response to these different stimuli.

In spite of the specific advantages described above, the PKA-based sensors for cAMP have some drawbacks. First, because of the multimeric nature of the sensor, equimolar concentration of R and C subunits is required in order to form a functional tetramer, and this is not easy to control in transfected cells. Second, the cAMP-dependent dissociation of R and C subunits occurs through a complex cooperative mechanism (Taylor et al. 2005), and, therefore, the kinetics of FRET change reported by the sensor may result slower as compared to the actual kinetics of cAMP changes. In addition, the C-YFP subunit is catalytically active. This may affect the level of cAMP inside the cell by, for example, hyperactivation of PKA-sensitive PDEs. In order to overcome these limitations, a new generation of unimo-lecular and catalytically inactive cAMP sensors has been designed.

0 0

Post a comment