Biochemical characterization and cellular imaging of a novel, membrane permeable fluorescent cAMP analog

Abstract

Background: A novel fluorescent cAMP analog (8-[Pharos-575]- adenosine-3', 5'-cyclic monophosphate) was characterized with respect to its spectral properties, its ability to bind to and activate three main isoenzymes of the cAMP-dependent protein kinase (PKA-Ialpha, PKA-IIalpha, PKA-IIbeta) in vitro, its stability towards phosphodiesterase and its ability to permeate into cultured eukaryotic cells using resonance energy transfer based indicators, and conventional fluorescence imaging.

Results: The Pharos fluorophore is characterized by a Stokes shift of 42 nm with an absorption maximum at 575 nm and the emission peaking at 617 nm. The quantum yield is 30%. Incubation of the compound to RIIalpha and RIIbeta subunits increases the amplitude of excitation and absorption maxima significantly; no major change was observed with RIalpha. In vitro binding of the compound to RIalpha subunit and activation of the PKA-Ialpha holoenzyme was essentially equivalent to cAMP; RII subunits bound the fluorescent analog up to ten times less efficiently, resulting in about two times reduced apparent activation constants of the holoenzymes compared to cAMP. The cellular uptake of the fluorescent analog was investigated by cAMP indicators. It was estimated that about 7 muM of the fluorescent cAMP analog is available to the indicator after one hour of incubation and that about 600 muM of the compound had to be added to intact cells to half-maximally dissociate a PKA type IIalpha sensor.

Conclusion: The novel analog combines good membrane permeability- comparable to 8-Br-cAMP - with superior spectral properties of a modern, red-shifted fluorophore. GFP-tagged regulatory subunits of PKA and the analog co-localized. Furthermore, it is a potent, PDE-resistant activator of PKA-I and -II, suitable for in vitro applications and spatial distribution evaluations in living cells.

Figure 1

Fluorescence absorption and emission spectra of the Pharos dye. The absorption maxima of the Pharos dye dissolved in buffer with the indicated pH values is at 575 nm, whereas the emission maxima are at 617 nm. All spectra exhibit a large Stokes shift of 42 nm.

Figure 2

Fluorescence spectra of 8-[ϕ-575]-cAMP bound to PKA R subunits. 150 nM 8-[ϕ-575]-cAMP was incubated with or without fourfold molar excess of PKA R subunits as indicated in the figure. The excitation spectra (a) were detected at Em 617 nm and Ex 610 nm-430 nm; the emission spectra (b) were detected at Ex 575 nm and Em 680 nm-580 nm. The experiments were repeated four times with similar results.

Figure 3

Competitive nucleotide binding assay using fluorescence polarization. Serial dilutions of cAMP or 8-[ϕ-575]-cAMP in the presence of 1 nM 8-Fluo-cAMP were prepared. 2.5 nM R subunit was added and fluorescence polarization was determined at Ex485 nm/Em535 nm. EC₅₀ values for cAMP and 8-[ϕ-575]-cAMP binding to the R subunit isoforms were deducted from the corresponding titration curves. Each data point represents the mean +/- S.E.M. from at least triplicate measurements.

Figure 4

8-[ϕ-575]-cAMP is a potent PKA activator in vitro. For determination of apparent activation constants, purified recombinant R subunits were allowed to form holoenzyme complexes with PKA-Cα (20 nM) as detailed in the methods section. Activation assays were performed by increasing (0.3 nM-10 μM) cAMP. To obtain apparent activation constants (K_act), the normalized activity of PKA-Cα was plotted against the logarithm of the cAMP (■) and 8-[ϕ-575]-cAMP (▲) concentration and fitted according to a sigmoid dose-response model (Graphpad Prism, variable slope). Each data point represents the mean ± S.D. of two measurements. Experiments were repeated two to three times with similar results.

Figure 5

Visualization of compounds in living cells. Intracellular imaging of 8-[ϕ-575]-cAMP (a-b) and of Pharos dye (c-d) in HEK293 (a,c) and in CHO (b,d) cells after 1 hour of treatment.

Figure 6

Co-localization of 8-[ϕ-575]-cAMP and GFP-tagged R subunits. COS-7 cells were transiently transfected with GFP-hRIα (a-c) and GFP-hRIIα (d-f), incubated with 500 μM 8-[ϕ-575]-cAMP for 30 minutes. Cells were fixed and fluorescence was imaged using confocal microscopy: (a,c) green fluorescence of GFP (b,e), red fluorescence of 8-[ϕ-575]-cAMP, (c,f) merged images. The scale bar indicates 10 μm.

Figure 7

Uptake of 8-[ϕ-575]-cAMP in intact cells. Kinetics of 8-[ϕ-575]-cAMP (a-d) and of Sp-5,6-DCI-cBIMPS (e-h) uptake in living cells. HEK293 cells were transfected for transient expression of the H30 indicator of cAMP and treated with 100 μM IBMX to inhibit phosphodiesterases. The ratio between the background subtracted emission intensities at 480 nm and 545 nm is plotted as a function of time. The vertical bars indicate the administrations of either 500 μM 8-[ϕ-575]-cAMP (a-d) or Sp-5,6-DCl-cBIMPS (e-h) and of 25 μM Forskolin, which activates adenylyl cyclase and saturates the FRET-based probe. The insets show the yellow fluorescent protein fluorescence at the beginning of the experiments and the regions of interest where FRET ratios are calculated.

Figure 8

8-[ϕ-575]-cAMP activates intracellular PKA. (a) For standard BRET experiments, COS-7-cells were co-transfected with the PKA type II sensor construct or negative control plasmids (bg) as indicated and grown for 48 hours. Cells were treated with the indicated amount of 8-[ϕ-575]-cAMP for 30 minutes, or mock treated (D-PBS). BRET signals were obtained after addition of the luciferase substrate DeepBlueC™ and detection of luciferase and fluorescence light emission using a multi-label reader. Shown is a representative experiment, repeated three times; data are mean ± S.E.M., performed with n = 6 replicates. (b) A BRET titration experiment was performed as described in the methods section. Briefly, cells were co-transfected with a constant amount of BRET donor (hRIIα-Rluc) and an increasing amount of acceptor DNA (GFP²-hCα) as indicated. Before BRET read-out, cells were incubated with 0.6 mM 8-[ϕ-575]-cAMP as described above. The BRET values of two independent experiments, each performed with n = 6 replicates, were background subtracted, normalized and plotted as mean ± S.E.M.