Search

Among cyclic nucleotide phosphodiesterases (PDEs), PDE6 is unique in serving as an effector enzyme in G protein-coupled signal transduction. In retinal rods and cones, PDE6 is membrane-bound and activated to hydrolyse its substrate, cGMP, by binding of two active G protein α-subunits (Gα*). To investigate the activation mechanism of mammalian rod PDE6, we have collected functional and structural data, and analysed them by reaction–diffusion simulations. Gα* titration of membrane-bound PDE6 reveals a strong functional asymmetry of the enzyme with respect to the affinity of Gα* for its two binding sites on membrane-bound PDE6 and the enzymatic activity of the intermediary 1 : 1 Gα* · PDE6 complex. Employing cGMP and its 8-bromo analogue as substrates, we find that Gα* · PDE6 forms with high affinity but has virtually no cGMP hydrolytic activity. To fully activate PDE6, it takes a second copy of Gα* which binds with lower affinity, forming Gα* · PDE6 · Gα*. Reaction–diffusion simulations show that the functional asymmetry of membrane-bound PDE6 constitutes a coincidence switch and explains the lack of G protein-related noise in visual signal transduction. The high local concentration of Gα* generated by a light-activated rhodopsin molecule efficiently activates PDE6, whereas the low density of spontaneously activated Gα* fails to activate the effector enzyme.

Phytochromes are modular photoreceptors of plants, bacteria and fungi that use light as a source of information to regulate fundamental physiological processes. Interconversion between the active and inactive states is accomplished by a photoinduced reaction sequence which couples the sensor with the output module. However, the underlying molecular mechanism is yet not fully understood due to the lack of structural data of functionally relevant intermediate states. Here we report the crystal structure of a Meta-F intermediate state of an Agp2 variant from Agrobacterium fabrum. This intermediate, the identity of which was verified by resonance Raman spectroscopy, was formed by irradiation of the parent Pfr state and displays significant reorientations of almost all amino acids surrounding the chromophore. Structural comparisons allow identifying structural motifs that might serve as conformational switch for initiating the functional secondary structure change that is linked to the (de-)activation of these photoreceptors., Phytochromes are photoreceptors that are present in plants, bacteria and fungi. Here the authors present crystal structures of the phytochrome Agp2 from Agrobacterium fabrum in the parent Pfr state as well as a functional Meta-F intermediate and discuss mechanistic implications for photoconversion.

Phytochromes are bimodal photoswitches composed of a photosensor and an output module. Photoactivation of the sensor is initiated by a double bond isomerization of the tetrapyrrole chromophore and eventually leads to protein conformational changes. Recently determined structural models of phytochromes identify differences between the inactive and the signalling state but do not reveal the mechanism of photosensor activation or deactivation. Here, we report a vibrational spectroscopic study on bathy phytochromes that demonstrates that the formation of the photoactivated state and thus (de)activation of the output module is based on proton translocations in the chromophore pocket coupling chromophore and protein structural changes. These proton transfer steps, involving the tetrapyrrole and a nearby histidine, also enable thermal back-isomerization of the chromophore via keto-enol tautomerization to afford the initial dark state. Thus, the same proton re-arrangements inducing the (de)activation of the output module simultaneously initiate the reversal of this process, corresponding to a negative feedback mechanism.

Dynein light chain 8 (DLC8) is a ubiquitous eukaryotic protein regulating diverse cellular functions. We show that the obligate intracellular parasite Toxoplasma gondii harbors 4 DLC8 proteins (TgDLC8a-d), of which only TgDLC8a clusters in the mainstream LC8 class. TgDLC8b-d proteins form a divergent and alveolate-specific clade. TgDLC8b-d proteins are largely cytosolic, whereas TgDLC8a resides in the conoid at the apical end of T. gondii. The apical location of TgDLC8a is also not shared by its nearly identical Eimeria (EtDLC8a), Plasmodium (PfDLC8), or human (HsDLC8) orthologs. Notwithstanding an exclusive conoid targeting, TgDLC8a exhibits a classical LC8 structure. It forms a homodimer by swapping of the β strands that interact with the antiparallel β' strands of the opposing monomers. The TgDLC8a dimer contains two identical binding grooves and appears to be adapted for multitarget recognition. By contrast, the previously reported PfDLC8 homodimer is shaped by binding of the β strand with the parallel β' strand and lacks such a distinct binding interface. Our comparisons suggest an unexpected structural and functional divergence of the two otherwise conserved proteins from apicomplexan parasites. Finally, we demonstrate that a phosphomimetic S88E mutation renders the TgDLC8a-S88E mutant monomeric and cytosolic in T. gondii, and its overexpression inhibits the parasite growth in human fibroblasts.

Photoreceptor cGMP Phsophodiesterase 6 (PDE6) is the effector molecule of visual signal transduction and mediates fast response of light signals. The rod holo-PDE6 comprises catalytic (α, β; each ~ 90 kDa) and two identical inhibitiory (γ; ~ 10 kDa) subunits. The catalytic subunits comprise N-terminal tandem GAF domains followed by C-terminal catalytic domains and isoprenylations for membrane-association. Contrary to activation of other tandem GAF comprising PDEs, PDE6 activation does not occur via cGMP-induced concerted conformational changes. Rather two copies of the α-subunit of retinal G-Protein (Gα*), transducin, activate PDE6 by partially displacing the inhibitory subunits. The activation of PDE6 has therefore been described as a “de-inhibition”. The affinity of Gα* to PDE6 and the enzymatic activity of the intermediary 1:1 complex is highly disputed, therefore a conclusive activation model is lacking so far. Our combined structural, enzymatic and computational investigations deal with the activation-mechanism of PDE6. Our cryo electron microscopy (EM) structure of PDEαβ catalytic core shows an elongated bell-shaped structure with symmetric side-by-side arrangement of the two subunits with flexible membrane-binding domains. A comparison with nearly full-length inactive PDE2A structure [1] suggests that less compaction of both subunits and higher degree of conformational freedom of the catalytic domains result in constitutive activation of PDE6αβ, which is kept inactive by the inhibitory γ subunits. Furthermore, the structure of PDE6 suggests Gα* binding-sites pointing to opposing faces. The enzymatic characterization using Gα* titration of the PDE6 however reveal striking asymmetry of the two catalytic subunits with a high and a low affinity binding site for Gα*. Occupancy of the PDE6 with one Gα* induces negligible activity, whereas occupancy with two copies of Gα* leads to full enzyme activity. Such an activation mechanism constitutes a “coincidence switch” that allows noise filtering (i.e., spontaneously produced Gα* do not activate PDE6). Our spatiotemporal simulation work indeed confirms that spontaneously generated Gα* lead to the formation of singly occupied PDE6 and only a high local concentration of Gα*, as produced by an active receptor (rhodopsin), leads to doubly Gα* occupied effector complex. Therefore the localized large concentration of Gα* combined with the asymmetric properties of PDE6 constitutes a “density switch” that allows suppression effector level noise and reliable reporting of single quantum events in rod photoreceptor cells.