An auto-inhibited state of protein kinase G and implications for selective activation

Curation statements for this article:
  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This reported crystal structure of nearly-full-length cGMP-dependent protein kinase β (PGK1β) provides new insights into how the activity of the PKG catalytic domain is held in check by intramolecular interactions between both the upstream regulatory cGMP-binding domains and the autoinhibitory segment and the catalytic domain, and how cGMP binding to the cGMP-binding domains can relieve these inhibitory constraints leading to an increase in catalytic activity. The structure of the activating PKGIα R177Q CNB-A domain mutant, which resembles a cGMP-bound wild type CNB-A domain, provides a nice explanation for how this point mutation activates PKG Iα and leads to the development of the TAAD (Thoracic Aortic Aneurysms and Dissections) syndrome. The work will be of specific interest to the cyclic nucleotide community, and to the broader signaling community in general.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)

This article has been Reviewed by the following groups

Read the full article See related articles

Abstract

Cyclic GMP-dependent protein kinases (PKGs) are key mediators of the nitric oxide/cyclic guanosine monophosphate (cGMP) signaling pathway that regulates biological functions as diverse as smooth muscle contraction, cardiac function, and axon guidance. Understanding how cGMP differentially triggers mammalian PKG isoforms could lead to new therapeutics that inhibit or activate PKGs, complementing drugs that target nitric oxide synthases and cyclic nucleotide phosphodiesterases in this signaling axis. Alternate splicing of PRKG1 transcripts confers distinct leucine zippers, linkers, and auto-inhibitory (AI) pseudo-substrate sequences to PKG Iα and Iβ that result in isoform-specific activation properties, but the mechanism of enzyme auto-inhibition and its alleviation by cGMP is not well understood. Here, we present a crystal structure of PKG Iβ in which the AI sequence and the cyclic nucleotide-binding (CNB) domains are bound to the catalytic domain, providing a snapshot of the auto-inhibited state. Specific contacts between the PKG Iβ AI sequence and the enzyme active site help explain isoform-specific activation constants and the effects of phosphorylation in the linker. We also present a crystal structure of a PKG I CNB domain with an activating mutation linked to Thoracic Aortic Aneurysms and Dissections. Similarity of this structure to wildtype cGMP-bound domains and differences with the auto-inhibited enzyme provide a mechanistic basis for constitutive activation. We show that PKG Iβ auto-inhibition is mediated by contacts within each monomer of the native full-length dimeric protein, and using the available structural and biochemical data we develop a model for the regulation and cooperative activation of PKGs.

Article activity feed

  1. Evaluation Summary:

    This reported crystal structure of nearly-full-length cGMP-dependent protein kinase β (PGK1β) provides new insights into how the activity of the PKG catalytic domain is held in check by intramolecular interactions between both the upstream regulatory cGMP-binding domains and the autoinhibitory segment and the catalytic domain, and how cGMP binding to the cGMP-binding domains can relieve these inhibitory constraints leading to an increase in catalytic activity. The structure of the activating PKGIα R177Q CNB-A domain mutant, which resembles a cGMP-bound wild type CNB-A domain, provides a nice explanation for how this point mutation activates PKG Iα and leads to the development of the TAAD (Thoracic Aortic Aneurysms and Dissections) syndrome. The work will be of specific interest to the cyclic nucleotide community, and to the broader signaling community in general.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 and Reviewer #2 agreed to share their names with the authors.)

  2. Reviewer #1 (Public Review):

    Here, the authors have followed up their prior work on the structures of individual domains of the cyclic GMP-dependent protein kinase Iα/β (PKG Iα/β) and have generated a crystal structure of residues 71-686 of PKGIβ bound to AMP-PNP:Mn2+ in the absence of cGMP, representing a nearly full-length protein lacking only the N-terminal leucine zipper dimerization domain. In general, the individual domain structures resemble those determined previously by the authors and other groups. The AI motif in the R domain occupies the active site, as expected, and is part of an extended interface involving the helical subdomain of CNB-A with αG-helix of the C-domain. In addition, the αA helix of CNB-B interacts with the activation loop, and the αB helix of CNB-B contacts a loop in the C-lobe of the C-domain, with the CNB-A and CNB-B interdomain helix interacting with pT532 in the activation loop. Together these four R:C contacts allow the R-domain to grip the C-domain between its two CNB domains, ensuring complete inhibition of catalytic activity. The structure of the autoinhibited C-domain is similar to that of the isolated PKG C-domain and that of the PKA C-domain, but there are differences that could prevent unwanted crosstalk between cGMP and cAMP, for instance by preventing PKA RIα from binding to the PKG Iβ C-domain. Subtle changes in the individual CNB-A/B structures as well as their relative orientation upon cGMP binding suggest a mechanism for releasing the autoinhibited C-domain allowing its activation, in which cGMP binding to the CNB-A phosphate-binding cassette alters its conformation so that it cannot bind to the C-domain but instead promotes R:R interactions. Finally, the authors solved the crystal structure of the isolated CNB-A domain of PKG Iα bearing an activating R177Q disease mutation, whose structure reveals that the mutant would not be able to adopt a closed conformation and instead closely mimics the cGMP-bound WT CNB-A, suggesting a mechanistic basis for the constitutive activation of the R177Q PKG Iα, which is linked to TAAD syndrome.

    This new structure of nearly full length PKG Iβ provides a significant advance in our understanding of how the cGMP-dependent protein kinase is autoinhibited in the absence of cGMP, and provides a plausible mechanism for how it is activated upon binding cGMP, as well as defining the structural basis through which crosstalk between PKA and PKG is precluded. The main uncertainty is whether the autoinhibited PKG Iβ normally resides in a "monomeric" or "dimeric" state. The crystal structure consists of a dimer in which the linker between the R and C domains is not visible making it impossible to determine whether the R domain of one monomer interacts with the C-domain of the other monomer in the dimer or rather with its own C-domain, i.e., is autoinhibition occurring in cis or trans. Based on the results of biochemical experiments in which WT and kinase-dead mutant PKG Iβ with and without an activating KR/EE AI region mutation were co-expressed, the authors concluded that autoinhibition occur predominantly in cis, consistent with the PKG Iβ 71-686 fragment being monomeric and inactive in solution. This conclusion seems reasonable, but the issue may only be fully resolved by a structure of the intact PKG Iβ dimer. If autoinhibition does occur in cis, this leaves open the question of why PKG 1 is dimeric, and this was not really illuminated by the present structures.

    The other point is that while the structure itself is a significant advance in our understanding of how PKG Iβ is autoinhibited, the paper would be strengthened by some additional analysis of the functional effects on kinase activity and cGMP regulation of mutating PKG Iβ at newly defined contact residues in the autoinhibited structure that the authors conclude are key to autoinhibition. As it stands, the only mutation that was analyzed functionally was the KR/EE AI region mutant, which as predicted was activating.

  3. Reviewer #2 (Public Review):

    This manuscript describes the R(CNB-A and CNB-B)/C structure of the cGMP dependent kinase, PKGIβ, in which the regulatory R- domain of the kinase is free of cyclic nucleotide, and thus in an auto-inhibited state. The authors present a model for the full activation of PKG, based on structural and functional data presented in the manuscript and on previous PKG structures. Overall, the manuscript is well-presented and comprehensive - figures are well-designed, and the manuscript is well-written, if not a little long.

  4. Reviewer #3 (Public Review):

    The authors describe the crystal structure of a large fragment of PKG Ib in an autoinhibited state. The structure includes both the regulatory (R) and catalytic (C) kinase domains, and shows in atomic detail how the regulatory cGMP binding domains and autoinhibitory segment bind the kinase to block its activity. A crystal structure of one of the cGMP binding domains bearing a disease-associated mutation (TAAD, Thoracic aortic aneurysms and dissections) provides an understanding of the mechanism by which the mutation leads to constitutive activation of PKG by inducing a conformation that resembles the cyclic nucleotide bound state. This interpretation is further supported by an NMR study of the mutant that reveals chemical shifts consistent with the "open" (nucleotide-bound) conformation. A structure-function study in which variants with mutations in one or both of the active sites and regulatory domain are co-expressed shows that autoinhibition occurs in cis; that is, in an intra-chain manner, rather than as part of a dimer as is likely present in the crystal. A SAXS experiment further supports this model. The authors propose a model for PKG activation, referencing the structures described here as well as prior crystal structures of the isolated kinase and regulatory domains as "snapshots" of distinct states in the autoinhibition-activation pathway. This is a careful and technically sound study that provides a first structural view of PKG autoinhibition. It also enables comparison to the related mechanism of regulation of protein kinase A, but this aspect of the manuscript could be much better developed.