Abstract
Motivation
Correlated Nuclear Magnetic Resonance (NMR) chemical shift changes identified through the CHEmical Shift Projection Analysis (CHESPA) and CHEmical Shift Covariance Analysis (CHESCA) reveal pathways of allosteric transitions in biological macromolecules. To address the need for an automated platform that implements CHESPA and CHESCA and integrates them with other NMR analysis software packages, we introduce here integrated plugins for NMRFAM-SPARKY that implement the seamless detection and visualization of allosteric networks.
Availability and implementation
CHESCA-SPARKY and CHESPA-SPARKY are available in the latest version of NMRFAM-SPARKY from the National Magnetic Resonance Facility at Madison (http://pine.nmrfam.wisc.edu/download_packages.html), the NMRbox Project (https://nmrbox.org) and to subscribers to the SBGrid (https://sbgrid.org). The assigned spectra involved in this study and tutorial videos using this dataset are available at https://sites.google.com/view/chescachespa-sparky.
Supplementary information
Supplementary data are available at Bioinformatics Online.
1 Introduction
Protein allostery is one of the primary mechanisms for regulating biological function. It relies on long-range couplings between orthosteric and remote allosteric sites (Boulton and Melacini, 2016). Such couplings typically arise from transitions between macromolecular states with distinct structural and/or dynamic profiles. When allostery relies on subtle conformational and/or dynamical changes, the underlying allosteric networks often remain elusive to traditional structural determination methods. Protein NMR chemical shift perturbations offer a robust approach to mapping otherwise elusive allosteric networks. Specifically, variations in 1H and 15N resonance frequencies are exquisitely sensitive not only to nearest-neighbor binding contacts but also to subtle, yet functionally relevant, allosteric transitions underlying long-range effects. Both CHESPA and CHESCA assume that the dynamic conformational equilibria underlying the allosteric couplings occur in the fast chemical-exchange regime. In this case, the observed NMR chemical shifts are linear, population-weighted averages of conformer-specific values. Hence, it is possible to utilize peak positions to monitor shifts in local conformational equilibria.
The growing cohort of CHESPA and CHESCA applications (Akimoto et al., 2013; Axe et al., 2014; Chen et al., 2017; Selvaratnam et al., 2011, 2012; Shao et al., 2020; VanSchouwen et al., 2015; Walker et al., 2019; Wang et al., 2019; Xu et al., 2019) illustrates the potential of these methods. To fully leverage this potential, both methods require a seamless integration with the PINE-SPARKY.2 and the I-PINE webserver (Lee and Markley, 2018; Lee et al., 2019) for semi-automated protein chemical shift assignment and validation through 3D NMR triple-resonance experiments for the same protein in multiple bound or mutated forms. To enable such integration, we developed plugins and a webserver for the implementation and visualization of CHESPA and CHESCA through one of the most popular biomolecular NMR analysis suites: NMRFAM-SPARKY (Lee et al., 2015) (Fig. 1 and Supplementary Fig. S1).
Fig. 1.
General scheme for the CHESPA- and CHESCA-SPARKY plugins. The plugins start with NMR spectra of an allosteric protein system in multiple states. Then the assigned spectra are loaded into the new GUIs for the CHES(P/C)A plugins. Results are automatically computed and presented in the CHES(P/C)A plugins. Further details available in Supplementary Information
2 Implementation
The CHESPA (two-letter-code cH; Supplementary Fig. S2, left panel) and CHESCA (two-letter-code ch; Supplementary Fig. S2, right panel) plugins enable the semi-automatic visualization of allosteric networks starting from unassigned NMR spectra (Supplementary Methods). To execute CHESPA-SPARKY, both reference and perturbation vectors need to be defined as well as parameters, such as 15N versus 1H weights, between-state and inter-peak chemical shift cutoffs. CHESCA-SPARKY set up is similar but relies on a library of differentially perturbed protein samples, along with additional parameters, such as correlation coefficient cutoff values. Both methods are executed locally with agglomerative clustering and singular value decomposition run by the NMRFAM server. Results are plotted with the integrated NDP-PLOT module and can be highlighted through PyMOL molecular visualization if a PDB structure is provided.
3 Results and conclusions
Both CHESPA- and CHESCA-SPARKY plugins are illustrated through the analysis of a set of 1H, 15N HSQC NMR spectra of the RIα subunit 119–244 construct of the cAMP-dependent protein kinase (protein kinase A, PKA) and its four complexes with cAMP and analogs. The results of the demonstration are available in the Supplementary Figures S3‒S8.
In conclusion, we have designed an integrated software platform that offers a seamless workflow from protein NMR spectra to allosteric network analysis and visualization. The new platform will enable the facile elucidation of allosteric mechanisms without having to export, compile and analyze multiple peak lists and chemical shift data through several nonintegrated scripts and software packages. Combined with the PINE-SPARKY.2 plugin, our CHES(P/C)A-SPARKY plugins significantly lowers the barrier for non-NMR specialists to map protein allosteric networks.
Supplementary Material
Acknowledgement
We thank Rashik Ahmed for helpful discussions.
Funding
This work was supported by the Canadian Institutes of Health Research Grant 389522 (to G.M.), the Natural Sciences and Engineering Research Council of Canada Grant RGPIN-2014-04514 (to G.M.), the National Institute of Health (P41GM103399 to J.L.M.; GM100310 and HL144130 to G.V.), the National Science Foundation (DBI-1902076 to W.L. and J.L.M.), and the University of Colorado Denver (STYPE 61193205 to W.L.).
Conflict of Interest: none declared.
Contributor Information
Hongzhao Shao, Department of Chemistry and Chemical Biology, Hamilton, ON L8S 4L8, Canada.
Stephen Boulton, Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, ON L8S 4L8, Canada.
Cristina Olivieri, Department of Biochemistry, Molecular Biology and Biophysics, Minneapolis, MN 55455, USA.
Hebatallah Mohamed, Department of Chemistry and Chemical Biology, Hamilton, ON L8S 4L8, Canada.
Madoka Akimoto, Department of Chemistry and Chemical Biology, Hamilton, ON L8S 4L8, Canada.
Manu Veliparambil Subrahmanian, Department of Biochemistry, Molecular Biology and Biophysics, Minneapolis, MN 55455, USA.
Gianluigi Veglia, Department of Biochemistry, Molecular Biology and Biophysics, Minneapolis, MN 55455, USA; Department of Chemistry, University of Minnesota, Minneapolis, MN 55455, USA.
John L Markley, Biochemistry Department, National Magnetic Resonance Facility at Madison, University of Wisconsin-Madison, Madison, WI 53717, USA.
Giuseppe Melacini, Department of Chemistry and Chemical Biology, Hamilton, ON L8S 4L8, Canada; Department of Biochemistry and Biomedical Sciences, McMaster University, Hamilton, ON L8S 4L8, Canada.
Woonghee Lee, Biochemistry Department, National Magnetic Resonance Facility at Madison, University of Wisconsin-Madison, Madison, WI 53717, USA; Department of Chemistry, University of Colorado Denver, Denver, CO 80217, USA.
References
- Akimoto M. et al. (2013) Signaling through dynamic linkers as revealed by PKA. Proc. Natl. Acad. Sci. USA, 110, 14231–14236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Axe J.M. et al. (2014) Amino acid networks in a (β/α) 8 barrel enzyme change during catalytic turnover. J. Am. Chem. Soc., 136, 6818–6821. [DOI] [PubMed] [Google Scholar]
- Boulton S., Melacini G. (2016) Advances in NMR methods to map allosteric sites: from models to translation. Chem. Rev., 116, 6267–6304. [DOI] [PubMed] [Google Scholar]
- Chen H. et al. (2017) Elucidation of a four-site allosteric network in fibroblast growth factor receptor tyrosine kinases. Elife, 6, e21137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee W., Markley J.L. et al. (2018) PINE-SPARKY.2 for automated NMR-based protein structure research. Bioinformatics, 34, 1586–1588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee W. et al. (2015) NMRFAM-SPARKY: enhanced software for biomolecular NMR spectroscopy. Bioinformatics, 31, 1325–1327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee W. et al. (2019) I-PINE web server: an integrative probabilistic NMR assignment system for proteins. J. Biomol. NMR, 73, 213–222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Selvaratnam R. et al. (2011) Mapping allostery through the covariance analysis of NMR chemical shifts. Proc. Natl. Acad. Sci. USA, 108, 6133–6138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Selvaratnam R. et al. (2012) The auto-inhibitory role of the EPAC hinge helix as mapped by NMR. PLoS One, 7, e48707. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shao H. et al. (2020) Mechanism of action of an EPAC1-selective competitive partial agonist. J. Med. Chem., 63, 4762–4775. [DOI] [PubMed] [Google Scholar]
- VanSchouwen B. et al. (2015) Mechanism of cAMP partial agonism in protein kinase G (PKG). J. Biol. Chem., 290, 28631–28641. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Walker C. et al. (2019) Cushing’s syndrome driver mutation disrupts protein kinase A allosteric network, altering both regulation and substrate specificity. Sci. Adv., 5, eaaw9298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang Y. et al. (2019) Globally correlated conformational entropy underlies positive and negative cooperativity in a kinase’s enzymatic cycle. Nat. Commun., 10, 799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xu Y. et al. (2019) Identifying coupled clusters of allostery participants through chemical shift perturbations. Proc. Natl. Acad. Sci. USA, 116, 2078–2085. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.