NMR Experiments for Studies of Dilute and Condensed Protein Phases: Application to the Phase-Separating Protein CAPRIN1.

Intrinsically disordered proteins (IDPs) or regions of intrinsic disorder in otherwise folded proteins (IDRs) play important roles in many different biological processes, including formation of biological condensates via liquid-liquid phase separation. NMR spectroscopy is a powerful tool for obtaining site-specific structural and dynamical information on IDPs/IDRs, and recent efforts have focused on the development of experiments for atomic-resolution studies of these molecules. These include triple-resonance experiments that are based on 13CO-direct detection of magnetization, exploiting increased sensitivity of cryogenically cooled probes. In order to evaluate the different classes of experiment for studies of IDRs or IDPs in both dilute and phase-separated environments, in particular at neutral and higher pHs where many of these proteins phase separate, we compared 13CO-detect versus 1Hα-detect experiments, showing that significant sensitivity gains are achieved via proton detection under the conditions of our experiments. A suite of 1Hα-detect experiments was subsequently developed for studies of IDPs/IDRs and applied to the dilute phase of a 103-residue disordered region of CAPRIN1 that phase separates at neutral pH. Residue-specific chemical shifts derived from our study enable the accurate prediction of the importance of the N-terminal Arg-containing region of this construct for promoting phase separation relative to other Arg-rich stretches of sequence, subsequently confirmed by mutagenesis. Our study emphasizes that the sequence positions of key residues can be a critical factor in controlling phase separation and highlights the unique role of NMR in establishing the relations between amino acid sequence and phase-separation propensity.

Hyperinsulinism-Causing Mutations Cause Multiple Molecular Defects in SUR1 NBD1.

The sulfonylurea receptor 1 (SUR1) protein forms the regulatory subunit in ATP sensitive K+ (KATP) channels in the pancreas. SUR proteins are members of the ATP binding cassette (ABC) superfamily of proteins. Binding and hydrolysis of MgATP at the SUR nucleotide binding domains (NBDs) lead to channel opening. Pancreatic KATP channels play an important role in insulin secretion. SUR1 mutations that result in increased levels of channel opening ultimately inhibit insulin secretion and lead to neonatal diabetes. In contrast, SUR1 mutations that disrupt trafficking and/or decrease gating of KATP channels cause congenital hyperinsulinism, where oversecretion of insulin occurs even in the presence of low glucose levels. Here, we present data on the effects of specific congenital hyperinsulinism-causing mutations (G716V, R842G, and K890T) located in different regions of the first nucleotide binding domain (NBD1). Nuclear magnetic resonance (NMR) and fluorescence data indicate that the K890T mutation affects residues throughout NBD1, including residues that bind MgATP, NBD2, and coupling helices. The mutations also decrease the MgATP binding affinity of NBD1. Size exclusion and NMR data indicate that the G716V and R842G mutations cause aggregation of NBD1 in vitro, possibly because of destabilization of the domain. These data describe structural characterization of SUR1 NBD1 and shed light on the underlying molecular basis of mutations that cause congenital hyperinsulinism.

Measurement of signs of chemical shift differences between ground and excited protein states: a comparison between H(S/M)QC and R1rho methods.

Carr-Purcell-Meiboom-Gill (CPMG) relaxation dispersion NMR spectroscopy has emerged as a powerful tool for quantifying the kinetics and thermodynamics of millisecond exchange processes between a major, populated ground state and one or more minor, low populated and often invisible 'excited' conformers. Analysis of CPMG data-sets also provides the magnitudes of the chemical shift difference(s) between exchanging states (|Deltavarpi|), that inform on the structural properties of the excited state(s). The sign of Deltavarpi is, however, not available from CPMG data. Here we present one-dimensional NMR experiments for measuring the signs of (1)H(N) and (13)C(alpha) Deltavarpi values using weak off-resonance R (1rho ) relaxation measurements, extending the spin-lock approach beyond previous applications focusing on the signs of (15)N and (1)H(alpha) shift differences. The accuracy of the method is established by using an exchanging system where the invisible, excited state can be converted to the visible, ground state by altering conditions so that the signs of Deltavarpi values obtained from the spin-lock approach can be validated with those measured directly. Further, the spin-lock experiments are compared with the established H(S/M)QC approach for measuring the signs of chemical shift differences. For the Abp1p and Fyn SH3 domains considered here it is found that while H(S/M)QC measurements provide signs for more residues than the spin-lock data, the two different methodologies are complementary, so that combining both approaches frequently produces signs for more residues than when the H(S/M)QC method is used alone.

Structural studies of FF domains of the transcription factor CA150 provide insights into the organization of FF domain tandem arrays.

FF domains are poorly understood protein interaction modules that are present within eukaryotic transcription factors, such as CA150 (TCERG-1). The CA150 FF domains have been shown to mediate interactions with the phosphorylated C-terminal domain of RNA polymerase II (phosphoCTD) and a multitude of transcription factors and RNA processing proteins, and may therefore have a central role in organizing transcription. FF domains occur in tandem arrays of up to six domains, although it is not known whether they adopt higher-order structures. We have used the CA150 FF1+FF2 domains as a model system to examine whether tandem FF domains form higher-order structures in solution using NMR spectroscopy. In the solution structure of FF1 fused to the linker that joins FF1 to FF2, we observed that the highly conserved linker peptide is ordered and forms a helical extension of helix alpha3, suggesting that the interdomain linker might have a role in orientating FF1 relative to FF2. However, examination of the FF1+FF2 domains using relaxation NMR experiments revealed that although these domains are not rigidly orientated relative to one another, they do not tumble independently. Thus, the FF1+FF2 structure conforms to a dumbbell-shape in solution, where the helical interdomain linker maintains distance between the two dynamic FF domains without cementing their relative orientations. This model for FF domain organization within tandem arrays suggests a general mechanism by which individual FF domains can manoeuvre to achieve optimal recognition of flexible binding partners, such as the intrinsically-disordered phosphoCTD.

Measuring the signs of 1H(alpha) chemical shift differences between ground and excited protein states by off-resonance spin-lock R(1rho) NMR spectroscopy.

Analysis of Carr-Purcell-Meiboom-Gill (CPMG) relaxation dispersion NMR profiles provides the kinetics and thermodynamics of millisecond-time-scale exchange processes involving the interconversion of populated ground and invisible excited states. In addition, the absolute values of chemical shift differences between NMR probes in the exchanging states, |Delta omega|, are also extracted. Herein, we present a simple experiment for obtaining the sign of (1)H(alpha) Delta omega values by measuring off-resonance (1)H(alpha) decay rates, R(1rho), using weak proton spin-lock fields. A pair of R(1rho) values is measured with a spin-lock field applied |Delta omega| downfield and upfield of the major-state peak. In many cases, these two relaxation rates differ substantially, with the larger one corresponding to the case where the spin-lock field coincides with the resonance frequency of the probe in the minor state. The utility of the methodology is demonstrated first on a system involving protein ligand exchange and subsequently on an SH3 domain exchanging between a folded state and its on-pathway folding intermediate. With this experiment, it thus becomes possible to determine (1)H(alpha) chemical shifts of the invisible excited state, which can be used as powerful restraints in defining the structural properties of these elusive conformers.

Identification of a collapsed intermediate with non-native long-range interactions on the folding pathway of a pair of Fyn SH3 domain mutants by NMR relaxation dispersion spectroscopy.

Recent 15N and 13C spin-relaxation dispersion studies of fast-folding mutants of the Fyn SH3 domain have established that folding proceeds through a low-populated on-pathway intermediate (I) where the central beta-sheet is at least partially formed, but without interactions between the NH2- and COOH-terminal beta-strands that exist in the folded state (F). Initial studies focused on mutants where Gly48 is replaced; in an effort to establish whether this intermediate is a general feature of Fyn SH3 folding a series of 15N relaxation experiments monitoring the folding of Fyn SH3 mutants N53P/V55L and A39V/N53P/V55L are reported here. For these mutants as well, folding proceeds through an on-pathway intermediate with similar features to those observed for G48M and G48V Fyn SH3 domains. However, the 15N chemical shifts extracted for the intermediate indicate pronounced non-native contacts between the NH2 and COOH-terminal regions not observed previously. The kinetic parameters extracted for the folding of A39V/N53P/V55L Fyn SH3 from the three-state folding model F<-->I<-->U are in good agreement with folding and unfolding rates extrapolated to zero denaturant obtained from stopped-flow experiments analyzed in terms of a simplified two-state folding reaction. The folding of the triple mutant was studied over a wide range of temperatures, establishing that there is no difference in heat capacities between F and I states. This confirms a compact folding intermediate structure, which is supported by the 15N chemical shifts of the I state extracted from the dispersion data. The temperature-dependent relaxation data simplifies data analysis because at low temperatures (< 25 degrees C) the unfolded state (U) is negligibly populated relative to I and F. A comparison between parameters extracted at low temperatures where the F<-->I exchange model is appropriate with those from the more complex, three-state model at higher temperatures has been used to validate the protocol for analysis of three-site exchange relaxation data.

A "three-pronged" binding mechanism for the SAP/SH2D1A SH2 domain: structural basis and relevance to the XLP syndrome.

The SH2 domain protein SAP/SH2D1A, encoded by the X-linked lymphoproliferative (XLP) syndrome gene, associates with the hematopoietic cell surface receptor SLAM in a phosphorylation-independent manner. By screening a repertoire of synthetic peptides, the specificity of SAP/SH2D1A has been mapped and a consensus sequence motif for binding identified, T/S-x-x-x-x-V/I, where x represents any amino acid. Remarkably, this motif contains neither a Tyr nor a pTyr residue, a hallmark of conventional SH2 domain-ligand interactions. The structures of the protein, determined by NMR, in complex with two distinct peptides provide direct evidence in support of a "three-pronged" binding mechanism for the SAP/SH2D1A SH2 domain in contrast to the "two-pronged" binding for conventional SH2 domains. Differences in the structures of the two complexes suggest considerable flexibility in the SH2 domain, as further confirmed and characterized by hydrogen exchange studies. The structures also explain binding defects observed in disease-causing SAP/SH2D1A mutants and suggest that phosphorylation-independent interactions mediated by SAP/SH2D1A likely play an important role in the pathogenesis of XLP.