Complete protein assignment from sets of spectra recorded overnight.

A flexible and scalable approach for protein NMR is introduced that builds on rapid data collection via projection spectroscopy and analysis of the spectral input data via joint decomposition. Input data may originate from various types of spectra, depending on the ultimate goal: these may result from experiments based on triple-resonance pulse sequences, or on TOCSY or NOESY sequences, or mixtures thereof. Flexible refers to the free choice of spectra for the joint decompositions depending on the purpose: assignments, structure, dynamics, interactions. Scalable means that the approach is open to the addition of similar or different experiments, e.g. larger proteins may require a wider selection of triple-resonance based experiments. Central to the proposed approach is the mutual support among the different spectra during the spectral analysis: for example, sparser triple-resonance spectra may help decomposing (separating) spin systems in a TOCSY or identifying unique NOEs. In the example presented, backbone plus side chain assignments of ubiquitin were obtained from the combination of either two or three of the following projection experiments: a 4D HCCCONH, a 4D HNCACO and a 3D HNCACB. In all cases, TOCSY data (4D HCCCONH) proved crucial not only for the side chain assignments, but also for the sequential assignment. Even when total recording time was reduced to about 10 h, nearly complete assignments were obtained, with very few missing assignments and even fewer differences to a reference.

Automated protein backbone assignment using the projection-decomposition approach.

Spectral projection experiments by NMR in conjunction with decomposition analysis have been previously introduced for the backbone assignment of proteins; various pulse sequences as well as the behaviour with low signal-to-noise or chemical shift degeneracy have been illustrated. As a guide for routine applications of this combined tool, we provide here a systematic analysis on different types of proteins using welldefined run-time parameters. As a second result of this study, the backbone assignment module SHABBA was extensively rewritten and improved. Calculations on ubiquitin yielded again fully correct and nearly complete backbone and CHß assignments. For the 128 residue long azurin, missing assignments mostly affect Hα and Hß. Among the remaining backbone (plus Cß) nuclei 97.5 % could be assigned with 1.0 % differences to a reference. Finally, the new SHABBA algorithm was applied to projections recorded for a yeast histone protein domain at room temperature, where the protein is subject to partial unfolding: this leads to unobservable resonances (about a dozen missing signals in a normal 15N-HSQC) and extensive degeneracy among the resonances. From the clearly observable residues, 97.5 % of the backbone and CHßresonances could be assigned, of which only 0.8 % showed differences to published shifts. An additional study on the protein MMP20, which exhibits spectral difficulties to an even larger extent, explores the limitations of the approach.

PRODECOMPv3: decompositions of NMR projections for protein backbone and side-chain assignments and structural studies.

UNLABELLED: PRODECOMP (projection decomposition) is an implementation of a multi-way decomposition algorithm for the analysis of two-dimensional projections of high-dimensional nuclear magnetic resonance spectra. The newest version, PRODECOMPv3, features a dramatic speedup, more reliable decompositions, a substantial reduction in memory demands, a new graphical user interface and integration into third-party software. These improvements extend the applicability of decompositions to novel types of NMR data on proteins, yielding backbone and side-chain assignments as well as structural information, and therewith enabling complete characterizations of proteins. AVAILABILITY: Program, short manual and an example calculation are freely available at www2.chem.gu.se/bcbp/nmr/prodecomp.html.

DIADECOMP: A new approach to analyze decompositions from projection spectroscopy.

We demonstrate for the first time a complete small protein characterization with the projection-decomposition approach, including full assignments as well as determination of the 3D fold. In TOCSY- and NOESY-type 4D experiments, pairing of signals from hydrogens and from their respective heavy atoms in decompositions represents a new problem. An approach, referred to as "DIADECOMP" (diagonal decomposition), is introduced to solve this problem; it consists of two separate decompositions of the input projections, differing in a 45° rotation of the spectral axes. While DIADECOMP requires a somewhat complex formulation, in practice it results in observing signals in the rotated decompositions that correspond to sums or differences of frequencies. When applied to a small protein, human defensin ß6, the analysis of a HCC(CO)NH-TOCSY with DIADECOMP results in largely unambiguous assignments of the aliphatic side chain groups. Furthermore, DIADECOMP applied to a 15N-HSQC-NOESY-15N-HSQC provides all expected short distances between amide groups (defined as all HN-HN distances <3.5Å in a reference structure). It is worth noting that short HN-HN distances unambiguously define α-helices, the alignment of ß-strands in sheets, as well as the presence of ß-bulges. This approach of using a minimal amount of NMR data, namely four projection experiments recorded in ∼2.5days, resulted for the human defensin ß6 in complete assignments and a backbone fold with a RMSD of the non-flexible structure of 0.6Å. Uniqueness of decompositions specifically from TOCSY- and NOESY-type 4D experiments is discussed.

Structural characterisation of a histone domain by projection-decomposition.

We demonstrate that two projection experiments, a (15)N-HSQC-NOESY-(15)N-HSQC and a (13)C-HSQC-NOESY-(15)N-HSQC, recorded for a histone domain from yeast, contain enough information to support a structural characterisation of the protein. At the temperature used, 298 K, the histone domain exhibits a very high extent of chemical shift degeneracy that is uncharacteristic for a fully folded domain. Nonetheless, a structured core of 67 residues, which is formed by three α-helices and a two-stranded ß-sheet is defined by this NOESY data; this core structure was shown earlier to be present at lower temperature. The above two experiments, which together required 18 h of instrument time, are part of a set of five projection experiments acquired during 2.5 days with the goal of complete characterisation of proteins, including full resonance assignment and structure.

Assignment of protein NMR spectra based on projections, multi-way decomposition and a fast correlation approach.

We present an approach for the assignment of protein NMR resonances that combines established and new concepts: (a) Based on published reduced dimensionality methods, two 5-dimensional experiments are proposed. (b) Multi-way decomposition (PRODECOMP) applied simultaneously to all acquired NMR spectra provides the assignment of resonance frequencies under conditions of very low signal-to-noise. (c) Each resulting component characterizes all spin (1/2) nuclei in a (doubly-labeled) CbetaH(n)-CalphaH-C'-NH-CalphaH-CbetaH(n) fragment in an unambiguous manner, such that sequentially neighboring components have about four atoms in common. (d) A new routine (SHABBA) determines correlations for all component pairs based on the common nuclei; high correlation values yield sequential chains of a dozen or more components. (e) The potentially error-prone peak picking is delayed to the last step, where it helps to place the component chains within the protein sequence, and thus to achieve the final backbone assignment. The approach was validated by achieving complete backbone resonance assignments for ubiquitin.