Global response of diacylglycerol kinase towards substrate binding observed by 2D and 3D MAS NMR

Escherichia coli diacylglycerol kinase (DGK) is an integral membrane protein, which catalyses the ATP-dependent phosphorylation of diacylglycerol (DAG) to phosphatic acid (PA). It is a unique trimeric enzyme, which does not share sequence homology with typical kinases. It exhibits a notable complexity in structure and function despite of its small size. Here, chemical shift assignment of wild-type DGK within lipid bilayers was carried out based on 3D MAS NMR, utilizing manual and automatic analysis protocols. Upon nucleotide binding, extensive chemical shift perturbations could be observed. These data provide evidence for a symmetric DGK trimer with all of its three active sites concurrently occupied. Additionally, we could detect that the nucleotide substrate induces a substantial conformational change, most likely directing DGK into its catalytic active form. Furthermore, functionally relevant interprotomer interactions are identified by DNP-enhanced MAS NMR in combination with site-directed mutagenesis and functional assays.


(A) Purity, functionality and structural homogeneity of ssNMR samples
The purity of DGK in DDM micelles was checked by absorption spectroscopy and SDS-PAGE (Fig. S1a), while BN-PAGE analysis offers a reliable assessment of the oligomeric state. It shows DGK exclusively in its trimeric form without any aggregates visible (Fig. S1b).
Additionally, a sucrose gradient was carried out. It indicates a homogenous protein reconstitution into the DMPC/DMPA liposomes (Fig. S1c). To verify the functionality of DGK in liposomes, we used a coupled enzyme assay as described before 1,2 . The activities are comparable to those reported previously 2 . They were highly reproducible from sample to sample (n=8, see Tab. S3: wtDGK in DMPC/DMPA in its apo state). To further investigate the quality of the proteoliposome sample, we carried out 1D 13 C and 15 N cross-polarization (CP) MAS as well as 2D 13 C-13 C proton driven spin diffusion (PDSD) spectra 3 (Fig. S2b).
Despite spectral overlap, which is characteristic for α-helical proteins, a fine structure is clearly observable. The 15 N-CP spectrum, for instance, shows a sharp band at approximately 107.0 ppm corresponding to 15 N glycine signals (Fig. S2a). The 2D 13 C-13 C PDSD spectrum was performed with a short carbon-carbon mixing time (20 ms) to provide one-bond correlations between aliphatic atoms (Fig. S2b). It serves as a fingerprint of the sample, evaluating structural homogeneity, resolution and secondary structure. The high spectral resolution demonstrates a homogenous sample preparation. Fig. S2c illustrates the superposition of the PDSD spectrum with the resonance assignment of the thermostable DGK mutant, published by Yang and co-workers 4 . Though this assignment was obtained using MAS NMR as well, significant deviations made a de-novo assignment of wild type DGK   Figure S3: Comparison of DGK embedded in different liposome compositions. (a) Superposition of 2D 13 C-13 C PDSD spectra of U-13 C, 15 N-wtDGK reconstituted into 90mol% DMPC/ 10mol% DMPA (black), 100mol% DMPC (green) and 90mol% DMPC/ 10mol% DMPG (yellow). The NMR spectra were recorded at 275 K. (b and c) Enlargement of the respective selected regions in the 2D 13 C-13 C PDSD spectra. The comparison reveals a similar fingerprint for DGK in all three liposome compositions. (d) Activity of DGK reconstituted in 90mol% DMPC/ 10mol% DMPA (black), 100mol% DMPC (green) and 90mol% DMPC/ 10mol% DMPG (yellow) demonstrating a similar activity in all three liposome compositions. The activity data were acquired at 30 °C.100% activity corresponds to the rate recorded with wtDGK in 90mol% DMPC/ 10mol% DMPA of 90 (± 9.9) µmol min -1 mg -1 . Experiments were repeated three times. The activity was calculated as the mean value. Error bars correspond to standard deviations. 2) recorded immediately after the incubation (black) and after 30 d (green) demonstrate that the fully saturated system is stable over a long period of time without any significant signs of degradation. Figure S5: 2D scalar-coupling based 1 H-15 N HETCOR (a), 1 H-13 C HETCOR (b) and 13 C-13 C TOBSY (c) of U-13 C, 15 N-DGK with residue-specific assignments. All residues, which could not be detected and assigned by dipolar-coupling based experiments are considered as possible candidates for detection by J-based experiments. INEPT and TOBSY were used for 1 H-15 N or 1 H-13 C heteronuclear polarization and 13 C-13 C homonuclear mixing, respectively. Peaks for Arg9 and Lys12 are highlighted green, since they could be assigned unambiguously. Peaks for the aromatic rings were folded in the indirect dimension to save experimental time. Amino acids that correspond to the His-tag are labelled by 'tag'.

(D) Secondary structure analysis
Based on the chemical shift assignment, a secondary structure analysis was carried out using the chemical shift index 6 . It is compared to the DGK X-ray structure (PDB 3ZE4, chain A) 7 and to the secondary structure obtained by MAS NMR for a thermostable DGK-mutant 4 in Fig. S6. All three structures feature substantial similarities, especially concerning the high α-helical content. However, there are few differences. The crystal structure shows small deviations around the interhelical turn (T) between helix 1 (H1) and the surface helix (SH), around the periplasmic loop (PL) between helix 1 (H1) and helix 2 (H2), as well as the cytoplasmic loop (CL) between helix 2 (H2) and helix 3 (H3). In subunit A of the crystal structure, the position of T and PL is slightly shifted upstream by two residues compared to the MAS NMR structures. Additionally, T is one residue longer and PL one residue shorter in the X-ray structure than in the MAS NMR structures. Concerning the position and/or length of the CL, all three structures vary from each other. CL is shifted from the residues 83-87 in the MAS NMR structure of wtDGK to the residues 81-85 in the MAS NMR structure of the thermostable mutant and to the residues 83-90 (subunit A) of the X-ray structure of wtDGK.
However, it has to be noted that the positions and lengths of the non-helical structures are even inconsistent between the three different subunits A, B and C within the crystal structure 4,7 .  6 . For Gly residues and residues without any assignment of Cβ, only Cα secondary shifts were considered. Strongly positive (≥ 1.5 ppm) values of the CSI indicate an α-helical structure, whereas negative or near-zero values imply deviations from helicity. In addition, the secondary structure of wild-type DGK determined by ssNMR is compared with the ssNMR structure of the thermostable mutant 4 and the crystal structure of wtDGK (PDB 3ZE4, chain A) 8 . Rectangles symbolize α-helical regions including the surface helix (SH) and the three transmembrane helices (H1-3), whereas solid lines reflect deviations from helicity including the interhelical turn (T), the periplasmic (PL) as well as the cytoplasmic loop (CL). Residues that were not resolved by ssNMR or by X-ray crystallography are illustrated by dashed lines. Differences between the secondary structures obtained by X-ray crystallography and ssNMR are highlighted in green.
ssNMR wild-type ssNMR mutant X-ray wild-type chain A   15  16  19  26  29  46  52  82  88  117   26  14  29  30  33  46  52  80  86  116   9  2 8  3 2  4 85 3  8 3  9  Afterwards, SDS is replaced by DDM, leading to mixed labelled DGK trimers, which can then be reconstituted into a lipid bilayer. Assembling [CN]-DGK from 13 C (gray) and 15 N (white) labelled monomers leads to 4 distinct trimer configurations, which are differently populated. The average number of NC interfaces per trimer is 1.5, of which only 50 % are unique (N → C vs. C → N). Due to this statistical distribution, the number of potential cross-protomer interactions involving residues with 13 C-labelled side chains on the one side and 15 N-labelled side chains on the other is small, making the use of DNP essential. (b) BN-PAGE analysis on DGK and its RxA mutants verifies successful trimer formation after disruption and mixing as shown in (a). Trimers are seen in DDM micelles (1), which are disrupted by SDS (2) but form again after detergent exchange to DDM (3). All RxA mutants feature a similar oligomerization behavior as the wild-type, suggesting that the respective arginines, which are all located in extramembranous regions of DGK, are not essential for the trimer formation. For each DGK sample a new gel was prepared and processed the same way. Lanes that were non-adjacent in the respective gels are separated by a black line. Full-length gels are presented in Supplementary Figure S9. (c) Coupled activity assay with wtDGK and RxA-DGK embedded into lipids, showing a reduction of activity for all Arg-mutants compared to the wild-type. This in turn displays the importance of all arginines for the catalytic activity. DGK trimers from DDM micelles (-SDS) and DGK trimers from DDM micelles after SDS treatment (±SDS) were reconstituted into lipid bilayers and then measured. Almost full activity (85%) is regained after detergent exchange and subsequent trimer formation, which is in agreement with previously reported activity data based on unfolding/refolding experiments on wtDGK 9 . The activity was measured three times and calculated as the mean value. Error bars correspond to standard deviations. (d) DNP-enhanced 15 N− 13 C-TEDOR spectra of mixed labelled trimers (U-13 C/U 15 N 12 C)-DGK and its Arg-mutants. A mixing time of 6.25 ms (24 rotor cycles) was used for all experiments, which is optimal for N-C distances between 2.5-3 Å. Three crosspeaks can be identified: Intra-protomer natural abundance N-CO contact (1), intra-residue natural abundance contact between ArgNη,ε -Cζ (3) and cross-protomer contacts between ArgNη,ε and GluCδ/AspCγ/AsnCγ (4). All spectra are normalized with respect to the N-CO crosspeak. A reduction in the intensity of resonance (2) is observed upon introducing the R9A, R81A and R92A mutation, while no significant effect is seen for R22A, R32A and R55A. These data show, that all three residues R9, R81 and R92 are involved in cross-protomer interactions. These findings are in principle compatible with the 3D crystal structure 8 , in which these residues have suitable locations at the interface to be involved in such interactions. Slices through the crosspeaks were created by integrating along the 15 N dimension from 68 to 84 ppm and 108 to 138 ppm, respectively