Short two-armed lanthanide-binding tags for paramagnetic NMR spectroscopy based on chiral 1,4,7,10-tetrakis(2-hydroxypropyl)-1,4,7,10-tetraazacyclododecane scaffolds

Michael D. Lee a, Matthew L. Dennis ab, Bim Graham *a and James D. Swarbrick *a
aMonash Institute of Pharmaceutical Sciences, Monash University, Parkville VIC 3052, Australia. E-mail: james.swarbrick@monash.edu; bim.graham@monash.edu
bCSIRO Biosciences Program, Parkville, VIC 3052, Australia

Received 15th October 2017 , Accepted 15th November 2017

First published on 15th November 2017


A new pair of enantiomeric two-armed lanthanide-binding tags have been developed for paramagnetic NMR studies of proteins. The tags produce large and significantly different paramagnetic effects to one another when bound to the same tagging site. Additionally, they are less sensitive to sample pH than our previous two-armed tag designs.


When suitably bound to a biomolecule, paramagnetic lanthanide ions can induce a range of useful effects in NMR spectra, including paramagnetic relaxation enhancement (PRE), residual dipolar couplings (RDCs) and pseudo-contact shifts (PCSs). These effects afford unique and valuable long-range restraints for structural and dynamic studies of proteins and carbohydrates,1 as well as their interactions with other biomolecules and ligands.2,3 PCSs are particularly attractive as they are easily measured as a simple chemical shift relative to a diamagnetic reference from sensitive 2D spectra. They provide long-range distance and angular information as defined by a metal-centred anisotropic magnetic susceptibility tensor (Δχ), thereby offering promise in the investigation of challenging systems such as large protein–protein complexes4 or membrane proteins,5 for which conventional (short-range) diamagnetic NMR experiments often yield limited useful information.

Because most proteins are unable to bind lanthanide ions specifically, there has been increased development of synthetic lanthanide-binding tags (LBTs) for the site-specific attachment of these ions in order to extend the scope of paramagnetic NMR spectroscopy in structural biology and drug discovery.6 Our most recent contributions to the field have included the development of LBTs based on the octadentate macrocyclic 1,4,7,10-tetrakis((S)-2-hydroxypropyl)-1,4,7,10-tetraazacyclododecane ((S)-THP) ligand, together with its mirror image, (R)-THP.7,8 We have reported enantiomeric pairs of both single-armed and two-armed THP-based LBTs (C7/87 and T1/2,8 respectively, in Fig. 1) for conjugation to either a single cysteine residue or a pair of cysteine residues (positioned in, for example, an i,i + 4 arrangement within an α-helix) via disulfide bond formation. DOTA-like LBTs9 possess significantly higher thermodynamic and kinetic stabilities than those based on well-known chelators such as IDA, NTA and DPA,10 meaning that they can be pre-formed prior to ligation and are also unaffected by the presence of other free metal ions in solution.


image file: c7cc07961c-f1.tif
Fig. 1 Structures of C7-Ln3+, T1-Ln3+ and the new T7-Ln3+ LBT. C8, T2 and T8 are the enantiomers of C7, T1 and T7, respectively.

Single-armed tags like C7/8 are of considerable utility, however their scope can be limited by varying degrees of inherent flexibility that average the paramagnetic effects, reducing the PCS coverage and complicating the analysis of PCSs associated with spins close to the metal centre. When the local protein structure is known, two-armed LBTs like the T1/2 pair,8 or the caged lanthanide NMR probes of Ubbink and co-workers,11 can be carefully positioned across turns in helices, strands in sheets or in some loops to give reliably larger paramagnetic effects than most single-armed tags, owing to the extra rigidity afforded by two points of attachment to the protein.

The T1/2 enantiomeric tags have performed well in our hands, giving rise to single protein-tag diastereomeric species and yielding single sets of NMR signals. Moreover, the two tags have the attractive feature of yielding distinct sets of structural restraints to one another, which is useful in helping to reduce ambiguity in structural solutions arising from the inherent symmetry of the Δχ-tensor. A drawback of these tags, however, is that the magnitude of their paramagnetic effects is highly pH-dependent. Thus, while large PCSs and RDCs can be measured at quite high pH (∼8), spectra recorded at pH 6.5 revealed significantly reduced effects. Given the preference for recording protein NMR at a lower pH when possible (to reduce the line broadening effects from fast solvent exchange), the pH-dependence of the T1/2 pair could impact on their general applicability and utility.

We now report a second enantiomeric pair of two-armed, THP-based LBTs, T7/8 (Fig. 1), which are two-armed variants of our previously developed C7/8 tags. By utilising two copies of the pyridyl disulfide-activated linker present within these single-armed tags, the T7/8 designs feature the shortest linker groups of any two-armed LBT reported to date. They are also the first tags of this type to feature chiral linkers. Like T1/2, the T7/8 tags are found to be highly complementary in nature, producing large, yet substantially different paramagnetic effects to one another. However, as detailed below, they have the advantage of being relatively insensitive to pH, allowing the measurement of PCSs and RDCs across a wider biologically useful pH range (pH 6.5–8).

The detailed synthesis of the T7/8-Ln3+ tags is described in the ESI. The Tm3+ and Yb3+ complexes of T7/8 and their Y3+ complexes (as a diamagnetic reference) were attached to a human ubiquitin E24C/A28C mutant protein (see ESI for details). In each case, a single set of PCSs was observed in the 15N-HSQC spectra of paramagnetically tagged samples (Fig. 2, Fig. S5 and S6, ESI), indicating a chirally pure, diastereomeric bound LBT. PCSs were measured as the change in chemical shift (in the 1H dimension) of nuclei between the paramagnetic and diamagnetic samples (Table S4, ESI) and Δχ-tensor parameters were determined by fitting the measured PCSs to the first structure of ubiquitin12 using NUMBAT13 (Table 1).


image file: c7cc07961c-f2.tif
Fig. 2 Superposition of part of the 15N-HSQC spectra of T7- (top) and T8- (bottom) tagged ubiquitin E24C/A28C, loaded with either Y3+ (blue), Tm3+ (green, with some assignments shown) or Yb3+ (red). Spectra were recorded at 25 °C, pH 8.0 and at 600 MHz. Selected PCSs are indicated with lines.
Table 1 Δχ-Tensor parameters for T7/8-tagged ubiquitin E24C/A28C and HPPK K76C/C80a
Protein Tag Ln3+ # PCS Δχax Δχrh Q
a The axial and rhombic components of the Δχ-tensors are in units of 10−32 m3. Standard deviations (in brackets) were determined from random removal of 10% of the PCSs and recalculating the Δχ-tensors 1000 times. See Table S2, ESI for metal ion coordinates and Euler angles.
Ubiq T7 Tm3+ 36 45.0 (0.9) 18.7 (0.6) 0.01
pH 8.0 Yb3+ 41 −11.8 (0.2) −4.1 (0.2) 0.06
T8 Tm3+ 28 −39.9 (1.6) −21.0 (0.9) 0.01
Yb3+ 37 13.7 (0.7) 6.8 (0.4) 0.04
Ubiq T7 Tm3+ 46 47.1 (0.4) 18.6 (0.4) 0.02
pH 6.5 Yb3+ 50 −12.1 (0.1) −6.4 (0.2) 0.04
T8 Tm3+ 32 44.0 (1.0) 11.8 (1.7) 0.01
Yb3+ 37 −6.8 (0.4) −2.3 (0.2) 0.06
HPPK T7 Tm3+ 36 47.4 (0.3) 19.9 (0.5) 0.02
pH 8.0 T8 Tm3+ 102 57.8 (1.1) 15.3 (0.7) 0.04
Yb3+ 108 −9.0 (0.2) −2.8 (0.1) 0.06


The Δχax components for the Tm3+ and Yb3+T7/8 complexes bound to ubiquitin E24C/A28C are more than twice the magnitude of those for the corresponding single-armed C7/8 complexes attached to a single cysteine ubiquitin A28C mutant,7 demonstrating the pronounced effect that two-point attachment has on rigidifying the lanthanide ion relative to the protein. Consistently low Q values (0.01–0.06) and excellent correlations between calculated and measured PCSs (Fig. S4, ESI) also attest to the high degree of tag rigidity.

The Δχ-tensor values indicate that a PCS of 0.2 ppm would be measureable for a spin located 49/47 Å and 31/33 Å from the metal centre (along the Z axis) of the Tm3+- and Yb3+-loaded T7/T8 tags, respectively. Conversely, spins as close as 14/15 Å and 11/14 Å to the Tm3+ and Yb3+ centres of the T7/T8 tags, respectively, were readily detected in 2D spectra and accurately predicted, illustrating the large spatial PCS coverage provided by the new LBTs. The T7-Tm3+ and -Yb3+ Δχ-tensors are of opposite sign, however the principal axes are orientated such that in both cases, the majority of the protein falls within a single lobe of each tensor, producing predominantly positive PCSs (Fig. 2) that are scaled approximately to the magnitude of the Δχ-tensor. In contrast, while the T8-Tm3+ Δχ-tensor is orientated such that the majority of the PCSs are positive, the orientation of the T8-Yb3+ Δχ-tensor is such that both positive and negative PCSs are observed. This variation in Δχ-tensor orientation likely derives from subtly different interactions between the tag and the protein, as proposed previously for the C7/8 and T1/2 tags.7,8

Comparison of the PCSs measured for equivalent spins with the T7 and T8 tags are shown in Fig. 3B. The PCSs for each enantiomeric pair show extremely poor correlation with one another (R2 of 0.006 and 0.018 for the Tm3+ and Yb3+ complexes, respectively), indicating that the tags produce distinct structural restraints from one another. Notably, the coefficients of determination are an order of magnitude lower than those observed for our previous two-armed tags, T1/2, attached to the same ubiquitin mutant (R2 of 0.28 and 0.29 for the Tm3+ and Yb3+ complexes, respectively).8 Evidently, for T7/8, the combination of closer proximity to the protein and the presence of additional chiral centres within the shorter linkers leads to greater differences in terms of their interactions with respect to the chiral protein surface.


image file: c7cc07961c-f3.tif
Fig. 3 (A) Isosurface representations of the Δχ-tensors for T7- (upper) and T8- (lower) tagged ubiquitin E24C/A28C, loaded with either Tm3+ (left) or Yb3+ (right) at pH 8. Blue/red surfaces indicate positive/negative PCSs of 1 ppm. (B) Correlation between PCSs measured with T7- and T8-tagged ubiquitin E24C/A28C at pH 8, loaded with either Tm3+ (black) or Yb3+ (red). Only PCSs that were measured with both tags are shown. The solid line represents perfect correlation. (C) Correlation between the PCSs measured at pH 8.0 and 6.5 for T7/8-tagged ubiquitin, loaded with either Tm3+ (black) or Yb3+ (red). Only PCSs that were assigned at both pH values are shown.

The structures of T7- and T8-tagged ubiquitin were modelled within Xplor-NIH,14 starting from the reported X-ray structure15 of [Eu-((S)-THP)(OH2)]3+ or its mirror image and allowing the tags to move during high-temperature simulated annealing runs, while keeping the backbone and the CB atoms of the protein fixed.16 The cloud of sterically allowable metal positions for each tag was found to be quite small (5 × 3 Å) (Fig. S9, ESI) and to lie on an arc ∼6 Å above and orthogonal to a line between the CA atoms of the two cysteine residues. The independently determined metal ion positions from the PCS data for both tags lie within the cloud, which is concordant with relatively rigid attachment to the protein.

The performance of the tags was further assessed using a second protein, the K76C/C80 mutant of S. aureus 6-hydroxymethyl-7,8-dihydropterin pyrophosphokinase (HPPK).17 We were able to quantitatively conjugate the T8 tag to HPPK and determine both the T8-Tm3+ and T8-Yb3+ Δχ-tensor properties (Table 1 and Table S5, Fig. S4, S7, S8, ESI). Tagging yields with T7 were of the order of 50% for the T7-Tm3+ complex. Nevertheless, we were readily able to determine a Δχ-tensor for T7-Tm3+-tagged HPPK (Table 1). In each case, large Δχ-tensors with excellent Q values and correlations between measured and back-calculated PCSs were observed.

PCSs of equivalent spins from both enantiomeric T7/8-Tm3+ tags on HPPK were compared and showed extremely poor correlation with each other (R2 of 0.033, Fig. S10, ESI), with the correlation of coefficient again being an order of magnitude lower than that for the T1/2 tags attached to the same HPPK mutant (R2 of 0.32 for the Tm3+ complexes).8 These results further demonstrate that significantly different restraints can be measured from the same conjugation site with the enantiomeric T7/8 tag pair.

The limited HPPK tagging yields with T7 highlights potential difficulties that may be encountered when using such short two-armed tags, likely due to local steric effects around the protein (we previously achieved quantitative labelling of the same HPPK mutant with both of the longer linker T1 and T2 tags8). For the determination of protein complexes using a single site of attachment,18 where only one of the T7/8 tags can be ligated, it may prove beneficial to rather employ one of the other two-armed LBTs to obtain a complementary and distinct set of structural restraints. Notably, the very low degree of correlation between the PCSs measured for T2-Tm3+- versusT8-Tm3+-tagged HPPK K76C/C80 (R2 = 0.017) suggests that the short and long two-armed THP-based tag designs are also highly complementary to one another.

The PCS-determined metal ion positions for the T7/T8-tagged HPPK samples were in good agreement with those modelled within Xplor-NIH (Fig. S9, ESI). Calculating the tensor using the fixed metal position of the nearest neighbour in the modelled cloud for T7 yielded an excellent fit, with a slight increase in Q (to 3.2%) and a ∼5% increase in the axial component of the Δχ-tensor. Taken together with the ubiquitin data, this suggests that the metal coordinates for the T7 and T8 tags are quite predictable when attached to an α-helix. We did not investigate potential changes in the local helical structure induced by attachment of the tags to the proteins (which could be done via NOE measurements on Y3+-tagged protein samples)10a as the realistic calculated metal positions indicated that helix integrity was maintained in all cases.

At 600 MHz, 1DHN RDCs up to 15 and 28 Hz were measured for T8-Tm3+/Y3+-tagged ubiquitin and HPPK, respectively, and alignment tensors determined by fitting the RDCs (Table S6, ESI) to structures of ubiquitin and HPPK using PALES19 (Fig. S11 and Table S3, ESI). The Δχax components derived from the RDCs are 96% (ubiquitin) and 89% (HPPK) as large as the PCS-determined Δχax components, and the principle components of the alignment and Δχ-tensors are oriented similarly (Fig. S12, ESI), both indicative of a high degree of tag rigidity.

Recently, Kuprov and co-workers addressed tag mobility and the interpretation of PCSs from local spins using a distributed paramagnetic centre model.20a,b We neither resorted to such analysis nor excluded PCSs from close spins to improve fits (as is often done for mobile single-armed LBTs)20c as the Q factors were excellent with all PCSs included; even PCSs over 5 ppm in magnitude fit well (Fig. S4, ESI) using a point metal centre approximation (Table 1). While we anticipate the new approach will not be necessary for most structural studies using these rigid two-armed LBTs, comparing the metal probability distributions derived from PCSs of spins within the tagged versus the untagged domain in a multidomain protein or complex may shed light on inter-domain mobility.20b

It has been observed that PCSs and Δχ-tensors of different lanthanide complexes, featuring metal ion-coordinating hydroxyl groups, can have different sensitivities to sample pH.7,8,21 To investigate if pH could influence the paramagnetic properties of the T7/8 tags, we also recorded 15N-HSQC spectra of the tagged ubiquitin samples at pH 6.5 (Fig. S5, S6 (ESI) and Table 1 and Table S4, ESI). Comparison of the PCSs measurable at both pHs (Fig. 3C) demonstrates that the T7-Tm3+ and -Yb3+ complexes are practically insensitive to changes in this pH range (R2 values both >0.99), whereas the T8-Tm3+ and -Yb3+ complexes were slightly more sensitive (R2 values of 0.91 and 0.77, respectively). This generally low sensitivity to pH contrasts to that observed with the T1/2 tags attached to the same site.8 The T1/2 and T7/8 Δχax components measured at pH 8 are quite comparable in magnitude (the T7/8 Δχax values are ∼10% larger). At pH 6.5, however, the T1/T2 Δχ-tensors are on average 40% smaller, while the Δχ-tensors of the T7/8 tags remain similar indicating that the T7/8 tags will be more applicable to large protein systems across a wider pH range.

While the T7/8 tag pair produces far more distinct paramagnetic data compared to the T1/2 tags, the Δχ-tensors of the shorter and longer armed tags on ubiquitin E24C/A28C and HPPK K76C/C80 are generally similar in magnitude at pH 8. Further efforts to shorten the linkers are probably not warranted as successful ligation may become increasingly difficult due to steric effects, coupled with the potential for deformation of the protein. Given that the T7/8 tags have produced large Δχ-tensors across a wider pH range than the T1/2 tags, coupled with their quite predictable metal positions, we anticipate that they should perform admirably in a range of investigations using paramagnetic NMR spectroscopy for large macromolecular structural studies.

An Australian Research Council Future Fellowship (FT1301100838) to BG is gratefully acknowledged.

Conflicts of interest

There are no conflicts to declare.

Notes and references

  1. (a) I. Baig, I. Bertini, C. Del Bianco and Y. Gupta, Biochemistry, 2004, 43, 5562–5573 CrossRef CAS PubMed ; (b) T. Zhang, H.-S. Lee, B. Imperiali and J. Prestegard, Protein Sci., 2008, 17, 1220–1231 CrossRef PubMed ; (c) J. Koehler and J. Meiler, Prog. Nucl. Magn. Reson. Spectrosc., 2011, 59, 360–389 CrossRef CAS PubMed ; (d) L. Russo, M. Maestre-martinez, S. Becker and C. Griesinger, J. Am. Chem. Soc., 2013, 135, 17111–17120 CrossRef CAS PubMed .
  2. Z. Wu, M. D. Lee, T. J. Carruthers, M. Szabo, M. L. Dennis, J. D. Swarbrick, B. Graham and G. Otting, Bioconjugate Chem., 2017, 28, 1741–1748 CrossRef CAS PubMed .
  3. W. N. Chen, C. Nitsche, K. B. Pilla, B. Graham, T. Huber, C. D. Klein and G. Otting, J. Am. Chem. Soc., 2016, 138, 4539–4546 CrossRef CAS PubMed .
  4. (a) K. D. Brewer, T. Bacaj, A. Cavalli, C. Camilloni, J. D. Swarbrick, J. Liu, A. Zhou, P. Zhou, N. Barlow, J. Xu, A. B. Seven, E. A. Prinslow, R. Voleti, D. Häussinger, A. M. J. J. Bonvin, D. R. Tomchick, M. Vendruscolo, B. Graham, T. C. Südhof and J. Rizo, Nat. Struct. Mol. Biol., 2015, 22, 555–564 CrossRef CAS PubMed ; (b) Y. Hiruma, M. a S. Hass, Y. Kikui, W.-M. Liu, B. Olmez, S. P. Skinner, A. Blok, A. Kloosterman, H. Koteishi, F. Löhr, H. Schwalbe, M. Nojiri and M. Ubbink, J. Mol. Biol., 2013, 425, 4353–4365 CrossRef CAS PubMed .
  5. D. J. Crick, J. X. Wang, B. Graham, J. D. Swarbrick, H. R. Mott and D. Nietlispach, J. Biomol. NMR, 2015, 61, 197–207 CrossRef CAS PubMed .
  6. (a) C. Nitsche and G. Otting, Prog. Nucl. Magn. Reson. Spectrosc., 2017, 98–99, 20–49 CrossRef CAS PubMed ; (b) W.-M. Liu, M. Overhand and M. Ubbink, Coord. Chem. Rev., 2014, 273–274, 2–12 CrossRef CAS .
  7. M. D. Lee, C. Loh, J. Shin, S. Chhabra, M. L. Dennis, G. Otting, J. D. Swarbrick and B. Graham, Chem. Sci., 2015, 6, 2614–2624 RSC .
  8. M. D. Lee, M. L. Dennis, J. D. Swarbrick and B. Graham, Chem. Commun., 2016, 52, 7954–7957 RSC .
  9. (a) C. Loh, K. Ozawa, K. L. Tuck, N. Barlow, T. Huber, G. Otting and B. Graham, Bioconjugate Chem., 2013, 24, 260–268 CrossRef CAS PubMed ; (b) T. Müntener, D. Häussinger, P. Selenko and F.-X. Theillet, J. Phys. Chem. Lett., 2016, 7, 2821–2825 CrossRef PubMed ; (c) Y. Hikone, G. Hirai, M. Mishima, K. Ionmata, T. Ikeya, S. Arai, M. Shirakawa, M. Sodeoka and Y. Ito, J. Biomol. NMR, 2016, 66, 99–110 CrossRef CAS PubMed .
  10. (a) J. D. Swarbrick, P. Ung, S. Chhabra and B. Graham, Angew. Chem., 2011, 50, 4403–4406 CrossRef CAS PubMed ; (b) J. D. Swarbrick, P. Ung, X. Su, A. Maleckis, S. Chhabra, T. Huber, G. Otting and B. Graham, Chem. Commun., 2011, 47, 7368–7370 RSC ; (c) X.-C. Su, B. Man, S. Beeren, H. Liang, S. Simonsen, C. Schmitz, T. Huber, B. A. Messerle and G. Otting, J. Am. Chem. Soc., 2008, 130, 10486–10487 CrossRef CAS PubMed .
  11. (a) P. H. J. Keizers, J. F. Desreux, M. Overhand and M. Ubbink, J. Am. Chem. Soc., 2007, 129, 9292–9303 CrossRef CAS PubMed ; (b) P. H. J. Keizers, A. Saragliadis, Y. Hiruma, M. Overhand and M. Ubbink, J. Am. Chem. Soc., 2008, 130, 14802–14812 CrossRef CAS PubMed ; (c) W. Liu, P. Keizers, M. Hass, A. Blok, M. Timmer, A. J. Sarris, M. Overhand and M. Ubbink, J. Am. Chem. Soc., 2012, 132, 17306–17313 CrossRef PubMed ; (d) W.-M. Liu, S. P. Skinner, M. Timmer, A. Blok, M. A. S. Hass, D. V. Filippov, M. Overhand and M. Ubbink, Chem. – Eur. J., 2014, 20, 6256–6258 CrossRef CAS PubMed .
  12. A. S. Maltsev, A. Grishaev, J. Roche, M. Zasloff and A. Bax, J. Am. Chem. Soc., 2014, 136, 3752–3755 CrossRef CAS PubMed .
  13. C. Schmitz, M. J. Stanton-Cook, X.-C. Su, G. Otting and T. Huber, J. Biomol. NMR, 2008, 41, 179–189 CrossRef CAS PubMed .
  14. C. D. Schwieters, J. J. Kuszewski, N. Tjandra and G. M. Core, J. Magn. Reson., 2003, 160, 66–74 CrossRef .
  15. C.-H. Huang, J. Hammell, S. J. Ratnakar, A. D. Sherry and J. R. Morrow, Inorg. Chem., 2010, 49, 5963–5970 CrossRef CAS PubMed .
  16. J. D. Swarbrick, P. Ung, M. L. Dennis, M. D. Lee, S. Chhabra and B. Graham, Chem. – Eur. J., 2016, 22, 1228–1232 CrossRef CAS PubMed .
  17. S. Chhabra, O. Dolezal, B. M. Collins, J. Newman, J. S. Simpson, I. G. Macreadie, R. Fernley, T. S. Peat and J. D. Swarbrick, PLoS One, 2012, 7, e29444 CAS .
  18. T. Saio, M. Yokochi, H. Kumeta and F. Inagaki, J. Biomol. NMR, 2010, 46, 271–280 CrossRef CAS PubMed .
  19. M. Zweckstetter, Nat. Protoc., 2008, 3, 679–690 CrossRef CAS PubMed .
  20. (a) G. T. P. Charnock and I. Kuprov, Phys. Chem. Chem. Phys., 2014, 16, 20184–20189 RSC ; (b) E. A. Suturina, D. Häussinger, K. Zimmermann, L. Garbuio, M. Yulikov, G. Jeschke and I. Kuprov, Chem. Sci., 2017, 8, 2751–2757 RSC ; (c) D. Shishmarev and G. Otting, J. Biomol. NMR, 2013, 56, 203–216 CrossRef CAS PubMed .
  21. M. Lelli, G. Pintacuda, A. Cuzzola and L. Di Bari, Chirality, 2005, 17, 201–211 CrossRef CAS PubMed .

Footnote

Electronic supplementary information (ESI) available: Synthesis and characterisation of T7-Ln3+ complexes, details for expression, purification and tagging of ubiquitin E24C/A28C and HPPK K76C/C80, details for protein NMR spectroscopy, Δχ-tensor and alignment tensor determination, 15N-HSQC spectra of differently tagged ubiquitin E24C/A28C and HPPK K76C/C80, Xplor-NIH modelled metal ion positions on ubiquitin and HPPK, full tables of Δχ-tensor and alignment tensor properties, correlations between measured and calculated PCSs and RDCs, correlations between PCSs measured with T7 and T8, Sanson-Flamsteed projections of Δχ-tensor and alignment tensor error analyses, experimental PCSs and RDCs. See DOI: 10.1039/c7cc07961c

This journal is © The Royal Society of Chemistry 2017