The manuscript by S. Tharayil et al. reports on a new modular type of tag based on 4-fluoro-2,6-dicyanopyridine for the binding of lanthanoids to proteins. The lanthanoid binding moiety is constructed in two steps and allows for the modification of the final tag, such that different coordination polyhedra of the tag can relatively easily be obtained, resulting in several different tensors with varying orientation and size of the anisotropic components. This strategy is certainly novel and may prove very useful for structural work with a minimum of synthetic effort.

The manuscript is concise and written in a clear way and is definitely suited for publication in Magnetic Resonance.

The major weak points are the poor characterisation of the chelate – lanthanoid complex, the open questions on the dimeric nature of the complex and the lack of information on the affinity of the new tags towards the lanthanoids.

I recommend, therefore, to accept the manuscript with minor changes that address the following points:

1) ¹H-DOSY spectra on the titration series of DCP-(L-Cys)₂ with YCl₃ should be performed in order to clarify the dimeric nature of the obtained shifts of the complex in slow exchange. (lines 378 ff.)

2) The aforementioned titration experiments have to be carried out with different protein concentrations and should then allow the (at least approximate) determination of an affinity constant. The titration spectra should all be provided in the SI.

3) In the list of crucial properties of an ideal lanthanoid binding site (lines 33ff) a high affinity between tag and metal is mandatory for a generally applicable system for structural work.

4) There is no information provided how the model of the complex for the rotamer libraries (line 166) was obtained (DFT?, force field?) and this holds also true for Figure S11. This should be included in the SI.

5) line 120 should read NMR: spectroscopy

6) In the SI a proper characterization of compounds 2 to 5 (pages S3, S4) needs to be provided (at least NMR, MS)

This paper represents a significant attempt at the generation of lanthanide binding sites on proteins with selectivity and limited tag mobility under biocompatible conditions.  The ability to generate several variants of the lanthanide tag after the initial labeling reaction is of high interest in the efficient localization of NMR PCS tensors for structural studies, and the potential ability to label selenocysteines may allow for fewer required host protein mutations.  The produced tags appear to be more permissive towards solution studies of labeled proteins than prior attempts, although room for improvement is still apparent, as noted by the need to stay at metal: protein ratios of < 0.6 to avoid sample precipitation (lines 217-219).

The cysteine labeling reaction for a surface exposed cysteine is noted to complete overnight at RT, and the selenocysteine reaction is noted to have completed in 10 minutes (lines 175-180).  This is stated to indicate that the reaction is selective for selenocysteine, but what is the actual timescale of the cysteine reaction?  Is there a sufficient difference to avoid replacement of native cysteine residues?  Additionally, buried cysteines are shown to label poorly or not at all with FDCP; is this the same for selenocys mutants?

The high salt requirement for the cyano group reaction with the free cysteine was noted. Was this to reduce aggregation of the protein in the reaction condition?

For the EPR experiments, several concerns were apparent.

First, in the NMR section, it is noted that metal: protein ratios > 0.6:1 resulted in significant protein precipitation, yet the EPR studies used stoichiometric or excess metal?  Were there issues with precipitation in these cases, and could any of the peaks presented in the DEER distributions reflect protein aggregates?  The authors indicate that multiple tags may be able to jointly coordinate additional lanthanides, but these results could also be explained by protein dimerization that leads to tags in close proximity. Would this problem be alleviated if the concentration of protein in the metal titration be lowered significantly?

As the authors acknowledge, the modulation depths of the EPR experiments used to check for tag mobility and lanthanide stoichiometry are very small.  The authors are correct in their assertion that while these tags do not appear to be highly efficient for DEER experiments, they do appear very promising for paramagnetic NMR.  Still, the pulsed EPR measurements are presented in this work as evidence for: the limited mobility of the lanthanide tags, the stoichiometry of the tags, and for the presence of expected distance distributions in each protein of interest.  The very small modulation depths shown limit the ability to determine accurate widths for the distance distribution, and also the ability to draw conclusions regarding stoichiometry, as the vast majority of the ensemble is of unknown state.  These experiments do indicate that at least some of the protein samples are tagged and in the expected structural state, but this section would benefit from an additional method for validation of coordination stoichiometry and tag mobility.