Facile access to C-glycosyl amino acids and peptides via Ni-catalyzed reductive hydroglycosylation of alkynes

C-Glycosyl peptides/proteins are metabolically stable mimics of the native glycopeptides/proteins bearing O/N-glycosidic linkages, and are thus of great therapeutical potential. Herein, we disclose a protocol for the syntheses of vinyl C-glycosyl amino acids and peptides, employing a nickel-catalyzed reductive hydroglycosylation reaction of alkyne derivatives of amino acids and peptides with common glycosyl bromides. It accommodates a wide scope of the coupling partners, including complex oligosaccharide and peptide substrates. The resultant vinyl C-glycosyl amino acids and peptides, which bear common O/N-protecting groups, are amenable to further transformations, including elongation of the peptide and saccharide chains.

Reviewer #3: Remarks to the Author: The manuscript by Xu, Yu and coworkers presents the development of an innovative method for nickel-catalyzed reductive additions to conbstruct hybrids of C-glycosyl motifs and amino acid derivatives. An optimized nickel (hydride) catalyst allowed for the chemo-and regio-selective hydroalkylation of alkynes derived from amino acids. Thereby innovative new structures could be accessed. The manuscript was very carefully prepared and is well written. Given the topical interest in organic electrocatalysis, along with the practical importance of aminations, I recommend publication of this fine manuscript after the following minor revision. 1) In the scope studies I do miss information on the selectivity when using more sensitive amino acids, such as tyrosine, lysine, serine, and cys.
2) The mechanism is proposed to involve a nickel hydrido species. Is there mechanistic support for this for the new method? 3) Is reductive homocoupling of the bromide 1a observed in Figure 2? 4) The method seems largely limited to terminal alkynes: Information on the use of internal alkynes would be helpful for the reader (even if not successful). 5) Are protecting group-free bromides and amino acids viable substrates?

I. Response to Decision Letter (NCOMMS-21-10854)
We are grateful to the reviewers for their kind comments and helpful suggestions that have helped us to improve our paper. As indicated in the point-by-point responses that follow, we have taken all these comments into account in the revised manuscript. The changes made in the revised manuscript are highlighted in yellow.

Reviewer #1 (Remarks to the Author):
In this work, Yu, Xu and coworkers reported a novel Ni-catalyzed hydroglycosylation of alkynes, affording a new class of C-glycosides. This reaction demonstrates a remarkable substrate scope and tolerates a broad array of sugar units. The successful preparation of 12 and 13 is particularly impressive, highlighting the power of this reaction. This method will be widely adopted to make C-glycosides that are of high medicinal value.
This reaction is challenging to accomplish. I was surprised that glycosyl bromides (prone to reduction) could be used as substrates. Moreover, the regioselectivity of this method, which gives 1,1-disubstituted alkenes is noteworthy. Conceivably, the alkene unit can serve as a handle for further derivatization, further adding to the utility of this method. I thus support the publication of this work, but have the following suggestions.

Response:
We are grateful to this reviewer for the encouraging comments.
First, as mentioned above, the regioselectivity to form vinyl Nickel species 1 is interesting.
The authors may provide a rationale for this hypothesis. Could this be because the hydrometallation has a high radical character?
Response: Thanks for the comment. Indeed, only 1,1-disubstituted alkene type products are obtained in the present reaction, whereas the corresponding 1,2-disubstituted regio-isomers have not been isolated. We attribute this regioselectivity to the "hydronickelation of terminal alkyne step (from int. B to int. C; Fig. 1b)", where the hydride is transferred to the terminal carbon of the alkyne due to the less bulkiness and thus faster insertion rate to this carbon.
However, it is not known about the radical character of the nickel hydride species (species B), and the resultant vinyl nickel species (int. C) might not possess radical character (a vinyl radical should be a highly reactive species), which could participate in the cross-coupling step with glycosyl bromide 1 to deliver int. D and E. The high valent species E might undergo a rapid reductive elimination to yield product 3. This plausible mechanism is depicted in Figure   1b.
Second, the authors made a keen observation about the fate of phosphine ligands, and proposed that they may not be involved in the catalytic cycle, but instead help to dissolve Ni or scavenge O2. However, it's also possible that phosphine ligands may just serve to stabilize Ni-catalyst by prevent it from decomposing (to metal nanoparticles). Lipshultz and Buchwald have suggested such a role of bystander phosphine ligands in their Cu-H catalyzed reactions. Third, it is interesting the authors didn't mention use of glucose derived bromides. They might be difficult substrates, but deserve mentioning in the scope to better define the limits and scope of this method.

Response:
Thanks for the expert suggestion. We have added a reaction with a glucose derived bromide. The reaction gives the desired glucosyl amino acid derivative 3ia in 52% yield, albeit with a poor β/α selectivity (β/α = 1:1; Fig. 4). A paragraph discussing briefly the β/α selectivity of C-glycosylation has been added (see Response to the first comment of reviewer #2).
Besides, other minor issues include the following: "prausible" in page 3 is a typo.
Response: Thanks. The typo has been corrected.
Reference 50 should be updated.

Response: Thanks. Ref. 50 has been updated.
Reviewer #2 (Remarks to the Author): In this work, Xu, Yu and co-workers reported a new method for the stereoselective synthesis of C-glycosides (C-glycosyl amino acids and peptides) via nickel-catalyzed reductive hydroglycosylation of alkynes. Intriguingly, they demonstrated that highly functionalized radicals (glycosyl radicals) could participate in nickel-catalyzed reductive hydroalkylations.
As a new C-glycosylation technology, it has the advantage of using readily available glycosyl bromides and unactivated alkynes as coupling partners. Furthermore, this reaction doesn't require organic zinc/magnesium and are tolerant of peptides and oligosaccharides (even tetra/pentasaccharide substrates), which is not possible with most previous transition metal Some points are raised for the authors to enhance the quality of this manuscript.

Response:
We are grateful to this reviewer for the encouraging comments.

Response:
Thanks for the suggestions. We have added one paragraph to discuss briefly the stereoselectivity of C-glycosylation following the suggestion of the reviewer. It reads: The attained stereoselectivity of the C-glycosylation could be attributed to the predominant conformation of the glycosyl radical intermediate, which is stabilized by the interaction of SOMO of the anomeric unpaired electron with lone pair of the ring oxygen and the σ* of the adjacent C2-O/C2-N bond 64-67 . Thus, a mannose-derived radical adopts preferentially a 4 C 1 conformation, leading to the 1,2-trans (α-selectivity) product in the Cglycosylation. A glucose-derived radical adopts a flexible B 2,5 conformation, thus the stereoselectivity of C-glycosylation could be shifted from 1,2-cis (α-selectivity) to 1,2-trans (β-selectivity) by using a bulkier protecting group on C2-OH; and for a glucosamine-derived radical bearing the bulky NPhth group at C2, exclusive 1,2-trans (β-selectivity) product is attained. Due to the lack of C5 substituent, a xylose-derived radical can adopt both the B 2,5 conformation and 1 C 4 conformation, thus resulting in a 1,2-trans (β-selectivity) dominated C- A study on the conformation-anomeric effect-stereoselectivity relationship in anomeric radical reactions. J. Am. Chem. Soc. 123, 11870-11882 (2001).

Response:
Thanks for the suggestion. We have performed a radical clock experiment to support the existence of glycosyl radical. Figure 6 and a paragraph are added. To probe the existence of the glycosyl radical species D (Fig. 1b), we conducted a radical clock experiment (Fig. 6). 35 Thus, δ-olefinic 1-bromo glucoside 6 and alkyne 2a were subjected to the standard Conditions II; the desired ring-closure product 7 was isolated in 33% yield with a mild diastereoselectivity (d.r. = 3:2). Though not conclusive, this result support the intermediacy of an anomeric radical species. The manuscript by Xu, Yu and coworkers presents the development of an innovative method for nickel-catalyzed reductive additions to conbstruct hybrids of C-glycosyl motifs and amino acid derivatives. An optimized nickel (hydride) catalyst allowed for the chemo-and regioselective hydroalkylation of alkynes derived from amino acids. Thereby innovative new structures could be accessed.
The manuscript was very carefully prepared and is well written.
Given the topical interest in organic electrocatalysis, along with the practical importance of aminations, I recommend publication of this fine manuscript after the following minor revision.

Response:
We are grateful to this reviewer for the encouraging comments. 1) In the scope studies I do miss information on the selectivity when using more sensitive amino acids, such as tyrosine, lysine, serine, and cys.
2) The mechanism is proposed to involve a nickel hydrido species. Is there mechanistic support for this for the new method? 3) Is reductive homocoupling of the bromide 1a observed in Figure 2?

Response:
The reductive homocoupling of bromide 1a, which would give 1,1-linked C-disaccharides, has not been observed in the reaction.
4) The method seems largely limited to terminal alkynes: Information on the use of internal alkynes would be helpful for the reader (even if not successful).
Response: Thanks for the suggestion. We have added two experiments with an internal alkyne (i.e., 2v), and the results are given in Fig. 5.

5) Are protecting group-free bromides and amino acids viable substrates?
Response: W have not tried protecting group-free glycosyl bromides and amino acids, because of the poor solubility of these substrates.

II. Other revisions
1) A "Data Availability" section has been added before the References.
2) Two more grants are added in the Acknowledgements, those are "Youth Innovation