Organocatalytic enantio- and diastereoselective cycloetherification via dynamic kinetic resolution of chiral cyanohydrins

Enantioselective approaches to synthesize six-membered oxacycles with multiple stereogenic centres are in high demand to enable the discovery of new therapeutic agents. Here we present a concise organocatalytic cycloetherification for the highly enantio- and diastereoselective synthesis of tetrahydropyrans involving simultaneous construction of two chiral centres, one of which is fully substituted. This method involves dynamic kinetic resolution of reversibly generated chiral cyanohydrins. A chiral bifunctional organocatalyst selectively recognizes a specific chair-like conformation of the intermediate, in which the small steric effect of the linear cyano group as well as its anomeric effect play important roles in controlling stereoselectivity. The products offer additional utility as synthetic intermediates because the cyano group can be further transformed into a variety of important functional groups. This strategy provides a platform to design efficient approaches to obtain a wide range of optically active tetrahydropyrans, which are otherwise synthetically challenging materials.

in accepting the above explanation. To me is more of a stereoelectronic effect of aryl/alky vs CN group, where aryl/alkyl preferred to stay at equatorial position. Which was exactly observed by them (Chem. Lett. 45, 2016, 1300-1303and by Gharpure (Eur. J. Org. Chem. 2014. Although, the results are very good from the synthetic point of view, I am not able to recommend this work for the publication in Nature Communication because of above reason. However, there are few more comments need to be address before submitting to another journal: 1) Authors should cite the Eur. J. Org. Chem. 3570-3574 (2014) in this they have prepared cis-2,6-Disubstituted Tetrahydropyrans with diastereoselectivity. 2) In figure-2b authors explained matched and mismatched TS, what happening when the same reaction performed with epimeric catalyst.
3) Further to confirm anomeric effect authors may have to synthesize 2,5-disubstituted furans under similar reaction conditions. 4) Did the authors observe any changes in selectivity in polar solvents (other than tabulated solvents) because they have the ability to reduces the lone pair (anomeric) interactions? Other nucleohiles which does not show anomeric effect (such as alkyle) need to explored. 5) In Table 2 all reactions were carried out about 24 hours. Is there any electronic effects of enone moiety on cyclization. 6) In figure 4 diastereoselectivity of structure 9 and 11 are missing. Have they maintained the same diastereoselectivity of structures 8 and 10? 7) In supporting information, please check HRMS data of compound 3k it has a difference between theoretical and experimental value is around 0.04 and NMR data of 1f and 1i k data showing solvent peaks.

Revision Request
(1) from Reviewer Answer: The synthesis of a tetrahydrofuran (THF) derivative was investigated, and Supplementary Figure 4 was added to the SI. However, under the present conditions, the stereoselectivities were not as good as those obtained from the reactions affording the tetrahydropyran (THP) derivatives. We believe the anomeric effect is essential for obtaining the excellent diastereoselectivities, but the small steric effect of the cyano group, decreasing 1,3-diaxial interactions in the six-membered oxacycle, also seems to be important.
Individual optimization of bifunctional catalysts is also necessary for obtaining higher enantioselectivity in THF synthesis. Thus, we consider the current protocol is useful for the construction of chiral THPs. Further optimization for the synthesis of THFs is currently underway, and we hope to investigate it as a part of our next work and report it in the near future. Answer: From the reaction using ketone 5, "<1%" means that sufficient product was not obtained and thus, the ee of product 6 could not be analysed. Because of this low yield, the corresponding cyanosilylation was carried out, and the ee of product 7 was analysed.
Additionally, we also investigated the reactions from the possible racemic intermediate. Because it is difficult to isolate the cyanohydrin because of the reversibility of its formation, we carried out the reactions using the corresponding racemic silyl ether, and Supplementary Figure 8 was added to the SI. However, the reaction did not afford the cyclic product and more than 80% of the starting material was recovered. Thus, the reaction mechanism was discussed by referring to the results indicated in Figure  Answer: The reaction from an aldehyde was carried out, and Supplementary Figure 5 was added to the SI. However, the diastereomeric ratio (dr) was not as good as that of the reactions from ketones under the present conditions. We believe the anomeric effect is essential for obtaining the excellent diastereoselectivities, but the small steric effect of the cyano group compared with another substituent on the same carbon also seems to be important to favor its axial position. Thus, we consider the current protocol is useful for the construction of a tetrasubstituted chiral centre, and supplements the conventional diastereoselective synthesis of less substituted derivatives (such as Eur. J. Org. Chem. 3570-3574 (2014)). Further optimization for the use of aldehyde substrates is currently underway, and we hope to investigate it as a part of our next work and report it in the near future.
Supplementary As mentioned below (Revision Request (4) from Reviewer #3), additional experiments show that polar solvents, which have the ability to reduce lone pair (anomeric) interactions, decrease the diastereoselectivity, and the use of a tertiary alcohol substrate bearing an alkyl group instead of the cyano group resulted in low diastereoselectivity. We believe that these results as well as the results indicated in Figure  Biophys. 318, 241-246 (1995). Answer: The reaction affording the 2,5-substituted THF was investigated as mentioned above (Revision Request (2) from Reviewer #1), and Supplementary Figure 4 was added to the SI. However, under the present conditions, the dr was not as good as those in the reactions affording the 2,6-substituted THPs. We believe the anomeric effect is essential for obtaining the excellent diastereoselectivities, but the small steric effect of the cyano group, decreasing 1,3-diaxial interactions in a six-membered oxacycle, also seems to be important. Thus, we consider the current protocol is useful for the construction of chiral THPs. Further optimization of the synthesis of THFs is currently underway, and we hope to investigate it as a part of our upcoming work and report it in the near future. We additionally investigated two polar solvents, CH3CN and EtOH, and they resulted in the decline of diastereoselectivity, while the high enantioselectivities were maintained.
To examine an alkyl group instead of the cyano group, the kinetic resolution of a substrate with a tertiary alcohol bearing a methyl group was investigated; the reaction resulted in significantly low diastereoselectivity with low to moderate enantioselectivities.
We consider these results support the importance of the anomeric effect by the cyano group in the control of stereoselectivity on the tetrasubstituted chiral centre.
Page 4, line 22, a sentence was modified as follows: Solvent optimization studies identified CH2Cl2, CHCl3, and hydrocarbon solvents as affording especially high stereoselectivities with good yields (Table 1, entries 1, 10-12), The reaction could also be carried out using a smaller amount of 2, and lower catalyst loading of 4a, giving the same excellent stereoselectivity, with a slight decrease in the yield (   In Table 2  Answer: All reactions shown in Table 2  Answer: Yes, 9 and 10 were obtained as single diastereomers, although 8 and 10 were diastereomer mixtures with respect to the stereochemistry of the secondary alcohol. For the sake of clarity, Figure 4 was corrected as follows. NMR data of 1f, 1i, and 1k data: For 1f and 1i, the solvents were further removed in vacuo, the NMR spectra were obtained again and added to the supporting information. We believe the NMR of 1k is now suitable, but if there are any other problems, we are willing to further investigate them.

Additional Revision
(1): During the revision process, another colleague participated in the additional experiments. Thus, we would like to add an additional author.
Author names (the main manuscript and the SI): The notes were modified as follows: "Naoki Yoneda, Yuki Fujii, Keisuke Asano* and Seijiro Matsubara*"

Results:
Subheadings were added as follows: "Optimization of reaction conditions." "Substrate scope."

Methods:
Section "Data availability" was added as follows:

Methods:
The subheading was corrected as follows: "General procedure for the asymmetric synthesis of tetrahydropyrans 3 To a 5-mL ..." ↓ "General procedure for the asymmetric synthesis of tetrahydropyrans 3.

REVIEWERS' COMMENTS:
Reviewer #1 (Remarks to the Author): In their revised manuscript, Dr Asano and co-workers have taken on board the comments and criticism of this reviewer (and others) and have produced an improved report of their interesting work, fully suitable for publication as an Article in Nature Communications . Thus, this reviewer commends them for being receptive to constructive input and duly supports the publication of their work in Nature Communications.

Reviewer #2 (Remarks to the Author):
I examined the revised manuscript and the correction table by the authors carefully and I found that the revision requests by the reviewers are well answered and the revision is considered to fulfill the requirements for the criteria of acceptance. I would like to recommend this revised manuscript to be published in the journal essentially as it stand.

Reviewer #3 (Remarks to the Author):
Considering the difficulty in controlling the stereselectivity in six-membered oxacycles, this strategy provides an efficient approach to a wide range of optically active tetrahydropyrans.The results are highly important. The authors have addressed many comments. But, my major concern where the authors claimed that the catalysts recognizes a specific chair-like conformation of the intermediate, in which a cyano-induced anomeric effect also plays a key role in the control of stereoselectivity is not addressed. To me is more of a stereoelectronic effect of aryl/alky vs CN group, where aryl/alkyl preferred to stay at equatorial position. The energy difference between axial and equatorial conformers (−∆Go): Ph (12 KJ/mol) and CN (0.8 KJ/mol) in substituted cyclohexane. The difference (−∆∆Go) is 11.2 KJ/mol which is more than sufficient to provide an exclusive formation of equatorial Ph and forcing CN to be axial. Moreover, another methyl keto substitution situated at cis to Ph group. Therefore, the analysis of 'cyano-induced anomeric effect plays a key role in the control of stereoselectivity' is not convincing to this reviewer. corresponds to an equatorial/axial ratio of ca. 13:1, the formation of 3n resulted in 19:1 dr in our reaction (see Figure 3); therefore, the anomeric effect also contributes to the selectivity. Thus, the manuscript was modified to state that the stereoselectivities are controlled by the small steric interaction  Page 4, line 76:

Response to Referees
"also favoured by the anomeric effect" was changed to ↓ "also favoured by the weak 1,3-diaxial interactions and anomeric effect" Page 6, line 131: "despite the small difference in size between a methyl group and a cyano group, reinforcing the strong influence of the anomeric effect" was changed to ↓ "despite the relatively small difference in size between methyl and cyano groups, reinforcing that assistance was provided by the anomeric effect" Page 9, line 182: "The anomeric effect induced by the cyano group also plays an important role" was changed to ↓ "The weak steric interaction and anomeric effect induced by the cyano group also play important roles"

Supplementary Figure 4:
"We believe the anomeric effect is essential for obtaining the excellent diastereoselectivities, but the small steric effect of the cyano group, decreasing 1,3-diaxial interactions in the six-membered oxacycle, also seems to be important." was changed to ↓ "Although we believe the anomeric effect also assists with obtaining the excellent diastereoselectivities, the small steric effect of the cyano group, which weakens 1,3-diaxial interactions in the six-membered oxacycle, seems to be essential."

Supplementary Figure 5:
"We believe the anomeric effect is essential for obtaining the excellent diastereoselectivities, but the small steric effect of the cyano group compared with another substituent on the same carbon also seems to be important to favor its axial position." was changed to ↓ "Although we believe the anomeric effect also assists with obtaining the excellent diastereoselectivities, the small steric effect of the cyano group compared with that of another substituent on the same carbon seems to be essential to favour its axial position."