Identification and structure determination of a third cyclobutane photodimer of thymidylyl-(3'.fwdarw.5')-thymidine: The trans-syn-II product

Abstract

Photolysis of thymidylyl-(3'-->5')-thymidine (TpT) has been previously reported by many workers to lead to only two cyclobutane dimers, the cis-syn and trans-syn-I dimers. This is curious in light of the fact that photolysis of thymidylyl-(3'-->5')-deoxyuridine (TpdU), which has a hydrogen in place of a methyl group at C6 of the 3'-thymidine, produces two trans-syn diastereomers. Recently, we discovered by way of X-ray crystallography that photolysis of the (Rp)-methyl phosphate ester of TpT leads to two trans-syn diastereomers, prompting us to reexamine the photochemistry of TpT. In this paper we show that sensitized photolysis of TpT also leads to the hitherto unknown trans-syn-II diastereomer in 2% yield. We also report the solution-state 1H NMR assignment of the trans-syn-II photodimer of TpT and its (Rp)-methyl phosphate ester by way of 2D homonuclear Hartmann--Hahn experiment, rotating frame nuclear Overhauser effect spectroscopy, and proton-detected 1H-31P correlation spectroscopy. Conformational analysis of three-bond 1H-1H, 1H-31P, and 13C-31P coupling constant data established a close similarity between the solution-state structures of the trans-syn-II photodimer of TpT and its (Rp)-methyl phosphate ester, with the crystal structure of the methyl phosphate ester.

An NMR and conformational investigation of thetrans-syn cyclobutane photodimers of dUpdT

“…Moreover, vicinal couplings in substituted and bridged cyclobutanes 26 -28 (in which the ring pucker and dihedral angles are well defined) indicate that the Karplus relationship can, with reasonable accuracy, predict dihedral angles in this type of system. Data presented here and by others 8,9,[11][12][13][14]18 indicate that trans-syn cyclobutane-type photodimers of a specific configurational class (i.e., trans-syn I or trans-syn II) adopt identical puckering modes (CB Ϫ for 1b-1d and CB ϩ for 2a-2d) and very similar ⌰ 1 values regardless of the cyclobutane ring substitution pattern. This similarity is manifested in a remarkably small (Ͻ0.…”

“…The major ( f 1 ϭ 0.83) form of the thymidine residue of 2b is also 3 T 4 (P 1 ϭ 24.3Њ; ⌽ ϭ 30.0Њ; ␦ ϭ 88.2Њ), while the minor form is 3 4 T. Pseudorotation parameters for 1b and 2b are very similar to those observed in several other trans-syn species (1c-1d and 2a, 2c-2d). 8,9,12,14 The data show remarkably similar results for a particular residue of a given diastereomeric form (i.e., trans-syn I or II) regardless of the cyclobutane ring substitution pattern. In each case, the high degree of conformational purity is evidence of very rigid deoxyribose ring systems.…”

“…8 Likewise, those of 2b are similar to 2a, 12 2c, 14 and 2d. 9 These data are evidence of a lack of C-5 substituent effect on the backbone conformation of these trans-syn photodimers.…”

“…In addition, 3 J U6,T6 values observed for 1b and 2b are similar to those observed for other trans-syn photodimers. 8,9,[11][12][13][14]18 ⌰ 1 values calculated from 3 J U6,T6 , using the Karplus equation, 19 are also consistent with a trans-syn configuration.…”

“…8,9,13,14,18 A general lack of conformational freedom is indicated by the single sugar-base NOE observed for each residue. These results are much different than those obtained for cis-syn photodimers, in which a SYN-ANTI equilibrium exists in either or both of the glycosidic bonds.…”