Received August 12, 1968

During an investigation of the association of nucleic acid derivatives with amino thiols it was observed that the C-5 hydrogen atom in uridine was readily exchangeable in aqueous base (1). In basic D₂O, hydrogen was exchanged for the C-5 deuterium atom.

The exchange occurred when an equimolar mixture of uridine and various bases in D₂O was held at room temperature (ca. 19^o C), while no exchange occurred in a 0.4 M uridine/D₂O solution (pH=5.3) left standing for three months. Using a typical base, 2-mercaptoethylamine (MEA), a new peak appeared after 24 hours in the ¹H NMR spectrum (Fig. 1b), midway between the peaks of the C-6 proton doublet (20. Raising the temperature to ca. 60^o C for 4 days, or allowing the solution to remain at room temperature for 21 days, resulted in the complete disappearance of the C-6 doublet peaks and the new single peak increased in intensity (Fig. 1c). Simultaneously one of the two sets of doublets (H₅, H₁^') at ca.= 5.9 ppm, also disappeared. The addition of MEA also produced a slight shift of 1-3 Hz in the chemical shifts of H₆ and H₁^', while the remainder of the ¹H NMR spectrum was unchanged (not shown). Other bases (NaOH, ethylamine and 2-aminoethanol) also produced shifts in the positions of H₆ and H₁^'.

Uridine-d₅ was isolated in 95% yield. A mass spectrum of uridine-d₅ gave a parent peak at m/e 249*. Uridine-d₄ and uridine gave parent peaks at m/e 248 and m/e 244, respectively. (Mass spectra were obtained on an AEI MS 12 mass spectrometer at 100^o C).

Chemical Shift Data from Figure 1:

	H6	H5	H1'
(a)	477.5, 469.5	358.0, 350.0	357.0, 353.0
(b)	474.5, 466.5, 471.0	359.0, 351.0	358.0, 354.0
(c)	470.0	----	359.0, 355.0

¹H NMR spectra were obtained on a Varian A-60 NMR spectrometer. Solution concentrations varied from 0.1 to 0.4 M in D₂O. Tiers salt, (3-Trimethyl-silyl)-propane sulfonic acid sodium salt, was used as an internal standard, = 0.0 ppm.

In order to determine the scope of this exchange, a number of related nucleophiles and bases were employed and uridine derivatives were studied ( Table 1).

Exchange was found to occur in basic solutions, but not in acidic solutions (MEA-HCl, ethanethiol and 2-mercaptoethanol), hence the reaction was considered to be a base catalyzed exchange. The five compounds studied (Table 1) suggest that a hydrogen atom at C-5 and a carbonyl group at C-4 were required for the exchange to occur. The difficulty of eliminating a methyl carbonium ion accounted for thymidine not incorporating deuterium at C-5.

The probable mechanism for this exchange is the Micheal 1,4 addition across the double bond, followed by a keto-enol tautomeric shift and then elimination of the C-5 hydrogen atom and the nucleophile.

The method of very mild selective deuterium substitution should prove useful to molecular biologists working with DNA or polynucleic acids containing the uridine ring system. Combined with the reverse exchange reaction this would also be particularly helpful for radioactive tritium labeling and subsequent removal.

The author thanks Dr. Peter Rankin, Chemistry Department, Georgetown University, for the mass spectral data, Dr. Daniel L. Klayman of this Institute for helpful discussions, and Mr. James Edson for technical assistance.

References

1. Other reported examples of C-5 tritium/deuterium-hydrogen exchange in these systems include:
(a) R. Shapiro and R. S. Klein, Biochemistry, 6, 3576 (1967)
(b) R. W. Chambers, J. Amer. Chem. Soc., 90, 2192 (1068)
(c) R. M. Fink, J. Biol. Chem., 238, 1764 (1963) and Arch. Biochem. Biophys., 107, 493 (1964)
(d) P. Beak and J. Bonham, J. Amer. Chem. Soc., 87, 3385 (1965)
(e) R. J. Cushley, S. R. Lipsky and J. J. Fox, Tetrahedron Letters, in press.

2. The ¹H NMR spectrum of uridine has been analyzed by:
(a) Varian Specra Catalog, Vol., 2, #535, Varian Associates, Palo Alto, California 94303
(b) C. D. Jardetzky and O. Jardetzky, J. Amer. Chem. Soc., 82, 222 (1960)