Bioconjugate Chem. 1003, 4,362-365
362
Bisulfite Ion-Catalyzed Transamination of Cytosine Residues with a,o-Alkanediamines: The Effect of Chain Length on the Reaction Kinetics John Molander,+Pertti Hurskainen,* Jari Hovinen,? Marjo Lahti,t and Harri Lonnberg'st Department of Chemistry, University of Turku, FIN-20500 Turku, Finland, and Wallac Oy, P.O. Box 10, FIN-20101 Turku, Finland. Received May 17, 1993" ~~
Pseudo-first-order rate constants for the bisulfite ion-catalyzed transamination of cytidine with 1,2ethanediamine, 1,3-propanediamine, l,Cbutanediamine, and l,&hexanediamine have been determined. Hydrolytic deamination has been shown to compete with transamination under acidic conditions, but is of minor importance at pH > 5.3 when the total concentration of diamine is greater than 0.2 mol dm-3. The dependence of the transamination rate on pH and the concentration of diamine and bisulfite ion indicates that the major reaction involves nucleophilic attack of the diamine monocation on the N3 protonated bisulfite adduct of cytidine. The effect of the chain length of the diamine on the rate of transamination is discussed, and the results are compared with those obtained by reacting singlestranded DNA with the same diamines and labeling the transaminated product with a europium chelate.
INTRODUCTION The bisulfite ion-catalyzed displacement of the C4 amino group of cytosine residues with 1,2-ethanediamine is a reaction used widely for the introduction of aliphatic amino groups into polynucleotides (1-6).These amino functions then serve as reaction centers and enable further derivatization of the polynucleotide via attachment of various reporter groups, including immunoreactive groups, fluorescent or spin labels, photochemical cross-linking agents, intercalators, alkylating groups, and catalytically active metal chelates. The transamination reaction proceeds by a rapid initial addition of bisulfite ion to the 5,g-double bond of the cytosine ring, followed by displacement of the 4-amino group with the entering amine (Scheme I) (7). Elimination of the bisulfite ion from the C5-C6 site then completes the reaction sequence. Hydrolytic deamination to uracil derivatives, which also proceeds via the same bisulfite adduct, competes with transamination at low pH (7-1 0 ) . Both the formation of the bisulfite adduct (11-14) and its deamination (13-16) have been studied in detail, whereas quantitative results on the transamination reaction are scanty. The kinetically relevant data are limited to the pH and temperature dependence of the reaction of cytosine, cytidine, and cytidine 5'-monophosphate with amino acids and dipeptides (8,9)and the rate constants of transamination of poly(C) with 1,Zethanediamine and 1,3-propanediamine at a fixed pH and bisulfite concentration (9). Moreover, semiquantitative data on the preparation of several N4-alkyl- and N4-arylcytosine nucleosides by transamination have been reported (7,lO). The present paper is aimed at giving a more clearcut picture of the kinetics of transamination of a cytosine residue with alkanediamines and the competition between transamination and deamination under various conditions. Since for some purposes transamination of polynucleotides with a,w-alkanediamines longer than 1,2-ethanediamine is advantagenous, comparative kinetic studies of transam+
University of Turku.
* WaUac 04.
Abstract published in Advance ACS Abstracts, September 1, 1993.
ination of cytidine with four different alkanediamines, uiz. 1,2-ethanediamine, 1,3-propanediamine, 1,Cbutanediamine, and l,Shexanediamine, have been performed. The results of these studies are compared with those obtained by treating single-stranded DNA with the same diamines in the presence of bisulfite ion and labeling the transaminated product with a europium chelate. EXPERIMENTAL PROCEDURES
Materials. The nucleosides, uiz. cytidine and uridine, were commercial products of Sigma. They were used as received after checking of the purity by HPLC. Of the alkanediamines employed, 1,2-ethanediamine was a product of Fluka (puriss. p.a.1, and 1,3-propanediamine, 1,4butanediamine, and 1,6-hexanediamine were products of Merck (>99 3' % 1. Sodium bisulfite (Mallinckrodt) waa of reagent grade (AnalR). Standard hydrogen chloride and sodium hydroxide solutions of Baker were used to adjust the hydrogen ion concentration of the reaction solutions. The ionic strength was adjusted with sodium chloride (Merck, p.a.). All solutions were prepared in degassed ice cold distilled water to minimize the decomposition of bisulfite ion. Kinetic Measurements. Reactions were initiated by dissolving 0.04 mmol of cytidine in 50 cm3 of a prethermostated reaction mixture. Aliquots (1cm3) withdrawn at appropriate intervals were cooled in an ice bath, and their pH was adjusted to 8 with sodium hydroxide to stop the reaction and to decompose the bisulfite adducts. The compositions of the aliquots were determined by HPLC on a Hypersil ODS column (4.6 X 250 mm, 5 pm) using UV detection (270 nm) and aqueous ammonium acetate (0.025 mol dm-3, flow rate of 1cm3min-l) as eluant. The areas of the cytidine and uridine signals were converted to concentrations by comparing them to those of calibration samples of known concentrations. First-order rate constants for the disappearance of the starting material were calculated from the integrated first-order rate law. These rate constants were bisected to the rate constants of the parallel transamination and deamination reactions with the aid of the mole fraction of uridine formed during the first half-life of the disappearance of cytidine. Since N4alkylated cytidines are known to undergo deamination 0 1993 American Chemical Society
Transamination of Cytosine Residues
Bioconlugate Chem., Vol. 4,
reaction was allowed to proceed for 6 h a t 60 "C. Transaminated DNAs were purified by dialysis against two exchanges of water, concentrated, and precipitated with ethanol. The precipitates were dissolved in aqueous sodium carbonate (100 mm3, 0.05 mol dm-3) containing 5 X lo4 mol dm-3 EDTA. The solutions were pipeted on 1.3 mg of the europium(II1) chelate of 4-[2-(Cisothiocyanatophenyl)ethyll-2,6-bis[N,N-bis(carboxymethyl)aminoethyllpyridine (18) and kept at room temperature overnight. Eu-labeled DNAs were purified on a Sephacryl S-400 column (7 X 460 mm) using a Tris/Tris.HC -buffer (0.010 mol dm-3, pH 8.5) containing sodium chloride (0.1 mol dm-9 as eluent. The extent of labeling was determined by releasing the europium(II1) ion in solution with the fluorescence enhancement solution and measuring the fluorescenceon a time-resolvedfluorometer (1234DELFIA Research fluorometer) against a europium(II1) chloride standard solution (19).The totalamount of DNA in each sample was estimated by measuring the UV absorbance at 260 nm and subtracting the contribution of the DNA bound chelate.
Scheme I
I R' 6
I
R' 2
S
I
I
R' 7
R' 4
No. 5, 1993 383
RESULTS AND DISCUSSION It has been shown previously (14)that protonation of I
kl
R'
1
3
5
less readily than cytidine (13,uridine is formed almost entirely from cytidine and not uia its transamination products. The sulfite ion concentration was determined before and after each kinetic run by iodometric titration. The first-order rate constant for the release of cytidine from its bisulfite adduct was determined spectrophotometrically under the conditions employed in HPLC analyses. Accordingly, 30 mm3 of a solution of cytidine mol dm-3) in a HS03-/S032- buffer (0.125/0.125 (5 X le3 mol dm3) was diluted with 3 cm3 of ammonium acetate solution (0.025 mol dm-3)in a spectrophotometer cell, and the decomposition of the bisulfite adduct was followed at 270 nm. The first-order rate constant was calculated by the method of Guggenheim. Determination of the Stability Constant of the Bisulfite Adduct of Cytidine. The stability constant of the bisulfite adduct of cytidine was determined by measuring the concentration of free cytidine as a function of bisulfite ion concentration in HS03-/S0s2- buffers having a fixed buffer ratio ([HS03-l/[S032-l 4/1,1/1,1/3, or 1/5). The measurements were carried out at 293 K by the HPLC technique described above. The total concentration of cytidine was kept constant (5 X 10-4 mol dm-3) within each series of measurements. Since the retention time of the bisulfite adduct was about 2 min, the adduct did not markedly decompose during the chromatographic separation (half-life 30 min). The stability constant, K, was obtained as the slope of eq 1
-
[Ctdl,,/[Ctdlfree = K[HSO
