1972 1991Synthetic Analogues of Polynucleotides. Part VIII. Analogues of Oligo-nucleotides containing Car boxymet hylt hymid ineBy M. D. Edge, A. Hodgson, A. S. Jones,' and R. T. Walker, Chemistry Department, The University ofAnalogues of oligonucleotides in which the nucleoside units are linked by acetate ester linkages (.O*CH,*CO,*)instead of the phosphodiester linkages of the natural compounds, have been synthesised. The compoundsobtained were thymidinylacetyl- (3'35') -thymidinylacetyl- (3'+5') -thymidinylacetyl- (3'35') -uridine, thymi-dinylacetyl- (3'35') -thymidinylacetyl- (3'35') -thymidinylacetyl- (3'+5') -thymidinylacetyl- (3'35') - uridi ne, andthymidinylacetyl- ( 3 ' 3 5') -thymidinylacetyl- (3'+ 5') -thymidinylacetyl- ( 3 ' 3 5') -thymidinylacetyl- ( 3 ' 3 5') cytidine.The last was also obtained labelled with tritium in the cytidine residue.An oligomer of similar structure to thesecompounds having a molecular weight between 4 x los and 4.7 x lo4 was obtained by copolymerising 3'-0-carboxymethylthymidine with 0.1 mol. equiv. of 2',3'-O-isopropylideneuridine and removing the blocking group.This oligomer and the other two uridine-containing oligomers gave a small but significant hypochromic effect whenmixed with polyadenylic acid in solution.Birmingham, Birmingham B15 2TTWE have previously described,2 the synthesis of ana-logues of trinucleoside diphosphates in which thephosphodiester linkages were replaced by acetate esterlinkages (O*CH,*CO,*). The compounds had the struc-ture ( I ; n = 1, X = uracil, cytosine, adenine, guanine,or hypoxanthine residue).In order to study furtherthe interaction of compounds of this type with poly-nucleotides and their effects in biological systems,oligomers of structure (I; n > 1, X = uracil or cyto-sine residue) have been synthesised. The techniquesused were essentially the same as those previouslyreported; the triphenylmethyl group was used toblock the 5'-hydroxy-group, the isopropylidene groupfor protecting vicinal 2'- and 3'-hydroxy-groups, thedimethylaminomethylene group for the 4-amino-groupof the cytosine residue, and the 2-cyanoethyl groupfor the carboxy-group. The 2-cyanoethyl group wasremoved by treating the ester with 4 mol. equiv. ofpotassium t-butoxide in dimethylformamide at 100"for 10 min instead of 2 mol.equiv. under the sameconditions for 2 h.2 Complete removal of the 2-cyano-ethyl group was achieved with only 5 hydrolysis ofthe internucleoside ester linkages instead of the 12hydrolysis obtained by the previous procedure. Forthe removal of the dimethylaminomethyleme groupfrom the cytosine residues, instead of the usual treatmentwith boiling ethanol,, the protected compound, afterchromatography in chloroform-ethanol was left ad-sorbed on the silica of a t.1.c. plate for 36 h at 20"and then eluted. This was a convenient procedurefor use on a small scale. It was not ascertained whetherthe effect was due to the silica, to the adsorbed ethanol,or to both. Both the triphenylmethyl and the iso-propylidene groups were removed by acidic hydrolysis.It was possible to remove the triphenylmethyl groupPart VII, M.J . Cooper, R. S. Goody, A. S. Jones, J. R.Tittensor, and R. T. Walker, J. Chem. SOC. (C), 1971, 3183.M. D. Edge and A. S. Jones, J . Chem. SOC. (C), 1971, 19331992 J.C.S. Perkin Iselectively in the presence of the isopropylidene group,as exemplified by the synthesis of thymidinylacetyl-(3‘+5’)-(2’,3’-0-isopropylidene)uridine dThd-a-Urd(iPj ;see Scheme for abbreviations.The final products of the syntheses were thymidinyl-in addition to the nucleosides and other intermediatessignals, their absence from their usual position indicatedthat the 5’-O-position was substituted. On the otherhand the terminal 5’-H resonated at the same field as theparent nucleoside and so its signal could be distinguishedacetyl(3‘+5‘)uridine (dThd-a-Urd), which Was used Tr-dThd-a-,-O.CH~,*CN __jl H+ dThd-a-1,-0-CH,.CN (1)DCC0I0(I)Ias a model for the interpretation of n.m.r. spectra,thymidinylacet yl- (3‘45’) -thymidin ylacet yl- (3’45’)-thymidinylacetyl-(3’+5’)-uridine (I; n = 2, X = uracilresidue) (dThd-a-,Urd) , thymidinylacetyl(3’+5’) -thy-midinylacet yl- (3’45’) -th yrnidin ylacet yl- (3’45‘) -thy-midinylacetyl-(3’+5’)-uridine (I; n = 3, X = uracilresidue) (dThd-a-,Urd), and the corresponding cyto-sine compound (I; n = 3, X = cytosine residue)(dThd-a-,Cyd) .The 3’4-5’ internucleoside acetateester linkages were in each case produced by condensationof the appropriate 3’-0-carboxymethyl derivative withthe appropriate derivative with a free 5’-hydroxy-groupby means of dicyclohexylcarbodi-imide in pyridineaccording to the Scheme shown.The compounds were characterised by their U.V.absorption spectra, by alkaline hydrolysis to theirconstituent nucleoside units, by elemental analysis,and by n.m.r.spectra. The presence of the 3’+5’acetate ester linkage was confirmed by the followingtwo observations. (a) The formation of the 3’+5’-acetate linkage caused the 5’-H to resonate at lowerfield than in the parent nucleoside. Although thesesignals could not be readily distinguished from other(1) 4- Tr-dThd-a-OH > Tr-dThd-a,-O*CH,,*CN (2)(2) __t dThd-a-,-O*CH*CN (3)(3) + Tr-dThd-a-OH ,- Tr-dThd-a-,-OfCH,*CN (4)(4) -+ Tr-dThd-a-,-OH (6)DMAM (a) DCCH+DCCBum-I ( 6 ) + Cyd(iP)-(b) SiO, plate + dThd-a-,Cyd(I; n = 3, X = cytosine residue)(4 H+DCCTr-dThd-a-OH + Urd(iP) __jc Tr-dThd-a-Urd(iP) (6)H+ Hf(6) dThd-a-Urd(iP) (7) dThd-a-Urd(7) + Tr-dThd-a-OH __+ Tr-dThd-a-1,-Urd(iP) (8)(8) + dThd-a-1,-Urd(iP) (9)(9) +Tr-dThd-a-OH ,-+ dThd-a-,Urd.DCCH+(a) DCC(I ; 12 = 2, X = uracil residue) (b) H+(a) DCC(9) + Tr-dThd-a-,-OH -* dThd-a-,Urd. L (b) H+(I; n = 3, X = uracil residue)ButO-(2) ___)L Tr-dThd-a-,OH (10)(10) + (7) + Tr-dThd-a-1,-Urd(iP)Abbreviations : Tr = 5’-0-triphenylmethyl, dThd = thymi-dine residue, -a- = 3‘+6’ acetate ester linkage (OCH,CO),Urd = uridine residue, Cyd = cytidine residue, DMAM =dimethylaminomethylene, iP = 2’,3’-O-isopropylidene, DCC =dic yclohex ylcarbodi-imide.SCHEME Synthesis of oligonucleotide analoguesDCCand the ratio of terminal nucleoside to total basesdetermined.With 5’-O-triphenylmethyl derivativesthis was not possible because the 5’-O-triphenylmethylgroup also caused a similar shift in the signal of the5’-H. (b) In a number of the oligomers the presenceof one NH signal per thymine or uracil residue indicatedthat none of the nucleosides was linked by N-3 of thebase. Details of the n.m.r. spectra are given in theTable.The cytosine-containing compound (I; n = 3, X =cytosine residue) was only obtained in a small amountso it was not characterised by n.m.r. Its structurewas established by the fact that it gave 4 mol of 3’-0-carboxymethylthymidine for every 1 mol of cytidineupon alkaline hydrolysis and that the U.V.absorptionspectrum showed that the cytosine residue was notacylated. A sample of this compound labelled withtritium in the cytidine residue was prepared in orde1972 1993to study its hybridisation with mouse satellite deoxy-ribonucleic acid, results of which will be publishedelsewhere .3X = uracil residue)of indefinite chain length (dThd-a-.Urd) was synthesisedby polymerising 3’-O-~arboxymethylthymidine with0.1 mol. equiv. of 2’,3’-0-isopropylideneuridine in thepresence of dicyclohexylcarbodi-imide and removingthe isopropylidene group by acidic hydrolysis. Theresulting oligomer was isolated by dialysing away thelow molecular weight material. Determination of theaverage chain length by measurement of the ratio ofAn oligomer of structure (I;Compound *UridineThymidinedThd-a- Urd (i P)dThd-a-UrddThd-a-,Urd( i P)dThd-a-,UrddThd-a-,UrdTr-dThd-a-OHTr-dThd-a-1,-O-CH,,.CNdThd-a-O*CH,,.CNdThd-a-,-O*CHsa-CNdThd-a-,-O.CH,,CNTr-dThd-a-1,-O*CH,I,.CN fThe oligomers dThd-a-,Urd, dThd-a-,Urd, anddThd-a-,Urd were examined for their interaction withpolyadenylic acid by determining the hypochromicityof the U.V.absorption obtained by mixing solutions ofthe complementary polymers. The results (see Figure)showed that in all three cases there was a small butdefinite hypochromic effect at 260 nm.With dThd-a-1,-Urd the effect was about 5 and with the other twooligomers about 3. Effects of similar magnitudewere also observed at 267 nm, and with dThd-a-,Urdthere was no observable shift in the A,. of the solutionsfrom that expected of a mixture of the two components.H- 6G i i Z s i Z7.85(d), 17.72(d) 7*72(~)7.86(d), 1 7.70(~), 17*74(d), 1 7*751~), 17*50(s), 17*60(d), 1 7-42(~), 37.80(s), 17-58(d), 1 7*42(~), 47-50(s), 17.5-7-4 f7.68(s), 17*70(s), 17-48(s), 17*71(s), 37*49(s), 3tN.m.r. data a (p.p.m.), number of protonsH-1‘Me ofRemarks r H-5 Thymidine Uridine H-5’ CH2*CN thymidine Me NH5.60(d), 1 5*77(d), 1 3*60(~), 25.67(d), 1 6.17(t) 1 542(d), 1 3.60br, 2 1.79(s), a 1.50(~), 3 H-6 for uridine +1.30(s), 3 thymidine = 2H5.61(d), 1 6.17(t), 1 5.78(d), 1 3.65br 1*77(s), 3 HCOnH, S 8.3(s), 1 H5*70(d), 1 6-20(t), 2 5*84(d), 1 Indistinct 1*80(s), 6 1-50(s), 31-30(s), 35.65(d) 6-15(m), 3 5*80(d) 3.60(m), 2 1.80(~), 9 11-5(s), 4 Uridine H-5 +5.65(d) 6-13(t), 4 5.i3(d) 3.50br, 2 1*79(s), 12 11.3(s), 5 Ditto6*25(t), 1 3.65(s), 2 1-81(s), H ll*3(s), 1H-1’ = 2H6*20(t), 1 1*45(s), 36-20(t), 2 2.90(t), 2 1.76(~), 3 11.5(s), 21*47(s), 36-23(t), 1 3.60br.2 2*90(t), 2 1.78(s), 36*20(t), 2 3*60br, 2 2.92(t), 2 1.79(s), 66.28(t), 3 3.60br, 2 2.92(t), 2 140(s),96.18(t), 4 2438(t), 2 1.76(m), 6 ll*4(s), 41*41(m), 3* For abbreviations see Scheme. t Merges with Ph,C protons. $ At 60 MHz.3’-O-~arboxymethylthymidine to uridine in an alkalinehydrolysate gave a value of 168 nucleoside units (2.e.mol.wt. 4-7 x lo4). This must be considered to be a0.10 0.30 0-50 0-70 0.90Molar fraction of adeninenucleotide in mixtureOptical density at 260 nm of mixtures of polyadenylic acid andcarboxymethylthymidine-containing analogues of oligo-nucleotides. Conditions as in text. A, dThd-a-,Urd ;B, dThd-a-,Urd ; C , dThd-a-,Urd0.280maximum value, however, because some chains probablydid not terminate in uridine residues. The molecularweight of the smallest molecules was probably notless than about 4-43 x lo3 because molecules of aboutthis molecular weight are able to diffuse through thedialysis membranes used.The dThd-a-,Urd was only partly soluble in aqueoussolutions, so for this experiment it was dissolved indimethylformamide and the solution was diluted withaqueous salt solution to give a dimethylformamideconcentration of 10.This procedure gave a moresatisfactory solution but the dimethylformamide mayhave reduced the hypochromicity. This result shouldbe compared with those obtained previously with anoligomer of structure dThd-a-1, in which a hypo-chromicity of 8 was obtained in the absence of di-methylformamide but at lower ionic strength.2 Theseresults confirm, therefore, that this type of oligonucleo-tide analogue can interact with polynucleotides insolution. In the case of dThd-a-,Urd and dThd-a-,-Urd the maximum hypochromicity appears to occurwhen the molar ratio of thymidine to adenosine residuesis approximately 1 : 1, whereas with dThd-a-,Urdthe maximum hypochromicity occurs at a much higherproportion of thymidine residues (2-3 : 1).Thissuggests that the shorter oligomers may be forming1 : 1 complexes whereas the longer oligomer is forming a2 : 1 complex, although this conclusion requires con-firmation. Such a conclusion is in accord with thebehaviour of the complementary oligonucleotides ;it has been shown that polyadenylic acid and a thymidinehexanucleotide form a 1 : 1 complex under conditionswhere polyadenylic acid and thymidine decanucleotideform a 1 : 2 complex.*3 P. M. B. Walker, to be published.R. Naylor and P. T. Gilham, Biochemistry, 1967, 5, 27221994 J.C.S. Perkin IThe oligomer dThd-a-,Urd showed some evidence of Carboxymethyl-5’-O-triphenylmethylthymidine (0.5 mmol)instability in the freeze-dried state (containing 20 and the foregoing compound (0.5 mmol) were con-water) when stored at -400 for 3 4 weeks.From densed together. Column chromatography of the productsmeasurements on the stability of a similar compound On gel was carried out by applying the compoundin solution,5 details of which will be published later, it dissolved in chloroform to the column and then elutingwith ethyl acetate followed by acetone-ethyl acetateat pH 7 is about 26 h. tained as a chromatographically homogeneous white powderthat the Of these oligomers at 370 (1 : 4). The required compound (350 mg, 52) was ob-EXPERIMENTALFor t.l.c., silica gel powder, MN Silica Gel UV254(Machery, Nagel lk Co., Germany), and for column chromato-graphy, Kieselgel, 0-05-0- 2 mm (70-325 Mesh ASTM),type 7734 (Merck, Germany) were used.General Techniques.-(a) Condensations.These werecarried out between the pyridinium salt of the appropriatecarboxylic acid (prepared from the sodium salt by use ofthe pyridinium form of Zeo Karb 225) and the appro-priate 5’-hydroxy-compound in dry pyridine a t a concen-tration of 0.5-1-0 mmol ml-I in the presence of dicyclo-hexylcarbodi-imide (4-5 mol. equiv.) a t about 20” forfor 18-24 h. Most of the excess of dicyclohexylcarbodi-imide was removed by extraction with cyclohexane orsometimes converted into dicyclohexylurea by additionof water. Dicyclohexylurea formed in this way or duringthe reaction was mainly removed by taking advantageof its low solubility in cold acetone.In most of thecases the required products were isolated by silica gelcolumn chromatography with a ratio of silica gel to com-pound of about 100: 1 by weight. The required productswere usually obtained as solids by precipitation fromacetone, or other suitable solvent, with light petroleum.This was carriedout by hydrolysis with acetic acid-water (4: 1) a t 100”for 15 min, or with formic acid-water (2 : 1) a t 20” for 4 h,or with formic acid (98) at 20” for 5-10 min. The lastprocedure caused selective removal of the triphenyl-methyl group in the presence of an isopropylidene group.This was carriedout either as previously described or by the use of 4 mol.equiv.of potassium t-butoxide in dimethylformamide a t100’ for 10 min.Determination of the triphenylmethyl group. This wascarried out by a procedure similar to but not identicalwith that described by Duffield and Nussbaum.s Thecompound was dissolved in a mixture of glacial aceticacid and conc. sulphuric acid (2: 1 v/v) and the opticaldensity was read a t 436 nm. Triphenylmethyl chloridewas used as standard. A 2.64 x mM-solution gave anoptical density of 1-00 in a 1 cm cell.2-Cyanoethyl Thymidinylacetyl- (3’+5‘) -thymidin- 3’-yl-acetate.-2-Cyanoethyl 5’-O-triphenylmethylthymidinyl-acetyl-( 3’+5’)-thymidin-3’-ylacetate (1.0 g) 2 wastreated with acetic acid to remove the triphenylmethylgroup. The product was isolated by silica gel columnchromatography, with ethyl acetate to elute triphenyl-methanol and acetone-ethyl acetate (4: 1) to elute theProduct (530 mg), which was obtained as a chromato-graphically pure white powder (Found: C, 51.0; H, 5.1;N, 11.2.C,,H,,N50,, requires C, 51.1; H, 5.2; N, ll-O),Amx. (H20) 267 nm (E 18.6 x lo3).2-Cyanoethyl 5’-0- Triphenylmethylthymidinylacetyl- (3’+5’)-thymidinylacetyl- (3’+5’)-thymidin-3’-ylacetate.- 3‘-0-(b) Removal of triphenylmethyl groups.(c) Removal of 2-cyanoethyl groups.A. S. Jones and M. MacCoss, unpublished results.(Found: C, 59.7; H, 5.6; N, 8.2. C5f61N,O19 requiresC, 60.0; H, 5-8; N, 8-5), A,, (EtOH) 266 nm ( E 25-5 xlo3). Alkaline hydrolysis gave 3’-O-~arboxymethylthymi-dine and 3’-O-carboxymethyl-5‘-O-triphenylmethylthymi-dine in the molar ratio of 2.1 : 1.The compound contained1 trityl group to 3-0 thymidine residues (the latter beingmeasured by U.V. absorption.2-Cyanoethyl Thymidinylacetyl- (3’+5’) -thymidinylacetyl-(3’+5’)-thymidin-3’-ylacetate.-The foregoing compound(200 mg) was treated with 98 formic acid to removethe triphenylmethyl group. The products (applied inpyridine solution) were fractionated on a silica gel column,which was eluted with ethyl acetate and then with acetone.The pvoduct (150 mg) was obtained as a chromatographicallyhomogeneous, white powder (150 mg) (Found: C, 51.2;H, 4.7; N, 10.4. C3,H17N,0,, requires C, 51.0; H, 5.1;N, 10*7y0), A,, (EtOH) 267 nm ( E 25-6 x lo3).Thymidinylacetyl- (3’+5’) - (2’, 3‘-O-isopropylidene) uridine.-3’-O-Carboxymethyl-5’-O-triphenylmethylthymidine (7g) and 2’, 3’-O-isopropylideneuridine (3-4 g) were con-densed together and the product was treated with 98formic acid to remove the triphenylmethyl group.Themixture, which consisted of a major component (whichgave 3’-O-~arboxymethylthymidine and 2’, 3’-O-isopro-pylideneuridine upon alkaline hydrolysis) and three minorcomponents, was fractionated on a silica gel columnelution with ethanol-chloroform (1 : 19 then 2 : 23).The product (2.6 g) was isolated in pure form as a whitesolid (Found: C, 51.2; H, 5.1; N, 10.1. C2,H,,,N4O,,requires C, 50.9; H, 5.3; N, 9-9yo), A,, (H,O) 264, Amin.232; A,,,. (alkali) 264, Amin. 246 nm.Thymidinylacetyl-(3’+5’)+widine.-The foregoing com-pound (200 mg) was treated with formic acid-water (2 : 1)to remove the isopropylidene groups.Removal of thesolvents left pure thymidinylacetyl-( 3’+5’)-uridine(140 mg) containing one mol. equiv. of formic acid (Found:C, 46.3; H, 4.6; N, 10.0. C21H2,N,0,2,HC02H requiresC, 46.2; H, 4.6; N, 9.8) (n.m.r. indicated the presence offormic acid). Alkaline hydrolysis gave 3’-O-carboxy-methylthymidine and uridine.Thyrnidinylacetyl- (3’+5’) -thymidinylacetyE (3’+5’) - (2’, 3’-O-isopropy1idine)uridine.-The foregoing compound (1- 13 g)was condensed with 3’-O-carboxymethyl-5’-U-triphenyl-methylthymidine (1.16 g). The product was isolatedby silica gel column chromatography with ethanol-chloro-form (2 : 23) as eluant. The triphenylmethyl groupwas then removed with 98 formic acid and the mixturewas fractionated on a silica gel column applied inpyridine solution and eluted with chloroform followedby ethanol-chloroform (3 : 22) to give the product (400mg) in pure form (Found: C, 50.6; H, 5.1; N, 9.9.C38H44N8018 requires C, 50.9; H, 5-2; N, 9.9).2-Cyanoethyl 5‘43- Triphenylrnethylthymidinylacetyl-(3’+5’) -thymidinylacetyZ- (3‘+5’) -thymidinylacetyl- (3’+5’) -thymidin-3‘ylacetate.-2-Cyanoethyl thymidinylacetyl-A.MI Duffield and A. L. Nussbaum, Analyt. Biochem., 7 M. H. Halford and A. S. Jones, J . Chem. Soc. (C), 1968,1964, 7, 254. 26671972 1995(3’+5’)-thymidinylacetyl-( 3’+5’)-thymidin-3’-ylacetate (48mg) was condensed with 3’-O-carboxymethyl-5‘-O-triphenyl-methylthymidine (30 mg).The mixture was fractionatedby silica gel column chromatography ethyl acetate thenacetone-ethyl acetate (1 : 1) as eluant. The productwas isolated as a white solid (40 mg) which contained1 triphenylmethyl group to every 3.95 thymine residues.An analytically pure sample (13 yield) was obtainedby t.1.c. ethanol-chloroform (2 : 23) followed bysilica column chromatography Found (after drying invacuo at 20”) : C, 56.0; H, 5.6; N, 8.4. C,,H7,N9O,,,3H,Orequires C, 56.0; H, 5.6; N, 8.4, Amax (CHCI,) 267 nm(E 34,650), 1 triphenylmethyl group to 3.95 thymine resi-dues. Alkaline hydrolysis gave 3’-O-carboxymethyl-thymidine and 3’-O-carboxymethyl-5’-O-triphenylmethyl-thymidine in a molar ratio of 2.9 : 1.Thymidinylacetyl- (3’+5’) -thymidinylacetyl- (3’+5’) -thy-~idinylacetyl-(3’+5’)-uradine.- 3’-O-Carboxymethyl-5’-0-triphenylniethylthymidine (54 mg) was condensed withthymidinylacetyl- (3’+5’)-thymidinylacetyl(3’+5’)-(2’, 3’-O-isopropylidene) uridine (85 mg) .The product wasisolated by silica gel column chromatography elutionwith ethanol-chloroform (2 : 23) and the protectinggroups were removed with formic acid-water (2 : 1).Triphenylmethanol was extracted with chloroformand the product (20 mg) was precipitated with ethylacetate from solution in dimethylformamide. Alkalinehydrolysis of this material gave 3’-O-~arboxymethylthymi-dine and uridine in a molar ratio of 2.9 : 1. It wascharacterised by 1i.m.r. spectroscopy (see Table).G O - Triphenylrnethylthymidinylacetyl- (3’+5’) -thymidinyl-acety l- ( 3’+5’) -thymidinyEacetyl- (3’+5’) -thymidinylacetyl-(3‘+5‘) - (2’, 3’-O-isopropylidene) uYidine.- 2-Cyanoethyl 5’-O-triphenylmethylthymidinylacetyl-(3’+5’) -thymidinyl-acetyl-( 3’+5’)-thymidin-3’-ylacetate (80 mg, 0.069 mmol)was treated with potassium butoxide (31.2 mg, 0.276 mmol)in dry dimethylformamide (20 ml) at 100” for 10 min to re-move the 2-cyanoethyl group.The resulting acid was con-densed with thymidinylacetyl-( 3’+5’)-(2’, 3’-O-isopropyl-idene)uridine (40 mg) in the usual way. The mixturewas fractionated by column chromatography on silicagel elution with chloroform, then ethanol-chloroform(1 : 49 then 1 : 19). The product (47 mg) was isolated as awhite powder Found (after drying in vacuo a t room tem-perature): C, 54.1; H, 5.3; N, 7.8.C,,H,,029Nl,,6H,0requires C, 54.3; H, 5.6; N, 8-0. Alkaline hydrolysisgave 3’-O-carboxymethyl-5’-O-triphenylmethylthymidine,3’-O-~arboxymethylthymidine, and 2’, 3’-O-isopropylidene-uridine in the molar ratio of 1.1 : 3.1 : 1.0.Thymidinylacetyl- ( 3’+5’) -thymidiny lacetyl- (3’+5’) -thy-Ynidinylacetyl- (3’+5’) -thymidinyZacetyl- (3’+5’) -uridine.-The foregoing compound (40 mg) was treated with formicacid-water (2 : 1) to remove the triphenylmethyl andisopropylidene groups. The mixture was extracted withchloroform to remove triphenylmethanol and with etherto remove formic acid, and the product (20 mg) was ob-tained chromatographically pure as a white solid. Alkalinehydrolysis gave 3’-0-~arboxymethylthymidine and uridinein a molar ratio of 3.8 : 1.This compound was character-ised by n.m.r. spectroscopy.An identical product was obtained by the condensationof 5’-O-triphenylmethylthymidinylacetyl- (3’+5’) -thymidin-3’-ylacetic acid with thymidinylacetyl-( 3’+5’) -thymi-dinyl- ( 3’+5’) - (2’, 3’-O-isopropylidene) uridine followed bytreatment with formic acid-water (2 : 1).4-N-Dimethylaminomethylene-2’, 3’-0-isopropyZidenecyti-dine.-This was obtained by a procedure described byZemlicka and Holy for the synthesis of similar compounds.2’, 3’-O-Isopropylidenecytidine (0.20 g) was dissolved indry dimethylformamide (3.5 ml) and dimethylformamidedimethyl acetal (0-36 ml) was added. The solution wasset aside a t room temperature for 18 h, then evaporatedunder reduced pressure.The residue was crystallised byaddition of ether to give the product (0.21 g), which wasrecrystallised from chloroform and light petroleum ; m.p.179-180°, Am= (EtOH) 316 (c 3.6 x lo4), Amin. 245 nm(E 1750) (Found : C, 53-6; H , 6.6; N, 16.2. Cl,H,,-N,O, requires C, 53.3; H, 6.5; N, 16-6y0), homogeneouson t.1.c. in ethanol-chloroform (2 : 23).Thymidinylacetyl- (3’+5’) -thymidinylacetyl- ( 3’+5’) -thy-midinylacetyl-( 3’+5‘)-thymidinylacetyl-(3’+5’) -cytidine.2-Cyanoethyl 5’-O-triphenylmethylthymidinyIacetyl-(3’+5’)-thymidinylacetyl- (3’+5’)-thymidinylacetyl-( 3’+5‘)-thymidin-3’-ylacetate (8 mg) was treated with potassiumt-butoxide in the usual way to remove the 2-cyanoethylgroup and then condensed with 4-N-dimethylamino-methylene-2’, 3’-O-isopropylidenecytidine ( 1.5 mg) in theusual way.After removal of excess of dicyclohexyl-carbodi-imide and dicyclohexylurea the mixture wassubjected t o t.1.c. in ethanol-chloroform (2 : 23). Thecompound was left adsorbed to the silica for 36 h at20’ then eluted with ethanol-chloroform (1 : 1). Theeluate was evaporated to dryness to give a white solidwhich had no absorption a t 310 nm. (This showed thatthe dimethylaminomethylene group had been removed bythe prolonged contact with the silica.) The other protect-ing groups were removed by treatment with formic acid-water (2 : 1) a t room temperature for 16 h and the product,dissolved in acetic acid-water (1 : 4), subjected to paperelectrophoresis for 9 kVh in O.O5~-acetate (pH 3.6).Twomajor components and two minor components weredetected. The former were eluted and again subjectedto paper electrophoresis. The faster-running componentremained homogeneous. This was eluted with aceticacid-water and salts were removed by dissolving in drydimethylformamide, filtering, and precipitating the productfrom solution with ethanol and ether; Amax. AcOH-H,O (1 : a) 268-269, Amin. 243 nm. Alkaline hydrolysisgave 3’-O-~arboxymethylthymidine and cytidine in themolar ratio of 4.1 : 1.A sample of this compound labelled with tritium in thecytidine residue was obtained by carrying out the synthesiswith labelled 4-N-dimethylaminomethylene-2’, 3’-O-iso-propylidenecytidine.The labelled compound was identicalwith the unlabelled compound with regard to chromato-graphic and electrophoretic mobility and U.V. absorptionspectrum; ca. 1 mg of material (specific activity of 236mCi mmol-l) was obtained.Polymerisation of 3‘-0-CarboxymethyZthyrnidine in thePresence of 2’,3’-O-Iso~ropylideneuridine.-3’-O-Carboxy-methylthymidine pyridinium salt (0.28 mmol) and2’, 3’-O-isopropylideneuridine (0.028 mmol) were exhaus-tively dried and dissolved in dry pyridine (1.5 ml); di-cyclohexycarbodi-imide (500 mg) was added and themixture was kept a t 20” for 16 h. The pyridine wasevaporated off in vacuo to leave an oil which was dissolvedin 98 formic acid (6 ml); water (3 ml) was added andthe mixture was kept a t 20’ for 4 h, filtered free fromJ. Zemlicka and A. Holy, Call. Czech. Chern. Comm.. 1967,32, 3169J.C.S. Perkin Idicyclohexylurea, and evaporated to dryness. The residuewas dissolved in dimethylformamide (20 ml) and dialysedagainst dimethylformamide (2 x 200 ml for 4 h ; 1 x 1-5 1for 16 h) at 20" and then against water (2 x 500 ml for4 h; 1 x 1 1 for 4 h) at 20'. The suspension insidethe dialysis bag was then freeze-dried to give the oligomer(14 mg, 120/,). Hydrolysis in alkali gave 3'-O-carboxy-methylthymidine and uridine in the molar ratio of 167 : 1.The product did not migrate on paper electrophoresis a tpH 6-8.Interaction with Polyadenylzc Acid.-This was studiedby measurement of the U.V. absorption of solutions con-taining the oligomers and polyadenylic acid (mol. wt.> l O 5 ; PL Biochemicals Inc.) in various proportions bythe standard procedures already used.', The solvent was0.3hl-sodium chloride, 0-O1M-glycylglycine (pH 6.3). Inthe case of dThd-a-,Urd the solution contained 10(v/v) of dimethylformamide and O.O3~-sodium citrate(instead of glycylglycine). The solutions were kept a t4" for 18 h, then allowed to warm up to 20", and theoptical density a t 260 and a t 267 nm was measured. Theresults (at 260 nm) are shown in the Figure.N.m.r. Spectra-The spectra (100 MHz) (Table) wererecorded for solutions in 2H,dimethyl sulphoxide withtetramethylsilane as internal reference.We thank the Medical Research Council and the CancerResearch Campaign for research grants and Messrs. ArthurGuinness, Son and Co. Ltd., for a research studentship(to M. D. E.).2/365 Received, 18th February, 197210 M. G. Boulton, A. S. Jones, and R. T. Walker, Biochim.Biophys. A d a , 1971, 248, 197
展开▼
