NUCLEIC ACIDS COMPONENTS AND THEIR ANALOGUES. LXVI. *
`HYDROLYSIS OF 5-AZACYTIDINE AND ITS CONNECTION
`WITH BIOLOGICAL ACTIVITY
`
`p. PrrHOV A, A. Pf SKALA, J. PITHA and F. SORM
`.Institute of Organic Chemistry and Biochemistry, Czechoslovak Academy of Sciences, Prague
`
`Received November 20th, 1964
`
`Hydrolysis of 5-azacytidine ([) in neutral and basic media affords the
`following products: 1-~-o-ribofuranosyl-3-guanylurea (III), guanidine,
`cx-o-ribofuro[l',2':4,5]-2-oxazolidone (JV) and o-ribose. As the intermediate,
`the N-formyl derivative of the compound III has been determined spectra(cid:173)
`photometrically. In acidic media, the hydrolysis affords 5-azacytosine (V) , 5-aza(cid:173)
`uracil ( VI) and o-ribose. In contrast to 5-azacytidine (/), the isolated products
`show a decreased biological activity. As shown by experiments and quantum
`chemical calculations, the biological activity of the compound I is probably
`due to the incorporation of a 5-azacytosine derivative into nucleic acids
`followed by a hydrolytic fission of the triazine ring.
`
`In the previous paper 1 of this Series, a study has been presented about structure
`and tautomerism of 5-azacytidine (I), a compound of remarkable cancerostatic,
`bacteriostatic2 and mutagenic properties. In connection with practical applications
`of 5-azacytidine, it was of interest to study its stability and the course of its hydro(cid:173)
`lytical destruction. Furthermore, two additional aspects have been taken into ac(cid:173)
`count: firstly, whether the biological activity is not due to some hydrolysis inter(cid:173)
`mediate (as found in this Institute in the case of 5-azauracil4), and, secondly, whether
`the mutagenic properties are not caused by an eventual incorporation of 5-azacyto(cid:173)
`sine derivatives into nucleic acids followed by a rapid hydrolysis of the instable
`heterocyclic nucleus. Finally, we have tried to compare the experimental results with
`results of simple quantum chemical calculations in the Ruckel approximation and
`to use this method for determination of properties by extrapolation in those cases
`where the experimental determination would be difficult.
`
`Experimental
`
`Melting points were taken on a heated microscope stage and are corrected. The analytical
`samples were dried for ten hours at 0·3 mm Hg and room tern perature.
`The nuclear magnetic resonance spectra were measured in deuterochloroform on the apparatus
`40 Mc of Czechoslovak origin equipped with superstabilisation. Calibration was performed
`
`* Part LXV.: This Journal 30, 2480 (1965).
`
`Vol. 30 (1965)
`
`CELGENE 2027
`APOTEX v. CELGENE
`IPR2023-00512
`
`2801
`
`

`

`Pit'hova, Piskala, Pit'ha, Sorm:
`
`by the usual technique of side bands (inner standard TMS). Concentration: 100 mg per 0·5 ml
`of deuterochloroform. Infrared spectra were measured on a spectrophotometer Zeiss (Jena)
`model UR-10 as usual. Ultraviolet spectra were taken by the technique described in the previous
`paper1 in a 0·lM borate buffer at 37°C. The quantum chemical calculations were performed
`with the use of parameters recommended in the literature5 •6 .
`Inhibition of the growth of Escherichia coli w~s determined as described previously 7 •
`Descending paper chromatography was performed on paper Whatman No 1 without any pre(cid:173)
`vious saturation in the following solvent systems: (S1 ) isobutyric acid-water-aqueous ammonia
`(66: 33 : 1 ·5), (S2 ) 1-butanol-acetic acid-water (5 : 2: 3), and (S 3 1-butanol-ethanol-water
`(40: 11 : 19). The spots were detected (D 1 ) in ultraviolet light, (D2 ) according to Reindel and
`Hoppe8 and (D3 ) with the sodium hydroxide-sodium nitroprusside-potassium ferricyanide
`reagent (by the latter reagent, the guanidine9 derivatives are detected). The vicinal cis-diol
`system of the ribosyl derivatives was detected with potassium periodate- benzidine10 (D4 ) .
`Some detections were performed with the use of radioactivity (D5). Paper electrophoresis was
`carried out on paper Whatman No 1 according to Markham and Smith in a 0·05M borate buffer
`(pH 9·25) for the period of thirty minutes at 30 V /cm.
`
`Isolation of the Hydrolysis Products
`
`A) A solution of 5-azacytidine (/; 0·122 g; 0·5 millimole) in lM aqueous ammonia (10 ml)
`was allowed to stand at 37°C for three hours and taken down under diminished pressure in a rota(cid:173)
`tory evaporator (bath temperature 30°C). The residual sirup was repeatedly coevaporated with
`three 10 ml portions of methanol, dissolved in methanol (2 ml) and precipitated with a solution
`of picric acid (0· 120 g) in ethanol (2 ml) to deposit immediately the picrate of _ribosylguanylurea
`Ill in the form of needles, m.p. 172- 174°C (decomposition). The next day, the needles were
`collected and washed with ethanol; yield, 0·163 g (70%). For C 13H 17N 7 0 12 (463·3) calculated:
`33·70% C, 3·71% H, 21·17% N; found: 33·96% C, 3·55~{ H, 21·21% N. Paper electrophoresis
`of the crude reaction mixture: anodic mobility (2·7 cm) of the compound III was determined
`by detections D 2 , D 3 and D 4 (no other compounds were found by these detections; l-~-o-ribo(cid:173)
`pyranosyl-3-guanylurea 11 did not move at all under analogous conditions). The compound III
`was obtained also by keeping 5-azacytidine (/; 0· 122 g; 0·5 milimole) with saturated methanolic
`ammonia (10 ml) and water (0·2 ml) at room temperature for five days and working up as above;
`yield, 0·095 g (41%) of the picrate of compound Ill, m.p. 172-174°C (decomposition).
`B) 5-Azacytidine (/; 0·244 g; 1 millimole) was heated to 60°C in 10 ml of a borate buffer
`(pH 9·28) for five hours. The solution was taken down under diminished pressure in a rotatory
`evaporator, the residue dissolved in a small amount of water and the solution chromatographed
`in the solvent system S3 on three sheets of paper Whatman No 3. By this technique, the mixture
`was separated into two bands (Rp values 0·09 and 0·31). The first broad band showed a weak
`absorption in ultraviolet light and co.uld be detected by procedures D 2 , D 4 and D 3 (violet colora(cid:173)
`tion). This band was not homogeneous as shown by rechromatography (solvent system S 1)
`-of the concentrated eluate and detection of the compound I (Rp value 0·58, a weak spot, detection
`procedure D 1) and III (Rp value 0·76, an intensive spot, detection procedures D 2 , D 3 and D 4 ).
`The second band was narrow, did not absorb in ultraviolet light and could be detected by procedu(cid:173)
`res D 2 (orange coloration) and D 3 . Rechromatography of the concentrated eluate of this band
`revealed the presence of guanidine (Rp value 0·85, detection procedures D 2 and D 3) and o-ribose
`(Rp value 0-55, detection procedure D 4 ). Moreover, guanidine was identified in solvent systems S2
`and S3 (Rp values 0·61 and 0·25, resp.) Precipitation with a solution of picric acid (0·050 g)
`in ethanol (1 ml) and keeping overnight afforded guanidine picrate which on recrystallisation
`from water melted at 333°C.(decomposition) without depression on admixture with an authentic
`specimen. Yield, 0·018 g(6%). For C 7 H 8 N 6 0 7 (288·2) calculated: 29-17% C, 2·80% H, 29·17~~ N;
`found: 29·45% C, 3·05% H, 28·94% N.
`
`2802
`
`Collection Czechoslov. Chem. Commun.
`
`

`

`_
`
`Nucleic Acids Components and their Analogues. LXVJ.
`
`C) A solution of 5-azacytidine (/; 2·44 g; 0·01 mole) in water (100 ml) was refluxed for five
`hours and then taken down under diminished pressure in a rotatory evaporator. The residue
`was dissolved in methanol (100 ml) and treated with 50 g of moist Dowex 50 W(H+) ion exchange
`resin, previously washed with water and methanol. After some minutes the resin was filtered
`off and washed with methanol. The filtrates were refiltered with charcoal (0·5 g) and evaporated
`under diminished pressure. The residual thick sirup was dissolved in boiling ethanol and the
`solution allowed to stand overnight. The crystals were collected (1 ·02 g) and the mother liquors
`concentrated to deposit additional 0· 125 g of the same product. Total yield of the compound IV
`1·245 g (65%); m.p. 169-170°C. For C6 H 9N05 (175·1) calculated 41 ·14% C, 5·18% H, 8·00% N;
`found: 41·12% C, 5·20% H, 8·18% N . In the carbonyl region of the infrared spectrum measured
`in a nujol suspension, a broad band at 1 727 cm - l with shoulders (1697 and 1 747 cm - l) was
`registered; no absorption band occurs in the amide II region. By the action of potassium periodate
`in a phosphate buffer (pH 6·8) at room temperature no consumption of the reagent was recorded
`in the course of eight hours. The compound reduced the Benedict solution at the boil. A short
`boiling with lM aqueous potassium hydroxide (2-3 minutes) led to a deep decomposition (black(cid:173)
`ening). Chromatography of the hydrolysate in the solvent system S3 revealed the presence of D-ri(cid:173)
`bose (RF value 0·20, detection procedure D 4 ). D-Ribose was found also in the mother liquors.
`(solvent system S3 , RF value 0·20, detection procedure D 4 ) after crystallisation of the compound
`JV. The collected ion exchange resin (vide supra) was washed with water and then eluted in a column
`with 3N-HC1 (300 ml). The eluate was evaporated under diminished pressure and the residue
`dried over sulfuric acid and potassium hydroxide to afford crude crystalline guanidine hydro(cid:173)
`chloride. The crystals were dissolved in water (20 ml), the solution was filtered and poured
`into a stirred solution of potassium picrate (2·68 g) in hot water (60 ml) to afford guanidine
`picrate. The next day, the precipitate was collected and washed with water, ethanol and ether;
`yield, 2·23 g (77%). The recrystallised sample melted at 333°C (water; decomposition) without
`depression on admixture with an authentic specimen. For C 7 H 8 N 6 O7 (288·2) calculated:
`29·17% C, 2·80% H, 29·17% N; found: 29·32% C, 2·93% H, 29·08% N . The above procedure
`was used also for the hydrolysis of 1-~-D-ribopyranosyl-5-azacytosine (0·244 g; 1 millimole)
`to give 0·096 g (55%) of the compound JV, m .p. 169-170°C, and 0·225 g (78%) of guanidine
`picrate. The mixture contained also n-ribose, as shown by chromatography (Rp value 0·20;
`detection procedure D 4) in the solvent system S3 •
`
`Analysis of the Hydrolytical Course by means of 5-Azacytidine-14C
`The labeled compound was dissolved (concentration 10- 4 M) in a borate buffer1 (0·lM), the
`solution allowed to stand at 37°C for the period given in Fig. 1 and the resulting mixture separated
`by paper chromatography in the solvent system S1 . The radioactive components of the mixture
`were determined either directly on the chromatographical paper or in the eluates. The direct
`measurement was performed in an automatic apparatus with the 41t geometry using the Geiger(cid:173)
`Miiller tube with a thin end-window. D-Ribose was determined in the eluates of radioactive
`zones by means of the orcinol reaction. The fission products (Table II) were identified by com(cid:173)
`parison with compounds obtained preparatively. In the quantitative determination, the radio(cid:173)
`active zones were eluted and the eluates evaporated; radioactivity of the residues was measured
`on a 21t windowless proportional flow counter in an infinitely thin layer. As shown by repeated
`chromatography of eluates and reelution, no considerable hydrolysis occurred during the work-up:
`thus, e.g., in the case of ribofuranosylguanylurea. Ill and guanidine, 90 and 100% recovery, resp.,
`was observed.
`
`3' ,5' -Di-O-acetyl-o:-o-ribofuro[l ',2' : 4,5]-2-oxazolidone (I Va)
`
`A stirred and ice-cooled solution of the compound IV (0·350 g; 2 millimole) in dry pyridine
`(5 ml) was treated with acetic anhydride (0·82 g). The mixture was allowed to stand in a stop-
`
`Vol. 30 (1965)
`
`2803
`
`

`

`Pi(hova, Piskala, Pit'ha, Sorm:
`
`0·8
`
`0·6
`
`0·4
`
`0-2
`
`0
`
`500
`
`1000
`
`1500
`
`2000
`
`Fig. 1
`Time Dependence of 5-Azacytidine-4- 14C Fission Products as Determined by Paper Chromato(cid:173)
`graphy
`A 5-Azacytidine-4-14C; U 1-Ribofuranosyl-3-guanylurea; G Guanidine; a borate buffer of the ionic
`strength 0·l-0·15; 1 pH 6·6; 2 pH 7·2; 3 pH 7·8; 4 pH 8·4; 5 water.
`
`pered flask at room temperature for 24 hours and taken down under diminished pressure in a rota(cid:173)
`tory evaporator (bath temperature 40°C). The residue was coevaporated with four 10 ml portions
`of methanol and then dissolved in chloroform (20 ml). The chloroform solution was washed
`subsequently with water, aqueous sodium hydrogen sulfate, water, a saturated solution of sodium
`hydrogen carbonate, dried over anhydrous sodium sulfate and evaporated under diminished
`pressure. The residual sirup was dissolved in benzene (5 ml) and the solution was precipitated
`with light petroleum (20 ml). The sirupous precipitate failed to crystallize. It was, therefore,
`dissolved in ether (50 ml), the solution concentrated and the concentrate treated with light
`petroleum (30 ml). The supernatant was decanted and the residue dried at 0· 1 mm Hg to afford
`0·385 g (74%) of a white foam. For C 10H 13NO 7 (259·2)-calculated: 46-34% C, 5·06% H, 5·40% N;
`found: 46·67% C, 5·20% H, 5·30% N. Infrared spectrum (chloroform): v(CO) 1 749 cm - l
`(five(cid:173)
`membered cyclic urethane), v(NH) 3 462 cm - l (a :five-membered cyclic urethane). The known
`cis-l',2',3',4'-tetrahydronaphtho-[1',2':4,5]-2-oxazolidone was measured for comparison under
`same conditions: v(CO) 1 762 cm - 1 and v(NH.) 3 455 cm - 1.
`
`3' ,5' -Di-O-benzoyl-a-ribofuro[l ',2' :4,5]-2-oxazolidone (/Vb)
`
`A stirred and cooled solution of the compound IV (0· 175 g; 1 millimole) in dry pyridine (3 ml)
`was treated with a solution of benzoyl chloride (0·42 g) in dry pyridine (2 ml), the mixture allowed
`to stand at room temperature for 24 hours, poured into cold water (20 ml) and extracted with
`chloroform (30 ml). The chloroform solution was washed subsequently with water, aqueous
`sodium hydrogen sulfate, water, a saturated aqueous solution of sodium hydrogen carbonate,
`
`2804
`
`collection czechoslov. Chem. Commun.
`
`

`

`Nucleic Acids Components and their Analogues. LXVI.
`
`Table I
`
`Hydrolytical Products of 5-Azacytidine (/)
`,Detected by Paper Chromatography or Obtained Preparatively
`
`Medium, Temperature, Time
`
`Products (Yield)
`
`Detection
`
`I, Ill, guanidine (Fig. 1)
`
`Water, 37°C, 0-2 200 min
`10- 1M Borate Buffers, pH 6·6; 7·2;
`I, Ill, guanidine (Fig. 1)
`7·8; 8·4; 37°C, 0-2 200 min
`1% Aqueous Ammonia, 37°C, 120 min Ill (91%)
`1% Aqueous Ammonia, 100°C,
`30min
`lM-HCl, 100°C, 30 min
`6M-HCI, 100°C, 30 min
`Water, 100°C, 300 min
`
`Ill (43%), guanidine (57%)
`I (51%), V (49%)
`I (5%), V (72%), VI (20%)
`isolated: 65% of IV, 77% of guanidine,
`o-ribose
`
`lM Aqueous Ammonia, 37°C,
`180 min
`Aqueous-Methanolic Ammonia,
`37°C, 5 days
`IO- 3M Borate Buffer,
`pH 9·3, 60°C, 300 min
`
`isolated: 70% of Ill
`
`isolated: 41% of Ill
`isolated: 6% of guanidine, I, Ill,
`o-ribose
`
`D1, Ds
`
`D1,Ds
`D1, Ds
`
`D1, Ds
`D1,Ds
`D1, Ds
`
`D2, D3, D4
`
`D2,D3
`
`D2,D3
`D1, D2,
`D3,D4
`
`dried over anhydrous sodium sulfate and evaporated under diminished pressure. The residual
`sirup was chromatographed on a thin layer (18 X 48 cm) of alumina (Brockmann activity
`II- III) in ethyl acetate. The zone (Rp value 0·65) absorbing in ultraviolet light was separated
`mechanically, eluted with ethyl acetate and the eluate evaporated. The residue which failed to
`crystallize was purified by precipitation of its solution in benzene (5 ml) with light petroleum
`(20 ml) followed by precipitation of the ethereal solution (10 ml) with light petroleum (40 ml).
`The final precipitate was dried at 0· 1 mm Hg to afford 0·245 g (66%) of a solid foam. Infrared
`spectrum (chloroform): v(CO) 1 726 cm - l (benzoate) and 1 777 cm - l (a five-membered cyclic
`urethane), v(NH) 3 462 cm - l (a five-membered cyclic urethane). The nuclear magnetic resonance
`spectrum (in deuterochloroform) showed a band at 5·46-r:, characteristic for the grouping
`CH2OCOC6 H5 . For C19H 17NO 7 (371·3) calculated: 3·77% N; found: 3·46% N.
`
`Dimethylaminomethylenecyanoguanidine
`
`A solution of cyanoguanidine (8-4 g; 0·1 mole) in hot absolute methanol (150 ml) was re(cid:173)
`fluxed with dimethylformamide dimethylacetal (13·1 g; 0·11 mole) for six hours under exclusion
`of atmospheric moisture (potassium hydroxide safe-guard tube). In the course of heating, the
`product began to separate. The next day, the crystals (9·6 g) were collected, washed with methanol
`and dried under diminished pressure. Concentration of mother liquors afforded an additional
`crop (1·5 g). Total yield, 11·1 g (80%). The product melted (quick heating) at 194-196°C (de(cid:173)
`composition). For C 5H 9N 5 (139·2) calculated: 43·15% C, 6·52% H, 50·33% N; found: 43·16% C,
`6·75% H, 50· 19% N.
`
`Vol. 30 119651
`
`2805
`
`

`

`Pit'hovd, Piskala, Pit'ha, Sorm:
`
`N-Formyl-N'-cyanoguanidine
`
`A solution of dimethylaminomethylenecyanoguanidine (5·56 g; 0·04 mole) in• 2M-HC1 (20 ml)
`was mechanically stirred at room temperature for 90 minutes. The reaction product separated
`from the solution. The mixture was cooled down in iced water and the product collected, washed
`with a little of cold water and dried in vacuo over concentrated sulfuric acid and potassium
`hydroxide; yield, 2·6 g. Evaporation of mother liquors afforded an additional crop (0·6 g)
`of the same product. Total yield 3·2 g (71 %). When heated to a higher temperature, the compound
`is changed slowly without any melting (up to 350°C). RF value: 0·71 (solvent system S2 , detection
`procedure D 1) . Ultraviolet spectrum is given in Table IL For C 3H 4N 4 0 (112·1) calculated:
`32·14% C, 3·64% H, 49·99% N; found: 32·11% C, 3·65% H , 49·69% N.
`
`Table II
`Ultraviolet Spectra
`
`Compound
`
`Buffer, pH
`
`Amax• nm
`
`log emax
`
`I
`
`1-~-o-Glucopyranosyl-3-guanylurea
`
`III
`Cyanoguanidine
`N-Formyl-N' -cyanoguanidine
`
`6·6
`7·2
`
`6·6
`7·2
`7·8
`8·4
`
`6·6
`water
`water
`
`244
`244
`
`225
`222
`222
`222
`
`225
`215
`238
`
`3·8a
`3·8a
`
`4·1 b
`4.3b
`4.3b
`4.3b
`
`b
`
`4·17
`4·24
`
`a Zero time extrapolation; b e at 244 nm negligible.
`
`Results and Discussion
`
`Results of the isolation experiments are shown in Table I and Fig. 1; time dependen(cid:173)
`ce of the ultraviolet spectrum is shown in Fig. 2. On the basis of these results, the
`simplified Scheme 1 of the hydrolysis may be written. In acidic media, hydrolysis
`of the glycosidic bond and deamination takes place (A). In neutral and basic media,
`the reaction sequence B occurs. In analogy to 5-azauracil, the first stage of hydrolysis
`in neutral and basic media consists in an attack in position 6, as unequivocally shown
`by isolation of 1-~-o-ribofuranosyl-3-guanylurea (III). The furanoid structure of the
`compound III follows from its electrophoretical behaviour in a borate buffer. As ex(cid:173)
`pected, the corresponding 1-~-o-ribopyranosyl-3-guanylurea exhibits a lower electro(cid:173)
`phoretical mobility. The fact that the lactol ring structures remain unchanged in both
`above cases points to the probability that also the ~-configuration at the glycosidic
`
`2806
`
`Collection Czechoslov. Chem. Commun.
`
`

`

`Nucleic Acids Components and their Analogues. LXVI.
`
`Fig. 2
`Time Dependence of the Molar Absorptivi,ty £max of 5-Azacytidine (/) at 244 nm
`1 pH 6·6; 2 pH 7·2; 3 pH 7·8.
`
`NH2
`)__
`N "'N
`
`~~~a
`(8) --
`
`~
`
`NH2
`
`H~~N
`HCO ~
`
`HOO a
`
`, NH2
`H2N~N
`
`HOCH2
`
`HN~O
`
`---
`
`HO
`
`OH
`
`fAJ
`
`HO
`
`OH
`
`II
`
`· HO
`
`OH
`
`Ill
`
`+
`
`NH2
`N~N
`
`Q_N ~O
`H
`
`V
`
`0
`NJlN
`
`~N~O
`H
`
`VI
`
`+
`
`o-ribose
`
`centra does not change. The N-formyl derivative II is assumed as intermediate
`of the hydrolysis I ~ Ill. Isolation or identification of this intermediate by paper
`chromatography failed, but the time dependence of the ultraviolet spectrum points
`to its presence. It may be noted (Fig. 2) that in the first stage of the hydrolysis an in(cid:173)
`crease of absorbance at 244 nm occurs. This increase can not be due to the formation
`
`Vol. 30 [19ii"5J
`
`2807
`
`

`

`Pi(hova, Piskala, Pit'ha, Sorm: ·
`
`of the compound III in view of the spectral properties of the compound III and the
`analogous glucosyl derivative. On the other hand, it is very probable that compounds
`of the type II would strongly absorb at the above wave length (cf. comparison
`of cyanoguanidine and N-formyl-N'-cyanoguanidine dates in Table II). Attempts
`to synthesize the compounds of the type II were fruitless; all the formylation ex(cid:173)
`periments with guanylurea led to the recovery of the starting compound or to the
`cyclic intermediate V. Notwithstanding, isolation of the compound V points to the
`reversibility of the reaction I~ II. The compound III as a relatively strong base
`is highly protonated in approximately neutral solutions.
`
`RockH1/
`
`0
`
`~
`
`1-1/ ~
`
`RO 0-CO
`IV, R = H
`/Va, R = CH 3CO
`/Vb, R = C6 H 5CO
`
`Hydrolysis of the compound I in boiling water afforded (in addition to guanidine
`and D-ribose) a good yield of Cl-D-ribofuro[l',2':4,5]-2-oxazolidone (IV), i.e., a cyclic
`N-glycosyl urethane. Such a type of compounds has not been described yet. The
`structure of the compound IV is suggested on the basis of elemental analysis, infrared
`spectra of the derivatives !Va and !Vb (a five-membered cyclic urethane), potassium
`periodate oxidation ( absence of a vicinal diol) and nuclear magnetic resonance
`spectra (presence of the grouping C6H 5 COOCH2 in the benzoate !Vb). Under
`analogous hydrolytic conditions, the same product IV has been obtained also from
`1-~-D-ribopyranosyl-5-azacytosine under isomerisation of the sugar moiety.
`For the quantitative study of the hydrolysis, sufficient numerical data are not
`available. The compound II could not be detected either by chromatography or auto(cid:173)
`radiography. As suggested by the time dependence of ultraviolet spectra ( vide infra),
`the compound II is present in a concentration exceeding 10%. The failure in isolation
`of the compound II is probably due to its instability: a decomposition takes place
`with the formation of compounds I and III in an unknown ratio. Results of concen(cid:173)
`tration measurements of compounds I and III shown in Fig. 1 constitute merely the
`upper limit of concentrations present in the reaction mixture. Since the attempted
`preparation of the compound III in a free state and sufficient purity failed, the actual
`composition of the reaction mixture could not be estimated precisely. With ,the as(cid:173)
`sumption that the compound II possesses log emax 4· 18 and that the amount of com(cid:173)
`pounds III and IV is negligible at the begin of hydrolysis, the rate of the pseudo(cid:173)
`molecular reaction I -+ II was estimated. As shown on comparison with data
`published in the previous paper12
`, the hydrolysis I-+ II proceeds (pH 6-6, calculated
`emax of the compound II 15400, k == 3·4. 10- 3 min- 1
`; pH 7·22, Gmax 15200, k =
`
`2808
`
`..:ollection czechoslov. Chem. Commun,
`
`

`

`Nucleic Acids Components and their Analogues. LXVI.
`
`= 4. 10- 3 min- 1) just a little faster than the analogous hydrolysis stage of 5-aza(cid:173)
`uracil.
`As shown by the tests with Escherichia coli (the technique, c/. 12), the compounds
`III, IV and guanidine are considerably less active than 5-azacytidine 2 which inhibits
`the growth by 50% at the concentration of 0·25 µg/ml (a 400 fold concentration is
`necessary in the case of the compound III and guanidine ).
`Quantum chemical calculations 5 •6 indicate a low density of n electrons in position
`·6 (0·744) of 5-azauracil; the hight reactivity of this position towards hydroxylic
`ions 1 •6 is in agreement with this finding. The electron density in position 6 of 5-aza(cid:173)
`cytosine is a little higher than in 5-azauracil, but considerably lower than in cytosine
`{Table III). The rate of hydrolysis of 5-azacytidine is comparable with the decomposi(cid:173)
`tion rate of 5-azauracil. It may be expected in analogy with 5-azauracil derivatives
`that substitution with o-ribose causes an increase of the rate of hydrolysis. Different
`values of hydrolytical rates of various 5-azauracil and 5-azacytosine derivatives are
`therefore in agreement with the considerably low density of n electrons.
`
`I
`I
`
`I
`
`I
`
`Compounds
`
`Table III
`Static lndicesa
`
`qi
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`Cytosineb
`V
`V(in VII)
`vc
`V (in VIII)
`
`1 ·639 0·796
`1·620 0·790
`1·623 0·783
`1·642 0·765
`1·634 0·771
`
`1·438
`1·441
`1 ·516
`1·708
`1·649
`
`0·828 1 ·169
`0·800 1·320
`0·789 1 ·323
`0·820 1 ·333
`0·799 1 ·331
`
`0·835 1·492 1·803
`0·746 1·484 1 ·798
`0·738 1·480 1·747
`0·742 1·456 1 ·535
`0·736 1·462 1 ·616
`
`Compounds
`
`Pij
`
`Fi
`
`Si
`
`12
`
`23
`
`34
`
`45
`
`56
`
`61
`
`27
`
`48
`
`Cytosineb
`V
`V(in VII)
`vc
`V(in VIII)
`
`0·375 0·433
`0·364 0·429
`0·360 0-420
`0·355 0·372
`0·357 0·387
`
`0·636
`0 ·643
`0·597
`0·445
`0·495
`
`0·525 0 ·758
`0·508 0·719
`0·506 0·718
`0·478 0·731
`0 ·490 0·724
`
`0·529 0 ·778 0-470
`0·572 0·785 0·483
`0·572 0·791 0·539
`0 ·553 0·816 0·717
`0·562 0·809 0·660
`
`-
`0-455
`0·441 0·254
`0-442 0·244
`0·448 · 0·260
`0·446 0·247
`
`a qi • electron density of the position i; Pii• order of the bond ij; Fi free valence, Si superdeloca(cid:173)
`lisibility for the position 6; b cf. 5; c in the imino form.
`
`In acidic media, 5-azacytidine is readily cleaved with the formation of o-ribose,
`again in agreement with the quantum chemical calculations. As indicated by results
`of A. and B. Pullman 5 •13 , hydrolysis of the glycosidic bond occurs the easier the lower
`
`Vol. 30 (1965)
`
`2,809
`
`

`

`Pit'hova, Piska/a, Pit'ha, Sorm:
`
`is the density of n electrons on the nitrogen atom. Comparison of values in Table III
`shows that an increased hydrolysis rate may be expected with the 5-azacytidine
`derivatives.
`The ready hydrolysis of 5-azacytidine (scheme 1, mechanisms A and B) may serve
`for explanation of the interesting biological properties of 5-azacytidine which acts.
`as a cancerostatic agent as well as mutagen. As shown in the previous paper1 on the
`basis of results of calculations in the Hiickel approximation, the mutagenic properties.
`are not caused by the amino-imino tautomerism; furthermore, the calculations.
`indicated that formation of pairs with purine bases was energetically favourable.
`Incorporation of the aza analogue into pairs may modify the chemical properties
`of the heterocycle; no considerable change of reactivity by the pair formation,however,
`is indicated on comparison of static indices of 5-azacytosine (V) in the amino and
`imino form 1 with those of the pairs 5-azacytosine-guanine (VI!) and 5-azacytosine(cid:173)
`adenine (VIII) (Table III). Incorporation of 5-azacytosine derivatives into nucleic
`acids followed by a rapid hydrolytical cleavage of the 5-azacytosine nucleus may lead
`to destruction of certain information units. Moreover, the hypothesis concerning
`the biological action of 5-azacytosine derivatives is supported by comparison of quan(cid:173)
`tu,n chemical° indices of 5-azacytosine and cytosine ( energetical indices were shown
`in the previous paper1 , static indices are listed in Table III): it may be seen that the
`most remarkable difference between both compounds consists in the character
`of position 6 important for the reactivity of the heterocyclic nucleus towards nucleo(cid:173)
`philic agents.
`
`VII
`
`VIII
`
`The authors wish to thank Dr Z. Samek ( Department of Molecular Spectroscopy of this Institute)
`for measurement and interpretation of nuclear magnetic resonance spectra. Analyses were per/ or med
`in the Analytical Department ( Head Dr J. Horticek) of this Institute by Mrs V. Rusova, Mrs E. Si-
`·
`povd and Mr V. Sterba.
`
`References
`
`1. Pifhova P., Piskala A., Pifha J., Sorm F.: This Journal 30, 1626 (1965).
`2. Sorm F ., Piskala A., Cihak A., Vesely J.: Experientia 20, 202 (1964).
`3. Pifhova P., Fucik V., Zadrazil S., Sormova Z., Sorm F.: This Journal, in the press.
`4. Cihak A., Skoda J., Sorm F.: This Journal 28, 3297 (1963).
`5. Pullman B., Pullman A.: Quantum Biochemistry. Interscience ~ublishers, New York 1963.
`6. Zahradnik R., Koutecky J., Jonas J., Gut J.: This Journal 28, 1499 (1963).
`
`2810
`
`Collection Czechoslov. Chem. Commun.
`
`

`

`Nucleic Acids Components and their Analogues. LXVI.
`
`7. Gut J ., Moravek J., Parkanyi C ., Prystas M ., Sorm F.: This Journal 24, 3154 (1959).
`8. Reindel F., Hoppe W.: Chem. Ber. 87, 1103 (1954).
`9. Roche J., Thosi N. Van, Hatt J. V .: Biochim. Biophys. Acta 14, 71 (1954).
`10. Viscontini M., Hoch D., Karrer S.: Helv. Chim. Acta 38, 642 (1955).
`11. Piskala A., Sorm F.: This Journal, in the press.
`12. Pifhova P., Piskala A ., Pifha J., Sorm F .: This Journal 35, 90 (1965).
`13. Pullman A. in the book: Electronic Aspec'ts of Biochemistry (B. Pullman, Ed.). Academic
`Press, New York 1964.
`
`Translated by J. Pliml.
`
`Note added in proof: In formulae IV, !Ve and !Vb, N should read NH.
`
`Pe3IOMe
`
`IT. ITHThroBa, A. ITHCKarra, M. ITHThra H <I>. lllopM: K0Mno11ellmbl llYK1ieu110BblX KUCllom
`u ux allallozu. LXVI. I'ui>pollu3 5-a~aqumuduua u ezo CBH3b c 6uo/lo2uq,ecKou aKmUBflocmbTO. rH.o:po(cid:173)
`JIH3 5-a3aQHTHJ:(HHa (I) B Heii:TpaJibHOH H IQeJIO'fHOH cpe.o:ax npHBOJ:(HT K cne.o:yEOIQHM npOJJ:YKTaM:
`l-~-D-pH6ocpypaH03HJI-3-ryaHHJIMO'IeBHHe (III), ryaHHJ:(HHY, Ct-D-pH6ocpypo[l ',2' :4,5]-2-0KCa30JI(cid:173)
`HJJ:OHY (IV) H D-pH603e. B Ka'IeCTBe npoMe)!(yTO'IHOro npo.o:yKTa 6bIJIO cneKTpOqJOTOMeTpH'IecKH
`o6Hapy)KeHO N-cpopMHJibHOe npOH3BOJ:(HOe coe.o:HHeHHH III. B KHCJIOH cpe.o:e 6bIJIH nony'IeHbl
`5-a3aQHT03HH (V), 5-a3aypaQHJI (VJ) H o-pu603a. B npoTHBorrono)!(HOCTb 5-a3aQIHHJJ:HHY (J),
`Bbl):(eneHHb1e npo.o:yKTbl o6na.o:aIOT TIOHH)KeHHOH 6HOJIOrH'IeCKOH aKTHBHOCTblO. KaK IlOKa3aHO
`OIIbITaMH H Ha OCHOBaHHH KBaHTOBO-XHMH'IeCKHX Bbl'IHCJieHHH, 6HOJIOrli'IeCKaH aKTHBHOCTb coe.o:H(cid:173)
`Hemrn I CB.!13aHa, llO-BHJ:(RMOMY, C BKJIIO'IeHHeM npoH3BOJ:(HOro 5-a3aQHT03HHa B HYKJieHHOBbie
`KRCJIOT.bl H C nocne.o:yIOIQHM pacIQermemi:eM TpHa3HHOBOro KOJibQa.
`
`Vol. 30 (1965)
`
`2811
`
`