Inhibition of methylation of nuclear ribonucleic acid in L1210 cells by tubercidin, 8-azaadenosine and formycin

The adenosine analogs tubercidin (7-deazaadenosine), formycin (7-amino-3-[β-d-ribofuranosyl] pyrazolo[4,3-d]pyrimidine) and 8-azaadenosine were examined for their effects on the synthesis and methylation of nuclear RNA in L1210 cells in vitro. Total RNA and DNA synthesis was affected to the greatest extent by tubercidin (IC₅₀ = 7 × 10^?6M) and to an insignificant degree by 8-azaadenosine and formycin; however, the effects of the latter two drugs, but not of tubercidin, were potentiated by 2'-deoxycoformycin, an inhibitor of adenosine deaminase. In the presence of 2'-deoxycoformycin, RNA synthesis was inhibited by 40 per cent at 1 × 10^?4 M 8-azaadenosine and by 50 per cent at 2 × 10^?4 M formycin, while DNA synthesis was inhibited less extensively. Alkaline hydrolysis of nuclear RNA labeled with [¹⁴C]uridine and l-[methyl-³H]methionine showed preferential inhibition of base methylation in mononucleotides, but not of 2′-O-methylation in dinucleotides, for all three drugs. This differential effect persisted to varying degrees in ?18S and 4S nuclear RNA separated by electrophoresis. The reduction in base methylation in 4S RNA was associated with seven of the eight methylated nucleosides in 4S RNA separated by two-dimensional thin-layer chromatography. These results indicate that tubercidin, 8-azaadenosine and formycin can preferentially inhibit the base methylation of nuclear RNA relative to its synthesis.     

Antiviral and antimetabolic activities of formycin and its N1-, N2-, 2'-O- and 3'-O-methylated derivatives.

Formycin A, formycin B, and the N₁-, N₂-, 2′-O- and 3′-O-methyl derivatives of formycin A, were all examined for activity against vaccinia, herpes simplex and vesicular stomatitis viruses in primary rabbit kidney cells. The susceptibilities of calf intestinal adenosine deaminase to all the formycin A derivatives, relative to those of some adenosine analogues, were measured in order to take into account the possible effects of intracellular deamination on the antiviral and cytotoxic effects of the formycin derivatives. Formycin B was found to be inactive in all assay systems. Formycin A exhibited significant antiviral activity only against vesicular stomatitis virus, but it also proved relatively toxic to the host cells, appreciably inhibiting cellular DNA and RNA sunthesis as measured by incorporation of labelled thymidine and uridine, respectively. Of the methylated analogues, N₁-methylformycin A (which was highly resistant to enzymatic deamination) and the 2′-O-metyhlderivatives of formycin A were totally inactive in all three viral assay systems. Only N₂-methylformycin exhibited relatively high activity againts vaccinia virus, was not toxic to the cells, and did not affect cellular DNA and RNA synthesis.     

Adenosine analogs and human platelets. Effects on nucleotide pools and the aggregation phenomenon.

The interaction between human blood platelets and adenosine and various adenosine analogs was examined. Effects were found both on the pools of nucleotides, as examined by the technic of high pressure liquid chromatography, and on the phenomenon of ADP-induced platelet aggregation. Although the normal ratio of ATP and ADP in platelets is about 1.5:1, after incubation with adenosine-8-¹⁴C, ATP and ADP were labeled in a ratio of about 7:1. This is consistent with a distribution of the nucleotides among storage and metabolic pools, with the adenosine-8-¹⁴C entering principally the metabolic pool. After 2 hr of incubation with 0–5 mM 2-fluoroadenosine (F-Ado), the concentration of F-ATP was approximately 12 μmoles/10¹¹ platelets. The ratio of F-ATP to F-ADP was approximately 7:1, indicating that it entered primarily the metabolic nucleotide pool. Also, during the first hr of incubation, as the F-ATP concentration increased, the ATP concentration fell. When F-ATP-containing platelets were treated with thrombin, an aggregator and storage granule releaser, the nucleotides released into the medium consisted principally of ATP and ADP in a ratio of about 0.8:1, with very little 2-fluoroadenine-containing nucleotides. After thrombin treatment, the washed platelet pellet contained most of the 2-fluoroadenine nucleotides, but with significant increases in the concentrations of F-AMP and F-ADP. This suggests that F-ATP can replace ATP as the energy donor for the aggregation and release phenomena.As reported elsewhere, adenosine strongly inhibits platelet aggregation induced by ADP. However, this inhibitory effect disappears after preincubation for about 30 min. If the preincubation is carried out in the presence of coformycin, a tight-binding inhibitor of adenosine deaminase (K_i ? 1 × 10^?10M), the inhibition of aggregation by adenosine is markedly prolonged, indicating that the loss of inhibition results from conversion of adenosine to inosine by adenosine deaminase. ADP-induced aggregation is powerfully inhibited by F-Ado, and the inhibition becomes more pronounced on prolonged incubation. This is consistent with the observation that F-Ado has very weak substrate activity with adenosine deaminase. The analog, N⁶-phenyladenosine, an inhibitor of adenosine kinase that does not form analog nucleotides in platelets, inhibits aggregation strongly, and the inhibition is maintained during incubation of 1 hr. Several other adenosine analogs only weakly inhibit ADP-induced aggregation even in the presence of coformycin. These include 2'-deoxyadenosine, 3'-deoxyadenosine (cordycepin), arabinosyladenine, and formycin A, a C-nucleoside. However, significant quantities of nucleotides of formycin A are formed in platelets in the presence of coformycin.     

In vivo inhibition of adenosine deaminase by 2'-deoxycoformycin in mouse blood and leukemia L1210 cells

Adenosine deaminase (ADA) activities in mouse whole blood, washed erythrocytes and L1210 cells were 0.48, 0.93 and 4.76 units/ml respectively. Methods were developed to determine the second-order association rate constant (k₁) of a tight-binding ADA inhibitor, deoxycoformycin (DCF), and ADA in mouse blood and L1210 cells in vivo. After i.v. injection of DCF, the inhibition of the enzyme was of a monophasic pseudo-first-order nature in blood and biphasic (with an initial lag of 3–5 min) in L1210 cells. In contrast, i.p. injection of DCF produced the opposite pattern, monophasic in L1210 cells and biphasic in blood. The apparent k₁ values determined from the linear portions of these curves were compared with the k₁ values obtained in vitro. The mean k₁ values in vivo were: 4.2 × 10⁴ and 1.4 × 10⁴M^?1 sec^?1 in blood after i.v. and i.p. injections, respectively, and 2.6 × 10³ and 2.2 × 10⁴ M^?1 sec^?1 in L1210 cells after i.v. and i.p. injections respectively. The k₁ values with either whole blood or L1210 in vitro (3.1 × 10⁴ and 5.5 × 10³ M^?1 sec^?1, respectively) were of the same order of magnitude as those obtained with these tissues in vivo. In contrast, the k₁ values were about 150 to 1400-fold higher when either blood hemolysates (4.8 x 10^?6M^?1 sec^?1) or homogenized L1210 cells (7.5 x 10^6?1 sec^?1) were used. The 150 to 1400-fold higher k₁ values for blood hemolysates and homogenized L1210 cells than for intact cell samples (whole blood or whole L1210 ascitic fluid) suggest that the cell membrane plays a role in the interaction of DCF and ADA in these cell lines. The similarity of the rates of association of DCF and ADA in vivo and in vitro for mouse blood and ascites L1210 cells suggests that data obtained in vitro may be used to estimate the k₁ values in in vivo conditions.     

Adenosine deaminase from human erythrocytes: Purification and effects of adenosine analogs

Adenosine deaminase (adenosine aminohydrolase, EC 3.5.4.4) has been purified about 3000-fold from human erythrocytes. The molecular weight of the enzyme was estimated to be 33,000. With the partially purified erythrocytic adenosine deaminase, K_m and V_max values relative to adenosine were: adenosine, 25 μM, 100 per cent; formycin A, 1000 μM, 753–850 per cent; 8-aza-adenosine, 130 μM, 310 per cent; 6-chloropurine ribonuclcoside, 1000 μM, 91 per cent; 2,6-diaminopurine ribonucleoside, 74 μM, 91 per cent; 2'-deoxyadenosine. 7 μm, 60 per cent; xylosyladenine, 33 μm, 62 per cent; arabinosyi adenine, 100 μM, 47 per cent; 3'-deoxyadenosine (cordycepin), 41 μM, 100 per cent; 3'-amino3'-deoxyadenosine. 133 μM, 89 per cent: 4'-thioadenosine, 13 μM, 43 per cent; and 6-methylselenopurine ribonucleoside, 27 μM, 88 per cent. Apparent K_ti values of reaction products and some adenosine analogs using adenosine as a substrate were as follows; inosine. 116 μM; 2'-deoxyinosine, 60 μM; guanosine, 140 μM; 2-fluoroadenosine, 60 μM; 2-fluorodeoxyadenosine. 19 μM; N⁶-methyladenosine, 17 μM; N₁-methyladenosine, 275 μM; 6-thioguanosine, 92 μM; 6-thioinosine, 330 μM; 6-methylthioinosine, 270 μM; arabinosyl 6-thiopurine, 360 μM; and coformycin, 0.01 μM. Tubercidin (7-deaza-adenosine) and toyocamycin were devoid both of substrate and inhibitor activity. Also. N⁷-methylinosine, N⁷-methylguanosine and dipyridamole (Persantin®) did not inhibit the enzymic activity.     

Inhibitors of purine nucleoside phosphorylase, C(8) and C(5') substitutions

The C(8) and C(5') positions of base and nucleoside substrates of human erythrocytic purine nucleoside phosphorylase (PNP) are promising sites for the development of an inhibitor of this enzyme. The substrate analog, 8-aminoguanine (8-AG), has the lowest dissociation constant (K_i = 0.2–1.2 μM) of any compound reported to date and V_max = 16 per cent (relative to guanine). Other C(8) substituents decreased the affinity for PNP and, with the exception of the methyl and sulfhydryl groups, abolished substrate activity. Halogens or a thiomethyl group at C(5') of inosine resulted in unchanged or improved affinities (K_i = 10–30 μM) and greatly decreased substrate activity (V_max < 1 per cent relative to inosine). The K_i of formycin B was reduced from 100 μM to 10 μM or less by substitution of a halogen at C(5'). Phosphorolysis of purine nucleosides was inhibited significantly by 8-AG in intact human erythrocytes and murine Sarcoma 180, L1210 and L5178Y cells. Although a K_i value of 17 μM was determined for 8-aminoguanosine, it was equally effective in inhibiting PNP activity in intact cells. The nucleoside was cleaved to 8-AG which was not a substrate for guanase or hypoxanthine-guanine phosphoribosyltransferase. Despite its low substrate activity (V_max < 0.2%). 5′-deoxy-5′-iodoinosine was cleaved by intact L1210 and L5178Y cells.     

Potentiation by 2'-deoxycoformycin of the inhibitory effect of xylosyladenine on nuclear RNA synthesis in L1210 cells in vitro

The effect of the adenosine deaminase inhibitor, 2'-deoxycoformycin (dCF), on the inhibitory effect of 9-β-d-xylofuranosyladenine (XA) on nuclear RNA synthesis was examined in L1210 cells in vitro. Pretreatment of cells for 15 min with a 100 per cent inhibitory dose (1 × 10^?6 M) of dCF resulted in approximately a 3- to 8-fold reduction in the 50 per cent inhibitory dose (id₅₀) of XA for [³H]uridine and [³H]thymidine incorporation into RNA and DNA respectively. The id₅₀ for XA for RNA synthesis vs DNA synthesis was 5-fold lower in the absence of dCF and 20-fold lower in the presence of dCF, indicating the greater sensitivity of RNA synthesis to this inhibitor. Fractionation of nuclear RNA into rRNA, non-poly(A) heterogeneous RNA and poly(A)heterogeneous RNA revealed the latter species of RNA to be less sensitive to XA in the absence of dCF; however, in the presence of dCF, all three species of nuclear RNA showed similar sensitivities. Nuclear polyadenylic acid synthesis was among the most sensitive RNA fractions to XA, and was also inhibited to a greater degree by pretreatment of cells with dCF. These results indicate that XA is potentiated markedly by inhibition of adenosine deaminase, and that deamination serves as a major catabolic route for this drug.     

Tight binding inhibitors--VI. Interactions of deoxycoformycin and adenosine deaminase in intact human erythrocytes and sarcoma 180 cells

The inactivation and reactivation of adenosine deaminase (ADA) by deoxycoformycin was studied in intact human erythrocytes and murine Sarcoma 180 cells in vitro. The second-order rate constant (k₁) for the association reaction between deoxycoformycin and intraerythrocytic ADA was calculated to be 5.1 × 10³M^?1 sec^?1. This is about 300 to 500-fold lower than the k₁ values determined either with hemolyzed human erythrocytes (k₁ = 1.4 × 10⁶ M^?1 sec^?1) or with partially purified human erythrocytic ADA (k₁ = 2.6 × 10⁶ M^?1 sec^?1). In intact erythrocytes only slight reactivation (<10 per cent) of the inhibited ADA (EI complex) was detectable over 24 hr, whereas with hemolysates about 50 per cent reactivation of the inhibited ADA was observed in about 25 hr (k₂ = 7.7 × 10^?6sec^?1). The k¹ values with intact and supernatant fractions from homogenized Sarcoma 180 cells were determined to be 1.1 × 10⁴M^?1 sec^?1 and 4.2 × 10⁶ M^?1 sec^?1 respectively. With intact Sarcoma 180 cells, negligible reactivation of ADA was seen during a 48-hr period. Preliminary studies indicate an important role for the erythrocytic nucleoside transport system on the apparent k₁ values and the rate of inactivation of ADA by deoxycoformycin in intact cells.     

Effects of adenosine analogs and adenine nucleotides on adenosine 5'-diphosphate-induced rat platelet aggregation

Various adenosine analogs and adenine nucleotides have been tested as inhibitors of ADP-induced aggregation of rat platelets. The potent inhibitors of human platelet aggregation, adenosine, 2-fluoroadenosine, 2-chloroadenosine, carbocyclic adenosine and N⁶-phenyl adenosine, had little effect on rat platelet aggregation (0–30 per cent inhibition). The effects of adenosine or its analogs on ADP-induced aggregation of cross-species platelet-rich plasmas (PRPs) (human platelets suspended in rat plasma or rat platelets in human plasma) were similar to those with the native PRPs, indicating that these species differences were due to intrinsic factors in the platelets and not in the plasma. When these analogs were tested in the presence of the cyclic AMP phosphodiesterase inhibitor papaverine, strong inhibiton of rat platelet ADP-induced aggregation was seen. 2′-Deoxyadenosine and 3′-deoxyadenosine were not inhibitory to ADP-induced aggregation of rat PRP even in the presence of papaverine. Adenosine 5′-tetraphosphate strongly inhibited both human and rat platelet aggregation. AMP, like adenosine, did not inhibit rat platelet aggregation but became strongly inhibitory in the presence of papaverine. This inhibitory effect was abolished by preincubating rat PRP with an adenylate cyclase inhibitor, 2′, 5′-dideoxyadenosine or adenosine deaminase. In the later case, however, if the adenosine deaminase inhibitor 2′-deoxycoformycin was included in the incubation mixture, the inhibition by AMP plus papaverine was similar to adenosine plus papaverine. About 50 per cent of [¹⁴C]AMP was converted to [¹⁴ C]adenosine in rat platelet-free plasma or PRP after a 10-min incubation. α,β-Methylene-ADP and β,γ-methylene-ATP (200 μM) inhibited rat platelet aggregation by 50 and 64 per cent, respectively. Cyclic AMP phosphodiesterase of rat and human platelets gave comparable K_m, and V_max values (K_m 0.53 and 0.21μM and V_max 6.0 and 6.7 pmoles/min/10⁷ platelets, respectively).     

Effects of 8-azaadenosine and formycin on cell lethality and the synthesis and methylation of nucleic acids in human colon carcinoma cells in culture

The cytocidal and biochemical effects of formycin and 8-azaadenosine in the presence and absence of the adenosine deaminase inhibitor, 2′-deoxycoformycin, were studied in human colon carcinoma (HT-29) cells in culture. Logarithmically growing cells were unaffected by 24-hr exposure to either 10^?6M formycin or 8-azaadenosine, but 1 to 1.4 log reductions in colony formation were produced by 10^?5M of each analog. In the presence of 10^?6M 2′-deoxycoformycin, a 3- and 30-fold potentiation of the cytocidal activity of 8-azaadenosine and formycin, respectively, was produced. Inhibition of DNA synthesis but not RNA synthesis by 8-azaadenosine paralleled its cytocidal activity; however, neither variable correlated closely with the cytotoxic effects of formycin. In addition, the methylation of nuclear RNA was unaffected by both drugs while the methylation of 5-methyl-deoxycytidine in DNA was inhibited to a lesser extent than DNA synthesis. Measurements of the incorporation of [³H]formycin and [³H]8-azaadenosine into nuclear RNA and DNA in the presence and absence of 2′-deoxycorformycin indicated that formycin substitution in RNA and DNA was enhanced 10- and 20-fold, respectively, while [³H]8-azaadenosine incorporation into both nucleic acids was increased 6- to 7-fold. These results suggest that the incorporation of formycin into nucleic acids, particularly DNA, correlates closely with its lethal effect on cell viability. On the other hand, the cytocidal activity of 8-azaadenosine more clearly parallels its inhibitory effect on DNA synthesis rather than its substitution into nucleic acids.     

Adenosine analogs and human platelets--II. Inhibition of ADP-induced aggregation by carbocyclic adenosine and imidazole-ring modified analogs. Significance of alterations in the nucleotide pools.

A number of adenosine analogs modified in the imidazole portion of the purine ring or in the carbohydrate moiety have been examined for their ability to inhibit the aggregation of human blood platelets induced by 10 μM ADP. Also, the effects of incorporation of adenosine analogs into the nucleotide pools, or of alteration of the natural adenine nucleotide levels, were studied. In accord with earlier findings, alterations in the ribose moiety of adenosine markedly diminished effectiveness in blocking ADP-induced aggregation. The C-nucleoside, formycin and its 1-methyl and 2-methyl derivatives, displayed little inhibitory activity and α-D-ribofuranosyl adenine and α-l-lyxofuranosyl adenine were without activity. Replacement of the 5'-hydroxyl group of the ribose by carboxyl, amino or S-methyl groups decreased capacity to inhibit aggregation. In contrast, carbocyclic adenosine (wherein an oxygen atom of the ribofuranosyl ring is replaced by a methylene group) retained full ability to inhibit ADP-induced aggregation. The effects of modifications in the imidazole portion of the purine ring of adenosine are complex. Tubercidin (7-deazaadenosine), 4-aminopyrazolo [3,4-d]pyrimidine ribonucleoside (4-APP-ribonucleoside) and 8-azaadenosine displayed negligible or weak inhibitory activity. However, analogs related to tubercidin, i.e. 6-aminotoyocamycin and sangivamycin (5-carboxamide tubercidin), as well as the 3-carboxamide derivative of 4-APP ribonucleoside, displayed inhibitory activity approaching that of adenosine. No relation was established between the incorporation of adenosine analogs into platelet nucleotide pools and their ability to inhibit ADP-induced aggregation. Analogs such as formycin, which in the presence of the adenosine deaminase inhibitor, deoxycoformycin, readily enter the platelet nucleotide metabolic pool with replacement of a substantial portion of the natural adenine nucleotides, cause only weak inhibitory effects, whereas analogs such as 6-aminotoyocamycin, and sangivamycin, which did not form analog nucleotides or formed only small quantities of the 5'-monphosphate nucleotides, display inhibitory activity comparable to that of adenosine. When platelets were incubated with 2-fluoroadenosine, a potent inhibitor of aggregation, large amounts of 2-fluoroadenosine, 5'-mono, di- and triphosphate nucleotides were formed with a coincident marked decrease in the ATP concentrations. When these platelets were washed free of extracellular 2-fluoroadenosine and resuspended in platelet-free plasma, normal aggregation was produced by ADP. However, incubation of this suspension for 10 min with 2-fluoroadenosine resulted in an almost complete inhibition of ADP-induced aggregation.     

Biochemical properties of the nucleoside of 3-amino-1,5-dihydro-5-methyl-1,4,5,6,8-pentaazaacenaphthylene (NSC-154020)

A study of 3-amino-1, 5-dihydro-5-methyl-l-β-d-ribofuranosyl-1, 4, 5,6,8- pentaazaacenaphthylene (NSC-154020), a “tricyclic” nucleoside with activity against certain experimental tumors, was undertaken to determine if it differed in biochemical properties from structurally related7- deazapurine nucleosides with established biological activity, such as sangivamycin. In cultured L1210 cells, [¹⁴C-methyl]-NSC-154020 was converted to a single metabolite with the properties of a 5'-monophosphate as shown by (a) similarity to AMP in migration on paper chromatograms and in retention time when subjected to high pressure liquid chromatography (h.p.l.c.) on an ion exchange column and (b) conversion to a compound with the properties of NSC-154020 upon treatment with alkaline phosphatase or 5'-nucleotidase. In cultured H.Ep.-2 cells, the principal metabolite of NSC-154020 was also the monophosphate. H.Ep.-2 cells contained in addition a variable amount of a second metabolite which also had the retention time (on h.p.l.c. analysis) of a monophosphate and which was converted by the action of alkaline phosphatase or 5'-nucleotidase to a compound that migrated like NSC-154020 upon chromatography in three solvent systems. This second metabolite is as yet unidentified. In crude extracts of L1210 cells, addition of adenosine or 6-(methylthio)purine ribonucleoside decreased the phosphorylation of NSC-154020. NSC-154020 was a substrate for adenosine kinase 110-fold purified from H.Ep.-2 cells; the K_m was 215 μM and the V_max was 1.8 times greater than that of adenosine. No ¹⁴C from labeled NSC-154020 was found in the polynucleotides of either H.Ep.-2 cells or L1210 cells grown for 24 hr in the presence of the labeled nucleoside. Several different studies failed to reveal any selective sites of action for NSC-154020. In cultured L1210 cells it inhibited synthesis of DNA, RNA and protein and reduced ribonucleotide pools, but with little selectivity. The incorporation of [¹⁴C]formate into polynucleotides was inhibited more severely than that of hypoxanthine; this is evidence for a blockade of purine synthesis de novo, an effect also produced by many other analogs of purines and nucleosides. Sangivamycin produced generally similar inhibitions of incorporation of formate and hypoxanthine. However, the cytotoxicity of NSC-154020 and sangivamycin to L1210 cells was not prevented or reversed by 5-amino-4-imidazolecarboxamide (AIC), adenine, guanine, hypoxanthine, uridine, or AIC + uridine; therefore, inhibition of de novo synthesis of purine and pyrimidine nucleotides is not a primary site of action of these compounds. Although the loci of action of NSC-154020 are not yet defined, the fact that it is not metabolized to polyphosphates indicates that its mechanism of action probably differs significantly from those of the related compounds, tubercidin and sangivamycin, which are converted to polyphosphates and incorporated into RNA and DNA.     

Successful chemoimmunotherapy of murine L1210 lymphatic leukemia with cyclophosphamide and mafosfamide-treated leukemia cells

Summary Balb/c × DBA/2 F₁ mice (CD2F₁ mice) bearing L1210 lymphatic leukemia (10 L1210 cells i.p. injected on day 0) were subjected to chemoimmunotherapy. They received 100 mg/kg of cyclophosphamide i.p. on day + 8 and 10⁶ or 10⁷ immunogenic L1210 cells treated in vitro with mafosfamide — ASTA Z7654 (L1210-Maf cells) i.p. or i.p. + s.c. on days 0, + 3, + 6, + 9, + 12 after the leukemia implantation.About 30% of leukemia-bearing mice receiving cyclophosphamide and L1210-Maf cells after L1210 inoculation were able to reject the leukemia (as compared with 0% after injection of L1210-Maf cells only or 5% after cyclophosphamide administration). Better results (54% of cured mice) were obtained if 10⁷ L1210-Maf cells were injected i.p. + s.c. beside cyclophosphamide. Biological response modifiers (BRM's): levamisole, BCG, bestatin did not improve these results in the doses used in the experiment.Augmentation of anti-L1210 therapeutic response is dependent on the administration of cyclophosphamide and L1210-Maf cels. Cyclophosphamide not only reduces the tumor burden but probably can potentiate the L1210-Maf dependent antitumor immunity as well.     

Ellipticine and derivatives induce breakage of L1210 cells DNA in vitro.

Part of the radioactive DNA extracted from prelabelled L1210 cells exposed in vitro to ellipticine, 9-hvdroxyellipticine, 2-methyl-9-hydroxyellipticinium and 9-aminoellipticine, migrate more slowly into an alkaline sucrose gradient. These effects occur in less than one hr of exposure, and can be detected at drug concentrations which lie respectively around 1000, 20, 100 and 500 ng per ml whereas the drug concentrations which inhibit by 50 per cent the rate of the in vitro cell growth are 242, 3.9, 13 and 53 ng per ml. The DNA breaks become rapidly undetectable when the cells' exposure to the drugs is discontinued.     

Failure of 6-thioGMP to inhibit guanylate kinase in intact cells.

Experiments have been performed that make necessary a modification of the hypothesis that 6-thioguanine produces cytotoxicity by a sequential blockade of the enzymes, ribosylamine-5-phos-phate: pyrophosphate phosphoribosyltransferase (PRPP-amidotransferase), inosinate dehydrogenase and guanylate kinase. L5178Y murine leukemia cells (in vivo and in vitro) were pretreated with 6-thio-guanine under conditions known to produce significant accumulations of the analog nucleoside 5'-monophosphate, 6-thioGMP, inhibition of purine de novo biosynthesis, and marked cytotoxicity. When these cells were incubated with [8-¹⁴C]guanine, no inhibition in the formation of guanine nucleotides was observed in comparison with control cells not treated with 6-thioguanine. Furthermore, measurement of changes in the intracellular concentrations of GMP, GDP and GTP did not provide evidence for the occurrence of a ‘cross-over’' between GMP and GDP in the presence of 6-thioGMP. Thus, the predicted accumulation of GMP resulting from the postulated blockade of guanylate kinase by 6-thioGMP did not occur. L5178Y cells incubated with guanine for periods of 1 hr or less accumulated concentrations of GDP and GTP that approximated the intracellular levels of ADP and ATP. Time studies were performed with human erythrocytes in which the rate of formation of nucleotides was compared in cells incubated with guanosine or 6-thioguanosine. With guanosine, the rapid formation of guanine nucleotides was observed with the ratio of GTP:GDP:GMP approximating the ratios of the adenine nucleotides of the erythrocytes, i.e. no accumulation of GMP was observed at any time period up to 6 hr. In contrast, with 6-thioguanosine, a rapid initial formation of 6-thioGMP was observed with a gradually accelerating formation of 6-thioGTP. After 2 hr, the concentration of 6-thioGMP decreased whereas the formation of 6-thioGTP achieved a velocity of about 0.007 μmole/ min/ml of erythrocytes. This velocity is about 2 per cent of that expected with saturating levels of GMP, if one assumes the intraerythrocytic activity of guanylate kinase to be 0.3 enzyme unit/ml of cells (enzyme unit: 1 unit catalyzes the conversion of 1 μmole substrate into product/min). These findings are in general agreement with the results of studies with purified guanylate kinase preparations described in the accompanying publication [R.L. Miller, D.L. Adamczyk, T. Spector, K.C. Agarwal, R.P. Miech and R.E. Parks Jr., Biochem. Pharmac. 26, 1573 (1977)], which indicate that the K_m value and V_max value of 6-thioGMP are approximately 2.0 mM and about 3 per cent of the V_max with GMP respectively. Therefore, it may be concluded that administration of 6-thioguanine does not cause significant inhibition of guanylate kinase. However, the poor reactivity of 6-thioGMP with guanylate kinase probably causes the marked intracellular accumulation of this analog nucleotide after administration of 6-thioguanine or its derivatives.     

Analog specific aberrancies in antifolate inhibition of L1210 cell dihydrofolate reductase

During studies with L1210 cells and a variety of folate analogs, large discrepancies were revealed between data on membrane transport, on inhibition of dihydrofolate reductase in cell-free extracts, and on inhibition of growth in culture for 10-oxa-, 10-benzyl- and 10-phenethyl-aminopterin, and for 3-deaza, 10-methyl-aminopterin. While aminopterin, 10-methyl (methotrexate)-, 10-ethyl- and 10-propyl-aminopterin were tight binding inhibitors (K_i: 2–3 × 10^?12M) of dihydrofolate reductase in cell-free extracts from L1210 cells, the other four analogs were only weak competitive inhibitors (K_i = 3–300 × 10^?8M). Similar differences among analogs were observed for inhibition of dihydrofolate reductase in cell-free extracts from Sarcoma 180 and Ehrlich cells, but not for this enzyme in microbial cell-free extracts. There were only small differences in the transport of all of the analogs by L1210 cells. Inhibition of L1210 cell growth in culture by 10-oxa-, 10-benzyl- and 10-phenethyl-aminopterin and by 3-deaza, 10-methyl-aminopterin, in contrast to the other analogs, was several orders of magnitude greater than that predicted from the data on dihydrofolate reductase inhibition. The extent of binding of 10-oxa-, 10-benzyl- and 10-phenethyl-aminopterin, and of 3-deaza and 10-methyl-aminopterin to dihydrofolate reductase in intact L1210 cells, in contradistinction to that seen for the cell-free enzyme preparations, approached that observed for methotrexate; these estimates of drug-enzyme interaction in situ were more predictive of the extent of inhibition by these analogs of L1210 cell growth in culture.     

Potentiation of the growth inhibitory effects of adenosine 3',5'-monophosphate analogues by homocysteine

The growth inhibitory effects of N⁶-monobutyryl adenosine 3',5' monophosphate (mbcAMP) and N⁶,O²'-dibutyryl adenosine 3',5' monophosphate (dbcAMP) towards Walker carcinoma in vitro are significantly potentiated by the addition of l-homocysteine to the culture medium. This effect is not seen with l-cysteine or when exogenous cAMP or prostaglandin E₂(PGE₂) replace the butyrylated cyclic nucleotide. Combinations of mbcAMP or dbcAMP and l-homocysteine significantly inhibit nucleic acid methylations. Both the butyrylated cyclic nucleotides cause an elevation of the intracellular level of S-adenosyl-l-homocysteine (SAH), a potent inhibitor of S-adenosyl-l-methionine (SAM) dependent methyltransferases, and this is significantly enhanced in combination with l-homocysteine. The increase in SAH level produced by such combinations is proportional to the inhibition of methyl group incorporation into 5-methyl cytosine and 7-methyl guanine. These results suggest that l-homocysteine potentiates accumulation of SAH in the presence of mbcAMP and dbcAMP and that the resultant inhibition of methylation accounts for the enhanced growth inhibition.     

Studies with a 2,4-diamino-5-(3',4'-dichlorophenyl)-6-methylpyrimidine (DDMP)-resistant L1210 leukemia cell line without cross-resistance to methotrexate

An L1210 leukemia cell line resistant to 2,4-diamino-5-(3',4'-dichlorophenyl)-6-methylpyrimidine (DDMP) (L1210/DDMP) was developed in vivo by treatment of tumor-bearing mice. Resistance to DDMP was confirmed by subsequent in vivo survival experiments and by in vitro dose-response curves. The L1210/DDMP line demonstrated little cross-resistance to another folate analog, methotrexate (MTX). This was confirmed both in vivo, with survival experiments, and in vitro, using dose-response curves. A statistical analysis of the in vivo data confirmed DDMP resistance with lack of MTX cross-resistance. Dihydrofolate reductase (DHFR) activity in the L1210/DDMP/R₅ line was no greater than in the parent cell line (L1210/S). and the K_m of DHFR for dihydrofolate was the same in the L1210/DDMP/R₅ and L1210/S lines. The K_i for DHFR of the L1210/DDMP/R₅ cell line versus the L1210/S cell line was increased 3.0-fold for MTX and 3.5-fold for DDMP. Total accumulation of [¹⁴C]DDMP was identical in the two cell lines. The explanation for the lack of MTX cross-resistance in the L1210/DDMP/R₅ line is unknown.     

Population analysis of the non linear red blood cell partitioning and the concentration-effect relationship of draflazine following various infusion rates

Aims To investigate the impact of the specific red blood cell binding on the pharmacokinetics and pharmacodynamics of the nucleoside transport inhibitor draflazine after i.v. administration at various infusion rates. It was also aimed to relate the red blood cell (RBC) occupancy of draflazine to the ex vivo measured adenosine breakdown inhibition (ABI). Methods Draflazine was administered to healthy volunteers as a 15-min i.v. infusion of 0.25, 0.5, 1, 1.5 and 2.5 mg immediately followed by an infusion of the same dose over 1 h. Plasma and whole blood concentrations were measured up to 120 h post dose, and were related to the ex vivo measured ABI, serving as a pharmacodynamic endpoint. The capacity-limited specific binding of draflazine to the nucleoside transporter located on the erythrocytes was evaluated by a population approach. Results The estimate of the population parameter typical value (%CV) of the binding constant K_d and the maximal specific binding capacity (B_max ) was 0.385 (3.5) ng ml⁻¹ plasma and 158 (2.1) ng ml⁻¹ RBC, respectively. The non-specific binding was low. The specific binding to the erythrocytes was a source of non-linearity in the pharmacokinetics of draflazine. The total plasma clearance of draflazine slightly decreased with increasing doses, whereas the total clearance in whole blood increased with increasing doses. The sigmoidal E_max equation was used to relate the plasma and whole blood concentration of draflazine to the ex vivo determined ABI. In plasma, typical values (%CV) of E_max, IC₅₀ and Hill factor were 81.4 (1.9)%, 3.76 (9.3) ng ml⁻¹ and 1.06 (3.4), respectively. The relationship in whole blood was much steeper with population parameter typical values (%CV) of E_max, IC₅₀ and Hill factor of 88.2 (2.0)%, 65.7 (2.8) ng ml⁻¹ and 4.47 (5.5), respectively. The RBC occupancy of draflazine did not coincide with the ex vivo measured ABI. The observed relationship between RBC occupancy and ABI was not directly proportional but similar for all studied infusion schemes. Conclusions The findings of this study show that the occupancy of the nucleoside transporter by draflazine should be at least 90% in order to inhibit substantially adenosine breakdown in vivo. On the basis of these findings it is suggested that a 15 min infusion of 1 mg draflazine followed by an infusion of 1 mg h⁻¹ could be appropriate in patients undergoing a coronary artery bypass grafting.     

Pathways of nucleotide metabolism in schistosoma mansoni. V. Adenosine cleavage enzyme and effects of purine analogues on adenosine metabolism in vitro

Schistosoma mansoni extracts have been found to possess an active anabolic pathway for nucleotide biosynthesis in which adenosine is cleaved to adenine followed by conversion of adenine to AMP via adenine phosphoribosyltransferase. A significant fraction of labeled adenosine was found to enter the nucleotide pool by this pathway; howeverm, most of the nucleoside was converted to nucleotides by a pathway which employs adenosine deaminase, purine nucleoside phosphorylase and hypoxanthine phosphoribosyltransferase enzymes. Formycin A has been found to be a potent blocker of adenosine cleavage when tested in worm extracts. Arabinosyl-6-mercaptopurine and 6-thioguanosine are inhibitors of worm adenosine deaminase, and formycin B and 6-thioguanosine were found to inhibit the purine nucleoside phosphorylase of this parasite. Combinations of arabinosyl-6-mercaptopurine with either formycin A or formycin B result in substantial blockage of adenosine utilization for nucleotide synthesis. These studies thus suggest that adenosine analogues in combination might be useful in vivo for the chemotherapy of schistosomiasis.