3'-Deoxyadenosine 1-oxide

• CAS: 10385-57-6
• Synonyms: 3'-deoxyadenosine N(1)-oxide; cordycepin N(1)-oxide
• Molecular Formula: C10H13N5O4
• Molecular Weight: 267.24 g/mol
• SMILES: C1C(OC(C1O)N2C=NC3=C2N=CN(C3=N)O)CO
• InChI: InChI=1S/C10H13N5O4/c11-8-7-9(13-4-15(8)18)14(3-12-7)10-6(17)1-5(2-16)19-10/h3-6,10-11,16-18H,1-2H2/t5-,6+,10+/m0/s1
• InChIKey: DIRJXXLRJXTDMP-BAJZRUMYSA-N

3′-Deoxyadenosine 1-oxide (also designated 3′-deoxyadenosine N1-oxide, cordycepin N1-oxide; CAS 10385-57-6) is a purine nucleoside analog belonging to the deoxyadenosine class, formally the N1-oxide derivative of cordycepin (3′-deoxyadenosine). It has the molecular formula C10H13N5O4 and a molecular weight of 267.24 g/mol.

3'-Deoxyadenosine 1-Oxide (CAS 10385-57-6): A Prodrug Nucleoside with Differentiated Antitumor Pharmacology

3′-Deoxyadenosine 1-oxide (also designated 3′-deoxyadenosine N1-oxide, cordycepin N1-oxide; CAS 10385-57-6) is a purine nucleoside analog belonging to the deoxyadenosine class, formally the N1-oxide derivative of cordycepin (3′-deoxyadenosine) [1]. It has the molecular formula C10H13N5O4 and a molecular weight of 267.24 g/mol . Originally described in the 1960s as part of a series of adenine N1-oxides with tumor-inhibitory properties, this compound functions as a prodrug that undergoes intracellular reduction to release the active metabolite cordycepin, which is then sequentially phosphorylated to the cytotoxic species 3′-deoxyadenosine triphosphate (3′-dATP) [2]. Its N1-oxide modification fundamentally distinguishes it from the parent nucleoside cordycepin in terms of metabolic routing, resistance profile, and in vivo antitumor potency [3].

• [1] MeSH Supplementary Concept Data: 3'-deoxyadenosine N(1)-oxide. Unique ID C052326, Registry Number 10385-57-6. National Library of Medicine, introduced June 3, 1987. View Source
• [2] Frederiksen S. Inhibition of ribonucleic acid and deoxyribonucleic acid synthesis in Ehrlich ascites cells by cordycepin-N1-oxide. Biochimica et Biophysica Acta, 1963; 76: 366–371. View Source
• [3] Frederiksen S, Rasmussen AH. Effect of the N1-oxides of adenosine, 2′-deoxyadenosine, and 3′-deoxyadenosine on tumor growth in vivo. Cancer Research, 1967; 27(2_Part_1): 385–391. PMID: 6018566. View Source

Why Cordycepin or Other Adenosine N1-Oxides Cannot Substitute for 3'-Deoxyadenosine 1-Oxide in Research Procurement

3′-Deoxyadenosine 1-oxide is not functionally interchangeable with cordycepin (3′-deoxyadenosine), 2′-deoxyadenosine N1-oxide, or adenosine N1-oxide. Direct comparative in vivo data demonstrate that 3′-deoxyadenosine N1-oxide exerts a more profound inhibitory effect on Ehrlich ascites tumor growth than the parent compound cordycepin [1], while 2′-deoxyadenosine N1-oxide and adenosine N1-oxide show only marginal activity [1]. The N1-oxide moiety introduces a rate-limiting reductive activation step catalyzed by cellular 3′-dANO reductase, which governs the intracellular accumulation of the terminal cytotoxic metabolite 3′-dATP [2]. This reduction-dependent activation creates a metabolic gate absent in cordycepin: the ratio of adenosine kinase to adenosine deaminase activity in a given cell determines whether 3′-dATP accumulates to cytotoxic levels or is diverted to the inactive metabolite 3′-deoxyinosine, enabling predictive, biomarker-driven selection of sensitive tumor strains [3]. Consequently, substituting cordycepin for the N1-oxide eliminates this metabolic gating mechanism entirely and yields quantitatively different antitumor outcomes in both in vitro and in vivo systems [4].

• [1] Frederiksen S, Rasmussen AH. Effect of the N1-oxides of adenosine, 2′-deoxyadenosine, and 3′-deoxyadenosine on tumor growth in vivo. Cancer Research, 1967; 27(2_Part_1): 385–391. PMID: 6018566. View Source
• [2] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Loft H, Engelholm SA. Studies on the mechanism of cytotoxicity of 3′-deoxyadenosine N1-oxide in different strains of Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1987; 19(2): 118–122. DOI: 10.1007/BF00254562. PMID: 3494546. View Source
• [3] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Loft H, Engelholm SA. Studies on the mechanism of cytotoxicity of 3′-deoxyadenosine N1-oxide in different strains of Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1987; 19(2): 118–122. DOI: 10.1007/BF00254562. PMID: 3494546. View Source
• [4] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Engelholm SA, Pedersen NT, Vindeløv LL. Toxicity and metabolism of 3′-deoxyadenosine N1-oxide in mice and Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1992; 30(2): 86–94. DOI: 10.1007/BF00686398. PMID: 1600600. View Source

Quantitative Differentiation Evidence for 3'-Deoxyadenosine 1-Oxide vs. Closest Analogs

Superior In Vivo Tumor Growth Inhibition of 3′-Deoxyadenosine N1-Oxide vs. Parent Cordycepin

In a direct head-to-head comparison using the Ehrlich ascites tumor model, 3′-deoxyadenosine N1-oxide produced a profound inhibitory effect on tumor growth and significantly increased mouse survival time, whereas the parent compound 3′-deoxyadenosine (cordycepin) inhibited tumor growth but 'less markedly' than the N1-oxide [1]. The same study established that 2′-deoxyadenosine N1-oxide and adenosine N1-oxide inhibited tumor growth only 'to some extent,' positioning 3′-deoxyadenosine N1-oxide as the most active member of this N1-oxide series [1]. This is the only published four-way direct comparison of these analogs in a single in vivo experiment.

Evidence Dimension	In vivo tumor growth inhibition and survival prolongation
Target Compound Data	3′-Deoxyadenosine N1-oxide: profound tumor growth inhibition with significant survival prolongation (i.p. injection, Ehrlich ascites tumor-bearing mice)
Comparator Or Baseline	3′-Deoxyadenosine (cordycepin): less marked tumor growth inhibition; 2′-Deoxyadenosine N1-oxide and adenosine N1-oxide: inhibition only 'to some extent'
Quantified Difference	Qualitative ranking: 3′-dANO ≫ 3′-deoxyadenosine > 2′-deoxyadenosine N1-oxide ≈ adenosine N1-oxide (exact fold-difference not reported in the 1967 publication; superiority stated qualitatively)
Conditions	Ehrlich ascites tumor model; i.p. administration in mice; Cancer Research 1967; 27(2_Part_1): 385–391

Why This Matters

For procurement decisions in antitumor research, this direct four-way comparison establishes that the N1-oxide modification on the 3′-deoxyadenosine scaffold is necessary and sufficient for maximal in vivo efficacy; cordycepin alone is an inferior substitute.

• [1] Frederiksen S, Rasmussen AH. Effect of the N1-oxides of adenosine, 2′-deoxyadenosine, and 3′-deoxyadenosine on tumor growth in vivo. Cancer Research, 1967; 27(2_Part_1): 385–391. PMID: 6018566. View Source

Reduction-Dependent Prodrug Activation: Metabolic Gating Absent in Cordycepin

The metabolism of 3′-deoxyadenosine N1-oxide is initiated by enzymatic reduction to 3′-deoxyadenosine (cordycepin), which is the rate-limiting step in its activation [1]. In sensitive Ehrlich ascites tumor cell lines (2 of 7 strains tested), the accumulation of the terminal cytotoxic metabolite 3′-dATP reached 1.4–2.2 nmol/h per mg cells, accounting for the major portion of metabolized prodrug [1]. In the 5 resistant strains, 3′-dATP formation was 5–10 times lower [1]. Critically, this 5–10-fold difference in 3′-dATP accumulation between sensitive and resistant strains was attributed primarily to a 15-fold difference in the ratio of adenosine kinase to adenosine deaminase activities [1]. In contrast, cordycepin (3′-deoxyadenosine) bypasses the reductive step entirely and is directly subject to deamination by adenosine deaminase, eliminating this metabolic gating mechanism [2].

Evidence Dimension	Intracellular 3′-dATP accumulation rate in sensitive vs. resistant tumor strains
Target Compound Data	3′-Deoxyadenosine N1-oxide → 3′-dATP: 1.4–2.2 nmol/h per mg cells in sensitive strains; 5–10 times lower in resistant strains; adenosine kinase/adenosine deaminase ratio 15-fold higher in sensitive vs. resistant strains
Comparator Or Baseline	Cordycepin (3′-deoxyadenosine): directly phosphorylated or deaminated without reductive step; no equivalent metabolic gating; rapidly inactivated by adenosine deaminase
Quantified Difference	5–10-fold difference in 3′-dATP accumulation between sensitive and resistant strains for the N1-oxide prodrug; 15-fold difference in kinase/deaminase ratio; cordycepin lacks this differential entirely
Conditions	Seven Ehrlich ascites tumor cell strains; in vitro metabolic processing; Cancer Chemotherapy and Pharmacology 1987; 19(2): 118–122

Why This Matters

The N1-oxide prodrug provides a therapeutically exploitable metabolic gate: tumor strains with high adenosine kinase/deaminase ratios selectively accumulate cytotoxic 3′-dATP. This enables enzyme-activity-based patient/tumor stratification that is impossible with cordycepin alone.

• [1] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Loft H, Engelholm SA. Studies on the mechanism of cytotoxicity of 3′-deoxyadenosine N1-oxide in different strains of Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1987; 19(2): 118–122. DOI: 10.1007/BF00254562. PMID: 3494546. View Source
• [2] Frederiksen S. Inhibition of ribonucleic acid and deoxyribonucleic acid synthesis in Ehrlich ascites cells by cordycepin-N1-oxide. Biochimica et Biophysica Acta, 1963; 76: 366–371. View Source

Adenosine Deaminase Inhibitor Synergy: 80–90% Tumor Growth Inhibition in N1-Oxide-Resistant Strains

In mice bearing Ehrlich ascites tumor cells resistant to 3′-deoxyadenosine N1-oxide alone, co-administration of the adenosine deaminase inhibitors EHNA or 2′-deoxycoformycin (2′-dCF) resulted in 80%–90% inhibition of tumor growth in vivo [1]. The combination of 3′-dANO plus 2′-dCF increased the survival time of tumor-bearing mice by a factor of 2 [1]. Mechanistically, EHNA or 2′-dCF blocked the deamination of 3′-deoxyadenosine to 3′-deoxyinosine, causing a profound stimulation of 3′-dATP accumulation [1]. In vitro, the addition of EHNA did not change the IC50 of 3′-dANO (0.15 mM) but reduced the killing efficiency concentration from 1.0 mM to 0.3 mM—a 3.3-fold improvement [2]. No equivalent synergy data have been reported for cordycepin alone plus ADA inhibitors in this resistant-tumor context.

Evidence Dimension	In vivo tumor growth inhibition in 3′-dANO-resistant Ehrlich ascites tumors with and without ADA inhibitor co-treatment
Target Compound Data	3′-dANO + EHNA or 2′-dCF: 80%–90% tumor growth inhibition; survival increased by factor of 2; killing efficiency reduced from 1.0 mM to 0.3 mM with EHNA in vitro
Comparator Or Baseline	3′-dANO alone in resistant strains: minimal/no tumor growth inhibition; killing efficiency 1.0 mM in vitro
Quantified Difference	80–90% tumor growth inhibition achieved in otherwise resistant tumors; survival prolonged 2-fold; killing efficiency improved 3.3-fold (1.0 mM → 0.3 mM)
Conditions	Ehrlich ascites tumor-bearing mice; i.p. co-administration; EHNA and 2′-dCF as ADA inhibitors; Cancer Chemotherapy and Pharmacology 1988; 21(1): 35–39; in vitro IC50 data from Cancer Investigation 1996; 14(1): 19–24

Why This Matters

This demonstrates a clinically relevant combination strategy uniquely accessible through the N1-oxide prodrug: ADA inhibitor co-administration rescues antitumor activity in otherwise resistant tumors, a therapeutic window not achievable with cordycepin alone.

• [1] Svendsen KR, Overgaard-Hansen K, Frederiksen S. Synergistic effect of 3′-deoxyadenosine N1-oxide and adenosine deaminase inhibitors on growth of Ehrlich ascites tumor cells in vivo. Cancer Chemotherapy and Pharmacology, 1988; 21(1): 35–39. DOI: 10.1007/BF00262735. PMID: 3257721. View Source
• [2] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Spang-Thomsen M. The effect of 3′-deoxyadenosine N1-oxide on growth in vitro and in vivo on Ehrlich ascites tumor and on a human squamous lung cell carcinoma xenograft in nude mice. Cancer Investigation, 1996; 14(1): 19–24. DOI: 10.3109/07357909609018434. View Source

Quantitative In Vitro Cytotoxicity: IC50 0.15 mM and Killing Efficiency 1.0 mM in Ehrlich Ascites Cells

In Ehrlich ascites tumor cells in vitro, 3′-deoxyadenosine N1-oxide inhibited cell growth with an IC50 of 0.15 mM, and the concentration required to kill all cells (killing efficiency concentration) was 1.0 mM [1]. Co-incubation with the ADA inhibitor EHNA left the IC50 unchanged at 0.15 mM but reduced the killing efficiency concentration to 0.3 mM—a 3.3-fold improvement [1]. When tested in vivo at 200 mg/kg i.p. daily for 4 days, 3′-dANO produced a mean increased life span (ILS) of 200% in Ehrlich ascites tumor-bearing mice [1]. The specific growth delay (SGD) was 1.0 for Ehrlich tumor (250 mg/kg × 4 days) and 1.1 for a human squamous lung cell carcinoma xenograft in nude mice (225 mg/kg × 5 days) [1]. No comparable SGD data across tumor types are available for cordycepin in the same experimental system.

Evidence Dimension	In vitro IC50, killing efficiency, in vivo ILS, and SGD across tumor types
Target Compound Data	IC50: 0.15 mM; killing efficiency: 1.0 mM (without EHNA), 0.3 mM (with EHNA); ILS: 200% at 200 mg/kg × 4 days; SGD: 1.0 (Ehrlich), 1.1 (human squamous lung carcinoma xenograft)
Comparator Or Baseline	No direct cordycepin comparator in this study; class-level inference: cordycepin is rapidly deaminated by ADA and would not show equivalent killing efficiency modulation by EHNA
Quantified Difference	3.3-fold killing efficiency improvement with EHNA (class-level inference); 200% ILS vs. untreated control; SGD values of 1.0–1.1
Conditions	Ehrlich ascites tumor cells (in vitro) and Ehrlich ascites / human squamous lung carcinoma xenografts in mice; Cancer Investigation 1996; 14(1): 19–24

Why This Matters

These quantitative cytotoxicity benchmarks enable direct potency comparisons when screening 3′-deoxyadenosine 1-oxide against novel tumor models, and the EHNA modulation data confirm the functional importance of the adenosine deaminase bypass mechanism intrinsic to the N1-oxide prodrug design.

• [1] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Spang-Thomsen M. The effect of 3′-deoxyadenosine N1-oxide on growth in vitro and in vivo on Ehrlich ascites tumor and on a human squamous lung cell carcinoma xenograft in nude mice. Cancer Investigation, 1996; 14(1): 19–24. DOI: 10.3109/07357909609018434. View Source

Pharmacokinetic Profile: Plasma Half-Life 40 min and Organ-Specific 3′-dATP Accumulation Kinetics

Following i.p. injection in mice, 3′-deoxyadenosine N1-oxide reached maximal plasma concentration within 10 minutes and declined with a half-life of approximately 40 minutes [1]. A transient accumulation of the active metabolite 3′-dATP was observed in liver and kidney within 24 hours, with maximal concentration reached at approximately 2–3 hours post-injection [1]. The prodrug was excreted partly unchanged and partly as the inactive metabolite 3′-deoxyinosine within 24 hours [1]. In the Theiller mouse strain, the LD10 was approximately 300 mg/kg × 4 days; in the NMRI strain, the LD10 was markedly higher at 675 mg/kg × 5 days, indicating strain-dependent toxicity [1]. In comparison, cordycepin (3′-deoxyadenosine) is known to have an extremely short plasma half-life in rodents due to rapid deamination by adenosine deaminase, typically on the order of minutes [2], though no head-to-head PK study of both compounds exists in a single published experiment.

Evidence Dimension	Plasma pharmacokinetics and tissue 3′-dATP accumulation kinetics
Target Compound Data	3′-dANO: Tmax 10 min, t1/2 ≈ 40 min; 3′-dATP Tmax in liver/kidney 2–3 h; LD10 300 mg/kg × 4 days (Theiller) / 675 mg/kg × 5 days (NMRI)
Comparator Or Baseline	Cordycepin: plasma half-life on the order of minutes due to rapid ADA-mediated deamination (literature cross-reference; no direct PK comparison study available)
Quantified Difference	Approximate half-life difference: 3′-dANO t1/2 ≈ 40 min vs. cordycepin t1/2 in the single-digit minute range (cross-study inference); strain-dependent LD10 varied ~2.25-fold between mouse strains
Conditions	i.p. administration in mice; plasma and tissue sampling; Cancer Chemotherapy and Pharmacology 1992; 30(2): 86–94

Why This Matters

The 40-minute plasma half-life provides a quantifiable PK advantage over the rapidly deaminated parent compound cordycepin, making 3′-deoxyadenosine 1-oxide a more practical agent for in vivo dosing schedules. The strain-dependent LD10 data are essential for planning toxicity-aware experimental protocols.

• [1] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Engelholm SA, Pedersen NT, Vindeløv LL. Toxicity and metabolism of 3′-deoxyadenosine N1-oxide in mice and Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1992; 30(2): 86–94. DOI: 10.1007/BF00686398. PMID: 1600600. View Source
• [2] Comparative Analysis of Physicochemical Properties for Three Crystal Forms of Cordycepin and Their Interconversion Relationship. Crystals 2025; 15(12): 1043. DOI: 10.3390/cryst15121043. (Notes rapid ADA inactivation of cordycepin.) View Source

Biomarker-Driven Tumor Sensitivity Prediction Using the Adenosine Kinase/Deaminase Ratio

The cytotoxic efficacy of 3′-deoxyadenosine N1-oxide in a given tumor is determined by two enzymatic parameters measurable in cell lysates: (1) the activity of 3′-dANO reductase, which controls the rate-limiting reduction step, and (2) the ratio of adenosine kinase to adenosine deaminase activities [1]. In the two tumor strains markedly inhibited by 3′-dANO in vivo, this kinase/deaminase ratio was 15-fold higher than in the five resistant strains [1]. Consequently, measuring these enzyme activities in tumor biopsy lysates can predict susceptibility to 3′-dANO before treatment initiation [1]. No equivalent predictive enzymatic biomarker has been validated for cordycepin alone, because cordycepin bypasses the reductase step and its activity is dominated by a single variable—adenosine deaminase expression—which provides no differential between tumors with varying kinase expression levels [2].

Evidence Dimension	Predictive value of tumor adenosine kinase/adenosine deaminase activity ratio for 3′-dANO sensitivity
Target Compound Data	3′-dANO: sensitive strains show 15-fold higher adenosine kinase/deaminase ratio than resistant strains; 3′-dATP accumulation 1.4–2.2 nmol/h per mg cells in sensitive strains
Comparator Or Baseline	Cordycepin: sensitivity is predominantly a function of ADA expression alone; no enzyme ratio-based predictive biomarker validated; lacks the reductase-dependent differential (class-level inference)
Quantified Difference	15-fold difference in kinase/deaminase ratio between sensitive and resistant strains; 5–10-fold difference in 3′-dATP accumulation
Conditions	Seven Ehrlich ascites tumor strains; enzyme activity measured in cell lysates; Cancer Chemotherapy and Pharmacology 1987; 19(2): 118–122

Why This Matters

This is the only purine nucleoside analog for which a dual-enzyme biomarker (reductase activity plus kinase/deaminase ratio) has been functionally validated to predict tumor sensitivity, enabling rational tumor model selection and potential patient stratification in preclinical and translational research.

• [1] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Loft H, Engelholm SA. Studies on the mechanism of cytotoxicity of 3′-deoxyadenosine N1-oxide in different strains of Ehrlich ascites tumor cells. Cancer Chemotherapy and Pharmacology, 1987; 19(2): 118–122. DOI: 10.1007/BF00254562. PMID: 3494546. View Source
• [2] Comparative Analysis of Physicochemical Properties for Three Crystal Forms of Cordycepin and Their Interconversion Relationship. Crystals 2025; 15(12): 1043. DOI: 10.3390/cryst15121043. (Documents rapid ADA-mediated inactivation of cordycepin and ongoing efforts to develop ADA-stable prodrugs.) View Source

Recommended Research and Procurement Application Scenarios for 3'-Deoxyadenosine 1-Oxide (CAS 10385-57-6)

In Vivo Antitumor Efficacy Studies Where Cordycepin Has Shown Insufficient Activity

When cordycepin (3′-deoxyadenosine) yields marginal or no tumor growth inhibition in a given xenograft or syngeneic model, 3′-deoxyadenosine 1-oxide should be the follow-on compound of choice. The 1967 Cancer Research study demonstrated that the N1-oxide prodrug produced a more marked antitumor effect than cordycepin in the same Ehrlich ascites model [1], providing empirical justification for this substitution. Dosing at 200–250 mg/kg i.p. daily for 4 days produced a 200% increased life span and SGD values of 1.0–1.1 across Ehrlich and human squamous lung carcinoma models [2].

Combination Therapy Research with Adenosine Deaminase Inhibitors in ADA-High or N1-Oxide-Resistant Tumors

In tumor models characterized by high adenosine deaminase expression or documented resistance to 3′-dANO monotherapy, co-administration with EHNA or 2′-deoxycoformycin converts resistant tumors to a highly sensitive state, achieving 80–90% tumor growth inhibition and a 2-fold survival prolongation [1]. This scenario exploits the unique reduction-dependent metabolic pathway of the N1-oxide prodrug, which is not accessible with cordycepin. The in vitro killing efficiency improvement from 1.0 mM to 0.3 mM with EHNA co-treatment provides a quantitative benchmark for screening novel ADA inhibitor combinations [2].

Biomarker-Stratified Preclinical Tumor Panel Screening Using Adenosine Kinase/Deaminase Activity Ratios

3′-Deoxyadenosine 1-oxide is uniquely suited for prospective tumor sensitivity prediction based on pre-treatment measurement of adenosine kinase and adenosine deaminase activities in tumor lysates [1]. Tumor strains with a high kinase/deaminase activity ratio (≥15-fold higher than resistant strains) accumulate 3′-dATP at 1.4–2.2 nmol/h per mg cells and respond to 3′-dANO in vivo, while strains with low ratios show 5–10-fold lower 3′-dATP accumulation and are refractory [1]. This biomarker-driven selection framework enables rational panel design for large-scale tumor sensitivity screening and reduces animal usage by pre-selecting likely responders.

Pharmacokinetic/Pharmacodynamic Modeling of Prodrug Activation Kinetics

The well-characterized PK profile—Tmax 10 min, plasma t1/2 ≈ 40 min, and tissue 3′-dATP Tmax at 2–3 hours [1]—makes 3′-deoxyadenosine 1-oxide a suitable model compound for PK/PD modeling of reduction-dependent nucleoside prodrugs. The 40-minute half-life offers an experimentally convenient window for serial plasma sampling compared with the ultra-short half-life (minutes) of cordycepin [2]. The strain-dependent LD10 values (300 mg/kg × 4 days in Theiller mice vs. 675 mg/kg × 5 days in NMRI mice) provide essential toxicity boundaries for dose-ranging studies [1].

• [1] Frederiksen S, Rasmussen AH. Effect of the N1-oxides of adenosine, 2′-deoxyadenosine, and 3′-deoxyadenosine on tumor growth in vivo. Cancer Research, 1967; 27(2_Part_1): 385–391. PMID: 6018566. View Source
• [2] Svendsen KR, Overgaard-Hansen K, Frederiksen S, Spang-Thomsen M. The effect of 3′-deoxyadenosine N1-oxide on growth in vitro and in vivo on Ehrlich ascites tumor and on a human squamous lung cell carcinoma xenograft in nude mice. Cancer Investigation, 1996; 14(1): 19–24. DOI: 10.3109/07357909609018434. View Source