3'-beta-Azido-2',3'-dideoxyuridine

• Molecular Formula: C9H11N5O4
• Molecular Weight: 253.22 g/mol
• InChI: InChI=1S/C9H11N5O4/c10-13-12-5-3-8(18-6(5)4-15)14-2-1-7(16)11-9(14)17/h1-2,5-6,8,15H,3-4H2,(H,11,16,17)/t5?,6-,8-/m1/s1
• InChIKey: ZSNNBSPEFVIUDS-KYVYOHOSSA-N

CS-87: Bone Marrow-Sparing NRTI

3′-beta-Azido-2′,3′-dideoxyuridine (synonyms: CS-87, AzddU, AZdU, Navuridine) is a synthetic pyrimidine nucleoside analogue belonging to the 2′,3′-dideoxynucleoside class of HIV-1 reverse transcriptase inhibitors [1]. It is the uridine congener of the first approved anti-HIV drug, 3′-azido-3′-deoxythymidine (AZT, zidovudine), differing by replacement of the thymine base (5-methyluracil) with uracil [2]. The compound was originally developed as a clinical candidate (entering Phase I/II evaluation [3]) based on its potent inhibition of HIV-1 replication in primary human peripheral blood mononuclear cells (PBMC) with an EC₅₀ range of 0.18–0.46 µM and, critically, limited toxicity toward human bone marrow progenitor cells, a key dose-limiting adverse effect of AZT [4].

1

Workflow HIV-1 reverse transcriptase chain-termination inhibition studies

2

Selection Bone marrow-sparing NRTI probe with uracil base; reduces thymidine kinase-driven activation in resting cells

3

Chemistry 3′-azido handle enables CuAAC click conjugation for activity-based probe development

4

Metabolism Unique intracellular diphosphohexose depot pathway for sustained nucleotide release research

• [1] Chu CK, Schinazi RF, Ahn MK, Ullas GV, Gu ZP. Structure-Activity Relationships of Pyrimidine Nucleosides as Antiviral Agents for Human Immunodeficiency Virus Type 1 in Peripheral Blood Mononuclear Cells. Journal of Medicinal Chemistry. 1989;32(3):612–617. doi:10.1021/jm00123a018. View Source
• [2] Boudinot FD, Schinazi RF, Gallo JM, et al. 3′-Azido-2′,3′-dideoxyuridine (AzddU): comparative pharmacokinetics with 3′-azido-3′-deoxythymidine (AZT) in monkeys. AIDS Research and Human Retroviruses. 1990;6(2):219–228. doi:10.1089/aid.1990.6.219. View Source
• [3] Zhu Z, Hitchcock MJ, Sommadossi JP. Cellular metabolism of 3′-azido-2′,3′-dideoxyuridine with formation of 5′-O-diphosphohexose derivatives by previously unrecognized metabolic pathways for 2′-deoxyuridine analogs. Molecular Pharmacology. 1990;38(6):929–938. View Source
• [4] NCATS Inxight Drugs. Navuridine (3′-Azido-2′,3′-dideoxyuridine). NCATS. Accessed 2026. View Source

Why CS-87 Cannot Be Replaced by AZT

Despite sharing a common 3′-azido-2′,3′-dideoxyribose scaffold, 3′-azido-2′,3′-dideoxyuridine (CS-87) diverges sharply from its closest structural analog, AZT (3′-azido-3′-deoxythymidine), in critical preclinical performance parameters that dictate experimental design and therapeutic window. AZT exhibits a severe liability in bone marrow toxicity that limits its sustained use, whereas CS-87 demonstrates markedly reduced human bone marrow cell toxicity at comparable anti-HIV potencies [1]. Furthermore, CS-87 displays a fundamentally different brain penetration profile—with a 3.7-fold higher brain-to-serum ratio at 50 mg/kg [2]—and a distinct intracellular metabolic fate, accumulating unique diphosphohexose derivatives not observed with AZT [3]. Simple substitution of CS-87 with AZT, 2′,3′-dideoxyuridine (which lacks the 3′-azido group and has negligible anti-HIV activity [4]), or 3′-fluoro-2′,3′-dideoxyuridine therefore invalidates comparative conclusions regarding hematological safety, CNS exposure, and long-term cellular pharmacology.

Property

CS-87 (AZddU)

AZT (zidovudine)

Thymidine kinase affinity

Low affinity (reported Ki/Km substantially higher); limits phosphorylation in resting cells

High affinity; efficient phosphorylation in many tissues

Intracellular metabolite profile

Forms diphosphohexose depot; sustained release of active monophosphate reported

Predominantly triphosphate; no diphosphohexose depot reported

Brain-to-serum distribution

Higher reported CNS exposure ratio in murine models

Lower CNS penetration; limited brain distribution

• [1] Chu CK, Schinazi RF. US Patent US4916122A: 3′-Azido-2′,3′-dideoxyuridine anti-retroviral composition. United States Patent and Trademark Office. 1990. View Source
• [2] Doshi KJ, Gallo JM, Boudinot FD, Schinazi RF, Chu CK. Comparative pharmacokinetics of 3′-azido-3′-deoxythymidine (AZT) and 3′-azido-2′,3′-dideoxyuridine (AZddU) in mice. Drug Metabolism and Disposition. 1989;17(6):590–594. PMID:2575492. View Source
• [3] Zhu Z, Hitchcock MJ, Sommadossi JP. Cellular metabolism of 3′-azido-2′,3′-dideoxyuridine with formation of 5′-O-diphosphohexose derivatives by previously unrecognized metabolic pathways for 2′-deoxyuridine analogs. Molecular Pharmacology. 1990;38(6):929–938. View Source
• [4] Chu CK, Schinazi RF, Ahn MK, Ullas GV, Gu ZP. Structure-Activity Relationships of Pyrimidine Nucleosides as Antiviral Agents for Human Immunodeficiency Virus Type 1 in Peripheral Blood Mononuclear Cells. Journal of Medicinal Chemistry. 1989;32(3):612–617. doi:10.1021/jm00123a018. View Source

CS-87: Quantitative Evidence vs Analogs

Bone Marrow Cytotoxicity: CS-87 vs AZT

In primary human peripheral blood mononuclear cells (PBMC) infected with HIV-1, 3′-azido-2′,3′-dideoxyuridine (CS-87) exhibited a median effective concentration (EC₅₀) of 0.18–0.46 µM [1]. The same study series reported that the structurally related and clinically approved comparator, 3′-azido-3′-deoxythymidine (AZT, zidovudine), displayed a substantially lower EC₅₀ of approximately 0.004 µM in the same PBMC assay system [2]. Thus, AZT is approximately 45- to 115-fold more potent on a molar basis in this primary cell model.

Bone Marrow Cytotoxicity

Head-to-head

≥20-fold lower cytotoxicity than AZT in human bone marrow cell cultures

Supports bone marrow-sparing HIV research models

Human CFU-GM assay; parent nucleoside comparison

Evidence Dimension	Anti-HIV-1 potency (EC₅₀)
Target Compound Data	0.18–0.46 µM
Comparator Or Baseline	AZT: ~0.004 µM
Quantified Difference	AZT is ~45–115× more potent
Conditions	HIV-1 infected human PBMC

Why This Matters

This potency differential establishes that CS-87 cannot be dose-matched with AZT in antiviral assays; procurement for head-to-head mechanistic studies must account for ~100-fold differences in effective concentration.

• [1] NCATS Inxight Drugs. Navuridine (3′-Azido-2′,3′-dideoxyuridine). NCATS. Accessed 2026. View Source
• [2] Mitsuya H, Weinhold KJ, Furman PA, et al. 3′-Azido-3′-deoxythymidine (BW A509U): an antiviral agent that inhibits the infectivity and cytopathic effect of human T-lymphotropic virus type III/lymphadenopathy-associated virus in vitro. Proceedings of the National Academy of Sciences. 1985;82(20):7096–7100. (AZT PBMC EC₅₀ cited within NCATS entry). View Source

Brain Distribution: CS-87 vs AZT

In human bone marrow cell (BMC) cultures, CS-87 demonstrates limited toxicity—a property explicitly contrasted with AZT in the original patent and subsequent mechanistic studies. The patent discloses that CS-87 exhibits “orders of magnitude less cytotoxicity than other anti-viral compounds such as AZT” [1]. Quantitative support comes from cellular metabolism studies: following incubation with 2–10 µM [³H]CS-87, the active triphosphate metabolite (AzdU-TP) was not detected in human BMC, whereas the monophosphate predominated [2]. This sharply contrasts with AZT, whose triphosphate accumulates in BMC and causes dose-limiting myelosuppression [3].

Brain Distribution

Head-to-head

3.7-fold higher brain-to-serum ratio (0.234 vs 0.064) in mice at 50 mg/kg IV; p<0.05

Supports CNS penetration research models

Uninfected murine PK study; single-dose IV

Evidence Dimension	Bone marrow toxicity
Target Compound Data	AzdU-TP undetectable in human BMC; limited toxicity
Comparator Or Baseline	AZT: significant BMC toxicity, dose-limiting myelosuppression
Quantified Difference	Qualitative: “orders of magnitude less”
Conditions	Human bone marrow progenitor cells in vitro

Why This Matters

For experimental systems requiring prolonged nucleoside exposure without confounding myelotoxicity, CS-87 provides a distinct advantage; researchers studying HIV replication in hematopoietic microenvironments should select CS-87 over AZT to avoid off-target bone marrow suppression.

• [1] Chu CK, Schinazi RF. US Patent US4916122A: 3′-Azido-2′,3′-dideoxyuridine anti-retroviral composition. United States Patent and Trademark Office. 1990. View Source
• [2] Zhu Z, Hitchcock MJ, Sommadossi JP. Cellular metabolism of 3′-azido-2′,3′-dideoxyuridine with formation of 5′-O-diphosphohexose derivatives by previously unrecognized metabolic pathways for 2′-deoxyuridine analogs. Molecular Pharmacology. 1990;38(6):929–938. View Source
• [3] Sommadossi JP, Carlisle R, Zhou Z. Cellular pharmacology of 3′-azido-3′-deoxythymidine with evidence of incorporation into DNA of human bone marrow cells. Molecular Pharmacology. 1989;36(1):9–14. View Source

Diphosphohexose Depot: Unique to CS-87

Following intravenous administration of 50 mg/kg to uninfected mice, the brain-to-serum concentration ratio for CS-87 was 0.234 ± 0.282, compared with 0.064 ± 0.025 for AZT (p < 0.05) [1]. This represents a 3.7-fold higher brain penetration for CS-87 at the lower dose. At a higher dose (250 mg/kg), the ratios were 0.103 ± 0.075 for CS-87 and 0.061 ± 0.015 for AZT, indicating that CS-87 maintains a relative brain exposure advantage across a range of doses [1].

Diphosphohexose Depot

Reported

20–30% of intracellular radioactivity as diphosphohexose conjugates after 48 h; t_1/2 14.3 h

Unique intracellular prodrug pathway for sustained nucleotide release research

Human PBMC/BMC incubation; [3H]AzdU

Evidence Dimension	Brain-to-serum concentration ratio
Target Compound Data	0.234 ± 0.282 (50 mg/kg); 0.103 ± 0.075 (250 mg/kg)
Comparator Or Baseline	AZT: 0.064 ± 0.025 (50 mg/kg); 0.061 ± 0.015 (250 mg/kg)
Quantified Difference	3.7-fold higher for CS-87 at 50 mg/kg
Conditions	Male CD-1 mice, single i.v. bolus, HPLC quantification

Why This Matters

In experimental models of HIV-associated neurocognitive disorders or studies requiring CNS antiviral activity, CS-87 offers significantly greater brain exposure than AZT, making it a more appropriate tool compound for investigating HIV replication in the central nervous system.

• [1] Doshi KJ, Gallo JM, Boudinot FD, Schinazi RF, Chu CK. Comparative pharmacokinetics of 3′-azido-3′-deoxythymidine (AZT) and 3′-azido-2′,3′-dideoxyuridine (AZddU) in mice. Drug Metabolism and Disposition. 1989;17(6):590–594. PMID:2575492. View Source

Resistance Mutation: CS-87 K70R vs AZT

In the same murine pharmacokinetic study, total body clearance (Clₜ) values for CS-87 and AZT were comparable at the 50 mg/kg intravenous dose: 1.27 L/hr/kg for CS-87 versus 1.38 L/hr/kg for AZT [1]. Half-life (t½) and steady-state volume of distribution (Vss) were also similar between the two compounds, with both exhibiting dose-dependent decreases in Clₜ and Vss at the higher 250 mg/kg dose, indicating saturable elimination [1].

Resistance Mutation

Reported

Transient K70R mutation in HIV-1 RT; non-overlapping with AZT multi-TAM pathway

Supports resistance selection study design

In vitro selection; genotypic analysis

Evidence Dimension	Total body clearance (Clₜ)
Target Compound Data	1.27 L/hr/kg
Comparator Or Baseline	AZT: 1.38 L/hr/kg
Quantified Difference	8% lower for CS-87
Conditions	Uninfected male CD-1 mice, 50 mg/kg i.v.

Why This Matters

The near-identical systemic clearance simplifies cross-study comparisons and dose extrapolations in murine models; procurement of CS-87 permits researchers to leverage the extensive AZT pharmacokinetic literature while benefiting from CS-87's distinct toxicity and brain penetration advantages.

• [1] Doshi KJ, Gallo JM, Boudinot FD, Schinazi RF, Chu CK. Comparative pharmacokinetics of 3′-azido-3′-deoxythymidine (AZT) and 3′-azido-2′,3′-dideoxyuridine (AZddU) in mice. Drug Metabolism and Disposition. 1989;17(6):590–594. PMID:2575492. View Source

Thymidine Kinase Affinity: CS-87 vs AZT

In human PBMC and BMC incubated with 2–10 µM [³H]CS-87, the predominant intracellular metabolite was the monophosphate (AzdU-MP, ~55–65% of total radioactivity), with diphosphate and triphosphate levels 10- to 100-fold lower [1]. Critically, CS-87 gave rise to two novel metabolites not previously described for 2′-deoxyuridine analogs: 3′-azido-2′,3′-dideoxyuridine-5′-O-diphosphoglucose and 3′-azido-2′,3′-dideoxyuridine-5′-O-diphospho-N-acetylglucosamine, which accumulated to 20–30% of total intracellular radioactivity after 48 hours [1]. These diphosphohexose conjugates exhibited a long intracellular half-life of 14.3 hours and persisted for at least 48 hours after drug removal [1]. No comparable diphosphohexose formation has been reported for AZT.

Thymidine Kinase Affinity

Head-to-head

Ki/Km = 71 vs 0.66 (AZT); 108-fold lower affinity

Tissue-selective activation research context

Human TK enzyme assay; substrate competition

Evidence Dimension	Intracellular metabolite profile
Target Compound Data	Diphosphohexose conjugates: 20–30% of total; t½ = 14.3 h
Comparator Or Baseline	AZT: no diphosphohexose conjugates reported
Quantified Difference	Unique pathway, 14.3 h intracellular retention
Conditions	Human PBMC and BMC, 2–10 µM [³H]CS-87, 48 h incubation

Why This Matters

The formation of long-lived diphosphohexose conjugates introduces a unique intracellular reservoir that may prolong antiviral activity beyond what plasma half-life predicts; this metabolic distinction has significant implications for experimental protocols involving washout periods and for interpreting sustained antiviral effects.

• [1] Zhu Z, Hitchcock MJ, Sommadossi JP. Cellular metabolism of 3′-azido-2′,3′-dideoxyuridine with formation of 5′-O-diphosphohexose derivatives by previously unrecognized metabolic pathways for 2′-deoxyuridine analogs. Molecular Pharmacology. 1990;38(6):929–938. View Source

CS-87: Application Advantages Over AZT

HIV Efficacy with Reduced Bone Marrow Toxicity

CS-87's 3.7-fold higher brain-to-serum ratio at 50 mg/kg compared to AZT [1] makes it the preferred nucleoside reverse transcriptase inhibitor for investigating HIV-1 replication in the central nervous system. Researchers modeling HIV-associated neurocognitive disorders, viral persistence in brain macrophages/microglia, or evaluating blood–brain barrier penetration of antiretroviral regimens should select CS-87 over AZT to achieve pharmacologically relevant CNS exposure without requiring prodrug derivatization.

CNS Distribution in Neuro-HIV Models

Given the undetectable levels of the active triphosphate in human bone marrow cells and the “orders of magnitude less cytotoxicity” compared to AZT [2][3], CS-87 is the appropriate tool compound for experiments requiring sustained antiretroviral pressure in hematopoietic microenvironments. This includes studies of HIV-induced myelosuppression, ex vivo bone marrow progenitor assays, and co-culture systems where AZT's myelotoxicity would confound interpretation.

Intracellular Nucleotide Prodrug Development

The unique formation of long-lived diphosphohexose conjugates (t½ = 14.3 h) in human PBMC and BMC [3] positions CS-87 as a valuable probe for investigating non-canonical nucleoside metabolism. Researchers interested in intracellular drug reservoirs, delayed washout pharmacokinetics, or the role of sugar-nucleotide conjugates in antiviral persistence should utilize CS-87 as a model substrate distinct from AZT and other 3′-azido nucleosides.

Non-TAM Resistance Selection Studies

CS-87's favorable intrinsic brain penetration (brain/serum = 0.234) has been successfully amplified to a relative brain exposure (rₑ) of 5.47 using dihydropyridine prodrug strategies [4]. This established track record makes CS-87 a preferred parent nucleoside for validating novel CNS-targeting prodrug approaches, chemical delivery systems, and nanoparticle formulations aimed at enhancing antiretroviral delivery to the brain.

Application

Selection Property

Validation Focus

Bone marrow-sparing NRTI research

Bone marrow progenitor cell tolerance

Hematopoietic CFU-GM/BFU-E endpoint review

Neuro-HIV CNS penetration studies

Brain-to-serum exposure ratio

CNS pharmacokinetic model validation

Intracellular nucleotide prodrug pathway research

Diphosphohexose depot formation

Intracellular PK/PD modeling

HIV resistance mutation profiling

K70R RT mutation selection

Non-TAM resistance mechanism characterization

• [1] Doshi KJ, Gallo JM, Boudinot FD, Schinazi RF, Chu CK. Comparative pharmacokinetics of 3′-azido-3′-deoxythymidine (AZT) and 3′-azido-2′,3′-dideoxyuridine (AZddU) in mice. Drug Metabolism and Disposition. 1989;17(6):590–594. PMID:2575492. View Source
• [2] Chu CK, Schinazi RF. US Patent US4916122A: 3′-Azido-2′,3′-dideoxyuridine anti-retroviral composition. United States Patent and Trademark Office. 1990. View Source
• [3] Zhu Z, Hitchcock MJ, Sommadossi JP. Cellular metabolism of 3′-azido-2′,3′-dideoxyuridine with formation of 5′-O-diphosphohexose derivatives by previously unrecognized metabolic pathways for 2′-deoxyuridine analogs. Molecular Pharmacology. 1990;38(6):929–938. View Source
• [4] Chu CK, Bhadti VS, Doshi KJ, et al. Brain targeting of anti-HIV nucleosides: synthesis and in vitro and in vivo studies of dihydropyridine derivatives of 3′-azido-2′,3′-dideoxyuridine and 3′-azido-3′-deoxythymidine. Journal of Medicinal Chemistry. 1990;33(8):2188–2192. doi:10.1021/jm00170a023. View Source