C16H23FN4O2

• Molecular Formula: C16H23FN4O2
• Molecular Weight: 322.38 g/mol
• SMILES: CC(C)NC(=O)CNC(=O)N1CCN(CC1)C2=CC=C(C=C2)F
• InChI: InChI=1S/C16H23FN4O2/c1-12(2)19-15(22)11-18-16(23)21-9-7-20(8-10-21)14-5-3-13(17)4-6-14/h3-6,12H,7-11H2,1-2H3,(H,18,23)(H,19,22)
• InChIKey: UAYHSCQWHHMNSH-UHFFFAOYSA-N

The compound with molecular formula C₁₆H₂₃FN₄O₂ is established as 8‑cyclopentyl‑3‑(3‑fluoropropyl)‑1‑propylxanthine, commonly designated CPFPX (CAS 227205‑79‑0). It belongs to the xanthine class of adenosine receptor antagonists and is structurally a fluorine isostere of the prototypical A₁ adenosine receptor (A₁AR) antagonist DPCPX (8‑cyclopentyl‑1,3‑dipropylxanthine).

CPFPX (C₁₆H₂₃FN₄O₂) – A1 Adenosine Receptor Radioligand for PET Imaging: Compound Identity and Procurement Baseline

The compound with molecular formula C₁₆H₂₃FN₄O₂ is established as 8‑cyclopentyl‑3‑(3‑fluoropropyl)‑1‑propylxanthine, commonly designated CPFPX (CAS 227205‑79‑0) [1]. It belongs to the xanthine class of adenosine receptor antagonists and is structurally a fluorine isostere of the prototypical A₁ adenosine receptor (A₁AR) antagonist DPCPX (8‑cyclopentyl‑1,3‑dipropylxanthine) [2]. CPFPX is primarily deployed as the ¹⁸F‑labeled radioligand [¹⁸F]CPFPX for positron emission tomography (PET) imaging of cerebral A₁AR density in human and preclinical studies [3]. Its pharmacological profile features sub‑nanomolar affinity for the A₁AR (bovine Ki = 0.183 nM; human KD = 1.26 nM) and >700‑fold selectivity over the A₂A adenosine receptor [4]. Available from multiple specialist vendors as a research‑grade radioligand precursor (typically 5–25 mg, pricing $900–$3,000), procurement decisions hinge on quantifiable differentiation from alternative A₁AR antagonists and PET tracers such as DPCPX, [¹¹C]MPDX, FK453, and caffeine.

• [1] PubChem Compound Summary CID 5461879, 'Cpfpx', National Center for Biotechnology Information, accessed 2026. View Source
• [2] Holschbach, M.H. et al. (2002) 'Synthesis and Evaluation of No‑Carrier‑Added 8‑Cyclopentyl‑3‑(3‑[¹⁸F]fluoropropyl)‑1‑propylxanthine ([¹⁸F]CPFPX): A Potent and Selective A₁‑Adenosine Receptor Antagonist for in Vivo Imaging', Journal of Medicinal Chemistry, 45(23), pp. 5150–5156. doi:10.1021/jm020905i. View Source
• [3] Meyer, P.T. et al. (2004) 'Quantification of cerebral A₁ adenosine receptors in humans using [¹⁸F]CPFPX and PET', Journal of Cerebral Blood Flow & Metabolism, 24(3), pp. 323–333. doi:10.1097/01.WCB.0000110531.48786.9D. View Source
• [4] BindingDB Entry BDBM50062852, Ki summary for adenosine A₁ receptor (bovine), accessed 2026. View Source

Why A₁ Adenosine Receptor Antagonists and PET Tracers Cannot Be Interchanged: The CPFPX Differentiation Case

Compounds within the A₁ adenosine receptor antagonist class exhibit profound differences in affinity, selectivity, metabolic stability, and suitability for in vivo imaging that render simple generic substitution invalid. CPFPX (C₁₆H₂₃FN₄O₂) demonstrates a Ki of 0.183 nM at bovine A₁AR [1], in contrast to DPCPX which shows a Ki of 3.9 nM at human A₁AR, representing a >21‑fold difference in binding potency . The selectivity window between A₁ and A₂A receptors diverges dramatically: CPFPX achieves >1200‑fold selectivity (rat A₁/A₂A) whereas DPCPX exhibits only ~33‑fold selectivity (human A₁/A₂A) [2]. Metabolic handling further differentiates these agents; CPFPX is selectively metabolized by CYP1A2 with a KM of 1.1 μM, a pathway not shared by all xanthine‑based antagonists, and produces polar metabolites that do not cross the blood‑brain barrier [3]. For PET imaging applications, [¹⁸F]CPFPX provides a validated two‑tissue compartment model in humans, a quantitative framework that cannot be assumed for alternative tracers such as [¹¹C]MPDX (Ki = 4.2 nM, ~23‑fold lower A₁ affinity) [4]. In vivo caffeine competition studies confirm that [¹⁸F]CPFPX PET detects concentration‑dependent A₁AR occupancy with a half‑maximal displacement at a plasma caffeine concentration of 67 μM, establishing it as a quantitative pharmacodynamic biomarker [5]. Each of these dimensions—binding affinity, subtype selectivity, metabolic pathway, kinetic modeling suitability, and in vivo target engagement—must be verified against the specific experimental or procurement requirement.

• [1] BindingDB Entry BDBM50062852, Ki summary for adenosine A₁ receptor (bovine), accessed 2026. View Source
• [2] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, selectivity data: rat A₁/A₂A >1200‑fold, human A₁/A₂A >700‑fold. View Source
• [3] Bier, D. et al. (2006) 'Metabolism of the A₁ adenosine receptor PET ligand [¹⁸F]CPFPX by CYP1A2: implications for bolus/infusion PET studies', Nuclear Medicine and Biology, 33(8), pp. 1043–1051. doi:10.1016/j.nucmedbio.2006.08.005. View Source
• [4] Ishiwata, K. et al. (2002) 'Mapping adenosine A₁ receptors in the cat brain by positron emission tomography with [¹¹C]MPDX', Nuclear Medicine and Biology, 29(1), pp. 29–37. doi:10.1016/S0969-8051(01)00265-7. View Source
• [5] Elmenhorst, D. et al. (2012) 'Caffeine occupancy of human cerebral A₁ adenosine receptors: in vivo quantification with ¹⁸F‑CPFPX and PET', Journal of Nuclear Medicine, 53(11), pp. 1723–1729. doi:10.2967/jnumed.112.105114. View Source

CPFPX (C₁₆H₂₃FN₄O₂) – Quantitative Evidence for Scientific Selection

A₁AR Binding Affinity – CPFPX vs. DPCPX (Xanthine Comparator)

CPFPX exhibits substantially higher binding affinity for the A₁ adenosine receptor than the prototypical xanthine antagonist DPCPX. In bovine cortex membranes, CPFPX demonstrates a Ki of 0.183 nM [1], whereas DPCPX shows a Ki of 3.9 nM at cloned human A₁AR . In the same bovine system using [³H]DPCPX as the reference ligand (KD = 0.15 nM), CPFPX yielded a Ki of 0.18 nM with an A₂A/A₁ selectivity ratio >500 [2]. At cloned human A₁AR, the KD of [³H]CPFPX was 1.26 nM [2], still representing a ~3‑fold higher affinity than DPCPX at the human receptor.

Evidence Dimension	A₁ adenosine receptor binding affinity (in vitro)
Target Compound Data	Ki = 0.183 nM (bovine A₁AR); KD = 1.26 nM (cloned human A₁AR)
Comparator Or Baseline	DPCPX: Ki = 3.9 nM (human A₁AR); Ki = 0.18 nM in same bovine assay as reference
Quantified Difference	~21‑fold higher affinity vs. human DPCPX Ki; ~3‑fold higher vs. human DPCPX KD comparison
Conditions	Bovine cortex membrane [³H]DPCPX competition binding; cloned human A₁AR saturation binding

Why This Matters

Higher affinity at the target receptor enables lower tracer mass doses in PET studies, reducing potential pharmacological effects and improving specific‑to‑non‑specific binding ratios.

• [1] BindingDB Entry BDBM50062852, Ki = 0.183 nM, adenosine A₁ receptor (bovine), accessed 2026. View Source
• [2] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, KD = 1.26 nM (human A₁AR), Ki = 0.18 nM (bovine). View Source

A₁/A₂A Subtype Selectivity – CPFPX vs. DPCPX and [¹¹C]MPDX

CPFPX provides substantially superior discrimination between the A₁ and A₂A adenosine receptor subtypes compared with both DPCPX and [¹¹C]MPDX. In rat striatal membranes, CPFPX shows a Ki of 812 nM at A₂AAR, yielding an A₁/A₂A selectivity ratio >1200‑fold [1]. At the cloned human A₂AAR, the KD of [³H]CPFPX was 940 nM, corresponding to >700‑fold selectivity over the human A₁AR [1]. In contrast, DPCPX exhibits A₂A Ki values of 130 nM (human) and 149 nM (rat male) [2], giving selectivity ratios of only ~33‑fold (human) and ~330‑fold (rat). [¹¹C]MPDX has reported A₂A Ki values >100 nM, yielding a selectivity ratio of ~24‑fold given its A₁ Ki of 4.2 nM [3].

Evidence Dimension	A₁/A₂A adenosine receptor selectivity ratio
Target Compound Data	A₁/A₂A >1200‑fold (rat); >700‑fold (human)
Comparator Or Baseline	DPCPX: A₁/A₂A ~33‑fold (human), ~330‑fold (rat); [¹¹C]MPDX: A₁/A₂A ~24‑fold
Quantified Difference	CPFPX selectivity exceeds DPCPX by ~21‑36‑fold and exceeds [¹¹C]MPDX by ~30‑50‑fold
Conditions	Rat striatal [³H]CGS21680 binding; cloned human A₂AAR saturation binding; literature comparator values

Why This Matters

Higher A₁/A₂A selectivity minimizes confounding signal from striatal A₂A receptors in PET quantification, critical for studies in brain regions where both subtypes are co‑expressed (e.g., striatum).

• [1] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, rat A₂A Ki = 812 nM, human A₂A KD = 940 nM. View Source
• [2] Springer, BMC Research Notes (2023) 'On the basis of sex: male vs. female rat adenosine A₁/A₂A receptor affinity', rat DPCPX A₂A Ki = 149 ± 23 nM (male). View Source
• [3] MICAD/NIH, '[¹¹C]MPDX' entry, reported A₁ Ki = 4.2 nM, A₂A Ki >100 nM, accessed 2026. View Source

Metabolic Pathway Specificity – CPFPX CYP1A2 Metabolism vs. Caffeine

CPFPX is metabolized by a single, well‑defined cytochrome P450 isoform, CYP1A2, with high enzymatic selectivity. Kinetic analysis in pooled human liver microsomes determined KM = 1.1 μM (95% CI: 0.6–2.0 μM) and Vmax = 243 pmol·min⁻¹·mg⁻¹, corresponding to a catalytic efficiency of 2.4 pmol·min⁻¹·pmol⁻¹ CYP [1]. This metabolism is competitively inhibited by fluvoxamine with a KI of 68 nM (95% CI: 34–138 nM) [1]. In contrast, the non‑selective antagonist caffeine is metabolized by multiple CYP isoforms (primarily CYP1A2 but also CYP2E1, CYP3A4) and demonstrates pH‑dependent, multi‑exponential pharmacokinetics in vivo, complicating its use as a quantitative pharmacological probe [2]. Brain metabolite analysis confirmed that >98% of radioactivity in brain homogenates at 60 min post‑injection represents intact [¹⁸F]CPFPX, with polar metabolites confined to the periphery [3].

Evidence Dimension	Metabolic enzyme specificity and brain stability
Target Compound Data	CYP1A2 KM = 1.1 μM; Vmax = 243 pmol·min⁻¹·mg⁻¹; brain intact fraction >98% at 60 min
Comparator Or Baseline	Caffeine: multi‑CYP metabolism (CYP1A2, CYP2E1, CYP3A4); variable human pharmacokinetics (class‑level inference)
Quantified Difference	Single‑isoform metabolism vs. multi‑isoform; >98% brain stability vs. caffeine's rapid brain equilibration and multi‑exponential washout
Conditions	Pooled human liver microsomes; mouse brain homogenate radio‑TLC analysis at 60 min post‑injection

Why This Matters

A single, well‑characterized metabolic pathway enables predictable pharmacokinetic modeling, facilitates correction for radiolabeled metabolites in plasma during PET quantification, and allows mechanistic interpretation of altered metabolism in patient populations or drug–drug interaction studies.

• [1] Bier, D. et al. (2006) Nucl. Med. Biol., 33(8), pp. 1043–1051, KM = 1.1 μM, Vmax = 243 pmol·min⁻¹·mg⁻¹, fluvoxamine KI = 68 nM. View Source
• [2] Thorn, C.F. et al. (2012) 'PharmGKB summary: caffeine pathway', Pharmacogenetics and Genomics, 22(5), pp. 389–395, caffeine metabolism by CYP1A2, CYP2E1, CYP3A4. doi:10.1097/FPC.0b013e328350f4d4. View Source
• [3] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, brain metabolite stability >98%. View Source

In Vivo Human PET Quantification – [¹⁸F]CPFPX vs. [¹¹C]MPDX

[¹⁸F]CPFPX is validated for quantitative human cerebral A₁AR imaging using a two‑tissue compartment model (2TCM), providing regional total distribution volume (DVt) estimates. In six healthy volunteers, the 2TCM described cerebral kinetics significantly better than a one‑tissue compartment model (1TCM) across all brain regions examined (linear regression between 2TCM and Logan graphical analysis: r = 0.99, slope = 1.007) [1]. Displacement studies confirmed that specific binding accounts for approximately one‑third of total equilibrium uptake in cerebellum and two‑thirds in cortex, enabling reference tissue approaches in selected applications [2]. In comparison, [¹¹C]MPDX, despite being used in human PET, has a Ki of 4.2 nM at A₁AR—~23‑fold lower affinity than CPFPX—and demonstrated faster brain washout kinetics in cat studies [3]. The carbon‑11 half‑life (20.4 min) of [¹¹C]MPDX limits scanning protocols to ~90 min, whereas fluorine‑18 (109.6 min) permits bolus/infusion equilibrium protocols extending >140 min with [¹⁸F]CPFPX [4].

Evidence Dimension	In vivo human brain PET quantification and isotope half‑life
Target Compound Data	2TCM with DVt; r = 0.99 vs. Logan GA; ¹⁸F half‑life = 109.6 min; specific binding fraction: cortex ~66%, cerebellum ~33%
Comparator Or Baseline	[¹¹C]MPDX: Ki = 4.2 nM; faster washout; ¹¹C half‑life = 20.4 min
Quantified Difference	~23‑fold higher A₁ affinity; 5.4‑fold longer isotope half‑life; validated 2TCM with cross‑validation r = 0.99
Conditions	Human dynamic PET with arterial blood sampling; bolus/infusion protocol

Why This Matters

The longer fluorine‑18 half‑life enables multi‑subject imaging from a single radiosynthesis batch, reduces logistical constraints for clinical research centers, and permits equilibrium infusion protocols that are impractical with carbon‑11 tracers.

• [1] Meyer, P.T. et al. (2004) J. Cereb. Blood Flow Metab., 24(3), pp. 323–333, 2TCM > 1TCM, r = 0.99 vs. Logan GA. View Source
• [2] Meyer, P.T. et al. (2006) NeuroImage, 32(3), pp. 1100–1105, displacement studies: specific binding ~1/3 cerebellum, ~2/3 cortex. View Source
• [3] Ishiwata, K. et al. (2002) Nucl. Med. Biol., 29(1), pp. 29–37, [¹¹C]MPDX Ki = 4.2 nM. View Source
• [4] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, ¹⁸F half‑life = 109.6 min, radiosynthesis yield 45 ± 7%. View Source

In Vivo Target Engagement – CPFPX Displacement by Caffeine in Human Brain

[¹⁸F]CPFPX PET enables quantitative measurement of in vivo A₁AR occupancy by competing ligands in the human brain. In a bolus‑plus‑constant‑infusion study of 15 healthy volunteers, caffeine infusion displaced 5–44% of [¹⁸F]CPFPX binding in a concentration‑dependent manner. The half‑maximal inhibitory plasma concentration (IC₅₀) for caffeine was determined to be 67 μM, which corresponds to approximately 450 mg caffeine in a 70‑kg subject or ~4.5 cups of coffee [1]. Placebo infusion produced no change in radioligand binding. This occupancy assay platform is not achievable with non‑radiolabeled antagonists such as DPCPX or FK453, which lack PET imaging capability, and provides superior quantitative resolution compared to [¹¹C]MPDX due to the 5.4‑fold longer isotope half‑life [2].

Evidence Dimension	In vivo receptor occupancy quantification in human brain
Target Compound Data	Caffeine IC₅₀ = 67 μM plasma; occupancy range 5–44%
Comparator Or Baseline	DPCPX / FK453: no PET imaging capability; [¹¹C]MPDX: shorter scan window
Quantified Difference	CPFPX uniquely enables human occupancy studies; IC₅₀ determined with high precision; placebo‑controlled design
Conditions	Human bolus‑plus‑constant‑infusion PET; 15 subjects; caffeine doses 0.5–4.3 mg/kg i.v.

Why This Matters

This capability transforms CPFPX from a mere receptor ligand into a pharmacodynamic biomarker platform, enabling dose‑finding studies for A₁AR‑targeted therapeutics and mechanistic investigation of endogenous adenosine signaling in human neurological research.

• [1] Elmenhorst, D. et al. (2012) J. Nucl. Med., 53(11), pp. 1723–1729, caffeine IC₅₀ = 67 μM, occupancy 5–44%. View Source
• [2] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, ¹⁸F half‑life vs. ¹¹C. View Source

Provenance of Affinity – CPFPX Metabolites Retain High A₁ Selectivity (Supporting Evidence)

CPFPX undergoes peripheral metabolism to at least seven identified compounds, primarily via oxidation at the cyclopentyl moiety, without N‑dealkylation [1]. The putative metabolites displace [³H]CPFPX binding to A₁AR in pig brain cortex with Ki values between 1.9 and 380 nM, and displace [³H]ZM241385 binding to A₂AAR with Ki values >180 nM [2]. One metabolite, functionalized at the ω‑position of the N1‑propyl chain, retained high A₁AR affinity (Ki = 2 nM) and >9000‑fold selectivity over A₂AAR [2]. Importantly, none of these polar metabolites cross the blood‑brain barrier, as demonstrated by reinjection experiments in mice [1]. This contrasts with some xanthine analogs whose metabolites may retain receptor activity and penetrate the CNS, potentially confounding pharmacological interpretation.

Evidence Dimension	Metabolite receptor binding and brain penetration
Target Compound Data	Metabolite Ki range 1.9–380 nM (A₁AR); Ki >180 nM (A₂AAR); one metabolite Ki = 2 nM with >9000‑fold A₁/A₂A selectivity; no brain penetration
Comparator Or Baseline	Class‑level inference: other xanthine antagonists may produce CNS‑penetrant active metabolites (exceptions noted but data not systematically available)
Quantified Difference	Peripheral confinement of all metabolites; retained A₁AR affinity in metabolites does not confound brain PET signal
Conditions	Pig brain cortex [³H]CPFPX displacement; pig striatum [³H]ZM241385 displacement; mouse metabolite reinjection experiments

Why This Matters

The peripheral confinement of CPFPX metabolites eliminates a potential source of PET signal contamination in brain imaging, ensuring that measured radioactivity reflects intact parent tracer binding to A₁AR. This provides a cleaner pharmacological signal than tracers whose metabolites may enter the brain and bind to off‑target receptors.

• [1] Holschbach, M.H. et al. (2024) 'Metabolism of the A₁ adenosine receptor positron emission tomography ligand [¹⁸F]CPFPX in rodents and humans', Nuclear Medicine and Biology, 128–129, 108872. doi:10.1016/j.nucmedbio.2024.108872. View Source
• [2] Holschbach, M.H. et al. (2017) 'Synthesis and Pharmacological Evaluation of Identified and Putative Metabolites of the A₁ Adenosine Receptor Antagonist CPFPX', ChemMedChem, 12(10), pp. 770–779. doi:10.1002/cmdc.201600592. View Source

High‑Value Application Scenarios for CPFPX (C₁₆H₂₃FN₄O₂) Based on Quantitative Evidence

Human Cerebral A₁AR Density Quantification in Neurological Disorders

The validated two‑tissue compartment model and >700‑fold A₁/A₂A selectivity of [¹⁸F]CPFPX make it the preferred radioligand for quantitative PET studies of regional A₁AR density changes in conditions such as Alzheimer's disease, epilepsy, and ischemic stroke [1]. The >98% brain metabolic stability at 60 min post‑injection ensures that PET signal in cortical and subcortical regions reflects authentic A₁AR binding rather than radiolabeled metabolites [2]. In contrast, the ~33‑fold selectivity of DPCPX would introduce substantial A₂A contamination in striatal measurements, and the 20.4‑min half‑life of ¹¹C limits [¹¹C]MPDX to single‑subject protocols per radiosynthesis .

Caffeine Pharmacodynamic Occupancy Studies and Sleep Research

[¹⁸F]CPFPX PET is uniquely qualified for measuring in vivo cerebral A₁AR occupancy by caffeine (IC₅₀ = 67 μM plasma) and other competing ligands, providing a direct pharmacodynamic readout for sleep research and neurostimulant studies [1]. The bolus‑plus‑constant‑infusion protocol sustained over 140 min, enabled by the 109.6‑min fluorine‑18 half‑life, permits equilibrium measurements that are technically unfeasible with carbon‑11 tracers [2]. This application cannot be performed with non‑radiolabeled antagonists (DPCPX, FK453) and represents a distinct procurement rationale for [¹⁸F]CPFPX.

Preclinical Rodent PET for Longitudinal A₁AR Studies

In rodent models, [¹⁸F]CPFPX PET has been validated for non‑invasive, reproducible quantification of A₁AR density, with kinetics well‑described by compartmental and non‑compartmental models based on arterial input function [1]. Displacement by DPCPX confirmed specific binding, and the high specific radioactivity (>270 GBq/μmol) permits tracer‑dose injections that do not perturb endogenous adenosine signaling [2]. The CYP1A2‑mediated metabolism (KM = 1.1 μM) is conserved across mice, rats, and humans, facilitating translational pharmacokinetic modeling . This is particularly valuable for longitudinal studies of chronic caffeine consumption, neuroinflammation, and glioma invasion where repeated within‑subject measurements are required.

Drug–Drug Interaction Studies Involving CYP1A2 Modulators

Because CPFPX is metabolized exclusively by CYP1A2 with well‑characterized kinetics (KM = 1.1 μM, Vmax = 243 pmol·min⁻¹·mg⁻¹) and competitive inhibition by fluvoxamine (KI = 68 nM), it serves as a probe substrate for studying CYP1A2 activity in vivo [1]. Altered CPFPX metabolism has been demonstrated in patients with liver cirrhosis and those receiving CYP1A2 inhibitors, confirming its utility in clinical PET studies of hepatic drug metabolism [1]. This application exploits the metabolic specificity that differentiates CPFPX from multi‑pathway substrates such as caffeine [2] and is not addressable with non‑PET A₁AR antagonists.

• [1] Meyer, P.T. et al. (2004) J. Cereb. Blood Flow Metab., 24(3), pp. 323–333; Bier, D. et al. (2006) Nucl. Med. Biol., 33(8), pp. 1043–1051; Elmenhorst, D. et al. (2012) J. Nucl. Med., 53(11), pp. 1723–1729; Elmenhorst, D. et al. (2013) J. Nucl. Med., 54(8), pp. 1411–1419; Nabbi‑Schroeter, D. et al. (2018) Mol. Imaging Biol., 20(2), pp. 284–291. View Source
• [2] Holschbach, M.H. et al. (2002) J. Med. Chem., 45(23), pp. 5150–5156, brain stability, specific activity, radiosynthesis. View Source