A Method for Thermal Generation of Aryloxyl Radicals at Ambient Temperatures: Application to Low-Density Lipoprotein (LDL) Oxidation This work was partly supported by the National Foundation for Cancer Research. We wish to thank M. C. Depew and J. K. S. Wan (Queen's University, Kingston, Canada) for their help in recording ESR spectra and D. Leek for the NMR measurements.

ARTS in chemistry: Aryloxymethyl hyponitrites provide the first aryloxyl radical thermal sources (ARTS) which generate a well-defined flux of aryloxyl radicals (ArO.) at room temperature [Eq. (1)]. These novel compounds can be used for quantitative studies on reactions of ArO. with biomolecules, as exemplified by the PhO.-mediated oxidation of human low-density lipoprotein. Hyponitrites, which are not subject to metal ion- or radical-induced decomposition,9, 10 decompose at ambient temperatures to give N2 and alkoxyl radicals. It appeared probable that aryloxyalkoxyl radicals would undergo very fast β-scission11 to yield aryloxyl radicals [Eq. (3)]. A synthetic route to aryloxyalkyl hyponitrites suitable for many different phenols was designed (Scheme 1)9, 12 and used to prepare 3 a (Ar=Ph; ARTS-Ph) as a potential phenoxyl radical source and 3 b (Ar=tocopheryl; ARTS-Toc) as a potential Toc. source.13 Decomposition rates of 3 a and 3 b were measured by 1H NMR spectroscopy and found to be almost identical, for example, at 37 °C: 9.1×10−4 s−1 for ARTS-Ph (in CD3CN) and 10×10−4 s−1 for ARTS-Toc (in CDCl3). ARTS-Ph decomposed at the same rate in CDCl3 and CD3CN:D2O (1:1) as in dry CD3CN, that is, its decomposition is solvent independent. The Arrhenius parameters for ARTS-Ph decomposition (based on measurements at three temperatures between 23 and 37 °C) were: EA=106 kJ mol−1 and log(A/s−1)=14.8. Synthesis of ARTSs. a) NaH in DMF, 0 °C, 30 min, ClCH2SCH3, 20 °C, 6 h; b) m-ClC6H4CO3H in CH2Cl2, 0 °C, 10 min; c) CH3COCl in CH2Cl2, 0→20 °C, 3 h; d) Ag2N2O2, 0→20 °C, 1–5 h. For steps a–c see ref. 12 and for step d see ref. 9. Physical data for new compounds are given in the Supporting Information. The expected decomposition pathways for an ARTS are outlined in Scheme 2. The initial geminate pair of alkoxyls, 4 sc, will partition between diffusion from the solvent cage (efficiency e) and an in-cage disproportionation to 5 and 6, with the latter probably decomposing to ArOH and formaldehyde. Under most circumstances the vast majority of free alkoxyls, 4, would be expected to undergo β-scission to yield the desired aryloxyls, ArO., with very little H-atom abstraction to form 6. Consistent with this scheme, thermolysis of 3 a gave PhOH, 5 a, formaldehyde, biphenols, phenoxyphenols, and polyphenols and thermolysis of 3 b gave TocH, 5 b, formaldehyde, tocopheryl quinone, and various tocopherol dimers.14 These products are consistent with the formation of at least some free ArO. from both ARTS, this was confirmed with 3 b by the direct detection of Toc. by ESR and UV/Vis spectroscopy (Figure 1). The maximum steady-state concentration of Toc. ([Toc.]mss) was solvent dependent, with 1.1 mM 3 b at 20 °C [Toc.]mss≈5.5 μM in chlorobenzene but was only half as large in cyclohexane and no Toc. could be detected in 1,4-cyclohexadiene. The rate constant for β-scission (kβ) in nonpolar solvents can now be estimated since the rate of β-scission in cyclohexane must be roughly equal to the rate of H-atom abstraction from the cyclohexane, ks×[c-C6H12]=1.2×106 M−1 s−1 15 thus ks×[c-C6H12]=1.2×106 M−1 s−115×9.3 M=1.1×107 s−1. More precisely, we estimate kβ≈1.8×107 s−1 in nonpolar solvents (see Supporting Information) and this reaction will be even faster in polar solvents.11, 15 Indeed, [Toc.]mss in cyclohexene containing 1 M methanol ([Toc.]mss=4 μM) was twice as large as in neat cyclohexene. It is likely that all 4 would undergo β-scission in chlorobenzene. The minimum cage escape efficiency (e) in chlorobenzene at 20 °C can therefore be calculated to be 5 % based on the 3 b decomposition rate constant (1.8×10−4 s−1), [Toc.]mss, and the smallest reported rate constant for the Toc./Toc. reaction (2k=1000 M−1 s−1).14 The presence of a TocH dimer as a minor impurity causes a dramatic increase in the apparent rate constant of the Toc./Toc. reaction14 and since this dimer may be formed during 3 b decomposition the true value of e is likely to be >5 %. Expected thermal decomposition pathways of ARTS. Thermal decomposition of ARTS-Toc. Top: Concentration of Toc. as determined by UV/Vis spectroscopy (λ=424 nm)14 during decomposition of 1 mM ARTS-Toc in chlorobenzene. Inset: UV/Vis spectra recorded at the maximum [Toc.]. Key: 20 °C (◊), 25 °C (□), 37 °C (▵), and 50 °C (○). Bottom: ESR spectrum recorded during decomposition of ARTS-Toc (10 mM) at room temperature in benzene. Hyperfine splitting constants determined by simulation: a(CH3)=0.644 mT, a(CH3)=0.492 mT, a(CH3)=0.104 mT, a(CH2)=0.162 mT, a(CH2)=0.016 mT (linewidth: 0.015 mT). The phenoxyl radical could not be detected during the thermal decomposition of 3 a at 23 °C, presumably because the PhO./PhO. reaction (2k=1–12×108 M−1 s−1)16 is so much faster than the Toc./Toc. reaction.14 However, e for 3 a, could be estimated by assuming that the in-cage reaction of 4 sc would yield equal amounts of 5 a and PhOH whereas GC analysis showed an excess of PhOH and biphenols over 5 a. From the “excess” phenol, e was estimated to be around 20 %. The aryloxyl radical-initiated peroxidation of LDL was chosen to illustrate a biologically relevant in vitro application of ARTS. To a freshly prepared LDL dispersion 3 a was added and incubated at 37 °C until decomposion was virtually complete. The TocH consumption curve and the cholesteryl ester hydroperoxide (CEOOH) formation curve are characteristic of TMP in that peroxidation is faster while TocH is present than after the TocH is consumed6 (Figure 2). Furthermore, at low initial 3 a concentrations where significant amounts of TocH remain after the 1 h incubation, the CEOOH was formed in a chain reaction, for example d[CEOOH]/d[PhO.]≈16 at an initial [3 a]=35 μM (see Supporting Information). That is, TMP is initiated in LDL by PhO. attack on TocH [Eq. (4)]. TMP explains the earlier observation that tyrosyl radicals generated by myeloperoxidase initiate LDL peroxidation in a process not inhibited by TocH.7b Concentration of cholesteryl ester hydroperoxides (CEOOH) (⧫) and α-tocopherol (TocH) (▪) in 1.8 μM LDL dispersed in aerated phosphate buffered saline (PBS; pH 7.4, 50 mM) after incubation for 1 h at 37 °C in the presence of the indicated initial ARTS-Ph concentrations. ARTS-Ph was added as a solution in DMSO, the amount of which did not exceed 1 % of the 300 μL LDL dispersion. (PhO.)aq+(TocH)LDL → (PhOH)aq+(Toc.)LDL At low 3 a concentration, approximately 0.2 molecules of TocH are consumed per PhO. generated (see Supporting Information). This value would be 0.5 if all the PhO. were destroyed by TocH and it implies that PhO./PhO. coupling reactions are probably important under the present conditions. Aryloxyl-radical induced oxidative stress has been largely ignored because of the lack of suitable precursors. The present synthesis of two ARTS overcomes this lack and provides a new tool for studying the effects of known fluxes of ArO. radicals on biologically relevant targets. Currently, we are designing a synthesis for water-soluble ARTSs which will enable the tyrosyl radical to be generated in a controlled manner. The Toc. EPR spectrum was recorded at room temperature under N2 on a Varian E104 spectrometer (9.5 GHz) with microwave power=2 mW, modulation amplitude=0.04 mT, modulation frequence=100 kHz, scan time=8 min. The hyperfine splitting constants were obtained using the ESR simulation program WINSIM.17 NMR spectroscopic data were recorded on a Bruker 400-DRX spectrometer. Freshly isolated LDL,18 was dispersed in aerated PBS, mixed with a known amount of a 3 a stock solution in DMSO and then incubated for 1 h at 37 °C with the usual analyses for CEOOH and TocH.19 Supporting information for this article is available on the WWW under http://www.angewandte.com or from the author. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.