Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Dec 5.
Published in final edited form as: Analyst. 2022 Dec 5;147(24):5643–5648. doi: 10.1039/d2an01527g

Synthesis and characterization of a biocompatible 13C1 isotopologue of trityl radical OX071 for in vivo EPR viscometry

Martin Poncelet 1,2, Thacien Ngendahimana 3, Teresa D Gluth 1,2, Emily H Hoblitzell 2,4, Timothy D Eubank 2,4, Gareth R Eaton 3, Sandra S Eaton 3, Benoit Driesschaert 1,2,5
PMCID: PMC9729415  NIHMSID: NIHMS1850261  PMID: 36373434

Abstract

We describe the synthesis, characterization, and application of an isotopologue of the trityl radical OX071, labeled with 13C at the central carbon (13C1). This spin probe features large anisotropy of the hyperfine coupling with the 13C1 (l=1/2), leading to an EPR spectrum highly sensitive to molecular tumbling. The high biocompatibility and lack of interaction with blood albumin allow for systemic delivery and in vivo measurement of tissue microviscosity by EPR.


Low-frequency electron paramagnetic resonance (EPR) associated with a paramagnetic spin probe allows for in vivo measurement of important physiological parameters such as oxygen, pH, thiol concentration, inorganic phosphate, microviscosity, etc16. Nitroxide and triarylmethyl (TAM or trityl) radicals are the two main classes of soluble spin probes used for biomedical in vivo EPR. Nitroxide spin probes were developed first and are available in a large variety of structures, but their use in vivo is generally hindered by their fast reduction, leading to EPR-silent hydroxylamines.

First claimed in patents by Nycomed Innovation in the ‘90s7, 8, Finland trityl (FT) and OX063 and their deuterated analogues; dFT and OX071 (Fig. 1) are among the most popular TAMs. They show unprecedented stability in biological media. For example, OX071 can be recovered from the bladder of a mouse after systemic delivery and reused again.9 In addition to their use as spin probes in vivo, TAM radicals have been extensively utilized for dynamic nuclear polarization10, 11 as spin labels for distance measurements in DNA or proteins using dipolar EPR spectroscopy12, 13 and as spin probes for in vitro detection of biothiols14 or superoxide radical anion15. Contrary to nitroxides which feature g and A anisotropies, making their EPR spectra sensitive to molecular tumbling, spectra of TAMs are generally poorly sensitive to molecular tumbling because of the small g-factor anisotropy and the absence of hyperfine coupling with the predominant 12C (nuclear spin I=0) isotope.

Figure 1.

Figure 1

Structures of Finland trityl (FT), OX063 and their deuterated analogues dFT and OX071 (also named OX063-d24), not sensitive to molecular tumbling and 13C1-dFT, 13C1-PTMTC sensitive to molecular tumbling.

We recently reported a 99% 13C1-labeled dFT (Fig. 1) that is highly sensitive to molecular tumbling.16, 17 This sensitivity arises from the strong anisotropy of the hyperfine coupling with the central 13C1 (A - A = 144 MHz, 51 G), which is not entirely averaged at room temperature in low viscosity media such as PBS. A perchlorinated trityl 99% 13C1 (13C1-PTMTC) with slightly higher anisotropy (A - A = 174 MHz, 62 G) was later reported.18, 19 Abnormal in vivo viscosity has been linked to many diseases, such as cancer. For example, a decrease of 1.1 cP has been reported in murine fibrosarcomas compared to normal tissue.20 EPR based on viscosity-sensitive TAM probes has the potential to non-invasively measure and image tissue microviscosity directly in vivo with application in diagnostic, therapy optimization, or as a tool to study biological mechanisms. However, both Finland- and perchlorinated-based trityl radicals are toxic upon systemic delivery and strongly bind to blood biomacromolecules such as albumin.21 Therefore, only the small, unbound fraction of the probe reports the media microviscosity,12,14 preventing practical applications. OX071, with twelve additional hydroxyethyl moieties, is highly hydrophilic and does not bind albumin. It is also highly biocompatible, with an LD50 of 8 mmol/kg in mice22. OX071 is used routinely with systemic delivery for EPR oximetry in multiple animal models.23, 24

Herein we report the synthesis and characterization of 13C1-OX071 and demonstrate its potential to measure tissue microviscosity in a murine model of breast cancer using L-band EPR spectroscopy and systemic delivery.

1. Synthesis of 13C1-OX071

The synthesis of 13C1-OX071 follows the synthetic strategy that was developed previously for OX063 and OX07125 (Scheme 1) with improvements for larger-scale synthesis and 13C labeling of the central carbon. The synthesis starts with the condensation of dimethyl acetonedicarboxylate with the commercially available tetrathiobenzene 1, leading to the formation of the thioketal 2. The enolizable hydrogens adjacent to the esters were exchanged by deuteriums with CH3OD/CH3ONa in tetrahydrofuran (THF). Subsequent reduction of the four ester groups by lithium aluminum hydride led to the tetra-alcohol 4. Protection of the alcohol groups was achieved by generation of the tert-butyl cation from isobutene and triflic acid in diethyl ether. We previously reported the use of THF as the solvent for the protection, but polymerization of THF under those conditions made the purification very tedious. The reaction in diethyl ether proved to proceed as fast as in THF and with a better yield while preventing any unwanted reaction from the solvent. The recrystallization in ethanol instead of column chromatography allows for large-scale production of 5. Monomer 5 was then converted to the aryl iodide 6 by treatment with a solution of freshly prepared LiTMP at −78°C, followed by the addition of iodine. The previous report used 2.5 equivalents of LiTMP and 2.5 equivalents of iodine.25 Those conditions lead to significant formation of a diiodinated derivative (up to 20%) that is difficult to separate from 6 by chromatography. Reducing the amount of LiTMP to 1.7 equivalents and the amount of iodine to 2 equivalents proved to be efficient, with minimal amounts of the diiodinated derivative formed. These conditions allow for the large-scale synthesis (>30 g per batch) of the key intermediate 6 in five steps with only one purification by column chromatography.

Scheme 1.

Scheme 1

Synthesis of the key intermediate 6.

The next step introduces the 13C central carbon, which is the key feature providing spectral sensitivity to molecular tumbling. The aryl iodide 6 was treated with 1.2 equivalent of sec-BuLi in n-hexane, then a solution of commercially available diethyl carbonate-(carbonyl-13C) was slowly added to afford the triarylmethanol 7 in 74% yield (Scheme 2). The reagent concentrations and nature of both base and solvent and the dropwise addition of the carbonate are critical for the reaction. These conditions favor the formation of the triarylmethanol 7 at the expense of the reaction intermediates such as diaryl ketone and aryl ester and avoid the thioketal ring opening described previously.26 Carboxyl groups were introduced by treatment of 7 with an excess of sec-BuLi in TMEDA at −30°C and dropwise transfer of this solution into a solution of TMEDA pre-saturated with carbon dioxide under constant CO2 bubbling. A mixture of triacid (>75%) and diacid (<15%), as well as traces of the monoacid and the starting material, was obtained. To allow for purification on silica gel for large-scale synthesis, carboxylic acids were esterified using iodomethane and sodium carbonate in DMF. Flash chromatography on silica gel afforded the pure triester 8 in 62% yield. The deprotection of the alcohol groups was achieved by treatment with formic acid for an hour at 60°C, leading to the formyl esters in a quantitative yield. The central carbon was converted to a radical by first generating the corresponding cation using triflic acid, followed by a one electron reduction with tin (II) chloride. The final hydrolysis of the methyl and formyl esters, afforded 13C1-OX071 in 90% yield over three steps.

Scheme 2.

Scheme 2

Synthesis of 13C1-OX071 from intermediate 6.

2. EPR characterization at X-band

The X-band spectrum of 13C1-OX071 sodium salt (13C1-OX071Na) at room temperature in water presents a doublet pattern due to the coupling of the unpaired electron with the central 13C1 (l=1/2) (Fig. 2). The hyperfine coupling of A=23.01±0.04 G (64.43±0.11 MHz) is similar to the one measured for 13C1-dFT (23.35±0.04 G, 65.38±0.11 MHz).16, 17 The residual 1% of 12C at the central carbon is responsible for the smaller sharper peak at the center of the spectrum. Peak-to-peak linewidths of the low- and high-field hyperfine components are 0.87±0.05 G and 0.97±0.05 G, respectively. Both widths are larger than those observed for 13C1-dFT (0.58±0.05 G and 0.64 ±0.05 G in PBS, respectively).17 This is expected for a larger spin probe exhibiting a longer rotational correlation time in the same solvent and temperature. Those linewidths are significantly larger than the single line of the residual 1% of 12C1 (LW=0.062±0.005G)27 and result from incomplete averaging of the anisotropy of the hyperfine coupling at room temperature in water.

Figure 2.

Figure 2

X-band CW EPR spectrum of 13C1-OX071Na (400 μM) in deoxygenated water at 22°C.

Next, 13C1-OX071Na was immobilized in trehalose to determine the spin Hamiltonian parameters. Figure 3 shows the X-band spectrum of 13C1-OX071Na at room temperature in trehalose. The extrema of the large Az are well resolved, allowing determination of Az=160±1 MHz (57.14±0.36 G) and gz=2.0027±0.0001. The two narrow lines at the center of the spectrum indicate small anisotropy in the x,y plane. Ax=Ay=18±1 MHz (6.43±0.36 G) and gx=2.0033±0.0001, and gy=2.0032±0.0001 were determined based on the X-band spectrum and confirmed at Q-band (see ESI Fig. S3). Additional sidebands are attributed to isotopologues with another 13C in addition to the 13C1.17 As expected, those parameters are very close to the ones of 13C1-dFT as the additional hydroxyethyl moieties do not affect the spin distribution significantly (See ESI table S4 for comparison).

Figure 3.

Figure 3

X-band CW spectrum of 13C1-OX071Na in trehalose at 20°C. Simulation parameters are 80% of A(C1) = [18 18 160] MHz, 10% A(C1) plus A(C2, 3,3’)=[29 29 30] MHz, and 10% of A(C1)+A(C4,4’,5) = [14 14 13] MHz. Hstrain parameters for components with 80, 10 and 10% weightings are [4.2 5.5 9] MHz, [7 10 7] MHz and [4 8 9] MHz, respectively. g=[2.0033 2.0032 2.0027]. The uncertainty in A and g values is 1 MHz and 0.0001, respectively.

Next, the spectra of 13C1-Ox071Na were recorded at room temperature in water with glycerol content ranging from 0 to 90% to increase the solvent viscosity. Figure 4 shows the progressive transition from a relatively isotropic doublet spectrum in water to an immobilized spectrum at 90% glycerol. The transition starts from the broadening of the peaks up to 45% glycerol to the resolution of the anisotropic features at higher glycerol content. Both low-field and high-field parts of the spectrum are affected similarly because of the low g anisotropy. Simulations using EasySpin28 and the A and g parameters measured from the immobilized spectrum in trehalose allow determination of the rotational correlation time (τr) for each solution (Table 1), assuming isotropic tumbling.

Figure 4.

Figure 4

X-band EPR spectra (Black) of 13C1-OX071 (400 μM) in deoxygenated water with 0%, 12.5%, 25%, 35%, 45%, 67.5%, 80%, 85% and 90% (V/V) glycerol at 22°C. The spectra were simulated (red dashed lines) using the chili function of EasySpin, using g=[2.0033, 2.0032, 2.0027] and A(MHz)=[18, 18, 160].

Table 1.

Influence of viscosity on τR and linewidth of 13C1-OX071.

Glycerol (%V/V)a Viscosity (cP) τr (ns)b Low-field peak-to-peak Linewidth (G)c High-field peak-to-peak Linewidth (G)c
0 0.96 0.47±0.01 0.87±0.01 0.97±0.01
12.5 1.43 0.66±0.01 1.30±0.03 1.46±0.02
25 2.28 1.12±0.02 2.30±0.01 2.53±0.05
35 3.52 1.83±0.01 3.70±0.04 3.96±0.01
45 5.82 2.93±0.03 6.59±0.04 7.43±0.06
67.5 25.6 12.3±0.1 - -
80 80.3 39.1±0.4 - -
85 140 73.5±0.2 - -
90 260 135±1 - -
a.

(%V/V) Glycerol in water at 22°C.

b.

determined using EasySpin simulations.

c.

Peak-to-peak linewidths measured from the spectrum.

13C1-OX071Na is a ~30% larger molecule than 13C1-dFT; this results in longer correlation times for 13C1-OX071 than for 13C1-dFT for a given viscosity (e.g., 0.47 ns versus 0.27 ns16 at 0% glycerol, 0.96 cP) and therefore explains the larger linewidths observed for 13C1-OX071. Indeed, the rotational correlation time can be related to viscosity by the Stokes-Einstein model τr=ηV/kBT where η is the viscosity, V is the volume of the molecule, kB is the Boltzmann constant, and T is the temperature. Fig. 5A shows the linear correlation between the viscosity and τr, determined by spectral simulation. Below 45% glycerol (<6 cP), the effect of increased viscosity and τr is a line-broadening. Therefore, the linewidths can be calibrated with respect to the viscosity (Fig. 5B, table S2). The linewidth is then an empirical spectral feature that can be used to determine the media viscosity or probe rotational correlation time.

Figure 5.

Figure 5

(A) τR (ns) from spectral simulations versus viscosity (cP). Linear fit leads to the equation τR(ns)=0.52⋅viscosity(cP)-0.05, R2=0.9993. (B) Measured ΔBpp linewidths for the low- and high‐field peaks versus viscosity (cP). Linear fit of the low-field peak leads to the equation ΔBpp(G)=1.13⋅viscosity (cP)-0.25, R2=0.9965.

The linewidth of TAM radicals is known to also be sensitive to the dissolved oxygen concentration and probe concentration (self-broadening). However, the effect of oxygen and concentration on the peak-to-peak linewidth (ca. 0.35 mG/mmHg pO2 and 5 mG/mM for OX071) is minimal by comparison with the line broadening induced by the media viscosity (1.13 G/cP).16, 17 Note that values of viscosity in vivo in a range of 5 cP have been reported using a nitroxide spin probe; this would result in a difference of linewidth of >5 G with 13C1-Ox071.20 The binding of 13C1-dFT to albumin and its toxicity upon systemic delivery limits its application for in vivo EPR tissue viscometry. To verify the absence of interaction between albumin and 13C1-OX071, we compared the spectrum of 13C1-OX071Na in water with the spectrum in the presence of one equivalent of bovine serum albumin (BSA). Figure 6A shows no significant changes on the X-band EPR spectrum of 13C1-OX071 when incubated with BSA demonstrating the absence of interaction with BSA. By contrast, 13C1-dFT incubated with one equivalent of BSA shows a drastic decrease in signal intensity for the doublet (>90%) and the presence of a second spectral component (Fig. 6B, shown with asterisks) corresponding to the probe bound to albumin, tumbling at a slower rate.

Figure 6.

Figure 6

(A) X-band EPR spectra of 13C1-OX071 500 μM in water with or without 500 μM of BSA. (B) X-band EPR spectra of 13C1-dFT 500 μM in water with or without 500 μM of BSA.

3. L-band in vivo EPR viscometry in tumor

Next, we demonstrated the application of 13C1-OX071 to measure tissue microviscosity in a live animal upon systemic delivery of the probe. First, the linewidth of the low-field peak of the doublet was calibrated at L-band with respect to viscosity using water/glycerol mixtures (see ESI table S3, Fig. S1, and S2). Note that for in vivo applications using CW L-band EPR, recording only one peak is more convenient as it allows for shorter acquisition times than recording the entire spectrum.

13C1-OX071 in saline was injected systemically at a dose of ~0.25 mmol/kg in a 107-day-old MMTV/PyMT mouse.29 Figure 7 shows the in vivo L-band EPR spectrum with the loop of the resonator positioned on top of mammary gland number four. The linewidth of 1.1±0.1 G corresponds to a tissue viscosity of 1.14±0.1 cP, in good agreement with literature data.20, 30 This result confirms that 13C1-OX071 can be used to measure tissue viscosity in vivo upon systemic delivery. The study of tissue microviscosity as a biomarker of cancer is beyond the scope of this paper and will be reported elsewhere.

Figure 7.

Figure 7

L-band EPR spectrum of 13C1-OX071 (10 mg in saline solution) injected retro-orbitally in a 107-day-old MMTV/PyMT mouse. The measured peak-to-peak linewidth is 1.1±0.1 G, corresponding to a viscosity of 1.14±0.1 cP.

Conclusions

In conclusion, we have reported an OX071 labeled 99% 13C at the central carbon. The spectral features are close to the previously reported derivative 13C1-dFT. However, the lack of binding to albumin and the high biocompatibility of OX071 allows for systemic delivery and non-invasive in vivo EPR viscometry, which was not possible with 13C1-dFT. 13C1-OX071 or derivatives could also find application in material sciences, in site-directed spin labeling applications, or as polarizing agents for dynamic nuclear polarization.

Supplementary Material

ESI

Acknowledgements

This work was partially supported by NIH grants (USA): R00EB023990 (BD), R21EB028553 (BD), R21GM143595 (BD), R01EB032321 (BD), R01CA194013 (TDE), R01CA262159 (GRE). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Conflicts of interest

The authors declare that there are no conflicts of interest regarding the publication of this research paper.

Electronic Supplementary Information (ESI) available: [details of any supplementary information available should be included here]. See DOI: 10.1039/x0xx00000x

Notes and references

  • 1.Dhimitruka I, Bobko AA, Eubank TD, Komarov DA and Khramtsov VV, J. Am. Chem. Soc, 2013, 135, 5904–5910. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Bobko AA, Eubank TD, Driesschaert B, Dhimitruka I, Evans J, Mohammad R, Tchekneva EE, Dikov MM and Khramtsov VV, Sci. Rep, 2017, 7, 41233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Marchand V, Levêque P, Driesschaert B, Marchand-Brynaert J and Gallez B, Magn. Reson. Med, 2017, 77, 2438–2443. [DOI] [PubMed] [Google Scholar]
  • 4.Marsh D, App. Magn. Reson, 2010, 37, 435–454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Bobko AA, Eubank TD, Voorhees JL, Efimova OV, Kirilyuk IA, Petryakov S, Trofimiov DG, Marsh CB, Zweier JL, Grigor’ev IA, Samouilov A and Khramtsov VV, Magn. Reson. Med, 2012, 67, 1827–1836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Clark A, Sedhom J, Elajaili H, Eaton GR and Eaton SS, Concepts Magn. Reson. Part A, 2016, 45A, e21423. [Google Scholar]
  • 7.Thaning M, WO9839277, 1998.
  • 8.Anderson S, Golman K, Rise F, Wikström H and Wistrand L-G, US5530140, 1996.
  • 9.Serda M, Wu Y-K, Barth ED, Halpern HJ and Rawal VH, Chem. Res. Toxicol, 2016, 29, 2153–2156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Mathies G, Caporini MA, Michaelis VK, Liu Y, Hu K-N, Mance D, Zweier JL, Rosay M, Baldus M and Griffin RG, Ang. Chem. Int. Ed, 2015, 54, 11770–11774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Lumata L, Merritt ME, Malloy CR, Sherry AD and Kovacs Z, J. Phys. Chem. A, 2012, 116, 5129–5138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Shevelev GY, Gulyak EL, Lomzov AA, Kuzhelev AA, Krumkacheva OA, Kupryushkin MS, Tormyshev VM, Fedin MV, Bagryanskaya EG and Pyshnyi DV, J. Ohys. Chem. B, 2018, 122, 137–143. [DOI] [PubMed] [Google Scholar]
  • 13.Yang Z, Liu Y, Borbat P, Zweier JL, Freed JH and Hubbell WL, J. Am. Chem. Soc, 2012, 134, 9950–9952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Tan X, Ji K, Wang X, Yao R, Han G, Villamena FA, Zweier JL, Song Y, Rockenbauer A and Liu Y, Ang. Chem. Int. Ed, 2020, 59, 928–934. [DOI] [PubMed] [Google Scholar]
  • 15.Liu Y, Song Y, De Pascali F, Liu X, Villamena FA and Zweier JL, Free. Rad. Biol. Med, 2012, 53, 2081–2091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Poncelet M and Driesschaert B, Ang. Chem. Int. Ed, 2020, 59, 16451–16454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Moore W, McPeak JE, Poncelet M, Driesschaert B, Eaton SS and Eaton GR, J. Magn. Reson, 2020, 318, 106797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Huffman JL, Poncelet M, Moore W, Eaton SS, Eaton GR and Driesschaert B, J. Ohys. Chem. B, 2021, 125, 7380–7387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Moore W, Huffman JL, Driesschaert B, Eaton SS and Eaton GR, App. Magn. Reson, 2022, 53, 797–808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Halpern HJ, Chandramouli GVR, Barth ED, Yu C, Peric M, Grdina DJ and Teicher BA, Cancer Res., 1999, 59, 5836–5841. [PubMed] [Google Scholar]
  • 21.Song Y, Liu Y, Liu W, Villamena FA and Zweier JL, RSC Adv., 2014, 4, 47649–47656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Krishna MC, English S, Yamada K, Yoo J, Murugesan R, Devasahayam N, Cook JA, Golman K, Ardenkjaer-Larsen JH, Subramanian S and Mitchell JB, Proc. Natl. Acad. Sci. U.S.A, 2002, 99, 2216–2221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Gertsenshteyn I, Epel B, Ahluwalia A, Kim H, Fan X, Barth E, Zamora M, Markiewicz E, Tsai H-M, Sundramoorthy S, Leoni L, Lukens J, Bhuiyan M, Freifelder R, Kucharski A, Giurcanu M, Roman BB, Karczmar G, Kao C-M, Halpern H and Chen C-T, Eur. J. Nucl. Med. Mol. Imaging, 2022, DOI: 10.1007/s00259-022-05889-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Epel B, Haney CR, Hleihel D, Wardrip C, Barth ED and Halpern HJ, Med. Phys, 2010, 37, 2553–2559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Poncelet M, Huffman JL, Khramtsov VV, Dhimitruka I and Driesschaert B, RSC Adv, 2019, 9, 35073–35076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hintz H, Vanas A, Klose D, Jeschke G and Godt A, J. Org. Chem, 2019, 84, 3304–3320. [DOI] [PubMed] [Google Scholar]
  • 27.The linewidth was recorded using a modulation amplitude of 30 mG and non-saturating power.
  • 28.Stoll S and Schweiger A, J. Magn. Reson, 2006, 178, 42–55. [DOI] [PubMed] [Google Scholar]
  • 29.Lin EY, Jones JG, Li P, Zhu L, Whitney KD, Muller WJ and Pollard JW, Am. J. Pathol, 2003, 163, 2113–2126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Velayutham M, Poncelet M, Eubank TD, Driesschaert B and Khramtsov VV, Molecules, 2021, 26, 2781. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ESI

RESOURCES