Over 20% (15)N Hyperpolarization in Under One Minute for Metronidazole, an Antibiotic and Hypoxia Probe.
Journal: 2017/June - Journal of the American Chemical Society
ISSN: 1520-5126
Abstract:
Direct NMR hyperpolarization of naturally abundant (15)N sites in metronidazole is demonstrated using SABRE-SHEATH (Signal Amplification by Reversible Exchange in SHield Enables Alignment Transfer to Heteronuclei). In only a few tens of seconds, nuclear spin polarization P(15)N of up to ∼24% is achieved using parahydrogen with 80% para fraction corresponding to P(15)N ≈ 32% if ∼100% parahydrogen were employed (which would translate to a signal enhancement of ∼0.1-million-fold at 9.4 T). In addition to this demonstration on the directly binding (15)N site (using J(2)H-(15)N), we also hyperpolarized more distant (15)N sites in metronidazole using longer-range spin-spin couplings (J(4)H-(15)N and J(5)H-(15)N). Taken together, these results significantly expand the range of molecular structures and sites amenable to hyperpolarization via low-cost parahydrogen-based methods. In particular, hyperpolarized nitroimidazole and its derivatives have powerful potential applications such as direct in vivo imaging of mechanisms of action or hypoxia sensing.
Relations:
Content
Citations
(18)
References
(42)
Grants
(4)
Drugs
(1)
Affiliates
(3)
Similar articles
Articles by the same authors
Discussion board
Journal of the American Chemical Society. Jul/5/2016; 138(26): 8080-8083
Published online Jun/19/2016

Over20% 15N Hyperpolarization in UnderOne Minute for Metronidazole, an Antibiotic and Hypoxia Probe

Abstract

Direct NMR hyperpolarizationof naturally abundant 15N sites in metronidazoleis demonstrated using SABRE-SHEATH(Signal Amplification by Reversible Exchange in SHield Enables AlignmentTransfer to Heteronuclei). In only a few tens of seconds, nuclearspin polarization P15N of upto ∼24% is achieved using parahydrogen with 80% para fraction corresponding to P15N ≈ 32% if ∼100% parahydrogen were employed (which wouldtranslate to a signal enhancement of ∼0.1-million-fold at 9.4T). In addition to this demonstration on the directly binding 15N site (using J2H-15N), we also hyperpolarized more distant 15N sites in metronidazole using longer-range spin–spincouplings (J4H-15N and J5H-15N). Taken together, these results significantly expand therange of molecular structures and sites amenable to hyperpolarizationvia low-cost parahydrogen-based methods. In particular, hyperpolarizednitroimidazole and its derivatives have powerful potentialapplications such as direct in vivo imaging of mechanismsof action or hypoxia sensing.

NMR hyperpolarization techniquesincrease nuclear spin polarization from thermal equilibrium levelsof 10–6–10–5 by ordersof magnitude—in some cases approaching 100%.14 Adequately chosen nuclear spinsites can retain their hyperpolarized (HP) state for many minutes,decaying exponentially back to thermal equilibrium; when this decayis slow enough, HP substances can be used as exogenous contrast agentsto probe metabolism and function in vivo.2,5,615N sites are particularlyinteresting because of their greater (e.g., compared to 13C) decay constants of up to 20 min.710 Moreover, 15N isotopic enrichmentis frequently performed using relatively straightforward and efficientchemistries and therefore could be significantly more cost-effectivethan production of 13C-enriched compounds such as 1-13C-pyruvate.11,12 Currently, the rapidly growingfield of HP 13C molecular contrast agents is dominatedby dissolution Dynamic Nuclear Polarization (d-DNP), which typicallyachieves 13C polarization levels well above 10%.13 However, for hyperpolarization of 15N sites, the efficiency of d-DNP has been limited to a few percentand currently requires hours of polarization time. To date, 15N-choline (P15N ≈ 3%in 2.5 h),14 pH sensors (P15N ≤ 3% in 2 h),15 and a calcium sensor (P15N < 2%)16 have been polarizedby d-DNP for potential use as injectable HP contrast agents.

Direct hyperpolarization of heteronuclei with NMR Signal AmplificationBy Reversible Exchange (SABRE17) is possiblevia SABRE-SHEATH (SABRE in SHield Enables Alignment Transfer to Heteronuclei).18,19 Unlike d-DNP, SABRE-SHEATH18,19 is scalable, providesefficient hyperpolarization in seconds, and uses very simple and inexpensivehardware.7,2015N SABRE-SHEATH of several biomoleculeshas been demonstrated (including nicotinamide,11,18 imidazole,21 and diazirines7), producing P15N of up to 10% in ≤1 min. It is our goal to expand therange of biomedically useful compounds amenable to SABRE-SHEATH, whichmay enable unprecedented molecular imaging and spectroscopy experimentsthat would provide direct visualization of biochemical processes.

Despite the success of 15N SABRE-SHEATH, so far thisnew approach has been limited to hyperpolarization of 15N sites directly binding to the polarization-transfer catalyst. Thispathway exploits short-range two-bond spin–spin couplings (J2H-15N) for polarizationtransfer from parahydrogen to 15N sites (Figure 1a) in a μT magnetic field(BT).18,19 Furthermore,sterically hindering groups (e.g., methyl groups) in the ortho-positionsof six-membered N-heterocycles significantly reduce the efficiencyof polarization transfer.22 These shortcomingslimit the use of 15N-SABRE-SHEATH.

Figure 1

(a) Schematic diagramof the SABRE process: coherent polarizationtransfer from parahydrogen-derived hydrides to 15N heteronuclei.(b) Molecular structure of metronidazole. (c) Schematicrepresentation of the AA′B spin system in the polarizationtransfer complex. Red circles represent naturally abundant 15N nuclei. (d) Most-probable metronidazole-bound structuresof the complex with relevant AA′B spin systems.

Here we overcome these limitations and demonstrate 15N SABRE-SHEATH on metronidazole, a member ofan importantgroup of antibiotics—the nitroimidazoles. Nitroimidazoleswere discovered in the 1950s as a class of antibiotics acting on awide range of anaerobic bacterial infections.23 Anaerobic conditions protect the organism’s employment ofpyruvate synthase, which consequently acts irreversibly on the nitroimidazolemoiety—leading to selective drug uptake by anaerobic bacterialcells in the host.23 The sensitivity ofnitroimidazole moieties to anaerobic conditions was laterexploited for hypoxia sensing in cancer. Despite a different mechanismof action23 in mammalian anaerobic cancercells, the end result is the same: the chemical reaction of the nitromoiety leads to irreversible binding in hypoxic cells.24,25 A number of nitroimidazole compounds were employed asmolecular contrast agents to probe hypoxia and upregulation of hypoxiainducible factor (HIF-1α) using Positron Emission Tomography(PET),26 and Magnetic Resonance (MR) imagingand spectroscopy.27 Hypoxia is a hallmarkof some aggressive forms of cancer,28,29 and its imagingcan be used as a biomarker for cancer screening and monitoring responseto treatment.26,30 The investigational use of fluoromisonidazole(FMISO),26 SR4554,27 and numerous other related nitroimidazole-based agentsfor hypoxia imaging has advanced well into clinical trials. Furthermore,the intravenous injection of up to 1.4 g/m2 of nitroimidazolederivatives is well tolerated in patients,27 which would certainly be sufficient for HP imaging.31

In the present SABRE-SHEATH hyperpolarization studies,we employedthe most efficient IrIMes polarization-transfer catalyst (using theestablished Ir catalyst precursor [IrCl(COD)(IMes)]; IMes = 1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene;COD = cyclooctadiene)32,33 and metronidazolewith 15N at natural abundance (∼0.364%). As thisabundance is low, the statistically most-abundant polarization-transferspecies contains only one 15N site. Therefore, the relevantspin system for polarization transfer contains only three spins: twomagnetically inequivalent protons and one 15N site (Figure 1c).22,34 Because metronidazole contains three nitrogen sites,it can form three types of AA′B spin systems, respectivelyinvolving two-bond, four-bond, or five-bond spin–spin couplings(Figure 1d). The 15N NMR signal from nitrogens only two bonds away from thenascent parahydrogen pair was enhanced by ∼72 000-foldat 9.4 T (corresponding to P15N ≈ 24%; see Supporting Information (SI) for additional experimental and analysis details) in 50 mMsolution of metronidazole in methanol-d4 using ∼80% parahydrogen (Figure 2b); these results may be extrapolated toexpect P15N ≈ 32% if∼100% parahydrogen were employed. Note that the 15N polarization was sampled ∼8 s after cessation of parahydrogenbubbling, and the measurement does not take into account polarizationdecay losses that have occurred during the sample transfer from thehyperpolarization setup into the bore of the 9.4 T NMR spectrometer.The P15N value of ∼24%is the highest 15N hyperpolarization level achieved byany technique to date by at least several-fold, and it is demonstrateddespite the presence of one or two ortho substituents, which havepreviously prevented the observation of SABRE-SHEATH enhancement ofsix-membered N-heterocycles.22 This highlevel of P15N is not far awayfrom the practical limit of 50% theoretically predicted in recentliterature.3436 Furthermore, investigations using more concentratedsolutions (∼150 mM, Figure 2c, and ∼100 mM, Figure S5) additionally enabled detection of relatively efficient SABRE-SHEATHhyperpolarization of the other two 15N sites (more weaklycoupled, via J4H-15N and J5H-15N couplings). Although these other two sites exhibited significantlymore modest P15N values of0.1–0.3% (i.e., “only” ∼1000-fold enhancements),it should be noted that neither the μT field nor the exchangerates governing the SABRE-SHEATH process were optimized for hyperpolarizationof these two remote 15N sites. The feasibility of 15N SABRE-SHEATH processes via long-range (up to J5H-15N) spin–spincouplings (in a manner similar to homonuclear SABRE17,32,37) enables 15N hyperpolarizationof a wider range of biomolecules via SABRE-SHEATH, because more 15N and potentially other heteronculear sites become amenableto SABRE-SHEATH.

Figure 2

(a) 15N NMR spectrum of thermally polarizedsignal reference,neat pyridine-15N (∼12.4 M). (b) 15Nspectrum of HP natural abundance 50 mM metronidazole at∼298 K. Note the appearance of NMR resonances as a doubletof quartets due to spin–spin coupling of 15N witharomatic and methyl protons (inset). (c) 15N spectrum ofHP natural abundance 150 mM metronidazole at ∼298K. (d) Optimization of HP metronidazole 15NNMR signal by varying the magnetic field in the shield (BT). (e) Optimization of HP metronidazole 15N NMR signal by varying the temperature (estimated values)of the sample. (f) 15N T1 signaldecay of HP metronidazole in methanol-d4 at 9.4 T. All HP 15N spectra shown were obtainedusing 80% parahydrogen gas.

It should also be emphasized that 15N NMR spectrawithhigh signal-to-noise ratios (SNRs >200 and >100) were recordeddespitethe low natural abundance of 15N and despite using only80% parahydrogen (Figure 2) and 50% parahydrogen (see SI spectra),respectively. Moreover, this SNR level is more than sufficient foroptimization of experimental conditions, i.e., the μT magneticfield for polarization transfer BT, temperature,and measurements of relaxation parameters (see Figure 2d–f). Furthermore, the reported 15N T1 (36 ± 1 s) relaxationis dominated by 15N Chemical Shift Anisotropy (CSA) athigh magnetic field (e.g., 9.4 T used here), and T1 is significantly extended (by several-fold) at the clinicallyrelevant fields of 1.5 T and below.7 Anexpected in vivo15N T1 of 1–2 min14 may besufficient for hypoxia sensing, because nitroimidazole-basedcompounds reach the hypoxic region within the first minute after intravenousinjection.38,39 Moreover, while other tissuesmay also absorb the injected hypoxia sensor,27,38,39 the 15N chemical shifts of thereduced form of the nitroimidazole moiety may be sufficientlydifferent to provide a good mechanism of contrast of hypoxic regionsversus surrounding tissues—mitigating the requirement for clearanceof the background signal (the subject of future work).

In conclusion,record-level 15N hyperpolarization wasachieved via SABRE-SHEATH in seconds for a representative compoundfrom the nitroimidazole class of antibiotics/contrastagents. Substituted five-membered N-heterocycles can be efficientlyhyperpolarized via SABRE-SHEATH even in the presence of ortho substituents. 15N SABRE-SHEATH of distant nitrogen sites via long-range spin–spincouplings is demonstrated. Moreover, high levels of P15N enable detection of HP 15N compoundseven at natural abundance and 50% parahydrogen (see SI for additional 15N spectra), which can be convenientlyutilized for screening of large libraries of compounds. Finally, theadvances presented here will likely be synergistically compatiblewith recent developments of heterogeneous SABRE40,41 and SABRE in aqueous media,20,4245 key elements for future biomedical translation of this work intoanimal models and ultimately clinical trials.

Acknowledgments

This work was supported byNSF under grants CHE-1058727, CHE-1363008,CHE-1416268, and CHE-1416432; NIH under T32EB001628, 1R21EB018014,and 1R21EB020323; DOD CDMRP BRP W81XWH-12-1-0159/BC112431; DODPRMRP awards W81XWH-15-1-0271 and W81XWH-15-1-0272; and ExxonMobilResearch and Engineering Company Knowledge Build. We thank Dr. FanShi for assistance with the SABRE catalyst.

Supporting Information Available

The Supporting Informationis available free of charge on the ACS Publications website at DOI: 10.1021/jacs.6b04784.

  • Additional experimentaldetails, table of calculatedvalues, and NMR spectra in Figures S1–S5 (PDF)

Supplementary Material

ja6b04784_si_001.pdfja6b04784_si_001.pdf

Author Contributions

# D.A.B. and R.V.S.contributed equally.

The authors declare nocompeting financial interest.

References

  • 1. NikolaouP.; GoodsonB. M.; ChekmenevE. Y.Chem. - Eur. J.2015, 21, 3156.[PubMed]
  • 2. KurhanewiczJ.; VigneronD. B.; BrindleK.; ChekmenevE. Y.; CommentA.; CunninghamC. H.; DeBerardinisR. J.; GreenG. G.; LeachM. O.; RajanS. S.; RiziR. R.; RossB. D.; WarrenW. S.; MalloyC. R.Neoplasia2011, 13, 81.[PubMed]
  • 3. BowersC. R.; WeitekampD. P.Phys. Rev. Lett.1986, 57, 2645.[PubMed]
  • 4. MewisR. E.Magn. Reson. Chem.2015, 53, 789.[PubMed]
  • 5. BrindleK. M.J. Am. Chem. Soc.2015, 137, 6418.[PubMed]
  • 6. GolmanK.; in’t ZandtR.; ThaningM.Proc. Natl. Acad. Sci. U. S. A.2006, 103, 11270.[PubMed]
  • 7. TheisT.; OrtizG. X.; LoganA. W. J.; ClaytorK. E.; FengY.; HuhnW. P.; BlumV.; MalcolmsonS. J.; ChekmenevE. Y.; WangQ.; WarrenW. S.Sci. Adv.2016, 2, e1501438.[PubMed]
  • 8. NonakaH.; HataR.; DouraT.; NishiharaT.; KumagaiK.; AkakabeM.; TsudaM.; IchikawaK.; SandoS.Nat.Commun.2013, 4, 2411.[PubMed]
  • 9. GabellieriC.; ReynoldsS.; LavieA.; PayneG. S.; LeachM. O.; EykynT. R.J. Am. Chem. Soc.2008, 130, 4598.[PubMed]
  • 10. PileioG.; CarravettaM.; HughesE.; LevittM. H.J. Am. Chem. Soc.2008, 130, 12582.[PubMed]
  • 11. ShchepinR. V.; BarskiyD. A.; MikhaylovD. M.; ChekmenevE. Y.Bioconjugate Chem.2016, 27, 878.[PubMed]
  • 12. OppenheimerN. J.; MatsunagaT. O.; KamB. L.J. Labelled Compd. Radiopharm.1978, 15, 191.[PubMed]
  • 13. Ardenkjaer-LarsenJ. H.; FridlundB.; GramA.; HanssonG.; HanssonL.; LercheM. H.; ServinR.; ThaningM.; GolmanK.Proc. Natl. Acad.Sci. U. S. A.2003, 100, 10158.[PubMed]
  • 14. CudalbuC.; CommentA.; KurdzesauF.; van HeeswijkR. B.; UffmannK.; JanninS.; DenisovV.; KirikD.; GruetterR.Phys. Chem. Chem. Phys.2010, 12, 5818.[PubMed]
  • 15. JiangW.; LumataL.; ChenW.; ZhangS.; KovacsZ.; SherryA. D.; KhemtongC.Sci. Rep.2015, 5, 9104.[PubMed]
  • 16. HataR.; NonakaH.; TakakusagiY.; IchikawaK.; SandoS.Chem. Commun.2015, 51, 12290.[PubMed]
  • 17. AdamsR. W.; AguilarJ. A.; AtkinsonK. D.; CowleyM. J.; ElliottP. I. P.; DuckettS. B.; GreenG. G. R.; KhazalI. G.; Lopez-SerranoJ.; WilliamsonD. C.Science2009, 323, 1708.[PubMed]
  • 18. TheisT.; TruongM. L.; CoffeyA. M.; ShchepinR. V.; WaddellK. W.; ShiF.; GoodsonB. M.; WarrenW. S.; ChekmenevE. Y.J. Am. Chem. Soc.2015, 137, 1404.[PubMed]
  • 19. TruongM. L.; TheisT.; CoffeyA. M.; ShchepinR. V.; WaddellK. W.; ShiF.; GoodsonB. M.; WarrenW. S.; ChekmenevE. Y.J. Phys. Chem.C2015, 119, 8786.[PubMed]
  • 20. TruongM. L.; ShiF.; HeP.; YuanB.; PlunkettK. N.; CoffeyA. M.; ShchepinR. V.; BarskiyD. A.; KovtunovK. V.; KoptyugI. V.; WaddellK. W.; GoodsonB. M.; ChekmenevE. Y.J. Phys. Chem.B2014, 118, 13882.[PubMed]
  • 21. ShchepinR. V.; BarskiyD. A.; CoffeyA. M.; TheisT.; ShiF.; WarrenW. S.; GoodsonB. M.; ChekmenevE. Y.ACS Sensors2016, .[PubMed]
  • 22. ShchepinR. V.; TruongM. L.; TheisT.; CoffeyA. M.; ShiF.; WaddellK. W.; WarrenW. S.; GoodsonB. M.; ChekmenevE. Y.J. Phys. Chem.Lett.2015, 6, 1961.[PubMed]
  • 23. UpcroftP.; UpcroftJ. A.Clin. Microbiol. Rev.2001, 14, 150.[PubMed]
  • 24. Kizaka-KondohS.; Konse-NagasawaH.Cancer Sci.2009, 100, 1366.[PubMed]
  • 25. MasakiY.; ShimizuY.; YoshiokaT.; TanakaY.; NishijimaK.-i.; ZhaoS.; HigashinoK.; SakamotoS.; NumataY.; YamaguchiY.; TamakiN.; KugeY.Sci. Rep.2015, 5, 16802.[PubMed]
  • 26. HendricksonK.; PhillipsM.; SmithW.; PetersonL.; KrohnK.; RajendranJ.Radiother. Oncol.2011, 101, 369.[PubMed]
  • 27. LeeC. P.; PayneG. S.; OregioniA.; RuddleR.; TanS.; RaynaudF. I.; EatonD.; CampbellM. J.; CrossK.; HalbertG.; TracyM.; McNamaraJ.; SeddonB.; LeachM. O.; WorkmanP.; JudsonI.Br. J. Cancer2009, 101, 1860.[PubMed]
  • 28. HanahanD.; WeinbergR. A.Cell2011, 144, 646.[PubMed]
  • 29. GilliesR. J.; VerduzcoD.; GatenbyR. A.Nat. Rev. Cancer2012, 12, 487.[PubMed]
  • 30. EvansS. M.; JenkinsW. T.; JoinerB.; LordE. M.; KochC. J.Cancer Res.1996, 56, 405.[PubMed]
  • 31. NelsonS. J.; KurhanewiczJ.; VigneronD. B.; LarsonP. E. Z.; HarzstarkA. L.; FerroneM.; van CriekingeM.; ChangJ. W.; BokR.; ParkI.; ReedG.; CarvajalL.; SmallE. J.; MunsterP.; WeinbergV. K.; Ardenkjaer-LarsenJ. H.; ChenA. P.; HurdR. E.; OdegardstuenL. I.; RobbF. J.; TroppJ.; MurrayJ. A.Sci. Transl. Med.2013, 5, 198ra108.[PubMed]
  • 32. CowleyM. J.; AdamsR. W.; AtkinsonK. D.; CockettM. C. R.; DuckettS. B.; GreenG. G. R.; LohmanJ. A. B.; KerssebaumR.; KilgourD.; MewisR. E.J. Am. Chem. Soc.2011, 133, 6134.[PubMed]
  • 33. ApplebyK. M.; MewisR. E.; OlaruA. M.; GreenG. G. R.; FairlambI. J. S.; DuckettS. B.Chem. Sci.2015, 6, 3981.[PubMed]
  • 34. BarskiyD. A.; PravdivtsevA. N.; IvanovK. L.; KovtunovK. V.; KoptyugI. V.Phys. Chem. Chem.Phys.2016, 18, 89.[PubMed]
  • 35. LevittM. H.J. Magn. Reson.2016, 262, 91.[PubMed]
  • 36. KnechtS.; PravdivtsevA. N.; HovenerJ.-B.; YurkovskayaA. V.; IvanovK. L.RSC Adv.2016, 6, 24470.[PubMed]
  • 37. EshuisN.; AspersR. L. E. G.; van WeerdenburgB. J.A.; FeitersM. C.; RutjesF. P. J. T.; WijmengaS. S.; TessariM.J. Magn. Reson.2016, 265, 59.[PubMed]
  • 38. KomarG.; SeppänenM.; EskolaO.; LindholmP.; GrönroosT. J.; ForsbackS.; SipiläH.; EvansS. M.; SolinO.; MinnH.J.Nucl. Med.2008, 49, 1944.[PubMed]
  • 39. ShiK.; SouvatzoglouM.; AstnerS. T.; VaupelP.; NüsslinF.; WilkensJ. J.; ZieglerS. I.J. Nucl. Med.2010, 51, 1386.[PubMed]
  • 40. ShiF.; CoffeyA. M.; WaddellK. W.; ChekmenevE. Y.; GoodsonB. M.Angew. Chem., Int. Ed.2014, 53, 7495.[PubMed]
  • 41. ShiF.; CoffeyA. M.; WaddellK. W.; ChekmenevE. Y.; GoodsonB. M.J. Phys. Chem. C2015, 119, 7525.[PubMed]
  • 42. HövenerJ.-B.; SchwaderlappN.; BorowiakR.; LickertT.; DuckettS. B.; MewisR. E.; AdamsR. W.; BurnsM. J.; HightonL. A. R.; GreenG. G. R.; OlaruA.; HennigJ.; von ElverfeldtD.Anal. Chem.2014, 86, 1767.[PubMed]
  • 43. ZengH.; XuJ.; McMahonM. T.; LohmanJ. A. B.; van ZijlP. C. M.J. Magn. Reson.2014, 246, 119.[PubMed]
  • 44. ShiF.; HeP.; BestQ. A.; GroomeK.; TruongM. L.; CoffeyA. M.; ZimayG.; ShchepinR. V.; WaddellK. W.; ChekmenevE. Y.; GoodsonB. M.J. Phys. Chem. C2016, 120, 12149.[PubMed]
  • 45. SpannringP.; ReileI.; EmondtsM.; SchlekerP. P. M.; HermkensN. K. J.; van der ZwaluwN. G. J.; van WeerdenburgB. J. A.; TinnemansP.; TessariM.; BlümichB.; RutjesF. P. J. T.; FeitersM. C.Chem. - Eur. J.2016, .[PubMed]
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.