Membrane-specific spin trap, 5-dodecylcarbamoyl-5-N-dodecylacetamide-1-pyroline-N-oxide (diC12PO): theoretical, bioorthogonal fluorescence imaging and EPR studies

Abstract

Membranous organelles are major endogenous sources of reactive oxygen and nitrogen species. When present at high levels, these species can cause macromolecular damage and disease. To better detect and scavenge free radical forms of the reactive species at their sources, we investigated whether nitrone spin traps could be selectively targeted to intracellular membranes using a bioorthogonal imaging approach. Electron paramagnetic resonance imaging demonstrated that the novel cyclic nitrone 5-dodecylcarbamoyl-5-N-dodecylacetamide-1-pyroline-N-oxide (diC₁₂PO) could be used to target the nitrone moiety to liposomes composed of phosphatidyl choline. To test localization with authentic membranes in living cells, fluorophores were introduced via strain-promoted alkyne-nitrone cycloaddition (SPANC). Two fluorophore-conjugated alkynes were investigated: hexynamide-fluoresceine (HYA-FL) and dibenzylcyclooctyne-PEG4-5/6-sulforhodamine B (DBCO-Rhod). Computational and mass spectrometry experiments confirmed the cycloadduct formation of DBCO-Rhod (but not HYA-FL) with diC₁₂PO in cell-free solution. Confocal microscopy of bovine aortic endothelial cells treated sequentially with diC₁₂PO and DBCO-Rhod demonstrated clear localization of fluorescence with intracellular membranes. These results indicate that targeting of nitrone spin traps to cellular membranes is feasible, and that a bioorthogonal approach can aid the interrogation of their intracellular compartmentalization properties.

Fig. 1.

Nitrone-based spin traps.

Figure 2.

EPR spectra of hydroxyl radical adduct generated from Fe²⁺/H₂O₂ in PBS with 50 mM phosphatidylcholine and 25 mM diC₁₂PO. Simulated parameters (dotted lines): g_x, g_y, g_z: 2.0086, 2.0099, 2.0020; a_Nx, a_Ny, a_Nz: 2 G, 1 G, 30 G; a_H: 13 G.

Fig. 3.

EPR spectra hydroxyl radical adduct of diC₁₂PO generated from A) Fe²⁺/H₂O₂ in PBS in the presence of 50% acetonitrile. a_N = 12.7 G; a_β−H = 10 G; and B) superoxide radical adduct generated from pyridine/H₂O₂, a_N = 12.4 G; a_β−H = 11.7 G. Simulated spectra are represented by dotted plots.

Fig. 4.

Calculated free energies (ΔG_rxn in kcal/mol) of 1,3-dipolar cycloaddition reaction of N-methyl-α-methylnitrone (MMN) with various alkene and alkyne dipolarophiles at the PCM(water)/B3LYP/6–31+G**//B3LYP/6–31G* level of theory.

Fig. 5.

Alkyne dipolarophile-fluorophore conjugates.A, HYA-FL; B, DBCO-Rhod.

Fig. 6.

Calculated free energies (ΔG_rxn in kcal/mol) of 1,3-dipolar cycloaddition reaction of DMPO and PBN with DBCO-Me (top) and HYA-Me (bottom).

Fig. 7.

(A) Fluorescence intensity of HYA-FL (50 μM) in the presence and absence of DMPO (50 μM) and/or Cu⁺ (250 μM). Fluorescence intensity of nitrones (50 μM), PBN and DMPO, and non-nitrone control (50 μM), LA, with differing concentrations of (B) HYA-FL or (C) DBCO-Rhod (0.05, 0.5, 5 and 50 μM) using the CuAAC and non-CuAAC reaction, respectively. Controls are HYA-FL or DBCO-Rhod solutions alone. All measurements were done in triplicate.

Fig. 8.

Flow cytometry detection using BAEC after 5 min and 10 min treatment with (A) HYA-FL (5 μM) under CuAAC reaction and (B) DBCO-Rhod (5 μM) in the presence and absence of PBN (50 μM) and/or S Buffer. HYA-FL was detected on the FITC channel while DBCO was detected using PE channel.

Fig. 9.

Quantitation of cellular fluorescence intensity using ImageJ analysis software (A); confocal microscopy images (B) and Z-stack images (C) of BAEC in the presence of HYA-FL and DBCO-Rhod using the procedure employed for flow cytometry.

Scheme 1.

Illustration of strategy for biorthogonal coupling of nitrones with fluorescent-cyclooctyne tag.

Scheme 2.

Scheme 3.

Conditions: i) 2.2 eq. C₁₂H₂₅NH₂, EDC/HOBt, CH₂Cl₂