Nuclear magnetic resonance at the picomole level of a DNA adduct

Abstract

We investigate the limit of detection for obtaining NMR data of a DNA adduct using modern microscale NMR instrumentation, once the adduct has been isolated at the picomole level. Eighty nanograms (130 pmol) of a DNA adduct standard, N-(2'-deoxyguanosin-8-yl)-2-acetylaminofluorene 5'-monophosphate (AAF-dGMP), in 1.5 μL of D₂O with 10% methanol-d₄, in a vial, was completely picked up as a droplet suspended in a fluorocarbon liquid and loaded efficiently into a microcoil probe. This work demonstrates a practical manual method of droplet microfluidic sample loading, previously demonstrated using automated equipment, which provides a severalfold advantage over conventional flow injection. Eliminating dilution during injection and confining the sample to the observed volume produce the full theoretical mass sensitivity of a microcoil, comparable to that of a microcryo probe. With 80 ng, an NMR spectrum acquired over 40 h showed all of the resonances seen in a standard spectrum of AAF-dGMP, with a signal-to-noise ratio of at least 10, despite broadening due to previously noted effects of conformational exchange. Even with this broadening to 5 Hz, a two-dimensional total correlation spectroscopy spectrum was acquired on 1.6 μg in 18 h. This work helps to define the utility of NMR in combination with other analytical methods for the structural characterization of a small amount of a DNA adduct.

Figure 1

Microdroplet loading method. (A) Schematic drawing of the injection device, with connection to microcoil NMR probe; the inset illustrates the elongated droplet in the PTFE tubing, surrounding by FC43. (B) Photograph of a conical glass vial (for visualization purposes; a plastic vial was used for AAF-dGMP) with a model sample (aqueous dye solution) being fully drawn into the 380 μm o.d. silica capillary of the injection device. (C) Structure of AAF-dGMP (numbered according to Evans et al.). Highlighted bonds are color coded to TOCSY spectrum annotations of Figure 3.

Figure 2

NMR spectra. (A) solvent blank, 40K scans, acquired in the microcoil. Solvent peaks are greyed. (B) AAF-dGMP, 80 ng, 100K scans acquired in the microcoil over 40 hours. Inset: 7.4 ppm peak from first 4 hours (start) and from last 4 hours (end). (C) Reference spectrum of 39 μg AAF-dGMP acquired using a conventional 5 mm NMR tube. The inset shows the aromatic region from spectra acquired at higher and lower temperatures. All of the signals have been assigned previously, for example the signals above 7 ppm are: 6.3 ppm (CH at ribose C-1), 4.2 ppm (CHs at ribose C-3 and C-4), 3.9 ppm (bridge CH₂ of fluorene and CH₂ at ribose C-5), and 2.2 ppm (acetyl, and CH₂ at ribose C-2).

Figure 3

A) aromatic region of a TOCSY spectrum of 35 μg AAFdG in 3 μL (20 mM) at 50 °C. The 1D spectrum aligned above it is 16 scans. B) aromatic region of TOCSY spectrum of 1.6 μg AAF-dGMP in 3 μL (1 mM) at 50 °C. The 1D spectrum aligned above it is 8k scans. Fluorene resonances are numbered as in Figure 1C.