Conformation-dependent lesion bypass of bulky arylamine-dG adducts generated from 2-nitrofluorene in epigenetic sequence contexts

Abstract Sequence context influences structural characteristics and repair of DNA adducts, but there is limited information on how epigenetic modulation affects conformational heterogeneity and bypass of DNA lesions. Lesions derived from the environmental pollutant 2-nitrofluorene have been extensively studied as chemical carcinogenesis models; they adopt a sequence-dependent mix of two significant conformers: major groove binding (B) and base-displaced stacked (S). We report a conformation-dependent bypass of the N-(2′-deoxyguanosin-8-yl)-7-fluoro-2-aminofluorene (dG-FAF) lesion in epigenetic sequence contexts (d[5′-CTTCTC#G*NCCTCATTC-3′], where C# is C or 5-methylcytosine (5mC), G* is G or G-FAF, and N is A, T, C or G). FAF-modified sequences with a 3′ flanking pyrimidine were better bypassed when the 5′ base was 5mC, whereas sequences with a 3′ purine exhibited the opposite effect. The conformational basis behind these variations differed; for -CG*C- and -CG*T-, bypass appeared to be inversely correlated with population of the duplex-destabilizing S conformer. On the other hand, the connection between conformation and a decrease in bypass for flanking purines in the 5mC sequences relative to C was more complex. It could be related to the emergence of a disruptive non-S/B conformation. The present work provides novel conformational insight into how 5mC influences the bypass efficiency of bulky DNA damage.

. Calculated monoisotopic mass and actual m/z measured by MALDI-TOF.S38.Table S2.Thermal and thermodynamic parameters of FAF-modified duplexes.S40.Table S3.List of oligonucleotide and primer sequences (5′→3′) used for the REAP and CRAB assays.S42.Table S4.Calculated and observed monoisotopic MW and m/z value of modified oligonucleotides.S43.Table S5.Calculated and observed monoisotopic MW and m/z value of modified oligonucleotides after digestion.

Preparation, Purification, and Characterization of Site-Specifically Modified Oligonucleotides
Briefly, the DNA-reactive N-acetoxy-N-(trifluoroacetyl)-7-fluoro-2-aminofluorene was synthesized via biomimetic activation from 2-fluoro-7-nitrofluorene to Nhydroyxyesters as previously described.(1,2) The activated species was then dissolved to 0.01 mg/μL in absolute ethanol (100 μL) and incubated with ~50 ODs of unmodified 16mer oligonucleotides in 300 μL of 10 mM sodium citrate buffer, pH 6.0 at 37 °C for 24 h.The reaction mixtures were syringe filtered and purified to >97% using reverse-phase high performance liquid chromatography (RP-HPLC) and a Phenomenex Luna C18 column (150 × 10 mm, 5.0 μm) (Phenomenex, Torrance, CA, USA) (Supplementary Figure S1).A gradient of 12.5-25% acetonitrile in triethylammonium acetate (TEAA) buffer (0.3 M, pH 4.5-6.0)for 25 minutes was used to separate the unmodified and modified oligos.The lower pH buffer was used to separate mono-adducts in oligos containing more than one G (i.e., -C # G * G-) (Supplementary Figure S2), whereas the higher pH buffer was sufficient for separating modified from unmodified oligos in sequences containing only one G (i.e., Lesion position was confirmed via enzymatic digestion and analyzed on a Shimadzu (Kyoto, Japan) AXIMA Performance matrix-assisted laser desorption-ionization time of flight mass spectrometer (MALDI-TOF/MS) equipped with a 50 Hz nitrogen laser operating in reflectron mode.Oligos (200 pmol) were incubated with snake venom phosphodiesterase I (0.2 units) for up to 10 minutes.Aliquots (1 μL) were removed at various time points and quenched on the MALDI plate by mixing with 1 μL of matrix (50% v/v 3-hydroxypicolinic acid (3-HPA; 150mg/mL) in dihydrogen ammonium citrate (DHAC; 50 mg/mL)).The MALDI-TOF results are shown in Supplementary Figure S3.It was determined by the digested molecular weights that, for both CGG and mCGG, peak 1 was the G2 adduct and peak 2 was the desired G1 adduct (Supplementary Figure S2).

F and 1 H NMR Parameters
T1 and T2 were measured to be ~0.5 s and 8-18 ms respectively for both signals using -mCGA-; these values are consistent with previously reported T1 and T2 for the B and S conformers.(3)D1 and AQ were therefore set to be 2.0 s and 0.2048 s respectively for 1 +  = 4 − 5 × 1 to ensure full relaxation.Spectra were collected using an average of 25000 scans, a 12500 Hz sweep width, and referenced to external trifluoroacetic acid (CF3CO2H).EXSY spectra were taken in phase-sensitive mode using a NOESY pulse sequence, spectral width 3333.3,1666.7 Hz, number of scans 256, and a mixing time of 100 ms.Imino 1 H NMR spectra were taken over 1024 scans with a sweep width of 6250 Hz.FIDs were processed with an exponential line broadening factor of 5 Hz for 1D 19 F NMR spectra and 2 Hz for 1D 1 H NMR spectra.

In Cell Lesion Bypass and Mutagenesis Assays
Briefly, the constructed M13 viral genomes were mixed with competitor genomes at a 50:1 ratio and transfected into E. coli strains by electroporation.Prior to transfection, HK82 (AlkB -) cells were made electrocompetent.(4)The M13 genomes were extracted from the amplified progeny using QIAprep M13 kit (Qiagen).The lesion region was PCR amplified with assay-specific primers followed by double digestion with XhoI and SphI endonucleases to obtain a DNA fragment (20mer/28mer for the adduct sequence and 23mer for the competitor).Analyses of DNA fragments were performed by LC-ESI-TOF-MS (AB Sciex, ABI4600), and all data represent the mean ± standard deviation (SD) of three independent experiments.Liquid chromatographic separation was achieved by using an Acclaim Polar Advantage II C18 column (2.1 × 250 mm; 3 μm) at a flow rate of 0.15 mL/min.Solvent A was 500 mM 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) in water, and solvent B was 500 mM HFIP in 50% methanol.A solvent gradient was carried out under the following conditions: 25% of B for 1 min, 25 to 50% of B over 2 min, 50 to 75% of B over 20 min, 75 to 100% of B over 1 min, 100% of B for 10 min, 80 to 25% of B over 1 min, and 25% B over 10 min.LC column oven was set at 35 °C during whole running time.ESI was conducted by using a needle voltage of 4.0 kV in a negative ion mode.A heated capillary was set at  represent the monoadducts on G2 and G1 respectively, and assignment of the peaks was determined by 3′ 5′ enzymatic digestion by snake venom phosphodiesterase (SVP) followed by MALDI-TOF analysis (see Table S1 and Figure S3).

Figure S17 .
Figure S17.UV melting temperature (Tm) of unmodified CGC (blue circles), FAF-modified CG*C (red squares), unmodified mCGC (green triangles), and FAF-modified mCG*C (purple triangles) as a function of primer elongation where the lesion position (n) is the 10mer.Data presented as the arithmetic mean (n=5) ± standard deviation (some error bars may be too small to be visible behind the symbols).

Figure S18 .
Figure S18.UV melting temperature (Tm) of unmodified CGT (blue diamonds), FAFmodified CG*T (red circles), unmodified mCGT (green squares), and FAF-modified mCG*T (purple triangles) as a function of primer elongation where the lesion position (n) is the 10mer.Data presented as the arithmetic mean (n=5) ± standard deviation (some error bars may be too small to be visible behind the symbols).

Figure S19 .
Figure S19.UV melting temperature (Tm) of unmodified CGA (blue triangles), FAFmodified CG*A (red diamonds), unmodified mCGA (green asterisks), and FAF-modified mCG*A (purple stars) as a function of primer elongation where the lesion position (n) is the 10mer.Data presented as the arithmetic mean (n=5) ± standard deviation (some error bars may be too small to be visible behind the symbols).

Figure S20 .
Figure S20.UV melting temperature (Tm) of unmodified CGG (blue crosses), FAF-modified CG*G (red x's), unmodified mCGG (green squares), and FAF-modified mCG*G (purple circles) as a function of primer elongation where the lesion position (n) is the 10mer.Data presented as the arithmetic mean (n=5) ± standard deviation (some error bars may be too small to be visible behind the symbols).

Figure S25 .
Figure S25.Diagram of LC-TOF-MS identification of the digestion product from the 58mer lesion-containing oligonucleotide using the CG*C sequence as an illustration (G*=dG-C8-FAF).

Figure S27 .
Figure S27.Diagram of PCR amplification for lesion containing M13 genome using the CG*C sequence as an illustration (G*=dG-C8-FAF).

Figure S28 .
Figure S28.Diagram of polyacrylamide gel of PCR products of lesion containing M13 genome using the CG*C sequence with FAF as an illustration.Section 1: 58mer Control.Section 2-4: 58mer CG*C ligation product (G*=dG-C8-FAF).15% Polyacrylamide gel of FAF-dG containing M13 genome and M13 genome.Two sets of primers were used to generate either 190mer or 230mer DNA product.

Figure S31 .
Figure S31.A typical LC-TOF-MS analysis of REAP and CRAB samples.eA is used for illustration.Left panel shows mutation patterns of eA and right shows the bypass efficiency of eA.(6)

Table S1 .
Calculated monoisotopic mass and actual m/z measured by MALDI-TOF

Table S2 .
Thermal and thermodynamic parameters of FAF-modified duplexes Template

Table S3 .
List of oligonucleotide and primer sequences (5′→3′) used for the REAP and CRAB assays