Synthesis and structure-activity relationship of peptide nucleic acid probes with improved interstrand-crosslinking abilities: application to biotin-mediated RNA-pulldown

Abstract

The development of interstrand-crosslinking (ICL) probes for the covalent targeting of DNA and RNA sequences of interest has been extensively reported in the past decade. However, most of the reactions reported so far induce the formation of a stable adduct that cannot be reverted, thus rendering these chemistries less useful in applications where the reversibility of the reaction is needed for further downstream processing of the targeted and isolated sequences, such as enzymatic amplification steps. In this work, we report on the reversibility of the furan-mediated ICL reaction. ICL formation can be conveniently triggered by either chemical (N-bromo succinimide, NBS) or luminous stimuli (visible light irradiation in presence of a photosensitizer) and quantitative reversion can be achieved by heating the crosslinked sample at 95 °C, while maintaining the structure of the DNA/RNA targets intact. As a proof-of-concept and showing the benefits of the ICL reversibility, we apply furan-mediated ICL to the pulldown of a target RNA strand from cell lysate.

Fig. 1. Schematic representation of the experiment proposed in this work. Furan-containing PNA probes equipped with a biotin tag are designed to be complementary to an RNA target in cell-lysate (i). Upon hybridization and subsequent light-irradiation in presence of a photosensitizer, the probe can covalently crosslink to the target (ii). To this solution, streptavidin (SAv)-containing magnetic beads are added, allowing the biotin probes to bind to SAv on the beads (iii). The beads are separated with a magnet and washed multiple times to remove non-crosslinked probes (iv). Finally, the beads are heated at 95 °C for 1 h, allowing for the release of the probe from the beads and simultaneous reversal of the ICL reaction (v).

Fig. 2. Structures and sequences of probes and monomers used in this study. (a) Monomer structures; (b) design and sequences of the probes for internal furan modification studies; (c) design and sequences of the probes for terminal furan modifications studies. r: l-arginine; O: 2-aminoethoxy(2-ethoxy)acetyl spacer; underlined nucleobases indicate mismatches.

Scheme 1. Synthetic pathways for the synthesis of M1 (a), M2 (b), and M3 (c). (i) 1-Benzyl-1H-pyrrole-2,5-dione, AcOEt, 75 °C, o.n., 75%; (ii) tert-butyl N-(2-Fmoc-aminoethyl)glycinate hydrochloride, DhBtOH, EDC·HCl, DIPEA, DMF, 0 °C to r.t., 2 h; 74%; (iii) TFA, DCM, 0 °C to r.t., 6 h, quantitative; (iv) HBTU, DIPEA, N,O-dimethylhydroxylamine, DMF, 0 °C to r.t., 45′, 71%; (v) LiAlH₄, THF, 0 °C to r.t., 20′, 86%; (vi) glycine tert-butyl ester hydrochloride, DIPEA, NaBH₃CN, AcOH, MeOH, r.t., 5 h, 81%; (vii) 2-(thymin-1-yl) acetic acid, DhBtOH, EDC·HCl, DIPEA, DMF, 0 °C to r.t., 2 h, 65%; (viii) TFA, CHCl₃, 0 °C, 4 h, 97%; (viii) CDI, THF, reflux, on, 91%; (ix) Boc₂O, TEA, DMAP, THF, 0 °C to r.t., on, 92%; (x) NBS, AIBN, CCl₄, reflux, on, 67%; (xi) NaN₃, DMF, 0 °C to r.t., 2 h, 67%; (xii) K₂CO₃, MeOH, r.t., 98%; (xiii) TFA, CHCl₃, 0 °C to r.t., quantitative yield; (xiv) tert-butyl 2-(5-bromo-2-oxo-4-(1H-1,2,4-triazol-1-yl)pyrimidin-1(2H)-yl)acetate, DBU, MeCN, r.t., 95%; (xv) KF, EtOH, reflux, 3 days, 50%; (xvi) TFA, CHCl₃, r.t., 3 h, quantitative yield; (xvii) tert-butyl N-(2-Fmoc-aminoethyl)glycinate hydrochloride, DhBtOH, EDC·HCl, DIPEA, DMF, 0 °C to r.t., 5 h, 71%; (xviii) TFA, CHCl₃, r.t., 4 h, quantitative yield.

Scheme 2. Introduction of the furan modified monomers in the final PNA sequences. (a) The maleimide-protected furan-containing monomer M1 was included in the PNAs on solid support (i). After cleavage (10% m-cresol in TFA, iia), the obtained PNA probes were submitted to retro Diels–Alder (DA) conditions to obtain PNA-1 series (pH 11.5, 90 °C, 2 h, iiia). (b) The unprotected building-block M2 was included through SPPS on the PNA probe (i). DA protection was subsequently performed directly on the solid support (iv), prior sequence cleavage (10% m-cresol, 10% thioanisol in TFA, iib). The PNA-2 series was obtained after post-cleavage retro-DA under acidic conditions (pH 3, 90 °C for 5 h, iiib). (c) Azide-containing monomer M3 was included on solid support through standard SPPS procedure (i). For PNA-3 series, an unprotected furan-containing alkyne 3-(furan-2-yl)-N-(prop-2-yn-1-yl)propanamide was included post-cleavage (10% m-cresol in TFA, iia), through CuAAC chemistry (v). For the PNA-4 series, the azide moiety of M3 was reduced (P(Me)₃, 2 × 10′, vi) prior the on-resin coupling of 2. Subsequent cleavage (10% m-cresol in TFA, iia) and retro-DA (pH 11.5, 90 °C for 2 h, iiia) delivered the desired probes.

Fig. 3. ICL yield (%) of PNA-1 series equipped with 1, towards DNA-1-XY (X = nucleobase facing 1; Y = nucleobase facing B), and representative PAGE analysis examples. (a) ICL% of PNA-1A. (b) ICL% of PNA-1C. (c) ICL% of PNA-1G. (d) ICL% of PNA-1T. The dashed line indicates the 60% yield threshold. Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, upon addition of 4.0 equivalents of NBS. Experiments were performed as single replicates.

Fig. 4. (A) ICL% of PNA-2 (a) and PNA-3 (b) series towards DNA-1-XY. Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, upon addition of 4.0 equivalents of NBS. The experiment leading to a confirmed ICL formation are indicated in the graph (2G-GC̲, 3A-C̲T, 3A-GC̲). (B) Examples of PAGE analysis for PNA-2G + DNA-1G-Y and PNA-3A + DNA-1Y. The reacting nucleobases are underlined. Experiments were performed as single replicates.

Fig. 5. (A) ICL% for PNA-5 (black), PNA-6 (pink) and PNA-7 (green), in presence of DNA-2-Z (Z = a single base, a 5-base overhang or a C-scan experiment). Z is indicated on the X-axis of the graph. The dashed line indicates a 60% ICL threshold. (B) Example of PAGE analysis of C-scan experiment in presence of PNA-5 and DNA-2-C1T4 (i), DNA-2-C2T₄ (ii), DNA-2-C3T₄ (iii), DNA-2-C4T₄ (iv), DNA-2-C1T₄ (v). Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, upon addition of 4.0 equivalents of NBS. * = ICL to cytosine at N − 1 position. Experiments were performed as single replicates.

Fig. 6. ICL% for the C-scan experiment of PNA-5T (black), PNA-6 (pink) and PNA-7T (green), in presence of DNA-2-Z (Z = C-scan experiment). Z is indicated on the X-axis of the graph. Experiments were performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, upon addition of 4.0 equivalents of NBS. Experiments were performed as single replicates.

Fig. 7. Comparison of ICL% under NBS and ¹O₂-mediated furan activation. (a) Comparison of terminal furan modifications included in PNA-5, PNA-6 and PNA-7, upon NBS and ¹O₂-mediated activation; (b) comparison of the internal modification included in PNA-1G and PNA-2G upon NBS and ¹O₂-mediated activation. The dashed line indicates 60% ICL% threshold. Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, upon addition of 4.0 equivalents of NBS or in presence of 2 μM MB concentration for the light-triggered setup (20′ light irradiation). Experiments were performed as single replicates.

Fig. 8. Expansion of the ICL reaction towards RNA targets and crosslink reversibility studies. (a) Representation of the ICL reaction and its reversion; (b) comparison between the ICL reaction of PNA-5 in presence of RNA-Z and DNA-2-Z. Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, in presence of 2 μM MB (20′ light irradiation). (c) Comparison between the ICL% of PNA-5 in PBS pH 7.4 and Cell lysate (55 million cells per mL) towards DNA-2-A and RNA-2-A. Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, in presence of 2 μM MB, eventually supplemented with cell lysate. (d) ICL reversion between PNA-5 and DNA-2-A. Chromatograms indicate the reaction before irradiation (blue trace), 20 minutes light irradiation (red line) and after heating at 95 °C for 10 (green) and 60 minutes (purple). Experiment performed in PBS buffer, pH 7.4 (100 mM NaCl, 10 mM phosphate), at 5 μM strand concentration, in presence of 2 μM MB (20′ irradiation). The insert shows the MALDI-TOF analysis of the purified DNA peak after ICL reversion, upon heating the sample at 95 °C for 1 hour; (e) PAGE analysis of the ICL reaction of PNA-5 in presence of RNA-C, and ICL reversion upon heating at 95 °C. Lane i: RNA:PNA duplex; lane ii: ICL reaction after 20′ irradiation in presence of MB 2 μM; lane iii: 5′ heating at 95 °C; lane iv: 10′ heating at 95 °C; lane v: 30′ heating at 95 °C; lane vi: 60′ heating at 95 °C; lane vii: RNA starting material. All experiments shown in the figures were performed in triplicate.

Fig. 9. Biotin-mediated RNA pull-down. (a) Schematic representation of the pull-down experiment proposed in this work. (b) Design of the sequences used in the pool-down approach. (c) Overview of the pull-down experiment results, performed in PBS buffer (pH 7.4) supplemented with cell lysate (SK-MEL-28 or MDA-MB-231 cell lines), at a final strands concentration of 5 μM. * indicates p < 0.05. For the furan-containing probes, the experiments were performed at 5 μM RhoB concentration. The sequence and probe design are reported in the left panel of the figure. k = l-lysine; r = l-arginine. All experiments shown in the figures were performed in triplicate.