Efficient DNA–Polymer Coupling in Organic Solvents: A Survey of Amide Coupling, Thiol-Ene and Tetrazine–Norbornene Chemistries Applied to Conjugation of Poly(N-Isopropylacrylamide)

A range of chemistries were explored for the efficient covalent conjugation of DNA to poly(N-isopropylacrylamide) (poly(NIPAM)) in organic solvents. Amide coupling and thiol–ene Michael addition were found to be ineffective for the synthesis of the desired products. However, the inverse electron-demand Diels–Alder (DAinv) reaction between tetrazine (Tz) and norbornene (Nb) was found to give DNA–polymer conjugates in good yields (up to 40%) in organic solvents (N,N-dimethylformamide, N,N-dimethylacetamide and N-methyl-2-pyrrolidone), and without the need for a catalyst. Methods for the synthesis of Tz-and Nb- functionalised DNA were developed, along with a post-polymerisation functionalisation strategy for the production of Tz-functionalised polymers.


Materials
2-(Dodecylthiocarbonothioylthio)-2-methylpropionic acid (DDMAT, 1) was synthesised using a previously published procedure and recrystallised from acetone/water. 1 N-isopropylacrylamide (NIPAM) was recrystallised from hexane and stored at 4°C. 2,2′-Azobis(2-methylpropionitrile) (AIBN) was purchased from Wako Pure Chemical Industries and recrystallised twice from methanol and stored in the dark at 4°C. Styrene was passed through a neutral alumina column prior to use to remove the radical inhibitor. 18 MΩ water was obtained using a MilliQ™ Simplicity system. Dialysis membranes were purchased from Spectra/Por® and soaked in 18 MΩ water before use. The DNA strands s0-NH2 and s0′ were purchased from Integrated DNA Technologies, Inc. and resuspended in 18 MΩ water to a concentration of 200 µM before use. Syringe filters were purchased from Gilson Scientific Ltd. Silica gel for column chromatography and all NMR solvents were purchased from Apollo Scientific Ltd. MOPS buffer consisted of 100 mM 3-(N-morpholino)propanesulfonic acid adjusted to the desired pH with 1 M HCl or NaOH. PBS consisted of 100 mM potassium phosphate -the desired pH was achieved by mixing appropriate amounts of the monobasic and dibasic versions (both at 100 mM). All other chemicals were purchased from Sigma-Aldrich Corporation and used as received. 3-Maleimidopropionic acid N-hydroxysuccinimide ester (5) was purchased from Alfa Aesar and used as received. NAP-5 and -10 sephadex columns were purchased from GE Healthcare. S-Propargyloxycarbonylphenylmethyl dithiobenzoate (4) was synthesised according to a previously published procedure. 5-oxo-5-(6-(6-(pyridin-2-yl)-1,2,4,5tetrazin-3-yl)pyridin-3-ylamino)pentanoic acid (9) and (4-(6-methyl-1,2,4,5-tetrazin-3yl)phenyl)methanol (6) were synthesised according to previously published procedures. 2,3 Silica gel was treated with EtSiCl3 according to a published procedure. 4 The DNA strands s0-NH2 and s0-MAAm were purchased from Integrated DNA Technologies Ltd. and resuspended in 18 MΩ water to a concentration of 200 µM prior to use. NAP-5 sephadex purification columns were purchased from GE Healthcare. Poly(NIPAM) containing a terminal norbornene group (P7) was synthesised according to a previously published procedure. 2 ZipTip® pipette tips were purchased from Merck Millipore. 4-Cyano-4-(dodecylthiocarbonothioylthio)pentanoic acid (7) was purchased from Sigma-Aldrich and used as received. Bio-Beads S-X1 for preparatory size exclusion chromatography were purchased from Bio-Rad Laboratories Inc.

Size exclusion chromatography (SEC)
DMF SEC data were obtained in HPLC grade DMF containing 1 mg mL −1 lithium bromide at 323 K, with a flow rate of 1.0 mL min −1 , on a set of two Varian PLgel 5 μm Mixed-D columns (7.5 mm diameter), with guard column. THF SEC data were obtained in HPLC grade THF containing 2 % triethylamine at 293 K, with a flow rate of 1.0 mL min −1 , on a set of two Varian PLgel 5 µm Mixed-D columns (7.5 mm diameter), with guard column. CHCl3 SEC data were obtained in HPLC grade CHCl3 at 293 K, with a flow rate of 1.0 mL min −1 , on a set of two Varian PLgel 5 µm Mixed-D columns (7.5 mm diameter), with guard column. SEC data were analysed using Cirrus SEC software calibrated using poly(methyl methacrylate) standards (690-271 400 Da) or poly(styrene) standards (162-371 100 Da).

IR spectroscopy
IR measurements were collected on a PerkinElmer Spectrum 100 FT-IR spectrometer. Solid samples were crushed and then applied to the FTIR sensor; liquid samples were applied as a small droplet.

UV-vis spectroscopy
UV-vis measurements were collected on a PerkinElmer Lambda 35 spectrometer using a Hellma TrayCell with a 1 mm path length adapter or, for cloud point measurements, with a quartz cell with a 1 cm path length. DNA solution concentrations were determined using UV-vis absorption measurements at 260 nm and the known extinction coefficient supplied by the manufacturer.

Mass spectrometry
ESI mass spectra were collected on a Bruker Esquire2000 ESI-MS machine using either methanol (for small molecules) or a 1:1 mixture of 2-propanol and 50 mM ammonium acetate (for DNA samples) as solvent. MALDI-ToF mass spectra were collected on a Bruker Ultraflex II MALDI-ToF machine using 3-hydroxypicolinic acid as the matrix.

Electrophoresis
Native polyacrylamide gel electrophoresis (PAGE) was carried out with 1 × Tris-Acetate EDTA (TAE) as running buffer at 4°C and constant voltage of 200 V, loading with glycerol/bromophenol blue loading buffer. All gels were run using a Bio-Rad Mini-Protean Tetra System apparatus, and visualised using SYBR Gold nucleic acid stain, purchased from Invitrogen, under UV transillumination with a UVITEC UVIdoc HD2 gel documentation system. Samples were diluted so that approximately 1 pmol of DNA was added to each lane of the gel (typically 10 µL of a 100 nM solution). Yields were estimated by densitometry using the Image-J image analysis package by taking the area under the peak of interest and dividing it by the area under all DNA-containing peaks.
1 × TAE buffer consisted of 40 mM Tris-acetate and 1 mM EDTA. 1 × TE buffer consisted of 10 mM Tris-HCl and 1 mM EDTA. The native loading buffer consisted of 25 % glycerol and 0.05 % bromophenol blue in 1 × TE buffer, and was diluted five-fold before use.

HPLC
HPLC analyses were performed on a Varian 920-LC™ integrated liquid chromatography system. Chromatography was performed on a Waters XBridge™OST C18 2.5 μm 4.6 × 50 mm column heated to 40°C (for DNA strands) or 24°C (for DNA-polymer conjugate). Flow rate was set at 1 mL min −1 with a linear gradient of the following buffers: Buffer A, 0.1 M triethylammonium acetate, 5 % acetonitrile, pH 7.0; buffer B, 0.1 M triethylammonium acetate, 70 % acetonitrile, pH 7.0. Fractions collected were combined and concentrated using an Eppendorf concentrator plus.

Removal of the RAFT end group using EPHP
The trithiocarbonate end group was removed from poly(NIPAM) as follows. 40 P1a (0.200 g, 0.01 mmol), 1-ethylpiperidine hypophosphite (EPHP) (0.060 g, 0.33 mmol) and AIBN (2 mg, 10 µmol) were dissolved in anhydrous DMF (5 mL) and transferred to an ovendried ampoule. The solution was rigorously degassed by three successive freeze-pumpthaw cycles and then stirred under nitrogen at 100 °C for 2 hours. Water (20 mL) was added and the solution dialysed (MWCO 1 kDa) against 18 MΩ water, with five water changes. The solution was freeze-dried to yield a white powder (0.150 g, 75 %), which was analysed by SEC using DMF as the eluent and PMMA calibration standards (Mn End group removal from poly(NIPAM) samples of various molecular weights was achieved using the same method, keeping the concentration and number of equivalents of EPHP and AIBN constant.

DNA-polymer conjugation using amide coupling reagents
For a full list of the coupling agents and solvents tested, see Table S1, Table S2 and Table  S3. A general procedure follows; unless otherwise stated, stock solutions were made up in the appropriate reaction solvent. The acid-functionalised polymer (1 µL, 10 mM in DMF), coupling agents (1 µL, 10 mM) and s0-NH2 (0.5 µL, 200 µM in water) were mixed and the solution topped up to 9.5 µL with the appropriate reaction solvent. DIPEA (0.5 µL, 20 mM) was added, and the mixture vortexed briefly then left overnight at room temperature. Water (70 µL) and 5 × glycerol loading buffer (20 µL) were added and the mixture analysed by 15 % native PAGE.
Anhydrous DMF (7.5 mL) was added via syringe and the flask cooled to 0°C with an ice bath. DIPEA (354 µL, 2.74 mmol) was then added via syringe, followed by dropwise addition of pentafluorophenyl trifluoroacetate (283 µL, 1.65 mmol). After one hour stirring at 0 °C, the flask was opened to the air and diethyl ether (30 mL) was added, followed by a 1 M solution of HCl (30 mL). The organic layer was collected and washed with water (2 × 30 mL) and brine (30 mL). The solvent was removed in vacuo to give a yellow oily residue, which was then purified by silica gel column chromatography, eluting with a mixture of ethyl acetate and pet. ether 40-60 (gradient from 5-10 % ethyl acetate).

Synthesis of poly(NIPAM) using NHS-and PFP-DDMAT
Polymerisation of NIPAM with 3 was conducted as follows. NHS-DDMAT, 3, (0.041 g, 0.09 mmol), NIPAM (1.000 g, 8.84 mmol) and AIBN (0.002 g, 0.01 mmol) were dissolved in 1,4-dioxane (1.5 mL) and transferred to an oven-dried ampoule. The mixture was subjected to three freeze-pump-thaw cycles and sealed under an atmosphere of nitrogen. It was then placed in an oil bath preheated to 65 °C. After 2 hours the ampoule was removed and the reaction quenched by opening it to the air and cooling with liquid nitrogen. The solution was poured into pet. ether 40-60 (80 mL) cooled in an ice bath and the precipitant collected by filtration. The product was then dissolved in THF (1 mL) and the process repeated 5 more times. Finally, the isolated solid was dissolved in THF (1 mL) and precipitated into diethyl ether (80 mL) cooled in an ice bath.

Removal of the trithiocarbonate group using AIBN and LPO
P2 (100 mg, 0.01 mmol), AIBN (187 mg, 1.14 mmol) and LPO (18 mg, 0.05 mmol) were dissolved in dry toluene (28 mL) and the solution degassed by three freeze-pump-thaw cycles and then sealed under nitrogen. The mixture was heated to 80 °C for five hours, then allowed to cool to room temperature. The solvent was removed in vacuo and the residue resuspended in THF (1 mL), which was then poured into pet. ether 40-60 (15 mL) cooled with an ice bath. The precipitated product was collected by filtration and dried under vacuum to give a white powder (71 mg, 71 %), which was analysed by DMF SEC using PMMA calibration standards (

Polymer end group removal using NaBH4
The trithiocarbonate group present in P1a-c was reduced using a previously reported procedure. 41 An example procedure follows. P1a (0.200 g, 0.03 mmol) was dissolved in water (20 mL) and sodium borohydride (0.378 g, 10.00 mmol) was added. The mixture was stirred at room temperature for 2.5 hours with vigorous stirring, then dialysed against 18 MΩ water for 4 days incorporating 5 water changes. The solution was then freeze-dried to obtain the thiol-terminated product (P4a) as a white solid (0.131 g, 66 %) and analysed by DMF SEC using PMMA calibration standards (

Ellman's assay
Ellman's assay was carried out as follows. Ellman's reagent (0.4 mg mL −1 ) was dissolved in potassium phosphate buffer solution (100 mM, pH 8.0) to give Ellman's solution. The polymer under investigation (1 mg) was dissolved in water (600 µL) which had been purged with nitrogen for 30 minutes. This solution was then diluted with 1.2 mL of potassium phosphate buffer solution (100 mM, pH 8.0). Finally, 200 µL of Ellman's solution was added and the solution mixed. The absorbance at 412 nm was recorded and the concentration of thiol determined by using the extinction coefficient of Ellman's solution (ε412 = 14 150 M −1 cm −1 ). Finally, the percentage incorporation of thiol was calculated by comparing the calculated concentration to the theoretical concentration (which assumes one thiol group per polymer chain).

Conjugation of P4a-c to s0-MAAm without catalyst
s0-MAAm (0.5 µL, 200 µM in water, 0.1 nmol) was added to thiol-terminated poly(NIPAM) (P4a-c) (10 µL, 1 mM in the reaction solvent) and the mixture shaken for 24 hours at 40 °C. The reaction mixture was then analysed by 15 % native PAGE. The reaction was also attempted with fewer equivalents of polymer and at room temperature.

Conjugation of P4a-c to s0-AAm
Conjugation of thiol-terminated poly(NIPAM) (P4a-c) to s0-AAm was attempted using identical conditions to those employed above for s0-MAAm. No product was observed under any of the conditions used.

Synthesis of s0-Mal using the bifunctional adapter, 5
s0-NH2 (1000 µL, 200 µM in water, 200 nmol), 5 (53.6 mg, 200 mmol) and DIPEA (35 µL, 200 mmol) were mixed in DMF (1000 µL) and the reaction shaken overnight at 40 °C. The excess small molecules were then removed using a NAP-10 Sephadex column and the collected solution concentrated in vacuo and purified by HPLC. The product was isolated as a single fraction, with an isolated yield of 63 % as quantified by UV-vis spectroscopy using the known extinction coefficient of the starting material DNA at 260 nm.

Conjugation of P4a-c to s0-Mal in organic solvents
s0-Mal (2 µL, 50 µM in water, 0.1 nmol) was mixed with P4a-c (1 µL, 10 mM in the reaction solvent) and DIPEA (0.5 µL, 20 mM in the reaction solvent) and the reaction solvent (DMF, THF, NMP, DMSO or MeCN -6.5 µL). The solution was left for 24 hours at room temperature and then analysed by 15 % native PAGE. The reaction mixture was also analysed by HPLC to assess the degree of degradation of the maleimide group.

Reaction of s0-Nb with 6
The DNA strand s0-Nb (14.3 µL, 7 µM in water) was added to a centrifuge tube and the solvent removed in vacuo. HPLC buffer (9 µl, 100 mM TEAA, 70 % MeCN) was added and then 6 (1 µL, 10 mM in DMSO). After twenty-four hours the mixture was purified by ZipTip to remove excess small molecules, and analysed by HPLC, which revealed a significant peak shift from the starting material.

Attempted modification of PFP-containing P2 with 6
Post-polymerisation modification of a polymer synthesised using CTA 2 was attempted as follows. Poly(NIPAM) (20 mg, 1 µmol), 6 (2 mg, 10 µmol) and TEA (0.5 mg, 5 µmol) were dissolved in anhydrous THF (0.25 mL) and stirred under nitrogen at 35 °C for two hours, then for a further fifteen hours at room temperature. The solution was purified by preparatory SEC (Bio-Beads S-X1) and the polymer isolated, dried and analysed by 1 H NMR spectroscopy and DMF SEC. Both indicated that no reaction had taken place.

Conjugation of s0-Nb to P7
s0-Nb (0.5 µL, 70 µM in water) was mixed with the reaction solvent (5.0 µL) and P7 (0.5 µL, 14 mM in DMF) and left at room temperature for forty-eight hours. The mixture was diluted with water (35 µL) and 5 × glycerol loading buffer (10 µL) and analysed by 15 % native PAGE. The yield was calculated by densitometry after staining with SYBR Gold and visualisation under UV transillumination.

Synthesis of s0-Tz
300 mM solutions of EDCI, HOBt and 10 were prepared in DMF and then mixed in equal proportions. 100 µL of s0-NH2 (200 µM in water) were added to a 1 mL centrifuge tube and the solvent removed in vacuo. 100 µL of the EDCI/HOBt/10 mixture were added, followed by 100 µL of potassium phosphate buffer (100 mM, pH 8.0). The solution was vortexed to mix and then heated at 40 °C for four hours, after which time small molecules were removed by passing the solution through a NAP-5 sephadex column, eluting with water. The sample was concentrated in vacuo and then purified by HPLC. The product was isolated as a single peak (6 %) and analysed by LC-MS. Expected mass 7 223.9 Da; observed 7 223.2 Da.

Conjugation of s0-Tz to P8
P8 (5 µL, 20/200/2000 µM in the reaction solvent) was mixed with s0-Tz (3.18 µL, 31.4 µM in water) and the reaction solvent (1.82 µL). After 48 hours, the mixture was diluted with 5 × glycerol loading buffer and analysed by 15 % native PAGE. The conjugate was observed as a broad low mobility band in up to 50 % yield (calculated by densitometry).    Figure S5 15 % native PAGE analysis of the DNA-polymer conjugation reactions detailed in Table S1, employing DCC and EDCI as amide coupling reagents. The only band observed was due to the starting material s0-NH2, indicating that the expected product had failed to form.  Table S2 Further coupling agents and solvents tested for the conjugation of P1a-c to s0-NH2. A * indicates that the coupling agent was not highly soluble in the solvent used. 100 equivalents of DIPEA as auxiliary base were used in all cases. Figure S6 15 % native PAGE analysis of the DNA-polymer conjugation reactions detailed in Table S2, employing HBTU and HATU as amide coupling reagents. The band due to the starting material s0-NH2 was still the main band observed; however, in the case of reactions 3a and 4a another, slow-migrating band was also observed, which was attributed to the DNA-polymer conjugate.  Figure S7 Densitometric analysis of lanes 3a and 4a in Figure S6. The DNA-polymer conjugate was clearly visible as a low-mobility band.  Figure S8 15 % Native PAGE analysis of control mixtures of s0 and s0-Mal with P4. Lanes: a) s0-Mal exposed to reaction solvent (MOPS pH 8); b) unfunctionalised s0; c) unfunctionalised s0 + P4; d) P4. No low mobility bands were observed, indicating that these are unlikely to be due to degradation products, binding of the PAGE dye by the polymer, or non-specific association of the DNA with the polymer.            Figure S18 15 % native PAGE analysis of the reaction mixtures detailed in Table S7. The product was expected to appear as a broad, low-mobility band but under no conditions was it observed.   Table S8. A broad, low-mobility band is just visible, possibly indicating the formation of the desired product.

Reaction # Coupling agent Solvent
Figure S20 Densitometric analysis of Figure S19, lane 6e. A small hump may be visible at low migration distance, which could be attributed to a very low yield of the DNA-polymer conjugate. Yields (given as percentage values) were estimated by comparing the areas under each peak. Figure S21 HPLC chromatogram of the reaction mixture during the synthesis of s0-AAm (black) from s0-NH2 (grey). A number of side products were also observed, but were successfully removed by HPLC purification. The product was identified by LC-MS.   Reactions were performed at 40 °C for sixteen hours. Figure S24 15 % native PAGE analysis of the crude reaction mixtures detailed in Table S10 showing that the conjugation of P4a to s0-Mal DNA did not work in various organic solvents. Splitting of the band due to the starting material implied degradation of the maleimide group was occurring. Under identical conditions in MOPS buffer the conjugate was produced in around 50 % yield.    Figure S28 HPLC analysis of the reaction of s0-Nb with the small molecule Tz 6. A clear peak shift was observed for both the endo and exo isomers, indicative of a successful reaction. Reaction conditions: s0-Nb (10 µM), 6 (1 mM), HPLC buffer (100 mM TEAA, 70 % MeCN), room temperature, 24 hours. Figure S29 UV-vis spectra showing the change in absorbance maxima when the Tz-containing alcohol 6 was added to s0-Nb DNA. Addition of the Tz caused a shift in the principal peak from 260 nm to 265 nm, and a new peak was observed at around 310 nm. 15 % native PAGE (right), stained with a nucleic acid-specific dye, was used to confirm the presence of DNA in the sample. Figure S30 1 H NMR spectrum of P6 showing the presence of the key peaks due to the polymer end groups. Trifluorotoluene was included as an external standard to assess the degree of incorporation of the PFP activated ester group. Solvent: CDCl3. Figure S31 19 F NMR spectrum of P6 showing the presence of the PFP activated ester peaks. Trifluorotoluene was included as an external standard to assess the degree of incorporation of the PFP activated ester group. Solvent: CDCl3. Figure S32 UV-vis spectrum of the tetrazine alcohol 6 (bottom, red) and P7 (top, orange). The polymer exhibited the characteristic Tz absorbances at 265 nm and 540 nm (insets), as well as a peak at 309 nm due to the trithiocarbonate group at the opposite end of the polymer chain (*). Figure S33 1 H NMR spectrum of P7. The peaks due to both end groups are clearly visible, and all the peaks due to the Tz group are accounted for. Integration of the Tz signals to the poly(NIPAM) NHCH peak at 4.00 ppm revealed that this group had been incorporated with an efficiency of 53 %. Solvent: CDCl3. Figure S34 THF SEC refractive index chromatogram (top) and UV-vis 2D colour map (bottom) of P7. The colour map clearly shows that the characteristic Tz peak at 540 nm is only associated with the polymer, confirming its successful incorporation at the chain end.