The impact of anionic polymers on gene delivery: how composition and assembly help evading the toxicity-efficiency dilemma

Cationic polymers have been widely studied for non-viral gene delivery due to their ability to bind genetic material and to interact with cellular membranes. However, their charged nature carries the risk of increased cytotoxicity and interaction with serum proteins, limiting their potential in vivo application. Therefore, hydrophilic or anionic shielding polymers are applied to counteract these effects. Herein, a series of micelle-forming and micelle-shielding polymers were synthesized via RAFT polymerization. The copolymer poly[(n-butyl acrylate)-b-(2-(dimethyl amino)ethyl acrylamide)] (P(nBA-b-DMAEAm)) was assembled into cationic micelles and different shielding polymers were applied, i.e., poly(acrylic acid) (PAA), poly(4-acryloyl morpholine) (PNAM) or P(NAM-b-AA) block copolymer. These systems were compared to a triblock terpolymer micelle comprising PAA as the middle block. The assemblies were investigated regarding their morphology, interaction with pDNA, cytotoxicity, transfection efficiency, polyplex uptake and endosomal escape. The naked cationic micelle exhibited superior transfection efficiency, but increased cytotoxicity. The addition of shielding polymers led to reduced toxicity. In particular, the triblock terpolymer micelle convinced with high cell viability and no significant loss in efficiency. The highest shielding effect was achieved by layering micelles with P(NAM-b-AA) supporting the colloidal stability at neutral zeta potential and completely restoring cell viability while maintaining moderate transfection efficiencies. The high potential of this micelle-layer-combination for gene delivery was illustrated for the first time. Supplementary Information The online version contains supplementary material available at 10.1186/s12951-021-00994-2.


Materials.
All chemicals were used as received unless stated otherwise. The chain transfer agent 2-(Butylthiocarbonothioylthio) propanoic acid (PABTC) was prepared following a previously reported procedure.

Instruments.
Nuclear magnetic resonance (NMR) spectroscopy. 1 H NMR (300 MHz) and DEPT 13 C (75 MHz) spectra were recorded on a Bruker AC 300 MHz spectrometer at 300 K. The delay time (d1) was set at 1 s for 1 H NMR and 2 s for DEPT 13 C. Chemical shifts (δ) are reported in ppm.
Size exclusion chromatography (SEC). SEC was conducted on one of two instruments. Dimethylacetamide (DMAc)-SEC was conducted using an Agilent 1200 series instrument equipped with differential refractive index (DRI) and UV/vis (DAD) detector. The liquid chromatography system used 1 × PSS GRAM 30 Å column (300 × 0.8 mm, 10 µm particle size) and 1 × PSS GRAM 1000 Å column (300 × 0.8 mm, 10 µm particle size). The DMAc eluent contained 0.21 wt.% LiCl as additive. Samples were run at 1 mL min −1 at 40 °C. Analyte samples were filtered through a polytetrafluoroethylene (PTFE) membrane with 0.45 μm pore size prior to injection. Poly(methyl methacrylate) (PMMA) narrow standards were used to calibrate the SEC system. The measurements in aqueous solution for P(NAM-b-AA) were carried out on a Jasco system equipped with a AS-2051 Plus autosampler, a DG-2080-53 degasser, a PU-980 pump, a RI-2031 Plus RI detector, a Jasco oven and a PSS SUPREMA guard/1000/30 Å (10 µm particle size). A mixture of 0.08 M Na2HPO4/0.05% NaN3 (pH 9) was used as an eluent at a flow rate of 1 mL min −1 and an oven temperature of 30 °C. PEG standards (400-800,000 g mol −1 ) were used to calibrate the system. Experimental Mn,SEC and Ð (Mw/Mn) values of synthesized polymers were determined using PSS WinGPC UniChrom GPC software. Microplate reader. Fluorescence intensity measurements for PrestoBlue, LDH assays and absorption measurements for hemolysis and aggregation assays were performed on the Infinite M200 PRO microplate reader (Tecan, Germany) with λEx / λEm used as indicated in the respective method sections and gain set to optimal. The combined EBA&HRA assay was conducted on the Cytation 5 multi-mode reader by BioTek, U.S.

Detailed Polymer Synthesis and Characterization.
Synthesis of P(DMAEAm)82 was performed as described before. [2] PABTC (11.9 mg, 5.0 × 10 -5 mol), DMAEAm (682.3 mg, 64.8 × 10 -3 mol), 1,4-dioxane (428.5 mg, 416.0 μL), DMAc (247.8 mg, 263.6 µL), V-65B (213.1 mg of a 1 wt.% solution in 1,4-dioxane, 2.1 mg, 8.2 × 10 -6 mol) and 1,3,5-trioxane (16.3 mg) as an external NMR reference were introduced to a vial equipped with a magnetic stirring bar which was sealed with a cap. The mixture was deoxygenated by bubbling argon through the solution for 10 min. The vial was then transferred to a thermostated oil bath set at 60 °C. After a polymerization time of 4 h, the flask was cooled to room temperature (RT) and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the crude polymer was precipitated three times from THF into -80 °C cold n-hexane. The polymer was dried under vacuum. Then, the polymer was dissolved in distilled water and lyophilized.
Synthesis of P(nBA). PABTC (230.40 mg, 9.7 × 10 -4 mol), nBA (12389.6 mg, 9.7 × 10 -2 mol), 1,4dioxane (4164.2 mg, 4042.9 μL), V-65B (1497.0 mg of a 0.5 wt.% solution in 1,4-dioxane, 7.5 mg, 2.9 × 10 -5 mol) and 1,3,5-trioxane (33.0 mg) as an external NMR reference were introduced to a vial equipped with a magnetic stirring bar which was sealed with a cap. The mixture was deoxygenated by bubbling argon through the solution for 20 min. The vial was then transferred to a thermostated oil bath set at 50 °C. After a polymerization time of 4 h, the flask was cooled to RT and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the solvent was removed and the crude polymer was precipitated three times from THF into cold MeOH/ H20 (75/25). Finally, the polymer was dried under vacuum.
The mixture was deoxygenated by bubbling argon through the solution for 20 min. The vial was then transferred to a thermostated oil bath set at 55 °C. After a polymerization time of 70 min, the flask was cooled to RT and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the crude polymer was precipitated three times from THF into -80 °C cold n-hexane. Finally, the polymer was dried under vacuum.
Synthesis of P(nBA86-b-tBA43-b-DMAEAm88). P(nBA86-b-tBA43) (504.7 mg, 3.0 × 10 -5 mol), DMAEAm (604.0 mg, 4.3× 10 -3 mol), 1,4-dioxane (1220 mg, 1184.5 μL), V-65B (101.9 mg of a 2.0 wt.% solution in 1,4-dioxane, 2.0 mg, 7.9 × 10 -6 mol) and 1,3,5-trioxane (7.3 mg) as an external NMR reference were introduced to a vial equipped with a magnetic stirring bar which was sealed with a cap. The mixture was deoxygenated by bubbling argon through the solution for 15 min. The vial was then transferred to a thermostated oil bath set at 55 °C. After a polymerization time of 70 min, the flask was cooled to RT and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the crude polymer was precipitated three times from THF into -80 °C cold n-hexane. The polymer was dried under vacuum. 1.2 × 10 -5 mol) and 1,3,5-trioxane (75.0 mg) as an external NMR reference were introduced to a vial equipped with a magnetic stirring bar which was sealed with a cap. The mixture was deoxygenated by bubbling argon through the solution for 10 min. The vial was then transferred to a thermostated oil bath set at 50 °C. After a polymerization time of 21 h, the flask was cooled to RT and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the crude polymer was precipitated three times from THF into -80 °C cold nhexane. The polymer was dried under vacuum. 1,4-dioxane, 1.34 mg, 5.18 × 10 -6 mol) and 1,3,5-trioxane (28.0 mg) as an external NMR reference were introduced to a vial equipped with a magnetic stirring bar which was sealed with a cap. The mixture was deoxygenated by bubbling argon through the solution for 10 min. The vial was then transferred to a thermostated oil bath set at 50 °C. After a polymerization time of 6 h, the flask was cooled to RT and exposed to air. 2-3 droplets of the polymerization mixture were used for 1 H NMR and SEC analysis. Afterward, the crude polymer was precipitated three times from chloroform into cold n-hexane. Finally, the polymer was dried under vacuum.

Boc-deprotection of P(nBA86-b-tBA43-b-DMAEAm88).
A sample of Boc-protected polymer was introduced to a 25 mL round-bottom flask equipped with a magnetic stirring bar and TFA/deionized water (97/3, v/v) was added to reach a concentration of 147 mg mL -1 . A small amount of THF was added to aid the solubility. The solution was stirred for 3 h at RT and the TFA was blown off overnight using compressed air. Subsequently, the crude deprotected polymer was precipitated three times from THF into -80 °C cold hexane. Finally, the deprotected polymer was dried under vacuum. 25 mL round-bottom flask equipped with a magnetic stirring bar and TFA/DMF (34/66, v/v) were added. The solution was stirred overnight at RT and quenched with sat. NaHCO3 to reach pH 7.
Then, the solution was dialyzed in water over 2 d with changing solution every 2 h for the first 8 h, and twice the following 2 d. After dialysis, the solution was concentrated under vacuum and lyophilized to yield the product.

Calculations for RAFT Polymerization.
The monomer conversion (p) was calculated from 1 H NMR data by comparing the integrals of vinyl peaks (5.5-5.75 ppm) against the external reference 1,3,5-trioxane (5.10 ppm) before and after polymerization. The theoretical number-average molar mass (Mn,th) was calculated with Equation S1:

Dynamic and Electrophoretic Light Scattering (DLS & ELS).
The hydrodynamic diameters and ζ-potential of the nano assemblies were monitored for three different sample preparations similar to as described before [3] by DLS or ELS using a Zetasizer Nano ZS (Malvern Instruments, Germany) with a He−Ne laser operating at a wavelength of 633 nm. The sample preparations were i) pure micelle solutions as obtained after dialysis, ii) micelle solutions mixed 3+1 with shielding polymer solution or buffer at different pH values, and iii) polyplexes of micelles mixed 3+1 with shielding polymer solution or buffer as control.
Regarding the pure micelle suspensions, no further sample preparation was necessary. Each sample was measured in triplicates at 25 °C with measurement duration of five times 60 s after an equilibration time of 60 s. The counts were detected at an angle of 173°. The mean particle size was approximated as the effective (z-average) diameter and the width of the distribution as the polydispersity index of the particles (PDI) obtained by the cumulants method assuming a spherical shape. The curves and data are presented in Figure S6 and Figure  The third class of samples, the polyplexes, was measured following polyplex preparation at N*/P 30 in 75 μL HBG buffer and mixing with 25 µL shielding polymer solution or HBG buffer as described in the polyplex preparation section. The hydrodynamic diameter and ζ-potential of the samples were measured as described in the paragraph above, but this time the samples were diluted 1:8. Data are expressed as mean ± SD of two by three measurements (n = 2).

Cryo Transmission Electron Microscopy (cryo-TEM).
The samples for cryo-TEM were prepared as described for the second preparation for DLS and Transfer to the microscope was performed with a Gatan cryo stage and the temperature was maintained below -172 °C at all times after vitrification.
The size of the micelles was determined using ImageJ, version 1.52. [4] Briefly, hexagonal arrangements of seven micelles each were identified and the distance between the core of the center micelle and the core of each micelle in a corner was measured. For the estimation of the size of the micellar core, the diameter of a circle drawn around the micellar core was measured. The results are presented as mean ± SD of all measurements of the respective sample.

Titration.
Titration of the polymers was conducted using a Metrohm OMNIS integrated titration system.
Where A1, A2, x0 and p are the initial value, the final value, the center and the power of the curve, respectively.

N*/P Ratio Calculations.
The N*/P ratio was defined as the ratio of the total amount of protonatable amines in polymer solution in relation to the total amount of phosphates in the pDNA solution.
The volume of polymer needed to prepare polyplexes with 15 µg mL -1 pDNA at different N*/P ratios was calculated as described by the following equations: Vtotal · P = Vpoly · Npoly For the heparin dissociation assay, heparin was added to the formed polyplex-EtBr mixtures using the dispenser of the microplate reader to obtain the indicated concentrations (Table S3). After each addition, the plate was shaken, incubated at 37 °C for 10 min and fluorescence intensity was measured.

Determination of Cytotoxicity.
For determination of cytotoxicity of the polymers, the PrestoBlue TM assay was performed with the L-929 cells based on ISO10993-5. In detail, cells were seeded at 0.1 × 10 6 cells mL -1 in growth medium (D10) containing 10 mM HEPES (D10H) in a 96-well plate without using the outer wells. Complete hemolysis (100%) was achieved using 1% Triton X-100 as positive control. Pure PBS was used as negative control (0% hemolysis). The hemolytic activity of the polycations was calculated as follows (Equation S7): Where ASample, ANegative control and APositive control are the absorption values of a given sample, the PBS treatment and the Triton X-100 treatment, respectively. A value less than 2% hemolysis rate was classified as non-hemolytic, 2 to 5% as slightly hemolytic and values > 5% as hemolytic.
To determine the cell aggregation, erythrocytes were isolated as described above. Subsequently, respectively. Experiments were run in technical triplicates and were performed with blood from three different blood donors.

Polyplex Uptake via Flow Cytometry.
To study the uptake of polymers over time in HEK293T cells, the cells were seeded at 0.2 × 10 6 cells mL -1 in D10H in 24-well plates, followed by incubation at 37 °C in a humidified 5% (v/v) CO2 atmosphere for 24 h and medium change to fresh D10H 1 h prior to treatment. The cells were treated with polyplexes with or without layering at N*/P 30 and a final pDNA concentration of 1.5 µg mL -1 for indicated time periods. The polyplexes were prepared as described above after labelling 1 µg pKMyc pDNA with 0.027 nmol YOYO-1 iodide. Subsequently, the polymer-pDNA-solutions were added to the cells, diluting the polyplexes 1:10 in cell culture medium.  Figure S19). MFI values of control cells can be found in Table S8. The experiments were performed at least three times and data are expressed as mean ± SD.

Polyplex Uptake via CLSM.
To study the uptake of polymers via CLSM, HEK293T cells were seeded at 0.2 × 10 6 cells mL -1 in D10H in 8-well slides (ibidi, Germany), followed by incubation at 37 °C in a humidified 5% Regarding the quantification of polyplexes out/inside organelles, the following settings were used to extract three different feature classes, nuclei, polyplexes and organelles (with the subclass: colocalized polyplexes). Regarding the nuclei, the Hoechst channel images were smoothed (size = 3) followed by segmentation using global thresholding (10,000-65,535) with a tolerance of 3% and watershed separation (count = 25). Objects larger than 20,000,000 µm 2 were counted as nuclei.
For the polyplexes, the YOYO-1 channel images were smoothed (size = 3) followed by background subtraction (radius of rolling ball = 30 µm), segmentation using global thresholding Where NPolyplexes, and NNuclei are the total counts of the respective feature classes of one repetition.
In case of the nuclei, the values were divided by three due to the acquisition of z-stack images with three slices.

Image Acquisition and Processing for the Calcein Release Assay via CLSM.
To image the intracellular distribution pattern of calcein in living cells, live cell imaging was performed using a LSM880, Elyra PS Regarding the quantification of calcein release, the images of the respective channels first had to be optimized regarding specific features (Hoechstnuclei, calceinextensive intracellular fluorescence). Therefore, the single channel images were made binary following special processing with the rolling ball background subtraction tool, automatic contrast enhancement and the setting of an automatic threshold. In case of the calcein channel, the "minimum" convolution filter was applied additionally before the contrast enhancement. The binary images were modified to eliminate small holes inside the feature areas. Subsequently, the processed images of both channels were combined using the "AND" combination mode, leaving only the nuclei with coincident calcein staining, representing cells with calcein release. These were then counted via "Analyze Particles" setting the threshold for the size to 30 square pixels/unit. The same step was     Depiction of exemplary intensity, number, and volume weighted plots and exponential decay correlation coefficients of single measurements. Numbers indicate different formulation batches.
Concentrations of the solutions were as indicated in Table S4. Figure S7. DLS measurements of (layered) micelles at different pH values.
Depiction of exemplary intensity, number, and volume weighted plots and exponential decay correlation coefficients of single measurements. Concentrations of the solutions were as indicated in Table S6. Figure S8. DLS measurements of (layered) polyplexes in HBG buffer.
Depiction of exemplary intensity, number, and volume weighted plots and exponential decay correlation coefficients of single measurements. Concentrations of the solutions were as indicated in Table S5.  (Table S4) of the respective polymer. Color code indicates polymer concentration in µg mL -1 . Figure S10. DLS/ELS measurements of different batches of (layered) polyplexes.
DLS/ELS measurements of (layered) polyplexes vs. (layered) micelles. Polyplexes were formed with 15 µg mL -1 pDNA at N*/P 30 (Table S5). Dots of the same color represent different assembly batches of the respective polymer (Table S4). Stars next to dots indicate dilution of polyplexes in HBG or water to concentrations below 100 µg mL -1 .
(A) EBA of polyplexes of pDNA and different batches of HAC-mic (Table S4)   and are classified as slightly hemolytic between 2% and 5%, and as non-or hemolytic if lower or higher than 2% or 5%, respectively. (C) Aggregation of indicated polymers was measured as light absorption by erythrocytes. Erythrocytes were washed and incubated as described in (B). 10 kDa relative to the sample value and represent mean ± SD (n = 3).
Microscopic Images of Treated HEK293T Cells. Figure S13. Influence of (layered) polyplexes on cell morphology.       followed by incubation with (layered) polyplexes (green) of YOYO-1-labeled pDNA and polymers at N*/P 30 in D10H for 1 h. Live cell imaging was performed following further staining of the cells with Hoechst 33342 for the nuclei (blue) and again with CMDR-PM.
Yellow dots indicate colocalization of CMDR-PM and YOYO-1. The images of the free polyplexes were generated by subtraction of the respective binary colocalization image from the binary polyplex image obtained during image analysis.
HEK293T cells were incubated with (layered) polyplexes of pDNA and polymers at N*/P 30 in D10H for 1 h. Just before the addition of polyplexes, calcein was added to a final concentration of 25 µg mL -1 . Cells were washed twice with warm FC-buffer before the