Vector-Free Deep Tissue Targeting of DNA/RNA Therapeutics via Single Capacitive Discharge Conductivity-Clamped Gene Electrotransfer

Abstract

Viral vector and lipid nanoparticle based gene delivery have limitations around spatiotemporal control, transgene packaging size, and vector immune reactivity, compromising translation of nucleic acid (NA) therapeutics. In the emerging field of DNA and particularly RNA-based gene therapies, vector-free delivery platforms are identified as a key unmet need. Here, this work addresses these challenges through gene electrotransfer (GET) of "naked" polyanionic DNA/mRNA using a single needle form-factor which supports "electro-lens" based compression of the local electric field, and local control of tissue conductivity, enabling single capacitive discharge minimal charge gene delivery. Proof-of-concept studies for "single capacitive discharge conductivity-clamped gene electrotransfer" (SCD-CC-GET) deep tissue delivery of naked DNA and mRNA in the mouse hindlimb skeletal muscle achieve stable (>18 month) expression of luciferase reporter synthetic DNA, and mRNA encoding the reporter yield rapid onset (<3 h) high transient expression for several weeks. Delivery of DNAs encoding secreted alkaline phosphatase and Cal/09 influenza virus hemagglutinin antigen generate high systemic circulating recombinant protein levels and antibody titres. The findings support adoption of SCD-CC-GET for vaccines and immunotherapies, and extend the utility of this technology to meet the demand for efficient vector-free, precision, deep tissue delivery of NA therapeutics.

Keywords: DNA and RNA vaccines; electric field focusing; nonviral gene therapy; nucleic acid electrotransfer; precision gene delivery platform.

Figure 1

Pulsed CC‐GET of naked DNA achieves high efficiency gene expression in mouse hindlimb skeletal muscle in vivo. a) Modeled maps of electric potential and derived electric field in the mouse hindlimb muscle tissue surrounding the CC‐GET probe (COMSOL Multiphysics) based on local conductivity matched to baseline measurements, for a 200 V pulse across the electrodes. Dashed boundary in the electric field map indicates CC‐GET threshold for gene expression (≈120 V cm⁻¹). b) CC‐GET probe showing “electro‐lens” established by the nonconductive space separating two Pt‐Ir electrodes (after electro‐lens patent application^{[ 19 ]}). c) Image of a shaved adult mouse hindlimb with the CC‐GET probe inserted and pDNA in 10% sucrose carrier (50 µL) injected (simulated electric field overlay in red). d,e) Whole‐mount low‐power LSM images of mCherry expression in mouse hindlimb flexor digitorium longus and gastrocnemius muscle groups from two mice 7 days following CC‐GET (CMVp‐mCHERRYnls pDNA; 2 µg µL⁻¹; 50 µL in 10% sucrose carrier; 1 min delay prior to delivery of 5 × 4 ms 50 mA/60 V square wave pulses). f) Transverse and g) longitudinal cryosections of the mouse gastrocnemius muscle 4 days after electrotransfer of a CAGp‐eGFP reporter plasmid (2 µg µL⁻¹; 50 µL; in 10% sucrose;10 × 100 µs pulses at 500 µs intervals, 50 mA/60 V) demonstrates widespread and high efficiency gene expression in the target muscle. h) fLuc – encoding dbCMVp‐fLuc dbDNA; 2 µg µL⁻¹; 50 µL in 10% sucrose was delivered into the mouse hindlimb, with electrotransfer with the CC‐GET probe 1 (5 × 4 ms, 50 mA) performed immediately (0 min), 1 min, or 5 min following injection (n = 10). A significant correlation between the achieved local conductance between the electrodes [mS] and the dbCMVp‐fLuc dbDNA incubation time prior to electrotransfer was recorded. Conductance during immediate CC‐GET (0.552 ± 0.025 mS) was significantly lower than conductance measured after 1 min (0.748 ± 0.042 mS; p = 0.012) and conductance increased further following a 5 min incubation of dbCMVp‐fLuc dbDNA (1.275 ± 0.087 mS; p < 0.001) prior to CC‐GET. This indicated that immediate pulse delivery achieves superior conductivity‐clamping and hence highest electric field strength for a fixed current pulse train. i) Luciferin–Luciferase bioluminescence imaging of dbCMVp‐fLuc dbDNA expression demonstrates that photon emission was inversely correlated to the delay between incubation time before CC‐GET pulse delivery (n = 30). Bioluminescence increase over the no‐GET control (2.15 × 10⁷ ± 8.05 × 10⁶) was highly significant p < 0.001 for each 0 min (2.79 × 10⁹ ± 3.55 × 10⁸), 1 min (1.22 × 10⁹ ± 1.80 × 10⁸), and 5 min (8.18 × 10⁹ ± 7.95 × 10⁷) incubation time before CC‐GET (bioluminescence data from 1, 3, and 7 days combined). Significantly elevated bioluminescence emission was detected using immediate post‐injection CC‐GET compared to 1 min (p = 0.014) or 5 min (p < 0.001) pulse delay after injection. Box plots show 25% and 75% percentile, median and mean (dashed lines), with 95% confidence intervals. One‐way ANOVA on ranks with Tukey post‐hoc test for all pairwise multiple comparison was performed on both data sets. Electro‐lens controlled GET described in patent application.^{[ 19 ]}

Figure 2

Single exponential decay CC‐GET enables field focused, titratable gene expression. a) Reference image of HEK293 cells expressing nuclear localized mCherry fluorescence 3 days following CC‐GET using a conventional 10 × 100 µs constant current (50 mA) pulse train delivered via a Digitimer DS5 stimulator (CMVp‐mCHERRYnls pDNA (2 µg µL) in 7.5% sucrose + 0.225% NaCl, pH 7.4), compared with single exponential decay CC‐GET with varying decay constants (τ), generated by D‐A control of an analog stimulator (DS5, Digitimer), driving the CC‐GET probe. This demonstrated titratable scaling of gene expression. The area of expression is a bioreporter for the suprathreshold electric field. b) Integrated pixel intensity in fluorescence arbitrary units (FAU) demonstrate that while the single 1 ms exponential decay pulse (3.504 ± 0.955 FAU) and the positive control using 10 × 100 µs pulses (4.009 ± 1.354 FAU) showed intermediate expression with no significant difference, the single “capacitor‐like” exponential decay with longer time constants significantly improves mCherry expression (τ 4 ms, 8.356 ± 2.058; p = 0.022 FAU; τ 10 ms 8.165 ± 1.674; p = 0.017; compared with τ 400 µs (1.312 ± 0.468 FAU; n = 5 per group; Kruskal‐Wallis one‐way analysis of variance on ranks with Tukey post‐hoc test for multiple comparisons). Box plots reflect 25% and 75% quartiles, with data overlay. Dashed lines show mean values; solid lines show the median and error bars outline the 95^th percentile confidence intervals. c) In vivo comparison of luciferase reporter expression using SCD‐CC‐GET versus conventional CC‐GET pulse train in the hindlimbs of a mouse with repeated bioluminescence readout out to 524 days (luciferase plasmid CAGp‐fLuc;1 µg µL⁻¹ in 10% sucrose delivered via a 2.2 µF capacitor charged to 120 V (SCD‐GET controller) driving the CC‐GET probe in the left hindlimb, compared with delivery of these pDNAs through the same CC‐GET probe, via a conventional 5 × 4 ms × 120 V square wave pulse train (Digitimer DS5 stimulator) in the opposite hindlimb muscle. Following i.p. luciferin injection, bioluminescence peak was measured (IVIS Spectrum CT imaging platform). d) Time course of repeated luciferase activity bioluminescence measurements in the hindlimbs of the mouse shown in c, at 1, 3, 7, 95, 145, 201, 263, 308, and 524 days; measured as photons per second per centimeter squared per steradian (p s⁻¹–cm² sr⁻¹). e) Repeated luciferin – luciferase bioluminescence recording on days 1, 3, 7, 14, 51, 108, 170, 215, 300, and 431 following SCD‐CC‐GET of the CAGp‐fLuc pDNA (10 µg at 0.5 µg µL⁻¹) with the tethered CC‐GET probe delivering a 120 V (red) or 250 V (blue) discharge from a 2.2 µF capacitor via the SCD‐GET controller in opposite legs (mean ± SEM; n = 3 per group; two‐way repeated measures ANOVA; p = 0.253); see Table S1 (Supporting Information) for individual data.

Figure 3

SCD‐CC‐GET drives efficient delivery of naked mRNA in skeletal muscle. a) Luciferin‐luciferase bioluminescence images from a representative mouse that received SCD‐CC‐GET (CC‐GET probe, 120 V, 2.2 µF via the SCD‐GET controller) of fLuc mRNA (left hindlimb; Trilink L‐7202 0.5 µg µL⁻¹; 30 µL in 10% sucrose) and the CAGp‐fLuc pDNA (right hindlimb; 0.5 µg µL⁻¹; 30 µL in 10% sucrose), 10 h, 24 h, 3 days, 7 days, and 2 weeks following GET. b) Peak photon flux over randomly paired hindlimbs from five mice receiving fLuc–pDNA (blue) and fLuc–mRNA (red) SCD‐CC‐GET (2.2 µF, 120 V). mRNA SCD‐CC‐GET resulted in a significantly stronger photon flux over the over the first 2 weeks (p < 0.001) due to quicker onset and substantially higher peak expression (total photon flux p s⁻¹ cm⁻² sr⁻¹) 10 h mRNA: 1.81 × 10⁸ ± 0.59 × 10⁷; pDNA: 1.87 × 10⁶ ± 3.70 × 10⁵; p = 0.171), 24 h (mRNA: 4.14 × 10⁸ ± 9.07 × 10⁷; pDNA: 5.83 × 10⁶ ± 1.71 × 10⁶; p = 0.007), 72 h (mRNA: 8.82 × 108 ± 2.09 × 10⁸; pDNA: 3.93 × 10⁷ ± 1.32 × 10⁷; p < 0.001), and 7 days (mRNA: 3.40 × 10⁸ ± 1.42 × 10⁸; pDNA: 2.80 × 10⁷ ± 8.59 × 10⁶; p = 0.027). Decline of mRNA expression was evident at 14 days, from peak at 3 days, compared to relatively stable pDNA‐based expression (mRNA: 4.31 × 10⁷ ± 1.23 × 10⁷; pDNA: 1.21 × 10⁷ ± 3.99 × 10⁶; p = 0.802) (n = 5; two‐way repeated measures ANOVA with Holm–Sidak post‐hoc test for multiple comparisons; Box plots reflect 25% and 75% quartiles, with data overlay. Dashed lines show mean values; solid lines show the median and error bars outline the 95^th percentile confidence intervals). Data from the representative mouse shown in A are highlighted by white frames. Figure S1 (Supporting Information, top row) hematoxylin and eosin (H&E) histochemistry of target gastrocnemius muscle at the 14 day timepoint indicates an absence of pathology.

Figure 4

SCD‐CC‐GET needle probe/syringe device achieves rapid onset kinetics for mRNA delivery and long‐term stable expression following pDNA delivery. a) SCD‐CC‐GET probe with in‐built capacitor and coaxial insulated needle with two conductive electrodes separated by the insulated electro‐lens element and b) SCD‐GET charger, for the SCD‐CC‐GET probe (after patent application^{[ 21 ]}). c) Close up of the SCD‐CC‐GET probe targeting the gastrocnemius muscle in the mouse hindlimb. d) Repeated intravital imaging of luciferin–luciferase bioluminescence at various timepoints from 3 h to ≈1 year following SCD‐CC‐GET probe‐mediated GET (2.2 µF, 200 V) of fLuc mRNA (right; 30 µL @ 0.5 µg µL⁻¹ in 10% sucrose) or fLuc pDNA (left; CAGp‐fLuc, 30 µL @ 2 µg µL⁻¹ in 10% sucrose) to BALB/c mouse hindlimb muscles. Images from 24 h onwards are from the same mouse. e) Time course of luciferase activity following fLuc mRNA SCD‐CC‐GET (red – 0.5 µg µL⁻¹ in 10% sucrose, 2.2 µF, 200 V) fLuc pDNA SCD‐CC‐GET (blue – 2 µg µL⁻¹; CAGp‐fLuc in 10% sucrose, 2.2 µF, 200 V) (n = 5) compared to fLuc mRNA no‐GET (cyan) and fLuc pDNA no‐GET (cream) controls (n = 3). Bioluminescence (total photon flux) was measured periodically at 1, 3, 7, 14, 28, 56, 80, 180, and 320 days confirming that GET is required for substantive gene expression, and that long‐term stable luciferase gene expression is achieved by pDNA SCD‐CC‐GET while mRNA SCD‐CC‐GET supports transient (weeks) albeit significantly higher expression levels (p < 0.001; t‐test) and rapid onset. mRNA‐mediated luciferase expression peaked at 6.68 × 10⁸ ± 3.19 × 10⁷ 3 days following SCD‐CC‐GET compared to pDNA driven expression reaching a peak of 3.91 × 10⁸ ± 4.17 × 10⁷ at 14 days. Data represent mean ± SEM; Table S2 (Supporting Information) for individual data. f) Onset of luciferase reporter expression following fLuc‐mRNA SCD–GET (0.5 mg mL⁻¹ in 10% sucrose, 2.2 µF, 200 V). Bioluminescence (peak photon flux p s⁻¹ cm⁻² sr⁻¹) was measured at 3 h (1.24 × 10⁸ ± 4.62 × 10⁷; n = 4), 12 h (1.24 × 10⁸ ± 6.82 × 10⁷; n = 6), 24 h (7.19 × 10⁸ ± 4.88 × 10⁷; n = 4), 48 h (9.47 × 10⁸ ± 1.83 × 10⁸; n = 4), and 96 h (4.99 × 10⁸ ± 6.71 × 10⁷; n = 4). Box plots show 25% and 75% percentile, median and mean (dashed lines), with 95% confidence intervals. ANOVA with Holm–Sidak post‐hoc test for multiple comparisons ^* p < 0.05; ^** p < 0.01, ^*** p < 0.001. g) Immunofluorescence detection of recombinant luciferase (fLuc – green) in BALB/c mouse gastrocnemius muscle 24‐h after mRNA SCD‐CC‐GET (2.2 µF, 200 V). Nuclei labeled with DAPI (blue).

Figure 5

Titratable gene dose delivery via SCD‐CC‐GET for naked NA vaccines. a) SeAP serum levels measured by ELISA at 3 or 5 days (separate cohorts n = 5 per treatment group) following GET (2.2 µF, 120 V) of 10 µg dbCMVp‐SeAP dbDNA (20 µL; 0.5 µg µL⁻¹) or 100 µg dbCMVp‐SeAP dbDNA (50 µL; 2 µg µL⁻¹) in CC‐GET carrier solution. Serum SeAP levels were proportionate to the DNA concentration at 3 days (10 µg dbCMVp‐SeAP dbDNA = 44.35 ± 7.46 pg µL⁻¹; 100 µg dbCMVp‐SeAP dbDNA = 136.81 ± 14.35; no DNA; no‐GET control = 2.83 ± 0.37) and increased by 5 days (10 µg dbCMVp‐SeAP dbDNA = 231.37 ± 26.00, 100 µg dbCMVp‐SeAP dbDNA = 327.91 ± 22.82; 100 µg; no‐GET control = 34.00 ± 6.90; ^*** p < 0.001 two‐way ANOVA with Holm–Sidak post‐hoc test for multiple comparison). b) dbDNA luciferase bioluminescence reporter (dbCMVp‐fLuc) was delivered in a cocktail by SCD‐CC‐GET to one hindlimb, alongside dbDNA encoding an hemagglutinin antigen against the influenza Cal/09 strain (dbCMVp‐Cal/09; both DNAs 2 µg µL⁻¹, 50 µL). Four GET levels were evaluated, spanning 60 V versus 250 V applied to 2.2 and 4.7 µF capacitors. Luciferase expression was measured by repeated bioluminescence imaging on day 1, 3, and 7 (n = 5) against a no electrotransfer control (n = 6). Within each individual treatment group, no significant differences in bioluminescence were detected between day 1, 3, and 7. As no significant difference was found between timepoints within experimental groups (two‐way ANOVA with Holm–Sidak post‐hoc test for multiple comparisons), data were combined. The statistically significant distinctions reflect combination of the data from the three timepoints within treatment groups (average peak photon flux p s⁻¹ cm⁻² sr⁻¹: no‐GET control 1.38 × 10⁷ ± 6.09 × 10⁶; 2.2 µF – 60 V 2.24 × 10⁸ ± 1.24 × 10⁸; 2.2 µF – 250 V 1.34 × 10⁹ ± 5.81 × 10⁸; 4.7 µF – 60 V 2.89 × 10⁸ ± 1.60 × 10⁸; 4.7 µF – 250 V 6.96 × 10⁸ ± 3.77 × 10⁸). Kruskal–Wallis one‐way ANOVA on Ranks with Dunn's post‐hoc test for multiple comparisons. c) Immunosorbant assay detection of the anti‐influenza Cal/09 hemagglutinin IgG in the serum of these animals collected 7 days post SCD–GET. No electrotransfer control 0.297 ± 0.122; n = 5; 2.2 µF – 60 V 0.412 ± 0.153; n = 4; 2.2 µF – 250 V 1.067 ± 0.166; n = 4; 4.7 µF – 60 V 0.550 ± 0.05; n = 4; 4.7 µF – 250 V 0.837 ± 0.195 n = 4. One‐way ANOVA with Bonferroni post‐hoc test for pairwise multiple comparison. Since no statistically significant difference was detected between the 4.7 µF charging voltages, combining the data for 4.7 µF SCD‐CC‐GET at 60 and 250 V (0.693 ± 0.108) yielded a statistically significant difference to the no‐GET control (^*Student t‐test with a two‐tailed p‐value of 0.0371). All box plots show 25% and 75% percentile, median and mean (dashed lines), with 95% confidence intervals. All experiments utilized the CC‐GET probe and SCD‐GET controller.

Figure 6

SCD‐CC‐GET electro‐lens focusing enables single pulse naked DNA/mRNA delivery in a needle/syringe‐style format. The SCD‐CC‐GET probe needle hub contains a small capacitor that on discharge enables production of a local electric field at the needle tip, driving efficient GET after local delivery of the NA therapeutic via the needle. The capacitor is charged remotely via a charging station (SCD‐GET charger), with the capacitor storing the energy until insertion of the needle into the target tissue. In a DNA/mRNA vaccine application, delivery could be intramuscular, equivalent to a conventional inoculation, where immediately following injection, the capacitor discharges to establish instantaneous GET. The “naked” therapeutic NA molecules are biologically inert unless bound to the target cells by this process, which is delineated by a critical electric field strength. a) Detail of the SCD‐CC‐GET needle hub showing the integrated capacitor, switch, charging pins and electro‐lens at the needle tip. b,c) Insertion of the SCD‐CC‐GET probe into the SCD‐GET charging station rapidly charges the device. d–f) Intramuscular SCD‐CC‐GET delivery, where the low conductivity carrier enhances the local electric field strength (conductivity‐clamping) for highly efficient gene delivery. SCD‐CC‐GET system described in patent.^{[ 21 ]}