key: cord-0886020-qlge9a8a
authors: Khare, Purva; Dave, Kandarp M.; Kamte, Yashika S.; Manoharan, Muthiah A.; O’Donnell, Lauren A.; Manickam, Devika S
title: Development of lipidoid nanoparticles for siRNA delivery to neural cells
date: 2021-09-22
journal: bioRxiv
DOI: 10.1101/2021.07.28.454207
sha: 766e260691241e374ed75ffde55cb29121021c86
doc_id: 886020
cord_uid: qlge9a8a

Lipidoid nanoparticles (LNPs) are the delivery platform in Onpattro, the first FDA-approved siRNA drug. LNPs are also the carriers in the Pfizer-BioNTech and Moderna COVID-19 mRNA vaccines. While these applications have demonstrated that LNPs effectively deliver nucleic acids to hepatic and muscle cells, it is unclear if LNPs could be used for delivery of siRNA to neural cells, which are notoriously challenging delivery targets. Therefore, the purpose of this study was to determine if LNPs could efficiently deliver siRNA to neurons. Because of their potential utility in either applications in the central nervous system and the peripheral nervous system, we used both cortical neurons and sensory neurons. We prepared siRNA-LNPs using C12-200, a benchmark ionizable cationic lipidoid along with helper lipids. We demonstrated using dynamic light scattering that the inclusion of both siRNA and PEG-lipid provided a stabilizing effect to the LNP particle diameters and polydispersity indices by minimizing aggregation. We found that siRNA-LNPs were safely tolerated by primary dorsal root ganglion neurons. Flow cytometry analysis revealed that Cy5 siRNA delivered via LNPs into rat primary cortical neurons showed uptake levels similar to Lipofectamine RNAiMAX—the gold standard commercial transfection agent. However, LNPs demonstrated a superior safety profile whereas the Lipofectamine-mediated uptake was concomitant with significant toxicity. Fluorescence microscopy demonstrated a time-dependent increase in the uptake of LNP-delivered Cy5 siRNA in a human cortical neuron cell line. Overall, our results suggest that LNPs are a viable platform that can be optimized for delivery of therapeutic siRNAs to neural cells. Graphical Abstract

25 Delivery of small interfering RNA (siRNA) is a promising strategy to treat pathologies as it 26 allows genetic manipulation with a greater degree of target specificity with fewer off-target effects 27

(1, 2). Lipidoid nanoparticles (LNPs) are the carriers in Onpattro, the first LNP-based RNAi drug 28 to gain FDA approval for the treatment of polyneuropathies caused by a rare and life-threatening 29 disorder, hereditary transthyretin-mediated amyloidosis (3) . The clinical utility of LNPs is further 30 reiterated by the recent full and emergency FDA approvals of the Pfizer-BioNTech (Comirnaty) 31

and Moderna COVID-19 mRNA vaccines, respectively (4, 5) . Based on the clinical success of 32 LNPs in delivering siRNAs to hepatic and non-hepatic targets (7) (8) (9) , we explored if LNPs could 33 effectively deliver siRNA to neural cells. Neural cells are targets for drug delivery in multiple CNS 34 disorders such as Alzheimer's disease, Parkinson's disease, ischemic stroke, peripheral nerve 35 injuries, neuropathic and inflammatory pain (9) (10) (11) (12) (13) . The key challenge associated with the delivery 36 of siRNA to neural cells are its poor accumulation and short duration of action inside cells. In this pilot study, we formulated siRNA-LNPs using C12-200 lipidoid and characterized their concentration of 400 nM and cationic ionizable lipidoid/siRNA w/w ratio was maintained at 5:1, 116 unless stated otherwise. 117 

The data is expressed as mean ± standard deviation (SD), wherever applicable. Comparative 264 statistical analyses were performed using either one-way, two-way ANOVA or One sample t and Wilcoxon tests using GraphPad Prism 9 (GraphPad Software, San Diego, CA). Bonferroni's 266 multiple comparisons test was performed for comparative analyses using one-way ANOVA, 267 wherever applicable. Tukey's and Šídák's multiple comparisons tests for statistical comparisons 268 were performed using two-way ANOVA, wherever applicable. Alpha was set at 0.05. biological activity. We also anticipated that neural cells may show a slower rate of particle uptake 294 compared to non-neuronal cells owing to limited endocytosis/other uptake pathways. We therefore 295 measured the changes in particle diameters and zeta potentials of the LNPs over a period of seven 296 days where the samples were refrigerated interim. Surface coating of nanoparticles using PEG-297 DMG has been reported to improve their pharmacokinetic profile by reducing the recognition by 298 the mononuclear phagocyte system (37-39). We studied the effect of PEG-DMG in stabilizing 299

LNPs by comparing LNPs prepared in the presence (+) and absence (-) of PEG-DMG. We 300 compared the particle parameters of blank LNPs and siRNA-loaded LNPs to study the effect of 301 siRNA encapsulation on the resulting particle characteristics. while not in use. Z-average particle diameters (a), dispersity indices (b) and zeta potentials (c) were measured on a Malvern Zetasizer Pro. Data are presented as mean ± SD of n=3 measurements. Statistical comparisons were made using one-way ANOVA or One sample t and Wilcoxon tests. *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.

increased to about 350 nm on day seven. In contrast, the diameters of LNPs (-PEG-DMG) were 331 about 900 nm post-preparation and increased to 4 µm after seven days (Figure 2a) . A similar trend 332 was observed for the PDI values of blank LNPs i.e., an initial PDI of 0.32 increased to 0.4 after 333 seven days for LNPs (+PEG-DMG) and an initial PDI of 0.5 increased to 1.8 after seven days for 334

LNPs (-PEG-DMG) (Figure 2b ). We observed a few statistically significant changes in the particle 335 diameters, PDI values and zeta potential of the prepared LNPs over the period of 7-days 336 Table 1 . Colloidal stability of blank-or siGFPloaded LNPs immediately after preparation (0 h) and upon storage at 37°C for 24 h measured using dynamic light scattering. SiGFP-LNPs were initially prepared in 10 mM citrate buffer and further diluted to a final siRNA concentration of 50 nM using 1x PBS at a final pH of 7.4 for particle size and PDI measurements and using de-ionized water for zeta potential measurements. The samples were stored at 37°C for 24 h. Z-average particle diameters (a), polydispersity indices (b) and zeta potentials (c) were measured on a Malvern Zetasizer Pro. Data are presented as mean ± SD of n=3 measurements. Tukey's and Šídák's multiple comparisons tests for statistical comparisons were performed using two-way ANOVA. *p < 0.05, ***p < 0.001, ****p < 0.0001 and ns: non-significant.

The colloidal stability of siGFP loaded-and blank LNPs was measured using dynamic light 375 scattering at 0 (immediately after preparation) and 24 h post-preparation upon storage at 37°C 376 that increased to about -9.1 mV 24 h-post storage at 37°C (Figure 3c) . Despite the apparent 405 numerical differences, it should be noted that the measured zeta potentials reflect electrostatically-406 neutral samples and a two-way ANOVA analysis revealed that the differences were not significant. 407 obtained for all the LNP samples. 415

The particle diameters of siGFP-LNPs (+PEG-DMG) containing 10% FBS was measured using 416 dynamic light scattering to determine the effect of serum proteins on the colloidal stability of 417

LNPs. We specifically chose siGFP-LNPs (+PEG-DMG) to study the effect of serum since the 418 presence of siRNA and PEG-DMG resulted in the maximum stability of LNPs with respect to 419 particle diameters and dispersity indices (Figure 2a and b) . We first measured the particle 420 diameter of a control sample containing 10% FBS in PBS (Figure 4a ) that showed an average 421 particle diameter of 19.6 nm whereas siGFP-LNPs (+PEG-DMG) in PBS showed an average 422 particle diameter of about 170 nm (Figure 4b) . We observed distinct size distribution by intensity 423 peaks for the blank 10% FBS as well as LNP samples (Figure 4a and b) . We then measured the 424 particle diameter of siGFP-LNPs (+PEG-DMG) that was supplemented with 10% FBS. MDA-MB-231 and BT-549 cells as these cell models are known to overexpress TRPV1, a 470 neuronal target for siRNA delivery (Figure 6a) (42) . We also studied the compatibility of C12-471 200/siTRPV1 LNPs with primary rat dorsal root ganglion (DRG) neurons (Figure 6b) 

After confirming that LNPs had little-to-no effect on cell viability, we next wanted to 494 determine if LNPs were taken up by neurons. Because of their commercial availability as well as 495 their non-mitotic nature similar to that of the primary rat DRG neurons, primary rat cortical 496 neurons were used to quantify the uptake levels of Cy5 siRNA using flow cytometry. profile for about 92% of the recorded events which indicated a single-cell suspension (Figure 7a , 505 right). We analyzed untreated cells and gated out the auto fluorescence (Figure 7b, left) . Thus, a 506 shift of signal to the right of the histogram gate (R1 region) was considered to signify that the cell 507 was Cy5-positive (+). The proportion of Cy5 (+) cells directly correspond to the % uptake of Cy5 508 siRNA in the neurons as represented by the individual histogram plots for the different groups 509 (Figure 7c) . 510 for the respective groups (f). The histograms are representative of quadruplicate samples. 520

Statistical comparisons were made using one-way ANOVA. ****p < 0.0001. 521

We then analyzed the % uptake of Cy5 siRNA encapsulated in LNPs prepared with (+) and 523 without (-) the inclusion of PEG-DMG. Around 92-94% (Figure 7c (yellow) ) and 81-83% ( Figure  524 LNPs (+PEG-DEMG) and Cy5 siRNA/Lipofectamine RNAiMAX complexes (Figure 7e and f) . 531

Furthermore, as seen in Figure 7e , we saw a greater uptake of LNPs with PEG-DMG as compared 532 to those without PEG-DMG. Particle sizes of LNPs play one of the major roles in determining 533 uptake levels and therefore, lower diameter particles may allow efficient cellular internalization. 534

Furthermore, a significant difference in the uptake of Cy5 siRNA was observed when transfected 535 with naked Cy5 siRNA and Cy5 siRNA-LNPs (+/-PEG-DMG) (Figure 7e and f) . This reiterates 536 the need for a safe and efficient transfection agent to maximize the uptake of siRNA in neurons, a 537 hard-to-transfect cell type. 538 539 Around 94-96% of the cells were Cy5 (+) for the Lipofectamine RNAiMAX group indicating 540 efficient uptake (Figure 7b right and f) . Charge-based interactions of cationic lipids with 541 negatively-charged cells allows for higher uptake as compared to neutral or negatively-charged 542 particles. About 44-46% of the cells were Cy5 (+) for the naked Cy5 siRNA transfection group 543 (Figure 7c (red) ). The difference between the uptake mediated by LNPs (+ PEG-DMG) and 544

Lipofectamine RNAiMAX was statistically non-significant (Figure 7f) . Despite the similar levels 545 of uptake among the LNP(+PEG-DMG) and lipofectamine groups, it must also be pointed out that 546 the uptake mediated by Lipofectamine RNAiMAX was accompanied with noticeable cell 547 stress/toxicity upon visual observation under a microscope whereas LNPs showed a superior safety 548 profile. 549

We studied the effect of exposure time on the uptake of LNPs by incubating the cells for two, 552 four or 24 hours. PEGylation is known to regulate the uptake kinetics of LNPs into cells (49). We 553 compared the differences in uptake for LNPs prepared with/without PEG-DMG. Although 554

Lipofectamine RNAiMAX is a gold standard transfection agent for RNA molecules, it is also quite 555 toxic to cells owing to its strong cationic nature. Therefore, it was unsurprising when we saw 556 changes in morphology indicating cell death, just 2 h after Lipofectamine RNAiMAX was added 557 to cortical neurons, while untreated cells continued to appear spindle-shaped and healthy ( Figure  558   8a) . 559 560 For this experiment, cells were incubated either with naked Cy5 siRNA, Cy5 siRNA LNP with 561 PEG-DMG, with Cy5 siRNA LNP without PEG-DMG, with Lipofectamine RNAiMAX Cy5 562 siRNA, or were left untreated. We qualitatively compared the fluorescence intensity among the 563 cells treated with the above samples. Cells treated with naked Cy5 siRNA showed less intense 564 fluorescent signals as compared to cells treated with Cy5 siRNA-LNPs (+/-PEG-DMG) for the 565 two-and four-hour incubation time points (Figure 8a, b) . However, we saw a noticeable increase 566 in siRNA uptake 24 h-post transfection (Figure 8c) . The next pivotal observation was the 567 difference in uptake from the LNPs (+/-PEG-DMG) groups. As mentioned earlier, PEG-DMG 568 plays a key role in determining the physical stability of LNPs as it allows to maintain lower particle 569 diameters by inhibiting particle aggregation (38). We also noted time-dependent differences in 570 (54). Although we did not find a significant difference in the encapsulation efficiencies of LNPs 605 prepared with and without PEG-DMG, the inclusion of PEG-DMG serves other vital roles i.e., 606 maintaining lower particle diameters by inhibiting particle aggregation and providing a stealth 607 effect for longer circulation times in vivo (37, 38). We further studied the physicochemical stability of the prepared LNPs and compared the stabilizing effect provided by the inclusion of PEG-DMG 609 and siRNA in the LNPs. We utilized siGFP as the model siRNA to compare the siRNA-loaded 610 and blank LNP counterparts prepared both with and without PEG-DMG. We observed a non-611 significant increase or change in the particle diameters, polydispersity indices and zeta potential 612 over a one-week storage period at 2-8 °C for all the samples (Figure 2) . Nevertheless, the particle 613 diameters differed significantly for blank vs. siRNA-loaded LNPs and LNPs with vs. LNPs without 614 PEG-DMG (Figure 2a) . A similar trend was also observed for the polydispersity indices of blank 615 and siRNA-loaded LNPs; both with and without PEG-DMG (Figure 2b) . Our data demonstrated 616 that both siRNA loading and PEG-DMG provided a stabilizing effect to the LNPs by maintaining 617 lower particle diameters and uniform dispersity indices over a period of seven-days. The 618 stabilizing effects of siRNA and PEG-DMG on the resulting particle diameters and dispersity 619 indices were noted when the LNPs were stored for 24 h at 37°C (Figure 3) . LNPs showed further 620 lower particle diameters (ca. 80.5 nm) when sizes were measured in the presence of serum (10% 621 FBS) suggestive of the additional serum-mediated stabilization (Figure 4) . 622 623 Although LNPs are deemed to be safe and are well-tolerated by most of the cells, determining 624 their safety and tolerability in neural cells was our primary aim. As shown in Figure 6 , LNPs 625 were deemed to be well-tolerated by primary DRG neurons as evident by >95% cell viabilities. 626

Increasing the dose of siRNA (and concomitantly the dose of the lipids) did not alter the safety 627 profile of LNPs (Figure 6c) RNAiMAX. There was a significant difference in the uptake of neurons transfected with naked 642

Cy5 siRNA as compared to the neurons transfected with the LNPs (both with and without PEG-643 DMG) emphasizing the need for an effective transfection agent for neural cell uptake (Figure 7f) . 644

We also performed fluorescence microscopy to qualitatively study the uptake of Cy5 siRNA 646 into neural cells. A technical caveat of such qualitative assessments is rooted in the fact that the 647 observed/apparent fluorescent intensities are not normalized to cell number. The cells appeared to 648 be healthy, and spindle-shaped in all the treatment groups except for the Lipofectamine 649 RNAiMAX group where cells appeared stressed and rounded as early as at the 2 h timepoint 650 ( Figure 8a) . As described earlier, we speculate that this is due strong cationic nature of 651 Lipofectamine RNAiMAX. We did not observe a greater uptake for cells treated with LNPs at 100 652 nM siRNA concentration (Supplementary Figure 2) . Nearly all the cells in the field showed Cy5 while the cationic Lipofectamine RNAiMAX-mediated uptake was concomitant with significant 672 cellular toxicity. Based on these findings, we conclude that LNPs are a safe carrier for siRNA 673 delivery to neural cells. We are currently screening a pre-existing LNP library prepared using 674 different lipidoid chemistries to identify LNP candidates for safe and efficient neuronal gene 675 knockdown. We anticipate that the results of these studies will set the foundation for using LNPs 676 for neural cell transfection in a variety of CNS diseases. While our approach validates using LNPs We are also thankful to Dr. Wilson Meng and express our special appreciation to Mr. Nevil 697 Abraham (Duquesne University) for his help with the qPCR experiments. We express our 698 appreciation to Mr. Duncan Dobbins (Duquesne University) for his assistance with the graphical 699 abstract. 700

Lipidic systems for in vivo siRNA delivery

Delivery of siRNA therapeutics: Barriers and 706 carriers

The 708 Onpattro story and the clinical translation of nanomedicines containing nucleic acid-based 709 drugs

Approves First COVID-19 Vaccine. FDA News Release

Without these lipid shells, there would be no mRNA vaccines for COVID-19

Delivery of siRNAs to dendritic cells using DEC205-targeted lipid nanoparticles to inhibit 717 immune responses

The efficiency of lipid nanoparticles with an 719 original cationic lipid as a siRNA delivery system for macrophages and dendritic cells

Degradable 722 lipid nanoparticles with predictable in vivo siRNA delivery activity

Exosomes as 728 drug delivery vehicles for Parkinson's disease therapy

Targeted drug delivery to the peripheral nervous system using 731 gene therapy

Alzheimer's disease: 734 Targeting the Cholinergic System

Extracellular vesicle-mediated 736 delivery of circular RNA SCMH1 promotes functional recovery in rodent and nonhuman 737 primate ischemic stroke models

Naturally Derived Membrane Lipids Impact Nanoparticle-Based 739

Lipid Nanoparticle 741 Technology for Clinical Translation of siRNA Therapeutics

How cationic lipids transfer nucleic acids into cells 744 and across cellular membranes: Recent advances

Nanotechnology for In vivo Targeted 751 siRNA Delivery

Lipid nanoparticle delivery systems for siRNA-based 753 therapeutics

Lipid-like 755 materials for low-dose, in vivo gene silencing

Systemic RNAi-mediated gene silencing in nonhuman primate and rodent myeloid cells

Inside out: optimization of lipid 761 nanoparticle formulations for exterior complexation and in vivo delivery of saRNA

Solid lipid nanoparticles and nanostructured lipid 765 carriers: Structure preparation and application

Lipid Nanoparticle Systems for Enabling Gene Therapies

Fluorescence-Based Solution Assay-RiboGreen Reagent Characterization

Jones et al. 773 -1998 -RNA Quantitation by Fluorescence-Based Solution As.pdf 774 26. Hartung JE, Gold MS. GCaMP as an indirect measure of electrical activity in rat 775 trigeminal ganglion neurons. Cell Calcium

Characterization of the SIM-A9 cell line as a model of 778 activated microglia in the context of neuropathic pain

DNA Polyplexes 781 of a Phosphorylcholine-Based Zwitterionic Polymer for Gene Delivery

Extracellular Vesicles 784 Derived from a Human Brain Endothelial Cell Line Increase Cellular ATP Levels

Effect of zeta potential on the properties of nano-drug delivery systems 787 -A review (Part 1)

Antifungal activity of chitosan nanoparticles and 789 correlation with their physical properties

Relationship between the size of nanoparticles and 791 their adjuvant activity: Data from a study with an improved experimental design

Physicochemical Properties Can Be Key Determinants of Mesoporous Silica Nanoparticle 795 Potency in Vitro

Role of physicochemical properties in nanoparticle toxicity

Rational engineering of physicochemical properties of 799 nanomaterials for biomedical applications with nanotoxicological perspectives

Physicochemical 802 characteristics of nanomaterials that affect pulmonary inflammation

Elsevier has created a COVID-19 resource centre with free information in English 806 and Mandarin on the novel coronavirus COVID-19 . The COVID-19 resource centre is 807 hosted on Elsevier Connect , the company ' s public news and information

Development of lipid 810 nanoparticle formulations of siRNA for hepatocyte gene silencing following subcutaneous 811 administration

Sheddable coatings for long-circulating 814 nanoparticles

On the 816 Formation and Morphology of Lipid Nanoparticles Containing Ionizable Cationic Lipids 817 and siRNA

Activation of 822 endogenous TRPV1 fails to induce overstimulation-based cytotoxicity in breast and 823 prostate cancer cells but not in pain-sensing neurons

Unique Characteristics of the Dorsal Root 827 Ganglion as a Target for Neuromodulation

Dorsal root ganglion 829 macrophages contribute to both the initiation and persistence of neuropathic pain

Peripheral tgf-β1 signaling is 833 a critical event in bone cancer-induced hyperalgesia in rodents

A two-stage 836 poly(ethylenimine)-mediated cytotoxicity: Implications for gene transfer/therapy

Flow cytometry and cell sorting of heterogeneous microbial 839 populations: The importance of single-cell analyses

The effects of 847 PEGylation on LNP based mRNA delivery to the eye

Influence of particle size 850 on the in vivo potency of lipid nanoparticle formulations of siRNA

A fast 857 and sensitive method for measuring the integrity of siRNA-carrier complexes in full 858 human serum

Microfluidic-Based 860 Manufacture of siRNA-Lipid Nanoparticles for Therapeutic Applications

Achieving long-term stability of lipid nanoparticles: 863 Examining the effect of pH, temperature, and lyophilization

RNA 866 Interference in Vitro and in Vivo Using a Novel Chitosan/siRNA Nanoparticle System

Cellular Uptake of siRNA-870

Loaded Nanocarriers to Knockdown PD-L1: Strategies to Improve T-cell Functions

Influence 875 of Polyethylene Glycol Lipid Desorption Rates on Pharmacokinetics and 876 Pharmacodynamics of siRNA Lipid Nanoparticles

Lipid Nanoparticles Enabling Gene Therapies: 879 From Concepts to Clinical Utility

Life Technologies. Attune ® acoustic focusing cytometer training guide

Live cell cycle analysis of Drosophila tissues using the attune acoustic focusing cytometer and vybrant dyecycle violet DNA stain

Addressing the problem of cationic lipidmediated toxicity: The magnetoliposome model

Inflammatory Pain: The Cellular Basis of Heat Hyperalgesia

Potentiation of capsaicin receptor activity by metabotropic ATP receptors as a possible mechanism for ATP-evoked pain and hyperalgesia

Pharmacotherapy for neuropathic pain in adults: Systematic review, meta-analysis and updated NeuPSig recommendations

GCaMP as an indirect measure of electrical activity in rat trigeminal ganglion neurons. Cell Calcium

An integrated cell isolation and purification method for rat dorsal root ganglion neurons

General overview of neuronal cell culture

Three-dimensional model of dorsal root

The precise volumes of the aqueous and ethanolic phases have been added to the revised manuscript in Table 1 . We have also appended the protocols to include details of the 'slow' and 'fast' mixing in the methods section under 'Preparation of siRNA-loaded LNPs (siRNA-LNPs)'. We treated cells with non-dialyzed LNPs and the concentration of ethanol in the transfection/treatment mixture was ca. 2% v/v. In our personal correspondence with Dr. Kathryn Whitehead (Carnegie Mellon University) ca. 2018, one of the pioneers in the development of siRNA-loaded LNPs, she indicated that non-dialyzed LNPs could be safely used on cells as long as the cells tolerated the formulations with no significant toxicity. We observed ~100% cell viabilities when cancer cell lines (MCF-7, MDA-MB-231 and BT-549) and primary DRG neurons were exposed to LNPs (Figure 6 ) and conclude that the non-dialyzed LNPs are safely tolerated by

We thank the reviewers for their insightful comments and helpful suggestions. We have addressed the comments and the individual responses to the questions are presented below. All new text (and figure titles) added in the revised manuscript in response to the reviewer's comments are highlighted in blue font.Reviewer #1: The authors give a specific neural application of LNPs which is a hot topic during the COVID-19 pandemic. There are some points to clarify before publication:1. Although a reference is cited in the section "Preparation of siRNA-loaded LNPs", this section can be described in more detail for the ease of understanding of readers out of the field. These parameters could be detailed: i) precise volume of ethanolic phase and aqueous phase used for the fabrication of LNPs should be cleared, ii) the 'slow mixing' and for the 'fast mixing' protocols should be explained in detail, iii) how the ethanol is removed from LNPs before application on the cells.As discussed above, we have used DLS measurements as a quantitative tool to demonstrate the stability of LNPs. We thank the reviewer for this important suggestion and are now establishing collaborative arrangements with the University of Pittsburgh to carry out cryo-TEM studies of LNPs and the results of these studies will be reported in a forthcoming manuscript.

In figure4, some of the graphs and the letters on these graphs are presented with the same color which makes it harder to follow and understand. This figure could be rebuilt with more clear structuring and coloring. Specifically, the graph lines with green color seem pixelized.This figure has been re-worked to improve clarity in the revised manuscript (Figure 7) . We would like to highlight a few things to explain the pixelization of the histogram. The green-colored histogram in Figure 7 shows the percentage of positive cells for Cy5 (Cy5 (+)) when the neurons were treated with Cy5 siRNA/Lipofectamine RNAiMAX complexes. As expected, Lipofectamine RNAiMAX showed the highest (94.9 %) percentage of Cy5 (+) cells. As a result, neurons treated with Lipofectamine RNAiMAX showed the greatest shift towards the right. Interestingly, we also observed two distinct peaks (a "bimodal" histogram) for this group with varying Cy5 intensities as compared to a "unimodal" histogram for neurons treated with LNPs (Figure 7c (yellow and brown)). As RNAiMAX is toxic to cells, it is likely that the first peak corresponds to Cy5 siRNA taken up by distressed cells (resulting in lower Cy5 intensities). The second peak showed a greater percentage Cy5 (+) cells demonstrating the increased uptake of Cy5 siRNA by the viable cells. The histogram corresponding the Lipofectamine groups appears to be pixelized as the histogram width is also broader for this treatment group compared to the LNP-treated neurons. Moreover, the histograms depicted in 

The 'blurriness' of the fluorescent microscopy image presented in Figure 8 (in the revised manuscript) is likely a result of the 'zoomed-out' micrographs presented in the figures. We have revised Figure 8 by including a 'zoomed-in' version of one the LNP groups (cells treated with 50 nM Cy5 siRNA-LNPs+PEG-DMG in panel d. This image depicts a clearer view of the cells wherein the Cy5 siRNA LNPs show diffuse fluorescence in the cytoplasm. The reviewer has rightly pointed out the difference in morphology of cells exposed to Lipofectamine RNAiMAX in panels b and c in Figure 8 . Lipofectamine RNAiMAX is a highly cationic lipid and a benchmark transfection agent for RNA molecules. Although a strong cationic charge serves to enhance the uptake of carrier molecules in cells, it concurrently mediates cell stress (5) . The observed changes in morphology of cells is a result of this Lipofectamine-mediated stress in this particular cell model, human cortical neurons.Reviewer #2: The author herein demonstrate performance of LNPs for siRNA delivery to neural cell. The study is interesting and deemed fit for publication in this journal. However, there are several limitations that must be addressed and discussed.1. The author must perform a reporter assay. Only uptake does not guarantee successful siRNA delivery.The reviewer makes an important comment here and we have indeed performed pilot experiments to study the knockdown of a therapeutic/neuron-relevant knockdown target. We had chosen not to include this data in the original submission due to the reasons discussed in the supplementary information, but this data is now included in the revised manuscript (Supplementary Figure 3) . Primary dorsal root ganglion (DRG) neurons isolated from rat trigeminal ganglia were used to study the knockdown of transient receptor potential cation channel subfamily V member 1 (TRPV1), a neuron-relevant knockdown target. SiRNA against TRPV1 (siTRPV1) was chosen as a proof-of-concept drug because it has been reported that TPRV1 becomes hyperactive in response to chronic inflammatory pain that reduces their threshold for activation and increases sodium, calcium and chloride fluxes (7, 8) . TRPV1 is a non-selective cation channel exhibiting high calcium permeability and is expressed in peripheral, central axon terminals in the spinal cord, C fibers and/or Aδ fibers (8) . Nearly 60% of the peptidergic primary nociceptors in the dorsal root ganglia and trigeminal ganglia express TRPV1 (8) . Primary DRG cultures were isolated using previously reported methods (9) . The inherent technical challenges associated with isolating DRG neurons/cultures resulted in low numbers of isolated cells but we still proceeded with the transfection study to determine if this TRPV1 target can be silenced using siTRPV1 delivered via LNPs (10) (11) (12) .A key limitation of this study is that the low numbers of neurons (2,500 cells/well) used will naturally have a low (or rather a very low) baseline expression of TRPV1 and therefore, this current setting does not allow us to optimally determine the effectiveness of siTRPV1 delivery via LNPs. Proceeding with this caveat, we cautiously discuss here the findings from this experiment. siTRPV1-LNPs (+PEG-DMG) were formulated using C12-200, an ionizable cationic lipid and helper lipids. The resulting mRNA levels post-transfection were determined using quantitative reverse transcription PCR. Our data showed a low 9% knockdown of TRPV1 when primary DRG cultures were treated with siTRPV1-LNPs that was similar to cells treated with the positive control, siTRPV1-Lipofectamine RNAiMAX complexes. Cells treated with naked siTRPV1-LNPs showed around 2.6% knockdown of TRPV1 whereas inverted (inv.) siTRPV1-LNPs showed about 3.2% TRPV1 knockdown. Despite the low levels of knockdown, the observed differences in % knockdown were significant (****p < 0.0001). As stated earlier in this section, this pilot study must be carefully interpreted due to the caveats associated with the low cell numbers and therefore, a lower baseline TRPV1 expression. Nevertheless, this data points out the safety and the potential of LNPs as delivery agents to silence therapeutically-relevant neuronal targets. Current efforts are underway in the laboratory to establish primary neuronal cultures with a higher yield and results of those studies will be reported in a forthcoming manuscript.

The cytocompatibility data of LNPs tested at increasing doses of siGFP LNPs is now included in the revised manuscript (Figure 6c ).

The intensity and autocorrelation function of DLS measurement is now included in the revised manuscript in Figure 4 . We have also compared the autocorrelation function at different time points and is now included in the revised manuscript as Table 2 The precise volumes of the aqueous and ethanolic phases have been added to the revised manuscript in Table 1 . We have also appended the protocols for 'slow' and 'fast' mixing in the methods section under 'Preparation of siRNA-loaded LNPs (siRNA-LNPs)'. The catalog numbers of the siRNAs are now included in the 'Materials' section of the revised manuscript.6. In LNP preparation, why the author chose to dilute instead of dialysis/similar technique to replace the citrate buffer with PBS? What's the final pH of the solution? This is important to report the zetapotential, as it is highly pH dependent. Depending on volume of citrate and PBS used for LNP preparation for different experiment their charge and size may get affected.In our personal correspondence with Dr. Kathryn Whitehead (Carnegie Mellon University) ca. 2018, one of the pioneers in the development of siRNA-loaded LNPs, she indicated that nondialyzed LNPs could be safely used on cells as long as the cells tolerated the formulations with no significant toxicity. We observed ~100% cell viabilities when cancer cell lines (MCF-7, MDA-MB-231 and BT-549) and primary DRG neurons were exposed to LNPs (Figure 6 ) and conclude that the non-dialyzed LNPs are safely tolerated by the cell models used in this study.The final pH of the solution is 7.4 (now included in the legend for Figure 2 in the revised manuscript).7. The method section must include statistical analysis. Two-way ANOVA and not one-way should be performed whenever applicable.We have now added statistical analysis in the 'Methods' section of the revised manuscript. We used either one-way ANOVA or two-way ANOVA along with One-Sample T-and Wilcoxon tests based on the recommendations from GraphPad Prism software.