key: cord-0745210-aab9t5uw
authors: Li, Min; Li, Sanpeng; Huang, Yixuan; Chen, Haixia; Zhang, Songya; Zhang, Zhicheng; Wu, Weigang; Zeng, Xiaobin; Zhou, Boping; Li, Bin
title: Secreted Expression of mRNA‐Encoded Truncated ACE2 Variants for SARS‐CoV‐2 via Lipid‐Like Nanoassemblies
date: 2021-07-18
journal: Adv Mater
DOI: 10.1002/adma.202101707
sha: e96aa22bec25db4a1edbea8cd897f03d19c78368
doc_id: 745210
cord_uid: aab9t5uw

The transfer of foreign synthetic messenger RNA (mRNA) into cells is essential for mRNA‐based protein‐replacement therapies. Prophylactic mRNA COVID‐19 vaccines commonly utilize nanotechnology to deliver mRNA encoding SARS‐CoV‐2 vaccine antigens, thereby triggering the body's immune response and preventing infections. In this study, a new combinatorial library of symmetric lipid‐like compounds is constructed, and among which a lead compound is selected to prepare lipid‐like nanoassemblies (LLNs) for intracellular delivery of mRNA. After multiround optimization, the mRNA formulated into core–shell‐structured LLNs exhibits more than three orders of magnitude higher resistance to serum than the unprotected mRNA, and leads to sustained and high‐level protein expression in mammalian cells. A single intravenous injection of LLNs into mice achieves over 95% mRNA translation in the spleen, without causing significant hematological and histological changes. Delivery of in‐vitro‐transcribed mRNA that encodes high‐affinity truncated ACE2 variants (tACE2v mRNA) through LLNs induces elevated expression and secretion of tACE2v decoys, which is able to effectively block the binding of the receptor‐binding domain of the SARS‐CoV‐2 to the human ACE2 receptor. The robust neutralization activity in vitro suggests that intracellular delivery of mRNA encoding ACE2 receptor mimics via LLNs may represent a potential intervention strategy for COVID‐19.

Messenger RNA (mRNA)-base therapeutics represents a promising therapeutic strategy for multiple clinical applications such as infectious disease vaccines, cancer immunotherapeutics, protein-replacement therapies, genome engineering, and regenerative medicine applications. [1] [2] [3] [4] [5] In-vitro-transcribed mRNA enables functional protein expression in cytoplasm without the risk of insertional mutagenesis. [1] [2] [3] [4] [5] Recently, two mRNA vaccines have been authorized under an emergency-use authorization for preventing COVID-19 infection. [6] Both vaccines and a couple of mRNA vaccine candidates, almost without exception, have employed the advanced lipid nanoparticle (LNP) technology as mRNA delivery system to generate SARS-CoV-2 vaccine antigens, thereby eliciting robust neutralizing antibodies against the virus. [7] [8] [9] [10] [11] [12] [13] [14] In an attempt to neutralize the SARS-CoV-2 virus from another perspective, we herein focus on human angiotensin converting enzyme 2 (ACE2), a cellular entry receptor for SARS-CoV-2. ACE2 (amino acids [aa] 1-805) is a type I transmembrane receptor that anchored to the plasma membrane through a single transmembrane helix. [15] As a cell surface receptor, ACE2 plays a vital role in SARS-CoV-2 entry into cells. [9, [16] [17] [18] Thus, blockage of SARS-CoV-2 binding to ACE2 provides an alternative strategy for controlling the COVID-19 pandemic. [9, 17] Apart from the membrane-bound form, ACE2 is also released via ectodomain shedding into the bloodstream in small amounts. [15, 19, 20] Soluble human ACE2 recombinant protein (aa 1-740) has already been tested for acute respiratory distress syndrome in phase 2 clinical trials, and has recently been demonstrated to be a potent neutralizing agent for blocking SARS-CoV-2 infections. [17, 21] More recently, engineered ACE2 variants with higher affinity to SARS-CoV-2 have been identified through protein engineering. [22] [23] [24] In addition, a series of de novo designed ACE2 proteins and peptides are reported to be potent SARS-CoV-2 inhibitor. [25] [26] [27] [28] In this study, we developed new binary lipid-like nanoassemblies (LLNs) for mRNA delivery in vitro and in vivo. We hypothesized that intracellular delivery of exogenous mRNA encoding truncated ACE2 variants (tACE2v mRNA) through LLNs (referred to hereafter as

Lipid nanoparticles (LNPs) have proven to be a versatile nanocarrier platform for siRNA and mRNA delivery. [1] [2] [3] [4] [5] Structurally, LNPs typically comprise four components including a primary lipid, two helper lipids (such as phospholipid and cholesterol), and a PEGylated lipid. [29] To maximize performance with the fewest components, we prepared dual-component LLNs and adopted a stepwise optimization strategy for determining the optimal formulation parameters (Figure 2a) . In the first step, a combinatorial library of symmetric lipid-like compounds with different side chains were synthesized via the condensation reaction between a para-or tri-substituted benzaldehyde derivative (termed pB and tB, respectively) with primary amine groups of organic nitrogen compounds, followed by reduction of Schiff base intermediates (Figure 2b ; Figures S1, Supporting Information).

To evaluate delivery efficiency, a capped and polyadenylated mRNA encoding the beta galactosidase (β-gal) reporter was complexed with each compound at an equal nitrogen:phosphate ratio (N:P ratio, also designated charge ratio). Preliminary screening showed that lipid-like compounds alone showed an almost background level of delivery capacity for the β-gal mRNA payload, suggesting that lipid-like compounds were ineffective without additional auxiliary components. Upon incorporation of an equimolar helper lipid 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine phospholipid (DOPE), however, two lipid-like compounds possessing two and three monounsaturated C 18 alkyl chains (pB-UC18 and tB-UC18), respectively, exhibited superior efficiency for mRNA delivery (Figure 2c ).

Adv. Mater. 2021, 33, 2101707 Figure 1 . Schematic illustration of lipid-like nanoassemblies (LLNs)-mediated intracellular mRNA delivery and potential application as decoy receptors for SARS-CoV-2. The human angiotensin I converting enzyme 2 (ACE2) gene encodes a transmembrane receptor ACE2 that implicates cellular entry of SARS-CoV-2. In-vitro-transcribed mRNA that encodes high-affinity truncated ACE2 variants (tACE2v mRNA) is encapsulated into LLNs to prevent mRNA degradation and facilitate tACE2v mRNA across cell membranes. Upon releasing into the cytosol from endosome, the tACE2v mRNA cargo initiates protein expression. Subsequently, tACE2v is secreted into the extracellular space, and acts as decoys to compete with the intact membranebound ACE2, thus blocking SARS-CoV-2 entry. www.advmat.de www.advancedsciencenews.com

The characterized top hits pB-UC18 and tB-UC18 both bearing unsaturated C 18 alkyl units (Table S1 , Supporting Information) were subsequently selected for optimizing the molar composition of formulations. In this step, the lipid-like compound-to-mRNA N:P ratio was maintained at 1:1, while the lipid-like compound-to-DOPE molar ratio in the formulation was varied. As shown in Figure 2c , both compounds formulated with DOPE at a 1:1 molar ratio showed the highest delivery efficiency among all LLNs tested. We next optimized the N:P ratio after the optimum composition (lipid-like compound:DOPE = 1:1, mol:mol) was determined. Maximum level of β-gal activity was detected when β-gal mRNA was delivered by the tB-UC18 at a N:P ratio of 1.5:1, which was about 1.6-fold higher than that delivered by a commercial transfection reagent Lipofectamine 2000 (Figure 2c , P < 0.05). Taken together, LLNs consisting of equimolar tB-UC18 and DOPE at tB-UC18-to-mRMA N:P ratio of 1.5:1 (termed tB-UC18 LLNs) outperformed all other formulations tested, and were used for subsequent in vitro and in vivo experiments. It is noteworthy that in this study only two components were used to encapsulate mRNA. Such binary nanocarrier enables minimized batch-tobatch variability during the formulation process. Moreover, the weight ratio of the primary ingredient tB-UC18 to mRNA (1.43:1) in the formulation calculated from the N:P ratio (1.5:1) was ≈14% or even lower than that in conventional lipid-based (10:1 or 20:1) when the same dose of mRNA was used. [30] [31] [32] [33] [34] [35] [36] This performance might be beneficial to broaden therapeutic window of tB-UC18 LLNs in vivo through a drastic reduction in nanocarrier usage.

Ionizable lipids interact with negatively charged mRNA via electrostatic attractions and self-assemble into nanocomplexes. [1] [2] [3] [4] [5] The gel retardation assay showed that the amount of mRNA migrated into an agarose gel decreased with the increase of N:P ratios (Figure 2d ). No obvious naked mRNA was detected when the N:P ratio was 1.5:1 or above (Figure 2d ). These results indicated that electrostatic interaction occurred between tB-UC18 and mRNA, and the majority of mRNA was encapsulated into LLNs at a high N:P ratio. To monitor the formation of nanocomplexes, the particle size and zeta potential of formulations with various N:P ratios were measured by dynamic light scattering. As shown in Figure 2e , the lipid-like compound tB-UC18 was able to assemble into nanoparticles of less than 400 nm at the N:P ratios ranging from 0.5:1 to 2:1, whereas too low and high N:P ratios were unfavorable to form nanoscale particles. The surface charge gradually increased from −35 mV at N:P ratio of 0.25:1 to 44 mV at N:P ratio of 16:1 ( Figure 2e ). Overall, these results demonstrated that the N:P ratio had a direct influence on the physicochemical properties of LLNs.

The optimized tB-UC18 LLNs showed an average diameter of about 300 nm with a polydispersity index of 0.17, as determined by dynamic light scattering (Figure 3a) . The hydrodynamic size was slightly larger than that obtained from transmission electron microscopy, where tB-UC18 LLNs revealed homogeneous core-shell structure around 200 nm in size (Figure 3a) . To evaluate the pH-sensitivity of tB-UC18 LLNs, the hydrodynamic size and surface charge were tested under different pH conditions. The particle size at pH of 9.8 remained constant, relative to the size at pH of 7.4 ( Figure 3b ). By contrast, LLNs at pH of 5.8 displayed an almost 200% pH-responsive swelling in size compared with that at pH of 7.4, along with a pH-induced charge inversion from −29 to 27 mV ( Figure 3b ). Such changes were considered as a favorable property for the endosomal escape of mRNA, [2, 4] and might attribute to the protonation of amine groups of tB-UC18 under acidic condition.

To assess the biocompatibility of nanocarriers, Cell Counting Kit-8 assays were performed in two mammalian cells. For 293T cells, nanocarriers displayed no significant toxicity at the optimal concentration for mRNA delivery (3 × 10 −6 m for tB-UC18 and 3 × 10 −6 m for DOPE), and displayed more than 75% cell viability at combined concentrations up to 32 × 10 −6 m for both tB-UC18 and DOPE (Figure 3c ). In the case of HeLa cells, we did not observe significant reduction in cell viability across the entire dose range (Figure 3c ). Hemolysis analysis revealed that nanocarriers did not cause visible hemolysis effect even at high dose (15 × 10 −6 m for both tB-UC18 and DOPE, Figure 3d) .

The systemic circulation and cellular uptake of mRNA were hindered by nuclease degradation, [1, 4] Thus, we next tested the metabolic stability of formulation in various conditions. The naked mRNA directly exposed to 10% fetal bovine serum (FBS) or RNase was subjected to rapid degradation within a few minutes, while mRNA encapsulated in nanoassemblies conferred resistance to degradation over a period of 12 h (Figure 3e) . Furthermore, over 5000-fold resistance to digestion by RNase was observed when mRNA was formulated into tB-UC18 LLNs ( Figure 3f ). Meanwhile, we investigated the serum stability of nanocarriers by monitoring changes of serum turbidity. We did not found any sign of serum-induced aggregation, demonstrating that nanocarriers remained stable over the period at room temperature in the presence of serum ( Figure 3g ). Furthermore, we stored formulations in 10% serum at 37 °C and monitored the long-term stability based on dynamic light scattering measurements and turbidity analysis. Both results indicated that formulations were thermostable without the addition of a PEGylated lipid ( Figure S4 , Supporting Information).

To establish mRNA expression profiles produced by tB-UC18 LLNs, we carried out dose response and time course studies, and found that both increasing of mRNA dose and prolonged exposure time increased levels of protein expression in both 293T (Figure 4a ) and HeLa cells (Figure 4b ). In addition to i) The top-performed compounds were selected from the resulting 20 lipid-like benzaldehyde derivatives using a given formulation. ii) The composition (lipid-like compound:DOPE, mol:mol) of LLNs were optimized, while keeping the N:P charge ratio fixed. iii) The optimal N:P ratio was determined with fixed composition. b) Chemical structures of two subclasses of lipid-like benzaldehyde derivatives used for assembling nanoparticles. Derivatives was obtained by varying side chains linked to the benzaldehyde scaffold, and named "tB-x" or "pB-x", where p represents para-substituted benzaldehyde derivative; t represents tri-substituted benzaldehyde derivative; x represents the R group in the side chains. c) β-gal activity of LLNs in 293T cells after delivery of 200 ng of β-gal mRNA using indicated formulations. β-gal activity (n = 3) was measured at 24 h post-transfection by a colorimetric assay, and normalized to the Lipo2ktreated group. Measurements were obtained from at least three biological replicates. d) Effect of N:P ratio on mRNA electrophoretic migration. For simplification, tB-UC18:DOPE:mRNA was used to describe the recipe of formulation, where tB-UC18:DOPE indicates molar ratio while tB-UC18:mRNA indicates N:P ratio. e) Effect of N:P ratio on particle size and zeta potential of LLNs. Particles in the gray dotted box were larger than 1000 nm. www.advmat.de www.advancedsciencenews.com measurement of β-gal enzyme activity, in situ staining was also utilized to validate protein expression. Similarly, a dosedependent response to β-gal mRNA delivered by tB-UC18 LLNs was visualized in both cell lines tested (Figure 4c ). To assess the broader applicability of our delivery platform, β-gal mRNA encapsulated in LLNs was replaced with eGFP mRNA, and fluorescence imaging of both cells was performed at different incubation time. As shown in Figure 4d , mRNA-encoded eGFP yielded a low level of fluorescence in 293T cells with shortterm incubation, while strong green fluorescence signals were detected at 24 h and lasted for longer than 3 days under the same measurement setting. Similar profiles were also observed in HeLa cells (Figure 4d) .

To visualize the cellular internalization of mRNA-loaded LLNs, mRNA was first labeled by a TAMRA-tagged oligo(dT) 17 complementary to the poly(A) tail of mRNA via hybridization ( Figure S5 , Supporting Information). Cells treated with nanoassemblies containing the TAMRA-labeled mRNA were then imaged 4 h postdelivery. A red dispersed fluorescent signal was observed inside the cells, indicating that LLNs were effectively internalized within hours (Figure 4e ). To further elucidate the endocytic mechanisms of internalization, three inhibitors including chloropromazine (CPZ), filipin, and 5-(N-ethyl-N-isopropyl)-amiloride (EIPA) were individually supplemented into cells to inhibit clathrin-mediated endocytosis, caveolae-mediated endocytosis, and macropinocytosis. [37] As shown in Figure 4f , both CPZ and EIPA caused dramatic inhibition of clathrin-dependent endocytosis and macropinocytosis in two cell lines tested, while inhibition of LLNs internalization events through caveolae-mediated endocytic pathway relied on cell types. These studies clearly demonstrated that LLNs were predominantly internalized by clathrin-mediated pathways and macropinocytosis.

Next, we explored the delivery efficacy of tB-UC18 LLNs in vivo. C57BL/6J mice (n = 3) was administrated with a single dose of tB-UC18 LLNs containing 0.5 mg kg −1 of firefly luciferase mRNA via different administration routes including intravenous (IV), intramuscular (IM), and subcutaneous (SC) injection. Four hours after administration, we noticed that the luciferase expression in mice upon delivery with mRNA-loaded LLNs varied significantly depending on routes of administration (Figure 5a-d) .

In the case of IM and SC injection, luminescence signal was observed at the injection site, whereas for IV injection, over 95% of luminescence signal originated from the spleen (Figure 5a d). In comparison with other spleen-targeted mRNA delivery system, [38] [39] [40] [41] tB-UC18 LLNs delivered systemically displayed comparable or higher tissue specificity for protein expression in vivo. Flow cytometric analysis of splenocytes showed that the percentage of macrophages, dendritic cells, B cells, and T cells that internalized particles was 31.7%, 28.5%, 5.4%, and 5.2%, respectively ( Figure S6 , Supporting Information). Together, these data suggested that exogenous mRNA encapsulated in tB-UC18 LLNs was successfully translated into functional proteins regardless of routes of administration.

Subsquently, we assessed safety profile of tB-UC18 LLNs in animal models. There were no obvious adverse effects with tB-UC18 LLNs at a 0.5 mg kg −1 or higher mRNA doses, as reflected by several serum chemistry indicators of liver (alanine transferase and aspartate aminotransferase) and kidney (creatinine and blood urea nitrogen) function (Figure 5e ). In addition, no significant morphological changes in tissue histology were observed even at high administration doses (1.0 mg kg −1 ) as compared to the untreated mice ( Figure S7 , Supporting Information). Collectively, hematological and histopathological analysis suggested that mRNA-loaded tB-UC18 LLNs were well tolerated in vivo.

Human ACE2 gene encodes a 805 amino-acid-long ACE2 receptor consisting of an ectodomain (aa 1-740), a single transmembrane helix (STH, aa 741-763), and a cytoplasmic domain (CPD, aa 764-805) (Figure 6a,b) . [15, 18] The ectodomain contains an N-terminal signal peptide (SP, aa 1-17), a peptidase domain (PD, aa 18-615), a neck domain (ND, also named dimerization domain, aa 616-726), and a cleavage region (CR, aa 727-740) connected to the single transmembrane helix (Figure 6a,b) . [18] It has been established that membrane-anchored ACE2 receptor mediates SARS-CoV-2 attachment and subsequent entry. [9, [16] [17] [18] To express high-affinity ACE2 decoy receptors for SARS-CoV-2 via our nanoscale mRNA delivery platform, we set out to engineer the ACE2 mRNA expression construct. Given that SARS-CoV-2 trimeric spike glycoprotein binds tightly to a dimeric version of ACE2 (Figure 6b ), [18] we first retrieved protein sequence of wild-type human ACE2 from UniProt (Accession number: Q9BYF1; Table S2 , Supporting Information), removed the C-terminal membrane anchor and cytosolic domain of the fulllength ACE2, and preserved its neck (dimerization) domain to facilitate tACE2v dimerization (Figure 6a ; Table S2 , Supporting Information). Each domain of the full-length human ACE2 protein and the thorough removal of transmembrane helices were further validated by hidden Markov models (Figure 6c ). [42] In addition, we introduced four point mutations (T27Y, L79T, N330Y, A386L) to the mRNA construct at the recognition interface between the receptor-binding domain (RBD) and ACE2 to increase its binding affinity (Figure 6a ,b; Table S2 , Supporting Information). [22] After codon usage optimization (Table S3 , Supporting Information) and in vitro transcription, we next packaged synthetic tACE2v mRNA into core-shell structured tB-UC18 LLNs (tACE-2vLLNs) and focused on its secreted expression in the mammalian cells. We treated 293T cells with tACE2vLLNs for 24 h and utilized the specific fluorogenic substrate to detect ACE2 released into the cell culture supernatant. [19] Compared to the Adv. Mater. 2021, 33, 2101707 Hemolysis activity was normalized to that of 1% Triton X-100. PBS buffer was used as a negative control. e) Serum and RNase stability of tB-UC18 LLNs. Samples were incubated with FBS (v/v, 10%) or RNase A (10 ng mL −1 ) at room temperature for different time points. f) Protective effect of nanoassemblies on nuclease digestion. β-gal mRNA was digested with different concentration of RNase A at room temperature for 10 min in the absence or in the presence of vehicles. The dotted line indicated the border between two separate gels. g) Turbidity measurement of tB-UC18 LLNs in 10% FBS at room temperature. The absorbance at 660 nm was recorded to monitor particle aggregation. background, free ACE2 was detected in the irrelevant mRNA-LLNs-treated and PBS-treated control samples (Figure 6d ), suggesting that a small amount of endogenous ACE2 shed their ectodomain into the extracellular space under normal physiological conditions. By contrast, over twofold higher ACE2 activity was detected in the tACE2vLLNs-treated group than both controls (Figure 6d ). To further confirm ACE2 activity, we preincubated samples with MLN-4760 (a potent and selective human ACE2 inhibitor) and found that the enzymatic activity of secreted ACE2 was inhibited by MLN-4760 in a dose-responsive manner ( Figure S8, Supporting Information) . These findings demonstrated that exogenous tACE2v mRNA delivered by tB-UC18 LLNs could be effectively translated into secreted tACE2v proteins. We subsequently tested the neutralization activity of mRNA-encoded tACE2v using a SARS-CoV-2 Surrogate Virus Neutralization Test. Neutralization titers revealed that the concentrated supernatant containing the secreted tACE2v was able to effectively block the binding of SARS-CoV-2 RBD to human ACE2 receptor, with an inhibition rate comparable to that of the positive control, SARS-CoV-2 neutralizing antibody (94% vs 96%, Figure 6e ). The potent neutralization activity of the secreted tACE2v demonstrated that tACE2v mRNA delivered via LLNs might serve as an antagonist for blocking SARS-CoV-2 cell entry.

We have designed 20 lipid-like compounds and identified a lead compound for mRNA delivery in vitro and in vivo. We found that the length of alkyl chain and the unsaturation of lipid-like compounds, as well as composition ratios dramatically affected LLN-mediated mRNA delivery. Multiround optimization yielded the optimal formulation tB-UC18 LLNs, which possessed 7-fold lower weight ratio (the key lipid:mRNA) than the typical literature ratio, but displayed 1.6-fold higher mRNA delivery efficiency than a commercial transfection reagent Lipofectamine 2000. It means that a lower dose of lipid-like compound is required for a given dose of mRNA, thus conferring minimal toxicity for in vivo applications. Core-shell-structured tB-UC18 LLNs rendered mRNA durable resistance toward nuclease degradation and boosted intracellular delivery via endocytic pathway and micropinocytosis. Systemic injection of tB-UC18 LLNs in mice led to more than 95% of protein expression from exogenous mRNA in the spleen, without causing significant hematological and histological changes.

Engineered ACE2 receptor mimics have been shown to be capable of neutralizing SARS-CoV-2. [21] [22] [23] [24] [25] [26] [27] [28] 43] To date, however, no peer-reviewed studies are available on mRNA-based decoy protein replacement therapy. We therefore generate tACE2v mRNA molecules through in vitro transcription of engineered ectodomain of human ACE2 gene. Intracellular delivery of synthetic tACE2v mRNA via our nanoscale platform achieves elevated expression of ACE2 decoys, and robust inhibitory action on the binding of SARS-CoV-2 RBD to the human ACE2 receptor in vitro. Further investigation on blockade of SARS-CoV-2 infection is required in an appropriate animal model. As the largest secondary lymphoid organ in the body, the spleen plays a crucial role in the defense against foreign pathogens in the blood. [44] Spleen-mediated secretion of tACE2v should provide both local and systemic protection against SARS-CoV-2 infection via the bloodstream. While tACE2v levels in multiple tissues throughout the body, especially in the lung, remain to be determined, it is expected that the nanoplatform will open a new avenue for further exploiting in vivo applications as an alternative antiviral strategy for COVID-19.

Chemicals and Reagents: Terephthalaldehyde was purchased from TCI. Benzene-1,3,5-tricarbaldehyde was from Jilin Chinese Academy of Sciences, Yanshen Technology. DOPE was ordered from Avanti Polar Lipids. Lipofectamine 2000 was from Thermo Fisher Scientific. β-galactosidase Assay Kit, In Situ β-galactosidase Staining, and Cell Counting Kit-8, nuclear staining solution Hoechst 33342, membrane probe DiO, DiD, ACK lysis buffer, ACE2 activity fluorometric assay kit, and ACE2 inhibitor MLN-4760 were obtained from Beyotime Biotechnology. Triton X-100, Eagle's minimum essential medium, Dulbecco's modified Eagle's medium (DMEM, high glucose), and RNase A were purchased from Sangon Biotech. Chloropromazine was purchased from Macklin Inc. Filipin was purchased from Meilunbio. 5-(N-ethyl-N-isopropyl)-amiloride was purchased from Aladdin. TAMRAlabeled oligos and SARS-CoV-2 Surrogate Virus Neutralization Test Kit were ordered from Genscript. β-gal mRNA was purchased from TriLink BioTechnologies. Firefly luciferase mRNA (fluc mRNA) was from APExBIO. eGFP mRNA was from Rhegen Bio. HiScribe T7 ARCA mRNA Kit (with tailing) and Monarch RNA Cleanup Kit were from New England Biolabs. FITC-conjugated anti-mouse CD3 antibody, PE-conjugated antimouse CD11c antibody, FITC-conjugated anti-mouse CD19 antibody, and PE-conjugated anti-mouse F4/80 antibody were ordered from BioLegend. All other chemicals were analytical grades or above.

Synthesis of Lipid-Like Compounds: To a solution of terephthalaldehyde (pB, 134 mg, 1 mmol) in ethanol/dichloromethane (5 mL, 2:1, v/v) was added R-NH 2 (2.4 mmol) and anhydrous sodium sulfate (2 mmol). The reaction mixture was allowed to stir at 35 °C for 24 h, filtered, and evaporated under vacuum to afford the Schiff base intermediate. The resulting Schiff base was suspended in ethanol/dichloromethane (5 mL, 4:1, v/v). Subsequently, NaBH 4 (76 mg, 2 mmol) was added to the mixture and stirred for 24 h at room temperature. The solvent was removed under reduced pressure, and NH 4 HCO 3 (64 mg) dissolved in ultrapure water (3 mL) was added. The resulting solution was then extracted three times with chloroform, and the organic layer was pooled and washed with ultrapure water. After the solvent was removed, the residue was chromatographed over flash silica chromatography system (SepaBean machine U200) with gradient elution of methanol and dichloromethane over 15 min to afford pB-R. The tB-R was synthesized with the similar method expect different amount of benzene-1,3,5tricarbaldehyde (tB, 1 mmol), R-NH 2 (3.6 mmol), and NaBH 4 (3 mmol) 6M17) . SARS-CoV-2 RBD, ACE2, and B 0 AT1 were colored in red, blue, and purple, respectively. A magnified view of the boxed region depicts a detail of recognition interface between RBD and ACE2, and mutational sites for tACE2v. Crucial residues involved in intermolecular interaction between RBD and ACE2 were highlighted in yellow. Residues (T27, L79, N330, and A386) used for generation of beneficial point mutations for high-affinity tACE2v were labeled at respective positions in green. c) Prediction of the transmembrane helix of the full-length human ACE2 protein (left) and its variant tACE2v (right) by TMHMM 2.0. d) Detection of the tACE2v protein expressed by LLNs-encapsulated tACE2v mRNA. The activity of tACE2v secreted into the extracellular fluid was detected by a fluorometric method. Data were obtained from three biological replicates. RFU, relative fluorescence unit. e) tACE2v-mediated blockage of the interaction between the human ACE2 and RBD of SARS-CoV-2. The concentrated supernatant containing the secreted tACE2v, positive control (SARS-CoV-2 neutralizing antibody), and negative control were prepared in triplicate.

Adv. Mater. 2021, 33, 2101707 was used. Lipid-like compounds were confirmed by means of HR-MS and 1 H NMR (Table S1 , Supporting Information).

In Vitro Transcription of tACE2v mRNA: Protein sequence of wild-type human ACE2 was retrieved from UniProt (aa 1-805, Accession number: Q9BYF1; Table S2 , Supporting Information). The codon-optimized open reading frame sequence for encoding a 732 amino-acid-long tACE2v (aa 1-732; Table S2 , Supporting Information) was provided in Table S3 (Supporting Information). Fixed 5′ and 3′ untranslated regions were constructed to flank the open reading frame and transcribed in vitro using HiScribe T7 ARCA mRNA Kit (with tailing) under the manufacturer's protocol. The capped and polyadenylated mRNA was purified using Monarch RNA Cleanup Kit, and quantified in microvolume by a multimode reader (BioTek, Synergy LX). [45] Self-Assembly of LLNs: LLNs were prepared at various N/P ratios by mixing of ethanol phase containing lipid-like compound and helper lipid with mRNA dissolved in aqueous phase. Typically, the desired quantity of mRNA stock was diluted in PBS, prior to the addition of ethanol phase. The lipid-like compound and the helper lipid DOPE were dissolved in ethanol in varied molar ratios. Two phases at an ethanol-to-aqueous volume ratio of 1:9 were homogenized by rapidly pipette mixing, and left at room temperature for 15 min to form LLNs. For in vivo injection, formulations were prepared at a final mRNA concentration of 0.1 mg mL −1 .

Cell Culture: Human embryonic kidney 293T cells and human cervical cancer HeLa cells were purchased from Stem Cell Bank, Chinese Academy of Sciences. Cells were maintained in DMEM (high glucose) supplemented with 10% (v/v) fetal bovine serum at 37 °C in 5% CO 2 environment.

β-Gal Activity Assay: 293T and HeLa cells were seeded in 96-well plates at density of 20 000 and 6000 cells per well per 90 µL, respectively, and left to adhere overnight. Cells were typically treated with 10 µL of LLNs containing 200 ng of β-gal mRNA. The same dose of mRNA was complexed with a commercial transfection reagent Lipofectamine 2000 (Lipo2k) according to manufacturer's instructions, and used as a control. Approximately 24 h postdelivery, β-gal activity was determined by a colorimetric assay. In brief, a volume of lysis buffer equal to the volume of medium (100 µL) was added to each well and incubated at ambient temperature for 15 min. Cell lysates (50 µL) was then transfer to a well containing equal volume (50 µL) of O-nitrophenyl-β-d-galactopyranoside (a colorimetric substrate for β-galactosidase) and incubated at 37 °C for 2 h. Next, the reaction was stopped by stop buffer, and the absorbance was recorded at 420 nm using a microplate reader (BioTek, Synergy LX). β-Gal activity was normalized to Lipo2k in order to enable comparability between different batches of cells. To assess time and dose effect, cells were exposed to LLNs (10 µL) containing 200 ng of β-gal mRNA for different incubation periods, and LLNs (10 µL) containing different mRNA doses for 24 h, respectively.

In Situ β-Gal Staining: Delivery procedures were performed in 96-well plates using the procedure described above. After 24 h postdelivery, cells were fixed at room temperature for 10 min and then rinsed with PBS for three times. Subsequently, 100 µL of X-Gal solution (a chromogenic substrate for β-galactosidase) was added for in situ β-galactosidase staining. After staining at ambient temperature for 2 h, cells were imaged by a fluorescence inverted microscope (Nikon, ECLIPSE Ts2R).

Fluorescence Imaging: In a separation experiment, eGFP mRNA instead of β-gal mRNA was loaded to LLNs using the above procedures. Following different incubation periods with formulations, the green fluorescent protein was detected using a fluorescence inverted microscope (Nikon, ECLIPSE Ts2R).

Measurement of Hydrodynamic Size and Zeta Potential: Hydrodynamic size and zeta potential of LLNs were measured at 25 °C using dynamic light scattering (Brookhaven Instruments Corporation, 90Plus PALS). Formulations were 5-and 135-fold diluted in nuclease-free water or 10% FBS (v/v) for size and zeta potential measurement, respectively. The effects of pH on the particle size and surface charge of LLNs were investigated by diluting of particles in different buffer (pH 5.8, 7.4, and 9.8) at room temperature. the imaging chamber for bioluminescent imaging using a small animal imaging system (PerkinElmer, IVIS Spectrum). After whole-body imaging, mice were sacrificed and organs were excised for ex vivo imaging.

Flow Cytometry: Fluorescent-labeled tB-UC18 LLNs were prepared by incorporation of a far-red fluorescent dye (DiD) into LLNs (tB-UC18:DOPE:DiD = 1.5:1.5:0.015, molar ratio). The spleen from C57BL/6J mice treated with a single intravenous injection of DiDlabeled tB-UC18 LLNs at 0.5 mg kg −1 were collected for flow cytometric analysis. Briefly, the spleen was mashed using the plunger of a 3 mL syringe and filtered through a 70 µm cell strainer. Red blood cells were removed by ACK lysis buffer. Splenocytes were then stained with FITCconjugated anti-mouse CD3 antibody, and PE-conjugated anti-mouse CD11c antibody, or FITC-conjugated anti-mouse CD19 antibody and PE-conjugated anti-mouse F4/80 antibody for 30 min at 4 °C in the dark. Flow cytometric data were acquired on a DxFLEX flow cytometer (Beckman Coulter) and analyzed using the CytExpert software (Beckman Coulter).

In Vivo Safety Evaluation: C57BL/6J mice received a single intravenous injection of formulation at a mRNA dose of 0.5 or 1 mg kg −1 of animal weight. At the designated time points, mice were euthanized, and whole blood was collected with K 2 -EDTA-coated tubes, and centrifuged for 10 min at 2000 × g. The supernatant was used to measure both liver and renal function. Meanwhile, the major organs of each mouse were harvested, fixed in 4% paraformaldehyde for 24 h, and processed into sections for hematoxylin-eosin staining by standard procedures. Mice without treatment were used as the negative control.

Assessment of ACE2 Activity: The cell culture supernatant (100 µL) from 293T cells treated with tB-UC18 LLNs containing 200 ng of tACE2v mRNA was harvested 24 h post-treatment. ACE2 activity was then detected by a fluorescence resonance energy transfer-based ACE2 substrate (MCA-YVADAPK(Dnp)-OH) according to the manufacturer's protocol with minor modifications. Briefly, cell-free supernatant (5 µL) was incubated in a black 96-well plate with a fluorescent substrate in a total volume of 50 µL of assay buffer. In a parallel experiment, serially diluted ACE2 inhibitor MLN-4760 was added to the assay buffer. Fluorescence was recorded in a kinetic mode for 60 min at room temperature using a plate reader (BioTek, Synergy LX) at an excitation wavelength 325 nm and emission wavelength 393 nm. The relative fluorescence unit (RFU) was plotted as a function of reaction time.

Neutralization Test: 293T cells seed in a 6-well plate were treated with tB-UC18 LLNs at a final tACE2v mRNA concentration of 2 µg mL −1 . The cell culture supernatant was harvested 48 h posttreatment and concentrated 35× using Amicon Ultra-4 centrifulgal filter with a 3 kDa molecular weight limit (Merck). Neutralization activity of tACE2v against SARS-CoV-2 was then evaluated by a SARS-CoV-2 Surrogate Virus Neutralization Test Kit according to the manufacturer's instructions. Briefly, samples were added to the horseradish peroxidaseconjugated-RBD solution (1:1000) at a volume ratio of 1:1 and incubated at 37 °C for 30 min. Mixture (100 µL) was transferred to the well precoated with the hACE2 protein and incubated at 37 °C for 15 min. Substrate reaction was then developed with TMB solution in the dark at room temperature for 15 min and quenched with stop solution. Absorbance at 450 nm was measured by a plate reader (BioTek, Synergy LX), and the inhibition rate was determined using the following calculations: Inhibition (%) = (1-OD 450_Sample /OD 450_Negative control ) × 100%.

Statistical Analysis: All statistical analyses were done with 95% confidence and carried out using GraphPad Prism 6 (* indicates p < 0.05, ** indicates p < 0.01, *** indicates p < 0.001). Data were expressed as mean ± SEM from three replicates unless otherwise stated. Statistical significance was analyzed using an unpaired, two-tailed Student's t-test, One-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test, or two-way ANOVA when applicable.

Supporting Information is available from the Wiley Online Library or from the author.

Proc. Natl. Acad. Sci. USA 2020

Proc. Natl. Acad. Sci

Proc. Natl. Acad. Sci. USA 2020

Transmission Electron Microscopy: A small drop of β-gal mNRA-loaded tB-UC18 (2.5 µL) was applied to a carbon-coated copper grid and allowed to dry overnight at room temperature. Transmission electron microscopy imaging was conducted on a Hitachi HT7700 TEM system at 80 kV accelerating voltage.Gel Retardation Assay: LLNs were formulated at a range of N:P ratios as described above. Naked β-gal mRNA was included as a control. Samples were loaded to a 1% agarose gel at 200 ng mRNA per well, electrophoresed at 120 V for 30 min in TAE buffer, and visualized by an imaging system (Thermo Fisher Scientific, iBright FL1000).Protection Assay: Naked mRNA (10 µL, 200 ng) or tB-UC18 LLNs containing 0.02 mg mL −1 of β-gal mRNA were incubated with FBS (5 μL, 30%, v/v) or RNase A (from 0 µg mL −1 to 150 µg mL −1 ) at room temperature for a given time point. The reactions were stopped by proteinase K at a final concentration of 1 mg mL −1 . Samples were then resolved on a 1% agarose gel and imaged an imaging system (Thermo Fisher Scientific, iBright FL1000).Susceptibility of LLNs to Serum-Induce Aggregation: This assay was conducted as previously described with minor modifications. [46] To monitor aggregation state of LLNs in serum environment, formulations containing 100 ng of β-gal mRNA were incubated with 10% FBS at room temperature or 37 °C in a clear 96-well plate. Absorbance at 660 nm was recorded using a microplate reader (BioTek, Synergy LX).Hemolysis Test: Hemolysis was carried out in a round-bottom 96-well plate through incubating vehicles (100 µL, 30 × 10 −6 m for both tB-UC18 and DOPE) with human red blood cells (100 µL, diluted to 4% in PBS, v/v) collected from an anonymous donor at 37 °C for 60 min. After centrifugation at 1000 × g for 5 min, the amount of hemoglobin released into the supernatant was quantified by measuring the absorbance at 540 nm wavelength using a microplate reader (BioTek, Synergy LX). PBS and 1% Triton X-100 were used to define baseline and 100% hemolysis. [47] Cell Counting Kit-8 Assays: Cell Counting Kit-8 assays were employed to assess the influence of LLNs on cell growth. Briefly, cells (20 000 per well for 293T cells, and 10 000 per well for HeLa cells) in DMEM medium (90 µL) were seeded overnight in 96-well plates, and then incubated with nanocarriers at varying concentration range (0, 2, 4, 8, 16, and 32 × 10 −6 m for both components). After 24 h incubation, CCK8 (10 µL) was added to medium and incubated for 1 or 2 h, and the absorbance was recorded at 450 nm using a microplate reader (BioTek, Synergy LX).Cellular Uptake: Fluorescently labeled mRNA was generated via hybridization of sixfold molar excess of TAMRA-tagged oligo(dT) 17 to the poly(A) tail of β-gal mRNA under the following conditions: heating at 95 °C for 2 min followed by cooling to 20 °C at a rate of 0.1 °C s −1 . TAMRAlabeled mRNA (200 ng) was then formulated into nanoassemblies and added to attached cells. After 4 h of treatment, cells were rinsed twice with PBS, and stain solution (100 µL) containing a membrane staining dye (Dio) and nuclei staining dye (Hoechst33342) was added for 10 min. Cells were then imaged by a fluorescence inverted microscope (Nikon, ECLIPSE Ts2R).Mechanisms of Cellular Internalization: To explore the internalization pathway of LLNs, cells were seeded in 96-well plates at a density of 20 000 cells per well, and allowed to attach overnight. Endocytosis inhibitors including chloropromazine (CPZ, 10 µg mL −1 ), filipin (1 µg mL −1 ), or 5-(N-ethyl-N-isopropyl)-amiloride (EIPA, 10 µg mL −1 ) were added to cells prior to adding LLNs. After 30 min of preincubation, cells were treated with LLNs-encapsulated β-gal mRNA (10 µL) for an additional 24 h. The β-galactosidase activity was then assayed using the procedures described above.In Vivo and Ex Vivo Bioluminescent Imaging: All experiments were conducted using female C57BL/6J mice (6-8 weeks old) from Charles River Laboratories. All animal experiments were carried out in accordance with guidelines approved by the Institutional Animal Care and Use Committee of Shenzhen People's Hospital. For in vivo bioluminescence imaging, mice received a single intravenous (IV), intramuscular (IM), or subcutaneous (SC) injection of tB-UC18 LLNs at a firefly luciferase (fluc) mRNA dose of 0.5 mg kg −1 . At 4 h postinjection, 150 mg kg −1 of d-luciferin potassium salt were intraperitoneally injected to the mice. Eight minutes later, mice were anesthetized by isoflurane and placed in

The authors declare no conflict of interest.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

lipid-like nanoassemblies, messenger RNA, SARS-CoV-2, spleen-targeted delivery systems, truncated ACE2 decoys