Boosting Cholesterol Efflux from Foam Cells by Sequential Administration of rHDL to Deliver MicroRNA and to Remove Cholesterol in a Triple-Cell 2D Atherosclerosis Model

www.small-journal.com 2105915 (1 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH ReseaRch aRticle Boosting Cholesterol Efflux from Foam Cells by Sequential Administration of rHDL to Deliver MicroRNA and to Remove Cholesterol in a Triple-Cell 2D Atherosclerosis Model Shifa Jebari-Benslaiman, Kepa B. Uribe, Asier Benito-Vicente, Unai Galicia-Garcia, Asier Larrea-Sebal, Izortze Santin, Iraide Alloza, Koen Vandenbroeck, Helena Ostolaza, and César Martín* S. Jebari-Benslaiman, A. Benito-Vicente, H. Ostolaza, C. Martín Biofisika Institute (UPV/EHU CSIC) and Department of Biochemistry and Molecular Biology University of the Basque Country UPV/EHU Leioa 48940, Spain E-mail: cesar[email protected] K. B. Uribe Center for Cooperative Research in Biomaterials (CIC biomaGUNE) Basque Research and Technology Alliance (BRTA) San Sebastián 20014, Spain DOI: 10.1002/smll.202105915 1. Introduction Cardiovascular disease (CVD), the leading cause of mortality in industrially developed countries,[1] is primarily caused by atherosclerosis, a progressive inflammatory disease of the arteries.[2–4] Atherosclerosis is characterized by an abnormal lipid and inflammatory cell accumulation in the intima, the subendothelial layer of large arteries.[5] During the development of atherosclerosis, several factors such as blood flow perturbation and hypercholesterolemia alter the function of both vascular endothelial cells (ECs) and smooth muscle cells (VSMCs), the main cellular components of arteries. These factors promote EC dysfunction and changes in VSMCs phenotype,[6,7] leading to vessel wall homeostasis perturbation and lipoprotein diffusion into the arterial wall.[8] Retention of cholesterol-rich lipoproteins in susceptible areas of the arterial vasculature favors low-density lipoprotein (LDL) oxidation, followed by endothelial activation, and macrophage infiltration, which internalize the modified LDL by a nonregulated mechanism leading to foam cell Cardiovascular disease, the leading cause of mortality worldwide, is primarily caused by atherosclerosis, which is characterized by lipid and inflammatory cell accumulation in blood vessels and carotid intima thickening. Although disease management has improved significantly, new therapeutic strategies focused on accelerating atherosclerosis regression must be developed. Atherosclerosis models mimicking in vivo-like conditions provide essential information for research and new advances toward clinical application. New nanotechnology-based therapeutic opportunities have emerged with apoA-I nanoparticles (recombinant/reconstituted high-density lipoproteins, rHDL) as ideal carriers to deliver molecules and the discovery that microRNAs participate in atherosclerosis establishment and progression. Here, a therapeutic strategy to improve cholesterol efflux is developed based on a two-step administration of rHDL consisting of a first dose of antagomiR-33a-loaded rHDLs to induce adenosine triphosphate-binding cassette transporters A1 overexpression, followed by a second dose of 1,2-dipalmitoyl-sn-glycero3-phosphocholine rHDLs, which efficiently remove cholesterol from foam cells. A triple-cell 2D-atheroma plaque model reflecting the cellular complexity of atherosclerosis is used to improve efficiency of the nanoparticles in promoting cholesterol efflux. The results show that sequential administration of rHDL potentiates cholesterol efflux indicating that this approach may be used in vivo to more efficiently target atherosclerotic lesions and improve prognosis of the disease. U. Galicia-Garcia, A. Larrea-Sebal Fundación Biofisika Bizkaia and Biofisika Institute (UPV/EHU, CSIC) Leioa 48940, Spain I. Santin Department of Biochemistry and Molecular biology University of the Basque Country UPV/EHU Leioa 48940, Spain I. Santin, I. Alloza, K. Vandenbroeck Biocruces Bizkaia Health Research Institute Barakaldo 48903, Spain I. Santin CIBER (Centro de Investigación Biomédica en Red) de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM) Instituto de Salud Carlos III Spain The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/smll.202105915. © 2022 The Authors. Small published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons AttributionNonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is pro perly cited, the use is noncommercial and no modifications or adaptations are made. Small 2022, 18, 2105915 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (2 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH differentiation.[9] As a result, atheroma plaque is developed and consequently, the vascular diameter is reduced thus increasing the incidence of cardiovascular events.[10–13] The role of atherosclerosis as the major risk factor leading to cardiovascular complications[14,15] highlights the importance of finding new therapeutic strategies to facilitate the reverse of lipid build-up in the plaques. Cholesterol efflux from macrophages is a key process in reverse cholesterol transport (RCT) and promotes the delivery of excess cholesterol from peripheral cells and tissues to the liver for excretion.[16] RCT occurs through the interaction of apolipoprotein A-I (apoA-I) contained in highdensity lipoproteins (HDL) with the adenosine triphosphatebinding cassette transporters A1/G1 (ABCA1/ABCG1).[17–19] Therefore, targeting RCT and ABCA1 could represent a valuable therapeutic approach to prevent atherosclerosis. The inversely associated relationship between low HDL cholesterol levels and CVD risk in epidemiologic studies highlighted the potential of HDL mimetics (recombinant/reconstituted HDL or rHDL) as a therapeutic tool and inspirational source for biomedical engineering.[20–23] rHDL, designed to mimic the atheroprotective function of endogenous pre-β HDL particles, are complexes of full-length apoA-I or apoA-I mimetic peptides and phospholipids. Their characteristics allow rapid mobilization of cholesterol from periphery to plasma and have shown to reduce atherosclerosis burden in animal models.[24–28] Based on these encouraging data, several rHDL have been designed and used in clinical trials for CVD therapy.[20,29] The discovery of apoA-I Milano, a natural apoA-I variant associated with very low levels of HDL and reduced atherosclerosis,[30,31] allowed formulating ETC-216. This rHDL is composed of apoA-I-Milano and phospholipids that mimics the properties of nascent HDL. In the phase II “The ApoA-I Milano Trial,” ETC-216 showed a mean 2.7% coronary atherosclerosis regression in acute coronary syndrome patients but also development of serious adverse effects in some patients thus its use was discontinued.[32] A new formulation containing apoA-I-Milano, denominated MDCO-216, was tested and proven to be safe in terms of immunostimulatory effects.[33] However, the phase I/II MILANO-Lipids, and Other Surrogate Biomarkers Trial (PILOT) study and the second phase II trial, MILANO-DRIVE demonstrated its lack of efficacy.[34] Another rHDL with ability to rapidly mobilize large amounts of cholesterol into the HDL fraction is CER-001. This rHDL is composed of recombinant human apoA-I and a lipid mixture composed of sphingomyelin and 1,2-dipalmitoyl-sn-glycero-3phospho-(1′-rac-glycerol). Although CER-001 did not cause any significant reduction in coronary atherosclerosis as evaluated in the CHI-SQUARE study,[35] it has been shown a U-shaped CER001 dose-response curve with the greatest atheroma regression occurring at a low concentration.[35] Finally, CSL-112 is a rHDL consisting of human plasma derived apoA-I and soybean phosphatidylcholine (apoA-I : lipid, mol ratio=1 : 55), which arose as an improvement of its predecessor, CSL-111. CSL-111 initially showed an enhanced cholesterol mobilization and improved anti-inflammatory markers, but was disfavored due to its hepatotoxicity.[36] On the contrary, CSL-112 was well tolerated and not associated with any significant alterations in liver or kidney function.[21] Moreover, CSL-112 has been found to enhance cholesterol efflux very efficiently[37] and its beneficial potential in reducing major adverse cardiovascular events will be assessed in the on-going large phase III AEGIS-II study (NCT03473223).[21,29] Importantly, the physicochemical characteristics of rHDLs should be taken into consideration to improve their efficiency on promoting cholesterol efflux.[38,39]. Very recently, the efficiency of cholesterol removal by rHDL mimicking different HDL maturation stages has shown to be higher with 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) rHDL, a nanoparticle that resembles nascent HDL. In fact, cholesterol efflux from macrophageand vascular smooth cell-derived foam cells induced by DPPC rHDL was even more efficient than that induced by rHDL with the lipid composition of CSL-112. Among the physicochemical characteristics of DPPC rHDL underlying this effect are the higher physical binding affinity of cholesterol for saturated long-chain-length phospholipids and the planar geometry of the DPPC lipid bilayer, which favors cholesterol transfer.[38–40] In addition, it has been shown that cholesterol transfer to DPPC rHDLs, which show a planar bilayer, is favored thermodynamically from the high curvature of the cell membrane promoted by ABCA.[40,41] Therefore, the use of well-formulated apoA-I nanoparticles constitutes a significant advance toward clinical application because at therapeutic doses they neither present toxicity nor immunogenicity, making them appropriate for therapeutic application.[42] Based on their biocompatibility, rHDL can be considered ideal carriers for the delivery of drugs and other therapeutic agents, as it has been shown for DNA and synthetic RNA, when incorporating cationic or zwitterionic lipids within rHDLs.[43,44] The discovery that multiple miRNAs participate in the progression of atherosclerosis and in the regulation of RCT by directly targeting ABCA1 has opened new opportunities in the use of nanotechnology-based miRNAs therapeutic platforms.[45,46] One of the most studied and well-known “target” of ABCA1 is miR-33a. As an intronic miRNA, miR-33a expression is linked to that of SREBP2 and both co-ordinately participate in the regulation of intracellular cholesterol levels.[47] Among the multiple roles in the regulation of cholesterol metabolism, miR-33a functionally regulates the activity of ABCA1 interfering with the protein expression. Both therapeutic and macrophagespecific miR-33 knockdown intervention to reduce miR-33 levels provided promising results in the past.[47] The advantages of using miRNAs rely both on their small repressive capacity on any single target gene, which is usually less than twofold and on their inherent ability to target multiple genes in the same biological pathway.[48,49] It has been shown that the inhibition of miRNAs using antisense oligonucleotides promotes RCT through upregulation of the ABCA1 gene.[45,46] Therefore, some studies have used cationic lipid/polymer-based nanoparticles for miRNA delivery in preclinical models,[50] while some others have delivered functional miRNA mimics into macrophages in chitosan nanoparticles to promote RCT in vivo.[51] Nguyen etal. showed that miR-33 can be delivered to naïve macrophages by chitosan nanoparticles and modulates the expression of its target gene, ABCA1, both in vitro and in vivo and have suggested that these miRNA containing nanoparticles can be used in vivo to target atherosclerotic lesions.[51] Small 2022, 18, 2105915 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (3 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH Given the huge complexity of atherosclerotic disease, the use of coculture systems reflecting the cellular complexity of atherosclerosis has been identified as an advantageous approach to in vitro research.[52] The importance of these in vitro studies relies on i) the demand to overcome the translational gap, ii) provide a suitable model to understand the physiological mechanisms underlying RCT, and iii) improve the efficiency of the nanoparticles in promoting cholesterol efflux, functional pathways that are central to the development of atherosclerosis. In that context, in this work we sought to develop a therapeutic strategy to improve RCT and cholesterol efflux based in a two-step administration of rHDL (Figure 1). In a first nanodisc administration, antagomiR-33a-loaded DPPC:cholesteryl ester (CE):1-palmitoyl-2-hydroxy-sn-glycero-3-phosphocholine (LPC) rHDLs have been used to induce the overexpression of ABCA1 transporter, and in a second step, DPPC rHDLs have been used to remove more efficiently cholesterol from foam cells. A triplecompartment cell culture 2D model comprising ECs, SMCs and foam cells has been used mimicking the atheroma plaque components in vitro.[53] This strategy allows independent isolation of each cellular compartment for downstream analysis without cell sorting. We find that sequential administration of rHDL potentiates cholesterol efflux compared to one-step administration of DPPC rHDL. The results indicate that delivery of antagomiR33a by rHDL efficiently upregulates ABCA1 expression and potentiates the cholesterol efflux induced by DPPC rHDL from foam cells in a 2D atheroma model. Overall, these data indicate that this therapeutic approach might be used in vivo to target more efficiently atherosclerotic lesions and improve prognosis of the disease. 2. Results 2.1. Development and Biophysical Characterization of DPPC:CE:LPC and DPPC rHDL HDL were reconstituted with DPPC:CE:LPC (75:20:5% mol) for antagomiR-33a delivery or DPPC alone for cholesterol efflux as indicated in the Experimental Section. ApoA-I:lipid ratio was optimized to 1:125 mol:mol in both compositions. Nanoparticle formation is illustrated in Figure 2A. Once reconstituted, rHDLs were purified by size exclusion chromatography. As shown in Figure 2B, rHDL showed a homogeneous peak centered at 11–13 mL, preceding the elution of free apoA-I at 15mL. Size distribution of nanodiscs determined by dynamic light scattering (DLS) showed an average diameter of the Small 2022, 18, 2105915 Figure 1. Schematic representation of the strategy to improve RCT and cholesterol efflux based in a two-step administration of rHDL. Overexpression of ABCA1 transporter is induced by a first administration with antagomiR-33a-loaded DPPC:CE:LPC rHDLs. A second step, which involves DPPC rHDLs administration, is used to remove more efficiently cholesterol from foam cells. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (4 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH particles of ≈10 nm (Figure 2C,F). Finally, α-helical structure content determined by circular dichroism (CD) showed higher α-helical content in rHDLs compared to free apoA-I (≈2.2–2.5 times, Figure2D and Table 1). These results indicate a correct reconstitution of rHDLs (Figure2B–D). Size and morphology of nanoparticles were determined by negative stain electron microscopy (NS-EM) (Figure2E), which showed a circular morphology, consistent with a discoid shape as shown by typically stacked nanoparticles resembling a “rouleaux” formation and, circular shapes presented by nanodiscs viewed from the top (Figure 2E). Longitudinal and transverse axes were 10.6±0.8 and 3.9±0.4nm, respectively. The longitudinal axis Small 2022, 18, 2105915 Figure 2. Development and biophysical characterization of rHDL. A) Schematic representation of rHDL reconstitution and purification by gel filtration chromatography on a Superdex 200 column. B) Gel filtration profiles of DPPC and DPPC:CE:LPC rHDLs, profiles were monitored by absorbance at 280nm. C) rHDL size profiles determined by DLS. No significant differences between the different rHDL compositions were determined. D) Circular dichroism of DPPC and DPPC:CE:LPC rHDLs and apoA-I protein in solution. θMRE: mean residue ellipticity. E) Representative rHDL transmission electron microscopy images. Magnification 100×. Scale bar of 50nm. F) rHDL diameter determined from DLS (hydrodynamic diameter) and NS-EM images. Size of rHDL (F) was measured as Feret diameter calculated from 1600 particles. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (5 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH was similar to the diameter range (8–10 nm) determined by both DLS and NS-EM (Figure2F). 2.2. Binding of miRNA to rHDL Next, DPPC:CE:LPC rHDL particles were loaded with a control miRNA as illustrated in Figure 3A. Binding efficiency of miRNA was assessed at different miRNA:apoA-I ratios. As shown in Figure3B, maximum miRNA binding was obtained at 1:1 mol:mol ratio. 2.3. Delivery of AntagomiR-33a by DPPC:CE:LPC rHDL into Foam Cells in Atheroma Plaque Model 2.3.1. 2D Atheroma Plaque Model Setup A modular coculture system, which facilitates the separation of each cellular compartment, was used to set up a three-cell 2D-atheroma model. The system allows coculturing ECs, VSMCs, and macrophage-derived foam cells mimicking the vascular compartment thus facilitating the study of dynamics and interaction of rHDL with foam cells. A schematic illustration of the model is shown in Figure 4A. The use of 0.4µm pored transwell inserts allows separation of VSMCs and ECs by a thin transwell membrane. This avoids translayer contamination by the other cell types while at the same time allows to individually isolate each culture layer. Barrier function was validated by determining transendothelial electrical resistance (TEER) of ECs. The calculated TEER value calculated by Ohm’s law, 24.6±1.2Ωcm2 confirmed integrity and permeability of the in vitro barrier and are in agreement with values reported by others.[54,55] As illustrated in Figure 4B, delivery of antagomiR-33a by DPPC:CE:LPC rHDL would promote miR-33a silencing and consequently upregulation of ABCA1 and ABCG1 transporters. 2.3.2. Uptake of DPPC:CE:LPC rHDL by ECs, VSMC, and Foam Cells We next aimed to address the delivery of antagomiR-33a by rHDL through the 2D atheroma model and determined the efficiency of rHDL uptake by ECs, VSMCs and foam cells in confluent populations. To establish the optimal condition for efficient microRNA delivery, different concentrations of rHDL were used (0–100µgmL−1) (Figure4C) and uptake of DiI-labeled DPPC:CE:LPC rHDL by the different cell types was assessed after 24 h by fluorescent microscopy and flow cytometry. As shown in Figure 4, foam cells were extremely avid internalizing DPPC:CE:LPC rHDL (Figure 4C, lower panel). On the other hand, VSMCs showed residual rHDL uptake (Figure4C, middle panel) while ECs internalized rHDL in a moderate way (Figure4C, top panel) compared to foam cells. To quantify rHDL uptake efficiency, cells from the different compartments were isolated after incubation with the nanoparticles and their uptake was analyzed by flow cytometry as described in the Experimental Section. As shown in Small 2022, 18, 2105915 Table 1. α-helical content of apoA-I and rHDL determined by CD. α-Helical content α-Helicity ratio rHDL/ apoA-I apoA-I 31.1±2.0 – DPPC 71.3±2.5a) 2.3±0.2a) DPPC:CE:LPC 67.4±3.4a) 2.2±0.1a) a)α-helical content calculated from ellipticity values measured at 222nm. R is the ratio of α-helicities between rHDLs and free protein. Data represent the mean±SD (n = 3). All measurements were performed independently three times and levels of significance were determined by a two-tailed Student's t-test. * p<0.01 compared to apoA-I. Figure 3. MicroRNA loading of DPPC:CE:LPC rHDL. A) Schematic representation of rHDL loading with microRNA. B) Efficiency of miRNA binding to rHDL determined by qRT-PCR. A mimic control (cel-miR-67 mature sequence) was used at miRNA:protein mol ratios ranging 0.001–10:1. Binding efficiency was calculated as described in the Experimental Section. Data represent the mean±SD of at least three independent measurements. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (6 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH Small 2022, 18, 2105915 Figure 4. DPPC:CE:LPC rHDL are efficiently delivered into foam cells in the 2D atheroma model. A) A schematic illustration of the 2D atheroma model. The system allows coculturing ECs, VSMCs, and macrophage-derived foam cells mimicking the vascular compartment. B) ABCA1 and ABCG1 upregulation in foam cells by the delivery of antagomiR-33a loaded rHDL, which promotes miR-33a silencing. C) Fluorescent images showing rHDL uptake in ECs (top panel), VSMC (middle panel), and foam cells (lower panel) at different rHDL concentrations (0–100µgmL−1). D) Uptake of DiI-labeled DPPC:CE:LPC rHDL concentrations (0–100µgmL−1) by the different cell types assessed by flow cytometry. Data represent the mean±SD of at least three independent measurements. Levels of significance were determined by a two-tailed Student's t-test. * p<0.01 compared to no rHDL addition (0µgmL−1). Scale bars 50µm. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (7 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH Figure4D, and confirming the results obtained by fluorescent microscopy, foam cells showed the highest ability in rHDL uptake, which resulted saturated at 50µgmL−1. Considering the fluorescence determined in foam cell uptake when treated with 100 µg mL−1 as 100%, rHDL uptake by ECs was 34% compared to foam cells while the uptake by VSMCs resulted residual (Figure4D). 2.3.3. Delivery of AntagomiR-33a by DPPC:CE:LPC rHDL to Foam Cells Intracellular delivery of miRNA by rHDL to foam cells was then assessed in the 2D atheroma model. First, delivery of celmiR-67 (a microRNA that is naturally and specifically expressed in C. elegans) was used to set up validation. DPPC:CE:LPC rHDL particles loaded with cel-miR-67 at 1:1 mol:mol ratio were added in the ECs compartment at different concentrations (0–100 µg mL−1 rHDL) and miRNA incorporation into foam cells was determined after 24h incubation with nanoparticles (Figure 5A). Total RNA was purified and intracellular celmiR-67 levels were quantified as indicated in the Experimental Section. As shown in Figure5A, delivery efficiency of miRNA reached maximum values at 100µgmL−1 rHDL. Similarly, a DPPC:CE:LPC rHDL dose-dependent (0–20 µg mL−1 rHDL) uptake assay was performed in macrophage-derived foam cells cultured alone in monolayer. In the absence of the vascular barrier simulated in the 2D atheroma model, the rHDL concentration required to achieve maximal delivery resulted five times lower (10µgmL−1 rHDL) compared with the ones required in the 2D atheroma model (Figure S1, Supporting Information). We next examined the ability of silencing miR-33a by delivering antagomiR-33a in DPPC:CE:LPC rHDL in the 2D atheroma model. Therefore, nanoparticles were loaded with antagomiR-33a and then, 50 µg mL−1 rHDL-antagomiR-33a were added into the ECs compartment and incubated for 48h. Delivery of antagomiR-33a to foam cells by rHDL reduced approximately four times the levels of endogenous miR-33a compared with cells treated with nanodisc carrying control miRNA (Figure5B). The extent of miR-33a silencing was also analyzed by determining the mRNA levels of ABCA1 and ABCG1, two known targets of miR-33a. As shown in Figure5C, treatment with antagomiR-33a loaded rHDL resulted in ≈2.5and twofold higher levels of ABCA1 and ABCG1 compared to control cells (treated with control miRNA), respectively. Similarly, upregulation of ABCA1 and ABCG1 protein levels in foam cells by delivering antagomiR-33a in DPPC:CE:LPC rHDL was confirmed by Small 2022, 18, 2105915 Figure 5. miRNA transfer capacity, miR-33a downregulation, and ABCA1/ABCG1 upregulation by miRNA delivery by DPPC:CE:LPC rHDL to foam cells. rHDLs were added to the cells in Opti-MEM to allow miRNA delivery, mir-33a repression, and protein upregulation. A) Intracellular delivery of miRNA by rHDL to foam cells assessed in the 2D atheroma model. DPPC:CE:LPC rHDL particles loaded with cel-miR-67 at 1:1 mol:mol ratio were added in the ECs compartment at different concentrations (0–100µgmL−1 rHDL). B) miR-33a silencing by delivering antagomiR-33a in DPPC:CE:LPC rHDL (50µgmL−1) in the 2D atheroma model. C) mRNA levels of ABCA1 and ABCG1 after delivery of antagomiR-33a by rHDL to foam cells. MiR-33a and mRNA levels were determined after incubation of rHDL with 2D atheroma model foam cells during 48h by qRT-PCR determined as described in the Experimental Section. D) Upregulation of ABCA1 and ABCG1 protein levels in foam cells by delivering antagomiR-33a in DPPC:CE:LPC rHDL. E) Expression levels of ABCA1 and ABCG1 determined by optical density. The data in (A), (B), (C), and (E) represent the mean±SD of at least three independent measurements. The data in (D) correspond to a representative western blot of n=3. Levels of significance were determined by a two-tailed Student's t-test. *p<0.01 compared to control miRNA. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (8 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH Western blot Figure5D. Expression levels of ABCA1 and ABCG1 determined by optical density were 4.0 ± 0.8 and 2.3 ± 0.6, respectively (Figure5E). 2.4. Cholesterol Efflux Promoted in Foam Cells within the 2D Atheroma Model Finally, the efficiency of cholesterol efflux induced by sequential administration of antagomiR-33a by DPPC:CE:LPC rHDL followed by DPPC rHDL was evaluated in macrophage-derived foam cells loaded with TopFluor Cholesterol (Figure 6A).[56] The efficiency of DPPC rHDL as cholesterol acceptor was compared with natural HDL (Figure6B). Additionally, a comparative study between cholesterol efflux in foam cells cultured in 1D and foam cells grown in the 2D atheroma model was performed. AntagomiR-33a-loaded DPPC:CE:LPC rHDL at 10µgmL−1 was used in foam cells cultured alone and, 50µgmL−1 in the 2D atheroma model. Nanoparticles were incubated for 48h with the cells to allow cargo delivery and ABCA1/ABCG1 upregulation. After cholesterol loading, DPPC rHDL (10µgmL−1) or HDL (10µgmL−1) were administered and incubated for 6h to allow cholesterol efflux (Figure6C). 2.4.1. HDL as Cholesterol Acceptor Incubation with miRNA-devoid rHDL (empty-rHDL) induced a cholesterol efflux to HDL of ≈5% in both 1D and 2D atheroma model cultured foam cells (Figure 6D). The contribution of upregulating ABCA1 and ABCG1 to cholesterol efflux was assessed by incubating the cells with TO901317, an LIVER X RECEPTOR (LXR) agonist, as an internal control of the assay.[57] As shown in Figure6D, TO901317 treatment increased significantly cholesterol efflux to HDL when compared to cells treated with empty-rHDL in both 1D and 2D cultured foam cells (Figure6D). Similarly, upregulation of ABCA1 and ABCG1 by delivery of antagomiR-33a promoted a significant enhancement of cholesterol efflux to HDL. On the other hand, treatment with TO901317 and antagomiR-33a in combination caused a three times higher cholesterol efflux compared to cells treated with empty-rHDL (Figure6D). Combination of TO901317 and antagomiR-33a-rHDL caused a synergistic upregulation of ABCA1 and ABCG1 protein levels in foam cells (Figure S2, Supporting Information). The cholesterol efflux was paralleled by a similar reduction in the intracellular cholesterol content (Figure S3A, Supporting Information). 2.4.2. DPPC rHDL as Cholesterol Acceptor DPPC as cholesterol acceptor showed a higher efficiency in terms of cholesterol efflux compared to HDL (Figure 6E). Treatment with DPPC:CE:LPC rHDL without antagomiR-33a (10µgmL−1 rHDL in 1D or 50µgmL−1 rHDL in the 2D atheroma model) induced a cholesterol efflux of ≈7.5% in both 1D cultured foam cells and 2D cultures (Figure 6E). Stimulation of LXR with TO901317 resulted in a significantly increased cholesterol efflux of 52 and 59% in 1D and 2D cultured foam cells, respectively, compared to cells treated with rHDL without antagomiR-33a (7.3±1.5 vs 11.1±1.3, and 8.3±1.8 vs 13.2±0.6, respectively) (Figure 6E). Delivery of antagomiR-33a by DPPC:CE:LPC rHDL into foam cells (10µgmL−1 rHDL in 1D or 50µgmL−1 rHDL in the 2D atheroma model) increased cholesterol efflux by 100 and 77% in 1D and 2D cultured foam cells, respectively (7.3±1.5 vs 14.6±0.8 and 8.3±1.8 vs 14.7±1.2, respectively). On the other hand, treatment combination with TO901317 and antagomiR-33a also caused a higher effect on cholesterol efflux than cells treated with the agonist and antagomiR-33a alone (Figure6E). The reduction in the intracellular cholesterol content when DPPC rHDL was used as cholesterol acceptor paralleled the cholesterol efflux to rHDL (Figure S3B, Supporting Information). In addition, data indicate that five times higher concentrations of antagomiR-33a loaded rHDL are required in foam cells grown in 2D atheroma model to achieve similar cholesterol efflux concentrations in those cultured in 1D. Being lipid-poor apoA-I the most efficient acceptor of cholesterol from macrophages in the arterial wall via ABCA1 its effects as cholesterol acceptor was also determined. As shown in Figure S4 in the Supporting Information, cholesterol efflux to apoA-I (10µgmL−1) in foam cells treated with empty-rHDL was 15.4±1.2%, in the presence of TO901317 was 21.5±2.1%, when cells were pre-treated with antagomiR-33a-loaded rHDL cholesterol efflux was 27.0±1.4%, and combination of TO901317 and antagomiR-33a-loaded rHDL caused a cholesterol efflux to apoA-I of 37.0 ± 2.8%. These results support that antagomir33a-loaded rHDLs induce ABCA1 in functional level. 3. Discussion Atherosclerotic plaque formation is a complex process in which macrophages play a key role in progression or regression of plaques.[1] During the last decades, disease management has improved significantly,[58] however research must now shift toward the necessity to develop new therapeutic strategies focused on accelerating atherosclerosis regression.[1] Although animal models provide essential information for atherosclerosis research, 2D culture and 3D multicellular atherosclerosis models allow mimicking in vivo-like conditions,[53,59] tackling the interaction between different cells in atherosclerotic plaques and, understanding the dynamics of new therapeutic approaches. Over the past five years, bionanomaterial-based strategies have emerged as therapeutic or theranostic agents for managing atherosclerosis.[60] Several nanoplatforms to direct delivery of pharmaceutical agents to atherosclerotic plaque-associated macrophages have shown to beneficially modulate disease process and improve outcomes. In example, core–shell nanoplatform composed of a poly(d,l-lactide-co-glycolide) efficiently deliver both siRNA against lectin-like oxidized low-density lipoprotein receptor-1 and atorvastatin to the atherosclerotic lesions and, exert a synergistic therapeutic effect on both endothelial cells and macrophages.[61] Similarly, mannose-functionalized dendrimeric nanoparticle (mDNP)-based platforms have been successfully Small 2022, 18, 2105915 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License www.advancedsciencenews.com www.small-journal.com 2105915 (15 of 15) © 2022 The Authors. Small published by Wiley-VCH GmbH Small 2022, 18, 2105915 [60] C. K. W.Chan, L.Zhang, C. K.Cheng, H.Yang, Y.Huang, X. Y.Tian, C. H. J.Choi, Small 2018, 14, 1702793. [61] Y.Zhao, H.Gao, J.He, C.Jiang, J.Lu, W.Zhang, H.Yang, J.Liu, J. Controlled Release 2018, 283, 241. [62] H. He, Q. Yuan, J. Bie, R. L. Wallace, P. J. Yannie, J. Wang, M. G. Lancina, 3rd, O. Y. Zolotarskaya, W. Korzun, H. Yang, S.Ghosh, Transl. Res. 2018, 193, 13. [63] H.He, J.Wang, P. J.Yannie, W. J.Korzun, H.Yang, S.Ghosh, Biomaterials 2020, 260, 120333. [64] J. E. Dahlman, C. Barnes, O. Khan, A. Thiriot, S. Jhunjunwala, T. E. Shaw, Y. Xing, H. B. Sager, G. Sahay, L. Speciner, A. Bader, R. L. Bogorad, H. Yin, T. Racie, Y. Dong, S. Jiang, D. Seedorf, A.Dave, K. S.Sandu, M. J.Webber, T.Novobrantseva, V. M.Ruda, A. K. R.Lytton-Jean, C. G.Levins, B.Kalish, D. K.Mudge, M.Perez, L.Abezgauz, P.Dutta, L.Smith, et al., Nat. Nanotechnol. 2014, 9, 648. [65] H. Yin, R. L. Kanasty, A. A. Eltoukhy, A. J. Vegas, J. R. Dorkin, D. G.Anderson, Nat. Rev. Genet. 2014, 15, 541. [66] O. F.Khan, P. S.Kowalski, J. C.Doloff, J. K.Tsosie, V.Bakthavatchalu, C. B.Winn, J.Haupt, M.Jamiel, R.Langer, D. G.Anderson, Sci. Adv. 2018, 4, eaar8409. [67] K. C. Vickers, B. T. Palmisano, B. M. Shoucri, R. D. Shamburek, A. T.Remaley, Nat. Cell Biol. 2011, 13, 423. [68] C. S. Thaxton, J. S. Rink, P. C. Naha, D. P. Cormode, Adv. Drug Delivery Rev. 2016, 106, 116. [69] J. S.Rink, S.Yang, O.Cen, T.Taxter, K. M.McMahon, S.Misener, A.Behdad, R.Longnecker, L. I.Gordon, C. S.Thaxton, Mol. Pharm. 2017, 14, 4042. [70] S.Yang, M. G.Damiano, H.Zhang, S.Tripathy, A. J.Luthi, J. S.Rink, A. V.Ugolkov, A. T.Singh, S. S.Dave, L. I.Gordon, C. S.Thaxton, Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 2511. [71] J. Robert, M. Lehner, S. Frank, D. Perisa, A. von Eckardstein, L.Rohrer, Arterioscler. Thromb. Vasc. Biol. 2013, 33, 2699. [72] L. Rohrer, P. M. Ohnsorg, M. Lehner, F. Landolt, F. Rinninger, A.vonEckardstein, Circ. Res. 2009, 104, 1142. [73] D.Perisa, L.Rohrer, A.Kaech, A.vonEckardstein, Biochim. Biophys. Acta 2016, 1861, 98. [74] K. Y.Fung, C.Wang, S.Nyegaard, B.Heit, G. D.Fairn, W. L.Lee, Front. Physiol. 2017, 8, 841. [75] S. Sankaranarayanan, J. F. Oram, B. F. Asztalos, A. M. Vaughan, S.Lund-Katz, M. P.Adorni, M. C.Phillips, G. H.Rothblat, J. Lipid Res. 2009, 50, 275. [76] M. P. Adorni, F. Zimetti, J. T. Billheimer, N. Wang, D. J. Rader, M. C.Phillips, G. H.Rothblat, J. Lipid Res. 2007, 48, 2453. [77] A. R.Tall, J. Intern. Med. 2008, 263, 256. [78] M. A. Kennedy, G. C. Barrera, K. Nakamura, A. Baldan, P. Tarr, M. C.Fishbein, J.Frank, O. L.Francone, P. A.Edwards, Cell Metab. 2005, 1, 121. [79] K.Nakamura, M. A.Kennedy, A.Baldan, D. D.Bojanic, K.Lyons, P. A.Edwards, J. Biol. Chem. 2004, 279, 45980. [80] J. J. Repa, G. Liang, J. Ou, Y. Bashmakov, J. M. Lobaccaro, I. Shimomura, B. Shan, M. S. Brown, J. L. Goldstein, D. J.Mangelsdorf, Genes. Dev. 2000, 14, 2819. [81] G. Liang, J. Yang, J. D. Horton, R. E. Hammer, J. L. Goldstein, M. S.Brown, J. Biol. Chem. 2002, 277, 9520. [82] C. R.Sirtori, L.Calabresi, G.Franceschini, Atherosclerosis 1999, 142, 29. [83] A.Murphy, J.Chin-Dusting, D.Sviridov, Clin. Lipidol. 2009, 4, 731. [84] G. Cavigiolio, B. Shao, E. G. Geier, G. Ren, J. W. Heinecke, M. N.Oda, Biochemistry 2008, 47, 4770. [85] N.Greenfield, G. D.Fasman, Biochemistry 1969, 8, 4108. [86] T.Klemetsrud, H.Jonassen, M.Hiorth, A. L.Kjoniksen, G.Smistad, Colloids Surf., B 2013, 103, 158. [87] J. A. Fernandez-Higuero, A. Etxebarria, A. Benito-Vicente, A. C.Alves, J. L.Arrondo, H.Ostolaza, M.Bourbon, C.Martin, Sci. Rep. 2015, 5, 18184. [88] J. Schindelin, I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, J. Y.Tinevez, D. J. White, V.Hartenstein, K.Eliceiri, P. Tomancak, A.Cardona, Nat. Methods 2012, 9, 676. [89] C.Vedhachalam, P. S.Chetty, M.Nickel, P.Dhanasekaran, S.LundKatz, G. H.Rothblat, M. C.Phillips, J. Biol. Chem. 2010, 285, 31965. [90] C. Vedhachalam, L. Liu, M. Nickel, P. Dhanasekaran, G. M.Anantharamaiah, S.Lund-Katz, G. H.Rothblat, M. C.Phillips, J. Biol. Chem. 2004, 279, 49931. 16136829, 2022, 13, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202105915 by CSIC Organizacion Central OM (Oficialia Mayor) (Urici), Wiley Online Library on [28/02/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License