WO2020107088A1 - Microfluidic process for obtaining stealth cationic liposomes - Google Patents

Abstract

The present certificate of addition describes a process for obtaining stealth liposomes in a microfluidic system, with high yield, based on chaotic advection without the formation of micelles, to which end the liposomes are generated in said step using 50% of a water-soluble organic solvent (such as ethanol). At the end of this process, the organic solvent remaining in the final formulation of the liposomes obtained is recovered using a vacuum centrifugal concentrator or distillation concentrator to remove same or reduce the amount thereof, without significantly altering the physical-chemical properties of the liposome structures.

Description

MICROFLUID PROCESS OF OBTAINING STEALTH LIPOSOMES

CATHONICS

Field of. invention:

[1] The present application is a certificate of addition to the patent application BR 10 2017 025862 9, filed on November 30, 2017 and entitled "Microfluidic process for obtaining cationic stealth liposomes, obtained cationic stealth liposomes and their use".

[2] This certificate of addition falls within the field of medical sciences, more specifically with regard to medicinal preparations characterized by special physical forms, and describes a process of obtaining Stealth liposomes in a microfluidic system using a water-miscible organic solvent with high yield based on chaotic advection without the formation of micelles. By using a high ethanol content (50%) when compared to BR 10 2017 025862 9 and dispensing with the use of salt (ionic strength), this certificate of addition avoids possible denaturation of sensitive molecules that may be associated with liposomes ( eg proteins), thus not limiting its applications.

[3] At the end of this process, water-soluble organic solvents (such as ethanol), which are in a concentration equivalent to that of the aqueous phase and remain in the final formulation of the obtained liposomes, are recovered by using a vacuum centrifuge concentrator. to remove or decrease it, without significantly altering the physical and chemical properties of liposomal structures.

[4] Gene therapy is a therapeutic strategy that it acts from the introduction of genetic material in the target cells, aiming at correction or inactivation of the genes responsible for a pathology. However, due to the size, anionic character and chemical instability of the genetic material in the face of cellular fluids, its interiorization in the cell nucleus (transfection) is difficult.

[5] In order to increase the rate of transfection in vitro and in vivo, several controlled delivery systems for nucleic acids have been studied. Among these, cationic liposomes stand out, whose structures are approximately spherical and formed from the self-aggregation of amphiphilic lipids in lamellae. Because the lamellae mimic cell membranes, liposomes have several biological characteristics, which include interactions with biomembranes and several cells, facilitating the internalization and delivery of drugs and nucleic acids. Due to their cationic character, these liposomes can interact electrostatically with genetic material (negatively charged), forming lipoplexes that have the ability to enter cells, as a result of the electrostatic interaction of lipoplexes with the anionic cell membrane.

[6] In addition to the cationic character, the liposomes in question have good compatibility and ease of chemical modification of their surface, which allows the production of liposomes covalently linked to polyethylene glycol (PEG) molecules.

[7] This new generation of liposomes is known as S tealth and is responsible for significantly increasing the circulation time of these nanocarriers in fluids due to the formation of a protective barrier promoted by PEG chains on its surface. In view of this, the optimization of processes for the production of these nanocarriers stabilized with PEG is necessary.

[8] In this context, microfluidics emerges as an advantageous technique for the production of high added value liposomes, working with systems that process small amounts of fluids (IO ^-9 to IO ^-18 liters), using channels with dimensions up to hundreds of micrometers. Recent advances in this science have created new perspectives for the area of gene delivery systems (gene delivery).

[9] The formation of liposomes in these microfluidic processes occurs by the self-aggregation of the lipid molecules through the contact of two soluble fluids with each other, one aqueous and the other organic (which may be alcoholic) and miscible in water (containing the lipids). As diffusion between species occurs, lipids become less and less soluble in the aqueous medium and begin to self-aggregate, forming fragments of lipid bilayers.

[10] As these lipid bilayer fragments grow, they come together in order to decrease the surface area of the hydrophobic ends, ultimately aggregating into spherical vesicles with a bilayer that separates an aqueous medium from the external medium.

[11] However, several technical problems are known in the state of the art involved in processes of obtaining Stealth liposomes using microfluidic devices, such as the undesirable formation of lipid films and micelles on the surface of said device, causing limitations in the control of final processing parameters.

[12] The document WO 2017/106799 Al describes a process for obtaining Lipidoids (aggregates that have, in addition to lipids, synthetic molecules for structuring in membrane) for the delivery of mRNA, which makes use of microfluidic mixers to perform advection chaotic. Among the phospholipids that can be used in this invention, mention is made of EPC, DOTAP and DOPE, as well as the PEG polymer associated with the lipid DSPE.

[13] However, the process proposed in this document does not form liposomes or Stealth liposomes, nor does it mention solvent recovery. In contrast, according to this certificate of addition, the organic solvent is completely removed by distillation, allowing the unilamellar liposomes obtained to be mixed with any other solutions for various uses, especially for pharmaceutical purposes. In addition, this process proposes the use of synthetic molecules with a different structure from the lipids normally used for the construction of liposomes.

[14] In US 2018/0028664 Al, lipid compositions for the release of active substances in vivo are described. This document aims at the formation of Lipid Nanoparticles (Lipid Nanoparticles), which can comprise phospholipids, such as DOPE, and modified lipids, such as DSPE-PEG. These lipid nanoparticles are obtained by microfluidization, such as chaotic advection, in which the solvent used is recovered by dialysis. In this case, the formation of these nanoparticles occurs through the simultaneous mixing of the lipids and the components to be encapsulated, in the mRNA case, for the formation of the nanoparticle. In this situation, the formation of liposomes or Stealth liposomes is not guaranteed.

[15] In contrast, the present certificate of addition provides for obtaining Stealth liposomes (whose structure differs from lipid nanoparticles), in which the solvent is recovered after chaotic advection by vacuum distillation.

[16] In US 2010/0022680 Al, reference is made to methods for the production of nanoparticles, such as liposomes, for the release of drugs. Among these methods, the use of microfluidic devices stands out, with the removal of solvents carried out by evaporation using vacuum.

[17] In this document, there are no references or scientific evidence that vacuum distillation or concentration maintains the properties of liposomes, as provided for in this certificate of addition.

[18] In turn, the article "Microfluidic manufacturing of phospholipid nanoparticles: stability, encapsulation efficacy and drug release" also refers to the preparation of liposomes that comprise phospholipids using microfluidic methods, such as, for example, using a low profile micro mixer chaotic.

[19] Unlike the present certificate of addition, the analysis of the physical-chemical, morphological and structural characteristics of these liposomes does not mention whether they are unilamellar. In addition, the process has a dialysis step to remove the non-encapsulated drug, without mentioning how the product obtained is purified.

[20] Finally, the article entitled "High-throughput manufacturing of size-tuned liposomes by a new microfluidics method using enhanced statistical tools for characterization ", published by Kastner et al. (2015), refers to the investigation of conventional LC production, using phospholipids 1, 2-dioleoil-sn-glycero-3 phosphatidylethanolamine (DOPE) and 1 , 2-dioleyl-3-trimethylammonium propane (DOTAP) in a microfluidic device called "Staggered Herringbone Micromixer" with 2 inputs (in this case, a microfluidic device based on chaotic advection) which was supplied by the company "Precision NanoSystem Inc." In the referred article the authors analyzed the effect of the flow rate ratio of ethanol and water introduced into the microchannel, in which the following flow rate ratios were evaluated a) 1 ethanol: 1 water; b) 1 ethanol: 3 water; c) 1 ethanol: 5 water, however, the authors of the article characterized LCs only using the Dynamic Light Scattering technique (Dynamic Light. Scattering, DLS), and the size distributions were not demonstrated in this study. res stated that they produced LC using the DOPE-DOTAP composition; which contradicts the knowledge of the literature {An Inverted Hexagonal Phase of Cationic Liposome-DNA Complexes Related to DNA Release and Delivery, Koltover, 1998), which clarifies that the formulation containing DOTAP / DOPE in the proportion 1: 1 does not provide only the formation exclusive of lipid bilayers, but the coexistence of the lamellar (La) (forming liposomes) and inverse hexagonal (HII) phases (another structure other than liposomes). With this, unstable and different structures are formed. In terms of application in the pharmaceutical area, this variation leads to differentiated interaction in the cells causing great variation. THE inclusion of EPC (50 mol%) allows the exclusive formation of lipid bilayers characterizing the formation of liposomes, as already proven by Balbino et al. (2012). Thus, although Kastner et al. (2015), who uses DOPE / DOTAP, cite what forms liposomes, there is no structural evidence to support the exclusive formation of liposomes.

[21] Unlike the 2015 article, the present certificate of addition of invention refers to an improved process of production of cationic S tealth liposomes without the presence of inverse hexagonal (HII), in which said process uses the phospholipid " egg phosphatidylcholine (EPC) "and the lipid DSPE-PEG (2000) associated with the DOTAP / DOPE formulation, in addition to a device with a different configuration.

[22] Therefore, no state-of-the-art document describes a microfluidic process for obtaining S tealth liposomes, involving a high amount of ethanol and its subsequent removal via distillation as now proposed in this certificate of addition.

[23] It is important to note that there is no investigation in the literature of LC of the Stealth type in microfluidic system based on chaotic advection. Publications and patents generally focus on the formation of "lipid nanoparticles", which is the joint synthesis of structures with nucleic acids, intended for gene therapy. In these studies, they even include lipids with PEG, but there is no comparison with the formation of liposomes. In addition, there is a technological bottleneck associated with the production of LC of the Stealth type, whose technical problem to be overcome is the simultaneous formation of micelles and lamellae, since the DSPE-PEG can be used to produce micellar system (Wu et al. 2013). In aqueous solution, during the diffusion process between DSPE-PEG and other phospholipids (EPC / DOTAP / DOPE), the times of micelle formation (IO ^-3 to IO ^-6 seconds) and phospholipid bilayer formation (10 ^{- 1} to 10 ^-3 seconds) are different (Evans and Wennerstrõm, 1999). As the time for the formation of micelles is shorter, the generation of micelles is favored. The diffusion processes used in microfluidic systems for the production of LC of the Stealth type are not very efficient due to the formation of nanoparticles similar to the micelles and possible accumulation of these particles along the channels, resulting in unfavorable physicochemical properties. Another crucial parameter in LC synthesis of the Stealth type is to incorporate the maximum amount of PEG into the phospholipid bilayer before PEG is converted into micelles, thus avoiding the micellar phase and promoting the lamellar phase to produce LC with adequate lipid-PEG insertion. Several factors influence the phase transition (from the lamellar to the micellar phase) of phospholipid-PEG conjugates. The main reason behind this phase transition has been shown to be the polymorphic property of phospholipids, which is the property of amphiphilic molecules that form aggregates of lipids under different conditions, such as water content, organic solution, temperature and pH (Hristova et al 1995). Among others, water content and ethanol content (as an organic solution) introduced into the system has a significant influence on the formation of LC of the Stealth type. For this reason, investigating the relationship between flows (Flow Rate Ratio - FRR, which is the ratio of the sum of lateral flows to the central flow) is very important and was considered for the solution proposed in the present invention addition certificate.

Brief description of the figures

[24] Figure 1 shows a geometric design of fishbone-like structures from a microfluidic device based on chaotic advection.

[25] Figure 2 shows flow configurations in tested microfluidic devices. (A) diffusion-based microfluidic device (D-MD); (B) microfluidic device based on chaotic advection with 3 entrances; (C) Microfluidic device based on chaotic advection with 2 inputs. The microfluidic devices shown in B and C were used for the formation of CL with variable FRR (flow rate) (between 1 and 10), while the D-MD in (A) was evaluated using only FRR 10. The TFR (total flow ratio) was adjusted to 120.12 pL / min in D-MD, while it varied between 100-1500 pL / min in other devices.

[26] Figure 3 shows the particle agglomeration possibly caused by the presence of DSPE-PEG (2000) along the central channel during LC production of the Stealth type in D-MD (A); The size distribution by intensity and volume of LCs of the Stealth type characterized by DLS (Dynamic Light Scattering) (B); Physico-chemical properties of LCs of the Stealth type produced in the same microfluidic device (C). The central flows (10.92 pL / min) and lateral flows (54.6 pL / min each entry) are phospholipids and PEG dispersed in ethanol and aqueous solution, respectively. The lines in each size distribution represent the profile of three independent replicates.

[27] Figure 4 shows the production of LC in microfluidic device based on chaotic advection. The first image was taken during the device's first herringbone-like units, while the second and third images are taken in the middle of the device. The first and second images were taken at 4X magnitude and the third image is taken at 10X. The arrows show the direction of the flow. The images were transformed from 2D objects into 3D with a computer program to facilitate visualization.

[28] Figure 5 shows the size distribution based on intensity and volume of conventional and Stealth LCs produced in CA-MD with 3 entries in FRR 1 and TFR 500 pL / min. The lines in each size distribution represent the profile of three independent replicates.

[29] Figure 6 shows the scattering intensity of X-rays at low angles of conventional and Stealth-type cationic liposomes produced by CA-MD with 2 and 3 inputs with a total flow rate of 500 pL / min. A: Measure taken before ethanol removal; B: After removing ethanol.

[30] Figure 7 shows morphological images of conventional and Stealth-type LCs obtained from electron cryo-microscopy (Cryo-TEM).

[31] Figure 8 demonstrates the efficiency of transfection of conventional and Stealth type LCs into PC3 cells. The efficiency of transfection of lipoplexes containing eGFP pDNA was measured by the percentage of cells (left panel) and the fluorescence intensity of GFP (right panel). The graphs show the mean and standard deviation of the samples; n = 9. (a) represents the statistical difference (p <0.05) between the sample mean and the control mean (CTRL) corrected by Dunnett's test; (c) represents the statistical difference (p <0.05) between the sample mean and cationic liposomes containing PEG (CL-PEG) corrected by Dunnett's test.

Brief description of the invention:

[32] The present invention relates to a process of obtaining Stealth liposomes in a high performance microfluidic system based on chaotic advection.

[33] Stealth liposomes are formed from egg phosphatidylcholine (EPC), 1,2-dioleyl-3-trimethylammonium propane (DOTAP) and 1,2-dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE) phospholipids, in addition to 1,2-distearyl-sn-glycero-3-phosphoethanolamine associated with polyethylene glycol (DSPE-PEG) in a molar ratio of 50: 25: 24: 1.

[34] The microfluidic device used was manufactured using the soft polydimethylsiloxane (PD S) lithography technique, in which two streams corresponding to the aqueous phase and the stream containing phospholipids were used together with DSPE-PEG (2000) dispersed in ethanol. Due to the presence of repetitive microchannels in the microfluidic device, an efficient mixing occurs to promote the formation of liposomes.

[35] At the end of this process, the water-soluble organic solvents (such as ethanol), remaining in the final formulation of the obtained liposomes, are recovered by using a vacuum centrifuge concentrator to remove or decrease its content, without changing significantly the physical and chemical properties of liposomal structures.

[36] Thus, in a first aspect, this Certificate of Addition refers to a microfluidic process of obtaining cationic liposomes and S tealth that comprises microfluidic device based on chaotic advection and the steps of:

(i) insertion of 5 to 25 mM of egg phosphatidylcholine,

1.2-dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE),

1.2-dioleyl-3-trimethylammonium propane (DOTAP) and 1,2-distearyl-sn-glycero-3-phosphoethanolamine with polyethylene glycol (DSPE-PEG) in the proportion ranging from 50: 20: 25: 5% to 50:24: 25: 1% EPC: DOPE: DOTAP: DSPE-PEG respectively, dispersed in 100% anhydrous ethanol in a microfluidic device, from a total flow rate ranging from 120 pL min ^-1 to 1500 pL min ^-1 , the resulting dispersion at the outlet of said microfluidic device comprising ethanol content in the range of 33.5 to 51%, preferably 50%;

(ii) means for recovering the excess organic solvent and adjusting the lipid concentration, in which the said means comprise conventional distillation or large-scale distillation, and when it is carried out by the latter the removal of the solvent occurs in a processing interval of 1 2 hours in a vacuum centrifuge concentrator.

[37] The said microfluidic device comprises 2 (two) or 3 (three) entrances and consists of two bases of poly dimethyl siloxane (PDMS) sealed with O2 plasma, where one base contains microchannels and the other base contains pools, constructed by the soft lithography technique using SU-8 photoresist, preferably in silicon wafer, in which the referred pools have fixed height along said microchannels that cause advection chaotic.

[38] In the realization of this Certificate of Addition in which the microfluidic device comprises 2 (two) inlets, the total flow of each inlet must be equal, resulting in individual flows of 60 to 1000 mΐ / min.

[39] In the concretization of this Certificate of Addition in which the microfluidic device comprises 3 (three) entries, the operating parameter FRR (flow rate ratio which consists of the sum of the lateral current flows, aqueous, by the flow of the central current, organic) should preferably have a value of 1 when the lipid dispersion is inserted into the central stream and aqueous phase on the sides, said aqueous phase comprising water or conventional buffer (such as PBS IX). The arrangement can be changed as long as the preferred 1: 1 ratio of aqueous phase and lipid dispersion is maintained. The lateral flows can vary from 30 to 500 mΐ / min and the central flow from 60 to 1000 mΐ / min, and the total production flow can operate from 100 to 2000 pL / min.

[40] The process as proposed here allows the exclusive formation of the lamellar phase (La) (forming liposomes) and prevents the formation of the inverse hexagonal phase (HII) (another structure other than liposomes).

[41] In a second aspect, this Certificate of Addition refers to the cationic Stealth liposomes obtained as described above and comprise from 50: 20: 25: 5% to 50: 24: 25: 1% EPC: DOPE: DOTAP : DSPE-PEG, preferably 50: 24: 25: 1%; and final lipid concentration of 25 to 10 mM, preferably 12.5 mM, in which said concentration of the final lipid dispersion can be modulated, as the operation removal of ethanol via vacuum concentrator or distillation does not alter the physico-chemical properties of liposomes to a critical concentration.

[42] Said liposomes have an average diameter ranging from 250 to 120 nm, preferably 180 nm; polydispersity index (PDI) ranging from 0.13 to 0.3, preferably 0.25; and zeta potential ranging from +45 to +55 mV, preferably +50 mV when produced in ultra pure water.

[43] In a third aspect, this Certificate of Addition refers to the use of cationic S tealth liposomes, as defined above, because they are applied in the agricultural, food, cosmetic, medical areas; in systems of sustained release of biopharmaceuticals, protein vectors, anti-cancer agents; and preferably in gene therapy and vaccination.

[44] Thus, the Stealth liposomes obtained have applications in several areas, especially the medical area, in systems of sustained release of biopharmaceuticals, including proteins, anticancer agents and mainly applications in gene therapy and vaccination.

Detailed description of the invention:

[45] The present application is a certificate of addition to the patent application BR 10 2017 025862 9 and this application refers to a process for obtaining Stealth type liposomes in a high-performance microfluidic system based on chaotic advection. The process also includes a step of recovering the organic solvent used by using a vacuum centrifuge concentrator to remove or decrease it, and also differs from the patent application BR 10 2017 025862 9 regarding the configuration of the microfluidic device used. Another difference in relation to the patent application BR 10 2017 025862 9 is that the use of high ionic strength is not necessary, such as the use of PBS buffer in high concentration (5X). In this addition certificate, Stealth liposomes are formed from a process with a higher ethanol content (50%), while in patent application BR 10 2017 025862 9 10% is used.

[46] This process comprises the step of inserting 25 mM egg phosphatidylcholine (EPC), 1,2-dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE) and 1,2-dioleyl-3-trimethylammonium propane ( DOTAP), as well as 1,2-distearyl-sn-glycero-3-phosphoethanolamine with polyethylene glycol (DSPE-PEG) in a proportion of 50: 24: 25: 1% EPC: DOPE: DOTAP: DSPE-PEG, respectively, dispersed in 100% anhydrous ethanol in a microfluidic device, from a total flow that varies from 120 pL min ^-1 to 1500 pL min ^-1 . This alcoholic stream is mixed with an aqueous stream (which does not require salt in its composition), in a proportion of 50% of each stream. The mixing takes place in a microfluidic device based on chaotic advection that allows operation at high productivity, unlike the patent application BR 10 2017 025862 9.

[47] At the end of the process, to remove excess and recover the organic solvent used, a vacuum centrifuge concentrator is used for a processing interval of 1 to 2 hours, without this step causing a significant change in terms of size , polydispersity, zeta potential and liposome morphology, in addition to maintaining biological activity. [48] This microfluidic device is constructed of poly dimethyl siloxane (PDMS) using the soft lithography technique, according to the methodology already described by Duffy and McDonald (1998).

[49] Therefore, through the process described here, using devices all in PDMS and allowing the mixing of side currents composed of aqueous solutions (with or without salt), followed by the step of removing the organic solvent by distillation, liposomes are obtained Stealth with excellent physical-chemical characteristics in reproducible processes, including controlled size and low polydispersity and unilamellar index.

[50] The developed technology makes it possible to apply Stealth liposomes in several areas, especially the medical field, in biopharmaceutical sustained release systems, including protein vectors, anticancer agents and mainly applications in gene therapy and vaccination.

[51] More details on the materials and methods comprised in carrying out the present certificate of addition of invention, including the manufacture of the microfluidic device, the formation of the cationic liposome in said device, as well as the removal of ethanol from the liposomal solution, follow below.

Materials

[52] 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), 1,2-dioleo-3-trimethylamine-propane (DOTAP) and egg phosphatidylcholine (EPC) were purchased from Lipoid (Ludwigshafen, Germany) and used as phospholipids to form cationic liposomes (LC) without further purification. 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N- [amino (polyethylene glycol) -2000] (DSPE-PEG (2000)) was used to form Stealth-type liposomes and purchased from Avanti Polar Lipids, Inc. (Alabama, USA). Ethanol was obtained from Labsynth (São Paulo, Brazil) and used after the dehydration process with molecular sieve to disperse phospholipids. The Sylgard® 184 Silicone Elastomer Kit, used to manufacture microfluidic devices, was purchased from Dow Corning (Auburn, MI, United States). Deionized water was obtained from Samtec Biotechnology (São Paulo, Brazil). The pDNA was amplified in Escherichia coli bacteria and purified using PureLink ™ HiPure Plasmid pDNA Purification Kit-Maxiprep K2100-07 (Invitrogen, Carlsbad, CA, United States). The quality and quantity of pDNA was measured spectrophotometrically with an ND-1000 NanoDrop UV-vis spectrophotometer (PeqLab, Erlangen, Germany).

Manufacture of microfluidic devices

[53] Chaotic advection-based microfluidic device (CA-MD), with 2 and 3 inputs, were microfabricated according to the method previously described by McDonald et al. (2000) using the Sylgard® 184 Silicone Elastomer Kit as a precursor material for polydimethylsiloxane (PDMS). The geometrical design of the main channel and herringbone-like units (Figure 1) of the microdevice were designed separately in AutoCAD software. After photolithography, the mask layouts were exposed to photo-plotting with a resolution of 8000 dpi, and the UV exposures were made in an aligned MJB-3 UV300 contact mask (Karl-Suss, Garching, Germany). The channel units in Ύ ' and fishbone, being two layers of PDMS, were irreversibly sealed using the O2 plasma surface activation technique (Plasma Technology PLAB SE80, plasma cleaner, Wrington, England). Before sealing two layers of PDMS, the surface of the layer containing the main channel was kept in KOH solution for 10 seconds to increase the sealing efficiency of the PDMS layers and then washed with ethanol to remove the unbound KOH solution. To strengthen the mechanical properties of the microdevice, it was covered with an additional layer of PDMS. Both 2-port and 3-port CA-MD have 40 herringbone-like subunits. A similar method was used for the microfabrication of the diffusion-based microfluidic device (D-MD, diffusion-based microfluidic device). The PDMS channels and the glass were irreversibly sealed by oxidizing the surfaces using an oxygen plasma cleaner (Plasma Technology PLAB SE80 plasma cleaner, Wrington, England). All D-MD channels had a rectangular cross section with a depth of 100 pm and a width of 140 pm.

Synthesis of cationic liposomes in microfluidic systems

[54] The production of conventional and stealth LCs was carried out using D-MD (as a control) and CA-MD with 2 and 3 inputs. EPC, DOTAP and DOPE with a 50:25:25% molar ratio were dispersed in anhydrous ethanol with a lipid concentration of 25 mM. For the production of LC of the Stealth type, the molar ratio of EPC, DOTAP, DOPE and DSPE-PEG (2000) was 50-25-24-1%. Before production, the lipid solution was sonicated for 15 min at 35 ° C (Ultrasonic Cleaner 8890, Cole-Parmer). THE lipid solution and deionized water were loaded into glass syringes with varying volumes (Hamilton, NV, United States). Syringe pumps (PHD Ultra, Harvard Apparatus, Holliston, MA, United States) were used to maintain constant infusion rates for each solution on the D-MD and CA-MD devices. The production of LC in D-MD was tested only in FRR 10 (Figure 2A). For the CA-MD device with 2 ports, the lipid and aqueous solutions were introduced through two ports with the same flow rate (Figure 2B). For the CA-MD device with 3 inputs, different flow configurations were tested and the effect of each configuration on the formation of LC was evaluated. The tested configurations of the CA-MD with 3 inlets were as follows: (1) phospholipid solution through the central inlet, aqueous solutions from the two side inlets with the relationship between the sum of the lateral flows by the central flow (Flow rate ratio) - FRR of 10; (2) phospholipid solution through the central inlet, aqueous solutions in the two side inlets with FRR of 1; (3) lipid solution at a side entrance, aqueous solutions at the central entrance and another at the lateral entrance (Figure 2C and 2D). The total flow rate (TFRs, total flow rates) varied between 100 and 1500 pL / min for each flow configuration using the CA-MD with 2 and 3 inputs. In our study, the FRR flow rate ratio is defined as the ratio between the sum of the lateral flows and the central flow, while the TFR (Total Flow Rate) is the total flow volume (pL) introduced in the input channels of the microfluidic device per minute and that consequently generates the production flow of the liposomes. During each production, the sample was collected during three different moments.

Ethanol removal from liposome solution

[55] The total concentration of ethanol in the final liposome solution was 50% for CA-MD with 2 entries and CA-MD with 3 entries, this generates an FRR 1. However, when FRR 10 was used, the content of ethanol was 10% for D-MD and CA-MD with 3 entries. To remove ethanol from the final solution, the vacuum centrifugal concentrator (Savant DNA 110 SpeedVac Concentrator, Marshall Scientific, New Hampshire, United States) was used using the average drying rate. The same procedure was also used for the concentration of liposomes for further analysis in techniques of Transmission Cryo-Microscopy (Cryo-TEM) and scattering of X-rays at low angles (SAXS), which require highly concentrated lipid formulation.

Chemical physical analysis

[56] The average hydrodynamic size weighted by the intensity and the polydispersity index (Pdl) of the LCs were measured using dynamic light scattering (Zetasizer NanoZS, Malvern, UK). The backscatter measurement configuration was used, where the detection was performed at a scattering angle of 173 °. The He / Ne laser wavelength and the power of the energy source for the analysis of the lipid particles were 633 nm and 4.0 mW, respectively. All samples were diluted before measurement. The zeta potential was determined using the Laser Doppler Anemometry technique. The measurement was performed in triplicate in water at 25 ° C (Zetasizer NanoZS, Malvern, United Kingdom).

Structural characterization using synchrotron SAXS [57] The structural information of the produced LCs, such as the fraction of multilamellar vesicles in relation to small unilamellar vesicles (SUVs, small unilamellar liposomes) and the size of the bilayer were obtained by means of X-ray scattering analysis at low angles (SAXS , Small Angle X-ray Scattering). The experiments were carried out on the SAXS1 beamline of the National Synchrotron Light Laboratory (LNLS) using a beam with an energy of 8.3 keV (l = 0.149 nm ^-1 ) and a sample-detector distance of 880 mm. The intensity of the scattered beam is displayed as a function of the momentum vector transfer module Q = 4n sen (Q) / l, where 2Q is the scattering angle and l is the radiation wavelength. The typical range of Q values was between 0.20 nm ^-1 and 5.0 nm ^-1 .

[58] The analyzes were performed using the models described in previous works (Balbino et al. 2012, Balbino et al. 2013). We assumed that the system consisted of two phases: a fraction of single bilayer vesicles and a fraction of multilayer vesicles. Therefore, the scattering intensity I (Q) can be written as follows:

I (Q) = P (Q) Seff (Q)

where P (Q) is the bilayer form factor and S _eff (Q) = fs _b + fm SMCT (Q) is the effective structure factor, where f _Sb is the simple bilayer fraction, f _m = (l- fs _b ) represents the fraction of vesicles in multilayers and SMCT (q) is the Modified Caillé structure factor.

Morphological analysis.

[59] The morphology of lipoplexes (LCs and pDNA) produced in bulk and the LCs produced by microfluidics were analyzed using Transmission Cryo-Microscopy (Cryo-TEM). An automated glazing system (Vitrobot Mark IV, FEI, Netherlands) was used for the preparation of the gratings. The grids were exposed to the incandescent discharge treatment (easiGlow discharge system, Pelco) with a current of 15 mA for 25 s to increase hydrophilicity. The samples were prepared at controlled temperature (22 ° C) and humidity (100%) to prevent sample evaporation. 3 µl of sample was placed on a 300 mesh carbon-coated copper grid (Ted Pella) with a deposition time of 3 s. Subsequently, the samples were evaluated with Jeol JEM-2100 operating at 200 kV with a defocus range of -2 to -4 pm. The images were obtained using a CMOS F-416 camera (TVIPS, Germany). Data acquisition was performed at the Electron Microscopy Laboratory of the National Nanotechnology Laboratory (LNNano, CNPEM, Brazil).

Transfection studies with PC-3 cells

[60] Caucasian prostate adenocarcinoma cells (PC-3) were maintained in RPMI medium (supplemented by 10% fetal bovine serum and 1% penicillin and streptomycin) under a humid atmosphere at 37 ° C and 5% CO2. The cells were divided using trypsin / PBS every 2 or 3 days. The transfection assay was performed on PC3 cells using the plasmid pEGFP-N1 (BD Biosciences Clontech, Paio Alto, USA) (4.7 kbp). The plasmid was amplified in Escherichia coli and purified using the PureLink ™ HiPure Plasmid pDNA Purification-Maxiprep K2100-07 Kit (Invitrogen, Carlsbad, CA, United States) following the manufacturer's instructions. The quality and quantity of pDNA were spectrophotometrically measured with an ND-1000 NanoDrop UV-vis spectrophotometer (PeqLab, Erlangen, Germany). For transfection experiments, 105 cells / well were added to a 12-well plate and allowed to adhere overnight. The next day, the cells were washed once with PBS and incubated with LC or LC Stealth type lipoplexes for 4 hours. After 4h, the medium was replaced with complete RPMI medium, and the cells were kept in the incubator for 24h. Then, the cells were washed with PBS, trypsinized and analyzed on the FACScalibur cytometer (BD, USA).

Statistical analysis

[61] All data obtained were expressed as the mean of triplicates with standard deviation (SD). The Tukey test was used to make multiple comparisons of values obtained for different flow rates.

Results and discussion

Synthesis of conventional and Stealth type LCs in a diffusion-based microfluidic device

[62] The production of LC in a diffusion-based microfluidic device (D-MD) using phospholipids DOPE / DOTAP / EPC was previously demonstrated by Balbino et al. 2013. In this procedure, the total flow rate (TFR) and the flow velocity were 120.12 pL / min and 143 mm / s, respectively, with a flow rate (FRR) ratio of 10 (leading to a final concentration 10% ethanol). The initial concentration of lipids in the ethanolic dispersion was adjusted to 25 M, while the lipid concentration in the produced liposome solution was 2.27 mM. LCs with mean diameter of 146.8 + 4.9 nm, Pdl 0.12 ± 0.04 and potential were obtained zeta 55.5 ± 2.1 mV, confirming the high reproducibility of this procedure. Physico-chemical parameters of LCs produced by D-MD were subjected to 7 months of storage at 8 ° C. In addition, previous work has shown that conventional LCs produced in this system were unilamellar and capable of transfecting human epithelial carcinoma cells (HeLa) and human prostate cancer cell line (PC-3) using plasmid DNA (pDNA) (Balbino et al 2012, Balbino et al 2013, Balbino et al., 2016, Balbino et al., 2017) and interference RNA (siRNA) (E§ et al. 2018).

[63] The importance of producing cationic liposomes modified with PEG polymer has been previously investigated (Kolate et al. 2014). In the first investigation, the D-MD was tested for the production of LCs of the Stealth type. Using the same conventional LC production process configuration developed by Balbino et al. 2013, adding DSPE-PEG (2000) (1% of the total lipid molar quantity) in the same lipid / ethanol dispersion in the central stream (Figure 2A). However, along the central microchannel, intense particle agglomeration was observed (Figure 3A), which was a significant factor in altering the final concentration and lipid composition of the produced LCs. This behavior was not observed in conventional LC production (without PEG) and a possible hypothesis is that the diffusion of molecules during the mixture between ethanol and water, facilitates aggregates or clusters of irregular particles (Figure 3A), in the presence of DSPE-PEG (2000). Despite the formation of aggregates along the channel, it was possible to produce nanoaggregates (Figure 3B and 3C). However, when characterized via Dynamic Light Dispersion (DLS) technique (Figure 3B), the size distribution by intensity revealed the presence of populations with different sizes (approx. 30 nm, 125 nm and the third with larger size). As the size distribution by intensity is sensitive to the presence of different populations (the diffused light is proportional to the diameter generated by the sixth power) (Egelhaaf et al. 1996), the relevance of different populations in size by volume was observed. In this distribution, the most relevant population is 30 nm, characterizing the formation of micelles possibly caused by PEG molecules (Figure 3B) (Johnson and Prud'homme, 2003).

[64] A critical issue associated with the production of LC of the Stealth type is the uncontrollable formation of micelles (Hu et al., 2012), since the PEG polymer has been used to form micelle particles for drug delivery applications ( Danafar et al. 2017). In aqueous solution, during the diffusion process, between the DSPE-PEG and other phospholipids, the formation of the micelle (10 ^-3 - 10 ^-6 s) occurs faster than the formation of the phospholipid bilayer (10 ^_1 - 10 ^-3 s ) (Evans and Wennerstrõm, 1999). In these circumstances, the diffusion processes employed in microfluidic systems for the production of LCs of the Stealth type are not considered efficient due to the possible formation of micelles and the consequent accumulation of aggregates along the channel, resulting in unfavorable physical-chemical properties.

[65] A crucial parameter for synthesis of Stealth type LC is to incorporate the maximum amount of PEG into the phospholipid bilayer (lamellar phase) before PEG is converted in micelles (micellar phase). Several factors play a role in the phase transition (from the lamellar to the micellar phase) of phospholipid-PEG conjugates. The main reason behind this phase transition has been shown to be the polymorphic property of phospholipids, which is the property of amphiphilic molecules that form aggregates of lipids under different conditions, such as water content, temperature and pH (Hristova et al. 1995 ). Another critical factor that can cause the formation of micelles is the chain length of the PEG / phospholipid conjugates, which is associated with the molecular weight (PM) of the PEG (Bedu-Addo et al., 1995). It has previously been shown that increasing the concentration of PEG to be incorporated into the surface of the bilayer can result in a micellar phase, which is more favorable than the bilayer phase in terms of energy (Hristova and Needham, 1995). In this investigation, PEG (PM = 2000) was used, but the diffusion-based microfluidic system (D-MD) was unable to produce micelle-free liposomes. As a possible solution to produce Stealth-type LC on D-MD without micelle formation, PEG with less molecular weight could be employed. However, a microfluidic solution was investigated by altering the mixing strategies to eliminate the formation of micelles and also alter the FRR. Therefore, we investigated the use of the chaotic advection-based microfluidic device (CA-MD) for conventional and stealth type LC synthesis.

Synthesis of conventional and stealth type LCs in a microfluidic device based on chaotic advection

[66] As the concomitant formation of micelles during the production of LCs of the Stealth type is a disadvantage for biological applications and the synthesis of LCs on diffusion-based microfluidic platforms requires an extended processing time (only 5-10 mL of LC production in one hour), the effect of chaotic advection on LC synthesis was investigated (Figure 4) . In this context, a microfluidic device based on chaotic advection (CA-MD) was produced, based on a reported device "Staggered Herringbone Micromixer" (Belliveau et al. 2012), with 2 and 3 entries to investigate the synthesis of conventional and like Stealth.

[67] Unlike diffusion-dominated mixing, recent studies have shown that the mixing process in non-turbulent flows can be significantly improved by changing the behavior of complex particles using chaotic advection, which is a state of the art invented from non-turbulent dynamics. linear flow (Aref, 1984; Aref, 2002; Vijayendran et al. 2003). In this study, the initial idea was to compare only the effect of the mixture (diffusion or chaotic advection), maintaining the FRR and the final lipid concentration (2.27 mM), as well as the amount of organic solvent (<10%) under conditions already established for the diffusion-based device (D-MD). Therefore, the same FRR (10), which was used for the D-MD, was used to operate the CA-MD with 3 inputs, varying the TFR from 100 to 1500 pL / min.

[68] In FRR 10, D-MD was suitable for conventional LC synthesis. Interestingly, in the same FRR, although there was no evidence of aggregation in the CA-MD (3 entries), the physical-chemical properties of conventional LCs worsened and the DLS technique detected an intense peak (close to the 10 nm region) possibly indicating aggregation disorganized, due to disturbance in diffusion patterns. In this case, the conventional LC produced in the CA-MD (TFR = 120.12 pL / min; FRR = 10) had a size close to ~ 110 nm, while the Pdl was greater than 0.5 and the zeta potential was significantly lower (~ 44 mV) than LCs produced in D-MD. For the production of LC-type Stealth in the CA-MD (3 inputs), no aggregation along the channel was observed; however, according to the DLS characterization, the physicochemical properties were similar to conventional LCs. Both in the conventional and in the Stealth type, the physico-chemical properties showed a large standard deviation and the presence of aggregates presenting sizes close to 10 nm. The results of DLS indicate the formation of disorganized structures that are not related to the conventional expected formation of LC, compromising biological applications and indicating low reproducibility of the process.

[69] As previously discussed, for the D-MD device, diffusion is predominant and strongly depends on the flow speed, which is the reason why there was a need to use low TFRs, to increase the residence time in the microchannels and allow the formation of an adequate concentration gradient along the microchannel. For CA-MD, since the mixing occurs by chaotic advection, the hypothesis was raised that this microfluidic device could work more efficiently at higher flow rates (TFR). Therefore, different TFRs were tested (up to 1500 pL / min) to find out if these results could be attributed to the flow rate. Unfortunately, DLS results have not changed significantly for each TFR tested. Stealth CLs (data not shown) were also produced using the same conditions, but the results were similar to those obtained with conventional CLs.

[70] To investigate the mixing effects of CA-MD (3 entries) on the physicochemical properties of conventional and Stealth LCs, changes were made to the system's FRR. The FRR was reduced from 10 to 1 (this means that the ethanol content at the end of the microfluidic device increased), maintaining the flow configuration (Figure S2-A), obtaining promising results in relation to the diameter, Pdl and zeta potential. Initially, the same TFR used on the D-MD (120.12 pL / min) was tested to confirm that the FRR had a significant effect on these properties. DLS results for conventional and Stealth LC have been significantly improved. In this TFR, the average diameter, Pdl and zeta potential of conventional LCs were 200 ± 14 nm, 0.120 ± 0.008 and 61.3 ± 2.7 mV, respectively. Comparing with conventional LCs produced in D-MD (FRR 10), the CLs produced by CA-MD were approximately 50 nm larger in size and 6 mV higher in zeta potential, while Pdl showed no significant difference. Compared with the properties of conventional LCs produced by CA-MD obtained in FRR 10, the average diameter and zeta potential of conventional LCs in FRR 1 increased significantly (from ~ 110 nm to ~ 200 nm and ~ 44 mV to 61 mV, respectively ). The most promising improvement was seen in Pdl. It may be possible to produce liposomes with Pdl as small as 0.03 in certain TFR, which makes these CLs injectable into the human body (Harashima et al. 1994; Woodle, 1995). [71] In the case of LCs of the Stealth type, the average diameter, the Pdl and the zeta potential were 168 ± 1 nm, 0.08 ± 0.03 and 57.0 ± 1.8 mV, respectively. The size distributions by intensity and volume of conventional and Stealth type LCs (Figure 5) indicated the presence of a typical population of liposomes and in the absence of other smaller peaks (which could indicate the formation of micelles). In addition, the zeta potential of Stealth-type LCs was relatively less than the zeta potential of conventional LCs produced in the same FRR and TFR. This behavior suggests the adequate formation of liposomes, and in the case of LC of the Stealth type, the DSPE-PEG conjugate (2000) was properly inserted into the structure of the liposome bilayer without any formation of micelles. Another interesting change was observed in the average diameter. Stealth-type LCs were approximately 30 nm smaller than the conventional ones, which can be caused by the insertion of PEG. Then, the TFRs were varied between 120.12 and 1250 pL / min; however, there was no significant change in the physicochemical properties of conventional and Stealth type LCs.

[72] In addition to the significant improvement in the physicochemical properties of conventional and Stealth-type LCs produced in FRR 1, the final lipid concentration of LCs also increased. The final concentration of LC lipids was 2.27 mM in LCs produced in D-MD and CA-MD with 3 inputs operating in FRR 10; however, in FRR 1, the concentration increased to 12.5 mM. In addition, CA-MD can operate the TFR of 1500 pL / min without any change in physicochemical properties for both lipid structures LCs and Stealth-LCs. In this way, the operation with FRR 1 increases mass productivity (g.mL ^-1 .min ^-1 ) about 70 times (5.5 times due to increased concentration "12.5 / 2.21" and 12.5 times due to increased TFR, 1500 / 120) compared to D-MD. However, although this decrease in FRR from 10 to 1 resulted in a higher final lipid concentration, this situation also caused a significant increase in the total amount of organic solvent (ethanol) in the final solution (from 9.09 to 50%), it is a concern for biological applications. It should be noted that for the analysis of DLS, all liposomes produced were diluted in water, reaching values equivalent to those used to measure CLs D-MD. This procedure was applied, since it is known that the increase in the concentration of ethanol can change the average diameter of the liposomes.

[73] In the investigation involving D-MD, the hypothesis was raised that FRR could significantly influence the formation of LC, since FRR 10 resulted in unfavorable physical-chemical properties and precipitation along the channel. In this way, a set of experiments was conducted, using the D-MD in which the FRR was decreased to 1 to confirm whether the FRR could improve the characteristics of the LCs. After DLS analysis, it was observed that micelle-like particles (less than 50 nm) do not appear in the size distribution graphs (data not shown). However, the mean diameter, Pdl, and the zeta potential of Stealth-type LCs were around 230 nm, 0.21 and 54 mV, respectively. In addition, the intensity and size distribution by volume confirmed the presence of peaks above 5000 nm, which could indicate an inadequate particle formation. Consequently, we conclude that CA-MD in FRR 1 was therefore more effective in synthesis of Stealth type LC compared to D-MD in FRR 1, therefore. Thus, investigations were continued using the CA-MD device.

[74] In general, the formation of nanoparticles is the result of the nanoprecipitation process, which occurs in contact between the two solvent streams and the aqueous phase. In the case of liposomes, the formation is mainly caused by the change in polarity due to the mixture caused by the geochannel of the microchannel, which includes the micropiscinas of the CA-MD. In addition, the surface area of the interface between the solvent and the aqueous phases increases significantly as the fluids are bent over each other, resulting in a chaotic mixture. The polarity can be easily increased along the microchannel, due to the intrinsic process of mixing ethanol and water, favoring the formation of liposomes. Thus, the alteration of the TFR as well as the FRR, which is directly associated with the ratio between aqueous phase and solvent, can have a great influence on the synthesis of liposomes. Changing the FRR changes the ethanol / water ratio and consequently leads to a change in the polarity of the final mixture, influencing the self-aggregation of lipids in bilayers. The rate of increase in polarity and its effect on the formation of liposomes were previously demonstrated (Bally et al ., 2012, Zhigaltsev et al., 2012). In addition, it is observed that conventional and Stealth-type LCs showed a larger mean size in FRR 1 (~ 170 nm) compared to FRR 10 (~ 120 nm). In FRR 10, the final concentration of ethanol was about 9.09%, resulting, therefore, in a reduction in the formation of larger liposomes due to the minimization of particle fusion and lipid exchange through the Ostwald ripening phenomenon. This phenomenon has been shown previously for the production of conventional LC using CA-MD (Kastner et al. 2014). Another explanation for the increase in size using CA-MD in FRR 1 is the presence of a higher ethanol content, as it is known that ethanol, as a solvent, interacts strongly with phospholipid bilayers at the lipid-water interface (composed mainly of polar head ) and not in the hydrocarbon interior. As a consequence of the incorporation of ethanol molecules in vesicle bilayers, liposomes can expand (Barry and Gawrisch, 1994).

[75] Another possible reason for the significant change in the physicochemical properties of both liposomes in different FRRs is the distance (W _f ) between 2 lateral aqueous phases and the central solvent phase. For FRR 10, W _f is extremely small and this indicates an intense contact between two currents until the particles enter the first subunits of the central channel similar to herringbone, which were placed at a distance of 4 mm from the crossing section of the entrances central and two sides. However, for FRR 1, W _f is considerably higher, since the proportion of ethanol is higher, and there is less contact between two flows before entering fishbone-like units. Until the beginning of the chaotic advection mixture, the diffusion along the distance of 4 mm can be considered predominant. Jahn et al. 2010 investigated the effect of FRR on diffusion-based microfluidic devices and as they increased FRR, the contact interface between the solvent and the aqueous phase became became closer and resulted in faster mixing and consequently smaller liposomes. The findings on the FRR effect are in agreement with previous studies (Kastner et al 2015; Zook and Vreeland, 2010; Jahn et al. 2010), however, chaotic advection as the predominant mixture is used in this study.

[76] As previously shown, the final lipid concentration of the LCs produced in the CA-MD (3 entries) in FRR 10 and FRR 1 was 2.27 and 12.5 mM, respectively. The increase in the concentration of phospholipids in the microfluidic system is another critical factor that can influence the adequate formation of liposomes in FRR 1. For diffusion-based microfluidic production (D-MD), this increase can impair the diffusion effectiveness, as it can interfere in the formation of several bilayer fragments and, consequently, form larger and polydispersed lipid vesicles. However, in microfluidic production based on chaotic advection, this increase did not negatively affect the formation of liposomes. The low FRR increases W _f at the entrance of the main channel (before chaotic advection begins), minimizing the contact between ethanol and the aqueous phase. Then, the chaotic advection mixture starts and the high percentage of ethanol can probably control the formation of the fragment of the bilayers, due to changes in polarity, favoring the formation of liposomes with adequate characteristics. However, the concentration of ethanol must be maintained up to a certain value, since the excessive amount of solvent can be incorporated into the bilayers of the vesicles and thus simply break them.

[77] Thus, in order to improve the system, the device microfluidic based on chaotic advection (CA-MS) was adapted to the same process conditions to generate the CA-MD with 2 inputs. The solvent: water ratio was kept identical to that obtained previously in 1: 1. Thus, different TFRs ranging from 100 to 1250 pL / min were tested. The particle size analysis by DLS showed results similar to those obtained with 3-input CA-MD, with the same FRR 1. For all TFRs tested, conventional LCs showed average diameter, Pdl and, zeta potential of 163 + 13 nm, 0 , 13 + 0.03 and 62 + 1 mM, respectively, while LCs of the Stealth type presented mean diameter, Pdl and, zeta potential of 153 + 5 nm, 0.11 + 0.01 and 60 + 2 mV, respectively. After t-test analysis, as expected, there was no statistical difference between conventional and Stealth type LCs, produced via CA-MD (3 entries) or CA-MD (2 entries). Although the zeta potential was slightly lower in LC of the Stealth type, the difference was not supported statistically. In addition, no peak was observed that could characterize structures corresponding to micelles (order of 10 - 30 nm), suggesting the adequate formation of liposomes. The particle diameter by both intensity and volume were similar to the LCs obtained with the CA-MD with 3 entries in the FRR 1. The production of conventional LCs in the CA-MD was previously demonstrated by Kastner et al. 2015; however, the liposomes produced were characterized only by the DLS technique, with no data associated with size distribution by intensity, volume or number. The conventional LCs presented average size, Pdl and zeta potential of approx. 180 nm, 0.2 and 60 mV, respectively, using the proportion of solvent / aqueous phase of 1. Unlike our study, they used the phospholipid formulation DOTAP / DOPE. According to the previous study (Koltover, 1998), the DOTAP / DOPE formulation does not allow the proper formation of lipid bilayers, as it can result in the coexistence of both the lamellar (L _a ) and inverted hexagonal (Hn) phases, which can cause liposome instability. This presence of L _a and Hn is relevant for pharmaceutical applications, since this variation leads to a different interaction in the cells, consequently, creating a considerable variation. In the same study by Koltover, 1998, the inclusion of EPC (50 mol%) resulted in the exclusive formation of lipid bilayers characterizing the formation of liposomes. Therefore, although the study by Kastner et al., 2015 states that liposomes are formed, there is no scientific evidence to support their claim. This certificate of addition reports that it is possible to produce LCs without the presence of the reverse hexagonal phase (Hn). In addition, these results were compared to the conventional diffusion-based process (D-MD). In addition, the technological solution now proposed shows the ability to produce LC of the Stealth type.

[78] Unlike conventional LC production, 1% of the DSPE-PEG conjugates (2000) was introduced into the system along with other phospholipids. According to the results of zeta potential, the inclusion of the lipid DSPE-PEG generated a decrease of ~ 5 mV, compared to conventional LCs, which may indicate the presence of PEG on the external surface of the LC.

[79] In most cases, the insertion of DSPE-PEG (2000) can significantly decrease the surface load of liposomes due to the Sstealth action (Frõhlich, 2012). However, this decrease also depends on the amount of PEG incorporated into the surface of the LCs, as well as the phospholipid formulation and the process used for incorporating PEG into liposomes. So far, the achievements of this Certificate of Addition have been conducted with 1 mol% of DSPE-PEG. To better understand the effect of the PEG polymer incorporated in the LCs, the experiments were repeated with the lipid DSPE-PEG at a concentration of 5%. First, the experiment was carried out using the CA-MD (3 entries) with FRR 10, and the results were similar to the LCs and LCs of the Stealth type with 1% DSPE-PEG). The results of the size distribution (DLS) showed different peaks below 10 nm, possibly belonging to micelle-like structures. As the FR-10 was reduced from 10 to 1 on the CA-MD (3 entries), the size results, again, were significantly improved. The mean size, Pdl and zeta potential of the SStealth type LCs (5% PEG) were 145.1 ± 1.5 nm, 0.11 ± 0.01 and 61.0 ± 1.8 mV, respectively. Statistical analysis showed no significant difference between LCs of the Stealth type produced in the same system with 1% PEG. However, when the CA-MD (2 inputs) was tested, LCs of the Stealth type had an average size of 144 ± 3 nm, Pdl of 0.12 ± 0.01 and zeta potential of 51.4 + 4.4 mV. Comparing with the CA-MD (3 inputs) in FRR 1, the size and the Pdl have not changed; however, a significant decrease in zeta potential was observed after the insertion of 5% PEG. The surface charge densities of conventional and Stealth LCs were 61.3 ± 2.7 mV and 62 ± 1 mM, respectively. With the inclusion of 5% of DSPE-PEG conjugates (2000), the surface charge density decreased to 51.4 ± 4.4 mV. This result shows that the increase in the amount of DSPE-PEG (2000) used to produce LCs of the Stealth type resulted in the incorporation of these molecules on the outer surface.

[80] Wolfram et al. (2014) incorporated PEG (1%) in their anionic liposomal formulation (PC, cholesterol, DPPEmPEG (2000) with a 6: 3: 1 molar ratio) and obtained a decrease of ~ 5 mV in the zeta potential value. Similar results were shown by Yang et al. (2007). They produced conventional liposomes using S100PC / CH (90:10 molar ratio) and the PEGylated liposome composed of S100PC / CH / MPEG (2000) -DSPE (90: 10: 5 molar ratio). Conventional liposomes had about zero surface charge, while PEG insertion (5%) reduced the potential by 20 mV. In addition, they observed that the liquid solution in which the liposomes are suspended also affects the surface charge of the liposomes. The zeta potential of PEGylated liposomes varied by ± 2-4 mV depending on the presence of the Tween 80 solution. Interestingly, the results were contrary in the study by Kim et al. (2010). Conventional LCs composed of DOTAP / DOPE / Chol / PEG in the proportion of 75: 20: 5: 0 had zeta potential of 15.4 + 7.5 mV, while the insertion of DSPE-PEG (2000) (70: 20: 5 : 5) increased the zeta potential to 25.3 ± 5.7 mV. In this formulation, DOTAP is the only phospholipid that provides cationic properties to LCs. Therefore, the 5% reduction in the concentration of DOTAP replaced by the DSPE-PEG conjugates (2000) should also decrease the surface charge density of the LCs, and with the proper insertion of PEG on the surface, the zeta potential total must have decreased further. However, the potential unexpectedly increased.

[81] As previously presented, the final percentage of ethanol present in the LC formulation was 50% in CA-MD (3 entries, FRR 1) and CA-MD (2 entries), which is considered a significant limitation for future biological applications. Most of the time, the removal of ethanol from the final lipid solution is conducted by the dialysis process. However, dialysis is a lengthy process (depending on the final volume of the solution, it can be up to days) and the work required to perform it successfully. Ethanol-type solvents are normally removed from the final solution via dialysis. However, for industrial productions, dialysis is not a viable process, as it requires a high volume of dialysis solution, as well as qualified researchers. Therefore, also in this Certificate of Addition, a new technological solution is proposed for the efficient removal of ethanol from the liposomal solution. For that, a vacuum centrifuge concentrator at low pressure and room temperature was used as an alternative distillation process. Briefly, the vacuum centrifuge concentrator uses a combination of centrifugal force, vacuum and heat to accelerate the evaporation rate of multiple small samples. However, this process can significantly change the structure or morphology and even biological activities of sensitive samples, such as pDNA and siRNA. Therefore, all physical-chemical, structural and morphological analyzes must be evaluated to control whether there have been any changes in these properties. Per for this reason, liposomal dispersions were produced and the effect of processing in the vacuum centrifuge concentrator to remove ethanol was evaluated. Thus, considering the evaporation rates of ethanol and residual water after ethanol removal, the percentage of ethanol remaining in the liposomal dispersion was insignificant and acceptable for further transfection studies. However, there were no significant changes in physical and chemical properties (Table 1).

Table 1: Comparison of the physical and chemical properties of liposomes before and after ethanol removal via vacuum centrifuge concentrator. Conventional LCs and Stealth LCs (1% DSPE-PEG (2000)) were obtained on CA-MD, FRR 1 with 2 inputs.

[82] It can be seen in Table 1 that the average size of conventional and Stealth LCs was almost similar shortly after production on the CA-MD inputs with 2 inputs. The difference in the average size after the ethanol removal process was not statistically significant. The most significant changes were observed in Pdl and zeta potential. Pdl increased from 0.1 to 0.24 for Stealth type LCs. Although this increase may be statistically significant in the case of LC of the Stealth type, it is still acceptable for transfection studies. The zeta potential of conventional LCs produced in this micro mixer is about 60 mV. The insertion of DSPE-PEG (2000) involves surface modification, therefore, decreases the zeta potential of colloidal systems. But we observed an additional decrease in the surface load of the LCs after using a vacuum centrifuge concentrator. This situation could be advantageous, since a very high potential can cause a toxic effect on animal cells (Frõhlich, 2012).

[83] Ethanol removal time and rate are essential factors that affect the evaporation rate of ethanol. As mentioned earlier, evaporation of ethanol occurs more quickly than that of water. Previous research (Balbino et al. 2013) has shown that a certain amount of ethanol residue (about 8%) in the final LC solution, and after our biological studies, there was no decrease in the transfection efficiency tested with the final formulation with a content of ethanol. Furthermore, the LC formulation did not show significant cytotoxicity to cells. As we continue to use the vacuum centrifuge concentrator, the LC solution begins to lose a significant amount of water and becomes more concentrated, which alters the properties of the liposomes. Therefore, it is essential to interrupt the vacuum centrifuge concentrator process in time to minimize the effect of the ethanol removal process on LCs. For this reason, an ethanol solution (100%) with 1 ml volume in an Eppendorf tube containing half the total volume of the LC samples (2 ml = 1 ml of water + 1 ml of ethanol) to be passed through the The ethanol removal process was placed in a vacuum centrifuge concentrator and, when 1 ml of ethanol solution was completely evaporated, the vacuum centrifuge concentrator was stopped and the samples were removed for further analysis. This simple optimization was considered sufficient to obtain LCs with minimal changes in their properties.

X-ray scattering analysis at low angles (SAXS)

[84] SAXS is used to study the structure of a variety of biological materials. It provides information about the shape, size and organization of colloids, liquid crystals, polymers and many other structures. Since liposomes are spherical vesicles that consist of phospholipid bilayers with amphiphilic properties, their arrangement in the presence of aqueous solution is important to form liposomes with a well-defined structure. This arrangement can vary significantly depending on the method of manufacture, as well as the type of phospholipids used in the formulation. The characterization of SAXS allows access, through modeling, to the electronic density through the membrane, providing information about the bilayer profile and lamellarity, among other structural parameters. The final structure of the liposome / DNA complexes will affect cell internalization and, consequently, the efficiency of transfection.

[85] Initially, the effect of the microfluidic mixer based on chaotic advection on the structure of cationic liposomes with / without DSPE-PEG (2000) was analyzed. No studies were found in the literature to explore the structural analysis of liposomes produced in this microdevice operated with chaotic advection. Figure 6A shows SAXS curves for LCs with and without PEGylated lipids produced in CA-MD microdevices with 2 and 3 inputs (2 different flow configurations) ". From fitted curves (not shown) using a model described in another study [Balbino et al. 2012 and Balbino et al. 2013], it is possible to infer that these systems are unilamellar (> 99%), and there is a small difference between their electronic density profile. The effect of the centrifuge concentrator was also evaluated vacuum, used to partially or completely remove ethanol from the final liposome formulation Figure 6B shows SAXS curves for the same systems after ethanol removal In the curves, it is possible to observe a reasonable difference in the curve profiles when compared to those obtained before ethanol removal. These differences are related to the ethanol leaving the bilayer, which changes the electronic density of the bilayer (bilayer form factor), as can be inferred from adjustments in curve (not shown). According to the modeling, it was also possible to verify that the lamellarity did not undergo significant changes (the systems maintained the multilamelity below 1%), showing that the systems are stable regarding this property even after the use of the vacuum centrifuge concentrator.

Morphological characterization of cationic liposomes

[86] Liposomal lamellarity (uni- or multi-) is of great importance in the efficiency of delivery of genetic material. In most applications, the unilamellar structure of LCs is highly recommended and has proved to be more efficient for delivering genetic material. Techniques such as DLS and SAXS can provide crucial information about this property of LCs through mathematical modeling. However, morphological analysis can provide additional information about the uni- or multilamellar characteristic of the produced liposomes. For this reason, the electron microscopy technique (Cryo-TEM) was used to characterize LCs and Stealth LCs.

[87] According to the images presented in Figure 7, all samples of LCs produced in the CA-MD (2 entries) proved to be unilamellar and supported the results obtained from DLS and SAXS. It was also observed that conventional and Stealth type LCs showed Pdl and low diameters between 150-200 nm, in agreement with the results obtained by DLS. Due to the small size of the PEG and its lesser influence on the physical and chemical properties of the liposomes, it was not possible to identify any morphological difference between conventional and Stealth type LCs. In addition, the insertion of DSPE-PEG into LC liposomes results in reduced rate of uptake by macrophages and prolonged time of circulation in the bloodstream. Consequently, it requires more time for a PEGylated LC to transfect cells, since PEG will delay the transfection process, decreasing the surface interaction between cells and LCs.

In vitro evaluation of the efficiency of lipoplex transfection

[88] The biological activity (transfection efficiency) of conventional and Stealth type LC formulations has been investigated in PC-3 cells. Prior to transfection studies, residual ethanol was removed to avoid toxic effects from conventional and Stealth-type LCs using the vacuum centrifuge concentrator. Conventional LCs showed a transfection level of approximately 20%, while Stealth-type LCs showed significantly lower transfection efficiency (~ 5%). The same pattern was observed for the fluorescence intensity of the GFP (Figure 8). The decrease in the level of transfection using LC of the Stealth type can be explained with the stealth effect of DSPE-PEG (2000) on the surface of the liposomes, preventing interaction with PC3 cells.

Conclusion

[89] Process productivity is an important issue in industrial production. This issue becomes more critical in the pharmaceutical industries, due to the complexity of the production system, as well as expensive materials that bring serious economic concerns. With the implementation of this Certificate of Addition it was possible to demonstrate that the high-yield microfluidic production of cationic liposomes, with a proportion of ethanol 50%, followed by ethanol removal operation in a vacuum concentrator in the formation allows the formation of cationic liposomes and also of the Stealth type (with up to 5% lipid-PEG). In the latter case, the presented process avoids the concomitant formation of micelles, which are undesirable for biological applications. According to the analyzes by Cryo- TE and SAXS, conventional and Stealth type LCs showed unilamellar structure and significant transfection efficiency in human prostate cell lines (PC-3). With a single microfluidic device (~ 4 cm x 7 cm), it would be possible to produce about 100 ml of LC in one hour. Considering 12 hours of continuous production, with a sufficient number of microfluidic devices that will occupy 1 m ² of area, it is possible to produce 400 L of LC, which is highly acceptable for industrial scale production. Due to the simplicity of the ethanol removal process, the same amount of LC can be processed using vacuum centrifuge concentrator adapted for industrial scale without changing any structural, morphological and biological properties of LCs. The proposed process is promising for the production of LC on an industrial scale, it is possible to provide safe, efficient and solvent-free non-viral vectors for gene delivery.

[90] Those skilled in the art will value the knowledge presented here and will be able to reproduce the invention in the modalities presented and in other variants, covered by the scope of the attached claims.

Claims

1. Microfluidic process for obtaining cationic liposomes and S tealth characterized by the fact that it comprises a microfluidic device based on chaotic advection and the steps of:

(i) insertion of 5 to 25 mM of egg phosphatidylcholine,

1.2-dioleoyl-sn-glycero-3-phosphatidylethanolamine (DOPE),

1.2-dioleyl-3-trimethylammonium propane (DOTAP) and 1,2-distearyl-sn-glycero-3-phosphoethanolamine with polyethylene glycol (DSPE-PEG) in the proportion ranging from 50: 20: 25: 5% to 50:24: 25: 1% EPC: DOPE: DOTAP: DSPE-PEG respectively, dispersed in 100% anhydrous ethanol in a microfluidic device, from a total flow rate ranging from 120 pL min ^-1 to 1500 pL min ^-1 , the resulting dispersion at the outlet of said microfluidic device having an ethanol content in the range of 33.5 to 51%, preferably 50%;

(ii) means for recovering the excess organic solvent and adjusting the lipid concentration, in which the said means comprise conventional distillation or large-scale distillation, and when it is performed by the latter the removal of the solvent occurs in a processing interval of 1 2 hours in a vacuum centrifuge concentrator.

2. Process, according to claim 1, characterized in that the said microfluidic device comprises 2 (two) or 3 (three) entries and is composed of two bases of poly dimethyl siloxane (PDMS) sealed with O2 plasma, in that one base contains microchannels and the other base contains pools, built by the technique of soft lithography employing photophoresis SU-8, preferably in silicon wafer, in which the above-mentioned pools have a fixed height along said microchannels that cause chaotic advection.

3. Process, according to claims 1 and 2, characterized by the fact that when the microfluidic device comprises 2 (two) inlets, the total flow of each inlet must be equal, resulting in individual flows from 60 to 1000 mΐ / min .

4. Process, according to claims 1 and 2, characterized by the fact that when the microfluidic device comprises 3 (three) inputs, the FRR (flow rate ratio) operating parameter should preferably have a value of 1 when the lipid dispersion and inserted in the central stream and aqueous phase on the sides, said aqueous phase comprising water or conventional buffer (such as PBS IX).

5. Process, according to claim 4, characterized by the fact that the arrangement can be changed as long as the preferred 1: 1 ratio of aqueous phase and lipid dispersion is maintained.

6. Process, according to claim 5, characterized by the fact that the lateral flows can vary from 30 to 500 mΐ / min and the central flow from 60 to 1000 mΐ / min, with the total production flow being able to operate 100 to 2000 pL / min.

Process according to any one of claims 1 to 6, characterized in that it allows the exclusive formation of the lamellar phase (La) (forming liposomes) and prevents the formation of the reverse hexagonal phase (HII) (another structure other than liposomes).

8. Cationic Stealth liposomes characterized by the fact that they are produced as defined in any of claims 1 to 7, and comprise from 50: 20: 25: 5% to 50: 24: 25: 1% EPC: DOPE: DOTAP: DSPE -PEG, preferably 50: 24: 25: 1%; and final lipid concentration of 25 to 10 mM, preferably 12.5 mM, wherein said concentration of the final lipid dispersion can be modulated.

9. Liposomes, according to claim 8, characterized by the fact that they have an average diameter ranging from 250 to 120 nm, preferably 180 nm; polydispersity index (PDI) ranging from 0.13 to 0.3, preferably 0.25; and zeta potential ranging from +45 to +55 mV, preferably +50 mV when produced in ultra pure water.

10. Use of cationic Stealth liposomes as defined in claims 8 or 9, characterized by the fact that they are applied in the agricultural, food, cosmetic, medical areas; in systems of sustained release of biopharmaceuticals, protein vectors, anti-cancer agents; and preferably in gene therapy and vaccination.