Microfluidic mixing devices for high-throughput lipid nanoparticle (LNP) preparation

Abstract

In one aspect, the disclosure relates to a microfluidic device, an apparatus comprising the same, and methods of use thereof for precise and high-throughput preparation of lipid nanoparticles (LNPs).

Description

[0001] Attorney Docket No. 046483-7491W01(04073)

[0002] TITLE OF THE INVENTION

[0003] Microfluidic Mixing Devices for High-Throughput Lipid Nanoparticle (LNP) Preparation

[0004] CROSS-REFERENCE TO RELATED APPLICATIONS

[0005] This application claims priority under 35 U.S.C. § 119(e) to U.S. Provisional Patent Application No. 63 / 717,568. filed November 7, 2024, which is incorporated herein by reference in its entirety.

[0006] STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0007] This invention was made with government support under TR002776 awarded by the National Institutes of Health. The government has certain rights in the invention.

[0008] REFERENCE TO AN ELECTRONIC SEQUENCE LISTING

[0009] The XML file named ”046483-7491 WO I - Sequence Listing.xml” created on November 6, 2025, comprising 2,050 bytes, is incorporated herein by reference in its entirety.

[0010] BACKGROUND

[0011] Bey ond their current clinical success as drug carriers in the SARS CoV-2 vaccines, lipid nanoparticles (LNPs) have demonstrated broad potential as delivery agents for gene editing cargo, immunotherapies, and many other clinically relevant applications. The behavior of an LNP in a biological setting is dictated by the structures of the individual lipid components of the LNP, their relative ratios, and how these structures are associated with a nucleic acid cargo. For example, the design of an LNP therapeutic which performs well for treating cystic fibrosis might entail an LNP which is compositionally totally distinct from an LNP which best transfects monocytes for a solid tumor therapy.

[0012] In the field of small molecule drug discovery', high-throughput screens are performed to identify which compounds in a library’ of compounds have some basic capacity’ to bind to a certain therapeutic “target,” usually a protein of interest. From there, a scientist might be able to determine how to modify a given compound to achieve more favorable “drug-like properties” and even to generate a structure-activity’ relationship (SAR), where inferences can be made from structural modifications to determine how they affect biological activity. Although advancements have been made in understanding how lipid structures affect particle structure and performance, there still exists a minimal understanding of the SAR of any given Attorney Docket No. 046483-7491W01(04073)

[0013] LNP. In other words, an LNP scientist cannot necessarily reasonably predict how altering either the structure of a set of lipids or their relative ratios would affect the biological behavior of an LNP formulated from those components.

[0014] Thus, the primary approach taken for determining optimal LNPs for any given function is to formulate and screen these LNPs in in vitro and in vivo models appropriate to the intended biological use case. However, the dimensionality of the LNP design space represents a significant challenge for researchers to overcome. The structural combinations in the major lipid components of an LNP (i.e., the ionizable lipid, helper or neutral lipid, steroid, and polymer conjugated lipid) easily exceed 108possibilities, over 105structures of which come from the ionizable lipid alone. Beyond structural variability, the relative ratios of these components in a given LNP formulation play a similarly significant role in defining the efficacy of that LNP for a particular biologic function, conservatively expanding the potential design space by a factor of 102dimensions. Thus, for any one biological function, determining an optimally performing LNP is akin to scanning for local minima in a geometry embedded in IO10dimensional space.

[0015] Considering the generally prescribed rule for acquisition of around 10 observations per dimension to perform statistically meaningful regression analysis, this would require 1011observations to properly assay this space. Although it is exceedingly likely that a large portion of these dimensions represent LNPs which do not perform meaningfully, which reduces the number of “intrinsic dimensions"’ meaningfully, determining an optimal LNP for a given function in vivo would still likely require screening of hundreds if not thousands of LNPs.

[0016] While techniques from combinatorial chemistry have been successfully applied to nanoparticle design and LNP barcoding allows for pooled screening of nanoparticles, there is not yet a way to controllably mix the different components of an LNP at a comparable throughput. Liquid handlers have been used to increase the throughput of LNP mixing, but from the perspective of consistency at the length scale of particle nucleation, pipette mixing intrinsically results in turbulent and heterogeneous mixing within each well, resulting in the potential evolution of different particles from batch to batch. Additionally, there is no pathway to scaling the production of particles beyond the discovery phase, as even the most advanced liquid handlers can generally only dispense up to 5 mL per sample.

[0017] There is thus a need in the art for microfluidic mixing devices and methods of use thereof which enable precise high-throughput preparation of LNPs. The present disclosure addresses this unmet need. Attorney Docket No. 046483-7491W01(04073)

[0018] BRIEF SUMMARY OF THE INVENTION

[0019] In one aspect, the disclosure provides a microfluidic device comprising a parallel array of microfluidic channels. In certain embodiments, each microfluidic channel independently comprises m occurrences of a first solution inlet, wherein each first solution inlet is in fluid communication with a first solution channel, and wherein m is 1, 2, 3, 4, 5, 6. 7, 8, 9, or 10. In certain embodiments, each microfluidic channel independently comprises n occurrences of a second solution inlet, wherein each second solution inlet is in fluid communication with a second solution channel, and wherein n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In certain embodiments, each microfluidic channel independently comprises a first combined solution channel which is in fluid communication with each of the first solution channels. In certain embodiments, each microfluidic channel independently comprises a second combined solution channel, which is in fluid communication with each of the second solution channels. In certain embodiments, each microfluidic channel independently comprises a third combined solution channel, which is in fluid communication with the first combined solution channel and the second combined solution channel. In certain embodiments, each microfluidic channel independently comprises an outlet which is in fluid communication with the third combined solution channel. In certain embodiments, each combined solution channel independently comprises at least one chaotic mixing region. In certain embodiments, each first solution channel comprises at least one chaotic mixing region. In certain embodiments, each second solution channel comprises at least one chaotic mixing region.

[0020] In another aspect, the disclosure provides a microfluidic processing apparatus. In certain embodiments, the microfluidic processing apparatus comprises the microfluidic device of the disclosure. In certain embodiments, the microfluidic processing apparatus comprises a collection plate comprising a plurality of wells. In certain embodiments, the microfluidic processing apparatus comprises a manipulator configured to hold the microfluidic device above the collection plate. In certain embodiments, the microfluidic processing apparatus comprises a motion control apparatus operatively coupled to the manipulator and configured to position the microfluidic device relative to the collection plate such that each outlet of the device is aligned with a well of the collection plate. In certain embodiments, the microfluidic processing apparatus comprises a controller configured to actuate the motion control apparatus to reposition the microfluidic device from a first configuration, in which an outlet of the device is aligned with a first well of the collection plate, to an alternate configuration, in which the outlet is aligned with a different well of the Attorney Docket No. 046483-7491W01(04073) collection plate. In certain embodiments, the microfluidic processing apparatus comprises a fluid control system operatively coupled to the microfluidic device and configured to regulate pressure applied to each inlet of the microfluidic device.

[0021] In another aspect, the disclosure provides a method for preparing a plurality of lipid nanoparticles (LNPs) using the microfluidic processing apparatus of the disclosure. In certain embodiments, the method comprises providing the microfluidic processing apparatus of the disclosure. In certain embodiments, the method comprises actuating the motion control apparatus to position the microfluidic device such than each outlet of the device is aligned with an empty well of the collection plate. In certain embodiments, the method comprises introducing a first solution into each first solution inlet of the microfluidic device to provide a combined first solution. In certain embodiments, the method comprises introducing a second solution into each second solution inlet of the microfluidic device to provide a combined second solution. In certain embodiments, the method comprises mixing the combined first solution and combined second solution to provide a formulation comprising a plurality of LNPs. In certain embodiments, the method comprises dispensing at least a portion of the formulation comprising the plurality of LNPs from each outlet to each respective well of the collection plate. In certain embodiments, the method comprises repeating steps (c)-(f) (z. e. , the actuating, introducing, mixing, and dispensing steps) after changing at least one parameter of the method. In certain embodiments, a parameter of the method comprises a flow rate through the microfluidic device. In certain embodiments, a parameter of the method comprises an identity of at least one component of at least one of the first solution or the second solution. For example, in certain non-limiting embodiments, modifying the identity7of a component of the at least one first solution comprises substituting ionizable lipid SM-102 for D-Lin-MC3-DMA. In certain embodiments, a parameter of the method comprises a concentration of at least one component of at least one of the first solution or the second solution.

[0022] BRIEF DESCRIPTION OF THE FIGURES

[0023] The drawings illustrate generally, by way of example, but not by way of limitation, various embodiments of the present application.

[0024] FIGs. 1A-1C: FIG. 1A: Overview of the iterative generation and screening of LNP libraries using LIBRIS. FIG. IB: Comparison of the number of unique formulations per hour possible using LIBRIS and other commercially available devices for microfluidic generation of LNPs. LIBRIS could be scaled beyond 103formulations by increasing the number of Attorney Docket No. 046483-7491W01(04073) mixing units using very large scale integration (VLSI). FIG. 1 A: Depiction of the scale invariance of parallelized microfluidic mixing, demonstrated with a parallelized chip comprising many of the same mixing units and resistors (SCALAR), as compared with other mixing techniques for LNP formulation.

[0025] FIGs. 2A-2F: Design and fabrication of the LIBRIS chip. FIG. 2A: Schematic of LIBRIS chip, depicting fluid entering one of six inputs (ii-6) and being distributed differentially through resistors Rt.m, where i indexes the fluidic input and m indexes the LNP generator. The six inputs combine in different combinations in each LNP generator (mis) to generate distinct formulations FI-F8. FIG. 2B: Labeled image of a single mixing unit. FIGs. 2C-2D: Circuit diagrams for the resistor network for a single (FIG. 2C) device and (FIG. 2D) fluidic input. FIG. 2E: The LIBRIS chip varies pressures Pt.n(top middle) to deliver a set of lipid structures at a given set of concentrations (top left) through a set of lithographically defined resistors (bottom), producing a library' of LNPs of the size 8n where n is the number of pressure settings (right). FIG. 2F : Though relative flow rates were varied, the total flow rate through the ethanol and aqueous mixers of each LNP generator remained the same across all pressures tested.

[0026] FIGs. 3A-3H: Characterization of mixing efficiency and relative flow rates across LIBRIS chip generators. FIG. 3A: Image of the fabricated LIBRIS chip. FIG. 3B: Fluorescent dyes were imaged flowing through the device, zi is schematized as containing the IL and cholesterol, while iz contains phospholipid and PEG lipid. FIGs. 3C-3E: Mixing efficiencies were quantitated (±SD; errors determined by fitting 1 - H to an exponential decay model) and compared using a one-way ANOVA followed by Tukey's HSD post-hoc test across the (FIG. 3C) ethanol (F= 0.001, P = 1.0), (FIG. 3D) aqueous (F= 0.006, P = 1.0), and (FIG. 3E) combined mixing regions of the chip (F= 0.004, P = 1.0). FIGs. 3F-3G: The assumed flow rate ratios for each mixing device (FIG. 3F) were compared using (FIG. 3G) pairwise distance analysis to the actual flow rate ratios. FIG. 3H: Correlation between pairwise distance matrices was compared using a Mantel test (R2= 0.96, P = 0.001).

[0027] FIGs. 4A-4G: Schematic overview, timing diagram, and collection consistency of custom plate robotic interface. FIG. 4A: Overview of a single collection run, where a user initiates the setup and the microcontroller synchronizes plate motion with pressure regulation. FIG. 4B: The life cycle of a single set of eight LNPs generated and collected by the plate robotic interface. FIGs. 4C-4D: After initiating, (FIG. 4C) a timing diagram shows the breakdown of the 2.6s waste cycle for a set of eight LNPs, (FIG. 4D) where a total of 90s is required to collect 32 LNPs. FIG. 4E: A labelled set of images of the plate robot. FIGs. 4F- Attorney Docket No. 046483-7491W01(04073)

[0028] 4G: Measurement of the precision of collection in (FIG. 4F) n = 20 iterations of movement on the same plate and (FIG. 4G) n = 20 iterations of movement across three separate plates; (FIG. 4F, left) Measurements of puncture marks across four collection positions for the same needle position were compared using a one-way ANOVA followed by a Tukey HSD post- hoc test (F = 1.5, P = 0. 19); (FIG. 4F, right) A one-sample t-test was performed to determine whether the radius of puncture was below the acceptable radius of 1mm (t = -572.4, **** = P « 0.0001). FIG. 4G: The variance of the puncture marks at the same positions across three plates (left) were compared similarly (F= 1.4, P = 0.27) and absolute position for each needle within the cassette was compared across plates (right, F= 2.2, P = 0.14).

[0029] FIGs. 5A-5D: Physicochemical characterization of the LNP libraries generated by the LIBRIS chip. FIG. 5A: Schematic of the potential design space assayed in this library. Across each lipid set, IL and cholesterol mole percent ratios vary directly proportionally with one another, which vary inversely with the phospholipid and PEG lipid ratios. A decrease in LNP size is shown with increased PEG concentration, schematized in the top right of the subpanel. FIG. 5B: DLS measurements of the size of each formulation; Size decays exponentially with the molar percentage of PEG (MC3: P < 0.0001, SM-102: P < 0.0001, C12-200: P < 0.0001). FIGs. 5C-5D: 95 / 96 particles and 93 / 96 particles fall within physicochemical benchmarks of encapsulation efficiency (EE%) (FIG. 5C) and polydispersity7index (PDI) (FIG. 5D).

[0030] FIGs. 6A-6C: In vitro and in vivo validation of differences between LNPs in libraries shown by physicochemical characterization. FIG. 6A: Heat map of the % of mCherry+ HepG2 cells across all 96 LNPs screened. The lowest and highest transfecting 02-200 LNP moved forward from the in vitro screen into in vivo testing. FIG. 6B: Representative IVIS images of the major organs extracted from one member of each cohort 6hr after LNP administration showing luciferase expression in the liver (H, heart; Lu, lungs: Li. liver; Sp, spleen; Ki, kidneys). FIG. 6C: Quantification and analysis of hepatic luciferase expression between each treatment group. Normality of each sample distribution was validated by a Shapiro-Wilk test (PBS: P = 0.5881, F4: P = 0.5913, F24: P = 0.3048, BF: P = 0.6586). Oneway ANOVA followed by Tukey's HSD test was used to compare total luminescence between all groups (*P < 0.05, **P < 0.01). n = 4 mice in PBS group, n = 6 in treatment groups.

[0031] DETAILED DESCRIPTION OF THE INVENTION

[0032] Reference will now be made in detail to certain embodiments of the disclosed subject matter, examples of which are illustrated in part in the accompanying drawings. While the Attorney Docket No. 046483-7491W01(04073) disclosed subject matter will be described in conjunction with the enumerated claims, it will be understood that the exemplified subject matter is not intended to limit the claims to the disclosed subject matter.

[0033] Throughout this document, values expressed in a range format should be interpreted in a flexible manner to include not only the numerical values explicitly recited as the limits of the range, but also to include all the individual numerical values or sub-ranges encompassed within that range as if each numerical value and sub-range is explicitly recited. For example, a range of "about 0.1% to about 5%" or "about 0.1% to 5%" should be interpreted to include not just about 0.1% to about 5%, but also the individual values (e.g., 1%, 2%, 3%, and 4%) and the sub-ranges (e.g., 0.1% to 0.5%, 1.1% to 2.2%, 3.3% to 4.4%) within the indicated range. The statement "about X to Y" has the same meaning as "about X to about Y," unless indicated otherwise. Likewise, the statement "about X, Y, or about Z" has the same meaning as "about X, about Y, or about Z," unless indicated otherwise.

[0034] In this document, the terms "a," "an," or "the" are used to include one or more than one unless the context clearly dictates otherwise. The term "or" is used to refer to a nonexclusive "or" unless otherwise indicated. The statement "at least one of A and B" or "at least one of A or B" has the same meaning as "A, B, or A and B." In addition, it is to be understood that the phraseology or terminology7employed herein, and not otherwise defined, is for the purpose of description only and not of limitation. Any use of section headings is intended to aid reading of the document and is not to be interpreted as limiting: information that is relevant to a section heading may occur within or outside of that particular section. All publications, patents, and patent documents referred to in this document are incorporated by reference herein in their entirety, as though individually incorporated by reference.

[0035] In the methods described herein, the acts can be carried out in any order, except when a temporal or operational sequence is explicitly recited. Furthermore, specified acts can be carried out concurrently unless explicit claim language recites that they be carried out separately. For example, a claimed act of doing X and a claimed act of doing Y can be conducted simultaneously within a single operation, and the resulting process will fall within the literal scope of the claimed process.

[0036] Description

[0037] Lipid nanoparticles (LNPs) have demonstrated broad potential for the delivery of nucleic acid cargos for gene therapy, immunotherapy, and vaccine applications. The lipid components of an LNP formulation, namely the ionizable lipid (IL), cholesterol, Attorney Docket No. 046483-7491W01(04073) phospholipid, and poly (ethylene glycol) (PEG)-conjugated lipid, is generally altered for each clinical application, considering desired tissue or cell tropism, immunogenicity, and other therapeutic-specific factors. Though each new clinical application requires a distinct LNP formulation, there do not yet exist sufficient design principles that relate the components of an LNP and its generation method to its structure and performance in vivo, necessitating the screening of large libraries of formulations in biological models.

[0038] The high dimensionality of the space of possible LNP formulations has motivated researchers to develop new approaches to generate and screen LNP formulations at high throughput. Combinatorial chemistry has been used to rapidly generate diverse lipids in parallel; as a result, there now exist over 105ionizable lipids alone. LNP barcoding can provide biodistribution data from the pooled screening of tens to hundreds of formulations in a single animal. However, no LNP formulation method thus far can match the throughput of these techniques. Robotic liquid handlers can formulate LNPs at the theoretical throughput of tens to hundreds of LNP formulations per hour but rely on turbulent pipette mixing which results in batch-to-batch v ariability and particle heterogeneity. Moreover, liquid handlers are typically limited to a maximum volume of 1 mL per sample, requiring that a different LNP generation method must be used to formulate LNPs beyond the scale of discovery, confounding the translation of LNPs discovered using this method to animal testing and to clinical evaluation.

[0039] Microfluidic mixing solves the issue of heterogeneous particle formulation by rapidly mixing particle components at the micrometer scale, resulting in particles with consistent physicochemical and biological properties. Using recently developed microfabricated parallelized architectures such as the silicon scalable lipid nanoparticle (generation) SCALAR chip, microfluidic mixing can also scale over three orders of magnitude of production scale, from the discovery phase (~10 mL / hr) to the production phase (~10 L / hr) without affecting LNP properties. While formulating LNPs using microfluidics has been automated to some degree, the process requires significant manual inter ention and is generally serial, producing only ~10 distinct formulations per hour.

[0040] In one aspect, the disclosure relates to methods useful to accelerate the rate of microfluidic LNP discovery, which provides in part, a platform for formulating distinct LNPs at a rate of 300 pL of one formulation every' three seconds, at least lOOx faster than conventional microfluidic approaches (FIGs. 1A-1C). This platform is referred to herein as Lipid nanoparticle Batch pro- duction via Robotically Integrated Screening (LIBRIS). The LIBRIS was designed using a parallelized microfluidic device architecture wherein multiple Attorney Docket No. 046483-7491W01(04073) fluidic inputs each bearing an LNP component are delivered to individual LNP generators in defined proportions determined by differentially encoded fluidic resistors and dynamically controlled pressure-driven flow. LIBRIS integrates a custom-designed and microfabricated silicon-glass chip with a microcontroller-operated plate robot and pressure manifold, enabling auto- mated control and direct collection of distinct outputs into dialysis cassettes for postnucleation processing. To evaluate the utility of this platform, a library of 96 distinct LNP formulations was generated, comprising three sets of 32 formulations each made from a different set of lipids. Three sets of 32 LNPs were generated in 90s each, and their physicochemical properties and transfection efficiencies were characterized, and in vivo performance of a subset of the library was evaluated.

[0041] Definitions

[0042] The term "about" as used herein can allow for a degree of variability in a value or range, for example, within 10%, within 5%, or within 1% of a stated value or of a stated limit of a range, and includes the exact stated value or range.

[0043] The term “anionic lipid” refers to any lipid that is negatively charged at physiological pH. These lipids include phosphatidylglycerol, cardiolipin, diacylphosphatidylserine, diacylphosphatidic acid, N-dodecanoylphosphatidylethanolamines, N- succinylphosphatidylethanolamines, N-glutarylphosphatidylethanolamines, lysylphosphatidylglycerols, palmitoyloleyolphosphatidylglycerol (POPG), and other anionic modifying groups joined to neutral lipids.

[0044] The term “array” as used herein refers to a two-dimensional arrangement of features, such as an arrangement of reservoirs, on a planar and / or flat surface.

[0045] The term “cationic lipid” refers to any of a number of lipid species that carry a net positive charge at a selected pH, such as physiological pH (e.g, pH of about 7.0). It has been found that cationic lipids comprising alkyl chains with multiple sites of unsaturation, e.g., at least two or three sites of unsaturation, are particularly useful for forming lipid particles with increased membrane fluidity. A number of cationic lipids and related analogs, which are also useful in the present disclosure, have been described in U.S. Patent Publication Nos. 20060083780 and 20060240554; U.S. Patent Nos. 5,208,036; 5,264,618; 5,279,833; 5,283,185; 5,753,613; and 5,785,992; and PCT Publication No. WO 96 / 10390, the disclosures of which are herein incorporated by reference in their entirety for all purposes. Non-limiting examples of cationic lipids are described in detail herein. In some cases, the cat-ionic lipids comprise a protonatable tertiary amine (e.g, pH titratable) head group, Cis Attorney Docket No. 046483-7491W01(04073) alkyl chains, ether linkages between the head group and alkyl chains, and 0 to 3 double bonds. Such lipids include, e.g., DSDMA. DLinDMA, DLenDMA. and DODMA.

[0046] The term '‘channel” as used herein refers to any type of conduit that defines a path for fluid to flow from one point to another point (e.g., elongated well or tube).

[0047] The term “chaotic mixing” as used herein refers to a process for enhancing fluid mixing which comprises creating a chaotic flow pattern for a liquid in motion. Unlike traditional mixing, which relies on turbulence or mechanical stirring, chaotic mixing uses precisely engineered geometries or flow conditions to stretch and fold fluid layers repeatedly.

[0048] The term “fluid communication” as used herein refers to a connection between elements which permits a fluid at least partially contained in a first element to flow to the second element.

[0049] The term '‘helper lipid” as used herein refers to a lipid capable of increasing the effectiveness of delivery of lipid-based particles such as cationic lipid-based particles to a target, preferably into a cell. The helper lipid can be neutral, positively charged, or negatively charged. In certain embodiments, the helper lipid is neutral or negatively charged. Nonlimiting examples of helper lipids include 1.2-distearoyl-sn-glycero-3-phosphocholine (DSPC), I,2-di-(9Z-octadecenoyl)-sn-glycero-3-phosphoethanolamine (DOPE), I-palmitoyl- 2-oleoyl-sn-glycero-3phosphocholin (POPC) and l,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC).

[0050] The term “independently selected from” as used herein refers to referenced groups being the same, different, or a mixture thereof, unless the context clearly indicates otherwise. Thus, under this definition, the phrase “X1, X2, and X3are independently selected from noble gases” would include the scenario where, for example, X1, X2, and X3are all the same, where X1. X2. and X3are all different, where X1and X2are the same but X3is different, and other analogous permutations.

[0051] The term “inlet” as used herein refers to a portion of a component through which a fluid can be drawn into a component, conduit, or fluid passageway.

[0052] The term “ionizable lipid” as used herein refers to a lipid (e.g., a cationic lipid) or lipidoid having at least one protonatable or deprotonatable group, such that the lipid is positively charged at a pH at or below physiological pH (e g., pH 7.4), and neutral at a second pH, preferably at or above physiological pH. It will be understood by one of ordinary' skill in the art that the addition or removal of protons as a function of pH is an equilibrium process, and that the reference to a charged or neutral lipid refers to the nature of the predominant species and does not require that all of the lipid be present in the charged or neutral form. Attorney Docket No. 046483-7491W01(04073)

[0053] Generally, ionizable lipids have a pKa of the protonatable group in the range of about 4 to about 7.

[0054] The term '‘lipid” refers to a group of organic compounds that include, but are not limited to, esters of fatty acids and are characterized by being insoluble in water, but soluble in many organic solvents. They are usually divided into at least three classes: (1) “simple lipids,” which include fats and oils as well as waxes; (2) “compound lipids.” which include phospholipids and glycolipids; and (3) “derived lipids” such as steroids.

[0055] The term “lipid particle” is used herein to refer to a lipid formulation that can be used to deliver an active agent or therapeutic agent, such as a nucleic acid (e.g. , mRNA), to a target site of interest. In the lipid particle of the disclosure, which is ty pically formed from a cationic lipid, a non-cationic lipid, and a conjugated lipid that prevents aggregation of the particle, the active agent or therapeutic agent may be encapsulated in the lipid, thereby protecting the agent from enzy matic degradation.

[0056] The term “microcontroller"’ as used herein refers to a CPU with additional function or structure, such as RAM, ROM, and or peripherals like I / O all embedded on a single chip, which can manipulate aspects of a device in response to electronic input(s) of a user.

[0057] The term “mRNA” or “messenger RNA” as used herein refers to a ribonucleic acid sequences which encodes a peptide or protein. In certain embodiments, the mRNA may comprise a “transcript"’ that is produced by using a DNA template and encodes a peptide or protein. Typically, mRNA comprises 5 -UTR, protein coding region and 3’-UTR. mRNA can be produced by in vitro transcription from a DNA template. Methods of in vitro transcription are known to those of skill in the art. For example, various in vitro transfer kits are commercially available. According to the present invention, mRNA can be modified by further stabilizing modifications and cap formation in addition to the modifications according to the invention.

[0058] The term “neutral lipid” or “helper lipid” refers to any of a number of lipid species that exist either in an uncharged or neutral zwitterionic form at a selected pH. At physiological pH, such lipids include, for example, diacylphosphatidylcholine, diacylphosphatidylethanolamine, ceramide, sphingomyelin, cephalin, cholesterol, cerebrosides, and diacylglycerols.

[0059] The term “non-cationic lipid” refers to any amphipathic lipid as well as any other neutral lipid or anionic lipid.

[0060] The term “nucleic acid” as used herein refers to a polymer containing at least two deoxyribonucleotides or ribonucleotides in either single- or double-stranded form and Attorney Docket No. 046483-7491W01(04073) includes DNA and RNA. DNA may be in the form of, e.g., antisense molecules, plasmid DNA, pre-condensed DNA, a PCR product, vectors (Pl. PAC, BAC, YAC. artificial chromosomes), expression cassettes, chimeric sequences, chromosomal DNA, or derivatives and combinations of these groups. RNA may be in the form of siRNA, asymmetrical interfering RNA (aiRNA), microRNA (miRNA), mRNA, tRNA, rRNA, tRNA, viral RNA (vRNA), and combinations thereof. Nucleic acids include nucleic acids containing known nucleotide analogs or modified backbone residues or linkages, which are synthetic, naturally occurring, and non-naturally occurring, and which have similar binding properties as the reference nucleic acid. Examples of such analogs include, without limitation, phosphorothioates, phosphoramidates, methyl phosphonates, chiral-methyl phosphonates, 2’- O-methyl ribonucleotides, and peptide-nucleic acids (PNAs). Unless specifically limited, the term encompasses nucleic acids containing known analogues of natural nucleotides that have similar binding properties as the reference nucleic acid. Unless otherwise indicated, a particular nucleic acid sequence also implicitly encompasses conservatively modified variants thereof (e.g., degenerate codon substitutions), alleles, orthologs, SNPs, and complementary sequences as well as the sequence explicitly indicated. Specifically, degenerate codon substitutions may be achieved by generating sequences in which the third position of one or more selected (or all) codons is substituted with mixed-base and / or deoxyinosine residues (Batzer et al., Nucleic Acid Res., 19:5081 (1991); Ohtsuka et al., J. Biol. Chem.. 260:2605-2608 (1985); Rossolini et al., Mai. Cell. Probes, 8:91-98 (1994)).

[0061] "Nucleotides" contain a sugar deoxyribose (DNA) or ribose (RNA), a base, and a phosphate group. Nucleotides are linked together through the phosphate groups. “Bases’’ include purines and pyrimidines, which further include natural compounds adenine, thymine, guanine, cytosine, uracil, inosine, and natural analogs, and synthetic derivatives of purines and pyrimidines, which include, but are not limited to, modifications which place new reactive groups such as, but not limited to, amines, alcohols, thiols, carboxylates, and alkyl halides.

[0062] Unless otherwise indicated, a particular nucleic acid sequence also implicitly encompasses conservatively modified variants thereof (e.g., degenerate codon substitutions), alleles, orthologs, SNPs, and complementary sequences as well as the sequence explicitly indicated. Specifically, degenerate codon substitutions may be achieved by generating sequences in which the third position of one or more selected (or all) codons is substituted with mixed-base and / or deoxyinosine residues (Batzer et al., Nucleic Acid Res., 19:5081 (1991); Ohtsuka et al. , J. Biol. Chem., 260:2605-2608 (1985); Rossolini et al., Mol. Cell. Attorney Docket No. 046483-7491W01(04073)

[0063] Probes, 8:91-98 (1994)).

[0064] The term “outlet’7as used herein refers to a portion of a component through which a fluid can be expelled from a component, conduit, or fluid passageway.

[0065] The term “pneumatic” as used herein refers to the use of any type of pressurized fluid and is not limited to compressed air.

[0066] The terms “polymer-conjugated lipid” and “conjugated lipid” are used interchangeably herein to refer to a lipid which is conjugated to one or more polymeric groups, which inhibits aggregation of lipid particles. Such lipid conjugates include, but are not limited to, polyamide oligomers (e.g., ATTA-lipid conjugates), PEG-lipid conjugates, such as PEG coupled to dialkyloxypropyls, PEG coupled to diacylglycerols, PEG coupled to cholesterol, PEG coupled to phosphatidylethanolamines, PEG conjugated to ceramides (e.g.. U.S. Pat. No. 5,885,613, the disclosure of which is herein incorporated by reference in its entirety for all purposes), cationic PEG lipids, and mixtures thereof. PEG can be conjugated directly to the lipid or may be linked to the lipid via a linker moiety. Any linker moiety suitable for coupling the PEG to a lipid can be used including, e.g., non-ester containing linker moieties and ester-containing linker moieties. In preferred embodiments, non-ester containing linker moieties are used.

[0067] The term “resistor” as used herein refers to a channel with defined fluidic resistance, determined by channel dimensions and fluid viscosity, inter alia, that regulates the flow rate at a given pressure. Analogous to electrical resistors, fluidic resistors follow the relation P=QRP=QR, where PP is pressure, QQ is flow rate, and RR is resistance. These resistors are precisely fabricated, often to nanometer-scale, enabling controlled mixing and flow modulation in microfluidic systems.

[0068] The term “substantially” as used herein refers to a majority of, or mostly, as in at least about 50%, 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.9%, 99.99%, or at least about 99.999% or more, or 100%. The term “substantially free of’ as used herein can mean having none or having a trivial amount of, such that the amount of material present does not affect the material properties of the composition including the material, such that the composition is about 0 wt% to about 5 wt% of the material, or about 0 wt% to about 1 wt%. or about 5 wt% or less, or less than, equal to, or greater than about 4.5 wt%, 4, 3.5, 3, 2.5, 2, 1.5, 1, 0.9, 0.8, 0.7, 0.6, 0.5, 0.4, 0.3, 0.2, 0. 1, 0.01, or about 0.001 wt% or less. The term “substantially free of’ can mean having a trivial amount of, such that a composition is about 0 wt% to about 5 wt% of the material, or about 0 wt% to about 1 wt%, or about 5 wt% or less, or less than, equal to, or greater than about 4.5 wt%, 4, 3.5, 3, 2.5, 2, 1.5, 1, 0.9, 0.8, 0.7, Attorney Docket No. 046483-7491W01(04073)

[0069] 0.6, 0.5, 0.4, 0.3, 0.2, 0.1, 0.01, or about 0.001 wt% or less, or about 0 wt%. The term “does not substantially;’ when used in conjunction with a process or event indicates that the process or event does not occur or occurs in a trivial amount, such that the process or event occurs about 0% or less than about 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, or about 5.0% of the maximal amount that the process or event could occur.

[0070] Microfluidic Devices

[0071] In one aspect, the disclosure provides a microfluidic device comprising a parallel array of microfluidic channels. In certain embodiments, each microfluidic channel independently comprises m occurrences of a first solution inlet, wherein each first solution inlet is in fluid communication with a first solution channel, and wherein m is 1, 2. 3, 4, 5, 6. 7, 8, 9, or 10. In certain embodiments, each microfluidic channel independently comprises n occurrences of a second solution inlet, wherein each second solution inlet is in fluid communication with a second solution channel, and wherein n is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10. In certain embodiments, each microfluidic channel independently comprises a first combined solution channel which is in fluid communication with each of the first solution channels. In certain embodiments, each microfluidic channel independently comprises a second combined solution channel, which is in fluid communication with each of the second solution channels. In certain embodiments, each microfluidic channel independently comprises a third combined solution channel, which is in fluid communication with the first combined solution channel and the second combined solution channel. In certain embodiments, each microfluidic channel independently comprises an outlet which is in fluid communication with the third combined solution channel. In certain embodiments, each combined solution channel independently comprises at least one chaotic mixing region. In certain embodiments, each first solution channel comprises at least one chaotic mixing region. In certain embodiments, each second solution channel comprises at least one chaotic mixing region.

[0072] In certain embodiments, each first solution inlet is independently in fluid communication with a first solution source which is pneumatically regulated by a microcontroller. In certain embodiments, each second solution inlet is independently in fluid communication with a second solution source which is pneumatically regulated by a microcontroller.

[0073] In certain embodiments, the device further comprises a collection plate beneath the parallel array of microfluidic channels and adjacent to the outlet. In certain embodiments, the collection plate has a number of wells which corresponds to the number of outlets in the Attorney Docket No. 046483-7491W01(04073) device. For example, in certain non-limiting embodiments, wherein the device comprises 8 parallel microfluidic channels, the collection plate has 8 wells aligned with the device outlets.

[0074] In certain embodiments, the collection plate is regulated by the microcontroller.

[0075] In certain embodiments, the microcontroller is configured to move the collection plate in response to a change in pressure. In certain embodiments, the response occurs in less than about 100 milliseconds.

[0076] In certain embodiments, the parallel array of microfluidic channels comprise at least one photolithographically etched layer bonded to a borosilicate glass surface.

[0077] In certain embodiments,

[0078] In certain embodiments, the microfluidic device comprises a plurality of functional layers arranged in a stacked configuration. In certain embodiments, the device comprises a mixing layer. In certain embodiments, the device comprises a resistor or herringbone layer. In certain embodiments, the device comprises a delivery layer. In certain embodiments, the device comprises a through-silicon-via (TSV) interconnect layer providing vertical fluid communication between layers.

[0079] In certain embodiments, the device comprises a mixing layer defining at least one mixing channel configured to combine liquid solutions (e.g., a first solution and a second solution).

[0080] In certain embodiments, the mixing layer has a channel depth of about 50, 55, 60, 65, 70, 75, 80. 85, 90, 95, 100, 105. 110, 115. or about 120 pm. In certain embodiments, the mixing layer may comprise junctions, serpentine segments, or expansion chambers that promote diffusive or convective mixing of incoming liquids.

[0081] In certain embodiments, the device further comprises a resistor layer or herringbone layer disposed downstream of the mixing layer. In certain embodiments, the resistor layer or herringbone layer comprises microstructured ridges or narrowed channel sections that generate local flow perturbations or increased hydraulic resistance. In certain embodiments, the resistor or herringbone layer has a channel depth of about 10, 15, 20, 25, or about 30 pm.

[0082] In certain embodiments, the herringbone structures are oriented at an angle of 30°, 35°, 40°, 45°, 50°, 55°, or about 60° relative to the longitudinal axis of the channel and repeat with a pitch of 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, 90, 95, or about 100 pm.

[0083] In certain embodiments, the device comprises a delivery layer in fluid communication with the resistor or herringbone layer. In certain embodiments, the delivery layer comprises one or more delivery channels or outlets configured to transport the mixed fluid to a downstream interface or to an outlet port of the device. In certain embodiments, the delivery Attorney Docket No. 046483-7491W01(04073) layer has a channel depth of about 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360, 370. 380, 390, 400, 410, 420. 430, 440, 450, 460, 470, 480, 490. or about 500 pm.

[0084] In certain embodiments, the device comprises a through-silicon-via (TSV) layer or vertical interconnect layer. In certain embodiments, this layer comprises one or more vias extending through a silicon or glass substrate to provide vertical fluid communication between different functional layers of the device or to connect the microfluidic channels to external inlets or outlets. In certain embodiments, the TSVs have a diameter of about 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, or about 200 pm and / or extend through a substrate of 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220. 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360, 370, 380, 390, 400. 410, 420, 430, 440, 450. 460, 470, 480, 490, or about 500 pm thickness.

[0085] In certain embodiments, m is 1. In certain embodiments, m is 2. In certain embodiments, m is 3. In certain embodiments, m is 4. In certain embodiments, m is 5. In certain embodiments, m is 6. In certain embodiments, m is 7. In certain embodiments, m is 8. In certain embodiments, m is 9. In certain embodiments, m is 10. In certain embodiments, n is

[0086] 1. In certain embodiments, n is 2. In certain embodiments, n is 3. In certain embodiments, n is

[0087] 4. In certain embodiments, n is 5. In certain embodiments, n is 6. In certain embodiments, n is

[0088] 7. In certain embodiments, n is 8. In certain embodiments, n is 9. In certain embodiments, n is

[0089] 10.

[0090] In certain embodiments, the device comprises a single microfluidic channel. In certain embodiments, the device comprises an array of 2 parallel microfluidic channels. In certain embodiments, the device comprises an array of 3 parallel microfluidic channels. In certain embodiments, the device comprises an array of 4 parallel microfluidic channels. In certain embodiments, the device comprises an array of 5 parallel microfluidic channels. In certain embodiments, the device comprises an array of 6 parallel microfluidic channels. In certain embodiments, the device comprises an array of 7 parallel microfluidic channels. In certain embodiments, the device comprises an array of 8 parallel microfluidic channels.

[0091] In certain embodiments, each occurrence of the first solution source independently comprises a lipid phase.

[0092] In certain embodiments, the lipid phase comprises an ionizable lipid. In certain embodiments, the lipid phase comprises cholesterol and / or a modified derivative thereof. In certain embodiments, the lipid phase comprises a neutral or helper lipid. In certain embodiments, the lipid phase comprises a polymer conjugated lipid. In certain embodiments, the lipid phase comprises ethanol. Attorney Docket No. 046483-7491W01(04073)

[0093] In certain embodiments, the lipid phase comprises a first lipid phase (ii). In certain embodiments, the lipid phase comprises a second lipid phase (z ). In certain embodiments, the lipid phase comprises a third lipid phase (is). In certain embodiments, the first lipid phase (ii) comprises at least one ionizable lipid and cholesterol and / or a modified derivative thereof. In certain embodiments, the second lipid phase (is) comprises ethanol. In certain embodiments, the third lipid phase (is) comprises at least one neutral lipid or helper lipid and a polymer- conjugated lipid.

[0094] In certain embodiments, each occurrence of the second solution source independently comprises an aqueous phase. In certain embodiments, the aqueous phase independently comprises a nucleic acid solution. In certain embodiments, the aqueous phase independently comprises a buffer solution. In certain embodiments, the aqueous phase comprises a first aqueous phase (14). In certain embodiments, the aqueous phase comprises a second aqueous phase (is). In certain embodiments, the aqueous phase comprises a third aqueous phase (is). In certain embodiments, the first aqueous phase (14) comprises a nucleic acid. In certain embodiments, the nucleic acid comprises DNA. In certain embodiments, the nucleic acid comprises RNA. In certain embodiments, the nucleic acid comprises mRNA. In certain embodiments, the nucleic acid comprises sgRNA. In certain embodiments, the nucleic acid comprises siRNA. In certain embodiments, the second aqueous phase (is comprises a buffer solution. In certain embodiments, the second aqueous phase (is) comprises citrate buffer. In certain embodiments, the citrate buffer comprises 10 mM citrate buffer at a pH of about 3. In certain embodiments, the third aqueous phase (is) comprises a buffer solution. In certain embodiments, the third aqueous phase (is) comprises citrate buffer. In certain embodiments, the citrate buffer comprises 10 mM citrate buffer at a pH of about 3. In certain embodiments, the second aqueous phase (is) and third aqueous phase (is) are identical.

[0095] In certain embodiments, the chaotic mixing region has an architecture selected from the group consisting of herringbone, ring, and serpentine.

[0096] Microfluidic Processing Apparatus

[0097] In another aspect, the disclosure provides a microfluidic processing apparatus. In certain embodiments, the microfluidic processing apparatus comprises the microfluidic device of the disclosure. In certain embodiments, the microfluidic processing apparatus comprises a collection plate comprising a plurality of wells. In certain embodiments, the microfluidic processing apparatus comprises a manipulator configured to hold the microfluidic device above the collection plate. In certain embodiments, the microfluidic Attorney Docket No. 046483-7491W01(04073) processing apparatus comprises a motion control apparatus operatively coupled to the manipulator and configured to position the microfluidic device relative to the collection plate such that each outlet of the device is aligned with a well of the collection plate. In certain embodiments, the microfluidic processing apparatus comprises a controller configured to actuate the motion control apparatus to reposition the microfluidic device from a first configuration, in which an outlet of the device is aligned with a first well of the collection plate, to an alternate configuration, in which the outlet is aligned with a different well of the collection plate. In certain embodiments, the microfluidic processing apparatus comprises a fluid control system operatively coupled to the microfluidic device and configured to regulate pressure applied to each inlet of the microfluidic device.

[0098] In certain embodiments, the collection plate comprises a microwell plate having an array of wells for receiving fluid dispensed from the microfluidic device. In certain embodiments, the microplate has a number of wells equivalent to the number of outlets in the microfluidic device. In certain embodiments, the microplate wells are aligned with the outlets of the microfluidic device.

[0099] In certain embodiments, the controller is configured to adjust the position of the microfluidic device automatically (e.g., by a timed program).

[0100] In certain embodiments, the controller is configured to adjust the position of the microfluidic device responsive to one or more stimuli selected from the group consisting of user input, volume measurement, optical feedback, and pressure measurements.

[0101] Methods

[0102] In another aspect, the disclosure provides a method for preparing a plurality of lipid nanoparticles (LNPs) using the microfluidic processing apparatus of the disclosure. In certain embodiments, the method comprises providing the microfluidic processing apparatus of the disclosure. In certain embodiments, the method comprises actuating the motion control apparatus to position the microfluidic device such than each outlet of the device is aligned with an empty well of the collection plate. In certain embodiments, the method comprises introducing a first solution into each first solution inlet of the microfluidic device to provide a combined first solution. In certain embodiments, the method comprises introducing a second solution into each second solution inlet of the microfluidic device to provide a combined second solution. In certain embodiments, the method comprises mixing the combined first solution and combined second solution to provide a formulation comprising a plurality of LNPs. In certain embodiments, the method comprises dispensing at least a portion of the Attorney Docket No. 046483-7491W01(04073) formulation comprising the plurality of LNPs from each outlet to each respective well of the collection plate. In certain embodiments, the method comprises repeating steps (c)-(f) (i.e.. the actuating, introducing, mixing, and dispensing steps) after changing at least one parameter of the method. In certain embodiments, a parameter of the method comprises a flow rate through the microfluidic device. In certain embodiments, a parameter of the method comprises an identi ty of at least one component of at least one of the first solution or the second solution. For example, in certain non-limiting embodiments, modifying the identity of a component of the at least one first solution comprises substituting ionizable lipid SM-102 for D-Lin-MC3-DMA. In certain embodiments, a parameter of the method comprises a concentration of at least one component of at least one of the first solution or the second solution.

[0103] In certain embodiments, each occurrence of the first solution source independently comprises a lipid phase.

[0104] In certain embodiments, the lipid phase comprises an ionizable lipid. In certain embodiments, the lipid phase comprises cholesterol and / or a modified derivative thereof. In certain embodiments, the lipid phase comprises a neutral or helper lipid. In certain embodiments, the lipid phase comprises a polymer conjugated lipid. In certain embodiments, the lipid phase comprises ethanol.

[0105] In certain embodiments, the lipid phase comprises a first lipid phase ( / ;). In certain embodiments, the lipid phase comprises a second lipid phase (12). In certain embodiments, the lipid phase comprises a third lipid phase (ii). In certain embodiments, the first lipid phase (z7) comprises at least one ionizable lipid and cholesterol and / or a modified derivative thereof. In certain embodiments, the second lipid phase (b) comprises ethanol. In certain embodiments, the third lipid phase (1 ) comprises at least one neutral lipid or helper lipid and a polymer- conjugated lipid.

[0106] In certain embodiments, each occurrence of the second solution source independently comprises an aqueous phase. In certain embodiments, the aqueous phase independently comprises a nucleic acid solution. In certain embodiments, the aqueous phase independently comprises a buffer solution. In certain embodiments, the aqueous phase comprises a first aqueous phase (b). In certain embodiments, the aqueous phase comprises a second aqueous phase (zb). In certain embodiments, the aqueous phase comprises a third aqueous phase (ie). In certain embodiments, the first aqueous phase (b) comprises a nucleic acid. In certain embodiments, the nucleic acid comprises DNA. In certain embodiments, the nucleic acid comprises RNA. In certain embodiments, the nucleic acid comprises mRNA. In certain Attorney Docket No. 046483-7491W01(04073) embodiments, the nucleic acid comprises sgRNA. In certain embodiments, the second aqueous phase (zb) comprises a buffer solution. In certain embodiments, the second aqueous phase (zb) comprises citrate buffer. In certain embodiments, the citrate buffer comprises 10 mM citrate buffer at a pH of about 3. In certain embodiments, the third aqueous phase (ze) comprises a buffer solution. In certain embodiments, the third aqueous phase (z<?) comprises citrate buffer. In certain embodiments, the citrate buffer comprises 10 mM citrate buffer at a pH of about 3. In certain embodiments, the second aqueous phase (zj) and third aqueous phase (z<j) are identical.

[0107] In certain embodiments, the ionizable lipid has a concentration ranging from about 0.1, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5. or about 10 mg / mL in the first solution. In certain embodiments, the ionizable lipid has a concentration of about 2.2 mg / mL in the first solution.

[0108] In certain embodiments, the cholesterol and / or modified derivative thereof has a concentration of about 0.01, 0.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, 0.50, 0.55, 0.60, 0.65, 0.70, 0.75, 0.80, 0.85, 0.90, 0.95, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, or about 5.0 mg / mL in the first solution. In certain embodiments, the cholesterol and / or modified derivative thereof has a concentration of about 0.6 mg / mL.

[0109] In certain embodiments, the neutral or helper lipid has a concentration of about 0.1, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0. 8.5, 9.0, 9.5, or about 10 mg / mL in the first solution. In certain embodiments, the neutral or helper lipid has a concentration of about 1 .0 mg / mL in the first solution.

[0110] Ionizable Lipids and / or Cationic Lipids

[0111] The scope of ionizable lipids contemplated for use in the present disclosure is not limited to ionizable lipids described herein. In the lipid nanoparticles of the disclosure, the cationic lipid or ionizable lipid may comprise, e.g., one or more of the following: (6Z,9Z,28Z,31Z)-heptatriaconta-6,9,28,31-tetraen-19-yl 4-(dimethylamino)butanoate (DLinMC3DMA), [(4-hydroxybutyl)azanediyl]di(hexane-6,l-diyl) bis(2 -hexyldecanoate) (ALC-0315), heptadecan-9-yl 8-{(2-hydroxyethyl)[6-oxo-6- (undecyloxy)hexyl]amino}octanoate (SM-102), l,l'-[[2-[4-[2-[[2-[bis(2- hy droxy dodecyl)amino] ethyl] (2-hy droxy dodecyl)amino] ethyl]- 1 - piperazinyl]ethyl]imino]bis-2-dodecanol (C12-200), l,2-dilinoleyloxy-N,N- dimethylaminopropane (DLinDMA), 1 ,2-dilinolenyloxy-N.N-dimethylaminopropane (DLenDMA), 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[l,3]-dioxolane (DLin-K-C2-DMA; Attorney Docket No. 046483-7491W01(04073)

[0112] “XTC2”), 2,2-dilinoleyl-4-(3- 45 dimethylaminopropyl)- l,3]-di oxolane (D Lin-K-C3-D MA). 2, 2-dilinoleyl-4-(4-dimethylaminobutyl)-[l,3]-di oxolane (DLin-K-C4-DMA), 2.2- dilinoley 1-5 -dimethylaminomethyl- [ l,3]-dioxane (DLin-K6-DMA), 2,2-dilinoleyl-4-N- methylpepiazino-[l,3]-dioxolane (DLin-K-MPZ), 2,2-dili-noleyl-4-dimethylaminomethyl- [1,3] -dioxolane (DLin-KDMA), l,2-dilinoleylcarbamoyloxy-3-dimethylaminopropane (D Lin-C-DAP). l,2-dilinoleyoxy-3-(dimethylaminoacetoxypropane (DLin-DAC), 1- 2dilinoleyoxy-3-morpholinopropane (DLin-MA), l,2-dilinoleoyl-3-dimethylaminopropane (DLinDAP), l,2-dilinoleylthio-3-dimethylaminopropane (DLin-2-DMAP), 1,2-dilinoleyloxy- 3 -trimethylaminopropane chloride salt (DLin-TMA.Cl), l,2-dilinoleoyl-3- trimethylaminopropane chloride salt (DLin-TAP.Cl), l,2-dilinoleyloxy-3-(N- methylpiperazino)propane (D Lin-MPZ), 3-(N,N-dilinoleylamino)-1.2-propanediol (D LinAP), 3-(N,N-dioleylamino)-l,2-propanedio (DOAP), l,2-dilinoleyloxo-3-(2-N,N- dimethylamino)ethoxypropane (D Lin-EG-D MA), N,N-dioleyl-N,N-dimethylanrmonium chloride (DODAC), l,2-dioleyloxy-N,N-dimethylaminopropane (DODMA), 1,2- distearyloxy-N,N-dimethylaminopropane (DSD MA). N-(l-(2,3-dioleyloxy)propyl)-N,N,N- trimethylammonium chloride (DOTMA), N,N-distearyl-N,N-dimethylammonium bromide (DDAB), N-(l-(2,3-dioleoyloxy)propyl)-N,N, N-trimethylammonium chloride (DOTAP), 3- (N-(N’,N’dimethylaminoethane)-carbamoyl)cholesterol (DC-Chol), N-(l,2- dimyristyloxyprop-3-yl)-N,N-dimethyl-N-hydroxyethyl ammonium bromide (DMRIE), 2,3- dioleyloxy-N-[2 (spermine-carboxamidoethyl]-N.N-dimethy 1-1- propanaminiumtrifluoroacetate (DOSPA), di octadecyl amidoglycyl spermine (DOGS), 3- dimethylamino-2-(cholest-5-en-3-beta-oxybutan-4-oxy)-l-(cis,cis-9,12- octadecadienoxy)propane (CLinDMA), 2-[5’-(cholest-5-en-3-beta-oxy)-3’-oxapentoxy)-3- dimethyl-l-(cis,cis-9',l-2’-octadecadienoxy) propane (CpLinDMA), N,N-dimethyl-3,4- dioleyloxy benzylamine (DMOBA), l,2-N,N’dioleylcarbamyl-3-dimethylaminopropane (DOcarbDAP), l,2-N,N’-dilinoleylcarbamyl-3-dimethylaminopropane (DLincarbDAP), or mixtures thereof. In certain embodiments, the cationic lipid is DLinDMA, DLin-K-C2-DMA (“XTC2"), or mixtures thereof. The ionizable lipids are not limited to those recited herein, and can further include ionizable lipids known to those skilled in the art, or described in PCT Application No. PCT / US2020 / 056255 and / or PCT Application No. PCT / US2020 / 056252, the disclosures of which are herein incorporated by reference in its entirety7.

[0113] The synthesis of cationic lipids such as DLin-K-C2-DMA (“XTC2 ’), DLin-K-C3- DMA, DLin-K-C4-DMA, DLin-K6-DMA, and DLin-K-MPZ, as well as additional cationic lipids, is described in U.S. Application Publication No. US 2011 / 0256175, the disclosure of Attorney Docket No. 046483-7491W01(04073) which is herein incorporated by reference in its entirety for all purposes. The synthesis of cationic lipids such as DLin-K-DMA, DLin-CDAP. DLin-DAC, DLin-MA, DLinDAP, DLin-S-DMA, DLin-2-DMAP, DLin-TMA.Cl, DLin-TAP.Cl, DLin-MPZ, DLinAP, DOAP, and DLin-EG-DMA, as well as additional cationic lipids, is described in PCT Application No. PCT / US08 / 88676, filed December 31, 2008, the disclosure of which is herein incorporated by reference in its entirety for all purposes. The synthesis of cationic lipids such as CLinDMA, as well as additional cationic lipids, is described in U.S. Patent Publication No. US20060240554, the disclosure of which is herein incorporated by reference in its entirety for all purposes.

[0114] Non-Cationic Lipid

[0115] In the nucleic acid-lipid particles of the present disclosure, the non-cationic lipid may comprise, e.g., one or more anionic lipids, helper lipids, and / or neutral lipids. In some embodiments, the non-cationic lipid comprises one of the following neutral lipid components: (1) cholesterol or a derivative thereof (2) a phospholipid (i.e., helper lipid or neutral lipid); or (3) a mixture of a phospholipid and cholesterol or a derivative thereof.

[0116] Examples of cholesterol derivatives include, but are not limited to, cholestanol, cholestanone, cholestenone, coprostanol, cholesteryl-2’ -hydroxy ethyl ether, cholesteryl-4’- hydroxybutyl ether, and mixtures thereof. The synthesis of cholesteryl-2'-hydroxyethyl ether is known to one skilled in the art and described in U.S. Patent Nos. 8.058,069, 8.492,359, 8,822,668, 9,364,435, 9,504,651, and 11 ,141,378, all of which are hereby incorporated herein in their entireties for all purposes.

[0117] Non-limiting examples of non-cationic lipids, such as phospholipids, include lecithin, phosphatidylethanolamine, lysolecithin, lysophosphatidylethanolamine, phosphatidylserine, phosphatidylinositol, sphingomyelin, egg sphingomyelin (ESM), cephalin, cardiolipin, phosphatidic acid, cerebrosides, dicetylphosphate, distearoylphosphatidylcholine (DSPC), dioleoy lphosphatidylcholine (DOPC), dipalmitoylphosphatidylcholine (DPPC), dioleoylphosphatidylglycerol (DOPG), dipalmitoylphosphatidylglycerol (DPPG), ioleoylphosphatidylethanolamine (DOPE), palmitoyloleoylphosphatidylcholine (POPC), palmitoyloleoylphosphatidylethanolamine (POPE), palmitoyloleyolphosphatidylglycerol (POPG), dioleoylphosphatidylethanolamine 4-(N-maleimidomethyl)-cyclohexane-l- carboxylate DOPE-mal), dipalmitoylphosphatidylethanolamine (DPPE), dimyristoylphosphatidylethanolamine (DMPE). distearoylphosphatidylethanolamine (DSPE), monomethylphosphatidylethanolamine, dimethylphosphatidylethanolamine, Attorney Docket No. 046483-7491W01(04073) dielaidoylphosphatidylethanolamine (DEPE), stearoyloleoylphosphatidylethanolamine (SOPE), lysophosphatidylcholine, dilinoleoylphosphatidylcholine, and mixtures thereof.

[0118] Other diacylphosphatidylcholine and diacylphosphatidylethanolamine phospholipids can also be used. The acyl groups in these lipids can be, for example, acyl groups derived from fatty' acids having C10-C24 carbon chains, e.g, lauroyl, myristoyl, palmitoyl, stearoyl, or oleoyl. Additional examples of non-cationic lipids include sterols such as cholesterol and derivatives thereof such as cholestanol, cholestanone, cholestenone, coprostanol, cholesteryl- 2’ -hydroxy ethyl ether, cholesteryl-4’-hydroxybutyl ether, and mixtures thereof. In certain embodiments, the phospholipid is DPPC, DSPC, or mixtures thereof.

[0119] Conjugated Lipid

[0120] In the nucleic acid-lipid particles of the present disclosure, the conjugated lipid that inhibits aggregation of particles may comprise, e.g., one or more of the following: a polyethyleneglycol (PEG) lipid conjugate, a polyamide (ATTA)-lipid conjugate, a cationic- polymer-lipid conjugates (CPLs), or mixtures thereof. In some embodiments, the nucleic acid-lipid particles comprise either a PEG-lipid conjugate or an ATTA-lipid conjugate.

[0121] PEG is a linear, water-soluble polymer of ethylene PEG repeating units with two terminal hydroxyl groups. PEGs are classified by their molecular weights; for example, PEG 2000 has an average molecular weight of about 2,000 daltons, and PEG 5000 has an average molecular weight of about 5.000 daltons. PEGs are commercially available from Sigma Chemical Co. and other companies and include, for example, the following: monomethoxypolyethylene glycol (MePEGOH), monomethoxypolyethylene glycolsuccinate (MePEGS), monomethoxypolyethylene glycolsuccinimidyl succinate (MePEG-S-NHS), monomethoxypolyethylene glycolamine (MePEG-NEb), monomethoxypolyethylene glycoltresylate (MePEG-TRES), and monomethoxypolyethylene glycolimidazolylcarbonyl (MePEG-IM). Other PEGs such as those described in U.S. Patent Nos. 6,774,180 and 7,053,150 (e.g., mPEG (20 KDa) amine) are also useful for preparing the PEG-lipid conjugates of the present disclosure. The disclosures of these patents are herein incorporated by reference in their entirety for all purposes. In addition, monomethoxypolyethyleneglycolacetic acid (MePEG-CEbCOOH) is particularly useful for preparing PEG-lipid conjugates including, e.g., PEG-DAA conjugates.

[0122] In certain embodiments, the PEG-lipid conjugate or ATTA-lipid conjugate is used together with a CPL. The conjugated lipid that inhibits aggregation of particles may comprise a PEG-lipid including, e.g., a PEG-diacylglycerol (DAG), a PEG dialkyloxypropyl (DAA), a Attorney Docket No. 046483-7491W01(04073)

[0123] PEG-phospholipid, a PEG-ceramide (Cer), or mixtures thereof. The PEGDAA conjugate maybe PEG-dilauryloxypropyl (C12), a PEG-dimyristyloxypropyl (C14). a PEG- dipalmityloxypropyl (Cis), a PEG-distearyloxypropyl (Cis), or mixtures thereof.

[0124] Additional PEG-lipid conjugates suitable for use in the disclosure include, but are not limited to, mPEG2000-l,2-diO-alkyl-sn3-carbomoylglyceride (PEG-C-DOMG). The synthesis of PEG-C-DOMG is described in PCT Application No. PCT / US08 / 88676, filed December 31, 2008, the disclosure of which is herein incorporated by reference in its entirety for all purposes. Yet additional PEG-lipid conjugates suitable for use in the disclosure include, without limitation, l-[8’-(l,2-dimyristoyl-3-propanoxy)-carboxamido-3’,6’- dioxaoctanyl] carbamoyl-methyl-poly(ethylene glycol) (2 KPEG-DMG). The synthesis of 2 KPEG-DMG is described in U.S. Patent No. 7,404,969. the disclosure of which is herein incorporated by reference in its entirety for all purposes.

[0125] The PEG moiety of the PEG-lipid conjugates described herein may comprise an average molecular weight ranging from about 550 daltons to about 10,000 daltons. In certain instances, the PEG moiety has an average molecular weight of from about 750 daltons to about 5,000 daltons (e.g., from about 1,000 daltons to about 5,000 daltons, from about 1,500 daltons to about 3,000 daltons, from about 750 daltons to about 3,000 daltons, from about 750 daltons to about 2,000 daltons, etc.). In some embodiments, the PEG moiety- has an average molecular weight of about 2,000 daltons or about 750 daltons.

[0126] In addition to the foregoing, it will be readily apparent to those of skill in the art that other hydrophilic polymers can be used in place of PEG. Examples of suitable polymers that can be used in place of PEG include, but are not limited to, polyvinylpyrrolidone, polymethyloxazoline, polyethyloxazoline, polyhydroxypropyl methacry lamide, polymethacrylamide and poly dimethylacrylamide, polylactic acid, poly glycolic acid, and derivatized celluloses such as hydroxymethylcellulose or hydroxy ethylcellulose.

[0127] EXAMPLES

[0128] Various embodiments of the present application can be better understood by reference to the following Examples which are offered by way of illustration. The scope of the present application is not limited to the Examples given herein.

[0129] Materials and Methods

[0130] Fabrication ofLIBRIS chip

[0131] Chips were fabricated in the Quattrone Nanofabrication Facility at the University of Attorney Docket No. 046483-7491W01(04073)

[0132] Pennsylvania. The chip design comprises four layers, designed in AutoCAD (San Rafael, CA): the mixing channels (Layer 1, depth: 85pm), the resistors and herringbones (Layer 2. depth: 20pm), the delivery channels (Layer 3, depth: 365pm), and the "through silicon vias" or interconnects between both sides of the wafer (Layer 4). In brief, using a Heidelberg 66+ laser writer (Heidelberg, Germany), chrome-coated soda lime photomasks (Telic Company, Santa Clarita, CA) were patterned, developed in AZ300 MIF developer (EMD Performance Materials Corp., Philadelphia, PA), and then the exposed pattern was etched into the chrome (Transene Company, Danvers, MA). Remaining photoresist was removed by PE Remover (DuPont, Wilmington, DE) under sonication at held at 65 °C for 10 min.

[0133] The general pattern for fabrication of each layer etched into a single 500pm thick, 100mm diameter silicon wafer (ID 775; University Wafer, South Boston. MA) is as follows: prepare wafer with resist adhesion promoter, spray coat resist, bake, rest, lithographically pattern, rest, develop, etch, perform metrology', adjust etching if feature depths are incorrect, and clean before moving to the next layer. In greater detail, at the beginning of each layer, the wafer was coated with 49% HF for 30s, then rinsed thoroughly with water, and dried on a 200 °C hotplate for 2 min. 81805 photoresist (Dow, Midland, MI) was mixed at a 1 :8 ratio with acetone, then spray-coated on the processed wafer to desired thickness using an AS8 AltaSpray Coater (SUSS MicroTec, Garching, Germany). Wafers were baked at 110 °C for approximately 5 min per 4 pm of resist immediately after spray coating, and then left to rest at room temperature for 30min for each step to prevent outgassing during exposure. The respective resist thicknesses for each layer 1-4 are 8pm, 8pm, 16pm, and 4pm. After spray coating, resting, and baking, a mask aligner (MA6, SUSS Microtec) was used to expose the coated wafer. Exposure times varied depending on resist thickness. After exposure, each wafer was developed in AZ300 MIF developer, then rinsed, dried at room temperature, and subsequently under- went a deep reactive ion etch (DRIE) (SPTS Rapier Si, Newport, UK) to the etch depths listed above. After etching, each wafer was profiled using an optical profilometer (Profilm 3D, Filmetrics, San Diego, CA), and cleaned using PE remover and Nano-Strip (VWR. Radnor, PA) heated to 110 °C. This process was repeated for each layer.

[0134] Layer 2 (resistors and herringbones) underwent a slight modification to the procedure to preserve the edges of the mixing channels where resist can delaminate during the exposure process. After cleaning from the end of Layer 1, the wafers were conformally coated with 500 nm of SiO2 using plasma-enhanced chemical vapor deposition (Oxford Plasma Lab 100 PECVD) (Oxford Instruments, Abingdon, UK). The layer process proceeded as described until after exposure, where the exposed 500nm layer of SiO2 was etched away using CF 4 Attorney Docket No. 046483-7491W01(04073) plasma in the reactive ion etcher (Oxford 80 Plus, Oxford Instru- ments, Concord, MA).

[0135] For Layer 4, a 4pm thick layer of SiO2 was deposited using PECVD on the delivery channel side (backside) of the wafer to prevent the breaking of vacuum lock under the final through silicon via DRIE step. Instead of promoting resist adhesion using 49% HF, which would rapidly dissociate this oxide layer, the wafer was coated with bis(trimethylsilyl)amine (HMDS) (Genesis Prime Oven, Genesis Systems, Davenport, IA) for 10 min.

[0136] The fully etched wafer, and two 100mm Borofloat 33 glass wafers (ID 517;

[0137] University Wafer), one of which was micromachined with a custom inlet-outlet hole setup using an IX-255 laser system (IPG Photonics, Oxford, MA), are left in Nano-Strip at 110 °C for 30 min. All wafers are spin-rinse dried (RENA Compass, RENA, Albany, OR), then stacked in a glass / silicon / glass arrangement, aligning the inlet-outlet glass holes with the etched delivery channels. The triple stacked wafer is anodically bonded at 900V and 400°C under 100N piston force using an EVG 510 Wafer Bonding System (EVG Group, Oberosterreich, Austria), then the wafer stack is let to cool to at least 180 °C before flipping the wafer stack over to repeat the bonding in the other direction, producing a bonded triple stack of wafers. The wafers are diced into single chips using a dicing saw (ADT 7100, ADT, Horsham, PA) fitted with a resin blade.

[0138] Circuit simulation

[0139] A replica of the chip was rendered in LTSpice Version 17.2.4. Pressure (Pa) was analogized as voltage for each respective fluidic inlet i across each set of pressures that were experimentally validated (Pi,n)- Flow rate (mnf’ / s) was analogized to current. Hydraulic or fluidic resistance (Pa x s / mm3) was analogized to electrical resistance. Hydraulic resistances were calculated from dimensions determined by measurements of each chip using optical profilometry (Profilm 3D). Pairwise distance analyses were performed between relative flow rate values derived from the LTSpice simulation and both predicted values and actual values. Correlation coefficients were calculated using a Mantel test.

[0140] Mechanical integration of pneumatic regulator setup with chip

[0141] A custom housing for the chip was designed in Fusion360 (Autodesk) and machined from 6061-T651 aluminum (Protolabs, Maple Plain, MN). 1 / 4" - 28 UNF threaded ports were included on the backside of the chip to allow for collection of inlets and outlets and viewing windows exposing all mixing channels were located on the front side of the chip. A nitrogen tank was connected to six dual-valve pressure controllers (Alicat Scientific, Tucson, AZ), via Attorney Docket No. 046483-7491W01(04073) a 7-port manifold for 1 / 16" outer diameter (OD) tubing (IDEX H&S, Oak Harbor, WA). Each regulator was connected to a pressurizable vessel (Elveflow, Paris, FR) via flat bottom 1 / 4"- 28 PEEK M-UNF fittings (IDEX H&S). These vessels contained the fluidic inputs, which were distributed to the separate delivery channels on chip through a pressure fit connection between the machined glass and a 1 / 16" OD ferrule (IDEX H&S). For the collection of outputs, eight M-UNF l / 4"-28 x Luer slip adapters (IDEX H&S) were threaded into the 1 / 4"- 28 M-UNF output ports of the aluminum housing, and 30G x 9mm low-dead volume needles (Air-Tite, Virginia Beach, VA) were press fitted to the adapters. An additional manifold was cut from l / 8"-thick clear acrylic and fit over the collars of the needles to better align and support the needles in the 4 x 2 array.

[0142] Design and fabrication of the plate robot

[0143] All parts of the skeleton of the robot, including robot legs, plate stage, and motor compartments, were designed in Fusion360. Lead screw nuts were printed using a Form2 SLA printer with Durable Resin (Formlabs, Somerville. MA). All other components were printed from PolyLite™ PETG filament (Polymaker, Shanghai, China). The top base plate which held both the chip and the xy z stage for the portable microscope was laser cut from 1 / 4" thick clear scratch-proof acry lic (McMaster-Carr, Robbinsville, NJ) with a 0.0075" tolerance. Additional parts, such as stainless steel and ceramic-coated 6061- T651 aluminum shafts (3 / 8" Diameter, 9" Long), lead screws (1018 Carbon Steel Precision Acme Lead Screw, Fast-Travel, Right-Hand, 3 / 8"-5 Thread, 5: 1 Speed), and stainless steel bearings (3 / 8" Bore, 3 / 8" x 5 / 8" x 1-19 / 32") were acquired (McMaster-Carr), and the skeleton of the robot was fabricated using additional bolts, nuts, and spacers from the GM Lab at the University of Pennsylvania.

[0144] The skeleton of the robot was fabricated by first assembling the central plate support as shown (FIG. 4E). This assembly was then mounted on two 12-inch linear motion shafts (Tapped Linear Motion Shaft, 52100 Alloy Steel, 3 / 8" Diameter, 8" Long, McMaster Carr), which were secured to H-supports at both ends using 16mm M4 screws (GM Lab). The top acrylic sheet was press fitted onto the H-supports to complete the primary framework. Motors for motion in the z-axis (two motors), y-axis, and x-axis were purchased from Digikey and assembled into their respective compartments (FIGs. 4A-4G).

[0145] Integration of robot motion with pneumatic regulator control: To control the robot, an Arduino Mega (Monza, IT) was utilized as the primary con- troller. X and y motors were controlled utilizing an Adafruit motor shield (v2, Adafruit Industries) with an external 12V Attorney Docket No. 046483-7491W01(04073)

[0146] 2A power supply. Two A4998 stepper motor drivers (Pololu Corporation, Las Vegas, Nevada) with a 1 / 4 microstep were used to control both z stepper motors, with an addition 12V 3 A power supply. To serially communicate with the five 50 psi PCD-series pressure regulators and one 100 psi PCD-series pressure regulators. A simplex architecture operating with an inverted transistor-transistor logic protocol at a baud rate of 19200 was employed, where the gauge pressure of the devices was read via analog pins on the Arduino Mega. Communication was performed through a 8-pin mini-DIN connector and 8- pin mini-DIN to breadboard adapter.

[0147] Development of a graphical user interface

[0148] An ESP32-C3 M5 stamp (Digikey) was used to generate a wifi router hosting a website developed in CSS, HTML, and Javascript. This stamp was connected to the primary Arduino Mega using an serial peripheral interface communication protocol.

[0149] Mixing characterization ofLIBRIS chip

[0150] Solutions of 10 kDa FITC-dextran (MilliporeSigma, Burlington, MA), 10 kDa rhodamine B isothiocyanate-dextran (MilliporeSigma), and 10 kDa Cascade Blue®-dextran were dissolved to 10 pM in ultrapure (UP) water, and filtered through 0.22 pm polyethersulfone or poly vinylidene fluoride filters (ThermoFisher). Each solution served as a separate fluidic input for each of the three ethanol or aqueous phase inputs. FITC-dextran being pressurized through ii and i4, rhodamine B-dextran through i2 and is, and Cascade Blue®-dextran through is and ie. The other three fluidic inputs (either aqueous or ethanol) comprised UP water. The image of the full chip was taken where all inputs ii and i4 carried FITC-dextran, i2 and is carried rhodamine B-dextran, and is and ie carried Cascade Blue- dextran (FIG. 2B). Each input was pressurized at the pressures used to generate the LNP libraries, resulting in a total flow rate of 300pL / min for the ethanol phase, and 900pL / min for the aqueous phase, thus 1.2mL / min through each combined mixing unit.

[0151] Fluorescence micrographs of the entire chip was taken using a fluorescence microscope (Leica. Wetzlar, Germany). Intensity profiles were measured for all three fluorescence channels at the same location on the chip using image analysis tools in Python.

[0152] Mixing efficiency was calculated using the equation described elsewhere herein. Solutions varied from H = 0 (completely unmixed, prior to any mixing cycles) to H= 1 (completely mixed). Above H = 0.8. or 1 - H = 0.2, solutions were considered completely mixed, validated by the fitting of an exponential decay model to each data set. Standard Atorney Docket No. 046483-7491W01(04073) errors were calculated from the model fit at 10 mixing cycles for the ethanol or aqueous SHMs or 15 mixing cycles for the LNP generating SHM.

[0153] Measurements of the relative flow rates in each channel were determined by analysis of the total intensities of each dye at the fully mixed stage (l-H<0.2, after 10 mixing cycles) of a channel when normalized to the intensity of the dye at an entirely unmixed stage (before 0 mixing cycles). Thus, for ii in mi, the relative intensity, 11, 1 , thus the normalized relative flow rate 0i,i is calculated: where each h,i is calculated from:

[0154] For a given fluorescence channel measurement from a given channel. To generate the images seen in FIG. 2B, images were stitched together and blended using Adobe Photoshop (San Jose, CA). All mixing analysis was performed from raw images.

[0155] LNP Formulation and Characterization

[0156] LNPs were formulated using the LIBRIS chip at the designated compositional ratios. For all three libraries (MC3, SM-102, C12-200), the IL and cholesterol were pooled in fluidic input ii and the phospholipid and PEG lipid were pooled together in fluidic input is. The stock solutions used to make the library comprised the lipids D-Lin-MC3-DMA (MedChemExpress, Monmouth Junction, NJ), l,2-distearoyl-sn-glycero-3-phosphocholine (DSPC, Avanti Polar Lipids, Al- abaster, AL), cholesterol (Sigma-Aldrich, St. Louis, MO), and DMG-PEG 2000 (Avanti Polar Lipids) at concentrations of 2.22 mg / mL, 0.55 mg / mL, 1.03 mg / mL. and 0.26 mg / mL respectively. The SM-102 library comprised SM-102 (Cayman Chemical, Ann Arbor, MI), DSPC, cholesterol, and DMG- PEG 2000 at concentrations of 2.22 mg / mL, 0.49 mg / mL, 0.93 mg / mL, and 0.24 mg / mL respectively. The C12-200 library comprised the lipids C12-200 (Cayman Chemical), l,2-dioleoyl-sn-glycero-3- phosphoethanolamine (DOPE, Avanti Polar Lipids), cholesterol, and 1,2-distearoyl-sn- glycero-3-phosphoethanolamine-N-[amino(polyethylene glycol)-2000] (DSPE-PEG 2000, Avanti Polar Lipids) at concentrations of 2.22 mg / mL, 0.66 mg / mL, 1.00 mg / mL, 0.38 mg / mL respectively. Input i2 comprised a solution of filtered ethanol.

[0157] The aqueous inputs i4, is, and ie comprised nucleic acid cargo, and two buffer inputs Attorney Docket No. 046483-7491W01(04073) of 10 mM citrate buffer at pH ~3 (Alpha Teknova, Inc., Hollister, CA) respectively. Measures of chip repeatability and scalability were conducted using a ssDNA cargo of the sequence 5'- ATG GTT CTA GCT T - 3' (SEQ ID NO: 1) (IDT, Coralville, IA) as the nucleic acid cargo with the SM-102 lipid set. Otherwise, the nucleic acid cargo consisted of either firefly luciferase or mCherry CleanCap® mCherry mRNA (5moU) (TriLink Biotechnologies San Diego, CA). The LIBRIS chip was primed for 10s using UP water, then either ethanol or citrate depending on the fluidic input, each for 30s. After priming, lipid stocks were loaded, then the microcontroller was initiated, pressurizing each input to the pressures referenced. Lines were flushed and flow equilibrated for 20s, the first sample was collected for 20s, then each set of LNPs was equilibrated between for 2.6s. After sample collection, the chip was flushed with 1% Alconox (Alconox Inc., Glenn Plains, NY) for 2min, then flushed with UP water for an additional 2min.

[0158] Each output was collected in a 0.3mL 20k MWCO Pierce Microdialysis Cassette (A50472, ThermoFisher Scientific, Waltham, MA), sitting in a well of 1.2 mL of lx phosphate buffered saline (PBS, ThermoFisher Scientific). The cassettes were arranged in a 4 x 2 pattern matching the chip output setup. After collection of all 32 outputs for a given set of fluidic inputs, the dialysis cassettes were transferred to new wells of 1.2 mL of PBS two additional times at intervals of 20 minutes. Samples were transferred from the cassettes and stored at 4 °C.

[0159] LNPs were diluted 1 :30 in lx PBS. and, using a DynaProPlate Reader III (Wyatt Technology, Santa Barbara, CA), dynamic light scattering was per- formed to quantify the hydrodynamic diameter (referred to as "size" through- out) and poly dispersity index (PDI). All sizes reported are the intensity -weighted mean hydrodynamic diameters (z-averages). Standard deviation is calculated based on measurement replicates. Relative encapsulation efficiency and RNA concentration were measured by reading out the results of a RiboGreen Quantit RNA assay kit (ThermoFisher) on an Infinite M Plex plate reader (Tecan, Miinnedorf, Switzerland).

[0160] In vitro characterization of mRNA transfection

[0161] HepG2 cells (HB-8065, ATCC, Manassas, VA) were cultured in DMEM (ThermoFisher) supplemented with 10% v / v fetal bovine serum (FBS, Ther- moFisher) and 1% v7 v Penicillin-Streptomycin (ThermoFisher). Cells were plated at 40k cells / lOOpL of media in a 96-well plate, then dosed after 24hr with either mCherry LNPs from the MC3, SM-102, or Cl 2-200 libraries at a dose of 10 ng mRNA / lOk cells. After another 24hr, cells Attorney Docket No. 046483-7491W01(04073) underwent preparation for flow cytometry as follows. Each well of cells was washed with 100 pL of 1 x PBS, then treated with 60 pL of 0.25% Trypsin-EDTA (ThermoFisher) for 6 min at 37 °C. Each well was diluted in 100 pL of DMEM, then spun dow n at 0.3 ref for 5 min. LIVE / DEAD™ Fixable Green Dead Cell Stain (ThermoFisher) w as diluted in lx PBS to a concentration of 1 pL of dye per 5mL of PBS. Media w as aspirated, and 50 pL of dye w as added to each well. Each well was agitated using a multichannel pipette and left to rest covered from light for 30 min at 25°C. 100 pL of lx TE buffer (ThermoFisher) was added to each well, and the plate was centrifuged at 0.3 ref for 5 min. TE was aspirated, and 70 pL of 2% v / v neutral buffered formalin (NBF, ThermoFisher) diluted in lx PBS w as added to each well, left to incubate covered from light for 10 min, then the plate was centrifuged at 0.6 ref for 5min. 150 pL of lx TE buffer was added to the cells, the cells were centrifuged again at 0.6 ref for 5min, then resuspended in 150 pL of lx PBS and stored at 4°C until run on the cytometer. A Guava® easyCyte™ HT System (Luminex, Austin, TX) was used to perform the cytometry' experiment.

[0162] In vivo characterization

[0163] Candidates w ere selected from the in vitro assay for high performance (C 12- 200 F4), for poor performance (Cl 2-200 F24), and from the previously optimized standard formulation for C12-200 for validation in vivo. Compositional ratios for each particle were calculated from the relative flow rates and input concentrations of F4 and F24 collected from the LIBRIS chip as they corresponded to their -in vitro performance. Ethanol stocks were mixed where all of the lipids w ere pooled at the calculated compositional ratios, and w ere precipitated together with 5-methoxyuridine CleanCap® Firefly Luciferase mRNA (Trilink) dissolved in pH~3 10 mM citrate buffer (ThermoFisher) through a single microfluidic mixing device as described in previous literature. The relative flow rates of the ethanol to aqueous phases w ere 1 :3 respectively, with a total flow- rate of 1.2 mL as dictated by displacement driven flow' using a syringe pump (Pump 33 DDS, Harvard Apparatus, Holliston, MA). Samples were collected in 20-kDa MWCO Slide- A-Lyzer™ dialysis cassettes (ThermoFisher) and dialyzed against lx PBS (ThermoFisher) for 1.5hr before continuing to analysis.

[0164] Female C57BL / 6 mice between six and eight weeks old (Jackson Labs, Bar Harbor, ME) were used for all animal experimentation. The temperature of the animal housing facility was 22±2 °C, held at a 12-hour dark / light cycle, with 40-70% air humidity. Randomly allocated mice for each group were injected intravenously (i.v.) at a dose of 0. Img mRNA / kg Attorney Docket No. 046483-7491W01(04073)

[0165] (2 pg per mouse, around lOOpl of volume through a tail vein injection). After 6 hr, mice were injected intraperitonealy (i.p.) with 200pl of D-luciferin potassium salt (ThermoFisher) at a concentration of 15mg / mL. Mice were sacrificed 10 min post i.p. injection, and their hearts, lungs, kidneys, livers, and spleens were dissected and imaged using a Lumina S3 in vivo imaging system (IVIS, PerkinElmer, Waltham, MA). Bioluminescence analysis was performed in Living Image 4.7.3 software (PerkinElmer).

[0166] Statistical analysis Statistical analyses and plotting were performed in Python using the scikit-leam or SciPy packages or in GraphPad Prism 10 with significance level a = 0.05. Data are represented as mean±s.d. One- way ANOVA followed by Tukey's HSD test or the LSD test followed by a Bonferroni correction were performed for multiple comparisons between groups.

[0167] Example 1: Integrated microfluidic LNP library generator leverages differentially encoded flow resistors across identical mixers

[0168] The LIBRIS chip comprises eight microfluidic LNP generators which mix three fluidic inputs (ii, i2, is) containing lipid components solubilized in ethanol and three fluidic inputs (i4, is, ie) containing RNA and citrate buffer (FIG. 2A). On the back side of the chip, a delivery7channel distributes each fluidic input to each of the eight generators on the front side of the chips using through-silicon vias (FIGs. 2A-2B). Upstream of each LNP generator, there are six unique lithographically defined resistors that control the relative flow rates of each of the six ethanol and aqueous inputs (FIG. 2A). First, the aqueous and ethanol inlets are separately mixed in two different staggered herringbone mixers (SHMs). After separate and complete mixing of both the aqueous and ethanol inputs, they combine to formulate LNPs in the final SHM (FIG. 2B).

[0169] The LIBRIS chip was designed using a lumped-element circuit model (FIGs. 2C-2D). Based on previous approaches, lithographically defined microfluidic flow resistors were incorporated upstream of each LNP generator (FIG. 2C), where the resistance of each resistor Ri,m is much greater than the resistance of the downstream SHMs such that the total resistance of each LNP generating device is approximately Ri,m, for each fluidic inlet i and LNP generator m. The fluidic resistance of each element in LIBRI8 was modeled, assuming a rectangular cross section of dimensions length L, width w, and height h (assuming here that h is the smallest dimension) using the relation. Atorney Docket No. 046483-7491W01(04073)

[0170] The resistance of each of the fluidic resistors Ri,mwas varied by varying the width of each resistor. Modifying the height across different devices would require many additional lithography steps and would be more sensitive to fabrication errors due to the strong nonlinear dependence of Rt.m to the limiting dimension h. The same length is maintained for each resistor to simplify connection of each generator to the delivery channels.

[0171] The chip architecture was modeled as a ladder geometry where the fluidic resistance of the delivery channel between each of the generators RDBI « Ri,m (FIG. 2D). Thus, we assume that the flow rate <Pi,mthrough each resistor is defined by the pressure applied to each inlet Pt divided by the resistance of each resistor Ri,m. This assumption was validated by simulating the circuit across the range of operating pressures and performing a pairwise distance analysis between the flow rates through each generator generated by the circuit simulation and by the values calculated using the assumptions above. A strong correlation of simulated and calculated relative flow rates were observed across all LNP generators (R2= 0.99) as well as simulated and actual relative flow rates (R2= 0.97). Thus, with an integrated pressure manifold, each inlet pressure P, can be rapidly modulated to vary the relative flow rate of the LNP components through the resistors Rtm to generate many distinct sets of eight formulations (FIG. 2E).

[0172] The LIBRIS chip was designed to maintain a constant total flow rate through each generator, even as the relative How rates of LNP components vary, allowing the lipid stoichiometry to change while preserving consistent mixing conditions (FIG. 2F). This function was achieved by using one of the aqueous and ethanol inputs to compensate for the flow rates from the other inputs are varied. The flow rates of each ethanol input (ii, i2, is) were designed to have a 1:3 ratio to their corresponding aqueous input (i-i, is, ie) to maintain the same ratio of IL from ii and mRNA from is across all LNP generators and to maintain the total flow rate ratio of ethanol and aqueous inputs at 1:3.

[0173] Example 2: Silicon and glass chip architecture permits highly precise featuring and chemical inertness at high pressures

[0174] The chip was fabricated by photolithographically paterning and then anisotropically etching four unique layers into a single 500 pm thick. 100 mm diameter silicon wafer that was anodically bond to glass wafers, adapting a previously disclosed fabrication strategy. Atorney Docket No. 046483-7491W01(04073)

[0175] Silicon and glass were chosen for their capacity to be patterned with sub-micron resolution, to be operated above 1000 psi. and to be compatible with a wide variety of solvents. Choosing to manufacture a device with features as small as those of the fluidic resistors in soft materials, such as PDMS, would result in uncertainty in flow rates due to channel deformation at typical operating pressures. After fabrication, chips were diced from finished wafers and then housed inside a custom aluminum manifold, machined by Protolabs™ with inlet ports and outlet ports spaced to fit a 96-well plate.

[0176] Example 3: Characterization of the LIBRIS chip reveals efficient and uniform mixing across all LNP generators and validates intended flow rate ratios

[0177] The mixing efficiency of each of the eight LNP generators were characterized and it was determined that both their mixing rates were the same and that the relative flow rates through each mixer were as designed (FIGs. 3A-3H). To perform this characterization, solutions of three w ater-soluble fluorescent dyes with minimal spectral overlap w ere used as each of the ethanol inputs (ii, i2, is) and the corresponding aqueous inputs ('14, is, ie). While operating the chip at the pressure setings later used to generate LNP libraries, the chip w as imaged using a fluorescence microscope, ensuring that intensity was within the linear dynamic range of the camera across all fluorescence channels (FIG. 3B).

[0178] To quantify the efficiency of mixing (H) for each device, a fluorescence micrograph was taken of each mixing portion, and three fluorescence channels were overlaid. To eliminate non-specific background, greyscale values of a brightfield image of the corresponding mixing portion w ere subtracted, and then the mean squared deviation of the dye intensity for each dye was integrated with respect to position y across the width w of the channel. The mixing efficiency was calculated from the fluorescence intensity values I at each pixel as follows:

[0179] Next, H was calculated for each of the three dyes used and report the arithmetic mean of all three values. When 1- H is ploted against the number of mixing cycles, the data fits an exponential decay model (FIGs. 2C-2E). Consistent with previous literature, a plateau in mixing efficiency was observed before the end of each of the SHM regions - 10 mixing cycles for the aqueous and ethanol SHMs and 15 mixing cycles for the LNP generating SHM. Atorney Docket No. 046483-7491W01(04073)

[0180] Sample mixing is considered complete at 1 - 77- 0.2. Within each group of ethanol, aqueous (FIGs. 3C-3D), and LNP generating SHMs (FIG. 3E). the mixing value of 1 - 77 < 0.2 was observed at the same mixing lengths, indicating that observed differences in LNP characteristics derive from differences in lipid ratios between generators, not differences in mixing efficiency.

[0181] From the same mixing experiments, the relative mixing ratios of the different LNP components through each LNP generator were evaluated. The resistances Rt.m through each of the three ethanol phase inputs i of the LIBRIS chip were chosen to follow the relation:

[0182] By doing so, the concentration of components can vary either directly or inversely proportionally across the eight generators depending on which inputs bear the components. This relation was qualitatively observed in the color gradient of the generator outputs across the chip, where the flow rates of green dye (<|>z.^, 1,4) and red dye ( 1,2, 1,5) flow rates trend inversely to the blue dye ( 13, (|> / ,<5) flow rate across the generators (FIG. 3B).

[0183] To quantify the relative ratios of each fluidic input in each mixer, the greyscale values of a brightfield image of each fluorescence micrograph were subtracted, the intensity of each dye relative to an unmixed portion of the device was normalized, and the intensity ratio of each dye between the fully mixed and unmixed portions of the device was calculated. Relative intensities were measured rather than laminar thicknesses due to the effects of boundary conditions on the fluidic velocity profde in a channel of rectangular cross section. To compare the difference between the predicted and actual flow rate ratios of the LIBRIS chip, flow rates through each mixing unit were normalized at each pressure seting and a pairwise distance analysis was performed on the relative flow rates through each input (FIG. 3F) of the calculated distribution (FIG. 2F and FIG. 3F) and actual distribution (FIG. 3G) as determined above, observing that the distributions of relative flow rates are nearly identical as compared by a Mantel test (R2= 0.96, P = 0.001) (FIG. 3H).

[0184] Example 4: Design and implementation of custom plate robotic interface allows for rapid collection of distinct LNPs to in-line dialysis

[0185] Though the LIBRIS chip rapidly generates many formulations from a single set of solutions simply by varying pressures, each of these outputs must be kept separate for further screening. Thus, an interface was designed and fabricated that coordinates the motion of a Attorney Docket No. 046483-7491W01(04073) custom plate robot with the control of a set of programmable pneumatic regulators, enabling disposal of undesired priming solution and collection of pure sample directly into on-plate dialysis cassettes once the chip has equilibrated (FIGs. 4A-4B).

[0186] The step-by-step process for LIBRIS to generate a set of distinct formulations is as follows: it initialize its flow under a new set of pressures, discards the output associated with this initialization, moves the plate to collect the eight outputs into individual dialysis cassettes, and collects -300 pL of each formulation, which takes around 22.5s or 2.75s per formulation (FIGs. 4C-4D). In this time per formulation, the integrated robotic system both equilibrates the chip and moves between each 4 x 2 array of dialysis cassettes to collect sets of eight samples. By integrating direct serial control of the regulators, the microcontroller initiates the pressure regulators in under 150 ms, and subsequent equilibration of the pressure regulators occurs within 1.5 s, as dictated by the proportional-integral-derivative (PID) controller. From the time of pressure initiation, stabilization of the fluidic lamina on chip occurs in 1.6 s, determined by analyzing the laminar flow versus time using inlets that contain fluorescent dye. To minimize the time it takes for the robot to move the chip to the next set of wells, motion of the plate from one position to another begins 600ms before the chip has finished equilibrating flow and enters the collection well at 2.75s (FIG. 4C). These exchange times between samples were achieved by minimizing on chip volume, minimizing tubing length and volumes, incorporating a silicon and glass architecture to improve the response time of the fluid flow to changes in pressure relative to mechanically softer devices made of soft polymers, and designing the plate robot to be as stiff as possible to improve its mechanical response time. The highest total dead volume of any output is -35 pL, comprising the volume of the chip outlets and outfitted needle. Thus, at a flow rate of 20pL / s, the dead volume of each output has been completely flushed between collections. The total collection time for 32 LNPs is -90s, with >10s of waste collection (FIG. 4D).

[0187] The precision of the positioning of the plate beneath the fixed LIBRIS chip was measured to ensure that no leaking or cross-contamination of the outputs would occur. Each chip output is connected to a needle which was positioned above the 2mm x 2.5mm aperture of a dialysis cassette situated in a 96-well plate (FIG. 4B). By covering each cassette with a disposable membrane, any backflow- or leakage from the cassette was prevented, and the sample was directly collected into dialysis. The precision of the plate positioning was measured by programming the robot to move through four collection positions 20 times in a row on the same 96-well plate and measuring the radius r of the puncture marks made by the collection needles (FIG. 4F). The diameter of the puncture marks across all cassettes Attorney Docket No. 046483-7491W01(04073) measured only 130 pm greater than the diameter of the needle, meaning that any centered needle is exceedingly unlikely to puncture outside of the acceptable radius of 1 mm from the center of the cassette (FIG. 4F). The plate positioning was further evaluated across multiple plates by measuring the puncture marks relative to a fixed position (FIG. 4G). No significant difference was observed across the variance of puncture marks at the same movement position across three plates (FIG. 4G).

[0188] Next, the consistency of total sample volume collection was evaluated between wells and the minimum collection volume possible to determine the versatility of the LIBRIS platform across volumes. Using 10s collection times, consistent sample volumes of 120pL were collected across all outputs, and the platform was capable of collecting a minimum volume of 10 pL per sample, enabling screening of physicochemical characteristics with minimal lipid and RNA use.

[0189] To improve the accessibility of this robotic system, a user-friendly graphical user interface (GUI) was designed which converts a set of input concentrations and a desired range of molar percentages for the LNP components into the correct input pressures. Since all parts except the motors and various structural components such as support rods and bearings (FIG. 4F), were designed and printed, the total cost for the single prototype robotic system is under $500 and would likely cost significantly less if produced at scale.

[0190] Example 5: Physicochemical characterization of 96 LNP library validates production of distinct LNPs

[0191] To validate the creation of LNPs with different physicochemical properties using the LIBRIS platform, a library of 96 LNPs was formulated. Each group is referred to by shorthand: IL-D-Lin-MC3-DMA (MC3), SM-102, and C12-200. MC3 and SM-102 were chosen given their clinical validation and to use C 12-200 given its status as a gold standard for transfection efficiency. The design space assayed during these experiments varies the molar percentages of IL and cholesterol in direct proportion, which both vary inversely with the concentration of phospholipid and PEG lipid (FIG. 5A). Using the LIBRIS chip, each of these lipids were formulated into a set of 32 LNPs, as determined by the stock concentrations and the relative flow rate ratios derived from the mixing experiments above. The hydrodynamic diameter (size) and poly dispersity index (PDI) of each sample was measured using dynamic light scattering, observing distinct sizes which correlate inversely with the percentage of PEG both within each set of eight LNPs and across all groups (FIG. 5B). Across all samples, only three formulations demonstrated a poly dispersity index (PDI) above Attorney Docket No. 046483-7491W01(04073)

[0192] 0.3 and only one formulation demonstrated an encapsulation efficiency (EE%) of lower than 70%. meaning that nearly all particles fall within acceptable physicochemical benchmarks (FIGs. 5C-5D). Unlike size, EE% and PDI did not correlate with the molar concentration of PEG.

[0193] To ensure that the differences in physicochemical characteristics between sets of eight LNPs were not simply an artifact of drift between samples, the chip was continuously operated to generate the same formulation 5 times. Across a representative sample of three SM-102 formulations, consistent particle size was measured between collections at the beginning and end of operation, indicating that particle characteristics do not meaningfully change within the operating time of the experiments performed. The physicochemical characteristics of LNPs generated were further compared using the SHM-based mixing technique against automated pipettes, meant to mimic conventional liquid handlers. It was demonstrate that using the LIBRIS chip, significantly smaller particles (~ 80nm vs. ~ 180nm) are generated with higher batch-to-batch consistency and lower, more consistent PDI than when using automated liquid handling. Additionally, it was shown that LNPs made using the LIBRIS platform compare in size and PDI to particles made using a control PDMS SHM, and a lOx parallelized SHM. Thus, as demonstrated herein, screening can be performed for an optimal LNP formulation at discovery' scale, then generate a formulation with identical properties using the same mixing architecture at market scale.

[0194] Example 6: In vitro and in vivo performance of LNPs from library validates differences observed in physicochemical characterization

[0195] The utility' of the LIBRIS chip was evaluated in the LNP screening process by measuring the transfection efficiency of particles across the previously described libraries (FIG. 6A). Using RNA concentrations determined during EE% characterization of the libraries, HepG2 cells were dosed w ith LNPs loaded with mCherry from the MC3, SM-102, and Cl 2-200 groups (FIG. 6A). Using flow cytometry', the percentage of mCherry+ cells and the mean fluorescence intensity (MFI) for n = 5 wells were determined, where a difference in LNP transfection efficiency contingent on formulation was observed (FIG. 6A). To determine which of the physicochemical characteristics had the most meaningful implications on transfection efficiency, a principal component analysis was performed, excluding the LNPs which failed to meet previous physicochemical benchmarks. It was observed that, excluding compositional variables, which varied in fixed increments between samples, either EE%, seen in the MC3 group, or size, seen in the SM-102 and Cl 2-200 groups, had the most significant Attorney Docket No. 046483-7491W01(04073) loading across the groups, a result generally consistent with the literature. To analyze the relationship between size, a variable with significant loading across all samples, and transfection efficiency, Pearson correlation analyses were performed between transfection efficiency and PEG lipid molar composition across each group, revealing that in the MC3 and SM-102 groups, PEG percentage and transfection efficiency correlate inversely, while in the C 12-200 group, transfection efficiency and PEG percentage correlate directly. To evaluate the relation between the in vitro performance of particles made using the chip and their in vivo performance, the in vivo hepatic transfection from the top performing LNP (Cl 2-200 F4) was quantified, the poorest performer (C12-200 F24), and the previously optimized C 12-200 formulation (FIG. 6B). Instead of formulating these using the LIBRIS chip, the results of the in vitro screen were cross-validated by determining the molar concentrations of each lipid corresponding to the formulations above (Table 1) then formulating each LNP using a single microfluidic mixing device as described previously. Separate cohorts of mice were treated intravenously (i.v.) with each of these formulations loaded with firefly luciferase mRNA at a total dose of 0. 1 mg / kg (FIG. 6C). Expression was observed almost entirely in the liver, consistent with previous literature (FIG. 6C). Total hepatic luminescence indicated around a two-fold in- crease in expression between F4 and F24, with F4 performing comparably to the previously optimized LNP formulation (FIG. 6C). Hepatic luciferase expression is reported here since it significantly outw eighed the expression of all other organs similarly across all groups. These results further validate the differences in composition generated in each library by the LIBRIS chip and suggest that there exist more biologically optimal LNPs discoverable by vary ing compositional ratios w ithin a single set of lipid structures.

[0196] Table 1. Physicochemical characteristics and molar compositions of LNPs tested in vivo

[0197] Enumerated Embodiments

[0198] The following exemplary' embodiments are provided, the numbering of which is not to be construed as designating levels of importance:

[0199] Embodiment 1 : microfluidic device comprising a parallel array of microfluidic channels, wherein each microfluidic channel independently comprises: Attorney Docket No. 046483-7491W01(04073)

[0200] (a) m occurrences of a first solution inlet, wherein each first solution inlet is in fluid communication with a first solution channel. wherein m is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10;

[0201] (b) n occurrences of a second solution inlet, wherein each second solution inlet is in fluid communication with a second solution channel, wherein n is 1, 2, 3, 4, 5, 6, 7, 8, 9. or 10;

[0202] (c) a first combined solution channel which is in fluid communication with each of the first solution channels;

[0203] (d) a second combined solution channel, which is in fluid communication with each of the second solution channels;

[0204] (e) a third combined solution channel, which is in fluid communication with the first combined solution channel and the second combined solution channel; and

[0205] (f) an outlet which is in fluid communication with the third combined solution channel; wherein each combined solution channel independently comprises at least one chaotic mixing region.

[0206] Embodiment 2: The device of Embodiment 1, wherein each first solution channel comprises at least one chaotic mixing region.

[0207] Embodiment 3: The device of Embodiment 1 or 2, wherein each second solution channel comprises at least one chaotic mixing region.

[0208] Embodiment 4: The device of any one of Embodiments 1 -3, wherein each first solution inlet is independently in fluid communication with a first solution source which is pneumatically regulated by a microcontroller.

[0209] Embodiment 5: The device of any one of Embodiments 1-4, wherein each second solution inlet is independently in fluid communication with a second solution source which is pneumatically regulated by a microcontroller.

[0210] Embodiment 6: The device of any one of Embodiments 1-5, w herein the device further comprises a collection plate beneath the parallel array of microfluidic channels and adjacent to the outlet.

[0211] Embodiment 7: The device of Embodiment 6, wherein the collection plate is regulated by the microcontroller.

[0212] Embodiment 8: The device of Embodiment 7, wherein the microcontroller is configured to move the collection plate in response to a change in pressure, optionally wherein the response occurs in less than about 100 milliseconds. Attorney Docket No. 046483-7491W01(04073)

[0213] Embodiment 9: The device of any one of Embodiments 1-6. wherein the parallel array of microfluidic channels comprise at least one photolithographically etched layer bonded to a borosilicate glass surface.

[0214] Embodiment 10: The device of any one of Embodiments 1-9, wherein m is 3.

[0215] Embodiment 11 : The device of any one of Embodiments 1-10, wherein n is 3.

[0216] Embodiment 12: The device of any one of Embodiments 1-1, wherein the device comprises an array of 8 parallel microfluidic channels.

[0217] Embodiment 13: The device of any one of Embodiments 2-12, wherein each occurrence of the first solution source independently comprises a lipid phase.

[0218] Embodiment 14: The device of Embodiment 13. wherein each lipid phase independently comprises at least one selected from the group consisting of an ionizable lipid, cholesterol and / or a modified derivative thereof, a neutral or helper lipid, a polymer conjugated lipid, and ethanol.

[0219] Embodiment 15: The device of Embodiment 13 or 14, wherein each lipid phase comprises a first lipid phase (zb), a second lipid phase (zb). and a third lipid phase (zb).

[0220] Embodiment 16: The device of Embodiment 15. wherein the first lipid phase (zz) comprises at least one ionizable lipid and cholesterol and / or a modified derivative thereof.

[0221] Embodiment 17: The device of Embodiment 15 or 16, wherein the second lipid phase (zb) comprises ethanol.

[0222] Embodiment 18: The device of any one of Embodiments 15-17. wherein the third lipid phase (is) comprises at least one neutral lipid or helper lipid and a polymer-conjugated lipid. Embodiment 19: The device of any one of Embodiments 1-18, wherein each occurrence of the second solution source independently comprises an aqueous phase.

[0223] Embodiment 20: The device of Embodiment 19. wherein each aqueous phase independently comprises at least one selected from the group consisting of a nucleic acid solution and a buffer solution.

[0224] Embodiment 21 : The device of Embodiment 19 or 20, wherein each aqueous phase comprises a first aqueous phase (zb), a second aqueous phase (zb), and optionally a third aqueous phase (is).

[0225] Embodiment 22: The device of Embodiment 21, wherein the first aqueous phase (zb) comprises a nucleic acid.

[0226] Embodiment 23 : The device of Embodiment 22, wherein the nucleic acid comprises DNA or RNA, optionally wherein the RNA comprises mRNA.

[0227] Embodiment 24: The device of any one of Embodiments 21-23, wherein the second Attorney Docket No. 046483-7491W01(04073) aqueous phase (z3) comprises a buffer solution, optionally wherein the second aqueous phase (zj) comprises citrate buffer, optionally wherein the citrate buffer comprises 10 mM citrate buffer at a pH of about 3.

[0228] Embodiment 25: The device of any one of Embodiments 21-24, wherein the third aqueous phase (ze) comprises a buffer solution, optionally wherein the third aqueous phase (i comprises citrate buffer, optionally wherein the citrate buffer comprises 10 mM citrate buffer at a pH of about 3.

[0229] Embodiment 26: The device of Embodiment 24 or 25, wherein the second aqueous phase (is) and third aqueous phase (A) are identical.

[0230] Embodiment 27: The device of any one of Embodiments 1-26, wherein the chaotic mixing region has an architecture selected from the group consisting of herringbone, ring, and serpentine.

[0231] Embodiment 28: A microfluidic processing apparatus comprising:

[0232] (a) the device of any one of Embodiments 1-27;

[0233] (b) a collection plate comprising a plurality of wells;

[0234] (c) a manipulator configured to hold the microfluidic device above the collection plate;

[0235] (d) a motion control apparatus operatively coupled to the manipulator and configured to position the microfluidic device relative to the collection plate such that each outlet of the device is aligned with a well of the collection plate;

[0236] (e) a controller configured to actuate the motion control apparatus to reposition the microfluidic device from a first configuration, in which an outlet of the device is aligned with a first well of the collection plate, to an alternate configuration, in which the outlet is aligned with a different well of the collection plate; and

[0237] (f) a fluid control system operatively coupled to the microfluidic device and configured to regulate pressure applied to each inlet of the microfluidic device.

[0238] Embodiment 29: The apparatus of Embodiment 28, wherein the collection plate comprises a microwell plate having an array of wells for receiving fluid dispensed from the microfluidic device.

[0239] Embodiment 30: The apparatus of Embodiment 28 or 29, wherein the controller is configured to adjust the position of the microfluidic device automatically (e.g., by a timed program) or responsive to one or more stimuli selected from the group consisting of user input, volume measurement, optical feedback, and pressure measurements.

[0240] Embodiment 31 : A method for preparing a plurality of lipid nanoparticles (LNPs), the Attorney Docket No. 046483-7491W01(04073) method comprising:

[0241] (a) providing the microfluidic processing apparatus of any one of Embodiments 28-30;

[0242] (b) actuating the motion control apparatus to position the microfluidic device such than each outlet of the device is aligned with an empty well of the collection plate;

[0243] (c) introducing a first solution into each first solution inlet of the microfluidic device to provide a combined first solution;

[0244] (d) introducing a second solution into each second solution inlet of the microfluidic device to provide a combined second solution;

[0245] (e) mixing the combined first solution and combined second solution to provide a formulation comprising a plurality of LNPs;

[0246] (f) dispensing at least a portion of the formulation comprising the plurality of LNPs from each outlet to each respective well of the collection plate; and

[0247] (g) repeating steps (c)-(f) after changing at least one parameter selected from the group consisting of:

[0248] (i) a flow rate through the microfluidic device;

[0249] (ii) an identity of at least one component of at least one of the first solution or the second solution; and

[0250] (iii) a concentration of at least one component of at least one of the first solution or the second solution.

[0251] Embodiment 32: The method of Embodiment 31, wherein each first solution independently comprises a lipid phase.

[0252] Embodiment 33: The method of Embodiment 32, wherein each lipid phase independently comprises at least one selected from the group consisting of an ionizable lipid, cholesterol and / or a modified derivative thereof, a neutral or helper lipid, a polymer conjugated lipid, and ethanol.

[0253] Embodiment 34: The method of Embodiment 33, wherein the ionizable lipid has a concentration ranging from about 0.1 mg / mL to about 10 mg / mL in the first solution, optionally wherein the ionizable lipid has a concentration of about 2.2 mg / mL in the first solution.

[0254] Embodiment 35: The method of Embodiment 33 or 34, wherein the cholesterol and / or modified derivative thereof has a concentration of about 0.01 mg / mL to about 5.0 mg / mL in the first solution, optionally wherein the cholesterol and / or modified derivative thereof has a concentration of about 0.6 mg / mL. Attomey Docket No. 046483-7491W01(04073)

[0255] Embodiment 36: The method of any one of Embodiments 33-35, wherein the neutral or helper lipid has a concentration of about 0. 1 mg / mL to about 10 mg / mL in the first solution, optionally wherein the neutral or helper lipid has a concentration of about 1.0 mg / mL in the first solution. Embodiment 37: The method of any one of Embodiments 31-36, wherein each second solution independently comprises an aqueous phase.

[0256] Embodiment 38: The method of Embodiment 37, wherein each aqueous phase independently comprises at least one selected from the group consisting of a nucleic acid solution and a buffer solution. The terms and expressions employed herein are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features show n and described or portions thereof, but it is recognized that various modifications are possible within the scope of the embodiments of the present application. Thus, it should be understood that although the present application describes specific embodiments and optional features, modification and variation of the compositions, methods, and concepts herein disclosed may be resorted to by those of ordinary skill in the art, and that such modifications and variations are considered to be within the scope of embodiments of the present application.

Claims

Attorney Docket No. 046483-7491W01(04073)CLAIMSWhat is claimed is:

1. A microfluidic device comprising a parallel array of microfluidic channels, wherein each microfluidic channel independently comprises:(a) m occurrences of a first solution inlet, wherein each first solution inlet is in fluid communication with a first solution channel, wherein m is 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10;(b) n occurrences of a second solution inlet, wherein each second solution inlet is in fluid communication with a second solution channel, wherein n is 1, 2, 3. 4, 5, 6, 7. 8, 9, or 10;(c) a first combined solution channel which is in fluid communication with each of the first solution channels;(d) a second combined solution channel, which is in fluid communication with each of the second solution channels;(e) a third combined solution channel, which is in fluid communication with the first combined solution channel and the second combined solution channel; and(f) an outlet which is in fluid communication with the third combined solution channel; wherein each combined solution channel independently comprises at least one chaotic mixing region.

2. The device of claim 1, wherein each first solution channel comprises at least one chaotic mixing region.

3. The device of claim 1 or 2, wherein each second solution channel comprises at least one chaotic mixing region.

4. The device of any one of claims 1-3, wherein each first solution inlet is independently in fluid communication with a first solution source which is pneumatically regulated by a microcontroller.Attomey Docket No. 046483-7491W01(04073)5. The device of any one of claims 1-4, wherein each second solution inlet is independently in fluid communication with a second solution source which is pneumatically regulated by a microcontroller.

6. The device of any one of claims 1-5, wherein the device further comprises a collection plate beneath the parallel array of microfluidic channels and adjacent to the outlet.

7. The device of claim 6, wherein the collection plate is regulated by the microcontroller.

8. The device of claim 7, wherein the microcontroller is configured to move the collection plate in response to a change in pressure, optionally wherein the response occurs in less than about 100 milliseconds.

9. The device of any one of claims 1-6. wherein the parallel array of microfluidic channels comprise at least one photolithographically etched layer bonded to a borosilicate glass surface.

10. The device of any one of claims 1-9, wherein m is 3.1 1. The device of any one of claims 1 -10, wherein n is 3.

12. The device of any one of claims 1-1, wherein the device comprises an array of 8 parallel microfluidic channels.

13. The device of any one of claims 2-12, wherein each occurrence of the first solution source independently comprises a lipid phase.

14. The device of claim 13, wherein each lipid phase independently comprises at least one selected from the group consisting of an ionizable lipid, cholesterol and / or a modified derivative thereof, a neutral or helper lipid, a polymer conjugated lipid, and ethanol.

15. The device of claim 13 or 14, wherein each lipid phase comprises a first lipid phase (z'z), a second lipid phase (z ), and a third lipid phase (is).Attorney Docket No. 046483-7491W01(04073)16. The device of claim 15, wherein the first lipid phase ( / / ) comprises at least one ionizable lipid and cholesterol and / or a modified derivative thereof.

17. The device of claim 15 or 16, wherein the second lipid phase (Z ) comprises ethanol.

18. The device of any one of claims 15-17, wherein the third lipid phase (z ) comprises at least one neutral lipid or helper lipid and a polymer-conjugated lipid.

19. The device of any one of claims 1-18, wherein each occurrence of the second solution source independently comprises an aqueous phase.

20. The device of claim 19, wherein each aqueous phase independently comprises at least one selected from the group consisting of a nucleic acid solution and a buffer solution.

21. The device of claim 19 or 20, wherein each aqueous phase comprises a first aqueous phase (14), a second aqueous phase (is), and optionally a third aqueous phase (is).

22. The device of claim 21, wherein the first aqueous phase ( ) comprises a nucleic acid.

23. The device of claim 22, wherein the nucleic acid comprises DNA or RNA, optionally wherein the RNA comprises mRNA.

24. The device of any one of claims 21-23, wherein the second aqueous phase (is) comprises a buffer solution, optionally wherein the second aqueous phase (is) comprises citrate buffer, optionally wherein the citrate buffer comprises 10 mM citrate buffer at a pH of about 3.

25. The device of any one of claims 21-24, wherein the third aqueous phase (is) comprises a buffer solution, optionally wherein the third aqueous phase (is) comprises citrate buffer, optionally wherein the citrate buffer comprises 10 mM citrate buffer at a pH of about 3.Attorney Docket No. 046483-7491W01(04073)26. The device of claim 24 or 25, wherein the second aqueous phase (is) and third aqueous phase (is) are identical.

27. The device of any one of claims 1-26, wherein the chaotic mixing region has an architecture selected from the group consisting of herringbone, ring, and serpentine.

28. A microfluidic processing apparatus comprising:(a) the device of any one of claims 1-27;(b) a collection plate comprising a plurality of wells;(c) a manipulator configured to hold the microfluidic device above the collection plate;(d) a motion control apparatus operatively coupled to the manipulator and configured to position the microfluidic device relative to the collection plate such that each outlet of the device is aligned with a well of the collection plate;(e) a controller configured to actuate the motion control apparatus to reposition the microfluidic device from a first configuration, in which an outlet of the device is aligned with a first well of the collection plate, to an alternate configuration, in which the outlet is aligned with a different well of the collection plate; and(f) a fluid control system operatively coupled to the microfluidic device and configured to regulate pressure applied to each inlet of the microfluidic device.

29. The apparatus of claim 28, wherein the collection plate comprises a microwell plate having an array of wells for receiving fluid dispensed from the microfluidic device.

30. The apparatus of claim 28 or 29, wherein the controller is configured to adjust the position of the microfluidic device automatically (e.g., by a timed program) or responsive to one or more stimuli selected from the group consisting of user input, volume measurement, optical feedback, and pressure measurements.

31. A method for preparing a plurality of lipid nanoparticles (LNPs), the method comprising:(a) providing the microfluidic processing apparatus of any one of claims 28-30;Attomey Docket No. 046483-7491W01(04073)(b) actuating the motion control apparatus to position the microfluidic device such than each outlet of the device is aligned with an empty well of the collection plate;(c) introducing a first solution into each first solution inlet of the microfluidic device to provide a combined first solution;(d) introducing a second solution into each second solution inlet of the microfluidic device to provide a combined second solution;(e) mixing the combined first solution and combined second solution to provide a formulation comprising a plurality of LNPs;(f) dispensing at least a portion of the formulation comprising the plurality of LNPs from each outlet to each respective well of the collection plate; and(g) repeating steps (c)-(f) after changing at least one parameter selected from the group consisting of:(i) a flow rate through the microfluidic device;(ii) an identity of at least one component of at least one of the first solution or the second solution; and(iii) a concentration of at least one component of at least one of the first solution or the second solution.

32. The method of claim 31. wherein each first solution independently comprises a lipid phase.

33. The method of claim 32, wherein each lipid phase independently comprises at least one selected from the group consisting of an ionizable lipid, cholesterol and / or a modified derivative thereof, a neutral or helper lipid, a polymer conjugated lipid, and ethanol.

34. The method of claim 33, wherein the ionizable lipid has a concentration ranging from about 0.1 mg / mL to about 10 mg / mL in the first solution, optionally wherein the ionizable lipid has a concentration of about 2.2 mg / mL in the first solution.

35. The method of claim 33 or 34, wherein the cholesterol and / or modified derivative thereof has a concentration of about 0.01 mg / mL to about 5.0 mg / mL in the first solution, optionally wherein the cholesterol and / or modified derivative thereof has a concentration of about 0.6 mg / mL.Attomey Docket No. 046483-7491W01(04073)36. The method of any one of claims 33-35, wherein the neutral or helper lipid has a concentration of about 0. 1 mg / mL to about 10 mg / mL in the first solution, optionally wherein the neutral or helper lipid has a concentration of about 1.0 mg / mL in the first solution.

37. The method of any one of claims 31-36, wherein each second solution independently comprises an aqueous phase.

38. The method of claim 37, wherein each aqueous phase independently comprises at least one selected from the group consisting of a nucleic acid solution and a buffer solution.

Images