WO2025232812A1

WO2025232812A1 - Biomolecule conjugated lipid nanoparticles

Info

Publication number: WO2025232812A1
Application number: PCT/CN2025/093310
Authority: WO
Inventors: Xi Zhu; Hong Chen; Jiaming Ren; Shin-Shay Tian; Xiaoping Zhao
Original assignee: Shanghai Vitalgen Biopharma Co Ltd
Current assignee: Shanghai Vitalgen Biopharma Co Ltd
Priority date: 2024-05-08
Filing date: 2025-05-08
Publication date: 2025-11-13
Anticipated expiration: 2026-11-08

Abstract

The present disclosure provides a biomolecule-conjugated lipid nanoparticle (LNP) and the method of preparing the biomolecule-conjugated LNP. Specifically, the present disclosure provides a nucleic acid delivery vector comprising a VHH-linker conjugate, wherein the VHH-linker conjugate comprises (a) a VHH, (b) a linker with a moiety Y and (c) a hinge, furthermore, the nucleic acid delivery vector comprises anchor-modified LNP, wherein the anchor-modified LNP comprises: (a) LNP; and (b) one or more anchor fragment, said anchor fragment comprises moiety X;wherein said moiety Y of the linker is capable of forming a linkage with said moiety X of the anchor fragment via bio-orthogonal click reaction.

Description

BIOMOLECULE CONJUGATED LIPID NANOPARTICLES

FIELD OF THE INVENTION

The disclosure belongs to the field of drug delivery. Specifically, the present application relates to a ligand-conjugated lipid nanoparticle (LNP) and the method of preparing the ligand-conjugated LNP.
SEQUNCE LISTING

The present disclosure includes a sequence listing as a part of the disclosure.

BACKGROUND OF THE INVENTION

mRNA has shown therapeutic potential in a range of applications, including viral vaccines, protein replacement therapies, cancer immunotherapies, cellular reprogramming and genome editing. To achieve therapeutic effects, mRNA molecules have to reach specific target cells and produce sufficient proteins of interest. However, targeted delivery and endosomal escape remain challenging for mRNA delivery systems.

Lipid nanoparticles as non-viral vectors are a key class of mRNA delivery system. They exhibit high delivery efficiency and low toxicity, which make them as the most commonly used non-viral vectors. Traditional four-component LNPs are limited to liver transfection which makes it difficult to reach other cells and organ tissues. Therefore, researchers are devoting time and effort to develop targeting strategies based on LNP. Compared to passive targeting lipid nanoparticles where drug accumulation is mediated via non-specific interaction with tissue biomacromolecules, active targeting relies on targeting ligands including small molecular ligands, carbohydrate, peptides, antibodies or aptamers which can specifically interact with receptors of target cells. Active targeting lipid nanoparticles demonstrated the strengths in multi-dimensional and multi-type precise targeting ranging from organ, cell to organelle-specific receptors. Depending on the physiochemical properties of these ligands, they can be conjugated to lipid nanoparticles via two approaches including 1) one-pot assembly of all structural lipids, helper lipids and targeting ligands; and 2) post-modification of targeting ligands into preformed lipid nanoparticles. Large molecular ligands, such as antibodies and aptamers which have high-order molecular structures and are fragile under the fabrication conditions, are usually introduced via post-modification into preformed lipid nanoparticles. A plain lipid nanoparticle is firstly prepared by the methods similar to one-pot assembly, while its surface is coated with functional groups (such as amine, carboxyl, NHS, Maleimide, DBCO, and N3) . Those functional groups can be introduced by one-pot assembly or post-insertion which is more complex in process.

Those lipid nanoparticles that can react with targeting ligands via organic reaction-based surface conjugation. WO2023287861A2 and WO2023056282A1 reported the primary routes of conjugation using thiol-based crosslinking (Maleimide-thiol, non-cleavable linkage and Pyridyl disulfide or S-S, cleavable linkage) . Although there has been great success to synthesize ligand-conjugated lipid nanoparticles, it remains as a challenge to produce those lipid nanoparticles especially with definable targetability, high quality and good translation reproducibility.

SUMMARY OF THE INVENTION

The present inventors developed a special site-specific conjugation method for ligand-conjugated LNP preparation. The method disclosed in the present application has at least of the following advantages: 1) allowing a one-step surface bioconjugation within a few hours, 2) being compatible and scalable with current LNP upstream and downstream manufacturing process, and/or 3) allowing site-specific conjugation of ligand.

In a first aspect, the present disclosure provides a delivery vector. Said vector comprises a ligand-linker conjugate.

In some embodiments, said ligand-linker conjugate comprises: (a) a ligand; (b) a linker comprising a moiety Y; and (c) a hinge. In further embodiments, said linker is connected to the ligand via a conjugation moiety. In further embodiments, said hinge comprises an amino acid sequence. In further embodiments, said hinge is positioned between the ligand and the linker.

In some embodiments, said vector is a nucleic acid delivery vector.

In some embodiments, said ligand is a VHH.

In some embodiments, said nucleic acid delivery vector is a ligand-conjugated lipid nanoparticle (LNP) . In some embodiments, said ligand-conjugated LNP comprises anchor-modified LNP. In some embodiments, said anchor-modified LNP comprises: (a) LNP; and (b) one or more anchor fragment, said anchor fragment comprises a moiety X. In some embodiments, said moiety Y of the linker is capable of forming a linkage with said moiety X of the anchor fragment via bio-orthogonal click reaction.

In some embodiments, said ligand is selected from the group consisting of Fab, scFv, VHH, or other active protein or peptide with a molecular mass under 50 kDa.

In some embodiments, said ligand is a VHH with a molecular mass of about 15kDa.

In some embodiments, said ligand is selected from the group consisting of scFv, VHH, or other active proteins or peptides with a molecular mass under 50 kDa. In some embodiments, said ligand is selected from the group consisting of VHH, or other active proteins or peptides with a molecular mass under 30 kDa. In some embodiments, said ligand is selected from the group consisting of VHH, or other active proteins or peptides with a molecular mass under 25 kDa. In some embodiments, the ligand is a nanobody (VHH) with a molecular mass of about 20 kDa. In some embodiments, the ligand is a nanobody (VHH) with a molecular mass of about 15 kDa.

In some embodiments, said VHH is modified with a sequence of the hinge followed by a conjugation-enabling moiety at the terminus.

In some embodiments, said hinge is a non-flexible hinge. In some embodiments, said hinge is a structural constraint hinge, a limited-mobile hinge or a controlled-flexible hinge. In some embodiments, the hinge is a rigid hinge or a restricted hinge.

In some embodiments, said hinge is not a G4S hinge.

In some embodiments, said hinge does not comprise a G4S motif.

In some embodiments, said hinge is a structurally constrained hinge comprising no more than 20%glycine (G) residues by amino acid composition.

In some embodiments, said hinge is a structurally constrained hinge comprising no more than 10%glycine (G) residues by amino acid composition.

In some embodiments, the controlled-flexibility or limited-mobility of the hinge is achieved through the following features:
The hinge comprises one or more rigidity-enhancing residue. Said rigidity-enhancing residue is
selected from the group consisting of proline (P) , tyrosine (Y) , phenylalanine (F) , tryptophan (W) , valine (V) , isoleucine (I) , and leucine (L) .

In further embodiments, said rigidity-enhancing residue is comprised in the hinge in a proportion of no less than 20%.

In some embodiments, the controlled-flexibility or limited-mobility of the hinge is achieved through the following features:
The hinge comprises one or more α-helix-stabilizing residue. Said α-helix-stabilizing residue is
selected from the group consisting of alanine (A) , leucine (L) , glutamate (E) , and methionine (M) .

In further embodiments, said α-helix-stabilizing residue is comprised in the hinge in a proportion of no less than 80%. Preferably, the proportion of α-helix-stabilizing residue is excluding α-helix disruptors like proline (P) and glycine (G) .

In some embodiments, the controlled-flexibility or limited-mobility of the hinge is achieved through the following features:
The hinge comprises one or more potential O-linked glycosylation sites.

In some embodiments, said potential O-linked glycosylation site comprises a serine (S) residue or a threonine (T) residue positioned adjacent to a proline (P) residue within the hinge. In some embodiments, said hinge comprises within a stretch of 10 contiguous amino acids at least two serine (S) and/or threonine (T) residues, each of which is located adjacent to a proline (P) residue. In some embodiments, said serine (S) and/or threonine (T) residues capable of O-linked glycosylation account for at least 20%, 30%, or 40%of the total amino acid residues within the hinge. In some embodiments, said hinge comprises a motif selected from SPSTPP (SEQ ID NO: 64) , PSTPPSP (SEQ ID NO: 65) , or other serine/threonine-rich motifs capable of O-glycosylation.

In some embodiments, the hinge comprises one or more hydrophilic residue. In some embodiments, said hydrophilic residue is selected from the group consisting of serine (S) , threonine (T) , asparagine (N) , glutamine (Q) , tyrosine (Y) , aspartic acid (D) , glutamic acid (E) , lysine (K) , arginine (R) , and histidine (H) .

In some embodiments, the structural composition of the hinge ensures backbone rigidity by restricting conformational freedom through steric constraints imposed by proline, aromatic residues, and branched-chain residues. In some embodiments, the combination of these rigidity-enhancing residues ensures structural stability, while the limitation on glycine content prevents excessive conformational flexibility and maintains the necessary balance between flexibility and rigidity. Alternatively, in some embodiments, the hinge contains strong α-helix-stabilizing residues, which enhance local rigidity by stabilizing the peptide backbone through hydrogen bonding.

In some embodiments, the hinge comprises 2 to 40 amino acids in length. In some embodiments, wherein the hinge comprises 5 to 25 amino acids in length. In some embodiments, the hinge does not comprise cysteine residues.

In some embodiments, the sequence of the hinge is computationally designed to optimize the balance between rigidity and solubility. In some embodiments, these sequences comprise rigidity-enhancing residues, such as proline (P) , alanine (A) , and leucine (L) , along with hydrophilic residues like glutamate (E) and lysine (K) to improve solubility. Examples of such sequence include KESGSVSSEQLAQFRSLD (SEQ ID NO: 4) , KEQPQVSSEQLAQFRPLD (SEQ ID NO: 13) , SPPRTSDPKNTP (SEQ ID NO: 14) , and EPPKRSTDNTPK (SEQ ID NO: 15) .

In some embodiments, the hinge forms a stable α-helical conformation, which enhances structural stability and conjugation efficiency. Suitable sequences of α-helical hinge include (EAAAK) n, where n is any integer between 2 and 6, the “EAAAK” is shown as SEQ ID NO: 59; and AEAAAKEAAAKA (SEQ ID NO: 6) .

In some embodiments, the sequence of the hinge is derived from naturally occurring human proteins. Such naturally occurring sequences are engineered to comprise 2 to 40 amino acids in length and to comprise at least 30%rigidity-enhancing residues, ensuring compatibility with site-specific conjugation methods.

In some embodiments, the sequence of hinge is derived from naturally occurring human proteins. In some embodiments, these sequences, found in inter-domain linkers of multi-domain proteins, provide the necessary conformation, flexibility, and stability for biological function. In some embodiments, these naturally occurring sequence is engineered to comprise 2 to 40 amino acids in length, and to comprise one or more of the following: (a) rigidity-enhancing residues, including P, Y, F, W, V, I, and L, at ≥20%, with G content <20%, (b) α-helix-stabilizing residues, including A, L, E, and M, at ≥80%, while excluding P and G to maintain structural stability and (c) potential O-linked glycosylation sites, including S and T positioned adjacent to P.

In some embodiments, the hinge is rich in proline. In some embodiments, the sequence of suitable proline-rich hinge is selected from the group consisting of (PA) n, (PPP) n, (PPG) n, where n is any integer between 2 and 15. Examples of such sequences include PAPAPAP (SEQ ID NO: 7) ; APAPAPAPAPAPAPA (SEQ ID NO: 8) ; VPPPPP (SEQ ID NO: 16) ; APGPPGPPG (SEQ ID NO: 17) ; PAPAPAPKE (SEQ ID NO: 9) ; APEKPPQPQPKEPP (SEQ ID NO: 10) ; APRRPPRPRPRRPP (SEQ ID NO: 11) ; KEPNQPPQPNPNQPD (SEQ ID NO: 12) .

In some embodiments, the sequence of the hinge is derived from an immunoglobulin hinge region while is truncated or engineered to exclude cysteine in the sequence. In some embodiments, the sequence of the hinge is derived from an upper or lower immunoglobulin hinge region with no cysteine. Suitable examples of such sequences include APELLGGP (SEQ ID NO: 18) , APEFLGGP (SEQ ID NO: 19) , APPVAGP (SEQ ID NO: 20) , and PAPELLGGPSVFLFPPKPKDTLMIS (SEQ ID NO: 21) . In some embodiments, said cysteine residue is linked to the ligand via a hinge derived from human IgA1, such as SPSTPPTPSPSTPP (SEQ ID NO: 1) .

In some embodiments, said hinge is a restricted hinge or a rigid hinge.

In some embodiments, said hinge comprises an amino acid sequence as described in any one of SEQ ID NO: 1, 4-12, 13-21 and 59-62.

In some embodiments, said cysteine residue is linked with said ligand via a rigid hinge of human IgA1.

In some embodiments, said hinge region comprises an amino acid sequence as described in SEQ ID NO: 1.

In some embodiments, said conjugation-enabling moiety comprises a cysteine residue, a peptide tag for enzymatic conjugation, or an unnatural amino acid bearing a bio-orthogonal reactive group.

In some embodiments, said conjugation-enabling moiety is a cysteine residue.

To facilitate controlled site-specific conjugation, in some embodiments, the C-terminal or N-terminal of the ligand is modified with a cysteine residue, and a hinge is inserted between the ligand and cysteine residue.

In some embodiments, said cysteine residue is reduced to provide free thiol group on its C-terminal.

In some embodiments, the cysteine residue is chemically reduced to generate ligand monomer with a free thiol (-SH) group, which is then site-specifically conjugated to a linker via a thiol-reactive functional group. In some embodiments, the thiol-reactive functional group is a maleimide group, a parafluoro group, an ene group, an yne group, a vinylsulfone group, a pyridyl disulfide group, a thiosulfonate group, and a thiol-bisulfone group. In some embodiments, the thiol-reactive functional group is a maleimide group.

In some embodiments, the hinge inserted between the ligand and cysteine residue is structurally constrained to provide controlled flexibility while maintaining a defined conformation. In some embodiments, the hinge exhibits a degree of rigidity that enhances structural stability and minimizes aggregation or dimerization post-conjugation.

In some embodiments, the hinge does not contain any cysteine residue to eliminate unintended disulfide bonding or uncontrolled polymerization.

In some embodiments, said ligand is modified with cysteine residue.

In some embodiments, said cysteine residue is linked with said ligand via a hinge.

In some embodiments, said ligand is modified with cysteine residue, which is linked with said ligand via a hinge.

The linker may be designed with different structural configurations to optimize conjugation efficiency and functionality. In preferred embodiments, a linker suitable for this invention can be a DBCO derivative, a BCN derivative, a TCO derivative, an azide derivative, or a tetrazine derivative. In specific embodiments, the linker has a structure as shown in the following table:

In some embodiments, the formation of said ligand-conjugated LNP is via bio-orthogonal click
reaction from said moiety X and said moiety Y.

In some embodiments, said anchor fragment comprises polyethylene glycol (PEG) amphiphilic polymer, polyethylene glycol (PEG) conjugated lipid (also known as PEG lipid) and/or polyethylene glycol (PEG) conjugated hydrophobic polymer.

In some embodiments, said anchor fragment comprises a lipid or a hydrophobic polymer linked to said moiety X via a PEG linker having a formula of – (OCH₂CH₂) _n –, wherein n is any integer ranging from 0 to 135, preferably 22 to 117, more preferably 45 to 90.

In some embodiments, said anchor fragment is a clickable polyethylene glycol (PEG) amphiphilic polymer, a clickable polyethylene glycol (PEG) conjugated lipid or a clickable polyethylene glycol (PEG) conjugated hydrophobic polymer.

In some embodiments, said anchor-modified LNP has said moiety X located on the outer surface of the LNP.

In some embodiments, said anchor fragment does not individually generate micelles in said self-assembly.

In some embodiments, said lipid of the anchor fragment include one or more lipid, the carbon number is more than 14 for each lipid. In some embodiments, said lipid of the anchor fragment is selected from the group consisting of di-stearoyl-phosphatidyl-ethanolamine (DSPE) , diphosphatidylglycerol (DPG) , dipalmitoyl phosphatidyl ethanolamine (DPPE) , distearoyl-rac-glycero (DSG) and cholesterol.

In some embodiments, said hydrophobic polymer of the anchor fragment is poly (lactic-co-glycolic acid) or poly (lactic acid) .

In some embodiments, said anchor fragment comprises the formula selected from:

wherein n ranges from 0 to 135, preferably 22 to 117, more preferably 45 to 90.

In some embodiments, the components constituting LNP comprise ionizable lipids, neutral lipids, steroids and PEG lipids.

In some embodiments, said neutral lipid is distearoylphosphatidylcholine (DSPC) or dioleoyl-phosphatidylethanolamine (DOPE) .

In some embodiments, said steroid is cholesterol, sitosterol or stigmasterol.

In some embodiments, said PEG lipid is 1, 2-Distearoyl-sn-glycero-3-phosphoethanolamine-Poly (ethylene glycol) (DSPE-PEG) , 1, 2-Distearoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DSG-PEG) , 1, 2-Dipalmitoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DPG-PEG) , 1, 2-Dipalmitoyl-sn-glycero-3-phosphoethanolamine-poly (ethylene glycol) (DPPE-PEG) , or 1, 2-Dimyristoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DMG-PEG) .

In some embodiments, said PEG lipid is 1, 2-distearoyl-sn-glycero-3-phosphoethanolamine-poly (ethylene glycol) (DSPE-PEG) .

In some embodiments, said ionizable lipid has formula (I) , or a salt, tautomer, or stereoisomer thereof,

wherein:
m and p are independently selected from any integer ranging from 3 to 8;
n is selected from any integer ranging from 2 to 4;
X is a bond, -C (O) O-, -OC (O) -, -OC (O) O-, or a biodegradable group;
R₁ is a hydrogen bond donor-containing group or hydrogen bond acceptor-containing group;
both of R₂ are same and selected from C₁-C₁₂ alkyl, substituted C₁-C₁₂ alkyl, C₂-C₁₂ alkenyl, substituted C₂-C₁₂ alkenyl, C₃-C₁₂ cycloalkyl and substituted C₃-C₁₂ cycloalkyl and combinations of thereof;
R₃ is selected from C₄-C₂₂ alkyl, substituted C₄-C₂₂ alkyl, C₄-C₂₂ alkenyl, substituted C₄-C₂₂
alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof; or R₃ is an acetal group ofwherein both of R₄ are same and selected from C₁-C₁₆ alkyl, substituted C₁-C₁₆ alkyl, C₂-C₁₆ alkenyl, substituted C₂-C₁₆ alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof.

In some embodiments, said ionizable lipid has formula (II) , or a salt, tautomer, or stereoisomer thereof,

wherein
q is selected from any integer ranging from 2 to 4.

In some embodiments, R₁ of said ionizable lipid is selected from hydroxyalkyl group having 1 to 5 carbon atoms, or optionally substituted amino alkylenyl group having 1 to 6 carbon atoms.

In some embodiments, R₁ of said ionizable lipid is selected from one of the following formulae:

wherein o is selected from 1, 2, 3, 4, and 5.

In some embodiments, R₂ of said ionizable lipid is selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl and C₂-C₁₂ alkenyl and combinations of thereof.

In some embodiments, R₂ is C₃-C₁₀ alkyl.

In some embodiments, R₃ of said formula (I) is selected from C₆-C₁₂ alkyl, C₆-C₁₂ cycloalkyl or C₆-C₁₂ alkenyl and combinations of thereof.

In some embodiments, R₃ is the same with R₂.

In some embodiments: a) each R₂ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains; and/or b) R₃ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains.

In some embodiments, R₂ and R₄ of said formula (II) are independently selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl and C₂-C₁₂ alkenyl and combinations of thereof.

In some embodiments, R₂ and R₄ are independently C₃-C₁₀ alkyl.

In some embodiments, R₄ is the same with R₂.

In some embodiments: a) each R₂ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains; and/or b) each R₄ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains.

In some embodiments, said carbon atom with hydrogen atom (s) substituted by one or two side chains is the second or more distant carbon atom counting from the junction.

In some embodiments, said one or two side chains are C₁-C₄ alkyls.

In some embodiments, each R₂, R₃ and/or each R₄ is independently selected from one of the following formulas:

In some embodiments, said ionizable lipid is selected from the group consisting of Compound Nos. : 002-011 and 013-103 as shown in Table 1.

Table 1. Chemical Structures of Representative Compounds

In some embodiments, said ionizable lipid is comprised in an amount of about 20-90 mol%of the total lipid content, preferably 20-70 mol%, 30-60 mol%, or 45-55 mol%. In some embodiments, the ratio of total lipid content to nucleic acid is also adjustable to achieve a nitrogen-to-phosphate (N/P) ratio ranging from about 3 to 10 or higher.

In some embodiments, said ionizable lipid is selected from the group consisting of alkylated amines, imidazolium-based lipids, guanidinium-functionalized lipids, piperazine-based lipids, and ester-or amide-linked lipids. In some embodiments, said ionizable lipids generally contain tertiary or quaternary amine groups that remain neutral at physiological pH (～7.4) but become protonated in acidic endosomal environments. In some embodiments, alkylated amine-based lipids include DLin-MC3-DMA, ALC-0315, and SM-102. Additionally, ester-linked lipids and amide-linked lipids introduce biodegradability to said LNP, ensuring controlled metabolism and clearance while maintaining high nucleic acid delivery efficiency.

In some embodiments, said ionizable lipid is comprised in an amount of about 40-60 mol%. In some embodiments, said neutral lipid is comprised in an amount of about 5-20 mol%. In some embodiments, said steroid is comprised in an amount of about 25-50 mol%. In some embodiments, said PEG lipid is comprised in an amount of about 0.1-5 mol%. In some embodiments, said anchor fragment is comprised in an amount of about 0.1-1 mol%. In the meanwhile, the components are comprised in said LNP of (a) and said anchor fragments of (b) totally as an amount of 100 mol%.

In some embodiments, said ionizable lipid is comprised in an amount of about 45-55 mol%. In some embodiments, said neutral lipid is comprised in an amount of about 7.5-15 mol%. In some embodiments, said steroid is comprised in an amount of about 35-45 mol%. In some embodiments, said PEG lipid is comprised in an amount of about 0.5-2 mol%. In some embodiments, and said anchor fragment is comprised in an amount of about 0.2-0.9 mol%. In the meanwhile, the components are comprised in said LNP of (a) and said anchor fragments of (b) totally as an amount of 100 mol%.

In some embodiments, said ionizable lipid is comprised in an amount of about 40-60 mol%, said neutral lipid is comprised in an amount of about 5-20 mol%, said steroid is comprised in an amount of about 25-50 mol%, said PEG lipid is comprised in an amount of about 0.1-5 mol%, and said anchor fragment is comprised in an amount of about 0.1-1 mol%, when taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.

In some embodiments, said ionizable lipid is comprised in an amount of about 50 mol%, said neutral lipid is comprised in an amount of about 10 mol%, said steroid is comprised in an amount of about 38-39.5 mol%, said PEG lipid is comprised in an amount of about 0.25-1.75 mol%, and said anchor fragment is comprised in an amount of about 0.25-1 mol%, when taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.

In some embodiments, said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 0.5-3 mol%, taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.

In some embodiments, said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 1 mol%, wherein the PEG lipid characterized in anchor fragment is equivalent to the PEG lipid characterized in LNP.

In some embodiments, said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 1 mol%, wherein the PEG lipid characterized in LNP is comprised in an amount of 0.

In some embodiments, said ionizable lipid is comprised in an amount of about 50 mol%. In some embodiments, said neutral lipid is comprised in an amount of about 10 mol%. In some embodiments, said steroid is comprised in an amount of about 38-39.5 mol%. In some embodiments, said PEG lipid is comprised in an amount of about 0.25-1.75 mol%. In some embodiments, and said anchor fragment is comprised in an amount of about 0.25-1 mol%. In the meanwhile, the components are comprised in said LNP of (a) and said anchor fragments of (b) totally as an amount of 100 mol%. Herein the amount of PEG lipid is the total amount of PEG lipid in the components constituting said LNP of (a) and said anchor fragments of (b) , thus including clickable PEG lipid (anchor PEG, PEG lipid characterized in anchor fragment) and PEG lipid without a clickable moiety (lipid PEG, PEG lipid characterized in LNP) . In some embodiments, about 5%, about 10%, about 15%, about 20%, about 25%, about 30%, about 35%, about 40%, about 45%, about 50%, about 60%, about 70%, about 80%, about 90%, or almost 100%of the total amount of PEG lipid is anchor PEG. In preferred embodiments, about 25%-50%of the total amount of PEG lipid is anchor PEG.

In some embodiments, said moiety X and said moiety Y are selected from the group consisting of the following click chemistry reactive partners: 1) azide and dibenzocyclooctyne (DBCO) , 2) azide and 4-dibenzocyclooctynol (DIBO) , 3) azide and biarylazacyclooctynone (BARAC) , 4) azide and alkyne, 5) tetrazine and trans-cyclooctene (TCO) , 6) tetrazine and cyclopropane, or 7) azide and bicyclononyne (BCN) , wherein said click chemistry reactive partners of each group are interchangeable between said moiety X and moiety Y.

In some embodiments, the molar ratio of the VHH-linker conjugate, defined by the amount of moiety Y present in the conjugate, relative to the total lipid content in the lipid nanoparticle formulation, is between 0.03 mol%and 0.5 mol%.

In some embodiments, the components constituting LNP comprises nucleic acid molecules.

In some embodiments, said nucleic acid molecules are RNAs.

In some embodiments, said nucleic acid molecules are selected from the group consisting of messenger RNA (mRNA) , guide RNA (gRNA) , a short interfering RNA (siRNA) , an RNA interference (RNAi) molecule, a microRNA (miRNA) , an antagomir, an antisense RNA, a ribozyme, a small hairpin RNA (shRNA) , or a mixture thereof.

In some embodiments, said RNAs are encapsulated in said anchor-modified LNPs.

In some embodiments, additional amino acid residues may be incorporated after the cysteine residue, specifically at the C-terminus of the cysteine residue, to further optimize conjugation efficiency and steric accessibility. For instance, in some embodiments, between 1 and 5 alanine residues may be added.

In some embodiments, a tag sequence can be inserted within the ligand to facilitate purification and characterization of the ligand. The tag may be positioned between the hinge and the original ligand, between the hinge and the terminal cysteine residue, or at other suitable locations within the ligand construct. In some embodiments, the tag sequence may comprise a polyhistidine tag (His-tag) , commonly consisting of six to ten histidine residues, which allows for affinity purification using nickel (Ni²⁺) or cobalt (Co²⁺) chelate chromatography. In other embodiments, the tag sequence may include other affinity purification motifs such as Strep-tag, FLAG-tag, HA-tag, or Avi-tag, which provide additional versatility in ligand purification and detection. In other embodiments, a short flexible sequence such as Gly-Ser repeats or short rigid spacers may be introduced adjacent to the tag sequence to ensure proper folding and accessibility of the ligand.

In some embodiments, the molar ratio of said moiety Y to the total lipid ranges from 0.062-0.5 mol%.

In a second aspect, the present disclosure provides a method of preparing ligand-conjugated lipid nanoparticle (LNP) or the first aspect, comprising: 1) providing (a) components constituting said LNP and (b) anchor fragments, thereby allows self-assembly of an anchor-modified LNP under appropriate conditions, wherein each of said anchor fragment comprises a moiety X; and 2) providing a ligand-linker conjugate, wherein said linker comprises a moiety Y, wherein said moiety Y is capable of forming a linkage with said moiety X of 1) via bio-orthogonal click reaction.

In some embodiments, said method further comprises: 3) contacting said anchor-modified LNPs of 1) with said ligand-linker conjugate of 2) , thereby allows the formation of said ligand-conjugated lipid nanoparticle (LNP) .

In further embodiments, said ligand is a VHH.

In some embodiments, said ligand is conjugated to said linker via a site-specific reaction between said linker and the conjugation-enabling moiety positioned downstream of the hinge.

In some embodiments, said site-specific reaction is selected from: a thiol–maleimide reaction, a thiol–vinylsulfone reaction, a thiol–para-fluorophenyl reaction, an enzymatic ligation reaction, or a bio-orthogonal click reaction involving a noncanonical amino acid.

In some embodiments, said conjugation-enabling moiety is a cysteine residue and said linker comprises a maleimide group that reacts with the thiol group of said cysteine residue.

In some embodiments, said bio-orthogonal click reaction is selected from a group consisting of nucleophilic ring-opening reactions, cycloaddition reactions, nucleophilic addition reactions, thiol-ene reactions, and Diels Alder reactions.

In some embodiments, the procedure 1) of said method further comprises providing nucleic acid molecules while providing the components constituting said LNP and said anchor fragments.

In some embodiments, said method further comprises: 4) purification of said ligand-conjugated LNP, wherein said purification allows removal of said ligand-linker conjugates of 2) that fail to contact with or form a linkage with said anchor-modified LNPs.

In some embodiments, said purification comprises the use of filtration device with filter core size smaller than 50 KDa～300 KDa.

In a third aspect, the present disclosure provides a ligand-linker conjugate, wherein the linker is designed to facilitate site-specific conjugation between a cysteine-modified ligand and a functionalized nanoparticle surface. The linker covalently bridges the ligand to the nanoparticle. In some embodiments, the ligand-linker conjugate is synthesized through reactions involving thiol-based chemistries, such as thiol-maleimide reaction, thiol-parafluoro reaction, thiol-ene reaction, thiol-yne reaction, thiol-vinylsulfone reaction, thiol-pyridyl disulfide reaction, thiol-thiosulfonate reaction, and thiol-bisulfone reaction. A schematic of ligand-linker conjugate construct is depicted in FIG. 1A.

In some embodiments, the linker comprises two key functional regions: (a) a thiol-reactive functional group, which enables covalent conjugation with the free thiol (-SH) of the ligand; (b) one or more moiety Y, which serves as a reactive click handle for bio-orthogonal click reaction with functionalized nanoparticles.

In some embodiments, the linker comprises one thiol-reactive functional group and one moiety Y. In other embodiments, the linker comprises one thiol-reactive functional group and two or more moieties Y, allowing for increased conjugation valency.

In some embodiments, the linker further includes a spacer between the thiol-reactive functional group and the moiety Y to modulate steric effects, enhance solubility, and improve linker flexibility. In some embodiments, the spacer is an alkyl chain with a carbon length ranging from C₂ to C₃₀, which may be saturated or unsaturated. In some embodiments, the spacer comprises a polyethylene glycol (PEG) _n chain, where n ethylene glycol (EG) units are incorporated, and n is any integer between 2 and 40. In some embodiments, the spacer comprises a polyglycerol, a polyoxazoline, or a poly (sarcosine) , with a degree of polymerization (DP, n) ranging from 2 to 40, wherein n is an integer. In some embodiments, the spacer region enhances hydrophilicity, minimizes aggregation and steric hindrance during conjugation.

In some embodiments, the linker may incorporate anionic or cationic functional groups to enhance its hydrophilicity, charge properties, and/or bio-compatibility. These charged groups may improve solubility, reduce aggregation, and/or optimize ligand conjugation efficiency in aqueous environments. In some embodiments, the linker structure may contain anionic groups such as Sulfonate (-SO₃ ^-) , Carboxylate (-COO^-) , Phosphate (-PO₄ ^2-) . In some embodiments, the linker may include cationic groups such as Ammonium (-NH₄ ⁺, -NR₃ ⁺) , Guanidinium (-C (=NH) NH₂ ⁺) , Imidazolium (-C₃H₄N₂ ⁺) . In some embodiments, charged functional groups can be covalently incorporated into the linker backbone or attached as terminal modifications, depending on the desired solubility and interaction properties.

In a fourth aspect, the present disclosure provides a targeted lipid nanoparticle (tLNP) comprising: a cationic lipid described herein, particularly as an ionizable cationic lipid; a neutral lipid, such as distearoylphosphatidylcholine (DSPC) or dioleoyl-phosphatidylethanolamine (DOPE) ; a steroid and analog thereof, such as cholesterol, sitosterol or stigmasterol; a hydrophilic polymer conjugated lipid, such as a PEG lipid, and an anchor fragment, such as a N3-PEG lipid; a ligand as described herein which have conjugated to the LNP surface through the click reaction of the linker described herein with said anchor fragment.

In some embodiments, the tLNP is loaded with one or more biologically active molecules, such as one or more nucleic acid molecules. In some embodiments, said nucleic acid molecule is therapeutic or prophylactic nucleic acid molecule.

In a fifth aspect, the present disclosure provides a pharmaceutical composition comprising the ligand-conjugated LNP of the first aspect, the ligand-conjugated LNP obtained by the method described herein (in the second aspect) , or the tLNP of the fourth aspect.

In a sixth aspect, the present disclosure provides a method of delivering a biologically active molecule to a cell of a subject, comprising administering the ligand-conjugated LNP of the first aspect or the ligand-conjugated LNP obtained by the method described herein (in the second aspect) .

In a seventh aspect, the present disclosure provides a method of delivering a biologically active molecule to a cell of a subject, comprising administering the ligand-conjugated LNP of the first aspect or the tLNP of the fourth aspect loaded with said biologically active molecule to the subject.

In an eighth aspect, the present disclosure provides a method of treating or preventing a disease or disorder in a subject in need thereof, comprising administering the ligand-conjugated LNP of the first aspect, a pharmaceutically effective amount of the ligand-conjugated LNP obtained by the method described herein (in the second aspect) , or the tLNP of the fourth aspect loaded with a therapeutic or prophylactic biologically active molecule to the subject.

In some embodiments, said molecule is a nucleic acid molecule. In some embodiments, said nucleic acid molecule is therapeutic or prophylactic nucleic acid molecule.

In a ninth aspect, the present disclosure provides use of the ligand-conjugated LNP of the first aspect, the ligand-conjugated LNP obtained by the method described herein (in the second aspect) , or the tLNP of the fourth aspect, or the pharmaceutical composition of the fifth aspect to treat and/or prevent a disease.

In a tenth aspect, the present disclosure provides use of the ligand-conjugated LNP of the first aspect, the ligand-conjugated LNP obtained by the method described herein (in the second aspect) , or the tLNP of the fourth aspect, or the pharmaceutical composition of the fifth aspect in the manufacture of a medicament.

Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrative embodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.
INCORPORATION BY REFERENCE

All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.
BRIEF DESCRIPTION OF THE DRAWING

The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are employed, and the accompanying drawings (also “figure” and “FIG. ” herein) , of which:

Figure 1A illustrates the structural configuration of the VHH conjugate. The VHH construct contains a C-terminal cysteine residue (Cys) positioned adjacent to a hinge region and His-tag. The VHH construct contains a bifunctional linker molecule, comprising maleimide (Mal) and click chemistry handles interconnected by a spacer domain, the linker facilitates conjugation. The VHH-linker conjugate is formed through mal-thiol reaction between the sulfhydryl group (-SH) of the VHH's cysteine residue and the maleimide moiety on the linker.

Figure 1B illustrates LC-MS data of anti-CD5 VHH with intact His-tag and hinge region (upper panel) and anti-CD5 VHH-linker conjugate (down panel) .

Figure 2 illustrates schematic illustration of targeted-LNP production via bio-orthogonal click reaction between the anchor on LNP surface and the VHH-linker conjugate.

Figure 3A illustrates transfection efficiency of tLNPs with different formulations by median fluorescence intensity (MFI) of Jurkat cells and HepG2 cells following ex vivo treatment (1μg per 2×10⁵ cells, 6 h) with LNPs of distinct formulations. The molar ratios of total PEG-lipid to total lipid and anchor PEG-lipid to total PEG-lipid of different formulations were A (0.5%, 100%) , B (1.0%, 100%) , C (1%, 50%) , D (1.0%, 25%) , E (1.5%, 50%) , F (1.5%, 25%) , respectively.

Figure 3B illustrates transfection efficiency of tLNPs with different formulations by frequency of GFP⁺ cells in CD3⁺ T cells and CD14⁺ monocytes isolated from PBMCs treated ex vivo under identical conditions. The molar ratios of total PEG-lipid to total lipid and anchor PEG-lipid to total PEG-lipid of different formulations were A (0.5%, 100%) , B (1.0%, 100%) , C (1%, 50%) , D (1.0%, 25%) , E (1.5%, 50%) , F (1.5%, 25%) , respectively.

Figure 4A illustrates impact of ligand density on tLNP targeting transfection efficiency by MFI of Jurkat cells and HepG2 cells following ex vivo treatment with tLNPs containing variable VHH-linker conjugate loads (1μg per 2×10⁵ cells, 6 h) . For those tLNP, the molar ratios of total PEG-lipid to total lipid ranged from 0.5 to 1.5%, and the molar ratios of anchor PEG-lipid to total PEG-lipid ranged from 25%to 100%.

Figure 4B illustrates impact of ligand density on tLNP targeting transfection efficiency by frequency of GFP⁺ cell in CD3⁺T cells and CD14⁺ monocyte cells isolated from PBMCs treated under identical conditions. For those tLNP, the molar ratios of total PEG-lipid to total lipid ranged from 0.5 to 1.5%, and the molar ratios of anchor PEG-lipid to total PEG-lipid ranged from 25%to 100%.

Figure 5 illustrates transfection efficiency of tLNPs fabricated via linker free method with varied PEG-lipid formulations by frequency of GFP⁺ Jurkat cells and HepG2 cells following ex vivo treatment with tLNPs (1μg per 2×10⁵ cells, 6 h) . Those tLNP were prepared by conjugating maleimide-functionalized LNPs (Mal-PEG-lipid) with cysteine-terminated VHHs (His-tag and hinge-containing constructs) through mal-thiol chemistry. The molar ratios of total PEG-lipid to total lipid and mal-PEG-lipid to total PEG-lipid were G (2%, 50%) , H (2%, 25%) , C (1%, 50%) , D (1.0%, 25%) , respectively.

Figure 6 illustrates transfection robustness comparison of tLNP prepared via distinct conjugation strategies by mean GFP⁺ cell frequency in Jurkat cells and HepG2 cells following ex vivo treatment with tLNPs (1μg per 2×10⁵ cells, 6 h) . The linker-free VHH group: VHH were conjugated to the surface of LNPs via mal-thiol reaction between the cysteine of VHH and Mal-PEG-lipid of LNP. The VHH-linker group: VHH were conjugated to the surface of LNPs via the click reaction between DBCO of VHH-linker conjugate and N₃-PEG-lipid of LNP. The molar ratios of total PEG-lipid to total lipid and mal or N3-PEG-lipid to total PEG were 1%and 50%, respectively.

Figure 7 illustrates in vivo transfection of anti-CD5 tLNP in hPBMC mice. Mice were injected with LNP (166, traditional liver-targeting formulation) , and tLNPs (176-179) at a dose of 2 mg/kg. 24 hours post i. v. administration, blood, spleen and liver cells were isolated and stained for a set of antibodies (anti-human CD45, anti-human CD3) . The percent of tdtomato⁺ cells were calculated from both CD45⁺CD3⁺ gated populations and CD3^-gate populations.

Figure 8 illustrates in vivo transfection of anti-CD5 tLNP in hCD5 mice. Mice were injected with LNP (263, traditional liver-targeting formulation) , and tLNPs (264, 265) respectively, at a dose of 2 mg/kg. 24 hours post i. v. administration, blood, spleen and liver cells were isolated and stained for a set of antibodies (anti-mouse CD45, anti-mouse CD3) . The percent of tdtomato⁺ cells were calculated from CD45⁺CD3⁺ gated populations and CD3^-gate populations, and CD45^-gate populations. The molar ratios of total PEG-lipid to total lipid of 264 and 265 were 1.5 and 1.0%, respectively. For both of them, the molar ratios of N3-PEG-lipid to total PEG were 50%, and the molar ratios of VHH to total lipid were 0.125%.

Figure 9 illustrates the storage stability of tLNP by the mean GFP⁺ cell frequency in Jurkat cells and HepG2 following ex vivo treatment with fresh and stored tLNPs (1μg per 2×10⁵ cells, 6 h) . tLNP were stored for 8 weeks under -80 ℃. For tLNP (201, 202, 211, 212) , the molar ratios of total PEG-lipid/total lipid were 1%, and the molar ratios of VHH to total lipid were 0.25% (201, 211) or 0.125% (202, 212) . For tLNP (196, 197, 206, 207) , the molar ratios of total PEG-lipid/total lipid were 1.5%, and the molar ratios of VHH to total lipid were 0.25% (196, 206) or 0.125% (197, 207) . The molar ratio of N3-PEG-lipid to total PEG-lipid were remained constant at 50%across all formulations.

Figure 10 illustrates the transfection tLNP with different types of anti-CD117 ligands. MFI value of HEL cells and hCD117 CHO cells after ex vivo treatment with tLNPs (1μg per 2×10⁵ cells, 6 h) . For formulation C and E, the molar ratios of total PEG-lipid to total lipid were 1%and 1.5%, respectively. The molar ratio of N3-PEG-lipid to total PEG-lipid were remained constant at 50%across all formulations. And the molar ratios of scFV or VHH to total lipid ranged from 0 to 0.5%.

Figure 11A illustrates the SDS-PAGE image of anti-CD7 VHH-linker conjugates with different hinge sequences.

Figure 11B illustrates the dimer ratios of anti-CD7 VHH-linker conjugates with different hinge sequences, which were determined by quantifying bands on SDS-PAGE gels using ImageJ software.

Figure 11C illustrates the binding activity of anti-CD7 VHH-linker conjugates with different hinge sequences.

Figure 12A illustrates the effect of hinge sequences of anti-CD7 VHH-linker on the transfection of tLNP by MFI value of Jurkat cells after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, across all formulations. And the molar ratios of VHH to total lipid ranged from 0 to 0.25%.

Figure 12B illustrates the effect of hinge sequences of anti-CD7 VHH-linker on the transfection of tLNP by MFI value of Raji cells after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, across all formulations. And the molar ratios of VHH to total lipid ranged from 0 to 0.25%.

Figure 12C illustrates the effect of hinge sequences of anti-CD7 VHH-linker on the transfection of tLNP by frequency of GFP⁺ cell in CD3⁺ T cells isolated from PBMCs treated under identical conditions. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, across all formulations. And the molar ratios of VHH to total lipid ranged from 0 to 0.25%.

Figure 12D illustrates the effect of hinge sequences of anti-CD7 VHH-linker on the transfection of tLNP by frequency of CD56⁺NK cells isolated from PBMCs treated under identical conditions. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, across all formulations. And the molar ratios of VHH to total lipid ranged from 0 to 0.25%.

Figure 12E illustrates the effect of hinge sequences of anti-CD7 VHH-linker on the transfection of tLNP by frequency of CD14⁺ monocyte cells isolated from PBMCs treated under identical conditions. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, across all formulations. And the molar ratios of VHH to total lipid ranged from 0 to 0.25%.

Figure 13A illustrates the effect of hinge sequences of anti-CD117 VHH on the transfection of tLNP by MFI value of hCD117 CHO cell after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 13B illustrates the effect of hinge sequences of anti-CD117 VHH on the transfection of tLNP by MFI value of wild CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 13C illustrates the effect of hinge sequences of anti-CD117 VHH on the transfection of tLNP by frequency of GFP⁺ cell in hCD117 CHO cells under identical conditions. Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 13D illustrates the effect of hinge sequences of anti-CD117 VHH on the transfection of tLNP by frequency of GFP⁺ cell in wild CHO cells under identical conditions. Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 14A illustrates the effect of none flexible hinge sequences of anti-CD117 VHH on the transfection of tLNP by MFI value of hCD117 CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 14B illustrates the effect of none flexible hinge sequences of anti-CD117 VHH on the transfection of tLNP by MFI value of wild CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 14C illustrates the effect of none flexible hinge sequences of anti-CD117 VHH on the transfection of tLNP by frequency of GFP⁺ cell in hCD117 CHO cells under identical conditions. Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 14D illustrates the effect of none flexible hinge sequences of anti-CD117 VHH on the transfection of tLNP by frequency of GFP⁺ cell in wild CHO cells under identical conditions. Across all formulations, the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG-lipid were remained constant at 1%and 50%, respectively, and the molar ratios of VHH to total lipid were 0.0625%.

Figure 15 illustrates the transfection of tLNP which were prepared through different methods by MFI value of Jurkat cells and HepG2 after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . The linker-free VHH group: VHH were conjugated to the surface of LNPs via mal-thiol chemistry reaction between the cysteine of VHH and Mal-PEG-lipid of LNP. The VHH-linker group: VHH were conjugated to the surface of LNPs via the click reaction between DBCO of VHH-linker conjugate and N₃-PEG-lipid of LNP. The molar ratios of total PEG-lipid to total lipid and Mal-or N3-PEG-lipid to total PEG were remained constant 1%and 50%, respectively, and the molar ratios of VHH to total lipid ranged from 0 to 0.5%.

Figure 16 illustrates the transfection of tLNP prepared through different methods by MFI value of Jurkat cells and HepG2 cells after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . The linker-free VHH group: VHH were conjugated to the surface of LNPs via mal-thiol reaction between the cysteine of VHH and Mal-PEG-lipid of LNP. The VHH-linker group: VHH were conjugated to the surface of LNPs via the click reaction between DBCO of VHH-linker conjugate and N₃-PEG-lipid of LNP. The molar ratios of total PEG-lipid to total lipid and Mal-or N3-PEG-lipid to total PEG were C (1%, 50%) , D (1.0%, 25%) , E (1.5%, 50%) , respectively, and the molar ratios of VHH to total lipid were 0.25%.

Figure 17A illustrates the effect of structures of linkers on the transfection of tLNP by MFI value of CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were remained constant 1%and 50%, respectively, and the molar ratios of VHH to total lipid ranged from 0.0625%to 0.25%.

Figure 17B illustrates the effect of structures of linkers on the transfection of tLNP by MFI value of hCD117 CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were remained constant 1%and 50%, respectively, and the molar ratios of VHH to total lipid ranged from 0.0625%to 0.25%.

Figure 17C illustrates the effect of structures of linkers on the transfection of tLNP by MFI value of Cyno CD117 CHO cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were remained constant 1%and 50%, respectively, and the molar ratios of VHH to total lipid ranged from 0.0625%to 0.25%.

Figure 17D illustrates the effect of structures of linkers on the transfection of tLNP by MFI value of HEL cells after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were remained constant 1%and 50%, respectively, and the molar ratios of VHH to total lipid ranged from 0.0625%to 0.25%.

Figure 18 illustrates the effect of hydrophobic structures of anchor fragment on the transfection of tLNP by MFI value of Jurkat cells and HepG2 cells after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . For PLA2k-PEG2k-N3 group, those tLNP was prepared with PLA2k-PEG2k-N3 as anchor PEG-lipid, and the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , D (1.0%, 25%) , G (2%, 50%) , H (2%, 25%) , respectively. For the DSPE-PEG2k-N3 group, those tLNP was prepared with DSPE-PEG2k-N3 as anchor PEG-lipid, and the molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG lipid were remained constant 1%and 50%, respectively.

Figure 19 illustrates the effect of PEG chain length of PEG-lipid on the transfection of tLNP by MFI value of Jurkat cells and HepG2 cells after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . The tLNP was prepared with DSPE-PEG2k/DSPE-PEG2k-N3, DSPE-PEG5k/DSPE-PEG5k-N3, DSPE-PEG4k/DSPE-PEG3.4k-N3 as non-anchor PEG-lipid/N3-PEG-lipd, respectively. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were I (0.5%, 50%) , C (1%, 50%) , G (2%, 50%) , respectively.

Figure 20 illustrates the effect of non-anchor PEG-lipid on the transfection of tLNP by MFI value of Jurkat cells and HepG2 cells after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . The tLNP was prepared with DSPE-PEG2k, DMG-PEG2k, TAP-PEG2k as non-anchor PEG-lipid, respectively. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , J (1.5%, 33%) , G (2%, 50%) , respectively.

Figure 21 illustrates the transfection of tLNP prepared through post-insertion or surface conjugation by MFI value of Jurkat cells and HepG2 cells after ex vivo treatment with anti-CD5 tLNPs (1μg per 2×10⁵ cells, 6 h) . Frequency of GFP⁺ cell in CD3⁺T cells and CD14⁺ monocyte cells isolated from PBMCs treated under identical conditions. Post insertion group, tLNP was prepared through adding VHH-PEG-lipid to LNP by post insertion. Surface conjugation group, tLNP were prepared through the click reaction between VHH-linker and the LNP with anchor on surface. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , E (1.5%, 50%) , respectively.

Figure 22A illustrates the transfection of anti-CD7 tLNP with different anti-CD7 VHH clones by MFI value of Jurkat cells, Raji and Daudi (non-targeting cells) after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . Data from tLNPs prepared with different anti-CD7 VHH clones but the same formulation were pooled. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were 1%and 50%, respectively.

Figure 22B illustrates the transfection of anti-CD7 tLNP with different anti-CD7 VHH clones by MFI value of Jurkat cells, Raji and Daudi (non-targeting cells) after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . Data from tLNPs prepared with different anti-CD7 VHH clones but the same formulation were pooled. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were 1.5%and 50%, respectively.

Figure 23A illustrates the transfection of anti-CD7 tLNP prepared with different anti-CD7 VHH clones by GFP⁺ cell frequency in CD3⁺ T cells, CD56⁺ NK cells and CD14⁺ monocyte cells from PBMC after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . Data from tLNPs prepared with different anti-CD7 VHH clones but the same formulation were pooled. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were 1%and 50%, respectively.

Figure 23B illustrates the transfection of anti-CD7 tLNP prepared with different anti-CD7 VHH clones by GFP⁺ cell frequency in CD3⁺ T cells, CD56⁺ NK cells and CD14⁺ monocyte cells from PBMC after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . Data from tLNPs prepared with different anti-CD7 VHH clones but the same formulation were pooled. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were 1.5%and 50%, respectively.

Figure 24 illustrates the transfection of anti-CD7 tLNP prepared with different ionizable lipids by GFP⁺ cell frequency in CD3⁺ T cells, CD56⁺ NK cells and CD14⁺ monocyte cells from PBMC after ex vivo treatment with anti-CD7 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG were 1%and 50%, respectively, and the molar ratios of VHH-linker to total lipid were 0.125%.

Figure 25 illustrates in vivo transfection of anti-CD7 tLNP in hCD7 mice. Mice were injected with liver-targeting LNP, and tLNPs with different formulations, at a dose of 1mg/kg. 24 hours post i. v. administration, blood and spleen cells were isolated and stained for a set of antibodies (anti-mouse CD45, anti-mouse CD3, anti-mouse CD56, anti-mouse Ly6C, anti-human CD7) . The percent of tdtomato⁺ cells calculated from T (CD45⁺CD3⁺) , CD7⁺T (CD45⁺CD3⁺CD7⁺) , NK (CD45⁺CD56⁺) and monocyte (CD45⁺Ly6C⁺) gated populations. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , E (1.5%, 50%) , respectively.

Figure 26 illustrates the transfection of anti-CD90 tLNP with different anti-CD90 VHH clones by MFI value of CHO cells, hCD90⁺ CHO cells and HEL, after ex vivo treatment with anti-CD90 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , E (1.5%, 50%) , respectively, and the molar ratios of VHH-linker to total lipid ranged from 0%to 0.5%.

Figure 27 illustrates the transfection of anti-CD117 tLNP with different anti-CD90 VHH clones by MFI value of CHO cells, hCD90⁺ CHO cells and HEL, after ex vivo treatment with anti-CD117 tLNPs (1μg per 2×10⁵ cells, 6 h) . The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were C (1%, 50%) , E (1.5%, 50%) , respectively, and the molar ratios of VHH-linker to total lipid ranged from 0%to 0.5%.

Figure 28 illustrates gene editing in HELs with CD117-tLNP (LNP 922) and CD90-tLNP (LNP-923) encapsulating CRISPR-AaCas12bMax mRNA and sgRNA. HEL were ex vivo treatment (0.5, 1, and 2ug/2E5 cells) for 24 hours with tLNP. Then the cells change to fresh media to culture for 3 days before DNA sequencing. The molar ratios of total PEG-lipid to total lipid and N3-PEG-lipid to total PEG of different formulations were 1%, and 50%, respectively, and the molar ratios of VHH-linker to total lipid were 0.125%.

DETAILED DESCRIPTION

While various embodiments of the invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed.
Definition

As used herein, including the appended claims, the singular forms of words such as “a” , “an” , and “the” , include their corresponding plural references unless the context clearly dictates otherwise.

As used herein, the term "and/or" refers to any one of the elements connected thereby or any combination of several of those elements.

In the context of the present application, unless being otherwise indicated, the wording “comprise” , and variations thereof such as “comprises” and “comprising” will be understood to imply the inclusion of a stated element, e.g., an amino acid sequence, a nucleotide sequence, a property, a step or a group thereof, but not the exclusion of any other elements, e.g., amino acid sequences, nucleotide sequences, properties and steps. When used herein the term “comprise” or any variation thereof can be substituted with the term “contain” , “include” or sometimes “have” or equivalent variation thereof. In certain embodiments, the wording “comprise” also includes the scenario of “consisting of” .

As used herein, the term “lipid nanoparticle” or its abbreviation “LNP” used herein refers to a drug carrier on the order of nanometers and composed of one layer of lipids. In the context of the present application, the term “lipid nanoparticle” also contemplates “lipid-polymer hybrid nanoparticle” which include both lipid portions and hydrophobic polymer portions.

As used herein, the term “ligand-functionalized lipid nanoparticle” , “ligand-conjugated lipid nanoparticle” refers to LNP (s) conjugated with ligand.

As used herein, the term “targeted LNP” “targeting LNP” or “tLNP” refers to “ligand-conjugated lipid nanoparticle” while the ligand in LNP is targeted.

As used herein, the term “ligand-linker conjugate” or “linker-modified ligand” refers to the ligand conjugated with linker. In certain embodiments, the method of conjugation is thiol-based reaction. In certain embodiments, the ligand is a nanobody (VHH) , and the “ligand-linker conjugate” is “VHH-linker conjugate” .

As used herein, the term “thiol-based chemistry” , “thiol-based crosslinking” , “thiol-based conjugation” or “thiol-based reaction” refers to a chemical reaction between a free thiol (-SH) group and other group. The thiol-based reaction could connect separated molecules or separated partition of same molecule. In certain embodiments, the thiol-based chemistry is thiol-maleimide reaction.

As used herein, the term “thiol-reactive functional group” refers to a reactive functional group capable of participating in a thiol-based reaction with an appropriate second reactive functional group, which the second reactive functional group is also a thiol-based functional group. The first and second click chemistry groups, or entities (e.g., molecules or moieties) comprising such groups, may be referred to as complementary. The first and second entities, e.g., molecules, that comprise complementary thiol-based functional groups could be referred to as thiol-based functional partners.

As used herein, the term “cysteine-modified ligand” refers to a ligand with cysteine. In preferred embodiments, the cysteine of ligand is on the outer face of the cysteine-modified ligand. In certain embodiments, the cysteine is used to release free thiol (-SH) group for thiol-based reaction.

As used herein, the term “click reaction” “click chemistry reaction” or “bio-orthogonal click reaction” refers to a simple, efficient, and highly selective method for chemical synthesis. It focuses on quickly and reliably "clicking" small molecules or groups together to form large, more complex molecules or groups. The key features of click chemistry include: (1) mild reaction conditions; (2) fast and efficient reactions; (3) simple by-products; (4) readily available materials.

As used herein, the term “click chemistry reactive group” “bio-orthogonal click chemistry group” “click chemistry group” “bio-orthogonal functional group” or “bio-orthogonal click chemistry handle” refers to a reactive functional group capable of participating in a click chemistry reaction with an appropriate second reactive functional group, which the second reactive functional group is also a click chemistry group. The first and second click chemistry groups, or entities (e.g., molecules or moieties) comprising such groups, may be referred to as complementary. The first and second entities, e.g., molecules, that comprise complementary click chemistry groups could be referred to as click chemistry partners. An entity or molecule comprising a click chemistry group may be referred to as “click-functionalized” , and a bond formed by reaction taken place between the complementary click chemistry partners may be referred to as a “click chemistry bond” .

As used herein, the term “anchor fragment” or “anchor” refers to the partition with click chemistry group of LNP. In certain embodiments, the anchor is on the out face of LNP trend for click reaction.

As used herein, the term “anchor-modified LNP” refers to LNP (s) modified with said anchor.

As used herein, the term “micelles” refers to a spherical aggregate of surfactant molecules that forms in a solution when the concentration of surfactants exceeds a certain threshold known as the critical micelle concentration (CMC) . These structures are characterized by having hydrophobic (water-repelling) tails pointing inward and hydrophilic (water-attracting) heads facing outward, which allows them to interact with both polar and nonpolar substances. In some embodiments, the molecule (s) forming micelles is/are both hydrophobic and hydrophilic.

As used herein, the term “linker” refers to an entity (e.g., molecule, group or moiety) between the ligand and the LNP. In some embodiments, the linker comprises a click chemistry group for connection of the ligand and the LNP.

As used herein, the term “ligand” refers to a molecule conjugated to the LNP. In some embodiments, the ligand is targeted ligand. In some embodiments, the ligand is a molecule binding to the receptor on the surface of target cell. In some embodiments, the ligand is a molecule binding to the receptor free in vivo or in vitro. As used in tLNP, the ligand is a molecule binding to the receptor on the surface of target cell.

As used herein, the term “click chemistry reactive partners” refers to the first and corresponding second click chemistry groups.

As used herein, the term “lipid” refers to a group of organic compounds that are poorly soluble in water, while soluble in nonpolar solvents. Lipids include, but are not limited to, esters of fatty acids.

As used herein, the term “steroid” refers to a molecule having a core with cyclopentane polyhydrogen phenanthrene. In some embodiments, the steroid in LNP is used for stability, anti-inframammary and improving delivery efficiency.

As used herein, the term “ionizable lipid” refers to charged amphiphiles consisting of a hydrophilic head group connected to a hydrophobic tail via a linker. Ionizable lipid is charged so that it can bind to charged molecules or entities such as membrane.

As used herein, the term “cationic lipid” refers to positively charged amphiphiles consisting of a hydrophilic head group connected to a hydrophobic tail via a linker. Cationic lipid is positively charged so that it can bind to negatively charged molecules or entities such as membrane.

As used herein, the term “polymer conjugated lipid” refers to a molecule comprising a lipid portion and a polymer portion.

As used herein, the term “PEG conjugated lipid” or “PEG lipid” refers to a molecule comprising a lipid portion and a polyethylene glycol portion.

As used herein, the term “clickable PEG lipid” refers to PEG lipid conjugated with click chemistry group. In preferred embodiments, the “clickable PEG lipid” is on the surface of the LNP.

As used herein, the term “helper lipid” in the context of LNP is an auxiliary lipid added to improve physicochemical properties of LNP, such as stability, fluidity, and drug loading capacity. Common helper lipids include neutral lipid.

As used herein, the term “neutral lipid” in the context of LNP is interchangeable and refers to lipids that are uncharged or stay in a neutral zwitterionic form at a selected pH.

As used herein, the term “anchor PEG” or “PEG lipid characterized in anchor (fragment) ” refers to the PEG lipid whose PEG chain is a part of the anchor fragment. When said moiety X is N3, the anchor PEG could be N3-PEG-lipid or PEG-N3.

As used herein, the term “lipid PEG” or “PEG lipid characterized in LNP” refers to the PEG lipid whose PEG chain is a part of the LNP excluding the anchor fragment.

As used herein, the term “alkyl” as used herein refers to a straight or branched hydrocarbon chain radical consisting solely of carbon and hydrogen atoms, which is saturated.

As used herein, the term “alkenyl” as used herein refers to a straight or branched hydrocarbon chain radical consisting solely of carbon and hydrogen atoms, which contains one or more double bonds.

As used herein, the term “stereoisomers” refer to compounds consisting of the same atoms bonded by the same bonds but having different three-dimensional structures.

As used herein, the term “tautomers” refer to structural isomers in which a proton shifts from one atom of a molecule to another atom of the same molecule.

As used herein, the term “side chain” refers to a group of atoms attached to the central carbon backbone of a molecule. In some embodiments, the side chain is particularly important in the structure and function of molecules such as amino acids, steroids, and other organic compounds.

As used herein, the term “nucleotide” , “oligonucleotide” and “nucleic acid” can be used interchangeable herein to refer to the base-sugar-phosphate unit consisting a nucleic acid sequence, e.g., a deoxyribonucleic acid (DNA) sequence or a ribonucleic acid (RNA) sequence.

As used herein, the term “conjugation moiety” refers to the entity between the VHH and the linker within a VHH-linker conjugate or ligand-linker conjugate. In some embodiments, this moiety results from a site-specific reaction between a conjugation-enabling moiety positioned at the terminus of the VHH (e.g., a cysteine residue) and a reactive group on the linker (e.g., a maleimide group) . The conjugation moiety thus forms a covalent bond that structurally connects the VHH to the linker in a defined orientation. In certain embodiments, the conjugation moiety comprises a thioether linkage generated via thiol-maleimide chemistry, but may also include alternative site-specific bonds formed through enzymatic or bio-orthogonal reactions. The design of the conjugation moiety is intended to preserve the antigen-binding activity of the VHH while enabling consistent and stable conjugate formation.

As used herein, the term “conjugation-enabling moiety” refers to a functional group incorporated at or near the terminus of a VHH or ligand that facilitates site-specific chemical conjugation to the linker. In some embodiments, the conjugation-enabling moiety comprises a cysteine residue that provides a free thiol group for selective coupling to a maleimide-functionalized linker. In other embodiments, the conjugation-enabling moiety may include a genetically encoded non-natural amino acid bearing a bio-orthogonal reactive group, or a short peptide tag (e.g., LAP) that is enzymatically recognized to mediate covalent bond formation with the linker. The conjugation-enabling moiety is typically positioned downstream of a hinge sequence to ensure spatial accessibility and structural compatibility.

As used herein, the term “hinge” refers to a molecule between the ligand and the compound providing a free thiol (-SH) group, preferably the compound is an amino acid comprising sulfur atom, more preferably the compound is cysteine. In some embodiments, the hinge is a structural spacer between said compound and the ligand. In some embodiments, the hinge is a peptide with controlled-flexibility providing steric hindrance between said compound and the ligand to make the compound more susceptible to exposure on the surface of the ligand-linker conjugate, furthermore, said exposure will propel subsequent thiol reaction without additional process.

As used herein, the term “non-flexible hinge” “structurally constrained hinge” “rigid hinge” “restricted hinge” or “controlled-flexible hinge” refers to the hinge with controlled-flexibility providing steric hindrance between these entities (e.g., molecules, groups or moieties) conjugated by the hinge. In this application, these terms refer to hinges comprising different types of amino acids residues with different contents.

As used herein, the term “rigidity-enhancing residue” refers to the molecule not trend for conformational change. In some embodiments, rigidity-enhancing residue refers to one or more amino acid with complex or rigid side chain. In some embodiments, the rigidity-enhancing residue is selected from the group consisting of phenylalanine (F) , tyrosine (Y) , tryptophan (W) , and proline (P) . In some embodiments, the rigidity-enhancing residue is selected from the group consisting of proline (P) , tyrosine (Y) , phenylalanine (F) , tryptophan (W) , valine (V) , isoleucine (I) , and leucine (L) .

As used herein, the term “α-helix-stabilizing residue” refers to the molecule trend to form intermolecular or intramolecular hydrogen bonding. In some embodiments, α-helix-stabilizing residue refers to one or more amino acid with polar side chain. In some embodiments, the α-helix-stabilizing residue is selected from the group consisting of serine (S) , threonine (T) , tyrosine (Y) , aspartate (D) , glutamate (E) , asparagine (N) , glutamine (Q) , lysine (K) , arginine (R) , and histidine (H) . In some embodiments, the α-helix-stabilizing residue is selected from the group consisting of alanine (A) , leucine (L) , glutamate (E) , and methionine (M) . As used herein, the term “α-helix disruptor” refers to the molecule disrupting hydrogen bonding. In some embodiments, the α-helix disruptor is selected from the group consisting of proline (P) and glycine (G) .

As used herein, the term “potential O-linked glycosylation sites” refers to amino acid sequence motifs or structural features within a polypeptide, such as a hinge, that are susceptible to post-translational O-linked glycosylation. In some embodiments, the potential O-linked glycosylation site comprises one or more serine (S) and/or threonine (T) residues, each of which is positioned adjacent to or within close proximity of a proline (P) residue. This spatial arrangement is known to favor enzymatic O-glycosylation, particularly in mucin-type glycoproteins, where the hydroxyl groups of serine or threonine residues serve as glycan attachment points. In some embodiments, potential O-linked glycosylation sites are defined as sequences containing at least two serine and/or threonine residues positioned adjacent to proline within a continuous stretch of 10 amino acids. In some embodiments, these S and/or T residues account for at least 20%, 30%, or 40%of the amino acids in the hinge region, thereby increasing the likelihood of O-glycosylation occurring. As used herein, inclusion of such potential glycosylation sites within the hinge region may contribute to improved solubility, reduced proteolytic degradation, or modulation of the local conformation of the conjugated ligand. In the context of the present invention, engineered hinges containing potential O-linked glycosylation sites are compatible with site-specific conjugation strategies and do not adversely impact the orientation or accessibility of the conjugation-enabling moiety (e.g., cysteine residue) .

As used herein, the term “hydrophilic residue” refers to the molecule with good solubility in water. In some embodiments, the hydrophilic residue refers to one or more amino acid which can form hydrogen bonds or ionic interactions with water molecules. In some embodiments, the hydrophilic residue comprises these amino acids with polar side chain and these charged amino acid at a selected pH such as acidic, neutral or alkaline pH. In some embodiments, the hydrophilic residue is selected from the group consisting of serine (S) , threonine (T) , asparagine (N) , glutamine (Q) , tyrosine (Y) , aspartic acid (D) , glutamic acid (E) , lysine (K) , arginine (R) , and histidine (H) .

As used herein, the term “peptide” , “polypeptide” and “protein” are used interchangeably herein to refer to a polymer of at least two amino acid residues linked by one or more peptide bonds.

As used herein, the term “subject” as used herein refers to an individual, preferably a vertebrate, more preferably a non-human mammal or human. The non-human mammal can be rodents such as murines, or non-human primates such as simians. The term “subject” may also encompass cells, tissues and progenies of a biological entity obtained in vivo or culture in vitro.

As used herein, the term “effective amount” or “therapeutically effective amount” refers to the quantity of a composition, e.g., a composition comprising the LNP that is sufficient to result in a desired activity when being delivered to a subject in need thereof. Said desired activity may encompass delaying the manifestation of a disorder, arresting or delaying the progression of a disorder, or alleviating at least one symptom of a disorder.

As used herein, the term “conjugate” , “connect” , “link” , “linkage” and “linked with…” and variations thereof such as “conjugated” , “connected” , “linked” are used interchangeably herein to refer to two or more moieties being physically or chemically associated or connected with one another to form a molecular structure that is sufficiently stable so that the moieties remain associated under the conditions in which the linkage is formed and, preferably, under the conditions in which the new molecular structure is used, e.g., physiological conditions. In certain preferred embodiments of the invention the linkage is a covalent linkage. In other embodiments the linkage is noncovalent. Moieties may be linked either directly or indirectly. When two moieties are directly linked, they are either covalently bonded to one another or are in sufficiently close proximity such that intermolecular forces between the two moieties maintain their association. When two moieties are indirectly linked, they are each linked either covalently or noncovalently to a third moiety, which maintains the association between the two moieties. In general, when two moieties are referred to as being linked by a “linking moiety” or “linking portion” , the linkage between the two linked moieties is indirect, and typically each of the linked moieties is covalently bonded to the linking moiety. Two moieties may be linked using a “linker” . A linker can be any suitable moiety that reacts with the entities to be linked within a reasonable period of time, under conditions consistent with stability of the entities (portions of which may be protected as appropriate, depending upon the conditions) , and in sufficient amount, to produce a reasonable yield. Typically, the linker will contain at least two functional groups, one of which reacts with a first entity and the other of which reacts with a second entity. It will be appreciated that after the linker has reacted with the entities to be linked, the term “linker” may refer to the part of the resulting structure that originated from the linker, or at least the portion that does not include the reacted functional groups. A linking moiety may comprise a portion that does not participate in a bond with the entities being linked, and whose main purpose may be to spatially separate the entities from each other.

As used herein, the term “site-specific conjugation” refers to a chemistry reaction based on specific sites. In some embodiments, the site-specific conjugation bases on the specific atoms or groups of entities (e.g., molecules, groups or moieties) . In some embodiments, the recognition of sites is based on enzyme or the corresponding ligand of said sites. In some embodiments, the site-specific conjugation is a click chemistry reaction or a thiol-based reaction.
Ligand-Conjugated LNP Technology

The present disclosure relates to ligand-functionalized lipid nanoparticles (being referred as “target LNPs” or “tLNPs” ) designed for targeted delivery of nucleic acids and therapeutic molecules. In some embodiments, the disclosed technology enables site-specific conjugation of targeting ligands onto the surface of lipid nanoparticles, thereby improving delivery efficiency, reducing off-target effects, and enhancing stability. Unlike conventional ligand-functionalized nanoparticles, which often rely on non-specific adsorption or uncontrolled chemical modifications, the present invention utilizes a precisely engineered conjugation strategy that ensures optimal ligand orientation and bioactivity. The disclosed system integrates key structural components, including an engineered targeting ligand, a linker facilitating site-specific conjugation, an anchor fragment incorporated into the LNP structure, and a lipid nanoparticle core encapsulating the therapeutic payload.

In some embodiments, the tLNP is generated through a multi-step process designed to achieve efficient ligand conjugation while preserving nanoparticle stability. The preparation begins with the synthesis of ligand-linker conjugates, where the targeting ligand, such as a VHH, scFv, Fab fragment, or peptide, is site-specifically modified with a linker via a conjugation moiety (e.g. a thioether linkage) . This modification allows for controlled attachment to a bio-orthogonal reactive anchor fragment pre-integrated into the lipid nanoparticle. In some embodiments, the conjugation is facilitated through a click-chemistry reaction between a reactive moiety on the linker and a complementary moiety on the anchor fragment. The resulting ligand-conjugated LNPs exhibit improved ligand presentation, reducing steric hindrance and maximizing binding affinity to target cells.

In contrast to conventional ligand-functionalized LNPs, which often employ direct ligand conjugation through maleimide-thiol chemistry or post-insertion methods, the present disclosure introduces an anchor fragment to mediate ligand attachment. In some embodiments, the anchor fragment is pre-incorporated into the lipid nanoparticle during formulation, eliminating the need for post-insertion ligand attachment, which may result in reduced efficiency and stability. The anchor fragment consists of a lipid-PEG structure functionalized with a bio-orthogonal reactive moiety, such as an azide or alkyne group, which selectively reacts with a complementary functional group on the ligand-linker conjugate. This pre-functionalized approach improves conjugation efficiency, maintains nanoparticle integrity, and minimizes ligand loss due to competitive exchange with plasma proteins.

In some embodiments, the tLNP design incorporates a hinge between the targeting ligand and the linker. The hinge provides controlled-flexibility, limited-mobility and spatial separation to ligand-linker conjugate, ensuring minimal interference of the internal structure of ligand and efficiency conjugation. Unlike conventional hinge, which are often composed of flexible glycine-rich or serine-rich motifs, the present disclosure introduces a structurally constrained hinge that enhance conjugation stability and reduce steric hindrance. The hinge may be derived from immunoglobulin hinge regions, proline-rich motifs, or computationally designed rigid structures, ensuring optimal ligand presentation without compromising structural integrity.

In some embodiments, the lipid nanoparticle core consists of ionizable lipids, structural lipids, sterols, and polymer-conjugated lipids. Ionizable lipids, selected based on their pKa properties, play a critical role in nucleic acid encapsulation and intracellular delivery. These lipids remain neutral at physiological pH but become protonated within endosomal compartments, facilitating endosomal escape of the payload. The choice of ionizable lipid composition influences LNP stability, biodistribution, and transfection efficiency. Structural lipids, such as phosphatidylcholine or DSPC, contribute to bilayer stability, while sterols, such as cholesterol or cholesterol analogs, modulate membrane rigidity and fluidity. Polymer-conjugated lipids, such as PEGylated lipids, provide steric stabilization, reducing opsonization and prolonging circulation half-life. In some embodiments, the LNP composition is optimized to achieve a balance between stability, delivery efficiency, and immune evasion.

In some embodiments, the disclosed ligand-functionalized lipid nanoparticles (tLNPs) enable precise delivery of therapeutic payloads by leveraging ligand-receptor interactions for targeted cellular uptake. Unlike untargeted LNPs, tLNPs achieve enhanced specificity and biodistribution, facilitating receptor-mediated endocytosis and intracellular release of nucleic acid therapeutics, including mRNA, siRNA, and CRISPR-Cas12b gene-editing constructs. In certain embodiments, tLNPs efficiently deliver gene-editing payloads to hematopoietic stem cells (HSCs) or immune cells such as T cells, B cells, and NK cells, enabling gene therapy and immunomodulation for cancer, autoimmune disorders, and infectious diseases. The conjugation strategy optimizes ligand density, improving receptor engagement, internalization efficiency, and therapeutic efficacy while minimizing off-target effects. The integration of site-specific conjugation, bio-orthogonal click chemistry, and engineered hinge sequences ensures tLNP stability, reproducibility, and controlled ligand orientation for enhanced delivery performance.
Bioactive Ligand

In some embodiments, the targeting ligand is a biomolecule capable of recognizing and binding to a specific cellular receptor or biomarker. The targeting ligand facilitates selective interaction between the conjugated lipid nanoparticles (LNPs) and the target cells. In certain embodiments, the targeting ligand may be an antibody, an antibody fragment, a receptor-binding peptide, a small molecule, an aptamer, or a carbohydrate moiety.

In some embodiments, the targeting ligand is an antibody, antigen binding fragment or an antibody-derived fragment, including but not limited to full-length antibodies, Fab, single-chain variable fragments (scFv) , heavy-chain-only antibodies (HCAb) , single-domain antibodies (sdAb) , VHH antibodies, Designed Ankyrin Repeat Proteins (DARPins) , fibronectin type III (FN3) domains, centyrins, or other engineered binding proteins. In certain embodiments, the targeting ligand is a single-domain antibody (sdAb) or a VHH antibody fragment.

In some embodiments, the targeting ligand is a non-protein biomolecule such as a folate derivative, an antibiotic mimetic, a vitamin, a carbohydrate, or an N-Acetylgalactosamine (GalNAc) moiety. In some embodiments, the targeting ligand is a peptide, which may be linear or cyclic, designed to bind specific cell-surface receptors.

In some embodiments, the targeting ligand is a protein or peptide with a molecular weight of less than 50 kDa, with a particular emphasis on ligands smaller than 30 kDa. A smaller ligand is preferable for conjugation to LNPs because it improves ligand accessibility, reduces steric hindrance, and facilitates efficient nanoparticle purification after conjugation.

In particular embodiments, the targeting ligand is a VHH antibody, which has a molecular weight of approximately 15 kDa. VHH fragments exhibit favorable biochemical properties, including high solubility, excellent stability, and low immunogenicity, making them well-suited for conjugation onto LNP surfaces. Unlike scFv, which requires the correct pairing of variable heavy (VH) and variable light (VL) chains, VHH molecules are naturally stable as single domains, allowing for more efficient production, purification, and conjugation. Experimental data (as described in Example 14) demonstrate that VHH-conjugated LNPs exhibit significantly higher targeted delivery efficiency compared to scFv-conjugated LNPs. This suggests that ligands with lower molecular weight are more favorable for efficient nanoparticle conjugation and targeted cellular uptake.

In some embodiments, the targeting ligand is designed to bind to a receptor expressed on the surface of specific cell types. In some embodiments, the targeting ligand binds to receptors present on hematopoietic cells, including T cells, B cells, natural killer (NK) cells, and hematopoietic stem cells (HSCs) .

In certain embodiments, the targeting ligand is selected to interact with T cell surface receptors, such as CD2, CD3, CD5, CD7 CD8. In some embodiments, the targeting ligand binds to HSC surface receptors, such as CD34, CD90 (Thy-1) , CD117 (c-Kit) .
Ligand Engineering for Conjugation

In some embodiments, the ligand is engineered to include a conjugation-enabling moiety positioned at either its N-terminal or C-terminal region to facilitate site-specific chemical conjugation. As used herein, a conjugation-enabling moiety refers to a functional group or peptide sequence that is capable of forming a covalent bond with a linker through a site-selective reaction. Suitable conjugation-enabling moieties include, but are not limited to, thiol-bearing residues (e.g., cysteine) , enzyme-recognizable tags (e.g., LAP tag for LplA-mediated ligation) , and non-natural amino acids incorporating bio-orthogonal functional groups such as azides, alkynes, or strained alkenes. The selection of a conjugation-enabling moiety may be tailored to the desired coupling chemistry and intended application, and it enables precise spatial control over ligand orientation and density on the nanoparticle surface.

In some embodiments, the ligand is modified by incorporating a cysteine residue at either the C-terminal or N-terminal region to enable site-specific conjugation. The introduction of a cysteine residue facilitates controlled thiol-based conjugation, thereby ensuring precise attachment of the ligand to the linker.

In some embodiments, the cysteine residue is chemically reduced to generate a ligand monomer containing a free thiol (-SH) group. The free thiol serves as a reactive handle for conjugation to a linker via thiol-reactive functional groups, including but not limited to maleimide, parafluoro, ene, yne, vinylsulfone, pyridyl disulfide, thiosulfonate, and thiol-bisulfone. In some embodiments, the maleimide-thiol reaction is preferred due to its high specificity and stability under physiological conditions.

In some embodiments, the conjugation-enabling moiety is an enzyme-recognizable tag that permits site-specific covalent attachment of a linker through an enzymatic reaction. Suitable tags include peptide sequences recognized by enzymes such as LplA (e.g., LAP tag) , transglutaminase (e.g., Q-tag) , sortase A (e.g., LPXTG motif) , and formylglycine-generating enzyme (e.g., aldehyde tag) . These tags enable mild and chemoselective conjugation at defined sites on the ligand, allowing for uniform orientation and preserving ligand activity. For example, the use of the LAP tag allows for selective conjugation via engineered LplA mutants to install a clickable moiety such as endo-BCN, which can subsequently participate in bio-orthogonal ligation to the LNP.

In other embodiments, the conjugation-enabling moiety comprises a non-natural amino acid bearing a bio-orthogonal functional group. Non-natural amino acids such as p-azidophenylalanine (AzF) , p-propargyloxyphenylalanine (PrF) , or bicyclononyne (BCN) -lysine can be incorporated into the ligand sequence via genetic code expansion. These modified residues introduce chemically reactive handles-such as azide, alkyne, or strained alkene groups-that enable site-specific conjugation through copper-catalyzed azide-alkyne cycloaddition (CuAAC) , strain-promoted azide-alkyne cycloaddition (SPAAC) , or tetrazine-trans-cyclooctene (TCO) ligation. The incorporation of such residues expands the chemical toolbox for orthogonal and site-specific ligand conjugation without requiring additional enzymatic processing.

In some embodiments, a hinge is inserted between the ligand and the cysteine residue. This hinge functions as a structural spacer, modulating the spatial orientation of the cysteine residue and improving site-specific conjugation efficiency. The hinge plays a pivotal role in ligand conjugation stability, steric accessibility, and overall structural integrity. A well-designed hinge can reduce interference from the C-terminal cysteine with the ligand’s internal disulfide bonds, thereby preserving ligand integrity, enhance steric accessibility of the conjugation site, ensuring efficient attachment of the ligand to the linker, and modulate structural flexibility, optimizing the ligand’s binding orientation post-conjugation.

In some embodiments, said hinge inserted between the ligand and cysteine residue is a structurally constrained hinge, designed to provide controlled-flexibility while maintaining a defined conformation. In some embodiments, the hinge exhibits a certain degree of rigidity to maintain structural stability while minimizing aggregation or dimerization post-conjugation.

In some embodiments, the controlled-flexibility and limited-mobility of the hinge is achieved through one of the following features:

(a) The hinge comprises rigidity-enhancing residues, including proline (P) , tyrosine (Y) , phenylalanine (F) , tryptophan (W) , valine (V) , isoleucine (I) , and leucine (L) , in a proportion of no less than 20%, and contains less than 20%glycine (G) . The combination of these rigidity-enhancing residues ensures structural stability, while the limitation on glycine content prevents excessive conformational flexibility and maintains the necessary balance between flexibility and rigidity.

(b) The hinge comprises strong α-helix-stabilizing residues, which enhance local rigidity by stabilizing the peptide backbone through hydrogen bonding. These residues include alanine (A) , leucine (L) , glutamate (E) , and methionine (M) , constituting at least 80%of the hinge composition. Additionally, the hinge does not contain α-helix disruptors, such as proline (P) and glycine (G) , which would interfere with the formation and stability of α-helical structures by disrupting backbone hydrogen bonding and compromising the overall rigidity of the hinge.

In some embodiments, the hinge comprises 2 to 40 amino acids in length. In some embodiments, wherein the hinge comprises 5 to 25 amino acids in length. In some embodiments, the hinge does not contain cysteine residues to eliminate unintended disulfide bonding or uncontrolled polymerization.

In protein and peptide structures, certain amino acids contribute significantly to rigidity by restricting backbone flexibility and stabilizing secondary structures. Among them, proline (Pro, P) is the most rigid due to its unique pyrrolidine ring, which directly connects to the backbone nitrogen. This structural constraint severely limits the(phi) dihedral angle, reducing flexibility and often introducing kinks in protein backbones. As a result, proline frequently disrupts α-helices and stabilizes β-turns, making it an essential element in rigid loops and structured hinges. Aromatic amino acids, including tyrosine (Tyr, Y) , phenylalanine (Phe, F) , and tryptophan (Trp, W) , also contribute to rigidity through their bulky, planar benzene or indole rings. These rings restrict rotational freedom around the peptide backbone and provide structural stability, especially in β-sheet cores and hydrophobic packing regions. Their presence enhances protein rigidity by promoting stable interactions such as π-π stacking and hydrophobic clustering. Branched-chain amino acids, such as valine (Val, V) , isoleucine (Ile, I) , and leucine (Leu, L) , introduce steric hindrance through their β-carbon branching, which restricts rotational flexibility and stabilizes β-strands. This structural constraint makes them essential in β-sheets, where they help maintain rigid frameworks.

Here is a classification of the most rigid amino acids along with their mechanisms of rigidity:

In some embodiments, said hinge comprises strong α-helix-stabilizing amino acids, which contribute to enhanced local rigidity by stabilizing the peptide backbone through intramolecular hydrogen bonding. These amino acids, including alanine (A) , leucine (L) , glutamate (E) , and methionine (M) , exhibit a high propensity to adopt α-helical conformations, thereby reinforcing structural integrity and reducing conformational disorder. The stabilization of the α-helix occurs through the formation of hydrogen bonds between the carbonyl oxygen of a given residue and the amide hydrogen of another residue located approximately three to four residues apart (i, i+4 spacing) along the peptide chain. This repetitive hydrogen bonding pattern strengthens the helical structure, restricting rotational freedom and minimizing structural fluctuations. Furthermore, hydrophobic interactions contributed by residues such as leucine and methionine enhance the packing stability of the helix, while charged residues such as glutamate facilitate electrostatic interactions that further stabilize the overall conformation. In some embodiments, by incorporating a high proportion of these α-helix-stabilizing amino acids, the hinge achieves a defined and rigid secondary structure, which provides mechanical stability while limiting excessive flexibility.

In some embodiments, said hinge comprises strong α-helix-stabilizing amino acids while does not contain α-helix disruptors, which are amino acids that interfere with the formation and stability of α-helical structures. Proline (P) and glycine (G) are the primary α-helix disruptors due to their unique structural properties. Proline, with its cyclic pyrrolidine ring, imposes rigid constraints on backbone dihedral angles, preventing the formation of the typical α-helical hydrogen bonding pattern and often introducing kinks or disruptions in helical regions. Glycine, on the other hand, lacks a side chain beyond a single hydrogen atom, resulting in extreme backbone flexibility that increases local conformational entropy and destabilizes ordered secondary structures such as α-helices. By excluding these α-helix disruptors, the hinge maintains a stable helical conformation, ensuring structural rigidity and minimizing undesired conformational fluctuations that could compromise its mechanical stability.

Here is a table summarizing the characteristics of α-Helix-stabilizing amino acids vs. α-Helix Disruptors

In some embodiments, the hinge does not contain glycine (G) or comprises no more than 40%glycine to ensure a balance between flexibility and structural stability. Glycine, due to its minimal steric hindrance and lack of a side chain beyond a single hydrogen atom, confers maximum conformational freedom to the peptide backbone. While glycine-rich regions can promote random coil structures and enhance flexibility by reducing steric constraints, an excessive proportion may compromise structural integrity, leading to instability, loss of defined conformation, and potential misfolding in functional applications.

Amino acid composition is a critical factor in hinge design, influencing the balance between rigidity and flexibility. Experimental data (e.g., Example 15 and Example 16) of the present disclosure demonstrate that the selection of an appropriate sequence of hinge significantly affects the efficiency of ligand conjugation and subsequent targeted delivery performance. Specifically, hinge such as Hinge-1 and Hinge-3 exhibited suboptimal conjugation efficiency, whereas Hinge-4, Hinge-5, Hinge-6, and Hinge-7 yielded superior conjugation stability and targeting efficiency. The data indicate that hinge sequences must incorporate a sufficient proportion of amino acids for structure stability to optimize conjugation efficiency and spatial orientation of the cysteine residue. Conversely, excessively flexible hinge may lead to suboptimal conjugation and increased dimerization, as the reactive thiol group may be sterically hindered or misaligned.

In some embodiments, the sequence of hinge is proline-rich, designed to introduce structural rigidity and minimize steric interference during conjugation. Proline (P) is a unique cyclic amino acid that imposes conformational constraints on peptide backbones, thereby limiting unnecessary flexibility and enhancing spatial orientation for site-specific conjugation. Representative sequences of proline-rich hinge include: (PA) n, (PPP) n, (PPG) n, where n = 2 to 15; VPPPPP (SEQ ID NO: 16) ; APGPPGPPG (SEQ ID NO: 17) ; SPSTPPTPSPSTPP (SEQ ID NO: 1) .

In some embodiments, the hinge is derived from immunoglobulin hinge regions, but is engineered to exclude cysteine residues to prevent unintended disulfide bonding. In some embodiments, the hinge is derived from an immunoglobulin upper or lower hinge region with no cysteine. Examples of sequences of Ig-derived hinge include: SPSTPPTPSPSTPP (derived from IgA1, SEQ ID NO: 1) ; APELLGGP (derived from IgG1, SEQ ID NO: 18) ; APEFLGGP (derived from IgG4, SEQ ID NO: 19) ; APPVAGP (derived from IgG2, SEQ ID NO: 20) ; PAPELLGGPSVFLFPPKPKDTLMIS (extended IgG-derived hinge sequence, SEQ ID NO: 21) .

In some embodiments, the hinge is a computationally designed hinge. With advancements in computational protein engineering, novel hinge have been developed to optimize the balance between rigidity and solubility. These computationally optimized sequences incorporate a mix of rigid amino acids, such as proline (P) , alanine (A) , and leucine (L) , combined with charged residues like glutamate (E) and lysine (K) to enhance solubility. Examples of sequences of computationally designed hinge include KESGSVSSEQLAQFRSLD (SEQ ID NO: 4) , SPPRTSDPKNTP (SEQ ID NO: 14) , TPPRSDNTKSPQ (SEQ ID NO: 60) , EPPKRSTDNTPK (SEQ ID NO: 15) , TPPKDSTNQSPR (SEQ ID NO: 61) , SPPRTNQETPKD (SEQ ID NO: 62) .

In some embodiments, the hinge adopts a stable α-helical conformation, which reinforces structural stability while maintaining sufficient steric accessibility for conjugation. These sequences typically contain amino acids that promote α-helix formation, such as alanine (A) , glutamic acid (E) , and lysine (K) . Representative sequences of α-helical hinge include: (EAAAK) n, where n = 2 to 6; AEAAAKEAAAKA.

In some embodiments, hydrophobic-hydrophilic alternating hinge are incorporated to balance structural rigidity and aqueous solubility. Representative sequences of hydrophobic-hydrophilic alternating hinge include: (AK) n, while n is any integer between 2 and 15 (alternating alanine (A) and lysine (K) ) ; (LQ) n, while n is any integer between 2 and 15 (alternating leucine (L) and glutamine (Q) ) .

Examples of rigid or semi-rigid hinges:

In some embodiments, the hinge is derived from naturally occurring human proteins, ensuring compatibility with site-specific conjugation methods. The inter-domain linker peptides of natural multi-domain proteins provide an ample source of potential linkers for novel fusion proteins, these linkers provide the conformation, flexibility and stability needed for a protein’s biological function in natural environment. In some embodiments, the naturally occurring hinge demonstrated controlled flexibility and limited mobility. In some embodiments, these sequences of naturally occurring hinge are engineered to be between 2 and 40 amino acids in length and to contain (a) one or more rigidity-enhancing residue, including P, Y, F, W, V, I, and L, constituting ≥20%, with glycine (G) content <20%; or (b) one or more α-helix-stabilizing residue, including A, L, E, and M, constituting ≥80%, while excluding α-helix disruptors P and G to maintain structural stability.

In some embodiments, additional amino acid residues may be incorporated after the cysteine residue, specifically at the C-terminus of the cysteine residue, to further optimize conjugation efficiency and steric accessibility. For instance, 1 to 5 (for example, 1, 2, 3, 4, or 5) alanine (A) residues may be added at the C-terminus of cysteine to fine-tune steric exposure and prevent aggregation. The incorporation of additional residues reduces steric clashes during conjugation, ensuring higher reaction efficiency.

In some embodiments, a tag sequence can be inserted within the ligand sequence to facilitate purification and characterization of the ligand. The tag may be positioned between the hinge sequence and the original ligand sequence, between the hinge and the terminal cysteine residue, or at other suitable locations within the ligand construct. In some embodiments, the tag sequence may comprise a polyhistidine tag (His-tag) , commonly consisting of six to ten histidine residues, which allows for affinity purification using nickel (Ni²⁺) or cobalt (Co²⁺) chelate chromatography. In other embodiments, the tag may include other affinity purification motifs such as Strep-tag, FLAG-tag, HA-tag, or Avi-tag, which provide additional versatility in ligand purification and detection. In other embodiments, flexible linkers such as Gly-Ser repeats or short rigid spacers may be introduced adjacent to the tag sequence to ensure proper folding and accessibility of the ligand.
Linkers for Bridging Ligand with LNP

In some embodiments, the present invention provides a ligand-linker conjugate, wherein the linker is specifically designed to facilitate site-specific conjugation between a modified ligand and a functionalized nanoparticle surface. The linker functions as a molecular bridge, ensuring covalent attachment of the ligand to the lipid nanoparticle (LNP) while preserving its targeting functionality. By leveraging bio-orthogonal conjugation chemistries, the disclosed linker designs enable highly specific, controlled, and reproducible conjugation of the ligand-linker conjugate to the LNP.

In accordance with the present disclosure, the linker comprises two distinct functional regions: (a) a reactive group capable of forming a covalent bond with a conjugation-enabling moiety present at the terminus of the ligand, thereby enabling site-specific conjugation of the ligand to the linker; and (b) one or more moiety Y, which serves as a reactive handle for bio-orthogonal click chemistry with a functionalized nanoparticle surface. In some embodiments, the reactive group of the linker is a thiol-reactive moiety suitable for conjugation to a cysteine residue on the ligand, such as maleimide, parafluoroaryl, ene, yne, vinylsulfone, pyridyl disulfide, thiosulfonate, or bisulfone groups. In other embodiments, the reactive group is selected based on the nature of the conjugation-enabling moiety, such as enzyme-recognizable sequences or non-natural amino acids bearing azide, alkyne, or strained alkene functionalities. Moiety Y ensures that the resulting ligand-linker conjugate is pre-functionalized for efficient and orthogonal attachment to lipid nanoparticles via a subsequent click chemistry reaction.

In one embodiment, the use of ligand-linker conjugates provides distinct advantages over traditional direct conjugation methods, particularly in terms of process robustness, batch consistency, and targeting efficiency. One of the primary benefits of using a pre-functionalized ligand-linker conjugate is the elimination of ligand dimerization, a common issue encountered in direct thiol-maleimide conjugation. When ligands with free cysteine residues are directly conjugated to nanoparticles, unintended dimerization and higher-order aggregation can occur due to uncontrolled thiol-thiol interactions. By contrast, the ligand-linker conjugate already contains a covalently linked functional handle and no free thiol groups, preventing dimer formation and ensuring a more controlled conjugation process.

Another key advantage is the increased batch-to-batch consistency in ligand conjugation and targeted delivery efficiency. The pre-functionalization of the ligand with a linker ensures that every batch of ligand-linker conjugate is chemically identical, reducing variability during the conjugation step. This significantly improves reproducibility across multiple production batches of targeted lipid nanoparticles (tLNPs) . In one embodiment, as demonstrated in Example 10, tLNPs prepared using the ligand-linker conjugate exhibited greater consistency in delivery efficiency, with minimal variation in target cell transfection rates. In contrast, tLNPs prepared using traditional thiol-maleimide conjugation with free-thiol ligands showed noticeable batch-to-batch variability, resulting in inconsistent target and non-target cell delivery.

In one embodiment, ligand-linker conjugates enhance overall targeting efficiency compared to linker-free conjugation methods. The incorporation of a well-defined linker structure provides better steric accessibility of the ligand at the nanoparticle surface, improving receptor recognition and binding. As evidenced by Example 17, across various ligand conjugation densities and formulations, tLNPs containing VHH-linker conjugates consistently demonstrated superior delivery to target cells while maintaining lower off-target effects. These findings suggest that linker-mediated conjugation not only improves the precision of site-specific attachment but also optimizes the functional presentation of the ligand, thereby increasing the overall efficiency of targeted delivery.

In some embodiments, the ligand-linker conjugate is synthesized through thiol-based chemistries, leveraging the reactivity of cysteine residues present in the modified ligand. Examples of thiol-specific conjugation reactions used for linker synthesis include thiol-maleimide reaction, thiol-parafluoro reaction, thiol-ene reaction, thiol-yne reaction, thiol-vinylsulfone reaction, thiol-pyridyl disulfide reaction, thiol-thiosulfonate reaction, and thiol-bisulfone reaction. These chemistries provide a versatile platform for covalent attachment of linkers to ligands, ensuring that the resulting conjugate does not contain free thiol groups, thus preventing undesired ligand dimerization or polymerization.

In some embodiments, the ligand-linker conjugate is synthesized through an enzyme-mediated reaction. In such embodiments, the linker is designed to include a corresponding enzyme-compatible substrate or reactive group capable of forming a covalent linkage with the enzyme-activated residue of the ligand. For example, when the conjugation-enabling moiety comprises a peptide sequence recognized by lipoic acid ligase (LplA) , such as the LAP motif, the linker may contain a lipoic acid analog (e.g., endo-BCN-pentanoic acid or alkyl azide derivatives) that serves as a substrate for enzymatic ligation. Similarly, in embodiments utilizing transglutaminase-recognized peptide tags (e.g., glutamine-containing sequences) , the linker may comprise a lysine mimic or related functional group suitable for enzymatic transamidation. In further embodiments, the conjugation-enabling moiety comprises a short peptide tag specifically recognized by a ligating enzyme. For example, when Sortase A is used, the ligand includes a C-terminal LPXTG motif (e.g., LPETG) , which is enzymatically cleaved between the threonine and glycine. The linker in such embodiments includes a terminal oligoglycine (e.g., GGG) or glycine derivative, which serves as the nucleophilic substrate for transpeptidation, resulting in covalent linkage between the ligand and the linker. Alternatively, when the conjugation-enabling moiety comprises a CXPXR motif recognized by formylglycine-generating enzyme (FGE) , the cysteine is enzymatically oxidized to formylglycine, which can then undergo selective reaction with hydrazide-or aminooxy-functionalized linkers via oxime or hydrazone ligation. These enzyme-mediated strategies provide high site-selectivity and mild reaction conditions, and are particularly useful for preserving ligand functionality and minimizing off-target modification. The resulting conjugate remains compatible with downstream click chemistry via moiety Y, thereby integrating enzyme-directed conjugation with nanoparticle surface bio-orthogonal functionalization.

In some embodiments, the conjugation-enabling moiety comprises an unnatural amino acid bearing a bio-orthogonal reactive group and the ligand-linker conjugate is synthesized through a bio-orthogonal reaction. In these embodiments, the linker contains a complementary chemical group that reacts selectively with the unnatural residue via bio-orthogonal ligation strategies, including but not limited to strain-promoted azide–alkyne cycloaddition (SPAAC) , oxime ligation, or tetrazine–trans-cyclooctene (TCO) click chemistry. This strategy ensures high precision and compatibility with living systems, and avoids the need for enzymatic activation or chemical reduction.

In some embodiments, the linker structure is further optimized by incorporating a spacer region between the thiol-reactive functional group and moiety Y. This spacer region plays a crucial role in modulating steric accessibility, linker solubility, and conjugation flexibility. The spacer provides additional conformational freedom for efficient ligand presentation on the nanoparticle surface, minimizing steric hindrance that may interfere with ligand-target interactions.

In some embodiments, the spacer region may be composed of an alkyl chain, where the carbon length ranges from C₂ to C₃₀, which may be saturated or unsaturated. In some embodiments, the spacer region may be composed of a polyethylene glycol (PEG) chain, where n ethylene glycol (EG) units are incorporated, and n is any integer between 2 and 40. In some embodiments, the spacer region may be composed of a hybrid spacer, which may combine both hydrophobic (alkyl) and hydrophilic (PEG) moieties, providing a balanced conjugation environment that optimizes ligand stability and LNP anchoring.

In some embodiments, the linker may be further engineered to include anionic or cationic functional groups, thereby modifying its hydrophilicity, charge properties, and bio-compatibility. In some embodiments, the linker structure may contain anionic groups, including but not limited to: sulfonate (-SO₃ ^-) , carboxylate (-COO^-) , phosphate (-PO₄ ^2-) . In some embodiments, the linker may contain cationic groups, including but not limited to: ammonium (-NH₄ ⁺, -NR₃ ⁺) , guanidinium (-C (=NH) NH₂ ⁺) , imidazolium (-C₃H₄N₂ ⁺) . In some embodiments, these charged functional groups may be covalently incorporated into the linker backbone or added as terminal modifications, depending on the desired solubility and bio-interaction properties.

In some embodiments, the moiety Y on the terminus of the linker serves as a reactive click handle for bio-orthogonal conjugation with a functionalized nanoparticle surface. In some embodiments, the formation of the ligand-conjugated LNP is mediated via a bio-orthogonal click reaction between a reactive moiety (moiety X) present on the LNP surface and a pre-functionalized ligand containing a complementary reactive moiety (moiety Y) . The use of bio-orthogonal click chemistry ensures that the conjugation reaction proceeds efficiently under physiological conditions.

In some embodiments, a key feature of bio-orthogonal click chemistry in the present invention is the utilization of reactive functional groups (moiety X and moiety Y) that selectively react with each other in a highly specific manner. In some embodiments, the conjugation reaction between moiety X and moiety Y is achieved via a cycloaddition reaction, nucleophilic ring-opening reaction, nucleophilic addition reaction, thiol-ene reaction, or Diels-Alder reaction. The choice of bio-orthogonal chemistry depends on factors such as reaction kinetics, stability of the conjugate, and compatibility with large-scale production. These reactions proceed under mild physiological conditions, minimizing potential damage to sensitive ligands or LNP components while ensuring rapid and stable conjugation.

In some embodiments, moiety X and moiety Y reactive pairs include azide and dibenzocyclooctyne (DBCO) , azide and 4-dibenzocyclooctynol (DIBO) , azide and biarylazacyclooctynone (BARAC) , azide and bicyclononyne (BCN) , tetrazine and trans-cyclooctene (TCO) , tetrazine and cyclopropane, and azide-alkyne click groups. The use of these selective click reactions ensures that the ligand-linker conjugate can be precisely anchored to the LNP surface, avoiding non-specific interactions and reducing unwanted aggregation. Importantly, the selection of click chemistry partners is interchangeable between moiety X and moiety Y, providing flexibility in the design of conjugation strategies.
Anchor Fragment for Functionalized Nanoparticles

The present disclosure provides compositions and methods for functionalizing lipid nanoparticles (LNPs) by incorporating anchor fragments. These anchor fragments play a critical role in enabling precise ligand attachment through bio-orthogonal click chemistry. Unlike conventional lipid-PEG modifications, which primarily serve to enhance nanoparticle stability, the anchor fragments disclosed herein are specifically designed for bio-orthogonal functionalization. By introducing a reactive moiety (moiety X) into the anchor fragment, this invention ensures controlled ligand orientation, minimizes non-specific binding, and enhances conjugation efficiency.

In some embodiments, the anchor fragments are incorporated into the LNP structure during nanoparticle formation rather than being introduced post-assembly via insertion techniques. In some embodiments, the anchor fragments are first embedded within the LNPs and then conjugated with the ligand-linker conjugate rather than being pre-conjugated to the ligand and subsequently inserted into the LNP. This approach was validated in Example 22, where tLNPs prepared using the post-insertion method were compared to those prepared via direct incorporation of anchor fragments during LNP formulation and surface conjugation of ligand. The study demonstrated that tLNPs generated through the post-insertion method exhibited significantly higher off-target effects, particularly in monocytes, whereas the direct incorporation of anchor fragments approach yielded superior targeting efficiency in Jurkat and primary T cells. These findings highlight the importance of using a surface conjugation strategy to maximize targeting efficiency and minimize off-target transfection.

In some embodiments, the anchor fragment consists of three key components: a hydrophobic moiety that integrates into the LNP core, a polyethylene glycol (PEG) spacer extending outward from the LNP surface, and a terminal bio-orthogonal functional group that serves as the reactive site for ligand attachment. Representative examples of lipid-PEG anchor fragments include DSPE-PEG-N3, DMG-PEG-N3, DPPE-PEG-N3, and DOPE-PEG-N3, all of which enable stable integration into the LNP lipid bilayer while providing a reactive azide group for click chemistry.

In some embodiments, in addition to lipid-based anchors, the anchor fragment may also comprise hydrophobic polymers. In some embodiments, the anchor fragment consists of a polymer-PEG system rather than a lipid-PEG system. Suitable examples of polymer-based anchor fragments include PLA-PEG-N3, PCL-PEG-N3, and PLGA-PEG-N3, which provide an alternative anchoring mechanism and allow for additional tunability in nanoparticle physicochemical properties. As demonstrated in Example 19, the delivery efficiency of PLA-based anchor fragments was evaluated alongside DSPE-based anchor fragments. The study found that while both anchor types supported targeted delivery, DSPE-PEG-N3 exhibited superior conjugation efficiency and lower off-target effects compared to PLA-PEG-N3. The results indicate that while polymer-PEG anchors can be effective, clickable lipid-PEG anchors are more advantageous in enhancing ligand presentation and bio-orthogonal conjugation efficiency.

In some embodiments, the PEG spacer can vary in molecular weight, typically ranging from 500 Da to 10,000 Da. In some embodiments, the PEG spacer has a molecular weight between 1,000 Da and 5,000 Da, with specific implementations including PEG2K, PEG3K, PEG3.4K, PEG4K, and PEG5K. The selection of PEG length directly influences the accessibility of the ligand for receptor binding and affects overall targeting efficiency. Example 20 explored the effect of different PEG lengths on targeting efficiency, where lipid-PEG and clickable lipid-PEG molecules ranging from 2K to 5K were all viable for targeted delivery

In some embodiments, the anchor fragment is functionalized with a reactive moiety (moiety X) to enable bio-orthogonal conjugation, which reacts selectively with a complementary moiety (moiety Y) on the ligand-linker conjugate. Suitable functional groups for moiety X include azide (-N₃) , alkyne (-C≡C-) , tetrazine, trans-cyclooctene (TCO) , cyclopropene, and strained alkyne derivatives. In some embodiments, the anchor fragment is further modified with anionic or cationic groups to optimize conjugation efficiency and improve the physicochemical stability of the LNP formulation. Anionic modifications include sulfonate (-SO₃ ^-) , carboxylate (-COO^-) , and phosphate (-PO₄ ^2-) groups, which enhance solubility and reduce non-specific aggregation. Cationic modifications include ammonium (-NH₄ ⁺, -NR₃ ⁺) , guanidinium (-C (=NH) NH₂ ⁺) , and imidazolium (-C₃H₄N₂ ⁺) groups, which can facilitate electrostatic interactions and modulate nanoparticle charge properties.

In some embodiments, the anchor fragment comprises between 0.01 and 5 mol%of the total LNP composition. In some embodiments, the anchor fragment comprises between 0.1 and 2 mol%, and in some cases, between 0.25 and 1 mol%. The ratio of the anchor fragment to the total PEG-lipid content significantly impacts conjugation efficiency. In some embodiments, the combined content of anchor fragment and PEG-lipid is between 0.5 and 2 mol%, while the anchor/total PEG-lipid ratio remains no less than 10%. In some embodiments, a higher anchor/total PEG-lipid ratio of at least 25%is maintained to optimize ligand conjugation and delivery efficiency.

In some embodiments, said anchor fragment comprises the formula selected from:

wherein n ranges from 0 to 135, preferably 22 to 117, more preferably 45 to 90.
Ionizable cationic lipids

The present disclosure provides a group of ionizable cationic lipids, which are suitable for preparing lipid nanoparticles described herein.

In addition to anchor fragments, the disclosed lipid nanoparticle (LNP) compositions incorporate ionizable lipids, which play a fundamental role in nucleic acid encapsulation, endosomal escape facilitation, and delivery efficiency optimization. Ionizable cationic lipids are essential components in lipid nanoparticle formulations. Typical structure of cationic lipids includes one or more amine group (s) which bear the positive charge. The ionizable cationic lipids can exist in a positively charged or neutral form depending on pH. The ionization of the cationic lipid affects the surface charge of the lipid nanoparticle under different pH conditions. This charge state can influence plasma protein absorption, blood clearance, and tissue distribution (Semple, S. C., et al., Adv. Drug Deliv Rev 32: 3-17 (1998) ) as well as the ability to form endosomolytic non-bilayer structures (Hafez, I. M., et al., Gene Ther 8: 1188-1196 (2001) ) critical to the intracellular delivery of nucleic acids.

In some embodiments, the lipid nanoparticle comprises an ionizable lipid in an amount ranging from approximately 20-90 mol%of the total lipid content. In preferable embodiments, the lipid nanoparticle comprises the ionizable lipid is in an amount ranging from approximately 20-70 mol%, 30-60 mol%, or 40-50 mol%of the total lipid content. In certain embodiments, the proportion of ionizable lipid is optimized between 50 mol%and 60 mol%to maintain a balance between particle stability and effective nucleic acid delivery. The ratio of total lipid content to nucleic acid content is also adjustable to achieve a desired nitrogen-to-phosphate (N/P) ratio, which can range from approximately 3 to 10 or higher. By modifying this ratio, the electrostatic interactions between the ionizable lipids and nucleic acid cargo can be fine-tuned, thereby influencing encapsulation efficiency, particle stability, and transfection performance.

The structural diversity of ionizable lipids allows for precise control over the physicochemical properties and performance characteristics of LNPs. In some embodiments, the ionizable lipids used in the present disclosure include, but are not limited to, alkylated amines, imidazolium-based lipids, guanidinium-functionalized lipids, piperazine-based lipids, and lipids incorporating ester, amide, or other functional groups. These ionizable lipids typically contain tertiary or quaternary amine groups that remain uncharged at physiological pH (～7.4) but become protonated in the acidic environment of endosomes, thereby facilitating endosomal escape via the proton sponge effect.

In some embodiments, the ionizable lipid is an alkylated amine-based lipid, containing primary, secondary, or tertiary amine groups that modulate charge interactions with nucleic acids. Representative examples include DLin-MC3-DMA, ALC-0315, and SM-102, which are widely employed in nucleic acid delivery. In other embodiments, the ionizable lipid is imidazolium-based, featuring an imidazolium ring system that undergoes pH-sensitive protonation and exhibits strong electrostatic interactions with nucleic acids, thereby providing enhanced stability under physiological conditions and promoting efficient cargo release upon cellular uptake.

In some embodiments, the ionizable lipid comprises guanidinium-functionalized groups, leveraging the strong hydrogen bonding capacity of guanidinium moieties to enhance RNA binding affinity and transfection efficiency. In other embodiments, the ionizable lipid is piperazine-based, incorporating a piperazine ring to provide a tunable pKa range and enhance membrane fusion properties, thereby supporting efficient endosomal escape and intracellular delivery.

In some embodiments, the ionizable lipid is ester-linked or amide-linked, where ester-linked amino head groups or amide-based backbones introduce biodegradability into the formulation. Such biodegradable lipid structures allow for controlled metabolism and clearance from the body, reducing potential toxicity while maintaining efficient nucleic acid delivery. The selection of ionizable lipids in the disclosed LNP formulations is based on their capacity to optimize encapsulation efficiency, enhance cellular uptake, and improve overall therapeutic efficacy.

The present disclosure also provides a lipid compound, comprising an amino head group, and two fatty acid or fatty alkyl tails, wherein at least one tail is a branched tail comprising an acetal group, wherein the carbon atom of the acetal group serves as the branching point of the branched tail and the two ether oxygens of the acetal group are connected to two hydrocarbyl chains.

In some embodiments, the ionizable cationic lipids of the present application is a compound having formula (I) , or a salt, tautomer, or stereoisomer thereof,

wherein:
m and p are independently selected from any integer ranging from 3 to 8;
n is selected from any integer ranging from 2 to 4;
X is a bond, -C (O) O-, -OC (O) -, -OC (O) O-, or a biodegradable group;
R₁ is a hydrogen bond donor-containing group or hydrogen bond acceptor-containing group;
both of R₂ are same and selected from C₁-C₁₂ alkyl, substituted C₁-C₁₂ alkyl, C₂-C₁₂ alkenyl, substituted C₂-C₁₂ alkenyl, C₃-C₁₂ cycloalkyl and substituted C₃-C₁₂ cycloalkyl and combinations of thereof;
R₃ is selected from C₄-C₂₂ alkyl, substituted C₄-C₂₂ alkyl, C₄-C₂₂ alkenyl, substituted C₄-C₂₂
alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof; or R₃ is an acetal group ofwherein both of R₄ are same and selected from C₁-C₁₆ alkyl, substituted C₁-C₁₆ alkyl, C₂-C₁₆ alkenyl, substituted C₂-C₁₆ alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof.

In some embodiments, the lipid compound of the first aspect is a compound having formula (II) , or a salt, tautomer, or stereoisomer thereof,

wherein q is selected from any integer ranging from 2 to 4.

In the structure as shown by Formula (I) and Formula (II) , R₁ and the central nitrogen atom (N) constitutes the head group of the lipid compound of the present application. Head group of a cationic lipid is positively charged at acidic environment and is essential for the delivery of nucleic acid payloads, as it interacts with the negatively charged phosphate groups of nucleic acids. Unlike double-stranded DNA or siRNA, in which only the phosphate backbone is exposed while the nucleobases are facing inward, mRNA has a more complex secondary structure, comprising single-and double-stranded regions with some nucleosides exposed to the surrounding environment, such as the solvent. Therefore, other interactions, such as hydrogen bonding, of the head group with the riboses and nucleobases of mRNA could contribute to an improved LNP expression. In the structure as shown by Formula (I) and Formula (II) , R₁ serves to provide a set of hydrogen bond donors and/or acceptors with varying hydrogen-bonding ability in the head group.

H-bond donor-or acceptor-containing groups can be those comprising nitrogen (N) , oxygen (O) , carbon (C) or fluorine (F) atoms serving as H-bond donor or H-bond acceptor. Exemplary H-bond donor-or acceptor-containing groups may include but not limited to hydroxyl-containing group, carboxyl-containing group, carbonyl-containing group, amide, imide, sulfoxide, and sulfonamide.

In preferred embodiments of the present application, R₁ is selected from a group consisting of following formulae:

wherein o is selected from 1, 2, 3, 4, and 5.

In specific embodiments, R₁ isand o is selected from 2, 3, or 4, preferably 2.

Aside from the headgroup, the ionizable cationic lipid of the present application comprises two hydrophobic tails. At least one of the two tails is a branched tail, which comprises an acetal group connecting two identical branching chains, resulting in a “symmetric acetal” .

In the compound of formula (II) , both tails are symmetric acetals and are represented by R₂ and R₄ groups, respectively. R₂ and R₄ are each independently selected from C₁-C₁₂ alkyl, substituted C₁-C₁₂ alkyl, C₂-C₁₂ alkenyl, substituted C₂-C₁₂ alkenyl, C₃-C₁₂ cycloalkyl and substituted C₃-C₁₂ cycloalkyl. For each of substituted alkyl and substituted alkenyl, it can comprise one or more substituents selected from deuterium (–D) , cyano (–CN) , halo, imino (=NH) , nitro (–NO₂) , oxo (=O) . In preferred embodiments, R₂ and R₃ are independently selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl and C₂-C₁₂ alkenyl, preferably C₃-C₁₀ alkyl, unbranched or branched. R₂ and R₄ can be same or different. In certain embodiments, R₂ and/or R₄ is linear alkyl. In preferred embodiments, at least one of R₂ and/or R₄ is branched. As a hyperbranched structure may augment the cone-shaped morphology, and therefore increase intracellular delivery efficiency, it is preferred in the present application. Accordingly, both R₂ and R₄ are preferably branched. For example, R₂ and/or R₄ may have one or more side chains. In specific embodiments, R₂ and R₄ are independently selected from 2-methylhexyl or 2, 6-dimethyloctyl.

In the compound of formula (I) , only one tail is a symmetric acetal and is represented by R₂ group. R₂ is selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl and C₂-C₁₂ alkenyl, preferably C₃-C₁₀ alkyl, unbranched or branched. As a hyperbranched structure is preferred in the present application, R₂ preferably is a branched chain. For example, R₂ may have one or more side chains. In specific embodiments, R₂ is selected from 2, 6-dimethyloctyl or 5-butylheptyl. For the other tail, it is preferred that R₃ is also a branched chain, more preferably having two identical branching tails. In specific embodiments, R₃ is selected from 5-butylheptyl or 4-propylhexyl.

The branched tail containing acetal group could easily been synthesized by conjugate R₂-OH or R₄-OH to the acetal group. In specific embodiments, the intermediate structure containing acetal group were synthesized using the following reactions:

In specific embodiments, R₂-OH or R₄-OH are branched alcohols, and the ionizable cationic lipid contains an hyperbranched structure. In specific embodiments, R₂-OH or R₄-OH are one of the following structures:

In specific embodiments, R₂-OH or R₄-OH are unsaturated alcohols, and the ionizable cationic lipid contains an unsaturated tail.

In the acetal-based branched tail, the part between the acetal branching position and the central nitrogen is referred as a “stem” region. The stem region comprises a linear chain comprising a biodegradable group.

The biodegradable group in the “stem” region can be -OC (O) -, -C (O) O-or -OC (O) O-. The biodegradable group can be cleaved in a biological environment. When the LNP formed by the ionizable lipid comprising the biodegradable group (s) is delivered into a biological environment, such as into a living organism, the biodegradable groups facilitate the metabolism and clearance of the lipid components after the payload has been delivered.

The length of the stem region is determined by the chain length before and after the biodegradable group. In some embodiments, m and p in Formula (II) is independently 3, 4, 5, 6, 7 or 8. In preferred embodiments, m is 4, 5 or 6, p is 5, 6 or 7. In some embodiments, n and q in Formula (II) is independently 2, 3, or 4. In most preferred embodiment, m = 5, n = 2, p = 5 or 7, q = 2, and X is -C (=O) O-, and the ionizable cationic lipid have one of the following structures:

In formula (II) , only one of the two tails is acetal group containing branched tails. In some embodiments, n and q in Formula (II) is 2, 3, or 4. In most preferred embodiment, m = 5, n = 2, and X is -C (=O) O-.

The compound of formula (I) has a non-acetal tail represented by the following formula: wherein R₃ is selected from C₆-C₂₂ alkyl, C₆-C₂₂ cycloalkyl or C₆-C₂₂ alkenyl, preferably C₆-C₁₂ alkyl, C₆-C₁₂ cycloalkyl or C₆-C₁₂ alkenyl, unbranched or branched.

R₃ can be linear or branched. In some specific embodiments, R₃ is a branched C₆-C₂₂ alkyl or a branched C₆-C₂₂ alkenyl, preferably a branched C₆-C₁₂ alkyl or a branched C₆-C₁₂ alkenyl. In some specific embodiments, when R₃ is a branched alkyl or alkenyl, R₃ could be the same with R₂. In some specific embodiments, R₃ is a linear C₆-C₂₂ alkyl or C₆-C₂₂ alkenyl, preferably a linear C₈-C₂₀ alkyl or a linear C₈-C₂₀ alkenyl, more preferably linear C₉-C₁₈ alkyl or linear C₉-C₁₈ alkenyl.

In specific embodiments, the lipid compound is selected from any one of following Compounds as shown in Table 1, including Compounds 002-011 and 013-051 which have a chemical structure of formula (II) , and Compounds 052-103 which have a chemical structure of formula (I) . In preferred embodiments, the lipid compound has a chemical structure of formula (II) and is selected from Compound 002 and Compound 030 in Table 1. In preferred embodiments, the lipid compound has a chemical structure of formula (I) and is selected from Compound 059, Compound 061, Compound 062, Compound 065 in Table 1.
Other lipid components of LNP

In addition to the ionizable cationic lipid, the lipid components comprised in the LNPs of the present application comprises a neutral lipid, a negative lipid, a polymer conjugated lipid, and a steroid.

Neutral lipid

The neutral lipid such as a phospholipid helps the LNP to bind to and cross the cell membrane. Exemplary neutral lipids include, but are not limited to, distearoyl-sn-glycero-phosphoethanolamine, distearoylphosphatidylcholine (DSPC) , dioleoylphosphatidylcholine (DOPC) , dipalmitoylphosphatidylcholine (DPPC) , dioleoyl-phosphatidylethanolamine (DOPE) , palmitoyloleoylphosphatidylcholine (POPC) , palmitoyloleoylphosphatidylethanolamine (POPE) , dioleoyl-phosphatidylethanolamine 4- (N-maleimidomethyl) -cyclohexane-1-carboxylate (DOPE-mal) , dipalmitoyl phosphatidyl ethanolamine (DPPE) , dimyristoylphosphoethanolamine (DMPE) , di stearoyl-phosphatidyl-ethanolamine (DSPE) , monomethyl-phosphatidylethanolamine (such as 16-O-monomethyl PE) , dimethyl-phosphatidylethanolamine (such as 16-0-dimethyl PE) , 18-1-trans PE, 1-stearoyl-2-oleoyl-phosphatidyethanolamine (SOPE) , hydrogenated soy phosphatidylcholine (HSPC) , egg phosphatidylcholine (EPC) , sphingomyelin (SM) , dimyristoyl phosphatidylcholine (DMPC) , dierucoylphosphatidylcholine (DEPC) , dielaidoyl-phosphatidylethanolamine (DEPE) , lecithin, phosphatidylethanolamine, lysolecithin, lysophosphatidylethanolamine, phosphatidylserine, phosphatidylinositol, sphingomyelin, egg sphingomyelin (ESM) , cephalin, cardiolipin, phosphatidicacid, cerebrosides, dicetylphosphate, lysophosphatidylcholine, dilinoleoylphosphatidylcholine, or mixtures thereof. It is understood that other phosphatidylcholine phospholipids can also be used. The acyl groups in these lipids are preferably acyl groups derived from fatty acids having C10-C24 carbon chains, e.g., lauroyl, myristoyl, palmitoyl, stearoyl, or oleoyl. In various embodiments, the molar ratio of the ionizable cationic lipid to the neutral lipid ranges from about 2: 1 to about 8: 1, preferably 5: 1.

Steroid

The steroid or steroid analogue. In certain embodiments, the steroid or steroid analogue is cholesterol, sitosterol or stigmasterol. In certain embodiments, the steroid or steroid analogue is cholesterol. In certain embodiments, the steroid or steroid analogue is beta-Sitosterol. In certain embodiments, the steroid or steroid analogue is Stigmastanol. In some of these embodiments, the molar ratio of the ionizable cationic lipid to cholesterol ranges from about 5: 1 to 1: 1, preferably 3: 1 to 1: 1.

Polymer conjugated lipid

In some embodiments, polymer-conjugated lipids serve as essential structural components that provide steric hindrance, thereby reducing opsonization and premature clearance by the immune system. In some embodiments, the polymer-conjugated lipid is a polyethylene glycol (PEG) lipid, also referred to as a PEGylated lipid. The PEGylated lipid comprises a hydrophobic lipid, such as a phospholipid, ceramide, or sterol derivative, conjugated to a PEG chain of varying molecular weight. In some embodiments, the PEGylated lipid is selected from the group consisting of pegylated diacylglycerol (PEG-DAG) , pegylated phosphatidylethanolamine (PEG-PE) , PEG succinate diacylglycerol (PEG-S-DAG) , pegylated ceramide (PEG-cer) , or a PEG dialkoxypropylcarbamate. Examples of specific PEGylated lipids include 1- (monomethoxy-polyethyleneglycol) -2, 3-dimyristoylglycerol (PEG-DMG) and 4-O- (2′, 3′-di (tetradecanoyloxy) propyl-1-O- (ω-methoxy (polyethoxy) ethyl) butanedioate (PEG-S-DMG) . The molecular weight of PEG in the polymer-lipid conjugate can be varied to achieve different steric shielding effects, typically ranging from 500 Da to 10 kDa, with preferred embodiments utilizing PEG chains of 2K, 3.4K, or 5K to optimize circulation time and prevent nanoparticle aggregation.

As demonstrated in Example 21, PEGylated lipids with different hydrophobic moieties are capable of supporting targeted delivery, including DSPE-PEG, DMG-PEG, and C14-TPA-PEG (N- (methoxy polyethylene glycol) -3- (4- (phenyl (4-tetradecylphenyl) amino) phenyl) propanamide) . These three classes represent PEGylated lipids with varying hydrophobic segments, namely long-chain lipids (C18, as in DSPE-PEG) , short-chain lipids (C14, as in DMG-PEG) , and aromatic lipid structures incorporating benzyl groups (as in C14-TPA-PEG) . In some embodiments, DSPE-PEG is preferred.

In some embodiments, alternative polymer-lipid conjugates are employed to enhance nanoparticle stability and modulate immune recognition. While PEG remains the most widely used polymer for conjugation, other hydrophilic polymers, such as poly (2-oxazoline) (POx) , poly (glycerol) (PG) , and poly (N- (2-hydroxypropyl) methacrylamide) (pHPMA) , have demonstrated potential advantages in reducing immune activation and improving circulation time. In some embodiments, polymer-conjugated lipids comprising zwitterionic moieties, such as phosphorylcholine-based lipids, are incorporated to minimize protein corona formation and improve nanoparticle stealth properties.

The selection of PEG length and density directly influences nanoparticle behavior in biological systems. In some embodiments, the PEG chain length is adjusted to control steric hindrance and ligand accessibility. Shorter PEG chains (e.g., PEG2K) provide moderate steric shielding while maintaining accessibility for ligand conjugation, whereas longer PEG chains (e.g., PEG5K) offer superior circulation stability but may hinder ligand accessibility. In some embodiments, the molar ratio of PEGylated lipids within the LNP formulation ranges from approximately 0.1-5 mol%, with preferred formulations incorporating approximately 0.5 mol%to 2 mol%PEG-lipid to balance stability and functionality.

In some embodiments, the incorporation of cleavable or degradable PEG-lipid conjugates is utilized to enhance intracellular delivery efficiency. Conventional PEGylation provides stability in circulation but may hinder cellular uptake and endosomal escape. To address this limitation, cleavable PEG-lipid conjugates incorporating disulfide, ester, or pH-sensitive linkages are introduced, allowing for PEG detachment under specific intracellular conditions. In some embodiments, stimuli-responsive PEG-lipid conjugates are designed to degrade in response to redox conditions, enzymatic activity, or acidic pH, thereby facilitating ligand exposure and improving endosomal escape efficiency.

Despite the advantages of PEGylation, repeated exposure to PEGylated nanoparticles has been associated with the accelerated blood clearance (ABC) effect, characterized by the rapid clearance of PEGylated formulations upon re-administration due to the formation of anti-PEG antibodies. In some embodiments, alternative polymer-lipid conjugates, such as zwitterionic polymers or POx-lipid conjugates, are incorporated to mitigate immunogenicity while retaining the steric stabilization benefits of polymer conjugation. These alternative polymer systems offer promising strategies to circumvent ABC-related limitations, providing sustained therapeutic efficacy upon repeated dosing.

In some embodiments, the composition of PEGylated lipids within LNPs is optimized to enhance the efficiency of ligand conjugation. The ratio of anchor fragments to total PEG-lipid content plays a critical role in determining ligand accessibility and conjugation efficiency. In some embodiments, the total PEG-lipid content in LNP formulations is maintained between 0.5 mol%and 2 mol%, with an anchor/PEG-lipid ratio no less than 10%. In some embodiments, the anchor/PEG-lipid ratio is maintained at no less than 25%to maximize ligand conjugation efficiency without compromising nanoparticle stability. These parameters ensure the optimal presentation of conjugation sites while preserving the structural integrity of the LNP system.

The clickable PEG amphiphilic polymer

The clickable PEG amphiphilic polymer can be a click group-PEGylated-linker-lipid or hydrophobic polymer. In various embodiments, the click group can be azide, DBCO, N3, TCO, tetrazine. The lipid is a DSPE, DPG, DPPE, DPG and other lipid. In various embodiments, the hydrophobic polymer is poly (lactic-co-glycolic acid) or poly (lactic acid) . The preferable clickable PEG conjugated lipid chose from those described herein. The molar ratio of the clickable PEG conjugated lipid ranges from about 0.1-1%.

In preferred embodiments, the nanoparticle of the present disclosure comprises the following in a mole ratio of ionizable lipid: DSPC: chol was 50: 10: 38 or 40: 10: 48, wherein total mole ratio is 98-99.5%: 1) the ionizable lipid of the present disclosure, preferable selected from table1; 2) a neutral lipid selected from DSPC or DOPE; 3) a steroid which is cholesterol.
Nucleic acid payload

The LNPs of the present application are particularly suitable for delivering nucleic acid molecules into cells.

The nucleic acid molecule can be either DNA or RNA or a mixture thereof, such as chimeric oligonucleotides. The nucleic acid molecule can comprise naturally occurring or modified polynucleotides. The nucleic acid molecule can be a coding sequence or non-coding sequence. The nucleic acid molecule can be DNA molecule such as plasmid DNA, closed-ended DNA or a mixture thereof. The nucleic acid molecule can be RNA molecule such as messenger RNA (mRNA) , guide RNA (gRNA) , a short interfering RNA (siRNA) , an RNA interference (RNAi) molecule, a microRNA (miRNA) , an antagomir, an antisense RNA, a ribozyme, a small hairpin RNA (shRNA) , or a mixture thereof.

Plasmid DNA or closed-ended DNA may have a length at a range of 500 to 500,000 base pairs. mRNA may have a length at a range of 200 to 100,000 base pairs. gRNA may have a length at a range of 30 to 1,000 base pairs. siRNAi, miRNA, or shRNA may have a length at a range of 15 to 1,000 base pairs.

The nucleic acid molecules can comprise modifications, such as one or more modifications to the backbone, one or more modifications to the base and/or one or more modifications to the sugar moiety.
Bio-orthogonal Click reaction

Bio-orthogonal click reaction offers numerous advantages, including the elimination of the need for additional reagents or organic solvents. It occurs efficiently in aqueous medium under physiological conditions of pH and temperature, and results in the formation of stable products without unwanted byproducts so that purification is not necessary. The reactants and products of bio-orthogonal click reaction will not react or interact with biological molecules. These characteristics of bio-orthogonal click reaction make it become a useful tool across various applications including imaging, drug delivery, 3D cell culture, and bioprinting. A number of bio-orthogonal click reactions have been developed, mainly including the Michael addition reaction between a thiol and an alkene (e.g., thiol-maleimide) , strain-promoted (Cu-free) azide-alkyne cycloaddition (SPAAC) , and the inverse-electron-demand Diels–Alder (IEDDA) reaction. The SPAAC reaction results in the formation of triazole, which is chemically stable and is not subject to enzymatic cleavage, making the bio-click chemistry product highly stable in biological systems. Cyclooctyne is an 8-carbon ring structure comprising an internal alkyne bond. The closed ring structure induces a substantial bond angle deformation of the acetylene, which is highly reactive with azide groups to form a triazole. To address the slow rate of the original cyclooctyne reaction, electron-withdrawing groups are attached adjacent to the triple bond. Examples of such substituted cyclooctynes include difluorinated cyclooctynes, azacyclooctyne and dibenzoazacyclooctyne (DBCO) . The reaction of DBCO with azide was reported to have exceptionally fast reaction kinetics and was used in a one-pot three-step protocol for site-specific modification of peptides and proteins. The Diels-Alder reaction has also been used for in vivo labeling of molecules. Rossin et al. reported a 52%yield in vivo between a tumor-localized anti-TAG72 (CC49) antibody carrying a trans-cyclooctene (TCO) reactive moiety and an ¹¹¹In-labeled tetrazine DOTA derivative.
Nanobodies

Nanobodies are single-domain antibodies of about 12-15kDa in size (about 110 amino acids in length) . Nanobodies can selectively bind to target antigens, like full-size antibodies, and have similar affinities for antigens. However. because of their much smaller size, they may be capable of better penetration into solid tumors. The smaller size also contributes to the stability of the nanobody, which is more resistant to pH and temperature extremes than full size antibodies. Single-domain antibodies were originally developed following the discovery that camelids (camels. alpacas, llamas) possess fully functional antibodies without light chains. The heavy-chain antibodies consist of a single variable domain (VHH) and two constant domains (CH2 and CH3) . Like antibodies, nanobodies may be developed and used as multivalent and/or bispecific constructs. Humanized forms of nanobodies are in commercial development that are targeted to a variety of target antigens, such as IL-6R, VWF, TNF, RSV RANKL, IL-17A &F and IgE (e. g, ABLYNXR, Ghent, Belgium) , with potential clinical use in cancer and other disorders

The plasma half-life of nanobodies is shorter than that of full-size antibodies, with elimination primarily by the renal route. Because they lack an Fc region, they do not exhibit complement dependent cytotoxicity.
Antibody Fragments

Antibody fragments are antigen binding portions of an antibody, such as F (ab') ₂, Fab', F(ab) ₂, Fab, Fv, sFv, scFv and the like. Antibody fragments which recognize specific epitopes can be generated by known techniques. Antibody fragments can be produced by digestion of the antibody molecule. For example, an approximate 100 kD fragment F (ab') ₂ fragment can be produced by enzymatic cleavage of antibodies with pepsin. This fragment can be further cleaved using a thiol reducing agent, and optionally a blocking group for the sulfhydryl groups resulting from cleavage of disulfide linkages, to produce an approximate 50 Kd Fab' monovalent fragment. Alternatively, an enzymatic cleavage using papain produces two monovalent Fab fragments and an Fc fragment directly. Other methods of cleaving antibodies, such as separation of heavy chains to form monovalent light-heavy chain fragments, further cleavage of fragments, or other enzymatic, chemical or genetic techniques may also be used, so long as the fragments bind to the antigen that is recognized by the intact antibody. A single chain Fv molecule (scFv) comprises a VI domain and a VH domain. The VL and VH domains associate to form a target binding site. These two domains are further covalently linked by a peptide linker (L) . A scFv molecule is denoted as either VL-L-VH if the VL domain is the N-terminal part of the scFv molecule, or as VH-L-VL if the VH domain is the N-terminal part of the scFv molecule. Methods for making scFv molecules and designing suitable peptide linkers are described.
Preparation of Nanoparticles

Nanoparticles can be made with mixing processes such as microfluidics and T-tube mixing of two fluid streams, one of which contains the therapeutic and/or prophylactic and the other has the lipid components.

Lipid compositions are prepared by combining a ionizable lipid, such as lipid according to formula (I) , helper lipids (such as DSPC) , a steroid (such as cholesterol) , a PEG lipid (such as DSPE-PEG2000) , and an anchor segment (such as DSPE-PEG2000-N3) , at concentrations of about 5 mg/mL and 25 mg/mL in ethanol. Solutions should be refrigerated for storage at, for example, -80 ℃.

Nanoparticles compositions including a therapeutic and/or prophylactic and a lipid component are prepared by combining the lipid solution with a solution including the therapeutic and/or prophylactic at lipid component to therapeutic and/or prophylactic wt: wt ratios between about 10: 1 about 20: 1. The lipid solution is injected using a microfluidic based system at flow rates between 0.25 mL/min and 2.0 mL/min into the therapeutic and/or prophylactic solution to produce a suspension with a water to ethanol ratio between about 3: 1.

For nanoparticle compositions including RNA or pDNA, solutions of the RNA or pDNA are diluted with 50 mM sodium citrate buffer at a pH between 3 and 4 to form a stock solution at concentrations of 0.1-0.5 mg/mL.

Nanoparticle compositions can be processed by dialysis to remove ethanol and achieve buffer exchange. Formulations were dialyzed twice against Tris-HCl Solution (20 mM, pH 7.4) at volumes about 2500 times that of the primary product using Slide-A-Lyzer cassettes (Thermo Fisher Scientific Inc., Rockford, Ill. ) with a molecular weight cutoff of 100 KD at 0-4 ℃ for 2 h. Then 40%sucrose solution was added, the final nanoparticle composition solution of 0.03 mg/mL to and 0.1 mg/mL (mRNA or pDNA) are generally obtained and stored at -80 ℃.
Characteristics of Nanoparticle Compositions

The physical and chemical properties of the lipid nanoparticle of the present application depend on the formulation of the LNPs.

For example, it has been found by the present inventors that the choice of the ionizable cationic lipid dramatically influences the size of the formulated LNPs. For example, the lipid nanoparticles have a mean diameter of from about 100 nm to about 350 nm, preferably 150 nm to 250 nm. In specific embodiments, the lipid nanoparticles have a mean diameter of about 100 nm, 110 nm, 120 nm, 130 nm, 140 nm, 150 nm, 160 nm, 170 nm, 180 nm, 190 nm, 200 nm, 210 nm, 220 nm, 230 nm, 240 nm, 250 nm, 260 nm, 270 nm, 280 nm, 290 nm or 300 nm.

Polydispersity index (PDI) is a parameter describing the distribution of molecular weight of the particles in an LNP formulation. A lager PDI describes a broader distribution of molecular weight. For example, the PDI of the formulation can be 0.10 to 0.50, preferably 0.20 to 0.45, more preferably 0.05 to 0.30.

In some embodiments, the lipid nanoparticles are substantially non-toxic or having acceptable toxicity to the subject. For example, the lipid nanoparticles of the present application have a reduced liver toxicity as compared to nanoparticles comprising a different ionizable lipid. The liver toxicity can be measured based known method in the art.

The lipid nanoparticle of the present disclosure can provide desirable transduction efficiency, which can be measured by e.g. the amount or expression level of the payload. In preferred embodiments, the lipid nanoparticles of the present application provide a comparable or increased amount of payload or expression of payload as compared to a nanoparticle comprising a different ionizable lipid.
Uses

In some embodiments, ligand-functionalized LNPs facilitate the targeted delivery of gene-editing components to specific cell populations, enabling precise genomic modifications for therapeutic applications. The disclosed tLNPs can be utilized for the in vivo and ex vivo delivery of various gene-editing nucleases, including CRISPR-associated nucleases (Cas9, Cas12, and Cas13) , RNA-guided base editors, prime editors, and transposon-based genome integration systems. In any of the foregoing embodiments, the gene-editing components may be delivered as mRNA encoding the nuclease, guide RNA, or a pre-assembled ribonucleoprotein (RNP) complex, depending on the intended therapeutic application.

In some embodiments, the disclosed tLNPs are particularly optimized for the delivery of CRISPR-Cas12b, a thermophilic class II nuclease with high specificity and low off-target activity. As demonstrated in Example 28, ligand-conjugated LNPs encapsulating Cas12b mRNA and guide RNA achieved efficient gene editing in target cells while maintaining minimal off-target effects. The optimized conjugation strategy enables targeted intracellular release of CRISPR components, reducing systemic toxicity and improving editing efficiency in primary cells. In any of the foregoing embodiments, ligand-conjugated LNPs may also be used to deliver newly emerging genome engineering technologies, including R2 retrotransposon-derived genome writers, transposase-based systems such as piggyBac and Sleeping Beauty, as well as integrase-based gene insertion technologies. These approaches expand the range of gene-editing applications beyond simple knockouts and enable precise site-specific insertions for therapeutic gene correction.

In some embodiments, ligand-conjugated LNPs can be used to edit hematopoietic stem and progenitor cells (HSPCs) , thereby enabling durable gene correction for inherited blood disorders. For example, the disclosed LNPs can be engineered to deliver CRISPR-Cas12b mRNA and guide RNA targeting HBG1/2 regulatory elements to reactivate fetal hemoglobin expression, providing a potential cure for β-thalassemia and sickle cell disease. The ability of ligand-conjugated LNPs to selectively target HSCs while avoiding non-specific uptake in other cell populations represents a significant advancement over traditional non-targeted gene-editing approaches.

In some embodiments, the disclosed ligand-conjugated LNPs enable the targeted delivery of gene-modifying components to immune cells, facilitating ex vivo and in vivo cell therapy applications. Engineered T cells, B cells, natural killer (NK) cells, and dendritic cells are widely used in immunotherapy, particularly for cancer and autoimmune disorders. However, traditional methods for genetic modification rely on viral vectors, which have limitations in manufacturing complexity, immunogenicity, and insertional mutagenesis risks. The present disclosure provides a non-viral alternative for precise gene delivery to immune cells, enabling rapid and scalable cell engineering.

In some embodiments, ligand-conjugated LNPs are used for in vivo CAR-T and CAR-NK cell engineering. The disclosed LNPs can be formulated to deliver mRNA encoding chimeric antigen receptors (CARs) to T cells or NK cells, thereby generating potent immune effector cells without the need for ex vivo manipulation. Example 23 demonstrates the targeted delivery of CD7-targeting LNPs to T and NK cells, achieving selective transfection with minimal monocyte uptake. This approach may be used to generate autologous or allogeneic CAR-T/NK cells in vivo, potentially reducing the time and cost associated with current cell therapy manufacturing processes.

In some embodiments, ligand-conjugated LNPs facilitate the delivery of cytokine mRNAs, immune checkpoint inhibitors, or other immunomodulatory agents to enhance anti-tumor immunity. Targeting immune cells such as dendritic cells or macrophages can enhance antigen presentation and promote robust anti-tumor immune responses. The disclosed LNPs may be conjugated with ligands that bind to CD3, CD5, CD7, CD8, CD19, or CD56, allowing for precise delivery to specific immune subsets.

In some embodiments, the disclosed ligand-conjugated LNPs are designed to target hematopoietic stem and progenitor cells (HSPCs) for the treatment of inherited and acquired hematological disorders. HSPCs reside in the bone marrow and give rise to all blood cell lineages, making them an ideal target for gene therapy approaches. Traditional gene therapies often rely on ex vivo lentiviral transduction followed by autologous transplantation, which is costly and requires myeloablative conditioning. The present disclosure provides a non-viral alternative using ligand-conjugated LNPs to deliver therapeutic payloads directly to HSCs in vivo.

In some embodiments, ligand-conjugated LNPs are engineered to deliver gene-editing payloads that correct mutations associated with sickle cell disease, β-thalassemia, or Fanconi anemia. Example 26 and Example 27 demonstrate the targeted delivery of CD90-and CD117-functionalized LNPs to HSCs, achieving selective uptake in hematopoietic progenitors while avoiding off-target transfection in other blood cell populations. Such targeted approaches enable precision gene correction without requiring bone marrow transplantation, offering a minimally invasive treatment option for patients with inherited blood disorders.

In some embodiments, the disclosed ligand-conjugated LNPs enable tumor-specific delivery of therapeutic agents, including RNA-based therapies, small molecules, and gene-editing tools. One of the major challenges in oncology is achieving selective tumor targeting while minimizing systemic toxicity. The present disclosure provides a strategy for tumor-selective delivery by conjugating LNPs with ligands that bind to tumor-associated antigens (TAAs) or receptors overexpressed on cancer cells.

In some embodiments, ligand-conjugated LNPs are designed to target HER2, EGFR, PD-L1, or CD44, all of which are overexpressed in various solid tumors. By incorporating an optimized ligand-linker conjugation strategy, the disclosed LNPs achieve high tumor selectivity, reducing off-target effects in healthy tissues. Example 21 demonstrates the use of different lipid-PEG to optimize ligand orientation and stability, further improving tumor targeting efficiency.

In some embodiments, ligand-conjugated LNPs deliver siRNA, miRNA, or antisense oligonucleotides (ASOs) to downregulate oncogenes, inhibit immune evasion, or restore tumor suppressor function. For example, siRNA-loaded LNPs targeting MYC or KRAS can selectively silence oncogenic pathways in tumor cells, inhibiting proliferation and inducing apoptosis. The modularity of the disclosed conjugation platform allows for rapid adaptation of ligand-functionalized LNPs for various oncogenic targets, expanding their applicability across different cancer types.

In some embodiments, the disclosed ligand-conjugated LNPs are designed to facilitate brain-targeted delivery by overcoming the blood-brain barrier (BBB) . The BBB presents a significant challenge in neurotherapeutic development, limiting the effectiveness of systemically administered drugs. The present disclosure provides ligand-functionalized LNPs conjugated with BBB-penetrating ligands such as transferrin, insulin receptor ligands, or low-density lipoprotein receptor (LDLR) ligands, enabling efficient transport across the BBB.

In some embodiments, ligand-conjugated LNPs are used to deliver gene therapies or neuroprotective agents for treating Alzheimer’s , Parkinson’s , and amyotrophic lateral sclerosis (ALS) . By leveraging ligand-directed delivery, the disclosed LNPs ensure precise neuronal targeting, improving therapeutic efficacy while minimizing peripheral side effects.

In some embodiments, the disclosed ligand-conjugated LNPs provide a targeted approach for modulating immune responses in autoimmune and inflammatory diseases. Traditional immunosuppressive therapies often result in broad immune suppression, leading to increased susceptibility to infections and other complications. The present disclosure enables precision delivery of immunomodulatory agents directly to pathogenic immune cell subsets, minimizing systemic side effects while enhancing therapeutic efficacy.

In some embodiments, ligand-conjugated LNPs are designed to selectively target autoreactive T cells, B cells, or antigen-presenting cells (APCs) involved in autoimmune diseases such as multiple sclerosis, rheumatoid arthritis, and systemic lupus erythematosus (SLE) . Ligands specific for CD4, CD19, CD20, or CD86 can be used to direct LNPs carrying siRNA, antisense oligonucleotides (ASOs) , or mRNA-encoded immunoregulatory proteins to these immune cell populations. By silencing key inflammatory cytokines such as TNF-α, IL-6, or IFN-γ, ligand-functionalized LNPs can effectively dampen excessive immune activation without broadly suppressing immune function.

In some embodiments, ligand-conjugated LNPs are used for the selective depletion or reprogramming of pathogenic immune cells in autoimmune diseases. For example, LNPs carrying mRNA encoding chimeric antigen receptors (CARs) can be delivered to regulatory T cells (Tregs) in vivo, enhancing their suppressive capacity against autoreactive immune responses. Similarly, targeted LNPs delivering Fas ligand (FasL) mRNA or apoptosis-inducing siRNA can selectively deplete hyperactive B cells or pathogenic T cells, restoring immune homeostasis in conditions such as type 1 diabetes or inflammatory bowel disease.

In some embodiments, ligand-functionalized LNPs are employed to target fibrotic cell populations and deliver therapeutic payloads aimed at reversing fibrosis and restoring tissue function. Fibrosis is a key pathological feature of chronic liver disease, pulmonary fibrosis, renal fibrosis, and cardiac remodeling following myocardial infarction. Conventional antifibrotic drugs often lack cell specificity, leading to systemic toxicity and suboptimal efficacy. The present disclosure provides ligand-conjugated LNPs engineered to selectively target activated fibroblasts, hepatic stellate cells, or myofibroblasts, thereby enabling precision delivery of therapeutic agents that modulate fibrotic signaling pathways.

In some embodiments, ligand-functionalized LNPs are conjugated with ligands specific for fibroblast activation protein (FAP) , integrin αvβ6, or PDGFR-β, which are upregulated in activated fibroblasts. These targeted LNPs can deliver siRNA, miRNA, or small-molecule inhibitors that suppress fibrogenic pathways such as TGF-β signaling, collagen synthesis, or extracellular matrix deposition. For instance, Example 25 demonstrates the targeted delivery of ligand-functionalized LNPs to myofibroblasts, achieving significant downregulation of fibrotic markers while reducing collagen deposition in affected tissues.

In some embodiments, ligand-conjugated LNPs are designed for cardiovascular applications, including targeted delivery of gene therapies or RNA-based therapeutics to endothelial cells, vascular smooth muscle cells, or cardiac myocytes. For example, ligand-functionalized LNPs carrying mRNA encoding vascular endothelial growth factor (VEGF) can promote angiogenesis in ischemic tissues, while siRNA-based therapies targeting PCSK9 can modulate lipid metabolism to treat hypercholesterolemia.

In some embodiments, ligand-conjugated LNPs provide a targeted gene delivery platform for the treatment of rare genetic disorders that currently lack effective therapeutic options. Many monogenic diseases require lifelong enzyme replacement therapy or invasive gene therapy approaches, which present financial burdens for patients. The disclosed ligand-functionalized LNPs enable systemic or organ-specific delivery of nucleic acids to correct underlying genetic defects, providing a minimally invasive alternative to traditional therapies.

In some embodiments, ligand-functionalized LNPs are used to treat lysosomal storage disorders, such as Gaucher disease, Fabry disease, or Pompe disease, by delivering mRNA encoding functional lysosomal enzymes directly to affected cells. Unlike enzyme replacement therapy, which requires frequent intravenous infusions, ligand-conjugated LNPs enable sustained intracellular production of the missing enzyme, improving therapeutic outcomes while reducing treatment burden.

In some embodiments, ligand-conjugated LNPs are developed for metabolic disorders, such as phenylketonuria (PKU) or urea cycle disorders. By selectively delivering gene-editing tools, such as CRISPR base editors or prime editors, to hepatocytes, the disclosed LNPs enable precise correction of disease-causing mutations in metabolic genes, potentially offering a one-time cure.
Examples
Summary of Materials and Antibodies

Table 2 Material Information

Table 3 Antibody Information

Example 1. Synthesis of ligand-linker conjugates

This example demonstrates the synthesis of a ligand-linker structure employing a modified anti-CD5 VHH (clone 5H10-1) designed for site-specific conjugation. The original anti-CD5 VHH sequence was engineered to include a C-terminal extension that added a histidine-tag, a hinge sequence, and a cysteine residue. The modified sequence was transiently expressed in Expi293 cells and purified to achieve a final purity of 84.87%. The sequence alterations facilitated the specific attachment of a click chemistry handle, enhancing the conjugation potential without compromising the VHH’s internal structure. The modified sequence is detailed in Table 4.
Table 4. Original and modified antibody sequence

A schematic illustration of structure of the modified ligand, linker, and the ligand-linker conjugate was shown in Figure 1A. The conjugation process involved reducing the modified VHH (like the molecule with the sequence shown as SEQ ID NO: 2) with 2-mercaptoethanol (2-MEA) and subsequently purifying the reduced VHH using a Zeba^TM desalt column (7K MWCO) . This preparation was then conjugated with DBCO-PEG₄-Maleimide at room temperature for three hours. The targeted attachment was confirmed to occur at the newly introduced cysteine via a mild reaction that preserves the internal disulfide bonds of the modified VHH. After conjugation, ultrafiltration with a 3kD molecular weight cutoff was employed to remove free linker molecules and to exchange the buffer to 10 mM PBS (pH 7.4) . The purified ligand-linker conjugate was then stored at -80℃.

Characterization of the conjugate through Size-Exclusion Chromatography (SEC) and Liquid Chromatography-Mass Spectrometry (LC-MS) confirmed the purity and the specific incorporation of DBCO. As shown in Figure 1B, mass spectrometry analysis revealed that among the three cysteine residues present in the modified VHH sequence, only a single cysteine was conjugated with the linker, resulting in an antibody-to-linker ratio of 1: 1. It is noteworthy that, due to the use of the hinge-5 sequence (SEQ ID NO: 1) in SEQ ID NO: 2, which contains multiple serine (S) and threonine (T) residues within a proline-rich context, partial glycosylation was observed. In sequences enriched with proline, serine and threonine residues are susceptible to O-linked glycosylation. As a result, the mass spectrometry profile of the modified VHH displayed multiple peaks corresponding to different glycosylation states. Nevertheless, regardless of the glycosylation level, each VHH species exhibited a consistent antibody-to-linker conjugation ratio of 1: 1 after the conjugating reaction. This result underscores the efficacy of the proline-rich hinge in exposing the C-terminal cysteine for precise conjugation in a controlled reductive environment, thereby maintaining the structural and functional integrity of the VHH.
Example 2: Preparation of Anchor-Modified Lipid Nanoparticles

This example illustrates the preparation of anchor-modified lipid nanoparticles using a clickable polyethylene glycol derivative as an anchor fragment. The anchor fragment can be a clickable amphiphilic polymer, a clickable PEG-conjugated lipid, or a clickable PEG-conjugated hydrophobic polymer. By introducing an anchor fragment with a click handle, these anchor-modified LNPs can be contacted with a ligand-linker conjugate, thereby allowing for the efficient formation of the desired ligand-conjugated lipid nanoparticle.

In this example, the anchor fragment with a click chemistry group used was DSPE-PEG-N3, a PEG lipid with a terminal azide group that enables click chemistry reactions. DSPE-PEG-N3 was incorporated into the lipid mixture used for LNP formulation. The lipid components for the ethanol phase included an ionizable cationic lipid, DSPC (a neutral lipid) , cholesterol (a steroid) , DSPE-PEG2000 (a PEG lipid) , and DSPE-PEG2000-N3 (a clickable PEG lipid) . All lipid materials were sourced from Avanti Polar Lipids. The ethanol phase was prepared by dissolving the specified lipids in ethanol. This included the ionizable cationic lipid, DSPC, cholesterol, DSPE-PEG2000, and DSPE-PEG2000-N3. RNA was dissolved in 50 mM sodium citrate buffer (pH 4.0) at a concentration of 0.1 mg/mL. The weight ratio of RNA to the ionizable cationic lipid was maintained at 1: 10.

The LNP formulation was carried out using a microfluidic technique. The lipid solution (ethanol phase) was injected at a rate of 1 mL/min, while the aqueous phase containing the RNA was injected at 3 mL/min into the microfluidic mixer. The volume ratio of the aqueous to the ethanol phase was maintained at 3: 1. The resulting mixture facilitated the formation of LNPs. After mixing, the LNP solution was subjected to dialysis to remove the ethanol and exchange the buffer to PBS (10 mM, pH 7.4) . Dialysis was conducted twice against a PBS solution using Slide-A-Lyzer cassettes (Thermo Fisher Scientific Inc. ) with a molecular weight cutoff of 100 KD at 4℃ for 2 hours. The LNPs were then concentrated using ultracentrifugation.

Characterization of LNPs: Particle size and polydispersity index (PDI) were determined using a Zetasizer Nano ZS (BeNano, Bettersize) , performed in both PBS and Tris-HCl buffers to confirm stability across different media. The RNA concentration within the LNPs was quantified via ultraviolet-visible spectroscopy, which involved recording the absorbance spectrum and calculating the RNA content based on its specific extinction coefficient. The differences in absorbance at 260 nm against a baseline of 330 nm were particularly noted to ensure accuracy in the measurement. Further, the encapsulation efficiency of RNA within the nanoparticles was evaluated using the QUANT-IT^TM RIBOGREEN RNA assay (Shanghai ShineGene Molecular Biotechnology Co., LTD. ) . For this assay, samples were first diluted in TE buffer and then transferred to a polystyrene 96-well plate, where they were incubated at 40℃ for 10 minutes. Subsequently, a 1: 200 diluted RIBOGREEN reagent was added to each well, and the fluorescence intensity was measured using a Molecular Devices i3max microplate reader at excitation and emission wavelengths of about 488 nm and 525 nm, respectively. The final assessment of RNA encapsulation was determined by comparing the fluorescence intensities of intact and disrupted samples (the latter achieved by the addition of Triton X-100) , thereby quantifying the percentage of free RNA in the sample.
Example 3. Production of tLNP via bio-orthogonal click reaction

This example demonstrates the synthesis of targeted lipid nanoparticles (tLNPs) that incorporate ligands on the surface through a bio-orthogonal click chemistry reaction. The procedure simply involves the mixing of a VHH-linker conjugate into a solution containing anchor-modified LNPs and ultrafiltration, a process designed to ensure ligand attachment under mild conditions. The overall process for preparing tLNPs is illustrated in Figure 2.

To begin, a precise quantity of the VHH-linker conjugate is introduced into a pre-prepared solution of anchor-modified LNPs. This LNP and solution with ligand-linker conjugate is mixed in a defined anchor to ligand-linker conjugate ratio and incubated at room temperature (RT) for 3 hours to facilitate the conjugating reaction. Following the incubation period, ultrafiltration is performed three times using a molecular weight cutoff of 100 KD to remove any unbound ligand and to exchange the buffer with 20 mM Tris-HCl (pH=7.4) . To stabilize the formulation for long-term storage, 40%sucrose is added to adjust the final buffer to a concentration of 20 mM Tris-HCl (pH=7.4) with 8%sucrose. The final tLNP formulation is then sterile-filtered through a 0.22 μm filter to ensure sterility. The filtered LNPs are aliquoted as required and stored at -80℃ for future use.
Example 4. Production of tLNP via thiol-maleimide reaction

This example outlines the production of targeted lipid nanoparticles (tLNPs) using a traditional thiol-maleimide reaction, serving as a comparative approach to the conjugating technologies developed in this invention. In this method, the ligand does not incorporate a linker structure but is directly conjugated to LNPs through a reaction between a maleimide group introduced into the LNPs and a thiol group of cysteine in the ligand’s sequence.

The LNPs equipped with a maleimide group are prepared following the procedure described in Example 2, with the substitution of DSPE-PEG-mal in place of the clickable polyethylene glycol derivative as the anchor fragment. Notably, in this instance, the ligand is not conjugated with a linker. The modified anti-CD5 VHH (with a cysteine introduced at the C-terminus) as detailed in Example 1 is used directly. Initially, the anti-CD5 VHH is reduced using 2-MEA and subsequently purified using a Zeba^TM desalt column (7K MWCO) . The reduced anti-CD5 VHH is then mixed with the maleimide-modified LNPs in a predefined anchor to ligand ratio and incubated at room temperature for 3 hours to enable the conjugating reaction. Following the incubation, ultrafiltration is performed three times using a molecular weight cutoff of 100 KD to remove any unbound ligand and to exchange the buffer with 20 mM Tris-HCl (pH=7.4) . To enhance the stability of the formulation for long-term storage, 40%sucrose is added to adjust the final buffer to a concentration of 20 mM Tris-HCl with 8%sucrose. The final tLNP formulation undergoes sterile filtration through a 0.22 μm filter to ensure sterility. The sterile-filtered LNPs are then aliquoted as needed and stored at -80℃ for future use.
Example 5. In vitro study with cell lines

This example describes the evaluation of targeted lipid nanoparticles (tLNPs) for their delivery efficiency using cell lines that express specific receptor, contrasted against unrelated cell lines lacking these receptors to assess off-target delivery efficiency. In our studies focusing on CD5 and CD7 targeting LNPs, Jurkat cells were employed as target cells, while Raji, Daudi, HEK293, and HepG2 served as non-target cells. Jurkat and Raji cells were cultured in RPMI-1640 medium supplemented with L-glutamine (ThermoFisher) , 10%fetal bovine serum, and 1%penicillin-streptomycin. Daudi, HEK293, and HepG2 cells were maintained in DMEM supplemented with 10%fetal bovine serum and 1%penicillin-streptomycin.

For LNPs targeting CD90 and CD117, the HEL cell line, a human erythroleukemia cell line, or engineered CHO cells overexpressing CD90 or CD117 receptors were used as target cells. Unmodified CHO cells were used as non-target cells. HEL cells were cultured in RPMI-1640 medium supplemented with L-glutamine (ThermoFisher) , 10%fetal bovine serum, and 1%penicillin-streptomycin. CHO cells were maintained in DMEM supplemented with 10%fetal bovine serum and 1%penicillin-streptomycin.

All cells were plated at a density of 200,000 cells per 250 μL of respective culture medium in 24-well plates. A specific amount of tLNP, encapsulating GFP mRNA as the reporter gene, was added to each well. After 6 hours of incubation, the cells were harvested by centrifugation at 300 g for 7 minutes. The cell pellets were then resuspended in 100 μL of PBS in preparation for flow cytometry analysis. Analysis was conducted using a CytoFLEX flow cytometer (Beckman Coulter) . The flow cytometry data were analyzed using CytExpert software to determine the percentage of GFP-or tdTomato-positive cells within the different cell lines.
Example 6: In Vitro Study with Human Peripheral Blood Mononuclear Cells (PBMCs)

This example outlines the methodology for assessing the delivery efficiency of targeted lipid nanoparticles (tLNPs) using human PBMCs, which include mainly T, B, NK, and monocyte cells. Human PBMCs were plated at a density of 200,000 cells per 250 μL of respective culture medium in 24-well plates. A predetermined quantity of tLNP, encapsulating either GFP or tdTomato mRNA as the reporter gene, was added to each well. Following 6 hours of incubation to allow for cellular uptake of the LNPs, cells were harvested by centrifugation at 300 g for 7 minutes. The cell pellets were then resuspended in 100 μL of PBS to prepare them for subsequent staining procedures.

For dead cell exclusion, the PBMC suspension was incubated with 1 μL of FVS620 when using GFP mRNA, or FVS520 for tdTomato mRNA, for 10 minutes. Excess LIVE/DEAD dye was then removed by washing the cells with cell staining buffer. The washed cells were resuspended in 100 μL of cell staining buffer and stained with a cocktail of fluorescently-labeled antibodies for 15-25 minutes at room temperature. This antibody cocktail included BV650 Mouse Anti-Human CD3 (SK7) , Alexa Fluor 700 Mouse Anti-Human CD14 (M5E2) , BV421 Mouse Anti-Human CD56 (NCAM16.2) , and APC-H7 Mouse Anti-Human CD45 (2D1) , enabling the identification and analysis of specific cell populations: T cells (CD45+CD3+) , monocytes (CD45+CD14+) , and NK cells (CD45+CD56+) . After staining, cells were washed again with cell staining buffer to remove excess antibodies and resuspended in 100 μL of cold cell staining buffer, then kept on ice until flow cytometric analysis. Analysis was conducted using a CytoFLEX flow cytometer (Beckman Coulter) . The flow cytometry data were analyzed using CytExpert software to determine the percentage of GFP-or tdTomato-positive cells within the different cell population in PBMC. This approach allowed for a detailed assessment of the delivery efficiency within distinct immune cell populations.
Example 7: In Vivo Study of Targeted Delivery Efficiency Using Mouse Models

We investigated the in vivo targeted delivery efficiency of targeted lipid nanoparticles (tLNPs) using different mouse models, including PBMC humanized mice, CD5-humanized mice, and CD7-humanized mice. To evaluate the specificity and efficiency of tLNP targeting, we administered tLNPs formulated with tdTomato mRNA intravenously at a dosage of 1～2 mg/kg. Twenty-four hours post-injection, the mice were deeply anesthetized and subjected to a comprehensive analysis of peripheral blood, spleen, and liver to assess both targeted and non-targeted cell delivery efficiencies.

For the peripheral blood samples, 200 μL of whole blood was collected from the orbital sinus into heparinizedtubes. To isolate blood cells, 100 μL of blood was treated with live-dead dye for 10 minutes to identify dead cells, followed by red blood cell lysis and subsequent washing. The resulting RBC-free cell pellet was then stained and analyzed via flow cytometry to assess delivery efficiency among various cell types. Spleen cells were prepared by mechanically grinding the tissue and passing it through a 70-μm strainer. Liver cells were isolated through a two-step collagenase perfusion process. Initially, the mice were anesthetized with isoflurane and subjected to perfusion using a liver perfusion medium, followed by a liver digestion medium. The liver was then dissected, and hepatocytes were released, filtered through a 100 μm mesh, and washed with ice-cold PBS. After red blood cell lysis and washing, these cells were stained with specific antibody cocktails tailored to each mouse model. These antibodies targeted human markers in humanized mice and specific mouse markers in CD5 and CD7 humanized models.

For flow cytometry, cell suspensions from the spleen and liver were stained with live-dead dye to identify dead cells, washed, and then stained with an antibody cocktail suitable for each model. Specifically, for the cells of hPBMC mice, the antibody cocktail was composed of BV650 Mouse Anti-Human CD3 (SK7) , Alexa Fluor 700 Mouse Anti-Human CD14 (M5E2) , BV421 Mouse Anti-Human CD56 (NCAM16.2) , and APC-H7 Mouse anti-Human CD45 (2D1) . For the hCD5 mice, the antibody cocktail was composed of PE/Cyanine7 anti-mouse CD45.2 Antibody, APC Hamster Anti-Mouse CD3e (145-2C11) , APC-Cy7 Rat Anti-CD11b (M1/70) and BV421 Rat Anti-Mouse F4/80 (T45-2342) . For the hCD7 mice, the antibody cocktail was composed of Fixable Viability Dye eFluor^TM 780, APC anti-mouse CD45, BV650 Hamster Anti-Mouse CD3e (145-2C11) , BV 510^TM anti-mouse CD4, PE/Cy7 anti-mouse CD8a, BV711 Rat Anti-Mouse CD19 (1D3) , Alexa488 anti-mouse NK-1.1, BV480 Rat Anti-mouse CD11b, BV 605^TM anti-mouse CD11c, PerCP-Cy5.5 Rat Anti-Mouse Ly-6C (AL-21) , and BV421 Mouse Ant-Human CD7 (M-T701) . After staining, cells were washed, resuspended in cold cell staining buffer, and analyzed on a CytoFLEX flow cytometer. CytExpert software was used to determine the percentage of tdTomato-positive cells within each cell population.
Example 8. Formulation Optimization of tLNP using an Anti-CD5 VHH

In this example, we demonstrate the formulation optimization of targeted lipid nanoparticles (tLNPs) using an anti-CD5 VHH, clone 5H10-1. Following the procedures described in Example 1, ligand-linker conjugates were prepared using a modified anti-CD5 VHH. Anchor-modified lipid nanoparticles were then prepared as outlined in Example 2, and CD5-targeted tLNPs encapsulating mRNA encoding GFP protein were formulated according to the method described in Example 3, employing compound 062 as the ionizable lipid. To enhance the targeting delivery efficiency, the molar ratios of PEG lipid characterized in anchor fragment (anchor PEG, DSPE-PEG2000-N3) and PEG lipid characterized in LNP (lipid PEG, DSPE-PEG2000) were adjusted in the formulation, along with the amount of VHH-linker conjugate. The delivery efficiency to target and non-target cells was evaluated using cell lines (Jurkat, HepG2) and human PBMCs.

In the preparation of Anchor-Modified Lipid Nanoparticles, six different formulation ratios were tested as seen in Table 5; total PEG-lipid (the sum of lipid-PEG and clickable lipid-PEG) varied from 0.5%to 1.5%, with anchor PEG accounting for 25%to 100%of the total PEG-lipid. The properties of the formulated CD5-targeting LNPs, including particle size, PDI, and mRNA encapsulation rates, are detailed in Table 6. The delivery efficiency of the reporter gene to target and non-target cells was assessed in cell lines and human PBMCs following the procedures described in Examples 5-6. In cell line studies, Jurkat cells served as the target cells, and HepG2 cells as non-target cells, with Mean Fluorescence Intensity (MFI) used as the reporting value. In PBMC studies, T cells were the target cells and monocytes (mo cells) the non-target cells, with the positivity rate serving as the reporting value.
Table 5. Compositions of different formulation

Table 6. Information on the formulations and characterization of different LNPs

To assess the impact of different formulations on targeted delivery efficiency, we categorized the tLNPs into six groups (A-F) based on their compositions and analyzed the delivery outcomes in target and non-target cells, as depicted in Figure 3. It was observed that the tLNPs exhibited minimal expression in non-target HepG2 cells, whereas significantly higher gene expression was noted in the target Jurkat cells. Notably, formulations B, C, and E demonstrated enhanced expression in target cells, whereas formulations A, D, and F showed markedly lower levels. This pattern was consistently replicated in studies with PBMCs, where tLNPs with formulations A, D, and F showed mild expression in primary T cells, while those with B, C, and E displayed substantial higher expression, particularly B and C, where some formulations achieved nearly 60%positivity in T cells with virtually no expression in monocytes. These results indicate that this technological approach can produce tLNPs with excellent targeted delivery capabilities. From the formulation data in Table 5, it is evident that the total PEG-lipid content in formulation A is 0.5%, whereas it is higher in B, C, and E. Similarly, the Anchor/total PEG-lipid ratio is 25%in D and F, while it exceeds this level in B, C, and E. Therefore, when using lipid-PEG and anchor PEG with PEG molecular weights of 2000, the preferred total PEG-lipid value should be above 0.5%, and the Anchor/total PEG-lipid ratio should also exceed 25%.

To explore the relationship between the amount of VHH-linker conjugate and targeted delivery efficiency, we classified the tLNPs based on varying VHH-linker conjugate amounts and charted the delivery outcomes to target and non-target cells as shown in Figure 4. The results indicated that when evaluated using cell lines, tLNPs within the range of 0.062%to 0.5%consistently demonstrated good targeted delivery efficiency (Figure 4A) . In experiments using primary T cells, tLNPs within the range of 0.062%to 1%also showed some level of targeted delivery efficiency, with the optimal performance occurring within the 0.062%to 0.5%range (Figure 4B) .
Example 9. Formulation Optimization of tLNP prepared with linker-free VHH via thiol-maleimide
reaction

To compare the tLNPs prepared by click reaction disclosed in this application with those prepared using the classical thiol-maleimide reaction, we employed the procedure described in Example 4 to produce tLNPs targeting CD5. Compound 062 was used as the ionizable lipid, and the tLNPs encapsulated mRNA encoding GFP protein. To enhance the targeted delivery efficiency, we adjusted the molar ratios of the anchor fragment (DSPE-PEG-mal) and lipid-PEG (DSPE-PEG2000) in the formulation (refer to Table 7) . VHH that not conjugated with a linker were used at different varied amount. The properties of the formulated CD5-targeting LNPs, including particle size, PDI, and mRNA encapsulation rates, are detailed in Table 8. The delivery efficiency to target and non-target cells was assessed using cell lines (Jurkat and HepG2) .
Table 7. Compositions of different formulation

Table 8. Information on the formulations and characterization of different LNPs

To evaluate the impact of different formulations on the targeted delivery efficiency, we charted the results based on various formulations and VHH amount. As shown in Figure 5, the CD5-targeted tLNPs demonstrated higher delivery efficiency in Jurkat cells compared to HepG2 cells. However, significant delivery was still observed in HepG2 cells, indicating notable off-target delivery associated with tLNPs produced by this method. Notably, LNP-150 exhibited a higher delivery efficiency to target cells, with a substantial difference between target and non-target cells. Thus, Formulation C was selected for further evaluation of the delivery efficiency and robustness of tLNPs prepared via the thiol-maleimide reaction with linker-free VHH.
Example 10. Comparison of robustness of different tLNP preparation strategy

To evaluate the robustness of tLNP preparation methods, we prepared several independent batches of CD5-targeted tLNPs using different strategies and measured their delivery efficiency to target and non-target cells. Using the method described in Example 4, we prepared tLNPs targeting CD5 through a classical thiol-maleimide reaction, employing a linker-free VHH structure. Five batches of tLNPs were produced using the optimized formulation condition obtained in Example 9. We also prepared six batches of tLNPs utilizing the technique disclosed in Example 1-3 using formulation C with two different amounts of VHH-linker conjugates. The delivery efficiencies to target cells (Jurkat) and non-target cells (HepG2) were assessed, with characteristics of the produced tLNPs detailed in the table 9. The analysis showed no significant differences between the two techniques in terms of particle size, PDI, and encapsulation efficiency.
Table 9. Information on the formulations and characterization of different LNPs

As shown in Figure 6, the batch-to-batch consistency in delivery efficiency was notably better for tLNPs prepared using VHH-linker conjugates via click reaction disclosed in this application compared to those prepared using the classical thiol-maleimide reaction with linker-free VHH. For instance, when using VHH-linker conjugates at 0.25%, all batches demonstrated a Jurkat positivity rate higher than 90%, while the positivity rate in HepG2 remained below 7%. Conversely, tLNPs prepared using the thiol-maleimide reaction with linker-free VHH exhibited significant batch-to-batch variability in both target and non-target cells.

Notably, when employing the thiol-maleimide reaction, the initial preparation scale was 0.2 mL. To further test the robustness of the preparation method of the classical thiol-maleimide reaction with linker-free VHH, a few more batches were prepared (Table 10) . Increasing the preparation volume to 0.4 mL with the same formulation resulted in a noticeable increase in particle size, as evidenced by LNP-163, which had a particle diameter of 441 nm. By reducing the VHH amount, we managed to decrease the particle size to under 200 nm, as shown in LNP-255 to LNP-257. However, further scaling up the production volume to 1.2 mL (LNP-174) under the same formulation led to an increase in particle size to 269 nm, indicating that the scalability of the thiol-maleimide reaction preparation method with linker-free VHH is relatively poor compared to the click reaction with the VHH-linker conjugates. This example highlights the superior robustness and consistency of the tLNP preparation technique with the VHH-linker conjugates, especially in terms of scalability and batch-to-batch delivery efficiency, making it more suitable for clinical and commercial applications where consistency is crucial.
Table 10. Characterization of different LNPs

Example 11. In vivo study in hPBMC-engrafted mice using CD5-targeting LNP

In this study, we evaluated the in vivo targeting delivery efficiency of targeted lipid nanoparticles (tLNPs) prepared using the VHH-linker conjugates disclosed in this application in hPBMC-engrafted NSG mice. Following the procedures described in Example 1, ligand-linker conjugates were prepared using a modified anti-CD5 VHH (clone 5H10-1) . Anchor-modified lipid nanoparticles were then prepared as outlined in Example 2, and CD5-targeted tLNPs encapsulating mRNA encoding tdTomato protein were formulated according to Example 3, using compound 062 as the ionizable lipid. Based on previous formulation experience, the total PEG-lipid was set within the range of 0.5%to 2%, with 50%of that being PEG lipid characterized in anchor fragment (anchor PEG) . For control purposes, non-targeted LNPs without conjugated ligand (LNP-166) were also prepared, using 1.5%DSPE-PEG without any anchor PEG. Information on the formulations and physicochemical characterization parameters of these LNPs is detailed in Table 11.
Table 11. Information on the formulations and characterization of different LNPs

In vivo experiments were performed as described in Example 7. The PBMC-engrafted mice were intravenously administered LNPs at 2 mg/kg, and blood and tissue samples were collected 24 hours later. The positivity rates of the reporter gene in T cells (CD3+) and non-target cells (CD3-) were analyzed. The results were shown in figure 7. Compared to the non-targeted LNP (LNP-166) , all LNPs conjugated with anti-CD5 VHH exhibited higher positivity rates in target cells. Notably, for the best formulation LNP-176, the percentage of tdTomato+ T cells in the peripheral blood, spleen, and liver were 27.5%, 30.2%, and 58.7%respectively. The percentage of tdTomato+ CD3-cells in blood, spleen, and liver was 1.4%, 0.8%, and 5.8%, respectively. Interestingly, LNP-179, despite having better delivery efficiency to target cells, also showed relatively higher expression in non-target cells, especially in the spleen. This suggests that total PEG-lipid levels might need to exceed 0.5%to achieve better targeting specificity.

Observations across all tested CD5-targeting LNPs indicated that the positivity rates of the reporter gene in T cells in the peripheral blood and spleen were similar, whereas the positivity rate in liver T cells was significantly higher. This is likely due to a graft-versus-host disease (GVHD) reaction in hPBMC-engrafted NSG mice, which might cause excessive activation of T cells, particularly in the liver.
Example 12. In vivo study in CD5-humanized mice using CD5-targeting LNP

In this study, we evaluated the in vivo delivery efficiency of CD5-targeting LNPs using CD5-humanized mice. In these hCD5 mice, exons 2 to 7 of the mouse Cd5 gene, which encode the extracellular domain, were replaced with the corresponding human CD5 exons, ensuring the expression of human CD5 in homozygous hCD5 mice. The humanization of CD5 did not alter the overall frequency or distribution of immune cell types in the spleen, blood, and lymph nodes, providing a model with normally functioning T cells and avoiding potential biases due to the overactivation often seen in PBMC models. This model also facilitates the evaluation of potential on-target/off-target delivery.

Following the methods described in Example 1, ligand-linker conjugates were prepared using a modified anti-CD5 VHH (clone 5H10-1) . Anchor-modified lipid nanoparticles were then prepared as outlined in Example 2, and CD5-targeted tLNPs encapsulating mRNA expressing tdTomato protein were formulated according to Example 3, using compound 062 as the ionizable lipid. Based on previous formulation experience, total PEG-lipid was set between 1%and 1.5%, with anchor/total PEG-lipid maintained at 50%. Additionally, a standard liver-targeting formulation (LNP-263) was used as a control, which included compound 062 as the ionizable component in a formulation ratio of ionizable lipid: DSPC: Chol: DMG-PEG=50: 10: 38.5: 1.5.
Table 12. Information on the formulations and characterization of different LNPs

In vivo studies, as described in Example 7, were conducted after intravenous injection of LNPs into hCD5 mice at 2 mg/kg. Blood and tissue samples were collected 24 hours post-injection for analysis. The positivity rates of the reporter gene in T cells (CD3+) and non-target cells (CD3-) were assessed, as illustrated in Figure 8. In peripheral blood and spleen, the liver-targeting LNP (LNP-263) showed higher reporter gene expression in CD3-cells compared to CD3+ cells. The two tested CD5 targeting LNPs reversed this trend, exhibiting higher delivery efficiency in T cells. In the liver, LNP-263 showed only 0.9%positivity in T cells, whereas 36.1%and 17.6%of non-target (CD3-) and parenchymal (CD45-) cells expressed the reporter gene, respectively. LNP-264 and LNP-265 increased the positivity rate of the reporter gene in liver T cells to 4-7%. Notably, 7-8%of non-target cells (CD3-) still expressed the reporter gene. One possible explanation is the presence of non-T cell populations expressing hCD5, such as Kupffer cells.
Example 13. Storage stability of tLNP

This example focuses on assessing the long-term storage stability of CD5-targeting LNPs prepared using VHH-linker conjugates via click reaction disclosed in this application. Different formulations of tLNP were stored at -80℃ for eight weeks. Subsequently, these tLNPs were utilized to transfect HepG2 and Jurkat cells as described in Example 5. As shown in Figure 9, the results demonstrated that the positivity rate in target cells (Jurkat) showed no significant difference between pre-storage and after eight weeks of cryopreservation. Moreover, except for LNP-202, the delivery to non-target cells (HepG2) also maintained very low levels after cryopreservation. These outcomes indicate that the tLNPs manufactured using this disclosed method exhibit excellent storage stability, maintaining their functional integrity and specificity even after extended periods of storage under deep-freeze conditions.
Example 14. Comparison of LNP surface bioconjugation of ligand with different Molecular weights

In this study, we evaluated the targeting efficiency of LNPs conjugated with ligands of different molecular weights using the tLNP preparation method disclosed in this application. We compared two types of antibodies targeting CD117 (c-kit) , a receptor highly expressed on hematopoietic stem cells (HSCs) , potentially for targeted delivery to HSCs. The antibodies used were a VHH (～16 kDa) and an scFv (～30 kDa) . Following the procedure described in Example 1, ligand-linker conjugates were prepared using a modified anti-CD117 VHH and a modified anti-CD117 scFv, the sequences of which are detailed in Table 13. Both the VHH and scFv sequences were engineered to include a C-terminal extension that added a histidine-tag, a non- (GGGGS) n hinge, and a cysteine residue. The “GGGGS” sequence is shown as SEQ ID NO: 63.
Table 13. Original and modified antibody sequence

Anchor-modified lipid nanoparticles encapsulating mRNA encoding GFP protein were then prepared as outlined in Example 2, and CD117-targeted tLNPs were formulated according to Example 3. Based on previous optimal formulation experiences, two ratios were used: Formulation C and Formulation E, containing 1%and 1.5%total PEG-lipid respectively, with Anchor/total PEG-lipid both at 50%. Compound 062 served as the ionizable lipid. During the ligand conjugation step, different amounts of ligands ranging from 0.125%to 0.5%were employed. Information on the formulations and physicochemical characterization parameters of these LNPs is detailed in Table 14.
Table 14. Information on the formulations and characterization of different LNPs

As described in Examples 5 and 6, these tLNPs were used to transfect HEL cells (human erythroleukemia cells) and hCD117 overexpressing CHO cells (hCD117 CHO) . The transfection results are displayed in Figure 10. Both VHH and scFv conjugations enhanced gene expression in target cells expressing CD117. It was observed that across different amounts of ligand-linker conjugates, VHH consistently outperformed scFv. For example, in HEL cells, the optimal LNP conjugated with VHH-linker showed mean fluorescence intensity (MFI) ten times higher than that of the optimal LNP conjugated with scFv-linker, after background subtraction. Furthermore, it is noteworthy that with scFv as the ligand, the delivery efficiency to target cells gradually increased with increasing ligand amount within the range of 0.125%to 0.5%. However, with VHH as the ligand, a dosage of 0.125%already exhibited the best delivery performance in HEL cells. These results suggest that using the ligand-linker conjugates via click reaction disclosed in this application, lower molecular weight antibody forms, particularly VHH, potentially achieve better targeting delivery outcomes.
Example 15. Comparison of Hinge Sequences for Optimized VHH Conjugation and Targeted Delivery
using anti-CD7 VHH

To assess how hinge sequences affect site-specific conjugation and targeted delivery, we designed a panel of anti-CD7 VHH constructs incorporating different hinge sequences. All constructs were derived from the same anti-CD7 VHH clone, with a C-terminal cysteine residue introduced for site-specific conjugation. This cysteine was linked to the VHH domain with no hinge, or through one of six different hinges (Hinge-1 through Hinge-6) , each varying in amino acid composition and predicted structural behavior. Table 15 lists the hinge sequences along with their composition profiles, including the percentage of rigidity-enhancing residues (P, Y, F, W, V, I, L) , glycine (G) , and α-helix-stabilizing residues (A, L, E, M) . Notably, Hinge-1 represents the widely used flexible (G₄S) ₃ motif with 80%of glycine. Hinge-3 contain >20%glycine and no rigidity-enhancing residues, and were thus defined as flexible hinges. Hinge-2 includes >20%rigidity-enhancing residues with limited glycine content (<10%) , while Hinge-4 comprises >80%α-helix-stabilizing residues, and both were categorized as restricted hinges. Hinge-5 and Hinge-6 feature ≥50%rigidity-enhancing residues and no glycine, and were thus defined as rigid hinges.
Table 15. The sequences and property of different hinges.

All VHH constructs containing Hinge-1 through Hinge-6 were transiently expressed and purified (sequences shown in Table 16) . Each construct was conjugated to a DBCO-PEG₄-maleimide linker via thiol-maleimide chemistry reaction, as described in Example 1. SDS-PAGE analysis of the resulting VHH-linker conjugates is shown in Figure 11A, and the percentage of dimer formation is presented in Figure 11B. Constructs without a hinge or with the G4S hinge (Hinge-1) exhibited high levels of dimerization (～40%) . In contrast, constructs containing non-G4S hinges showed significantly lower dimer levels. Notably, VHHs with restricted hinges (Hinge-2 and Hinge-4) exhibited ～20%dimer content, lower than flexible hinge constructs. Rigid hinge constructs (Hinge-5 and Hinge-6) showed the lowest dimer levels, approximately 10%. These findings suggest that hinges with greater structural constraint spatially separate the terminal cysteine from the VHH core, thereby reducing its interference with VHH folding and disulfide bond formation. This spatial decoupling may improve conjugation efficiency and reduce mispaired disulfides. Furthermore, ELISA assays evaluating antigen binding to CD7 (Figure 11C) revealed that the VHH-linker conjugate without any hinge had the strongest binding affinity (lowest EC₅₀) , while the addition of hinge sequences modestly reduced binding, possibly due to steric effects introduced by the hinge.
Table 16. Anti-CD7 VHH sequence with different hinges.

Anchor-modified lipid nanoparticles encapsulating GFP mRNA were formulated following Example 2, and CD7-targeting tLNPs were produced per Example 3 using Formulation C (1%total PEG-lipid, 50%anchor/PEG-lipid ratio) , with Compound 062 as the ionizable lipid. During conjugation, VHH-linker conjugates were applied at three different molar percentages (0.0625%, 0.125%, 0.25%) . The composition and physicochemical properties of the resulting LNPs are provided in Table 17.
Table 17. Information on the formulations and characterization of different LNPs

As outlined in Examples 5 and 6, these tLNPs were applied to transfect Jurkat (target) and Raji (non-target) cells. As shown in Figures 12A-B, all VHH-conjugated LNPs induced strong GFP expression in Jurkat cells, while showing minimal activity in Raji cells. Importantly, hinge rigidity had a clear influence on conjugation behavior across varying ligand densities. Constructs lacking hinges or incorporating flexible hinges (Hinge-1 and Hinge-3) showed a positive correlation between conjugation density and delivery efficiency in Jurkat cells. However, VHHs containing restricted or rigid hinges (Hinge-2, Hinge-4, Hinge-5, Hinge-6) displayed higher transfection at lower ligand densities, suggesting that structurally constrained hinges (comprise restricted hinge and rigid hinge) allow more efficient delivery even at reduced ligand loading.

PBMCs were also used to evaluate tLNP targeting across T cells, NK cells (as target populations) , and monocytes (non-targets) . As shown in Figures 12C-E, across all three conjugation densities, constructs with structurally constrained hinges consistently achieved higher GFP expression in target cells compared to those with no hinge or flexible hinges. For instance, Hinge-2 and Hinge-5 yielded T cell positivity rates approaching 60%at intermediate ligand density, versus 7–30%for constructs without hinge or with flexible hinges. Moreover, while flexible hinge groups showed improved delivery at higher ligand densities (0.25%) , rigid and restricted hinges achieved optimal delivery at lower conjugation levels (0.0625–0.125%) . Importantly, the enhanced performance of structurally constrained hinges cannot be attributed to increased antigen-binding affinity, as demonstrated in the ELISA results. Instead, the improved delivery may arise from enhanced conjugation efficiency and favorable spatial orientation of the ligand on the LNP surface-mediated by the structural properties of the hinge sequence. These findings highlight the critical role of hinge design in optimizing targeted LNP delivery.
Example 16. Evaluation of Hinge Sequence on Targeted Delivery Efficiency of CD117-targeting tLNPs

To further investigate the influence of hinge sequences on targeted delivery efficiency, we employed a single-domain antibody (VHH) directed against the hematopoietic stem cell marker CD117 (clone H. 2346B13) . All constructs shared the same VHH framework, with a C-terminal cysteine residue introduced to facilitate site-specific conjugation. The cysteine was connected to the VHH either directly (no hinge) or via one of five distinct hinge sequences (Hinge-1, Hinge-2, Hinge-4, Hinge-5, and Hinge-7) , each selected to represent different structural profiles. Table 18 provides the full amino acid sequences of hinges and corresponding compositional features, including the percentage of rigidity-enhancing residues (P, Y, F, W, V, I, L) , glycine content, and the fraction of α-helix-stabilizing residues (A, L, E, M) . Hinge-1 consisted of the widely utilized flexible motif (G₄S) ₃, known for high glycine content and structural flexibility. Hinge-2 and Hinge-4 were categorized as restricted hinges, with Hinge-2 possessing 22%of rigidity-enhancing residues and minimal glycine (<10%) , and Hinge-4 containing over 80%α-helix-stabilizing residues. Hinge-5 and Hinge-7, each comprising approximately 50%rigidity-enhancing residues and no glycine, were classified as rigid hinges.
Table 18. The sequences and property of different hinges.

All VHH constructs were transiently expressed and purified (see sequences in Table 19) . Following reduction of the C-terminal cysteine, site-specific conjugation to a DBCO-PEG₄-maleimide linker was performed via thiol-maleimide chemistry as outlined in Example 1. Anchor-modified lipid nanoparticles encapsulating mRNA encoding GFP were prepared in accordance with Example 2, and CD117-targeted tLNPs were formulated per the procedure in Example 3 using Formulation C (1%total PEG-lipid with a 50%anchor PEG ratio) . Compound 062 was used as the ionizable lipid. The final LNP compositions and their physicochemical characteristics are summarized in Table 20.
Table 19. Anti-CD117 VHH sequence with different hinges.

Table 20. Information on the formulations and characterization of different LNPs

We evaluated the delivery efficiency of these CD117-targeted LNPs using human CD117-expressing CHO cells as the positive target population and wild-type CHO cells as the negative control. As described in Examples 5 and 6, the tLNPs were applied for in vitro transfection, and results are presented in Figure 13. All VHH-conjugated tLNPs showed elevated MFI (Figure13 A and B) and positivity (Figure13 C and D) in CD117-positive CHO cells compared to wild-type CHO cells, confirming target-specific delivery. Notably, the G4S-hinged construct outperformed the no-hinge construct, suggesting that even flexible hinge insertion can enhance accessibility of the C-terminal cysteine for conjugation. The nature of the hinge sequence substantially influenced the delivery outcome. Constructs bearing restricted or rigid hinges (Hinge-2, Hinge-4, Hinge-5, Hinge-7) consistently outperformed those with no hinge or a G4S hinge in terms of delivery efficiency in target cells. Among them, rigid hinge-containing constructs (Hinge-5 and Hinge-7) achieved the highest delivery levels, slightly surpassing those with restricted hinges. These findings suggest that spatial restriction provided by rigid linkers may help better orient the conjugated VHHs on the LNP surface for receptor engagement.

Building on these observations, we further engineered an additional set of rigid hinge sequences with enhanced biophysical characteristics. As summarized in Table 21, these newly designed hinges (Hinge-8 to Hinge-12) are characterized by complete absence of glycine residues and a proportion of rigidity-enhancing residues ranging from 33.3%to 50%. Moreover, each of these hinges incorporates hydrophilic residues such as lysine (K) , arginine (R) , aspartic acid (D) , and glutamic acid (E) , enhancing solubility without compromising structural integrity. The corresponding hinge-modified constructs were expressed based on the same anti-CD117 VHH clone (H. 2346B13) , with full amino acid sequences provided in Table 22.
Table 21. The sequences and property of different hinges.

Table 22. Anti-CD117 VHH sequence with different hinges.

The targeting delivery efficiency of these CD117-targeted tLNPs was evaluated using human CD117-expressing CHO cells as the positive target population and wild-type CHO cells as the negative control. Following the procedures described in Examples 5 and 6, the tLNPs were applied for in vitro transfection studies, and the results are presented in Figure 14. As expected, the unconjugated LNPs exhibited comparable levels of transfection in both target and non-target cells, indicating a lack of inherent targeting capability. In contrast, all VHH-conjugated tLNPs demonstrated a substantial increase in MFI (Figure 14A and B) and percentage of positive cells (Figure 14C and D) specifically in CD117-positive CHO cells compared to wild-type CHO cells, confirming efficient and selective targeted delivery.

Notably, tLNPs conjugated with VHHs containing the newly engineered hinges (Hinge-8 to Hinge-12) achieved target cell transfection efficiencies comparable to those observed with the Hinge-5 construct, while maintaining very low off-target delivery in wild-type cells. These results suggest that hinge sequences devoid of glycine and enriched with at least 33.3%rigidity-enhancing residues can effectively support targeted delivery performance of tLNPs. Furthermore, the incorporation of hydrophilic residues carrying either positive or negative charges did not impair the targeting efficiency, indicating flexibility in hinge design for future optimization efforts.
Example 17. Comparison of targeted delivery efficiency of LNP conjugated with VHH-linker conjugates
or linker-free VHH

This example evaluates the targeted delivery efficiency of LNPs using two different conjugation strategies: with VHH-linker conjugates and with linker-free VHH. The technology protected in this application utilizes a chemical linker structure connected to the ligand, which leverages a cysteine residue introduced at the end of the ligand sequence to incorporate a click chemistry group. In this study, we employed an optimized formulation to conjugate both linker-equipped anti-CD5 VHH and linker-free anti-CD5 VHH to LNPs, comparing the delivery efficiency of the resulting tLNPs. The linker structure used was DBCO-PEG₄-Maleimide, and tLNPs were prepared following the procedure described in Example 3. tLNPs without a linker structure were prepared using the method outlined in Example 4.

We investigated the effects of varying VHH conjugation amounts under the same anchor-modified LNP formulation ratios or different formulation ratios but with identical VHH conjugation amounts. The best-performing formulation ratio from the optimization studies (Formulation C) was used, which includes 1%total PEG-lipid, anchor/total PEG-lipid at 50%, with VHH dosages ranging from 0.031%to 0.5%. Information on the formulations and physicochemical characterization parameters of these LNPs is detailed in Table 23. Delivery efficiency to target and non-target cells was assessed using the methods described in Example 6.
Table 23. Information on the formulations and characterization of different LNPs

Results illustrated in Figure 15 show that both tLNPs, whether conjugated with a linker or not, significantly enhanced delivery efficiency to target cells while maintaining very low delivery rates to non-target cells. Across the tested range of VHH conjugation amounts, the group with the VHH-linker conjugates consistently demonstrated higher delivery efficiency to target cells compared to the linker-free VHH group. Furthermore, we compared the differences between VHH-linker and linker-free VHH under various formulations while keeping the VHH amounts constant. Total PEG-lipid was tested at 1%and 1.5%, with Anchor/total PEG-lipid ratios of 25%and 50%. Information on the formulations and physicochemical characterization parameters of these LNPs is detailed in Table 24. Results, as shown in Figure 16, indicate that under all three different formulations, tLNPs with the VHH-linker structure consistently exhibited higher delivery efficiencies to Jurkat target cells and maintained relatively lower deliveries to non-target HepG2 cells. These findings suggest that employing VHH with a linker structure significantly enhances the targeted delivery efficiency of tLNPs, providing a substantial improvement over linker-free VHH approaches.
Table 24. Information on the formulations and characterization of different LNPs

Example 18. Comparison of LNP Surface Bioconjugation Using Ligands Modified with Linkers of
Different Structures

This example investigates the influence of linker structural variation on the targeted delivery efficiency of tLNPs. Specifically, we evaluated multiple linker designs with distinct chemical backbones to conjugate anti-CD117 VHH ligands to the surface of lipid nanoparticles. Each linker possessed a maleimide group at one end, which facilitated site-specific conjugation to the thiol group on the C-terminal cysteine residue of the VHH. The other end of the linker featured a bio-orthogonal click chemistry group, such as DBCO, enabling covalent attachment to azide-functionalized anchor lipids embedded in the LNP surface. The linkers varied in their internal structural elements, including aliphatic chains of different lengths, PEG-based chains, and amide linkages. These internal motifs were incorporated to modulate flexibility, hydrophilicity, and steric spacing between the ligand and LNP surface.
Table 25. Chemical structure of different linker.

VHH-linker conjugates were synthesized using the procedure described in Example 1, employing an anti-CD117 VHH clone as the targeting moiety. Subsequent conjugation to anchor-modified LNPs was carried out using the methodology described in Examples 2 and 3 to produce CD117-targeted tLNPs. The physicochemical characteristics of the resulting formulations-including particle size, polydispersity index (PDI) , and mRNA encapsulation efficiency-are summarized in Table 26.
Table 26. Information on the formulations and characterization of different LNPs

The targeted delivery efficiency of these formulations was assessed in vitro using a panel of cell lines. HEL cells, human CD117-expressing CHO cells, and cynomolgus CD117-expressing CHO cells were employed as target cells, while wild-type CHO cells served as the non-target control. As shown in Figure 17, all linker-modified tLNPs exhibited high gene expression levels in target cells and minimal delivery in non-target cells, confirming the specificity of the system. Importantly, despite the differences in the chemical structures connecting the maleimide and DBCO groups-including variations in carbon chain length, the presence or absence of amide bonds, and the inclusion of PEG spacers-no substantial differences were observed in the delivery efficiency across the different linker groups. These results indicate that, within the context of the tested designs, the chemical structure of the linker backbone does not significantly affect the overall delivery performance of the resulting tLNPs, provided the terminal functional groups remain reactive and the linker length is sufficient to ensure effective ligand presentation on the LNP surface.
Example 19. Comparison of anchor fragment with different hydrophobic structures

In this implementation example, we evaluate the impact of different hydrophobic structures in the clickable anchor fragments on the delivery efficiency using the tLNP preparation technology protected in this application. As described in Example 2, Anchor-Modified Lipid Nanoparticles were prepared. Differing from the method described in Example 2, we also used a PEGylated polymer PLA (2000 Da) , equipped with a click chemistry group at the end of the PEG segment, named PLA2K-PEG2K-N3. PLA, being a biocompatible hydrophobic polymer, replaced the hydrophobic DSPE lipid as the hydrophobic part of the anchor fragment. We replaced DSPE-PEG2K-N3 with PLA2K-PEG2K-N3 and prepared anchor-modified LNP with four different ratios. Additionally, different amounts of VHH were used to prepare CD5-targeted tLNP. As a control, tLNPs using DSPE-PEG2K-N3 as the anchor fragment were also prepared. Information about the prepared CD5-targeting LNP, including particle size, PDI, and mRNA encapsulation rates, is detailed in Table 27.
Table 27. Information on the formulations and characterization of different LNPs

The targeted delivery efficiency of the reporter gene was evaluated using cell lines. In these cell line studies, Jurkat cells served as target cells, and HepG2 cells as non-target cells. As shown in Figure 18, tLNPs using PLA2K-PEG2K-N3 exhibited higher transfection in Jurkat as compared with HepG2 cells. However, across four different formulations and two different VHH conjugation amounts, tLNPs using PLA2K-PEG2K-N3 did not perform as well as those using DSPE-PEG2K-N3 in terms of delivery efficiency to target cells. This difference was more pronounced at lower VHH conjugation amounts. Additionally, tLNPs using DSPE-PEG2K-N3 exhibited lower off-target delivery in non-target cells. These results suggest that clickable polymer-PEG with PLA could also be used as anchor fragments. Clickable lipid-PEG with DSPE as an anchor fragment can enhance targeting delivery efficiency more effectively than clickable polymer-PEG with PLA.
Example 20. Comparison of lipid-PEG and anchor-PEG with different PEG lengths

In this implementation example, we utilized the tLNP preparation technology protected in this application to assess the impact of different lengths of PEG in lipid-PEG and anchor fragments (anchor-PEG) on delivery efficiency. Following the method described in Example 2, Anchor-Modified Lipid Nanoparticles were prepared with three different formulation ratios, as shown in Table 28. Differing from Example 2, we utilized DSPE-PEG and DSPE-PEG-N3 molecules of varying lengths (2K, 4K, 5K for DSPE-PEG and 2K, 3.4K, 5K for DSPE-PEG-N3) . CD5-targeted tLNPs were prepared with two different conjugation amounts of VHH-linker conjugates. Information about the prepared tLNPs, including particle size, PDI, and mRNA encapsulation efficiency, is presented in Table 29.
Table 28. Compositions of different formulation

Table 29. Information on the formulations and characterization of different LNPs

Jurkat cells served as target cells and HepG2 cells as non-target cells to evaluate the targeted delivery efficiency of the reporter gene. As shown in Figure 19, for lipid-PEG and anchor-PEG molecules ranging from 2K to 5K in length, effective delivery in target cells and very low delivery in non-target cells were achieved. In the groups with DSPE-PEG5K/DSPE-PEG5K-N3 and DSPE-PEG4K/DSPE-PEG3.4K-N3, good targeting delivery results were observed with Formulations I and C. Notably, Formulation I used 0.5%total PEG-lipid. This contrasts with findings from comparative studies in Example 8, where using 0.5%total PEG-lipid (DSPE-PEG2K/DSPE-PEG2K-N3) resulted in weaker targeting efficiency. This suggests that longer PEG chains might allow reducing the total PEG-lipid level to as low as 0.5%. Furthermore, no delivery effects were observed in Jurkat cells when using Formulation G in the groups DSPE-PEG5K/DSPE-PEG5K-N3 and DSPE-PEG4K/DSPE-PEG3.4K-N3. In Formulation G, the total PEG-lipid ratio was 2%. This indicates that when using DSPE-PEG as lipid-PEG, the total PEG-lipid ratio needs to be below 2%to avoid inhibiting targeting efficiency.
Example 21. Comparison ofdifferent lipid-PEG molecules

In this implementation example, we utilized the tLNP preparation technology protected in this application, employing different lipid-PEG molecules to prepare tLNPs. Following the procedures described in Example 2, Anchor-Modified Lipid Nanoparticles were prepared using three different formulation ratios, as detailed in Table 30. Differing from Example 2, three types of lipid-PEG molecules with distinct lipid structures were used: DSPE-PEG2K, DMG-PEG2K, and TPA-PEG2K. Two different conjugation amounts of VHH were used to prepare CD5-targeted tLNPs. Information on the prepared tLNPs, including particle size, PDI, and mRNA encapsulation efficiency, is presented in Table 31.
Table 30. Compositions of different formulation

Table 31. Information on the formulations and characterization of different LNPs

Jurkat cells served as target cells and HepG2 cells as non-target cells to evaluate the targeted delivery efficiency of the reporter gene. As depicted in Figure 20, using DSPE-PEG2K resulted in excellent delivery efficiency to Jurkat cells and very low off-target delivery to HepG2 cells. When using DMG-PEG2K, although the MFI was higher in Jurkat cells than in HepG2 across all three formulation ratios and two VHH densities, indicating some level of targeted delivery, the difference between target and non-target cells was less pronounced than in the DSPE-PEG2K group. Notably, in Formulation C (Total PEG-lipid = 1%) , an increase in HepG2 expression was observed at both VHH conjugation amounts. Increasing the Total PEG-lipid amount significantly reduced off-target delivery. Using TPA-PEG2K, excellent delivery efficiency to Jurkat cells and baseline levels of off-target delivery to HepG2 cells were observed in Formulation G (Total PEG-lipid = 2%) . These results suggest that the optimal Total PEG-lipid percentage for achieving the best targeting delivery results varies with different lipid-PEG molecules, with potential ranges from 0.5%to 2%being reasonable depending on the specific lipid-PEG used. This illustrates the importance of selecting the appropriate lipid-PEG type and concentration to optimize tLNP formulation for specific targeting requirements.
Example 22. Comparison of tLNPs Prepared by Surface Conjugation versus Post-Insertion Technique

In this example, we compared the delivery efficiency of targeted lipid nanoparticles (tLNPs) prepared using the surface conjugation method using a VHH-linker conjugate by the click reaction protected in the application with those prepared using a conventional post-insertion technique. The post-insertion technique involves incubating pre-formed lipid nanoparticles with functionalized lipid conjugates (e.g., DSPE-PEG-VHH) under defined conditions to facilitate spontaneous membrane insertion, thereby incorporating targeting ligands onto the LNP surface.
Protocol for Post-Insertion Preparation:
1. LNP Preparation:

LNPs were formulated using a lipid mixture comprising DSPC, cholesterol, and DSPE-PEG2000 in molar ratios of 50: 10: 38.5: 0.9 or 50: 10: 38.5: 1.35. Lipids were dissolved in absolute ethanol to yield final DSPE-PEG concentrations of 0.221 mg/mL or 0.331 mg/mL, respectively. A 0.2 mL aliquot of this lipid solution was rapidly mixed with 0.6 mL of mRNA solution (0.1 mg/mL) using a microfluidic mixing device. The resulting nanoparticle suspension was immediately diluted with 1.6 mL of 10 mM phosphate-buffered saline (PBS) and incubated at room temperature for 30 minutes. LNPs were purified via three cycles of ultrafiltration (Amicon, 100 kDa) with PBS buffer exchange, followed by volume adjustment to 0.6 mL. The final DSPE-PEG concentrations in the purified LNPs were 0.0737 mg/mL or 0.110 mg/mL, depending on the initial formulation.
2. Preparation of DSPE-PEG-VHH Conjugates:

To prepare the conjugates for post-insertion, DSPE-PEG-N₃ (4 mg) was dissolved in 0.1 mL ethanol, generating a 40 mg/mL stock solution. For micelle formation, 1 μL of the stock solution was diluted into 100 μL PBS, yielding a working concentration of 0.4 mg/mL. A 10 μL aliquot of VHH-DBCO solution (3.2 mg/mL) was mixed with 10 μL of the micellar solution to achieve a DSPE-PEG-N₃: VHH-DBCO molar ratio of 1: 1.5 (final DSPE-PEG-N₃ concentration =0.2 mg/mL) . The click reaction proceeded at room temperature for 16 hours, yielding DSPE-PEG-VHH conjugates.
3. Post-Insertion into LNPs:

To perform the insertion, DSPE-PEG-VHH conjugates (5, 10, or 20 μL, corresponding to 0.2 mg/mL DSPE-PEG equivalents) were added to 0.2 mL of pre-formed LNPs. This corresponds to final conjugation densities of 0.062%, 0.125%, and 0.25%of total lipid molarity. The insertion reaction was carried out by incubating the mixture at 45 ℃ for 1 hour with gentle agitation. Free, non-incorporated conjugates were removed via three cycles of ultrafiltration (Amicon, 100 kDa) using 20 mM Tris-HCl (pH 7.4) , and the final volume was adjusted to 0.2 mL.

tLNPs were generated using both post-insertion and surface conjugation techniques under equivalent lipid compositions corresponding to Formulations C and E, with conjugated VHH densities ranging from 0.062%to 0.25%. The physicochemical properties of the resulting tLNPs-including particle size, polydispersity index (PDI) , and mRNA encapsulation efficiency-are summarized in Table 32. As shown in Table 32, the method of preparation significantly impacted the physical characteristics of the nanoparticles. Particle sizes ranged from approximately 100 to 1033 nanometers, while PDI values varied from 0.1 to 0.5, indicating differences in particle uniformity. Encapsulation efficiency also varied widely across formulations, ranging from 47%to 92%.
Table 32. Information on the formulations and characterization of different LNPs

The targeting delivery efficiency was evaluated using both cell lines and PBMCs. As depicted in Figure 21, under two different formulations and four levels of VHH, tLNPs prepared using the post-insertion technique showed increased expression in Jurkat cells and primary T cells but were significantly lower than those prepared using the surface conjugation technique. Moreover, it is noteworthy that tLNPs prepared via the post-insertion technique exhibited substantially higher reporter gene expression in monocytes compared to those prepared via surface conjugation. This study demonstrates that using the surface conjugation technique, along with the preparation processes disclosed in this application, results in tLNPs with significantly better targeting delivery efficiency than those prepared by the post-insertion technique. This highlights the advantages of the disclosed methods in achieving precise and efficient targeting, especially when using optimized formulations and conjugation strategies.
Example 23: In Vitro Characterization of tLNP Targeting CD7 with Different Anti-CD7 VHH Clones

We prepared tLNPs using seven different anti-CD7 VHH clones to further validate the tLNP preparation method protected in this application. The original anti-CD7 VHH sequences were engineered to include a C-terminal extension consisting of a histidine-tag, a hinge sequence, and a cysteine residue. The sequences of these VHH clones are listed in Table 33. The VHHs were transiently expressed in Expi293 cells and purified, and VHH-linker conjugates were prepared following the method described in Example 1.
Table 33. Modified anti-CD7 VHH sequence

In this study, two optimized formulations, Formulation C and Formulation E, were used to prepare anchor-modified LNPs following the procedure described in Example 2. Subsequently, tLNPs were prepared with different amounts of VHH using the method described in Example 3. Information on the prepared CD7-targeting LNPs, including particle size, polydispersity index (PDI) , and mRNA encapsulation efficiency, is summarized in Table 34.
Table 34. Information on the formulations and characterization of different LNPs

The targeted delivery efficiency of the reporter gene was evaluated using cell lines and human PBMC. Jurkat cells were used as target cells, while two B lymphoma cell lines (Raji and Daudi) served as non-target cells. As shown in Figure 22, across almost all VHH clones tested, tLNPs prepared with Formulations C (Figure 22A) and E (Figure 22B) demonstrated strong gene expression in Jurkat cells while maintaining extremely low expression levels in B cells over a VHH amounts range of 0.031%to 0.5%. Further validation was conducted using human PBMCs, where T cells and NK cells, which highly express CD7, were designated as target cells, while monocytes were used as non-target cells. Consistent with the cell line results, tLNPs prepared with Formulations C (Figure 23A) and E (Figure 23B) exhibited clear targeting efficiency within the 0.031%to 0.5%VHH amount range. The highest targeting delivery efficiency was observed in the VHH dosage range of 0.062%to 0.25%. These findings demonstrate that the tLNP preparation technology protected in this application exhibits excellent versatility across different VHH clones, effectively enabling targeted delivery with strong specificity and minimal off-target effects.
Example 24: InVitro Characterization of tLNP Targeting T and NK Cells with Different Ionizable
Lipids

In this study, we evaluated the targeted delivery efficiency of tLNPs prepared using different ionizable lipids. Anchor-modified lipid nanoparticles were prepared following the method described in Example 2, with the exception that five different ionizable lipids were used. These included three novel ionizable lipid structures (Compound 061, Compound 062, and Compound 104) , an FDA-approved ionizable lipid (SM-102) , and a clinically tested ionizable lipid (CIN16645) . The formulation followed the previously optimized Formulation C ratio. Anti-CD7 VHH was used as the targeting ligand, and tLNPs were prepared according to the method described in Example 3, encapsulating mRNA encoding GFP protein for targeted delivery to T and NK cells. Information on the prepared CD7-targeting LNPs is summarized in Table 35.
Table 35. Information on the formulations and characterization of different LNPs

To assess in vitro targeting efficiency, peripheral blood PBMCs were used, with T and NK cells serving as target cells and monocytes as non-target cells. As shown in Figure 24, all tested ionizable lipids resulted in significantly higher delivery efficiency in target cells compared to non-target cells. Notably, the different ionizable lipids exhibited varying levels of delivery efficiency in target cells. Among them, the three novel ionizable lipids, Compound 061, Compound 062, and Compound 104, demonstrated superior delivery efficiency in target cells compared to SM-102 and CIN16645. These results highlight the impact of ionizable lipid selection on tLNP targeting performance and suggest that the novel ionizable lipid candidates may provide enhanced delivery efficiency for targeted applications.
Example 25: In Vivo Characterization of tLNP Targeting T and NK Cells with Different Formulations

To further evaluate the in vivo delivery efficiency of CD7-targeting tLNPs, we utilized CD7-humanized mice. In these hCD7 mice, the exons of the murine Cd7 gene encoding the extracellular domain were replaced with human CD7 exons, allowing for the detection of human CD7 expression in homozygous hCD7 mice. Following the procedure described in Example 1, ligand-linker conjugates were prepared using a modified anti-CD7 VHH (clone S-VHH12) . Anchor-modified lipid nanoparticles were then prepared as described in Example 2, and CD7-targeted tLNPs encapsulating tdTomato mRNA were formulated using the method described in Example 3. Compound 062 was used as the ionizable lipid. Based on previous formulation experience, total PEG-lipid content was set at 1%and 1.5%, with Anchor/total PEG-lipid ratios maintained at 50%. Additionally, as a control, we used a liver-targeting formulation (LNPL1090) , in which compound 062 served as the ionizable lipid. The formulation for this liver-targeting LNP was ionizable lipid: DSPC: Chol: DMG-PEG = 50: 10: 38.5: 1.5. The composition details and physicochemical properties of these LNPs are summarized in Table 36.
Table 36. Information on the formulations and characterization of different LNPs

The in vivo study was conducted following the method described in Example 7. After intravenous administration of LNPs into hCD7 mice, blood and tissue samples were collected 24 hours post-injection for analysis. The positivity of the reporter gene in T cells, NK cells, and non-target cells was evaluated. Since only a subset of T cells expresses hCD7 in this animal model, we also analyzed tdTomato expression specifically in CD7⁺ T cells.

As shown in Figure 25, liver-targeting LNPs did not exhibit significant reporter gene expression in T cells or NK cells in peripheral blood and spleen. In the spleen, the expression of the reporter gene in CD7⁺ T cells targeted by liver-specific LNPs was lower than in non-target monocytes. In comparison, all four tested CD7-targeting tLNPs exhibited excellent reporter gene expression in target cells in peripheral blood and spleen. These results indicate that the CD7-targeting tLNPs prepared using the VHH-linker conjugate via click reaction efficiently deliver cargo to T and NK cells, demonstrating strong specificity and superior targeting capability compared to classic LNP formulation.
Example 26: InVitro Characterization of tLNP Targeting HSC Cells with Different Anti-CD90 VHH
Clones

In this study, we prepared tLNPs using three different anti-CD90 VHH clones to evaluate their targeting efficiency. CD90 is highly expressed on hematopoietic stem cells (HSCs) , making CD90-targeting LNPs a promising tool for HSC-specific delivery. The anti-CD90 VHH sequences were engineered to include a C-terminal extension consisting of a histidine-tag, a hinge sequence, and a cysteine residue. The sequences of the three VHH clones are listed in Table 37. These VHHs were transiently expressed in Expi293 cells and purified, and VHH-linker conjugates were prepared following the method described in Example 1.
Table 37. Modified anti-CD90 VHH sequence

For the tLNP preparation, two optimized formulations, Formulation C and Formulation E, were used. Anchor-modified LNPs were prepared following the procedure described in Example 2. tLNPs were then formulated using the method described in Example 3, with varying amounts of VHH. The characteristics of the CD90-targeting LNPs, including particle size, PDI, and encapsulation efficiency, are summarized in Table 38.
Table 38. Information on the formulations and characterization of different LNPs

The targeting delivery efficiency of the reporter gene was assessed in cell lines. CHO cells served as non-target cells, while CD90-overexpressing CHO cells and HEL cells (a human erythroleukemia cell line) were used as target cells. As shown in Figure 26, when using Formulations C and E, nearly all tested VHH clones exhibited strong expression in target cells over a VHH conjugation range of 0.062%to 0.5%, while maintaining minimal expression in non-target cells. Among the three different VHH clones tested, the optimal delivery efficiency was observed at a VHH conjugation range of 0.125%to 0.5%. This finding contrasts with previous studies on CD5-targeting and CD7-targeting tLNPs, where optimal targeting efficiency was achieved at lower VHH conjugation levels (0.062%to 0.25%) . This discrepancy suggests that the optimal VHH conjugation density may vary depending on the target cell type and receptor characteristics, highlighting the importance of fine-tuning ligand density for different targeting applications. And among the three different VHH clones tested, the clone H. 2346A5 show highest MFI value in both CD90CHO cells and HEL cells.
Example 27. InVitro Characterization ofCD117-Targeted tLNPs Prepared with Different VHH Clones

In this study, targeted lipid nanoparticles (tLNPs) were formulated using three distinct anti-CD117 VHH clones to evaluate their comparative targeting efficiency. CD117 (also known as c-Kit) is a receptor tyrosine kinase that is highly expressed on hematopoietic stem cells (HSCs) , making it a promising target for stem cell-specific delivery applications. Each of the three anti-CD117 VHH sequences was engineered to include a C-terminal extension comprising a hexahistidine tag (His-tag) for purification, a proline-rich hinge sequence (Hinge-5) to enhance structural rigidity, and a terminal cysteine residue to enable site-specific conjugation. The full amino acid sequences of these VHH clones are provided in Table 39.
Table 39. Modified anti-CD117 VHH sequence

The VHH constructs were transiently expressed in Expi293 cells and subsequently purified. VHH-linker conjugates were synthesized using the thiol-maleimide chemistry described in Example 1. For the preparation of tLNPs, two previously optimized formulations, Formulation C and Formulation E, were employed. Anchor-modified lipid nanoparticles were prepared according to the method described in Example 2, followed by ligand conjugation and tLNP formulation as described in Example 3. Different ligand conjugation levels were tested, ranging from 0.125%to 0.5%. The physicochemical characteristics of the resulting CD117-targeted LNPs, including particle size, polydispersity index (PDI) , and mRNA encapsulation efficiency, are summarized in Table 40.
Table 40. Information on the formulations and characterization of different LNPs

To assess the delivery performance, delivery efficiency was evaluated using multiple cell lines. Wild-type CHO cells served as non-target controls, while CD117-overexpressing CHO cells and HEL cells (a human erythroleukemia line known to express CD117 endogenously) were used as target cells. As shown in Figure 27, tLNPs formulated with either Formulation C or E exhibited strong, target-specific reporter gene expression in both CD117⁺ CHO and HEL cells across a VHH conjugation density range of 0.125%to 0.5%. In contrast, minimal gene expression was observed in wild-type CHO cells, confirming the targeting specificity of the tLNPs.

Among the three VHH clones tested, the clone H. 2346B13 demonstrated the highest mean fluorescence intensity (MFI) in CD117-overexpressing CHO cells, indicating superior targeting and delivery in this engineered model. Interestingly, in HEL cells, a different VHH clone achieved the highest MFI, suggesting that optimal clone selection may vary depending on the cellular context and surface expression of CD117. These results are consistent with findings in previous experiments targeting CD90, where optimal delivery efficiency was similarly observed at VHH conjugation levels between 0.125%and 0.25%.
Example 28: In Vitro Gene Editing of HSC Cells Using Targeted LNPs Loaded with CRISPR/Cas12b

In this study, we evaluated the gene editing capability of tLNPs using the tLNP preparation technology protected in this application. The tLNPs were designed to deliver gene-editing components and were formulated to encapsulate mRNA encoding CRISPR-AaCas12bMax and sgRNA targeting the HBG1/2 promoter. Editing of this gene is expected to induce indel mutations, thereby re-activating fetal globin expression, which has the potential to serve as a therapeutic approach for beta-thalassemia and sickle cell disease.

We utilized AaCas12bMax, a highly active variant of Alicyclobacillus acidiphilus-derived Cas12b (AaCas12b) , as the gene-editing enzyme. The amino acid sequence of AaCas12bMax is provided in Table 41. The mRNA encoding AaCas12bMax was synthesized via in vitro transcription (IVT) following the protocol below: The coding sequence of the AaCas12bMax was inserted into the plasmid to form the DNA template which contains sequences of the target protein, 5’ and 3’ UTRs, and a T7 promoter upstream of the 5’ UTR. PCR products which contain a poly A tail from the above plasmid is used for in vitro transcription (IVT) . T7 RNA polymerase recognizes the T7 promoter of the DNA template and initiates the in vitro mRNA transcription. The 5’ cap structure of the in vitro transcribed mRNA was capped by the addition of cap analogs during IVT and N1-methyl-pseudo-U was used to substitute UTPs. All IVT reactions were performed at 37 ℃ for 2 hours. And DNA template is removed by DNase I at 37 ℃ for 30 minutes, then column-purification performed to obtain full length IVT mRNA.

The sgRNA was designed to target the LRF binding motif at -200 region of the HBG1/HBG2 gene promoters. The sgRNA was optimized with chemical modifications and the spacer was extended to 23 nt. Optimized sgRNA was synthetized by GenScript Biotech Corp. and dissolved to 100 μM with water.
Table 41. The sequence of CRISPR-AaCas12bMax and sgRNA

To achieve precise gene editing in HSCs, we prepared CD117-and CD90-targeting LNPs encapsulating CRISPR-AaCas12bMax mRNA and sgRNA. Ligand-linker conjugates were prepared following Example 1, using a modified anti-CD117 VHH (Clone H. 2346B13) and a modified anti-CD90 VHH (Clone H. 2346A5) . Anchor-modified lipid nanoparticles were prepared as described in Example 2. tLNPs encapsulating CRISPR mRNA and sgRNA were formulated according to Example 3, maintaining an AaCas12bMax mRNA: sgRNA weight ratio of 1: 1. The final LNP concentration was determined based on the total RNA concentration in the formulation. The detailed composition and physicochemical properties of the CD117-and CD90-targeting LNPs are summarized in Table 42.
Table 42. Information on the formulations and characterization of different LNPs

HEL cells were exposed to LNP-922/923 (1-4 μg/mL dose) at a 4E+05/mL cell concentration for 24 hours. HEL cells were cultured for another 48h and harvested for genomic DNA extracting. Libraries were constructed by amplifying the region surrounding the on-target sgRNA binding site and sequenced by next-generation sequencing (NGS) . NGS results were analyzed with Cas-Analyzer (www. rgenome. net/cas-analyzer) for indels assessment. The indel frequencies revealed that both LNP-922 and LNP-923 exhibited high and dose-dependent editing efficiencies, indicating its potentiality for in vivo editing.

While preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. It is not intended that the invention be limited by the specific examples provided within the specification. While the invention has been described with reference to the aforementioned specification, the descriptions and illustrations of the embodiments herein are not meant to be construed in a limiting sense. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the invention. Furthermore, it shall be understood that all aspects of the invention are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention. It is therefore contemplated that the invention shall also cover any such alternatives, modifications, variations or equivalents. It is intended that the following claims define the scope of the invention and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

A nucleic acid delivery vector comprising a VHH-linker conjugate, wherein the VHH-linker conjugate comprises:

a) a VHH;

b) a linker connected to the VHH via a conjugation moiety, wherein the linker comprises a moiety Y; and

c) a hinge comprising an amino acid sequence, wherein the hinge is positioned between the VHH and the linker.
The nucleic acid delivery vector according to claim 1, wherein said nucleic acid delivery vector is a ligand-conjugated lipid nanoparticle (LNP) , and said ligand-conjugated LNP comprises anchor-modified LNP, said anchor-modified LNP comprises:

a) LNP; and

b) one or more anchor fragment, wherein said anchor fragment comprises moiety X;

wherein said moiety Y of the linker is capable of forming a linkage with said moiety X of the anchor fragment via bio-orthogonal click reaction.
The nucleic acid delivery vector according to any one of claims 1-2, wherein said VHH is a heavy-chain only antibody with a molecular mass under 30kDa.
The nucleic acid delivery vector according to any one of claims 1-3, wherein said VHH is a heavy-chain only antibody with a molecular mass of about 15kDa.
The nucleic acid delivery vector according to any one of claims 1-4, wherein said VHH is modified with a sequence of the hinge followed by a conjugation-enabling moiety at the terminus, and wherein said sequence of the hinge does not comprise a G4S motif.
The nucleic acid delivery vector according to any one of claims 1-5, wherein said hinge is a structurally constrained hinge comprising no more than 20%glycine (G) residues by amino acid composition.
The nucleic acid delivery vector according to any one of claims 1-6, wherein said hinge is a structurally constrained hinge comprising no more than 10%glycine (G) residues by amino acid composition.
The nucleic acid delivery vector according to any one of claims 5-7, wherein said hinge comprises one or more rigidity-enhancing residue.
The nucleic acid delivery vector according to claim 8, wherein said rigidity-enhancing residue is selected from the group consisting of proline (P) , tyrosine (Y) , phenylalanine (F) , tryptophan (W) , valine (V) , isoleucine (I) , and leucine (L) .
The nucleic acid delivery vector according to any one of claims 8-9, wherein said rigidity-enhancing residue is comprised in said hinge in a proportion of no less than 20%.
The nucleic acid delivery vector according to any one of claims 5-7, wherein said hinge comprises one or more α-helix-stabilizing residue.
The nucleic acid delivery vector according to claim 11, wherein said α-helix-stabilizing residue is selected from the group consisting of alanine (A) , leucine (L) , glutamate (E) , and methionine (M) .
The nucleic acid delivery vector according to any one of claims 11-12, wherein said α-helix-stabilizing residue is comprised in said hinge in a proportion of no less than 80%.
The nucleic acid delivery vector according to any one of claims 5-7 or 11-13, wherein said hinge comprises no α-helix disruptors.
The nucleic acid delivery vector according to claim 14, wherein said α-helix disruptor is selected from the group consisting of proline (P) and glycine (G) .
The nucleic acid delivery vector according to any one of claims 5-7, wherein said hinge comprises one or more potential O-linked glycosylation sites.
The nucleic acid delivery vector according to claim 16, wherein said potential O-linked glycosylation site comprises a serine (S) residue or a threonine (T) residue positioned adjacent to a proline (P) residue within the hinge.
The nucleic acid delivery vector according to any one of claims 16-17, wherein said hinge comprises within a stretch of 10 contiguous amino acids at least two serine (S) and/or threonine (T) residues, each of which is located adjacent to a proline (P) residue.
The nucleic acid delivery vector according to any one of claims 16-17, wherein said serine (S) and/or threonine (T) residues capable of O-linked glycosylation account for at least 20%, 30%, or 40%of the total amino acid residues within the hinge.
The nucleic acid delivery vector according to any one of claims 5-7, wherein said hinge comprises a motif selected from SPSTPP (SEQ ID NO: 64) , PSTPPSP (SEQ ID NO: 65) , or other serine/threonine-rich motifs capable of O-glycosylation.
The nucleic acid delivery vector according to any one of claims 5-20, wherein said hinge comprises one or more hydrophilic residue.
The nucleic acid delivery vector according to claim 21, wherein said hydrophilic residue is selected from the group consisting of serine (S) , threonine (T) , asparagine (N) , glutamine (Q) , tyrosine (Y) , aspartic acid (D) , glutamic acid (E) , lysine (K) , arginine (R) , and histidine (H) , preferably said hydrophilic residue is selected from the group consisting of lysine (K) , arginine (R) , aspartic acid (D) , and glutamic acid (E) .
The nucleic acid delivery vector according to any one of claims 5-22, wherein said hinge comprises none cysteine residues.
The nucleic acid delivery vector according to any one of claims 5-23, wherein said hinge is a restricted hinge or a rigid hinge.
The nucleic acid delivery vector according to any one of claims 5-24, wherein said hinge comprises an amino acid sequence as described in any one of SEQ ID NO: 1, 4-12, 13-21 and 59-62.
The nucleic acid delivery vector according to any one of claims 5-25, wherein said conjugation-enabling moiety is linked with said ligand via a rigid hinge of human IgA1.
The nucleic acid delivery vector according to any one of claims 5-26, wherein said hinge comprises an amino acid sequence as described in SEQ ID NO: 1.
The nucleic acid delivery vector according to any one of claims 1-27, wherein said conjugation-enabling moiety comprises a cysteine residue, a peptide tag for enzymatic conjugation, or an unnatural amino acid bearing a bio-orthogonal reactive group.
The nucleic acid delivery vector according to any one of claims 1-28, wherein said conjugation-enabling moiety is a cysteine residue.
The nucleic acid delivery vector according to any one of claims 1-29, wherein said linker comprises the formula selected from one of them:
The nucleic acid delivery vector according to any one of claims 1-30, wherein said anchor fragment comprises polyethylene glycol (PEG) amphiphilic polymer, polyethylene glycol (PEG) conjugated lipid (also known as PEG lipid) and/or polyethylene glycol (PEG) conjugated hydrophobic polymer.
The nucleic acid delivery vector according to any one of claims 1-31, wherein said anchor fragment comprises a lipid or a hydrophobic polymer linked to said moiety X via a PEG linker having a formula of – (OCH2CH2) _n –, wherein n is any integer ranging from 1 to 135.
The nucleic acid delivery vector according to any one of claims 1-32, wherein said anchor-modified LNP has said moiety X located on the outer surface of the LNP.
The nucleic acid delivery vector according to any one of claims 31-33, wherein said lipid of the anchor fragment is di-stearoyl-phosphatidyl-ethanolamine (DSPE) , diphosphatidylglycerol (DPG) , dipalmitoyl phosphatidyl ethanolamine (DPPE) , distearoyl-rac-glycero (DSG) or cholesterol.
The nucleic acid delivery vector according to any one of claims 31-34, wherein said hydrophobic polymer of the anchor fragment is poly (lactic-co-glycolic acid) or poly (lactic acid) .
The nucleic acid delivery vector according to any one of claims 1-35, wherein said anchor fragment comprises the formula selected from:

wherein n is any integer ranging from 22 to 117.
The nucleic acid delivery vector according to any one of claims 1-36, wherein components constituting LNP comprise ionizable lipids, neutral lipids, steroids and/or PEG lipids.
The nucleic acid delivery vector according to claim 37, wherein said neutral lipid is distearoylphosphatidylcholine (DSPC) or dioleoyl-phosphatidylethanolamine (DOPE) .
The nucleic acid delivery vector according to claim 37, wherein said steroid is cholesterol, sitosterol or stigmasterol.
The nucleic acid delivery vector according to claim 37, wherein said PEG lipid is 1, 2-Distearoyl-sn-glycero-3-phosphoethanolamine-Poly (ethylene glycol) (DSPE-PEG) , 1, 2-Distearoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DSG-PEG) , 1, 2-Dipalmitoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DPG-PEG) , 1, 2-Dipalmitoyl-sn-glycero-3-phosphoethanolamine-poly (ethylene glycol) (DPPE-PEG) , or 1, 2-Dimyristoyl-sn-glycero-3-glycerol-poly (ethylene glycol) (DMG-PEG) .
The nucleic acid delivery vector according to claim 37, wherein said ionizable lipid has formula (I) , or a salt, tautomer, or stereoisomer thereof,

wherein:

m and p are independently selected from any integer ranging from 3 to 8;

n is selected from any integer ranging from 2 to 4;

X is a bond, -C (O) O-, -OC (O) -, -OC (O) O-, or a biodegradable group;

R₁ is a hydrogen bond donor-containing group or a hydrogen bond acceptor-containing group;

both of R₂ are same and selected from C₁-C₁₂ alkyl, substituted C₁-C₁₂ alkyl, C₂-C₁₂ alkenyl, substituted C₂-C₁₂ alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof;

R₃ is selected from C₄-C₂₂ alkyl, substituted C₄-C₂₂ alkyl, C₄-C₂₂ alkenyl, substituted C₄-C₂₂ alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof; or R₃ is an acetal group ofwherein both of R₄ are same and selected from C₁-C₁₆ alkyl, substituted C₁-C₁₆ alkyl, C₂-C₁₆ alkenyl, substituted C₂-C₁₆ alkenyl, C₃-C₁₂ cycloalkyl, substituted C₃-C₁₂ cycloalkyl and combinations of thereof.
The nucleic acid delivery vector according to claim 37 or 41, wherein said ionizable lipid has formula (II) , or a salt, tautomer, or stereoisomer thereof,

wherein q is selected from any integer ranging from 2 to 4.
The nucleic acid delivery vector according to claim 41 or 42, wherein R₁ of said ionizable lipid is selected from hydroxyalkyl group having 1 to 5 carbon atoms, or optionally substituted amino alkenyl group having 1 to 6 carbon atoms.
The nucleic acid delivery vector according to any one of claims 41-43, wherein R₁ of said ionizable lipid is selected from one of the following formulae:

wherein o is selected from 1, 2, 3, 4, and 5.
The nucleic acid delivery vector according to any one of claims 41-44, wherein R₂ of said ionizable lipid is selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl, C₂-C₁₂ alkenyl and combinations of thereof.
The nucleic acid delivery vector according to any one of claims 41-45, wherein R₂ is C₃-C₁₀ alkyl.
The nucleic acid delivery vector according to any one of claims 41-46, wherein R₃ of said formula (I) is selected from C₆-C₁₂ alkyl, C₆-C₁₂ cycloalkyl, C₆-C₁₂ alkenyl and combinations of thereof.
The nucleic acid delivery vector according to any one of claims 41-47, wherein R₃ is the same with R₂.
The nucleic acid delivery vector according to any one of claims 41-48, wherein:

a) each R₂ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains; and/or

b) R₃ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains.
The nucleic acid delivery vector according to any one of claims 42-49, wherein R₂ and R₄ of said formula (II) are independently selected from C₂-C₁₂ alkyl, C₃-C₁₂ cycloalkyl, C₂-C₁₂ alkenyl and combinations of thereof.
The nucleic acid delivery vector according to any one of claims 41-50, wherein R₂ and R₄ are independently C₃-C₁₀ alkyl.
The nucleic acid delivery vector according to any one of claims 41-51, wherein R₄ is the same with R₂.
The nucleic acid delivery vector according to any one of claims 41-52, wherein:

a) each R₂ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains; and/or

b) each R₄ has at least one carbon atom with hydrogen atom (s) substituted by one or two side chains.
The nucleic acid delivery vector according to claim 49 or 53, wherein said carbon atom with hydrogen atom (s) substituted by one or two side chains is the second or more distant carbon atom counting from the junction.
The nucleic acid delivery vector according to claim 54, wherein said one or two side chains are C₁-C₄ alkyls.
The nucleic acid delivery vector according to any one of claims 41-55, wherein each R₂, R₃ and/or each R₄ is independently selected from one of the following formulas:
The nucleic acid delivery vector according to any one of claims 37 or 41-56, wherein said ionizable lipid is selected from the group consisting of Compound Nos: 002-011 and 013-103 as shown in Table 1.
The nucleic acid delivery vector according to any one of claims 37-57, wherein said ionizable lipid is comprised in an amount of about 40-60 mol%, said neutral lipid is comprised in an amount of about 5-20 mol%, said steroid is comprised in an amount of about 25-50 mol%, said PEG lipid is comprised in an amount of about 0.1-5 mol%, and said anchor fragment is comprised in an amount of about 0.1-1 mol%, when taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.
The nucleic acid delivery vector according to any one of claims 37-58, wherein said ionizable lipid is comprised in an amount of about 50 mol%, said neutral lipid is comprised in an amount of about 10 mol%, said steroid is comprised in an amount of about 38-39.5 mol%, said PEG lipid is comprised in an amount of about 0.25-1.75 mol%, and said anchor fragment is comprised in an amount of about 0.25-1 mol%, when taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.
The nucleic acid delivery vector according to any one of claims 31-59, wherein said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 0.5-3 mol%, taking the components constituting said LNP of (a) and said anchor fragments of (b) totally as 100 mol%.
The nucleic acid delivery vector according to claim 60, wherein said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 1 mol%, wherein the PEG lipid characterized in anchor fragment is equivalent to the PEG lipid characterized in LNP.
The nucleic acid delivery vector according to claim 60, wherein said PEG lipid of the components constituting said LNP and the anchor fragment is collectively comprised in an amount of about 1 mol%, wherein the PEG lipid characterized in LNP is comprised in an amount of 0.
The nucleic acid delivery vector according to any one of claims 1-62, wherein said moiety X and said moiety Y are selected from the group consisting of the following click chemistry reactive partners: 1) azide and dibenzocyclooctyne (DBCO) , 2) azide and 4-dibenzocyclooctynol (DIBO) , 3) azide and biarylazacyclooctynone (BARAC) , 4) azide and alkyne, 5) tetrazine and trans-cyclooctene (TCO) , 6) tetrazine and cyclopropane, or 7) azide and bicyclononyne (BCN) , wherein said click chemistry reactive partners of each group are interchangeable between said moiety X and moiety Y.
The nucleic acid delivery vector according to any one of claims 1-63, wherein the molar ratio of the VHH-linker conjugate, defined by the amount of moiety Y present in the conjugate, relative to the total lipid content in the lipid nanoparticle formulation, is between 0.03 mol%and 0.5 mol%.
The nucleic acid delivery vector according to any one of claims 2-64, wherein components constituting LNP comprise nucleic acid molecules.
The nucleic acid delivery vector according to claim 65, wherein said nucleic acid molecules are RNAs.
The nucleic acid delivery vector according to claim 65 or 66, wherein said nucleic acid molecules are selected from the group consisting of messenger RNA (mRNA) , guide RNA (gRNA) , a short interfering RNA (siRNA) , an RNA interference (RNAi) molecule, a microRNA (miRNA) , an antagomir, an antisense RNA, a ribozyme, a small hairpin RNA (shRNA) , and a mixture thereof.
The nucleic acid delivery vector according to claim 66 or 67, wherein said RNAs are encapsulated in said anchor-modified LNPs.
A method of preparing the nucleic acid delivery vector of any one of claims 1-68, comprising:

1) providing (a) components constituting said LNP and (b) said anchor fragments, thereby allows self-assembly of said anchor-modified LNP under appropriate conditions, wherein each of said anchor fragment comprises a moiety X;

2) providing said VHH-linker conjugate, wherein said linker comprises a moiety Y, wherein said moiety Y is capable of forming a linkage with said moiety X of 1) via bio-orthogonal click reaction;

3) contacting said anchor-modified LNPs of 1) with said VHH-linker conjugate of 2) , thereby allows the formation of said ligand-conjugated lipid nanoparticle (LNP) .
The method according to claim 69, wherein said VHH is conjugated to said linker via a site-specific reaction between said linker and the conjugation-enabling moiety positioned downstream of the hinge.
The method according to claim 70, wherein said site-specific reaction is selected from: a thiol–maleimide reaction, a thiol–vinylsulfone reaction, a thiol–para-fluorophenyl reaction, an enzymatic ligation reaction, or a bio-orthogonal click reaction involving a noncanonical amino acid.
The method according to claim 70, wherein said conjugation-enabling moiety is a cysteine residue and said linker comprises a maleimide group that reacts with the thiol group of said cysteine residue.
The method according to claim 69, wherein said bio-orthogonal click reaction is selected from a group consisting of nucleophilic ring-opening reactions, cycloaddition reactions, nucleophilic addition reactions, thiol-ene reactions, and Diels Alder reactions.
The method according to any one of claims 69-73, wherein said method further comprising: 4) purification of said ligand-conjugated LNP, wherein said purification allows removal of said VHH-linker conjugates of 2) that fail to contact with or form a linkage with said anchor-modified LNPs.