Injectable artificial antigen presenting cells for immune therapy

Abstract

In various aspects and embodiments, the present disclosure provides artificial antigen presenting cells (aAPCs) that are suitable for parenteral administration for immune therapy. In various embodiments, the aAPCs are effective to activate or inhibit target T cells in vivo, including CD8+ or CD4+ T cells. The aAPCs according to this disclosure provide a shelf-stable nanoparticle platform for immune therapies. The shelf-stable nanoparticle platform in various embodiments controls particle size and particle chemistry, ligand design and ligand density, and aAPC aggregation propensity, among other features.

Description

[0001]Attorney docket: NEX-013PC / 107578-5013 INJECTABLE ARTIFICIAL ANTIGEN PRESENTING CELLS FOR IMMUNE THERAPY PRIORITY This Application claims priority to, and the benefit of, US Provisional Application No. 63 / 454,361 filed March 24, 2023, which is hereby incorporated by reference in its entirety. SEQUENCE LISTING The instant application contains a Sequence Listing which has been submitted in XML format via EFS-Web and is hereby incorporated by reference in its entirety. Said XML copy, created on March 25, 2024, is named NEX-013PC_107590-5013_Sequence_Listing and is 61,440 bytes in size. BACKGROUND An antigen-presenting cell (APC) processes and displays antigenic peptides in complexes with major histocompatibility complex (MHC) proteins on their surfaces. Effector cells, such as T-cells, recognize these peptide-MHC (pMHC) complexes through cell-surface receptors, such as T-cell receptors (TCRs). Dendritic cells (DCs) are an example of an antigen presenting cell that can be stimulated to effectively present antigen and support expansion of immune effector cells, thereby activating a cytotoxic response towards an antigen. In some immunotherapies, DCs are harvested from a patient and either pulsed with an antigen or transfected with a viral vector. Upon transfusion back into the patient these activated cells present tumor antigen to effector lymphocytes (e.g. CD4+ T cells, CD8+ T cells, and B cells). When successful, this therapy initiates a cytotoxic response against cells expressing antigens (including tumor antigens). However, there remains a need for shelf- stable pharmaceutical compositions that can effectively present peptide antigens to T cells in vivo (i.e., artificial antigen presenting cells, or aAPCs), to either activate or inhibit T cells in an antigen-specific fashion. This disclosure meets these and other objectives. DB1 / 145613893.1 1 Attorney docket: NEX-013PC / 107578-5013 SUMMARY OF THE DISCLOSURE In various aspects and embodiments, the present disclosure provides artificial antigen presenting cells (aAPCs) that are suitable for parenteral administration for immune therapy. In various embodiments, the aAPCs are effective to activate or inhibit target T cells in vivo, including CD8+ or CD4+ T cells. The aAPCs according to this disclosure provide a shelf- stable nanoparticle platform for immune therapies. The shelf-stable nanoparticle platform in various embodiments controls particle size and particle chemistry, ligand design and ligand density, and aAPC aggregation propensity, among other features. In various aspects and embodiments, the present invention provides shelf-stable compositions and methods for activating or inhibiting antigen-specific T cells in a patient. In addition to shelf-stable properties, the nanoscale aAPCs described herein are designed to provide pharmacodynamic advantages, including with respect to circulating properties, biodistribution, and degradation kinetics. These advantages can result from physical parameters including particle size, surface charge, polydispersity index, polymer composition, ligand conjugation chemistry, ligand density, and peptide loading, among others. In some embodiments, the aAPCs have a polypeptide ligand density that avoids aggregation potential as well as steric constraints from an abundance of ligands on the surface, without loss of activity and / or potency. In some embodiments, the aAPCs persist in peripheral blood circulation sufficiently long to allow distribution to target tissues, including trafficking to lymphoid organs (e.g., lymph nodes) via blood / lymph exchange and / or trafficking to tumors and / or trafficking to target organs. In some embodiments, the aAPCs are suitable for subcutaneous administration. In one aspect, the disclosure provides aAPCs comprising poly(lactic acid)- polyethylene glycol (PLA-PEG) or poly(lactic acid-co-glycolic acid)-polyethylene glycol (PLGA-PEG) co-polymers. In these embodiments of the disclosure, the aAPCs have advantages in stability and ligand density, for example. In these embodiments, the disclosure provides an aAPC suitable for parenteral administration (including subcutaneous administration in some embodiments), and which comprises PLA-PEG or PLGA-PEG DB1 / 145613893.1 2 Attorney docket: NEX-013PC / 107578-5013 copolymers and one or more polypeptide ligands conjugated to PEG polymers through a thioether bond or other conjugation chemistry. The polypeptide ligands comprise HLA ligands (Human Leukocyte Antigen ligands) presenting a peptide antigen, and optionally one or more signal 2 ligands. Signal 2 ligands include signal 2 ligands for T cell activation and / or expansion, or T cell inhibition. In various embodiments, about 40% or less by weight of the copolymers have a functional group for polypeptide ligand coupling. The HLA ligands in various embodiments may be HLA Class I and / or Class II molecular complexes, or portions thereof comprising an antigen-binding cleft. In some embodiments, the HLA molecular complexes are monomeric or dimeric, and may contain additional heterologous sequences, such as immunoglobulin sequences. HLA-fusions (e.g., HLA-Immunoglobulin fusions) in some embodiments provide additional advantages in stability, TCR binding affinity, and / or potency for T cell activation or inhibition. In various embodiments, the aAPCs comprise HLA class I ligands for presentation of peptide antigens to CD8+ T cells (e.g., for activation and / or expansion of CD8+ cells, or inhibition of CD8+ cells). In some embodiments, the HLA class I ligand comprises at least two fusion proteins. A first fusion protein comprises a first HLA class I α chain and a first immunoglobulin heavy chain, and a second fusion protein comprises a second HLA class I α chain and a second immunoglobulin heavy chain. The first and second immunoglobulin heavy chains associate to form the HLA class I molecular complex (e.g., associate through disulfide bonds). The HLA class I molecular complex comprises a first HLA class I peptide binding cleft and a second HLA class I peptide binding cleft. Alternatively, the aAPC comprises HLA class II ligands for presentation of peptide antigens to CD4+ T cells (e.g., for activation and / or expansion of CD4+ cells, or inhibition of CD4+ cells). In some embodiments, the HLA class II molecular complex comprises at least four fusion proteins. Two first fusion proteins comprise (i) an immunoglobulin heavy chain and (ii) an extracellular domain of an HLA class II β chain. Two second fusion proteins comprise (i) an immunoglobulin light chain and (ii) an extracellular domain of an HLA class II α chain. The two first and the two second fusion proteins associate to form the HLA class II molecular complex. The extracellular domain of the HLA class II β chain of each first DB1 / 145613893.1 3 Attorney docket: NEX-013PC / 107578-5013 fusion protein and the extracellular domain of the HLA class II α chain of each second fusion protein form an HLA class II peptide binding cleft. Peptide antigens are bound to an antigen binding cleft of the antigen presenting complex. Peptide antigens for immune therapy of oncological disease, infectious diseases, and autoimmune diseases are described herein. In some embodiments, the peptide antigen does not induce aggregation of the aAPC. That is, peptide antigens are selected with low aggregation propensity when loaded onto aAPCs. Peptides with low aggregation propensity in some embodiments are selected using an Aggregation Potential Score (APS). In some embodiments, the polypeptide ligands comprise a signal-2 ligand that is a co-stimulatory ligand, e.g., for activation and / or expansion of target T cells. Exemplary co- stimulatory ligands include agonists for any one of CD28, 4-1BB, CD27, OX-40, CD30, ICOS, and LIGHT, among others. Co-stimulatory ligands can induce activation and / or expansion of CTLs or Tregs in various embodiments. In other embodiments, and particularly where aAPCs are intended to inhibit target T cells, the polypeptide ligands do not comprise any signal 2 ligand or the polypeptide ligands comprise inhibitory ligands, which induce tolerance or apoptosis of target T cells. In various embodiments, the inhibitory ligand is an agonist for Fas, TGF-β, or PD-1. Agonistic ligands can include natural agonistic ligands (or engineered variants thereof, including immunoglobulin fusions as described), or in some embodiments, antibody agonists. In some embodiments, the co-inhibitor ligand is PD-L1 (or immunoglobulin fusion thereof) or FasL (or immunoglobulin fusion thereof). Antibody agonists can be full monoclonal antibodies, or portions or fragments comprising antigen- binding sequences, such as Fab, Fab', F(ab')2or scFv. In various embodiments, the signal 1 and signal 2 ligands can be combined in homodimeric or heterodimeric constructs. For example, the HLA ligand can comprise fusion of HLA extracellular domains to an immunoglobulin Fc region, such as IgG4 Fc region, which can be dimerized (e.g., through disulfide bonds) with a signal 2-Immunoglobulin (Ig) fusion (i.e., a heterodimeric Ig fusion construct). In still other embodiments, the HLA ligand comprises a fusion to a signal 2 ligand. For example, an HLA extracellular domain can be fused at its C-terminus to an immunoglobulin Fc region, and fused at its N-terminus to a DB1 / 145613893.1 4 Attorney docket: NEX-013PC / 107578-5013 signal 2 ligand, to prepare homodimeric ligands with both signals dimerized. In some embodiments, the signal 2 ligand comprises a single chain antibody (e.g., scFv) or agonistic portion of a natural ligand. In some embodiments, the co-inhibitory ligand is an agonist antibody for Fas. In some embodiments, the agonist antibody for Fas is an IgG4 antibody based on clone CH11. As demonstrated herein, this anti-Fas antibody has activity when it is able to crosslink multiple Fas receptors, such as when the antibody is conjugated to a nanoparticle. In some embodiments, the aAPC further comprises one or more cytokines that support T cell activation and / or expansion or T cell inhibition. The one or more cytokines or functional portion thereof may be conjugated to the aAPC as a polypeptide ligand. Alternatively, the cytokine or functional portion thereof may be fused to a signal 1 or signal 2 polypeptide ligand (which can optionally be presented in homodimeric or heterodimeric Ig fusion constructs as described herein). In some embodiments, the cytokine is encapsulated by the copolymers, and will release cytokine locally in targeted environments (e.g., in lymphoid organs, tumor, or target tissue or organ). Examples of cytokines that may be used include IL-1β, IL-2, IL-4, IL-7, IL-10, IL-12, IL-15, and gamma interferon. For example, IL-2 can be employed with a co-stimulatory signal 2 ligand. In some embodiments, the aAPC comprises a tolerogenic cytokine as a polypeptide ligand. In various embodiments, the one or more peptide antigens are tumor or cancer associated antigens, such as tumor-derived antigens, tumor-specific antigens, and neoantigens. In some embodiments, the target peptide antigens include at least one that is associated with or derived from a pathogen, such as a viral, bacterial, fungal, or parasitic pathogen. In some embodiments, the one or more target peptide antigens is an “autoantigen”, meaning that it is associated with an autoimmune disease or reaction. In various embodiments, the aAPC is comprised in a pharmaceutical composition suitable for administration to a subject. The pharmaceutical composition may have one or more excipients such as buffering agents, surfactants, preservative agents, polymers, bulking agents, and stabilizers. DB1 / 145613893.1 5 Attorney docket: NEX-013PC / 107578-5013 In some aspects and embodiments, the present disclosure provides aAPCs suitable for parenteral administration (including subcutaneous administration), and which have low aggregation propensity. In these aspects, the aAPC comprises a polymeric or lipid nanoparticle comprising a polyethylene glycol (PEG) sheath and one or more polypeptide ligands conjugated to PEG (e.g., the PEG terminus), which can be through a thioether bond or other functional group. The polypeptide ligands comprise HLA Class I or Class II ligands presenting a peptide antigen and optionally one or more signal 2 ligands. The peptide antigen for presentation to T cells does not induce aggregation of the aAPC. For example, in some embodiments, the peptide antigen does not have an exposed Cysteine and / or the peptide antigen has one or more exposed glycine residues or exposed charged residues. In some embodiments, the peptide antigen does not have any Cysteine residue, and contains one or more charged residues (e.g., 1, 2, or 3 charged residues). In some aspects and embodiments, aggregation potential is estimated in silico by determining the mean Aggregation Potential Score (APS) of antigen peptide residues within the HLA antigen binding cleft. In other aspects, the invention provides a method for immunotherapy. The method comprises administering the aAPC or pharmaceutical composition thereof as described herein to a subject in need of treatment. In various embodiments, the subject has cancer or an infectious disease, and the aAPCs comprise a co-stimulatory ligand. In some embodiments, the peptide antigen is selected in a personalized basis for a cancer patient, based on an analysis of the patient's tumor. In some embodiments, the nano-aAPCs are used as a booster vaccine, after adoptive T cell therapy, in which naive T cells from the patient, TILs, or T cells from an HLA-matched donor are expanded ex vivo and administered to the patient. In some embodiments, the subject has an autoimmune condition, and the aAPCs comprise a co-inhibitory signal or do not contain a signal 2 ligand. In some embodiments, the autoimmune condition is type 1 diabetes. Generally, the aAPC or pharmaceutical composition thereof is parenterally administered. For example, the aAPC or pharmaceutical composition thereof is administered by intravenous administration, intra-arterial administration, subcutaneous administration, intradermal administration, intra-lymphatic administration, intramuscular administration, or DB1 / 145613893.1 6 Attorney docket: NEX-013PC / 107578-5013 intratumoral administration. In some embodiments, the aAPC composition is administered by subcutaneous administration. In other aspects of this disclosure, polypeptide ligands for immune therapy (including as polypeptide ligands for aAPCs) are disclosed. Such polypeptide ligands include an anti- Fas agonistic antibody having an IgG isotype (e.g., IgG4), which can be conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen (e.g., associated with an autoimmune disease). In other embodiments, the polypeptide ligand is a dimeric PD-L1 ligand comprising an activating portion of PD-L1, such as amino acids residues F19 to T239 of human PD-L1. Each PD-L1 activating fragment may be fused directly or indirectly through a linker at its C-terminus to an IgG Fc region (e.g., IgG4), and the ligand may be dimerized by disulfide bonds in the Fc region. The dimeric PD-L1 ligand can be conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen and used to drive tolerance to the antigen. In other embodiments, the polypeptide ligand is a dimeric FasL ligand comprising an activating portion of FasL, such as amino acids P132 to L279 of human FasL fused directly or indirectly through a linker at its N-terminus to a dimerized IgG-Fc region (e.g., IgG4). The dimerized Fc region can be conjugated to nanoparticles through a Cys-containing linker in some embodiments. The nanoparticles may further present HLA- peptide antigen ligands to drive apoptosis of antigen-specific T cells. In other embodiments, the polypeptide ligand is a tolerogenic ligand comprising an activating fragment of PD-L1, such as amino acids F19-T239 of human PD-L1, fused directly or through a linker to HLA- immunoglobulin fusion protein. Such ligands can be conjugated to nanoparticles as disclosed herein and used for immunotherapy (to drive tolerance in targeted T cells). In other embodiments, the polypeptide ligand is a co-stimulatory ligand comprising an anti-CD28 agonistic scFv conjugated to an HLA-immunoglobulin fusion protein, providing homodimeric ligands comprising signal 1 and co-stimulatory signal 2 ligands. The scFv may be fused to the HLA sequence through the heavy or light chain sequence. This invention and various embodiments are further described through the following detailed description. DESCRIPTION OF THE FIGURES DB1 / 145613893.1 7 Attorney docket: NEX-013PC / 107578-5013 FIGs. 1A and 1B show the size and surface charge of nanoparticles prepared according to this disclosure. FIG. 1A shows TEM images of naked particles, protein- conjugated nanoparticles, and peptide loaded nanoparticles. FIG. 1B shows (from left to right) the size distribution, average size of particles, polydispersity index (PDI), and surface charge. FIG.2 is an illustration showing conjugation of thiolated ligands to PEG-maleimide functional groups on nanoparticles. FIGs. 3A and 3B show the effect on ligand density (bars) and nanoparticle size (squares) upon altering %PEG-mal on the surface of the nanoparticles (FIG. 3A) or by altering the maleimide:thiol ratio during the coupling reaction (FIG.3B). FIGs.4A and 4B show that use of PLGA-PEG or PLA-PEG do not lead to significant differences in size (bars) or PDI (dot) (FIG. 4A) or protein density (FIG. 4B) (ligands per NP shown by the bar, and µg protein at 10 OD shown by the dot). FIGs.5A-D show that the use of PLGA-PEG or PLA-PEG do not lead to significant differences in antigen-specific CD8+ T cell stimulation. FIG. 5A shows percent IFNγ+ CD8+ T cells. FIG.5B shows percent TNFα+ CD8+ T cells. FIG.5C shows percent IL2+ CD8+ T cells. FIG.5D shows percent CD107a+ CD8+ T cells. FIG.6 shows that antigen-loaded aAPCs according to embodiments of this disclosure traffic to lymph nodes, spleen, and tumor. FIGs 7A and 7B show that systemically administered aAPCs increased antigen- specific T cells in the spleen (A) and tumor (B) of tumor-bearing mice (B16-OVA, with implanted OT-1 T cells). Splenocytes from mice that received peptide-loaded aAPC had greater killing potential as compared to peptide with Complete Freund’s Adjuvant (CFA). FIG.7C. N=2 per arm. In vitro killing assay performed with splenocytes harvested on day 22. FIGs. 8A-C shows that T cells recovered from lymph nodes and tumors have a phenotype consistent with persistence and strong anti-tumor effect. FIG.8A shows T cells (specific for ovalbumin antigen) recovered from lymph nodes in a non-disease model. These DB1 / 145613893.1 8 Attorney docket: NEX-013PC / 107578-5013 T cells exhibited central memory and effector memory phenotype. FIG. 8B quantifies antigen-specific T cells in tumors in a melanoma model (B16F10 cells), 63% of which had at least three effector functions. FIG. 8C quantifies gp100-specific T cells in tumors in a melanoma model, and showing that antigen-specific CTLs recovered from the tumor exhibited a effector memory and central memory phenotype. FIG.9 shows that aAPCs loaded with gp100 antigen extended survival in B16 mouse model (x-axis is days from tumor implantation). FIG.9 shows that aAPC activate anti-tumor T cells in vivo, and in this model, clear lung metastases. FIGs. 10A-C show the construction of an agonistic anti-Fas antibody ligand. FIG. 10A shows construction of an IgG4 antibody based on the variable domain of CH11 clone (which is IgM isotype). The IgG4 antibody is crosslinked using an anti-IgG4 antibody for in vitro testing. FIG. 10B and 10C shows that the anti-Fas agonistic ligand has little to no activity in the absence of crosslinking but shows robust activity upon crosslinking. FIG.11 shows that aAPCs with PD-L1-Ig Signal 2 ligands rapidly inhibit antigen- specific killing of peptide-loaded target cells. FIG.12 shows that aAPCs with anti-Fas Signal 2 ligands rapidly eliminate antigen- specific T cells. FIG.13 shows that anti-Fas aAPCs eliminate up to 90% of MART-1-specific T cells by Day 13 in a mouse model, and that the effect is dose dependent. FIG. 14 is a diagram showing a process for selecting antigens for presenting on aAPCs to mitigate aggregation driven by antigen loading. FIG. 15A and 15B illustrate modelling and scoring of aggregating and non- aggregating peptides, respectively. DETAILED DESCRIPTION In various aspects and embodiments, the present disclosure provides artificial antigen presenting cells (aAPCs) that are suitable for parenteral administration for immune therapy. In various embodiments, the aAPCs are effective to activate or inhibit target T cells in vivo, including CD8+ or CD4+ T cells. The aAPCs according to this disclosure provide a shelf- DB1 / 145613893.1 9 Attorney docket: NEX-013PC / 107578-5013 stable nanoparticle platform for immune therapies. The shelf-stable nanoparticle platform in various embodiments controls particle size and particle chemistry, ligand design and ligand density, and aAPC aggregation propensity, among other features. In various aspects and embodiments, the present invention provides shelf-stable compositions and methods for activating or inhibiting antigen-specific T cells in a patient. In addition to shelf-stable properties, the nanoscale aAPCs described herein are designed to provide pharmacodynamic advantages, including with respect to circulating properties, biodistribution, and degradation kinetics. These advantages result from physical parameters including particle size, surface charge, polydispersity index, polymer composition, ligand conjugation chemistry, ligand density, and peptide loading, among others. In some embodiments, the aAPCs have a polypeptide ligand density that avoids aggregation potential as well as steric constraints from an abundance of ligands on the surface, without loss of activity and / or potency. In some embodiments, the aAPCs persist in peripheral blood circulation sufficiently long to allow distribution to target tissues, including trafficking to lymphoid organs (e.g., lymph nodes) via blood / lymph exchange and / or trafficking to tumors and / or trafficking to target organs. In some embodiments, the aAPCs are suitable for subcutaneous administration. In one aspect, the disclosure provides aAPCs (and pharmaceutical compositions thereof) comprising poly(lactic acid)-polyethylene glycol (PLA-PEG) or poly(lactic acid- co-glycolic acid)-polyethylene glycol (PLGA-PEG) co-polymers. PLA-PEG and PLGA- PEG nanoparticles can be prepared by nanoprecipitation using known processes. In these embodiments of the disclosure, the aAPCs have advantages in stability and ligand density, among other things. In these embodiments, the disclosure provides an aAPC suitable for parenteral administration (including subcutaneous administration in some embodiments), and which comprises PLA-PEG or PLGA-PEG copolymers and one or more polypeptide ligands conjugated to PEG polymers (e.g., conjugated to the PEG terminus) through a thioether bond or other conjugation chemistry (such as conjugation through an amine). The polypeptide ligands comprise HLA ligands (Human Leukocyte Antigen ligands) presenting a peptide antigen, and optionally one or more signal 2 ligands. Signal 2 ligands are described elsewhere herein and include signal 2 ligands for T cell activation and / or expansion, or T DB1 / 145613893.1 10 Attorney docket: NEX-013PC / 107578-5013 cell inhibition. In various embodiments, about 40% or less by weight of the copolymers have a functional group for polypeptide ligand coupling. In various embodiments, about 30% or less, or about 25% of less, or about 20% or less by weight of the copolymers have a functional group for polypeptide ligand coupling. In some embodiments, from about 15% to about 35% by weight of the copolymers have a functional group for polypeptide ligand coupling. In various embodiments, PEG-Mal (by weight) or other PEG-functional group is less than about 10%, or less than about 7%, such as about 5% of the co-polymer weight. As illustrated in FIG.2, only a portion of the copolymers comprise a functional group for ligand coupling. In some embodiments, the functional group is a maleimide functional group at the PEG terminus, with other PEG groups being inert (e.g., comprising an alkyl ether end cap, such as a methyl ether end cap). The end capped PEG may also be referred to herein as mPEG. The reaction between sulfhydryl groups and maleimide is well known in the art. See, for example, João MJM et al., Bioconjugation with Maleimides: A Useful Tool for Chemical Biology, Chemistry (August 2018). Other conjugation chemistries (including amine conjugation) can be used and are well known. Alternative conjugation chemistries are described for example in U.S. Patent No. 10,435,668, which is hereby incorporated by reference in its entirety. For example, molecules can be directly activated with a variety of chemical functionalities, including nucleophilic groups, leaving groups, or electrophilic groups. Activating functional groups include alkyl and acyl halides, amines, sulfhydryls, aldehydes, unsaturated bonds, hydrazides, isocyanates, isothiocyanates, ketones, and other groups known to activate for chemical bonding. Alternatively, a molecule can be bound to a nanoparticle through the use of a small molecule-coupling reagent. Non-limiting examples of coupling reagents include carbodiimides, maleimides, N-hydroxysuccinimide esters, bischloroethylamines, bifunctional aldehydes such as glutaraldehyde, anyhydrides and the like. In other embodiments, a molecule can be coupled to a nanoparticle through affinity binding such as a biotin-streptavidin linkage or coupling. In an exemplary embodiment, the nanoparticles have a PLGA copolymer that can be tuned for a specific biodegradation rate in vivo (by adjusting the LA:GA ratio and / or DB1 / 145613893.1 11 Attorney docket: NEX-013PC / 107578-5013 molecular weight of the PLGA polymer). In exemplary embodiments, the PLGA is based on a LA:GA ratio of from 20:1 to 1:20, including compositions of L / G of: 5 / 95, 10 / 90, 15 / 85, 20 / 80, 25 / 75, 30 / 70, 35 / 65, 40 / 60, 45 / 55, 50 / 50, 55 / 45, 60 / 40, 65 / 35, 70 / 30, 75 / 25, 80 / 20, 85 / 15, 90 / 10, or 95 / 5. PLGA degrades by hydrolysis of its ester linkages. The time required for degradation of PLGA is related to the ratio of monomers: the higher the content of glycolide units, the lower the time required for degradation as compared to predominantly lactide units. In some embodiments, the PLGA is 50 / 50 L / G. In some embodiments, the aAPCs may further comprise PLGA or PLA polymers (in addition to the copolymers). In some embodiments, the PLGA or PLA polymer has a molecular weight in the range of about 15 kDa to about 35 kDa, or from about 15 kDa to about 25 kDa. In an exemplary embodiment, the PLGA or PLA polymer has a molecular weight of about 20 kDa. In various embodiments, the PLGA or PLA polymer is present at about 5 wt% to about 40 wt%, or about 5 wt% to about 25 wt%, based on the total weight of the polymers and co-polymers. In various embodiments, the ratio of PEG-maleimide groups to mPEG limits the density of the polypeptide ligands conjugated through, for example, sulfhydryl groups. In various embodiments, the aAPC has about 10 to about 500 polypeptide ligands per particle. In some embodiments, the aAPC has about 50 to about 400 polypeptide ligands per particle. In some embodiments, the aAPC has about 100 to about 300 polypeptide ligands per particle. In various embodiments, by refining the peptide density, aAPC particles can be prepared with potent T cell effector properties (activation or inhibition), while avoiding undesirable properties including but not limited to aAPC aggregation and steric limitations. In some embodiments, other means can also be employed to control polypeptide ligand density on the particles, such as by limiting the amount of thiolated polypeptides during the coupling reaction. In various embodiments, the PLA or PLGA portion of the copolymers have molecular weights of from about 15 kDa to about 50 kDa. In some embodiments, the PLA or PLGA portion of the copolymers have molecular weights in the range of about 15 kDa to DB1 / 145613893.1 12 Attorney docket: NEX-013PC / 107578-5013 about 35 kDa, or from about 15 kDa to about 25 kDa. In an exemplary embodiment, the PLA or PLGA portion of the copolymers have a molecular weight of about 20 kDa. In various embodiments, the PEG portions of the copolymers have molecular weights in the range of about 2 kDa to about 10 kDa, or in the range of about 2 kDa to about 7 kDa. In an exemplary embodiment, the PEG portion of the copolymers have molecular weights in the range of about 2 kDa and about 5 kDa. In some embodiments, the PEG portions having a functional group for polypeptide ligand coupling have molecular weights of about 5 kDa, and the PEG portions without functional groups for ligand coupling have molecular weights of about 3 kDa. In such embodiments the smaller mPEG moieties (as compared to PEG- functional group) limit steric constraints on polypeptide ligand binding to targets and / or improve conjugation efficiency. Thus, in exemplary embodiments, the PLA or PLGA portions of the copolymers have molecular weights of about 20 kDa, the PEG-functional group portions of the co-polymers have molecular weights of about 5 kDa, and the mPEG portions of the co-polymers have molecular weights of about 3 kDa. In various embodiments, the aAPC has a diameter of from about 50 nm to about 150 nm. In some embodiments, the aAPC has a diameter of from about 50 nm to about 130 nm. In some embodiments, the aAPC has a diameter of from about 50 nm to about 120 nm. In some embodiments, the aAPC has a diameter of from about 50 nm to about 100 nm or from about 50 nm to about 75 nm. In exemplary embodiments, the aAPC has a diameter of about 60 nm, about 80 nm, about 90 nm, about 100 nm, about 110 nm, or about 120 nm. In some embodiments, the aAPC population has a size distribution with polydispersity index (PDI) of less than 0.2. In various embodiments, the aAPC has a surface charge of from about 0 to -15 mV, or from about 0 to about -10 mV. For example, the aAPC may have a surface charge of from about -2.5 mV to about -10 mV. The aAPC size and surface charge allows for desired circulating and biodistribution properties, and in some embodiments provides advantages in particle stability. The HLA ligands in various embodiments may be HLA Class I and / or Class II molecular complexes, or portions thereof comprising an antigen-binding cleft. In some DB1 / 145613893.1 13 Attorney docket: NEX-013PC / 107578-5013 embodiments, the HLA molecular complexes are monomeric or dimeric, and may contain additional heterologous sequences, such as immunoglobulin sequences. Alternative heterologous sequences include dimerizing amino acid sequences such as c-fos or c-jun, or monomeric amino acid sequences such as albumin. HLA-fusions (e.g., HLA- Immunoglobulin fusions) in some embodiments provide additional advantages in stability, TCR binding affinity, and / or potency for T cell activation or inhibition. In various embodiments, the aAPCs comprise HLA class I ligands for presentation of peptide antigens to CD8+ T cells (e.g., for activation and / or expansion of CD8+ cells, or inhibition of CD8+ cells). In some embodiments, the HLA class I ligand comprises at least two fusion proteins. A first fusion protein comprises a first HLA class I α chain and a first immunoglobulin heavy chain, and a second fusion protein comprises a second HLA class I α chain and a second immunoglobulin heavy chain. The first and second immunoglobulin heavy chains associate to form the HLA class I molecular complex (e.g., associate through disulfide bonds). The HLA class I molecular complex comprises a first HLA class I peptide binding cleft and a second HLA class I peptide binding cleft. Alternatively, the aAPC comprises HLA class II ligands for presentation of peptide antigens to CD4+ T cells (e.g., for activation and / or expansion of CD4+ cells, or inhibition of CD4+ cells). In some embodiments, the HLA class II molecular complex comprises at least four fusion proteins. Two first fusion proteins comprise (i) an immunoglobulin heavy chain and (ii) an extracellular domain of an HLA class II β chain. Two second fusion proteins comprise (i) an immunoglobulin light chain and (ii) an extracellular domain of an HLA class II α chain. The two first and the two second fusion proteins associate to form the HLA class II molecular complex. The extracellular domain of the HLA class II β chain of each first fusion protein and the extracellular domain of the HLA class II α chain of each second fusion protein form an HLA class II peptide binding cleft. In various embodiments, the immunoglobulin sequence of the HLA polypeptide ligand (i.e., HLA-Ig) is a partial heavy chain sequence comprising the hinge region to support dimerization. In some embodiments, the HLA-Ig fusion construct contains no variable region sequences. For example, the HLA extracellular domain sequences (i.e., HLA DB1 / 145613893.1 14 Attorney docket: NEX-013PC / 107578-5013 class I alpha chain extracellular domain) can be fused to an Ig constant region sequence above the hinge region to provide a dimeric HLA. For example, an HLA or antigen presenting portion thereof may be conjugated to a CH1 portion of each IgG heavy chain. All IgG molecules consist of two identical heavy chains (constant and variable regions) joined together by disulfide bonds in the hinge region (upper and lower). For example, in some embodiments, an HLA molecule or antigen presenting complex is fused to the CH1 (N- terminal end of the Ig heavy chain above the hinge region), thereby creating a dimeric fusion protein that is smaller than one created by fusion to the ends of full antibody heavy chains, due to lack of any VH and VL light chain sequences. Thus, such constructs would further include CH2 and CH3 domains. Such a construct provides manufacturing advantages, as well as exhibits less potential for immunogenicity. In some embodiments, these constructs also display sufficient binding cooperativity for efficient T cell activation or inhibition. In some embodiments, the IgG sequence is an IgG4 sequence, and the immunoglobulin sequence may comprise or consist of (or consist essentially of) the amino acid sequence of SEQ ID NO: 23. In some embodiments, the HLA ligands are HLA class I ligands, and are optionally HLA-A, HLA-B, HLA-C, or HLA-E ligands. In embodiments, the HLA ligand comprises associated beta 2 microglobulin (β2M) polypeptide. In various embodiments, the HLA ligand (e.g., as presented by an HLA-Ig) corresponds to an allele selected from HLA- A*02:01, HLA-A*01:01, HLA-A*02:05, HLA-A*02:06, HLA-A*02:12, HLA-A*03:01, HLA-A*11:01, HLA-A*24:02, and HLA-B*07:02. In some embodiments, the HLA ligand is modified with cysteines that form a disulfide bond across the alpha helices that make up the peptide binding groove, to increase complex stability with bound peptide. Such Cysteines in some embodiments are substituted at positions 84 and 139 of the extracellular domain. These modifications bridge the F-pocket, where the C-terminus of the peptide binds. In some embodiments, the HLA ligands are HLA-A*02:01 ligands (IMGT Accession No. HLA00005). In some embodiments, the HLA ligand is modified with cysteines that form a disulfide bond across the alpha helices that make up the peptide binding groove, to increase complex stability with bound peptide. For example, the HLA-A*02:01 ligand may be modified with cysteines at positions 84 and 139 as shown in SEQ ID NO: DB1 / 145613893.1 15 Attorney docket: NEX-013PC / 107578-5013 26, to form the disulfide bond across the alpha helixes that make up the peptide binding groove. Accordingly, in some embodiments, the HLA-Ig ligand comprises the sequence of SEQ ID NO: 26, optionally having from one to five amino acid modifications selected from substitutions, deletions, and insertions, with the proviso that the positions of 84 and 139 with respect to SEQ ID NO: 26 are Cysteine. In other exemplary embodiments, the HLA class I (e.g., HLA-A*02) have W51C and G175C substitutions (e.g., with respect to SEQ ID NO: 26) forming a disulfide bond at end of the peptide groove where the N-terminus of the peptide binds. In other exemplary embodiments, the HLA class I (e.g., HLA-A*02) have F22C and S71C substitutions (e.g., with respect to SEQ ID NO: 26) forming a disulfide bond within one of the alpha helices that borders the peptide binging groove. In various embodiments, the recombinant HLA class I ligand is associated with a β2 microglobulin protein. The amino acid sequence of β2 microglobulin is provided herein as SEQ ID NO: 8. In various embodiments, derivatives of β2 microglobulin may be employed, for example, having from 1 to 10 or from 1 to 5 amino acid modifications independently selected from substitutions, deletions, and insertions. In some embodiments, the HLA ligands are HLA class II ligands. In various embodiments, the HLA-ligands are HLA-DR, HLA-DP, or HLA-DQ. In some embodiments, the HLA class II ligands comprise immunoglobulin fusions of HLA alpha and beta chains to antibody heavy and light chains as already described, thereby creating a dimeric HLA-class II antigen presenting complex. In some embodiments, the aAPCs contain HLA-E ligands, which are optionally HLA-E-Ig (e.g., as described). In some embodiments, the HLA-E ligands are engineered to reduce or eliminate interaction with NKG2A / CD94. HLA-E is a non-classical MHC Class I molecule. HLA-E is represented by only two principal alleles. Given this low polymorphism, HLA-E ligands may be adaptable to create a nearly universal aAPC platform. However, HLA-E has a dual role in both the innate and adaptive immune systems. The role of HLA-E in the innate immune response is to present peptides of other HLA class I molecules to inhibit Natural Killer (NK) cell-mediated lysis via recognition by NKG2A / CD94. NK cells sense DB1 / 145613893.1 16 Attorney docket: NEX-013PC / 107578-5013 the presence of HLA-E presenting self-peptides, and thereby receive inhibitory signals through the NKG2A / CD94 complex (inhibiting NK-mediated lysis). HLA-E can also bind and present peptide sequences for recognition by T-cells (e.g., CD8+ T cells) (the adaptive immune response). Notably, the HLA-E molecule binds NKG2A / CD94 through a binding surface that overlaps with the binding surface for interacting with the T-cell Receptor (“TCR”). In various embodiments, the NK cell deactivating function of HLA-E is decoupled from the T cell activating function, by engineering point mutations that affect only the HLA- E binding to NKG2A / CD94, but not the binding of HLA-E to the TCR. These point mutations enable the redirection of HLA-E for modulating HLA-E-restricted T cells, while avoiding HLA-E exhaustion of Natural Killer cells. In accordance with other aspects and embodiments, HLA-E amino acid substitutions are implemented to provide a stable peptide- binding cleft for presentation of bound antigen to HLA-E-restricted T cells. In some embodiments, a recombinant HLA-E ligand comprises an amino acid sequence that has at least 90% sequence identity to the amino acid sequence of SEQ ID NO: 21 (an HLA-E extracellular domain), and having a substitution to Cysteine at the amino acids corresponding to Y84 and A139 of SEQ ID NO: 21. These substitutions allow for the formation of a disulfide bond that stabilizes the peptide binding cleft. In some embodiments, the amino acid sequence has at least 95% sequence identity, or at least 97% sequence identity, or at least 98% sequence identity, or at least 99% sequence identity to the amino acid sequence of SEQ ID NO: 21. In some embodiments, the amino acid sequence has one or more amino acid modifications with respect to SEQ ID NO: 21 that reduce or eliminate interaction with NKG2A / CD94. In some embodiments, the amino acid modifications are selected from a substitution of D162 and a substitution of E166 with respect to SEQ ID NO: 21. In various embodiments, the substitutions do not include acidic side chains. For example, in some embodiments, the substitution at D162 is selected from D162A, D162G, D162L, D162V, D162I, D162S, D162T, D162M, D162N, and D162Q; and the substitution at E166 is selected from E166A, E166G, E166L, E166V, E166I, E166S, E166T, E166M, E166N, and DB1 / 145613893.1 17 Attorney docket: NEX-013PC / 107578-5013 E166Q. In some embodiments, the recombinant HLA-E ligand comprises the substitutions D162A and E166A with respect to SEQ ID NO: 21. In some embodiments, the HLA-E ligand comprises the substitution to Cys at Y84 and A139, as well as the substitution of D162 and E166. In some embodiments, engineered HLA-E polypeptide ligands are coupled to nanoparticles with co-stimulatory ligands to activate antigen-specific HLA-E restricted T cells against one or more tumor or infectious disease antigens. In other embodiments, engineered HLA-E polypeptide ligands are coupled to nanoparticles with co-inhibitory ligands to inhibit antigen-specific HLA-E restricted T cells against one or more autoantigens. In various embodiments, the immunoglobulin heavy chain sequences fused to the HLA (Class I or II) can be any isotype, and in some embodiments are IgG. In some embodiments, the isotype is selected from IgG1, IgG3, IgG2β, IgG2α, and IgG4. In some embodiments, the immunoglobulin sequences are IgG4 Fc sequences. In some embodiments, the IgG4 Fc domain comprises an amino acid sequence that has at least 90% sequence identity to the amino acid sequence of SEQ ID NO: 23. In some embodiments, the IgG4 Fc domain comprises an amino acid sequence that has at least 95%, or at least 97%, or at least 98%, or at least 99% sequence identity to SEQ ID NO: 23. In some embodiments, the recombinant HLA ligand comprises a linker between the HLA amino acid sequence, and the immunoglobulin sequences (e.g., IgG4 Fc domain). In some embodiments, the linker is a flexible linker, such as a linker that is predominately glycine and serine amino acid residues. An exemplary flexible linker comprises the amino acid sequence of SEQ ID NO: 24. Alternatively, linkers can be selected from flexible and rigid peptide linkers. Flexible linkers are predominately or entirely composed of small and / or polar residues such as Gly, Ser, and Thr. An exemplary flexible linker comprises (GlyxSer)n linkers, where x is from 1 to 10 (e.g., from 2 to 6), and n is from 1 to about 10, and in some embodiments, is from 2 to about 6. In exemplary embodiments, x is from 2 to 4, and n is from 2 to 4. Due to their flexibility, these linkers are substantially unstructured. More rigid linkers include polyproline or poly Pro-Ala motifs and α-helical linkers. Generally, linkers of varying rigidity can be predominately composed of amino acids selected from Gly, Ser, DB1 / 145613893.1 18 Attorney docket: NEX-013PC / 107578-5013 Thr, Ala, and Pro. Exemplary linker sequences contain at least 5 amino acids, and may be in the range of 5 to 30 amino acids or in the range of 5 to 20 amino acids. Peptide antigens are bound to an antigen binding cleft of the antigen presenting complex. Optionally, an antigenic peptide can be covalently bound to a peptide binding cleft. If desired, a peptide tether can be used to link an antigenic peptide to a peptide binding cleft. For example, crystallographic analyses of multiple class I MHC molecules indicate that the amino terminus of β2M is very close, approximately 20.5 Angstroms away, from the carboxyl terminus of an antigenic peptide resident in the MHC peptide binding cleft. Thus, using a relatively short linker sequence, approximately 13 amino acids in length, one can tether a peptide to the amino terminus of β2M. If the sequence is appropriate, that peptide will bind to the MHC binding groove. Peptide antigens for immune therapy of oncological disease, infectious diseases, and autoimmune diseases are described herein. In some embodiments, the peptide antigen does not induce aggregation of the aAPC. That is, peptide antigens are selected with low aggregation propensity when loaded onto aAPCs. Peptides with low aggregation propensity in some embodiments are selected using an Aggregation Potential Score (APS), which can be assessed as described elsewhere herein. For example, a peptide with low aggregation potential when loaded onto HLA ligands (e.g., HLA-A ligands) and modelled in silico shows a mean APS of peptide residues 1 and 3-5 that is less than about 0.07. In some embodiments, the peptide antigen does not have an exposed Cysteine residue, and / or the peptide has one or more exposed glycine residues and / or charged residues (e.g., amino acids Glu, Asp, Lys, Arg, and His). In certain embodiments, the peptide antigen does not have more than 2, 3, or 4 hydrophobic residues, such as those selected from Phe, Val, Leu, and Ile. Further, by limiting the density of HLA ligands (as already described), aggregation propensity can be limited. In some embodiments, the HLA ligand is HLA-A (e.g., HLA-A*02:01), and positions 1 and 3-5 contain at least one or at least two independently selected from Gly and charged residues (Glu, Asp, Lys, Arg, and His) and do not contain more than one exposed hydrophobic residue (e.g., Phe, Val, Leu, and Ile). DB1 / 145613893.1 19 Attorney docket: NEX-013PC / 107578-5013 In some embodiments, the polypeptide ligands comprise a signal-2 ligand that is a co-stimulatory ligand, e.g., for activation and / or expansion of target T cells. Exemplary co- stimulatory ligands include agonists for any one of CD28, 4-1BB, CD27, OX-40, CD30, ICOS, and LIGHT, among others. Co-stimulatory ligands can induce activation and / or expansion of CTLs or Tregs in various embodiments. In other embodiments, and particularly where aAPCs are intended to inhibit target T cells, the polypeptide ligands do not comprise any signal 2 ligand or the polypeptide ligands comprise co-inhibitory ligands, which induce tolerance or apoptosis of target T cells. In various embodiments, the co-inhibitory ligand is an agonist for Fas, TGF-β, or PD-1. Agonistic ligands can include natural agonistic ligands (or engineered variants thereof, including immunoglobulin fusions as described), or in some embodiments, antibody agonists. In some embodiments, the co-inhibitor ligand is PD-L1 (or immunoglobulin fusion thereof) or FasL (or immunoglobulin fusion thereof). Exemplary PD-L1-Ig and Fc-FasL fusion protein ligands are described herein (SEQ ID NO: 1 and SEQ ID NO: 4, respectively). In certain embodiments the immunoglobulin fusion sequences are IgG4 and variants thereof as described herein, and may comprise the amino acid sequence of SEQ ID NO: 23. Embodiments of these constructs are described in more detail elsewhere herein. Antibody agonists can be full monoclonal antibodies, or portions or fragments comprising antigen-binding sequences, such as Fab, Fab', F(ab')2 or scFv. In various embodiments, the signal 1 and signal 2 ligands can be combined in homodimeric or heterodimeric constructs (e.g., homodimeric or heterodimeric Ig fusion constructs). For example, the HLA ligand can comprise fusion of HLA extracellular domains to an immunoglobulin Fc region, such as IgG4 Fc region, which can be dimerized (e.g., through disulfide bonds) with a signal 2-Immunoglobulin (Ig) fusion (i.e., a heterodimeric Ig fusion construct). In still other embodiments, the HLA ligand comprises a fusion to a signal 2 ligand. For example, an HLA extracellular domain can be fused at its C-terminus to an immunoglobulin Fc region (e.g., IgG4 Fc as already described), and fused at its N- terminus to a signal 2 ligand, to prepare homodimeric ligands with both signals dimerized. In some embodiments, the signal 2 ligand comprises a single chain antibody (e.g., scFv) or agonistic portion of a natural ligand. See SEQ ID NOS: 6 and 7, which comprise an anti- CD28 scFv fused to the N-terminus of HLA-Ig. SEQ ID NO: 6 employs the scFv in VH- Linker-VL orientation, and SEQ ID NO: 7 employs the scFv in VL-Linker-VH orientation. DB1 / 145613893.1 20 Attorney docket: NEX-013PC / 107578-5013 See also SEQ ID NO: 5, demonstrating fusion of PD-L1 activating fragment to the N- terminus of HLA-A-IgG4. In some embodiments, the co-stimulatory ligand is an agonistic antibody against CD28, which is optionally a humanized or human monoclonal antibody or a scFv based thereon. For example, the anti-CD28 antibody may be an IgG isotype (e.g., IgG4), and may be as described in US Patent No.10,632,193, which is hereby incorporated by reference in its entirety. In some embodiments, one, two, three, or more complementarity determining regions (CDRs) are based on mouse 9.3 mAb (Tan et al. J. Exp. Med. 1993177:165). In some embodiments, the antibody has the full set of heavy chain and / or full set of light chain CDRs of 9.3 mAb. For example, in some embodiments the heavy chain variable region contains one, two or three of the following CDRs, which optionally may each be modified by one, two, or three amino acid substitutions: CDR1 (DYGVH, SEQ ID NO: 9), CDR2 (VIWAGGGTNYNSALMS, SEQ ID NO: 10), and CDR3 (DKGYSYYYSMDY, SEQ ID NO: 11). In some embodiments, the light chain contains one, two, or three of the following CDRs, which each may be modified by one, two, or three amino acid substitutions: CDR1 (RASESVEYYVTSLMQ, SEQ ID NO: 12), CDR2 (AASNVES, SEQ ID NO: 13), and CDR3 (QQSRKVPYT, SEQ ID NO: 14). Heavy chain variable and Light Chain amino acid sequences for exemplary humanized anti-CD28 agonistic antibodies are provided herein as SEQ ID NOS: 15 to 20. In some embodiments, the anti-CD28 antibody (or portion thereof) binds to the same or overlapping epitope as 9.3 mAb, or binds the same or overlapping epitope as an antibody having CDR1, CDR2, and CDR3 of 9.3 mAb. Antibodies with the same or overlapping epitope can be selected by any suitable technique, including competitive immunoassays, using, for example, Surface Plasmon Resonance (Biacore). Alternative CDR sequences, variable regions, or CD28-binding ligands may be employed in various embodiments. Alternative ligands, CD28 epitopes, and anti-CD28 antibodies are described in U.S. Pat. Nos.7,612,170, 6,987,171, and 6,887,466, for example, and these disclosures are hereby incorporated by reference in their entireties. DB1 / 145613893.1 21 Attorney docket: NEX-013PC / 107578-5013 In some embodiments, the co-inhibitory ligand is an agonist antibody for Fas. In some embodiments, the agonist antibody for Fas is an IgG4 antibody based on clone CH11. SEQ ID NOS: 2 and 3 exemplify heavy chain variable region and light chain sequences for an IgG4 anti-Fas agonistic antibody. As demonstrated herein, this anti-Fas antibody has activity when it is able to crosslink multiple Fas receptors, such as when the antibody is conjugated to a nanoparticle. In some embodiments, the heavy and light chain sequences of SEQ ID NOS: 2 and 3 are humanized and comprise the CDRs of SEQ ID NOS: 2 and 3, shown below in Table 1: Table 1: Anti-Fas IgG4 Variable Light (SEQ ID NO: 3) Variable Heavy (SEQ ID NO: 2) DVVMTQSPLSLPVSLGDQASISC EVQLQQSGPELVKPGASVKISCKASGYT Y ant region and the constant region may be any isotype. In some embodiments, the antibody constant region is human IgG4 or variant thereof. In some embodiments, the constant region comprises one or more hinge stabilizing mutations, which may be introduced in the CH chain (e.g., S241P). In some embodiments, the antibody ligand comprises a constant region and the constant region comprises one or more mutations suitable for chemically coupling the antibody to a solid support. The one or more mutations suitable for coupling create an unpaired cysteine. An exemplary mutation in an IgG4 constant region is S473C. Other changes to the constant region include those modifications to reduce Fc gamma receptor binding. For example, the CH chain may be modified at L248, e.g., L248E. DB1 / 145613893.1 22 Attorney docket: NEX-013PC / 107578-5013 In some embodiments, polypeptide ligands based on antibodies may be minimized such that the ligand is more suitable for functional attachment to nanoparticles. For example, the antibody may be an antibody fragment, such as F(ab')2or Fab, or is a single chain antibody (scFv), or other antigen-binding antibody fragment. For example, the antibody fragment can be a scFv of the humanized mAb described herein or other agonistic anti-CD28 antibody. In some embodiments, the antibody (signal 2 ligand) is a scFv comprising or consisting essentially of the antigen binding loops formed by the VH and VL chains of the monoclonal antibody (e.g., anti-CD28). scFv antibody constructs may comprise one or several (2, 3, 4, or 5) VH and VL hypervariable region chains (the portion of each chain that together form the 3-D antigenic epitope binding pockets) linked together in head-head or head-tail configurations by short peptide linkers. In some embodiments, these constructs are fused to an HLA-Ig sequence as described, to create homodimeric constructs. In some embodiments, other ligand-binding formats are used to produce the co- stimulatory or inhibitory ligand, including peptides, aptamers, and AdNectins. The various formats for target binding include a single-domain antibody, a recombinant heavy-chain- only antibody (VHH), a single-chain antibody (scFv), a shark heavy-chain-only antibody (VNAR), a microprotein (cysteine knot protein, knottin), a DARPin, a Tetranectin, an Affibody; a Transbody, an Anticalin, an Affilin, a Microbody, a peptide aptamer, a phylomer, a stradobody, a maxibody, an evibody, a fynomer, an armadillo repeat protein, a Kunitz domain, an avimer, an atrimer, a probody, an immunobody, a triomab, a troybody, a pepbody, a UniBody, a DuoBody, a Fv, a Fab, a Fab', a F(ab')2, a peptide mimetic molecule, or a synthetic molecule, or as described in US Patent Nos. or Patent Publication Nos. U.S. Pat. No.7,417,130, US 2004 / 132094, U.S. Pat. No.5,831,012, US 2004 / 023334, U.S. Pat. Nos. 7,250,297, 6,818,418, US 2004 / 209243, U.S. Pat. Nos. 7,838,629, 7,186,524, 6,004,746, 5,475,096, US 2004 / 146938, US 2004 / 157209, U.S. Pat. Nos. 6,994,982, 6,794,144, US 2010 / 239633, U.S. Pat. No. 7,803,907, US 2010 / 119446, and / or U.S. Pat. No.7,166,697, the contents of which are hereby incorporated by reference in their entireties. See also, Storz MAbs.2011 May-June; 3(3): 310-317. DB1 / 145613893.1 23 Attorney docket: NEX-013PC / 107578-5013 In some embodiments, the aAPC further comprises one or more cytokines that support T cell activation and / or expansion or T cell inhibition. The one or more cytokines or functional portion thereof may be conjugated to the aAPC as a polypeptide ligand. Alternatively, the cytokine or functional portion thereof may be fused to a signal 1 or signal 2 polypeptide ligand (which can optionally be presented in homodimeric or heterodimeric Ig fusion constructs as described herein). In some embodiments, the cytokine is encapsulated by the copolymers, and will release cytokine locally in targeted environments (e.g., in lymphoid organs, tumor, or target tissue or organ). Examples of cytokines that may be used include IL-1β, IL-2, IL-4, IL-7, IL-10, IL-12, IL-15, and gamma interferon. For example, IL-2 can be employed with a co-stimulatory signal 2 ligand. In some embodiments, the aAPC comprises a tolerogenic cytokine as a polypeptide ligand. An exemplary tolerogenic cytokine is IL-10. In some embodiments, the ligands further comprise one or more homing ligands for lymphoid organs. For example, an exemplary homing ligand is CD62L. In some embodiments, ligands (which may or may not be polypeptide ligands) are included to target the aAPCs to a tissue or organ or interest, such as the pancreas, intestine, lungs, liver, muscle, skin, etc. Suitable peptide ligands or other ligands can be selected based on information in the art. In various embodiments, the one or more peptide antigens are tumor or cancer associated antigens, such as tumor-derived antigens, tumor-specific antigens, and neoantigens. T cells specific for tumor associated antigens are often very rare, and in many cases undetectable, in the peripheral blood of healthy individuals. Further, the cells are often of a naive phenotype. See, Quintarelli et al., Cytotoxic T lymphocytes directed to the preferentially expressed antigens of melanoma (PRAME) target chronic myeloid leukemia. Blood 2008; 112: 1876-1885. This is often a distinction observed between viral-specific and tumor antigen specific T cells. In accordance with these embodiments, the injectable aAPC of this disclosure can activate and / or expand such T cells in vivo, to produce an anti-tumor immune response. DB1 / 145613893.1 24 Attorney docket: NEX-013PC / 107578-5013 “Tumor-associated antigens” or “cancer specific antigens” include unique tumor or cancer antigens expressed exclusively by the tumor or malignant cells from which they are derived, shared tumor antigens expressed in many tumors but not in normal adult tissues (oncofetal antigens), and tissue-specific antigens expressed also by the normal tissue from which the tumor arose. Tumor associated antigens can be, for example, embryonic antigens, antigens with abnormal post-translational modifications, differentiation antigens, products of mutated oncogenes or tumor suppressors, fusion proteins, or oncoviral proteins. A variety of tumor-associated antigens are known in the art. Oncofetal and embryonic antigens include carcinoembryonic antigen and alpha-fetoprotein (usually only highly expressed in developing embryos but frequently highly expressed by tumors of the liver and colon, respectively), MAGE-1 and MAGE-3 (expressed in melanoma, breast cancer, and glioma), placental alkaline phosphatase sialyl-Lewis X (expressed in adenocarcinoma), CA-125 and CA-19 (expressed in gastrointestinal, hepatic, and gynecological tumors), TAG-72 (expressed in colorectal tumors), epithelial glycoprotein 2 (expressed in many carcinomas), pancreatic oncofetal antigen, 5T4 (expressed in gastriccarcinoma), alphafetoprotein receptor (expressed in multiple tumor types, particularly mammary tumors), and M2A (expressed in germ cell neoplasia). Mutated oncogene or tumor-suppressor gene products include Ras and p53, both of which are expressed in many tumor types, Her-2 / neu (expressed in breast and gynecological cancers), EGF-R, estrogen receptor, progesterone receptor, retinoblastoma gene product, myc (associated with lung cancer). Fusion proteins include BCR-ABL, which is expressed in chromic myeloid leukemia. Oncoviral proteins include HPV type 16, E6, and E7, which are found in cervical carcinoma. Tissue-specific antigens include melanotransferrin and MUC1 (expressed in pancreatic and breast cancers); CD10 (previously known as common acute lymphoblastic leukemia antigen, or CALLA) or surface immunoglobulin (expressed in B cell leukemias and lymphomas); the α chain of the IL-2 receptor, T cell receptor, CD45R, CD4+ / CD8+ (expressed in T cell leukemias and lymphomas); prostate specific antigen and prostatic acid- phosphatase (expressed in prostate carcinoma); GP100, MelanA / Mart-1, tyrosinase, DB1 / 145613893.1 25 Attorney docket: NEX-013PC / 107578-5013 gp75 / brown, BAGE, and S-100 (expressed in melanoma); cytokeratins (expressed in various carcinomas); and CD19, CD20, and CD37 (expressed in lymphoma). Tumor-associated antigens also include altered glycolipid and glycoprotein antigens, such as neuraminic acid-containing glycosphingolipids (e.g., GM2 and GD2, expressed in melanomas and some brain tumors); blood group antigens, particularly T and sialylated Tn antigens, which can be aberrantly expressed in carcinomas; and mucins, such as CA-125 and CA-19-9 (expressed on ovarian carcinomas) or the underglycosylated MUC-1 (expressed on breast and pancreatic carcinomas). Tumor-associated antigens are further disclosed in US Patent 11,007,222, which is hereby incorporated by reference. In some embodiments, the target peptide antigens include at least one that is associated with or derived from a pathogen, such as a viral, bacterial, fungal, or parasitic pathogen. For example, at least one peptide antigen may be associated with tuberculosis (TB), HIV (human immunodeficiency virus), HTLV (human T-lymphotropic virus) type 1, hepatitis (e.g., A, B, C, or D) cytomegalovirus (CMV), Epstein-Barr virus (EBV), HPV, influenza, herpes virus (e.g., HSV 1 or 2, or varicella zoster), and Adenovirus. CMV, for example, is the most common viral pathogen found in organ transplant patients and is a major cause of morbidity and mortality in patients undergoing bone marrow or peripheral blood stem cell transplants. In some embodiments, the one or more target peptide antigens is an “autoantigen”, meaning that it is associated with an autoimmune disease or reaction. In some embodiments, aAPCs carrying tolerogenic ligands induce tolerance in antigen-specific T cells to the target antigen. In some embodiments, aAPCs carrying apoptotic signals (e.g., Fas ligand or agonist anti-Fas antibody), will induce specific apoptosis of antigen-specific T cells. Autoantigens can be involved in autoimmune diseases such as type 1 diabetes, Goodpasture's syndrome, multiple sclerosis, Graves' disease, myasthenia gravis, systemic lupus erythematosus, rheumatoid arthritis, pemphigus vulgaris, Addison's disease, dermatitis herpetiformis, celiac disease, Crohn’s disease, and Hashimoto's thyroiditis, vitiligo, among others. DB1 / 145613893.1 26 Attorney docket: NEX-013PC / 107578-5013 In some embodiments, peptide antigens for presentation by the HLA ligands are determined in a personalized manner, as described in US 10,098,939 and US 2020 / 0291381, which are hereby incorporated by reference in their entireties. For example, sequencing data can provide information about both shared as well as personalized targets for immunotherapy, such as for cancer. In principle, mutant proteins are foreign to the immune system and are putative tumor-specific antigens. Indeed, sequencing efforts have defined hundred if not thousands of potentially relevant immune targets. Studies have shown that T cell responses against these neo-epitopes can be found in cancer patients or induced by cancer vaccines. Mutation catalogues derived from whole exome sequencing provide a starting point for identifying such neo-epitopes. Using HLA binding prediction algorithms (Srivastava, PLoS One 4, e6094 (2009), it has been predicted that each cancer can have up 7-10 neo-epitopes. A similar approach estimated hundreds of tumor neo-epitopes. Neoepitopes predicted from DNA or RNA sequencing of a patient’s tumor, can be tested for their activation potential in association with the HLA ligand, by determining whether (or to what extent) an aAPC carrying the predicted antigen in association with the HLA ligand is able to activate T cells from the subject. Similar assays can be employed to identity relevant autoantigens. In some embodiments, for treatment of type 1 diabetes, peptide antigens can be antigens from ICA, insulin, G6, GAD2, GAD65, insulinoma antigen-2, HSP, IGRP, imogen- 38, PDX1, ZnT8, CHGA, and IAAP. See, e.g., Han, S, et al. Novel autoantigens in type 1 diabetes. Am J Transl Res.2013; 5(4): 379–392. Non-limiting examples of peptide antigens include those shown in the following Table 2. Table 2: Peptide Antigens PeptideName DB1 / 145613893.1 27 Attorney docket: NEX-013PC / 107578-5013 SLFVLGLFL CS1.239247.M0004 (SEQ ID NO: 44) 3 In vario , utical composition suitable for administration to a subject. The pharmaceutical composition may have one or more excipients such as buffering agents, surfactants, preservative agents, polymers, bulking agents, and stabilizers. Buffering agents are used to control the pH of the composition. Surfactants are used to stabilize proteins, inhibit protein aggregation, inhibit protein adsorption to surfaces, and assist in protein refolding. Exemplary surfactants include Tween 80, Tween 20, Brij 35, Triton X- 10, Pluronic F127, and sodium dodecyl sulfate. Preservatives are used to prevent microbial growth. Examples of preservatives include benzyl alcohol, m-cresol, and phenol. Bulking agents are used during lyophilization to add DB1 / 145613893.1 28 Attorney docket: NEX-013PC / 107578-5013 bulk. Hydrophilic polymers such as dextran, hydroxyl ethyl starch, polyethylene glycols, and gelatin can be used to stabilize proteins. Polymers with nonpolar moieties such as polyethylene glycol can also be used as surfactants. Protein stabilizers can include polyols, sugars, amino acids, amines, and salts. Suitable sugars include sucrose and trehalose. Amino acid stabilizers include histidine, arginine, glycine, methionine, proline, lysine, glutamic acid, and mixtures thereof. Proteins like human serum albumin can also competitively adsorb to surfaces and reduce aggregation of the active agent. Particular formulation components can serve multiple purposes. For example, histidine can act as a buffering agent and an antioxidant. Glycine can be used as a buffering agent and as a bulking agent. In some embodiments, the pharmaceutical composition is lyophilized. In some aspects and embodiments, the present disclosure provides aAPCs suitable for parenteral administration (including subcutaneous administration), and which have low aggregation propensity. In these aspects, the aAPC comprises a polymeric or lipid nanoparticle comprising a polyethylene glycol (PEG) sheath and one or more polypeptide ligands conjugated to PEG (e.g., the PEG terminus), which can be through a thioether bond or other functional group. The polypeptide ligands comprise HLA Class I or Class II ligands presenting a peptide antigen and optionally one or more signal 2 ligands (as already described). The peptide antigen for presentation to T cells does not induce aggregation of the aAPC. As disclosed herein, certain peptides when loaded onto aAPCs will induce aggregation due to the character of the exposed surface. Analyzing the peptide sequence alone does not provide a clear indication for aggregation potential in the context of the HLA. The aggregation potential occurs at the level of the peptide-loaded nanoparticle as opposed to a protein complex alone in solution. In some embodiments, the peptide antigen does not have an exposed Cysteine and / or the peptide antigen has one or more exposed glycine residues or exposed charged residues. In some embodiments, the peptide antigen does not have any Cysteine residue, and contains one or more charged residues (e.g., 1, 2, or 3 charged residues). In various embodiments, the peptide antigen in the HLA binding cleft has one or more exposed amino acids selected from glycine, aspartic acid, glutamic acid, lysine, arginine, and histidine; and does not have an exposed cysteine. In some embodiments, positions 1 and 3 to 5 of the peptide satisfy these DB1 / 145613893.1 29 Attorney docket: NEX-013PC / 107578-5013 criteria. In this context, the term “exposed” means that the side chain of the amino acid is surface exposed. In some embodiments, the HLA is HLA-A, and in some embodiments is HLA-A2, such as HLA-A*02:01. In some aspects and embodiments, aggregation potential is estimated in silico by determining the mean Aggregation Potential Score (APS) of antigen peptide residues within the HLA antigen binding cleft. In some embodiments, for example with respect to HLA-A (e.g., HLA-A2, such as HLA-A*02:01), peptide positions 1 and 3-5 are evaluated for APS and an average score determined for these positions. For example, a score of less than about 0.07 is indicative of a non-aggregating peptide. Protein modeling and aggregation propensity analyses can be performed using Discovery Studio 2021 (DS2021, Dassault Systèmes BIOVIA, Discovery Studio Modeling Environment, Release 2021, San Diego CA). For example, the MODELLER program, or similar program, can be used for homology or comparative modeling of protein three-dimensional structures (e.g., HLA with bound peptide). Spatial aggregation propensity (SAP) can be calculated according to available tools, and which are included in the DS2021 package and otherwise commercially available. The spatial aggregation propensity (SAP) score synthesizes the solvent accessible area of a residue side chain with a hydrophobicity score based on a scale developed by Black and Mould (1991) where glycine has a value of zero and other residues are scaled positively if more hydrophobic and negatively if less hydrophobic. This score, coupled with specific peptide features already described identifies peptides likely to drive aggregation of proteins and by extension HLA-decorated nanoparticles when complexed with the HLA ligands. The use of aggregation propensity on in silico models (as well as with experimentally determined structures) is a known methodology. However, these tools are conventionally focused on antibody screening to find antibodies with good drug-like properties or to engineer out biophysical liabilities. While comparative modeling and APS tools are in common use, their application to peptide-HLA complexes, especially in the context of nanoparticle coating and aggregation of these particles, has not been described or suggested. Further, the significant impact of the peptide in the HLA binding pocket on particle DB1 / 145613893.1 30 Attorney docket: NEX-013PC / 107578-5013 aggregation was surprising, given that the peptide makes up a relatively small portion of the structure and the overall exposed surface of the aAPC. In some embodiments, the process is initiated by generating molecular models of candidate relevant peptide-HLA complexes. The protein model includes the relevant peptides positioned in the HLA binding groove. The MODELLER algorithm as implemented in Discovery Studio 2021 can be used, or other suitable software. Multiple template structures can be used for the modeling steps, and manual sequence alignments are performed on each model. PDB accessions for template structures used in HLA-A*02:01 modeling include: 4L29, 1I7R, 5EUO, 1TVB, 6OPD, 6TRO, 6AMT, 6AM5, 2GT9, and 3OXS. For example, using the static models, Aggregation Scores can be calculated. In some embodiments, radii of 5 or 10 Angstroms are selected for calculating scores. This parameter determines how many nearby amino acid residues to include in the aggregation potential calculation. This is similar to a sliding window for protein sequence parameter calculations. Depending on the size of the model being analyzed, the radius can give an appropriately smoothed surface map that is not overly noisy. The protocol calculates exposed surfaces for each residue and calculates an aggregation potential score based on the averaging of each residue and those in the defined radius surrounding the residue. The numerical score (APS) is recorded with the sequence and structure information and is used to generate a surface map based on the solvent-accessible surface of the model. This provides an easy visualization of the shape of the surface coupled with the scored aggregation potential. In some instances, manual inspection of these results is sufficient to bin a peptide as an aggregator or non-aggregator. In accordance with some embodiments, the APS of residues 1 and 3-5 of HLA-A peptide complex are averaged. If this score is above a current threshold of 0.07, this is considered a factor for aggregation. The presence of an exposed Cys residue (as determined by inspection of the model structures) is considered another factor for aggregation. Presence of charged residues and Gly residues are considered factors against aggregation. The total DB1 / 145613893.1 31 Attorney docket: NEX-013PC / 107578-5013 number of factors in favor or against aggregation can be used to determine whether the peptide-HLA complex is considered a potential aggregator or not. In some embodiments, the nanoparticles are polymeric nanoparticles comprising poly(lactic acid)-polyethylene glycol (PLA-PEG) or poly(lactic acid-co-glycolic acid)- polyethylene glycol (PLGA-PEG) copolymers and one or more polypeptide ligands conjugated to PEG through a thioether bond. Such polymeric nanoparticles can be as already described. Alternative polymers that can be used in connection with the aAPC platforms described herein include one or more of cyclodextrin-containing polymers, cationic cyclodextrin-containing polymers, poly(D,L-lactic acid-co-glycolic acid) (PLGA), poly(caprolactone) (PCL), ethylene vinyl acetate polymer (EVA), poly(lactic acid) (PLA), poly(L-lactic acid) (PLLA), poly(glycolic acid) (PGA), poly(L-lactic acid-co-glycolic acid) (PLLGA), poly(D,L-lactide) (PDLA), poly(L-Lactide) (PLLA), PLGA-b-poly(ethylene glycol)-PLGA (PLGA-bPEG-PLGA), PLLA-bPEG-PLLA, PLGA-PEG-maieimide (PLGA-PEG-mal), PLA-PEG-maleimide, poly(D,L-lactide-co-caprolactone), poly(D,L- Lactide-co-caprolactone-co-glycolide), poly(D,L-lactide-co-PEO-co-D,L-lactide), poly(D,L-lactide-co-PPO.about.co-D,L-lactide), polyalkylcyanoacralate, polyurethane, poly-L-lysine (PLL), hydroxypropyl methacrylate (HPMA), polyethyleneglycol, poly-L- glutamic acid, poly(hydroxy acids), polyanhydrides, polyorthoesters, poly(ester amides), polyamides, poly(ester ethers), polycarbonates, polyalkylenes such as polyethylene and polypropylene, polyalkylene glycols such as poly(ethylene glycol) (PEG), polyalkylene oxides (PEO), polyalkylene terephthalates such as poly(ethylene terephthalate), polyvinyl alcohols (PVA), polyvinyl ethers, polyvinyl esters such as poly(vinyl acetate), polyvinyl haiides such as poly(vinyl chloride) (PVC), polyvinylpyrrolidone, polysiloxanes, polystyrene (PS), polyurethanes, derivatized celluloses such as alkyl celluloses, hydroxy alkyl celluloses, cellulose ethers, cellulose esters, nitro celluloses, hydroxypropylcellulose, carboxymethylcellulose, polymers of acrylic acids, such as polymethylmethacrylate) (P MA), poly(ethyl(meth)acrylate), poly(butyl(meth)acrylate), poly (isobutyl (meth)acrylate), poly(hexyl(meth)acrylate), poly(isodecyl(meth)acrylate), poly(lauryl(meth)acrylate), poly(phenyl(meth)acrylate), poly(methyl acrylate), poly(isopropyl acrylate), polyiisobutyl DB1 / 145613893.1 32 Attorney docket: NEX-013PC / 107578-5013 acrylate), poly(octadecyl acrylate) (poly acrylic acids), and copolymers and mixtures thereof, polydioxanone and its copolymers, polyhydroxyalkanoates, polypropylene fumarate), polyoxymethylene, poloxamers, poly(ortho)esters, poly(butyric acid), poly(valeric acid), poly(lactide-co-caprolactone), trimethylene carbonate, polyvinylpyrrolidone, polyorthoesters, polyphosphazenes, and polyphosphoesters, dendrimers and derivatives thereof, and blends and / or block copolymers of two or more such polymers. In other embodiments, the nanoparticles are lipid nanoparticles comprising a PEG- conjugated lipid. Exemplary PEG lipids are selected from one or more of a PEG-modified phosphatidylethanolamine, a PEG- modified phosphatidic acid, a PEG-modified ceramide, a PEG-modified dialkylamine, a PEG- modified diacylglycerol, and a PEG-modified dialkylglycerol. A PEG lipid may be selected from PEG-c-DOMG, PEG-DMG, PEG-DLPE, PEG-DMPE, PEG-DPPC, PEG-Cholesterol, PEG tocopherol, or a PEG- DSPE lipid. In some embodiments, the lipid nanoparticles further comprise a cationic or ionizable lipid, a neutral lipid or phospholipid, and a structural lipid. Exemplary structural lipids can be selected from one or more of cholesterol, fecosterol, sitosterol, ergosterol, campesterol, stigmasterol, brassicasterol, tomatidine, ursolic acid, and tocopherols (e.g., alpha tocopherol). In some embodiments, the structural lipid is cholesterol. In some embodiments, the LNP comprises one or more phospholipids. Exemplary phospholipids are selected from the group consisting of cardiolipins, sterol modified lipids (modified with a cholesterol moiety attached at the sn-2 carbon of the glycerol backbone), mixed-acyl glycerophospholipids, and symmetrical acyl glycerophospholipids. Head groups for acyl glycerophospholipids include, for example, phosphatidic acid, lysophosphatidic acid, phosphatidylcholine, phosphatidylethanolamine, phosphatidylglycerol, phosphoinositides, and phosphatidylserine. Exemplary phospholipids are selected from 1,2-dilinoleoyl-sn- glycero-3-phosphocholine (DLPC), 1,2-dimyristoyl-sn-glycero-phosphocholine (DMPC), 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 1,2-dipalmitoyl-sn-glycero-3- phosphocholine (DPPC), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), 1,2- diundecanoyl-sn-glycero-phosphocholine (DUPC), 1-palmitoyl-2-oleoyl-sn-glycero-3- phosphocholine (POPC), 1,2-di-O-octadecenyl-sn-glycero-3-phosphocholine (18:0 Diether DB1 / 145613893.1 33 Attorney docket: NEX-013PC / 107578-5013 PC), 1-oleoyl—2-cholesterylhemisuccinoyl-sn-glycero-3-phosphocholine (OChemsPC), l- hexadecyl-sn-glycero-3-phosphocholine (C16 Lyso PC), 1,2-dilinolenoyl-sn-glycero-3- phosphocholine, 1,2-diarachidonoyl-sn-glycero-3-phosphocholine, l,2-didocosahexaenoyl- sn-glycero-3-phosphocholine, l,2-dioleoyl-sn-glycero-3-phosphoethanol amine (DOPE), l,2-diphytanoyl-sn-glycero-3-phosphoethanolamine (ME 16.0 PE), l,2-distearoyl-sn- glycero-3-phosphoethanolamine, l,2-dilinoleoyl-sn-glycero-3-phosphoethanolamine, l,2- dilinolenoyl-sn-glycero-3-phosphoethanolamine, l,2-diarachidonoyl-sn-glycero-3- phosphoethanolamine, l,2-didocosahexaenoyl-sn-glycero-3-phosphoethanolamine, l,2- dioleoyl-sn-glycero-3-phospho-rac-(l -glycerol) sodium salt (DOPG), and sphingomyelin. In some embodiments, the nanoparticle is a lipid nanoparticle that further comprises a polynucleotide for expression in target T cells. In some embodiments, the polynucleotide (e.g., mRNA) encodes a cytokine, including a cytokine described herein (e.g., IL-2 or IL- 10). In some embodiments, sulfhydryl groups on the polypeptide ligands are coupled to PEG-maleimide functional groups of the lipid nanoparticles. In various embodiments, the lipid nanoparticle aAPC has about 10 to about 500 polypeptide ligands or about 50 to about 400 polypeptide ligands, or about 100 to about 300 polypeptide ligands. Polypeptide ligand density can generally be controlled by reducing the proportion of PEG moieties having a maleimide group for conjugation as already described, or by modifying the level of PEG lipid in the composition. The lipid nanoparticle aAPC has a diameter of from about 50 nm to about 150 nm, or a diameter of from about 60 nm to about 130 nm, or a diameter of from about 80 nm to about 120 nm. In various embodiments, the aAPC has a diameter of about 80 nm, about 90 nm, about 100 nm, about 110 nm, or about 120 nm. The lipid nanoparticle aAPC has a surface charge of from about 0 to -15 mV, or from about 0 to about -10 mV, or a surface charge of from about -2.5 mV to about -10 mV. The aAPC population has a size distribution with polydispersity index (PDI) of less than 0.2. In various embodiments, the aAPC is comprised in a pharmaceutical composition as described and may be optionally lyophilized. In various embodiments, the aAPCs and DB1 / 145613893.1 34 Attorney docket: NEX-013PC / 107578-5013 pharmaceutical compositions described herein are useful for treating patients with infectious diseases, cancer, or autoimmune diseases, or to provide prophylactic protection to immunosuppressed patients. In other aspects, the invention provides a method for immunotherapy. The method comprises administering the aAPC or pharmaceutical composition thereof as described herein to a subject in need of treatment. In accordance with this aspect, the aAPC and compositions thereof described herein are used for immunotherapy. In various embodiments, the subject has cancer or an infectious disease, and the aAPCs comprise a co-stimulatory ligand. Infectious diseases that can be treated include those caused by bacteria, viruses, prions, fungi, parasites, helminths, etc. Such diseases include human papilloma virus (HPV) (and related cancers), AIDS, adult T-cell leukaemia / lymphoma (ATL), hepatitis, CMV infection, and post-transplant lymphoproliferative disorder (PTLD). CMV, for example, is the most common viral pathogen found in organ transplant patients and is a major cause of morbidity and mortality in patients undergoing bone marrow or peripheral blood stem cell transplants. This is due to the immunocompromised status of these patients, which permits reactivation of latent virus in seropositive patients or opportunistic infection in seronegative individuals. Current treatment focuses on the use of antiviral compounds such as gancyclovir, which have drawbacks, the most significant being the development of drug-resistant CMV. PTLD occurs in a significant fraction of transplant patients and results from Epstein- Barr virus (EBV) infection. EBV infection is believed to be present in approximately 90% of the adult population in the United States. Active viral replication and infection is kept in check by the immune system, but, as in cases of CMV, individuals immunocompromised by transplantation therapies lose the controlling T cell populations, which permits viral reactivation. This represents a serious impediment to transplant protocols. EBV may also be involved in tumor promotion in a variety of hematological and non-hematological cancers. There is also a strong association between EBV and nasopharyngeal carcinomas. Thus a prophylactic treatment with EBV-specific T cells offers an excellent alternative to current therapies. DB1 / 145613893.1 35 Attorney docket: NEX-013PC / 107578-5013 Cancers that can be treated according to this disclosure include melanoma, carcinomas, e.g., colon, head and neck cancer, duodenal, prostate, breast, lung, ovarian, ductal, hepatic, pancreatic, renal, endometrial, stomach, dysplastic oral mucosa, polyposis, invasive oral cancer, non-small cell lung carcinoma, transitional and squamous cell urinary carcinoma etc.; neurological malignancies, e.g., neuroblastoma, gliomas, etc.; hematological malignancies, e.g., chronic myelogenous leukemia, childhood acute leukemia, non- Hodgkin's lymphomas, chronic lymphocytic leukemia, malignant cutaneous T-cells, mycosis fungoides, non-MF cutaneous T-cell lymphoma, lymphomatoid papulosis, T-cell rich cutaneous lymphoid hyperplasia, bullous pemphigoid, discoid lupus erythematosus, lichen planus, etc.; and the like. See, e.g., Mackensen et al, Int. J. Cancer 86, 385-92, 2000; Jonuleit et al., Int. J. Cancer 93, 243-51, 2001; Lan et al., J. Immunotherapy 24, 66-78, 2001; Meidenbauer et al, J. Immunol. 170(4), 2161-69, 2003. In some embodiments, the subject has a solid tumor, which can be Stage I, Stage II, Stage III, or Stage IV cancer. In some embodiments, the cancer is metastatic and / or recurrent, and / or is nonresectable. In some embodiments, the patient is refractory or only partially responsive to chemotherapy and / or immune checkpoint inhibitor therapy. In some embodiments, the invention provides a method for treating cancer, including those cancers identified above, through administration of the pharmaceutical composition described herein to activate T-cells having anti-tumor activity. In some embodiments, the therapy is provided together with one or more immune checkpoint inhibitors, such as Nivolumab, Pembrolizumab, and Ipilimumab. In some embodiments, the additional therapy is anti-CTLA4 or anti-PD1, or anti-PD-L1. The additional therapy or checkpoint inhibitor may be administered separately through its conventional regimen, or may be administered as an additional ligand to the nanoparticles described herein, or attached to a separate population of nanoparticles. In some embodiments, the one or more immune checkpoint inhibitors are provided as initial therapy, and therapy with the aAPCs described herein initiated subsequently, for example, after from about 1 to about 8 weeks of checkpoint inhibitor therapy, or after about 2 to about 4 weeks of checkpoint inhibitor therapy. In some embodiments, the one or more checkpoint inhibitors are provided concomitantly with the nanoparticle therapy, for example at initiation of therapy and about every two weeks, or at initiation of therapy and about every two weeks for the one or more checkpoint inhibitors DB1 / 145613893.1 36 Attorney docket: NEX-013PC / 107578-5013 and about every four weeks for the nanoparticle therapy. In some embodiments, the patient is resistant or shows only a partial or transient response to checkpoint inhibitor therapy, and the aAPCs described herein enhance tumor regression in these patients. In still other embodiments, for cancers that are typically resistant to immune checkpoint inhibitor therapy, the compositions described herein expand the successful use of checkpoint inhibitors to such cancers. In some embodiments, the peptide antigen is selected in a personalized basis for the patient, based on an analysis of the patient's tumor. For example, a process described by Ionov Y., A high throughput method for identifying personalized tumor-associated antigens, Oncotarget 1(2):148-155 (2010) (which is hereby incorporated by reference) may be used, or other process. In these embodiments, the nanoparticles can be provided (on an “off-the shelf” basis), and tumor antigens selected and loaded in a personalized basis. Other processes for selecting peptides based on their ability to activate and expand relevant T cells (including in a personalized manner in some embodiments) are described in US Patents 10,987,412 and 10,098,939, which are hereby incorporated by reference in their entireties. In some embodiments, the nano-aAPCs are used as a booster vaccine, after adoptive T cell therapy, in which naive T cells from the patient or T cells from an HLA-matched donor are expanded ex vivo and administered to the patient. The nano aAPC composition may be administered from 1 to about 10 times over the course of from 4 months to about 1 year to enhance cancer immunity in these embodiments. In some embodiments, the subject has an autoimmune condition, and the aAPCs comprise a co-inhibitory signal or do not contain a signal 2 ligand. In some embodiments, the autoimmune condition is type 1 diabetes. In some embodiments, the autoimmune condition is vitiligo. Autoimmune diseases that can be treated include systemic lupus erythematosus, rheumatoid arthritis, multiple sclerosis, Crohn's disease, ulcerative colitis, psoriasis, myasthenia gravis, Goodpasture's syndrome, Graves' disease, pemphigus vulgaris, Addison's disease, dermatitis herpetiformis, celiac disease, Sjogren’s Disease, Hashimoto's thyroiditis, alopecia, ankylosing spondylitis, scleroderma, and HTLV-1-Associated Myelopathy (HAM) / Tropical Spastic Paraparesis (TSP), among others. DB1 / 145613893.1 37 Attorney docket: NEX-013PC / 107578-5013 Generally, the aAPC or pharmaceutical composition thereof is parenterally administered. For example, the aAPC or pharmaceutical composition thereof is administered by intravenous administration, intra-arterial administration, subcutaneous administration, intradermal administration, intra-lymphatic administration, intramuscular administration, or intratumoral administration. In some embodiments, the aAPC composition is administered by subcutaneous administration. In other aspects of this disclosure, polypeptide ligands for immune therapy (including as polypeptide ligands for aAPCs) are disclosed. Such polypeptide ligands include an anti-Fas agonistic antibody having an IgG isotype (e.g., IgG4), which can be conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen (e.g., associated with an autoimmune disease). In some embodiments, the anti-Fas antibody comprises a heavy chain of SEQ ID NO: 2, optionally having from one to ten, or from one to five, amino acid substitutions. Amino acid substitutions are optionally in the Fc domain, and may include amino acid substitutions with known pharmacological or stability advantages, or advantages in removing potential immunogenicity. In some embodiments, the amino acid sequence of SEQ ID NO: 2 is humanized and comprises the same complementarity determining regions (CDRs) of SEQ ID NO: 2 (see Table 1) with no more than 1, 2, or 3 amino acid substitutions. In some embodiments, the anti-Fas antibody comprises a light chain of SEQ ID NO: 3 (see Table 1), optionally having from one to ten, or from one to five, amino acid substitutions. In some embodiments, the amino acid sequence of SEQ ID NO: 3 is humanized and comprises the same CDRs of SEQ ID NO: 3, with no more than 1, 2, or 3 amino acid substitutions. In other embodiments, the polypeptide ligand is a dimeric PD-L1 ligand comprising an activating portion of PD-L1, such as amino acids residues F19 to T239 of human PD-L1. Alternatively, an additional 20 residues can be included (collectively on one or both termini), or up to 10 amino acids deleted from F19 to T239. Each PD-L1 activating fragment may be fused directly or indirectly through a linker at its C-terminus to an IgG Fc region (e.g., IgG4), and the ligand may be dimerized by disulfide bonds in the Fc region. Suitable linkers are described elsewhere herein, and include flexible linkers such as Gly Ser linkers. The dimeric DB1 / 145613893.1 38 Attorney docket: NEX-013PC / 107578-5013 PD-L1 ligand can be conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen and used to drive tolerance to the antigen. An exemplary PD-L1-IgG4 is shown in SEQ ID NO: 1. In other embodiments, the polypeptide ligand is a dimeric FasL ligand comprising an activating portion of FasL, such as amino acids P132 to L279 of human FasL fused directly or indirectly through a linker at its N-terminus to a dimerized IgG-Fc region (e.g., IgG4). Alternatively, an additional 20 residues can be included (collectively on one or both termini), or up to 10 amino acids deleted from P132 to L279. The dimerized Fc region can be conjugated to nanoparticles through a Cys-containing linker in some embodiments. An exemplary construct according to these embodiments is provided by SEQ ID NO: 4. The nanoparticles may further present HLA-peptide antigen ligands to drive apoptosis of antigen-specific T cells. In other embodiments, the polypeptide ligand is a tolerogenic ligand comprising an activating fragment of PD-L1, such as amino acids F19-T239 of human PD-L1, fused directly or through a linker to HLA-immunoglobulin fusion protein. Alternatively, an additional 20 residues can be included (collectively on one or both termini), or up to 10 amino acids deleted from F19 to T239 of human PD-L1. These embodiments provide a homodimeric construct that is homodimeric for both signal 1 and inhibitory signal 2 ligands. See SEQ ID NO: 5. Such ligands can be conjugated to nanoparticles as disclosed herein and used for immunotherapy (to drive tolerance in targeted T cells). In other embodiments, the polypeptide ligand is a co-stimulatory ligand comprising an anti-CD28 agonistic scFv conjugated to an HLA-immunoglobulin fusion protein, providing homodimeric ligands comprising signal 1 and co-stimulatory signal 2 ligands. See SEQ ID NOS: 6 and 7. The scFv may be fused to the HLA sequence through the heavy or light chain sequence. For example, SEQ ID NO: 6 employs the scFv in VH-Linker-VL orientation, and SEQ ID NO: 7 employs the scFv in VL-Linker-VH orientation. The singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. DB1 / 145613893.1 39 Attorney docket: NEX-013PC / 107578-5013 As used in the specification and in the claims, the open-ended transitional phrases “comprise(s),” “include(s),” “having,” “contain(s),” and variants thereof require the presence of the named features / steps and permit the presence of other features / steps. These phrases should also be construed as disclosing the closed-ended phrases “consist of” or “consist essentially of” that permit only the named features / steps and unavoidable impurities, and exclude other features / steps. As used herein, the term “about” means ±10% of a numerical value, unless the context requires otherwise. The term “identity” refers to the similarity between a pair of sequences (nucleotide or amino acid). Identity is measured by dividing the number of identical residues by the total number of residues and multiplying the product by 100 to achieve a percentage. Thus, two copies of exactly the same sequence have 100% identity, but sequences that are less highly conserved and have deletions, additions, or replacements may have a lower degree of identity. Those skilled in the art will recognize that several computer programs, such as those that employ algorithms such as BLAST, are available for determining sequence identity. BLAST nucleotide searches are performed with the NBLAST program, and BLAST protein searches are performed with the BLASTP program, using the default parameters of the respective programs. The term CDR refers to a complementarity-determining region. CDRs are part of the variable chains in immunoglobulins (antibodies). A set of CDRs constitutes a paratope. This invention is further illustrated by the following non-limiting examples. EXAMPLES Example 1: Construction of Injectable Artificial Antigen Presenting Cells Ligands for aAPCs, including signal 1 ligands (HLA) and signal 2 ligands are constructed with free sulfhydryl groups for coupling to particles. See US 10,632,193, which is hereby incorporated by reference in its entirety. Naked polymer nanoparticles were prepared from PLGA-PEG or PLA-PEG copolymers by nanoprecipitation using known DB1 / 145613893.1 40 Attorney docket: NEX-013PC / 107578-5013 processes. PLGA and PLA portions of the block copolymer were about 20 kDa. PEG portions of the block co-polymers were either 3000 Da or 5000 Da. Specifically, PEG polymers having functional groups for ligand conjugation were designed to be longer (5000 Da), while the inert PEG polymers were shorter (3000 Da) to reduce steric effects. Experiments shown below employ NANOASSEMBLR IGNITE (Precision Nanosystems) for nanoprecipitation of particles. Nanoparticles include a population of copolymers having a terminal maleimide functional group (PEG-maleimide) for ligand coupling. Thiolated ligands are conjugated to naked particles through the terminal maleimide functional groups and purified by known techniques. See US 10,632,193. HLA ligands on the nanoparticles are loaded by incubating with excess peptide antigen. Peptide-loaded nanoparticles can be separated using an SEC column for example. Using these processes, particles can be obtained having a size and surface charge suitable for parenteral administration for immunotherapy. FIG. 1A shows TEM images of naked particles, protein-conjugated nanoparticles, and peptide loaded nanoparticles. As shown in FIG. 1B, the particles have an average diameter of about 100 nm. The size with ligands conjugated and peptide loaded are slightly larger. The polydispersity index was shown to be less than 0.2. The surface charge was shown to be within 0 and -10 mV. The targeted range of ligand density was 100 to 400 ligands per particle. Experiments were conducted to vary the density of PEG-maleimide functional groups on the surface of the particle, within the ranges determined to allow for stable particles (1% to 10% by weight). Specifically, particles were prepared with PLGA-PEG where a portion of the PEG termini contained maleimide functional groups. Maleimide:Thiol (Mal:Thiol) in the conjugation reaction was kept constant at 1:1. These data are summarized in Table 3 below. Table 3 1% 2.5% 5% 5% 7.5% 10% DB1 / 145613893.1 41 Attorney docket: NEX-013PC / 107578-5013 In a separate experiment, ligand density was controlled by varying the ratio of Maleimide functional groups on the nanoparticles to Thiol groups on ligands during the coupling reaction (Mal:Thiol). PEG-Mal was kept constant at 5%. These data are summarized in Table 4 below. Table 4 1:1 1:2 1:3 1:2 1:3 1:4 Experiment 3 Experiment 4 5% PEG-Maleimide polymer was selected as a good candidate for nanoparticle design in terms of stability, particle size, and ligand density. The data in Tables 3 and 4 is plotted in FIGs. 3A and 3B, showing the trends in particle density and size. 5% PEG- maleimide beads for PLGA-PEG in these examples corresponds to a 3:1 ratio (by weight) of PLGA-mPEG : PLGA-PEG-maleimide. For PLA-PEG, 5% PEG-maleimide corresponds to a ratio of about 5.67 : 1 (by weight) of PLA-mPEG : PLA-PEG-maleimide. Nanoparticles were constructed with PLGA-PEG and PLA-PEG nanoparticles for comparison. As shown in FIGs. 4A and 4B, differences were not significant. PLA-PEG showed a small reduction in size and protein density. PLGA-PEG and PLA-PEG nanoparticles having signal 1 ligands (HLA-Ig ligands loaded with MART-1 antigen) and signal 2 ligands (antibody agonist against CD28) were assayed for their ability to stimulate antigen-specific CD8+ T cells. As shown in FIGs. 5A-5D, no significant differences were seen between PLGA-PEG and PLA-PEG in terms of T cells positive for INFγ (FIG. 5A), TNFα (FIG. 5B), IL-2 (FIG. 5C), and CD107a (FIG. 5D). T cells prepared by enrichment and expansion (“AIM-ACT”) were used as a reference. See US Patent No.10,987,412 and US Patent 11,007,222, which are hereby incorporated by reference in their entireties. DB1 / 145613893.1 42 Attorney docket: NEX-013PC / 107578-5013 PLGA-PEG and PLA-PEG aAPCs loaded with CMV peptide antigen (D14) were tested for the ability to drive expansion of antigen-specific CD8+ T cells. As shown in the following Table 5 (using HLA_CMV tetramer stain), there were no significant differences between PLGA-PEG and PLA-PEG nanoparticles. Table 5 D14 Specificity CD8+ CMV (%) Ctrl (non-specific tetramer stain) (%) PLGA-PEG CMV 69.15 0.30 PLGA-PEG and PLA-PEG aAPC loaded with CMV peptide antigen were examined for their ability to generate T cell memory phenotypes essentially as follows. PBMCs were isolated from whole blood, and CD8+ T cells isolated using a commercially available kit. Cells were suspended at 1 x 106per mL in cytokine mix (see U.S. Patent No.11,007,222, which is hereby incorporated by reference). Cells (1 x 106) were resuspended with aAPCs (~10 µg) in 16 mL total volume.160 µL of the cell / aAPC suspension was plated per well in a 96 well round bottom plate. At day 4, cells were fed with cytokine mix. At day 7, cells were collected and counted and phenotypes were evaluated. As shown in the following Table 6, there were no significant differences in the ability of PLGA-PEG and PLA-PEG nanoparticles to generate T memory cells. CMV-loaded nanoparticles expanded CMV-specific CD8+ T cells with >90% memory phenotype. Table 6: D14 Memory Phenotypes Tcm% Tn / Tscm% Tem% Temra% As shown in FIG. 6, antigen-loaded aAPCs prepared essentially according to this example, when administered systemically to tumor-bearing mice, traffic to lymph nodes, spleen, and tumor (right panel), as compared to naked nanoparticles (left panel). aAPCs DB1 / 145613893.1 43 Attorney docket: NEX-013PC / 107578-5013 included MHC class I dimers (immunoglobulin fusion) and anti-CD28 agonist antibody ligands. Further, systemically administered aAPCs resulted in an increase of antigen-specific T cells in the spleen and tumor, and these T cells had greater killing potential as compared to peptide and CFA. FIGS. 7A-7C. As shown in FIG. 8A-C, the phenotype of T cells recovered from lymph nodes and tumors (after aAPC administration) have a phenotype consistent with persistence and strong anti-tumor effect. Further, aAPC administration (loaded with GP100 antigen) extended overall survival in B16 model (see FIG.9). Example 2: Construction of anti-Fas Agonistic Antibody Ligand CH11 is an IgM activating antibody against human Fas. The antibody demonstrates cytolytic activity on human cells that express Fas. The variable domains from CH11 were grafted to an IgG4 framework described in US Patent 10,632,193. See FIG.10A. The heavy chain and light chain amino acid sequences are provided as SEQ ID NO: 2 and SEQ ID NO: 3, respectively. The activity of the anti-Fas antibody (with and without dimerization by an anti-IgG4 antibody) were assayed in vitro for induction of apoptosis of CD8+ T cells (105cells per well). As shown in FIG.10B, while the pentameric CH11 induces a strong apoptotic effect on CD8+ T cells, non-crosslinked anti-Fas IgG4 has negligible effects. However, once crosslinked (here through dimerization with an anti-IgG4 antibody), the apoptotic effect was significant. These results suggest that anti-Fas antibody (such as that derived from clone CH11) may be effective for inducing apoptosis specifically of antigen-specific T cells by conjugation to nanoparticles having selected Signal 1 ligands (e.g., targeting autoreactive T cells). Example 3: Injectable aAPC bearing PD-L1 or anti-Fas Signal 2 ligands Injectable aAPCs were created as described in Example 1, and with PD-L1 Signal 2 ligands (SEQ ID NO: 1). As shown in FIG.11, these aAPCs rapidly inhibit antigen-specific killing of peptide-loaded target cells. Specifically, MART-1-specific CD8+ T cells were incubated for 90 minutes with 50 µg / mL of aAPCs. After washing, cells were incubated with peptide-loaded target cells for 4 hours and antigen-specific killing was assessed by caspase 3 / 7 assay. A 50% reduction of antigen-specific cytotoxic killing was observed. DB1 / 145613893.1 44 Attorney docket: NEX-013PC / 107578-5013 Injectable aAPCs were created as described in Example 1, and with anti-Fas Signal 2 ligands (as described in Example 2). As shown in FIG.12, these aAPCs rapidly eliminate antigen-specific T cells. FIG.12 shows 50% elimination of MART-1 specific T cells after a 4-hour incubation in vitro. Non-target T cells were not affected. For example, aAPC presenting a survivin peptide did not impact the number of MART-1 specific T cells. As shown in FIG.13, the anti-Fas aAPCs eliminate up to 90% of MART-1-specific T cells by Day 13 in a mouse model. This effect is dose dependent. Example 4: Selection of Peptide Antigens for aAPCs It was observed that certain peptide antigens, when loaded onto nanoparticles (aAPCs), would induce aggregation of the particles. This effect was evaluated to determine the properties of peptide antigens suitable for loading onto nano-aAPCs without inducing an aggregation effect. Protein modeling and aggregation propensity analyses were performed using Discovery Studio 2021 (DS2021), a software product from BIOVIA (Dassault Systemes BIOVIA, Discovery Studio Modeling Environment, 2021, San Diego CA). Calculation of aggregation propensity scores based on Trout’s method is included in DS2021. The MODELLER program (UCSF, San Francisco CA) was used for homology and comparative modeling of protein three-dimensional structures. The process for identifying peptide antigens with an aggregation propensity when presented by HLA ligands on nanoparticles is shown diagrammatically in FIG.14. First, molecular models of the relevant peptide-HLA complexes are generated. HLA- A:0201 structure models include: 4L29, 1I7R, 5EUO, 1TVB, 6OPD, 6TRO, 6AMT, 6AM5, 2GT9, and 3OXS. Modeling of the complexes is based on generating a standard comparative protein model of the relevant peptides positioned in the HLA binding groove. The MODELER algorithm as implemented in Discovery Studio 2021 was employed. Multiple template structures are used for the modeling steps, and manual sequence alignments are performed on each model. Next, using the top-scoring completed static model, the ‘Calculate Aggregation Scores’ protocol as implemented in Discovery Studio 2021 is run. Radii of 5 and 10 DB1 / 145613893.1 45 Attorney docket: NEX-013PC / 107578-5013 Angstroms was selected as radii for calculating scores. This parameter determines how many nearby amino acid residues to include in the aggregation potential calculation. Depending on the size of the model being analyzed, the radius can give an appropriately smoothed surface map that is not overly noisy. The protocol calculates exposed surface for each residue of the peptide antigen and calculates an aggregation potential score based on the averaging of each residue and those in the defined radius surrounding the residue. The numerical aggregation potential score (APS) is recorded with the sequence and structure information and is used to generate a surface map based on the solvent-accessible surface of the model. This provides a visualization of the shape of the surface coupled with the scored aggregation potential. Peptides can be binned as aggregators or non-aggregators by manual inspection or computationally. By evaluating numerous peptide antigens, inferences were prepared to guide the evaluation of candidate peptides. Here, the APS of residues 1 and 3-5 were averaged. If above a threshold of 0.07, this is considered a factor for aggregation. The presence of an exposed Cys residue (as determined by inspection of the model structures) is considered another factor for aggregation. Presence of charged residues and Gly residues are considered factors against aggregation. The total number of factors in favor or against aggregation can determine whether the peptide-HLA complex is binned as a potential aggregator or not. FIG. 15A and 15B illustrate modelling and scoring of aggregating and non- aggregating peptides, respectively. This process was used for a set of known aggregators and non-aggregators as summarized in Table 7 below. Table 7: Aggregation Analysis Peptide Name Known Predicted DB1 / 145613893.1 46 Attorney docket: NEX-013PC / 107578-5013 CMTWNQMNL WT1.235.M0010 x x (SEQ ID NO: 47) FLDRFLSCM CYCLIN.227.M0013 REFERENCES B. Webb, A. Sali. Comparative Protein Structure Modeling Using Modeller. Current Protocols in Bioinformatics 54, John Wiley & Sons, Inc., 5.6.1-5.6.37, 2016. M.A. Marti-Renom, et al. Comparative protein structure modeling of genes and genomes. Annu. Rev. Biophys. Biomol. Struct.29, 291-325, 2000. A. Sali & T.L. Blundell. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol.234, 779-815, 1993. A. Fiser, R.K. Do, & A. Sali. Modeling of loops in protein structures. Protein Science 9.1753-1773, 2000. Chennamsetty, N., et al. Design of therapeutic proteins with enhanced stability. Proc. Natl. Acad. Sci. USA 2009, 106(29), 11937-11942. Chennamsetty, N., et al. Aggregation-Prone Motifs in Human Immunoglobulin G. J. Mol. Biol.2009, 391, 404-413. Chennamsetty, N., et al. Prediction of Aggregation Prone Regions of Therapeutic Proteins. J. Phys. Chem. B 2010, 114, 6614-6624. DB1 / 145613893.1 47 Attorney docket: NEX-013PC / 107578-5013 Schütz C, et al. Antigen-specific T cell Redirectors: a nanoparticle based approach for redirecting T cells. Oncotarget.2016 Oct 18;7(42):68503-68512. Schappert A, et al. Soluble MHC class I complexes for targeted immunotherapy. Life Sci.2018 Sep 15;209:255-258. Turtle CJ, Riddell SR. Artificial antigen-presenting cells for use in adoptive immunotherapy. Cancer J.2010;16(4):374-381. Black SD, Mould DR. Development of hydrophobicity parameters to analyze proteins which bear post- or cotranslational modifications. Anal Biochem.1991 Feb 15;193(1):72-82. Riley TP, et al. Structure Based Prediction of Neoantigen Immunogenicity. Front Immunol.2019 Aug 28;10:2047. Liu J, et al. Structural insights into the binding of hepatitis B virus core peptide to HLA-A2 alleles: towards designing better vaccines. Eur J Immunol. 2011 Jul;41(7):2097-106. Halabelian L et al. Class I major histocompatibility complex, the trojan horse for secretion of amyloidogenic β2-microglobulin. J Biol Chem.2014 Feb 7;289(6):3318- 27. Coles CH, et al. T cell receptor interactions with human leukocyte antigen govern indirect peptide selectivity for the cancer testis antigen MAGE-A4. J Biol Chem. 2020 Aug 14;295(33):11486-11494. Buslepp J, et al. T cell activity correlates with oligomeric peptide-major histocompatibility complex binding on T cell surface. J Biol Chem. 2001 Dec 14;276(50):47320-8. Borbulevych OY, et al. Structures of MART-126 / 27-35 Peptide / HLA-A2 complexes reveal a remarkable disconnect between antigen structural homology and T cell recognition. J Mol Biol.2007 Oct 5;372(5):1123-36. Borbulevych OY, et al. Increased immunogenicity of an anchor-modified tumor- associated antigen is due to the enhanced stability of the peptide / MHC complex: implications for vaccine design. J Immunol.2005 Apr 15;174(8):4812-20. DB1 / 145613893.1 48 Attorney docket: NEX-013PC / 107578-5013 SEQUENCES SEQ ID NO: 1 – PD-L1-linker-IgG4 FTVTVPKDLYVVEYGSNMTIECKFPVEKQLDLAALIVYWEMEDKNIIQFVHGEEDLKVQH SSYRQRARLLKDQLSLGNAALQITDVKLQDAGVYRCMISYGGADYKRITVKVNAPYNKIN QRILVVDPVTSEHELTCQAEGYPKAEVIWTSSDHQVLSGKTTTTNSKREEKLFNVTSTLR INTTTNEIFYCTFRRLDPEENHTAELVIPELPLAHPPNERTGGGGSGGGGSGGGGSESKY GPPCPPCPAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGV EVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQ PREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDG SFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 2 – CH11 Heavy Chain-IgG4 EVQLQQSGPELVKPGASVKISCKASGYTFTdynmhWVKQSHGKSLEWIGyiypynggtgy nqkfksKATLTVDNSSSTAYMELRSLTSEDSAVYYCARsyyamdyWGQGTSVTVSSASTK GPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVLQSSGLYS LSSVVTVPSSSLGTKTYTCNVDHKPSNTKVDKRVESKYGPPCPPCPAPEFEGGPSVFLFP PKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVS VLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVS LTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFS CSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 3 – CH11 Light Chain DVVMTQSPLSLPVSLGDQASISCrsskslvhsngntylhWYLQKPGQSPKLLIYkvsnrf SGVPDRFSGSGSGTDFTLKISRVEAEDLGVYFCsqsthvppaFGGGTKLEIKRTVAAPSV FIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSL SSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC SEQ ID NO: 4 – Cys linker-IgG4Fc-Linker-FasL DNSLCLSLGGGGSESKYGPPCPPCPAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVD VSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSN KGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNG QPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLSLSL GKGGGGSGGGGSGGGGSPSPPPEKKELRKVAHLTGKSNSRSMPLEWEDTYGIVLLSGVKY KKGGLVINETGLYFVYSKVYFRGQSCNNLPLSHKVYMRNSKYPQDLVMMEGKMMSYCTTG QMWARSSYLGAVFNLTSADHLYVNVSELSLVNFEESQTFFGLYKL SEQ ID NO: 5 – PD-L1-Linker-HLA-A*02:01(C84,C139)-Linker- IgG4 FTVTVPKDLYVVEYGSNMTIECKFPVEKQLDLAALIVYWEMEDKNIIQFVHGEEDLKVQH SSYRQRARLLKDQLSLGNAALQITDVKLQDAGVYRCMISYGGADYKRITVKVNAPYNKIN QRILVVDPVTSEHELTCQAEGYPKAEVIWTSSDHQVLSGKTTTTNSKREEKLFNVTSTLR DB1 / 145613893.1 49 Attorney docket: NEX-013PC / 107578-5013 INTTTNEIFYCTFRRLDPEENHTAELVIPELPLAHPPNERTGGGGSGGGGSGGGGSGSHS MRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAASQRMEPRAPWIEQEGPEYWDGET RKVKAHSQTHRVDLGTLRGCYNQSEAGSHTVQRMYGCDVGSDWRFLRGYHQYAYDGKDYI ALKEDLRSWTAADMCAQTTKHKWEAAHVAEQLRAYLEGTCVEWLRRYLENGKETLQRTDA PKTHMTHHAVSDHEATLRCWALSFYPAEITLTWQRDGEDQTQDTELVETRPAGDGTFQKW AAVVVPSGQEQRYTCHVQHEGLPKPLTLRWGGGGSGGGGSGGGGSESKYGPPCPPCPAPE FEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKTKPRE EQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPP SQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVD KSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 6 – αCD28scFv (VH-linker-VL)-Linker-HLA- A*02:01(C84, C139)-Linker-IgG4 EVKLQQSGPGLVKPSETLSLTCTVSGFSLSDYGVHWVRQAPGKGLEWLGVIWAGGGTNYN SALMSRKTISKDNSKSQVSLKMSSVTAADTAVYYCARDKGYSYYYSMDYWGQGTLVTVSS GGGGSGGGGSGGGGSDIELTQSPDSLAVSLGERATINCRASESVEYYVTSLMQWYQQKPG QPPKLLIFAASNVESGVPDRFSGSGSGTDFTLTISSLQAEDVAMYFCQQSRKVPYTFGGG TKVEIKRGGGGSGGGGSGGGGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDA ASQRMEPRAPWIEQEGPEYWDGETRKVKAHSQTHRVDLGTLRGCYNQSEAGSHTVQRMYG CDVGSDWRFLRGYHQYAYDGKDYIALKEDLRSWTAADMCAQTTKHKWEAAHVAEQLRAYL EGTCVEWLRRYLENGKETLQRTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLTWQRD GEDQTQDTELVETRPAGDGTFQKWAAVVVPSGQEQRYTCHVQHEGLPKPLTLRWGGGGSG GGGSGGGGSESKYGPPCPPCPAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQE DPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLP SSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPEN NYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 7 – αCD28scFv (VL-linker-VH)-Linker-HLA- A*02:01(C84, C139)-Linker-IgG4 DIELTQSPDSLAVSLGERATINCRASESVEYYVTSLMQWYQQKPGQPPKLLIFAASNVES GVPDRFSGSGSGTDFTLTISSLQAEDVAMYFCQQSRKVPYTFGGGTKVEIKRGGGGSGGG GSGGGGSEVKLQQSGPGLVKPSETLSLTCTVSGFSLSDYGVHWVRQAPGKGLEWLGVIWA GGGTNYNSALMSRKTISKDNSKSQVSLKMSSVTAADTAVYYCARDKGYSYYYSMDYWGQG TLVTVSGGGGSGGGGSGGGGSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAA SQRMEPRAPWIEQEGPEYWDGETRKVKAHSQTHRVDLGTLRGCYNQSEAGSHTVQRMYGC DVGSDWRFLRGYHQYAYDGKDYIALKEDLRSWTAADMCAQTTKHKWEAAHVAEQLRAYLE GTCVEWLRRYLENGKETLQRTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLTWQRDG EDQTQDTELVETRPAGDGTFQKWAAVVVPSGQEQRYTCHVQHEGLPKPLTLRWGGGGSGG GGSGGGGSESKYGPPCPPCPAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQED PEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPS SIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENN YKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 8 – Beta-2-microglobulin DB1 / 145613893.1 50 Attorney docket: NEX-013PC / 107578-5013 IQRTPKIQVYSRHPAENGKSNFLNCYVSGFHPSDIEVDLLKNGERIEKVEHSDLSFSKDW SFYLLYYTEFTPTEKDEYACRVNHVTLSQPKIVKWDRDM SEQ ID NO: 9 – Anti-CD28 antibody HC CDR1 DYGVH SEQ ID NO: 10 – Anti-CD28 antibody HC CDR2 VIWAGGGTNYNSALMS SEQ ID NO: 11 – Anti-CD28 antibody HC CDR3 DKGYSYYYSMDY SEQ ID NO: 12 – Anti-CD28 antibody LC CDR1 RASESVEYYVTSLMQ SEQ ID NO: 13 – Anti-CD28 antibody LC CDR2 AASNVES SEQ ID NO: 14 – Anti-CD28 antibody LC CDR3 QQSRKVPYT SEQ ID NO: 15 – Humanized anti-CD28 Heavy Chain Variable Region EVKLQQSGPGLVKPSETLSLTCTVSGFSLSDYGVHWVRQAPGKGLEWLGVIWAGGGTNYN SALMSRKTISKDNSKSQVFLKMNSLTAADTAVYYCARDKGYSYYYSMDYWGQGTLVTVSS SEQ ID NO: 16 – Humanized anti-CD28 Heavy Chain Variable region EVKLQQSGPGLVKPSETLSLTCTVSGFSLSDYGVHWVRQAPGKGLEWLGVIWAGGGTNYN SALMSRKTISKDNSKSQVSLKMSSVTAADTAVYYCARDKGYSYYYSMDYWGQGTLVTVSS SEQ ID NO: 17 – Humanized anti-CD28 Heavy Chain Variable Region EVKLQQSGPGLVKPSETLSLTCTVSGFSLSDYGVHWVRQAPGKGLEWLGVIWAGGGTNYN SALMSRVTISKDNSKSQVSLKLSSVTAADTAVYYCARDKGYSYYYSMDYWGQGTLVTVSS SEQ ID NO: 18 – Humanized anti-CD28 Light Chain DB1 / 145613893.1 51 Attorney docket: NEX-013PC / 107578-5013 DIELTQSPDSLAVSLGERATINCRASESVEYYVTSLMQWYQQKPGQPPKLLIFAASNVES GVPDRFSGSGSGTNFTLTISSLQEEDVAMYFCQQSRKVPYTFGGGTKVEIK SEQ ID NO: 19 – Humanized anti-CD28 Light Chain DIELTQSPDSLAVSLGERATINCRASESVEYYVTSLMQWYQQKPGQPPKLLIFAASNVES GVPDRFSGSGSGTNFTLTISSLQAEDVAMYFCQQSRKVPYTFGGGTKVEIK SEQ ID NO: 20 – Humanized anti-CD28 Light Chain DIELTQSPDSLAVSLGERATINCRASESVEYYVTSLMQWYQQKPGQPPKLLIFAASNVES GVPDRFSGSGSGTDFTLTISSLQAEDVAMYFCQQSRKVPYTFGGGTKVEIK SEQ ID NO: 21 – HLA-E extracellular domain GSHSLKYFHTSVSRPGRGEPRFISVGYVDDTQFVRFDNDAASPRMVPRAPWMEQEGSEYW DRETRSARDTAQIFRVNLRTLRGYYNQSEAGSHTLQWMHGCELGPDGRFLRGYEQFAYDG KDYLTLNEDLRSWTAVDTAAQISEQKSNDASEAEHQRAYLEDTCVEWLHKYLEKGKETLL HLEPPKTHVTHHPISDHEATLRCWALGFYPAEITLTWQQDGEGHTQDTELVETRPAGDGT FQKWAAVVVPSGEEQRYTCHVQHEGLPEPVTLRW SEQ ID NO: 22 – Engineered HLA-E extracellular domain (mutations shown in bold and underline) GSHSLKYFHTSVSRPGRGEPRFISVGYVDDTQFVRFDNDAASPRMVPRAPWMEQEGSEYW DRETRSARDTAQIFRVNLRTLRGCYNQSEAGSHTLQWMHGCELGPDGRFLRGYEQFAYDG KDYLTLNEDLRSWTAVDTCAQISEQKSNDASEAEHQRAYLEATCVAWLHKYLEKGKETLL HLEPPKTHVTHHPISDHEATLRCWALGFYPAEITLTWQQDGEGHTQDTELVETRPAGDGT FQKWAAVVVPSGEEQRYTCHVQHEGLPEPVTLRW SEQ ID NO: 23 – IgG4 Fc ESKYGPPCPPCPAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWY VDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISK AKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVL DSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 24 – Flexible Linker GGGGSGGGGSGGGGS SEQ ID NO: 25 – HLA-E-Fc Fusion (HLA-E mutations shown in bold and underline, IgG4 Fc shown in capital italics, flexible linker shown in lowercase italics) GSHSLKYFHTSVSRPGRGEPRFISVGYVDDTQFVRFDNDAASPRMVPRAPWMEQEGSEYW DRETRSARDTAQIFRVNLRTLRGCYNQSEAGSHTLQWMHGCELGPDGRFLRGYEQFAYDG KDYLTLNEDLRSWTAVDTCAQISEQKSNDASEAEHQRAYLEATCVAWLHKYLEKGKETLL DB1 / 145613893.1 52 Attorney docket: NEX-013PC / 107578-5013 HLEPPKTHVTHHPISDHEATLRCWALGFYPAEITLTWQQDGEGHTQDTELVETRPAGDGT FQKWAAVVVPSGEEQRYTCHVQHEGLPEPVTLRWggggsggggsggggsESKYGPPCPPC PAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKT KPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVY TLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSR LTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK SEQ ID NO: 26 -- HLA-A*02:01(C84, C139)-linker-IgG4 GSHSMRYFFTSVSRPGRGEPRFIAVGYVDDTQFVRFDSDAASQRMEPRAPWIEQEGPEYW DGETRKVKAHSQTHRVDLGTLRGCYNQSEAGSHTVQRMYGCDVGSDWRFLRGYHQYAYDG KDYIALKEDLRSWTAADMCAQTTKHKWEAAHVAEQLRAYLEGTCVEWLRRYLENGKETLQ RTDAPKTHMTHHAVSDHEATLRCWALSFYPAEITLTWQRDGEDQTQDTELVETRPAGDGT FQKWAAVVVPSGQEQRYTCHVQHEGLPKPLTLRWGGGGSGGGGSGGGGSESKYGPPCPPC PAPEFEGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQEDPEVQFNWYVDGVEVHNAKT KPREEQFNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVY TLPPSQEEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSR LTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLCLSLGK DB1 / 145613893.1 53

Claims

Attorney docket: NEX-013PC / 107578-5013 CLAIMS:

1. An artificial antigen presenting cell (aAPC) suitable for parenteral administration, comprising: poly(lactic acid)-polyethylene glycol (PLA-PEG) or poly(lactic acid-co-glycolic acid)-polyethylene glycol (PLGA-PEG) copolymers and one or more polypeptide ligands conjugated to PEG through a thioether bond; wherein: the polypeptide ligands comprise HLA ligands presenting a peptide antigen and optionally one or more signal 2 ligands, and about 40% or less by weight of the copolymers have a functional group for polypeptide ligand coupling.

2. The aAPC of claim 1, wherein about 15% to about 35% by weight of the copolymers have a functional group for polypeptide ligand coupling.

3. The aAPC of any one of claims 1 or 2, wherein the PLA or PLGA portion of the copolymers have a molecular weight of from about 15 kDa to about 50 kDa, or from about 15 kDa to about 35 kDa, or from about 15 kDa to about 25 kDa.

4. The aAPC of claim 3, wherein the PLA or PLGA portion of the copolymers have a molecular weight of about 20 kDa.

5. The aAPC of claim 3 or 4, wherein the PEG portion of the copolymers have molecular weights in the range of about 2 kDa to about 10 kDa, or in the range of about 2 kDa to about 7 kDa.

6. The aAPC of claim 5, wherein the PEG portion of the copolymers have molecular weights in the range of about 2 kDa and about 5 kDa. DB1 / 145613893.1 54Attorney docket: NEX-013PC / 107578-5013 7. The aAPC of claim 6, wherein the PEG portions having a functional group for polypeptide ligand coupling have molecular weights of about 5 kDa, and the PEG portions without functional groups for ligand coupling have molecular weights of about 3 kDa.

8. The aAPC of any one of claims 1 to 7, wherein sulfhydryl groups on the polypeptide ligands are coupled to PEG-maleimide functional groups.

9. The aAPC of any one of claims 1 to 8, wherein the aAPC has about 10 to about 500 polypeptide ligands.

10. The aAPC of claim 9, wherein the aAPC has about 50 to about 400 polypeptide ligands.

11. The aAPC of claim 10, wherein the aAPC has about 100 to about 300 polypeptide ligands.

12. The aAPC of any one of claims 1 to 11, wherein the aAPC has a diameter of from about 50 nm to about 150 nm.

13. The aAPC of claim 12, wherein the aAPC has a diameter of from about 60 nm to about 130 nm.

14. The aAPC of claim 13, wherein the aAPC has a diameter of from about 80 nm to about 120 nm.

15. The aAPC of claim 13, wherein the aAPC has a diameter of about 60 nm, 80 nm, about 90 nm, about 100 nm, about 110 nm, or about 120 nm.

16. The aAPC of any one of claims 1 to 15, wherein aAPC has a surface charge of from about 0 to -15 mV, or from about 0 to about -10 mV. DB1 / 145613893.1 55Attorney docket: NEX-013PC / 107578-5013 17. The aAPC of claim 16, wherein the aAPC has a surface charge of from about -2.5 mV to about -10 mV.

18. The aAPC of any one of claims 1 to 17, wherein the aAPC is a population of aAPCs having a size distribution with polydispersity index (PDI) of less than 0.

2.

19. The aAPC of any one of claims 1 to 18, wherein the HLA ligands are HLA class I ligands, and are optionally HLA-A, HLA-B, HLA-C, or HLA-E ligands.

20. The aAPC of claim 19, wherein the HLA ligand comprises beta2 microglobulin (β2M).

21. The aAPC of claim 19 or 20, wherein the HLA ligands are HLA-A ligands, and optionally HLA-A*02:01 ligands.

22. The aAPC of any one of claims 19 to 21, wherein the HLA ligand has a peptide- binding cleft stabilized with a disulfide bond.

23. The aAPC of any one of claims 19 to 22, wherein the peptide antigen does not induce aggregation of the aAPC.

24. The aAPC of claim 23, wherein the mean Aggregation Potential Score (APS) of peptide residues 1 and 3-5 is less than about 0.

07.

25. The aAPC of claim 23 or 24, wherein the peptide antigen does not have an exposed Cysteine.

26. The aAPC of any one of claims 23 to 25, wherein the peptide antigen has one or more exposed glycine residues or exposed charged residues. DB1 / 145613893.1 56Attorney docket: NEX-013PC / 107578-5013 27. The aAPC of any one of claims 19 to 26, wherein the HLA ligand is dimeric, and comprises a fusion to an immunoglobulin Fc region, which is optionally IgG4 isotype.

28. The aAPC of any one of claims 19 to 26, wherein the HLA ligand comprises a fusion to an immunoglobulin Fc, which is optionally IgG4 isotype, and is optionally dimerized with a signal 2-Fc fusion.

29. The aAPC of any one of claims 19 to 27, wherein the HLA ligand comprises a fusion to a signal 2 ligand.

30. The aAPC of claim 29, wherein the signal 2 ligand comprises a single chain antibody, which is optionally scFv.

31. The aAPC of any one of claims 1 to 18, wherein the HLA ligands are HLA class II ligands, and are optionally HLA-DR, HLA-DP, or HLA-DQ.

32. The aAPC of claim 31, wherein the HLA class II ligands comprise immunoglobulin fusions of HLA alpha and beta chains to antibody heavy and light chains.

33. The aAPC of any one of claims 1 to 32, wherein the polypeptide ligands comprise a co-stimulatory ligand.

34. The aAPC of claim 33, wherein the co-stimulatory ligand is an agonist for CD28, 4- 1BB, CD27, OX-40, CD30, ICOS, and LIGHT.

35. The aAPC of claim 34, wherein the co-stimulatory ligand is an agonistic antibody against CD28.

36. The aAPC of claim 35, wherein the aAPC further comprises one or more cytokines that support T cell activation and / or expansion conjugated to the aAPC as a polypeptide ligand, fused to a signal 1 or signal 2 polypeptide ligand, or encapsulated by the copolymers. DB1 / 145613893.1 57Attorney docket: NEX-013PC / 107578-5013 37. The aAPC of any one of claims 1 to 32, wherein the polypeptide ligands do not comprise any signal 2 ligand.

38. The aAPC of any one of claims 1 to 32, wherein the polypeptide ligands comprise co-inhibitory ligands.

39. The aAPC of claim 38, wherein the co-inhibitory ligand is an agonist for Fas, TGF- β, or PD-1.

40. The aAPC of claim 39, wherein the co-inhibitory ligand is an agonist antibody for Fas.

41. The aAPC of claim 40, wherein the agonist antibody for Fas is an IgG4 antibody based on clone CH11.

42. The aAPC of claim 39, wherein the co-inhibitory ligand is FasL, which is optionally an immunoglobulin Fc fusion.

43. The aAPC of claim 39, wherein the co-inhibitory ligand is PD-L1, which is optionally an immunoglobulin Fc fusion.

44. The aAPC of claim 43, wherein an activating fragment of PD-L1 is fused to an HLA class I ligand, which is optionally dimerized by fusion to an immunoglobulin Fc.

45. The aAPC of any one of claims 41 to 44, wherein the aAPC further comprises a tolerogenic cytokine as a polypeptide ligand, fused to a signal 1 polypeptide ligand, or encapsulated by the copolymers, wherein the tolerogenic cytokine is optionally IL-10.

46. The aAPC of any one of claims 1 to 45, wherein the polypeptide ligands further comprise one or more homing ligands for a tissue or organ. DB1 / 145613893.1 58Attorney docket: NEX-013PC / 107578-5013 47. The aAPC of any one of claims 1 to 46, wherein the aAPC is comprised in a pharmaceutical composition suitable for administration to a subject.

48. The aAPC of claim 47, wherein the pharmaceutical composition is lyophilized.

49. An artificial antigen presenting cell (aAPC) suitable for parenteral administration, comprising: polymeric or lipid nanoparticles comprising a polyethylene glycol (PEG) sheath and one or more polypeptide ligands conjugated to PEG through a thioether bond; wherein: the polypeptide ligands comprise HLA Class I or Class II ligands presenting a peptide antigen and optionally one or more signal 2 ligands, and the peptide antigen does not induce aggregation of the aAPC.

50. The aAPC of claim 49, wherein the HLA ligands are HLA class I ligands, and are optionally HLA-A, HLA-B, HLA-C, or HLA-E ligands.

51. The aAPC of claim 50, wherein the HLA ligand comprises beta2 microglobulin (β2M).

52. The aAPC of claim 51, wherein the HLA ligands are HLA-A ligands, and optionally HLA-A*02:

01.

53. The aAPC of any one of claims 49 to 52, wherein the HLA ligand has a peptide- binding cleft stabilized with a disulfide bond.

54. The aAPC of any one of claims 49 to 53, wherein the mean Aggregation Potential Score (APS) of peptide residues 1 and 3-5 is less than about 0.

07. DB1 / 145613893.1 59Attorney docket: NEX-013PC / 107578-5013 55. The aAPC of any one of claims 49 to 54, wherein the peptide antigen does not have an exposed Cysteine.

56. The aAPC of any one of claims 49 to 55, wherein the peptide antigen has one or more exposed glycine residues or exposed charged residues.

57. The aAPC of any one of claims 49 to 56, wherein the nanoparticles are polymeric nanoparticles comprising poly(lactic acid)-polyethylene glycol (PLA-PEG) or poly(lactic acid-co-glycolic acid)-polyethylene glycol (PLGA-PEG) copolymers and one or more polypeptide ligands conjugated to PEG through a thioether bond.

58. The aAPC of claim 57, wherein about 40% or less by weight of the copolymers have a functional group for polypeptide ligand coupling.

59. The aAPC of claim 58, wherein about 15% to about 35% by weight of the copolymers have a functional group for polypeptide ligand coupling.

60. The aAPC of any one of claims 57 to 59, wherein the PLA or PLGA portion of the copolymers have a molecular weight of from about 15 kDa to about 50 kDa, or from about 15 kDa to about 35 kDa, or from about 15 kDa to about 25 kDa.

61. The aAPC of claim 60, wherein the PLA or PLGA portion of the copolymers have a molecular weight of about 20 kDa.

62. The aAPC of claim 60 or 61, wherein the PEG portion of the copolymers have molecular weights in the range of about 2 kDa to about 10 kDa, or in the range of about 2 kDa to about 7 kDa.

63. The aAPC of claim 62, wherein the PEG portion of the copolymers have molecular weights in the range of about 2 kDa and about 5 kDa. DB1 / 145613893.1 60Attorney docket: NEX-013PC / 107578-5013 64. The aAPC of claim 63, wherein the PEG portions having a functional group for polypeptide ligand coupling have molecular weights of about 5 kDa, and the PEG portions without functional groups for ligand coupling have molecular weights of about 3 kDa.

65. The aAPC of any one of claims 49 to 56, wherein the nanoparticles are lipid nanoparticles comprising a PEG-conjugated lipid.

66. The aAPC of claim 65, wherein the lipid nanoparticles further comprise a cationic or ionizable lipid, a neutral lipid or phospholipid, and a structural lipid such as a cholesterol or cholesterol moiety.

67. The aAPC of any one of claims 49 to 66, wherein sulfhydryl groups on the polypeptide ligands are coupled to PEG-maleimide functional groups.

68. The aAPC of any one of claims 49 to 67, wherein the aAPC has about 10 to about 500 polypeptide ligands.

69. The aAPC of claim 68, wherein the aAPC has about 50 to about 400 polypeptide ligands.

70. The aAPC of claim 69, wherein the aAPC has about 100 to about 300 polypeptide ligands.

71. The aAPC of any one of claims 49 to 70, wherein the aAPC has a diameter of from about 50 nm to about 150 nm.

72. The aAPC of claim 71, wherein the aAPC has a diameter of from about 60 nm to about 130 nm.

73. The aAPC of claim 72, wherein the aAPC has a diameter of from about 80 nm to about 120 nm. DB1 / 145613893.1 61Attorney docket: NEX-013PC / 107578-5013 74. The aAPC of claim 71, wherein the aAPC has a diameter of about 60 nm, about 80 nm, about 90 nm, about 100 nm, about 110 nm, or about 120 nm.

75. The aAPC of any one of claims 49 to 74, wherein aAPC has a surface charge of from about 0 to -15 mV, or from about 0 to about -10 mV.

76. The aAPC of claim 75, wherein the aAPC has a surface charge of from about -2.5 mV to about -10 mV.

77. The aAPC of any one of claims 49 to 76, wherein the aAPC is a population of aAPCs having a size distribution with polydispersity index (PDI) of less than 0.

2.

78. The aAPC of any one of claims 49 to 77, wherein the HLA ligand is dimeric, and comprises two fusion to an immunoglobulin Fc, which is optionally IgG4 isotype.

79. The aAPC of any one of claims 49 to 77, wherein the HLA ligand comprises a fusion to an immunoglobulin Fc, which is optionally IgG4 isotype, and is optionally dimerized with a signal 2-Fc fusion.

80. The aAPC of any one of claims 49 to 78, wherein the HLA ligand comprises a fusion to a signal 2 ligand.

81. The aAPC of claim 80, wherein the signal 2 ligand comprises a single chain antibody, which is optionally scFv.

82. The aAPC of any one of claims 49 to 81, wherein the polypeptide ligands comprise a co-stimulatory ligand.

83. The aAPC of claim 82, wherein the co-stimulatory ligand is an agonist for CD28, 4- 1BB, CD27, OX-40, CD30, ICOS, and LIGHT. DB1 / 145613893.1 62Attorney docket: NEX-013PC / 107578-5013 84. The aAPC of claim 83, wherein the co-stimulatory ligand is an agonistic antibody against CD28.

85. The aAPC of any one of claims 49 to 84, wherein the aAPC further comprises one or more cytokines that support T cell activation and / or expansion conjugated to the aAPC as a polypeptide ligand, fused to a signal 1 or signal 2 polypeptide ligand, or encapsulated by the copolymers.

86. The aAPC of any one of claims 49 to 78, wherein the polypeptide ligands do not comprise any signal 2 ligand.

87. The aAPC of any one of claims 49 to 81, wherein the polypeptide ligands comprise co-inhibitory ligands.

88. The aAPC of claim 87, wherein the co-inhibitory ligand is an agonist for Fas, TGF- β, or PD-1.

89. The aAPC of claim 88, wherein the co-inhibitory ligand is an agonist antibody for Fas.

90. The aAPC of claim 89, wherein the agonist antibody is an IgG4 antibody based on clone CH11.

91. The aAPC of claim 88, wherein the co-inhibitory ligand is FasL, which is optionally an immunoglobulin Fc fusion.

92. The aAPC of claim 88, wherein the co-inhibitory ligand is PD-L1, which is optionally an immunoglobulin Fc fusion. DB1 / 145613893.1 63Attorney docket: NEX-013PC / 107578-5013 93. The aAPC of claim 92, wherein an activating fragment of PD-L1 is fused to an HLA class I ligand, which is optionally dimerized by fusion to an immunoglobulin Fc.

94. The aAPC of any one of claims 86 to 93, wherein the aAPC further comprises a tolerogenic cytokine as a polypeptide ligand, as a fusion to a signal 1 or signal 2 polypeptide ligand, or encapsulated by the copolymers, wherein the tolerogenic cytokine is optionally IL-10.

95. The aAPC of any one of claims 49 to 94, wherein the polypeptide ligands further comprise one or more homing ligands.

96. The aAPC of any one of claims 49 to 95, wherein the aAPC is comprised in a pharmaceutical composition suitable for administration to a subject.

97. The aAPC of claim 96, wherein the pharmaceutical composition is lyophilized.

98. A method for immunotherapy, comprising, administering the aAPC or pharmaceutical composition thereof of any one of claims 1 to 97 to a subject in need of treatment.

99. The method of claim 98, wherein the subject has cancer or an infectious disease, and the aAPCs comprise a co-stimulatory ligand.

100. The method of claim 99, wherein the subject has a hematological malignancy.

101. The method of claim 99, wherein the subject has a solid tumor, which is optionally Stage I, Stage II, Stage III, or Stage IV.

102. The method of any one of claims 99 to 101, wherein the subject has a human papilloma virus (HPV)-associated cancer. DB1 / 145613893.1 64Attorney docket: NEX-013PC / 107578-5013 103. The method of any one of claims 98 to 102, wherein the subject is administered an immune checkpoint inhibitor therapy prior to, during, or after administration of the aAPC or pharmaceutical composition thereof.

104. The method of claim 98, wherein the subject has an autoimmune condition, and the aAPCs comprise a co-inhibitory signal or do not contain a signal 2 ligand.

105. The method of claim 104, wherein the autoimmune condition is type 1 diabetes.

106. The method of claim 104, wherein the autoimmune condition is selected from vitiligo, systemic lupus erythematosus, rheumatoid arthritis, multiple sclerosis, Crohn's disease, ulcerative colitis, psoriasis, myasthenia gravis, Goodpasture's syndrome, Graves' disease, pemphigus vulgaris, Addison's disease, dermatitis herpetiformis, celiac disease, Sjogren’s Disease, Hashimoto's thyroiditis, alopecia, ankylosing spondylitis, scleroderma, and HTLV-1-Associated Myelopathy (HAM) / Tropical Spastic Paraparesis (TSP), among others.

107. The method of any one of claims 98 to 106, wherein the aAPC or pharmaceutical composition thereof is parenterally administered.

108. The method of claim 107, wherein the aAPC or pharmaceutical composition thereof is administered by intravenous administration, subcutaneous administration, or intramuscular administration.

109. The method of claim 108, wherein the aAPC or pharmaceutical composition thereof is administered subcutaneously.

110. A method for making an artificial Antigen Presenting Cell, comprising: providing an in silico HLA ligand comprising an antigen binding cleft, the antigen binding cleft comprising a candidate peptide antigen; DB1 / 145613893.1 65Attorney docket: NEX-013PC / 107578-5013 preparing a surface map may of the antigen binding clef comprising the candidate peptide antigen, the surface map providing an aggregation potential score for each residue of the candidate peptide antigen and surrounding amino acid residues; selecting a candidate peptide that is below a threshold aggregation potential; and loading the selected peptide onto aAPCs.

111. An anti-Fas agonistic antibody having an IgG isotype, and conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen.

112. A dimeric PD-L1 ligand comprising amino acids residues F19 to T239 of human PD-L1 fused directly or indirectly through a linker at its C-terminus to a dimerized IgG Fc region, the dimeric PD-L1 ligand conjugated to nanoparticles together with an HLA ligand presenting a peptide antigen.

113. A dimeric FasL ligand comprising amino acids P132 to L279 of human FasL fused directly or indirectly through a linker at its N-terminus to a dimerized IgG-Fc region, the dimerized Fc region conjugated to nanoparticles through a Cys-containing linker, the nanoparticles further presenting HLA-peptide antigen ligands.

114. A tolerogenic ligand comprising amino acids F19-T239 of human PD-L1 used directly or through a linker to HLA-immunoglobulin fusion protein.

115. A co-stimulatory ligand comprising an anti-CD28 agonistic scFv conjugated to an HLA-immunoglobulin fusion protein. DB1 / 145613893.1 66

Images