Precise surface functionalization of PLGA particles for human T cell modulation

The biofunctionalization of synthetic materials has extensive utility for biomedical applications, but approaches to bioconjugation typically show insufficient efficiency and controllability. We recently developed an approach by building synthetic DNA scaffolds on biomaterial surfaces that enables the precise control of cargo density and ratio, thus improving the assembly and organization of functional cargos. We used this approach to show that the modulation and phenotypic adaptation of immune cells can be regulated using our precisely functionalized biomaterials. Here, we describe the three key procedures, including the fabrication of polymeric particles engrafted with short DNA scaffolds, the attachment of functional cargos with complementary DNA strands, and the surface assembly control and quantification. We also explain the critical checkpoints needed to ensure the overall quality and expected characteristics of the biological product. We provide additional experimental design considerations for modifying the approach by varying the material composition, size or cargo types. As an example, we cover the use of the protocol for human primary T cell activation and for the identification of parameters that affect ex vivo T cell manufacturing. The protocol requires users with diverse expertise ranging from synthetic materials to bioconjugation chemistry to immunology. The fabrication procedures and validation assays to design high-fidelity DNA-scaffolded biomaterials typically require 8 d. The protocol describes the fabrication of DNA scaffolds, the bioconjugation of biomolecules with complementary DNAs, conjugate assembly onto the DNA scaffolds and their immunomodulatory effect on primary human T cells in culture. Steric hindrance typically limits the use of orthogonal chemistry and covalent surface attachment strategies, whereas this DNA hybridization-based approach maintains control over the loading of each biomolecule species. The protocol describes the fabrication of DNA scaffolds, the bioconjugation of biomolecules with complementary DNAs, conjugate assembly onto the DNA scaffolds and their immunomodulatory effect on primary human T cells in culture. Steric hindrance typically limits the use of orthogonal chemistry and covalent surface attachment strategies, whereas this DNA hybridization-based approach maintains control over the loading of each biomolecule species. The assembly and organization of functional cargo using synthetic DNA scaffolds on biomaterials enable precise presentation of modulatory signals to immune cells.

• The protocol describes the fabrication of DNA scaffolds, the bioconjugation of biomolecules with complementary DNAs, conjugate assembly onto the DNA scaffolds and their immunomodulatory effect on primary human T cells in culture.
• Steric hindrance typically limits the use of orthogonal chemistry and covalent surface attachment strategies, whereas this DNA hybridization-based approach maintains control over the loading of each biomolecule species.

Introduction
Synthetic materials have been widely engineered to present biomolecules to engage cellular receptors and control cell behaviors for disease modulation [1][2][3][4] .In particular, immunotherapies show potential as treatment options for conditions including some types of cancers and autoimmune diseases [5][6][7][8][9] .In both clinical use and preclinical models, these treatments are mostly administered as in vivo immunomodulatory agents, such as antigens, antibodies and cytokines, or as cellular therapies involving ex vivo stimulation and/or engineering to control disease [10][11][12][13][14][15][16] .Immune cells in fact respond to signals from cell-cell synapses and the extracellular space to determine their phenotype, fate and behaviors [17][18][19] .Therefore, methods capable of precisely controlling the signals presented to immune cells may enable the engineering of cell therapeutic products with improved therapeutic efficacy or other benefits 3,20,21 .
Although the immobilization of stimulatory ligands on biomaterial surfaces can mimic the natural signals for immune cell programming [22][23][24] , the efficient and controllable conjugation of multiple ligands on synthetic surfaces is a major challenge of traditional chemical approaches [25][26][27][28] .Thus, we developed a synthetic short DNA-scaffold strategy for surface biofunctionalization 20 .This plug-and-play approach can precisely control the density and ratios of multiple functionalities with rapid surface assembly.This biofunctionalization approach can be used in various applications and requires the careful assembly of synthetic materials, oligonucleotides and proteins.Here, we provide the step-by-step description to fabricating DNA-scaffolded particles, engineering complementary DNA (cDNA)-conjugated biomolecules and applying these materials to activate human primary T cells ex vivo.This protocol further provides detailed methods and quality control assays to ensure a high fidelity of functional biomaterials and an optimal activation of human T cells.

Applications
We initially tested the approach to present agonistic αCD3 and αCD28 antibodies onto biodegradable polymeric microparticles composed of poly-lactic-co-glycolic acid (PLGA).These immune cell-engaging particles (ICEps) activate T cell receptor and co-stimulatory receptors for human T cell ex vivo expansion, which is a key step for manufacturing T cell-based therapies 5,29 .Due to the biodegradable and biocompatible properties of ICEps, they did not need to be removed from ex vivo cultures compared with using commercially available magnetic particles (e.g., Dynabeads) 20 .The quantitative control of αCD3 and αCD28 antibodies showed an impact on both T cell expansion fold and phenotypic outcomes-in terms of differentiation fate and exhaustion-which are critical aspects for therapeutic uses 20,[30][31][32] .In addition, these materials can be administered in vivo to control immune cell activities and can be tailored for both localized delivery-such as intratumor or subcutaneous injectionand systemic delivery through intravenous administration 20,33 .For example, logic-gated CAR-T cells have been engineered to recognize dual antigens to minimize 'off-tumor' toxicity, and we engineered microparticles presenting synthetic antigens to prime these T cells to target tumor-specific antigens 34 .With the intratumoral injection of antigen-functionalized microparticles, we were able to restrict the activation of these logic-gated CAR-T cells locally to minimize systemic toxicity 20 .While this protocol will focus primarily on the quality control of ICEp fabrication and the uses in vitro, readers are encouraged to consult the original report on this technology for additional details on in vivo use 20 .
The customizability of this approach facilitates a wide range of other applications where the precision control of multiple biomolecules is needed, for example, targeted drug delivery, gene engineering and tissue remodeling [35][36][37][38] .An effective intracellular delivery of gene regulatory-or-editing molecules must overcome various barriers at the tissue, cell and intracellular (e.g., endosomes and lysosomes) levels, which can be facilitated using different biological functionalities 1,2,39 .Similarly, the precision density control of ligands for cellular receptors involved in tissue remodeling-for example, integrin and adhesion signaling-can provide avenues for tissue engineering 35 .The approach is also adaptable for drug loading within the particle core, and polymers with different degradation profiles can be leveraged Protocol for controlled release 33,40 .Particle size can be varied across multiple length scales, enabling systemic delivery or localized retention 41,42 .Through the joint engineering of the DNA scaffold and underlying polymer, this approach can be reformulated to fit multiple biological challenges, and thus displays unprecedented levels of control for cell modulation and therapeutic applications.

Development of the protocol
This protocol describes the fabrication of (1) DNA-scaffolded PLGA particles, (2) bioconjugation of biomolecules (e.g., antibodies) with cDNA, (3) cDNA-biomolecule conjugate assembly onto DNA scaffolds and (4) primary human T cell activation and phenotyping using ICEp (Fig. 1).The high controllability of surface functionalization requires a dense layer of DNA scaffolds built on the particle surface, which depends on the efficient conjugation of the PLGA-poly(ethyleneglycol)-maleimide (PLGA-PEG-mal) with thiolated DNA (thiol-DNA) and is susceptible to poor reagent quality and improper reaction conditions (Fig. 2).Thus, we developed a framework for testing PLGA-PEG-DNA conjugation efficiency among different lots of precursor materials and correlated this with the DNA-scaffold density of the resultant particles.After validating successful polymer conjugation, PLGA-PEG-DNA batches bearing different DNA sequences can be mixed at select ratios, which will reflect the final DNA-scaffold ratio on particles.
Generating cDNA-biomolecule conjugates requires careful design to preserve the activity of the biomolecule during conjugation and surface attachment [43][44][45] (Fig. 3).For example, in antibody conjugation, tris(2-carboxyethyl)phosphine hydrochloride (TCEP) is used to selectively reduce hinge-region disulfide bonds to free thiol groups for thiol-DNA conjugation 46 .The TCEP molar excess and the reaction duration are important parameters for maintaining antibody function 47 .After DNA conjugation, a critical concern is the removal of unreacted DNA, which can compete for surface loading in later steps and thus requires affinity-based chromatography methods for purification due to electrostatic interactions between DNA and antibody.After the rapid surface assembly of purified antibody-DNA (Ab-DNA) conjugates, a flow cytometry-based method is provided to quantify the particle surface loading (Fig. 4).
When using ICEps for T cell activation, we found that culture seeding conditions, including the cell density, particle-to-cell ratio and surface ratio of stimulatory biomolecules all influence T cell expansion and the resultant phenotype (Fig. 5).For example, in our original report, we enriched either memory or effector T cell fates through the control of particle compositions

Protocol
including the ratiometric control of agonistic αCD3 and αCD28 antibodies on the particle surfaces 20 .Here, we intend to highlight the influence of these parameters on T cell activation and manufacturing so that they can be taken into consideration for related research.While the focus is on using 2 µm (mean diameter) ICEps for T cell activation, this fabrication protocol is compatible with multiple particle size scales; thus, we have provided protocol modifications throughout.

Comparison with other methods
An often-used approach for surface functionalization is through covalent conjugation between functional groups on the synthetic material and the biomolecule using a bifunctional linker

Protocol
(e.g., PEG linker with N-hydroxysuccinimide (NHS) and mal groups at the end sites) 26,27,48 .However, the efficiency of this method is severely limited by surface steric hinderance and the instability of the functional groups [49][50][51][52][53] .While orthogonal chemistries provide an additional dimension of control for immobilizing multiple biomolecules species, they still suffer from the same limitations inherent to covalent surface attachment strategies 49,50 .Further, it becomes increasingly difficult to tune the surface stoichiometry of multiple biomolecule species as characteristics of the biomolecule heavily influence their attachment-including molecular weight and charge 20,54 .In comparison, our DNA hybridization-based approach reaches the theoretical surface saturation limit while simultaneously maintaining independent control over the loading of each biomolecule species.Another surface functionalization approach to load multiple cargos is to use streptavidin handles 21,22 .We previously found that the ratiometric control is largely affected by the molecular weight and charge of the cargo, where the species with highest surface affinity always outcompeted the others.Also, the maximal density of smaller molecules may be bottlenecked by the size of streptavidin.

Limitations
There are three areas of limitations in adapting DNA-scaffolded materials: (1) the broad skillset and equipment required to combine synthetic materials, bioconjugation methods and biological applications; (2) the variations in precursor material quality; and (3) the many steps involved throughout the whole protocol.We have adapted existing technologies commonly available in biological research laboratories for characterizing fabricated materials (e.g., gel electrophoresis, Nanodrop spectrophotometers, flow cytometers, etc.).Most polymers, linkers and synthetic DNAs are commercially sourced to facilitate user adoption.For precursor quality, we have identified that PLGA-PEG-mal was the main source of quality variation, possibly due to reactant impurities remaining in the purchased polymer; therefore, we have included methods for evaluating this precursor quality.

Particle size
This protocol can be adapted for fabricating spherical particles across varying size scales while maintaining the functionality of the DNA approach (Fig. 2g,h).Here, PLGA-PEG-DNA serves as the sole surfactant, which correlates well with particle size control.Different quantification methods are needed for size quantification; micron-scale requires microscopy whereas nano-scale requires either Zetasizer or Nanosight.The protocol exhibits minimal batch-tobatch variation, although large particle sizes are associated with greater size distribution  AF488, full: 20 nM/ OD 550 , 1/2: 10 nM/OD 550 , 1/4: 5 nM/OD 550 , 1/8: 2.5 nM/OD 550 ).A linear trend was determined using a one-way ANOVA (F 1,15 = 3,944, P < 0.0001) and inter-Ab-DNA input P values were determined by one-way ANOVA (F 4,15 = 1,151, P < 0.0001) followed by Tukey's post hoc test.d, A schematic of the ratiometric control of surface cargos by the input ratio of Ab-DNA cargos at the hybridization-based assembly step.e, Flow cytometry-based quantification of particles (R:G of 1:1) that are hybridized with different ratios of Ab-DNA (αCD28-compR-AF488 and αCD3-compG-AF647) with a constant total amount of 20 nM/OD.Data are normalized MFI to the maximal loading capacity from particles with only one sequence of scaffold (R:G of 1:0 or R:G of 0:1).P values were determined by multiple two-tailed paired t-tests.f, A schematic of the ratiometric control of surface cargos by the DNA-scaffold ratio of different sequences.g, Flow cytometry histograms of particles fabricated with varying ratios of DNA scaffolds (R:G of 9:1, R:G of 1:1, R:G of 1:9) and hybridized with equal input amount of cargos (αCD28-compR-AF488 and αCD3-compG-AF647) in excess.h, Normalized MFI of histograms in g to the maximal loading capacity from particles with only one sequence of scaffold (R:G of 1:0 or R:G of 0:1).P values were determined by multiple two-tailed paired t-tests.Data in c,e and h represent mean ± s.d. of n = 4 experimental replicates from two independent experiments.

Protocol
variance, which has been reported with probe-sonication methods 55 .Thus, alternative methods for better size control could be evaluated for compatibility with the DNA-scaffolding method, including postfabrication size filtration, differential centrifugation or even alternatives to probe sonication such as microfluidic droplet generators or electrospray fabrication.

Surface density control
There are two methods for controlling the surface density of biomolecules: varying the DNA-scaffold density during particle fabrication or limiting the input quantity of

Protocol
cDNA-biomolecule conjugates during hybridization 20 .The first method was demonstrated previously by varying the molar excess of thiol-DNA to PLGA-PEG-mal during polymer-DNA conjugation, while keeping the input amount of PLGA-PEG-mal constant for particle fabrication 20 .The second method of density control involves titrating the cDNA-biomolecule below the surface saturation level (Fig. 4a-c), which is more convenient as it shares the same particle formulation and is used within this protocol.

Surface ratiometric control
Ratiometric control of biomolecules is achieved similarly to density control during either particle fabrication or surface hybridization.The surface ratios of scaffold DNA sequences are controlled by the input mixtures of PLGA-PEG-DNAs during particle fabrication.Thus, the addition of excess cDNA-biomolecules will present the biomolecules in a ratio defined by the scaffold DNA ratio (Fig. 4f-h).By contrast, the hybridization method involves inputting a predefined ratio of cDNA-biomolecules below the saturation level of each respective scaffold-DNA sequence, allowing the input biomolecule stoichiometry to define the surface ratio outcome (Fig. 4d,e).

Particle core loading
While not described within the procedural section, an additional functionality of our material is the capacity for core-loading biomolecules and tracking dyes.Fluorescent dye can be preconjugated to PLGA (e.g., AlexaFluors) and mixed during particle fabrication for in vitro or in vivo tracking.Biomolecules of interest can be loaded into the core for slow release via a double-emulsion procedure 40,56 .

Particle biodegradability
We have adopted PLGA due to its biocompatibility and tunable degradation, as degradation rates can be controlled by varying the chain lengths or lactic-to-glycolic acid ratios 40,56 .Different polymers with varying stabilities can alter the release rates of core-loaded biomolecules 57,58 ; we have shown that other polymers, such as poly-lactic acid, are also compatible with the DNA-approach technology but requires additional optimization.

Protein-DNA conjugation
There are many protein bioconjugation chemistries available, which should be balanced with conjugation efficiency, cost and maintenance of biomolecule activity 43,44,48,59,60 .Alternatively, a protein tag (e.g., SNAP-tag) can be incorporated at an optimal site of the protein to link with the functional group of the DNA 61,62 .To note, it is necessary to validate protein bioactivity post-conjugation through assays relevant to the biological function.

Protein-DNA storage
The purification procedure for removing unreacted DNA typically results in low Ab-DNA concentration, which reduces its stability.Further, long-term storage in solution is not advisable due to the risk of protein degradation [63][64][65] .Lyophilization has been used to improve long-term protein storage and is also used here to facilitate increased protein concentrations after resuspension-this can improve stability and minimize particle hybridization volumes as described later in 'Particle surface loading of antibody'.Biomolecules that are unstable or sensitive to freezing will require protein-specific bioactivity assays to verify minimal bioactivity loss and to decide whether lyophilization is appropriate.Previously, spin-concentrator columns were used to increase protein concentration, but this resulted in substantial protein loss onto the concentrator membrane, and this was more apparent when using DNAs labeled with charged fluorescent dyes.

T cell sourcing and expansion using ICEp
Peripheral blood mononuclear cells can be isolated from whole-blood or leukapheresis products and can be used without further purification or processed in a variety of ways to collect desired T cell fractions 29,66,67 .Cells can be separated on a variety of markers using Protocol commercially available positive or negative selection binding kits.To further enhance population purity and/or collect T cell subsets, such as regulatory T cells or naive T cells, FACS can be used.Isolated cells can be stimulated using a combination of T cell receptor and co-stimulatory activating proteins and growth factors.The former is provided via ICEps presenting agonistic αCD3 and αCD28 antibodies, while mitogenic cytokines (e.g., IL-2) are provided as soluble supplementation in the medium.For the latter, while we are providing cytokine in the media, we and others have identified advantages for surface delivery of growth factors, which is compatible with ICEp technology 20,22,68 .Various cell-culture parameters using ICEp can influence overall expansion and should be optimized for each cell type and experimental timeline, including: (1) choice of the culture plate, (2) cell seeding and maintenance densities, (3) cytokine concentrations and ( 4) the particle-to-cell ratio 69 .Following expansion, T cells may be analyzed using flow cytometry.

Critical controls
There are numerous controls that are important in (1) determining PLGA and DNA quality, (2) surface loading of biomolecules onto the PLGA-DNA scaffold and (3) biomolecule activity after DNA conjugation.Determining the quality of the precursors for PLGA-PEG-DNA fabrication requires gel electrophoresis; thus, we suggest using commercially synthesized oligos to serve as an unreacted DNA band control.This serves to identify the unreacted DNA fraction within the PLGA-PEG-DNA lanes, enabling the calculation of DNA consumption during conjugation, which is used as a proxy for PLGA conjugation efficiency.
For quantifying particle biomolecule loading, it is necessary to have a fluorescent standard ladder when using a plate spectrophotometer or, when using flow cytometry, have both unhybridized and saturated single-color particle controls.The fluorescent biomolecule used in either case should match the biomolecule hybridized onto particles.For flow cytometry, batch-to-batch variation in particle size could result in dissimilar fluorescence intensities, thus control and experimental particles should come from the same common stock.
For the biological activity of Ab-DNA conjugates, cell-staining titrations should be compared with unmodified antibody controls and measured via flow cytometry to detect changes over time or between conjugation batches.To minimize variation, a large batch of Ab-DNA should be aliquoted and either frozen or lyophilized immediately after conjugation.Smaller aliquots from this stock could serve as standards when comparing with new conjugations.Similarly, when loading particle with biomolecule-DNA conjugates, it may be beneficial to hybridize a large batch of particles and lyophilize them in aliquots for each future experiment.

Biological materials
• Peripheral blood mononuclear cells are isolated from leukapheresis products collected from healthy donors (StemCell Technologies) ▲ CAuTIon For working with primary human blood products, the appropriate approvals, trainings,and safety procedures should be followed according to institutional guidelines.

Protocol
Freezing medium, 10% (vol/vol) DMSO in FBS 1.In a sterile BSC, combine 22.5 mL of heat-inactivated FBS and 2.5 mL of DMSO. 2. Sterile-filter using a 0.22 µM filter and store at 4 °C.

PLGA-PEG-DNA conjugate synthesis
• TIMInG 2 d ▲ CrITICAL The following describes the synthesis of 500 nmol PLGA-PEG-DNA using commercially synthesized PLGA-PEG-mal and thiol-DNA precursors.For validated DNA-sequence options, see Table 1.Repeat the procedure for each desired oligo sequence.1. Use a micropipette to transfer 500 nmol of thiol-DNA into a 1.5 mL microcentrifuge tube (DNA tube).2. Add 100 µL of 500 mM TCEP (100× molar excess to thiol-DNA) to reduce any interstrand disulfide bonds and incubate for 1.5 h at 37 °C.3. Prepare a Glen size-exclusion desalting column that is appropriately sized for the DNA-tube volume using 10 mM EDTA in 10 mM Tris-HCl (1× TE, pH 7.5) for buffer exchange washes, per the manufacturer's instructions.4. Buffer exchange the TCEP-reduced thiol-DNA into 1× TE (pH 7.5) using the prepared Glen column to collect the DNA-containing flow-through.▲ CrITICAL STEP The exchange buffer should not contain any chemical groups that react with the selected conjugation chemistry.EDTA prevents disulfide reformation following reduction.5.For DNA precipitation, aliquot the thiol-DNA into 1.5 mL tubes (precipitation tubes) with ~400 µL per tube.To each 400 µL tube, add 50 µL of 3 M sodium acetate (pH 5.0) and 1.3 mL of ethanol (200 proof); thoroughly mix and vortex after each addition.Cool tubes at −20°C for 30 min.6. Centrifuge the precipitation tubes at 18,000g for 10 min at 4 °C.Remove the supernatant and either air dry or use a pressurized air line to further dry the DNA pellet.7. Resuspend the DNA pellet in one precipitation tube with 200 µL of TE.Combine this volume into another precipitation tube and repeat until all tubes are resuspended in a total of 200 µL (targeting ~2.5 mM DNA if DNA loss was minimal during the preceding steps).▲ CrITICAL STEP DNA should be resuspended in less than 200 µL to be compatible with the optimized reaction conditions later.Adjust this volume appropriately and reoptimize if needed.8. Measure the absorbance at 260 nm (A 260 ) of a diluted sample of DNA using a Nanodrop.
Reference Table 1 for the relevant extinction coefficients and calculate the stock concentration using Beer's law: where Nanodrop path length is 1 cm and the extinction coefficients used here are in M/cm.    2 for the necessary equations and constants for constructing the template.An example template is provided in Supplementary Table 1.▲ CrITICAL STEP The PLGA-PEG-mal:DNA ratio should be optimized for each new polymer lot.▲ CrITICAL STEP Use the PLGA-PEG-mal molecular number average instead of weight average due to the distribution of different polymer chain lengths.The number average here is specific to our PLGA-PEG-mal lot.10.Allow the PLGA-PEG-mal container to warm to room temperature before opening.
▲ CrITICAL STEP Allowing the container to warm to room temperature before opening to avoid water condensation, which can hydrolyze the functional group.11.Weigh the calculated amount of PLGA-PEG-mal and add DMF to achieve a 30 mg/mL solution.12. Add the solutions to a 15 mL tube in the following order, referring to the volumes in the reaction template: (1) extra TE buffer, (2) DNA solution, (3) triethylamine, (4) extra DMF and ( 5) PLGA-PEG-mal DMF solution.Vortex to mix. 13.Wrap the top of the tube with parafilm and shake overnight using an orbital shaker at room temperature.14.Use nitrogen or other inert gas line to back-fill the stock container of PLGA-PEG-mal.15.Wrap the container with parafilm before putting back into −20 °C storage.16.The next day, briefly vortex the PLGA-PEG-DNA reaction tube and aliquot into 1.5 mL tubes with ~500 µL into each tube.As ratiometric particles may be desired, it is recommended to premix PLGA-PEG-DNA bearing different sequences at a specified ratio before drying, ensuring that 100 nmol of total PLGA-PEG-DNA is aliquoted per tube.▲ CrITICAL STEP The downstream fabrication protocol uses 100 nmol of PLGA-PEG-DNA; thus, aliquoting 500 µL equates to a theoretical 100 nmol of PLGA-PEG-DNA (assuming 200 µM was the target PLGA-PEG-mal reaction concentration).Premixing the different PLGA-PEG-DNA sequences before drying ensures more precise control over the mixture ratio, whereas later the volumes may be difficult to control due to solvent evaporation.17.Dry the PLGA-PEG-DNA aliquots in a vacuum centrifuge at 70 °C for 2-3 h.18.Once dried, store at −20 °C.
■ PAuSE PoInT Dried PLGA-PEG-DNAs are stable for over a year.PLGA-PEG-DNA can be stable if dissolved in organic solvent, although any aqueous solutions should be avoided as this will lead to hydrolysis of either the PLGA ester linkages or the thiol-mal bond.19.Urea-polyacrylamide gel electrophoresis (PAGE) is used to verify PLGA-PEG-DNA conjugation (Fig. 2b,c).20.Prepare ~20 µL of a 0.2 µM solution of PLGA-PEG-DNA (diluted in 1× TE) and dilute to 0.1 µM using 20 µL of 2× urea-PAGE loading buffer.Similarly, make a 0.1 µM dilution of pure DNA (in loading buffer) used for the reactions.22.Heat the sample for 3 min at 70 °C.23.During heating, prepare a urea-PAGE gel by loading a vertical gel chamber with 1× TBE buffer and pre-running the gel for 10 min at 120 V. Use a syringe or pipette to clean the melted gels in each lane using TBE buffer within the chamber.24.Load 1 pmol (~10 µL) of 0.1 µM sample in triplicate alongside 1 pmol of control pure DNA lanes.Run the gel for 1.5 h at 120 V. 25.Prepare a 25 mL of 1× Sybr Gold (10,000× dilution) in 1× TBE. 26.Dispense into a wide disposable glass dish, cover the dish with the lid and protect from light.27.After the gel has finished running, release the gel from the cast and transfer to the 1× Sybr Gold solution.28.Place onto an orbital shaker at room temperature for 5-10 min protected from light.29.Rinse the stained gel with 1× TBE and transfer into a new glass dish containing buffer to prevent gel dehydration.30.Image the gel using a gel-doc reader or laser scanner.31.Import the gel image into ImageJ.After adjusting brightness and contrast, perform gel densitometry analysis as described by the ImageJ operational manual (see 'Software') 70 .32. Use the intensity of the top PLGA-PEG-DNA band and the lower, unreacted DNA to calculate the efficiency of the reaction using the equation below and record to track batch variation.▲ CrITICAL STEP Disulfide bonds can form between the thiol-DNA and can appear in the gel above the unreacted thiol-DNA (Fig. 2b).We typically do not include the disulfide band intensity since it is negligible relative to the main unreacted thiol-DNA band.

PLGA particle fabrication
• TIMInG 6 h ▲ CrITICAL This procedure describes the fabrication of 2 µm particles bearing a maximally dense surface DNA scaffold at 1:1, R:G DNA sequence ratios (for sequence information, see Table 1), where R and G are different DNA sequences.This procedure assumes that 100 nmol of PLGA-PEG-DNA was dried in Step 18 with a 1:1 mixture of DNA G and R sequences (PLGA-PEG-G and PLGA-PEG-R, respectively).100 nmol of PLGA-PEG-DNA generates ~100 OD 550 in 400 µL volume (40 OD 550 in 1 mL) or ~2 × 10 9 particles.For fabricating particles of other target diameters, refer to Table 3 for modifying reagent amounts within this section and to the 'Anticipated results' section for representative morphologies and size distributions (Fig. 2g,h  Protocol 37. Resuspend the 1:1 R:G PLGA-PEG-DNA tube from Step 18 with 100 µL of water and 100 µL of EtOAc.Reuse this pipette tip whenever transferring PLGA-PEG-DNA for a given sequence ratio (switch if using a different sequence ratio).38.Place the PLGA-PEG-DNA tube into the bath sonicator for 10 min or until fully resuspended.39.Transfer the PLGA-PEG-DNA into the 15 mL fabrication tube in 100 µL increments to reduce material loss inside the pipette tip.40.To wash the PLGA-PEG-DNA tube, add 300 µL of water and 100 µL of 5× particle fabrication buffer (see 'Reagent setup').41.Using the saved PLGA-PEG-DNA pipette tip, transfer this solution into the fabrication tube.
If the pipette tip gets clogged, briefly pipette the EtOAc fraction within the fabrication tube to dissolve the clog.42.Sonicate the fabrication tube and vortex until mixed.Place the fabrication tube on ice.43.Place a magnetic stir plate with a 250 mL beaker and a stir magnet into a fume hood.This will be needed after probe sonication after Step 50.44.Prepare a 50 mL conical tube partially filled with ice to act as a secondary container for the fabrication tube during probe sonication.Set up a vortexer, 0.2% (wt/vol in water) PVA and separate ice container near the probe sonicator.45.For the sonication setup, clean the sonication microtip probe using 70% (vol/vol in water) ethanol and allow to dry.▲ CrITICAL STEP Ensure that the sonication program is set to the recommended settings (Box 1).Sonication will need to pause halfway through, so if your sonicator does not allow for this function then adjust the number of cycles accordingly.46.Vortex the reaction tube and place into the 50 mL secondary ice container.47.Position the sonication probe into the fabrication tube solution, avoiding the tube walls.48.Initiate the sonication program, moving the microtip throughout the solution to ensure a more homogeneous sonication.After two cycles, pause sonication and vortex the reaction tube before finishing the remaining cycles.49.Immediately after sonication add 9 mL of 0.2% (wt/vol) PVA into the fabrication tube, invert to mix, then vortex.50.Dispense the contents of fabrication tube into the 250 mL beaker from Step 43 and turn on the magnetic stirrer for ~2.5 h without any heating.▲ CrITICAL STEP This step will evaporate the EtOAc residue.For larger volumes, use a rotary evaporator.◆ TroubLESHooTInG 51.After 2.5 h, place a 40 µm filter onto a 50 mL conical tube and pour the particle solution through the filter.Use a micropipette to transfer any remaining solution.52.Centrifuge the particle tubes at 225g for 10 min.▲ CrITICAL STEP If nanoparticles were fabricated, then after Step 52 the supernatant will contain the nanoparticles while any large particle contaminants will be contained within the pellet.If larger microparticles (>2 µm) were fabricated, then proceed as written without protocol modification.

Sonication settings for particle fabrication
The following sonication settings were chosen for the S-4000 probe sonicator (Qsonica).The settings should be adjusted for other sonication systems and yielded particles should be quality checked to match the characteristics described within this protocol.1.Total energy: 230-250 J 2. Amplitude: 30 3. Pulse sequence timing: 5 s on, 10 s off 4. Total sonication time: 25 s (5 total pulse sequences) Protocol 53.Discard the supernatant and resuspend in 2 mL of TE containing 0.1% (vol/vol) Tween 20 (TE-Tween) using a micropipette.▲ CrITICAL STEP If nanoparticles were fabricated, then collect the supernatant and discard any visible pellet after Step 52.For all subsequent nanoparticle centrifugation steps in this protocol, spin at 16,000g for 10 min.54.Distribute the 2 mL into smaller microcentrifuge tubes and centrifuge at 6,000g for 5 min.55.Resuspend each tube in 200 µL of TE-Tween.56.Spin again at 6,000g for 5 min, resuspending again in 200 µL TE-Tween.During the final resuspension, combine all tubes into a single tube with a total volume of ~400 µL TE-Tween.57.Prepare small sample for Nanodrop quantification.Since the stock concentration is large, use a larger dilution volume to allow for sufficiently large pipetting volumes from the stock solution (~0.5-1 µL).Assuming a successful fabrication yield of ~100 OD 550 in 400 µL, use the dilution example below to generate a dilution of ~0.5 OD 550 : (A) Generalized dilution equation used: B) (100 OD 550 stock) × (X µL stock sampled) = (0.5 OD 550 target concentration) × (100 µL total dilution volume); X µL stock sampled = 0.5 µL (C) (Total dilution volume) -(X µL stock sampled) = (volume of TE-Tween to dilute stock sample); volume of TE-Tween to dilute stock sample = 99.5 µL (D) Dilution factor = (total dilution volume)/(volume of stock sampled); dilution factor = 200 ▲ CrITICAL STEP Nanodrop particle absorbance is linear between 0.2 and 1.0 at OD 550 , so the estimated dilution fold would need to be adjusted accordingly.The estimate of 0.5 used above is an appropriate initial target as some amount of error will probably maintain the measured range between 0.2 and 1.0.▲ CrITICAL STEP Microparticles settle quickly, creating a concentration gradient and a particle pellet over time.Whenever handling microparticles, ensure the tubes are sufficiently resuspended.58.After using an appropriate buffer (TE-Tween) to blank the Nanodrop, measure the OD 550 of the diluted sample and solve for the stock concentration via the equation below and using the dilution factor calculated in Step 57.If the measured OD 550 is below 0.2, then remake the dilution using a lower dilution factor.(A) (Stock OD 550 ) = (dilution factor) × (measured OD 550 of diluted sample) 59.Using the equation from Step 57, set aside a small sample of diluted particles to generate ~20 µL at 5-10 OD 550 .Save this sample for imaging and size quantification later.▲ CrITICAL STEP For nanoparticles, refer to Step 99 for the necessary sample amount and the dilution concentration for Zetasizer measurements.60.To each ~400 µL tube of particles, add 100 µL of 5% (wt/vol) PVA and mix.61.In a secondary container, prepare a small volume of liquid nitrogen.62. Flash freeze the tubes by submerging in liquid nitrogen below the cap level using a tube holder (e.g., long forceps).

Place the frozen tubes into a lyophilization chamber for 24 h with the tube caps open.
■ PAuSE PoInT Lyophilized particles can be stored for up to 2 years at −20 °C.Particles stored after 2 years should be reassessed for DNA-scaffold density (see 'Particle surface DNA loading analysis')

Particle surface DNA loading analysis
• TIMInG 4 h ▲ CrITICAL This protocol describes the quantification of particle scaffold DNA density and relative ratio of DNA sequences via the detection of hybridized, fluorescently labeled cDNA (5′ end label) using a plate spectrophotometer.The procedure assumes particles are taken from lyophilized stock.The total particle amount required for fluorescent detection varies depending on the particle size since each formulation has a different nM/OD 550 loading capacity.Thus, the fluorescence detection limit of the spectrophotometer should be used to predict the amount of particles needed to adapt this method for other particle sizes.

Remove the lyophilized particles from
Step 63 onto a disposable weigh boat.65.Use a razor blade to cut a small fraction of the particle for OD 550 measurement.Target a concentration of 20 OD 550 in 100 µL and readjust later after OD 550 quantification is made.▲ CrITICAL STEP 20 OD 550 in 100 µL was chosen to ensure that the particle signal will be above the signal detection limit for our spectrophotometer.Additionally, if users are not careful during pipetting steps there could be substantial particle loss, which is mitigated by increasing the initial particle quantity.66. Resuspend the particle sample in 500 µL of water for 5 min.67.Centrifuge the particles at 6,000g for 5 min.
▲ CrITICAL STEP Since the particles were lyophilized in a nonvolatile buffer, the buffer salts are still contained in the pellet.Water should be used to resuspend to prevent high concentrations of buffer salts.68.Remove the supernatant and wash with 500 µL of TE-Tween.69.Repeat spinning and washing one more time with the final resuspension in 100 µL of TE-Tween.70.Make a sample dilution in a separate tube.71.Measure the diluted sample OD 550 .Use the OD 550 to calculate the stock tube concentration.▲ CrITICAL STEP If there is less than 20 OD 550 in 100 µL, repeat Steps 64-67 to resuspend a newly cut portion of the particle as described previously and add to the existing particle volume after sufficient wash steps described in this step.Repeat Steps 70-71.72.Calculate the hybridization component volumes according to Table 4, assuming a particle concentration target of 20 OD 550 in 100 µL total hybridization volume.▲ CrITICAL STEP The total loading capacity of cDNA of high-scaffold density 2 µm particles approaches 75-150 nM of cDNA per OD 550 depending on batch-to-batch variation 20 .For the 1:1 R:G particle here, both compR and compG DNAs will maximally load between 37.5 and 75 nM/OD 550 , respectively.cDNA should be loaded at three times the maximal theoretical loading capacity (~225 nM/OD 550 for each cDNA on the 1:1 particle) to ensure surface saturation regardless of particle batch variability.▲ CrITICAL STEP 200 nm particle loading approaches 1,000-2,000 nM of cDNA per OD 550 , whereas the 8 µm particle loading approaches 10-20 nM of cDNA per OD 550 , depending on the batch variation.The loading capacity should be determined for different particle sizes before experimental use.This should be adjusted in Table 4 for calculating hybridization reaction conditions depending on the particle size used.

Protocol
▲ CrITICAL STEP The hybridization buffer will constitute half of the total volume.The remaining half will be used for particle volume and cDNA.If 100 µL has not been reached, calculate the volume for TE-Tween to fill the remainder.The total hybridization volume may exceed the target volume depending on the concentration of reagents, so extra TE-Tween may not be required (seen as a negative or zero value for the extra TE-Tween calculation).73.Transfer a quantity of particles into a microcentrifuge tube such that, once diluted, it will result in 20 OD 550 in 100 µL (hybridization tube).If the volume of particles needed in Step 72 exceeds 50 µL, then centrifuge particles and remove supernatant until 50 µL of volume remains.74.To the hybridization tube, add 50 µL 2× hybridization buffer, cDNAs and extra TE-Tween (if needed).75.Mix the solution using a micropipette followed by bath sonication for 15 s to ensure particle dispersion.76.Incubate particles on a shaker for 30 min at 37 °C.▲ CrITICAL STEP Particle hybridization is achieved in less than 2 min, although to ensure surface saturation we hybridized for 30 min.During this time, settling occurs at high particle concentrations, which is more apparent when using larger-diameter microparticles.If this is substantial, vortex the particles halfway through their incubation period.▲ CrITICAL STEP Particles will be loaded onto the plate in PBS-DMSO, so the ladder should be made in the same buffer.78.After hybridization, add 400 µL of TE-Tween and centrifuge at 6,000g for 5 min at 4 °C.79.Remove supernatant and wash twice more.80. Use 120 µL TE-Tween for the final resuspension.▲ CrITICAL STEP After particles have been hybridized, all centrifugation steps should occur at 4 °C to minimize dehybridization of loaded cargos.▲ CrITICAL STEP It is important to remove the majority of the supernatant to prevent background signal.It is additionally important to not disturb the pellet during any steps, as this will reduce the total signal detected during later steps.81.Add 50 µL of hybridized particles (particle replicate tubes) into two separate centrifuge tubes-these will be used for repeated measures.82.With the remaining 20 µL volume, dilute a small volume for OD 550 calculation to determine the concentration in the particle replicate tubes.This value will be needed to calculate the final DNA nM/OD 550 .83. Centrifuge replicate particle tubes at 6,000g for 5 min and remove 45 µL of supernatant from each.84.Add 45 µL of DMSO to each particle tube to dissolve particles.▲ CrITICAL STEP 5 µL of wash buffer should be remaining after supernatant removal to reduce particle loss.If previous wash steps were not thorough, the 5 µL of remaining supernatant could include background DNA signal.The 45 µL of removed supernatant can be saved and measured to determine the background fluorescence contribution.85.For replicate measurements, add 90 µL of PBS into the microwell plate from Step 77 and 10 µL of dissolved particles.86.Resuspend all wells thoroughly and do not generate bubbles.Protocol 87.Read the fluorescence of the microplate on a microplate spectrophotometer in top-down mode with settings in accordance with the respective fluorophores used.▲ CrITICAL STEP Filters should be carefully selected to minimize signal crossover between fluorophores.Other settings, such as channel voltage, should be optimized for each machine.88.For fluorescence analysis, average the blank PBS-DMSO wells and subtract from all wells.
Create a linear best-fit curve for the fluorescent ladder lanes.▲ CrITICAL STEP Since the ladder fluorescent signal could be widely different than the measured particle signal, ensure that the ladder range used for generating the best-fit curve are within one-to-two dilution steps away from the measured particle signal to increase accuracy.89.Calculate the fluorophore concentration of each well using the best-fit curve above.Correct for sample dilution by dividing each well fluorescence concentration by 1/10 of the OD 550 value determined in Step 82 to determine stock nM/OD 550 .▲ CrITICAL STEP Particles were diluted tenfold in Step 85.This factor needs to be corrected for the OD 550 in the plate.90.Average the nM/OD 550 values from each well and report as the mean ± s.e.m.Calculate the surface ratio between R:G signals using the equation below.The ratio of cDNAs is reflective of the ratio of the scaffold DNAs:

Particle size quantification
• TIMInG 2 h ▲ CrITICAL Microparticle size distributions are assessed using confocal microscopy imaging (option A).While brightfield requires less material preparation, confocal imaging of fluorescent particles produces defined silhouettes and reduces off-target quantification of debris; thus, confocal imaging is recommended for accurate size quantification.The selected magnification should be used to provide a sufficient field-of-view to capture a large number of particles, while still maintaining visualization of small-diameter particles.Since nanoparticle fabrication may be of interest, we suggest the use of dynamic light scattering instruments such as Zetasizer (option B).Since Zetasizer does not rely on fluorescence measurements, unlike the confocal microscopy method, nanoparticles do not need to be hybridized with fluorescent cDNA and can be analyzed immediately after Step 56.
Option A (microparticle size quantification using confocal microscopy): 91.Particles must first be hybridized using saturating levels of fluorescent cDNA as described in Steps 72-76 and 78-80.A small amount of particles are needed for imaging (~5-10 OD 550 in 30 µL), so adjust starting particle amount to minimize particle waste.92.Pipette 10 µL of diluted, fluorescent particle (target ~5-10 OD 550 ) onto a clear microscope slide and overlay a coverslip.93.Seal the coverslip corners with clear nail polish.94.After the corners have partially dried and flattened, seal the sides of the slips by connecting each corner with nail polish.This will prevent sample drying and allow for slide inversion on the microscope if needed.95.Visualize particles under confocal microscopy (Fig. 2g).96.Adjust laser power and exposure settings for the relevant laser line, being careful to avoid photobleaching.Height focus should be set using the fluorescence channel.97.Acquire at least five representative images.98. Analyze images using ImageJ to determine particle diameters.Size distribution curves can be generated in software such as Graphpad (Fig. 2h).
Protocol 100.Dispense an appropriate volume into a disposable cuvette and perform size analysis using the Zetasizer and the manufacturer's instructions.Intensity-weighted size distributions and other variation metrics, such as the average diameter or polydispersity index (PDI), can be exported and visualized within software such as GraphPad (Fig. 2h).

Antibody conjugation with complementary DNA
• TIMInG 1 d ▲ CrITICAL This procedure describes the conjugation of antibodies with amine-labeled cDNA using an NHS-PEG-mal linker at a 2 mg antibody scale.This protocol does not change whether the DNA is labeled, but for most applications we recommend a dyeless DNA.If a dye-labeled DNA is used, special attention should be placed to the charge of the dye; we have found that positively charged dyes may have increased association with the antibody and thus leads to purification difficulties.Ab-DNA can be labeled for quantification purposes after purification if required (see 'Preparation of antibodies for surface loading quantification').101.Calculate the volume needed for 2 mg of antibody and prepare a Glen size-exclusion column that is appropriately sized for the antibody volume, as per the manufacturer's instructions.Buffer exchange washes should be 10 mM EDTA in 1× PBS, Ca 2+ /Mg 2+ free (PBS-EDTA).▲ CrITICAL STEP Ensure that the buffer does not contain any amine-groups (e.g., Tris) as this will compete to react with NHS reagent used later.102.Buffer exchange the antibody into PBS-EDTA per Glen column manufacturer's instructions and collect into a new tube (reaction tube).103.Measure the antibody A 280 using a Nanodrop with an appropriate dilution.Place the antibody at 4 °C.The protein concentration can be calculated using the following equations: • (A 280 × dilution)/1.33= (mg/mL antibody) • ((mg/mL antibody) × 1,000)/155 = nmol antibody, where 155 is the antibody molecular weight (kDa).104.Calculate volume of amine-cDNA needed for 4× molar excess relative to antibody in Step 103.Move this volume into a new tube (DNA reaction tube).The following equation can be used: • nmol DNA needed = 4 × (nmol antibody) • mL of DNA needed = (nmol DNA needed)/(µM DNA stock) ▲ CrITICAL STEP Here, a subsaturating amount of DNA-as determined using SDS-PAGE immediately after DNA conjugation without purification-was used to prioritize Ab-DNA purity over conjugation efficiency (Fig. 3b,c).Higher amounts of DNA could be used to improve the Ab-DNA yield as long as the removal of unreacted DNA is confirmed.Importantly, the ratio of DNA to biomolecule should be optimized for every new biomolecule and linker.105.Calculate the mg of NHS-PEG-mal for 20× molar excess relative to DNA from Step 104.
Dissolve linker in a small volume of DMSO, with at least 30 µL per 0.8 mg of linker.106.Add 20× molar-excess-dissolved linker to the DNA tube and incubate for 1 h at 37 °C.If the reaction DMSO volume exceeds 5% (vol/vol), add HEPES (100 mM, pH 7.2) until 5% DMSO is reached.107.When the DNA-PEG-mal reaction from the previous step is nearly complete, dilute TCEP to 5 mM in PBS-EDTA.108.Calculate the volume of 5 mM TCEP needed for 4.5× molar excess relative to antibody amount determined in Step 103.109.Add this TCEP volume into the antibody tube and incubate for 1 h at 37 °C.110.Afterwards place the antibody at 4 °C.▲ CrITICAL STEP A lower molar excess can be used but may require longer incubation; longer timing or increased molar excess can result in different reduction cleavage products.

Protocol
111.Precipitate the DNA as described in Step 5.During the precipitation, a second Glen column should be equilibrated to PBS-EDTA.The final volume after DNA precipitation will be 200 µL, so prepare an appropriately sized Glen column.112.After 30 min at −20 °C, centrifuge the DNA reaction tube at 18,000g for 10 min at 4 °C.113.Remove the supernatant and resuspend in 200 µL PBS-EDTA.114.Use the Glen column to buffer exchange to PBS-EDTA to remove any excess unreacted linker from the DNA-PEG-mal.115.Determine the DNA-PEG-mal concentration from the Nanodrop A 260 and Beer's Law.
Reference Table 1 for the relevant extinction coefficients.116.Add 4× molar excess of DNA-PEG-mal into the antibody tube and incubate for 1 h at 37 °C.
Afterwards, place the antibody reaction tube at 4 °C overnight.

▲ CrITICAL
The following steps are required for removal of free, unreacted DNA-PEG-mal from the Ab-DNA conjugate, which can compete for surface hybridization.117.Use a ring-stand clamp to suspend a resin gravity column over a liquid waste container Assemble the column by placing the column filter at the bottom end nearest the exit port and capping the bottom.Vortex a bottle of Protein G resin beads and add 1.5 mL of the bead suspension followed by a sufficient volume of 1× Protein G binding buffer (binding buffer, diluted in water) to fill the column.▲ CrITICAL STEP 1.5 mL of suspension results in ~0.75 mL column volume (CV) of resin after liquid drainage.Varied resin amounts can be used depending on the amount of protein being purified.118.Remove the column cap and allow for the buffer to drain.When ~3/4 of the column height remains, cap the bottom, and wait 20 min for the resin to settle.119.Place a second column filter into the column and push until above the binding resin.▲ CrITICAL STEP Do not trap bubbles beneath the filters as this can slow the elution of the column.120.Add 5 CVs (~3.75 mL) of binding buffer and allow to drain.121.Remove the waste container under the column and replace with a 15 mL capture conical tube.122.Remove the capture tube and replace with another 15 mL conical tube.123.Add the Ab-DNA from the first capture tube.124.Repeat twice more by loading the flow through to ensure maximum column binding.125.Discard the last flow through.126.Place a waste container underneath the filter column and wash with 10 CVs of binding buffer.127.While the column is washing, label ~13 1.5 mL tubes: 5 for the acidic elutions, 3 for the neutral and 5 for the basic.128.Add 55 µL of 10× acidic elution neutralization buffer into each acidic elution tube and 55 µL 10× of basic elution neutralization buffer into each basic elution tube.▲ CrITICAL STEP If different elution volumes are captured per tube, the volume of neutralization buffer should be adjusted to achieve a final 1× concentration.129.Add 3 CV (2.5 mL) of acidic elution buffer into the column and begin capturing 500 µL of flow through into each acidic capture tube.130. Mix each tube afterwards to ensure the neutralization buffer has mixed into the flow through.131.After all acidic buffer has passed, add 3 CV of binding buffer and capture a third of the volume into each of the neutral tubes.132.After all the binding buffer has eluted, add 3 CV of basic elution buffer and capture 500 µL of flow through into each basic capture tube.133.Mix each tube afterwards to ensure the neutralization buffer has mixed into the flow through.134.Place a waste container underneath the column and add 5-10 CV of binding buffer.Protocol 135.After draining, cap the bottom and add binding buffer to cover just above the top resin.136.Label and store at 4 °C if subsequent purifications are needed.137.Quantify the A 260 and A 280 of each elution tube using a Nanodrop (Fig. 3d).138.Dispose all tubes where the A 280 indicates minimal protein recovery (<5-10% of the original theoretical protein amount) and also dispose when the A 280 /A 260 ratio is less than 0.9.▲ CrITICAL STEP This step is the most critical for improving the purity of the final Ab-DNA.The A 280 /A 260 ratio can slightly vary, although the tubes that primarily contain the unbound DNAs should have a ratio much less than 1.0.139.Dialyze the Ab-DNA with 1× PBS using a 50 mL dialysis column (10K MWCO) and place onto an orbital shaker at 4 °C as per the manufacturer's instructions.140.Swap the 1× PBS after 2 h and 4 h cumulative time.141.After the final swap, dialyze overnight.142.The next day, collect the Ab-DNA from the dialysis column and store at 4 °C.
▲ CrITICAL STEP This step removes the glycine and other buffer components that may inhibit downstream quantifications and purifications.The glycine must be removed if additional Fc-affinity column purifications are needed, otherwise the antibody cannot bind to the resin.143.Use a microBCA kit to determine the protein concentration (in mg/mL) within the Ab-DNA conjugate according to the manufacturer's recommendations.144.The DNA concentration within the Ab-DNA is required for hybridization calculations, but this requires additional steps to calculate since both the antibody and the DNA independently contribute to both A 260 and A 280 .Refer to Box 2 to solve for the DNA concentration within the Ab-DNA.▲ CrITICAL STEP If a dye-labeled Ab-DNA was used, then the A 260 (DNA component of Ab-DNA) can be estimated on a plate spectrophotometer using a standard fluorescent curve of known cDNA-dye concentrations and comparing the fluorescence of a known dilution of Ab-DNA.145.Use urea-PAGE to confirm that free DNA has been removed from the Ab-DNA conjugate (Fig. 3e).

Box 2
Equations for determining the DNA concentration within Ab-DNA solution Protocol 146.Prepare dilutions of Ab-DNA and pure DNA, run the gel and analyze according to Steps 19-32.▲ CrITICAL STEP If a dye-less DNA was used for conjugation, urea-PAGE must be performed to later stain the DNA with Sybr Gold, which is not compatible with SDS-PAGE gels.If a dye-labeled DNA was used, then SDS-PAGE gel is recommended as the antibody bands are more clearly defined.◆ TroubLESHooTInG 147.Calculate the Ab-DNA purity with the equation below.If the sample is not pure (e.g., purity <0.95), then the purification Steps 120-145-using the column saved from Step 136-must be repeated before proceeding: • Ab-DNA purity = (intensity of Ab-DNA band)/((intensity of Ab-DNA band) + (intensity of DNA band)) 148.Prepare a sufficiently sized Glen column using 0.1 M TEAA (pH 7.0) as the exchange buffer.149.Buffer exchange the Ab-DNA into the TEAA and aliquot into separate tubes for lyophilization.150.Label the estimated protein and DNA amount in each tube to calculate the new concentrations when later resuspending.▲ CrITICAL STEP TEAA is a volatile buffer and thus does not leave salts after lyophilization, which could damage the proteins at high concentrations.151.Freeze the Ab-DNA using liquid nitrogen as described in Steps 61-63 and lyophilize overnight.152.The next day, resuspend an Ab-DNA aliquot in 0.1 µm-filtered PBS so that the concentration of the antibody is at least 6.5 µM, using the concentrations determined in Steps 143-144 to determine the new concentrations after resuspension.Store remaining aliquots at −20 °C.▲ CrITICAL STEP Higher resuspension concentrations (>1 mg/mL) are important for the stability of proteins and to have more reasonable volumes to work with during particle hybridization.To note, resuspending the Ab-DNA at too high concentrations could result in protein aggregation 71,72 .Thus, concentrations between 1 and 10 mg/mL are recommended, which is comparable to the concentrations of the purchased antibody stocks used within this protocol.■ PAuSE PoInT Lyophilized proteins are stable at −20 °C for over 2 years.The antibody can remain stable at 4 °C for over a year.The shelf-life of other proteins should be assessed and monitored.153.(Optional) Sodium azide (0.05%) can be added to a desired concentration to limit microbial growth once resuspended and stored at 4 °C.

Preparation of antibodies for surface loading quantification
• TIMInG 2 h ▲ CrITICAL Flow cytometry can be immediately used to verify the ratio between protein species on particle surfaces if the conjugated cDNA was labeled with a fluorescent dye.If unlabeled cDNA was used for conjugation, NHS-dye labeling of the antibody is first required.Below we describe the labeling of αCD28-compR.The procedure is identical for labeling αCD3-compG but with a different fluorophore.To reduce nonspecific interactions between the Ab-DNA and the particle, we recommend using negatively charged dyes for Ab-DNA labeling.154.For later quantification, record the A 260 and A 280 of the Ab-DNA.Calculate the ratios R1 = A 280 /(µM antibody) and R2 = A 260 /(µM DNA), where the respective antibody and DNA concentrations are known from Step 152 after resuspension from lyophilized stock.155.Resuspend NHS-Alexafluor-488 (AF488) to 2 mM in DMSO.156.Aliquot ~50 µg (0.32 nmol) of purified αCD28-compR (~100 µL at 0.5 mg/mL).157.Add 8× molar excess of 2 mM AF488 into the antibody aliquot and react for 1 h at 37 °C.158.Record the final reaction volume.▲ CrITICAL STEP Do not exceed 5% (vol/vol) DMSO to reduce protein denaturing.
To prevent this, either make a more concentrated stock of NHS-dye or dilute with HEPES (100 mM, pH 7.2).

Protocol
dosing.While in vitro use is described later, particles for in vivo use should be concentrated using centrifugation to a desired volume suitable for localized or systemic injections as previously demonstrated 20 .

Quantification of antibody loading onto particles using flow cytometry
• TIMInG 4-6 h ▲ CrITICAL This procedure quantifies the microparticle surface loading of αCD28-compR-AF488 and αCD3-compG-AF647 using flow cytometry.Blank and single-antibody loaded particles are made using the method described in Steps 163-169 and are used for compensation controls and downstream calculations.Use single DNA-sequence-scaffolded particles (R or G only) for single-color controls to saturate the surface with their respective antibody species.The plate spectrophotometer used in the previous section 'Particle surface DNA loading analysis' can be used as an alternative quantification tool, although this uses prohibitively more material compared with flow cytometry.170.Perform flow cytometric analysis on particles from Step 169.Reference Box 3 for performing surface-loading analysis using software such as FlowJo.We have included representative flow cytometry fluorescence histograms and calculated surface loadings for particles hybridized using a variety of loading methods as described in the section 'Experimental design' (Fig. 4).Example data from Steps 161-169 that used the 1:1 R:G surface scaffold and particle surface-saturating amount of Ab-DNA is provided (Fig. 4h).

T cell enrichment from leukapheresis products
• TIMInG 2 h ▲ CrITICAL This procedure describes the isolation of either CD4 + or CD8 + T cells from leukapheresis blood product using commercial negative selection beads.171.In a sterilized BSC, isolate CD4 + or CD8 + T cells from leukapheresis blood using the EasySep Enrichment kit per the manufacturer's instructions.Wash steps should be performed using sterile-filtered PBS-FBS wash buffer (see 'Reagent Setup').When required, cells should be centrifuged at 300g for 5 min at 4 °C.

Quantification of particle surface loading using flow cytometry
Flow cytometry should be performed including single-color and blank controls.The mean fluorescent intensities (MFIs) can be used to calculate the antibody surface occupancy from the below equations.Optional normalization to the particle DNA loading can be performed for more convenient comparisons of surface ratios between particles batches when comparing particles with varied total surface protein density.This method is useful for comparing surface DNA ratios but cannot be used to compare DNA densities between batches. 1. Surface occupancy of antibody (% αCD3) • % αCD3 = (MFI αCD3 -MFI blank particle )/(MFI single-color-control αCD3 -MFI blank particle ), where the MFI is the signal coming from the respective channel as the αCD3 antibody dye.Repeat for % αCD28 using relevant values 2. Ratio of αCD3:αCD28 • If % αCD3 > % αCD28, then the ratio of αCD3:αCD28 is (% αCD3/% αCD28):1 • If % αCD3 < % αCD28, then the ratio of αCD3:αCD28 is 1:(% αCD28/% αCD3) 3. Optional: normalization of surface occupancy • % αCD3 norm = (% αCD3)/(% αCD3 + % αCD28), repeat similar calculation for normalizing % αCD28 using relevant values Protocol 172.After cells have been enriched, spin down the cells at 300g for 5 min at 4 °C.Calculate the volume to resuspend cells between 10 and 50 × 10 6 cells/mL.Remove the supernatant and resuspend in sterile freezing medium to the desired concentration.Aliquot 1 mL of cells into each liquid nitrogen compatible freezing vial and place into a polyethylene CoolCell.
Immediately transfer the CoolCell into the −80 °C freezer overnight.Transfer freezing vials to liquid nitrogen storage the following day.
■ PAuSE PoInT T cells can be stored in liquid nitrogen for over a year and thawed when needed.

T cell expansion using ICEp
• TIMInG ~11 d ▲ CrITICAL This procedure describes CD4 + T cell culturing using ICEps, which is identical for CD8 + T cells.Cells will be expanded in a 96-well (flat-bottom) culture plate throughout, although they can be transferred to larger well-plate volumes as long as the appropriate cell concentrations are maintained.ICEps should be prepared sterilely with αCD3 and αCD28 1 d before T cell activation as described in the previous section, 'Particle surface loading of antibody'.The quantity of particles required should be determined before T cell activation to reduce material waste.Complete T cell media (media) should contain 100 U/mL hIL-2.173.Centrifuge αCD3-and αCD28-loaded ICEps at 6,000g for 5 min at 4 °C.174.In a BSC, carefully remove supernatant and resuspend to 1 OD 550 (~20 × 10 6 particles/mL) in media.OD 550 can be measured to verify desired particle concentration.▲ CrITICAL STEP We will seed 25,000 T cells per 96-well plate, so 1.25 µL of particles (at 1 OD 550 ) will eventually be added to each well for 1× particle to cell excess.Additional particle amounts can be added, although the total well volume should stay consistent between conditions.175.Warm media in a 37 °C water bath.176.Aliquot 9 mL of warmed media into a 15 mL tube.177.Remove a CD4 + enriched T cell vial from liquid nitrogen storage and thaw in the water bath.
Just before fully thawing, move the vial into the BSC.178.Gently pipette to resuspend the cell pellet and transfer the volume into the 9 mL of warmed media to dilute the DMSO.179.Spin cells at 300g for 5 min at 4 °C.180.Remove the supernatant and resuspend cells in 10 mL of media and count the cells.
Depending on the number of T cell conditions, dilute an appropriate volume of cells in media to ~0.278 × 10 6 cells/mL.181.After mixing, pipette 90 µL of cells per well in a 96-well plate.182.Thoroughly mix the ICEps without generating bubbles and add an appropriate volume to each well (1.25 µL of 1 OD 550 for 1× particle-to-cell).Occasionally resuspend stock ICEps to prevent particle settling.183.Add additional media for a total well volume of ~100 µL.The cells are now approximately at 0.25 × 10 6 cells/mL.184.Using a multichannel pipette, gently mix all wells to thoroughly distribute ICEps and cells.
Transfer the seeded culture plate into a sterile incubator set to 37 °C and 5% CO 2 .185.After 24 h (day 1) visualize the plate under a bright-field microscope to observe cell clustering and look for any signs of contamination.186.After 48 h (day 2), double the well volume using prewarmed media (~100 µL) by dispensing around the well perimeter, attempting not to disturb the cell clusters.Cells are typically not ready to be split at this day due to a freezing-related growth delay.187.On day 4, resuspend the T cell wells and take a small sample for counting.188.Calculate the volume containing 25,000 cells and reseed this volume into an unused well.
Add media for a total well volume of 100 µL.189.Track the cell expansion fold between well splitting and repeat every 2 d until growth slows or a predetermined end point has been reached (Fig. 5a,b).▲ CrITICAL STEP At this concentration, a 2 d splitting procedure lets cells expand upwards of 10-15 fold without filling the entire well.Other plating conditions requires different schedules.190.At the experiment endpoint, stain and fix cells for flow cytometry analysis (Fig. 5c-e).

Troubleshooting
Troubleshooting advice can be found in Table 5.

Ab-DNA conjugation and purification
To chemically modify antibodies with minimal activity loss, a selective reduction protocol using a precise molar excess of TCEP (4.5×) is used to generate free thiols for cDNA attachment (Fig. 3a,b).Therefore, it is important to accurately measure the antibody concentration to determine the TCEP dose.During antibody handling, Ca 2+ /Mg 2+ -free PBS buffer supplemented with 10 mM EDTA is used to keep the thiol groups from oxidizing and reforming disulfide linkages 75 .SDS-PAGE serves as a handy tool to check the extent of antibody reduction and fractionation after Ab-DNA conjugation (Fig. 3b).Linker-attached cDNA with mal functionalization (DNA-PEG-mal) is given in excess to ensure a high yield of DNA-Ab conjugates.However, the excess amount of unreacted DNA needs to be removed to avoid competition in the downstream surface hybridization step.Hence, we titrated the DNA molar excess to the antibody and found that a range of 4-6× is the minimal excess for the highest conjugation efficiency, as determined by the saturating trend for the conjugation reaction (Fig. 3c).Fc affinity-based chromatography was found to be the only method that effectively removed unreacted DNA.To ensure a high recovery of the costly antibodies and a complete removal of excess DNA, we provide guidelines to determine the appropriate elution fractions to collect from the purification (Fig. 3d) and to check for residual free-DNA after purification (Fig. 3e) (see 'Ab-DNA purification').

Particle surface functionalization and quantification
The ratiometric and density control of one or more functionalities on particle surfaces are achieved through cargo-directed (Fig. 4a-e) and scaffold-directed (Fig. 4f-h) strategies.For the former method, one or multiple functionalities (cDNA-protein cargos) are hybridized onto the surfaces with a total input amount below the predetermined loading capacity of the cargo (Fig. 4a-e).The density and relative ratio of different cargos are adjusted by the input mixture ratio before surface hybridization (Fig. 4d,e).For the latter method, one or multiple functionalities are hybridized onto particles with different densities of DNA scaffolds (with one or more sequences), and the loading input of the cargos are controlled at 3× molar excess to the predetermined loading capacity of each cargo (Fig. 4f-h).Flow cytometry enables the precise ratiometric quantification of surface-decorated cargos resulting from either method.While the cargo-directed method does not require unique particle scaffold formulations, it can have limitations in precision when cargos with different chemical properties (e.g., size and charge) are co-loaded 20 .Currently, we focus on using this method to characterize the relative densities and ratios of biomolecules, although the absolute numbers could be further quantified by establishing standard titrations in the future.

Exemplified human T cell activation ex vivo
DNA-scaffolded PLGA microparticles (2 µm) were coated with cDNA-conjugated agonistic antibodies αCD3 and αCD28 at various ratios to provide stimulatory and co-stimulatory signals for human T cell activation and ex vivo expansion, which is a key step for T cell manufacturing 29 .As reported previously, the ratiometric control of αCD3 to αCD28 had an impact on T cell expansion fold (Fig. 5a), and here we found that the particle to cell excess also affected cell expansion 20 (Fig. 5b).Additionally, we expect that the size of the particles and the stability of the polymer may also matter for cell activation, so these chemiophysical parameters-in addition to details of material-cell interactions-would need to be systematically investigated among multiple T cell donors when adopting this type of material for T cell manufacturing 24,76,77 .Other than cell quantity, cell quality that is directly associated with the therapeutic efficacy after infusion into patients should also be evaluated.Here, we exemplified a flow cytometry-based immune profiling to evaluate cell differentiation (memory and effector fates, Fig. 5c,d) and exhaustion (co-expression of inhibitory receptors, Fig. 5c,e).While cell phenotyping provides important metrics relating to cell quality, the functionality of manufactured cells should also be evaluated in vivo in related animal models 78 .

Fig. 1 |
Fig.1| Schematic of the fabrication protocol for precision ICEps.The precise functionalization of immunomodulatory signals on synthetic material surfaces is enabled by attaching DNA handles on both components and associating them via DNA hybridization.This protocol involves (i) synthesizing particles with dense surface DNA scaffolds (with one or multiple sequences) through emulsion-based fabrication using polymer-DNA amphiphiles as surfactants, (ii) conjugating the cDNA to the immunomodulatory biomolecules with minimal bioactivity loss and complete removal of free DNA, and (iii) loading cDNA-biomolecule conjugates on particle surfaces through one-step hybridization.Here, ICEps are exemplified for their use in human T cell ex vivo expansion, which is highlighted with essential details in (iv) T cell isolation and (v) cell culture and activation that can impact phenotypic outcome of cell products.TCR, T cell receptor.

2 Fig. 2 |
Fig. 2 | Quality control of PLGA particles with dense DNA scaffolds.a, A schematic of the synthesis of polymer-DNA amphiphiles and their quality check via gel electrophoresis-an essential step in achieving a high-DNAscaffold density upon particle fabrication.b-d, Urea-PAGE of PLGA-PEG-DNA conjugates from the synthesis reactions using varying molar ratios of thiol-DNA to PLGA-PEG-mal (b), different lots of PLGA-PEG-mal (c) and different lots of thiol-DNA (d).The total DNA input into each lane was controlled at 1 pmol.e, Normalized PLGA-PEG-DNA amount to the total DNA amount in each lane of gel images in b and c using densitometry analysis in ImageJ.PLGA Lot-1 and Lot-2 were fit using exponential plateau (R 2 = 0.9983, root mean square error (RMSE) 0.0057) and exponential growth (R 2 = 0.9956, RMSE 0.0041) models, respectively.f, Surface-loading capacity of fluorescently labeled cDNA on microparticles (2 µm diameter) fabricated using PLGA-PEG-mal from different lots of PLGA-PEG-mal and thiol-DNA in c and d.Data are mean ± s.e.m of n = 3 technical replicates.g, Representative confocal microscope images (40× magnification) of particles fabricated using different protocols that yield different sizes and hybridized with Cy3-labeled cDNA.Scale bar, 20 µm.h, Size distribution of particles measured using Zetasizer (size A: mean diameter 220.42 nm, mean diameter range 201.54-234.35nm, PDI range 0.120-0.176,n = 3 independent samples) or shown in g using ImageJ analysis (size B: mean diameter 1.90 µm, mean diameter range 1.74-2.10µm, s.d.average 1.01 µm, s.d.range 0.99-1.01,n = 3 independent samples; size C: mean diameter 7.78 µm, mean diameter range 6.57-8.61µm, s.d.average 3.75 µm, s.d.range 3.40-3.89µm, n = 4 independent samples).Size frequencies were fit with gaussian distribution curves ((adjusted R 2 , RMSE): size A (0.8323, 3.343); size B (0.9462, 1.720); size C (0.6897, 2.543)) and the shaded regions represent error envelopes of ±1 s.d. for each discrete frequency bin.

Fig. 3 |
Fig. 3 | Protocol and quality checkpoints of Ab-DNA conjugation and purification.a, A schematic of Ab-DNA conjugation through selective reduction of antibody hinge-region disulfides and Fc affinity-based chromatography for purification to remove excess, unreacted DNA.b, SDS-PAGE of Ab-DNA conjugates from selective reduction by TCEP treatment at 4.5× molar excess (lane 1) versus full reduction by β-mercaptoethanol (β-ME) treatment (lane 2).c, Densitometric analysis of Ab-DNA bands from urea-PAGE of reactions with varying ratios of DNA to antibody input.Data were fit using a sigmoidal doseresponse curve (R 2 = 0.9827, RSME 0.8078).Data represent mean ± s.e.m of n = 3 technical replicates.d, Example heat maps depicting the A 280 or A 280 /A 260 ratios of Ab-DNA conjugates eluted from Fc affinity-based chromatography columns.e, Urea-PAGE of Ab-DNA conjugates with and without purification.

Fig. 4 |
Fig. 4 | Density and ratiometric control of cargos co-loaded onto particle surfaces.a, A schematic of titrating the input amount of cDNA cargo to control the surface density on particles.b,c, Flow cytometry histograms (b) and normalized mean fluorescence intensities (MFIs) (c) of particles hybridized with varying input amounts of Ab-DNA (αCD28-compR-AF488, full: 20 nM/ OD 550 , 1/2: 10 nM/OD 550 , 1/4: 5 nM/OD 550 , 1/8: 2.5 nM/OD 550 ).A linear trend was determined using a one-way ANOVA (F 1,15 = 3,944, P < 0.0001) and inter-Ab-DNA input P values were determined by one-way ANOVA (F 4,15 = 1,151, P < 0.0001) followed by Tukey's post hoc test.d, A schematic of the ratiometric control of surface cargos by the input ratio of Ab-DNA cargos at the hybridization-based assembly step.e, Flow cytometry-based quantification of particles (R:G of 1:1) that are hybridized with different ratios of Ab-DNA (αCD28-compR-AF488

Protocol 9 .
Create a reaction template in Excel to facilitate reagent calculations for synthesizing the PLGA-PEG-DNA.Refer to Table

77 .
During particle incubation, generate fluorescent-cDNA standard curves in a black-walled microwell plate.(A) For each fluorescent cDnA: (i) Start with 200 µL of a 2 µM DNA concentration in 9% (vol/vol) DMSO in PBS (PBS-DMSO).(ii) Remove 100 µL to perform twofold serial dilutions until reaching the limit of detection for the plate spectrophotometer, leaving 100 µL per well.(iii) Separately, make blank wells containing 100 µL PBS-DMSO for background subtraction.(iv) Cover the well-plate top and set aside to protect from light.
). 33.Weigh 50 mg of unmodified PLGA 50:50 (38-54 kDa, PLGA) into a 15 mL tube (fabrication tube).34.Use a glass pipette to add 400 µL of EtOAc into the tube.▲ CrITICAL STEP Keep EtOAc-containing tubes open for as little time as possible to minimize evaporation-this will reduce the size variability between batches.Do not hold tubes near the liquid as this may contribute to heating.35.Wrap the tube with parafilm and place vertically on a shaker table overnight to dissolve.36.The next day, place stock tubes of EtOAc, water and fabrication tubes on ice to reduce evaporation when opened.

Table 4 | Particle surface hybridization of complementary DnA
a When loading antibody onto 2 µm particles, the loading capacity is 20 nM/OD 550(ref.20).These values must be determined for each biomolecule species and particle size.200 nm diameter nanoparticle cDNA loading capacity is between 1,000 and 2,000 nM/OD 550 and the 8 µm particles load between 12 and 20 nM/OD 550 , depending on batch-tobatch variation.Antibody loading capacity has not been determined for the nanoparticles or 8 µm particles.b If this volume exceeds 50 µL, centrifuge this particle volume as previously described and remove supernatant until 50 µL remains.Adjust the calculation for Extra TE-Tween volume accordingly.c This value represents the total loading capacity of all represented cDNAs.If ratiometric particles are used, this value should be distributed between each sequence based on the particle scaffold DNA surface ratio.

Table 5 | Troubleshooting table
PEG-DNA conjugation efficiency should be tracked.If one batch failed or had low conjugation efficiency, the reaction should have been redone and the poor conjugate should not have been used to fabricate particles Polymer concentration too high after diluting DMSO-degraded particles, leading to reaggregation After diluting the DMSO-degraded particles, do not exceed 1-2 OD 550 per 100 µL.Minimize wait time before plate-reader analysis Improper particle handling during spin-down steps or dilutions During supernatant removal steps, ensure that particles are not accidently removed.Ensure thorough mixing before any dilutions or aliquoting Particle scaffold fidelity is impaired due to particle age or mishandling During lyophilized particle resuspension, select the proper solution to not increase salt concentration.Monitor particle DNA loading over time, as the mal-thiol bond can hydrolyze, leading to a less dense scaffold Improper ratio-mixture of PLGA-PEG-DNAs before particle fabrication When drying polymers at a fixed ratio, ensure that the volumes mixed are accurate and no liquid is stuck inside the pipette tip 100 Particle sizes are highly variable between batches Incorrect amount of unmodified PLGA Ensure the unmodified PLGA amount is within ±1% weight target between batches Improper mixture of particle fabrication components Ensure that all components are thoroughly mixed before emulsification steps.Pay extra attention to the pipetting volumes for viscous or volatile components Evaporation of EtOAc during mixing Cool all liquid reagents on ice before mixing and avoid heat transfer from hands by holding tubes away from bottom.Reduce time that volatile tubes are opened 146 Low Ab-DNA gel band intensity Poor conjugation efficiency Titrate DNA amount to find optimal concentration; if no conjugation is observed, verify quality of the cDNA and/or use fresh TCEP.Ensure quality of the NHS-PEGmal linker and keep in proper storage conditions Incorrect staining or gel imaging procedure Verify staining reagent is compatible with selected gel type.Ensure the correct pmol of biomolecule was loaded in the lanes.Make sure that the gel-doc voltage and filter channel is appropriate for the dye used 170 Low antibody signal on particle surface Poor or incorrect dye labeling Increase the dye-to-antibody ratio and verify the reaction calculations are correct.If still low signal, verify the NHS-dye concentration or repeat labeling reaction DNA impurities leading to competition for hybridization Ensure that the Ab-DNA purity is above 95% for removing unreacted DNA in Step 147; perform additional column purifications and increase stringency on A 280 /A 260 cutoffs for elution collection 190 Poor T cell expansion Donor variation Evaluate multiple donors as some may just have poor expansion at baseline.Compare with a gold-standard expansion reagent, such as Dynabeads, to ensure the biomaterial is not at fault Low antibody activity due to improper handling Verify antibody structural integrity using SDS-PAGE.Perform cell-staining studies using stock Ab-DNA and comparing binding with unmodified controls (flow cytometry).Perform new conjugation with newly purchased antibody if the stock unmodified antibody quality is suspected Incorrect number of particles given It is critical to not lose particles during wash steps.Before adding to culture, remeasure the stock particle OD 550 to ensure the correct volume of particles are added