HaloTag display enables quantitative single-particle characterization and functionalization of engineered extracellular vesicles

Extracellular vesicles (EVs) play key roles in diverse biological processes, transport biomolecules between cells, and have been engineered for therapeutic applications. A useful EV bioengineering strategy is to express engineered proteins on the EV surface to confer targeting, bioactivity, and other properties. Measuring how incorporation varies across a population of EVs is important for characterizing such materials and understanding their function, yet it remains challenging to quantitatively characterize the absolute number of engineered proteins incorporated at single-EV resolution. To address these needs, we developed a HaloTag-based characterization platform in which dyes or other synthetic species can be covalently and stoichiometrically attached to engineered proteins on the EV surface. To evaluate this system, we employed several orthogonal quantification methods, including flow cytometry and fluorescence microscopy, and found that HaloTag-mediated quantification is generally robust across EV analysis methods. We compared HaloTag-labeling to antibody-labeling of EVs using single vesicle flow cytometry, enabling us to quantify the substantial degree to which antibody labeling can underestimate the absolute number of proteins present on an EV. Finally, we demonstrate use of HaloTag to compare between protein designs for EV bioengineering. Overall, the HaloTag system is a useful EV characterization tool which complements and expands existing methods.


Synthesis of Citrate-Stabilized AuNPs
Gold nanoparticles (AuNPs) were synthesized by a variation of the Frens/Turkevich method. 2 All used glassware was pre-washed with aqua regia and triple rinsed with water before synthesis.HAuCl4 hydrate (42.2 mg, 0.11 mmol) was dissolved in water (99 mL).The clear, yellow solution was brought to a rolling boil with stirring and allowed to boil for 15 min.To the boiling solution a solution of tribasic sodium citrate dihydrate (115.3 mg, 0.43 mmol) in water (1 mL) was added quickly.Addition of the citrate solution was followed by the typical loss of solution color followed by a greying of solution until it turned black and ultimately took on a wine-red color.The solution was allowed to continue to reflux for 15 minutes after citrate solution addition.The winered solution was then allowed to slowly cool to room temperature and filtered through a medium glass frit to obtain the colloid.

Preparation of TEM Samples
To prepare the samples for transmission electron microscopy (TEM) microscopy, carbon-coated, copper TEM grids were used as a support film.However, these grids are naturally hydrophobic, making it difficult for aqueous solutions of particles to uniformly spread and adhere to the grid surface.Therefore, grids were glow discharged on a Pelco EasiGlow Glow Discharge system for 30 seconds to deposit a charge onto the grids, rendering them hydrophilic.Next, 5 μL of the AuNP solution was placed onto the grid, and excess solution was absorbed by Whatman filter paper.The sample was then left to settle for 1-2 minutes to dry.TEM micrographs of the synthesized AuNPs were collected using a Hitachi HD-2300 Dual EDS Cryo STEM operated at a voltage of 200kV.

Characterization of Citrate-Stabilized AuNPs
Citrate-stabilized AuNPs were characterized by TEM for gold core size (Figure S1A-B).Raw micrographs were worked up with ImageJ software. 1Scale bars were set and the images were then converted to 8-bit.A gaussian blur of 2 was applied and the image threshold was adjusted to create a binary image of black particles on a white background.Binary images were then converted to a mask and the watershed function was applied to separate particles in close proximity.Particle areas were then collected from ImageJ, confirmed for overlap with raw images, and converted to diameters using the equation  = 2 * sqrt(Abs/) under the assumption of a circular projection of a spherical particle shape.The population of measured particles (N=133) was assumed to be gaussian and described with a mean of 12.3 ± 1.3 nm (error represents standard deviation around the mean).UV/visible spectroscopy was collected on the colloid to corroborate the average nanoparticle size as well as get a nanoparticle concentration via reported empirical equations (Figure S1C-D). 3UV/vis was collected using a microdrop cell with a 1 mm path length.From the spectrum a calculated average size of 12.1 nm was determined in good agreement with the measured TEM value.From the UV/vis a calculated concentration of 23 nM AuNPs was determined using the reported extinction coefficient at 450 nm for 12 nm gold AuNPs.

Synthesis of HT-ligand
Supplementary Scheme 1: Synthetic route to coupling-ready ligand for HaloTag protein, HTligand. 4,5e synthesis of HT-ligand was accomplished in four steps from 2-(2-aminoethoxy)ethanol following literature procedures (Scheme S1). 4,5Each intermediate was isolated via standard phase silica gel chromatography by cited literature procedures.Briefly, the amine of 2-(2-aminoethoxy)-ethanol was protected using Boc2O in ethanol. 4The unprotected alcohol was then deprotonated with sodium hydride and alkylated with 1-bromo-6-chlorohexane in a mixture of THF and DMF. 4 The Boc protecting group was then removed with 14% TFA in DCM (v/v) and then worked up with K2CO3 in methanol to yield a free amine. 4The resulting free amine was further alkylated with succinic anhydride in dichloromethane in the presence of triethylamine to afford peptide coupling-ready HT-ligand. 5The product was purified using flash chromatography on silica gel using 40:10:2 DCM/CH3OH/NH4OH(aq) and TLC was monitored using an eluent of 20:10:1 DCM/CH3OH/NH4OH(aq) and visualized with ninhydrin stain.Supplemental Figure 2: 1 H NMR (top) and 13 C NMR spectra of HT-ligand in CDCl3.The inset in the 1 H spectrum is a zoom-in of the peaks between 1.2-3.8ppm.Blue Xs represent peaks which were ignored from residual DCM and an unknown minor impurity that co-eluted.

Solid-Phase Peptide Synthesis
Supplementary Scheme 2. Reaction flow for solid-phase peptide synthesis for the CKEEE-LA and CKEEE-HT peptides.
Peptides were synthesized via solid-phase peptide synthesis (SPPS) using 2chlorotritylchloride resin pre-loaded with S-trityl-L-cysteine (0.401 mmol/g, ChemImpex).Syntheses were carried out at a 0.1 mmol scale (ca.250 mg of resin).The resin was swelled in a PolyPrep ® chromatography column (BIO-RAD, 20 mL, 0.8×4.0cm) in dichloromethane (15 mL) for 90 min and transferred to a glass peptide synthesis chamber affixed with a medium frit filter and valve selector between vacuum and high-pressure N2 inlet.The resin was then washed with dichloromethane (3x10 mL) followed by anhydrous DMF (3x10 mL).Deprotections were carried out using a solution of piperidine in anhydrous DMF (10 mL, 20% v/v piperidine) with N2 bubbling for 7 min.After deprotection the resin was washed five times with anhydrous DMF.Each washing involves the addition of 10 mL of anhydrous DMF, vigorously bubbling the mixture with N2 gas, and removal of solvent with vacuum assistance.Amino acid couplings were performed with 4 eq.(0.4 mmol) of the relevant Fmoc-protected amino acid and 4 eq.(0.4 mmol) of HBTU in a solution of 2,4,6-collidine in anhydrous DMF (10 mL, 20% v/v 2,4,6-collidine, mixed fresh).Amino acid/HBTU solutions were allowed to sit for ≥15 min before use.Coupling solutions were added to the deprotected peptide and allowed to bubble with N2 gas to mix for 20 min.After coupling, the same washing procedure was used as after deprotections.For each synthesis couplings followed the order L-lysine, and 3 x glutamic acid.After final deprotection, a coupling was then performed in the same way with either then 4 eq.levulinic acid or HT-ligand.The resin was then washed with anhydrous DMF (3x10mL) and DCM (3x10mL) and transferred back to the PolyPrep ® column.The DCM was then removed from the resin with positive N2 pressure.Full cleavage and deprotection of the CKEEE-HT was accomplished with a solution of 95:2.5:2.5 TFA/H2O/TIPS (20 mL, v/v/v).Full cleavage and deprotection of the CKEEE-LA was accomplished with a solution of 95:2.5:2.5 TFA:H2O:TES (20 mL, v/v/v).The cleavage solutions were mixed via rotation for 120 min.The filtrate was then collected in a flask and the solvent was removed in vacuo.A small amount of oil with suspended white solids was obtained.Upon addition of diethyl ether (100 mL), white solids immediately crash out from the crude mixtures.The white solid was agitated in the ether and allowed to settle and the diethyl ether was decanted.The white solids were then washed in the same way an additional time and then dissolved in water (8 mL) and filtered through a 0.20 μm hydrophilic filters.The filtered solutions were frozen with liquid N2 and lyophilized to yield flocculent white powders.Peptides were characterized ESI-MS and used without further purification (Figure S3).The cleavage of CKEEE-LA from resin lead to apparent reduction of the terminal ketone to an alcohol, likely by the TES.CKEEE-HT: ESI-MS: m/z calcd for