Amphiphilic Polyphenylene Dendron Conjugates for Surface Remodeling of Adenovirus 5

Abstract Amphiphilic surface groups play an important role in many biological processes. The synthesis of amphiphilic polyphenylene dendrimer branches (dendrons), providing alternating hydrophilic and lipophilic surface groups and one reactive ethynyl group at the core is reported. The amphiphilic surface groups serve as biorecognition units that bind to the surface of adenovirus 5 (Ad5), which is a common vector in gene therapy. The Ad5/dendron complexes showed high gene transduction efficiencies in coxsackie‐adenovirus receptor (CAR)‐negative cells. Moreover, the dendrons offer incorporation of new functions at the dendron core by in situ post‐modifications, even when bound to the Ad5 surface. Surfaces coated with these dendrons were analyzed for their blood‐protein binding capacity, which is essential to predict their performance in the blood stream. A new platform for introducing bioactive groups to the Ad5 surface without chemically modifying the virus particles is provided.

Thin-layer chromatography (TLC) was performed on Alugram Sil G/UV254 plates from Macherey-Nagel and substances were detected under UV light at 254 nm or 366 nm. Column chromatography was performed applying Macherey Nagel silica gel with particle size of 0.04-0.063 mm or 0.063-0.2 mm. Sizeexclusion chromatography was carried out using Sephadex® LH-20 in DMF.

H-NMR and 13 C-NMR were recorded on a Bruker
Avance III 300 MHz, Avance III 500 MHz or Avance III 700 MHz spectrometer in deuterated solvents like CD2Cl2, MeOD and DMSO-d6. 13 C-NMR were recorded in j-modulated spin-echo (JMOD) mode. Spectra were analyzed in either MestReNova or Topspin.

Field Desorption Mass Spectrometry (FD-MS)
. FD mass spectra of precursors were recorded on a VG Instruments ZAB 2-SE-FPD using an 8 kV accelerating voltage. Figure S1. Reaction scheme of building blocks. (A) AB4 building block 1, synthesized based on modified protocols from Morgenroth et al. [1] and (B) synthesis of surface building block 2 based on modified protocols from Stangenberg et al. [2] 1,3-Bis(4-bromophenyl)propan-2-one (20) 1,3-Bis(4-bromophenyl)propan-2-one (20) was synthesized according to the literature with modified protocol for purification. [3] Briefly, in a dry two neck round-bottom flask equipped with a dropping funnel dicyclohexylcarbodiimide (DCC) (9.6 g, 46.5 mmol) and 4-(dimethylamino)pyridine (DMAP) (1.42 g, 11.6 mmol) were dissolved in 100 mL dry dichloromethane. After degassing with argon for 30 min, pbromophenylacetic acid (10.0 g, 46.5 mmol) in 100 mL dry dichloromethane was added dropwise and the reaction mixture was stirred for 24 h at room temperature. Then, the resulting N,N'-dicyclohexylurea was filtered off and the organic layer was washed with 10% hydrochloric acid and water. The crude product was purified by column chromatography using a mixture of cyclohexane and dichloromethane (1:2) to afford 20 as a white solid (4.85 g, 57%). All spectral data was in agreement with the literature. [ General procedure P1 for Sonogashira-Hagihara coupling:

Synthesis of building blocks
The synthesis of ethynylated aryl compounds was modified from previously reported methods. [2,4] Aromatic bromo compound (1 equiv), ethynyl derivative (1.1 equiv per bromine on the bromo compound) and triphenylphosphine (0.1 equiv) were dissolved in a mixture of 1,4-dioxane and triethylamine (2:1, 40 mL per gram bromo compound). After degassing with argon for 30 min, bis(triphenylphosphine)palladium(II)chloride (Pd(Ph3)P)2Cl2) (0.05 equiv) and copper iodide (0.1 equiv) were added. The reaction mixture was stirred under reflux at 85 °C and argon atmosphere for 15 h. The reaction mixture was cooled to room temperature, the palladium catalyst was filtered off and the fitration residue was washed with dichloromethane. The solvents were removed in vacuo and the residue was dissolved in dichloromethane. The organic layer was washed with water, dried over sodium sulfate and the solvent was evaporated. The crude product was purified by column chromatography.

Synthesis of PPD3
PPD3 was synthesized in a divergent way as previously reported. All spectral data was in agreement with the literature. [ (13) Imine 11 (300 mg, 0.65 mmol) and TIPS-acetylene (12) (137 mg, 169 µL, 0.75 mmol, 1.15 equiv) were dissolved in 10 mL THF and 2 mL triethylamine. After degassing, Pd(Ph3)P)2Cl2 (45.8 mg, 65.3 µmol, 0.1 equiv) and copper iodide (24.9 mg, 131 µmol, 0.2 equiv) were added. The reaction mixture was stirred at room temperature under argon atmosphere for 15 h. Then, it was filtered and the filtrate was diluted with dichloromethane. The organic layer was washed with water, dried over sodium sulfate and purified by column chromatography using a mixture of cyclohexane and dichloromethane (1:2) to obtain 13 as a yellow solid (309 mg, 92%).

Amine-biphenyl-G2-(PSpen)4 7
Imine-biphenyl-G2-(PSpen)4 6 (110 mg, 36.6 µmol) was dissolved in 3 mL THF and 1 mL 2 N hydrochloric acid were added. After stirring at room temperature under argon atmosphere for 5 min, 0.5 mL of concentrated hydrochloric acid were added and stirred for further 20 min. Then, ethyl acetate and water were added. The organic layer was separated, washed twice with water and dried over sodium sulfate.
After concentration in vacuo the crude mixture was purified by column chromatography using a mixture of cyclohexane and THF (3:1) to afford compound 7 as a light brown solid (76 mg, 73%).

Synthesis of polystyrene nanoparticles
Amine functionalized polystyrene nanoparticles were synthesized via the previously reported direct miniemulsion protocol [6] . Cetyl trimethyl ammonium chloride was used a cationic surfactant to stabilize the dispersion and 2-aminoethyl methacrylate hydrochloride (2 wt% to styrene) was copolymerized with styrene. The dispersion was purified via centrifugation and dialysis. A detailed protocol is described in previous reports. [7] (Diameter Ø: 98 ± 10 nm, ζ-Potential: + 49 mV)

Coating of liposomes and nanoparticles with dendron 8 or dendrimer
Dendron 8 or the amphiphilic dendrimer was dissolved in DMSO at a concentration of 20 mg mL -1 .

Human plasma/serum
Human blood serum and plasma was obtained from six (serum) or ten (plasma) healthy donors at the Transfusion Center of the University Clinic of Mainz, Germany, pooled and stored at 20 °C. Citrate was used as an anticoagulant for plasma preparation.

Protein corona preparation
Liposomes coated with dendron 8 and dendrimer as well as nanoparticles (1 mg

Pierce assay
The Pierce 660 nm Protein Assay was used to determine the protein concentration. The assay was performed according to the manufacturer's instruction. The absorbance was measured with a Tecan infinite plate reader.

SDS PAGE
Proteins (2-3 μg in 26 µL) were loaded on a NuPage 10% Bis-Tris protein gels. Samples were mixed with 4 μL of NuPage Sample Reducing Agent and 10 μL of NuPage LDS Sample Buffer. Electrophoresis was carried out for 1 h at 120 V and gels were stained with Pierce Silver Staining Kit according to the manufacturer's instruction. All components were obtained from Thermo Fisher.

In solution digestion
Digestion of corona proteins was performed according to former instruction. [8], [9] Briefly, SDS was removed from the protein samples with Pierce detergent removal columns (Thermo Fisher). Afterwards, the proteins were precipitated overnight using ProteoExtract protein precipitation kit (CalBioChem) according to the manufacturer's instructions. The resulting proteins pellet was re-suspended in RapiGest SF (Waters Cooperation) dissolved in ammonium bicarbonate (50 mM) buffer. Proteins were reduced with dithiothreitol (Sigma, 5 mM, 45 min at 56 °C) and alkylated with idoacetoamide (Sigma, 15 mM, 60 min at room temperature). A ratio between protein:trypsin (50:1) was used and the digestion was carried out over 16 h at 37 °C. The reaction was quenched with 2 µL hydrochloric acid (Sigma).

Liquid chromatography coupled to mass spectrometry (LC-MS analysis)
Peptide samples were diluted with 0.1% formic acid and 50 fmol µL -1 Hi3 Ecoli (Waters Cooperation) was added for absolute protein quantification. [10] LC-MS measurements were performed with a Synapt G2-Si mass spectrometer coupled to a nanoACQUITY UPLC. A NanoLockSpray source was used in positive ion mode for electrospray ionization (ESI). Data-independent acquisition (MS E ) experiments were carried out and the Synapt G2-Si was operated in resolution mode. For data acquisition and processing MassLynx 4.1 and peptides/proteins were identified with Progenesis QI (2.0). The human database was downloaded from Uniprot modified with the sequence information of Hi3 Ecoli standard for absolute quantification. Processing parameters for peptide and protein identification were applied as described in detail in previous reports. [11] The absolute amount of each protein was determined in fmol based on the TOP3/Hi3. [12] Each measurement was performed in technical duplicates or triplicates.       were studied. By X-ray reflectivity measurements it was found that the interaction between the amphiphilic dendrimers and the zwitterionic lipids is mainly electrostatic. They proved that upon adsorption of the dendrimer towards the lipid surface, the monolayer remains intact. [13] As DOPE is also a zwitterionic lipid with a similar structure, we assume that the interaction between the amphiphilic dendron or dendrimer and the DOPE liposomes are also electrostatically driven. An incorporation of the lipophilic n-propyl group was not observed.

Materials and Instruments
Confocal laser scanning microscopy was performed using Leica TCS SP5. CellTiter-Glo ® Cell Viability Assay was purchased from Promega and luminescence intensities were measured on a Glomax Multi 96-well plate reader from Promega.

Instruments
Fluorescent imaging was carried out by a fluorescent microscope MF52 (Guangzhou Micro-shot

DLS and zeta potential at diluted conditions
Dynamic light scattering was used to determine interaction between Ad5 and dendrons by means of measuring the polydispersity index (PDI) and the hydrodynamic diameter of the particles. Complex formation was performed in a volume of 30 µL phosphate buffer (5mM, pH7.4) with 5×10 8 Ad5 particles.
Dendron was added in defined ratios to Ad5, then mixed and incubated for 40 min. After transfer to a cuvette, it was filled up with PB to a total volume of 0.9 mL. All samples were measured at 25 °C and an angle θ = 90°. For intensive cleaning of the cuvette, ethanol and acetone was used to avoid measurement errors by dust particles.
Zeta potential was used to determine the charge on the surface of Ad5 or complexes of Ad5 and dendrons. All samples were prepared the same as DLS and measured at 25 °C.

DLS and zeta potential at high Ad5 concentration and high ratios
The surface charge of vector particles was measured using a ZetaSizer Nano-ZS ( Table S4. Size and zeta potential of Ad5/dendron complexes. Ad5 vector with a concentration of 1x10 11 vp/mL was incubated with propargyl-dendron 8 at the ratios 1:200k (Ad5:Dendron) and 1:1000k as well as biotin-dendron 9a at the ratio 1:200k in 50 mM HEPES buffer pH 7.4 for 15 min. We observed an increase in size when mixing dendron with Ad5 and for Ad5 + propargyl-dendron 8 at the ratio of 1:1000k and biotin-dendron 9a at a ratio 1:200k a second peak was observed (Fig. S26).

Sample
Size (  We observed an increase in size when mixing dendron with Ad5 and for Ad5 + propargyldendron 8 at the ratio of 1:1000k and biotin-dendron 9a at a ratio 1:200k a second peak was observed. To assess whether this peak is related to unbound dendron in the mixture, free dendron was measured at same concentrations, which is shown in (B). For free dendron-conjugates we observed a size of about 100 nm which can be explained by assembly processes of the dendron in buffer solution due to its amphiphilic nature. Thus, we assume that the second peak (for Ad5 + propargyl-dendron 8 at the ratio of 1:1000k and biotin-dendron 9a at a ratio 1:200k) is related to unbound dendron which means that the Ad5 vector is saturated at these ratios.

Transduction in CAR-negative cell line CHO-K1
For transduction assays, 24-well plates were used containing 5 × 10 4 cells per well, which were seeded the day before transduction. Cells were cultured in DMEM/F12 medium with 10% FBS and 1% penicillin/streptomycin (PS, 10 000 IU penicillin and 10 000 μg/mL streptomycin) at 37 ºC in a humidified atmosphere containing 5% CO2. Dendron conjugates were added to Ad5 in defined ratios (

Investigating the importance of a ligand in CuAAC
Stock

Kinetic binding analysis [14]
The following chapter is reproduced with permission from ACS Nano 2019, 13, 8749-8759, https://pubs.acs.org/doi/10.1021/acsnano.9b01484. Further permissions related to the material within this chapter excerpted should be directed to the ACS journal.
The interaction between biotin-dendron 9a and Ad5 was studied by Bio-Layer Interferometry assays (BLI) from Octet96 (Pall ForteBio, CA, USA). In order to receive a significant signal for this binding event, we have immobilized biotin-dendron 9a at the sensor surface and applied Ad5 as binding molecule. To immobilize the dendron at the surface of streptavidin-coated biosensors, we used biotin-dendron 9a. The basic experiment contains four steps: Step 1 included hydration of the biosensor to record the baseline.
Step 3: Washing and establishing the baseline.
Step 4: Association of the Ad5.
A significant interaction signal could be seen even in the presence of only 2 pM Ad5. The KD (equilibrium dissociation constant) determined by this method is 1.27x10 -12 M. We believe that this very strong binding could be a result of multivalent interactions between the large virus particles providing large numbers of binding sites and the sensor surface densely coated with dendrons. These results clearly support that there is a strong binding between biotin-dendron 9a and Ad5 viruses. Figure S32. Workflow for dendron loading and dendron-Ad5 interaction assay [14] (adapted with permission from ACS Nano 2019, 13, 8749-8759, https://pubs.acs.org/doi/10.1021/acsnano.9b01484).