Surface Cross-Linking by Macromolecular Tethers Enhances Virus-like Particles’ Resilience to Mucosal Stress Factors

Virus-like particles (VLPs) are emerging as nanoscaffolds in a variety of biomedical applications including delivery of vaccine antigens and cargo such as mRNA to mucosal surfaces. These soft, colloidal, and proteinaceous structures (capsids) are nevertheless susceptible to mucosal environmental stress factors. We cross-linked multiple capsid surface amino acid residues using homobifunctional polyethylene glycol tethers to improve the persistence and survival of the capsid to model mucosal stressors. Surface cross-linking enhanced the stability of VLPs assembled from Acinetobacter phage AP205 coat proteins in low pH (down to pH 4.0) and high protease concentration conditions (namely, in pig and mouse gastric fluids). Additionally, it increased the stiffness of VLPs under local mechanical indentation applied using an atomic force microscopy cantilever tip. Small angle X-ray scattering revealed an increase in capsid diameter after cross-linking and an increase in capsid shell thickness with the length of the PEG cross-linkers. Moreover, surface cross-linking had no effect on the VLPs’ mucus translocation and accumulation on the epithelium of in vitro 3D human nasal epithelial tissues with mucociliary clearance. Finally, it did not compromise VLPs’ function as vaccines in mouse subcutaneous vaccination models. Compared to PEGylation without cross-linking, the stiffness of surface cross-linked VLPs were higher for the same length of the PEG molecule, and also the lifetimes of surface cross-linked VLPs were longer in the gastric fluids. Surface cross-linking using macromolecular tethers, but not simple conjugation of these molecules, thus offers a viable means to enhance the resilience and survival of VLPs for mucosal applications.


Table of Contents
We finally emphasize that in our investigation of the crosslinking reaction, no aggregation was observed up to a VLP concentration of 4 mg/ml and a molar excess of the modifying PEG molecule to AP205 monomer of 10x, as determined by dynamic light scattering (DLS).It is important to note that this assertion assumes a low excess of dimethyl sulfoxide (DMSO), the organic solvent used to dissolve the PEG molecules.A low DMSO excess is achieved when a 10-100 mM PEG solution is used for the reaction.Conversely, when a 1.0 mM PEG solution is used, the excess of DMSO   VLP.The VLPs bPEG9-AP205, bPEG5-AP205, BS3-AP205 and Sulfo-NHS-AP205 were prepared at three molar ratios, 1x, 5x and 10x.In lane 1, the MW band in the range 15-20 kDa corresponds to AP205 monomer.The same band appears in all the other lanes, expect that in lanes 5 and 6 versus 3 and lanes 7 and 8 versus 6, respectively for bPEG9-AP205 VLP and bPEG5-AP205 VLP, the bands were shifted to a higher molecular weight consistent with an increasing number of PEGylated AP205 monomers at higher molar ratio from mass spectroscopy.This statement does not apply to BS3-AP205 VLP and Sulfo-NHS-AP205 VLP which appear to be saturated already at 1x molar ratio.In cases of bPEG9-AP205 VLP and bPEG5-AP205 VLP, MW bands in the range 25-37 kDa appeared and corresponded to crosslinked AP205 dimers.Similarly, with bPEG9-AP205 VLP and bPEG5-AP205 VLP, the MW bands around 50 kDa appeared and corresponded to AP205 trimers.Bands corresponding to higher order oligomers were apparent for these VLPs.
Unlike PEG-crosslinked AP205 VLPs, in lanes associated with BS3-AP205 VLP bands only in the dimeric region appeared while with Sulfo-NHS the band strength in the dimeric regions is similar to the native AP205 VLP.

S5. Mass spectrometric characterization of PEG-crosslinked AP205 VLPs
Electrospray ionization mass spectroscopy (ESI-MS) -"AP205 monomeric region" From these measurements, the average MW of AP205 monomer was measured to be 15768 Da.
Reaction with BS(PEG9) at 1x molar ratio resulted in a fraction of AP205 monomers remaining either native (i.e., no PEG molecule was conjugated to them), or conjugated to one PEG molecule.
Increasing the molar ratio to 5x decreased the population of native AP205 monomers, while peaks at doubly-and triply-PEGylated AP205 monomers appeared.At 10x molar ratio, there was no population of native AP205 monomers, and we found a maximum of four PEG molecules per AP205 monomer in the monomeric region.
The variation of average MW between the individual peaks corresponded to the MW of one-end conjugated BS(PEG9).
The reactions with Bis-dPEG13-NHS ester, Bis-dPEG17-NHS ester and Bis-dPEG25-NHS ester followed a similar behavior; however, at 10x molar ratio, a maximum of three PEG molecules per AP205 monomer was detected.

ESI-MS of AP205 VLP
In           Nano ultra-performance liquid chromatography coupled to mass spectroscopy (NanoUPLC-MS/MS) analysis of bPEG9-AP205 VLP -"AP205 dimeric region" The results showed that five out of seven lysine residues together with the N-terminus were modified by one-end conjugated BS(PEG9) with the molecular formula C22O12H40 (MW 496.3 Da) (Table S5-6).
In agreement with the results of reducing SDS-PAGE, the LC-MS/MS analysis revealed the presence of "inter"-AP205 monomer crosslinking.In this case, two AP205 monomers were connected by PEG9 crosslinker with the molecular formula C22O11H38 (MW 478.3 Da).The cross linkages were between K4 and K61 as well as between K4 and K94.More details in below: The raw data were analyzed by Byonic 4.0 (Protein Metrics, USA) with the consideration of mass tolerance of 5 ppm at MS1 level, and 20 ppm at MS2 level and a score higher than 200.To confirm the results, Spectrum Identification Machine (SIM-XL) was also used for crosslink analysis 1 .
For crosslinked peptides, the results from SIM-XL showed that there were two intra-linked peptides at A2-K4 and K94-K101 (Figure S5-7 and Table S5-7).In addition, two cross-linked peptides were observed, namely A2/K4-K61 and A2/K4-K94.The software did not confirm either N-terminus A2 or lysine K4 participated in the crosslinking reaction with the other lysine residue (K61 or K94) due to the lack of key ions.However, if the major linkage was between N-terminus and K61 or K94, K4 residue would have been cleaved off by trypsin.Then, the results would have shown a linkage between 2 AN 4 K-crosslinked with the partner peptide.But the results showed 2 AN 4 KPMQPITSTAN 15 K linked to the other peptide.Therefore, the data suggested that most likely, K4 instead of A2 was linked to the other lysine.Byonic results agreed with the assignment from the SIM-XL software.
The location of the afore-mentioned N-terminus and lysine residues are noted in the sequence below: M. 1.The chemical format for one-end conjugated PEG9 adduct on lysine is C22O12H40 and the exact mass is 496.26.For PEG9-crosslinked peptides, the mass is the sum of two peptides plus 478.26 (C22O11H38).2. The MS/MS spectra could not confirm which amino acid was modified by PEG9 due to the lack of b1 and b2 ions.3. The lysine was not labeled with PEG9.
Figure S5-7: The scheme of crosslinked peptides from bPEG9-AP205 VLP via SIM search engine.The fasta sequence that was used to perform the search started from A at the N-terminus.Therefore, the position of modified amino acids had 1 position less than the full sequence.The identified peptides are summarized in Table 2.

Native AP205 VLP
The colloidal stability of AP205 VLP was narrow, e.g., at pH  6.0, the VLP was readily aggregated.
At pH 2.0, the VLP was somewhat stable.Therefore, the stable pH for AP205 VLP in PBS is pH 7.4.
AP205 VLP + BS(PEG)25 100x AVE increases by 10-100x, leading to partial aggregation of reaction products.Despite this drawback of using a 1.0 mM PEG solution, it is a potentially better PEG solution concentration due to easier handling and higher accuracy in maintaining the stoichiometry of the reaction, especially considering that our work involved very small VLP volumes (100-200 µl).After crosslinking, the product was thoroughly cleaned using ZebaSpin Desalting columns to remove excess PEG molecules.Subsequently, it was filtered by centrifugation (0.22 µm) to eliminate aggregates formed either due to DMSO excess or during ZebaSpin Desalting treatment.S3.Morphology of PEG-crosslinked AP205 VLPsUsing transmission electron microscopy (TEM), we investigated the morphology of PEGcrosslinked AP205 VLPs.The images showed a spherical geometry after PEG-crosslinking with BS(PEG9), BS(PEG13), BS(PEG17), BS(PEG21) and BS(PEG25).
the figure below, the mass spectrum of AP205 VLP is divided into three regions corresponding to the molecular weight of AP205 monomer (MW 15768 Da), dimer (MW 31536 Da), and trimer (MW 47304) which are indicated on each subfigure (respectively, left, middle and right).
AP205 direct mix PBS pH2 AVeComparison of colloidal stability between native and PEG-crosslinked AP205 VLPsThe figure below shows the effect of pH on colloidal stability of native and PEG-crosslinked AP205 VLPs.The data suggested that AP205 VLP was aggregated in pH 6.0-3.0.The origin of this aggregation is electrostatics.As a function of pH, the surface charge density of AP205 coat protein monomers varies leading to instability of the VLP.In the cases of bPEG9-AP205 VLP and bPEG21-AP205 VLP, the VLPs remained stable up to pH 4.0.Therefore, PEG-crosslinking increased the colloidal stability at low pH.

Figure S7- 2 .
Figure S7-2.Enzymatic stability of native and PEGylated VLPs in pig gastric fluid at pH 5.5.Agarose gel of AP205 VLP, bPEG9-AP205 VLP, bPEG17-AP205 VLP and bPEG25-AP205 VLP in PBS and in pig gastric fluid, respectively on the left and right of the gel.VLP incubation times included 5 min, 30 min, 1 hour and 1.5 hours at 37°C.M stands for marker.

Figure S7- 4 .
Figure S7-4.Enzymatic stability of native and PEGylated VLPs in pig gastric fluid at pH 3.0.(a) Agarose gel of AP205 VLP, bPEG9-AP205 VLP, bPEG17-AP205 VLP, bPEG25-AP205 VLP and mPEG25-AP205 VLP incubated with pig gastric fluid (pH 3.0) at time points 0 min, 1 min, 2 min, 5 min, 15 min, 30 min, 60 min and 120 min in consecutive lanes.M stands for marker.(b) Reduction in gray value intensity (calculated using ImageJ) as a function of time in the period 0-5 min.(c) Degradation rate estimated from the slop of the reduction in gray value with respect to time.

Figure S8- 1 :
Figure S8-1: The accumulation height and the mean fluorescence intensity (MFI) of the translocated native and PEG-crosslinked AP205 VLPs on nasal monodonor tissue 1 (MD0860) (a) and on nasal monodonor tissue 2 (MD0871) (b).In the experiments with tissue 1, PBS was added as a control.In the experiments with tissue 2, no liquid was added as control.

Table S5 -
1: Summary of the number and relative abundance of PEG9-conjugated AP205 monomers.

Table S5 -
2: Summary of the number and relative abundance of PEG13-conjugated AP205 monomers. 1[ m o lecu le A P 2 0 5 m o n o m er ]

Table S5 -
3: Summary of the number and relative abundance of PEG17-conjugated AP205 monomers. 1[ m o lecu le A P 2 0 5 m o n o m er ]

Table S5 -
4: Summary of the number and relative abundance of PEG21-conjugated AP205 monomers.

Table S5 -
7: Summary of PEG9-cross-linked peptides from SIM-XL search engine Intra-linked

Table S6 -
1: Summary of the measurements from FigureS6-1.Cumulant and Contin analysis algorithms were used for data analysis.