Engineering hepatitis B virus core particles for targeting HER2 receptors in vitro and in vivo

Hepatitis B Virus core (HBc) particles have been studied for their potential as drug delivery vehicles for cancer therapy. HBc particles are hollow nano-particles of 30–34 nm diameter and 7 nm thick envelopes, consisting of 180–240 units of 21 kDa core monomers. They have the capacity to assemble/dis-assemble in a controlled manner allowing encapsulation of various drugs and other biomolecules. Moreover, other functional motifs, i.e. receptors, receptor binding sequences, peptides and proteins can be expressed. This study focuses on the development of genetically modified HBc particles to specifically recognise and target human epidermal growth factor receptor-2 (HER2)-expressing cancer cells, in vitro and in vivo, for future cancer therapy. The non-specific binding capacity of wild type HBc particles was reduced by genetic deletion of the sequence encoding arginine-rich domains. A specific HER2-targeting was achieved by expressing the ZHER2 affibodies on the HBc particles surface. In vitro studies showed specific uptake of ZHER2-ΔHBc particles in HER2 expressing cancer cells. In vivo studies confirmed positive uptake of ZHER2-ΔHBc particles in HER2-expressing tumours, compared to non-targeted ΔHBc particles in intraperitoneal tumour-bearing mice models. The present results highlight the potential of these nanocarriers in targeting HER2-positive metastatic abdominal cancer following intra-peritoneal administration.


Supporting Materials
Ampicillin sodium, AIM-terrific broth base including trace elements, Tris base Ultra-Pure, EDTA disodium and DTT dithiothreitol were obtained from ForMedium™ (UK). Calcium chloride (anhydrous) and urea (carbamide) were obtained from Melford (UK). Imidazole, Glycerol,N,N,N',sodium chloride, agarose, ribonuclease A (from bovine pancreas) and Bromophenol Blue sodium salt were obtained from Sigma Life Science (UK). cOmplete™ ULTRA Tablets, glass vials Protease Inhibitor Cocktail and cOmplete™ His-Tag Purification Resin were from Roche (Germany).
Sodium hydrochloride and methanol were obtained from Fisher Scientific (UK). Acetic acid ≥99.0% (T) and skimmed milk powder were obtained from Fluka Analytical (Switzerland).

Determination of HER2 receptors expression levels in cancer cell lines
Cells were cultured in complete media in 25 cm 2 canted-neck tissue culture flasks. When cells reached 80% confluency, cells were washed twice with PBS buffer; trypsinised and 1 x 10 6 cells were transferred to assay tubes for the subsequent immunostaining. Cells were fixed in 4 % PFA for 10 min at room temperature, washed in PBS buffer and incubated with 150 µl of blocking buffer (1% BSA in PBS buffer) for 30 min at room temperature. After this period of time, cells were washed with PBS buffer and incubated with 20 µl of PE Mouse Anti-Human HER-2neu antibody (BD Biosciences, USA) or 20 µl of Isotype Mouse BALB/c IgG 1 , κ for 30 min in dark at room temperature. Finally, cells were washed with PBS buffer. PE fluorescence was analysed by flow cytometry using a BD FACS Calibur™ flow cytometer (BD Biosciences, USA). A total of 10,000 cells were gated and fluorescence was analysed in triplicates for each condition using the FL-2 detector and BD CellQuest software (BD Biosciences, USA).

Fluorescence labelling of HBc particles
All HBc particles were fluorescently labelled with Alexa Fluor™ 488 dye. Fluorescence was measured with excitation at 485 nm and emission at 520 nm, using BMG FLUOstar Omega fluorometer. As shown in Figure S2, HBc particles showed higher fluorescence intensity than ΔHBc and Z HER2 -ΔHBc particles ( Figure S2).

In vivo monitoring of highly-expressing HER2 (HER2 (+++)) xenograft tumour models by bioluminescence imaging
In order to study the biodistribution of HBc particles in an in vivo model, two tumour mice models were prepared; intraperitoneal and mammary fat pad tumour models. NSG mice were inoculated with luciferase-expressing MDA-MB-435-MLE cells, either intraperitoneally or subcutaneously into the mammary fat pad, to generate intraperitoneal or mammary fat pad tumour models, respectively. Tumour growth in each model was monitored by bioluminescence in vivo imaging, every 7 days after inoculation. Ventral images of a representative NSG mouse from both tumour models are shown in Figure S3. The intraperitoneal model showed tumour growths at multiple sites around the abdominal cavity, including the spleen, intestines and male reproductive organs ( Figure S3A). Meanwhile, localised bioluminescent signals were observed at mammary fat pads on both sides in the mammary fat pad model ( Figure S3B).

administered [ 99m Tc(CO) 3 ] + particles by gamma counting
Blood sampling at early time points demonstrated rapid clearance from the blood with values of ~50% and ~30%ID in blood detected at 30 min-post injection, in intraperitoneal and mammary fat pad tumour models, respectively ( Figure S4A). At 24 h, these values dropped to ~30% and ~10%ID in blood for intraperitoneal and mammary fat pad tumour models, respectively ( Figure S4A). High radioactivity eliminated into urine was observed for [ 99m Tc(CO) 3 ] + in both intraperitoneal and mammary fat pad tumour models with ~18-28%ID eliminated/mouse ( Figure S4B). A similar excretion profile with HBc particles in faeces was observed for [ 99m Tc(CO) 3 ] + in both models.
[ 99m Tc(CO) 3 ] + showed a similar accumulation profile in systemically-administered mammary fat pad tumour model (Figure S8C, grey bars) to that obtained in intraperitoneal model.
In the whole body SPECT/CT imaging of mammary fat pad tumour-bearing mice, [ 99m Tc(CO) 3 ] + exhibited a similar biodistribution profile to 99m Tc-ΔHBc and 99m Tc-Z HER2 -ΔHBc, where by most of the signals were retained in the tumour ( Figure S6A). No significant radioactivity signals were observed in any organs except bladder, which indicated that some of the particles reached the systemic circulation prior to elimination in the urine. As shown in Figure S6B, [ 99m Tc(CO) 3 ] + also resulted in a similar excretion profile to HBc particles, with 33.38 ± 5.66% and 1.36 ± 0.11%ID eliminated/mouse in the urine and faeces, respectively.

Blood clearance, excretion profiles and organ biodistribution profile of systemicallyadministered HBc particles by gamma counting
Quantitative studies by gamma counting were performed to assess the pharmacokinetic profile and organ biodistribution of HBc particles 24 h post-injection. Radiolabelled 99m Tc-ΔHBc and 99m Tc-Z HER2 -ΔHBc particles were injected via the tail vein (systemic route) in both of the tumour models. Blood sampling at early time points demonstrated fast clearance from the blood with values of ~7.0-7.5% and ~4.0-5.5%ID in blood detected in blood at 30 min-post injection, in the intraperitoneal ( Figure S7A) and mammary fat pad ( Figure S7B) tumour models, respectively. At 24 h, these values dropped to ~4.0-5.0% and ~0.5-1.5%ID in blood for the intraperitoneal ( Figure S7A) and mammary fat pad ( Figure S7B) tumour models, respectively. No significant difference in the blood clearance profiles between the 99m Tc-ΔHBc and 99m Tc-Z HER2 -ΔHBc particles was observed. Relatively low levels of radioactivity suggest that both HBc particles have a short blood circulation time in mice. Animals were also housed in metabolic cages for urine and faeces collection for 24 h to assess excretion profile. Approximately ~10-20% and ~1.0-2.0%ID was eliminated in urine and faeces within 24 h for both particles (p > 0.05) (Figure S7C, D).
Tumour uptake of systemically-administered HBc particles exhibited low uptake with ~ 0.5% and ~0.2% achieved in the intraperitoneal and mammary fat pad tumour model, respectively, with no significant differences observed between 99m Tc-ΔHBc and 99m Tc-Z HER2 -ΔHBc ( Figure S8).

Excretion profiles of locally-administered HBc particles by gamma counting
In the intraperitoneal tumour-bearing mice model, approximately ~4-10% and ~0.2-2.0 %ID was eliminated into urine and faeces after 24 h for both particles with no significant differences between the type of particles (p > 0.05) (Figure S9A). In contrast, in the mammary fat pad tumour-bearing mice model, higher radioactivity was observed in urine for both HBc particles after 24 h (~17-22%) than found with the intraperitoneal model.

Figure S1. HER2 expression levels in cancer cell lines by flow cytometry. (A)
Fluorescence intensity histograms (FL2) and (B) Median fluorescence intensity (MFI) analyses. Cell lines were stained with PE mouse anti-human HER2/neu primary antibody for assessing the expression level of HER2 receptors on cell membrane surface. PE mouse antihuman IgG1, κ isotype was used to account for non-specific binding to the IgG. HER2  WT-HBc, ΔHBc and Z HER2 -ΔHBc particles were fluorescently labelled with Alexa Fluor™ 488 dye and measured for fluorescence at 485nm and 520 nm excitation and emission wavelengths, respectively. Measurements were performed at 25˚C using a BMG FLUOstar Omega fluorometer. Error bars are too small to be seen. Higher labelling efficiency was obtained for WT-HBc than ΔHBc and ZHER2-ΔHBc particles.        Mice were injected with 99m Tc-ΔHBc (black bars) or 99m Tc-Z HER2 -ΔHBc (grey bars), intraperitoneally or intratumourally in the (A) IP or (B) MFP mouse tumour model, respectively, at a dose of 50 µg protein/mouse (4-6 MBq per mouse). Excretion profile at 24 h post injection is expressed as mean of %ID eliminated/mouse ± SD (n=3).