1. Introduction
Intracellular delivery of proteins or nucleic acids through a protein carrier is a potentially game-changing strategy for therapeutics. Common challenges that have to be addressed for this approach are the effective uptake of the protein–cargo carrier into the cell, the vector stability, the intracellular release of the cargo and the risks associated with the vector’s eventual cytotoxicity. To develop a delivery agent for therapeutics, several important steps should be taken into account, such as an initial cell-surface binding followed by internalization, efficient release from the endosome into the cytosol and intracellular trafficking towards the compartment of interest, for example, the nucleus [
1].
Adenovirus virions are the most commonly used vehicles that have been developed so far for gene therapy [
2,
3,
4] or cancer vaccine purposes [
5] and more recently as vaccine platforms in the context of the SARS-CoV-2 pandemic [
6]. The adenoviral virions are built from three major structural proteins; 240 copies of the hexon trimeric protein form the icosahedral capsid that also comprises 12 pentons located at the capsid vertices. Each penton consists of five copies of the penton protein connected to a fiber protein [
7]. Of particular interest is the ability of co-expressed penton plus fiber proteins to self-assemble into dodecahedral particles that could be used as “minimal” vectors for gene therapy [
8]. These viral-like particles (VLP)s can also be used to deliver drugs [
9], enzymes or large proteins such as immunoglobulins [
10] in cells [
11]. Engineered adenovirus capsid proteins, such as penton base capsomers of Ad2 and Ad5, have been shown to facilitate the entry of various membrane-impenetrable molecules including genes [
12], siRNA [
13] and corroles [
14]. Moreover, the adenoviral hexon protein, when fused into a polyethylenimine-plasmid complex, can enhance the nuclear delivery of DNA due to its nuclear homing capability [
15].
The number of reports on the intracellular uptake of adenovirus fiber proteins is considerably smaller than the ones focusing on virions, penton VLPs or penton capsomers. The soluble full-length fiber of Ad5 demonstrated high uptake levels into mammalian cells in the absence of the entire viral capsid, utilizing an actin-mediated and temperature-independent entry pathway. A potential route suggested could be through the heparan sulfate glycosaminoglycans (HS-GAGs) receptors found in cell lines. The fiber protein can assemble into a complex with protamine-condensed plasmid DNA and facilitate gene transfer [
16]. Given the aforementioned report and the exceptional stability of the fiber proteins, we deemed that they merit further investigation as stable potential cargo carriers for cellular delivery.
Adenovirus type 2 fibers are homo-trimers composed of an N-terminal domain attached to the viral capsid, a central fibrous shaft and a C-terminal domain that binds to the cell receptor, namely the Coxsackie and Adenovirus Receptor, termed CAR [
17,
18,
19]. The globular C-terminal “head” or “knob” domain is necessary for the trimerization of the fiber and might act as a registration signal that directs its correct folding and assembly [
20,
21,
22]. Synthetic fiber shaft peptides corresponding to shaft sequences do not fold and assemble into the native shaft fold in isolation; instead, the peptides self-assemble into amyloid-type fibrils [
23,
24,
25]. The shaft segment adopts a trimeric, highly intertwined fold termed the “triple beta-spiral” that conveys onto the fiber protein its exceptional stability [
26,
27]. The protein is resistant to SDS, temperature, proteases and denaturants [
22].
For gene therapy and delivery applications, the targeting of specific cell types is highly desirable. To target vectors onto specific cell types except for the CAR receptor, the knob CAR-binding domain should be substituted with other cell-targeting motifs. However, by removing the knob, the fiber loses its trimerization ability, and an alternative trimerization motif has to be used. Well-known trimerization motifs include triple-coiled coils [
28] and the beta-hairpin trimerization motif from phage T4 fibritin [
29]. Recombinant adenovirus virions have been previously reported [
30], where the entire knob domain of the Ad5 fiber was deleted and replaced by a helical trimerization motif from MoMuLV envelope glycoprotein, a Myc-epitope and a 6 x-His-tag. In the context of these virions, knobless recombinant Ad5 fibers were capable of efficient trimerization and internalization through their His-tag interaction with an anti-His single-chain antibody variant displayed on the cell surface. Bifunctional adapter molecules have been engineered that enable the targeting of the “mosaic” adenovirus fiber to alternative cellular receptors distinct from CAR. A metabolically biotinylated fiber mosaic Ad was fabricated through the incorporation of a biotin acceptor peptide in the sequence of the fiber protein. The hybrid protein containing a luciferase-expressing gene was effectively complexed with an epidermal growth factor (EGF)–streptavidin. The fiber mosaic virion was then retargeted to EGF receptor (EGFR)-expressing cells and exhibited increased infectivity [
31].
In the context of structural studies, fiber constructs have been previously designed to comprise a stable shaft segment (residues 319–392) N-terminally fused to the small non-cell targeting trimerization domain, the fibritin foldon, with or without the natural linker that connects the shaft to the globular head [
32]. The resulting “shortened” construct comprised four shaft repeats plus the foldon domain. The shaft repeats adopted their correct folding and trimerization ability into nanorods, as was verified by solving the crystal structure of the chimeric proteins [
33]. Moreover, they were found to be stable and could withstand extreme conditions, such as SDS, temperature and digestion by proteases.
In the present study, we aimed to design and study a series of almost full-length fiber constructs destined for potential application as protein carriers for intracellular delivery. The approach is again based on the principle of replacing the globular head with the foldon trimerization motif [
33], thus removing the CAR receptor specificity. Hybrid fiber constructs were fabricated comprising of a much longer shaft segment, Met 61-Gly 392 with or without the natural linker sequence in addition to the foldon trimerization motif.
The first generation of constructs was engineered as follows: a foldon domain was placed at the C-terminus to enable correct trimerization and a 6x His-tag for purification purposes was placed at the N-terminus. A second generation included an additional 14-peptide following the His-tag, which acts as a docking station and enables the binding of biotinylated molecules. The protein was overexpressed along with a biotin ligase enzyme with the addition of both IPTG and biotin for in vivo biotinylation. Protein biotinylation in vivo can be achieved by co-expressing in bacterial cells the protein of interest fused to a 14-amino-acid biotinylation sequence, (G-L-N-D-I-F-E-A-Q-K-I-E-W-H) and the BirA ligase. The BirA ligase specifically biotinylates the lysine side-chain of the biotinylation sequence by joining it with the carboxyl group of the biotin molecule [
34,
35]. The constructs designed and studied are summarized in
Scheme 1.
The constructs were evaluated for their expression and production in E. coli cells and ability to fold into a trimeric, rod-like conformation. They were subsequently studied for their stability in regard to temperature and protease digestion following their purification.
We report below that the His-tag conveys stability to the initial constructs; additionally, the His-tag biotinylated construct folds into thermally and protease-stable fibrous nanorods that can be internalized into mammalian cells and are not cytotoxic. Moreover, they can bind to proteins and nanoparticles through the biotin–avidin interaction and mediate their delivery to cells. We discuss the potential implications for their use as stable delivery vehicles.
2. Materials and Methods
2.1. Chemical Reagents
Chemical reagents and buffers were purchased from Sigma-Aldrich and Biotin Coated Plates and SuperSignal WestPico Plus chemiluminescent substrate were purchased from Thermo-Scientific. Q-Sepharose and Ni-NTA Sepharose columns were purchased from GE Healthcare. Carbon/Formvar electron microscopy copper grids were purchased from Agar Scientific. AlexaFluor594 FluoroNanogold-Streptavidin was purchased from Nanoprobes.
2.2. Cell Cultures
HeLa (human cervical cancer cells) and NIH3T3 (mouse fibroblast cells) were from the cell bank at the Institute of Molecular Biology and Biotechnology (IMBB), FORTH. Cells were cultured at 37 °C in a 5% humidified CO2 incubator in Dulbecco′s Modified Eagle’s–Medium (DMEM) growth medium (pH 7.4) from Gibco (Billings, MT, USA) supplemented with 10% Fetal Bovine Serum (FBS) purchased from Gibco and 50 μg·mL−1 gentamycin (Applichem, Darmstadt, Germany). Cell fluorescent dyes and NHS-Fluorescein were purchased from Thermo-Scientific, Waltham, MA, USA and nuclear dye DAPI was purchased from Molecular Probes (Eugene, Oregon).
2.3. Antibodies and Plasmids
Penta-His Antibody BSA-free was from Qiagen (Hilden, Germany) and secondary antibody anti-mouse IgG-alkaline phosphatase was purchased from Sigma-Aldrich (St. Louis, MO, USA). Streptavidin conjugated with alkaline phosphatase and streptavidin conjugated with HRP were from Merck (Darmstadt, Germany).
2.4. Molecular Cloning
The pET28a plasmid used was from Novagen (Darmstadt, Germany). Deep Vent polymerase and restriction enzymes used for cloning were purchased from New England Biolabs (Ipswich, MA, USA).
2.5. Design of the Chimeric DNA Constructs
The NoLinker and Linker protein constructs were fabricated with molecular cloning techniques using as template the pT7.7 vector containing the foldon trimerization domain sequence (Gly457–Leu483), as previously described [
33]. For the NoLinker protein, the DNA sequence corresponding to shaft residues Met61-Gly 392 was amplified with primers 5′-GACACCTCCCACCATATGCTTGCGCTTAAA-3′ and 5′-GGTAAGTTTGTCATCATTGGATCCTCCTATTGTAAT-3′ and inserted between the Nde
I and Bam
HI of the vector. For the Linker protein, where the shaft Met61- Gly 392 and the foldon sequences are connected with the natural fiber linker residues Asn393–Lys398 [
32], the following primers were used: 5′-GACACCTCCCACCATATGCTTGCGCTTAAA-3′ and 5′-TTGTCCACAGGGATCCTTTGTCATCATTTTTGT-3′. For both constructs, two extra residues (Gly- Ser) were used as a result of the cloning strategy. The chimeric protein constructs Linker-His [L-H] and Linker-His-Biotin [LHB] were generated using the following primers: Forward: 5′-GGCCGAATTCATGCTTGCGCTT-3′ and Reverse: 5′-GCCCCTCGAGTCTATCCTATTGTAATGGC-3′. [L-H] and [LHB] were digested with the enzymes SmaI/ClaI and EcoRI/XhoI accordingly before being ligated with the pET28a vectors.
The L-H encoding sequence was subcloned in the plasmid vector pET28a (+) in order to be in frame with the N-terminally located His-tag sequence. Plasmid pET28a (+) was modified for the addition of the biotinylation encoding site after the N-terminal sequence that encodes the 6 × His-tag sequence, and the new plasmid generated was termed pET28a-bio. The DNA fragment encoding the biotinylation site was amplified with Polymerase Chain Reaction with the use of the forward primer (FW): 5′-GGCGCATATGTCCGGCCTG-3′ and reverse primer (RV): 5′-GCGCGAATTCTTCGTGCCA-3′. The amplified biotinylation fragment and the pET28a plasmid were double digested with the NdeI and EcoRI restriction enzymes. The sequence encoding the Linker protein was finally subcloned into the pET28a-bio plasmid as described above.
2.6. Recombinant Protein Expression and In-Vivo Biotinylation of the Chimeric Proteins
The plasmid vectors containing the gene of interest were transformed into BL21 (DE3) E. coli cells. The vector containing the LHB protein gene was co-transformed into E. coli cells with the pBirAcm vector encoding BirA Ligase in order to catalyze biotin binding to the translated proteins. The transformed cell cultures of the E. coli strain BL21 (DE3) carrying the respective vectors were grown at 37 °C in 1 L of LB (Luria Bertani) medium with the corresponding antibiotics (kanamycin for PET28 and chloramphenicol for pBirA plasmid) until the culture density reached an OD600 value of 0.5–0.6. Protein overexpression was induced by 1 mM IPTG plus 1.2 mg/mL biotin dissolved in 10 mL of 10 mM bicine buffer pH: 8 in the case of the LHB construct when biotinylation was desired. Cell cultures were incubated for 4 h at 37 °C aerated by vigorous shaking at 250 rpm. Samples were collected before and after IPTG induction. Cells were harvested by cell culture centrifugation at 6000 rpm for 15 min.
2.7. Inclusion Bodies Lysis and Purification of the Recombinant Proteins
The cell pellet was resuspended in 20 mL of lysis buffer containing 20 mM Tris-HCl pH: 9, 100 mM NaCl, 2% Triton, 1× cOmplete™ Protease Inhibitor Cocktail solution, 1 mM EDTA and 5% glycerol. The bacterial cell lysis was performed by three freeze–thaw cycles at −80 °C and a subsequent submersion in a 37 °C water bath. Addition of 2 mg/mL lysozyme at a final concentration with 1 h incubation at 37 °C followed with subsequent three freeze–thaw cycles. The cell suspension was incubated for 30 min at room temperature while shaking with DNase I at a final concentration of 100 μg/mL and 100 mM MgSO4. The cell suspension was finally passed through a French press at 1000 psi in order to achieve the maximum lysis of any remaining intact cells.
The majority of the expressed protein was stored in the inclusion body fraction. The pellet of inclusion bodies after lysis was washed under shaking for 10 min with urea 1 M, then Triton 1% and 20 mM Tris-HCl pH: 9, followed by centrifugation in order to remove non-specifically bound impurities. The resultant pellet was resuspended at the initial volume in 6 M urea, 20 mM Tris-HCl pH 9, NaCl 0.5 M, Imidazole 10 mM resuspension buffer and was stirred overnight. The remaining insoluble products were subsequently separated by centrifugation at 7000 rpm for 30 min and discarded. The proteins were purified from the cell lysate by Ni-NTA affinity chromatography column. 2 mL of bulk Ni-NTA beads was used as stationary phase. The beads were deposited in a gravity flow column and the supernatant was passed twice. Then, the beads were washed with 8 volumes of two washing buffers with increasing imidazole concentration (10 and 30 mM). The protein was eluted with increasing imidazole concentration solutions (50, 75, 100, 150, 250 and 500 mM). Elution and washing buffers also contained 0.5 M NaCl to reduce non-specific binding of proteins.
2.8. Protein Refolding-Dialysis
Following purification, the elution fractions containing the protein of interest were pooled and sealed in a 10–14 kDa molecular weight cut-off dialysis tubing cellulose membrane. The membrane was transferred to a beaker containing 2 L of the desired exchange buffer. Phosphate buffer 20 mM with 0.2 M NaCl was used for the further labeling of the protein with NHS-fluorescein. Tris-HCl pH: 9, 0.2 M NaCl and 5% glycerol were used for all the other assays. In order to achieve exhaustive dialysis, the membrane was immersed initially into the dialysis buffer for 1 h and subsequently relocated into a fresh dialysis buffer and left for buffer exchange overnight at 4 °C. The dialyzed samples were concentrated using an Amicon centrifugal filter (30 kDa molecular weight cut-off) in order to achieve increase in protein concentration.
2.9. Western Blot
The purified proteins were analyzed by SDS-PAGE and Western Blot in order to detect the successful incorporation of His-tag sequences and biotin and for the distinction of the potential degradation products. For His-tag, the mouse Penta-His Antibody BSA-free (Qiagen) was used as primary antibody and for the biotin molecule, streptavidin conjugated with alkaline phosphatase (Sigma-Aldrich, St. Louis, MO, USA) was solely used for detection. Briefly, purified proteins were separated in a 7.5% SDS-PAGE and transferred to PVDF membranes. The membrane was blocked by mild shaking in 5% w/v skimmed milk in 1× phosphate-buffered saline (PBS) for 1 h at room temperature. After blocking, the PBS-milk was removed and the membrane was soaked in a solution of PBS-milk containing the primary antibody diluted 1:2000 and incubated overnight at 4 °C under mild shaking. After 3 washes by shaking for 15 min in a solution of 1× PBS with 0.04% Tween 20, the blots were incubated in PBS-milk containing the secondary antibody (anti-mouse IgG-alkaline phosphatase) diluted 20,000-fold for 2 h at room temperature under mild shaking. Three washes followed and the protein bands were visualized with the addition of 176 μL/10 mL NBT-BCIP substrate (Sigma) in alkaline phosphatase buffer (0.1 M Tris-HCl pH: 9, 0.1 M NaCl, 5 mM MgCl2). The same methodology was used for the detection of the fibritin foldon motif. The T4 fibritin anti-wac Antibody (Cusabio) was used as the primary antibody at the ratio of 1:1000 and anti-rabbit IgG conjugated with alkaline phosphatase as a secondary antibody at a ratio of 1:20,000. For biotin detection, the blot was incubated with streptavidin conjugated with alkaline phosphatase for 1 h at room temperature and was subsequently washed with PBS-Tween20. Following a rapid wash with alkaline phosphatase buffer, the chemiluminescent detection was performed with the addition of the BCIP-NBT substrate. The reaction was stopped by rinsing and substituting the reaction buffer with distilled water.
2.10. TEM Observation
Transmission Electron Microscopy after negative staining was used to observe the conformation of the LHB protein following purification and refolding. A protein sample solution (concentration 20 μg/mL) of 8 μL was deposited onto a carbon/formvar-covered copper grid. The sample solution was left onto the grid for 2 min and the excess solution was removed using a Whatman filter paper. Subsequently, 8 μL of 1% Uranyl acetate was applied, left for 2 min and the excess solution was removed using a filter paper. Samples were observed in a JEOL JEM 2100 High-Resolution microscope, operating at 200 kV (University of Crete, Biology Department).
2.11. Thermostability Evaluation with SDS-PAGE and Coomassie Blue Staining
10 μL of the LHB protein at a 0.4 mg/mL concentration that was previously stored at 4 °C was mixed with 5 μL sample buffer containing 0.1% SDS. The sample was immediately frozen at −20 °C. 10 μL of LHB protein of the same concentration was left at RT for 30 min and subsequently mixed with 5 μL sample buffer containing 0.1% SDS. The sample was also immediately frozen at −20 °C. 10 μL of the LHB protein at a 0.3 mg/mL concentration that was previously stored at 4 °C was added into a preheated Eppendorf tube of 1.5 mL capacity which was placed in a heat-block in order to equilibrate in the desired temperature. Starting from 37 °C, the aforementioned procedure was repeated while the temperature was increased reaching specific temperature points every 30 min. The protein sample was incubated at the desired temperature for 30 min. After the 30-min incubation, 5 μL sample buffer (0.1% SDS) was added and the sample was frozen at −20 °C. Temperature points for 30 min incubation each were the following: 4, 25, 37, 50, 60, 70, 80, 100 °C. The sample at 100 °C was prepared twice in two different tubes. In the first, the sample buffer added contained 0.1% SDS and in the other, it was 2%. The samples were loaded on a 7.5% SDS-PAGE gel and electrophoresed in the cold room at a constant voltage of 150 V; they were subsequently stained with Coomassie Brilliant Blue R-250 for the detection of the folded and unfolded protein samples.
2.12. Chymotrypsin Digestion
The susceptibility of the LHB protein to proteases was assessed with the chymotrypsin digestion assay. Chymotrypsin was incubated with the LHB protein (conc: 0.2 mg/mL) at protease-to-protein molar ratios of 1:200 and 1:50, for 10 min at RT. As a control, an untreated protein sample was used. The activity of the protease was quenched with the addition of 2% SDS sample buffer and immediate freezing at −20 °C. The samples were loaded on a 7.5% SDS-PAGE gel and electrophoresed at 180 V. Thawed protein samples to be electrophoresed were either unboiled to assess the digestion effect on the trimeric form of the protein or heated at 100 °C for 5 min before SDS-PAGE to evaluate the digestion effect on the monomer form. Unboiled digested samples were electrophoresed in the cold room whereas the boiled samples were at RT conditions. Western Blot analysis was performed as described above, using the anti-His and the foldon wac antibody to detect the terminal end(s) where the protease starts trimming.
2.13. Protein Binding with NHS-Fluorescein and Nanoparticles Conjugated with Fluorophores
0.2 mg/mL of protein in 20 mM phosphate buffer pH: 9, NaCl: 0.2 M was conjugated with a 15-fold molar excess of NHS-Fluorescein (N-Hydroxysuccinimide (NHS)-ester Fluorescein) and incubated for 2 h at room temperature. The unbound fluorescein was removed by exhaustive overnight dialysis in 20 mM phosphate buffer pH: 9, NaCl: 0.2 M and the protein was stored at 4 °C. 4-fold of the biotinylated and NHS-fluorescein labeled LHB protein (80 μg/mL) was incubated with the labeled NPs for 1 h at room temperature. The excess of the unbound streptavidin fluorescent NPs was removed by incubating the mixture into a biotin-coated plate (Thermo-Scientific). LHB labeled with NHS-fluorescein was used for the binding assay of the streptavidin-Alexa-NP for confocal microscopy co-localization observations.
2.14. LHB Protein Linkage with Active Enzymes through the Biotin-Streptavidin Bond
The biotinylated LHB protein is interacting with the streptavidin–alkaline phosphatase (strept-AP) and the streptavidin–horseradish peroxidase (strept-HRP) conjugates through the biotin–streptavidin non-covalent bond. Samples of various ratios of LHB:strept-AP or LHB:strept-HRP (1:3, 1:1, 3:1) were incubated for 2 h at RT. LHB had a fixed concentration at 0.4 mg/mL. The optimal binding ability was verified in a 1% agarose gel with the addition of a native protein sample buffer (without SDS). The gel was run in the cold room at 65 V.
2.15. Cell Internalization and Trafficking of the Proteins and Protein-NP Conjugates
NIH3T3 fibroblast and HeLa cell lines were cultured at 37 °C, 5% CO2 in DMEM supplemented with 10% fetal bovine serum and 50 μg/mL gentamycin. 8 × 104 HeLa or NIH3T3 fibroblast cells were seeded for 24 h in a 24-well plate after addition of a 13 mm TC Coverslip at the bottom of the well. The culture medium (DMEM) was replaced and a 50 μL protein sample containing 50 μg/mL of the fluorescein labeled protein or protein-NPs conjugate, diluted in 0.5 mL DMEM, was added for overnight incubation. Culture media was aspirated, and the cells were carefully washed two times with PBS 1×. The cells were fixed with 4% formaldehyde for 15 min, washed twice with PBS 1× and then the coverslip was mounted on a microscope coverslip containing a drop of the DAPI nuclear staining dye. The internalization and subcellular localization of the proteins was assessed with a Leica SP8 inverted confocal microscope at excitation/emission wavelengths of 488/520 nm for the Fluorescein–protein conjugate, excitation/emission of 360/460 nm for DAPI nucleus stain and excitation/emission of 594/617 nm for the Alexafluor594-bound NPs.
2.16. MTT Cell Proliferation Assay
The cell viability in the presence of the LHB protein only was studied by monitoring the conversion of Thiazolyl Blue Tetrazolium Bromide reagent (MTT) into formazan by the mitochondrial dehydrogenases of the living cells. NIH3T3 fibroblast and HeLa cells with concentrations of 7 × 103 cells/well were cultured in a 96 plate for 24 h. Removal of the medium was followed by treatment of the cells with increasing concentrations (10–800 μg per well) of the protein, suspended in a total volume of 200 μL of culture medium. Cells that were not treated with the protein served as control. After 48 h incubation, the medium was carefully removed and replaced with 90 μL of fresh medium and 10 μL of MTT (5 mg/mL) dissolved in PBS 1×. The cells were incubated for 4 h to allow the development of the purple formazan products and the MTT-culture medium was substituted with 100 μL of isopropanol-DMSO 1:1 solution. The formazan crystals were allowed to dissolve for 15 min at 37 °C. The absorbance was measured at 570 nm in a Synergy HTX BioTEK Plate Reader.
2.17. Measurement of the Internalization Efficacy and Enzymatic Activity of LHB-Strept-AP
The LHB protein at a concentration of 50 μg/mL was incubated with the streptavidin-Alkaline Phosphatase (AP) conjugate at a 1:1 ratio. The final volume of the mixture was 50 μL in 20 mM Tris-HCl pH: 9, 0.2 M NaCl. After 1-h incubation, the mixture was transferred to a biotin-coated plate for 30 min in order to remove the excess of the streptavidin-AP conjugate. The enzyme-linked LHB was added to overnight precultured NIH3T3 cells (initial cell density 8 × 104/well) in a 24 well plate. HeLA cells were not used for this assay due to the high inherent production of AP in this cell line. The conjugates were incubated with 500 μL of DMEM at 37 °C overnight and the following day, the medium was removed and the cells were washed two times with PBS 1×. The internalization and the enzymatic activity of Alkaline phosphatase were assessed by the addition of 7 μL of the substrate BCIP-NBT in each well containing 300 μL of PBS. The reaction product has a blue color and is insoluble in water, therefore rendering the cells visible with a dark coloring where the product is located. As controls, untreated cells and the streptavidin-AP alone at the same concentration were used. The cells were visualized with an optical microscope.
2.18. Measurement of the Internalization Efficacy and Enzymatic Activity of LHB-Strept-HRP
The LHB–streptavidin–Horseradish peroxidase (LHB-strept-HRP) was formed and incubated with the cells similarly as described for the LHB–streptavidin–Alkaline Phosphatase conjugate. HeLa cancer cells were also used for these assays since there is no inherent production of HRP in this cell line. After 24-h incubation of the conjugates with the cells, the enzymatic activity and internalization efficacy of the LHB conjugate was assessed by 3,3′-diaminobenzidine (DAB) staining and luminescence detection. Untreated cells and streptavidin-HRP alone at the same concentration were used as controls. Each sample and control was assessed in triplicates. For observation using light microscopy, NIH3T3 and HeLa cells were washed twice with PBS to remove excess uninternalized conjugates. 0.1% of Triton X-100 was added for 5 min and the cell washing was followed with PBS prior to DAB staining. HRP activity was detected by the brownish precipitation formed following the addition of 300 μL of 0.05% DAB in the presence of 1 μL H2O2 (30%). The precipitates were left to form for several hours, and the cells were observed with an optical microscope. Quantitative measurement of the HRP enzymatic activity was performed by luminescence detection. The enzymatic activity and transfer ability of LHB linked with strept–HRP was quantified after the 24 h incubation of the cells with the conjugates. LHB:strept–HRP ratio was 1:1, the total volume of the conjugate was 50 μL and the concentration of LHB was 50 μg/mL. Following incubation, the cells were washed twice with PBS. Subsequently, 300 μL of PBS-EDTA 1 mM were added to the well and the well plate was incubated at 37 °C for 5 min for cell detachment. The cells were detached by intense pipetting from the bottom of the well and transferred to a 1.5-milliliter tube. The procedure was repeated for the complete detachment of the cells. The tubes were centrifuged at 5500 rpm for 5 min. The supernatant was removed and the cells were lysed with 30 μL of lysis buffer (25 mM Tris-Phosphate buffer pH: 7.8, 2 mM EDTA, 10% glycerol, 1% Triton X-100, 2 mM DTT) for 30 min. Luminol substrate and H2O2 (SuperSignal WestPico Plus chemiluminescent substrate) were added at equal volumes (25 μL) and the solution was immediately transferred to a 96-well plate for measuring the luminescence in a Synergy HTX BioTEK Plate Reader.
2.19. Statistical Analysis
Statistical analysis was performed using the Student’s t-test. Statistical significance was indicated with the value of probability (p) (* p < 0.05, ** p < 0.01, and *** p < 0.001).
4. Discussion
Fibrous nanostructured objects, including protein nanoneedles and self-assembling peptides, can be used as platforms for biomedical and nanotechnological applications. Elegant examples include protein nanoneedles [
43,
44,
45] and amyloid-type or prion peptides [
46,
47,
48]. These stable nanofibrous biomaterials targeted for such applications should not trigger cytotoxic responses. Moreover, they should adopt a well-defined conformation stable and resistant to harsh conditions. In this study, we designed and studied adenovirus protein fiber constructs for potential applications such as carriers for the delivery of protein and nanoparticle cargoes. We used as a basic structural framework the fibrous shaft segment of the adenovirus fiber protein comprising residues 61–392, connected to the fibritin foldon trimerization motif at the C-terminal end with or without the natural linker sequence (Asn-Lys-Asn-Asp-Asp-Lys-Gly-Ser) between them. These constructs [L] and [NoL] (B and C respectively in
Scheme 1) were overexpressed in the form of inclusion bodies and subsequently were refolded into trimers; however, they were found to be partially digested during the overexpression/purification and/or refolding process. The addition of a 6× His-tag at the N-terminal end of the protein construct C leads to construct [LH] (D in
Scheme 1) and improved the efficiency of the purification process through a one-step Ni-NTA affinity column and most importantly, stabilized the protein against proteolytic degradation during the overexpression/purification and/or refolding processes. As a next step, a fourteen-amino-acid biotinylation sequence was inserted immediately after the N-terminal, His-tagged end of the construct in order to enable the attachment of a biotin moiety in vivo. The protein construct [LHB] (E in
Scheme 1) was overexpressed along with a biotin ligase enzyme (BirA ligase), and the successful attachment of the biotin molecule was verified with a streptavidin-Alkaline phosphatase conjugate in a Western blot. The [LHB] proteins were overexpressed in abundant quantities in
E. coli and were subsequently studied in order to assess their structure and stability. They were found to self-assemble into a trimeric nanorod conformation as observed with TEM microscopy; furthermore, they were found to be thermally stable up to 70 °C. Thermostability is an important parameter to take into consideration when designing protein carriers for delivery purposes. Any delivery agent should be able to withstand the temperature of the human body or at least stay stable and not rapidly aggregate until the carrying therapeutic molecule reaches its target in the cell. In vitro digestion experiments with chymotrypsin were carried out; at high protease-to-protein ratios, the LHB proteins could be digested from their N-terminus. In an intracellular context, they would presumably release their N-terminally attached protein cargo into the cell cytoplasm. The biotinylation site added to the chimeric protein Linker-His-Biotin (LHB) enables the addition of a biotin molecule during the overexpression process into the bacterial cells and the subsequent in vivo biotinylation. Advantages that are associated with the in vivo biotinylation in contrast to the in vitro biotinylation are the uniform modification of the protein that excludes the use of intervening chemistry. Additionally, there is no need to separate modified products from unreacted ones and moreover, the purification of the ligase enzyme from the overexpressed protein is avoided. Although the LHB protein does not contain the CAR receptor binding motif located in the natural globular head of the fiber, it can internalize into cell lines such as HeLa and NIH3T3 fibroblasts cells. This confirms previous studies that reported internalization of the fiber protein in the absence of the virion and indicates that another currently undeciphered mechanism is allowing the internalization of the fiber; however, deciphering the exact internalization mechanism is beyond the scope of the present study. MTT assays showed no significant cytotoxicity of the LHB protein towards the aforementioned mammalian cell lines. As a next step, we sought to exploit the streptavidin-biotin strong non-covalent bond in order to bind and transfer biotinylated moieties into the cell. The protein LHB construct was able to transfer into mammalian cells (HeLa and NIH3T3 fibroblasts) a streptavidin-conjugated gold nanoparticle with an attached fluorophore. Most importantly, two streptavidin-conjugated model proteins, alkaline phosphate and horseradish peroxidase can be delivered in the cell cytoplasm by the LHB carrier and retain their enzymatically active form.
Taken together, a series of constructs were fabricated to examine possible parameters for the design and engineering of an optimal molecule carrier. The optimal carrier chosen was the LHB protein, an N-terminally His tagged, in vivo biotinylated fibrous construct comprising the adenovirus fiber shaft residues 61–392 connected to the fibritin foldon motif. The experimental results show that LHB proteins are structurally stable, non-cytotoxic, and display functionalities that can be rationally engineered onto the initial protein structural framework. The work presented in this study aimed at establishing the proof-of-principle for the use of adenovirus fiber-based proteins as nanorods for the delivery of nanoparticles and model proteins; future experimental avenues would also include the exploration of the proteins used as therapeutic protein carriers. Of note, the shaft domain carries exposed loops including the extended loop comprising residues 346–355, offering the possibility of site-directed mutagenesis or sequence replacement for targeted anchoring. The display of heterologous viral epitopes towards vaccine development can also be envisaged, the most rational choice being the replacement of these extended loops on the surface of the LHB protein.
The long-term prospects will be the development of multifunctional and modular fibrous nanorod platforms that can be tailored to the application at the sequence level. For example, they could target a specific type of cell or deliver a therapeutic molecule and an imaging nanoparticle at the same time. The present study reports a proof-of-principle for the development of a basic designable scaffold for future applications in nanomedicine. The viral origin of the recombinant protein prompted us to investigate cytotoxicity in two model cell lines, NIH3T3 and HeLa, with no significant cytotoxicity detected for protein concentrations ten times higher than the standard LHB concentrations used in the experiments.
E. coli is the most common and widely accepted system for protein engineering and structural biology investigations. Nevertheless, for any future application destined for use in humans, the endotoxin effects associated with proteins produced in bacterial cultures have to be considered; as such, endotoxin-free expression systems, for example,
E. coli strains [
49]
Lactobacillus lactis [
50] and
Pichia pastoris [
51] could be considered. Moreover, besides the widely used mammalian cells, plant expression systems could be considered as well [
52]. The proven thermostability of the constructs reported here guarantees safe delivery until any cargo reaches its target, and additionally, the thermostability of a carrier significantly lowers its handling and shipping costs. Last but not least, they can serve as protein-only platforms, attached or not to nanoparticles, that could display heterologous viral epitopes and specifically target immune cells towards the development of future vaccines.