Acta Biomaterialia 131 (2021) 472–482 Contents lists available at ScienceDirect Acta Biomaterialia journal homepage: www.elsevier.com/locate/actbio Full length article Multifunctional antibody-conjugated coiled-coil protein nanoparticles for selective cell targeting Marcos Gil-Garcia, Salvador Ventura∗ Institut de Biotecnologia i de Biomedicina and Departament de Bioquímica i Biologia Molecular, Universitat Autònoma de Barcelona, Bellaterra, Barcelona 08193, Spain a r t i c l e i n f o a b s t r a c t Article history: Nanostructures decorated with antibodies (Abs) are applied in bioimaging and therapeutics. However, Received 21 February 2021 most covalent conjugation strategies affect Abs functionality. In this study, we aimed to create protein- Revised 20 June 2021 based nanoparticles to which intact Abs can be attached through tight, specific, and noncovalent interac- Accepted 23 June 2021 tions. Initially considered waste products, bacterial inclusion bodies (IBs) have been used in biotechnology Available online 27 June 2021 and biomedicine. However, the amyloid-like nature of IBs limits their functionality and raises safety con- Keywords: cerns. To bypass these obstacles, we have recently developed highly functional α-helix-rich IBs exploit- Functional inclusion bodies ing the natural self-assembly capacity of coiled-coil domains. We used this approach to create spherical, Coiled-coil submicrometric, biocompatible and fluorescent protein nanoparticles capable of capturing Abs with high Protein nanoparticles affinity. We showed that these IBs can be exploited for Ab-directed cell targeting. Simultaneous deco- Antibody ration of the nanoparticles with two different Abs in a controllable ratio enabled the construction of a Cell targeting bispecific antibody mimic that redirected T lymphocytes specifically to cancer cells. Overall, we describe an easy and cost-effective strategy to produce multivalent, traceable protein nanostructures with the po- tential to be used for biomedical applications. Statement of significance Functional inclusion bodies (IBs) are promising platforms for biomedical and biotechnological applica- tions. These nanoparticles are usually sustained by amyloid-like interactions, which imposes some lim- itations on their use. In this work, we exploit the natural coiled-coil self-assembly properties to create highly functional, nonamyloid, and fluorescent IBs capable of capturing antibodies. These protein-based nanoparticles are successfully used to specifically and simultaneously target two unrelated cell types and bring them close together, becoming a technology with potential application in bioimaging and im- munotherapy. © 2021 The Author(s). Published by Elsevier Ltd on behalf of Acta Materialia Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 1. Introduction nant protein (Fig. S1). For decades, these refractile aggregates have been considered waste products [3]. However, it has gradually been The use of microorganisms as cell factories to produce recom- revealed that IBs might contain a significant degree of properly binant proteins often leads to the formation of protein inclusion folded and functional proteins [4,5]. These nanometric protein par- bodies (IBs) in the cytoplasm [1,2]. The formation of such protein- ticles are mechanically stable [6,7] and their production and pu- based nanoclusters is the consequence of an inside-the-cell assem- rification are cost-effective and fully scalable, making them attrac- bly reaction consisting in the establishment of noncovalent inter- tive and ready-to-use functional nanostructured materials [8–10]. molecular interactions between different copies of the recombi- In this regard, IBs are increasingly used in biotechnological and biomedical applications, i. e., as reusable biocatalysts [11–13] or in cancer therapy, since they can penetrate cells and release antitu- ∗ Corresponding author. moral polypeptides [14,15]. E-mail addresses: marcos.gil.garcia@uab.cat (M. Gil-Garcia), salvador.ventura@uab.cat (S. Ventura). https://doi.org/10.1016/j.actbio.2021.06.040 1742-7061/© 2021 The Author(s). Published by Elsevier Ltd on behalf of Acta Materialia Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 The mechanical stability of IBs relies on the amyloid-like nature pression was induced by treatment with 1 mM isopropyl β-D-1- of the interactions that stabilize these nanoparticles [16–18]. This thiogalactopyranoside (IPTG) for 6 h at 30 °C. The cells were col- amyloidogenic character raises two critical concerns. First, the for- lected after centrifugation at 50 0 0 g for 20 min and the pelleted mation of nonnative intermolecular β-sheet conformations comes cells were resuspended in 400 μL of buffer A (50 mM Tris and at the expense of the native structure, implying that a significant 100 mM NaCl, pH 7.4) supplemented with 4 μL of 100 mM PMSF fraction of the protein inside IBs becomes necessarily inactivated and 6 μL of 10 mg/mL lysozyme. Then, the cells were incubated [18]. Second, together with the release of adequately folded moi- for 1 h at 37 °C, cooled on ice and sonicated for 3 min at 15% eties, the liberation of toxic oligomeric β-sheet species from these amplitude in 1 s cycles. 4 μL of Nonidet P40 (NP-40) was added nanostructures cannot be disregarded. To offer an alternative to to the sonicated cells, and the mixture was incubated for 1 h at classical amyloid-like IBs, we have recently developed functional 4 °C. Next, 12 μL of 1 M MgSO4 and 10 μL of 1 mg/mL DNase I IBs exploiting the antiparallel coiled-coil fold of the ZapB protein. were added and the mixture was incubated for 45 min at 37 °C. This domain promotes intracellular self-assembly into nonamyloid, The mixture was centrifuged at 15,0 0 0 g for 15 min at 4 °C, and nontoxic, α-helix-rich IBs. The native-like nature of the interactions the pellet was washed with buffer A containing 0.5% Triton X-100. that sustain these IBs may preclude the inactivation of ZapB-fused All incubations were performed under agitation. Purified IBs were globular domains during the assembly reaction, resulting in highly washed three times with buffer A, and the remaining detergent active nanometric particles [19]. This high functionality seems to was removed. The purity of the IBs was confirmed by SDS-PAGE, be a common property of coiled-coil stabilized IBs [11,13,20–22]. and the protein concentration was estimated by measuring the ab- Active ZapB-based IBs can be produced in high yields and easily sorbance of IBs dissolved in guanidine hydrochloride at 280 nm purified to homogeneity [19], thus they are amenable for use in using a Specord 200 Plus spectrophotometer (Analytik Jena, Ger- preparative applications. many). The globular nature of proteins confers the main advantage For the production of soluble and His-tagged Z-domain, E. onto protein-based materials, allowing the alteration of material coli BL21 (DE3) (Invitrogen, USA) competent cells were trans- functionality by genetic redesign to fit the intended application formed with the corresponding plasmid and grown aerobically in [23]. We explored the potential of ZapB-based protein nanopar- LB medium supplemented with 50 μg/mL kanamycin. After an ab- ticles to generate functionalized nanostructures by creating a tri- sorbance OD600 of 0.6 was reached, protein expression was in- partite fusion protein consisting of ZapB, as the assembling unit, duced by treatment with 1 mM IPTG for 16 h at 20 °C. The cells sequentially connected to the green fluorescent protein (GFP) and were collected after centrifugation at 50 0 0 g for 20 min and resus- the Z-domain, as the active moieties (ZapB-GFP-Z). GFP is widely pended in buffer A containing 20 mM imidazole and 1 mM PMSF. used for in vivo imaging [24], whereas the Z-domain [25] is an The solution was sonicated on ice, and the supernatant was col- engineered analog of the B-domain of Staphylococcus aureus pro- lected after centrifugation at 15,0 0 0 g for 30 min, and the protein tein A with high affinity (in the nanomolar range) for IgGs [26– was purified by a 5 mL HisTrap FF column (GE Healthcare, USA) 28]. The binding of this allα–helix protein to the Fc region of an using an ÄKTA chromatograph (GE Healthcare, USA). IgG is mainly mediated by hydrophobic and noncovalent interac- α-Synuclein was produced, purified and aggregated as previ- tions involving residues located at helices 1 and 2 (Fig. S2) [29,30]. ously described in Pujols et al. [37]. Considering this, the objective was to create traceable fluorescent submicrometric particles that can be easily decorated with an an- 2.2. Cell fractionation tibody of interest and directed to a specific cellular antigen. Bispecific antibodies (BsAbs) are mAbs engineered to recognize The distribution of the expressed ZapB-GFP-Z protein in E. coli and bind to two different epitopes simultaneously. BsAbs are in- was analyzed by SDS-PAGE. After protein expression, the cells creasingly used in cancer therapy since they can recognize tar- were centrifuged at 50 0 0 g for 20 min and resuspended in buffer gets in cancer cells and immune cells, helping to bring the two A. Then, the cells were lysed by sonication and centrifuged at cell types together and facilitating tumor cell destruction [31–33]. 15,0 0 0 g for 15 min at 4 °C, and the soluble fraction (supernatant) However, the generalized use of these BsAbs in the clinic is hin- was separated from the insoluble fraction (pellet). The insoluble dered by the high costs and low yields of production and the short fraction was resuspended in buffer A to be the same volume as half-life of these molecules in vivo [34]. Dual-targeting nanopar- the soluble fraction, and both fractions were heated for 10 min ticles conjugated to two different monoclonal antibodies (mAbs) at 98 °C. Finally, 10 μL of the different fractions were loaded onto [35] or binding proteins [36] are being developed to bypass these SDS-PAGE gels. obstacles and generate new therapeutic applications. We show how ZapB-GFP-Z IBs behave as spherical and fluorescent protein 2.3. Preparation of antibody-loaded IBs nanoparticles that, when decorated with two different antibodies, promote selective interactions between different cell types, specif- For the preparation of the IgG-IB complexes, 5 μM ZapB-GFP-Z ically redirecting T lymphocytes to tumoral cells. and 5 μM ZapB-GFP IBs were incubated in the presence of 1 μg of IgG for 45 min at room temperature. Due to the nature of the 2. Materials and methods noncovalent but strong interaction between the Z-domain and the IgGs, the preparation of antibody-decorated IBs was performed by 2.1. Protein production and purification simple mixing. Then, the IBs were harvested by centrifugation at 15,0 0 0 g for 20 min and resuspended in buffer A. Next, five wash- The ZapB-GFP-Z gene fragment (the DNA sequence is presented ing steps were performed to remove unbound antibody molecules in the Supplementary Information) was cloned into a pET-28a vec- and prevent nonspecific binding. tor between the NcoI and BamHI restriction sites. For the competition assay with ZapB-GFP-Z IBs and soluble Z- For the production of ZapB-GFP and ZapB-GFP-Z IBs, Es- domain protein, 1 μg of IgG labeled with Alexa Fluor 555 was in- cherichia coli (E. coli) BL21 (DE3) (Invitrogen, USA) competent cubated in the presence of 5 μM ZapB-GFP-Z IBs, either with an cells were transformed with the corresponding plasmids and excess (50 μM) or without soluble Z-domain. After incubation for grown aerobically in Luria-Bertani broth (LB) medium supple- 45 min at room temperature, the samples were centrifuged, and mented with 100 μg/mL ampicillin and 50 μg/mL kanamycin, re- the presence of the IgGs labeled with Alexa Fluor 555 in the insol- spectively. After absorbance OD600 of 0.6 was reached, protein ex- uble fraction (corresponding to the IBs) was analyzed by measuring 473 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 their fluorescence in a Jasco FP-8200 fluorescence spectrofluorom- 2.9. Scanning electron microscopy (SEM) and dynamic light eter (Jasco Corporation, Japan). scattering (DLS) For the determination of the antibody-binding capacity of the ZapB-GFP-Z IBs, 0.5 μg of IgG labeled with Alexa Fluor 555 was in- Scanning electron microscopy (SEM) was used to analyze the cubated in the presence of increasing concentrations of ZapB-GFP-Z morphology of the undecorated and antibody-decorated ZapB-GFP- IBs (from 0.5 to 20 μM) for 45 min. Then, the GFP and Alexa Fluor Z IBs. Specifically, 10 μL of 5 μM IBs with and without antibodies 555 fluorescence of the IBs was measured using a Jasco FP-8200 resuspended in H2 O were deposited on silicon wafers (Ted Pella fluorescence spectrofluorometer (Jasco Corporation, Japan). Inc., USA), air-dried and observed using a Merlin scanning electron microscope (Zeiss Merlin, Germany) operating at 2 kV. Dynamic light scattering (DLS) was used for quantitative size 2.4. Stability in bovine serum determination of 5 μM ZapB-GFP-Z IBs undecorated or decorated with antibodies. For the determination of the stability of the ZapB- ZapB-GFP-Z IBs were incubated in bovine serum at room tem- GFP-Z IBs, the size of the nanoparticles was monitored for 15 days. perature for 15 days. Then, samples obtained at different time The size of the IBs was determined using a Zetasizer Nano S90 points were loaded onto SDS-PAGE gels to evaluate the integrity (Malvern Instruments Limited, UK) at 25 °C. of the ZapB-GFP-Z protein fusion. 2.10. Congo red (CR) binding assay 2.5. Epifluorescence microscopy A Congo red (CR) assay was performed to determine the pres- Ten microliters of 5 μM decorated IBs in buffer A were de- ence of amyloid structures. For the CR assay, 5 μM ZapB-GFP-Z IB posited on top of microscopy glass slides and covered with cover- was resuspended in buffer A and mixed with CR to a final concen- slips. Fluorescence was observed using an Eclipse Ts2R-FL inverted tration of 10 μM CR. Aggregated α-synuclein was used as a positive microscope (Nikon, Japan) with a C-LED385 filter for eFluor 450 control for protein amyloid aggregation. Optical absorption spectra fluorescence (excitation (390/38 nm) and emission (475/90 nm)), were recorded in the range from 375 to 700 nm in a Specord 200 C-LED470 filter for GFP fluorescence (excitation (470/40 nm) and Plus spectrophotometer (Analytik Jena, Germany). The spectrum of emission (534/55 nm)) and C-LED525 filter for Alexa Fluor 555 flu- the protein alone was acquired to subtract protein scattering. orescence (excitation (525/50 nm) and emission (597/58 nm)). 2.11. Fluorescence spectra 2.6. Confocal microscopy The emission spectrum of Alexa Fluor 555 IgG was obtained by Ten microliters of the solution containing 5 μM ZapB-GFP-Z IBs recording the emitted fluorescence between 560 and 700 nm using decorated with two different IgGs labeled with eFluor 450 and an excitation wavelength of 555 nm. Alexa Fluor 555 were deposited on top of microscopy glass slides The emission spectrum of eFluor 450 IgG was obtained by and covered with coverslips. Then, the fluorescence intensity of the recording the emitted fluorescence between 440 and 660 nm using different fluorophores (eFluor 450, GFP and Alexa Fluor 555) was an excitation wavelength of 410 nm. observed upon excitation at a specific wavelength for each fluo- The emission spectrum of the ZapB-GFP-Z IBs in buffer A was rophore in a Leica SP5 confocal fluorescence microscope (Leica Mi- obtained by recording the emitted fluorescence between 500 and crosystems, Germany). Undecorated ZapB-GFP-Z IBs at a concen- 600 nm. The excitation spectrum was obtained by exciting the tration of 5 μM were used as negative controls, as no fluorescence samples at a range of 40 0–50 0 nm. signal corresponding to eFluor 450 or Alexa Fluor 555 fluorophores Three spectra were accumulated at 25 °C using a Jasco FP-8200 was observed. fluorescence spectrofluorometer (Jasco Corporation, Japan) with slit widths of 5 nm, a 0.5 nm interval, and a 10 0 0 nm/min scan rate for each sample. 2.7. Fourier transform infrared spectroscopy (FTIR) For the thermal unfolding of unloaded and IgG-loaded ZapB- GFP-Z IBs, the GFP fluorescence signal was recorded in a range of ZapB-GFP-Z IBs at 5 μM were washed with H2 O to remove the 25–95 °C with an increasing heat rate of 1 °C/min and slit widths presence of salts, placed on ATR crystals and dried under N2 flow. of 5 nm and 600 rpm. ZapB-GFP-Z IBs were excited at 495 nm and The experiment was carried out in a Bruker Tensor 27 FTIR (Bruker emission was recorded at 514 nm using a Jasco FP-8200 fluores- Optics, USA) supplied with a Specac Golden Gate MKII ATR acces- cence spectrofluorometer (Jasco Corporation, Japan). sory. The spectrum consisted of 32 acquisitions measured at a res- olution of 1 cm−1 . The data were acquired and normalized using 2.12. Toxicity assay OPUS MIR Tensor 27 software (Bruker Optics, USA). The IR spec- trum was fitted employing the nonlinear peak-fitting equation in HeLa cells and MRC-5 cells were seeded into 96-well plates and PeakFit package v4.12 (Systat Software, USA). The area for each cultured for 24 h in Dulbecco’s Modified Eagle Medium (DMEM) Gaussian curve was calculated in the amide I region from 1700 to and Minimum Essential Medium α (MEM α), respectively, supple- 1600 cm−1 using the second derivative deconvolution method with mented with 10% Fetal Bovine Serum (FBS). The ZapB-GFP-Z IBs PeakFit package v4.12 (Systat Software, USA). were resuspended in PBS at pH 7.4 and added at a concentration ranging from 2 to 12 μM. PBS was used alone as a control. 2.8. Circular dichroism (CD) spectroscopy For the evaluation of the potential antibody-associated ZapB- GFP-Z IBs toxicity, HeLa cells were seeded into 96-well plates and CD spectroscopy was used to analyze the secondary structure cultured in DMEM supplemented with 10% FBS for 24 h. ZapB-GFP- content of the ZapB-GFP-Z IBs. Specifically, 5 μM ZapB-GFP-Z IBs Z IBs loaded with anti-EGFR IgGs were resuspended in PBS at pH were resuspended in buffer A, and the far-UV CD spectrum was 7.4 and added at a concentration ranging from 2 to 12 μM. PBS recorded at a 1 nm bandwidth, scan speed of 100 nm/min and alone and anti-EGFR IgGs were used as controls. response time of 1 s using a Jasco-815 spectropolarimeter (Jasco In the case of the potential toxicity induced by ZapB-GFP-Z IBs Corporation, Japan) at 25 °C. loaded with anti-CD3 IgGs, T cells (Jurkat cell line) were seeded 474 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 into 96-well plates cultured in Roswell Park Memorial Institute a FITC laser to detect the green fluorescence of GFP. Fluorescence (RPMI) 1640 + GlutaMAX medium supplemented with 10% FBS for intensities of cell-bound IBs were analyzed and quantified using 24 h. ZapB-GFP-Z IBs decorated with anti-CD3 IgGs were resus- FlowJo (BD Biosciences, USA). pended in PBS at pH 7.4 and added at a concentration ranging from For the evaluation of the localization of HeLa and T cells in the 2 to 12 μM. PBS alone, anti-CD3 IgGs and undecorated ZapB-GFP-Z presence of dual-decorated (anti-EGFR and anti-CD3 IgGs) ZapB- IBs at 10 μM were used as controls. GFP-Z IBs, HeLa cells were cultured on a 35 mm cell culture dish Treated and control cells were incubated at 37 °C in 5% CO2 (SARSTEDT AG & Co, Germany) to a confluence of 70–80%. Then, for 72 h. Then, 10 μL of PrestoBlue cell viability reagent (Thermo the initial medium was replaced with T cells and dual antibody- Fisher Scientific) was added and incubated for 30 min. Cell viabil- decorated 10 μM ZapB-GFP-Z IBs (anti-EGFR and anti-CD3 IgG). ity was determined by quantifying the fluorescence with a Victor3 Cells with decorated IBs were incubated for 1 h at 37 °C and fluorescent plate reader (Perkin Elmer, USA). 5% CO2 and washed five times with fresh medium. ZapB-GFP-Z IBs decorated with only anti-EGFR IgGs were used as controls. 2.13. Statistical methods Then, trypsin was added to the solution and neutralized with FBS- supplemented medium. Then, the cells were centrifuged, resus- All experimental values are presented as the means ± standard pended in PBS at pH 7.4 and analyzed using a FACSCanto flow cy- error of the mean (SEM). Measurements are taken in duplicate. tometer (BD Biosciences, USA). FSC and SSC signals were measured GraphPad Prism software (GraphPad Software, USA) was used to for the detection of both cell types, and untreated HeLa and T cells calculate the mean and SEM values. were used alone as controls. Finally, the percentage and quantifi- cation of the cells were determined using FlowJo (BD Biosciences, 2.14. Specific cell targeting by decorated IBs USA). HeLa cells were cultured on a 35 mm cell culture dish (SARST- 3. Results EDT AG & Co, Germany) to a confluence of 70–80%. The initial DMEM was replaced with fresh medium containing 10 μM ZapB- 3.1. ZapB-GFP-Z forms nonamyloid α-helix-rich IBs GFP-Z-anti-EGFR IgG IB labeled with Alexa Fluor 555 and prepared as previously described in the “Preparation of antibody-loaded IBs” To generate nonamyloid fluorescent IBs with antibody capturing section. Then, HeLa cells were incubated with ZapB-GFP-Z-anti- activity, ZapB was N-terminally fused to two consecutive globular EGFR IgG IBs for 1 h at 37 °C and 5% CO2 , and ZapB-GFP-Z-IgG anti- domains: GFP and the Z-domain. This created a tripartite fusion rabbit IBs were used as controls. The incubated cells were washed protein named ZapB-GFP-Z. The Z-domain [25] is an engineered five times with fresh medium to remove unbound molecules. analog of the B domain of Staphylococcus aureus protein A, consists T cells were maintained in RPMI 1640 + GlutaMAX medium of 58 residues (6.6 kDa), and folds into three α-helices. Therefore, supplemented with 10% FBS. T cells were harvested and resus- the fusion protein consisted of sequentially ordered coiled-coil, all- pended in fresh medium containing 5 μM ZapB-GFP-Z-anti-CD3 β, and all α- domains (Fig. S3). IgG IB labeled with eFluor 450. The cells were incubated with ZapB-GFP-Z was expressed in E. coli and located in the insolu- ZapB-GFP-Z-anti-CD3 IgG IBs for 1 h at 37 °C and 5% CO2 , precip- ble cell fraction (Fig. S4), from which IBs composed of assembled itated and washed five times with fresh medium. ZapB-GFP-Z-IgG copies of the fusion protein were purified to homogeneity (Fig. S5) anti-rabbit and anti-EGFR IBs were used as controls. at a yield of 98.5 ± 5.5 mg/L. The secondary structure of the IBs For the binding of unrelated cell types, HeLa cells were cul- was analyzed using Fourier Transform Infrared Spectroscopy (FTIR). tured on a 35 mm dish cell culture (SARSTEDT AG & Co, Germany) We recorded the infrared spectrum of the ZapB-GFP-Z IBs in the to a confluence of 70–80%. Then, the initial medium was replaced amide I region (170 0–160 0 cm−1 ), corresponding to the absorp- with T cells and dual antibody-decorated 10 μM ZapB-GFP-Z IBs tion of the main chain carbonyl group. As shown in Fig. 1A, the (anti-EGFR labeled with Alexa Fluor 555 and anti-CD3 labeled with primary contributor to the ZapB-GFP-Z IBs spectrum was a signal eFluor 450). The cells with antibody-decorated IBs were incubated located at 1654 cm−1 (44% of the area), assigned to the α-helix for 1 h at 37 °C and 5% CO2 and washed five times with fresh conformation. This outcome is reasonable since the ZapB and Z- medium. IBs decorated with only anti-EGFR IgGs were used as con- domains both consist of α-helices. A second band at 1631 cm−1 ac- trols. counts for 35% of the spectrum area and likely represents the low- Finally, the cells incubated with decorated IBs were visualized frequency coiled-coil signal [38] and the β-sheet signal of the GFP using an Eclipse Ts2R-FL inverted microscope (Nikon, Japan), and β-barrel. The third band at 1676 cm−1 (21% of the area) possibly is images were acquired using the appropriate filters and brightfield. attributable to the mixed contribution of structurally different do- mains at high frequencies. Although the far-UV Circular Dichroism 2.15. Flow cytometry assay (CD) spectrum of the ZapB-GFP-Z IBs is influenced by the β-barrel structure of GFP, the two main α-helical signals are still evident, ZapB-GFP-Z IBs were decorated with anti-EGFR IgG labeled with with a minimum at 224 nm and an inflection at 210 nm, in good Alexa Fluor 555 as previously described. Then, HeLa cells were in- agreement with the FTIR data (Fig. S6) and with the previous sec- cubated with the decorated IBs for 1 h, pelleted and washed five ondary structure analysis of ZapB IBs and ZapB-GFP IBs [19]. These times to discard unbound molecules. HeLa cells incubated with results corroborate the native-like secondary structure of the ZapB- ZapB-GFP IBs (previously incubated with anti-EGFR IgGs), ZapB- GFP-Z polypeptides embedded in these IBs. GFP-Z IBs decorated with anti-rabbit IgG, ZapB-GFP-Z IBs alone, Scanning Electron Microscopy (SEM) revealed that the ZapB- and buffer alone were used as controls. For the T cell experiments, GFP-Z IBs were of submicrometric size and displayed a spheri- the procedure was the same as previously described for HeLa cells, cal shape (Fig. 1B). Dynamic light scattering (DLS) data revealed decorating ZapB-GFP-Z IBs with anti-CD3 IgG. In this case, T cells a moderately polydisperse size distribution with an average diam- incubated with ZapB-GFP IBs (previously incubated with anti-CD3 eter of 560 ± 157 nm (Fig. S7) that remained stable in size for at IgGs), ZapB-GFP-Z IBs decorated with anti-rabbit IgG, ZapB-GFP-Z least 15 days (Fig. S8). This submicrometric size is the result of the IBs alone, and buffer alone were used as controls. high order assembly of multiple copies of the ZapB-GFP-Z fusion A total of 10,0 0 0 cells incubated with IBs were analyzed using protein inside the bacteria, already before their purification. To as- a FACSCanto flow cytometer (BD Biosciences, USA) equipped with sess whether ZapB-GFP-Z IBs were stable under close to physiolog- 475 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 Fig. 1. Conformational properties of the ZapB-GFP-Z IBs. (A) FTIR absorbance spectrum of the ZapB-GFP-Z IBs in the amide I region (dashed line). The different component bands (solid lines) were obtained by deconvolution of the spectrum using PeakFit software. (B) SEM image of a ZapB-GFP-Z IB. (C) CR absorbance spectra of the ZapB-GFP-Z IBs (solid line) and buffer alone (dashed line). ical conditions, they were incubated with bovine serum for up to good agreement with our previous studies [19], with excitation and 15 days. The SDS-PAGE analysis revealed that the fusion protein in emission maxima at 495 nm and 514 nm, respectively. To confirm these IBs remained stable for at least 10 days at room temperature the expected bifunctional activity of these IBs, we incubated ZapB- (Fig. S9). GFP-Z IBs with an IgG antibody labeled with Alexa Fluor 555 at An amyloid-like nature for ZapB-GFP-Z IBs was excluded using room temperature for 45 min. Then, the IBs were precipitated and one of the most common dyes to detect amyloid-like structures, washed five times to minimize nonspecific binding, and the fluo- the Congo red (CR) [39]. The CR absorption spectrum is sensitive rescence of these nanoparticles was analyzed by fluorescence mi- to the presence of amyloid structures, displaying a redshift in their croscopy. The ZapB-GFP-Z IBs emitted the characteristic GFP green presence, caused by the cooperative binding of CR molecules along fluorescence and the labeled antibody red fluorescence, with a per- the fibril axis [40]. As shown in Fig. 1C, the spectra of CR dye in fect overlap between both signals (Fig. 2). When ZapB-GFP IBs the presence and absence of ZapB-GFP-Z IBs were very similar. In without the Z-domain were used as controls, green fluorescence contrast, when CR was incubated with aggregated α-synuclein, the was evident, but we could not detect red fluorescence coming from displacement of the spectrum was evident (Fig. S10). the labeled IgG (Fig. S13), indicating that the Z-domain was essen- Finally, the thermal stability of the fusion protein within the tial for the binding of IgG to the IBs. The ability of the ZapB-GFP-Z ZapB-GFP-Z IBs was analyzed by monitoring the changes in the IBs to bind antibodies was tested in the presence and in the ab- GFP fluorescence intensity upon heating. As shown in Fig. S11, the sence of 10-fold molar excess of soluble Z-domain. As shown in melting curve is indicative of cooperative GFP unfolding with a Tm Fig. 3A, the binding of these IBs to antibodies labeled with Alexa of 74.5 ± 3.0 °C. Fluor 555 was strongly reduced in the presence of the competing Overall, the biophysical analysis indicates that ZapB-GFP-Z IBs soluble Z-domain, as it was evident by the much lower fluores- are α-helix-rich and stable nanometric spherical particles, likely cence intensity. This indicates that, as intended, the binding of the sustained by coiled-coil interactions, and devoid of amyloid-like in- antibodies to the IBs is mediated by the Z-domain. termolecular β-sheet structures. To determine the IgG loading capacity of these IBs, we per- formed a titration assay by measuring the fluorescence of the la- 3.2. ZapB-GFP-Z IBs are bifunctional beled IgG bound to the IBs. As shown in Fig. 3B, the GFP fluo- rescence of these IBs increases linearly with the concentration. In The conformational traits of ZapB-GFP-Z IBs suggest that the contrast, the binding of the Alexa Fluor 555-labeled IgG reached a different domains in the protein fusion might maintain their native plateau at 2 μM ZapB-GFP-Z IB, revealing a conventional saturation structure in the assembled state. Therefore, they were expected binding curve, with an apparent binding capacity of 130 ng of an- to emit green fluorescence and bind antibodies simultaneously. To tibody per μg of IB and 867 ng of antibody per μg of Z-domain in confirm that the spectral properties of the GFP embedded in the the IB (Fig. 3C). ZapB-GFP-Z IBs were preserved, the excitation and emission spec- We characterized the size and morphology of these nanoparti- tra were recorded. As shown in Fig. S12, the GFP located in the cles when loaded with antibodies using DLS and SEM. The pres- IBs displays the characteristic excitation and emission profiles, in 476 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 Fig. 2. Bifunctional activity of the ZapB-GFP-Z IBs. Representative fluorescence microscopy images of (A) GFP fluorescence and (B) IgG labeled with Alexa Fluor 555 fluores- cence. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.) ence of the IgGs increased the average diameter of the IBs to cells were labeled with anti-EGFR decorated IBs (Fig. S16), whereas 724.8 ± 292.7 nm, while they kept the spherical shape (Fig. S14). no significant labeling was observed in the other cases. Similarly, The stability of the IBs constituent fusion proteins was not signifi- when ZapB-GFP IBs were incubated with the anti-EGFR antibody, cantly altered by the presence of the IgGs, as it was deduced from and these IBs were added to HeLa cells, the cells remained unla- the cooperative melting curve of GFP, with a Tm of 75.0 ± 2.1 °C, beled. Furthermore, when ZapB-GFP-Z IBs loaded with anti-EGFR fairly similar to that of undecorated IBs (Fig. S11). antibody were incubated with cells deficient in EGFR expression, such as T lymphocytes, the IBs did not recognize them (Fig. S17). 3.3. ZapB-GFP-Z IBs are biocompatible Finally, we wanted to discard any antigen-specific toxicity of the antibody-decorated IBs. ZapB-GFP-Z IBs loaded with anti-EGFR an- Once the functionality of ZapB-GFP-Z IBs was confirmed, we tibody were incubated with HeLa cells. As shown in Fig. S18, these sought to determine whether they induce toxicity in human cells functionalized IBs did not induce any detectable antigen-associated since it would preclude any further cellular assay. HeLa cells were toxicity in the HeLa cells, which displayed viability values similar incubated with different concentrations of ZapB-GFP-Z IBs (from 0 to the cells treated with buffer or anti-EGFR soluble antibody. to 12 μM), and we monitored their viability using a PrestoBlue® Overall, we can conclude that ZapB-GFP-Z IBs can be used to fluorescent assay. As it can be observed in Fig. 4A, ZapB-GFP-Z IBs target specific human cell epitopes and that this activity depends did not show associated toxicity at any of the assayed concen- both on the presence of the Z-domain and the identity of the con- trations. The IBs toxicity was further tested in the MRC-5 nontu- jugated antibody, without any antigen-associated toxicity. The in- mor human lung fibroblast line. As shown in Fig. 4B, ZapB-GFP-Z trinsic green fluorescence of these nanoparticles allows to localize IBs were innocuous to MRC-5 cells at all the assayed concentra- them easily without need for post-assembly IBs labeling. tions, corroborating that these functional protein-based nanoparti- cles have optimal biocompatibility. 3.5. Dual antibody binding by the ZapB-GFP-Z IBs 3.4. Antibody-decorated ZapB-GFP-Z IBs specifically target human In principle, these bifunctional IBs can be endorsed with mul- cancer cells tivalency simply by loading different IgG antibodies onto the Z- domain. To confirm this possibility, ZapB-GFP-Z IBs were simul- We aimed to assess whether ZapB-GFP-Z IBs can target spe- taneously incubated with two different antibodies labeled either cific antigens in living cells once they have been loaded with an- with Alexa Fluor 555 (red fluorescence) or eFluor 450 (blue flu- tibodies through their Z-domain. First, we decorated ZapB-GFP-Z orescence). As shown in Fig. 6, ZapB-GFP-Z nanoparticles emit- IBs with an anti-EGFR antibody labeled with Alexa Fluor 555 as ted three different fluorescence signals, green from GFP, and red described above. Anti-EGFR antibodies recognize epidermal growth and blue fluorescence from the antibodies, with a perfect over- factor receptor (EGFR) overexpressed on the membrane of epithe- lap, indicating that both antibodies were bound to the IBs, as in- lial cancer cells such as HeLa cells [41,42]. We incubated the anti- tended. This dual antibody-loading was confirmed using confocal EGFR antibody-loaded ZapB-GFP-Z IBs with HeLa cells. The pres- microscopy. A good overlap of the three fluorescence signals in ence of IBs attached to the cell membranes was corroborated us- the dual-decorated IBs was observed, without detecting blue or red ing fluorescence and brightfield microscopy (Fig. 5). We observed fluorescence signals in the undecorated ZapB-GFP-Z IBs (Fig. S19). a good overlay between green and red fluorescence signals, emit- Moreover, controlling the proportion of the antibodies in the initial ted from the GFP and the antibody, respectively, indicating that mixture allowed us to obtain IBs decorated with a defined ratio of the nanoparticles maintained their bifunctionality in the assay. In these molecules (Fig. S20). contrast, when HeLa cells were incubated in the presence of ZapB- Overall, the data indicate that ZapB-GFP-Z IBs can be function- GFP-Z IBs loaded with an anti-rabbit antibody, IBs binding to cells alized with the desired proportion of two, and likely more, differ- was not detected (Fig. S15), indicating that the recognition of HeLa ent antibodies. cells by IBs is specific and depends on the presence of anti-EGFR antibodies anchored to their Z-domains. 3.6. Dual antibody-conjugated ZapB-GFP-Z IBs direct T lymphocytes The binding of IBs to cells was also investigated using flow cy- to HeLa cells tometry (FC). We incubated ZapB-GFP-Z IBs alone and loaded with either anti-EGFR or anti-rabbit antibodies with HeLa cells, followed Once we confirmed the ability of ZapB-GFP-Z IBs to capture dif- by a washing step. The cells were immediately analyzed by FC, ferent antibodies simultaneously, we wondered whether they can taking advantage of GFP fluorescence using a FITC emission de- be used as scaffolds to bring two different cell types together. To tector. The analysis revealed that approximately 80% of the HeLa this end, we first analyzed the ability of the ZapB-GFP-Z IBs to tar- 477 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 get T lymphocytes when they were decorated with anti-CD3 IgG labeled with eFluor 450. This anti-CD3 antibody recognizes the TCR/CD3 complex of T cells [43]. ZapB-GFP-Z IBs loaded with anti- CD3 IgG were incubated with T cells for one hour and washed to prevent nonspecific binding. As shown in Fig. 7, the presence of conjugated IBs attached to T cells was evident in the fluorescence and brightfield microscopy images. In addition, there was a per- fect overlap between the green and blue fluorescence emitted from the GFP and anti-CD3 antibody, respectively. When T cells were incubated with ZapB-GFP-Z IBs loaded with anti-rabbit IgG, these nanoparticles did not recognize the T lymphocytes (Fig. S21). The binding of IBs to T cells was also investigated using FC, as previously described with HeLa cells. As shown in Fig. S22, the analysis indicated that approximately 60% of T cells were labeled when incubated with ZapB-GFP-Z IBs conjugated to the anti-CD3 antibody. In contrast, T cells incubated with IBs loaded with non- specific antibodies, unloaded IBs or IBs incubated with the anti- CD3 antibody but devoid of the Z-domain remained unlabeled and were nonfluorescent. To confirm that ZapB-GFP-Z IBs decorated with anti-CD3 an- tibody did not induce antigen-specific toxicity, T cells were in- cubated with different concentrations of anti-CD3-loaded IBs for 72 h. As is shown in Fig. S23, these functionalized IBs did not in- duce any significant toxicity at the tested concentrations. The data indicate that, as shown for HeLa cells, the binding to T lymphocytes is controlled by the specificity of the bound antibody. Next, we loaded ZapB-GFP-Z IBs simultaneously with two antibod- ies, namely, fluorescently labeled anti-EGFR and anti-CD3. These dual-antibody decorated IBs were incubated with HeLa and T cells for one hour, and the culture was washed to remove unbound par- ticles. Brightfield microscopy analysis showed abundant circular T cells and polygonal HeLa cells connected by fluorescent IBs (Figs. 8 and S24). In contrast, when HeLa cells and T cells were incubated in the presence of ZapB-GFP-Z IBs loaded only with anti-EGFR IgG, the IBs targeted HeLa cells but did not capture T lymphocytes (Fig. S25). The IBs mediated colocalization of the two cell types was fur- ther studied using FC. As shown in Fig. S26, the dual decoration of ZapB-GFP-Z IBs with anti-EGFR and anti-CD3 antibodies promoted the association of T cells and HeLa cells. However, in the absence of anti-CD3 antibody, the ZapB-GFP-Z IBs failed to promote this cell association. A quantitative analysis indicated that dual anti-EGFR and anti-CD3 antibody decoration of ZapB-GFP-Z IBs increased the number of T cells associated to HeLa cells by at least 14-fold, when Fig. 3. Antibody-binding capacity of the ZapB-GFP-Z IBs. (A) Alexa Fluor 555 flu- compared with IBs decorated only with anti-EGFR antibody (Table orescence intensity of ZapB-GFP-Z IBs incubated with labeled IgG in the presence S1). and absence of excess of competing soluble Z-domain. (B) Fluorescence intensity of Overall, the data indicate that dual antibody-loaded ZapB-GFP- GFP of IgG-loaded ZapB-GFP-Z IBs as a function of the IBs concentration. (C) Alexa Z IBs can simultaneously bind at least two different antigens and Fluor 555 fluorescence intensity of IgG-bound to ZapB-GFP-Z IBs as a function of approach unrelated cell types. the IBs concentration. Fig. 4. Cytotoxicity induced by the ZapB-GFP-Z IBs. HeLa (A) and MRC-5 (B) cell viability was tested at different concentrations (from 2 to 12 μM) of ZapB-GFP-Z IBs. IBs were added to the cells and incubated for 72 h. PBS buffer was used as the control. 478 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 Fig. 5. Specific binding of functionalized ZapB-GFP-Z IBs to human cells. Representative microscopy images of green fluorescence emitted from the ZapB-GFP-Z IBs (A), red fluorescence emitted from anti-EGFR IgG labeled with Alexa Fluor 555 (B), HeLa cells with brightfield microscopy (C), and the three images merged (D). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 4. Discussion negligible antigen-specific toxicity when they are loaded with an- tibodies, a feature that IgG-decorated ZapB-GFP-Z IBs fulfill. Functional IBs are becoming promising sources of active protein Selective targeting of defined cell types has emerged as a safe for biotechnological and biomedical purposes [44]. These protein alternative to conventional therapies, minimizing toxic side effects. nanoparticles are easy to produce and purify at preparative yields The ability of BsAbs to simultaneously recognize two different epi- and have been applied in tissue engineering, protein delivery, as topes is increasingly exploited in immuno-oncology, gradually dis- vaccines, and for catalysis [45–48]. placing monovalent therapies, since this capacity allows redirection We have recently succeeded in exploiting the ability of a nat- of immune cells to selected cancer cell types [51]. ural coiled-coil domain to self-assemble inside bacterial cells to Dual-targeting nanoparticles that combine two different mAbs generate highly active IBs [19]. These α-helix-rich protein nanopar- are receiving increasing attention as alternatives to BsAbs for can- ticles overcome the drawbacks of classical β-sheet-rich, amyloid- cer immunotherapy [35]. The FDA has approved many monovalent like IBs. Here, we address whether this inside-the-cell coiled-coil- mAbs for targeted cancer treatment, and these mAbs are manufac- driven assembly strategy can be used to create multifunctional IBs tured at a much lower cost than BsAbs. In addition, mAb immobi- containing two different folded and active globular domains. lization in a nanoscopic assembly results in significant stabilization The tripartite ZapB-GFP-Z fusion protein was successfully ex- relative to short-lived soluble BsAbs. Current methods for develop- pressed in E. coli cells as IBs. These nanoparticles have been pu- ing multivalent antibody nanostructures rely on covalent strategies rified to homogeneity, and they display a spherical submicrometric that require consecutive chemical conjugations steps [52], which size that remains stable in solution for more than a week, sug- can potentially modify the antibody structure and function. In con- gesting suitability for their use in nanotechnological or biomedical trast, in ZapB-GFP-Z protein nanoparticles, antibody decoration re- applications [49,50]. In this study, we showed that these IBs are lies on the strong, specific, and noncovalent interaction between sustained by native-like α-helical interactions and that the activ- the Z-domain and the Fc-region of IgG, as shown in the compe- ity of the all-β GFP and all α- Z-domain moieties is preserved, as tition binding assay, which is not expected to alter the antibody evidenced by the perfect overlap between the green fluorescence affinity/specificity for its target. of GFP and the red or blue fluorescence emitted from the labeled Once we demonstrated the ability of antibody decorated ZapB- IgG antibody attached to the Z-domain. This outcome suggests a GFP-Z IBs to recognize different cell types specifically, we ratio- generic ability of proteins to keep their active structures within nalized that our multifunctional nanoparticles can be a convenient ZapB-based IBs, as indicated by the cooperative melting of the GFP way to promote dual epitope targeting when decorated with two moiety. Importantly, ZapB-GFP-Z IBs keep a submicrometric size different specific antibodies and, indeed, dual-loaded nanoparticles when bound to antibodies, without their shape or the constituent effectively redirect T cells to cancer cells. proteins stability being altered. The intrinsic fluorescence of these nanostructures makes them Biomaterials for in vivo applications need to be nontoxic to hu- traceable in vivo, without the need for fluorescent dyes, and they man cells [19]. As intended, generating IBs without amyloid char- can be simultaneously loaded with two, and likely more, antibod- acteristics resulted in innocuous and biocompatible IBs potentially ies in a user-defined stoichiometry. These features, together with suitable for biomedical purposes, as it has been corroborated with the high yield of production confer them advantages, relative to different cell lines. Another requirement for future in vivo uses is BsAbs, from a biotechnological point of view. 479 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 Fig. 6. Multifunctional ZapB-GFP-Z IBs decorated with two different antibodies. Representative microscopy images of GFP fluorescence (A), red fluorescence emitted from IgG labeled with Alexa Fluor 555 (B), blue fluorescence emitted from IgG labeled with eFluor 450 (C), and the three images merged (D). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Fig. 7. Specific binding of bifunctional ZapB-GFP-Z IBs to T lymphocytes. Representative microscopy images of GFP fluorescence (A), blue fluorescence emitted from anti-CD3 IgG labeled with eFluor 450 (B), and merging of these images and brightfield image (C). (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.) 480 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 Fig. 8. Dual IgG-decorated ZapB-GFP-Z IBs direct T cells expressing CD3 to HeLa cells expressing EGFR. Representative microscopy images of GFP fluorescence (A), red fluorescence emitted from anti-EGFR IgG labeled with Alexa Fluor 555 (B), blue fluorescence emitted from anti-CD3 IgG labeled with eFluor 450 (C), and merging of these three images and brightfield (D). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Overall, the present proof-of-concept study describes a novel Acknowledgments kind of protein-based nanoparticles with potential applications in biomedicine, although the in vivo usefulness of this approach We are grateful to the UAB Microscopy Service of Barcelona and needs to be further validated. to Servei de Cultius Cel•lulars, Producció d’Anticossos i Citometria (SCAC) for technical advice. We thank Samuel Peña-Díaz for pro- viding us with purified and aggregated α-synuclein protein. The 5. Conclusion graphical abstract and Fig. S1 were created with Biorender.com. In this work, we created fluorescent nonamyloid IBs able to Supplementary materials capture antibodies with high affinity. These submicrometric, spher- ical, and α-helix-rich protein nanoparticles are innocuous for hu- Supplementary material associated with this article can be man cells. They can potentially recognize any cell type of interest found, in the online version, at doi:10.1016/j.actbio.2021.06.040. when loaded with the proper antibody. Furthermore, these func- References tional IBs can be easily functionalized by adding different anti- bodies simultaneously, promoting the association of unrelated cell [1] A. Schrödel, A. De Marco, Characterization of the aggregates formed during types, acting as mimics of BsAbs. Their production is cost-effective recombinant protein expression in bacteria, BMC Biochem. 6 (2005) 1–11, and fully scalable, and therefore, they are convenient sources of in doi:10.1186/1471- 2091- 6- 10. [2] N. Ferrer-Miralles, J. Domingo-Espín, J. Corchero, E. Vázquez, A. Villaverde, Mi- vivo traceable protein nanoparticles for potential biomedical appli- crobial factories for recombinant pharmaceuticals, Microb. Cell Fact.. 8 (2009) cations. 1–8. 10.1186/1475-2859-8-17. [3] F. Baneyx, M. Mujacic, Recombinant protein folding and misfolding in Es- cherichia coli, Nat. Biotechnol. 22 (2004) 1399–1407, doi:10.1038/nbt1029. [4] U. Rinas, E. Garcia-Fruitós, J.L. Corchero, E. Vázquez, J. Seras-Franzoso, Funding A. Villaverde, Bacterial inclusion bodies: discovering their better half, Trends Biochem. Sci. 42 (2017) 726–737, doi:10.1016/j.tibs.2017.01.005. [5] A. De Marco, N. Ferrer-Miralles, E. Garcia-Fruitós, A. Mitraki, S. Peternel, U. Ri- This work was funded by the Spanish Ministry of Economy and nas, M.A. Trujillo-Roldán, N.A. Valdez-Cruz, E. Vázquez, A. Villaverde, Bacterial Competitiveness (BIO2016–78,310-R) and by the Spanish Ministry inclusion bodies are industrially exploitable amyloids, FEMS Microbiol. Rev. 43 of Science and Innovation (PID2019–105017RB-I00) to S.V and by (2019) 53–72, doi:10.1093/femsre/fuy038. [6] E. García-Fruitós, E. Rodríguez-Carmona, C. Díez-Gil, R.M. Ferraz, E. Vázquez, ICREA, ICREA-Academia 2015 and 2020 to S.V. M.G.G. was sup- J.L. Corchero, M. Cano-Sarabia, I. Ratera, N. Ventosa, J. Veciana, A. Villaverde, ported by the Spanish Ministry of Science and Innovation via a Surface cell growth engineering assisted by a novel bacterial nanomaterial, doctoral grant (FPU16/02465). Adv. Mater. 21 (2009) 4249–4253, doi:10.10 02/adma.20 090 0283. [7] G. Margreiter, P. Messner, K.D. Caldwell, K. Bayer, Size characterization of inclu- sion bodies by sedimentation field-flow fractionation, J. Biotechnol. 138 (2008) 67–73, doi:10.1016/j.jbiotec.2008.07.1995. Declaration of Competing Interest [8] V.D. Jäger, R. Lamm, K. Küsters, G. Ölçücü, M. Oldiges, K.E. Jaeger, J. Büchs, U. Krauss, Catalytically-active inclusion bodies for biotechnology-General con- cepts, optimization, and application, Appl. Microbiol. Biotechnol. 104 (2020) The authors declare that they have no competing interests. 7313–7329, doi:10.10 07/s0 0253- 020- 10760- 3. 481 M. Gil-Garcia and S. Ventura Acta Biomaterialia 131 (2021) 472–482 [9] A. Villaverde, J.L. Corchero, J. Seras-Franzoso, E. Garcia-Fruitós, Functional pro- [32] H. Kantarjian, A. Stein, N. Gökbuget, A.K. Fielding, A.C. Schuh, J.M. Ribera, tein aggregates: just the tip of the iceberg, Nanomedicine 10 (2015) 2881–2891 A. Wei, H. Dombret, R. Foà, R. Bassan, Ö. Arslan, M.A. Sanz, J. Bergeron, (London, England), doi:10.2217/nnm.15.125. F. Demirkan, E. Lech-Maranda, A. Rambaldi, X. Thomas, H.A. Horst, M. Brügge- [10] P. Singhvi, A. Saneja, S. Srichandan, A.K. Panda, Bacterial inclusion bodies: a mann, W. Klapper, B.L. Wood, A. Fleishman, D. Nagorsen, C. Holland, Z. Zim- treasure trove of bioactive proteins, Trends Biotechnol. 38 (2020) 474–486, merman, M.S. Topp, Blinatumomab versus chemotherapy for advanced acute doi:10.1016/j.tibtech.2019.12.011. Lymphoblastic Leukemia, N. Engl. J. Med. 376 (2017) 836–847, doi:10.1056/ [11] R. Kloss, M.H. Limberg, U. Mackfeld, D. Hahn, A. Grünberger, V.D. Jäger, NEJMoa1609783. U. Krauss, M. Oldiges, M. Pohl, Catalytically active inclusion bodies of L- [33] C. Rader, Bispecific antibodies in cancer immunotherapy, Curr. Opin. Biotech- lysine decarboxylase from E. coli for 1,5-diaminopentane production, Sci. Rep. nol. 65 (2020) 9–16, doi:10.1016/j.copbio.2019.11.020. 8 (2018) 1–11, doi:10.1038/s41598- 018- 24070- 2. [34] H. Liu, A. Saxena, S.S. Sidhu, D. Wu, Fc engineering for developing therapeu- [12] U. Krauss, V.D. Jäger, M. Diener, M. Pohl, K.E. Jaeger, Catalytically-active inclu- tic bispecific antibodies and novel scaffolds, Front. Immunol. 8 (2017) 1–15, sion bodies-carrier-free protein immobilizates for application in biotechnology doi:10.3389/fimmu.2017.0 0 038. and biomedicine, J. Biotechnol. 258 (2017) 136–147, doi:10.1016/j.jbiotec.2017. [35] A.K. Kosmides, J.W. Sidhom, A. Fraser, C.A. Bessell, J.P. Schneck, Dual target- 04.033. ing nanoparticle stimulates the immune system to inhibit tumor growth, ACS [13] V.D. Jäger, R. Lamm, R. Kloß, E. Kaganovitch, A. Grünberger, M. Pohl, J. Büchs, Nano 11 (2017) 5417–5429, doi:10.1021/acsnano.6b08152. K.E. Jaeger, U. Krauss, A synthetic reaction cascade implemented by colocal- [36] M.C. Lo Giudice, F. Meder, E. Polo, S.S. Thomas, K. Alnahdi, S. Lara, K.A. Dawson, ization of two proteins within catalytically active inclusion bodies, ACS Synth. Constructing bifunctional nanoparticles for dual targeting: improved grafting Biol. 7 (2018) 2282–2295, doi:10.1021/acssynbio.8b00274. and surface recognition assessment of multiple ligand nanoparticles, Nanoscale [14] M. Pesarrodona, T. Jauset, Z.V. Díaz-Riascos, A. Sánchez-Chardi, M.E. Beaulieu, 8 (2016) 16969–16975, doi:10.1039/c6nr05478a. J. Seras-Franzoso, L. Sánchez-García, R. Baltà-Foix, S. Mancilla, Y. Fernández, [37] J. Pujols, S. Peña-Díaz, M. Conde-Giménez, F. Pinheiro, S. Navarro, J. San- U. Rinas, S. Schwartz, L. Soucek, A. Villaverde, I. Abasolo, E. Vázquez, Target- cho, S. Ventura, High-throughput screening methodology to identify alpha- ing antitumoral proteins to breast cancer by local administration of functional synuclein aggregation inhibitors, Int. J. Mol. Sci. 18 (2017), doi:10.3390/ inclusion bodies, Adv. Sci. 6 (2019), doi:10.10 02/advs.20190 0849. ijms18030478. [15] M.V. Céspedes, O. Cano-Garrido, P. Álamo, R. Sala, A. Gallardo, N. Serna, [38] T. Heimburg, J. Schünemann, K. Weber, N. Geisler, FTIR-spectroscopy of mul- A. Falgàs, E. Voltà-Durán, I. Casanova, A. Sánchez-Chardi, H. López-Laguna, tistranded coiled coil proteins, Biochemistry 38 (1999) 12727–12734, doi:10. L. Sánchez-García, J.M. Sánchez, U. Unzueta, E. Vázquez, R. Mangues, 1021/bi983079h. A. Villaverde, Engineering secretory amyloids for remote and highly selective [39] R. Sabaté, S. Ventura, Cross-β-sheet supersecondary structure in amyloid folds: destruction of metastatic foci, Adv. Mater. 1907348 (2019) 1–9, doi:10.1002/ techniques for detection and characterization, Methods Mol. Biol. 932 (2013) adma.201907348. 237–257, doi:10.1007/978- 1- 62703- 065- 6_15. [16] S. Ventura, A. Villaverde, Protein quality in bacterial inclusion bodies, Trends [40] A. Espargaró, S. Llabrés, S.J. Saupe, C. Curutchet, F.J. Luque, R. Sabaté, On the Biotechnol. 24 (2006) 179–185, doi:10.1016/j.tibtech.20 06.02.0 07. binding of congo red to amyloid fibrils, Angew. Chem. Int. Ed. Engl. 59 (2020) [17] M. Carrió, N. González-Montalbán, A. Vera, A. Villaverde, S. Ventura, Amyloid- 8104–8107, doi:10.1002/anie.201916630. like properties of bacterial inclusion bodies, J. Mol. Biol. 347 (2005) 1025–1037, [41] D. Li, F. Chen, J. Ding, N. Lin, Z. Li, X. Wang, Knockdown of HIP1 expression doi:10.1016/j.jmb.2005.02.030. promotes ligand-induced endocytosis of EGFR in HeLa cells, Oncol. Rep. 38 [18] M. Morell, R. Bravo, A. Espargaró, X. Sisquella, F.X. Avilés, X. Fernàndez- (2017) 3387–3391, doi:10.3892/or.2017.6025. Busquets, S. Ventura, Inclusion bodies: specificity in their aggregation process [42] N. Normanno, A. De Luca, C. Bianco, L. Strizzi, M. Mancino, M.R. Maiello, and amyloid-like structure, Biochim. Biophys. Acta-Mol. Cell Res. 1783 (2008) A. Carotenuto, G. De Feo, F. Caponigro, D.S. Salomon, Epidermal growth fac- 1815–1825, doi:10.1016/j.bbamcr.20 08.06.0 07. tor receptor (EGFR) signaling in cancer, Gene 366 (2006) 2–16, doi:10.1016/j. [19] M. Gil-Garcia, S. Navarro, S. Ventura, Coiled-coil inspired functional inclusion gene.2005.10.018. bodies, Microb. Cell Fact. 19 (2020) 1–16, doi:10.1186/s12934- 020- 01375- 4. [43] M.S. Kuhns, H.B. Badgandi, Piecing together the family portrait of TCR-CD3 [20] V.D. Jäger, R. Kloss, A. Grünberger, S. Seide, D. Hahn, T. Karmainski, M. Pi- complexes, Immunol. Rev. 250 (2012) 120–143, doi:10.1111/imr.120 0 0. queray, J. Embruch, S. Longerich, U. Mackfeld, K.E. Jaeger, W. Wiechert, M. Pohl, [44] E. García-Fruitós, E. Vázquez, C. Díez-Gil, J.L. Corchero, J. Seras-Franzoso, I. Rat- U. Krauss, Tailoring the properties of (catalytically)-active inclusion bodies, Mi- era, J. Veciana, A. Villaverde, Bacterial inclusion bodies: making gold from crob. Cell Fact. 18 (2019) 1–20, doi:10.1186/s12934- 019- 1081- 5. waste, Trends Biotechnol. 30 (2012) 65–70, doi:10.1016/j.tibtech.2011.09.003. [21] M. Diener, B. Kopka, M. Pohl, K.E. Jaeger, U. Krauss, Fusion of a coiled-coil do- [45] S.T.T. Schetters, W.S.P. Jong, L.J.W. Kruijssen, H.B. van den, B. van Saparoea, main facilitates the high-level production of catalytically active enzyme inclu- S. Engels, W.W.J. Unger, D. Houben, J.M.M. den Haan, J. Luirink, Y. van sion bodies, ChemCatChem 8 (2016) 142–152, doi:10.1002/cctc.201501001. Kooyk, Bacterial inclusion bodies function as vehicles for dendritic cell- [22] R. Lamm, V.D. Jäger, B. Heyman, C. Berg, C. Cürten, U. Krauss, K.E. Jaeger, mediated T cell responses, Cell. Mol. Immunol. 17 (2020) 415–417, doi:10.1038/ J. Büchs, Detailed small-scale characterization and scale-up of active YFP in- s41423- 019- 0298- x. clusion body production with Escherichia coli induced by a tetrameric coiled [46] M.V. Céspedes, Y. Fernández, U. Unzueta, R. Mendoza, J. Seras-Franzoso, coil domain, J. Biosci. Bioeng. xxx (2020), doi:10.1016/j.jbiosc.2020.02.003. A. Sánchez-Chardi, P. Álamo, V. Toledo-Rubio, N. Ferrer-Miralles, E. Vázquez, [23] W. Wang, S. Ventura, Prion domains as a driving force for the assembly S. Schwartz, I. Abasolo, J.L. Corchero, R. Mangues, A. Villaverde, Bacterial of functional nanomaterials, Prion 14 (2020) 170–179, doi:10.1080/19336896. mimetics of endocrine secretory granules as immobilized in vivo depots for 2020.1785659. functional protein drugs, Sci. Rep. 6 (2016) 1–10, doi:10.1038/srep35765. [24] R.M. Hoffman, Application of GFP imaging in cancer, Lab. Invest. 95 (2015) [47] R. Kloss, T. Karmainski, V.D. Jäger, D. Hahn, A. Grünberger, M. Baumgart, 432–452, doi:10.1038/labinvest.2014.154. U. Krauss, K.E. Jaeger, W. Wiechert, M. Pohl, Tailor-made catalytically active [25] M. Tashiro, R. Tejero, D.E. Zimmerman, B. Celda, B. Nilsson, G.T. Montelione, inclusion bodies for different applications in biocatalysis, Catal. Sci. Technol. 8 High-resolution solution NMR structure of the Z domain of staphylococcal pro- (2018) 5816–5826, doi:10.1039/c8cy01891j. tein A, J. Mol. Biol. 272 (1997) 573–590, doi:10.1006/jmbi.1997.1265. [48] J. Seras-Franzoso, C. Dez-Gil, E. Vazquez, E. Garca-Fruits, R. Cubarsi, I. Ratera, [26] C. Chen, Q.L. Huang, S.H. Jiang, X. Pan, Z.C. Hua, Immobilized protein ZZ, an J. Veciana, A. Villaverde, Bioadhesiveness and efficient mechanotransduction affinity tool for immunoglobulin isolation and immunological experimentation, stimuli synergistically provided by bacterial inclusion bodies as scaffolds for Biotechnol. Appl. Biochem. 45 (2006) 87, doi:10.1042/ba20060055. tissue engineering, Nanomedicine 7 (2012) 79–93, doi:10.2217/nnm.11.83. [27] W. Choe, T.A. Durgannavar, S.J. Chung, Fc-Binding Ligands of immunoglobu- [49] Š. Peternel, R. Komel, Isolation of biologically active nanomaterial (inclusion lin G: an overview of high affinity proteins and peptides, Materials 9 (2016) bodies) from bacterial cells, Microb. Cell Fact. 9 (2010) 1–16, doi:10.1186/ (Basel, Switzerland), doi:10.3390/ma9120994. 1475- 2859- 9- 66. [28] A. Konrad, A.E. Karlström, S. Hober, Covalent immunoglobulin labeling through [50] L. Gifre-Renom, J. Seras-Franzoso, D. Rafael, F. Andrade, O. Cano-Garrido, a photoactivable synthetic Z domain, Bioconjug. Chem. 22 (2011) 2395–2403, F. Martinez-Trucharte, E. Ugarte-Berzal, E. Martens, L. Boon, A. Villaverde, doi:10.1021/bc20 0 052h. G. Opdenakker, S. Schwartz, A. Arís, E. Garcia-Fruitós, The biological po- [29] M. Ultsch, A. Braisted, H.R. Maun, C. Eigenbrot, 3-2-1: Structural insights from tential hidden in inclusion bodies, Pharmaceutics 12 (2020), doi:10.3390/ stepwise shrinkage of a three-helix Fc-binding domain to a single helix, Pro- pharmaceutics12020157. tein Eng. Des.Sel. PEDS 30 (2017) 619–625, doi:10.1093/protein/gzx029. [51] A.F. Labrijn, M.L. Janmaat, J.M. Reichert, P.W.H.I. Parren, Bispecific antibodies: a [30] A.C. Braisted, J.A. Wells, Minimizing a binding domain from protein A, PNAS 93 mechanistic review of the pipeline, Nat. Rev. Drug Discov. 18 (2019) 585–608, (1996) 5688–5692, doi:10.1073/pnas.93.12.5688. doi:10.1038/s41573- 019- 0028- 1. [31] D. Skokos, J.C. Waite, L. Haber, A. Crawford, A. Hermann, E. Ullman, R. Slim, [52] M. Arruebo, M. Valladares, Á. González-Fernández, Antibody-conjugated S. Godin, D. Ajithdoss, X. Ye, B. Wang, Q. Wu, I. Ramos, A. Pawashe, L. Canova, nanoparticles for biomedical applications, J. Nanomater. 2009 (2009), doi:10. K. Vazzana, P. Ram, E. Herlihy, H. Ahmed, E. Oswald, J. Golubov, P. Poon, 1155/2009/439389. L. Havel, D. Chiu, M. Lazo, K. Provoncha, K. Yu, J. Kim, J.J. Warsaw, N.S. Oristian, C.J. Siao, D. Dudgeon, T. Huang, T. Potocky, J. Martin, D. MacDonald, A. Oyejide, A. Rafique, W. Poueymirou, J.R. Kirshner, E. Smith, W. Olson, J. Lin, G. Thurston, M.A. Sleeman, A.J. Murphy, G.D. Yancopoulos, A class of costimulatory CD28- bispecific antibodies that enhance the antitumor activity of CD3-bispecific an- tibodies, Sci. Trans. Med. 12 (2020), doi:10.1126/scitranslmed.aaw7888. 482