1. Introduction
Fibroblast growth factor receptors (FGFRs) constitute a family of four receptor tyrosine kinases (RTKs) that, together with their cognate ligands, fibroblast growth factors (FGFs), are involved in the transmission of signals from the extracellular environment to the cell interior. FGFRs are composed of a large, highly glycosylated extracellular region including acidic box (AB) and three immunoglobulin-like domains D1, D2 and D3, of which D2 and D3 are involved in FGF binding, whereas AB and D1 fulfill regulatory functions. FGFRs are anchored to the plasma membrane by a single transmembrane helix that participates in receptor dimerization. The intracellular part of FGFRs comprises a regulatory juxtamembrane region and a tyrosine kinase (TK) domain directly involved in the phosphorylation cascade [
1,
2]. In the prevailing model, inactive FGFRs are predominantly monomeric and dimerize upon FGF binding, resulting in large conformational changes of the receptors, leading to the transphosphorylation of key tyrosine residues within TK. These modifications provide binding sites for several proteins involved in cellular signaling, initiating the propagation of the signals through phospholipase C gamma (PLCγ), signal transducer and activator of transcription (STAT), phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR), and Ras–Raf–MEK–ERK. The FGFR-dependent signaling controls pivotal cellular processes, such as division, motility, metabolism and death [
3,
4,
5,
6,
7].
Unbalanced FGF/FGFR signaling is responsible for several developmental diseases and is a hallmark of numerous tumors [
4,
8,
9]. Therefore, cells tightly balance FGF/FGFR units through several regulatory mechanisms that fine-tune the amplitude and duration of signals. For example, the stability of FGF/FGFR complexes determines downstream signaling and the induced cellular responses [
10]. FGF/FGFR activity is modulated by a number of extracellular and plasma membrane-embedded proteins, some of which act as essential co-receptors [
11,
12,
13,
14]. Furthermore, several feedback loops participate in the downregulation of FGF/FGFR signaling [
15,
16]. FGFR endocytosis, which directs receptors to lysosomes for degradation, further shapes the kinetics and specificity of the signals transmitted [
6,
16,
17].
The modulation of the spatial distribution of FGFR in the plasma membrane has recently emerged as an additional mechanism for shaping FGFR activity and cell fate. Firstly, the specificity of FGFR signaling is adjusted by the oligomeric state of the receptor. Several reports indicated that FGFRs form oligomers in the absence of ligands and these complexes are partially signaling active [
18]. Secondly, the duration of FGFR-dependent signals depends on the residence time of the receptor on the cell surface. Shifting FGFR1 from monomer to dimer with a bivalent engineered antibody stimulates clathrin-mediated endocytosis and lysosomal degradation of FGFR1, shortening the receptor’s exposure at the surface [
19,
20]. High-order oligomerization of FGFR1 with a tetravalent engineered antibody or with FGF1 oligomers further enhances the efficiency of receptor endocytosis and accelerates its lysosomal targeting [
21]. Importantly, FGFR1 clustering alters the mechanisms of receptor internalization by activating clathrin-independent, dynamin-dependent endocytic routes [
21]. Thirdly, tetravalent and pentavalent FGF1 oligomers alter FGFR1 activity by promoting mitogenic signaling and inhibiting cell metabolic activity [
22].
All above-mentioned effects were largely dependent on the architecture of the oligomeric ligand applied to crosslink FGFR1. Thus, it is highly likely that yet unexplored FGFR1 activities and receptor-dependent cellular effects can be triggered by novel multivalent FGFR1 ligands. Here, we report a novel strategy to generate oligomeric FGFR1 ligands by employing bivalent fusions of FGF1 with the Fc fragment of human IgG1 and GFPpolygons-protein G as oligomerization scaffolds.
2. Materials and Methods
2.1. Antibodies and Reagents
The primary antibodies directed against ERK1/2 (#9102) and phospho-ERK1/2 (pERK1/2; #9101) were from Cell Signaling (Danvers, MA, USA). Anti-tubulin primary antibody (#T6557) was from Sigma-Aldrich (St. Louis, MO, USA), anti-FGF1 primary antibody (sc-55520) was from Santa Cruz Biotechnology (Dallas, TX, USA) and anti-human IgG (Fc) antibody coupled to HRP (#4-10-20) was from KPL (Gaithersburg, MA, USA). Secondary antibodies coupled to HRP were from Jackson Immuno-Research Laboratories (Cambridge, UK).
2.2. Cells
The human osteosarcoma cell line (U2OS) was obtained from American Type Culture Collection (ATCC), and U2OS, stably expressing FGFR1 (U2OS-R1), were obtained as described previously [
23]. Cells were cultured in 5% CO
2 atmosphere at 37 °C in Dulbecco’s Modified Eagle’s Medium (Biowest, Nuaille, France) supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA), antibiotics mix (100 U/mL penicillin and 100 μg/mL streptomycin) (Thermo Fisher Scientific), for U2OS-R1, additionally supplemented with 1 mg/mL geneticin (Thermo Fisher Scientific). Murine embryonic fibroblasts (NIH3T3) were from ATCC and were cultured in Dulbecco’s Modified Eagle’s Medium (Gibco) supplemented with 2% bovine serum (Thermo Fisher Scientific) and antibiotics mix (100 U/mL penicillin and 100 μg/mL streptomycin). Cells were grown in 5% CO
2 atmosphere at 37 °C.
2.3. Recombinant Proteins
Plasmid pET28a_HisTag-GFPpoly_protG was a kind gift from the Jung lab, Department of Chemistry, National University in Daejeon, South Korea. The GFPpoly-ProtG (GFPpG) oligomers were expressed in E. coli BL21(DE3)-RIL strain (Agilent Technologies, Santa Clara, CA, USA). Cells were grown at 37 °C until OD600 = 0.8, then protein expression was induced by addition of 1 mM IPTG, followed by incubation of cells at 16 °C for 16 h. GFPpG oligomers were purified using nickel-charged resin, eluted with 50 mM Tris, 150 mM NaCl, 250 mM imidazole, pH 8.0. Next, GFPpG variants were separated by Blue Native-PAGE (BN-PAGE). Mixture of proteins were applied in maximum volume into gradient gel (4–13%). After the electrophoretic separation the bands representing each of GFPpG oligomers were cut out under UV light and transferred to buffer containing 20 mM Tris, 150 mM glycine, 0.02% SDS, pH 8.3 with overnight shaking at 4 °C. Then, separated GFPpG variants were dialyzed into PBS buffer. The identity and the purity of the proteins were confirmed by SDS-PAGE and Western blotting. Pure GFPpG oligomers were stored with the addition of 5% glycerol at −20 °C.
Genetic constructs for expression of Fc-FGF1 and FGF1-Fc were prepared using restriction free cloning technique. In the first part of this method the FGF1 sequence was amplified in a regular polymerase chain reaction (PCR). Primers were obtained from Sigma Aldrich and their sequences were sequentially: 5′ GAGCCTCTCCCTGTCTCCG GGTAAAATGGCTAATTACAAGAAGCCCAAAC 3′ (forward) and 5′ GAACCACTGAA CAAATGGCACTAGGTTAATCAGAAGAGACTGGCAGGGGG 3′ (reverse) for the Fc-FGF1, and 5′ GTCAGTAACGACTGGTGTCCACTCCGCTAATTACAAGAAGCCCAAACTCC 3′ (forward) and 5′ ACGGTGGGCATGTGTGAGTTTTGTCATCAGAAGAGACTGGCAGGG GGAGA 3′ (reverse) for the FGF1-Fc. The pET3c_FGF1 construct was used as a template. Amplification reactions were performed in a total volume of 50 µL. Obtained PCR product was used as a megaprimer in a second amplification reaction, introducing FGF1 in front or behind the Fc fragment coding sequence in pLEV113 plasmid. Amplification reactions were also performed in a total volume of 50 µL. Both PCR reactions were performed with use of Phusion Flash High-Fidelity PCR Mater Mix (Thermo Fisher Scientific). Obtained products were subjected to DpnI digestion (Thermo Fisher Scientific), and the correctness of genetic constructs was confirmed by DNA sequencing. The FGF1-Fc-FGF1 genetic construct was prepared via gene synthesis combined with standard molecular biology methods. Recombinant proteins were expressed in CHO-S cells. Cells were grown in serum-free PowerCHO-2CD medium (Lonza, Basel, Switzerland) supplemented with 8 mM L-glutamine (Biowest, Riverside, MO, USA) and antibiotic mix (100 U/mL penicillin and 100 µg/mL streptomycin) (Thermo Fisher Scientific). One day prior to transfection, CHO-S cells were seeded at 1.8 × 106 cells/mL. On the transfection day, CHO-S cells were centrifuged and cell pellet was re-suspended in protein- and serum-free ProCHO4 medium. Plasmids encoding Fc-FGF1 or FGF1-Fc or FGF1-Fc-FGF1 (1.25 μg DNA per 1 × 106 cells) and PEI (5 μg per 1 × 106 cells) were diluted separately in 150 mM NaCl, mixed, and incubated at RT for 10 min prior addition to cells. Cells were incubated under standard conditions for 4 h (37 °C, 110 rpm, 8% CO2). After this time, the cell culture was diluted with an equal volume of PowerCHO-2CD supplemented with 4 mM L-glutamine and antibiotic mix (200 U/mL penicillin and 200 µg/mL streptomycin) to obtain the cell density of about 1 × 106 cells/mL, and incubated at 32 °C. The next day, the cell culture was supplemented with 8 mM L-glutamine and finally harvested at day 7. Fc-FGF1, FGF1-Fc and FGF1-Fc-FGF1 were purified from culture supernatants using the HiTrap MabSelect SuRe column (GE Healthcare, Chicago, IL, USA). Both proteins were eluted with 0.01 M sodium citrate, pH 3.5, and neutralized with 1 M Tris-HCl, pH 9.0. Then, proteins were dialyzed into PBS buffer using the Desalting HiTrap column (GE Healthcare). The identity and the purity of the proteins were confirmed by SDS-PAGE and Western blotting.
2.4. Activation of FGFR1 and Receptor-Downstream Signaling Cascades
To analyze the impact of the recombinant proteins on FGFR1 activation and initiation of receptor-downstream signaling cascades, serum starved NIH3T3 cells were incubated for 15 min at 37 °C with FGF1 (100 ng/mL), as control, Fc-FGF1 (50, 100 ng/mL) or FGF1-Fc (50, 100 ng/mL) in the presence of heparin (10 U/mL). Cells were lysed in Laemmli buffer and subjected to SDS-PAGE and Western blotting.
An analogous experiment was performed for the FGF1 and GFPpG multivalent complexes. Serum starved NIH3T3 cells were incubated for 15 min at 37 °C with Fc-FGF1 or FGF1-Fc (both at 100 ng/mL) and GFPpG variants (at equimolar concentrations to the Fc) in the presence of heparin (10 U/mL). Cells were lysed in Laemmli buffer and subjected to SDS-PAGE and Western blotting.
2.5. Fluorescence Microscopy
For the analysis of Fc-FGF1 and FGF1-Fc cellular uptake, early endosomes in U2OS (control) and U2OS-R1 cells were labeled with Rab5a-RFP (CellLight Early Endosomes-RFP, Thermo Fisher Scientific). Next, serum-starved cells were incubated with each recombinant protein (both at 500 ng/mL) in the presence of heparin (10 U/mL) at 37 °C, 5% CO2 for 30 min. After this time, the internalization was stopped by cooling down cells on ice. Next, cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton in PBS. Zenon AF-488 (Thermo Fisher Scientific) was used for labeling of Fc-FGF1 and FGF1-Fc and NucBlue Live (Thermo Fisher Scientific) was used to fluorescently mark nuclei. The Zenon AF-488 signal was visualized with a 450/490 nm bandpass excitation filter and a 500/550 nm bandpass emission filter. The cell light early endosomes-RFP signal was visualized with a 540/552 nm bandpass excitation filter and a 575/640 nm bandpass emission filter. NucBlue Live signal was visualized with 335/383 nm bandpass excitation filter and 420/470 nm emission filter.
For the analysis of FGF1 and GFPpG multivalent complexes cellular uptake serum starved U2OS-R1 cells were incubated with each of GFPpG variant (5 µg/mL) and Fc-FGF1 or FGF1-Fc (at equimolar concentrations to the protein G in GFPpG) in the presence of heparin (10 U/mL). After 20 min of incubation on ice, the cells were moved to 37 °C, 5% CO2 for 45 min to allow for internalization. After this time, the internalization was stopped by cooling down cells on ice. Next, cells were fixed with 4% paraformaldehyde. DyLight 550 NHS Ester (Thermo Fisher Scientific) was used for the labeling of Fc-FGF1 and FGF1-Fc and NucBlue Live (Thermo Fisher Scientific) was used to fluorescently mark nuclei. GFP signal was visualized with a 450/490 nm bandpass excitation filter and a 500/550 nm bandpass emission filter. DyLight 550 signal was visualized with a 540/552 nm bandpass excitation filter and a 575/640 nm bandpass emission filter. NucBlue Live signal was visualized with 335/383 nm bandpass excitation filter and 420/470 nm emission filter. Wide-field fluorescence microscopy was carried out using Zeiss Axio Observer Z1 fluorescence microscope (Zeiss, Oberkochen, Germany). Images were taken using LD-Plan-Neofluar 40 × /0.6 Korr M27 objective and Axiocam 503 camera. Images were processed with Zeiss ZEN 2.3 software (Zeiss, Oberkochen, Germany) and Adobe Photoshop (Adobe, San Jose, CA, USA).
2.6. Flow Cytometry
U2OS-R1 cells were seeded onto 12-well plates (1 × 105 cells per well) in full medium and left to attach overnight. Then, medium was removed, cells were washed with PBS buffer and starved with serum-free medium for 4 h. Next, plates were cooled on ice. Fluorescent GFPpG variants (500 ng/mL) alone or combined with Fc-FGF1 or FGF1-Fc (at equimolar concentrations to the protein G in GFPpG) were added to the cells in the presence of heparin (10 U/mL), in a serum-free medium supplemented with 1% BSA. After 20 min of incubation on ice, the cells were moved to 37 °C for 30 min to allow for internalization. Then, the medium was removed and the cells were washed with serum-free medium supplemented with 0.2% BSA pH 3.5 (three times, 5 min) and then with PBS buffer (three times, 1 min). Cells were subsequently detached with 10 mM EDTA in PBS buffer, pH 8.0, harvested by centrifugation and resuspended in PBS supplemented with 1% BSA. Cells were analyzed using a Agilent Technologies, Santa Clara, CA, USA) and NovoExpress software (ACEA Biosciences, San Diego, CA, USA).
2.7. Blue Native-PAGE
The formation of Fc-FGF1 or FGF1-Fc complexes with GFPpG was analyzed with Blue Native PAGE (BN-PAGE). The experiments were performed with Fc-FGF1 (5 µg), FGF1-Fc (5 µg) and their mixtures with various oligomeric forms of GFPpG (from dimer to pentamer) (5 µg) incubated for 10 min at RT. Proteins were separated using 4–13% BN-PAGE gradient gels and imaged under UV light.
2.8. BLI Measurements
Binding analyses of FGF1 variants (Fc-FGF1 and FGF1-Fc) to GFPpG oligomers (from dimer to pentamer) was performed using bio-layer interferometry (BLI) with ForteBio Octet K2 (Pall ForteBio, San Jose, CA, USA). The Fc-FGF1 (10 µg/mL) or FGF1-Fc (10 μg/mL) was chemically immobilized on Amine Reactive Second-Generation (AR2G) biosensors (Pall ForteBio, San Jose, CA, USA). The measurements were conducted at 25 °C for each GFPpG variant (20 μg/mL) in PBS buffer supplemented with 0.2% BSA and 0.05% Triton X-100. Association and dissociation phases were monitored by 300 s.
4. Discussion
Although most of RTKs activate similar intracellular pathways, the elicited cellular responses largely differ between individual RTK members. This phenomenon can be attributed to discrete differences in the amplitude and duration of signals transmitted via distinct intracellular pathways [
28,
29]. A number of regulatory mechanisms shape RTK-dependent signals, determining fate of cells. Regulation of RTK distribution in the plasma membrane is one of strategies for fine-tuning RTK-dependent signaling [
10]. We have previously reported that by using engineered oligomeric FGFR1 ligands it is possible to control activity, endocytosis and cellular level of the receptor [
21,
22]. Importantly, the architecture of the oligomeric ligands (binding site on the receptor, affinity, oligomeric state, oligomerization scaffold used) played an important role in determining the outcome of FGFR1 modulation [
21,
22]. Therefore, we have sought to develop novel strategies to generate multivalent FGFR1 ligands.
So far, we have been able to obtain FGFR1 ligands having up to five receptor binding sites [
22]. In this report, we present an approach that allows the efficient assembly of FGFR1-targeting oligomers with custom-selected, higher number of receptor binding sites. In principle, using this strategy it is possible to generate ligands with very high number of receptor binding sites, as it can be modulated both by the oligomeric state of the scaffold (GFPpG) and the valency the of the ligand (in our case, bivalent FGF1 fusions with the Fc, tetravalent FGF1-Fc-FGF1, etc.). Kim et al. described the efficient isolation of GFPpG composed of over ten units. Although we also observed the assembly of higher oligomeric states of GFPpG, we were able to efficiently separate high quantities of oligomers up to the pentamer. A slight modification of the GFPpG expression and isolation protocol would likely allow to isolate these higher state oligomers, but it was beyond the scope of this work.
Dikov et al. produced functional Fc-FGF1 in E. coli cells, implicating the feasibility of the strategy undertaken to fuse FGF1 with Fc [
30]. In order to enable glycosylation of the Fc fragment we decided to produce FGF1 fusions with the Fc in CHO cells. We prepared N- and C-terminal fusions of FGF1 with the Fc to allow the assembly of FGFR1-targeting oligomers of distinct architecture. We were able to produce in mammalian cells and purify biologically active Fc-FGF1 and FGF1-Fc. Interestingly, we observed that large fractions of Fc-FGF1 and FGF1-Fc were trapped inside CHO cells, limiting overall yield. Although the engineered expression constructs are equipped with a functional signal peptide directing Fc-FGF1 and FGF1-Fc to the secretory route, it is likely that the intrinsic ability of FGF1 to employ unconventional secretion mechanisms partially outcompetes the ER targeting [
19]. Consequently, the incorporation of a large Fc tag into FGF1 may limit the efficiency of its unconventional secretion, resulting in cytosolic localization of part of recombinant proteins. Clearly, further studies are required to understand in detail the limited efficiency of Fc-FGF1 and FGF1-Fc secretion.
The combination of Fc-tagged FGF1 and GFPpG resulted in the self-assembly of FGF1 oligomers formed through the Fc-protG interaction. Depending on the applied GFPpG oligomeric form, FGF1 complexes consisting of up to ten growth factor molecules were obtained, demonstrating the feasibility of the selected strategy. BLI experiments revealed that FGF1-Fc, Fc-FGF1 and GFPpG interactions are characterized by low koff, implicating high stability of the formed FGF1 oligomers. Notably, all developed FGF1 oligomers retained their biological activity, as they were able to stimulate FGFR1-dependent signaling pathways. Importantly, multivalent FGFR-targeting complexes caused clustering of the receptor on the cell surface, affecting cellular trafficking of the receptor. Similar effect we have recently observed for multivalent galectin-3, an extracellular lectin that binds sugar chains of FGFR1 [
31,
32].
FGF/FGFR signaling system is kept under tight control to ensure balanced stream of signals allowing for proper execution of metabolic and developmental processes. Developed multivalent complexes presented in this study can be used for fine-tuning FGFR signaling to evoke desired cellular activity, which may be helpful in understanding fundamentals of FGF/FGFR biology. Furthermore, some of multivalent FGFR-targeting complexes may turn out to be useful in clinics, e.g., in wound healing. Certainly, more in-depth studies are required to test the applicability of multivalent FGFR-targeting complexes.