1. Introduction
Skeletal diseases are a major global health problem, affecting individuals across all demographic groups and posing significant challenges to healthcare systems in terms of both direct medical costs and indirect costs associated with disability and lost productivity [
1]. These diseases, which can be attributed to various factors such as genetic mutations, metabolic disorders, and tumor processes, underscore the complexity of maintaining skeletal health. Among these pathologies, X-linked hypophosphatemia (XLH) stands out as a paradigmatic example, characterized by aberrant phosphate metabolism and skeletal deformities [
2].
Dysregulation of fibroblast growth factor (FGF) 23 signaling is central to the pathogenesis of XLH. FGF23, a member of the fibroblast growth factor (FGF) superfamily, consists of a class of 22 polypeptides with critical roles in embryonic and postnatal development, as well as cell cycle regulation [
3]. It exerts pivotal control over phosphate and vitamin D metabolism primarily within bone tissue [
4]. Dysfunction in FGF23 signaling pathways, whether due to genetic abnormalities or other factors, can lead to systemic disorders such as XLH, the most common form of hereditary rickets, characterized by skeletal abnormalities and metabolic perturbations [
5,
6,
7], thereby emphasizing the critical role of FGF23 in skeletal homeostasis [
5].
Emerging research has highlighted the significance of endoplasmic reticulum (ER) stress in the pathogenesis of various diseases, including atherosclerosis, diabetes, and obesity [
8]. The ER, responsible for protein synthesis and folding, plays a central role in maintaining cellular homeostasis. Under conditions of metabolic dysregulation or pathological stress, the ER can undergo stress-induced dysfunction, triggering the unfolded protein response (UPR) and potentially leading to cellular dysfunction or apoptosis, as summarized in
Figure 1.
The UPR augments ER size to enhance protein folding capacity, initiated by IRE1α, PERK, and ATF6 [
9]. Unresolved ER stress can trigger IRE1α-induced apoptosis via hyperactivation and homo-oligomerization, involving mitochondrial protective proteins and proapoptotic factors like Bcl-2-like protein 4 (BAX) and Bcl-2-like protein 7 (BAK), leading to controlled cell death [
10]. ER stress-mediated apoptosis is implicated in numerous chronic diseases [
11]. If cells survive ER stress, antioxidant responses and ER chaperones are upregulated [
10]. Upregulated calcium-binding proteins, including calreticulin (CALR), Grp94, 78 kDa glucose-regulated protein (Grp78), and ER protein 72 (Erp72), are part of the ATF6 signaling cascade [
12,
13,
14].
Mitochondrial dysfunction due to metabolic changes induces mitochondrial stress as part of the stress response [
15]. In addition, mitofusin 2 (MFN2) regulates ER-mitochondria contact sites, affecting ER shape and inter-organelle interactions [
16,
17].
Understanding the intricate interplay between FGF23 dysregulation, ER stress, and cellular responses is crucial for elucidating the molecular mechanisms underlying skeletal pathologies such as XLH. Despite its rarity, XLH serves as a model system for exploring fundamental biological processes underlying bone diseases, offering insights that may have broader implications for skeletal health.
The present study investigates the impact of FGF23 overexpression on cellular structures and stress responses using SaOS-2 cells as a model system, revealing that FGF23 alters the structure of mitochondria and the rough ER (rER) and their interactions. SaOS-2 cells, a well-established human osteosarcoma cell line, are used in scientific research for the investigation of bone biology and associated processes. These cells closely mimic the functional properties of osteoblasts, the primary cells responsible for bone formation [
18]. Additionally, SaOS-2 cells have been used to assess the cytotoxicity of various compounds and their effects on cell proliferation, apoptosis, and the unfold protein response (UPR) [
19,
20]. By elucidating these mechanisms, this work contributes to the broader effort to improve diagnostic strategies and develop targeted interventions aimed at preserving bone health and enhancing patient outcomes.
2. Materials and Methods
2.1. Cell Culture
2.1.1. SaOS-2 Cells
Human SaOS-2 cells (#89050205, Sigma-Aldrich, Taufkirchen, Germany) were cultivated in T75 cell culture flasks in growth medium containing Minimal Essential Medium Eagle (M7278, Sigma-Aldrich, Taufkirchen, Germany) supplemented with 10% FBS (F2442, Sigma-Aldrich, Taufkirchen, Germany), 1% penicillin/streptomycin (PS-B, Capricorn-Scientific, Ebsdorfergrund, Germany), 1% Glutamax (#35050-061; Thermo Fisher Scientific, Darmstadt, Germany) and 0.55 ug/mL puromycin (ant-pr-1, Thermo Fisher Scientific, Darmstadt, Germany) in a humidified incubator at 37 °C and 5% CO
2. The growth medium was refreshed every 3 days. When approx. 80–90% confluence was reached, the cells were either taken into the experiment (see
Figure 2) or split to be cultured as required. For splitting, the cells were first digested with 2 mL trypsin with EDTA (#T4049, Sigma-Aldrich, Taufkirchen, Germany) for about 5 min; the successful digestion was verified under the microscope. Subsequently, 4 mL of medium (see above) were used, and after a 5 min centrifugation at 1000 rpm (Beckmann Coulter, Krefeld, Germany), 1:4 splitting was performed to obtain 4 new culture flasks from one T 75 culture flask (#833911002, Sarstedt, Nümbrecht, Germany).
2.1.2. Generation of FGF23 Overexpressing and Control Empty Vector Cells
The cells were generated by the Medical Proteome Center (MPC) in cooperation with the Department of Molecular GI-Oncology (MGO, both Ruhr-University Bochum, D-44801 Bochum, Germany). In brief, stable overexpression of FGF23 in SaOS-2 cells was induced by lentiviral transduction. A FGF23-6 His coding gene (derived from #RC210127, Origene, Herford, Germany) was induced into the lentiviral plasmid pLJM (#19319, Addgene, Teddington, UK), while non-coding pLJM was used as an empty vector control (EV). The pLJM vectors were mixed with the lentiviral envelope plasmid pCMV-VSV-G (#8454, Addgene, Teddington, UK), the packaging plasmid gag/pol (Addgene #14887, Teddington, UK) (ratio 2:1:2), HBS buffer, and 125 mM CaCl2 solution and added to the medium of lentivirally infected packaging cells (Human Embryonic Kidney cells, HEK). After an incubation of 24 h, the medium was replaced, followed by a second incubation of 24 h. Finally, the medium was removed from the packaging cells, sterile filtered (0.45 µm), and added to SaOS-2 cells at 50% confluence for another overnight incubation. To select transduced cells, 0.5 µg/µL pyromycin was added to the culture medium. Initial experiments showed no significant difference between wild-type and empty vector cells.
2.1.3. Human-Embryonic-Kidney (HEK) Cells
Three days after the last medium refreshment, the hole medium, from now on referred to as “supernatant” from SaOS-2 cells, was transferred to HEK cells. After two hours of incubation, RNA from the HEK cells was isolated and subsequently measured by real-time PCR (see below, “Quantitative reverse transcription polymerase chain reaction (qPCR)”). On the one hand, the supernatants of SaOS-2 cells overexpressing FGF23 were used for the experiment, while the supernatants of empty vector SaOS-2 cells were used in the control groups. HEK cells were cultured in T75 cell culture flasks (#833911002, Sarstedt, Nümbrecht, Germany) in growth medium consisting of Dulbecco’s Minimal Essential Medium Eagle (#M4655, Sigma-Aldrich, Taufkirchen, Germany) supplemented with 10% FCS (#FBS-11A, Capricorn-Scientific, Ebsdorfergrund, Germany), 1% penicillin/streptomycin (PS-B, Capricorn-Scientific, Ebsdorfergrund, Germany), and 1% sodium pyruvate (#NPY-B, Capricorn-Scientific, Ebsdorfergrund, Germany) in a humidified incubator at 37 °C and 5% CO2. Growth medium was refreshed every 3 days.
2.2. Transmission Electron Microscopy (TEM)
Confluent SaOS-2 cells were washed with PBS and fixed for one hour with 2% glutaraldehyde (#G5882, Sigma-Aldrich, Taufkirchen, Germany) in 0.2 mol/L cacodylate buffer (#AGR1503, agarscientific, Essex, UK) (pH 7.2) at room temperature. This was followed by incubation with 2% osmium tetroxide (#0972B-6, Polyscience Inc., Hirschberg, Germany) for 90 min. The cells were then collected using a cell scraper and fixed in agar blocks. Samples were then dehydrated in an ascending ethanol series, starting with 50% ethanol, followed by incubation in 70% ethanol, 1% uranyl acetate (#21447, Polyscience Inc., Hirschberg, Germanyand 1% phosphotungstic acid (#455970, Sigma-Aldrich, Taufkirchen, Germany) overnight at 4 °C. Dehydration continued the next day with an ascending ethanol series (80–100%). The samples were carefully transferred into epoxy resin (consisting of glycidether (#21045.02, Serva, Heidelberg Germany), methylnadic anhydride (#29452.02, Serva, Heidelberg, Germany), 2-dodecenyl succinic anhydride (#20755.01, Serva, Heidelberg Germany), and 2,4,6-tris (dimethylaminomethyl)phenol (#36975.01, Serva Heidelberg, Germany) and incubated in propylene oxide (#807027, Sigma-Aldrich, Taufkirchen, Germany). After polymerization, ultrathin slices (50–70 nm) were cut using an Ultracut E Reichert-Jung (Leica Microsystems GmbH, Wetzlar, Germany) with a DiATOME histo diamond knife (45°, 6 mm, MX559; Diatome AG, Switzerland). Images were captured with a Zeiss Leo 910 (Zeiss, Oberkochen, Germany) equipped with a digital CCD camera and analyzed using ImageJ 1.51 s (National Institutes of Health, USA).
2.3. Evaluation and Measurement of the TEM Images
To get a good overall impression, images of SaOS-2 cells were taken from different sections, and various parameters were considered for morphological characterization. In particular, the expansion of the cell organelles was quantified. The (1) area in µm
2 with a spatial calibration of the corresponding scale bar in each picture. (2) FeretMin as the smallest possible distance between two parallel tangents on the sides of the rER at any angle. (3) aspect ratio defined as:
(4) The length of the outside boundary of the selection as perimeter. (5) circularity, calculated by:
In addition, (6) roundness as:
were considered. A total of 698 ER structures and 787 mitochondria were selected from 188 images using ImageJ 1.51 s software (National Institutes of Health, USA) and measured in a blinded manner. An unpaired
t-test was performed for significance testing between the two cell lines, and values with
p < 0.05 were considered significant.
2.4. Measurement of ER-Mitochondria Contacts
To assess ER-mitochondria contacts, the distance from the ER to the mitochondria was measured with the “length tool” in ImageJ (
Figure 3, yellow lines, numbered). First, the outline of the cell organelle was marked and circled along the contact zone. Measurements were taken at intervals of 25 nm where the length (distance) between the two structures was determined. All measurements were carried out blinded. Although the contact zones have been measured over a large part of their extent (see
Figure 3), only contact zones up to 25 nm were used for the evaluation, as these are relevant as described by Csordás et al. 2006 [
21]. In total, 44 contact zones with 1025 measurements were performed in FGF23 overexpressing cells (FGF23) and 28 contact zones with 442 measurements in the control (EV) cells. An illustrative example of these measurements is provided in
Figure 3 for better visualization.
2.5. Western Blots
For Western blot analysis, confluent SaOS-2 cells were detached from the culture flasks using EDTA-Trypsin (#T4049, Sigma-Aldrich, Taufkirchen, Germany), and subsequently proteins were extracted using cell lysis (CL) (#FNN0011, ThermoFisher Scientific, Darmstadt, Germany) buffer containing additionally protease inhibitors (#11873580001, Roche, Mannheim, Germany). All procedures were conducted on ice to preserve protein integrity. The solution was gently shaken for 30 min at 4 °C, followed by centrifugation at 12,000 rpm for 10 min, and the resulting supernatant was collected for further western blot analysis. Protein concentrations were determined using the Pierce™ BCA Protein Assay Kit (# 23225, Thermo Fisher Scientific, Darmstadt, Germany). Prior to protein separation, samples were heated at 90 °C for five minutes. A 4× Laemmli sample buffer (#161-0747, BioRad, Feldkirchen, Germany) was added to the samples. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis at 100 V, followed by transfer onto PVDF membranes using the Trans-Blot
® Turbo™ Transfer System (#1704150, BioRad. Feldkirchen, Germany) according to the manufacturer’s protocol. Membranes were then blocked for at least one hour using either 1× phosphate-buffered saline (PBS) with 1% RotiBlock (#A151 Roth, Karlsruhe, Germany) or 5% BSA (#fraction V; Sigma-Aldrich, Taufkirchen, Germany) or 5% skim milk, depending on the primary antibody used. Primary antibodies (
Table 1) were incubated overnight at 4 °C, followed by incubation with secondary antibodies (
Table 1) for 2 h at room temperature. Protein bands were visualized using Western Blotting Luminol Reagent (#sc-2048, Santa Cruz Biotechnology, Heidelberg, Germany). For semiquantitative analysis, the band intensities of the proteins of interest were normalized to the loading control. All values were then further normalized to the mean of the normalized empty vector SaOS-2 cell signals. Western blot analysis was performed on tissue samples from at least three FGF23 overexpressing and three empty vector samples.
2.6. Immunofluorescence
For staining, coverslips were placed in a Petri dish containing Minimal Essential Medium (see “Cell culture–
Section 2.1”) and incubated in a humidified incubator at 37 °C to equilibrate. Subsequently, previously trypsinized SaOS-2 cells from culture were added to the coverslips and cultivated under the same conditions for 1 to 3 days to measure apoptosis. Following incubation, the medium was discarded, and coverslips were washed with PBS before fixation with 4% PFA. Staining was carried out according to a standard protocol, with 0.1% PBS-Triton for permeabilization and triple washes in PBS between all steps. Primary antibodies (phalloidin 1:50, #ab176757, abcam; cleaved caspase 3 1:100, #9664, Cell signalling, Leiden, Netherlands) were incubated overnight at 4 °C, followed by incubation with secondary antibodies (anti-rabbit IgG (H+L) highly cross-adsorbed secondary antibody, 1:200, #A-21206, Thermo Fisher Scientific, Darmstadt, Germany) at room temperature for 2 h. Nuclear staining was performed using 1 mg/mL DAPI for 30 min at room temperature. After mounting, samples were allowed to dry, and images were captured using a Keyence microscope (BZ-X800, Keyence Corporation, Osaka, Japan).
Apoptotic cells could be identified by cleaved caspase 3. To determine the apoptosis rate, the percentage of apoptotic cells (caspase 3 positive) in the total number of cells (DAPI positive) was calculated and analyzed in a blinded manner.
2.7. Quantitative Reverse Transcription Polymerase Chain Reaction (qPCR)
For qRT-PCR analysis, confluent SaOS-2 cells were detached from the culture flasks using EDTA-Trypsin, followed by centrifugation at 1200 rpm for 5 min to collect the cell pellets on ice. RNA isolation from the cell pellet was carried out using the NucleoSpin RNA kit, followed by reverse transcription into complementary deoxyribonucleic acid (cDNA) using the GoScript™ Reverse Transcription Mix, Oligo (dT) (#A2791, Promega, Madison, WI, USA), as per the manufacturer’s instructions.
The qRT-PCR was performed with a CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA), employing GoTag qPCR Master Mix (Cat# A6001, Promega, Walldorf, Germany), 100 ng cDNA per reaction, and the appropriate primers (
Table 2). The qRT-PCR analyses were performed in at least quadruplicate, using GAPDH and actin as housekeeping genes.
The 2
−ΔΔct method was applied [
22], using the mean ct-value normalized to the housekeeping genes GAPDH and actin, as well as the empty vector cells as reference.
2.8. Statistical Evaluations
Statistical data analysis was performed with Prism 7.0 (GraphPad Inc., La Jolla, CA, USA). Data are represented as means of independent experiments ± standard deviation (SD) as indicated. Significance testing was conducted using Student’s t-test or two-way ANOVA with a Sidak post-hoc test. Results with p < 0.05 were considered statistically significant. The ROUT method with a Q-value of 1% was employed to identify outliers.
4. Discussion
Our study reveals significant alterations in SaOS-2 cells induced by FGF23. On the one hand, we found morphological differences in cellular organelles such as the rough endoplasmic reticulum (rER) and mitochondria, which are significantly enlarged and elongated in SaOS-2 cells overexpressing FGF23. In addition, the contact zones between these organelles are notably more pronounced, suggesting a closer interaction under the influence of FGF23. This morphological enlargement and increased organelle contact may indicate an adaptation to the increased protein synthesis and secretion demands imposed by FGF23 overexpression. Further molecular studies may indicate that not only elevated FGF23 expression but also secretion is associated with higher apoptosis rates after 24–72 h of culture, and upregulated protein levels are associated with ER stress and apoptosis. These findings suggest a mechanistic link between FGF23 signaling and cellular morphology, potentially implicating FGF23 in the regulation of cellular homeostasis and stress response pathways.
While the relationship between stress and other endocrine growth factors has been explored in the literature, FGF23 remains relatively unexplored. Previous research has linked ER stress to the regulation of FGF15, FGF19, and FGF21, with FGF21 also associated with mitochondrial stress [
15]. However, the role of FGF23 in stress pathways has been largely overlooked, although it is known to be associated with inflammation, renal and cardiovascular disease, and proinflammatory pathways [
23].
The morphological abnormalities of the rER and the mitochondria already indicate cellular stress. Underscored by the closer rER-mitochondrial contact zones, which in 2021, Kumar et al. have shown to be a sign of ER stress [
17], as well as the higher apoptosis rate. The literature highlights the critical role of mitochondria-endoplasmic reticulum contact sites (MERCS) in regulating cellular homeostasis and responding to various stress conditions, including tumor microenvironmental stress and ischemic/hypoxic stress [
24,
25,
26], but to date nothing is known about MERCS in the context of X-linked hypophosphatemia (XLH).
Another finding was the elevated FGF23 protein concentration and secretion. In a broader context, endocrine FGFs exert significant influence over various metabolic processes such as bile acid, glucose, fatty acid, phosphate, and vitamin D homeostasis [
27]. Both FGF23 and ER stress have been implicated as contributing factors to conditions like overweight and obesity, which are often found in XLH [
7]. ER stress, for instance, influences obesity by dampening energy expenditure, suggesting its potential as a therapeutic target for obesity treatment [
28,
29]. FGF23, on the other hand, shows positive associations with central and abdominal obesity [
30,
31]. Notably, the induction of ER stress by FGF23 may play a role in the severe obesity observed in patients with XLH, suggesting FGF23 as a potential trigger for metabolic dysregulation and increased fat accumulation.
Further protein analysis revealed significant increases in CHOP, XBP1 (both spliced and unspliced forms), GRP94, eIF2α, and BAX, indicating activation of the unfolded protein response (UPR) and subsequent apoptotic pathways. The protein CHOP (C/EBP homologous protein) is a key mediator of ER stress-induced apoptosis. Its role in promoting cell death in response to prolonged ER stress has been well documented [
1]. Elevated CHOP levels in our study suggest that FGF23 overexpression induces a severe ER stress response leading to apoptosis.
XBP1 (X-box binding protein 1) exists in two forms: unspliced (XBP1u) and spliced (XBP1s). XBP1s is a potent transcription factor activated during the UPR to increase ER protein folding capacity and restore homeostasis [
32]. The significant increase in both forms of XBP1 in our study indicates a robust UPR activation, perhaps aimed at coping with the increased protein load due to FGF23 overexpression.
GRP94 (glucose-regulated protein 94) is a chaperone involved in protein folding within the ER. Its upregulation is a hallmark of ER stress and functions to stabilize and correctly fold nascent proteins under stress conditions [
33]. The increased GRP94 levels in our study indicate an attempt to cope with the increased protein synthesis demands induced by FGF23.
Phosphorylation of eIF2α (eukaryotic initiation factor 2 alpha) is a critical event in the UPR that leads to a reduction in global protein synthesis while selectively translating UPR-specific mRNAs, such as ATF4, to alleviate stress [
34]. The elevated eIF2α levels in our results suggest an adaptive response to reduce the protein load entering the ER.
BAX (Bcl-2-associated X protein) is a long-known pro-apoptotic member of the Bcl-2 protein family [
35]. Its upregulation triggers mitochondrial outer membrane permeabilization (MOMP), leading to cytochrome c release and activation of the apoptotic cascade [
36]. The increased BAX expression in our study underscores the activation of apoptotic pathways due to prolonged ER stress and mitochondrial dysfunction.
To gain a more comprehensive understanding of the precise signaling cascade and the various pathways activated during ER stress, additional Western blots were attempted to provide a more exhaustive depiction of the protein response. Unfortunately, despite numerous attempts and adjustments to protocols, certain persistent challenges remained unresolved. Considerable efforts were made to modify protocols based on prior literature and manufacturer’s guidelines. However, difficulties were encountered with several antibodies targeting key proteins associated with ER stress, apoptosis, and mitochondrial dynamics, including phospho-IRE1alpha, ATF6, phospho-PERK, BiP/Grp78, Calreticulin/CALR3, XBP1, cleaved caspase 3, OMA1, and OPA1.
Nevertheless, these findings are consistent with the role of FGF23 in bone metabolism and its dysregulation leading to skeletal pathologies like X-linked hypophosphatemia (XLH). The observed ER stress and mitochondrial alterations can be interpreted as cellular responses to FGF23 overexpression, mirroring conditions seen in XLH where phosphate homeostasis is disrupted. The increased levels of UPR markers such as CHOP and GRP94 in our study highlight the cellular stress response, which is critical for understanding the pathophysiology of XLH and related disorders.
Moreover, the possible secretion of excess FGF23 from SaOS-2 cells suggests that overexpression not only affects intracellular pathways but also has paracrine effects, potentially influencing neighboring cells and contributing to the systemic nature of XLH. In future experiments, a direct measurement of FGF23 levels should be considered in addition to the indirect approach used in this study.
Future research should focus on elucidating the precise molecular mechanisms by which FGF23 induces ER stress and apoptosis, as well as exploring potential therapeutic targets to mitigate these effects. Understanding the broader implications of FGF23 overexpression on bone and systemic health could lead to improved treatments for skeletal diseases.
5. Conclusions
Our study reveals key insights into the cellular effects of FGF23 overexpression in SaOS-2 cells, highlighting a significant impact on organelle morphology, cellular stress responses, and apoptosis. We observed marked enlargement and elongation of the rough endoplasmic reticulum (rER) and mitochondria, alongside increased contact zones between these organelles, suggesting a closer interaction and adaptation to elevated protein synthesis demands. The upregulation of ER stress markers such as CHOP, XBP1, GRP94, and eIF2α, coupled with increased apoptosis rates, underscores the activation of the unfolded protein response (UPR) and the subsequent induction of apoptosis. These findings link FGF23 signalling, ER stress, and mitochondrial dysfunction, potentially contributing to the pathophysiology of X-linked hypophosphatemia (XLH).
Given the challenges encountered with certain molecular analyses, future research should aim to clarify the precise signaling pathways involved in FGF23-induced ER stress and apoptosis. Further investigation into the paracrine effects of FGF23 and its systemic implications could lead to novel therapeutic strategies for managing XLH and related skeletal disorders.
The study further implicates FGF23 in broader metabolic disturbances, possibly linking it to the obesity commonly observed in XLH patients. The evidence of ER stress and mitochondrial abnormalities provides a basis for exploring the systemic effects of FGF23 overexpression, which may extend beyond bone metabolism to affect overall cellular homeostasis and stress responses.