Osteogenic impact of pro-apoptotic caspase inhibitors in MC3T3-E1 cells

Caspases are proteases traditionally associated with inflammation and cell death. Recently, they have also been shown to modulate cell proliferation and differentiation. The aim of the current research was to search for osteogenic molecules affected by caspase inhibition and to specify the individual caspases critical for these effects with a focus on proapoptotic caspases: caspase-2, -3, -6, -7, -8 and -9. Along with osteocalcin (Ocn), general caspase inhibition significantly decreased the expression of the Phex gene in differentiated MC3T3-E1 cells. The inhibition of individual caspases indicated that caspase-8 is a major contributor to the modification of Ocn and Phex expression. Caspase-2 and-6 had effects on Ocn and caspase-6 had an effect on Phex. These data confirm and expand the current knowledge about the nonapoptotic roles of caspases and the effect of their pharmacological inhibition on the osteogenic potential of osteoblastic cells.

www.nature.com/scientificreports www.nature.com/scientificreports/ Apoptotic caspases are activated during differentiation. The activity levels of all investigated caspases were significantly different at the monitored time points (P < 0.05 for all caspases, see Supplement 1). Furthermore, each caspase displayed changes in their activity level at different stages of MC3T3-E1 cell differentiation. At the proliferation stage (between days 0 and 7), caspase-8 activity (Fig. 2B) significantly increased (caspase-8 activity was significantly lower at time point 0 than at all later time points). However, the activity of caspase-9 ( Fig. 2C) decreased in the following period (caspase-9 activity was significantly lower after 14 days of the experiment than it was at time points 0 and 7). Caspase-3/7 activity significantly ( Fig. 2A) increased at the mineralization stage, between days 14 and 21 (Casp 3/7 activity was significantly higher after 21 days of the experiment than it was at all other time points).
The orientation scale of MC3T3-E1 cell differentiation according to Choi 9 is shown in Fig. 2D.
Caspase inhibition changes osteogenic gene expression. In the inhibition experiments, differentiated MC3T3-E1 cells were first challenged by the general caspase inhibitor Z-VAD-FMK, and after 6 days, their osteogenic profile was compared to that of the control (DMSO treated). Of the panel of 84 genes, the expression of two genes, Bglap (the gene encoding osteocalcin) and Phex (phosphate-regulating neutral endopeptidase, X-linked gene), was significantly changed, more than 2-fold (Fig. 3A). The decrease in the regulation of these genes after caspase inhibition was confirmed by real-time PCR (P < 0.05) (Fig. 3B,C). To determine which apoptotic caspases were involved in Bglap and Phex regulation, individual caspase inhibitors were tested (Fig. 4). A statistically significant decrease in Bglap expression was observed after the inhibition of caspase-2 (P < 0.05), caspase-6 (P < 0.001) and caspase-8 (P < 0.01) (Fig. 4A,C,D). In contrast, the inhibition of caspase-3/7 caused a significant (P < 0.01) increase in Bglap expression (Fig. 4B). Similarly, the expression of Phex was also increased (P < 0.05) after the inhibition of caspase-3/7 (Fig. 4E). A decrease in the expression of Phex was found after the inhibition of caspase-6 (P < 0.01) and caspase-8 (P < 0.05) (Fig. 4F,G). The results of the individual caspase inhibition experiments are summarized in Fig. 4H. Moreover, we observed an almost 2-fold decrease (1.87-fold, P < 0.05) in Alpl gene expression after general caspase inhibition (Fig. 3A). This decrease was slightly under the PCR Array threshold, which was based on a −/+2-fold change. Along with the downregulated expression of Alpl detected by PCR Arrays, alkaline phosphatase activity also decreased in the FMK inhibitor-treated group (Fig. 3D,E).

Discussion
Pharmacological pan-caspase inhibition has recently been reported to significantly affect the expression of osteocalcin, a major marker of osteoblastic differentiation 4 . Furthermore, the possible engagement of proapoptotic caspases in cell differentiation has been reviewed 12 . Additionally, the nonapoptotic activation of caspases was demonstrated in MC3T3-E1 cells 4 , the most common in vitro models for osteoblastic lineage.
The differentiation of MC3T3-E1 cells is commonly achieved by the exposure of precursor cells to ascorbic acid 13,14 . Ascorbic acid-stimulated MC3T3-E1 cells synthesize and organize the collagenous matrix and undergo mineralization in manner very similar to that of bone in vivo 15 . Nevertheless, MC3T3-E1 cells cultured in AA/ βGP-free medium can spontaneously undergo the differentiation process induced by cell-cell contact but are unable to completely differentiate 16 . The reason is attributed to collagen I transcription being independent of ascorbic acid induction; however, ascorbic acid is required for the deposition of collagen in the extracellular matrix 17 . Recently, ascorbic acid was demonstrated to modulate signaling pathways in several cell types 18,19 . Therefore, to determine the impact of ascorbic acid induction on the expression of major osteogenic markers, the profile of the stimulated MC3T3-E1 cells was compared to that of the nondifferentiated MC3T3-E1 cells.
Differentiated MC3T3-E1 cells are characterized by high expression levels of alkaline phosphatase (Alp) and increasing activity during the early matrix production period 9,10 . As expected, the level of Alpl was found to be higher in the induced cells. In terms of other genes with changed osteogenic gene expression, the most prominently altered was Tnfsf11, the gene for RANKL. The increased production of RANKL was connected with more effective osteoblast differentiation 20 . Similarly, increased expression of the transcription factor Sp7 had an effect on osteoblast specification 21 , and Igf-1 was expressed in conjunction with MC3T3-E1 cell differentiation 13 . Additionally, in both groups of cultured cells, with and without AA/βGP, several genes were constantly expressed at high levels (Table 1). Thus, after ascorbic acid stimulation of the MC3T3-E1 cells, the changes in osteogenic www.nature.com/scientificreports www.nature.com/scientificreports/ markers were mostly quantitative. Therefore, the main difference in the two cell groups is not based on their rates or states of differentiation but on their mineralized matrix production, as previously reported 15 .
Since one of the most important osteoblastic markers, osteocalcin, was downregulated by general caspase inhibition 4 , the MC3T3-E1 cells were tested to obtain details on the impact of the individual caspases critical for this effect. Widely used specific FMK inhibitors 1 were applied, and caspase-2,-3,-6,-7,-8, and-9 were investigated as major candidates with a broader spectrum of functions, in addition to apoptosis 2 .
The most important decrease in osteocalcin expression was caused by the inhibition of caspase-8. This caspase was previously determined as a molecule with a significant effect on MC3T3-E1 cell differentiation; however, osteocalcin was not evaluated in the previous study 5 . Furthermore, the inhibition of caspase-2, which can be a substrate for caspase-8 in nonapoptotic events 22 , also caused a decrease in osteocalcin expression. However, the www.nature.com/scientificreports www.nature.com/scientificreports/ inhibition of the intrinsic apical caspase, caspase-9, did not have any significant impact on osteocalcin expression in the MC3T3-E1 cells. These results were supported by the bioluminescence-based quantification of caspase activation during MC3T3-E1 cell differentiation. While the activation of caspase-9 ceased after the proliferation phase, the level of active caspase-8 increased, and this increased level was maintained until the final differentiation stage. An analysis of caspase activation during MC3T3-E1 cell differentiation was performed, for the first time, and the results indicated different roles for individual caspases during the progression of osteoblastic differentiation. These observations thus point to extrinsic components being crucial for the decreased osteocalcin levels after general caspase inhibition.
Among the executive caspases, caspase-6 inhibition displayed a negative effect on osteocalcin mRNA levels. The role of caspase-6 in differentiation has not yet been determined. However, caspase-6 is known to be involved in immune B-cell differentiation and to act in Tnf-alpha induction in macrophages 23,24 . Caspase-6 can cleave Satb1 25 , a protein from the same family as Satb2, which regulates osteoblastic differentiation 26 . Moreover, a lower level of Satb2 protein resulted in decreased levels of Runx2, Osx, Bsp and osteocalcin mRNA 27 . Recently, caspase-dependent cleavage of Satb2 was observed during the differentiation processes (preprinted work) 28 .
On the other hand, the inhibition of caspase-3/7 had an opposite effect on osteocalcin expression in the MC3T3-E1 cells, indicating specific roles for caspase-3 and caspase-7 connected with their major functions in apoptosis. Caspase-3 and caspase-7 are known to be connected with the intrinsic apoptotic pathway; however, along with caspase-9, they have distinct roles in this process 29 . Unfortunately, due to overlapping substrate specificity 30 , FMK inhibition of caspase-3 and caspase-7 cannot be performed separately.
Decreased expression of Phex was observed as was that of osteocalcin. Phex is a transmembrane molecule expressed by osteoblasts and its deletion is linked to hypophosphatemia 31,32 . In the individual caspase inhibition experiments, Phex expression was downregulated as a consequence of caspase-8 and caspase-6 inhibition, similar to osteocalcin.
In osteoblasts, Phex is downregulated by exposure to parathyroid hormone (PTH) 33 . In fact, PTH induces the differentiation of MC3T3-E1 cells by activating the Wnt/β-catenin pathway 34 . Caspases-8,-6 and-3 are known to directly lead to β-catenin proteolysis in vitro 35 . Caspases may thus link these pathways.
In conclusion, pharmacological caspase inhibition in osteoblastic MC3T3-E1 cells can decrease the expression of two clinically important osteogenic factors: osteocalcin and Phex. In both cases, the nonapoptotic effects of caspase-8 and caspase-6 were involved in this process. In vivo, deregulated levels of osteocalcin or Phex have a great influence on bone homeostasis and disorders. For example, mutations in the Phex gene are connected with rachitic changes in the bone growth plate, and this effect is mimicked by caspase inhibitors 36 . Osteocalcin is clinically important not only in bone regulation but also in many other tissues where it has a mediating role 37 . The regulation of Phex expression by caspase inhibition was described for the first time as was the specification of individual caspases in osteocalcin and Phex pathways. The apparent contribution of proapoptotic caspases and the possibility of their targeted inhibition can be considered in future therapeutic strategies.
For the experiments, passages 6-20 were used. The cells were seeded and cultured for 21 days in nondifferentiation and differentiation conditions in parallel. Differentiation medium was prepared as described above but with the addition of 10 mM β-glycerolphosphate (βGP) and 50 µg/ml ascorbic acid (AA). The cultured cells were passaged several times to avoid the formation of a dense collagenous matrix. After 3 weeks of culture, the cells were either fixed on plates with 4% PFA and stained with alizarin red, lysed in RLT buffer or used for caspase inhibition experiments, as described below.
Caspase inhibition. After 21 days, differentiated cells were used for caspase inhibition. The cells were seeded at a density of 5000 cells per cm 2 and cultured for 6 days (as reported previously 4 ) in the presence of caspase inhibitors: general caspase inhibitor Z-VAD-FMK and inhibitors for individual apoptotic caspases (Caspase-2,-3,-6,-8,-9). Pharmacological inhibitors (FMK001, FMK003, FMK004, FMK006, FMK007, and FMK008, R&D Systems) were added to the culture medium at a concentration of 100 μM, according to the manufacturer's instructions and previous studies 4 . Controls were generated using DMSO as an inhibitor vehicle at the same concentration. The medium with different treatments was changed every 2 days. The effectiveness of the individual caspase inhibitors was verified with Caspase-Glo assays (Supplement 2), and the constant inhibition effect of the FMK inhibitors in cells was confirmed previously 38 RNA isolation, real-time PCR, PCR Arrays. Cells for RNA isolation were lysed in 350 µl of RLT buffer with β-mercaptoethanol. RNA was isolated by using a RNeasy Mini Kit (Qiagen). First-strand cDNA was transcribed using Super Script VILO (Invitrogen); qPCR was performed in 10 μl of a final reaction mixture containing the one-step GB Ideal PCR Master Mix (Generi Biotech), Osteocalcin (Mouse Bglap, Mm03413826_mH), and Phex (Mouse Phex, Mm00448119_m1) expression was detected by using a TaqMan Gene Expression Assay (Thermo Fisher Scientific). The expression levels were calculated using the ∆∆CT method with normalization based on actin levels (Mouse Actb, Mm02619580_g1).
Osteogenic-related gene expression was analyzed by RT 2 Profiler PCR Array Mouse Osteogenesis (PAMM026Z, Qiagen), which allows for expression comparisons of 84 genes. (2020) 10:7489 | https://doi.org/10.1038/s41598-020-64294-9 www.nature.com/scientificreports www.nature.com/scientificreports/ Caspase activity assay. Caspase activity was measured with Caspase-Glo bioluminescence assays (Promega) using a laboratory-built luminometer suitable for single-cell caspase detection, as described previously 39 . Currently, a Caspase-Glo reaction kit is available only for caspase-3/7,-8 and-9. The caspase-3/7,-8 and-9 activity levels were measured during the differentiation of MC3T3-E1 cells on day 1 and then periodically after every 7 days until the end of the differentiation phase on day 21. After being harvested by trypsinization, the cells were centrifuged and suspended in fresh medium. Under phase-contrast microscopic inspection, 50 cells were collected using an ICSI micromanipulator (Eppendorf) equipped with a 20-μm o.d. holding micropipette (Microtech) and transferred to 5-μL glass vials filled with 4 μL of Caspase-Glo; this reagent contained 60 μM MG-132 proteasome inhibitor (Promega) to suppress the background signal. The vials were incubated at room temperature, and the signal intensity was periodically measured until the bioluminescence emission reached a steady state. All measurements were performed in triplicate separately for each caspase.
Staining for alkaline phosphatase. Statistical analysis. Real-time PCR data were assessed using the ∆∆CT method. Three independent biological samples were analyzed, and the reactions were performed in triplicate for each sample. Expression level was related to that of the control cells, and the differences were evaluated using a t-test. The threshold for statistical significance was set to P < 0.05.
The expression data of the PCR Array were statistically evaluated by the Qiagen Data Analysis Center using a t-test, as recommended by the producer. Three independent biological samples were analyzed for each stage. The housekeeping genes included Actb, B2m, Gapdh, Gusb, and Hsp90ab1. Significance was determined as P < 0.05, and the threshold for fold changes was ±2.
Caspase activity levels were statistically analyzed in a Statistica software system (version 13, Dell Inc., 2016) using one-way ANOVAs, and Tukey HSD was used as a posthoc test. The threshold for significance was set to P < 0.05. Three independent samples were analyzed for each time point.