Bone marrow stromal cells (BMSCs CD45‐/CD44+/CD73+/CD90+) isolated from osteoporotic mice SAM/P6 as a novel model for osteoporosis investigation

Abstract Available therapies aimed at treating age‐related osteoporosis are still insufficient. Therefore, designing reliable in vitro model for the analysis of molecular mechanisms underlying senile osteoporosis is highly required. We have isolated and characterized progenitor cells isolated from bone marrow (BMSCs) of osteoporotic mice strain SAM/P6 (BMSCSAM/P6). The cytophysiology of BMSCSAM/P6 was for the first time compared with BMSCs isolated from healthy BALB/c mice (BMSCBALB/c). Characterization of the cells included evaluation of their multipotency, morphology and determination of specific phenotype. Viability of BMSCs cultures was determined in reference to apoptosis profile, metabolic activity, oxidative stress, mitochondrial membrane potential and caspase activation. Additionally, expression of relevant biomarkers was determined with RT‐qPCR. Obtained results indicated that BMSCSAM/P6 and BMSCBALB/c show the typical phenotype of mesenchymal stromal cells (CD44+, CD73+, CD90+) and do not express CD45. Further, BMSCSAM/P6 were characterized by deteriorated multipotency, decreased metabolic activity and increased apoptosis occurrence, accompanied by elevated oxidative stress and mitochondria depolarisation. The transcriptome analyses showed that BMSCSAM/P6 are distinguished by lowered expression of molecules crucial for proper osteogenesis, including Coll‐1, Opg and Opn. However, the expression of Trap, DANCR1 and miR‐124‐3p was significantly up‐regulated. Obtained results show that BMSCSAM/P6 present features of progenitor cells with disturbed metabolism and could serve as appropriate model for in vitro investigation of age‐dependent osteoporosis.


| INTRODUC TI ON
Osteoporosis represents the most common bone disease in elderly patients of both sexes and all races, occurring worldwide. 1 Moreover, its incidents are more frequent in well-developed and ageing societies. 2 Notably, more than 200 million citizens worldwide currently suffer from osteoporosis, and ~8.9 million fractures are caused by osteoporosis-related fractures. 3 Furthermore, the specialists estimate that within 50 years, osteoporosis will reach the scale of global epidemy. The sudden termination of physical activity of osteoporotic patients, especially those suffering from bone fractures, becomes an even higher sociological cost that eliminates them from social life. 4 Therefore, the investigation of novel concepts improving knowledge about the molecular mechanism of osteoporosis occurrence is critical for developing new therapeutic strategies and very much needed.
Osteoporosis is characterized by low bone mass as a result of impaired bone mineralisation, leading to reduced bone mechanical properties, increasing fracture risk. 5 The pathophysiological mechanism of osteoporosis includes the deterioration of bone metabolisms. It is a consequence of several factors, including the advantage of bone resorption over bone formation process. The impairment of bone remodelling is caused by an imbalance between osteoclasts and osteoblasts, that is between bone-degrading and boneproducing cells. 6 The recruitment of osteoclasts and osteoblast at the bone remodelling site requires activation of the plethora of molecular signals including hormones, cytokines, growth factors and non-coding RNAs, including long non-coding RNA (lncRNA) and microRNA (miRNA), which mediates the interaction between bone cells and progenitor cells. 7 Moreover, the recruitment of bone marrow stromal cells (BMSCs), which are a source of progenitor cells at the bone remodelling site, guarantees a supportive role during new bone formation.
Bone marrow-derived stromal cells (BMSCs) are multipotent stem cells with self-renewal capacity. 8 The population of showing features of BMSCs was described for the first time by Alexander Friedenstein and colleagues. 9,10 Since that time, the knowledge regards BMSCs biology and nature are still extensively investigated. 8,11 According to the current statement of International  12 The self-renewal potential, associated with increased proliferative capacity, shed a promising light for various clinical application of BMSCs transplants. Numerous studies identified the molecular mechanisms of BMSCs, emphasizing their beneficial effects in the course of fractures bone regeneration. [13][14][15] The progenitor cells of bone marrow express critical markers for new bone formation, which included alkaline phosphatase (ALP), bone morphogenetic protein 2/4 (BMP-2/4), osteoprotegerin (OPG), receptor activator of nuclear factor B (RANK), RANK ligand (RANKL), osteocalcin (OCL), osteopontin (OPN), winglesstype MMTV integration site family (Wnt) proteins and signalling through parathyroid hormone receptors. 16,17 The pro-regenerative potential of BMSCs is also mediated by their paracrine activity and shedding the extracellular microvesicles (ExMV's), which are particularly rich in growth factors, miRNAs or ln-cRNAs. [18][19][20] Moreover, various miRNA and lncRNA have been shown recently to be involved in the mediation of balance between cell populations of osteoblast-like or osteoclast-like nature. 7,21 Recent data suggest that the immunomodulatory activity of BMSCs makes them an even more promising therapeutic tool in terms of cell-based therapies in osteoporosis treatment. 22 However, the metabolic imbalance associated with osteoporosis affects the activity of BMSCs. The cells are losing their valuable biological properties, such as proliferative activity and multipotency.
Moreover, BMSCs isolated from osteoporotic patients show apoptotic phenotype and accumulation of oxidative stress factors, which seriously reduce their viability. BMSCs isolated from osteoporotic rat show increased expression of several markers related to adipogenesis and simultaneously reduced expression of master regulators essential for bone formation. 23,24 Thus, BMSCs are currently extensively investigated since understanding their molecular nature under osteoporosis might bring us closer to understanding the molecular mechanism involved in osteoporosis development.
For that reason, in this study, for the first time, we have isolated BMSCs derived from osteoporotic mice strain SAM/P6 (BMSC SAM/ P6 ) and described it as a genuine and relevant in vitro model, allowing determination of the molecular basis of osteoporosis development Current models rely on BMSCs isolated from ovariectomized rats 23,24 or patients with osteoporosis 17,25,26 However, still the molecular aspects of BMSCs cytophysiology affected by osteoporosis has not been fully elucidated. Here, we have characterized BMSC SAM/P6 proliferative and metabolic activity and determined the expression of common phenotype markers, critical for stemness.
Moreover, using cytometric-based tests, we have confirmed lowered metabolism of the cells, associated with depolarization of the mitochondrial membrane, intracellular accumulation of reactive oxygen species (ROS), accompanied with down-regulation of mitofusin 1 (MFN-1) protein expression in osteoporotic BMSCs compared with BMSCs isolated from healthy tissue. Additionally, we have evaluated the multipotency of BMSC SAM/P6 and determined the expression profile of bone-related markers (lncRNA-miRNA-mRNA axis). The molecular pattern of miRNAs expression, for example miR-21-5p or miR-124-3p in osteoporotic murine BMSCs has not been previously evaluated by other authors. Moreover, the analysed miRNAs were referred to the expression of lncRNA (DANCR1) and mRNAs, including Runx-2 (runt-related transcription factor 2), Trap (tartrateresistant acid phosphatase) or Opn (osteopontin). Obtained results were compared with BMSCs isolated from healthy BALB/c mice (BMSC BALB/c ). Here, we characterized novel bone marrow multipotent stem cells that could be used in future research regarding osteoporosis, especially attributed to ageing.

| Isolation procedure and propagation of bone marrow-delivered progenitor stem cells
The bone marrow-derived stromal cells were isolated from long bones of mice collected from lower limbs. After removal, bones were washed twice in Hank's Balanced Salt Solution (HBSS) with 1% addition of antibiotics (P/S-penicillin and streptomycin). The distal parts of every bone were cut out. Following that, bone marrow was isolated by its flushing from the medullary canal with an insulin syringe U-40 (29G X 1/2′′ needle) filled with HBSS as described previously. 27 The cells were collected and centrifuged two times (300 × g, 4 min).
Subsequently, the isolated cells were counted by Muse ® Count & Viability Kit (Merck®; cat. no.: MCH100102, Poznan, Poland). The procedure was carried out following protocol provided by the manufacturer. Further, the cells were inoculated on the 24-well dishes at density 800 000 cells/well and suspended in 500 µL of complete growth medium (CGM), consisted of Ham's F-12 Nutrient Mixture (F-12) supplemented with 15% of foetal bovine serum (FBS) and 1% of antibiotics (P/S). The cultures were propagated in sterile conditions using CO 2 incubator with constant parameters: 5% CO 2 , 37°C and 95% humidity. After 24 hours of culture, the media were removed and replaced by the fresh media in order to eliminate hematopoietic cell lineage. 28 All reagents used for cell cultures (media, HBSS, antibiotics, FBS) were derived from Sigma-Aldrich (Poznan, Poland).
During propagation culture condition, growth pattern, as well as cells morphology were monitored using Axio Observer A1-inverted microscope (Zeiss, Oberkochen, Germany) and documented with Canon PowerShot digital camera (Woodhatch, UK). The photographs were taken under 100× and 400× magnification.

| Analysis of BMSCs metabolic activity
The analysis of BMSC BALB/c and BMSC SAM/P6 metabolic activity was carried out by the use of well-established Alamar Blue test. After five days of cultures propagation, the cells were washed once with HBSS and 350 µL of CGM with 10% addition of resazurin dye solution (Tox8-1KT, Sigma-Aldrich, Munich, Germany) was added.
Subsequently, the cultures were incubated for 2 hours in 37°C (5% CO 2 and 95% humidity). After incubation, the supernatant was removed and transported to the 96-well dish in six repetitions. The absorbance was measured at the wavelengths of 600 and 690 nm.
The metabolic activity of BMSCs was calculated using formula:

| Immunocytochemical detection of CD44, CD73, CD90 and CD45
In order to characterize the isolated cells, surface markers typical for BMSCs were stained. After reaching ~80% of confluency, the

| Chondrogenic, osteogenic and adipogenic differentiation of BMSCs
In order to prove the multipotent abilities of isolated BMSCs, chondrogenic, osteogenic and adipogenic differentiation of cultures were performed with differentiation media. The fresh chondrogenic and adipogenic media were changed twice a week and maintained for 7 days.
The fresh osteogenic medium was changed twice a week. The osteogenic conditions were maintained for 10 days. After the differentiation process, the cultures were collected for subsequent analyses.

| Evaluation of BMSCs extracellular matrix composition and neutral lipids staining after differentiation conditions
The protocol of extracellular matrix staining was described previously. 21,29 Briefly, differentiated cultures of BMSCs were fixed with

| Immunocytochemical detection of RUNX-2, OPN and TRAP
The procedure of RUNX-2, OPN and TRAP protein staining using confocal microscopy was mentioned in Section 2.3. and was described previously in detail. 21
The whole procedure was performed accordingly to manufacturer's protocol and our previous experiment. 35 Briefly, Caspase buffer was diluted 10× in DEPC-treated water and the MultiCaspase Reagent Stock Solution was diluted in 50 µL of DMSO. Other reagents were prepared for the analysis as it was described elsewhere. 35 According to manufacturer instructions, the analysis of caspases activation was based on membrane permeable VAD-peptide that can detect multiple caspases for example caspase 1, −3, −4, −5, −6 −7, −8 and −9.
Stained samples were incubated for 30 min (37°C, 95% humidity, 5% CO 2 ) and 150 µL of Muse TM 7-AAD Working Solution was added in order to detect dead cells. The caspases activity profile was measured using Muse TM Cell Analyzer. Each analysis was performed in triplicate. The gating procedure of cells' populations was based on the positive and negative controls. 32-34

| Analysis of BMSCs reactive oxygen species activation
The analysis of reactive oxygen species activation (ROS) was measured using the Muse™ Oxidative Stress Kit (Merck ® ; cat. no.: MCH100111, Poznań, Poland). The staining procedure was performed accordingly to producer's protocol and described previously. 32 Briefly, after trypsinisation 10 µL of cells were added to 190 µL of Muse Oxidative Stress Working Solution and incubated 30 minutes in 37°C. The staining reagent was provided by manufacturer and based on dihydroethidium (DHE), which is widely used for ROS detection in many cell cultures. 36 Then, the oxidative stress was measured using Muse TM Cell Analyzer. Each analysis was performed in triplicate. The gating procedure of cells' populations was based on the positive and negative controls. 32-34

| Analysis of BMSCs mitochondrial membrane depolarisation status
The measurements of mitochondrial membrane depolarisation were

| Analysis of mRNA, miRNA and lncRNA expression
The transcripts levels for selected mRNA, miRNA and lncRNA were evaluated using reverse transcription quantitative polymerase chain reaction (RT-qPCR). After experiment, cultures were homogenized using 1 mL of Extrazol ® (Blirt DNA, Gdańsk, Poland).
The isolation procedure of RNA was performed accordingly to manufacturer's protocol. After isolation, total RNA was diluted in molecular grade water (Sigma-Aldrich, Poznan, Poland). The quantity and purity was evaluated spectrophotometrically at 260 and 280 nm wavelength (Epoch, BioTek, Bad Friedrichshall, Germany). B2m (beta-2-microglobulin). The transcript levels for miRNA were calculated in relation to a snU6 gene. The list of used primers are enclosed in Table 1.

| Statistical analyses
Experimental values are presented as means of obtained from at least three technical repetitions and they supplemented with standard deviation (±SD). The statistical calculations and data presentation was done with GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA). The data were analysed using Student's t test.
Differences were considered as statistically significant at P < .05.

| Characterization of isolated BMSCs-growth pattern, metabolic activity and multipotency
Isolated BMSCs, derived both from SAM/P6 and BALB/c mice, were successfully cultured in the monolayer system in sterile plastic dishes and maintained in the CO 2 incubator with constant conditions: 37°C, 5% CO 2 and 95% of humidity ( Figure 1A). was reduced (P <.05), and culture growth was impeded ( Figure 1A and B).

TA B L E 1 The list of primers used for RT-qPCR
Immunocytochemical staining showed that BMSC SAM/P6 and BMSC BALB/c expressed typical cell surface confirming their mesenchymal origin (CD44, CD73 and CD90) and did not express CD45 characteristic for haematopoietic cells ( Figure 1C). Importantly, the expression of CD44 and CD90 in BMSC SAM/P6 decreased (P < .05 and P < .001, respectively), compared with BMSC BALB/c ( Figure 1D).
After reaching around 80% of confluency, the cultures were  The BMSCs isolated from SAM/P6 mice were characterized by lowered expression of RUNX-2 (P < .05) and OPN (P < .001) proteins, key factors regulating osteogenic potential of progenitor cells ( Figure 2B and D). Simultaneously, the expression of an osteoclastic marker, that is TRAP was increased (P < .001) in BMSC SAM/P6, while BMSC BALB/c cultures showed reversed phenotype. (Figure 2F).

| The BMSCs ultrastructure and expression of osteogenic markers
Obtained results are consistent with decreased deposition of calcium in ECM formed by BMSC SAM/P6 (Figure 1E and H).

| Decreased expression of osteogenic markers is characteristic for BMSC derived from mice with osteoporotic phenotype
The analysis performed by the use of RT-qPCR technique indicated that BMSC SAM/P6 are characterized by the reduced level of transcripts associated with proper osteogenesis and bone homeostasis.

| Increased levels of Trap and osteogenesisdependent non-coding RNAs distinguish BMSC derived from osteoporotic mice
The RT-qPCR analysis of mRNAs and non-coding RNAs associated with osteoclastogenesis and bone loss confirmed the osteoporotic phenotype of BMSC SAM/P6 cultures. The expression level of noncoding DANCR1 (differentiation antagonizing non-protein coding RNA 1) was significantly up-regulated (P < .05) in BMSC SAM/ P6 ( Figure 4A). Moreover, it has been shown that the expression of Trap (tartrate-resistant acid phosphatase) in BMSC SAM/P6 was significantly elevated (P < .01), compared with BMSC BALB/c ( Figure 4B).
The expression profile of transcripts determined in BMSC BALB/c and BMSC SAM/P6 was also presented as a heatmap in Supporting Information ( Figure S1).

| BMSC from osteoporotic mice are characterized by apoptotic phenotype and increased oxidative stress
It has been shown that BMSCs delivered from osteoporotic SAM/ P6 mice were characterized by a lower ratio of viable cells (P < .01; Figure 5A and B) and a greater ratio of dead cells (P < .01; Figure 5A and C), compared with BMSCBALB/c. Moreover, BMSCSAM/ P6 had a significantly greater ratio of cells that undergo apoptosis (P < .01; Figure 5A and D). The analysis performed by the use of RT-qPCR technique showed no differences in the expression of important markers associated with programmed cell death, that is pro-apoptotic Bax (Bcl-2-associated X protein) and anti-apoptotic Bcl-2 (B-cell lymphoma 2) ( Figure 5F and G). However, BMSC SAM/P6 expressed more transcript for Mmp-9 (P < .05; metalloproteinase 9), an additional marker of apoptotic cells. 37 The examination of oxidative stress in the isolated BMSCs indicated an increased accumulation of reactive oxygen species (ROS) in BMSC SAM/P6 (P < .01; Figure 5H and J). In turn, the reactivity of ROS in BMSC BALB/c was lesser (P <.01; Figure 5H and I). Furthermore, it has been shown that BMSC SAM/P6 accumulates more transcripts for Igf-1 (P < .001; insulin-like growth factor 1). In correspondence with increased ROS, the up-regulated Igf-1 may be associated with the pro-inflammatory activity of progenitor cells derived from bone marrow of SAM/P6 mice ( Figure 5K).

| Decreased viability of BMSC from osteoporotic mice can be a caspaseindependent process
The flow cytometry-based measurements showed that BMSC SAM/ P6 were characterized by a lower cell ratio with activated caspases (P < .01; Figure 6A and D). However, the ratio of viable cells was decreased (P < .01) in BMSC SAM/P6 ( Figure 6A and B) and ratio of dead cells was increased in BMSC SAM/P6 (P < .001; Figure 6A and C), compared with BMSC BALB/c . Obtained data suggested that the deteriorated viability of BMSC SAM/P6 is not a result of caspase-dependent processes.

| BMSCs with osteoporotic phenotype are characterized by mitochondrial membrane depolarisation and impaired dynamics of mitochondrial network
The mitochondrial membrane depolarisation status indicated on el-  The photographs of cultures that undergo adipogenesis were taken using confocal microscopy in order to visualize cells' nuclei (DAPI) and lipid droplets (LipidTox). The confocal images were captured under 630-fold magnification, and the scale bar is equal 40 µm. The stainings intensity measured in differentiated cultures was presented as bar graphs (H, I and J). Significant differences between groups are indicated with asterisks: * P < .05, ** P < .01, *** P < .001. Non-significant differences are marked as ns

| D ISCUSS I ON
impaired phenotype, what affected disturbed recruitment and reduced regenerative potential. 26 Numerous studies confirmed that increased patient age correlates with decreased beneficial properties of progenitor cells, namely "stemness" depending on lowered self-renewal potential and causing defective extracellular matrix formation. 41,42 Moreover, progenitor cells with senescence phenotype exhibit seriously deteriorated multilineage differentiation potential, limiting Our previous study has shown that increased patient age corresponds with adipogenic and osteoclasts like phenotype of multipotent stromal cells, associated with high expression of p53. 43 However, the mechanism of that phenomenon is still poorly investigated. For that reason, in this study, we investigated the expression of critical transcripts involved in regulating osteogenesis on mRNA, miRNA, and lncRNA. We have found, that BMSC SAM/P6 expressed higher osteoclasts related transcripts including Trap and long non-coding RNA DANCR1 together with elevated expression of miR-7a-5p, miR-17-5p,

F I G U R E 3
The analysis of genes' transcriptomes (mRNA) associated with osteogenesis and bone homeostasis. The measurements were performed with RT-qPCR technique, calculated using RQMAX method and presented in a log scale. The analysed targets were Runx-2 (A); Coll-1 (B); Opg (C); Ocl (D) and Opn (E). Significant differences between groups are indicated with asterisk: *P < .05, **P < .01, ***P < .001.
Non-significant differences are marked as ns F I G U R E 4 The analysis of genes' transcriptomes (mRNA and non-coding RNA) associated with osteoporosis and bone loss. The measurements were performed with RT-qPCR technique, calculated using RQMAX method and presented in a log scale. The analysed targets were DANCR1 (A); Trap (B); miR-21-5p (C); miR-124-3p (D); miR-7a-5p (E); miR-17-5p (F); miR-145-3p (G); miR-203a (H) and miR-223-3p (I). Significant differences between groups are indicated with asterisk: *P < .05, **P < .01, ***P < .001. Non-significant differences are marked as ns within resorption sites. The increased expression of TRAP was characteristic for BMSC SAM/P6 and determined both at mRNA, as well as protein level. TRAP activity correlates not only with resorptive activity of osteoclasts, but also might be implicated in autoimmune disorders. 56 The complex role of TRAP has been explained by its key role in both bone homeostasis and the immune system. 56  . The representative graph of Western blot (E). The mitochondrial dynamics was evaluated by protein level of MFN-1 (F) and PINK1 (G). Significant differences between groups are indicated with asterisk: *P < .05, **P < .01, ***P < .001. Non-significant differences are marked as ns patients. Moreover, Zannata et al showed that RUNX-2 regulates bone formation and remodelling throughout life, and its expression profile is age-dependent and correlates with bone mineral density (BMD). 76 This study indicates that BMSC SAM/P6 exhibit a high similarity with progenitor cells isolated from osteoporotic human patients, thus becoming a novel model for in vitro study to develop new and efficient therapeutic strategies for age-related (senile) osteoporosis.
The SAM/P6 model of osteoporosis has a valuable impact on the preclinical examinations of age-related osteoporosis and might help to develop more effective strategies of treatment. Here, we have performed profound characteristic of BMSC SAM/P6 cytophysiology, with particular attention on self-renewal and multilineage potential.
The BMSC SAM/P6 show features of ageing and senescent cells, with lowered pro-regenerative function, related to decreased osteogenic potential and enhanced accumulation of a lipid vacuoles. Given the limited access to human cells with age-related and senescence phenotype typical for senile osteoporosis, the BMSC SAM/P6 can be used successfully as a reliable model to explore and establish novel agents for osteoporosis treatment.

ACK N OWLED G EM ENTS
The Moreover, acknowledgements go to Ariadna Pielok for the help with animals' killings, as well as Klaudia Marcinkowska for the guidance during in vitro procedures.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used in this study are available from the first author and corresponding author on reasonable request.