Osteoblasts Cultured from Osteoporotic Bone: A Comparative Investigation on Human and Animal‐Derived Cells

Abstract

In vitro studies on pathophysiology and innovative treatments of many orthopaedic diseases, based on the investigations of cells from pathologic skeletal tissues, greatly improve basic knowledge of osteoporosis. Primary osteoblast (OB) cultures derived from osteopenic bone from different species (human, rat, sheep) were compared to assess the differences that should be taken into account when performing in vitro biocompatibiliy tests or investigating pharmacological and physical treatments. Primary OB were isolated from osteopenic patients and animals by well‐established methods and their metabolism was assessed with or without 1,25(OH)₂D₃. The greatest significant differences were observed between rat and human cells both under basal conditions and after 1,25(OH)₂D₃ stimulation. In addition, the response to 1,25(OH)₂D₃ stimulation of OBs from osteopenic rats was significantly different from that of human and sheep OB cultures, in terms of NO, OC, IL‐6, and TGF‐β1. Cells derived from osteopenic sheep behaved much more similarly to those from humans, except for a significant difference in terms of TGF‐β1 observed both under basal conditions and after stimulation.

• Osteoblasts
• Human
• Rats
• Sheep
• Osteoporosis

Introduction

The possibility of culturing primary osteoblast cells not only from embryonic bones but also from adult subjects has greatly improved in vitro studies on pathophysiology and innovative treatments of many orthopaedic diseases through investigation of cells from pathologic skeletal tissues (Diaz et al., [1998]; Fedarko et al., [1995]; Sell et al., [1998]).

Because of the wide diffusion of osteoporosis among the ageing population (Ding, [2000]), cultures derived from osteopenic bone have been used by researchers to improve basic knowledge and treatment of such a pathology (Anselme et al., [2000]; Battmann et al., [1997]; Egrise et al., [1999]; Evans et al., [1990]; Fini et al., [2001a], [2001b]; Kassem et al., [1994]; Lomri and Marie, [1990]; Manoglass, [2000]; Marie et al., [1989], [1992], [1993]; Modrowski et al., [1992]; Walsh et al., [2000]; Wong et al., [1994]). Moreover, in recent years an increasing interest has arisen in the use of osteopenic bone‐derived osteoblasts to conduct in vitro tests on biomaterial behaviour prior to animal experimentation (Fini et al., [2001c]; Martini et al., [2001]; Torricelli et al., [2000a], [2001]). A growing number of prosthetic devices, in fact, have been implanted in osteoporotic patients (Fini et al., [1997]) due to the increasing age of the population. Bone rarefaction, source reduction of mesenchymal stem cells and local mediator impairment due to osteoporosis may all affect bone‐biomaterial osteointegration in these patients (Bruder et al., [1998]; Hu, [1997]; Kienapfel et al., [1999]).

On the basis of the researchers' preferences, habits, facilities and knowledge, cells are cultured from aged or postmenopausal patients and various different ovariectomized laboratory animal species, where previous densitometric and/or histomorphometric investigations have demonstrated the development of an osteopenic state secondary to the loss of estrogenic hormones. Apart from bone biopsies in patients (Battmann et al., [1997]; Evans et al., [1990]; Kassem et al., [1994]; Lomri and Marie, [1990]; Marie et al., [1989], [1992], [1993]; Walsh et al., [2000]; Wong et al., [1994]), the rat is, to the current authors' knowledge, the animal species most widely used as donor of osteoblasts for in vitro studies on osteoporosis (Egrise et al., [1999]; Fini et al., [2001a], [2001b], [2001c]; Modrowski et al., [1992]; Torricelli et al., [2000a], [2001]). In addition, since an increasing use is being made of sheep as a model of osteoporosis (Bellino, [2000]; Chavassieux et al., [2001]; Fini et al., [2000]; Giavaresi et al., [2001a]; Martini et al., [2001]; Thorndike and Turner, [1998]), osteoblasts derived from osteopenic sheep bone also appear to be a promising alternative when conducting research on metabolic bone diseases (Torricelli et al., [2000b]).

However, the number of in vitro methods may be restricted by the choice of the cell type and differences between the adopted models may prevent any comparison of data, thus affecting reliability of in vitro experimental results (Kirkpatrick and Mittermayer, [1990]). A survey of the available literature has highlighted that no information exists on interspecies differences between primary osteoblast cultures, in particular, as far as osteopenic bone‐derived cells are concerned.

The aim of the present study was to culture osteoblasts from patients, rats and sheep with established osteoporosis and to evaluate differences between these three species making reference to the most frequently used parameters of cell proliferation, differentiation and synthetic activity. Moreover, for the sake of completeness, the current findings were also compared with those obtained in a previous study on interspecies differences between osteoblasts derived from the healthy bone of patients, rats and sheep (Torricelli et al., [2003]).

Materials and Methods

To collect human bone specimens informed consent was obtained, and procedures approved by the Ethical Committee of the Rizzoli Orthopedic Institute were followed.

The animal study was performed following European and Italian Law on animal experimentation and in accordance with the Animal Welfare Assurance No. A5424‐01 by the National Institute of Health (NIH‐Rockville Maryland USA). The experimental protocol was sent to the Italian Ministry of Health.

Human Osteoblasts

Human osteoblasts (hOB) were isolated sterilely from small specimens of trabecular bone derived from the osteopenic bone tissue of patients undergoing fracture fixation. Lumbar DEXA performed in the selected donors before surgery revealed osteoporosis, and these individuals were therefore characterized as osteoporotic patients.

Rat Osteoblasts

Specimens of trabecular bone were obtained sterilely from the femoral condyles of osteopenic Sprague Dawley female rats aged 13 months. Osteopenia was obtained 4 months after bilateral ovariectomy in 10‐month‐aged rats and was confirmed by densitometric, histomorphometric and biomechanical tests, as described in a previous paper (Giavaresi et al., [2002]). Immediately after euthanasia and in aseptic conditions, femurs were cleaned of soft tissues and the cortical area was removed with a bone cutter from the left condyle in order to expose the trabecular tissue used for osteoblast (rOB) cultures.

Ovine Osteoblasts

Mongrel sheep, aged 3 to 5 years at the beginning of the study, 68 ± 7 kg b/w, were used. Osteopenia was obtained 24 months after bilateral ovariectomy and was confirmed by densitometric, histomorphometric and biomechanical tests, as described in previous papers (Fini et al., [2000]; Giavaresi et al., [2001a], [2001b]). Iliac crest bone biopsies from osteopenic sheep were used to prepare osteoblast (sOB) cultures. Biopsies were obtained sterilely in the vertical direction and performed with a counter‐rotating biopsy needle, having an internal diameter of 8 mm, positioned in an area situated 2 cm behind the anterosuperior iliac spine and 2 cm below the summit of the iliac crest.

Cell Cultures

Trabecular bone fragments from patients, rats and sheep were put in DMEM:F12 serum‐free culture medium and immediately processed to obtain primary cultured osteoblasts. Briefly, bone fragments were repeatedly washed with DMEM:F12 serum‐free medium, and digested in F12 medium with 1mg/ml collagenase for 90 min. at 37°C. The enzymatic reaction was stopped by adding an equal volume of medium with 10% FCS, and the supernatant containing the released cells was collected. Washing and collecting were repeated three times. The cells obtained were pelletted by centrifugation, resuspended, seeded in culture flasks (75 ml), cultured in DMEM medium containing 10% FCS and antibiotics (penicillin 1000 U/ml, streptomycin 10 mg/ml), and incubated at 37°C in a humidified 95%air/5%CO₂ atmosphere. The three different derivation osteoblasts were characterized to assess their osteoblast phenotype (Torricelli et al., [2000a], [2000b], [2002]). The cells were fed every 3 days and released at 80% confluence with 0.05% (w/v) trypsin and 0.02% (w/v) EDTA. Cell suspensions were counted and plated in 24‐ and 96‐well plates at the density of 1 × 10⁴ cells/ml in DMEM supplemented with 50 µg/ml ascorbic acid and 10^{− 8} M β‐glycerophosphate. A supplement of 10^{− 9} M 1,25(OH₂)D₃ was added to half of the wells of each cultures. Cultures were maintained in the same conditions as described above for 72 hours. No bacterial or fungal contamination was found during this period.

At the end of the experiment, supernatants from 24‐well‐plates were collected and aliquots dispensed in Eppendorf tubes for storage at − 70°C. They were then assayed for C‐terminal procollagen type I (PICP), Prolagen‐C enzyme (Immunoassay kit, Metra Biosystem, CA, USA), Interleukin‐6 (IL‐6, Human IL‐6 Immunoassay kit, Biosource International, CA, USA) and Transforming Growth factor‐β1 (TGF‐β1, Quantikine human TGF‐β1 Immunoassay, R&D Systems, MN, USA). Nitric Oxide (NO, Sigma colorimetric assay, St. Louis, MO, USA), Alkaline Phosphatase activity (ALP, Sigma Kinetic method kit, St. Louis, MO, USA) and Osteocalcin (OC, Novocalcin enzyme Immunoassay kit, Metra Biosystem, CA, USA), were tested on the supernatants immediately after collection. Results were corrected for cell number.

The MTT test (Sigma UK) was performed to assess cell proliferation: 80 µl of MTT solution (5 mg/ml in phosphate buffer) and 720 µl of medium were added to the cell monolayers, and the multi‐well plates were incubated at 37°C for a further 4 h. After discarding the supernatants, the dark blue crystals of formazan were dissolved by adding DMSO (800 µl), and were quantified spectrophotometrically at 550 nm. Results were reported as optical density (OD).

Statistical Analysis

Statistical evaluation of data was performed using the software package SPSS/PC⁺ Statistics™ 10.1 (SPSS Inc., Chicago, IL USA). Data are reported as mean ± standard deviations (SD). After verifying the absence of normal distribution, the non‐parametric Kruskall‐Wallis test was done to compare results for the three species. The Mann‐Whitney test was done to assess differences between percentage variations obtained from 1,25(OH)2D3‐stimulated and unstimulated normal (Torricelli et al., [2003]) or osteopenic osteoblasts.

Results

All parameters tested for the different species in osteopenic conditions, with and without 1,25(OH)₂D₃ stimulation, are reported in Tables 1 and 2. Under unstimulated conditions (Table 1), cell viability expressed by the MTT test was quite similar for the three groups (MTT: 1.51–1.97 OD). ALP, OC, NO, PICP and IL‐6 were significantly lower in rOB than in hOB, while there was no significant difference between sOB and hOB values except for TGF‐β1. After 1,25(OH)2D3 stimulation (Table 2), all cultures showed a minor reduction in proliferation rate which was similar for the three different groups. ALP, OC, NO and TGF‐β1 increased, while PICP and IL‐6 decreased. Even after 1,25(OH)₂D₃ stimulation the greatest significant differences were observed between rOB and hOB cultures.

Table 1. Cell Viability and Synthetic Activity of Primary Osteoblasts Derived from Osteopenic Human (hOB), Rat (rOB) and Sheep (sOB) Bone Without 1,25(OH)₂D₃ Stimulation (Mean ± SD, n = 10 Replicates)

Osteopenic cultures	MTT (OD 550 nm)	ALP (IU/l)	OC (ng/ml)	NO (µM)	PICP (ng/ml)	IL‐6 (pg/ml)	TGF‐β1 (pg/ml)
hOB	1.53 ± 0.03	41.5 ± 1.4^a	23.4 ± 1.5^b	3.58 ± 0.14^a	27.5 ± 0.7^a	185 ± 35^a	1066 ± 23^d
rOB	1.71 ± 0.20	12.8 ± 1.7	13.5 ± 1.3	2.48 ± 0.21	10.2 ± 0.5	32 ± 6	641 ± 125
sOB	1.97 ± 0.08	37.0 ± 1.1	22.2 ± 2.9^c	3.20 ± 0.14	17.4 ± 1.4	94 ± 4	442 ± 19

Kruskall‐Wallis Multiple Comparison test: ^a, hOB versus rOB (p < 0.01); ^b, hOB versus rOB (p < 0.05); ^c, sOB versus rOB (p < 0.05); ^d, hOB versus sOB (p < 0.01).

Table 2. Cell Viability and Synthetic Activity of Primary Osteoblasts Derived from Osteopenic Human (hOB), Rat (rOB) and Sheep (sOB) Bone with 1,25(OH)₂D₃ Stimulation (Mean ± SD, n = 10 Replicates)

Osteopenic cultures	MTT (OD 550 nm)	ALP (IU/l)	OC (ng/ml)	NO (µM)	PICP (ng/ml)	IL‐6 (pg/ml)	TGF‐β1 (pg/ml)
hOB	1.51 ± 0.03	68.3 ± 0.7^a	43.4 ± 2.3	4.08 ± 0.29^a	23.0 ± 0.9^a	182 ± 34^a	1267 ± 58^c
rOB	1.67 ± 0.13	36.3 ± 1.5	49.7 ± 2.9	2.54 ± 0.06	10.0 ± 0.7	27 ± 4	944 ± 176
sOB	1.95 ± 0.09	47.0 ± 1.6	36.0 ± 1.4^b	3.33 ± 0.10	14.0 ± 2.3	88 ± 2	454 ± 14

Kruskall‐Wallis Multiple Comparison test: ^a, hOB versus rOB (p < 0.01); ^b, sOB versus rOB (p < 0.01); ^c, hOB versus sOB (p < 0.01).

Table 3 indicates the percentage variations between vitamin 1,25(OH)₂D₃‐stimulated and unstimulated primary osteoblasts from osteopenic and normal bone (Torricelli et al., [2003]) of different origin. Taking into account the percentage variations seen in osteopenic cultures, no significant differences were observed between species in terms of MTT, ALP and PICP. On the contrary, osteoblasts from osteopenic rats showed a significantly different response to 1,25(OH)₂D₃‐stimulation in terms of NO, OC, IL‐6 and TGF‐β1 when compared to hOB and sOB cultures. Comparison with percentage variations of normal cultures demonstrated that osteoblasts from osteopenic bone had different sensitivity to 1,25(OH)₂D₃ stimulation, but there were no differences in response between species in terms of MTT, PICP and TGF‐β1.

Table 3. Percentage Variations Between Vitamin 1,25(OH)₂D₃‐Stimulated and Unstimulated Primary Osteoblasts from Osteopenic Human (hOB), Rat (rOB) and Sheep (sOB) Bone

Parameters	Species			Kruskall‐Wallis
	Humans^1	Rats^2	Sheep^3
Osteopenic cultures
MTT	− 2.81 ± 0.78	− 4.09 ± 6.33^a	− 2.59 ± 5.02	ns
ALP	34.51 ± 15.44^a	41.98 ± 10.29	25.62 ± 5.62^a	ns
NO	12.40 ± 2.91^b	2.59 ± 0.29^a	4.06 ± 4.78	1 vs 2: p < 0.05
OC	66.36 ± 8.81^a	258.15 ± 39.83^a	57.45 ± 5.39^a	1,3 vs 2: p < 0.05
PICP	− 9.97 ± 6.24	− 7.95 ± 7.72^a	− 24.60 ± 7.69^c	ns
IL‐6	− 7.72 ± 2.95^a	− 17.13 ± 0.57^a	− 8.96 ± 2.93^a	1,3 vs 2: p < 0.05
TGF‐β1	23.58 ± 3.88 ^d	50.95 ± 38.15 ^c	2.82 ± 6.18 ^a	2 vs 3: p < 0.05
Normal cultures
MTT	− 2.66 ± 0.67	− 0.26 ± 5.23	− 2.91 ± 2.09	ns
ALP	49.85 ± 1.05	42.84 ± 11.91	34.25 ± 6.00	1 vs 3: p < 0.05
NO	15.87 ± 11.34	4.96 ± 0.94	3.28 ± 6.35	1 vs 2,3: p < 0.05
OC	97.8 ± 8.15	151.82 ± 10.00	60.24 ± 6.43	2 vs 3: p < 0.05
PICP	− 9.98 ± 1.85	− 13.76 ± 9.04	− 13.24 ± 5.91	ns
IL‐6	− 5.49 ± 3.39	− 19.29 ± 1.28	− 3.28 ± 2.28	1,3 vs 2: p < 0.05
TGF‐β1	21.06 ± 1.72	23.08 ± 3.64	24.47 ± 4.51	ns

Mann‐Whitney test between Osteopenic and Normal cultures: ^a, p < 0.0005; ^b, p < 0.001; ^c, p < 0.005; ^d, p < 0.05.

Discussion

The results of the present study, based on cell cultures from osteopenic bone, showed the greatest significant differences between rat and human cells (ALP, OC, NO, PICP, IL‐6) both under basal conditions and after 1,25(OH)₂D₃ stimulation. Cells from osteopenic sheep behaved much more similarly to human cells, with the sole exception of TGF‐β1 which revealed a significant difference under basal conditions and after stimulation.

A similar previous study performed using cells from healthy bone tissue also made it possible to compare osteopenic with normal cells in terms of different responsiveness to 1,25(OH)₂D₃ stimulation.

The current findings confirmed that the behaviour of osteopenic versus healthy sheep cells paralleled data obtained with human osteopenic cells. In both species 1,25(OH)₂D₃ stimulation always led to a significant reduction of some parameters when compared to those observed in healthy cells. Such a finding would appear to confirm lower responsiveness to 1,25(OH)₂D₃ stimulation of osteopenic versus healthy cells.

On the contrary, in this case too, rat osteoblasts showed less homogeneity than human osteoblasts in terms of response to 1,25(OH)₂D₃ stimulation of osteopenic versus healthy cells, and a significant increase in OC and PICP was observed.

The great advantage of in vitro cultures is the use of human cells or tissues, which allows researchers to overcome the differences existing between laboratory animals and human beings. On the other hand, heterogeneity and intraindividual variations confirmed by sequential biopsies in osteoporotic patients, together with the multifactorial aetiology of the disease, may affect reliability of data when using human cultures (Byers et al., [1997]; DeVernejoul et al., [1987]; Wong et al., [1994]). Osteoblasts from healthy and osteoporotic human patients are, for example, greatly influenced by drug intake and systemic pathologies. Apart from the type of bone turnover (high, low or normal), various different osteoblast dysfunctions have been observed histologically and morphometrically in the trabecular bone of osteoporotic patients: a reduced number of osteoblasts, lower activity and efficiency, either singularly or in combination (Byers et al., [1997]; Oelzner et al., [1999]).

This may probably explain the preference for standardized and homogeneous animal models also for in vitro studies, particularly for investigations on osteoporosis. Rat‐derived osteoblasts are more frequently used on account of the deep knowledge and exhaustive characterization of the ovariectomized rat model (Kalu, [1991]; Yamazaki and Yamaguchi, [1989]). Moreover, it is well known that the rat develops an osteopenic state starting from postoperative day 15 after ovariectomy (Thompson et al., [1995]). In recent years the sheep has also become a popular model for the study of osteoporosis probably because researchers now tend to avoid the use of dogs and non‐human primates in experimental studies as much as possible (Bellino, [2000]; Chavassieux et al., [2001]; Fini et al., [2000]; Giavaresi et al., [2001a]; Thorndike and Turner, [1998]). The sheep has been demonstrated to develop progressive osteopenia from 6 to 24 months after ovariectomy (Fini et al., [2000]).

Experimental surgical models are fundamental for the study of biocompatibility and biofunctionality of new materials. However, it should be remembered that in vivo experimentation requires much time and involves high costs. In addition, from an ethical point of view, current regulations recommend that the number of animals to be used should be reduced to the minimum. Consequently, it becomes mandatory to improve in vitro experimentation through validation of more appropriate, and therefore more reliable, in vitro experimental models. For these reasons, primary cultures of osteoblasts derived not only from healthy but also from pathologic bone, are now being increasingly used, and they depict the real in vivo situation in the best way.

To the present authors' knowledge, no data exist concerning differences between various species in terms of osteoblast activity, differentiation and proliferation. The current findings allow researchers to assess better the parameters tested to characterize cultures used, for example in studies on biocompatibility and in vitro osteointegration, on the basis of the species adopted.

The current findings could therefore be useful for researchers who select in vitro models to conduct studies on osteoporosis and obtain data prior to in vivo bone implantation.

Acknowledgments

Financial support for this research was partially provided by IRCCS “Istituti Ortopedici Rizzoli” Bologna, Italian Minister of Health, strategic project “Fratture osteoporotiche,” and Fondazione Cassa di Risparmio. Moreover, the authors would like to thank Patrizio Di Denia, Claudio Dal Fiume, Nicola Corrado, Franca Rambaldi, Alberto Iannì, and Patrizia Nini (Experimental Surgery Department, Rizzoli Orthopaedic Institute) for their technical assistance.