Table 1.

Effects of tanshinone on bone health.

Studies	Study Design	Changes with Tanshinone Treatment
Cell Culture Studies
Liu et al. [27]	Cell: hPDLSC from the premolars of 20 donors without oral or systematic diseases (10 men and 10 women aged 12–25 years old) Model: osteogenic differentiation Treatment: 2.5 and 5 μM of T-IIA for 48 h Negative control: Untreated cells Positive control: N.A.	↑ osteogenic differentiation ↑ OCN, OPN, Runx2, ALP gene expression vs. negative control
Qian et al. [28]	Cell: BM-MSCs isolated from the tibia and femur of BALB/cJ mice (4–6 weeks old) Model: dexamethasone-induced osteogenic differentiation Treatment: 1, 5, 10 and 20 μM of T-IIA for 7 days Negative control: Untreated cells Positive control: N.A.	↑ ALP expression vs. negative control ↑ OPN, OPG, collagen 1 and ↓ RANKL vs. negative control ↑ Runx2 and BMP 4 vs. negative control
Kim and Kim [35]	Cell: C2C12 cells Model: BMP-2-induced osteoblast differentiation Treatment: 2.5, 5, 10 and 20 μM of T-IIA for 7 days Negative control: Untreated cells Positive control: N.A.	↑ BMP-2-induced osteoblast differentiation and ALP production ↑ activation of osteogenic genes (ALP, OCN, Runx2, BMP-2, 4, 6, 7, 9)
Wang et al. [36]	Cell: MC3T3-E1 cells Model: osteoblast differentiation Treatment: 5, 10, 15 and 20 μM of T-IIA for 48 h Negative control: Untreated cells Positive control: N.A.	↑ Runx2, Osx expression and ALP activity vs. negative control ↑ BMP-2 protein expression vs. negative control ↑ activation of JNK pathway vs. negative control
Li et al. [38]	Cell: MC3T3-E1 cells Model: dexamethasone-induced osteoblast apoptosis Treatment: 1 μM of T-IIA for 24 h Negative control: Untreated cells Positive control: N.A.	↓ dexamethasone-induced osteoblast apoptosis by inhibiting Nox4 expression
Zhu et al. [37]	Cell: Osteoblast cells from 10 weeks old female Wnt1sw/sw mice Model: H2O2-induced osteoblast apoptosis Treatment: 1.5 mg/mL of T-IIA for 24 h Negative control: PBS-treated cells Positive control: 1.5 mg/kg alendronate	↑ osteoblasts viability and ↓ osteoblast apoptosis ↓ ALP, H2O2, ROS, SOD, TBARS and RNS levels vs. osteoblasts from mice treated with alendronate ↓ caspace-3, and Apaf-1 expression and ↑, Bcl-2, TRAF 1, IAP and p53 expression vs. osteoblasts from mice treated with alendronate ↓ activation of NF-κB phosphorylation, NF-κB activity, TNF-α, iNOS and COX2 expression vs. osteoblasts from mice treated with alendronate ↓ p65, IKK-β and IκBα vs. osteoblasts from mice treated with alendronate
Cheng et al. [39]	Cell: RAW264.7 cells and BMMCs isolated from the femoral bone marrow of 8-week-old C57BL/6 mice. Model: RANKL-induced osteoclastogenesis Treatment: 1, 2 or 5 μg/mL of T-IIA for 7 days Negative control: Untreated cells Positive control: N.A.	↓ osteoclastogenesis and osteoclast function vs. negative control ↓ TRAP+ cells (in a dose-dependent manner) vs. negative control ↓ numbers and size of actin ring structures (a characteristic feature of mature osteoclasts during osteoclastogenesis) vs. the negative control ↓ RANKL-induced osteoclast differentiation at early stages vs. negative control ↓ osteoclastogenesis-related genes (TRAP, MMP- 9, cathepsin K, CTR, and TRAF6) vs. negative control ↓ RANKL-induced activation of the NF-κB, MAPK and Akt pathways ↓ ERK, JNK, c-Fos and Akt expression in osteoclasts
Lee et al. [29]	Cell: Primary osteoblasts from calvarial cells of ICR newborn mice/bone marrow cells obtained from tibiae of 6- to 7-week-old ICR mice (coculture) Model: M-CSF-induced osteoclast differentiation Treatment: 0.5, 1 and 2.5 μg/mL of T-IIA, T-I, C-T, D-T for 3 days (osteoblasts) and 6-7 days (osteoclasts) Negative control: Untreated cells Positive control: N.A.	↓ TRAP-positive multinuclear cells vs. negative control ↓ viability of bone marrow cells following treatment with T-I, C-T, D-T NS in the viability of bone marrow cells following treatment with T-IIA
Panwar et al. [30]	Cell: mononuclear cells from human bone marrow and bone marrow cells from femur and tibia from 4 months old mice Model: M-CSF and RANKL-induced osteoclastogenesis Treatment: 1 and 3 μM of T06 for 72 h Negative control: DMSO (1%)-treated cells Positive control: N.A.	↓TRAP-stained osteoclasts and toluidine-stained resorption events in mouse and human osteoclasts vs. negative control ↓ total eroded surface in human and mouse osteoclasts vs. negative control ↓ CTx-1 expression vs. negative control NS for osteoclastogenesis in bone marrow mononuclear cells
Kim et al. [34]	Cell: MC3T3-E1 cells and bone marrow cells isolated from the long bone of 7-weeks-old ICR male mice Model: M-CSF and RANKL-induced osteoclastogenesis Treatment: 1 μg/mL of T-IIA, T-I, D-T, and C-T for 7 days Negative control: DMSO-treated cells Positive control: N.A.	↓ osteoclastogenesis in all tanshinone isoforms vs. negative control NS for osteoblastogenesis in all tanshinone isoforms vs. negative control
Kwak et al. [31]	Cell: Bone marrow cells from tibia and femur and mouse calvariae from pericranium of 5 weeks old male ICR mice. Model: M-CSF and RANKL-induced osteoclastogenesis Treatment: 10 μg/mL of T-IIA for 4 days Negative control: M-CSF-treated cells Positive control: N.A.	↓ RANKL-mediated osteoclast differentiation vs. negative control ↓ c-Fos and NFATc1 expression induced by RANKL
Kim et al. [33]	Cell: Calvarial osteoblasts from the new bone of ICR mice and bone marrow cells from tibiae of 6–7 weeks old ICR mice (Mouse bone marrow cells and calvarial osteoblast coculture) Model: M-CSF and RANKL-induced osteoclastogenesis Treatment: 2.5–20 μg/mL of T-IIA for 15 min–20 h Negative control: Untreated cells Positive control: N.A.	↓ osteoclast differentiation, osteoclast fusion, actin ring formation and resorption area vs. negative control ↓ osteoclast differentiation-related genes (calcitonin receptor, c-Src, and integrin β3) ↓ activation of ERK, Akt and NF-κB signal transduction pathways
Kwak et al. [32]	Cell: Calvarial osteoblasts and bone marrow cells isolated from the femur and tibias of 5-weeks-old ICR male mice (Mouse bone marrow cells and calvarial osteoblast coculture) Model: TNF-α, IL-1α or LPS-induced osteoclast differentiation Treatment: 5 μg/mL of T-IIA for 7 days Negative control: Untreated cells Positive control: N.A.	↓ osteoclast differentiation vs. negative control ↓ LPS-induced RANKL and OPG expression in osteoblasts vs. negative control ↓ LPS-mediated COX-2 expression and LPS-induced PGE2 in osteoblasts
Animal studies
Zhang, et al. [45]	Animals: 48 male Wistar rats (2 months old) Experimental model: Relapse stage after orthodontic mesial movement of maxillary first molar tooth Treatment: 0.36, 0.72 and 1.44 mg/day of T-IIA for 4 weeks (localised gingival mucosa injection) Negative control: Normal saline injection Positive control: N.A.	↓ recurrence distance and percentage of tooth movement by regulating osteoclast activity with ↑ OPG/osteoclast differentiation factor NS in body weight changes
Yang, et al. [46]	Animals: 40 female Wistar rats (1 month old, 97 ± 3 g Disease model: N.A. Treatment: 22 mg/kg/day of T-IIA for 8 weeks (oral) Negative control: untreated rats Positive control: N.A. Comparative groups: 16.8 mg/kg/day of resveratrol and T-IIA (11 mg/kg/day) + resveratrol (8.4 mg/kg/day) for 8 weeks	↑ serum OCN, whole-body and femoral BMD, maximum femoral load and histomorphometric indices (trabecular width, trabecular separation degree, Tb.N and trabecular area) vs. untreated rats ↓ serum TRAP vs. untreated rats NS in body weight
Wang et al. [36]	Animals: 40 male C57BL/J6 mice (3 months old) Disease model: Open osteotomy at femur diaphysis Treatment: 10, 20 and 30 mg/kg/day of T-IIA for 4 weeks (oral) Negative control: Mice with open osteotomy received with methanol Normal control: No treatment or surgery Positive control: N.A.	↑ callus area, callus intensity, BV1/TV, TMD and BMD vs. negative control
Yao et al. [41]	Animals: 24 male C57BL/J6 mice (2 months old) Disease model: PE particle-induced calvarial osteolysis Treatment: 1 and 2 μg/g/day of T-IIA for 21 days (periosteum injection) Negative control: PE-induced mice treated with PBS Normal control: Sham treated with PBS Positive control: N.A.	↓ number of pits and percentage of porosity of the skull vs. negative control ↑ BV/TV and BMD vs. negative control ↓ TRAP (+) osteoclasts, RANKL, OSCAR, CTX-1 and ↑ OPG vs. negative control
Kwak et al. [32]	Animals: ICR mice Disease model: LPS-induced bone loss Treatment: 5 μg/g of T-IIA (i.p.) on the day before LPS induction, the day 1, 3, 5 and 7 after LPS induction Negative control: Mice with LPS-induced bone loss Normal control: Mice with PBS treatment Positive control: N.A.	↓ LPS-induced bone loss vs. the negative control ↓ LPS-induced osteoclast formation and loss of cancellous bone vs. the negative control
Zhu et al. [37]	Animals: 64 female Wnt1sw/sw mice with osteoporosis (2.5 months old, 32–40 g) Disease model: Spontaneous WNT1 mutation for osteogenesis imperfecta Treatment: 10 mg/kg of T-IIA (i.p.)/day for 6 weeks (i.p.) Negative control: Osteoporotic mice with PBS injection Normal control: N.A. Positive control: 10 mg/kg alendronate for 6 weeks	↑ stiffness, ultimate strength, elastic modulus, proline/amide 1, phosphate/amide 1 and phosphate/proline vs. negative control and greater potency than positive control
Zhang et al. [42]	Animals: 40 male C57BL/J6 mice (2 months old) Disease model: STZ-induced diabetic osteoporosis Treatment: 10 and 30 mg/kg of T-IIA 3 times per week for 8 weeks (i.p.) Negative control: Diabetic mice with corn oil (vehicle) Normal control: Non-diabetic mice with corn oil Positive control: 2 mg/kg aliskiren, 3 times a week for 8 weeks	↑ bone mass of trabecular bone vs. negative control ↑ BMD/TV, BV/TV, BA/TA, Conn.D and ↓ SMI vs. negative control
Wang et al. [40]	Animals: 32 female Sprague Dawley rats (3 months old) Disease model: OVX-induced osteoporosis Treatment: 10 mg/kg/day of T-IIA for 2 weeks (i.v.) Negative control: untreated OVX rats Normal control: Sham-operated rats Positive control: N.A.	↑ bone volume, trabecular number, trabecular thickness and ↓ trabecular separation vs. negative control
Cheng et al. [39]	Animals: 18 female C57BL/6 mice (2 months old) Disease model: OVX-induced osteoporosis Treatment: 10 mg/kg T-IIA for 6 weeks (i.p.) Negative control: OVX with normal saline Normal control: Sham with normal saline Positive control: N.A.	↓ trabecular bone loss vs. negative control ↑ BS/TV, BV/TV, BMD, Tb.N and ↓ Tb.Pf vs. negative control
Panwar et al. [30]	Animals: 29 female C57BL/6 mice (8 months old, 25g) Disease model: OVX-induced osteoporosis Treatment: 40 mg/kg/day of T06 for 3 months (oral) Negative control: OVX with an unknown vehicle Normal control: Sham with an unknown vehicle Positive control: N.A.	↑ N.Ob/B.Pm ↓ Tb.Sp and ↑ BV/TV, Tb.N vs. negative control ↑ ALP-positive osteoblasts vs. negative control. ↑ P1NP concentration vs. negative control NS for CatK-positive osteoclast numbers/bone surface and osteoclast numbers/bone perimeter
Cui et al. [43]	Animals: 32 female Sprague-Dawley (4 months old) Disease model: OVX-induced osteoporosis Treatment: 200 mg/kg/day of total tanshinone (5 mg/kg/day of T-IIA and 16 mg/kg/day of C-T) for 10 weeks (oral) Negative control: OVX with oral deionised water Normal control: Sham with oral deionised water Positive control: 30 μg/kg/day of 17α-37 for 10 weeks	↑ BV/TV, Tb.Th and ↓ OCS/BS in LV4 vs. negative control ↑ BV/TV, Tb.Th and ↓ OCS/BS, MAR and BFR/ BV vs. negative control
Zhou et al. [44]	Animals: 30 male and 30 female KM mice (3 months old, 30 ± 5 g) Disease model: retinoic acid-induced osteoporosis Treatment: 40–160 mg/kg/day of tanshinone for 14 days (oral) Negative control: Untreated osteoporotic mice Normal control: sham-operated normal mice Positive control: 3 mg/kg/day of vitamin D3 for 14 days	↑ cortical bone thickness and Tb.N with active epiphysis ↑ serum estradiol levels with ↓ serum phosphate, ALP and TRAP levels vs. negative control NS for serum calcium and OCN

Abbreviations: ↑, increase or upregulate; ↓, decrease or downregulate; % ES/BS, percentage eroded surface; % trench surface/BS, percentage trench surface per bone surface; Akt, protein kinase B; ALP, alkaline phosphatase; Apaf-1, apoptotic protease-activating factor 1; BA/TA, bone area fraction; Bcl-2, B-cell lymphoma 2; BFR/BV, bone formation rate per unit of bone volume; BMD, bone mineral density; BMD/TV, BMD over total volume; BMMCs, bone marrow mononuclear cells; BM-MSCs, bone marrow mesenchymal stem cells; BMP, Bone morphogenetic proteins; BS/TV, Bone surface area/total value; BV/TV, bone volume/total volume; BV1/TV, low-density bone volume/Callus total volume; Conn.D, connectivity density; COX2, cyclooxygenase-2; c-Src, proto-oncogene tyrosine-protein kinase Src, C-T, cryptotanshinone; CTR, calcitonin receptor; CTX-1, cross linked C-telopeptide of type I collagen; D-T, 15,16-dihydrotanshinone; ERK, extracellular signal-related kinase; hPDLSC, human periodontal ligament stem cells; IAP, inhibitor of apoptosis protein, IL-1 α, interleukin 1 alpha; IκBα, inhibitor of NF-κB α; IKK-β, IκB kinase-β; iNOS, inducible nitric oxide synthase; i.p., intraperitoneal injection; i.v., intravenous injection; JNK, c-Jun N-terminal kinase; LPS, lipopolysaccharide; LV4, fourth lumbar vertebrae; MAPK, mitogen-activated protein kinase; MAR, mineral apposition rate; M-CSF, macrophage colony-stimulating factor; MMP-9, matrix metalloproteinase 9; N.A., not available; NFATc1, nuclear factor of activated T-cells cytoplasmic 1; NF-κB, nuclear factor kappa B; Nox4, NADPH oxidase 4; NS, not significant; OCN, osteocalcin; OCS/BS, percent osteoclast surface; OPG, osteoprotegerin; OPN, osteopontin; OSCAR, osteoclast associated receptor; Osx, osterix; OVX, ovariectomy; P1NP, procollagen-1 N-terminal peptide; PBS, phosphate buffer saline; PE, polyethylene; PGE2, prostaglandin E2; pNF-κB, phosphorylated NF-κB; PTM, proximal tibial metaphysis; RANKL, receptor activator of nuclear factor κB ligand; RNS, reactive nitrogen species; ROS, reactive oxygen species; Runx2, runt-related transcription factor 2; SMI, structure model index; SOD, superoxide dismutase; STZ, streptozocin; T-I, tanshinone I; T-IIA, tanshinone II A; T06, tanshinone IIA sulfonic sodium; TBARS, thiobarbituric acid reactive substances; Tb.N, trabecular number; Tb.pf, trabecular pattern factor; Tb.Th., trabecular thickness; TMD, tissue mineral density; TNF-α, tumor necrosis factor alpha; TRAP, Tartrate-resistant acid phosphatase; TRAF-1 and 6, Tumor necrosis factor receptor associated factor 1 and 6.