Abstract: BACKGROUND: At present, human placental mesenchymal stem cells are rarely used in bone tissue engineering research, and the use of human placental mesenchymal stem cells in the treatment of osteoporotic fractures is likely to open new horizons for their therapeutic methods.  
OBJECTIVE: To investigate the mechanism of human placenta-derived mesenchymal stem cells in the treatment of tree shrews with osteoporotic fracture, and the effect of different drug delivery methods on the efficacy of tree shrews with osteoporotic fracture.
METHODS:  Bilateral ovaries and uterus were removed from female tree shrews to simulate postmenopausal osteoporosis, and they were naturally fed for 180 days. The osteoporotic fracture tree shrew model was established by performing a right femoral fracture. The 24 tree shrews with osteoporotic fracture were randomly divided into four groups. We adjusted the concentration of the human placenta-derived mesenchymal stem cells to 1×109 L-1. In the tail vein injection group, human placenta-derived mesenchymal stem cells were injected with 1 mL in the tail vein. In the tail vein combined with fracture injection group, human placenta-derived mesenchymal stem cells were injected with 0.5 mL in the tail vein and 0.5 mL at the fracture site. In the tail vein combined with intraperitoneal injection group, human placenta-derived mesenchymal stem cells were injected with 0.5 mL in the enterocoelia and 0.5 mL in the tail vein. In the model group, 1 mL physiological saline was injected through tail vein. From the 3rd day after the fracture, tree shrews were injected once a week and three times in a row. Eight weeks after the last treatment, bone mineral density was tested in tree shrews of each group. Three-point bone biomechanics and hematoxylin-eosin staining experiments were performed on the femurs of tree shrews in each group. The expression levels of osteocalcin, estrogen, bone alkaline phosphatase and tartrate resistant acid phosphatase were measured by ELISA in all groups. The mRNA expression levels of bone morphogenetic protein-2, osteoprotegerin and receptor activator of nuclear factor-κB ligand were measured by quantitative real time polymerase chain reaction.  
RESULTS AND CONCLUSION: (1) Eight weeks after the last treatment, compared with the model group, the other three groups showed increased bone mineral density, maximum load, structural stiffness and energy absorption, with tail vein injection group showed the most significant increase. (2) Serum estrogen, bone alkaline phosphatase and osteocalcin expression levels in tail vein injection group, tail vein combined with fracture injection group, and tail vein combined with intraperitoneal injection group increased significantly compared with the model group. Serum tartrate resistant acid phosphatase levels decreased significantly in the tail vein injection group, tail vein combined with fracture injection group, and tail vein combined with intraperitoneal injection group compared with the model group. These serum detection results showed the most obvious improvement in tail vein injection group. (3) Hematoxylin-eosin staining showed significant improvement of pathological changes in tail vein injection group and tail vein combined with fracture injection group. (4) The expression level of osteoprotegerin and bone morphogenetic protein-2 mRNA in tail vein injection group, tail vein combined with fracture injection group, and tail vein combined with intraperitoneal injection group increased to varying degrees compared with model group; the tail vein combined with fracture injection group had the highest level. (5) The expression level of receptor activator of nuclear factor-κB ligand mRNA in tail vein injection group, tail vein combined with fracture injection group, and tail vein combined with intraperitoneal injection group decreased to different degrees compared with model group, with tail vein combined with fracture injection group having the lowest level. (6) It is concluded that human placenta-derived mesenchymal stem cell transplantation can effectively improve the symptom of tree shrews with osteoporotic fracture, increase bone formation indexes such as bone alkaline phosphatase, osteocalcin and osteoprotegerin, decrease bone resorption indexes such as tartrate resistant acid phosphatase and receptor activator of nuclear factor-κB ligand, and improve bone mineral density and bone biomechanics of tree shrews. Tail vein injection group had the best results for systemic treatment and tail vein combined with fracture injection group showed better results for treatment at the fracture site.

Key words: placental mesenchymal stem cell, tree shrew, osteoporotic fracture, efficacy, administration route, animal model