Aging and menopause reprogram osteoclast precursors for aggressive bone resorption

Bone Research ARTICLE www.nature.com/boneres OPEN Aging and menopause reprogram osteoclast precursors for aggressive bone resorption Anaïs Marie Julie Møller1,2,3, Jean-Marie Delaissé1,2,4,5,6, Jacob Bastholm Olesen1,4, Jonna Skov Madsen Troels Bechmann2,8, Silvia Regina Rogatto2,7 and Kent Søe 1,2,4,5,6,9 2,3 , Luisa Matos Canto7, 1234567890();,: Women gradually lose bone from the age of ~35 years, but around menopause, the rate of bone loss escalates due to increasing bone resorption and decreasing bone formation levels, rendering these individuals more prone to developing osteoporosis. The increased osteoclast activity has been linked to a reduced estrogen level and other hormonal changes. However, it is unclear whether intrinsic changes in osteoclast precursors around menopause can also explain the increased osteoclast activity. Therefore, we set up a protocol in which CD14+ blood monocytes were isolated from 49 female donors (40–66 years old). Cells were differentiated into osteoclasts, and data on differentiation and resorption activity were collected. Using multiple linear regression analyses combining in vitro and in vivo data, we found the following: (1) age and menopausal status correlate with aggressive osteoclastic bone resorption in vitro; (2) the type I procollagen N-terminal propeptide level in vivo inversely correlates with osteoclast resorption activity in vitro; (3) the protein level of mature cathepsin K in osteoclasts in vitro increases with age and menopause; and (4) the promoter of the gene encoding the dendritic cell-specific transmembrane protein is less methylated with age. We conclude that monocytes are “reprogrammed” in vivo, allowing them to “remember” age, the menopausal status, and the bone formation status in vitro, resulting in more aggressive osteoclasts. Our discovery suggests that this may be mediated through DNA methylation. We suggest that this may have clinical implications and could contribute to understanding individual differences in age- and menopause-induced bone loss. Bone Research (2020)8:27 ; https://doi.org/10.1038/s41413-020-0102-7 INTRODUCTION Bone is continuously being turned over and repaired throughout life. This occurs through a process called bone remodeling, consisting of a tight coordination and balance between bone resorption and bone formation.1,2 In this process, bone-resorbing osteoclasts (OCs) and bone-forming osteoblasts (OBs) play a central role. To maintain bone mass throughout adulthood, OBs must replace the precise amount of bone removed by OCs. This link between them, necessary to balance out their activities, is termed “coupling”.3,4 However, with age, bone resorption slowly begins to exceed new bone formation during remodeling. Women gradually lose bone mass from the age of ~35, but at menopause, the bone resorption rate increases further, the bone formation rate decreases, and consequently, bone loss is accelerated, making women more prone to osteoporosis.5–7 The reason for the increase in OC activity has been studied extensively, and a link to a reduced level of estrogen and possibly to an increased level of follicle stimulating hormone has been shown.8–12 Bone turnover can be detected using bone biomarkers, such as serum C-terminal telopeptide of type I collagen (CTX) for bone resorption and serum procollagen type I N propeptide (PINP) for bone formation.13 In general, the CTX level is significantly elevated in women with osteoporosis (postmenopausal) compared with nonosteoporotic postmenopausal women, while the PINP level is less elevated and may even be reduced.14 Thus, these bone biomarkers can reveal the loss of coupling between bone resorption and formation that eventually leads to osteoporosis. Osteoporosis dramatically affects human health as a major cause of fracture worldwide and is strongly associated with both premature death and morbidity, the latter especially in terms of pain and disability.7 Osteoporosis is a very common condition and is associated with a substantial healthcare burden.7 Previous studies have shown that estrogen affects OCs in both mice and humans. The decrease in estrogen following ovariectomy/ menopause triggers the increased expression of macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor kappa-Β ligand (RANKL) in OB-lineage cells.15–17 Since M-CSF and RANKL are both key cytokines driving osteoclastogenesis, a drop in estrogen will indirectly boost the formation of OCs. Estrogen also directly affects OCs, e.g., by reducing the expression of cathepsin K (CatK), a key factor in organic bone matrix degradation.18–20 These effects are thought to occur due to the presence or 1 Clinical Cell Biology, Lillebaelt Hospital, University Hospital of Southern Denmark, 7100 Vejle, Denmark; 2Department of Regional Health Research, University of Southern Denmark, 5230 Odense M, Denmark; 3Department of Clinical Biochemistry and Immunology, Lillebaelt Hospital, University Hospital of Southern Denmark, 7100 Vejle, Denmark; 4 Clinical Cell Biology, Department of Pathology, Odense University Hospital, 5000 Odense C, Denmark; 5Department of Clinical Research, University of Southern Denmark, 5230 Odense M, Denmark; 6Department of Molecular Medicine, University of Southern Denmark, 5230 Odense M, Denmark; 7Department of Clinical Genetics, Lillebaelt Hospital, University Hospital of Southern Denmark, 7100 Vejle, Denmark; 8Department of Oncology, Lillebaelt Hospital, University Hospital of Southern Denmark, 7100 Vejle, Denmark and 9 OPEN, Odense Patient data Explorative Network, Odense University Hospital, 5000 Odense C, Denmark Correspondence: Anaïs Marie Julie Møller () or Kent Søe () Received: 16 January 2020 Revised: 6 March 2020 Accepted: 7 April 2020 © The Author(s) 2020 Aging reprograms pre-OCs for aggressive bone erosion AMJ Møller et al. 2 absence of estrogen in the microenvironment. However, in recent years, different observations have indicated that factors such as sex, the menopausal status, and age affect the properties of OCs themselves, in a manner unrelated to the microenvironment and presence or absence of ligands/receptors. This has mainly been investigated using OCs differentiated from peripheral blood mononuclear cells (PBMCs)21–23 and bone marrow-derived OCs,23 which are both widely accepted models for the generation and characterization of human OCs in vitro. Several studies have indicated that in vitro-generated OCs act and/or respond in a sexdependent manner in humans.24–28 In addition, sex-dependent differences in the resorption mode of OCs in vitro have also been suggested.29 Aging and/or menopause have also been found to affect OC formation both in vivo and in vitro.28 First, the OC progenitor pool has been reported to increase with advancing age in humans30,31 and mice.16,32 However, in mice, this increase was only observed with stimulation by OB-derived cytokines, such as interleukin-3, granulocyte-macrophage c (...truncated)