In Vitro Modeling of Diurnal Changes in Bone Metabolism

Abstract

There is evidence that bone health is closely linked to a functioning circadian rhythm. Most of the evidence comes from mice, which may exhibit some species-specific differences from humans due to their nocturnal lifestyle. To address the current lack of human model systems, the present study aimed to develop an in vitro model system that can represent diurnal changes in bone metabolism. The model is based on co-cultured SCP-1 and THP-1 cells that serve as osteoblast and osteoclast precursors, respectively. Diurnal effects were induced by replacing the FCS in the differentiation medium with human serum pools (HSPs) obtained in the morning, noon, or evening. The model system was tested for cell viability, gene expression, and osteoblast and osteoclast function. The replacement of the FCS with the HSPs increased viability and induced expression changes in circadian clock genes in the model system. Resulting alterations in osteoblast and osteoclast function led to a gradual increase in mineral density and stiffness when 3D co-cultures were differentiated in the presence of the HSPs collected in the morning, noon, or evening, respectively. Here, we present for the first time an in vitro model that can present diurnal changes in bone metabolism in the form of a snapshot. With the simple use of HSPs, this model can be used as a platform technique to investigate bone function in various situations, taking into account the time of day.

Keywords: BMAL1; CLOCK; circadian rhythm; in vitro model; osteoblast; osteoclast.

Figure 1

Expression of circadian clock genes in conventional bone cell co-culture is not affected by the time of day. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 2% fetal calf serum. At day 21 of culture, total mRNA was collected in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm) to investigate expression of genes (qRT-PCR) involved in the core loop of circadian rhythm: (A) BMAL1, (B) CLOCK, (C) NPAS2, and their negative feedback regulators (D) CRY1, (E) PER1, and (F) PER2. Gene expression levels were determined using the ΔΔCt method and normalized as z-score. Data are displayed as box plots with individual data points (N = 4, n = 2) and the median. Groups were compared with the non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test when p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 2

Functional testing of the modified medium composition. (A) Schematic overview of the experimental setup. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 2% fetal calf serum (FCS) or 1% respective 2% human serum pool (HSP). For the HSP, equal volumes of human serum from 10 donors (N = 10) were pooled. On day 21 of culture, viability was assessed by (B) mitochondrial activity (resazurin conversion assay) and (C) total protein content (Sulforhodamin B staining). On day 14 of culture, (D) alkaline phosphatase (ALP) activity was measured photometrically as an early osteogenic marker. On day 21 of culture, (E) carbonic anhydrase II (CAII) and (F) tartrate-resistant acidic phosphatase (TRAP5b) activity were measured photometrically as an early and late osteoclast marker, respectively. All data were normalized as z-scores and are displayed as box plots with individual data points (N = 3, n = 3). Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test whenever p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 3

Highest cell proliferation in bone cell co-cultures differentiated in the presence of human serum collected in the evening. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 1% human serum pools (N = 10) obtained in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm). Cell growth was assessed by changes in (A) mitochondrial activity (resazurin conversion assay) and (B) total protein content (Sulforhodamin B staining). Changes over time are displayed as line charts (median ± 95% CI of N = 4, n = 3) and summarized as the area under the curve (AUC). On day 21 of culture, total mRNA was collected in the morning and the expression of genes involved in the proliferation was investigated by qRT-PCR (ΔΔCt) and normalized as a z-score: (C) MIK67, (D) TPX2, and (E) TOP2A. Data are displayed as box plots with individual data points (N = 4, n ≥ 2) and the median. Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test whenever p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 4

Highest osteoblast function in bone cell co-cultures differentiated in the presence of human serum collected in the evening. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 1% human serum pools (N = 10) obtained in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm). On day 21 of culture, total mRNA was collected in the morning, and gene expression levels of (A) RUNX2 and (B) SP7 (Osterix) were investigated by qRT-PCR (ΔΔCt). (C) Alkaline phosphatase (ALP) activity was measured as an early osteogenic function on days 7, 14, and 21 of culture. Formed mineralized matrix was visualized by (D) von Kossa and Alizarin red staining (representative images) and quantified (E) by image analysis (ImageJ 1.47v) or (F) photometrically, respectively. All data are displayed as box plots with individual data points (N = 4, n ≥ 2 or 3) and the median. Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test whenever p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 5

Highest osteoclast function in bone cell co-cultures differentiated in the presence of human serum collected in the morning. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 1% human serum pools (N = 10) obtained in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm). The levels of the macrophage colony-stimulating factor (M-CSF), receptor activator of NF-κB ligand (RANKL), and Osteoprotegerin (OPG) in the serum pools were determined by dot blot (n = 3). (A) Image of signal intensity and (B) its quantification by ImageJ 1.47v analysis (heat map). (C) On day 21 of culture, total mRNA was collected in the morning, and the gene expression levels of the key osteoclastic transcription factor NFATc1 were determined by qRT-PCR (ΔΔCt). After 21 days of differentiation, (D) carbonic anhydrase II (CAII) and (E) tartrate-resistant acidic phosphatase (TRAP5b) activities were measured as early and late osteoclast markers, respectively. All data are displayed as box plots with individual data points (N = 4, n = 2 or 3) and the median. Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test whenever p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 6

Highest scaffold density and stiffness with bone cell co-cultures differentiated in the presence of human serum collected in the evening. Three-dimensional co-cultures of THP-1 and SCP-1 cells were osteogenically differentiated with media containing 1% human serum pools (N = 10) obtained in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm). After 21 days of culture, mineral density of the scaffolds was determined by computer tomography. (A) Three-dimensional reconstructions of the cultured scaffolds representative for the morning, noon, and evening group. (B) The bone mineral density of the cultured scaffolds was quantified using the ImageJ 1.47v software. (C) Scaffold stiffness (E-modulus) was determined using a ZwickiLine Z 2.5TN material testing machine. All data were normalized as z-scores and are displayed as box plots with individual data points (N = 4 or 5, n = 3). Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test whenever p < 0.05. All p-values below 0.1 are displayed in the graphics.

Figure 7

The expression of circadian clock genes in the newly established bone cell co-culture shows changes characteristic for the time of day. THP-1 and SCP-1 cells were co-cultured in osteogenic differentiation medium containing 1% human serum pools (N = 10) obtained in the morning (7–8 am), noon (1–2 pm), or evening (7–8 pm). On day 21 of culture, total mRNA was collected in the morning, and the expression of genes involved in the circadian rhythm core loop (scheme) was investigated by qRT-PCR: (A) BMAL1, (B) CLOCK, (C) NPAS2, and their negative feedback regulators (D) CRY1, (E) PER1, and (F) PER2. Gene expression levels were determined using the delta-delta Ct (ΔΔCt) method and normalized as z-scores. Data are displayed as box plots with individual data points (N = 4, n = 2) and the median. Groups were compared with a non-parametric Kruskal–Wallis test followed by Dunn’s multiple comparison test when p < 0.05. All p-values below 0.1 are displayed in the graphics.