Background
Bone metabolic diseases are caused by a disruption in the metabolic balance between osteoblasts and osteoclasts [
1]. The diagnosis and prognosis of bone metabolism disorders primarily rely on the assessment and analysis of bone mineral density using dual-energy X-ray absorptiometry (DEXA) [
2,
3]. However, These tools present hysteresis in their evaluation of bone metabolism. The current evaluation methods for bone metabolism and its prognosis fall significantly short of meeting practical needs in terms of their evaluation and predictive capabilities. Indeed, these tools underestimate the risk of bone metabolic disorders and associated fractures and account for only a fraction of the total fracture risk [
4‐
6]. Therefore, observing and predicting bone metabolism more accurately is necessary to manage fractures and other diseases associated with bone metabolic disorders [
7].
Since bone metabolic disorders are strongly associated with the balance between osteoblasts and osteoclasts, biomarkers reflecting the activity of these cells can potentially reflect the current level of bone metabolism [
8]. For example, the N-terminal propeptide of type I procollagen (P1NP) detects bone formation states [
9,
10], while the β-collagen degradation product (β-CTX) reflect bone resorption states [
11,
12]. These bone turnover markers have therefore been utilized for the diagnosis and management of various diseases, such as osteoporosis, Paget's disease, fibrous dysplasia, hypophosphatemia, primary hyperparathyroidism, and chronic kidney disease-mineral bone disease [
13], as well as in assessing and predicting treatment response, bone mass, and fracture risk alongside other indicators of bone metabolism [
8,
14]. However, differences in biological variability of individuals affect the accuracy of bone turnover markers [
15,
16], and, in disease state, such as dermatological disorders and hepatic fibrosisbone, turnover markers result in diminished assessment efficiency [
17,
18]. This ultimately affects the diagnostic and predictive ability of these bone turnover markers.
Therefore, it is imperative to explore new indicators capable of comprehensively reflecting bone metabolism, such as within the realm of osteoimmunology, to enhance the prediction and evaluation of bone metabolic disorders, as well as the effective management of adverse outcomes in patients with such conditions. Osteoimmunology focuses on the interaction between the immune system and bones [
19] and explores bone metabolism from the intersection of these two fields.
Studies have found that the ratio of lymphocytes to neutrophils, monocytes, and platelets can indicate osteoporosis [
20‐
22]. Circulating T lymphocyte subsets can predict bone morphology [
23]. We have previously demonstrated a significant correlation between total T lymphocytes, CD8
+ T lymphocytes, and bone mineral density (BMD) [
24]. The findings suggest that lymphocytes and their secreted cytokines may serve as potential biomarkers for predicting and evaluating bone metabolism and its prognosis. Therefore, in this study, we analysed lymphocyte subsets, cytokines, and bone metabolism indicators (P1NP and β-CTX) in inpatients at our medical centre, and determined the correlations among these indicators to identify new potential indicators for the evaluation and prediction of bone metabolism.
Methods
Study design and ethics statement
This retrospective cross-sectional study was conducted at the Department of Geriatrics and Institute of Respiratory and Critical Medicine, Eighth Medical Center, PLA General Hospital following the ethical standards of the Declaration of Helsinki. The Ethics Committee of the Eighth Medical Center of PLA General Hospital approved this study (approval number: 309202109091005). All participants were aware of the study objectives and content and provided signed informed consent.
Study participants and their basic characteristics
This study included 162 patients hospitalized in the Department of Geriatrics between December 23, 2021, and May 18, 2022. The inclusion criteria were as follows: patients over the age of 40 years with disorders of bone metabolism and immune system, most of which occurred above the age of 40. The exclusion criteria were as follows: (1) primary immune dysfunctional disease, severe infection, acquired immune deficiency, or allogeneic blood transfusion within 3 months; (2) use of immunomodulatory agents and hormone treatment within 1 year; (3) malignant tumours, major organ failure, or organ transplantation; (4) history of orthopaedic diseases characterized by bone damage; (5) other diseases that cause abnormal bone metabolism and immune regulation. General information about each participant, including the age, sex, height, weight, and body mass index (BMI), was collected. The medical history of each participant was recorded. Informed consent was obtained from the participants for the collection of these data.
Detection of lymphocyte subsets
At admission, a 2-mL blood sample was taken from all participants in a fasting state. The samples were analysed using gating lymphocytes by CD45 staining versus side scatter (SSC), including total T lymphocytes (CD3
+ CD45
+), immature T lymphocytes (CD4
+CD8
+), suppressor/cytotoxic T lymphocytes (CD3
+CD8
+), helper/inducer T lymphocytes (CD3
+CD4
+), B lymphocytes (CD3
−CD19
+), and natural killer (NK) cells (CD3
−CD16
+CD56
+) [
25,
26].
First, 8 μL mixed antibodies (BD Biosciences, San Jose, CA, USA, No. 66299) were added to the bottom of absolute counter tubes (BD Multitest 6-Color TBNK Reagent, No. 662967), including CD45 (APC-Cy5.5), CD3 (FITC-A), CD4 (PE-Cy7-A), CD8 (APC-Cy7-A), CD19 (APC-A), and CD16CD56 (PE-A). Next, 50 μL of the blood sample was added to the bottom of the absolute counter tubes using a reverse pipette before vortex mixing. The mixture was incubated at room temperature (25 °C ± 1 °C) in the dark for 15 min. Subsequently, 450 μL of haemolysin (BD Biosciences, No. 349202) (FACS Lysing solution:distilled water = 1:10) was added to each tube, mixed well, and incubated at room temperature in the dark for 15 min. Finally, the absolute counts of the lymphocyte subsets were analysed and obtained using the FACSCanto Plus flow cytometer (BD Biosciences), and the ratio of each cell was calculated.
Detection of cytokines
The blood samples were centrifuged (1,000 × g for 10 min) to collect plasma within 4 h after collection and stored at –20 °C until use. The levels of interferon-gamma (IFN-γ), tumour necrosis factor-alpha (TNF-α), IFN-α, interleukin-1 beta (IL-1β), IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, and IL-17 were determined using the 12 cytokines detection kit (multiple microglobulin immunofluorescence luminescence, China Qingdao Riskell Biotechnology Co., LTD.).
First, 25 μL of plasma, experimental buffer, capture microsphere antibody (RAISECARE, Qingdao, China, No. R701002), and detection antibody (RAISECARE, No. R701002) were sequentially added to the sample tube and incubated at room temperature while shaking for 2 h (400–500 rpm) in the dark. Subsequently, 25 μL of Streptavidin Phycoerythrin (SA-PE) (RAISECARE, No. R701002) was added to the tube and incubated at room temperature while shaking for 30 min (400–500 rpm) in the dark. Then, 1 mL washing buffer was added to the tube and vortexed at 1,500 r for 5 min. Next, the supernatant liquid was slowly poured out, and 70 μL washing buffer (RAISECARE, No. R701002) was added. Finally, the levels of cytokines were obtained using the FACSCanto Plus flow cytometer (BD Biosciences) and the test results were acquired using LEGENDplex Data Analysis software (v8.0).
Detection of bone turnover markers
The plasma samples for bone turnover marker assays were prepared as for cytokine assays except that they were stored at –80 °C until use. The P1NP level was detected using a P1NP detection kit (Roche Diagnostics, Switzerland, No. 03141071–190), while the β-CTX content was detected using a serum-CTX detection kit (Roche Diagnostics, No. 11972308–122). Both assays are chemiluminescence immunoassays. According to the manufacturer's instructions, the reaction solution and plasma samples were aspirated into the measurement cell of a fully automatic electrochemiluminescence analyser (Cobas e601, Roche Diagnostics), and the detection results of P1NP and β-CTX were automatically obtained. In addition, the assay performance was verified using the manufacturer's control samples before the assay, according to the manufacturer's instructions.
As the study population comprised inpatients in a single centre in China, the reference range of bone metabolism indicators was implemented following the "Expert Consensus on the Clinical Application of Biochemical Indicators of Bone Metabolism (2019)" promulgated in China [
27]. The PINP reference range was 31.7–70.7 ng/mL for females, while the average was 21–78 ng/mL. The bone formation states were altered when P1NP exceeded the reference range. The reference range of β-CTX detection was as follows: (1) the mean value for premenopausal females was 0.299 ng/mL; (2) mean value for postmenopausal females was 0.556 ng/mL; (3) mean value in males aged 30–50 years was 0.3 ng/mL; (4) mean value for males aged 50–70 years was 0.304 ng/mL; (5) mean value for males aged > 70 years was 0.394 ng/mL. The bone resorption states were altered when the β-CTX level exceeded the 95% confidence interval.
Statistical analysis
GraphPad Prism 9 software (San Diego, CA, USA) was used to process, analyse, and visualize the data, and P < 0.05 indicated a statistical difference. The numbers represent enumeration data; continuous variables not normally distributed are represented as quartiles [50% (25–75%)], whereas those normally distributed as mean ± standard deviation. Furthermore, normally distributed continuous variables were analysed using the analysis of variance, whereas abnormally distributed data were analysed using Kruskal–Wallis test.
The effects of lymphocyte subsets and cytokines on bone metabolism in participants with different bone metabolism profiles were analysed using the principal component analysis (PCA). The principal components with eigenvalues > 1 were selected. Subsequently, the eigenvalues were compared with the mean of the corresponding principal components. Moreover, the correlation between lymphocyte subsets and cytokines in patients with bone metabolic disorders was analysed, and the differences in different bone metabolism conditions were explored.
Additionally, the partial receiver operating characteristic (ROC) package in Rstudio3.6.4 was used to evaluate the overall efficacy of differentially expressed lymphocytes and cytokines in distinguishing bone metabolic status. The ROC curve reflected the diagnostic ability of tested cells and factors. The area under the curve (AUC) was > 0.5, indicating that the test item had diagnostic ability. Statistical significance was set at P < 0.05. The ggplot2 package in Rstudio3.6.4 was used for visualization.
Discussion
This study analysed the relationship between bone metabolism and the immune system and showed that the absolute count of total T lymphocytes and percentage of CD4
+ T lymphocytes and the levels of IL-17 and IL-2 are one of the principal components mediating bone formation disorders. This aligns with previous findings showing the regulatory role of T cells and CD4
+ T lymphocytes in osteoblast activity through the secretion of various cytokines (including TNF-α and RANKL) [
28,
29]. Similarly, IL-2 and IL-17 can also inhibit osteogenic activity by regulating RANKL or inducing inflammatory activity [
30,
31]. These factors collectively mediate the activity of osteoblasts. Interestingly, IL-17 can also directly induce the formation of osteoclasts or indirectly promote osteoclast activity by enhancing the expression of pro-resorptive factors [
32,
33]. This demonstrates the complexity and heterogeneity of the immune system's regulation of bone metabolism. Figures
3 and
4 demonstrate the heterogeneity in the correlation of immune cells with cytokines in bone formation and bone resorption disorder. Additionally, we observed that the total T lymphocytes, IL-12p70, and IL-2 were the primary components mediating bone resorption disorders, and have been reported to mediate bone loss or osteoclast activity [
34‐
36]. In summary, the emergence of these complex regulatory networks provides new possibilities for exploring and regulating bone metabolism.
To identify potential biomarkers for assessing bone metabolism, we explored the differentially expressed lymphocytes and cytokines and showed an increase in B lymphocytes and IL-12p70 that led to a decrease in P1NP levels. This was consistent with the conclusion of previous studies that B lymphocytes inhibit osteogenic differentiation [
37]. We also observed a synergistic increase in IL-8 and P1NP levels. It is possible that IL-8 activated T lymphocytes, which further led to bone loss and increased compensatory bone formation [
38]. However, previous studies have only reported that IL-12p70 could stimulate Th1 and Th2 cells to inhibit the formation of osteoclasts by secreting the IFN-γ, without discussing its relationship with decreased bone formation [
39]. In addition, we observed that the increase of T and B lymphocytes led to the reduction in the β-CTX levels. T and B lymphocytes are involved in the induction of osteoclastic activity in various inflammatory and immunological diseases [
40‐
42]. Notably, in a relatively stable state, T lymphocytes inhibit osteoclast formation [
41,
43,
44]; similarly, B lymphocytes can inhibit osteoclast formation by secreting TGF-β or limiting bone resorption under certain pathological conditions [
45,
46]. In this study, we found that a decrease in the level of IL-6 led to the reduction of the β-CTX levels. Reduced IL-6 levels may reduce osteoclast formation and β-CTX levels. The positive correlation between IL-6 and osteoclast activity has been previously confirmed [
47]. Surprisingly, the increase and decrease in the percentage of NK cells were accompanied by an increase in the β-CTX levels; however, previous studies only reported that NK cells negatively affect osteoclasts [
48]. Therefore, we intend to further expand the sample size in future studies to clarify this unexpected finding.
Finally, we analysed the diagnostic and evaluation capabilities of these differentially expressed lymphocytes and cytokines using the ROC curve.
The ROC curve is used in bone metabolism research to determine the diagnostic ability of tested factors. For example, chitinase 3-like protein is used to determine the occurrence and development of osteoporosis [
49]. The IL-6 is used to assess the risk of occurrence of fragility fractures [
50]. In this study, the absolute count of B lymphocytes and level of IL-12p70 could distinguish a reduction in the states of bone resorption, while IL-8 levels could determine the hyper states of bone resorption. Additionally, the absolute count of total T lymphocytes and the absolute count and percentage of B lymphocytes can distinguish a reduction in the o states of bone formation. Simultaneously, the percentage of NK cells can predict both the hyperactivity of osteoclasts and a decrease in the states of bone formation. These lymphocytes and cytokines are potential bone metabolic biomarkers to assess the bone metabolic status of patients and their response to therapy. Overall, this study explored the relationship among bone metabolism, lymphocyte subsets, and cytokines at the clinical level, providing new potential perspectives and tools for understanding and evaluating bone metabolism.
This study has some limitations. First, this study was conducted in a single centre, which may affect the generalizability of the results. Therefore, we intend to include larger samples and conduct multicentre studies in future studies. Second, this was a retrospective analysis, and the outcome may be affected by non-observational factors; therefore, we intend to design prospective research based on the results of this study to assess the predictive and evaluation capabilities of the identified indicators. Third, the bone metabolism index is subject to several confounding factors, and it can only reflect bone metabolism rather than the bone metabolism outcome. Therefore, in subsequent studies, we intend to include multiple evaluation indicators of bone metabolism activity, such as BMD. Finally, this study only involved participants from northern China. Bone metabolism status differs among populations worldwide [
51].
Conclusions
This study revealed a link between bone metabolism and the immune system and showed that the decreased P1NP levels were associated with increased absolute count of B lymphocytes and IL-12p70 levels. The increased P1NP levels were associated with increased IL-8 levels. Similarly, the reduced β-CTX levels were associated with decreased IL-6 levels, increased absolute count of T and B lymphocytes, increased percentage of B lymphocytes, and reduced percentage of NK cells. In addition, the absolute count of T lymphocytes, level of IL-12p70, level of IL-8, absolute count and percentage of B lymphocytes, and percentage of NK cells may be used to evaluate different bone metabolic s tatus. These findings suggest potential biomarkers for bone metabolic disease prevention and diagnosis.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.