Osteoclasts Link Dysregulated Peripheral Degradation Processes and Accelerated Progression in Alzheimer’s Disease

Abstract

Background:

The amyloid-β (Aβ) enhances the number and activity of blood monocyte-derived osteoclasts (OCs). Individuals with osteoporosis (OP) face an increased risk of developing dementia or Alzheimer’s disease (AD). Despite this association, the contribution of bone-resorbing OCs to the progression of AD pathology remains unclear.

Objective:

Our objective was to investigate the potential impacts of OCs on the development of AD pathology.

Methods:

We conducted targeted analysis of publicly available whole blood transcriptomes from patients with AD to characterize the blood molecular signatures and pathways associated with hyperactive OCs. In addition, we used APP23 transgenic (APP23 TG) AD mouse model to assess the effects of OCs pharmacological blockade on AD pathology and behavior.

Results:

Patients with AD exhibited increased osteoclastogenesis signature in their blood cells, which appears to be positively correlated with dysfunction of peripheral clearance of Aβ mediated by immune cells. Long-term anti-resorptive intervention with Alendronate inhibited OC activity in APP23 mice, leading to improvements in peripheral monocyte Aβ-degrading enzyme expression, Aβ-deposition, and memory decline.

Conclusions:

Our findings suggest that OCs have a disease-promoting role in the development and progression of AD, possibly linked to their modulation of peripheral immunity. These findings guide future research to further elucidate the connection between OP and AD pathogenesis, highlighting the potential benefits of preventing OP in alleviating cognitive burden.

Keywords

Alzheimer’s disease amyloid-β burden monocyte osteoclast osteoporosis

INTRODUCTION

The intricate interplay between different physiological systems plays a pivotal role in the development of diseases. As disease progresses, the risk of fractures [1] and osteoporosis (OP) [2] tends to increase in patients with Alzheimer’s disease (AD). In addition, patients with OP and those with low bone mineral density (BMD) seem to be more susceptible to developing dementia [3 –5]. These observations suggest the existence of a potential biological causal link between bone destruction and AD pathology. Osteoclasts (OCs), a type of multinucleated specialized bone-resorbing cell, serve as key players in both physiological bone remodeling and pathological bone destruction. The activation of OCs can be positively regulated by several physical states, including aging [6], physical inactivity [7, 8], and hormonal withdrawal [9], all of which have been associated with an increased risk of AD and dementia [10 –12]. Furthermore, previous evidence has indicated the accumulation of OCs in young amyloid precursor protein transgenic (APP TG) mice [13]. However, the extent to which OCs contribute to the pathological progression of AD remains unknown.

OCs originate from the monocyte/macrophage hematopoietic lineage [14], with circulating monocytes being the major precursors of mature OCs. Alterations in monocyte homeostasis may reflect osteoclast differentiation, activation, and bone resorption potential under pathological state [15 –17]. On the other hand, peripheral monocytic dysfunction has been observed in patients with AD and is thought to potentially contribute to disease progression [18 –20]. This perspective is supported by evidence of monocytes and monocyte-derived macrophages (MDMs) effectively clearing amyloid-β (Aβ) [21, 22], a process impaired in patients with AD. Thus, peripheral monocytes appear to be centrally located in the bone-brain pathology connection.

As members of the monocytic family, OCs are presumed to possess immunomodulatory capabilities, an aspect that has only recently begun to be understood [23, 24]. OCs can initiate T cell responses in both immunosuppressive and activated ways, depending on the host state. In the context of chronic inflammation, OCs exhibit a remarkable ability to induce the differentiation of TNFα-producing CD4⁺ T cells [25]. Interestingly, both TNFα and activated T cells have been shown to inhibit the degradation of Aβ by MDMs [26]. A hypothesis emerging from this suggests that in the context of AD, excessive osteoclastogenesis activity may disrupt peripheral immune homeostasis, especially in monocytes with Aβ clearance capabilities. In this study, we aimed to unravel the potential involvement of OCs in the development of AD through the expression profiling of blood samples, as well as to preliminarily investigate the beneficial effects of OC-targeted interventions on the AD brain.

MATERIALS AND METHODS

Data accession

Transcriptome profiles of blood from individuals of mild cognitive impairment (MCI)/AD were downloaded from the Gene Expression Omnibus (GEO) with accession numbers: GSE140829, GSE63061, GSE97760, and GSE168813, respectively. GSE56815 series included gene expression profiling of circulating monocytes from 40 extremely high and 40 extremely low BMD subjects were downloaded to identify OP-related gene signatures. Human protein microarray dataset for correlation analysis contains 50 serum samples from individuals of MCI with accession number: GSE74763.

Gene set variation analysis

Gene Set Variation Analysis (GSVA) is a non-parametric, unsupervised method used to estimate the enrichment changes of gene sets between samples in an expression dataset [27]. We used the R Bioconductor opensource package gsva (version 1.40.1) to calculate the enrichment scores for each sample in the gene set. GSVA was run on each dataset separately. A minimum of 5 genes was required for each signature.

GSVA gene sets

To evaluate the osteoclastogenic potential in blood of patients with AD and healthy controls, we got OC differentiation-related gene sets (systematic name: M13018, M24618) from Molecular Signature Database (MSigDB, http://www.gsea-msigdb.org/gsea/msigdb/collections.jsp). Gene set (systematic name: M24957) associated with Aβ clearance disorders were also obtained for subsequent analysis. Pathway enrichment scores were subsequently calculated.

Identification of differentially expressed genes

Differentially expressed gene (DEG) analysis was performed using Limma for public array datasets. For DEGs between low BMD and high BMD circulating monocytes, up- and downregulated DEGs were identified by a fold change≥1.5 or≤0.5 and a false discovery rate adjusted p value of < 0.05. For identification of DEGs in AD blood with high and low OC differentiation gene signature, transcripts with p value < 0.05 were considered as significantly changed. The Gene Ontology (GO) enrichment analyses were performed using Metascape [28].

Gene Set Enrichment Analysis

Gene Set Enrichment Analysis (GSEA) (http://software.broadinstitute.org/gsea/index.jsp) was used to determine whether a priori-defined gene sets would show statistically significant differences in expression between groups. AD blood samples were divided into two groups according to the GSVA score of OC differentiation gene signature, and the collection of c7.all.v7.4.symbols.gmt gene sets were downloaded from MSigDB to evaluate relevant pathways and molecular mechanisms.

Mice

APP23 transgenic mice (B6.Cg-Tg(Thy1-APP)3Somm/J) and non-transgenic control mice were maintained at Children’s Hospital of Chongqing Medical University Animal Care Centre. All mice were maintained under SPF conditions on a 12-h light-dark cycle, and provided food and water ad libitum. The genotype of the mice was confirmed by PCR using DNA from tail tissues. Both female and male mice were used. All animal experiments were conducted in accordance with the Chongqing Science and Technology Commission guidelines and approved by the Chongqing Medical University Animal Care Committee.

Alendronate treatment

4-month-old APP23 mice were treated for 10 months with 0.1 mg/kg/week alendronate (ALN) sodium trihydrate (Sigma-Aldrich, St. Louis, MO, USA) or the equivalent volume of vehicle by subcutaneous injection. To assess the effect of ALN treatment on OCs, mice were euthanized and blood was collected transcardially for determination of serum C-telopeptide of type I collagen (CTX-1) levels using an ELISA kit.

Monocyte isolation

For splenic monocyte preparation, ex vivo spleens were pressed through a 70-μm cell strainer to obtain a suspension containing monocytes. The cell pellets were resuspended in 3 mL of 45% Percoll and overlaid on top of 4.5 mL of 65% Percoll. Gradient centrifugation at 500 g for 20 min at 4°C, the mononuclear cells at the interface were collected and washed with PBS. To obtain monocytes, mononuclear cell suspension was purified using the Mouse Monocyte Isolation Kit (Biolegend) following the manufacturer’s protocol.

Western blotting

Extraction of total protein of cells or tissue was performed as we described previously [29]. 30μg of total protein was boiled with 4×sample buffer at 95°C for 5 min. Followed by, the samples were separated on 10% SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad). The membranes were then incubated with 5% fat-free milk in TBS contains 0.1% Tween-20 for 1 h at room temperature to block non-specific binding site. The target proteins were immunoblotted with primary antibody overnight at 4°C to neprilysin (NEP, 1 : 2000), insulin degrading enzyme (IDE, 1 : 2000), IL-1β (1 : 2000), IL-10 (1 : 2000), glial fibrillary acidic protein (GFAP, 1 : 1000), ionized calcium binding adaptor molecule-1 (Iba-1, 1 : 1000), β-actin (1 : 2000). On the next day, the membranes were washed and incubated with anti-mouse HRP or anti-rabbit HRP secondary antibody (1 : 4000, 2 h at room temperature). The protein band was visualized in the Bio-Rad Imager. Band intensities were quantified using Fiji (NIH).

Immunohistochemistry staining

After behavioral testing, mice were euthanized and half of their brains were immediately frozen for protein extraction, while the other half was fixed in freshly prepared 4% paraformaldehyde (PFA) in 0.1 M phosphate-buffered saline (PBS, pH 7.4) for 24 h. The fixed brains were then dehydrated in 30% sucrose until they sank to the bottom, followed by sectioning into 30μm thick coronal slices using a Leica instrument. The slices were incubated in 3% H₂O₂ for 30 min to remove residual peroxidase activity. Subsequently, the sections were blocked with 10% BSA and incubated overnight at 4°C with mouse monoclonal antibody 4G8 (1 : 500). Sections were mounted onto slides, and plaques were visualized by the ABC and DAB method and counted by microscopy at ×40 magnification. The mean plaque count per slice was recorded for each mouse as described previously [29].

Contextual and cued fear conditioning

All animals were habituated to the operant environment for 3 d before fear conditioning began. Mice were placed into a sound-attenuating shock chamber with white light illumination (context A). Mice were conditioned with 3 presentations (at an average interval of 60s) of white noise acoustic stimuli (cues, continuous 30s) that were co-terminated with foot shocks (0.5 mA, 1 s). After the last tone-shock pairing, mice remained in the shock chamber for 1 min before returning to their home cages. After 24 h, the mice returned to the same environment (context A) for 5 min to assess contextual fear learning. After another 24 h, the mice were placed in a new environment (context B) and allowed to explore freely for 2 min, followed by a single presentation of white noise to assess cued fear learning. Freezing behaviors (no movement other than breathing) were manually counted frame by frame by colleagues blinded to experimental groups. Freezing (%) was calculated by dividing the freezing time by the context or cue duration.

Statistics

Data are mean±SEM. Statistical analyses were performed with GraphPad Prism (version 9.2.0) using an unpaired or paired two-tailed Student’ s t-test for two groups. ANOVA was used for more than two groups followed by the Holm-Šidák post hoc test. p < 0.05 were considered statistically significant.

RESULTS

An enriched osteoclastogenesis signature in AD blood

We carried out a transcriptomic analysis of two large public microarray blood datasets from both patients with AD and healthy controls (Table 1). To examine osteoclastogenesis signaling on an individual patient level, we carried out GSVA using the informative gene signatures related to monocyte-OC differentiation. Signature for positive regulation of OC differentiation (SPROD) was enriched in patients with AD but not MCI compared to healthy blood (Fig. 1A). The GSVA score of SPROD was positively correlated with individual stage from health to disease (Fig. 1A). Although similar alterations were not identified in another independent dataset (The AddNeuroMed Cohort), we found that SPROD enrichment scores in AD blood was upregulated compared to MCI, but not compared to healthy blood (Fig. 1B). Additionally, a multinucleation signature was significantly enriched in blood of these MCI individuals from the AddNeuroMed Cohort (Fig. 1C). The multinuclear OC differentiation signature comprises a smaller gene set than SPROD and is closely associated with monocyte specialization into multinucleated cells [30]. We further investigated two additional small-sample blood datasets (Table 1). Compared to healthy individuals, patients with AD exhibited upregulation of SPROD in both datasets (Fig. 1E).

Fig. 1

AD blood is characterized by enrichment of osteoclast differentiation signatures. A–E) GSVA plots of blood enrichment scores by AD cohorts for gene signatures for positive regulation of osteoclast differentiation (A, B, E), multinuclear osteoclast differentiation (C, D). F, G) Illustration of sample source of dataset and acquisition of osteoporosis-related monocyte signature gene set. H, I) GSVA plots of blood enrichment scores by AD cohorts for gene signatures for osteoporosis-related monocyte. Unpaired two-sided t-test between groups, one-way ANOVA with Holm-Šidák’s multiple comparisons test, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001. aAD, advanced Alzheimer’s disease; LOAD, late-onset Alzheimer’s disease.

Table 1

Transcriptome (Alzheimer’s disease)

	Accession no	Description	Samples	Sex (y)	Platform	Reference
AD1	GSE140829	UCLA, UCSF;Whole blood	204 AD134 MCI249 HC	104 F (73.05), 100 M (72.9);62 F (71.89), 72 M (74.4);139 F (73.36), 110 M (74.05)	GPL15988 Illumina HumanHT-12V4.0 expression beadchip	https://www.biorxiv.org/content/10.1101/2019.12.13.875112v1(Pre-print)
AD2	GSE63061	AddNeuroMed Cohort;Whole blood	139 AD109 MCI134 HC	85 F (78.2), 54 M (77.5);65 F (77.4), 44 M (79.3);81 F (74.8), 53 M (75.98)	GPL10558 Illumina HumanHT-12V4.0 expressionbeadchip	PMID: 26343147
AD3	GSE97760	Research Institute at Nationwide Childrens Hospital;Whole blood	9 AD10 HC	9 F (79.3);10 F (72.1)	GPL16699 SurePrint G3 Human GE v2 8x60K Microarray	PMID: 25079797
AD4	GSE168813	Universidad Autánoma de ManizalesWhole blood	5 AD10 HC	15 F (76.3)	GPL17303 Ion Torrent Proton	PMID: 34484988

AD, Alzheimer’s disease; MCI, mild cognitive impairment; HC, healthy control; F, female; M, male.

To further validate the identified active osteoclastogenesis signatures detected in AD blood samples, next analysis was made using DEGs obtained from blood monocytes in an OP study [31] (Fig. 1F, G). We refer to this gene set as the OP-related monocyte signature (i.e., 14 down- and 35 upregulated). The OP-related monocyte signature in the blood of patients with AD (Fig. 1H, I) or MCI (Fig. 1I) showed an increase compared to healthy individuals, and GSVA score was positively correlated with disease stage (Fig 1H, I). These data indicate that the Alzheimer’s peripheral blood profiles are pre-set to a state favoring monocyte-osteoclast differentiation, possibly reflecting hyperactive osteoclastogenesis in patients with AD.

Correlation between serum OC marker TRACP5b and neurodegeneration markers in an MCI cohort

Tartrate-resistant acid phosphatase 5b (TRACP5b) can be secreted by OCs and released into the circulation as a marker indicating the number of OCs [32, 33]. We next investigated the relationship between TRACP5b and the markers of AD-related neurodegeneration in MCI subjects using serum protein microarray data (GSE74763) [34]. Although serum TRACP5b levels were slightly upregulated in patients with MCI compared to age-matched healthy controls, there was no statistical difference (data not shown). Glial activation and neuroinflammation occur as early as the prodromal phase of AD pathology [35]. We identified three known glial cell-related markers (i.e., GFAP [36], S100β [37, 38], and TSPO [39]) associated with AD pathology or cognitive impairment in serum microarray data. Serum TRACP5b levels in MCI were significantly positively correlated with GFAP, S100β and TSPO (Fig. 2A-C). Additionally, TRACP5b also correlated positively with serum PDGFR-β reflecting pericyte injury [40]. Neurofilament (Nf) and β-synuclein were also identified in serum microarray data, and both were used to indicate neurodegeneration [41, 42]. Serum TRACP5b correlated strongly positively with Nf but not with β-synuclein (Fig. 2E, F). IL-12 and IL-1β were confirmed to be upregulated in the peripheral blood of patients with AD [43], and we also found that TRACP5b was positively correlated with both serum inflammatory factors (Fig. 2G, H).

Fig. 2

Correlation between serum osteoclast marker TRACP5b and different neurodegeneration markers. A-G) Pearson correlation analysis between serum TRACP5b and serum GFAP, S100β, TSPO, PDGFR-β, β-synuclein, Nf, IL-12A, and IL-1β in MCI cohort. ^**p < 0.01, ^****p < 0.000.

Stratification of blood samples based on osteoclastogenesis signature uncovers dysregulation of Aβ degradation

To explore the relevant mechanisms affected by altered osteoclastogenesis in an unbiased manner, we next divided patients with AD into two groups according to the GSVA score of SPROD (i.e., top 25% : high group, bottom 25% : low group) (Fig. 3A). SPROD-related DEGs (high versus low; p < 0.05), including 133 down- and 204 upregulated, were identified. Enrichment analyses of upregulated DEGs revealed enrichment for GO functions associated with cell development and mitochondrion localization (Fig. 3B, left), which may be consistent with the increased OC differentiation signature observed, as the process of monocyte-OC differentiation involves enhanced mitochondrial biogenesis [44]. Notably, among the top 8 terms identified in the downregulated DEGs, half were enriched in biological processes related to protein degradation (Fig. 3B, right), i.e., ‘positive regulation of hydrolase activity’, ‘regulation of small molecule metabolic process’, ‘regulation of protein depolymerization’, and ‘proteasomal protein catabolic process’.

Fig. 3

Associations between osteoclast differentiation and Aβ degradation signatures in AD blood. A) Illustration of the grouping of AD blood samples. B) GO analysis of top 8 pathways by significance comparing differentially expressed genes in AD blood from patients with high and low osteoclast differentiation signature scores. C, D) GSVA plots of blood enrichment scores by patients with different osteoclast differentiation signatures for negative regulation of amyloid-β clearance gene signature, and Pearson correlation analysis between GSVA scores of positive regulation of osteoclast differentiation and negative regulation of amyloid-β clearance gene signatures (C right, D). E) GSEA results for AD blood with high and low osteoclast differentiation signature scores using ImmuneSigDB gene sets. Unpaired two-sided t-test between groups, ^***p < 0.001, ^****p < 0.000.

Furthermore, we also observed a significant enrichment of signature for negative regulation of Aβ clearance (SNRAC) among patients with high SPROD scores (Fig. 3C). In all patients with AD, SPROD GSVA scores were positively correlated with SNRAC GSVA scores (Fig. 3C), a correlation that was also observed in another dataset (Fig. 3D). It has been reported that activated T cells and cytokines IFN-γ and TNF-α can inhibit the degradation of Aβ by mononuclear phagocytes [26]. GSEA was performed on high SPROD and low SPROD blood transcripts using previously identified immune-related gene sets (gsea-msigdb, C7). Changes associated with T cell activation were found in high SPROD score blood, including positive enrichments for T cell activation-related upregulated gene sets, and a negative enrichment for IFN-γ and TNF-α-treated macrophage downregulated gene set (Fig. 3E). These findings collectively suggest that AD blood transcripts with high osteoclastogenesis signature are associated with dysregulation of immune-mediated Aβ clearance.

Anti-resorptive treatment improves Aβ-degrading enzyme expression in monocytes of APP23 TG mice

We next investigated the effects of OCs on the expression of proteins associated with Aβ clearance in peripheral monocytes using an AD mouse model of APP23 TG. We observed a significant upregulation of serum CTX-1, a bone resorption marker, in 14-month-old TG mice compared to non-transgenic (NTG) mice (Fig. 4B). Bisphosphonates (BPs) are widely used to inhibit bone resorption processes. BP ALN has a high affinity for hydroxyapatite and acts an agent that block osteoclastogenesis [45 –47]. We treated TG mice with ALN (Fig. 4A), and the effect was confirmed by reduced serum CTX-1 (Fig. 4B). Aβ clearance involves several enzymes, such as IDE and NEP [48]. Immunoblotting revealed upregulation of NEP expression levels in the splenic monocytes of ALN-treated TG mice (Fig. 4C, E). In addition, we observed that ALN treatment upregulated the expression of the anti-inflammatory cytokine IL-10 but not the pro-inflammatory IL-1β in AD mouse monocytes (Fig. 4C, F, G). IL-10 has been reported to enhance Aβ degradation [26], suggesting that OCs in AD may negatively regulate monocyte-related Aβ clearance.

Fig. 4

Effects of anti-resorptive agent ALN on monocytes of APP23 mice. A) Timeline of ALN treatment. B) Validation of the inhibitory effect of ALN on osteoclasts based on serum CTX-1 level. C–G) Western blots and quantification for IDE, NEP, IL-1β, and IL-10 from monocyte protein extract of control (NTG) and APP23 TG mice treated with vehicle or ALN; IDE, NEP, IL-1β and IL-10 levels are normalized to β-actin internal controls and expressed as a fold change of the NTG group. One-way ANOVA with Holm-Šidák’s multiple comparisons test, ^*p < 0.05, ^**p < 0.01.

Anti-resorptive treatment ameliorates pathology and memory decline in APP23 TG mice

In view of the effects of OC blocking in modulating the expression of Aβ-degrading enzymes in peripheral monocytes, we next assessed the impact of this intervention on AD pathology. We observed a reduction of Aβ plaques in the cortex and hippocampus of 14-month-old APP23 mice after long-term treatment with ALN (Fig. 5A, B). Activation of astrocytes is usually accompanied by increased expression of their marker GFAP [49] as one of the pathological changes in AD mouse models [50]. We detected a more than 3-fold upregulation of GFAP expression in the hippocampus homogenates of APP23 mice compared to age-matched NTG controls (Fig. 5C, D). However, ALN-treated APP23 mice showed significantly lower GFAP expression (Fig. 5C, D). Additionally, activated microglia are generally believed to be involved in the neuroinflammation of AD along with astrocytes [51]. Consistent with changes in GFAP expression, we observed lower levels of the microglial marker Iba-1 in the hippocampal homogenate protein of ALN-treated APP23 mice (Fig. 5C, D).

Fig. 5

ALN ameliorates pathology and fear memory deficits in APP23 mice. A, B) Amyloid pathology is reduced for 14-month-old APP23 TG mice in ALN treatment; scale bar, 500μm. C, D) Western blots and quantification for GFAP and Iba-1 from hippocampal homogenates of control (NTG) and APP23 TG mice treated with vehicle or ALN; GFAP and Iba-1 levels are normalized to β-actin internal controls and expressed as a fold change of the NTG group. E, F) Freezing behavior presented as the percentage of time mice are frozen during contextual testing (E) and cued testing (F). One-way ANOVA with Holm-Šidák’s multiple comparisons test, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001.

To assess the effects of ALN intervention on memory-related behavior in AD mice, we conducted a context- and cue-dependent fear conditioning experiment. Mice were trained to associate context and tone (cues) with unconditioned foot shocks. In the contextual fear memory test, APP23 mice showed a significant associative memory deficit, i.e., a reduction in freezing time, while the ALN-treated APP23 mice were comparable to the NTG control in terms of freezing time (Fig. 5E). Similar alterations were also observed in cued fear memory tests (Fig. 5F), in which mice were placed in a new environment and presented with tones. Together, these data suggest that long-term pharmacological intervention against OCs appears to help slow pathological progression as well as behavioral deterioration in AD mice.

DISCUSSION

In this study, we propose that OCs, traditionally thought to be involved solely in bone resorption, may be implicated in the pathological progression of AD. We used transcriptomic analysis to characterize the molecular features of blood in AD, and consistently identified differences in the enrichment of gene signatures associated with osteoclastogenesis between diseased and healthy blood samples across multiple AD cohorts. Given the inherent pathway of monocyte-to-OC differentiation in the bloodstream [14], these blood-derived insights may offer some understanding of individual OC differentiation status. Our analysis revealed a stronger signature for osteoclastogenesis in the blood of patients with AD, aligning directionally with previous reports linking AD to an increased risk of fractures and OP [1, 2]. Interestingly, Aβ appears to promote OC activation [52] but inhibit differentiation of pre-OCs into OCs [53]. However, a recent study has demonstrated that within the inflammatory microenvironment, Aβ actually promotes OC differentiation through a Trem2-dependent signaling pathway [54]. Considering the systemic inflammation observed in patients with AD [55], it seems plausible that the positive effect of Aβ on OC differentiation is dependent on individual inflammatory status. Overall, our data further solidify the association between AD and human OC differentiation.

We evaluated the dataset using GSVA, allowing us to quantify individual levels of osteoclast differentiation characteristics and identify patterns of pathway enrichment between subgroups with different levels. The established relationship between immunity and AD reveals an immunosuppressive milieu in the brain and peripheral immune microenvironment of AD patients, including hyporesponsive microglia and monocytes to Aβ [18–20 , 56], expansion of an immunosuppressive neutrophil subset [57], and an increase in regulatory T cells in the preclinical stage [58, 59]. Breaking immune tolerance may serve as an intervention strategy for AD [60]. Of these, functionally limited monocytes have the most direct association with AD pathology due to their characterized capacity for Aβ clearance [21, 22]. A previous study showed that nearly 40%–60% of brain Aβ flows into the peripheral system for clearance, while spleen mononuclear phagocytes are mainly responsible for peripheral Aβ clearance [61]. Therefore, changes in peripheral or splenic monocyte homeostasis may be sufficient to influence the progression of AD pathology. Aβ-degrading enzyme is expressed in neurons, microglia and monocytes [62]. A previous study found that in AD mice, CD11b⁺ monocytes transfected with human soluble NEP could be recruited around dense Aβ plaques in the brain and completely prevent plaque development [63]. Therefore, changes in Aβ-degrading enzymes in peripheral monocytes may contribute to AD progression by affecting their capacity to clear Aβ. We serendipitously discovered that the downregulated blood DEGs in AD blood with high osteoclastogenic activity were largely associated with protein degradation-related processes. This observation suggests that dysregulation of peripheral Aβ clearance pathways. Notably, a previous study has reported that hip fracture results in phagocytosis inhibition and increases TNFα production by classical monocytes [64]. Given that fracture healing involves OC differentiation and activation, this finding also implies a potential link between OCs and peripheral monocyte dysfunction. Given these observations, we propose that OCs exert a regulatory effect on peripheral monocytes, potentially impacting peripheral Aβ clearance by inhibiting monocyte-mediated Aβ degradation, however this may need to be verified by direct functional experiments in the future.

Human OCs have the ability to suppress T cell proliferation [65], and in tumors, OCs promote the formation of an immunosuppressive microenvironment [66]. This OC-mediated adaptive immune negative regulation may also be present in the context of AD. From these findings, interventions targeting OCs may play a role in coordinating the entire peripheral immune network. Further studies are needed to elucidate the exact mechanisms by which OCs influence immune regulation in AD and explore their therapeutic implications.

The APP23 mouse exhibits a relatively slow progressive cerebral amyloid deposition, reflecting a decline in Aβ clearance as the animals age [67], which closely resembles the development of sporadic AD in humans. We performed long-term pharmacological depletion of OCs in APP23 mice and observed an improvement in Aβ pathology and cognitive impairment at old ages. However, future work may involve introducing OP-related environmental factors to young AD mice, assessing the impact of these factors on AD pathology while simultaneously blocking OC signaling, for a more comprehensive evaluation of the interaction between OCs and AD pathology.

Given that the excessive activation of OCs can occur in specific populations such as bedridden individuals, postmenopausal women, and patients with inflammatory bone diseases, which often precedes the onset of AD, early identification of such risks can facilitate interventions to prevent or slow the progression of AD. Additionally, investigating the association between the history of taking anti-resorptive medication and the risk of AD may shed light on these findings.

AUTHOR CONTRIBUTIONS

Bin Wu (Conceptualization; Data curation; Investigation; Visualization; Writing – original draft); Mulan Chen (Investigation; Methodology; Writing – review & editing); Ling Meng (Data curation; Investigation; Methodology; Writing – review & editing); Qiuyun Tian (Investigation; Methodology; Writing – review & editing); Zhifang Dong (Conceptualization; Funding acquisition; Project administration; Supervision; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

The authors have no acknowledgments to report.

FUNDING

This work was supported by grants from the National Natural Science Foundation of China (32371030 and 82071395) and the CQMU Program for Youth Innovation in Future Medicine (W0044).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Baker

, Cook

, Arrighi

, Bullock

(2011) Hip fracture risk and subsequent mortality among Alzheimer’s disease patients in the United Kingdom, 1988–2007. Age Ageing 40, 49–54.

Wang

, Wu

, Tee

, Lo

(2018) Medical comorbidity in Alzheimer’s disease: A nested case-control study. J Alzheimers Dis 63, 773–781.

Amouzougan

, Lafaie

, Marotte

, Denarie

, Collet

, Pallot-Prades

, Thomas

(2017) High prevalence of dementia in women with osteoporosis. Joint Bone Spine 84, 611–614.

Kostev

, Hadji

, Jacob

(2018) Impact of osteoporosis on the risk of dementia in almost 60,000 patients followed in general practices in Germany. J Alzheimers Dis 65, 401–407.

Xiao

, Ghatan

, Mooldijk

, Trajanoska

, Oei

, Gomez

, Ikram

, Rivadeneira

, Ikram

(2023) Association of bone mineral density and dementia. Neurology 100, e2125–e2133.

Henriksen

, Leeming

, Byrjalsen

, Nielsen

, Sorensen

, Dziegiel

, Martin

, Christiansen

, Qvist

, Karsdal

(2007) Osteoclasts prefer aged bone. Osteoporosis Int 18, 751–759.

Baecker

, Tomic

, Mika

, Gotzmann

, Platen

, Gerzer

, Heer

(2003) Bone resorption is induced on the second day of bed rest: Results of a controlled crossover trial. J Appl Physiol 95, 977–982.

McGee-Lawrence

, Carey

, Donahue

(2008) Mammalian hibernation as a model of disuse osteoporosis: The effects of physical inactivity on bone metabolism, structure, and strength. Am J Physiol Regul Integr Comp Physiol 295, R1999–2014.

Compston

(2001) Sex steroids and bone. Physiol Rev 81, 419–447.

10.

Scarmeas

, Luchsinger

, Brickman

, Cosentino

, Schupf

, Xin-Tang

, Gu

, Stern

(2011) Physical activity and Alzheimer disease course. Am J Geriatr Psychiatry 19, 471–481.

11.

Kivimäki

, Singh-Manoux

, Pentti

, Sabia

, Nyberg

, Alfredsson

, Goldberg

, Knutsson

, Koskenvuo

, Koskinen

, Kouvonen

, Nordin

, Oksanen

, Strandberg

, Suominen

, Theorell

, Vahtera

, Väänänen

, Virtanen

, Westerholm

, Westerlund

, Zins

, Seshadri

, Batty

, Sipilä

, Shipley

, Lindbohm

, Ferrie

, Jokela

(2019) Physical inactivity, cardiometabolic disease, and risk of dementia: An individual-participant meta-analysis. BMJ 365, l1495.

12.

Barth

, Crestol

, de Lange

, Galea

LAM

(2023) Sex steroids and the female brain across the lifespan: Insights into risk of depression and Alzheimer’s disease. Lancet Diabetes Endocrinol 11, 926–941.

13.

Cui

, Xiong

, Hong

, Jung

, Li

, Liu

, Yan

, Mei

, Feng

, Xiong

(2011) APPswe/Aβ regulation of osteoclast activation and RAGE expression in an age-dependent manner. J Bone Miner Res 26, 1084–1098.

14.

Pereira

, Petretto

, Gordon

, Bassett

JHD

, Williams

, Behmoaras

(2018) Common signalling pathways in macrophage and osteoclast multinucleation. J Cell Sci 131, jcs216267.

15.

Jin

, Wei

, Yang

, Du

, Wan

(2014) Mitochondrial complex I activity suppresses inflammation and enhances bone resorption by shifting macrophage-osteoclast polarization. Cell Metab 20, 483–498.

16.

Purton

, Lee

, Torok-Storb

(1996) Normal human peripheral blood mononuclear cells mobilized with granulocyte colony-stimulating factor have increased osteoclastogenic potential compared to nonmobilized blood. Blood 87, 1802–1808.

17.

Liu

, Dvornyk

, Lu

, Shen

, Lappe

, Recker

, Deng

(2005) A novel pathophysiological mechanism for osteoporosis suggested by an in vivo gene expression study of circulating monocytes. J Biol Chem 280, 29011–29016.

18.

, Huang

, Ou

, Rembach

, Fowler

, Avula

, Horton

, Doecke

, Villemagne

, Macaulay

, Maruff

, Fletcher

, Guymer

, Wiley

, Masters

(2016) Innate phagocytosis by peripheral blood monocytes is altered in Alzheimer’s disease. Acta Neuropathol 132, 377–389.

19.

Chen

, Tian

, Shen

, Cheng

, Fan

, Sun

, He

, Sun

, Bu

, Zeng

, Liu

, Deng

, Xu

, Chen

, Wang

(2020) Amyloid-beta uptake by blood monocytes is reduced with ageing and Alzheimer’s disease. Transl Psychiatry 10, 423.

20.

Bradshaw

, Chibnik

, Keenan

, Ottoboni

, Raj

, Tang

, Rosenkrantz

, Imboywa

, Lee

, Von Korff

, Morris

, Evans

, Johnson

, Sperling

, Schneider

, Bennett

, De Jager

(2013) CD33 Alzheimer’s disease locus: Altered monocyte function and amyloid biology. Nat Neurosci 16, 848–850.

21.

Yan

, Kim

, Xiao

, Ma

, Czerniewski

, Liu

, Rawnsley

, Yan

, Randolph

, Epelman

, Lee

, Diwan

(2022) Peripheral monocyte-derived cells counter amyloid plaque pathogenesis in a mouse model of Alzheimer’s disease. J Clin Invest 132, e152565.

22.

Simard

, Soulet

, Gowing

, Julien

, Rivest

(2006) Bone marrow-derived microglia play a critical role in restricting senile plaque formation in Alzheimer’s disease. Neuron 49, 489–502.

23.

Madel

, Ibáñez

, Wakkach

, de Vries

, Teti

, Apparailly

, Blin-Wakkach

(2019) Immune function and diversity of osteoclasts in normal and pathological conditions. Front Immunol 10, 1408.

24.

Park-Min

, Lorenzo

(2022) Osteoclasts: Other functions. Bone 165, 116576.

25.

Ibáñez

, Abou-Ezzi

, Ciucci

, Amiot

, Belaïd

, Obino

, Mansour

, Rouleau

, Wakkach

, Blin-Wakkach

(2016) Inflammatory osteoclasts prime TNFα-producing CD4+T cells and express CX3 CR1. J Bone Miner Res 31, 1899–1908.

26.

Yamamoto

, Kiyota

, Walsh

, Liu

, Kipnis

, Ikezu

(2008) Cytokine-mediated inhibition of fibrillar amyloid-beta peptide degradation by human mononuclear phagocytes. J Immunol 181, 3877–3886.

27.

Hänzelmann

, Castelo

, Guinney

(2013) GSVA: Gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 14, 7.

28.

Zhou

, Zhou

, Pache

, Chang

, Khodabakhshi

, Tanaseichuk

, Benner

, Chanda

(2019) Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun 10, 1523.

29.

Dong

, Han

, Li

, Bai

, Wang

, Tu

, Peng

, Zhou

, He

, Wu

, Tan

, Liu

, Wu

, Zhou

, Jin

, Zhang

, Sacktor

, Li

, Song

, Wang

(2015) Long-term potentiation decay and memory loss are mediated by AMPAR endocytosis. J Clin Invest 125, 234–247.

30.

Miyamoto

, Suda

(2003) Differentiation and function of osteoclasts. Keijo J Med 52, 1–7.

31.

Zhou

, Gao

, Xu

, Shen

, Tian

, Deng

(2018) A novel approach for correction of crosstalk effects in pathway analysis and its application in osteoporosis research. Sci Rep 8, 668.

32.

Rissanen

, Suominen

, Peng

, Halleen

(2008) Secreted tartrate-resistant acid phosphatase 5b is a Marker of osteoclast number in human osteoclast cultures and the rat ovariectomy model. Calcified Tissue Int 82, 108–115.

33.

Halleen

, Tiitinen

, Ylipahkala

, Fagerlund

, Väänänen

(2006) Tartrate-resistant acid phosphatase 5b (TRACP 5b) as a marker of bone resorption. Clin Lab 52, 499–509.

34.

DeMarshall

, Nagele

, Sarkar

, Acharya

, Godsey

, Goldwaser

, Kosciuk

, Thayasivam

, Han

, Belinka

, Nagele

(2016) Detection of Alzheimer’s disease at mild cognitive impairment and disease progression using autoantibodies as blood-based biomarkers. Alzheimers Dement (Amst) 3, 51–62.

35.

Nordengen

, Kirsebom

, Henjum

, Selnes

, Gísladóttir

, Wettergreen

, Torsetnes

, Grøntvedt

, Waterloo

, Aarsland

, Nilsson

LNG

, Fladby

(2019) Glial activation and inflammation along the Alzheimer’s disease continuum. J Neuroinflammation 16, 46.

36.

Pereira

, Janelidze

, Smith

, Mattsson-Carlgren

, Palmqvist

, Teunissen

, Zetterberg

, Stomrud

, Ashton

, Blennow

, Hansson

(2021) Plasma GFAP is an early marker of amyloid-β but not tau pathology in Alzheimer’s disease. Brain 144, 3505–3516.

37.

Cox

, Allerhand

, Ritchie

, Muñoz Maniega

, Valdés Hernández

, Harris

, Dickie

, Anblagan

, Aribisala

, Morris

, Sherwood

, Abbott

, Starr

, Bastin

, Wardlaw

, Deary

(2018) Longitudinal serum S100β and brain aging in the Lothian Birth Cohort 1936. Neurobiol Aging 69, 274–282.

38.

Peskind

, Griffin

, Akama

, Raskind

, Van Eldik

(2001) Cerebrospinal fluid S100B is elevated in the earlier stages of Alzheimer’s disease. Neurochem Int 39, 409–413.

39.

Tournier

, Tsartsalis

, Ceyzériat

, Fraser

, Grégoire

, Kövari

, Millet

(2020) Astrocytic TSPO upregulation appears before microglial TSPO in Alzheimer’s disease. J Alzheimers Dis 77, 1043–1056.

40.

Van Hulle

, Ince

, Okonkwo

, Bendlin

, Johnson

, Carlsson

, Asthana

, Love

, Blennow

, Zetterberg

, Scott Miners

(2024) Elevated CSF angiopoietin-2 correlates with blood-brain barrier leakiness and markers of neuronal injury in early Alzheimer’s disease. Transl Psychiatry 14, 3.

41.

Preische

, Schultz

, Apel

, Kuhle

, Kaeser

, Barro

, Gräber

, Kuder-Buletta

, LaFougere

, Laske

, Vöglein

, Levin

, Masters

, Martins

, Schofield

, Rossor

, Graff-Radford

, Salloway

, Ghetti

, Ringman

, Noble

, Chhatwal

, Goate

, Benzinger

TLS

, Morris

, Bateman

, Wang

, Fagan

, McDade

, Gordon

, Jucker

(2019) Serum neurofilament dynamics predicts neurodegeneration and clinical progression in presymptomatic Alzheimer’s disease. Nat Med 25, 277–283.

42.

Mohaupt

, Pons

, Vialaret

, Delaby

, Hirtz

, Lehmann

(2022) β-Synuclein as a candidate blood biomarker for synaptic degeneration in Alzheimer’s disease. Alzheimers Res Ther 14, 179.

43.

Swardfager

, Lanctôt

, Rothenburg

, Wong

, Cappell

, Herrmann

(2010) A meta-analysis of cytokines in Alzheimer’s disease. Biol Psychiatry 68, 930–941.

44.

Park-Min

(2019) Metabolic reprogramming in osteoclasts. Semin Immunopathol 41, 565–572.

45.

Nancollas

, Tang

, Phipps

, Henneman

, Gulde

, Wu

, Mangood

, Russell

, Ebetino

(2006) Novel insights into actions of bisphosphonates on bone: Differences in interactions with hydroxyapatite. Bone 38, 617–627.

46.

Fisher

, Rogers

, Halasy

, Luckman

, Hughes

, Masarachia

, Wesolowski

, Russell

, Rodan

, Reszka

(1999) Alendronate mechanism of action: Geranylgeraniol, an intermediate in the mevalonate pathway, prevents inhibition of osteoclast formation, bone resorption, and kinase activation in vitro. Prac Natl Acad Sci U S A 96, 133–138.

47.

Yano

, Tsutsumi

, Soga

, Lee

, Trepel

, Osada

, Neckers

(2008) Inhibition of Hsp90 activates osteoclast c-Src signaling and promotes growth of prostate carcinoma cells in bone. Prac Natl Acad Sci U S A 105, 15541–15546.

48.

Qiu

, Walsh

, Ye

, Vekrellis

, Zhang

, Podlisny

, Rosner

, Safavi

, Hersh

, Selkoe

(1998) Insulin-degrading enzyme regulates extracellular levels of amyloid beta-protein by degradation. J Biol Chem 273, 32730–32738.

49.

Sofroniew

, Vinters

(2010) Astrocytes: Biology and pathology. Acta Neuropathol 119, 7–35.

50.

Oakley

, Cole

, Logan

, Maus

, Shao

, Craft

, Guillozet-Bongaarts

, Ohno

, Disterhoft

, Van Eldik

, Berry

, Vassar

(2006) Intraneuronal beta-amyloid aggregates, neurodegeneration, and neuron loss in transgenic mice with five familial Alzheimer’s disease mutations: Potential factors in amyloid plaque formation. J Neurosci 26, 10129–10140.

51.

Hansen

, Hanson

, Sheng

(2018) Microglia in Alzheimer’s disease. J Cell Biol 217, 459–472.

52.

, Yang

, Teguh

, Zhou

, Xu

, Rong

(2016) Amyloid β peptide enhances RANKL-induced osteoclast activation through NF-κB, ERK, and calcium oscillation signaling. Int J Mol Sci 17, 1683.

53.

Yang

, Li

, Chen

, Feng

, He

, Liu

, He

, Wang

, Chen

, Xie

, Rong

(2020) Amyloid β peptide promotes bone formation by regulating Wnt/β-catenin signaling and the OPG/RANKL/RANK system. FASEB J 34, 3583–3593.

54.

Weng

, Wang

, Li

, Feng

, Xu

, Wang

(2021) Trem2 mediated Syk-dependent ROS amplification is essential for osteoclastogenesis in periodontitis microenvironment. Redox Biol 40, 101849.

55.

Holmes

, Cunningham

, Zotova

, Woolford

, Dean

, Kerr

, Culliford

, Perry

(2009) Systemic inflammation and disease progression in Alzheimer disease. Neurology 73, 768–774.

56.

Monsonego

, Maron

, Zota

, Selkoe

, Weiner

(2001) Immune hyporesponsiveness to amyloid beta-peptide in amyloid precursor protein transgenic mice: Implications for the pathogenesis and treatment of Alzheimer’s disease. Proc Natl Acad Sci U S A 98, 10273–10278.

57.

Dong

, Lagarde

, Xicota

, Corne

, Chantran

, Chaigneau

, Crestani

, Bottlaender

, Potier

, Aucouturier

, Dorothée

, Sarazin

, Elbim

(2018) Neutrophil hyperactivation correlates with Alzheimer’s disease progression. Ann Neurol 83, 387–405.

58.

, Duan

, Mo

, Xiao

, Huang

, Chen

, Xiang

, Yang

, Chen

, Xu

(2020) Mild cognitive impairment patients have higher regulatory T-Cell proportions compared with Alzheimer’s disease-related dementia patients. Front Aging Neurosci 12, 624304.

59.

Saresella

, Calabrese

, Marventano

, Piancone

, Gatti

, Calvo

, Nemni

, Clerici

(2010) PD1 negative and PD1 positive CD4+T regulatory cells in mild cognitive impairment and Alzheimer’s disease. J Alzheimers Dis 21, 927–938.

60.

Baruch

, Rosenzweig

, Kertser

, Deczkowska

, Sharif

, Spinrad

, Tsitsou-Kampeli

, Sarel

, Cahalon

, Schwartz

(2015) Breaking immune tolerance by targeting Foxp3(+) regulatory T cells mitigates Alzheimer’s disease pathology. Nat Commun 6, 7967.

61.

, Chen

, Tan

, Zeng

, He

, Wang

, Bu

, Wang

(2022) Physiological clearance of Aβ by spleen and splenectomy aggravates Alzheimer-type pathogenesis. Aging Cell 21, e13533.

62.

Britschgi

, Wyss-Coray

(2007) Immune cells may fend off Alzheimer disease. Nat Med 13, 408–409.

63.

Lebson

, Nash

, Kamath

, Herber

, Carty

, Lee

, Li

, Szekeres

, Jinwal

, Koren

, Dickey

, Gottschall

, Morgan

, Gordon

(2010) Trafficking CD11b-positive blood cells deliver therapeutic genes to the brain of amyloid-depositing transgenic mice. J Neurosci 30, 9651–9658.

64.

Baëhl

, Garneau

, Lorrain

, Viens

, Svotelis

, Lord

, Cabana

, Larbi

, Dupuis

, Fülöp

(2016) Alterations in monocyte phenotypes and functions after a hip fracture in elderly individuals: A 6-month longitudinal study. Gerontology 62, 477–490.

65.

, Lu

, Qian

, Zheng

, Zhang

, Bi

, He

, Liu

, Xu

, Gao

, Yi

(2014) Human osteoclasts are inducible immunosuppressive cells in response to T cell-derived IFN-γ and CD40 ligand in vitro. J Bone Miner Res 29, 2666–2675.

66.

, Acharya

, Feng

, Wen

, Zhong

, Zhang

, Munshi

, Qiu

, Tai

, Anderson

(2016) Osteoclasts promote immune suppressive microenvironment in multiple myeloma: Therapeutic implication. Blood 128, 1590–1603.

67.

Kuo

, Beach

, Sue

, Scott

, Roher

(2001) The evolution of Aβ peptide burden in the APP23 transgenic mice: Implications for Aβ deposition in Alzheimer disease. Mol Med 7, 609–618.