Abstract

   As an enticing bone anabolic target, short‐term inhibition of
   Schnurri‐3 (SHN3) resulted in high‐bone mass due to augmented
   osteoblast activity. However, no studies are conducted to identify
   natural products targeting SHN3 inhibition. Herein, a screening
   strategy for the discovery of marine compounds that facilitate
   osteoblast differentiation by targeting SHN3 silencing is presented.
   One leading quinolinone alkaloid, viridicatol (VDC), isolated from
   deep‐sea‐derived fungus, vigorously promotes osteogenic differentiation
   via the Wnt/SHN3 signaling pathway in osteoblasts, thereby preventing
   osteoporosis while enhancing bone‐fracture healing in a mouse model.
   Subsequently, the SDSSD (Ser, Asp, Ser, Ser, Asp) is further employed
   to engineer bone‐targeting nanovesicles (BT‐NVs) for the optimal
   delivery of VDC to osteoblasts, which mitigates the bone loss observed
   in a severe osteogenesis imperfecta model. Hence, these results
   initially uncover a promising marine natural product, VDC, targeting
   the Wnt/SHN3 pathway for the treatment of bone loss and highlighting
   its translational potential in clinical applications.

   Keywords: bone loss, bone‐targeting nanovesicles, osteoblasts,
   Schnurri‐3, viridicatol
     __________________________________________________________________

   Viridicatol (VDC), isolated from the deep‐sea‐derived fungus, enhances
   osteogenesis in vitro and in vivo by targeting Wnt/SHN3 signaling
   pathway. It can be delivered via engineered nanovesicles to mitigate
   bone loss of osteoporosis, fracture repair, and osteogenesis
   imperfecta.

   graphic file with name ADVS-12-2416140-g009.jpg

1. Introduction

   Recent advancements in osteoanabolic therapies have been witnessed,
   particularly with the advent of the anti‐SOST biologic romosozumab.
   Nevertheless, there remains a substantial demand for further progress
   in this domain, as all current anabolic treatments possess considerable
   limitations.^[ [44]^1 ^] For instance, the anti‐sclerostin therapy of
   romosozumab may be associated with an elevated risk of cardiovascular
   events, and its anabolic effects diminish after one year of
   treatment.^[ [45]^2 ^] Additionally, anabolic therapies employing
   parathyroid hormone (PTH) analogs including teriparatide and
   abaloparatide are confined to a maximum duration of two years on
   account of discoveries that high doses of PTH were correlated with the
   development of osteosarcoma in rodent studies.^[ [46]^3 ^] Notably,
   contemporary agents display substantial site‐specific disparities in
   their osteoanabolic efficacy, with PTH analogues demonstrating
   significantly greater effectiveness in enhancing bone formation in the
   spine compared to distal locations like the distal radius.^[ [47]^4 ^]
   Hence, fractures in patients with severe osteoporosis or osteogenesis
   imperfecta (OI), caused by a decreased strength of bone, require a
   diverse assortment of mechanistically distinct osteoanabolic
   therapies.^[ [48]^5 ^]

   The Wnt signaling pathway, encompassing Wnt ligands, receptors,
   intracellular components, transcription factors, and antagonists, plays
   a crucial role in bone development, formation, and homeostasis.^[
   [49]^6 ^] Upon activation by Wnt ligands, the canonical Wnt signaling
   stabilizes β‐catenin, which translocates to the nucleus to activate
   target genes essential for osteoblast differentiation and function.
   Schnurri‐3 (SHN3), a large zinc finger adapter protein encoded by
   Hivep3, acts as a potent cell‐intrinsic negative regulator of
   osteoblast activity. SHN3 inhibits the Wnt signaling pathway by
   selectively suppressing the phosphorylation of specific ERK substrates
   such as p90RSK and GSK3β, rather than directly inactivating ERK itself.
   This regulation is critical because reducing ERK activity through the
   Mek1/2‐Het genotype reverses the high bone mass observed in SHN3
   haploinsufficiency, indicating that SHN3 suppresses ERK activity in
   vivo.^[ [50]^7 ^] Consequently, the Wnt/SHN3 pathway has garnered
   significant attention due to its regulatory influence on osteoblast
   differentiation and bone mass maintenance.^[ [51]^7 , [52]^8 ^] Our
   prior studies discovered that mice deficient in SHN3 exhibit
   adult‐onset osteosclerosis, characterized by increased bone mass
   resulting from enhanced osteoblast activity.^[ [53]^8 , [54]^9 ^] More
   recently, SHN3 has been shown to play a crucial role in the treatment
   of bone‐fracture,^[ [55]^9 , [56]^10 ^] osteoporosis,^[ [57]^8 ,
   [58]^11 ^] rheumatoid arthritis,^[ [59]^12 ^] and OI^[ [60]^13 ^] in
   animal models. Osteoblast‐lineage cells exhibit high sensitivity
   following reduced SHN3 expression levels. Notably, a mere 30% decrease
   in SHN3 mRNA levels in adult mice resulted in a significant, ≈50%,
   increase in bone mass.^[ [61]^8 , [62]^14 ^] Partial SHN3 inhibition
   can lead to significant increases in bone mass, facilitated through an
   increased rate of bone formation. This suggests that compounds capable
   of blocking SHN3 expression or activity serve as promising anabolic
   agents for the treatment of bone loss diseases. Bone‐targeting AAV or
   exosome‐mediated silencing of SHN3 expression^[ [63]^8 , [64]^10 ,
   [65]^11 ^] marks the dawn of treating bone loss diseases. Still,
   currently, there are no literature reports on the targeted delivery of
   natural small molecules to silence SHN3. Targeting the Wnt/SHN3 pathway
   can serve as a powerful new strategy for treating bone loss diseases by
   enhancing bone formation.

   The deep‐sea environment with its structurally diverse array of
   secondary metabolites is promising reservoir of new drug candidates.^[
   [66]^15 ^] Numerous marine drugs are currently applied in clinical
   trials including ziconotide, cytarabine, and trabectedin. Despite the
   significant potential of marine natural products as effective drugs for
   the treatment of bone loss diseases, research on these marine products
   is limited.^[ [67]^16 ^] In this study, the osteogenic differentiation
   ability of 251 compounds from deep‐sea‐derived fungi were investigated
   for osteogenic activity. Among these compounds, viridicatol (VDC), a
   viridicatin‐type alkaloids with a m‐hydroxyl groups, was found to
   promote early osteogenic differentiation and late mineralization, while
   inhibiting SHN3. Previous studies have revealed that VDC exhibits
   various biological activities, including anti‐inflammatory, antitumor,
   and antifungal activities. This compound has been biosynthesized for
   research given that it exhibits beneficial biological activity;
   moreover, the enzyme that catalyzes meta hydroxylation rarely occurs
   naturally.^[ [68]^17 ^] Despite these beneficial properties exhibited
   by VDC, its underlying mechanism in bone metabolic diseases, and its
   therapeutic potential remains unclear. Consequently, we first report
   that VDC directly targets SHN3 and exhibits strong osteogenic activity
   in vitro and in vivo by activating the Wnt/SHN3 signaling pathway. To
   validate VDC's therapeutic potential, we assessed the osteogenic
   efficacy of VDC in vivo models of osteoporosis, fracture, and OI. Given
   the severity of OI, we developed a novel bone‐targeting drug delivery
   system using modified homologous membrane derivatives, resulting in the
   creation of bone‐targeting nanovesicles (BT‐NVs) for drug delivery.
   This method allows for the precise delivery of VDC to bone tissue,
   thereby effectively suppressing the SHN3 gene and enhancing the
   treatment of OI (Scheme [69] 1 ).

Scheme 1.

   Scheme 1
   [70]Open in a new tab

   Schematic diagram of the potential mechanism by which viridicatol
   (VDC), derived from deep‐sea fungus, enhances osteoblastogenesis
   through modulation of the Wnt/SHN3 signaling pathway. Bone‐targeting
   nanovesicles (BT‐NVs) were engineered from MC3T3‐E1 cells using a
   multi‐step process, including ultrasonic lysis, density gradient
   centrifugation, and hydrophobic surface modification. VDC was
   subsequently loaded into BT‐NVs via ultrasonic encapsulation, improving
   targeted delivery to osteoblasts. Once delivered, VDC enhances
   osteogenic activity by regulating key mediator SHN3 in the Wnt
   signaling pathway, manifested by promoting the expression of key genes
   involved in osteogenic differentiation, such as upregulating the
   expression of alkaline phosphatase (Alpl), Runt‐related transcription
   factor 2 (Runx2), Osterix (SP7), Integrin binding sialoprotein (Ibsp),
   Osteocalcin (Bglap), and downregulating the expression of human
   immunodeficiency virus type I enhancer binding protein 3 (Hivep3),
   ultimately promoting bone formation. This innovative approach holds
   significant promise for treating osteoporosis, facilitating fracture
   repair, and addressing osteogenesis imperfecta.

2. Results

2.1. VDC Promotes Osteogenesis Differentiation and Inhibits Hivep3 Expression

   To investigate the effects of natural marine products on osteogenic
   differentiation, we screened 251 compounds derived from deep‐sea fungi
   and measured alkaline phosphatase (Alp) activity as an indicator
   (Figure [71] 1A). 19 compounds showed notable osteogenic activity, with
   VDC being especially potent (Figure [72]1B). VDC significantly enhanced
   Alp activity while inhibiting Hivep3 expression (Figure [73]S1,
   Supporting Information). Additionally, VDC modulated the expression of
   key genes associated with various phases of osteogenesis, including
   Runx2, Sp7, Atf4, Ibsp, Sema3a, Sema3e, and Spp1 (Figure [74]S2,
   Supporting Information). Consequently, VDC was chosen for further
   investigation given its relative potency and yield. VDC exhibited
   robust Alp activity (Figure [75]1D,E) and osteoblast mineralization
   (Figure [76]1F) in a concentration‐dependent manner among the MC3T3‐E1
   cells without cytotoxicity (Figure [77]1C). The EC[50] value for
   promoting bone mineralization was 5.204 µm (Figure [78]1G).
   Bone‐derived mesenchymal stem cells (BMSCs) can provide an effective
   therapeutic strategy for preventing age‐related osteoporosis.^[ [79]^18
   ^] Subsequently, we investigated the ability of VDC to directly induce
   BMSCs differentiation into osteoblasts, chondroblasts, and adipocytes.
   VDC significantly promoted the mineralization of BMSCs in a
   concentration‐dependent manner, with an EC[50] value of 4.132 µm
   (Figure [80]1H,I). The adipogenic and chondrogenic differentiation of
   BMSCs exposed to the same concentrations of VDC were assessed through
   oil red O and alcian blue staining. The results demonstrated that VDC
   induced chondrogenic differentiation (Figure [81]1J,K), but had no
   effect on adipogenic differentiation among BMSCs (Figure [82]1L,M).
   These findings suggest that VDC effectively regulates BMSCs
   differentiation, promoting osteogenic mineralization and cartilage
   formation. Additionally, the TRAP staining results indicated that VDC
   did not exhibit concentration‐dependent inhibition of
   osteoclastogenesis activity within the concentration range of 1 to
   10 µm. Consequently, we concluded that VDC exerts its effects primarily
   on osteoblasts rather than osteoclasts (Figure [83]S3, Supporting
   Information).

Figure 1.

   Figure 1
   [84]Open in a new tab

   VDC promotes the osteogenesis activity of MC3T3‐E1 cells and BMSCs. A)
   A schematic figure showing the steps involved in screening for
   osteogenic activity of VDC. B) The 2D and 3D structures of VDC. C) The
   cytotoxicity of VDC (0–40 µm) among MC3T3‐E1 cells as determined by the
   CCK‐8 assay. (n = 5). D) Representative images of ALP activity
   analysis. Under osteogenesis induction conditions, the MC3T3‐E1 cells
   were treated with different concentrations of VDC (1, 2.5, 5, and
   10 µm) for 5 days, and E) the quantitative analysis of VDC was
   calculated from the optical density values (n = 3). F) Representative
   images of alizarin red S staining. Under the osteogenesis induction
   conditions, the MC3T3‐E1 cells were treated with different
   concentrations of VDC (1, 2.5, 5, 10, 15, and 20 µm) or Purmorphamine
   (PM, 1 µm) for 21 days. The PM served as the positive control. Scale
   bar, 100 µm. G) The EC[50] values of osteogenic mineralization (n = 5).
   H) Representative images of alizarin red S staining. The BMSCs were
   treated with different concentrations of VDC (1, 2.5, 5, 10, 15, and
   20 µm) for 10 days under adipogenesis induction conditions. Scale bar,
   100 µm. I) The EC[50] values of osteogenic mineralization (n = 5). J)
   Representative images of alcian blue staining. The BMSCs were treated
   with different concentrations of VDC (1, 5, and 10 µm) for 21 days
   under adipogenesis induction conditions. Scale bar, 100 µm. K) The
   quantitative results of the alcian blue staining assay (n = 3). L)
   Representative images of Oil red O staining. The BMSCs were treated
   with different concentrations of VDC (1, 5, and 10 µm) for 10 days
   under adipogenesis induction conditions. Scale bar, 100 µm (n = 3). M)
   The quantitative results of Oil red O staining assay. Normal culture
   medium, differentiation medium, and 0.1% DMSO served as the negative
   control (NC), differentiation (Diff.), and solvent group (DMSO),
   respectively. Data represent mean ± SD, ns: not significant, *p < 0.05,
   **p < 0.01, ***p < 0.001, ****p < 0.0001 versus the DMSO group by
   one‐way ANOVA with Tukey's post‐hoc test.

2.2. VDC Induces Osteogenesis via Wnt Signaling

   RNA‐seq was employed to explore the mechanisms underlying
   osteoblastogenesis in MC3T3‐E1 cells treated with or without VDC for 3,
   5, and 8 days. The volcano plot analysis revealed significant
   differential expression of genes (DEGs) in VDC‐treated cells compared
   to the control cells at various time points. Specifically, 1747, 470,
   and 828 genes were upregulated on days 3, 5, and 8, respectively,
   whereas 1074, 446, and 989 genes were downregulated on the same days
   (Figure [85] 2A). To illustrate the function of DEGs, gene ontology
   (GO) analysis was performed on three levels: molecular function,
   cellular component, and biological process. The VDC‐treated group
   showed significant enrichment in pathways related to the Wnt signaling
   pathway, bone mineralization, ossification, and bone morphogenesis, in
   comparison to the control group on days 3, 5, and 8. Notably, the
   functions of the Wnt signaling pathway throughout the process, along
   with its optimal P and richness, are vital (Figure [86]2B). Kyoto
   Encyclopedia of Genes and Genomes (KEGG) analysis was conducted to
   explore the regulated pathways involving DEGs. The mechanisms of
   anti‐osteoporosis were predominant in the Wnt, PIK3‐AKT, Notch, mTOR,
   and MAPK signaling pathways (Figure [87]2C). Based on the results from
   GO, KEGG, and disease‐related biological knowledge analyses, we
   postulated that the Wnt signaling pathway may be the primary mediator
   of the VDC treatment effects on skeletal diseases. The Venn diagram
   (Figure [88]2D) was used to identify the intersection of DEGs within
   the Wnt signaling pathway at days 3, 5, and 8. The analysis revealed
   that Hivep3, Dkk2, Camk2b, Wnt4, and Wif1 were consistently
   differentially expressed across these time points. During osteogenic
   differentiation, Hivep3, Dkk2, Camk2b, Wnt4, and Wif1 show dynamic
   expression changes crucial for osteoblast regulation.^[ [89]^7 ,
   [90]^19 ^] Hivep3 influences ERK activity in Wnt signaling, affecting
   osteoblast differentiation. Dkk2 modulates bone homeostasis by
   inhibiting Wnt signaling via LRP5/6. Camk2b interacts with Wnt pathways
   through calcium signaling to regulate bone metabolism. Wnt4 promotes
   osteoblastogenesis by activating Wnt signaling, while Wif1 inhibits it
   by binding to Wnt ligands. Together, these genes regulate bone
   metabolism and contribute to bone mineralization, ossification, and
   morphogenesis through VDC‐mediated signaling, as shown by GO analysis.
   The heatmap illustrating the DEGs at the intersection of the three‐time
   points is presented in Figure [91]2E. VDC treatment led to upregulation
   of Dkk2, Camk2b, Wnt4, and Wif1, while inhibiting the expression of
   Hivep3. GO and KEGG analysis revealed significant activation of the Wnt
   signaling pathway on days 3, 5, and 8, suggesting it may mediate VDC
   treatment effects on skeletal diseases. After deduplicating DEGs (37 on
   day 3, 8 on day 5, and 13 on day 8), a protein‐protein interaction
   (PPI) network diagram was constructed to investigate the crucial
   protein targets within the Wnt signaling pathway at these specified
   time points. The findings from the analyses of the PPI regulatory
   networks (Figure [92]2F) revealed interactions among potential
   therapeutic targets at these time points, comprising 47 nodes and 280
   edges. Notably, Hivep3 exhibited the strongest correlation and occupied
   the central position within the network graph. Consequently, Hivep3 was
   identified as the primary target of VDC within the Wnt signaling
   pathway.

Figure 2.

   Figure 2
   [93]Open in a new tab

   VDC induces osteogenesis via Wnt signaling pathway. A) Volcano plots
   showing the DEGs between control and VDC groups of 3, 5 and 8 days.
   Blue indicates downregulated genes; Red indicates upregulated genes. B)
   Gene Ontology (GO) enrichment analyses of the DEGs in RNA‐seq data
   showing the key upregulated biological process in VDC groups of 3, 5,
   and 8 days. C) The KEGG analysis results. D) Venn diagram showing the
   overlapping DEGs of the Wnt signaling pathway at 3, 5, and 8 days. E)
   Heatmap illustrating the overlap DEGs of the Wnt signaling pathway at
   3, 5, and 8 days. F) The PPI regulatory network of the Wnt signaling
   pathway.

2.3. VDC Stimulates the Expression of Osteoblast‐Specific Proteins by
Activating the Wnt/SHN3 Signaling Pathway

   Our previous studies demonstrated that SHN3 negatively regulates bone
   mass by inhibiting ERK activity while also activating downstream
   β‐catenin via GSK‐3β to promote osteogenesis.^[ [94]^20 ^] To validate
   that the Wnt/SHN3 signaling pathway was activated through VDC treatment
   in BMSCs, we performed immunofluorescence staining to detect the level
   of β‐catenin and its co‐localization with osteocalcin (OCN) in BMSCs.
   Immunofluorescence results showed that the OCN and β‐catenin levels
   were significantly upregulated in BMSCs treated with VDC, and it did
   not dependent on concentration or time (Figure [95] 3A,B).

Figure 3.

   Figure 3
   [96]Open in a new tab

   VDC stimulates the expression of osteoblast‐specific proteins by the
   Wnt/SHN3 pathway. A) Representative images and corresponding
   quantitative analysis of immunofluorescence staining results of OCN and
   β‐catenin treated with 1, 5, and 10 µm VDC for 5 days (n = 5). Scale
   bar, 50 µm. B) Representative images and quantitative analysis of
   immunofluorescence staining of OCN and β‐catenin treated with VDC
   (5 µm) for 3 days, 5 days, and 8 days (n = 5). Scale bar, 50 µm. C)
   Representative images of western blot showing the expression levels of
   β‐catenin, NF‐κB, GSK‐3β, phosphorylated (p)‐GSK‐3β, ERK1/2, p‐ERK1/2,
   Slit3 and normalized to the expression of GAPDH. Under induction
   conditions (50 µg mL^−1 ascorbic acid and 5 mm β‐glycerophosphate),
   MC3T3‐E1 cells were cultured with VDC (1, 5, 10 µm) or 0.1% DMSO
   (control group) for 5 days. The solvent control group (Control) was
   treated with 0.1% DMSO (n = 3). D) Quantitative analysis of the
   relative protein expression for β‐catenin, NF‐κB, p‐GSK‐3β and p‐ERK1/2
   and Slit3. Data represent mean ± SD, ns: not significant, *p < 0.05,
   **p < 0.01, ***p < 0.001, ****p < 0.0001 versus the Control group by
   one‐way ANOVA with Tukey's post‐hoc test.

   Additionally, SHN3 controls GSK3β activity and the levels of β‐catenin
   and via ERK in Wnt signaling^[ [97]^5 , [98]^7 ^] And the NF‐κB
   controls the differentiation or activity of osteoblasts.^[ [99]^21 ^]
   Therefore, we used western blot analysis to assess the effects of VDC
   on the expression levels of β‐catenin, NF‐κB, GSK‐3β,
   phosphorylated(p)‐GSK‐3β, ERK1/2, p‐ERK1/2, Slit3, and SHN3. Treatment
   with VDC for 5 days in BMSCs resulted in the upregulation of β‐catenin,
   p‐GSK‐3β, and p‐ERK1/2 expression levels, while the expression levels
   of NF‐κB, total ERK1/2, and GSK‐3β remained unchanged
   (Figure [100]3C,D). And it also revealed that VDC significantly
   inhibits SHN3 protein expression, thereby validating its mechanistic
   role in regulating SHN3 activity (Figure [101]S4, Supporting
   Information). The significant modulation of these proteins by VDC
   underscores its role in regulating osteogenic activity via the Wnt/SHN3
   signaling pathway.

2.4. VDC‐Mediated Silencing of SHN3 Promotes Osteogenic Activity

   To further validate the regulatory effect of VDC on Hivep3, we used
   quantitative polymerase chain reaction (qPCR) to assess the expression
   level of Hivep3 at time points of 0, 6, 12, 24, and 48 h, as well as at
   3, 5, and 8 days. Under osteogenic induction conditions, the expression
   profiles of the Hivep3 gene initially increased gradually, before
   sharply decreasing in the early stages of osteogenic differentiation.
   This observation demonstrated the negative regulatory role of Hivep3 on
   the osteogenic differentiation process of osteoblasts. Upon exposure to
   VDC (5 µm) for 3, 5, and 8 days, VDC exhibited a time‐dependent
   inhibition of Hivep3 expression, while promoting the expression of
   alkaline phosphatase (Alpl, marker gene for early osteogenic
   differentiation) and osteocalcin (Bglap, marker gene for late
   osteogenic differentiation) in a time‐dependent manner (Figure [102]
   4A).

Figure 4.

   Figure 4
   [103]Open in a new tab

   SHN3 is a key target for VDC‐induced osteoblast activity. A) The effect
   of VDC on the expression levels of key genes Hivep3, Alpl, and Bglap
   among MC3T3‐E1 cells. The VDC (5 µm) was added to MC3T3‐E1 cells and
   cultured for the indicated durations (0 h, 6 h, 12 h, 24 h, 48 h, 3
   days, 5 days, 8 days) under induction conditions (n = 6). B) A diagram
   showing the BMSCs extraction procedure. (C) Representative images of
   alizarin red S staining assay for osteogenic mineralization in BMSCs.
   The VDC (5 µm) was added to BMSCs for induction for 8 days under
   osteogenic induction conditions, which were subjected to alizarin red
   staining and microscopic images were recorded. D) Quantitative analysis
   of Alizarin red staining (n = 5). E) qPCR validation of the knockout
   efficiency of SHN3 and the inhibitory effect of VDC on SHN3. The VDC
   (5 µm) was added in BMSCs for induction 5 days under osteogenic
   induction conditions (n = 6). F) The laplace diagram was produced
   directly by the SAVES server. G) The binding modes of VDC and SHN3 in a
   3D workspace cartoon, as well as their detailed enlarged images. H) 2D
   combination mode diagram of VDC with SHN3. The H‐bonding was expressed
   as a pink arrow. Normal culture medium and 0.1% DMSO served as the
   negative control group (NC) and solvent control group (control),
   respectively. I) Molecular dynamics simulation analysis of SHN3 and
   VDC. Root mean square deviation (RMSD) plot of VDC with respect to SHN3
   protein. Data represent mean ± SD, ns: not significant, *p < 0.05,
   **p < 0.01, ***p < 0.001, and ****p < 0.0001 versus the DMSO group by
   one‐way ANOVA with Tukey's post‐hoc test.

   To further explore the role of SHN3 in VDC‐induced osteoblast
   differentiation, BMSCs were extracted from 3‐week‐old SHN3^−/− knockout
   and wild‐type (WT) mice and cultured in vitro. The cells were treated
   with 5 µm VDC. qPCR and alizarin red S staining analyses were performed
   (Figure [104]4B). In the absence of SHN3, both KO control and WT‐VDC
   significantly promoted the osteogenic mineralization of BMSCs compared
   to WT control. However, VDC's osteogenic mineralization effect on BMSCs
   was diminished without SHN3, suggesting that SHN3 is crucial for
   VDC‐induced osteoblast mineralization (Figure [105]4C–E).

   In a study by Jones,^[ [106]^7 ^] a D‐domain motif within SHN3 was
   found to mediate the interaction with and inhibition of ERK activity
   and osteoblast differentiation. However, the 3D structure of SHN3
   remains currently unknown due to its high molecular weight (>260 kDa)
   and low expression level. To explore whether VDC can bind with SHN3, we
   simulated SHN3 protein structure using the de novo prediction method.
   We focused on amino acids 327–1183 (NCBI Reference Sequence:
   NP_0 011 21186.1) of SHN3 for prediction and evaluated the model
   quality by constructing Laplace plots. The Laplace diagram, generated
   through the SAVES server, indicated that 90.7% of the protein residues
   fell within the allowed red and yellow maximum regions, demonstrating
   high reliability for molecular docking (Figure [107]4F). Our analysis
   focused on the protein‐ligand interactions, identifying and
   categorizing all functional residues based on their specific
   interactions. Various groups of residues were found to participate in
   the interactions between SHN3 and VDC, including the hydrogen bond
   formed by GLU863 of SHN3 with the ligand. These interactive forces
   generated a binding energy of −7.7 kcal mol^−1 for the protein‐ligand
   complex, indicating it was a strong binding force (Figure [108]4G,H).
   Moreover, molecular dynamics simulations to investigate the
   interactions between VDC and SHN3.The root mean square deviation (RMSD)
   plot illustrates the SHN3 protein remained relatively stable, whereas
   the ligand exhibited fluctuations before stabilizing in the final RMSD.
   Post‐dynamic simulation, a more stable binding conformation was
   established on the original basis, indicating enhanced conformational
   and interaction stability of the SHN3 and VDC complexes
   (Figure [109]4I). In comparison to the initial binding conformation,
   VDC has experienced a certain degree of deviation (Figure [110]S5A,
   Supporting Information). This shift suggests dynamic adjustments in the
   ligand's position relative to the protein during the simulation. The
   Ligand Root Mean Square Fluctuation (L‐RMSF) is a valuable metric for
   characterizing positional changes of ligand atoms, while the Root Mean
   Square Fluctuation (RMSF) provides insights into local flexibility
   along the protein backbone. As shown in Figure [111]S5B (Supporting
   Information), the RMSF values of amino acid residues involved in
   interactions with the small molecule (highlighted with green lines) are
   generally higher, indicating significant conformational rearrangements
   during binding. Consistent with this observation, the L‐RMSF values of
   the small molecule's atoms are also relatively high (Figure [112]S5C,
   Supporting Information), further supporting the notion of substantial
   mobility and structural adaptation during the interaction. Throughout
   the simulation, protein‐ligand interactions within 5 Å of VDC were
   monitored, as illustrated in the Interaction Fractions plot (Figure
   [113]S5D, Supporting Information). These interactions include hydrogen
   bonds, hydrophobic contacts, and water‐mediated bridges, underscoring
   the complexity and diversity of the binding mechanism. Notably, several
   surrounding amino acids—GLN 438, LEU 440, GLU 863, PRO 873, LYS 880,
   GLU 883, GLU 935, HIS 984, and GLU 987—played critical roles in
   stabilizing the binding of VDC. This stability supports the hypothesis
   that VDC effectively targets SHN3, thereby modulating the Wnt/SHN3
   pathway and promoting osteogenic differentiation.

2.5. VDC Ameliorates OVX‐Induced Bone Loss and Promotes Fracture Healing

   The absence of SHN3 effectively restores bone remodeling processes,
   mitigating bone loss associated with osteoporosis, facilitating
   fracture repair, and treating OI.^[ [114]^8 , [115]^9 , [116]^10 ,
   [117]^11 , [118]^12 , [119]^22 ^] To further investigate whether
   VDC‐mediated silencing of SHN3 offers therapeutic benefits for these
   osteolytic diseases, we assessed the osteogenic effects of VDC in vivo
   using ovariectomized (OVX), fractured, and OI mouse models.

   VDC was tested for its effects on estrogen‐induced osteoporosis in mice
   after ovariectomy. The schematic diagram of the OVX model is shown in
   (Figure [120] 5A). C57BL/6J mice were injected with VDC (5 mg kg^−1) or
   DMSO for 6 weeks, with no significant weight changes or deaths observed
   (Figure [121]S6A, Supporting Information). HE staining revealed no
   abnormalities in the liver, spleen, and kidneys, indicating VDC was
   safe for use (Figure [122]S6B, Supporting Information). The OVX surgery
   resulted in uterine atrophy and weight loss, successfully modeling
   osteoporosis (Figure [123]S6C,D, Supporting Information).

Figure 5.

   Figure 5
   [124]Open in a new tab

   VDC delays bone loss in a mouse model of osteoporosis. A) Schematic
   diagram of the mouse OVX model. B) 3D reconstructed images of mouse
   femur after Micro‐CT scanning. Scale bar, 1 mm (long femur) and 100 µm
   (frontal view, superior norma and cortical bone). C) Quantitative
   measurements of bone microstructure‐related parameters: BV/TV, Tb.N,
   Tb.Sp, Tb.Th, and Cs.Th (n = 8). D) Representative images and BV/TV
   quantitative of Von Kossa stained bone tissue in the fifth lumbar
   spine. Scale bar, 250 µm. E) Representative fluorescence images of the
   double‐labeled resin sections of vertebral bone with calcein. Scale
   bar, 25 µm (n = 5). F) Quantitative analysis of the dynamic bone
   reconstruction: MS/BS, MAR, and BFR/BS. G) Representative images of the
   decalcified bone stained with Toluidine blue. Scale bar, 100 µm. H)
   Quantitative analysis of Toluidine blue staining results for the lumbar
   spine and femur (n = 5). I) Representative images of the decalcified
   bone stained with TRAP staining. Scale bar, 100 µm. J) Quantitative
   measurements of TRAP staining of the lumbar spine and femur (n = 5).
   Data represent mean ± SD, ns: not significant, *p < 0.05, **p < 0.01,
   ***p < 0.001, and ****p < 0.0001 versus the OVX group by one‐way ANOVA
   with Tukey's post‐hoc test.

   CT scan data of femurs indicated that the OVX+VDC group exhibited
   higher bone volume fraction (BV/TV) and trabecular bone number (Tb. N)
   compared to the OVX‐only group (Figure [125]5B,C). Moreover, Von Kossa
   staining of the 5th lumbar spine bone tissue showed increased bone mass
   with VDC treatment (Figure [126]5D). Fluorescence double labeling
   results demonstrated improvements in the mineralized surface to bone
   surface ratio (MS/BS), mineral appositional rate (MAR), and bone
   formation rate to bone surface ratio (BFR/BS) in the OVX+VDC group
   (Figure [127]5E,F). These dynamic bone parameters demonstrate an
   overall increase in bone formation after VDC treatment. Additionally,
   Toluidine blue (Figure [128]5G) and TRAP staining (Figure [129]5I)
   revealed that VDC enhanced the osteoblast surface/bone surface ratio
   (Ob.S/BS) in the 5th lumbar vertebrae and femur (Figure [130]5H), with
   no significant effect on the osteoclast count/bone surface ratio
   (N.Oc.S/BS) (Figure [131]5J). Collectively, these static and dynamic
   bone parameters suggest that VDC stimulated bone formation in vivo by
   enhancing osteoblast activity, thereby preventing bone loss in OVX
   mice.

   To evaluate the pharmacological effects of VDC on fractures, we
   established an open fracture model in 8‐week‐old female C57BL/6J mice.
   The mice then underwent postoperative administration of VDC‐containing
   matrix gel (5 mg kg^−1) or 3.5% DMSO matrix gel with injections every 2
   days for 14 days (Figure [132] 6A). CT scans showed that VDC
   significantly increased callus volume and bone mass (Figure [133]6B,C).
   The results of Safranin O/Fast Green staining further demonstrated that
   VDC increased cartilage and bone areas, underscoring its potential to
   promote fracture healing (Figure [134]6D,E). Additionally, results from
   toluidine blue and TRAP staining of fractures demonstrated that VDC
   primarily exerts its effects on osteoblasts rather than osteoclasts
   (Figure [135]6F–I).

Figure 6.

   Figure 6
   [136]Open in a new tab

   VDC promotes fracture healing. A) A schematic diagram of the mouse
   fracture model. B) Representative 3D reconstructed images obtained
   through Micro‐CT scanning of mouse femoral fractures. Scale bar, 1 mm.
   C) Quantification of callus volume and BV/TV of fractures (n = 5). D)
   Representative images showing the results of the Safranin O/Fast Green
   staining. Scale bar, 1 mm (Overall view) and 200 µm (enlarged view). E)
   Quantitative analysis of bone area, soft tissue area, and chondrocyte
   area (n = 5). F) Representative images of the decalcified bone stained
   with Toluidine blue. Scale bar, 50 µm (n = 5). G) Quantitative
   measurements of Toluidine blue staining (n = 5). H) Representative
   images of decalcified bone subjected to TRAP staining. Scale bar,
   50 µm. I) Quantitative measurements of TRAP staining (n = 5). Data
   represent mean ± SD, ns: not significant, *p < 0.05, **p < 0.01,
   ***p < 0.001, and ****p < 0.0001 versus the control group
   (0.1%DMSO+PBS) by an unpaired two‐tailed Student's t‐test.

2.6. Targeted Delivery of VDC for Treating OI

   To investigate the therapeutic potential of VDC in treating OI, a
   Col1a2 ^oim/oim mice model was used. This model has been demonstrated
   to be appropriate for cases of moderate to severe OI.^[ [137]^13 ,
   [138]^23 ^] Due to the severity of OI mice (OIM), achieving the desired
   therapeutic effect can be challenging. To enhance tissue accumulation
   and the efficacy of VDC, we explored the development of a
   bone‐targeting system.

   Cellular nanovesicles (NVs), fabricated through direct secretion or
   physicochemical methods, are increasingly favored for drug delivery due
   to their excellent biocompatibility.^[ [139]^24 ^] Exosomes, a
   naturally secreted NV, are extensively investigated due to their
   structure, composition, and unique role in intercellular communication.
   However, despite their promising therapeutic potential, the clinical
   application of exosomes remains constrained by complex production
   processes and low yields. However, advancements in nanotechnology have
   enabled the production of NVs through various physical and chemical
   methods, resulting in significantly higher yields compared to naturally
   secreted exosomes.^[ [140]^25 ^] Moreover, effective targeted
   modifications of these NVs enhance delivery efficiency to specific
   organs and maximize their carrier potential.^[ [141]^26 ^] In this
   study, we used SDSSD (Ser, Asp, Ser, Ser, Asp), a proven bone‐targeting
   peptide, to engineer BT‐NVs for the precise delivery of VDC.^[ [142]^27
   ^]

   BT‐NVs extracted from MC3T3‐E1 cells were produced through ultrasonic
   lysis, followed by density gradient centrifugation, and then
   hydrophobic modification. Subsequently, VDC was loaded into BT‐NV using
   ultrasonic to prepare BT‐NVs‐VDC (Figure [143] 7A). The
   characterization of BT‐NVs‐VDC was then performed. Transmission
   electron microscopy (TEM) revealed that NVs, BT‐NVs, and BT‐NVs‐VDC
   exhibited cup‐shaped morphology. Notably, the BT‐NVs‐VDC group
   exhibited increased fluorescence compared to the other groups,
   indicating successful VDC loading. Dynamic light scattering (DLS)
   analysis showed that the NVs were NVs with an average size of 127.60 nm
   and a zeta potential of ≈−13.87 mV. Additionally, no significant
   changes were observed in size or zeta potential following subsequent
   modifications (Figure [144]7B,C). The VDC‐loading efficiency was
   ≈28.67% and ≈82.51% of VDC was released from PBS after 48 h
   (Figure [145]7D,E). The stability of BT‐NVs in vitro and in vivo was
   assessed using PBS supplemented with 10% FBS. The results indicated
   that BT‐NVs remained stable over 8 days (Figure [146]7F; Figure
   [147]S7, Supporting Information).

Figure 7.

   Figure 7
   [148]Open in a new tab

   Preparation and characterization of BT‐NVs‐VDC. A) A schematic
   representation of the preparation process of BT‐NVs‐VDC. B) Size
   distribution and TEM images of NVs, BT‐NVs and BT‐NVs‐VDC. C) The
   average surface zeta potential of NVs, BT‐NVs and BT‐NVs‐VDC (n = 5).
   D) Loading efficiency of VDC in BT‐NVs (n = 6). E) In vitro drug
   release profiles under PBS (pH 7.4) conditions (n = 3). F) The
   stability of BT‐NVs. Changes in size and zeta potential of BT‐NVs in
   PBS during the 8 days of treatment (n = 3). G–I) Flow cytometry
   analysis of DiR‐labeled NVs or BT‐NVs binding to MC3T3‐E1, RAW264.7 and
   BMSC cells after a 1 or 4 h incubation (n = 3). J,K) Representative
   confocal microscopy images of MC3T3‐E1 cells incubated with DiR‐labeled
   NVs or BT‐NVs for 4 h and the mean fluorescence intensity analysis.
   Scale bar, 20 µm. (n = 5). L) Cytotoxicity of NVs, BT‐NVs, and
   BT‐NVs‐VDC against MC3T3‐E1 cells after a 24 h incubation at indicated
   concentrations (n = 5). Data represent mean ± SD, ns: not significant,
   *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 by one‐way
   ANOVA with Tukey's post‐hoc test.

   We then investigated the efficiency of the bone‐targeting peptide by
   comparing the uptake of DIR‐labeled NVs and BT‐NVs in MC3T3‐E1 cells
   after incubation for 1 or 4 h. Flow cytometry revealed stronger
   fluorescent signals in the BT‐NVs group compared to the NVs group
   (Figure [149]7G,H). However, no significant differences were observed
   in BMSCs and RAW264.7 cells (Figure [150]7I). Subsequently,
   osteoblast‐specific uptake was further validated using confocal laser
   scanning microscopy (CLSM) (Figure [151]7J,K). CCK‐8 assays indicated
   that these three NVs exhibited optimal cytocompatibility
   (Figure [152]7L). In summary, we have developed an effective
   bone‐targeting delivery system.

   Before OIM treatment, we assessed the in vivo distribution effects of
   NVs and BT‐NVs. Analysis of the IVIS Spectrum revealed that the
   bone‐targeted modification was effective, increased accumulation in
   bone tissues such as the femurs (Figure [153] 8A,B; Figure [154]S8,
   Supporting Information). Based on these observations, we proceeded with
   in vivo experiments (Figure [155]8C). As anticipated, the OIM exhibited
   a severe osteopenic phenotype and spontaneous bone fractures compared
   to the WT group during the 4‐week treatment period. Additionally, the
   therapeutic effects of free VDC and NVs‐VDC were limited, due to
   insufficient effective concentrations at the target organs caused by
   blood diffusion during long‐term circulation. Conversely, BT‐NVs‐VDC
   significantly mitigated the osteopenic symptoms and reduced spontaneous
   fracture occurrences in OIM (Figure [156]8D‐G; Figure [157]S9,
   Supporting Information). Subsequent studies further validated the
   effect of VDC on osteoblasts in OIM (Figure [158]8H–J). Collectively,
   these bone parameters indicate that VDC promotes bone formation by
   enhancing osteoblast activity in vivo, thereby mitigating bone loss and
   reducing the incidence of spontaneous fractures in OIM.

Figure 8.

   Figure 8
   [159]Open in a new tab

   VDC delays bone loss in a mouse model of osteogenesis imperfecta and
   reduces the occurrence of spontaneous fractures. A) The biodistribution
   of DiR‐labelled NVs or BT‐NVs at a specified time points after
   intravenous injection. B) Radiant efficiency of DiR‐labelled NVs or
   BT‐NVs in femurs (n = 3). C) Schematic diagram of the mouse OI model.
   D) Reconstruction of µCT data reflected spontaneous bone fracture in
   OIM groups. Scale bar, 1 mm. E) The number of spontaneous bone fracture
   femurs in the OIM groups. F) Four weeks after intravenous injection
   into 1‐month‐old male mice, femoral trabecular bone mass was evaluated
   with µCT. Scale bar, 100 µm. G) Representative 3D reconstructions and
   corresponding quantitative analysis are presented (n = 5). H)
   Representative images of decalcified bone subjected to the Toluidine
   blue staining. Scale bar, 50 µm. I) Quantitative measurements of
   Toluidine blue staining in the femurs, (n = 5). J) Representative
   images of the decalcified bone stained with TRAP staining. Scale bar,
   50 µm. K) Quantitative measurements of TRAP staining in the femurs
   (n = 5). Data represent mean ± SD, ns: not significant, *p < 0.05,
   **p < 0.01, ***p < 0.001, and ****p < 0.0001 by one‐way ANOVA with
   Tukey's post‐hoc test.

3. Discussion

   The skeletal system of healthy adults is regulated by a dynamic
   equilibrium maintained by osteoclasts and osteoblasts. Osteoblasts play
   a crucial role in bone reconstruction, and their quantity as well as
   functionality is closely linked to various osteolytic diseases,
   including osteoporosis, osteonecrosis, rheumatoid arthritis, and
   osteoarthritis. Based on current literature reviews^[ [160]^15 ,
   [161]^16 , [162]^28 ^] on marine natural products exhibiting bone
   metabolism activity, we found that only 36 MNPs exhibited inhibitory
   effects on osteoclastogenesis, with only 14 MNPs demonstrating the
   ability to induce osteogenic differentiation. However, due to the
   limited yield, low bioavailability, and unclear mechanisms of these
   marine compounds, their clinical translation is limited, with no
   natural marine drugs available for clinical treatments. In this study,
   the active target compound VDC is investigated for the first time to
   assess its effects on osteoblast activity both in vivo and in vitro,
   along with elucidating its mechanism of action. The high yield (1.6 g/8
   kg), strong osteogenic activity, definite molecular mechanisms, and
   therapeutic targets of VDC demonstrate its promising clinical
   application prospects. Additionally, VDC provides a marine molecular
   tool for combating bone metabolic diseases.

   Wnt signaling regulates liver metabolism and regeneration, lung tissue
   repair and metabolism, hair follicle renewal, hematopoietic system
   development, and osteoblast maturation. In recent years, potential
   regulatory factors of the Wnt/β‐catenin signaling pathway have been
   identified, including SHN3, Twa1, FOXKs, ICAT, and Kdm2a/b. In
   osteoblasts, SHN3 plays various roles in regulating bone homeostasis by
   inhibiting osteoblast activity. These findings suggest that traditional
   osteoblast‐targeted therapies could be effective for treating low bone
   mass diseases and highlight the Wnt/SHN3 pathway as a promising target
   for inducing therapeutic bone metabolism responses. Notably, we used de
   novo sequencing to predict the large protein structure of SHN3 and
   determined the binding mode and energy between VDC and SHN3, providing
   a promising strategy for future research on molecular inhibitors of
   this protein. In vitro experiments of SHN3 deficiency have demonstrated
   that SHN3 is a key target for VDC‐induced osteoblasts, marking the
   first study to identify a natural small molecule ligand for SHN3.

   However, this study was characterized by several limitations. First,
   there is potential to further optimize the material design. For
   example, engineering modifications to the NVs could help regulate the
   osteogenic microenvironment, or functional groups on VDC could be
   further modified to enhance its pharmacological efficacy. This approach
   could potentially inhibit osteoclast formation while simultaneously
   promoting osteogenic activity, achieving a double effect.^[ [163]^29 ^]
   However, it may introduce challenges associated with drug loading,
   release, as well as the potential loss of VDC active groups. Therefore,
   for future translational research, a direct NVs formulation is advised,
   with the VDC providing a promising foundation for developing drugs for
   bone diseases. Second, due to the unavailability of a SHN3 antibody for
   western blotting, our investigation on the impact of VDC on SHN3
   protein was limited. This limitation arises from the significant size
   of SHN3, consisting of 2348 amino acids. Additionally, the low
   expression levels of SHN3 hindered the acquisition and purification of
   the protein in substantial quantities. Consequently, the 3D structure
   of SHN3 remains unclear. While we have attempted to predict the
   structure of SHN3, our findings necessitate further experimental
   evidence to validate, such as cryoelectron microscopy. Before assessing
   the SHN3 protein, procurement of a substantial quantity of its
   expression remains a significant challenge. Our investigation reveals
   that XCL‐5 can significantly promote the mRNA transcription of SHN3
   (Figure [164]S1, Supporting Information), presenting a promising avenue
   for further exploration. Finally, the conventional tail vein injection
   method used in OIM heightens the risk of spontaneous fractures.
   Specifically, this risk can be mitigated by using anesthetic induction,
   or delivering modified particles through the nasal‐brain pathway.

   Collectively, the quinolinone alkaloid VDC, obtained from
   deep‐sea‐derived fungus, significantly enhances osteogenic
   differentiation through the Wnt/SHN3 signaling pathway. It provides
   potential therapeutic benefits for osteoporosis treatment and fracture
   healing. Through this study, we also engineered SDSSD‐modified
   osteoblast membrane NVs containing VDC for targeted drug delivery to
   osteoblasts, offering a potential treatment option for OI. Our findings
   highlight the potential of using VDC to target the Wnt/SHN3 pathway for
   treating osteoporosis, fractures, and OI in clinical settings.

4. Experimental Section

Materials

   Alpha‐modified minimal essential medium (α‐MEM), fetal bovine serum
   (FBS), and penicillin/streptomycin were procured from Gibco, while
   β‐glycerophosphate (S0942), DMSO (34 869), Alizarin Red S (A5533), Oil
   Red O (1320‐06‐5) Alcian blue (33864‐99‐2) were obtained from Sigma.
   Toluidine blue (6586‐04‐5) was acquired from Macklin. Na[2]CO[3]
   (497‐19‐8), NaHCO[3] (144‐55‐8), and MgCl[2] (7791‐18‐6) were purchased
   from Sinopharm Chemical Reagent Co., Ltd. Ascorbic acid (60374ES60) was
   sourced from Yeasen. Vazyme's HiScript II Q RT SuperMix ([165]R22301),
   trizol (R40101AA), and ChamQ Universal SYBR qPCR Master Mix
   ([166]Q71102) were gained from Vazyme for use in the study. The
   antibodies β‐Catenin (15B8, 1:1000), GSK‐3β (D5C5Z, 1:1000),
   phospho‐GSK‐3β (Ser9) (D3A4, 1:1000), NF‐κB (D14E12, 1:1000), ERK1/2
   (137F5, 1:1000), phospho‐ERK1/2 (D13.14.4E, 1:2000) and GAPDH (D4C6R,
   1:1000) were acquired from Cell Signaling Technology. The SHN3 antibody
   (HPA005728, 1:200) was purchased from Merck, while the Slit3 antibody
   (AF3629, 1:400) was obtained from R&D. Cell Counting Kit‐8 (CCK‐8,
   CT01A) was supplied by Cellcook. The primer sequences implemented in
   the qPCR were furnished by Origene. The organic reagents, such as
   chloroform, isopropanol, ethanol, and others, were all obtained from
   Sinopharm.

Cell and Animals

   The MC3T3‐E1 cell line (clone 4; CRL‐2593) was derived from the
   American Type Culture Collection, Rockville, MD, USA. Three‐ to
   six‐week‐old C57BL/6J mice were used to obtain BMSCs and bone marrow
   monocytes (BMMs). MC3T3‐E1 cells and BMSCs were plated in a normal
   culture medium (α‐MEM, 10% FBS, and 1% penicillin/streptomycin) and
   incubated at 37 °C in a 5% CO[2] incubator. BMMs were cultured in a
   normal culture medium with 25 ng mL^−1 M‐CSF at 37 °C and 5% CO[2].
   Passage‐2 BMSCs, passage‐0 BMMs, and passage‐5 MC3T3‐E1 cells were used
   in this experiment.

   The C57BL/6J mice were obtained from Gempharmatech Co., Ltd. Col1a2
   ^oim/oim mice were obtained from the Jaxson Laboratory (B6C3Fe
   a/a‐Col1a2^oim /J, Stock No: 0 01815, Bar Harbor, ME, USA). SHN3 ^−/−
   mice were described in the previous study.^[ [167]^9 ^] All mice were
   housed in up to five per cage under a 12 h light‐dark cycle with chow
   ad libitum in the laboratory animal center at Xiamen University. All
   mouse experiments were handled according to the protocols approved by
   the Institutional Animal Care and Use Committee of Xiamen University
   Laboratory Animal Center.

Source and Identification of VDC

   All 251 compounds from deep‐sea‐derived fungi were provided by
   professor Xianwen Yang. VDC was isolated from the deep‐sea‐derived
   fungus Penicillium solitum MCCC 3A00215. It was obtained as a white
   amorphous powder and identified by comparison of NMR and HRESIMS data
   with literature references. HRESIMS m/z 252.0744 [M − H]^−; ^1H‐NMR