Abstract

   The brain-bone regulatory system regulates skeletal homeostasis via
   bioactive neuropeptides, yet the underlying mechanism remains elusive.
   Here, we report the role of the neuropeptide VF (NPVF, VPNLPQRF-NH[2])
   in enhancing both angiogenesis and osteogenesis in a rat skeletal
   system and the potential pathways involved. An in vitro study revealed
   that NPVF not only promotes migration and angiogenesis of human
   umbilical vein endothelial cells (HUVECs) by activating NPFFR1, which
   leads to upregulation of miR-181c-3p and downregulation of Argonaute1
   (AGO1), but also mediates osteogenic differentiation of bone
   mesenchymal stem cells (BMSCs) via the Wnt/β-catenin signaling pathway.
   To improve the stability and bioavailability and thus efficacy of NPVF
   as a promoter of in vivo bone regeneration, we genetically engineered
   amyloid-NPVF-fusion proteins and utilized them as self-assembling
   nanofiber coatings to treat bone defects in a rat calvarial defect
   model. We found that a porous hydroxyapatite scaffold loaded with the
   NPVF peptide-fused amyloid coating substantially enhanced angiogenesis
   and site-specific fresh bone in-growth when implanted in calvarial
   defects. Taken together, our work uncovered a previously undefined
   crosstalk between the brain and bone by unveiling the role of NPVF in
   bone tissue and demonstrated a viable method for promoting bone tissue
   repairs based upon self-assembling NPVF-containing protein coatings.

   Subject terms: Peptides, Preclinical research
     __________________________________________________________________

   The neuropeptide NPVF stimulates bone formation via the Wnt signaling
   pathway and hydroxyapatite scaffolds coated with NPVF stimulates
   angiogenesis and osteogenesis in a rat calvaria defect model.

Introduction

   Bone homeostasis involves bone formation by osteoblasts and bone
   destruction by osteoclasts; these processes are interconnected and
   tightly regulated, assuring the maintenance of skeletal health^[34]1.
   Bone homeostasis has been found to be closely associated with the
   brain-bone regulatory system, primarily through various bioactive
   neuropeptides, such as substance P (SP), neuropeptide Y (NPY), and
   calcitonin gene-related peptide (CGRP)^[35]2–[36]4. Recent studies have
   revealed that these neuropeptides are closely involved in osteogenesis
   and angiogenesis^[37]2,[38]3,[39]5,[40]6, two hallmark events in bone
   formation^[41]7,[42]8.

   The neuropeptide FF (NPFF) system, first identified in bovine brain
   extract in 1985^[43]9,[44]10, has been found to be involved in a
   variety of physiological processes, including food intake, blood
   pressure regulation, memory, insulin release, neural regeneration,
   metabolic disease, and cardiovascular activity^[45]11–[46]16, via
   activation of a G[i/o] protein-coupled receptor located in the nervous
   system^[47]17,[48]18. Despite these advances, no previous studies have
   reported a connection between the NPFF system and bone homeostasis.
   However, as an increasing number of neuropeptides have been recently
   discovered to participate in the brain-bone regulatory system, the
   tantalizing possibility has been raised that the NPFF system may have a
   largely ignored yet important physiological role in the skeletal
   system.

   In this work, using NPVF (a representative of the NPFF family, it is
   named NPVF after the abbreviation for its first and last amino acid.)
   as a model system, we probed the possible roles of the NPFF family in
   both osteogenesis and angiogenesis during bone formation. We revealed
   that NPVF could indeed promote both osteogenesis and angiogenesis in a
   rat skeletal system (Fig. [49]1a). Building on these discoveries, we
   next explored the potential application of the NPVF peptide for in vivo
   bone regeneration. Because the application of peptide drugs is often
   hampered by their structural vulnerability in serum and rapid clearance
   via renal filtration^[50]19, we genetically engineered NPVF-amyloid
   protein fusions and utilized them as self-assembling nanofiber coatings
   to treat bone defects in a rat calvarial defect model (Fig. [51]1b). We
   found that implantation of a porous HAp scaffold containing NPVF
   peptide-fused amyloid coatings could significantly enhance angiogenesis
   and fresh site-specific bone in-growth in calvarial bone. In short, our
   studies clearly revealed a crosstalk between the brain and bone by
   unveiling the exact molecular role of NPVF in bone tissue formation,
   and demonstrated the feasiblilty of a cell-free and growth factor-free
   method for efficient treatment of bone repairs based upon
   self-assembling NPVF-containing protein coatings.

Fig. 1. Unveiling the role and mechanism of NPVF in mediating crosstalk
between the brain and bone and application of self-assembling NPVF-containing
protein coatings for in vivo bone defect repair.

   [52]Fig. 1
   [53]Open in a new tab

   a Schematic image depicting the promoting effect of the neuropeptide
   NPVF on both angiogenesis and osteogenesis. NPVF is an eight-amino acid
   neuropeptide secreted by the human hypothalamus, and it regulates
   various physiological processes including bone homeostasis maintenance
   by activating its G[i/o] protein-coupled receptor located in the
   nervous system (left panel). The amino acid sequence of NPVF is
   presented in the upper box. The right panel shows the signaling pathway
   by which NPVF promotes osteogenic differentiation of BMSCs and
   vascularization of HUVECs. NPVF binds preferentially to NPFFR1 in the
   central nervous system, promotes angiogenesis of HUVECs via the
   miR-181c-3p/AGO1 pathway, and mediates osteogenic differentiation of
   BMSCs via the Wnt/β-catenin signaling pathway. b Illustration of bone
   fracture repair through the implantation of a CsgA-NPVF
   nanofiber-coated porous hydroxyapatite scaffold in a rat calvarial
   defect model. The CsgA amyloid can self-assemble into nanofibers to
   form a robust proteinaceous coating on the surface of given substrate.
   Coupling the coat-forming property of CsgA with NPVF induced new bone
   formation. Displaying the engineered CsgA-NPVF functional nanofibers on
   the HAp scaffold substantially facilitated bone regeneration. The right
   panel shows the simulated structure of the CsgA-NPVF monomer revealed
   by molecular dynamics simulations.

Results

NPVF induces HUVEC migration and angiogenesis by activating NPFFR1

   Inspired by recent discoveries that neuropeptide systems might be
   involved in vascular development and angiogenesis in bone
   homeostasis^[54]5,[55]6,[56]20, we set out to assess whether the NPVF
   peptide, a well-known representative of the NPFF neuropeptide family,
   participates in bone homeostasis. Note that the NPFF system exerts a
   variety of physiological functions via selective activation of the
   G[i/o] protein-coupled receptors Neuropeptide FF Receptor 1 (NPFFR1)
   and NPFFR2^[57]21,[58]22, and NPVF is a highly selective ligand of
   NPFFR1^[59]21. Therefore, we first investigated the expression level of
   the NPFF1R in HUVECs by qPCR. The average Ct value for NPFFR1 was
   19.20 ± 0.22, indicating that NPFFR1 was highly expressed in HUVECs
   (Fig. [60]2a). Next, we explored the role of NPFFR1 in HUVECs using the
   selective agonist NPVF. A Cell Counting Kit-8 (CCK-8) assay showed that
   low dose of NPVF (0.01, 0.1 and 1 nM) had no toxic effect on the
   proliferation of HUVECs, while high dose of NPVF (10 nM) had a slight
   inhibition on the proliferation of HUVECs (Fig. [61]2b). Hence, low
   doses of NPVF (0.01, 0.1, and 1 nM) were chosen for subsequent
   experiments. We next studied the effect of NPVF on the migration
   capacity of HUVECs by performing wound healing (36 h) and transwell
   migration (24 h) assays. The wound healing area of HUVECs in the NPVF
   group was much larger than that in the control group, and the NPFFR1
   antagonist RF9 significantly counteracted the healing-promoting effect
   of NPVF (Fig. [62]2c-d). The results of the transwell assay also
   revealed that RF9 could neutralize the NPVF-induced migration of HUVECs
   (Fig. [63]2e-f). These results together indicated that NPVF could
   promote HUVEC migration and proliferation. Next, we explored the effect
   of NPVF on angiogenesis using a tube formation assay (6 h). As shown in
   Fig. [64]2g-i, the NPVF group had more loop structures and a higher
   number of branch points than the control group, while RF9 reversed the
   NPVF-induced promotion of HUVEC tube formation. Finally, qPCR and
   enzyme-linked immunosorbent assay (ELISA) were employed to detect the
   expression of angiogenesis-associated genes and proteins, respectively,
   in HUVECs. The qPCR results revealed that the expression of VEGF, EGF,
   and PDGF in HUVECs remarkably increased after NPVF treatment, and the
   addition of RF9 could offset this promotion effect.
   (Figs. [65]2j-[66]2l). ELISA analysis of the protein expression level
   of VEGF yielded the same conclusion. (Fig. [67]2m). In short, our
   experimental results suggested that NPVF significantly promotes
   migration and angiogenesis by activating NPFFR1 in HUVECs.

Fig. 2. NPVF promotes the migration and tube formation of HUVECs.

   [68]Fig. 2
   [69]Open in a new tab

   a The expression level of NPFFR1 in HUVECs. Results are presented as
   means ± S.E.M. of four independent experiments. b The proliferation of
   HUVECs treated with NPVF was evaluated by the CCK-8 assay. Results are
   presented as means ± S.E.M of four independent experiments. c, d Wound
   healing of HUVECs (36 h). Results are presented as means ± S.E.M of
   four independent experiments. The P values are 0.0005 (left) and 0.0106
   (right). ^*P < 0.05, one-way ANOVA. e, f Transwell migration of HUVECs
   (24 h). Results are presented as means ± S.E.M. of four independent
   experiments. The P values are 0.0004 (left) and 0.001 (right).
   ^*P < 0.05, one-way ANOVA. g–i Tube formation by HUVECs (6 h). Results
   are presented as means ± S.E.M of four independent experiments. P
   values from left to right are 0.0003, 0.0007, 0.0008, and 0.0038.
   ^*P < 0.05, one-way ANOVA. j–l The gene expression levels of PDGF, EGF,
   and VEGF in HUVECs in the presence or absence of NPVF ± RF9 as
   determined by qPCR. Results are presented as means ± S.E.M of four
   independent experiments, and the P values from left to right are
   0.0002, 0.0044, 0.0015, 0.009, 0.0042, and 0.0111. ^*P < 0.05, one-way
   ANOVA. m The protein expression level of VEGF in HUVECs in the presence
   or absence of NPVF ± RF9. Results are presented as means ± S.E.M of
   four independent experiments, and the P values are 0.0004 (left) and
   0.0004 (right). ^*P < 0.05, one-way ANOVA.

NPVF-induced migration and angiogenesis of HUVECs is mediated by MiR-181c-3p

   MicroRNAs (miRNAs) are a family of evolutionarily conserved endogenous
   gene repressors present in all living organisms. They are extensively
   involved in the post-transcriptional regulation of gene expression in
   various biological contexts, including the systemic control of many
   biological processes, such as osteogenesis and
   angiogenesis^[70]23–[71]26. For example, miR-497/195 regulates bone
   homeostasis by regulating the coupling of angiogenesis and
   osteogenesis^[72]27, and miR-136-3p regulates vascularization by
   targeting PTEN^[73]28.

   Blood vessel invasion plays a vital role in new bone regeneration by
   providing oxygen, nutrients, cells, and hormones^[74]29–[75]31. Hence,
   to further explore the molecular mechanism underlying NPVF promotion of
   angiogenesis of HUVECs, we performed miRNA sequencing (miRNA-seq) of
   HUVECs after NPVF treatment (48 h). The miRNA-seq results showed that
   six miRNAs were upregulated and thirteen miRNAs were downregulated in
   the NPVF-treated HUVECs compared with untreated HUVECs (Fig. [76]3a).
   Among all the miRNAs with upregulated expression, we were particularly
   interested in miR-181c, the target gene of which plays a substantial
   role in endothelial cell function^[77]32,[78]33. Thus, we further
   investigated its expression level by qPCR. Our results showed that the
   expression level of miR-181c-3p in HUVECs was remarkably upregulated
   under NPVF induction (Fig. [79]3b). We further validated the effect of
   miR-181c-3p on HUVECs by performing a series of experiments, namely a
   wound healing experiment (36 h), transwell assay (24 h), tube formation
   assay (6 h), qPCR, and ELISA analysis, where cells were treated with
   miR-181c-3p or NC-mimics (as negative control). The wound healing
   experiment and transwell assay both demonstrated that HUVECs in the
   miR-181c-3p group exhibited better migration and proliferation
   capacities than those in the control group (Fig. [80]3c-f). In
   addition, quantitative analysis of tube formation revealed that the
   miR-181c-3p group formed more loop structures and branch points,
   suggesting higher angiogenic activity (Fig. [81]3g-h, Fig. S[82]1a).
   Moreover, the qPCR and ELISA (48 h) results indicated that expression
   of angiogenesis-associated markers (VEGF, EGF, PDGF, and FGF) was
   significantly increased in HUVECs after miR-181c-3p treatment (Fig.
   S[83]1b-f). Based on the above results, we inferred that miR-181c-3p
   mediates the NPVF-induced migration and angiogenesis in HUVECs.

Fig. 3. miR-181c-3p mediates the NPVF neuropeptide-induced migration and tube
formation of HUVECs.

   [84]Fig. 3
   [85]Open in a new tab

   a Heatmap of miRNA sequencing results for HUVECs treated with NPVF
   (48 h). The color represents the relative expression value of miRNA
   genes. b The gene expression level of miR-181c-3p in HUVECs after
   treatment with NPVF. Results are presented as means ± S.E.M. of four
   independent experiments, and the P value is 0.0005. ^*P < 0.05,
   Student’s t-test. c, d Wound healing of HUVECs (36 h). Results are
   presented as means ± S.E.M. of four independent experiments, and the P
   value is 0.0004. ^*P < 0.05, Student’s t-test. e, f Transwell migration
   of HUVECs (24 h). Results are presented as means ± S.E.M. of four
   independent experiments, and the P value is 0.0001. ^*P < 0.05,
   Student’s t-test. (g, h) Tube formation by HUVECs (6 h). Results are
   presented as means ± S.E.M. of four independent experiments, and the P
   value is 0.0001. ^*P < 0.05, Student’s t-test. (i) Construction of the
   wild-type AGO1 3′-UTR and mutant AGO1 3′-UTR plasmids. The potential
   target sites of miR-181c-3p in the AGO1 3′-UTR are shown. j Normalized
   luciferase activity of the vectors after miR-181c-3p overexpression in
   HEK293T cells. Results are presented as means ± S.E.M. of six
   independent experiments, and the P value is 0.0039. ^*P < 0.05,
   Student’s t-test. k The level of AGO1 was remarkedly downregulated
   after miR-181c-3p overexpression in HEK293T cells. Results are
   presented as means ± S.E.M. of four independent experiments, and the P
   value is 0.0005. ^*P < 0.05, Student’s t-test. (l, m) AGO1 siRNA was
   used to knock down the expression of AGO1 protein in HUVECs. Results
   are presented as means ± S.E.M. of four independent experiments, and
   the P values are 0.0004 (left) and 0.0001 (right). ^*P < 0.05, one-way
   ANOVA. n The wound healing of HUVECs. (o) Transwell migration of
   HUVECs. p Tube formation by HUVECs.

   MiRNAs generally function by base pairing with the 3’-untranslated
   regions (3’-UTRs), thereby downregulating the post-transcriptional
   expression of mRNAs. Bioinformatic analysis of miR-181c-3p showed that
   it might target the 3’-UTR sequence of AGO1 (Fig. [86]3i). The
   wild-type and mutant AGO1 3′-UTR sequences were thus cloned into the
   psiCHECK2 luciferase vector to explore whether miR-181c-3p directly
   regulates AGO1 production. Transfection of miR-181c-3p mimic
   significantly inhibited the luciferase activity of the wild-type 3′-UTR
   group. In contrast, the luciferase activity of the mutant 3′-UTR group
   was unaffected by the miR-181c-3p mimic (Fig. [87]3j). Moreover,
   transfection of the miR-181c-3p mimic distinctly decreased the
   expression level of AGO1 in HEK293T cells, as compared with those
   transfected with the miR-NC (Normal Control) mimic (Fig. [88]3k).

   We further investigated the relationship between AGO1 and NPVF in
   HUVECs. NPVF treatment substantially reduced the AGO1 gene and protein
   expression levels in HUVECs (Fig. S[89]1g-h). Next, we probed the exact
   role of AGO1 in the signaling pathway in HUVECs by knocking down AGO1
   using small interfering RNA (siRNA). A western blotting assay revealed
   that the control siRNA (NC siRNA) did not affect the protein expression
   level of AGO1 in HUVECs, while AGO1 siRNA distinctly downregulated it
   (Fig. [90]3l-m). Moreover, AGO1 siRNA distinctly promoted the migration
   and tube formation of HUVECs, which mimicked the effect of NPVF
   (Fig. [91]3n-p, Fig. S[92]1i-k). Collectively, these results indicated
   that NPVF regulates cell migration and angiogenesis by activating
   NPFFR1, which then mediates the miR-181c-3p/AGO1 pathway.

The WNT/β-catenin signaling pathway mediates the NPVF-induced osteogenic
differentiation of BMSCs

   We next explored whether the NPVF peptide mediates osteogenesis,
   another hallmark event of bone regeneration. Indeed, preliminary
   studies revealed that NPVF promoted the osteogenic differentiation of
   BMSCs in a dose-dependent manner (Fig. S[93]2). To probe the underlying
   mechanism, we carried out transcriptome analysis of BMSCs in the
   presence and absence of NPVF. Heatmap and KEGG pathway enrichment
   analysis indicated that the WNT pathway was highly enriched in the
   presence of NPVF (Fig. [94]4a-b), implying that the WNT pathway might
   play a key role in the NPVF-induced osteogenic differentiation of
   BMSCs. A western blotting assay revealed that NPVF could distinctly
   increase the level of β-catenin in BMSCs, while the NPFFR1 antagonist
   RF9 and the WNT antagonist JW74 both abolished the upregulation of
   β-catenin in NPVF-treated BMSCs (Fig. [95]4c). Alizarin Red S (ARS, 21
   days) and alkaline phosphatase (ALP, 7 days) assays were then carried
   out to explore the role of the Wnt/β-catenin pathway in NPVF-induced
   osteogenic differentiation of BMSCs. Both the staining and
   corresponding quantitative results indicated that NPVF remarkably
   promoted extracellular mineralization and ALP activity in BMSCs, while
   these effects were abolished in the presence of either RF9 or JW74
   (Fig. [96]4d, Fig. S[97]3). In conclusion, these results revealed that
   NPVF promotes osteogenic differentiation of BMSCs via the Wnt/β-catenin
   signaling pathway.

Fig. 4. The WNT/β-catenin pathway mediates the NPVF-induced osteogenic
differentiation of BMSCs.

   [98]Fig. 4
   [99]Open in a new tab

   a The results of miRNA sequencing of BMSCs treated with NPVF. The color
   represents the relative expression value of miRNA genes. b KEGG
   enrichment analysis of BMSCs after treatment with NPVF. c The protein
   expression of β-catenin in BMSCs. Results are presented as means ±
   S.E.M. of four independent experiments, and the P values from left to
   right are 0.0013, 0.0754, and 0.092. ^*P < 0.05, one-way ANOVA. d ARS
   and ALP staining of BMSCs undergoing osteogenic differentiation.

CsgA-NPVF scaffolds enhance bone regeneration in vivo

   After verifying that NPVF can induce angiogenesis and osteogenesis of
   HUVECs, we further explored the application potential of NPVF in bone
   defect repair. We chose to construct a functional recombinant protein,
   CsgA-NPVF, by fusing the N-terminus of NPVF with the amyloid protein
   CsgA, a major protein component of Escherichia coli curli biofilm,
   which is well known for its universal ability to form coatings around a
   given substrate through amyloid self-assembly (Fig. [100]5a). This
   strategy offers a facile way to display functional peptides on amyloid
   scaffolds and overcomes the problems of serum instability and low
   bioavailability of peptide drugs in preclinical application. After
   induction of bacterial expression and protein purification, we
   demonstrated by Transmission Electron Microscopy (TEM) that
   NPVF-containing nanofibers could only be formed when expressed in
   fusion with CsgA (Fig. S[101]4).

Fig. 5. Preparation and characterization of the CsgA-NPVF nanofiber coating.

   [102]Fig. 5
   [103]Open in a new tab

   a Schematic diagram depicting the induction of recombinant protein
   expression by Isopropyl ß-D-1-thiogalactopyranoside (IPTG) and the
   preparation of nanofiber coatings on a porous 3D HAp scaffold. b Congo
   red staining of the purified CsgA and CsgA-NPVF nanofibers. c Assembly
   kinetics of CsgA and CsgA-NPVF monomer solutions monitored by ThT
   fluorescence assay. The imidazole buffer used to elute the target
   proteins during protein purification was used as a control for Congo
   red staining and ThT fluorescence assay. d AFM images showing the
   morphology of the self-assembled protein nanofibers. e Comparison of
   the adsorption behavior of the CsgA and CsgA-NPVF nanofibers on HAp
   substrates measured by QCM. f SEM images showing the surface morphology
   of the 3D HAp scaffold before and after the coating process.

   We performed Congo red staining (Fig. [104]5b) and a Thioflavin T assay
   (Fig. [105]5c) to verify the amyloid features of both CsgA and
   CsgA-NPVF; their fibrous structures were further confirmed by atomic
   force microscopy (AFM) imaging (Fig. [106]5d). Quartz crystal
   microbalance (QCM) tests indicated strong adsorption of the CsgA-NPVF
   nanofibers to the HAp substrate (Fig. [107]5e). After overnight
   immersion in fresh purified protein solution, we observed that the
   CsgA-NPVF protein could form robust nanofiber coatings on the surface
   and internal interfaces of a porous 3D HAp scaffold. Morphological
   changes recorded by scanning electron microscopy (SEM) imaging
   validated the successful formation of CsgA-NPVF nanofiber coatings on
   the scaffold surfaces (Fig. [108]5f). To further explore the stability
   of the recombinant protein coating, we performed in vitro protein
   release experiments using simulated body fluid (SBF). Both uncoated
   scaffolds and CsgA-NPVF-coated scaffolds were immersed in SBF solution
   (pH=7.4) and incubated at 37 °C for two months. His-tag labelled
   CsgA-NPVF proteins released into the SBF solution was detected using a
   commercially available enzyme-linked immunosorbent assay (ELISA) kit.
   During the 60-day test, the amount of supernatant protein in the
   CsgA-NPVF groups was almost the same as that in the control group (Fig.
   S[109]5), indicating that the CsgA-NPVF coatings had good stability.

   Owing to its good biocompatibility and surface adhesion capacity, CsgA
   has been used in various medical applications, such as surface
   modification of cell culture scaffolds and bone implants^[110]34,
   secretion of CsgA-fused cytokine for the treatment of inflammatory
   bowel disease (IBD)^[111]35. Based on previous studies, we reasonably
   speculated that fusion with CsgA could improve the stability of NPVF
   peptides while preserving their biological activity. Before animal
   experiments, we tested the cytocompatibility of different scaffolds
   using BMSCs and found that compared with cells grown on the empty HAp
   scaffolds and CsgA-coated scaffolds, cells grown on the
   CsgA-NPVF-coated scaffolds had the best adhesion capacity and the
   highest viability (Fig. S[112]6).

   We next evaluated the efficacy of the NPVF-containing protein-coated
   HAp scaffolds in assisting bone regeneration in vivo using a standard
   rat calvarial defect model. We first set out to investigate the
   expression level of NPFFR1 in rat calvarias by immunohistochemistry
   (IHC). The results indicated that NPFFR1 was highly expressed in rat
   calvarias (Fig. S[113]7), which was consistent with the results from
   the in vitro experiment (Fig. [114]2a). Furthermore, at 12 weeks after
   transplantation, micro-CT reconstruction data indicated that more fresh
   bone formation formed into the direct 3D space of the CsgA-NPVF
   protein-coated scaffolds in the defects than in those treated with the
   control groups (bare HAp scaffolds and CsgA protein-coated scaffolds)
   (Fig. [115]6a). Quantitative analysis revealed that both the bone
   mineral density (BMD, 1.538 g/cm^3 VS 1.276 g/cm^3) and the bone
   volume/tissue volume (BV/TV, 69.069% VS 50.142%) ratio in the CsgA-NPVF
   group were higher than those in the control groups (Fig. [116]6b-c).
   Hematoxylin & eosin (H&E) staining and Masson staining all demonstrated
   enhanced osteoblastic marker expression and bone matrix formation
   occurred in the 3D space of the CsgA-NPVF protein-coated scaffolds in
   the defects than in those treated with the control groups (bare HAp
   scaffolds and CsgA protein-coated scaffolds (Fig. [117]6d-e). Moreover,
   IHC analysis revealed much higher expression levels of CD31 and COL I
   in the bone matrix formation for the CsgA-NPVF group (Fig. [118]6f-g).
   It should be noted that HAp scaffolds are a well-recognized and
   standard biomaterial used as bone graft substitutes, providing an
   objective evaluation of bone regeneration^[119]36–[120]39. The
   HAp-CsgA-NPVF group showed significantly more new bone in-growth and
   neovascularization in the defect area than the pure HAp scaffold group
   and HAp-CsgA group, suggesting that our cell-free and growth
   factor-free composite is superior to this traditionally employed
   scaffold. Overall, the CsgA-NPVF-coated scaffold enhanced bone
   regeneration via coupling of osteogenesis and angiogenesis in rats.

Fig. 6. CsgA-NPVF nanofiber-coated scaffolds promote bone defect repair in a
rat calvarial defect model.

   [121]Fig. 6
   [122]Open in a new tab

   a 3D reconstructed (left) and 2D sagittal (right) micro-CT images of
   the bone defects. b, c The quantification of BMD and BV/TV in the new
   bone. Results are presented as means ± S.E.M. of five animals, and the
   P values from left to right are 0.0009, 0.0001, 0.0008, and 0.0003.
   ^*P < 0.05, one-way ANOVA. d Histological analysis of new bone
   in-growth using H&E (blue arrows indicate new bone formation). e
   Masson’s histological analysis of new bone formation (green arrows show
   less mature fresh bone and yellow arrows point to mature bone tissue).
   f Histological observation of CD31 staining (red arrow shows positive
   staining for CD31). g Immunohistochemical analysis of COL I staining
   (red arrows indicate positive staining for COL I).

Discussion

   Neuropeptides are the messengers in the brain-bone axis. They are
   released from sensory nerve endings and exert distinct functions on
   skeletal system, including BMSC differentiation^[123]40,[124]41,
   cartilage metabolism^[125]2,[126]4, fracture healing^[127]42,[128]43,
   and bone-fat equilibrium^[129]20. Despite advances in elucidating the
   functions of these neuropeptides, no previous studies have reported a
   connection between the NPFF system and bone homeostasis. In this study,
   we demonstrate for the first time that the NPVF peptide (a
   representative of the NPFF system) significantly promotes both
   angiogenesis and osteogenesis. In addition, we identified the signaling
   pathways through which NPVF participates in maintaining bone
   homeostasis. The finding of a crosstalk between the brain and bone
   suggests that other neuropeptides may also play an important role in
   bone homeostasis and that NPVF could potentially be used as a peptide
   drug to treat bone fractures. Moreover, the identification of the
   critical pathways involved in mediating angiogenesis and osteogenesis
   paves the way for further investigation and exploitation of NPVF in
   therapeutic applications. Notably, NPVF represents a type of endogenous
   peptide that exhibits promising therapeutic efficacy. The main
   advantages of peptides include their safety, the high target affinity,
   targeting a wide range of molecules, low toxicity and low
   immunogenicity^[130]44–[131]46. The functional coatings based on the
   NPVF peptide-containing fusion proteins might avoid problems including
   immune rejection and donor-donor variability where existing xenogenous
   cell-laden tissue engineering strategy frequently
   meets^[132]47–[133]49.

   Beyond the analysis of the molecular regulatory pathway, we explored
   the application of the NPVF peptide in the pre-clinical treatment of
   bone defects. Although there is increased interest in peptide
   therapeutics, the bioavailability of these peptides is usually hampered
   by their structural vulnerability in the serum environment and rapid
   clearance via renal filtration^[134]50,[135]51. To resolve these
   issues, we took advantage of a programmable self-assembling CsgA
   nanofiber coating when developing a strategy for the pre-clinical
   application of the NPVF peptide. CsgA is an amyloid protein that
   assembles into a fibrous coating on a given substrate in solution.
   Another distinctive feature of CsgA is that the amyloidogenic core of
   the assembled structure does not significantly diverge from that of the
   original amyloid backbone structure when domains with the appropriate
   size and conformation are present^[136]52. In our work, we fused the
   NPVF peptide to the C-terminus of the CsgA subunit to fabricate a
   porous HAp scaffold with a functional CsgA-NPVF nanofiber coating. It
   is well recognized that porous HAp scaffolds are highly biocompatible,
   osteoconductive, and osseointegrative, and they have excellent
   mechanical properties, which makes them the most commonly used
   biomaterial in bone tissue repair^[137]53,[138]54, both for scientific
   research and clinical applications^[139]55–[140]59. The porous HAp
   scaffold we selected is a cell culture insert that has already been
   commercialized in the field of bone regeneration. The experimental
   results showed that the CsgA-NPVF nanofiber-coated HAp scaffold
   exhibited drastically increased osteoinductivity and
   angiogenic-inductivity compared with those of the uncoated scaffolds in
   a rat cranial defect model, demonstrating its high pre-clinical
   application value.

   Note that NPAF (three versions, including the NPSF peptide containing
   18 amino acid residues, the 8 amino acid short version of NPAF and a 37
   amino acid extended version of RFRP-1, had been described in the
   literature) is also an important family member of NPFF
   system^[141]21,[142]60. We initially chose the shorter polypeptide
   (NPVF) in our study rather than the longer polypeptide (NPSF) because
   we rationalize that the fusion of the shorter polypeptide onto CsgA
   would not disrupt the self-assembly of CsgA protein and thus ensure
   their fiber coating formation abilities. In addition, our series of
   previous works mainly focused on the osteo-inductivity of NPVF and the
   in-depth molecular mechanism (manuscript to be submitted), the selected
   NPVF had already been validated with specific functions and perfectly
   met the need of current studies. However, it should be noticed that
   neither NPSF nor NPAF should be ignored or excluded from the future
   study of NPFF family, where they might also play critical role. It
   should be noted that the NPVF itself is an important member in the
   family of neuropeptide. This work aimed at introducing the possibility
   of utilizing NPVF in the skeletal system which widens the application
   of neuropeptides. Combined with our previous findings that NPFF
   activation of NPFF2 receptor stimulates neurite outgrowth in Neuro 2 A
   cells via activating the ERK signaling pathway^[143]18, we will further
   work on how this neuropeptide regulates bone metabolism and the
   feedback of the skeletal system to the brain in the future.

   In summary, we have revealed the role and potential mechanism of
   neuropeptides in regulating bone formation and developed a
   generalizable and viable strategy based on neuropeptide-containing
   protein coatings for bone repair. Recalling the role of neuropeptides
   in regulating angiogenesis, it is reasonable to speculate that
   neuropeptides would also provide an effective treatment for various
   human diseases that are closely related to angiogenesis, such as
   cancer, obesity, cardiovascular diseases, and chronic inflammation. It
   is also worth noting that the CsgA nanofiber coating method employed in
   our work showed great biocompatibility and robustness. The CsgA coating
   technique alleviates the problem of enzymatic degradation in the
   physiological environment seen with other frequently used techniques
   including oral administration and intravenous injection, and it offers
   long-term protection of the functional peptide, thus substantially
   enhancing peptide drug bioavailability. Notably, many other CsgA fusion
   proteins with well-defined biological functions can be precisely
   engineered using a simple modular genetic design, offering
   possibilities to functionally modify various graft or scaffold
   materials in the field of tissue engineering.

Methods

Study design

   The objective of this study was to reveal the role and potential
   mechanism of NPVF in regulating bone formation. We discovered a close
   relationship between the neuropeptide NPVF and bone formation, and
   uncovered the underlying molecular mechanisms. Building on these
   discoveries, we adopted an amyloid protein coating technique to display
   the NPVF peptide in 3D HAp scaffolds for in vivo treatment of bone
   defects. HUVECs and BMSCs were used as target cells for the evaluation
   of NPVF in regulating angiogenesis or osteogenesis, respectively. For
   the in vitro evaluation of NPVF in regulating angiogenesis, we carried
   out series of in vitro tests including wound healing assay, transwell
   assays and ELISA of VEGF in HUVECs. For revealing of the molecular
   mechanism miRNA sequencing and dual-luciferase reporter assay were
   performed. The result indicated the miR-181c-3p was the center mediator
   which downregulation of AGO1. The evaluation of osteogenesis was
   performed by PCR, western blots and ARS in BMSCs. Further experiments
   revealed that NPVF promotes osteogenic differentiation of BMSCs via the
   Wnt/β-catenin signaling pathway.

   To improve the stability and bioavailability and thus efficacy of NPVF
   as a promoter of bone regeneration, we genetically engineered
   amyloid-NPVF-fusion proteins and utilized them as self-assembling
   nanofiber coatings to treat bone defects in a rat calvarial defect
   model. For the preparation of functional coatings, we constructed a
   pET22b-CsgA-NPVF plasmid and transformed it into the BL21 (DE3) E. coli
   competent cell for recombinant protein purification. The amyloid
   characteristics of CsgA-NPVF were verified by Congo red staining and
   Thioflavin T assay, and the fibril morphology was demonstrated by AFM
   imaging. QCM technique was adopted to reveal the adsorption of proteins
   on HAp scaffolds. Morphological changes of the scaffold after coating
   process were observed by SEM imaging. For in vivo studies, animals were
   randomized to three groups, and no mice were excluded during the
   experiment. Mice were treated by an operator who was blinded to
   treatment groups. All tests were replicated by at least three times,
   and all replicates were successful.

Cell culture

   The embryonic kidney 293 T cells (HEK93T) and human BMSCs were
   purchased from the cell bank of the Chinese Academy of Sciences
   (Shanghai, China) and cultured in Dulbecco’s Modified Eagle’s Medium
   (Gibco, Grand Island, NY) and α Minimum Essential Medium (α-MEM,
   Gibco), respectively, containing 10% fetal bovine serum (FBS)
   (Invitrogen, Carlsbad, CA) and 1% penicillin/streptomycin (Invitrogen)
   at 37 °C with 5% CO[2]. HUVECs were purchased from Procell (Wuhan,
   China) and cultured in endothelial cell medium (ECM, ScienCell, CA).
   The miRNA sequencing of HUVECs and the mRNA sequencing of BMSCs were
   performed by LC-bio (Hangzhou, China).

Cell proliferation

   CCK-8 (Beyotime, Jiangsu, China) was used to detect the proliferation
   of BMSCs and HUVECs after NPVF treatment. In brief, a 96-well plate was
   seeded with 5×10^3 cells per well in 100 μL medium at day 0. On days 1,
   3, 5 and 7, the medium was refreshed with 10 μL CCK-8 solution and
   90 μL medium, then the plate was incubated in an incubator for 2 h.
   Finally, the absorbance values of the supernatants of each well were
   detected at 450 nm and the medium was refreshed.

Angiogenesis and migration of HUVECs

   The wound healing assay was conducted by seeding HUVECs at 5×10^5 per
   well in a 6-well culture plate. Then, the confluent cells were
   scratched using a sterile pipette tip. Cells were then incubated in
   conditional medium and images of the wounds were captured at 0 h and
   36 h. Image J was used to calculate the rates of wound healing.
   Transwell assays were performed to test the migration of HUVECs after
   treatment with NPVF, RF9 (Selleck, Shanghai, China), miR-NC mimic,
   miR-181c-3p mimic, NC siRNA, and AGO1 siRNA (Ruibo, Guangdong, China).
   In brief, 2×10^4 cells per well were seeded in the upper chambers of a
   transwell plate with 200 μL medium and 700 μL of conditional medium was
   added to the lower chambers. After incubation 24 h at 37 °C, the cells
   were washed and fixed with 4% paraformaldehyde; cells in the upper
   surface of the membrane were rubbed away carefully and cells in the
   lower surface were stained by crystal violet. Finally, the total number
   of cells in six randomly chosen lower surfaces were counted under a
   microscope. The tube formation assay was carried out by adding 50 μL
   per well Matrigel (BD Bioscience, San Jose, CA) to 96-well plates and
   allowing gelatinization to proceed at 37°C for 30 min. Then, 4×10^4
   HUVECs/well were seeded on the surface of the Matrigel. A LEICA
   microscope was used to capture tube formation images 6 h later, and the
   number of complete capillaries and nodes in each hole was counted.

Enzyme-linked immunosorbent assay

   HUVECs were incubated in a 6-well plate. After reaching 80% confluence,
   cells were incubated in serum-free medium for 48 h. Then an ELISA kit
   (Neobioscience, Shenzhen, China) was used to measure the amount of VEGF
   in the supernatant. The absorbance values at 450 nm were used to
   calculate the VEGF content using a standard curve.

Osteogenesis assay

   To investigate the effect of NPVF, RF9, and JW74 (Selleck) on
   osteogenic differentiation of BMSCs, cells were incubated in
   conditional medium for a specific amount of time. In brief, 2×10^5
   cells were seeded in 6-well plates with α-MEM. After cells reached 80%
   confluence, osteogenic medium (Cyagen, Suzhou, China) was refreshed to
   induce osteogenic differentiation every two days. ARS staining and ALP
   staining were performed on days 21 and 14, respectively. A LEICA
   microscope was used to capture images. The quantification of Alizarin
   red staining was performed according to the protocol of
   Hexadecylpyridinium chloride monohydrate (Solarbio, Beijing, China),
   the OD value was detected at 562 nm by a multifunctional microplate
   reader after the dissolution of mineralization nodes. The
   quantification of ALP staining was performed by an Alkaline phosphatase
   assay kit (Jiancheng, Nanjing, China) in the manuscript.

Quantitative real-time polymerase chain reaction

   BMSCs or HUVECs were incubated in 6-well plates with conditional medium
   for 48 h, and then total miRNA or mRNA was extracted based on the
   manufacturer’s protocols (EZBioscience, USA). Reverse transcriptase
   reactions consisted of purified RNA and 50 nM RT primer. QPCR was
   performed on an MX3005P system using an SYBR Premix Ex Taq protocol
   (EZBioscience). GAPDH and U6 were used as the internal genes for
   normalization of mRNA and miRNA expression, respectively. The primers
   are listed in Table [144]S1.

Western blot analysis

   In brief, BMSCs or HUVECs were lysed with RIPA lysis buffer (1 mM PMSF,
   Beyotime). The BCA protein assay kit (Beyotime) was used to determine
   the protein concentrations according to the manufacturer’s protocols.
   An equal amount of protein (35 μg) was electrophoresed on a 10% sodium
   dodecyl sulfate-polyacrylamide gel, then transferred to a PVDF
   membrane. The membrane was blocked with 5% (w/v) milk and incubated
   overnight at 4°C with primary antibodies against GAPDH, AGO1, or
   β-catenin (1:1000, CST, Shanghai, China). Then, the PVDF membrane was
   washed and incubated with an HRP-conjugated second antibody (1:5000,
   CST). Finally, the membrane was washed and reacted with an ECL
   detection kit (Thermo Scientific, Waltham, MA). Scanning densitometry
   was used to quantify the signals.

Dual-luciferase reporter assay

   The wild-type and the mutant fragments of the AGO1 3′-UTR sequence
   including the predicted miR-181c-3p target sites (Positions 3932-3938)
   were directly synthesized (Obio, Shanghai, China) and individually
   cloned into the 3′ end of the psiCHECK2 luciferase vector (Promega;
   Madison, WI). The Lipofectamine 3000 reagent (Life Technologies) was
   used to transfect plasmids into the HEK293T cells with the miR-181c-3p
   mimic or miR-NC (Ruibo). A dual-luciferase reporter assay kit (Promega)
   was used to detect the activities of the Renilla and firefly
   luciferases at 48 h after transfection. Renilla luciferase activity was
   normalized to the firefly luciferase activity to calculate the relative
   luciferase activities.

Protein purification and coating fabrication

   To create the plasmid for CsgA-NPVF expression, the gene sequence of
   NPVF was synthesized by GENEWIZ and then inserted into the linearized
   pET22b-CsgA plasmid via one-step Gibson assembly to obtain
   pET22b-CsgA-NPVF. The pET22b-CsgA and pET22b-CsgA-NPVF plasmids were
   transformed into BL21 (DE3) E. coli competent cells individually.
   Protein expression and purification were conducted following the
   protocols described in a previous study^[145]52. Congo red staining and
   the ThT assay were performed to characterize the amyloid structure of
   the proteins following the methods reported in another study^[146]61.
   Then the protein solutions were incubated on mica surfaces for 2 hours
   and washed with deionized water for AFM imaging. Samples were dried
   under a constant flow of nitrogen gas before observation. Tapping mode
   AFM was performed on an Asylum MFP-3D atomic force microscope (Asylum
   Research) using microcantilevers (k = 26 N/m and ν~ 300 kHz). Next, QCM
   tests were applied to examine the adsorption capacities of the protein
   nanofibers on the HAp surface. Nanofiber solutions at an initial
   concentration of 0.5 mg·ml^-1 were introduced into HAp-coated QCM chips
   at a flow rate of 0.02 mL/min using a four-channel Ismatec ICP-N4
   peristaltic pump, followed by washing with PBS buffer at the same rate.
   After incubation in the protein solution overnight, the modified 3D HAp
   scaffolds were washed with copious amounts of deionized water and dried
   using a vacuum drying oven, then coated with an 8–10-nm gold layer
   using a sputter coater for SEM observation. SEM images were taken using
   a JSM 7800 scanning electron microscope operated at a 5-kV accelerating
   voltage in secondary electron imaging mode. The 3D HAp scaffolds we
   used were purchased from INNOTERE company (111CC4), the scaffold
   consists of synthetic calcium phosphate (mainly α-tricalcium phosphate)
   and nanocrystalline calcium deficient hydroxyapatite. The scaffolds
   offer interconnected porosity (3D version), high bioactivity, ease of
   handling, and high mechanical stability, making them ideal substrates
   for cell cultures in the field of bone regeneration.

TEM imaging

   For TEM imaging, a 10 μL droplet of protein was directly deposited on a
   TEM grid (Zhongjingkeyi Technology, EM Sciences) for 5 min. The
   excessive solution was wicked away with pieces of filter paper and the
   samples were rinsed twice with ddH2O by placing 10 μL ddH2O on TEM grid
   and quickly wicking it off with filter paper. The samples were then
   negatively stained with 10 μL 2 wt% uranyl acetate for 1 min and dried
   for 15 min under an infrared lamp. TEM images were obtained on a Tecnai
   G2 Spirit T12 transmission electron microscope operated at 120 kV
   accelerating voltage.

In vitro protein release experiment

   After the coating process, both uncoated HAp scaffolds and
   CsgA-NPVF-coated scaffolds were immersed in 10 mL SBF solution (pH=7.4)
   and incubated at 37 °C for two months. His-tag labelled CsgA-NPVF
   proteins released into the solution were detected using a commercially
   available His tag ELISA detection kit (Genscript, Nanjing, China).
   ELISA experiments were performed according to the instruction at 1, 3,
   7, 14, 30 and 60 days after immersion.

SEM assay and immunofluorescent staining

   SEM assays and immunofluorescent staining were performed to observe the
   morphology and adhesion of BMSCs on the scaffolds. In brief, after
   being seed with BMSCs (1.0 × 10^4) in 48-well plates for 48 h,
   scaffolds were fixed with glutaraldehyde and dehydrated by gradient
   concentrations of ethanol (30, 40, 50, 60, 70, 80, 90, 100 v/v%). Then,
   the scaffolds were observed by SEM. For immunofluorescent staining, the
   BMSCs co-incubated with scaffolds were stained with
   4’,6-diamidino-2-phenylindole (DAPI, blue) and rhodamine-phalloidin
   (red) to directly observe subtle cytoskeletal changes under a LEICA
   microscope.

Animal experiments

   With the approval from the Animal Research Committee at Shanghai Sixth
   People’s Hospital, 20 8-week-old male Sprague-Dawley rats (250 g) were
   acquired and randomly divided into three groups for animal experiments.
   Chloral hydrate sodium (250 mg·kg^-1) administered by intraperitoneal
   injection was used to anaesthetize rats. A 1.5–2.0 cm sagittal incision
   was cut on the scalp, then an electric trephine was used to make two
   full-thickness 5-mm diameter defects on both sides of the skull. The
   scaffolds (5 mm in diameter and 2 mm in thickness) were randomly
   implanted into the cranial defects. After surgery, silk sutures were
   used to carefully suture the soft tissues. All rats were sacrificed at
   week 12 and the calvarias were harvested.

Micro-CT scanning and histomorphometry

   The rat skulls were scanned by a micro-CT scanner (Skyscan, Kontich,
   Belgium). Tomograms were reconstructed with the 3D creator software
   (Skyscan), and the CTAn image analysis software was used to calculate
   the BV/TV ratio and the BMD (g·cm^-3) according to the reconstructed
   images. Next, the samples were decalcified, embedded, and cut (5-μm
   sections). The sections were then stained with H&E or Masson trichrome,
   or immunohistochemical staining was performed (NPFFR1, CD31, VEGF, and
   OPN). A LEICA microscope was used to capture all images.

Statistics and reproducibility

   All data are presented as means ± standard error of the mean (S.E.M.).
   Student’s t-test or one-way ANOVA was used to determine the differences
   between groups with Bonferroni’s correction in SPSS 18 (IBM, Armonk,
   NY). *P < 0.05 was considered statistically significant. All the data
   in this article is reproducible. All experiments were of at least three
   replicates with detailed information listed in each experiment.

Reporting summary

   Further information on research design is available in the [147]Nature
   Portfolio Reporting Summary linked to this article.

Supplementary information

   [148]Peer Review File^ (3.8MB, pdf)
   [149]Supplementary Information^ (1.1MB, pdf)
   [150]42003_2023_4567_MOESM3_ESM.pdf^ (85.6KB, pdf)

   Description of Additional Supplementary Files
   [151]Supplementary Data^ (188.9KB, xlsx)
   [152]Reporting Summary^ (1.9MB, pdf)

Acknowledgements