RESEARCH ARTICLE

Kangfuxin Accelerates Extraction Socket Healing by

Promoting Angiogenesis Via Upregulation of CCL2
in Stem Cells
Zheng-Rong Gao,1 Ying-Hui Zhou,1,2 Ya-Qiong Zhao,1 Jie Zhao,1 Qin Ye,1 Shao-Hui Zhang,1
Yao Feng,1 Li Tan,1 Qiong Liu,1 Yun Chen,1 Ze-Yue Ouyang,1 Jing Hu,1
Marie Aimee Dusenge,1 Yun-Zhi Feng,1 and Yue Guo1
1
Department of Stomatology, the Second Xiangya Hospital, Central South University, Changsha, China
2
National Clinical Research Center for Metabolic Diseases, Hunan Provincial Key Laboratory of Metabolic Bone Diseases, and Department of
Metabolism and Endocrinology, The Second Xiangya Hospital of Central South University, Changsha, China

ABSTRACT
Kangfuxin (KFX) shows potential in wound healing, but its role in socket healing is unclear. This research finds increased bone mass,
mineralization, and collagen deposition in KFX-treated mice. Mouse bone marrow mesenchymal stem cells, human periodontal lig-
ament stem cells (hPDLSCs), and human dental pulp stem cells (hDPSCs) are treated with KFX under osteogenic induction. RNA-
sequencing reveals upregulated chemokine-related genes, with a threefold increase in chemokine (C-C motif) ligand 2 (Ccl2). The
conditioned medium (CM) of hPDLSCs and hDPSCs treated with KFX promotes endothelial cell migration and angiogenesis. Ccl2
knockdown abolishes CM-induced endothelial cell migration and angiogenesis, which can be reversed by recombinant CCL2 treat-
ment. KFX-treated mice showed increased vasculature. In conclusion, KFX increases the expression of CCL2 in stem cells, promoting
bone formation and mineralization in the extraction socket by inducing endothelial cell angiogenesis. © 2023 American Society for
Bone and Mineral Research (ASBMR).

KEY WORDS: KANGFUXIN; EXTRACTION SOCKET HEALING; STEM CELLS; CCL2; ANGIOGENESIS

Introduction reconstruction to achieve maximal therapeutic efficacy with

minimal side effects.(7-9) Bu-Shen-Gu-Chi-Wan was shown to
decrease the expression of the proinflammatory cytokines and
T he incidence of delayed socket healing after tooth extraction
is approximately 7.5%.(1) Delayed healing and dimensional
loss of the alveolar ridge make subsequent treatment difficult
stimulate alveolar bone remodeling in experimental periodonti-
tis.(10) Icariin suppresses bone loss by inhibiting osteoclasts and
because they further impair occlusal function and have negative enhancing osteocalcin (OCN) expression in the rat tooth extrac-
aesthetic consequences.(2) Anorganic bovine bone, guided bone tion model.(11) Yunnan Baiyao has potential in promoting osteo-
regeneration, platelet-rich plasma, and autogenous bone grafts genesis via upregulating the expression of OCN and alkaline
are the most common methods used to stimulate socket heal- phosphatase (ALP). Meanwhile, its safety in socket healing has
ing.(2,3) However, immunologic rejection, bacterial infection, been confirmed in clinical research.(12) Kangfuxin (KFX), the
increased risk of donor site morbidity,(3,4) and the high cost ethanol extract of Periplaneta americana L., is a TCM that has
(ranging from $450 to $2500 USD per treatment) are important been widely used for wound healing, gastroprotection, duode-
factors to consider.(5) Establishing an effective, easily implemen- nal ulcer, and diabetic foot ulcer;(13-16) it can stimulate prolifera-
ted method to accelerate socket healing after tooth extraction is tion and migration and regulate the cell cycle in bone marrow
a major goal in dentistry. mesenchymal stem cells (BMSCs).(17) Thus, it is possible that
Drugs have also been used to promote cell proliferation and KFX can promote socket healing.
differentiation, along with bone remodeling in experimental Stem cells (SCs) play an important role in bone regeneration,
and clinical settings.(6) Recently, there has been increasing inter- and BMSCs have been used in tissue engineering research for
est in using traditional Chinese medicine (TCM) for bone decades.(18) BMSCs can not only directly stimulate bone

Received in original form February 4, 2023; revised form May 10, 2023; accepted May 18, 2023.
Address correspondence to: Yue Guo, PhD, and Yun-Zhi Feng, PhD, Department of Stomatology, The Second Xiangya Hospital, Central South University,
139 Renmin Middle Road, Changsha, Hunan 410011, China. E-mail: [email protected] and [email protected]
Additional Supporting Information may be found in the online version of this article.
Z-RG and Y-HZ contributed equally to this work.
Journal of Bone and Mineral Research, Vol. 00, No. 00, Month 2023, pp 1–14.

DOI: 10.1002/jbmr.4860
© 2023 American Society for Bone and Mineral Research (ASBMR).

1 n
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
formation but also indirectly regulate osteogenesis by promot- were randomly allocated to two groups with three mice each:
ing angiogenesis, inhibiting osteoclast differentiation and func- KFX and normal saline (NS) groups. Mice were weighed and
tion, and suppressing inflammation.(19-22) We previously anesthetized with 1% pentobarbital sodium (7.5 μL/g). The max-
showed that in insulin receptor substrate 1 (IRS-1)-deficient mice, illary first molars (M1) were extracted and pictures of mice max-
BMSCs promoted bone formation via the IRS-1/miR-342/colla- illa before and after the surgery were taken. After surgery, the
gen type I alpha 2 (Col1α2) axis in a temporally specific man- extraction sockets were continuously scrubbed with 0.2 mL
ner.(23) BMSCs were used to promote bone formation in a KFX or NS twice daily by using sterile swabs. At 1, 3, 5, 7,
maxillary defect along with alveolar bone formation.(21,24,25) 14, and 28 days after tooth extraction, mice were euthanized
Stem cells in the oral cavity, including periodontal ligament stem by 1% pentobarbital sodium overdose, and we prepared images
cells (PDLSCs) and dental pulp stem cells (DPSCs), have high of extraction healing. In addition, at 5, 7, 14, and 28 days, the
osteogenic differentiation potential and have been used in bone maxillae samples were collected for the subsequent assay.
regeneration.(26,27) PDLSCs and the conditioned medium
(CM) from PDLSCs cultures decreased tumor necrosis factor Cell culture and CM preparation
alpha (TNF-α) mRNA expression in periodontal tissue and
enhanced the mineralization of healthy periodontal ligament Human umbilical vein endothelial cells (HUVECs) were pur-
remnants, thus contributing to bone formation.(28,29) In a cranio- chased from American Type Culture Collection (Manassas, VA,
facial defect repair model, DPSCs had comparable bone regener- USA). Mandibular mouse BMSCs (mBMSCs) were isolated from
ation and mineralization efficacy with BMSCs and could be the mandible as previously described.(31) Briefly, mice were
obtained from the oral cavity by a minimally invasive euthanized with an overdose of pentobarbital sodium, and the
procedure.(30) mandible was removed. After removing the soft tissues and
In this study, we investigated whether KFX can accelerate teeth, the mandible was crushed with scissors and flushed with
bone healing in a mouse tooth extraction model. We also exam- complete growth medium (alpha-minimal essential medium
ined the mechanism of action of KFX in regulating bone remo- [α-MEM]; Gibco, Grand Island, NY, USA) supplemented with
deling using in vitro and in vivo models. 10% fetal bovine serum (FBS, Gibco) and 1% penicillin and strep-
tomycin to obtain mBMSCs, which were seeded in culture dishes.
hPDLSCs and hDPSCs were isolated as previously described.(26,27)
Materials and Methods Tooth samples from human subjects were collected with the
approval of the Ethics Committee of The Second Xiangya Hospi-
Animals
tal of Central South University. Briefly, dental pulp, periodontal
All animal experiments in this study were approved by the ligaments, and gingival tissues were separated from the
Animal Ethics Committee of Central South University and extracted wisdom tooth, then digested for 1 hour at 37
C in a
conformed to ARRIVE guidelines (approval no. 2021407). solution of 3 mg/mL collagenase type I (#17018029, Gibco) and
C57BL/6 mice (2-month-old female) were purchased from the 4 mg/mL dispase (#D4693, Sigma-Aldrich, St. Louis, MO, USA).
SJA Laboratory (Changsha, China) and maintained in a The dissociated cells were seeded in culture plates with α-MEM
pathogen-free vivarium on a soft-food diet under a 12-hour containing 20% FBS and cultured for 3 days. The medium was
dark/light cycle at the Laboratory Animal Center of The Second replaced with α-MEM containing 10% FBS for continuous subcul-
Xiangya Hospital. After a 1-week acclimation period, the mice ture, and the cells were maintained in complete growth medium;
cells from passages 3 to 5 were used for experiments. Osteogenic
induction was carried out when cells reached 70% to 80%
Table 1. Small Interfering RNA Sequences confluence.(23)
Name Sequence (50 to 30 ) KFX and the sterile ethanol extract of P. americana L. (10)
were provided by GoodDoctor Pharmaceutical Group Co
siCCL2 Sense: GAAUCUGCAGCUAACUUAUTT
(Chengdu, China). KFX contains 0.72 mg/mL alanine, whereas
Antisense: AUAAGUUAGCUGCAGAUUCTT
the sterile ethanol extract of P. americana L. contains 10 times
scr-siRNA Sense: UUCUCCGAACGUGUCACGU
that amount; thus, alanine can be considered as an index of
Antisense: ACGUGACACGUUCGGAGAA
KFX concentration in a sample. KFX was used in experiments

Table 2. Primer Sequences for Quantitative Real-Time PCR

Gene Forward primer Reverse primer
Mouse
Runx2 CGGGCTACCTGCCATCAC GGCCAGAGGCAGAAGTCAGA
Col1α2 CAGAACATCACCTACCACTGCAA TTCAACATCGTTGGAACCCTG
Ccl2 CCACTCACCTGCTGCTACTCATTC CTTCTTTGGGACACCTGCTGCTG
Cxcl12 TCTGAAAATCCTCAACACTCCA CAGGTACTCTTGGATCCACTTT
β-actin CAACGAGCGGTTCCGATG GCCACAGGATTCCATACCCA
Human
RUNX2 CCTTCCAGACCAGCAGCACTC CCATCAGCGTCAACACCATCATC
COL1A2 CCGTGGCAGTGATGGAAGTGTG GCAGGACCAGCGTTACCAACAG
CCL2 AAGTCTCTGCCGCCCTTCTG GCGAGCCTCTGCACTGAGAT
CXCL12 TCAACACTCCAAACTGTGCCCTTC TCCACTTTAGCTTCGGGTCAATGC
GAPDH CAGGAGGCATTGCTGATGAT GAAGGCTGGGGCTCATTT

n2 GAO ET AL. Journal of Bone and Mineral Research

15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

(Figure legend continues on next page.)

n
3
KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING
Journal of Bone and Mineral Research
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
with mice after tooth extraction. To study the effect of KFX on observed under fluorescence microscopy (Eclipse C1, Nikon,
SCs, the sterile ethanol extract of P. americana L. was used in cell Tokyo, Japan).
cultures after dilution to 0.0008 mg/mL and was added to the
osteogenic induction medium in the KFX-treated group. HUVECs Scanning electron microscopy (SEM)
were treated with 0.0008 mg/mL KFX to study the direct effect of
KFX on HUVECs. After 2 days, mRNA and protein were extracted The surface of the extraction socket after treatment with KFX or
from the cells, and the CM of SCs were collected and stored at NS was examined by SEM. Briefly, alveolar bone was dissected
80
C until use. For Western blotting, CM was concentrated free of soft tissues and fixed with 2.5% glutaraldehyde (#G1102,
using a 3-kDa ultrafiltration centrifuge tube (UFC5003, Millipore, Servicebio, Wuhan, China). After rinsing with phosphate-
Billerica, MA, USA) at 12,000g and 4
C for 1 hour. The CM was buffered saline (PBS) and post-fixing in a solution of 1% osmium
concentrated approximately five times. tetroxide, the samples were dehydrated in a graded series of eth-
Small interfering RNA (siRNA) targeting chemokine (C-C motif) anol concentrations (50%, 60%, 70%, 80%, 90%, and 100%) and
ligand 2 (siCCL2) and scrambled negative control siRNA (scr- sputtered with thin gold film before examination with a Regulus
siRNA) were purchased from BioeGene (Shanghai, China) 8100 microscope (Hitachi, Tokyo, Japan) at 3.0 kV, 30  6000
(Table 1). For gene silencing experiments, hPDLSCs and hDPSCs magnification, and 12-mm working distance.
were seeded in 12-well plates at 1  105 cells/well 1 day before
transfection. Lipofectamine 2000 (#11668019; Invitrogen, Carls- Micro-computed tomography (μCT)
bad, CA, USA) and siRNA were separately added to 100 μL
Collected maxillae were removed, fixed with 4% formaldehyde
serum-free medium (Opti-MEM I, Invitrogen) and mixed after
for 48 hours, and stored in 70% ethanol at 4
C for bone mass
5 minutes. The 200-μL mixture was loaded into the wells of the
measurements. The maxillary tissue was scanned with a μCT
plate. For assessing the effect of CCL2, the human recombinant
scanner (Cheetah, YXLON International GmbH, Hamburg,
CCL2 (rhCCL2) (#HY-P7237; MedChemExpress, Monmouth Junc-
Germany) at an isotropic nominal resolution of 18 μm at 80 kV
tion, NJ, USA) were added into HUVECs in different concentra-
and 55.6 mA, with an exposure time of 763 seconds. Quantita-
tions (1, 5, 10, 50, and 100 ng/mL) to find the effective
tive analysis of bone regeneration in sockets was performed
concentration. Then, the indicated cells were treated with the
using V.G Studio v3.0 (Volume Graphics GmbH, Heidelberg,
effective concentration (100 ng/mL) of rhCCL2.
Germany). Bone volume/tissue volume ratio (BV/TV), trabecular
thickness (Tb.Th), trabecular number (Tb.N), and trabecular sepa-
Cell cytotoxicity and proliferation assay ration (Tb.Sp) were determined from 3-dimensional μCT images.

The effect of KFX on the viability and proliferation of mBMSCs, Histochemistry, immunohistochemistry, and
hPDLSCs, and hDPSCs was assayed using the Cell Counting
histomorphometry
Kit-8 (CCK-8) assay (Dojindo, Kumamoto, Japan) and 5-ethynyl-
20 -deoxyuridine (EdU) assay (Ribobio Co., Guangzhou, China) Alveolar bone was fixed in 4% of paraformaldehyde for 48 hours
according to the manufacturer’s instructions. For the CCK-8 and then decalcified in 10% EDTA (pH 7.4) for 28 days. Hematox-
assay, SCs were cultured with different concentrations (0.00016, ylin and eosin (H&E) staining was performed as previously
0.0008, 0.004, 0.02, 0.1, and 0.5 mg/mL) of KFX in 96-well plates described.(32) Decalcified 4-μm-thick alveolar bone sections were
for 72 hours to assess the cytotoxicity. SCs were cultured with deparaffinized with xylene and rehydrated in a graded series of
concentrations of 0, 0.00016, 0.0008, and 0.004 mg/mL of KFX, alcohol solutions. H&E staining (#G1120, Solarbio, Beijing,
and HUVECs were treated with 0.0008 mg/mL KFX to measure China) was performed according to the manufacturer’s instruc-
the proliferative effect at 24, 48, and 72 hours. At each time tions, and immunolabeling was performed according to a stan-
point, fresh 10% FBS α-MEM containing 10% CCK8 solution was dard protocol. Decalcified 4-μm-thick alveolar bone sections
added to each well. After incubation at 37
C for 4 hours, the were permeabilized with 0.1% Triton X, then incubated in 3%
absorbance value was measured at 450 nm. For the EdU assay, bovine serum albumin (BSA) solution for 1 hour at room temper-
SCs were cultured with different concentrations of KFX for ature, followed by overnight incubation at 4
C with primary anti-
72 hours. Then the hPDLSCs and hDPSCs were incubated with bodies against alpha smooth muscle actin (α-SMA) (1:50,
50 μM EdU at 37
C for 4 hours, and mBMSCs were incubated Proteintech, Rosemont, IL, USA), cluster of differentiation
with 10 μM EdU at 37
C for 24 hours. After fixation with 4% poly- 31 (CD31) (1:100; Abcam, Cambridge, UK), and endomucin
formaldehyde and permeabilized with 0.5% TritonX-100, the (1:100, Bioss, Beijing, China). The next day, the samples were
cells were stained by Apollo and Hoechst 33342 solutions at incubated with fluorophore-conjugated secondary antibodies
room temperature for 30 minutes. The stained cells were in the dark at room temperature for 1 hour. Samples incubated

(Figure legend continued from previous page.)

Fig. 1. KFX promotes gingival recovery, socket healing, and mineralization after tooth extraction. (A) Representative intraoperative image of the socket
defect created in the maxilla (upper jaw) of a mouse after extraction of the first molar. (B) Representative gingival images in the KFX and NS groups at
14 and 28 days post extraction. (C) Representative μCT images of new alveolar bone formation in each group at 14 and 28 days post extraction. (D) Ana-
lyses of BV/TV, Tb.Th, Tb.N, and Tb.Sp of the new bone area in MRs at 14 days and palatal roots at 28 days. (E) H&E staining. (F) SEM view of surface details
and collagen structure. Red arrow = collagen; green arrow = crystal structure. (G) Picrosirius red staining for type I collagen (red) and type III collagen
(green). (H) Goldner trichrome staining for mineralized bone (green) and new bone (red). Red boxes indicate the defect regions. Data are presented as
the mean  SD (n = 6). BV/TV = bone volume/tissue volume; DR = distal root; H&E = hematoxylin and eosin; KFX = kangfuxin; MR = mesial root; NS
= normal saline (negative control); PR = palatal root; Tb.N = trabecular number; Tb.Sp = trabecular separation; Tb.Th = trabecular thickness.

n4 GAO ET AL. Journal of Bone and Mineral Research

15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

(Figure legend continues on next page.)

n
5
KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING
Journal of Bone and Mineral Research
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
with only the secondary antibody served as the control. The sam- Transwell migration assay
ples were counterstained with 40 ,6-diamidino-2-phenylindole
(1:1000, Thermo Fisher Scientific, Waltham, MA, USA) for A transwell permeable support with an 8-μm polycarbonate
10 minutes at 37
C and visualized by fluorescence microscopy membrane insert (Costar; Corning, Corning, NY, USA) was used
(Eclipse C1; Nikon, Tokyo, Japan). to evaluate HUVEC migration. HUVECs were starved in serum-
The maxillae were fixed with 4% formaldehyde for 48 hours free RPMI 1640 medium for 12 hours at 37
C and 5% CO2, and
and stored in 70% ethanol at 4
C. Undecalcified sections were approximately 3–5  104 cells/mL in 200 μL of RPMI 1640
stained with picrosirius red and Goldner trichrome using com- medium containing 0.2% FBS were loaded into the upper cham-
mercial kits (#G1018 and #G1064, Servicebio) according to the ber of the insert, while the lower chamber was filled with 600 μL
manufacturer’s instructions. of 1% FBS RPMI 1640 medium containing CM, KFX, or CM
+ rhCCL2, respectively. Cells were incubated at 37
C and 5%
CO2 for 12 and 24 hours. Attached cells were fixed with 4% para-
formaldehyde for 30 minutes, washed twice with PBS, and
RNA-sequencing (RNA-seq) and bioinformatic analysis
stained with 0.1% crystal violet solution. To minimize bias,
Total RNA was extracted from mBMSCs using TRIzol reagent migrated cells were counted in at least three randomly selected
(Invitrogen) according to the manufacturer’s instructions. Each fields at 100 magnification under a microscope.
group included three mBMSCs samples. The dsDNA HS Assay
Kit for Qubit (#12640ES76, Yeasen, Shanghai, China) was used Tube formation
to determine the concentration and purity of RNA samples; the
Agilent High Sensitivity DNA Kit (#5067-4626, Agilent Technolo- HUVECs were starved in serum-free RPMI 1640 medium at 37
C
gies, Santa Clara, CA, USA) was used to evaluate the quality of in a humidified atmosphere of 5% CO2 for 12 hours, and
the library; and the Agilent 2100 Bioanalyser (Agilent Technolo- 5  104 cells/well were loaded in 48-well plates precoated with
gies) was used to quantify library integrity and size. The mRNA 150 μL/well Matrigel matrix (#354234, Corning). After incubation
was purified through two rounds of hybridization to Dynal Oligo for 12 hours in 10% FBS RPMI 1640 medium containing CM, KFX,
beads (#N411-03, Vazyme Biotech, Nanjing, China). After remov- or CM + rhCCL2, respectively, tube morphology was examined
ing rRNAs and applying fragmentation buffer (#AM8740, Invitro- and photographed using an optical microscope (Zeiss, Wetzlar,
gen), cDNA was synthesized from the fragments using random Germany). The length and density of tubular structures were
primers. The purified fragments were amplified by PCR. Library quantified.
preparation was performed using the VAHTS mRNA-seq V3
Library Prep Kit for Illumina (#NR611, Vazyme Biotech). For func- Quantitative real-time PCR (qRT-PCR)
tional analyses, differentially expressed genes (DEGs) were
mapped to terms in the Gene Ontology (GO) database (and clas- Total RNA was extracted from mBMSCs, hPDLSCs, and hDPSCs
sified into biological process [BP], cellular component [CC], and using TRIzol reagent (Invitrogen). The absorbance of the RNA
molecular function [MF] categories) and Kyoto Encyclopedia of was measured at 260 nm (NanoDrop 2000, Thermo Fisher Scien-
Genes and Genomes (KEGG) database with p < 0.05 and fold tific). The RNA was reverse-transcribed using the PrimeScript RT
change >2 as the threshold. Raw sequencing reads have been kit (Takara, Otsu, Japan) and qRT-PCR performed with the SYBR
deposited to NCBI GenBank under BioProject PRJNA777735. Premix Ex Taq II kit (Takara) using the primers in Table 2
(Shanghai Generay Biotech Co., Shanghai, China). Cycling condi-
tions for PCR (LightCycler Real-Time PCR System, Roche, Basel,
Wound-healing assay Switzerland) were as follows: 95
C for 30 seconds and 40 cycles
at 95
C for 5 seconds and 60
C for 20 seconds. Melting curves
HUVECs were seeded in 12-well plates at a density of 2  105 were used to confirm primer specificity, and the 2ΔΔCt method
cells/well and cultured to 100% confluence. After washing twice was used to calculate target gene expression levels. The mouse
with PBS, the cells were starved in serum-free Roswell Park and human internal control genes were β-actin and glyceralde-
Memorial Institute (RPMI) 1640 medium (Thermo Fisher Scien- hyde 3-phosphate dehydrogenase, respectively.
tific) for 12 hours at 37
C and 5% CO2. A scratch was made
through the cell monolayer with a 200-μL pipet tip, and after
Western blotting
two washes with PBS, 1 mL of 10% FBS RPMI 1640 medium sup-
plemented with CM, KFX, or CM + rhCCL2, respectively, was Western blotting was performed as previously described.(33)
added for 6 and 12 hours at 37
C and 5% CO2. The cells were Briefly, cells were lysed with radioimmunoprecipitation assay
photographed and the migration area of the HUVECs was mea- buffer (#78501, Thermo Fisher Scientific) containing protease
sured from the images. and phosphatase inhibitors. Protein concentrations were

(Figure legend continued from previous page.)

Fig. 2. DEGs identified by RNA-sequencing. (A) Heat map and (B) volcano map of DEGs between KFX- and NS-treated mBMSCs. Red areas and dots in the
figure represent genes with high expression, whereas blue or green areas represent genes with low expression. (C) Significantly enriched GO terms in the
BP, CC, and MF categories. (D) Significantly enriched KEGG pathways of 25 DEGs. (E–M) Effects of KFX and NS on Runx2, Col1α2, and Ccl2 (E–G) mRNA and
(H–M) protein levels in mBMSCs, hPDLSCs, and hDPSCs. Data are presented as mean  SD (n = 3). BP = biological process; CC = cellular component; Ccl2
= chemokine (C-C motif) ligand 2; Col1α2 = collagen type I alpha 2; DEG = differentially expressed gene; GO = Gene Ontology; hDPSCs = human dental
pulp stem cells; hPDLSCs = human periodontal ligament stem cells; KEGG = Kyoto Encyclopedia of Genes and Genomes; KFX = kangfuxin; MF = molec-
ular function; mBMSCs = mouse mandibular bone marrow mesenchymal stem cells; NS = normal saline; Runx2 = runt-related transcription factor 2.

n6 GAO ET AL. Journal of Bone and Mineral Research

15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

(Figure legend continues on next page.)

n
7
KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING
Journal of Bone and Mineral Research
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
determined with BCA Assay Kits (#P0010, Beyotime, Beijing, with the NS group (Fig. 1C, D); there was no difference in MR
China), and proteins in the samples were separated by sodium bone mass between the two groups at 28 days (Supplemental
dodecyl sulfate–polyacrylamide electrophoresis on a 12% gel Fig. S1). No significant differences were observed between the
and transferred to a polyvinylidene fluoride membrane (EMD NS and NI groups in the MR bone mass at 14 days
Millipore) that was blocked with 5% skim milk powder for 1 hour (Supplemental Fig. S1). Histologic analysis of bone specimens
at room temperature and incubated overnight at 4
C in 3% BSA revealed differences in bone structure and bone mass between
(#9048-46-8, Sigma-Aldrich) in Tris-buffered saline with Tween KFX and NS control group by H&E staining (Fig. 1E). At 14 days,
20 supplemented with primary antibodies against runt-related immature cross-linked collagen fibers were observed in the NS
transcription factor 2 (RUNX2) (#ab23981, Abcam), COL1A2 group by SEM, whereas the KFX group showed crystallization
(#DF3549, Affinity Biosciences, Cincinnati, OH, USA), CCL2 on the collagen surface, which was observed only after 28 days
(#DF7577-BP, Affinity Biosciences), C-X-C motif chemokine in the control mice. Meanwhile, at the 28-day time point, the
ligand 1 (CXCL12) (#AF5166, Affinity Biosciences), and β-actin socket site was covered by mature, compacted, and regularly
(#AM1021B, Abcepta, Suzhou, China). Protein bands were arranged collagen fibers in the KFX group (Fig. 1F). Type I colla-
detected using a peroxidase-conjugated anti-rabbit or mouse gen (COLI) appeared bright red or strong orangish yellow,
secondary antibody (1:5000, Bioss) and visualized by enhanced whereas type III collagen (COLIII) appeared green on picrosirius
chemiluminescence (GE Healthcare, Little Chalfont, UK). red staining and polarization microscopy. At 14 days, there was
obvious collagen fiber formation in the KFX group; at 28 days,
Statistical analysis there were more red and orangish yellow areas in the KFX group
but more green areas in the NS group (Fig. 1G). Further analysis
All values are presented as mean  standard deviation. Statisti- with Goldner’s trichrome staining, with red representing incom-
cal analyses were performed using Prism software v8.0 plete calcification of the bone matrix and green representing
(GraphPad, La Jolla, CA, USA). Results were analyzed by Student’s mineralized bone matrix, identified more mineralized alveolar
t test, one-way analysis of variance (ANOVA) with Tukey’s multi- bone at the socket site in the KFX group at 14 and 28 days
ple comparisons tests. p < 0.05 was considered significant. (Fig. 1H).

Results KFX alters the expression of chemokine-related genes but

not osteogenic differentiation markers
KFX promotes tooth socket healing in a mouse tooth
To detect the cytotoxic effect of KFX, SCs were treated with KFX
extraction model
(0.00016, 0.0008, 0.004, 0.02, 0.1, and 0.5 mg/mL) for 72 hours
We established a tooth extraction model by extracting the M1 of and detected by the CCK-8 assay. In mBMSCs, concentrations
mice (Fig. 1A). There was no significant difference in body weight of 0.00016, 0.0008, and 0.004 mg/mL had no difference com-
between KFX and the NS control group (Supplemental Fig. S1). pared with 0 mg/mL, whereas concentrations of 0.02, 0.1, and
Based on the healing condition of the gingiva at 1, 3, 5, 7 0.5 mg/mL showed cytotoxicity. In hPDLSCs and hDPSCs, only
(Supplemental Fig. S1), 14, and 28 days (Fig. 1B), no significant concentrations of 0.1 and 0.5 mg/mL showed cytotoxicity,
differences were observed between the two groups during the whereas others showed no significant difference compared with
first 3 days. Compared with the NS group, the results of wound 0 mg/mL. Then, a non-cytotoxic concentration was selected to
opening distance and thickness of gingival epithelium showed verify the proliferation effect of KFX on cells. Both the CCK-8
that the gingival healing in the KFX group was better at 5, 7, assay and EdU assay showed that KFX did not affect the prolifer-
and 14 days. Both groups were completely healed in 28 days. ation of mBMSCs, hPDLSCs, and hDPSCs (Supplemental Fig. S2).
There was no significant difference in gingival healing between To determine whether KFX promotes extraction socket heal-
NS and the non-intervention control (NI) group (Supplemental ing via oral cavity-specific SCs, mBMSCs were cultured and trea-
Fig. S1). μCT scanning showed greater alveolar bone mass in ted with 0.0008 mg/mL KFX or PBS under osteogenic induction
the extraction socket in the KFX group than in the NS group. for 3 days, and DEGs were identified by RNA-seq. There were
KFX treatment increased the bone volume/tissue volume ratio 25 DEGs (|fold change| > 2 and p < 0.05) including 21 that were
(BV/TV) and trabecular number (Tb.N), and significantly upregulated and 4 that were downregulated in the KFX
decreased trabecular separation (Tb.Sp) of the mesial root group compared with the NC group (Fig. 2A, B). To assess the bio-
(MR) at 14 days and of the palatal root (PR) at 28 days compared logical significance of these DEGs, we carried out GO and KEGG

(Figure legend continued from previous page.)

Fig. 3. KFX-CM of hPDLSCs and hDPSCs enhances HUVECs migration and angiogenesis. (A, B) Effects of KFX and NS on the protein levels of RUNX2,
COL1A2, and CCL2 in mBMSCs, hPDLSCs, and hDPSCs grown in CM. (C) Representative bright-field light micrographs of the wound-healing assay using
HUVECs treated with KFX-CM or NC-CM from hPDLSCs and hDPSCs cultures for 6 and 12 hours. (D) Quantitative analysis of migration rates in (C). (E, G)
Migratory capacity of HUVECs under different treatments evaluated with the transwell assay. (F, H) Quantitative analysis of migrated cells in (E, G). (I) Angio-
genic capacity of HUVECs in KFX-CM and NC-CM assessed with the tube formation assay. (J) Quantitative analysis of the total tube length in (I). (K) Immu-
nofluorescence micrographs of CD31 (green) and Emcn (red) expression in the socket; nuclei are stained with DAPI (blue). (L) α-SMA expression in sockets.
Data are presented as mean  SD (n = 3 per group). α-SMA = alpha-smooth muscle actin; CD31 = cluster of differentiation 31; CM = conditioned
medium; DAPI = 40 ,6-diamidino-2-phenylindole; Emcn = endomucin; HUVECs = human umbilical vein endothelial cells; hDPSCs = human dental pulp
stem cells; hPDLSCs = human periodontal ligament stem cells; KFX-CM = kangfuxin-treated conditioned medium; NC-CM = PBS-treated conditioned
medium; SCs = stem cells.

n8 GAO ET AL. Journal of Bone and Mineral Research

15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

(Figure legend continues on next page.)

n
9
KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING
Journal of Bone and Mineral Research
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
pathway analyses. The most enriched GO terms in the BP cate- and angiogenesis, albeit not significantly (Supplemental
gory were chemokine-mediated signaling pathway and chemo- Fig. S4A–D). To exclude the promoting effect of the possible
taxis; the most enriched CC terms were extracellular space and residual KFX in CM on endothelial cell function, we treated
extracellular region; and the most enriched MF terms were C-C HUVECs with KFX or PBS directly. Compared with the PBS
chemokine receptor type 2 (CCR2) binding, chemokine activity, group, CCK-8 assay indicated that KFX did not affect the prolifer-
and CCR binding (Fig. 2C). In the KEGG pathway analysis, the ation of the HUVECs (Supplemental Fig. S5A). In addition, the
most enriched pathways were cytokine-cytokine receptor inter- wound-healing assay (Supplemental Fig. S5B), transwell assay
action and chemokine signaling pathway (Fig. 2D). Among the (Supplemental Fig. S5C), and tube formation assay (Supple-
chemokine-related genes Cxcl1, Ccl2, Ccl5, Ccl7, and Ccl12, Ccl2 mental Fig. S5D) indicated that HUVECs cultured with KFX
showed the greatest increase in expression in mBMSCs after showed lower motility and tube formation. Immunofluorescence
KFX treatment (Table 3) at both the mRNA (Fig. 2E) and protein labeling of α-SMA, CD31, and endomucin in the alveolar bone of
(Fig. 2H, K) levels, as determined by qRT-PCR and Western blot- KFX- and NS-treated mice confirmed that the former group had
ting, respectively. In contrast, the mRNA and protein expression more blood vessels and type H vessels (Fig. 3K, L). These results
of the osteogenic differentiation biomarkers RUNX2 and COL1A2 indicated that KFX promotes endothelial cell migration and
did not differ between groups (Fig. 2E, H, K), consistent with the angiogenesis.
RNA-seq results. Additionally, the mRNA (Fig. 2F, G) and protein
(Fig. 2I, J, L, M) levels of RUNX2, COL1A2, and CCL2 were upregu-
Migration and angiogenesis in HUVECs are inhibited by
lated in hPDLSCs and hDPSCs derived from the KFX group com-
pared with those from the NC group. These results suggest that Ccl2 knockdown in CM
CCL2 plays an important role in KFX-induced bone formation. To confirm the pro-angiogenic role of CCL2, we used siCCL2 to
deplete the transcript in hPDLSCs and hDPSCs and evaluated
CCL2 upregulation in KFX-treated SCs promotes migration HUVEC migration and angiogenesis. Compared with cells treated
and angiogenesis in HUVECs with scr-siRNA, Ccl2 mRNA levels were reduced by Ccl2 knock-
down (Fig. 4A). Western blot analysis showed that CCL2 protein
Because chemokine-related genes were found to be involved in expression was increased in CM of hPDLSCs and hDPSCs trans-
socket healing under KFX treatment, we analyzed the CM of fected with scr-siRNA under KFX treatment, whereas for cells
mBMSCs, hPDLSCs, and hDPSCs treated with 0.0008 mg/mL transfected with siCCL2, CCL2 expression was suppressed in
KFX (KFX-CM) or PBS (NC-CM) under osteogenic induction for the presence of KFX (Fig. 4B, C). The wound-healing (Fig. 4D–G),
2 days by Western blotting. The CCL2 protein level was increased transwell (Fig. 4H–K), and tube formation (Fig. 4L–O) assays
by 40% in mBMSC, 60% in hPDLSCs, and 95% in hDPSCs CM in showed that Ccl2 knockdown significantly impaired HUVEC
cultures treated with KFX compared with the NC group (Fig. 3A, migration and angiogenesis induced by KFX treatment. The
B). In addition, the expression of CXCL12 mRNA in oral cavity– 50 ng/mL and 100 ng/mL of rhCCL2 protein promoted the
specific SCs and the expression of CXCL12 protein in CM of oral migration of HUVECs (Fig. 5A, B). The deceased cell migration
cavity–specific SCs were detected (Supplemental Fig. S3A, B). (Fig. 5C–J) and tube formation (Fig. 5K–N) of HUVECs after CCL2
No significant difference was found between the KFX and NC knockdown could be reversed by treatment with rhCCL2, which
groups, which was consistent with the RNA-sequence results further confirms the role of CCL2 in HUVECs migration and
(Fig. 2B). To verify whether CCL2 expressed by KFX-treated SCs angiogenesis. These results indicate that KFX promotes endothe-
accelerates socket healing by promoting angiogenesis in endo- lial cell migration and angiogenesis via CCL2 signaling.
thelial cells, HUVECs were treated with the CM. In the wound-
healing assay at 6 and 12 hours (Fig. 3C, D) and transwell migra-
tion assay at 12 and 24 hours (Fig. 3E, H), the migratory capacity Discussion
of HUVECs treated with KFX-CM of hPDLSCs and hDPSCs was
increased compared with cells treated with NC-CM. The angio- In this study, we demonstrated that KFX has therapeutic poten-
genic potential of HUVECs was evaluated with the tube forma- tial for healing soft and hard tissues in murine extraction sockets.
tion assay. Compared with the NC group, longer capillary-like To clarify the mechanism of action of KFX in this process,
tubular structures were observed in KFX-CM-treated HUVECs mBMSCs were treated with 0.0008 mg/mL KFX, and DEGs were
(Fig. 3I); quantitative analysis showed that the lengths were identified by RNA-seq. We found that Ccl2 expression was upre-
increased by 35% and 100% with KFX-CM of hPDLSCs and gulated in mBMSCs, hPDLSCs, and hDPSCs and their CM after
hDPSCs, respectively, compared with treatment with NC-CM KFX treatment, which enhanced the migratory and angiogenic
(Fig. 3J). The KFX-CM of mBMSCs increased HUVEC migration capacities of HUVECs. The numbers of blood vessels and type H

(Figure legend continued from previous page.)

Fig. 4. Ccl2 knockdown abolishes KFX-induced HUVECs migration and angiogenesis. (A) Ccl2 mRNA expression in hPDLSCs and hDPSCs transfected with
siCCL2 or scr-siRNA. (B, C) CCL2 expression in hPDLSCs and hDPSCs cultured in KFX-CM and NC-CM after siRNA-mediated knockdown for 48 hours con-
firmed by Western blotting. (D, F) Wound-healing assay using HUVECs. (E, G) Quantitative analysis of migration rates in (D, F). (H, J) Migratory capacity of
HUVECs assessed with the transwell assay. (I, K) Quantitative analysis of migrated cells in (H, J). (L, N) Angiogenic capacity of HUVECs was further confirmed
by tube formation assays. (M, O) Quantitative analysis of the total tube length (L, N). Data are presented as mean  SD (n = 3 per group). CCL2 = chemo-
kine (C-C motif) ligand 2; CM = conditioned medium; HUVECs = human umbilical vein endothelial cells; hDPSCs = human dental pulp stem cells; hPDLSCs
= human periodontal ligament stem cells; KFX = kangfuxin; NS = normal saline; siCCL2 = small interfering RNA targeting CCL2; scr-siRNA = scrambled
negative control small interfering RNA.

n 10 GAO ET AL. Journal of Bone and Mineral Research

 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. 5. The rhCCL2 protein reverses HUVECs migration and angiogenesis inhibited by Ccl2 knockdown. (A) Representative cell migration of HUVECs after rhCCL2
treatment (1, 5, 10, 50, and 100 ng/mL) for 24 hours was detected using the transwell assay. (B) Quantitative analysis of migration rates in (A). (C, E) Wound-
healing assay using HUVECs. (D, F) Quantitative analysis of migration rates in (C, E). (G, I) Migratory capacity of HUVECs assessed with the transwell assay. (H,
J) Quantitative analysis of migrated cells in (G, I). (K, M) Angiogenic capacity of HUVECs was further confirmed by tube formation assays. (L, N) Quantitative analysis
of the total tube length (K, M). Data are presented as mean  SD (n = 3 per group). CM = conditioned medium; HUVECs = human umbilical vein endothelial
cells; hDPSCs = human dental pulp stem cells; hPDLSCs = human periodontal ligament stem cells; KFX = kangfuxin; NS = normal saline; rhCCL2 = recombinant
human chemokine (C-C motif) ligand 2; siCCL2 = small interfering RNA targeting CCL2; scr-siRNA = scrambled negative control small interfering RNA.

vessels were also increased at the socket site. Collectively, these extended those findings in our study that showed that KFX pro-
results indicate that KFX induces CCL2 expression in SCs to pro- moted the healing of gingiva after tooth extraction
mote angiogenesis in vascular endothelial cells, ultimately accel- (Supplemental Fig. S1). We confirmed here by μCT analysis that
erating socket healing. KFX accelerated socket healing; SEM (Fig. 1) showed collagen for-
KFX has been used in TCM for wound healing, as a gastropro- mation and crystallization, reflecting mineralized and mature
tective agent, and for ulcer treatment.(13,15,16) Others have collagen.(34) These results suggest that KFX promotes bone
shown that KFX can promote wound healing,(13) and we repair by inducing the formation of bone extracellular matrix.

Journal of Bone and Mineral Research KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING 11 n
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 3. Chemokine-Related Differentially Expressed Genes instead exerts effects via an indirect mechanism, which might
Gene Fold change Direction of change p Value be associated with the alteration of cellular chemotaxis caused
by changes in CCL2. In addition, it is worth noting that the
Ccl2 3.758596 Up 0.003759 CXCL12, a common chemokine-related factor, did not show a
Ccl7 3.331355 Up 1.01E06 significant difference, which might be related to different path-
Ccl5 2.390792 Up 0.046266 ways regulated by different drugs.
Ccl12 2.307059 Up 0.042549 Ccl2 is located on chromosome 17 (q11.2) and encodes a
Cxcl1 2.136252 Up 0.012977 76-amino acid protein with a size of 13 kDa that is involved in
the migration of endothelial cells and monocytes.(41,44) Ccl2
directly promotes angiogenesis, and upregulation of Ccl2 signal-
In fact, KFX was previously shown to facilitate collagen formation ing may enhance early fracture healing and bone regenera-
through the TGF-β1/Mothers against decapentaplegic homolog tion.(45,46) We showed that KFX treatment increased CCL2
2 (Smad2)/Smad4 signaling pathway.(35) We also analyzed the protein expression in oral cavity–specific SCs and their
type of collagen that was deposited by picrosirius red staining CM. Bone is a highly vascularized connective tissue, and angio-
and found that mice treated with KFX had more COLI and less genesis accelerates bone repair because blood vessels deliver
COLIII than those treated with NS (Fig. 1H). During bone forma- oxygen and nutrients that promote bone cell survival and activ-
tion, COLIII is expressed in newly synthesized bone and is later ity; conversely, poor vascularization, including decreases in
replaced by COLI,(36) which constitutes >90% of bone matrix pro- blood vessel area, distribution, and number in the extraction
tein. Goldner trichrome staining showed more mineralized bone socket, can delay bone regeneration.(47) Thus, therapeutic strate-
in the KFX group. In another study, alizarin red S staining gies for bone regeneration should focus on enhancing vascular-
revealed a large area of calcium nodules in osteoblasts incubated ization.(18) Chemokines potentially mediate vessel development
in KFX.(17) Additionally, KFX suppressed inflammation and apo- during osteogenesis.(48) Type H vessels are a type of capillary
ptosis and promoted collagen production in human dermal characterized by high expression of CD31 and endomucin and
fibroblasts.(7) Interestingly, there was no statistically significant are closely associated with osteogenesis.(49) Type H vessels are
difference between groups in terms of new bone formation in present in alveolar bone.(50) P. americana L. is the main active
the MR, whereas in the PR, more newly synthesized bone was component of KFX, which has been shown to improve local
observed in the KFX group, whereas the NS group showed per- blood circulation and promote wound healing.(17,35) SCs have
sistent bone defects at 28 days. We conclude that the healing the capacity to stimulate angiogenesis. Therefore, KFX may
rates of the two root types differ, with the MR healing more rap- accelerate socket healing by enhancing the angiogenic potential
idly than the PR. of HUVECs via the upregulation of CCL2 in SCs. This was sup-
BMSCs, PDLSCs, and DPSCs are oral cavity–specific SCs that ported by our observation that knocking down Ccl2 expression
have an important function in bone-defect healing and tissue abrogated the ability of KFX to promote HUVEC migration and
regeneration.(28,30,37) We hypothesized that KFX might promote angiogenesis (Fig. 4), which could be reversed after the supple-
socket healing by affecting SCs. Of these, BMSCs have been mentation with exogenous rhCCL2 protein (Fig. 5). In addition,
proven to play a crucial role in alveolar bone regeneration. Du our study also found that direct intervention of KFX can inhibit
and colleagues showed that BMSCs improved tissue and bone HUVEC migration and angiogenesis, suggesting that the increase
regeneration in periodontal alveolar bone regeneration by inhi- of HUVEC migration and angiogenesis was not caused by the
biting the expression of TNF-α, interferon-gamma (IFN-γ), and direct effect of KFX on HUVECs, which was similar to the study
interleukin-1beta (IL-1β).(20) Unlike BMSCs, PDLSCs and DPSCs of Xie and colleagues.(51) CM of mBMSCs in the KFX-treated
are confined to the alveolar socket site, and some are removed group did not induce HUVEC migration (Supplemental Fig. S4),
with the extracted tooth. We therefore used KFX-treated which may be attributable to a structural difference between
mBMSCs for the RNA-seq analysis and validated the results of human and mouse CCL2 or the incompatibility of mouse CCL2
RNA-seq analysis in mBMSCs, hPDLSCs, and hDPSCs simulta- and the receptor expressed by HUVECs. A previous study
neously. We identified five upregulated chemokine-related reported that the sequence homology between human CCL2
genes that potentially mediate the effect of KFX on bone repair and mouse CCL2 is only 55%.(44)
(Fig. 2A and Table 3). Chemokines are involved in the immune It should be noted that the therapeutic potential of KFX was
response, development, inflammation, and angiogenesis.(38) performed in a mouse tooth extraction model. To evaluate the
CCR2 expressed on the surface of monocytes/macrophages, leu- clinical applicability of KFX, the ability of KFX to promote extrac-
kocytes, smooth muscle cells, and endothelial cells have six tion socket healing should be further investigated in human sub-
ligands, including CCL2, CCL7, CCL8, CCL12, CCL13, and jects. In addition, we speculate that KFX may also have
CCL16.(39-42) CCL2, CCL7, and CCL12 were identified in the DEGs. therapeutic potential for infected extraction sockets. Pro-
Of these, CCL2 binds more strongly to the receptor,(39) and CCL2/ inflammatory cytokines such as interleukin-6 (IL-6), IL-1β, and
CCR2 signaling is essential in the early stage of fracture heal- TNF-α rise in inflamed gingival tissues, stimulating osteoclast
ing.(43) The osteogenic differentiation biomarkers RUNX2 and activity and resulting in alveolar bone resorption.(52) KFX has
COL1A2 showed no difference in expression in mBMSCs been demonstrated to have potential in suppressing inflamma-
between the KFX and NC groups but were upregulated in tion by inhibiting the expression of IL-6, IL-1β, and TNF-α.(53)
hPDLSCs and hDPSCs (Fig. 2). KFX was shown to increase OCN Therefore, the therapeutic role of KFX in the infected extraction
protein levels but inhibit ALP activity in osteoblasts in a socket is worthy of further study.
concentration-dependent manner.(17) This may reflect a differ- In summary, the results of this study highlight a new applica-
ence in the ability of KFX to promote osteogenic differentiation tion for an agent that is widely used in TCM. KFX was shown to
in different cell types. Thus, KFX may not promote alveolar bone promote alveolar bone remodeling after tooth extraction by
healing by directly inducing osteogenic differentiation but increasing CCL2 expression in BMSCs, PDLSCs, and DPSCs,

n 12 GAO ET AL. Journal of Bone and Mineral Research

 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
leading to bone regeneration and angiogenesis in the extraction 5. Schneider A, Burr R, Garbis N, Salazar D. Platelet-rich plasma and the
socket. shoulder: clinical indications and outcomes. Curr Rev Musculoskelet
Med. 2018;11(4):593–597.
6. Mao A, Mooney D. Regenerative medicine: current therapies and
Acknowledgments future directions. Proc Natl Acad Sci U S A. 2015;112(47):14452–
14459.
This study was supported by grants from the National Natural 7. He Y, Zhao W, Dong Z, et al. A biodegradable antibacterial
Science Foundation of China (nos. 81800788 and 81773339); Sci- alginate/carboxymethyl chitosan/Kangfuxin sponges for promoting
ence and Technology Department of Hunan Province, China blood coagulation and full-thickness wound healing. Int J Biol Macro-
(nos. 2018SK52511 and 2017WK2041); the Fund for Xiangya Clin- mol. 2021;167:182–192.
ical Medicine Database of Central South University 8. Han Y, Sun H, Zhang A, Yan G, Wang X. Chinmedomics, a new strat-
(no. 2014-ZDYZ-1-16); and the Open Sharing Fund for Large- egy for evaluating the therapeutic efficacy of herbal medicines. Phar-
macol Ther. 2020;216:107680.
scale Instruments and Equipment of Central South University.
9. Gao ZR, Feng YZ, Zhao YQ, et al. Traditional Chinese medicine pro-
motes bone regeneration in bone tissue engineering. Chin Med.
Author Contributions 2022;17(1):86.
10. Yang H, Wen Q, Xue J, Ding Y. Alveolar bone regeneration potential
Zheng-Rong Gao: Methodology; investigation; project of a traditional Chinese medicine, Bu-Shen-Gu-Chi-Wan, in experi-
administration; writing – original draft; writing – review and mental periodontitis. J Periodontal Res. 2014;49(3):382–389.
editing. Ying-Hui Zhou: Methodology; investigation; project 11. Liu J, Mattheos N, Su C, et al. The effects of icariin on wound healing
administration; writing – original draft; writing – review and of extraction sites with administration of zoledronic and dexametha-
sone: a rat model study. J Oral Pathol Med. 2018;47(2):198–205.
editing. Ya-Qiong Zhao: Investigation; validation; software.
Jie Zhao: Software; validation; investigation. Qin Ye: Software; 12. Zhang K, Wang X, Zhang W, Zhao JZ, Dong H. Effect of Yunnan
Baiyao capsules on the socket healing of impacted mandibular
investigation; validation. Shao-Hui Zhang: Data curation;
third molar extraction. Zhonghua Kou Qiang Yi Xue Za Zhi. 2012;
formal analysis. Yao Feng: Data curation; formal analysis. 47(4):199–202.
Li Tan: Data curation; formal analysis. Qiong Liu: Data cura-
13. Li LJ, Xu XH, Yuan TJ, Hou J, Yu CL, Peng LH. Periplaneta Americana L
tion; formal analysis. Yun Chen: Formal analysis; visualization. as a novel therapeutics accelerates wound repair and regeneration.
Ze-Yue Ouyang: Formal analysis; visualization. Jing Hu: For- Biomed Pharmacother. 2019;114:108858.
mal analysis; visualization. Marie Aimee Dusenge: Formal 14. Lu S, Wu D, Sun G, et al. Gastroprotective effects of Kangfuxin against
analysis; visualization. Yun-Zhi Feng: Conceptualization; water-immersion and restraint stress-induced gastric ulcer in rats:
supervision; resources; funding acquisition. Yue Guo: Concep- roles of antioxidation, anti-inflammation, and pro-survival. Pharm
tualization; supervision; resources; funding acquisition. Biol. 2019;57(1):770–777.
15. Shen Y, Sun J, Niu C, et al. Mechanistic evaluation of gastroprotective
effects of Kangfuxin on ethanol-induced gastric ulcer in mice. Chem
Peer Review Biol Interact. 2017;273:115–124.
16. Qu K, Li Y, Liang Y, et al. KangFuXin liquid in the treatment of diabetic
The peer review history for this article is available at https:// foot ulcer: a systematic review and meta-analysis. Evid Based Com-
www.webofscience.com/api/gateway/wos/peer-review/10. plement Alternat Med. 2019;2019:3678714.
1002/jbmr.4860. 17. Huang YF, Li LJ, Gao SQ, et al. Evidence based anti-osteoporosis
effects of Periplaneta Americana L on osteoblasts, osteoclasts, vascu-
lar endothelial cells and bone marrow derived mesenchymal stem
Data Availability Statement cells. BMC Complement Altern Med. 2017;17(1):413.
18. Wu V, Helder MN, Bravenboer N, et al. Bone tissue regeneration in the
The data that supports the findings of this study are available in oral and maxillofacial region: a review on the application of stem cells
the supplementary material of this article. and new strategies to improve vascularization. Stem Cells Int. 2019;
2019:6279721.
Disclosures 19. Wang J, Liu S, Li J, Zhao S, Yi Z. Roles for miRNAs in osteogenic differ-
entiation of bone marrow mesenchymal stem cells. Stem Cell Res
Ther. 2019;10(1):197.
The authors declare no conflicts of interest.
20. Du J, Shan Z, Ma P, Wang S, Fan Z. Allogeneic bone marrow mesen-
chymal stem cell transplantation for periodontal regeneration.
References J Dent Res. 2014;93(2):183–188.
21. Zhang M, Yu W, Niibe K, et al. The effects of platelet-derived growth
1. Akashi M, Kishimoto M, Kusumoto J, Yakushijin K, Matsuoka H, factor-BB on bone marrow stromal cell-mediated vascularized bone
Komori T. Delayed socket healing after dental extraction in patients regeneration. Stem Cells Int. 2018;2018:3272098.
undergoing myelosuppressive chemotherapy for hematological
malignancy: incidence and risk factors. J Oral Maxillofac Surg. 2018; 22. Xu R, Shen X, Si Y, et al. MicroRNA-31a-5p from aging BMSCs links
76(10):2057–2065. bone formation and resorption in the aged bone marrow microenvi-
ronment. Aging Cell. 2018;17(4):e12794.
2. Chappuis V, Araújo MG, Buser D. Clinical relevance of dimensional
bone and soft tissue alterations post-extraction in esthetic sites. Per- 23. Guo Y, Tang CY, Man XF, et al. Insulin receptor substrate-1 time-
iodontol. 2017;73(1):73–83. dependently regulates bone formation by controlling collagen Iα2
expression via miR-342. FASEB J. 2016;30(12):4214–4226.
3. Papi P, Di Murro B, Tromba M, Passarelli P, D’Addona A, Pompa G. The
use of a non-absorbable membrane as an occlusive barrier for alveo- 24. Xu H, Zhou S, Qu R, et al. Icariin prevents oestrogen deficiency-
lar ridge preservation: a one year follow-up prospective cohort study. induced alveolar bone loss through promoting osteogenesis via
Antibiotics. 2020;9(3):110. STAT3. Cell Prolif. 2020;53(2):e12743.
4. McNeill EP, Zeitouni S, Pan S, et al. Characterization of a pluripotent 25. Liu L, Guo S, Shi W, et al. Bone marrow mesenchymal stem cell-
stem cell-derived matrix with powerful osteoregenerative capabili- derived small extracellular vesicles promote periodontal regenera-
ties. Nat Commun. 2020;11(1):3025. tion. Tissue Eng, Part A. 2021;27(13–14):962–976.

Journal of Bone and Mineral Research KANGFUXIN ACCELERATES EXTRACTION SOCKET HEALING 13 n
 15234681, 0, Downloaded from https://asbmr.onlinelibrary.wiley.com/doi/10.1002/jbmr.4860 by Central South University, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
26. Seo BM, Miura M, Gronthos S, et al. Investigation of multipotent post- 41. França C, Izar M, Hortêncio M, et al. Monocyte subtypes and the CCR2
natal stem cells from human periodontal ligament. Lancet. 2004; chemokine receptor in cardiovascular disease. Clin Sci. 2017;131(12):
364(9429):149–155. 1215–1224.
27. Gronthos S, Mankani M, Brahim J, Robey PG, Shi S. Postnatal human 42. White G, Iqbal A, Greaves D. CC chemokine receptors and chronic
dental pulp stem cells (DPSCs) in vitro and in vivo. Proc Natl Acad Sci inflammation—therapeutic opportunities and pharmacological chal-
U S A. 2000;97(25):13625–13630. lenges. Pharmacol Rev. 2013;65(1):47–89.
28. Nagata M, Iwasaki K, Akazawa K, et al. Conditioned medium from 43. Ishikawa M, Ito H, Kitaori T, et al. MCP/CCR2 signaling is essential for
periodontal ligament stem cells enhances periodontal regeneration. recruitment of mesenchymal progenitor cells during the early phase
Tissue Eng Part A. 2017;23:367–377. of fracture healing. PLoS One. 2014;9(8):e104954.
29. Pei X, Wang L, Chen C, Yuan X, Wan Q, Helms J. Contribution of the 44. Van Coillie E, Van Damme J, Opdenakker G. The MCP/eotaxin sub-
PDL to osteotomy repair and implant osseointegration. J Dent Res. family of CC chemokines. Cytokine Growth Factor Rev. 1999;10(1):
2017;96(8):909–916. 61–86.
30. Lee Y, Chan Y, Hsieh S, Lew W, Feng S. Comparing the osteogenic 45. Hankenson K, Gagne K, Shaughnessy M. Extracellular signaling mole-
potentials and bone regeneration capacities of bone marrow and
cules to promote fracture healing and bone regeneration. Adv Drug
dental pulp mesenchymal stem cells in a rabbit calvarial bone defect
Delivery Rev. 2015;94:3–12.
model. Int J Mol Sci. 2019;20(20):5015.
46. Korbecki J, Kojder K, Simi
nska D, et al. CC chemokines in a tumor: a
31. Yang X, Yang K, An Y, Wang L, Luo G, Hu X. MicroRNA-705 regulates
the differentiation of mouse mandible bone marrow mesenchymal review of pro-cancer and anti-cancer properties of the ligands of
stem cells. PeerJ. 2019;7:e6279. receptors CCR1, CCR2, CCR3, and CCR4. Int J Mol Sci. 2020;21(21):
8412.
32. Chen Y, Guo Y, Li J, et al. Endoplasmic reticulum stress remodels alve-
olar bone formation after tooth extraction. J Cell Mol Med. 2020; 47. Filipowska J, Tomaszewski K, Niedźwiedzki Ł, Walocha J,
24(21):12411–12420. Niedźwiedzki T. The role of vasculature in bone development, regen-
eration and proper systemic functioning. Angiogenesis. 2017;20(3):
33. Li J, Guo Y, Chen Y, et al. miR-124-3p increases in high glucose 291–302.
induced osteocyte-derived exosomes and regulates galectin-3
expression: a possible mechanism in bone remodeling alteration in 48. Gilchrist A. Chemokines and bone. Handb Exp Pharmacol. 2020;262:
diabetic periodontitis. FASEB J. 2020;34(11):14234–14249. 231–258.
34. Feng P, Wei P, Shuai C, Peng S. Characterization of mechanical and 49. Kusumbe A, Ramasamy S, Adams R. Coupling of angiogenesis and
biological properties of 3-D scaffolds reinforced with zinc oxide for osteogenesis by a specific vessel subtype in bone. Nature. 2014;
bone tissue engineering. PLoS One. 2014;9(1):e87755. 507(7492):323–328.
35. Chen Z, Hu Y, Li J, et al. A feasible biocompatible hydrogel film 50. Yan Z, Wang X, Zhou Y, et al. H-type blood vessels participate in alve-
embedding Periplaneta Americana extract for acute wound healing. olar bone remodeling during murine tooth extraction healing. Oral
Int J Pharm. 2019;571:118707. Dis. 2020;26(5):998–1009.
36. Miedel E, Brisson B, Hamilton T, et al. Type III collagen modulates frac- 51. Huang J, Yin H, Rao SS, et al. Harmine enhances type H vessel forma-
ture callus bone formation and early remodeling. J Orthop Res. 2015; tion and prevents bone loss in ovariectomized mice. Theranostics.
33(5):675–684. 2018;8(9):2435–2446.
37. Aghaloo TL, Chaichanasakul T, Bezouglaia O, et al. Osteogenic poten- 52. Yamaguchi M, Fukasawa S. Is inflammation a friend or foe for ortho-
tial of mandibular vs. long-bone marrow stromal cells. J Dent Res. dontic treatment?: inflammation in orthodontically induced inflam-
2010;89(11):1293–1298. matory root resorption and accelerating tooth movement. Int J Mol
38. Brylka L, Schinke T. Chemokines in physiological and pathological Sci. 2021;22(5):2388.
bone remodeling. Front Immunol. 2019;10:2182. 53. Xue P, Wang L, Xu J, et al. Temperature-sensitive hydrogel for rectal
39. Hao Q, Vadgama J, Wang P. CCL2/CCR2 signaling in cancer patho- perfusion improved the therapeutic effect of Kangfuxin liquid on
genesis. Cell Commun Signal. 2020;18(1):82. DSS-induced ulcerative colitis mice: the inflammation alleviation
40. Zlotnik A, Yoshie O. The chemokine superfamily revisited. Immunity. and the colonic mucosal barriers repair. Int J Pharm. 2020;589:
2012;36(5):705–716. 119846.

n 14 GAO ET AL. Journal of Bone and Mineral Research