![Effects of titanium nanosurfaces on development of osteocytic network for osseointegration . (A) Scanning electron microscopy (SEM) images for titanium mini-implants with micro-roughened (MR) or nano-roughened (NR) surface, and a photograph showing implant placements on the right or left upper first molar region in an 11-week-old rat. (B) Reverse torque curve around the inflection point (black triangle) indicating osseointegration strength. (C) Osseointegration strength evaluated by reverse torque test for the MR and NR titanium implants 4 weeks postoperatively. (D) Representative images of nondecalcified resin-embedded histological sections with Villanueva-Goldner staining and the bone-implant contact ratio on peri-implant bone tissue around the MR or NR titanium implants 4 weeks postoperatively. (E) Representative SEM images of resin-cast etched sections of the supporting bone region or the bone-implant interface region in peri-implant bone tissue around the MR or NR titanium implants 4 weeks postoperatively, together with results of the aspect ratio and circularity of osteocytes (F) and the number of dendrites per osteocyte on the supporting bone region (G) or the bone-implant interface region (H) in peri-implant bone tissue. Arrowheads in (D) indicate spots of mineralized bone deposition on the implant surface (asterisk). Data are presented as means ± standard deviation (SD) ( N = 7 in [C], N = 6 in [D], N = 35 in [F and G], N = 14 in [H]). Asterisks indicate statistically significant differences ( P < 0.05; Student's t-test, Welch's t-test or Mann-Whitney U-test); n.s., no significant difference. Symbols: † , implant or implant surface; §, newly formed mineralized bone tissue around the MR titanium implants; ¶, newly formed mineralized bone tissue around the NR titanium implants. Note: (E) dense osteocyte networks in the supporting bone region (double arrowheads) and osteocytes adherent to the implant surface with numerous dendrites (triangles) in the bone-implant interface region in peri-implant bone tissue around the NR surface are evident.](https://www.researchgate.net/profile/Xindie-He/publication/355433906/figure/fig5/AS:11431281178486764@1690947038688/Effects-of-titanium-nanosurfaces-on-development-of-osteocytic-network-for_Q320.jpg)
Content uploaded by Xindie He
Author content
All content in this area was uploaded by Xindie He on Aug 02, 2023
Content may be subject to copyright.
Acta Biomaterialia 151 (2022) 613–627
Contents lists available at ScienceDirect
Acta Biomaterialia
journal homepage: www.elsevier.com/locate/actbio
Full length article
Titanium nanotopography induces osteocyte lacunar-canalicular
networks to strengthen osseointegration
Xindie He
a
, Masahiro Yam ada
a , ∗, Jun Watanab e
a
, Watcharaphol Tiskratok
a
,
Minoru Ishibashi
a
, Hideki Kitaura
b
, Itaru Mizoguchi
b
, Hiroshi Egusa
a
,
c
,
∗
a
Division of Molecular and Regenerative Prosthodontics, Toh oku University Graduate School of Dentistry, Sendai, Miyagi, Japan
b
Division of Orthodontics and Dentofacial Orthopedics, To hoku University Graduate School of Dentistry
c
Center for Advanced Stem Cell and Regene rative Research, Tohoku University Graduate School of Dentistry, Sendai, Miyagi, Japan
a r t i c l e i n f o
Article history:
Received 18 April 2022
Revised 9 August 2022
Accepted 12 August 2022
Available online 20 August 2022
Keywo rds:
Bone quality
Bone turnover
Gap junctions
Nanotopography
Cellular projections
3D culture
Surface modification
a b s t r a c t
Osteocyte network architecture is closely associated with bone turnover. The cellular mechanosensing
system regulates osteocyte dendrite formation by enhancing focal adhesion. Therefore, titanium surface
nanotopography might affect osteocyte network architecture and improve the peri-implant bone tissue
quality, leading to strengthened osseointegration of bone-anchored implants. We aimed to investigate
the effects of titanium nanosurfaces on the development of osteocyte lacunar-canalicular networks and
osseointegration of dental implants. Alkaline etching created titanium nanosurfaces with anisotropically
patterned dense nanospikes, superhydrophilicity, and hydroxyl groups. MLO-Y4 mouse osteocyte-like cells
cultured on titanium nanosurfaces developed neuron-like dendrites with increased focal adhesion as-
sembly and gap junctions. Maturation was promoted in osteocytes cultured on titanium nanosurfaces
compared to cells cultured on machined or acid-etched micro-roughened titanium surfaces. Osteocytes
cultured in type I three-dimensional collagen gels for seven days on nano-roughened titanium surfaces
displayed well-developed interconnectivity with highly developed dendrites and gap junctions compared
to the poor interconnectivity observed on the other titanium surfaces. Even if superhydrophilicity and
hydroxyl groups were maintained, the loss of anisotropy-patterned nanospikes reduced expression of gap
junction in osteocytes cultured on alkaline-etched titanium nanosurfaces. Four weeks after placing the ti-
tanium nanosurface implants in the upper jawbone of wild-type rats, osteocytes with numerous dendrites
were found directly attached to the implant surface, forming well-developed lacunar-canalicular networks
around the nano-roughened titanium implants. The osseointegration strength of the nano-roughened ti-
tanium implants was significantly higher than that of the micro-roughened titanium implants. These data
indicate that titanium nanosurfaces promote osteocyte lacunar-canalicular network development via nan-
otopographical cues and strengthen osseointegration.
Statement of significance
The clinical stability of bone-anchoring implant devices is influenced by the bone quality. The osteocyte
network potentially affects bone quality and is established by the three-dimensional (3D) connection of
neuron-like dendrites of well-matured osteocytes within the bone matrix. No biomaterials are known
to regulate formation of the osteocyte network. The present study provides the first demonstration that
titanium nanosurfaces with nanospikes created by alkali-etching treatment enhance the 3D formation
of osteocyte networks by promoting osteocyte dendrite formation and maturation by nanotopographic
cues, leading to strengthened osseointegration of titanium implants. Osteocytes attached to the titanium
nanosurfaces via numerous cellular projections. The success of osteocyte regulation by nanotechnology
paves the way for development of epoch-making technologies to control bone quality.
©2022 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
∗Corresponding authors at: Division of Molecular and Regenerative Prosthodon-
tics, Toh oku University Graduate School of Dentistry, 4-1 Seiryo-machi, Aoba-ku,
Sendai 980-8575, Japan.
E-mail addresses: masahiro.yamada.a2@tohoku.ac.jp (M. Yama da) ,
egu@tohoku.ac.jp (H. Egusa) .
https://doi.org/10.1016/j.actbio.2022.08.023
1742-7061/© 2022 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
1. Introduction
Osteocytes are the most abundant cells in bone tissue. These
cells orchestrate bone remodeling by controlling osteoblast and os-
teoclast activities in response to mechanical stress [1] . Mature os-
teocytes have neuron-like stellate shapes with polygonal or circu-
lar cell bodies and multiple long dendritic processes [2] . The os-
teocytes form a lacunar-canalicular network connecting the neigh-
boring osteocytes within the bone tissue and osteoblasts and os-
teoclasts on the bone surface [3] . Their morphology is correlated
with the alignment of the bone apatite c-axis and collagen fibers
[4] which influences bone quality [ 5 , 6 ]. In addition, the morphol-
ogy and alignment of osteocytes affect their mechanosensory na-
ture [7] . The cells function as mechanosensors and transducers
that initiate bone remodeling in response to mechanical stim-
ulation [8] . The osteocyte lacunar-canalicular network is closely
linked to the mechanoresponse of bone tissues [9] . Degeneration
of this network impairs fluid and solute transport and mechanore-
sponses of bone tissue, resulting in skeletal fragility with aging
[10] . Moreover, osteocytes may be involved in the remodeling of
peri-implant bone tissue under functional occlusal loading [11] . Os-
teocytes in peri-implant bone tissue contact implant surfaces with
their dendritic processes at the bone-implant interface region [12] ,
which is located within a few micrometers of the implant sur-
face. The arrangements of osteocytes and mineralized collagen fib-
rils in the bone-implant interface region are similar to those at
the bone-osteocyte interface [13] . In addition, osteocytes at the
bone-implant interface region might form lacunar-canalicular net-
works with neighboring osteocytes. These networks might extend
in peri-implant supporting bone adjacent to the interface by form-
ing interconnectivity [14] . These facts suggest that control of os-
teocyte dendrite development at the bone-implant interface re-
gion might develop the lacunar-canalicular network and lead to
enhanced quality of peri-implant bone tissue.
Surface topography affects the osseointegration capability of
implants [15] . Titanium implants with micro-roughened surfaces
accelerate and enhance osseointegration by promoting osteoblas-
tic cell differentiation [16] . One of the mechanisms underlying the
regulation of osteoblastic differentiation by surface topography is
the control of osteoblast morphology and formation of focal adhe-
sion plaques, resulting in promotion of cellular differentiation via
activation of mechanotransduction [17] . Nano-roughened surfaces
facilitate the formation of focal adhesion plaques in osteoblas-
tic cells more than the micro-roughened surfaces [18] . Osteocytes
in the bone-implant interface region can be closely interdigitated
with the topographic features of micro-roughened surfaces [ 12 , 19 ].
In contrast, osteocytes with poor dendritic processes have been ob-
served around implants with machined surfaces [19] . These obser-
vations suggest that implant surface properties may regulate the
morphology and function of osteocytes adherent to the implant
surface. However, implant surfaces intentionally designed to reg-
ulate osteocyte lacunar-canalicular networks in peri-implant bone
tissue have not yet been introduced [20] .
An alkali-etching treatment with immersion in a concentrated
hot sodium hydroxide solution creates superhydrophilic and rough-
ened titanium surfaces [21] . Nanotopographic patterns can be de-
signed by modifying the immersing protocol [21] . Alkali-etched
titanium nanosurfaces with numerous anisotropically-patterned
nanospikes can change the morphology of primary gingival and
dermal fibroblasts from spindle to circular form, with the cells ex-
hibiting multiple projections, numerous focal adhesions, and en-
hanced synthesis of fibroblastic extracellular matrix [ 21 , 22 ]. The
titanium nanosurface tunes macrophages into M1-like types char-
acterized by truly circular cell shapes with numerous cell projec-
tions and production of factors that inhibit osteoclastogenesis [23] .
Interestingly, turning of cell morphology and functions by the ti-
tanium nanosurface is attributed to nanotopographic cues, rather
than other physicochemical properties, such as superhydrophilicity
[23] .
Based on this background, we hypothesized that the titanium
nanosurface with anisotropically-patterned nanospikes enhances
the peri-implant bone quality by facilitating development of an
osteocyte lacunar-canalicular network, which strengthens the os-
seointegration of dental implants. The purpose of this study was
to investigate the effects of titanium nanosurfaces on the develop-
ment of an osteocyte lacunar-canalicular network and osseointe-
gration capability.
2. Material and methods
2.1. Titanium sample preparation
Commercially pure grade 2 titanium discs (13 or 20 mm di-
ameter, 1 mm thick), titanium screws (1.4 mm diameter, 4.0 mm
length), and titanium cylinders (1.0 mm diameter, 3.0 mm length)
were purchased from Nishimura Co., Ltd. (Fukui, Japan). The as-
prepared titanium samples with smooth-machined (SM) surfaces
were washed by ultrasonication in a series of ethanol and dis-
tilled water (DW) solutions after acetone cleaning. Titanium sam-
ples with acid-etched surfaces were prepared as representative os-
seointegrated micro-roughened (MR) surfaces by immersion of the
cleaned SM titanium samples in 67% (w/w) sulfuric acid solution
(FUJIMA Wako Pure Chemical Corporation, Osaka, Japan) at 120 °C
for 75 s. Three types of nanotopographical titanium surfaces were
prepared based on previously reported alkaline etching protocols
[ 21 , 22 ]. Briefly, the cleaned SM titanium samples were boiled in a
5, 10, or 15 M sodium hydrate solution for 24 h at 60 or 90 °C, re-
spectively. After boiling, the titanium samples were washed with
DW and dried overnight under dark and ambient conditions. The
titanium samples were then sintered in a furnace with a linearly
increasing temperature of 5 °C / min to 600 °C, with the final tem-
perature maintained for 1 h, followed by natural cooling after sin-
tering. Boiling with a 5, 10, or 15 M sodium hydrate solution pro-
vided different nano-roughened (NR) surfaces, namely NR_a, NR,
and NR_b titanium surfaces, respectively. All samples were auto-
claved immediately before use in vitro or in vivo .
2.2. Analyses of titanium surface properties
The surface topography of titanium surfaces was evaluated us-
ing scanning electron microscopy (SEM) with a model JSM-6390LA
microscope (JEOL Ltd., Tokyo, Japan) and by laser microscopy us-
ing a Talysur f PGI 1250A microscope (AMETEK Taylo r Hobson, Le-
icester, UK). The horizontal pattern and distribution of spikes on
the MR or NR surfaces were evaluated on their vertex extraction
images from the secondary electron SEM images ( Fig. S1A ) using
a WinRoof image analyzer (MITANI Corporation, Tokyo, Japan) as
previously described [23] . Briefly, based on the centroidal Voronoi
tesselation, the Voronoi diagrams of the vertex-extracted images
were prepared by dividing the area in a way that the vertex of each
spike became the centroid of the area ( Fig. S1A ). Vertex density
was calculated by measuring the number of vertices per square
micrometer. Anisotropy in the vertex distribution pattern was cal-
culated by dividing the mean of the tesselated areas on each sur-
face by the corresponding standard deviation. Elemental analysis of
titanium surfaces was performed using an EX-94300S4L1Q energy-
dispersive spectroscopy (EDX) system (JEOL Ltd.) incorporated in
the scanning electron microscope. The vertical roughness param-
eters of the arithmetical mean height (Sa) and maximum peak
height (Sp) were measured at a measurement length of 50 μm on
each titanium surface using the laser microscope after removing
waviness by approximating a cubic polynomial.
614
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
To evaluate the chemical properties of the titanium surface,
Fourier transform infrared (FTIR) spectra indicating functional
groups on titanium discs were analyzed using an IRT70 0 0 lin-
ear array imaging microscope (JASCO Corporation, Tokyo , Japan).
Micro-reflection spectra were recorded in the range of 40 0 0-20 0 0
cm
−1 at a spectral resolution of 4 cm
−1 with 500 accumulations
and a 50 μm
2 aperture. Background correction was performed
based on the surface spectrum of SM.
To evaluate the titanium surface energy, contact angles for 30
μL of DW droplets were measured on each titanium surface us-
ing a CA-X sessile drop machine (Kyowa Interface Science Co. Ltd.,
Saitama, Japan). The contact angles were measured using the θ/ 2
method, which is based on the calculation of twice the value of the
angle of the straight line connecting the left and right endpoints
of the droplet to the solid surface ( Fig. S1B ). Water contact an-
gles > 90 °and < 90 °were defined as hydrophobic and hydrophilic,
respectively, and a water contact angle < 10 °was defined as super-
hydrophilic [23] .
An ultrathin longitudinal section of NR titanium discs was pre-
pared using the ion-milling method for metal specimens as pre-
viously described [23] . Briefly, an NR titanium disc was bonded
with a dummy lining material using epoxy resin, shaped using a
cutting machine, thinned along the longitudinal sectional direction
using mechanical polishing, and ultra-thinned on a fixing mesh us-
ing a precision ion polishing system (PIPS) 691 ion-milling ma-
chine (Gatan, Pleasanton, CA, USA). Bright-field observation and
elemental analysis of the ultrathin longitudinal titanium sections
were performed by transmission electron microscopy (TEM) using
a model HF-20 0 0 microscope system (Hitachi High-Tech Corpora-
tion, Tokyo, Japan) equipped with a JEM-2100F EDX spectroscope
(JEOL Ltd.) at an acceleration voltage of 200 keV.
2.3. Osteocyte cell culture
MLO-Y4 cells were expanded on a culture-grade polystyrene
plate in alpha-MEM (FUJIMA Wako Pure Chemical Corporation)
containing 10% fetal bovine serum (Japan Bioserum, Hiroshima,
Japan), 100 U/mL penicillin, and 100 μg/mL streptomycin solution
(FUJIFILM Wako Pure Chemical Corporation) at 37 °C in a humidi-
fied atmosphere of 5% CO
2
. Titanium discs with SM, MR, NR, NR_a,
or NR_b surfaces were placed in 24- or 12-well polystyrene culture
plates. After cells reached 80% confluency, they were detached with
a solution containing 0.25% trypsin and 1 mM EDTA (FUJIMA Wako
Pure Chemical Corporation).
For the monolayer culture experiment, the cells were seeded
on polystyrene culture plates and titanium discs at a density of
1.5 ×10
4 cells/cm
2 in alpha-MEN with the same supplements as
above. The cells were cultured for 1, 3, or 7 d at 37 °C in humidified
5% CO
2
.
For the three-dimensional (3D) culture experiment, the cells
were mixed with type I collagen gels consisting of acid-extracted
native type I collagen (Cell matrix type I-A; Nitta Gelatin, Osaka,
Japan) containing a neutralization buffer with 200 mM HEPES, 50
mM NaOH, and 220 mM NaHCO
3
(Nitta Gelatin) and 10-fold con-
centrated Ham’s F12 culture media (Nitta Gelatin) at a cell den-
sity of 2.0 ×10
6 cells/mL. Two hundred microliters of the cell-
containing collagen gel was placed on polystyrene culture plates
and titanium discs and cultured in 10 0 0 μL of alpha-MEN with
the same supplements as above at 37 °C in humidified 5% CO
2
. The
culture media was replaced every 3 days.
2.4. Immunofluorescent staining
On days 1 and/or 7 of the monolayer and 3D cultures, the cells
were fixed in 10% neutral buffered formalin for 15 min, followed
by washing with phosphate-buffered saline (PBS) three times. In
the 3D culture, the collagen gel-containing cells were gently sepa-
rated from the disc, and the cells in the gel and on the disc were
independently subjected to the following staining procedure. The
cells were permeabilized in 0.1% Triton X-100 for 15 min and then
incubated in PBS containing 3% bovine serum albumin (BSA) and
0.1% Triton X-100 for 60 min to block nonspecific antibody re-
actions. Next, the cells were incubated in a 1/200 dilution of
anti-rabbit paxillin monoclonal antibody (ab32084; Abcam, Cam-
bridge, UK), 1/200 dilution of mouse anti-connexin 43 monoclonal
antibody (sc-271837; Santa Cruz Biotechnology, Dallas, TX, USA),
or 1/200 dilution of anti-mouse podoplanin monoclonal antibody
conjugated with Alexa Fluor 488 (eBio53-5381-82; Thermo Fisher
Scientific, Waltham, MA, USA) overnight at 4 °C. After washing with
PBS at least three times, the cells were incubated in an Alexa
Fluor 488-conjugated secondary antibody solution consisting of a
1/10 0 0 dilution of goat anti-rabbit immunoglobulin G (IgG) H&L
antibody (ab150081; Abcam) for paxillin or a 1/10 0 0 dilution of
goat anti-mouse IgG (H + L) antibody (A-11001; Thermo Fisher Sci-
entific) for connexin 43 at room temperature for 30 min. Next, the
cells were incubated in a 1/400 dilution of rhodamine phalloidin
solution (Thermo Fisher Scientific) for 90 min to stain F-actin. Af-
ter washing three times, the cells in the collagen gel or on the disc
were mounted on a glass-bottom dish (Matsunami Glass Ind., Ltd.,
Osaka, Japan) with a mounting agent containing 4’,6-diamidino-
2-phenylindole (DAPI) for staining nuclei of cells (Vector Labo-
ratories, Inc., Burlingame, CA, USA) and observed using a model
LSM 780 confocal laser microscope (Carl Zeiss, Jena, Germany). Cell
morphometry was performed on confocal laser microscopy images
using ImageJ software (National Institutes of Health, Bethesda, MD,
USA).
2.5. Reverse transcription-polymerase chain reaction (RT-PCR)
analysis
The total RNA was extracted at days 1, 3, and/or 7 from the
monolayer and 3D cultures by scraping the collagen or disc sur-
face in a TRIzol reagent (Ambion/Life Technologies, Carlsbad, CA,
USA) for RT-PCR analysis. The extracted total RNA was purified us-
ing an RNAeasy® Mini Kit (QIAGEN, Hilden, Germany) followed by
DNase I treatment (Thermo Fisher Scientific). Complementary DNA
(cDNA) was synthesized using a PrimeScript
TM II 1
st Strand cDNA
Synthesis Kit (TaKaRa Bio, Shiga, Japan). Messenger RNA (mRNA)
expression was determined using the StepOnePlus Real-Time PCR
system (Applied Biosystems, Thermo Fisher Scientific) and Thun-
derbird® SYBR® qPCR Mix (Toyobo, Osaka, Japan) for the SYBR
green-based PCR reaction. The target gene expression levels were
quantitatively analyzed using the comparative cycle time ( CT)
method. Glyceraldehyde 3-phosphate dehydrogenase ( Gapdh ) was
used as a housekeeping gene. The primers used are listed in Sup-
plementary Table 1.
2.6. Western blotting assay
On day 7, the cells from the 3D cultures were lysed by ultra-
sonic homogenization using ice-cold RIPA lysis buffer (Pierce, Rock-
ford, IL, USA) and then denatured at 95 °C for 5 min using Laemmli
sample buffer (Bio-Rad, Hercules, CA, USA). The total protein con-
centration of each sample was measured using a protein colori-
metric assay (Pierce 660 nm protein assay; Thermo Fisher Scien-
tific). Total protein samples were transferred onto polyvinylidene
fluoride (PVDF) membranes. The membranes were then blocked
with 5% BSA for 1 h at room temperature, followed by incuba-
tion with a 1/20 0 0 dilution of mouse anti-connexin 43 monoclonal
antibody (sc-271837; Santa Cruz Biotechnology) or a 1/10 0 0 dilu-
tion of mouse anti- β-actin monoclonal antibody (8H10D10; Cell
Signaling Technology, Beverly, MA, USA) at 4 °C overnight. Next,
615
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
the membrane was incubated with horseradish peroxidase (HRP)-
conjugated secondary antibody solution consisting of a 1/200
diluted anti-mouse-IgG κbinding protein (sc-516102; Santa Cruz
Biotechnology) at room temperature for 2 h. Signals were de-
tected with Immobilon Western Chemiluminescence HRP substrate
(Merck Millipore, Burlington, MA, USA) using an ImageQuant LAS-
500 image analyzer (GE Healthcare Japan, Tokyo, Japan).
2.7. Animal surgery
Eleven-week-old male Sprague-Dawley rats weighing 320–400
g were used for the in vivo experiment and implanted with two
types of titanium implants, i.e., a screw and cylinder, depending
on the kind of analysis. Screw-type implants were used for the re-
moval torque test and histological and histomorphometrical anal-
yses of nondecalcified resin-embedded sections with Villanueva-
Goldner staining. Cylinder-type implants were used for SEM anal-
ysis of nondecalcified resin-embedded sections to evaluate osteo-
cyte morphology in peri-implant tissue because this type of im-
plant enabled a clear distinction between the bone and implant in-
terface region and the supporting bone region in peri-implant bone
tissue.
After anaesthetizing via subcutaneous administration of 0.15
mg/kg medetomidine hydrochloride, 2 mg/kg midazolam, and 2.5
mg/kg butorphanol tartrate, the bilateral upper first molars were
gently extracted with a needle while avoiding root fracture. MR
or NR titanium screw- or cylinder-type implants were bilaterally
placed in an extraction socket using the following procedure. The
screw-type implant was placed into a bony hole made on the
septa interalveolaria with a #70 endodontic K file, leaving the
screw head exposed without suturing. The cylinder-type implant
was press-fitted into an extraction socket of the mesial root after
scraping with a #50 endodontic K file and was then submerged in
the gingiva with a 5-0 nylon monofilament suture. Titanium im-
plants with two surface types were randomly placed on each side.
Neither implant touched the opposite mandibular tooth. Biome-
chanical tests and histological assessments were performed after
4 weeks, which was sufficient for complete bone healing in the rat
upper jawbone [ 24 , 25 ].
Seven animals were used for the removal torque test. Besides
this, six and five animals were used for histological examination of
nondecalcified sections with Villanueva-Goldner staining and SEM
analysis to evaluate osteocyte morphology, respectively. The animal
experiments strictly followed the protocol approved by the Institu-
tional Laboratory Animal Care and Use Committee of Tohoku Uni-
versity (protocol No. 2019DnA-046-02).
2.8. Removal torque test
The maxillae containing screw-type implants were rinsed with
PBS. The tissue specimens were immediately embedded in au-
topolymerized acrylic resin (Ostron II; GC Corporation, Tokyo ,
Japan) with the implant head exposed. After fixing the polymer-
ized resin pedestal embedded in the maxillae, the removal torque
value of the implant was measured using a model DIS-RL05 torque
tester (Sugisaki Meter Co., LTD, Ibaraki, Japan). The precision flat-
blade screwdriver bit connected to the torque tester was posi-
tioned perpendicular to the long axis of the implant, with the
adaptation of the blade to the notch of the implant head. The re-
verse torque value loaded onto the implant was measured by con-
tinuously rotating the driver bit in the reverse direction until the
implant rotated. In the obtained reverse torque curve, the value
of the inflection point that led to a rapid reduction after a linear
increase was regarded as the removal torque value that destroyed
osseointegration. Its absolute values were defined as the osseointe-
gration strength. A reverse torque curve with no inflection resulted
in a rapid reduction after a linear increase and was defined as an
incorrect reverse torque curve. Samples showing an incorrect re-
verse torque curve were excluded from the analysis.
2.9. Processing of resin-embedded nondecalcified histological sections
and histological evaluations
Maxillary bones containing the implants were immersed in 70%
ethanol after fixation in 10% neutral buffered formalin. Next, the
samples were subjected to en bloc staining using Villanueva Os-
teochrome Bone Stain Solution (Polysciences, Inc., Warrington, PA,
USA) for 2 weeks, with daily degassing at –1 Pa for 30 min. Af-
ter washing and dehydrating with a series of ethanol, acetone,
and xylene solutions, the samples were embedded in polymethyl-
methacrylate (FUJIFILM Wako Pure Chemical Corporation) or epoxy
resin (Oken Epok 812, Okenshoji Co., Ltd., Tokyo , Japan) for histo-
logical and histomorphometric evaluation or SEM analysis, respec-
tively. The samples were sectioned at the center of the screw- and
cylinder-type implant in the longitudinal and cross-sectional direc-
tions, respectively, using an SP1600 saw microtome (Leica Camera
AG, Wetzlar, Germany). Next, each section was ground to a thick-
ness of 50 μm with a series of water-resistant sandpapers of up to
2400 grit.
The polymethylmethacrylate resin-embedded longitudinal sec-
tions of the screw-type implants and the surrounding peri-implant
bone tissue were further stained with Villanueva-Goldner stain to
color mineralized (green) or non-mineralized (red-purple) tissue
after optical microscopy. The bone-implant contact ratio for the
evaluation of osseointegration strength was measured by calculat-
ing the attachment area of green-colored bone tissue onto the im-
plant surface relative to the total surface area of the bony part of
the implant.
For SEM evaluation of the morphologies of osteocytes and their
lacunar-canalicular networks in peri-implant bone tissue, epoxy
resin-embedded cross sections of the cylinder-type implants and
the surrounding peri-implant bone tissue underwent resin-cast
etching [26] . The surface of each polished section was etched
with 37% phosphoric acid for 10 s, and then immersed in 5%
sodium hypochlorite for 5 min to demineralize and decompose
the bone matrix. After rinsing with DW, the sections were air-
dried overnight. The surfaces of the sections were sputter-coated
with gold/palladium alloy and observed using an XL30 SEM sys-
tem (Philips, Eindhoven, Netherlands) at an acceleration voltage of
10 keV. Observed cells in the bone tissue region mainly include
osteocytes and osteoblasts, which can be clearly distinguished by
their morphological features [27] . Osteocytes exhibit a round or el-
liptical cell shape with numerous elongated cell processes extend-
ing from their cell surface. In contrast, osteoblasts have a smooth
surface, a polygonal or rounded shape with a bulge in the center,
and a small number of short microvilli [27] . Osteocyte morpholo-
gies and dendrite formation were evaluated using Image J software.
For peri-implant bone tissue, the bone tissue region in direct con-
tact with the implant surface and the region at least 50 μm away
from the surface were defined as the bone-implant interface region
and the supporting bone region [ 16 , 28 ], respectively.
2.10. Statistical analysis
For surface analyses, three independent samples of each tita-
nium surface were subjected to a series of quantitative measure-
ments. For FTIR, EDX, and TEM analyses, multiple samples were
analyzed. Data of one representative sample are shown. Except for
cytomorphometry of MLO-Y4 cells cultured on each substrate, all
616
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
culture experiments were performed in at least three independent
cell batches on multiple days ( N = 3). Gene expression analyses
were performed three times, and representative data sets are pre-
sented after confirming consistency ( N = 3). Cytomorphometry of
cell shapes and paxillin expression was analyzed in single cells in
randomly selected confocal laser microscopic images in multiple
cultures on each substrate ( N = 10). Seven or six individual ani-
mals with two different types of implants placed bilaterally on the
upper jawbone were subjected to a removal torque test ( N = 7)
or histomorphometric evaluations with Villanueva-Goldner stain-
ing ( N = 6). In the histomorphometric evaluations of osteocytes
analyzed from SEM images obtained from five individual animals,
35 or 14 cells in the supporting bone region or bone-implant in-
terface region were analyzed for each implant type ( N = 35, or
14) . The sample size of all animal experiments was determined us-
ing G
∗Power 3.1.9.7 (Heinrich-Heine-Universität Düsseldorf, Düssel-
dorf, Germany). One-way analysis of variance (ANOVA) was used to
assess differences among multiple experimental groups. Two-way
ANOVA was used to assess the interactions between differences in
the substrate types and the culturing period or the loading time.
When appropriate, post-hoc Tukey’s honestly significant difference
(HSD) test or Bonferroni’s multiple comparison test was used. Stu-
dent’s t-test, Welch’s t-test, or Mann-Whitney U-test were used for
comparisons between the two groups. The statistical method was
applied based on the results of the Shapiro-Wilk and Levene’s tests,
which analyze the normality and homoscedasticity, respectively. P
< 0.05 was considered statistically significant. All statistical anal-
yses were performed using SPSS Statistics version 21 (IBM Japan,
Ltd., Tokyo, Japan).
3. Results
3.1. Topographic and physicochemical characteristics of titanium
nanosurfaces
SEM of SM titanium surfaces revealed a flat appearance to-
gether with shallow machine grooves. In contrast, MR titanium
surfaces had numerous sharp ridges and micro-pits ( Fig. 1 A , ar-
rowheads). NR titanium surfaces contained multiple nanospikes
and crevasses connecting holes with a sponge-like inner network
( Fig. 1 A , double arrowheads). The vertex distribution analyzed us-
ing the Voronoi diagram ( Fig. S1A ) demonstrated that the vertex
density of NR titanium surfaces was greater than that of MR ti-
tanium surfaces ( Fig. 1 B ) ( P < 0.05, Student’s t-test). Anisotropy
of the vertex distribution was not different between the surfaces
( Fig. 1 B ) ( P > 0.05, Student’s t-test). TEM demonstrated that NR
titanium surfaces featured a superficial porous layer and underly-
ing titanium substrate ( Fig. 1 C ). The superficial porous layer had
a sponge-like structure with numerous nanospikes ( Fig. 1 D , arrow-
heads), composed of amorphous titanium oxide ( Fig. 1 E-a ) contain-
ing sodium atoms ( Fig. 1 E-b ), in contrast to the complete titanium
metal crystal lattice detected on the titanium substrate ( Fig. 1 F-a
and F-b ).
MR titanium surfaces had the maximum vertical roughness pa-
rameters, such as Sa and Sp, at micron levels compared to sub-
micron values on the SM and NR titanium surfaces ( Fig. 1 G ) ( P <
0.05, Tukey’s HSD test). Sa and Sp values of the NR titanium sur-
faces remained two-times higher than the values of SM surfaces
( P < 0.05, Tukey’s HSD test). FTIR spectra featured peaks of the
hydroxyl group for the NR titanium surface, but not SM or MR
titanium surfaces ( Fig. 1 H ). NR titanium surfaces had water con-
tact angles < 10 °(superhydrophilic) in contrast with approximately
100 °(hydrophobic) on SM and MR titanium surfaces ( Fig. 1 I ) ( P <
0.05, Tukey’s HSD test).
3.2. Effects of titanium nanosurfaces on osteocyte morphology
MLO-Y4 osteocyte cells cultured for a day on polystyrene cul-
ture plates and the SM titanium surfaces displayed polygonal cell
shapes with lamellipodium-like structures and wide cellular pro-
jections slightly expressing paxillin, the adaptor protein of focal ad-
hesion ( Fig. 2 A , arrowheads). Osteocytes on MR titanium surfaces
differed in their small and triangular shapes, but were similar to
those on SM surfaces and polystrene in their wide cellular projec-
tions with little paxillin expression ( Fig. 2 A ). In contrast, osteocytes
on the NR titanium surface had a stellate cell shape with elongated
multiple cellular projections expressing intensive paxillin signals
on day 1 ( Fig. 2 A , double arrowheads). Cell morphometry demon-
strated that osteocytes on the NR titanium surface had a perimeter
> 25%, greater than the perimeter of osteocytes on the other sub-
strates ( Fig. 2 B ) ( P < 0.05, Tukey’s HSD test), despite similar areas
and Feret diameters to those on the SM surfaces. Paxillin expres-
sion per osteocyte was higher on NR titanium surfaces than on the
other surfaces ( Fig. 2 C ) ( P < 0.05, Tukey’s HSD test).
3.3. Effects of titanium nanosurfaces on osteocyte maturation
Gene expression of the podoplanin ( Pdpn ) dendrite marker in
cultured osteocytes started to increase on day 7 on all substrates
and was highest on the NR titanium surface ( Fig. 3 A ) ( P < 0.05,
Bonferroni’s multiple comparison). Expression of the Gja1 gene en-
coding gap junction connexin 43 in osteocytes on the NR titanium
surface was low on days 1 and 3, but increased at day 7 to a
level that was slightly higher than that on the other substrates
( Fig. 3 A ) ( P < 0.05, Bonferroni’s multiple comparison). Expression
was downregulated or changed only marginally over time.
Confocal laser microscopy observations of the osteocyte cul-
tured on the NR titanium surface on day 7 demonstrated a neuron-
like cell appearance with markedly elongated cellular processes
and strongly expressed podoplanin ( Fig. 3 B , double arrowheads),
in contrast to the expanded polygonal cell shapes on polystyrene
and SM titanium surfaces. Osteocytes on the MR titanium surface
were elongated and developed cellular projections to some extent,
but remained far from the neuron-like shape, with less podoplanin
expression detected ( Fig. 3 B , arrowheads). Osteocytes on the NR
titanium surface at day 7 extensively expressed connexin 43 in
the cell body and in overlapping regions of the cellular projections
( Fig. 3 C , double arrowheads). In contrast, cells on the other sub-
strates showed limited expression of connexin 43 ( Fig. 3 C , arrow-
heads). Podoplanin or connexin 43 expression per osteocyte was
more than three times greater on the NR titanium surface than on
the other substrates ( Fig. 3 B and C ) ( P < 0.05, Tukey’s HSD test).
Expression of osteocyte maturation gene markers, such as phos-
phate regulating gene with dentin matrix protein 1 ( Dmp1 ), matrix
extracellular phosphoglycoprotein ( Mepe ), and fibroblast growth
factor 23 ( Fgf23 ), and homologies to endopeptidases on the X-
chromosome ( Phex ) in the osteocyte culture on day 7 was higher
on the NR titanium surface than on the other substrates ( Fig. 3 D )
( P < 0.05, Tukey’s HSD test), except for Phex expression on the MR
surface.
3.4. Effects of titanium nanosurfaces on osteocyte morphology and
maturation during 3D culture
Osteocytes cultured in 3D collagen gel for 7 days attached to
the underlying polystyrene or titanium substrate ( Fig. 4 A ). Simi-
lar to the monolayer osteocyte culture at day 7, the 3D-cultured
osteocytes attached to NR titanium surfaces displayed neuron-like
cell shapes with elongated cellular projections that intensively ex-
pressed intensive connexin 43 at day 7 ( Fig. 4 A , double arrow-
heads). In contrast, cells on the other substrates remained polyg-
617
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 1. Topographic and physicochemical characteristics of titanium nanosurfaces .
(A) Representative secondary scanning electron microscopy images of titanium discs with smooth-machined (SM), micro-roughened (MR), and nano-roughened (NR) surfaces.
(B) Vertex density and anisotropy in vertex distribution patterns of MR and NR surfaces together with the corresponding typical analyzed images of vertex extraction and
Voronoi diagram, respectively. Bright-field images (C and D), selected area electron diffraction images (E-a and F-a), and energy-dispersive X- ray spectroscopy profile (E-b
and F-b) of ultrathin titanium longitudinal sections of NR titanium surfaces examined by transmission electron microscopy. Data are from each dashed circle spot showing
the
corresponding typical features on the superficial layer (E-a and E-b) and titanium base (F-a and F-b). (G) Vertical roughness parameters of arithmetical mean height
(Sa) and maximum peak height (Sp), (H) Fourier transform infrared (FTIR) spectra and (I) wa ter contact angles with the corresponding water droplet image of each type
of titanium surface. Black arrowheads on FTIR spectra (H) indicate peaks corresponding to hydroxy l groups (-OH). Data are presented as means ±standard deviation (SD)
( N = 3). Different letters or asterisks indicate statistically significant differences ( P < 0.05; Tukey’s honestly significant differences test or Student’s t-test); n.s. indicates no
significant difference. Note: Sharp ridges and numerous micro-pits are evident on the MR titanium surface (arrowheads in A). Multiple nanospikes are observed on the NR
titanium surface (double arrowheads in A and D).
618
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 2. Effects of titanium nanosurfaces on osteocyte morphology .
(A) Representative confocal laser microscopy images of F-actin, nuclei, and paxillin in MLO-Y4 cells cultured on polystyrene culture plates (Poly) or titanium discs with
smooth-machined (SM), micro-roughed (MR) and nano-roughened (NR) surfaces at day 1. Cytomorphometric parameters of area, perimeter, and Feret diameter (B) and
numbers of paxillin per osteocyte (C) were measured in each image measured using Image J software. Data are presented as means ±standard deviation (SD) ( N = 10 ).
Different letters indicate statistically significant differences ( P < 0.05; Tukey’s honestly significant differences test). Note: (A) cells on the NR titanium surface develop
multiple elongated cellular projections with highly expressed paxillin signals (double arrowheads). In contrast, fewer cellular projections with less expressed paxillin signals
are evident on the other surfaces (arrowheads).
onal with less connexin 43 expression. Osteocytes in the 3D col-
lagen gel culture on NR titanium surfaces formed well-developed
interconnectivity with extensive connexin 43 expression ( Fig. 4 B ,
double arrowheads), in contrast to cell aggregation with a small
amount of connexin 43 expression in the 3D culture on other
substrates ( Fig. 4 B , arrowheads). Western blotting analysis demon-
strated that NR titanium surfaces markedly increased the connexin
43 expression of osteocytes cultured in the 3D collagen gel on day
7 compared to that of osteocytes cultured on the other substrates
( Fig. 4 C ). The expression of Pdpn, Gja1, Gjc1 (encoding connexin
45), and Dmp1 in osteocyte 3D cultures on day 7 was higher on
NR titanium surfaces than that of cells on the other substrates, ex-
cept for Pdpn expression in the 3D culture on polystyrene ( Fig. 4 D )
( P < 0.05, Tukey’s HSD test).
3.5. Effects of titanium nanosurfaces on development of osteocytic
network for osseointegration
Screw and cylinder types of implants had surface properties
similar to those of the culture experimental titanium discs treated
with the corresponding surface modifications ( Fig. S2 ). MR or
NR titanium screw implants placed on the upper jawbones in
the first molar regions ( Fig. 5 A ) were evaluated for osseointegra-
tion strength 4 weeks postoperatively using a biomechanical test
to measure the reverse torque value at loss of osseointegration
( Fig. 5 B ). The osseointegration strength of NR titanium implants
was 1. 5 times higher than that of MR titanium implants ( Fig. 5 C )
( P < 0.05, Student’s t-test). In contrast, histology of nondecalcified
resin-embedded samples demonstrated that both MR and NR tita-
nium implants allowed direct bone deposition on the implant sur-
face to a comparable extent ( Fig. 5 D , arrows). The bone-implant
contact ratio was approximately 50% for both the MR and NR tita-
nium implants ( Fig. 5 D , histogram) ( P > 0.05, Student’s t-test).
SEM observation on resin-cast etched histological sections 4
weeks postoperatively demonstrated that the osteocytes showed
elliptic shapes with a little dendrite formation in the peri-implant
supporting bone region around the MR titanium wire implants
( Fig. 5 E, arrowheads ). In contrast, the well-developed complex
networks were formed by polygonal or circle osteocytes with
numerous dendrites in the peri-implant supporting bone region
around the NR titanium implants ( Fig. 5 E, double arrowheads ).
Cytomorphometry on SEM images demonstrated the lower aspect
ratio ( Fig. 5 F ) ( P < 0.05, Mann-Whitney U-test) and the higher cir-
cularity of cell body ( Fig. 5 F ) ( P < 0.05, Mann-Whitney U-test) to-
gether with the greater number of dendrites ( Fig. 5 G ) ( P < 0.05,
Mann-Whitney U-test) in peri-implant supporting bone around the
NR titanium implants than those around the MR titanium im-
plants. Osteocytes at bone-implant interface region attached on the
619
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 3. Effects of titanium nanosurfaces on osteocyte maturation .
(A) Expressions of genes encoding a dendrite marker podoplanin ( Pdpn ) and a gap junction marker connexin 43 ( Gja1 ) relative to glyceraldehyde 3-phosphate dehydrogenase
( Gapdh ) analyzed by reverse transcription-polymerase chain reaction (RT-PCR) in MLO-Y4 cells cultured on a polystyrene culture plate (Poly) or titanium discs with smooth-
machined (SM), micro-roughed (MR), or nano-roughened (NR) surfaces at days 1, 3, and 7. Representative confocal laser microscopy images of F-actin, nuclei, and podoplanin
(B) or connexin 43 (C), and expression of each target molecule per osteocyte in MLO-Y4
cells on each culture substrate at day 7. (D) Gene expression of osteocytic maturation
markers dentin matrix protein 1 ( Dmp1 ), matrix extracellular phosphoglycoprotein ( Mepe ), fibroblast growth factor 23 ( Fgf23 ), and phosphate regulating endopeptidase
homolog x-linked ( Phex ) relative to Gapdh were analyzed by reverse transcription-polymerase chain re action (RT-PCR) in MLO-Y4 cells on each culture substrate at day
7. Data are presented as means ±standard deviation (SD) ( N = 3). Dashed circles in (A) indicate no significant differences ( P < 0.05; Bonferroni’s multiple comparison).
Different letters indicate statistically significant differences ( P < 0.05; Tukey’s honestly significant differences test).
620
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 4. Effects of titanium nanosurfaces on osteocyte morphology and maturation in 3D culture .
Representative confocal laser microscopy images of F-actin, nuclei, and connexin 43 on MLO-Y4 cells adherent to a polystyrene culture plate (Poly) or titanium discs with
smooth-machined (SM), micro-roughed (MR), or nano-roughened (NR) surfaces on day 7 (A) and in cells in a three-dimensional (3D) collagen gel over the corresponding
substrate (B). (C) Wes tern blot of expression of connexin 43 in MLO-Y4 cells within a 3D collagen gel over each culture substrate on day 7. (D) Gene expression of podoplanin
( Pdpn ), connexin 43 ( Gja1
), connexin 45 ( Gjc1 ), and dentin matrix protein 1 ( Dmp1 ) relative to glyceraldehyde 3-phosphate dehydrogenase ( Gapdh ) analyzed by reverse
transcription-polymerase chain reaction (RT-PCR) in MLO-Y4 cells in a 3D collagen gel over each culture substrate on day 7. Data are presented as means ±standard
deviation
(SD) ( N = 3). Different letters indicate statistically significant differences ( P < 0.05; Tukey’s honestly significant difference test). Note: A well-developed 3D osteocyte network
with strong connexin 43 signals in the collagen gel is evident over the NR titanium surface (double arrowheads in B). In contrast, connexin 43 signal was less in the collagen
gel over the other substrates (arrowheads in B).
621
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 5. Effects of titanium nanosurfaces on development of osteocytic network for osseointegration .
(A) Scanning electron microscopy (SEM) images for titanium mini-implants with micro-roughened (MR) or nano-roughened (NR) surface, and a photograph showing implant
placements on the right or left upper first molar region in an 11-week-old rat. (B) Reverse torque curve around the inflection point (black triangle) indicating osseointegra-
tion strength. (C) Osseointegration strength evaluated by rever se torque test for the MR and NR titanium implants 4 weeks postoperatively. (D) Representative images of
nondecalcified resin-embedded histological sections with Villanueva-Goldner staining and the bone-implant contact ratio on peri-implant bone tissue
around the MR or NR
titanium implants 4 weeks postoperatively. (E) Representative SEM images of resin-cast etched sections of the supporting bone region or the bone-implant interface region
in peri-implant bone tissue around the MR or NR titanium implants 4 weeks postoperatively, together with results of the aspect ratio and circularity of osteocytes (F) and
the number of dendrites per osteocyte on the supporting bone region (G) or the bone-implant interface region (H) in peri-implant bone tissue. Arrowheads in (D) indicate
spots of mineralized bone deposition on the implant surface (asterisk). Data are presented as means ±standard deviation (SD) ( N = 7 in [C], N = 6 in [D], N = 35 in [F
and G], N = 14 in [H]). Asterisks indicate statistically significant differences ( P < 0.05; Student’s t-test, We lch’s t-test or Mann-Whitney U-test); n.s., no significant difference.
Symbols: † , implant or implant surface; §, newly
formed mineralized bone tissue around the MR titanium implants; ¶, newly forme d mineralized bone tissue around the NR
titanium implants. Note: (E) dense osteocyte networks in the supporting bone re gion (double arrowheads) and osteocytes adherent to the implant surface with numerous
dendrites (triangles) in the bone-implant interface region in peri-implant bone tissue around the NR surface are evident.
622
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
MR titanium implant surface with a limited number of dendrites
( Fig. 5 E, arrows ). In contrast, the NR titanium implants allowed
osteocytes to attach the implant surface with a lot of dendrites
( Fig. 5 E, triangles ), which were much higher in number per os-
teocyte than those on the MR titanium implant implants ( Fig. 5 H )
( P < 0.05, Welch’s t-test).
3.6. Effects of nanotopography on promotion of osteocyte network
formation by titanium nanosurface
To determine which surface properties of the NR titanium
surface affect osteocyte maturation, the NR_a and _b titanium
nanosurfaces were created by altering the surface modification
protocol for the NR titanium surface. The NR_a and _b titanium
nanosurfaces were similar in elemental ( Fig. 6 A , EDX profile) and
chemical ( Fig. 6 D ) compositions, and surface wettability ( Fig. 6 E ) to
the NR titanium surfaces, but differed in surface nanotopography
( Fig. 6 A-C ). The NR_a titanium surface contained numerous shorter
nanospikes that were distributed more densely but less anisotrop-
ically than those on the NR surfaces ( Fig. 6 A-C and S1A ). In con-
trast, the NR_b titanium surface retained crevasses but lacked most
of the nanospikes ( Fig. 6 A and B ), and were flatter than the NR sur-
face ( Fig. 6 C ).
The expression levels of Gjc1 and Dmp1 in osteocytes cultured
on NR titanium surfaces at day 7 were consistently higher than
those of osteocytes on the polystyrene and SM titanium surfaces
( Fig. 6 F ) ( P < 0.05, Tukey’s HSD test). Furthermore, osteocytes on
the NR_a and NR _b titanium surfaces displayed the same levels of
Pdpn and Dmp1 gene expression as those of osteocytes on the NR
titanium surfaces ( P > 0.05, Tukey’s HSD test). However, NR_a and
NR _b titanium surfaces did not display upregulated Gjc1 expres-
sion compared to the level on the SM surface ( P > 0.05, Tukey’s
HSD test).
4. Discussion
In this study, MLO-Y4 cells cultured directly on NR titanium sur-
faces showed satellite shaped with strong paxillin expression on
day 1, whereas the cells on the SM or MR titanium surfaces had
polygonal shapes with short cell processes and less paxillin ex-
pression ( Fig. 2 A and B ). Elongated MLO-Y4 cells typically present
osteocytic phenotypes characterized by a neuron-like stellate mor-
phology with long dendrites and high expression of osteocyte mat-
uration markers [29] . Paxillin binds intracellularly to connexin 43,
and this interaction is linearly correlated to osteocyte gap junc-
tion formation [30] . The expression of the Pdpn dendrite elonga-
tion marker in MLO-Y4 cells on the NR titanium surfaces was as
low as that of cells on the other substrates at days 1 and 3, but
increased to the highest level on day 7 at both the gene and pro-
tein levels ( Fig. 3 A and B ). In addition, Dmp1, Mepe, Fgf23, and Phex
maturation gene markers were upregulated on NR titanium sur-
faces more than the upregulation on the other substrates on day 7
( Fig. 3 D ). Furthermore, increased connexin 43 expression by osteo-
cytes attached to NR titanium surfaces was evident compared to
the limited expression in cells on the other substrates ( Fig. 3 A and
C ). Connexin 43 is essential in the osteocyte network. The protein
forms gap junction channels for cell-to-cell connections [31] . These
observations indicate that NR titanium surfaces promote matura-
tion of osteocytes by inducing neuron-like elongated dendrite for-
mation through the activation of paxillin expression and devel-
opment of cell-to-cell connections together with the gap junction
channels.
Interestingly, the formation of an osteocyte network with con-
nexin 43 expression was observed even in the 3D collagen gel cul-
ture over NR titanium surfaces ( Fig. 4 B and C ). In contrast, cells
cultured in the collagen gel over the polystyrene, SM, or MR ti-
tanium surfaces merely aggregated and showed few connections
( Fig. 4 B and C ). Part of the cell population within the collagen gel
attached to the underlying substrate. The adherent cells on NR ti-
tanium surfaces also developed cell-to-cell connections with strong
connexin 43 signals, in contrast to weak signals in the cells adher-
ing to the other substrates ( Fig. 4 A ). To gether with gap junction-
encoding genes Gja1 and Gjc1 , the expression of Pdpn and Dmp1
in osteocytes within the 3D collagen gel was upregulated on NR
titanium surfaces compared to the expression in cells on other ti-
tanium surfaces. These observations indicate that the NR titanium
surface promotes osteocyte maturation and the formation of cellu-
lar networks, even in osteocytes adjacent to the adherent cells on
the surface.
Connexin 43 serves as the core of the osteocyte network by
forming gap junction channels to exchange substances or mechan-
ical signals between cells participating in the network, as well as
hemichannels connecting cells with the extracellular microenviron-
ment [ 32 , 33 ]. In peri-implant tissue, osteocytes form a lacunar-
canalicular network extending from the bone-implant interface re-
gion to the surrounding supporting bone by connecting adjacent
cells with each other [14] . If osseointegration is established by di-
rect bone deposition from the implant surface to the surrounding
bony wall ( i.e ., contact osteogenesis) [34] , the osteocyte network
might begin to develop from the bone-implant interface region.
Alkali-etched titanium surfaces are prone to inducing contact os-
teogenesis [35] . Interestingly, NR titanium implants in the upper
jawbone promoted the attachment of osteocytes to the implant
surface via dozens of dendrites ( Fig. 5 E and H ). The number of
dendrites attached to NR titanium surfaces was double or more
than the number on MR titanium implants ( Fig. 5 E and H ), and
markedly greater than those reported in previous studies on vari-
ous types of implant surfaces, such as machined titanium surfaces
[11] , sandblasted titanium surfaces [ 12 , 36 ], laser-abraded titanium
surfaces [ 13 , 37 ], 3D-printed microporous titanium alloy surfaces
[14] and zirconium-doped cobalt chrome alloy surfaces [38] . The
expression of focal adhesion proteins induces the opening of gap
junctions to promote cell-to-cell communication [39] . In addition,
we observed that osteocytes in the peri-implant supporting bone
around NR titanium implants were characterized by a polygonal or
circular cell body with numerous elongated dendrites ( Fig. 5 E-G ),
which is the typical active form of osteocytes under mechanical
loading [11] . In contrast, osteocytes around MR titanium implants
tended to be in a static form with an elliptical shape and less den-
drite formation ( Fig. 5 E-G ). These findings suggest that the NR tita-
nium implants promoted the maturation of osteocytes adherent on
the surface into the mature and active form by activating the for-
mation of focal adhesions. More importantly, like a domino effect,
the influence of this 2D osteocyte regulation by the NR titanium
surface might spread to the adjacent osteocytes, facilitating the
development of an active osteocyte lacunar-canalicular network in
peri-implant bone tissue through the gap junctions ( Fig. 7 ).
Bone minerals are distributed along the osteocyte lacunar-
canalicular network [40] , and the architecture correlates with bone
properties as a material [41] . Although the MR surfaces with
spikes that exceed 10 μm ( Fig. 1 G ) inherently enhance osseoin-
tegration strength by the under-cut effects on the bony wall [42] ,
NR titanium implants displayed nearly twice the osseointegration
strength as that of MR titanium implants ( Fig. 5 C ). However, NR
titanium implants did not increase the bone-implant contact ra-
tio beyond approximately 50% ( Fig. 5 D ), which is within the val-
ues for other conventional implant surfaces, including MR implants
[43–45] . Bone quality-related factors including bone apatite orien-
tation is closely related to the architecture of the osteocyte net-
work [4] and correlated with osseointegration strength [46] . These
observations suggest that the NR titanium surface might increase
623
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 6. Effects of nanotopography on promotion of osteocyte network formation by titanium nanosurface .
(A) Representative secondary scanning electron microscopy (SEM) images and energy-dispersive X-r ay spectroscopy profile on titanium discs with the nano-roughened (NR)
surface or NR_a or NR_b surface subtypes. (B) Vertex density and anisotropy in ve rtex distribution patterns of each titanium nanosurface together with the corresponding
typical analyzed images of vertex extraction and Voronoi diagram, respectively. (C) Vertical roughness parameters of arithmetical mean height (Sa) and maximum peak height
(Sp), (D) Fo urier transform infrared spectra and (E) water contact angles with the corresponding water droplet image of each
titanium nanosurface. (F) Gene expressions of
connexin 45 ( Gjc1 ), podoplanin ( Pdpn ), and dentin matrix protein 1 ( Dmp1 ) relative to glyceraldehyde 3-phosphate dehydrogenase ( Gapdh ) analyzed by reverse transcription-
polymerase chain reaction in MLO-Y4 cells cultured on a polystyrene culture plate (Poly) or titanium discs with the NR, NR_a, or NR_b surface at day 7. Black arrowheads
on FTIR spectra (D) indicate spectra peak corresponding to hyd roxyl groups (-OH). Data are presented as mean ±standard deviation (SD) ( N = 3). Different letters indicate
statistically significant differences ( P < 0.05; Tu key’s honestly significant differences test).
624
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
Fig. 7. Titanium nanotopography-mediated promotion of osteocytic maturation and development of the osteocytic network in peri-implant bone tissue .
A scheme showing the possible mechanism underlying the promotion of osteocyte maturation and development of osteocyte network in peri-implant bone tissue around a
titanium implant with nanospikes. The titanium nanosurface promotes the maturation of osteocytes adherent on the surface into active forms. Subsequently, the influence
of this two-dimensional osteocyte regulation by the titanium nanosurface eventually spreads to the adjacent osteocytes for the development of an active osteocyte lacunar-
canalicular network in peri-implant bone tissue through the gap junctions. In contrast, titanium implants
with ge neral surfaces do not support the progressive development
of osteocyte networks because of reduced osteocyte attachment and maturation. The developed osteocyte network around titanium implants with nanosurfaces might
contribute to the enhancement of peri-implant bone quality, which is characterized by the promotion of bone formation under mechanical stress transmitted from the
implant body.
peri-implant bone strength by enhancing bone quality-related fac-
tors, but not bone mass.
We also explored the surface properties involved in the pro-
motion of osteocyte maturation. Nanotopography of NR titanium
surfaces created with alkali-etching typically revealed multiple
nanospikes with higher anisotropy of vertex distribution patterns
lined by sponge-like inner networks ( Fig. 1 A-D ) composed of amor-
phous sodium titanate ( Fig. 1 E ). These surfaces were relatively flat
on the micron scale to a level comparable with that of SM surfaces
( Fig. 1 G ). In contrast, MR titanium surfaces displayed micron-scale
irregularities without nanospikes ( Fig. 1 A and B ) and much higher
vertical roughness values than SM or NR surfaces ( Fig. 1 G ). NR ti-
tanium surfaces were superhydrophilic ( Fig. 1 I ) with a hydroxyl
group ( Fig. 1 H ), in contrast to the hydrophobic nature and ab-
sence of hydroxyl groups on SM and MR titanium surfaces ( Fig. 1 H
and I ). Alkali-etching can confer topographical, crystallographic,
and physicochemical characteristics to titanium, regardless of the
shape of the material [ 21 , 23 ]. Superhydrophilicity, which is related
to the presence of hydroxyl groups on biomaterials, promotes pro-
tein adsorption and cellular attachment [47–50] . The nanotopog-
raphy of other alkali-etched titanium nanosurfaces was character-
ized by low but dense nanospikes with isotropic vertex distribu-
tion patterns (NR_a) or residual crevasses with loss of nanospikes
(NR_b) ( Fig. 6 A-C ). NR_a and NR_b alkali-etched titanium nanosur-
faces had almost the same physicochemical properties ( Fig. 6 D and
E ) but different surface topographies ( Fig. 6 A and B ). NR titanium
surfaces featured upregulated expression of gap junction-related
genes in osteocytes compared with the expression in osteocytes
on other alkali-etched titanium nanosurfaces ( Fig. 6 F ). These ob-
servations suggest that nanotopography of the NR titanium surface
is essential for the promotion of osteocyte network formation.
In contrast, the effects of other physicochemical or crystallo-
graphic properties on osteocyte maturation are mostly adjunctive.
Focal adhesions are formed on the vertices of a roughened sur-
face [51] via the concentrated adsorption of adhesion proteins [52] .
However, in the present study, the NR titanium surface showed
lower vertex density of nanospikes but higher anisotropy of ver-
tex distribution patterns and vertical roughness than the NR_a ti-
tanium surface ( Fig. 1 B and C ). Osteoblast differentiation of human
mesenchymal stem cells is promoted on substrates with dense
and anisotropically distributed nanopatterns. This effect disappears
when the density of the nanopatterns is reduced [53] . Nonlinear
anisotropically distributed nanopatterns promote neuronal differ-
entiation in neuroblastic cell lines co-cultured with nerve growth
factors [ 54 , 55 ]. Interestingly, the osteocyte lacunar-canalicular net-
work is similar in complexity to the neural network in the brain
[56] . Osteocyte dendrite formation is controlled by the same tar-
get gene as that for neuron dendrite formation [57] . Nanospikes
with dense and anisotropic distribution patterns on the NR tita-
nium surface may contribute to the promotion of osteocytic matu-
ration. Determination of surface properties that promote osteocytic
maturation would be of great interest for future research.
This study describes novel findings in regard to a titanium nan-
otechnology that enhances osseointegration capabilities by regulat-
625
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
ing the development of osteocytic 3D networks. Nanotechnology-
mediated regulation of osteocytes is expected to inform develop-
ment of pioneering technology to control bone quality and will
contribute to advancement in the fields of biomaterial science and
bone biology.
5. Conclusion
Titanium nanosurfaces with anisotropically patterned and
dense nanospikes promote the development of osteocyte lacunar-
canalicular networks by nanotopographical cues. These nanosur-
faces enhance the osseointegration strength of titanium implants.
Declaration of Competing Interest
All authors have no conflict of interest in this research.
Those authors have no conflict of interest.
CRediT authorship contribution statement
Xindie He: Investigation, Methodology, Formal analysis, Investi-
gation, Visualization, Funding acquisition, Writing – original draft.
Masahiro Yamada: Conceptualization, Investigation, Methodology,
Formal analysis, Funding acquisition, Writing –review & editing.
Jun Watanabe: Investigation, Methodology, Formal analysis, Writ-
ing –review & editing. Watcharaphol Tiskratok: Methodology,
Writing –review & editing. Minoru Ishibashi: Funding acquisition,
Writing –review & editing. Hideki Kitaura: Methodology, Writing
–review & editing. Itaru Mizoguchi: Methodology, Writing –re-
view & editing. Hiroshi Egusa: Conceptualization, Supervision, For-
mal analysis, Funding acquisition, Writing –review & editing.
Acknowledgments
We thank the Instrumental Analysis Group Graduate School of
Engineering Tohoku University for technical assistance for TEM and
FTIR analyses. This work was supported by Grant-in-Aids for Scien-
tific Research (B: 20H03874, M.Y. and H.E.; C: 20K10067, M. I., M.Y.
and H. E.) from the Japan Society for the Promotion of Science and
Foundation Nakao for Worldwide Oral Health (2021-2022: H.E.).
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi: 10.1016/j.actbio.2022.08.023 .
References
[1] Y. Uda, E. Azab, N. Sun, C. Shi, P.D . Pajevic, Osteocyte mechanobiology, Curr.
Osteoporos Rep. 15 (4) (2017) 318–325
.
[2] H. Chen, T. Senda, K.Y. Kubo , The osteocyte plays multiple roles in bone re-
modeling and mineral homeostasis, Med. Mol. Morphol. 48 (2) (2015) 61–68
.
[3] L.F. Bonewald, Summary–Osteocytes and mechanotransduction, J. Muscu-
loskelet. Neuronal Interact. 5 (4) (20 05) 333–334
.
[4] T. Ishimoto, K. Kawahara, A. Matsugaki, H. Kamioka, T. Nakano, Quantitative
Evaluation of Osteocyte Morphology and Bone Anisotropic Extracellular Matrix
in Rat Femur, Calcif. Tissue Int. 10 9 (4) (2021) 434–4 4 4
.
[5] A . Sekita, A . Matsugaki, T. Ishimoto, T. Nakano, Synchronous disruption of
anisotropic arrangement of the osteocyte network and collagen/apatite in
melanoma bone metastasis, J. Struct. Biol. 197 (3) (2017) 260–270
.
[6] T. Ishimoto, T. Nak ano, Y. Umakoshi, M. Yamamoto, Y. Ta bata , Degree of biolog-
ical apatite c-axis orientation rather than bone mineral density controls me-
chanical function in bone regenerated using recombinant bone morphogenetic
protein-2, J. Bone Miner. Res . 28 (5) (2013) 117 0– 1179
.
[7] R.G. Bacabac, D. Mizuno, C.F. Schmidt, F.C. MacKintosh, J.J. Van Loon,
J. Klein-Nulend, T.H. Smit, Round versus flat: bone cell morphology, elasticity,
and mechanosensing, J. Biomech. 41 (7) (2008) 1590–1598
.
[8] S. Burra, D.P. Nicolella, W.L. Francis, C.J. Freitas, N.J. Mueschke, K. Poole,
J.X. Jiang, Dendritic processes of osteocytes are mechanotransducers that in-
duce the opening of hemichannels, Proc. Natl. Acad. Sci. U. S. A. 107 (31) (2010)
13648–13653
.
[9] A.F. van Tol , V. Schemenz, W. Wagermaier, A. Roschger, H. Razi, I. Vitienes,
P. Fratzl, B.M. Willie, R. Weinka mer, The mechanoresponse of bone is closely
related to the osteocyte lacunocanalicular network architecture, Proc. Natl.
Acad. Sci. U. S. A. 117 (51) (2020) 32251–32259
.
[10] C.A. Schurman, S.W. Verbruggen, T. Alliston, Disrupted osteocyte connectivity
and pericellular fluid flow in bone with aging and defective TGF-beta signaling,
Proc. Nat l. Acad. Sci. U. S. A. 11 8 (25) (2021)
.
[11] M. Sasaki, S. Kuroshima, Y. Ao ki, N. Inaba, T. Sawase, Ultrastructural alterations
of osteocyte morphology via loaded implants in rabbit tibiae, J. Biomech. 48
(15) (2015) 4130–4141
.
[12] Z. Du, S. Ivanovski, S.M. Hamlet, J.Q. Feng, Y. Xiao, The Ultrastructural Rela-
tionship Between Osteocytes and Dental Implants Following Osseointegration,
Clin. Implant Dent. Rela t. Res. 18 (2) (2016) 270–280
.
[13] F.A. Shah, X. Wa ng, P. Thomsen, K. Grandfield, A. Palmquist, High-resolution vi-
sualization of the osteocyte lacuno-canalicular network juxtaposed to the sur-
face of nanotextured titanium implants in human, ACS Biomater. Sci. Eng. 1 (5)
(2015) 305–313
.
[14] F.A . Shah, A . Snis, A . Matic, P. Thomsen, A. Palmquist, 3D printed Ti6Al4V
implant surface promotes bone maturation and retains a higher density of
less aged osteocytes at the bone-implant interface, Acta Biomater. 30 (2016)
357–367
.
[15] R.A. Gittens, R. Olivares-Navarrete, Z. Schwartz, B.D. Boyan, Implant osseointe-
gration and the role of microroughness and nanostructures: lessons for spine
implants, Act a Biomater. 10 (8) (2014) 3363–3371 .
[16] T. Ogawa, I. Nishimura, Different bone integration profiles of turned and
acid-etched implants associated with modulated expression of extracellular
matrix ge nes, Int. J. Oral Maxillofac. Implants 18 (2) (2003) 200–210
.
[17] J.A. Luckanagul, K. Metavarayuth, S. Feng, P. Maneesaay, A.Y. Clark, X. Yang,
A.J. Garcia, Q. Wang, Tobacco mosaic virus functionalized alginate hy drogel
scaffolds for bone regenerati on in rats with cranial defect, ACS Biomater. Sci.
Eng. 2 (4) (2016) 606–615
.
[18] R.A. Gittens, R. Olivares-Navarrete, A. Cheng, D.M. Anderson, T. McLach-
lan, I. Stephan, J. Geis-Gerstorfer, K.H. Sandhage, A.G. Fedorov, F. Rupp,
B.D. Boyan, R. Tannenbaum, Z. Schwartz, The ro les of titanium surface mi-
cro/nanotopography and wettability on the differential response of human os-
teoblast lineage cells, Acta Biomater. 9 (4) (2013) 6268–6277 .
[19] F.A. Shah, P. Stenl und, A. Martinelli, P. Thomsen, A. Palmquist, Direct communi-
cation between osteocytes and acid-etched titanium implants with a sub-mi-
cron topography, J. Mater. Sci. Mater. Med. 27 (11) (2016) 167
.
[20] R.S. Liddell, E. Ajami, Y. Li, E. Bajenova, Y. Yan g, J.E. Davies, The influence of
implant design on the kinetics of osseointegration and bone anchorage home-
ostasis, Act a Biomater. 121 (2021) 514–526
.
[21] E. Kato, K. Sakurai, M. Yamada, Periodontal-like gingival connective tissue at-
tachment on titanium surface with nano-ordered spikes and pores created by
alkali-heat treatment, Dent. Mater. 31 (5) (2015) e116–e130
.
[22] M. Yam ada , E. Kato, A. Yamamoto, K. Sakurai, A titanium surface with
nano-ordered spikes and pores enhances human dermal fibroblastic extracel-
lular matrix production and integration of collagen fibers, Biomed. Mater. 11
(1) (2016) 015010
.
[23] N. Kartikasari, M. Yam ada , J. Wata nabe , W. Tiskratok, X. He, Y. Kamano,
H. Egusa, Titanium surface with nanospikes tunes macrophage polarization
to produce inhibitory factors for osteoclastogenesis through nanotopographic
cues, Ac ta Biomater. 137 (2022) 316–330
.
[24] M. Haga, N. Fujii, K. Nozawa-Inoue, S. Nomura, K. Oda, K. Uoshima, T. Maeda,
Detailed process of bone remodeling after achievement of osseointegration in
a ra t implantation model, Anat Rec (Hoboken) 292 (1) (2009) 38–47
.
[25] S. Kuroshima, B.L. Kovacic, K.M. Kozlo ff, L.K. McCauley, J. Yamashita, Intra-oral
PTH administration promotes tooth extraction socket healing, J. Dent. Res. 92
(6) (2013) 553–559
.
[26] A .A . Gorustovich, Imaging resin-cast osteocyte lacuno-canalicular system at
bone-bioactive glass interface by scanning electron microscopy, Microsc. Mi-
croanal. 16 (2) (2010) 132–136
.
[27] K. Abe, H. Hashizume, T. Ushiki, An EDTA-KOH method to expose bone cells
for scanning electron microscopy, J. Electron. Microsc. (Tokyo) 41 (2) (1992)
113–115
.
[28] F. Butz, T. Ogawa, T. L. Chang, I. Nishimura, Three-dimensional bone-implant
integration profiling using micro-computed tomography, Int. J. Oral Maxillofac.
Implants 21 (5) (2006) 687–695
.
[29] I. Kalajzic, B.G. Matthews, E. Torreggiani, M.A. Harris, P. Divieti Pajevic, S.E. Har-
ris, In vitro and in vivo approaches to study osteocyte biology, Bone 54 (2)
(2013) 296–306
.
[30] D. Zhang, C. Zhou, Q. Wang, L. Cai, W. Du, X. Li, X. Zhou, J. Xie, Extracellu-
lar Matrix Elasticity Regulates Osteocyte Gap Junction Elongation: Involvement
of Paxillin in Intracellular Signal Transduction, Cell. Physiol. Biochem. 51 (3)
(2018) 1013–1026
.
[31] L.I. Plotkin, T. Bellido, Beyond gap junctions: Connexin43 and bone cell signal-
ing, Bone 52 (1) (2013) 157–166
.
[32] S. Burra, D.P. Nicolella, W.L. Francis, C.J. Freitas, N.J. Mueschke, K. Poole,
J.X. Jiang, Dendritic processes of osteocytes are mechanotransducers that in-
duce the opening of hemichannels, Proc. Nat . Acad. Sci. USA 107 (31) (2010)
13648–13653
.
[33] M.C. Moorer, J.P. Stains, Connexin43 and the intercellular signaling network
regulating skeletal rem odeling, Curr Osteoporosis Reports 15 (1) (2017) 24–31
.
[34] T. Berglundh, I. Abrahamsson, N.P. Lang, J. Lindhe, De novo alveolar bone for-
mation adjacent to endosseous implants, Clin. Oral. Implants Res. 14 (3) (2003)
251–262
.
626
X. He, M. Ya mada , J. Watan abe et al. Acta Biomaterialia 151 (2022) 613–627
[35] T. Ueno, N. Tsukimura, M. Yam ada , T. Ogawa, Enhanced bone-integration ca-
pability of alkali- and heat-treated nanopolymorphic titanium in micro-to–
nanoscale hierarchy, Biomaterials 32 (30) (2011) 7297–7308
.
[36] Z. Du, Y. Xiao, S. Hashimi, S.M. Hamlet, S. Ivanovski, The effects of im-
plant topography on osseointegration under estrogen deficiency induced os-
teoporotic conditions: Histomorphometric, transcriptional and ultrastructural
analysis, Acta Biomater. 42 (2016) 351–363
.
[37] F.A. Shah, M.L. Johansson, O. Omar, H. Simonsson, A. Palmquist, P. Thomsen,
Laser-modified surface enhances osseointegration and biomechanical anchor-
age of commercially pure titanium implants for bone-anchored hearing sys-
tems, PLoS One 11 (6) (2016) e0157504
.
[38] F.A . Shah, E. Jergeus, A. Chiba, A. Palmquist, Osseointegration of 3D printed
microalloyed CoCr implants-addition of 0.04% Zr to CoCr does not alter bone
material properties, J. Biomed. Mater. Res. A 10 6 (6) (2018) 1655–1663
.
[39] N. Batra, S. Burra, A.J. Siller-Jackson, S. Gu, X. Xia, G.F. Web er, D. DeSi-
mone, L.F. Bonewald, E.M. Lafer, E. Sprague, M.A. Schwartz, J.X. Jiang, Me-
chanical stress-activated integrin alpha5beta1 induces opening of connexin 43
hemichannels, Proc. Nat. Acad. Sci. US A 109 (9) (2012) 3359–3364
.
[40] B. Hesse, P. Varga, M. Langer, A. Pac ureanu, S. Schrof, N. Mannicke, H. Suho-
nen, P. Maurer, P. Cloetens, F. Peyrin, K. Raum, Canalicular network morphol-
ogy is the major determinant of the spatial distribution of mass density in hu-
man bone tissue: evidence by means of synchrotron radiation phase-contrast
nano-CT, J. Bone Miner. Res. 30 (2) (2015) 346–356 .
[41] M. Kerschnitzki, P. Kollmannsberger, M. Burghammer, G.N. Duda,
R. Wein kamer , W. Wagermaier, P. Fratzl, Architecture of the osteocyte
network correlates with bone material quality, J. Bone Miner. Res. 28 (8)
(2013) 1837–1845 .
[42] M. Hasegawa, J. Saruta, M. Hirota, T. Taniyama, Y. Sugita, K. Kubo , M. Ishijima,
T. Ikeda, H. Maeda, T. Ogawa, A newly created meso-, micro-, and nano-scale
rough titanium surface promotes bone-implant integration, Int. J. Mol. Sci. 21
(3) (2020)
.
[43] N. Tsukimura, T. Ueno, F. Iwasa, H. Minamikawa, Y. Sugita, K. Ishizaki, T. Ikeda,
K. Nakagawa, M. Ya mad a, T. Ogawa, Bone integration capability of alkali-
and heat-treated nanobimorphic Ti-15Mo-5Zr-3Al, Acta Biomater. 7 (12) (2011)
4267–4277
.
[44] M. Yama da, T. Ueno, N. Tsukimura, T. Ikeda, K. Nakagawa, N. Hori, T. Suzuki,
T. Ogawa, Bone integration capability of nanopolymorphic crystalline hydrox-
yapatite coated on titanium implants, Int J Nanomedicine 7 (2012) 859–873
.
[45] M. Yam ada , T. Ueno, H. Minamikawa, T. Ikeda, K. Nakagawa, T. Ogawa, Ear-
ly-stage osseointegration capability of a submicrofeatured titanium surface
created by microroughening and anodic oxidation, Clin. Oral. Implants Res. 24
(9) (2013) 991–1001
.
[46] S. Kuroshima, T. Nakano, T. Ishimoto, M. Sasaki, M. Inoue, M. Yasutake,
T. Sawase, Optimally oriented grooves on dental implants improve bone quality
around implants under repetitive mechanical loading, Acta Biomater. 48 (2017)
433–4 4 4
.
[47] F. Rupp, R.A. Gittens, L. Scheideler, A. Marmur, B.D. Boyan, Z. Schwartz,
J. Geis-Gerstorfer, A review on the wettability of dental implant surfaces I: the-
oretical and experimental aspects, Acta Biomater. 10 (7) (2014) 289 4–2906
.
[48] J.C.M. Souza, M.B. Sordi, M. Kanazawa, S. Ravindran, B. Henriques, F.S. Silva,
C. Aparicio, L.F. Cooper, Nano-scale modification of titanium implant surfaces
to enhance osseointegration, Acta Biomater. 94 (2019) 112–131
.
[49] K.M. Hotchkiss, G.B. Reddy, S.L. Hyzy, Z. Schwartz, B.D. Boyan, R. Olivares–
Navarrete, Titanium surface characteristics, including topography and wetta-
bility, alter macrophage activation, Acta Biomater. 31 (2016) 425–434
.
[50] A. Aoyagi, M. Hata, R. Matsukawa, Y. Imanishi, J. Take be, Physicochemical prop-
erties of anodized-hydrothermally treated titanium with a nanotopographic
surface structure promote osteogenic differentiation in dental pulp stem cells,
J. Prosthodont. Res. 65 (4) (2021) 474–481
.
[51] G. Nardone, J. Oliver-De La Cruz, J. Vrbsky, C. Martini, J. Pribyl, P. Skladal,
M. Pesl, G. Caluori, S. Pagliari, F. Martino, Z. Maceckova, M. Hajduch, A. Sanz–
Garcia, N.M. Pugno, G.B. Stokin, G. Forte, YA P regulates cell mechanics by con-
trolling focal adhesion assembly, Nature Commun. 8 (2017) 15321
.
[52] E. Gongadze, D. Kabaso, S. Bauer, T. Slivnik, P. Schmuki, U. van Rienen,
A. Iglic, Adhesion of osteoblasts to a nanorough titanium implant surface, Int.
J. Na nomed. 6 (2011) 1801–1816 .
[53] M.J. Dalby, N. Gadegaard, R. Tare , A. Andar, M.O. Riehle, P. Herzyk, C.D. Wilkin-
son, R.O. Oreffo, The control of human mesenchymal cell differentiation using
nanoscale symmetry and disorder, Nat. Mater. 6 (12) (20 07) 997–1003
.
[54] I. Tonazzini, S. Meucci, P. Fara ci, F. Beltram, M. Cecchini, Neuronal differentia-
tion on anisotropic substrates and the influence of nanotopographical noise on
neurite contact guidance, Biomaterials 34 (25) (2013) 6027–6036
.
[55] I. Tonazzini, A. Cecchini, Y. Elgersma, M. Cecchini, Interaction of SH-SY5Y cells
with nanogratings during neuronal differentiation: com parison with primary
neurons, Adv. Healthc. Mater. 3 (4) (2014) 581–587
.
[56] P.R . Buenzli, N.A. Sims, Quantifying the osteocyte network in the human skele-
ton, Bone 75 (2015) 144–150
.
[57] J.S. Wang, T. Kamath, C.M. Mazur, F. Mirzamohammadi, D. Rotte r, H. Hojo,
C.D. Castro, N. Toka vani ch, R. Patel, N. Govea, T. Enishi, Y. Wu, J. da Silva
Martins, M. Bruce, D.J. Brooks, M.L. Bouxsein, D. Tokar z, C.P. Lin, A. Abdul, . . .
M.N. We in, Control of osteocyte dendrite formation by Sp7 and its target ge ne
osteocrin, Nat . Commun. 12 (1) (2021) 6271
.
627
