Abstract
Implant-associated infections due to the formation of bacterial
biofilms pose a serious threat in medical healthcare, which needs
effective therapeutic methods. Here, we propose a
multifunctional nanoreactor by spatiotemporal ultrasound-driven tandem
catalysis to amplify the efficacy of sonodynamic and chemodynamic
therapy. By combining piezoelectric barium titanate with polydopamine
and copper, the ultrasound-activated piezo-hot carriers transfer easily
to copper by polydopamine. It boosts reactive oxygen species production
by piezoelectrics, and facilitates the interconversion between Cu2^+
and Cu^+ to promote hydroxyl radical generation via Cu^+ -catalyzed
chemodynamic reactions. Finally, the elevated reactive oxygen species
cause bacterial membrane structure loosening and DNA damage.
Transcriptomics and metabolomics analysis reveal that intracellular
copper overload restricts the tricarboxylic acid cycle, promoting
bacterial cuproptosis-like death. Therefore, the polyetherketoneketone
scaffold engineered with the designed nanoreactor shows excellent
antibacterial performance with ultrasound stimulation and promotes
angiogenesis and osteogenesis on-demand in vivo.
Subject terms: Implants, Biomedical materials, Tissues, Soft materials,
Biomaterials
__________________________________________________________________
Implantation-associated infections often lead to infections. Here, the
authors propose a piezo-based nanoreactor to achieve US-excited tandem
catalysis, endowing the polyetherketoneketone bone scaffold with
on-demand antibacterial and osteogenic capacities.
Introduction
Bone implantation utilizing inert substrates (e.g., metals, polymers,
and ceramics) has emerged as a highly efficacious clinical stratagem in
the realm of long bone fracture stabilization, spinal refurbishment,
and replacement of arthritic joints^[50]1,[51]2. Despite advancements
in minimally invasive surgery and aseptic techniques,
implant-associated infections (IAIs) continue to pose significant
challenges in medical care and overall well-being. The existing
clinical interventions are largely restricted to antibiotics and the
physical removal of infected tissues or implants^[52]3,[53]4.
Unfortunately, the efficacy of systemically administered antibiotics is
compromised by bacterial drug resistance and limited drug penetration
at the infection site. The development of smart coatings, which can
either inherently identify and disrupt biofilm formation on implant
surfaces or be activated remotely to provide antibacterial effects as
needed, holds immense promise for effectively addressing
IAIs^[54]5–[55]7.
Nanomaterials responsive to external stimuli with high tissue
penetration and bio-safety, such as microwave and ultrasound (US), have
been extensively studied for antibacterial treatments^[56]8–[57]10.
Piezoelectric nanomaterials, originating from the non-centrosymmetric
crystalline architecture, have the innate capacity to engender an
indigenous electric field and piezo-potential owing to the
piezoelectric phenomenon, which induces catalytic behavior and
generates reactive oxygen species (ROS) when subjected to US
excitation^[58]11. Large piezo-potential and fast charge separation
ability are two key factors for achieving effective sonodynamic therapy
(SDT) in piezo-based sonosensitizer^[59]12. Therein, the hot carriers
(electron-hole pairs) possessing higher energy above the thermal
equilibrium, emerge as pivotal architects of surface redox dynamics.
However, the mismatch between the ephemeral lifetimes of these hot
carriers (on the order of femtoseconds to nanoseconds) and long
timescales of chemical reactions (spanning milliseconds to seconds)
constitutes a fundamental impediment that gives rise to feeble
catalytic efficacy^[60]13,[61]14. To surmount this challenge,
researchers have conceived metal/semiconductor (Schottky)
heterostructures, allowing hot electrons from metal surfaces with
enough energy to transport across the Schottky barrier and be trapped
at the conduction band (CB) of semiconductors^[62]15,[63]16, prolonging
the lifetime of hot electrons and enhancing catalytic efficiency.
Nevertheless, concerns regarding the potential biotoxicity of metals
necessitate further exploration and investigation.
Combining SDT with chemodynamic therapy (CDT), which catalyzes the
conversion of overexpression of H[2]O[2] into hydroxyl radicals (•OH)
via transition metal catalysts (TMCs)^[64]17,[65]18, may substantially
boost ROS generation. Therein, Copper-based TMCs based on valence
transition between Cu^+ and Cu^2+ have shown effective CDT effect.
Besides, the anomalous copper accumulation in cells has been reported
to directly bind to the tricarboxylic acid cycle (TCA) pathway with
lipid-acylated components, consequently causing cuproptosis in tumor
cells^[66]19, providing new insights into copper-mediated bacterial
death. For example, copper-doped polyoxometalate clusters can treat
biofilm-associated infection by inducing bacterial cuproptosis-like
death^[67]20. However, major limitations of current Cu-based TMCs
include the easy oxidization and exhaustion of Cu^+. Moreover, some
bacteria have developed various strategies to combat copper
accumulation, including upregulating copper transport systems or
producing copper-binding proteins that sequester excess Cu ions^[68]21.
Therefore, developing Cu-based TMCs to yield explosive ROS via the
synergy of SDT and CDT, changing membrane fluidity and permeability of
bacteria, could facilitate bacterial copper overload and related
metabolic interference. However, Cu ions have dose-dependent
antibacterial performance and cytotoxicity^[69]22, and whether Cu ions
can induce cuproptosis-like bacterial death in a concentration lower
than the minimal inhibitory concentration (MIC) and non-cytotoxic
extracellular aggregation remains to be explored. To delve deeper into
the intrinsic relationship between SDT and CDT synergistically
enhancing ROS production, tandem catalysis, which enables rapid and
selective synthesis of therapeutic molecules in live cells via
TMCs-mediated chemical transformation^[70]23–[71]25, holds great
potential to involve the sequential catalytic processes and achieve
synergetic therapeutic effects. However, few previously reported
Cu-based TMCs can be spatiotemporally triggered under external stimuli,
especially ultrasound with high tissue penetration and safety for the
human body.
Here, we proposed a multifunctional nanoreactor based on copper-armed
piezoelectric sonosensitizer to achieve spatiotemporal US-driven
piezo-hot carriers that assisted valence state interconversion between
Cu^2+ and Cu^+. This approach enabled tandem catalysis between SDT and
CDT, ultimately leading to efficient bacteria killing and addressing
IAIs (Fig. [72]1). Specifically, a modified barium titanate (BT)-based
(mBT) piezoelectric sonosensitizer with good SDT effect was
encapsulated by polydopamine (pDA) with conjugated π structure serving
as the “electron aspirator” to extract US-activated piezo-hot carriers
outsides pDA (pBT) (Fig. [73]1b). The quantum transport characteristic
of multilayer stacked pDA was elucidated using the Keldysh
non-equilibrium Green’s function formalism and density functional
theory (NEGF-DFT). Subsequently, the electrons were transferred to
Cu^2+ that chelated with pDA (CpBT), initiating the oxidizing reaction
of Cu^2+ to Cu^+, which converted endogenous H[2]O[2] into •OH.
Subsequently, 3D-printed polyetherketoneketone (PEKK) bone implants,
commonly utilized in orthopedics, were engineered with CpBT
nanoreactors and hydroxyapatite (HA) to obtain the PH-CpBT scaffold.
Notably, the PH-CpBT scaffold displayed excellent antibacterial
performance against Staphylococcus aureus (S. aureus) with US
stimulation by disturbing bacterial membrane homeostasis through
elevated levels of ROS and promoting intracellular flux of Cu^+ into
bacteria. Transcriptomics and metabolomics analysis revealed that
intracellular Cu accumulation inhibited the TCA cycle by the
combination of Cu^+ and lipoylated enzymes, eventually causing
bacterial cuproptosis-like death. Moreover, the elevated levels of ROS
generated by SDT and CDT of CpBT induced DNA and protein damage in
bacteria. Besides, the PH-CpBT scaffold accelerated bone regeneration
by promoting the angiogenesis by Cu ions and osteogenic differentiation
of pre-osteoblasts by releasing Ca and P ions (Fig. [74]1c). We
envisioned that the piezo-based nanoreactors triggering tandem
catalysis to amplify therapeutic effect of SDT and CDT on bone implant
surface can inhibit bacteria effectively, thus providing an on-demand
and noninvasive way in response to the challenge of IAIs.
Fig. 1. Design strategy for US-induced tandem catalysis to trigger robust SDT
and CDT.
[75]Fig. 1
[76]Open in a new tab
a Fabrication of CpBT nanoreactors and PH-CpBT bone scaffold. b Driven
by US-activated tandem catalysis, CpBT nanoreactors can trigger cascade
reactions between SDT and CDT. S. aureus was eradicated through
cuproptosis-like death and cell oxidative stress mediated by ROS. c In
an infected bone implant model, the PH-CpBT bone scaffold showed good
antibacterial, angiogenesis, and osteogenesis multifunctionality in
vivo, synergistically.
Results and discussion
Characterization of CpBT nanoreactors
The preparation process of CpBT nanoreactors was shown in Fig. [77]1a.
The transmission electron microscopy (TEM) image of CpBT revealed a
thickness of pDA of approximately 7 nm, and uniform distribution of Ti
and Cu was displayed by the elemental mapping (Fig. [78]2a). The
lattice fingers in high-resolution TEM image of CpBT (Fig. [79]2b)
agreed well with the d-spacings of (101) planes of pure BT. The
scanning electron microscopy (SEM) image showed CpBT possessed a
uniform size distribution with an average size of several hundred
nanometers (Supplementary Fig. [80]1a). X-ray diffraction (XRD)
patterns of mBT, pBT and CpBT (Fig. [81]2c) exhibited pure perovskite
structure, conforming to the standard pattern of pure BT. The Raman
spectrum of mBT (Supplementary Fig. [82]1b) displayed a first-order
Raman spectrum, with a peak at 305 cm^−1 suggesting the tetragonality.
The average size of CpBT was around 210 nm and the zeta potential was
−21 mV by the dynamic light scattering method, suggesting good
stability in water (Supplementary Fig. [83]2). By the switching
spectroscopy piezoresponse force microscopy (SS-PFM), the amplitude and
phase images of mBT exhibited a standard butterfly amplitude curve and
a near-complete phase shift of 180° at ±15 V, demonstrating the
complete domain switching and strong intrinsic piezoelectric response
(Fig. [84]2d). To assess the piezoelectricity of mBT and pure BT, the
temperature dependence of the relative permittivity was measured
(Fig. [85]2e). The relative permittivity of mBT at body temperature
(37-43 °C) was much higher than pure BT, indicating the higher
piezoelectricity of mBT. Additionally, the Landau free energy modeling
revealed that at temperature of 37-43 °C, the polarization anisotropy
energy and energy barrier (<0.3 J cm^−3) were relatively lower than
that of pure BT ( > 1 J cm^−3) (Fig. [86]2f, Supplementary Figs. [87]3
and [88]4), which led to a small energy barrier for polarization
rotation among T < 001> and O < 110> states^[89]26, hence inducing the
enhanced piezoelectric performance over the body temperature.
Fig. 2. Structural characterizations and US-activated tandem catalysis of
CpBT nanoreactors.
[90]Fig. 2
[91]Open in a new tab
a TEM image and elemental mapping of CpBT. Similar TEM images were
obtained for more than three times experiments. b. HRTEM image of CpBT
with lattice fringes corresponding to the (101) plane. Similar TEM
images were obtained for more than three times experiments. c XRD
patterns. d SS-PFM measurement. e The temperature dependence of
relative permittivity of mBT and pure BT measured at 1 kHz. f
Free-energy profiles for pure BT and mBT at 37 °C. g Comparison of MB
degradation by mBT, pBT, CpBT, and CpBT+H[2]O[2] under US stimulation.
h The reaction mechanism of Cu^+ detection by neocuproine and the
related UV-Vis absorbance spectra of neocuproine treated with CpBT and
US stimulation. i XPS of Cu 2p[3/2] for CpBT before and after US
stimulation (the insets show Cu LMM spectra). j Peak ratio of Cu^+ to
Cu^2+ at 570 eV and 932 eV in XPS spectra of CpBT before and after US
stimulation. k •OH generation by ESR. l The proposed mechanism of
enhanced ROS generation by US-activated tandem catalysis of SDT and CDT
based on CpBT nanoreactors.
US-activated tandem catalysis of CpBT
To evaluate US-activated SDT and CDT performance, methylene blue (MB)
was used to measure in vitro •OH generation. Over time and with
increasing content of CpBT, CpBT gradually degraded MB under US
stimulation (Supplementary Fig. [92]5a and b). Significantly,
CpBT+H[2]O[2] + US further enhanced the degradation of MB, while
CpBT+H[2]O[2] without US stimulation displayed poor degradation
(Fig. [93]2g), highlighting the importance of US-mediated SDT and CDT
in improving •OH generation. Additionally, we validated the production
of singlet oxygen (^1O[2]) using the fluorescent singlet oxygen sensor
green (SOSG) probe, which showed a significant increase in ^1O[2]
production with CpBT under US stimulation (Supplementary Fig. [94]5c
and [95]5d). To figure out the enhanced ROS production in the presence
of H[2]O[2], neocuproine, a Cu^+-specific sequestering agent, was used
as an indicator. In the detection of Cu^+, colorless neocuproine
typically formed yellow complexes [Cu(neocuproine)[2]]^+ (Fig. [96]2h),
showing maximum absorption at 452 nm. Neocuproine solution treated with
CpBT+US turned yellow (Fig. [97]2h), providing strong evidence of the
formation of Cu^+ on CpBT nanoreactor during US stimulation. Besides,
the characteristic absorption peak increased slowly with US time
increased, suggesting that the production of Cu^+ was time-dependent
(Supplementary Fig. [98]6a). Without piezo-electrons produced by mBT,
the reduction of Cu^2+ to Cu^+ under US stimulation failed
(Supplementary Fig. [99]6b and [100]6c). Besides, XPS spectra of CpBT
before and after US stimulation were carried out (Fig. [101]2i). The
higher peak at ~935 eV in Cu 2p[3/2] spectra was assigned to Cu^2+,
accompanied by the characteristic Cu^2+ shakeup satellite peaks
(938–945 eV). The lower peak at ~932 eV suggested the presence of Cu^+
or Cu^0 species. Furthermore, the Cu LMM Auger spectra at ~570 eV
confirmed the presence of Cu^+ after US. Notably, the integral area
ratio of Cu^+ to Cu^2+ after US was significantly enhanced at 935 eV
(from 0.28:1 for fresh CpBT to 0.67:1 for used CpBT) and at 570 eV
(from 0.33:1 for fresh CpBT to 0.5:1 for used CpBT) (Fig. [102]2j),
indicating that part of surface Cu^2+ species were reduced to Cu^+
species during US stimulation. Furthermore, the electron spin resonance
(ESR) technique was performed. The typical equal peaks with 1:2:2:1
represented •OH generation (Fig. [103]2k), and an apparent three-line
spectrum with the peak intensity of 1:1:1 belonged to the signal of
^1O[2] for both pBT and CpBT under US stimulation (Supplementary
Fig. [104]5e). Therefore, US-excited Cu^2+/Cu^+ conversion of CpBT was
proposed for tandem catalysis of SDT and CDT. Specifically, the
electrons generated by sonosensitizer mBT under US stimulation were
transported to Cu^2+ by “electron aspirator” pDA, which enabled the
reduction of Cu^2+ to Cu^+ and transferred H[2]O[2] to •OH
(Fig. [105]2l) via Fenton-like reactions.
Mechanism of US-activated tandem catalysis
The localized orbital locator (LOL) -π variant function was employed to
investigate the delocalization channel of π electrons in pDA
complicated system^[106]27,[107]28, and the isosurface maps of LOL-π
for dopamine (DA), dihydroxyindole (DHI), and DHI oligomers were
illustrated in Supplementary Fig. [108]7. The fully connected LOL-π
isosurfaces of the six-membered rings in all DHI oligomers indicated a
strong conjugation and the evident delocalization of π electrons in
this region. Moreover, the π electron delocalization remained prominent
in the layered aggregation system, with the LOL-π isosurfaces
reflecting the extensive presence of delocalized π electrons over the
six-membered rings (Supplementary Fig. [109]8). To elucidate carrier
transportation between mBT and pDA, the NEGF-DFT was used^[110]29. The
quantum transport architecture devices containing Au electrodes and
16-layer stacked DHI and DA acting as the scattering region were
constructed (Supplementary Fig. [111]9a). The calculated potential
energy distribution of the quantum transport architecture along the
XY-plane was illustrated in Supplementary Fig. [112]9b. The
transmission function, T(E, V[b]), were calculated using the
Landauer-Büttiker formula^[113]30:
[MATH:
T(E,Vb)=<
/mo>Tr[ΓL(E)GR(E)ΓR(E)GA(E)] :MATH]
1
where G^R(E) and G^A(E) represented the retarded and advanced Green’s
functions of the scattering region. V[b], Γ[L](E), and Γ[R](E) were the
bias voltage, the linewidth functions of the left and right Au
electrodes describing the coupling between electrodes and the
scattering region, respectively. The calculated electronic structure
showed that the constructed 16-layer stacked DHI and DA exhibited
semiconducting properties (Figs. [114]3a and [115]3b). Due to its
smaller electronic bandgap (2.61 eV) compared to mBT (3.16 eV,
according to Fig. [116]3e), pDA served as a favorable charge transport
medium during US stimulation, promoting the electron transfer from mBT
to Cu ions. Notably, the transmission spectra (Fig. [117]3c)
demonstrated a zero transmission coefficient near the HOMO and LUMO,
with no observable opened transport channel in the real-space
scattering states below the LUMO energy level (<0.1 eV, 2 in
Fig. [118]3d). The absence of transmission around the HOMO and LUMO
energy levels concurred with the observations of a non-overlapping
interlayer charge density in LOL-π isosurfaces within the layered
aggregation structure. The carrier transmissions occurred beyond HOMO
and LUMO, and the corresponding real-space scattering states of
16-layer stacked DHI and DA-based device at different energy levels
clearly indicated the presence of available interlayer transport
channels (Fig. [119]3d). According to valence band-XPS spectra
(Fig. [120]3f), the valence band maximum (VBM) and the conduction band
minimum (CBM) of mBT were calculated to be 1.90 eV and -1.26 eV, which
was enough to trigger the redox reaction of -OH/•OH. Besides, the
finite element analysis (Fig. [121]3g) demonstrated that the higher
piezo-potential of mBT (0.035 V) compared with pure BT (0.010 V)
(Supplementary Fig. [122]10) with the cavitation pressure of 10^8 Pa,
endowing full energy for US-induced hot carriers to transport. To
assess the electron transmission in pDA-Cu^2+/Cu^+ composites, the
charge density difference between pDA-Cu^+ and pDA-Cu^2+ complexes with
different pDA configurations was calculated (Fig. [123]3h), showing
that the oxidation of Cu^+ or the reduction of Cu^2+ led to significant
charge transport in pDA-Cu^2+/Cu^+ composites. The negative charge
density differences based on the natural population analysis of Cu ion
[Cu^+-Cu^2+] further confirmed the reduction of charge on Cu ion in
pDA-Cu^+ composites compared to the pDA-Cu^2+ composites. These
findings provided strong evidence for charge transfer to Cu ion,
consistent with the observations from the charge density difference
isosurfaces. Thus, the involvement of pDA as a charge transport medium
was essential for the charge transfer during the oxidation process of
Cu^+ or the reduction process of Cu^2+. As a result, US-induced hot
carriers by mBT with large piezo-potential injected into the enveloped
pDA, which underwent interlayer carrier transport through high-energy
scattering states, and were eventually transferred to Cu^2+ ions at the
surface, enabling the reduction of Cu^2+, eventually displaying the
intrinsic ability to assemble catalytic process for spatiotemporal
control over cascade reactions like in living systems (Fig. [124]3i).
Fig. 3. Mechanism of US-activated tandem catalysis for enhancing SDT and CDT.
[125]Fig. 3
[126]Open in a new tab
a Energy levels and b the corresponding density of states (DOS) of
16-layer stacked DHI and DA. c The transmission spectra of the 16-layer
stacked DHI and DA-based transport architecture device. d The real
space scattering states of the 16-layer stacked DHI and DA system at
different energy levels. All the figures share the same color bar. e
UV-Vis diffuse reflectance spectra of mBT. f VB-XPS spectra of mBT and
CpBT. g Finite element method simulation for piezo-potential
distribution on the surface of mBT. h The charge difference value
achieved from natural population analysis of Cu ion (Cu^+-Cu^2+).
Isosurface map of charge density difference computed from first
principles for DHI-Cu^+-Cu^2, Dimer-Cu^+-Cu^2, Trimer-Cu^+-Cu^2, and
Tetramer-Cu^+-Cu^2 (Isosurface in 0.003 e Å^−3). Violet and yellow
colors correspond to positive and negative differences, respectively.
The silver gray sphere represented Cu ion. i The mechanism of
US-activated tandem catalysis of SDT and CDT.
Antibacterial activity of PH-CpBT scaffold in vitro
The 3D-printed PEKK bone scaffold (PEKK, P) was designed and modified
with pDA (Pp) (Supplementary Fig. [127]11), and loaded with nano-HA
(PH) and CpBT nanoreactors (PH-CpBT). Each rod within the PEKK scaffold
had a diameter of approximately 300 μm, with an approximate gap of 200
μm between adjacent rods (Supplementary Fig. [128]12a). The CpBT was
uniformly distributed on the scaffold (Supplementary Fig. [129]12b and
[130]12c). The Young’s modulus of PH-CpBT measured 248 Mpa, closely
matching the range of human trabecular bone (6.9 to 199.5 Mpa)
(Supplementary Fig. [131]13a and b). Furthermore, CpBT improved the
hydrophilicity properties of the PEKK scaffold (as indicated by water
contact angle, Supplementary Fig. [132]13c), facilitating the incipient
adhesion of osteocytes. To assess the stability of CpBT coating on
PH-CpBT with US stimulation, the released CpBT was detected by
inductively coupled plasma (ICP) (Supplementary Fig. [133]14). Process
A contained the released Cu ions and CpBT from scaffolds, and process B
contained the released Cu ions only. It showed that the concentration
of Cu of two processes by ICP was similar, indicating that CpBT on
PH-CpBT surface was stable due to the strong adhesion of pDA^[134]31.
Subsequently, in vitro antibacterial ability of different scaffolds
against S. aureus and Escherichia coli (E. coli) was assessed
(Supplementary Fig. [135]15). Under US stimulation, PH-CpBT with
H[2]O[2] addition exhibited an efficient antibacterial rate of 99.90%
for S. aureus (Fig. [136]4a) and 99.95% for E. coli (Fig. [137]4b). In
the absence of US, PH-CpBT displayed inefficient antibacterial
properties for the unactivated SDT. Compared with Pp+US, PH + US, and
PH-pBT, bacteria cultivated with PH-pBT+US, PH-CpBT+US, and
PH-CpBT+H[2]O[2] photographed by SEM displayed varying degrees of
deformation (Supplementary Fig. [138]16), while those in
PH-CpBT+H[2]O[2] + US exhibited wrinkled membranes, indicating the
inactivation of bacteria. The bacterial biofilm was severely disrupted
after PH-CpBT+H[2]O[2] + US treatment, with wafery and dispersive
features and intense red fluorescence observed, indicating that a mass
of bacteria in biofilm were killed (Fig. [139]4c). To observe
microscopic changes in dead bacteria, Bio-TEM images of bacteria were
examined (Supplementary Fig. [140]17a). Bacteria incubated with
PH-CpBT+H[2]O[2] + US showed incomplete walls and cytoplasmic leakage
in both E. coli (Fig. [141]4d) and S. aureus (Fig. [142]4e), and
bacteria with H[2]O[2] + US maintained intact morphology. Besides, the
element mapping of bacteria showed intracellular copper content was
distinct in PH-CpBT+H[2]O[2] + US compared with US + H[2]O[2] and
PH-CpBT+H[2]O[2] (Supplementary Fig. [143]17b), indicating an increased
intracellular copper inward flow due to the membrane permeability
alteration induced by evaluated ROS (Fig. [144]4f and Supplementary
Fig. [145]18). Previous reports showed that Cu ions have dose-dependent
antibacterial activity^[146]22. To find out the antibacterial effect of
Cu ions in this work, Cu ions released from PH-CpBT scaffolds with or
without US stimulation were analyzed by an ICP spectrometer
(Supplementary Fig. [147]19). The total amount of released Cu ions was
3-7 μg L^-1 for each PH-CpBT scaffold, much lower than MIC of Cu^2+
ions (630 ug L^-1 for S. aureus and 63-630 ug L^-1 for E.
coli)^[148]32. Furthermore, S. aureus or E. coli. were co-incubated
with the leaching solution of PH-CpBT after US stimulation (30 min) for
24 h (Supplementary Fig. [149]20a), finding that the antibacterial rate
was lower than 50%, indicating that the antibacterial performance by
only released Cu ions was poor. Besides, bacteria were also incubated
with PH-CpBT+US and tetrathiomolybdate (TTM, a copper chelator). In
this situation, Cu ions cannot get into the bacteria, resulting in an
antibacterial rate of 70% due to ROS attack (Supplementary
Fig. [150]20b and [151]20e-g). Furthermore, mBT and Cu ions coating on
Pp without pDA as interlayer exhibited a lower antibacterial rate than
PH-CpBT (Supplementary Fig. [152]20c), suggesting that the electron
transport within pDA was a key factor for the reduction of Cu^2+. These
results proved that the antibacterial mechanism was more intricate than
the release of Cu ions. In addition, we compared the antibacterial
activity of CpBT with other piezoelectric materials (including BT, ZnO,
MoS[2] and TiO[2]), finding that CpBT showed better antibacterial
activity (Supplementary Fig. [153]21). The leaking of bacterial
proteins was detected using the protein leakage assay, and PH-CpBT
exhibited the highest protein leakage, indicating significant membrane
permeability alteration and cytoplasmic leakage after US-activated
tandem catalysis (Supplementary Fig. [154]22a). Besides, the total
bacterial DNA was quantified by Bacterial DNA Kit for different groups
(Supplementary Fig. [155]22b). Notably, only a few intact DNA was
detected in PH-CpBT+H[2]O[2] + US, indicating severe damage to
bacterial DNA due to the initiated ROS storm.
Fig. 4. Antibacterial activity of PH-CpBT scaffold in vitro.
[156]Fig. 4
[157]Open in a new tab
The antibacterial results of Pp, PH, PH-pBT, and PH-CpBT against a S.
aureus and b E. coli with or without US stimulation. c 3D CLSM images
of S. aureus biofilms by Live/Dead staining following incubation on
different scaffolds with or without US stimulation. A representative
image of three biological replicates from each group was shown. The
microstructure and element mapping of Cu of d S. aureus and e E. coli
observed by Bio-TEM. A representative image of three biological
replicates from each group was shown. f The fluorescence image of
intracellular ROS in S. aureus with DCFH probe. A representative image
of three biological replicates from each group was shown. Volcano plot
for the distribution of DEGs in g PH-CpBT+H[2]O[2] and h PH-CpBT+US
compared with control. KEGG enrichment of top 20 relevant pathways in
response to i PH-CpBT+H[2]O[2] and j PH-CpBT+ H[2]O[2] + US. k Left,
heat map showing the differential expression of genes of interest.
Right, the death pathways by ROS-induced oxidative stress and
Cu-induced cuproptosis-like death. l A correlation analysis between
differential metabolites and differentially expressed genes. Red
rectangles indicated increased metabolites, and green rectangles
indicate decreased metabolites; red circles indicated up-regulated
genes, and green ovals indicated down-regulated genes. DBT:
Dihydrolipoamide Branched Chain Transacylase E2, GCSH: Glycine Cleavage
System Protein H, DLST: Dihydrolipoamide S-Succinyl transferase, DLAT:
Dihydrolipoamide S-Acetyltransferase. a, b n = 3 biologically
independent samples; ANOVA followed by Tukey’s multiple comparisons;
data were presented as mean values ± standard deviations (SD); error
bars = SD. Significant differences between groups were indicated as
****p < 0.0001, ***p < 0.001, **p < 0.01, and *p < 0.05. Source data
are provided as a Source Data file.
Antibacterial mechanism of PH-CpBT scaffold
To gain insights into the comprehensive antibacterial effects of
PH-CpBT on S. aureus, RNA-sequencing analysis was conducted. Volcano
plots revealed that out of a total of 2622 expressed genes in both
groups, and 642 genes were downregulated and 669 genes were upregulated
in PH-CpBT+H[2]O[2] without US (Fig. [158]4g) compared with Control.
Furthermore, PH-CpBT+H[2]O[2] + US exhibited 1152 significantly
differentially expressed genes (DEGs), with 546 genes downregulated and
606 genes upregulated (Fig. [159]4h). Correlation analysis confirmed
the comparability of three groups (Supplementary Fig. [160]23). The
DEGs were conditioned to Kyoto Encyclopedia of Genes and Genomes (KEGG)
annotation analysis, showing that both PH-CpBT+H[2]O[2] and
PH-CpBT+H[2]O[2] + US had significant genes primarily enriched in
membrane transport, folding and degradation, translation, lipid
metabolism, and energy metabolism (Supplementary Fig. [161]24). KEGG
pathway enrichment analysis and generated bubble plots depicting the
top 20 metabolic pathways elucidated PH-CpBT+H[2]O[2] had significant
effects on S. aureus in ribosome, citrate cycle (TCA cycle),
biosynthesis of various secondary metabolisms, purine metabolism,
amonoacyl-tRNA biosynthesis, and RNA degradation (Fig. [162]4i).
However, PH-CpBT+H[2]O[2] + US predominantly affected the citrate cycle
(TCA cycle), ribosome, C5-Branched dibasic acid metabolism,
biosynthesis of various secondary metabolisms, purine metabolism,
amonoacyl-tRNA biosynthesis, and protein export (Fig. [163]4j).
Notably, TCA cycle pathway exhibited a high enrichment index and a
small p-value in both enrichment analyses, which led us to delve into
the concept of cuproptosis. Cuproptosis related to a type of cell death
triggered by excessive copper that entered the cell and got reduced to
the more toxic Cu^+ state. This process led to the aggregation of four
enzymes linked to the lipoic acid in the TCA cycle, along with
disruptions in Fe-S cluster protein functionality^[164]19.
Interestingly, upregulation or downregulation of related genes was
observed in PH-CpBT+H[2]O[2] + US and PH-CpBT+H[2]O[2], leading to a
phenomenon named bacterial cuproptosis-like death^[165]20. These gene
changes were more pronounced in PH-CpBT+H[2]O[2] + US (Fig. [166]4k).
This can be attributed to the eruptive ROS production in
PH-CpBT+H[2]O[2] + US facilitated by the synergistic SDT and CDT
effect, which not only compromised the integrity of the bacterial cell
wall and membrane but also affected the proton transport channel,
leading to increased influx of copper ions into the bacterial cytoplasm
(Fig. [167]4d and e). Consequently, copper overload transpired,
resulting in changes in associated genes and a more pronounced
cuproptosis-like death in PH-CpBT+H[2]O[2] + US. An intriguing
observation was the relatively inconspicuous upregulation of the
RS07245 gene (a reductase, similar to FDX1) in PH-CpBT+H[2]O[2] + US.
This gene was responsible for the reduction of Cu^2+ to Cu^+, which was
crucial to cuproptosis-like death process^[168]33. The limited
upregulation of it in PH-CpBT+H[2]O[2] + US may be attributed to the
higher content of Cu^+ released by CpBT nanoreactors. It was known that
Cu^+ exhibited stronger cuproptosis toxicity, which explained the
greater observation of cuproptosis-like death in bacteria in
PH-CpBT+H[2]O[2] + US. In results, CpBT achieved the destruction of
bacterial biofilm, induced DNA damage and protein leakage of bacteria
through the toxicity of ROS-related oxidative stress, and the
inhibition of TCA cycle by cuproptosis-like effects, ultimately
culminating in complete sterilization. To present the changes in
metabolites and the relevant genes in the TCA cycle more clearly, we
conducted targeted detection of TCA cycle metabolites. We observed the
lowest levels of metabolites in PH-CpBT+H[2]O[2] + US compared with
Control and PH-CpBT+H[2]O[2] (Supplementary Fig. [169]25).
Figure [170]4l vividly illustrated the changes in TCA cycle metabolites
and the genes influenced by them. In short, under the influence of ROS,
bacteria with PH-CpBT+H[2]O[2] + US experienced an increase in
intracellular Cu, which bound with four lipoylation enzymes (DLAT,
GCSH, DBT, and DLST)^[171]19,[172]34, leading to their deactivation and
consequent reduction in metabolites throughout the pathway.
Additionally, due to the formation of more lipoylation proteins, lipoic
acid was reduced^[173]35, resulting in the upregulation of the genes
that regulated lipoic acid synthesis^[174]36. Ultimately, the decreased
metabolites, through negative feedback regulation, increased the
expression of genes regulating the TCA cycle. Therefore, through
metabolomics and transcriptomics, it was evident that copper-induced
cuproptosis-like death, causing an overall reduction in metabolism, and
in synergy with ROS, effectively killed bacteria.
In vivo antibacterial capability
To assess the antibacterial performance in a live setting, S.
aureus-contaminated scaffolds were implanted into femoral condyle bone
defects in the experimental Sprague Dawley (SD) rats (Fig. [175]5a and
Supplementary Fig. [176]26). The baseline weights of all groups were
equivalent before surgery (Supplementary Fig. [177]27). Detection of
IAIs in clinical scenarios typically relied on routine blood tests
conducted on the second day after surgery. Therefore, we performed US
stimulation at the implant site from the first to 6th-day post-surgery
to replicate the real-life scenario in vivo (Fig. [178]5a). On the
7th-day post-surgery, the animals were sacrificed to gather the femurs
for bacteriological and histological analysis. Visual examination
revealed the presence of secretions and pus at the implant site of Pp
and PH, and partial mitigation were observed in PH-pBT, PH-CpBT (US-),
and PH-mBT/Cu (Supplementary Fig. [179]28). By comparison, the implant
site of PH-CpBT+US and vancomycin (Van.) exhibited smooth tissue
healing without secretion and pus formation, suggesting the clearance
of bacterial infection. The order of bacterial colonies on agar plates
and the turbidity of the Luria-Bertani (LB) medium after cultivation
were as follows: Pp ≈ PH > PH-pBT ≈ PH-CpBT (US-) > PH-mBT/Cu > PH-CpBT
≈ Van, suggesting that PH-CpBT exhibited favorable in vivo
antibacterial characteristics rivaling Van (Figs. [180]5b and [181]5c).
For PH-mBT/Cu, the lack of pDA as “electron aspirator” to initiate the
reduction of Cu^+, the antibacterial performance by SDT and released
Cu^2+ only was inefficient. For PH-pBT, no induction of copper-induced
cuproptosis-like bacterial death occurred due to the absence of Cu
ions, also exhibited moderate antibacterial effect. Hematoxylin and
Eosin (H&E) revealed typical signs of bone tissue infection in Pp and
PH, characterized by a large number of lymphocytes, monocytes, and
neutrophil infiltrations into the tissues (Fig. [182]5d).
Semi-quantitative analysis corroborated that PH-CpBT and Van. exhibited
the least neutrophils, affirming efficient disinfection and minimal
inflammatory response (Fig. [183]5e). Furthermore, Giemsa staining
suggested the presence of numerous bacteria (Figs. [184]5f and
[185]5g). Only a few bacteria were observed in PH-CpBT and Van.,
further supporting the effective bactericidal effect of PH-CpBT in in
vivo environment and its potential for clinical sterilization.
Fig. 5. The treatment of implant infection with PH-CpBT in vivo.
[186]Fig. 5
[187]Open in a new tab
a Schematic illustration of infected modified PEKK scaffolds and animal
experimental treatment in therapeutic. b Photographs of bacterial
colonies and turbid liquid. c Quantitative analysis of bacterial turbid
liquid by OD 600. A representative image of three biological replicates
from each group was shown. d H&E images and e semi-quantification of
neutrophil in the infected bone tissues surrounding the implants. The
red arrows represented neutrophils, and the green arrows represented
lymphocytes. A representative image of three biological replicates from
each group was shown. f Giemsa staining images and g
semi-quantification of bacteria in the infected bone tissues
surrounding the implants. The black arrows represented bacteria. A
representative image of three biological replicates from each group was
shown. c, e, g n = 3 biologically independent samples; ANOVA followed
by Tukey’s multiple comparisons; data were presented as mean
values ± SD; error bars = SD. Significant differences between groups
were indicated as ****p < 0.0001, ***p < 0.001, **p < 0.01, and
*p < 0.05. Source data are provided as a Source Data file.
Bone regeneration in vitro and in vivo
The central objective of antibacterial treatment was to enhance bone
ingrowth. To verify cell adhesion and proliferation, various scaffolds
were cultured with MC3T3-E1 cells in vitro. Cell viability by Cell
Counting Kit-8 (CCK-8) assay showed almost no cytotoxicity in all
scaffolds (Supplementary Fig. [188]29). Besides, all scaffolds except
for P and Pp showed similar proliferation rates (Supplementary
Fig. [189]30). With 3 days of culture, cells efficiently infiltrated
the entire volume of PH-CpBT, and their actin cytoskeleton exhibited an
elongated morphology throughout the scaffold, indicating good
biocompatibility and cell adhesion (Supplementary Fig. [190]31). SEM
images (Supplementary Fig. [191]32) showed superior cell spreading and
tight adhesion to PH-CpBT, and cells exhibited more extended filopodia
than other scaffolds due to the excellent hydrophilicity of PH-CpBT.
Following 7 and 14 days of osteogenic induction, activity was analyzed
(Supplementary Fig. [192]33), suggesting the significant osteogenic
differentiation of PH-CpBT. Similar findings were observed in Alizarin
Red S (ARS) staining (Supplementary Fig. [193]34), where PH-CpBT
exhibited a significant increase in calcium nodules after 14 and 21
days of incubation. The number of tube formations and junctions for
tubulogenesis of human umbilical vein endothelial cells (HUVECs) was
higher in PH-CpBT (Supplementary Fig. [194]35), underlining its
advantageous angiogenic properties.
At 4th and 8th weeks post-implantation in vivo, the rat femur was
scanned by micro-computed tomography (Micro-CT). 3D reconstructions of
the femoral condyle and bone defect areas, facilitated by Imaris
software, uncovered substantial bone loss around bone defects in Pp and
PH, attributed to osteolysis induced by bacterial proliferation and
spread (Supplementary Fig. [195]36). Conversely, PH-CpBT exhibited
robust in vivo osteogenicity, evident from the extent of new bone
formation (red section) (Fig. [196]6a, Supplementary Movie [197]1).
Over time, new bone in PH-CpBT effectively filled the scaffold pores,
including the middle of the scaffold. The presence of gaps in the top
and front views of the new bone indicated the scaffold’s suitability
for bone growth. Although Van. exhibited similar antibacterial ability,
PH-CpBT outperformed it by highlighting the good osteogenic ability.
Based on micro-CT quantitative data, parameters such as bone
volume/total volume (BV/TV), bone mineral density (BMD), trabecular
number (Tb.N), trabecular separation (Tb.Sp), and trabecular thickness
(Tb.Th) were most distinguished in PH-CpBT (Fig. [198]6b-f). This
suggested that excellent antibacterial and osteogenic abilities were
essential for PH-CpBT in vivo. To evaluate the rate of new bone
deposition, calcitonin and alizarin red through intraperitoneal
injection were used to mark the bone formation line. PH-CpBT exhibited
the largest distance and area between the two calcium deposition lines
(Fig. [199]6g). Quantitative analysis, including mineral apposition
rate (MAR), revealed that PH-CpBT stimulated faster bone deposition
compared to the other groups (Fig. [200]6h).
Fig. 6. The number and speed of new bone growing into the scaffold.
[201]Fig. 6
[202]Open in a new tab
a Mico-CT of femoral condyle, from up to down, reconstruction of the
defect and new bone ingrowth in the scaffolds, top view of new bone,
side view of new bone. The green and pink cylinder showed the new bone
growth for the first four weeks and the next four weeks, respectively.
A representative image of four (4 W) and six (8 W) biological
replicates from each group was shown. Quantitative statistics of bone
regeneration related index in 3D reconstruction by micro-CT including b
BV/TV, c BMD, d Tb. N, e Tb. Sp, and f Tb. Th. g Calcitonin (green) and
alizarin red (red) marked new bone. A representative image of three
biological replicates from each group was shown. h Quantitative
statistics of MAR from 4th week to 6th week. b–f n = 4 biologically
independent samples (4 W), n = 6 biologically independent samples
(8 W); h n = 4 biologically independent samples. b–h ANOVA followed by
Tukey’s multiple comparisons; data were presented as mean values ± SD;
error bars = SD. Significant differences between groups were indicated
as ****p < 0.0001, ***p < 0.001, **p < 0.01, and *p < 0.05. Source data
are provided as a Source Data file.
New bone ingrowth: histopathology, histochemistry, and immunofluorescence
The undecalcified implant-bone tissue underwent H&E staining and
Goldner trichrome staining after 4th and 8th weeks of implantation
(Fig. [203]7a). The H&E staining clearly showcased heightened new bone
formation encircling the implant, alongside more pronounced bone
ingrowth within the pores of PH-CpBT. Quantitative evaluation of H&E
staining showed that PH-CpBT had a higher percentage of new bone area
(60.1% at 8 weeks) compared to other groups and Van. (Fig. [204]7b).
Goldner trichrome staining demonstrated PH-CpBT exhibited more new
mineralized bone surrounding implant (Fig. [205]7a), and quantitative
analysis confirmed highest amount of new tissues in PH-CpBT
(Fig. [206]7c). These two staining demonstrated that bone ingrowth was
primarily observed at the edge of the scaffolds, gradually penetrating
deeper inside the porous scaffolds over time, and the fine structure of
new bone resembled the surrounding normal bone at the scaffold edge,
tightly attaching to PH-CpBT. Histological analysis of harvested organs
(heart, liver, spleen, lung, and kidney) from the rats in PH-CpBT
showed no organic changes, indicating satisfactory biosafety
(Supplementary Fig. [207]37). Blood routine testing also revealed
favorable indicators in the experimental group (Supplementary
Fig. [208]38), confirming the satisfactory biocompatibility and minimal
side effects of PH-CpBT. Moreover, immunohistochemistry demonstrated
that the expression of tumor necrosis factor-α (TNF-α) decreased in
PH-CpBT (Supplementary Fig. [209]39), confirming the remarkable
antibacterial effect of PH-CpBT in eradicating infection and reducing
inflammation in vivo, which contributed to accelerated bone
regeneration and vascularization. The ingrowth of bone and blood
vessels at the defect site was evaluated using four-color
immunofluorescence (DAPI, CD31, BMP-2, and RUNX2) based on tyramide
signal amplification (TSA). PH-CpBT exhibited the strongest
fluorescence intensity (Fig. [210]7d). Immunofluorescence analysis
pinpointed a higher density and fluorescence intensity of CD31 in
PH-CpBT, signifying an augmented level of vascularization facilitated
by copper (Fig. [211]7e). The activity of osteogenesis-associated
proteins RUNX2 and BMP-2 exhibited distinct upregulation in PH-CpBT
(Figs. [212]7f and [213]7g), indicating the implant’s ability to
promote bone formation, meeting the prerequisites for optimal bone
growth within the implants. The upregulated expression of these
osteogenesis-related factors can be attributed to the antibacterial
ability of PH-CpBT under US and Cu loading, which effectively
facilitated osteogenesis, demonstrating the significant osteogenic
promotion of PH-CpBT. In general, this study provided a promising
strategy for designing multifunctional bone implants with simultaneous
highly effective antibacterial and osteogenic capacities based on
US-activated tandem catalysis, offering a reference for future clinical
applications.
Fig. 7. The microstructure and mechanism of new bone ingrowth.
[214]Fig. 7
[215]Open in a new tab
a H&E and Goldner staining of undecalcified tissue around the implant
at 4th and 8th weeks post-surgery. A representative image of three
biological replicates from each group was shown. b Quantification of
H&E in the implants at 8th weeks. c Quantification of Goldner positive
area in the implants at 8th weeks. d Immunofluorescence staining of
CD31, BMP-2, and RUNX2 surrounding tissues of implants at 8th weeks
post-surgery. A representative image of three biological replicates
from each group was shown. Quantification of e CD31, f BMP-2, and g
RUNX2 immunoreactivity within the implants. b, c, e, f, g n = 3
biologically independent samples; ANOVA followed by Tukey’s multiple
comparisons; data were presented as mean values ± SD; error bars = SD.
Significant differences between groups were indicated as
****p < 0.0001, ***p < 0.001, **p < 0.01, and *p < 0.05. Source data
are provided as a Source Data file.
Methods
Chemicals
Barium carbonate (99%), titanium dioxide (99%), calcium carbonate
(99.9%), zirconium dioxide (99.99%), cupric sulfate (99%), and nano HA
were from Sinopharm Chemical Reagent Co., Ltd. (China). Dopamine
hydrochloride (DA-HCl), tri(hydroxymethyl) amino methane hydrochloride
(Tris-HCl, 1 M), methylene blue (MB), vancomycin, Calcein, and ARS were
from Sigma Chemical Co. (USA). SOSG was from Invitrogen (USA). The
Bacterial Live/Dead Bac Light viability kit was from Thermo Fisher
(USA). Calcein/PI Cell Viability/Cytotoxicity Assay Kit, CCK-8 kit, and
ARS dye were purchased from Beyotime (China).
Synthesis of CpBT nanoreactors
The mBT was prepared in the following steps: Weigh the elemental
components according to the chemical formula
Ba[0.90]Ca[0.10]Ti[0.91]Zr[0.09]. Ball-mill the weighed components in
nylon jars for 24 hours. Calcined the obtained powders at 1260 ^oC for
3 hours. Sand-milled the calcined powders at 2000 revolutions per
minute for 4 hours to obtain mBT. Then, 1 mg mL^-1 mBT was dispersed in
the Tris-HCl solution (10 mM, pH = 8.5) and sonicated for 40 min.
DA-HCl was added and sonicated for another 10 min and stirred for
2 hours. After centrifuging and washing with DI water for three times,
pBT nanoparticles were obtained. Then, 10 uM cupric sulfate solution
was added to pBT and stirred for 2 hours, and then were collected by
centrifugation and washed with DI water for three times to obtain CpBT
nanoreactor.
Preparation of PH-CpBT scaffold
3D printing technology utilizing fused deposition modeling was employed
to fabricate PEKK scaffolds. The design of the structures was
accomplished using Materialise 3-Matic software, resulting in bone
scaffolds measuring 3 mm in diameter and 4 mm in height for in vivo
experiments, as well as 10 mm in diameter and 1 mm in height for in
vitro experiments involving cells and bacteria. Medical-grade PEKK
filaments were extruded into the deposition bin of the 3D printer,
enabling layer-by-layer preparation of the scaffolds according to
pre-determined shapes at a temperature of 200 ^oC. The PEKK scaffold
was stirred in Tris-HCl solution (10 mM, pH = 8.5) containing
3 mg ml^-1 DA-HCl for 24 hours to obtain PEKK@pDA (Pp). Then Pp
scaffold was immersed in HA solution (2 mg ml^-1) for 12 hours to get
Pp@HA (PH). The scaffolds of PH-pBT and PH-CpBT were constructed by
soaking PH scaffolds in different solutions (2 mg ml^-1) for 12 hours.
Characteristics
The crystal structures were unveiled through XRD utilizing Cu Kα
radiation (λ = 1.5406 Å) (Empyrean, Malvern Panalytical, UK). The SEM
(SUPRA 55, Carl Zeiss AG, Germany) was used to examine the surface
morphologies. High-resolution portraits and the related lattice fringes
were captured by HRTEM (Talos F200i, FEI, USA). Raman spectra were
examined by Raman spectroscopy with an excitation source of 532 nm
(Invia Reflex, Renishaw, UK). The chemical state and valence band were
obtained by XPS (K-Alpha + , Thermo Fisher, USA). The switching
spectroscopy piezo-response force microscopy loops were collected by a
commercial atomic force microscope (MFP−3D, Asylum Research, UK). The
temperature-dependent dielectric constant (ε[r]-T) was obtained via an
LCR meter (TH2816A, Tonghui, China). The 5982 Universal testing machine
(Instron, USA) was used to perform the compression test. The contact
angle of the surfaces of scaffolds was measured by A JY-82C contact
angle apparatus (Dingsheng Testing Equipment Co. Ltd., China). Then,
the concentration of Cu was obtained by an inductively coupled
plasma-optical emission spectrometer (ICP-OES, model 5100, Agilent,
USA). Cu^2+ ions released from PH-CpBT scaffolds were analyzed by an
ICP spectrometer (ICP-MS, 7850, Agilent, USA). Specifically, PH-CpBT
scaffolds were immersed in a 0.9 % NaCl solution at 37 ± 1 °C with the
surface-area to -volume ratio was 3 cm^2 mL^-1 according to the
international standard ISO 10993-12. Triplicate samples were used to
obtain an average value with standard deviation.
Detection of ROS in vitro
The ROS generated from the samples when exposed to US stimulation were
tested using MB, SOSG, and ESR. To assess •OH production, different
samples were mixed with a solution containing MB and subjected to US
stimulation (1 MHz, 1.0 W cm^-2, 50% duty cycle) with or without
H[2]O[2] (50 μM). The changes in absorption of MB at 664 nm before and
after US stimulation were recorded using the ultraviolet and visible
spectrophotometer (UV-5200, METASH, China). Similarly, the presence of
singlet oxygen was confirmed by the fluorescence intensity of SOSG at
525 nm using a fluorescence spectrophotometer (F-7000, Hitachi, Japan).
The identification of ROS species was performed using an ESR
spectrometer (JES-FA200, JEOL, Japan). For trapping the singlet oxygen,
we utilized 2,2,6,6-tetramethylpiperidine (TEMP) at a concentration of
50 mM, while for detecting •OH, we employed
5,5-dimethyl-1-pyrroline-N-oxide (DMPO) at a concentration of 0.1 mM.
Detection of Cu^+ of CpBT
Neocuproine (Aladdin, China) was selected as an indicator for detecting
in-situ Cu^+ production. Briefly, 1.04 mg of neocuproine was dissolved
in 5 mL of ethyl alcohol and then diluted five times with ultrapure
water. Subsequently, the buffer solution at pH 6.5 was prepared by
dissolving KH[2]PO[4] and NaOH in ultrapure water. CpBT dispersion at a
concentration of 400 μg mL^-1 was prepared by dissolving itself in
ultrapure water, and then the final working solution, consisting of
0.75 mL of buffer solution, 1.0 mL of CpBT solution, and 0.8 mL of
neocuproine solution, was irradiated under sonication for 0-12 min
(1 MHz, 1.0 W cm^-2, 50% duty cycle). Finally, the absorption of the
reaction solution was measured by a UV-vis absorbance spectrometer at
452 nm.
Simulation details
The dimer, trimers, and tetramers structures of pDA employed in the
current study were extracted from the most stable geometry among all
the structures generated using a brute-force algorithmic
generator^[216]37. The layered aggregates consistent of the stacked
dopamine (DA) and dihydroxyindole (DHI) via π-π interactions^[217]31.
The pDA-Cu^2+/Cu^+ composites were constructed by chelating Cu^2+/Cu^+
ions near the dihydroxyl site^[218]38,[219]39. Unless otherwise
specified, the Becke three-parameter Lee-Yang-Parr (B3LYP) functional
was employed to optimize the molecular structure and generate the
wavefunction, together with the def2-TZVP basis set. The BLYP
functional combined with the 6−31+g (d, p) basis set was employed to
optimize the structures of the layered aggregates and analyze the
wavefunction. To account for weak interactions in the layered
aggregates systems, the dispersion correction DFT-D3 with Becke-Johnson
damping (D3BJ) was implemented^[220]40. The Pipek-Mezey method^[221]41
was chosen to localize occupied molecular orbitals (MOs) and
differentiate σ and π characters. For investigating the delocalization
channel of π electrons, the LOL-π variant of the localized orbital
locator (LOL) function was utilized^[222]27,[223]28. All quantum
chemistry calculations were performed using the Gaussian16
program^[224]42, while wavefunction analysis was conducted using the
Multiwfn 3.8 (dev) code^[225]43. Isosurface graphs were generated using
the VMD program for improved visual representation. To investigate the
quantum transport properties of multilayer stacked pDA, we designed a
vertical transport architecture device with Au electrodes being the
source and drain contacts, and pDA was acted as the transport channel
region. The simulations were performed using the first-principles
methods, with the combination of the NEGF-DFT^[226]29, conducted in the
first-principles quantum transport software package Nanodcal^[227]44.
The local density approximation (LDA) was embraced to portray the
exchange and correlation function, and atomic cores were defined by the
standard norm-conserving nonlocal pseudo potentials^[228]45. A double-ϛ
polarized (DZP) atomic orbital basis^[229]46 was conducted for Au metal
electrode and pDA to expand all physical quantities with a kinetic
energy cutoff of 4500 eV. Furthermore, a k-point mesh of 3 × 3 × 1 and
17 × 17 × 1 was applied to sample the first Brillouin zone for
integrations in the reciprocal space of the scattering region (pDA) and
Au electrode, respectively. In addition, self-consistent calculations
were converged until each component of the density matrix have declined
to 10^−5 Hartree. Employing the Landau-Ginsburg-Devonshire
phenomenological model, the Landau free energy was expressed as
△G = α[1](P[1]^2 + P[2]^2 + P[3]^2) + α[11](P[1]^4 + P[2]^4 + P[3]^4) +
α[12](P[1]^2P[2]^2 + P[2]^2P[3]^2 + P[1]^2P[3]^2) + α[111](P[1]^6 + P[
2]^6 + P[3]^6) + α[112][P[1]^4(P[2]^2 + P[3]^2) + P[2]^4(P[1]^2 + P[3]^
2) + P[3]^4(P[1]^2 + P[2]^2)] + α[123]P[1]^2P[2]^2P[3]^2, where α[1],
α[11], α[12], α[111], α[112] and α[123] represented Landau energy
coefficients. Finite element method calculations were carried out with
COMSOL Multiphysics 5.4 with a model of a piezoelectric device based on
a steady-state study.
Culturing of bacteria
Gram-positive S. aureus (ATCC 25923) and Gram-negative E. coli (ATCC
25922) were cultured in a sterile Luria-Bertani (LB) medium (10 g L^−1
of back to-tryptone,10 g L^−1 of NaCl, and 5 g L^−1 of bacto-yeast
extract). The bacterial counts were obtained from the spread plate of
different samples.
Antibacterial assessment in vitro
The effectiveness of Pp, PH, PH-pBT, and PH-CpBT against S. aureus and
E. coli with and without US stimulation was assessed using the spread
plate method. The different scaffolds were exposed to bacterial
suspensions (2×10^7 CFU mL^-1) in 48-well plates for specified
durations and subjected to US stimulation (1 W cm^-2) for 9 mins or
left untreated. Bacterial growth was cultured on agar plates at 37 °C
for 18 h using the spread plate method to quantify colony-forming units
(CFUs). To simulate the high H[2]O[2] environment in vitro,
PH-CpBT+H[2]O[2] was supplemented with 50 μM H[2]O[2]. Furthermore,
under identical conditions, antibacterial experiments were conducted to
compare copper chelator (TTM, 20 μM, Aladdin, China), copper release
from the scaffold, the coating of mBT and copper ions (PH-mBT/Cu), and
similar materials (commercial TiO[2], ZnO, MoS[2], and BT, Aladdin,
China). The adherent bacteria on PEKK were fixed, dried, dehydrated,
and coated with gold. SEM (ZEISS Gemini 300, Germany) was employed to
examine bacterial morphology and integrity. Bacteria treated with US
and PH-CpBT+H[2]O[2] + US were collected, cryo-centrifuged at high
speed for 2 mins, embedded into blocks, and sliced into 50 nm thick
sections using an ultrathin microtome (EM UC7, Leica, Germany).
Subsequently, TEM images were captured after staining the samples with
uranyl acetate and lead citrate, and placing on copper grids. After 5
days of bacterial cultivation on various scaffolds, a live/dead
staining assay was conducted employing the Live/Dead BacLight viability
kit. The data were collected by the confocal laser scanning microscopy
(CLSM, N-SIM S, Nikon, Japan). Moreover, intracellular ROS levels were
measured using the ROS assay kit (Beyotime, China) four hours after
coculturing with different scaffolds, and the outcomes were also
visualized using CLSM. The assessment of bacterial DNA damage induced
by various samples utilized the Bacterial DNA Kit (Beyotime, China).
After treatment, bacteria underwent collection through centrifugation
at 5000× g for 5 mins at 4 °C. Following the manufacturer’s
instructions, the Bacterial DNA kit was employed for the purification
of intact DNA strands from the bacteria. Subsequently, complete DNA
fragments were subjected to quantitative analysis using a UV-vis
spectrophotometer. This comprehensive methodology aimed to discern and
quantify the extent of DNA damage caused by the diverse samples. The
supernatant obtained by centrifugation was detected by BCA Protein
Quantitation Assay (Beyotime, China), and the bacterial leakage protein
was quantified by a microplate reader.
Transcriptome analysis
S. aureus was cultivated with PH-CpBT with and without US stimulation.
The bacteria were collected by cryo-centrifugation during the
logarithmic phase and immediately frozen in liquid nitrogen. Three sets
of biological replicates were carried out under identical conditions:
the control group (US-1, US-2, US−3), the experimental group 1
(PH-CpBT+H[2]O[2]-1, PH-CpBT+H[2]O[2]-2, PH-CpBT+H[2]O[2]−3), and the
experimental group 2 (PH-CpBT+H[2]O[2] + US-1, CpBT+H[2]O[2] + US-2,
CpBT+H[2]O[2] + US−3). Total RNA isolation and cDNA library
construction were executed according to the manufacturer’s
instructions, followed by sequencing on an Illumina HiSeq platform at
Majorbio Bio-pharm Technology Co., Ltd. The expression quantification
results were subjected to analysis using DESeq2 (Version 1.24.0)
software to discern DEGs with a screening threshold of |log2FC | ≥ 1
and a p-value < 0.05. We utilized Kyoto Encyclopedia of Genes and
Genomes (KEGG, [230]http://www.genome.jp/kegg/) databases to elucidate
the biological implications of DEGs and explain gene functional
differences between samples. Furthermore, we conducted KEGG pathway
analysis by KOBAS. The significance of KEGG pathways was evaluated
through a Fisher’s exact test.
Metabonomics analysis
The sample preparation for the metabolomic analysis was performed as
described in transcriptome analysis, and metabolomics analysis was
performed under the same conditions for each group of six biological
replicates. The experimental procedure was as follows: preparation of
the central carbon metabolite standard solution; pre-processing of the
standard curve, treatment of the bacterial cell precipitate samples,
and qualitative and quantitative detection of target compounds in the
samples using LC-ESI-MS/MS (UHPLC-Qtrap) with the ExionLC AD system;
Chromatographic analysis was performed using the Waters HSS T3 column
(2.1 ×150 mm, 1.8 μm); Mass spectrometry analysis was conducted using
the SCIEX QTRAP 6500+ in both positive and negative modes. Finally, the
Sciex OS quantitative software was utilized for the automatic
identification and integration of ion fragments to analyze the data.
Biocompatibility assessment in vitro
Mouse osteoblastic MC3T3-E1 cells (GNM15) were purchased from the Cell
Bank of the Chinese Academy of Sciences (Shanghai, China). The HUVECs
(CP-H082) were purchased from Procell Life Science & Technology Co.,
LTD. (Wuhan, China). MC3T3-E1 cells cultured with α minimum essential
medium (α-MEM, Gibcom, USA), including 10% fetal bovine serum (FBS,
Gibcom, USA) and 1% penicillin-streptomycin solution. The cultures were
maintained in a humidified atmosphere incubator with 5% CO[2] at 37 °C
for in vitro cytocompatibility and osteogenic differentiation studies.
Cell proliferation and spreading assay
Cell viability and proliferation were quantified through a CCK-8 assay.
Various scaffolds were co-cultured with cells, and the optical density
(OD) at 450 nm was gauged on days of 1, 3, and 5 using a microplate
reader. Moreover, cells were seeded onto different scaffolds for 3
days, and their viability was gauged using a Calcein/PI Cell
Viability/Cytotoxicity Assay Kit, followed by observation using laser
scanning CLSM. MC3T3-E1 cells were seeded onto different scaffolds in a
24-well plate for 24 h, and the cells were stained with fluorescein
isothiocyanate-phalloidin and 4,6-diamidino-2-phenylindole, then the
arrangement of F-actin and the cell nuclei were then observed using
CLSM.
Osteogenic differentiation and angiogenesis in vitro
MC3T3-E1 cells were seeded on the surface of different scaffolds. After
24 h, the culture medium α-MEM was replaced with an osteoinductive
medium containing 10 mM β-glycerophosphate, 50 µg mL^-1 ascorbic acid,
and 10 nM dexamethasone (all from Sigma, USA) to prompt the osteogenic
commitment of MC3T3-E1 osteoblasts. This osteoinductive medium was
renewed every 3 days. On days 7 and 14, the activity of ALP was
assessed using a 5-bromo-4-chloro−3-indolyl phosphate/nitro-blue
tetrazolium (BCIP/NBT) ALP color development kit (Beyotime, China). For
the evaluation of calcified extracellular matrix, cells were fixed and
treated with ARS dye on days 14 and 21. Then subsequent to the removal
of excess dye, images were captured using a scanner. HUVECs were
cultured for in vitro angiogenesis investigation. The suitable cells
and samples were introduced into the μ-Slide 15 well 3D (Ibidi,
Germany) and allowed to incubate for 6 hours. Afterward, the cells were
stained with Calcein and subsequently imaged using a fluorescence
microscope.
In vivo experiments
Male SD rats (200–220 g, about two-months old) were bought from the
Beijing Huafukang Bioscience Cojnc. Animal experimentation in this
study received ethical approval from the Laboratory Animal Ethics
Committee of West China Hospital, Sichuan University (IACUC number
20221216022). All rats were raised in 25 ± 3 °C (temperature), 60-70%
(humidity), and 12 h light/dark cycle conditions for two weeks before
the experiments. SD rats (n = 8 per group) were divided randomly into
seven groups: Pp, PH, PH-pBT, PH-mBT/Cu, PH-CpBT(-), PH-CpBT, and Van..
Except for PH-CpBT(-) and Van., other groups underwent ultrasound
intervention. To create an implant-related contamination model, the
engineered PEKK implants (⦶ 3 mm × 4 mm) were immersed in S. aureus (2
× 10^7 CFU ml^-1) solution at 37 °C for 4 hours. Subsequently, a rat
lateral femoral condyle model for cylindrical bone defect repair was
created, and the implants with S. aureus were implanted in the
bilateral femoral condyles. After 1st, 4th, and 8th weeks, the rats
were sacrificed to assess the antibacterial effect and new bone
formation.
Antibacterial activity in vivo
Starting from the first to sixth days after surgery, rats were exposed
to US stimulation (1 MHz, 1.0 W cm^-2, 50% duty cycle) for 8 mins under
gas anesthesia induction. On the seventh day post-surgery, the rats
were sacrificed, and the implants were removed and placed in sample
collection tubes containing PBS to collect bacteria. Dilutions were
cultured on agar plates to assess implant infection. Moreover, bone
tissue around the scaffolds was gathered for histological analysis,
including H&E and Giemsa staining. These sections were observed and
captured using an inverted microscope (Olympus BX53, Japan).
The volume of new bone and bone growth rate
To evaluate the volume of new bone and trabecular thickness within the
implants, femoral condyles were scanned using the Quantum GX Micro CT
(PerkinElmer, USA). The 3D reconstruction of the CT images was
generated by Imaris 9.9 (BitPlane, Oxford Instruments). Following
reconstruction, parameters such as BV, Tb. Th, % BS/BV, % BV/TV, Tb. N,
and BMD were quantified using Skyscan NRecon software. For the
assessment of new bone growth rate, ARS (30 mg kg^-1) and Calcein
(20 mg kg^-1) were intraperitoneally injected into the rats at 4th and
6th week after implantation. Infected rats were euthanized at 8th week
post-surgery, and their implants and surrounding bone were fixed,
sliced, and examined using CLSM.
The microstructure of new bone ingrowth: histomorphometry,
immunohistochemistry, and immunofluorescence
Histological sections were generated parallel to the long axis of the
implants around the undecalcified femoral condyle before
decalcification. Sections were ground to 100 μm thickness, and slides
were polished and stained with H&E and Goldner. Additionally, other
bone samples without implants were fixed, decalcified, dehydrated, and
embedded in paraffin. The tissues were cut into sections, being
prepared for IHC staining (TNF-α, 1:200, Abcam ab1793) and 4-color
immunofluorescence staining with the following primary antibodies
(Abcam, UK): CD31 (1:500, ab182981), RUNX2 (1:200, ab236639), BMP-2
(1:200, ab214821), and DAPI (1:1000, ab285390) at 4 °C overnight.
Finally, 4-color immunofluorescence staining was performed and
visualized with Vectra Polaris (PerkinElmer, USA).
Statistics analysis
Data were presented as mean values ± standard deviations (SD); error
bars = SD. Statistical analysis was performed using ANOVA followed by
Tukey’s multiple comparison with statistical significance assigned at
*P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001.
Reporting summary
Further information on research design is available in the [231]Nature
Portfolio Reporting Summary linked to this article.
Supplementary information
[232]Supplementary Information^ (7MB, pdf)
[233]Peer Review File^ (10.6MB, pdf)
[234]41467_2024_45619_MOESM3_ESM.pdf^ (132.1KB, pdf)
Description of Additional Supplementary Files
[235]Supplementary Movie 1^ (40.6MB, 7z)
[236]Reporting Summary^ (2.7MB, pdf)
Source data
[237]Source Data^ (44.2MB, xlsx)
Acknowledgements