Abstract
Epithelial malfunction rescue is the decisive step involved in complete
trachea repair; however, this step remains challenging due to the harsh
tracheal environment and unclear pathogenesis, which still made current
bioengineered trachea transplants receive fatal complications. Herein,
bacterial infection-induced neutrophilic oxidative stress imbalance and
epithelial stemness loss were identified as the pathogenic factors.
Targeting pathogenesis, multiplexed hydrogels with adhesive and
anti-fouling Janus sides, anti-swelling and anti-bacteria properties
are constructed to adapt in mucous and causative agent-rich trachea
environments. In two epithelial injury models and two tracheal
transplantation-related epithelial deficiency models, the hydrogels
blockade oxidative stress-innate immune cascade axis, reactivate
epithelial mucociliary regenerative ability to rescue epithelial
malfunction with stenosis-free mucociliary epithelium regeneration.
Importantly, the versatility of hydrogel is validated via its
integration with routine bioengineered vascular and cartilage
transplants, wherein the regenerated pseudostratification epithelium,
cartilage and vascularization resemble native-like trachea, resulting
in the complete tracheal repair including structure and respiratory
function reinvigoration. Our research provides insights into epithelial
interface diseases and guides related biomaterials design.
Subject terms: Tissue engineering, Transplantation, Biomedical
materials
__________________________________________________________________
Epithelial malfunction is an unmet critical challenge in trachea
repair. Here, Chen et al. developed a multiplexed hydrogel to promote
tracheal repair with native-like epithelium in pre-clinical models,
exhibiting clinical translation potential.
Introduction
Tracheal stenosis or injury, which represents a major public health
concern, usually requires complete tracheal repair^[62]1–[63]4, but the
tremendous difficulty in enabling the composition, structure and
functions of the repaired trachea to match those of natural tracheal
tissue continues to exist. Thus, complete tracheal repair remains an
intractable but critical challenge. The trachea generally consists of
cartilage, blood vessels and epithelial layers, which play critical
roles in maintaining airway opening, supplying oxygen and nutrition and
inhibiting airway bacterial adherence, respectively^[64]5–[65]8. To
achieve complete tracheal repair, the regeneration of cartilage, blood
vessels and epithelial layers needs to occur in a synchronous manner.
Nevertheless, current attentions have been paid only to cartilage
regeneration^[66]9–[67]12, and the complex pathogenesis and harsh
tracheal environment of tracheal injury have discouraged researchers to
carry out functional epithelial layer repair^[68]13–[69]17. Therefore,
epithelial deficiency is ubiquitous in current bioengineered tracheal
transplantation patients and in the ageing population; moreover, this
deficiency often causes tracheal epithelial malfunction and severe
complications and respiratory disturbances^[70]18,[71]19. Thus,
complete trachea repair is still hard to reach, and engineered tracheal
substitutes fail to meet clinical demands. In light of the vital
functions^[72]20–[73]22, rescuing epithelial layer malfunctions is
highly important for decreasing fatal complications and inhibiting
respiratory disturbances in common clinical scenarios. Unfortunately,
repeated polluted airflow stimulation and mucus layer barrier lead to
the failure of current epithelial transplantation methods and drug
treatment in rescuing epithelial malfunctions, as well as the failure
of complete trachea repair^[74]23–[75]26.
In this study, we utilize clinical samples and identify innate immune
and epithelial regenerative ability (or stemness) losses as two
dominant pathogenesis of epithelial malfunction (Fig. [76]1,
Supplementary Fig. [77]1). Building on these insights, we devise a
tracheal lumen-hydrogelation strategy, in which tracheal
environment-adapted and multiplexed hydrogels with Janus surfaces are
constructed to target pathogenesis, rescue epithelial malfunction, and
eventually achieve the complete trachea repair. In such multiplexed
self-adaptable hydrogels, one Janus surface could quickly adhere to the
diseased area in the fully mucous airway environment, and another Janus
surface is armed with a favorable anti-fouling ability to generate a
stable and degradable protection shield against causative agents (such
as breathing pollutants) in the air for epithelial rehabilitation.
Under this circumstance, the activation and infiltration of MPO^+
neutrophils will be hampered, effectively restraining innate immune
hyperactivation and preventing epithelial malfunction in pathogenesis.
Concurrently, ROS scavenging by the L-arginine burst is expected to
further resist inflammation and promote tracheal repair. Moreover, the
slow release of non-covalent tannic acid is anticipated to promote the
immigration and differentiation of CK14^+ basal cells^[78]27–[79]30.
Fig. 1. Neutrophilic oxidative stress hyperactivated innated immune and
impaired epithelial regeneration.
[80]Fig. 1
[81]Open in a new tab
a Procedures for collecting and analyzing clinical samples. b
Representative image of epithelial malfunction trachea. Scale bar,
200 μm; and 10 μm in magnified regions. HE staining exhibited gross
view. Neutrophils were stained with MPO (orange), oxidative stress were
characterized by ROS (red), and macrophages were stained with CD206
(marker of anti-inflammatory M2 phenotype, green) and CD86 (marker of
pro-inflammatory M1 phenotype, red). DAPI (white), cell nuclei. c
representative image of healthy human tracheal HE staining exhibited
the gross view of native tracheal epithelial structure. Scale bar,
200 μm. d Representative IF staining images of tracheal innate immune
in native healthy trachea. Scale bar, 50 μm; and 10 μm in magnified
regions. The white dashed line represented the basement membrane (the
junction line between the epithelial layer and the submucosal layer).
Neutrophils were stained with MPO (orange), oxidative stress were
characterized by ROS (red). Cell nuclei, DAPI (white). macrophages were
stained with CD206 (green, anti-inflammatory phenotype) and CD86 (red,
pro-inflammatory phenotype). Cell nuclei, DAPI (white). e Quantitative
statistics of MPO^+ cells number per high power field (HPF) in b and d.
f quantitative statistics of CD86^+ cells number per HPF in b and d. g
constructed epithelial injury model in CD177 knockout mice to block the
neutrophils-initiated oxidative stress and evaluate epithelial
function. Representative HE images of WT group and CD177^-/- group at 7
days after epithelial injury model. Elements were created in BioRender.
Yi, C. (2025) [82]https://BioRender.com/m4mpwtc. Scale bar, 50 μm.
n = 3 in epithelial malfunction group and n = 2 in native human trachea
group. n = 6 in mice experiments. Source data are provided as a Source
Data file. Data are presented as mean ± SD. All error bars represent
SD. p values calculated using a two-tailed unpaired t-test.
Depending on these designs, such multiplexed self-adaptable hydrogels
quickly adhere to the diseased wet area, form a stable, anti-swelling,
anti-bacterial, innate immune-regulating barrier, and promote the
differentiation of CK14^+ basal cells into de novo multifunctional
epithelial mucociliary elevator in both rabbit acute and recurred
epithelial injury model (Supplementary
Fig. [83]1)^[84]15,[85]31–[86]33. Especially in two rabbit
bioengineered tracheal transplantation models (partial tracheal defect
model and long circumferential tracheal defect model), the hydrogels
also result in complete trachea repair with structural and functional
reinvigoration after integration with routine vascular and cartilage
repair routes. Intriguingly, the degradation of such multiplexed
self-adaptable hydrogels maintains a similar pace with the trachea
regeneration, thus guaranteeing no respiration disturbance and high
safety during repair. The generated trachea shapes into a native-like
cartilage-vessel-epithelium structure with healthy respiratory
functions, and no stenosis complications are found. Overall, this
pioneering work provides a solid foundation to complete structural and
functional repair of trachea, which highlights epithelial malfunction
rescuing, and the good compatibility with existing bioengineered
tracheal construction techniques demonstrates potential for the high
clinical translation potential.
Results
The screening of pathogenesis targets in clinics
To identify the pivotal pathogenesis, we compared the pathological
changes between epithelial malfunction tissues in patients and normal
tracheal tissues in healthy donors (Fig. [87]1a, Supplementary
Table [88]1, [89]2). Hematoxylin eosin (HE) and immunofluorescence (IF)
staining confirmed epithelial malfunction at the site of injury (Fig.
[90]1a–f). After epithelial injury, repetitive pollutants transmitted
via breathing can lead to persistent infiltration and activation of
MPO^+ neutrophils in the diseased area. The MPO^+ neutrophils further
released large amounts of reactive oxygen species (ROS), and induced
CD86^+ proinflammatory macrophage (M1 macrophage) infiltration
(Fig. [91]1c). Under physiological conditions (Fig. [92]1c, d),
neutrophils play an important role in immune defense, in which process
neutrophils destroy phagocytosed microbes in a ROS-dependent manner and
then undergo apoptosis and disappear after resolution of the
inflammatory stimulus^[93]34–[94]36. While continuous infiltration and
activation of neutrophils were observed in area of epithelial
malfunction, thereby indicating the possibility that bacterial
infection and injury-related stress induced neutrophils and downstream
innate immune hyperactivation, which ultimately caused epithelial
malfunction (Fig. [95]1b–f).
Therefore, the hypothesis was raised that due to epithelial injury or
deficiency, the tracheal tissue underwent injury stress and would be
constantly exposed to a polluted environment, thereby leading to
recurrent infection and neutrophil infiltration. The continuous
activation and infiltration of neutrophils hyperactivated the innate
immune system, thus causing epithelial malfunction. To validate this
hypothesis, we first analyzed the number of tracheal bacteria in
tracheas in the sham group and the epithelial injury group
(Supplementary Fig. [96]2). The results aligned with our hypothesis
that the bacterial burden of both Gram^+ and Gram^- pathogens was
greater at the site of epithelial injury compared to healthy tracheas.
Furthermore, we evaluated the role of hyperactivated oxidative stress
and bacterial infection via co-culture models (Supplementary
Fig. [97]3). Pseudomonas aeruginosa (P.a.), which is a commonly
observed bacterium at sites of tracheal injury, was selected for
co-culture with epithelial cells either with or without oxidative
stress stimulation (Supplementary Fig. [98]3a). An MOI (multiplicity of
infection) of 5 was considered to indicate mild bacterial burden,
whereas an MOI of 20 was considered to indicate severe bacterial
burden. When the MOI was 5, ROS were demonstrated to exert a positive
bactericidal effect and protect the function of the epithelium to some
extent (Supplementary Fig. [99]3b, c). However, when the MOI reached
20, the bactericidal effect of ROS alone was weakened, thus indicating
that it was unable to protect the epithelium (Supplementary
Fig. [100]3b, e). Subsequently, neutrophil-bacteria-epithelial cells
co-culture model was established (Supplementary Fig. [101]3e). Via
co-culture and ROS scavenging experiments, neutrophils were
demonstrated to be capable of killing bacteria. However, the ROS
released during this process could also impair the function of
epithelial cells (Supplementary Fig. [102]3f–h). At an MOI of 20, the
rescue effect of ROS scavenging on epithelial function was markedly
attenuated compared with that at an MOI of 5, thereby indicating that
bacterial-mediated epithelial cytotoxicity also plays a substantial
role in epithelial malfunction (Supplementary Fig. [103]3f–h).
To further determine the role of neutrophilic ROS in epithelial
multifunctional differentiation (Supplementary Fig. [104]4a). We
utilized ESR spectra and demonstrated neutrophils co-cultured with P.a.
released ROS (e.g., superoxide radicals), and the addition of an ROS
scavenger eliminated ROS in the culture supernatant (Supplementary
Fig. [105]4b). The staining results demonstrated that the supernatant
of the neutrophil-P.a. co-cultured medium directly impaired the
differentiation of epithelial organoids, manifesting as a malfunctional
phenotype (including reduced thickness, decreased basal cells, and the
loss of mucociliary structures; Supplementary Fig. [106]4c–j). ROS
scavenging protected organoid differentiation at an MOI of 5 but failed
to restore functionality at an MOI of 20 (Supplementary Fig. [107]4d,
e). The abovementioned results preliminarily confirmed our hypothesis
in vitro; specifically, the increased bacterial burden observed at the
injured tracheal site drove persistent ROS release by neutrophils but
had limited bactericidal efficacy. Both imbalanced ROS levels and
bacterial secretions were identified as two key contributors to
epithelial malfunction (which included the loss of epithelial
barrier/multi-differentiation function).
To further validate this hypothesis in vivo, we constructed an
epithelial injury model in neutrophil function-deficient mice
(C57BL/6N-Cd177^em1C/Cya, or CD177^−/− mice), where CD177 is an
important surface marker for neutrophils, and its knockout impairs the
functions of neutrophils and disfavors innate immune
hyperactivation^[108]37–[109]40. It is found that CD177 knockout
promoted epithelial regeneration and increased the differentiation
ratio of functional cilia cells (Supplementary Fig. [110]5), indicating
that the function-deficiency of neutrophils in a bacteria-free
environment could rescue epithelial malfunction (Fig. [111]1g). All of
the abovementioned results reveal that bacterial infection and further
neutrophilic oxidative stress could hyperactivate innate immune system
and impair epithelial regenerative ability to aggravate epithelial
malfunction, with these effects potentially representing therapeutic
targets for epithelial malfunction.
Tracheal environment-matched and multiplexed self-adaptable hydrogel
synthesis
Hydrogels have shown a wide application domain after rational design
and engineering procedures according to the specific
demands^[112]41–[113]45. To target this specific pathogenesis, we
designed multiplexed self-adaptable hydrogels (LA@PA-TA@C) featuring
high stability, Janus adhesion, anti-swelling and anti-bacteria to
reprogram the innate immune and rescue epithelial regenerative ability
(Fig. [114]2a, Supplementary Fig. [115]1). Such multiplexed hydrogels
adapting to tracheal environment were constructed via a hybrid
crosslinking strategy combining dynamic multiple non-covalent
self-assembly and photopolymerization (Fig. [116]2b). Specifically, the
chitin (C)-assisted tannic acid (TA) self-assembly via hydrogen bonding
was firstly enforced to form C@TA hydrogel. Concurrently,
poly-(Aspartic acid) (PA) network is activated by NHS via replacing
-COOH (Fig. [117]2b, c), and then loaded with L-arginine (LA) to obtain
L@PA hydrogels. Immediately afterwards, unilateral photopolymerization
between the C@TA hydrogel and L@PA via amide bonds was enacted to
eventually obtain LA@PA-TA@C hydrogels (Fig. [118]2b, d). The Janus
structure is observed in LA@PA-TA@C with C@TA and L@PA serving as the
upper anti-adhesion side and bottom tissue-adhesion side, respectively,
as indicated by the structural and composition differences between the
two Janus sides (Fig. [119]2e, f).
Fig. 2. Design, fabrication, and characterization of multiplexed
self-adaptable LA@PA-TA@C hydrogel.
[120]Fig. 2
[121]Open in a new tab
a Hydrogel design principles inspired by pathogenesis and adapted to
environment. b the synthesis and fabrication process of LA@PA-TA@C
hydrogel. c ^1H NMR spectra evolution of N-acryloyl aspartic acid and
PA. d ATR-FTIR spectra of different groups of hydrogels (PA, PA-C,
LA@PA-C, LA@PA-TA@C). e representative SEM images of adhesive PA
hydrogel compared with Janus adhesive LA@PA-TA@C hydrogel (upper
anti-adhesive surface: TA@C, bottom adhesive surface: LA@PA). Scale
bar, 200 μm. f Element distribution spectrum (EDS) and images of upper
and bottom surface of LA@PA-TA@C hydrogel. Scale bar, 200 μm. g–i in
vivo stability. g stress value at 75% strain of LA@PA-C and LA@PA-TA@C
hydrogel treated with or without 0.2% lactic acid for 1 h. h
representative frequency sweep rheological plots of LA@PA-TA@C hydrogel
treated with or without 0.2% lactic acid for 1 h. i, in vivo stability
and degradation of LA@PA-TA@C hydrogel. Scale bar, 10 mm. j–m the
adhesion principle and effect of Janus adhesive LA@PA-TA@C hydrogel. j
adhesion principle of PA. k gross view and test of adhesion strength.
*, LA@PA-TA@C hydrogel was treated with double-side adhesion for
adhesion strength test. l Janus adhesive performance of two sides of
LA@PA-TA@C hydrogel. m gross view and cross-sectional image of
LA@PA-TA@C hydrogel adhered in the tracheal lumen of the pig, and
adhesion to rabbit trachea under extreme stress-induced deformation. n
the swelling curve of different groups of hydrogels. o swelling
performance of LA@PA-TA@C hydrogel in different water environment (PBS,
PBS + 5% trypsin, PBS + 0.2% lactic acid, and PBS + 200 μM H[2]O[2]). p
anti-bacterial principle and effect. Anti-Gram^+ Staphylococcus aureus
test was performed via co-culturing with hydrogels. Anti-bacterial test
of Gram^- E. coli and multidrug-resistant Mycobacterium abscessus
(19977) were performed via antibacterial ring test. n = 3 samples per
group, where each sample was from an individual experiments. Source
data are provided as a Source Data file. Data are presented as
mean ± SD. All error bars represent SD. p values calculated using
one-way ANOVA followed by Tukey’s post hoc test.
Janus adhesion, anti-swelling and anti-bacteria assessments
In general, metabolic acid accumulation in the epithelial malfunction
site presents a threat to hydrogel stability. Intriguingly, the
mechanical strength of the LA@PA-TA@C hydrogels was reinforced by
lactic acid (Fig. [122]2d, h, Supplementary Fig. [123]6), and the
LA@PA-TA@C hydrogels were able to resist rapid degradation in vivo
(Fig. [124]2i). Beyond that, LA@PA-TA@C hydrogels are imparted with
other distinctive properties. They display good Janus adhesiveness
through covalent amide cross-linking between NHS ester groups in bottom
tissue-adhesive PA hydrogels and amino residues on tissues
(Fig. [125]2j, Supplementary Fig. [126]7), and both LA loading and
crosslinking with C@TA failed to influence the tissue adhesion strength
of PA network (Fig. [127]2k, Supplementary Movie [128]1). Unilateral
modification of LA@PA with TA@C imparted the two sides of LA@PA-TA@C
hydrogels with different adhesive performance on tissues
(Fig. [129]2l), in which the PA-originated Janus adhesion property was
inherited, whereas C-originated anti-adhesion owning to no
tissue-binding groups is introduced. All of these properties allowed
the LA@PA-TA@C hydrogels to be adhered on the tracheal lumen, even if
the trachea was subjected to extreme bending or stretching motions
(Fig. [130]2m, Supplementary Movie [131]2).
Both the upper C and bottom PA hydrogels feature high swelling
ratios^[132]46,[133]47, usually disfavoring repair. Appealingly,
noncovalent interactions with TA inhibited C hydrogel swelling, and the
cross-linking between C and PA also resisted the swelling of C and PA
to some extent. Therefore, these anti-swelling actions rendered
L@PA-TA@TA hydrogels to harvest a much lower swelling rate of less than
110% in different aqueous media, and such anti-swelling properties were
maintained even when challenged by acidic, trypsin, or oxidative stress
environments (Fig. [134]2n, o). Meanwhile, the LA@PA-TA@C hydrogel
could be degraded in vivo within 14 days, exhibiting biodegradability
(Supplementary Fig. [135]8).
Due to the anti-bacterial ability of C (including the effect of
rupturing bacterial cell walls)^[136]45 and TA (including the effect of
interfering with bacterial cell membrane synthesis)^[137]28,
co-culturing and antibacterial ring experiments confirmed the
antibacterial efficacy of the LA@PA-TA@C hydrogel against clinically
common bacteria, such as, Gram^+ Staphylococcus aureus, Gram^-
Escherichia coli, and even multidrug-resistant Mycobacterium abscessus
(Fig. [138]2p). The antibacterial ability of the LA@PA-TA@C hydrogel
was further verified in an ex vivo trachea explant model and an in vivo
epithelial injury model. The results demonstrated that after the
LA@PA-TA@C hydrogel adhered to the surface of the ex vivo trachea
explants, it was able to inhibit bacterial proliferation (Supplementary
Fig. [139]9). Similarly, when the tracheal epithelial injury model
treated with the LA@PA-TA@C hydrogel for 7 days was subjected to tissue
homogenization and plate coating, the bacterial quantity in the
LA@PA-TA@C-treated tracheas was significantly lower than that in the
epithelial injury group and the sham group (Supplementary Fig. [140]2).
Oxidative stress-innate immune cascade axis programming
TA has been demonstrated to scavenge ROS^[141]28, and LA can serve as a
reducing agent^[142]48,[143]49. Based on these characteristics, this
multiplexed Janus hydrogels exhibited a good oxidative stress
scavenging ability (Supplementary Fig. [144]10a, b), as represented by
a significant decrease in superoxide radicals and singlet oxygen
signals; additionally, the results of the DPPH and ABTS scavenging
tests revealed almost 100% free radical scavenging (Supplementary
Fig. [145]10c, d). This can be attributed to the continuous release of
LA and TA within 14 days (Supplementary Fig. [146]10e). In particular,
the TA@C-LA@PA hydrogels adapted to oxidative stress environment to
achieve the differential TA release manner (Supplementary
Fig. [147]10f). To further assess the ROS scavenging ability, an in
vitro oxidative stress model was established via the addition of
H[2]O[2]. LA@PA-TA@C hydrogels were observed to downregulate
intracellular ROS levels and stabilize the mitochondria membrane
(Supplementary Fig. [148]10g–i), and the oxidative stress scavenging
protected 3T3 cells and increased cell viability to above 95%
(Supplementary Fig. [149]10j, Supplementary Fig. [150]11). Furthermore,
we utilized macrophages to validate the regulatory capacity of
LA@PA-TA@C on the innate immune system. Both IF staining (Supplementary
Fig. [151]12a, b) and flow cytometry (Supplementary Fig. [152]12c, d)
analyses demonstrated that LA@PA-TA@C treatment upregulated CD206
expression (Supplementary Fig. [153]12a, d) and induced an increased
elongation ratio of morphology (Supplementary Fig. [154]12b), which
manifested as a characteristic anti-inflammatory M2 macrophage
phenotype. Additionally, when challenged under oxidative stress
conditions (via the addition of 200 μM H[2]O[2], Supplementary
Fig. [155]12e–h), the LA@PA-TA@C hydrogel exhibited comparable efficacy
in promoting M2 polarization.
These results indicated that LA@PA-TA@C hydrogels could remodel the
oxidative stress environment, and especially modulate the phenotype of
innate immune-associated cells such as macrophages, thereby
successfully reprogramming oxidative stress-innate immune cascade axis.
In vitro epithelial multidirectional regenerative ability reshaping
After validating the ability of the LA@PA-TA@C hydrogel to reprogram
the oxidative stress-innate immune axis, we explored whether LA@PA-TA@C
hydrogel could rescue epithelial basal cells in an oxidative
stress-rich environment (Fig. [156]3a). Consistent with the
aforementioned findings, oxidative stress led to epithelial cell death
and tight junction destruction, but the LA@PA-TA@C hydrogels protected
the integrity and barrier function of epithelial cells (Fig. [157]3b,
c, Supplementary Fig. [158]13). Scratching experiments with H[2]O[2]
stimulation revealed that LA@PA-TA@C hydrogels accelerated epithelial
cell migration and wound healing within 48 h (Fig. [159]3d, e,
Supplementary Fig. [160]14), and the IF staining of ZO-1 and Ki67
biomarkers revealed epithelial proliferation and tight junction
maintenance (Fig. [161]3d, f, Supplementary Fig. [162]15). Furthermore,
the results of the qPCR assay revealed the upregulation of epithelial
function-associated gene upregulations (CK14, ZO-1, and E-Cadherin) and
the downregulation of epithelial malfunction and apoptosis-associated
genes including α-SMA and Caspase 9 in the collected epithelial basal
cells (Fig. [163]3g–k).
Fig. 3. LA@PA-TA@C hydrogel rescued epithelial regenerative ability and
promoted mucociliary differentiation of epithelial organoids in vitro.
[164]Fig. 3
[165]Open in a new tab
a The rabbit primary tracheal epithelial cells were co-cultured with
different hydrogels (PA, PA-C, LA@PA-C, LA@PA-TA@C), 200 μM H[2]O[2]
was added to simulate the hyperactivated innate immune environment.
Elements were created in BioRender. Yi, C. (2025)
[166]https://BioRender.com/m4mpwtc. b ZO-1 (white) and cell nuclei
(DAPI, blue) IF staining of epithelial cells for tight junction marker
expression after 3 days culture. Scale bar, 20 μm. Pink arrows
represent empty areas after cell apoptosis. c quantitative analysis of
the ZO-1 expression in b n = 4 samples per group, and the results were
representative of 2 independent experiments. d–f Rabbit primary
epithelial cells scratching assay under oxidative stress after 48 h. d
IF staining of ZO-1 (marker of tight junctions, red), Ki67 (marker of
proliferation, green), and cell nuclei (DAPI, blue). The circles
represent empty areas after cell apoptosis. scale bar, 200 μm. e
statistical analyses of the residual wound area in d. f quantitative
analysis of the Ki67 expression in D. n = 4 samples per group in b, f,
and the results were representative of 2 independent experiments. g–k
qPCR revealed gene expression in primary epithelial cells after
co-cultured with hydrogels under oxidative stress for 3days. g CK14,
stem-related gene. h E-Cadherin, and i ZO-1 were tight
junctions-related genes. j α-SMA, epithelial dedifferentiation-related
gene. k Caspase 9, apoptosis-related gene. Group setting: Blank (no
hydrogel and no H[2]O[2]), +200 μM H[2]O[2] group (no hydrogel and add
200 μM H[2]O[2]). PA, PA-C, LA@PA-C, and LA@PA-TA@C represented groups
co-cultured with corresponding hydrogels under 200 μM H[2]O[2]
environment[.] n = 6 samples per group in g–k from two individual
repeated experiments. l–o LA@PA-TA@C hydrogel rescued regenerative
ability and promoted mucociliary differentiation of epithelial
organoids. l the epithelial organoids culture model. The organoids were
induced by rabbit primary tracheal epithelial cells on air-liquid
interface (ALI) and were co-cultured with different hydrogels under
oxidative stress for 21 days. Elements were created in BioRender. Yi,
C. (2025) [167]https://BioRender.com/m4mpwtc. m SEM images of organoids
in different groups at 21 days. n staining for exhibiting organoid
differentiation co-culture with PA hydrogel under oxidative stress for
21 days; o, co-culture with LA@PA-TA@C hydrogel under oxidative stress
for 21 days. n, o gross morphology (HE staining), tight junction, TJ
(ZO-1, green), mucociliogenic (PAS, purple; AC-Tub, green), and
epithelial stemness (CK-14, red). Cell nuclei (DAPI, blue). Scale bar
in (l–o), 20 μm; and 10 μm in o magnified regions. n = 3 samples per
group in l–o, where each sample was from an individual experiments.
Source data are provided as a Source Data file. Data are presented as
mean ± SD. All error bars represent SD. p values calculated using ANOVA
followed by Tukey’s post hoc test.
To assess the epithelial regenerative ability in an oxidative stress
environment, an epithelial organoid model at the air-liquid interface
(ALI) was established (Fig. [168]3l, Supplementary Fig. [169]13,
[170]16). The epithelial cells exhibited a malfunctional phenotype of
squamous metaplasia, characterized by disrupted tight junction and cell
apoptosis (Fig. [171]3m, n). By contrast, intact epithelial barriers
and well-differentiated cilia were observed, wherein native-like
pseudostratified columnar epithelial structures with ZO-1^+ tight
junction, AC-Tub^+ cilia cells, PAS^+ goblet cells, and CK14^+ basal
cells were obtained (Fig. [172]3o). These results demonstrated
LA@PA-TA@C hydrogels activate and enhance the multidirectional
regenerative ability of CK14^+ basal cells.
Respiratory function rescue in an acute epithelial injury model
To simulate real-world clinical tracheal injury
scenarios^[173]9,[174]15,[175]31–[176]33, a rabbit tracheal epithelial
scraping injury model was established (Fig. [177]4a, b). The epithelium
and most of the submucosa at the injury site were scraped away on Day 0
post operation. On Day 7, the entire trachea exhibited hyperactivated
innate immunity, extensive epithelial malfunction, and stenosis
(Fig. [178]4c), which closely resembled the pathological changes of
human epithelial malfunction.
Fig. 4. Utilizing LA@PA-TA@C hydrogel to rescue respiratory function and
investigating biological mechanism in rabbit model of acute epithelial
injury.
[179]Fig. 4
[180]Open in a new tab
a diagram of epithelial acute injury modeling and hydrogel treatment. b
the timeline of observation and processing. c validated the
establishment of the model to simulate human tracheal epithelial
malfunction. Evaluate epithelial defects and squamous malfunction (HE),
oxidative stress-mediated innate immune hyperactivation (ROS/iNOS for
oxidative stress, CD86/CD206 for macrophages in innate immune, Tunel
for apoptosis). Cell nuclei (DAPI, blue). Scale bar, 1 mm; and 100 μm
in magnified regions. n = 3. d Survival curves of rabbits treated by
different hydrogel groups (PA, PA-C, LA@PA-C, LA@PA-TA@C), n = 6. e
evaluated the tracheal patency by bronchoscopy and full-length gross
view on Day 28 post-operation. f, g respiratory frequency monitoring. f
respiratory rate changes during 28 days post-operation, pink line
showed the respiratory rate of healthy rabbits. n = 6 rabbits per
group. Each rabbit has 3 points per day on figure represented collected
respiratory rate at three time points (morning, noon, and evening)
until death or sample collection. g respiratory rate on Day 28
post-operation. Each rabbit was recorded breathing 5 times a day, n = 4
per group, and each point represents one record. h–l tracheal samples
were harvested on Day 14, and RNA sequencing was used to investigate
the biological mechanism (n = 3). h Pearson’s correlation coefficient
analysis between samples: green, LA@PA-TA@C group; yellow, blank group
(untreated). i Volcano plot analysis of differentially expressed genes
(DEGs) between the blank control and LA@PA-TA@C treatment. Fold change
≥ 2 and p value ≤ 0.05. p values calculated using two-tailed unpaired
t-test. j Heatmap of DEGs. The genes are clustered into four groups by
unsupervised clustering: cluster 1 and cluster 4 represent genes that
are downregulated in the LA@PA-TA@C group, while cluster 2 and cluster
3 represent genes that are upregulated in the LA@PA-TA@C group. k
Biological process enrichment analysis via unsupervised clustering of
DEGs. Biological process enriched in gene cluster 1 and gene cluster 4
were downregulated, while biological process enriched in gene cluster 2
and gene cluster 3 were upregulated. l KEGG pathway enrichment analysis
showing potential pathways in cluster 1 (purple, upregulated) and
cluster 2 (pink, downregulated). Source data are provided as a Source
Data file. Data are presented as mean ± SD. Error bars or shaded areas
represent SD. p values calculated using one-way ANOVA test.
Subsequently, corresponding treatments were initiated for the different
groups. Rabbit death occured in all of the groups due to severe
respiratory shortness symptoms except LA@PA-TA@C group (Fig. [181]4d).
On day 28, bronchoscope examination shows severe stenosis and adhered
sputum in the PA group (Fig. [182]4e, Supplementary Fig. [183]17a). In
contrast, LA@PA-TA@C treatment resulted in good tracheal patency,
native-like smooth breathing, and no sputum retention, which
consequently enabling complete trachea repair including structure and
function to effectively rescue respiratory function (Fig. [184]4f, g,
Supplementary Fig. [185]17b, Supplementary Movie [186]3). The high
biocompatibility and biosafety paves a solid foundation to clinical
translation (Supplementary Fig. [187]18, [188]19).
LA@PA-TA@C induced trachea repair mechanism analysis
Whole transcriptome RNA sequencing (RNA-Seq) performed on Day 14
demonstrated the good agreement within the group (Fig. [189]4h), where
a significant difference in the transcriptome profiles between
LA@PA-TA@C group and blank control group (model only with no treatment)
was observed (Fig. [190]4i). The enrichment analysis identified both
upregulated and downregulated gene sets in LA@PA-TA@C group
(Supplementary Fig. [191]20, [192]21), and all of these differential
expression genes (DEG) were classified into four distinct clusters via
machine learning-based unsupervised cluster analysis (Fig. [193]4j, k).
Compared to the blank group, cluster 1 (purple) and cluster 4 (green)
genes were downregulated genes in the LA@PA-TA@C group, while cluster 2
(pink) and cluster 3 (blue) genes were upregulated genes in the
LA@PA-TA@C group. In cluster 2, the signaling pathways associated with
the oxidative-reduction process and glutathione metabolism were found
to be upregulated by LA@PA-TA@C. In cluster 4, there was an enrichment
in biological processes that LA@PA-TA@C downregulated inflammatory
responses. In cluster 3, crucial biological processes associated with
the regeneration of the pseudostratified columnar ciliated epithelium
were found to be upregulated by LA@PA-TA@C. These findings clearly
indicated that LA@PA-TA@C hydrogels effectively fulfil their design
goals by inhibiting the neutrophil-initiated oxidative stress-innate
immune cascade axis, and improving epithelial regenerative ability,
including that of cilium. (Fig. [194]4k). Furthermore, an additional
cluster (cluster 1, a cluster downregulated in the LA@PA-TA@C group) of
genes that was intricately involved in the biological process of
collagen fibril organization closely correlates with tracheal stenosis
complication (Fig. [195]4k, l), such as cell adhesion molecules,
ECM-receptor interaction, focal adhesion, and PI3K-Akt, etc.. They
indicated that LA@PA-TA@C hydrogels can inhibit ECM deposition to
impede stenosis.
Further validate the mechanism involved in reprogramming the
neutrophil-initiated oxidative stress-immune cascade axis using such
multiplexed self-adaptable hydrogels. Many MPO^+ROS^+ neutrophils were
observed to still be activated at the site of injury in the control
group, but were extinguished in the LA@PA-TA@C group (Supplementary
Fig. [196]22a–c). The presence of M1 macrophages (CD86^+CD206^-) and
Tunel^+ apoptosis cells within 1 - 4 weeks indicated persistent
hyperactivation of the innate immune in the control groups
(Supplementary Fig. [197]22d–h, Supplementary Fig. [198]23). By
contrast, the activation of neutrophils and ROS in the LA@PA-TA@C group
were significantly dampened within 1 week (Supplementary
Fig. [199]22a), and the numbers of M1 macrophages and apoptotic cells
considerably decline after 4 weeks (Supplementary Fig. [200]23).
Moreover, the up-regulation of pro-regenerative CD206^+ M2 macrophages
at 1 week and 2 weeks post-operation confirmed that the
neutrophil-initiated oxidative stress-immune cascade was interrupted,
thus reshaping into a pro-regenerative environment (Supplementary
Fig. [201]22d–h).
In vivo epithelial regenerative ability reinvigoration
In addition to reprogramming the innate immune cascade axis, another
pathogenesis hampering epithelial regeneration, i.e., epithelial
regenerative ability or stemness impairment, was reinvigorated with
such Janus hydrogels. To verify it, 42 DEGs were firstly identified
after intersecting all the DEGs with the epithelial regeneration and
tracheal epithelium-related gene set in public databases
(Fig. [202]5a). Compare to the blank control, the DEGs associated with
epithelial stemness (including KRT5 and TP63), epithelial barrier
function (including CDH23 and SHH), and ciliated cell differentiation
(ULK4) were significantly increased after LA@PA-TA@C hydrogel treatment
at both the transcriptome and translational protein levels
(Fig. [203]5b, Supplementary Fig. [204]24, [205]25). Further gene set
enrichment analysis (GSEA) revealed that four pivotal signaling
pathways associated with cilial cell differentiation and Notch model
pathways were upregulated in LA@PA-TA@C hydrogel group (Fig. [206]5c),
which thus undertook to regulate tracheal epithelial regeneration and
differentiation.
Fig. 5. LA@PA-TA@C hydrogel rescued epithelial regenerative ability and
regulated epithelial regeneration patterns in vivo.
[207]Fig. 5
[208]Open in a new tab
a Venn diagram of tracheal epithelial regeneration-associated DEGs.
DEGs were firstly identified after intersecting all the DEGs with the
epithelial regeneration and tracheal epithelium-related gene set in
public databases. b heatmap of the DEGs in a. c GSEA enrichment plot
showing the top-ranked epithelial regeneration-associated pathways were
upregulated in LA@PA-TA@C group. d the columnar mucociliary structure
of native tracheal epithelium. Cilia cells (AC-Tub, blue arrow), goblet
cells (MUC5ac, PAS staining, green arrow), basal cells (CK14). Cell
nuclei (DAPI, blue). e rabbit trachea samples were harvested and
stained at indicated timepoints. Epithelial regeneration patterns in
rabbit epithelial model were detected after treated by different
hydrogel groups on the 1-, 2-, and 4-weeks post-operation. f
quantitative statistics of regenerated epithelial thickness in e. g–p
explore the patterns of epithelium proliferation and mucociliary
differentiation. g the spatial distribution of proliferating cells
(PCNA, marker of proliferation, organ; Tunel, marker of apoptosis,
green; DAPI, cell nuclei, blue). h quantitative statistics of PCNA
expression. i quantitative statistics of Tunel expression 4-weeks
post-operation. j basal cell migration contributed to epithelial
barrier reconstructing (CK14, marker of epithelial stemness, red; ZO-1,
marker of epithelial barrier function, green; DAPI, cell nuclei, blue).
k quantitative statistics of ZO-1 expression. l quantitative statistics
of CK14 expression. m functional cilia regeneration. Gross morphology
(HE staining), ciliogenic (AC-Tub, green), and epithelial plasticity
(CK-14, red). n quantitative statistics of cilia-related AC-Tub
expression. o functional regeneration of goblet cells. Gross morphology
(HE staining), mucogenic (PAS, purple; MUC5ac, green), and epithelial
stemness (CK-14, red). Cell nuclei (DAPI, blue). p quantitative
statistics of MUC5ac expression. The dashed line represents the
basement membrane. n = 4. Scale bar, 50 μm. Source data are provided as
a Source Data file. Data are presented as mean ± SD. All error bars
represent SD. p values calculated using one-way ANOVA followed by
Tukey’s post hoc test.
The native healthy epithelium is characterized by functional cilial
cells and goblet cells, and renewal-obligated basal cells
(Fig. [209]5d). The LA@PA-TA@C hydrogel treatment resulted in the
complete epithelial coverage within 1 week, featured a pseudolaminar
structure. After 4 weeks, the structure, cellular composition, and
thickness of regenerated epithelium resembled those of the native
epithelium (Fig. [210]5e, f, Supplementary Fig. [211]26). Meanwhile,
the control group failed to achieve epithelial coverage at the early
stage, and eventually escalated into epithelial malfunction
(Fig. [212]5e, f). In the LA@PA-TA@C group, residual PCNA^+ CK14^+
epithelial basal cells were activated, proliferated, and migrated
towards the injured site (Fig. [213]5g–i), thereby indicating that
stratified epithelial cells accumulated and communicated with ZO-1^+
expressing tight junctions to restore the epithelial barrier function
within 1 week (Fig. [214]5j–l). Additionally, CK14^+ epithelial cells
were observed to continuously differentiate into native-like AC-Tub^+
cilia cells (Fig. [215]5m, n) and MUC5ac^+PAS^+ goblet cells
(Fig. [216]5o, p), and only a small proportion of proliferatively
active CK-14^+ basal cells were observed to reside above the basement
membrane for self-renewal, forming native-like epithelial regenerative
patterns. Notably, the epithelial regeneration rate was consistent with
LA@PA-TA@C hydrogel degradation (Fig. [217]2i), thus indicating that
LA@PA-TA@C could perform epithelial barrier functions without
respiratory disturbance before the occurrence of epithelial rescue.
Stenosis complications
Tracheal stenosis is the main fatal complication that frequently occurs
after epithelial malfunction, and is characterized by α-SMA^+
fibroblast proliferation and extracellular matrix (ECM)
deposition^[218]50–[219]52. After rescuing epithelial malfunction, the
ability of the multiplexed hydrogels to eliminate epithelial
malfunction-caused tracheal stenosis was further assessed
(Supplementary Fig. [220]27). The LA@PA-TA@C group demonstrated less
ECM deposition (Supplementary Fig. [221]27a–c, Supplementary
Fig. [222]28a–c), less α-SMA^+ fibroblast proliferation (Supplementary
Fig. [223]27d, e, Supplementary Fig. [224]28d–f), indicating with no
stenosis complications via IF and Masson trichrome staining. Sirius red
staining revealed that the collagen distribution direction curve
exhibited a native-like bimodal pattern, indicating the ECM direction
in the LA@PA-TA@C group was more regular and orderly, approaching to
the collagen I/III ratio in the normal submucosa layer (Supplementary
Fig. [225]27f–h). The other groups displayed typical remodeled fibrous
scars or partially-bulging granulation with abundant isotropic and
horizontally-oriented collagen bundles (Supplementary Fig. [226]27g),
wherein α-SMA^+ fibroblasts uncontrollably overgrew and finally led to
stenosis (Supplementary Fig. [227]27d). RNA-seq analysis also indicated
that ECM deposition related biological process was down-regulated in
the LA@PA-TA@C group (Fig. [228]4j–l), and the key downregulated genes
included COL1, VWF, THBS3, SPP1, IGF, and SOX9 (Supplementary
Fig. [229]27i). Furthermore, the RNA-seq results were validated at the
translational protein level by IF staining, and the results revealed
that the expression of these genes in the submucosal region was
inhibited at the protein level, which aligns with the results at the
transcriptome level (Supplementary Fig. [230]27j, k). These results
indicated that the LA@PA-TA@C hydrogels downregulated ECM deposition at
the injury site and thereby significantly reduced the incidence of
adverse tracheal stenosis complications.
LA@PA-TA@C rescued epithelial functions and prevented stenosis in a rabbit
model of recurred circumferential epithelial injury
Tracheal injury with epithelial malfunction tends to result in
nonhealing and recurrent stenosis, whereas in clinical practice,
re-patency treatment via electrocautery, biopsy forceps scraping, and
stent implantation can only temporarily alleviate the symptoms of
stenosis. To better mimic clinical scenarios, we constructed a recurred
circumferential tracheal epithelial injury model (Supplementary
Fig. [231]29a–c). On Day 7 post-circumferential injury, the rabbits
were divided into different groups: one group underwent re-patency and
received LA@PA-TA@C hydrogel treatment (LA@PA-TA@C group), and the
blank control group only underwent re-patency (Blank control)
(Supplementary Fig. [232]29b). Rabbit death occurred in the blank
control group, whereas the survival rate of the LA@PA-TA@C group was
observed to be 100% (Supplementary Fig. [233]29d). Further assessment
of epithelial regeneration at 2 weeks after re-patency (Supplementary
Fig. [234]29e–h) and tissue section staining revealed that LA@PA-TA@C
treatment regenerated the epithelium with a native-like pseudolaminar
structure (Supplementary Fig. [235]29e), which featured PAS^+ goblet
cells (Supplementary Fig. [236]29f) and AC-Tub^+ cilia cells
(Supplementary Fig. [237]29g). The number of CK14^+ epithelial basal
cells was increased in the LA@PA-TA@C group, thereby indicating
positive tissue regeneration (Supplementary Fig. [238]29e, h). These
phenotypes aligned with the acute partial epithelial injury model
(Fig. [239]5), thus indicating that LA@PA-TA@C could rescue epithelial
functions in various scenarios.
We further evaluated recurrent stenosis complications in this model at
2 weeks post re-patency (Supplementary Fig. [240]29i, k). As previously
mentioned (Supplementary Fig. [241]27), tracheal stenosis is a fatal
complication of epithelial malfunction and is characterized by α-SMA^+
fibroblast proliferation and ECM deposition. Tissue section staining
revealed that the LA@PA-TA@C group demonstrated no stenosis
complications with native-like submucosal thickness (Supplementary
Fig. [242]29i, k). Moreover, the expression of α-SMA (Supplementary
Fig. [243]29i) and the collagenous area (Supplementary Fig. [244]29j)
were downregulated in the LA@PA-TA@C group, thereby indicating that the
hydrogel could prevent tracheal stenosis even in recurred injured
epithelial clinical scenarios.
Bioengineered tracheal transplantation for complete tracheal repair
After rescuing epithelial malfunction, we integrated the aforementioned
multiplexed self-adaptable Janus hydrogels with a standardized vascular
and cartilage repair protocol, and investigated whether the integrated
transplants could address the long-standing challenge (i.e., epithelial
malfunction) of complete tracheal repair in the thoracic field. In this
standardized route, engineered cartilage (TEC) was constructed and
embedded directly into the anterior cervical muscle to promote
cartilage maturation and routine prevascularization for 28 days
(Supplementary Fig. [245]30). Afterwards, the integrated trachea
transplants (TEC/LA@PA-TA@C) were obtained through allowing the
LA@PA-TA@C hydrogels to adhere to the vascularized TEC (Fig. [246]6a,
Supplementary Fig. [247]30, Supplementary Movie [248]4 and
Supplementary Movie [249]5). At 28 days post-tracheal transplantation,
the integrated transplants exhibited a survival rate above 83%, while
the survival rate was less than 18% in the control group (TEC group)
(Fig. [250]6b). Simultaneously, the respiratory status of the rabbits
in TEC/LA@PA-TA@C group was demonstrated to be stable, and the
respiratory function of these rabbits progressively approached that of
native rabbits (Fig. [251]6c). Notably, TEC/LA@PA-TA@C revealed
tracheal patency without stenosis, and exhibits de novo structures
resembling native cartilage at the transplantation site, while the
control group shows stenosis with no cartilage degradation
(Fig. [252]6d–f, Supplementary Fig. [253]31, Supplementary
Movie [254]6). No neutrophil-initiated oxidative stress-immune cascade
was observed in the TEC/LA@PA-TA@C group, as indicated by no neutrophil
activation, lower ROS production, and less M1 macrophages
(Supplementary Fig. [255]32). The regeneration of the tracheal
pseudostratified columnar epithelium was observed (Fig. [256]6g–l,
Supplementary Fig. [257]33), wherein the epithelium coverage ratio
exceeded 98% (Fig. [258]6j) with over 85% mature cilia coverage
(Fig. [259]6i, l). Due to the protective effect of TEC/LA@PA-TA@C and
regenerated epithelial tissues, both lacunae and stroma deposits
resembled to natural cartilage in tracheal grafts (Fig. [260]6m–o,
Supplementary Fig. [261]34). In the TEC group, the overactivation of
the innate immune system directly impaired epithelial regeneration and
caused tracheal stenosis; in particular, the engineered TEC cartilage
was destroyed, which further led to fatal tracheal collapse.
Fig. 6. Multiplexed self-adaptable LA@PA-TA@C hydrogel promoted completed
tracheal repair.
[262]Fig. 6
[263]Open in a new tab
a flow diagram of rabbit bioengineered tracheal reconstruction via TEC
and LA@PA-TA@C hydrogel on Day 28. b survival curves of rabbits treated
by groups (TEC, TEC/LA@PA-TA@C), n = 6. c respiratory frequency
monitoring during 28 days post-reconstruction, area between pink lines
shows the respiratory rate of healthy rabbits. d patency evaluation of
different bioengineered tracheal grafts (TEC, TEC/LA@PA-TA@C) via CT
scan and bronchoscope on Day 56 (28 days post-reconstruction). The
arrow indicated the grafts. e gross view, lumen exhibition, and
clinical grading. Scale bar in (d, e), 2 mm. f quantitative analysis of
bioengineered tracheal patency. g–l assessment of epithelial functions
in the reconstructed bioengineered trachea area, general view (HE),
cilia regeneration (AC-Tub, green), epithelial stemness (CK-14, red),
and cell nuclei (DAPI, blue). Scale bar was indicated in figures. g
reconstructed trachea via only TEC (area in dashed box) on Day 56. h
reconstructed bioengineered trachea via TEC/LA@PA-TA@C (area in dashed
box) on Day 56. i SEM images, the blue area represented cilia cells,
the yellow area represented goblet cells, the red area represented
undifferentiated basal cells, and the gray area represented fibroplasia
without epithelium. j quantitation of epithelium coverage rate in g, h.
k quantitation of epithelial thickness in g, h. l quantitation of cilia
cells coverage rate in g, h. m–o assessment of tracheal cartilage
regeneration. Alcian blue staining exhibited cartilage matrix deposit.
m reconstructed via TEC. n reconstructed bioengineered trachea via
TEC/LA@PA-TA@C. Scale bar, 1 mm. o Col II immunohistochemical (IHC)
staining and 3D surface plot analysis. The yellow peaks in 3D surface
plot represented the Col II area and intensity in IHC images, while the
white area indicated no Col II deposition. Scale bar, 100 μm. n = 4 in
TEC group, n = 5 in TEC/LA@PA-TA@C group, and n = 3 in native trachea
control group. Source data are provided as a Source Data file. Data are
presented as mean ± SD. All error bars represent SD. p values
calculated by one-way ANOVA with Tukey’s post hoc test for multiple
comparisons (≥3 groups) or two-tailed unpaired t-test for two-group
comparisons.
The efficacy of LA@PA-TA@C was further investigated in a long
circumferential tracheal defect model to confirm its clinical
translational capabilities. Chondrocytes were encapsulated in a
methacrylated gelatine (GelMA) hydrogel and moulded into 6 mm internal
diameter rings. These rings were stacked into a tubular construct and
subsequently implanted into the anterior cervical muscle to promote
cartilage maturation and routine pre-vascularization for 28 days,
thereby ultimately forming a tissue-engineered trachea (TET). The
LA@PA-TA@C hydrogel was subsequently adhered to the TET luminal surface
to create TET/LA@PA-TA@C grafts, which were then orthotopically
transplanted to construct a long circumferential tracheal defect model
(Fig. [264]7a). The survival rate of the TET/LA@PA-TA@C grafts was
100%, whereas the survival rate was 33.3% in the TET group
(Fig. [265]7b). A comprehensive evaluation of epithelial and cartilage
regeneration was performed via histological and IF staining analyses
(Fig. [266]7, Supplementary Fig. [267]35). The TET/LA@PA-TA@C grafts
demonstrated successful reconstruction of the tracheal wall
architecture, which featured the following characteristics. 1) Complete
re-epithelialization with a pseudostratified columnar epithelium lining
the graft lumen was demonstrated (Fig. [268]7c, d). Via IF
quantification, functional epithelial restoration was evidenced by
increased CK14^+ basal cells and AC-Tub^+ ciliate cells (Fig. [269]7e,
f). 2) Cartilage integrity preservation revealed that the chondrocyte
lacunae were embedded in sulfated Col II-rich extracellular matrix,
thereby phenocopying the native cartilage morphology (Fig. [270]7e, g).
While in the TET group, no epithelial regeneration was observed, let
alone cilia cells, and the cartilage was also severely degraded
(Fig. [271]7c, e–g), indicating a lethal risk. Meanwhile, the liver
function and kidney function were assessed on day 28 post orthotopic
transplantation. Both liver function (AST, ALT) and kidney function
(CREA and UA) were no difference with native rabbits, indicating the
biosafety of LA@PA-TA@C hydrogel (Supplementary Fig. [272]36).
Fig. 7. LA@PA-TA@C hydrogel realized completed tracheal repair in a long
circumferential tracheal defect model.
[273]Fig. 7
[274]Open in a new tab
a flow diagram of rabbit long circumferential tracheal defect tracheal
reconstruction via tissue-engineering trachea (TET) and LA@PA-TA@C
hydrogel on Day 28. NT, native trachea. b survival curves of rabbits
treated by different graft groups (TET, TEC/LA@PA-TA@C). c, d rabbit
trachea samples of different graft groups were harvested and stained to
exhibit the regeneration of tracheal structure. HE staining exhibited
gross morphology, Alcian blue staining exhibited cartilage and
epithelium. The white dashed line represented the epithelial basement
membrane. The dark dashed line represented the cartilage region. Scale
bar, 500 μm. c TET. d TET/LA@PA-TA@C. e Representative magnified IF
staining images of the middle section of the grafts (TET group, and
TET/LA@PA-TA@C) to evaluate tracheal epithelium-cartilage structure.
Epithelium (CK14, red), cartilage (Col II, yellow), and cell nuclei
(DAPI, blue). Scale bar, 500 μm. f assessment and quantitative
statistics of epithelial thickness and cilia regeneration. cilia
regeneration (AC-Tub, green), epithelial plasticity (CK-14, red), and
cell nuclei (DAPI, blue). The white dashed line represented the
epithelial basement membrane. Scale bar, 50 μm. g representative Alcian
blue staining images and quantitative statistics of cartilage lacunae
number. Scale bar, 50 μm. n = 6. Source data are provided as a Source
Data file. Data are presented as mean ± SD. All error bars represent
SD. p values calculated using one-way ANOVA test with Tukey’s post hoc
test.
In brief of above two tracheal transplantation models, such multiplexed
self-adaptable hydrogel represented a pioneering approach to rescue
tracheal epithelial malfunction, which promoted the complete structural
and functional repair of trachea through even using the simplest
routine vascular and cartilage regeneration and re-epithelialization
techniques.
Discussion
Complete tracheal repair with native-like structure and functions
represent Goldbach’s conjecture in the field of thoracic
surgery^[275]2, and tissue engineered tracheal construction is
considered a feasible way^[276]9–[277]11. However, there is still an
unmet requirement for specialized and standardized treatment for
tracheal epithelial rehabilitation after tracheal transplantation in
clinical practice. Tracheal epithelial malfunction is usually
accompanied by severe complications, such as sputum accumulation,
tracheal stenosis, and tracheal collapse^[278]9–[279]11. Clinical
treatment can only address apparent symptoms, but fails to resolve the
pathogenesis of these complications^[280]53–[281]57. Unresolved
epithelial malfunction typically induces these complications to persist
and inflict patients with repeated fatal risks^[282]13,[283]14, thus
imposing a substantial burden on medical resources. The rescue of
epithelial structure and function is a precondition of complete trachea
repair. Current explorations to restore the tracheal epithelium have
relied on the grafts of the epidermis, buccal epithelium, bioengineered
epithelium, or epithelial cells^[284]6,[285]16,[286]17,[287]58–[288]64.
However, the disputed pathogenesis of tracheal epithelial malfunction
and deficiencies of targeted interventions have easily caused these
epithelial grafts evolved into squamous metaplasia or other
malfunctional phenotypes and shed, let alone the cost and ethical
considerations. Therefore, it is necessary to find specific therapies
for epithelial malfunction. In this study, we provided information on
the pathogenesis of epithelial malfunction and constructed the
multiplexed self-adaptable Janus hydrogels to address this issue.
Different from other epithelial interface environments^[289]25,[290]51,
the harsh tracheal environment includes sputum, pollutants, and
microbes, and was also suffered by the mechanical stress caused by
inhaled air. Patients with epithelial injuries or deficiencies are
unable to receive routine wound rehabilitation care and are in a
persistent state of innate immune hyperactivation. In this process, the
bacterial infection and neutrophil- induced oxidative stress to
directly damage epithelial regenerative ability and leads to epithelial
malfunction. The mechanical/physical stress caused by inhaled air could
also incorporate and intensify oxidative stress and immune response.
Overall, bacterial infection, oxidative stress, and innate immune
hyperactivation constituted a cascade reaction, and caused a further
vicious circle. Notably, the multiplexed self-adaptable hydrogels can
target this pathogenesis to rescue epithelial malfunction. First, the
hydrogels adapted to the full-mucus environment and formed a stable and
degradable barrier for epithelial rehabilitation, which isolated
foreign bodies and sterilized most of the bacteria. Second, the
hydrogels adapted to the oxidative stress- innate immune cascade,
blockaded breathing pollutants-caused MPO^+ neutrophil activation, and
scavenge oxidative stress. Third, the hydrogels adapted to epithelial
regenerative patterns, and promoted CK14^+ basal cells to immigrate and
differentiate to a multifunctional epithelial mucociliary elevator. In
this study, the multiplexed self-adaptable hydrogels effectively
addressed three crucial factors in the pathogenesis, including
bacterial infection, neutrophilic oxidative stress, and epithelial
regenerative ability. In both two models, i.e., epithelial acute injury
and recurred injury, the hydrogels successfully prevented stenosis
complications and achieved functional cilia, representing a promising
therapy for tracheal epithelial malfunction. In the epithelial
bioengineered tracheal transplantation model, the hydrogels
successfully achieved structural and functional mucociliary epithelium
regeneration and promoted completed tracheal repair, thus representing
a universal approach for tracheal epithelial rehabilitation after
tracheal transplantation.
The activation of the innate immune system can affect the adaptive
immune system; for example, neutrophils and macrophages may activate
lymphocytes and plasma cells. Thus, we also examined adaptive immunity.
Compared with native conditions, in both the human epithelial
malfunction trachea model and the rabbit epithelial injury malfunction
model, T-cell infiltration and activation were increased, as was B-cell
infiltration (Supplementary Fig. [291]37). Furthermore, LA@PA-TA@C
hydrogel treatment inhibited the activation of adaptive immunity in
rabbit acute epithelial model, which was possibly due to the fact that
the hydrogel blocked innate immune hyperactivation.
In summary, epithelial stemness loss and the bacterial
infection-induced neutrophil-initiated innate immune hyperactivation
have been identified as the pathogenesis of epithelial malfunction.
With targeting them, a multiplexed self-adaptable Janus hydrogel has
been constructed via the hybrid photopolymerization between assembled
TA@C hydrogels and NHS-activated LA@PA hydrogels. The two Janus sides
imparted LA@PA-TA@C hydrogels with high tissue adhesion and favorable
anti-fouling ability, which collaborated with acid-reinforced
stability, anti-swelling and oxidative stress scavenging properties to
adapt harsh tracheal environment (e.g., rich mucous, pollutants,
bacteria etc.) in airway. In both in vitro primary epithelial organoid
model and in vivo epithelial injury model, such multiplexed
self-adaptable hydrogels reprogrammed oxidative stress-innate immune
cascade axis, and promoted multifunctional epithelial mucociliary
regeneration. Consequently, they invigorated epithelial mucociliary
regenerative ability, modulated epithelial regeneration patterns to
accomplish epithelial rehabilitation. Moreover, they inhibited
fibroblast activations and rearranged native-like grid ECM networks,
which persistently addressed tracheal stenosis complications from the
roots. Especially after integrating with a standardized routine
vascular and cartilage repair protocol, a luminal
hydrogel-functionalized bioengineered tracheal graft was obtained to
rescue epithelial malfunctions and promote complete tracheal repair
with native trachea-like structure and functions. Therefore, such
multiplexed self-adaptable Janus hydrogels presented a targeted
approach to various clinical scenarios of epithelial malfunction.
Methods
Ethical statement
This study protocol was approved by the Ethics Committee of Shanghai
Pulmonary Hospital (K23-235), and written informed consent was obtained
from all individual participants recruited in this study. The study
design and implementation adhere to all relevant regulations concerning
the use of human research participants and are conducted in accordance
with the standards set forth in the Declaration of Helsinki.
Patient population and sample collection
Patients with tracheal epithelial malfunction and stenosis complication
were diagnosed and confirmed through bronchoscopy, and biopsy samples
were collected from the pathological sites for histopathological
examination. The healthy control group was composed of tracheal tissue
obtained from healthy lung transplant donors, with exclusion of
diseases such as infection, inflammation, or sarcoidosis. Based on the
observed stenosis, typing and grading were performed according to the
Freitag criteria (Supplementary Table [292]1).
Animals
New Zealand rabbits were supplied by Shanghai Jiagan Experimental
Animal Company (Shanghai, China). Both female and male New Zealand
rabbits were 3 months old, with an average weight of 2 kilograms.
C57BL/6N-Cd177^em1C/Cya mice were get from professor Z.J.L. of the
Shanghai Tenth People’s Hospital. C57BL/6, female and male mice, 8
weeks of age, were purchased were purchased from GemPharmatech in
Jiangsu, China. Mice were maintained in specific pathogen-free
conditions in microisolator cages, and all the rabbits were housed
under conventional (CV) level conditions. Animals were treated by
following the guidelines for the care and use of animals (National
Research Council and Tongji University). All animals were fed ad
libitum. The procedures for the use of animals were approved by the
Ethics Committee of Shanghai Pulmonary Hospital. All applicable
institutional and governmental regulations concerning the ethical use
of animals were followed.
Primary epithelial cell culture and epithelial organoid differentiation
The rabbit trachea was obtained from after euthanasia. Then trachea was
washed three times by PBS and incubated for 30 min at 37 °C with
dispase (Stem Cell, Canada). After incubation, the mucosal layer was
separated from the trachea and further digested overnight. The digested
liquid was collected and filtered through a 70 μm filter
(Sigma-Aldrich, USA), and then cell suspension was centrifuged at 200 g
for 10 min. Cells were resuspended by epithelial culture medium
consisting of DMEM (Gibco, USA) and F-12 (Gibco, USA) with
penicillin-streptomycin (Solarbio, Beijing, China) and 10% FBS (Gibco,
USA) supplemented with 5 μm Y27632 (Macklin, China), 0.125 ng/mL EGF
(PeproTech, China), 5 μg/mL insulin (Sigma, USA), 0.1 nm cholera toxin
(Sigma, USA), 250 ng/mL amphotericin B (Solarbio, Beijing, China), and
10 μg/mL gentamycin (Gibco, USA), seeded in 3 cm culture dish, and
cultured at 37 °C with 5% CO[2] with three changes of medium per week.
Epithelial cells were passaged to the third generation (P3). Then the
cells were seeded on Matrigel-coated transwell (Stem Cell, Canada) to
form the air-liquid interface (ALI) culture model, and induced by
PneumaCult™ culture media (Stem Cell, Canada) for 3 weeks to form
epithelial organoids. In the process of organoid induction, 200 μm
H[2]O[2] was added to the culture medium to simulate oxidative stress.
Neutrophil isolation and culture
Neutrophils were isolated by density gradient centrifugation from the
peripheral blood of rabbits with bronchiectasis (Lymphoprep 1858).
Isolated rabbit neutrophils were used immediately for in vitro
experiments. Neutrophils were cultured at 37 °C with 5% CO[2]. The
cells were seeded in 24-well flat bottom plates with a concentration of
1.5 × 10^6/well.
3T3 cells and RAW 264.7 macrophages culture
3T3 cells (No. SCC-220911) and RAW 264.7 macrophages (No. SCC-211800)
from Servicebio, China. Cells were cultured at 37 °C with 5% CO[2]. The
cells were seeded in 24-well flat bottom plates with a concentration of
1.5 × 10^6/well for testing.
Bacterial strain and culture
The strain used in this study was Pseudomonas aeruginosa (P.a.) and was
get from professor J.X. Department of Respiratory and Critical Care
Medicine, Shanghai Pulmonary Hospital. Unless otherwise stated, all the
bacteria in the represented study were cultivated in 3 ml of sterile
lysogeny broth (LB) medium for 13 h (37 °C). Subsequently, they were
diluted by LB at a ratio of 1:200 and agitated for an additional 2 h.
Synthesis and preparation of LA@PA-TA@C hydrogel
Chitin, tannic acid, aspartic acid, L-arginine, and other chemicals and
solvents used were all purchased from MACKLIN (Shanghai, China).
First, TA@C hydrogel was prepared as previous reported. Chitin powder
was immersed in 36.7 wt% NaOH solution at room temperature overnight,
and subsequently, the suspension was repeated freeze-thaw to promote
chitin dissolution. Then, ddH[2]O was added into suspension, and
freeze-thaw for serval cycles to obtain a 4.31 wt% transparent chitin
solution. Besides, 10 wt% tannic acid solution were prepared. The 9:1
chitin-tannic acid solution was mixed. Tannic acid mediated chitin
self-assembly and form hydrogel in 65 °C for overnight. The hydrogel
was neutralized and washed thoroughly with water.
Then, N-acryloyl aspartic acid was synthesized via aspartic acid.
0.45 g N-acryloyl aspartic acid was dissolved in 0.55 g MES solutions.
To replace the carboxyl group with NHS ester and form the PA precursor,
50 mg each of 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide (EDC) and
N-hydroxy succinimide (NHS) were added, followed by an overnight
reaction. For modification on one side of TA@C hydrogel, a mixture
comprising 1 mg photoinitiator (IRGACURE 2959), 50 mg L-arginine, and
1 g PA precursor was introduced prior to inducing the reaction through
UV irradiation at a wavelength of 365 nm for a duration of 30 min to
yield LA@PA-TA@C hydrogel.
Characterization of LA@PA-TA@C hydrogel
Fourier transform infrared (FTIR) spectra
A FT-IR spectrometer (Nicolet iS 10, Thermo Fisher Scientific) was
employed to measure the FTIR spectra of hydrogels. The hydrogels were
freeze-dried, ground into powder, and pressed into tablet. The measured
wavenumber ranges from 4000 to 500 cm^–1.
^1H NMR spectra
The hydrogel precursor was obtained with 5 mg, and the monomer
structure was analyzed by NMR spectroscopy (AVANCENEO600M, USA).
SEM examination
The hydrogels were freeze dried and then coated with a thin gold layer
by using an auto sputter fine coater (JFC 1600, JEOL, Tokyo, Japan)
before imaging. The morphology of the hydrogels and simples was
assessed by using field emission scanning electron microscopy (SEM,
JSM-7001F, JEOL, Japan).
Element mapping and EDS examination
Element mapping was tested on both side of LA@PA-TA@C hydrogel via a
field-emission Magellan microscopy (ULTIMATELYMAX 40, UK).
Compression tests of hydrogels
A nonmechanical analyzer (Instron-6800, UK) was used to evaluate the
hydrogel mechanical strength via compression testing. 6 samples per
group (hydrogel treated with or without 0.2% lactic acid for 1 h),
which were all 1 cm in diameter were compressed to the maximum
deformation strain at a rate of 2.00 mm/min. The stress-strain curve of
the hydrogels plotted.
Rheology analysis
The rheological properties of different hydrogel were characterized on
the stress-controlled rheometer (HAAKE RheoStress 6000, Thermo
Scientific, USA). LA@PA-TA@C hydrogel diameter was 10 mm, and had been
treated with or without 0.2% lactic acid for 1 h. Under the 1 Hz
oscillatory frequency with a fixed 10 N shear stress, the storage (G′)
and loss (G″) moduli were recorded.
Adhesive performance
The LA@PA-TA@C hydrogels (both sides were endowed adhesive ability via
LA@PA) were adhered to the surfaces of different materials and recorded
by iPhone camera. Adhesive strength were tests. Fresh rat skin was cut
into a 3 cm × 5 mm in size, and two pieces of rat skin were adhered
with hydrogel (5 mm × 5 mm in size). The adhesive stress was tested via
an EZ-LX electronic universal testing machine (Shimadzu, Japan) with a
speed of 10 mm/min.
Swelling ratio test
The swelling ratio were evaluated by immersing the hydrogels with PBS.
The hydrogel samples (10 mm in diameter and 5 mm in height) were
weighed after immersed for different time points. The swelling ratio
was defined by the following formula (1). Further, the hydrogels were
immersed in PBS with 5% trypsin, 0.2% lactic acid, and 200 μM H[2]O[2]
to test anti-swelling property.
[MATH: Swelling
ratio=Weight−Weight0hWeight0h×1
00% :MATH]
1
Anti-bacterial property test
The anti-bacterial property of hydrogel against S. aureus (CMCC26003),
E. coli (CMCC44102), and multiresistant Myco. abscessus (19977) were
tested. The bacteria were got from professor H.Q.C. Department of
Respiratory and Critical Care Medicine, Shanghai Pulmonary Hospital.
The colonies formed on the agar medium were counted, and the
antibacterial rings of hydrogels were recorded.
Release curve
The hydrogel was immersed in the solution and the absorbance of the
solution was measured at 200 nm and 320 nm, representing the release of
LA and TA, respectively.
EPR test
Take 20 μL xanthine solution and 20 μL xanthine oxidase PBS solution,
followed by the addition of 10 μL 200 mM BMPO solution. Subsequently,
add 50 μL buffer or 50 μL hydrogel precursor solution. Finally, collect
samples for testing after a reaction time of 10 min.
DPPH scavenging and ABTS scavenging test
The antioxidant activity of hydrogels was assessed via the scavenging
ability of the 1,1-diphenyl-2-picrylhydrazyl (DPPH) and Diammonium
2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonate) (ABTS) free radical.
A 100 μM DPPH or ABTS solution was prepared. The hydrogels were ground
into pastes and prepared to produce a hydrogel dispersion of 100 mg/mL
with ethanol. Then, a DPPH or ABTS solution and a hydrogel dispersion
were mixed in the dark. After incubation for 1 h, the absorbance of the
mixture was measured via a UV–vis spectrophotometer.
Cell culture and intracellular ROS scavenging
3T3 cells and RAW 264.7 macrophages were cultured via medium consisting
of DMEM (Gibco, USA), penicillin-streptomycin (Solarbio, Beijing,
China), and 10% FBS (Gibco, USA). The cells were co-culture with
different group hydrogel via transwell (Stem Cell, Canada) for 3 days,
and 200μm H[2]O[2] was added to the culture medium to simulate
oxidative stress.
The intracellular ROS levels were examined via ROS assay kit (Beyotime,
China). fluorescence microscope (Olympus IX73; Olympus, Japan) and flow
cytometry were performed to characterize the intra ROS levels by
measuring the fluorescence intensity of DCF (Ex = 488 nm, and
Em = 525 nm).
Scratch test
Primary epithelial cells were planted in a six-well tissue culture
plate and grew for up to 70‒80% confluence in the plate. The cell
monolayer in each well was then scratched with a sterile plastic tip,
and the cells were washed twice with PBS and incubated in epithelial
culture medium co-culture with different group hydrogel via transwell
(Stem Cell, Canada). 200 μm H[2]O[2] was added to the culture medium to
simulate oxidative stress. The cells and scratched wound were imaged at
0, 24, and 48 h, and the area of wound closure was calculated with
ImageJ 1.8 software (National Institutes of Health) for three
independent replicate experiments.
Acute epithelial injury model and hydrogel treatment in rabbit
Firstly, the rabbit epithelial injury model was built and validated.
After anesthesia by Zoletil (50 mg/kg, Virbac, France), the skin and
fascia were sequentially separated and incised until the trachea was
fully exposed in New Zealand rabbits. A nylon brush was inserted to
scrape off the inner epithelium (1 cm × 1.5 cm in size) to simulate a
clinical tracheal epithelium injury scenario. The incision was then
closed, and the fascia and skin were sutured by layer. Trachea tissue
was obtained on the 0 and 7 days after scraping and then pathological
sections and HE staining were performed to validated the successful
establishment of the epithelial injury model.
Hydrogel treatment: using above methods to scrape off the inner
epithelium (1 cm × 1.5 cm in size) and different groups of hydrogels
(1.5 cm × 2 cm in size) were subsequently adhered to cover the injury
site. Then the incision was closed, and the fascia and skin were
sutured by layer. Group setting: only epithelial injury model with no
hydrogel treated (Blank control), PA, PA-C, LA@PA-TA@C, LA@PA-TA@C.
Rabbits with an average body weight of 2.5 kg were randomly divided
into different groups. n = 6 rabbits in each group for one experiment,
and the experiments were repeated independently for three times.
Recurred circumferential tracheal epithelial injury model
After anesthesia by Zoletil (50 mg/kg, Virbac, France), the skin and
fascia were sequentially separated and incised until the trachea was
fully exposed in New Zealand rabbits. A nylon brush was inserted to
scrape off the circumferential inner epithelium (1.5 cm in length) to
simulate a clinical tracheal epithelium injury scenario. The incision
was then closed, and the fascia and skin were sutured by layer. On day
7 post circumferential injury model, the rabbits were anesthesia, and
the skin and fascia were sequentially separated and incised until the
trachea was fully exposed once again. A nylon brush was inserted to
scrape the inner lumen, scraping off the neoplasm at the modeling site
(we called re-patency). Rabbits were then divided into different
groups: one underwent re-patency and received LA@PA-TA@C hydrogel
treatment, the blank control group only underwent re-patency, sham
group, only received sham operation without epithelial injury. n = 6
rabbits in each group for one experiment, and the experiments were
repeated independently for two times.
Bioengineered tracheal graft construction
Rabbit auricular cartilage samples, measuring 4 cm × 6 cm in size, were
harvested under anesthesia induced by Zoletil (50 mg/kg, Virbac,
France). The tissue was sterilized by spraying with 75% ethanol
(Macklin, Shanghai, China) and subsequently washed three times with
phosphate-buffered saline (PBS, Solarbio, Beijing, China) solution
containing 2% penicillin-streptomycin (Solarbio, Beijing, China). On a
clean bench, the attached skin and fascia were carefully peeled off
from the cartilage. The cartilage was then cut into pieces and digested
in 0.25% trypsin (Gibco, USA) at 37 °C for 2 h. Following
neutralization and removal of trypsin, the cartilage pieces were
immersed in 0.15% type II collagenase (Gibco, USA) and incubated at
37 °C overnight. The digested cartilage was filtered through a 70 μm
pore filter (Sigma-Aldrich, USA) to obtain a primary chondrocyte cell
suspension, which was then collected by centrifugation. Primary
autologous chondrocytes (P0) were implanted at a density of 10^6/ 10 cm
diameter culture dish and passaged to the third generation (P3). The
expanded chondrocytes were centrifuged at 200 g for 10 min to pellet
the cells. The cells were resuspended in a 10% solution of
methylacrylyl anhydride gelatin (GelMA, Engineering For Life, Suzhou,
China) and then cast into specific molds. The constructs were
photocrosslinked under ultraviolet light to form stable TEC. The TEC
was further embedded directly into the anterior cervical muscle to
promote cartilage maturation and finish prevascularization for 28 days.
LA@PA-TA@C hydrogel was synthesized following a specific chemical
protocol and then applied to the inner wall of the TEC to construct a
bioengineered TEC/LA@PA-TA@C tracheal graft. n = 6 rabbits in each
group for one experiment, and the experiments were repeated
independently for two times.
The construction method of tissue-engineering trachea (TET) was
basically the same as the above. The expanded P3 chondrocytes were
centrifuged at 200 g for 10 min to pellet the cells. The cells were
resuspended in a 10% solution of methylacrylyl anhydride gelatin
(GelMA, Engineering For Life, Suzhou, China) and then cast into
specific ring-shape molds with 6 mm internal diameter. These rings were
stacked into a tubular construct and then implanted anterior cervical
muscle to promote cartilage maturation and routine prevascularization
for 28 days, ultimately forming TET. n = 3 rabbits in each group for
one experiment, and the experiments were repeated independently for two
times.
Rabbit trachea transplantation
After anesthesia induced by Zoletil (50 mg/kg, Virbac, France), the
skin and fascia in New Zealand rabbits were sequentially separated and
incised until the trachea was fully exposed.
In tracheal partial tracheal wall defect model. The TEC graft embedded
before was then isolated from adjacent tissue carefully with the
vascular pedicle kept at one end. Subsequently, the tracheal
TEC/LA@PA-TA@C graft was prepared, and tracheal graft via only TEC was
used as the control group. An anterolateral trachea was cut off
(10 mm × 10 mm in size) and different groups of bioengineered graft
were transplanted to defect area. The incision was then closed, and the
fascia and skin were sutured by layer.
In long circumferential tracheal defect model. The TET graft embedded
before was then isolated from adjacent tissue carefully with the
vascular pedicle kept at one end. Subsequently, the tracheal
TET/LA@PA-TA@C graft was prepared, and tracheal graft via only TET was
used as the control group. A circumferential trachea was cut off (15 mm
in length) and different groups of bioengineered graft were
transplanted to defect area via end-to-end Anastomosis to simulate
tracheal transplantation clinical scenario. The incision was then
closed, and the fascia and skin were sutured by layer.
Respiratory function monitoring
Respiratory rates of the animals were monitored every 2 days. The
respiratory rate was recorded at three different time points during the
day: morning (8:00 AM), midday (12:00 PM), and evening (6:00 PM), to
account for any diurnal variations in breathing patterns. Prior to each
recording session, animals were allowed to acclimate to the testing
environment for a period of 15–20 min to minimize stress and ensure
they were in a calm state. Recorded the number of breaths taken by the
animal over the 10 min observation period during which animals were in
a quiet, undisturbed state. The total number of breaths recorded during
the three time points was averaged as the respiratory rate for the day.
RNA-sequencing and data analysis
RNA-Sequencing (RNA-Seq) was finished via standard sequencing
procedures (Cloud-seq Company, China) on 14 days post operation in
rabbit epithelial injury model. The blank group referred specifically
to the untreated blank samples (only received epithelial injury model.
In Fig. [293]4, the upregulated genes referred to those that show
significantly higher expression in the LA@PA-TA@C group relative to the
blank group, while the downregulated genes were those with
significantly lower expression in the LA@PA-TA@C group compared to the
blank group. Genes were considered to be expressed in a sample if value
was greater than or equal to that of 1 in the sample. Differentially
expressed genes (DEGs) were defined as fold change ≥ 2 and P
value ≤ 0.05. The raw sequencing data initially underwent quality
assessment utilizing FastQC. Gene ontology analysis was performed by
using DAVID and REVIGO ([294]https://david.ncifcrf.gov;
[295]http://revigo.irb.hr/). Target gene screening was based on
GeneCards dataset ([296]https://www.genecards.org/). n = 3 biologically
independent samples. The visualization of these differentially
expressed genes was achieved through heatmaps generated in R and
Sangerbox. Additionally, GO and pathway (KEGG) analyses were conducted
with the aid of Gene Set Enrichment Analysis (GSEA) software to further
elucidate the biological implications of the gene expression data.
CT scanning
Prior to the CT scanning, animals were fasted for 12 h. After being
anesthetized by Zoletil (50 mg/kg, Virbac, France), the animals were
positioned supine on the CT scanning table with their heads secured in
a custom-made immobilization device to maintain a consistent and
reproducible scanning position.
Bronchoscopy
Rabbits were fasted for 12 h prior to the bronchoscopy examination.
Anesthesia was performed using Zoletil (50 mg/kg, Virbac, France) to
ensure a stable plane of anesthesia throughout the procedure. The
rabbits were positioned in dorsal recumbency with the neck slightly
extended to facilitate access to the trachea. Then a flexible
bronchoscope (diameter and length appropriate for the size of the
rabbit) equipped with a light source and a camera was used for the
examination. The tongue was gently pulled forward to expose the
glottis. The bronchoscope was inserted through the mouth, passed over
the glottis, and advanced into the trachea. The tracheal lumen and
mucosa were inspected for any abnormalities and relative images and
videos were captured using the bronchoscope’s camera for documentation
and further analysis.
ROS production (Flow cytometry method)
Cells were digested and collected (1 × 10^5 per group), resuspended
with buffer (PBS containing 2% FBS), and collected in 1.5 ml EP tubes.
200 μL Fc-Block was firstly added and incubated at 4 °C for 10 min, and
then add buffer to terminate the reaction. Subsequently, 2 μL DCFH-DA
probe were added and incubated at 4 °C for 20 min. Finally, cells were
washed for 2 times with buffer and resuspended. The samples were tested
using a FACSVerse flow cytometer (BD Biosciences, Franklin Lakes, USA).
Compare with the blank control group to distinguish the positive cells.
Quantitative real-time PCR
Total RNA was extracted from cultured cells using the TRIzol reagent
(Takara Bio, Kyoto, Japan). The integrity and purity of the extracted
RNA were assessed using a NanoDrop 2000 spectrophotometer (Thermo
Fisher Scientific). Complementary DNA (cDNA) was synthesized from 1 μg
of total RNA using a commercial reverse transcription kit (Takara Bio).
Quantitative PCR was performed to assess the expression levels of ZO-1,
CK14, E-Cadherin, α-SMA, and Caspase 9. Specific primers were designed
to target these genes and are listed in Supplementary Table [297]3.
Histological and immunohistochemical analysis
After euthanizing the animals, the samples were carefully isolated,
harvested, and fixed in 4% paraformaldehyde for 24 h. The specimens
were then dehydrated with a graded ethanol series and embedded in
paraffin. after that, 8 μm thick sections were obtained. Hematoxylin
and eosin (H&E), alcian blue, sirius red, masson trichromatic,
mitochondrial membrane potential staining (JC-1), and periodic
acid-schiff (PAS) staining were applied using corresponding staining
kit (Solarbio, China). Staining was performed under recommendations of
the kits. Using immunofluorescent staining, the expression of MPO,
PCNA, TUNEL, CD86, CD206, CD80, AC-Tub, MUC5ac, ZO-1, CK14, Ki-67,
α-SMA and SOX9 was detected under a standard IF protocol (Supplementary
Table [298]4). The standard IF protocol is listed as follows: all
samples were washed with PBS for 10 min, permeabilized with 1% Triton
X-100 (Sigma-Aldrich, USA) for 5 min, washed with PBS, and then blocked
with 5% BSA (Sangon Biotech, Shanghai, China) in PBS. The sections were
incubated with primary antibodies overnight at 4 °C. They were then
washed three times with PBS, followed by incubation with the
fluorescent secondary antibodies for 1 h. DAPI was used to visualize
the nuclei. Samples from three independent experiments were examined
with a fluorescence microscope (Olympus IX73; Olympus, Japan). ImageJ
1.8 software (National Institutes of Health) was used to quantify the
number of positive numbers or areas in each view.
Statistical analysis
All the results were exhibited as the mean ± standard deviation (SD)
values. p values calculated by one-way ANOVA with Tukey’s post hoc test
for multiple comparisons (≥3 groups) or two-tailed unpaired t-test for
two-group comparisons. Statistical analysis was conducted with GraphPad
Prism software (8.0), and p < 0.05 was considered to indicate
statistical significance.
Reporting summary
Further information on research design is available in the [299]Nature
Portfolio Reporting Summary linked to this article.
Supplementary information
[300]Supplementary Information^ (20.6MB, pdf)
[301]41467_2025_61135_MOESM2_ESM.docx^ (16KB, docx)
Description of Additional Supplementary Files
[302]Supplementary Movie 1^ (1MB, mp4)
[303]Supplementary Movie 2^ (3.3MB, mp4)
[304]Supplementary Movie 3^ (1.4MB, mp4)
[305]Supplementary Movie 4^ (3.2MB, mp4)
[306]Supplementary Movie 5^ (2.4MB, mp4)
[307]Supplementary Movie 6^ (952.4KB, mp4)
[308]Reporting Summary^ (1.3MB, pdf)
[309]Transparent Peer Review file^ (83.5MB, pdf)
Source data
[310]Source Data^ (5.1MB, xlsx)
Acknowledgements