Content uploaded by Steven Garry Wise
Author content
All content in this area was uploaded by Steven Garry Wise on May 10, 2021
Content may be subject to copyright.
RESEARCH ARTICLE
www.advhealthmat.de
Silk Fibroin Scaffold Architecture Regulates Inflammatory
Responses and Engraftment of Bone Marrow-Mononuclear
Cells
Nianji Yang, Matthew J. Moore, Praveesuda L. Michael, Miguel Santos, Yuen Ting Lam,
Shisan Bao, Martin K. C. Ng, Jelena Rnjak-Kovacina, Richard P. Tan,*
and Steven G. Wise*
Despite being one of the most clinically trialed cell therapies, bone
marrow-mononuclear cell (BM-MNC) infusion has largely failed to fulfill its
clinical promise. Implanting biomimetic scaffolds at sites of injury prior to
BM-MNC infusion is a promising approach to enhance BM-MNC engraftment
and therapeutic function. Here, it is demonstrated that scaffold architecture
can be leveraged to regulate the immune responses that drive BM-MNC
engraftment. Silk scaffolds with thin fibers and low porosity (LP) impairs
immune activation in vitro compared with thicker fiber, high porosity (HP)
scaffolds. Using the authors′established in vivo bioluminescent BM-MNC
tracking model, they showed that BM-MNCs home to and engraft in greater
numbers in HP scaffolds over 14 days. Histological analysis reveals thicker
fibrous capsule formation, with enhanced collagen deposition in HP
compared to LP scaffolds consistent with substantially more native CD68+
macrophages and CD4+T cells, driven by their elevated pro-inflammatory M1
and Th1 phenotypes, respectively. These results suggest that implant
architecture impacts local inflammation that drives differential engraftment
and remodeling behavior of infused BM-MNC. These findings inform the
future design of biomimetic scaffolds that may better enhance the clinical
effectiveness of BM-MNC infusion therapy.
Dr. N. Yang, M. J. Moore, Dr. P. L. Michael, Dr. M. Santos, Dr. Y. T. Lam,
Prof.S.Bao,Dr.R.P.Tan,Prof.S.G.Wise
School of Medical Sciences
Faculty of Health and Medicine
The University of Sydney
Sydney, NSW , Australia
E-mail: richard.tan@sydney.edu.au; steven.wise@sydney.edu.au
Dr. N. Yang, M. J. Moore, Dr. P. L. Michael, Dr. M. Santos, Dr. Y. T. Lam,
Prof.S.Bao,Dr.R.P.Tan,Prof.S.G.Wise
Charles Perkins Centre
The University of Sydney
Sydney NSW , Australia
The ORCID identification number(s) for the author(s) of this article
can be found under https://doi.org/./adhm.
DOI: 10.1002/adhm.202100615
1. Introduction
Bone marrow-mononuclear cell (BM-MNC)
infusion is one of the most widely used
cell therapies in clinical trials over the last
two decades. BM-MNCs capacity to regen-
erate and repair a number of damaged
tissues has led to their study in a wide
range of intractable diseases and condi-
tions including myocardial infarction,[1–4]
limb ischemia,[5,6 ] stroke,[7,8 ] and spinal
cord injury.[9] However, poor cell sur-
vival and engraftment after homing to
sites of injury has resulted in disap-
pointing clinical outcomes.[10,11 ] BM-MNC
fate and behavior is closely guided by
biophysical cues present within the lo-
cal tissue microenvironment.[12–14 ] Within
necrotic or injured tissue these cues are
often degraded or absent, impairing BM-
MNC engraftment and limiting therapeu-
tic potential.[15] Replacing lost biophysical
cues by implanting biomaterial scaffolds
at sites of injury, prior to infusion, is one
promising strategy to enhance BM-MNC
engraftment and function. The optimal design of these scaffolds
is reliant on better understanding of the in vivo responses of BM-
MNCs toward implanted biomaterials.
Prof.M.K.C.Ng
Sydney Medical School
The University of Sydney
Sydney NSW , Australia
Prof.M.K.C.Ng
Department of Cardiology
Royal Prince Alfred Hospital
Sydney NSW , Australia
Dr.J.Rnjak-Kovacina
Graduate School of Biomedical Engineering
University of New South Wales
Sydney NSW , Australia
Prof. S. G. Wise
The University of Sydney Nano Institute (Sydney Nano)
The University of Sydney
Sydney NSW , Australia
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (1 of 14)
www.advancedsciencenews.com www.advhealthmat.de
In our previous work, we developed a novel mouse model us-
ing bioluminescent imaging to track the temporal and spatial
distributions of BM-MNC infusions non-invasively and in real-
time.[16] Using this model, we observed that BM-MNC preferen-
tially homed to sites of injury and inflammation due to a large
proportion of their cell fraction being comprised of immune
cell phenotypes. Furthermore, the model allowed for the high-
throughput evaluation of BM-MNC behavior toward multiple
implanted scaffolds simultaneously. Using electrospinning, we
studied the effects of scaffold material composition on BM-MNC
engraftment. Scaffolds fabricated from polycaprolactone (PCL)
demonstrated enhanced BM-MNC homing and engraftment fol-
lowing blending with collagen (PCL/collagen). These studies re-
vealed that inflammation was a key driver enhancing the hom-
ing of BM-MNCs and that in combination with implanted scaf-
folds comprising natural materials, BM-MNC engraftment could
be increased.
Naturally derived materials typically support better cell at-
tachment and proliferation than synthetic polymers of simi-
lar architecture. As a natural biomaterial, silk fibroin derived
from Bombyx mori cocoons is known for its tunable mechan-
ical and physical properties, making it widely used in numer-
ous tissue engineering applications.[17–20 ] In previous work, we
showed that altered manufacturing of electrospun silk fibroin
produces implantable scaffolds with divergent fiber thicknesses
and porosities.[21] In the context of small diameter vascular
grafts, silk constructs with HP more rapidly polarized local
macrophages into an anti-inflammatory phenotype, leading to
faster graft re-endothelialization and improved long-term remod-
eling. These findings suggest that in addition to the biological
composition of the scaffold, biophysical cues inherent to the ma-
terial architecture could be leveraged to regulate local immune
responses and direct BM-MNC homing and engraftment follow-
ing exogenous infusion.
In this study, we examined the differential immune responses
arising from electrospun silk fibroin scaffolds with distinct fiber
thicknesses and porosities, in the context of BM-MNC infusion
and subsequent engraftment. Comparing high porosity (HP) and
low porosity (LP) formulations of electrospun silk, we first eval-
uated in vitro immune responses by seeding scaffolds with ei-
ther RAW 264.7 murine macrophages or Jurkat human T lym-
phocyte cells, as representative cellular contributions from the
innate and adaptive immune system, respectively. When chal-
lenged with pro-inflammatory stimuli, both cell types showed
suppressed immune activation when seeded on LP scaffolds
compared to HP scaffolds. Using our bioluminescent BM-MNC
tracking mouse model, we further showed that HP scaffolds en-
hanced BM-MNC recruitment and engraftment compared to LP
scaffolds. Immunostaining revealed this to be consistent with the
presence of substantially more CD68+macrophages and CD4+T
cells in HP scaffolds compared to LP scaffolds. Further exami-
nation of immune cell phenotype showed that HP scaffolds in-
creased the proportion of pro-inflammatory (M1) macrophages
and Type I helper T (Th1) cells, while LP scaffolds elevated anti-
inflammatory (M2) and regulatory T (Treg) cells. These immune
regulated BM-MNC engraftment responses led to strikingly dif-
ferent remodeling effects, including increased cell infiltration
and collagen deposition in HP scaffolds. This is the first study
to comprehensively evaluate and reveal differential host immune
responses toward distinct electrospun silk architectures. Impor-
tantly, this study highlights silk scaffold architecture as a crucial
consideration in regulating the engraftment of BM-MNC infu-
sions and justifies the investigation of additional scaffold features
that could further enhance and/or direct BM-MNC therapeutic
function.
2. Results
2.1. Scaffold Characterization
Fiber diameter of electrospun silk scaffolds was characterized us-
ing scanning electron microscopy (SEM, Figure 1A). Quantifica-
tion of fiber thickness revealed HP scaffolds were composed of
larger fibers than LP scaffolds (2.15 ±0.35 vs 0.30 ±0.06 µm,
p<0.0001, Figure 1B). Porosity analysis showed that HP scaf-
folds were significantly more porous compared to LP scaffolds
(45.07 ±1.159% vs 23.30 ±1.166%, p<0.0001, Figure 1C).
Tensile testing was conducted to evaluate the mechanical prop-
erties of each scaffold. Stress–strain curves indicated HP scaf-
folds are more ductile compared to LP scaffolds (Figure 2A).
HP scaffolds were also stiffer and stronger compared to LP
scaffolds, indicated by a relatively higher Young’s modulus
(7.5 ±0.4 vs 5.6 ±0.3 MPa, respectively, p<0.0001, Figure 2B)
and ultimate tensile strength (UTS, 1.5 ±0.05 vs 1.2 ±0.2 MPa,
respectively, p<0.05, Figure 2C), statistically different, but within
the same order of magnitude.
This was consistent with Fourier transformed infra-red spec-
troscopy surface analysis conducted to determine differences
in the silk chemical structure of both scaffolds. Specifically,
the degree of beta-sheets was evaluated through Fourier self-
deconvolution (FSD) of the amide I infrared bands (Figure 2D),
which showed higher beta-sheet crystallinity for HP scaffolds
(35.97 ±1.39% vs 30.70 ±2.09%, p<0.05, Figure 2E).
2.2. In Vitro Immune Cell Characterization
To determine the effects of scaffold architecture on immune cell
activation, LP and HP scaffolds were seeded with either murine
macrophage RAW 246.7 or human T lymphocyte Jurkat cells in
the presence of inflammatory stimuli. Each cell type was chosen
to represent key cellular elements of the innate and adaptive im-
mune systems, respectively.
Cell morphology was used as an initial indicator of inflamma-
tory activation. Representative actin stains showed RAW 264.7
cells with increased spreading following inflammatory stimula-
tion when plated on tissue culture plate (TCP) controls and HP
scaffolds, but not when plated on LP scaffolds (Figure 3A). Quan-
tification of cell spreading showed no differences in RAW 246.7
cell area between LP and HP without stimulation. However, fol-
lowing stimulation, the average size of RAW 246.7 cells increased
1.3-fold when seeded on HP scaffolds compared with those
seeded on LP scaffolds (470.710 ±5.415 vs 363.827 ±28.024 µm2,
p<0.01, Figure 3B). Moreover, significantly higher levels of
pro-inflammatory tumor necrosis factor (TNF)-𝛼secretion was
detected from stimulated RAW 246.7 cells seeded on HP scaf-
folds compared to LP scaffolds (368.733 ±91.650 vs 0 pg mL−1,
p<0.001, Figure 3C).
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (2 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 1. Characterization of electrospun LP and HP silk scaolds. A) Representative SEM images of fiber diameter of LP and HP silk scaold at ×
magnification. Scale bar represents m. B) Quantification of fiber thickness, each dot representing one fiber measurement (****p<.). C)
Analysis of porosity of LP and HP, shown as percentage (****p<.).
Figure 2. A) Representative stress–strain curves of LP and HP silk scaolds. B) Young’s modulus of both LP and HP silk scaolds (****p<.). C)
Ultimate tensile strength (UTS) of LP and HP silk scaolds (*p<.). D) Representative amide I and amide II spectra (– cm−)ofLPand
HP silk scaolds. E) Quantification of beta-sheet crystallinity of LP and HP silk scaolds (*p<.).
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (3 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 3. In vitro immune cell characterization of electrospun LP and
HP silk scaolds. A) Representative F-actin staining images of Raw cells
seeded on silk scaolds. Scale bar represents m. B) Quantification
of Raw cells spreading with and without LPS stimulation (**p<.,
***p<., ****p<.). C) ELISA of TNF𝛼with Raw cells seeded
on silk scaolds (**p<., ***p<.). D) Representative Calcein
staining images of Jurkat cells seeded on silk scaolds. Scale bar repre-
sents m. E) Quantification of Jurkat cells spreading with and without
PMA/iono stimulation (***p<., ****p<.). F) ELISA of IL-
with Jurkat cells seeded on silk scaolds (****p<.).
Similarly, representative calcein stains showed that activated
Jurkats appeared larger and more clustered on both TCP con-
trols and HP scaffolds, but not on LP scaffolds (Figure 3D).
Quantification of cell spreading showed no difference in Jurkat
cell area between LP and HP in the absence of immune stim-
ulation. However, following stimulation, the average size of Ju-
rkat cells increased 1.7-fold when seeded on HP scaffolds com-
pared to those seeded on LP scaffolds (369.536 ±39.872 vs
210.047 ±11.918 µm2,p<0.0001, Figure 3E). This was consis-
tent with significantly higher levels of interleukin (IL)-2 secretion
observed in Jurkats seeded on HP scaffolds compared to LP scaf-
folds (898.6 ±84.22 vs 0 pg mL−1,p<0.0001, Figure 3F). Together
these experiments revealed that both macrophages and T cells
have suppressed immune responses following pro-inflammatory
activation when seeded on LP silk compared to HP silk.
2.3. Bioluminescent Tracking Model
Bioluminescent BM-MNCs harvested from FVB-L2G transgenic
mice constitutively expressing a firefly luciferase reporter were
injected into WT mice implanted with HP and LP scaffolds.
Serial bioluminescence imaging was conducted over 14 days
to non-invasively determine levels of BM-MNC engraftment
within each implanted scaffold (Figure 4A). Temporal biolu-
minescence curves showed that HP scaffolds had significantly
increased bioluminescence compared to LP scaffolds at day
7 (49 140 ±4352 vs 32 995 ±4804 photon/s/cm2/steradian,
p<0.05, Figure 4B). Total area-under-the-curve (AUC) analysis
showed that HP scaffolds had a 1.6-fold increase in biolumines-
cence compared to LP scaffolds over the 14 days (495 325 ±40 664
vs 319 002 ±36 380 photon/s/cm2/steradian*day, p<0.05, Fig-
ure 4C). Representative images of bioluminescence at 1, 7, and
14 days showed that BM-MNC engraftment occurred at the areas
of scaffold implantation (Figure 4D). These results showed that
total BM-MNC engraftment over 14 days was higher within HP
scaffolds.
2.4. Macrophage Responses
Macrophage responses toward silk scaffolds were assessed by
immunostaining for the total macrophage population and po-
larized phenotypes within the capsule and scaffolds. HP scaf-
folds showed increased CD68+expression compared to LP scaf-
folds at day 14 post implantation (3963 ±224.9 vs 2115 ±175.7,
p<0.0001) (Figure 5A), indicating a greater number of
recruited macrophages. Further analysis into polarization/
phenotype was conducted using the MHC Class II (MHCII)
and CD206 markers, respectively. HP scaffolds showed in-
creased M1 polarized macrophages compared to LP scaf-
folds (Figure 5B), indicated by a 32-fold increase in MHCII
(17.47 ±2.517% vs 0.5282 ±0.1601%, p<0.0001). Conversely,
LP scaffolds showed higher M2 macrophage polarization indi-
cated by a 1.6-fold increase in CD206 compared to HP scaffolds
(20.40 ±4.192% vs 7.908 ±1.513%, p<0.05, Figure 5C). The state
of macrophage driven immune responses is best represented by
the overall ratio of M2/M1 polarization. The M2/M1 ratio was
strikingly higher in LP scaffolds compared to HP silk scaffolds
(61.22 ±19.20% vs 0.51 ±0.12%, p<0.01, Figure 5D), indicat-
ing a more anti-inflammatory microenvironment in LP scaffolds
and more pro-inflammatory in HP scaffolds. Representative im-
ages showed increased M1 macrophage staining occurring pri-
marily within the body of HP scaffolds. In contrast, increased
M2 macrophage staining found in LP scaffolds was limited to
the scaffold periphery (Figure 5E). These results suggest that
intrinsic architectural differences of each scaffold were driving
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (4 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 4. Bioluminescence tracking model of in vivo BM-MNC. A) Experimental procedure of BM-MNC bioluminescence tracking model. B) Biolumi-
nescence measurements of silk scaold implanted mice with BM-MNC at , , and days. C) Area under the curve (AUC) analysis of bioluminescence
measurements of both LP and HP scaolds over days (*p<.). D) Representative IVIS images of day , , and are also shown below the curve.
opposing macrophage responses consistent with BM-MNC en-
graftment.
2.5. T Cell Responses
Similar assessments of T cell responses were conducted by
immunostaining total T cell numbers and phenotype within
the capsule and scaffolds. HP scaffolds showed a signifi-
cantly higher number of total T cells indicated by a twofold
increase in CD4 compared to HP silk scaffolds (1635 ±
185.2 vs 835.3 ±96.72 mm−2,p<0.01, Figure 6A). Surprisingly,
LP scaffolds showed an increased number of pro-inflammatory
Th1 cells compared to HP scaffolds (58.10 ±2.963% vs 45.39 ±
1.946%, p<0.01, Figure 6B). Further immunostaining showed a
striking increase of anti-inflammatory regulatory T cells (Treg) in
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (5 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 5. In vivo analysis of macrophages within both the capsule and scaolds. A) Quantification of CD+macrophages as the number of counts
within both the capsule area and LP and HP scaolds (****p<.). B) Quantification of M macrophage polarization represented as a ratio of
MHCII+CD+/CD+stained area within both the capsule and LP and HP scaolds (****p<.). C) Quantification of M macrophage polariza-
tion represented as a ratio of CD+CD+/CD+stained area within both the capsule and LP and HP scaolds (*p<.). D) Quantification of
M:M macrophage polarization represented as a ratio of MHCII+CD+/CD+CD+stained area within both the capsule and LP and HP scaolds
(**p<.). Data is represented as mean ±standard error of the mean, t-test was performed between two groups. E) Representative images of cross
section of both LP and HP silk scaolds; CD stained in green; MHCII and CD stained in red. White dotted lines indicate the interface of scaolds.
Scale bar represents m.
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (6 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 6. In vivo analysis of T cells within both the capsule and scaolds. A) Quantification of CD+T cells as the number of counts within both the
capsule area and LP and HP scaolds (**p<.). B) Quantification of Th cells represented as a ratio of T-bet+CD+/CD+stained area within both
the capsule and LP and HP scaolds (**p<.). C) Quantification of Treg cells represented as a ratio of FoxP+CD+/CD+stained area within both
the capsule and LP and HP scaolds (****p<.). D) Quantification of Treg:Th cells represented as a ratio of FoxP+CD+/Tbet+CD+stained
area within both the capsule and LP and HP scaolds (****p<.). Data is represented as mean ±standard error of the mean, t-test was performed
between two groups. E) Representative images of cross section of both LP and HP silk scaolds; CD stained in green; T-bet and FoxP stained in red.
White dotted lines indicate the interface of scaolds. Scale bar represents m.
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (7 of 14)
www.advancedsciencenews.com www.advhealthmat.de
LP scaffolds, indicated by a 11-fold increase in FoxP3+expression
compared to HP scaffolds (12.04 ±1.228% vs 1.442 ±0.5954%,
p<0.0001, Figure 6C). To determine the overall state of T cell ac-
tivation a Treg/Th1 ratio was calculated. Treg/Th1 ratio showed
a significant decrease in HP scaffolds compared to LP scaffolds
indicating a more immune active T cell microenvironment in
HP scaffolds and more immune suppressive in LP scaffolds
(3.479 ±1.577 vs 21.48 ±2.791, p<0.0001, Figure 6D). Rep-
resentative images showed T cells localized only to the scaffold
periphery (Figure 6E).
2.6. Cytokine Analysis
Further evaluation of the immune response to LP and HP
silk was conducted by staining classical pro-inflammatory (IL-
1𝛽and TNF-𝛼) and anti-inflammatory (transforming growth
factor [TGF]-𝛽and IL-10) cytokines. For pro-inflammatory cy-
tokines, HP scaffolds showed a significant increase in both
IL-1𝛽(2.160 ±0.3997% vs 10.72 ±1.754%, p<0.01, Figure
S1A, Supporting Information) and TNF-𝛼(2.461 ±0.4277%
vs 9.997 ±2.404%, p<0.05, Figure S1B, Supporting In-
formation) compared to LP silk. For anti-inflammatory cy-
tokines, there was no significant difference observed in TGF-𝛽
(6.084 ±0.9157% vs 9.499 ±1.651%, Figure S2A, Supporting In-
formation) or IL-10 (5.584 ±1.455% vs 5.726 ±0.7322%, Figure
S2B, Supporting Information) between HP and LP silk. This in-
dicated that HP silk induced more pro-inflammatory responses
that LP silk.
2.7. Cell Infiltration
To evaluate the cellular remodeling effects of silk scaffold inflam-
mation and BM-MNC engraftment, Haematoxylin and Eosin
(H&E) staining was performed. HP scaffolds showed deeper cell
infiltration (53.61 ±2.237 vs 0.2168 ±0.1182 µm, p<0.0001,
Figure 7A) as well as greater total cells (170.1 ±11.64 vs 11.32 ±
9.627 mm−2,p<0.001, Figure 7B) compared to LP scaffolds after
14 days.
H&E staining was also used to measure fibrous encapsu-
lation surrounding implanted scaffolds. Fibrous capsule area
and thickness were more extensive for HP scaffolds com-
pared to LP scaffolds. Mean capsule area percentage for HP
silk scaffolds was significantly greater than LP silk scaffolds
(114.3 ±11.49% vs 38.22 ±4.067%, p<0.01, Figure 7C). Sim-
ilarly, HP silk scaffolds had a significantly higher mean cap-
sule thickness compared to LP scaffolds (80.95 ±0.7174 vs
29.97 ±3.797 µm, p<0.001, Figure 7D). Representative H&E im-
ages of scaffold cross sections showed that capsules surrounding
HP scaffolds were highly cellularized (Figure 7D).
2.8. Collagen Deposition
Further evaluation of functional BM-MNC driven remodeling in
LP and HP scaffolds was conducted by assessing collagen deposi-
tion using Picro-sirius Red (PSR) staining. HP scaffolds showed
a 3.7-fold increase in ratio of collagen/capsule area compared to
LP scaffolds (12.7 ±2.7% vs 4.1 ±2.0%, p<0.01, Figure 8A).
Polarized light microscopy was used on PSR stains for closer ex-
amination of collagen type. Type I collagen (thick fibers) appears
as hues of red/yellow/orange, where type III (thin fibers) appears
as green.[22,23 ] HP scaffolds contained a significantly higher per-
centage of type I collagen in their capsule compared to LP scaf-
folds (12.28 ±1.246% vs 3.982 ±0.9791%, p<0.01, Figure 8B).
In contrast, minimal type III collagen was observed in either scaf-
fold. Representative PSR images showed that collagen deposition
occurs mainly throughout the fibrous capsule and not within the
scaffold body (Figure 8C), which is consistent with the localiza-
tion of type I collagen in polarized light microscopy images (Fig-
ure 8D).
3. Discussion
Enhancing the engraftment of BM-MNC infusions through the
use of implantable scaffolds is a promising approach to improve
therapeutic efficacy and clinical feasibility. While it is well-known
that BM-MNC infusions naturally home and engraft toward ar-
eas of tissue inflammation and injury,[11,24] scaffold implanta-
tion alone stimulates independent inflammatory mechanisms
driving characteristic foreign body responses.[25] Although the
roles particular implant features play in altering these immune
responses has been well-documented,[26] their possible impact
on systemic BM-MNC infusion and engraftment have not been
studied. A more nuanced understanding of the associations be-
tween implant-induced local inflammation and its influence on
the engraftment of circulating BM-MNC is critical to designing
implantable scaffolds with enhanced therapeutic impact. In this
study, we examined distinct architectural features of electrospun
silk scaffolds and determine whether they could serve as biophys-
ical cues to regulate the local immune responses that drive en-
graftment of BM-MNC infusions.
As products of separate manufacturing processes described in
our previous work,[21] we produced two architecturally distinct
scaffold formulations of silk fibroin with either thin fibers and
LP or thick fibers and HP. Consistent with our prior charac-
terization, both scaffolds show striking architectural differences
in both fiber thickness and porosity which translate into differ-
ences in mechanical properties that remain in the same order
of magnitude but are statistically distinct. HP silk was found to
be marginally stiffer and stronger than LP silk, consistent with an
observed increase in 𝛽-sheet crystallinity. Despite being manufac-
tured from chemically distinct solvents, surface chemistry analy-
sis showed no residual traces of 1,1,1,3,3,3-hexafluoro-2-propanol
(HFP) on HP scaffolds, removing the possibility of adverse re-
active chemical groups that could potentially cause more promi-
nent immune responses. On this basis, we inferred that differen-
tial cellular responses observed in our study were not confounded
by opposing chemical groups/structures within either scaffold
surface, but rather were a direct consequence of the striking bio-
physical differences intrinsic to each scaffold structure. While
it is extremely challenging to decouple fiber thickness, porosity,
and stiffness to identify a single biophysical cue with the most
impact on local inflammation, both scaffolds were manufactured
with architectures different enough from one another to deter-
mine whether biophysical cues were involved to any degree in
modulating local immune cells.[27]
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (8 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 7. Cell infiltration in subcutaneously implanted scaolds. A) Quantification of depth of cell infiltration in subcutaneously implanted LP and HP
scaolds (****p<.). B) Quantification of infiltrated cell numbers within scaold (***p<.). C) Quantification of fibrous capsule area in
subcutaneously implanted LP and HP scaolds (**p<.). D) Quantification of capsule thickness (***p<.). E) Representative images of cross
section of the scaold stained with H&E. Black dotted lines indicate the interface between scaolds and fibrotic capsule. Red arrowheads indicate
infiltrated cells in scaolds. Scale bar represents m.
Studying in vitro immune cell cultures of macrophages and T
cells revealed that scaffold architecture was a major regulator of
immune activation. In the absence of external pro-inflammatory
stimuli, no differences were observed in either HP or LP scaffolds
compared to TCPs, when assessing changes in basal cell mor-
phology and pro-inflammatory cytokine release. This suggested
that neither of the silk scaffolds possessed overt immune reactiv-
ity. Only after challenging cultures with their respective classical
inflammatory stimuli (lipopolysaccharide [LPS] for macrophages
and phorbol 12-myristate 13-acetate [PMA]/ionomycin for T
cells) were differences in immune responses revealed. LP scaf-
folds prevented immune activation entirely in both immune cell
types. This suggested scaffold architecture did not actively stimu-
late inflammation but instead, primed cells for downstream im-
mune activation. Previous biophysical signaling studies suggest
two potential mechanisms for this effect. Macrophages are well-
known mechanosensory cells[28] and when cultured on PCL vas-
cular grafts, showed a positive correlation between fiber thick-
ness and phenotype polarization.[29] Macrophages were found to
polarize more toward pro-inflammatory M1 phenotypes on thin-
ner PCL fibers and more toward anti-inflammatory M2 pheno-
types on thicker PCL fibers in the absence of pro-inflammatory
stimuli. Additionally, the direction of macrophage polarization
was opposite to our findings. This potentially suggests that
fiber composition in addition to its thickness regulates immune
cell activation. Additionally, spatial confinement, such as those
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (9 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Figure 8. Collagen deposition within both the capsule and scaolds. A) Quantification of a ratio of collagen/capsule area (**p<.). B) Quantification
of proportions of type I&III collagen within the capsule area (**p<.). Data is represented as mean ±standard error of the mean. C) Representative
Picro-sirius red stained images of cross section of both LP and HP silk scaolds taken by light microscopy; collagen stained in red, muscle fibers stained
in yellow, nuclei stained in brown. Black dotted lines indicate the interface of scaolds. D) Representative Picro-sirius red stained images of cross sections
of both LP and HP silk scaolds taken by polarized light microscopy; type I collagen stained in red/orange/yellow, type III collagen stained in green.
White dotted lines indicate the interface of scaolds. Scale bar represents m.
imposed by microporous substrates and micropatterning, has
been demonstrated to reduce macrophage spreading, thus down-
regulating LPS-activated pro-inflammatory cytokine expression
and downstream phagocytic activities.[30] This suggests that
when cells become lodged in the pores of LP scaffolds, the sur-
face area needed for cells to sufficiently change morphology is
limited and therefore serves to suppress activation.
While these effects have so far only been reported in
macrophages, previous mechanobiology studies suggest that the
suppression of T cell activation observed in our study may be
explained by similar mechanisms. T cells are exposed to a wide
range of biomechanical environments that vary considerably in
the body, ranging from soft tissue like the bone marrow to stiffer
peripheral inflamed tissue. In response, T cell receptors (TCRs)
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (10 of 14)
www.advancedsciencenews.com www.advhealthmat.de
act as mechanosensory receptors which can alter activation in
response to external forces and tissue stiffness.[31] Studies in-
vestigating T cells plated on polyacrylamide gels with increasing
Young’s modulus showed that stiffness positively regulated mi-
gration, morphological activation, and cell-cycle-related gene ex-
pression. It further showed that the sensitivity of TCRs to sub-
strate stiffness was as low as 0.5–100 kPa,[32] suggesting that
differences in LP and HP, while relatively minor, were suffi-
cient to regulate Jurkat activation accordingly. Also similar to
macrophages, actin polymerization of the cytoskeleton is a key
downstream event of TCR activation and thus spatial confine-
ment in LP scaffolds could have similarly affected Jurkat acti-
vation. While our data does not isolate a particular underlying
mechanism, our findings collectively show that silk scaffolds do
not stimulate inflammation but rather modulate activation mech-
anisms in both Jurkats and RAW cells, demonstrating the archi-
tectural influence on adaptive and innate immunity in vitro.
Using our previously established BM-MNC in vivo tracking
model we evaluated homing and engraftment to HP and LP
scaffolds. Bioluminescence curves quantified differences in BM-
MNC engraftment over time within each scaffold. AUC analysis
showed that overall engraftment over 14 days was significantly
greater in HP scaffolds. Looking closer at the temporal trends
of the bioluminescent curves, we found the greatest differences
in engraftment occurred within the first 7 days, which is known
to be the acute to subacute phase of the foreign body response
where the immune landscape is most rapidly changing.
Further analysis of the inflammatory microenvironment at day
14 post-implantation showed strikingly different immune cell re-
modeling of both scaffolds consistent with these levels of BM-
MNC engraftment. Elevated macrophage recruitment and M1
phenotype was observed in the HP scaffolds in comparison to
LP scaffolds which showed higher levels of M2 macrophages.
More interesting were the differences in macrophage localization
within either scaffold. In LP scaffolds, macrophages appeared
limited to the scaffold periphery with an enriched M2 pheno-
type. In contrast, HP scaffolds showed macrophages uniformly
infiltrated within the scaffold body in a predominantly M1 phe-
notype. Interestingly, numerous studies have reported opposing
observations often linking enhanced macrophage infiltration to
M2 polarization and less inflammatory phagocytic activity.[ 33–35]
As most of these studies involve synthetic materials, this sup-
ports that idea that traditional biophysical cues controlling im-
mune activation may be regulated differently in silk materials.
This could in part be explained by the lack of silk immune re-
activity found in our in vitro experiments. More importantly, the
heighted pro-inflammatory macrophage profile of HP scaffolds
was likely responsible for improved engraftment of BM-MNCs.
M1 macrophages are characteristically known to facilitate im-
mune cell recruitment through the secretion of chemokines in-
cluding CXCL1, CXCL3, and CXCL5 as well as sustaining im-
mune activation through cytokines such as IL-1𝛽, IL-6, and TNF-
𝛼.[36] Given the cross-talk between innate and adaptive immune
responses, these macrophages responses were likely interacting
synergistically with T cells to sustain higher levels of inflamma-
tion and BM-MNC engraftment.
HP scaffolds showed higher levels of total T cells; however,
in contrast with macrophage responses, polarization trends be-
tween the two scaffolds were less clear. LP scaffolds unex-
pectedly showed higher relative levels of pro-inflammatory Th1
cells. However, LP scaffolds also showed higher levels of anti-
inflammatory Treg cells compared to HP scaffolds. The mag-
nitude of this difference was also greater than the difference
in Th1 cells between LP and HP scaffolds. This was sugges-
tive of an enhanced immunosuppressive environment around LP
scaffolds and that enhanced BM-MNC engraftment in HP scaf-
folds may have been the result of decreased anti-inflammatory
T cell function rather than elevated pro-inflammatory function.
Tregs function as potent resolvers of both adaptive and innate in-
flammation. Tregs primarily secrete IL-4 and IL-13 which stimu-
late the production of anti-inflammatory IL-10 in macrophages,
inducing M2 polarization and stemming immune cell recruit-
ment. However, whether these effects are critical to BM-MNC
engraftment is still unknown. Previous studies have suggested
that T cells are not traditionally necessary for locally induced for-
eign body responses toward materials. Studies using T cell de-
ficient mice implanted with Elasthane 80A (PEU), silicone rub-
ber and poly(ethylene terephthalate), showed that T cells serve
as a redundant source of IL-4 and IL-13 and do not influence
remodeling outcomes such as macrophage fusion and fibrous
encapsulation.[37] As commonly discussed throughout this study,
it should be noted that these previous works used synthetic ma-
terials. This could again suggest that T cells may trigger ad-
ditional and/or entirely separate activation cascades intrinsic
to their interactions with silk. The decreased immunosuppres-
sive functions of T cells observed within HP scaffolds provided
yet another pro-inflammatory cue advantageous for BM-MNC
engraftment.
Cytokine analysis showed that HP silk increased pro-
inflammatory cytokine expression, while no differences were ob-
served in anti-inflammatory cytokine expression. This suggested
that the observed differences in both macrophage and T cell pop-
ulations could be explained by increased inflammation around
HP scaffolds. This contrasts with previous notions that LP scaf-
folds stimulated anti-inflammatory signaling to suppress inflam-
mation. Together, these differential immune responses estab-
lished microenvironments that impacted BM-MNC engraftment.
Tissue remodeling of the scaffolds potentially further aug-
mented these engraftment differences. HP scaffolds showed a
greater extent of cell infiltration and fibrous encapsulation. In-
creased cell infiltration was expected with increased porosity
which, from an architectural standpoint, provides increased sur-
face area for cell engraftment to occur. However, contrary to tra-
ditional foreign body capsules which are typically smaller when
cell infiltration is improved, HP scaffolds showed larger capsules.
This was most likely a result of the enhanced engraftment of
BM-MNCs which are well-known to facilitate tissue remodeling
and scar formation.[38] The most notable cellular contributors
reported are the abundance of stem cell fractions within BM-
MNCs that possess the potential to differentiate into a multitude
of cell lineages,[39] including collagen-producing fibroblasts.[38,40]
The increased deposition of type I collagen in particular was also
more favorable for improved BM-MNC engraftment which is a
well-suited substrate commonly used in biomimicry of the bone
marrow niche. Additionally, in the context of functional tissue
repair, improved collagen deposition and remodeling is critical
to healing and regeneration of damaged organs. These results
support the beneficial associations between improved BM-MNC
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (11 of 14)
www.advancedsciencenews.com www.advhealthmat.de
engraftment within HP scaffolds and the resulting enhanced
therapeutic outcomes.
4. Conclusion
The use of implantable scaffolds to enhance the in situ engraft-
ment of BM-MNC infusions is a promising approach to improve
therapeutic efficacy and clinical feasibility. In addition to its mate-
rial composition, scaffold architecture provides biophysical cues
that can be leveraged to drive local immune responses and fur-
ther improve BM-MNC engraftment. Electrospun silk fibroin
scaffolds with key architectural differences result in significant
influential effects on the engraftment of BM-MNC infusions. LP
silk with thinner fibers and increased porosity inhibits immune
activation of macrophages and T cells. In contrast, HP silk with
thick fibers and increased porosity facilitates immune activation
in these cells. These effects translate to similar outcomes on lo-
cal inflammation when both scaffolds are implanted in vivo. LP
scaffolds stimulate anti-inflammatory M2 macrophage and Treg
T cell responses compared to HP scaffolds that promote pro-
inflammatory M1 macrophage and Th1 T cell activity. This en-
hanced pro-inflammatory microenvironment in HP scaffolds re-
sults in increased BM-MNC engraftment and heightened levels
of tissue remodeling with increased type I collagen deposition.
These results show that scaffold architecture can be used to reg-
ulate BM-MNC engraftment and potential therapeutic outcomes.
Consideration of architecture and the integration of biophysical
cues are vital to the design of implantable scaffolds aimed at im-
proving the engraftment of BM-MNC infusions. These findings
inform the future design of biomimetic scaffolds that may better
enhance the clinical effectiveness of BM-MNC infusion therapy.
5. Experimental Section
Materials:All reagents were purchased from Sigma-Aldrich (St. Louis,
MO, USA) unless otherwise stated. All antibodies and ELISA kits were
purchased from Abcam (Cambridge, UK). B. mori silk cocoons were pur-
chased from Tajima Shoji Co., Ltd., Yokohama, Japan.
Silk Preparation:Silk fibroin was isolated from B. mori silk cocoons
as previously reported.[] Briefly, g of silk cocoons were boiled in L
of . sodium carbonate for min to remove sericin. The extracted
silk fibroin was thoroughly washed in ultrapure water, dried overnight, and
then dissolved in a . lithium bromide solution at % w/v for h at
°C. The dissolved silk solution was transferred to SnakeSkin dialysis
tubing ( MWCO, Thermo Fisher Scientific, Waltham, MA, USA) and
dialyzed against L of ultrapure water for h. Water was changed nine
times during dialysis. Debris and flocculent were removed by centrifuga-
tion ( ×g, °C, min, repeated twice). Concentration of the final
silk solution was measured by drying . mL and weighing the remaining
solid material. This protocol produced on average an .% w/v aqueous
silk solution. Silk solution was stored at °C (to be used within month),
or lyophilized for h and stored at °C.
Electrospinning:Scaolds with HP and LP were electrospun as pre-
viously described.[] HP scaolds were electrospun from lyophilized silk
dissolved in HFP at % w/v. LP scaolds were electrospun from aqueous
silk solution diluted with water to % w/v, then mixed with polyethylene
oxide (PEO, % w/v in water) in a : ratio (silk:PEO, v/v). The volume of
each solution used for electrospinning ( mL HP; mL LP) was adjusted
to account for the dierence in concentration to produce scaolds with ap-
proximately the same amount of silk. Solutions were loaded into syringes
(Terumo Corp., Tokyo, Japan) and pumped through a . mm diameter
tube to a . mm diameter needle connected to a kV power supply. A
mm diameter mandrel rotating at rpm was positioned cm from
the needle tip. A syringe pump (Harvard Apparatus, Holliston, MA, USA)
was used to control the flow rate ( mL h−HP; . mL h−LP). Rela-
tive humidity was maintained between –%.
Scaffold Post-Treatments:Scaolds were cross-linked by water anneal-
ing in a humidified vacuum chamber (Wheaten) at room temperature for
h. Relative humidity within the chamber reached % within h once
under vacuum conditions. Cross-linked scaolds were soaked in L ul-
trapure water overnight with gentle stirring to remove residual HFP and
PEO, in HP and LP scaolds, respectively.Scaolds were then immersed in
% (v/v) ethanol for min to increase the crystallinity, washed briefly in
water, then air-dried and stored at room temperature. Scaolds were cut
to standard sample sizes using a Stiefel mm diameter circular biopsy
punch. Samples were then sterilized by UV radiation for min on each
side and washed thrice in sterile PBS prior to use.
Scanning Electron Microscopy:Fiber diameter was quantified from
scanning electron micrographs. Samples were prepared by gold sputter
coating under argon atmosphere to create a nm surface layer. Images
were taken with a Neoscope JCM- SEM (JEOL, Tokyo, Japan) under
high vacuum at an accelerating voltage of kV and ×magnification.
dierent fibers were measured from five dierent scaolds each for
LP and HP silk, using ImageJ software (version .r).
Porosity Measurement:For porosity measurements, five scaolds each
for both LP and HP silk scaolds were embedded in JB- resin as previously
published.[] Samples were dehydrated with gradients of ethanol, fol-
lowed by the incubation of infiltration and embedding solution. The resin
blocks were cut into m cross sections for quantification. For porosity
measurement, cross-sectional images were converted into grayscale and
the ratio of white (scaold)/black (pores) was represented as percentage
porosity.
Fourier Transformed Infra-Red Spectroscopy:The secondary structure of
silk scaolds was determined by Fourier transform spectroscopy in at-
tenuated total reflectance mode using a Bruker Alpha spectrometer (MA,
USA). Infrared spectra were collected by averaging a total of scans at
aspectralresolutionofcm
−within the wavenumber range of –
cm−. amide I bands (– cm−) were deconvoluted using
a FSD algorithm by adopting Gaussian line profiles. The relative contri-
bution of each secondary structure was determined by integration of the
corresponding fitted peaks obtained from FSD calculations. The position
of each peak was further confirmed by double dierentiation of the amide
I envelopes and peak assignment was as follow:[, ] side chains (–
cm−), 𝛽-sheets (– and – cm−), random-coils
(– cm−), alpha-helix (– cm−), and beta-turns (–
cm−).[, ] The contribution of 𝛽-sheets was determined by calcu-
lating the relative areas of the corresponding peaks in the area-normalized
deconvoluted amide I profile.
Mechanical Testing:Uni-axial tensile properties were measured on an
Instron model (Instron, Melbourne, Australia) equipped with a N
load cell. Silk scaolds were incubated in phosphate buered saline (PBS)
at °C for h, then cut into rectangular strips . mm wide with a
mm gauge length. Digital calipers were used to confirm sample dimen-
sions prior to each test. A water bath filled with PBS at °C simulated
physiological conditions. Force was applied at a constant crosshead speed
of mm min−until sample failure. Force and extension were recorded us-
ing Bluehill software (version .) and converted to stress and strain.
Young’s modulus was calculated from the gradient of the initialstrictly lin-
ear region of stress–strain curves using linear regression. UTS was taken
as the maximum stress recorded before failure.
In Vitro Culture and Stimulation:RAW . murine macrophage
cells were cultured in Corning T- flasks in Dulbecco’s modified Ea-
gle’s medium supplemented with % (v/v) fetal bovine serum (FBS),
U mL−penicillin, and g mL−streptomycin at °C, % CO.
Cells were subcultured at ≈% confluency. Jurkat human T-lymphocytes
were cultured in upright Corning T- flasks in RPMI- medium sup-
plemented with % (v/v) FBS, U mL−penicillin, and g mL−
streptomycin. Jurkat cells were subcultured every – days depending on
cell density by replacing ≈% of the cell suspension volume with fresh
medium. RAW . cells were stimulated using LPS ( ng mL−).
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (12 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Jurkat cells were stimulated using a combination of phorbol -myristate
-acetate (PMA, ng mL−) and ionomycin ( g mL−).
Cell Morphology and ELISA:RAW . and Jurkat cells were seeded
onto scaolds and TCP at . ×and . ×cells/mL, respectively,
in -well-plates. Cells were given h to settle prior to stimulation and
stimulants remained in the cell culture medium for the duration of the
experiments.
Macrophage morphology analysis was conducted at h post-
stimulation. RAW . seeded scaolds were fixed in % paraformalde-
hyde (PFA) and F-actin stained using ActinRed (Thermo Fisher MA, USA).
For T cell morphology analysis, Jurkat cells were incubated for min
in Calcein AM live cell dye (BD Pharmingen, Franklin Lakes,
NJ, USA) prior to scaold seeding and stimulation. T cell imaging was
conducted at min post-stimulation on a Zeiss Axio Vert.A fluores-
cent microscope. Average cell area was measured using ImageJ software
(version .r).
ELISA kits were used to quantify murine IL-𝛽(cat. no. ab) se-
cretion in RAW . cells after -h stimulation and human IL- (cat. no.
ab) secretion in Jurkat cells after -h stimulation. Cell supernatant
was collected, centrifuged at ×gfor min to remove cell debris prior
to analysis. Results with a negative value after correcting for the blank were
assigned a value of zero.
In Vivo Bioluminescent Tracking Model:CBL/ male mice aged
weeks were obtained from Animal BioResources (Mossvale, NSW, Aus-
tralia) and upon arrival, housed in a self-ventilated OptiMice caging sys-
tem with -h light/dark cycles, constant temperature of °C, and ad libi-
tum access to food and water. During surgery, anesthesia was induced and
maintained with % methoxyflurane. The dorsal surface of each mouse
was shaved, sterilized with betadine solution, and washed with sterile PBS.
Two mm incisions were made in the skin to create two subcutaneous
pockets, as previously described.[] A single mm biopsy punched ster-
ilized scaold was inserted into each incision so that each mouse carried
one HP and one LP scaold. Incisions were closed with - silk sutures
(Ethicon, J&J, USA). A total of four mice were used for tracking (n=
scaolds per group).
BM-MNC homing and engraftment to implanted scaolds were non-
invasively tracked using bioluminescence imaging in an in vivo imaging
system (IVIS; Perkin Elmer, USA) as previously described.[] Briefly, BM-
MNCs were harvested from the hindlimbs of donor FVB-LG transgenic
mice (Jackson Laboratories, USA) expressing firefly luciferase and isolated
using Ficoll density gradient centrifugation. BM-MNCs were then injected
( ×cells in L) into the tail vein of CBL/ mice immediately
after being implanted with scaolds. Mice were IVIS imaged at days ,
, and post implantations. Bioluminescence was measured in units of
radiance (photon/s/cm/steradian) and quantified as mean radiance val-
ues within a predefined region of interest surrounding each scaold. All
in vivo experiments were approved by the Sydney Local Heath District An-
imal Welfare Committee (protocol number /A) and conducted in
accordance with the Australian Code of Practice for the Care and Use of
Animals for Scientific Purpose.
Histology and Immunohistochemistry:Mice were euthanized by cervi-
cal dislocation and skin biopsies ( mm × mm) together with scaf-
fold implants were collected for histological analysis at day . Biop-
sies/scaolds were fixed in % PFA for h, dehydrated through an ethanol
gradient, embedded in paran, and sectioned at m. H&E and PSR
staining was used to assess cell infiltration/capsule thickness and col-
lagen deposition, respectively. Immunostaining was conducted with pri-
mary antibodies for total macrophages using CD (:, ab),
M macrophages using MHC Class II (MHC II, :, ab), and
M macrophages using CD (:, ab). Primary antibodies used
for T cell staining were CD for total T cells (:, ab), FoxP for
regulatory T cells (:, ab), and T-bet for Type I helper T cells
(:, ab). Primary antibodies for staining inflammatory mark-
ers were IL- (:, ab), IL-𝛽(:, ab), TNF-𝛼(:,
ab), and TGF-𝛽 (:, ab). Secondary antibodies used were
Alexa Fluor (:, ab) and Alexa Fluor (:, ab).
Immunostains were counterstained with DAPI mounting media (Fluo-
roshield, Sigma-Aldrich) to label cell nuclei. All images were taken with a
Zeiss Axio Scan.Z. Polarized images of PSR stained sections were taken
with a Leica DM.
Image Analysis:Quantitative image analysis was performed using Im-
ageJ software (version .r). Capsule thickness, depth of cell infiltration,
fibrous encapsulation, and the number of cells within scaolds were mea-
sured from H&E images. Capsule thickness was measured as the shortest
distance in a straight line from the outer capsule perimeter to the scaold
along the long axis (five measurements per side) for each sample. Depth
of cell infiltration was taken as the shortest distance from a cell within
the scaold to the nearest scaold edge (ten measurements per sample).
Fibrous encapsulation was measured as the capsule area divided by the
scaold boundary length, to account for dierences in scaold size. The
number of cells within scaolds was counted by first isolating the haema-
toxylin color (i.e., cell nuclei) by color deconvolution (H&E), then using
the “analyze particles” function to count cells.
Collagen deposition was measured from brightfield PSR images. The
proportion of collagen types present was determined from polarized PSR
images using color thresholds to separate hues. The following hue defini-
tions were used as previously described:[] red – and –, orange
–, yellow –, and green –. Hues in the range – were
considered interstitial space and non-collagen tissue.
Total macrophage infiltration was determined by the number of CD+
cells within the scaold and capsule. Macrophage polarization was as-
sessed by comparing the proportions of M and M macrophages. M
macrophages were quantified as the percentage of MHCII+CD+stained
area to CD+stained area in the total area of the scaold along with cap-
sule. Similarly, M macrophage was measured as a ratio of CD+CD+
to CD+. T cells were quantified by measuring the number of CD posi-
tively stained cells in the scaold and capsule areas. Th cells were quan-
tified as a percentage of Tbet+CD+stained area to CD+stained area in
the total area of the scaold along with capsule. Similarly, Treg cells were
measured as a ratio of FoxP+CD+to CD+.
Statistics:Statistical analysis were performed using GraphPad Prism
(version .. for Windows, GraphPad Software, La Jolla, CA, USA). Sig-
nificant dierences were determined by unpaired Student’s t-tests for two-
mean data sets, and one-way analysis of variance (ANOVA) with Holm–
Sidak post-hoc multiple comparisons for data sets with more than two
means. For data sets with two independent variables a two-way ANOVA
with post-hoc multiple comparisons were used. Statistical significance
was set at p<.. Data was presented as mean ±standard error of the
mean.
Supporting Information
Supporting Information is available from the Wiley Online Library or from
the author.
Acknowledgements
N.Y. and M.J.M. contributed equally first to this manuscript. R.P.T. and
S.G.W. contributed equally first to this manuscript. The authors acknowl-
edge the facilities as well as scientific and technical assistance at the Aus-
tralian Center for Microscopy and Microanalysis, The University of Sydney.
This work was supported by the National Health and Medical Research
Council (APP; S.G.W., M.K.C.N., and J.R.-K.) and funding from the
Sydney Local Health District (S.G.W. and M.K.C.N.). The authors also ac-
knowledge the financial support of E. Brackenreg.
Conflict of Interest
The authors declare no conflict of interest.
Data Availability Statement
Research data are not shared.
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (13 of 14)
www.advancedsciencenews.com www.advhealthmat.de
Keywords
biophysical cues, bone marrow mononuclear cells, electrospinning, in-
flammation, silk fibroin
Received: March ,
Published online:
[] V. Schächinger, S. Erbs, A. Elsässer, W. Haberbosch, R. Hambrecht,
H. Hölschermann, J. Yu, R. Corti, D. G. Mathey, C. W. Hamm, N. Engl.
J. Med. 2006,355, .
[] J. M. Hare, J. E. Fishman, G. Gerstenblith, D. L. D. Velazquez, J. P.
Zambrano, V. Y. Suncion, M. Tracy, E. Ghersin, P. V. Johnston, J. A.
Brinker, JAMA, J. Am. Med. Assoc. 2012,308, .
[] D. Sürder, R. Manka, V. L. Cicero, T. Moccetti, K. Rufibach, S. Soncin,
L. Turchetto, M. Radrizzani, G. Astori, J. Schwitter, Circulation 2013,
127, .
[] F. Cao, D. Sun, C. Li, K. Narsinh, L. Zhao, X. Li, X. Feng, J. Zhang, Y.
Duan, J. Wang, Eur. Heart J. 2009,30, .
[] B. Pignon, M.-A. Sevestre, L. Kanagaratnam, G. Pernod, D. Stephan,
J. Emmerich, C. Clement, G. Sarlon, C. Boulon, C. Tournois, Circ. J.
2017,81, .
[] S. P. Weem, M. Teraa, G. De Borst, M. Verhaar, F. Moll, Eur. J. Vasc.
Endovasc. Surg. 2015,50, .
[] K. Prasad, A. Sharma, A. Garg, S. Mohanty, S. Bhatnagar, S. Johri, K.
K. Singh, V. Nair, R. S. Sarkar, S. P. Gorthi, Stroke 2014,45, .
[] S. I. Savitz, V. Misra, M. Kasam, H. Juneja, C. S. Cox Jr., S. Alderman,
I. Aisiku, S. Kar, A. Gee, J. C. Grotta, Ann. Neurol. 2011,70, .
[] M. V. P. Mendonça, T. F. Larocca, B. S. de Freitas Souza, C. F. Villarreal,
L. F. M. Silva, A. C. Matos, M. A. Novaes, C. M. P. Bahia, A. C. d. O.
M. Martinez, C. M. Kaneto, Stem Cell Res. Ther. 2014,5, .
[] C. Toma, M. F. Pittenger, K. S. Cahill, B. J. Byrne, P. D. Kessler, Circu-
lation 2002,105, .
[] M. Hofmann, K. C. Wollert, G. P. Meyer, A. Menke, L. Arseniev, B.
Hertenstein, A. Ganser, W. H. Knapp, H. Drexler, Circulation 2005,
111, .
[] S. Gobaa, S. Hoehnel, M. Roccio, A. Negro, S. Kobel, M. P. Lutolf,
Nat. Methods 2011,8, .
[] K. H. Nakayama, L. Hou, N. F. Huang, Adv. Healthcare Mater. 2014,
3, .
[] C.-H. Wang, T.-M. Wang, T.-H. Young, Y.-K. Lai, M.-L. Yen, Biomaterials
2013,34, .
[] L. Li, X. Chen, W. E. Wang, C. Zeng, Stem Cells Int. 2016,2016,
.
[] R. P. Tan, B. S. Lee, A. H. Chan, S. C. G. Yuen, J. Hung, S. G. Wise, M.
K. Ng, Acta Biomater. 2017,53, .
[] G. S. Perrone, G. G. Leisk, T. J. Lo, J. E. Moreau, D. S. Haas, B. J.
Papenburg, E. B. Golden, B. P. Partlow, S. E. Fox, A. M. Ibrahim, S. J.
Lin, D. L. Kaplan, Nat. Commun. 2014,5, .
[] E. C. Filipe, M. Santos, J. Hung, B. S. Lee, N. Yang, A. H. Chan, M. K.
Ng, J. Rnjak-Kovacina, S. G. Wise, JACC: Basic Transl. Sci. 2018,3, .
[] Y. Sang, M. Li, J. Liu, Y. Yao, Z. Ding, L. Wang, L. Xiao, Q. Lu, X. Fu,
D. L. Kaplan, ACS Appl. Mater. Interfaces 2018,10, .
[] X. Wang, Z. Ding, C. Wang, X. Chen, H. Xu, Q. Lu, D. L. Kaplan, J.
Mater. Chem. B 2018,6, .
[] A. H. Chan, E. C. Filipe, R. P. Tan, M. Santos, N. Yang, J. Hung, J. Feng,
S. Nazir, A. J. Benn, M. K. Ng, J. Rnjak-Kovacina, S. G. Wise, Sci. Rep.
2019,9, .
[] L. Junqueira, G. Montes, E. Sanchez, Histochemistry 1982,74, .
[] R. Koren, E. Yaniv, D. Kristt, J. Shvero, V. Veltman, I. Grushko, R. Fein-
messer, J. Sulkes, R. Gal, Acta Histochem. 2001,103, .
[] A.Y.Sheikh,S.A.Lin,F.Cao,Y.Cao,K.E.VanDerBogt,P.Chu,C.
P. Chang, C. H. Contag, R. C. Robbins, J. C. Wu, Stem cells 2007,25,
.
[] E. Mariani, G. Lisignoli, R. M. Borzì, L. Pulsatelli, Int. J. Mol. Sci. 2019,
20, .
[] A. Vishwakarma, N. S. Bhise, M. B. Evangelista, J. Rouwkema, M. R.
Dokmeci, A. M. Ghaemmaghami, N. E. Vrana, A. Khademhosseini,
Trends Biotechnol. 2016,34, .
[] J. Rnjak-Kovacina, A. S. Weiss, Tissue Eng., Part B 2011,17, .
[] N. Jain, J. Moeller, V. Vogel, Annu. Rev. Biomed. Eng. 2019,21, .
[] Z. Wang, Y. Cui, J. Wang, X. Yang, Y. Wu, K. Wang, X. Gao, D. Li, Y. Li,
X.-L. Zheng, Biomaterials 2014,35, .
[] N. Jain, V. Vogel, Nat. Mater. 2018,17, .
[] A. de la Zerda, M. J. Kratochvil, N. A. Suhar, S. C. Heilshorn, APL
Bioeng. 2018,2, .
[] M. Saitakis, S. Dogniaux, C. Goudot, N. Bufi, S. Asnacios, M.
Maurin, C. Randriamampita, A. Asnacios, C. Hivroz, Elife 2017,6,
e.
[] A. Mantovani, S. Sozzani, M. Locati, P. Allavena, A. Sica, Trends Im-
munol. 2002,23, .
[] Q. Z. Zhang, W. R. Su, S. H. Shi, P. Wilder-Smith, A. P. Xiang, A. Wong,
A. L. Nguyen, C. W. Kwon, A. D. Le, Stem cells 2010,28, .
[] S. Oghumu, S. Varikuti, C. Terrazas, D. Kotov, M. W. Nasser, C. A.
Powell, R. K. Ganju, A. R. Satoskar, Immunology 2014,143,.
[] C. Atri, F. Z. Guerfali, D. Laouini, Int. J. Mol. Sci. 2018,19, .
[] A. Rodriguez, S. R. MacEwan, H. Meyerson, J. T. Kirk, J. M. Anderson,
J. Biomed. Mater. Res., Part A 2009,90A, .
[] H. Möllmann, H. M. Nef, S. Kostin, C. von Kalle, I. Pilz, M. Weber, J.
Schaper, C. W. Hamm, A. Elsässer, Cardiovasc. Res. 2006,71, .
[] E. L. Herzog, L. Chai, D. S. Krause, Blood 2003,102, .
[] S. E. Dunsmore, S. D. Shapiro, J. Clin. Invest. 2004,113, .
[] D. N. Rockwood, R. C. Preda, T. Yücel, X. Wang, M. L. Lovett, D. L.
Kaplan, Nat. Protoc. 2011,6, .
[] N. Yang, R. P. Tan, A. H. Chan, B. S. Lee, M. Santos, J. Hung, Y.
Liao,M.M.Bilek,J.Fei,S.G.Wise,ACS Biomater. Sci. Eng. 2020,6,
.
[] X. Cui, B. G. Soliman, C. R. Alcala-Orozco, J. Li, M. A. Vis, M. Santos,
S. G. Wise, R. Levato, J. Malda, T. B. Woodfield, Adv. Healthcare Mater.
2020,9, .
[] R. P. Tan, A. H. Chan, K. Lennartsson, M. M. Miravet, B. S. Lee, J.
Rnjak-Kovacina, Z. E. Clayton, J. P. Cooke, M. K. Ng, S. Patel, S. G.
Wise, Stem Cell Res. Ther. 2018,9, .
[] L. Rich, P. Whittaker, J. Morphol. Sci. 2017,22,.
Adv. Healthcare Mater. 2021, © Wiley-VCH GmbH
2100615 (14 of 14)
A preview of this full-text is provided by Wiley.
Content available from Advanced Healthcare Materials
This content is subject to copyright. Terms and conditions apply.