JID: ACTBIO

ARTICLE IN PRESS [m5G;August 2, 2021;15:11]

Acta Biomaterialia xxx (xxxx) xxx

Contents lists available at ScienceDirect

Acta Biomaterialia
journal homepage: www.elsevier.com/locate/actbio

Full length article

Evaluation of the in vitro cytotoxicity and modulation of the

inflammatory response by the bioresorbable polymers
poly(D,L-lactide-co-glycolide) and poly(L-lactide-co-glycolide)
Lucy Geddes a, Efrosyni Themistou b, James F. Burrows c, Fraser J. Buchanan a,
Louise Carson c,∗
a
School of Mechanical and Aerospace Engineering, Queens University Belfast, Ashby Building, Belfast, BT9 5AG, Northern Ireland, UK
b
School of Chemistry and Chemical Engineering, Queens University Belfast, David Keir Building, Belfast, BT9 5AG, Northern Ireland, UK.
c
School of Pharmacy, Queens University Belfast, Belfast, BT9 7BL, Northern Ireland, UK

a r t i c l e i n f o a b s t r a c t

Article history: Bioresorbable polymers composed of poly(D,L-lactide-co-glycolide) (PDLLGA) and poly(L-lactide-co-

Received 3 December 2020 glycolide) (PLLGA) have become increasingly popular for the preparation of bone substitute constructs.
Revised 21 July 2021
However, there are reports of a delayed inflammatory reaction occurring months or years after implanta-
Accepted 22 July 2021
tion. Due to the long polymer degradation times, in vitro tests carried out at physiological temperature,
Available online xxx
37°C, tend to assess only the short-term biocompatibility of these materials. The aim of this work is to de-
Keywords: velop an in vitro protocol that can be used to assess the long-term cytotoxicity of bioresorbable polymers
Poly(lactide-co-glycolide) in a time efficient manner. This study used a previously developed and validated accelerated degradation
Biocompatibility protocol to obtain samples of PDLLGA and PLLGA at increasing levels of degradation. Samples were then
Cytotoxicity applied to standard ISO 10993-5 direct contact cytotoxicity testing and it was found that PDLLGA samples
Quasi Vivo showed increasing levels of cytotoxicity at the later stages of degradation, with PLLGA samples demon-
Lactate
strating significantly less cytotoxic behaviour. Following concern that accumulation of acidic degradation
Hydroxycarboxylic acid receptor 1
HCA1
products in a closed multi-well culture environment could overestimate cytotoxicity, we developed and
validated a new dynamic flow culture methodology, for testing the cytotoxicity of these degradable ma-
terials, by adapting a commercial “organ on a chip” flow culture system, Quasi Vivo®. In addition to
cytotoxicity testing, we have carried out profiling of inflammatory cytokines released by cells in response
to degraded PDLLGA and PLLGA, and have suggested mechanism by which lactide-based bioresorbable
materials could modulate the inflammatory response through the G-protein coupled receptor (GPCR), hy-
droxycarboxylic acid receptor 1 (HCA1 ).

Statement of significance

Bioresorbable materials naturally disintegrate over time when implanted into the body. They are often
used to make screws and clips for repair of broken bones. Unfortunately, some patients can react badly
to the material, resulting in painful inflammation. Biomaterials scientists are interested in developing
materials that are more compatible with the body. However, it is very difficult to predict the long-term
compatibility of bioresorbable materials in the lab. In our study, we have developed a method that will
allow us to study the effects of the materials as they continue to break down. This will help us under-
stand why the materials may cause inflammation, and will support research into the development of new
and improved materials for bone repair.
© 2021 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Bioresorbable polymers composed of poly(lactide) (PLA),

∗
Corresponding author. poly(glycolide) (PGA) and their related copolymers are commonly
E-mail address: l.carson@qub.ac.uk (L. Carson). used in orthopaedic applications to help regenerate and repair

https://doi.org/10.1016/j.actbio.2021.07.049
1742-7061/© 2021 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Please cite this article as: L. Geddes, E. Themistou, J.F. Burrows et al., Evaluation of the in vitro cytotoxicity and modulation of the
inflammatory response by the bioresorbable polymers poly(D,L-lactide-co-glycolide) and poly(L-lactide-co-glycolide), Acta Biomaterialia,
https://doi.org/10.1016/j.actbio.2021.07.049
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
injured bone tissue. Upon implantation, polymers degrade via
hydrolytic degradation, gradually transferring the load to the bone
tissue, with metabolic pathways used to excrete the degradation
products from the body [1–3]. The main benefit of bioresorbable
polymers is the ability to tailor their physicochemical and me-
chanical properties to meet the desired needs of the application.
Poly(L-lactide) (PLLA) is commonly used for orthopaedic appli-
cations, however, reports of long polymer degradation times,
in excess of 5 years, has led to research into faster degrading
polymers, allowing replacement by regenerating tissue over more
realistic timeframes [4–8]. There is growing interest into the
copolymerisation of poly(D,L-lactide) (PDLLA) and PLLA with PGA
forming poly(D,L-lactide-co-glycolide) (PDLLGA) and poly(L-lactide-
co-glycolide) (PLLGA). The inherent properties of the amorphous
PDLLGA and semi-crystalline PLLGA, give the copolymers different
characteristics and their mechanical, physiochemical and degrada-
tion properties can be controlled by altering their monomer unit
ratios.
The in vivo host response to implanted biomaterials is impor-
tant to be considered during the development of bioresorbable
polymers. Upon implantation, a host reaction is triggered and re-
sults in a cascade of regulated events. Following implantation, a
provisional matrix is formed by protein adsorption on the sur-
face of the biomaterial and is rich in cytokines, chemoattractants
and growth factors. Acute inflammation follows and is charac-
terised by the presence of neutrophils, whose role involves the
removal of cellular debris via phagocytosis and the expression of
cytokines, which are involved in the recruitment of macrophages,
such as macrophage inflammatory protein-1α (MIP-1α ) and mono-
cyte chemoattractant protein-1 (MCP-1) [9–13]. This phase is fol-
lowed by chronic inflammation, where inflammatory cells, such as
monocytes, are recruited to the site of implantation. The acute and
chronic inflammation phases are relatively short, usually lasting no
longer than two weeks. The resolution of these phases leads to
the formation of granulation tissue, which can be identified by the
presence of macrophages, fibroblasts, and the proliferation of small
blood vessels. The foreign body reaction (FBR) is identified by the
presence of macrophages and foreign body giant cells at the sur-
face of the implant, together with the components of the gran-
ulation tissue. The cellular composition of the FBR is largely de-
termined by the form, topography, and surface-to-volume ratio of
the implanted biomaterial and can persist for its lifetime. The final
stage of the healing process is fibrous capsule formation surround-
ing the biomaterial. This acts as protective barrier isolating the FBR
and the biomaterial from the host tissue [13,14]. This is the body’s
natural healing mechanism and is essential for the successful in-
tegration of the implanted biomaterial with its host environment.
The progression of these stages occurs via complex cellular inter-
actions, with crosstalk between different cascade systems control-
ling the cellular response and behaviour. Development of a better
understanding of these complex interactions will enhance the pro-
gression of suitable bioresorbable polymers for use as medical de-
vices.
Biocompatibility is used to describe the desired interaction of
a biomaterial with its host environment, whereby the biomate-
rial or degradation products should not cause any adverse reac-
tions or cytotoxicity after implantation [15]. PLA, PGA, and their
related copolymers are widely considered biocompatible, however,
throughout the literature there is evidence of a delayed inflamma-
tory reaction occurring months or years after implantation. These
reactions generally present as a non-specific FBR and can range
in severity [5,16–20]. The occurrence of this reaction is not well
understood as it is not specific to polymer type, implant loca-
tion or subject [5]. Numerous reasons for the reported incom-
patibility have been hypothesised, which include the presence of
crystalline particle debris, the accumulation of degradation prod-
2
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
ucts, the acidic environment created by the release of acidic by-
products, the geometry of the implant and the implant location
[21–24]. However, it is difficult to determine the aetiology of this
response as it is difficult to isolate the factors that contribute to
this response.
Medical devices must undergo a series of tests to ensure they
are safe for clinical use, with these tests regulated by the Interna-
tional Organisation for Standardisation (ISO) [25,26]. Cytotoxicity
analysis is required for every medical device and is generally the
first step of in vitro biocompatibility testing. These tests provide a
highly effective method to determine the cytotoxicity of a material
or substance on living cells, however, their use is limited as they
cannot be used to determine the cause of cellular death. Testing
of the in vitro immune response at the early stages of biomaterial
development is currently not incorporated within standard proce-
dures. Gaining insight into the in vitro cell response to biomaterials
would allow for early detection and prediction of potential adverse
reactions. Protein-based arrays can be used as a useful method to
evaluate how biomaterials interact with different cells within the
FBR [27]. The profiling of cytokines expressed by cells in contact
with biomaterials can be determined using antibody arrays and an
enzyme-linked immunosorbent assay (ELISA).
In vitro studies play an important role in the prediction of in
vivo behaviour by allowing rapid screening of biomaterials prior to
animal trials. However, in vitro tests tend to assess the short-term
effects, with the long-term biocompatibility assessed using in vivo
trials. The development of extended in vitro biocompatibility test-
ing for degradable biomaterials would be beneficial, as it would
allow the early identification of potential issues before progressing
to animal or clinical studies and would also advance the under-
standing of why adverse reactions occur. Despite clinical studies
demonstrating adverse reactions to bioresorbable polymers, there
are few in vitro studies that corroborate these results. The aim of
this work is to develop an in vitro protocol that can be used to as-
sess the cytotoxicity of bioresorbable polymers in a time efficient
manner.
Current in vitro ISO standard tests for cytotoxicity are carried
out at 37°C in cell culture media (or other suitable diluent), in
closed, static multi-well culture vessels. These tests only assess the
short-term biocompatibility, whereas late-stage degradation, can
take months to occur at physiological temperature. This timescale
is incompatible with cell culture-based assays.
In this study, we have used a previously developed and vali-
dated accelerated degradation protocol to obtain polymer samples
at increasing levels of degradation [28]. This utilised an elevated
temperature of 47°C, i.e. 10°C above physiological temperature.
This previous study also compared elevated temperatures of 57°C
and 70°C and concluded 47°C was the optimum temperature owing
to the fact that it was below the polymer’s Tg, and produced con-
sistent results over practicable timescales. Briefly, polymers were
incubated in sterile phosphate buffered saline at 47°C, and samples
were removed at predetermined time points, informed by physico-
chemical characterisation. The samples were then applied to stan-
dard ISO 10993-5 direct contact cytotoxicity testing [29]. Addition-
ally, we have developed a dynamic flow culture methodology for
testing these degradable materials, following concern that accumu-
lation of acidic degradation products in closed multi-well culture
could overestimate cytotoxicity. In an in vivo physiological environ-
ment, diffusion of degradation products will occur, and there will
be an exchange of tissue fluid in the peri-implant microenviron-
ment. To replicate the in vivo environment and increase physiolog-
ical relevance, we have adapted a commercial “organ on a chip”
flow culture system, Quasi Vivo®, for testing the polymer samples.
In addition to cytotoxicity testing, we have carried out profil-
ing of inflammatory cytokines released by cells in response to de-
graded PDLLGA and PLLGA, and have proposed a suggested mecha-
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
nism by which lactide-based bioresorbable materials could module
the inflammatory response through the G-protein coupled receptor
(GPCR), Hydroxycarboxylic acid receptor 1 (HCA1 ).
2. Materials and methods
2.1. Materials
PDLLGA 85:15 (PURASORB PDLG 8531, Corbion Biomaterials, the
Netherlands) and PLLGA 85:15 (PURASORB PLG 8531, Corbion Bio-
materials, the Netherlands) in pellet form were used for this study.
Materials were stored in a freezer (-20°C) in sealed packaging until
use.
2.2. Sample preparation
PDLLGA and PLLGA pellets were compression moulded (Collins
P200P), at 180°C for 6.5 minutes, into sheets with a volume of
100 × 100 × 1 mm3 . PLLGA compression moulded sheets were
then annealed in an oven for 4 hours at 100°C (PDLLGA was not
annealed, as this was considered unnecessary due to its amor-
phous nature). A laser cutting machine (FB 1800 50W) was used
to cut the PDLLGA and PLLGA sheets into disc-shaped samples with
an 8mm diameter. Samples were sterilised using electron beam ra-
diation (Steris, Ireland), with a dosage of 20 kGy. For experimen-
tation under static cell culture conditions positive control mate-
rial was prepared by swelling poly(2-hydroxyethyl methacrylate)
(pHEMA) in a solution of 0.1% benzalkonium chloride (BAK) in
phosphate buffered saline (PBS), and sterilising by autoclave. BAK
is a known cytotoxic substance and hydrogel polymers soaked in
BAK have been validated as a suitable positive control material for
cytotoxicity testing [30]. The negative controls were non-degraded
PDLLGA samples and non-degraded PLLGA samples.
2.3. In vitro degradation
PDLLGA and PLLGA samples were separated into three groups,
the first group was used for characterisation, the second for
tests on L929 murine fibroblasts and the third for tests on
RAW264.7 murine macrophages. Under sterile conditions, samples
were placed into transwell inserts (Millicell, PIHP01250), immersed
in 10 mL of PBS buffer (pH 7.4) and degraded in an oven at 47°C.
PDLLGA sample were degraded for 5, 7, 10 and 12 days and PLLGA
was degraded for 28, 42, 56 and 70 days.
2.4. Characterisation
Polymers were characterised in a previous study by the au-
thors [28]. PDLLGA and PLLGA were assessed for changes in the
pH of the degradation media (PBS buffer), percentage mass loss,
percentage swelling and molecular weight loss during degradation
at 47°C. Briefly, the pH of the degradation media (PBS buffer) was
recorded at each time point. Prior to degradation, the initial mass
of the samples was recorded. At each time point the samples (n=5)
were retrieved from the PBS buffer and rinsed with distilled wa-
ter. Surface moisture was removed, and the weight was recorded
(wet mass). The dry mass was then determined after samples were
dried in a vacuum oven for 72 hours at 30°C and 600 mmHg.
Eqs. 1 and 2 were used to determine the percentage mass loss and
swelling of the samples, respectively.
m0 − md
% Mass Loss = x100 (1)
m0
mw − md
% Swel l ing = x100 (2)
md
Where, m0 = initial mass; md = dry mass and mw = wet mass
3
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
The number average molecular weight (Mn ) of the samples was
determined by gel permeation chromatography (GPC).
2.5. Cell culture
L-929 murine fibroblasts (ATCC CCL-1) were maintained in Ea-
gle’s minimum essential medium (MEM) containing 10% foetal
bovine serum (FBS) and 50 U/mL penicillin/streptomycin (pen
strep). RAW264.7 (ATCC TIB-71) murine macrophages were main-
tained in Dulbecco’s minimum essential medium (DMEM) contain-
ing 10% FBS and 50 U/mL pen strep. Both cell lines were cultured
under standard conditions of 37°C and 5% CO2 .
2.6. Cell viability assay
2.6.1. Static conditions
Cells were cultured in a 24-well plate for 24 hours in 1 mL
of media to allow for a near confluent monolayer. The degraded
samples, in the transwell inserts, were placed into the wells of
the tissue culture plate, and cells incubated for a further 24 hours
at 37°C. After this time, the transwell inserts were removed from
the 24-well plate. The supernatants were removed from the cells,
placed into microcentrifuge tubes, and stored at -80°C for later
analysis (i.e. antibody arrays and ELISA analysis described below).
500 μL of fresh MEM or DMEM were then added to the cells. All
experimental steps were carried out under aseptic conditions in a
Class II microbiological safety cabinet. All cell culture experiments
were carried out in triplicate.
2.6.2. Development of flow conditions for cell culture using quasi
vivo® QV500 system
The Quasi Vivo® QV500 system (Kirkstall) is a flow system
that can be used to culture cells in a way that better reflects the
dynamic nature of the physiological environment. This approach
utilises interconnected chambers with cell culture media pumped
through the system, allowing cells to be cultured under flow condi-
tions. The components of the QV500 system were connected under
sterile conditions in a Class II microbiological safety cabinet. The
system is designed to recirculate media; however, this was adapted
for the current study to allow for a single pass of media though the
system. Nine Quasi Vivo® chambers were connected in series and
each sample examined in triplicate (Fig. 1). A peristaltic pump (SP-
minipump compact, Shenchen Baoding) with variable speed con-
trol was used with the QV500 system. The pump was calibrated
in order to achieve a flowrate of 75 μL/min; this flowrate was se-
lected as it was used by Nithiananthan et al. [31] when using the
QV500 system with fibroblast cell lines.
2.6.2.1. Flow conditions using quasi vivo® QV500 system. Prior to
use, glass coverslips (13 mm diameter) were sterilised in ethanol,
rinsed in PBS, primed in MEM, and placed into a 24-well plate.
L929 fibroblasts were cultured, in static conditions, on coverslips
in 1 mL of MEM at 37°C with 5% CO2 , for 24 hours. As per man-
ufacturer’s instructions, before starting the experiment, the QV500
system was washed with PBS buffer for 20 minutes, then primed
with cell culture media (MEM) for 20 minutes, with the peristaltic
pump set to its maximum flowrate. The media was then removed
from the system. Fig. 2 shows step-by-step images of the next
stages of the experiment, wherein material samples are loaded
into each QV500 chamber. The closed system is then transferred
to the incubator and operated with a flowrate of 75 μL/min for 24
hours. After this time, coverslips were removed from the system
and placed into a 24-well plate, with 500 μL of fresh MEM added,
in preparation for the MTT cytotoxicity assay.
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
Fig. 1. Schematic representation of QV500 system (A) and real image of QV500 system (B).
Fig. 2. Step-by-step images of Quasi-Vivo® system set-up. The empty chamber (A) is opened and a coverslip containing cultured L929 fibroblasts is placed into the chamber
(B). A transwell insert, either empty (cells only) or containing the negative control or sample is then placed on top of the cells (C), 1 mL of MEM is added (D) and the
chamber is closed (E). This is repeated until all chambers contain samples.
2.6.3. MTT cytotoxicity assay
The viability of the cells is determined by the ability of the
mitochondrial enzyme succinate dehydrogenase to metabolically
reduce the yellow water soluble 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazoliumbromid (MTT) into blue/ violet formazan crys-
tals. The intensity of the blue/ violet colour is directly related to
the number of viable cells [29,32]. Following the treatments de-
scribed in Sections 2.6.1 and 2.6.2, 100 μL of a filter-sterilised MTT
(5 mg/mL) solution was added to the cells in the 24-well plates.
The plates were wrapped in tin foil to protect from light and
placed in the incubator at 37°C for 2 hours. Media was then aspi-
rated, and the formazan crystals formed were dissolved in 500 μL
of dimethyl sulfoxide (DMSO). The optical density was measured at
a wavelength of 570 nm using a FLUOstar Omega microplate reader
(BMG Labtech). Eq. 3 was used to monitor the changes in cell vi-
ability. As stated by ISO 10993-5, if the cell viability is less than
70% the material has cytotoxic potential. The lower the percentage
viability the greater the cytotoxic potential of the material. To cal-
culate the percentage viability all the sample values are relative to
a “cells only” blank control, as per Equation 3.
ODs
Cel l V iabil ity (% ) = × 100%
ODb
where: ODs is the optical density of the test sample extracts and
ODb is the optical density of the “cells only” blanks.
2.7. Antibody array
The relative expression levels of 40 mouse cytokines from
selected samples were determined using a Proteome ProfilerTM
Mouse Cytokine Array, Panel A (R&D Systems). Four samples of cell
conditioned media were selected based on the results of the MTT
cytotoxicity assay, in Section 3.2. These included the supernatants
collected from L929 fibroblasts and RAW264.7 macrophages after
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
contact with PDLLGA degraded for 10 days and PLLGA degraded
for 42 days. These time points were selected at which the mate-
rials began to show slight reduction in cell viability. The controls
used for this study were the supernatants collected from the L929
fibroblasts and RAW264.7 macrophages cultured in cell medium
only. Supernatants from triplicate samples were pooled and pre-
pared according to the manufacturer’s array procedure. The cy-
tokine array was conducted according to the manufacturer’s in-
structions and chemiluminescence was detected using the Molec-
ular Imager® ChemiDocTM XRS+ imaging system (BioRad). Images
were analysed using Image Studio Lite version 5.2 software.
2.8. ELISA
MCP-1 and MIP-1α , were selected for continued analysis using
the DuoSet® ELISA Development System (R&D Systems), based on
the results of the antibody array, in Section 3.5. Cell conditioned
media, collected under static conditions (Section 2.6.1), were anal-
ysed. These samples included the supernatants collected from L929
fibroblasts and RAW264.7 macrophages in contact with PDLLGA
(3) non-degraded and degraded for 7, 10 and 12 days at 47°C and
PLLGA non-degraded and degraded for 28, 42, 56 and 70 days at
47°C. The controls used were the supernatants collected from the
L929 fibroblasts and RAW264.7 macrophages cultured in cell cul-
ture medium only. The samples collected from the L929 fibrob-
lasts were analysed for MCP-1 and the samples collected from the
RAW264.7 macrophages were analysed for MIP-1α . The ELISA was
carried out according to the manufacturer’s instructions. All ELISA
analysis was performed in duplicate wells, on three experimental
replicates. The optical density was determined using a FLUOstar
Omega microplate reader (BMG Labtech), measured at wavelengths
of 450 nm and 570 nm. Wavelength correction was calculated by
subtraction of the reading at 570 nm from the reading at 450 nm.
4
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
The calibration curve was generated using four-parameter logistic
(4-PL) curve fit.
2.9. GPCR (HCAR1 ) activation and calcium mobilisation assays
GPCR activation was monitored via calcium signalling using the
proprietary Fluo-4 Direct TM Calcium Assay Kit. This kit contains
a Ca2+ indicator, which is loaded into the cells and produces a
large fluorescence increase upon binding Ca2+ . To prevent extru-
sion of the indicator out of the cell by organic anion transporters,
probenecid is included in the kit formulation, and this helps to
reduce the baseline signal. Media removal and washing steps are
not required with the Fluo-4 Direct TM Calcium Assay Kit, as back-
ground fluorescence attributable to the growth media is eliminated
by the addition of a suppression dye.
Cells were seeded at a density of 5 × 104 cells/well in black
sterile 96-well plates with optical bottom, in 100 μl of the ap-
propriate culture media, and allowed to recover for 24 hours. The
Fluo-4 Direct TM indicator was reconstituted as per kit instructions
under aseptic conditions and added to each well in an equal vol-
ume to culture media. The plate was wrapped in aluminium foil
to protect from light, then incubated for 30 minutes at 37°C, and a
further 30 minutes at room temperature.
To confirm GPCR receptor activation in response to lactate (and
the presence of HCA1 ) 15 μM sodium L-lactate was added to in-
dicator loaded cells, and the fluorescence was monitored using a
BMG Fluostar Optima® spectrofluorimeter, at wavelengths of 494
nm excitation and 516 nm emission. The response to 10 μL samples
of the polymer degradation media (as described in Section 2.3)
was also evaluated, with degradation time points of 12 days for
PDLLGA, and 70 days for PLLGA selected for examination. This anal-
ysis was carried out in triplicate. Sterile, calcium free PBS was used
as a vehicle control and was not found to have any effect on recep-
tor activation (Supplementary Information, Figure S1).
2.10. Statistics
Statistical analysis was carried out using GraphPad Prism 5.0
software. The results are expressed as the mean ± standard devi-
ation. Statistical analysis for the static conditions MTT assay and
ELISA was performed using a one-way ANOVA with a post-hoc
Tukey analysis. Statistical significance was set at p < 0.05. Statisti-
cal analysis for the Quasi Vivo flow conditions was performed us-
ing a t-test. Statistical significance was set at p < 0.05.
3. Results
3.1. Visual inspection
PDLLGA underwent significant changes to its appearance dur-
ing degradation and changes occurred without considerable mass
loss, where, after 12 days, only 6% mass was lost from the samples
(Fig. 3a). After 5 days at 47°C the samples became visibly damaged,
after 10 days the samples appeared visibly swollen with a hard
exterior and viscous liquid interior, and after 12 days the sam-
ples were a viscous liquid. In contrast, the appearance of PLLGA
(Fig. 3b) remained the same until 56 days at 47°C, where the sam-
ples changed from a translucent off-white colour to opaque and
bright white. Throughout degradation, the samples appeared volu-
metrically unchanged, however, when mass loss occurred, after 56
days, the samples became fragile and disintegrated into a powder
on handling.
3.2. MTT cytotoxicity assay for static conditions
The direct cytotoxicity assay (conducted as per ISO 10993-5) on
pre-degraded PDLLGA (Fig. 4a) showed similar trends in both fi-
5
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
broblasts and macrophages. High cell viability values, between 86
– 106%, were obtained in both cell lines for the 5- and 7- days
samples, which showed no significant difference to the negative
control. However, a significant decrease compared to the negative
control, was observed after 10 days, as viability was reduced to
53% and 14% in the fibroblasts and macrophages, respectively. Af-
ter 12 days, viability was reduced to 0% in both cell lines, therefore
demonstrating a high degree of cytotoxicity. When PLLGA degraded
for 42- and 70- days was placed in contact with the fibroblasts,
cell viability reduced to 58% and 63%, respectively (Fig. 4b). These
values are below the 70% viability threshold that is stated by ISO
10993-5, below which the material displays cytotoxic potential. In
contrast, when PLLGA was placed onto the macrophages (Fig. 4b),
there was no cytotoxic effect observed at any time point, with high
cell viability values between 84–112%.
3.3. Cytotoxicity and polymer degradation
The degradation characteristics (change in pH of the degrada-
tion media, percentage mass loss, percentage swelling and change
in the number averaged molecular weight (Mn ) value) of PDLLGA
and PLLGA, degraded at 47°C, were measured [28] and compared
to the MTT assay results in Fig. 4. PDLLGA and PLLGA degrade via
a mechanism of bulk degradation, which can be observed when
degradation proceeds in two distinct phases; firstly, a reduction in
molecular weight, followed by mass loss. Fig. 5 shows that after 5
days, there was a significant reduction in the Mn value of PDLLGA
from 27 800 g/mol to 3380 g/mol indicating significant change in
the degree of polymerization of the polymer and break down of
its chains to oligomers and monomers, with a mass loss of only 6%
after 12 days. There was a slight decrease in the pH of the degra-
dation media from pH 7.4 to 6.3 after 12 days. However, there was
significant polymer swelling after 10 days (97%) and 12 days (57%),
which corresponds to the reduction in cell viability in both the fi-
broblasts and macrophages at these time points. Fig. 6 shows that
the Mn of PLLGA decreased steadily with degradation time from
135 0 0 0 g/mol to 22 70 0 g/mol after 42 days and to 20 0 0 g/mol af-
ter 56 days, indicating that the high molecular weight chains have
almost completely broken down to shorter chains. There was ini-
tially a small decrease in mass, 2%, after 42 days, followed by a
larger reduction in mass, 19%, after 56 days, and a further reduc-
tion to 38% after 70 days. There was a slight decrease in the pH
of the degradation media from pH 7.4 to 6.0 after 42 days, and
a significant drop to pH 2.8 after 56 days. Swelling in PLLGA did
not occur immediately but increased with degradation time with a
swelling ratio value of 43% after 70 days, with a corresponding cell
viability of 63% in the L929 fibroblasts and 103% in the RAW264.7
macrophages.
3.4. MTT cytotoxicity assay using quasi vivo® flow conditions
Fig. 7 compares the viability of fibroblasts in contact with
PDLLGA degraded for 10 days and PLLGA degraded for 42 days, un-
der static and flow conditions using the Quasi Vivo® QV500 sys-
tem. The samples were selected based on the results of the MTT
assay, in Section 3.2, as after 10 days and 42 days, PDLLGA and
PLLGA, respectively, began to show cytotoxic potential towards the
fibroblasts. Fig. 7a shows that PDLLGA was cytotoxic after 10 days
using both static and flow conditions, with cell viability reduced
to 46% and 16%, respectively. Significant differences were observed
between the control and 10 days sample for both sets of condi-
tions. PLLGA showed cytotoxic potential, under static conditions,
after 42 days, with cell viability reduced to 58% (Fig. 7b). How-
ever, cell viability was improved to 69% using the Quasi Vivo flow
conditions and the results showed no significant difference to the
control, suggesting that after 42 days degradation, the material
JID: ACTBIO
ARTICLE IN PRESS [m5G;August 2, 2021;15:11]

L. Geddes, E. Themistou, J.F. Burrows et al. Acta Biomaterialia xxx (xxxx) xxx

Fig. 3. Samples appearance and mass loss of a) PDLLGA and b) PLLGA after degradation at 47°C and drying.

Fig. 4. Cell viability of L929 fibroblasts and RAW264.7 macrophages versus degradation time of a) PDLLGA and b) PLLGA at 47°C. (∗ indicates significant difference to the
negative control (p<0.05), red line represents ISO 10993-5 cytotoxicity threshold, at 70% viability relative to blank control (cells only); Percentage cell viability values of the
samples may be greater than 100% in cases where the sample cells proliferate more than the “cells only” blank control).

showed less potential to induce cytotoxic effects under flow condi-
tions than those observed under the standard static experimental
conditions.
3.5. Antibody array analysis
Fig. 8a shows the relative expression levels of the inflam-
matory cytokines detected in conditioned media from fibrob-
lasts and macrophages after contact with PDLLGA degraded for
10 days. For fibroblasts, nine cytokines were detected in the
control, including MCP-1, tissue inhibitor of metalloproteinase-
1 (TIMP-1), macrophage colony-stimulating factor (M-CSF), stro-
mal cell-derived factor-1 (SDF-1), keratinocyte chemoattractant
(KC), interleukin-1α (IL-1α ), interferon-inducible T cell alpha
chemoattractant (I-TAC), tumour necrosis factor-α (TNF-α ), inter-
feron gamma (IFN-γ ), and eleven cytokines were detected in
the treated samples, with the addition of interleukin-17 (IL-17)
and interleukin-7 (IL-7) to those already observed in the con-
trol. The results showed a downregulation in the relative ex-
pression levels of MCP-1 and KC in the treated samples. For
the macrophages, eight cytokines were detected, which included
MIP-1α , macrophage inflammatory protein-1β (MIP-1β ), SDF-
1, interferon-inducible protein-10 (IP-10), interleukin-1ra (IL-1ra),
MCP-1, TNF-α and C-X-C motif chemokine ligand 9 (CXCL9). How-
ever, only MIP-1α and SDF-1 were detected in the treated sam-
ples. The results shows a pronounced downregulation of the rel-
ative expression levels of MIP-1α , MIP-1β , IP-10, IL-1ra, MCP-
1 and TNF-α . Fig. 8b shows the relative expression levels of
the cytokines released from the fibroblasts and macrophages af-
ter contact with PLLGA degraded for 42 days. In the fibroblasts,
four cytokines were detected in the control and treated sam-
ples, these include MCP-1, TIMP-1, M-CSF and SDF-1. The re-
sults show an upregulation of MCP-1, a downregulation of TIMP-
1 and SDF and no change in M-CSF in the treated samples. In
the macrophages, four cytokines were also detected in the con-
trol and treated samples, which include MIP-1α , SDF-1, IL-1ra and
MIP-1β . The results show that there was an increase in the rel-
ative expression levels of MIP-1α , SDF-1, IL-1ra and MIP-1β in
the treated samples compared to the control. In both PDLLGA
and PLLGA, MCP-1 was the most abundant cytokine in the fi-
broblasts and MIP-1α was the most abundant cytokine in the
macrophages.

6
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
Fig. 5. Relationship of cell viability and degradation time with a) pH of degradation media, b) Mass Loss, c) Swelling and d) Mn of PDLLGA.
3.6. ELISA analysis
Fig. 9 shows the concentrations of the cytokines MCP-1 and
MIP-1α , in the fibroblast and macrophage conditioned media, re-
spectively, after contact with PDLLGA and PLLGA. Concentrations
of the cytokines MCP-1 and MIP-1α increased when in contact
with non-degraded PDLLGA and PDLLGA degraded for 7 days, in
both cell lines (Fig. 9a). However, after 10 days, the concentration
of the cytokines decreased. The concentration of MCP-1 in the fi-
broblasts decreased by 49% after 10 days and 28% after 12 days
relative to the control. The concentration of MIP-1α decreased by
88% and 89% compared to the control, after 10 and 12 days, re-
spectively. Statistical analysis showed that there was no significant
change in the concentration of MCP-1 during the experiment, but
there was a significant difference in the concentration of MIP-1α
in the macrophages after 10 and 12 days.
Somewhat different trends were observed when the fibroblasts
and macrophages were in contact with PLLGA (Fig. 9b). The con-
centration of MCP-1 in the fibroblasts decreased with degradation
time (Fig. 9bi). There was no significant difference in the concen-
tration of MCP-1 in the non-degraded PLLGA sample compared to
the control, however, significant differences were observed after
28, 42, 56 and 70 days. There was a 42% reduction in the concen-
tration of MCP-1 after 28 days and a 60% reduction observed after
70 days. There was no statistically significant difference in the con-
centration of MIP-1α in the macrophages after contact with PLLGA
at any time point (Fig. 9bii). However, elevated levels of MIP-1α
7
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
were observed after 28 days and 70 days, with values increasing
by 240% and 383%, respectively, compared to the control.
3.7. GPCR (HCAR1 ) activation and calcium mobilisation assays
analysis
When a ligand binds to its G-protein coupled receptor (GPCR),
it undergoes a conformational change, which facilitates the acti-
vation of specific intracellular pathways. Guanosine diphosphate
(GDP) bound to the Gα protein subunit is replaced by guanosine
triphosphate (GTP). This initiates the dissociation of the Gα sub-
unit from the Gβγ subunit, which in turn opens calcium ion chan-
nels and a surge of calcium is released into the cytoplasm [33].
The calcium mobilisation assay detects this influx of calcium using
a fluorescent dye (Fluo-4).
This assay was used to determine whether GPCRs of the fibrob-
lasts and macrophages, are activated in the presence of the degra-
dation products released from PDLLGA and PLLGA during degrada-
tion in PBS buffer at 47°C. Observation of receptor activation would
strongly point towards activation of the hydroxycarboxylic acid re-
ceptor 1 (HCA1 ), the agonist for which is lactic acid/lactate, and
this is the only constituent of the degradation products known to
activate GPCRs. Receptor activation is observed, in Fig. 10, upon the
addition of the degradation media of PDLLGA and PLLGA degraded
for 12- and 70- days, respectively. The curves show a sharp in-
crease in fluorescence, which gradually declines after reaching a
maximum. This is characteristic of GPCR activation [33]. Control
JID: ACTBIO
ARTICLE IN PRESS [m5G;August 2, 2021;15:11]

L. Geddes, E. Themistou, J.F. Burrows et al. Acta Biomaterialia xxx (xxxx) xxx

Fig. 6. Relationship of cell viability and degradation time with a) pH of degradation media, b) Mass Loss, c) Swelling and d) Mn of PLLGA.

Fig. 7. Comparison of the cell viability of L929 fibroblasts using static conditions compared to Quasi Vivo flow conditions when in contact with a) PDLLGA degraded for 10
days and b) PLLGA degraded for 42 days. (∗ indicates significant difference to the control p<0.05, Percentage viability is calculated relative to blank controls (cells only);
Percentage cell viability values of the samples may be greater than 100% in cases where the sample cells proliferate more than the “cells only” blank control).

8
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
Fig. 8. Relative expression levels of inflammatory cytokines detected in the supernatants collected from L929 fibroblasts and RAW264.7 macrophages after contact with
PDLLGA pre-degraded for 10 days and PLLGA pre-degraded for 42 days, and controls collected from the supernatants of L929 fibroblasts and RAW264.7 macrophages that
had no contact with the polymers.
curves for PBS (blank) and sodium L-lactate can be found in Sup-
plementary Information (Figure S1).
4. Discussion
4.1. Correlating cytotoxicity with degradation behaviour
The biocompatibility of bioresorbable polymers is generally
tested using short-term in vitro cytotoxicity tests before proceeding
to in vivo trials [34]. These studies have shown polymers composed
of lactide and glycolide demonstrate satisfactory results [35,36].
Yang et al. [37] carried out an MTT assay on L929 fibroblasts in-
cubated for up to 7 days with extract medium prepared using
PDLLGA (70:30). The results showed no cytotoxic effects at any
time point. However, this study did not account for how poly-
9
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
mer cytotoxicity may change during a more prolonged degradation
timeframe. Therefore, long-term cytotoxicity studies are needed to
assess the effects of these polymers and their degradation products
in the physiological environment. Previously, Sung et al. [38] com-
pared the biocompatibility of the fast degrading PDLLGA to slow
degrading poly(ε -caprolactone) (PCL), by culturing mouse aortic
smooth muscle cells on PDLLGA and PCL polymer scaffolds and as-
sessed the in vitro cell viability after degradation for 1, 7, 14,21 and
28 days at 37°C. The results showed that the cell viability in the
fast-degrading PDLLGA was lower than that of the slow degrading
PCL. It was suggested that the reduction in cell viability was due to
the increase in acidification of the local environment during poly-
mer degradation. Despite the results showing differences in the cy-
totoxicity of the two polymers, the study did not consider that the
fast-degrading PDLLGA was significantly more advanced in terms of
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
Fig. 9. Concentration of MCP-1 and MIP-1α in the supernatants collected from L929 fibroblasts and RAW264.7 macrophages in contact with PDLLGA and PLLGA after
degradation at 47°C. (∗ indicates significant difference to the control).
the relative course of its degradation, than the slow degrading PCL
after the 28 day period. For a true comparison, it is important to
evaluate the polymers at similar levels of degradation, considering
for example, mass and molecular weight loss. However, the slow
polymer degradation times at 37°C can make it difficult and im-
practical to obtain polymer samples for study at adequate extents
of degradation. One method to achieve this is to degrade the poly-
mers at an increased temperature. Ignatius et al. [39], evaluated
the biocompatibility of poly(D,L lactide) (PDLLA) 70:30 and PLLGA
90:10 using a series of in vitro tests, including an MTT assay on
BALB 3T3 cells. The MTT assay was performed using the extracts
of PDLLA and PLLGA prepared in PBS buffer for 10 days at 37°C
and 70°C. The extracts prepared at 70°C were used to simulate the
degradation products released after the long-term degradation of
the polymers. To exclude pH concerns during the assay and to fo-
cus on the degradation products, samples were neutralised to pH
7.4. The results of the assay found that cell proliferation was sig-
nificantly reduced when in contact with high concentrations of the
extracts prepared at 70°C. Although the conclusions of the study
found both polymers showed acceptable biocompatibility overall,
it was suggested that high concentrations of degradation products
were toxic to the cell lines.
In vitro cytotoxicity tests are carried out by placing the test
material or extract solution directly onto the cultured cells [34].
Extract tests are useful in determining the cytotoxicity of leach-
able substances released from the biomaterial, however, they do
10
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
not provide any information on how a material interacts directly
with the cells. To the best of the authors knowledge, no in vitro
study has demonstrated how the cytotoxicity of polymers com-
posed of lactide and glycolide change during degradation, using
a direct contact test. Direct contact tests are highly sensitive and
can be used to detect even small levels of cytotoxicity in samples.
One limitation of direct contact studies is the potential for me-
chanical damage to the cell monolayer, therefore great care must
be taken with experimental technique when placing the material
in contact with the cells. To overcome this, we have developed a
protocol for pre-degrading polymers while cradled within a tran-
swell insert, which is then transferred directly from degradation
media to cell culture. This method has been validated against con-
ventional direct contact methodology, showing no significant dif-
ference in results, but with the advantage of convenience during
sample handling, especially for degraded samples that become fri-
able or gel-like, and that are difficult to keep intact on handling
(Supplementary information Figure S2). The aim of this study was
to develop an in vitro protocol that can be used to assess the cy-
totoxicity of bioresorbable polymers, at different stages of degra-
dation, in an accelerated period. The increased temperature ac-
celerated degradation methodology, validated in a previous study
[28], was used to assess the long-term cytotoxicity of PDLLGA
and PLLGA. Both polymers were pre-degraded in PBS buffer at
47°C, before being placed on L929 fibroblasts and RAW264.7
macrophages.
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
Fig. 10. Calcium mobilisation assay of L929 fibroblasts and RAW264.7 macrophages response to polymer degradation media a) PDLLGA pre-degraded at 47°C for 12 days b)
PLLGA pre-degraded at 47°C for 70 days.
The results of this work demonstrate differences in the cyto-
toxic potential of PDLLGA and PLLGA. The cytotoxicity of PDLLGA
increased significantly at the later stages of degradation, after 12
days with 0% cell viability in both the fibroblasts and macrophages,
whereas PLLGA demonstrated low levels of cytotoxicity against fi-
broblasts after 42 and 70 days, with 61% and 64% cell viabil-
ity, respectively. The correlation of these results to the degrada-
tion behaviour data is shown in Figs. 5 and 6. PDLLGA showed
a significant reduction in cell viability, in both fibroblasts and
macrophages, at the later stages of degradation. Fig. 5 shows that
there were no significant reductions in the pH of the degrada-
tion media or mass of the polymer after 12 days. However, a large
reduction in the Mn was observed after 5 days, with substantial
polymer swelling occurring after 10 days. This indicates that low
molecular weight polymer chains are being trapped within the
polymer matrix. It is likely that these acidic degradation products
are released from the polymer during incubation with the cells,
overwhelming the local microenvironment at the polymer/cell in-
terface and causing a significant reduction in cell viability. In com-
parison, PLLGA did not become toxic to the macrophages but
showed cytotoxic potential against the fibroblasts after 42 days and
70 days. As shown in Fig. 6, there was no significant mass loss after
42 days, but there was a gradual decline in the molecular weight
of the polymer. This suggests that low molecular weight chains are
retained within the polymer matrix. These acidic chains are sub-
sequently released during incubation with the fibroblasts, causing
a reduction in cell viability. After 56 days, mass is lost from the
polymer and a significant decrease in the pH of the degradation
media demonstrates the release of the acidic degradation products
from the polymer matrix. Therefore, when this sample is in con-
tact with fibroblasts, there is no reduction in cell viability as there
11
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
is no accumulation of acidic degradation products within the poly-
mer matrix. After 70 days, further mass loss occurs, however, there
is an increase in polymer swelling. Swelling occurs via an influx of
the acidic degradation media, and a reduction in cell viability oc-
curs when the acidic degradation products are released from the
polymer matrix upon incubation with the fibroblasts. It is unclear
why the cytotoxic effect was only observed in the fibroblasts and
not the macrophages. However, it should be noted that the cyto-
toxic effect observed for PLLGA was significantly lower than that
of PDLLGA, which suggests PDLLGA has greater potential to dis-
play cytotoxicity compared to PLLGA. The molecular weight dis-
tributions, from a previous study by Geddes et al. [28], revealed
that PDLLGA is more acidic than PLLGA at the later stages of degra-
dation as high amounts of oligomers and monomers are retained
within the polymer matrix of PDLLGA. This could explain the dif-
ferences in cytotoxicity of the two polymers.
The temperature accelerated degradation methodology vali-
dated in a previous study [28], can be used as a method to pre-
degrade specimens at the accelerated temperature (47°C) where
physiological temperature (37°C) would be impractical due to the
time duration required. It should be noted however that the model
is only valid up to the point where mass loss occurs, so beyond
this point it will become a less reliable tool. The model devel-
oped found that the rate of hydrolysis at 47°C was increased 3.7-
fold in PDLLGA and 21.9-fold in PLLGA compared to 37°C. Polymer
degradation time is dependent on many factors, such as geometry,
composition, crystallinity and implant location. It can, therefore, be
difficult to give exact degradation times for every scenario and it
is acknowledged that, in the in vivo environment, factors such as
phagocytosis combined with enzymic action may play a role in the
later stages of degradation [40] . Few in vivo animal studies or in-
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
deed clinical studies have monitored PLLGA degradation up to the
point of full resorption. Of four studies identified for PLLGA in vivo
degradation, none continued to monitor up to full resorption, so
there is no definitive answer as to when this occurs. What can be
determined from the literature [40–43], is that PLLGA is likely to
be substantially resorbed within 24 months in vivo. It should be
noted, however, that additives such as ciprofloxacin [40] or trical-
cium phosphate (TCP) [41,43] may influence polymer degradation
rate. Our previous work, using an in vivo rabbit model, showed a
slight, but non-significant reduction in degradation rate in PLLGA
screws loaded with TCP, compared to screws composed of PLLGA
polymer alone [47].
Throughout the literature, an inflammatory foreign body reac-
tion for PDLLGA has been observed to occur between 14–243 days
[21,41,44–48]. Fewer incidences of foreign body reactions have
been reported for PLLGA, with some reporting no foreign body re-
actions [40,42], and others reporting incidences of cyst formation
occurring between 730–1460 days [41,43]. Based on the results of
this in vitro study, the degradation times extrapolated to 37°C, in-
dicate adverse reactions would be predicted to occur 37 days after
implantation for PDLLGA and after 920 days for PLLGA. This ex-
trapolation method corroborates the results observed in this study
with the adverse reactions noted by clinical case reports in the lit-
erature [21,41,43–48].
4.2. Development of an improved method to test for cytotoxicity
In Section 4.1, fibroblasts and macrophages were cultured in
traditional static 24-well plates. An MTT assay was carried out, af-
ter PDLLGA and PLLGA were placed on the cultured cells and in-
cubated for 24 hours. Multiwell plates, used in this way, have be-
come a standard in cell culture experiments. However, static con-
ditions are limited as they do not resemble the dynamic nature
of the physiological environment. The QV500 system was inves-
tigated as a technique that could be used to improve upon the
static multiwell plate conditions, and better mimic in vivo con-
ditions. The transwell methodology used for the static conditions
was adapted using the QV500 system, as the dimension of the
chambers (15 mm internal diameter, 10 mm height) are similar to
those of the standard 24-well plate [49,50]. The system was con-
nected in series to allow for a continuous flow of media through
each chamber. Therefore, degradation products, released into the
media, are continuously carried away and replenished with fresh
media, as would be the case in vivo, where degradation products
would diffuse from the vicinity of the implanted material follow-
ing a concentration gradient, and where there would be a con-
tinuous exchange of tissue fluid. In this study, samples were as-
sessed for cytotoxicity using the QV500 system and compared to
the response observed under static conditions (Fig. 7). The results
demonstrate that when PDLLGA reaches a certain stage of degrada-
tion it becomes inherently cytotoxic to the fibroblasts, even under
flow conditions. This suggests that the amount of acidic degrada-
tion products released from PDLLGA exceeds the clearing capacity
of the flow system, resulting in a reduction in cell viability. PLLGA
showed improved cell viability using the flow system with no sig-
nificant difference between the degraded sample and the control,
whereas under static conditions a cytotoxic response was observed.
This indicates that the flow of fresh media through the chamber
provides sufficient clearance of the acidic degradation products re-
leased from PLLGA. The flow system is more indicative of the true
in vivo scenario where diffusion through, and replenishment of tis-
sue fluid in the peri-implant environment will result in the clear-
ance of acidic degradation products.
Throughout the literature there are reports that implanting
polymers composed of lactide and glycolide in bone regions
with poor vascularity can result in an adverse reaction occurring
12
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
[8,23,51,52]. This study has demonstrated that methodology using
flow conditions can reveal an improved cytotoxicity profile of one
polymer (PLLGA) but have no impact on another (PDLLGA). Poly-
mers in the poly(lactide-co-glycolide) family degrade via heteroge-
neous bulk degradation. However, each polymer within this fam-
ily exhibits different degradation characteristics which are depen-
dent on a number of factors, such as polymer morphology, com-
position and configurational structures, and the geometry of the
device [53,54]. Future research should consider how different poly-
mer degradation mechanisms and characteristics will result in dif-
ferences of physiological response.
4.3. Further evaluation of the inflammatory response
The aim of the final stage of this study was to gain a better
understanding of the in vitro inflammatory response of fibroblasts
and macrophages to PDLLGA and PLLGA at progressive stages of
degradation. The results of the antibody array (Fig. 8) and ELISA
(Fig. 9) demonstrate a general downregulation of inflammatory cy-
tokines at degradation levels where PDLLGA and PLLGA begin to
show cytotoxic potential. Alternatively, when no cytotoxicity was
observed, there was either no significant change or an upregu-
lation in the expression of the inflammatory cytokines. This was
contrary to our expectations, as it was anticipated that materials
displaying cytotoxic potential would initiate an inflammatory re-
sponse. Surprisingly, these results suggest an immunosuppressive
or anti-inflammatory effect of PDLLGA and PLLGA. Sung et al. [38],
compared the biocompatibility of fast degrading PDLLGA to the
slow degrading PCL. The polymer scaffolds were implanted subcu-
taneously in wild-type 129/SvEv mice, harvested after 7, 14, 21 and
28 days and assessed for the density of inflammatory cells within
the polymer. It was found that a significantly reduced number of
inflammatory cells migrated into PDLLGA than PCL and it was sug-
gested that the acidic environment, created by PDLLGA, inhibited
cell migration. However, it is not clear what inflammatory cells
were detected. Britland et al. [55] found a dose dependent rela-
tionship between lactate concentration and cell viability using an
MTT assay. It was observed that cell viability was reduced when
cells were treated with a lactate concentration of 10 mM for hu-
man umbilical vein endothelial cells (HUVEC) and between 10 and
50 mM for 3T3 fibroblasts. However, the authors were not able to
determine the concentration threshold for when lactate made this
change from being a positive influence on the healing process to
becoming a detrimental presence.
In light of our results discussed above, we formed the hypoth-
esis that the degradation products of PDLLGA and PLLGA must be
active in an immunomodulatory capacity. Lactate/lactic acid, a ma-
jor degradation product, has been found to be biologically active
and involved in the signalling pathways and regulatory functions
in many physiological and pathological conditions [56,57]. Lactate
is the conjugate base of lactic acid and is formed in the body as
the end product of anaerobic or aerobic glycolysis [55]. Anaerobic
glycolysis yields lactate under hypoxic conditions in normal cells,
whereas, in aerobic glycolysis (Warburg effect [58]) lactate is pro-
duced, in the presence of oxygen, in rapidly dividing cells, for ex-
ample, in cancer cells or during wound healing [56,59]. Monocar-
boxylate transporters (MCTs) are used to transport lactate across
cell membranes, which is carried out by a series of reactions, in-
volving the conversion of lactate to pyruvate and alterations to
the intercellular NADH/NAD+ ratio [59,60]. Lactate has tradition-
ally been considered as an energy-rich metabolic fuel, but more
recently its role as an important cell signalling molecule has been
come to light. It was observed for the first time in 2009 that HCA1
(also known as G-protein coupled receptor-81 or GPR81) could
be activated by L-lactate [56,57,59,61]. The D-lactate enantiomer
has also been reported as a agonist/partial agonist of the receptor
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
[61–63]. The stimulation of HCA1 by lactate results in cell sig-
nalling by the downregulation of cAMP and the activation of non-
canonical, β -arrestin-dependent signalling pathways [56,57]. This
method of signalling is different to MCT-driven transport and its
discovery has led to significant interest in the role of lactate as a
cell signalling molecule through the activation of HCA1 . Lactate has
been shown to have an important role in the modulation of the
immune response and inflammation of various pathophysiological
conditions. The activation of HCA1 has been shown to have an im-
munosuppressive effect on the regulation of intestinal inflamma-
tion by signalling the downregulation of inflammatory cytokines
in colonic macrophages and dendritic cells [64]. The presence of
lactate generated via aerobic glycolysis in cancer cells has been
found to promote tumour growth and metastasis by acting as an
immunosuppressant [59,65,66]. Lactate has also been shown to in-
crease collagen deposition and promote angiogenesis by inducing
the release of proangiogenic factors, such as VEGF [67–70]. The
therapeutic effect of poly(lactide-co-glycolide) (PLGA) polymers in
wound healing and as a drug delivery system has been reported
in the literature. Porporato et al. [69], reported that wound heal-
ing was accelerated in two strains of mice, by delivering a rate-
controlled release of lactate using PDLLGA 50:50. It was found that
a nearly 60% decrease in excisional wound closure, 10 days post
injury, and a 2.2-fold increase in angiogenesis in the mice treated
with the polymer occurred. Similar results were also observed by
Chereddy et al. [71], with a 75% wound closure observed after 10
days, when the injury was treated with 1 mg of PLGA nanopar-
ticles, however, it is unclear which enantiomer of lactide was
used. They also observed a downregulation in the inflammatory
expressions of glutathione peroxidase (GPx) and nuclear factor-κ B
(NFκ B). These observations reported in the literature, and the re-
sults reported in this present study (inflammatory cytokine pro-
filing, and media from PDLLGA and PLLGA degradation studies in-
ducing GPCR receptor activation, as demonstrated by calcium mo-
bilisation), have important implications for lactic acid-based biore-
sorbable materials used for medical devices, such as orthopaedic
fixation screws. These devices have to date been chosen for their
physical and mechanical characteristics, however their effects on
host tissues may be much further reaching, given the proven cell
signalling role of lactate. Devices made from these polymers have
the potential to be bio-active and influence inflammation and heal-
ing at the site of implantation.
The results in this study support what has been revealed in
the literature in relation to the immunosuppressive effect of lac-
tate. During the degradation of PDLLGA and PLLGA a downregula-
tion in the inflammatory cytokines in fibroblasts and macrophages
was observed. The results from the calcium mobilisation assays
demonstrate that this observed effect is likely to be due to the ac-
tivation of the GPCR, HCA1 , by the polymer degradation product,
lactate/lactic acid, which in turn acts as a signalling molecule to
downregulate the cytokines. This effect is likely to have an impact
on the in vivo implantation of PLGA and PLA polymers. It is possi-
ble that as a polymer degrades, in vivo, the lactate released initially
has a therapeutic anti-inflammatory effect on the surrounding tis-
sue. However, at the later stages of degradation, significant poly-
mer erosion leads to a large release of acidic degradation products,
which could be considered the ‘tipping-point’ at which lactate lev-
els overwhelm the polymer/tissue environment.
5. Conclusion
The results observed in this study demonstrate the occurrence
of a delayed inflammatory reaction at the late stages of polymer
degradation. It can be difficult to extrapolate in vitro results to in
vivo applications as the conditions are very different. However, the
use of the increased temperature, 47°C, to accelerate the degrada-
13
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
tion of PDLLGA and PLLGA has demonstrated great potential for the
assessment of bioresorbable polymers with long degradation times
as the results extrapolated to 37°C corroborate with the clinical
data of adverse reactions occurring, in vivo, in the literature. Efforts
were also made in this work to improve upon the static conditions
generally used for in vitro research. The results validate the feasi-
bility of using the QV500 system as a method to better represent
the in vivo flow environment. This study has shown differences in
the cytotoxicity of PDLLGA and PLLGA and proposed explanations
for the observed response. It is recommended that future research
into orthopaedic devices focuses on PLLGA.
The final section of this work demonstrates the immunosup-
pressive and anti-inflammatory role of lactate. To the best of the
authors knowledge, the effect of lactate in the clinical outcomes
of PLA and PLGA used for orthopaedic devices has not been con-
sidered in the literature, due to the recent discovery of the ligand
activity of lactate at HCA1 receptor. It is likely lactate plays a sig-
nificant role in the cell signalling and regulatory functions at the
site of polymer implantation. It is therefore important future re-
search considers this effect when analysing the biological response
of these polymers.
Declaration of Competing Interest
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared to
influence the work reported in this paper.
Acknowledgments
Corbion Biomaterials, The Netherlands, is acknowledged for the
kind gift of poly(D, L-lactide-co-glycolide) and poly(L-lactide-co-
glycolide). This work has been funded by the Department for the
Economy, Northern Ireland, via a PhD studentship awarded to LG.
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.actbio.2021.07.049.
References
[1] M.S. Taylor, A.U. Daniels, K.P. Andriano, J. Heller, Six bioabsorbable polymers:
In vitro acute toxicity of accumulated degradation products, J. Appl. Biomater.
5 (1994) 151–157.
[2] J. Bergsma, Late degradation tissue response to poly(L-lactide) bone plates and
screws, Biomaterials 16 (1995) 25–31.
[3] P.B. Maurus, C.C. Kaeding, Bioabsorbable implant material review, Oper. Tech.
Sports Med. 12 (2004) 158–160.
[4] R.E. Cameron, A. Kamvari-Moghaddam, Synthetic bioresorbable polymers, in:
Degrad. Rate Bioresorbable Mater. Predict. Eval., 2008: pp. 43–66.
[5] K. Ginjupalli, G.V. Shavi, R.K. Averineni, M. Bhat, N. Udupa, N.P. Upadhya,
Poly(α -hydroxy acid) based polymers: a review on material and degradation
aspects, Polym. Degrad. Stab. 144 (2017) 520–535.
[6] D.G. Weir NA, F.J. Buchanan, J.F. Orr, D.F. Farrar, Degradation of poly-L- lac-
tide. Part 2: increased temperature accelerated degradation, Proc. IME H J. Eng.
Med. 218 (2004) 321–330.
[7] D. Farrar, Bioresorbable polymers in orthopaedics, Biomaterials 6 (2005) 1–4.
[8] J.C. Middleton, A.J. Tipton, Synthetic biodegradable polymers as orthopedic de-
vices, Biomaterials 21 (20 0 0) 2335–2346.
[9] J.C.Y. Chan, K. Burugapalli, J.L. Kelly, A.S. Pandit, in: Influence of Clinical Ap-
plication on Bioresorbability: Host Response, Elsevier, 2008, pp. 267–318. in:
Degrad. Rate Bioresorbable Mater..
[10] J.M. Anderson, A. Rodriguez, D.T. Chang, Foreign body reaction to biomaterials,
Semin. Immunol. 20 (2008) 86–100.
[11] Z. Sheikh, P.J. Brooks, O. Barzilay, N. Fine, M. Glogauer, Macrophages, for-
eign body giant cells and their response to implantable biomaterials, Materials
(Basel) 8 (2015) 5671–5701.
[12] E. Mortaz, S.D. Alipoor, I.M. Adcock, S. Mumby, L. Koenderman, Update on neu-
trophil function in severe inflammation, Front. Immunol. 9 (2018) 1–14.
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
[13] A.R. Amini, J.S. Wallace, S.P. Nukavarapu, Short-term and long-term effects of
orthopedic biodegradable implants, J. Long. Term. Eff. Med. Implants. 21 (2011)
93–122.
[14] J.M. Anderson, G. Cook, B. Costerton, S.R. Hanson, A. Hensten-Pettersen,
N. Jacobsen, R.J. Johnson, R.N. Mitchell, M. Pasmore, F.J. Schoen, M. Shirtliff,
P. Stoodley, Host reactions to biomaterials and their evaluation, in: Biomater.
Sci. (1996) 293–304.
[15] I.C.C. de Moraes Porto, in: Polymer Biocompatibility, in: Polymerization, InTech,
2012, pp. 47–62, doi:10.5772/47786.
[16] C.G. Ambrose, T.O. Clanton, Bioabsorbable implants: review of clinical experi-
ence in orthopedic surgery, Ann. Biomed. Eng. 32 (2004) 171–177.
[17] S. Konan, F.S. Haddad, A clinical review of bioabsorbable interference screws
and their adverse effects in anterior cruciate ligament reconstruction surgery,
Knee 16 (2009) 6–13.
[18] W.H. De Jong, J.E. Bergsma, J.E. Robinson, R.R.M. Bos, Tissue response to par-
tially in vitro predegraded poly- L -lactide implants, Biomaterials 26 (2005)
1781–1791.
[19] O.M. Böstman, H.K. Pihlajamäki, Late foreign-body reaction to an intraosseuous
bioabsorbable polylactic acid screw, J. Bone Jt. Surg. 80-A (1998) 1791–1794.
[20] K.A. Athanasiou, G.G. Niederauer, C.M. Agrawal, Sterilization, toxicity, biocom-
patibility and clinical applications of polylactic acid /polyglycolic acid copoly-
mers, Biomaterials 17 (1996) 93–102.
[21] S. Konan, F.S. Haddad, The unpredictable material properties of bioabsorbable
PLC interference screws and their adverse effects in ACL reconstruction
surgery, Knee Surgery, Sport. Traumatol. Arthrosc. 17 (2009) 293–297.
[22] J.H. Kwak, J.A. Sim, S.H. Kim, K.C. Lee, B.K. Lee, Delayed intra-articular inflam-
matory reaction due to poly-L-lactide bioabsorbable interference screw used
in anterior cruciate ligament reconstruction, arthrosc, J. Arthrosc. Relat. Surg.
24 (2008) 243–246.
[23] P.K. Givissis, S.I. Stavridis, P.J. Papagelopoulos, P.D. Antonarakos,
A.G. Christodoulou, Delayed foreign-body reaction to absorbable implants in
metacarpal fracture treatment, Clin. Orthop. Relat. Res. 468 (2010) 3377–3383.
[24] P. Tuompo, E.K. Partio, H. Pätiälä, K. Jukkala-Partio, E. Hirvensalo, P. Rokka-
nen, Causes of the clinical tissue response to polyglycolide and polylactide im-
plants with an emphasis on the knee, Arch. Orthop. Trauma Surg. 121 (2001)
261–264.
[25] J.M. Anderson, Future challenges in the in vitro and in vivo evaluation of bio-
material biocompatibility, Regen. Biomater. 3 (2016) 73–77.
[26] W. Li, J. Zhou, Y. Xu, Study of the in vitro cytotoxicity testing of medical de-
vices (Review), Biomed. Rep. (2015) 617–620.
[27] A. Lock, J. Cornish, D.S. Musson, The role of in vitro immune response assess-
ment for biomaterials, J. Funct. Biomater. 10 (2019) 31.
[28] L. Geddes, L. Carson, E. Themistou, F. Buchanan, A comparison of the in-
creased temperature accelerated degradation of Poly(D,L-lactide-co-glycolide)
and Poly(L-lactide-co-glycolide), Polym. Test. 91 (2020) 106853.
[29] International Organisation for Standardisation, Biological Evaluation of Medical
Devices - Part 5: Tests for in vitro cytotoxicity, 2009.
[30] C.K. Postnikoff, R. Pintwala, S. Williams, A.M. Wright, D. Hileeto, M.B. Gorbet,
Development of a curved, stratified, in vitro model to assess ocular biocom-
patibility, PLoS One 9 (2014) e96448.
[31] S. Nithiananthan, A. Crawford, J.C. Knock, D.W. Lambert, S.A. Whawell, Physi-
ological fluid flow moderates fibroblast responses to TGF-β 1, J. Cell. Biochem.
118 (2017) 878–890.
[32] X. Wang, Y. Xia, L. Liu, M. Liu, N. Gu, H. Guang, F. Zhang, Comparison of MTT
assay, flow cytometry, and RT-PCR in the evaluation of cytotoxicity of five
prosthodontic materials, J. Biomed. Mater. Res. - Part B Appl. Biomater. 95 B
(2010) 357–364.
[33] J. Caers, K. Peymen, N. Suetens, L. Temmerman, T. Janssen, L. Schoofs, I. Beets,
Characterization of G protein-coupled receptors by a fluorescence-based cal-
cium mobilization assay, J. Vis. Exp. (2014) 1–10.
[34] S. Anju, N. Prajitha, V.S. Sukanya, P.V. Mohanan, Complicity of degradable poly-
mers in health-care applications, Mater. Today Chem. 16 (2020) 100236.
[35] E. Díaz, I. Puerto, I. Sandonis, S. Ribeiro, S. Lanceros-Mendez, Hydrolytic degra-
dation and cytotoxicity of poly(lactic- co -glycolic acid)/multiwalled carbon
nanotubes for bone regeneration, J. Appl. Polym. Sci. 137 (2020) 48439.
[36] J.R. Sarasua, N. López-Rodríguez, E. Zuza, S. Petisco, B. Castro, M. Del Olmo,
T. Palomares, A. Alonso-Varona, Crystallinity assessment and in vitro cytotox-
icity of polylactide scaffolds for biomedical applications, J. Mater. Sci. Mater.
Med. 22 (2011) 2513–2523.
[37] Y. Yang, J. Fang, W. Liu, Y. Zhao, T. Huang, J. Cui, Z. Zhou, In vitro degradation
behavior and biological properties of a novel maleated poly (D, L-lactide-co-g-
lycolide) for biomedical applications, J. Macromol. Sci. Part B Phys. 58 (2019)
209–218.
[38] H. Sung, C. Meredith, C. Johnson, Z.S. Galis, The effect of scaffold degrada-
tion rate on three-dimensional cell growth and angiogenesis, Biomaterials 25
(2004) 5735–5742.
[39] A.A. Ignatius, L.E. Claes, In vitro biocompatibility of bioresorbable polymers:
poly (L, DL-lactide) and poly (L-lactide-co-glycolide), Biomaterials 17 (1996)
831–839.
[40] J. Tiainen, Y. Soini, E. Suokas, M. Veiranto, P. Törmälä, T. Waris, N. Ashammakhi,
Tissue reactions to bioabsorbable ciprofloxacin-releasing polylactide-polygly-
colide 80/20 screws in rabbits’ cranial bone, J. Mater. Sci. Mater. Med. 17
(2006) 1315–1322.
14
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
[41] H.E. Bourke, L.J. Salmon, A. Waller, C.S. Winalski, H.A. Williams, J.M. Linklater,
A. Vasanji, J.P. Roe, L.A. Pinczewski, Randomized controlled trial of osteocon-
ductive fixation screws for anterior cruciate ligament reconstruction: a com-
parison of the calaxo and milagro screws, Arthrosc. - J. Arthrosc. Relat. Surg.
29 (2013) 74–82.
[42] A.S. Xue, J.C. Koshy, W.M. Weathers, E.M. Wolfswinkel, Y. Kaufman, S.E. Sharabi,
R.H. Brown, M.J. Hicks, L.H. Hollier, Local foreign-body reaction to commercial
biodegradable implants: an in vivo animal study, Craniomaxillofac. Trauma Re-
constr. 7 (2014) 27–34.
[43] C.L. Cox, K.P. Spindler, J.P. Leonard, B.J. Morris, W.R. Dunn, E.K. Reinke, Do
newer-generation bioabsorbable screws become incorporated into bone at two
years after ACL reconstruction with patellar tendon graft? J. Bone Jt. Surg. 96
(2014) 244–250.
[44] A.C. Stahelin, A. Weiler, H. Rufenacht, R. Hoffmann, A. Geissmann, R. Feinstein,
Clinical degradation and biocompatibility of different bioabsorbable interfer-
ence screws : a report of six cases, 13 (1997) 238–244.
[45] G. Lajtai, I. Noszian, K. Humer, F. Unger, G. Aitzetmuller, E. Orthner, Serial mag-
netic resonance imaging evaluation of operative site after fixation of patellar
tendon graft with bioabsorbable interference screws in anterior cruciate liga-
ment reconstruction, Arthroscopy 15 (1999) 709–718.
[46] C.L. Cox, K.C. Homlar, J.L. Carey, K.P. Spindler, CALAXO osteoconductive inter-
ference screw: the value of postmarket surveillance, J. Surg. Orthop. Adv. 19
(2010) 121–124.
[47] I. Palmer, S.A. Clarke, F.J. Buchanan, Enhanced release of calcium phosphate
additives from bioresorbable orthopaedic devices using irradiation technology
is non-beneficial in a rabbit model: an animal study, Bone Joint Res. 8 (2019)
266–274.
[48] J. Dujardin, H. Vandenneucker, J. Bellemans, Tibial cyst and intra-articu-
lar granuloma formation after anterior cruciate ligament reconstruction us-
ing polylactide carbonate osteoconductive interference screws, arthrosc, J.
Arthrosc. Relat. Surg. 24 (2008) 238–242.
[49] G. Mattei, S. Giusti, A. Ahluwalia, Design criteria for generating physiologically
relevant in vitro models in bioreactors, Processes 2 (2014) 548–569.
[50] P. Miranda-Azpiazu, S. Panagiotou, G. Jose, S. Saha, A novel dynamic multicel-
lular co-culture system for studying individual blood-brain barrier cell types
in brain diseases and cytotoxicity testing, Sci. Rep. 8 (2018) 1–10.
[51] O. Böstman, H. Pihlajamäki, Clinical biocompatibility of biodegradable or-
thopaedic implants for internal fixation: a review, Biomaterials 21 (20 0 0)
2615–2621.
[52] O.M. Böstman, H.K. Pihlajamäki, Adverse tissue reactions to bioabsorbable fix-
ation devices, Clin. Orthop. Relat. Res. (20 0 0) 216–227.
[53] S. Li, M. Vert, Biodegradation of Aliphatic Polyesters, in: Degrad. Polym. Princ.
Appl., Springer, Dordrecht, 1995.
[54] S. Li, Hydrolytic degradation characteristics of aliphatic polyesters derived from
lactic and glycolic acids, J. Biomed. Mater. Res. (Appl. Biomater.) 48 (1998)
342–353.
[55] S. Britland, O. Ross-Smith, H. Jamil, A.G. Smith, K. Vowden, P. Vowden, The
lactate conundrum in wound healing: clinical and experimental findings indi-
cate the requirement for a rapid point-of-care diagnostic, Biotechnol. Prog. 28
(2012) 917–924.
[56] K.K. Chereddy, V.L. Payen, V. Préat, PLGA polymer: from a classic drug car-
rier to a novel therapeutic activity contributor, J. Control. Release. 289 (2018)
10–13.
[57] S. Sun, H. Li, J. Chen, Q. Qian, Lactic acid: no longer an inert and end-product
of glycolysis, Physiology 32 (2017) 453–463.
[58] M.V. Liberti, J.W. Locasale, The Warburg effect: how does it benefit cancer
cells? (vol 41, pg 211, 2016), Trends Biochem. Sci. 41 (2016) 287.
[59] T.P. Brown, V. Ganapathy, Lactate/GPR81 signaling and proton motive force
in cancer: role in angiogenesis, immune escape, nutrition, and Warburg phe-
nomenon, Pharmacol. Ther. 206 (2020) 107451.
[60] L. Ippolito, A. Morandi, E. Giannoni, P. Chiarugi, Lactate: a metabolic driver in
the tumour landscape, Trends Biochem. Sci. 44 (2019) 153–166.
[61] C. Liu, J. Wu, J. Zhu, C. Kuei, J. Yu, J. Shelton, S.W. Sutton, X. Li, J.Y. Su,
T. Mirzadegan, C. Mazur, F. Kamme, T.W. Lovenberg, Lactate inhibits lipolysis
in fat cells through activation of an orphan G-protein-coupled receptor, GPR81,
J. Biol. Chem. 284 (2009) 2811–2822.
[62] W. Wagner, K.D. Kania, A. Blauz, W.M. Ciszewski, The lactate receptor
(HCAR1/GPR81) contributes to doxorubicin chemoresistance via ABCB1 trans-
porter up-regulation in human cervical cancer HeLa cells, J. Physiol. Pharmacol.
68 (n.d.) 555–564.
[63] X. Castillo, K. Rosa, M.T. Wyss, K. Drandarov, A. Buck, L. Pellerin, B. Weber,
A probable dual mode of action for both L- and D-lactate neuroprotection in
cerebral ischemia, J. Cereb. Blood Flow Metab. 35 (2015) 1561–1569.
[64] P. Ranganathan, A. Shanmugam, D. Swafford, A. Suryawanshi, P. Bhattacharjee,
M.S. Hussein, P.A. Koni, P.D. Prasad, Z.B. Kurago, M. Thangaraju, V. Ganapathy,
S. Manicassamy, GPR81, a cell-surface receptor for lactate, regulates intestinal
homeostasis and protects mice from experimental colitis, J. Immunol. (2018)
1781–1789.
[65] C.L. Roland, T. Arumugam, D. Deng, S.H. Liu, B. Philip, S. Gomez, W.R. Burns,
V. Ramachandran, H. Wang, Z. Cruz-Monserrate, C.D. Logsdon, Cell surface lac-
tate receptor GPR81 is crucial for cancer cell survival, Cancer Res. 74 (2014)
5301–5310.
JID: ACTBIO
ARTICLE IN PRESS
L. Geddes, E. Themistou, J.F. Burrows et al.
[66] A. Calcinotto, P. Filipazzi, M. Grioni, M. Iero, A. De Milito, A. Ricupito, A. Cova,
R. Canese, E. Jachetti, M. Rossetti, V. Huber, G. Parmiani, L. Generoso, M. Santi-
nami, M. Borghi, S. Fais, M. Bellone, L. Rivoltini, Modulation of microenviron-
ment acidity reverses anergy in human and murine tumor-infiltrating T lym-
phocytes, Cancer Res. 72 (2012) 2746–2756.
[67] Q.P. Ghani, S. Wagner, M.Z. Hussain, T.K. Hunt, Role of ADP-ribosylation in
wound repair. The contributions of Thomas K. Hunt, MD, Wound Repair Re-
gen. 11 (2003) 439–444.
[68] J. Song, K. Lee, S.W. Park, H. Chung, D. Jung, Y.R. Na, H. Quan, C.S. Cho,
J.H. Che, J.H. Kim, J.H. Park, S.H. Seok, Lactic acid upregulates VEGF expression
in macrophages and facilitates choroidal neovascularization, Investig. Ophthal-
mol. Vis. Sci. 59 (2018) 3747–3754.
15
[m5G;August 2, 2021;15:11]
Acta Biomaterialia xxx (xxxx) xxx
[69] P.E. Porporato, V.L. Payen, C.J. De Saedeleer, V. Préat, J.P. Thissen, O. Feron,
P. Sonveaux, Lactate stimulates angiogenesis and accelerates the healing of su-
perficial and ischemic wounds in mice, Angiogenesis 15 (2012) 581–592.
[70] T.K. Hunt, R.S. Aslam, S. Beckert, S. Wagner, Q.P. Ghani, M.Z. Hussain, S. Roy,
C.K. Sen, Aerobically derived lactate stimulates revascularization and tissue re-
pair via redox mechanisms, Antioxidants Redox Signal 9 (2007) 1115–1124.
[71] K.K. Chereddy, R. Coco, P.B. Memvanga, B. Ucakar, A. Des Rieux, G. Vander-
meulen, V. Préat, Combined effect of PLGA and curcumin on wound healing
activity, J. Control. Release. 171 (2013) 208–215.