Biofabrication

PAPER • OPEN ACCESS

Functional evaluation of prevascularization in one-stage versus two-

stage tissue engineering approach of human bio-artificial muscle
To cite this article: D Gholobova et al 2020 Biofabrication 12 035021

View the article online for updates and enhancements.

This content was downloaded from IP address 178.117.91.38 on 12/06/2020 at 18:04

 Biofabrication 12 (2020) 035021 https://doi.org/10.1088/1758-5090/ab8f36

Biofabrication

PAPER

Functional evaluation of prevascularization in one-stage versus

OPEN ACCESS
two-stage tissue engineering approach of human bio-artificial
RECEIVED
13 July 2019 muscle
REVISED
6 March 2020 D Gholobova1, L Terrie1, K Mackova2,3,4,5, L Desender1, G Carpentier6, M Gerard1, L Hympanova2,3,4,5,
ACCEPTED FOR PUBLICATION J Deprest2,3,4 and L Thorrez1,7
1 May 2020
1
Tissue Engineering Lab, Department of Development and Regeneration, KU Leuven, E. Sabbelaan 53, 8500, Kortrijk, Belgium
PUBLISHED 2
9 June 2020 Centre for Surgical Technologies, Group Biomedical Sciences, KU Leuven, Leuven, Belgium
3
Department of Development and Regeneration, Woman and Child, Group Biomedical Sciences, KU Leuven, Leuven, Belgium
4
Pelvic Floor Unit, University Hospitals KU Leuven, Leuven, Belgium
Original content from 5
Institute for the Care of Mother and Child, Third Faculty of Medicine, Charles University, Prague, Czech Republic
this work may be used 6
under the terms of the Laboratoire de Recherche sur la Croissance Cellulaire, la Réparation et la Régénération Tissulaires (CRRET), Faculté des Sciences et
Creative Commons Technologie, Université Paris-Est, 94000, Créteil, France
Attribution 3.0 licence. 7
Author to whom any correspondence should be addressed.
Any further distribution
E-mail: lieven.thorrez@kuleuven.be
of this work must
maintain attribution to
the author(s) and the title Keywords: prevascularization, tissue engineering, skeletal muscle
of the work, journal Supplementary material for this article is available online
citation and DOI.

Abstract
A common shortcoming of current tissue engineered constructs is the lack of a functional
vasculature, limiting their size and functionality. Prevascularization is a possible strategy to
introduce vascular networks in these constructs. It includes among others co-culturing target cells
with endothelial (precursor) cells that are able to form endothelial networks through
vasculogenesis. In this paper, we compared two different prevascularization approaches of
bio-artificial skeletal muscle tissue (BAM) in vitro and in vivo. In a one-stage approach, human
muscle cells were directly co-cultured with endothelial cells in 3D. In a two-stage approach, a one
week old BAM containing differentiated myotubes was coated with a fibrin hydrogel containing
endothelial cells. The obtained endothelial networks were longer and better interconnected with
the two-stage approach. We evaluated whether prevascularization had a beneficial effect on in vivo
perfusion of the BAM and improved myotube survival by implantation on the fascia of the
latissimus dorsi muscle of NOD/SCID mice for 5 or 14 d. Also in vivo, the two-stage approach
displayed the highest vascular density. At day 14, anastomosis of implanted endothelial networks
with the host vasculature was apparent. BAMs without endothelial networks contained longer and
thicker myotubes in vitro, but their morphology degraded in vivo. In contrast, maintenance of
myotube morphology was well supported in the two-stage prevascularized BAMs. To conclude, a
two-stage prevascularization approach for muscle engineering improved the vascular density in the
construct and supported myotube maintenance in vivo.

1. Introduction localized skeletal muscle repair includes muscle flap

transplantation. Therefore, autologous muscle tissue
Skeletal muscle tissue has a high regenerative capacity is transplanted from a donor site to the damaged
that is able to replace damaged muscle tissue upon muscle. Unfortunately, this is associated with donor
injury [1, 2]. A central role in muscle regeneration site morbidity and poor flap survival [3]. To limit this
is played by the muscle progenitor cells known as problem, microsurgery is used to include vasculature
satellite cells. However in cases of volumetric muscle to allow the reconstitution of blood supply in the tis-
loss, a failure of muscle repair and scar tissue forma- sue. In turn, this provides oxygen and nutrients, and
tion is often observed, resulting in functional muscle removes waste products throughout the transplanted
impairment [2]. The current treatment strategy for tissue [4, 5].

© 2020 The Author(s). Published by IOP Publishing Ltd

Biofabrication 12 (2020) 035021 D Gholobova et al

To avoid the need of donor site tissue and perfusion is not as fast as with in vivo prevasculariz-
related injuries, tissue engineered skeletal muscle ation using microsurgery to connect the construct to
holds promise as an alternative treatment. Skeletal the host vasculature. However, vascularization speed
muscle tissue engineering aims to create in vitro is not the sole factor to consider [16]. Functionality
muscle tissue that mimics the structure and func- of the newly formed vessels will determine the success
tion of in vivo muscle, and thus offers the potential of a tissue-engineering strategy as well. More specific-
to repair the damaged muscle area. Similar to muscle ally, regression of in vitro formed networks should be
flaps, one of the current limitations of avascular tissue avoided [17].
engineered constructs is the lack of a blood supply in Still, only a limited number of implanta-
the initial phase after implantation since passive dif- tion studies focused on pre-vascularized tissue
fusion is limited to 150–200 µm [4, 6, 7]. In response engineered skeletal muscle have been conducted
to the hypoxia that arises when having a tissue engin- so far. A prevascularized fibrin matrix, obtained
eered graft with wall thickness greater than the dif- using a microsurgically created arteriovenous-loop
fusion limit, host vessels invade the implanted tissue. approach, was injected with expanded primary myo-
However, the rate of spontaneous vascular ingrowth blasts [18]. This prevascularization approach pro-
is limited to ~5 µm h−1 [8], which cannot address moted myoblast survival and preservation of the
the metabolic need in time leading to necrosis in the myogenic phenotype, but no differentiation into
central area of the graft. As a result, successful use of myotubes occurred. In another approach, a pre-
tissue-engineered constructs in the clinic is limited to vascularized construct containing HUVECs (human
thin or avascular tissues, such as skin or cartilage, in umbilical vein endothelial cells), C2C12 mouse myo-
which this rate of neovascularization from the host is blasts and fibroblasts was implanted, which resulted
sufficient. Yet, in other tissue grafts such as muscle, in enhanced survival and prolonged functionality of
vascularization will be of key importance to achieve the skeletal muscle construct [12, 19]. The used poly-
successful transplantation. This remains an import- L-lactic acid (PLLA)/polylactic-glycolic acid (PLGA)
ant hurdle to overcome if the aim is to increase the 3D porous scaffolds however over time release acidic
size of engineered tissue constructs as the lack of oxy- degradation products, which decrease cell survival
gen and nutrient supply constrains the size of viable [20]. Thus, this approach may not represent the
constructs. most optimal approach for tissue engineering vas-
Prevascularization is the formation of capillary- cularized skeletal muscle. Similarily, an engineered
like networks before implantation and is aimed to flap composed of human adipose-derived microvas-
overcome this diffusion limit. As for engineering cular endothelial cells and mesenchymal stem cells
(immature) blood vessels in skeletal muscle con- seeded on the same PLLA/PLGA scaffold was used for
structs, a few in vitro approaches have been described, abdominal wall defect repair [21]. The graft was cul-
such as co-culturing endothelial cells (EC) and myo- tured in vitro until a small capillary net was formed.
genic progenitors on different scaffolds and hydro- This was folded around an exposed femoral artery
gels, combining endothelial cell sheets with myoblast and vein of the recipient rat. After 21 d, the grafts
cell sheets and engineering vasculature using micro- with the arteries and veins were transferred to an
fabrication techniques [4, 5, 9, 10]. Besides applic- abdominal full-thickness wall defect. This resulted
ations in regenerative medicine, such a vascularized in a highly vascularized, well-integrated muscle flap.
muscle model can be used to study mechanisms However, further reports on functional muscle regen-
underlying myogenesis, vasculogenesis and vasculo- eration using this approach are missing.
myogenic interactions. Another disadvantage in the above mentioned
This prevascularization is based on de novo blood prevascularization studies is the use of solid scaffolds.
vessel formation known as vasculogenesis and the The skeletal muscle tissue-engineering approach used
advantages of this strategy are fourfold. First, the in our lab is based on previous work of the Vanden-
presence of this preformed endothelial network fur- burgh lab and uses hydrogels, allowing cells to self-
ther enhances host vessel ingrowth upon implant- organize and migrate [22, 23]. In addition, a hydro-
ation [11, 12]. Second, having a prevascularized gel better mimicks the physiological environment of
construct bears the advantage that the preformed developing skeletal muscle which has been shown to
microvascular network only has to anastomose to influence the proliferation of muscle progenitor cells
the host vasculature after implantation to achieve [24]. In this approach, cells isolated from a human
rapid construct vascularization [13]. Third, this pre- muscle biopsy are expanded to several millions of
vascularization improves perfusion and survival of muscle cells (myoblasts and a minority of fibroblasts).
the construct upon implantation [11, 12]. Fourth and These are cast in silicone molds in the presence of a
last, in addition to forming blood vessels, endothelial natural hydrogel. After one week, this results in the
cells are known for their paracrine activity on formation of a bio-artificial muscle (BAM) contain-
developing tissue, including skeletal muscle [14–16]. ing multinucleated myotubes, well-aligned in the dir-
Although an in vitro prevascularization approach ection of the attachment points. The BAM is lim-
reduces the time needed for complete vascularization, ited in size (2 cm long, ±1 mm thick) due to the

2
Biofabrication 12 (2020) 035021 D Gholobova et al

absence of a vascular network. To increase the size Glutamax and pyruvate supplemented with 0.1% col-
of the BAM, which would expand its applications lagenase, type II (Sigma) and 4 mg ml−1 dispase II
and utility, we already succeeded in prevascularizing a (Roche Diagnostics). After the incubation, isolated
BAM in vitro in a previous study [10]. This was done cells were collected by filtering through a 100 µm cell
by seeding HUVECs together with muscle cells in a strainer (Falcon) and fragments were incubated again
fibrin hydrogel followed by one week of tissue con- to digest the whole tissue. Isolated cells were pooled,
traction and cell differentiation. This is a direct 3D centrifuged for 5 min at 200 × g and resuspended in
co-culture approach, defined as a one-stage prevas- SkGM. Muscle cells were split at 60%–70% conflu-
cularization approach in this manuscript (BAM-1s). ence and used in experiments at 12 doublings. In the
Despite detailed in vitro analysis, our previous study isolated muscle cell population, myoblast percentage
lacked an in vivo part in wich the functionality of the and fusion index were determined as described before
preformed endothelial networks was tested. In this [10, 25].
follow-up study we aim to (i) further improve the pre- Muscle cells that were used for tissue engineer-
vascularization with minimal loss of myotube forma- ing of BAMs, and subsequently for implantation,
tion in the BAM based on a novel two-stage BAM pre- expressed the far-red fluorescent protein miRFP670
vascularization approach (BAM-2s) and (ii) assess in [26]. The expression of miRFP670 was obtained by
vivo behaviour of prevascularized BAMs in a mouse transducing 5 × 105 muscle cells in 3.15 ml SkGM
model. for 48 h with 350 µl 1.21 × 106 pg p24 ml−1 CMV-
miRFP670 cDNA-containing HIV1-derived retro-
viral particles. Retroviral vectors were produced as
2. Methods
previously described by the Leuven Viral Vector
Core (Full description in [27] and with improve-
2.1. Cell culture media
ments in [28]). The transfer plasmid pCH-CMV-
The medium that was employed for the isolation
miRFP670-3xflag-IRES-blasti was made by cloning
and proliferation of myogenic progenitors (skeletal
the miRFP670 sequence (RefSeq KX421097.1) into
muscle growth medium, SkGM) was composed of
a pCH-CMV-eGFP-3xflag-Ires-blasti plasmid. All
high glucose Dulbecco’s Modified Eagle Medium
cloning steps were verified by sequencing. This resul-
(DMEM, Gibco), 10% fetal bovine serum (FBS,
ted in the expression of triple flag-tagged miRFP670
Thermo Fisher), 50 µg ml−1 gentamicin (Life Tech-
under the control of a cytomegalovirus (CMV) pro-
nologies), and 1% Ultroser solution (Pall corpora-
motor and with simultaneous expression of a blasti-
tion). Differentiation medium (SkFM) composed of
cidin resistance cDNA. Lentiviral vector particle
DMEM with high glucose containing 10 ng ml−1
number was estimated by quantifying the surface pro-
hEGF, 10 µg ml−1 insulin, 50 µg ml−1 BSA and
tein p24 with enzyme linked immunosorbent assay
50 µg ml−1 gentamicin was used to promote form-
(ELISA) (HIV-1 core profile ELISA, DuPont), which
ation of multinucleated myotubes.
was determined to be 1.21 × 106 pg p24 ml−1 .

2.2. Cell culture 2.3. Tissue-engineering of BAMs

Green Fluorescent Protein (GFP)-labeled HUVECs Three different types of BAMs were used:
(GFP-HUVECs, Angio-proteomie) were seeded in (i) BAMs containing only muscle cells
cell culture flasks which were precoated for 1 h at (2×106 muscle cells construct−1 ) cultured for
37 ◦ C with gelatin (0.1% gelatin, Millipore) and cul- 2 weeks in SkGM/SkFM (termed BAM), (ii) the
tured in endothelial growth medium (EGM-2 with one-stage prevascularized BAM with muscle cells
bullet kit, Lonza). HUVECs were split at 90% conflu- and endothelial cells (1.4×106 muscle cells and
ence and used in the experiment at passage 7. 0.6 × 106 HUVECs construct−1 ) cultured for 1 week
Human skeletal muscle cells were isolated from a in EGM-2 (termed BAM-1s) and (iii) the two-stage
fresh human muscle tissue biopsy, of an adult male prevascularized BAM with initially only muscle cells
aged 60, obtained from the Human Body Donation (2×106 muscle cells construct−1 ) cultured for 1 week
programme of KU Leuven University as described in SkGM/SkFM followed by an additional embed-
in [25]. The donor provided written consent (for ding step in a fibrin hydrogel containing 2 × 106 GFP
research and educational purpose) prior to death. labeled HUVECs (termed BAM-2s) (schematic rep-
Briefly, isolated tissue was cut in strips of approx- resentation in figure 1). In previous experiments, the
imately 2 mm × 10 mm using sterile forceps and initial seeding density of in total 2 × 106 cells has been
scalpel after removing excess connective tissue and set based on the obtained aligment and dense form-
fat. Before isolation of cells, muscle strips were pinned ation of myofibers [10]. For the BAM-1s, this same
under tension into a sylgard coated 6-well plate to initial cell density appeared to result in best myofiber
maintain cell survival while avoiding muscle atrophy. formation and the total amount of cells was divided
Two days after pinning the strips, enzymatic diges- over the two cell types. The ratio of cells used was
tion was performed by incubating the muscle strips described to be the best trade-off between maximal
at 37 ◦ C for 1 h in DMEM high glucose with myogenesis while having interspersed endothelial

3
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 1. Schematic representation of tissue engineering strategies. (top) BAM represents bio-artificial muscle in a fibrin hydrogel
containing aligned myotubes, differentiated for 2 weeks in SkGM and SkFM medium. (middle) BAM-1s is a prevascularization
strategy based on co-culture of endothelial and muscle cells in a fibrin hydrogel, cultured for 1 week in EGM medium which
results in aligned myotubes with interspersed endothelial networks. (bottom) BAM-2s is a prevascularization strategy by
embedding a one-week old BAM in a new fibrin hydrogel containing endothelial cells (EC). Culturing this construct for an
additional week in EGM medium, results in a myotube bundle surrounded by an endothelial coat containing an endothelial
network.

networks in [10]. Finally, for the BAM-2s the ini- (NOD/SCID) mice (Janvier) (n = 12 for BAM, n = 8
tial amount of cells (muscle cells) was kept constant for BAM-1s and n = 4 for BAM-2s). Imbalance
(2×106 ) as this allows the direct comparison with in numbers is related to the surgical procedure in
the BAM in terms of myofibers while endothelial cell which a lower number of animals with the BAM-
numbers were increased to 2 × 106 HUVECs. The 2s constructs survived. The higher number of the
latter was optimized separately in terms of maximal BAM as compared to the BAM-1s and BAM-2s is
endothelial network formation extent. because the BAM condition was used as a refer-
Tissue-engineering of BAMs was described in ence in each animal. All animals were anesthet-
[10, 25]. Briefly, cells were mixed with 500 µl throm- ized and maintained with isoflurane (0.5%–2%, 3–
bin (4 U ml−1 , Bio Phar Laboratories LLC). The mix- 5 l min−1 ) and oxygen while being kept warm on
ture was cast into 25-mm-long silicone rubber molds a heating pad during surgery and recovery. The fas-
with end attachment sites spaced 20 mm. Then, to cia of the latissimus dorsi muscle was scraped with
form a fibrin hydrogel (1 mg ml−1 ) containing the a sterile scalpel until minor bleeding was observed.
cells, 500 µl fibrinogen (2 mg ml−1 , Merck Chemic- Both sides of the BAM were secured to the muscle
als) was added and the cell-gel mix was mixed. Fol- with non-resorbable polypropylene suture (Ethicon)
lowing 2 h incubation at 37 ◦ C, SkGM (for BAM) parallel to the spine. After implantation, meloxicam
or EGM-2 medium (for BAM-1s) supplemented with (2 mg kg−1 ; Meloxidyl CEVA) and buprenorphine
fibrinolysis inhibitors aprotinin (92.5 µg ml−1 Carl (0.1 mg kg−1 ; Vetergesic) (for 2 d) were admin-
Roth) and tranexamic acid (400 µM, Sigma) was istered subcutaneously. Animals were euthanized 5
added. For BAM, the medium was switched to dif- or 14 d later. Construct and underlying host tissue
ferentiation medium (SkFM) two days after casting. were explanted to capture both construct mainten-
To obtain the third type of BAM, BAM-2s, 1 week old ance and host vessel ingrowth as well as construct
BAMs were embedded in 1 ml fibrin (created similar integration into the host respectively. Host muscle
as above) hydrogel containing 2 × 106 HUVECs and without direct contact with the implanted constructs
cultured for another week in EGM-2. Medium was (gastrocnemius) was explanted each time and served
replaced every 2 d and BAMs were kept in culture for 7 as a negative control for all graft-specific stainings.
or 14 d prior to thickness measurement and fixation. Explants were washed in PBS for 5 min and fixed
using 4% formaldehyde solution (w/v) for 6 h at room
2.4. Thickness measurements temperature (RT).
Cross-sectional thickness of BAMs in culture and
post-implantation was measured with a sterile micro-
meter at day 7 or day 14 after casting. 2.6. Whole mount immunofluorescence staining
To assess in vitro BAM formation, the 3 types of BAM
2.5. In vivo implantation of BAMs (BAM, BAM-1s, BAM-2s) were washed (3 × 5 min
All animal experiments were evaluated and approved in PBS) and then removed from the attachment sites.
by the Animal Ethics Committee of the KU Leuven Then, they were pinned on Styrofoam to preserve
(P099/2017) and were performed according to inter- their original shape during fixation in 4% formalde-
national guidelines. BAMs were implanted subcu- hyde for 1 h at room temperature (RT) and stored
taneously on the fascia of the latissimus dorsi muscle at 4 ◦ C in PBS. Explanted constructs from the in
for a period of 14 d in 6–7 weeks old non- vivo implantation study were fixed for 6 h at RT in
obese diabetic severe combined immunodeficiency 4% formaldehyde and stored at 4 ◦ C in PBS in the

4
Biofabrication 12 (2020) 035021 D Gholobova et al

dark. For tropomyosin immunohistochemical stain- 2.9. Data analysis

ing, samples were fixed a second time, in methanol Myotube analysis and endothelial network analysis
(−20 ◦ C, 20 min) immediately prior to staining. in in vitro constructs and after implantation were
Next, the samples were permeabilized in blocking performed as previously reported in Gholobova et al
buffer containing 0.2% Triton-X-100 (Sigma) and [10]. For myotube formation 20 µm z-projections
1% bovine serum albumin (BSA, Sigma) in PBS for were manually analyzed for different parameters
3 h at RT. Subsequently, the samples were incubated of myotube formation: myotube alignment, length,
overnight at 4 ◦ C with a monoclonal mouse anti- density and diameter. Myotube alignment was rep-
body against tropomyosin (Sigma, T9283, 1:100 in resented as the mean of the standard deviation of the
blocking buffer) or with polyclonal rabbit antibody angles of the myotubes. A lower number thus indic-
against collagen IV (Abcam, ab6586, 1:400 in block- ates better alignment. For endothelial network ana-
ing buffer). To stain capillaries, the explanted BAM lysis, 60 µm z-projections were automatically ana-
constructs were incubated overnight at 4 ◦ C with lyzed by a customized version of the ‘Angiogenesis
a polyclonal rabbit antibody against CD 31 (Invit- Analyzer’, an ImageJ plugin created by Gilles Carpen-
rogen, PA16301, 1:30 in blocking buffer). Next, the tier [10, 30]. Moreover, a customized algorithm was
human BAMs and explants were washed and incub- developed that was able to detect non-perfused and
ated with a polyclonal goat anti-mouse secondary perfused vessels. This tool is able to measure several
antibody (Alexa Fluor 633, A-11 059 or A-21 063, parameters of endothelial networks of which we used:
Invitrogen, 1:200), a polyclonal goat anti-rabbit sec- total length of endothelial network (master segments,
ondary antibody (Alexa Fluor 633, A21070, Thermo branches and isolated segments), % master segments
Fisher) or a polyclonal donkey anti-rabbit secondary (length of master segments, longest elements giv-
antibody (Alexa Fluor 488, A-21 206, Life Technolo- ing origin to branches, divided by total network
gies) for 3 h in the dark, followed by incubation with length), % branching length (length of intercon-
DAPI (Life Technologies, 0.1 µg ml−1 in PBS) for 1 h nected segments and branches, elements delimited
at RT. Samples were stored in PBS in the dark until by two junctions, or a junction and one extremity,
visualized. divided by total network length) and % isolated
segments (length of isolated segments, binary lines
which are not branched, divided by total network
2.7. Confocal imaging length).
Confocal imaging and data analysis of in vitro BAMs
and in vivo BAMs were performed as described 2.10. Detection of anastomosis
in [10]. Briefly, BAMs and explanted tissue were Two different approaches were used to evaluate ana-
visualized by confocal microscopy (Zeiss LSM710). stomosis and perfusion of the implanted endothelial
Per BAM, 5 z-stacks were acquired with Plan- networks with host vessels; in the first approach
Apochromat 25x/0.8, WD 0.57 mm objective every 1.25 µg g−1 body mass rhodamine-labelled Ulex
5 µm. Since explanted BAMs at day 14 were sur- europaeus agglutinine-I (UEA-I; Vectorlabs) was
rounded by host tissue, the distance to the con- injected intravenously in the tail vein, circulation
struct exceeded the working distance of the object- was allowed for 2 min, whereafter the animal was
ive. Therefore, explanted tissue was embedded in euthanized by cervical dislocation. UEA-I specifically
liquid 4% agarose in distilled water, cooled until targets human endothelial cells. Presence of UEA-I
the agarose solidified and cut in 200 µm sections positive human endothelial networks was confirmed
using a vibratome (Microm HM 650 V, VWR) to with confocal imaging of explanted constructs. In
allow visualization within the whole explanted tis- the second approach, human GFP labeled endothelial
sue (in collaboration with Prof. Veerle Baekelandt, networks in the explanted constructs were examined
KU Leuven). More detailed images were acquired for the presence of auto-fluorescent red blood cells by
with Plan-Apochromat 63x/1.2, WD 0.49 mm. Three confocal microscopy.
dimensional reconstructions of stacked images were
generated with Imaris 9.2.0 software (Bitplane). 2.11. Quantitative real-time PCR
RT-qPCR was performed to determine the expres-
sion of the different myosin heavy chain isoforms
2.8. Histology (MYH1, MYH3, MYH8) and myogenin (MyoG) in
4% formaldehyde fixed BAMs and explanted con- the three types of human BAMs to reflect the develop-
structs were dehydrated through an alcohol series and mental state of the myotubes in the tissue-engineered
embedded in paraffin. Tissue was sectioned at a thick- constructs. To normalize for the amount of cells,
ness of 5 µm with a microtome (Leica RM2125 RTS). three reference genes were used: GAPDH, HSP90AB1
Tissue slices were stained with Martius Scarlet Blue and RPL13A. RT-qPCR was performed as described
(MSB) staining [29] to evaluate the extracellular mat- previously [31]. Briefly, BAMs were washed with
rix deposition in BAMs before and after implantation. PBS and lyzed by sonication. After centrifugation
Murine muscle was used as positive control. (5 min, 2600 g), high quality total RNA was isolated

5
Biofabrication 12 (2020) 035021 D Gholobova et al

Table 1. List of primers used for RT-qPCR.

Gene Orientation Primer sequence (5′ —3′ ) Amplicon size (bp)

MYH1 Forward GGG AGA CCT AAA ATT GGC TCA A 106
Reverse TTG CAG ACC GCT CAT TTC AAA
MYH3 Forward CTT GTG GGC GGA GGT CTG G 119
Reverse AGC TAT GCC GAA CAC TTC CAT
MYH8 Forward ACA TTA CTG GCT GGC TGG AC 143
Reverse ACC TTT CTT CGC GCT GCT AT
MyoG Forward GTG TGT AAG AGG AAG TCG GTG TC 90
Reverse GAA GGC CTC ATT CAC CTT CTT
GAPDH Forward TCA AGA AGG TGG TGA AGC AGG 168
Reverse ACC AGG AAA TGA GCT TGA CAA A
HSP90AB1 Forward AGA AAT TGC CCA ACT CAT GTC C 75
Reverse ATC AAC TCC CGA AGG AAA ATC TC
RPL13A Forward CCT GGA GGA GAA GAG GAA AGA GA 126
Reverse TTG AGG ACC TCT GTG TAT TTG TCA A

from supernatant using the PureLink RNA Mini Kit and thickness analyses as mean ± standard deviation.
∗
(Ambion #12 183 018A) as evaluated by A260/A280. p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.
2 µg of isolated RNA was reverse transcribed with the
qScript cDNA SuperMix (Quantabio #95 048-100). 3. Results
PCR was performed with GeneAmp® PCR System
9700 (Applied Biosystems). Gene-specific primers 3.1. In vitro BAM formation
(table 1) were designed using NCBI/Primer-Blast and First, cells isolated from a human muscle biopsy
Primer3plus. Primer efficiency was determined by were expanded and characterized. This muscle cell
serial dilution of cDNA during RT-qPCR and veri- population, previously described in [31] contained
fied to be >90%. RT-qPCR was performed with the 82 ± 5% desmin-positive myoblasts (n = 6) at
LightCycler® 480 Instrument (Roche) using Perfecta 12 doublings. Myoblasts had a fusion index of
Sybr Green Supermix (Quantabio #95 054-500). All 80 ± 15% (n = 6) as determined by tropomy-
samples, 10 µl reaction volume, were run in technical osin staining, forming myotubes with 10 ± 4 nuc-
duplicates. 40 cycles were run with a DNA dissoci- lei per myotube (figure S1 (available online at
ation step (95 ◦ C, 15 s), a primer annealing step and stacks.iop.org/BF/12/035021/mmedia)). To make
an amplification step (60 ◦ C, 45 s). Melting curve bio-artificial muscle with only muscle cells (BAM),
analysis was performed to verify the amplification 2 × 106 muscle cells were mixed in a fibrin extra-
product. Relative mRNA expression was statistically cellular matrix (1 mg ml−1 ) and cultured in skeletal
analyzed with relative expression software tool (REST muscle growth medium (SkGM) for 2 d, followed by
2009, Qiagen). The REST software calculates whether skeletal muscle differentiation medium (SkFM) for
gene expression is significantly different between 12 d. The cell-fibrin hydrogel mix contracted around
the sample and control group using bootstrap two attachment points and formed a 2 cm long
randomization [32]. muscle bundle (schematic representation in figure 1,
macroscopic image in figures 2(a) and (d)), contain-
2.12. Statistics ing aligned multinucleated myotubes (figure 3(a))
Number of replicates refers to number of analyzed with a thickness of 1.25 ± 0.2 mm (n = 21).
images for microscopic analyses or to number of
BAMs for thickness analyses. D’Agostino & Pear- 3.2. In vitro myotube formation in prevascularized
son normality test and Bartlett’s test were used to BAMs
verify normality of the data and equality of variances, Two different prevascularization strategies of BAMs
respectively. Normally distributed data with equal were evaluated. The first approach was previously
variances were analyzed by an unpaired student t-test evaluated in vitro and described in [10]. This co-
when two groups were compared. For comparing sev- culture strategy, here termed the 1-stage prevascu-
eral normally distributed groups, a one-way ANOVA larization (or BAM-1s) was made by co-culturing
was used with a Bonferroni multiple comparison muscle cells with endothelial cells (HUVECs) in a
post test. For groups that were not normally distrib- 70:30 ratio and a total number of 2 × 106 cells in
uted and/or had unequal variances, a non-parametric fibrin hydrogel (schematic representation in figure 1,
Mann-Whitney test was used for comparison between macroscopic image in figure 2(a) right side). As we
two groups while Kruskal-Wallis test followed by a described previously, culturing HUVECs in skeletal
Dunn’s post test was performed for multiple com- muscle media results in poor HUVEC survival, there-
parisons. All values of analyzed images and relative fore we cultured BAM-1s in EGM-2, an optimized
expression data were displayed in Whisker boxplots medium for endothelial cells [10]. After one week

6
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 2. Implantation of BAM, BAM-1s and BAM-2s constructs. Macroscopic images of BAM, BAM-1s and BAM-2s at day 0 of
implantation (a) and (d), at day 5 (b) and (e) and 14 (c) and (f) of post-implantation. BAM constructs were sutured on the left
fascia of the latissimus dorsi muscle, BAM-1s or BAM-2s on the right fascia of the latissimus dorsi muscle. (g), (h) and (i) show
enlarged images of explanted constructs with surrounding host tissue at day 14 post-implantation. Scale bars represent 0.5 cm
(a)–(f) and 0.2 cm (g)–(i).

Figure 3. In vitro myotube formation and prevascularization. Fluorescent confocal images of in vitro tissue-engineered skeletal
muscle (BAM) with only muscle cells (BAM), a one-stage co-culture BAM with GFP labeled HUVECs (green) (BAM-1s) and a
two-stage co-culture BAM with an endothelial coat (BAM-2s). The 3 types of BAMs show presence of aligned multinucleated
(DAPI, blue) myotubes (red) (a), (b) and (d). In BAM-1s, endothelial networks (green) are interspersed between the aligned
myotubes (b). These endothelial networks are positive for collagen IV (c), in red). In BAM-2s, interconnected endothelial
networks are located in the endothelial coat surrounding the myotube bundle (d), (e) and are negative for collagen IV (f). Scale
bars represent 100 µm (a), (b), (d) and (e)) and 50 µm (c) and (f).

in culture, the cell—fibrin hydrogel contracted to a (figure 2(a) right side). No significant difference in
muscle bundle with interspersed endothelial network BAM-1s thickness was present in comparison to
and a total thickness of 1.34 ± 0.15 mm (n = 16) BAM. However, as reported before [10], myotube

7
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 4. Morphological analysis of myotubes in BAMs in vitro and in vivo. Myotube formation and morphology was quantified
with the following parameters: myotube length (a), alignment (b), diameter (c) and myotube density (d). These parameters were
determined in BAM, BAM-1s and BAM-2s at 0, 5 and 14 d post implantation (PI). Myotube analysis was performed based on
tropomyosin staining for all conditions. Statistical significances for comparisons in a BAM condition between different time
points are indicated below the graphs and for comparisons between different conditions at a specific time point are indicated
above the graphs. For multiple comparisons a nonparametric Kruskal-Wallis test with Dunn’s post test corrected for multiple
testing was used. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

formation in BAM-1s is compromised in the EGM in myotube diameter and density compared to
medium resulting in decreased myotube diameter, BAM [10].
length and density compared to BAM (compare To circumvent this compromised myotube form-
figure 3(a) with 3(b) and figures 4(a), (c) and (d). ation, we also explored a two-stage prevascularization
No difference in myotube alignment was observed approach (BAM-2s). First, 2 × 106 muscle cells
(figure 4(b)). RT-qPCR analysis of pre-implantation were mixed with a fibrin hydrogel (1 mg ml−1 ) to
BAM, BAM-1s and BAM-2s for myosin heavy chain engineer a one-week old BAM as described before
(MYH) isoform expression showed a presence of less [10]. Then, we embedded the BAM in a fibrin
mature myotubes (figure 5) in BAM-1s compared to hydrogel containing 2 × 106 HUVEC cells, fur-
BAM reflected by a 2.5 and 4.7-fold lower expres- ther referred to as the endothelial coat. In each
sion level of respectively the adult isoform MYH1 tissue-engineering stage, a 2 × 106 cell ml−1 ratio
and the perinatal isoform MYH8. Also, a respect- was used, as was done with BAM and BAM-1s.
ively 3-fold and 1.4-fold lower expression level was This construct was cultured for another week in
observed for the embryonic isoform MYH3 and myo- EGM-2 to stimulate endothelial network formation
genin, indicating a lower myogenic differentiation in (schematic representation in figure 1, macroscopic
BAM-1s. Earlier reports on the in vitro evaluation image in figure 2(d) right side). Endothelial cells
of this co-culture BAM-1s also described a decrease in this coat layer spontaneously formed endothelial

8
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 5. Myogenic gene expression. Relative expression of MYH isoforms and myogenin in pre-implantation BAMs, direct
co-culture BAMs (BAM-1s) and BAMs with an endothelial coat with HUVECs (BAM-2s). Expression levels of these genes were
quantified by RT-qPCR as the fold changes relative to gene expression in BAMs with only muscle cells (represented with Whisker
box plots). To compare the three conditions, a nonparametric Kruskal-Wallis test with Dunn’s post test corrected for multiple
testing was used. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

networks, and this layer was 117 ± 12 µm thick (e)). In BAM-1s, endothelial networks were inter-
(n = 6) one week after casting (figure 3(e)). The spersed between aligned myotubes. In BAM-2s,
advantage of this procedure is that differentiation the endothelial networks (figure 3(e)) penetrated
of each cell type occurred in its own specialized the myotube bundle occasionally in the transition
medium, resulting in better survival and morpho- area between the BAM and the endothelial coat
logy of each cell type. The resulting BAM-2s was (figure 3(d)). When comparing endothelial networks
1.5 ± 0.15 mm thick (n = 16, figure 2(d)), which was in BAM-1s and BAM-2s, a higher network complex-
significantly thicker than BAM and BAM-1s. After ity could be observed in BAM-2s (figures 3(b), (e)
14 d in culture, we visualized myotube formation and 6). For all measured parameters, BAM-2s out-
using confocal microscopy (figure 3(d)). Myotube performed BAM-1s (figure 6). This means that in
density was significantly higher in BAM-2s com- BAM-2s, the endothelial networks were longer, more
pared to BAM-1s and similar to BAM (figure 4(d)). branched with less isolated segments not particip-
Myotube length and diameter in vitro was highest ating in a network compared to BAM-1s. Although
in BAM (figures 4(a) and (c)). No differences were the endothelial network characteristics were better in
observed for myotube alignment between the 3 dif- BAM-2s than BAM-1s, they were negative for colla-
ferent BAM types (figure 4(b)). Regarding MYH iso- gen IV, a basement membrane protein important in
form expression in BAM-2s, no difference with BAM vascular maturity and stabilization [33] (figure 3(f)).
was observed for MYH1. A lower expression of Networks in BAM-1s were positive for collagen IV
MYH3, MYH8 and myogenin was observed, respect- (figure 3(c)) but they regressed after one week in vitro,
ively 45, 3.2 and 3.3-fold lower than in myotubes in a clear indication of the poor stability of the network
BAM (figure 5). There were no significant differences (figure S2).
in MYH or myogenin expression in BAM-2s versus
BAM-1s (figure 5).
3.4. In vivo implantation and explantation
Next, we evaluated whether different strategies of pre-
3.3. In vitro endothelial network formation in vascularization affected the constructs in terms of
BAM-1s and BAM-2s (i) myotube survival, (ii) myotube maturation and
Endothelial networks were formed through self- (iii) functionality of the endothelial networks in vivo.
assembly and self-organization in BAM-1s and Therefore, we implanted the three BAM types on
BAM-2s fibrin constructs (figures 3(b), (d) and the fascia of the latissimus dorsi muscle on the back

9
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 6. Morphological analysis of endothelial networks in vitro and in vivo in BAMs. Endothelial network formation was
quantified with the following parameters: (a) total length/100 µm2 , (b) % master segments (frequently branching long segments),
(c) % branching and (d) % isolated segments. These parameters were determined in BAM, BAM-1s and BAM-2s at 0, 5 and 14 d
post implantation (PI). Network analysis was performed based on GFP fluorescence for BAM-1s and BAM-2s (containing
GFP-labeled HUVECs) and CD31-staining for BAM. Statistical significances for comparisons in a BAM condition between
different time points, are indicated below the graphs and for comparisons between different conditions at a specific time point are
indicated above the graphs. For multiple comparisons a nonparametric Kruskal-Wallis test with Dunn’s post test corrected for
multiple testing was used. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001.

of NOD/SCID mice. The prevascularized constructs host tissue at day 14. To visualize myotube charac-
(BAM-1s and BAM-2s) were implanted on the right teristics throughout the whole construct at day 14
fascia of the latissimus dorsi muscle, while BAMs PI, the tissue was cut into 200 µm sections with a
(without prevascular networks) were implanted on vibratome.
the left fascia of the latissimus dorsi (figures 2(a)
and (d)) as a reference. Both on day 5 and day 14 3.5. In vivo myotube maintenance
post implantation (PI), differences in vascularization Myotube survival and maturation in vivo were evalu-
between constructs could be observed macroscopic- ated morphologically at 5 and 14 d after implantation.
ally (figures 2(b) and (e), (c) and (f), (h) and (i)). Myotubes were still present at day 14 of implantation
While the appearance of the muscle cell only BAM in all conditions (figure 7). However, compared to
construct did not change (white), BAM-1s and BAM- pre-implantation, a reduction in myotube length
2s were red, pointing to the presence of blood (fig- in all conditions was shown at day 5 which was
ures 2(b), (c) and (e), (f); right side). In contrast even more prominent at day 14 (figure 4(a)). At
to day 5, the different types of BAM were integ- this time point, myotube length was decreased by
rated with host tissue at day 14. Therefore, BAM 40% (BAM), 61% (BAM-1s) and 47% (BAM-2s).
constructs were explanted together with surrounding Similarly, myotube alignment decreased over 14 d,

10
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 7. Confocal images of myotube and endothelial network in BAMs 5 and 14 d post-implantation. Fluorescent confocal
images of musce cell only BAMs explanted after 5 (a) or 14 d (d) in vivo showing myotubes (red), nuclei (DAPI, blue) and
endothelial cells (CD31, green). Images with the same setup are shown for BAM-1s explanted after 5 (b) or 14 (e) days and
BAM-2s explanted after 5 (c) or 14 (f) days. For BAM-1s and BAM-2s, endothelial cells are labeled with GFP. Scale bars represent
100 µm.

Figure 8. In vivo perfusion of prevascularized endothelial networks at 5 and 14 d post-implantation. Fluorescent confocal images
of implanted BAM at day 5 and day 14 with only muscle cells (BAM) and CD 31 positive capillaries (green) (a) and (d), of a
one-stage co-culture BAM with GFP labeled HUVECs (BAM-1s) (green) (b) and (e) and a two-stage co-culture BAM with an
endothelial coat containing GFP labeled HUVECs (green) (BAM-2s) (c) and (f). At day 5 post-implantation prevascularized
BAMs contain GFP labeled endothelial vessels (green) that are perfused with red blood cells (red dots, asterisks (b) and (c)). At
day 14 GFP labeled endothelial vessels (green) are UEA-I positive (orange) (e) and (f), showing perfusion of the newly formed
blood vessels due to anastomosis between host vessels and implanted networks. Muscle cell only BAMs contain some CD31
positive capillaries but no perfusion of the construct with UEA-I at day 14 was observed (d). Scale bars represent 20 µm (a)–(c)
and 100 µm (d)–(f).

11
Biofabrication 12 (2020) 035021 D Gholobova et al

Figure 9. Extracellular matrix remodeling in BAMs. Martius Scarlet Blue (MSB staining) of muscle cell-only BAM, one-stage
co-culture BAM (BAM-1s) and two-stage co-culture BAM (BAM-2s) at day 0 of implantation and day 5 and 14
post-implantation. The extracellular matrix (ECM) of in vitro BAMs consist mainly of fibrin (light pink) (a)–(c) with some
collagen deposition (blue). Aligned myotubes are shown in dark pink/red. At day 5 post-implantation (d)–(f), the ECM still
consists mainly of fibrin, while at day 14 post-implantation, fibrin is being remodeled to collagen. Red blood cells are stained in
orange and are present in capillaries at day 5 and day 14 in BAM-1s and BAM-2s (e)–(f), (h)–(i)). Scale bars represent 20 µm,
dashed line in (c), (f) and (i) indicates the border between the endothelial coat and the muscle bundle.

depicted in figure 4(b) as an increase in misalignment. BAM-1s, endothelial vessels in BAM-2s surrounded
On the other hand, there was an overall decrease from the myotube bundle (figures 7(b) and (c)). However,
day 0 to day 14 PI in myotube diameter for BAM at day 14, vascular structures derived from HUVECs
and BAM-1s conditions, while this was not the case were also observed between myotubes in BAM-2s
for BAM-2s. Finally, myotube density decreased in (figure 7(f)). Although there was a general decrease
all conditions from day 0 to day 14 PI although this in the vascular density in both prevascularized con-
effect was least prominent in the BAM-2s condition structs, pointing to degeneration of part of the net-
(67.8% (BAM), 61% (BAM-1s) and 33,4% (BAM- work (figure 6(a)), the remaining network was more
2s), figures 4(d) and 7(d), (e) and (f)). interconnected (figures 6(c) and (d)). To assess the
functionality of the networks, a rhodamin-labelled
3.6. In vivo vascularization and anastomosis human EC specific agglutinin (UEA-I) was injected
In vivo vascularization of BAMs was assessed 5 and intravenously. Detection of this dye within the BAMs
14 d after implantation. GFP labeled endothelial points to a functional network which is anastomosed
vessels, which originated from the HUVECs, were with the host blood vessels [5]. The latter could also
present in BAM-1s and BAM-2s at day 5 and 14 be verified by the presence of red blood cells in the
(figures 6, 7 and 8). Ingrowth of host blood vessels GFP labeled networks. At day 5, in both BAM-1s and
in muscle cell only BAM was visualized by CD31- BAM-2s, the majority of endothelial vessels (>90%)
staining. There was a higher vascular density, determ- contained red blood cells, while only sparse presence
ined by total network length per 100 µm2 , in the of UEA-I-positive endothelial networks was observed
prevascularized BAMs compared to muscle cell only (figures 8(b) and (c)). This pointed to early anastom-
BAM (figures 6(a) and 7). After 5 d in vivo, the osis with little perfusion. At day 14, a part of the GFP-
location of the endothelial vessels remained sim- labeled endothelial vessels in BAM-1s and BAM-2s
ilar as to the situation in vitro for BAM-1s and were UEA-I positive (figures 8(e) and (f), S3, supple-
BAM-2s: while interspersed between myotubes in mentary Videos S1 and S2) and contained red blood

12
Biofabrication 12 (2020) 035021 D Gholobova et al

cells (figures 8(b) and (c), supplementary Video S3), myotube formation in EGM was also described by
showing functionality of implanted networks . others [12] but is unavoidable in the one-stage
approach since EC survival is severly compromised
3.7. Extracellular matrix remodeling after in SkGM/SkFM. In this follow-up study, the goal was
implantation to test the functionality of the preformed endothelial
Extracellular matrix (ECM) remodeling in vivo was networks after implantation in vivo.
evaluated by Martius Scarlet Blue (MSB) which To avoid the reported suboptimal myotube form-
stains muscle tissue red, fibrin pink/purple, colla- ation in BAM-1s, we established another prevascu-
gen blue and red blood cells orange. The ECM of larization strategy, consisting of two stages, termed
in vitro cultured BAMs mainly consists of fibrin BAM-2s. In the first stage, muscle cell-only BAMs
(figures 9(a)–(c)) with limited deposition of collagen. were cultured for one week in SkGM/SkFM, allowing
No significant ECM remodeling had occurred by day for optimal myotube formation. In a second stage, an
5 PI (figures 9(d)–(f)). The presence of myotubes as endothelial cell coat was added in which endothelial
seen with confocal microscopy (figure 7) was con- networks formed during the second week. Then, the
firmed with MSB staining (figure 9). The presence BAM-2s was cultured in EGM, supporting survival
of blood vessels in BAM-1s and BAM-2s constructs and differentiation of the ECs. This two-stage strategy
was also confirmed with MSB staining; red blood cells indeed resulted in less compromised myotube form-
stained in orange were clearly detectable and located ation, while supporting endothelial network forma-
within the BAM-1s or surrounding the BAM-2s (fig- tion even stronger than in the one-stage approach.
ures 9(e), (f), (h) and (i)). After 14 d in vivo, fib- Also these BAM-2s were tested in an in vivo settting
rin ECM in BAM and BAM-1s was remodeled into in this study.
a collagen ECM (figures 9(g) and (h)). In BAM-2s, a We are aware that there are multiple differences
difference could be observed between the inner part between the one-stage and two-stage prevasculariza-
of the construct containing mainly myotubes with tion protocols since both strategies have been optim-
traces of fibrin and collagen versus the endothelial ized separately to produce the best possible co-culture
coat which showed primarily collagen as ECM BAM. For example, the prolonged culture time of
(figure 9(i)). the muscle cells in the two-stage approach resulted
in better myotube formation, as also reported else-
4. Discussion where [38]. However, applying the same culture time
in the one-stage approach was detrimental for the ECs
At present, the use of muscle flaps to treat volu- since culturing BAM-1s longer than 1 week resulted in
metric muscle loss has significant drawbacks [2]. endothelial network degradation (figure 2(s)).
Tissue engineered skeletal muscle holds promise Vascularization is key for muscle development
for an alternative treatment. The tissue-engineering and function. Therefore, we assessed the myogenic
approach used in this paper further builds on gene expression in the three types of BAMs before
a muscle tissue engineering approach, called bio- implantation using quantitative real-time PCR.
artificial muscle (BAM) containing aligned myotubes Myosin heavy chain isoforms reflect the develop-
[23, 34]. Still, many challenges towards clinical as mental state of muscle fibers [39]. Relative expression
well as in vitro testing applications remain. One of levels of MYH1, MYH3, MYH8 and myogenin were
them is the limited size of the constructs, imposed by analyzed to assess the influence of our vasculariz-
the limit of passive diffusion of nutrients and gases. ation strategy on muscle development. Myogenin
Prevascularization is being explored as a method to expression is indicative for early myogenic differen-
allow in vitro perfusion and to enhance in vivo sur- tiation and myofiber formation [31] while MYH3,
vival and engraftment of tissue-engineered constructs MYH8 and MYH1 are expressed in embryonic, peri-
[12, 35, 36]. In this paper, we have evaluated two pre- natal and adult stages, respectively. Analysis showed
vascularization strategies of tissue-engineered bio- that the significantly higher MYH8, MYH3 and/or
artificial muscle (BAM) to tackle the limitation. myogenin levels indicate a more active myofiber
The first approach was previously evaluated formation in BAM compared to BAM-1s or BAM-
in vitro [10]. In this one-stage prevascularization 2s. This is not surprising given the optimal condi-
approach (BAM-1s) a direct co-culture of endothelial tions for myofiber formation provided by the skeletal
cells and muscle cells resulted in aligned multinuc- muscle fusion medium in this condition. No signific-
leated myotubes with interspersed endothelial net- ant difference in adult isoform MYH1 expression was
works. The natural hydrogel fibrin was used as a observed between the three conditions pointing out
scaffold, because of its pro-angiogenic and stiffness that pre-implantation, the endothelial networks were
characteristics [37]. As reported before [10] myotube not able to promote muscle fiber maturation. The
formation was suboptimal in the endothelial growth BAM-2s approach was explored with the intention to
medium (EGM) in which BAM-1s was cultured improve myogenesis and/or vasculogenesis compared
when compared to BAMs cultured in skeletal muscle to the BAM-1s strategy. Although this was successful,
cell-specific media (SkGM/SkFM). The impaired the BAM-2s still underperformed compared to the

13
Biofabrication 12 (2020) 035021 D Gholobova et al

BAM regarding myogenesis, as shown by the image Optimizing fibroblast—endothelial cell ratios and/or
analysis and qPCR data. This can be explained by evaluation of other mesenchymal-derived cells as
the need to maintain the BAM-2s in EGM for one mural cells may still improve prevascularization out-
week. Myogenic cell lineages are typically cultured in come in the two-stage approach. In this study we have
a high-glucose basal medium with at least 10% FBS used HUVECs as endothelial cell source. HUVECs
while the EGM-2 medium only contains 5% FBS. are easy to isolate, abundant and well characterized.
The negative impact on myogenic cell lineages when Furthermore, they are capable of endothelial net-
defining co-culture media has been described by us work formation in vitro [5, 10, 12]. Other examples
and others [10, 12, 40]. Still, the BAM-2s provided of endothelial cell sources able to form endothelial
the best trade-off between vasculogenesis and myo- networks are microvascular endothelial cells [52],
genesis in a tissue engineered muscle construct. embryonic heart endothelial cells [53], pluripotent
Characterization of endothelial networks in vitro cell derived endothelial cells [54] and blood out-
indicated that the most interconnected networks with growth endothelial cells [55] and are worthwhile for
the highest vascular density were observed in BAM- further investigation in prevascularization strategies.
2s. Indeed, an increased endothelial cell density in To evaluate whether prevascularization improves
the fibrin hydrogel was already described to result in vivo myotube survival and tissue perfusion we
in an enhanced number of branches and total ves- implanted BAMs with and without endothelial net-
sel length [13], and this is consistent with the high works on the fascia of the latissimus dorsi muscle
endothelial cell density in the coat of the BAM-2s of NOD/SCID mice for 5 or 14 d. In all condi-
approach. Also, our previous study showed improved tions, myotubes were still observed at day 14 of
endothelial network formation in vitro when a higher implantation in the construct. However, myotube
density of endothelial cells was used [10]. A well- length, density and alignment decreased in all con-
organized endothelial network in vitro is important ditions in vivo. Myotube length was decreased by
for the in vivo vascularization and tissue integration 40% (BAM), 61% (BAM-1s) and 47% (BAM-2s)
in the host tissues as shown before [5, 19]. One week over 14 d while myotube density was decreased by
after implantation, the extent of vascularization in 67.8% (BAM), 61% (BAM-1s) and 33,4% (BAM-2s).
the co-culture flaps was greater than in the myotube This deterioration in myotube characteristics over
only flaps. Vascularization and efficient perfusion are time could be caused by the initial lack of perfu-
also challenges in cardiac tissue engineering. Perfus- sion. The detrimental effects of this initial hypoxia
able microvascular constructs containing patterned state on the myotubes are not completely counter-
microchannels with endothelial cells integrated with acted, as also reported elsewhere [56]. In addition,
the coronary vasculature of infarcted rat hearts to a the supply of nutrients and oxygen in the first days
greater degree than non-perfusable constructs at 5 d in vivo was limited compared to the in vitro situ-
post-implantation [41]. ation where constructs were completely submerged
Besides analysis of morphological parameters, in a nutrient-rich medium. Finally, the decrease may
collagen IV deposition was assessed. Collagen IV also be explained by the occurrence of xenogen-
is the main component of the vascular basement eic rejection. Although NOD/SCID mice lack func-
membrane matrix. Basement membrane matrix tional T and B cells, have impaired macrophage
assembly is a critical step in vessel maturation and function and lack the complement system [57], a
vessel stability as observed in in vitro and develop- low level of natural killer (NK) cell activity remains
mental studies [33, 42–44]. Endothelial networks present [58–60]. Macrophage, neutrophil and mast
in BAM-1s were surrounded by a thin layer that cell invasion may also result in rejection, as seen with
contained collagen IV, suggesting the presence of implantation of human mesenchymal stem cells on
a basement membrane, in contrast to endothelial poly(ε-caprolactone) in a NOD/SCID model [60].
networks in BAM-2s. In BAM-1s, endothelial cells Immunodeficient animal models are also deprived
were mixed with muscle cells containing both myo- of pro-regenerative effects of the immune response
blasts and fibroblasts [10], in contrast to BAM-2s during skeletal muscle repair [61–63]. Inflammat-
where only endothelial cells were present in the coat ory cells, such as macrophages and neutrophils, are
region. Several groups have successfully shown that recruited and start producing growth factors and
addition of mesenchymal-derived cells, such as fibro- cytokines important for satellite cell activation, pro-
blasts, pericytes and/or smooth muscle cells improved liferation and differentiation and capillary ingrowth
endothelial network maturation and in vitro/in to injured muscle [2]. For clinical application, use
vivo network stabilization in tissue engineered con- of human autologous muscle and endothelial cells
structs [45–49]. These cells are able to differentiate would eliminate xenogeneic responses. In our study,
into supportive mural cells around newly formed the overall decrease in myotube density was least
endothelial networks in co-culture settings with present for the BAM-2s which we hypothesize to
endothelial cells [47, 50, 51]. Pericytes for example are be due to (i) the improved myotube characteristics
known for their role in stimulating basement mem- before implantation compared to the BAM-1s and
brane matrix deposition by endothelial cells [33]. (ii) the pre-vascularization which was absent in the

14
Biofabrication 12 (2020) 035021 D Gholobova et al

BAM. An implantation time exceeding 14 d may even fibrin degradation was observed over two weeks in
improve myotube maturation as reported previously acellular fibrin clots [69] and collagen deposition
[64]. As the ultimate aim was to develop a muscle was shown in tissue engineered skeletal muscle with
tissue engineering strategy with a prevascularization a fibrin ECM [64]. Recent work from our group
approach balancing between optimal vasculogenesis showed that replacement of the degrading fibrin ECM
and myogenesis, the comparison between BAM-1s vs through deposition of autologous collagen enabled
BAM-2s is crucial. We can conclude that for three out the maintenance of the BAM without fibrinolysis
of four measured parameters for myotube character- inhibition, while increasing the visco-elastic proper-
istics, the two-stage prevascularization approach out- ties of the BAMs [70]. This ECM remodeling can be
performed the one-stage approach. attributed primarily to fibroblasts that make up 30%
Next to examining the effects of prevascular- of the cell population in the primary human skeletal
ization on myotube maintenance in vivo, we also muscle cells used for making the BAMs. Furthermore,
evaluated the effect on in vivo vascularization. since electrical stimulation has been demonstrated
Prevascularized BAMs showed a significantly higher to increase collagen deposition as well, maturating
vascular density in vivo compared to BAMs with myotubes may also contribute to collagen deposition
only muscle cells, as reported previously by oth- [71]. The same can be said about ECs which have
ers in tissue-engineered constructs [12, 19]. In this been shown to locally degrade or deposit matrix pro-
study, already at day 5 of implantation anastom- teins upon sprouting [72]. The differences in ECM
osis of the implanted networks with host vessels we observed between the inner myotube part versus
was observed by the presence of red blood cells in the outer endothelial coat of the BAM-2s may be
the human endothelial networks. Anastomosis was explained by the presence of the vasculature. Taken
further confirmed by perfusion of a fluorescent dye together, many cell types may contribute to colla-
injected in the tail vein, which colocalized with the gen deposition and the results obtained in this work
transplanted endothelial cells. Similar to our findings, are a combined effect of the presence of fibroblasts
evidence of anastomosis of endothelial networks in a and the increased vascularization and myogenesis in
fibrin gel (without muscle cells) at day 5 PI was found and around the implants. ECM remodeling should
[13]. This anastomosis speed may not be sufficient for be controlled, since too extensive collagen deposition
oxygen-sensitive tissues such as cardiomyocytes caus- might result in fibrotic scarring. This would ulti-
ing ischemia that leads to cell death [65]. Improving mately interfere with the engraftment and graft func-
the structure and maturation of in vitro networks and tion. In this context, the elimination of an inflam-
optimizing the mesenchymal-derived cell/endothelial matory trigger and resolution of inflammation are
cell ratio as proposed before [19, 35, 47, 49], addition important for preventing fibrosis [73]. Furthermore,
of exogenous growth factors (e.g. VEGF, FGF and characterizing the presence of invading immune cells
PDGF) to promote in vivo vascularization [66] and could help understand the evolution of ECM remod-
optimizing fibrin concentration, type and ultrastruc- eling and engraftment [74, 75]. From our results, no
ture [37, 67] may further reduce anastomosis time sign of fibrotic tissue was observed around implanted
and increase vascularization. Despite anastomosis BAMs on histological sections at day 14. Together
of HUVEC endothelial vessels, shown by red blood with the high vascular density and perfusion at day
cell presence, no perfusion of intravenously injected 14 this proves successful engraftment of BAM-1s and
agglutinin was observed at day 5, suggesting a restric- BAM-2s as opposed to the muscle cell only BAM,
ted perfusion. It has been observed with laser speckle which was not perfused.
imaging that thrombi occurred early after anastom- In conclusion, we tested two prevascularization
osis of implanted endothelial vessels with host vessels strategies of bio-artificial skeletal muscle and evalu-
[68]. By day 7 however, functional perfused vessels ated construct survival and engraftment in the host
were reported [68]. In our experiments, onset of full tissue after implantation in vivo. The two-stage pre-
perfusion as demonstrated with fluorescent agglu- vascularization approach resulted in an endothelial
tinin occurred between day 5 and 14 after implant- network that became functional in vivo and suppor-
ation and showed functionality of the endothelial ted best myotube survival. Although further improve-
networks. No differences were found in endothelial ments can still be done, this strategy opens perspect-
network parameters between the one-stage and the ives for engineering larger skeletal muscle constructs.
two-stage approach constructs.
Fibrin degradation is controlled in vitro by modu-
lating the fibrinolysis through fibrinolysis inhibitors. Acknowledgments
However, in vivo fibrinolysis occurs, which results in
the rapid degradation of fibrin ECM. Nevertheless, The authors are grateful to Stephanie De
fibrin BAMs maintained structural integrity over 14 d Vleeschauwer, Erna Dewil and Marleen Lox,
and fibrin ECM was remodeled to collagen as shown Aernout Luttun, Sander Craps, Greetje Vande Velde,
by MSB staining in constructs explanted after 14 d. Veerle Baekelandt and Joris Van Asselberghs for
This is in line with previous in vivo studies in which technical assistance, Rik Gijsbers for miRFP670

15
Biofabrication 12 (2020) 035021 D Gholobova et al

cDNA-containing HIV1-derived retroviral vec- 2014 An engineered muscle flap for reconstruction of large
tor, Evie Vereecke for human muscle samples, soft tissue defects Proc. Natl. Acad. Sci. USA 111 6010–5
[6] Thorrez L, Shansky J, Wang L, Fast L, VandenDriessche T,
Laurens Willaert and Rune Peyskens for myotube
Chuah M, Mooney D and Vandenburgh H 2008 Growth,
morphological analysis and Sigrid Vanryckeghem differentiation, transplantation and survival of human
for administrative support. This work was funded skeletal myofibers on biodegradable scaffolds Biomaterials
by Fonds Wetenschappelijk Onderzoek—Vlaanderen 29 75–84
[7] Folkman J and Hochberg M 1973 Self-regulation of growth
(1.5.298.17 N and mandate to Lisanne Terrie); and
in three dimensions J. Exp. Med. 138 745–53
AFM-Telethon trampoline grant (19802). [8] Orr A W 2003 Thrombospondin signaling through the
calreticulin/LDL receptor-related protein co-complex
Conflict of Interest stimulates random and directed cell migration J. Cell. Sci.
116 2917–27
[9] Cerino G, Gaudiello E, Grussenmeyer T, Melly L, Massai D,
The authors confirm that there are no known con- Banfi A, Martin I, Eckstein F, Grapow M and Marsano A
flicts of interest associated with this publication and 2016 Three dimensional multi-cellular muscle-like tissue
there has been no significant financial support for this engineering in perfusion-based bioreactors Biotechnol.
Bioeng. 113 226–36
work that could have influenced its outcome.
[10] Gholobova D, Decroix L, Van Mulyder V, Desender L,
Gerard M, Carpentier G, Vandenburgh H and Thorrez L
Data availability 2015 Endothelial network formation within human
tissue-engineered skeletal muscle Tissue Eng. Part A
21 2548–58
The raw data required to reproduce these find- [11] Laschke M W and Menger M D 2015 Prevascularization in
ings are available to download from https:// tissue engineering: current concepts and future directions
data.mendeley.com/datasets/6p47mbpr9f/draft?a=e8 Biotechnol. Adv. 34 112–21
0a2562-a984-4dad-8adb-6bd3060ecac5 and https:// [12] Levenberg S et al 2005 Engineering vascularized skeletal
muscle tissue Nat. Biotechnol. 23 879–84
data.mendeley.com/datasets/8gkvvy8j9h/draft?a =11 [13] Chen X, Aledia A S, Ghajar C M, Griffith C K, Putnam A J,
94a3ca-b02e-4627-b1a5-5283b8698986. The pro- Hughes C C W and George S C 2009 Prevascularization of a
cessed data required to reproduce these find- fibrin-based tissue construct accelerates the formation of
ings are available to download from https://data. functional anastomosis with host vasculature Tissue Eng.
Part A 15 1363–71
mendeley.com/datasets/595jwcnnkm/draft?a=8ff32c [14] Abou-Khalil R, Mounier R and Chazaud B 2010 Regulation
59-4c2f-4312-8c91-18c654ed6a6f. of myogenic stem cell behaviour by vessel cells: the ‘ménage à
trois’ of satellite cells, periendothelial cells and endothelial
Author contributions cells Cell Cycle 9 892–6
[15] Christov C et al 2007 Muscle satellite cells and endothelial
cells: close neighbors and privileged partners Mol. Biol. Cell
DG and LTh contributed to conception and designed 18 1397–409
the project. DG performed in vitro experiments. DG, [16] Post M J, Rahimi N and Caolo V 2013 Update on
KM, LH, LTe, JD were responsible for in vivo exper- vascularization in tissue engineering Regen. Med.
8 759–70
iments. DG, LD, LTe performed data collection. DG, [17] Hanjaya-Putra D, Shen Y-I, Wilson A, Fox-Talbot K, Khetan
Lte, MG and GC performed data analysis and GC per- S, Burdick J A, Steenbergen C and Gerecht S 2013
formed customized software development. DG and Integration and Regression Of Implanted Engineered
MG performed interpretation and statistical analysis. Human Vascular Networks During Deep Wound Healing
Stem. Cells Transl. Med. 2 297–306
DG and LTe wrote the manuscript with input from all [18] Bach A D, Arkudas A, Tjiawi J, Polykandriotis E, Kneser U,
authors. LTh supervised the project. All authors read Horch R E and Beier J P 2006 A new approach to tissue
and approved the manuscript. engineering of vascularized skeletal muscle J. Cell Mol. Med.
10 716–26
[19] Koffler J, Kaufman-Francis K, Yulia S, Dana E, Daria A P,
ORCID iD Landesberg A and Levenberg S 2011 Improved vascular
organization enhances functional integration of engineered
L Thorrez  https://orcid.org/0000-0002-5896-6823 skeletal muscle grafts Proc. Natl. Acad. Sci.
108 14789–94
[20] Higgins S P, Solan A K and Niklason L E 2003 Effects of
References polyglycolic acid on porcine smooth muscle cell growth and
differentiation J. Biomed. Mater. Res. 67A 295–302
[1] Musarò A 2014 The basis of muscle regeneration Adv. Biol. [21] Freiman A, Shandalov Y, Rosenfeld D, Shor E, Ben-David D,
2014 1–16 Meretzki S, Levenberg S and Egozi D 2018 Engineering
[2] Järvinen T A, Järvinen T L, Kääriäinen M, Kalimo H and vascularized flaps using adipose-derived microvascular
Järvinen M 2005 Muscle injuries: biology and treatment Am. endothelial cells and mesenchymal stem cells J. Tissue Eng.
J. Sports Med. 33 745–64 Regen. Med. 12 e130–41
[3] Lee K T and Mun G H 2014 A systematic review of [22] Vandenburgh H, Shansky J, Del Tatto M and Chromiak J
functional donor-site morbidity after latissimus dorsi muscle 1999 Organogenesis of skeletal muscle in tissue culture
transfer Plast. Reconstr. Surg. 134 303–14 Methods Mol. Med. 18 217–25
[4] Cittadella Vigodarzere G and Mantero S 2014 Skeletal [23] Powell C A, Smiley B L, Mills J and Vandenburgh H H 2002
muscle tissue engineering: strategies for volumetric Mechanical stimulation improves tissue-engineered human
constructs Front Physiol. 5 362 skeletal muscle Am. J. Physiol. Cell Physiol. 283 1557–65
[5] Shandalov Y, Egozi D, Koffler J, Dado-Rosenfeld D, [24] Boonen K J M, Rosaria-Chak K Y, Baaijens F P T, van der
Ben-Shimol D, Freiman A, Shor E, Kabala A and Levenberg S Schaft D W J and Post M J 2009 Essential environmental

16
Biofabrication 12 (2020) 035021 D Gholobova et al

cues from the satellite cell niche: optimizing proliferation of its assembly during early development Development
and differentiation Am. J. Physiol. Physiol. 296 C1338–45 131 1619–28
[25] Gholobova D, Gerard M, Terrie L, Desender L, Shansky J, [45] Carosio S, Barberi L, Rizzuto E, Nicoletti C, Del Prete Z and
Vandenburgh H and Thorrez L 2019 Co-culture method to Musarò A 2013 Generation of eX vivo-vascularized Muscle
obtain endothelial networks within human Engineered Tissue (X-MET) Sci. Rep. 3 1420
tissue-engineered skeletal muscle Myogenesis 1889 169–83 [46] Brady M A, Lewis M P and Mudera V 2008 Synergy between
[26] Shcherbakova D, Baloban M, Emelyanov A, Brenowitz M, myogenic and non-myogenic cells in a 3D tissue-engineered
Guo P and Verkhusha V 2016 Bright monomeric craniofacial skeletal muscle construct J. Tissue Eng. Regen.
near-infrared fluorescent proteins as tags and biosensors for Med. 2 408–17
multiscale imaging Nat. Commun. 7 1–12 [47] Jeon J S, Bersini S, Whisler J A, Chen M B, Dubini G, Charest
[27] Geraerts M, Micheils M, Baekelandt V, Debyser Z and J L, Moretti M and Kamm R D 2014 Generation of 3D
Gijsbers R 2005 Upscaling of lentiviral vector production by functional microvascular networks with human
tangential flow filtration J. Gene Med. 7 1299–310 mesenchymal stem cells in microfluidic systems Integr. Biol.
[28] Ibrahimi A et al 2009 Highly efficient multicistronic 6 555–63
lentiviral vectors with peptide 2A sequences Hum. Gene [48] Sacharidou A, Stratman A N and Davis G E 2011 Molecular
Ther. 20 845–60 mechanisms controlling vascular lumen formation in
[29] Lendrum A C, Fraser D S, Slidders W and Henderson R 1962 three-dimensional extracellular matrices Cells Tissues Organs
Studies on the character and staining of fibrin J. Clin. Pathol. 195 122–43
15 401–13 [49] Gökçinar-Yagci B, Uçkan-Çetinkaya D and Çelebi-Saltik B
[30] Carpentier G 2012 Angiogenesis Analyzer for ImageJ—Gilles 2015 Pericytes: properties, functions and applications in
Carpentier Research tissue engineering Stem. Cell Rev. Rep. 11 549–59
(http://image.bio.methods.free.fr/ImageJ/?Angiogenesis- [50] Ding R, Darland D C, Parmacek M S and Amore P A D 2004
Analyzer-for-ImageJ) Endothelial-mesenchymal interactions in vitro reveal
[31] Gholobova D, Gerard M, Decroix L, Desender L, Callewaert molecular mechanisms of smooth muscle/pericyte
N, Annaert P and Thorrez L 2018 Human tissue-engineered differentiation Stem. Cells Dev. 13 509–20
skeletal muscle: a novel 3D in vitro model for drug [51] Lesman A, Koffler J, Atlas R, Blinder Y J, Kam Z and
disposition and toxicity after intramuscular injection Sci. Levenberg S 2011 Engineering vessel-like networks within
Rep. 8 1–14 multicellular fibrin-based constructs Biomaterials
[32] Pfaffl M W, Horgan G W and Dempfle L 2002 Relative 32 7856–69
expression software tool (REST) for group-wise comparison [52] Nör J E, Peters M C, Christensen J B, Sutorik M M, Linn S,
and statistical analysis of relative expression results in Khan M K, Addison C L, Mooney D J and Polverini P J 2001
real-time PCR Nucleic Acids Res. 30 36 Engineering and characterization of functional human
[33] Stratman A N, Malotte K M, Mahan R D, Davis M J and microvessels in immunodeficient mice Lab. Invest. 81 453–63
Davis G E 2009 Pericyte recruitment during vasculogenic [53] Van der Schaft D W J, Van Spreeuwel A C C, Van Assen H C
tube assembly stimulates endothelial basement membrane and Baaijens F P T 2011 Mechanoregulation of
matrix formation Blood 114 5091–101 vascularization in aligned tissue-engineered muscle: a role
[34] Vandenburgh H, Shansky J, Benesch-Lee F, Barbata V, Reid J, for vascular endothelial growth factor Tissue Eng. Part A
Thorrez L, Valentini R and Crawford G 2008 Drug-screening 17 2857–65
platform based on the contractility of tissue-engineered [54] Sriram G, Tan J Y, Islam I, Rufaihah A J and Cao T 2015
muscle Muscle Nerve 37 438–47 Efficient differentiation of human embryonic stem cells to
[35] Costa-Almeida R, Granja P L, Soares R and Guerreiro S G arterial and venous endothelial cells under feeder- and
2014 Cellular strategies to promote vascularisation in tissue serum-free conditions Stem. Cell Res. Ther. 6 261
engineering applications Eur. Cells Mater. 28 51–67 [55] Van Beem R T, Verloop R E, Kleijer M, Noort W A, Loof N,
[36] Laschke M W and Menger M D 2012 Vascularization in Koolwijk P, Van der Schoot C E, Van Hinsbergh V W M and
tissue engineering: angiogenesis versus inosculation Eur. Zwaginga J J 2009 Blood outgrowth endothelial cells from
Surg. Res. 48 85–92 cord blood and peripheral blood: angiogenesis-related
[37] Morin K T and Tranquillo R T 2013 In vitro models of characteristics in vitro J. Thromb Haemost 7 217–26
angiogenesis and vasculogenesis in fibrin gel Exp. Cell Res. [56] Martin N R W, Aguilar-Agon K, Robinson G P, Player D J,
319 2409–17 Turner M C, Myers S D and Lewis M P 2017 Hypoxia
[38] Madden L, Juhas M, Kraus W E, Truskey G A and Bursac N impairs muscle function and reduces myotube size in tissue
2015 Bioengineered human myobundles mimic clinical engineered skeletal muscle J. Cell. Biochem. 118 2599–605
responses of skeletal muscle to drugs Elife 4 1–14 [57] Shultz L et al 1995 Multiple defects in innate and adaptive
[39] Wang L, Shansky J and Vandenburgh H 2013 Induced immunologic function in NOD/LtSz-scid mice J. Immunol.
formation and maturation of acetylcholine receptor clusters 154 180–91
in a defined 3D bio-artificial muscle Mol. Neurobiol. [58] Christianson S 1997 Enhanced human CD4+ T cell
48 397–403 engraftment in beta2-microglobulin-deficient NOD-scid
[40] Wragg N M, Mosqueira D, Blokpeol-Ferreras L, Capel A, mice J. Immunol. 158 3578–86
Player D J, Martin N R W, Liu Y and Lewis M P 2019 [59] Yoshino H et al 2000 Natural killer cell depletion by
Development of a 3D tissue-engineered skeletal muscle and anti-asialo GM1 antiserum treatment enhances human
bone co-culture system Biotechnol. J. 15 1900106 hematopoietic stem cell engraftment in NOD/Shi-scid mice
[41] Redd M A, Zeinstra N, Qin W, Wei W, Martinson A, Wang Y, Bone Marrow Transplant. 26 1211–6
Wang R K, Murry C E and Zheng Y 2019 Patterned human [60] Xia Z, Ye H, Choong C, Ferguson D J P, Platt N, Cui Z and
microvascular grafts enable rapid vascularization and Triffitt J T 2004 Macrophagic response to human
increase perfusion in infarcted rat hearts Nat. Commun. 10 mesenchymal stem cell and poly(ε-caprolactone)
584 implantation in nonobese diabetic/severe combined
[42] Jain R K 2003 Molecular regulation of vessel maturation Nat. immunodeficient mice J. Biomed. Mater. Res. Part A
Med. 9 685–93 71 538–48
[43] Kalluri R 2003 Basement membranes: structure, assembly [61] Laumonier T and Menetrey J 2016 Muscle injuries and
and role in tumour angiogenesis Nat. Rev. Cancer strategies for improving their repair J. Exp. Orthop 3 15
3 422–33 [62] Juhas M, Abutaleb N, Wang J T, Ye J, Shaikh Z, Sriworarat C,
[44] Poschl E, Brachvogel U, Schlötzer-Schrehardt B, Saito K, Qian Y and Bursac N 2018 Incorporation of macrophages
Ninomiya Y and Mayer U 2004 Collagen IV is essential for into engineered skeletal muscle enables enhanced muscle
basement membrane stability but dispensable for initiation regeneration Nat. Biomed. Eng. 2 942–54

17
Biofabrication 12 (2020) 035021 D Gholobova et al

[63] Sadtler K et al 2016 Developing a pro-regenerative [69] Wolbank S, Pichler V, Crawford J F, Meinl A, van Griensven
biomaterial scaffold microenvironment requires T helper 2 M, Goppelt A and Redl H 2015 Non-invasive in vivo
cells Science 352 366–70 tracking of fibrin degradation by fluorescence imaging J.
[64] Gilbert-Honick J, Iyer S R, Somers S M, Lovering R M, Tissue Eng. Regen. Med. 9 973–6
Wagner K, Mao H Q and Grayson W L 2018 Engineering [70] Thorrez L, DiSano K, Shansky J and Vandenburgh H 2018
functional and histological regeneration of vascularized Engineering of human skeletal muscle with an autologous
skeletal muscle Biomaterials 164 70–79 deposited extracellular matrix Front Physiol. 9 1–11
[65] Zhang M, Methot D, Poppa V, Fujio Y, Walsh K and Murry C [71] Rahimi N, Swennen G, Verbruggen S, Scibiorek M, Molin D
E 2001 Cardiomyocyte grafting for cardiac repair: graft cell G and Post M J 2014 Short Stimulation of
death and anti-death strategies J. Mol. Cell. Cardiol. Electro-Responsive PAA/Fibrin Hydrogel Induces Collagen
33 907–21 Production Tissue Eng. Part C: Methods 20 703–13
[66] Kaully T, Kaufman-Francis K, Lesman A and Levenberg S [72] Daub J T and Merks R M H 2013 A Cell-Based Model of
2009 Vascularization-the conduit to viable engineered tissues Extracellular-Matrix-Guided Endothelial Cell Migration
Tissue Eng. Part B: Rev. 15 159–69 During Angiogenesis Bull. Math. Biol. 75 1377–99
[67] Kaijzel E L, Koolwijk P, Van Erck M G M, Van Hinsbergh V [73] Van Linthout S, Miteva K and Tschöpe C 2014 Crosstalk
W M and De Maat M P M 2006 Molecular weight fibrinogen between fibroblasts and inflammatory cells Cardiovasc. Res.
variants determine angiogenesis rate in a fibrin matrix in 102 258–69
vitro and in vivo J. Thromb Haemost [74] Bloise N et al 2018 Ether-oxygen containing electrospun
4 1975–81 microfibrous and sub-microfibrous scaffolds based on
[68] White S, Pittman C, Hingorani R, Arora R, Esipova T, poly(butylene 1,4-cyclohexanedicarboxylate) for skeletal
Vinogradov S, Hughes C, Choi B and George S 2014 muscle tissue engineering Int. J. Mol. Sci. 19 3212
Implanted cell-dense prevascularized tissues develop [75] Yuan T, Xiao Y, Fan Y, Liang J and Zhang X 2017 The
functional vasculature that supports reoxygenation after degradation and local tissue effects of collagen hydrogel and
thrombosis Tissue Eng. Part A 20 2316–28 sponge implants in muscle Polym. Test 62 348–54

18