Biomaterials 31 (2010) 8889e8901

Contents lists available at ScienceDirect

Biomaterials
journal homepage: www.elsevier.com/locate/biomaterials

Tissue-engineered conduit using urine-derived stem cells seeded bacterial

cellulose polymer in urinary reconstruction and diversion
Aase Bodin a, b, Shantaram Bharadwaj b, Shaofeng Wu b, Paul Gatenholm a, b, Anthony Atala b,
Yuanyuan Zhang b, *
a
Chalmers University of Technology, BBV laboratory, Sweden
b
Wake Forest Institute for Regenerative Medicine, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157, USA

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

Article history: The objective of this study was to generate bacterial cellulose (BC) scaffolds seeded with human urine-
Received 17 May 2010 derived stem cells (USC) to form a tissue-engineered conduit for use in urinary diversion. Microporous BC
Accepted 30 July 2010 scaffolds were synthesized and USC were induced to differentiate into urothelial and smooth muscle cells
Available online 25 August 2010
(SMC). Induced USC (106 cells/cm2) were seeded onto BC under static and 3D dynamic (10 or 40 RPM)
conditions and cultured for 2 weeks. The urothelial cells and SMC derived from USC formed multilayers
Keywords:
on the BC scaffold surface, and some cells infiltrated into the scaffold. The urothelium derived from USC
Tissue-engineered urinary conduit
differentiation expressed urothelial markers (uroplakin Ia and AE1/AE3) and the SMC expressed SMC
Mesenchcymal stem cells
Urinary tract
markers (a-smooth muscle actin and desmin). In addition, USC/BC scaffold constructs were implanted
Cellematrix infiltration into athymic mice, and the cells were tracked using immunohistochemical staining for human nuclear
Cellulose antigen. In vivo, the cells appeared to differentiate and express urothelial and SMC markers. In conclu-
sion, porous BC scaffolds allow 3 dimensional growth of USC, leading to formation of a multilayered
urothelium and cellematrix infiltration. Thus, cell-seeded BC scaffolds hold promise for use in tissue-
engineered urinary conduits for urinary reconstruction.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction
Currently, bladder cancer is the second most common urologic
malignancy in the United States, after prostate cancer. According to
the American Cancer Society, almost 71,000 new cases of bladder
cancer were diagnosed in 2009, and 14,000 patients with this type
of cancer died [1]. The chance of developing bladder cancer is about
1 in 27 for men and 1 in 85 for women. In order to treat malig-
nancies that have invaded the bladder muscle, surgical resection of
the tumor, followed by the creation of a continent urinary reservoir
using segments of the small or large intestine is often necessary. A
number of surgical techniques have been used to perform urinary
diversion, but the most common techniques are continent diver-
sion and conduit diversion. In continent diversion, a continent
stomal reservoir with a segment of intestine is created for cathe-
terization, and this allows the patient to empty the reservoir via
catheter as needed throughout the day. In conduit diversion
a section of intestine, usually ileum is used to create a conduit
that collects urine and allows it to drain through a stoma in the
* Corresponding author. Tel.: þ1 336 713 1189; fax: þ1 336 713 7290.
E-mail address: yzhang@wfubmc.edu (Y. Zhang).
abdominal wall. The urine is then collected in a bag outside the
body, and the bag must be emptied periodically. However, both
continent diversion and conduit diversion can cause several
potential complications, including hyperchloremic metabolic
acidosis with hypokalemia, urinary tract infections, stone forma-
tion, malignancy, bowel obstruction, skin breakdown around the
stoma, stenosis of the stoma, impaired renal function, or damage to
the upper urinary tract resulting from urine reflux.
Tissue engineering technology may provide an alternative
approach for building a functional urinary conduit to store urine for
patients with bladder cancer who require total cystectomy. Various
biodegradable scaffolds seeded with cells have been introduced for
bladder reconstruction or urinary conduit procedures. Most of
these biomaterials are biodegradable, including natural collagen
materials i.e. bladder submucosa, (BSM) [2], small intestine
submucosa (SIS) [3] or collagen type I matrix [4] and synthetic
polymers such as polyglycolic acid (PGA), poly (lactic-co-glycolic
acid) (PLGA)[5,6] and biocarbon [7]. Most degradable biomaterials,
promote cellular interaction and tissue development, and possess
adequate mechanical and physical properties. However, natural
collagen scaffolds cannot maintain a robust physical structure in an
in vivo environment when used in total or subtotal bladder
replacement, resulting in graft collapse, contraction, formation of

0142-9612/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biomaterials.2010.07.108

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8890 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901
fibrosis, and shrinkage of the new bladder, with resultant decreased
bladder capacity [8]. A biomaterial with a retained hollow struc-
ture, anti-fibrosis properties, and a 3D porous microstructure for
graft cell seeding would be highly desirable for creating a viable
tissue-engineered urinary conduit.
Tissue-engineered urinary conduits and tissue-engineered
bladders are similar in structure, but different in function. Both
constructs require a hollow structure lined with urothelium on
the lumen side. A new bladder or conduit covered with urothelial
tissue can prevent stone, calcification, or scar formation caused by
the naked muscle layer or biomaterial being directly exposed to
urine [9,10]. However, tissue-engineered urinary conduits are
mainly used to store urine and they are emptied via catheterization
in continent diversion or a channel in conduit diversion. Unlike
engineered bladder tissue, an engineered conduit does not need to
contract to empty itself, and thus the development of a thick
muscular wall is not required. Furthermore, tissue-engineered
bladders must be constructed using a degradable biomaterial to
allow the necessary muscle contractility to develop, but this type of
contractility is not required for the development of functional
tissue-engineered conduits. Although either resorbable or non-
resorbable materials can be used in tissue-engineered urinary
conduits, a non-resorbable bacterial cellulose material can be an
attractive alternative.
Bacterial cellulose is a natural biopolymer and consists of
a network of nanofibrils which possess high mechanical strength
[11]. This material is highly hydrophilic, non-cytotoxic, and stable
within a wide range of temperatures and pH levels. Although
bacterial cellulose polymer is non-degradable, it is an FDA-
approved material used clinically in wound dressing [12], and it has
been evaluated to repair cartilage, provide drug delivery, and
replace dura mater or small-caliber blood vessels [13e16]. Thus,
this biomaterial might be a good candidate as scaffold for fabri-
cating a urinary conduit.
Another critical factor in generating tissue-engineered urinary
conduits for patients with bladder cancer is the cell source. Tradi-
tionally, cells obtained from bladder biopsies are a main cell source
for cell-based tissue engineering in urinary tract reconstruction.
However, healthy cells might not be available in patients with
bladder cancer or chronic inflammation. Cancer-free urothelial cells
or stem cells from the upper urinary tract provide an alternative
cell source. Our recent study demonstrated that urine-derived cells
can be collected from the upper urinary tract. These cells possess
stem cell features, including self-renewal and multipotential
differentiation [unpublished data]. Urine-derived stem cells from
the upper urinary tract can be differentiated into urothelial and
smooth muscle cells. It would be a simple and low-cost approach to
harvest cells from patients who already have a nephrostomy tube in
place.
The goal of this study was to produce a tissue-engineered
urinary conduit structure with 3D urothelial mucosa using autol-
ogous urine-derived stem cells, cultured within 3D porous bacterial
cellulose under dynamic culture conditions. This cell-based tissue-
engineered conduit may be useful for patients with end-stage
bladder diseases who need bladder reconstruction.
2. Materials and methods
2.1. Production of 3-D porous bacterial cellulose
Bacterial cellulose scaffolds with various pores sizes were prepared by adding
sterile paraffin particles of 3 different size ranges (90e150 mm, 150e300 mm and
300e500 mm) to a tubular fermentation vessel containing Acetobacter xylinum [17].
After seven days of bacterial culture, porous cellulose tubes formed. These were
harvested and purified until no paraffin was observed under light microscopy or
infrared spectroscopy. The tubes were cut open to form a bacterial cellulose polymer
sheet 5 mm in thickness. The sheets were mounted on an inserter to yield a culture
Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
surface of 1.6 cm2 [18]. These cellulose scaffolds were sterilized by autoclaving in
deionised water at 120  C, 1 bar for 20 min and were stored at 4  C until needed.
2.2. Endotoxin analysis
Bacterial cellulose samples with various pore sizes were analyzed for endotoxin
using the LAL test. Testing was done by the accredited clinical bacteriology labora-
tory at Sahlgrenska University Hospital (Gothenburg, Sweden) according to SS-EN
ISO 15189 (Medical laboratories e Particular requirements for quality and compe-
tence). One gram (wet weight) of bacterial cellulose was cut into 1  1 cm fragments
and then autoclaved (120  C, 20 min) in 40 ml of pyrogen-free water. Sterile
pyrogen-free water served as the control. The water was removed from each
cellulose sample, mixed with the reconstituted LAL reagent, placed in the incubating
spectrophotometer, and automatically monitored over time for the appearance of
turbidity. The assay was measured at 340 nm wavelength and had a sensitivity range
of 0.005e100.0 EU/ml.
2.3. Scanning electron microscopy
3D porous bacterial cellulose scaffolds were prepared with different pore size
ranges using paraffin particles as porogens. These cellulose scaffolds were frozen in
liquid nitrogen and freeze-dried for 24 h at 52  C (Heto PowerDry PL3000). The
dried material was coated with gold and mounted for analysis. Pore size, intercon-
nections between pores, and the cross-sectional area of bacterial cellulose were
assessed using electron microscopy performed with a Zeiss DSM 940A electron
microscope (Zeiss, Germany) operated at 10 kV.
2.4. Tensile testing
Tensile strength and elastic modulus were measured using an Instron 5565A
instrument (Norwood, MA) with a 100N load cell. Porous bacterial cellulose was cut
into strips with uniform dimensions (0.2 mm in thickness, 3 cm in length, 12 mm in
width). The exact thickness of each sample was measured using a slide caliper and
imported into the material testing software (Bluehill 2, Norwood, MA). The strips
were clamped to the instrument between pieces of sandpaper to prevent slipping.
The crosshead speed was set at 5 mm/s and the test was stopped when the force
decreased by 40% after the onset of failure. Zero stress was recorded when the strip
of cellulose was stretched. From the stressestrain curves, Young’s modulus was
calculated at 20% strain for all samples [11]. All measurements were performed in
distilled water at 37  C. Seven specimens from each of the 4 different pore size
groups (5 mm (control), 90e150 mm, 150e300 mm and 300e500 mm) were tested.
2.5. Cell harvesting
The Wake Forest University Health Sciences Institutional Review Board
approved the use of tissue from human ureters and human urine samples in this
study. Urothelial and smooth muscle cells were harvested from sections of surgically
discarded ureter tissue. The tissue was digested with collagenase type IV at 37  C for
30 min. Urothelial cells were scraped from the ureter mucosa and cultured in ker-
atinocyte-serum free medium (KFSM) supplemented with bovine pituitary extract,
epidermal growth factor (EGF), and cholera toxin. For isolation of smooth muscle
cells, the remaining ureter was minced into pieces (approximately 0.1 cm3 each)
after connective tissues were removed. These pieces of tissue were plated in 100 mm
tissue culture dishes with high glucose DMEM containing 10% fetal bovine serum
(FBS). Culture medium was changed every 2 days until cells reached 80e90%
confluence.
To isolate human urine-derived cells, urine was collected in the clinic via
nephrectomy tubes or open renal pelvic surgery and was immediately transferred to
the laboratory for cell isolation and culture using previously reported techniques
[19]. Briefly, four urine samples (2.5e10 ml) from the upper urinary tract of patients
(17 months-18 years) were used. Each urine sample was centrifuged at 500g for
5 min and the supernatant was removed. The cell pellet was gently resuspended in
3 ml of mixed media composed of KSFM and embryo fibroblast medium (EFM) (1:1
ratio) [19] and plated in 24 well-plates at 1  104 cells/well. Individual clones
appeared 3e5 days after implantation. Each clone was plated into a single well of
a 24-well dish, and when the well reached 70e80% confluency, the cells were
trypsinized and transferred into 6-well dishes. When the cultures reached 70e80%
confluency once more, the cells were transferred to a 100 mm culture dish for
expansion. Before seeding the cells onto the scaffolds their viability was assessed
using tryptan blue. Less than 1% of the cells were seen to be non-viable.
2.6. Characterization of urine-derived cells from upper urinary tract
To determine the phenotype of the urine-derived cells, immunocytochemistry
for the pericyte/mesenchymal stem cell surface markers monoclonal rabbit CD146
(abcam, Cambridge, MA) monoclonal, mouse NG2 (abcam, Cabridge, MA), rabbit
PDGF-rb (abcam, Cambridge, MA) monoclonal, mouse STRO-1(R&D, Minneapolis,
MN), goat polyclonal CD31 (Santa cruz, Santa Cruz, CA) and monoclonal mouse CD73
(abcam, Cambridge, MA) was performed. Briefly, the cells were cultured in chamber
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901
slides until 50% confluence was reached, and then the cells were fixed in 3.7%
paraformaldehyde for 15 min. Non-specific binding was blocked using a serum-free
blocker (Dako). Cells were then incubated overnight with 1:100 dilution of each
primary antibody followed by 1:200 dilution of FITC secondary antibody against
rabbit or mouse for 1 h at room temperature. Finally the cells were mounted in PI
mounting media (Vector Laboratories, Burlingame, CA) before visualization under
fluorescent microscope.
2.7. Urothelial and smooth muscle differentiation
Within 3 weeks of culture, 1  106 cells were expanded from a single urine-
derived clone collected from the upper urinary tract. The cells were then induced to
differentiate into urothelial and smooth muscle cells using specific differentiation
media. Differentiation to a urothelial phenotype was performed in medium
composed of KSFM supplemented with 2% FBS and 30 ng/ml EGF. Differentiation to
a smooth muscle phenotype was induced with myogenic differentiation medium
composed of DMEM and EFM with 10% FBS, 2.5 ng/ml transforming growth factor
b (TGF-b1) and 5.0 ng/ml platelet-derived growth factor BB (PDGF-BB) for 14 days.
2.8. In vitro seeding of porous bacterial cellulose
In order to determine which scaffold pore size and which shaker speed was
optimal for cell attachment and growth, urine-derived cells were seeded at a density
of 1  106/cm2 onto cellulose scaffolds with pore sizes of 5 mm (control),
90e150 mm, 150e300 mm, and 300e500 mm. The seeded scaffolds were placed onto
an orbital shaker (Belly Button, Stovall Life Science Inc., Greensboro, NC) at
a constant inclination and speed (either 10 or 40 rpm) [18]. After 7 days, the cell-
seeded scaffolds were processed for histology (hematoxylin and eosin [H&E],
40 ,6-diamidino-2-phenylindole [DAPI], and Masson’s trichrome stains). Urothelial
and smooth muscle differentiation of urine-derived stem cells were evaluated with
immunohistochemical staining. Briefly, smooth muscle cells were identified using
a-smooth muscle actin monoclonal antibody (Sigma, St Louis, MO) and desmin
mouse antibody (D33, DakoCytomation, Carpinteria, CA). To identify urothelial cells,
a monoclonal antibody to cytokeratins AE1/AE3 (DakoCytomation, Carpinteria, CA)
and a goat polyclonal uroplakin-Ia antibody (C-18, Santa Cruz Biotechnology) were
used. The antibodies were hybridized to biotinylated anti mouse (AE1/AE3), bio-
tinylated anti goat (Uroplakin), rhodamine conjugated anti mouse (desmin and
actin). DAB (peroxidise substrate kit) was used for the uroplakin and AE3/AE1.
Fig. 1. Morphology of three different porous bacterial cellulose matrixes. Scanning electron micrographs (SEM) of the surface of 90e150 mm (upper row left), 150e300 mm
(upper row middle) and 300e500 mm (upper row right) bacterial cellulose matrices. Cross sections of 90e150 mm (second row left), 150e300 mm (second row middle) and
300e500 mm (second row right) cellulose matrixes. Images at 100 and 50 respectively.
Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8891
All samples were mounted with mounting media containing DAPI (Vector Labora-
tories, Burlingame, CA).
Layered co-cultures of urothelial cells and smooth muscle cells derived from the
cultures of urine-derived stem cells were prepared as described previously [20].
Briefly, smooth muscle cells were seeded onto the matrix and 12e24 h later, uro-
thelial cells were seeded on top of the smooth muscle cell layer. The co-culture was
incubated in static conditions for one day. Then, the co-cultures were transferred to
rotating dynamic conditions (10 or 40 RPM) [18] and maintained this way for 7 days.
The cell-seeded cellulose scaffolds were then harvested and processed for histology
and immunohistochemistry.
2.9. Implantation of cell-seeded scaffolds in vivo
Urine-derived stem cells were seeded onto porous cellulose at a density of
1  106/cm2 and these scaffolds were maintained in either static or 3D dynamic
culture conditions for 7 days. The cell-seeded bacterial cellulose scaffolds were then
implanted subcutaneously into 6-week-old female athymic mice under aseptic
conditions, as described previously [21,22]. The cell-seeded grafts, 1.6 cm2, 0.2 mm
thickness, were folded once facing the seeded cells outwards and implanted into the
right and left side of each mouse (n ¼ 3). The retrieved scaffold sizes were 1.6 cm2,
0.4 mm thickness. As a control, unseeded porous bacterial cellulose was implanted
in an identical fashion (n ¼ 3). The skin incision was closed with 6-0 Ethilon suture
material. The grafts were harvested one month after implantation, fixed in 10%
neutral buffered formalin, dehydrated and embedded in paraffin. Cellular infiltration
and the number of cell layers in each graft were assessed using hematoxylin and
eosin staining, DAPI staining, and Masson’s trichrome staining on 5 mm sections.
To identify human cells within the grafted tissues, immunohistochemical
staining was performed using a monoclonal antibody specific for human nuclei
(Calbiochem, San Diego, CA) along with DAPI. Immunohistochemical staining was
performed as described to identify the phenotypes of the cells present in each graft.
2.10. Statistical analysis
Statistical analysis was carried out on the measurements of the mechanical
properties of each scaffold. All values are shown as the mean  standard deviation
(SD). Comparisons between different groups were made using one-way analysis of
variance (ANOVA) with a Turkey test for significance. Differences were considered
8892 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901

study are well below this threshold value and thus, these scaffolds
should not induce complications such as fever in vivo.

3.2. Scanning electron microscopy

The porous walls of the bacterial cellulose scaffolds are

composed of a network of nanofibrils and spherical interconnected
pores. We have previously shown that the diameter of these
nanofibrils ranges from 50 to 500 nm, which is similar to that of
natural collagen fibres [17,24]. The structure of the interconnected
pores in 3 different samples (90e150 mm, 150e300 mm, and
300e500 mm in pore size) is shown in Fig. 1.

3.3. Tensile testing

Our measurements indicate that as pore size increases, the

Fig. 2. Young’s modulus of porous BC material decreases with increase in pore size.
A statistical significance between the 5 mm and 90e150 mm samples was observed
when compared with the larger pore sizes, P < 0.05. However, there were no signif-
icant differences between the larger pore sizes (150e300 mm versus 300e500 mm),
P > 0.05.
Young’s modulus of the scaffold decreases (Fig. 2). Young’s modulus
was significantly different between the control group (5 mm) and
each group of porous scaffold (P < 0.05). However, there were no
statistically significant differences in Young’s modulus between the
group containing pores of 150e300 mm and the group containing
pores of 300e500 mm in size (p > 0.05).
3.4. Pericyte/mesenchymal stem cell surface marker expression

significant at p < 0.05. All statistical calculations were performed using GraphPad
Prism software.
3. Results
3.1. Endotoxin results
The endotoxin levels in the water incubated with the porous
cellulose scaffold were measured to be 0.1  0.04 endotoxin units
(EU)/ml. The limit for endotoxin set by the FDA for medical devices
is 0.5 EU/ml (based on a 40 mL rinse) [23]. This data shows that the
endotoxin levels in the porous bacterial cellulose generated in this
A single urine-derived stem cell clone could be expanded to one
million cells in 3 weeks. Once adequate cell numbers were reached,
the cells were induced to differentiate into urothelial and smooth
muscle cells using cell lineage-specific inductive media. With time,
the morphology of the cells changed from a compact oval shape
into a cobblestone shape when the cells were placed into urothe-
lial-specific media, and into a spindle shape when placed in smooth
muscle-specific media (Fig. 3).
Immunocytochemistry revealed that urine-derived stem cells
(p4) isolated from the upper urinary tract expressed several
established pericyte/mesenchymal stem cell surface markers
(Fig. 4). Cells expressed specific and intense membrane staining for

Fig. 3. Morphology of urine-derived stem cells obtained from upper urinary tract (USC-UUT) with differentiation. Crystal violet staining of a single clone of USC-UUT prior to and
after differentiation to myogenic (USC-SMC) and urothelial cell (USC-UC) lineage for 14 days. The shape of non-treated USC-UUT changed from an oval to a spindle shape with the
addition of myogenic medium and to a cuboidal shape with the addition of uro-epithelial medium. Scale bar shown is 100 mm.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901 8893

Fig. 4. Characterization of urine-derived stem cells from upper urinary tract (USC-UUT) for expression of pericyte markers. Immunofluorescence staining of USC-UUT (p4) for
pericyte/mesenchymal markers (a) CD146, (b) NG2, (c) PDGF-rb, (d) CD31, (e) STRO-1 and (f) CD73. Specific membrane staining of markers (green) is shown by arrows. Cell nucleus
was counterstained red using PI. Images are at 400 and scale bar shown is 50 mm.

CD146, STRO-1 and CD 73 (Fig. 4a, arrows) but only weak to
moderate staining for NG2 and PDGF-rb (Fig. 4band c). Overall,
60e70% of the cells expressed these specific markers. Moreover,
these markers shared regions of colocalization (data not shown).
3.5. Cell culture on porous bacterial cellulose in vitro
When seeded on bacterial cellulose scaffolds and grown in 3D
rotational culture, cells grew to a uniform depth of 4e7 layers
compared to an uneven depth of 2e5 layers in static culture in all
groups (Table 1). In addition, rotational culture promoted cell-
scaffold infiltration (Fig. 5). However, no statistically significant
differences in cell growth, cell-scaffold infiltration, or cell differ-
entiation were found between the 10 rpm and 40 rpm rotational
cultures. Furthermore, no differences in cell infiltration into the
scaffolds with different pore sizes (90e150 mm, 150e300 mm and
300e500 mm) could be seen when rotational culture conditions
were used (data not shown).
We observed that layered co-cultures of urothelial and smooth
muscle cells generated from urine-derived stem cells resulted in up
to 14 cell layers with increasing cellematrix infiltration into the
cellulose scaffold compared with cultures consisting of urine-
derived stem cells alone (Fig. 5). Importantly, the number of cells
expressing urothelial cell markers (uroplakin-Ia, AE1/AE3) in
co-culture is nearly double that seen in mono-culture, as verified
by immunohistochemical staining after 7 days of culture (Figs. 6
and 7). However, it appears that there is no difference in urothe-
lial and smooth muscle cell marker (a-SM actin and desmin)
expression between cells grown in static and dynamic culture
conditions (Figs. 6e8).
Bacterial cellulose scaffolds seeded with human urine-derived
stem cells or a co-culture of urothelial and smooth muscle cells
differentiated from urine-derived stem cells were implanted into
athymic mice, the presence of both urothelial and smooth muscle
cells could be seen one month after implantation (Table 2). When
the implants were removed from the mice and examined, we
observed that more cells expressed urothelial cell markers in the
co-cultures compared to cultures of urine-derived stem cells or
urothelial cells alone (Table 2). A small amount of collagen around
the explants for both unseeded (data not shown) and seeded BC

Table 1
Impact of dynamic culture and co-culture on USC growth and differentiation in vitro.

Study groups UC þ SMCa UCb USCb USC-SMC þ USC-UCa

Culture conditions Static Dynamic Static Dynamic Static Dynamic Static Dynamic
Cell growth in vitro
Cell layers 2e4 4e5 2e3 4e5 2e4 5e6 2e5 5e7
Cell infiltration 0e10% 10% 0e10% 10% 0e10% 10% 10% 20e30%
Cell differentiation
Uroplakin þþþþ þþþþ þþþ þþþ þþþ þþþ þþþþ þþþþ
AE3/AE1 þþþþ þþþþ þþþ þþþ þþþ þþþ þþþþ þþþþ
ASMA þþþþ þþþþ e e e e þþþþ þþþþ
Desmin þþþþ þþþþ e e e e þþþþ þþþþ

e No cells; þþþ 75% stained cells; þþþþ 100% stained cells. UC: urothelial cells; SMC: smooth muscle cells; USC: urine derived stem cells; USC-SMC: smooth muscle
differentiated urine derived stem cell; USC-UC: urothelial differentiated urine derived stem cell.
a
Co-culture.
b
Mono-culture; dynamic ¼ 10 rpm.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8894 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901

Fig. 5. Effect of extracellular matrix formation as assessed by Masson’s trichrome staining with varying culture conditions in vitro. Co-culture of UC and SMC (top row), UC (second
row), urine-derived stem cells (third row) and co-culture of USC induced to UC and SMC (last row). Collagen stained blue, confirmed that the cells did not produce any extracellular
matrix after 7 days of culture. Insets are magnified images of areas boxed in dotted lines at 400.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901 8895

Fig. 6. Expression of urothelial-specific marker, AE1/AE3 on seeded BC scaffold in vitro. Staining was performed on scaffolds cultured using static and rotational culture conditions.
Co-culture of UC and SMC (top row), UC (second row), urine-derived stem cells (third row) and co-culture of USC induced to UC and SMC (last row). Sections were counterstained
with Hematoxylin. Urothelial cells are stained brown and muscle cells are stained pink. Insets are magnified images of areas boxed in dotted lines at 400.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8896 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901

Fig. 7. Expression of urothelial-specific marker, Uroplakin-Ia on seeded BC scaffold in vitro. Staining was performed on scaffolds cultured using static and rotational culture
conditions. Co-culture of UC and SMC (top row), UC (second row), urine-derived stem cells (third row) and co-culture of USC induced to UC and SMC (last row). Urothelial cells
stained brown and muscle cells stained pink. Hematoxylin was used as a counter stain. Insets are magnified images of areas boxed in dotted lines at 400.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901 8897

Fig. 8. Expression of smooth muscle-specific markers on seeded BC scaffold assessed by immunofluorescence analysis. Co-culture of UC and SMC and co-culture of USC induced to
UC and SMC, showed positive staining for the two myogenic markers: a-smooth muscle actin (aSMA, top and third row) and desmin (second and last row). The tissue sections were
counterstained with DAPI. Images are at 100. Insets are magnified images of areas boxed in dotted lines at 400.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8898 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901
Table 2
Effect of co-culture on USC growth and differentiation in vivo.
Cell differentiation UC þ SMC UC USC USC-SMC þ USC-UC
Human nucleus þ þ þþþ þþþþ
Uroplakin þ þ þþþ þþþþ
AE3/AE1 þ þ þþþ þþþþ
ASMA þþ þþþþ
e e
Desmin þþþ e e þþþ
e No stain; þ 25% stained cells; þþ 50% stained cells; þþþ 75% stained cells; þþþþ
100% stained cells.
was observed with Masson’s trichrome staining (Fig. 9). This
corresponds to our previous published data showing that the
material does not elicit any fibrotic capsule [24].
4. Discussion
The urinary tract is a unique system consisting of tubules and
hollow organs designed to transport, store, and eliminate urine.
In recent years, a number of studies have shown that various
biomaterials can be used to repair defects in the urinary tract
using tissue engineering-based strategies. Most biomaterials used
in urological tissue engineering are degradable. However, these
biodegradable materials often cause problems when they are
used improperly. These problems include fibrosis formation,
hollow organ (i.e. bladder) collapse, graft contraction, or tubular
organ (e.g. urethra or ureter) stricture resulting from inflamma-
tory reactions to the biomaterial. Although non-degradable
materials may cause complications such as displacement, erosion,
stone formation, and infection [25,26], suitable non-degradable
materials such as polyethylene, polyamides (i.e. nylon), poly-
esters, polyethereesters, and silk for surgical sutures have been
utilized in special applications or for sling procedures to treat
stress urinary incontinence.
As a non-degradable material, bacterial cellulose is an attractive
candidate for creating a tissue-engineered conduit because it is
highly hydrophilic and causes little fibrosis when implanted [24].
This polymer is biosynthesized as a network of nanofibrils. The
fibril entanglement and hydrogen bonding within the cellulose
network provides high mechanical strength and a large surface area
[11]. It was approved by the FDA in 1996 for applications including
wound dressing [12]. When implanted subcutaneously in rats,
bacterial cellulose does not elicit fibrosis or induce proliferation of
giant cells [24]. Bacterial cellulose has been shown to remain intact
for 90 days when implanted subcutaneously [24], and when it is
used as a material for blood vessel replacement, it remains in place
for at least 13 months (our unpublished data). In this study, we
prepared bacterial cellulose polymers with 3D interconnecting
pore microstructures. We hypothesized that the hydrophilicity and
interconnecting pore structure of the bacterial cellulose would
facilitate cell attachment and spreading of cells throughout the
scaffold to support the formation of a multilayered urothelium-like
structure. It has been shown that the optimal pore size for
biomaterials designed for tissue engineering ranges from 50 to
500 mm [27], and thus, we created scaffolds with pore sizes of
90e150 mm, 150e300 mm, or 300e500 mm to determine the
optimal scaffold configuration for urothelial tissue engineering
applications.
In order for many biomaterial-based repairs within the urinary
tract to be successful, the material must be covered with urothelial
cells to prevent stone formation and development of scar tissue,
particularly when larger defects are repaired in this manner.
Even non-degradable biomaterials might be used to form tissue-
engineered conduits for urinary tract reconstruction if the material
is covered with cells to form a urothelial lining. Autologous cells are
Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
always desirable for use in tissue engineering because the cells are
obtained from the individual undergoing treatment, and the risk of
Control graft rejection is eliminated. Our previous study [19] demonstrated
e that a subpopulation of cells isolated from naturally voided urine
e possesses the features of progenitor/stem cells, including expres-
e sion of MSC/pericyte cell surface markers and clonogenic, multi-
potential, and plastic adhesive capacity. These cells can give rise to
e
e
urothelial, smooth muscle, endothelial and interstitial cells [19]. In
the current study, we demonstrated that urine-derived stem cells
from the upper urinary tract can be isolated and expanded to
a large number in vitro and that they have the same features as
urine-derived stem cells from voided urine, including expression of
the percicyte/mesenchymal stem cell surface makers CD146, NG2,
PDGF-rb, CD73 and STRO-1.
Importantly, however, these cells have the capability to differ-
entiate into urothelial and smooth muscle cells. Thus, we hypoth-
esized that, when combined with appropriate scaffold materials,
urine-derived stem cells could be effectively used in urological
tissue engineering. In particular, we wished to determine if these
cells would be able to generate a complete urothelial layer when
seeded on a non-degradable bacterial cellulose scaffold in order to
produce a tissue-engineered urinary conduit that could be used in
patients undergoing cystoplasty. Specifically, urothelial differenti-
ation of urine-derived stem cells could be used to form a multilay-
ered barrier structure for fully covering the scaffold, thereby
preventing urinary salt encrustation on the exposed surface, stone
formation, and urinary tract infection. To test this hypothesis, we
seeded urine-derived stem cells from the upper urinary tract onto
a porous bacterial cellulose scaffold under dynamic culture condi-
tions to generate a cell-based urinary conduit. Porous bacterial
cellulose provided a 3D cell growth environment in vitro. Fully
differentiated urothelial and smooth muscle cells could be derived
from differentiation of the urine-derived stem cells using conduc-
tive media, and the differentiated cells were able to grow and
infiltrate the porous bacterial cellulose scaffold. These cells formed
a multilayered tissue structure in vitro under dynamic culture
conditions. These results suggest that both urine-derived stem cells
from the upper urinary tract and bacterial cellulose scaffolds can
be used to engineer specific urinary structures. These cells can
be harvested from the upper urinary tract using a simple, safe,
noninvasive, and low-cost procedure for patients who already have
a nerophectomy tube in place. Importantly, urine-derived stem
cells from the upper urinary tract can be a potential cell source for
patients with advanced bladder cancer who need cystectomy but
are tumor-free in the upper urinary tract. About 2e4% of patients
with bladder cancer develop upper urinary tract tumors due to the
multifocal nature of bladder transitional cell carcinomas [28,29].
Therefore, a careful screening of patients with bladder cancer
would be required to ensure that non-malignant cells are used for
this strategy.
Further, we were able to demonstrate that scaffold pore size did
not significantly affect cell growth. However, we chose to use the
300e500 mm range for future studies because this range would
allow adequate space for cell growth and extracellular matrix
secretion and remodeling, as well as for ingrowth of blood vessels
from the native tissue after implantation. We also found that
dynamic culture conditions promoted consistent cell growth and
led to the formation of more urothelial and muscle layers. Cellular
infiltration into the bacterial cellulose matrix was also enhanced in
dynamic compared to static cultures, and this observation is
consistent with our previous study [18].
The speed of dynamic culture also affects cell proliferation and
multilayer formation on scaffold matrices. For example, it has been
shown that when bladder cells are seeded on a collagen matrix
such as decellularized bladder submucosa and cultured in
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901 8899

Fig. 9. Histology of in vivo implanted porous BC with co-cultured with USC induced to SMC and UC following 1 month in nude mouse. Histochemical staining for hematoxylin and
eosin (top left), DAPI (top right) and Masson’s trichrome (second row left) are shown. Immunohistochemical staining for Human nucleus (second row, right), AE3/AE1 (third row,
left), aSMA (third row, right), Uroplakin-Ia (last row, left) and Desmin (last row, right) are also shown. Graft Magnification 200 for bright field and 400 for fluorescence.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
8900 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901
dynamic conditions at 40 RPM, cell layer formation is enhanced
compared to both 10 RPM and static culture conditions [18].
However, in the present study, dynamic culture conditions at
10 RPM promoted cell growth and multilayer formation as well as
40 RPM culture conditions. One reason for this might be that
bladder submucosa matrix is soft and can absorb the extra force
from rapid rotational culture, while the bacterial cellulose polymer
is relatively rigid. Hence, a rotation speed of 10 RPM might be
better for cell seeding on a bacterial cellulose matrix than 40 RPM.
This indicates that each type of biomaterial requires standardiza-
tion of dynamic culture parameters before cell seeding. However,
dynamic culture conditions would still be preferred to static
culture conditions, as nutrients are constantly available to growing
cells.
Epithelialestromal cell interactions in co-cultures play an
important role in cell growth and are an efficient means of
promoting cell growth, cellematrix infiltration, and cell differen-
tiation. The cellecell communication present in co-culture condi-
tions facilitates cell signaling and thus promotes epithelialization
[20]. In the present study, layered co-cultures of urothelial and
smooth muscle differentiated urine-derived stem cells showed
better cell growth and cellematrix penetration and epithelializa-
tion compared to mono-culture conditions, consistent with our
previous report [18]. The multilayered structure covered the entire
surface of the polymer scaffold, with smooth muscle cells infil-
trating the scaffold to a large extent. However, it appears that
dynamic culture alone does not enhance urothelial and smooth
muscle differentiation in this model.
In vitro data in this study demonstrated that the cells derived
from USC survive and maintain both urothelial and smooth muscle
markers one month after implantation, indicating that differenti-
ated grafted cells could retain bladder cell phenotypes in in vivo
environment. Urothelial barrier function of the differentiated USC
in vitro and in vivo will be investigated in a future study.
5. Conclusions
Cell-based tissue engineering may offer an alternative to ileal
conduit surgery for both continent and conduit diversions, allowing
patients to evacuate urine and preventing risks of further compli-
cations for patients with bladder cancer who need cystoplasty. In the
present study, we demonstrated that a 3D cell growth system
provides optimal culture conditions and mimics some in vivo
parameters. This 3D cell culture system includes three key elements:
a 3D porous biomaterial, a co-culture of different cell types, and
dynamic culture conditions. Autologous urine-derived stem cells
collected from the upper urinary tract are a valuable cell source.
Bacterial cellulose scaffolds with porous microstructures accom-
modated co-culture of differentiated urine-derived stem cells under
dynamic culture conditions. The cells generated a uniform multi-
layered urothelial-like structure. Our results indicate that a 3D
culture system gives favorable conditions for urine-derived stem cell
differentiation in porous bacterial cellulose scaffolds and may assist
in the development of a tissue-engineered urinary conduit, which
would be particularly useful for patients with bladder cancer who
need bladder replacement.
Acknowledgements
This work was funded by WFIRM and the Swedish Govern-
mental Agency for Innovation Systems (VINNOVA), who is greatly
acknowledged. The authors AB and PG would also like to thank
Arterion AB for making this study possible. Moreover does AB
express gratitude to the BBV project at Chalmers for providing lab
Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.
space. Finally, the authors would like to thank Dr. Jennifer Olson
and Ms. Karen Klein for their editorial assistance.
Appendix
Figures with essential color discrimination. Figs. 3e9 in this
article are difficult to interpret in black and white. The full color
images can be found in the on-line version, at doi:10.1016/j.
biomaterials.2010.07.108.
References
[1] Cancer facts and figures 2009. Atlanta: AmericanCancerSociety; May 2009.
[2] Yoo JJ, Meng J, Oberpenning F, Atala A. Bladder augmentation using allogenic
bladder submucosa seeded with cells. Urology 1998;51:221e5.
[3] Drewa T. The artificial conduit for urinary diversion in rats: a preliminary
study. Transplant Proc 2007;39:1647e51.
[4] Becker C, Olde Damink L, Laeufer T, Brehmer B, Heschel I, Jakse G. ‘UroMaix’
scaffolds: novel collagen matrices for application in tissue engineering of the
urinary tract. Int J Artif Organs 2006;29:764e71.
[5] Atala A, Bauer SB, Soker S, Yoo JJ, Retik AB. Tissue-engineered autologous
bladders for patients needing cystoplasty. Lancet 2006;367:1241e6.
[6] Oberpenning F, Meng J, Yoo JJ, Atala A. De novo reconstitution of a functional
mammalian urinary bladder by tissue engineering. Nat Biotechnol 1999;17:
149e55.
[7] Kobashi LI, Raible DA. Biocarbon urinary conduit: laboratory experience and
clinical applications. Urology 1980;16:27e32.
[8] Zhang Y, Frimberger D, Cheng EY, Lin HK, Kropp BP. Challenges in a larger
bladder replacement with cell-seeded and unseeded small intestinal
submucosa grafts in a subtotal cystectomy model. BJU Int 2006;98:
1100e5.
[9] Frey P, Lutz N, Leuba AL. Augmentation cystoplasty using pedicled and de-
epithelialized gastric patches in the mini-pig model. J Urol 1996;156:608e13.
[10] Lutz N, Frey P. Enterocystoplasty using modified pedicled, detubularized,
de-epithelialized sigmoid patches in the mini-pig model. J Urol 1995;154:
893e8.
[11] Backdahl H, Helenius G, Bodin A, Nannmark U, Johansson Bengt R, Risberg B,
et al. Mechanical properties of bacterial cellulose and interactions with
smooth muscle cells. Biomaterials 2006;27:2141e9.
[12] Wojciech Czaja AK, Stanislaw Bielecki, Malcolm Brown Jr R. Microbial cellu-
losedthe natural power to heal wounds. Biomaterials 2006;27:145e51.
[13] Mello LR, Feltrin LT, Fontes Neto PT, Ferraz FA. Duraplasty with biosynthetic
cellulose: an experimental study. J Neurosurg 1997;86:143e50.
[14] Klemm D, Schumann D, Udhardt U, Marsch S. Bacterial synthesized cellulose e
artificial blood vessels for microsurgery. Prog Polym Sci 2001;26:1561e603.
[15] Svensson A, Nicklasson E, Harrah T, Panilaitis B, Kaplan DL, Brittberg M, et al.
Bacterial cellulose as a potential scaffold for tissue engineering of cartilage.
Biomaterials 2005;26:419e31.
[16] Serafica G, Damien C, Oster GA, Lentz K, Hoon R. Microbial cellulose materials
for transdermal drug delivery of drugs. USA: Application: WOWO: (Xylos
Corporation; 2006. p. 17.
[17] Bäckdahl H, Esguerra. M, Delbro. D, Risberg. B, Gatenholm. P. Engineering
microporosity in bacterial cellulose scaffolds. J Tissue Eng Regen M 2008;2:
320e30.
[18] Liu Y, Bharadwaj S, Lee SJ, Atala A, Zhang Y. Optimization of a natural collagen
scaffold to aid cellematrix penetration for urologic tissue engineering.
Biomaterials 2009;30:3865e73.
[19] Zhang Y, McNeill E, Tian H, Soker S, Andersson KE, Yoo JJ, et al. Urine derived
cells are a potential source for urological tissue reconstruction. J Urol
2008;180:2226e33.
[20] Zhang Y, Kropp BP, Moore P, Cowan R, Furness 3rd PD, Kolligian ME, et al.
Coculture of bladder urothelial and smooth muscle cells on small intestinal
submucosa: potential applications for tissue engineering technology. J Urol
2000;164:928e34. discussion 34e5.
[21] Reddy P, Barrieras D, Wilson G, Bägli D, McLorie G, Khoury A, et al. Regen-
eration of functional bladder substitutes using large segment acellular matrix
allografts in a porcine model. J Urol 2000;164:936e41.
[22] Zheng MH, Chen J, Kirilak Y, Willers C, Xu J, Wood D. Porcine small intestine
submucosa (SIS) is not an acellular collagenous matrix and contains porcine
DNA: possible implications in human implantation. J Biomed Mater Res Part B
2005;73B:61e7.
[23] FDA. Guidline on validation of the limbulus amebocyte lysate test as an end-
product endotoxin test for human and animal parenteral drugs, biological
products and medical devices medical devices; 1987. 49.
[24] Helenius G, Backdahl H, Bodin A, Nannmark U, Gatenholm P, Risberg B.
In vivo biocompatibility of bacterial cellulose. J Biomed Mater Res A
2006;76:431e8.
[25] Grise P, Droupy S, Saussine C, Ballanger P, Monneins F, Hermieu JF, et al.
Transobturator tape sling for female stress incontinence with poly-
propylene tape and outside-in procedure: prospective study with 1 year of
 A. Bodin et al. / Biomaterials 31 (2010) 8889e8901 8901

minimal follow-up and review of transobturator tape sling. Urology
2006;68:759e63.
[26] Jakus SM, Shapiro A, Hall CD. Biologic and synthetic graft use in pelvic
surgery: a review. Obstet Gynecol Surv 2008;63:253e66.
[27] Smith IO, Liu XH, Smith LA, Ma PX. Nanostructured polymer scaffolds for
tissue engineering and regenerative medicine. Wiley Interdiscip Rev
Nanomed Nanobiotechnol 2009;1:226e36.
[28] Millan-Rodriguez F, Chechile-Toniolo G, Salvador-Bayarri J, Huguet-Perez J,
Vicente-Rodriguez J. Upper urinary tract tumors after primary superficial
bladder tumors: prognostic factors and risk groups. J Urol 2000;
164:1183e7.
[29] Slaton JW, Swanson DA, Grossman HB, Dinney CP. A stage specific approach to
tumor surveillance after radical cystectomy for transitional cell carcinoma of
the bladder. J Urol 1999;162:710e4.

Descargado para Anonymous User (n/a) en Italian Hospital of Buenos Aires de ClinicalKey.es por Elsevier en mayo 01, 2021. Para uso
personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2021. Elsevier Inc. Todos los derechos reservados.