Bi-layered carboxymethyl cellulose-collagen vitrigel dual-surface adhesion-prevention membrane

Carbohydrate Polymers 285 (2022) 119223

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

Bi-layered carboxymethyl cellulose-collagen vitrigel dual-surface
adhesion-prevention membrane
Yue Wang a, Kei Kanie a, Toshiaki Takezawa b, Miki Horikawa a, b, Kyoshiro Kaneko a,
Ayako Sugimoto a, Aika Yamawaki-Ogata c, Yuji Narita c, Ryuji Kato a, d, e, *
a

Department of Basic Medicinal Sciences, Graduate School of Pharmaceutical Sciences, Nagoya University, Tokai National Higher Education and Research System,
Furocho, Chikusa-ku, Nagoya, Aichi 464-8601, Japan
Division of Biomaterial Sciences, Institute of Agrobiological Sciences, National Agriculture and Food Research Organization, 1-2 Owashi, Tsukuba, Ibaraki 305-8634,
Japan
c
Department of Cardiac Surgery, Nagoya University Graduate School of Medicine, Tokai National Higher Education and Research System, 65 Tsurumai-cho, Showa-ku,
Nagoya, Aichi 466-8550, Japan
d
Institute of Nano-Life-Systems, Institutes of Innovation for Future Society, Nagoya University, Tokai National Higher Education and Research System, Furocho, Chikusaku, Nagoya, Aichi 464-8601, Japan
e
Institute of Glyco-core Research (IGCORE), Nagoya University, Tokai National Higher Education and Research System, Furocho, Chikusa-ku, Nagoya, Aichi 464-8601,
Japan
b

A R T I C L E I N F O

A B S T R A C T

Keywords:
Bi-layered carboxymethyl cellulose-collagen
vitrigel membrane
Cell-adhesion inhibition
Wound regeneration
Adhesion-inhibitor
Collagen hydrogel
Anti-adhesion membrane

During wound regeneration, both cell adhesion and adhesion-inhibitory functions must be controlled in parallel.
We developed a membrane with dual surfaces by merging the properties of carboxymethyl cellulose (CMC) and
collagen using vitrification. A rigid membrane was formed by vitrification of a bi-layered CMC and collagen
hydrogel without using cross-linking reagents, thus providing dual functions, strong cell adhesion-inhibition with
the CMC layer, and cell adhesion with the collagen layer. We referred to this bi-layered CMC-collagen vitrigel
membrane as “Bi-C-CVM” and optimized the process and materials. The introduction of the CMC layer conferred
a “tough but stably wet” property to Bi-C-CVM. This enables Bi-C-CVM to cover wet tissue and make the
membrane non-detachable while preventing tissue adhesion on the other side. The bi-layered vitrification pro­
cedure can expand the customizability of collagen vitrigel devices for wider medical applications.

1. Introduction
Recent tissue-engineering research reported that cell-derived medi­
cal devices have higher therapeutic effectiveness than conventional
products with simple materials (Imashiro & Shimizu, 2021). One
advantage of cell-derived products is the multiple functions in one
product that balances complex in vivo regeneration events, as in some
cases, contrary cellular events must be controlled in parallel.
Wound regeneration is a complex regeneration process that requires
delicate control of contrary cellular events to minimize side effects.
Although cell adhesion enhancement is essential for tissue repair, un­
wanted tissue adhesion may also occur. Tissue adhesion is a critical lifethreatening side effect after various surgeries (Kheilnezhad &

Hadjizadeh, 2021; Park et al., 2020). In abdominal surgeries, post­
operative peritoneal adhesion increases the risk of chronic pain and
intestinal obstruction (Hu et al., 2020; Ward & Panitch, 2011). Post
cardiac surgeries, postoperative intrapericardial adhesion increases the
critical risk of complications (Kikusaki et al., 2021; Wang et al., 2021).
Such unwanted adhesion forces patients to undergo a further thor­
acolaparotomy. To prevent such postoperative adhesion, a physical
barrier using anti-adhesive polymer membranes has been proposed
(Chandel et al., 2021; Hellebrekers et al., 2000; Mayes et al., 2020). To
provide anti-adhesive properties, adhesion-preventive membranes are
designed to be hydrophilic and water-absorbent. Such membranes are
wet and slippery, and can therefore detach from the target position after
the closure of the surgical site, losing the expected barrier effect. If a

Abbreviations: CMC, Carboxymethyl cellulose; Bi-C-CVM, Bi-layered CMC-collagen vitrigel membrane; CVM, Collagen vitrigel membrane; EAB, Elongation at
breaks.
* Corresponding author at: Department of Basic Medicinal Sciences, Graduate School of Pharmaceutical Sciences, Nagoya University, Tokai National Higher
Education and Research System, Furocho, Chikusa-ku, Nagoya, Aichi 464-8601, Japan.
E-mail address: (R. Kato).
/>Received 4 September 2021; Received in revised form 12 December 2021; Accepted 1 February 2022
Available online 8 February 2022
0144-8617/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( />

Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Fig. 1. Procedures for formulating Bi-C-CVM.
The initial bi-layer image with brighter color indicates layered CMC sol and collagen hydrogel before vitrification: CMC (green), CMC sol; Col (pink), collagen

hydrogel. The dark-colored bi-layer image represents rehydrated vitrigel. The fluorescent image indicates the bi-layer structure of Bi-C-CVM: CMC1 (green), Col
(red). The white bar in the image indicates scale bar = 100 μm.

membrane can control contrary functions by maintaining the antiadhesive function and stably attaching to the wound, it can overcome
the drawbacks of the current adhesion-prevention membranes.
CMC is a biodegradable material used in membranes for post­
operative adhesion (Ditzel et al., 2012; Kelekci et al., 2004; Park et al.,
2011; Lim et al., 2010; Bae et al., 2004) such as Seprafilm® (Helle­
brekers et al., 2000). However, such CMC membranes are unable to
provide cell adhesion for reducing detachment from the wound, and
controlling the degradation of stable, mechanically strong membranes
remains challenging. If a membrane absorbs water and rapidly dis­
perses, its functionality is hindered. Cross-linking is a common strategies
´ et al., 2018; Mohamed et al.,
used to control CMC degradation (Csapo
2020; Zennifer et al., 2021). However, chemical cross-linking is not
favored due to safety concerns. For safer cross-linking, enzymatic crosslinking has been reported (Cai et al., 2018). However, simpler mem­
brane formulating methods are preferred for manufacturing of medical
devices.
Vitrification that concerted the white of boiled egg into a transparent
and hard state was first discovered in a process in which bound water
was partially removed after free water was completely evaporated by
slow drying at 4 ◦ C for more than 10 days (Takushi et al., 1990).
Takezawa et al. produced collagen vitrigel by gelation of collagen sol,
vitrification of collagen gel and rehydration of vitrified collagen gel, and
clearly defined vitrigel as “a gel in a stable state produced by rehydra­
tion after the vitrification of a traditional hydrogel” (Takezawa et al.,
2004). The collagen vitrigel membrane (CVM) is an established
biomaterial used for cellular scaffolds formulated without any crosslinking reagent, with an excellent balance of both biocompatibility
and mechanical properties. This high-density collagen fibril-membrane

is formulated by rehydration after the vitrification of collagen hydro­
gel. During vitrification, a collagen hydrogel is dried by complete water
release (Takezawa et al., 2004). Loss of free water effectively increases
collagen intramolecular interactions to form a film composed of highdensity collagen fibrils. These intramolecular interactions are retained
even after rehydration to form a “wet but rigid” film, which completely
differs from the original hydrogel. The CVM was first used as a cell
culture substratum. A nylon membrane-framed CVM was useful for
fabricating three-dimensional culture models with paracrine (Takezawa
et al., 2007). A plastic cylinder-framed CVM (CVM chamber,

commercially available as ad-MED Vitrigel®) facilitated the exposure of
chemicals to culture models fabricated on CVM (Oshikata-Miyazaki &
Takezawa, 2016; Uzu & Takezawa, 2020; Yamaguchi et al., 2013).
Moreover, the attractive properties of CVM prompted its use in tissue
engineering and clinical applications including skin regeneration (Aoki
et al., 2014), articular cartilage reconstruction (Maruki et al., 2017), and
corneal repair (Chae et al., 2015). Although the properties of CVM are
ideal for tissue engineering, controlling the cell adhesive-inhibitory
property to prevent adhesion is difficult.
To design an adhesion-prevention membrane to control contrary
cellular events in wound regeneration, we hypothesized that “bi-layered
vitrification of CMC and collagen hydrogel can create a dual-surface
with advantageous properties, thus enabling one membrane to control
dual contrary cell-adhesion functions”. We developed a reproducible
protocol for formulating a bi-layered CMC-collagen vitrigel membrane
(Bi-C-CVM) that can provide dual surfaces for contrary functions: cell
adhesion-inhibitory function with the stable CMC layer and retention of
the cell-adhesion function with the collagen layer. The biological and
mechanical properties of Bi-C-CVM were evaluated by comparison with
CVM to evaluate the novel properties generated by CMC-collagen bilayer vitrification.

2. Materials and methods
2.1. Cells and cell-adhesion assay
Normal human dermal fibroblasts (NHDF; KF-4109, Kurabo Ltd.,
Osaka, Japan) were stained with Calcein-AM solution (148504-34-1,
Dojindo Laboratories Ltd., Kumamoto, Japan), seeded on the samples
(CVM, Bi-C-CVM, polytetrafluoroethylene (PTFE) sheet, or polystyrene
(PS) surface of well plates) in serum-free medium, incubated on the
sample for 2 h, washed three times with Dulbecco's phosphate-buffered
saline (D-PBS; 14249-24, Nacalai Tesque, Inc., Kyoto, Japan), and then
the remaining cells were counted by microplate reader (Fluoroskan
Ascent; Labsystems Ltd., Helsinki, Finland). The detailed cell culture and
cell-adhesion assay protocol is described in the supplementary data.

2


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Table 1
Molecular profiles of examined CMCs (CMC1–CMC5).
Type

Mw

DS

Concentration

Product no.

Lot no.

Provider

CMC1
CMC2
CMC3
CMC4
CMC5

250,000
250,000
250,000
90,000
700,000

0.7
0.9
1.2
0.7
0.9

2% w/v*
2% w/v*
2% w/v*
2% w/v*
1% w/v*

419,311
419,303
419,281
419,273
419,338

MKCD5149
MKCB9856
MKBZ8581V
MKCG4437
MKCD0622

Sigma-Aldrich
Sigma-Aldrich
Sigma-Aldrich
Sigma-Aldrich
Sigma-Aldrich

*

Concentrations are set to the maximum condition which CMCs can be completely dissolved in D-PBS after autoclaving.

2.2. Chemical components

2.3. Formulation of Bi-C-CVM

The chemical reagents used in this study are listed in Table S1 with
their compound identification numbers.

The preparation scheme of Bi-C-CVM is illustrated in Fig. 1. To form

the collagen bottom layer hydrogel, native collagen solution (5 mg
mL− 1) in 1 mM HCl at pH 3.0 (IAC-50, Koken Corporation Ltd., Tokyo,
Japan) was diluted with D-PBS to a concentration of 2 mg mL− 1 on ice.

Fig. 2. Comparison of CMC types to formulate Bi-C-CVM. (a) Representative fluorescent microscopic images of remaining adherent cells after washing. The white bar
in each image indicates scale bar = 1 mm. (b) Bar plot representation of the quantified percentage of remaining adherent cells after washing, with an SD of N = 3.
CVM, PTFE sheet, and bare PS surface were compared.
3


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Fig. 3. Effect of CMC sol conditions on Bi-C-CVM. (a, c, e) Representative fluorescent microscopic images of the remaining adherent cells after washing. The white
bar in each image indicates scale bar = 1 mm. (b, d, f) Bar plot representation of the quantified percentage of remaining adherent cells after washing, with an SD of N
= 3. CVM, PTFE sheet, and bare PS surface were compared.

The pH value of the diluted collagen solution before gelation was 6.78,
which was measured using a pH meter (S220, Mettler-Toledo, Colum­
bus, OH, USA). The dissolved collagen solution was kept at 0.22 mL
cm− 2 in containers which size is fit for their further application, and
incubated at 37 ◦ C for 2 h until complete gelation. CMC (detailed data
listed in Table 1) was dissolved in Milli-Q purified water and autoclaved
(121 ◦ C, 15 min) to form CMC sol. The CMC sol was layered on the
collagen hydrogel layer, and the two layers were vitrified together in an
air dryer (AA1-AD, As One Corporation, Osaka, Japan) at 10 ◦ C in a lowtemperature incubator (IN604, Yamato Scientific Ltd., Tokyo, Japan) for
48 h to produce Bi-C-CVM. The Bi-C-CVM was washed three times (10
min soaking each time) with D-PBS to rehydrate the sample, wash off the
fragile CMC layer surface, and stabilize the surface. After washing, the

sample was redried in a low-temperature incubator at 10 ◦ C for 24 h and
stored until further analysis. This form was designated as “redried form.”
For density measurements, the weight of Bi-C-CVM was measured by a
microbalance (ATX84, Shimadzu Co., Ltd., Kyoto, Japan).
For cell-adhesion assays, the samples were vitrified in a 12-well
plate, and the redried samples were rehydrated in D-PBS for 10 min in
the well before the assay. D-PBS was removed, and a cell suspension was
seeded onto the rehydrated sample. For tensile strength measurement,
samples were vitrified in an 8-well plate, and the washed membrane was
removed from the well using tweezers and redried on a PTFE sheet. The
membrane was rehydrated with D-PBS for 10 min before the measure­
ment. This form was named “rehydrated membrane.” To pick up the
membrane by tweezers, a white nylon membrane (HybondTM-N+,
RPN1782B, GE Healthcare, Little Chalfont, UK) was kept under the
hydrogel for support.

2.5. Collagen staining
Bi-C-CVM and CVM were stained for 30 min using Collagen Stain Kit
(K-61, Cosmo Bio Co., Ltd., Tokyo, Japan) and then repeatedly washed
with D-PBS until the color of the supernatant disappeared. Stained im­
ages were acquired by phase-contrast microscopy (TS100-F, Nikon Co.,
Ltd., Tokyo, Japan) equipped with a digital camera (CFI Achromat ADL
10XF, Nikon Co., Ltd) at 10× magnification.
2.6. Tensile strength measurements
The rehydrated membranes were measured using a tensile test ma­
chine (AGS-5KNJ, Shimadzu Co., Ltd.) and Trapezium 2 software (Shi­
madzu Co., Ltd.). The detailed protocol is described in the
supplementary data.
2.7. Characterization of Bi-C-CVM and CVM
The contact angle, swelling ratio, water retention, in vitro degrada­

tion, tissue adhesion, and handling properties were measured to char­
acterize Bi-C-CVM and CVM. The detailed protocols are described in the
supplementary data.
3. Results
3.1. Selection of optimum CMC type for Bi-C-CVM
To obtain a stable CMC layer on Bi-C-CVM, five types of CMCs with
different molecular profiles (Table 1) were examined. First, all CMCs
were vitrified without the collagen hydrogel but all samples formed
fragile flake-like membranes that dispersed immediately in water.
Therefore, the vitrification of only CMC cannot form a stable membrane.
However, we found that with the bi-layered vitrification process with
collagen hydrogel (Fig. 1), all samples formulated visually-similar vit­
rigel membranes that did not disperse after rehydration. From the
sectioned image, the clearly formed bi-layer structure and their thick­
ness of Bi-C-CVM was confirmed. In the cell-adhesion assay, CMC1

2.4. Scanning electron microscopy (SEM) characterization
Bi-C-CVM (without washing) and CVM (with washing) were lyoph­
ilized using vacuum deposition equipment (VE-2030, Vacuum Device
Ltd., Osaka, Japan), and coated with osmium plasma (OPC-40, Japan
laser corporation Ltd., Aichi, Japan). Samples were observed by SEM
(JSM-7610F, JEOL Ltd., Tokyo, Japan) at 10,000× magnification.

4


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

(a)
Bi-C-CVM

CMC1
Col

CVM

Col

Bi-C-CVM

CMC1
Col

CVM

Col

(b)

Fig. 4. Surface analysis of CMC1 layer on Bi-C-CVM. (a) SEM images of the top surface of Bi-C-CVM (CMC1 layer, without washing) and CVM. (b) Phase contrast
microscopy images of collagen staining of the top surface of Bi-C-CVM (CMC1 layer) and CVM. The white bar in each image indicates scale bar = 1 mm.

showed the best cell adhesion-inhibitory performance (Fig. 2), which
was comparable with that of PTFE, one of the most frequently used
materials that inhibit cell adhesion. Thus, the bi-layered vitrification
and optimum CMC type selection are essential for formulating a Bi-CCVM with cell adhesion-inhibitory function. CMC1 was selected for
further Bi-C-CVM evaluation. As the basic property, the average density
of the dried state Bi-C-CVM with CMC1 was 1.6 ± 0.2 g/cm3 (N = 3).

´n-Colo
´n et al.,
affecting the structure and properties of CVM (Caldero
2012), we examined the vitrification time (Fig. 3e–f). Our results
showed that the cell adhesion-inhibitory performance of Bi-C-CVM was
stable after 2 days of vitrification. Moreover, the finalized Bi-C-CVM
formation procedure was highly reproducible.

3.2. Optimization of CMC1 vitrification conditions for Bi-C-CVM

To confirm whether the CMC1 layer existed stably on the collagen
layer even after Bi-C-CVM rehydration, we assessed the CMC1 layer
surface using SEM (Fig. 4a). The CMC1 layer surface was different from
that of CVM, although a clear structural pattern was not observed
because the sample was never washed to retain the surface. To further
examine the presence of the CMC1 layer on Bi-C-CVM, collagen was
stained (Fig. 4b). No collagen with pink staining was detected on the
CMC1 layer surface of Bi-C-CVM. Collagen staining of all CMC surfaces
with different types of CMCs (Fig. S1a), concentrations of CMC1
(Fig. S1b), and volumes of CMC1 sol (Fig. S1c) showed that collagen
exposure was found in most samples that lost cell adhesion-inhibitory
performance. Therefore, although CMC layer of Bi-C-CVM can absorb
water and partially disperse, it reproducibly functions as a cell adhesioninhibitory layer under the optimum conditions determined in this study.

3.3. Surface analysis of the CMC layer on Bi-C-CVM

Vitrification conditions were optimized using CMC1. First, the CMC1
sol concentration was optimized (Fig. 3a–b). The desired performance
was achieved at a concentration of more than 1.5% wt. To make its

performance stable and reproducible, we selected 2.0% wt as the opti­
mum concentration.
Second, we examined the volume of the CMC layer (Fig. 3c–d) and
showed that thinner CMC1 layers could not sustain the cell adhesioninhibitory effect, suggesting that the rehydrated CMC1 layer can be
lost during washing before cell seeding. This result is reasonable because
the CMC1 layer is attached to the collagen layer through only simple
vitrification without any cross-linking reagents. However, the data
showed that such defects can be overcome by thickening the CMC1
layer. The maximum cell adhesion-inhibitory performance was achieved
with 2.6-mm thick CMC1 sol, which was selected as the optimum layer
volume.
Third, since the vitrification condition is an important parameter

3.4. Investigation of layering procedure of Bi-C-CVM
After evaluating the conditions and effects of the CMC-layer of Bi-C5


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Fig. 5. Development of a two-step procedure for Bi-C-CVM. (a) Schematic illustration of one-step and two-step procedures for forming Bi-C-CVM. (b) Bar plot
indicating the percentage of remaining adherent cells after washing, with an SD of N = 3. Images above the bar plot indicate the representative fluorescent
microscopic images of remaining adherent cells after washing. The white bar in each image indicates the scale bar = 1 mm. CVM was compared.

CVM, we investigated the tuning of the collagen layer in Bi-C-CVM. As
CVM is uniquely tough, we examined whether we can enhance the
mechanical property of Bi-C-CVM by controlling the collagen layer.
Therefore, we modified the bi-layer formulation procedure for Bi-CCVM. The one-step procedure established in previous sections was
feasible; however, the thickening of the collagen bottom layer was a

challenge as the vitrification of bi-layered hydrogels, particularly those
containing a CMC layer, take longer to dry. By separating the vitrifica­
tion step of the bottom layer (collagen hydrogel) and the top layer (CMC
sol), a two-step procedure was developed (Fig. 5a). Based on the cell
adhesion-inhibitory performances of Bi-C-CVM formulated with both
procedures, the result indicated that the two-step procedure sustained
good cell adhesion-inhibitory performance (Fig. 5b), enabling the
development of a Bi-C-CVM with a thickened collagen layer for subse­
quent mechanical property measurements.

3.5. Tensile strength evaluation of Bi-C-CVM
To characterize the mechanical property of Bi-C-CVM, we measured
the tensile strength in the rehydrated “wet” state (Fig. 6a). Bi-C-CVM (1Col) extended up to ~20%–30% of its total length (20 mm) (Fig. 6a
upper right). This data also revealed that the mechanical strength of BiC-CVM was nearly equivalent to CVM and suggests that the CMC layer
has little an effect on the mechanical strength of Bi-C-CVM, and the
collagen layer is the key parameter for controlling strength.
We next doubled the thickness of the collagen hydrogel layer using
our two-step procedure and measured their property (Fig. 6a bottom
right); the maximum endurable tensile force was increased nearly 7.5fold (up to 1.0–1.5 N). The tensile strain nearly doubled in the Bi-CCVM (2-Col) compared to CVM (2-Col). These data suggest that
although the CMC layer did not contribute to the strength, it enhanced
6


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Fig. 6. Tensile strength evaluation of rehydrated form Bi-C-CVM. (a) Tensile stretching curves of Bi-C-CVM (1-Col), Bi-C-CVM (2-Col), Bi-C-CVM (10-Col), singlelayered CVM and double-layered CVM. (b) Averaged EABs of CVMs and Bi-C-CVMs. (c) Averaged tensile stresses of CVMs and Bi-C-CVMs.

the stretchable property of Bi-C-CVM.

We further examined the thickening of the collagen layer using a
two-step procedure and found that Bi-C-CVM (10-Col) could achieve a
tensile force of 4–14 N (Fig. 6a). Interestingly, the thickness of Bi-C-CVM
(10-Col) in the rehydrated state was 120.4 ± 5.2 μm, whereas that of BiC-CVM (1-Col) was 79.5 ± 9.2 μm. These data indicated that during 10
repeated collagen hydrogel layering, the newly layered collagen
hydrogel merged with the bottom prior-vitrificated layer during vitri­
fication to form a denser membrane with small thickness increase.
To compare the mechanical property with the previously reported
anti-adhesive membranes, we converted the stretching profile into two
indices: averaged elongation at break (EAB) (Fig. 6b) and averaged
tensile stresses (Fig. 6c). EAB results over 130% were equivalent to those
of previously reported crosslinked membranes using CMC, whereas the
tensile stress values from Bi-C-CVMs were nearly 10-fold (2-Col) and 40fold (10-Col) higher than those of the cross-linked membranes (Cai et al.,
2018). We also confirmed that its cell adhesion-inhibitory function on
CMC layer had no negative effect by thickening the collagen layer
(Fig. S2). These results suggest that our Bi-C-CVM formulated from
simple vitrification achieved unique and superior mechanical properties
compared to those of other membranes formulated with cross-linking
agents. From the collagen layer thickening effect, our data also sug­
gest that the property control is feasible using our two-step procedure
(Fig. 5a).

3.6. Characterization of Bi-C-CVM for adhesion-prevention membrane
application
We further characterized the Bi-C-CVM properties to investigate its
applicability as an adhesion-prevention membrane (Fig. 7).
First, the redried form of Bi-C-CVM was evaluated upon physical and
visual examination, being semitransparent with a flexible and tough
texture (Fig. 7a). The Bi-C-CVM was difficult to tear by hand, but easily
cut using scissors, and was easy to roll without breaking. This handling

property was equivalent to that of the CVM; however, it was slightly
tougher due to the additional thickness of the CMC layer.
Second, the moisture-absorbent performance of the redried form of
Bi-C-CVM was evaluated. The CMC layer side showed higher wettability
and rapidly absorbed D-PBS (Fig. 7b, Video S1). The contact angle
measurements revealed that, after 1 min, the CMC side of Bi-C-CVM
quickly absorbed D-PBS and became hydrophilic (Fig. 7b). These re­
sults clearly indicated that the CMC layer added the novel property of
moisture-absorbance to Bi-C-CVM.
Third, the moisture-retaining performance of rehydrated Bi-C-CVM
was evaluated. The Bi-C-CVM remained wet for 5 h owing to its
moisture-retaining property of CMC layer (Fig. 7c, Video S2). From the
measurements of swelling ratio and water retention ratio, it was found
that although both membranes absorbed water within 3 min, Bi-C-CVM
has 10-fold higher water absorption capacity than CVM (Fig. 7d).
Furthermore, when rehydrated form of CVM dried quickly within 1 h,
the rehydrated form of Bi-C-CVM retained 80% of absorbed water. These
7


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Fig. 7. Characterization of redried form Bi-C-CVM for adhesion-prevention membrane application. (a) Image of the CVM and Bi-C-CVM. (b) The contact angle of
CVM and Bi-C-CVM (CMC side and Col side) (N = 3). (c) Water-retaining performance of CVM and Bi-C-CVM. (d) Swelling ratio (left) and water retention ratio (right)
of CVM and Bi-C-CVM (N = 3). (e) Tissue attachment and re-positioning performance of CVM and Bi-C-CVM. The collagen side of Bi-C-CVM is facing the wet tissue.
(f) Adhesion-prevention performance of CVM and Bi-C-CVM using the wet tissue model. (g) Tissue adhesion performance of Bi-C-CVM (CMC side or Col side) under
pull-down force (100 G, 1 min) (N = 3). *Student's t-test p < 0.05. (h) Time-course in vitro degradation of Bi-C-CVM (N = 3).

8


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

moisture-retaining performances enabled Bi-C-CVM to achieve the
“tough but stably wet” property for more feasible handling. Even in
highly wet conditions, Bi-C-CVM was tough enough to withstand
tweezer manipulation.
Fourth, the attachment performance of Bi-C-CVM was evaluated. The
redried form of Bi-C-CVM was flexible enough to attach to the wet tissue
with their collagen-side and could smoothly cover the three-dimensional
surface with the help of their moisture-absorbent property (Fig. 7e,
Video S3). Even with such feasible attachment, it was stable enough that
considerable effort was required for them to be detached by tweezers.
However, interestingly, when we tried to re-position the membranes
within a short period, Bi-C-CVM could be feasibly re-positioned,
whereas CVM shrunk rapidly and became crumpled, hindering repositioning. This result further suggest that Bi-C-CVM achieved a
“tough but stably wet” property due to the CMC layer, which enabled
attachment with flexibility, allowing physicians to reconsider its
positioning.
Fifth, the adhesion-prevention performance of the redried form of BiC-CVM was evaluated with a brief tissue model. CVM and Bi-C-CVM
were sandwiched between wet tissue samples, and their performances
were compared after 5 h (Fig. 7f, Video S4). Tissues adhered tightly to
both sides of the CVM. However, for Bi-C-CVM, the bottom tissue
adhered tightly to the collagen layer, whereas the top tissue adjacent to
the CMC layer detached easily. To quantify the strength of adhesive
ability on both sides of Bi-C-CVM, after 1 h of attachment to wet tissue,

tissues were pulled down by 100 G for 1 min using a centrifuge. Even
with such severe tissue tearing stress, the tissues on the collagen side of
Bi-C-CVM remained 5-fold more than those on the CMC side (Fig. 7g).
These data strongly suggest that the developed Bi-C-CVM achieved the
dual cell-adhesion functions stated in our hypothesis and suggests its
potential as a candidate material for an adhesion-preventive barrier with
lower risk of detachment.
Furthermore, to estimate the in vivo degradation potency, the timecourse in vitro degradation was measured (Fig. 7h). In the presence of
collagenase, Bi-C-CVM degraded within 1 d; however, with only me­
dium or trypsin, the degradation was slow, and 40% (wt/wt) remained
within 1 week.

preventive devices (Li et al., 2020); thus, our dual-surface membrane BiC-CVM may overcome these limitations.
In tensile strength evaluation, Bi-C-CVM showed a “J” curve
stretching property essential for artificial blood vessel grafts (Sonoda
et al., 2003). This “J” curve stretching property is important for avoiding
compliance mismatch between the native artery and artificial grafts,
which is a critical cause when using small-diameter artificial graft;
however, it is difficult to achieve by using hydrogel-based materials
(Aussel et al., 2017; Yang et al., 2020). When we compared Bi-C-CVM
data with that of porcine blood vessels, our Bi-C-CVM showed a nearly
equivalent “J” curve property (Fig. S3a). We also found that Bi-C-CVM
(10-Col) can maximize the tensile stress up to 2.2 MPa, which is stron­
ger than that of porcine blood vessels (average 1.7 MPa) (Fig. S3b).
Therefore, together with the reported hemocompatibility effect of CMC
(Basu et al., 2018), our Bi-C-CVM showed a great potential for artificial
graft application, and its unique mechanical property with simple des­
ignability will enable wider medical applications.

4. Discussion

This study was partially supported by JSPS KAKENHI, grant numbers
23680055 to RK, and 18K14061 and 20K05227 to KK1 (Kanie). This
work was also supported by the Toyoaki Scholarship Foundation to KK1,
and Tatematsu Foundation to KK.

5. Conclusions
We developed a biocompatible membrane with dual surfaces for
controlling contrary cell-adhesion functions with one membrane, named
as Bi-C-CVM. By simple bi-layered vitrification of CMC and collagen
hydrogels, the difficulty of controlling CMC degradation for membrane
formation was overcome, resulting in a rigid membrane without using
cross-linking agents and supporting our hypothesis. Our Bi-C-CVM
exhibited novel properties exerted by the CMC layer while retaining
the advantages of CVM. Therefore, it may be applicable for controlling
more complex regenerative events in vivo such as postoperative adhesion
prevention. Furthermore, by taking advantage of its unique mechanical
properties and its dual-surface functions, Bi-C-CVM might be considered
as a basal material to formulate a tube-like structure for developing
novel-type artificial blood vessels.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.carbpol.2022.119223.
Funding

We developed a dual-functional membrane device for controlling
contrary cellular events in wound regeneration by combining the ad­
vantages of CMC and CVM. Using a simple but robust bi-layer vitrifi­
cation procedure, we developed Bi-C-CVM that retains the advantageous
properties of both CMC and collagen without using any cross-linking
agents. Our combinatorial application of CMC and collagen in one bilayered membrane will accelerate not only the use of CVM devices but

also other medical applications of CMC-based membranes (Rahman
et al., 2021) and collagen-based membranes (Sayani & Raines, 2014).
Uncontrolled detachment of adhesion-prevention membranes after
the closure of a surgical site is a practical and serious risk. Most animal
models of postoperative adhesion are designed using small animals by
creating abdominal defects between small spaces with flat surfaces (e.g.,
abdominal walls, or organ surfaces); therefore, the membrane detach­
ment risk in humans (larger space between more structurally complex
surfaces) is difficult to determine. However, even in small animals,
Hellebrekers et al. (2000) showed that 11 of 20 PolyActive™ were de­
tached from the wound, suggesting the potential for such risk. Our data
indicated the novel potential of the “dual-surface adhesion-prevention
membrane,” which can attach and cover the wound on one side, while
preventing adhesion on the other. Cauterizing the surface negatively
affects the opposite intact peritoneum surface within 1 h, damaging
protective mesothelial cells (Suzuki et al., 2015). Therefore, not only
providing barrier effect but also rapid ensealing of post-operative tissue
surface is important for improving adhesion prevention. There are still
few reports that propose such wound protection effect for adhesion

CRediT authorship contribution statement
Yue Wang: Data curation, Investigation, Methodology, Validation,
Visualization, Writing – original draft, Writing – review & editing. Kei
Kanie: Conceptualization, Methodology, Funding acquisition, Project
administration, Supervision, Visualization, Writing – review & editing.
Toshiaki Takezawa: Conceptualization, Methodology. Miki Hori­
kawa: Conceptualization, Data curation, Validation, Methodology.
Kyoshiro Kaneko: Conceptualization, Data curation, Validation,
Methodology. Ayako Sugimoto: Data curation, Methodology. Aika
Yamawaki-Ogata: Methodology. Yuji Narita: Conceptualization,

Methodology. Ryuji Kato: Conceptualization, Funding acquisition,
Methodology, Project administration, Supervision, Writing – review &
editing.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.

9


Y. Wang et al.

Carbohydrate Polymers 285 (2022) 119223

Acknowledgments

Li, Z., Liu, L., & Chen, Y. (2020). Dual dynamically crosslinked thermosensitive hydrogel
with self-fixing as a postoperative anti-adhesion barrier. Acta Biomaterialia, 110,
119–128. />Lim, R., Stucchi, A. F., Morrill, J. M., Reed, K. L., Lynch, R., & Becker, J. M. (2010). The
efficacy of a hyaluronate-carboxymethylcellulose bioresorbable membrane that
reduces postoperative adhesions is increased by the intra-operative coadministration of a neurokinin 1 receptor antagonist in a rat model. Surgery, 148(5),
991–999. />Maruki, H., Sato, M., Takezawa, T., Tani, Y., Yokoyama, M., Takahashi, T., Toyoda, E.,
Okada, E., Aoki, S., Mochida, J., & Kato, Y. (2017). Effects of a cell-free method using
collagen vitrigel incorporating TGF-β1 on articular cartilage repair in a rabbit
osteochondral defect model. Journal of Biomedical Materials Research - Part B Applied
Biomaterials, 105(8), 2592–2602. />Mayes, S. M., Davis, J., Scott, J., Aguilar, V., Zawko, S. A., Swinnea, S., Peterson, D. L.,
Hardy, J. G., & Schmidt, C. E. (2020). Polysaccharide-based films for the prevention
of unwanted postoperative adhesions at biological interfaces. Acta Biomaterialia,
106, 92–101. />Mohamed, S. A. A., El-Sakhawy, M., & El-Sakhawy, M. A. M. (2020). Polysaccharides,

protein and lipid -based natural edible films in food packaging: A review.
Carbohydrate Polymers, 238, Article 116178. />carbpol.2020.116178
Oshikata-Miyazaki, A., & Takezawa, T. (2016). Development of an oxygenation culture
method for activating the liver-specific functions of HepG2 cells utilizing a collagen
vitrigel membrane chamber. Cytotechnology, 68(5), 1801–1811. />10.1007/s10616-015-9934-1
Park, J. S., Lee, J. H., Han, C. S., Chung, D. W., & Kim, G. Y. (2011). Effect of hyaluronic
acid-carboxymethylcellulose solution on perineural scar formation after sciatic nerve
repair in rats. Clinics in Orthopedic Surgery, 3(4), 315–324. />cios.2011.3.4.315
Park, H., Baek, S., Kang, H., & Lee, D. (2020). Biomaterials to prevent post-operative
adhesion. Materials, 13(14), 3056. />Rahman, M. S., Hasan, M. S., Nitai, A. S., Nam, S., Karmakar, A. K., Ahsan, M. S.,
Shiddiky, M. J. A., & Ahmed, M. B. (2021). Recent developments of carboxymethyl
cellulose. Polymers, 13, 1345. />Sayani, C., & Raines, R. T. (2014). Collagen-based biomaterials for wound healing.
Biopolymers, 101(8), 821–833. />Sonoda, H., Takamizawa, K., Nakayama, Y., Yasui, H., & Matsuda, T. (2003). Coaxial
double-tubular compliant arterial graft prosthesis: Time-dependent morphogenesis
and compliance changes after implantation. Journal of Biomedical Materials Research Part A, 65(2), 170–181. />Suzuki, T., Kono, T., Bochimoto, H., Hira, Y., Watanabe, T., & Furukawa, H. (2015). An
injured tissue affects the opposite intact peritoneum during postoperative adhesion
formation. Scientific Reports, 5, 7668. />Takezawa, T., Ozaki, K., Nitani, A., Takabayashi, C., & Shimo-Oka, T. (2004). Collagen
vitrigel: A novel scaffold that can facilitate a three-dimensional culture for
reconstructing organoids. Cell Transplantation, 13(4), 463–473. />10.3727/000000004783983882
Takezawa, T., Takeuchi, T., Nitani, A., Takayama, Y., Kino-oka, M., Taya, M., &
Enosawa, S. (2007). Collagen vitrigel membrane useful for paracrine assays in vitro
and drug delivery systems in vivo. Journal of Biotechnology, 131(1), 76–83. https://
doi.org/10.1016/j.jbiotec.2007.05.033
Takushi, E., Asato, L., & Nakada, T. (1990). Edible eyeballs from fish. Nature, 345(24),
298–299. />Uzu, M., & Takezawa, T. (2020). Novel microvascular endothelial model utilizing a
collagen vitrigel membrane and its advantages for predicting histamine-induced
microvascular hyperpermeability. Journal of Pharmacological and Toxicological
Methods, 106, Article 106916. />Wang, X., Liu, Z., Sandoval-Salaiza, D. A., Afewerki, S., Jimenez-Rodriguez, M. G.,
Sanchez-Melgar, L., Güemes-Aguilar, G., Gonzalez-Sanchez, D. G., Noble, O.,
Lerma, C., Parra-Saldivar, R., Lemos, D. R., Llamas-Esperon, G. A., Shi, J., Li, L.,

Lobo, A. O., Fuentes-Baldemar, A. A., Bonventre, J. V., Dong, N., & RuizEsparza, G. U. (2021). Nanostructured non-Newtonian drug delivery barrier prevents
postoperative intrapericardial adhesions. ACS Applied Materials & Interfaces, 13(25),
29231–29246. />Ward, B. C., & Panitch, A. (2011). Abdominal adhesions: Current and novel therapies.
Journal of Surgical Research, 165(1), 91–111. />jss.2009.09.015
Yamaguchi, H., Kojima, H., & Takezawa, T. (2013). Vitrigel-eye irritancy test method
using HCE-T cells. Toxicological Sciences, 135(2), 347–355. />toxsci/kft159
Yang, Y., Zhao, X., Yu, J., Chen, X., Chen, X., Cui, C., Zhang, J., Zhang, Q., Zhang, Y.,
Wang, S., & Cheng, Y. (2020). H-bonding supramolecular hydrogels with promising
mechanical strength and shape memory properties for postoperative antiadhesion
application. ACS Applied Materials and Interfaces, 12(30), 34161–34169. https://doi.
org/10.1021/acsami.0c07753
Zennifer, A., Senthilvelan, P., Sethuraman, S., & Sundaramurthi, D. (2021). Key advances
of carboxymethyl cellulose in tissue engineering & 3D bioprinting applications.
Carbohydrate Polymers, 256, Article 117561. />carbpol.2020.117561

We thank Dr. Ayumi Oshikata-Miyazaki in the Division of Biomate­
rial Sciences, Institute of Agrobiological Sciences, National Agriculture
and Food Research Organization for assisting the vitrigel formation
experiments. We are also thankful for the thoughtful discussion of Dr.
Kazunori Shimizu, and Dr. Hiroyuki Honda in the Graduate School of
Engineering, Nagoya University. Lastly, we thank Sugimoto Shokuniku
Sangyo Co., Ltd. for providing porcine blood vessels.
References
Aoki, S., Takezawa, T., Miyazaki-Oshikata, A., Ikeda, S., Nagase, K., Koba, S., Inoue, T.,
Uchihashi, K., Nishijima-Matsunobu, A., Kakihara, N., Hirayama, H., Narisawa, Y., &
Toda, S. (2014). Collagen vitrigel membrane: A powerful tool for skin regeneration.
Inflammation and Regeneration, 34(3), 117–123. />inflammregen.34.117
Aussel, A., Montembault, A., Malaise, S., Foulc, M. P., Faure, W., Cornet, S., Aid, R.,
Chaouat, M., Delair, T., Letourneur, D., David, L., & Bordenave, L. (2017). In vitro
mechanical property evaluation of chitosan-based hydrogels intended for vascular

graft development. Journal of Cardiovascular Translational Research, 10(5–6),
480–488. />Bae, J. S., Jin, H. K., & Jang, K. H. (2004). The effect of polysaccharides and
carboxymethylcellulose combination to prevent intraperitoneal adhesion and
abscess formation in a rat peritonitis model. Journal of Veterinary Medical Science, 66
(10), 1205–1211. />Basu, P., Narendrakumar, U., Arunachalam, R., Devi, S., & Manjubala, I. (2018).
Characterization and evaluation of carboxymethyl cellulose-based films for healing
of full-thickness wounds in normal and diabetic rats. ACS Omega, 3(10),
12622–12632. />Cai, X., Hu, S., Yu, B., Cai, Y., Yang, J., Li, F., Zheng, Y., & Shi, X. (2018).
Transglutaminase-catalyzed preparation of crosslinked carboxymethyl chitosan/
carboxymethyl cellulose/collagen composite membrane for postsurgical peritoneal
adhesion prevention. Carbohydrate Polymers, 201(2), 201–210. />10.1016/j.carbpol.2018.08.065
Calder´
on-Col´
on, X., Xia, Z., Breidenich, J. L., Mulreany, D. G., Guo, Q., Uy, O. M.,
Tiffany, J. E., Freund, D. E., McCally, R. L., Schein, O. D., Elisseeff, J. H., &
Trexler, M. M. (2012). Structure and properties of collagen vitrigel membranes for
ocular repair and regeneration applications. Biomaterials, 33(33), 8286–8295.
/>Chae, J. J., McIntosh Ambrose, W., Espinoza, F. A., Mulreany, D. G., Ng, S., Takezawa, T.,
Trexler, M. M., Schein, O. D., Chuck, R. S., & Elisseeff, J. H. (2015). Regeneration of
corneal epithelium utilizing a collagen vitrigel membrane in rabbit models for
corneal stromal wound and limbal stem cell deficiency. Acta Ophthalmologica, 93(1),
e57–e66. />Chandel, A. K. S., Shimizu, A., Hasegawa, K., & Ito, T. (2021). Advancement of
biomaterial-based postoperative adhesion barriers. Macromolecular Bioscience, 21(3),
1–34. />´ Varga, N., Janov´
Csap´
o, E., Szokolai, H., Juh´
asz, A.,
ak, L., & D´
ek´
any, I. (2018). Crosslinked and hydrophobized hyaluronic acid-based controlled drug release systems.

Carbohydrate Polymers, 195, 99–106. />carbpol.2018.04.073
Ditzel, M., Deerenberg, E. B., Komen, N., Mulder, I. M., Jeekel, H., & Lange, J. F. (2012).
Postoperative adhesion prevention with a new barrier: An experimental study.
European Surgical Research, 48(4), 187–193. />Hellebrekers, B. W. J., Trimbos-Kemper, G. C. M., van Blitterswijk, C. A., Bakkum, E. A.,
& Trimbos, J. B. M. Z. (2000). Effects of five different barrier materials on
postsurgical. Human Reproduction, 15(6), 1358–1363. />humrep/15.6.1358
Hu, Q., Xia, X., Kang, X., Song, P., Liu, Z., Wang, M., Lu, X., Guan, W., & Liu, S. (2020).
A review of physiological and cellular mechanisms underlying fibrotic postoperative
adhesion. International Journal of Biological Sciences, 17(1), 298–306. https://doi.
org/10.7150/ijbs.54403
Imashiro, C., & Shimizu, T. (2021). Fundamental technologies and recent advances of
cell-sheet-based tissue engineering. International Journal of Molecular Sciences, 22(1),
1–18. />Kelekci, S., Yilmaz, B., Oguz, S., Zergeroǧlu, S., Inan, I., & Tokucoǧlu, S. (2004). The
efficacy of a hyaluronate/carboxymethylcellulose membrane in prevention of
postoperative adhesion in a rat uterine horn model. Tohoku Journal of Experimental
Medicine, 204(3), 189–194. />Kheilnezhad, B., & Hadjizadeh, A. (2021). A review: Progress in preventing tissue
adhesions from a biomaterial perspective. Biomaterials Science, 9(8), 2850–2873.
/>Kikusaki, S., Takagi, K., Shojima, T., Saku, K., Fukuda, T., Oryoji, A., Anegawa, T.,
Fukumoto, Y., Aoki, H., & Tanaka, H. (2021). Prevention of postoperative
intrapericardial adhesion by dextrin hydrogel. General Thoracic and Cardiovascular
Surgery, 69, 1326–1334. />
10