1 s2.0 S0928493114000186 Main

Collagen tissue treated with chitosan solutions in carbonic acid for
improved biological prosthetic heart valves

Marat O. Gallyamov
a,b,
, Ivan S. Chaschin
b
, Marina A. Khokhlova
a
, Timofey E. Grigorev
b
, Natalia P. Bakuleva
c
,
Irina G. Lyutova
c
, Janna E. Kondratenko
c
, Gennadii A. Badun
d
, Maria G. Chernysheva
d
, Alexei R. Khokhlov
a,b
a
Faculty of Physics, Lomonosov Moscow State University, Leninskie gory 12, Moscow 119991, Russian Federation
b
Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Vavilova 28, Moscow 119991, Russian Federation
c
Bakulev Scientic Center for Cardiovascular Surgery of the Russian Academy of Medical Sciences, Roublyevskoe Sh. 135, Moscow 121552, Russian Federation
d
Radiochemistry Division, Faculty of Chemistry, Lomonosov Moscow State University, Leninskie gory 12, Moscow 119991, Russian Federation
a b s t r a c t a r t i c l e i n f o
Article history:
Received 26 September 2013
Received in revised form 16 December 2013
Accepted 5 January 2014
Available online 11 January 2014
Keywords:
Biological prosthetic heart valve
Bovine pericardium
Calcication
Chitosan
Carbonic acid
Antimicrobial activity
Calcication of bovine pericardium dramatically shortens typical lifetimes of biological prosthetic heart valves
and thus precludes their choice for younger patients. The aimof the present work is to demonstrate that the cal-
cication is to be mitigatedby means of treatment of bovine pericardiuminsolutions of chitosan incarbonic acid,
i.e. water saturated with carbon dioxide at high pressure. This acidic aqueous uid unusually combines antimi-
crobial properties with absolute biocompatibility as far as at normal pressure it decomposes spontaneously
and completely into H
2
Oand CO
2
. Yet, at high pressures it can protonate and dissolve chitosan materials with dif-
ferent degrees of acetylation (in the range of 1633%, at least) without any further pretreatment. Even exposure
of the bovine pericardium in pure carbonic acid solution without chitosan already favours certain reduction in
calcication, somewhat improved mechanical properties, complete biocompatibility and evident antimicrobial
activity of the treatedcollagentissue. The reasonmay be due to highextractionability of this peculiar compressed
uidic mixture. Moreover, exposure of the bovine pericardiuminsolutions of chitosanincarbonic acid introduces
even better mechanical properties and highly pronounced antimicrobial activity of the modied collagen tissue
against adherence and biolmformation of relevant Gram-positive and Gram-negative strains. Yet, the most im-
portant achievement is the detected dramatic reduction incalcication for suchmodiedcollagentissues inspite
of the fact that the amount of the thus introduced chitosan is rather small (typically ca. 1 wt.%), which has been
reliably detectedusing original tritiumlabelling method. We believe that these improved properties are achieved
due to particularly deep and uniform impregnation of the collagen matrix with chitosan from its pressurised
solutions in carbonic acid.
2014 Elsevier B.V. All rights reserved.
1. Introduction
There is strong interest in improving properties of biological pros-
thetic heart valves, which demonstrate certain advantages as compared
with mechanical heart valves [15]. The major functional part of the
bioprosthetic valves is a soft, elastic and durable collagen tissue made
of either bovine pericardium or porcine aortic valve. For surgery appli-
cations the tissue is to be crosslinked mainly by glutaraldehyde (GA)
to ensure mechanical stability andabsence of any foreign-body immune
response of a patient [15]. Such a leaet of the valve is akin in its
structure and properties to the native one being replaced. As far as
bioprosthetic valves better emulate haemodynamic properties of native
human valves as compared with mechanical substitutes, this leads to
the much less damage of red blood cells and a lower risk of blood clot
formation [15].
Nevertheless, still in the majority of surgery cases mechanical heart
valves remain to be the best or the only possible choice. Indeed, ac-
cording to the recent review of Siddiqui et al., more than 250,000 sub-
stitute valves are implanted each year worldwide, of which roughly
55% are mechanical heart valves and only 45% are biological prosthetic
heart valves [6]. A decisive drawback of the bioprosthetic valves is
too short service life therefore they are mainly not recommended for
young patients [18]. Calcication of the collagen tissue is among the
main causes of failure of the bioprosthetic valves [9,10]. This calcica-
tion is considered to be induced by residual unreacted aldehyde groups
remaining in the crosslinked tissue and interacting with the compo-
nents of the blood plasma, though the total concentration of such
groups is not really high [11]. GA-crosslinked collagen is a spongy-like
matrix with many internal voids and cavities, which tend to be lled
with the calcium-containing deposit [12,13] mainly in the form of
calcium phosphate salts [14].
Many efforts were directed towards the development of methods to
mitigate the calcium deposition. In order to reduce susceptibility of the
GA-crosslinked collagen tissue to calcication, researchers in this area
Materials Science and Engineering C 37 (2014) 127140
Corresponding author. Tel.: +7 495 9391430.
E-mail address: glm@spm.phys.msu.ru (M.O. Gallyamov).
0928-4931/$ see front matter 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.msec.2014.01.017
Contents lists available at ScienceDirect
Materials Science and Engineering C
j our nal homepage: www. el sevi er . com/ l ocat e/ msec
mainly aimed at masking or inactivating of residual free aldehyde
groups. In some cases they also tried additionally to ll the interstitial
(inter-brils or inter-tropocollagen) space with some ller. Both low-
molecular-weight and polymer materials were tested. Among success-
fully tested low-molecular-weight anticalcication agents one should
mention amino-containing -amino oleic acid (AOA) [1517] and
L-arginine [18], amino-containing [19] and other diphosphonates [20],
urazole with sodium borohydride [21] as well as citric acid [22]
and some surfactants [23]. Among tested polymer materials promising
results were obtained with generally biocompatible poly(ethylene
oxide) [2426], other biocompatible polymers [27], amino-containing
polyacrylamide [28] as well as with glycosaminoglycans such as
hyaluronic acid [29] and heparin [30,31]. In general, according to subcu-
taneous implantation express-tests in rats, polymeric modiers [2431]
demonstrated higher efciency in the reduction of calcication in com-
parison with their low-molecular-weight counterparts [1523]. Yet, the
treatment procedure with polymers was typically rather complex and
in some cases an additional step of a special chemical modication
was required to ensure polymer grafting [24,25,29].
Following this general paradigm, in a set of papers Chanda proposed
and successfully implemented with co-workers simpler procedure of
covering the bioprostheses surface with a chitosan lm. The procedure
is based on direct exposure of a collagen tissue in solutions of chitosan
assuming binding between chitosan amino-groups and residual free
aldehyde groups in the crosslinked matrix [3238]. But Chanda worked
with water-soluble chitosan [32] such as a material with degree of
N-acetylation (DA) of ca. 50% [3941]. Yet, commercially available
chitosans with normal DAs of ca. 1535% are soluble only in acidic
media [42]. There are many possible ways to prepare water-soluble
chitosan materials (at neutral pH conditions). But a chitosan lm
adsorbed from such solutions would hardly be sufciently stable in
a blood plasma stream (pH value of ca. 7.4) and most probably should
excessively swell there and eventually delaminate.
Other authors formed chitosan lms on collagen matrices from
commonly used solutions in acetic acids with subsequent alkaline
neutralisation. Its purpose is to convert chitosan into a water-insoluble
form thus ensuring mechanical stability of the adsorbed lms [4346].
But there are some possible disadvantages of using acetic acid as a sol-
vent for treatment of implants. First of all, acetic acid deeply impreg-
nates a collagen tissue and may even disturb organization of collagen
brils [4749]. Consequently, deterioration of mechanical properties
of the matrix may be expected. This is why Shanthi and Rao [43]
attempted to use the lowest possible concentrations of acetic acid solu-
tions. Further, residual traces of this solvent in the modied collagen
tissue are not only cytotoxic [45] but potentially allergenic. Indeed,
it is well-known that in some occasional cases surprisingly strong
quasi-allergenic reactions may be observed, when acetic acid acts
most probably as a hapten [5055]. In these cases it is responsible for
pronounced intolerance or hypersensitivity observed, including positive
skin-prick tests [5055].
Our task was to nd another solvent, which is capable to dissolve
chitosan and yet its residual traces are absolutely compatible with the
human body. Previously, we tested the possibility to apply biocompati-
ble and non-allergenic supercritical (sc) CO
2
as a solvent for chitosan,
but the achieved solubility was too low for real applications [56].
Further, working withthe mixtures of water and CO
2
in a closed ves-
sel at high pressure we found out rather good solubility of chitosan not
in CO
2
saturated with water, but, on the contrary, in water saturated
with pressurised CO
2
, i.e. in carbonic acid [57]. Indeed, previously,
authors of Ref. [58] have already demonstrated that it is possible to
dissolve chitosan with different DAs in carbonic acid.
The aim of the present work is to develop a procedure of the forma-
tion of chitosan coatings on a bovine pericardium tissue from solutions
of chitosan in carbonic acid as well as to study the structure and main
properties of the modied collagentissue including stressstrainbehav-
iour and afnity towards calcium deposits, bacteria and cells. Taking
into account very peculiar properties of this polysaccharide [59], we ex-
pected not only mitigation of calcication, but also improved biocom-
patibility and introduced remarkable antimicrobial activity (including
suppression of possible biolm [6062] formation) of the collagen
tissue with the formed chitosan coating to be used as a material for
biological prosthetic heart valves.
2. Experimental
2.1. Materials
We tested several chitosan samples with different DAs supplied
by Sigma-Aldrich: #448869 (1624% DA), #419419 (2930% DA),
#448877 (2627% DA), #417963 (3031% DA), #c3646 (2833% DA),
#48165 (2629% DA). All of them demonstrated certain solubility in
carbonic acid with the achievable concentrations of up to a few g L
1
.
The best solubility was detected for the chitosan of a low molecular
weight grade (catalogue number: #448869), which was selected for
the further experiments and used without any purication.
Using gel permeation chromatograph (Agilent 1200) calibrated
with pullulan standards (from 1.08 to 710 kg mol
1
) we determined
the molecular weight of this chitosan sample: M
w
= 210 kg mol
1
,
M
n
= 80 kg mol
1
(25 C, aqueous buffer solution, 0.2 M acetic acid,
0.15 M ammonium acetate, 1 ml min
1
), which correlated well with
M
= 80 kg mol
1
as determined by us from viscosity measurements
(25 C, aqueous solution, 0.3 M acetic acid, 0.2 M sodium acetate,
using MarkKuhnHouwink equation with previously described coef-
cients according to Ref. [63]). The chitosan sample had DA of 1624%
according to our data of potentiometric titration and IR spectroscopy
(Thermo Nicolet IS5 FT-IR).
In the experiments we used CO
2
of high purity (N99.997%; Linde Gas
Rus, Russia) and freshly puried Milli-Q water (Milli-Q Synthesis).
Collagen matrices of bovine pericardium were picked out, GA-
stabilised and sterilised in accordance with technological regulations
approved for surgery practice in A.N. Bakulev Scientic Center for
Cardiovascular Surgery. In general, this treatment includes stabilisation
by 0.625% GA aqueous solution with intermediate washing with aque-
ous solution of 1% sodium dodecyl sulphate and with HEPES buffer.
After washing with sterile saline solution, the collagen samples were
immersed into a high pressure vessel for subsequent coating with
chitosan.
2.2. High pressure setup
The experimental setup was the same, as it was previously used for
deposition of chitosan from solutions in scCO
2
or carbonic acid onto
mica as described before [56,57]. Briey, the setup consists of a high
pressure generator equipped with pressure sensors and a thermostati-
cally controlled stainless steel high pressure vessel (inner volume of
30 ml), which are connected together by a set of capillaries. The setup
can sustain the pressures of up to 80 MPa (as limited by the vessel).
2.3. Preparation of chitosan solution in carbonic acid
The preparation of chitosan solution in carbonic acid was performed
as described before [57] and illustrated in Fig. 1. We put 150 mg of chi-
tosan powder into the high pressure vessel. The half of the vessel
(15 ml) was lled with freshly prepared Milli-Q water. Then it was
closed and lled with liquid CO
2
up to high pressure (30 MPa) at room
temperature (2325 C) and left at these conditions for dissolution
andequilibration of CO
2
inwater andchitosan in the thus generated car-
bonic acid with periodic agitating by a magnetic stirrer. The total time of
dissolution was selected in the range of several days (714 days, but no
signicant difference was observed regarding the properties of the
obtained solutions).
128 M.O. Gallyamov et al. / Materials Science and Engineering C 37 (2014) 127140
2.4. Coating bovine pericardium with chitosan from previously prepared
chitosan solution in carbonic acid
After 714 days of chitosan dissolution we decompressed the vessel
to the normal pressure. During decompression liquid CO
2
phase left the
vessel but water phase with dissolved chitosan remained in the vessel.
Next, the vessel was opened and several collagen samples withdifferent
sizes (0.5 5 cm
2
or 1 1 cm
2
) in approximately similar total amount
(as determined by their total surface area) were put into the water
phase, see Fig. 1. Then the vessel was closed and tightened again and
was lled with liquid CO
2
up to the same high pressure of 30 MPa at
room temperature (2325 C). The density of liquid CO
2
at pressures
and temperatures of the exposure is equal to about 0.970.98 g/cm
3
,
as calculated using the NIST Thermophysical Properties of Pure Fluids
software. That means it is still smaller as compared with water density.
So, liquid CO
2
phase is located in the upper half of the vessel. Therefore,
the collagen samples were exposed to chitosan solution in carbonic
acid remaining in the lower half of the reactor and spontaneous adher-
ence of the chitosan to the collagen took place (Fig. 1). The total expo-
sure time was 2 h. After that the vessel was opened and the collagen
matrices coated with chitosan were retrieved (Fig. 1). All loosely
bound chitosan molecules were removed by subsequent washing the
collagen matrices with distilled water. The thus prepared samples
were kept in 0.9% NaCl solution.
2.5. Reference bovine pericardiumsamples treated in carbonic acid without
chitosan
For comparison, we prepared dummy collagen samples, which
were treated in pure carbonic acid without chitosan at the same condi-
tions (30 MPa, room temperature (2325 C), 2 h). The procedure
was exactly the same as illustrated in Fig. 1 and described above
(Section 2.3.) just without putting chitosan into the high pressure
vessel.
2.6. Determination of the amounts of chitosan introduced into the collagen
matrices by tritium label
We found out that the amounts of chitosan typically introduced into
collagen matrices from the prepared solutions in carbonic acid were
rather small. Therefore, we applied tritiumlabelling (with [
3
H]chitosan)
as a precise method to determine the chitosan at collagen loadings. [
3
H]
Chitosan was obtained by tritium thermal activation method [64] as
illustrated in Fig. 2.
The initial chitosan (30 mg) was uniformly distributed on the walls
of a glass reactor. The reactor was lled up with 1 Pa gaseous tritium-
hydrogen mixture with 17.8% tritium content after vacuumization.
The tungsten lament in the centre of the reactor was heated by the
electric current up to 1940 K to generate tritium atoms for 10 s. After
the reaction with tritium atoms the target was retrieved from the reac-
tor. The labelled samples contained labile tritiumin exchange positions
of the macromolecule (Fig. 2). The purication from the labile tritium
was carried out by keeping the chitosan within several days in distilled
water at 4 C with removal of water on a rotary evaporator (repeating
3 times) and subsequent washing with ethanol by means of centrifuga-
tion (repeating 5 times). According to measurements, radioactivity of
the prepared [
3
H]chitosan was 54.3 MBq, the specic radioactivity
was 8.1 MBq mg
1
(9.5 TBq mmol
1
recalculated to 100% tritium in
initial reaction gas and chitosan M
w
= 210 kg mol
1
).
A sample of [
3
H]chitosan solution was mixed with the initial
(unlabelled) chitosan dissolved in 17% acetic acid as to obtain a product
with the specic radioactivity of 27 kBq mg
1
. Further, acetic acid was
removed by a freeze drying. This labelled chitosan was dissolved in car-
bonic acid and applied to coat collagen tissue using the normal proce-
dure as illustrated by Fig. 1 and described above (Sections 2.3 and 2.4).
After coating of the collagen tissue with the labelled chitosan and
subsequent washing, the excess of liquid was removed with a lter
paper and the composite tissue sample was placed in a round bottom
ask equipped with a backow condenser. Then 1 ml of 54% nitric
Fig. 1. Scheme of modication of bovine pericardium in solutions of chitosan in carbonic acid, room temperature.
acid (density 1.33 g cm
3
) was added to the ask, which afterwards
was heated up to boiling. Bothcollagen and chitosan were decomposing
with bubbling gas (CO
2
and oxides of nitrogen) coming out and the
solution eventually became transparent. After decomposition of the
matrix the boiling of the solution was performed during 10 min. The so-
lution was cooled down to room temperature, and its radioactivity was
measured. Then 0.1 ml of the solution was placed in a vial with 10 ml of
a scintillation cocktail Ultima Gold (Perkin Elmer) and nally the radio-
activity was detected by a liquid scintillation spectrometer RackBeta
1215 (LKB, Finland). Control experiment was conducted using boiling
[
3
H]chitosan without collagen tissue according to the procedure de-
scribed above.
2.7. ATR-IR analysis of the surface of the collagen samples
Attenuated total reectance infrared spectroscopy (ATR-IR) mea-
surements were performed using Thermo Nicolet IS5 FT-IR device
with ID5 ATR accessory (diamond crystal) for treated and untreated
bovine pericardium samples, which were completely dried before the
measurements.
2.8. FE-SEM analysis of the surface of the collagen samples
The outer surfaces of collagen matrices with and without chitosan
coating were studied as is by eld-emission scanning electron micro-
scope (FE-SEM) HITACHI SU 8000 (Hitachi, Japan) without any metallic
sputtering [65] to preserve native surface structure. Images were ac-
quired in a secondary electron mode at 0.51.0 kV landing voltages
and at working distance of 810 mm.
2.9. TEM analysis of the structure of the collagen samples
Ultra-thin sections of the air-dried collagen matrices were cut with a
diamond knife (Diatome Ultra 35) on a Reichert-Jung ultra-microtome
(Germany) at room temperature. Bright-eld micrographs were ob-
tained on a LEO 912 AB (Germany) transmission electron micro-
scope (TEM) at an accelerating voltage of 100 kV (LaB
6
cathode) in a
digital format (with information density of 2048 2048 points) and
analysed using an adapted image processing software FemtoScan
(Advanced Technologies Centre, Russia)
2.10. Tensile testing of collagen samples with and without chitosan coating
We used tensile testing to compare mechanical properties of the
following bovine pericardium samples: i) control samples, i.e. initial,
just as stabilised by 0.625% GA; ii) dummy samples treated in carbonic
acid without chitosan; and iii) chitosan-coated samples treated in chi-
tosan solution in carbonic acid. For the testing, the collagen samples
were initially cut into rectangular pieces with lateral sizes of approxi-
mately 5 0.5 cm
2
. Two types of samples were cut out with respect
on their axial or radial orientation relatively the main orthogonal direc-
tions of bovine pericardium. After that treatment in carbonic acid with
or without chitosan was performed. Untreated samples were used as
controls for comparison. Stressstrain curves were recorded with a
universal testing machine Zwick/Roell Z 2.5/TN1S (500 N load
cell, constant cross-head speed of 50 mm/min, effective gauge length
of 20 mm). Wet samples as just retrieved from saline solution were
tested in both main orthogonal directions of the collagen tissue. In
total, for each set sixteeneighteen samples (per sample type/per
direction) were tested.
2.11. In vivo calcication tests of collagen samples with and without
chitosan coating
Due to fast metabolism of rats we adopted subcutaneous tests on
them as a convenient model of in vivo express-testing with relatively
short characteristic times of calcication, see Fig. 3. Collagen samples
of ca. 1 1 m
2
lateral size were implanted subcutaneously into Wistar
rats (110 5 g). The protocol was approved by A.N. Bakulev Scientic
Center for Cardiovascular Surgery in compliance with internal guide-
lines for care and use of laboratory animals.
There were ten rats per sample type. Each rat received simultaneous-
ly two samples: a control sample (bovine pericardium, initial, just as
stabilised by 0.625% GA) and either a dummy sample (pericardium
treated in carbonic acid without chitosan) or a chitosan-coated sam-
ple (pericardium treated in chitosan solution in carbonic acid). The
Fig. 2. Scheme of preparation of [
3
H]chitosan. The sign T indicates all available positions for isotopic exchange (before and after the removal of labile tritium). The substitution degree for
chitosan is not higher than one atom per a monomer unit.
operations were performed under thiopental anaesthesia. After implan-
tation the animals were kept in normal conditions receiving food ad
libitum. The diet includes calciumregulator ergocalciferol (vitamin D2).
After 4 months since the implantation, the collagen samples were
retrieved from the rats. There were no traces of any inammatory
reaction observable. The amount of calcium in the samples was deter-
mined by an atomic absorption spectrophotometer AAS-5100/Zeeman
(Perkin-Elmer, Germany).
2.12. Biocompatibility of collagen samples with and without chitosan
coating
The biocompatibility tests were performed as illustrated in Fig. 4.
We compared freshly prepared extracts of: i) control samples (bovine
pericardium, initial, just as stabilised by 0.625% GA), ii) dummy
samples (pericardium treated in carbonic acid without chitosan) and
iii) chitosan-coated samples (pericardium treated in carbonic acid
with dissolved chitosan). The extracts of the samples were prepared
in 0.9% NaCl solution at 37 C during (24 2) hours with the relation-
ship between sample surface and extracting medium volume as of
3 cm
2
: 1 ml.
The test cells (mouse broblast line NIH/3 T3) were cultured in
Dulbecco's modied Eagle's medium (DMEM) supplemented with 10%
foetal calf serum, 50 g/ml gentamicin, 2 mM L-glutamine and 1 mM
HEPES. The cells (12 10
5
cells ml
1
) were seeded in 96-well culture
plates (Corning-Costar, USA) and were incubated during (24 2)
hours at 37 C in a humidied atmosphere containing (5 1) % CO
2
.
Under these conditions (80 10) % conuence was achieved.
Thus prepared extracts and corresponding controls were introduced
into the wells with the test cells (eight wells per sample) in the amount
of 100 l per well andwere incubated during (24 2) hours at 37 C in
a humidied atmosphere containing (5 1) %CO
2
. The DMEMwithout
serumwas used as a negative control. Zinc nitrate solution (9.95 mg Zn
in 12 wt.% nitric acid, 200 times diluted with 0.9% NaCl solution) was
used as a positive control. Additionally, 0.9% NaCl solution was used as
a reagent control.
After (24 2) hours of incubation of the extracts/controls in
the wells with test cells we removed contents of the wells and washed
the wells with phosphate-buffered saline. We performed an estimation
of the culture microscopically for the presence of morphological
changes and/or a decrease in the density of cells using vital dye trypan
blue. We used following reference for interpretation of the results of the
cytotoxic reactions on the mouse broblasts NIH/3 T3. The reactivity
was considered as severe if the monolayer had been almost complete-
ly destroyed (more than 70% lysis). The reactivity was considered as
slight if not more than 20% of the cells were round, loosely attached,
and without intracytoplasmic granules (not more than 20% lysis).
The reactivity was considered as none if there were discrete
intracytoplasmic granules observed and no lysis. The positive control
induced severe reaction. Whereas reaction was absent (none reac-
tivity) for both the negative and the reagent controls.
2.13. Antimicrobial properties of collagen samples with and without
chitosan coating
In order to test the ability of bacteria towards adherence and biolm
[6062] formation on collagen tissues we compared i) control samples
(GA-stabilised bovine pericardium as prepared), ii) dummy samples
(GA-stabilised bovine pericardium treated in carbonic acid without
chitosan) and iii) chitosan-coated samples (GA-stabilised bovine
pericardium treated in carbonic acid with dissolved chitosan). Due
to noticeable levels of GA residually realised fromGA-stabilised pericar-
dium, it demonstrate strong activity against bacteria (yet, in a human
body after implantation this activity would cease rather fast). Therefore,
GA-stabilised pericardium is not an appropriate control for testing and
comparing antimicrobial properties of the samples, which are found
to be highly biocompatible. Thus, we used a non-stabilised allogeneic
tissue without any GA-treatment as a control. Additionally, we used
GA-stabilised bovine pericardium with additional washing in aqueous
ammonium solution to inactivate residual GA as an additional control.
The results for both controls were the same. As the test cultures we
selected clinical strains of both Gram-positive bacteria: methicillin-
resistant Staphylococcus aureus (MRSA), Staphylococcus haemolyticus,
and Bacillus cereus, and Gram-negative ones: Pseudomonas aeruginosa,
Escherichia coli, Enterobacter cloacae, and Klebsiella pneumoniae. Testes
with Gram-positive Candida albicans (fungi) were also performed. For
a particular strain we repeatedly performed several tests in order to ob-
tain statistically signicant data (ANOVA) and to determine standard
deviations of mean values to be measured.
The samples being tested of ca. 1 1 cm
2
lateral size were im-
mersed in the suspensions of the bacterial cultures (10
6
cells ml
1
,
McFarland standard [66]) and were inoculated at 37 C for 4 h. This
Fig. 3. Scheme of in vivo express-testing on calcication: subcutaneous implantation of modied and as prepared (control) GA-stabilised bovine pericardiumsamples inrats. Modication
of bovine pericardium was performed in solutions of carbonic acid with or without dissolved chitosan (Fig. 1).
time is sufcient for adherence but not for bacterial growth. Further, we
used these samples to estimate the amounts of CFUs in them. For that
we triturated each sample in 1 ml of sterile 0.9% sodium chloride solu-
tion with sterile glass powder and prepare homogenates, which were
placed on a semi-solid nutrient MuellerHinton medium. We incubated
the inoculations of the test and control samples overnight at 37 C.
After incubation we counted the numbers of colonies both for the test
and control samples. For comparison purposes index of adhesion was
determined as I = log (CFU), where CFUs were counted per sample.
In order to estimate the ability of bacteria to form biolms on the
collagen tissue we used the following modication of the above-
described procedure. After inoculation of the samples in the microbial
suspensions (4 h) the samples with the adhered microbes were addi-
tionally stored during 24 h at 37 C in sterile Petri dishes at the surface
of sterile multi-layered gauze tissue abundantly wetted withsterile 0.9%
sodiumchloride solution. After this additional microbial growing during
the 24-h incubation, the CFU numbers and corresponding indexes of
survivability, I = log (CFU), were determined per each sample.
3. Results and discussion
3.1. Coating bovine pericardium with chitosan from solutions in carbonic
acid: amounts of bound chitosan and changes in tissue morphology
We detected dissolution of the low molecular weight chitosan
#448869 in water saturated with CO
2
at room temperature (25 C)
and at 30 MPa pressure up to concentrations of, at least, several g L
1
.
Indeed, solubility of CO
2
in water increases with increasing pressure
and at 30 MPa (25 C) is about 3 mol% [6769]. The pH value of the
water saturated with pressurised CO
2
decreases with increasing pres-
sure and at the conditions of our experiment is about 2.8 (carbonic
acid) [70,71]. This value of pH should be sufcient for protonating of
amino groups of chitosan and its dissolution [42]. The dissolution of chi-
tosan took place in a closed pressurised vessel during several days. After
decompression we visually observed disappearance of chitosan powder
previously put into the vessel and formation of slightly viscous solution.
The exposure to high pressure does not signicantly affect characteris-
tics of the dissolved macromolecules. We tested with gel permeation
chromatography (GPC) that molecular weight of chitosan after such
an exposure was somewhat (not signicantly) decreased and its lm-
forming ability was not impaired.
After decompression, the values of pHof the solutionwere easy to be
measured and they were found to uctuate aroundthe value of 4 during
many subsequent hours eventually demonstrating increase up to the
value of 5.5. Therefore, the decompression resulted in a spontaneous in-
crease of the pH value from ca. 3 to ca. 4. At these conditions dissolved
chitosan at the concentration of several g L
1
demonstrated a tendency
towards partial precipitation, which could be observed visually with a
characteristic time of several tens of minutes. Thus, after decompression
the amount of protons in the solution is already insufcient to keep
macromolecules dissolved. Apparently, chitosan carbonate is unstable
and macromolecules tend to become uncharged (deprotonated) at the
Fig. 4. Scheme of biocompatibility tests on mouse broblast line NIH/3T3. Extracts of control samples (initial GA-stabilised bovine pericardium), dummy samples (pericardiumtreated in
carbonic acid without chitosan) and chitosan-coated samples (pericardium treated in carbonic acid with dissolved chitosan) were compared.
pH close to neutral values even without being neutralised by a base.
Therefore, by the moment of implantation one may expect absence of
any tendency towards solubility and thus sufcient stability of the pre-
pared from such solutions chitosan coating in a blood stream(pH value
of ca. 7.4).
Taking into account previously reported results on pericardium
immersed in chitosan solutions in water [3238] or in acetic acid
[4346], we expected that dissolved chitosan macromolecules in car-
bonic acid should become bound to the collagen tissue placed into
high pressure reactor (Fig. 1). Indeed, during exposure of pericardium
in such conditions the amino groups of chitosan macromolecules
may form Schiff base bonds with residual free GA groups (Fig. 1). Addi-
tionally, these cationic macromolecules may form links with anionic
sites in the collagen tissue. Therefore, mild additional crosslinking
of the GA-stabilised collagentissues by chitosan is expected. This should
affect stressstrain behaviour of the tissue and its afnity towards
calcium deposit, bacteria and cells. Here, we address those properties
in details. Yet, the possible inuence of carbonic acid itself on the prop-
erties of collagen tissue is unknown (it was never tested before to the
best of our knowledge). Therefore, we have to compare our results
with dummy samples of pericardium exposed to pure carbonic acid
without chitosan at the same conditions (2 h in high pressure vessel).
It was impossible to detect the amounts of chitosan typically intro-
duced into the collagen tissue from such solutions by gravimetric
analysis in any reliable manner. Fortunately, applying the tritium
labelling method (Fig. 2), we manage to determine those amounts
with rather reasonable precision and moderate data scattering. Thus,
we detected that the integral amount of chitosan as typically adsorbed
fromthe prepared solutions in carbonic acid (Fig. 1) onthe collagenma-
trix is equal to about (1.0 0.1) wt% of the matrix (the value is the
mean standard deviation). This is an amount of chitosan, which is
bound sufciently irreversibly to sustain thorough washing with dis-
tilled water after deposition. The standard deviation here was obtained
statistically with several different samples tested.
Next question is where this chitosan is localised. The typical ATR-
IR spectra for treated and untreated pericardium samples are pre-
sented in Fig. 5. For the initial untreated just GA-stabilised bovine
pericardium one can reveal typical peaks, reported in the literature
[72,73]: 32923305 cm
1
(OH, NH, amide A), 3078 cm
1
(amide B),
2928 cm
1
(CH
2
), 16511633 cm
1
(C_O triple peak [74]: within
triple helix, glycine, amide I), 1538 cm
1
(amide II), 1449 cm
1
,
1336 cm
1
, and 1234 cm
1
(NH, proline, amide III). Exposure to pure
carbonic acid mainly results in partial or complete disappearance
of the small bands at 2854 cm
1
(CH
2
) and 1744 cm
1
(C_O), which
may be indicative of extraction of residual glutaraldehyde or some
phospholipids. Treatment with chitosan solution in carbonic acid notice-
ably increases typical spectral features of polysaccharides [75,76]:
1159 cm
1
(bridge C\O\C), 1077 cm
1
(C\O), and 1032 cm
1
(C
6
\OH). Also somewhat increased peak at 2874 cm
1
and shoulder
at ~1377 cm
1
are due to the CH
3
groups of acetylated monomer units
of the introduced chitosan. In general, the spectrum is quite typical for
chitosan-coated pericardium (or collagen) as reported in the literature
[44,73]. It is interesting to note that characteristic chitosan peak at
~3440 cm
1
exists only as a poorly resolved shoulder and most proba-
bly is mainly shifted to the amide A band. This may indicate that amino
groups of chitosan participate strongly in hydrogen and electrostatic
bonding with the collagen matrix. Remaining peak at 1744 cm
1
,
which typically disappears (at least, partially) after treatment with
pure carbonic acid without chitosan, may indicate on less effective
extraction of residual glutaraldehyde in the presence of chitosan in
solution due to a reaction between them. Thus, the obtained ATR-IR re-
sults have illustrated that there is some small amount of chitosan in the
near-surface layer of the bovine pericardiummatrix within the depth of
about a micron.
We used FE-SEM to compare outer surface structure of the collagen
matrices withandwithout chitosancoatings, see Fig. 6. Thus, we studied
the surface structure of the initial collagen tissue (just as stabilised by
0.625% GA) and the same tissue coated by chitosan(exposed to chitosan
solution in carbonic acid). Also, we studied dummy samples, i.e. colla-
gen matrices exposed to pure carbonic acid without chitosan at the
same conditions and exposure times.
Comparing the typical images presented (Fig. 6) one can see that the
structure of the surface of initial GA-stabilised collagen matrix (Fig. 6a)
is similar to the structure of the surface of matrix, which was treated by
pure carbonic acid without chitosan (Fig. 6b). One can clearly discern
the individual collagen brils with a periodicity of 6070 nm, which
corresponds well to the characteristic D-period of the collagen brils
[77]. There is hardly any noticeably difference in the images of surfaces
of the samples treated and untreated in carbonic acid (Fig. 6a and
Fig. 6b). Therefore, we can conclude that the treatment in pure carbonic
acid (without chitosan) does not affect the structure of the surface of
pericardium signicantly.
On the contrary, the surface of collagen matrices after coating with
chitosan looks quite different in FE-SEM(Fig. 6c). Indeed, one can notice
two features. First, the characteristic periodicity of the collagen brils is
not discernible anymore (Fig. 6c). Second, protruding collagen brils
appear as somewhat thicker strands (Fig. 6c). Yet, they are quite visible
and thus not covered with any thick chitosan lm. Therefore, we may
conclude, that if there is any chitosan lm at the outer surface of the
collagen matrix, its thickness is of several tens of nanometres, at most.
Such a thin lmwould be capable of masking the periodicity of collagen
brils. But it would not cover and hide the individual strands of collagen
brils, which are indeed quite visible in Fig. 6c. The volume of such a
thin lm is much smaller as compared with the total amount of intro-
duced chitosan. Thus we proved that only a minor fraction of chitosan
may be adsorbed on the outer surface of the collagen matrix. Therefore,
the chitosan is mainly localised somewhere inside.
To reveal the localisation of the chitosan inside the collagen matrix,
we applied ultra-microtomy in conjunction with subsequent TEM
analysis of the ultra-thin sections. The chitosan phase in the matrix
is rather small (about 1% only). It has the same density and the same
electron contrast as the collagen phase. Thus, it is rather hard to be dis-
tinguished in the TEMimages. Therefore, we exploited silver nanoparti-
cles as a special marker of the chitosan. For that we adsorbed chitosan
on the matrices from solutions in acids with pre-reduced (by chitosan)
and stabilised (by chitosan) silver nanoparticles. The nanoparticles
were prepared from a silver nitrate precursor using chitosan as both a
reducer and a stabiliser [78]. Thus, one type of samples was the collagen
tissue, exposed to the pre-formed solution of chitosan with silver
3600 3400 3200 3000 2800 1800 1600 1400 1200 1000 800
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
1
3
7
7
c
b
a
1
0
3
2
1
0
7
7
1
1
5
9
1
2
3
4
1
3
3
6 1
4
0
0
1
4
4
9
1
5
3
8
1
6
5
1
-
1
6
3
3
1
7
4
4
2
8
5
4
2
8
7
4
2
9
2
8
3
4
4
0
3
0
7
8
3
2
9
2
-
3
3
0
5
a
b
s
o
r
b
a
n
c
e
wave number, cm
-1
initial GA-pericardium
GA-pericardium, after carbonic acid
GA-pericardium, chitosan-coated
Fig. 5. ATR-IR spectra for initial bovine pericardium as GA-stabilised without any further
treatment (a), GA-stabilised bovine pericardium treated in pure carbonic acid without
chitosan (b) and GA-stabilised bovine pericardium coated with chitosan from solution
in carbonic acid (c).
nanoparticles in carbonic acid (at high pressure). The second type
of samples for comparison was the collagen tissue, exposed to the pre-
formed solution of chitosan with silver nanoparticles in 2% acetic
acid (at normal pressure). The corresponding images are presented
in Fig. 7.
In Fig. 7 one can see the chitosan phase as marked by Ag-
nanoparticles. This phase forms rather straight channels in the collagen
matrix (highlighted by outlines in Fig. 7) for both types of the samples.
The net of the channels is particularly developed for the sample treated
with chitosan dissolved incarbonic acid (Fig. 7a). These channels appar-
ently represent pathways of chitosan penetration into the matrices. One
may also notice that the collagen tissue looks signicantly less porous
(less spongy-like) after the exposure to chitosan solution at high pres-
sure (Fig. 7a) as compared with the exposure to chitosan solution at
Fig. 6. FE-SEMimages of the outer surface of bovine pericardium: a) initial GA-stabilised collagen matrix, b) the collagen matrix exposed to pure carbonic acid without chitosan, and c) the
collagen matrix with chitosanadsorbedfromsolutions incarbonic acid. Images inthe upper rawwere recordedat 25 k magnication. Rectangles here indicate areas rescanned at higher
magnication (50 k, corresponding images in the lower raw).
Fig. 7. TEM images of ultra-thin sections of bulk bovine pericardium: a) collagen matrix treated by chitosan solution in carbonic acid at high pressure, and b) collagen matrix treated by
chitosan solution in acetic acid at normal pressure. Ag nanoparticles are explored as a marker of chitosan phase. Channels of the chitosan phase are highlighted by outlines.
normal pressure (Fig. 7b). Therefore, some structural changes do occur
in the collagen matrix during deposition of chitosan from solutions in
carbonic acid. Those changes are most probably related to the inuence
of high pressure of the solvent. We believe that macropores collapse
under the inuence of high pressure. The dissolved chitosan penetrates
via the narrower channels formed by borders of the collapsed pores
(Fig. 7a). Further, the chitosan effectively glues the walls of the col-
lapsed pores, so as they remain collapsed after decompression. Indeed,
the chitosan amino groups should form Schiff base bonds with residual
aldehyde groups of the GA-stabilised collagen matrix [32]. Due to this
inuence of chitosan glue, the collagen matrix becomes more solid
after treatment with chitosan solution in carbonic acid having notice-
ably less pores and voids (Fig. 7a). Such morphology changes should
manifest themselves in an alteration of stressstrain properties of the
modied bovine pericardium, which is indeed observed experimentally
(see below).
Thus, it was revealed that when introduced from solutions in car-
bonic acid chitosan mainly does not just cover the outer surface of a col-
lagen matrix but penetrates deeply into its bulk. Here it xes
macropores, which are collapsed due to high pressure applied. The ma-
trix becomes more monolithic being effectively reinforced by the chito-
san. Therefore, it can be assumed that the collagen matrix with chitosan
introduced from solutions in carbonic acid should demonstrate im-
proved mechanical properties.
3.2. Stressstrain behaviour of bovine pericardium with and without
chitosan coating
Mechanical properties of collagen appear to be a critical factor in
determining the long-term durability of GA-treated materials for bio-
logical prosthetic heart valves [79]. As far as we detected peculiar mor-
phological changes in the collagen tissue after its treatment in chitosan
solutions in carbonic acid there were strong reasons to expect changes
in stressstrain behaviour.
We tested mechanical strength properties of initial GA-stabilised bo-
vine pericardium and compared them with the properties of the same
tissue additionally treated either in chitosan solution in carbonic
acid or in pure carbonic acid without chitosan. The obtained results
are summarised in Table 1. It is known, that type I collagen tissue
including bovine pericardiumdemonstrates different stressstrain be-
haviour in two major orthogonal directions, which are determined
mainly by orientation of bril growth [8082]. Therefore, the data in
Table 1 are presented for the tested samples, which were cut out in
both orthogonal directions (the corresponding numbers are indicated
with signs of || and ).
For the initial GA-stabilised bovine pericardium we detected values
of ca. 15 (||)/7 () MPa for ultimate tensile strengths and ca. 41% for cor-
responding maximal strains (Table 1), which both are in general agree-
ment with the literature data [30,33,8084]. From the data of Table 1
one can also notice that the treatment of the collagen matrix in carbonic
acid (with or without chitosan) results in an increase of the tensile
strength (pay attention to the values for || direction). The increase is rel-
atively small (1114% in total), but ANOVAreveals signicant difference
(p b 0.05). The corresponding strains at maximal stress increase also
(for the same || direction). Particularly signicant increase (about 50%
in total) is detected for the samples coated with chitosan.
Concerning elastic behaviour, it is known, that stressstrain curves
for bovine pericardium [82] and for some other collagen tissues
[28,85] typically have two distinct regions: of small strains (up to
a fewper cents) and of high strains (up to maximal) with two quite dif-
ferent effective Young modules. We measured both of the values and
marked them as E
low
and E
high
, respectively (Table 1). The initial elastic
moduli (E
low
) were found to be reduced as a result of the exposure of
the matrices to pure carbonic acid without chitosan. This is probably
due to an extraction of some components of the tissue. But it is restored
or preserved in the presence of chitosan in the solution, apparently, due
to some reinforcement or suppression of the extraction. The nal elastic
moduli (E
high
) were found to be always reduced as a result of the expo-
sure of the matrices to carbonic acid with or without chitosan. Here
again this is most probably due to extraction of some components of
the tissue. One should mention that early used GA-xation at elevated
pressures was known to produce, on the contrary, stiffer tissues,
which typically demonstrated faster valve deterioration [6]. Differently,
our method benecially produces softer tissue with good tensile
strength. Therefore an improved durability is to be expected.
Thus, we may conclude that the deposition of chitosan on bovine
pericardium from solutions in carbonic acid renders the collagen tissue
being more exible, less rigid, with somewhat improved stressstrain
properties. Therefore, a risk factor of a fatigue failure of bioprostheses
is expected to be reduced. Additionally, the obtained results in general
support our hypothesis about reinforcement of the collagen matrices
by the adsorbed chitosan.
3.3. Calcication of bovine pericardiumwith and without chitosan coating:
in vivo tests
We performed subcutaneous calcication tests on rats (Fig. 3),
which demonstrated that treatment of bovine pericardium in solutions
of chitosan in carbonic acid dramatically mitigates calcication process.
The data are summarised in Table 2. Each rat received jointly two pieces
of bovine pericardium: modied one and reference (control) one just
as prepared. Therefore, in order to reduce data scattering and more
convincingly compare results of different experimental sessions, we
present in Table 2 the ratios of the mean calcium content in the two
types of pieces. The mean amount of calcium in chitosan-coated colla-
gen matrices detected after 4-month implantation test in rats was ca.
110 times lower as compared with the mean amount detected in nor-
mal (untreated) GA-stabilised tissue at the same conditions. For each
particular rat calcication in coated samples was always dramatically
lower than in control samples, which were implanted pairwise.
Table 1
Summary of stressstrain behaviour characteristics of bovine pericardium samples cut out in two main orthogonal directions. Control samples (initial, just as stabilised by 0.625% GA),
dummy samples (treated in carbonic acid without chitosan) and chitosan-coated samples (treated in chitosan solution in carbonic acid) were tested.
Number of samples Ultimate tensile strength,
ts
, MPa Rupture strain, , % Initial Young modulus, E
low
, MPa Final Young modulus, E
high
, MPa
Bovine pericardium coated with chitosan from solutions in carbonic acid
18 ||: 17 3 ||: 62 2 ||: 6.0 1.2 ||: 41 2
18 : 7 2 : 42 2 : 4.7 0.7 : 23 2
Bovine pericardium treated in pure carbonic acid without chitosan
16 ||: 17 3 ||: 46 2 ||: 3.6 1.2 ||: 57 3
18 : 9 3 : 41 2 : 2.9 1.3 : 33 3
Initial bovine pericardium, as prepared by GA-stabilisation
40 ||: 15.1 0.7 ||: 41 2 ||: 5.7 0.3 ||: 71 1
40 : 7.3 0.4 : 41 1 : 4.6 0.2 : 37 1
The || and symbols indicate two main orthogonal directions of bovine pericardium samples
Even dummy samples exposed to pure carbonic acid without any
chitosan demonstrated suppressed afnity towards calcium deposits,
though to a lesser degree: the reduction here was six-fold only.
In both cases of the treatment in carbonic acid (with and without
chitosan), though the data scattering was rather large, ANOVA revealed
signicant difference (p b 0.05) in comparison with corresponding
controls on the just GA-stabilised (untreated) tissue.
A possible reason for the reduction of calcication after treatment
in pure carbonic acid seems to be related to an effective removal of re-
sidual cellular components, which is conrmed by our preliminary his-
tological observations. Indeed, during the exposure in the high pressure
vessel there is a highly mobile and uctuating mixture of polar liquid
water and non-polar liquid subcritical CO
2
. This polar/non-polar mix-
ture seems to be highly effective in extraction of polar, non-polar and,
particularly, amphiphilic cellular residues.
Yet, treatment of bovine pericardium in chitosan solutions in
carbonic acid is much more effective to mitigate calcication. It is im-
portant to notice that chitosan is known as a biodegradable material.
Nevertheless, apparently it suppresses calcication during the whole
period of 4 month subcutaneous tests on rats, which is relatively quite
a long period taking into account the fast metabolismof rats. Otherwise
such a strong reduction of calcication would not be observed. In order
to explain this long-lasting effect we may presume that chitosan chem-
ically bound to collagen matrix via Schiff base bonds with residual free
GA groups is rather stable material with respect to withstanding possi-
ble biodegradation.
Our number of the 110 times reduction of calcication for the
chitosan-treated samples is comparable to the values reported by
Chanda et al. [3234]. Yet, Chanda deposited specially prepared
water-soluble chitosan [3234] from solutions in water at neutral pH,
which should hardly be sufciently stable on the bioprosthetic valves
in neutral blood stream during long-term operation in a heart. Our
chitosan coating is adsorbed from carbonic acid at high pressure. It
spontaneously loses charge after decompression: macromolecules
become deprotonated. Thus, without any neutralisation with an alkali,
it is not anymore soluble in neutral media such as blood plasma and
should be sufciently stable here. Therefore, we believe that this method
is more promising as compared to the one previously patented by us
[86], where 2040 fold reduction of calcication was achieved via depo-
sition of an electrostatically-stabilised bilayer of chitosan derivatives
from aqueous solutions at normal pressure [87] (the corresponding
numbers for calcication in such samples are also presented in Table 2
for comparison).
In the present method the inuence of high pressure seems to be
signicant. We believe that the high pressure promotes penetration of
chitosan solutions into very deep pores of the collagen matrix thus
ensuring particularly thorough modication. Typically, pores of bovine
pericardium are well lled with aqueous solutions as far as their walls
are hydrophilic and are well wetted with water. Nevertheless, some
small part of the pores may remain non-wetted due to local hydropho-
bic moieties, entrapped bubbles, or pores radii being too small. Applied
external pressure (30 MPa) forces the chitosan solutions to wet all
those pores. The chitosan being forced to penetrate into pores at high
pressure lls effectively an inter-bril and inter-tropocollagen space.
Besides, not only are the residual free aldehyde groups efciently
neutralised all over the whole pericardium bulk, but also the pores,
voids and cavities in the tissue are collapsed at high pressure and
stapled with the chitosan glue. The combination of all these factors
seems to result in the particularly high degree of longed-for calcium
mitigation observed. Of course subcutaneous tests in rats are only a
convenient model system with fast characteristic times, which does
not completely represent real conditions in the blood stream of
the heart. The results obtained with such express-testing procedures
may be somewhat biassed due to the absence of the circulatory blood
ow. They may be also strongly affected by inammatory reactions.
Nevertheless, in all tests we systematically observed signicantly
(ANOVA) reduced calcication intensity, particularly for samples coated
with chitosan (two orders of magnitude reduction). Thus, these chal-
lenging results justify further development and approbation of the
promising approach proposed.
3.4. Biocompatibility of bovine pericardium with and without chitosan
coating
Residual cytotoxicity due to unbound or loosely bound glutaralde-
hyde is very typical for GA-stabilised bovine pericardium [88]. We
tested cytotoxicity of extracts from the treated and untreated GA-
stabilised tissues against culture of mouse broblast line NIH/3T3 as
a model system, see Fig. 4. Treatment of the GA-stabilised bovine peri-
cardium was performed in carbonic acid with or without chitosan. The
results are summarised in Table 3.
We revealed that, indeed, the just as prepared GA-stabilised bovine
pericardium demonstrated a minor but quite noticeable cytotoxicity.
Yet, in overall lysis was insignicant (less than 20% of the conuence
was affected). On the contrary, the samples treated in carbonic acid
(either with or without chitosan) have not induced any cytotoxic reac-
tion at all. No lysis was observed.
Apparently, mixture of polar liquid water and non-polar liquid sub-
critical CO
2
is highly effective in extracting of any cytotoxic residues, in-
cluding unreacted GA. This is conrmed by the ATR-IR data (see Fig. 5
and corresponding discussion in the text above). Chitosan is well
known as a highly biocompatible polymer [42], thus it is just natural
that it suppresses cytotoxicity of any residual GA groups masking
them and binding to them. In our scheme chitosan is adsorbed on
Table 2
Relative reduction incalcication for different bovine pericardiumsamples: results of subcutaneous calcication tests onrats (Fig. 3). Normalised amount of calciumis calculated as a ratio
of the detected mean calcium content in the modied samples to the detected mean calcium content in the reference bovine pericardium samples just as prepared by GA-stabilisation
(in 4 months after joint implantation of the two types of sample pieces to each rat).
Bovine pericardium:
Coated with chitosan from solutions in
carbonic acid
Post-treated in pure carbonic acid
without chitosan
Coated with chitosan-based bilayer from aqueous/acetic
acid solutions according to patent [86]
Normalised amount of calcium, % 0.9 17 36
Table 3
Results of biocompatibility tests for extracts from bovine pericardium samples and controls with respect to mouse broblast line NIH/3T3.
Extracts from samples of bovine pericardium: Controls:
Coated with chitosan from solutions
in carbonic acid
Post-treated in pure carbonic
acid without chitosan
Initial, as prepared by
GA-stabilisation
DMEM, negative Zn in HNO
3
, positive NaCl solution, reagent
Reactivity: None None Slight None Severe None
(No lysis) (No lysis) (Lysis b 20%) (No lysis) (Lysis N 70%) (No lysis)
bovine pericardium from solutions in carbonic acid. This solvent pro-
duces no cytotoxic residues in the modied tissue in distinct from
some other solvents for chitosan, such as acetic acid. From this view-
point the observed complete biocompatibility of the modied samples
is quite expectable. Of course, this test is rather simple and may be con-
sidered as a rst step only, but the results obtained are promising and
they justify our approach in general.
3.5. Antimicrobial activity of bovine pericardiumwith and without chitosan
coating
We compared antimicrobial activity both of the chitosan-coated
samples (exposed to chitosan solution in carbonic acid) and dummy
samples (exposed to pure carbonic acid without chitosan). As far as
there is an evident residual cytotoxicity of GA-stabilised bovine pericar-
dium (Table 3) this material cannot serve as an appropriate control for
comparison. This residual cytotoxicity was found to affect adherence
of bacteria strongly. Indeed, both chitosan-coated and dummy samples
do not have any cytotoxicity at all (Table 3). Therefore, we used a non-
stabilised allogeneic tissue without any GA-treatment as a positive
control. Similar results were obtained if instead of the allogeneic tissue
we used GA-stabilised bovine pericardium specially post-treated with
ammonium solutions for inactivation of residual unbound GA.
Most relevant [6] clinical strains of Gram- positive and Gram-
negative bacteria were tested, which are mainly responsible for bacte-
rial postoperative complications (preliminary results of the testing
were partially published in our previous paper [89]). An articial inocu-
lation of the collagen samples was performed during 4 h. Further, CFU
numbers were calculated for the samples. The results are summarised
in Table 4 (marked as just after inoculation). This 4-h inoculation is
not sufcient for biolm formation (it is only a test for adherence).
Therefore, for some strains (of the most interesting cases) we per-
formed additional testing with prolonged 24-h incubation after the
initial inoculation. The bacteria may growand demonstrate their ability
to form biolms during this prolonged incubation. The detected after
that CFU numbers are also summarised in Table 4 (marked as after
incubation).
Fromthe data obtained (Table 4) we can drawthe following conclu-
sions. First of all, at the stage of initial inoculation, adherence of all the
tested strains to the collagen tissue is noticeably reduced for the bovine
pericardium samples treated in carbonic acid with or without chitosan
as compared with the control of the allogeneic tissue (or ammonium-
inactivated GA-stabilised pericardium). Complete inhibition of both
tested types of Gram-positive Staphylococcus just after adsorption is ob-
served for the samples with the chitosan coating already at the stage of
initial inoculation. Certain reduction in adhesion at the adherence stage
is observed also for the same samples with respect to P. aeruginosa and
K. pneumoniae. One can also notice reduced adhesion of Gram-negative
K. pneumoniae and E. coli to the collagen tissue treated in pure carbonic
acid.
We can propose following explanations of the observed regularities.
Firstly, we have to explainreduced adhesionof the microbes at the stage
of initial inoculation at the surfaces of dummy samples treated in pure
carbonic acid without chitosan. Using ATR-IR (see Fig. 5 and corre-
sponding discussion in the text) we found out that carbonic acid
may extract efciently some residual phospholipids. Our preliminary
histological data (not shown here) also demonstrate signicantly re-
duced amount of residual cellular material in the bulk of pericardium
matrix after its exposure to pure carbonic acid. This efciency of the
extraction may be explained by the peculiar properties of pressurised
H
2
O (polar)/CO
2
(non-polar) mixture.
Therefore, the amount of possible anchoring sites for the patho-
genic microbes is reduced. Further, such a treatment with carbonic
acid may result in an increased negative charge of the pericardium
surface [90,91]. Therefore, there would be an electrostatic repulsion be-
tween the sample surface and either teichoic acids or lipopolysaccha-
ride molecules in the bacterial shell [92,93]. Thus, both types of
bacteria should encounter certain repulsion from the sample surface.
In particular, for Gram-negative bacteria localisation of negative charges
in the outer layer of the outer membrane shell should be particularly
close to the presumably negatively charged pericardium surface
resulting in a particularly effective repulsion. Thus, we believe that the
reduced amount of anchoring sites and possibly increased negative
charge are the reasons for the detected reduced afnity of the tissue
treated in pure carbonic acid towards Gram-negative bacteria.
Further, we should remember that there is a thin chitosan lm
adsorbed at the surface of bovine pericardium treated in chitosan
solution in carbonic acid, see Fig. 6 and discussion above. Strong anti-
bacterial properties of chitosan and its lms are well-known [9497],
including reported ability to inhibit biolm formation [98]. It is also
well documented, that chitosan is predominantly effective against
Gram-positive bacteria [97100], in particular, against Staphylococcus
[97,101,102]. This explains our observation of complete inhibition of
both tested strains of Staphylococcus after adsorption at the stage of ini-
tial inoculation. According to the prevailing literature model, negatively
charged teichoic acids are the preferred targets for chitosan in the cell
wall of Gram-positive bacteria [103] when it is acting as a membrane
perturbant [104]. Thus, the polycationic nature of chitosan is a major
factor contributing to its antimicrobial activity.
Even more intriguing are the results demonstrating strong inhibition
of the majority of the tested strains at the modied collagen tissue
during the stage of 24-h incubation after the initial 4-h inoculation
(Table 4). Indeed, even if the microbes do adhere more or less succes-
sively, they do not survive at the surface during subsequent incubation.
Table 4
Summary of antimicrobial properties: dummy samples (bovine pericardiumtreated in carbonic acid without chitosan), chitosan-coated samples (bovine pericardiumtreated in chitosan
solution in carbonic acid) and biocompatible allogeneic tissue (as a control). CFU numbers were detected just after 4-h inoculation and after additional subsequent 24-h incubation. The
mean values and standard deviations are presented.
Gram-positive Gram-negative
S. aureus S. haemolyticus B. cereus C. albicans Ps. aeruginosa E. coli E. cloacae K. pneumoniae
I = log CFU I = log CFU I = log CFU I = log CFU I = log CFU I = log CFU I = log CFU I = log CFU
Bovine pericardium coated with chitosan from solutions in carbonic acid
Just after inoculation (No CFU)
a
(No CFU)
a
2.92 0.02 2.85 0.01 1.67 0.05 3.46 0.01 2.68 0.01 1.70 0.06
After incubation (No CFU)
a
(No CFU)
a
2.93 0.02 1.64 0.03 0.95 0.05 0.30 0.04 (No CFU)
a
(No CFU)
a
Bovine pericardium treated in pure carbonic acid without chitosan
Just after inoculation 3.18 0.01 3.74 0.01 2.56 0.03 2.53 0.02 2.25 0.03 1.92 0.01 3.23 0.01 1.40 0.06
After incubation 2.59 0.08 (No CFU)a 2.34 0.04 (No CFU)a
Biocompatible allogeneic tissue (control)
Just after inoculation ~5 ~5 ~5 ~5 ~5 ~5 ~5 ~5
After incubation ~5 ~5 ~5 ~5 ~5 ~5 ~5 ~5
a
Complete inhibition.
Thus, we detected that E. cloacae and K. pneumoniae were completely
inhibited at the surface of chitosan-coated collagen after rather suc-
cessful inoculation. Further, C. albicans and, particularly, E. coli and
P. aeruginosa also demonstrated strongly reduced survivability. Both
Staphylococcus were completely inhibited already at the stage of initial
inoculation. The only strain among tested, which survive at the surface
of chitosan-coated collagen, is B. cereus. Similar effect was observed
with the dummy samples treated in pure carbonic acid without
chitosan. At the surfaces of these samples two strains, E. coli and
K. pneumoniae, also did not survive after rather successful inoculation
of the tissue.
The reason for this effect is not yet clear. But we believe that the
main contribution may be due to the excessive charging of the modied
surfaces. One can expect locally increased negative charge of the surface
of collagen treated with pure carbonic acid and positive charge of the
surface of chitosan-coated samples due to an overcharging effect (see
the discussion above). These locally distributed charges may electro-
statically disturb and disrupt the microbial membranes. Additional con-
tribution may be due to highly effective removal of any residual cellular
material from the both types of samples as a result of treatment in car-
bonic acid with or without chitosan. The effective elimination of the in-
oculated microbes indicates on suppression of their tendency to form
biolms.
Thus, on one hand, treatment in carbonic acid with or without chito-
san renders the surface of bovine pericardium highly biocompatible
(Table 3). The biocompatibility of such tissues is much higher as com-
pared with the just GA-stabilised bovine pericardiumand is comparable
to the biocompatibility of the allogeneic tissue or GA-stabilised bovine
pericardium post-treated with ammonium solutions. On the other
hand, in spite of this biocompatibility, bovine pericardium treated in
carbonic acid with or without chitosan demonstrates much stronger re-
sistance against articial inoculation with microbes as compared with
the allogeneic tissue or GA-stabilised bovine pericardium post-treated
with ammonium solutions.
Therefore, the proposed method of modication of bovine pericar-
dium is capable to improve its antimicrobial properties. Particularly ef-
fective is the chitosan adsorbed from solutions in carbonic acid against
Gram-positive bacteria (Staphylococcus) apparently due to rupture of
their membranes during initial interaction and rst contact with the
outer chitosan lm coating the surface of the modied bovine pericar-
dium. Other strains may adhere at the surface of the chitosan-coated
samples more or less successively but mainly do not survive during
subsequent incubation. Therefore, the developed composite material
seems to be very promising for biomedical applications attractively
combining absolute biocompatibility with strong antimicrobial activity.
This antimicrobial activity is expected to be long lasting since it is
induced by a lm of a high molecular weight compound, which is in-
soluble in plasma medium and cannot be washed out. Our data on cal-
cication (Table 2) demonstrate that the effect of chitosan is prolonged
in spite of its known biodegradability. Thus, for a long time the risks of
bacterial endocarditis as well as operation-related bacterial aftereffects
should be suppressed.
4. Conclusions
Saturation of water with CO
2
at high pressure results in formation of
anacidic medium(carbonic acid) withpHvalue of about 3 (at pressures
up to 30 MPa and roomtemperature). Such pressures may be routinely
generated with typical sc CO
2
-extraction equipment commonly used
nowadays in practice of biomedical laboratories in research centres
and hospitals. Carbonic acid is capable to dissolve chitosan materials
with different DAs (1633%, at least) up to concentrations of a few mg
per ml. GA-stabilised bovine pericardium post-treated in carbonic acid
(with or without dissolved chitosan) demonstrates a set of improved
properties, which are highly relevant to the usage of this collagen tissue
as a material for biological prosthetic heart valves.
Further, mixture of polar liquid water and non-polar liquid subcriti-
cal CO
2
apparently results inaneffective extractionof remaining free GA
molecules, phospholipids, unbound protein moieties and other residual
cellular fragments. Indeed, already GA-stabilised bovine pericardium
post-treated in pure carbonic acid (without dissolved chitosan) demon-
strates somewhat improved mechanical properties (it is softer but
durable), enhanced (complete) biocompatibility and six-fold increased
resistivity to calcications in model express-tests (subcutaneous
implantation in rats) as compared to the initial tissue just after GA-
crosslinking. An adhesion and surviving of several tested bacterial
strains on such a dummy tissue is suppressed.
Moreover, GA-stabilised bovine pericardium post-treated with chi-
tosan fromsolutions in carbonic acid demonstrates even more advanced
and highly promising properties. The stressstrain behaviour of the ma-
trix is improved to a higher degree as compared to the dummy samples
apparently due to a collapse of the pores (voids, cavities) at high pres-
sure and sticking together of their wall with the chitosan coating.
Together with masking of residual aldehyde groups with chitosan that
result in extremely strong reduction of calcication (two orders of mag-
nitude reduction in subcutaneous tests on rats). There is also adsorbed
chitosan lm at the outer surface of the matrix with the thickness of
tens of nanometres. As a result, suchsamples alsodemonstrate complete
biocompatibility combining it with most evident antimicrobial activity.
The treatment with chitosan solutions in carbonic acids effectively pre-
vents biolm formation at the collagen surface.
Acknowledgments
The authors are very grateful to Alexey S. Kashin (N.D. Zelinsky
Institute of Organic Chemistry RAS), Dr. Sergei S. Abramchuk (A.N.
Nesmeyanov Institute of Organoelement Compounds) and Dr. Inesa V.
Blagodatskikh (A.N. Nesmeyanov Institute of Organoelement Com-
pounds) for their help with FE-SEM, TEM and GPC experiments and to
Prof. Boris V. Lokshin for fruitful discussions on ATR-IR. This work was
supported by the Russian Foundation for Basic Research (projects No
13-03-00378-a and No 12-03-31863-mol_a).
References
[1] L.H. Edmunds, Thrombotic and bleeding complications of prosthetic heart valves,
Ann. Thorac. Surg. 44 (1987) 430445.
[2] S.C. Cannegieter, F.R. Rosendaal, E. Brit, Thromboembolic and bleeding complica-
tions in patients with mechanical heart valve prostheses, Circulation 89 (1994)
635641.
[3] P.D. Stein, J.S. Alpert, H.I. Bussey, J.E. Dalen, A.G.G. Turpie, Antithrombotic therapy
in patients with mechanical and biological prosthetic heart valves, Chest 119
(2001) 220S227S.
[4] Sh.H. Rahimtoola, Choice of prosthetic heart valve for adult patients, J. Am. Coll.
Cardiol. 41 (2003) 893904.
[5] D.S. Bach, Choice of prosthetic heart valves: update for the next generation, J. Am.
Coll. Cardiol. 42 (2003) 17171719.
[6] R.F. Siddiqui, J.R. Abraham, J. Butany, Bioprosthetic heart valves: modes of failure,
Histopathology 55 (2009) 135144.
[7] A. Carpentier, From valvular xenograft to valvular bioprosthesis: 19651970, Ann.
Thorac. Surg. 48 (1989) S73S74.
[8] A. Starr, C.L. Fessler, G. Grunkemeier, G.-W. He, Heart valve replacement surgery:
past, present and future, Clin. Exp. Pharmacol. Physiol. 29 (2002) 735738.
[9] F.J. Schoen, J.J. Collins, L.H. Cohn, Long-term failure rate and morphologic correla-
tions in porcine bioprosthetic heart valves, Am. J. Cardiol. 51 (1983) 957964.
[10] P. Sinha, D. Zurakowski, T.K.S. Kumar, D. He, Ch. Rossi, R.A. Jonas, Effects of
glutaraldehyde concentration, pretreatment time, and type of tissue (porcine
versus bovine) on postimplantation calcication, J. Thorac. Cardiovasc. Surg. 143
(2012) 224227.
[11] F.J. Schoen, J.W. Tsao, R.J. Levy, Calcication of bovine pericardium used in cardiac
valve bioprostheses, Am. J. Pathol. 123 (1986) 134145.
[12] R.J. Levy, F.J. Schoen, F.S. Sherman, J. Nichols, M.A. Hawley, S.A. Lund, Calcication
of subcutaneously implanted Type I collagen sponges, Am. J. Pathol. 122 (1986)
7182.
[13] G.M. Bernacca, W.R. Dimitri, A.C. Fisher, T.G. Mackay, D.J. Wheatley, Chemical
modication of bovine pericardium and its effect on calcication in the rat subder-
mal model, Biomaterials 13 (1992) 345352.
[14] D. Mavrilas, A. Apostolaki, J. Kapolos, P.G. Koutsoukos, M. Melachrinou, V. Zolota, D.
Dougenis, Development of bioprosthetic heart valve calcication in vitro and in
animal models: morphology andcomposition, J. Cryst. Growth205(1999) 554562.
[15] J.P. Gott, Pan-Chih, L.M.A. Dorsey, J.L. Jay, G.K. Jett, F.J. Schoen, J.-M. Girardot,
R.A. Guyton, Calcication of porcine valves: a successful new method of
antimineralization, Ann. Thorac. Surg. 53 (1992) 207216.
[16] W. Chen, F.J. Schoen, R.J. Levy, Mechanism of efcacy of 2-amino oleic acid for in-
hibition of calcication of glutaraldehyde-pretreated porcine bioprosthetic heart
valves, Circulation 90 (1994) 323329.
[17] P. Zilla, D. Bezuidenhout, P. Human, Carbodiimide treatment dramatically potenti-
ates the anticalcic effect of alpha-amino oleic acid on glutaraldehyde-xed aortic
wall tissue, Ann. Thorac. Surg. 79 (2005) 905910.
[18] K.S. Jee, Y.S. Kim, K.D. Park, Y.H. Kim, Anovel chemical modication of bioprosthetic
tissues using L-arginine, Biomaterials 24 (2003) 34093416.
[19] C.L. Webb, J.J. Benedict, F.J. Schoen, J.A. Linden, R.J. Levy, Inhibition of bioprosthetic
heart valve calcication with aminodiphosphonate covalently bound to residual
aldehyde groups, Ann. Thorac. Surg. 46 (1988) 309316.
[20] R.J. Levy, J. Wolfrum, F.J. Schoen, M.A. Hawley, S.A. Lund, R. Langer, Inhibition of cal-
cication of bioprosthetic heart valves by local controlled-release diphosphonate,
Science 228 (1985) 190192.
[21] A.E. Trantina-Yates, P. Human, M. Bracher, P. Zilla, Mitigation of bioprosthetic heart
valve degeneration through biocompatibility: in vitro versus spontaneous endo-
thelialization, Biomaterials 22 (2001) 18371846.
[22] H. Gulbins, A. Goldemund, I. Anderson, U. Haas, A. Uhlig, B. Meiser, B. Reichart,
Preseeding with autologous broblasts improves endothelialization of
glutaraldehyde-xed porcine aortic valves, J. Thorac. Cardiovasc. Surg. 125
(2003) 592601.
[23] A. Abolhoda, S. Yu, J.R. Oyarzun, J.R. McCormick, J.D. Bogden, Sh. Gabbay, Calcica-
tion of bovine pericardium: glutaraldehyde versus no-react biomodication, Ann.
Thorac. Surg. 62 (1996) 169174.
[24] K.D. Park, W.K. Lee, J.Y. Yun, D.K. Han, S.H. Kim, Y.H. Kim, H.M. Kim, K.T. Kim,
Novel anti-calcication treatment of biological tissues by grafting of sulphonated
poly(ethylene oxide), Biomaterials 18 (1997) 4751.
[25] Sh. Aravind, W. Paul, S.C. Vasudev, Ch.P. Sharma, Polyethylene glycol (PEG)
modied bovine pericardiuim as a biomaterial: a comparative study on immuno-
genicity, J. Biomater. Appl. 13 (1998) 158165.
[26] S.C. Vasudev, Th. Chandy, Ch.P. Sharma, The antithrombotic versus calcium
antagonistic effects of polyethylene glycol grafted bovine pericardium, J. Biomater.
Appl. 14 (1999) 4866.
[27] M. Dahm, J. Rhe, B. Berchthold, D. Prfer, O. Prucker, B.-J. Chang, A. Wallrath, H.
Oelert, Ultrathinpolymer monolayers for promotionof cell growthonbioprosthetic
materials evolution of a new concept to improve long term performance of bio-
logic heart valves, Biomed. Mater. Eng. 14 (2004) 419425.
[28] A. Oosthuysen, P.P. Zilla, P.A. Human, Ch.A.P. Schmidt, D. Bezuidenhout, Bioprosthetic
tissue preservation by lling with a poly(acrylamide) hydrogel, Biomaterials 27
(2006) 21232130.
[29] R. Ohri, S.K. Hahn, A.S. Hoffman, P.S. Stayton, C.M. Giachelli, Hyaluronic acid
grafting mitigates calcication of glutaraldehyde-xed bovine pericardium,
J. Biomed. Mater. Res. 70A (2004) 328334.
[30] W.K. Lee, K.D. Park, D.K. Han, H. Suh, J.-Ch. Park, Y.H. Kim, Heparinized bovine peri-
cardium as a novel cardiovascular bioprosthesis, Biomaterials 21 (2000)
23232330.
[31] K.S. Jee, J.W. Bae, J.S. Shin, D.K. Han, K.D. Park, Heparin immobilized small intestinal
submucosa for cardiovascular applications, Macromol. Symp. 249250 (2007)
120123.
[32] J. Chanda, Anticalcication treatment of pericardial prostheses, Biomaterials 15
(1994) 465469.
[33] J. Chanda, S.B. Rao, M. Mohanty, A.V. Lal, C.V. Muraleedharan, G.S. Bhuvaneshwar,
M.S. Valiathan, Prevention of calcication of tissue valves, Artif. Organs 18 (1994)
752757.
[34] J. Chanda, Prevention of calcication of heart valve bioprostheses: an experimental
study in rat, Ann. Thorac. Surg. 60 (1995) S339S342.
[35] J. Chanda, R. Kuribayashi, T. Abe, Use of the glutaraldehyde-chitosan-treated por-
cine pericardium as a pericardial substitute, Biomaterials 17 (1996) 10871091.
[36] R. Kuribayashi, J. Chanda, T. Abe, Efcacy of the chitosan posttreatment in calcica-
tion prevention of the glutaraldehyde-treated porcine aortic noncoronary cusp im-
planted in the right ventricular outow tract in dogs, Artif. Organs 20 (1996)
761766.
[37] J. Chanda, R. Kuribayashi, T. Abe, Prevention of calcication in glutaraldehyde-
treated porcine aortic and pulmonary valves, Ann. Thorac. Surg. 64 (1997)
10631066.
[38] J. Chanda, R. Kuribayashi, T. Abe, Heparin in calcication prevention of porcine
pericardial bioprostheses, Biomaterials 18 (1997) 11091113.
[39] T. Sannan, K. Kurita, Y. Iwakura, Studies on chitin, 2) effect of deacetylation on
solubility, Makromol. Chem. 177 (1976) 35893600.
[40] S.-i. Aiba, Studies on chitosan: 3. Evidence for the presence of random and block
copolymer structures in partially N-acetylated chitosans, Int. J. Biol. Macromol.
13 (1991) 4044.
[41] Y.W. Cho, Y.N. Cho, S.H. Chung, G. Yoo, S.W. Ko, Water-soluble chitin as a wound
healing accelerator, Biomaterials 20 (1999) 21392145.
[42] M. Rinaudo, Chitin and chitosan: properties and applications, Prog. Polym. Sci. 31
(2006) 603632.
[43] C. Shanthi, K.P. Rao, Chitosan modied poly(glycidyl methacrylatebutyl acrylate)
copolymer grafted bovine pericardial tissueanticalcicationproperties, Carbohydr.
Polym. 44 (2001) 123131.
[44] G.M. Nogueira, A.C.D. Rodas, R.F. Weska, C.G. Aimoli, O.Z. Higa, M. Maizato, A.A.
Leiner, R.N.M. Pitombo, B. Polakiewicz, M.M. Beppu, Bovine pericardium coated
with biopolymeric lms as an alternative to prevent calcication: in vitro calcica-
tion and cytotoxicity results, Mater. Sci. Eng. C 30 (2010) 575582.
[45] A.C.D. Rodas, R. Polak, P.H. Hara, E.I. Lee, R.N.M. Pitombo, O.Z. Higa, Cytotoxicity and
endothelial cell adhesion of lyophilized and irradiated bovine pericardium modi-
ed with silk broin and chitosan, Artif. Organs 35 (2011) 502507.
[46] R. Polak, A.C.D. Rodas, D.L. Chicoma, R. Giudici, M.M. Beppu, O.Z. Higa, R.N.M.
Pitombo, Inhibition of calcication of bovine pericardium after treatment with
biopolymers, E-beam irradiation and in vitro endothelization, Mater. Sci. Eng. C
33 (2013) 8590.
[47] F.S. Steven, The Nishihara technique for the solubilization of collagen application
to the preparation of soluble collagens from normal and rheumatoid connective
tissue, Ann. Rheum. Dis. 23 (1964) 300301.
[48] G.C. Na, L.J. Butz, R.J. Carroll, Mechanism of in vitro collagen bril assembly, J. Biol.
Chem. 261 (1986) 1229012299.
[49] M.-M. Giraud-Guille, Liquid crystalline order of biopolymers in cuticles and bones,
Microsc. Res. Techn. 27 (1994) 420428.
[50] B. Przybilla, J. Ring, Anaphylaxis to ethanol and sensitisation to acetic acid, Lancet
321 (1983) 483.
[51] M. Sticherling, J. Brasch, H. Brning, E. Christophers, Urticarial and anaphylactoid
reactions following ethanol intake, Br. J. Dermatol. 132 (1995) 464467.
[52] W.H. Boehncke, H. Gall, Ethanol metabolite acetic acid as causative agent for type-I
hypersensitivity-like reactions to alcoholic beverages, Clin. Exp. Allergy 26 (1996)
10891091.
[53] M. Sticherling, J. Brasch, Alcohol: intolerance syndromes, urticarial and anaphylac-
toid reactions, Clin. Dermatol. 17 (1999) 417422.
[54] Y. Nakagawa, Y. Sumikawa, T. Nakamura, S. Itami, I. Katayama, T. Aoki, Urticarial
reaction caused by ethanol, Allergol. Int. 55 (2006) 411414.
[55] B. Wthrich, Allergic and intolerance reactions to wine, Allergologie 34 (2011)
427436.
[56] I.S. Chaschin, T.E. Grigorev, M.O. Gallyamov, A.R. Khokhlov, Direct deposition of
chitosan macromolecules on a substrate from solutions in supercritical carbon
dioxide: solubility and conformational analysis, Eur. Polym. J. 48 (2012) 906918.
[57] M.A. Khokhlova, M.O. Gallyamov, A.R. Khokhlov, Chitosan nanostructures depos-
ited from solutions in carbonic acid on a model substrate as resolved by AFM,
Colloid Polym. Sci. 290 (2012) 14711480.
[58] Y. Sakai, K. Hayano, H. Yoshioko, T. Fujieda, K. Saito, H. Yoshioka, A novel method
of dissolving chitosan in water for industrial applications, Polymer J. 33 (2002)
640642.
[59] R.A.A. Muzzarelli, Chitins and chitosans for the repair of wounded skin, nerve,
cartilage and bone, Carbohydr. Polym. 76 (2009) 167182.
[60] J.W. Costerton, Z. Lewandowski, D.E. Caldwell, D.R. Korber, H.M. Lappin-Scott,
Microbial biolms, Annu. Rev. Microbiol. 49 (1995) 711745.
[61] W.M. Dunne, Bacterial adhesion: seen any good biolms lately? Clin. Microbiol.
Rev. 15 (2002) 155166.
[62] D. Pavithra, M. Doble, Biolm formation, bacterial adhesion and host response on
polymeric implantsissues and prevention, Biomed. Mater. 3 (2008) 034003.
[63] A. Gamzazade, M. Chlenov, A. Kriger, E. Frontchek, The study of the relationship be-
tween intrinsic viscosity and molecular weight of chitosan and its carboxymethyl
ether, in: H. Struszczyk, H. Pospieszny, A. Gamzazade (Eds.), Chitin and Chitosan,
Polish Chitin Society, Lodz, Poland, 1999, pp. 5158.
[64] G.A. Badun, M.G. Chernysheva, A.L. Ksenofontov, Increase in the specic radioactiv-
ity of tritium-labeled compounds obtained by tritium thermal activation method,
Radiochim. Acta 100 (2012) 401408.
[65] A.S. Kashin, V.P. Ananikov, A SEM study of nanosized metal lms and metal
nanoparticles obtained by magnetron sputtering, Russ. Chem. Bull. 60 (2011)
26022607.
[66] J. McFarland, An instrument for estimating the number of bacteria in suspensions
usedfor calculating the opsonic indexandfor vaccines, JAMA49 (1907) 11761178.
[67] H. Teng, A. Yamasaki, M.-K. Chun, H. Lee, Solubility of liquid CO
2
in water at
temperatures from 278 K to 293 K and pressures from 6.44 MPa to 29.49 MPa
and densities of the corresponding aqueous solutions, J. Chem. Thermodyn. 29
(1997) 13011310.
[68] L.W. Diamond, N.N. Akinev, Solubility of CO
2
in water from 1.5 to 100 C and
from 0.1 to 100 MPa: evaluation of literature data and thermodynamic modelling,
Fluid Phase Equilib. 208 (2003) 265290.
[69] N. Spycher, K. Pruess, J. Ennis-King, CO
2
H
2
O mixtures in the geological sequestra-
tion of CO
2
. I. Assessment and calculation of mutual solubilities from 12 to 100 C
and up to 600 bar, Geochim. Cosmochim. Acta 67 (2003) 30153031.
[70] B. Meyssami, M.O. Balaban, A.A. Teixeira, Prediction of pH in model systems pres-
surized with carbon-dioxide, Biotechnol. Prog. 8 (1992) 149154.
[71] K.L. Toews, R.M. Shroll, C.M. Wai, N.G. Smart, pH-dening equilibrium between
water and supercritical CO
2
. Inuence on SFE of organics and metal-chelates,
Anal. Chem. 67 (1995) 40404043.
[72] M.H. Santos, R.M. Silva, V.C. Dumont, J.S. Neves, H.S. Mansur, L.G.D. Heneine, Ex-
traction and characterization of highly puried collagen from bovine pericardium
for potential bioengineering applications, Mater. Sci. Eng. C 33 (2013) 790800.
[73] A. Faikrua, R. Jeenapongsa, M. Sila-asna, J. Viyoch, Properties of -glycerol
phosphate/collagen/chitosan blend scaffolds for application in skin tissue engi-
neering, Sci. Asia 35 (2009) 247254.
[74] W. Sun, M. Sacks, G. Fulchiero, J. Lovekamp, N. Vyavahare, M. Scott, Response of
heterograft heart valve biomaterials to moderate cyclic loading, J. Biomed. Mater.
Res. 69A (2004) 658669.
[75] S. Noppakundilograt, K. Sonjaipanich, N. Thongchul, S. Kiatkamjornwong, Syntheses,
characterization, and antibacterial activity of chitosan grafted hydrogels and associ-
ated mica-containing nanocomposite hydrogels, J. Appl. Polym. Sci. 127 (2013)
49274938.
[76] F.G. Pearson, R.H. Marchessault, C.Y. Liang, Infrared spectra of crystalline polysac-
charides. V. Chitin, J. Polym. Sci. 43 (1960) 101116.
[77] D.J.S. Hulmes, Building collagen molecules, brils, and suprabrillar structures,
J. Struct. Biol. 137 (2002) 210.
[78] D. Wei, W. Qian, Facile synthesis of Ag and Au nanoparticles utilizing chitosan as a
mediator agent, Colloids Surf. B: Biointerfaces 62 (2008) 136142.
[79] V.J. Ferrans, T.L. Spray, M.E. Billingham, W.C. Roberts, Structural changes in
glutaraldehyde-treated porcine heterografts used as substitute cardiac valves,
Am. J. Cardiol. 41 (1979) 11591184.
[80] B. Mendoza-Novelo, D.I. Alvarado-Castro, J.L. Mata-Mata, J.V. Cauich-Rodrguez, A.
Vega-Gonzlez, E. Jorge-Herrero, F.J. Rojo, G.V. Guinea, Stability and mechanical
evaluation of bovine pericardium cross-linked with polyurethane prepolymer in
aqueous medium, Mater. Sci. Eng. C 33 (2013) 23922398.
[81] D.A.D. Parry, The molecular and brillar structure of collagen and its relationship
to the mechanical properties of connective tissue, Biophys. Chem. 29 (1988)
195209.
[82] G. Arcidiacono, A. Corvi, T. Severi, Functional analysis of bioprosthetic heart valves,
J. Biomech. 38 (2005) 14831490.
[83] J.M.G. Pez, E. Jorge-Herrero, Assessment of pericardium in cardiac bioprostheses:
a review, J. Biomater. Appl. 13 (1999) 351388.
[84] D.M. Snchez, D.M. Gaitn, A.F. Len, J. Mugnier, J.C. Briceo, Fixation of vascular
grafts with increased glutaraldehyde concentration enhances mechanical proper-
ties without increasing calcication, ASAIO J. 53 (2007) 257262.
[85] M.C. Vinci, G. Tessitore, L. Castiglioni, F. Prandi, M. Soncini, R. Santoro, F. Consolo, F.
Colazzo, B. Micheli, L. Sironi, G. Polvani, M. Pesce, Mechanical compliance and im-
munological compatibility of xative-free decellularized/cryopreserved human
pericardium, PLoS ONE 8 (2013) e64769.
[86] L. A. Bokerija, N. P. Bakuleva, V. T. Kostava, A. I. Gamzazade, M. O. Gallyamov,
A. R. Khokhlov, (Bakulev Scientic Center for Cardiovascular Surgery, RAMS),
Method for chemical treatment of xenopericardium, RU Patent No. 2 384 348,
2010.
[87] A.I. Gamzazade, N.P. Bakuleva, E.M. Belavtseva, M.O. Gallyamov, Electron microsco-
py of the coating morphology of pericardum tissue with chitosan ionogen deriva-
tives, Bull. Russ. Acad. Sci. Phys. 73 (2009) 468470.
[88] E. Gendler, S. Gendler, M.E. Nimni, Toxic reactions evoked by glutaraldehyde-xed
pericardiumand cardiac valve tissue bioprosthesis, J. Biomed. Mater. Res. 18 (1984)
727736.
[89] I.G. Lyutova, N.M. Anuchina, N.P. Bakuleva, V.T. Kostava, I.S. Chaschin, Antimi-
crobial properties of bioprostheses coated with nanostructured low-molecular-
weight chitosan, Bull. Bakulev Sci. Cent. Cardiovasc. Surg. 14 (2013) 5257
(in Russian).
[90] M.R. Bet, G. Goissis, S. Vargas, H.S. Selistre-de-Araujo, Cell adhesion and cytotoxic-
ity studies over polyanionic collagen surfaces with variable negative charge and
wettability, Biomaterials 24 (2003) 131137.
[91] G. Goissis, A. de F. Giglioti, D.M. Braile, Preparation and characterization of an
acellular bovine pericardium intended for manufacture of valve bioprostheses,
Artif. Organs 35 (2011) 484489.
[92] A. Terada, A. Yuasa, T. Kushimoto, S. Tsuneda, A. Katakai, M. Tamada, Bacterial
adhesion to and viability on positively charged polymer surfaces, Microbiology
152 (2006) 35753583.
[93] A.B. Vishu Kumar, M.C. Varadaraj, L.R. Gowda, R.N. Tharanathan, Characterization
of chito-oligosaccharides prepared by chitosanolysis with the aid of papain and
Pronase, and their bactericidal action against Bacillus cereus and Escherichia coli,
Biochem. J. 391 (2005) 167175.
[94] E.I. Rabea, M.E.-T. Badawy, Ch.V. Stevens, G. Smagghe, W. Steurbaut, Chitosan
as antimicrobial agent: applications and mode of action, Biomacromolecules 4
(2003) 14571465.
[95] M.S. Benhabiles, R. Salah, H. Lounici, N. Drouiche, M.F.A. Goosen, N. Mameri, Anti-
bacterial activity of chitin, chitosan and its oligomers prepared from shrimp shell
waste, Food Hydrocoll. 29 (2012) 4856.
[96] C. Mohamed, K.A. Clementine, M. Didier, L. Grard, D.C.M. Nolle, Antimicrobial
and physical properties of edible chitosan lms enhanced by lactoperoxidase
system, Food Hydrocoll. 30 (2013) 576580.
[97] S. Yorsaeng, O. Pornsunthorntawee, R. Rujiravanit, Preparationand characterization
of chitosan-coated DBDplasma-treatednatural rubber latexmedical surgical gloves
with antibacterial activities, Plasma Chem. Plasma Process. 32 (2012) 12751298.
[98] E.M. Costa, S. Silva, F.K. Tavaria, M.M. Pintado, Study of the effects of chitosan upon
Streptococcus mutans adherence and biolm formation, Anaerobe 20 (2013) 2731.
[99] D. Raafat, H.-G. Sahl, Chitosan and its antimicrobial potential a critical literature
survey, Microb. Biotechnol. 2 (2009) 186201.
[100] J. Huang, H. Jiang, M. Qiu, X. Geng, R. Yang, J. Li, Ch. Zhang, Antibacterial activity
evaluation of quaternary chitin against Escherichia coli and Staphylococcus aureus,
Int. J. Biol. Macromol. 52 (2013) 8591.
[101] L.Ch. Chen, W.D. Chiang, W.Ch. Chen, H.H. Chen, Y.W. Huang, W.J. Chen, Sh.B. Lin,
Inuence of alanine uptake on Staphylococcus aureus surface charge and its
susceptibility to two cationic antibacterial agents, nisin and low molecular weight
chitosan, Food Chem. 135 (2012) 23972403.
[102] J.R. Rodrguez-Nez, J. Lpez-Cervantes, D.I. Snchez-Machado, B. Ramrez-Wong,
P. Torres-Chavez, M.O. Cortez-Rocha, Antimicrobial activity of chitosan-based lms
against Salmonella typhimuriumand Staphylococcus aureus, Int. J. Food Sci. Technol.
47 (2012) 21272133.
[103] G. Geisberger, E.B. Gyenge, D. Hinger, A. Kch, C. Maake, G.R. Patzke, Chitosan-
thioglycolic acid as a versatile antimicrobial agent, Biomacromolecules 14 (2013)
10101017.
[104] D. Raafat, K. von Bargen, A. Haas, H.-G. Sahl, Insights into the mode of action of chi-
tosan as an antibacterial compound, Appl. Environ. Microbiol. 4 (2008) 37643773.

1 s2.0 S0928493114000186 Main

Hochgeladen von

Dokumentinformationen

Originalbeschreibung:

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

1 s2.0 S0928493114000186 Main

Hochgeladen von

Copyright:

Verfügbare Formate

Collagen tissue treated with chitosan solutions in carbonic acid for

improved biological prosthetic heart valves

Das könnte Ihnen auch gefallen