Photocrosslinkable chitosan as a biological adhesive
Katsuaki Ono,1Yoshio Saito,2Hirohumi Yura,2Keiichi Ishikawa,3Akira Kurita,4Toshihiro Akaike,5 Masayuki Ishihara4
1National Defense Medical College,Department of Surgery II,3-2Namiki,Tokorozawa,Saitama,359-8513Japan
2NeTech Inc.,KSP East Wing,502Sakado3-2-1,Takatsu,Kawasaki,Kanagawa,213-0012Japan
3National Defense Medical College,Department of Surgery I,3-2Namiki,Tokorozawa,Saitama,359-8513Japan
4National Defense Medical College,Research Institute,Division of Biomedical Engineering,3-2Namiki,Tokorozawa, Saitama,359-8513Japan
5Tokyo Institute of Technology,Department of Biomolecular Engineering,Faculty of Bioscience and Biotechnology,4259 Nagatsuta,Midori,Yokohama,226Japan
Received19April1999;revised24June1999;accepted16July1999
Abstract:A photocrosslinkable chitosan to which both azide and lactose moieties were introduced(Az-CH-LA) was prepared as a biological adhesive for soft tissues and its effectiveness was compared with that of fibrin glue.Intro-duction of the lactose moieties resulted in a much more water-soluble chitosan at neutral pH.Application of ultra-violet light(UV)irradiation to photocrosslinkable Az-CH-LA produced an insoluble hydrogel within60s.This hydro-gel firmly adhered two pieces of sliced ham with each other, depending upon the Az-CH-LA concentration.The binding strength of the chitosan hydrogel prepared from30–50mg/ mL of Az-CH-LA was similar to that of fibrin glue.Com-pared to the fibrin glue,the chitosan hydrogel more effec-tively sealed air leakage from pinholes on isolated small intestine and aorta and from incisions on isolated trachea. Neither Az-CH-LA nor its hydrogel showed any cytotoxicity in cell culture tests of human skin fibroblasts,coronary en-dothelial cells,and smooth muscle cells.Furthermore,all mice studied survived for at least1month after implantation of200␮L of photocrosslinked chitosan gel and intraperito-neal administration of up to1mL of30mg/mL of Az-CH-LA solution.These results suggest that the photocrosslink-able chitosan developed here has the potential of serving as a new tissue adhesive in medical use.©2000John Wiley& Sons,Inc.J Biomed Mater Res,49,289–295,2000.
Key words:chitosan;chitin;biological adhesive;photo-crosslink
INTRODUCTION
Biological adhesives are used for tissue adhesion, hemostasis,and sealing of the leakage of air and body fluids during surgical procedures.Various studies have been carried out,and chemically crosslinkable gelatin1–3and cyanoacrylate polymer4,5have been de-veloped as such biological adhesives.However,resor-cinol,formaldehyde,or carbodiimides generated through the crosslink reaction of gelatin,as well as formaldehyde produced by degradation of cyanoacry-late,have been found to be toxic.Commercially avail-able fibrin glue,which contains fibrinogen,thrombin, factor XIII,and protease inhibitor,utilizes the blood coagulation system for sealing tissue and currently is used in clinics.6–9However,despite its useful proper-ties of high adhesiveness and rapid gelation,fibrin glue has a disadvantage in its industrial production. Furthermore,when using these biological materials it is difficult to prevent infectious contaminations. Polysaccharides,such as chitosan,having a hydro-gel-forming property are considered to be more ad-vantageous in their application as adhesive materials. Chitin is a linear homopolymer of1,4␤-linked N-acetyl-D-glucosamine,and chitosan is a partially N-deacetylated chitin.Chitosan already has been pro-posed as a biomedical material.10,11It has been ob-served to accelerate wound healing12,13and has a hemostatic potential.14,15In addition,chitosan is bio-degradable and nontoxic.16Chitosan also is known for its immunologic activity,that is,its macrophage acti-vation,17its cytokine production,18and its inhibition of infection.19
We designed a new photocrosslinkable chitosan molecule that contains both lactose moieties and pho-toreactive azide groups and may be used as a biologi-cal adhesive for medical purposes.The present study was undertaken to assess the capacity of hydrogel for-
Correspondence to:M.Ishihara
©2000John Wiley&Sons,Inc.CCC0021-9304/00/020289-07
mation by photocrosslinking of the chitosan molecules to provide a biological adhesive and its safety in medi-cal uses.The adhesive property of the hydrogel was evaluated by measuring the binding strength between two ham slices,and we tested its sealing strength rela-tive to the prevention of air leakages from a tracheal incision and from pinholes of isolated pig intestine and aorta by puncturing them with a small needle.In addition,the toxicity of the hydrogel also was studied in vitro and in vivo.
MATERIALS AND METHODS Preparation of photocrosslinkable chitosan
Figure1shows the chemical structure of photocrosslink-able chitosan derivative(Az-CH-LA).Both azide and lactose moieties were introduced into the chitosan molecule through a two-step condensation reaction.The chitosan used had a molecular weight of800–1,000kDa,with an80% degree of deacetylati
on(Yaizu Suisankagaku Industry Co., Ltd.,Shizuoka,Japan).For the first condensation reaction, chitosan(125g)was added to3L(pH=4.75)of50m M of TEMED(N,N,NЈ,NЈ-tetramethylethylenediamine;Wako Pure Chemical Industry,Ltd.,Japan)containing56.25mL of concentrated HCl.Then EDC[32.5g,1-ethyl-3-(3-dimethylaminopropyl)carbodiimide;Wako Pure Chemical Industry]and4-O-␤-D-galactopyranosyl-(1,4)-D-gluconic acid20(0.25g of lactobionic acid;Wako Pure Chemical In-dustry)were added.21This mixture was stirred at room tem-perature for24h,followed by an ultrafiltration membrane passing unreacted substances below10kDa(Diaflo Ultrafil-ters YM10,Amicon,Ontario,Canada).Finally,a powder of lactose-linked chitosan(CH-LA)was obtained by freeze-drying.It has been estimated by a conventional phenol-sulfuric acid colorimetric assay for carbohydrate that about 2%of amino groups in the chitosan molecule were replaced by lactobionic acid.
Subsequently the product(CH-LA,1g),EDC(0.35g),and 4-azidobenzoic acid(0.2g;Tokyo Chemical Industry Co., Ltd.,Japan)were added to100mL of50m M of the TEMED solution.The mixture was stirred at room temperature for 72h.Condensation and purification steps to obtain the pho-tocrosslinkable chitosan(Az-CH-LA)were carried out as de-scribed above.It has been estimated by specific absorbance (275nm)and NMR or FI-IR spectrum of azidobenzoic acid (data not shown)that about2.5%
of amino groups in the chitosan molecule were replaced by azidobenzoic acid. Measurement of water solubility and gelation time
Chitosan,CH-LA,40%lactose-introduced chitosan,and Az-CH-LA were dissolved separately in water(0.7and1 mg/mL)and the pH adjusted to3with0.1N of HCl.At pH 3,chitosan and its derivatives completely dissolved in water. The pH of each solution gradually was raised by addition of 0.1N of NaOH,and the pH at which precipitation occurred (insoluble chitosan derivatives)was measured using a pH meter.The concentrations of the chitosan derivatives did not decrease below70%of the original concentration(1mg/mL) upon addition of0.1N of NaOH.
Viscous Az-CH-LA aqueous solutions were gelatinized by UV irradiation.Azide groups(−N3)are known to release N2 upon UV irradiation and to be converted into highly reactive nitrene groups.Nitrene groups are supposed to interact quickly with each other or with amino groups of the chito-san to generate azo groups(−N=N−),thus causing gelation. To measure gelation times,600␮L of30mg/mL of Az-CH-LA aqueous solution were put into tissue culture dishes (each35×10mm,Falcon).The dishes were irradiated with UV light(at a distance of2cm;4W,254nm UV-tube light; Vilber Lourmat,France)for various time periods to obtain insoluble hydrogels.Formed hydrogels were rinsed thor-oughly with water to remove any water-soluble materials and subsequently dried with air blowing.The weight of the
remaining dry gels in the dishes then was determined. Measurement of binding strength
Two square slices of ham(each2×5cm and2mm thick) were placed flatly side-by-side with the2-cm side of each closely touching.Chitosan gel prepared from10,30,or50 mg/mL of Az-CH-LA aqueous solution then was applied to the boundary between the two slices over the2-cm side (with a thickness of2mm)to adhere the slices.After UV irradiation for90s,the adhered ham preparations were hung on a hook with a thread attached to the unglued2-cm side of one ham slice,and subsequently they were pulled by hanging various weights on the lower2-cm side.When the ham separated into the original two slices at a certain weight,the cross-sectional area of the adhered chitosan gel (perpendicular to the larger surface of ham)was measured. The binding strength was calculated using the following for-
mula:binding strength=A×B,where A is the cross-sectional area(cm2)of residual chitosan gel and B is the weight(g)at which the ham preparations separated into two slices.The binding strength of fibrin glue(Beriplast P, HoechstиMarionиRoussel)also was measured in a similar way after the equal volume mixture of thrombin and fibrino-gen solutions was applied for5min to bind both ham slices. Measurement of sealing strength
The small intestine,trachea,and thoracic aorta(10mm in diameter)were removed from farm pigs and th
e removed tissues immersed in phosphate-buffered saline.Experiments were performed within5h after removal of those tissues. Suturing around a small catheter connected to a syringe ligated one end of each of these tissue tubes,and a mercurial sphygmomanometer(Okose Inc.,Japan)was inserted into the other end.A pinhole in the small intestine or in the aorta was made by needle puncture(1.2mm in diameter),and an incision(5mm in length)was made in the trachea using an operation knife.One drop(about30␮L)of10,15,20,or30 mg/mL of Az-CH-LA aqueous solution then was applied to the pinhole or incision.After UV irradiation for90s,each tissue tube was placed underwater and air infused into the tissue tube from a syringe until bubbles were observed due to leaking from the pinhole or incision.The pressure re-quired to burst the seal of the pinhole or incision was mea-sured using the sphygmomanometer to obtain the sealing strength.The sealing strength of the fibrin glue also was measured in a similar way after application of a fibrinogen solution followed by rubbing a thrombin solution over the pinhole or the incision for5min.
In vitro and in vivo tests for the cytotoxicity of
Az-CH-LA solutions and chitosan hydrogels
Human skin fibroblasts,coronary smooth muscle cells, and coronary endothelial cells(from4th to8th pa
ssages) were purchased from Takara Biomedical(Tokyo,Japan). The fibroblasts and smooth muscle cells were grown in DMEM(Life Technologies Oriental Inc.,Tokyo,Japan) supplemented with10%heat-inactivated fetal bovine serum (FBS)and antibiotics(100U/mL of penicillin G and100␮g/mL of streptomycin)in an atmosphere of5%CO2in air and100%relative humidity.The cells routinely were sub-cultured at a split ratio of1/10.The coronary endothelial cells were grown in medium199(Life Technologies Oriental Inc.)supplemented with10%heat-inactivated FBS,antibiot-ics(100U/mL of penicillin G and100␮g/mL of streptomy-cin),and10ng/mL of FGF-2(R&D Systems,Minneapolis, Minnesota).
For in vitro testing of the cytotoxicity of Az-CH-LA,the cells were plated at an initial density of5,000cells/well in 100␮L of culture medium in each well of a96-well tissue culture plate(Falcon).After5h of culturing,implying that cells were adhered to the plate,100␮L of culture media containing various concentrations of Az-CH-LA were added,after which cells were grown for another2days. After the incubation,the used medium was removed and 100␮L of freshly prepared medium,including10␮L of WST-1reagent(Cell Counting Kit,Dojindo,Kumamoto,Ja-pan),were added to each well.Subsequently,after1h of incubation the optical densities were read at450nm using the Immuno Mini plate reader(Nunc InterMed,Tokyo,Ja-pan).
For in vitro testing of the toxicity of the chitosan hydrogel, 100␮L of30,20,10,or0mg/mL of Az-CH-LA aqueous solution were spotted on each well of a12-well tissue culture plate(Falcon)and converted into an insoluble hydrogel by 90s of UV irradiation.Subsequently cells were plated with an initial density of50,000cells/well(each well containing1 mL of the medium in the12-well tissue culture plate)and grown for2days.After the incubation,the used medium was removed and500␮L of freshly prepared medium in-cluding50␮L of WST-1reagent were added to each well. After1h of incubation,using a UV-VIS spectrophotometer (UV-1200;Shimazu,Japan),the optical density was read at 450nm.
In order to determine the in vivo toxicity of Az-CH-LA aqueous solutions,0.2,0.5,and1mL of30mg/mL of Az-CH-LA aqueous solution were intraperitoneally injected into each of six etherized mice(C57BL/6,CLEA Japan Inc., Tokyo,Japan).Furthermore,photocrosslinked chitosan gels were implanted subcutaneously on the backs of the six mice. Briefly,the mice were anesthetized with an intraperitoneal injection of pentobarbital(12mg/kg,Dainippon Pharma-ceutical Co.,Ltd.,Japan).Skin incisions(about1cm long) were made on the backs of the mice and the photo-crosslinked chitosan gels formed from200␮L of30mg/mL of Az-CH-LA aqueous solution were implanted subcutane-ously.The incisions then were closed with nylon sutures. The Az-CH-LA aqueous solutions were not sterilized or checked for contamination of endotoxin,but all surgical pro-cedures were done aseptically.The survival of the mice was determined on day30after the implantation.
RESULTS
Water-solubility of CH-LA and Az-CH-LA and gelation of Az-CH-LA
Native chitosan at1mg/mL was insoluble at pH4and higher.Chitosan to which2%lactobionate was introduced (CH-LA)exhibited a good aqueous solubility at neutral pH and lower while CH-LA with40%introduced lactobionate had a good aqueous solubility at pH values below13.The introduction into CH-LA of4-azidobenzoic acid by2.5% showed no additional change in water-solubility(Table I). When the same experiment was performed using0.7mg/ mL of aqueous solutions of these chitosan derivatives at pH 3,the same pH-dependent water solubilities were observed (data not shown).
Figure2shows that insoluble hydrogels were formed from30mg/mL of Az-CH-LA aqueous solutions,their yields depending on the time of UV irradiation.Upon UV irradiation,formation of insoluble hydrogels was completed
291
PHOTOCROSSLINKABLE CHITOSAN AS ADHESIVE
within60s.With increasing intensity of UV light,the time for converting into the insoluble hydrogels was f
urther shortened(<10s,data not shown).The insoluble chitosan gel retained a high water content.
Binding and sealing strength of the insoluble chitosan gel after irradiation
Figure3shows the binding strength of insoluble chitosan gels between two slices of ham.The binding strength in-creased with increasing Az-CH-LA concentrations up to50 mg/mL.The hydrogel obtained from50mg/mL of Az-CH-LA exhibited a higher strength(43g/cm2)than fibrin glue (40g/cm2).In these experiments,detachment of the chitosan gels from the ham surface has not been observed whereas the fibrin glue readily separated from the ham surfaces.As shown in Figures4,5,and6,chitosan hydrogels very effec-tively prevent air leakage from pinholes on isolated small intestine and aorta as well as from incisions on isolated tra-chea showing a concentration dependence of the Az-CH-LA used.The mean of the so-called bursting pressure(as a mea-sure of sealing strength)for the small intestine was61±1 mmHg(mean±SD),with the chitosan gel formed from30 mg/mL of Az-CH-LA solution,which was higher than that of the fibrin glue(51±6mmHg)(Fig.4).Similarly,the chitosan hydrogel could seal air leakage from the pinholes on the aorta and from the incision on the trachea up to mean bursting pressures of200±18mmHg and103±24cmH2O, respectively(Figs.5,6).
Table II summarizes the sealing strength of the chitosan gel formed from30mg/mL of Az-CH-LA solutio
n in com-parison with the fibrin glue at pinholes on the small intestine and the aorta and at the incisions of the trachea.The sealing strength of chitosan gels in the prevention of air leakage from the pinholes in or the incisions of these tubes was significantly higher than that of the fibrin glue.Thus chito-san hydrogels prepared from30mg/mL of Az-CH-LA were capable of sealing the air leakage from tissue preparations and in all cases were superior to the fibrin glue.
In vitro cytotoxicity testing of Az-CH-LA and its chitosan gel
It has been found to be difficult for cells to adhere to and grow on chitosan hydrogels having a high water content.22
In our experiments,human skin fibroblasts,coronary smooth muscle cells,and endothelial cells did not adhere to and grow on immobilized chitosan gels prepared by irra-diation of100␮L of30,20,15,and10mg/mL of Az-CH-LA aqueous solutions,probably due to their hyperhydrous structure(data not shown).However,these cells did grow normally beside the immobilized hydrogels(Fig.7).Further-more,the addition of400,200,100,50,and25␮g/mL of Az-CH-LA to the culture medium did not influence the ad-hesion and proliferation of these cells(data not shown). These results suggest that Az-CH-LA aqueous solutions and photocrosslinked hydrogels do not cause toxicity to cells. In vivo toxicity testing of Az-CH-LA and its
chitosan gel
Mice(n=6)with subcutaneously chitosan gels prepared from200␮L of Az-CH-LA aqueous solutions(30mg/mL) were all confirmed to be alive for at least1month(data not shown).No chitosan gel visibly could be detected at the site of implantation after1month.Furthermore,mice(n=3)intraperitoneally administered0.25,0.5,and1mL of the30 mg/mL Az-CH-LA aqueous solutions all survived at least1 month(data not shown).
DISCUSSION AND CONCLUSIONS Chitosan is a cationic polymer derived by deacety-lation of chitin.10,11Chitosan has been shown to accel-erate wound healing,12,13to have hemostatic14,15and antibacterial properties,19and to stimulate macro-phages.17Furthermore,chitosan is nontoxic,biocom-patible,and biodegradable.16Although many at-tempts have been made to develop chitosan/chitin biomaterials,only a few biomedical materials have been commercialized,such as Beschitin(Unichika Co., Ltd.,Kyoto,Japan).10One of the main reasons it has not been commercialized to a greater extent is the dif-ficulty in manufacturing chitosan/chitin biomaterials due to their poor water solubility under mild process-ing conditions.We found that the introduction of lac-tose moieties into chitosan(CH-LA)provided a much better water solubility at neutral pH values(Table I). In the present study,with its application in mind as a biological adhesive for soft tissues,we prepared a photocrosslinkable CH-LA w
ith photoreactive azide groups(Az-CH-LA).
Photocrosslinkable chitosan hydrogels have several properties that are superior or comparable to fibrin glue.Chitosan hydrogels are very soft and flexible, and the sealing strength of hydrogels prepared from 30mg/mL of Az-CH-LA solutions was found to be higher than that of fibrin glue in preventing air leak-age from pinholes on small intestine and blood vessels and also from incisions in trachea(Table II).This was so even though the binding strength of the hydrogel between two ham slices was found to be almost equal to that of fibrin glue.Chitosan hydrogels have been found not to detach from the surface of ham slices or blood vessels due to their firm adhesion to the tissues. It has been found,rather,that the hydrogel itself was ruptured by a strong pulling force.To the contrary, the fibrin glue was found to rupture at the interface between the glue and tissue surface.Since azide
TABLE II
Summary of Air-Sealing Strength of Chitosan Hydrogel
and Fibrin Glue
Sealing Strengthreactive materials studies
Small Intestine
mm Hg
Aorta
mm Hg
Trachea
cm H2O Chitosan gel from
30mg/mL of Az-CH-LA61±1200±18103±24 Fibrin glue51±680±2057±20 For each data point n=3.

版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。