Toxicology in Vitro 24 (2010) 721–727

Contents lists available at ScienceDirect

Toxicology in Vitro

journal homepage: www.elsevier .com/locate / toxinvi t

In vitro biocompatibility of chitosan porous skin regenerating templates(PSRTs) using primary human skin keratinocytes

C.K. Lim a, N.S. Yaacob b, Z. Ismail c, A.S. Halim a,*

a Reconstructive Sciences Unit, School of Medical Sciences, Universiti Sains Malaysia, 16150 Kubang Kerian, Kelantan, Malaysiab Department of Chemical Pathology, School of Medical Sciences, Universiti Sains Malaysia, 16150 Kubang Kerian, Kelantan, Malaysiac Chemistry Department, Faculty of Science and Technology, Universiti Pendidikan Sultan Idris, 35900 Tanjung Malim, Perak, Malaysia

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

Article history:Received 8 July 2009Accepted 12 January 2010Available online 15 January 2010

Keywords:BiocompatibilitypNHEK culturesChitosan porous skin regeneratingtemplatesCytocompatibilityGenotoxicityPro-inflammatory cytokines

0887-2333/$ - see front matter Crown Copyright � 2doi:10.1016/j.tiv.2010.01.006

* Corresponding author. Tel.: +60 09 7663141; fax:E-mail address: ashalim@kb.usm.my (A.S. Halim).

Biopolymer chitosan (b-1,4-D-glucosamine) comprises the copolymer mixture of N-acetylglucosamineand glucosamine. The natural biocompatibility and biodegradability of chitosan have recently highlightedits potential use for applications in wound management. Chemical and physical modifications of chitosaninfluence its biocompatibility and biodegradability, but it is unknown as to what degree. Hence, the bio-compatibility of the chitosan porous skin regenerating templates (PSRT 82, 87 and 108) was determinedusing an in vitro toxicology model at the cellular and molecular level on primary normal human epider-mal keratinocytes (pNHEK). Cytocompatibility was accessed by using a 3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyl tetrazolium bromide (MTT) assay from 24 to 72 h. To assess the genotoxicity of the PSRTs, DNAdamage to the pNHEK was evaluated by using the Comet assay following direct contact with the variousPSRTs. Furthermore, the skin pro-inflammatory cytokines TNF-a and IL-8 were examined to evaluate thetendency of the PSRTs to provoke inflammatory responses. All PSRTs were found to be cytocompatible,but only PSRT 108 was capable of stimulating cell proliferation. While all of the PSRTs showed someDNA damage, PSRT 108 showed the least DNA damage followed by PSRT 87 and 82. PSRT 87 and 82induced a higher secretion of TNF-a and IL-8 in the pNHEK cultures than did PSRT 108. Hence, basedon our experiments, PSRT 108 is the most biocompatible wound dressing of the three tested.

Crown Copyright � 2010 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Chitin is a b-(1,4)-D-linked polymer of N-acetylglucosamine thatis usually extracted from shellfish waste or the cell walls fromcrustaceans and arthropods. When the monomers of chitin aredeacetylated to form a mixture of N-acetylglucosamine and gluco-samine, it is called chitosan (Illum, 1998). Wound dressing in por-ous structures, such as Integra™ has been typically used as thewound dressing for partial and full-thickness wounds. The biocom-patibility and biodegradability of chitosan have made it a poten-tially useful pharmaceutical material especially for woundmanagement (Diegelmann et al., 1996; Halim et al., 1998). In thisstudy, chitosan was porous-structured using different acidic sol-vents and neutralized either by ethanol serial-hydration or sodiumbicarbonate (NaHCO3) methods before lyophilization. However,various chemical modifications aimed at further developing thechitosan into various forms might adversely influence the biocom-patibility and biodegradability of chitosan. Therefore, an evalua-tion of the biocompatibility of chemically and physically

010 Published by Elsevier Ltd. All r

+60 09 7645126.

modified chitosan must be performed to ensure that the final prod-uct is biocompatible as a wound management product.

In vitro, cells respond to the location and manner in which thebiomaterials are applied. Fibroblasts or keratinocytes, which arethe main cell constituents of the dermal and epidermal layersrespectively, can be cultured to investigate the biocompatibilityof chitosan derivatives as wound dressings. The in vitro toxicologymodel simplifies the system due to its ability to assess the cellularand molecular reactions outside an organism (Schreier et al., 1993;Cha et al., 1996). Additionally, an in vitro system often has highersensitivity than an in vivo assessment (Ulreich and Chavapilin,1983).

At the cellular level, the effect of chitosan formulations wasevaluated with a 3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyl tetra-zolium bromide (MTT) assay, using primary human keratinocytes(Lim et al., 2007). Traditional assays have measured cytotoxicityin vitro by means of either an end-stage event such as permeabilityof cytoplasmic membranes of dead and dying cells or metabolicparameters such as cell division or an enzymatic reaction (Hankset al., 1996). The MTT assay is a simple, metabolic activity basedassay which is used to determine detrimental intracellular effectson mitochondria. The MTT assay utilized the selective ability ofviable cells to reduce tetrazolium bromide into purple formazan

ights reserved.

722 C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727

crystals which are only soluble in organic solvents. Previous re-ports have shown that MTT assays adequately assess cytocompat-ibility, give results that are easily interpreted and are cost effective(Janvikul et al., 2005; Mei et al., 2005; Lim et al., 2007; Neamnarket al., 2007).

There are numerous assays that have been used to detectdeoxyribonucleic acid (DNA) damage, including the Ames, micro-nuclei, mutations and structural chromosomal aberrations assays.However, a more recent and useful approach for assessing DNAdamage is the comet assay (Singh et al., 1988). A general guidelineprovided by Tice et al. (2000) is that a cytotoxicity assay should beperformed prior to the comet assay, and a cell viability of above70% is required in order to exclude false positive results, due toapoptosis. There are many excellent reviews on the use of the co-met assay on aquatic animal cells to assess the DNA damagecaused by genotoxicants (Nacci et al., 1996; Mitchelmore and Chip-man, 1998; Cotelle and Ferard, 1999).

In addition to DNA damage assays, expression analysis of pro-inflammatory cytokines such as tumor necrosis factor-a (TNF-a)and interleukin-8 (IL-8) is another advantage of the in vitro assaythat ensures the biocompatibility of the biomaterials with humanskin cells (Allen et al., 2001).

In this study, the newly developed chitosan derivatives in theform of porous skin regenerating templates (PSRTs) were testedfor biocompatibility in a direct-contact method with primary nor-mal human epidermal keratinocyte (pNHEK) cultures. Cytotoxicity,genotoxicity and inflammation were assessed using the MTT assay,the comet assay and analysis of the pro-inflammatory cytokines(TNF-a and IL-8), respectively.

2. Materials and methods

2.1. Chitosan porous skin regenerating templates (PSRTs)

PSRTs were obtained from the Standards and Industrial Re-search Institute of Malaysia (SIRIM) and the Advanced MaterialsResearch Centre (AMREC) in Malaysia. Sterilization proceduresusing ethylene oxide were carried out for each PSRT according tothe International Standards Organization (ISO) guidelines (Part10993–7:1995: Ethylene Oxide Sterilization Residuals) beforebeing tested on primary cell cultures. Sterilized PSRTs were thenquarantined at room temperature in a dark and dry place for onemonth before being used for biocompatibility evaluation.

Chitosan was dissolved in 1% (v/v) acetic acid (PSRT 82 and 108)or 1% (v/v) lactic acid (PSRT 87) to prepare 2% (w/v) chitosan solu-tion. This was followed by an addition of 4 g glycerol as the plasti-cizer in all PSRTs. Neutralization was performed with twotechniques to achieve pH 6.2, serial hydration with 95% ethanolto 20% ethanol in PSRT 82, or NaHCO3 in PSRT 87 and PSRT 108.The chitosan solution was then molded in polytetrafluoroethylene(PTFE). During the porous structure fabrication process, the chito-san solution was frozen at �20 �C followed with lyophilization in afreeze dryer for 24 h. The following PSRTs were used in this study:

(a) PSRT 82: Neutralization of a chitosan solution using ethanol.(b) PSRT 87: Chitosan solution in lactic acid and a polyvinyl

alcohol as an additive.(c) PSRT 108: Neutralized by NaHCO3. The freezing temperature

before lyophilization was �20 �C.

2.2. Establishment of pNHEK cultures

The establishment of pNHEK cultures was performed in accor-dance with the procedures stated for human keratinocyte culturesby Gibco� (Catalog: 10744) with some modifications. Skin samples

were obtained from consenting patients who had undergone elec-tive surgery at the Hospital Universiti Sains Malaysia (USM), Ku-bang Kerian, Kelantan after being approved by the Human EthicCommittee of Universiti Sains Malaysia. Freshly obtained skin sam-ples were immediately immersed in a defined keratinocyte serum-free media (DK-SFM) (Gibco�) containing gentamicin.

Skin samples were rinsed with Dulbecco’s phosphate bufferedsaline (D-PBS) (Gibco�) containing gentamicin (20 lg/ml) and thencut into 4 � 4 mm2 sized pieces. The fragmented tissues were im-mersed in dispase (2.4 units/ml DK-SFM) for 18 h at 4 �C. After theincubation, the epidermal layer was lifted from the dermal layer.Keratinocytes in the epidermal layer were trypsinized with 0.25%trypsin–EDTA for 15 min at 37 �C, during which time the epidermallayer was aspirated in trypsin–EDTA every 3 min to facilitate thedissociation of the cells. Following incubation, serum supple-mented Dulbecco’s minimal eagle medium (DMEM) was added tostop the activity of the trypsin–EDTA. Cells were then centrifugedat 1000 rpm for 6 min. The resulting cell pellets were resuspendedin Epilife� medium (Cascade Biologics, Inc.) medium supple-mented with human keratinocyte growth supplements (HKGS-V2) (Cascade Biologics) and antibiotics (penicillin, streptomycin,amphotericin) and incubated at 37 �C and 5% CO2. The primaryseeding density was adjusted to 2 � 105 viable cells/ml.

Upon reaching 70% cell confluence, the pNHEK cultures weresubcultured by removing the medium, and the monolayer of cellswas trypsinized with 0.25% trypsin–EDTA. The cells were periodi-cally observed under the microscope until the majority of the cellswere dislodged from the flasks. Serum-supplemented DMEM wasadded, and the cells were pelleted and resuspended at 6 � 104 via-ble cells/ml in supplemented Epilife� medium. All subsequentexperiments were carried out by using the pNHEK at the secondpassage (P2).

2.3. Treatments of pNHEK cultures via direct contact

A direct-contact test was performed in accordance to the out-lines stated in the ISO, 10993-Part 5 (1992) with some modifica-tions. Each PSRT was equally cut into 3 � 3 mm2 squares andwashed with D-PBS to remove any ethylene oxide residue andpotentially toxic substances. Cells at P2 were seeded into a 24-wellflat bottom plate at a final concentration of 5 � 104 viable cells/ml.The seeded plates were then incubated at 37 �C and 5% CO2 untilthe cells achieved approximately 60% confluence in each well.The medium in each well was then discarded, and the PSRTs pieceswere gently placed onto the pNHEK monolayer in each well usingforceps. Freshly supplemented Epilife� medium was then addedaccordingly into each well. Each chitosan derivative was assayedin triplicate at 24, 48 and 72 h. The use of low density polyethylene(LDPE) was included as the negative control for comparison in eachsubsequent experiment.

2.4. Cell viability assay using the colorimetric tetrazolium salt (MTT)

The MTT reagent was purchased from Sigma–Aldrich (Catalog:M2128). The MTT reagent was dissolved in D-PBS (without calciumand magnesium) to a final concentration of 5 mg/ml and filter-ster-ilized before use. At each time interval (24, 48 and 72 h) of thepost-treatment phase, pieces of PSRTs were carefully removedfrom each well (24-well plate) using forceps. A total of 100 ll ofMTT solution was then added to each well. Incubation with MTTwas performed for 4 h in the dark at 37 �C supplemented with 5%CO2. After the incubation, the culture medium in each well wasremoved and the formazan crystals were dissolved with 1 ml ofdimethyl sulfoxide (DMSO) (Amresco, ACS grade). For analysis,100 ll was transferred from each well (24-well plate) into a96-well flat bottom plate, and read at 570 nm with a reference

C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727 723

wavelength of 690 nm using an enzyme-linked immunosorbent as-say (ELISA) reader (Brand: SUNRISE�). Each PSRT was performedindependently in triplicate. The cytotoxicity of chitosan PSRTs onthe pNHEK cultures was calculated using the following formula:

OD of treated cells� OD of background controlOD of negative controlðLDPEÞ � OD of background control

� 100%:

PSRT 82

PSRT 87 PSRT 108

Cytotoxicity of PSRTs relative to LDPE

0

20

40

60

80

100

120

140

24h 48h 72h

Time Intervals (Hours)

Cel

l Via

bilit

y (%

)

** *

Fig. 1. Average percentages of cell viability (mean ± SD) accessed by MTT from 24to 72 h compared with the LDPE negative control (n = 3) and *p < 0.05.

2.5. Comet assay for genotoxicity assessment

This experiment was conducted according to the guidelines de-scribed by Tice et al. (2000) with some modifications. Only cultureswith a cell viability of greater than 70% in the cytotoxicity assaywere further assessed for genotoxicity by the comet assay. Trevi-gen’s CometAssay™ lysis solution (Catalog: 4250–050-01) waschilled at 4 �C for 20 min before use. Low melting point agarose(LMAgarose) was melted at 37 �C. Cells at a density of1 � 105 cells/ml were then combined with the melted LMAgaroseat a ratio of 1:10 (v/v), and 75 ll was immediately pipetted ontoa CometSlide™ (Trevigen�). The slides were placed flat at 4 �C inthe dark for 30 min in a high humidity environment until the gelsolidified. The slides were then immersed in the pre-chilled lysissolution and left on ice at 4 �C for 60 min followed by the additionof a freshly-prepared alkaline solution at pH > 13 (0.012 g/mlNaOH pellets, 5 ll/ml of 200 mM EDTA and 995 ll dH2O) for60 min at room temperature in the dark.

To perform the alkaline electrophoresis, the slides were trans-ferred from the alkaline solution to a horizontal electrophoresisapparatus, placed flat onto a gel tray and centered between theelectrodes. The alkaline electrophoresis solution at pH > 13 (12 g/l NaOH pellets, 2 ml/l of 0.5 M EDTA at pH 8, adjusted to 1 l withdH2O) was then poured into the tray to cover the slide. The voltagewas set to about 1 V/cm, and the electrophoresis was performedfor 40 min at a constant 300 mA. Later, the slides were rinsed sev-eral times in dH2O and then immersed in 70% ethanol for 5 min.The slides were air-dried and stained with the Trevigen CometAs-say� silver staining solution (Catalog: 4254–200-K). In this exper-iment, the comet tail length for each pNHEK treatment (intriplicates) was scored at 24 h and 72 h post-treatment. ThepNHEK cells treated with cyclophosphamide monohydrate andorganotin polyvinylchloride (PVC) served as the positive controlsand treatment with LDPE served as the negative control. As anadditional negative control, pNHEK cells without any treatmentwere included (cultured cells only). By using a light microscopewith image analysis software (Raxvision Biowizard 4.0), 100 ran-domly selected comets of each treatment in triplicate were scoredby measuring the comet tail length (lm). The comet tail length wasmeasured using a microscope micrometer from the demarcatingline of the nucleus to the most distant observed DNA fragment.The longer the comet tail length, the greater the DNA damage thathad occurred in each treatment of the pNHEK cells.

2.6. Pro-inflammatory cytokines (TNF-a and IL-8)

The supernatants from each treated pNHEK were obtained at24, 48 and 72 h, and the amounts of secreted human TNF-a andIL-8 from the pNHEK were quantitatively measured by ELISA. Pro-tocols were performed according to the manufacturer’s (BLK Diag-nostics) instructions. Microwell strips were pre-coated withantibodies against human TNF-a and IL-8, washed twice withapproximately 300 ll/well of washing buffer (PBS with 1%Tween-20) and tapped on an absorbent pad to remove the excessbuffer. Standard dilutions for each cytokine were prepared by add-ing 100 ll of sample diluent (protein matrix) into 7 wells in dupli-

cates, followed by 100 ll of pre-reconstituted standards for TNF-aand IL-8 to make serial dilutions ranging from 8–500 pg/ml forTNF-a to 16–1000 pg/ml for IL-8. Only 100 ll of the sample diluentwas left in the blank wells.

The absorbance of each microwell was read at 450 nm as theprimary wavelength and at 620 nm as the reference wavelength.Concentrations (in pg/ml) of TNF-a and IL-8 in each treatmentcondition were determined from the respective standard curve.For the treatment samples that had been diluted according the gi-ven instructions, the concentrations were multiplied by the dilu-tion factor (�2). In this experiment, the LDPE negative control,and the cultured cells only negative control were included. ThepNHEK cells treated with organotin-PVC served as the positivecontrol, and culture medium alone served as the backgroundcontrol.

2.7. Statistical analyses

Data are expressed as the mean and standard deviation (SD) inall experiments. The Kruskal–Wallis one-way analysis of varianceand Wilcoxon signed-rank tests were implemented for statisticalanalysis. Differences were regarded as significant at p < 0.05.

3. Results

3.1. Cytotoxicity by MTT

Viability of the pNHEK in the LDPE negative control was slightlyless than that in the cultured cells only negative control based onthe higher optical density (OD) values in the control wells. How-ever, the minor difference between the number of viable cells inthe LDPE negative control and the cultured cells only negative con-trol wells was not statistically significant (p > 0.05) (data notshown). Thus, the viability of the pNHEK cells after treatment withthe PSRTs is more precise if compared to the negative control. Noneof the PSRTs induced severe cytotoxicity in the pNHEK cultures atany of the time intervals; however, PSRT 108 sustained a signifi-cantly better proliferation rate compared with PSRT 87 and PSRT82 at all time points (Fig. 1). The lowest viability percentage wasachieved by PSRT 82 at 72 h post-treatment (78.73 ± 2.64%).

The color of the medium in the presence of PSRT 108 and PSRT82 remained unchanged. In contrast, culture medium with PSRT 87during direct-contact experiments was yellow on day three post-treatment.

724 C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727

3.2. Genotoxicity by the comet assay

Genotoxic substances lead to DNA strand breaks, and the bro-ken DNA migrates from the nucleus to the cytoplasm and out ofthe cell under electrophoretic circumstances. By measuring thedistance that the broken DNA migrates from the cell nucleus (taillength), a relatively reliable genotoxicity assessment of a bio-sub-stance can be carried out (Singh et al., 1988; Cotelle and Ferard,1999; Kamer and Rinkevich, 2002; Muller et al., 2002; Lee and Ste-inert, 2003; Wang et al., 2003). There was no significant differencein the resulting DNA tail length of both positive controls in thisstudy. Therefore, in this experiment, organotin-PVC was chosenas the positive control and LDPE was chosen as the negative con-trol. The pNHEK cells treated with PSRT 108 were observed to havea tail length comparable to the LDPE negative control at 24 and72 h post-treatment (Fig. 2). However, significant genotoxicitywas detected for PSRT 82 and PSRT 87 at 24 h post-treatment(p < 0.05) with approximately three times the tail lengths of theLDPE negative control. Both were even greater at 72 h at which

Genototxicity of PSRTs via Comet assay

0

50

100

150

200

72 hrs24 hrsTime (Hours)

PSRT 82 PSRT 87 PSRT 108 Positive Control (Organotin-PVC)Negative Control (LDPE)Negative Control (cultured cells only)Positive control (Cyclophosphamide monohydrate)

**

Tail

Len

gth

(µm

)

Fig. 2. Means ± SD of the comet tail length (lm) of single pNHEK cells from each ofthe treatment conditions. At 24 and 72 h, treatment with PSRT 82 and PSRT 87significantly induced longer tail length in a cell compared with LDPE negativecontrol, respectively (n = 3) and *p < 0.05.

TNF-alpha expression for pNH

0

50

100

150

200

250

300

350

24 hours 48 hours 7Time (hours)

Con

cent

rati

ons

(pg/

ml

*

Fig. 3. Concentrations (mean ± SD) of TNF-a (pg/ml) secreted into the pNHEK culture sinduced TNF-a production at 72 h compared with LDPE negative control. TNF-a secretio48 h (n = 3) and *p < 0.05.

time the tail lengths were approximately five times that of the neg-ative control (LDPE).

3.3. Skin pro-inflammatory cytokines (TNF-a and IL-8)

The pNHEK cultures constitutively released TNF-a and IL-8 in atime-dependent manner. This phenomenon could be seen in bothnegative controls (LDPE and cultured cells only), which both exhib-ited slight increases in the concentration of TNF-a and IL-8 from 24to 72 h.

For TNF-a expression (Fig. 3), the lowest secretion level oc-curred in the cultured cells only negative control and the amountwas comparable to the pNHEK cells treated with PSRT 108 at alltime intervals. Meanwhile, the LDPE negative control expressedapproximately 19 pg/ml more TNF-a than PSRT 108. Among thetreatments, the pNHEK cells treated with organotin-PVC secretedthe greatest amount of TNF-a (approximately four times the nega-tive control) at 72 h, followed by PSRT 82 and PSRT 87.

The amount of IL-8 expression was higher than the expressionof TNF-a in the pNHEK cultures. There was a difference of approx-imately 460 pg/ml between the positive and LDPE negative con-trols (Fig. 4). Treatment with PSRT 108 contributed to a lowerexpression of both cytokines compared with PSRT 87 and PSRT82 at all time points. At 72 h post-treatment with PSRT 108, IL-8expression was observed to decrease significantly, by 160 pg/ml(p < 0.05) from 48 h. In contrast, treatment with PSRT 82 signifi-cantly (p < 0.05) increased the IL-8 secretion from 24 to 72 hpost-treatment. However, IL-8 secretion under treatment withPSRT 87 was not significantly (p > 0.05) enhanced or reducedthroughout the time intervals, but it was maintained at higher lev-els than the LDPE negative control.

4. Discussion

In this study, the MTT cell viability assay was used as an effi-cient and easy way to investigate the cytocompatibility of testmaterials on cultured cells (Carmichael et al., 1987; Regina et al.,1998; Kim et al., 2005; Lim et al., 2007). PSRT 108, 87 and 82 (intheir final forms) were placed in direct contact with pNHEK cul-tures to assess the cytotoxicity of these three biomaterials. Accord-ing to Saw et al. (2005), the direct-contact method has proven to be

EK treated with PSRTs

2 hours

PSRT 82

PSRT 87

PSRT 108

Positive Control (Organotin-PVC)

Negative Control (LDPE)

Negative Control (cultured cellsonly)

*

upernatant at 24, 48 and 72 h. Treatment with PSRT 82 and PSRT 87 significantlyn was further enhanced in positive control at 72 h post-treatment compared with

IL-8 expression for pNHEK treated with PSRTs

0

200

400

600

800

1000

1200

1400

24 hours 48 hours 72 hoursTime intervals (Hours)

Con

cent

rati

ons

(pg/

ml

PSRT 82

PSRT 87

PSRT 108

Positive Control (Organotin-PVC)

Negative Control (LDPE)

Negative Control (culturedcells only)

**

Fig. 4. Concentrations (mean ± SD) of IL-8 (pg/ml) secreted into the pNHEK culture supernatant at 24, 48 and 72 h. Level of IL-8 secretion under treatment with PSRT 82 wassignificantly elevated from 24 to 72 h. In contrast, PSRT 108 markedly reduced IL-8 production from 24 to 72 h (n = 3) and *p < 0.05.

C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727 725

a more sensitive and reliable method to determine the cytotoxicityof biomaterials as compared to the indirect-contact method.

Low density polyethylene (LDPE) was used as negative controlin this study as recommended in the ISO, 10993-Part 5 (1992).However, the minor difference between the number of viable cellsin the LDPE negative control and the cultured cells only negativecontrol was not significantly different statistically (results notshown). This finding might be due to the physical impact fromthe weight of the LDPE though this is not significant. Therefore,in order to obtain the net cytotoxic effect of the PSRTs in this study,a comparison between treated cells and the LDPE negative controlwould be ideally performed under conditions in which the PSRTshave a similar physical impact on the cultured cells as LDPE. Thus,the impact of the physical weight can be eliminated from the cal-culation. PSRTs in this study were prepared by various techniquessuch as the use of solvents and neutralization techniques. Aceticacid and lactic acid were implemented as the solvent for the chito-san powder, as chitosan is soluble in aqueous acidic medium (Sash-iwa and Shigemasa, 1999). Introduction of polyvinyl alcohol intothe chitosan solution in the PSRT 87 served to enhance mechanicalstrength and hydrophilicity of the PSRT. Neutralization of theacidic chitosan solution with glycerol was performed by serial-hydration with ethanol or, with sodium bicarbonate in order toeliminate the residual acidic residues before casting into the por-ous structure.

Several transformed keratinocyte culture models have beenused to study carcinogenesis, differentiation, apoptosis and cell cy-cle regulation (Yuspa et al., 1994). Our focus here is on primary nor-mal (non-transformed) keratinocyte culture models becausecutaneous toxicity testing of new drugs, cosmetic products andother chemicals requires phenotypically normal cell systems (Aco-sta et al., 1985), as the loss of differentiated function in transformedcell types undoubtedly confounds the interpretation of the results(Flint, 1990). In addition, Sharpe and Fisher (1990) further pointedout that the human keratinocytes cultures are far more sensitivethan fibroblasts due to the fact that the keratinocytes are culturedin a serum-free environment, whereas fibroblasts cultures requireapproximately 10% serum in the culture medium.

Culturing of the pNHEKs with PSRT 108 and PSRT 82 resulted inno physical changes of PSRTs after three days in the culture med-ium. The color of the medium in the wells with PSRT 108 and PSRT82 remained unchanged. These data agree with the findings pub-lished by Sundararajan and Howard (1999) who stated that sam-

ples washed by serial-hydration with ethanol exhibited nosignificant shape changes. According to Rathke and Hudson(1994), scaffolds hydrated in sodium hydroxide will exhibit shrink-age and distortion due to base-induced changes in crystallinity andassociated structural stress. Additionally, the base-induced shrink-age may also result from numerous entrapped air bubbles. PSRT108 stimulated the most cell proliferation and differentiation inthe cytotoxicity experiment. Moreover, PSRT 108 exceeded 100%cell viability compared with the LDPE negative control. This findingmay be associated with the extra cells derived from cell differenti-ation, observed by the increase in the absorbance values from ele-vated total formazan crystal. According to Watson et al. (2004) andSanmano et al. (2005), cultured keratinocytes in a confluent statemay undergo differentiation.

Color change of the medium can indicate the shifting of the pHin a cell culture system with phenol red, which acts as a pH indica-tor in the medium. Thus, the presence of lactic acid in PSRT87 orother newly formed chemical complexes from chitosan and lacticacid may be partially extracted by the culture medium during di-rect-contact experiments and may lead to slightly acidic culturemedium on day three post-treatment. This effect may explain thereduced cell viability of the pNHEK cells treated with PSRT 87 at72 h. The use of ethanol in the neutralization step for PSRT was re-ported to have a negligible effect on bacterial adhesion that couldbe attributed to the contact-dependent characteristics of chitosan(Sarasam et al., 2008). In this experiment, pNHEK cultures treatedwith PSRT 82 had decreasing cell viability percentages from thefirst to the third day of treatment, potentially due to entrappedethanol residues, or newly formed chemical complexes with chito-san itself in the three-dimensional (3D) structure of the PSRT thatcan serve as a cytotoxic agent in the cell culture system.

Although all PSRTs were cytocompatible at the cellular level, butthe long-term usage of the PSRTs as a wound dressing biomaterialwould still be questionable if no further experiments were per-formed (Keong and Halim, 2009). Thus, in vitro screening at themolecular level was performed using single cell gel electrophoresis(the comet assay). According to Leroy et al. (1996), the comet assayis sensitive enough to assay the DNA damage that is caused bygenotoxic substances at the level of the individual cell with a rela-tively small number of any type of eukaryotic cells. Many research-ers have reported that the comet assay is rapid and is also the mostsensitive assay as compared to the micronucleus and Ames tests(Sasaki, 1999; Iwahori et al., 2004; Lah et al., 2005; Lah et al.,

726 C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727

2008; Demma et al., 2009). Measurements of the tail length of lib-erated damaged DNA correlate well with the mutagenic and carcin-ogenic properties of environmental pollutants (Mitchelmore andChipman, 1998; Cotelle and Ferard, 1999; Muller et al., 2002). A sig-nificant increase in DNA damage was associated with high toxicity.Similar results were obtained when monitoring the toxic and geno-toxic effects in some plants (Gichner et al., 2006). The comet assaywas successfully used to detect the DNA strand breaks of bone mar-row cells treated with fungal-derived high molecular weight chito-san (HMWC) (Uen et al., 2008). A general guideline for the cellviability assay is that, by the end of the exposure period, the dosesin the test must not decrease the cell viability compared to the con-trol culture in vitro by more than 30% (Tice et al., 2000; Poli et al.,2002). In other words, the cell viability of the PSRTs in this studymust achieve at least 70% of viable cells compared to the LDPE neg-ative control or cultured cells only negative control before they aresubmitted for the comet assay. This avoids false positive resultssuch as apoptosis that may be due to cytotoxicity. In this genotox-icity experiment, PSRT 87 and PSRT 82, both of which werecytocompatible, displayed a genotoxic phenomenon at 72 h post-treatment. In fact, the damaged DNA migrated almost three times(PSRT 82) and four times (PSRT 87) further compared with the LDPEnegative control at 24 h post-treatment. In contrast, the DNA dam-age in cells exposed to PSRT 108 was comparable to the LDPE neg-ative control at all time points. This result is probably the outcomeof activating the DNA repair system or the death of severely dam-aged cells compensated for by cell division, which likely occurredin the pNHEK cells treated with PSRT 108. In a genotoxicity studyconducted by Yuan et al. (2009), the chitosan-oligosaccharideswere reported to have prominent protective effects on the DNAdamage and to prohibit the apoptosis of pancreatic islet cells. Inboth positive controls of this experiment, although there was nei-ther an obvious increase nor a decrease in the mean tail length ateither time point, the positive controls still maintained the longestmean tail length expected for damaged cells without activation of aDNA repair system.

Cytokines play roles in regulating the immunologic responsesand host responses to infectious agents and inflammatory stimuli.The use of keratinocyte cell cultures have been gaining acceptanceas an alternative to animal studies for such research (Roguet, 1998;Bernstein and Vaughan, 1999). IL-8 is a powerful neutrophil attrac-tant and is commonly produced by keratinocytes after external irri-tation (Mohamadzadeh et al., 1994; Yen et al., 1996; Ushio et al.,1999). Release of IL-8 is preceded by the release of TNF-a (Lugerand Schwarz, 1990). Toxic substances may stimulate higher expres-sion levels of both TNF-a and IL-8. Both TNF-a and IL-8 expressionwere found to be comparable in the cultured cells only negativecontrol and pNHEK cells treated with PSRT 108. However, treatmentwith PSRT 108 resulted in a lower expression level of TNF-a com-pared to the LDPE negative control at all time points. This may bedue to the chemical components in PSRT 108 that help to alter theIL-8 receptor binding by TNF-a, which induces IL-8 production (Al-len et al., 2001). Alternatively, the leachable chitosan chemical mayform a complex with itself to inhibit the secretion of TNF-a, perhapsthrough the binding at the TNF-a receptors (Kim et al., 2002). PSRT87 and PSRT 82 exhibited higher expression of both TNF-a and IL-8than the LDPE negative control, indicating that both may have led toa pro-inflammatory reaction in the in vitro-cultured pNHEK cell.This phenomenon could be due to leachable products or chemicalcomplexes present in the culture medium.

5. Conclusion

PSRT 82, 87 and 108 were all cytocompatible for 72 h at the cel-lular level in vitro. When screened in detail at the molecular level,

PSRT 87 and PSRT 82 revealed a genotoxic effect at 24 h post-treat-ment and provoked a greater inflammatory response than PSRT108. This study determined that PSRT 108 is the most biocompat-ible wound dressing tested, followed by PSRT 87 and PSRT 82.

Acknowledgments

This work was supported by a grant (No. 03-03-01-0000-PR0071/05) from the Intensification of Research in Priority AreaProgram (IRPA), Ministry of Science, Technology and Innovation(MOSTI) Malaysia.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.tiv.2010.01.006.

References

Acosta, D., Sorensen, E.M.B., Anuforo, D.C., Mitchell, D.B., Ramos, K., Santone, K.S.,Smith, M.A., 1985. An in vitro approach to the study of target organ toxicity ofdrugs and chemicals. In Vitro Cellular and Developmental Biology 21, 495–504.

Allen, D.G., Riviere, J.E., Monteiro-Riviere, N.A., 2001. Cytokine induction as ameasure of cutaneous toxicity in primary and immortalized porcinekeratinocytes exposed to jet fuels, and their relationship to normal humanepidermal keratinocytes. Toxicology Letters 119, 209–217.

Bernstein, I.A., Vaughan, F.L., 1999. Cultured keratinocytes in in vitrodermatoxicological investigation: a review. Toxicology and EnvironmentalHealth B 2, 1–30.

Carmichael, J., Degraff, W., Gazdar, A., Minna, J., Mitchell, J., 1987. Evaluation of atetrazolium-based semiautomated colorimetric assay: assessment ofchemosensitivity testing. Cancer Research 47, 936–942.

Cha, D., O’Brien, D., O’Toole, E.A., Woodley, D.T., Hudson, L.G., 1996. Enhancedmodulation of keratinocyte motility by TGF-relative to epidermal growth factor.Investigative Dermatology 106, 590–597.

Cotelle, S., Ferard, J.F., 1999. Comet assay in genetic ecotoxicology: a review.Environmental and Molecular Mutagenesis 34, 246–255.

Demma, J., Engidawork, E., Hellman, B., 2009. Potential genotoxicity of plantextracts used in Ethiopian traditional medicine. Ethnopharmacology 122, 136–142.

Diegelmann, R.F., Dunn, J.D., Lindbald, W.J., Cohen, I.K., 1996. Analysis of the effectsof chitosan on inflammation, angiogenesis, fibroplasia and collagen depositionin polyvinyl alcohol sponge implants in rat wounds. Wound Repair andRegeneration 4, 48–52.

Flint, O.P., 1990. In vitro toxicity testing: purpose, validation and strategy.Alternatives to Laboratory Animals 18, 11–18.

Gichner, T., Patkova, Z., Szakova, J., Demnerova, K., 2006. Toxicity and DNA damagein tobacco and potato plants growing on soil polluted with heavy metals.Ecotoxicology and Environmental Safety 65, 425–426.

Halim, A.S., Stone, C.A., Devaraj, V.S., 1998. The hyphecan cap: a biological fingertipdressing. Injury 29 (4), 261–263.

Hanks, C.T., Wataha, J.C., Sun, Z., 1996. In vitro models of biocompatibility: a review.Dental Materials 12, 186–193.

Illum, L., 1998. Chitosan and its use as a pharmaceutical excipient. PharmaceuticalResearch 15, 1326–1331.

ISO, 10993-5, 1992. Biological Evaluation of Medical Devices. Part 5: Tests forCytotoxicity: In Vitro Methods. National Organization for Standardization,Ctineva.

Gichner, T., Patkova, Z., Szakova, J., Demnerova, K., 2006. Toxicity and DNA damagein tobacco and potato plants growing on soil polluted with heavy metals.Ecotoxicology and Environmental Safety 65, 425–426.

Iwahori, K., Miyata, N., Yamamoto, Y., Sone, D., Aoyama, K., 2004. Normalization ofDNA damage indices in comet assay for comparing with various mutagenicitytests. Japanese Journal of Water Treatment Biology 40, 63–70.

Janvikul, W., Uppanan, P., Prateepasen, R., Swasdison, S., Muzarelli, Ricardo., 2005.Biocompatibility of chitin derivatives in vitro. Journal of Metals, Materials andMinerals 15, 59–62.

Kamer, I., Rinkevich, B., 2002. In vitro application of the comet assay for aquaticgenotoxicity: considering a primary culture versus a cell line. Toxicology inVitro 16, 177–184.

Keong, L.C., Halim, A.S., 2009. In vitro in biocompatibility assessment forbiomedical-grade chitosan derivatives in wound management. InternationalJournal of Molecular Sciences 10, 1300–1313.

Kim, D.H., Lee, S.H., Kim, K.N., Kim, K.M., Shim, I.B., Lee, Y.K., 2005. Cytotoxicity offerrite particles by MTT and agar diffusion method for hyperthermicapplication. Magnetism and Magnetic Materials 293, 287–292.

Kim, M.S., Sung, M.J., Seo, S.B., Yoo, S.J., Lim, W.K., Kim, H.M., 2002. Water-solublechitosan inhibits the production of pro-inflammatory cytokine in humanAstrocytoma cells activated by Amyloid b peptide and interleukin-1b.Neuroscience Letters 321, 105–109.

C.K. Lim et al. / Toxicology in Vitro 24 (2010) 721–727 727

Lah, B., Zinko, B., Tisler, T., Logar, M., 2005. Genotoxicity detection in drinking waterby Ames test, Zimmermann test and Comet assay. Acta Chimica Slovenica 52,341–348.

Lah, B., Vidic, T., Glasencnik, E., Cepeljnik, T., Gorjang, G., Marinsek-Logar, R., 2008.Genotoxicity evaluation of water soil leachates by Ames test, comet assay andpreliminary Tradescantia micronucleus assay. Environmental Monitoring andAssessment 139, 107–118.

Lee, R.F., Steinert, S., 2003. Use of the single cell gel electrophoresis/comet assay fordetecting DNA damage in aquatic (marine and freshwater) animals. MutationResearch 544, 43–64.

Leroy, T., Van Hummelen, P., Anard, D., Castelain, P., Kirsch-Volders, M., Lauwerys,R., Lison, D., 1996. Evaluation of three methods for the detection of DNA single-strand breaks in human lymphocytes: alkaline elution, nick translation, andsingle-cell gel electrophoresis. Toxicology and Environmental Health 47, 409–422.

Lim, C.K., Halim, A.S., Lau, H.Y., Ujang, Z., Hazri, A., 2007. In vitro cytotoxicologymodel of oligo-chitosan and N,O-carboxymethyl chitosan using primary normalhuman epidermal keratinocytes cultures. Applied Biomaterials andBiomechanics 5, 82–87.

Luger, T.A., Schwarz, T., 1990. Evidence for an epidermal cytokine network.Investigative Dermatology 95, 104S–110S.

Mei, N., Chen, G., Zhou, P., Chen, X., Shao, Z.Z., 2005. Biocompatibility of poly(E-caprolactone) scaffold modified by chitosan: the fibroblast proliferation in vitro.Journal of Biomaterials Applications 19, 323–339.

Mitchelmore, C.L., Chipman, J.K., 1998. DNA strand breakage in aquatic organismsand the potential value of the comet assay in environmental monitoring.Mutation Research 399, 135–147.

Mohamadzadeh, M., Muller, M., Hultsch, T., Enk, A., Saloga, J., Knop, J., 1994.Enhanced expression of IL-8 in normal human keratinocytes and humankeratinocytes cell line HaCaT in vitro after stimulation with contact sensitizers,tolerogens and irritants. Experimental Dermatology 3, 298–303.

Muller, P., Stock, T., Bauer, S., Wolff, I., 2002. Genotoxicological characterization ofcomplex mixtures: Genotoxic effects of a complex mixture of perhalogenatedhydrocarbons. Mutation Research 515, 99–109.

Nacci, D.E., Cayula, S., Jackim, E., 1996. Detection of DNA damage in individual cellsfrom marine organisms using the single cell gel assay. Aquatic Toxicology 35,197–210.

Neamnark, A., Sanchavanakit, N., Pavasant, P., Bunaprasert, T., Supaphol, P.,Rujiravanit, R., 2007. In vitro biocompatibility evaluations of hexanoylchitosan film. Carbohydrate Polymers 68, 166–172.

Poli, P., de Mello, M.A., Buschini, A., Mortara, R.A., de Albuquerque, C.N., da Silva, S.,Rossi, C., Zucchi, T.M.A.D., 2002. Cytotoxic and genotoxic effects of megazol, ananti-Chagas’ disease drug, assessed by different short-term tests. BiochemicalPharmacology 64, 1617–1627.

Rathke, T.D., Hudson, S.M., 1994. Review of chitin and chitosan as fiber and filmformers. Macromolecular Chemistry and Physics 34, 375–437.

Regina, Z., Youxin, L., Thomas, K., 1998. Biocompatibility testing of ABA tri-blockcopolymers consisting of poly(L-lactic-co-glycolic acid) A blocks attached to acentral poly(ethylene oxide) B block under in vitro conditions using differentL929 mouse fibroblasts cell culture models. Controlled Release 56, 249–258.

Roguet, R., 1998. Use of skin cell cultures for in vitro assessment of corrosion andcutaneous irritancy. Cell Biology and Toxicology 15, 63–75.

Sanmano, B., Mizoguchi, M., Suga, Y., Ikeda, S., Ogawa, H., 2005. Engraftment ofumbilical cord epithelial cells in athymic mice. In an attempt to improvereconstructed skin equivalents used as epithelial composite. Journal ofDermatological Science 37, 29–39.

Sarasam, A.R., Brown, P., Khajotia, S.S., Dmytryk, J.J., Madihally, S.V., 2008.Antibacterial activity of chitosan-based matrices on oral pathogens. Journal ofMaterials Science: Materials in Medicine 19, 1083–1090.

Sasaki, Y.F., 1999. Detection of mouse organ-specific genotoxicity by the cometassay. Environmental Mutagenesis Research 21, 131–140.

Sashiwa, H., Shigemasa, Y., 1999. Chemical modification of chitin and chitosan 2:preparation and water soluble property of N-acylated or N-alkylated partiallydeacetylated chitins. Carbohydrate Polymers 39, 127–138.

Saw, T.Y., Cao, T., Yap, A.U.J., Ng, M.M.L., 2005. Tooth slice organ culture andestablished cell line culture models for cytotoxicity assessment of dentalmaterials. Toxicology in Vitro 19, 145–154.

Schreier, T., Degen, E., Baschong, W., 1993. Fibroblast migration and proliferationduring in vitro wound healing: A quantitative comparison between variousgrowth factors and a low molecular weight blood dialysate used in the clinic tonormalise impaired wound healing. Research in Experimental Medicine 193,195–205.

Sharpe, G.R., Fisher, C., 1990. Time-dependent inhibition of growth of humankeratinocytes and fibroblasts by cyclosporine A: effect on keratinocytes attherapeutic blood levels. British Journal of Dermatology 123, 207–213.

Singh, N.P., McCoy, M.T., Tice, R.R., Schneider, E.L., 1988. A simple technique for thequantitation of low levels of DNA damage in individual cells. Experimental CellResearch 175, 184–191.

Sundararajan, V.M., Howard, W.T.M., 1999. Porous chitosan scaffolds for tissueengineering. Biomaterials 20, 1133–1142.

Tice, R.R., Agurell, E., Anderson, D., Burlinson, B., Hartmann, A., Kobayashi, H.,Miyamae, Y., Rojas, E., Ryu, J.C., Sasaki, Y.F., 2000. Single cell gel/comet assay:Guidelines for in vitro and in vivo genetic toxicology testing. Environmental andMolecular Mutagenesis 35, 206–211.

Uen, Y.H., Wang, Y.J., Hsu, T.M., Chen, M.H., Chan, H.Y., Ho, Y.S., Lin, S.Y., 2008.Effects of fungal-derived high molecular weight chitosan on 5-fluorouacil-induced adverse reactions. Bioactive and Compatible Polymers 23, 458–472.

Ulreich, J.B., Chavapilin, M., 1983. In vitro toxicity testing: a quantitative microassay.In: Brown, S.A. (Ed.), Cell Culture Test Methods. ASTM, Philadelphia., pp. 102–113.

Ushio, H., Nohara, K., Fujimaki, H., 1999. Effect of environmental pollutants on theproduction of pro-inflammatory cytokines by normal human dermalkeratinocytes. Toxicology Letters 105, 17–24.

Wang, H.L., Zhang, D.F., Luo, Z.F., Lou, Y., Li, Q.R., Wang, Y.Z., 2003. In vitro study ongenotoxicity of dichloromethane extracts of valerian (DEV) in humanendothelial ECV304 cells and the effect of vitamins E and C in attenuating theDEV-induced DNA damages. Toxicology and Applied Pharmacology 188, 36–41.

Watson, A., Baker, C., Bailey, J., Fray, T., Markwell, P., 2004. A high yield method forgrowing primary canine keratinocytes. The Veterinary Journal 168, 81–86.

Yen, H.T., Chiang, L.C., Wen, K.H., Chang, S.F., Tsai, C.C., Yu, C.L., Yu, H.S., 1996.Arsenic induces interleukin-8 expression in cultured keratinocytes. Archives ofDermatological Research 288, 716–717.

Yuan, W.P., Liu, B., Liu, C.H., Wang, X.Y., Zhang, M.S., Meng, X.M., Xia, X.K., 2009.Antioxidant activity of chitosan-oligosaccharides on pancreatic islet cells instreptozotocin-induced diabetes in rats. World Journal of Gastroenterology 15,1339–1345.

Yuspa, S.H., Dlugosz, A.A., Cheng, C.K., Denning, M.F., Tennenbaum, T., Glick,A.B., Winberg, W.C., 1994. Role of oncogenesis and tumor suppressorgenes in multi-stage carcinogenesis. Journal of Investigative Dermatology103, 90S–95S.