release of metronidazole from electrospun poly(l-lactide-co-d/l-lactide) fibers for local...
TRANSCRIPT
Rpt
MMBa
b
c
d
a
A
R
R
1
A
K
E
P
M
B
P
P
D
D
0d
d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 179–188
Available online at www.sciencedirect.com
jo u rn al hom epa ge : www.int l .e lsev ierhea l th .com/ journa ls /dema
elease of metronidazole from electrospunoly(l-lactide-co-d/l-lactide) fibers for local periodontitisreatment
arkus Reisea,∗, Ralf Wyrwab, Ulrike Müllerb, Matthias Zylinskia, Andrea Völpela,atthias Schnabelrauchb, Albrecht Bergb, Klaus D. Jandtc, David C. Wattsc,d,
ernd W. Siguscha
Department of Conservative Dentistry, Jena University Hospital, Friedrich-Schiller-Universität Jena, GermanyDepartment of Biomaterials, INNOVENT e.V., Jena, GermanyInstitute of Materials Science and Technology (IMT), Friedrich-Schiller-University Jena, GermanyUniversity of Manchester, School of Dentistry, Manchester M15 6FH, UK
r t i c l e i n f o
rticle history:
eceived 18 July 2011
eceived in revised form
0 December 2011
ccepted 11 December 2011
eywords:
lectrospun fibers
olymeric fibers
etronidazole
iomaterials
olylactide
eriodontitis
rug delivery system
ental materials
a b s t r a c t
Objectives. We aimed to achieve detailed biomaterials characterization of a drug delivery sys-
tem for local periodontitis treatment based on electrospun metronidazole-loaded resorbable
polylactide (PLA) fibers.
Methods. PLA fibers loaded with 0.1–40% (w/w) MNA were electrospun and were characterized
by SEM and DSC. HPLC techniques were used to analyze the release profiles of metronidazole
(MNA) from these fibers. The antibacterial efficacy was determined by measuring inhibition
zones of drug-containing aliquots from the same electrospun fiber mats in an agar diffusion
test. Three pathogenic periodontal bacterial strains: Fusobacterium nucleatum, Aggregatibacter
actinomycetemcomitans and Porphyromonas gingivalis were studied. Cytotoxicity testing was
performed with human gingival fibroblasts by: (i) counting viable cells via live/dead staining
methods and (ii) by exposing cells directly onto the surface of MNA-loaded fibers.
Results. MNA concentration influenced fiber diameters and thus w/w surface areas: diame-
ter being minimal and area maximal at 20% MNA. HPLC showed that these 20% MNA fibers
had the fastest initial MNA release. From the third day, MNA release was slower and nearly
linear with time. All fiber mats released 32–48% of their total drug content within the first 7
days. Aliquots of media taken from the fiber mats inhibited the growth of all three bacterial
strains. MNA released up to the 28th day from fiber mats containing 40% MNA significantly
decreased the viability of F. nucleatum and P. gingivalis and up to the 2nd day also for the resis-
tant A. actinomycetemcomitans. All of the investigated fibers and aliquots showed excellent
cytocompatibility.
Significance. This study shows that MNA-loaded electrospun fiber mats represent an inter-
esting class of resorbable drug delivery systems. Sustained drug release properties and
cytocompatibility suggest
tal diseases.
© 2011 Academy
∗ Corresponding author at: Schlossgasse 2, 07743 Jena, Germany. Tel.: +4E-mail address: [email protected] (M. Reise).
109-5641/$ – see front matter © 2011 Academy of Dental Materials. Puoi:10.1016/j.dental.2011.12.006
their potential clinical applicability for the treatment of periodon-
of Dental Materials. Published by Elsevier Ltd. All rights reserved.
9 151 40417944.
blished by Elsevier Ltd. All rights reserved.
l s 2 8 ( 2 0 1 2 ) 179–188
Fig. 1 – The principle of electro-spinning to produce mats ofPLA-fibers.
180 d e n t a l m a t e r i a
1. Introduction
For treatment of periodontal disease, there is a need for anoptimal local drug delivery system since the widespread sys-temic administration of antibiotics might cause undesiredside effects or favor the development of resistances.
The use of antibacterial biomaterials becomes increas-ingly important in medical and dental science. Especially inthe field of conservative dentistry, the elimination of bacte-ria and plaque is foundational for effective treatment [1,2].For instance, the conventional treatment of periodontitis byscaling and root planning is advantageously accompaniedby the adjuvant administration of antibiotics [3–5]. Antibac-terial drug compounds can be applied by systemic or localadministration. Compared to systemic drug delivery the localadministration of drugs in periodontology is considered tobe more effective, since the pathogen-specific drug can beplaced directly in the periodontal pocket achieving effectiveconcentrations. In addition the risk of undesired side effectscaused by high systemic doses or resistance developmentcan be reduced [6,7]. For effective elimination of pathogenicbacteria, the antibiotic agent has to be available in the peri-odontal pocket in adequate concentrations for a sufficientlylong period of time. It is therefore necessary to use localdelivery systems that control the release of their agents andguarantee lasting drug concentrations in the pocket in spiteof high sulcular fluid rates.
Non-resorbable drug carrier systems such as tetracycline-loaded fibers are placed from seven up to ten days inthe periodontal sulcus. In this period concentrations up to1300 �g/ml in the sulcus fluid can be maintained. However,the insertion of such non-resorbable fibers is time consum-ing and when their removal is required this incurs the risk oftissue damage.
Many resorbable drug delivery systems were developedduring recent decades, such as drug loaded hydroxypropy-lcellulose films [8], which were first described by Noguchiet al. (1984), or drug carrying gels such as Elyzol® (DumexGmbH, Bad Vilbel, Germany) dental gel, based on meltedglycerol mono-oleate [9–12]. However, also for these sys-tems, the periodontal milieu often poses the major problemthat the required period of drug exposure (7–10 days)cannot be achieved [9,13]. Also in the field of periodon-tal surgery – as in the transplantation of a mucousmembrane [14] – resorbability of the scaffold materialis important to avoid inflammatory effects and surgicalremoval.
Therefore the aim of this study was to investigate aresorbable drug reservoir, which releases essential amountsof its ingredients within an adequate period of time. Apossible drug delivery system based on electrospinning ofpolylactide was developed whereby mats of electrospun fiberscontaining the antibiotic metronidazole (MNA) were gen-erated having a large surface area per volume ratio. Fibermats incorporating different proportions of MNA (from 0.1 to
40.0%, w/w) were created to investigate release characteris-tics and determine the concentrations necessary for effectiveantibacterial action when placed in an appropriate host envi-ronment.2. Materials and methods
2.1. Electrospinning
Poly(l-lactide-co-d/l-lactide) 70/30 (Resomer LR 708,Boehringer Ingelheim, Germany) was used for electro-spinning (Fig. 1). A weight-average molecular weight of1.5 × 106 g mol−1 was determined for the polymer by gelpermeation chromatography using CHCl3 as solvent andpolystyrene as external standard. All solvents used for elec-trospinning purposes were of HPLC grade (Sigma–Aldrich,Germany). Micronized 2-(2-methyl-5-nitro-1H-imidazol-1-yl)ethanol (Metronidazole, MNA, Ph Eur 6.0 specification)was purchased from FARGON GmbH & Co. KG (Barsbüttel,Germany). Tetrahydrofuran, chloroform, dichloromethane,and acetone were tested for their suitability to dissolve poly(l-lactide-co-d/l-lactide) (PLA) and MNA for its subsequent usefor electrospinning. Acetone was chosen because of its goodsolubilizing of both PLA and MNA, its low boiling pointand its established use in dental adhesive applications. Wedetermined that a 3–5% (w/w) polymer solution, dependingon the MNA content, was necessary to spin the copolymerunder the conditions described below, to obtain similar fiberdiameters. A homogeneous solution was prepared by slowstirring of appropriate amounts of PLA and MNA in acetone atroom temperature for 3 h using a magnetic stirrer at 250 rpm.The obtained clear and viscous solutions were transferreddirectly into a 5 ml plastic syringe. The PLA/MNA mixture wasthen deployed in the electrospinning process using a customdesigned electrospinning apparatus. This incorporated anadjustable high-voltage power supply (ESV-100; IngenieurbüroG. Fuhrmann, Leverkusen, Germany) and an infusion pump(LA-100, Landgraf Laborsysteme, Germany). The syringe wasconnected by a 35 cm PTFE tube to a stainless-steel straight-end hollow needle (0.4 mm) under conditions adapted fromthose that have been previously described [15,16]. A mirroredglass surface (20 cm × 20 cm; glass thickness 2 mm) was used
as the electrically grounded plate to collect the drug loadedfibrous mat. The needle was connected to the ESV-100 DCpower supply adjusted to 20 kV. The syringe was mounted
d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 179–188 181
Table 1 – Applied materials, process parameters and mean diameters of the electrospun MNA-loaded PLA-fibers[compare with Fig. 5].
PLA (mg) MNA (mg) MNAconc.
(w/w) %
Acetone (g) PLAconc.(w/w) %
Target distance(cm)
Relativehumidity (%)
Temp.(◦C)
Mean fiberdiameter (�m)
200 0.2 0.1 3.81 5.3 16 29 25.9 1.20203 1.1 0.5 3.84 5.3 16 29 26.5 1.06198 2.0 1.0 3.81 5.2 16 29 26.4 1.15190 10.1 5.0 3.81 5.0 16 29 26.9 0.89
85 9.5 10 3.06 2.8 18.5 30 25.9 0.74
vfdard(Uaort2
2
2TeZaiwewo
2TScetv
2T1hdpT
101 25.0 20 3.03 3.3
132 57.0 30 2.98 4.4
133 89.0 40 2.94 4.5
ertically against the collector and the sample solution wased to the nozzle at a constant flow rate of 1.5 ml h−1. Theistance between the needle tip and the mirror was definedt 16–18.5 cm. The formation of smooth fibers with diametersanging from 0.6 to 1.2 �m was observed. Diameters wereetermined by software analysis of microscopic images
Image-Pro Plus 5.0, Media Cybernetics Inc., Silver Spring, MD,SA) and scanning electron microscopy (SEM). The appliedmounts of PLA, MNA and solvent, together with parametersf the electrospinning process and mean diameters of theesulting fibers are shown in Table 1. The overall area ofhe obtained electrospun fiber matrices was approximately5 cm2, and the thickness ranged between 70 and 100 �m.
.2. Material characterization
.2.1. Scanning electron microscopy (SEM)he drug-loaded fiber morphology was examined by field-mission scanning electron microscopy (SEM) (Supra 55VP;eiss, Oberkochen, Germany). For SEM, a Si wafer was useds substrate for the fibers and Au was sputtered on the spec-mens to ensure sufficient electrical conductivity. The imagesere taken using an InLens-detector with 5 keV excitation
nergy. The aim of this microscopic analysis was to determinehether the MNA concentration influenced the fiber structurer diameter.
.2.2. Energy dispersive X-ray spectroscopy (EDX)he SEM was equipped with an EDX-system (Quantax withi(Li)-detector, Bruker, Berlin, Germany). The fiber mats wereoated with evaporated carbon. For the measurements, anxcitation energy of 2 keV was used. Spectra were taken fromhe crossed fiber regions of the mat to get a larger excitationolume for EDX.
.2.3. Differential scanning calorimetry (DSC)hese experiments were performed on dry fiber mats in a Pyris
apparatus (PerkinElmer). The samples were subjected to a◦ ◦ ◦
eating scan from 25 C up to 200 C at 10 C/min. After coolingown to 25 ◦C at a rate of 10 ◦C/min, another heating run waserformed from that temperature up to 200 ◦C at 10 ◦C/min.he first scans are those analyzed here.
18.5 32 21.6 0.6418.5 30 23.5 0.9818.5 32 22.1 1.06
2.3. Generation of aliquots
2.3.1. PBS aliquotsTo evaluate the release of MNA from fibers placed in liquidmedia and its antibacterial efficiency, aliquots were producedfor HPLC analysis and agar diffusion tests. Due to the varyingfiber density, the weight of the individual fiber mats contain-ing different percentages of MNA (0%, 0.1%, 0.5%, 1%, 5%, 10%,20%, 30%, 40%, w/w) had to be standardized to 2 mg (±100 �g).Therefore rectangular parts of appropriate sizes (4–100 mm2)were cut of each mat with a scalpel and weighed with amicroscale (Genius ME215S; Satorius, Germany). Afterwardsthese weighted parts from the fiber mats were placed into 24-well plates and exposed to 2 ml of PBS (phosphate bufferedsaline/Invitrogen, Germany). Aliquots for study were removedon successive days (1–7) and weeks (day 14, 21 and 28) forHPLC analysis and agar diffusion tests. After each aliquotremoval the fiber mats were rinsed with 1 ml of PBS and wettedwith new liquid medium. For the investigation of initial MNArelease characteristics, another set of aliquots was produced.Therefore MNA loaded fiber mats (1%, 10%, 20% and 40%, w/w)of 2 mg were rinsed with 500 �l PBS from each side.
2.3.2. DMEM aliquotsFor cytotoxicity tests (MTT-test, Neutral Red uptake) usinghuman gingival fibroblasts, aliquots from fiber mats with dif-ferent MNA concentrations (0.1–40.0%, w/w) were generated asdescribed above for PBS aliquots. Instead of PBS, DMEM (Dul-becco’s Modified Eagle Medium; Invitrogen, Germany) with10% of fetal calf serum was used as aliquot medium. The plateswere cultivated at 37 ◦C with 5% CO2.
All aliquots were stored in Eppendorf tubes at −20 ◦C untilthey were used for further tests.
2.4. Drug release profiles
2.4.1. HPLCDetermination of MNA concentrations in the test aliquotswas carried out by high-performance liquid chromatography(HPLC). The HPLC system consisted of a Shimadzu systemwith a column oven CTO-10AC, a series pump LC-10AT, a sol-vent degasser DGU-14A, an autosampler, a dual wavelength
detector SPD-10A and Shimadzu CLASS-VP v 5.0 software.A LiChrosorb R18 column, 5 �m, 25 mm × 4.0 mm (GöhlerHPLC-Analysentechnik, Chemnitz, Germany) was used. Theseparation was carried out under isocratic elution with
l s 2
182 d e n t a l m a t e r i adouble distilled water/acetonitrile (95:5, v/v) at a flow rateof 1 ml min−1 with UV detection at 254 and 320 nm whichwas applied for analysis. Acetonitrile was HPLC grade andpurchased from Fisher Scientific (Fisher Scientific GmbH,Schwerte, Germany). The column temperature was ambientand an injection volume of 10 �l was used. The buffer solu-tions of drug eluting experiments were stored at 4 ◦C untilthose were used.
2.5. Antibacterial efficacy
2.5.1. Bacterial strainsFor this study three pathogenic periodontal bacterial strainswere used: Fusobacterium nucleatum (ATCC 10953/DSM 20482),Aggregatibacter actinomycetemcomitans (NCTC 9710/DSM 8324)and Porphyromonas gingivalis (ATCC 33277/DSM 20709). Forthe following test procedures, suspension cultures of thesebacterial strains were cultivated in nutrient solution (Oxoid,Germany) that was enriched with vitamin K for 24 h at37 ◦C under anaerobic conditions. After centrifugation of thebacterial strains they were rinsed twice with PBS and in phys-iological saline resuspended to an optical density of 0.1 at640 nm, that equals a bacterial density of 108 bacteria/ml.
2.5.2. Agar diffusion assayThis method of testing was prepared to evaluate the antibac-terial efficiency of the test aliquots. 100 �l of each bacterialsuspension (optical density: 0.1) was spread onto a Petri dishwith Schaedler agar (Oxoid, Germany) enriched with 1% ofvitamin K and 10% of sheep blood. Holes of 8 mm in diam-eter were punched-out of the agar and filled with 100 �l ofeach test aliquot. Aliquots from fibers without MNA were usesas negative control. After incubation time of 48 h at 37 ◦C thediameters of the inhibition zones were measured. Six speci-mens were tested for each aliquot.
2.6. Statistical analysis of antibacterial action
The diameters of affected areas can vary with concentrationand time for each of the three bacterial species investigated.This dependence was analyzed statistically, as follows. A two-way analysis of variance (ANOVA) was conducted using SPSSstatistical software (IBM; version 19), by means of univariateanalysis in a general linear model, to determine F-ratios forconcentration and time, and their interaction, for each speciesseparately. This was followed by application of the post hocScheffé test to determine homogenous subsets (alpha = 0.05) forboth concentration and time.
2.7. Cytocompatibility
2.7.1. Cell culturesFor cytotoxicity tests human gingival fibroblasts (HGFs) wereused (Ethics Commission Jena: 1881-10/06). These cells were
cultivated in DMEM with 10% of fetal calf serum and 0.1% ofAAS (antibiotic antimycotic solution) at 37 ◦C with 5% of CO2.To investigate the cytotoxicity of the fiber mats the neutral reduptake test and the MTT-test were used.8 ( 2 0 1 2 ) 179–188
2.7.2. Neutral red uptake testHere, HGFs were pipetted into 96-well microplates with acell density of approximately 10,000 cells per well and cov-ered with 100 �l of the test aliquot (prepared as describedabove). After two days of incubation time (37 ◦C, 5% CO2) theviability of the cells was analyzed by measuring their neu-tral red (Sigma–Aldrich, Germany) uptake using a microplatereader (Lambda Scan 200, BmT GmbH, Meerbusch-Osterath,Germany) with wavelength of 540 nm.
2.7.3. MTT-testTo evaluate the influence of the test aliquots on the viability ofthe HGFs, the cells were treated with the water-soluble dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium-bromide(MTT) (Roche, Germany). Due to the glycolysis equivalenttransformation of the colorant MTT, the viability of the cellscould be analyzed by measuring the optical density.
For both cytotoxicity tests DMEM was used as negative con-trol substance. Positive control tests already existed with HGFsthat were performed previously at the University of Jena, asalready published [17,18].
2.7.4. Direct exposure of cells on MNA-loaded fiber matsFiber mats containing different percentages of MNA (0%, 1.0%,5.0%, 10.0%, 20.0% and 40.0%, w/w) were clamped into Cell-Crown (Scaffdex Oy, Finland) cell culture inserts and sterilizedwith ethanol (70%, v/v). After rinsing with 2 ml of PBS, a cellsuspension (10,000 cells/cm2) was applied onto each fiber matthat was then covered with 2 ml of DMEM medium. After thecultivation period of 48 h the fiber mats were taken out the cellculture inserts and after rinsing with PBS stained with a live-dead colorant (12 ml of PBS + 12 �l of fluorescein diacetate [19](vital) + 16 �l ethidium bromide [EtBr] (avital)) and evaluatedmicroscopically.
3. Results
3.1. Morphology of the electrospun fibers
The drug loaded fiber mats showed a smooth structure of thesingle fibers (Fig. 2A). The amount of MNA incorporated in thefibers had a distinctive influence on their diameters, possiblycaused by changes in viscosity. Mean diameters of the fibersrange between 0.64 �m and 1.20 �m (Table 1; Fig. 5). Fibersloaded with 20.0% MNA (w/w) featured the smallest mean fiberdiameters while fibers with higher and lower MNA concen-trations exhibited larger diameters. SEM images of the fibermats, that were cultivated in PBS for 28 days for the prepara-tion of aliquots, showed neither signs of fiber break-up nor ofdisruption of the fiber mat assembly itself (Fig. 2B).
3.1.1. EDXEDX was measured over an area of 50 �m × 50 �m on the fibers
containing 40.0% (w/w) MNA (Supplementary InformationFigure SI). It showed clearly the nitrogen peak between 0.35and 0.45 keV. Besides this signal, larger signals of C and O werealso detected mainly arising from the PLA matrix.
d e n t a l m a t e r i a l s 2 8
Fig. 2 – SEM images of electrospun MNA-loaded PLA fiberm
3FP4p
released effective concentrations within the first 24 h to inhibit
ats: (A) before and (B) after 28 days in PBS.
.1.2. DSCig. 3 shows the DSC thermograms of pure PLA and
LA/metronidazole mixtures containing MNA of 10.0, 30.0 and0.0% (w/w). A single glass-transition step as shown for theure PLA occurs in each sample. With addition of MNA aFig. 3 – DSC thermograms of pure PLA and PLA/MNA mix
( 2 0 1 2 ) 179–188 183
second phase transition was detected in the range of themelting point of MNA of 159–163 ◦C, indicating the presenceof crystallized MNA. The endothermic peak increased withhigher concentrations of MNA, consistent with crystallizationof MNA within the electrospun fibers.
3.2. Drug release profiles
3.2.1. HPLCFig. 4 shows the cumulative drug release of fiber mats loadedwith different percentages of MNA (1.0%, 5.0%, 10.0%, 20.0%,30.0% and 40.0%, w/w). It is evident that all fibers released atleast 20% of their total amount of MNA within the first 48 h.The inverse association between lowest mean fiber diame-ter and maximal initial MNA release by rinsing is apparentin Fig. 5. Mats of fibers with small mean diameters (20.0%and 10.0% MNA, w/w) were characterized by a fast initial drugrelease. For example, the fiber mats loaded with 20.0% MNAreleased 21% of their total drug quantity by immediate rinsingand 43% within the first 48 h in PBS. By contrast, fibers withlarger mean diameters (1.0% and 40.0%, w/w) released onlysmall initial drug amounts when rinsed.
After a fast initial release phase during the first two days,the following days showed a more linear release of activeagent. Fig. 6 shows the cumulative release of MNA from thefibers loaded with 1.0% MNA (w/w) in a plot using the squareroot of time as x-coordinate.
3.3. Antibacterial efficacy
3.3.1. Agar diffusion assayThis test was conducted to analyze the antibacterial effec-tiveness of the aliquots from drug loaded fibers containingdifferent concentrations of MNA on F. nucleatum, A. actino-mycetemcomitans and P. gingivalis (Supplementary InformationFigure SII).
F. nucleatum: Fibers containing 1.0% (w/w) MNA or more
the growth of F. nucleatum. Aliquots with antibacterially effec-tive drug concentrations were released from day 6 up to day28 by the fibers loaded with 40.0% (w/w) MNA.
tures containing 10.0%, 30.0% and 40.0% MNA (w/w).
184 d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 179–188
Fig. 4 – Drug release profiles of PLA fibers loaded with different (w/w) percentages of MNA: (A) 1.0%, 5.0% and 10.0%; (B)20.0%, 30.0% and 40.0%. The initial (0 day) MNA release values were determined by immediate rinsing of the surfaces of the
d 40
fiber mats, for those fibers loaded with 1.0%, 10.0%, 20.0% anA. actinomycetemcomitans: Only aliquots of the first day,
from the drug loaded fibers with 10.0% (w/w) MNA ormore, and aliquots of the second day from fibers with40.0% (w/w) of MNA contained sufficient amounts ofFig. 5 – Variation of PLA fiber characteristics with MNAconcentration. Upper section: mean fiber diameters. Lowersection: initial MNA release after immediate rinsing of thefiber mats. Minimal fiber diameter, at 20% MNA,corresponded to maximal MNA release.
.0% (w/w) MNA.
drug to create inhibition zones up to 26 mm in diame-ter.
P. gingivalis: Aliquots of the first day, from fibers loaded with0.1% up to 40.0% (w/w) of MNA, inhibited the growth of P. gingi-valis. Effective amounts of drug were released from the fibersloaded with 5.0% (w/w) of MNA or more up to the 6th day,fibers with 30.0% and 40.0% (w/w) of MNA up to the 28th day.The summed areas of the inhibition zones that were measuredat day 1–7, day 14, 21 and day 28, as shown in Fig. 7, indicatethe effectiveness of the aliquots against the tested bacterialstrains.
3.4. Statistical analysis of antibacterial action
Highly significant variances were observed, for each species,in both concentration and time (Table 2). Some interactioneffects were also apparent between concentration and time.Homogenous subsets were observed for the concentrations
and times shown in Table 3.Fig. 6 – Cumulative drug release profile of the fibers loadedwith 1.0% (w/w) MNA with the square-root of time.
d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 179–188 185
Table 2 – The F-statistics for the dependent variables: concentration and time, plus their interaction, for theinhibition-zone diameters of three bacterial species.
Bacterial species F. nucleatum A. actinomycet P. gingivalis
F-ratio: concentration 316.79 729.49 1100.29F-ratio: time 348.59 2774.27 489.11F-ratio: conc. × time (interaction term) 28.70 503.05 27.90
Fig. 7 – Summed areas of inhibition zones for threebacterial species, produced by aliquots of PLA fibers loadedwith MNA, as a function of MNA concentration. Standardd
3
Uoetocem
H
Table 3 – Lists of homogenous subsets of inhibition-zonediameters, depending upon the variables: concentrationand time, for three bacterial species.
Bacterial species F. nucleatum A. actinomycet P. gingivalis
Concentration 0.1–1.0 0.1–5.0 0.1–0.55.0–20.0 10.0 1.0
30.0 20.0 5.0–10.040.0 30.0 20.0
40.0 30.040.0
Time 1.0 1.0 1.02.0 2.0 2.03.0 3.0–28.0 3.0–5.04.0 4.0–7.0
Fcat
eviations were small but are indicated, where possible.
.5. Cytocompatibility
sing the neutral red uptake test, the number of living cellsf a negative control sample was compared to the onesxposed to the aliquots (Fig. 8). It was evident that none ofhe aliquots from MNA-loaded fibers with different amountsf MNA (1.0–40.0%, w/w) caused a reduced number of livingells. Also the results of the MTT test (Fig. 9) showed that HGFsxposed to aliquots from the fibers did not impair cell viability
ore than 8%.Fig. 10 shows images from the microscopic analysis of theGFs that were exposed directly onto the surface of the drug
ig. 8 – HGF cell viability (%), measured by neutral red uptake, aftontacting test fibers from 1 up to 28 days. The cells were incubaliquot number (d): (A) for MNA concentrations 1.0–10.0% and (B)aken from the fiber mats on successive days (1–7) and then for s
5.0–28.0 14.0–21.021.0–28.0
loaded fiber mats with different concentrations of MNA andcolored with FDA and EtBr. The fibroblasts are green coloredby FDA and exhibit their typical structure when adhered toa surface. This clearly suggests the biocompatibility of thefibers. In summary, these investigations have shown that nei-ther the test aliquots nor the drug-loaded fibers themselveshad a negative influence on the HGF cells.
4. Discussion
Electrospinning has proven itself as a very effective methodfor the fabrication of drug loaded fibers. For three main rea-sons poly(l-lactide-co-d/l-lactide) (70/30) was used as matrixmaterial: First, it is a bioresorbable material; second, it fea-
tures suitable mechanical stability and third, it is approvedfor medical purposes. The form stability of the mats as wellas their constituent fibers (Fig. 2) in a liquid medium such aser exposure to aliquots of DMEM media—taken from mediated for 2 days in each case. Viability is plotted versus
for MNA concentrations 20.0–40.0%. DMEM aliquots wereuccessive weeks (days 7–28).
186 d e n t a l m a t e r i a l s 2 8 ( 2 0 1 2 ) 179–188
Fig. 9 – HGF cell viability (%), measured by MTT transformation, after exposure to aliquots of DMEM media—taken frommedia contacting test fibers from 1 up to 28 days. The cells were incubated for 2 d in each case. Viability is plotted versusaliquot number (d): (A) for MNA concentrations 1.0–10.0% and (B) for MNA concentrations 20.0–40.0%. DMEM aliquots were
for s
taken from the fiber mats on successive days (1–7) and thenPBS over a test period of 28 days is an advantage that supportstheir clinical usability in the periodontal sulcus. A too-rapiddegradation of the PLA fibers to lactic acid could lead to anacidosis and interfere with wound healing. The mechanicalstability of the fiber mats is also essential for to resist themechanical stress caused by the high production of sulcusfluid [9]. The values obtained from the HPLC measurementsfor the fibers loaded with 1.0% MNA (w/w) are very close to alinear regression with square-root of time, as seen in Fig. 6.This suggests that the liberation of MNA at this concentration
is by a diffusion process. The major impact of the fiber diam-eters on the initial MNA release is demonstrated in Fig. 5. Ithas been shown that fibers loaded with 20.0% MNA have theFig. 10 – Human gingival fibroblasts (HGFs) after direct exposurepercentages of MNA: (A) 0%; (B) 1.0%; (C) 5.0%; (D) 10.0%; (E) 20.0%structures are apparent in all cases.
uccessive weeks (days 7–28).
smallest mean fiber diameter while fiber diameters increasewith higher and lower MNA concentrations. Since the fibermats were standardized by weight, mats from fibers with smalldiameters feature larger real surfaces. This explains the fact,that fiber mats loaded with 10.0% and 20.0% MNA (w/w) releasehigh drug amounts via rinsing with PBS. With higher degree ofcrystallization (Fig. 3) and decreased active surface areas, fibermats loaded with a higher concentration of MNA (e.g. 40.0%,w/w) release initially only minor amounts of drug by rinsing.From the third day all tested fiber mats show a more linear
release profile of MNA which is probably consistent with drugrelease by diffusion. Changing parameters of the electrospin-ning process could allow modification of the diameters of theto the surface of electrospun PLA fibers containing w/w; (F) 40.0%. Scale-bars are 100 or 50 �m. Typical healthy
2 8
fid
srgoMpi[avrTswrztglviifis4
bagsoceMe
tnbpoa
PentotMcM4
tcs
r
d e n t a l m a t e r i a l s
bers whereby the drug release could be controlled furtherepending on the application purpose.
Recent investigations, published by Soscia et al. in 2010,how that similar drug loaded electrospun fibers incorpo-ating the antibiotic chloramphenicol are able to reduce therowth of selected bacterial strains by more than 93% with-ut showing a cytotoxic impact on mammalian cells [20].NA was chosen for this study since its efficiency against
eriodontal bacteria was shown in several clinical stud-es and is the drug of choice for many periodontologists21–23]. The minimal inhibitory concentrations (MIC) of MNAgainst A. actinomycetemcomitans (8–>256 �g/ml) and P. gingi-alis (0.031–1.5 �g/ml) can be found in the literature and are aeliable standard to evaluate the sensitivity to MNA [22,24–26].hese values correlate well with the results of the agar diffu-ion assay of this study (Fig. 7). The sensitivity of F. nucleatumas expected to be similar to P. gingivalis since both bacte-
ial stains are strictly anaerobic. Measurements of inhibitionones confirm these expectations by showing that the antibac-erial effect of MNA against F. nucleatum is slightly below P.ingivalis. Also it has been noted that the drug release lastsonger then the chosen time period of 28 days for release obser-ation. Longer time periods of drug release will be consideredn further studies. It is interesting to note that the antibacterialmpact against F. nucleatum and P. gingivalis of the aliquots frombers with 5.0%, 10.0% and 20.0% (w/w) of MNA induce similarized summed areas of inhibition, while the aliquots from the0.0% (w/w) MNA-loaded fibers are significantly higher (Fig. 7).
Since A. actinomycetemcomitans is a facultative anaerobicacterium, it is very resistant to MNA. Nevertheless, highmounts of MNA released from fibers are able to inhibit therowth of this bacterial strain, as reflected in the agar diffu-ion assays. For an effective elimination of the most relevantbligate anaerobic periodontal pathogenic bacterial strains,oncentrations of 1 �g/ml MNA in the sulcus fluid are nec-ssary. In this in vitro study it was shown that the releasedNA concentrations from the fiber mats represent therapeutic
fficient doses.The antibacterial effect of the released MNA on F. nuclea-
um and P. gingivalis even after 28 days shows that there is noeed to slow down the drug release by using more hydropho-ic MNA derivatives such as MNA benzoate. MNA benzoate, aro-drug of MNA is used for example as the active ingredientf the biodegradable ELYZOL® gel to extend the duration of thentibacterial effect [10,27].
Looking on the intended clinical use of the MNA-loadedLA fibers, the main advantage of such a local drug deliv-ry systems is the relatively small amount of drug that isecessary to achieve effective concentrations in the periodon-al sulcus. Moreover, these smaller amounts avoid the riskf systemic side effects as well as the development of resis-ances. For example, a highly loaded fiber mat with 40.0% ofNA (w/w) of the size 50 mm × 50 mm and a weight of 35 mg
ontains 14 mg of MNA. In comparison, the total amount ofNA when administered systemically adds up to 8400 mg (3×
00 mg daily over 7 days).
The cytotoxicity tests conducted demonstrated that nei-her the MNA nor the polylactide fiber-matrix show anyytotoxic effect on HGFs. The structure of the cells (Fig. 10)howed that regardless of the MNA concentration the
( 2 0 1 2 ) 179–188 187
fibroblasts attached onto the surface of the fiber mats.This observation leads to the conclusion that, despite itshydrophilicity, the percentage of MNA (1.0% up to 40.0%,w/w) might have only a minor influence on cell adhesion.Good cell adhesion might be another advantage of the fibermats enabling them to remain within the periodontal pocketdespite high sulcus-fluid rates.
5. Conclusion
This study has shown that MNA can be integrated intothe matrix of electrospun polylactide fibers. Despite thepresent encouraging results, optimization of the syntheticmatrix material, by for example using polymer blends, couldlead to further improvements of the release characteristics.Possible materials include polylactones such as poly-epsilon-caprolactone, as used for electrospinning by the group ofZamani et al. [28]. Also the insertion of different antibioticagents into the polylactide matrix of the fibers should bethe main focus for further in vitro studies. For example theuse of MNA and amoxicillin, also known as the “Van Winkel-hoff Cocktail” is a promising combination [29,30]. Sustainedrelease of MNA and its antibacterial efficiency on periodontalpathogenic bacteria, as well as the high level of biocompat-ibility shown in this study, are beneficial characteristics forthe usability of electrospun polylactide fibers as a drug-releasesystem for local periodontitis treatment.
Acknowledgements
We appreciate the financial support of the EU (EFRE) andthe Thuringian Ministry for Economy, Labor and Technology(2007FE0115). We greatly acknowledge Dr. M. Frigge for SEMinvestigations and M. Herz for DSC measurements (both fromINNOVENT).
MR is very thankful for the scholarship Deutschland-stipendium from the German Federal Ministry of Education andResearch as well as Heraeus.
DCW gratefully acknowledges the support of the Alexandervon Humboldt Foundation (Bonn, Germany) by a HumboldtResearch Award.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,in the online version, at doi:10.1016/j.dental.2011.12.006.
e f e r e n c e s
[1] Jandt KD, Sigusch BW. Future perspectives of resin-baseddental materials. Dent Mater 2009;25:1001.
[2] Balamurugan A, Balossier G, Laurent-Maquin D, Pina S,Rebelo AH, Faure J, et al. An in vitro biological and
anti-bacterial study on a sol–gel derived silver-incorporatedbioglass system. Dent Mater 2008;24:1343.[3] Sigusch BW, Güntsch A, Pfitzner A, Glockmann E. Enhancedroot planing and systemic metronidazole administration
l s 2
therapy to full-mouth scaling and root planing in the
188 d e n t a l m a t e r i a
improve clinical and microbiological outcomes in a two-steptreatment procedure. J Periodontol 2005;76:991.
[4] Sigusch B, Beier M, Klinger G, Pfister W, Glockmann E. A2-step non-surgical procedure and systemic antibiotics inthe treatment of rapidly progressive periodontitis. JPeriodontol 2001;72:275.
[5] Hanes PJ, Purvis JP. Local anti-infective therapy:pharmacological agents. A systematic review. AnnPeriodontol 2003;8:79.
[6] Kinane DF, Radvar M. A six-month comparison of threeperiodontal local antimicrobial therapies in persistentperiodontal pockets. J Periodontol 1999;70:1.
[7] Rams TE, Slots J. Local delivery of antimicrobial agents inthe periodontal pocket. Periodontol 2000 1996;10:139.
[8] Noguchi T, Izumizawa K, Fukuda M, Kitamura S, Suzuki Y,Ikura H. New method for local drug delivery usingresorbable base material in periodontal therapy. Bull TokyoMed Dent Univ 1984;31:145.
[9] Soskolne WA. Subgingival delivery of therapeutic agents inthe treatment of periodontal diseases. Crit Rev Oral Biol Med1997;8:164.
[10] Al-Mubarak SA, Karring T, Ho A. Clinical evaluation ofsubgingival application of metronidazole 25%, andadjunctive therapy. J Int Acad Periodontol 2000;2:64.
[11] Sato S, Fonseca MJ, Ciampo JO, Jabor JR, Pedrazzi V.Metronidazole-containing gel for the treatment ofperiodontitis: an in vivo evaluation. Braz Oral Res2008;22:145.
[12] Stellini E, Migliorato A, Mazzoleni S, Mottola A, Lombardi L,Favero GA. Topical treatment of peri-implantitis withmetronidazole dental gel 25%. Clinical analysis andmicrobiological control. Minerva Stomatol 2000;49:59.
[13] Shifrovitch Y, Binderman I, Bahar H, Berdicevsky I,Zilberman M. Metronidazole-loaded bioabsorbable films aslocal antibacterial treatment of infected periodontalpockets. J Periodontol 2009;80:330.
[14] Roccuzzo M, Bunino M, Needleman I, Sanz M. Periodontalplastic surgery for treatment of localized gingival recessions:a systematic review. J Clin Periodontol 2002;29(Suppl. 3):178.
[15] Kluger PJ, Wyrwa R, Weisser J, Maierle J, Votteler M, Rode C,et al. Electrospun poly(d/l-lactide-co-l-lactide) hybridmatrix: a novel scaffold material for soft tissue engineering.J Mater Sci Mater Med 2010;21:2665.
[16] Wyrwa R, Finke B, Rebl H, Mischner N, Quaas M, Schäfer J,et al. Design of plasma surface-activated, electrospunpolylactide non-wovens with improved cell acceptance. Adv
Biomater 2011:B165.[17] Sigusch BW, Völpel A, Braun I, Uhl A, Jandt KD. Influence ofdifferent light curing units on the cytotoxicity of variousdental composites. Dent Mater 2007;23:1342.
8 ( 2 0 1 2 ) 179–188
[18] Sigusch BW, Pflaum T, Völpel A, Schinkel M, Jandt KD. Theinfluence of various light curing units on the cytotoxicity ofdental adhesives. Dent Mater 2009;25:1446.
[19] Greenstein G, Polson A. The role of local drug delivery in themanagement of periodontal diseases: a comprehensivereview. J Periodontol 1998;69:507.
[20] Soscia D, Raof N, Xie Y, Cady NC, Gadre A. Antibiotic-loadedPLGA nanofibers for wound healing applications. AdvBiomater 2010:B83.
[21] Pavia M, Nobile CG, Bianco A, Angelillo IF. Meta-analysis oflocal metronidazole in the treatment of chronicperiodontitis. J Periodontol 2004;75:830.
[22] Poulet PP, Duffaut D, Lodter JP. Metronidazole susceptibilitytesting of anaerobic bacteria associated with periodontaldisease. J Clin Periodontol 1999;26:261.
[23] Silva MP, Feres M, Sirotto TA, Soares GM, Mendes JA, FaveriM, et al. Clinical and microbiological benefits ofmetronidazole alone or with amoxicillin as adjuncts in thetreatment of chronic periodontitis: a randomizedplacebo-controlled clinical trial. J Clin Periodontol2011;38:828.
[24] Kulik EM, Lenkeit K, Chenaux S, Meyer J. Antimicrobialsusceptibility of periodontopathogenic bacteria. JAntimicrob Chemother 2008;61:1087.
[25] Eick S, Pfister W. Efficacy of antibiotics againstperiodontopathogenic bacteria within epithelial cells: anin vitro study. J Periodontol 2004;75:1327.
[26] Eick S, Seltmann T, Pfister W. Efficacy of antibiotics to strainsof periodontopathogenic bacteria within a single speciesbiofilm—an in vitro study. J Clin Periodontol 2004;31:376.
[27] Salvi GE, Mombelli A, Mayfield L, Rutar A, Suvan J, Garrett S,et al. Local antimicrobial therapy after initial periodontaltreatment. J Clin Periodontol 2002;29:540.
[28] Zamani M, Morshed M, Varshosaz J, Jannesari M. Controlledrelease of metronidazole benzoate from polyepsilon-caprolactone electrospun nanofibers for periodontaldiseases. Eur J Pharm Biopharm 2010;75:179.
[29] Winkel EG, Van Winkelhoff AJ, Timmerman MF, Van derVelden U, Van der Weijden GA. Amoxicillin plusmetronidazole in the treatment of adult periodontitispatients. A double-blind placebo-controlled study. J ClinPeriodontol 2001;28:296.
[30] Sgolastra F, Petrucci A, Gatto R, Monaco A. Effectiveness ofsystemic amoxicillin/metronidazole as an adjunctive
treatment of aggressive periodontitis: a systematic reviewand meta-analysis. J Periodontol 2011,doi:10.1902/jop.2011.110432.