University of Groningen
Antibiotic-loaded poly(trimethylene carbonate) degradation, release and
staphylococcal biofilm inhibition
Kluin, Otto Samuel
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2016
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Kluin, O. S. (2016). Antibiotic-loaded poly(trimethylene carbonate) degradation, release and staphylococcal
biofilm inhibition [Groningen]: Rijksuniversiteit Groningen
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 18-06-2017
Asurface-erodingantibiotic
deliverysystembasedon
poly(trimethylenecarbonate)
KluinOS,VanderMeiHC,BusscherHJ,NeutD.
Biomaterials2009,30(27):4738-42.
ReproducedwithpermissionofElsevier
43
Biodegradable delivery systems that do not produce acidic compounds
during degradation are preferred for local antibiotic delivery in bone
infections in order to avoid adverse bone reactions. Poly(trimethylene
carbonate) (PTMC) has good biocompatibility, and is such a polymer. The
objectiveofthisinvitrostudywastoexplorethesuitabilityofPTMCasan
antibiotic releasing polymer for the local treatment of bone infections.
Degradation behaviour and corresponding release profiles of gentamicin
and vancomycin from slowly degrading PTMC168 and faster degrading
PTMC339 discs were compared in the absence and presence of a lipase
solution. Gentamicin release in the absence of lipase was diffusioncontrolled,whilevancomycinreleasewaslimited.SurfaceerosionofPTMC
only occurred in the presence of lipase. Both antibiotics were released in
high concentrations from PTMC in the presence of lipase through a
combination of surface erosion and diffusion. This illustrates the major
advantage of surface-eroding biodegradable polymers, allowing release of
largerantibioticmoleculeslikevancomycin.
44
Bone and soft tissue infections, such as osteomyelitis or the infected
diabetic foot are difficult to treat [Gristina et al. 1985; James et al. 2008].
Usually blood flow is poor under these pathological conditions, and
systemically applied antibiotics are unable to achieve the local
concentrations needed to eradicate the infecting bacteria, which are, in
addition,protectedbytheirbiofilmmodeofgrowth[Stewartetal.2001].In
order to treat these infections, local drug delivery systems are needed to
achieve effective levels of antibiotics. Nowadays, research on antibiotic
delivering carriers for the local treatment of infections is concentrated on
biodegradablepolymers[Kanellakopoulou&Giamarellos-Bourboulis2000;
Smith2005;Tayloretal.1994;Liuetal.2002].Biodegradablecarriersnot
onlysupplyahighlocaldrugconcentrationforafixedamountoftime,but
can also release their entire antibiotic content [Kanellakopoulou &
Giamarellos-Bourboulis2000].Moreover,thereisnonecessityforasecond
surgical intervention to remove the carrier. The most frequently
investigated biodegradable antibiotic carriers consist of polylactic acid
(PLA) and/or polyglycolic acid (PGA) [Kanellakopoulou & GiamarellosBourboulis 2000]. However, the use of PLA and PGA implants in bone is
controversial: bone resorption can develop as a result of a decrease in pH
whenthepolymerdegrades[Tayloretal.1994].Therefore,thesepolymers
maynotbeidealforthelocaltreatmentofboneinfections.Forthispurpose,
carriers that do not produce acidic degradation products are highly
preferable. This requirement can be fulfilled by the use of biodegradable
and biocompatible poly(trimethylene carbonate) (PTMC) [Albertsson &
45
Eklund 1995; Pêgo et al. 2003], which degrades in water, CO2 and
propanediol.
EnzymesplayanimportantroleintheinvivodegradationofPTMC,
whichcanbesimulatedinvitrobytheuseofalipasesolution[Zhangetal.
2006b].Infact,intheabsenceoflipase,PTMChardlydegradesinvitro,and
this sharply contrast with the high degradation rates observed in vivo
[Zhang et al. 2006b]. Degradation of PTMC is characterised by surface
erosion,whichoccursfasterforPTMCofhighermolecularweight[Pêgoet
al. 2003; Zhang et al. 2006b]. Drug release kinetics from surface-eroding
polymersshowacloserelationshiptothepolymerdegradationrate[Langer
& Peppas 2003], but whether drug release kinetics from PTMC are
influencedbyitsmolecularweightneedstobedetermined.
Antibiotics incorporated in local delivery systems ideally
demonstrate broad and sufficient antibacterial efficacy, marked water
solubilityfacilitatingitsrelease,andchemicalandthermalstability[Wahlig
& Dingeldein 1980]. Aminoglycosides in general and gentamicin in
particular are preferred antibiotics, both from a bacteriological and
physico-chemical point of view. However, the increasing resistance to
gentamicin of staphylococci in bone infections [Sheftel et al. 1985;
Tentolouris et al. 1999] necessitates the use of different antibiotics.
Vancomycin is the antibiotic of last resort used for the treatment of
clinicallyresistantbacteria,likemethicillin-resistantStaphylococcusaureus
(MRSA). Nevertheless, the use of vancomycin in non-biodegradable drug
carriers like bone cement is limited because it demonstrates poor release
[Klekampetal.1999],duetoitshighmolecularweight,impedingdiffusion
through the polymer matrix [Liu et al. 2002]. PTMC, however, is not
46
expectedtohamperthereleaseofhighmolecularweightdrugsbecauseof
itsdegradationbysurfaceerosion.
The aim of this study is to explore the suitability of biodegradable
PTMC of different molecular weights as an antibiotic delivery system,
focusingonthedegradationprocessandcorrespondingantibioticreleaseof
gentamicinandvancomycin.
47
Materials
High purity, polymerisation grade 1,3-trimethylene carbonate (TMC) was
purchased from Boehringer Ingelheim (Ingelheim, Germany). The
antibiotics gentamicin sulphate (molar mass 723 g/mol) and vancomycin
hydrochloride (molar mass 1486 g/mol) were purchased from Sigma–
Aldrich (St. Louis, MO, USA). Lipase from Thermomyces lanuginosus
(EC3.1.1.3,minimum100.000units/g,theaqueoussolutioncontains2wt%
enzyme concentrate, 0.5 wt% CaCl2 and 25 wt% propylene glycol) was
purchased from Sigma (St. Louis, MO, USA) and used as received.
Tetrahydrofuran(THF),anhydroushexane,andanalyticalgradechloroform
werepurchasedfromMerck(Darmstadt,Germany),andusedasreceived.
Antibioticloadingofthepolymers
TwotypesofPTMCwithdifferentmolecularweightswereusedtoprepare
antibiotic-loadedpolymerfilms:lowmolecularweightPTMC(PTMC168,Mw
=168x103g/mol,Mw/Mn=1.15,intrinsicviscosityat27°Cinchloroform
2.2dl/g)andhighmolecularweightPTMC(PTMC339,Mw=339x103g/mol,
Mw/ Mn = 1.17, intrinsic viscosity at 27°C in chloroform 4.3 dl/g). The
polymers were synthesised, purified, and characterised as previously
described [Zhang et al. 2004]. Briefly, PTMC solutions were prepared by
48
dissolving 2 g of polymer in 75 mL THF. To these solutions, 0.20 g of
antibiotic (gentamicin sulphate or vancomycin hydrochloride) was added,
andafter5minofultrasonicmixingfollowedbymagneticstirringfor24h,
thesolutionscontainingtheantibioticdispersionwereprecipitatedintoan
excess of anhydrous hexane. The composite was then collected and dried
undervacuumatroomtemperature.
PTMC films with dispersed antibiotic particles incorporated were
prepared by compression moulding at 77°C, using stainless steel moulds
and a laboratory press (Carver Inc., Wabash, IN, USA). Of each polymer,
three separate films (approximate thickness 500 mm) containing either
gentamicin sulphate or vancomycin chloride were prepared. From these
films, discs with a diameter of 5 mm were punched. Films without
antibioticswerepreparedaswell.
The initial antibiotic contents of the discs were determined in
triplicate using a fluorescence polarisation immunoassay (AxSYM, Abbott
Laboratories, Abbott Park, IL, USA) after appropriate dilution. To this end,
thediscsweredissolvedin3mLchloroformandextractedfivetimeswith3
mLwatertominimisetheresidualantibioticinthechloroformtolessthan
5%. The aqueous fractions were collected, and their gentamicin or
vancomycincontentmeasured.
Degradationexperiments
To investigate the degradation behaviour, gentamicin-loaded PTMC168 and
PTMC339discswereincubatedin5mLphosphatebufferedsaline(PBS)orin
a lipase-containing solution at 37°C. The mass, thickness and diameter of
the specimens were determined before incubation and at predetermined
49
time points during the degradation process. After one day, the discs were
blotted dry on filtration paper, analysed and re-immersed in the
degradationmedia.Thisprocedurewasrepeatedafter2,3,7,10,14and28
days.Degradationexperimentswereperformedintriplicate.
Molecular weights were determined after 0, 2, and 7 days of
incubation using non-loaded PTMC168 and PTMC339 discs. Weight average
(Mw) and number average (Mn) molecular weights, molecular weight
distributions (Mw/Mn) and intrinsic viscosities of the polymers were
determined by gel permeation chromatography, as previously described
[Pêgo et al. 2003]. Changes in the surface structure of gentamicin-loaded
PTMCdiscsuponincubationinPBSandlipasesolutionswerevisualisedby
scanningelectronmicroscopy(SEM).After0,2,7and14days,thepolymer
specimenswererinsedwithdemineralisedwateranddriedusingfiltration
paper.Thesamplesweresputter-coatedwitha3nmthickgold/palladium
layerforexaminationat2.0kVusingaJEOL6301FfieldemissionSEM.
Antibioticrelease
PTMC168 and PTMC339 discs loaded with gentamicin or vancomycin were
immersedin5mLPBSorinlipasesolutionsandincubatedfor4weeksat
37°C.After1,2,3,7,10,14and28daysofincubation,150 mLaliquotsof
the media were taken, and the antibiotic concentrations were indirectly
determinedbymeasuringbacterialinhibitionzones[Joostenetal.2005].A
gentamicin-sensitive S. aureus A20734 (minimal inhibitory concentration
for gentamicin 0.50 mg/mL) was used to determine the concentrations of
gentamicinreleased,whileavancomycin-sensitiveS.aureus5298(minimal
50
inhibitory concentration for vancomycin 0.064 mg/mL) was used for
vancomycin concentration measurements. Inhibition zone measurements
werechosenfordeterminingtheantibioticrelease,inordertoensurethat
the antibiotics had retained their activity throughout the entire process of
PTMCpreparation.
Staphylococci were cultured aerobically from cryopreservative
beads (Protect Technical Service Consultants Ltd., United Kingdom) onto
blood agar plates at 37°C overnight. A 18 h preculture of both bacterial
strains was made in Tryptone Soya Broth (TSB). These precultures were
dilutedtoaconcentrationofapproximately108bacteriapermLinorderto
inoculate TSB agar plates with a sterile cotton swab. The thickness of the
agarwasabout5mm.Tenminutesafterinoculation,15mLofPBSorlipase
aliquots were transferred to the centre of each plate, and the plate was
subsequently incubated aerobically at 37°C. After overnight incubation,
clearareasaroundthepositionofthesampledropletindicatedtheabsence
ofbacterialgrowth.Thediametersoftheseinhibitionzonesweremeasured
intwoperpendiculardirectionsandrelatedwiththediametersofinhibition
zones around droplets of PBS or lipase solutions with known antibiotic
concentrations, in order to calculate the antibiotic concentration in the
aliquots.
Statisticalanalysis
Dataarepresentedasmeanvalueswithstandarddeviations.TheStudentttest for independent samples was used, and a 95% (p<0.05, two-tailed)
confidenceintervalwasconsideredsignificant.
51
Degradationexperiments
Initially,PTMCspecimenshadathicknessof646±88μm,adiameterof5.2
±0.1mm,andaninitialmassof16.7±1.0mg.BothPTMC168andPTMC339
didnotdegradeinPBS,anditsmassandthicknessremainedmoreorless
constant during the entire experimental period (see figure 1). Small initial
increaseswereobservedduetoswelling.PTMCinlipasesolutionsshowed
significant(p<0.05)decreasesinbothmassandthickness,whichwereboth
strongerforhighmolecularweightPTMC339thanforlowmolecularweight
PTMC168. During incubation, Mw and Mn changed somewhat for both
polymers,resultinginmoleculardistributionsMw/Mnof1.12and1.23for
PTMC339 and PTMC168, respectively. Since significant loss of mass and
decreaseinthicknessoccurredwhilethemolecularweightremainedmore
or less constant, the enzymatic break down of PTMC in lipase solutions
must be considered as a surface erosion process. Ignoring the initial
increaseinthicknessduetowateruptake,erosionratesof12mm/dayand
70mm/daycanbecalculatedforPTMC168andPTMC339,respectively.
Theerosionofthesurfaceofthegentamicin-loadedPTMCdiscswas
visualised by SEM. While in PBS no changes in surface structure could be
observed(micrographsnotshown),enzymaticdegradationofthepolymer
surfaceisalreadyseenafter2daysinlipasesolutions.Figure2showsthat
during the enzymatic degradation process, a significant number of pores
have been formed in the discs. This illustrates that the enzymatic surface
52
erosionisnotahomogeneousprocess.Notethatthesurfaceofthepolymer
matrix becomes markedly less porous for the lower molecular weight
PTMC168discsthanforthehighermolecularweightPTMC339discs.
!
relative mass
!!
1,5
1
0,5
0
relative thickness
0
10
20
0
10
20
30
1,5
1
0,5
0
time (days)
30
Figure 1: Relative mass and thickness during incubation of gentamicinloaded PTMC discs (with 7.4 wt% gentamicin) in PBS (light grey symbols)
or lipase (black symbols) solutions. Mass is relative to the one of dry
specimens16.7mgforPTMC168(opensymbols)andPTMC339discs(closed
symbols),whiletheinitialthicknessamounted646mm.Eachlineisalinear
fit of the experimental points. Error bars denote the SD over triplicate
measurementswithseparatelypreparedspecimens.
53
rated in the polymer matrices) was determined in PBS
as well as in lipase solutions (Fig. 4). The gentamicin
e incorporated per disc amounted 1252 ! 160 mg on
or 7.4 wt% and 8.3 wt% for PTMC168 and PTMC339 discs,
vely, while for vancomycin hydrochloride these values
delivery systems depend upon: (1) diffusion through the polymer
matrix; (2) matrix degradation and erosion; and (3) a combined
degradation/diffusion process. Thus, diffusion and degradation
govern the process of antibiotic release and the observed antibiotic
release profiles from these polymer matrices strongly depend on
PTMC168
PTMC339
day0
day2
Figure 2: Scanning electron micrographs of the surfaces of gentamicin-
nning electron micrographs of the surfaces of gentamicin-loaded PTMC discs before and after incubation in lipase solutions at 37 # C. Antibiotic-loaded PTMC films were
rom PTMC168 (left) and PTMC339 (right), the top images show the surfaces of the discs before exposure to the enzyme solution, while the bottom images depict the
ter 2 days of incubation. Arrows indicate gentamicin particles.
loaded PTMC discs before and after incubation in lipase solutions at 37°C.
Antibiotic-loaded PTMC films were prepared from PTMC168 (left) and
PTMC339 (right), the top images show the surfaces of the discs before
exposure to the enzyme solution, while the bottom images depict the
surfacesafter2daysofincubation.Arrowsindicategentamicinparticles.
54
Antibioticreleaseexperiments
Antibioticrelease(relativetothetotalamountsofantibioticsincorporated
in the polymer matrices) was determined in PBS (figure 3) as well as in
lipase solutions (figure 4). The gentamicin sulphate incorporated per disc
amounted1252± 160mgonaverage,or7.4wt%and8.3wt%forPTMC168
andPTMC339discs,respectively,whileforvancomycinhydrochloridethese
values were 1370 ± 235 mg, corresponding with 9.2 wt% and 10.0 wt%,
respectively. Generally, higher amounts of gentamicin and vancomycin are
releasedintoPBSfromPTMC168thanfromPTMC339(p<0.05).Analysisofthe
relative amounts of antibiotics released versus t1/2 (see also figure 3)
revealed a first-order release profile in PBS (linear correlation coefficients
R2 equal 0.88 and 0.87 for gentamicin from PTMC168 and PTMC339 discs,
respectively). Vancomycin is not released from PTMC168 according to firstorderkinetics(correlationcoefficient0.04,datanotshown),likelybecause
diffusion through the PTMC matrix is restricted. Whereas vancomycin
releasefromthehighermolecularweightPTMC339wasbelowdetection.
Enzymatic degradation of PTMC in a lipase-containing solution
(figure 4) caused a significant higher release (p<0.05) of gentamicin and
vancomycin than in PBS. Both antibiotics were released at significantly
(p<0.05) higher rates from PTMC339 than from PTMC168. Note, that within
theexperimentalaccuracy,releaseduringthefirst3daysfromPTMC339can
beconsideredtoincreaselinearlywithtimeforbothantibiotics(correlation
coefficients R2=0.99) according to a zero-order release profile. However,
release during the entire experiment also carries the characteristics of a
first-order release, and linear correlation coefficients R2 between the
amounts released and t1/2 exceed 0.86 (except for release of vancomycin
from PTMC168). This suggests that antibiotic release from PTMC under
55
enzymatic degradation is controlled by a combination of first-order and
zero-orderreleasekinetics.
40
Mt/M∞ (%)
30
0
0
10
2
4
time½ (days½)
6
20
30
time (days)
50
40
10
0
20
0
20
30
10
Mt/M∞ (%)
50
Mt/M∞ (%)
!!
30
20
10
0
0
10
20 time (days) 30
Figure 3: Cumulative antibiotic release expressed as a percentage of the
totalamountofgentamicin(topgraph,lightgreysymbols)andvancomycin
(bottom graph, light blue symbols) incorporated from antibiotic-loaded
discs prepared from PTMC168 (open symbols) and PTMC339 discs (closed
symbols)duringincubationinPBSat37°C.Gentamicinreleaseispresented
asafunctionoftimeaswellasoftime1/2(inset).Eachlineisalinearfitof
the experimental points. Error bars denote the SD over triplicate
measurementswithseparatelypreparedspecimens.
56
Mt/M∞ (%)
300
Mt/M∞ (%)
2
4
½
time (days½)
6
100
0
0
300
Mt/M∞ (%)
50
0
100
0
150
200
10
Mt/M∞ (%)
!!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!!
!
!
!
!
20
time (days)
30
150
100
50
0
200
0
2
4
time½ (days½)
6
100
0
0
10
20
30
time (days)
Figure 4: Cumulative antibiotic release expressed as a percentage of the
total amount of gentamicin (top graph, black symbols) and vancomycin
(bottom graph, blue symbols) incorporated from antibiotic-loaded discs
preparedfromPTMC168(opensymbols)andPTMC339discs(closedsymbols)
duringincubationinlipaseenzymesolutionsat37°C.Releaseispresented
asafunctionoftimeaswellasoftime1/2(inset).Eachlineisalinearfitof
the experimental points. Error bars denote the SD over triplicate
measurementswithseparatelypreparedspecimens.
57
The antibiotic release rates from biodegradable polymer based delivery
systemsdependupon:(1)diffusionthroughthepolymermatrix;(2)matrix
degradation and erosion; and (3) a combined degradation/diffusion
process. Thus, diffusion and degradation govern the process of antibiotic
release and the observed antibiotic release profiles from these polymer
matricesstronglydependontherelativeratesofdegradationanddiffusion.
Diffusionoccurswhenantibioticcompoundspassfromwithinthepolymer
matrix to its surroundings, and yields a first-order release kinetic. In firstorder release kinetics, the release rate decreases, since the antibiotic
compounds have to travel progressively longer distances and therefore
require a longer diffusion time to release. This is clinically undesirable,
because therapeutic effectiveness of antibiotics will not continue once
release rates fall below certain levels. In case of a combined
degradation/diffusion process, however, degradation of the polymer itself
enhancesthedrugreleasetoaconsiderableextent.
To ourknowledge, no literature exists describing antibiotic release
from PTMC, and although the idea of using PTMC as a drug carrier is not
new, release under surface-eroding conditions has never been evaluated
[Zhang & Zhuo 2005; Zhang et al. 2006a]. In the current study, release of
gentamicin and vancomycin from PTMC with different molecular weights
wascomparedinPBSandinalipasesolution,facilitatingsurfaceerosion.In
PBS, gentamicin release from PTMC168 was higher than from PTMC339 and
diffusion-controlledfrombothmatrices,sincepolymerdegradationdidnot
occur.WhereasgentamicinreleasefromPTMC339inPBSamounted4%in4
58
weeks, vancomycin release from PTMC339 stayed below detection. In fact,
vancomycin failed to diffuse out of PTMC, likely because this large
compoundistrappedwithinthepolymermatrix[Klekampetal.1999].
Polymer degradation can be divided into bulk degradation and
surfaceerosion.Inbulkdegradation,thepolymerdegradesthroughoutthe
matrix.Insurfaceerosion,materialislostfromthepolymersurfaceandthe
erosion rate is directly proportional to the external surface area. For drug
delivery applications, surface erosion is more desirable, as a surfaceeroding polymer can provide constant, zero-order and easily controllable
release rates. In vivo and in lipase solutions, PTMC degrades through
surfaceerosion,sincethethicknessofPTMCdiscsdecreasesparalleltoits
totalmassandadecreaseinmolecularweightisnegligible.
Although zero-order drug release profiles theoretically apply for
surface-eroding polymers [Siepmann & Göpferich 2001], analysis of the
release of both gentamicin and vancomycin under surface-eroding
conditionsindicatesacombinationofzero-andfirst-orderreleasekinetics,
yieldingconsistentlyhigherlevelsofantibioticreleasethanintheabsence
of lipase, i.e. in PBS. Zero-order release is a characteristic that is strongly
linked to the dimension of the surface-eroding polymer [Von Burkersroda
et al. 2002]. Exclusive zero-order release kinetics can therefore only be
obtainedwhenthesurfacearearemainsrelativelyconstantduringerosion
[Langer&Peppas1981],whichisnotthecaseforourPTMCdiscs.Figure5
presents the antibiotic release from PTMC in lipase as a function of time
after correction for the lost surface area, indicating zero-order release
kinetics (correlation coefficients R2exceed 0.87), except for vancomycin in
PTMC168.
59
3000
3000
Mt/A (µg/mm²)
Mt/A (µg/mm²)
2000
2000
1000
1000
00
00
10
10
20
20
time (days)
(days)
time
30
!
Figure 5: Cumulative antibiotic release (Mt) divided by the mean surface
area correctedfordecreaseintimeduetoerosion(A)fromPTMC168(open
symbols)andPTMC339discs(closedsymbols)inlipasesolution,plottedasa
function of time. Black symbols represent release from gentamicin-loaded
discs and blue symbols release from vancomycin loaded discs. The plots,
which are linearly correlated (R2 = 0.98 and 1.00 for gentamicin and
vancomycin release from PTMC339, respectively and R2 = 0.88 for
gentamicinreleasefromPTMC168)signifiesaconstantamountofantibiotic
beingreleasedpersurfacearea,indicatingzero-orderreleaseprofile.
60
High molecular weight PTMC339 is a promising biodegradable local
antibioticdeliverysystem,permittinghighantibioticlevelsforanextended
period, as a result of its surface erosion. Moreover, also relatively large
antibiotics like vancomycin can be effectively released, as surface erosion
allows for the release of large and small compounds, unlike in diffusioncontrolled release system where release of larger molecules may be
restrictedbymatrixinteractions.
61
62