Effect of tooth bleaching agents on protein content and mechanicalproperties of dental enamel

Hunida M. Elfallah a, Luiz E. Bertassoni a, Nattida Charadram b,c, Catherine Rathsam b, Michael V. Swain a,⇑a Department of Biomaterials Science, Faculty of Dentistry, University of Sydney, New South Wales, Australiab Institute of Dental Research, Westmead Millennium Institute and Westmead Centre for Oral Health, New South Wales, Australiac Restorative Department, Faculty of Dentistry, Naresuan University, Phitsanulok, Thailand

a r t i c l e i n f o

Article history:Received 8 September 2014Received in revised form 24 March 2015Accepted 28 March 2015Available online 1 April 2015

Keywords:Tooth bleachingEnamelHardnessElastic modulusCreep

a b s t r a c t

This study investigated the effect of two bleaching agents, 16% carbamide peroxide (CP) and 35% hydro-gen peroxide (HP), on the mechanical properties and protein content of human enamel from freshlyextracted teeth. The protein components of control and treated enamel were extracted and examinedon sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE). Marked reduction of the pro-tein matrix and random fragmentation of the enamel proteins after bleaching treatments was found. Themechanical properties were analyzed with Vickers indentations to characterize fracture toughness, andnanoindentation to establish enamel hardness, elastic modulus and creep deformation. Results indicatethat the hardness and elastic modulus of enamel were significantly reduced after treatment with CPand HP. After bleaching, the creep deformation at maximum load increased and the recovery uponunloading reduced. Crack lengths of CP and HP treated enamel were increased, while fracture toughnessdecreased. Additionally, the microstructures of fractured and indented samples were examined with fieldemission gun scanning electron microscopy (FEG-SEM) showing distinct differences in the fracture sur-face morphology between pre- and post-bleached enamel. In conclusion, tooth bleaching agents can pro-duce detrimental effects on the mechanical properties of enamel, possibly as a consequence of damagingor denaturing of its protein components.

! 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Dental enamel, as the hardest tissue in the body, acts as aprotective covering of teeth and can withstand a wide range offunctional and non-functional loads. Enamel is a hierarchical bio-composite made of 94–96 wt.% inorganic content, 1 wt.% organicmatrix and 4–5% water [1]. The inorganic content is comprised oflong thin hydroxyapatite crystals tightly packed together with aprotein glue (hydrophobic enamelin) to form enamel rods [1].These rods are encapsulated by a thin protein-rich organic sheath(primarily hydrophilic ameloblastin) around 0.8–1 lm thick [2].Although the organic contents of enamel comprise less than 1%,they play a significant role in determining the mechanical behaviorof enamel [3]. This organic content helps to define three-dimen-sional cleavage planes to deflect cracks which prevent fracture

from progressing through enamel and allows limited movementbetween the rods during stress [2,4].

The mechanical properties of enamel are dependent on its com-position and its structural organization [5]. The hardness andmodulus of elasticity of the rods have been reported to be muchhigher than that of the organic sheath [6]. The highest values ofhardness and modulus of elasticity have been found at the occlusalsurface and decrease toward the dentin-enamel junction (DEJ) asmineral density decreases and the organic matrix increases [7,8].In addition, the fracture toughness of enamel ranges from0.4 MPa m1/2 in the outer enamel to 2.37 MPa m1/2 in the innerregion near the DEJ [9,10]. This increase is primarily due to anincrease of the organic matrix and different rod orientations(decussation) which causes crack propagation resistance by theformation of protein bridges, crack deflection and microcracks [9].

Tooth bleaching has become a highly popular esthetic dentaltreatment as it is the easiest and the least destructive procedurefor treatment of tooth discoloration. At present, there are threemain tooth bleaching techniques being used: in-office bleaching,dentist supervised home bleaching and over the counter (OTC)bleaching products [11]. Contemporary bleaching techniques use

http://dx.doi.org/10.1016/j.actbio.2015.03.0351742-7061/! 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Biomaterials Science, Sydney Dental Hospital, Schoolof Dentistry, University of Sydney, 2 Chalmers St, Surry Hills, NSW 2010, Australia.Tel.: +61 29351 8357.

E-mail address: mswain@mail.usyd.edu.au (M.V. Swain).

Acta Biomaterialia 20 (2015) 120–128

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /ac tabiomat

hydrogen peroxide (HP) or its precursor carbamide peroxide (CP)as an active ingredient [11]. Bleaching peroxide penetrates throughenamel and dentin eventually reaching the pulp in the course ofwhich it undergoes chemical breakdown forming oxygen free radi-cals. These free radicals are highly unstable and capable of oxidiz-ing and disintegrating a wide range of organic and inorganicmaterials including chromophores [12]. Despite their ability to dis-integrate pigmented molecules in tooth structure, bleaching agentscan also attack the organic and inorganic contents of tooth struc-ture [13].

Morphological alterations of enamel surface after bleachingtreatment can be seen as an increase in surface roughness and ero-sion [14,15]. Changes of the chemical composition of enamel andmineral loss has also been reported [15,16]. Micro- and nano-me-chanical investigations of enamel subjected to bleaching agentshave shown that HP significantly decreases the hardness andmodulus of elasticity of enamel [17–19], as well as the fracturetoughness and wear resistance [20,21]. These changes in themechanical properties of enamel could be due to both, loss of min-eral content as well as denaturation and degradation of the organicmatrix by the oxidation reaction [18,19,21]. However, the mecha-nism(s) of action by which the bleaching peroxide induces detri-mental effects on the mechanical properties of enamel has notbeen fully identified. Therefore, this study aims to analyze theinfluence of bleaching agents on the nanomechanical propertiesof enamel and enamel matrix protein/peptides along with theunderlying mechanisms in an attempt to understand the structuralalterations that accompany these changes. The nanomechanicalproperties of enamel were investigated before and after bleachingusing a nanoindentation system. The fracture toughness and crackbehavior of enamel were quantitated with Vicker’s microhardnesstest then examined under field emission gun scanning electronmicroscopy (FEG-SEM). The effects of the bleaching agents on theenamel matrix content were analyzed on SDS–polyacrylamide gelelectrophoresis (SDS–PAGE).

2. Materials and methods

2.1. Specimen preparation

Twenty-four fully erupted human third molars were obtainedfrom healthy patients aged 18–40 according to protocols approvedby the Sydney West Area Health Service (Ref. No. HREC/11/WMEAD/115). The extracted teeth were cleaned and embeddedin a cold-curing epoxy resin (Epofix, Struers, Denmark). Upon cur-ing the occlusal surface was ground flat with silicon carbide paperswith grit sizes 240, 400, 600 and 1200 under continuous waterirrigation to avoid overheating. Specimens were then polished byspecimen polisher (RotoPol-22, Struers, Denmark) with 9 lm and1 lm diamond polishing pastes then lastly with 0.04 lm colloidalsilica. After each polishing cycle, specimens were ultrasonicallycleaned (Unisonics, Australia) for 5 min and stored in Hank’s bal-anced salt solution HBSS (Sigma–Aldrich, Germany) at room tem-perature between the bleaching treatments. Prior to indentationexperiments, the specimens were examined under optical micro-scope to exclude excessively scratched and cracked surfaces. Onehalf of the specimen was covered with nail varnish (control) andthe other half was treated with one of the bleaching agents.

Two bleaching agents were used in this study: 16% carbamideperoxide (CP) home bleaching gel (Polanight, SDI Limited,Australia); and 35% hydrogen peroxide (HP) in-office bleachingagent (Polaoffice+, SDI Limited, Australia). The whitening treat-ments of enamel samples were performed at room temperatureand according to the manufacturers’ instructions as outlined below.

Briefly, for the CP group, 16% CP gel was applied on the toothsurface for 90 min. Samples were then rinsed thoroughly withwater, dried and stored in HBSS. This process was repeated dailyfor 14 days. For the HP bleaching group, 35% HP was applied onthe surface of specimen for 10 min, rinsed under running waterand dried. The treatment was repeated 3–4 times in a single ses-sion. The mechanical tests were conducted within 24 h after appli-cation of bleaching agents.

2.2. Hardness and elastic modulus

The indentation experiment was conducted using an UltraMicro Indentation System (UMIS-2000, CSIRO, Australia), equippedwith a three-sided Berkovich indenter tip calibrated on a fused sil-ica standard sample of known properties. Specimens (n = 12) werenanoindented using a maximum load of 10 mN. A total of 15indents separated from each other by 20 lm were positioned onthe occlusal surface of each specimen along its buccal aspect forboth the bleached and non-bleached halves of each tooth sample.The UMIS system software (Ibis, Fisher-Cripps laboratories,Australia) was used for the subsequent calculation of the elasticmodulus and hardness of the samples tested. The nanoindentationhardness (H) is the contact pressure of the indenter divided by theprojected contact area (A) of the sample at maximum load (Pmax)which can be estimated from [22]:

H ¼ Pmax=A ð1Þ

whereas elastic modulus of the specimen Es can be calculated from[22]:

1=Er ¼ ð1$ v2s Þ=Es þ ð1$ v2

i Þ=Ei ð2Þ

where Er is the indentation elastic modulus (reduced modulus), Ei

and vi are the elastic modulus and Poisson’s ratio of the diamondindenter 1070 GPa and 0.07 respectively. The Poisson’s ratio, vs,for enamel is 0.3 [23]. The data were analyzed according to Oliverand Pharr [22] and the hardness and elastic modulus values wereaveraged from the results of 15 indents.

2.3. Load dependent mechanical properties of enamel

To investigate the effect of tooth bleaching agents on subsurfaceenamel and its depth dependent mechanical properties, a range ofloads 5, 10, 20 and 50 mN were applied to control and bleachedenamel samples. Hardness and elastic modulus were obtained fol-lowing the methods described above.

2.4. Creep behavior of enamel

The approach for determination of creep behavior of control andbleached enamel was the same as that used by He and Swain [24].A 250 mN force was applied using the UMIS nanoindentation sys-tem with a 900 s holding time at maximum load. Another 900 shold period was included during unloading at 2% of the peak loadto analyze the creep recovery of enamel [24]. One-increment stepduring loading and unloading was used in order to reduce the vis-coelastic effects that result from gradual loading and unloading[25]. The difference between the displacement during creep atmaximum load and creep recovery at minimum load representsthe amount of unrecovered deformation during a hold period atconstant force. In addition, the relative recovery of the materialin relation to the deformation was calculated as a percentage bythe relationship:

% Recovery ¼ h2=h1 & 100 ð3Þ

H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128 121

where h1 is the indentation depth displacement during hold atmaximum load and h2 is the displacement recovery during theholding time at 2% of the maximum load upon unloading. This rela-tive recovery indicates the amount of the deformation that occurredduring loading that was recovered to its normal state once the loadswere almost entirely released.

2.5. Fracture toughness and crack behavior

Fracture toughness (K1C) of the material may be calculated fromthe size of radial cracks about a Vicker’s hardness impression usingthe following expression [26]:

K1C ¼ xðE=HÞ1=2 P=c3=2 ð4Þ

where P is the applied load, c is the crack length, and x is a geomet-ric constant of the indenter x = 0.016, E and H are Young’s modulusand hardness of the enamel.

Hardness was measured with a Vicker’s microhardness tester(Shimadzu, Japan) and a load of 9.8 N was applied. Five indenta-tions were made on the polished occlusal surface of each specimen.Three specimens were used and indentations were performed bothon the control and on the bleached surface. The indentations wereviewed with a Leica microscope and the crack length was mea-sured using Image J (version 1.44, National Institutes of Health,USA).

2.6. Scanning electron microscopy

On completion of the enamel stress tests, the fracture surfaceand crack patterns were examined by FEG-SEM. Tooth sampleswere dehydrated with graded ethanol concentrations, mountedon a sample holder with carbon tape and sputter coated with goldabout 15 nm thick. Observation was performed under high res-olution FEG-SEM (Zeiss Ultra plus, Germany). Images wereobtained using a secondary electron detector at an acceleratingvoltage of 10.00 kV.

2.7. Extraction and purification of enamel proteins

After bleaching treatments, the enamel from each group wasseparated from dentin and pulverized. The protein extraction tech-nique was similar to the protocol used previously [27]. The pulver-ized enamel was demineralized with 10% acetic acid with proteaseinhibitor (Protease Inhibitor Cocktail P8340, Sigma–Aldrich) for24 h at 4 "C under constant agitation. The first supernatant wascollected by centrifugation at 4000g for 5 min and stored at$20 "C. The precipitated demineralized enamel was suspended in4 M guanidine-HCl in 0.05 M TRIS with protease inhibitor for24 h at 4 "C under constant agitation. These extraction procedureswere repeated three times. The collected supernatant was purifiedby running through OASIS-HLB reversed-phase liquid chro-matography (OASIS#-HLB Plus, Waters; Ireland) and freeze-dried(Ilshin Lab co., Korea). The protein aggregates from each groupwere suspended in 200 ll MQH2O and the total protein yieldedfrom each group was quantified spectrophotometrically (Bradfordprotein assay, Thermo Scientific; USA).

The enamel protein extracts were separated by SDS–PAGE on a16.5% Tricine gel with a broad range standard molecular weightmarker (BIORAD; USA). The gel was stained with Coomassie bril-liant blue G250 for 30–40 min. The stained gel was scanned andanalyzed using Image J (version 1.44; National Institutes ofHealth).

2.8. Statistical analysis

All data are presented as mean ± SD. Statistical significance wasdetermined by comparison between CP and HP bleached groupsand a control group. The data were analyzed using ANOVA and aTukey/Kramer test with the confidence level set at 5%. In the pre-sent study, a P-value 6 0.0001 was considered as statisticallysignificant.

3. Results

3.1. Hardness and elastic modulus

The nanoindentation elastic modulus and hardness were calcu-lated as a function of penetration depth from the load–displace-ment curves. Typical behavior of the bleached versus controlenamel is shown (Fig. 1). In the CP group, elastic modulus andhardness were significantly reduced by approximately 6.5% and13.7% respectively (Table 1). Although the decrease in hardnessand elastic modulus is small, the differences were statistically sig-nificant (P 6 0.0001). At the same time, the HP group presented asignificant (P 6 0.0001) reduction of '10% in hardness and 5.5%in elastic modulus in comparison with the control (Table 1).There was no significant difference between the two treatmentgroups.

3.2. Load dependent mechanical properties of enamel

The hardness and elastic modulus of control enamel decreasefrom about 7.5 GPa and 121 GPa at 5 mN load to 4 GPa and96 GPa at 50 mN. There is more than 45% reduction in hardnessand about 20% reduction in modulus of elasticity as the depth ofpenetration increases (Table 2). A similar reduction of themechanical properties with increasing contact depth was also seenin the bleached enamel (Fig. 2). The difference in enamel mechani-cal properties between the control and the two bleached groups isgreater under lower loads. Furthermore, as the load and contactdepth increase the difference among the three groups decreases.

3.3. Creep deformation behavior

The nanoindentation creep data of triplicate samples from eachgroup were averaged and plotted (Fig. 3a). The control and CP trea-ted enamel showed a similar creep deformation of 80 nm and83 nm respectively, while the HP bleached enamel had a signifi-cantly higher creep deformation under constant loading of about96 nm displacement at maximum load.

Conversely, the creep recovery data (Fig. 3b) show the highestvalues for control enamel with 73% of the creep deformation

0

2

4

6

8

10

12

0 0.1 0.2 0.3 0.4

Load

(mN

)

Depth (µm)

HP

CP

Control

Fig. 1. Load–displacement curve of control, CP and HP bleaching groups.

122 H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128

during loading recovered at unloading. On the other hand, HPbleached enamel had the least recovery with about 49 nmunrecovered deformation and only 48% relative recovery. The CPbleached enamel; on the other hand presented 66% relativerecovery.

3.4. Fracture toughness and crack behavior

The indentations and the crack patterns of the control andbleached enamel were examined for measurement of crack length.The control enamel yielded an average crack length of67.9 ± 14.8 lm and average fracture toughness 1.3 ± 0.5 MPa m1/

2. The fracture toughness of the CP enamel was significantlyreduced (P 6 0.0001) to 0.7 ± 0.2 MPa m1/2 with the averageobserved crack length of 100 ± 18 lm. The HP bleached enamel,fracture toughness was also significantly reduced to0.8 ± 0.3 MPa m1/2 (P 6 0.0001) with an average crack length of86.2 ± 20 lm. On the bleached specimens, these cracks wereobserved to be longer with multiple secondary cracking anddetachment of the adjacent enamel. Additionally, in the bleachedgroups the cracks tended to propagate between the enamel rodsalong the least resistant inter-rod sheath (Fig. 4).

3.5. Scanning electron microscopy results

The microstructures of fractured and indented samples wereexamined with FEG-SEM. Cracks arising from the Vicker’s indenta-tions of control enamel are shown in Fig. 4a. At higher magnifica-tion, crack deflection, crack tip splitting and crack bridges appearas indications of resistance of enamel structure to crack prop-agation. In the bleached enamel, secondary multiple cracks arisefrom the indenter impression which propagate through the inter-rod enamel as indicated by the semi-circular like crack path abouteach rod (Fig. 4b). This crack pattern suggests a weaker connectionbetween the rods in the bleached enamel.

The fractured surface of control enamel in Fig. 5a illustrates theline of fracture propagating through the rods and inter-rod compo-nents, irrespective of their orientation. However, in bleachedenamel, fracture propagation appeared in an arch-like pattern pre-dominantly through the inter-rod enamel as shown in Fig. 5b.Cracking and dislodgement of the inter-rod and rod enamel werealso seen. Under higher magnification, the SEM image of controlenamel clearly shows the tightly packed nature of hydroxyapatitecrystals found within both the rod structure and the inter-rodenamel (Fig. 6a). Comparatively in Fig. 6b, the bleached enamel

appears more ‘loosely’ attached, with increased spacing and poros-ity between the crystals.

3.6. Protein analysis and gel electrophoresis

The protein component of enamel from bleached and controlgroups was extracted and quantitated by Bradford assay(Table 3) then examined on SDS–PAGE (Fig. 7a). SDS–PAGE gelsshow a reduction of high and low molecular weight proteins afterthe application of bleaching agents, with the most remarkablereduction occurring in the lower molecular weight proteins below6 kDa (Fig. 7a). This is supported by the Image J plot analysis ofSDS–PAGE gel staining intensity (Fig. 7b) and the Bradford proteinassay results (Table 3).

4. Discussion

4.1. Hardness and elastic modulus

The average elastic modulus and hardness for the controlenamel at 10 mN load are consistent with the values reported inprevious studies [7,28,29] where the elastic modulus and hardnessranged from 120 GPa and 6.4 GPa at the cusp tip to 47 GPa and2.7 GPa near the DEJ [7]. In the current study, the area of enamelinvestigated was just below the occlusal surface and parallel tothe rod direction where the highest values of hardness and modu-lus of elasticity are reported [7].

Both HP and CP bleaching agents resulted in significant reduc-tion of enamel hardness and modulus of elasticity. These changesin mechanical properties are presumed to be due to dissolutionof the hydroxyapatite [16,30–32] and destruction of the proteinmatrix by the peroxide free radicals [16,18,19,21]. Zimmermanet al. claimed this reduction of the mechanical properties afterwhitening treatment is due to denaturing or loss of the proteinstructure as evidenced by loss of fluorescence particularly at thesurface enamel [19]. In another study by Jiang et al. [18] usingRaman/fluorescence spectroscopy, the authors found significantreduction of fluorescence intensity in the HP treated group sug-gesting that the organic matrix of enamel may be greatly affectedby tooth bleaching [18].

In the current study, there was no significant difference in thechanges of the mechanical properties between the two peroxideconcentrations used. This is in agreement with the results of DeAbreu et al. [33] who found that the tooth bleaching agents pro-duce significant enamel microhardness loss irrespective of the per-oxide concentration, pH or time of exposure. Hence, the significantreduction of the hardness and elastic modulus of dental enamelafter application of bleaching agents are suggested to be primarilydue to loss or destruction of the protein layer and are not as depen-dent on the concentration or pH of the whitening products.

4.2. Load-dependent mechanical properties

For control and bleached enamel the values of hardness andmodulus of elasticity decrease as the load and associated depth

Table 1The average and standard deviations of the hardness and elastic modulus of enamelwith the P-values for the control, CP and HP bleached.

Group Control CP HP P-valueCP

P-valueHP

Elasticmodulus

108 ± 2.8 101 ± 6.9 102 ± 4.7 <.0001 <.0005

Hardness 5.8 ± 0.3 5.0 ± 0.7 5.2 ± 1.1 6.0001 6.0001

Table 2Hardness (H) and elastic modulus (E) of control and bleached enamel under different loads with their contact depth in nanometres.

Load (mN) Control CP HP

Depth H E Depth H E Depth H E

5 0.19 ± 0.01 7.5 ± 1.2 121 ± 4.3 0.22 ± 0.01 5.6 ± 0.2 99.4 ± 5.9 0.19 ± 0.03 7.0 ± 1.4 116.5 ± 810 0.3 ± 0.02 5.8 ± 0.3 108 ± 2.8 0.34 ± 0.03 5.0 ± 0.7 101 ± 6.9 0.33 ± 0.03 5.2 ± 1.1 102.1 ± 4.720 0.46 ± 0.01 5.6 106.6 ± 4.3 0.52 ± 0.02 4.6 ± 0.3 99 ± 6 0.49 ± 0.02 5.1 ± 0.3 96.9 ± 3.450 0.8 ± 0.02 4 ± 0.2 96 ± 3.3 0.84 ± 0.04 3.6 ± 0.3 89.3 ± 3.3 0.83 ± 0.02 3.7 ± 0.2 91.6 ± 4

H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128 123

of penetration increase. This load and depth dependence of enamelmechanical properties was attributed to its hierarchical structure.As the load increases, more rods are involved and the volume frac-tion of the sheath protein increases which leads to reduction of themechanical properties under the indenter [34].

The load-dependent mechanical properties were also seen inthe bleached enamel. This indicates that the reduction of hardnessand modulus values is higher for the surface layer of enamel andreturns to normal as the depth increases. Similar results wereestablished in a previous study by Ushigome et al. [15] whoindented a cross-section of a bleached enamel surface. They foundthat the reduction in hardness was greatest at the outer 2 lm andwas less significant at deeper regions up to 20 lm below the sur-face. Consequently, we support the perspective that the mostdeleterious effect of tooth whitening products is limited to the sur-face layer of enamel.

4.3. Creep deformation and recovery

By considering its anisotropic structure dental enamel has con-siderable creep deformation under constant load and upon unload-ing. From the current results, control enamel exhibited steadycreep deformation during time held at maximum load. In a recentstudy comparing the creep behavior of enamel with hydroxyap-atite, the authors found that enamel has much greater creep defor-mation, which indicates that creep behavior of enamel principallyoriginates from its protein components [24]. This deformationbehavior can be explained by the Tension-Shear Chain (TSC) model

of Ji and Gao [35]. Under constant loading the force of the indentergenerates tension on the mineral crystallites that is transferred tothe protein between the crystals and in the rod sheath as shear.The shearing of the protein is argued to be due to unfolding of itsdomain structure and enables limited sliding of the hydroxyapatitecrystals [35]. Moreover, during unloading the stretched proteinstend to return to their normal form by refolding of the domain(sacrificial) bonds which results in relative recovery of the defor-mation upon unloading. Similarly, Schneider et al. [36] have pro-posed that the creep behavior of enamel is due to flow of theprotein in the inter-rod sheath layer under stress application. Ithas been suggested that the thickness and shear properties of theprotein layer are the significant parameters that endow enamelwith its inelastic deformation behavior [37].

The higher creep deformation seen in bleached enamel; there-fore, can be attributed to the conformational changes that areinduced by the oxidation reaction of the bleaching agent. Thesestructural changes of the protein may lead to denaturing or lossof the protein domain bonds which during hold at maximum load,results in increased creep deformation. In addition, loss of the pro-tein can lead to movement of the mineral crystallites due to greaterslippage at protein–mineral interface that manifests as highercreep deformation. This was observed in SEM images as increasedspacing between the crystals (Fig. 7b).

For the unloading period, bleached enamel manifested limitedrecovery of the creep deformation in comparison with the controlenamel. This may arise because refolding of the protein domainbond does not take place due to conformational changes or damagecaused by peroxide free radicals.

0 1 2 3 4 5 6 7 8 9

10

0 20 40 60

Har

dnes

s (G

Pa)

Load (mN)

Control

CP

HP

a

0

20

40

60

80

100

120

140

0 20 40 60

Elas

tic m

odul

us (G

Pa)

Load (mN)

Control

CP

HP

b

Fig. 2. Plot of (a) hardness and (b) elastic modulus for control and bleached enamelat different indentation loads.

0

0.02

0.04

0.06

0.08

0.1

0.12

0 200 400 600 800

ht (µ

m)

t (sec)

control

CP

HP

a

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 200 400 600 800

ht (µ

m)

t (sec)

control

CP

HP

b

Fig. 3. (a) The creep behavior of enamel for the different groups, and (b) Therecovery behavior of control enamel, CP and HP bleached. All tests were made witha Berkovich indenter at 250 mN load.

124 H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128

4.4. Fracture toughness and crack behavior

The average fracture toughness of control enamel was1.3 ± 0.5 MPa m1/2 which is within the range of the reportedenamel toughness [2,38,39]. Although in this study fracture tough-ness was determined using Vicker’s microindentation approachbased on direct measurement data from crack length, the valuesobtained were within the range of the results from the most recentenergetic methodology [40]. Using continuous stiffness measure-ments of the depth sensing indentation approach, fracture tough-ness values reported were in agreement with previous literature[41]. The resistance of enamel to wear and cracking is crucial inorder to survive its physiological functions. The toughness charac-teristics of enamel are attributed to its anisotropic structure, whichis made of highly mineralized enamel rods embedded in a protein-rich organic matrix [39]. It has been shown that the fracture tough-ness of enamel is inhomogeneous and spatially anisotropic. Dentalenamel has higher resistance to cracks that are perpendicular tothe rod direction than cracks parallel to the rod direction [39].Cracks initiating at the surface enamel toward DEJ are deflectedby the decussated rods and continue growth about the tooth’speriphery away from dentin and pulp [42]. By comparing the frac-ture toughness of enamel with its primary constituent hydroxyap-atite (HAP), it has been found that enamel is much tougher thanHAP. This was attributed to the structural organization of enamel

rods and its organic matrix where the stress about the crack tipis transferred from the harder mineralized crystals to the softerorganic layer around them [2,43]. The higher fracture resistanceof enamel has been shown to be partly due to stretching of theorganic matrix forming protein bridges to resist separation whichexert crack closure stress [38]. This mechanism is argued to beformed by unfolding of the structural domains ‘‘sacrificial bonds’’of these protein bridges between the mineral and protein orbetween different portions of the protein itself [38,44]. In addition,unfolding of this bond allows a limited slippage at the mineral-pro-tein interface which in turn increases dissipation of fracture energy[35]. This crack resistance characteristic was reduced in bleachedenamel which appeared as multiple longer cracking that primarilyextended within the inter-rod region. These results were also seenin a previous study where 10% CP resulted in significant reductionof enamel fracture toughness due to alterations of the organicmatrix [21]. Similar results have been found by Garrido et al.[20] after application of 38% HP bleaching agent. Their results havealso shown increased wear rate and reduced fracture toughnessafter bleaching. However, a partial recovery of the reduced fracturetoughness has been observed after one week storage in artificialsaliva [20].

4.5. Scanning electron microscopy analysis

Cracks found in bleached enamel propagate a greater distancein an arch-like pattern as they grow through the inter-rod enameldemonstrating a weaker matrix surrounding the rods (Fig. 4b). In

10µm 1µm

a

b

Fig. 4. Vicker’s indentations of control enamel (a), higher magnification of thecracks (inset) showing the crack bridging, and (b) bleached enamel.

10µm

R

IR

R

IR

2µm

a

R

b

IR

R R

IR

2 µm

Fig. 5. Fracture surface of (a) control enamel shows the path of fracture (dottedline) including rods (R) and interrod (IR) enamel and (b) fractured enamel surfaceafter bleaching showing the arched-like pattern of the interrod enamel (dotted line)with cracking of the inter-rod enamel (arrows) and dislodgement of the rods.

H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128 125

addition, the fractured surface of bleached enamel (Fig. 5b) clearlyshows cracking and dislodgement of the rod and inter-rod enamelindicating changes in the properties of the protein glue that holdthe crystals together. By comparing Fig. 6a and b, the bleachedenamel shows increased spacing and porosity between the HAPcrystals which may be a result of loss or destruction of the proteinglue [45]. In contrast for control enamel, the highly tortuous pathformed by rod/inter-rod enamel (Fig. 5a) contributes to crackdeflection and reduction of stress at the crack tip, thereby actingas a toughening mechanism [38].

The extension of the cracks from the corners of indentationsand in the fractured specimens of enamel at the inter-rod regionis most pronounced following bleaching. This observation maypossibly arise because the protein remnants of the inter-rod areprimarily hydrophilic ameloblastin rich [46] and as such the perox-ide bleaching could more readily penetrate, interact and fragmentsuch proteins [47]. The protein remnants binding the crystallites

within the rods together on the other hand are primarilyhydrophobic enamelin [48] and as such the penetration of the per-oxide is far more difficult resulting in less complete dissolution andbreak up of such proteins. This would explain why the extent offracture toughness degradation ('50%) is far more significant thanthe reduction in elastic modulus and hardness.

4.6. Protein analysis and gel electrophoresis

Control enamel presents greater protein content than bleachedenamel (Table 1). Analysis of the protein extracted from controlenamel by SDS–PAGE shows low (' 5 kDa) and high ('90 kDa)molecular weight proteins. The distinct reduction of protein inten-sity along with the lower shift of low molecular weight protein wasobserved in CP and HP bleaching groups compared to the control.This suggests the slightly greater ability of CP bleaching agent todegrade enamel protein (Fig. 7). The urea contained in CP bleachinggel is credibly responsible for this protein degradation by affectinghydrogen-bond of matrix proteins [49].

In mature enamel, the residual matrix proteins play a functionalrole in mechanical properties of enamel [5]. Dental enamel is muchtougher than its major components HAP [2]. The increase in tough-ness is due to the residual protein components that have profoundsoftening effect and provide a degree of ductility to an otherwisebrittle mineralized structure [50]. Mechanically, the residualmatrix proteins could prevent crack propagation by defining threedimensional cleavage planes that deflect cracks laterally and allowlimited movement between adjacent rods. This could reduce stres-ses and increase enamel’s resistance to fracture [2]. Moreover, it

1µm

a

b

Fig. 6. Higher magnification of enamel structure (a) control enamel showing thehydroxyapatite crystals tightly packed together forming the rod structure sur-rounded by the inter-rod enamel (b) bleached enamel shows increased porositiesand spacing between the crystals (insets).

Table 3Concentration of protein extracted from control and bleached enamel measured bythe Bradford protein assay.

Group Enamel weight(g)

Protein recovered(lg/ll)

% of protein in enamel(wt.)

Control 2.86 5.272 0.036HP 1.89 1.527 0.016CP 2.87 1.044 0.007

Fig. 7. (a) SDS–PAGE of protein extracts from control enamel, HP and CP bleachedenamel. (b) Intensity plot of the SDS–PAGE gel of extracted enamel proteins in (a) asanalyzed by Image J (version 1.44; National Institutes of Health). Intensity of lowmolecular weight enamel proteins recovered from bleached groups is lowercompared to control.

126 H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128

has been shown that enamel matrix proteins have the ability toregulate the electrostatic properties of enamel surface and thuscan buffer the HAP against acid attack [51]. Therefore, even a slighteffect on organic composition of enamel could lead to significantchanges in its mechanical properties.

The results from this study clearly show the deleterious effect oftooth bleaching products on the mechanical properties and matrixproteins of dental enamel. In particular, the organic content ofenamel has been shown to be the responsible structure in regulat-ing enamel mechanical properties and viscoelasticity [37].Removal or degradation of enamel organic matrix has demon-strated significant reduction of its mechanical properties [5]. Thecurrent results support the notion that H2O2 in tooth bleachingproducts gives rise to unstable oxygen free radicals that are ableto induce structural changes in the matrix proteins locatedbetween the crystallites and in the rod sheath area. Tooth bleach-ing agents have induced significant alteration of the interrod pro-teins which was seen as increased cracking and reduced fracturetoughness. The semicircular crack paths seen under FEG-SEMdemonstrate the reduced presence of the hydrophilic sheath pro-teins to resist fracture.

Accordingly, the ability of enamel matrix proteins to enableshear and stress dissipation as well as subsequent recovery proper-ties is compromised. Therefore, destruction or denaturation ofthese organic components leads to a decrease in the ability ofenamel for stress-redistribution and fracture resistance.

One of the limitations of this study is that the enamel surface ofthe specimens was ground and polished before bleaching. This pro-cedure was done to provide a more even surface for the accuracy ofthe indentations. However, sample preparation can induce residualstresses in the surface enamel which consequently alter cracklength [52].

Another limitation is the use of HBSS solution for storing teethafter bleaching instead of artificial or human saliva which mayinduce enamel remineralization and recovery. It has been foundthat bleaching agents could breakdown or remove salivary proteinthat tightly bound to enamel lesions which consecutively enhancethe remineralization of the early carious lesions [53]. Various sali-vary proteins have been found to maintain oral health by protect-ing against tooth wear, controlling microbial adherence, inducingion transport, and enhance remineralization [54]. In addition, sal-iva plays an important role in re-establishing the characteristicsof the enamel and reduction of its roughness after microabrasion[55]. However, many previous studies have used artificial salivaor remineralizing solution after bleaching to simulate clinicalsituation and have also found significant reduction of the mechani-cal properties [56–59]. From this study further research exploringthe roles that salivary proteins may potentially play in repair ofbleaching damaged enamel would be recommended.

In summary, tooth bleaching products, which have been shownto induce strong oxidation reaction of the protein/peptide compo-nents of enamel, compromise its mechanical properties and as aconsequence, reduce its ability to resist deformation and withstandcontact induced damage during function.

5. Conclusions

Significant reduction of enamel hardness, modulus of elasticityand fracture toughness occur after whitening treatment regardlessof the type of tooth whitening gels. These effects, which are consis-tent with the degradation of the matrix protein by the bleachingagents, result in a significant decrease in the mechanical propertiesof enamel. In addition, the bleaching agents increased the creepdeformation and reduced the creep recovery of enamel under con-stant loading which is also consistent with damage to the

deformable matrix protein by the peroxide free radicals. Changesin protein structures caused by the H2O2 appear to be responsiblefor modifications in the structure and the altered pattern and char-acteristics of the fracture surface of bleached enamel. Therefore,tooth whitening products potentially produce detrimental effectson the mechanical properties of dental enamel as a consequenceof destruction or denaturing of the matrix proteins. Accordingly,the treatment protocol for tooth discoloration, particularly toothbleaching should be re-evaluated to prevent the undesirablealtered enamel properties.

Disclosures

This study has no conflicts of interest.

Acknowledgments

The authors would like to thank Professor Roberto Braga forvaluable technical assistance and Dr.Bill Kahler for kindly provid-ing teeth. Scanning electron microscopy was performed with assis-tance from staff at the Australian Centre for Microscopy &Microanalysis (ACMM) at the University of Sydney. Partial supportfor this research from an Australian Research Foundation grant anda scholarship for HME by Ministry of higher education and scienti-fic research, Libya are gratefully appreciated.

Appendix A. Figures with essential color discrimination

Certain figures in this article, particularly Figs. 1–3, and 7, aredifficult to interpret in black and white. The full colour imagescan be found in the on-line version, at http://dx.doi.org.10.1016/j.actbio.2015.03.035.

References

[1] Nanci A. Enamel: composition, formation, and structure. In: Nanci A, editor.Ten Cate’s oral histology: development, structure, and function. St.Louis: Mosby; 2003. p. 145–91.

[2] White S, Luo W, Paine M, Fong H, Sarikaya M, Snead M. Biological organizationof hydroxyapatite crystallites into a fibrous continuum toughens and controlsanisotropy in human enamel. J Dent Res 2001;80(1):321–6.

[3] He L, Swain M. Understanding the mechanical behaviour of human enamelfrom its structural and compositional characteristics. J Mech Behav BiomedMater 2008;1(1):18–29.

[4] Hassan R, Caputo A, Bunshah R. Fracture toughness of human enamel. J DentRes 1981;60(4):820–7.

[5] Baldassarri M, Margolis HC, Beniash E. Compositional determinants ofmechanical properties of enamel. J Dent Res 2008;87(7):645–9.

[6] Ge J, Cui F, Wang X, Feng H. Property variations in the prism and the organicsheath within enamel by nanoindentation. Biomaterials 2005;26(16):3333–9.

[7] Cuy J, Mann A, Livi K, Teaford M, Weihs T. Nanoindentation mapping of themechanical properties of human molar tooth enamel. Arch Oral Biol2002;47(4):281–91.

[8] He B, Huang S, Jing J, Hao Y. Measurement of hydroxyapatite density andKnoop hardness in sound human enamel and a correlational analysis betweenthem. Arch Oral Biol 2010;55(2):134–41.

[9] Bajaj D, Arola D. Role of prism decussation on fatigue crack growth andfracture of human enamel. Acta Biomat 2009;5(8):3045–56.

[10] Xu HHK, Smith DT, Jahanmir S, Romberg E, Kelly JR, Thompson VP, et al.Indentation damage and mechanical properties of human enamel and dentin. JDent Res 1998;77(3):472–80.

[11] Joiner A. The bleaching of teeth: a review of the literature. J Dent2006;34(7):412–9.

[12] Minoux M, Serfaty R. Vital tooth bleaching: biologic adverse effects-a review.Quintessence Int 2008;39(8):645–59.

[13] Jiang T, Ma X, Wang Y, Zhu Z, Tong H, Hu J. Effects of hydrogen peroxide onhuman dentin structure. J Dent Res 2007;86(12):1040–5.

[14] Fu B, Hoth Hannig W, Hannig M. Effects of dental bleaching on micro- andnano-morphological alterations of the enamel surface. Am J Dent2007;20(1):35–40.

[15] Ushigome T, Takemoto S, Hattori M, Yoshinari M, Kawada E, Oda Y. Influenceof peroxide treatment on bovine enamel surface-cross-sectional analysis. DentMater 2009;28(3):315–23.

H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128 127

[16] Efeoglu N, Wood D, Efeoglu C. Thirty-five percent carbamide peroxideapplication causes in vitro demineralization of enamel. Dent Mater2007;23(7):900–4.

[17] Hairul Nizam B, Lim C, Chng H, Yap A. Nanoindentation study of humanpremolars subjected to bleaching agent. J Biomech 2005;38(11):2204–11.

[18] Jiang T, Ma X, Wang Y, Tong H, Shen X, Hu Y, et al. Investigation of the effects of30% hydrogen peroxide on human tooth enamel by Raman scattering andlaser-induced fluorescence. J Biomed Opt 2008;13(1):014019.

[19] Zimmerman B, Datko L, Cupelli M, Alapati S, Dean D, Kennedy M. Alteration ofdentin–enamel mechanical properties due to dental whitening treatments. JMech Behav Biomed Mater 2010;3(4):339–46.

[20] Garrido M, Giráldez I, Ceballos L, Gómez del Río M, Rodríguez J.Nanotribological behaviour of tooth enamel rod affected by bleachingtreatment. Wear 2011;271(9–10):2334–9.

[21] Seghi R, Denry I. Effects of external bleaching on indentation and abrasioncharacteristics of human enamel in vitro. J Dent Res 1992;71(6):1340–4.

[22] Oliver W, Pharr G. An improved technique for determining hardness andelastic modulus using load and displacement sensing indentationexperiments. J Mater Res 1992;7(6):1564–83.

[23] DJ Haines. Physical properties of human tooth enamel and enamel sheathmaterial under load. J Biomech. 1968; 1, 2: 117–118, IN17-IN22, 119–125.

[24] He L, Swain M. Nanoindentation creep behavior of human enamel. J BiomedMater Res 2009;91A(2):352–9.

[25] Yang S. Analysis of nanoindentation creep for polymeric materials. J AppPhysic. 2004;95(7):3655.

[26] Anstis G, Chantikul P, Lawn B, Marshall D. A critical evaluation of indentationtechniques for measuring fracture toughness: i, direct crack measurements. JAm Ceram Soc 1981;64(9):533–8.

[27] Fincham AG, Belcourt AB, Termine JD. Molecular composition of the proteinmatrix of developing human dental enamel. J Dent Res 1983;62(1):11–5.

[28] Jeng Y, Lin T, Hsu H, Chang H, Shieh D. Human enamel rod presents anisotropicnanotribological properties. J Mech Behav Biomed Mater 2011;4(4):515–22.

[29] Lippert F, Parker D, Jandt K. Susceptibility of deciduous and permanent enamelto dietary acid-induced erosion studied with atomic force microscopynanoindentation. Eur J Oral Sci 2004;112(1):61–6.

[30] Al-Salehi S, Wood D, Hatton P. The effect of 24h non-stop hydrogen peroxideconcentration on bovine enamel and dentine mineral content andmicrohardness. J Dent 2007;35(11):845–50.

[31] Cavalli V, Rodrigues LKA, Paes-Leme AF, Soares LES, Martin AA, Berger SB, et al.Effects of the addition of fluoride and calcium to low-concentrated carbamideperoxide agents on the enamel surface and subsurface. Photomed Laser Surg.2011;29(5):319–25.

[32] Jiang T, Ma X, Wang Z, Tong H, Hu J, Wang Y. Beneficial effects ofhydroxyapatite on enamel subjected to 30% hydrogen peroxide. J Dent2008;36(11):907–14.

[33] De Abreu D, Sasaki R, Amaral L, Fiorio F, Basting R. Effect of home-use and in-office bleaching agents containing hydrogen peroxide associated withamorphous calcium phosphate on enamel microhardness and surfaceroughness. J Esthet Rest Dent 2011;23:158–68.

[34] Zhou J, Hsiung L. Depth-dependent mechanical properties of enamel bynanoindentation. J Biomed Mater Res 2007;81A(1):66–74.

[35] Ji B, Gao H. Mechanical properties of nanostructure of biological materials. JMech Phys Solids 2004;52(9):1963–90.

[36] Schneider G, He L, Swain M. Viscous flow model of creep in enamel. J Appl Phys2008;103(1):014701.

[37] Xie Z, Swain M, Munroe P, Hoffman M. On the critical parameters that regulatethe deformation behaviour of tooth enamel. Biomaterials2008;29(17):2697–703.

[38] Bajaj D, Arola D. On the R-curve behavior of human tooth ename. Biomaterials2009;30:4037–46.

[39] Xu H, Smith D, Jahanmir S, Romberg E, Kelly J, Thompson V, et al. IndentationDamage and Mechanical Properties of Human Enamel and Dentin. J Dent Res.1998;77(3):472–80.

[40] Rodríguez J., Garrido M., Ceballos L. Fracture toughness measurements onbovine enamel by indentation techniques. In: ICF13, 2013.

[41] Garrido M, Giráldez I, Ceballos L, Rodríguez J. On the possibility of estimatingthe fracture toughness of enamel. Dent Mater 2014;30:1224–33.

[42] Yahyazadehfar M, Bajaj D, Arola DD. Hidden contributions of the enamel rodson the fracture resistance of human teeth. Acta Biomater 2013;9(1):4806–14.

[43] Bajaj D, Nazari A, Eidelman N, Arola D. A comparison of fatigue crack growth inhuman enamel and hydroxyapatite. Biomaterials 2008;29:4847–54.

[44] Fantner G, Hassenkam T, Kindt J, et al. Sacrificial bonds and hidden lengthdissipate energy as mineralized fibrils separate during bone fracture. NatMater 2005;4(8):612–6.

[45] Cavalli V, Giannini M, Carvalho RM. Effect of carbamide peroxide bleachingagents on tensile strength of human enamel. Dent Mater 2004;20:733–9.

[46] Uchida T, Tanabe T, Fukae M, Shimizu M, Yamada M, Miake K, et al.Immunochemical and immunohistochemical studies, using antisera againstporcine 25 kDa amelogenin, 89 kDa enamelin and the 13–17 kDanonamelogenins, on immature enamel of the pig and rat. Histochem CellBiol. 1991;96(2):129–38.

[47] Hegedüs C, Bistey T, Flóra-Nagy E, Keszthelyi G, Jenei A. An atomic forcemicroscopy study on the effect of bleaching agents on enamel surface. J Dent1999;27(7):509–15.

[48] Fukae M, Tanabe T. Nonamelogenin components of porcine enamel in theprotein fraction free from the enamel crystals. Calcif Tissue Int1987;40(5):286–93.

[49] Goldberg M, Arends J, Jongebloed WL, Schuthof J, Septier D. Action of ureasolutions on human enamel surfaces. Caries Res 1983;17(2):106–12.

[50] Robinson C, Lowe NR, Weatherell JA. Amino acid composition, distribution andorigin of ‘‘tuft’’ protein in human and bovine dental enamel. Arch Oral Biol1975;20(1):29–42.

[51] Lubarsky G, D’Sa R, Deb S, Meenan B, Lemoine P. The role of enamel proteins inprotecting mature human enamel against acidic environments: a double layerforce spectroscopy study. Biointerphases 2012;7(1):1–8.

[52] Hassan R, Caputo A, Bunshah R. Fracture toughness of human enamel. J DentRes 1981;60(4):820–7.

[53] Iizuka J, Mukai Y, Taniguchi M, Mikuni-Takagaki Y, Ten Cate JM, Teranaka T.Chemical alteration by tooth bleaching of human salivary proteins thatinfiltrated subsurface enamel lesions—experimental study with bovine lesionmodel systems. Dent Mater J 2014;33(5):663–8.

[54] van Nieuw Amerongen A, Bolscher JGM, Veerman ECI. Salivary proteins:protective and diagnostic value in cariology? Caries Res 2004;38(3):247–53.

[55] Pini NIP, Lima DANL, Sundfeld RH, Ambrosano GMB, Aguiar FHB, Lovadino JR.In situ assessment of the saliva effect on enamel morphology aftermicroabrasion technique. Braz J Oral Sci 2014;13(3):187–92.

[56] Azer S, Azer S, Machado C, Sanchez E, Rashid R. Effect of home bleachingsystems on enamel nanohardness and elastic modulus. J Dent2009;37(3):185–90.

[57] Basting R, Rodrigues Jr A, Serra M. The effect of 10% carbamide peroxide,carbopol and/or glycerin on enamel and dentin microhardness. Oper Dent2005;30(5):608–16.

[58] de Oliveira R, Paes-Leme A, Giannini M. Effect of carbamide peroxide bleachinggel containing calcium or fluoride on human enamel surface microhardness.Braz Dent J 2005;16(2):103–6.

[59] Pinto C, de Oliveira R, Cavalli V, Giannini M. Peroxide bleaching agent effectson enamel surface microhardness, roughness and morphology. Braz Oral Res2004;18(4):306–11.

128 H.M. Elfallah et al. / Acta Biomaterialia 20 (2015) 120–128