Unstimulated human saliva samples were obtained from a total of 10 healthy participants (6 male, 4 female). In order to avoid interferences arising from the introduction of exogenous agents into the oral environment, participants were requested to collect all saliva available, i.e., (“whole”) saliva expectorated from the mouth, into a plastic universal collection tube immediately after waking in the morning on a pre-selected day. Each participant was also requested to refrain completely from oral activities (i.e., eating, drinking, tooth-brushing, oral rinsing, smoking, etc.) during the short period between awakening and sample provision (ca. 5 min). Approval for the collection of saliva specimens from human participants in this study was obtained from the University of Bolton Research Ethics Committee.

Each collection tube contained sufficient sodium fluoride (15 μmol.) in order to ensure that metabolites were not produced or consumed via the actions of micro-organisms or their enzymes present in whole saliva during periods of sample preparation and/or storage. Saliva specimens were transported to the laboratory on ice and then centrifuged immediately (4000 r.p.m. for 30 min.) on their arrival to remove cells and debris, and the resulting supernatants were stored at –20˚C for a maximum duration of 18 hr prior to treatment with tooth-whitening product supernatants and/or solutions, further storage at –20˚C as detailed below, and ¹H NMR analysis. The pH values of each salivary supernatant were determined prior to ¹H NMR analysis (these ranged from 6.45 - 7.44).

The tooth-whitening product investigated here consisted of a dual-syringe system, the first of which contained 12.5% (w/w) H₂O₂ (the active bleaching agent) and 2.50% (w/w) carbopol, and the second 95.0% (v/v) of an amino-alcohol bleaching accelerant, the remaining 5.0% (v/v) being water (Figure 1(a)). For the purpose of clinical application, the components in the first syringe are pre-mixed with those of the second in a 4:1 volume ratio so that the final applied H₂O₂ content is 10.0% (w/v). The pH value of the undiluted accelerant solution was 12.56.

For the clinical application of this dual-syringe toothwhitening product, a specially-designed mouth retractor is employed to retract the lips away from the teeth, and then the wy10 bleaching gel/activator mixture is painted onto tooth surface enamel as a thin coat of varnish [Figure 1(b)] in order to ensure that the bleaching agent does not gain access to areas outside the tooth surface (adequate further steps are taken to ensure that the bleaching agent components are isolated from all exposed dentine and soft tissues). This tooth-whitening product is employed with a thermal diffusion system² which is mounted over the retractor and serves to elevate the temperature of the bleaching gel on the enamel surfaces up to 38˚C. The application is applied once-daily (of total length 40 min, and involving a further bleaching episode after 20 min with a repeated application (replenished) volume of the 4:1 tooth-whitening gel:activator solution mixture). The final pH value of the resulting 4:1 (v/v) tooth-whitening gel:amino-alcohol activator product admixture is 10.0.

Approximately 100 mg pre-recorded masses of the 12.5% (w/w) H₂O₂ bleaching gel formulation present in syringe 1 were added to a 10.00 ml volume of HPLC-grade water, the mixture rotamixed and then centrifuged (3500 r.p.m. at ambient temperature) for a 30 min. period. The clear supernatant (of pH value 5.53) was removed and then employed to treat salivary supernatant samples and aqueous solutions containing sodium pyruvate in the manners outlined below.

The 95.0% (v/v) bleaching activator (accelerant) material present in syringe 2 was diluted 1:10 with HPLCgrade water so that its final concentration was 9.50% (v/v), and this solution was then directly utilised for additions to human salivary supernatant specimens. This 9.50% (v/v) activator solution had a pH value of 11.24, and the addition of only 38 µl of this to 0.18 ml volumes of human salivary supernatants (the mixture made up to a final volume of 0.594 ml with ²H₂O and H₂O as detailed below) did not elevate the final pH value of the latter and the final reaction mixture above 7.70, i.e. the buffering capacity of the much larger volume salivary supernatant effectively countered the non-neutral (alkaline) pH value of the activator.

Three separate aliquots (0.18 ml) of each salivary supernatant sample were removed and 76 µl volumes of the supernatant arising from the above 12.5% (w/v) H₂O₂ (syringe 1) formulation was then added to the first, 38 µl of the above 9.50% (v/v) amino-alcohol accelerant formulation solution was added to the second, and a combination of 76 µl of the above H₂O₂-containing supernatant and 38 µl of the accelerant solution were added to the third; to each of these mixtures, a 0.30 ml aliquot of ²H₂O (field frequency lock) was then added and their final volume was then adjusted to 0.594 ml with HPLCgrade H₂O. The mixtures were thoroughly rotamixed and then equilibrated at a temperature of 37˚C for a 30 min. period and then stored at –20˚C for a maximum duration of 72 hr prior to ¹H NMR analysis. Further 0.18 ml aliquots of each salivary supernatant sample treated with a 0.30 ml volume of ²H₂O and 0.114 ml of HPLC-grade H₂O [both previously “sparged” with helium (He) gas for a 30 min period] and then equilibrated and stored in the same manner served as controls.

Aqueous solutions containing sodium pyruvate (5.00 × 10^–3 mol·dm^–3) (Sigma Chemical Co. Ltd., Poole, Dorset, UK) were prepared in 5.00 × 10^–2 mol·dm^–3 phosphate buffer (pH 7.00) which was rigorously deoxygenated by purging with helium (He) gas prior to use (30 min at ambient temperature). 1.00 ml aliquots of these solutions were treated with 0, 19, 38, 57, 76 and 95 µl volumes of the above H₂O₂ solution, the samples thoroughly rotamixed and equilibrated at a temperature of 37˚C for 30 min, and then stored at a temperature of –20˚C for a period of 72 hr prior to ¹H NMR analysis. Matching pyruvate-containing solutions treated with equivalent volumes of HPLC-grade H₂O in place of the oral rinse, equilibrated at 37˚C (30 min), and stored in the same manner served as controls. This technique served as an effective means to estimate the content of this active tooth-whitening agent present in the tooth-whitening gel (syringe 1) formulation.

Typically, the total 0.594 ml volume of each sample was placed in a 5-mm diameter NMR tube, and then 6.0 µl of a 1.00 × 10^–2 mol·dm^–3 solution of trimethylsilyl-[2,2,3, 3-²H₄] propionate (TSP) in ²H₂O [serving as an internal chemical shift reference (δ = 0.00 ppm) and quantitative ¹H NMR standard] was added.

Proton (¹H) NMR measurements on the above pre-prepared human salivary supernatant samples were conducted on a Bruker Avance AX-600 spectrometer (Queen Mary University of London Facility, London, UK) operating at a frequency of 600.13 MHz in quadrature detection mode and a probe temperature of 300 K. Each spectrum corresponded to 32 free induction decays (FIDs), 5.6 μs pulses, a 1 s pulse repetition rate, 32,768 (subsequently zero-filled to 65,536) data points and spectral width 10,500 Hz. The intense residual H₂O/HOD signal (δ = 4.80 ppm) was suppressed via gated decoupling during the delay between pulses. Exponential linebroadening functions of 0.30 Hz were routinely applied to the FIDs prior to Fourier transformation, and all spectra were manually-phased and baseline-corrected. Spectra were acquired in an automated manner using a sample changer for continuous sample delivery.

Where appropriate, the broad protein resonances present in control and tooth-whitening product-treated salivary supernatant samples were suppressed by the Hahn spin-echo sequence (D[90˚x-t-180˚y-t-collect]) which was repeated 128 times (t = 68 ms). Chemical shifts were referenced to internal TSP. Where present, the methyl group resonances of acetate (singlet, δ = 1.920 ppm), alanine (doublet, δ = 1.487 ppm) and lactate (doublet, δ = 1.330 ppm) served as secondary internal chemical shift references for the saliva samples investigated.

For single-pulse, one-dimensional (1D) ¹H NMR spectra acquired on control and H₂O₂-containing tooth whitening gel supernatant-treated solutions of pyruvate, samples were prepared by the addition of a fixed volume (0.10 ml) of a solution of TSP in ²H₂O (1.00 × 10^–2 mol·dm^–3) to a 0.50 ml volume of the reaction mixture. Typical pulsing conditions were 64 free induction decays (FIDs) using 32,768 data points, 72˚C pulses and a 3 s pulse repetition rate to allow full spin-lattice (T₁) relaxation of the protons in the samples investigated. Chemical shifts were referenced to TSP (internal; final concentration 1.67 × 10^–3 mol·dm^–3), and exponential line-broadening functions of 0.30 Hz were again employed for purposes of processing.

The identities of biomolecule resonances present in the salivary ¹H NMR spectra acquired were routinely assigned by a consideration of chemical shift values, coupling patterns and coupling constants. The relative intensities of selected signals therein, and those of the phosphate-buffered aqueous pyruvate solutions treated with increasing volumes (0, 19, 38, 57, 76 and 95 µl) of the above solution prepared from the H₂O₂-containing toothwhitening gel, were determined by electronic integration via the spectrometer’s proprietory software (XWIN-NMR), and the concentrations of components detectable were determined by comparisons of their resonance areas with that of TSP (final concentration 1.00 × 10^–4 mol·dm^–3 in the salivary supernatants). Maintenance of the exact integral regions for each spectrum acquired was ensured.

Since the protein concentration of human saliva (1.40 g·dm^–3 - 6.40 g·dm^–3 [11]) is much lower than that of human blood plasma (65 g·dm^–3 - 83 g·dm^–3 [12]), the minimal macromolecular broad resonance envelop was not found to interfere with the observation and integration of each of the sharp low-molecular-mass resonances, and hence all of the salivary metabolite concentrations documented here represent those derived from such spectra. Furthermore, it should also be noted that the biomolecule concentrations determined in this investtigation reflect only the non-macromolecular-bound portion of these components and hence are expected to be somewhat lower than their total salivary levels.

Two-dimensional shift-correlated ¹H-¹H NMR (COSY) spectra of human salivary supernatants were acquired using the standard sequence of Aue et al. [13] with 2048 data points in the t₂ dimension, 256 increments of t₁, a 3.00 s relaxation delay, and 64 transients.

The statistical significance of differences observed between salivary pyruvate, methionine, lactate and trimethylamine concentrations was determined by a paired sample t-test performed on untransformed data (based on 9 degrees-of-freedom for n = 10 participants).

Figure 2(a) shows a typical high-field (0.72 ppm - 2.46 ppm) region of the 600 MHz single-pulse ¹H NMR spectrum of an untreated human salivary supernatant sample, and corresponding partial spectra acquired subsequent to the in vitro treatment of this sample with specified aliquots of the tooth-whitening amino-alcohol activator solution, bleaching gel formulation supernatant, and a combination of equivalent added volumes of the activator and bleaching gel supernatant solutions, are shown in Figures 2(b)-(d) respectively (the volumes of each solution or supernatant added are specified above in Section 2).

As previously documented [10], 600 MHz single-pulse ¹H NMR spectra of control (untreated) human salivary supernatant samples contained many prominent, sharp resonances ascribable to a wide range of low-molecularmass components. Indeed, signals assignable to short-chain organic acid anions (e.g., acetate, isoand n-butyrates, formate, fumarate, lactate, propionate, pyruvate, succinate, n-valerate and 3-D-hydroxybutrate), carbohydrates

such as glucose and galactose (both α- and β-anomers) are observable. The organic acid anions detectable are products arising from microbial metabolic pathways, and hence these agents (either individually or several or more in concert), particularly their salivary concentrations, may serve as chemotaxonomic indicators of microbial activity in the oral environment. Indeed, the pathogenic microorganism P. gingivalis generates high levels of acetate, n-butyrate and propionate, together with smaller amounts of iso-butyrate, iso-valerate, phenylacetate and succinate [14,15].

Also present are one or more broad resonances located at 2.04 ppm which are assignable to the acetamido methyl group protons (-NHCOCH₃) of N-acetylsugars located in the molecularly-mobile branching carbohydrate side-chains of salivary glycoproteins. This assignment is consistent with those previously made for the ¹H NMR spectra of other biofluids [16]. This broad resonance overlaps several sharper acetamido-CH₃ group ¹H resonances attributable to low-molecular-mass N-acetylsugars such as N-acetylneuraminate and N-acetylglucosamine saccharide fragments, which conceivably arise from the actions of bacterial-derived neuraminidase and hyaluronidase respectively, the latter towards ground substance hyaluronate [17]. These resonances may also be assignable to N-acetyl amino acids such as N-acetylglycine or N-acetylglutamate.

In addition, both ethanol and methanol were detectable in a large proportion of the human saliva samples subjected to ¹H NMR analysis. Although ethanol is a bacterial-derived catabolite (for example, arising from carbohydrate metabolism by Streptococcus mutans) [18], the methanol present is putatively derived from the passive or direct inhalation of cigarette smoke in which this alcohol is present, a consequence of the combustion of tobacco lignin which contains many methoxy substituents in its complex, macromolecular aromatic structure.

Two-dimensional ¹H-¹H COSY NMR spectra of typical human salivary supernatant samples showed clear connectivities between the multiplet resonances present. For example, linkages between the alanine-CH₃ and α-CH signals (doublet, δ = 1.487 ppm and quartet, δ = 3.765 ppm respectively), propionate-CH₃ and -CH₂ group signals (triplet, δ = 1.04 ppm and quartet, δ = 2.17 ppm respectively), the n-butyrate-CH₃, β-CH₂ and α-CH₂ group resonances (triplet, δ = 0.91 ppm, multiplet, δ = 1.56 ppm and triplet, δ = 2.15 ppm respectively), the ethanolCH₃ and -CH₂ groups (triplet, δ = 1.21 ppm and quartet, δ = 3.68 ppm), the lactate-CH₃ and -CHOH group resonances (doublet, δ = 1.330 ppm and quartet, δ = 4.130 ppm respectively) and the tyrosine aromatic ring protons (doublets located at 6.88 and 7.17 ppm) (data not shown).

Clearly, addition of the H₂O₂-containing product gives rise to the complete disappearance of the pyruvate-CH₃ group signal (singlet, δ = 2.388 ppm), an observation reproducible in all saliva samples tested in this manner (n = 10) [Figures 2(a) and (c)]. Further expanded 1.88 ppm - 2.46 ppm regions of these spectra are exhibited in Figures 3(a) and (c) respectively. These data are fully consistent with the oxidative consumption of salivary pyruvate by H₂O₂ present in the formulation (Equation (1)). Indeed, previous investigations have demonstrated that pyruvate acts as a powerful biofluid electron-donor (i.e., a water-soluble antioxidant) and is oxidatively decarboxylated to acetate and CO₂ on reaction with H₂O₂. The mean (±SEM) pre-treatment salivary pyruvate concentration was 2.19 (±1.07) × 10^–3 mol·dm^–3 for the n = 10 samples collected in this study.

An alternative mechanism for the generation of acetate and CO₂ from pyruvate involves 1) the prior generation of ^ŸOH radical from the interaction of “catalytic” redoxactive low-molecular-mass metal ions such as Fe(II), Cu(I) and Co(II) present in human saliva with the H₂O₂ tooth-whitening agent, and 2) direct reaction of this aggressively-reactive oxygen radical species with this α-keto acid anion “scavenger” (Equations (2) and (3)), although with much larger concentrations of available H₂O₂, this oxidation process is, of course, much more likely to proceed in accordance with Equation (1). In addition to both endogenous and exogenous (dietary) sources of these metal ions, the mean concentrations of copper and cobalt ions in human saliva are 1.25 (±0.20) × 10^–6 (mean ± SEM) [19] and 3.7 × 10^–7 mol·dm^–3 [20] respectively, the latter representing the mean of 10 out of 37 salivary determinations in which it was detectable, and presumably predominant as Co(II) in view of thermodynamic considerations and the availability of Co(III)-reducing electrondonors in this biofluid.

M⁽ⁿ⁺¹⁾⁺ + e^– → Mⁿ⁺ (2)

Mⁿ⁺ + H₂O₂ → M⁽ⁿ⁺¹⁾⁺ + ^ŸOH + OH^–(3)

Since ¹H NMR-detectable concentrations of acetate were detectable as a minor contaminant present in the tooth-whitening gel (syringe 1) supernatant, spectra of the H₂O₂-containing bleaching product-treated salivary supernatants showed elevations in the intensities of resonances ascribable to this agent. Of course, smaller increases in the acetate-CH₃ group signal intensity observed

following treatment are attributable to the H₂O₂-mediated oxidative decarboxylation of pyruvate described above.

Further H₂O₂-induced modifications to the ¹H NMR profiles of human saliva included 1) the oxidative consumption of lactate (-CH₃ group doublet, δ = 1.33 ppm, -CHOH proton quartet, δ = 4.13 ppm), which is explicable by the attack of ^ŸOH radical on this metabolite to primarily generate pyruvate (Equation (5)) which is then consumed by excess H₂O₂ (Equation (1)), or further ^ŸOH radical (Equation (4)) [as noted above, ^ŸOH radical can be generated from Fenton or pseudo-Fenton reaction processes involving trace levels of “catalytic” metal ions (Equations (2) and (3)); 2) significant elevations in the salivary concentrations of formate, an organic acid anion which is known to arise from oxidative damage to carbohydrate species such as glucose, and/or hyaluronate or N-acetyl glycoprotein N-acetylsugar species by ^ŸOH radical produced in the same manner as that described in 1) above; 3) the oxidation of trimethylamine (TMA) (singlet, δ = 2.91 ppm) to its corresponding N-oxide, a process represented by equation 6. Mean ± SEM percentage decreases observed in the TSP-normalised intensities of the lactate-CH₃ and trimethylamine-N(CH₃)₃ resonances were 21% ± 5% (p < 0.05) and 45% ± 9% (p < 0.01) respectively, whereas that for the increase in the formate signal was 42% ± 11% (p < 0.02). The mean ± SEM before-treatment concentrations of lactate, trimethylamine and formate were (4.43 ± 2.17) × 10^–3, (9.17 ± 3.04) × 10^–5 and (5.72 ± 2.51) × 10^–3 mol·dm^–3 respectively.

Moreover, for a limited number of the salivary supernatant samples examined (n = 3), we also found that the amino acid methionine (-S-CH₃ group singlet resonance, δ = 2.13 ppm) was also consumed, a process which generated its primary oxidation product methionine sulphoxide (with a characteristic -SO-CH₃ group singlet signal located at δ = 2.725 ppm) as previously described [21] and fully consistent with Equation (7). For these 3 samples, the mean ±SEM methionine concentration was 1.60 ± 0.44 × 10^–4 mol·dm^–3, and the mean percentage decrease in its 2.13 ppm -S(CH₃) group signal was 93% ± 4% (p < 0.005 for n = 3 specimens).

Treatment of human salivary supernatant specimens with a combination of the H₂O₂-containing bleaching gel supernatant (derived from syringe 1) with the diluted bleaching accelerant solution (from syringe 2) gave rise to similar modifications to the ¹H NMR profile of this biofluid [Figures 2(d) and 3(d)], specifically the complete oxidative consumption of pyruvate, accompanied by smaller reductions in the intensities of the lactate signals and increases in that of formate, together with even smaller decreases in that of the methionine-S-CH₃ resonance, although the latter observation was again limited to the small number of samples in which it was ¹H NMR-observable (n = 3).

As expected, addition of the tooth bleaching aminoalcohol accelerant alone to each salivary supernatant sample did not give rise to the ¹H NMR spectral modifications which were observed in those treated with either the H₂O₂-containing bleaching gel supernatant or a combination of an equivalent volume of this solution (76 µl) with an additional 38 µl of the 9.50% (v/v) solution of the tooth-whitening accelerant (apart, of course, from the appearance of signals ascribable to the added accelerant and an acetate contaminant also present in its solution). A Typical example of the spectra acquired from these samples is shown in Figures 2(b) and 3(b).

However, a sharp singlet resonance located at ca. 2.06 ppm was markedly and reproducibly diminished in intensity on treatment of each saliva sample with the bleaching activator solution [Figures 2(b) and 3(b)], and this observation was also notable in samples that were treated with a combination of this activator solution and the H₂O₂ gel-derived supernatant [Figure 1(d)], but not in those to which the H₂O₂-containing supernatant was added alone [Figures 2(c) and 3(c)]. As noted above, this particular signal is ascribable to the acetamido group (-NHCOCH₃) protons of low-molecular-mass N-acetylsugars or N-acetyl amino acids. This observation is not simply explicable, but if this resonance is assignable to an N-acteyl amino acid, the decrease in its intensity may arise from a condensation reaction between the carboxylate group (s) of these biomolecules (including the side-chain carboxylate group of N-acetylglutamate if this species is responsible for the signal) and the free amino group of the bleaching accelerant added, although this reaction is unlikely to proceed at the 37˚C equilibration temperature; however, it may be assisted by one or more salivary enzymes.

Moreover, also noteworthy were elevations in the electronically-integrated intensities of a resonance assignable to the malodorous biomolecule dimethylamine (singlet, δ = 2.78 ppm), an observation which is presumably attributable to the displacement and hence mobilisation of this species (which is protonated and hence positively-charged at salivary pH values which are close to neutrality) from negatively-charged binding sites present in salivary macromolecules by a relatively large excess of this added positively-charged amino-alcohol accelerating agent. Such binding-sites could include the carboxylate groups of aspartate or glutamate residues in salivary proteins, and/or that of glucuronyl residues in salivary polysaccharides or glysoaminoglycans such as soft tissue-derived hyaluronate.

Aqueous 5.00 × 10^–3 mol·dm^–3 standard solutions of pyruvate (n = 3) were treated with a 76 µl volume of the supernatant solution derived from the H₂O₂-containing tooth-whitening gel (Section 2) in order to provide further information regarding the nature and mechanism of this agent’s ability to oxidise this α-keto acid anion in human saliva. ¹H NMR analysis revealed that reaction of pyruvate (initially 5.00 × 10^–3 mol·dm^–3) with this bleaching agent gave rise to an essentially stoichiometric transformation of the α-keto acid anion substrate to acetate and CO₂. Indeed, >54% of the pyruvate was oxidatively decarboxylated to acetate and CO₂ under the experimenttal conditions employed, and after making allowances for the presence of a small amount of pyruvate as pyruvate hydrate () in aqueous solution, and asming that the reaction goes to completion, the expected level of acetate production from Equation (1) was 55.9%.

Pyruvate hydrate has a singlet resonance (δ = 1.50 ppm) of relatively low intensity which was also present in all ¹H NMR spectra acquired on aqueous pyruvate solutions. This resonance was also completely removed from the spectrum subsequent to treatment with the H₂O₂- containing product evaluated here.

Plots of the ratio of the concentration of acetate generated to that of the initial (pre-H₂O₂-treated) pyruvate plus pyruvate hydrate concentrations versus the added tooth-whitening product supernatant volume or corresponding (predicted) added H₂O₂ concentration ([H₂O₂]) were linear with a zero intercept (Figure 4), data demonstrateing that this technique serves as an excellent method for the determination of this agent in such commerciallyavailable tooth-whitening products. Indeed, the correlation coefficient (r value) obtained was 0.9988. The relationship between this ratio and added H₂O₂ concentration is described by equation 8, where [acetate]₀, [pyruvate]₀ and [pyruvate hydrate]₀ represent the initial concentrations of acetate, pyruvate and pyruvate hydrate respecttively [the initial concentration of pyruvate hydrate was also determined by the electronic integration of its singlet resonance (δ = 1.50 ppm) and normalisation of its intensity to that of the TSP internal standard of known concentration]. The [acetate]₀ value (the initial, untreated pyruvate solution acetate concentration) was corrected for a contribution from small amounts of an acetate contaminant present in the H₂O₂ bleaching gel supernatant solution employed for these studies (i.e. this contribution was dependent on the volume of supernatant added).

The estimated percentage H₂O₂ content of the gel present in syringe 1 of the tooth-whitening formulation investigated was found to be 12.37% ± 0.35% (w/w) (mean ±SEM) for n = 5 repeat determinations, a value in close agreement with that of its specified value [12.50% (w/w)].

Multicomponent ¹H NMR investigations of the oxidising

ability of the tooth-whitening formulation investigated here towards salivary biomolecules demonstrated that critical salivary electron-donors are readily consumed by H₂O₂ therein. Indeed, it is clear that any H₂O₂ which “escapes” from the tooth-whitening site to saliva (or alternative oral fluids), or that which inadvertently comes into contact with this oral fluid, will be effectively consumed by salivary electron-donors which therefore serve to protect soft tissues against any sensitivity or damage inducible by this tooth-whitening agent or any reactive oxygen radical species derivable therefrom. These further oxidants include hydroxyl radical (^ŸOH) arising from Fenton-type reactions [22] involving any low-molecularmass redox-active transition metal ions available in human saliva and serving as catalytic sources, mono-deprotonated H₂O₂ (), which is more bleaching-active than its protonated precursor [23], (although is only generated in significant amounts at alkaline pH values since H₂O₂’s first pKa value is 11.65), and perhaps also superoxide anion () generated from the single-electron oxidation of H₂O₂ and/or.

Indeed, the salivary concentrations of pyruvate (a two electron donor) range from (0.10 - 10.60) × 10^–3 mol·dm^–3 [17], and the mean level of thiols (a single electron supplier) therein is 3.6 × 10^–5 mol·dm^–3 [24]. After making appropriate allowances for thermodynamic equilibria, and the rate of each reaction involved under physiological conditions, data acquired here indicates that further reductants present in human saliva (e.g., urate, ascorbate, thiocyanate anion and the amino acids cysteine and methionine), those at relatively low concentrations, will also be at least partially effective with regard to the neutralisation of any adverse effects exertable by H₂O₂ during tooth-whitening episodes conducted with this particular product.

However, it should be noted that selected single electron-donors present in human saliva such as thiols and ascorbate also have the capacity to reduce higher oxidation state metal ions to their lower ones [for example, reduction of Fe(III) to Fe(II)] in accordance with Equation (2), so that the latter can then take part in Fenton or pseudo-Fenton reaction systems which generate the aggressively-reactive ^ŸOH radical (Equation (3)).

All of the above salivary metabolites serve to act as biomolecular protectants against any adverse sensitivity reactions and/or toxic effects exertable by the leakage of H₂O₂ to soft tissue areas which, without an adequate form of control, are accessible by this particular biofluid.

Interestingly, methionine is known to be an essential precursor required for the production of volatile sulphur compounds (VSCs) by Gram-negative anaerobic bacteria in the oral environment; i.e., it plays a major role in the development of halitosis. Moreover, malodorous TMA has been implicated as a major factor involved in recalcitrant oral malodour. However, their involvement in these conditions will not be discussed further here.

Currently, there are four methods for the administration of such tooth-whitening agents, specifically 1) that administered by a dental clinician, a process which can involve the employment of high H₂O₂ concentrations, which commonly range from 35% - 40% (w/w)—this treatment is frequently supplemented with the use of a light and/or heat source, 2) that supervised by a dental clinician in which a customised bleaching tray containing high levels of CP, often ranging from 35% - 45% (w/w), is placed into the mouth of the patient for periods of up to 1.00 hr in the clinician’s surgery, 3) a home-use tooth-whitening kit provided by a dentist (otherwise known as “Home” or “Nightguard” bleaching) and applied by the patient [usually a 10% - 22% (w/w) CP formulation] present in a customised application tray, and 4) products which are purchasable by consumers in pharmacies and other retail outlets (known as “over-the-counter” products); these often contain H₂O₂ or CP [up to an EU legal limit of 0.10% (w/w or w/v)], or peroxoborates or peroxosulphates, but more recently this range of products has expanded to include those which contain chlorine dioxide ()-generating systems, especially those which involve the acidulation of aqueous chlorite solutions. The product investigated here can be directly applied in the dental surgery [under 1) and 2) above], but is predominantly for “at-home” use as in 3) above.

Both a series of small clinical investigations and case studies have provided evidence that a 10% (w/w) CPcontaining gel formulation applied in a bleaching tray overnight, (i.e. the “Nightguard” vital bleaching technique), gave rise to a predictable level of tooth whitening [25-32], as did H₂O₂-containing strips [33] and a “Powerbleaching” technique employing 35% (w/v) H₂O₂ coupled with or without activation by UV light and/or heating stages [34,35].

Tooth sensitivity is a common deleterious effect of external tooth-whitening processes [36], and particularly noteworthy in this context is the observation that with the use of products containing 10% (w/v) CP, 15% - 65% of patients reported an enhanced level of tooth sensitivity [37,38]. Moreover, increased incidences of tooth sensitivity (from 67% - 78%) have been noted following toothwhitening episodes with H₂O₂ coupled with a thermal enhancement process [39].

It is well known that tooth sensitivity usually lasts for up to four days post-bleaching [39,40]. However, this adverse effect can be prolonged for up to 39 days [37, 38]. Furthermore, a clinical study that involved comparisons of two differing 10% (w/w) CP brands revealed that 55% of a large number of patients investigated found tooth sensitivity and/or gingival irritation, and 20% of those who had experienced such deleterious effects discontinued the treatment in view of their discomfort [37].

Although mechanisms that could account for such tooth sensitivities during or following such external toothwhitening episodes remain speculative, one in vitro study revealed that H₂O₂ can penetrate enamel and dentine, and also penetrate the pulp chamber [41].

High-resolution, high-field ¹H NMR spectroscopy is a technique which offers many advantages over alternative time-consuming, labour-intensive analytical methods for evaluating the fate of H₂O₂ (or alternative peroxo-adduct tooth-whitening agents) in oral fluids since 1) it permits the rapid, non-invasive and simultaneous examination of a very wide range of components present in biofluids, and 2) it generally requires little or no knowledge of sample composition prior to analysis. Furthermore, chemical shift values, coupling patterns and coupling constants of resonances present in the ¹H NMR spectra of such multicomponent systems provide much valuable information regarding the molecular nature of both endogenous and exogenous agents therein.