Fluctuations in surface pH of maturing rat incisor enamel are a result of cycles of H+-secretion by ameloblasts and variations in enamel buffer characteristics

Bone 60 (2014) 227–234

Contents lists available at ScienceDirect

Bone journal homepage: www.elsevier.com/locate/bone

Original Full Length Article

Fluctuations in surface pH of maturing rat incisor enamel are a result of cycles of H+-secretion by ameloblasts and variations in enamel buffer characteristics Helle H. Damkier a, Kaj Josephsen b, Yoshiro Takano c, Dirk Zahn d, Ole Fejerskov a, Sebastian Frische a,⁎ a

Department of Biomedicine, Aarhus University, Denmark Department of Dentistry, Aarhus University, Denmark c Section of Biostructural Science, Graduate School of Tokyo Medical and Dental University, Japan d Lehrstuhl für Theoretische Chemie/Computer Chemie Centrum, Friedrich-Alexander Universität Erlangen-Nürnberg, Germany b

a r t i c l e

i n f o

Article history: Received 1 July 2013 Revised 18 November 2013 Accepted 16 December 2013 Available online 25 December 2013 Edited by: Mark Johnson Keywords: Ameloblasts Hydroxyapatite V-ATPase Vinblastine FR167356

a b s t r a c t It is disputed if ameloblasts in the maturation zone of the enamel organ mainly buffer protons released by hydroxyapatite (HA) crystal growth or if they periodically secrete protons to create alternating acidic and alkaline conditions. The latter hypothesis predicts alternating pH regimes in maturing enamel, which would be affected by pharmacological interference with ameloblast H+-secretion. This study tests these predictions. Colorimetric pH-indicators and ratiometric fluorometry were used to measure surface pH in maturation zone enamel of rat incisors. Alternating acidic (down to pH 6.24 ± 0.06) and alkaline zones (up to pH 7.34 ± 0.08) were found along the tooth coinciding with ameloblast morphological cycles. Underlying the cyclic pattern, a gradual decrease in pH towards the incisal edge was seen. Vinblastine or FR167356 (H+-ATPase-inhibitor) disturbed ameloblast acid-secretion, especially in the early parts of acidic zones. 3− would be Enamel surface pH reflects the titration state of surface PO3− 4 -ions. At the pH-values observed, PO4 protonated (pKa N 12) and HA dissolved. However, by molecular dynamics simulations we estimate the pKa of at an ideal HA surface to be 4.3. The acidic pH measured at the enamel surface may thus only dissolve HPO2− 4 is less electrostatically shielded. During repeated alkaline/ non-perfect domains of HA crystals in which PO3− 4 acidic cycles, near-perfect HA-domains may therefore gradually replace less perfect HA-domains resulting in near-perfect HA-crystals. In conclusion, cyclic changes in ameloblast H+-secretion and the degree of enamel maturation determine enamel surface pH. This is in accordance with a hypothesis implicating H+-ATPase mediated acid-secretion by ameloblasts. © 2013 Elsevier Inc. All rights reserved.

Introduction Acid–base homeostasis of the intracellular compartment is crucial for the maintenance of most cellular functions in the body. Intracellular homeostasis is obtained by extensive transport of acid/base equivalents between the intracellular and the extracellular compartment. The acid/ base status of the extracellular compartment is in turn maintained constant by appropriate function of the kidney and the lungs. However, the acid/base status of specialized parts of the extracellular space may also be controlled by nearby cells, as is the case for the lacunae formed by osteoclasts during bone resorption. In addition, acid/base conditions play

Abbreviations: RA, ruffle-ended ameloblasts; SA, smooth-ended ameloblasts; HA, hydroxyapatite; GBHA, glyoxal bis(2-hydroxyanil). ⁎ Corresponding author at: Department of Biomedicine, Aarhus University, Vilh. Meyers Allé 3, Universitetsparken Bygn. 1234, 8000 Århus C, Denmark. E-mail address: [email protected] (S. Frische). 8756-3282/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.bone.2013.12.018

an important role for the mineralization process of biological apatite, exemplified by developing enamel [1]. Tooth enamel mineralization depends on a delicate cellular control of the ionic composition and pH of the fluid surrounding the growing crystals. Control of these parameters is most likely mediated by the enamel organ cells and not least the ameloblasts (see review [2]). As the enamel organ is effectively inaccessible to direct physiological examinations, information on pH of the enamel in different stages of enamel formation is scarce. Before it became appreciated that the ameloblasts during maturation appear as either ruffle-ended (RA) or smooth-ended (SA) ameloblasts [3], the extracellular pH during the secretory stage was claimed to be 7.3–7.4 whereas the pH in the more mature enamel ranged from 8.0 to 8.5 [4]. However, studies by Takano et al. introduced a calcium-chelator dye, glyoxal bis(2-hydroxyanil) (GBHA), which demonstrated distinct red banding along the enamel at regular intervals corresponding to the location of SA bands in rat, bovine, porcine and monkey teeth [5]. Using colorimetric indicators it was shown, that this staining reflects areas of alkaline conditions [6].

228

H.H. Damkier et al. / Bone 60 (2014) 227–234

The two different morphological states of ameloblasts (RA and SA) reflect a cyclic cellular modulation process common to several species [7]. In rats, this process is characterized by the ameloblasts going through a cycle lasting about 8 h, during which the cells are in the state of RA for about 4 h before abruptly changing to become a SA type of cell for about 2 h and then gradually rebuilding the highly infolded (ruffled border) cell membrane towards the enamel. The modulation process proceeds in a wavelike pattern from the start of the maturation zone towards the incisal edge resulting in transverse bands of RA and SA conformation moving along the enamel [2]. Much effort has been invested in understanding the cellular mechanisms behind this modulation process and the functional implications. Sasaki, Takagi & Suzuki used 3 different pH staining solutions on bovine incisors and showed a banding of alternating acidic (pH ~ 6) and neutral (pH ~ 7) enamel zones and associated them with RA and SA cell bands [6]. They suggested that the acidic extracellular environment might be a result of protons released by the growth of hydroxyapatite crystals during maturation. According to this hypothesis, which was further advanced in several reviews [2,8], an important role of the ameloblasts is to transport bicarbonate to the maturing enamel to buffer the excess protons. A number of studies have aimed to support this hypothesis [9–14]. However, we have recently presented data on ion transporters in secretion and maturation stage ameloblasts, which have resulted in a new functional hypothesis of the enamel organ during maturation [15]. According to our hypothesis, H+ is pumped by H+-ATPases in the ruffled border towards the maturing enamel surface, which therefore becomes acidic under the RA. This hypothesis does not preclude that protons are stoichiometrically released by the appositional growth of the enamel crystals during the long lasting process of enamel maturation [2]. This passive proton release may induce a low pH in bulk enamel in the early nucleation and precipitation phase probably facilitating the gradual matrix proteolyses [2]. However, this passive proton release cannot explain the steep fluoride gradient observed in the outermost 100–150 μm of the human enamel when F− is given during tooth development [1,3,16], and the gradual reduction in content of carbonate, magnesium etc. in surface enamel crystals during the maturation phase [1]. According to our hypothesis, the composition of surface enamel crystals is the result of an energy-demanding active process based on intricate cellular collaboration and cyclic changes of ameloblast morphology and function resulting in alternating partial dissolution and precipitation of enamel crystal subdomains by cyclic changes in enamel pH [15]. As pointed out above, the knowledge of enamel surface pH is scarce and based on traditional colorimetric methods or analysis of sectioned teeth. Based on the current knowledge, which has limited spatial resolution, it appears that pH is higher under the SA, than under the RA. This is in accordance with our hypothesis, but as the colorimetric measurements may be influenced by other factors including the concentration of the indicator, more precise knowledge of enamel surface pH is needed. In the present study the hypothesis of ameloblast proton secretion described above is tested in two ways: 1) The spatial variability and range of pH on the surface of maturing enamel of rat incisors are measured by a ratiometric fluorescent pH indicator (BCPCF) and conventional colorimetric pH indicators. The hypothesis predicts lower pH under RA than SA. 2) The spatial variability in pH and GBHA staining is assessed in vinblastine and H+-ATPase-inhibitor-treated (FR167356) rats. The hypothesis predicts that both treatments should result in higher pH under RA since the acid-secretion of RA is predicted to be reduced. Materials and methods Enamel staining with GBHA and colorimetric pH-indicators Eight male Wistar rats (bw 274 ± 24 g) were anesthetized with 20% chloral hydrate (0.2 ml/100 g bw) and decapitated. The mandibles

were removed and cleaned of soft tissues. The lower incisors with adhering enamel organs were then carefully dissected from the surrounding alveolar bone [17]. In each animal one tooth was used for GBHA staining and one for pH-indicators. For GBHA staining the periodontal connective tissue adhering to the enamel organ were pulled off with a pair of fine tweezers. Each incisor was stained at room temperature for 7–8 min by immersion in 2 ml solution containing 3.75% GBHA (Sigma Chemical Co., St. Louis) and 1.275% sodium hydroxide dissolved in 75% ethanol. Subsequently the incisors were briefly rinsed in 75% ethanol, the enamel organ peeled off the enamel surface, further rinsed in 96% ethanol, and air dried. For staining with pH-indicators the enamel organ was gently wiped from the enamel surface with lightly moistened ear cotton sticks. Immediately after this, the incisors were soaked into 2 ml solutions of either methyl red (0.1% in 96% ethanol) or bromocresol purple (0.1% in 20% ethanol) for 2 min. Excess solution was removed and the teeth air dried. Color standards at different pH values (ranging from 4 to 7.5) were made from adding 200 μl pHindicator to 4 ml of PBS buffer stem solutions adjusted with hydrochloric acid or sodium hydroxide. All incisors were photographed shortly after staining and air drying with a Nikon D1 digital camera attached to a Wild M8 Macroscope. Enamel surface pH measurements using ratiometric fluorometry A series of pilot experiments were performed to optimize the labeling and ratiometric pH measurement procedure. For the final series of experiments, incisors from 3 untreated rats (bw 197 ± 8 g), 5 rats (bw 160 ± 5 g) treated with vinblastine (0.5 mg/rat) and 3 rats (bw 208 ± 7 g) treated with H+-ATPase blocker FR167356 [18] (50 mg/rat) were investigated. Vinblastine and FR167356 were given by IP injection 4 h before the animals were anesthetized with isoflurane and decapitated. Animal experiments were conducted according to animal experiment license # 2012-15-2934-00240 issued by the Danish Animal Experiment Authorities. One lower incisor was removed and stained with GBHA as described above. The contralateral incisor was removed and cleaned as described above for colorimetric pH indicators. The enamel was air dried and quickly formed an opaque region with a sharply defined boundary. This opaque white boundary was marked with a scalpel and served as a reference point. The tooth was loaded for 60 s in 50 μM 2′,7′-bis(3carboxypropyl)-5(6)-carboxyfluorescein (BCPCF, Molecular Probes) dissolved in saline. Excess dye was gently removed by using a filter paper. Each tooth was separately placed on a glass coverslip and mounted on an inverted microscope. The dye was excited using 495 nm and 440 nm light from a monochromator (Till Photonics) and the light emitted at 510 to 535 nm was recorded by a 12-bit cooled monochrome CCDcamera system (IMAGO, Till photonics). For overview images, exposure time was for 50 and 100 ms, for 495 and 440 nm images respectively, using a 1× objective. For detailed measurements of RA and SA zones, a 10 × objective was used and exposure time was reduced to 1 ms for both wavelengths. QED InVivo imaging software (Media Cybernetics) was used to control wavelength, light exposure time, and binning. The distances from the start of the maturation zone and the center of each SA-band were measured on overview images. The SA-bands were numbered from the start of the maturation zone and high magnification images were obtained from each. For each SA-band, regions of interest were selected at 5 positions: A: in the newly formed RA close to the SA, B: in the terminal part of the SA band, C: in the middle of the band, D: in the newly formed SA and E: in the RA just preceding the SA band (Fig. 1E). The images were analyzed using ImageJ Software. The fluorescence intensities following excitation with both 495 and 440 nm within each region were measured. For translation of fluorescence to pH, background was subtracted from both images and the ratio of emission following excitation at the two wavelengths (495/440 ratio) was calculated. The ratios were calibrated using HEPES buffered saline (145 mM

H.H. Damkier et al. / Bone 60 (2014) 227–234

229

E

A SA

secr

RA

GBHA

SA

SA

10

SA

SA

SA

I495/I440

8 6

RA

RA

RA

RA

RA

RA

4 2

B

A BC D E

secr

Methyl red

SA

SA

SA

pH

0 6

SA

7

8

8.0

RA

RA

RA

RA

7.5

C

RA

Bromocresol purple

RA

SA

RA

SA

pH

SA

secr

7.0 6.5

RA

D

6.0 SA secr

SA

SA

5.5 RA

RA

BCPCF

RA

A B C D E

A B C D E

A B C D E

SA1

SA2

SA3

Fig. 1. Labial views of rat mandibular incisors stained with GBHA (A), colorimetric pH indicators methyl red (B) bromocresol purple (C) and ratiometric pH-probe BCPCF (D). Inserts in (B) and (C) show a series of standard buffers with the relevant pH indicator. The apical ends with the secretory zone (secr) are to the left, the yellow pigmented incisal portions to the right (only visible in (A), (B) and (C)). In the maturation zone different stained bands are visible in all stainings corresponding to regions of ruffle-ended (RA) and smooth-ended (SA) ameloblasts. Based on high magnification images (left insert in (E)) of individual bands of BCPCF labeled teeth and a standard curve (right insert in (E)), pH values were measured in the 5 areas A-E indicated in the right insert in (E) The histogram in (E) shows the distribution of pH-values across the SA-bands classified with SA1 closest to the border between the secretory zone and the maturation zone of the enamel organ.

Na+, 3.6 mM K+, 1.8 mM Ca2+, 0.8 mM Mg2+, 138.6 mM Cl−, 5.5 mM glucose, 10 mM HEPES, 2 mM PO3− 4 ) adjusted to five pH values ranging from 5 to 7.5 (Fig. 1E, calibration curve). A set of control experiments were performed in which teeth were mounted in three different positions to certify that positioning of the tooth did not influence the measurement (not shown). Computational estimation of apparent pKa of HPO24 − at the surface of hydroxyapatite To explore the difference in pKa between hydrogenphosphate in aqueous solution and at the apatite surface numerical experiments using molecular dynamics simulations were performed based on earlier studies related to apatite–water interfaces [19,20]. Two systems were prepared as slight modifications of the (001) apatite–water interface reported in Hochrein and Zahn [20]. Model a) (Eq. (1)) considers the solvation of a single HPO2− ion in the water phase, whereas model b) 4 (Eq. (2)) describes PO3− 4 solvation in water, while one of the phosphate ions of the apatite surface is protonated. Both models include an extra Ca2+ ion dispersed in the water phase to maintain charge neutrality. On the basis of the average energetic difference of both systems, and knowledge of the pKa value for hydrogenphosphate in aqueous solution taken from the literature, we can thus derive a ‘local’ pKasurf value for the apatite surface (Eq. (3)): 2−

pKaaq

3−

HPO4aq þ H2 O → PO4aq þ H3 O 2−

pKasurf

3−

þ

ð1Þ þ

HPO4surf þ H2 O → PO4surf þ H3 O

ð2Þ

  hE i−hEb i pKasurf −pKaaq ¼ ΔpK ¼ − log10 exp − a kB T

ð3Þ

where bEaN and bEbN refer to the average energy levels of systems a) and b), respectively, which were each sampled from 1000 ps molecular dynamics runs.

Results GBHA staining indicates position of SA bands The GBHA staining showed a distinct pattern of well-demarcated red bands running transversely or obliquely across the enamel surface (Fig. 1A). These bands reflect the position of SA at the time of decapitation. The bands in the early part of the maturation zone stained homogeneously, whereas the more incisally located bands stained weaker and exhibited 2 distinct narrow lines of high staining intensity separated by a wider zone showing a weak or no staining reaction. Further towards the incisal edge, the bands fainted away. Colorimetric pH-indicators show high pH in SA relative to RA bands Enamel surfaces stained with the pH indicator methyl red (Fig. 1B) exhibited several yellow bands running more or less perpendicular to the long axis of the tooth indicating the position of SA-bands as they correspond precisely to the location of the SA bands in the GBHA staining of the contralateral teeth. These bands were slightly wider than the GBHA red staining bands. When comparing with the color standards of pH the reaction in the SA bands indicated a neutral pH around 7.0. The wider bands separating these yellow bands stained orange/pink indicative of a pH around 6. The colors were more pronounced in the early part of the maturation zone and decreased gradually in intensity in the incisal direction. The stain of the pH indicator bromocresol purple (Fig. 1C) was also most distinct in the early half of the maturation zone and barely visible in the incisal half. The SA-bands demonstrated two somewhat diffuse narrow zones of purple color separated by a wider pale blue colored zone. When comparing with the pH standards, the purple zones expressed a pH value of about pH 6.5–7.0 while the pH value of the pale blue color was close to pH 7.5. The color of the RA bands between the SA bands had a color indicative of a pH value between 5.5 and 6.0.

H.H. Damkier et al. / Bone 60 (2014) 227–234

Surface enamel pH decreases with distance from the secretion/maturation border in SA-bands and late parts of RA-bands, —but not in the early parts of RA-bands To understand the apparent trend in pH value of a particular phase in the ameloblast modulation cycle towards the incisal end, we plotted the pH values in SA-phase, early RA-phase and late RA-phase as a function of the distance from the secretion/maturation border (Fig. 2). This analysis showed that the decrease in pH in area C (center of SAband, see Fig. 1E) as a function of increasing distance from the secretion/maturation border was statistically significant (Fig. 2A). In contrast, in the early RA-phase (area A in Fig. 1E) this correlation was not present (Fig. 2B). In the late part of the RA-phase (area E in Fig. 1E), the correlation with distance was statistically significant (Fig. 2B). The pH at the enamel surface thus appeared to be influenced significantly by the degree of enamel maturation, which increases as a function of distance from the start of the maturation zone. In SA- and late RA-phase, the effect of the enamel maturation was strong enough to impose a statistical significant correlation, whereas in early RAphase, the effect of the morphological state of the ameloblasts facing the enamel appear so dominating, that any role of the degree of maturation could not be measured. Following vinblastine-treatment, pH decreases with distance from the secretion/maturation border in both late and early parts of RA-bands In vinblastine-injected animals, statistically significant correlations between enamel surface pH and distance from the secretion/maturation border were found in the SA-phase (Fig. 3A), the early RA-phase (Fig. 3B) and the late RA-phase (Fig. 3C). Vinblastine had a very dramatic effect on the pattern of SA-bands as seen by GBHA-staining (Fig. 3D) in comparison to control animals (Fig. 2D). In vinblastine-treated animals, the SA bands were in general wider and showed much larger variability in width and shape. Following H+-ATPase inhibition, pH decreases with distance from the secretion/maturation border in both late and early parts of the RA-phase In incisors from rats treated with the H+-ATPase inhibitor FR167356, the correlations between pH and the position of the bands were also statistically significant in the SA-phase (Fig. 4A), the early RA-phase

9

SA-band p=0.0275

8

pH

In BCPCF preparations (Fig. 1D) all teeth exhibited banding patterns similar to what is found in the GBHA stained teeth. SA bands could be identified as two parallel zones of high fluorescent intensity separated by a low intensity fine band corresponding to the center of a SA band (Figs. 1D & E). The reasons for the large differences in fluorescence intensity are unknown. However, since the ratiometric pH measurement is independent of the concentration of the pH-sensitive probe it is possible to measure the pH in the sub-zones within each SA and RA band (Fig. 1E). When considering the pH values in the 5 subzones across an SA-band, a remarkable uniform pattern is observed (Fig. 1E). Within each band the pH was highest in the center of the SA-band, subzones C–D and lowest in area A representing newly formed RA-zones. In the first SA-band to be found at the beginning of enamel maturation (SA1), the pH varies between 6.65 ± 0.22 (in area A) and 7.34 ± 0.08 (in area D), but in the more incisally located SA-bands, pH becomes lower and ranges between 6.24 ± 0.06 (Area A) and 6.64 ± 0.18 (Area C) in the SA3 band. Although the SA-bands can be classified as SA1, SA2 etc. within a single tooth, the absolute position of the band with respect to the border between the secretion and maturation zone differs between teeth and especially between animals because of the cyclic modulation of ameloblasts between phases of RA- and SA-states.

A

7 6 5 0

B

2000

4000

μm

9

6000

8000

RA-early p=0.3117

8

pH

Ratiometric pH-measurements show pH-values between 6.24 and 7.34 in maturation zone enamel

7 6 5 0

C

2000

4000

μm

9

6000

8000

RA-late p=0.0421

8

pH

230

7 6 5 0

2000

4000

μm

6000

8000

D

Fig. 2. Surface enamel pH measurements as a function of distance from the secretion/ maturation border of incisors from control rats (A–C). Smooth ended ameloblast phases (Area C) (A). Early part of phases of ruffle-ended ameloblasts (B). Late part of phases of ruffle-ended ameloblasts (C). GBHA-stainings of incisors from control rats (D).

(Fig. 4B) and the late RA-phase (Fig. 4C). The SA-bands appeared slightly wider than in control animals but the GBHA staining pattern (Fig. 4D) was not dramatically disturbed as seen in vinblastine-treated rats. The drop in pH across an SA-band decreases towards the incisal edge From the initial analysis in Fig. 1 it was evident that the pH appeared to be lower in the early RA-phase (area A), than in the late RA-phase (area E) preceding a particular SA-phase. We therefore plotted the difference (ΔpH) between corresponding areas E and areas A as a function of the distance from the secretion/maturation border. In control rats, a very significant correlation was found between ΔpH and the distance from the secretion/maturation border, with the absolute value of ΔpH decreasing towards the incisal end (Fig. 5A). No significant correlations between ΔpH and the distance from the secretion/maturation border

H.H. Damkier et al. / Bone 60 (2014) 227–234

A

SA-band p=0.0068

9

A

231

9

SA-band p=0.0039

8

pH

pH

8 7

7 6 6 0

2000

4000

μm

B

6000

RA-early p=0.0201

9

5

8000

0

B

2000

4000

μm

9

6000

8000

RA-early p=0.0154

8

pH

pH

8 7

7 6

6 0

2000

4000

μm

C

6000

RA-late p=0.0063

9

5

8000

0

C

2000

4000

μm

9

6000

8000

RA-late p=0.0046

8

pH

pH

8 7

7 6

6 0

2000

4000

μm

6000

8000

D

5 0

2000

4000

μm

6000

8000

D

Fig. 3. Surface enamel pH measurements as a function of distance from the secretion/ maturation border of incisors from vinblastine-treated rats (A–C). Smooth-ended ameloblast phases (area C) (A). Early part of phases of ruffle-ended ameloblasts (B). Late part of phases of ruffle-ended ameloblasts (C). GBHA-stainings of incisors from vinblastinetreated rats (D).

were seen in neither the vinblastine nor the H+-ATPase-blocker-treated rats (Figs. 5B and C). It should be noted that the distance between corresponding areas E and A in vinblastine-treated rats was larger than in the other groups, due to the broadening of the SA-bands. The apparent pKa of surface HPO2− estimated by molecular dynamics 4 HPO2− 4 deprotonation at the apatite surface was found to be facilitated by 47 ± 5 kJ/mol in comparison to bulk aqueous solution. This

Fig. 4. Surface enamel pH measurements as a function of distance from the secretion/ maturation border of incisors from FR167376-treated rats (A–C). Smooth ended ameloblast zones (area C) (A). Early part of phases of ruffle-ended ameloblasts (B). Late part of phases of ruffle-ended ameloblasts (C). GBHA-stainings of incisors from FR167376treated rats (D).

results in a ΔpKa value of –8.23 ± 0.8. Literature values for pKa of HPO2− in aqueous solution vary between 12.3 [21] and 12.8 [22]. By 4 using the average of these values (12.55) and assuming an error of 0.3 in this estimate, the apparent pKa value of HPO2− at the crystal surface 4 is estimated as pKasurf = pKaaq + ΔpK = 4.3 ± 1.1.

232

H.H. Damkier et al. / Bone 60 (2014) 227–234

A

1.5

Control p=0.0006

Δ pH

1.0 0.5 0.0 -0.5

of maturation. The reasons for these differences are unknown, but – importantly – the ratiometric pH measurement is very robust to differences in probe concentration. This is clearly illustrated in the discrepancy between the patterns in intensity and pH illustrated in Fig. 1. Phosphate may be released from enamel in RA zones

-1.0 -1.5 0

B

2000

4000

μm

1.5

8000

Vinblastine p=0.2830

1.0

Δ pH

6000

0.5 0.0 -0.5 -1.0 -1.5 0

C

2000

4000

μm

8000

FR167356 p=0.7225

1.5 1.0

Δ pH

6000

0.5 0.0 -0.5 -1.0 -1.5 0

2000

4000

μm

6000

8000

Fig. 5. The difference in pH (ΔpH) between corresponding late RA-phases and early RAphases surrounding one SA-band as a function of the distance from the secretion/maturation border. ΔpH in control rats (A). ΔpH in vinblastine-treated rats (B). ΔpH in FR167376treated rats (C).

Discussion Enamel surface pH is determined by the state of the adjacent ameloblasts and the degree of enamel maturation This study represents the first measurements of the pH variations in the surface of maturing dental enamel using ratiometric fluorescent pH indicators. This technique has the advantage, compared to traditional colorimetric methods, that the pH value measured is independent of the concentration of the dye. In addition, the spatial resolution of our measurements is higher than previously reported data. In accordance with previous measurements we find a low pH under RA in comparison to the pH prevailing under the SA. However, the spatial resolution of our measurements also allows us to conclude that the pH under SA-bands and the late part of RA-zones decrease with increasing degree of enamel maturation. It is thus a major conclusion of this study, that the pH of surface enamel at a particular position in the maturation zone is determined by two factors: 1) the current morphological state of the ameloblasts facing the enamel and 2) the distance from the secretion/ maturation border. The latter factor reflects the degree of enamel maturation (degree of mineralization) and represents as such the accumulated result of the number of RA/SA cycles, to which the enamel has been subjected at the given position. The incisally located SA-bands are less homogeneously stained in the GBHA preparations and the colometric pH-data also shows a fading staining of SA-bands with increased degree of maturation. Similarly the BCPCF staining varied in intensity within each band and with the degree

The signals recorded from GBHA, colorimetric pH-indicators and BCPCF are all sensitive to the chemical environment on the enamel surface. Accordingly, the differences in color/fluorescence along the teeth represent spatial differences in the chemical environment. GBHAstaining was originally suggested to reflect the amount of free Ca2 + ions [5], but careful studies have shown that the staining intensity is also very sensitive to the concentration of phosphate (HPO24 − and H2PO− 4 ) [23]. In fact it appears that in the presence of equimolar concentrations of phosphate and Ca2+, the “Ca2+”-signal recorded from GBHA staining vanishes [23]. Based on 45Ca tracer studies, Ca2+ concentrations have been reported to differ between the SA-bands and zones of RA, with the highest Ca2 + concentration seen in SA-bands in some studies [5] and in RAzones in others [24]. The concomitant appearance of marked GBHA staining in SA bands has been interpreted to support this. However, the apparent difference in 45Ca-accumulation between SA and RA may also be interpreted to reflect different degrees of isotopic exchange [25]. In the light of the marked differences in ameloblast morphology between RA and SA resulting in different permeabilities of the epithelium and the – at least transient – disappearance of tight junctions at the transition between the RA and SA phase [26], it may be likely that the ability for Ca2+ isotopes to be exchanged in surface enamel differs between RA and SA regions. If the results of 45Ca-tracer experiments are interpreted to largely reflect differences in isotope exchange, the GBHA staining pattern may instead reflect differences between RA and SA regions in the ratio between soluble Ca2+ and phosphate. The discrepancy pointed out by Moran et al. [25], that the low pH under RA is expected to result in more free Ca2 + but does not stain with GBHA, could be resolved if the concentration of phosphate is increased more than the concentration of Ca2 + under RA as a consequence of the low pH. Release of phosphate under RA would in fact be a prediction from the main hypothesis underlying this study: The low pH under RA is a result of active proton-secretion from the ameloblasts and may result in partial dissolution of “immature subdomains” along the surface of enamel crystals. The pKa of enamel crystal phosphate is a key parameter in the maturation process As pointed out above, the results of this study shows pH of surface enamel to be determined by ameloblast activity (SA/RA phase) and the degree of enamel maturation. The effect of the ameloblasts on enamel surface pH can be considered to be addition of acid in the RA phase and addition of base in the SA phase [4,15], but the mechanism behind the effect of the degree of enamel maturation on enamel surface pH is unknown. It is tempting to suggest that phosphate dynamics might explain the phenomenon. In Fig. 6A, the equilibria of the phosphate species under consideration in the solution and in the crystal are depicted. The pKa values for the equilibria in solution indicate, − that at the pH ranges observed, only HPO2− 4 and H2PO4 will be available in relevant concentrations. In the hydroxyapatite crystal, PO3− 4 is known to be the dominant species and dissolution of hydroxyapatite by acid is considered to be initiated by protonation of PO3− to HPO2− [27]. Ac4 4 cordingly, it is fair to simplify the equilibria in the crystal to only involve 2− PO3− 4 and HPO4 . Values for the pKa of this equilibrium are reported between 12.3 [21] and 12.8 [22]. As pointed out earlier [19], this essentially precludes the direct transfer of PO3− from solution to hydroxyapatite 4 crystals at neutral pH, since only negligible amounts of PO3− 4 compared

H.H. Damkier et al. / Bone 60 (2014) 227–234

A

Solution pKa = 7.21

pKa 12.5

PO O43- +

H+

PO43- + H+ Ca2+

pKa

pKa = 2.12

H+

H2PO4- + H+

H3PO O4

HPO42- + H+

H2PO O4- + H+

H3PO O4

HPO42- +

Crystal

B CO 2

HCO 3-

C

HCO 3- + H+

H2PO4-

H2PO4Ca2+ HPO42- + H+ HPO42Ca2+ PO43- + H+ Ca2+ Ca2+

present in real enamel. Overall, the apparent pKa of the crystal surface HPO2− will tend to fall during maturation due to increasing perfection 4 of the crystals including the removal of carbonate from the crystals [1]. The degree of enamel maturation is thus reflected in this apparent pKa value and we therefore propose this parameter to mediate the correlation between measured pH and the distance from the secretion/ maturation border. Enamel crystal structure determines the consequences of titration cycles by ameloblasts

4.3

HCO 3-

233

Ca2+

HCO 3+ CO 2 HPO42- + H+ HPO42-

Ca2+ PO43- + H+ Ca2+ Ca2+

Fig. 6. Schematic figure illustrating the dynamics of phosphate equilibria in solution and enamel crystal (A) and the effect of ameloblasts on phosphate dynamics in the RA phase (B) and SA phase (C).

to HPO2− would be present in solution. Further, hydroxyapatite would 4 be unstable at neutral pH if the apparent pKa of HPO2− when incorpo4 rated in the hydroxyapatite crystal surface is not substantially lower than the reported solution pKa-values. The ions in the crystal are bound together by electrostatic interactions partially canceling out the charges of each ion. For the PO34 − ions, the electrostatic shielding by Ca2 + in the crystal lattice results in a reduced probability for H+ to bind the PO34 −, leading to an increase in the value of H+-activity required to protonate 50% of the available PO3− 4 ; in other words, a lower apparent pKa of HPO2− 4 . In accordance with early in vitro studies [28], mature enamel is not dissolved in the oral cavity above the critical pH of 5.5 [29]. The apparent pKa of HPO2− at the crystal surface, where protonation will occur, can 4 therefore be predicted to be lower than 5.5, but is difficult to measure experimentally [30]. However, by molecular simulations we assessed phosphate (de)protonation at the apatite surface resulting in an estimated apparent pKa value of HPO2− 4 at the surface of a perfect hydroxyapatite crystal of 4.3 ± 1.1. Since our simulations refer to carbonatefree, ideal hydroxyapatite, this value represents a lower limit for the apparent pKa values of HPO2− in real enamel. Earlier studies of the re4 lationship between pH and dissolution rates of hydroxyapatite showed, that the rate-limiting step in the dissolution process involves the disengagement of ions at dissolution sites at the surface and that the dissolution rate was accelerated at pH-values below 5 [31]. Moreover, solutions of phosphoric acid were less effective in dissolving hydroxyapatite than solutions of organic acids [31], implying a crucial role for phosphate dynamics at the hydroxyapatite surface for the stability of hydroxyapatite in acidic environments. The local degree of crystal perfection, which determines the degree of electrostatic shielding, will determine the specific pKa value for any given HPO24 − ion at the enamel surface. Therefore, a distribution of pKa values with a mean somewhat above 4.3 will be expected to be

The observed pH at a given point along the tooth is thus the result of the buffering power and pKa of the enamel surface HPO2− 4 and the titration state of this buffer by the overlying ameloblasts. Figs. 6B and C summarize these titration reactions in RA and SA phase respectively. The net result is partial enamel crystal dissolution with release of phosphate and Ca2+ in RA (Fig. 6B) and reposition of ions onto the crystal surfaces in SA (Fig. 6C). It should be noted, that in the model depicted in Fig. 6, exchange or transport of phosphate across the ameloblasts is not included, although this epithelium is known from 32P-tracer studies to have a permeability to phosphate which appear to differ between SA and RA phases [32]. Molecular dynamics simulations have shown the dissolution of hydroxyapatite by acid to appear by initial protonation of the PO3− and 4 subsequent protonation of the OH− [27]. Since the solution pKa values for these ions are approximately 12.5 and 14 respectively, one would expect the protonation to take place in the opposite order. This illustrates the crucial ability of electrostatic shielding in the crystal lattice to influence the apparent pKa of the ionic species involved [27]. Another interesting effect based on a similar shift of the crystal to a more stable state is the effect of incorporation of fluoride in enamel. It is well known that replacement of hydroxyl ions in enamel by F− leads to a more caries-resistant enamel [33] due to improved electrostatic interactions in the crystal [27,34]. The dominant factor for pH differs between the early and late part of the RA phase In the early RA-phase, no correlation between pH and distance from the secretion/maturation border is seen in control rats. Accordingly, we conclude that ameloblast activity relative to the degree of enamel maturation is the major determinant of surface enamel pH in the early RAphase. Since the RA-state is characterized by well-defined tight junctions, which first disappear at transition to the SA-state, the enamel is effectively sealed off from the extracellular space during an RA-phase. Acid added in the early part of an RA-phase may thus interact with the buffer-groups of the enamel during the middle and late parts of the phase. An unsuspected finding in this study is, that at a given distance from the secretion/maturation border, pH appears to be lower in the initial than the late part of the RA-phase (as seen in Fig. 5A). We think that this may indicate protonation and release of phosphate to be slow processes, which result in a gradual increasing buffering capacity along an RA-phase. If this increased buffer capacity gradually catches up with the ameloblasts H+-secretory capacity, an increase in pH will be seen with the endpoint influenced by the pKa of the buffer groups – phosphates – in the enamel at the given degree of maturation. Pharmacological interference with ameloblasts affects enamel surface pH patterns In vinblastine-treated animals, ameloblasts cannot return to the RAstate from the SA-state [35]. The RA present after 4 h of vinblastinetreatment are therefore most likely formed at least 4 h ago, and thus represent entirely the “late” parts of the RA-phase. This could explain why a correlation between pH and distance to the secretion/maturation border is seen in the “early” RA-phase of vinblastine-treated animals

234

H.H. Damkier et al. / Bone 60 (2014) 227–234

(Fig. 3B), but not in control animals. Since the SA-bands are much broader in vinblastine-treated animals (Fig. 3D), corresponding areas “A” and “E” (as defined in Fig. 1E) at a particular SA-band are further apart in vinblastine-treated animals than in controls. The enamel in the corresponding areas “A” and “E” on each side of an SA-band thus differ more with respect to the degree of enamel maturation in vinblastine-treated animals than in controls. This may contribute to the absence of the correlation between ΔpH across an SA-band and distance from the secretory/maturation border in vinblastine-treated animals (Fig. 5B). Since a correlation between pH in the early RA-phase and distance to the secretion/maturation border is present in rats treated with H+ATPase-blocker, but absent in control rats, H+-ATPase inhibition likely has an effect on the lowering of enamel surface pH in the early RAphase. This may be due to an inhibiting effect on the ameloblasts ability to secrete protons, or an indirect effect mediated by putting a halt to the ameloblast modulation cycle in analogy to the effect of vinblastine. Based on the available data, we cannot distinguish between these possibilities, although the latter possibility may explain the persistence of low pH under RA after H+-ATPase inhibition. If H+-ATPase activity was blocked and the enamel surface was not covered by the tight layer of RA, a rapid jump in pH due to the entry of base from the extracellular fluid as seen in SA-bands would be expected. This was not seen. However, if proton-secretion and conversion between SA and RA were both inhibited as a consequence of FR167356-injection, the pattern of alternating alkaline and acidic zones would persist, except that the surface enamel pH would be dominated by the degree of enamel maturation in all parts of the RA-zones. This scenario is in accordance with the results of this study. Based on the absence in ameloblasts of expression of TCIRG (ATP6V0a3) [12] [unpublished results from our lab], which is essential for osteoclast proton-secretion, proton-secretion by ameloblasts has recently been questioned [12]. However, TCIRG (a3) is only one of the 4 homologous isoforms of a-subunits known in the mammalian VATPase protein complex [36]. Evidence has previously been presented for the presence of the a1-isoform (ATP6V0A1) in ameloblasts [15], and also the presence of immunoreactivity for ATP6V1B2 [12], supports earlier findings of the H+-ATPase in ameloblast plasma membrane [37]. Conclusion: Ruffled-ended ameloblasts lower enamel surface pH The present measurements show the pH values in surface enamel of incisors from control, vinblastine- and H+-ATPase-blocker-treated rats in accordance with the integrative functional model of the enamel organ implicating H+-ATPase mediated acid-secretion by ruffle-ended ameloblasts. Acknowledgments Astellas Research Technologies Co (Ibaraki, Japan) is thanked for providing FR167356 for this study. We thank Inger Merete Paulsen, Pia Kjærgaard and Peter Aakær for the technical assistance. We also thank one of the reviewers for the very helpful and detailed comments. The study was partly supported by a grant from the Bagger-Sørensen Foundation. References [1] Takagi T, Ogasawara T, Tagami J, Akao M, Kuboki Y, Nagai N, et al. pH and carbonate levels in developing enamel. 1998;38:181–7. [2] Smith CE. Cellular and chemical events during enamel maturation. Crit Rev Oral Biol Med 1998;9:128–61. [3] Josephsen K, Fejerskov O. Ameloblast modulation in the maturation zone of the rat incisor enamel organ. A light and electron microscopic study. J Anat 1977;124:45–70. [4] Lyman GE, Waddell WJ. pH gradients in the developing teeth of young mice from autoradiography of [14C]DMO. Am J Physiol 1977;232:F364–7.

[5] Takano Y, Crenshaw MA, Bawden JW, Hammarström L, Lindskog S. The visualization of the patterns of ameloblast modulation by the glyoxal bis(2-hydroxyanil) staining method. J Dent Res 1982:1580–7 [Spec No]. [6] Sasaki S, Takagi T, Suzuki M. Cyclical changes in pH in bovine developing enamel as sequential bands. Arch Oral Biol 1991;36:227–31. [7] Smith CE, McKee MD, Nanci A. Cyclic induction and rapid movement of sequential waves of new smooth-ended ameloblast modulation bands in rat incisors as visualized by polychrome fluorescent labeling and GBHA-staining of maturing enamel. Adv Dent Res 1987;1:162–75. [8] Simmer JP, Fincham AG. Molecular mechanisms of dental enamel formation. Crit Rev Oral Biol Med 1995;6:84–108. [9] Lacruz RS, Nanci A, Kurtz I, Wright JT, Paine ML. Regulation of pH during amelogenesis. Calcif Tissue Int 2009. http://dx.doi.org/10.1007/s00223-009-9326-7. [10] Lacruz RS, Hilvo M, Kurtz I, Paine ML. A survey of carbonic anhydrase mRNA expression in enamel cells. Biochem Biophys Res Commun 2010;393:883–7. http://dx.doi.org/ 10.1016/j.bbrc.2010.02.116. [11] Bronckers A, Guo J, Lyaruu D, Denbesten P, Zandieh-Doulabi B. Immunolocalization and western blotting of the anion exchanger pendrin in ameloblasts. Eur J Oral Sci 2012;120:369–72. http://dx.doi.org/10.1111/j.1600-0722.2012.00978.x. [12] Bronckers ALJJ, Lyaruu DM, Bervoets TJ, Medina JF, Denbesten P, Richter J, et al. Murine ameloblasts are immunonegative for Tcirg1, the v-H-ATPase subunit essential for the osteoclast plasma proton pump. Bone 2012;50:901–8. http://dx.doi.org/ 10.1016/j.bone.2011.12.019. [13] Bronckers ALJJ, Lyaruu DM, Jansen IDC, Medina JF, Kellokumpu S, Hoeben KA, et al. Localization and function of the anion exchanger Ae2 in developing teeth and orofacial bone in rodents. J Exp Zool B Mol Dev Evol 2009;312B:375–87. http://dx.doi.org/10.1002/jez.b.21267. [14] Bronckers A, Kalogeraki L, Jorna HJN, Wilke M, Bervoets TJ, Lyaruu DM, et al. The cystic fibrosis transmembrane conductance regulator (CFTR) is expressed in maturation stage ameloblasts, odontoblasts and bone cells. Bone 2010;46:1188–96. http://dx.doi.org/10.1016/j.bone.2009.12.002. [15] Josephsen K, Takano Y, Frische S, Praetorius J, Nielsen S, Aoba T, et al. Ion transporters in secretory and cyclically modulating ameloblasts: a new hypothesis for cellular control of preeruptive enamel maturation. Am J Physiol Cell Physiol 2010;299:C1299–307. http://dx.doi.org/10.1152/ajpcell.00218.2010. [16] Richards A, Fejerskov O, Baelum V. Enamel fluoride in relation to severity of human dental fluorosis. Adv Dent Res 1989;3:147–53. [17] Josephsen K. A technique for isolating the enamel organ of the rat incisor for histologic studies. Scand J Dent Res 1974;82:229–38. [18] Niikura K, Takano M, Sawada M. A novel inhibitor of vacuolar ATPase, FR167356, which can discriminate between osteoclast vacuolar ATPase and lysosomal vacuolar ATPase. Br J Pharmacol 2004;142:558–66. http://dx.doi.org/10.1038/sj.bjp.0705812. [19] Zahn D. Mechanisms of calcium and phosphate ion association in aqueous solution. Z Anorg Allg Chem 2004;630:1507–11. http://dx.doi.org/10.1002/zaac.200400151. [20] Zahn D, Hochrein O. Computational study of interfaces between hydroxyapatite and water. Phys Chem Chem Phys 2003;5:4004–7. http://dx.doi.org/10.1039/b306358e. [21] Lide DR. CRC handbook of chemistry and physics. 78th ed. Boca Raton, New York: CRC Press; 1997. [22] Hollemann, Wiberg. Lehrbuch der Anorganischen Chemie. 101st ed. Berlin: deGruyter; 1995. [23] Wolters GH, Pasma A, Konijnendijk W, Bouman PR. Evaluation of the glyoxal-bis-(2hydroxyanil)-method for staining of calcium in model gelatin films and pancreatic islets. Histochemistry 1979;62:137–51. [24] McKee MD, Warshawsky H. Modification of the enamel maturation pattern by vinblastine as revealed by glyoxal bis(2-hydroxyanil) staining and 45calcium radioautography. Histochemistry 1986;86:141–5. [25] Moran RA, Deaton TG, Bawden JW. Problems associated with estimation of net calcium uptake during enamel formation using 45Ca. J Dent Res 1995;74:698–701. [26] Kallenbach E. Access of horseradish peroxidase (HRP) to the extracellular spaces of the maturation zone of the rat incisor enamel organ. Tissue Cell 1980;12:165–74. [27] Hochrein O, Zahn D. On the molecular mechanisms of the acid-induced dissociation of hydroxy-apatite in water. J Mol Model 2011;17:1525–8. http://dx.doi.org/ 10.1007/s00894-010-0855-9. [28] Ericsson Y. Enamel-apatite solubility. Acta Odontol Scand 1949;8:1–139. [29] Fejerskov O, Kidd E, editors. Dental caries: the disease and its clinical management. Copenhagen: Blackwell/Munkgaard; 2008. [30] Christoffersen J, Christoffersen MR. Kinetics of dissolution of calcium hydroxyapatite: V. The acidity constant for the hydrogen phosphate surface complex. J Cryst Growth 1982;57:21–6. [31] Margolis HC, Moreno EC. Kinetics of hydroxyapatite dissolution in acetic, lactic, and phosphoric acid solutions. Calcif Tissue Int 1992;50:137–43. [32] Crenshaw MA, Takano Y. Mechanisms by which the enamel organ controls calcium entry into developing enamel. J Dent Res 1982:1574–9 [Spec No]. [33] Armstrong WD, Brekhus PJ. Chemical composition of enamel and dentin. II. Fluorine content; 1938. [34] Zahn D, Hochrein O. The role of substitutional defects in order/disorder phenomena of OH− ions in hydroxyapatite. Z Anorg Allg Chem 2006;632:79–83. http://dx.doi.org/ 10.1002/zaac.200500350. [35] Akita H, Kagayama M, Sato R. Light and electron microscopy of the effects of colchicine and vinblastine on maturing rat ameloblasts in vivo. Arch Oral Biol 1983;28:263–71. [36] Miranda KC, Karet FE, Brown D. An extended nomenclature for mammalian V-ATPase subunit genes and splice variants. PLoS One 2010;5:e9531. http://dx.doi.org/ 10.1371/journal.pone.0009531.t001. [37] Lin HM, Nakamura H, Noda T, Ozawa H. Localization of H(+)-ATPase and carbonic anhydrase II in ameloblasts at maturation. Calcif Tissue Int 1994;55:38–45.