Microdistribution of lead in human teeth using microbeam synchrotron radiation X-ray fluorescence (μ-SRXRF)

Abstract

Lead (Pb) exposure is known to be associated with adverse effects on human health, especially during the prenatal period and early childhood. The Pb content in teeth has been suggested as a useful biomarker for the evaluation of cumulative Pb exposure. This study was designed to employ the microbeam synchrotron radiation X-ray fluorescence technique to determine the microdistribution of Pb within the tooth to evaluate the reliability of the technique and the effectiveness of tooth Pb as a biomarker of Pb exposure. The results showed that in the incisor sample, Pb primarily deposited in secondary dentine region close to the pulp and secondarily at enamel exterior. In addition, Pb colocalised with Zn, indicating a positive correlation between Pb and Zn. By contrast,in the two molar samples, Pb accumulated principally in the pulp, and secondarily in the enamel. At the same time, Pb in these two molar samples colocalised with Ca instead of Zn as was observed in the incisor sample. Several batches of line scans further confirmed the conclusions. The feasibility of using microbeam synchrotron radiation X-ray fluorescence to determine the microdistribution of Pb in teeth and of using the tooth Pb, especially in dentine, as a biomarker was discussed.

Introduction

Lead (Pb) exposure has been known to be hazardous and poses irreversible health risks even at low levels among general population. The exposure can result from environmental conditions, for example the wide industrial usage in leaded gasoline, lead-based paint, ceramics, etc.^[1] It can also result from dietary habits, for example the contamination of tap water in the old lead-containing pipes and of comestibles from a lead-polluted area.^[2] The adverse effects can be on the nervous (e.g. encephalopathy), haematopoietic (e.g. anaemia), renal (e.g. nephropathy), cardiovascular (e.g. hypertension), and reproductive system (e.g. infertility).¹ These deleterious manifestations can be more severe among paediatric population, especially for the nervous system, than adults, because the developing nervous system in foetuses and young children lacks a well-developed blood-brain barrier system and is more vulnerable to Pb exposure.^[3,4] Pertinent environmental and epidemiological studies also support this conclusion.^[5-7] One of the most widely applied biomarker for Pb exposure assessment is the blood Pb. However, once Pb is absorbed into the blood compartment, its amount decreases with a mean biological half-life of about 40 days,^[5] so it reflects only the recent exposure.

To accurately evaluate historical exposure in early periods of development, teeth have been used to extract temporal information of long-term exposure because the residence time of Pb in calcified tissues ranges from 10 to 30 years. Teeth composed of compact calcified tissues have slower turnover rates than calcified tissues of trabecular bones have, so they are expected to have slower release rates of Pb and longer residence time,^[5] making it an excellent indicator to monitor Pb exposure. Whole tooth Pb has been considered as a biomarker for Pb exposure.^[8] Pb may accumulate unevenly within the tooth because of different affinities to different dental parts. Therefore, a thorough understanding of Pb accumulation in tooth is necessary to indicate undue exposure and learn the environmental conditions and dietary habits. In fact, not only Pb but also other elements have heterogeneous distributions in different anatomical parts in tooth, which has been acknowledged by the scientific community.^[9-12]

It has been hypothesised that Pb levels at each point along some lines in tooth represent Pb exposure at the specific time when that part of tooth is being mineralised.^[13] A typical human incisor transect is shown in Fig. 1 and the internal anatomic structure is given in Fig. 2. When the formation of the incisor starts, the enamel (composed of calcium hydroxyapatite crystals) and the primary dentine (PD) (consisting of shaped tubules containing about 75% calcium hydroxyapatite) deposit at the enamel–dentine junction (EDJ). Because the enamel is fixed early in growth and only the surface experiences remineralization (a deposition process of Ca and P ions from saliva to the enamel surface),^[14] it can record the information of in utero Pb exposure. However, such a buildup may come from either internal Pb exposure during the enamel mineralization or external surface contamination (e.g. ion-exchange with saliva) in childhood.^[15] Although there have been observations that the exterior of enamel contains higher amount of Pb than other parts do,^[16-18] the concentration in the exterior of enamel is usually much (an order of magnitude) lower in most cases than that in dentine, especially in the dentine area close to pulp.^[15] Unlike enamel and PD that store information from the foetal period and childhood, circumpulpal dentine (CD), later retreated to secondary dentine (SD), grows with a gradually decreased speed after the root is formed and is able to provide information of long-term Pb exposure throughout adulthood immediately after the root is formed. One of the dissimilarities is that although the neonatal line normally exists in all human primary teeth, it is only occasionally present in permanent first molars.^[19] As a result, most studies focused on incisors of deciduous teeth. Both types of tooth of adult teeth were investigated in our study.

Figure 1.

Tooth sample of an incisor from a 56-year-old woman with three scanned lines marked as line 1, 2, and 3. The arrow indicates the scanned direction.

Figure 2.

Schematic diagram of the incisor anatomic structure edited based on Arora et al.’s work.^[13] The bold in-italic words denote the detailed parts of incisor, the un-bold in-italic words indicate the two lines related to temporal development of incisor, and the un-bold italic words specify the direction of the development of each part of incisor. The arrow on neonatal line represents the growth direction. PD, primary dentine; CD/SD, circumpulpal dentine/secondary dentine; EDJ, enamel–dentine junction; Pre-,prenatal; Post-,postnatal; E, enamel; P, pulp.

A variety of techniques have been applied to analyse the whole tooth Pb concentration; however, the measurements of the spatial microdistribution in each part of the tooth structure are limited to only a few techniques. In vitro analyses by inductively coupled plasma-mass spectrometry (ICP-MS) coupled with laser ablation (LA-ICP-MS)^[15] and with electrothermal vaporization (ETV-ICP-MS)^[20] were used to quantify the Pb microdistribution within human tooth. Non-invasive in vitro techniques, such as microbeam synchrotron radiation X-ray fluorescence (μ-SRXRF), have also been used for this purpose.^[9,18] Although LA-ICP-MS has a relatively low detection limit, its slightly destructive nature affects histopathological examinations of the samples after the elemental analysis. The heterogeneity of the sample volume ablated could introduce large uncertainty in quantification.^[21] In addition, the certified analytical standards for comprehensive quantification and method validation are deficient.^[11] To contrast, μ-SRXRF is non-destructive, permitting the sample to be prepared simply and preserved longer, which allows for repeated analysis and other analysis^[22] such as stained analysis for histopathological examinations after measurements. Moreover, the spatial resolution for μ-SRXRF has been enhanced to the micro-scale (for example with a beam dimension of less than 1 μm in the beamline of Advanced Photon Source at Argonne National Laboratory) owing to the superior X-ray beam brightness coupled with high-class focusing optics; the detection limit is lowered to the ppm level (e.g. 0.5 ppm for Ca) because of the increase signal-to-noise ratio.^[9,18,23] With its nondestructive feature, superior special resolution, and relatively short acquisition time, μ-SRXRF is an important method to be employed in the study of microdistribution of both metal elements and non-metal elements.

The principle aims of this work were (1) to verify the feasibility of using μ-SRXRF to quantify Pb content within human tooth, (2) to obtain the spatialmapping of themicrodistribution of Pb and other elements, and (3) to use the microdistribution to provide information on the temporal Pb exposure. The usefulness of tooth Pb and its distribution as a valid biomarker for Pb exposure in early childhood (postnatal) and adulthood (cumulative) were also discussed. Information regarding foetal period (prenatal) could not be examined for this report because the samples used were permanent tertiary teeth.

Materials and methods

Sample preparation

Tooth samples consist of one mandibular central/lateral incisor (Fig. 1) from a 56-year-old woman and two third molars (Fig. 3) from a 29-year-old man. The molar has a similar structure to the incisor (Fig. 4). The ethics approval was issued by institutional review board (IRB) at T.H. Chan School of Public Health, Harvard University. No large growth of cementum covering the root was found, so there was no hypercementosis. The teeth were cross-sectioned along the vertical median plane with an Isomet low speed dental saw (Buehler, Lake Bluff, IL, USA) with a diamond blade. A longitudinal 1 mm-thick slice was cut through the centre of the tooth crown, and the slice was polished using sand paper with ultrapure water as a lubricant in order to obtain a smooth and plane surface. The final thickness was near 0.6mm. Only one slice of each tooth was analysed.

Figure 3.

Tooth samples of a molar from a 29-year-old male (panel (a): molar sample 1, panel (b): molar sample 2) with three scanned lines marked as line 1, 2, and 3.

Figure 4.

Schematic diagram of the molar anatomic structure with the dark blue dash line being enamel–dentine junction.

Experimental setup for line positioning and scanning

The elemental analysis was performed by the μ-SRXRF 8-BM-B beamline at the Division of Advanced Photon Source, Argonne National Laboratory. The schematic setup is shown in Fig. 5. The incident beam energy of 16.3 keV was selected for Pb (Lα: 10.6 keV, Lβ: 13.2 keV) by the crystal monochromator, and the pixel size was determined by the pinhole. The processed tooth samples were first decontaminated by 50% alcohol and de-ionised water and then mounted properly by tape placed at the sample edge to the holding brackets that served as the image plane. An appropriate mounting should be flat in order to assure an even scan over the surface without scanning the tape. The image plane had an accuracy of 1 μm with a 2 axis (x,y) remote-controlled stage.

Figure 5.

Actual (up) and schematic (down) experimental setup (http://www.stefan.vogt.net/research_interests.html).

The tooth samples were first scanned on a scale of 100 μm to obtain the mapping images. The data for each pixel with a colour scale for the mapping images were measured for 2 s for all nine images (three images for each one of the three samples), with 1 s being the measurement of the pixel itself and 1 s moving from pixel to pixel. The incisor sample had three mapping images with matrices of 61 × 41 pixels, 61 × 31 pixels, and 71 × 29 pixels, respectively. The three mapping images of molar no.1 had matrices of 61 × 13 pixels, 75 ×32 pixels, and 5 ×21 pixels, respectively, and those of molar no. 2 had matrices of 87 × 31 pixels, 62 × 30 pixels, and 61 × 31 pixels, respectively. Then, the mapping images of the metal elements were obtained via appropriate fitting, and the three lines for each tooth sample were determined by the mapping image and the coordinates of the images. The corresponding line scanning on a scale of 50 μm was performed for the microdistribution evidenced on the tooth diagrams (Figs 1 and 3) inwardly along line 1 and line 3 and outwardly along line 2. The lines were determined after consulting the dentist, reviewing a few previous studies,^[10,14,22] and taking the anatomy of the teeth into account. The total scanned length of the incisor sample was 6, 6, and 7mm for line 1, line 2, and line 3, respectively;while that of molar no. 1 was 4.4, 7.4, and 7.5mm, and that of molar no. 2 was 5.6, 7.6, and 6.1mm for line 1, line 2, and line 3, respectively. The intensity of Pb signal at different position was acquired. The microdistribution information through this very plot of intensity was inferred against position.

Imaging processing and peak fitting

Each pixel denotes a signal intensity of a length of 100 μm for line positioning and 50 μm for line scanning. All the spectra of each pixel were analysed using MAPS, a custom-built software package that is written in IDL and able to produce elemental microdistribution based on the input full-spectrum-per-pixel XRF data.^[24] Elemental concentrations were obtained by fitting the spectrum of each scanned point at a 100 μm. The quality control of the XRF setup, including the alignment, optimization, and calibration of the system, was done using a thin film XRF reference sample RF4-200-S1749 with known concentrations of seven elements including Pb, Cu, Fe, and Ca. The lateral homogeneity of all elements deposited on these reference samples has been tested with μ-XRF mappings, and the deviation was smaller than 1% over the entire sample area. The concentration was given in the unit of μg/cm² to correct for the density differences of the samples. Each line scanning produced a single data file in comma separated value (.csv) format. The L_α peaks of Pb were fitted by MATLB with a Gaussian peak combined with an exponential background and the goodness of fitting was presented by chi-squared values as described by Specht et al.^[25] The uncertainty of the fitting can be obtained from MATLAB, and the best way is to add the error bars for the data points in each figure. However, more than 1000 figures were fitted in this study and adding error bars to every single point in every single figure would be ponderous and unrealistic. We postulated that the uncertainty was reasonably small based on our fitting results.

Results and discussions

Incisor

The Pb microdistribution of the transects across the sectioned incisor showed that a clearly demarcated zone of high Pb is evident in SD and interestingly has the same distribution with Zn but a different distribution with Ca (Fig. 6). The images in Fig. 6(a) also showed a fair amount of Pb accumulates at the enamel surface, but the level is much lower than that in the SD based on the brightness of these two regions. The level of Ca decreased from exterior to interior, and a sharp decrease was evidenced near the EDJ, which agreed well with the content of the hydroxyapatite in each part of tooth. Such a phenomenon is actually similar for all types of tooth and is independent of donor characteristics.^[26] However, the distribution of Pb remains controversial. Our result showed that Pb was conspicuous in SD and had the same distribution with Zn as stated previously, which agrees well with Bellis et al.’s works on both goat teeth^[21] and human teeth,^[15] with Ide-Ektessabi et al.’s work^[22] and with Anjos et al.’s work.^[26] In fact, Ca and Pb compete for the binding sites in the body via isovalent replacements in hydroxyapatite (Marques et al.^[12]) and dietary deficiencies of Ca and P can result in an increased Pb absorption.^[21] Our results contradict the results of Guerra et al.’s work on primary teeth of children by using SRXRF, which stated that Pb and Ca have a similar distribution and Pb content decreases from enamel to CD,^[18] and that later Pb and Cu similarly amassed in superficial enamel.^[10] Such a discrepancy may implicate that the displacement of Pb to Ca and Cu outstrips that to Zn in the early stage of tooth development, while it becomes laggard in adult tooth. As for which tooth section Pb accumulates in, although we observed the existence in enamel, specifically in superficial enamel, consistent with Guerra et al.’s result of primary incisors in 2014^[10] and permanent incisors,^[27] the concentration was much lower than that in SD. Pb primarily resides in SD region because this region is adjacent to the pulp tissue contained within the pulp cavity filled with vascularised blood that contains Pb upon exposure.^[21] The tubules in dentine make the dentine permeable and create the routes for Pb to diffuse.^[12] In addition, the odontoblastic cells and associated blood vessels in this region to make the rate of Pb exchange rate from deposits in SD higher than that in PD nearer to the EDJ, which has been corroborated.^[28,29] Another way for Pb to accumulate in SD is the incorporation of the hydroxyapatite of proximal SD. Of note, Pb deposition in SD is not affected by prenatal Pb exposure and SD continues to accrue Pb as long as the tooth is vital, so it is able to serve as a measure of integrated lifetime Pb exposure up to the time it is shed.^[13] This accumulation of Pb arises from Pb in circulation during adulthood when CD is remodeled to SD.^[15] As to the colocalization with Zn, the Zn content in CD remains practically constant across all animals^[21] and is slightly higher in the inner part of teeth.^[9] Such a similarity did not result from spectral overlap as it is known that the Zn Kα peak, and the Pb Kα peak were far away from each other. Although the reason of the colocalization remains unfathomed hitherto, this positive correlation between Pb and Zn may have biological relevance^[10] related to the metabolism of the pulp and to the odontoblast forming the dentine.^[22] More studies are requisite to delve into this phenomenon.

Figure 6.

Microdistributions and line positioning of three metals of concern [(a) line 1 positioning, (b) line 2 positioning, and (c) line 3 positioning] of the incisor from a 56-year-old woman. The brighter the pixel is, the higher content the metal is as can be seen from the relative concentration index below. The blue dotted line shows the enamel–dentine junction (EDJ). The green dashed line with arrow shows the scanning direction and scanning position, corresponding to the red dashed line in Figs 1 and 3.

The predetermined lines (line 1, 2, and 3 in Fig. 1) can be positioned from the images in Fig. 6 based on the coordinates of pixels. The chi-squared values rescaled by the degree of freedom for all the spectral fittings of concern are close to 1, with the average chi-squared and standard deviation of the chi-squared value for these fits being 1.02 ± 0.15. Each scanning of the three chosen lines (Fig. 7) showed several spikes. The anatomic regions parallel to the abscissa were obtained via the thickness of each region. The thickness of enamel varies from region to region but shows no sexual dimorphism and has a thickness of about 1mm^[30] that is assumed here, so roughly 20 points can be scanned in the enamel region using μ-SRXRF with a resolution of 50 μm. Likewise, regardless of the tooth type, the average thickness of dentine and pulp for 45–59 years old is 4.69±0.39 and 0.84±0.42mm, respectively, while that for 15–29 years old is 4.45 ± 0.54 and 1.13 ± 0.43mm, respectively.^[31] All the thickness of each part assumed in this study is summarised in Table 1.

Figure 7.

Spatial distributions of Pb along three scanned lines [(a) line 1, (b) line 2, and (c) line 3] across the longitudinally sectioned transect of the incisor from the 56-year-old woman. PD, primary dentine; SD, secondary dentine; E, enamel; P, pulp.

Table 1.

The presumed thickness of each part of the tooth samples based on Siddiqui et al.’s^[25] and Harris et al.’s work^[22]

Tooth sample	Thickness (mm)
	Enamel	Dentine	Pulp
Incisor (56 years old)	1.00	4.69	0.84
Molar (29 years old)		4.45	1.13

Line 1 in Fig. 7(a) was scanned inwardly, and two peaks were observed with the first being at the enamel surface and the second being in SD. Line 2 in Fig. 7(b) was scanned outwardly with one broad peak being across the region of SD and part of PD and one small sharp peak being at the enamel surface. Line 3 in Fig. 7(c) was scanned inwardly with only one peak in the SD region. The highest signal intensity corresponds to the maximum concentration shown on top of the images in Fig. 6. The maximum concentration of Pb of the three scanned regions was 5.04, 3.13, and 9.6 μg/cm², respectively, for line 1, line 2, and line 3. Again, the conclusions drawn from Fig. 6 corroborated that the majority of Pb deposited in SD with minor locates at the enamel surface. Because the Pb deposits more in the SD (adulthood) than in the enamel region (childhood), this person received more exposure in adulthood than in childhood.

Molar

The results based on the images of both molar samples were very similar. The Pb microdistribution mapping of the transects across the sectioned molar samples is shown in Figs 8-10. As can be seen in the figures, pulp and enamel are the main regions where Pb primarily resides in. This partially agrees with what was seen for the incisor sample. The microdistribution of Pb in the molar samples was relatively homogeneous throughout the enamel region compared with that in the incisor sample and more analogous to that of Ca than to that of Zn. This is quite contradictory to the conclusions we just came to previously, saying that Ca and Pb compete for the binding sites in the body, and dietary deficiencies of Ca and P can result in an increased Pb absorption.^[21] On the other hand, both Ca and Pb are bone-seeking elements. With the similarity between bone and teeth, it is not surprising to see the same pattern for Ca and Pb distribution in a tooth sample. The reasons for the differences between the conclusions drawn from the incisor and molar samples are not clear, and more studies are required in this area.

Figure 8.

Microdistributions and line positioning of three metals of concern of molar no. 1 [(a) line 1 positioning, (b) line 2 positioning, and (c) line 3 positioning] with the relative concentration index below.

Figure 10.

Microdistributions and line positioning of three metals of concern of molar no. 2 [(a) line 1 positioning, (b) line 2 positioning, and (c) line 3 positioning] with the relative concentration index below.

The line scanning results of both molar samples are shown in Figs 9 and 11. The distributions are also alike, showing a consistency across different molars of the same person. Pb tends to accumulate in the pulp region and SD–pulp junction region, similar to the conclusions drawn by the incisor sample but with the signal intensity being much lower. This is another reason that most studies focus on incisors instead of molars aside from lacking of the neonatal line for molars as stated previously. The maximum concentrations of Pb of the three scanned regions on molar no. 1 were 0.133, 0.702, and 1.23 μg/cm², respectively, for line 1, line 2, and line 3, while for molar no. 2, the maximums were 0.327, 1.5, and 0.839 μg/cm². The Pb concentrations of both molar samples were substantially lower than that of the incisor sample.

Figure 9.

Spatial distributions of Pb along three scanned lines [(a) line 1, (b) line 2, and (c) line 3] across the longitudinally sectioned transect of the molar no. 1 from the 29-year-old man. PD, primary dentine; SD, secondary dentine; E, enamel; P, pulp.

Figure 11.

Spatial distributions of Pb along three scanned lines [(a) line 1, (b) line 2, and (c) line 3] across the longitudinally sectioned transect of the molar no. 2 from the 29-year-old man. PD, primary dentine; SD, secondary dentine; E, enamel; P, pulp.

We need to treat the results with caution because the incisor and molar samples are from different individuals, and we are not aware of their backgrounds, including dietary habits and living environment. If the samples were from the same person, these tooth samples would have the same physiological environment and hence the differences of Pb micro-distribution would be more likely originate from different types of tooth.

Conclusions

Human teeth, especially incisor, can record the information of Pb exposure at different stages of development (prenatal, postnatal, and adulthood). Our study has analysed the adult teeth that reflected the exposure history in the adulthood. One of the modern analytic techniques able to ascertain the exposure and retrospectively reconstruct the timing of the exposure and the long-term exposure is μ-SRXRF. There have been quite a few applications of the technique in tooth^{[9,12,14,18,22,26,32]} and cell analyses, e.g. Cu and Zn,^[33] studies of the mapping of Pb distribution in different types of tooth in modern people, to the authors’ knowledge, has not been seen yet because the majority focused on incisors. Occasionally, a few studies have investigated molars, however, for archaeological purposes.^[14,32] Hence, the study presented can provide very useful information for the environmental epidemiology of Pb. The beam of μ-SRXRF can reach micrometer scale with enhanced resolution, which is very helpful in the visualization of metal microdistribution. Line scanning confirms the conclusions drawn from the microdistribution maps. For the incisor sample, a clearly delineated zone of high Pb was manifested in postnatally formed SD (with significant but less amount in enamel surface) and Pb and Zn shared similar distributions in the teeth, both of which were well explained. For the molar samples, the microdistribution of Pb is comparatively identical throughout the enamel region and more similar to that of Ca than to that of Zn. Both the accumulation region and the colocalization with other metals for molar samples are different from the conclusions drawn from the incisor sample, so the mechanisms of Pb deposition in different tooth types may be different. The individual to whom the incisor belonged was concluded to receive Pb exposure postnatally. With regard to the other individual to whom the molars belonged to, only the Pb exposure in adulthood was evaluated. The maximum concentrations of Pb in both incisor and molar samples were measured. For the incisor sample, the concentration was 5.04, 3.13, and 9.6 μg/cm², for regions used to define line 1, line 2, and line 3, respectively. By contrast, for molar no. 1, a much lower value of 0.133,0.702, and 1.23 μg/cm², respectively, was measured. Molar no. 2 also had a relatively low concentration of 0.327,1.5, and 0.839 μg/cm², respectively, for the three lines. The pitfall of this study was that no statistical comparison holds because of the limitations of the amount of tooth samples and to only one measure per location; the variability of the data might be augmented.